Therapeutic significance of exosomes as a drug delivery system Application: Parkinson’s disease
1
stexaminer: dr. Maarten H. K. Linskens 2
ndexaminer: prof. dr. Arjan Kortholt
Student: Ilya Ivanov
Study program: MSc Molecular Biology and Biotechnology Student number: S2961806
Date of submission: 12.11.18
Index
Abstract………..…..3
1. Introduction to exosomes………3
1.1 Exosome biogenesis and release………..…………....3
1.2 Composition and loading of exosomes………...………...…...4
1.3 Exosomes and their biological functions and applications………..5
1.3.1 Cell-to-cell communication……….………..5
1.3.2 Immunological response………5
1.4 Cellular recognition and internalization of exosomes……….6
2. Exosomes for drug delivery………6
2.1 Advantages and limitations of exosomes over other drug delivery systems………6
2.2 Exogenous drug loading methods………6
2.3 The significance of exosome surface molecules and their modifications………8
3. Blood-brain barrier (BBB)………...9
3.1 Challenges of crossing the BBB………..9
3.2 Manipulating exosomes for a better BBB permeability………...10
4. Parkinson’s disease (PD) and therapeutic approaches for its treatment………....11
4.1 Catalase-loaded exosomes……….11
4.2 Delivery of dopamine to the brain………15
4.3 Downregulation of α-synuclein expression with RNA interference………17
Conclusion………20
References………20
Abstract
Exosomes are nanovesicles secreted by all cell types and can be found in all body fluids.
They carry a cell-specific composition of membrane phospholipids and biomolecules.
Exosome biogenesis and composition is mediated by a number precise mechanisms. In this review exosomes will be addressed as drug delivery vesicles for treatment of Parkinson’s disease. PD targeting requires the crossing of the blood-brain barrier and exosomes are able to do so. Various therapeutic cargo molecules can be loaded into exosomes ex vivo, such as curcumin, dopamine or mRNA. Exosomes appear to be a prominent drug delivery system due to their biocompatability, versatility, and the ability to preserve cargo catalytic activity and to cross the BBB. This last one is essential for all CNS diseases, including PD.
Different therapeutic strategies for PD treatment and therapeutic cargo incorporation in exosomes for drug delivery to the brain are the main topics addressed in this review.
1. Introduction to exosomes
Extracellular vesicles (EVs) are subcellular structures secreted by all cells; they can be found in all body fluids. There are three types of EVs: exosomes (30-100 nm) [5],
microvesicles (100-500 nm) and apoptotic vesicles (500-1000 nm) [1]. This review will focus on exosomes as potential drug delivery vehicles. Exosomes were initially considered cells’ mechanism of waste disposal but it was recently noticed that exosomes carry a mix of biomolecules that reflect intracellular molecular composition. That makes exosomes the key to understanding the intracellular environment and the state of a cell.
1.1 Exosome biogenesis and release
The biogenesis of exosomes begins with endocytosis. Early endosomes combine and become bigger, transitioning into a late endosome. The inward budding of the late endosomal membrane leads to the formation of intraluminal vesicles (ILVs). Endosomes containing ILVs are called multivesicular bodies (MVBs). As can be seen from Fig. 1 and the mechanism of ILV formation, the ILVs have lipid bilayer that is topologically identical to the cell plasma membrane. Inside of ILVs closely resembles cytosolic environment. The process of ILV formation is mediated by endosomal sorting complex required for transport
(ESCRT). This complex consists of 4 smaller protein complexes: ESCRT0, ESCRTI, ESCRTII and ESCRTIII. This machinery is involved in the invagination of late endosomal membrane with the aid of curvature-inducing factors [2]. However, a second, ESCRT-independent, mechanism was found to be involved in MVB formation [27]. Endosomal membrane contains lipid rafts (more details in 1.3), that are regions of membrane enriched in
cholesterol, ganglioside GM3, and sphingomyelin. Neutral sphingomylinase2 (nSMase2) is able to convert sphingolipids to ceramide. Ceramide induces coalescence and domain- induced budding of late endosomal membrane thus promoting ILV formation [28].
Figure 1. Biogenesis of MVBs and release of exosomes and other EVs [2]
After ILV formation is finished MVBs are carried by motor proteins to the periphery of a cell, where they fuse with plasma membrane (Fig. 1). The fusion process is mediated by SNAREs and tethering factors. After being released into the extracellular space ILVs become exosomes [23].
1.2 Composition and loading of exosomes
Exosomes are composed of many different kinds of enzymes, protein complexes, nucleic acids.
The main phospholipids found in the exosomal membrane include phosphatidylcholines, phosphatidylamines, and phosphatidylserines. Additionally, cholesterol, ganglioside GM3, and sphingomyelins have been found in exosomal membranes of most cell types. This lipid composition resembles lipid rafts, membrane regions present in cell membrane [30]. Lipid rafts, previously mentioned in 1.1, are a feature of cell membrane and, therefore, are also present in exosomal membrane. However, exosomal membrane seems to have higher amount of protein than the cell membrane. So, exosomal lipid rafts are only resembing the lipid rafts of cell membrane [5].
There is a list of proteins that are always present in exosomes as they are essential for the maintanance of exosomal microenvironment. HSPs, or heat-shock proteins, are found in all types of exosomes and are important for when exosome experiences stress conditions.
Other common types of proteins present in exosomes include tetraspanins (CD9, CD63 etc.), histones, integrins, small GTPases, annexins and actins [6].
It’s not fully understood yet how proteins are sorted into ILVs. There are several methods for protein sorting, of which the most important one is the ubiquitin-dependent sorting by ESCRT machinery [23]. However, not all exosomal proteins were found to be ubiquitinated, which suggests another mechanism of protein uptake into exosomes [5]. Plasma
membrane anchor tags, such as myristoylation, prenylation and palmitoylation can target proteins towards the site of membrane budding for ILV formation. Therefore, not only ubiquitination but also posttranslational lipid modifications can serve as a sorting signal for protein loading into exosomes [2].
