Isolation of three different sizes of exosomes in an Asian population with different retinal diseases before and after treatment: preliminary results

ABSTRACT

Exosomes are membranous structures measuring between 40–120 nm that are secreted by various cells of the human body into the body fluid system. Exosomes contain proteins, mRNA, miRNA, and signaling molecules, and physiologically they assist in the intercellular transport of proteins and RNA molecules. In this study, we used an immunoaffinity filter paper platform combined with scanning electron microscopy and microfluidic systems to detect the size of exosomes within the aqueous humor. Eight aqueous humor samples showed three distinct sizes of exosomes that were significantly different on scanning electron microscopy(P < 0.01). We further used nanoparticle tracking analysis to assess the size distribution of exosomes within the aqueous humor. We found significantly different distributions of exosomes between patients with three different ocular diseases and patients with normal cataracts as controls. An obvious peak of exomeres(size around 35 nm)was found in the patients with central retinal vein occlusion and vitreous hemorrhage. Flare-ups of large exosomes(size 90–120 nm)were found in the patients with the inflammatory ocular disease pars planitis. No obvious peaks in exomeres or large exosomes were found in the control group. There was a high association between the distribution of exosomes and the pathogenesis of ocular diseases. After intravitreal anti-vascular endothelial growth factor treatment, the aqueous humor from the patients with neovascular diseases showed a significant reduction in exosomes in nanoparticle tracking analysis. These findings suggest that at least three distinct sizes of exosomes exist in the aqueous humor:(1)exomeres:<35 nm;(2)small exosomes:60–80 nm; and (3)large exosomes:90–120 nm. Different sizes of exosomes may have different implications in normal or diseased eyes.

Research Highlights

• Three different sized exosomes were identified in aqueous humor.

• The distribution of exosome size was significantly different between the patients with inflammatory and neovascularization retinal diseases.

• After intravitreal anti-vascular endothelial growth factor treatment, the aqueous humor from patients with neovascular diseases showed a significant reduction in exosomes in nanoparticle tracking analysis.

Graphical Abstract

KEYWORDS:

• Exosome
• aqueous humor
• nanoparticle tracking analysis
• ocular disease
• exomere

1. Introduction

Exosomes are heterogeneous extracellular vesicles (EVs) which play a role in intercellular communication by transporting proteins, mRNA, miRNA, and signaling molecules [1–4]. Even though exosomes and microvesicles, also known as ectosomes, are both EVs, they have different sizes and biogenesis prior to secretion in response to various cellular activation or stress [5]. Recent findings have indicated that the size of an exosome is correlated with its function [6,7]. However, the distribution of different subtypes of exosomes in ophthalmic diseases has not been fully elucidated. Identifying the mechanisms underlying intercellular communication via EVs, as well as the interaction targets between cells, may lead to novel therapies involving EVs to treat refractory retinal diseases [8,9]. Previous studies have shown that exosomes derived from retinal pigment epithelium (RPE) cells can promote endothelial cell migration and tube formation, which may account for the formation of choroidal neovascularization (CNV) [10,11]. In this study, we used an immunoaffinity filter paper platform combined with scanning electron microscopy (SEM) and nanoparticle tracking analysis (NTA) to detect the size of exosomes within the aqueous humor [12]. The aim of this study was to identify potential early diagnostic methods and therapeutic interventions for blindness-causing vitreoretinal diseases.

2. Material and methods

2.1 Patients and aqueous samples

This study was conducted according to the Declaration of Helsinki and was approved by the Chung Shan Medical University Hospital Institutional Review Board (IRB number: CSMUH No: CS2–19048; Approved on 2019/7/26). All participants in this study signed informed consent forms and agreed to have their data published. We obtained aqueous samples during intravitreal injections, before cataract surgery, or for the measurement of exosomes for retinal diseases from 26 July 2019 to 25 July 2021 at Chung Shan Medical University Hospital. The experiments in this study were conducted using 16 aqueous samples collected from patients with the following diseases: wet age-related macular degeneration (AMD), diabetic macular edema (DME), polypoidal choroidal vasculopathy (PCV), dry AMD with geographic atrophy, myopic CNV, proliferative diabetic retinopathy with tractional retinal detachment, central retinal vein occlusion, vitreous hemorrhage, and pars planitis. In addition, two samples were obtained from patients with senile cataracts as the control group. All of the 18 samplings were well tolerated by the patients with no adverse events.

