Poloxamer-188 Exacerbates Brain Amyloidosis, Presynaptic Dystrophies, and Pathogenic Microglial Activation in 5XFAD Mice

Abstract

Background:

Alzheimer's disease (AD) is initiated by aberrant accumulation of amyloid beta (Aβ) protein in the brain parenchyma. The microenvironment surrounding amyloid plaques is characterized by the swelling of presynaptic terminals (dystrophic neurites) associated with lysosomal dysfunction, microtubule disruption, and impaired axonal transport. Aβ-induced plasma membrane damage and calcium influx could be potential mechanisms underlying dystrophic neurite formation.

Objective:

We tested whether promoting membrane integrity by brain administration of a safe FDA approved surfactant molecule poloxamer-188 (P188) could attenuate AD pathology in vivo.

Methods:

Three-month-old 5XFAD male mice were administered several concentrations of P188 in the brain for 42 days with mini-osmotic pumps. After 42 days, mice were euthanized and assessed for amyloid pathology, dystrophic neurites, pathogenic microglia activation, tau phosphorylation, and lysosomal / vesicular trafficking markers in the brain.

Results:

P188 was lethal at the highest concentration of 10mM. Lower concentrations of P188 (1.2, 12, and 120μM) were well tolerated. P188 increased brain Aβ burden, potentially through activation of the γ-secretase pathway. Dystrophic neurite pathology was exacerbated in P188 treated mice as indicated by increased LAMP1 accumulation around Aβ deposits. Pathogenic microglial activation was increased by P188. Total tau levels were decreased by P188. Lysosomal enzyme cathepsin D and calcium-dependent vesicular trafficking regulator synaptotagmin-7 (SYT7) were dysregulated upon P188 administration.

Conclusion:

P188 brain delivery exacerbated amyloid pathology, dystrophic neurites, and pathogenic microglial activation in 5XFAD mice. These effects correlated with lysosomal dysfunction and dysregulation of plasma membrane vesicular trafficking. P188 is not a promising therapeutic strategy against AD pathogenesis.

1. INTRODUCTION

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by aberrant accumulation of misfolded Aβ protein in the form of amyloid plaques in the brain parenchyma [1]. The microenvironment surrounding amyloid plaques is characterized by aberrant swelling of presynaptic axons and terminals (dystrophic neurites) associated with microtubule disruption, lysosomal dysfunction, and impaired axonal transport as an apparent consequence of Aβ toxicity [2]. These dystrophic neurites are sites where synaptic damage and further exacerbation of amyloid pathology occur [2-4]. The initiating mechanism governing dystrophic neurite formation is not clear, although Aβ-induced plasma membrane damage and aberrant calcium influx are potential candidates [5-9]. In particular, Aβ fibrils and oligomers may interact with the plasma membrane of neurons in dystrophic neurites and potentially reduce membrane integrity [5-7]. Amyloid aggregates promote calcium influx through the neuronal plasma membrane in vitro [8]. Calcium overload was observed at dystrophic neurites in proximity of Aβ plaques in vivo [9]. Increased calcium was directly linked to alteration of neuritic morphology in the same model [9]. We hypothesized that promoting membrane integrity with a membrane resealing agent might reduce the formation of dystrophic neurites.

Poloxamer-188 (P188) is an 8400 kDa non-ionic linear copolymer with surfactant properties capable of resealing damaged membranes by incorporation in the phospholipid bilayer [10, 11]. P188 is a well-tolerated molecule approved by the FDA to reduce blood viscosity in blood transfusion and is largely used in cosmetic, pharmaceutical, and industrial applications [12]. P188 showed beneficial effects when delivered to mouse models of several neurodegenerative disorders. P188 promoted survival of motor neurons in SOD1 transgenic models of Amyotrophic Lateral Sclerosis (ALS) [13]. Moreover, P188 administration reduced loss of dopaminergic neurons, attenuated α-synuclein accumulation, reduced lysosomal dysfunction, promoted synaptic integrity, and mitigated microgliosis in mouse models of Parkinson’s disease (PD) [14, 15]. In vitro studies demonstrated that P188 reversed the membrane permeabilization effects of Aβ oligomers and increased the survival of neural cells [16].

In the current study, we investigated whether membrane-sealing agent P188 could attenuate the deleterious effects of Aβ plaques on dystrophic neurites in vivo. To test this hypothesis, we delivered P188 via osmotic mini pump in the brains of 3-month-old 5XFAD male mice when amyloid pathology and dystrophic neurites were well established [17]. After 42 days, we euthanized the mice and measured amyloid burden, neuritic dystrophies, and pathogenic microglial activation around amyloid deposits. To our surprise, P188 treatment strongly exacerbated amyloid plaque pathology, dystrophic neurites, and microglial activation in the brains of 5XFAD mice. We also found evidence that P188 may promote activation of γ-secretase cleavage of amyloid precursor protein (APP) into Aβ. In addition, P188 caused a decrease in total tau levels in the same mice. P188 promoted lysosomal dysfunction, as demonstrated by increased pro-cathepsin D levels, and induced dysregulation of calcium-dependent vesicular trafficking regulator synaptotagmin-7 (SYT7). These data suggest that treatment with P188 in the brain is not a suitable therapeutic strategy to attenuate dystrophic neurites. In fact, P188 delivery strongly exacerbated amyloid pathology in a dose dependent fashion.

