Background
The sign of the initiation of puberty is the reactivation of the hypothalamic-pituitary–gonadal axis (HPGA) [
1]. Central precocious puberty (CPP) is due to the early release of gonadotropin-releasing hormone (GnRH), which causes HPGA to activate prematurely. Precocious puberty could accelerated bone development, result in premature discontinuation of linear growth [
2], and increase the risk of type 2 diabetes [
3,
4], obesity, cardiovascular disease [
5], and breast cancer [
6,
7]. The prevalence of CPP is 5 to 10 times higher prevalence in girls than in boys [
8]. Globally, at least 0.2% of women experienced earlier puberty each year [
9]. However, the pathogenesis of CPP is not completely known and remains to be studied.
The microbiota-gut-brain axis (MBGA) refers to that gut microbiota affects the central nervous system by regulating intestinal neural, endocrine, and immunologic pathways. Moreover, this manner is often bidirectional [
10]. The role of gut microbiota on the host is not limited to modulating the host immunity, nervous and hormones [
11], but also regulating intestinal epithelial cells the blood brain barrier [
12], and the production and degradation of neuroactive compounds [
13], such as Nitrogen Monoxide (NO) [
14]. Microbial metabolites involved in MBGA include γ-aminobutyric acid (GABA), serotonin, butanoate, cortisol, and quinolinic acid [
15]. With the inclusion of extensive studies, the mechanism of interaction between gut microbiota and brain is becoming more and more clear. It’s worth noting that even though the relationship between precocious puberty and gut microbiota has been investigated [
16], the complex association between gut microbiota, metabolites, and CPP is largely unknown.
In this study, we dissected gut microbiome and metabolomics profiles from 150 participants to explore the correlations between gut microbiome features, metabolic features, and CPP. Bioinformatics and statistical analyses, including the comparisons of alpha and beta diversity, abundances of microbes, were undertaken on the 16S rRNA gene sequences and metabolism profiling. We revealed the widespread alterations of gut microbes and metabolites in CPP, which were involved in nitric oxide synthesis pathway. The results provided novel insights into recognizing potential therapeutic molecular markers for CPP.
Methods
Participants
In total, 150 fecal and blood samples (91 CPP patients, 59 healthy controls) were recruited from the Hainan women and Children’s medical center, Hainan Medical University. Patients who satisfied the following criteria were enrolled in the CPP group: (1) Girls younger than 10 years old. (2) Complying with the diagnostic criteria in the diagnosis and treatment guidelines of CPP: a) Secondary sexual characteristics appeared before 8 years in girls and progressed according to the normal developmental routine. (b) With evidence of gonadal development. (c) The height growth spurt during the development. (d) The gonadotropin elevated to pubertal level and luteinizing hormone-releasing hormone (LHRH) provocation test was positive. (e) The bone age was advanced 1 year more than the chronological age.
Participants were excluded when they met any of the following criteria. (a) Patients with other systemic diseases, including underlying diseases with clinical impacts (such as serious diseases of the heart, liver, kidney, lungs, and brain), tumors, abnormal glucose metabolism, immunodeficiency, and suffering tuberculosis, hepatitis B and C and other diseases within half a year. (b) Any history of other drugs, such as various antibiotics, used other than immune-pharmaceuticals (for instance, prednisone, tacrolimus cyclosporin, and cyclophosphamide) within 3 months. (c) Patients undergoing major gastrointestinal, inflammatory bowel diseases, long-term constipation, or diarrhea. (d) Coexisting other connective tissue diseases (such as Sjogren’s syndrome and overlap syndrome).
Sample collection and preparation
The subjects’ feces (greater than 400 mg) were collected into a sterile preservation tube with a sterile spoon, the Bristol Stool Scale scores were recorded. Then the feces were immediately placed into a – 80 °C freezer for cryopreservation for testing gut microbial and metabolite. About 3 ml of whole blood samples of the same subjects were collected using heparin anti-coagulant tubes. After staying still for 30 min at room temperature, all samples were centrifuged at 1300–2000 g for 10 min at 4 °C. After removing the upper plasma (not less than 0.3 ml), the samples were flash-frozen in liquid nitrogen followed by and preserved at − 80 °C to detect blood metabolites.
DNA library construction and sequencing
After extracting genomic DNA from the fecal samples using CTAB or SDS methods, the V4 variable region of 16S rDNA was amplified by PCR utilizing primers specific for Barcode and high-fidelity DNA polymerase. The library was quantified by Qubit and Q-PCR after construction by using TruSeq® DNA PCR-Free Sample Preparation Kit. Sequencing of the V4 variable region was performed through Illumina Miseq after the library was qualified.