2.2 Isolation of exosomes from the aqueous humor using immuno-affinity modified filter paper and further examinations using SEM

Exosomes are heterogeneous in size, specific content, and biogenesis route. To bind them, paper devices are coated with capture molecules such as anti-CD63 antibodies. CD63 is a member of the tetraspanin family (which includes CD9, CD63, CD81 and CD82) that is enriched on exosome membranes [13]. We used Grade 5 Whatman® qualitative filter paper (WHA1005110, Merck, Darmstadt, Germany), and the surface of the paper sheet test zone was activated by a brief process with oxygen plasma and then conjugated to anti-human CD63 biotin-conjugated antibodies (AHN16.1/46-4-5, 215–030, Ancell Corporation, Stillwater, MN) using 3-mercaptopropyl trimethoxysilane (175617, Sigma-Aldrich, St. Louis, MI), N-γ-maleimidobutyryloxy succinimide ester (63175, Sigma-Aldrich), and NeutrAvidin (31000, Thermo Scientific, Waltham, MA). After washing and blocking the test zone, we spotted 25 µL of aqueous humor onto the test zone (rate: 5 µL/min), and then washed the test zone with PBS containing 1% (w/v) BSA. The captured EVs were fixed with 10 ul 0.5× Karnovsky’s fixative solution for 10 min, and then rinsed twice with PBS for 5 min. The samples were then dehydrated through 35% (10 min), 50% (2 × 10 min), 70% (2 × 10 min), 95% (2 × 10 min), and 100% (4 × 10 min) ethanol. The samples were then subjected to critical point drying and sputter-coated with palladium/gold, and examined using a scanning electron microscope with a low electron acceleration voltage (~5 kV) . The SEM model used in this study was a Hitachi S-4300 scanning electron microscope (Hitachi Ltd., Tokyo, Japan) at the Instrument Technology Research Center (Hsinchu, Taiwan). The details of this protocol were described in our previous study [12].

2.3 Determining the exosome sizes in aqueous humor using NTA

NTA was developed to examine the size of nanoparticles including exosomes using the properties of light scattering and Brownian motion. In NTA, a laser beam with a width of approximately 50 μm is passed through a chamber filled with particles suspended in liquid. The light scattered by the particles is captured and recorded by a microscope camera. The recorded video is then analyzed using NTA software with the Stokes-Einstein equation to calculate the hydrodynamic diameter of individual particles. In this study, NTA was performed on a NanoSight NS300 system (Malvern Panalytical Ltd., Malvern, UK), and the results were displayed as a particle size distribution plot (Y axis: Concentration; X axis: Size) or 2D intensity vs. size scatter plot (Y axis: Intensity; X axis: Size).

2.4 Statistical analysis

SPSS software version 16 (IBM Inc., Armonk, NY) was used for data analysis. Results are expressed as mean ± standard deviation (SD), and P < 0.05 was considered significant.

3. Results

EVs have been shown to be involved in intercellular communication in ocular diseases. Due to the blood-brain barrier (BBB) and blood-retinal barrier (BRB), the pathogenesis of ocular diseases is influenced less by EVs in the systemic circulation. Thus, understanding EVs within ocular fluids has become increasingly important in major eye diseases leading to blindness such as diabetic retinopathy and AMD.