Exosomes are known to contain nucleic acids such as mRNA and miRNA. This fact places them as a potential RNA delivery vehicles. Essential differences between total RNA profile of parental cells and of exosomes were noticed. This means that some mRNA may be enriched in exosomes but not in the parental cells. A conclusion is that exosomal nucleic acid composition is also regulated by a sorting mechanism [6]. This sorting mechanism, in contrast to sorting of proteins, is ESCRT-independent, and depends on ceramide. We have seen already that nSMase2 is able to convert sphingolipids present in lipid raft-like regions into ceramide. Some RNAs can be delivered to raft-like regions by RNA-binding proteins (RBP). A2B1 binds to miRNA and has an affinity for ceramide. Therefore, RNA loading is dependent on the interaction of RNA or RNA-RBP complex with lipid raft-like region of MVB membrane. In this way, RNA is present at the site of ILV formation [23].
1.3 Exosomes and their biological functions and applications
1.3.1 Cell-to-cell communication
The hypothesis that exosomes propagate intercellular communication is based on cell- specific content of exosomes and its targeted delivery to recipient cells. Moreover, studies have shown that the delivery of functional biomolecules to cells results in a phenotypical response [7]. It has been shown that tumor-derived exosomes can promote angiogenesis [29]. The same exosomes can also target epithelial cells and result in aberrant mitosis and increased cell proliferation leading to the progression of tumor [8]. Exosomes in tumor microenvironment (TME) can have an adverse effect on CD8+ T-cells. Exosomes of developed tumor cells are able to carry immunosuppressive molecules and make tumor cells escape immune system surveillance. This results in inhibition of T-cell proliferation, suppressed activity of CD8+ T-cells and inhibition of killer cells anti-tumor function [8].
These findings indicate that tumor cells can work towards their common goal (such as tumor proliferation and escaping immune response) by secretion of exosomes expressing specific biomolecules.
1.3.2 Immunological response
Exosomes initially were believed to be used for “waste disposal”. In fact, exosomes can have a very important role in immune regulation. For example, B-cell (and later on dendritic cell (DC)) derived exosomes are known to express MHC class II on their surface [6]. B-cells are able to uptake antigens, which then bind to MHC class II. B-cells then secrete exosomes that express antigen loaded MHC class II on their surface. Antigen is delivered to CD4+ T-cells and induces antigen-specific immune response [5].
1.4 Cellular recognition and internalization of exosomes
Exosome targeting is cell-specific, therefore cells should have ways to recognize and uptake exosomes. Recognition of exosomes can happen in three ways, as proposed by McKelvey et al. Recognition by free-floating leads to the opsonization, i.e. the defensive action of
immune system. Opsonization means that exosomes are internalized by macrophages.
Exosomes express large array of cytokines that may attract nearby T-cells and other immune cells. This results in internalization by phagocytosis [13].
Recognition by adhesion is governed by integrins that undergo a transition from low to high affinity state. Integrins are then able to oligomerize causing lymphocytes, that have affinity for oligomerized integrins, to bind to the integrin-bound exosomes. This approach could be partially involved in T-cell recognition of exosomes [13].
A good example of antigen-based recognition would be activation of T-cells. DC-derived or B lymphocyte-derived exosomes have MHC class II molecule bound to specific antigen.
Such exosomes are able to activate antigen-specific T-cells and evoke an immune response [13]. The fact that all three examples of exosome recognition include interactions with immune cells further supports the notion that exosomes play very important
immunological role.
Internalization of exosomes by recipient cells happens either by direct fusion of membrane and exosome contents release into the cytosol or by endocytosis. Endocytosis can be
mediated by clathrin, dynamin or caveolin. Exosomes can also be internalized by phagocytosis or macropinocytosis [24].
2. Exosomes as drug delivery vehicles
2.1 Advantages and limitations of exosomes over other drug delivery systems
Several properties make EVs such great drug delivery system. While being protected by lipid bilayer, their cargo is stable in the cytosolyc environment. Exosomes prove to be especially good for delivering small hydrophilic molecules, like mRNAs or dopamine [6].
Their structure allows them to bypass many biological barriers and even the toughest one to pass, the blood-brain barrier. By expressing ligands on their surface exosomes can be utilized for targeted delivery by making use of ligand-receptor specific interactions.
However, exosome-based drug delivery has its limitations. First, not much is known about the mechanism of BBB crossing. This limits knowledge-based discovery of modifications for increased BBB permeability. Second, there is lack of understanding of potential side- effects caused by therapeutic cargo [6]. In addition, current EV isolation and purification techniques are still far away from the quality required for conduction of human clinical trials [6]. Also, cargo loading is much more efficient with small molecules and nucleic acids which significally narrows the scope of pharmaceutical application of exosomes.
2.2 Exogenous drug loading methods
If exosomes are utilized for drug delivery then they should contain some therapeutic cargo incorporated inside them. Clearly, the chances of such cargo being loaded by the cell’s endogenous sorting machinery (e.g. ESCRT) are very low. This calls for the development of methods for the loading of an external cargo. Some of these methods are presented in this section. They can be divided into two main groups: loading in vivo and in vitro. There are
three general approaches for exosomal drug loading described by Elena V. Batrakova and Myung Soo Kim (Fig. 2).
