α-synuclein conformational antibodies fused to penetratin are effective in models of Lewy body disease.

OBJECTIVE: Progressive accumulation of α-synuclein (α-syn) has been associated with Parkinson's disease (PD) and Dementia with Lewy body (DLB). The mechanisms through which α-syn leads to neurodegeneration are not completely clear; however, the formation of various oligomeric species have been proposed to play a role. Antibody therapy has shown effectiveness at reducing α-syn accumulation in the central nervous system (CNS); however, most of these studies have been conducted utilizing antibodies that recognize both monomeric and higher molecular weight α-syn. In this context, the main objective of this study was to investigate the efficacy of immunotherapy with single-chain antibodies (scFVs) against specific conformational forms of α-syn fused to a novel brain penetrating sequence.
METHOD: We screened various scFVs against α-syn expressed from lentiviral vectors by intracerebral injections in an α-syn tg model. The most effective scFVs were fused to the cell-penetrating peptide penetratin to enhance transport across the blood-brain barrier, and lentiviral vectors were constructed and tested for efficacy following systemic delivery intraperitoneal into α-syn tg mice. RESULT: Two scFVs (D5 and 10H) selectively targeted different α-syn oligomers and reduced the accumulation of α-syn and ameliorated functional deficits when delivered late in disease development; however, only one of the antibodies (D5) was also effective when delivered early in disease development. These scFVs were also utilized in an enzyme-linked immunosorbent assay (ELISA) assay to monitor the effects of immunotherapy on α-syn oligomers in brain and plasma.
INTERPRETATION: The design and targeting of antibodies for specific species of α-syn oligomers is crucial for therapeutic immunotherapy and might be of relevance for the treatment of Lewy body disease.

Background

Progressive accumulation of α‐synuclein (α‐syn) in the CNS and peripheral organs has been associated with the pathogenesis of Parkinson's disease (PD), Dementia with Lewy body (DLB), multiple system atrophy (MSA), and other related neurodegenerative disorders.1, 2, 3, 4, 5, 6 α‐syn can adopt multiple conformations, including various oligomeric species, under physiological and pathological conditions; however, it is not completely clear which aggregated forms of α‐syn are toxic. Aggregated forms of α‐syn were shown to induce toxicity in dopaminergic neurons in vivo7 and several different oligomeric morphologies were shown to each have different toxic mechanisms and effects on cells.8, 9 Oligomeric but not fibrillar forms of α‐syn were shown to be toxic to neuronal cells,10 and toxic oligomeric α‐syn forms have been identified in living cells11 and in human plasma from PD patients.12

Intracellular targeting and clearance of oligomeric α‐syn provided better protection against mammalian cell toxicity induced by α‐syn overexpression compared to targeting and clearance of monomeric α‐syn.13 Clearly different aggregated forms of α‐syn are likely to be a contributing factor in PD; however, it is not clear what form of α‐syn represents the best therapeutic target. Since α‐syn has protective functions in neurons including chaperone activity (reviewed in14), an effective α‐syn based therapeutic for PD should selectively target toxic oligomeric α‐syn species and not interfere with cellular functions of other beneficial forms of α‐syn.

Both active and passive immunization strategies have been utilized in animal models and human patients to alleviate the effects of toxic protein aggregation associated with various neurodegenerative diseases.15, 16, 17, 18, 19 Initial immunotherapeutic studies showed that vaccination with the full human α‐syn protein in α‐syn tg mouse models of DLB decreases accumulation of aggregated α‐syn and reduces neurodegeneration.20 Along these lines, active immunization with small peptides that mimic α‐syn epitopes (AFFITOPEs) have shown to be effective at reducing α‐syn accumulation and deficits in three different tg models of synucleinopathy.21, 22 Likewise, passive immunization with monoclonal antibodies that recognize a C‐terminus epitope of α‐syn have been shown to be most effective at reducing accumulation and propagation of α‐syn in tg mouse models of PD.17, 23, 24, 25 Moreover, antibodies against the N‐terminus of α‐syn have been shown to be effective at clearing α‐syn aggregates, reducing α‐syn propagation and diminishing motor deficits in an AAV‐α‐syn model.26, 27 Together, these reports support the value of immunotherapy against α‐syn as a disease‐modifying option for synucleinopathies.

Most of these studies have utilized antibodies that recognize both monomeric and aggregated α‐syn and given the potential toxicity of oligomeric α‐syn it is important to develop more specific therapeutics. Single‐chain antibody fragments (scFVs) are an attractive strategy for immuno‐therapeutics due to their small size, ability to target selected species utilizing phage display strategies and to be genetically manipulated with targeting sequences to enhance intracellular trafficking (reviewed in28). In this context, we have previously shown that a scFv targeting α‐syn oligomers (D5) fused to a CNS targeted peptide (apoB) reduced the accumulation of α‐syn and facilitated the clearance of oligomers via the ESCRT pathway.15

While for the previous study15 we focused on investigating the effects of the scFV in PDGF‐α‐syn wt transgenic model of Lewy body disease; for the present follow‐up study, we compared the efficacy of various scFVs that recognized monomeric versus oligomeric α‐syn10, 29 in a different mouse model that display PD‐like pathology and deficits (mThy1‐α‐syn Line 61).30 These scFVs were fused to a novel cell permeating peptide, also known as penetratin. Penetratin is a 16 amino acid peptide derived from the Antennapedia homoeodomain of Drosophila. The peptide is able to translocate cellular and nuclear lipid bilayers in an energy independent, nonreceptor mediated, nonendocytic vesicle manner.31 Pentratin fused to cargo proteins have been shown to translocate the BBB and even deliver proteins including scFVs to neurons in the CNS.32 In contrast to the low‐density lipoprotein (LDL) receptor‐binding domain of ApoB that we previously used, penetratin will not target the scFV antibody to the lysosome for degradation within the cell once it is endocytosed.15 For this reason, we fused the penetratin cell peptide along with a secretory signal derived from the CD5 protein33 for secretion of the scFV antibodies for delivery to the mouse.

For this study, we administered different scFvs (D5, D10, H10) fused to penetratin via intraperitoneal (IP) lentiviral injection at two different time points; at 3 months of age before substantial α‐syn accumulation occurs to test whether the scFvs have potential value as prophylactics to prevent disease, and at 6–9 months of age to test whether the scFvs have value as treatments to ameliorate α‐syn aggregation and toxicity during later stages of the disease. We show that in the PD mouse model, whereas both of the scFVs were effective at reducing α‐syn pathology and restoring neuronal health when administered late during the course of the disease, only one (D5) of these was also effective at reducing α‐syn pathology and toxicity during very early stages of disease progression. These results suggest that the most effective therapeutic targets may change with disease progression.

Methods

Construction of lentivirus vectors

The 10H and D5 scFvs recognize two distinct α‐syn oligomeric variants.10, 29 While D5 binds to dimers and tetramers,10 10H binds higher molecular weight aggregates corresponding to trimers and hexamers.10 These scFvs did not bind to monomeric α‐syn or cross‐react with β‐amyloid aggregates.29 In contrast, the third anti‐α‐syn scFv, D10 recognizes monomeric, oligomeric, and fibrillar forms.29, 34 The cDNAs for the three scFvs against α‐syn (D5, 10H, and D10)10 were PCR amplified and cloned into the third‐generation self‐inactivating lentivirus vector plasmid35 with the CMV promoter driving expression and the secretory signal from the human CD5 gene33 and the penetratin sequence producing the vectors lentivirus (LV)‐Pen‐D5, LV‐Pen‐10H, and LV‐Pen‐D10.