2. MATERIALS AND METHODS

2.1. Mice

5XFAD (B6SJL) harboring human APP mutations (Swedish K670N/M671L, Florida I716V, and London V717I) and human PSEN1 mutations (M146L and L286V) were generated and maintained as described previously [2, 18]. Specifically, male transgenic animals were crossed to first generation B6SJL hybrid females. Male mice were used exclusively in this study to avoid known gender differences in amyloid accumulation in 5XFAD mice [19].

Mice were housed in the facilities of the Center for Comparative Medicine at Northwestern University Chicago. All experimental procedures were approved by the IACUC office of Northwestern University under animal protocol IS00001530.

2.2. Mini-osmotic Pump Implantation for Poloxamer-188 (P188) Delivery into the Right Lateral Ventricle

Three-month-old transgene positive males underwent mini-osmotic pump implantation to deliver poloxamer-188 (P188) in the right lateral ventricle. Surgery was performed using Model 900 Small Animal Stereotaxic Instrument, KOPF Instruments. Mini-osmotic pump model 2006 and brain infusion kit 3 model 0008851 were purchased from Alzet.

Animals were anesthetized by isofluorane inhalation (ISOTHESIA, ndc 11695-6776-2, Henry Schein) delivered with a Basic Small Animal Anesthesia Device model R500IE, RWD Life Science Co. Ltd. After checking for the absence of reflexes, the mouse head was shaved, and skin was disinfected with 70% ethanol wipes and betadine scrubs. The right lateral ventricle was located using the following stereotaxic coordinates: AP: −0.6, ML: +1.2, DV: −2.0 (Fig. 1A). A small hole was made in the skull, and the brain infusion kit was implanted in the right lateral ventricle and anchored to the skull with instant adhesive gel. The mini-osmotic pump was connected to the infusion kit and placed subcutaneously in the dorsal area of the mouse.

Fig. (1). P188 brain delivery reduces survival and brain weight in 5XFAD mice.

A (left): Exposed skull of mouse with head held in stereotaxic apparatus for insertion of cannula for mini-osmotic pump implantation. A (right): Stereotaxic coordinates used for cannula implantation in the right lateral ventricle. B: Mouse brain demonstrating successful delivery of blue dye solution into both brain ventricles using a single site injection at coordinates in A. C: Survival curve for 5XFAD mice treated with different concentrations of P188 or vehicle delivered by osmotic minipump. Note that only the 10mM P188 concentration showed reduced survival. D: Body weight of all groups of animals at 4.5 months, the age of brain harvest. E: Brain weight of all groups of animals. F: Brain weight:body weight ratio for all groups of animals. No significant differences between treatments were noted for any weight parameter. 5XFAD transgenic treated with vehicle (VHL n=6); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=7); 5XFAD transgenic treated with 12μM P188 (12μM n=6); 5XFAD transgenic treated with 120μM P188 (120μM n=8); 5XFAD transgenic treated with 10mM P188 (10mM n=4). Data expressed as mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

Pumps were filled with either artificial cerebrospinal fluid (aCSF) or different concentrations of 10% sterile-filtered poloxamer-188 solution (P188) (cat # P5556, SIGMA Life Science) diluted in aCSF. Solution was delivered at a speed of 0.15μl/h over 42 days. Mouse health and survival were monitored daily for 42 days.

Animals were randomly assigned to the following groups:

• VHL = 6 5XFAD male mice administered aCSF only (vehicle);

• 1.2μM = 7 5XFAD male mice administered 1.2μM poloxamer-188 in aCSF;

• 12μM = 6 5XFAD male mice administered 12μM poloxamer-188 in aCSF;

• 120μM = 8 5XFAD male mice administered 120μM poloxamer-188 in aCSF;

• 10mM = 4 5XFAD male mice administered 10mM poloxamer-188 in aCSF.

Decreasing concentrations of P188 were selected starting from the published 10mM dose [13]. We observed high mortality with the 10mM concentration; therefore, three lower doses were implemented that were well tolerated. Accuracy of the stereotaxic coordinates was tested in one animal by injecting blue dye into the implanted brain infusion kit. This mouse was euthanized immediately and the brain harvested to assess successful delivery of the dye to both ventricles (Fig. 1B).

2.3. Necropsy and Organ Collection

Animals were euthanized with a lethal injection of ketamine (100mg/kg) and xylazine (15 mg/kg) 42 days post-surgery. After checking for the absence of reflexes, animals underwent transcardial perfusion with ice-cold perfusion buffer (20mg/ml phenylmethylsulfonyl fluoride, 5mg/ml leupeptin, 200nM sodium orthovanadate, and 1M Dithiothreitol in 1X PBS). Brains were harvested and weighed. Left hemi brains were fixed in 10% formalin and cryopreserved in 30% sucrose / 1X PBS for immunofluorescence microscopy analysis. Right hemi brains were dissected on ice to isolate the cortex, hippocampus, cerebellum, and mid brain. After dissection, right hemi brain regions were flash frozen in liquid nitrogen and stored at −80 °C for biochemical analyses.