The first approach is based on isolation and purification of native exosomes with subsequent in vitro loading of the drug molecules. There are actually many possible strategies for cargo loading in vitro. The strategy should be chosen based on type of cargo molecules. Molecules with low molecular weight like curcumin or dopamine can be loaded into exosomes by incubation at room temperature or by freeze/thaw cycles. The loading capacity varies from 7.2% to 11.7%. The mild loading capacity can be explained by a large array of molecules already being present inside exosomes [1]. Higher loading capacity can be achieved by increasing the cargo concentration to maximize the concentration gradient and enhance the diffusion across the lipid bilayer [21]. Loading efficiency can be increased twice by the addition of saponin when loaded by incubation at RT [16].
Large and hydrophobic molecules can be more efficiently loaded into exosomes by sonication or extrusion. These methods result in the formation of temporary pores and possible lipid bilayer reformation. The formation of such pores allows large molecules to enter exosomes [16].
Exogenous RNA can be loaded into exosomes by electroporation [Alvarez et al]. Higher temperature and precomplexation of RNA with cationic liposomes can improve RNA loading [1].
All in vitro loading strategies are different from how exosomes are naturally loaded. They can therefore have an indirect impact on the final structure of exosomes. It is not clear why but different different strategies of cargo loading into exosomes can have an effect on the rate of cargo release from exosomes (Fig. 4C) [16].
Another way to load cargo is to make an organism produce the cargo protein and load it in vivo (Fig. 2C). Ndfip1 is an enzyme that binds to WW domain and performs ubiquitination.
By making use of this reaction Sterzenbach et al. managed to load Cre recombinase in vivo.
Coexpression of WW-Cre and Ndfip1 in host cells resulted in a successful ubiquitination of Cre recombinase and its loading into exosomes. WW tag doesn’t interfere with Cre
localization or activity [12]. By using this Ndfip1 reactivity it could be possible to load other proteins into exosomes in a similar way.
Figure 2. Cargo loading approaches. (A) Through external loading of cargo into puriofied exosomes. (B)
Through loading of drug into parental cells and subsequent incorporation. (C) Through engineering of recombinant cells that produce the drug and load it into exosomes. [1]
2.3 The significance of exosome surface molecules and their modifications
Exosomes possess an intrinsic ability to cross various biological barriers and thus can deliver their contents to most cells in an organism. Exosomes contain many different molecules on their surface. Exosome surface can be modified for better recipient cell targeting, stability or as a fluorescent biomarker to measure its distribution. The role of exosome surface composition and its modifications are explained in this section.
Exosomal surface molecules can be used for therapeutic purposes. Application of exosome native surface glycosphingolipid glycan for treatment of Alzheimer’s disease was suggested by Yuyama et al. In their study they took advantage of the affinity of amyloid-β for
glycosphingolipid glycan groups expressed on the surface of neuroblastoma exosomes.
Amyloid-β build up in the brain drives the pathogenesis of Alzheimer’s disease. Amyloid-β bound NB exosomes delivered amyloid-β to microglial immune cells in the brain, which degraded amyloid-β together with exosomes. Taking advantage of exosomal surface phospholipids can develop into a strategy for targeting AD symptoms in the future [32].
Targeting to certain regions in a body can be achieved by expressing certain peptides on the surface of exosomes. Alvarez-Erviti et al. was the first group to come up with a strategy of exosome modification to target the brain. It was achieved by host cells expressing Lamp2b protein fused to rabies virus glycoprotein (RVG). Lamp2b is an exosomal
membrane protein that usually gets incorporated into exosomal membrane. RVG peptide causes homing of exosomes to the brain. They obtained exosomes that express Lamp2b on their surface fused to RVG and achieved better targeting to the brain [23]. A group of researchers in 2014 noticed that peptides attached to N-terminus of LAMP2b are
susceptible to proteolytic degradation in acidic conditions. They realized that glycosylation motif GNSTM can protect peptides from such degradation. This potentially leads to more RVG expressed on exosomal surface and a better drug delivery [25].
3. Blood-brain barrier (BBB)
BBB governs a highly restricted passage of biomolecules from bloodstream to the brain.
Structure of BBB can be seen in Fig. 3. Big part of BBB is basal lamina, which is located on the brain side and separates endothelial cells from astrocytes. It is comprised of laminin, heparin sulfate, and collagen. Astrocytes provide support for endothelial cells and also contribute to BBB overall structure. Inbetween the endothelial cells is the space called tight junctions. These tight junctions are formed by claudin, occludin and ZO-1 [15].
Figure 3. Structure of the blood-brain barrier.
Acquired from: https://www.emf.ethz.ch/en/knowledge/topics/health/blood-brain-barrier/
3.1 Challenges of crossing the BBB
Endothelial cell lining is the main barrier that prevents the flow of pathogen molecules to the brain. This barrier is maintained due to the presence of tight junctions between
endothelial cells. Their function is to keep endothelial cell lining intact and to prevent free passage of molecules to the brain. Tight junctions are also responsible for exceptionally high electrical resistance across BBB and prevent entry of toxic substances and pathogens.
BBB not only restricts passage to the brain but also takes a role in excretion by expressing ABC transporters on the endothelial cells. These transporters provide a second line of defense against unwanted molecules passing to the brain and remove molecules that
BBB is so preventive because brain is a very preserved microenvironment and many molecules can have an adverse effect on brain function. The time spent in the blood flow before crossing the BBB is called circulation time and is often maximized when it comes to therapeutic applications. The reason is that there are macrophages present in blood that internalize nanoparticles (a process called opsonization) that appear to be pathogenic, meaning they express specific antigen peptides that trigger an immune response. The bigger the circulation time the longer the nanoparticles can reside in the blood without undergoing opsonization. This is likely the main disadvantage of using liposomes or exosome-like engineered nanoparticles – they are cytotoxic and undergo rapid drug clearance by mononuclear phagocyte system (MPS) [16]. That’s why the research now is focused on the application of naturally produced exosomes in order to avoid opsonization and increase biocompatability.