The LV‐control and LV‐α‐syn, expressing human α‐syn, have been previously described.36 The LV‐D5 and LV‐D5‐apoB viruses have been previously characterized.37 LV were prepared by transient transfection in 293T cells.38

Establishment of a neuronal cell line expressing α‐syn and scFv and in vitro experiments

For these experiments, we used the rat neuroblastoma cell line B103. This model was selected because overexpression of α‐syn in these cells results in mitochondrial alterations, reduced cell viability, defective neurite outgrowth and abnormal accumulation of oligomeric α‐syn.36 For all experiments, cells were infected with LV expressing wt α‐syn at a multiplicity of infection of 20. Cells were coinfected with LV‐Pen‐D5, LV‐Pen‐D10, LV‐Pen‐10H or empty vector (LV‐control). After infection, cells were incubated in a humidified, 5% CO2 atmosphere at 37°C. All experiments were conducted in triplicate to ensure reproducibility.

Cultured supernatent containing Pen‐D5, Pen‐10H, Pen‐D10, or control were prepared by BBS transfection of 293T HEK cells with the plasmids pLV‐PenD5, pLV‐Pen‐10H, pLV‐Pen‐D10, or pLV‐Control. After 48 h transfection, supernatent was collected and filtered with a 0.22 μmol/L cellular acetate filter (Corning, NY) to remove cellular debris. Protease inhibitor cocktail was added to the supernatent (Roche, Indianapolis, INx) to prevent protease degradation of scFvs during 4°C storage.

For coculture analysis, 5 × 104 neuroblastoma B103 cells were plated onto poly L‐lysine coated glass coverslips or onto 12 well culture inserts containing a 0.4 μm PET membrane (Fisher Scientific, Waltham, MA) as previously described.24 Cultures were incubated separately for 6 h to allow cells to attach and then cocultured for 24 h. For one set of experiments cells in glass coverslips in the bottom were double infected with LV‐α‐syn and either LV‐Pen‐D5, LV‐Pen‐10H, LV‐Pen‐D10, or LV‐control. For another set, cells in glass coverslips in the bottom were double infected with LV‐α‐syn and neuronal cells in the top (insert) were infected with LV‐Pen‐D5, LV‐Pen‐10H, LV‐Pen‐D10, or LV‐control. At the conclusion of all experiments, coverslips were washed twice with ice‐cold phosphate‐buffered saline (PBS) and then fixed with 4% PFA for further analysis.

Immunoblot analysis

Cells were infected with LV for 72 h and then lysed in TNE buffer (50 mmol/L Tris‐HCl, pH 7.4, 150 mmol/L NaCl, 1 mmol/L EDTA; all from Sigma‐Aldrich, St. Louis, MO) containing 1% Nonidet P‐40 (Calbiochem, San Diego, CA) with protease and phosphatase inhibitor cocktails (Roche). Total cell extracts were centrifuged at 6000g for 15 min, and the protein concentration of supernatants was assayed with a BCA protein assay kit (Pierce Biotechnology, Waltham, MA). For Western blot analysis, 20 μg of lysate per lane was loaded into 4–12% Bis‐Tris SDS‐PAGE gels and blotted onto polyvinylidene fluoride membranes. Blots were incubated with antibodies against α‐syn (Millipore), V5 epitope tag (Life Technologies, Carlsbad, CA) and actin (Millipore, San Diego, CA) followed by secondary antibodies tagged with horseradish peroxidase (Santa Cruz Biotechnology, Inc. Santa Cruz, CA), visualized by enhanced chemiluminescence and analyzed with a Versadoc XL imaging apparatus (BioRad, Irvine, CA). Analysis of actin levels was used as a loading control.

α‐syn scFV ELISA

The phage capture ELISA protocol used to detect the various oligomeric targets has been described previously.39, 40 The capture scFvs were bound to the wells of high‐binding ELISA plates (Costar, Corning, NY) for 30–60 min at 37°C. All incubation steps were carried out at 37°C for 30–60 min. These scFvs were produced using HB2151 bacteria cells containing the plasmid of interest. The details of our scFv expression protocol has also been published previously.39, 40 After the incubation period the plates were then washed three times with 0.1% PBS with Tween‐20. Nonspecific binding sites were blocked using 2% milk followed by addition of the mice brain tissue (0.1–0.5 mg/mL) or sera (1/100 v/v dilution). Bound antigens were detected using our 40 mmol carboxyl biotinylated D10 detection phage (The protocols for phage production and biotinylation have been published earlier39, 40). Next, avidin‐HRP was added to the wells followed by the SuperSignal ELISA Femto Maximum Sensitivity Substrate (Thermo Scientific, Waltham, MA, USA). Binding intensities were measured using the Wallac Victor2 microplate reader and signal ratios calculated relative to the controls (PBS or normal mice) using these raw absorbance values.

Transgenic mouse lines and injections of lentiviral vectors

For this study, mice overexpressing α‐syn from the murine Thy1 promoter (Line 61) (6 month old, male, C57/Bl6) were utilized.30 This model was selected because mice from this line develop intraneuronal α‐syn aggregates distributed through the brain similar to PD and DLB41 resulting in neurodegeneration, motor, and nonmotor deficits.42, 43 A cohort of n = 24 non‐tg (n = 6 per group) and 24 tg (n = 6 per group) (LV‐control, LV‐Pen‐D5, LV‐Pen‐D10, and LV‐Pen‐10H) were injected with 3 μL of the lentiviral preparations (2.5 × 107 TDU) into the hippocampus (using a 5 μL Hamilton syringe). Briefly, as previously described,44 mice were placed under anesthesia on a Koft stereotaxic apparatus and coordinates (hippocampus: AP 2.0 mm, lateral 1.5 mm, depth 1.3 mm) were determined as per the Franklin and Paxinos atlas.45 The lentiviral vectors were delivered using a Hamilton syringe connected to a hydraulic system to inject the solution at a rate of 1 μL every 2 min. To allow diffusion of the solution into the brain tissue, the needle was left for an additional 5 min after the completion of the injection. Mice survived for 6 weeks after the lentiviral injection.

At first to evaluate the effects of the penetratin peptide to improve the trafficking of the single‐chain antibodies into the CNS non‐tg (6 month old, male, C57/Bl6) mice (n = 6, per group) were injected intraperitoneal (IP) with 100 μL of the lentiviral preparations (2.5 × 107 TDU) with LV‐control, LV‐D5 (alone), LV‐Pen‐D5, and LV‐D5‐apoB and analyzed 4 weeks after. To determine the central efficacy of the systemically delivered scFvs, briefly as previously described mice were injected IP with 100 μL of the lentiviral preparations (2.5 × 107 TDU) (LV‐control, LV‐Pen‐D5 or LV‐Pen‐10H). Mice were separated by age into two groups as follows: early group (3 months old) non‐tg (n = 6, LV‐control) and α‐syn tg (LV‐control, LV‐Pen‐D5 or LV‐Pen‐10H, n = 4 per group) and late group (6–9 months old) non tg (LV‐control, n = 8), α‐syn tg (LV‐control, LV‐Pen‐D5 or LV‐Pen‐10H, n = 5 per group). Mice survived for 3 months after the lentiviral injection.

Following NIH guidelines for the humane treatment of animals, blood was drawn and mice were anesthetized with chloral hydrate and flush‐perfused transcardially with 0.9% saline. Brains and peripheral tissues were removed and divided sagitally. The right hemibrain was postfixed in phosphate‐buffered 4% PFA (pH 7.4) at 4°C for 48 h for neuropathological analysis, whereas the left hemibrain was snap‐frozen and stored at −70°C for subsequent protein analysis.