2.4. Antibody Staining for Immunofluorescence Microscopy

Left hemi brains were cut into 30μm free-floating coronal sections using an HM 430 freezing-sliding microtome. Sections were stored in cryopreservative buffer (30 % sucrose / 30 % ethylene glycol in 1X PBS) at −20 °C. The day of the experiment, five comparable coronal sections, including the dorsal hippocampus, were collected for each mouse (bregma −1.06 to −2.30). Sections were washed 3 times in TBS buffer for 5 minutes and then incubated in 16mM glycine / 0.25% triton TBS solution for 1 hour at room temperature. After 3 x 5 minute TBS washes, sections were incubated in 5% donkey serum / 0.25% triton TBS solution for 1 hour. Sections were incubated overnight at 4 °C with primary antibodies diluted in 1% BSA / 0.25% triton TBS solution (1% BSA buffer) (Table 1). The following day, sections were rinsed 3 x 10 minutes in 1% BSA buffer. Secondary antibodies in 1% BSA buffer were added to the sections for 1 hour at room temperature in the dark (Table 2). After 3 x 5 minute TBS washes, sections were mounted on diamond white glass microscope slides (cat # 1358W, Globe Scientific Inc.) using Prolong Gold antifade reagent (cat #P36930, Invitrogen) and 24 x 40mm No 1.5 gold seal cover glasses (cat # 3421, Thermo Scientific).

Table 1.

Summary of primary antibodies

Antibody	Vendor / Donor	Catalog #	Dilution	Application
Aβ-3D6 (mouse) Lot 1397A	Elan Pharmaceuticals	-	1:1000	IF
LAMP1 (rat)	DHSB	1D4B	1:1000	IF
IBA1 (rabbit)	ABCAM	1778846	1:500	IF
CD68 (rat)	BIORAD	MCA 1957GA	1:250	IF
APP/Aβ-6E10 (mouse)	BIOLEGEND	8030001	1:1000	WB
PS1 NT (rabbit) *	Dr. Gopal Thinakaran	-	1:2000	WB
BACE1 (rabbit)	ABCAM	108394	1:1000	WB
APOE HJ6.8	Dr. David Holtzman	-	1:1000	WB
CATHEPSIN-D (rabbit)	ABCAM	75852	1:1000	WB
SYT7 (rabbit)	SYNAPTIC SYSTEMS	105 173	1:2000	WB
TAU5 (mouse)	MILLIPORE	MAB361	1:1000	WB
PHF1 (mouse)	Dr. Peter Davies	-	1:1000	WB
p-Tau-Ser404 (rabbit)	Cell Signaling Technology	20194	1:1000	WB
p-Tau-Ser396 (rabbit)	Invitrogen	44-752G	1:1000	WB
BETA-ACTIN (mouse)	SIGMA	A5441	1:5000	WB
GAPDH (rabbit)	Cell Signaling Technology	2118	1:1000	WB

^*Used without boiling step. IF (immunofluorescence). WB (western blot).

Table 2.

Summary of secondary antibodies

Antibody	Vendor	Catalog #	Dilution	Application
Donkey anti-mouse Alexa Fluor 647	Thermo Fisher	A31571	1:750	IF
Donkey anti-rat Alexa Fluor 488	Thermo Fisher	A21208	1:750	IF
Donkey anti-rabbit Alexa Fluor 568	Thermo Fisher	A10042	1:750	IF
Goat anti-rabbit peroxidase	Vector Laboratories	PI-1000	1:5000	WB
Horse anti-mouse peroxidase	Vector Laboratories	PI-2000	1:5000	WB

IF (immunofluorescence). WB (western blot).

Aβ (3D6) and LAMP1 staining (Fig. 2) were performed on the same sections. Staining for IBA1 and CD68 were performed on a separate set of sections together with 3D6.

Fig. (2). P188 brain delivery increases amyloid pathology and dystrophic neurites in 4.5 month old 5XFAD mice.

A: Representative images of immunofluorescence microscopy staining for extracellular Aβ (3D6 antibody; red) and intracellular LAMP1 (green) in coronal cortical sections of mice from all P188 treatment groups (concentrations indicated at top). B: Quantification of Aβ (3D6) immunostaining in the cortex of P188 treated 5XFAD mice. Each point represents the average of cortical % area Aβ (3D6) immunoreactivity using 5 sections per mouse. Note the dose-dependent increase of % area of Aβ immunostaining for P188 treatments compared to vehicle. C: Quantification of LAMP1 staining in the cortex of P188 treated 5XFAD mice. Each point represents the average of cortical % area of LAMP1 immunoreactivity using 5 sections per mouse. Note the significant increase of % area of LAMP1 immunostaining for the 120μM P188 concentration and trends for the other concentrations compared to vehicle. 5XFAD transgenic treated with vehicle (VHL n=5); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=7); 5XFAD transgenic treated with 12μM P188 (12μM n=5); 5XFAD transgenic treated with 120μM P188 (120μM n=5). Data expressed as mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

2.5. Microscopy and Image Analysis

Imaging work was performed at the Northwestern University Center for Advanced Microscopy, generously supported by NCI CCSG P30 CA060553, awarded to the Robert H Lurie Comprehensive Cancer Center. Images were captured with a Nikon Ti2 Widefield microscope using a 10X air objective (0.3 objective numerical aperture).

