Salivary lipid metabolism in periodontitis patients with spleen-stomach dampness-heat syndrome

Abstract

Background

Spleen-stomach damp-heat syndrome is one of the most common syndrome types in periodontitis from traditional Chinese medicine theory. However, its pathological mechanism is still uncertain. Tissue metabolism is driven by microbes in the host and its microenvironment. Hostmicrobe-metabolism is an interacting and closely related complex. Lipid metabolomics can find lipid metabolites in disease or healthy state, which is beneficial to explore the metabolic process and change mechanism of lipids that may be involved in organisms in healthy or disease state from the perspective of systems biology.

Methods

In this study, 10 patients in the periodontitis group (CP), 10 patients in the periodontitis with spleen-stomach dampness-heat syndrome group (SP) and 10 patients in the healthy group (H) were recruited for participation, whose unstimulated saliva was collected. The differential metabolites between the groups were detected by ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) and screened out based on the variable importance in projection (VIP) combined with the P-value and fold change (FC) value of univariate analysis. Finally, KEEG pathway enrichment analysis was performed on these differential metabolites.

Results

A total of 1131 metabolites were detected in saliva in this study. 497 metabolites were significantly up-regulated in periodontitis, mainly plasma-membrane-associated lipids, unsaturated fatty acids and oxidized lipids. Compared with the healthy group, the lipid metabolism pathways of periodontitis with or without spleen-stomach dampness-heat syndrome group were mainly characterized by significant enrichment of glycerophospholipid metabolism and unsaturated fatty acid metabolism such as arachidonic acid metabolism.

Conclusion

Compared with periodontally healthy patients, periodontitis with or without spleen-stomach dampness-heat syndrome can cause changes in lipid metabolism in saliva samples of patients. These metabolites are mainly plasma membrane lipids, unsaturated fatty acids and oxidized lipids quality. These lipids may be potential biomarkers of periodontitis. The downstream metabolites of unsaturated fatty acids in the saliva samples of patients with periodontitis and spleen-stomach dampness-heat syndrome were abnormal, and the oxidized lipid (±)5-HETE was significantly abnormal. We speculate that this may be related to the increased state of oxidative stress in saliva samples in disease states.

Introduction

Periodontitis refers to an inflammatory and destructive disease of periodontal tissue caused by microorganism-dominated multi-factors. According to previous reports, there are about 20–50% of the world’s population suffers from periodontitis [1]. According to the Global Burden of Disease Study, severe periodontitis affects approximately 1 billion people worldwide in 2019 [2]. This disease not only impairs the function of the oral chewing system, but also seriously affects the health of the whole body. How to effectively control the progress of periodontitis is a global public health problem that needs to be solved urgently.

The etiology of periodontitis is very complex. Some studies believe that its etiology can be attributed to Spleen-stomach dampness-heat, Kidney essence deficiency, or blood deficiency from the perspective of traditional Chinese medicine [3]. Spleen-stomach dampness-heat syndrome in periodontitis is one of the common syndromes. Its main symptoms include swollen, red, and bleeding gums, persistent bad breath, and severe constipation, which seriously affects the daily life of many people.

Lipids are fundamental building blocks of cells, which perform and regulate a variety of cellular responses. Ferguson et al. found that lipid mediator metabolites are associated with periodontal inflammation status, with differences in lipid mediator profiles between healthy gums and periodontitis and post-treatment periodontitis [4]. Pro-inflammatory lipid mediators have been reported to be significantly increased in periodontal inflamed tissue and gingival crevicular fluid [4]. Alterations of these metabolites may affect the microbial composition and inflammatory signaling in periodontitis, thus playing a key role in the mechanism of periodontitis and related systemic diseases. Research confirms that saliva can be used as a non-invasive diagnostic tool for assessing lipid profiles [5]. Related studies have shown an association between periodontitis and salivary lipid levels [6]. However, many studies have yet to show periodontitis-associated lipid metabolites at the level of the whole organism. As a complete system in the human body, lipid responses may involve changes at the level of lipid metabolism networks, thus requiring the analysis of overall lipids. Metabolomics is the comprehensive study of small-molecule metabolites (< 1,500 Da) within biological systems, which enables the systematic profiling of endogenous metabolites, such as lipids, amino acids, and organic acids, to uncover metabolic dysregulation associated with disease states or therapeutic interventions. Liquid Chromatograph-Mass Spectrometer technology(LC-MS) is a chromatographic analysis method using liquid as a mobile phase that combines the separation power of liquid chromatography with the sensitive detection and structural elucidation capabilities of mass spectrometry.In metabolomics, LC-MS is widely employed for high-resolution separation, accurate mass measurement, and quantification of metabolites in complex biological matrices. “Based on this, this study used UPLC-MS/MS method to detect lipid metabolism changes in periodontitis, periodontitis with spleen-stomach dampness-heat syndrome and healthy people.

