Porphyromonas gingivalis (W83) Infection Induces Alzheimer’s Disease-Like Pathophysiology in Obese and Diabetic Mice

Abstract

Background:

Periodontal disease(s) and metabolic illnesses negatively impact the quality of life and, eventually mental health.

Objective:

This study investigated the effect of Porphyromonas gingivalis (W83) oral infection on the development of Alzheimer’s disease (AD) pathophysiology in a wild-type obese, diabetic (db/db) mouse model.

Methods:

The db/db mice were either orally infected with P. gingivalis and Fusobacterium nucleatum or sham infected for 16 weeks. The presence of amyloid-β (Aβ) and neurofibrillary tangles (NFTs) were assessed using a silver impregnation technique and subsequently by immunohistochemistry for tau and neuroinflammation. The mRNA abundance of a panel of 184 genes was performed using quantitative real-time PCR, and the differentially expressed genes were analyzed by Ingenuity Pathway Analysis.

Results:

While no Aβ plaques and NFTs were evident by silver impregnation, immunohistochemistry (glial cell markers) of the P. gingivalis-infected mice tissue sections exhibited neuroinflammation in the form of reactive microglia and astrocytes. Anti-tau immunopositivity, in addition to cells, was prominent in thickened axons of hippocampal CA neurons. The mRNA abundance of crucial genes in the insulin signaling pathway (INSR, IGF1, IRS, IDE, PIK3R, SGK1, GYS, GSK3B, AKT1) were upregulated, potentially exacerbating insulin resistance in the brain by P. gingivalis oral infection. Increased mRNA abundance of several kinases, membrane receptors, transcription factors, and pro-inflammatory mediators indicated hyperactivation of intracellular cascades with potential for tau phosphorylation and Aβ release in the same infection group.

Conclusion:

P. gingivalis W83 infection of db/db mice provides a disease co-morbidity model with the potential to reproduce AD pathophysiology with induced periodontal disease.

Keywords

Alzheimer’s disease diabetes genes insulin resistance

INTRODUCTION

Chronic periodontal disease is one of the most prevalent human chronic diseases that affect the host’s immune response and the periodontal tissues, which if left untreated, leads to eventual tooth loss [1]. Porphyromonas gingivalis is considered a keystone pathogen of the disease and is observed in gingival plaque samples of chronic periodontitis patients [2]. The encapsulated serotypes of P. gingivalis such as W83 (K1 serotype) are highly effective at causing severe disease because of their ability to stimulate the immune system by inducing several pro-inflammatory cytokines including (IL)-1β, IL-6, IL-17, interferon (IFN)-γ, and tumor necrosis factor (TNF)-α in macrophages and dendritic cells [3]. The pathogenic strains can also directly enter the systemic circulation following toothbrushing of inflamed gingivae [4], as revealed through microbial analysis of the blood samples from the infected host [5]. Lipopolysaccharide (LPS) is also reported to be present in the blood samples of patients with periodontal disease [6]. Growing evidence suggests that periodontopathogens play a role in the modulation of systemic and non-systemic pathological manifestations, especially metabolic syndrome (MetS), which is characterized by the presence of insulin resistance [7] and Alzheimer’s disease (AD) [8, 9].

AD is a neurodegenerative disease characterized by two microscopical lesions, namely amyloid-β (Aβ) plaques and intraneuronal neurofibrillary tangles (NFTs) bound to hyperphosphorylated tau protein in the fronto-temporal cerebral cortices [10]. The insoluble Aβ deposition and the development of NFT tau protein have a direct connection with the intracerebral innate immune response and complement activation [11]. Progression and development of Aβ links with microbial infections [12, 13], and the inflammatory mediator (cytokines) release that always follows the microbial entry into the host [3, 14]. While the role of P. gingivalis infection in the development of Aβ deposits is unfolding, it is unclear how the amyloid-β protein precursor processing could lead to Aβ deposition.

Numerous studies implicate the association of obesity/type-2 diabetes (T2D) with periodontal disease [15]. Active periodontal disease can potentially affect glycemic control leading to T2D (an example of MetS). T2D patients with periodontitis tend to have fewer teeth, and deeper periodontal pockets containing P. gingivalis [16]. The ability of P. gingivalis to cause periodontal tissue damage and initiate periodontal pockets is due predominantly to its two main virulence factors, which are gingipains and LPS [17]. LPS mediates the secretion of pro-inflammatory cytokines (IL-1β, IL-6, and IL-8) contributing to hyperglycemia and insulin resistance [18]. Elevated levels of lipid A within the LPS macromolecule can inhibit neutrophil apoptosis and prevent macrophage phagocytosis of apoptotic cells thereby providing a safe sanctuary for P. gingivalis survival within infected neutrophils [19]. P. gingivalis LPS can potentially contribute to the high circulatory level of LPS and is commonly found in individuals suffering from obesity and T2D [7]. Although LPS in blood plasma in MetS is predominantly of gastrointestinal origin [7], the contribution of LPS from P. gingivalis in the circulating blood in MetS is likely, but its overall impact remains unclear.

Growing evidence suggests that periodontitis and MetS together exacerbate the onset of AD [20, 21]. Currently, longitudinal studies on the intricate association between periodontal disease, MetS, and AD are scarce. The obesity/type-2 diabetes (db/db) mouse is a relevant laboratory model for MetS as they develop identifiable obesity around 3-4 weeks of age with an elevated plasma insulin beginning at 10–14 days. Thus, db/db mice provide a valuable co-morbidity model that shows clinical signs of periodontitis and MetS for investigating AD. We initiated the study to test our hypothesis that chronic induction of P. gingivalis via the oral route in db/db mice would stimulate the formation of the pathological lesions in the presence of simulated predisposing conditions that co-exist during AD onset and progression [22 –25]. Our main objective in this study was to assess the effect of P. gingivalis W83 mediated oral infection on Aβ plaque deposition and NFT formation and to identify the specific molecular pathways that serve as common molecular drivers during inflammation and insulin resistance in vivo in db/db mice.

MATERIALS AND METHODS

Mice and infection regime

Obese and type-2 diabetic mice (db/db mice) aged five weeks were purchased from Jackson Laboratories, Bar Harbor, MA, USA. Mice at six weeks were randomly assigned to any of the three groups, and periodontal disease was induced by orally infecting with 10⁹ colony-forming unit (CFU/ml) P. gingivalis W83 (n = 9), and Fusobacterium nucleatum (ATCC 49256) (n = 3) and uninfected (sham mice n = 6 as carrier control). The sham mice (sham-infection and/or control group) received the carboxymethyl cellulose (vehicle control) only and the F. nucleatum bacterial infection group (an irrelevant bacterium for AD infection was another control group) for the same experimental duration. Mice were orally infected with the capsulated P. gingivalis strain W83 (K1 serotype) and F. nucleatum and sham, every other week, three times a week for 16 weeks to maintain a continuum of the original infection. At the end of the experimental period, mice were euthanized, and tissues (whole brain, for this study) collected for further analysis. One cerebral hemisphere was fixed in 10%neutral buffered formalin for microscopy and immunohistochemistry and the other cerebral hemisphere was placed into RNA-later and stored at –80°C for gene expression analysis. Brain samples from AD transgenic Tg2576 mice with the Swedish mutation (n = 3) were obtained from Prof. Roxane O. Carare, Faculty of Medicine, University of Southampton, UK, as secondary users of tissue, to act as positive controls for Aβ insoluble plaques and for immunohistochemistry for resident central nervous system inflammatory cell responses.

All experimental procedures were performed in accordance with the guidelines of the University of Florida Institutional Animal Care and Use Committee (IACUC, protocol #201710038) and the University of Central Lancashire, UK (AWERB #RE1718 and RE19/02 (db/db mice) and RE/17/06 for Tg2576 mice brains).

Light microscopy

The formalin-fixed brain specimens were processed in an automated tissue processor (Shandon Citadel 2000, Thermo Scientific) and embedded in paraffin wax as per routine methods. The tissue blocks with temporal lobe inclusive of the hippocampus were sectioned (5 mm in thickness) using the Leica RM2235 microtome and collected on Superfrost+^® glass microscope slides (Leica UK). Rehydrated paraffin wax sections from the hippocampus of the AD transgenic Tg2576 mice (positive controls) and db/db mice (test specimens) were screened for argyrophilic Aβ_1–_40/42 insoluble plaques using the modified Methenamine silver stain described previously [26].

Immunohistochemistry

Following dewaxing and alcohol dehydration steps, the brain tissue sections from the Tg2576 and the db/db mice were treated with 0.003%H₂O₂ in methanol for 20 min at room temperature to quench endogenous peroxidase activity. The sections were washed in running tap water. For optimal antigen-binding, an antigen retrieval step was included.