Cortex from each brain section was highlighted with the polygon tool in ImageJ-Fiji software and saved in a new file separately from the rest of the brain with edit-clear outside command. All relevant channels from the same sections were processed with identical ROI. Brain cortex files were analyzed with the following macros:

• 3D6 / LAMP1 sections:

• setAutoThreshold("Triangle dark");

• setOption("BlackBackground", false).

• run("Convert to Mask");

• run("Analyze Particles…", "size=20-Infinity summarize");

• IBA1 / CD68 sections:

• run("Subtract Background…", "rolling=50");

• setAutoThreshold("Triangle dark");

• setOption("BlackBackground", false);

• run("Convert to Mask");

• run("Measure").

Percent area values from five sections per mouse were averaged in excel. GraphPad Prism 8 software was used for statistical analysis.

Protein Sample Extraction

15 to 35mg of cortical tissue was minced and sonicated using Ultrasonic Liquid Processor XL-2000 (Misonix) with 150 to 350ul of Radio Immunoprecipitation Assay (RIPA) buffer containing protease inhibitor (cat # 535140, Calbiochem) and phosphatase inhibitor (cat # 78427 Thermo Fisher Scientific) cocktails. Samples were centrifuged at 14,000 rpm for 45 minutes at 4 °C. Supernatants were assayed for protein concentration with the Pierce BCA assay kit (Thermo Fisher Scientific).

Western Blot Analysis

20μg of cortical homogenate was mixed with 4x Laemmli Sample buffer (cat # 1610747, Bio-Rad) and heated at 95 °C for 10 minutes. Samples were separated on 4-12% Bis-Tris Criterion XT Precast Gel (cat # 3450125, Bio-Rad) using MES SDS running buffer (cat # NP0002-02, Invitrogen). Protein was transferred onto 0.45μm nitrocellulose membrane (cat # 1620167, Bio-Rad) using the Trans-Blot Turbo System (Bio-Rad). Membranes were incubated in 0.1% Ponceau solution to assess transfer quality. After 3 x 5 minute 0.1% Tween-TBS buffer washes, the membrane was blocked in 5% milk for 1 hour. Primary antibodies diluted in 5 % milk were added to the membrane overnight at 4 °C (Table 1). The next day, the membrane was washed 3 x 5 minutes in 0.1% Tween-TBS buffer. Secondary antibodies diluted in 5% milk were added to the membrane for 1 hour at room temperature (Table 2). After 2 x 5 minute 0.1% Tween-TBS buffer and 1 x 5 minute TBS buffer washes, the membrane was incubated in Super Signal West Femto Chemiluminescent Substrate (cat # 34096 Thermo Scientific) for 5 minutes. Super Signal West Pico Chemiluminescent Substrate (cat # 34580 Thermo Scientific) was used for loading controls only. Membranes were imaged using the Fluor-Chem R System (Protein Simple). Densitometric analysis was performed using AlphaView Software (Protein Simple).

Statistical Analysis

All statistical analyses were performed using GraphPad Prism 8 software. The one-way ANOVA followed by Turkey’s multiple comparisons analysis was used to determine differences between groups. Results are presented as means with standard deviations, and a p-value equal to or lower than 0.05 (*) was considered significant.

3. RESULTS

3.1. 10mM P188 Brain Delivery Decreases Survival in 5XFAD Mice

The effects of P188 brain delivery on general health and survival were monitored daily in 5XFAD male mice compared to vehicle treated animals. Only male mice were used in this study to reduce variability since 5XFAD mice display gender differences in brain amyloid burden [19].

High dose P188 (10mM) brain delivery promoted a sharp decline in survival. In particular, survival dropped to 50% and to 25% after one and two days post-surgery, respectively (Fig. 1C). No animal survived 10mM brain delivery of P188 to 42 days post-surgery (Fig. 1C). All animals treated with lower doses (1.2μM, 12μM, and 120μM) of P188 or vehicles survived to the 42 days post-surgery endpoint (Fig. 1C).

Body weight and brain weight were assessed in all the groups 42 days after surgery. No significant differences were observed in body and brain weight among all groups (Fig. 1D and 1E). When normalizing brain weight to body weight, no significant differences were observed upon P188 treatment (Fig. 1F). We conclude that intracerebral ventricle delivery of P188 is well tolerated in mice, at least up to the 120μM dose for 42 days of treatment.

3.2. P188 Brain Delivery Increases Amyloid Deposition and Dystrophic Neurite Formation in 5XFAD Mice

Next, we assessed the effect of P188 brain delivery on amyloid pathology in the cortex of 5XFAD mice. In particular, we measured amyloid beta (Aβ) burden and dystrophic neurite formation around Aβ plaques.

High doses of P188 (12μM and 120μM) promoted a strong increase in Aβ accumulation in the cortex of 5XFAD mice compared to vehicle treatment, as demonstrated by increased area of Aβ (3D6) immunoreactivity (Fig. 2A and 2B). The 120μM P188 dose increased Aβ immunoreactive area compared to the 1.2μM dose as well. No effect on Aβ pathology was observed in the P188 1.2μM group compared to vehicle (Fig. 2A and 2B).

