Changes of gut mycobiota in the third trimester of pregnant women with preeclampsia

Abstract

Objective

Studies recently acknowledged that the fungal community in the gut plays an important role in many inflammatory diseases of noninfectious origin. but the role of gut fungi in the pathogenesis of preeclampsia (PE) remains unknown.

Methods

We performed a case-case–control study to compare the gut mycobiota of PE, pregnancy with chronic hypertension (PCH), and the normal group (Normal pregnancy group without any underlying disease) by internal transcribed spacer sequencing. In addition, LC/MS was used to explore the relationship between the fecal metabolites and gut mycobiota.

Results

Compared with the PCH and the normal group, α diversity (represented the species abundance in a single sample) of mycobiota were lower in the PE group, but there was no statistically significant difference among these three groups. However, Linear discriminant analysis Effect Size (LEfSe) analysis found 3 differentially abundant fungal taxa in PE group when compared with the normal group. The gut metabolites of PE patients were significantly different from PCH and the normal group. Choline metabolism molecule glycerophosphocholine was the most discriminant metabolite between PE and the normal group. Correlation analysis found that Candida spp.was positively correlated with glycerophosphocholine which increased in PE.

Conclusion

We found that gut mycobiota changed in the third trimester of pregnant women with preeclampsia.

Keywords:

• Gut fungal
• preeclampsia
• glycerophosphocholine
• Candida spp.

Introduction

Gestational hypertension, preeclampsia (PE), and pregnancy with chronic hypertension (PCH) and other pregnancy-induced hypertension disease (PHD) are some of the most common diseases during pregnancy, threatening the health of mothers and neonates. PE, occurs most often after 20 weeks of gestation combined with proteinuria, affects 3–8% of pregnancies, and is associated with an increased risk of adverse pregnancy outcomes [1]. At present, the treatment of PE is mainly symptomatic treatment such as spasmolysis, sedation, and hypotension. Timely delivery is the only way to remove the threat of PE to pregnant women, which often leads to premature birth and may cause poor prognosis of neonates such as hypoxic-ischemic encephalopathy, respiratory distress syndrome and Retinal hemorrhage of neonate [2]. Therefore, it is of great clinical significance to predict the occurrence of PE in advance and find new targets for PE treatment to improve the quality of life of mothers and infants.

Researchers have recently identified that gut microbiota imbalance is associated with preeclampsia [3–5]. However, the fungal community, the mycobiota, playing an undeniably important role in association with the human host [6–9], has been seriously neglected in the development of PE. Over the past few years, research has demonstrated that mycobiota have substantial effects on the host immune responses [10]. Many studies have confirmed that mycobiota dysbiosis had local and systemic effects on immunity and inflammation [6,7,11]. Recently, another study further discovered a rich genetic diversity of Candida albicans in patient's gut with inflammatory bowel disease, where strains with high immune-cell-damaging capacity aggravated intestinal inflammation in vivo through IL-1β-dependent mechanisms [9]. Furthermore, researchers retrieved fecal metagenomic data sets from 7 previous publications and established an additional in-house cohort. They found that six species of fungi were enriched in the colorectal cancer (A rambellii, Cordyceps sp. RAO-2017, Erysiphe pulchra, Moniliophthora perniciosa, Sphaerulina musiva, and Phytophthora capsici) and one specie was depleted (A kawachii) [12]. These findings reveal that mycobiota dysbiosis have detrimental effects on host immunity and highlight new diagnostic and therapeutic targets for diseases.

In this study, we enrolled three groups of pregnant women, including PE, pregnancy with chronic hypertension (PCH), and the normal group. Internal transcribed spacer regions (ITS) Sequencing was used to characterize the gut mycobiota. In addition, untargeted metabolomics was also performed to provide a basic description of microbial metabolism. We described the characteristic of mycobiota associated with PE, detected potential correlations between microbes and metabolite, and also speculated on the pathogenesis of PE.

Methods

Ethical considerations

All patients signed informed consent before sampling, and the Ethics Committee of Women and Children’s Hospital of Chongqing Medical University reviewed and passed this protocol.

