Increased expression of the neuroplastin 65 protein is involved in neurofibrillary tangles and amyloid beta plaques in Alzheimer’s disease

Abstract

BACKGROUND

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder currently lacking effective therapeutic interventions. Pathological hallmarks of AD include intracellular neurofibrillary tangles (NFTs) and extracellular amyloid beta (Aβ) plaques. Neuroplastin 65 (NP65), highly expressed in the brain, has been previously shown to mitigate cognitive impairments and decrease Aβ plaques in the AD mouse model, suggesting that NP65 is involved in AD neuropathology. However, direct evidence linking NP65 expression to AD pathogenesis in human brain remains absent.

AIM

To quantify NP65 isoform expression gradients across distinct neuroanatomical regions in the healthy brain and investigate the alterations of NP65 expression in the AD brain.

METHODS

Immunohistochemical, immunofluorescent and western blot analyses were used to investigate NP65 expression in 19 postmortem brains (AD = 10, controls = 9). Double immunostaining with 6E10 and or phosphorylated-microtubule-associated protein tau (AT-8, a marker for NFT) markers was performed to assess NP65 colocalization with Aβ plaques and NFTs.

RESULTS

In controls, NP65 was highly expressed in a wide-range of brain areas. AD cases showed significantly increased NP65 immunoreactivity across multiple brain regions, including the frontal and temporal cortex, hippocampus, and cerebellum, compared to controls. Western blot analysis consistently confirmed significantly elevated NP65 expression in the hippocampus of AD patients relative to controls. Double immunostaining demonstrated partial colocalization of NP65 with NFTs and Aβ plaques in AD brain tissue.

CONCLUSION

Our findings demonstrate a significant increase of NP65 protein, which colocalizes with NFTs and Aβ plaques in AD brains, providing direct evidence supporting a critical role of NP65 expression in the neuropathological mechanisms of this disease.

Key Words: Neuroplastin 65 protein; Alzheimer’s disease; Neurofibrillary tangles; Amyloid beta; Hippocampus

Core Tip: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder, with pathological hallmarks including intracellular neurofibrillary tangles and extracellular amyloid beta (Aβ) plaques. Synaptic membrane glycoprotein neuroplastin 65 (NP65) has been previously shown to mitigate cognitive impairments and decrease Aβ plaques in the AD mouse model. However, direct evidence linking NP65 expression to AD pathology in human brain remains absent. The present study shows that NP65 is highly expressed in a wide-range of brain areas and is significantly increased in AD cases. In addition, we found colocalization of NP65 with neurofibrillary tangles and Aβ plaques in AD brain. Thus, this study provides direct evidence of NP65 involvement in AD pathology.

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disease that is clinically characterized by progressive cognitive deficits. The number of AD patients will rise to 82 million by 2050, and become a global crisis[1]. Clinically, the principal manifestations of AD patients are the progressive decline of memory, language and mental disorders, and the progressive weakening of cognition, including temporal and spatial judgment, and finally dementia[2]. Whole brain atrophy occurs, which is characterized by a smaller volume and thinner cortex, all of which are accompanied by amyloid beta (Aβ) protein deposits throughout the brain neuropil. Other histopathological features of AD include neurofibrillary tangles (NFTs) caused by hyperphosphorylation of tau protein within neurons, diffuse inflammatory necrotic foci, and loss of neurons [3]. Available therapies, such as cholinesterase blockers and N-methyl-D-aspartate receptor agonists, can only relieve symptoms but cannot block the progress of AD. Recently, several amyloid-binding antibodies have been approved for patient treatment in the United States (e.g., lecanemab, donanemab, and aducanumab), and clinical trials are underway on other Aβ binding antibodies and immunization protocols[4]. Therefore, targeting the study of Aβ production, deposition and degradation is helpful in developing new therapeutic methods [5]. Interestingly, several studies have shown that certain cell adhesion molecules (CAMs), such as L1 and neuroplastin (NP) 65, are involved in the formation of Aβ plaques in APP/PS1 mice (a commonly used AD model of cerebral amyloidosis)[6-8], indicating a potential role of CAMs in the pathogenesis of AD.

The NP isoforms NP65 and NP55 are encoded by a single gene (NPTN in humans, Nptn in rodents) and result from alternative splicing of the mRNA [9]. Both isoforms are single-spanned transmembrane proteins belonging to the immunoglobulin superfamily with two (NP55) and three (NP65) Ig domains, respectively. NP65 and NP55 are solely distinguished by the absence of the Ig1 module in NP55 but the presence in NP65[9]. In rodents, NP55 is widely localized in various tissues, whereas NP65 is mainly expressed in the brain and retina[10-13]. In the human brain, NP65 is strongly expressed in the cerebral cortex, striatum, hippocampus, hypothalamus and cerebellum[10,14]. Notably, NP55 is the major isoform in the rodent cerebellum, but NP65 is the major isoform in the human cerebellum[10,15]. Functionally, NP65, as a synaptic glycoprotein, is related to synaptic plasticity, and involved in neuropsychiatric diseases[12,16]. Our serial reports have shown that NP65 is related to anxiety in mice, as NP65 deficient mice clearly display anxiety-like behaviors, in several paradigms including the hole-board test and marble burying test, open-field test and light-dark transition test[17-19]. Additionally, a single nucleotide polymorphism in the NP locus is reported to be associated with cortical thickness and intellectual ability in adolescents[20]. Furthermore, our previous studies have shown that NP65 deficiency improves learning and memory in wild-type mice[17,18]. Therefore, these reports suggest a role for NP65 in neuropsychiatric disease.