16S rRNA gene sequencing data analysis
All the raw 16 s rRNA sequencing data were processed using QIIME software [
17]. The sequences with 97% resemblance for each sample were clustered into operational taxonomic units (OTUs) through Usearch algorithm [
18]. Then based on the reference sequence of the Silva database [
19], the OTUs representative sequence was used for species annotation using the UCLUST algorithm [
18]. The Chao index measured in OTU was used to evaluate alpha diversity. Beta diversity was calculated through the Bray–Curtis and was used to build principal coordinate analysis (PCoA). ANOSIM test was carried out to visualize and compare the differences in beta diversity between CPP and healthy control groups.
To identify the metabolomic features of subjects' fecal and blood samples, untargeted metabolomic analysis methods were performed using an ultra-performance liquid chromatography system with quadrupole-time-of-flight mass spectrometry (UPLC-QTOFMS), which was used to measure polar metabolites, such as organic acids.
LC–MS/MS analysis
All samples were separated by Ultra-high-performance liquid chromatography (UHPLC) reversed separation on the Agilent 1290 Infinity UHPLC. The conditions for detection were as follows: the temperature was 25 °C, the flow rate was 0.5 mL/min, and the sample injection volume was 2 μL. A mobile phase consisting of a binary solution was used: Mobile phase A contained water, 25 mM ammonium acetate and 25 mM ammonia, and Mobile phase B consisted of acetonitrile. The gradient elution process was as below: 95% B for 0–0.5 min, the concentration of B from 95 to 65% linearly in 0.5–7 min, B from 65 to 40% linearly for 7–8 min, B was kept at 40% for 8–9 min; B was changed from 40 to 95% linearly for 9–9.1 min, lastly, B remained at 95% for 9.1–12 min. The samples were placed within an autosampler at 4 °C throughout the analysis. To avoid the influences due to the signal fluctuations arising from the detection of the instrument, the samples were analyzed continuously in random order. QC samples were inserted into the sample cohort to monitor and assess the solidity of the system and the credibility of experimental data.
Quadrupole–time-of-flight conditions
Positive and negative ions were separated through hydrophilic interaction chromatography (HILIC), followed by UHPLC separation, then a Triple TOF® 6600 (AB SCIEX) was intended for the mass spectrometer. The ESI operating conditions were as below: nebulization pressure (Gas1) was set as 60, adjuvant air pressure (Gas2) was 60, curtain gas was 30, ion source temperature was 600 ℃, the ion spray voltage was 5500 V in the positive ion mode and − 5500 V in the negative ion mode, the m/z range of TOF MS and daughter ion scanning were 0.20 s/spectra and 0.05 s/spectra, respectively. Secondary mass spectrometry was acquired by information-dependent acquisition (IDA) with the high sensitivity mode: the declustering potential (DP) was ± 60 V (positive and negative mode), the collision energy was 35 ± 15 eV, excluding isotopes within 4 Da, and the candidate ions to monitor per cycle: 10.
Random forest classification
To identify biomarkers in gut microorganisms, fecal metabolites, and blood metabolites that could be used to distinguish CPP patients from the population, a random forest model based on gut microorganisms, fecal metabolites, and blood metabolites was constructed using the R package “randomForest” (Version 4.7–1.1) to identify the important features. The combined dataset of the CPP group and healthy control group was randomly split into the training set and testing set with a ratio of 7:3. In addition, the Boruta algorithm of the R package “Boruta” (Version 7.0.0) was used to select gut microorganisms, fecal metabolites, and blood metabolites that could make significant contributions to the classification, and based on the selected features constructed a model. The area under curve (AUC) of the receiver operating characteristic (ROC) curve was plotted using the R package “pROC” (Version 1.18.0) to evaluate the model performance.
The pathways of the gut microbiome and the activity of gut-brain modules were predicted by PICRUSt2 [
20]. Differences in pathways abundances between the CPP and healthy control groups were calculated by
t-test, and
p-values were corrected applying Benjamini-Hochberg (BH) adjustment.
According to the Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolite database [
21], we used hypergeometric tests to perform the functional annotation of fecal and blood metabolites, and Benjamini-Hochberg (BH) adjustment was applied to correct
p-values. The pathways were considered as significantly enriched only if the number of their corresponding differential metabolites was at least 3 and their corrected
p-values were less than 0.05.
We used correlation analysis to identify differential metabolites associated with differential microorganisms. The Spearman correlation coefficient was calculated. Only the differential metabolites with the correlation coefficients R2 ≥ 0.3 or R2 ≤ − 0.3, and BH corrected p-values < 0.05 were selected as metabolites potentially affected by microorganisms.
