Effective Refractive Index and Lipid Content of Extracellular Vesicles

Subscriber access provided by - Access paid by the | UCSB Libraries

Interface Components: Nanoparticles, Colloids, Emulsions, Surfactants, Proteins, Polymers

Effective refractive index and lipid content of extracellular vesicles revealed using optical waveguide scattering and fluorescence microscopy Deborah L. M. Rupert, Mokhtar Mapar, Ganesh Vilas Shelke, Karin Norling, Mathias Elmeskog, Jan O. Lotvall, Stephan Block, Marta Bally, Björn Agnarsson, and Fredrik Höök Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b04214 • Publication Date (Web): 20 Jun 2018 Downloaded from http://pubs.acs.org on June 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Effective refractive index and lipid content of extracellular vesicles revealed using optical waveguide scattering and fluorescence microscopy Déborah L. M. Rupert1,#, Mokhtar Mapar1,#, Ganesh Vilas Shelke2, Karin Norling, Mathias Elmeskog1, Jan O. Lötvall2, Stephan Block1, Marta Bally1,3, Björn Agnarsson1, Fredrik Höök1* 1)

2)

Department of Applied Physics, Chalmers University of Technology, Gothenburg, Sweden

Krefting Research Centre, Department of Internal Medicine and Clinical Nutrition, University of Gothenburg, Gothenburg, Sweden

3)

Institut Curie, Centre de Recherche, CNRS, UMR168, Physico-Chimie Curie, Paris, France #) These authors contributed equally

1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

ABSTRACT Extracellular vesicles (EVs) are generating a growing interest due to the key roles they play in various biological processes and because of their potential use as biomarkers in clinical diagnostics and as efficient carriers in drug-delivery and gene-therapy applications. Their full exploitation, however, depends critically on the possibility to classify them into different subpopulations, a task that in turn relies on efficient means to identify their unique biomolecular and physical signatures. Due to the large heterogeneity of EV samples, such information remains rather elusive and there is accordingly a need for new and complementary characterization schemes that can help expanding the library of distinct EV features. In this work, we used surface-sensitive waveguide scattering microscopy with single EV resolution to characterize two subsets of similarly sized EVs that were pre-separated based on their difference in buoyant density. Unexpectedly, the scattering intensity distribution revealed that the scattering intensity of the high density (HD) population was on average a factor of 3 lower than that of the low density (LD) population. By further labeling the EV samples with a self-inserting lipidmembrane dye, the scattering and fluorescence intensities from EVs could be simultaneously measured and correlated at the single particle level. The labelled HD sample exhibited not only lower fluorescence and scattering intensities but also lower effective refractive index (n ~ 1.35) compared with the LD EVs (n ~1.38), indicating that both the lipid and protein content was indeed lower in the HD EVs. While separation in density gradients of similarly sized EVs is usually linked to differences in biomolecular content, we suggest based on these observations that the separation rather reflects the ability of the solute of the gradient to penetrate the lipid membrane enclosing the EVs; i.e. the two gradient bands are more likely due to differences in membrane permeability than to differences in biomolecular content of the EVs.


Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

INTRODUCTION Extracellular vesicles (EVs) are spherical nanoscopic vesicles secreted by most cell types [1-3] that consist of an encapsulating membrane composed of lipids and membrane proteins that protect an aqueous lumen, often containing both genetic material and soluble proteins [4]. Microvesicles and exosomes are two different types of EVs that differ by their formation pathway; microvesicles originate from direct budding of the cell membrane, while exosomes are vesicles of cytoplasmic origin that are released into the extracellular environment through fusion of multivesicular bodies with the plasma membrane.[5] EVs and in particular exosomes have been identified to play key roles in intercellular communication processes such as modulation of the immune system, inflammation reactions and tissue regeneration.[6-10] In addition, they are also considered promising candidates in clinical diagnostics of several health disorders such as cancer [11-13], as well as for drug-delivery and gene-therapy applications.[14-17] The biomolecular composition of EVs depends on various parameters such as cellular origin, formation pathway and size and is presently subject to thorough investigations.[14] Current approaches used to distinguish microvesicles from exosomes rely on various physicochemical criteria such as size (30-120 nm for exosomes and 100 nm-1 µm for microvesicles) and on the presence of specific markers, such as tetraspanin membrane proteins (CD63, CD9, CD81) and luminal proteins (ALIX, TGS101) for exosomes, and external phosphatidyl serine lipids (PS) for microvesicles.[18-21] To illustrate the complexity of the subject, it is worth noting that many of these markers were found to be present on both exosomes and microvesicles, although with different degrees of relative abundance. For example, specific markers such as CD9, CD63 CD81 [19, 22, 23] and various nucleic acid components (miRNAs, RNA and DNA) [24, 25] were reported in both exosomes and microvesicles, while their relative


