3×Tg-AD Mice Overexpressing Phospholipid Transfer Protein Improves Cognition Through Decreasing Amyloid-β Production and Tau Hyperphosphorylation

Abstract

Background:

Phospholipid transfer protein (PLTP) belongs to the lipid transfer glycoprotein family. Studies have shown that it is closely related to Alzheimer’s disease (AD); however, the exact effect and mechanism remain unknown.

Objective:

To observe the effect of PLTP overexpression on behavioral dysfunction and the related mechanisms in APP/PS1/Tau triple transgenic (3×Tg-AD) mice.

Methods:

AAV-PLTP-EGFP was injected into the lateral ventricle to induce PLTP overexpression. The memory of 3×Tg-AD mice and wild type (WT) mice aged 10 months were assessed using Morris water maze (MWM) and shuttle-box passive avoidance test (PAT). Western blotting and ELISA assays were used to quantify the protein contents. Hematoxylin and eosin, Nissl, and immunochemistry staining were utilized in observing the pathological changes in the brain.

Results:

3×Tg-AD mice displayed cognitive impairment in WMW and PAT, which was ameliorated by PLTP overexpression. The histopathological hallmarks of AD, senile plaques and neurofibrillary tangles, were observed in 3×Tg-AD mice and were improved by PLTP overexpression. Besides, the increase of amyloid-β42 (Aβ₄₂) and Aβ₄₀ were found in the cerebral cortex and hippocampus of 3×Tg-AD mice and reversed by PLTP overexpression through inhibiting APP and PS1. PLTP overexpression also reversed tau phosphorylation at the Ser404, Thr231 and Ser199 of the hippocampus in 3×Tg-AD mice. Furthermore, PLTP overexpression induced the glycogen synthase kinase 3β (GSK3β) inactivation via upregulating GSK3β (pSer9).

Conclusion:

These results suggest that PLTP overexpression has neuroprotective effects. These effects are possibly achieved through the inhibition of the Aβ production and tau phosphorylation, which is related to GSK3β inactivation.

Keywords

Alzheimer’s disease 3×Tg AD mice GSK3β PLTP

INTRODUCTION

Alzheimer’s disease (AD) is an insidious, progressive degenerative disease of the central nervous system (CNS) characterized by deficits in episodic memory, working memory, and executive function. It is a serious threat to the health of elderly people [1 –3]. Despite the existence of many hypotheses about its pathogenesis of AD, effective treatment for it has not been elucidated at present. From the literature, the development of AD is linked to the accumulation of extracellular amyloid-β protein (Aβ) and intracellular tau phosphorylation [4, 5].

Phospholipid transfer protein (PLTP), as one of the key proteins in lipid and lipoprotein metabolism, is responsible for the transportation of phospholipids, diacylglycerol, α-tocopherol, cerebroside, and lipopolysaccharides in plasma and peripheral tissues [6]. The activity of PLTP in the metabolism of lipoproteins, especially high density lipoprotein (HDL), plays a major role in atherogenesis [7]. PLTP is widely distributed in the CNS [8] and is considered to be associated with many CNS disease, such as multiple sclerosis [9], Down syndrome [10], AD [8 , 12], and anxiety [13, 14]. Previous studies in mice have shown that PLTP deficiency deteriorates cognitive function and promotes the progression of AD, effects which seem to be related to the interference with Aβ metabolism [15 –17]. On the other side, cellular research confirmed that recombinant PLTP protein inhibits tau phosphorylation. Although there were some contrary observations [18], these previous findings suggest that PLTP could be playing a beneficial role in the pathogenesis of AD and might be a promising treatment target. However, there is no direct evidence that shows that PLTP can improve AD.

A past cellular study reported that PLTP increases phosphorylation of inactive glycogen synthase kinase-3β (GSK3β) at serine 9 [19]. GSK3β is a key enzyme in signal transduction and is widely distributed in the brain [20]. Studies have shown that GSK3β is involved in both tau phosphorylation and Aβ metabolism [21 –23]. However, it is not clear whether PLTP regulation of Aβ metabolism is also related to GSK3β. Besides, the pieces of experimental evidence showing whether PLTP can improve AD through GSK3β mediated Aβ metabolism and tau phosphorylation are limited. In the present study, we investigated the effects of PLTP overexpression on AD and revealed the underlying neurochemical mechanism.

MATERIALS AND METHODS

Reagents and antibodies

The adeno-associated virus-serotype 2/9 vectors (AAV2/9) carrying PLTP gene and coding EGFP (AAV-PLTP) was constructed by GeneChem Company (Shanghai, China) for the overexpression in mice. The empty AAV vectors coding EGFP (AAV-Null) were used as the control. Primary antibodies against PLTP (1:500), Presenilin 1 (PS1; 1:5,000), β-Site amyloid precursor protein-cleaving enzyme-1 (BACE1; 1:1,000), Aβ₄₂ (1:1,000), tau protein (1:5,000), tau [Phospho-Ser404 (1:1,000), Ser199 (1:10,000), Ser214 (1:1,000) and Thr231 (1:1,000)], glycogen synthase kinase 3β (GSK3β; 1:1,000), GSK3β (Phospho-Ser9; 1:10,000) and GSK3β (Phospho-Tyr216; 1:1,000) were purchased from Abcam (Cambridge, MA, USA) and the antibody against amyloid-β protein precursor (AβPP; 1:1,000) was purchased from cell signaling technology (Beverly, CA, USA). Antibodies against β-actin (1:1,000) and GAPDH (1:1,000) were purchased from ZSGB-BIO (Beijing, China). Silver nitrate and Mayer’s Hematoxylin Solution were purchased from Sigma-Aldrich. 3, 3’-diaminobenzidine (DAB) staining reagent were purchased from ZSGB-Bio (Beijing, China). Aβ₄₀ and Aβ₄₂ immunosorbent assay (ELISA) kits were purchased from Shanghai Blue Gene Biotech (Shanghai, China). BCA Protein Quantitation Kits were purchased from Solarbio (Shanghai, China).