In order to understand the differences in quantity and characteristics of EVs between different ocular conditions, we obtained 16 aqueous humor samples from patients with different retinal diseases (wet AMD, DME, PCV, geographic atrophy, myopic CNV, proliferative diabetic retinopathy with tractional retinal detachment, central retinal vein occlusion, and vitreous hemorrhage) and two aqueous humor samples from patients before undergoing cataract surgery as controls. The diagnoses of all diseases were confirmed by retinal color fundus photography, optical coherence tomography, and fluorescent angiography (Figure 1) [14]. Exosomes from each 25 ul aqueous humor samples were captured on anti-CD63 antibody-immuno-affinity modified filter paper, and their physical diameters were measured using SEM. Although measuring the diameter of exosomes on SEM is not completely accurate, the SEM images showed that the exosomes within the aqueous humor could be approximately separated into three subtypes (Figure S1).

Figure 1. Aqueous humor exosomes in different ophthalmic diseases. Patient A: wet age-related macular degeneration (wet AMD); patient B: diabetic macular edema (DME); patient C: polypoidal choroidal vasculopathy (PCV); patient D: geographic atrophy; patient E: myopic choroidal neovascularization (myopic CNV). Patient F: proliferative diabetic retinopathy with tractional retinal detachment. Patient G: central retinal vein occlusion. Patient H: vitreous hemorrhage. Retinal color fundus photos (A1-H1), optical coherence tomography (A2-H2), and fluorescent angiography (A3-H3) were used to make the diagnosis×. Exosomes from aqueous humor were captured on anti-CD63 antibody-immuno-affinity modified filter paper, then images were obtained by scanning electron microscopy (A4-H4).

* Figure s1 and s3 images are similar to the figures that previously reported in Hsiao et al. (Diagnostics 2021, 11(7), 1276., Reference [14]),because we use the different samples from the same patients. Fundoscopic and fluorescein angiographic examinations of the same patients have similar presentation and they represent different timing respectively.

We then performed NTA in the patients with central retina artery occlusion without treatment Figure 2(a,e), retinal artery occlusion with treatment Figure 2(b,f), pars planitis Figure 2(c,g) vitreous hemorrhage Figure 2(d,h) and central retinal vein occlusion before or after treatment (Figure S2(a). The figures show the distributions of particle size in the aqueous fluid. In the patient with central retinal vein occlusion without treatment, the major peak was at 68 nm (red arrow, Figure 2(a), and the total concentration was 2.62 × 10⁹ ± 9.97×10⁸ particles/ml. In the patients with central retinal vein occlusion without treatment, the major peaks were at 65 and 72 nm (Figure S2(a, c)), and the total concentrations were 1.57 × 10⁹ ± 8.24×10⁷ and 1.93 × 10⁹ ± 1.58×10⁸ particles/ml. In the patient with ocular inflammatory disease (pars planitis) without treatment, the total concentration was 4.47 × 10⁹ ± 9.22×10⁸ particles/ml. The peak concentrations of nanoparticles were at 113 and 159 nm (red arrow, Figure 2(c)), which represents large exosomes that are known to play an important role in the IL-2/STAT5 signaling pathway and cause an inflammatory response in the pathogenesis of ocular diseases. In the patient with a vitreous hemorrhage, the total concentration was 1.18 × 10⁹ ± 5.02×10⁸ particles/ml (Figure 2(d)), and the peak concentrations of nanoparticles were at 50 and 88 nm (red arrow, Figure 2(d)). These peaks represent exomeres that play an important role in the mTOR signaling pathway, and cause a neovascularization response in the pathogenesis of ocular diseases. Comparing the particle sizes in these five patients, the patients with central retinal vein occlusion (blue circle, Figure 2(e)), central retinal vein occlusion with or without vitreous hemorrhage before treatment (Figure S2(a, c)) and vitreous hemorrhage (blue circle, 2 H) had remarkably higher concentrations of particles <100 nm than the patient with ocular inflammatory disease (blue circle, Figure 2(g)). We also took serial aqueous humor samples for analysis. In the patient with central retinal vein occlusion, 1 month after an anti-vascular endothelial growth factor (VEGF) injection (Figure 2(b)), the total concentration had decreased to 2.90 × 10⁸ ± 2.02×10⁸ particles/ml compared to the untreated sample 2.62 × 10⁹ ± 9.97×10⁸ particles/ml (Figure 2(a)). Compared to before treatment (blue circle, Figure 2(e)), the concentration of particles with a size <100 nm was significantly reduced after treatment (blue circle, Figure 2(f)). In addition, the peak exosome size was 48 nm after treatment, and this size represents exomeres (<50 nm) which are thought to play an important role in neovascularization in ocular diseases (Figure 2(b)). Similar results were also found in the patient with central retinal vein occlusion, and a reduction in exosome concentration was found 1 week after treatment with anti-VEGF antibodies (Figure S2(a, b); 1.57 × 10⁹ ± 8.24×10⁷ to 6 × 10⁸ ± 4.54×10⁷; Figure S2(c, d); 1.93X10⁹ ± 1.58×10⁸ to 3.55 × 10⁸ ± 3.42×10⁷). The major reductions in exosome size were Exo-S (50–80 nm) and Exo-L (80–120 nm) (Figure S2(e)). In the control group, two major clusters had the same distribution pattern, with one cluster at 100 nm and the other at approximately 200 nm (red circle, Figure 2(i,j) Compared with the control group, the disease groups had a clearly different pattern Figure 2(e-j).