Patients and Sample Collection

This case-to-case study was performed at Chongqing Health Center for Women and Children in Southwest China. The research was divided into three groups, including late-onset PE, PCH, and the normal group. PE and PCH were diagnosed according to the current guidelines. PE: (1) systolic blood pressure (SBP)/diastolic blood pressure (DBP) ≥ 140/90 mmHg on two occasions for at least 4 h with previously normal blood pressure; (2) proteinuria ≥ 300 mg/24h urine collection; and (3) in the absence of proteinuria, new onset of any of the following: platelet count <100,000/µL; serum creatinine concentration > 1.1 mg/dL or a doubling in the absence of other renal disease; elevated blood concentrations of liver transaminases to twice normal concentration; pulmonary edema; and cerebral or visual symptoms. Additionally, preeclampsia with severe features was diagnosed with any of the following fifindings: SBP/DBP ≥ 160/110 mmHg on two occasions at least 4 h apart; platelet count < 100,000/µL; elevated blood concentrations of liver transaminases to twice normal concentration; severe persistent right upper quadrant or epigastric pain; serum creatinine concentration > 1.1 mg/dL or a doubling in the absence of other renal disease; pulmonary edema; and new-onset cerebral or visual disturbances. PCH: hypertensive pregnant women without albuminuria before 20 weeks of pregnancy. [13]. The exclusion criteria were as follows: 1) Other underlying diseases except PE and PCH, such as obesity, lipid metabolic disorders, inflammatory bowel disease, gestational diabetes and hypothyroidism. 2) administration of antibiotics, antifungal drugs or probiotic treatment during pregnancy. 3) Fungal foods such as mushrooms are prohibited for one month prior to specimen collection. The fecal specimens were collected after hospitalization but before childbirth (gestational weeks > 37 weeks) Fecal samples were collected with special feces collection containers, a 76 × 20 mm tube with a screw cap attached scoop, each of which was preloaded with 5 ml antiputrefactiva (MGIEasy Stool Sample Collection Kit, Shenzhen Huada Zhizao Technology) and stored at −80 °C for further processing. Six pregnant women who met the inclusion and exclusion criteria were included in each group.

DNA extraction and PCR amplification

Total microbial genomic DNA was extracted from fecal samples using the E.Z.N.A.^® soil DNA Kit (Omega Bio-tek, Norcross, GA, U.S.) according to manufacturer’s instructions. The detection of mycobiota relied on sequencing the fungal internal transcribed spacer (ITS) ITS1 and ITS2 regions. All samples were amplified in triplicate. The PCR product was extracted from 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer’s instructions and quantified using Quantus™ Fluorometer (Promega, USA). Purified amplicons were pooled in equimolar amounts and paired-end sequenced on an Illumina MiSeq PE300 platform/NovaSeq PE250 platform (Illumina, San Diego, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China).

Diversity analysis of intestinal flora

Bioinformatic analysis of the gut mycobiota was carried out using the Majorbio Cloud platform (https://cloud.majorbio.com). Based on the operational taxonomic units (OTUs) information, rarefaction curves and α diversity indices including observed OTUs, Chao1 richness, Shannon index, and Good’s coverage were calculated with Mothur v1.30.1 [14]. The Shannon and Simpson diversity indices were used to estimate the α diversity, which represented the species abundance in a single sample. βdiversity, the rate and extent of diversity change in species that occurs on a gradient from one biotope to another. The β diversity was used to evaluate differences in the species complexity in the samples based on the weighted UniFrac metric. Based on the UniFrac phylogenetic distance, the test of the significance of the clustering of samples in the study was carried out by one-way analysis of similarities (ANOSIM). The similarity among the mycobiota communities in different samples was determined by principal coordinate analysis (PCoA) based on Bray-Curtis dissimilarity using Vegan v2.5-3 package. The PERMANOVA test was used to assess the percentage of variation explained by the treatment along with its statistical significance using Vegan v2.5-3 package[14]. The linear discriminant analysis (LDA) effect size (LEfSe) [15] (http://huttenhower.sph.harvard.edu/LEfSe) was performed to identify the significantly abundant taxa (phylum to genera) of fungal among the different groups (LDA score > 2, p < .05).