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 31

distribution between exosomes and microvesicles remains to be determined.[26] Such observations challenge the currently established classification of sub-populations of EVs and illustrate the need for additional physicochemical criteria to aid the identification of both unique and common properties of EVs originating from the same or different cellular pathways. Protocols to aid discrimination of sub-populations of EVs have so far been mostly focused on immune-recognition assays to identify the presence of specific biological markers on their surface.[27-30] Western Blot and enzyme-linked immunosorbent assays (ELISA) were used to confirm the presence of EVs by screening for the presence of specific membrane-bound biomarkers [31-33]. Specific biomarkers on the outer surface of EVs can also be detected with flow cytometry by first binding EVs to micron-sized beads functionalized with capturing antibodies followed by detection using additional fluorescently labeled antibodies,[34, 35] while electron microscopy combined with gold-immunostaining can provide direct observation of single EVs carrying specific markers of interest.[36, 37] In yet another approach, Wyss et al. utilized fluorescence correlation spectroscopy (FCS) and fluorescence-based immunostaining to determine the average amount of surface exposed biomarkers on EVs.[38] As a complement to biomolecular content, physical properties such as size, structure and density have also been reported to signal specific formation pathways.[39] Such properties have been determined by combining established and newly developed characterization methods, such as dynamic light scattering (DLS)[40, 41], nanoparticle tracking analysis (NTA)[42, 43], electron microscopy[37, 44, 45], atomic force microscopy (AFM)[46, 47], surface plasmon resonance (SPR)[48-50], quartz crystal microbalance (QCM)[51] and fluorescence correlation spectroscopy (FCS)[38, 52].


Page 5 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Since first applied to characterize EVs by Dragovic et al. in 2011, [53] NTA has grown enormously in popularity as a means to determine the size distribution and concentration of suspended EVs. The operation principle of NTA is based on visualizing individual suspended nanoparticles by detecting the light scattered by the particles upon laser illumination. Particle tracking to determine their Brownian motion then yields their size distribution.[53] Besides operation in scattering mode, NTA can also be operated in florescence mode, which allows for determining the size distribution of sub-populations of EV being fluorescently labeled using antibodies or alternative ligands.[53] Labeling with florescent lipophilic tracers which self-insert into lipid bilayer membranes represents an alternative means to visualize all lipid vesicles present in EV samples. Such a labeling approach was utilized by Hoen et al. to determine the concentration and size distribution of EVs using high sensitivity flow cytometry and by Kanwar et al. to visualize exosomes present in blood samples after capture on the microfluidic wall via immuno-recognition [54]. Another important, yet less investigated, physical characteristic of EVs is their refractive index. EVs have dimensions in the order of the wavelength of visible light, which should in principle make it possible to quantify refractive index distributions from light scattering intensity profiles detected using e.g. NTA[55]. Van der Pol et al. proposed to take advantage of the scattered light intensity of single particles detected with NTA, as a means to determine the effective refractive index of individual EVs using monodisperse polystyrene beads with known refractive index for calibration purposes[43]. A limitation of this approach is, however, that the scattering intensity measured from suspended single particles is subject to significant uncertainties, because of i) intensity variations due to the random motion the particles across the imaging focal plane ii) uneven illumination of the observation volume and iii) differences in