Figure 2. The distribution of particle size in aqueous fluid of patients with central retinal vein occlusion, ocular inflammatory and vitreous hemorrhage. The Y-axis represents exosome concentration (particles/ml) and the X-axis represents the size of each exosome (in nanometers). (a) Patient with central retinal vein occlusion without treatment. The major peak was at 68 nm width (red arrow). The total concentration was 2.62 × 10⁹ ± 9.97×10⁸ particles/ml. (b) One month after the patient with central retinal vein occlusion received an anti-vascular endothelial growth factor injection. After treatment, the total concentration decreased to 2.90 × 10⁸ ± 2.02×10⁸ particles/ml, and the major peak was at 48 nm width (red arrow). (c) Patient with ocular inflammatory disease(pars planitis) without treatment. The total concentration was 4.47 × 10⁹ ± 9.22×10⁸ particles/ml. The peak concentrations of nanoparticles were at 113 and 159 nm (red arrow). (d) Patient with vitreous hemorrhage. The total concentration was 1.18 × 10⁹ ± 5.02×10⁸ particles/ml. The peak concentrations of nanoparticles were at 50 and 88 nm (red arrow). Nanoparticle tracking analysis showing exosome size (X axis) and signal intensity (Y axis) in the aqueous humor from patients with different ocular conditions. Each sample was tested three times. The patient with central retinal vein occlusion (e,g) had a large number of exomeres, however the number obviously decreased after treatment. (g) The patient with ocular inflammation (pars planitis) had mainly large exosomes. (h) The exosomes in the patient with vitreous hemorrhage were mostly exomeres. The control group had two clusters of exosomes that were different compared with the disease group. In the control group, there were two major clusters with the same distribution pattern, which one cluster at 100 nm and the other at approximately 200 nm. Compared with the control group, the disease groups had a clearly different pattern.

4. Discussion

This is the first study to isolate and analyze different sizes of exosomes in the aqueous humor from patients with different retinal diseases before and after treatment. Compared to current exosome separation technologies, our isolation method using a paper-based immuno-affinity assay is cost-effective, precise, requires only 25 µL of liquid sample, and only takes 1 hour to process. We also used NTA to analyze exosomes within the aqueous humor from patients with different ocular diseases.

