Periodontal Status and Saliva Metabolic Signature in Patients with Alzheimer’s Disease

Abstract

Background:

Characterizing the periodontal status of patients with Alzheimer’s disease (AD), investigating differences in salivary metabolism between patients with and without AD under the same periodontal conditions, and understanding how it is related to oral flora are critical.

Objective:

We aimed to examine the periodontal condition of patients with AD and to screen salivary metabolic biomarkers from the saliva of individuals with and without AD with matched periodontal conditions. Furthermore, we aimed to explore the possible relationship between salivary metabolic changes and oral flora.

Methods:

In total, 79 individuals were recruited into the experiment for periodontal analysis. Especially, 30 saliva samples from the AD group and 30 from healthy controls (HCs) with matched periodontal conditions were selected for metabolomic analysis. The random-forest algorithm was used to detect candidate biomarkers. Among these, 19 AD saliva and 19 HC samples were selected to investigate the microbiological factors influencing the alterations in saliva metabolism in patients with AD.

Results:

The plaque index and bleeding on probing were considerably higher in the AD group. Further, Cis-3-(1-carboxy-ethyl)-3,5-cyclohexadiene-1,2-diol, dodecanoic acid, genipic acid, and N, N-dimethylthanolamine N-oxide were determined as candidate biomarkers, based on the area under the curve (AUC) value (AUC = 0.95). The results of oral-flora sequencing showed that dysbacteriosis may be a reason for the differences in AD saliva metabolism.

Conclusion:

Dysregulation of the proportion of specific bacterial flora in saliva plays a vital role in metabolic changes in AD. These results will contribute to further improving the AD saliva biomarker system.

Keywords

Alzheimer’s disease metabolomics microbiota periodontitis saliva

INTRODUCTION

Saliva is a complex extracellular fluid critical for lubrication, digestion, and oral health. Hundreds of components are present in saliva, including various electrolytes, enzymes, and antimicrobial agents. Many substances enter saliva from the blood capillaries. Therefore, saliva is a “mirror” of the physiological condition of the body [1]. There is evidence that saliva is sometimes functionally equivalent to plasma as a diagnostic fluid for systemic diseases [2]. For instance, the saliva-based glucose test has been proven to be a substitute for blood tests for monitoring diabetes [3]. Additionally, saliva collection is non-invasive and convenient and has advantages over blood collection [4]. Some systemic diseases can affect oral health by affecting the quality and quantity of saliva [5]. Alzheimer’s disease (AD) presents a heavy burden on the society and family of the patients [6]. The latest data show that AD prevalence will increase nine-fold globally by the year 2050 according to the biological definition of AD [7]. Thus, biomarkers for AD diagnosis are of great importance. Previous studies have demonstrated changes in salivary metabolism in patients with AD, and they screened some metabolites with significant differences compared with those in healthy people, such as arginine, taurine, and ornithine [8 –10]. Moreover, it has been reported that clinical periodontal indicators, including the probing depth (PD), clinical attachment level (CAL), percentage of bleeding on probing (BOP), and plaque index (PI) of patients with AD were worse than patients without AD [11, 12]. In addition, periodontitis-related salivary microbiota has been shown to aggravate AD [13 –15]. Therefore, the combined analysis of the differences in oral microbes and saliva metabolism in such patients may be an essential tool to investigate the onset and progression of AD.

Metabolomics is an analytical method for identifying and quantifying metabolites that are widely used in the saliva field. Analyzing changes in the saliva metabolism of patients with AD will help us explore the changes in the oral environment, thereby further revealing their impact on the onset and progression of Alzheimer’s disease. Compared to alternative analytical platforms in nontargeted analysis, liquid chromatography– mass spectrometry (LC-MS) has advantages with regard to velocity, sensitivity, comparatively effortless sample preparation, and expansive dynamic range [16]. Therefore, LC-MS was used to compare changes in the saliva metabolism between patients with AD and healthy controls (HCs). Moreover, we explored the changes in periodontal conditions in patients with AD and the microbial factors that may cause changes in AD salivary metabolomics.