Materials and methods

Study participants

Patients who presented to the Department of Stomatology of the First Affiliated Hospital of Jinan University from January 2022 to January 2023 were likely selected in this study. A total of 30 volunteers were recruited and divided into periodontitis (CP group, n = 10), periodontitis with spleen-stomach dampness-heat syndrome (SP group, n = 10) and healthy individuals (H group, n = 10). All periodontitis subjects were diagnosed with generalized/localized periodontitis, Stage II or III, Grade B, C or D, which was mainly defined by the criteria of Zhou M [7] et al. Spleen-stomach dampness-heat syndrome is mainly defined according to the symptoms. H group consisted mainly of individuals without periodontitis and self-reported general health.

The following patients will be excluded from our study: ① Any subjects diagnosed with metabolic diseases, such as hyperthyroidism or cancer; ② Subjects with a history of taking antibiotics, immunosuppressive drugs, etc. three months before the experiment; ③ Subjects who have received periodontal treatment such as supragingival scaling within 1 year; ④ Subjects who be pregnant or breastfeeding. All study subjects were required to fast for at least 8 h, and 2 ml of saliva was collected between 9:00 a.m. and 11:00 a.m. under the condition that they could not eat, rinse, brush their teeth and do other oral movements at least 2 h before sampling. The collected samples were placed in a − 80 °C freezer immediately after centrifugation. All participants signed a written informed consent and agreed to use their own saliva samples for this study. This study has been approved by the relevant ethics committee (approval number Kyk-2022-014).

Sample extraction

Remove saliva samples from − 80 °C freezer and place them on ice to thaw completely. The samples were then vortexed for 10 s to mix, and the samples were transferred to correspondingly numbered centrifuge tubes, and 1 mL of lipid extraction solution containing internal standard (MTBE: MeOH = 3:1, v/v) was added. After vortexing for 15 min, add 100 µL of water and vortex for 1 min. After centrifugation for 10 min at 4 °C, transfer 500 µL of the upper organic layer to the corresponding numbered centrifuge tube and concentrate to complete dryness. Before LC-MS/MS analysis, 200 µL of mobile phase B (acetonitrile/isopropanol) was added, vortexed for 3 min, and centrifuged for 3 min. The analytical sequences of samples were randomized by source and group to minimize the batch effect.

UPLC chromatographic acquisition and tandem mass spectrometry conditions

The instrument system is LC-ESI-MS/MS system (UPLC, ExionLC AD, https://sciex.com.cn/; MS, QTRAP^® System, https://sciex.com/). The liquid conditions mainly include: ① Column: Thermo Accucore™ C30 (2.6 μm, 2.1 mm*100 mm i.d.); ② Mobile phase A: acetonitrile/water (60/40, V/V) (containing 0.1% formic acid, 10 mmol /L ammonium formate); ③ Mobile phase B: acetonitrile/isopropanol (10/90, V/V) (containing 0.1% formic acid, 10 mmol/L ammonium formate); ④ Mobile phase gradient: 0 min for A/B(80:20, V/V)at 0 min, 70:30 V/V at 2.0 min, 40:60 V/V at 4 min, 15:85 V/V at 9 min, 10:90 V/V at 14 min, 5:95 V/V at 15.5 min, 5:95 V/V at 17.3 min, 80:20 V/V at 17.3 min, 80:20 V/V at 20 min; ⑤ The flow rate is 0.35 ml/min; the column temperature is 45 °C; the injection volume is 2 µl. Tandem mass spectrometry analysis conditions mainly refer to the research of Han et al [8].