Our recent studies have shown that NP65 deficiency alleviates cognitive deficits and decreases Aβ plaque load and neuroinflammation in APP/PS1 mice[8]. Given that NP65 is highly expressed in the cerebral cortex, hippocampus and cerebellum in the human brain, it is necessary to explore whether NP65 is involved in the pathogenesis of AD. Previously, NP65 immunoreactivity was only investigated in the hippocampus of AD patients, showing an increase in NP65 immunoreactivity at the early stage but not the late stage of AD when compared to controls[21]. Therefore, it is necessary to further determine whether NP65 is involved in the pathogenesis of AD. Here, we thoroughly investigated the expression and distribution of NP65 in postmortem human brain and further investigated whether the expression of NP65 is altered in distinctive regions of the AD brain, such as the hippocampus, frontal and temporal cortex, and cerebellum. Interestingly, we found a significant increase in NP65 in a wide range of brain regions in AD and partial co-localization of NP65 with Aβ plaques and NFTs in the AD brain. These results provide direct evidence that NP65 is involved in the pathogenesis of AD, indicating a new potential therapeutic target for AD.

MATERIALS AND METHODS

Human postmortem brain and tissue preparation

Human postmortem brains were obtained from the NeuroBioBank of Central South University, which receives willed body donations with consent[22]. This NeuroBioBank is supported by the Red Cross Society of China and provides cadavers for medical education and research. In total ten cases were included in the AD group, of which six cases were definitely diagnosed as AD according to their medical records, and the other four cases were recorded as dementia and clinically regarded as AD patients. Furthermore, using the specific 6E10 antibody against Aβ plaques and the phosphorylated-microtubule-associated protein tau (p-Tau) antibody against NFTs[3], all ten brains were confirmed as AD cases with NFTs and Aβ plaques (Supplementary Figures 1 and 2). Age-matched (controls: 72.33 ± 10.33 years, AD: 79 ± 7.85 years, t = -0.109, ns) cases of older patients who did not die of neurological diseases, including cancer and cardiovascular diseases, were included as controls and were also confirmed by the 6E10 and p-Tau antibodies to have negative results (Supplementary Table 1). The present study was approved by the Ethics Committee for Research and Education at Xiangya School of Medicine in 2020, approval No. 2020KT-37, in compliance with the Code of Ethics of the World Medical Association (Declaration of Helsinki).

The brain was dissected within 15 hours after death. One half of the brain was freshly frozen at -80 °C for western blot or other molecular biology detection, while the other half was immersed in neutral formalin for 20 days for section staining. To prepare brain sections, the cerebral hemisphere and hippocampus were cut coronally into slices (1.0 cm thick); the cerebellum was sagittally cut into slices (1.0 cm thick); and the brainstem was cut horizontally into slices (0.5 cm thick). These brain slices were then placed in 15%, and 30% sucrose in 0.01 mol phosphate buffer (PBS) until they sank to the bottom at -4 °C. They were then cut into sections (20 μm thick) using a cryostat (CM 1950, Leica, Heidelberg, Germany). These sections were stored in a cryoprotectant at -20 °C before use.

Immunohistochemical staining

The procedures for human brain sectioning were described previously and modified[23,24]. In brief, after being rinsed with PBS, the human brain sections were pretreated with 5.0% hydrogen peroxide in methanol for 30 minutes at room temperature and then washed with PBS to neutralize endogenous peroxidase activity. To expose the antigen, the sections were treated with 95% formic acid for 10 minutes (for the 6E10 antigen), or treated with 2 × saline sodium citrate buffer in a water bath at 65 °C for 60 minutes (for the NP65 antigen). The sections were then treated with 0.01M Tris-HCI for 10 minutes and immersed in a blocking solution (5% horse serum and 0.5% Triton X-100 in PBS) on a shaking rocker for 2 hours at room temperature. Finally, the sections were incubated with primary antibodies, including mouse anti-6E10 (1:1000, Abcam, United States), rabbit anti-p-Tau (1:1000, Abcam, United States) and goat anti-NP65 (1:200, R and D Systems, United States) at 4 °C overnight (Table 1). The sections were then incubated with biotinylated horse anti-goat secondary immunoglobulin G (IgG) or goat anti-mouse/rabbit IgG (1:200, Vector Laboratories, United States) for 1 hour and further with an avidin-biotin-horseradish peroxidase complex kit (1:200, Vector Laboratories, United States) for 2 hours at room temperature. Immunoreactivity was visualized using 0.03% hydrogen peroxide and 0.05% 3, 3’-diaminobenzidine. The sections were dehydrated with ascending ethanol, cleared with xylene, and covered with a mounting medium. Negative controls included pre-immune serum replacing the primary antibody, and omission of the secondary antibody.