Animal care

All experiments described were carried out in strict accordance with good animal practice according to NIH recommendations, and all procedures for animal use were approved by the Institutional Animal Care and Use Committee at the University of California at San Diego (UCSD) under protocol #S02221.

Immunocytochemical and neuropathological analyses

Analysis of the scFV (V5 tagged) penetration in the CNS and α‐syn accumulation was performed in serially sectioned, free‐floating, blind‐coded vibratome sections from α‐syn tg and non‐tg mice treated with LV‐PenD5, LV‐Pen‐D10, LV‐Pen‐10H, and LV‐control vectors by immunocytochemistry.36 For detection of the scFv, sections were incubated overnight at 4°C with an anti‐V5 antibody and for α‐syn with an antibody against total α‐syn (affinity purified rabbit polyclonal, Millipore),46 followed by biotinylated goat anti‐rabbit IgG (Vector Laboratories, Inc. Burlingame, CA), Avidin D‐HRP (ABC Elite; Vector) and detection with the diaminobenzidine to determine the levels of V5 or α‐syn immunoreactivity in the neuropil in the neocortex, hippocampus, striatum, and substantia nigra. To detect aggregated α‐syn sections were pretreated with proteinase K (PK) for 8 min at 10 μg/mL as previously described41 and immunostained with antibody against total α‐syn. For each case, three sections were analyzed using a digital video microscope (Olympus BX51, Tokyo, Japan) with the Image‐Pro Plus analysis system integrated. Results were averaged and expressed as corrected optical density or as % of the neuropil occupied by V5 or α‐syn immunostained structures.

To determine if the treatments with the various scFv expressed from lentiviral vectors ameliorated the neurodegenerative alterations associated with the expression of α‐syn, briefly as previously described,36 blind‐coded, 40‐μm‐thick vibratome sections from mouse brains fixed in 4% paraformaldehyde were immunolabeled with the mouse monoclonal antibodies against NeuN (neuronal marker, Millipore) or glial fibrillary acidic protein (GFAP, astroglial marker, Millipore).36 After overnight incubation with the primary antibodies, sections were incubated with biotinylated secondary antibody and reacted with diaminobenzidine. All sections were processed under the same standardized conditions. The immunolabeled blind‐coded sections were evaluated using a digital video microscope (Olympus BX51) for NeuN sections were analyzed with the Stero‐Investigator software using the disector method.47 For the GFAP immunostained sections the Image‐Pro Plus analysis program system was utilized and results expressed as corrected optical density.36

To determine if the scFV against different conformations of α‐syn (Pen‐D10, Pen‐D5 and Pen‐10H) colocalized with endogenous or transgenic α‐syn double labeling experiments were performed with vibratome sections from brains of the non‐tg (n = 4) and α‐syn tg (n = 4) mice. For this purpose, purified conditioned media from 293T cells expressing control, Pen‐D10, Pen‐D5 or Pen‐10H with antibodies against V5 to detect the scFv and α‐syn (SYN1, BD Biosciences, San Diego, CA). Followed by secondary antibodies tagged with fluorescein isothiocyanate (FITC) to detect α‐syn and Tyramide red to detect the V5 tag. Sections were imaged with a Zeiss, Jenna, Germany 63X (N.A. 1.4) objective on an Axiovert 35 microscope (Zeiss) with an attached MRC1024 LSCM system (BioRad).46 Double‐labeled sections were analyzed with the Image J program to determine the level of colocalization in 10 random fields in the neocortex either in the neuropil (non‐tg) or neuronal inclusions (α‐syn tg). Double labeling experiments were performed to investigate the colocalization between α‐syn or MAP2 (neuronal marker) and the scFV in neuronal cells in vitro and in vivo in the brain sections from non‐tg and α‐syn tg mice. For this purpose, tissues were incubated with antibodies against α‐syn (rabbit polyclonal), MAP2 (Millipore), and V5 (mouse monoclonal, to detect the scFv) followed by secondary antibodies tagged with FITC to detect α‐syn and Tyramide red to detect the V5 tag. All sections were processed simultaneously under the same conditions and experiments were performed twice in order to assess the reproducibility of results. Sections were imaged with a Zeiss 63X (N.A. 1.4) objective on an Axiovert 35 microscope (Zeiss) with an attached MRC1024 LSCM system (BioRad).46

Adhesive removal and horizontal beam test

Motor response to sensory stimuli was measured with a stimulation test in mice.42 A small adhesive label was placed on the snout of the mouse, and the time to make contact and remove the stimulus was recorded. To remove the stimulus, animals would raise both forelimbs toward their face and swipe off the stimulus with both forepaws. Each animal received two trials, and the trials are run consecutively with a 15 s interval. All testing was performed in the animal's home cage, and cage mates were temporarily removed during testing because they can interfere with stimulus removal. Time to removal of stimulus (latency) was recorded with a maximum of 60 sec allowed before the experimenter removed the label. The round beam test allows the assessment of gait and balance impairments through distance traveled in an allotted amount of time over a round beam placed horizontally. As previously described,48 three consecutive trials, 1 min each, were run in 1 day. The number of foot slippages were recorded, and errors on the beam are calculated as foot slips/distance traveled.

Statistical analysis

All experiments were performed blind‐coded and in triplicate. Values in the figures are expressed as means ± SEM. To determine the statistical significance, values were compared using the one‐way ANOVA with post hoc Dunnet when comparing the scFv‐treated samples to LV‐control treated samples. Additional comparisons were done using Tukey–Kramer or Fisher post hoc tests. The differences were considered to be significant if P values were less than 0.05.

Results

scFvs against morphologically distinct α‐syn aggregates recognize α‐syn in the α‐syn tg mouse brain

The D10, D5, and 10H scFv antibodies were initially developed against morphologically distinct forms of α‐syn.10, 13, 15, 29, 40, 49 D5 reacts primarily to synthetically generated α‐syn dimers and tetramers; whereas 10H reacts to synthetically generated trimers and hexamers.49 In contrast, D10 binds to both the monomer form of α‐syn as well as oligomeric and fibrillar forms.49 In order to determine if these antibodies bind to α‐syn in a mouse model of PD/DLB, we used them for immunohistochemistry in the Line 61 α‐syn tg mouse (Fig. 1). In non‐tg mice, endogenous murine α‐syn was detected in small punctate structures throughout the neuropil (Fig. 1A). This was recognized by all three scFvs (D10, D5, and 10H) as detected by the epitope tag V5 on the scFV (Fig. 1A and C). In contrast, in the α‐syn tg mouse which expresses human α‐syn as well as the endogenous mouse α‐syn primarily intracellularly, α‐syn was detected both in punctate structures as well as in larger lewy body‐like structures as previously described for the Line 61 α‐syn tg mouse (Fig. 1B and D).30 The D10 scFv did not appear to bind to the Lewy bodies, whereas both the D5 and 10H antibodies recognized Lewy bodies in the Line 61 α‐syn mouse model of PD/DLB. These in vivo findings are consistent with previous in vitro cell‐free studies showing that D5 and 10H binds aggregated α‐syn only while D10 binds to the monomeric, oligomeric and fibrilar forms of α‐syn.10, 29, 34 In the brain sections, monomeric α‐syn will be the predominant form while the oligomeric forms will be present at much lower levels so the D10 antibody may not be binding to Lewy bodies because it is bound by the monomeric α‐syn throughout the brain.