MATERIALS AND METHODS

Experimental design and study participants

In this cross-sectional study, 43 patients diagnosed with AD and 36 HCs with matched age and gender were recruited from West China Hospital and West China School of Stomatology from the year 2021 to 2022 (Supplementary Table 1). This study was conducted in accordance with the tenets of the Declaration of Helsinki. All procedures were approved by the Ethical Committee of West China Hospital, Sichuan University (approval number: 1672). All included participants provided written informed consent. Patients with AD were diagnosed according to the criteria established by the National Institute of Neurological and Communication Disorders and the Stroke-Alzheimer’s Disease and Related Disorders Association [17]. The Mini-Mental State Examination (MMSE) was used to assess patients with AD. Those receiving anti-inflammatory drugs, antibiotics, probiotics, glucocorticoids, or basic periodontal treatment within a 3-month period and those who had other serious systemic diseases (i.e., heart diseases, diabetes, and stroke) were not included. Full-mouth periodontal screening was performed by one investigator who was blinded to the study allocation to assess CAL, PI, BOP, PD, and tooth loss. Participants were also categorized by their periodontal health status according to the 2018 classification [18]. Saliva samples were collected from patients who met the inclusion criteria and provided signed informed consent. Prior to saliva sampling, participants were instructed to avoid smoking, eating, drinking, and performing oral hygiene activities for 2 h in the morning. A 50 ml tube was used to collect approximately 3 mL of spontaneous and unstimulated saliva from each participant. The samples of saliva were centrifuged for 1 min at 4°C at 1,000×g for removing insoluble substances and dietary residues. Then, the collected supernatant was frozen at –80°C for LC-MS and 16 S ribosomal RNA (16 S rRNA) sequencing analysis. Finally, we obtained 30 saliva samples from the AD group and 30 saliva samples with basically matching periodontal conditions and indices from the HC group and conducted metabolomic analyses (Supplementary Table 2, Supplementary Figure 2). Then, 19 saliva samples each from patients with AD and HCs with matching periodontal indices were selected from these 60 samples for 16 S rRNA sequencing to explore the relationship between oral metabolism and oral flora (Supplementary Figure 5).

Sample preparation and LC-MS detection

Methanol/acetonitrile (1 : 1 v/v), 800μL, and 0.02 mg/mL L-2-chlorophenylalanine were combined with saliva samples (200μL) as an internal reference. The mixture was vortexed for 30 s, followed by 30 min of low-temperature ultrasonic extraction (5°C, 40 kHz) and preservation at –20°C for 30 min. Then, the supernatant was removed from the samples and put in an injection vial after they had been spun at 13,000×g and 4°C for 15 min. To conduct quality control (QC), 20μl of the supernatant from each sample was separated.

An ultra-high-performance liquid chromato-graphy-tandem Fourier transform mass spectrometer UHPLC-Q Exactive HF-X system (Thermo Fischer Scientific, Waltham, MA, USA) with electrospray ionization techniques was used to conduct the MS studies.

Chromatographic condition: we used the Acquity UPLC HSS T3 (100 mm) 2.1 mm i.d., 1.8 m (Waters Corporation, Milford, MA, USA) chromatographic column; the injection volume was 2μL; the column temperature was set at 40°C; the specific conditions for mobile phase A were 95% water + 5% acetonitrile (containing 0.1% formic acid), and the specific conditions for mobile phase B were 47.5% acetonitrile + 47.5% isopropanol + 5% water (containing 0.1% formic acid).

The MS conditions are presented below. Electrospray ionization was used to gather samples, with both positive and negative ion scanning modes. The data was collected using both positive and negative scan modes. QC samples were prepared by combining extracts from all samples. QC sample had the same processing, testing, and volume parameters as did in the experimental samples. Every 5–15 experimental samples were put into a QC sample during instrumental analysis to assess the stability of the detection procedure.

Metabolite identification and data analysis

The initial results were then analyzed using ProgenesisQI metabolomics processing software (Waters Corporation), including baseline filtering, peak identification, peak alignment, and other processing methods to obtain retention time, mass-to-charge ratio, and peak intensity. Next, we searched and identified the characteristic peaks in the software and then matched the MS and MS/MS information with the corresponding data in the metabolic database one by one. MS error was required to be < 10 ppm, and metabolite identification was determined on MS match scores. The main databases used in the analysis are available at the following links: http://www.hmdb.ca/ and https://metlin.scripps.edu/.