Metabolite extraction and ultra performance liquid chromatography/time of flight mass spectrometry analysis

50 mg solid samples were accurately weighed, and the metabolites were extracted using a 400 µL methanol:water (4:1, v/v) solution [16–18]. The mixture was allowed to settle at −20 °C and treated by High throughput tissue crusher Wonbio-96c (Shanghai wanbo biotechnology co., LTD) at 50 Hz for 6 min, then followed by vortex for 30s and ultrasound at 40 kHz for 30 min at 5 °C. The samples were placed at −20 °C for 30 min to precipitate proteins. After centrifugation at 13000 g at 4 °C for 15 min, the supernatants were carefully transferred to sample vials for liquid chromatography-tandem mass spectrometry/mass spectrometry (LC-MS/MS) analysis. a pooled quality control sample (QC) was prepared by mixing equal volumes of all samples. The QC samples were disposed and tested in the same manner as the analytic samples. It helped to represent the whole sample set, which would be injected at regular intervals (every 8 samples) in order to monitor the stability of the analysis. Details of untargeted metabolomics referred to the manufacturer’s instructions. After ultra-performance liquid chromatography/time of flight mass spectrometry (UPLC-TOF/MS) analyses, the raw data were imported into the Progenesis QI 2.3 (Nonlinear Dynamics, Waters, USA) for peak detection and alignment. MS/MS fragments spectra and isotope ratio difference with searching in reliable biochemical databases as Human metabolome database (HMDB) (http://www.hmdb.ca/) and Metlin database (https://metlin.scripps.edu/).

Metabolite statistical analysis

A multivariate statistical analysis was performed using ropls (Version1.6.2, http://bioconductor.org/packages/release/bioc/html/ropls.html) R package from Bioconductor on Majorbio Cloud Platform (https://cloud.majorbio.com). Principle component analysis (PCA) using an unsupervised method was applied to obtain an overview of the metabolic data, general clustering, trends, or outliers were visualized. All of the metabolite variables were scaled to unit-variances prior to conducting the PCA. Orthogonal partial least squares discriminate analysis (OPLS-DA) was used for statistical analysis to determine global metabolic changes between comparable groups. All of the metabolite variables were scaled to pareto Scaling prior to conducting the OPLS-DA. The model validity was evaluated from model parameters R2 and Q2, which provide information for the interpretability and predictability, respectively, of the model and avoid the risk of over-fitting. Variable importance in the projection (VIP) was calculated in OPLS-DA model. p values were estimated with One-way ANOVA statistical analysis.

Differential metabolites analysis

Statistically significant among groups were selected with VIP value more than 1 and p value less than .05. Differential metabolites among two groups were summarized, and mapped into their biochemical pathways through metabolic enrichment and pathway analysis based on database search (KEGG, http://www. genome.jp/keg/). These metabolites can be classified according to the pathways they involved or the functions they performed. Enrichment analysis was usually to analyze a group of metabolites in a function node whether appears or not. The principle was that the annotation analysis of a single metabolite develops into an annotation analysis of a group of metabolites. Scipy. stats (Python packages) (https://docs.scipy.org/doc/scipy/) was exploited to identify statistically significantly enriched pathway using Fisher’s exact test.

Result

The basic Characteristics of the PE, PCH and the normal group

Six subjects with complete medical record were recruited for each group, meeting the inclusion and exclusion criteria. As shown in Table 1, there were no differences in age, BMI, or gestational weeks among the three groups. Alanine transaminase, creatinine, and uric acid may be elevated during pregnancy in pregnant women with PE.

Table 1. baseline characteristics of PE, PCH and normal group women.