To analyze dystrophic neurites, we used anti-LAMP1 immunostaining as a marker for aberrant lysosomal accumulation in neuritic dystrophies around plaques [20]. P188 delivery promoted a significant increase in LAMP1 immunoreactivity around Aβ plaques in mice treated with the 120μM concentration when compared to vehicles (Figs. 2A and 2C). Lower doses of P188 (1.2μM and 12μM) promoted an increase in LAMP1 immunoreactivity that, however, was not statistically significant (P = 0.0674 and 0.0723, respectively) (Figs. 2A and 2C). These results demonstrate that P188 increases amyloid deposition and dystrophic neurite formation in 5XFAD mice in a dose-dependent manner.

3.3. P188 Brain Delivery Increases Aβ and PS1-NTF in 5XFAD Mice

To gain insight into the mechanism by which P188 promoted amyloid pathology in the brain, we investigated the pathways for Aβ production in the cortex of P188 treated 5XFAD animals compared to vehicle treated mice.

First, we measured protein levels of human Amyloid Precursor Protein (APP), β-secretase-derived C-terminal fragment (β-CTF), and Aβ in all groups by western blot using the anti-Aβ 6E10 antibody. As expected, no signal was detected for either of the three species in the non-transgenic wild type (WT) animals (Figs. 3A and 3B). Confirming the immunofluorescence findings (Fig. 2), Aβ levels were significantly increased for both the 12μM and 120μM P188 groups compared to vehicles when expressed as either percent of VHL or as a ratio to APP or β-CTF (Figs. 3A and 3B). No changes were detected for β-CTF levels when expressed as either percent of VHL or as a ratio to APP (Figs. 3A and 3B). These findings suggest either increased γ-secretase processing of APP, higher plaque formation or reduced Aβ clearance upon P188 treatment. Finally, APP levels were decreased in the P188 120μM mice compared to the other groups (Figs. 3A and 3B), suggesting either downregulation of APP transgene expression or neuronal loss.

Fig. (3). P188 brain delivery increases Aβ and PS1-NTF in 4.5 month old 5XFAD mice.

A: Western blot for APP (6E10 antibody) and GAPDH loading control. Molecular weight marker positions (kDa) are shown at the right and positions of APP, β-CTF, and Aβ are indicated at the left. # denotes non-specific bands. B: Quantification of APP, β-CTF, and Aβ immunoblot signal expressed as % VHL (upper panel) and quantification of β-CTF normalized to APP and Aβ normalized to APP and β-CTF (lower panel). Note that APP levels decreased while Aβ levels increased at the higher P188 concentrations. Aβ normalized to APP and β-CTF also increased at higher P188 concentrations, indicating elevated γ-secretase processing, increased plaque formation, or reduced Aβ clearance. C: Western blot for PS1 full length, PS1-NTF and actin loading control. Molecular weight marker positions (kDa) are shown at the right and positions of PS1 full length and PS1-NTF are indicated at the left. # denotes non-specific band. D: Quantification of full-length PS1 (left, % VHL), PS1-NTF (middle; % VHL), and PS1-NTF/PS1 full length (right). Note the significant increases in PS1-NTF/PS1 full length in the absence of changes in PS1 full length. 5XFAD transgenic treated with vehicle (VHL n=5); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=5); 5XFAD transgenic treated with 12μM P188 (12μM n=5); 5XFAD transgenic treated with 120μM P188 (120μM n=5). Data expressed at mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

Next, we assessed the main enzymes involved in the amyloidogenic cleavage of APP: β-secretase (BACE1) and the active subunit of γ-secretase (presenilin). To test γ-secretase cleavage, we measured levels of full length versus active N-terminal fragment (NTF) of presenilin-1 (PS1). Consistent with the 6E10 data, we found an increase for PS1-NTF when quantified as a ratio to full length PS1 for the 12μM and 120μM dose groups compared to the other groups (Figs. 3C and 3D). A non-significant increase in PS1-NTF protein levels expressed as percent VHL was also detected for the 12μM and 120μM doses (Figs. 3C and 3D). All groups displayed no changes in full length PS1 protein levels (Figs. 3C and 3D). BACE1 protein levels were also not changed upon P188 administration (Figs. S1A and S1B). Finally, we measured Apolipoprotein-E (APOE) levels using the anti-APOE antibody HJ6.8 [21]. APOE protein levels were not changed upon P188 treatment (Figs. S1C and S1D). Taken together, our results suggest that P188 treatment may increase Aβ production by elevating γ-secretase activity via increased PS1-NTF/PS1 full-length ratio.

3.4. P188 Brain Delivery Exacerbates Pathogenic Microglial Activation Around Plaques in 5XFAD Mice

Aβ plaques promote pathogenic microglial activation in the brain (reviewed in [22, 23]). To test the effect of P188 on microglial activation, we measured levels of ionized calcium-binding adaptor molecule (IBA1), a general marker of microglia, and a cluster of differentiation-68 (CD68), which is increased in activated microglia in the cortex of P188-treated mice by immunofluorescence microscopy.

Levels of activated microglia marker CD68 [24] were significantly increased in both 12μM and 120μM P188-treated mice compared to VHL. The 12μM dose showed a significant CD68 elevation when compared to the 1.2μM dose as well. In addition, levels of general microglia marker IBA1 [25] were significantly increased only in the 12μM group when compared to VHL mice (Figs. 4A and 4B). The muted effect at 120μM P188 may be due to reduced APP and Aβ at this dose (Fig. 3). No effect was observed for either marker in the 1.2μM P188 group when compared to VHL (Fig. 4A-C). Our results suggest that P188 increases microglial activation in 5XFAD mice, probably as a consequence of increased amyloid deposition.