Data processing

Qualitative analysis was carried out according to the retention time (RT) of the detected substance, the information of the parent-child ion pair and the secondary structure profile secondary spectrum data. Metabolite quantification was performed using triple quadrupole mass spectrometry in multiple reaction monitoring modes. Open the mass spectrometry file of the sample off-machine through the MultiaQuant software, and perform the integration and calibration of the chromatographic peaks. The peak area of each chromatographic peak represents the relative content of the corresponding substance. The raw mass spectral data obtained were processed with Analyst software.

Statistical and biological information analysis

The data of the study population were analyzed by SPSS, and p value < 0.05 was considered statistically significant. Welch’s t-test (for normally distributed data) or Mann-Whitney U test (non-parametric data) was applied to compare metabolite levels between cohorts. One-way ANOVA with Tukey’s post hoc test was used for comparisons across three groups. The Benjamini-Hochberg false discovery rate (FDR) method was employed to adjust p-values. Batch effect correction was performed using the ComBat algorithm. Mass spectral data were normalized using R software. The correlation between groups was analyzed using Pearson correlation. Using the MetaboAnalyst R software, the OPLS-DA method was used to eliminate variables irrelevant to the study, an OPLS-DA model was established to analyze the metabolome data, and the scores of each group were drawn to show the differences between the groups. In addition, according to the VIP value of each metabolite, the influence strength and explanatory power of each metabolite accumulation difference on the classification and discrimination of each group of samples were measured. VIP combined with FC and P value to screen differential metabolites. Finally, KEGG metabolic pathway enrichment analysis was performed for each differential metabolite.

Results

Baseline characteristics of the study population

As shown in Table 1, there was no significant difference in age, gender among CP group, SP group and H group (P > 0.05). There were significant statistical differences in CAL and PD among CP group, SP group and H group (P < 0.05).

Table 1 Social demographic characteristics, periodontal parameters of periodontitis and healthy groups

Salivary lipid metabolomics

In this study, UPLC-MS/MS technology was used to detect widely targeted metabolomes in the saliva of CP, SP and H groups. After baseline filtering, peak identification, integration, normalization and data correction, a total of 1131 metabolites were detected in the study. In order to detect experimental reliability, we set the detection process to 4 quality control (QC) samples. Pearson correlation analysis was performed on the QC samples (Fig. 1), and the result showed|r|>0.95, indicating the stability of the instrument during the experiment and the repeatability of metabolite extraction and detection good. Principal component analysis (PCA) was used to observe the trend of overall metabolic distribution and the magnitude of differences among all samples. In the PCA three-dimensional result score chart of each group (Fig. 2), PC1 is the first principal component in the multidimensional data matrix, the second principal component in the PC2 multidimensional data matrix, and PC3 is the third principal component in the multidimensional data matrix. As shown in Fig. 2, the distribution between CP group, SP group and H group showed a certain aggregation trend, while the distribution between CP group and H group showed a certain separation trend, suggesting that PCA model could not completely distinguish the characteristics of lipid metabolism among the three groups.

Fig. 1

Pearson correlation score plot for saliva QC samples. This figure comprises four samples(mix01, mix02, mix03, mix04) arranged in a 2 × 2 grid. Each subplot displays pairwise correlations between technical replicates, with data points tightly clustered along the diagonal line (red) representing perfect correlation (y = x). The Pearson correlation coefficients (r) are annotated in the upper-right corners of each subplot, ranging from ​0.9956 to 0.9966, indicating exceptionally high inter-sample agreement

Fig. 2

PCA score plot between CP, SP and H groups. This figure illustrates the separation of experimental groups (CP, TC, H) in principal component space through three subplots (a, b, c). PC1, PC2, and PC3 correspond to the first, second, and third principal components in the multidimensional data matrix, respectively

The OPLS-DA score plot showed that the lipid metabolites of the CP, SP and H groups were clearly separated (Fig. 3), that is, the lipid metabolism of the CP group, the SP group and the H group was different from each other.