Variate	PE	PCH	Normal	p-value
				PE VS Normal	PE VS PCH	PCH VS Normal
Maternal age	29.00 ± 2.09	28.50 ± 3.78	28.66 ± 2.42	.80	.78	.93
BMI (kg/m2)	28.69 ± 3.13	26.80 ± 4.31	27.78 ± 2.64	.60	.41	.65
Gestational week	37.60 ± 1.69	38.73 ± 1.48	40 ± .68	.07	.24	.06
BP at fecal collection
Systolic	144.33 ± 14.36	134.33 ± 21.46	117 ± 12.49	.006	.36	.12
Diastolic	95.16 ± 10.90	90.16 ± 8.65	69.83 ± 11.10	.003	.4	.005
24-h proteinuria collection (g)	3.78 ± 1.52	.16 ± .01	NA	.039	NA	NA
FBG (mmol/L)	4.6 ± .20	4.35 ± .39	4.31 ± .31	.08	.18	.87
TG (mmol/L)	4.31 ± 1.20	2.98 ± .98	NA	NA	.08	NA
TCH (mmol/L)	7.03 ± 2.02	6.00 ± .91	NA	NA	.32	NA
HDL-C (mmol/L)	2.02 ± .36	1.47 ± .21	NA	NA	.15	NA
LDL-C (mmol/L)	3.36 ± 1.37	3.53 ± .67	NA	NA	.81	NA
HCT (%)	38.21 ± 2.97	36.81 ± 3.47	37.43 ± 1.86	.59	.47	.71
AST (U/L)	24.16 ± 19.28	10.83 ± 3.43	10.66 ± 3.14	.12	.12	.93
ALT (U/L)	24.67 ± 6.02	16.50 ± 1.87	15.83 ± 3.37	.01	.01	.68
LDH (mmol/L)	204 ± 44.95	170.50 ± 34.67	189.33 ± 32.43	.52	.18	.35
CRE (umol/L)	61.33 ± 5.32	48.33 ± 4.45	49.83 ± 7.85	.01	.01	.69
UA (umol/L)	469.00 ± 28.17	315.67 ± 81.17	289.50 ± 47.68	<.001	.001	.51

PE: preeclampsia; PCH: pregnancy with chronic hypertension; N: normal group; BP: blood pressure; FBG: fasting blood glucose;TG: triglyceride; TCH: total cholesterol; HDL-C: high-density lipoprotein; LDL-C: low density lipoprotein;HCT: hematokrit; AST: oxaloacetic transaminase; ALT: Alanine transaminase; LDH: lactic dehydrogenase; CRE: creatinine; UA: uric acid.

Gut mycobiota in PE, PCH, and the normal group

Research have demonstrated that Ascomycota, Zygomycota, and Basidiomycota phyla dominate in healthy intestinal mycobiota [19–21]. Inflammatory bowel disease (IBD) [7], allergic airway inflammation [22] and alcohol abuse-induced cirrhosis [23] are associated with the increased abundance of Ascomycota, especially Candida spp. In our study, 1001 OTUs were identified and annotated based on an open-source, universal microbiome bioinformatics platform, QIIME2. Chi-square test was used for statistical analysis and p < .05 was considered to be statistically significant. At the phylum level the majority of the OTUs belong to Ascomycota (88.99% PE vs. 78.37% PCH vs. 70.04% Control, p = .27), followed by Basidiomycota (7.16%PE vs. 20.97% PCH vs. 10.16% Control, p = .08) and Mortierellomycota (2.16% PE vs. .01% PCH vs. .21% Control, p = .07) (Figure 1(a)).

Figure 1. Mycobiota composition and α diversity analysis of PE, PCH and normal group. a) Mycobiota composition of PE, PCH and normal group. b) α diversity analysis of PE, PCH and normal group. PE: preeclampsia, PCH: pregnancy with chronic hypertension, N: normal group.

The α diversity and β diversity of gut mycobiota

Though the Shannon index of gut mycobiota in the PE group was slightly lower than PCH and the normal groups, there was no statistical significance (Figure 1(b)). PCoA was used to evaluate β diversity. The horizontal and vertical coordinates are the two principal components respectively, and the percentage in parentheses is the percentage of variables that can be explained by the principal component. As shown in Figure 2(a), though ANOSIM analysis showed no statistical significance between PE and the normal group in mycobiota (p = .23), a distinct clustering of the fungal composition was observed between the PE group and the normal group according to the first principal component interpretation (PC1) and PC3 scores, which accounted for 29.05% and 15.45% (Figure 2(a), p = .232). Notably, a separation between the PE and PCH in gut mycobiota could be observed from PC1 and PC2 scores that accounted for 39.31% and 20.62%. ANOSIM analysis suggested that the gut fungal composition of the two groups was significantly different (Figure 2(b), p = .036).