Animals

The triple transgenic AD [B6; 129-Psen1^tm1Mpm Tg (APPSwe, tauP301L) 1Lfa/Mmjax, 3×Tg-AD] mice and wild type (WT) non-transgenic mice were purchased from The Jackson Laboratory. In this study, 24 male 3×Tg-AD mice aged 10 months were used, while 12 WT mice of the same age and sex were used as controls. All mice were housed in individually ventilated cages (IVC) with a 12-h light/12-h dark cycle (7:00 am-7:00 pm) and were given food and water ad libitum in the SPF animal facilities at the Institute of Pharmacology of Shandong First Medical University (Tai’an, China). All experiments were performed in accordance with the protocols approved by the Laboratory Animals’ Ethic Committee of Shandong First Medical University.

Stereotaxic injection of the AAV virus and experimental schedule

Adeno-associated virus (AAV) shows promise for the treatment of brain disorders in clinical trials because of its effectiveness for transgene expression in vivo. Studies have suggested that multiple brain sections show transduction after intracerebroventricular injections (IVC) of AAV particles [24]. Such sections include the cerebral cortex and hippocampus, which are closely related to learning and memory [25]. To elucidate the role of PLTP overexpression in AD, the ICV injection of AAV particles was performed as follows. The mice were first anesthetized under intraperitoneal chloral hydrate (300 mg/kg) and then localized by a stereotactic instrument for operation. After incising the cranial midline, the periosteum was removed and bregma and lambda points identified. The injection site was measured relative to the bregma: 0.6 mm posterior to the bregma and 1.2 mm laterally from the sagittal suture. The skull was perforated on the both sides by microdrill and an automatic Hamilton syringe was used to inject 2μl AAV-PLTP (8×10¹¹ VP) or AAV-Null (8×10¹¹ VP) 2.2 mm below the dura mater (0.2μl/min). Among 3×Tg-AD mice, twelve were injected with AAV-PLTP and another twelve with AAV-null. All the WT mice were injected with AAV-null. To avoid spillover, syringes were kept in place for 10 min before withdrawal after each injection. The incision was sutured, and mice were put back into their cages and kept for thirty days. Combined with our pre-experimental results and references [24], the protein was highly expressed in the third week of transfection. Therefore, in this day, from day 22nd to 30th after IVC injection of AAV particles, behavioral experiments such as Morris water maze (MWM) and passive avoidance test (PAT) were carried out. One hour after the last behavioral test (i.e., the PAT), the mice were sacrificed under decapitation and brains collected. The hemispheres of three mice in each group were collected for pathological staining. The cerebral cortex and hippocampus of the rest of the mice were isolated from the hemispheres and stored at –80°C, and the cerebral cortex and hippocampus of six mice in each group were used for western blot and ELISA assay.

Morris water maze

To examine spatial learning and memory, the MWM was conducted in this study from day 22nd to 27th in mice as described [26]. Simply put, the mice were individually placed into an open swimming arena (diameter 120 cm, height 50 cm, water depth 30 cm) to locate an invisible underwater platform (1 cm under the water) for five days, which was designated as the navigation trial. The maximum swimming time was 60 s, and mice were allowed to stay on the platform for 10 s. If the mouse did not escape by swimming to the platform within 60 s, the escape latency, which was defined as the time spent in locating the hidden platform, was recorded as 60 s. On the sixth day of the MWM, the probe trial was conducted as follows. The platform was removed, and the mice were placed from the opposite side of the previous platform quadrant (i.e., the target quadrant) to swim for 60 s. Time at which the mouse first reached the location of the target quadrant, the numbers of crossing into the target quadrant, and the swimming distance were recorded. The tracking information was processed by the Topscan Package (Clever Sys Inc., USA).

Passive avoidance test

To evaluate passive avoidance learning in mice, the PAT was used in a light/dark shuttle-box from the 28th day to the 30th day as previously described with small modifications [27]. The first day of PAT was designated adaptation day. Each mouse was placed from in the bright compartment and was allowed to explore between the dark compartment and the bright compartment for 2 min. The second day of PAT was designated as a training day, each mouse was still placed at the bright compartment to explore between the two compartments for 2 min, but when the mouse entered the dark compartment, it immediately received the electric shock (0.6 mA, 2 s). The third day of PAT was designated as a test day, each mouse was placed from at the bright compartment to explore two compartments for 2 min, similar to the first day, but the mouse entered the dark compartment without electric shock. The following indicators, the latency and entries into the dark compartment, and the duration in the dark compartment of mice, were collected. The tracking information was processed by the Topscan Package (Clever Sys Inc., USA).

Hematoxylin and eosin (H&E) and Nissl staining

As previously described [26], the fresh brain tissues were fixed in 4%paraformaldehyde, embedded with paraffin, cut into 3-μm sections on a microtome (Leica, Nussloch, Germany), and then stained with H&E or Nissl (toluidine blue) staining. In H&E staining, the sections were stained in hematoxylin and eosin for 3–5 min and 5 min, respectively. In Nissl staining, the slides were stained in Toluidine Blue solution for 3 min. All the sections were observed under the microscope (Nikon Eclipse E200).