Antigen retrieval

Rehydrated paraffin wax sections were optimized for antigen-antibody binding for the calcium-binding protein anti-Iba1 (Invitrogen^®, cat no 13269248), which is specifically expressed by microglia, and the glial fibrillary acidic protein (GFAP) expressed by astrocytes (ab7260, AbCam, UK) using 10 mM citric acid buffer (pH 6.1) for 15 min at 750W power. Following thorough washings, sections were equilibrated, for 5 min, in PBS (0.1 M PBS, pH 7.2), and the non-specific antibody binding was controlled by incubating tissue sections for 30 min in blocking solution containing 0.1%normal horse serum (Vectastain kit, PK 4002) in PBS. AD transgenic Tg2576 mice tissue sections acted as positive controls for neuroinflammation. The negative antibody control sections were incubated in the block solution, whereby the primary antibody was omitted. All other tissue sections were incubated in a humidity chamber, overnight, at 4°C in the primary antibodies (rabbit anti-GFAP, ab7260, 1/2000) and (rabbit anti-Iba1, 1/150) diluted in blocking solution. The next day, the sections were washed in PBS (x3 for 5 min each) and incubated for 60 min at room temperature in the biotin labelled anti-rabbit IgG (from kit PK-4001) diluted 1/200 in block solution. From here, the rabbit peroxidase kit (Vector laboratories, Perterborough, UK) and the DAB kit (SK-4100) were used as per manufacturers’ instructions. The sections were lightly counterstained in Mayer’s haemalum (LAMB/170D) before dehydrating them in graded alcohols and clearing in xylene. The sections were mounted under a coverslip and examined using the Nikon Eclipse E200 Microscope. Images were captured using the DS-L2 v.441 Software (Nikon, UK). The immunohistochemistry procedure was similar to above for mouse anti-tau (clone AT8, Invitrogen ^® ) except antigen retrieval was omitted. The block solution contained PBS-0.1%triton X-100, horse serum 0.01%) and the mouse anti-tau antibody was diluted 1/20. The secondary detection antibody used was the biotin labelled anti-mouse IgG (from kit PK-4002) diluted 1/200 in block solution.

Gene expression analysis

Total RNA extraction

Total RNA was extracted from ∼25 mg brain tissue (fronto-temporal) following a two-step procedure. In step 1, the tissue sample was ground in 1 ml TRIzol^TM reagent (ThermoFisher Scientific Inc.), and total RNA was obtained following the manufacturer’s instructions. In step 2, the purification of the total RNA obtained from step 1 was performed following the GenElute mammalian total RNA Miniprep Kit (Sigma-Aldrich Corp.) according to the manufacturer’s instructions. Total RNA was subjected to DNAse I (Sigma-Aldrich Corp.) treatment to eliminate genomic DNA contamination. Column purification of the RNA was performed using GenElute mammalian total RNA kit (Sigma-Aldrich Corp.). Total RNA was finally suspended in 50μl 0.1%diethylpyrocarbonate (DEPC) treated water and stored at –80°C. The quality and quantity of the total RNA were assessed in a NanoDrop-ND1000 Spectrophotometer (Thermo Fisher Scientific Inc). The cDNA synthesis was performed with 1μg of total RNA using the RevertAid H minus first-strand cDNA synthesis kit (Fermentas GmbH, St. Leon-Rot, Germany) following the manufacturer’s protocol. The final volume of cDNA was adjusted to 150μl with nuclease-free water.

Quantitative real-time PCR (qPCR) array

The expressions of a panel of 192 genes were evaluated using a custom-designed qPCR array (Thermo Fisher Scientific Inc), performed in a QuantStudio 5 real-time PCR system (Applied Biosystems). The qPCR arrays (96-well format x 2) enabled mRNA screening of a total of 184 target genes and eight internal controls. The target genes screened included the common markers of the inflammatory immune cascade, oxidative stress, glucose metabolism, and insulin pathways. This panel of genes was previously used to capture the intracellular signaling pathways associated with the chronic inflammation induced by P. gingivalis LPS treatment [27]. For the qPCR array run, 25μl cDNA (after 1:5 dilution) from each replicate within a single treatment group were pooled. The qPCR was performed on a 20μl reaction mixture per well containing 1μl pooled cDNA, 9μl water, and 10μl SYBR green master mix (Applied Biosystems). The thermal cycling conditions were 94°C for 30 s, followed by 60°C for 1 min, for 40 cycles. In this experiment, a CT value of 35 was considered as the cut-off limit. The relative quantities of the target genes were normalized using the geometric mean of the relative quantities of three internal control genes (beta-actin (ACTB), hypoxanthine-guanine phosphoribosyl transferase 1 (HPRT1), and beta-2 microglobulin (B2M). Briefly, average ΔCt was calculated as the difference of Ct values of any target gene minus the geometric average of the Ct value of the three reference genes. Then, fold change was calculated as 2^{(–average ΔCt target gene)}/2^{(–average ΔCt reference gene)}. The differentially expressed genes between any two treatments were identified from the ratio of fold change values with a cut-off±2.0 fold.

Gene ontology and pathway analysis

Gene ontology (GO) of the differentially expressed genes was performed using the database for annotation, visualization and integrated discovery (DAVID), a web-based online (http://david.abcc.ncifcrf.gov/) bioinformatics tool [28]. The major annotation clusters of GO term were identified based on the enrichment score calculated by the DAVID tool’s default setting. For pathway analysis, the fold change values were analyzed using Qiagen Ingenuity Pathway Analysis (IPA) and the relevant canonical pathways were identified using the default setting in the software. The statistical probability (p-value) of the observed number of genes affecting a particular biological function was calculated based on Fisher’s Exact Test. This p-value indicated the statistical probability of the observed number of genes affected out of the total number of genes evaluated in the PCR array for the biological function. The correlation between the relationship direction and gene expression was determined by calculating the Z-score following the formula Z = (N⁺–N^–)/√N where N⁺ represents the number of genes whose expression follows the same direction while N–represents the number of genes whose expression follows an opposite direction of the expression of a particular gene compared to that already available in the IPA knowledge database. N indicates the total number of genes affected. A high stringency Z-score between≥+2.0 or≤–2.0 was applied to identify the most relevant signaling pathways.

RESULTS

Brain histology

Hematoxylin and Eosin morphology staining of the formalin fixed, paraffin wax embedded, and rehydrated temporal lobe tissue sections from Tg2576, P. gingivalis W83, F. nucleatum infected, and sham mice showed they were well preserved (not shown). Silver impregnation of the tissue sections from the Tg2576 mouse brains demonstrated numerous Aβ plaques, which were of variable sizes, randomly distributed, within the fronto-temporal cortices including the hippocampus (Fig. 1a). The argyrophilic Aβ plaques and NFTs were absent in the hippocampus or in the fronto-temporal cortices of db/db brain tissue sections, from the P. gingivalis W83, F. nucleatum infected, and sham mice groups (Fig. 1b-d).

Fig. 1

Rehydrated paraffin wax embedded brain tissue sections from the hippocampus of the AD transgenic Tg2576 (ADtg) mice and obese diabetic mice stained with methenamine silver to demonstrate insoluble amyloid-β (Aβ) plaques. a) AD transgenic mouse (positive control for Aβ plaques). Methenamine silver shows numerous argyrophilic plaques of variable size in the hippocampus in close proximity to the CA neurons. Some blood vessels stand out following silver impregnation. b) Sham infected (sham) obese diabetic mice. c) F. nucleatum (FN). d) P. gingivalis (Pg W83). Images b-d) from obese diabetic mice completely lacked Aβ plaques.

Immunohistochemistry

Neuroinflammation

Microgliosis. The negative controls whereby the primary antibody was omitted remained negative (Fig. 2a-d). In theTg2576 mice (Fig. 2e, arrows), and the sham (Fig. 2f, arrows) and the F. nucleatum infected (Fig. 2g, arrows) activated microglia were below the threshold of the numbers (not counted) observed in the P. gingivalis W83 infected mice brain tissue sections (Fig. 2h, i, arrows).

Fig. 2

Rehydrated paraffin wax tissue sections and order of brains as for Fig. 1. This image shows immunostaining with rabbit anti-Iba-1 antibody to detect activated microglia. Images a-d) show little staining as a negative control when the primary antibody was omitted in all mice categories. Images e-i) show variable degree of activated microglia scattered within the hippocampus as denoted by arrows. The sham-infected (f) and the F. nucleatum infected (g) mice, showed only a few resting microglia. The P. gingivalis W83 mice demonstrated considerably more reactive microglial cell distribution with branched processes within the hippocampus (h, i, arrows).