3.2 Manipulating exosomes for a better BBB permeability
It’s possible for exosomes to cross BBB and it was proven by detecting tumor-derived exosomes and microvesicles of glioma-bearing mice circulating in blood. This finding suggests that exosomes have an intrinsic ability for crossing the BBB [16]. However, it’s still unclear what the precise mechanism of BBB crossing is. It was proposed that exosomes could be internalized into MVBs of the first layer of endothelial cells to then be released and internalized by the second layer of cells and so on until reaching the brain [31].
One of the possibilities to locally increase BBB permeability is to disrupt the tight junctions by sonication. The problem is that this approach raises the risk of unwanted substances entering the brain. A much more friendly approach is by taking advantage of transcytotic pathway of nanoparticles binding to endothelial cells [15].
Transcytosis can be of two types: adsorptive and receptor-mediated. In case of adsorptive transcytosis exosomes would come in contact and bind to the plasma membrane of
endothelial cells. Nanoparticles with positively charged surface are more likely to adsorb rather than nanoparticles with negatively charged or neutral surface [15]. Receptor- mediated approach is based on ligand-receptor specific interactions. Endocytosis that follows can be mediated by clathrin or caveolin. With clathrin-mediated endocytosis 200 nm endosomes are budding off in clathrin enriched membrane regions. pH of these endosomes lowers as they progress to the late stage and the contents are subjected to lysosomal degradation [15]. Caveolin-mediated endocytosis occurs in lipid rafts and results in vesicles of much smaller diameter (80 nm). Caveolin-coated vesicles avoid degradation and their fate depends on the cell type [15].
Better BBB permeability can be achieved by optimization of several properties of drug carrying vesicles. First, nanoparticles smaller than 200 nm are more likely to cross BBB [15]. As it was mentioned previously, positively charged surface molecules increase the chance of adsorptive transcytosis. Nanoparticles can be coated with PEG corona, which leads to increased circulation time and decreased uptake by immune system [15].
However, it was reported that PEGylated liposomes lost their long-circuation property on the second week of systemic administration to mice. This creates a problem associated with possible long-term treatment and increaseed opsonization of PEGylated nanocarrier vesicles [11].
Surface modifications to enhance receptor-mediated transcytosis have been used,
including ligands for Lf, LRP1 and Tf receptors [9]. An important concept to note is avidity.
It’s recommended to avoid introducing too many different ligands on the exosomal surface as in that case they will all hinder on another. Good endocytosis can be achieved by a balanced expression of necessary ligands.
4. Parkinson’s disease (PD) and therapeutic approaches for its treatment
PD is the second most common disease in neurodegenerative group after Alzheimer’s disease. Neurodegeneration is the process of accumulated damage received by neurons in CNS. It’s symptoms include difficulty while walking, shaking as well as sleeping, thinking, and behavioral problems. Dementia is usually one of the symptoms of PD in an advanced stage [17]. The cost of brain diseases in 2014 in the US reached USD$789 billion. As neurodegenerative diseases are known to progress and worsen with age and average life expectancy in the developed countries is only increasing it is expected for these costs to rise even higher [15].
At this point there is no cure for PD. It’s still unclear what the cause of PD is but it is
considered to involve both genetic and environmental factors. Neurons in substantia nigra are dying and this causes motor disorders (thus the name – neurodegeneration).
Neurodegeneration leads to decreased dopamine levels. Therefore, dopamine supply appears to be an attractive strategy to delay the progression of PD. A lot of reactive oxygen species (ROS) were found in the brain lesion associated with PD pathology. ROS are
dangerous as they are damaging the tissue around them and can lead to loss of neural cells.
Normally the elimination of ROS is done by antioxidants or reductases. Supplying a strong antioxidant can result in reduced ROS level and delayed PD progression. One of the key features of PD is an increased accumulation of α-synuclein in Lewy bodies. An increased production of this protein together with shape altering S129 phosphorylation cause the formation of accumulated α-synuclein aggregates in Lewy bodies. This process is closely associated with PD at the moment [18]. Gene therapy or RNA interference strategies seem appealing as a way to reduce the amount of α-synuclein.
4.1 Catalase-loaded exosomes
Catalase is an enzyme that catalyzes decomposition of hydrogen peroxide to oxygen and water. Catalase’s main function is the reduction of oxidative stress and elimination of ROS.
Catalase is one of the strongest antioxidants and can be used for reduction of ROS in PD patients.
Injecting catalase intravenously is possible but it’s a large hydrophobic protein so it will be challenging for bulk catalase to cross BBB. Therefore, catalase will probably have long circulation time meaning that it will most likely undergo opsonization. Incorporating catalase into exosomes will preserve its catalytic activity and prevent degradation. Haney et al. showed in their research that catalase-loaded nanoparticles were able to reach various cell types in vitro [19]. In their esearch they mixed catalase with PEI-PEG block copolymer. This resulted in nanoparticles with catalase-polymer complex core and PEG corona, termed nanozymes. Nanozymes were loaded into macrophage cells by incubation for 1 h. [19].
They investigated the ability of macrophage-derived nanozymes to reach three types of
Nanozyme transfer from macrophage cells to BMVEC, bovine brain microvessel endothelial cells (BBMEC), and neurons was modelled with and without (2 mm distance) the physical contact of the two cell lines. To test whether nanozyme transfer was contact dependent for BBMEC and neuron cells adhesion inhibitors were added, namely trypsin and locostatin.
Nanozyme transfer to these cell types was not affected by adhesion inhibitors.
Nevertheless, close contact seems to improve nanozyme transfer from macrophage cells to BMVEC [19]. Nanoparticle accumulation in recipient cells was measured by flow cytometry.