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 31

track length leading to statistical uncertainties of the intensity value. A common origin for these uncertainties is associated to the three-dimensional motion of the particles. Although the measurement accuracy can be partially improved by considering the highest scattering intensity measured for each particle trace[43], these uncertainties would be drastically reduced upon surface-immobilization of the nanoparticles and visualization of the scattered signal with surface-sensitive microscopy. With the aim to provide new insights with respect to relative scattering intensity distribution, effective refractive index as well as lipid and protein content of EVs, we have in this work explored the possibility to use a newly developed waveguide-based scattering and fluorescence microscopy setup[56] to investigate two subsets of EVs that were separated from a human mast cell line according to their difference in buoyant density, corresponding to 1.15 ± 0.06 g/cm3 and 1.27 ± 0.03 g/cm3 for low density (LD) and high density (HD) EV samples, respectively. Given that both EV subsets had similar size, EVs of higher buoyancy should imply a higher effective refractive index and therefore a higher scattering intensity compared to EVs of lower buoyant density. However, the analysis revealed that the difference in scattering intensity distribution between the LD and HD population contradicts the expected refractive index difference. In fact, complementary measurements using fluorescence labeling of the lipid membrane of the EVs using the self-inserting membrane dye PKH26 to estimate the lipid-to-protein ratio of the EVs, lead us to suggest that the separation based on buoyant density rather originates from differences in membrane permeability than significant differences in the biomolecular content of the two subsets of EVs.


Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

RESULTS The heart of the microscopy setup used in this work is a waveguide chip made of a planar 400 nm thick silica core encapsulated by a 4 μm thick CYTOP cladding layer on each side, as schematically shown in Fig. 1a. Linearly polarized laser light (TE,  ~532nm) is coupled to the chip via a single mode optical fiber aligned to the facet of the chip. To obtain a sensing surface, part of the upper cladding layer is partially removed to achieve physical access to the silica core, thereby creating a shallow well for carrying out observation in an aqueous solutions (Fig. 1a). The close refractive index matching between the cladding layer (CYTOPTM, n = 1.34) and that of water (n = 1.33) allows for a non-disruptive symmetric propagation of the evanescent wave along the solid-liquid interface.[57] As described previously, this configuration results in an illumination profile on the sensor surface with an exponentially decaying sensing depth of ~120 nm and sufficiently low background scattering to resolve the scattering signal from objects as faint as single sub 80 nm lipid vesicles.[56] Further, by using optical filters, fluorescence-based read-out of labeled objects, such as EVs, can be conducted in parallel to the label-free scattering operation mode (Fig. 1b).


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 31

Figure 1: a) Schematic illustration of the waveguide microscopy setup: a water-immersion objective placed in the droplet positioned in the opening to the core of the waveguide chip made it possible to observe single surface-bound EVs both in scattering and fluorescence mode. The waveguide chip consists of a 400 nm silica core layer embedded in a 4 µm thick symmetric CYTOP cladding. The coupling of an optical fiber to a facet of the chip induced a planar singlemode wave propagating in the cladding layer to reach the solution-filled well created by local removal of CYTOP layer. b) In a typical experiment, EVs and polystyrene beads were adsorbed on the glass core of the waveguide chip in sequential injection steps and their respective scattering and fluorescence intensities were recorded: 1) polystyrene calibration beads observed in scattering mode, 2) unlabeled EV particles observed in scattering mode and 3) PKH26 fluorescently labeled EVs particles observed in both fluorescence (top) and scattering (bottom) mode (scale bar: 2 µm). In order to compare the scattering intensity distributions of the LD and HD populations, the different EV suspensions were adsorbed on different waveguide chips and illuminated at a wavelength, , of 532 nm at a fix power of ~1 mW (see Materials and Methods for details). To compensate for possible variations in illumination intensity between different chips, monodispersed polystyrene-sulfate beads with a modal diameter determined using AFM to ~63 nm /FWHM ~12 nm (see Fig. 2a), and a refractive index of ~1.59 [58] , were used to calibrate the intensity measurements. The size distributions of both EV samples were determined using NTA and were fairly similar, with modal diameters/standard deviations of ~97 nm / ~64 nm and ~109 nm / ~93 nm for the LD and HD population, respectively (Fig. 2a). To avoid false identification of particles due to signal overlap at high surface coverage, the adsorption of EVs and polystyrene beads was deliberately interrupted before the coverage reached 500 entities per field of view. The scattering intensities of adsorbed EVs and the polystyrene beads were then analyzed (see Materials and Methods) as summarized in the scattering intensity histograms shown in Fig. 2b.


Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 2: a) Normalized size distributions of unlabeled LD and HD EV samples (measured with NTA operated in light scattering mode) and polystyrene beads (measured with AFM). b) Scattering intensity distributions of calibration beads and unlabeled LD and HD EV samples plotted on semi-logarithmic scale. Although the size of the polystyrene beads was smaller than that of the EVs, their scattering intensities were somewhat larger (Fig. 2b), which is attributed to a significantly higher effective refractive index of the beads compared to that of EVs. In contrast, despite fairly similar size distributions of the two EV samples, a clear difference between the scattering intensity distributions was observed: using the peak positions (modal values) of the scattering intensity distributions (Fig 2b) and assuming that they correlate with the modal sizes of the two EV populations (Fig 2a), the LD sample had, despite its lower buoyant density, a ~3 times higher scattering intensity than the HD sample. Extraction of the physical parameters that determine the magnitude of the scattering intensity, such as size, refractive index, shape, depth of the illumination profile etc. from this type of scattering intensity data is a challenging task that is not yet theoretically resolved.[56] Rayleigh-


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 31

Gans-Debye (RGD) approximation, however, provides a fairly good estimate on the scattering from optically soft spheres with a diameters up to 240 nm (see Fig. S1 in Supporting Information). In the RGD approximation the dependence of scattering intensity, Is, on nanoparticle radius, r, and refractive index, np, when surrounded by a medium of refractive index, nm, is given by (see Supporting Information for derivation): IS 

1 n 2 6  2  r  n 2 r 6 4  nm

(1)

where n = np – nm. Under the assumption that EVs can be considered as spheres with a homogeneous effective refractive index, np,EV, and that the modal values of their scattering intensities and size distribution can be used as representative values, Eq. 1 may be used to estimate the effective refractive index of the EVs using calibration beads of known refractive index as a reference:

np,EV

1/ 6   I  rbead s,EV    nm  np,bead   I s,bead  rEV    

3

(2)

Here np,bead represents the difference between refractive index of a calibration bead and water,

I s,EV and I s,bead are the average scattering intensities for EVs and calibration beads, respectively, and rbead and rEV are the modal radius of the calibration bead and EVs, respectively. Before applying this model on the EV sample, it was first validated by measuring the scattering intensities from two types of nanospheres of known size and refractive index. The size distribution of suspended polystyrene beads (n=1.591, from ThermoFisher Scientific) and silica spheres (n=1.43-1.46, from Sigma-Aldrich) were determined using NTA (Fig. 3a), while the scattering intensities of individual particles were recorded after adsorption to the surface of a waveguide chip (Fig. 3b). Since the two different samples were subsequently adsorbed on the


Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

sensor surface, they could be safely distinguished (inset Fig. 3b) on the surface despite a significant overlap in scattering intensity (Fig. 3b). With the polystyrene bead as the reference scaterrer, the modal values for the sizes (106 and 144 nm for latex and silica, respectively) and scattering intensities (1.43×105 and 1.62×105 for latex and silica, respectively) of the two populations, yield a refractive index of 1.44 for the silica beads (obtained using Eq. 2 with np,EV replaced for that of silica). This is in excellent agreement with expectations, confirming the validity of the method for determination of refractive index values, if beads with a known size and refractive index are used as reference.

Figure 3: a) Normalized size distributions of silica (orange) and latex (green) beads measured with NTA operated in light scattering mode. b) Waveguide-scattering intensity distributions from surface adsorbed silica (orange) and latex (green) beads. The inset image, 20 μm accross, shows the scattering signal from the two subsequently adsorbed populations (the silica beads are identified using green circles and latex beads using orange circles). Scattering objects found on the surface prior to adsorption were not included in the analysis (unmarked scattering objects).