Astrogliosis. Anti-GFAP immunostaining of the AD transgenic Tg2576 mice brain sections demonstrated the presence of abundant, reactive, astrocytes in close vicinity to the Aβ plaques (Fig. 3a box, b, where p represents the Aβ_1–42 plaque core and is the area within the box from Fig. 3a). The sham-infected mice brain tissue sections (Fig. 3c box, d area within the box in 3c) demonstrated resting astrocytes. F. nucleatum infected mouse hippocampus demonstrated significant numbers (not counted) of activated, astrocytes (Fig. 3e box, f area within the box in 3e). The P. gingivalis W83 infected mice brain tissue sections (Fig. 3g area within the smaller box is represented in h and i is the area from the larger box in 3g) also demonstrated an abundance of reactive astrocytes almost equivalent (not counted) to those observed in F. nucleatum infected mice tissue sections (Fig. 3e, f).

Fig. 3

Rehydrated paraffin wax tissue sections and order of brains as for Fig. 1. Immunolabelling of the GFAP. The AD transgenic mice brain sections (a, b) demonstrate numerous reactive astrocytes in close vicinity to the amyloid plaques (a box, and b where p represents the amyloid plaque core and is the area within the box from a). The sham brain sections (c box, d area within the box in c) demonstrated resting astrocytes. F. nucleatum infected mice hippocampus (e, f) demonstrated significant numbers of activated astrocytes (e box, f area within the box in e). The P. gingivalis W83 brain sections, (g box, h and i are areas from box in g) also demonstrated abundant reactive astrocytes.

Mouse anti-tau (AT8) immunostaining for detecting NFTs

As expected, the AD transgenic Tg2576 mice exhibited an intense immunopositive staining for tau within numerous CA neuronal cell bodies. The axon hillock of some neurons was also immunostained (Fig. 4a, arrows). Overall, the tau staining in the sham (non-infected) group of mice was weak (Fig. 4b and arrow). In the F. nucleatum mice, a weak tau immunostaining was observed within the CA neurons without evidence of staining within the axon hillocks (Fig. 4c). The P. gingivalis W83 infected mice demonstrated tau immunostaining within the CA neurons (Fig. 4d) and equally stronger tau positivity within thickened bundles of axonal hillocks in the hippocampal CA neurons (Fig. 4e boxed area shown in f, arrows point to thickened axonal hillocks).

Fig. 4

Rehydrated paraffin wax tissue sections and order of brains as for Fig. 1. This image shows immunostaining with mouse anti-tau (AT8) antibody to detect neurofibrillary tangles. a) The AD transgenic mice exhibited an intense immunopositive staining within numerous CA neuronal cell bodies within the hippocampus. The axon hillock of some neurons was also immunostained (a and arrows). b) Sham infected mice demonstrated some inconspicuous stating in the axon hillock (b and arrow). c) In the F. nucleatum mice, a weak, largely diffuse tau immunostaining was observed within the CA neurons without evidence of staining within the axon hillock (c). d) The images d to f) are tissue sections from P. gingivalis W83 infected mice demonstrating tau immunostaining within the CA neurons and more stronger tau positive and thickened bundles of axon hillock(s) in the hippocampus (d, e box, f arrows).

Gene expression in brains using qPCR array

Of the total 192 genes evaluated on the qPCR array, the mRNA of 184 genes were detectable. The mRNA abundance of a total of 96 genes in the cerebral hemisphere was differentially expressed due to the P. gingivalis W83 induction of obese diabetic mice (Table 1). The primary group of genes upregulated (by > 2 fold) by P. gingivalis W83 infection within the extracellular space of the cells includes the growth factor IGF1 and several cytokines (IL1A, IL1B, CSF1, LTA, and CXCL10). Other genes upregulated in the extracellular space include IGFBP3, CRP, and RETN. In the plasma membrane, genes upregulated include transmembrane receptor (IGF1R, TLR4, TLR6, ICAM1, VCAM1, IL6R, IL1R2, CD28, CD80), G-protein coupled receptor (ADORA1, CCR2, GLP1R, GLP2R), and membrane transporters (LDLR). Besides, the mRNA abundance of two crucial kinases (INSR, KDR) and plasma membrane-associated enzymes (ACE, ADA, DPP4, PTPRC) and other molecules (DOK2, CLTC, MYD88, CAP1) were also upregulated by > 2 fold. As expected, within the cytoplasm, the major group of molecules affected were kinases (AKT1, PIK3R1, PIK3R2, PRKCA, PRKCB, PRKCD, JAK2, etc.) and enzymes (Table 1). The mRNA abundance of 12 transcription factors (FOXO1, NFKB2, NFATC2, CTNNB1, FOS, SREBF1, SREBF2, TRP53, STAT3, JUN, SMAD7, and CBL) were also upregulated whilst two transcription factors (CREB1 and STAT6) were downregulated. Other genes differentially expressed within the nucleus included ligand-dependent nuclear receptor (NR4A2), nuclear kinase (GSK3B), enzymes (APC, AEBP1), and other molecules (CDC27).

Table 1

Differential expression of genes caused by the P. gingivalis W83 (PG) induction of the db/db mice versus the F. nucleatum (FN) or sham infection in the cerebral hemisphere region of the brain (fold change values≥2.0 and ≤–2.0 folds in the P. gingivalis group and in the corresponding groups