16S rRNA gene sequencing analysis

We used 1% agarose gel electrophoresis was used to further analyze the isolated genomic DNA. Based on the chosen sequencing region, the matching particular primers were synthesized. To ensure that the results of subsequent data analysis are highly accurate and reliable, two conditions must be met during the DNA amplification process: 1) use a relatively low number of cycles to amplify as much as possible and 2) ensure that each sample is amplified at the same time. The numbers of cycles were identical. During this process, we conducted pre-experiments by randomly selecting representative samples to guarantee that most of the sample could eventually be used to amplify the appropriate concentration of products with the least number of cycles. Polymerase chain reaction (PCR) was performed using TransGen AP221-02 (TransStart Fastpfu DNA polymerase; TransGene Biotech, Beijing, China; PCR machine: ABI GeneAmp® 9700, Thermo Fischer Scientific). All samples were analyzed at the formal experimental conditions. Each sample was examined three times. The PCR final mixture product of the same sample was detected using 2% agarose gel electrophoresis. To recover the product of PCR, the AxyPrepDNA Gel Recovery Kit (Axygen Inc., San Francisco, CA, USA) was employed for slicing the gel, followed by Tris-HCl elution and 2% agarose electrophoresis detection. Then, a MiSeq library (Illumina Inc., San Diego, CA, USA) was constructed and sequenced using MiSeq. The obtained PE reads were sampled after sequencing using MiSeq, and the QC and filtering of the paired-end reads were carried out according to the sequencing quality. Then, the optimized data were processed using sequence denoising methods (i.e., DADA2/Deblur) to obtain related representative sequences and corresponding abundance information based on amplicon sequence variant (ASV) conditions. Thus, the taxonomic, community diversity, and species difference analyses were carried out for the representative sequence and abundance information under ASV conditions.

Quality and risk of bias assessment of included studies

All the procedures in this study were performed in accordance with the STROBE checklist.

Statistical analysis

Partial least-squares discriminant analysis (PLS-DA) and orthogonal partial least-squares discriminant analysis (OPLS-DA) were performed to determine the overall variance between the two groups. Then, random forest (RF)-R (version 3.3.1) was applied to screen candidate AD salivary metabolism-related biomarkers and estimate the classification error by out-of-bag error. To assess diagnostic effectiveness, the receiver operating characteristic curve (ROC) was generated. The Kruskal– Wallis test was used to detect the difference between the periodontal statuses of the two groups using IBM SPSS Statistics 26 (IBM Corp., Armonk, NY, USA).

RESULTS

Discovery of the AD signature periodontal characteristics

We performed periodontal screening in 43 patients with AD and 36 HCs (mean age, 75.2 and 72.9 years, respectively; Table 1). MMSE scores significantly distinguished the two groups.The percentages of BOP and PI were higher in the AD than in the HC group, and the differences were statistically significant. PD, CAL, and number of remaining teeth between the two groups were not significantly different (Table 2, Supplementary Figure 1). Moreover, the AD group consisted of 13, 20, and 10 cases of stage III periodontitis, stage IV periodontitis, and edentulousness, respectively. The HC group included 18, 10, and 8 individuals with stage III periodontitis, stage IV periodontitis, and edentulousness, respectively. Moreover, we found no statistically significant variations in periodontal conditions between them after performing the Kruskal– Wallis test (Table 3). At the same time, we obtained consistent results using the same approach for 60 samples in metabolomics and 16 S rRNA-involved periodontal conditions (Supplementary Tables 2 and 3). It has been reported that there is no difference in salivary metabolites and bacterial diversity between different stages of periodontitis [10, 19]. In this study, we took the periodontal-specific parameters BOP, PI, CAL, PD, and number of teeth as a reference, conduct metabolomics and bacterial diversity studies, and tried to match the periodontal parameters of the two groups.