Table 1 Antibodies and kits used in this study.

Antibody	Manufacturer	Product code	Host species	Dilution
Htr3A	Abcam	ab13897	Goat	1:300
Aβ	Novus	2-62566	Mouse	1:500
APP	Abcam	ab32136	Rabbit	1:300
BACE-1	Santa	AF10748	Rabbit	1:200
p-Tau	Abcam	ab62639	Mouse	1:1000
594-conjugated affinipure anti-goat IgG	Jackson	705-585-003	Donkey	1:500
488-conjugated affinipure donkey anti-rabbit IgG	Jackson	711-545-152	Donkey	1:500
488-conjugated affinipure donkey anti-mouse IgG	Jackson	715-545-150	Donkey	1:500
Horse anti-mouse/rabbit/goat IgG antibody	Vector	BA-1300-2.2	Horse	1:500
ABC-HRP kit	Vector	PK-6100	-	1:500

Htr3A: Hydroxytryptamine receptor 3A; Aβ: Amyloid beta; BACE-1: β-site amyloid-precursor-protein-cleaving enzyme 1; p-Tau: Phosphorylated-microtubule-associated protein tau; IgG: Immunoglobulin G; ABC: Avidin-biotin-horseradish peroxidase complex; HRP: Horseradish peroxidase.

Immunofluorescent staining

In addition, immunofluorescent staining of NP65 alone or with 6E10 or p-Tau (AT-8, a marker for NFTs) was performed according to a previous report[24]. For this staining, the sections were treated with a 95% sodium citrate solution (pH 6.0) in a water bath at 65 °C for 60 minutes and then cooled to room temperature. After being blocked with 7% donkey serum in 0.1% Triton X-100 for 3 hours at room temperature, the sections were incubated with primary antibodies, including goat-anti NP65 or antibody cocktails: NP65 with 6E10 or p-Tau overnight at 4 °C. After thoroughly washing with PBS, the sections were incubated in fluorescein isothiocyanate (FITC)-conjugated or Cy3-conjugated anti-goat or anti-mouse and anti-rabbit secondary IgG (1:1000, Jackson Immuno-Research Laboratories, United States) for 2 hours at room temperature, and counterstained with diamidino-phenyl-indole (Beyotime Biotechnology, Shanghai, China). Primary and secondary antibodies were diluted with 7% donkey serum in 0.1% Triton X-100 in 0.01 mol PBS. All sections were treated with 0.1% Sudan black to block autofluorescence before being covered with VECTASHIELD^® antifading mounting media (Vector Laboratories, United States). A negative control was set up by omitting the primary antibody.

Visualization and quantification of immunostaining

For semi-quantification of NP65 expression, two blind investigators performed the analysis of immunohistochemical and immunofluorescent staining, respectively. The images were captured under a 200 × or 400 × magnification using a microscope (Olympus FV1000, Tokyo, Japan) and semi-quantitatively analyzed with ImageJ software v1.51 (National Institutes of Health, Bethesda, MD, United States). The optical density (OD) of NP65 immunoreactivity was determined by subtracting the background of the designated region of interest (ROI) using ImageJ software v1.51 (National Institutes of Health, Bethesda, MD, United States). In immunohistochemical analysis, the upper 40 and lower 255 threshold were set to subtract the background of ROI; in immunofluorescence analysis, the upper 75 and lower 255 threshold were set to subtract the background of ROI. The averaged OD of 3 selected ROIs was used as the OD of each individual. Signal intensity per field was then calculated. The immunoreactivity data are derived from three cases with three slides of each individual.