Gene	Location	Type of molecule	PG versus FN	PG/sham	FN/sham
Insulin like growth factor 1 (IGF1)	Extracellular space	Growth factor	+3.47	+2.25	–1.54
C-X-C motif chemokine ligand 10 (CXCL10)		Cytokine	+2.30	+1.78	–1.29
Interleukin 1 alpha (IL1A)		Cytokine	+2.22	+1.90	–1.17
Lymphotoxin alpha (LTA)		Cytokine	+2.17	+2.69	+1.24
Colony stimulating factor 1 (CSF1)		Cytokine	+2.13	+2.06	–1.03
Interleukin 1 beta (IL1B)		Cytokine	+2.07	+2.63	+1.27
Matrix metallopeptidase 2 (MMP2)		Peptidase	+3.35	+3.62	+1.08
Insulin degrading enzyme (IDE)		Peptidase	+2.19	+1.77	–1.24
Resistin (RETN)		Other	+6.11	+1.53	–3.99
Insulin like growth factor binding protein 3 (IGFBP3)		Other	+3.42	+3.42	+1.00
C-reactive protein (CRP)		Other	+2.02	+2.56	+1.27
Toll like receptor 4 (TLR4)	Plasma membrane	Transmembrane receptor	+7.43	+7.21	–1.03
Interleukin 6 receptor (IL6R)		Transmembrane receptor	+6.05	+3.42	–1.77
Intercellular adhesion molecule 1 (ICAM1)		Transmembrane receptor	+5.32	+3.43	–1.55
interleukin 1 receptor type 2 (IL1R2)		Transmembrane receptor	+4.73	+2.27	–2.08
CD 80 molecule (CD80)		Transmembrane receptor	+2.78	+2.01	–1.39
CD 28 molecule (CD28)		Transmembrane receptor	+2.73	+1.93	–1.41
Vascular cell adhesion molecule 1 (VCAM1)		Transmembrane receptor	+2.15	+1.79	–1.20
Toll like receptor 6 (TLR6)		Transmembrane receptor	+2.15	+2.13	–1.01
Insulin like growth factor 1 receptor (IGF1R)		Transmembrane receptor	–3.01	–4.31	–1.43
Glucagon like peptide 1 receptor (GLP1R)		G-protein coupled receptor	+5.96	+2.07	–2.88
Glucagon like peptide 2 receptor (GLP2R)		G-protein coupled receptor	+4.61	+2.32	–1.99
C-C motif chemokine receptor 2 (CCR2)		G-protein coupled receptor	+3.67	+2.10	–1.74
Adenosine Receptor A1 (ADORA1)		G-protein coupled receptor	+2.03	+4.12	+2.03
Low density lipoprotein receptor (LDLR)		Transporter	+3.50	+3.48	–1.01
LDL receptor related protein 2 (LPR2)		Transporter	–3.08	–2.07	+1.49
Insulin receptor (INSR)		Kinase	+3.35	+1.33	–2.52
Kinase insert domain receptor (KDR)		Kinase	+2.89	+1.96	–1.47
Fibroblast growth factor receptor 4 (FGFR4)		Kinase	+2.58	+1.92	–1.35
Epidermal growth factor receptor (EGFR)		Kinase	+2.58	+1.26	–2.06
Protein tyrosine phosphatase receptor type C (PTPRC)		Phosphatase	+2.74	+1.63	–1.68
Angiostensin I converting enzyme (ACE)		Peptidase	+5.36	+3.47	–1.54
Dipeptidyl peptidase 4 (DPP4)		Peptidase	+2.21	+1.65	–1.34
Adenosine deaminase (ADA)		Enzyme	+2.88	+2.24	–1.28
Docking Protein 2 (DOK2)		Other	+4.90	+3.18	–1.54
Clathrin heavy chain (CLTC)		Other	+3.33	+2.76	–1.20
MTD88 innate immune signal transduction adaptor (MYD88)		Other	+2.08	+3.43	+1.65
Cyclase associated actin cytoskeleton regulatory protein 1 (CAP1)		Other	+2.05	+2.84	+1.38
Glycogen synthase 2 (GYS2)	Cytoplasm	Enzyme	+4.30	+6.42	+1.49
Hydroxysteroid 17-beta dehydrogenase 4		Enzyme	+2.42	+4.70	+1.94
KRAS proto-oncogene, GTPase (KRAS)		Enzyme	+2.30	+2.44	+1.06
Insulin receptor substrate 1 (IRS1)		Enzyme	+2.20	+1.93	–1.14
Nitric oxide synthase 2 (NOS2)		Enzyme	–4.64	–4.25	+1.09
Mitogen-activated protein kinase 7 (MAPK7)		Kinase	+3.75	+1.58	–2.37
Phosphoinositide-3-kinase regulatory subunit 2 (PIK3R2)		Kinase	+3.63	+2.23	+1.62
Component of inhibitor of nuclear factor kappa B kinase complex (CHUK)		Kinase	+3.38	+2.15	–1.57
ATP serine/threonine kinase 1 (AKT1)		Kinase	+2.87	+3.79	+1.32
Protein kinase cAMP-dependent type I regulatory subunit 1A (PRKAR1A)		Kinase	+2.74	+3.05	+1.11
Mitogen-activated protein kinase kinase kinase 1 (MAP3K1)		Kinase	+2.69	+1.65	–1.63
Protein kinase C beta (PRKCB)		Kinase	+2.58	+2.43	–1.06
Phosphoenolypyruvate carboxykinase 2, mitochondrial (PCK2)		Kinase	+2.56	+4.25	+1.66
Protein kinase cAMP-activated catalytic subunit alpha (PRCAKA)		Kinase	+2.55	+4.83	+1.90
Adenylate Kinase 2 (AK2)		Kinase	+2.45	+3.93	+1.61
Protein kinase C beta (PRKCA)		Kinase	+2.41	+2.32	–1.04
Raf-1 proto-oncogene, serine/threonine kinase (RAF1)		Kinase	+2.39	+2.63	+1.10
Phosphoinositide-3-kinase regulatory subunit 1 (PIK3R1)		Kinase	+2.37	+2.80	+1.18
Serum/glucocorticoid regulated kinase 1 (SGK1)		Kinase	+2.33	+1.68	–1.39
Interleukin 1 receptor associated kinase 4 (IRAK4)		Kinase	+2.25	+1.22	–1.85
Protein kinase C delta (PRKCD)		Kinase	+2.20	+2.23	+1.01
Serum/glucocorticoid regulated kinase 3 (SGK3)		Kinase	+2.13	+1.96	–1.09
Janus kinase 2 (JAK2)		Kinase	+2.13	+1.32	–1.62
Inhibitor of nuclear factor kappa B kinase subunit beta (IKBKB)		Kinase	+2.01	+2.66	+1.32
Caspase 7 (CASP7)		Peptidase	+2.53	+4.11	+1.62
Protein phosphatase 2 scaffold subunit A alpha (PPP2R1A)		Phosphatase	+2.08	+4.45	+2.14
Glucose 6 phosphatase catalytic subunit (G6PC)		Phosphatase	–2.53	–1.07	+2.37
Adenosine A2a receptor (ADORA2A)		G-protein coupled receptor	+2.34	+2.35	+1.01
Adenosine A2a receptor (ADORA2B)		G-protein coupled receptor	+2.12	+2.15	+1.01
Adaptor related protein complex 1 subunit sigma 2 (AP1S2)		Transporter	+3.90	+1.91	–2.04
Adaptor related protein complex 1 subunit sigma 1 (AP1S1)		Transporter	+2.29	+4.07	+1.78
Uncoupling protein 2 (UCP2)		Transporter	+2.18	+1.30	–1.68
Growth factor receptor bound protein 10 (GRB10)		Other	+4.05	+1.36	–2.97
Tubulin beta class I (TUBB)		Other	+2.68	+3.14	+1.17
GRB2 associated binding protein 1 (GAB1)		Other	+2.40	+2.08	–1.16
SOS Ras/Rho guanine nucleotide exchange factor 2 (SOS2)		Other	+2.27	+1.85	–1.23
Growth factor receptor bound protein 2 (GRB2)		Other	+2.14	+3.81	+1.78
CRK proto-oncogene, adaptor protein (CRK)		Other	+2.01	+2.51	+1.25
Suppressor of cytokine signaling 1 (SOCS1)		Other	–2.21	+1.49	+3.30
Forkhead box O1 (FOXO1)	Nucleus	Transcription regulator	+4.00	+3.35	–1.20
Nuclear factor kappa B subunit 2 (NFKB2)		Transcription regulator	+4.68	+3.77	–1.24
Nuclear factor of activated T cells 2 (NFATC2)		Transcription regulator	+2.82	+3.25	+1.15
Catenin beta 1 (CTNNB1)		Transcription regulator	+2.71	+1.91	–1.42
Fos proto-oncogene, AP-1 transcription factor subunit (FOS)		Transcription regulator	+2.71	+2.10	–1.29
Sterol regulatory element binding transcription factor 1 (SREBF1)		Transcription regulator	+2.69	+2.93	+1.09
Tumor protein p53 (TP53)		Transcription regulator	+2.62	+3.13	+1.20
Sterol regulatory element binding transcription factor 2 (SREBF2)		Transcription regulator	+2.31	+2.89	+1.25
Signal transducer and activator of transcription 3 (STAT3)		Transcription regulator	+2.31	+3.15	+1.36
Jun proto-oncogene, AP-1 transcription factor subunit (JUN)		Transcription regulator	+2.21	+3.35	+1.52
SMAD family member 7 (SMAD7)		Transcription regulator	+2.10	+1.74	–1.21
Cbl proto-ocogene (CBL)		Transcription regulator	+3.04	+1.59	–1.91
Signal transducer and activator of transcription 6 (STAT6)		Transcription regulator	–3.41	–2.23	+1.53
cAMP responsive element binding protein 1 (CREB1)		Transcription regulator	–4.76	–2.23	+2.13
Nuclear receptor subfamily 4 group A member 2 (NR4A2)		Ligand-dependent nuclear receptor	+2.45	+3.55	+1.45
Glycogen synthase kinase 3 beta (GSK3B)		Kinase	+4.78	+2.42	–1.97
APC regulator of WNT signaling pathway (APC)		Enzyme	+2.73	+1.26	–2.16
AE binding protein 1 (AEBP1)		Peptidase	+4.00	+2.29	–1.75
Cell division cycle 27 (CDC27)		Other	+2.34	+1.71	–1.37

Annotation clustering and gene ontology

An evaluation of the panel of differentially expressed genes revealed the major annotation clusters and gene ontologies associated with the molecular, biological and cellular processes altered in the cerebral hemisphere of the db/db mice induced with P. gingivalis W83. A total of 13 annotation clusters with an enrichment score of≥2.0 were scored, of which four most relevant clusters with the highest scores are presented in Table 2. Cluster 1 (enrichment score 7.97) includes the well-represented gene ontology (GO) terms relating to protein phosphorylation, kinase activity, transferase activity, intracellular signal transduction, ATP binding and nucleotide bindings indicating that the genes differentially expressed by the P. gingivalis W83 induction are likely to affect these intracellular processes (Fig. 5a). Cluster 2 (enrichment score 4.06) relates to the inflammatory immune response (Fig. 5b). Two other highly enriched annotation clusters (Cluster 3 and 4) include GO terms relating to insulin and insulin receptor signaling pathways (Fig. 5c, d).

Table 2

The major gene annotation clusters of gene ontology (GO) affected by the genes differentially expressed in the mice brain due to the P. gingivalis W83 induction compared to those infected with F. nucleatum

Annotation	Enrichment score	GO term	Gene count	%	p
Cluster 1	7.970	GO:0006468∼protein phosphorylation	25	26.32	1.3E-15
		GO:0016301∼kinase activity	24	25.26	8.2E-13
		GO:0016310∼phosphorylation	22	23.16	4.4E-12
		GO:0004672∼protein kinase activity	20	21.05	4.8E-11
		GO:0035556∼intracellular signal transduction	14	14.74	1.5E-07
		GO:0018105∼peptidyl-serine phosphorylation	9	9.47	4.5E-07
		GO:0046777∼protein autophosphorylation	10	10.53	4.6E-07
		GO:0004674∼protein serine/threonine kinase activity	14	14.74	4.8E-07
		GO:0005524∼ATP binding	23	24.21	1.2E-05
		GO:0000166∼nucleotide binding	26	27.37	2.3E-05
		GO:0016740∼transferase activity	22	23.16	2.9E-05
Cluster 2	4.060	GO:0006954∼inflammatory response	13	13.68	2.2E-07
		GO:0006955∼immune response	9	9.47	8.6E-05
		GO:0005125∼cytokine activity	5	5.26	2.9E-02
Cluster 3	8.640	GO:0008286∼insulin receptor signaling pathway	9	9.47	2.1E-10
		GO:0048009∼insulin-like growth factor receptor signaling pathway	6	6.32	2.6E-09
		GO:0032869∼cellular response to insulin stimulus	9	9.47	1.8E-08
Cluster 4	4.300	GO:0048009∼insulin-like growth factor receptor signaling pathway	6	6.32	2.6E-09
		GO:0005159∼insulin-like growth factor receptor binding	5	5.26	5.4E-07
		GO:0005158∼insulin receptor binding	6	6.32	1.2E-06
		GO:0046326∼positive regulation of glucose import	5	5.26	2.5E-05
		GO:0045725∼positive regulation of glycogen biosynthetic process	4	4.21	5.9E-05
		GO:0014065∼phosphatidylinositol 3-kinase signaling	4	4.21	2.9E-04
		GO:0043491∼protein kinase B signaling	4	4.21	1.0E-03