Fig. 3

OPLS-DA score plot between CP, SP and H groups. This figure presents three Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA) score plots (a, b, c) comparing metabolic profiles across experimental groups.The x-axis displays predictive component scores (p [1]), with horizontal spread reflecting inter-group differences. The y-axis shows orthogonal component scores (pcorr [1]), with vertical spread indicating intra-group variability

In the S-plot, the red dots represent significant differences in metabolites (VIP ≥ 1), while the green dots represent no significant differences in metabolites (VIP < 1). In addition, the horizontal and vertical coordinates in the figure represent the cocorrelation coefficient and correlation coefficient score between principal components and metabolites, respectively. As shown in the S-plot in Fig. 4, there were differences in intergroup lipid metabolism among the three groups.

Fig. 4

S-plot between CP, SP and H groups. This figure presents three orthogonal partial least squares-discriminant analysis (OPLS-DA) S-plots comparing metabolic profiles across experimental groups. a: CP vs. SP, b: SP vs. H, a: CP vs. H. The x-axis represents the covariance coefficient between the principal component and metabolites. The y-axis represents the correlation coefficient between the principal component and metabolites. Red points indicate metabolites with VIP values ≥ 1. Green points indicate metabolites with VIP values < 1

Table 2 Differences in lipid metabolite distribution between groups

The screening criteria for saliva differential lipid metabolites were: (1) FC ≥ 2 and FC ≤ 0.5; (2) VIP ≥ 1.0; (3) P > 0.05. Based on this criterion, 465 differential lipid metabolites were screened between the SP and CP groups (Table 2 and Fig. 5.1b), 591 differential lipid metabolites were screened between the SP and H groups (Table 2 and Fig. 5.2), 352 differential lipid metabolites were screened between the CP and H groups (Table 2 and Fig. 5.3b), and their expressions were all up-regulated. It can be seen that the different lipid metabolites in each group are mainly glycerolipids, glycerolides, sphingolipids and fatty acyl lipids (Fig. 5.1a, Fig. 5.2a and Fig. 5.3a). According to log 2FC values, the top 20 different lipids among the groups were listed, mainly glycerophospholipid lipids and fatty acids (Fig. 5.1c, Fig. 5.2c and Fig. 5.3c). The results of this study showed that the unsaturated fatty acid metabolism such as arachidonic acid metabolism and its downstream oxidized lipid metabolite 5-HETE expression in saliva of periodontitis patients with spleen-stomach dampness-heat syndrome were significantly up-regulated compared with that of patients with simple periodontitis.

Fig. 5.1

Pie chart, volcano plot, top Fc bar chart between CP and SP groups. In the volcano plot, red dots denote upregulated metabolites, while gray dots represent non-significant metabolites (VIP ≥ 1, p < 0.05). In the fold change bar plot, the x-axis represents log2FC of differential metabolites, and the y-axis indicates their counts. Red bars signify upregulated metabolite abundance

Fig. 5.2

Pie chart, volcano plot, top Fc bar chart between SP and H groups

Fig. 5.3

Pie chart, volcano plot, top Fc bar chart between CP and H groups

Differential metabolite KEGG classification and enrichment pathway changes

Differentially expressed lipid metabolites were annotated into six major metabolic categories: cellular processes, environmental information processing, genetic information processing, human disease, metabolism, and organismal systems. Combined with the differential metabolite KEGG pathway enrichment bubble chart (Fig. 6.1,6.2 and 6.3), it was found that the abnormal lipid metabolism in the CP group and SP group were mainly manifested in glycerophospholipids and fatty acid metabolism (such as linoleic acid metabolism, arachidonic acid metabolism) compared with that in H group.