There are several important findings to this study. First, using a paper-based immunoaffinity assay and analysis of SEM images, three significantly different sizes of exosomes were found among eight aqueous humor samples from patients with eight different retinal diseases (wet AMD, DME, PCV, dry AMD with geographic atrophy, myopic CNV, proliferative diabetic retinopathy with tractional retinal detachment, central retinal vein occlusion, and vitreous hemorrhage). The exosomes were classified according to their diameter as being large (80–120 nm), small (60–80 nm), and exomeres (<50 nm). This is consistent with the findings of Zhang et al., who identified three exosome subpopulations and classified them as large exosome vesicles (Exo-L, 90–120 nm), small exosome vesicles (Exo-S, 60–80 nm), and exomeres (~35 nm). The authors suggested that their distinct biological functions were due to their diverse organ biodistribution patterns [13]. Using NTA in our study, we found a higher distribution of exomeres in the patients with retinal diseases related to neovascularization subsequent to retinal ischemia (i.e. central retinal vein occlusion and vitreous hemorrhage), and a higher distribution of large exosomes in the patient with retinal disease related to inflammation (pars planitis). This corresponds to Zhang’s study, in which exomeres were selectively enriched in proteins involved in metabolism, coagulation, and hypoxia, whilst Exo-L containing proteins were involved in IL-2/STAT5 signaling pathways, a complex signaling pathway of inflammation and immunomodulation [6,15]. In the current study, NTA showed different size distributions among the patients with central retinal vein occlusion, pars planitis, vitreous hemorrhage, and the control group. We speculate that different sizes of exosomes are distinct nanoparticle subsets which play different roles in the pathophysiology of retinal diseases. In addition, a recent study showed that unprocessed aqueous humor samples were suitable for NTA and single-particle interferometric reflectance imaging sensor (SP-IRIS) analysis [16]. The ability to use unprocessed aqueous humor samples to differentiate between different diseases may allow for the wider application of aqueous humor samples.

In addition, we demonstrated a change in size distribution after retinal disease treatment. There were significant reductions in the concentrations of all sizes of exosomes after anti-VEGF treatment in the patient with central retinal vein occlusion (Table 1). Moreover, exomeres still represented the highest peak after anti-VEGF treatment. This may suggest that exomeres play an important role in the pathophysiology of ischemic retinal diseases. Previous studies have shown that exosomes may contain structural RNA, small ribosomal RNA, fragmented transfer RNA, microRNA and mRNA [1–4]. Other studies have also shown that exosomes may be involved in physiological and pathological processes by transmitting intercellular information and material in cancer patients [17–19]. In some studies, exosomes have been shown to promote angiogenesis related to neoplastic processes, protect limbs from ischemic injury, and regulate myocardial cell repair and functional recovery in stroke patients [20]. In retinal diseases, a previous study reported a link between oxidative stress and low-grade inflammation in the pathophysiology of AMD [21]. Drusen deposition in the macula occurs in early AMD, and the subsequent increase in atrophic changes can lead to geographic atrophy, while a gradual increase in CNV can lead to wet AMD. Recent studies have reported shared molecular machinery as well as substantial crosstalk between exosome biogenesis and autophagy, a lysosomal-dependent degradation and recycling pathway [22]. In one study, drusen in AMD donor eyes contained markers of autophagy and exosomes, and the authors speculated that increased autophagy and the release of intracellular proteins via exosomes from aged RPE cells may contribute to the formation of drusen [23]. Furthermore, an in vitro study demonstrated that oxidative stress in RPE cells increased the secretion of exosomes, VEGF receptor mRNA, and VEGF receptors in the membrane, and promoted angiogenesis in endothelial cells [24], a process related to the development of CNV in wet AMD. Our findings may suggest that differently sized exosomes may carry different translational messages, thereby influencing the development of different retinal diseases. In geographic atrophy, the progressive expansion of atrophic RPE cells, overlying retina and underlying choriocapillaris forms a large and sharply defined area in the macula. The cause of geographic atrophy is unclear even though it has been studied extensively, and currently no treatment is available to halt or reverse its progression. Neuronal cells have been shown to use autophagic degradation and exosome secretion to eliminate protein aggregates, thereby reducing proteotoxicity. Autophagy or lysosomal dysfunction may cause insufficient degradation of intracellular protein aggregates, and cells may alleviate the proteotoxic stress by enhancing exosome secretion [22]. However, secreted exosomes with unwanted protein aggregates may be taken up by neighboring neurons, and this may be responsible for the propagation of the disease [25]. Taken together, we speculate that the increased secretion of exosomes by stressed RPE cells may be taken up by neighboring RPE cells which are relatively healthy, and that this contributes to the expansion of geographic atrophy. With regards to retinal diseases involving neovascularization and an inflammatory response, the secretion of EVs is complicated by various intercellular reactions. Although we found a change in size distribution of exosomes between the samples before and after treatment, further studies with more specimens are needed to elucidate the mechanisms underlying the interactions. In addition, ocular diseases such as diabetic macular edema involve intraocular VEGF-related neovascularization and inflammatory processes such as interleukin-1. This further complicates the composition and characteristics of EVs in aqueous humor, and therefore further studies with more samples and consistent conditions are needed to investigate these issues.