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 31

With the model validated, the scattering intensities and size distributions obtained from EVs and polystyrene reference beads (Figs. 2a and 2b) were used with Eq. 2 to estimate the effective refractive index of the LD and HD EVs. The LD population have a refractive index of ~1.38, which is in good agreement with the value of 1.37 determined using NTA for EVs from human urine.[43] However, with a refractive index of biomolecules being roughly 1.5, this suggests that only ~12% of the volume of the EV is made up of biological molecules. Taking into consideration that the sensing depth of ~120 nm of the evanescent illumination is similar to the modal diameter of the LD population (~97 nm) but significantly larger than that of the polystyrene beads (modal diameter ~63nm), the volume occupancy of the biomolecular content might approach 15 - 20%[56]. However, even if all of this material consisted of proteins (with a density of ~1.4 to l.5 g/cm3 [59]) rather than lipids (with a density of ~1 g/cm3), the maximum density would still not be larger than ~1.1 g/cm3, which is lower than the buoyant density of ~1.15 g/cm3. The rough approximation used to analyze the scattering response, which neither takes into account the evanescent illumination nor the physical distribution of the biomolecules within the EVs, demonstrates convincing agreement with previous reports[43] suggesting that EVs are indeed relatively dilute entities. This conclusion is even more striking if the same analysis is applied to the HD population, which suggests that this population has a lower effective refractive index (~1.35) than the LD population. Under the assumption that the lipid material is located in the membrane surrounding the EVs, this is in contrast to the expectations from their significantly higher (~1.27 g/cm3) buoyant density. To gain further insight into this peculiar observation, we made use of the possibility to simultaneously observe the scattering and florescence signals of single EVs. Since the density of


Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

lipids is significantly lower (~1 g/cm3) than that of proteins (~1.5 g/cm3) the buoyant density of EVs is expected to depend on their biomolecular content. For this purpose, the lipid membrane of the EVs was modified with PKH26, an aliphatic fluorophore initially designed for high affinity fluorescence labeling of lipid membrane of cells[60] and recently extended to be used as a fluorescent label for lipid membranes of EVs.[61, 62] To evaluate the capacity of PKH26 to target lipid-based nanoparticles, LD and HD EVs and synthetic lipid vesicles composed of 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipids were labeled with PKH26 (see Materials and Methods section) and characterized using NTA in scattering and fluorescence mode. All three samples displayed similar size distributions in both scattering and fluorescence mode (Fig. S2), indicating that PKH26 labeling is not strongly size dependent and as efficient for EVs as for the synthetic POPC system. In waveguide microscopy, surface-immobilized particles remain in the focal plane throughout the analysis. In contrast to NTA, this enables simultaneous detection of both fluorescence and scattering intensities of every individual nanoparticle. While the intensity distribution obtained from the NTA analysis revealed broad and featureless distributions (see Fig S2), the intensity signals of EVs detected using waveguide microscopy displayed log-normal distributions in both scattering and fluorescence mode (Fig. 4). Since the scattering intensity is dependent on both the size and refractive index of the EVs (Eq. 1), and since the fluorescent signal is expected to depend on size only, the positive correlation between fluorescence and scattering intensities within each EV population (Fig. 4) suggests that the total amount of fluorescent lipids increase with the size of the particles, and that the lipid-to-protein ratio is not strongly dependent on the EV size. However, it is clear from Fig. 4 that not only the scattering intensity of the HD population is lower than that of the LD population (right axis: scattering distributions, see also Fig. 2), but also that


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 31

the fluorescent signal of the HD population is lower (top axis: fluorescent distributions) despite their somewhat larger size (Fig. S2). The latter observation can be attributed either to a higher protein-to-lipid ratio of the HD sample or less efficient labeling of the membrane of the EVs. However, since the refractive index of lipids is lower than that of proteins, a higher protein-tolipid ratio contradicts the lower scattering intensity observed for the HD population. Although differences in labeling efficiency cannot be excluded, the correlation observed between fluorescence and scattering intensities within each EV population in the logarithmic representation (Fig. 4) should primarily depend on their effective refractive index (Eq. 1), irrespective of labeling efficiency. Hence, the fact that the HD population has a somewhat lower slope (slope ~1.48) than that of the LD population (slope ~1.74), combined with the fact that the slightly smaller LD population (Fig. 2a) also has the highest scattering intensity (Fig. 2b and Fig. 4), suggest that the LD sample has a higher density core.


Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 4: Correlation between the scattering and fluorescence intensity distributions of PKH26labeled EV populations, obtained using the scattering intensity distribution of polystyrene beads (see Fig. 1) as reference to compensate for chip-to-chip differences in illumination efficiency both for the scattering and fluorescence data (see Materials and Methods). Main: log-log plot of

Iscatt vs. Ifluo for EVs individually visualized both in scattering and fluorescence mode. (top: fluorescence intensity distribution, right: scattering intensity distribution.) It is in the context of this analysis worthwhile to keep in mind that the EVs were upon derivation of Eq. 1 assumed to be spherical with a homogeneous effective refractive index. Inspection of Fig. 3 shows that the slope of the log(Iscatt) versus log(Ifluo) curve is between 1.4 and 1.8 for both EV samples, which is lower than the slope of 3 that is expected if the lipids (dyes) are distributed in the shell ( r2) while other biomolecules are homogeneously throughout the particle ( r3,, Eq. 1). A slope approaching 2, as observed here, is in better agreement with previous observations of the relation between Iscatt and Ifluo for fluorescently labeled hollow lipid vesicles,[56] in which case Iscatt is expected to scale with the volume square, i.e. r4l2, where l is the thickness of the membrane, while Ifluo is expected to scale with the area, i.e. r2. Even if the core of the LD population might be somewhat denser that the HD population, this suggests that not only lipids but also the majority of other biomolecules are primarily located in the outer membrane of the EVs. This does not, however, violate the estimation of the effective refractive index made above, since the permittivity of shell-like lipid vesicle structures can to a good approximation be represented by a sphere with an effective permittivity and corresponding refractive index.[63]

DISCUSSION Quantitative descriptions of the dependence of nanoparticle scattering intensity on size, structure and refractive index (distribution) remains an unresolved theoretical challenge, in particular in


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 31

the context of evanescent wave illumination of surface-bound nanoparticles. Nevertheless, the simplistic analytical model proposed above, together with comparative measurements using well-defined polystyrene beads and the use of a self-inserting fluorescent membrane dye for comparative scattering and fluorescent measurement at the single EV level, demonstrate that i) the two EV subsets analyzed in this work are optically faint objects with effective refractive index between 1.35 (HD) and 1.38 (LD), ii) most of the biomolecular content, which makes up 10 to 20% of the total mass of the EVs, is primarily located in their outer membrane, and iii) both the lipid and the protein content of the LD sample is higher than that of the HD sample. Taken together, this contradicts the intuitive interpretation of the difference in the buoyant density, namely that the amount of biomolecular content per total EV volume should be larger for the HD sample. There is, however, a quite plausible explanation to this discrepancy, which becomes clear if one considers how two clearly separated fractions of EVs in a density gradient emerge. Lipid membranes are expected to be impermeable to both sucrose and iodixanol which were here used to generate the density gradient, and consequently, the solute density that matches that of the EVs should correspond to the mass of both the biomolecules and the water, divided by the volume enclosed by the (solute-impermeable) membrane. Taken the density of lipids of ~1 g/cm3 and typical biomolecules of ~l.5 g/cm3, the average density of 1.15 g/cm3 obtained for the LD population is a fairly reasonable buoyant density for extracellular vesicles with a water-rich lumen. However, a buoyant density approaching 1.3 g/cm3 suggests a very little, if any, water in the lumen of the vesicles, an observation which contradicts their weak scattering signals and accompanied low effective refractive index. Hence, a much more likely explanation to the high buoyant density of the HD population is that the outer membrane of the vesicles belonging to the HD sample is simply permeable to the solute(s) used to generate the