Western blot analysis

For detecting levels of NP65 protein in the hippocampus from AD (n = 3) and controls (n = 3), the block of hippocampus (15-20 g) was taken out from the -80 °C refrigerator and rinsed with PBS. It was then cut into pieces. The cell membrane protein of the hippocampus was extracted using a membrane protein extraction kit (Beyotime, China), according to the manufacturer’s instructions. In brief, the tissues were lysed with reagent A with 1 mmol/L phenylmethylsulfonyl fluoride buffer in an ice bath for 10 minutes. They were then centrifuged at 14000 × g at 4 °C for 10 minutes and the supernatant was collected. The supernatant was further centrifuged at 700 × g at 4 °C for 30 minutes. The pellet was collected and dissolved with reagent B in an ice bath for 10 minutes. It was then centrifuged at 14000 × g at 4 °C for 5 minutes. The supernatant was the membrane protein and was collected for protein detection. After determining the protein concentration, 15 μg of protein samples were electrophoretically separated in 10% sodium dodecyl sulfate polyacrylamide gels and transferred to polyvinylidene fluoride membranes. These membranes were blocked with 5% bovine serum albumin in Tris-buffered saline containing 0.05% Tween-20 for 1 hour at room temperature. They were then incubated overnight at 4 °C with goat anti-NP65 (1:1000, R and D Systems, Minnesota, United States) and rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (1:2000, Beyotime, Shanghai, China). The membranes were then washed and incubated with anti-rabbit IgG-horseradish peroxidase and anti-goat IgG-horseradish peroxidase (1:1000, Beyotime) for 2 hours at room temperature. Immunoreactive bands were visualized and images were analyzed quantitatively using ImageJ software (NIH, Bethesda, United States). Targeted protein levels were normalized to those of glyceraldehyde-3-phosphate dehydrogenase from two separate experiments.

Statistical analysis

Statistical analyses were carried out using SPSS 22.0 (IBM SPSS, Chicago, IL, United States) for Windows. All data were expressed as mean ± SEM. An unpaired Student’s t-test was used to evaluate the differences between two groups. Statistical significance was set at P < 0.05.

RESULTS

Distribution and cellular location of NP65 in human brain

Although the distribution of NP65 in some regions of the human brain has been reported[10,14,21], detailed information on NP65 in the entire brain remains unclear. In the present study, we first investigated the distribution of NP65 in the aged human brain (from individuals who died of no neurological or psychiatric disorders), as controls for AD cases. In the hippocampal formation, using both immunohistochemistry and immunofluorescence staining, we found that NP65 expression was most abundant in the dentate gyrus (DG) and cornu ammonis (CA) 1 regions, while moderate in the CA2-3 regions of the hippocampus and the hilar part of the DG, as demonstrated by statistical analysis (Figure 1). This distribution is similar to that in the rodent hippocampus[12], whereas it is different to a previous report showing that CA1-CA4 exhibited a rather uniform NP65 immunostaining in the human hippocampus[10]. In the CA1, prominent NP65 staining was found in the stratum pyramidale, moderate staining in the stratum radiatum, and faint staining in the stratum oriens (Figure 1A-E). This distribution differs to that in the rodent brain, as it was reported that strong NP65 immunoreactivity was found in the stratum oriens and stratum radiatum of the rodent CA1[12]. In the DG, the most intense immunoreactivity was found in the stratum granulosum (SG) and stratum moleculare and less intense in the stratum plexiforme (Figure 1A-E), similar to previous reports in rodent and human brains[12,14]. At the cellular level, NP65 immunoreactivity was highly localized on the cell membranes of granular and pyramidal neurons and in the neuropil (Figure 1C and H). Strikingly, NP65 immunoreactive materials appeared as discrete knots in the cell membranes of granular neurons (Figure 1C and H indicated by arrow), strikingly similar to the NP55 immunoreactivity in Purkinje cell somata of the rat cerebellar cortex[15]. In the stratum pyramidale and the neuropil of the hippocampus, NP65 immunoreactivity showed punctate, plexus and laminar patterns in the CA1 region (Figure 1C and H).

Figure 1 Expression and cellular localization of neuroplastin 65 immunoreactivity in human hippocampal formation. A: Neuroplastin 65 (NP65) immunofluorescent staining in a coronal section of the human hippocampal formation. A representative view of the human hippocampal formation at a low magnification (scale bars = 3 mm); B and C: NP65 immunofluorescent staining in a coronal section of the human hippocampal formation. Representative micrographs of NP65 immunoreactivity in the cornu ammonis 1 (CA1) region and dentate gyrus at a high magnification view. Note the immunoreactive knots (indicated by an arrow) in the stratum granulosum; D and E: Quantification of NP65 immunofluorescent staining in the sublayers of CA1 and DG and hippocampal areas; F: A representative scan of the human hippocampal formation (scale bars = 3 mm); G and H: Representative micrographs of NP65 immunoreactivity in the CA1 region and dentate gyrus at a high magnification view. Note the immunoreactive knots (indicated by an arrow) in the stratum granulosum; I: Quantification of NP65 immunohistochemical staining in hippocampal areas. DG: Dentate gyrus; CA: Cornu ammonis; HIL: Hilus; SUB: Subiculum; EC: Entorhinal cortex; SG: Stratum granulosum; SM: Stratum moleculare; SPl: Stratum plexiforme; SLM: Stratum lacunosum-molecular; SR: Stratum radiatum; SP: Stratum pyramidale; SO: Stratum oriens.