Fig. 6.1

Scatter plot for KEGG enrichment results between CP and SP groups. The y-axis labels metabolic pathways, while the x-axis shows the ratio of differential metabolites to total metabolites per pathway. Dot size correlates with metabolite enrichment magnitude in each pathway

Fig. 6.2

Scatter plot for KEGG enrichment results between SP and H groups

Fig. 6.3

Scatter plot for KEGG enrichment results between CP and H groups

Discussion

In recent years, traditional Chinese medicine has attracted widespread attention due to its unique theoretical viewpoints. Chinese medicine focuses on distinguishing the main syndrome of disease and seeking the most effective treatment. In this article, Spleen-stomach damp-heat syndrome is a common symptom of periodontitis. Due to dietary excess, such patients accumulate heat in their spleen and stomach, eventually leading to swollen gums and bad breath. However, the difficulty in clarifying the syndrome differentiation standards and efficacy indicators of traditional Chinese medicine has affected the credibility of the efficacy of traditional Chinese medicine and hindered the development potential of traditional Chinese medicine. Therefore, it is crucial to screen non-invasive, sensitive, and reliable biomarkers, and to study and explain the mechanism of traditional Chinese medicine in preventing and treating periodontitis for the diagnosis and treatment of periodontitis. In recent years, metabonomics research techniques have proven to be valuable analytical tools in the early diagnosis of many diseases [9]. Lipids are one of the main components in saliva, and their composition and content change in response to changes in the microenvironment. Saliva has been reported as an alternative diagnostic fluid for serum. Salivary lipids are important for maintaining oral health, and their increased expression has been confirmed to be closely related to oral diseases such as dental caries and periodontitis. Alterations in these lipids and related mediators may influence the microbial composition in periodontitis and thus mediate the progression of periodontitis [10, 11]. Therefore, more and more scholars have begun to perform metabolomics analysis on oral saliva to determine unique metabolite information. These potential biomarkers can accurately diagnose the current disease state, which helps to effectively prevent and treat periodontitis. Seonghye Kim et al. [12] used proton nuclear magnetic resonance spectroscopy to perform metabolomics analysis on saliva samples from 92 healthy controls and 129 periodontitis patients and verified 5 metabolite biomarkers of periodontitis: Ethanol, taurine, isovalerate, butyrate, and glucose. KUBONIWA et al. [13] identified metabolites such as cadaverine, 5-oxoproline, and histidine as potential indicators of periodontal inflammation. These markers are expected to be used to assess the severity of periodontal inflammation and provide a basis for monitoring disease activity in patients with periodontitis. In terms of the results of this study, if the plasma membrane lipids, unsaturated fatty acids, and oxidized lipids in saliva are significantly increased or at a high level compared to the previous ratio, the individual may be at risk of developing periodontitis. This prediction can prompt individuals to carry out targeted prevention early in life.

Previous studies have reported that metabolites such as carbohydrates, lipids, and nucleotides are altered in saliva in a state of periodontal inflammation [14, 15]. Recent studies on salivary lipid profiles have shown an association between periodontitis and salivary lipid levels, and that these lipid mediator profiles are associated with periodontal inflammation. However, most of these studies focus on triglycerides, cholesterol, unsaturated fatty acids, etc., and few of them have an overall large-scale understanding of the changes of salivary lipids in disease states. Therefore, in this study, UPLC-MS/MS technology was used to analyze the salivary lipid profiles of the CP, SP and H groups, and to screen potential lipid metabolites in the CP and SP groups on a large scale.