Fig. 5

Gene annotation clusters of gene ontology (GO) term relating to kinases (a) and inflammatory immune response (b), insulin/IGF-1 signaling (c), and insulin/IGF-1 receptors (d) and the candidate gene/s associated with the GO term. The GO analysis was performed based on the 96 differentially expressed genes caused by the P. gingivalis W83 induction in the db/db mice compared to those treated with the Fusobacterium nucleatum (control). Two major clusters (Cluster 1 and 2) relating to kinases and inflammatory immune response and two clusters (cluster 3 and 4) relating to insulin/IGF-1signaling pathways are shown.

Intracellular pathways affected in P. gingivalis infected db/db mice

The list of most relevant intracellular signaling pathways and the number of gene affected within each of these pathways due to P. gingivalis W83 induction of mice versus F. nucleatum and sham control groups are presented in Table 3. The pathway analysis revealed that classical signaling associated with inflammatory immune response (NF-kB signaling, neuroinflammatory signaling, toll-like receptor signaling, LPS mediated MAPK signaling, IL1 and IL6) upregulated in the P. gingivalis W83 infected mice brains. Besides, pathways for cell growth, survival, and proliferation (IGF-1 signaling, PI3K/AKT, PTEN, and mTOR signaling pathways) and cellular metabolism and insulin signaling (PPAR, insulin receptor signaling, insulin secretion pathways, and T2D signaling) were affected in the P. gingivalis W83 induced mice. As expected, these pathways were mostly unaffected when assessed in sham and the F. nucleatum control mice. These observations confirm the specificity of the response generated by P. gingivalis W83 infection specific to these db/db mice in vivo.

Table 3

Canonical pathways affected by the P. gingivalis W83 (PG) infection of db/db mice versus either the F. nucleatum (FN) or sham infected brains

Signaling pathways	Genes	PG/FN				PG/Sham				Sham/FN
		Affected genes	Direction	Z-score	p	Affected genes	Direction	Z-score	p	Affected genes	Direction	Z-score	p
Inflammatory Cascade
NF-kB	179	25	↑	+4.600	3.0e–32	23	↑	+4.379	3.7e–30	7	↓	–1.890	4.1e–10
Neuroinflammation	315	27	↑	+3.530	6.2e–29	23	↑	+2.711	2.2e–24	6	↓	–0.816	6.1e–07
LPS stimulated MAPK	83	14	↑	+3.207	1.5e–19	15	↑	+3.350	3.3e–22	0	↔	–	–
PI3K/AKT	189	20	↑	+1.500	2.4e–23	23	↑	+2.065	1.4e–29	3	↔	–	–
Toll-like receptor	77	12	↑	+3.000	1.8e–16	10	↑	+2.646	1.3e–13	3	↔	–	–
Interleukin 1	95	11	↑	+3.317	1.1e–13	9	↑	+3.000	4.2e–11	2	↔	–	–
Interleukin 6	126	20	↑	+4.472	5.0e–27	19	↑	+4.359	2.4e–26	3	↔	–	–
mTOR	216	10	↑	+2.530	1.3e–08	10	↑	+2.530	3.9E–09	2	↔	–	–
Diabetes
Insulin-like growth factor1	106	20	↑	+3.771	1.2e–28	19	↑	+3.500	7.0e–28	3	↔	–	–
Insulin receptor	145	20	↑	+2.683	9.8e–26	16	↑	+2.000	5.1e–20	5	↓	–1.342	3.2e–07
Insulin secretion	267	13	↑	+2.496	3.9e–11	12	↑	+2.309	1.4e–10	5	↑	+0.447	6.5e–06
Type-2 diabetes	163	13	↑	+2.714	7.4e–14	11	↑	+2.530	1.1e–11	4	↔	–	–
Other
PPAR	105	13	↓	–3.600	2.2e–16	15	↓	–3.873	1.4e–20	3	↔	–	–
PTEN	138	18	↓	–2.840	9.1e–23	16	↓	–2.673	2.2e–20	4	↔	–	–

DISCUSSION

This study aimed to investigate the effect of Porphyromonas gingivalis (W83) oral infection on the development of AD pathophysiology in an obese, diabetic (db/db) mouse model that had been maintained under chronic infection (16 weeks) for its specific effect on the host. The Aβ insoluble plaque hallmark of AD in a wild-type young mouse model of periodontitis following an oral infection with the highly virulent strain of P. gingivalis W83 (capsular K1 serotype) has been previously reported [29]. This was the first report which demonstrated classical human AD look-alike Aβ plaque deposits in the hippocampus of a mouse model of periodontal disease, which would be expected to be demonstrated by neutral histology stains such as silver impregnation used in the present study.

Encouraged by the Ilievski et al. [29] data, and based on the knowledge that periodontal disease, cardiovascular disease, and diabetes (MetS) represent a greater risk of dementia for individuals who have suffered multiple co-morbidities in their lives [8 , 30], we used a co-morbid disease db/db mouse model to reproduce the cardinal hallmark lesions of AD, following an oral infection with the same strain of P. gingivalis (W83) [29]. However, in db/db mice, the classical AD Aβ plaques and NFT formation were not observed in the hippocampus or any other anatomical regions of the brain following silver impregnation of tissue sections. Regarding anti-tau immunostaining, our observations are similar to the Díaz-Zúñiga et al. [14] study in which oral infections of rats with several capsular P. gingivalis strains of variable serotypes and virulence showed the absence of the classical AD Aβ plaques shown by Ilievski et al. [29]. An underlying reason behind the lack of Aβ deposition in the present study could be that the mice were obese and of diabetic disposition and were not of the healthy wild-type phenotype. A high level of circulating insulin is a hallmark of insulin resistance and could prevent Aβ and NFT accumulation [31]. Insulin can potentially increase the activities of anti-amyloidogenic proteins such as insulin-degrading enzyme (IDE) and inhibit the phosphorylation of tau by glycogen synthase kinase-3β (GSK3B) activity [31, 32]. This led us to change the course of our investigation into gene array methodology for early clues to AD pathophysiological changes in the oral infection mediated periodontal model for AD.

However, before moving on, Díaz-Zúñiga et al. [14] observed thickened axons in the hippocampal neurons of P. gingivalis (capsular forms) infected rat brains which they suggested were equivalent to the human AD NFT lesion [14]. In the present study, following anti-tau immunostaining we also observed thickened axon hillocks of cortical hippocampal neurons reported by Díaz-Zúñiga et al. [14], which contrasts with the explanation offered by Pandini et al. [31] and Cholerton et al. [32]. Díaz-Zúñiga et al. [14] and Ilievski et al. [29] also observed neuroinflammation with the capsulated P. gingivalis strains as originally reported by Poole et al. [33] in their P. gingivalis (strain FDC 381, non-capsulated, K0 serotype) mono-infected apolipoprotein knockout mouse model. In the db/db mouse model, F. nucleatum infected mouse hippocampi demonstrated significant numbers of activated, astrocytes but not microglia. This was an interesting observation showing that even a bridging bacterium like F. nucleatum can stimulate astrocytes, but a much stronger immune stimulation was provided by P. gingivalis (W83) virulence factors that led to the activation of microglia. Thus, our results concur with previous observations which showed that an oral infection with P. gingivalis in rodents leads to an inflammatory intracerebral response. This is also reported in humans after an infectious episode [34], which has been associated with cognitive deficit [35].

In support of further inflammatory activity resulting from P. gingivalis (W83) infection, gene expression analysis of a panel of 184 genes that were altered due to P. gingivalis (W83) oral infection demonstrated that the pathways of insulin signaling and neuroinflammatory immune response mediated by PI3K/AKT1 were hyperactivated in the brain. In addition, AKT1/GSK-3B hyperactivation can also promote a pro-inflammatory state through phosphorylation of STAT3, a transcription factor required to induce pro-inflammatory cytokines [36]. An increase in the mRNA abundance of STAT3 gene and several pro-inflammatory cytokines (IL1A and IL1B) genes in the present study also agree with previous reports for P. gingivalis infection mediated cytokine release [3, 14]. These results demonstrate the potential of P. gingivalis W83 induced inflammation in a co-morbidity mouse model of AD.