Table 1

Cohort demographic characteristics

	HC	AD
N	36	43
Age (mean±SD)	72.9±6.9	75.2±9.1
Male/female (n)	23/13	26/17
MMSE (mean±SD)	–	15.9±6.0
Smoking (n)	10/36	19/43

AD, Alzheimer’s disease; HC, healthy control; MMSE, Mini-Mental State Examination; SD, standard deviation.

Table 2

Periodontal clinical parameters of participants

	HC	AD
PI (mean±SD)	29.6±12.7	72.2±21.8
BOP (mean±SD)	47.9±21.4	54.1±28.6
PD (mean±SD)	2.7±0.5	2.9±1.1
CAL (mean±SD)	3.3±1.0	3.8±1.7
No. of teeth (mean±SD)	16.6±11.7	13.3±10.8

AD, Alzheimer’s disease; BOP, bleeding on probing; CAL, clinical attachment level; HC, healthy control; PD, probing death; PI, plaque index; SD, standard deviation.

Table 3

Periodontal status of participants

	Stage III	Stage IV	Edentulous	Total
HC	18	10	8	36
AD	13	20	10	43
Total	31	30	18	79

p = 0.195. AD, Alzheimer’s disease; HC, healthy control.

Altered salivary metabolomic pattern in AD

Metabolic test results revealed that the PLS-DA and OPLS-DA score plots separated the saliva samples into two clusters (Fig. 1A, B), and 152 differential metabolites in the AD group and HC were picked if the variable importance for projection (VIP) was more than one. To explore the metabolites that significantly contributed to the clustering between the groups, a VIP plot was generated from the OPLS-DA results with a threshold of 2. Therefore, 35 discriminating variables were obtained (Fig. 1C), including cis-3-(1-carboxy-ethyl)-3,5-cyclohexadiene-1,2-diol (VIP = 4.03), genipic acid (VIP = 5.06), N,N-dimethylethanolamine N-oxide (VIP = 4.03), 12-hydroxydodecanoic acid (VIP = 3.83), and dodecanoic acid (VIP = 2.29). At the same time, we found that there was no significant metabolic difference among different periodontitis grades in all participants (Supplementary Figure 3).

Fig. 1

Sample comparison analysis results. A) PLS-DA (R² X = 0.502; R² Y = 0.987; Q² = 0.854). B) OPLS-DA plot (R² X = 0.41; R² Y = 0.94; Q² = 0.63). C) Expression profile and VIP of metabolites. Contributions to metabolite differences between the two groups are indicated by the length of the bars. The longer the length, the more obvious the difference between the two groups. The color of the bar represents the p-value. The darker the color, the smaller the p-value. On the right, *p < 0.05, **p < 0.01, ***p < 0.001. AD, Alzheimer’s disease; HC, healthy control; OPLS-DA, orthogonal partial least-squares discriminant analysis; PLS-DA, partial least-squares discriminant analysis.

RF models deciphered possible salivary biomarkers for AD

We further employed an RF algorithm to screen the candidate salivary biomarkers for AD. Fifteen metabolites were selected as candidate biomarkers of the test set, with an out-of-bag value of 0.225 (Fig. 2a, b). Combining the results of the training set (Fig. 2c, d), two sets overlapping to yield four metabolites showed higher reliability as biomarkers, including cis-3-(1-carboxy-ethyl)-3,5-cyclohexadiene-1,2-diol, dodecanoic acid, genipic acid, and N,N-dimethylethanolamine N-oxide (Fig. 3A–D). The ROC value of the four metabolites was 0.95, suggesting good predictability (Fig. 3E, F).

Fig. 2

RF results of the test set (a, b) and training set (c, d). The importance of variables is proportional to the MDA value. MDA, mean decrease accuracy; RF, random forest.

Fig. 3

RF model of candidate metabolite biomarkers for AD. A– D) Candidate metabolite biomarkers. E, F) Receiver operating characteristic analysis of the four candidate biomarkers and their combinations. AD, Alzheimer’s disease; HC, healthy control.