Figure 1

Double immunocytochemical analysis with scFvs against α‐syn detect α‐syn aggregates in the brains of a tg mouse. Brain vibratome sections from 6‐month‐old (A) non‐tg and (B) α‐syn tg mice were double labeled with antibodies against α‐syn (detected with a secondary FITC tagged) and the scFvs Pen‐D10, Pen‐D5, or Pen‐10H followed by Anti‐V5 (detected with Tyramide Red) and imaged with the laser scanning confocal microscope. Images are from the frontotemporal cortex. The scFVs as detected with the antibody against the V5 tag colocalized with α‐syn in the brains of tg mice. Open boxes (white margins) in the merged panels indicate areas of higher magnification in detail panels. Arrows indicate intraneuronal α‐syn inclusions. (C) Computer aided image analysis of the % of Anti‐V5 (scFV) colocalizing with total neuropil associated α‐syn immunofluorescence in the fronto‐temporal cortex. (D) Computer aided image analysis of the % of Anti‐V5 (scFV) colocalizing with intracellular α‐syn immunofluorescence in the fronto‐temporal cortex. (n = 4, mice per group) #P < 0.05 when comparing α‐syn tg treated with lentivirus (LV)‐control vs LV‐scFVs by one‐way ANOVA post hoc Tukey–Kramer. Scale bar represents 10 μm.

scFvs against morphologically distinct α‐syn aggregates reduce the accumulation of α‐syn in an in vitro neuronal model

We previously showed that targeting D5 reactive oligomeric α‐syn aggregates in a mouse model of PD reduced pathology and neurodegeneration.15 In the previous study, the D5 scFV was introduced into the mouse by ip injection of lentivirus. The D5 scFV contained an apoB LDL receptor‐binding domain to facilitate transfer across the BBB and to also facilitate clearance of the bound oligomeric α‐syn aggregates through the ESCRT‐mediated pathway. Here, we utilized the homeodomain sec/pen domain peptide, penetratin, to facilitate transfer across the BBB. Penetratin enables the scFV antibodies to both enter and exit neuronal cells through a nontraditional secretion pathway.31, 50, 51, 52

All three scFvs (D10, D5, and 10H) recognize morphologically distinct forms of α‐syn, so to examine the ability of these different antibodies to block the accumulation of α‐syn in an in vitro neuronal model, we coinfected B103 neuronal cells with a lentivirus vector overexpressing α‐syn along with a lentivirus vector expressing either the Pen‐D10, Pen‐D5 or Pen‐10H (Fig. 2A). All three scFv coding vectors expressed equivalent amounts of the antibodies, however, in this model, only Pen‐D5 and Pen‐10H were able to reduce the accumulation of α‐syn in the neuronal cells (Fig. 2B and C). The vector expressing Pen‐D10 did not appear to have an effect on the neuronal accumulation of α‐syn. These results are consistent with previous results showing that intracellular clearance of oligomeric α‐syn reduces cellular toxicity more efficiently than clearance of monomeric α‐syn.13

Figure 2

In vitro analysis of the effects of the Pen‐scFvs against α‐syn expressed from lentivirus vectors reducing the accumulation of α‐syn in a neurons in a chamber system. (A) Diagrammatic representation of experiment depicting the B103 neuronal cells and coinfection with lentivirus (LV)‐α‐syn and LV‐Pen‐D10, LV‐Pen‐D5 or LV‐Pen‐10H for 48 h and then evaluated by immunocytochemistry for levels of α‐syn and scFVs (V5 tag). (B) Coverslips were fixed and double immunostained with antibodies against α‐syn (red) and the scFv (V5 epitope tag, green) and imaged with the laser scanning confocal microscope. (C) Relative fluorescence was analyzed to determine levels of α‐syn and scFv immunoreactivity expressed as pixel intensity. (D) Diagrammatic representation of experiment depicting the B103 neuronal cells that were cocultured with cells in the lower chamber on coverslips infected with LV‐α‐syn and cells in the upper chamber infected with LV‐Pen‐D10, LV‐Pen‐D5 or LV‐Pen‐10H for 48 h. Chambers were separated by a 0.22 μm membrane to allow for the passage of only proteins and small molecules. (E) Coverslips were fixed and double immunostained with antibodies against α‐syn (red) and the scFv (V5 epitope tag, green). (F) Relative fluorescence was analyzed to determine levels of α‐syn and scFv immunoreactivity expressed as pixel intensity. *Indicates statistical significance P < 0.05 compared to cell expressing α‐syn and treated with LV‐control. Experiments were performed in triplicate. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 10 μm.

To examine the ability of extracellular scFvs to reduce the accumulation of α‐syn in an in vitro neuronal model, we separated the α‐syn expressing cells from the scFv expressing cells by a 0.4 μm membrane that allows only the passage of proteins but prevents the contact of the cells (Fig. 2D). The Pen‐D10, Pen‐D5, and Pen‐10H lentivector expressing cells were cultured in the top chamber and the α‐syn expressing cells were cultured in the bottom chamber. Analysis of the cells in the lower chamber showed similar levels of uptake of the scFvs for Pen‐D10, Pen‐D5, and Pen‐10H. Both Pen‐D5 and Pen‐10H significantly reduced the accumulation of α‐syn similar to the coinfection experiment (Fig. 2E and F). In contrast to the coinfection experiment, Pen‐D10 also significantly reduced the accumulation of α‐syn in the recipient cells although to a lesser extent than Pen‐D5 or Pen‐10H (Fig. 2E and F). It is interesting to note that the remaining accumulation of α‐syn in the cells following exposure to the various scFVs is unevenly distributed; this may be due to the distinct oligomers recognized by the different antibody fragments and the subcellular localization of those oligomers. Taken together, these studies are consistent with the concept that D5 and 10H are more effective than D10 at reducing α‐syn accumulation and that the addition of the penetratin sequence does not alter activity.

Effects of scFvs against α‐syn at reducing accumulation of α‐syn following intracerebral injection to α‐syn tg mice

To further confirm in vivo which of the scFvs where most effective at reducing α‐syn accumulation, we injected the lentivirus expressing Pen‐D10, Pen‐D5, or Pen‐10H or a control virus into the hippocampus of α‐syn tg mouse30 and analyzed the brains 6 weeks later. The hippocampus was selected because it is easily accessible, displays extensive α‐syn accumulation in this model and is an area affected in patients with PD/DLB. Mice were examined by immunocytochemistry with antibodies against V5 to detect the scFvs and total α‐syn. All three vectors expressed similar levels of scFv with accumulation observed in neuronal cell bodies and throughout the neuropil in the hippocampus (Fig. 3A and B). Similar to previous reports, α‐syn was observed throughout the neuropil in the hippocampus and the cortex in the α‐syn tg mouse in the neuronal cell bodies and extending into the neurites in mice that received that received LV‐control.47 α‐syn tg mice that received the lentivectors expressing either Pen‐D5 or Pen‐10H showed significant reduction in accumulated α‐syn in the hippocampus (Fig. 3C and D); whereas, mice that received the Pen‐D10 lentivirus showed only a slight reduction in α‐syn that was not significant. These results are similar to those observed when we coinfected neuronal cells in vitro. On the basis of these and the in vitro results, we chose to continue experiments in vivo in the α‐syn tg mouse using only the Pen‐D5 and Pen‐10H scFv vectors.