In the present study, P. gingivalis infection mediated the activation of insulin secretion, insulin receptor signaling, and IGF-1 signaling pathways. These interconnected pathways involve multiple signaling routes, potentially contributing to AD pathophysiology (Fig. 6). The mammalian brain is a highly insulin-sensitive organ, which heavily depends on insulin to utilize glucose as an energy source and support cognitive functions [37]. Endogenous insulin released by the pancreas crosses the blood-brain barrier to enter the neural tissue of the brain, but our results suggest a plausible activation of a localized insulin secretion in the hippocampus following the P. gingivalis infection. Our gene expression pathway analysis identified 11 genes (GLP1R, INSR, JAK2, PIK3R1, PIK3R2, PRKACA, PRKAR1A, PRKCA, PRKCB, PRKCD, and STAT3) whose mRNA abundance indicated activation of a localized insulin secretion mechanism in the hippocampal region. One potential route of insulin secretion in the brain is via glucagon-like peptide 1 (GLP-1) and protein kinase [38]. During a hyperglycemic state, GLP-1 in the brain is able to cross the blood-brain barrier [39] and inhibit glucose utilization and cause insulin secretion [40, 41]. In this study, P. gingivalis induction resulted in a higher mRNA abundance of GLP1R and several kinases and their receptors and lower mRNA abundance of G6P, a key glucose utilization marker. The adult neuronal cells of the hippocampus were previously reported to secrete insulin [42]. A localized synthesis of insulin in the brain may be a compensatory mechanism linked to the insulin resistance that has the potential to develop following P. gingivalis infection and its mechanism of action warrants further research.

Fig. 6

Schematic outline of four major pathways (PI3K/AKT pathway, MAPK pathway, Insulin signaling pathway and TNF pathway) affected in the db/db mice due to the P. gingivalis W83 induction. These interconnected pathways lead to altered activities of kinases (PI3K/AKT1, RAS/MEKK1)), affect insulin signaling (IRS1) and nutrient metabolism (AKT1), alter oxidative stress (NOS2), transcriptional activation pro-inflammatory genes (NFKB, AP1, cFOS), all potentially contributing to Alzheimer’s disease pathophysiology (Aβ plaque and tau neurofibrillary tangles). The mRNA abundance of the markers as found in this study to increase (up arrow) or decrease (down arrow) following the P. gingivalis W83 treatment compared to those of the F. nucleatum infection are shown. The stimulatory and inhibitory relationship of the molecules are shown as arrow and hammer signs, respectively.

AD is a neurodegenerative disease being linked to MetS often characterized by insulin resistance, oxidative stress, and neuroinflammation [43 –45]. As identified through pathway/GO analyses, a higher level of insulin activity in the brain potentially causing insulin resistance was also evident from the increased mRNA levels of key markers of the insulin/insulin-like growth factor-1 (IGF-1) pathways. For example, the P. gingivalis infection increased the mRNA abundance of IGF1, insulin receptor (IR), IGF-1 receptor (IGF1R), and IGF binding protein (IGFBP3) genes. While the shared IR/IGF1R system in the plasma membrane of hippocampal cells is crucial for normal insulin signaling, binding of IGFBP-3 to IGF-1 can potentially contribute to glucose intolerance and insulin resistance in the brain [46]. An interesting protein in the IR/IGF1R system is the insulin receptor substrate 1 (IRS-1), a docking protein, which is a primary checkpoint for the insulin signaling pathway and glucose metabolism regulation [43]. The normal physiological activation of the insulin signaling is achieved by binding insulin to the insulin receptor within the extracellular matrix resulting in the phosphorylation of IRS-1 at the tyrosine residue. However, phosphorylation on multiple serine residues can inhibit IRS-1 activity and result in insulin resistance [47]. P. gingivalis infection increased the mRNA abundance of the IRS1 gene. Whether an increased level of IRS1 mRNA due to the P. gingivalis infection in this study is associated with phosphorylation at the serine residues, thereby contributing to the insulin resistance requires further investigation.

The interaction of insulin/IGF-1 to the IR/IGF1R receptor system triggers the release of IRS-1/2 and subsequent activation of the phosphatidylinositol 3-kinase type I (PI3K)/AKT pathways [43]. Our gene expression profiling indicated P. gingivalis infection activates multiple pathways that potentially lead to hyperactivation of the PI3K/AKT pathway genes. These include numerous mRNA markers of the IR/IGF1R, LPS/TLR, and MAPK pathways (Fig. 6). The PI3K/AKT1 pathway plays a vital role in the metabolism of nutrients and maintains cellular homeostasis essential for cell growth, survival, and differentiation [48].

An impaired insulin/PI3K/AKT1 pathway activity can contribute to the onset of neurodegenerative diseases [48, 49]. During the cellular oxidative stress, PI3K, together with AKT1, triggers reactive oxygen species (ROS) production [50, 51]. The hyperactivation of PI3K/AKT pathway results in lower nitric oxide (NO) production, higher ROS activities, and a higher level of pro-inflammatory cytokines, which are known to culminate in the pathophysiology of AD [43]. A conspicuous decrease in the mRNA abundance of nitric oxide synthase 2 (NOS2) by P. gingivalis infection observed in this study further suggests potentially a lower level of NO synthesis. Other mechanisms of AKT-mediated response include affecting glucogenesis, lipolysis, and mTOR-mediated protein synthesis, leading to impaired autophagy and synaptic plasticity associated with AD pathophysiology [49]. The pathway analysis also revealed inhibition of the phosphatase and tensin homolog (PTEN), a tumor suppressor molecule that catalyzes the conversion of PIP3 to PIP2 and reverses the PI3K/AKT activities. Inhibition of the PTEN pathway is likely to contribute to the hyperactivation of PI3K/AKT pathways.

In the present study, the mRNA abundance of AKT1 and its three crucial downstream cascades triggered by the action of GSK3B, FOXO1, and NFKB2 genes as they were upregulated by P. gingivalis infection. Hyperactivity of AKT1 may lead to excessive phosphorylation of GSK-3B, FOXO1, and NF-kB, resulting in several consequences that affect metabolic and cognitive functions [48]. AKT1/GSK-3B hyperactivation causes abnormal phosphorylation of tau, leading to NFT formation, a hallmark of AD pathophysiology [43, 52]. In this study, the mRNA abundance of both AKT1 and GSK3B genes were increased by the P. gingivalis infection, and a higher level of phosphorylated tau was also evident in the hippocampal neurons with thickened bundles of axon hillock(s) in the hippocampus. P. gingivalis infection inducing greater phosphorylated tau was also reported by other recent studies [29, 53]. We have previously demonstrated that P. gingivalis LPS has the potential to increase the mRNA abundance of GSK3B, STAT3, and CREB1 genes in the IMR-32 neuronal cell model [27]. In this study, while the mRNA abundance of GSK3B, STAT3 genes increased due to P. gingivalis infection, that of CREB1 gene decreased. The P. gingivalis mediated induction of pro-inflammatory responses is likely to include LPS and gingipain virulence factors [9], and it is a multi-component signaling cascade that is likely to be induced that may negatively impact on tau hyperphosphorylation.

The activation of the AKT1/FOXO1cascade results in enhanced transcriptional activity of FOXO1 that can lead to hyperglycemia and ROS production, and thus insulin resistance and oxidative stress [54]. FOXO1 can also activate c-Jun N-terminal kinase and inhibit Wingless (Wnt) pathways and can contribute to Aβ plaque formation and phosphorylation of tau, potentially leading to neurodegeneration [54]. The FOXO1 within the nucleus activates the transcription of many inflammatory mediators. In the present study, the mRNA abundance of FOXO1, and that of a few downstream molecules such as pro-inflammatory cytokines (IL1B), wound healing factor (VEGF), and adhesion molecule (VCAM1) were increased. Together these observations suggest that P. gingivalis infection does indeed hyperactivate the AKT1/FOXO1 cascade and could potentially exacerbate the insulin resistance and oxidative stress pathways.

In the present study, the mRNA abundance of the NF-kB cascade (NFKB2, IKBKB) and pro-inflammatory cytokines (LTA, IL1A, IL1B, IL6R, CSF1) were upregulated by the P. gingivalis infection. The AKT1/NF-kB signaling plays a crucial role in the inflammatory response, oxidative stress, activation of microglia, and apoptotic cell death, thereby contributing to neurodegeneration [55]. The NF-kB has both neuroprotective and neurodegenerative roles in AD. One of the mechanisms through which NF-kB contributes to AD is mediated by activation of β-site APP cleaving enzyme 1 (BACE1). An excessive cleavage of amyloid-beta protein precursor leads to Aβ plaque deposition [56]. The gene expression pathway analysis further indicated that the PPAR pathway was downregulated due to the P. gingivalis infection in db/db mice. Inhibition of PPAR activities may, therefore, lead to the higher activity of BACE1 [57].