Haney et al. managed to show effective transfer of catalase nanoparticles from macrophage cells to BMVEC, neurons, or astrocytes. They showed that cellular contact favors the
transport but adhesion is not necessary. The main limitation of their work is the absence of in vivo trials. Nanoparticles are known to be susceptible to opsonization, hence it’s not clear whether these nanoformulations will reach brain cells in mouse PD model.
Figure 4. Levels of RITC-labelled catalase nanozyme accumulation in (A) BMVEC, (B) Neurons, (C) Astrocytes
measured by flow cytometry [19].
4 years later the same group of researchers did a similar experiment but with several important changes. They were also aiming at the delivery of catalase to the brain but this time decided to take a different approach. First of all, instead of PEG-coated nanoparticles they used exosomes. Although in their 2011 work they showed low cytotoxicity of
nanozymes drug delivery using exosomes would probably prove to be more biocompatible.
Macrophage-derived exosomes were loaded with catalase in vitro via incubation at RT (with and without saponin), freeze/thaw cycles, sonication, and extrusion [16]. Catalase- loaded exosomes were called exoCATs. ExoCATs were tested on their loading efficiency, catalase release rate and antioxidant activity.
The results of loading efficiency are shown in Fig. 5A. Loading with sonication and extrusion resulted in the highest loading efficiency, possibly because catalase is a large protein (≈240kDa) and simple permeation will not result in a good catalase loading.
Extensive reformation of exosomes membrane under sonication and extrusion methods allowed catalase to bypass the lipid bylayer [16].
Fig. 5B shows enzymatic activity of exoCATs that was done to evaluate the ability of
exosomes to preserve catalase activity. Catalase, loaded by sonication and extrusion, shows the highest enzymatic activity, followed by catalase, loaded with freeze/thaw cycles and incubation at RT. These findings further prove the data from Fig. 5A.
Catalase release was measured with dialysis membranes with 2000 kDa cutoff. Catalase loaded at RT without saponin showed the fastest release while exosomes loaded by sonication only released 40% catalase in 24h (Fig. 5C). Thus, sonication of exosomes resulted in high loading efficiency and sustained, prolonged catalase release [16].
Figure 5. Results of loading efficiency (A), catalytic activity of catalase (B) and rate of catalse release from
exosomes (C) [16]
The researchers in this study did not do any modifications to the exosome surface, they relied on the intrinsic ability of exosomes to cross the BBB. Hence, it was not entirely clear which interactions took place in exosome recognition. Rate of exosomal uptake by PC12 recipient cells was measured by measuring fluorescence. This experiment revealed that exosomes loaded by sonication showed the highest uptake rate. Freeze/thaw loaded exosomes and exosomes loaded by incubation at RT both didnt result in high uptake by PC12 cells (Fig. 6). It is possible that due to membrane reformation of exosomes loaded by sonication they were accepted at a higher rate than the exosomes loaded by other methods.
Membrane reorganization could reveal certain hydrophobic regions or integral proteins that resulted in improved interactions with recipient cell membrane [16].
Figure 6. Uptake of exosomes by PC12 recipient cells. Exosomes were stained with DIL dye and fluorescence
was measured by Shimadzu RF5000 spectrophotometer. The amount of catalase was normalized for the total protein content [16]
ExoCAT neuroprotective ability was measured by MTT assay and the results are presented in Fig. 6. The extent of neuron protection is shown as the percentage of survived neurons relative to the initial number of neurons. Neuroprotection is the highest for exoCATs loaded by sonication (Fig. 6). An addition of saponin during incubation at RT significantly improves neuroprotection by exoCATs. All other loading methods did not result in exoCATs with high neuroprotective activity, including catalase alone [16].
Figure 6. Neuroprotection evaluated on cells pre-incubated with 6-OHDA (C) [16].
Overall, exosomes loaded by sonication, extrusion, and incubation at RT with saponin showed high loading efficiency, preservation of catalase enzymatic activity as well as prolonged, sustained release. It was hypothesized that saponin may selectively remove some of integral membrane molecules, like cholesterol, and thus create pores for successful catalase diffusion [16]. ExoCATs loaded by permeation with saponin and by sonication provided high loading and sustained release. These two types of exoCATs also showed decreased brain inflammation and improved neuronal survival in vivo (Fig. 6). Potentially, exosomes loaded by permeation with saponin are superior to exosomes loaded by
sonication as they could have a more uniform and intact surface morphology. Such morphology can decrease their recognition by immune system and increase their blood circulation time for future clinical trials [16].
4.2 Delivery of dopamine to the brain
Neurodegeneration in substantia nigra is followed by dopamin deficiency in PD patients.
Dopamine supply serves as an effective strategy to increase dopamine level in the brain and increase neuronal survival. Delivery of bulk dopamine to the brain seems hardly possible due to the rapid clearance by immune system. It was confirmed in 2015 by Pahuja et al. that dopamine-loaded nanoparticles are less cytotoxic than bulk dopamine.
Administration of dopamine-loaded nanoparticles significantly reduces dopamine
opsonization and ensures dopamine delivery to SH-SY5Y cells induced with 6-OHDA [20].
Although incorporation of dopamine into nanoparticles reduced dopamine opsonization, NPs are still usually recognized by macrophages and risk to be internalized. Tumor-derived vesicles originating in mice brain were found in blood circulation. This indicates that
exosomes are able to penetrate the BBB without any additional enhancements [16].
The research group of Qu et al. used this ability of exosomes and isolated bone marrow derived macrophage exosomes. These exosomes were loaded with dopamine by incubation at RT for 24h. They used saturated dopamine solution and achieved loading of 16%. Free dopamine was removed by ultracentrifugation. Dopamine is a relatively small and
hydrophilic molecule, therefore incubation at RT seems like a good strategy to load dopamine into exosomes [21].