Differential metabolites involved in Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathways

The nervous system, signaling molecules, and functional interaction pathways in the KEGG drew our attention (Fig. 4). These nervous system-related metabolic pathways enriched four differential metabolites (i.e., acetylcholine, endomorphin-2, PBG2, and L-glutamine). Except for endomorphin-2, the levels of the remaining three metabolites were lower in the AD group. Simultaneously, we discovered that the eight differential metabolites in the amino acid metabolic pathway were mainly common oral microbial host cometabolites and were mainly related to the metabolism of tryptophan, arginine, and cysteine (Fig. 5). It indicates that specific bacteria in the mouth might have a critical function in the changes in saliva metabolism in patients with AD.

Fig. 4

The results of KEGG pathway analysis (a) and KEGG enrichment analysis (b). KEGG pathway enrichment analysis refers to the enrichment analysis of the selected metabolic set. When the p-value is < 0.05, this pathway was considered to be significantly enriched. KEGG, Kyoto Encyclopedia of Genes and Genomes.

Fig. 5

Products in the amino acid cometabolism pathway of oral microbial hosts. AD, Alzheimer’s disease; HC, healthy control.

Comparison of oral microbial diversity

Studies have shown that changes in the intestinal and salivary flora have a certain impact on the occurrence and development of AD [20 –22]. Our extensive metabolic results also suggest that oral microbes may cause corresponding changes in salivary metabolism in patients with AD. We collected 38 saliva samples again in the same way and performed microbial diversity analysis to explore the changes in oral microbes in such patients. Alpha diversity showed that no significant difference was observed in the diversity of oral flora between the AD and HC groups (Fig. 6a, b). However, differences in bacterial abundance were found at the genus-specific level between the AD and HC groups (Fig. 7a, b) for Streptococcus , Leptotrichia, Gemella, and Haemophilus. Meanwhile, we found no significant differences in the comparative analysis of species at the genus level for all participants at different periodontitis stages (Supplementary Figure 4).

Fig. 6

Community diversity analysis. a) Comparison of alpha diversity between the groups. b) Gene-level PCA results of the two groups. AD, Alzheimer’s disease; HC, healthy control; PCA, principal components analysis.

Fig. 7

a) Comparative analysis of two groups of species at the genus level. b) Community Circos Diagram. The Circos sample-species relationship diagram describes the corresponding relationship between samples and species. *0.01 < p≤0.05, **0.001 < p≤0.01. AD, Alzheimer’s disease; HC, healthy control.

DISCUSSION

In this study, we analyzed the periodontal status of 79 participants and found that the PI and percentages of BOP of patients with AD were considerably higher than those of HCs. At the same time, the PI and BOP indices are directly related to oral hygiene [23]. A healthy oral environment greatly impacts the life quality of AD patients [24, 25], suggesting that oral hygiene in patients with AD requires attention. In addition, plaque is the initiating factor of periodontitis, and the bleeding index is an important indicator of periodontal inflammation. Higher plaque and bleeding indices will undoubtedly further aggravate periodontal damage.

In our research, we compared alterations in salivary metabolism and its relationship with bacterial diversity between patients with and without AD under the matched periodontal indexes. We performed metabolomic research in saliva samples to screen four biomarkers. The dodecanoic acid and genipic acid levels were higher in patients with AD. Interestingly, both acids have antibacterial effects [26, 27]. We speculate that the upregulation of these two metabolites is a self-protective mechanism in patients with AD. Analogs of cis-3-(1-carboxy-ethyl)-3,5-cyclohexadiene-1,2-diol were revealed as products of the microbial metabolism pathway based on KEGG, which once again reveals the important influence of microbes on the saliva metabolism of patients with AD. N,N-dimethylethanolamine N-oxide may be the oxide of N,N-dimethylethanolamine, a controversial precursor of acetylcholine [28].

Some metabolic pathways identified in the KEGG enrichment analysis attracted our attention. For example, in the neuroactive ligand-receptor interaction pathway, we found that acetylcholine and endomorphin-2 levels were clearly distinct comparing the two groups. The acetylcholine’s level is downregulated in the AD group, and loss of acetylcholine is directly related to cognitive decline [29, 30]. Endomorphin-2 protects the mouse brain from intracellular amyloid-β toxicity [31]. However, its expression level was lower in the AD group. These two salivary metabolites confirm the neuropathological changes in patients with AD.