In the cerebral cortex, the overall distribution of NP65 was uneven and showed regional differences, inconsistent with one report showing no obvious regional differences[10]. With regard to the layers of the neocortex, the most intense immunoreactivity was found in layers II and V/VI (Figure 2), while other layers were moderately labelled (Figure 2A, B and D), consistent with the distribution pattern of NP65 in the rat neocortex[12]. In the dorsolateral frontal cortex, the prefrontal cortex (PFC) exhibited prominent immunostaining relative to the precentral gyrus (Figure 2A and D). Notably, prominent NP65 immunoreactivity was localized at the apical dendrites (indicated by an arrow in Figure 2A) and cell membranes of the pyramidal neurons in layer V of the PFC (Figure 2A). In addition, the white matter underlying the cerebral cortex showed no immunoreactivity (Figure 2A and D).

Figure 2 Neuroplastin 65 immunoreactive pattern in the dorsolateral frontal cortex of individuals with Alzheimer’s disease and age-matched controls. A: Representative Neuroplastin 65 (NP65) immunohistochemical staining micrographs of a coronal section in the dorsolateral prefrontal cortex (dPFC) from controls (CON) and Alzheimer’s disease (AD), with an inset indicating the sampling position. Note the strongly positive apical dendrites (indicated by an arrow) in layer V; B: Quantification of NP65 immunoreactivity in sublayers of the dPFC in CON; C: Comparison of NP65 immunoreactivity in layer II and V of the dPFC between AD and CON; D: Representative NP65 immunohistochemical staining micrographs of a coronal section in the precentral gyrus of CON and AD, with an inset indicating the sampling position. Note the strongly positive apical dendrites (indicated by an arrow) in layer V of the AD brain; E: Comparison of NP65 immunoreactivity in layer II and V of the precentral gyrus between AD and CON. The nucleus is stained with hematoxylin. PFC: Prefrontal cortex; CON: Controls; AD: Alzheimer’s disease; PG: Precentral gyrus.

In the lentiform nucleus, the putamen and globus pallidus displayed intense NP65 immunoreactivity (Figure 3A and B), which differs greatly from the distribution pattern in the rat lentiform, showing strong staining in the putamen but faint in the globus pallidus[12]. The high-power view showed that NP65 immunoreactivity was strong and punctate in the putamen, while it formed more dense networks in the globus pallidus (Figure 3A and B). In the superior colliculus of the midbrain, prominent NP65 staining was present in the substantia nigra, moderate staining was present in the superior colliculus, and faint staining was present in the red nucleus and reticular formation. There was no staining in the crus cerebri (Figure 3C). The high-power view showed that NP65 immunostaining was strongly positive in punctate and fibers in the substantia nigra, while there were less intense positive fibers in the red nucleus (Figure 3D). The general NP65 immunoreactivity in the brain stem was obviously faint compared with that in the cerebral hemisphere (Figure 3A-F). In addition, in the basilar part of the pons, a few nerve fibers and the pontine nucleus were moderately stained, while the longitudinal pyramidal tract was devoid of any immunoreaction (Figure 3F).

Figure 3 Expression of neuroplastin 65 immunoreactivity in human lentiform nucleus, midbrain and pons. A: A representative immunohistochemical staining micrograph in the lentiform nucleus, with an inset indicating the position of the sampling (scale bars = 5 mm); B: A representative immunohistochemical staining micrograph in the lentiform nucleus, with an inset indicating the position of the sampling (higher power views, scale bars = 200 μm). Note that neuroplastin 65 immunoreactivity shows strong punctate staining in the putamen, while there are dense networks in the globus pallidus; C: An immunohistochemical staining micrograph in the midbrain at a low magnification view, with an inset indicating the position of the sampling (scale bars = 5 mm); D: An immunohistochemical staining micrograph in the midbrain at a low magnification view, with an inset indicating the position of the sampling (higher power view, scale bars = 200 μm). Note that there is positive punctate staining and fibers in the substantia nigra and less intense positive fibers in the red nucleus; E: A low magnification view of the pons, with an inset indicating the position of the sampling (scale bars = 5 mm); F: A low magnification view of the pons, with an inset indicating the position of the sampling (higher power views, scale bars = 200 μm). The nucleus is stained by hematoxylin. iCtx: Insular cortex; Pu: Putamen; GPe: External part of the globus pallidus; GPi: Internal part of the globus pallidus; IC: Internal capsule; CC: Crus cerebri; SN: Substantia nigra; RN: Red nucleus; SC: Superior colliculus; RF: Reticular formation; BPP: Basilar part of the pons; TP: Tegmentum of the pons; PT: Pyramidal tract; CT: Corticopontine tract; MCP: Middle cerebellar peduncle.