Figure 3

Immunocytochemical analysis of the effects of the Pen‐scFvs against morphologically distinct α‐syn species following stereotaxic intracerebral injection of lentivirus vectors in α‐syn tg mouse. Six‐month‐old α‐syn tg and non‐tg mice received hippocampal injections of lentivirus (LV) and were sacrificed 6 weeks later and serial sections prepared. (A) Immunocytochemical analysis with an antibody against the V5 tag (to detect scFvs) and (B) α‐syn in brain sections from non‐tg and α‐syn tg mice that received a unilateral stereotaxic of LV‐control, LV‐Pen‐D10, LV‐Pen‐D5 or LV‐Pen‐10H. Sections were analyzed with digital bright field video microscope, panels in the top represent an overview at low magnification (20X). Panels in the bottom are higher magnification (400X) of the hippocampus at the injection site. (C) Image analysis of levels expressed as corrected optical density for anti‐V5 immunoreactivity and (D) anti‐α‐syn immunoreactivity in the hippocampus, respectively. *Indicates statistical significance P < 0.05 compared to non‐tg to α‐syn tg treated with LV‐control by one‐way ANOVA with post hoc Dunnet's. #P < 0.05 when comparing α‐syn tg treated with LV‐control vs LV‐scFVs by one‐way ANOVA post hoc Tukey–Kramer. Scale bar in upper panels represents 250 μm and in lower panel 25 μm.

Systemic delivery of lentiviruses expressing scFvs against distinct α‐syn species reduces α‐syn accumulation depending on the time of administration

Stereotaxic delivery of the viral vectors expressing the scFvs into the α‐syn tg mice was effective at reducing the accumulation of α‐syn in the hippocampus; however, in order to deliver the antibodies throughout the whole CNS, we proposed to deliver the scFvs by the blood across the blood–brain barrier. The addition of the penetratin sequence has previously been shown to facilitate the transport of scFvs from the blood to the CNS across the blood‐brain barrier,32 so we delivered the lentivirus vectors by a single i.p. injection to allow for the liver and spleen to express the antibodies into the blood.53 Comparison of delivery of the LV‐Pen‐D5 to the control LV‐D5 or LV‐control showed significantly greater CNS penetration of the scFV that contained the penetratin sequence (Fig. S1). Increased uptake was observed in the neocortex and the hippocampus and was similar to the uptake observed by the addition of the ApoB LDL‐R transport tag (Fig. S1).37 Two different age groups of α‐syn tg mice were utilized in this study: early–representing mice from 3 months of age, and late–representing mice from 6 to 9 months of age. The early group of α‐syn tg mice were chosen as this is a time point typically before the observance of significant accumulation of neuronal α‐syn and neuronal degeneration.54 In contrast by 6–9 months of age, the Line 61 α‐syn tg mice have significant intraneuronal α‐syn accumulation and early neuronal degeneration.30, 42 Thus the early group represents prophylactic treatment prior to the accumulation of α‐syn, whereas the late group represents a therapeutic approach following the accumulation of α‐syn in the CNS. Both groups of mice were analyzed 3 months after the lentivirus injection.

Immunocytochemistry with an antibody against V5 showed that compared to LV‐control into non‐tg and α‐syn tg mice, delivery of the Pen‐D5 and Pen‐10H lentivectors by i.p. injection to α‐syn‐tg mice resulted in widespread accumulation of both Pen‐D5 and Pen‐10H 3 months after injection in both early and late groups of mice (Fig. 4). scFv immunoreactivity was observed across the whole CNS in particular in the hippocampus and neocortex (Fig. 4A), two areas that show accumulation of α‐syn in this tg mouse model of DLB/PD.30

Figure 4

Immunocytochemical analysis of levels of CNS accumulation of Pen‐scFv antibody (V5) following systemic delivery of a lentivirus vector expressing Pen‐scFvs against α‐syn. The non‐tg and α‐syn tg mice (early and late groups) received a single intraperitoneal injection of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H. Three months after injection, mice were sacrificed and serial sections were prepared. (A) Sections were immunostained with antibodies against V5 to identify the scFv. Upper panels are an overview of the brain sections at low magnification (20X), the lower panels are higher magnification (400X) areas (dotted squares) of the neocortex and hippocampus. Arrows indicate neuronal immunostaining. (B) Image analysis of levels of scFv immunoreactivity (V5 immunoreactivity) expressed as corrected optical density from early group (3 month old at start and 6 month old at the end of the experiment) mice. (C) Image analysis of levels of scFv immunoreactivity (V5 immunoreactivity) expressed as corrected optical density of sections from late group (6–9 month old at start and 9–12 month old at the end of the experiment) mice. *Indicates statistical significance P < 0.05 compared to non‐tg mice. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 200 μm in low power images and 40 μm in high power images.

To determine if the Pen‐scFV colocalized with α‐syn, brain sections from mice treated with LV‐control, LV‐Pen‐D5, or LV‐Pen‐D10 were double labeled with antibodies against the V5 tag of the Pen‐scFV and human α‐syn. Compared to mice injected with the LV‐control, mice treated with either Pen‐D5 or Pen‐D10 displayed staining of the Pen‐scFV colocalized to α‐syn immunostaining (Fig. 5A). The Line 61 α‐syn tg mouse exhibits strong intraneuronal α‐syn accumulation so to verify that the Pen‐scFV was localizing to the neurons, brain sections were double labeled for the V5 tag of the Pen‐scFV and the neuronal dendritic marker MAP2. Mice that received the Pen‐D5 or Pen‐D10 showed significant colocalization of the Pen‐scFV with the neuronal marker compared to mice that received LV‐control (Fig. 5B).

Figure 5

The Pen‐scFV antibodies colocalize with α‐synuclein and neurons in the brain. Vibratome brain sections from non‐tg and α‐syn tg that received i.p. injections of lentivirus (LV)‐control or LV‐Pen‐D5 or LV‐Pen‐10H were double immunofluorescence labeled with antibodies against human α‐syn or MAP2 and the Pen‐scFV (V5) and analyzed with the laser scanning confocal microscope. Representative neurons from the frontal cortex are presented, the panels in red corresponds to the scFV (anti‐V5 antibody) and in green to the α‐syn or MAP2. (A) Double immunolabeling for the Pen‐scFV (V5) (red) and human α‐syn (green) with nuclei (DAPI, blue) showing colocalization between the two markers in tg mice treated with Pen‐D5 and Pen‐10H but not with the control. (B) Double immunolabeling for the Pen‐scFV (V5) (red) and the neuronal marker MAP2 (green) with nuclei (DAPI, blue). In tg mice treated with Pen‐D5 or Pen‐10H the scFV colocalizes to MAP2‐positive neurons. Scale bar = 10 μm.

Next, we analyzed by immunocytochemistry the effects of the scFvs on α‐syn accumulation in the absence or following PK digestion. As expected, in the non‐tg mice treated with the LV‐control there was mild labeling of the neuropil. In the α‐syn tg mice treated with LV‐control, there was extensive accumulation of α‐syn in the neocortex, hippocampus, striatum, and substantia nigra (Fig. 6A). In contrast, mice that received Pen‐D5 or Pen‐10H at the early age showed a significant reduction in all these brain regions (Fig. 6B–D). Mice aged 9–12 months that received Pen‐D5 or Pen‐10H also displayed a reduction in α‐syn accumulation but to a lesser extent compared to mice that received it earlier (Fig. 6B–D). Furthermore, PK‐resistant α‐syn, representing α‐syn aggregates, showed similar reductions in α‐syn‐tg mice that were treated early (Fig. 7). In contrast, the Pen‐D5 and Pen‐10H scFvs appeared to have a greater effect on reducing PK‐resistant α‐syn in the older mice in the neocortex, hippocampus, and substantia nigra than reducing total α‐syn. In the substantia nigra, the LV‐Pen‐D5 was more effective at reducing the accumulation of both total α‐syn and PK‐resistant α‐syn in the young mice compared to mice treated with LV‐Pen‐10H (Figs. 6E and 7E). In older mice, the two vectors appeared to be equally effective. Similarly, LV‐Pen‐D5 was more effective at reducing PK‐resistant α‐syn in the striatum in young mice compared to treatment with LV‐Pen‐10H while both vectors appeared to be equally effective in older mice in the same region (Fig. 7C).