Page 17 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

gradient. With the density of the lumen of the EVs being equilibrated with the external solution, the buoyant density determined from centrifugation in a density gradient would correspond to the average density of the lipids and other biomolecules, a number that for a typical mixture of lipids, proteins and nucleic acids in an EV should be close to the observed average value of 1.27 g/cm3. We thus suggest from this analysis that the physical reason to the appearance of two clearly distinguishable bands of EVs in the density gradient has one dominating origin: a difference in permeability of the outer membrane to the solute of the gradient. While EVs with leaky membranes might still contain valuable information regarding origin and formation pathway, they are not likely to be as biologically functional as EVs with intact outer membranes and thus intact lumen content, which has been reported to contain, among others, genetic material susceptible, for example, to impact the phenotype of recipient cells. To conclusively verify if this interpretation is correct one should in future work put emphasis on i) independent size determination of each individual immobilized EVs, thus avoiding correlations based ensemble-averaged values, ii) the use of multiple labeling compounds to target membranes, proteins and nucleic acids, iii) the development of theoretical models capable of fully describing the relative contribution of the physical properties of adsorbed soft dielectric nanoparticles to the scattering intensity when illuminated by an exponentially decaying evanescent field and iv) solute permeability measurements of different EV populations, preferably with single vesicle resolution. Despite the need for improved instrumentation and theoretical models to gain complete understanding about the physical differences between individual EVs, we conclude from this investigation that the waveguide setup contributes a new means to unambiguously distinguish extracellular vesicles of different types based both on differences in scattering intensity and on the correlation between the scattering intensity and the


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 31

efficiency of membrane labeling with self-inserting dyes. This is a unique asset that distinguishes waveguide microscopy from more commonly utilized EV characterization techniques. For example, in the case of NTA, which is an excellent tool for the determination of the size of suspended nanoparticles, the size of an individual nanoparticle cannot be directly correlated with its scattering (or fluorescence) intensity (see Supporting Information), while waveguide scattering microscopy offers a means to directly correlate scattering and fluorescence signals at the level of individual EVs. The growing need to identify different sub-populations of EVs characterized by the presence of a specific biomarker on their surface makes the implementation of an antibody-coated sensor surface for the selective capture of EVs an additional natural extension of this work. By operating the waveguide at multiple excitation and detection wavelengths, this opens up possibilities for combined scattering and fluorescence based multiplexing, which with single EV resolution should challenge both conventionally used ELISA and flow-cytometry based approaches. By taking advantage of the possibility of temporally resolving the binding process by counting the number of newly surface-bound EVs versus time under diffusion limited conditions, one could also determine the bulk concentration of sub-populations of EVs with very high accuracy, as uncertainties with respect to film thickness, dn/dC value and surface-induced contraction of EVs encountered when using ensemble averaging techniques like surface plasmon resonance (SPR) would be reduced.[50, 64]

MATERIAL AND METHODS Liposomes. Lipid vesicles were prepared by the lipid film hydration and extrusion method[65]. Briefly, 5 mg 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), (Avanti Polar Lipids,


Page 19 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Inc, USA) were dissolved in a methanol/chloroform mixture. Lipid vesicles were prepared by evaporating the solvents first under gentle rotation at reduced pressure (200 mbar) in a 50°C water bath for 30 minutes and then under vacuum overnight. The thin lipid film was then rehydrated in 1 ml of phosphate buffered saline (PBS: 10 mM phosphate buffer, 2.7 mM KCl, 137 mM NaCl, pH=7.4) by vortexing. The resulting suspension was extruded 11 times through two 100 nm polycarbonate membranes (Whatman, UK) and stored at 4°C until usage. Extracellular vesicles. Extracellular vesicles were isolated by differential ultracentrifugation steps from the supernatant of human mast cell line, HMC-1.2 cultured in conditioned media supplemented with exosome-depleted FBS[66]. Briefly, cell supernatants were centrifuged to remove apoptotic bodies and microvesicles (20,000 × g) before pelleting the remaining EV (120, 000 × g). The EV-enriched suspension was further separated into two subsets of different buoyant density using an density gradient and centrifuged for 16 hours at 4 °C in a swing bucket rotor (SW41-Ti rotor) (180,000 × g). Density steps were prepared by mixing “50% iodixanol” solution and “homogenization medium” (HM) in various proportions. Briefly, 50% (w/v) iodixanol was prepared by mixing 5 parts of OptiPrep (Axis-Shield, Norway) (containing 60% iodixanol) with 1 part of 0.25 M sucrose, 6 mM EDTA, 60 mM Tris-HCl at pH 7.4. The components of HM were 0.25 M sucrose, 1 mM EDTA and 10 mM Tris-HCl at pH 7.4. The use of sucrose and iodixanol mixture was to ensure the osmolality of the solution remain in the range of 295–310 mOsm. The separated fractions, named low and high density EVs, were then isolated from the buoyant gradient according to their difference in RNA profile, PBS washed and stored in PBS buffer at -80 °C. The two subsets of EVs have different buoyant densities  : 1.09-1.21 g/cm3 and 1.24-1.31 g/cm3 for low density (LD) EVs and high density (HD) EVs respectively. Lässer et al. showed that both subsets are composed of lipids, proteins and genetic material and that both contain commonly used