Brain targeting by exosomes and their distribution in the brain was modelled on bEnd.3 and SH-SY5Y cells. In bEnd.3 model of BBB the concentration of dopamine delivered by blood exosomes was 1.02±0.15mmol/g, while the concentration of free dopamine was non- detectible. Free dopamine accumulated in kidney, liver, and lungs whereas dopamine from blood exosomes achieved 15-fold increase in dsitribution compared to free dopamine 6 hours after the start. Free dopamine levels are much lower also because of opsonization [21].
An experiment with SH-SY5Y cells model showed that exosomal uptake happened through energy-dependent endocytosis, which involved clathrin- and caveolin-dependent processes and macropinocytosis [21]. Exosomes appeared to facilitate dopamine transport across BBB. This transport could be inhibited by an addition of NaN3 or at low temperature (Fig.
7).
Figure 7. Accumulated clearance volume obtained in BBB in vitro model with time [21]
One of the goals of the research was to investigate the interactions underlying exosome recognition by target cells. Transferrin is an abundant protein present in blood that binds to its receptor, TfR. There is TfR present on the surface of blood-derived exosomes as well as on the surface of endothelial cells. Qu et al. explored the interaction between bEnd.3 cells and blood-derived exosomes and elucidated it was dependent on transferrin-TfR binding. With the help of ZDOCK rigid-body protein docking program they discovered that the binding occurs between transferrin dimer and two TfRs forming a tetramer (Fig. 8) [21].
Figure 8. Transferrin dimer bound to two TfR proteins predicted by ZDOCK doicking program. Transferrin
shown in surface (left) and in ribbon (right) structure [21]
For the assessment of therapeutic efficacy in vivo clinical mice were divided into 5 groups.
The groups are: sham control (healthy mice, treated with 0.9% saline), mice treated with empty blood exosomes, free levodopa, free dopamine, and dopamine-loaded exosomes.
Activity was measured as dopamine level relative to healthy mice (sham control). From fig.
9 we can see that DA exosomes treatment was the most successful in elevating dopamine level, resulting in over 55% dopamine compared to sham control [21].
Figure 9. Relative dopamine level in mice compared to sham control [21]
DA-loaded exosomes were also the best treatment to elevate the amount of such important enzymes as superoxide dismutase and glutathionine peroxidase, important antioxidant enzymes. This suggests that dopamine-loaded blood exosomes “can help improve dopaminergic neurons and ameliorate disease phenotype in a mouse model of PD” [21].
This study conducted by Qu et al. presents increased dopamine level in mouse model of PD achieved by systemic injection of dopamine loaded blood exosomes. It’s possible that dopamine could be the cause for dopaminergic neurogenesis, meaning that the neural cells started producing endogenous dopamine after this treatment [21]. Such treatment has other indirect positive effects like an increase in antioxidant enzymes resulting in lowered oxidative stress. In the end, blood exosomes were able to target bEnd.3 cells due to
transferrin-TfR interactions and achieved decreased oxidative stress and increased number of neurons and dopamine.
4.3 Downregulation of α-synuclein expression with RNA interference
It is known that PD is characterized by accumulation of mutated α-synuclein aggregates in Lewy bodies. The first externally loaded cargos for drug delivery by NPs were RNA
molecules. They are small molecules and have a very broad target scope. Delivery of RNA molecules in liposomes dates back to 1978 [22].
In their research Cooper et al. attempted to reduce the expression of α-synuclein in the brain by systemic injections of siRNA. An advantage of this strategy is that it is versatile.
Any protein in theory can be the target for RNAi, therefore RNAi has wide range of applications (cancer, cardiovascular diseases etc.).
α-synuclein siRNA was loaded into DC-derived exosomes post-isolation. The most effective
strategy previously mentioned in 2.3. The effect of siRNA delivery was tested on healthy mice and transgenic mice with human α-synuclein mimic.
In the healthy mice model the researchers assessed siRNA ability to reduce endogenous α- synuclein in mouse brain. The brain was analyzed for the amount of α-synuclein and α- synuclein mRNA 3 and 7 days after injection (Fig. 10). mRNA levels were analyzed by qPCR relative to GAPDH and protein levels were analyzed by western blot relative to actin [18].
Both were measured in midbrain, striatum, and cortex. mRNA and α-synuclein levels are decreased in all three brain regions by approximately 50%. The exception is α-synuclein level in cortex, where it’s 75% of the control (Fig. 10). The explanation for that could be slower turnover of cortical α-synuclein [18]. 7 days after the injection α-synuclein and mRNA levels didn’t change much meaning most of RNAi activity happened in the first 3 days post-injection (Fig. 10). These findings suggest that the major therapeutic effect is achieved within 3 days and further waiting for higher efficiency is not very practical. On the one hand, the major therapeutic effect is achieved within 3 days. On the other hand, there is not much room for the improvement of therapeutic efficacy.
Figure 10. Amount of α-synuclein and α-synuclein mRNA shown as percentage of control 3 (A, B) and 7 (C, D)
days after injection in midbrain, striatum and cortex [18]
In the experiment with healthy mice they used mouse α-synuclein and not human α- synuclein. The mice were healthy and α-synuclein didn’t form aggregates. It still wasn’t confirmed that siRNA will be as effective in reducing aggregated α-synuclein levels as it was with normal mouse α-synuclein. Therefore, the next experiment was done on
transgenic (Tg13) mice with human α-synuclein mimic. In human PD patients α-synuclein is phosphorylated at S129 and formes aggregates [26]. They used mice model with S129D α-synuclein. This mutation mimics the phosphorylation and S129D α-synuclein forms aggregates in mouse brain.