Immunochemistry staining

According to the literature [28], the brain embedded in paraffin was cut into 5-μm sections. The sections were put through xylene dewaxing and graded alcohol dehydration. The antigens were recovered in citric acid buffer fluid boiled for 15 min, and the activity of endoperoxidase blocked by 3%H₂O₂ for 25 min. The sections were then sealed by 3%Bovine Serum Albumin (BSA) for 30 min. They were then incubated with primary antibody (rabbit anti-Aβ₄₂, 1:100 dilution in PBS-T containing 1%BSA) overnight at 4°C and stained with a secondary antibody with horseradish peroxidase (HRP) (goat anti rabbit IgG, 1 : 100 dilution in PBS-T containing 1%BSA). Immunoreactivity was visualized by the DAB Detection Kit according to the manufacturer’s instructions. Finally, the sections were counterstained with hematoxylin, dehydrated, and mounted by a neutral gum, and cover slipped. The stained sections were photographed using a microscope (Nikon Eclipse E200).

Silver staining

As previously described [15], the sections embedded in paraffin were stained in pre-warmed 10%silver nitrate solution for 15 min, and then incubated in ammonium silver solution at 40°C for 30 min. Staining was stop using 1%ammonium hydroxide solution and the sections were in turn fixed, dehydrated, cleared, and mounted with a resinous medium. This was finally followed by visualizing the sections with a microscope.

ELISA assay

The level of Aβ₄₀ and Aβ₄₂ in the hippocampus and cerebral cortex were quantified by ELISA Assay (Shanghai Bluegene Biotech Co., Ltd., Shanghai, China) according to previous protocol [29]. Briefly, the brain was homogenized in 0.01 M phosphate buffer saline (PBS) containing protease inhibitor phenylmethanesulfonyl fluoride (PMSF) and centrifuged at 10,000 g for 10 min at 4°C. The supernatant was collected and analyzed using ELISA kits following the manufacturer’s protocol. The protein quantification was performed using the bicinchoninic acid (BCA) assay. The colorimetric reaction was conducted and the absorbance at 450 nm was read using a multifunctional microplate reader (TECAN, Switzerland). The results were normalized to protein concentrations.

Western blot

As reported previously [26], total protein was extracted from the tissues. Expressions of AβPP, PS1, BACE1, tau, tau (pSer214), tau (pThr231), tau (pSer199), tau (pSer404), GSK3β, GSK3β (pSer9), and GSK3β (pTyr216) were determined by western blot. GAPDH or β-actin expression was used as an internal reference. The BCA method was used to measure the protein concentrations with the Enhanced BCA Protein Assay Kit (Solarbio, Shanghai, China) according to the manufacturer’s instructions. The protein signals were visualized and sensitometry analyses conducted using the Image-Pro Plus software version 6.0 (Media Cybernetics Corp, Bethesda, MD, USA).

Statistical analysis

All quantitative data were expressed as the means±standard deviation (SD) and analyzed using GraphPad Prism version 8 (GraphPad Software Inc., San Die CA, USA). Comparisons among multiple groups were performed by one-way ANOVA, and p values of < 0.05 were considered statistically significant.

RESULTS

PLTP overexpression was induced in 3×Tg-AD mice by AAV mediated PLTP gene transfection

The levels of PLTP were deleted by western blot in cerebral cortex and hippocampus to determine the PLTP expression in 3×Tg-AD mice with and without IVC injection of AAV-PLTP. The results showed that PLTP expression of both cerebral cortex and hippocampus in 3×Tg-AD with AAV-null were downregulated compared with WT. However, the expression of PLTP in 3×Tg-AD with AAV-PLTP was significantly increased compared with 3×Tg-AD with AAV-null (Fig. 1A-C). This indicated that PLTP overexpression was successfully induced.

Fig. 1

PLTP overexpression were induced in 3×Tg-AD mice by AAV mediated PLTP gene transfection. A) Representative images by western blotting for the expression of PLTP in the hippocampus and cerebral cortex of WT or 3×Tg-AD mice treated with AAV-null or AAV-PLTP. B, C) Quantitative analysis of PLTP. The values shown are means±SEM, normalized by GAPDH to WT mice treated with AAV-null. ^##p < 0.01, ^#p < 0.05 versus WT mice plus AAV-null; ^**p < 0.01 versus 3×Tg-AD mice with AAV-null; n = 12.

PLTP overexpression reversed the cognition deficits in 3×Tg-AD mice

To evaluate the effect of PLTP overexpression on cognitive function in mice, we performed MWM and PAT. During the navigation trail of MWM, the escape latency of all mice was gradually shortened with the increase in training. From the second day of training, the escape latency in 3×Tg-AD mice with AAV-null showed a marked increase compared to WT mice, while from the third day, PLTP overexpression could significantly reduce escape delay in 3×Tg-AD mice (Fig. 2A, B). In the probe trail of MWM, the escape latency was evidently longer (Fig. 2C) and the number of crossings into the target quadrant was less (Fig. 2D) in 3×Tg-AD with AAV-null mice than WT mice. PLTP Overexpression reversed the two observations. Moreover, the swimming distance of all mice did not change significantly with or without PLTP overexpression (Fig. 2E) indicating that all mice had the same physical status. Similar to MWM, in the PAT, the entrance latency to the dark compartment was tremendously shorten (Fig. 2F), and the entries into the dark compartment and the dark time (Fig. 2G) and the entrance latency to the dark compartment (Fig. 2H) were obviously increased in 3×Tg-AD mice with AAV-null than WT mice, and these observations were significantly reversed following PLTP overexpression.

Fig. 2

PLTP overexpression reversed the cognition deficits in 3×Tg-AD. Cognition function were assessed using Morris water maze (MWM) and passive avoidance test (PAT). In the acquisition trial of MWM, all mice were tested for 5 consecutive days to locate the hidden platform; the escape latency (A) and swimming paths (B) were recorded. During the probe trial, the platform was removed, and the escape latency that is the time entering the previous platform area for the first time (C) and entries into the platform area (D) and the swimming distance (E) were recorded. In the PAT, the latency into dark compartment (F), the duration in dark compartment (G), and the entries into dark compartment (H) were recoded. Values shown represent the means±SEM. ^##p < 0.01, ^#p < 0.05 versus WT mice with AAV-null; ^*p < 0.05 versus 3×Tg-AD mice with AAV-null; n = 12.