In the cerebellum, NP65 immunostaining was present in the cerebellar cortex, and absent in the medulla. Strikingly similar to a previous postmortem report[10], strong immunoreactivity was found in the granule cell layer (GCL) and the molecular layer (ML) but not in the Purkinje cell layer (PCL) (Figure 4A-C). This distribution is also similar to that of NP55 previously observed in the mouse cerebellum[15]. In the GCL, strong immunoreactivity appeared blocky or granular (indicated by an arrow in Figure 4C), seemingly representing synaptic glomeruli, as previously reported[15].

Figure 4 Neuroplastin 65 immunoreactive pattern in the cerebellum derived from individuals with Alzheimer’s disease and age-matched controls. A: Low-power microphotograph showing the regional distribution of neuroplastin 65 (NP65) immunoreactivity in the cerebellar anterior lobe of controls (CON) and Alzheimer's disease (AD). The inset indicates the position of the sampling (scale bars = 2.5 mm); B: Low-power views in the rectangle frame in Figure 4A showing the regional distribution of NP65 in the cerebellar cortex of CON and AD (scale bars = 200 μm); C: Higher-power views in the ML, PCL and GCL of the rectangle frame in Figure 4B. Note that intense NP65 immunoreactivity is shown in a blocky or granular pattern (indicated by arrows) in the GCL, while it is punctate in the ML. There is a significant increase in immunoreactivity in AD compared with CON. The nucleus is stained with hematoxylin (scale bars = 50 μm); D: Quantification of NP65 immunoreactivity in the cerebellar cortex between AD and CON. CON: Controls; AD: Alzheimer's disease; GCL: Granular cell layer; PCL: Purkinje cell layer; ML: Molecular layer.

Increased expressions of NP65 protein in AD cases

Exploring the immunohistochemical and immunofluorescent approaches, we observed that the expression levels of NP65 were significantly elevated in the hippocampus, temporal and frontal cortex, and cerebellum in the AD group compared to controls. However, the cellular localization of NP65 immunoreactive deposits in AD had not changed. Intracellular immunoreactivity, as previously reported[21], was not found.

In the hippocampal formation, we found that NP65 immunoreactivity in the DG, CA1 and CA3 of the hippocampus in the AD group was significantly increased compared to the control group (Figure 5A-E), partially consistent with Ilic’s report showing increased NP65 expression at an early stage of AD[21]. Notably, the apical dendrites of the pyramidal cells in CA3 showed strongly positive immunoreactivity in AD cases (Figure 5C, indicated by an arrow). Moreover, western blot analysis was performed to validate the immunohistochemical staining results and showed a significant increase in NP65 protein in the AD hippocampus compared with the controls (Figure 5F and G).

Figure 5 Neuroplastin 65 immunoreactivity pattern in the hippocampal formation and temporal cortex derived from individuals with Alzheimer’s disease and age-matched controls. A: Low-power microphotographs display the regional distribution of neuroplastin 65 (NP65) in the hippocampal formation, entorhinal cortex and fusiform gyrus of controls (CON) and Alzheimer’s disease (AD) (scale bars = 2 mm); B-D: High-power microphotographs of cornu ammonis (CA) 1, CA3 and dentate gyrus, respectively. Note the strong positive staining of apical dendrites of the stratum pyramidale of CA3 (indicated by arrow in C) in AD (scale bars = 50 μm); E: Quantification of NP65 immunoreactivity in the hippocampal formation, entorhinal cortex and fusiform gyrus between AD and CON; F and G: NP65 immunoreactive bands from total membrane protein of the hippocampus in AD and CON, and quantitative analysis showing NP65 levels in the hippocampus of AD and CON (data from two independent experiments). NP65: Neuroplastin 65; DG: Dentate gyrus; CA: Cornu ammonis; EC: Entorhinal cortex; SG: Stratum granulosum; SM: Stratum moleculare; SR: Stratum radiatum; SP: Stratum pyramidale; FG: Fusiform gyrus; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; Neuroplastin 65 (NP65); CON: Controls; AD: Alzheimer’s disease.

In the inferior and medial temporal cortex, the entorhinal cortex (EC) of the para-hippocampal gyrus displayed higher immunoreactivity while the fusiform gyrus showed comparable immunoreactivity in the AD group compared with the controls (Figure 5A-E). In the dorsolateral-frontal cortex, NP65 immunoreactivity in layer II of the dorsolateral PFC and in layers II and V of the precentral gyrus was significantly increased in the AD group relative to the controls (Figure 2A and C-E). Note that strongly positive apical dendrites in layer V are more visible in AD than in controls (Figure 2A and D, indicated by an arrow).

In the cerebellar cortex, the ML and the GCL showed a significant increase in NP65 immunoreactivity, while the PCL showed unchanged immunoreactivity in the AD group when compared to the controls (Figure 4D). Strikingly, the GCL of the AD group showed more intense, and blocky or granular immunoreactivity compared to the controls (Figure 4C, indicated by an arrow). Taken together, the present results clearly show that NP65 expression is significantly increased in wide areas of the AD brain.