Figure 2. PCoA analysis of PE, PCH and normal group. a) PCoA analysis between PE and normal group. b) PCoA analysis between PE and PCH group. PE: preeclampsia, PCH: pregnancy with chronic hypertension, N: normal group.

Fungal taxa differences in PE, PCH, and the normal group

In order to further investigate which taxa served as biomarkers among the groups, LEfSe was applied to explore the significant changes and relative richness of the fungal community. we found the relative abundances of the phylum Ascomycota, class Sordariomycetes, order Glomerellales were increased in the PE group compared to the normal group (Figure 3(a,b)), while Eurotiomycetes and Dothideomycetes were higher in the PE group than those in the PCH group (log LDA score > 3.5, p < .05) (Figure 3(c,d)).

Figure 3. LEfSe analysis PE, PCH and normal group. a) LEfSe analysis between PE and normal group. b) LEfSe Bar between PE and normal group. c) LEfSe analysis between PE and PCH group. D) LEfSe Bar between PE and PCH group. PE: preeclampsia, PCH: pregnancy with chronic hypertension, N: normal group. g_: genus, s_: species, f_: family, o_: order.

The metabolite characteristics of the PE, PCH, and the normal group

As fecal metabolome plays an important role during host-microbiome interaction, untargeted metabolomics analysis of fecal samples was applied to interrogate whether changes in gene abundance of gut mycobiota led to distinct alterations of gut metabolites among these groups. A total of 1227 metabolites were detected from all samples. OPLS-DA was conducted to analyze the metabolites of PE, PCH, and the normal groups. A distinct clustering of the cationic metabolite (Pantothenic Acid include, ndolelactic acid and Fucoxanthin, and so on) was observed between the PE group and the normal group according to the PC1 and PC2 scores, which accounted for 15.00% and 16.10% (Supplementary Figure 1a). Similar to cationic metabolites, there were significant differences in anionic metabolites (Methylnoradrenaline, L-Arginine and Taurine, and so on) between PE group and the normal control group (PC1 14.90%, PC2 15.60%) (Supplementary Figure 1b). In addition, the metabolites of PE group and PCH group were also significantly different. As shown in Supplementary Figure 2a, according to the PC1 (18.00%) and PC2 (14.80%) scores, the cationic metabolites of PE group and PCH group could be clearly distinguished. The anionic metabolites could also be distinguished, which PC1 accounted for 20.80% and PC2 accounted for 12.00% (Supplementary Figure 2b).

The variable importance in the projection (VIP) generated in OPLS-DA revealed the discriminating metabolites (VIP > 1 and p < .05) among these groups. Compared with the normal group, glycerophosphocholine were clearly increased in PE group, while other metabolites galactosylhydroxylysine and N2-Malonyl-D-tryptophan with highest VIP value were significantly decreased (Figure 4). Although PE and PCH are all belong to hypertensive diseases of pregnancy, their intestinal metabolites are quite different. As shown in Figure 5, coprocholic acid was significantly higher in the PE group, while 20-COOH-leukotriene E4 was most abundant in PCH group (Figure 5). KEGG pathway analysis showed that choline metabolism, ether lipid metabolism, and glutathione metabolism enriched in the PE group compared with the normal group (Figure 6(a)). It is worth noting purine metabolism was significantly enriched in the PE group compared with PCH group, which may account for the higher uric acid in PE patients [24] (Figure 6(b).

Figure 4. VIP analysis of PE and normal group. PE: preeclampsia, N: normal group. PE group included fb1-3, fb1-2, fb1-18, fb1-22, fb1-26, fb1-33. Normal group included fb1-47, fb1-48, fb1-52, fb1-53, fb1-54. *p < .05. The legend in the upper left corner belongs to the heat map. The red represents metabolites with increased abundance and the blue represents metabolites with decreased abundance. The legend in the upper right corner belongs to the VIP map.