Fig. (4). P188 brain delivery exacerbates microglial activation around plaques in 4.5 month old 5XFAD mice.

A: Representative images of immunofluorescence microscopy staining for Aβ (3D6; blue), IBA1 (red) and CD68 (green) in coronal cortical sections of mice from all P188 treatment groups (doses indicated at top). B: Quantification of IBA1 immunostaining in the cortex of P188 treated 5XFAD mice. Each point represents the average of cortical % area IBA1 immunoreactivity using 5 sections per mouse. Note the significant increase of % area IBA1 immunostaining for the 12μM P188 dose. C: Quantification of CD68 immunostaining in the cortex of P188 treated 5XFAD mice. Each point represents the average of cortical % area CD68 immunoreactivity using 5 sections per mouse. Note the significant increase of % area CD68 immunostaining for the 12μM and 120μM P188 doses. 5XFAD transgenic treated with vehicle (VHL n=5); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=7); 5XFAD transgenic treated with 12μM P188 (12μM n=5); 5XFAD transgenic treated with 120μM P188 (120μM n=5). Data expressed at mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

3.5. P188 Brain Delivery Reduces Total Tau Levels in 5XFAD Mice

Tau hyper-phosphorylation is another hallmark of AD. 5XFAD mice do not develop tau neurofibrillary tangles but accumulate early tau phosphorylation epitopes in the brain [26]. We tested the effect of P188 treatment on total tau and Ser396 and Ser404 phospho-tau levels in the brain. High doses of P188 (12μM and 120μM) promoted a significant reduction of total tau using the TAU5 antibody compared to vehicle (Figs. 5A and 5B). However, no significant changes were detected when measuring levels of tau phosphorylated at Ser396 or Ser404 either separately or together using the PHF1 antibody (Figs. 5A and 5B). Ratios of phosphorylated tau over total tau were also not significantly affected by treatment, although increased trends were observed for the 12μM and 120μM doses (Fig. 5C). Together with reduced APP (Figs. 3A and B), the decreased total tau levels at 12mM and 120mM suggest that P188 is neurotoxic at these doses.

Fig. (5).

P188 brain delivery reduces total tau levels in 4.5 month old 5XFAD mice but does not affect tau phosphorylated at Ser396 or Ser404. A: Western blots for TAU5, PHF1, pTau(Ser396) and pTau(Ser404) together with GAPDH or actin loading controls. B: Quantification of western blots in A normalized to loading control and expressed as percent of VHL. C: Quantification of western blots in A expressed as absolute ratio of each pTau signal normalized to total tau (TAU5) signal. Note that total tau (TAU5) levels were decreased for 12μM and 120μM P188 concentrations (B), while phosphorylated tau levels were not significantly changed at any P188 concentration (B, C). 5XFAD transgenic treated with vehicle (VHL n=5); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=5); 5XFAD transgenic treated with 12μM P188 (12μM n=5); 5XFAD transgenic treated with 120μM P188 (120μM n=5). Data expressed at mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

3.6. P188 Brain Delivery Triggers Lysosomal Abnormalities in 5XFAD Mice

To gain insights into the potential mechanism responsible for the effect of P188 on brain pathology, we focused on lysosomal abnormalities, which have been reported to occur early in AD pathogenesis [27]. To do so, we measured levels of pro- and mature-forms of the lysosomal protease cathepsin D. As previously reported [2], inactive pro-cathepsin D is increased in the cortex of 5XFAD compared to wild type mice. P188 treatment exacerbated pro-cathepsin D accumulation at the highest 120μM dose when compared to the VHL and 1.2μM groups (Figs. 6A and 6B). The active mature-cathepsin D isoform was also increased by the 120μM dose (Figs. 6A and 6B). Finally, P188 administration promoted a significant increase in the mature- / pro-cathepsin D ratio at 120μM dose when compared to VHL treated mice (Figs. 6A and 6B). These observations suggest that P188 administration impairs lysosomal function in the brains of 5XFAD mice, at least at the 120μM dose.

Fig. (6). P188 brain delivery increases cathepsin D and decreases synaptotagmin-7 in 4.5 month old 5XFAD mice.

A: Western blots for cathepsin-D and actin loading control. Molecular weight marker positions (kDa) are shown at the right and positions of pro- and mature- cathepsin D are indicated at the left. # denotes non-specific band. B: Quantification of inactive pro-cathepsin D (left), active mature- cathepsin D (middle), and mature/pro cathepsin D ratio (right). Pro- and mature-cathepsin D levels are normalized to actin and expressed as percent of VHL. Note that both pro- and mature-cathepsin D are increased by 120μM P188, as well as the mature/pro cathepsin D ratio. C: Western blots for SYT7 and GAPDH loading control. Molecular weight marker positions (kDa) are shown at the right and positions of large and small SYT7 isoforms are indicated at the left. D: Quantifications of SYT7 small (left) and large (middle) isoforms in C normalized to GAPDH and expressed as percent of VHL. SYT7 large/small isoform ratio is also shown (right). Note the non-significant trend for increased SYT7 small isoform for all doses, the decrease in SYT7 large isoform for 120μM dose, and the reduced SYT7 large/small isoform ratio for 12μM and 120μM doses. 5XFAD transgenic treated with vehicle (VHL n=5); 5XFAD transgenic treated with 1.2μM P188 (1.2μM n=5); 5XFAD transgenic treated with 12μM P188 (12μM n=5); 5XFAD transgenic treated with 120μM P188 (120μM n=5). Data expressed as mean +/− standard deviation. * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001.