Statistical analysis
All statistical analysis and charting were performed using R software (version 4.1.2). Kolmogorov–Smirnov test was used to check the normal distribution of the data. Wilcoxon’s rank-sum test was used to calculate the differences in metabolite and microbial abundances, which are non-normally distributed. The Spearman correlation coefficient was determined by the R package “corrplot”, with a corrected significance threshold of p < 0.05. Additionally, partial least square discriminant analysis (PLS-DA) was used to analyze the between-group difference. BH-adjusted p-values < 0.05 was considered statistically significant.
Discussion
In this study, we have discovered the alterations in the characteristics of gut microbiota, fecal and blood metabolites in patients with CPP, and identified some microbial and metabolite candidates that may be useful for CPP treatment. Genus Bifidobacterium and Streptococcus were highly enriched in CPP, both of which were associated with signal transduction of GnRH [
32,
37]. Fecal metabolites M333T401_1 (Estrone sulfate), M367T29 (3-dehydroepiandrosterone sulfate), M465T86 (Androsterone glucuronide), M329T110 (11beta-hydroxyprogesterone) and M345T107 (Corticosterone) showed a significant difference between CPP and healthy controls, and these five metabolites with a weak overall up-regulated tendency mediated steroid metabolism. Among of them, Estrone sulfate and Androsterone glucuronide were highly enriched in CPP, which correlated with GnRH activation [
38,
39]. While Corticosterone was highly enriched in healthy controls, possibly due to its role in downregulating follicle-stimulating hormone and luteinizing hormone [
40]. The blood metabolites M450T211 (Glycochenodeoxycholate), M466T255 (Glycocholate), M443T355 (Trihydroxycoprostane), M426T226 (Cholic acid), and M498T160 (Taurochenodeoxycholic acid) showed significant differences in CPP and healthy controls. These five metabolites with a generally downregulated trend mediated the primary bile acid biosynthesis pathway, among which the organic acid Glycocholate could promote the absorption of sex hormones [
41]. The up-regulation of Glycocholate in healthy control group meant that Glycocholate in CPP patients may be degraded because of promoting the release of sex hormones.
In addition to a brief description of changes in gut microbes, fecal, and blood metabolites in CPP, our study also identified CPP-related KEGG metabolic pathways and neuroendocrine GBMs. Previous studies have reported that Tetracycline biosynthesis, Bisphenol degradation, Lysosome, Flavonoid biosynthesis, and NO synthesis pathway were associated with the treatment and pathogenesis of CPP [
24‐
28]. During the functional annotations of fecal and blood metabolites, we discovered the dysfunctional pathways in disease progression and quantified the roles of metabolites in pathways using differential abundance scores. Such as cholesterol metabolism and primary bile acid biosynthesis enriched in blood metabolites, which were mediated by upregulated blood metabolites, while cholesterol and bile acids were discovered to be contributed to the treatment of precocious puberty [
31].
Conventional studies on the pathogenesis of CPP have focused on host genetics and peripheral factors [
42], and several studies have analyzed the gut microbiota in CPP [
16], but only identified dysregulated gut microbiota and did not analyze downstream pathways. This study not only revealed dysregulated microbiota and metabolites in CPP through analysis of microbiome and metabolome, but also linked them together, aiming to characterize the influence of microbiota on the body using metabolites, providing a new perspective for the diagnosis and treatment of CPP. Together, the study of CPP requires the integration of multiple omics data to comprehensively depict the dynamic changes of response factors in the disease process and discover the pathogenesis.
However, we also realized several limitations of this study. Although 16 s rRNA sequencing was widely used to characterize microbial communities, it existed limitations in explaining complete genetic information compared to metagenomic sequencing. Nevertheless, 16 s rRNA gene sequencing was mature technology and was enough for massive research. Furthermore, for the studies of metabolites, candidate microorganisms need to be further cultured to judge the origin of metabolites more accurately. Precocious puberty was often related to obesity [
43], but the population collected in this study did not contain people with obesity which made it impossible to explore the co-occurrence effect of microorganisms, metabolites, and obesity on CPP, but this direction deserves to be studied.
Conclusions
In conclusion, we integrated for the first time microbiomics and metabolomics to characterize systematic changes in gut microbes, fecal and blood metabolites in CPP patients. We revealed the microbial and metabolite features associated with CPP, interpreted the correlation between the two in the setting of CPP, and developed a predictive model to distinguish and diagnose for CPP.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.