Figure 6

Systemic delivery of lentivirus vector expressing Pen‐scFvs against α‐syn reduces CNS accumulation of total α‐syn. The non‐tg and α‐syn tg mice (early and late groups) received a single intraperitoneal injection of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H. Three months after injection, mice were sacrificed and serial sections were prepared for immunocytochemical analysis with an antibody against total α‐syn and imaged with a bright field digital video microscope. (A) Representative images of the neocortex, hippocampus, striatum and substantia nigra of the early (3 months old at start) group of non‐tg and α‐syn tg mice immunostained with an antibody against α‐syn. (n) neuropil (arrows) neuronal cell bodies. Image analysis of levels of α‐syn immunoreactivity expressed as corrected optical density of sections from early (3 month old) mice and late (6–9 month old) mice in the (B) neocortex, (C) striatum, (D) hippocampus, and (E) substantia nigra. *Indicates statistical significance P < 0.05 compared to non‐tg mice. #Indicates statistical significance P < 0.05 compared to α‐syn tg mice treated with LV‐control. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 40 μm in high power images.

Figure 7

Effects of systemically delivered lentivirus vectors expressing Pen‐scFvs against α‐syn on the accumulation of proteinase K (PK)‐resistant α‐syn in the CNS. The non‐tg and α‐syn tg mice (early and late groups) received a single intraperitoneal injection of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H. Three months after injection, mice were sacrificed and serial sections were pretreated with PK followed by immunocytochemical analysis with an antibody against total α‐syn and imaged with a bright field digital video microscope. (A) Representative images of the neocortex, hippocampus, striatum and substantia nigra of the early (3 month old at start) group of non‐tg and α‐syn tg mice. (arrowheads) dystrophic neurites and (arrows) neuronal cell bodies. (B) Image analysis of levels of PK‐resistant α‐syn immunoreactivity expressed as corrected optical density of sections from early (3 month old) mice and late (6–9 month old) mice in the (B) neocortex, (C) striatum, (D) hippocampus, and (E) substantia nigra. #Indicates statistical significance P < 0.05 compared to α‐syn tg mice treated with LV‐control. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 40 μm in high power images.

Next, we wanted to confirm the neuropathology by an independent biochemical method. For this purpose, we used the scFvs as capture antibodies in an ELISA assay to screen for the levels of D5 or 10H morphologic specific α‐syn following systemic treatment with the Pen‐scFvs. Brain homogenates from each of the age and treatment groups was added to wells containing the D5 or 10H capture antibody and then detected with a pan‐α‐syn antibody. As expected, α‐syn tg mice treated with LV‐control contained significantly greater levels of D5 and 10H morphologic α‐syn in both early and late age groups (Fig. 8). In contrast, mice treated with the LV‐Pen‐D5 showed significantly reduced levels of both D5 and 10H α‐syn morphologic forms in the early and late age groups treated (Fig. 8A–D). LV‐Pen‐10H treated α‐syn tg mice showed reduced levels of 10H morphologic α‐syn in early and late age groups; however, only reduced levels of D5 morphologic α‐syn were observed in the late age group (Fig. 8A–D). Early age group mice showed a reduction in D5 α‐syn oligomers only when treated with LV‐Pen‐D5 but not when treated with LV‐Pen‐10H; however, treatment of the same mice later in age with the LV‐Pen‐10H was effective at reducing the D5 oligomers as measured in the ELISA (Fig. 8A and B). In contrast, the early age group treated with LV‐Pen10H showed a reduction in the 10H morphologic α‐syn levels in the brain homogenates (Fig. 8A).

Figure 8

Analysis by ELISA of the levels of distinct α‐syn oligomeric species in the CNS of mice following systemic treatment with Pen‐scFv against α‐syn. α‐syn tg or non‐tg mice received a single intraperitoneal injection of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H. Three months after injection, mice were sacrificed and whole brain homogenates were prepared and analyzed by ELISA. The D5 scFv was used as a capture antibody to assay (A) early (3 month old) mice and (B) late (6–9 month old) mice. The 10H scFv was used as a capture antibody to assay (C) early (3 month old) mice and (D) late (6–9 month old) mice. *Indicates statistical significance P < 0.05 compared to non‐tg mice. #Indicates statistical significance P < 0.05 compared to α‐syn tg mice treated with LV‐control. One‐way ANOVA with post hoc Tukey–Kramer.

scFvs against distinct α‐syn aggregates ameliorates neurodegeneration and behavioral deficits in α‐syn tg mice

We have previously shown that the Line 61 α‐syn tg mouse suffers from neuronal degeneration in the CA3 of the hippocampus and increased astrogliosis similar to patients with PD/DLB.55 In fact, compared to non‐tg LV‐control treated mice, we observed reduced numbers of NeuN‐positive pyramidal cells in the CA3 region of the hippocampus of LV‐control treated α‐syn tg treated mice and this was reversed in mice treated with either LV‐Pen‐D5 or LV‐Pen‐10H treated at either the early or late age time point (Fig. 9A–C). Astrogliosis was observed in both the frontal cortex and the hippocampus as increased immunohistochemical staining of GFAP in the LV‐control treated α‐syn tg mice. This was partially alleviated by treatment with the LV‐Pen‐D5 or LV‐Pen‐10H in both the early and late group of mice in the frontal cortex and the hippocampus (Fig. 9D–F). This suggests that delivery of the Pen‐D5 or Pen‐10H not only reduces the accumulation of α‐syn but also ameliorates the neurodegeneration and partially the astrogliosis in the α‐syn tg mouse model of PD/DLB.

Figure 9

Effects of systemic delivered lentivirus vectors expressing Pen‐scFvs against α‐syn on neuronal and astroglial pathology in a α‐syn tg mouse. The α‐syn tg and non‐tg mice received intraperitoneal injections of lentivirus (LV)‐Pen‐D5, LV‐Pen‐10H or LV‐control. Three months after injection, mice were sacrificed and brain sections were immunostained with antibodies against the neuronal (NeuN) and astroglial cell glial fibrillary acidic protein (GFAP) markers and imaged with a bright field digital video microscope. (A) Representative images from early group (3 month old at start) non‐tg and α‐syn tg mice immunostained with antibodies against the neuronal marker NeuN. The upper panel is a low magnification (20X) overview, while the lower power is a higher magnification (400X) view of the field in the hippocampus (CA3) marked with a solid line rectangle. Stereological estimates (dissector method) of total NeuN‐positive neuronal counts measured in the (B) frontal cortex and (C) hippocampus of early (3 month old at start) and late (6–9 month old at start) mice. (D) Representative images from early group (3 month old at start) of non‐tg and α‐syn tg mice immunostained with antibodies against the astroglial marker GFAP. The upper panel is a low magnification (20X) overview, whereas the lower power is a higher magnification (400X) view of the field in the hippocampus (CA3) marked with a solid line rectangle. Image analysis of levels of GFAP immunoreactivity expressed as corrected optical density of sections from the (E) frontal cortex and (F) hippocampus. *Indicates statistical significance P < 0.05 compared to non‐tg mice. #Indicates statistical significance P < 0.05 compared to α‐syn tg mice treated with LV‐control. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 200 μm in low power images and 40 μm in high power images.