Simultaneously, we found that changes in the relative abundance of specific bacterial genera in the mouth of patients with AD may cause corresponding changes in the oral metabolic pathways detected with metabolomics and in bacterial diversity. Tryptophan can be used as a precursor for endogenous biochemical reactions in the human body to synthesize tryptamine, serotonin, and melatonin. Oral gram-negative bacteria Prevotella intermedia, Fusobacterium nucleatum, and Porphyromonas gingivalis can degrade tryptophan to produce a range of indole and phenolic compounds [32, 33]. Moreover, 3-methyldioxyindole, indole-3-ethanol, and 3-methylindole (Fig. 5D–F) were enriched in tryptophan metabolic pathways, and their contents were lower in the AD group, in consistency with the result that the proportion of gram-negative bacteria was lower in the AD according to the oral microbial diversity test (Fig. 7). In patients with AD, through the interaction between oral microorganisms and the host, the proportion of gram-negative bacteria is reduced, and the synthesis of tryptamine, melatonin, and serotonin, which are beneficial to the brain, is increased. As the proportion of oral gram-negative bacteria in patients with AD was decreased, the level of the cysteine-methionine metabolite N-formylmethionine in the common metabolic pathway of gram-negative bacteria was correspondingly reduced (Fig. 5B).

Arginine is mainly present in the metabolic pathway of Streptococcus, and its metabolite ammonia is an important source of oral alkali production; thus, arginine has a significant impact on maintaining the stability of oral flora [35, 36]. Decreases in the levels of the related metabolites L-glutamine and N-acetyl-L-glutamate 5-semialdehyde in the arginine biosynthesis pathway in patients with AD may reflect an imbalance of oral microbes. For example, the Streptococcus ratio significantly increased in the AD group but the abundance of gram-negative bacteria Haemophilus and Leptotrichia significantly decreased (Fig. 7).

Our study has certain limitations. Subsequent sample sizes for saliva metabolomics analysis need to be increased. At the same time, more longitudinal studies are needed to explore the correlation mechanism between changes in periodontal levels and changes in salivary metabolism in patients with AD. Mechanisms related to cometabolism between oral microbes and the host in patients with AD and its influence on the development of AD should be further elucidated.

In general, we explored the changes in the periodontal status of patients with AD and screened four saliva biomarkers under matched periodontal indexes that can distinguish patients with AD from individuals without AD through a machine-learning method. In addition, we verified that dysregulation of the proportion of specific bacterial flora in saliva plays a vital role in metabolic changes in AD. We believe that these results will contribute to the further improvement of the AD saliva biomarker system and open the door to predicting the mechanism of AD occurrence and development through oral metabolism in a combined bacterial environment.

Footnotes

ACKNOWLEDGMENTS

We would like to express our gratitude to all of the study participants and their families.

FUNDING

Supported by the Natural Science Foundation of Sichuan Province (2022NSFSC0773).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the study’s conclusions are accessible from the corresponding author upon request. Due to privacy and ethical concerns, the data is not publicly available.

The supplementary material is available in the electronic version of this article: .

References

McGeer

, Lee

, Kennedy

, McGeer

(2020) Saliva diagnosis as a disease predictor. J Clin Med 9, 377.

Zhang

, Cheng

, Li

, Zhang

, Yi

, Xu

, Zhou

(2016) Saliva in the diagnosis of diseases. Int J Oral Sci 8, 133–137.

Viswanath

, Choi

, Lee

, Kim

(2017) Recent trends in the development of diagnostic tools for diabetes mellitus using patient saliva. TrAC Trends Anal Chem 89, 60–67.

Vernerová

, Kujovská Krčmov

, Melichar

, Švec

(2021) Non-invasive determination of uric acid in human saliva in the diagnosis of serious disorders. Clin Chem Lab Med 59, 797–812.

Proctor

, Shaalan

(2021) Disease-induced changes in salivary gland function and the composition of saliva. J Dent Res 100, 1201–1209.

(2023) 2023 Alzheimer’s disease facts and figures. Alzheimers Dement 19, 1598–1695.