To better dissect the relationship between P188 administration and lysosomal dysfunction, we performed a correlation analysis between P188 concentration and cathepsin D levels (Figs. S2A-C). We observed a direct correlation between the dose of P188 and the levels of both the pro- and mature- isoforms of cathepsin D, as well as for the ratio of mature- / pro- cathepsin D. These data suggest that the concentration of P188 affects the lysosomal machinery progressively.

3.7. P188 Brain Delivery Decreases Presynaptic Calcium-dependent Vesicular Trafficking Protein Synaptotagmin-7 in 5XFAD Mice

Next, we investigated the effect of P188 on the presynaptic calcium-dependent vesicular trafficking protein, synaptotagmin-7 (SYT7). SYT7 regulates the presynaptic activity and membrane repair by lysosomal fusion in response to calcium [28-30]. The 120μM dose of P188 caused a significant decrease of the large isoform of SYT7 (75kDa), while a non-significant trend toward an increase for the most abundant small isoform (45kDa) was observed at all P188 doses (P = 0.0704, 0.0584 and 0.0861 for 1.2μM, 12μM, and 120μM, respectively (Figs. 6C and 6D) [29]. P188 promoted a dose-dependent decrease in the ratio of the large to the small SYT7 isoform (Figs. 6C and 6D). This data suggests a P188-induced dose-dependent proteolytic degradation of the large SYT7 isoform or alternative STY7 large isoform RNA splicing, which may result in abnormal vesicular trafficking and impaired lysosomal function at the presynaptic terminal.

4. DISCUSSION

In the current study, we report that brain delivery of membrane-sealing agent P188 increased Aβ deposition, dystrophic neurites, and pathogenic microglial activation around amyloid plaques in the brains of 5XFAD male mice. P188 treatment may have exacerbated amyloid accumulation in the brain by promoting γ-secretase processing of APP via elevated PS1-NTF/PS1 full-length ratio. Total tau levels were reduced by P188 treatment. P188 administration promoted lysosomal abnormalities, as evidenced by pro-cathepsin D upregulation. The large isoform of the calcium-dependent vesicular trafficking protein SYT7 appeared to undergo proteolytic degradation or alternative RNA splicing as a result of P188 treatment and may explain the deleterious effects of P188, at least in part. We did not investigate whether female 5XFAD mice had similar responses to P188, a topic for future study.

P188 is a surfactant copolymer with a good safety profile [12]. Several studies have described its therapeutic potential in regenerative medicine. In particular, P188 membrane resealing properties effectively promoted cellular healing in traumatic brain injury, ischemia reperfusion injury, heart failure, muscular dystrophies, and wound healing [10, 11, 31-50]. P188 administration showed promising effects in animal models of neurodegenerative disorders such as Parkinson’s disease and amyotrophic lateral sclerosis [13-15]. In vitro studies demonstrated that P188 reversed the membrane permeabilization effects of Aβ oligomers, thus increasing neural cell survival [16].

Dystrophic neurites around amyloid plaques are a key hallmark of AD (reviewed in [51]). In particular, Aβ plaques promote lysosomal dysfunction, cytoskeletal abnormalities, and disruption of axonal trafficking in the surrounding pre-synaptic processes [2, 20]. At these sites, synaptic loss and further Aβ accumulation drive the progression of AD pathology [2-4]. The exact mechanisms by which amyloid plaques trigger axonal dystrophies are still obscure. Aβ-induced plasma membrane damage and aberrant calcium influx are potential candidates [5-8]. In this study, we tested whether promoting membrane resealing by P188 administration could ameliorate dystrophic neurite formation around amyloid plaques.

First, we observed that lateral ventricle infusion of P188 at 10mM concentration was extremely toxic to mice. All animals treated with 10mM P188 died before the final time point. This finding is in striking contrast with previous literature. In fact, 10mM P188 was safely delivered to G93ASOD1 animals in a previous study [13]. Ours is the first report of a severe side effect of P188 brain delivery. The discrepancy between our results and those of Riehm et al. [15] is unclear. We hypothesize that 5XFAD mice might be particularly vulnerable to P188 treatment due to their extremely early and rapid brain pathology. Indeed, brain inflammation and dystrophic pathology are well established at 3 months of age in 5XFAD mice [17]. When utilizing lower doses of P188 (1.2uM, 12uM, and 120uM), no effect on survival up to 42 days was observed. Moreover, P188 administration at these doses had no effect on body and brain weight in our mice.