PLTP overexpression ameliorates histopathological damage of brain in 3×Tg-AD mice

H&E and Nissl staining were applied to reveal the neuronal damage. The results showed that the number of cortical neurons and hippocampus cells was decreased, the nuclei were concentrated, and the cells were swollen and deformed in 3×Tg-AD mice with AAV-null compared with WT mice. These changes were improved after overexpression of PLTP (Fig. 3).

Fig. 3

PLTP overexpression ameliorated histopathological damage of brain in 3×Tg-AD mice. Neuronal injuries of the cerebral cortex and hippocampal CA1 sections were examined by H&E and Nissl staining (upper and lower two rows of each stain) with or without PLTP overexpression. The neurons in WT mice appeared regularly spaced within the layers with stained cytoplasm and nuclei located approximately at the center of the soma (white arrows). Damaged neurons in 3×Tg-AD mice were darkly stained and exhibited shrunken and triangulated in the neuronal bodies, cytoplasmic swelling and deformation (black arrows). n = 3.

PLTP overexpression relieves amyloid plaques and neurofibrillary tangle (NFT) in 3×Tg-AD mice

Together with amyloid plaques, NFTs are considered the major contributor to the symptoms of AD. Thus, we sought to determine whether PLTP overexpression regulates these two characteristic pathological hallmarks of AD through immunochemistry staining with Aβ₄₂ and Bielschowsky silver stain. We observed more amyloid plaques in the cerebral cortex and hippocampus of 11-month-old 3×Tg-AD mice with AAV-null than in WT mice (Fig. 4A-C). The density of amyloid plaques in 3×Tg-AD mice with AAV-PLTP was found to be significantly lower than in 3×Tg-AD mice with AAV-null (Fig. 4A-C). Similarly, more NFT was observed in the cerebral cortex and hippocampus of 3×Tg-AD mice with AAV-null than in WT mice and 3×Tg-AD mice with AAV-PLTP (Fig. 4A, D, E).

Fig. 4

PLTP overexpression relieved amyloid plaques and neurofibrillary tangles (NFTs) in 3×Tg-AD mice. A) Amyloid plaques and NFTs of the cerebral cortex and hippocampal CA1 sections were observed by immunochemistry staining with Aβ₄₂ and Bielschowsky silver stain separately (upper and lower two rows of each stain). Plaques (red arrows) and NFTs (yellow arrows) were obviously found in the 3×Tg-AD mice with AAV-null. B, C) Quantification of immunochemistry staining with Aβ₄₂ in the cerebral cortex and hippocampal CA1 sections. D, E) Quantification of NFTs in the cerebral cortex and hippocampal CA1 sections by Bielschowsky silver staining. The values shown are means±SEM, ^##p < 0.01, ^#p < 0.05 versus WT mice plus AAV-null; ^**p < 0.01, ^*p < 0.05 versus 3×Tg-AD mice with AAV-null; n = 3.

PLTP overexpression regulates AβPP processing and Aβ generation in 3×Tg-AD mice

Amyloid plaques are composed of Aβ, a predominantly 40–42-amino acid peptide (Aβ₄₀ and Aβ₄₂) formed through consecutive proteolytic cleavages of the AβPP) by the β-site AβPP cleaving enzyme (BACE1) and γ-secretase. Presenilin1 (PS1) on the other hand is one of the four integral membrane protein components of the mature γ-secretase complex. In this study, AβPP metabolic pathway proteins were deleted to allow for clear determination of the effect of PLTP overexpression on the metabolism and deposition of Aβ. As a result, the soluble Aβ₄₀ and Aβ₄₂ were upregulated in the cortex and hippocampus of 3×Tg-AD mice with AAV-null compared with WT mice (Fig. 5A, B). The overexpression of PLTP reversed all the changes in hippocampus (Fig. 5A, B) and only Aβ₄₂ in the cerebral cortex of 3×Tg-AD mice (Fig. 5C, D). Similarly, compared with WT mice, AβPP and PS1 were increased in the cortex and hippocampus of 3×Tg-AD mice with AAV-null and the PLTP overexpression reversed this increase in both cerebral cortex and hippocampus (Fig. 5E-I). No statistically significant difference in BACE1 was observed between the groups (Fig. 5E).

Fig. 5

PLTP overexpression regulated AβPP processing and Aβ generation in 3×Tg-AD mice. A-D) The level of Aβ₄₀ and Aβ₄₂ were detected by ELISA assay in the cerebral cortex and hippocampal of WT or 3×Tg-AD mice treated with AAV-null or AAV-PLTP. E) Representative images by western blotting for the expression of AβPP and PS1 in the cerebral cortex and hippocampal of WT or 3×Tg-AD mice treated with AAV-null or AAV-PLTP. F-I) Quantitative analysis of AβPP and PS1. The values shown are means±SEM, normalized by β-actin to WT mice treated with AAV-null. ^##p < 0.01, ^#p < 0.05 versus WT mice plus AAV-null; ^**p < 0.01, ^*p < 0.05 versus 3×Tg-AD mice with AAV-null; n = 6.

PLTP overexpression inhibits hyperphosphorylation of tau protein in 3×Tg-AD mice

NFT is formed by the accumulation of the hyperphosphorylated protein tau; thus, we deleted tau protein and shut off tau phosphorylation to deeper understand the effect of PLTP overexpression on AD. Form our results, tau protein and phosphorylation of tau protein at Ser404 (tau (pSer404)) and Ser214 (tau (pSer214)) were markedly increased in both the cerebral cortex and hippocampus of 3×Tg-AD mice with AAV-null compared with WT mice (Fig. 6A, B, F, G, I, and K). Tau (pSer231) and tau (pSer199) was found only in the hippocampus of the 3×Tg-AD mice (Fig. 6A, C, E). Further, in the hippocampus of 3×Tg-AD mice, PLTP overexpression reversed the increase in tau (pSer199), tau (pThr231), and tau (pSer404) were but did not show any significant effect on tau (pSer214) (Fig. 6A-F). Different from the hippocampus, PLTP overexpression only inhibited the high expression of tau (pSer404) in the cerebral cortex of 3×Tg-AD mice (Fig. 6A, G-K).