Scheltens

, De Strooper

, Kivipelto

, Holstege

, Chételat

, Teunissen

, Cummings

, van der Flier

(2021) Alzheimer’s disease. Lancet 397, 1577–1590.

Liang

, Liu

, Zhang

, Jiang

, Xing

, Zhang

(2015) Metabolomics-based screening of salivary biomarkers for early diagnosis of Alzheimer’s disease. RSC Adv 5, 96074–96079.

François

, Karpe

, Liu

, Beale

, Hor

, Hecker

, Faunt

, Maddison

, Johns

, Doecke

, Rose

, Leifert

(2021) Salivaomics as a potential tool for predicting Alzheimer’s disease during the early stages of neurodegeneration. J Alzheimers Dis 82, 1301–1313.

10.

Kim

, Kim

, Song

, Lee

, Kim

, Chung

(2021) Metabolic phenotyping of saliva to identify possible biomarkers of periodontitis using proton nuclear magnetic resonance. J Clin Periodontol 48, 1240–1249.

11.

Martande

, Pradeep

, Singh

, Kumari

, Suke

, Raju

, Naik

, Singh

, Guruprasad

, Chatterji

(2014) Periodontal health condition in patients with Alzheimer’s disease. Am J Alzheimers Dis Other Demen 29, 498–502.

12.

Laugisch

, Johnen

, Buergin

, Eick

, Ehmke

, Duning

, Sculean

(2021) Oral and periodontal health in patients with Alzheimer’s disease and other forms of dementia— A cross-sectional pilot study. Oral Health Prev Dent 19, 255–261.

13.

, Zhang

, Huang

, Qian

, Tan

, Qian

, Zhuang

, Zou

, Li

, Yan

(2022) Periodontitis-related salivary microbiota aggravates Alzheimer’s disease via gut-brain axis crosstalk. Gut Microbes 14, 2126272.

14.

Jungbauer

, Stähli

, Zhu

, Auber Alberi

, Sculean

, Eick

(2022) Periodontal microorganisms and Alzheimer disease— a causative relationship? Periodontol 2000 89, 59–82.

15.

Teles

, Collman

, Mominkhan

, Wang

(2022) Viruses, periodontitis, and comorbidities. Periodontol 2000 89, 190–206.

16.

Want

(2018) LC-MS untargeted analysis. Methods Mol Biol 1738, 99–116.

17.

Dubois

, Feldman

, Jacova

, DeKosky

, Barberger-Gateau

, Cummings

, Delacourte

, Galasko

, Gauthier

, Jicha

, Meguro

, O’Brien

, Pasquier

, Robert

, Rossor

, Salloway

, Stern

, Visser

, Scheltens

(2007) Research criteria for the diagnosis of Alzheimer’s disease: Revising the NINCDS– ADRDA criteria. Lancet Neurol 6, 734–746.

18.

Chapple

, Mealey

, Van Dyke

, Bartold

, Dommisch

, Eickholz

, Geisinger

, Genco

, Glogauer

, Goldstein

, Griffin

, Holmstrup

, Johnson

, Kapila

, Lang

, Meyle

, Murakami

, Plemons

, Romito

, Shapira

, Tatakis

, Teughels

, Trombelli

, Walter

, Wimmer

, Xenoudi

, Yoshie

(2018) Periodontal health and gingival diseases and conditions on an intact and a reduced periodontium: Consensus report of workgroup 1 of the 2017 World Workshop on the Classification of Periodontal and Peri-Implant Diseases and Conditions. J Periodontol 89, S74–S84.

19.

Plachokova

, Andreu-Sánchez

, Noz

, Fu

, Riksen

(2021) Oral microbiome in relation to periodontitis severity and systemic inflammation. Int J Mol Sci 22, 5876.

20.

Varesi

, Pierella

, Romeo

, Piccini

, Alfano

, Bjorklund

, Oppong

, Ricevuti

, Esposito

, Chirumbolo

, Pascale

(2022) The potential role of gut microbiota in Alzheimer’s disease: From diagnosis to treatment. Nutrients 14, 668.

21.