Individuals with periodontal disease have an increased risk for cardiovascular pathologies and AD. Subjects with predisposing factors like MetS, obesity, and diabetes, are at greater risk of developing periodontal disease and AD [58]. The db/db mouse model tested here provided a unique opportunity to explore the associations between obesity, diabetes, and periodontitis on the development of AD. Furthermore, it can be inferred from these observations that P. gingivalis W83 infection induces a distinct inflammatory response because of a chronic oral infection during T2D which has the potential to lead to the development of AD. Our observations from the present study provide further support to the notion that AD is an inflammatory disease affected by peripheral (systemic) [23 , 60], and intracerebral inflammatory events (localized) which ultimately impact on cognitive decline [61, 62].

Consequences of the P. gingivalis W83 in a non-db/db model and prevention

Currently, there are only two experimental models of periodontitis with P. gingivalis (W83) infection in non-db/db rodents, which have both been shown to produce early hallmark lesions [14, 29]. In these studies, neuroinflammation was an unequivocal outcome post infection with various strains of P. gingivalis. Advancing age and apolipoprotein E gene allele 4 inheritance, which are significant risk factors for AD, are associated with increased susceptibility to infection. This highlights potential preventive initiatives for AD which can lower the risk of neuroinflammation in individuals with and without co-morbidities. Hence, there is a critical need to initiate future studies that focus on exploring the preventive measures which could help us in attenuation of consequences induced by P. gingivalis infection.

Footnotes

ACKNOWLEDGMENTS

SK and SKS in 2017 and again with BB and RW and SKS in 2018 received PreViser awards from the Oral and Dental Research Trust. In addition, SK also acknowledges having received a TC White Young Researcher award (2019). SC acknowledge the support from faculty seed grant (UFCD), University of Florida, Gainesville, FL, USA.

Authors’ disclosures available online ().

References

Chapple

, Van der Weijden

, Doerfer

, Herrera

, Shapira

, Polak

, Madianos

, Louropoulou

, Machtei

, Donos

, Greenwell

, Van Winkelhoff

, Eren Kuru

, Arweiler

, Teughels

, Aimetti

, Molina

, Montero

, Graziani

(2015) Primary prevention of periodontitis: Managing gingivitis. J Clin Periodontol 42(Suppl 16), S71–S76.

How

, Song

, Chan

(2016) Porphyromonas gingivalis: An overview of periodontopathic pathogen below the gum line. Front Microbiol 7, 53.

Vernal

, León

, Silva

, van Winkelhoff

, García-Sanz

, Sanz

(2009) Differential cytokine expression by human dendritic cells in response to different porphyromonas gingivalis capsular serotypes. J Clin Periodontol 36, 823–829.

Silver

, Martin

, McBride

(1977) Experimental transient bacteraemias in human subjects with varying degrees of plaque accumulation and gingival inflammation. J Clin Periodontol 4, 92–99.

Païssé

, Valle

, Servant

, Courtney

, Burcelin

, Amar

, Lelouvier

(2016) Comprehensive description of blood microbiome from healthy donors assessed by 16S targeted metagenomic sequencing. Transfusion 56, 1138–1147.

DeLeon-Pennell

, de Castro Brás

, Lindsey

(2013) Circulating Porphyromonas gingivalis lipopolysaccharide resets cardiac homeostasis in mice through a matrix metalloproteinase-9-dependent mechanism. Physiol Rep 1, e00079.

Troseid

, Nestvold

, Rudi

, Thoresen

, Nielsen

, Lappegard

(2013) Plasma lipopolysaccharide is closely associated with glycemic control and abdominal obesity. Diabetes Care 36, 3627–3632.

Poole

, Singhrao

, Kesavalu

, Curtis

, Crean

StJ

(2013) Determining the presence of periodontopathic virulence factors in short-term postmortem Alzheimer’s disease brain tissue. J Alzheimers Dis 36, 665–677.

Dominy

, Lynch

, Ermini

, Benedyk

, Marczyk

, Konradi

, Nguyen

, Haditsch

, Raha

, Griffin

, Holsinger

, Arastu-Kapur

, Kaba

, Lee

, Ryder

, Potempa

, Mydel

, Hellvard

, Adamowicz

, Hasturk

, Walker

, Reynolds

, Faull

RLM

, Curtis

, Dragunow

, Potempa

(2019) Porphyromonas gingivalis in Alzheimer’s disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Sci Adv 5, eaau3333.

10.

Alzheimer

, Stelzmann

, Schnitzlein

, Murtagh

(1995) An English translation of Alzheimer’s 1907 paper, “Uber eine eigenartige Erkankung der Hirnrinde”. Clin Anat 8, 429–431.

11.

Veerhuis

, Nielsen

, Tenner

(2011) Complement in the brain. Mol Immunol 48, 1592–603.

12.

Kumar

, Choi

, Washicosky

, Eimer

, Tucker

, Ghofrani

, Lefkowitz

, McColl

, Goldstein

, Tanzi

, Moir

(2016) Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer’s disease. Sci Transl Med 8, 340ra72.

13.

Eimer

, Vijaya Kumar

, Navalpur Shanmugam

, Rodriguez

, Mitchell

, Washicosky

, György

, Breakefield

, Tanzi

, Moir

(2018) Alzheimer’s disease-associated β-amyloid is rapidly seeded by herpesviridae to protect against brain infection. Neuron 99, 56–63.e3.

14.

Díaz-Zúñiga

, More

, Melgar-Rodríguez

, Jiménez-Unión

, Villalobos-Orchard

, Muñoz-Manríquez

, Monasterio

, Valdés

, Vernal

, Paula-Lima

(2020) Alzheimer’s disease-like pathology triggered by Porphyromonas gingivalis in wild type rats is serotype dependent. Front Immunol 11, 588036.

15.

Chaffee

, Weston

(2010) Association between chronic periodontal disease and obesity: A systematic review and meta-analysis. J Periodontol 81, 1708–1724.

16.

Campus

, Salem

, Uzzau

, Baldoni

, Tonolo

(2005) Diabetes and periodontal disease: A case-control study. J Periodontol 76, 418–425.

17.

Mysak

, Podzimek

, Sommerova

, Lyuya-Mi

, Bartova

, Janatova

, Prochazkova

, Duskova

(2014) Porphyromonas gingivalis: Major periodontopathic pathogen overview. J Immunol Res 2014, 476068.

18.

Chiu

H-C

, Fu

MM-J

, Yang

T-S

, Fu

, Chiang

C-Y

, Tu

H-P

, Chin

Y-T

, Lin

F-G

, Shih

K-C

(2016) Effect of high glucose, Porphyromonas gingivalis lipopolysaccharide and advanced glycation end-products on production of interleukin-6/-8 by gingival fibroblasts. J Periodontal Res 52, 268–276.

19.

Murray

, Wilton

(2003) Lipopolysaccharide from the periodontal pathogen Porphyromonas gingivalis prevents apoptosis of HL60-derived neutrophils in vitro. Infect Immun 71, 7232–7235.

20.

Holmes

, Cotterell

(2009) Role of infection in the pathogenesis of Alzheimer’s disease. CNS Drugs 23, 993–1002.

21.

Marques

, Sousa

, Palha

(2013) Blood-brain-barriers in aging and in Alzheimer’s disease. See comment in PubMed Commons below. Mol Neurodegener 8, 38.

22.

van Oijen

, de Jong

, Witteman

, Hofman

, Koudstaal

, Breteler

(2007) Atherosclerosis and risk for dementia. Ann Neurol 61, 403–410.

23.

Demmer

, Norby

, Lakshminarayan

, Walker

, Pankow

, Aaron Folsom

R AR

, Mosley

, Beck

, Lutsey

(2020) Periodontal disease and incident dementia: The Atherosclerosis Risk in Communities Study (ARIC). Neurology 95, e1660–e1671.

24.

Dioguardi

, Crincoli

, Laino

, Alovisi

, Sovereto

, Mastrangelo

, Russo

, Muzio

(2020) The role of periodontitis and periodontal bacteria in the onset and progression of Alzheimer’s disease: A systematic review. J Clin Med 9, 495.

25.

, Cavuoto

, Biddiscombe

, Pike

(2020) Diabetes mellitus increases risk of incident dementia in APOEɛ4 carriers: A meta-analysis. L. J Alzheimers Dis 74, 1295–1308.

26.

Cole

, Neal

, Singhrao

, Jasani

, Newman

(1993) The distribution of amyloid plaques in the cerebellum and brain stem in Down’s syndrome and Alzheimer’s disease: A light microscopical analysis. Acta Neuropathol 85, 542–552.

27.

Bahar

, Singhrao

(2021) An evaluation of the molecular mode of action of trans-resveratrol in the Porphyromonas gingivalis lipopolysaccharide challenged neuronal cell model. Mol Biol Rep 48, 147–156.