Figure 11. S129D α-synuclein (WB) and mRNA (PCR) levels in Tg13 mice 7 days after the injection compared
to untreated (control) Tg13 mice. Measured in midbrain, striatum, and cortex [18]
mRNA levels are reduced significantly in all three brain regions to approximately 50%
compared to mRNA level in untreated mice. Protein levels also decreased to about 70% of the control (Fig. 11). To ensure these results weren’t random Cooper et al. did another experiment in which they did everything the same except siRNA had mutation in one base pair. The results of CsiRNA RVG-exosomes treatment of Tg13 mice is presented in Fig. 12.
As can be seen, CsiRNA didn’t cause any reduction in mRNA and S129D α-synuclein levels, confirming that S129D α-synuclein downregulation was due to RNA interference of siRNA [18].
Figure 12. S129D α-synuclein (WB) and mRNA (PCR) levels in Tg13 mice 7 days after the injection of control
siRNA compared to untreated Tg13 mice. Measured in midbrain, striatum, and cortex [18]
Erviti et al. delivered BACE1 siRNA in DC-derived exosomes to mouse brain and obtained a 60% reduction in both BACE1 and BACE1 mRNA expression [23].
It was shown that systemic injections of α-synuclein siRNA can significantly reduce the levels of mouse α-synuclein as well as human S129D α-synuclein in aggregates. The success siRNA systemic injections opens a possibility for clinical trials and the development of therapeutic strategies for other applications, such as Alzheimer’s or Huntington’s diseases.
Conclusion
Drug delivery nowadays is at the point when there are plenty of applications but
therapeutic approaches are still in development. Exosomes appear to be an effective drug delivery model due to their low cytotoxicity, prolonged circulation time and targeted drug delivery.
Parkinson’s disease is the 2nd most common neurodegenerative disease with no cure found today. There are certain strategies that can help to halt the progression of the disease but their delivery is very inefficient due to poor BBB permeability by foreign molecules.
Exosomes offer a good platform for the delivery of therapeutic agents, not only because they can cross BBB but they also preserve the activity of cargo drug [19].
Exosomes can be isolated from the host organism and purified, though purification strategies are not very efficient. Exosomes can be loaded in vitro with different kinds of cargo molecules, e.g. proteins, mRNA. A good way to load small hydrophilic molecules is with incubation at RT, whereas large proteins are better incorporated by sonication or extrusion [16]. Addition of saponin during incubation at RT can increase exosomal loading.
mRNA or miRNA are best loaded with electroporation.
It was presented that the therapeutic strategies of using exosomes for drug delivery to relieve PD symptoms were successful. Exosomal drug delivery shows better result in delivering catalase to the relevant PD brain regions than bulk catalase. Widespread distribution of cargo in the brain leads to diverse positive effects. Experiments on mice show no abnormalities caused by the administration of exosome drug nanoformulations.
With more knowledge of exosomal biology and blood-brain barrier physiology exosomes can be applied in clinical trial for PD patients.
Exosomes are an effective platform for drug delivery, showing better results in recent years than liposomes, exosome-like nanoparticles or bulk therapeutic agents. Due to their intrinsic ability to cross BBB, preservative function and immense robustness exosomes promise to be a success in the field of drug delivery in future.
References
1. Batrakova, E. and Kim, M. (2015). Using exosomes, naturally-equipped nanocarriers, for drug delivery. Journal of Controlled Release, 219, pp.396-405.
2. Dreyer, F. and Baur, A. (2016). Biogenesis and function of exosomes and
extracellular vesicles. Lentiviral Vectors and Exosomes as Gene and Protein Delivery Tools, 1448, pp.201-213
3. Bunggulawa, E., Wang, W., Yin, T., Wang, N., Durkan, C., Wang, Y. and Wang, G.
(2018). Recent advancements in the use of exosomes as drug delivery systems. Journal of Nanobiotechnology, 16(1).
4. Hessvik, N. and Llorente, A. (2017). Current knowledge on exosome biogenesis and release. Cellular and Molecular Life Sciences, 75(2), pp.193-208.
5. Keller, S., Sanderson, M., Stoeck, A. and Altevogt, P. (2006). Exosomes: From
biogenesis and secretion to biological function. Immunology Letters, 107(2), pp.102- 108.
6. van Dommelen, S., Vader, P., Lakhal, S., Kooijmans, S., van Solinge, W., Wood, M. and Schiffelers, R. (2012). Microvesicles and exosomes: Opportunities for cell-derived membrane vesicles in drug delivery. Journal of Controlled Release, 161(2), pp.635- 644.
7. Jansen, F. and Li, Q. (2017). Exosomes as Diagnostic Biomarkers in Cardiovascular Diseases. Advances in experimental medicine and biology, 998, pp. 61-68
8. Whiteside, T.L. (2016). Tumor-derived exosomes and their role in cancer progression. Advanced clinical chemistry, 74, pp.103-141
9. Liu, D., Cheng, Y., Cai, R., Wang, BD, W., Cui, H., Liu, M., Zhang, B., Mei, Q. and Zhou, S.
(2018). The enhancement of siPLK1 penetration across BBB and its anti glioblastoma activity in vivo by magnet and transferrin co-modified
nanoparticle. Nanomedicine: Nanotechnology, Biology and Medicine, 14(3), pp.991- 1003.
10. McNatt, M., McKittrick, I., West, M. and Odorizzi, G. (2007). Direct Binding to Rsp5 Mediates Ubiquitin-independent Sorting of Sna3 via the Multivesicular Body Pathway. Molecular Biology of the Cell, 18(2), pp.697-706.
11. Ishida, T., Kashima, S. and Kiwada, H. (2008). The contribution of phagocytic activity of liver macrophages to the accelerated blood clearance (ABC) phenomenon of PEGylated liposomes in rats. Journal of Controlled Release, 126(2), pp.162-165.