Chen

, Liao

, Xia

, Liu

, Jones

, Haran

, McCormick

, Sampson

, Alam

, Ye

(2022) Gut microbiota regulate Alzheimer’s disease pathologies and cognitive disorders via PUFA-associated neuroinflammation. Gut 71, 2233–2252.

22.

Chen

, Xu

, Wu

, Cao

, Li

, Hou

, Wang

, Liu

, Ji

, Zhang

, Li

(2022) A comparison of the composition and functions of the oral and gut microbiotas in Alzheimer’s patients. Front Cell Infect Microbiol 12, 942460.

23.

Dommisch

, Skora

, Hirschfeld

, Olk

, Hildebrandt

, Jepsen

(2019) The guardians of the periodontium-sequential and differential expression of antimicrobial peptides during gingival inflammation. Results from in vivoand in vitro studies. J Clin Periodontol 46, 276–285.

24.

Sureda

, Daglia

, Argüelles Castilla

, Sanadgol

, Fazel Nabavi

, Khan

, Belwal

, Jeandet

, Marchese

, Pistollato

, Forbes-Hernandez

, Battino

, Berindan-Neagoe

, D’Onofrio

, Nabavi

(2020) Oral microbiota and Alzheimer’s disease: Do all roads lead to Rome? Pharmacol Res 151, 104582.

25.

Gao

, Chu

, Young

FYF

(2020) Oral health and care for elderly people with Alzheimer’s disease. Int J Environ Res Public Health 17, 5713.

26.

Plachokova

, Andreu-Sánchez

, Noz

, Fu

, Riksen

(2021) Oral microbiome in relation to periodontitis severity and systemic inflammation. Int J Mol Sci 22, 5876.

27.

Dickson

, Tenon

, Svilar

, Fança-Berthon

, Lugan

, Martin

, Vaillant

, Rogez

(2018) Main human urinary metabolites after Genipap (Genipa americana L.) juice intake. Nutrients 10, 1155.

28.

Dayrit

(2015) The properties of lauric acid and their significance in coconut oil. J Am Chem Soc 92, 1–15.

29.

Zahniser

, Chou

, Hanin

(1977) Is 2-dimethylaminoethanol (deanol) indeed a precursor of brain acetylcholine? A gas chromatographic evaluation. J Pharmacol Exp Ther 200, 545–559.

30.

Moreira

, Sale

MGF

, Di Lorenzo

(2017) Towards timely Alzheimer diagnosis: A self-powered amperometric biosensor for the neurotransmitter acetylcholine. Biosens Bioelectron 87, 607–614.

31.

Sun

, Zhang

, Li

, Yao

, Liu

, Chen

, Luo

, Wang

, Gong

, Li

, Luo

(2022) Acetylcholine deficiency disrupts extratelencephalic projection neurons in the prefrontal cortex in a mouse model of Alzheimer’s disease. Nat Commun 13, 998.

32.

Cui

, Wang

, Dong

, Wu

, Xiao

, Hu

, Chai

, Zhang

(2011) Morphine protects against intracellular amyloid toxicity by inducing estradiol release and upregulation of Hsp70. J Neurosci 31, 16227–16240.

33.

Mogilnicka

, Bogucki

, Ufnal

(2020) Microbiota and malodor-etiology and management. Int J Mol Sci 21, 2886.

34.

Chen

, Niu

, Lv

(2022) Multi-omics insights reveal the remodeling of gut mycobiome with P. gingivalis. Front Cell Infect Microbiol 12, 937725.

35.

Chakraborty

, Burne

(2017) Effects of arginine on Streptococcus mutans growth, virulence gene expression, and stress tolerance. Appl Environ Microbiol 83, e00496.

36.

Nascimento

, Alvarez

, Huang

, Browngardt

, Jenkins

, Sinhoreti

, Ribeiro

APD

, Dilbone

, Richards

, Garrett

, Burne

(2019) Metabolic profile of supragingival plaque exposed to arginine and fluoride. J Dental Res 98, 1245–1252.

37.

Carda-Diéguez

, Moazzez

, Mira

(2022) Functional changes in the oral microbiome after use of fluoride and arginine containing dentifrices: A metagenomic and metatranscriptomic study. Microbiome 10, 159.