Fig. 6

PLTP overexpression inhibited hyperphosphorylation of tau protein in 3×Tg-AD mice. A) Representative images by western blotting for the expression of tau and pTau in the cerebral cortex and hippocampal of WT or 3×Tg-AD mice treated with AAV-null or AAV-PLTP. B-K) Quantitative analysis of tau and the ratio of pTau and tau. The values shown are means±SEM, normalized by GAPDH to WT mice treated with AAV-null. ^###p < 0.001, ^##p < 0.01, ^#p < 0.05 versus WT mice plus AAV-null; ^*p < 0.05 versus 3×Tg-AD mice with AAV-null; n = 6.

PLTP overexpression promoted the phosphorylation of GSK3β at serine 9 in 3×Tg-AD mice

To further reveal the mechanism of PLTP overexpression on Aβ metabolism and tau phosphorylation, GSKβ and phosphorylation of GSK3β at serine 9 (pGSK3β_Ser9) and Tyrosine 216 (pGSK3β_Tyr216) were detected. As a result, the ratio of pGSK3β_Ser9 and GSK3β was declined dramatically in both the cerebral cortex and hippocampus of 3×Tg-AD mice with AAV-null compared with WT mice; these changes were relieved by PLTP overexpression (Fig.7A-C). On the other hand, the increase of the ratio of pGSK3β_Tyr216 and GSK3β was shown only in hippocampus of 3×Tg-AD mice with AAV-null compared with WT mice; that was reversed by PLTP overexpression (Fig. 7A, D, E).

Fig. 7

PLTP overexpression promoted the phosphorylation of GSK3β at serine 9 in 3×Tg-AD mice. A) Representative images by western blotting for the expression of GSKβ, phosphorylation of GSK3β at serine 9 (GSK3β (pSer9)) and at tyrosine 216 (GSK3β (pTyr216)) in the cerebral cortex and hippocampal of WT or 3×Tg-AD mice treated with AAV-null or AAV-PLTP. B-E) Quantitative analysis of the ratio of GSK3β (pSer9)/GSK3β and GSK3β (pTyr216)/GSK3β. The values shown are means±SEM, normalized by GAPDH. ^##p < 0.01 versus WT mice plus AAV-null; ^*p < 0.05 versus 3×Tg-AD mice with AAV-null; n = 6.

DISCUSSION

PLTP belongs to the lipid transfer glycoprotein family and is widely distributed in the brain. Numerous recent studies have found that PLTP is involved in the etiology of AD [8, 30]. Nonetheless, there is a need for an in-depth study of the biological effects of PLTP on AD and its mechanisms. Our study demonstrated that PLTP could improve learning and memory of 3×Tg-AD mice, as a validated transgenic model of AD. This is likely through inhibition of the production of toxic Aβ and hyperphosphorylation of tau protein via the promotion of GSK3β inactivation.

3×Tg-AD mice is one of the most advanced preclinical tools available and has been widely employed in the study of mechanisms underlying AD and the development of treatment [31]. This is because it possesses both amyloid plaques and NFT-like pathology in a progressive and age-dependent manner. So, 3×Tg-AD mice were used in this study to examine relationship between PLTP and Aβ metabolism and tau protein phosphorylation. Interestingly, we observed the downregulation of PLTP for the first time in the 3×Tg-AD mice aged 11 months which presented with the decline in cognitive function and the pathological changes of plaques and NFT as other studies in the literature had reported [32].

Recombinant AAV vectors, as a delivery tool for gene transduction in vitro and in vivo studies, has grown by leaps and bounds of gene therapy in recent years [33]. Presently, thousands of serotypes of AAV have been found, some of which are also used in gene delivery to the CNS. In the present study, AAV2/9, a more effective serotype transducing neurons and glial cells in multiple regions of the brain [24, 25], were utilized in mediating the PLTP gene transduction. Results showed that AAV-mediated PLTP gene transduction significantly increased PLTP level and improved the cognitive impairment of 3×Tg-AD mice. Moreover, our results also showed that PLTP upregulation can improve the characteristic pathological lesion of AD mice, including amyloid plaques and NTF, which were previously uncertain, although several studies support that PLTP deficiency deteriorate AD [15 –17]. So, this further supported PLTP protects AD and might be as a target for AD treatment.

Aβ, as a component of senile plaques, is one of the key molecules in the pathology of AD, which is generated by sequential cleavages of AβPP [34]. Previous studies have proved that PLTP deficiency significantly promotes Aβ generation by upregulating AβPP and its secretase (PS1 and BACE1) [15]. Here, we found that PLTP overexpression downregulated the level of Aβ₄₀ and Aβ₄₂ of the hippocampus, and only Aβ₄₂ in the cerebral cortex in 3×Tg-AD mice. This suggests that the role of PLTP in the cerebral cortex and hippocampus might be slightly different. However, further studies are needed to elucidate whether the hippocampus more sensitive to PLTP than the cerebral cortex. Moreover, PLTP overexpression decreased the AβPP and PS1 of the hippocampus and cerebral cortex in 3×Tg-AD mice but did not affect BACE1.This was inconsistent with the data in previous study about PLTP deficiency mice [15]. Thus, PLTP possibly have a stronger regulation on AβPP and PS1 than BACE1 as supported by a study that showed the existence of a cross-linked interaction between AβPP and PLTP [16].