Figure 5. VIP analysis of PE and PCH group. PE: preeclampsia, PCH: pregnancy with chronic hypertension. PE group included fb1-3, fb1-2, fb1-18, fb1-22, fb1-26, fb1-33. PCH group included fb1-10, fb1-19, fb1-20, fb1-25, fb1-27, fb1-28. *p < .05. The legend in the upper left corner belongs to the heat map. The red represents metabolites with increased abundance and the blue represents metabolites with decreased abundance. The legend in the upper right corner belongs to the VIP map.

Figure 6. KEGG analysis of PE and normal group. a) KEGG analysis between PE and normal group. b) KEGG analysis between PE and PCH group. PE: preeclampsia, PCH: pregnancy with chronic hypertension, N: normal group. HD: human diseases, M: metabolism, EIP: environmental infommation processing, OS: organismal systems.

Correlation between metabolite and fungal

Correlations between the abundances of metabolites and the relative abundances of the fungal were calculated. we found that Candida spp. was positively correlated with glycerophosphocholine, a metabolite involved choline metabolism. In addition, xanthosine, an intermediate product of purine metabolism, was also related with Candida spp. (Figure 7). Taken together, these results indicate the possibility of an adverse impact caused by redundant Ascomycota in pregnant women with PE. However, currently, little is known about the relationship between Ascomycota and PE as well as the mechanism by which these metabolites (glycerophosphocholine, and xanthosine) interact with the gut mycobiota. Future work should be conducted to verify the role of Ascomycota in the gut environment of PE.

Figure 7. Correlation analysis between metabolite and fungus. The correlation heatmap is used to analyze the correlation between fungal classification and metabolites. *p < .05.

Discussion

In this study, the changes of gut mycobiota of pregnant women with PE were analyzed and compared with those of chronic hypertension complicated with pregnancy and normotensive, uncomplicated pregnant women. We found that though there was no significant difference in composition of gut mycobiota between PE and control group, the abundance of Ascomycota in PE group significantly increased. LEfSe was applied to demonstrate that relative abundances of the phylum Ascomycota, class Sordariomycetes, order Glomerellales were enriched in PE group. Notably, untargeted metabolomics analysis showed that choline metabolism, ether lipid metabolism and glutathione metabolism enriched in PE group, which were closely related to Ascomycota. Since the main population of this study was preeclampsia, PCH and normal pregnancy were the controls. Therefore, PCH and the normal group were never compared in these analyses.

Though, previous studies have suggested that the pathogenesis of preeclampsia includes deep placental defects, oxidative and endoplasmic reticulum stress, angiotensin receptor type 1 autoantibodies, platelet and thrombin activation, endovascular inflammation, and endothelial dysfunction, the exact cause of PE remains unknown [1,25]. Since we now know that microbes especially gut microbiota can affect each of these four general mechanisms(oxidative and endoplasmic reticulum stress, platelet and thrombin activation, endovascular inflammation, and endothelial dysfunction) [4]. Yangyu Zhao demonstrated that gut microbiota dysbiosis led to an increase in plasma lipopolysaccharide and trimethylamine-N-oxide levels, then, participated in the development of PE [26]. Other studies have also shown that Short-chain fatty acids accompanying changes in the gut microbiome contribute to the development of hypertension in patients with preeclampsia [27].