28.

Dennis

Jr , Sherman

, Hosack

, Yang

, Gao

, Lane

, Lempicki

(2003) DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 4, P3.

29.

Ilievski

, Zuchowska

, Green

, Toth

, Ragozzino

, Le

, Aljewari

, O’Brien-Simpson

, Reynolds

, Watanabe

(2018) Chronic oral application of a periodontal pathogen results in brain inflammation, neurodegeneration and amyloid beta production in wild type mice. PLoS One 13, e0204941.

30.

Stein

, Desrosiers

, Donegan

, Yepes

, Kryscio

(2007) Tooth loss, dementia and neuropathy in the Nun study. J Am Dent Assoc 138, 1314–1322.

31.

Pandini

, Pace

, Copani

, Squatrito

, Milardi

, Vigneri

(2013) Insulin has multiple antiamyloidogenic effects on human neuronal cells. Endocrinology 154, 375–387.

32.

Cholerton

, Baker

, Craft

(2013) Insulin, cognition, and dementia. Eur J Pharmacol 719, 170–179.

33.

Poole

, Singhrao

, Chukkapalli

, Rivera

, Velsko

, Kesavalu

, Crean

StJ

(2015) Active infection of Porphorymonas gingivalis and infection-induced complement activation in ApoE-/-mice brains. J Alzheimers Dis 43, 67–80.

34.

Dunn

, Mullee

, Perry

, Holmes

(2005) An association between dementia and infectious disease: Evidence from a case-control study. Alzheimer Dis Assoc Disord 19, 91–94.

35.

Holmes

, El-Okl

, Williams

, Cunningham

, Wilcockson

, Perry

(2003) Systemic infection, interleukin 1beta and cognitive decline in Alzheimer’s disease. Neurol Neurosurg Psychiatry 74, 788–789.

36.

Jope

, Cheng

, Lowell

, Worthen

, Sitbon

, Beurel

(2017) Stressed and inflamed, can GSK3 be blamed? Trends Biochem Sci 42, 180–192.

37.

Spinelli

, Fusco

, Grassi

(2019) Brain insulin resistance and hippocampal plasticity: Mechanisms and biomarkers of cognitive decline. Front Neurosci 13, 788.

38.

Meloni

, DeYoung

, Lowe

, Parkes

(2013) GLP-1 receptor activated insulin secretion from pancreatic β-cells: Mechanism and glucose dependence. Diabetes Obes Metab 15, 15–27.

39.

, Gong

, Liu

, Wu

, Barrett

, Aylor

, Liu

(2020) Brain endothelial cells regulate glucagon-like peptide 1 entry into the brain via a receptor-mediated process. Front Physiol 11, 555.

40.

Knauf

, Cani

, Perrin

, Iglesias

, Maury

, Bernard

, Benhamed

, Grémeaux

, Drucker

, Kahn

, Girard

, Tanti

, Delzenne

, Postic

, Burcelin

(2005) Brain glucagon-like peptide-1 increases insulin secretion and muscle insulin resistance to favor hepatic glycogen storage. J Clin Invest 115, 3554–3563.

41.

Sandoval

, Sisley

(2015) Brain GLP-1 and insulin sensitivity. Mol Cell Endocrinol 418(Pt 1), 27–32.

42.

Kuwabara

, Kagalwala

, Onuma

, Ito

, Warashina

, Terashima

, Sanosaka

, Nakashima

, Gage

, Asashima

(2011) Insulin biosynthesis in neuronal progenitors derived from adult hippocampus and the olfactory bulb. EMBO Mol Med 3, 742–754.

43.

Bedse

, Di Domenico

, Serviddio

, Cassano

(2015) Aberrant insulin signaling in Alzheimer’s disease: Current knowledge. Front Neurosci 9, 204.

44.

, Song

, Leng

(2015) Link between type 2 diabetes and Alzheimer’s disease: From epidemiology to mechanism and treatment. Clin Interv Aging 10, 549–560.

45.

de la Monte

(2019) The full spectrum of Alzheimer’s disease is rooted in metabolic derangements that drive type 3 diabetes. Adv Exp Med Biol 1128, 45–83.

46.

Silha

, Gui

, Murphy

(2002) Impaired glucose homeostasis in insulin-like growth factor-binding protein-3-transgenic mice. Am J Physiol Endocrinol Metab 283, E937–E945.

47.

Yarchoan

, Toledo

, Lee

, Arvanitakis

, Kazi

, Han

L-Y

, Louneva

, Lee

VM-Y

, Kim

, Trojanowski

, Arnold

(2014) Abnormal serine phosphorylation of insulin receptor substrate 1 is associated with tau pathology in Alzheimer’s disease and tauopathies. Acta Neuropathol 128, 679–689.

48.

, Na

, Li

, Chen

(2020) Roles of the PI3K/AKT/mTOR signalling pathways in neurodegenerative diseases and tumours. Cell Biosci 10, 54.

49.

Gabbouj

, Ryhänen

, Marttinen

, Wittrahm

, Takalo

, Kemppainen

, Martiskainen

, Tanila

, Haapasalo

, Hiltunen

, Natunen

(2019) Altered insulin signaling in Alzheimer’s disease brain - Special emphasis on PI3K-Akt Pathway. Front Neurosci 13, 629.

50.

Antico Arciuch

, Galli

, Franco

, Lam

, Cadenas

, Carreras

, Poderoso

(2009) Akt1 intramitochondrial cycling is a crucial step in the redox modulation of cell cycle progression. PLoS One 4, e7523.

51.

Franke

, Yang

, Chan

, Datta

, Kazlauskas

, Morrison

, Kaplan

, Tsichlis

(1995) The protein kinase encoded by the Akt proto-oncogene is a target of the PDGF-activated phosphatidylinositol 3-kinase. Cell 2, 727–736.

52.

de La Monte

(2014) Type 3 diabetes is sporadic Alzheimer’s disease: Mini-review. Eur Neuropsychopharmacol 24, 1954–1960.

53.

Haditsch

, Roth

, Rodriguez

, Hancock

, Cecere

, Nguyen

, Arastu-Kapur

, Broce

, Raha

, Lynch

, Holsinger

, Dominy

, Ermini

(2020) Alzheimer’s disease-like neurodegeneration in Porphyromonas gingivalis infected neurons with persistent expression of active gingipains. J Alzheimers Dis 75, 1361–1376.

54.

Manolopoulos

, Klotz

, Korsten

, Bornstein

, Barthel

(2010) Linking Alzheimer’s disease to insulin resistance: The FoxO response to oxidative stress. Mol Psychiatry 15, 1046–1052.

55.

Jha

, Jha

, Kar

, Nand

, Swati

, Goswami

(2019) Nuclear factor kappa β as a therapeutic target for Alzheimer’s disease. J Neurochem 150, 113–137.

56.

O’Brien

, Wong

(2011) Amyloid precursor protein processing and Alzheimer’s disease. Annu Rev Neurosci 34, 185–204.

57.

Wójtowicz

, Strosznajder

, Jeżyna

, Strosznajder

(2020) The novel role of PPARalpha in the brain: Promising target in therapy of Alzheimer’s disease and other neurodegenerative disorders. Neurochem Res 45, 972–988.

58.

Beck

, Moss

, Morelli

, Offenbacher

(2018) Periodontal profile class is associated with prevalent diabetes, coronary heart disease, stroke, and systemic markers of C-reactive protein and interleukin-6. J Periodontol 89, 157–165.

59.

Leira

, Carballo

, Orlandi

, Aldrey

, Pías-Peleteiro

, Moreno

, Vázquez-Vázquez

, Campos

, D’Aiuto

, Castillo

, Sobrino

, Blanco

(2020) Periodontitis and systemic markers of neurodegeneration: A case-control study. J Clin Periodontol 47, 561–571.

60.

, Chen

, van der Sluis

LWM

, Schuller

, Tjakkes

(2021) Mediation analysis of white blood cell count on the association between periodontal inflammation and digit symbol substitution test scoring cognitive function among older U.S. adults. J Gerontol A Biol Sci Med Sci 76, 1309–1315.

61.

Heneka

, Carson

, El Khoury

, Landreth

, Brosseron

, Feinstein

, Jacobs

, Wyss-Coray

, Vitorica

, Ransohoff

, Herrup

, Frautschy

, Finsen

, Brown

, Verkhratsky

, Yamanaka

, Koistinaho

, Latz

, Halle

, Petzold

, Town

, Morgan

, Shinohara

, Perry

, Holmes

, Bazan

, Brooks

, Hunot

, Joseph

, Deigendesch

, Garaschuk

, Boddeke

, Dinarello

, Breitner

, Cole

, Golenbock

, Kummer

(2015) Neuroinflammation in Alzheimer’s disease. Lancet Neurol 14, 388–405.