Tau, as a microtubule-associated protein, was believed to be another key factor for AD. However, the neurotoxic one is the abnormally phosphorylated tau, which is the main component in paired helical filaments, and thus, NFTs [32]. Previous studies only confirmed that the recombinant PLTP inhibited tau phosphorylation in vitro [19]. In the present study, we proposed that PLTP inhibited tau phosphorylation in vivo, and 3×Tg-AD mice were considered to be the best animal model to simulate human AD. These proved that the anti-AD effect of PLTP is associated with not only the regulation of Aβ metabolism but also inhibition of tau phosphorylation. Tau protein can be phosphorylated at multiple sites; the tau (pSer199) and tau (pThr231) are among the most critical phosphorylation sites leading to a transformation from tau protein into an inhibitory molecule that sequesters normal microtubule-associated proteins [35]. Modification of phosphorylation in tau-T231 may play a role in formation of NFTs in the AD brain, the phosphorylation at tau-S404 is credited as one of the earliest molecular targets in AD [36], and the phosphorylation at tau-S214 was found to prominently stain in intraneuronal and extracellular NFTs, and involve the information of NFTs [37]. Consistent with our results, a previous study in human neuronal cells found that PLTP could reduce tau phosphorylation at threonine 231 and serine 199 [19]. In this study, we found PLTP to also inhibit the tau phosphorylation at serine 404, but had little effect on tau phosphorylation at serine 214, suggesting that PLTP could be having a protective effect in the early stage of AD. Furthermore, the tissue-specific changes were found in the inhibition of tau phosphorylation, and this was more pronounced in the hippocampus than in the cerebral cortex. Thus, it is possible that the hippocampus is more sensitive to PLTP than the cerebral cortex. This finding could be attributed to the distribution of PLTP. A prior study suggested a higher PLTP mRNA expression in the hippocampus than that in the cortex in the human brain [8]. However, the lack of direct evidence in mice calls for further studies. Meanwhile, more researchers believe that AD is a multifactor disease and a targeted treatment at one aspect alone might not confer vivid advantages [38]. Hence, PLTP is a promising treatment for AD because of the regulation of both Aβ and tau.

Finally, we preliminarily investigated the regulation mechanism of PLTP on the Aβ metabolism and tau phosphorylation. Previous studies suggested that recombinant PLTP could promote the inactivation of GSK3β and inhibit tau phosphorylation [19]. GSK3β, a serine/threonine protein kinase, has been reported to be involved in the pathogenesis of AD [20]. GSK3β contributes to neurodegeneration by various mechanisms [20]. It was originally thought that GSK3β could promote abnormal hyperphosphorylation of tau protein and accelerate the AD pathology process [39]. Concurrently, AβPP and PS1, as GSK3-β substrates, were reported to cause higher production and subsequent deposition of Aβ in the AD brain [40, 41]. GSK3-β function can be regulated by phosphorylation and de-phosphorylation on different sites. For instance, auto-phosphorylation on tyrosine 216 mediates GSK3-β (GSK3β (pTyr216)) activation, while phosphorylation on serine sites 9 (GSK3β (pSer9)) leads to its inhibition [42]. So here, GSK3β, GSK3β (pTyr216) and GSK3β (pSer9) were deleted. We found that PLTP overexpression upregulated the inactive form of GSK3β-GSK3β (pSer9) without affecting GSK3β in both the cerebral cortex and the hippocampus of 3×Tg-AD mice. However, GSK3β active form-GSK3β (pTyr216) was found the downregulation induced by PLTP overexpression only in the hippocampus of 3×Tg-AD mice. These results suggested that the regulation of PLTP on the Aβ production and tau hyperphosphorylation could be mediated by GSK3β. Moreover, the effect of PLTP is different in cerebral cortex and hippocampus, which has been proved again. In addition, more mechanisms may be involved in the effect of PLTP’ anti-AD, especially lipid metabolism, because PLTP is a lipid metabolism related protein after all; however, there is no in-depth study on how PLTP regulates lipid metabolism in the CNS. And it is believed that with the unveiling of the veil of AD, especially after the application of proteomics [43] and transcriptomics [44], the anti-AD mechanism of PLTP will gradually become clear.

In summary, this study has provided a solid demonstration that PLTP overexpression can recover cognition dysfunction in 3×Tg AD mice by regulating the Aβ process and tau protein phosphorylation. This process is most likely induced by the inactivation of GSK3β. And compared with the cerebral cortex, PLTP plays a stronger role in the hippocampus. Therefore, PLTP possesses a neuroprotective effect and can be a therapeutic target for AD (Fig. 8).

Fig. 8

The possible mechanisms whereby PLTP overexpression improves AD. PLTP overexpression decreases the production of toxic Aβ₄₀ and Aβ₄₂ by inhibiting AβPP and PS1, and downregulates tau Phosphorylation at Ser 214, Thr231, and Ser 404, most likely via increasing the phosphorylation of GSK3β at serine 9 to induce its inactivation. Therefore, PLTP possess neuroprotective effect, and can be a therapeutic agent or target for AD in the further.

Footnotes

ACKNOWLEDGMENTS

This work was supported by grants from the Natural Science Foundation of China (81441111, 81601229) and the Key R & D Plan of Shandong Province (2019GSF108037), Academic Promotion Programme of Shandong First Medical University (2019LJ003 and 2019QL011), and Taishan Scholars Foundation of Shangdong Province (ts201511057). The authors thank Nikoli Peacher for language editing.

Authors’ disclosures available online ().

References

Malik

(2017) Treatments in Alzheimer’s disease. J Neurol 264, 416–418.

Soria Lopez

, González

, Léger

(2019) Alzheimer’s disease. Handb Clin Neurol 167, 231–255.

Alzheimer’s Association (2016) 2016 Alzheimer’s disease facts and figures. Alzheimers Dement 12, 459–509.