Next, we tested the effect of P188 delivery on brain amyloid burden. We observed a striking dose dependent increase of Aβ plaques in the brain cortex. When assessing dystrophic neurite pathology, we observed a strong increase in LAMP1-positive neuritic dystrophies around amyloid plaques in the cortex. These observations suggest that P188 increased amyloid deposition and consequently exacerbated the formation of dystrophic neurites. It is also possible that P188 administration promoted lysosomal abnormalities in dystrophies, thus enhancing amyloid deposition. In support of this hypothesis, endo-lysosomal dysfunction was described as a potential early event in AD pathology occurring before Aβ accumulation in pyramidal neurons of pre-symptomatic Down syndrome and AD patients [27]. Moreover, the endosomal-lysosomal compartment is a major site for Aβ production [52]. When assessing levels of the enzymes and proteins involved in Aβ production, we found a strong increase in the active N-terminal fragment (NTF) of presenilin-1 (PS1) at 12μM and 120μM doses. No changes were detected for BACE1 and APOE in the same animals. These findings suggest a potential direct effect of P188 on the γ-secretase pathway. An indirect effect by which P188-induced lysosomal accumulation promoted APP cleavage through the γ-secretase pathway in the lysosome is also plausible. In fact, several components of the γ-secretase complex have been detected in the lysosomal membrane together with APP [53]. In addition, PS1 mutations may affect lysosomal function in the context of AD [54, 55]. P188 treatment might have disturbed feedback regulation between PS1 and lysosomal maturation, causing exacerbation of dystrophic neurite formation and amyloid pathology simultaneously.

Amyloid plaques are efficiently surrounded by microglial cells in AD brains (reviewed in [22, 23]). At the initial stages of amyloid pathology, microglia may play a protective role against Aβ induced dystrophic pathology and promote clearance of amyloid aggregates [56, 57]. Aging and increased amyloid deposition appear to reduce microglial functions to both mitigate Aβ-induced neuritic dystrophy formation and clear amyloid aggregates [56, 57]. Microglia may become chronically activated and induce neuronal damage through the release of inflammatory chemokines (reviewed in [58]). We tested the effect of P188 treatment on microglial activation around amyloid plaques. We found a strong exacerbation of pathogenic microglial activation upon P188 brain delivery. In particular, immunoreactivity for pathogenic microglial marker CD68 [24] was increased at 12μM and 120μM doses. Pathogenic microglial activation strictly corresponded to the increase in amyloid burden. We hypothesize that P188 effects on microglial activation are secondary to exacerbation of amyloid pathology.

Dystrophic neurites are characterized by extensive disruption of microtubular architecture and impaired axonal transport [2]. Hyperphosphorylation and pathogenic accumulation of tau protein are key hallmarks of AD (reviewed in [59]). In fact, tau phosphorylation promotes its disengagement from microtubules and reduces microtubule stability [60]. Recent studies have shown that dystrophic neurites are ideal sites for pathogenic tau accumulation [61]. 5XFAD mice do not display tau pathology, but the accumulation of early phosphorylated epitopes of tau (Ser396 and Ser404) has been reported [26]. P188 treatment promoted a significant decrease of total tau at 12μM and 120μM doses. No change was detected in the levels of phosphorylated tau at Ser396 and Ser404 either as an absolute value or as a ratio to total tau. Although we did not assess phosphorylation of tau at other sites, we speculate that reduction of total tau could suggest an increased disruption of microtubules and potential neurodegeneration upon P188 exposure.

To better characterize endo-lysosomal function in the brain of our animals, we measured levels of lysosomal protease cathepsin D. Defective activation of this enzyme has been linked to the development of dystrophic neurites in AD mouse models [2]. Moreover, cathepsin D is involved in the clearance of both Aβ and tau [62-64]. We showed that levels of inactive pro-cathepsin D increased at the highest 120μM P188 dose. At the same dose, active mature-cathepsin D was also increased. These findings suggest that P188-induced lysosomal dysfunction is a consequence of exacerbated amyloid pathology.

Finally, we assessed the effect of P188 on SYT7, a calcium sensor that regulates vesicular fusion to the plasma membrane in response to calcium influx [29, 30]. This SNARE complex protein is crucial in promoting membrane repair by means of lysosomal exocytosis [28] and is localized both at the presynaptic plasma membrane and at lysosomes in neurons [29, 65, 66]. We found a strong decrease in the large isoform of SYT7 at both 12μM and 120μM doses. These data raise the possibility that P188 might have interfered with vesicular trafficking dynamics at the presynaptic plasma membrane, further exacerbating lysosomal accumulation and dysfunction. Moreover, SYT7 levels and localization may be regulated by APP and presenilin [66, 67]. It is possible that SYT7 abnormalities were triggered by dysregulation of the γ-secretase pathway. Alternatively, the decreased large isoform over the small isoform could suggest proteolytic degradation of the large isoform due to neurodegeneration or altered RNA splicing of the STY7 large isoform.

CONCLUSION

Brain delivery of membrane resealing agent P188 strongly exacerbated AD pathology in a well-established mouse model of AD. Its deleterious effects may be initiated by dysregulation of the γ-secretase pathway and/or vesicular trafficking. According to our study, we conclude that P188 should not be considered as a therapeutic agent against AD. Further characterization of plasma membrane dynamics governing APP processing and lysosomal maturation could offer better therapeutic avenues. In fact, targeting specific vesicular trafficking pathways at the plasma membrane might be a more successful approach against Aβ-induced dystrophic neurite pathology in AD.