Studies have found a link between gut bacteria and PE. Professor Huang revealed reduced gut bacteria diversity among pregnant women with PE. Women with preeclampsia had more opportunistic pathogens, particularly Fusobacterium and Veillonella, and less beneficial bacteria, including Faecalibacterium and Akkermansia [3]. Another study also acquired a similar conclusion, and their LEfSe analysis found 17 differentially abundant taxa between the PE and healthy groups. Jing Wang et al. also found, from the second trimester (T2) to the third trimester (T3), there was an obvious alteration in the gut microbiota. The gut microbiota of PE patients in T3 was significantly different from that of the control group [26]. However, they have overlooked another important role in the gut: the fungi. Molecular evidence suggests that animals and fungi have been co-evolving for a billion years, which had provided opportunities for fungi to greatly influence the evolution of animals and their immune systems [28]. With the rapid development of deep sequencing and computer technology, researchers found that fungal microbiota—the mycobiota have substantial effects on the host immune responses, though its proportion in the gut is much lower than that of bacteria. Shotgun metagenomics suggests that fungi constitute .01%–.1% of the human gut microbiome [11]. There are more than 50 species of human intestinal fungi, mainly including the following genera: Candida, yeast, Aspergillus, Cryptococcus, Malassezia, Cladosporium, galactosus and Trichospora [7]. Studies have shown that dendritic cells (DC cells), macrophages, and natural killer T cells recognize carbohydrate components such as glucan or mannan in fungal cell walls through their pattern recognition receptors (PRRs), mediating intestinal fungal immune tolerance [10]. When mycobiota flora is disturbed, TLRs (TLR2 and TLR4) receptors on DC cells recognize associated antigens and activate MYD88 or TRIF pathways to secrete IL-23, IL-1β, and IL-6, driving immune-assisted T1 immune cells (Th1 cells) and Th17 immune cells to respond. Meanwhile, IL-6 can inhibit the immunomodulatory properties of Treg cells and inhibit the differentiation of Treg cells, leading to the imbalance of Treg/Th17 immune cells. Meanwhile, some scholars believe that Candida albicans specific Th17 cells can promote the immunopathology of distant organs through cross-reaction with xenoantigen [8]. It is worth noting that the imbalance of Treg/Th17 immune cells is one of the mechanisms leading to the pathogenesis of PE [29]. Our study found that relative abundances of the phylum Ascomycota, class Sordariomycetes, order Glomerellales were enriched in PE group. Glomerellales is a common opportunistic pathogenic fungus, which could activate cellular immunity [30].

In addition, our study examined the metabolome of maternal feces samples to identify altered metabolite pathways that may provide new insights into the underlying biology of pre-eclampsia. The comprehensive analysis method combining gut microbiota diversity analysis and metabolome improve the comprehensiveness and reliability of the research method, making the research results more scientific. Through metabolic correlation analysis, we found that choline metabolism was closely related to fungi, while other metabolisms such as ether lipid metabolism and glutathione metabolism had little relationship with fungi and might be related to bacteria. We found that choline metabolism enriched in PE group and the abundance of glycerophosphocholine was significantly higher in patients with PE. At the same time, we also found that glycerophosphocholine was positively correlated to Ascomycota. It was reported that glycerophosphocholine could be converted into trimethylamine via gut microbiota, a precursor of trimethylamine oxide [31]. Yangyu Zhao demonstrated that the plasma trimethylamine-N-oxide (TMAO) concentrations of PE patients were higher than those of the healthy controls [5]. Previous studies have identified the TMA-TMAO pathway as a novel crossover between diet, gut microbiota, atherosclerosis, and the risk of thrombosis [32]. TMAO activates MAPK and NF-KB signals in vascular smooth muscle cells and endothelial cells, leading to inflammatory gene expression and adherence of endothelial leukocytes [33], which were significantly correlated with the occurrence of PE [5]. Therefore, we hypothesized that gut mycobiota are involved in the development of PE by affecting choline metabolism.

Conclusion

In summary, our findings first demonstrated that Ascomycota, especially Candida spp. could impact the abundance of glycerophosphocholine, which is related to the pathogenesis of PE. The compound mentioned above is expected to be an indicator to help predict the occurrence of PE. This research could help to provide new insights into the ongoing pathophysiological processes of this disease and could eventually lead to the identification of new drug targets.

Authors’ contribution

We thank HZ and YS designed the study, YS and CL performed the experiments. QH and WZ analyzed data. HZ wrote this manuscript. All authors read and approved the final manuscript.

Disclosure statement

The authors declare that they have no competing interests.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author (CL). The data are not publicly available due to their containing information that could compromise the privacy of research participants.

Additional information

Funding

This study was supported in part by Women and Children’s Hospital of Chongqing Medical University (Grant number 2020YJQN02).