Tam

, Pasternak

(2012) Amyloid and Alzheimer’s disease: Inside and out. Can J Neurol Sci 3, 286–298.

Gong

, Grundke Iqbal

(2010) Targeting tau protein in Alzheimer’s disease. Drugs Aging 27, 351–365.

Albers

, Vuletic

, Cheung

(2012) Role of plasma phospholipid transfer protein in lipid and lipoprotein metabolism. Biochim Biophys Acta 1821, 345–357.

Chowaniec

, Skoczyńska

(2018) Plasma lipid transfer proteins: The role of PLTP and CETP in atherogenesis. Adv Clin Exp Med 27, 429–436.

Vuletic

, Jin

, Marcovina

, Peskind

, Moller

, Albers

(2003) Widespread distribution of PLTP in human CNS: Evidence for PLTP synthesis by glia and neurons, and increased levels in Alzheimer’s disease. J Lipid Res 44, 1113–1123.

Vuletic

, Kennedy

, Albers

, Killestein

, Vrenken

, Lütjohann

, Teunissen

(2014) Cerebrospinal fluid apolipoprotein E and phospholipid transfer protein activity are reduced in multiple sclerosis; relationships with the brain MRI and CSF lipid variables. Mult Scler Relat Disord 3, 533–541.

10.

Krapfenbauer

, Yoo

, Kim

, Cairns

, Lubec

(2001) Differential display reveals downregulation of the phospholipid transfer protein (PLTP) at the mRNA level in brains of patients with Down syndrome. Life Sci 68, 2169–2179.

11.

Kuerban

, Shibata

, Komatsu

, Ohnuma

, Arai

(2010) Genetic association between PLTP gene polymorphisms and Alzheimer’s disease in a Japanese population. Dement Geriatr Cogn Disord 30, 78–82.

12.

Vuletic

, Peskind

, Marcovina

, Quinn

, Cheung

, Kennedy

, Kaye

, Jin

, Albers

(2005) Reduced CSF PLTP activity in Alzheimer’s disease and other neurologic diseases; PLTP induces ApoE secretion in primary human astrocytes in vitro. J Neurosci Res 80, 406–413.

13.

Desrumaux

, Mansuy

, Lemaire

, Przybilski

, Le Guern

, Givalois

, Lagrost

(2018) Brain vitamin E deficiency during development is associated with increased glutamate levels and anxiety in adult mice. Front Behav Neurosci 12, 310.

14.

Desrumaux

, Risold

, Schroeder

, Deckert

, Masson

, Athias

, Laplanche

, Le Guern

, Blache

, Jiang

, Tall

, Desor

, Lagrost

(2005) Phospholipid transfer protein (PLTP) deficiency reduces brain vitamin E content and increases anxiety in mice. FASEB J 19, 296–297.

15.

Wang

, Yu

, Chen

, Cui

, Luo

, Ma

, Jiang

, Qin

(2014) PLTP deficiency impairs learning and memory capabilities partially due to alteration of amyloid-β metabolism in old mice. J Alzheimers Dis 39, 79–88.

16.

Tong

, Sun

, Tian

, Zhou

, Wang

, Zhang

, Zhan

, Zhao

, Kuerban

, Li

, Wang

, Jin

, Fan

, Guo

, Han

, Qin

, Chui

(2015) Phospholipid transfer protein (PLTP) deficiency accelerates memory dysfunction through altering amyloid precursor protein (APP) processing in a mouse model of Alzheimer’s disease. Hum Mol Genet 24, 5388–5403.

17.

Desrumaux

, Pisoni

, Meunier

, Deckert

, Athias

, Perrier

, Villard

, Lagrost

, Verdier

, Maurice

(2013) Increased amyloid-β peptide-induced memory deficits in phospholipid transfer protein (PLTP) gene knockout mice. Neuropsychopharmacology 38, 817–825.

18.

Mansuy

, Baille

, Canet

, Borie

, Cohen-Solal

, Vignes

, Perrier

, Chevallier

, Le Guern

, Deckert

, Lagrost

, Givalois

, Desrumaux

(2018) Deletion of plasma Phospholipid Transfer Protein (PLTP) increases microglial phagocytosis and reduces cerebral amyloid-β deposition in the J20 mouse model of Alzheimer’s disease. Oncotarget 9, 19688–19703.

19.

Dong

, Albers

, Vuletic

(2009) Phospholipid transfer protein reduces phosphorylation of tau in human neuronal cells. J Neurosci Res 87, 3176–3185.

20.

Lauretti

, Dincer

, Praticò

(2020) Glycogen synthase kinase-3 signaling in Alzheimer’s disease. Biochim Biophys Acta Mol Cell Res 1867, 118664.

21.

DaRocha-Souto

, Coma

, Pérez-Nievas

, Scotton

, Siao

, Sánchez-Ferrer

, Hashimoto

, Fan

, Hudry

, Barroeta

, Serenó

, Rodríguez

, Sánchez

, Hyman

, Gómez-Isla

(2012) Activation of glycogen synthase kinase-3 beta mediates β-amyloid induced neuritic damage in Alzheimer’s disease. Neurobiol Dis 45, 425–437.

22.

Zhang

, Huang

, Yan

, Jin

, Zhou

, Shi

, Jin

(2018) Diabetes mellitus and Alzheimer’s disease: GSK-3β as a potential link. Behav Brain Res 339, 57–65.

23.

Zhu

, Zhang

, Zhao

, Shu

, Wang

, Yang

, He

, Wang

, Ying

(2018) GSK3β-mediated tau hyperphosphorylation triggers diabetic retinal neurodegeneration by disrupting synaptic and mitochondrial functions. Mol Neurodegener 13, 62.

24.