12. Sterzenbach, U., Putz, U., Low, L., Silke, J., Tan, S. and Howitt, J. (2017). Engineered Exosomes as Vehicles for Biologically Active Proteins. Molecular Therapy, 25(6), pp.1269-1278.
13. McKelvey, K., Powell, K., Ashton, A., Morris, J. and McCracken, S. (2015). Exosomes:
Mechanisms of Uptake. Journal of Circulating Biomarkers, 4, p.7.
14. Hood, J. (2016). Post isolation modification of exosomes for nanomedicine applications. Nanomedicine, 11(13), pp.1745-1756.
15. Ceña, V. and Játiva, P. (2018). Nanoparticle crossing of blood–brain barrier: a road to new therapeutic approaches to central nervous system diseases. Nanomedicine, 13(13), pp.1513-1516.
16. Haney, M.J., Klyachko, N.L., Zhao, Y., Gupta, R., Plotnikova, E.G., He, Z., Patel, T., Piroyan, A., Sokolsky, M., Kabanov, A.V., Batrakova, E.V. (2016). Exosomes as drug delivery vehicles for Parkinson’s disease therapy. Journal of control release, 207, pp.18-30
17. Sveinbjornsdottir, S. (2016). The clinical symptoms of Parkinson’s disease. Journal of neurochemistry, 139, pp.318-324
18. Cooper, J., Wiklander, P., Nordin, J., Al-Shawi, R., Wood, M., Vithlani, M., Schapira, A., Simons, J., El-Andaloussi, S. and Alvarez-Erviti, L. (2014). Systemic exosomal siRNA delivery reduced alpha-synuclein aggregates in brains of transgenic mice. Movement Disorders, 29(12), pp.1476-1485.
19. Haney, M., Zhao, Y., Li, S., Higginbotham, S., Booth, S., Han, H., Vetro, J., Mosley, R.,
20. Pahuja, R., Seth, K., Shukla, A., Shukla, R., Bhatnagar, P., Chauhan, L., Saxena, P., Arun, J., Chaudhari, B., Patel, D., Singh, S., Shukla, R., Khanna, V., Kumar, P., Chaturvedi, R.
and Gupta, K. (2015). Trans-Blood Brain Barrier Delivery of Dopamine-Loaded Nanoparticles Reverses Functional Deficits in Parkinsonian Rats. ACS Nano, 9(5), pp.4850-4871.
21. Qu, M., Lin, Q., Huang, L., Fu, Y., Wang, L., He, S., Fu, Y., Yang, S., Zhang, Z., Zhang, L.
and Sun, X. (2018). Dopamine-loaded blood exosomes targeted to brain for better treatment of Parkinson's disease. Journal of Controlled Release, 287, pp.156-166.
22. Magee, W.E., Cronenberger, J.H., Thor, D.E. (1978). Marked Stimulation of Lymphocyte-mediated Attack on Tumor Cells by Target-directed Liposomes Containing Immune RNA. Cancer research, 38, pp.1173-1176
23. Alvarez-Erviti, L., Seow, Y., Yin, H., Betts, C., Lakhal, S., Wood, M.J.A. (2011). Delivery of siRNA to the mouse brain by systemic injection of targeted exosomes. Nature biotechnology, 29, pp.141-145
24. Villarroya-Beltri, C., Baixauli, F., Guitiérrez-Vázquez, C., Sánchez-Madrid, F.,
Mittelbrunn, M., (2015). Sorting it out: regulation of exosome loading. Semin cancer biology, 28, pp.3-13
25. Hung, M.E. and Leonard, J.N. (2015). Stabilization of exosome-targeting peptides via engineered glycosylation. The journal of biological chemistry, 290, pp.8166-8172 26. Anderson, J.P., Walker, D.E., Goldstein, J.M., de Laat, R., Banducci, K., Caccavello, R.J.,
Barbour, R., Huang, J., Kling, K., Lee, M., Diep, L., Keim, P.S., Shen, X., Chataway, T., Schlossmacher, M.G., Seubert, P., Schenk, D., Sinha, S., Gai, W.P. (2006).
Phosphorylation of Ser-129 is the dominant pathological modification of α- synuclein in familial and sporadic Lewy body disease. The journal of biological chemistry, 281, pp.29739-29752
27. Stuffers, S., Wegner, C.S., Stenmark, H., Brech, A. (2009). Multivesicular endosome biogenesis in the absence of ESCRTs. Traffic, 10, pp.925-937
28. Trajkovic, K., Hsu, C., Chiantia, S., Rajendran, L., Wenzel, D., Wieland, F., Schwille, P., Brügger, B., Simons, M. (2008). Ceramide triggers budding of exosome vesicles into multivesicular endosomes. Science, 319, pp.1244-1247
29. Visconti, L., Nelissen, K., Deckx, L., van der Akker, M., Adriaesen, W., Daniels, L., Matheï, C., Linsen, L., Hellings, N., Stinissen, P., Buntinx, F. (2014). Prognostic value of circulating cytokines on overall survival and disease-free survival in cancer patients. Future medicine, 8, pp.297-306
30. Subra, C., Laulagnier, K., Perret, B., Record, M. (2007). Exosome lipidomics unravels lipid sorting at the level of multivesicular bodies. Biochimie, 89, pp.205-212
31. Record, M., Subra, C., Silvente-Poirot, S., Poirot, M. (2011). Exosomes as intercellular signalosomes and pharmacological effectors. Biomedical pharmacology, 81,
pp.1171-1182
32. Yuyama, K., Sun, H., Sakai, S., Mitsutake, S., Okada, M., Tahara, H., Furukawa, J., Fujitani, N., Shinohara, Y., Igarashi, Y. (2014). Decrease amyloid-β pathologies by intracerebral loading of glycosphingolipid-enriched exosomes in Alzheimer model.
The journal of biological chemistry, 289, pp.24488-24498