Table 1. Patient information. The concentrations of EVs decreased after anti-VEGF treatment in the patient with central retinal vein occlusion. In addition, the size distribution of EVs changed after anti-VEGF treatment.

Diagnosis	Age	Right/Left eye	Onset EVs	After anti-VEGF EVs	Changes in different diameter of exosome
Central retinal vein occlusion with vitreous hemorrhage	>80	Left	1.57 × 10^9 ± 8.24 × 10^7	6 × 10^8 ± 4.54 × 10^7	In Fig S2 E
Central retinal vein occlusion	61–80	Left	1.93 × 10^9 ± 1.58 × 10^8	3.55 × 10^8 ± 3.42 × 10^7	In Fig S2 F
Central retinal vein occlusion	41–60	Right	2.62 × 10^9 ± 9.97 × 10^8	2.90 × 10^8 ± 2.02 × 10^8	N/A

The P value of onset EVs vs After anti-VEGF EVs is 0.0073.

There are several limitations to this study. The number of cases was small, although each test was repeated three times in NTA. Only one patient had each retinal disease, therefore the exosome characteristics described in this study cannot sufficiently represent the whole disease spectrum. In addition, we could only obtain 100 µL of aqueous humor during each sampling. However, we believe that our isolation method will increase understanding of the characteristics and roles of exosomes in the aqueous humor of patients with different ocular diseases.

Conclusion

Our solid findings of the different sizes and distribution of exosomes in various retinal diseases before and after treatment may provide a novel perspective for molecular studies of ocular diseases and clinical application including the diagnosis and treatment of such diseases. Our preliminary results, combined with the technique of isolating exosomes using immuno-affinity filter paper with tetraspanin antibodies and exosome diameter assessment using NTA, may prompt further investigations of the potential role of exosomes in response to diverse cellular activation or stress.

Author contributions

Conceptualization: Sung-Yu Wu, Connie Chen, Chao-Min Cheng, Chihchen Chen, Min-Yen Hsu; Investigation: Jyh-Cheng Liou, Sung-Yu Wu, Yu-Chien Hung, Chao-Min Cheng, Chihchen Chen, Min-Yen Hsu; Drafting and Writing the Manuscript: Sung-Yu Wu, Yu-Chien Hung, Connie Chen, Min-Yen Hsu; Funding acquisition: Min-Yen Hsu.

Availability of data and material

The data that support the findings of this study are available on request from the corresponding author.

Ethics approval and consent to participate

This study was approved by the review board of the Ethics Committee of Chung Shan Medical University Hospital [IRB number: CSMUH No: CS2–19048]. Informed consent was obtained from all participants included in this study.

Acknowledgments

We thank Dr. Yu-Ping Wang (Department of Radiology, Chiayi Branch, Taichung Veterans General Hospital) for his assistance in creating the illustration for the Graphical Abstract. We also thank ATS Medical Editing for editing the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21655979.2023.2297320

Additional information

Funding

This study was supported in part by intramural grants from Chung Shan Medical University Hospital [Grant No. CSH-2022-C-003], and the National Science and Technology Council of Taiwan (111-2636-E-040-001, Columbus program of MOST Young Scholar Fellowship; 111-2314-B-040-026-MY2).