Gholizadeh

, Tharmalingam

, Macaldaz

, Hampson

(2013) Transduction of the central nervous system after intracerebroventricular injection of adeno-associated viral vectors in neonatal and juvenile mice. Hum Gene Ther Methods 24, 205–213.

25.

Hammond

, Leek

, Richman

, Tjalkens

(2017) Cellular selectivity of AAV serotypes for gene delivery in neurons and astrocytes by neonatal intracerebroventricular injection. PLoS One 12, e0188830.

26.

Wang

, Zhang

, Xu

, Fu

, Wang

, Chen

, Xu

, Gao

, Zhang

(2020) The phosphodiesterase-4 inhibitor roflumilast, a potential treatment for the comorbidity of memory loss and depression in Alzheimer’s disease: A preclinical study in APP/PS1 transgenic mice. Int J Neuropsychopharmacol 23, 700–711.

27.

Zhang

, Yu

, Zhang

, Chen

, Dong

, Lu

, Song

, Zhou

(2018) Porphyromonas gingivalis lipopolysaccharide induces cognitive dysfunction, mediated by neuronal inflammation via activation of the TLR4 signaling pathway in C57BL/6 mice. J Neuroinflammation 15, 37.

28.

Jung

, Park

, Lee

, Park

, Kong

, Rho

, Hur

, Yang

, Park

(2015) Development of fluorescent probes that bind and stain amyloid plaques in Alzheimer’s disease. Arch Pharm Res 38, 1992–1998.

29.

, Dong

, Gong

, Wu

, Jin

, Shi

(2015) Icariin decreases both APP and Aβ levels and increases neurogenesis in the brain of Tg2576 mice. Neuroscience 24, 29–35.

30.

Demeester

, Castro

, Desrumaux

, De Geitere

, Fruchart

, Santens

, Mulleners

, Engelborghs

, De Deyn

, Vandekerckhove

, Rosseneu

, Labeur

(2000) Characterization and functional studies of lipoproteins, lipid transfer proteins, and lecithin: Cholesterol acyltransferase in CSF of normal individuals and patients with Alzheimer’s disease. J Lipid Res 41, 963–974.

31.

Oddo

, Caccamo

, Shepherd

, Murphy

, Golde

, Kayed

, Metherate

, Mattson

, Akbari

, LaFerla

(2003) Triple-transgenic model of Alzheimer’s disease with plaques and tangles: Intracellular Abeta and synaptic dysfunction. Neuron 39, 409–421.

32.

Ontiveros-Torres

MÁ

, Labra-Barrios

, Díaz-Cintra

, Aguilar-Vázquez

, Moreno-Campuzano

, Flores-Rodríguez

, Luna-Herrera

, Mena

, Perry

, Florán-Garduño

, Luna-Muñoz

, Luna-Arias

(2016) Fibrillar amyloid-β accumulation triggers an inflammatory mechanism leading to hyperphosphorylation of the carboxyl-terminal end of tau polypeptide in the hippocampal formation of the 3×Tg-AD transgenic mouse. J Alzheimers Dis 52, 243–269.

33.

Naso

, Tomkowicz

, Perry

, 3rd, Strohl

(2017) Adeno-Associated virus (AAV) as a vector for gene therapy. BioDrugs 31, 317–334.

34.

Siddiqi

, Majid

, Malik

, Alam

, Khan

(2019) Amyloid oligomers, protofibrils and fibrils. Subcell Biochem 93, 471–503.

35.

Foidl

, Humpel

(2018) Differential hyperphosphorylation of Tau-S199, -T231 and -S396 in organotypic brain slices of Alzheimer mice. A model to study early Tau hyperphosphorylation using okadaic acid. Front Aging Neurosci 10, 113.

36.

Chong

, Ng

, Koh

, Chye

(2018) Tau proteins and tauopathies in Alzheimer’s disease. Cell Mol Neurobiol 38, 965–980.

37.

Augustinack

, Schneider

, Mandelkow

, Hyman

(2002) Specific tau phosphorylation sites correlate with severity of neuronal cytopathology in Alzheimer’s disease. Acta Neuropathol 103, 26–35.

38.

Yang

, Liu

, Wu

, Kou

, Shen

(2021) A review on α-mangostin as a potential multi-target-directed ligand for Alzheimer’s disease. Eur J Pharmacol 15, 173950.

39.

Hernandez

, Lucas

, Avila

(2013) GSK3 and tau: Two convergence points in Alzheimer’s disease. J Alzheimers Dis 33, S141–S144.

40.

, Wu

, Zou

, Wang

, Zhou

, Kinoshita

, Zhang

, Yang

, Cai

, Woodgett

, Song

(2013) Inhibition of GSK3β-mediated BACE1 expression reduces Alzheimer-associated phenotypes. J Clin Invest 123, 224–235.

41.

Twomey

, McCarthy

(2006) Presenilin-1 is an unprimed glycogen synthase kinase-3beta substrate. FEBS Lett 580, 4015–4020.

42.

Pajak

, Songin

, Strosznajder

, Gajkowska

(2009) Alzheimer’s disease genetic mutation evokes ultrastructural alterations: Correlation to an intracellular Abeta deposition and the level of GSK-3beta-P (Y216) phosphorylated form. Neurotoxicology 30, 581–588.

43.

Zhou

, Wang

, Xiao

, Su

, Zheng

, Zhang

, Wang

, Xu

, Yang

, Hoi

MPM

(2019) Memantine improves cognitive function and alters hippocampal and cortical proteome in triple transgenic mouse model of Alzheimer’s disease. Exp Neurobiol 28, 390–403.

44.

Zheng

, Zhang

, Zhao

, Chen

, Jia

, Cao

, Shen

, Ni

, Song

(2019) Transcriptomic insights into the response of the olfactory bulb to selenium treatment in a mouse model of Alzheimer’s disease. Int J Mol Sci 19, 2998.