Compared with the H group, the CP group detected up-regulated expression of 352 differential lipids, the SP group detected up-regulated expression of 591 differential lipids, and these metabolites were mainly plasma membrane lipids, as well as unsaturated fatty acids and oxidized lipids. Plasma membrane lipids, one of the principal components of the cytoplasmic membrane, contain abundant phosphatidylethanolamine (PE), sphingomyelin (SM), and phosphatidylcholines (PC). Plasma membrane lipids determine the elasticity, fluidity, and permeability of cell membranes and also mediate intercellular transport and signal transduction mechanisms across oral and systemic tissues. In addition, the composition of plasma membrane lipids has been implicated in the pathophysiology of metabolic disorders. Previous studies have confirmed that there are abnormal expressions of PC, PE et al. in periodontal pocket exudates [16]. IIZUKA et al. found that the cell membranes of some subgingival bacteria associated with periodontitis contained large amounts of plasma membrane lipids [17]. Since the initial interaction of bacteria with oral tissue involves the formation of hydrophobic bonds, which can be stabilized by a lipid-rich environment, elevated lipid levels in saliva may predispose periodontitis-causing bacteria to the periodontal colonization in the organization [18]. We speculate that the abnormality of various lipid metabolites in the glycerophospholipid signaling pathway may be partly mediated by the persistence of periodontal pathogens.

Linoleic acid and arachidonic acid belong to the omega-6 series of unsaturated fatty acids. Under conditions of oxidative stress, this type of fatty acid is converted into various pro-inflammatory compounds, which play an important role in activating cellular signaling pathways associated with inflammation [19]. Lipoxygenase (LOX) can catalyze arachidonic acid to produce oxidized lipids such as 5-hydroxyeicosatetraenoic acid (5-HETE) [20]. As reported by Chen Jiao and colleagues, a significant increase in salivary unsaturated fatty acids (e.g., arachidonic acid) was observed in individuals with periodontitis [21]. Barnes et al. [22] found that the purine degradation metabolites level of periodontitis patients increased and the characteristics of ω-3 (docosapentaenoic acid) and ω-6 fatty acids (linoleic acid and arachidonic acid) increased. Similar to these study, our metabolomic analysis revealed abnormal fatty acid metabolism (e.g., linoleic acid metabolism, arachidonic acid metabolism, α-linolenic acid metabolism) in the CP group. And oxidation products of the fatty acid LOX metabolic pathway 5 -HETE expression was significantly up-regulated, suggesting that this pathway may play an important role in the pathogenesis of periodontitis. Huang et al. pointed out that the metabolites of omega-6 series of unsaturated fatty acids, especially arachidonic acid, can be used as important indicators of periodontitis inflammatory response and oxidative stress state [23]. In their study, it was found that local redox reactions in periodontitis respond to oxidative stress and inflammation by regulating fatty acid metabolism. Excess lipids in periodontitis can promote oxidative damage caused by oxidative stress [24]. It has been suggested that redox reactions and fatty acid imbalances are one of the hallmarks of bone destructive diseases, which is supported by some studies in periodontitis [25]. By reducing the proportion of arachidonic acid, the activity of osteoclasts can be inhibited and the resorption of alveolar bone can be slowed down [26]. Therefore, the abnormality of unsaturated fatty acids of ω-6 series and related metabolic pathways may be related to oxidative stress in periodontitis. The result of this study showed that the unsaturated fatty acid metabolism such as arachidonic acid metabolism and its downstream oxidized lipid metabolite 5-HETE expression in saliva of periodontitis patients with spleen-stomach dampness-heat syndrome was significantly up-regulated compared with that of patients with simple periodontitis. We speculate that this may be related to the altered oxidative stress state in the salivary fluid samples during the disease state.

However, our study could not confirm that abnormal lipid metabolism is a cause or an outcome of periodontitis. At the same time, the sample size of this study may limit the detection rate of some metabolites. Therefore, further research on the relationship between lipid metabolism and periodontitis is still needed in the future.

Conclusion

Compared with periodontally healthy patients, periodontitis with or without spleen-stomach dampness-heat syndrome can cause changes in lipid metabolism in saliva samples of patients. These metabolites are mainly plasma membrane lipids, unsaturated fatty acids and oxidized lipids quality. The downstream metabolites of unsaturated fatty acids in the saliva samples of patients with periodontitis and spleen-stomach dampness-heat syndrome were abnormal, and the oxidized lipid (±)5-HETE was significantly abnormal. We speculate that this may be related to the increased state of oxidative stress in saliva samples in disease states.