NP65 colocalized with the NFT and Aβ plaques in AD brain

For Aβ and NFTs, which are the major neuropathological features of AD, we investigated whether NP65 is involved in Aβ and p-Tau pathology. Interestingly, double-immunofluorescent staining of NP65 with p-Tau (a marker for NFTs) showed that NP65 immunoreactive punctate structures partially colocalized with the NFTs in the EC of AD (Figure 6A). In addition, the double-immunofluorescent staining of NP65 with 6E10 (a marker for Aβ plaques) displayed that NP65 immunostaining also partially colocalized with diffuse and punctate Aβ plaques in the DG (Figure 6B, indicated by an arrow). Together, we show for the first time that NP65 partially colocalizes with NFTs and Aβ plaques in the AD human brain.

Figure 6 Neuroplastin 65 colocalized with the neurofibrillary tangles and amyloid beta plaques in Alzheimer’s disease brain. A: The double immunostaining microphotograph of neuroplastin 65 (NP65) with AT-8 (a marker for neurofibrillary tangles) showed that NP65-positive puncta partially colocalized with phosphorylated-microtubule-associated protein tau (yellow, indicated by arrow) in the entorhinal cortex of Alzheimer’s disease brain, scale bar = 30 μm; the left panel shows a higher power view in the rectangle frame, scale bar = 10 μm; B: The double immunostaining microphotograph of NP65 with 6E10 (a marker for amyloid beta) exhibited that NP65-positive puncta partially colocalized with amyloid beta plaques (yellow, indicated by arrow) in the cornu ammonis 3 region of Alzheimer’s disease brain, scale bar = 200 μm; The left panel shows a higher power view in the rectangle frame, scale bar = 30 μm. The nucleus was stained with 4’,6-diamidino-2-phenylindole (blue). NP65: Neuroplastin 65; EC: Entorhinal cortex; DG: Dentate gyrus; SM: Stratum moleculare; SG: Stratum granulosum; Spl: Stratum plexiforme; DAPI: 4’,6-diamidino-2-phenylindole.

DISCUSSION

In this study, we showed that NP65 is significantly increased in a wide range of postmortem AD human brain regions, such as the frontal and temporal cortex, hippocampus, DG, and cerebellar cortex. Importantly, we found that NP65 was partly colocalized with NFTs and Aβ plaques in AD. Combined with our previous study showing a decreased Aβ plaques in APP/PS1 mice when NP65 is deficient[8], these results provide evidence that NP65 is involved in the pathology of AD.

Recently, some neural CAMs have been reported to be involved in the pathogenesis of AD in mice. For example, L1, as a membrane CAM, shows an increase in the cerebrospinal fluid of AD patients[25]. In addition, L1 is reported to be decreased in the hippocampus of APPswe mice, and upregulation of L1 or the L1-70 fragment results in a decrease in Aβ plaques in APP/PS1 or APPswe mice (mouse models of AD)[7,26]. These results suggest that L1 expression may be negatively associated with Aβ plaque formation. Interestingly, NP65, another member of the CAMs, has been reported to play crucial roles in neuronal survival, neurite outgrowth, synapse formation, hippocampal CA1 synaptic plasticity, and cognitive function as well as neuronal response following ischemic injury in mice[12,18,27]. Notably, our recent study showed that the deficiency of NP65 resulted in alleviated cognitive deficits, decreased Aβ plaques and reduced microglial activation and neuroinflammation in APP/PS1 mice[8]. Additionally, it was reported that the NP65 immunoreactive product was significantly increased in the hippocampus of AD patients with a short duration[21]. Therefore, these results suggest the NP65 might be involved in the pathogenesis of AD.

The distribution of NP65 in the brain shows species differences in rats, mice and humans[10,12,14,15]. Despite there being more reports on NP65 distribution in rodent brains[11,12,15], only two groups have reported on the NP65 distribution in some regions of the human brain[10,14]. To determine whether NP65 is involved in AD pathogenesis, it is necessary to clarify the detailed characteristics of its expression in sublayers of the cortex, its cellular and subcellular distribution, and its distribution in a wide range of regions in the human brain[10]. Interestingly, the present study showed that NP65 immunoreactivity is most prominent in the PFC. Moreover, strongly-positive apical dendrites were visible in layer V of the PFC. These results are inconsistent with a previous report that showed no regional differences in the cerebral cortex[10]. In the cerebellum, the most intense NP65 immunoreactivity was found in the ML and GCL. The blocky immunoreactive product may be synaptic glomeruli in the GCL, as previously reported in the human cerebellum or NP55 in the rat cerebellum[10,15]. In the hippocampal formation, the most prominent NP65 immunoreactivity was found in the SG and stratum molecular of the DG, consistent with a previous report[10,14]. Unexpectedly, we found prominent punctate structures in the SG, greatly different to the tiny punctate structures in other brain regions. As NP65 can be localized to both pre- and post-synaptic structures[13,28], this large punctate structure may be a synapse indicating the projection of the EC to the granular cells. Notably, we have identified for the first time that there is abundant NP65 immunolabelling in the substantia nigra in the midbrain, implying a potential role in Parkinson’s disease. At the cellular level, NP65 was localized on the cell membrane and dendrites of some principal neurons, such as pyramidal neurons in the hippocampus, pyramidal neurons in layer II and V of the cerebral cortex, and localized in the neuropil. It showed punctate, blocky, and continuous plexuses, positive nerve fibers and a dense network in appearance. The medulla and projection fibers in the brainstem were always devoid of any immunoreaction.