To determine if the reduction in brain α‐syn and reduced neurodegeneration had functional consequences, behavioral analysis with adhesive removal test as a measure of coordination and motor control42, 56 was performed. α‐syn tg mice examined in the adhesive removal test showed a nearly twofold increase in time to remove the sticker compared to age‐matched non‐tg mice for both the early and late groups (Fig. 10A and B). Treatment with LV‐Pen‐D5 was able to completely reverse the behaviorally deficit in the α‐syn tg mice when used to treat the early group, but not the late group (Fig. 10A and B). In contrast, LV‐Pen‐10H was able to reverse the behavioral deficit observed in the late group, but not in the early group (Fig. 10A and B). Further analysis was performed with the horizontal round beam test. Compared to non‐tg controls, the α‐syn tg mice treated with LV‐control displayed increased number of errors in the early and late groups (Fig. 10C and D). Treatment with LV‐Pen‐D5 was able reverse the behaviorally deficit in the α‐syn tg mice when used to treat the early and late groups (Fig. 10C and D). The LV‐Pen‐10H was able to ameliorate the behavioral deficit observed in the late group, but not in the early group (Fig. 10C and D). Thus, the two scFvs had significantly different effects on behavioral outcomes depending on the age of the mice that were treated.

Figure 10

Effects of systemic treatment with a lentivirus vector expressing Pen‐scFvs against α‐syn at ameliorating motor deficits in α‐syn tg mice. The α‐syn tg and non‐tg received a single intraperitoneal injections of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H and 3 months later, motor coordination was assessed with the adhesive removal test. (A) Early group (3 month old at start) mice were examined at the end of the treatment by placing a small adhesive sticker on their nose and the time to removal was recorded. Compared to LV‐control, non‐tg mice, the α‐syn tg mice displayed delayed removal of the adhesive that was improved by the treatment with the LV‐Pen‐D5. (B) Late group (6–9 month old at start) mice were examined similarly at the end of the treatment. Compared to LV‐control non‐tg mice, the α‐syn tg mice displayed delayed removal of the adhesive that was improved by the treatment with the LV‐Pen‐10H. Gait and balance were assessed with the round beam. (C) Early group mice were examined at the end of the treatment and assessed for number of missed foot placements on the beam as errors/10 cm. (D) Late group mice were examined at the end of treatment and assessed for the number of foot placement errors as errors/10 cm. *Indicates statistical significance P < 0.05 when compared with non‐tg controls. #Indicates statistical significance P < 0.05 when compared to α‐syn tg mice that received LV‐control. One‐way ANOVA with post hoc Dunnet's.

Antibodies against distinct α‐syn aggregates can be used to track the effects of scFvs in serum

As shown in the previous section, the scFvs can be utilized as a therapeutic agent when delivered systemically with the LVs or for detection of selected aggregated species of α‐syn in homogenates and fluids by ELISA. In order to determine if the ELISA assay used in the previous section could be utilized as a biomarker assessment for blood following treatment, serum from both mice groups (early and late) were taken following treatment with LV‐control, LV‐Pen‐D5, or LV‐Pen‐10H and assayed for D5 and 10H morphologic α‐syn. Early (3–6 month) α‐syn tg mice showed a significant increase in the D5 morphologic α‐syn in the serum compared to non‐tg mice (Fig. 11A). This was reduced by treatment with either LV‐PenD5 or LV‐Pen10H (Fig. 11A). In contrast, late mice (9–12 month) did not show elevated levels of D5 morphologic α‐syn above those of non‐tg mice (Fig. 10B). The levels of 10H morphologic α‐syn in the serum were markedly higher than non‐tg mice in the early mice group (twofold) and still significantly higher in the late mice group (1.5‐fold) (Fig. 11C and D). Treatment of either the early or late group of mice with LV‐Pen‐D5 or LV‐Pen‐10H reduced levels of the 10H morphologic α‐syn in the serum to levels similar to non‐tg mice (Fig. 11C and D).

Figure 11

The scFVs antibodies against distinct α‐syn oligomers can be detected by α‐syn ELISA in the serum and monitor treatment effect. The α‐syn tg or non‐tg mice received a single intraperitoneal injection of lentivirus (LV)‐control, LV‐Pen‐D5 or LV‐Pen‐10H. Three months after injection, serum was collected and assayed by ELISA using the scFvs as capture antibodies. The D5 scFv was used as a capture antibody to assay (A) early (3 month old) mice and (B) late (6–9 month old) mice. The 10H scFv was used as a capture antibody to assay (C) early (3 month old) mice and (D) late (6–9 month old) mice. *Indicates statistical significance P < 0.05 compared to non‐tg mice. #Indicates statistical significance P < 0.05 compared to α‐syn tg mice treated with LV‐control. One‐way ANOVA with post hoc Tukey–Kramer.

Discussion

This study showed that scFvs against oligomeric α‐syn (D5, 10H) fused to penetratin to enhance transport across the BBB were effective at reducing the accumulation of α‐syn and ameliorating neurodegenerative and functional deficits when systemically delivered with lentiviral vectors. One of the scFvs (D5) was more effective when delivered early in disease development, whereas the other antibody (10H) reduced the α‐syn and related deficits when administered to mice at later stages (Table 1). α‐syn plays a critical though poorly understood role in PD.1, 2, 3, 4, 5, 6 While the presence of insoluble fibrilar aggregates of α‐syn in Lewy bodies and neurites are hallmark characteristics of PD, smaller oligomeric aggregates of α‐syn have increasingly been implicated in toxicity and spread of pathology.57, 58, 59, 60, 61 We previously demonstrated that selectively targeting and clearing an oligomeric form of α‐syn recognized by the D5 scFv fused to the BBB targeting sequence of apoB reduced α‐syn pathology and restored neuronal health.15 At the cellular level, the mechanism for clearance of the α‐syn/scFv complex involved targeting to the autophagy/lysosomal pathways via ESCRT pathway. Here, we expand on these studies by alternatively targeting two morphologically different α‐syn aggregate species, recognized by the antibody fragments D5 and 10H.10, 29 We utilized a novel peptide tag, the sec/pen tag from the homeodomain protein, which enables transfer into and out of cells,31, 50, 51, 52 to facilitate transfer across the BBB. We administered the scFvs to two different age PD mice cohorts: one 3 months of age before significant α‐syn deposition has taken place to determine preventative effects of the scFvs; and a second 6–9 months of age after significant α‐syn accumulation has occurred to determine therapeutic benefits of the scFvs. Interestingly, while both the D5 and 10H scFvs provided substantial therapeutic benefit when administered to mice from 6 to 9 months of age, D5 showed greater preventative effects than 10H when administered to mice at 3 months of age.