In the present study, intracellular NP65 immunoreactivity was not detected throughout all sections. Interestingly, a previous report showed NP65 immunoreactivity in the somata of neurons of the anterior hypothalamus in normal human brain[10]. This may reflect the complexity of NP65 expression in different subregions and different types of neurons, as there is no brain atlas of NP65 distribution in the human brain. Taken together, the present results show that NP65 is differentially expressed in divergent neurons and is most abundant in the PFC, hippocampal formation, cerebellar cortex, and substantia nigra in the midbrain throughout the human brain.

Next, NP65 expression was detected in the AD and control human brain. Interestingly, we found that the NP65 immunoreactive products were significantly increased in a wide range of brain regions in the AD group compared with the controls, such as the hippocampal formation, frontal and EC, and cerebellar cortex. Furthermore, western blot analysis further confirmed this increase of NP65 protein in the AD hippocampus compared to controls. At the subcellular level, there was no difference in the localization of NP65 immunoreactivity between the AD and control groups. In all AD cases, NP65 was not found to be ectopically expressed in the cytoplasm. Inconsistently, a previous report showed that intracellular NP65 immunoreactivity was found in neurons of the subicular pyramidal layer of the hippocampal formation[21]. This discrepancy may be due to the different methods, severity of AD and altered trafficking and degradation of NP65 protein in AD pathology. Remarkably, the NP65 immunolabelled apical dendrites in layer V in the PFC and apical dendrites of pyramidal cells in CA3 were more prominent and clearly visible in AD cases compared to controls. Additionally, the granular cell layer of the cerebellar cortex exhibited more intense immunoreactivity and appeared blocky in AD cases than in controls. Therefore, the present results clearly show a significant increase in the expression of NP65 in vast brain areas of AD cases.

Finally, we explored whether NP65 is related to Aβ plaques and NFTs, which are the neuropathological hallmarks of AD. More importantly, we found for the first time that NP65 was partly colocalized with NFTs and Aβ plaques in the AD human brain, confirmed by double-immunofluorescent staining. These results indicate that NP65 may affect the formation of Aβ plaques and NFTs in AD pathology.

Our recent studies have shown that NP65 knockout alleviates cognitive deficits and reduces Aβ plaques and neuroinflammation in APP/PS1 mice[8]. Notably, NP65 knockout reverses the increase in 5-hydroxytryptamine receptor 3A (Htr3A) levels in APP/PS1 AD model[8]. Combined with the previous study showing that Htr3A interneurons contribute to Aβ production and inhibiting Htr3A expression decreases Aβ plaques in APP/PSI mice[23], it is thus hypothesized that NP65 deficiency decreases Aβ plaques by downregulating Htr3a expression in APP/PS1 mice. Recent studies have proved that NP65/55 is required for plasma membrane Ca²⁺ adenosine triphosphatase expression and essential auxiliary subunits of plasma membrane Ca²⁺ adenosine triphosphatase and key regulators of Ca²⁺ clearance[29,30]. Ca²⁺ imbalance of homeostasis could promote brain function deficits and accumulation of Aβ in AD patients. Given that Htr3 is an ionotropic receptor with Ca²⁺ permeability[31], it is possibly that NP65 could affect Ca²⁺ imbalance of homeostasis in Htr3A interneurons, which contribute to Aβ generation. However, how NP65 likely affects the formation of NFTs and Aβ in AD pathology by regulating Htr3A interneurons remains unknown. Taken together, our present findings suggest that NP65 may be involved in the formation of Aβ plaques and NFTs in human AD pathology. In the near future, using 3xTg-AD mice that develop Aβ plaques and NFTs[32], we anticipate examining how the up- or down-regulation of NP65 level affects the formation of Aβ plaques and NFTs and the underlying mechanisms. In addition, using co-immunoprecipitation and protein mass spectrometry to determine the interacting proteins with NP65 may identify potential mechanisms of NP65 involved in AD pathology.

CONCLUSION

In conclusion, our results show that the NP65 protein is significantly increased and colocalized with NFTs and Aβ plaques in the brains of individuals with AD. Combined with our previous report on AD mice[8], our findings suggest that NP65 is involved in the formation of Aβ plaques in AD, making it a potential target for the treatment of this disease.