Table 1

Summary of effects of lentivirus (LV)‐Pen‐D5 or LV‐Pen‐10H treatment on α‐syn tg mice treated early (3 month) or late (6–9 months) in age

	D5 (Early)	D5 (Late)	10H (early)	10H (Late)
SYN	Yes	Trend	Yes	Trend
SYN PK+	Yes	Trend	Yes	Trend
10H Elisa	Yes	Yes	Yes	Yes
D5 Elisa	Yes	Yes	No	Yes
Behavior	Yes	No	No	Yes
Blood 10H	Trend	Yes	Trend	Yes
Blood D5	Yes	No	Yes	No
NeuN	Yes	Yes	No	Yes
GFAP	Yes	Yes	Yes	Trend

Although some evidence suggests that larger species of α‐syn are toxic including proto‐fibrils, there is significant evidence suggesting small oligomers (dimers – hexamers) cause toxicity to neurons.57, 58, 59, 60, 61 Additionally, evidence shows that small oligomers are involved in the cell‐to‐cell propagation of synuclein in PD/DLB.62, 63, 64, 65, 66 Both D5 and 10H reactive oligomeric α‐syn variants are selectively present in human PD brain tissue, but not to the plaques or tangles in AD or cognitively normal tissue,40 and therefore both these protein variants are attractive therapeutic targets. It is important to point out that while both D5 and 10H scFvs recognize different naturally occurring oligomeric α‐syn species in human PD brain tissue, they do not interact with monomeric or fibrillar forms.10, 29 Supporting the contention of the importance of targeting oligomerized α‐syn, when we tested both in vitro and in vivo (via intracerebral injection) an scFv (D10) that recognizes primarily monomeric α‐syn showed no significant effects at reducing α‐syn accumulation. Therefore, these results suggest that the beneficial effects observed in the mice treated with D5 or 10H are related to recognizing specific oligomeric α‐syn aggregates. In fact, D5 reacts primarily with in vitro generated dimers and tetramers while the 10H reacts with similarly generated trimers and hexamers.10, 29 This may explain why the Pen‐D5 is more effective than the Pen‐10H in treating the early age group mice while both are effective at treatment in late age group (Table 1).

Cell penetrating peptides (CPP) like penetratin have been used successfully to deliver peptides, proteins, and even siRNA to cells.32, 67, 68 In fact, several instances of CPP linked to the delivery of cargo across the blood–brain barrier have been reported. Penetratin is a 16 amino acid sequence derived from the homeodomain of the Drosophila Antennapedia homeodomain.51, 52 Recently, it was shown that when inserted into the middle of the scFV between the two variable domains, it could facilitate transport of the antibody across the blood‐brain barrier.32 We fused the penetratin sequence to the N‐terminus of the scFV and observed significantly greater uptake of the antibody across the blood–brain barrier. In addition, we added a secretory signal sequence from CD533 to allow for secretion from the lentivector‐infected cells of the liver and spleen53 following i.p. injection. More recently, a modified form of penetratin, labeled PenetraMax, was able to facilitate transport across the small intestine into the blood stream.69 Thus, with additional modifications, it may be possible to deliver the scFv orally for delivery across the intestinal epithelium, through the blood stream and finally across the blood–brain barrier for action in the CNS.

The scFvs were administered to 3–4‐month‐old mice to study their value at preventing α‐syn deposition during which D5 provided a much stronger benefit than 10H. Mice in the early age group treated with D5 showed a sharp decrease in PK‐resistant α‐syn inclusions in the substantia nigra, whereas the 10H treated mice showed no benefit. Early stage mice treated with both D5 and 10H showed significant reductions in α‐syn in the neuropil in the fronto‐parietal cortex and hippocampus and slight reductions in axonal dystrophy in the basal ganglia. The D5 but not 10H‐treated mice in the early age group showed a sharp decrease in D5‐reactive α‐syn aggregates while both treatments showed decreases in 10H‐reactive α‐syn levels where the D5‐treated mice had sharper decreases in both D5‐ and 10H‐reactive oligomeric α‐syn aggregates in the hippocampus. The younger mice treated with D5 also showed a significant gain in NeuN staining in the hippocampus. Therefore, while both D5 and 10H provide substantial benefit as a therapeutic to mice that already have significant α‐syn pathology present, D5 provides a much more significant benefit compared to 10H as a prophylactic in preventing neuronal damage in mice in the early stages of PD. As expected, total levels of α‐syn increased slightly in the control treatment group between the early and late stage groups, as did axonal dystrophy, number of synuclein‐positive cells, levels of D5‐ and 10H‐reactive oligomeric α‐syn in total brain homogenates as well as the hippocampus and cortex.

The results obtained here provide evidence that different therapeutic strategies may be most beneficial during different stages of PD. α‐syn is a complex molecule that can occur in a variety of different morphological aggregate and posttranslationally modified states.70, 71 The role of these different protein variants in PD is still unclear. Here, we show that selectively targeting different α‐syn variants provides differential therapeutic effects. Where the D5 scFv provides substantial therapeutic benefit when administered either early or late during disease progression; the 10H scFv provides benefit primarily when administered during later stages of α‐syn deposition. Both D5‐ and 10H‐reactive oligomeric α‐syn variants can be detected in serum providing readily available biomarkers that can assist in following the progression of PD and determining what are the most appropriate therapeutic strategies. We show here the D5‐reactive α‐syn variants are present in serum in the early stage mice but not the later stage mice, whereas the 10H‐reactive variants have high variability in the early stage mice, with more consistent if lower levels in the later stage mice. Previously, the D5 and 10H antibodies have been used to classify patient sera as presumptive PD with D5 the more effective diagnostic antibody.40 In fact, conformational specific α‐syn antibodies developed through immunization of mice with in vitro developed fibrils showed differential immunohistochemistry of brains of patients with PD, MSA, and DLB.72 Thus, morphological‐specific α‐syn antibodies may be a more specific method for detecting stages of PD and/or differentiating synucleinopathies such as DLB, MSA or PD.

In summary, we have developed two antibodies fused to a cell‐penetrating peptide, penetratin, that are effective at reducing the accumulation of α‐syn in the CNS of α‐syn tg mice. The two antibodies were developed against different conformations of α‐syn and when delivered to α‐syn tg mice as either a prophylactic or as a treatment during disease development, they showed different effects. Thus, the development of conformational specific antibodies for α‐syn or other aggregating proteins may be warranted, and more than one antibody may prove to be useful depending on the application during disease development. To that end, the use of these antibodies for diagnostics may also prove to be beneficial to diagnose patients along the disease development pathway allowing for personalized medical intervention.

Author's Contributions

B. S., S. W., E. M., and M. S. conceived of the study and participated in its design. B. S., S. W., E. R., E. V., W. Z., M. M., J. F., and A. A. performed the experiments. E. M. completed the statistical analysis. B. S., S. W., E. M., and M. S. drafted the manuscript. All authors read and approved the final manuscript.

Conflict of Interest

Arizona State University has a patent pending on behalf of MS for the D5, D10, and 10H scFV antibodies.

Figure S1. Immunocytochemical analysis of levels of CNS accumulation of D5 scFv antibody (V5) following systemic delivery of a lentivirus vector expressing either PenSecD5, D5‐apoB or D5 alone. The non‐tg received a single intraperitoneal injection of LV‐control, LV‐D5, LV‐Pen‐D5 or LV‐D5‐apoB. Four weeks after injection, mice were sacrificed and serial sections were prepared. (A) Sections were immunostained with antibodies against V5 to identify the scFv. Upper panels are an overview of the brain sections at low magnification (20X), the lower panels are higher magnification (400X) areas (dotted squares) of the neocortex and hippocampus. Arrows indicate neuronal immunostaining. (B) Image analysis of levels of scFv immunoreactivity (V5 immunoreactivity) in the neocortex expressed as corrected optical density. (C) Image analysis of levels of scFv immunoreactivity (V5 immunoreactivity) in the hippocampus expressed as corrected optical density. *Indicates statistical significance P < 0.05 compared to LV‐D5 injected mice. One‐way ANOVA with post hoc Tukey–Kramer. Scale bar represents 200 μm in low power images and 40 μm in high power images.

Click here for additional data file.