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 31

EVs markers such as the CD63 tetraspanin membrane protein.[67] However, the two subsets exhibited different transcriptomics, lipidomics and proteomics profiles, suggesting that the two subsets may have different biological functions. The total protein concentration was determined using the Pierce bicinchoninic acid (BCA) protein assay kit according to the manufacturer’s instructions. The total protein concentrations of stock solutions were 0.65 ± 0.02 mg/mL and 0.21 ± 0.02 mg/mL for LD and HD density EV suspensions respectively. Fluorescence labeling of liposomes and EVs with PKH26. Lipid-based objects present in liposome and EV suspensions were fluorescently labeled with the self-inserting membrane dye PKH26 (MINI26, Sigma-Aldrich). A volume of 1 or 2 µL of sample suspension was diluted in diluent C (Sigma-Aldrich) to final volume of 50 µL and mixed with a PKH26 solution composed of 1µL of PKH26 (diluted in ethanol to a 100 µM concentration) and 49 µL of diluent C. During the labeling procedure, the suspensions were composed of 1 µM of PKH26 and 100 µg/mL lipid mass for the liposome suspension and 6.5 µg/mL and 2.1 µg/mL total protein content for LD and HD suspensions respectively. The mixtures were incubated for 5 minutes at 4°C in the dark. Size exclusion columns MicroSpin S-200 HR (GE Healthcare) were employed to remove unbound PKH26 fluorophores from the labeled suspensions. Prior to separation, the TRIS-EDTA buffer present in the column was exchanged with PBS buffer following the manufacturer’s recommendations i.e. by spinning 3 times 300 µL of PBS through the column at 740 x g for 2 minutes. 100 µL of labeling solution were then loaded in a column and spun at 740 x g for 2 minutes at 4°C. The filtered PKH26 labeled vesicles were collected with an average recovery volume of 99%. The impact of the column filtration step on the sample suspensions was assessed with NTA in scattering and fluorescence mode resulting into a slight change in size distribution and an average loss of 50 % of the particles (data not shown).


Page 21 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

AFM characterization of the calibration beads. Sulfate polystyrene beads (S37202, Molecular Probes, ThermoFisher Scientist) were used to compensate for the differences in illumination conditions between different waveguide chips. Their size distribution was measured using AFM imaging in Tapping Mode (Bruker Dimension Icon AFM, Bruker Corporation, Billerica, MA) after electrostatically adsorbing the beads on a silicon surface as described in the following. The silicon surfaces were first treated with oxygen plasma to induce negative charges on the surface. Then a layer of aluminium chloride hydroxide positive polyions was adsorbed on the surface from 2 w% aqueous solution for 2 minutes to guarantee the electrostatic attraction of negatively charged sulfate polystyrene beads to the surface. Beads were adsorbed from a 0.2 w% solution of calibration beads in Milli-Q water (Millipore, France) on the surface for at least 1 min, followed by rinsing with Milli-Q water for another 30 seconds and drying using pressurized nitrogen. AFM images were recorded in air using standard tapping mode cantilevers (NSC37/AlBS, Mikromasch, Sofia, Bulgaria) and analyzed using a home-written MatLab script (MathWorks Natick, MA) that locates the highest point of each recorded bead as described previously[68]. The so-determined size distribution of calibration beads followed a Gaussian distribution with an average bead diameter of 63 nm and a full width at half maximum of 12 nm. Nanoparticle Tracking Analysis. The size distributions of liposomes and EV suspensions were measured with Nanoparticle Tracking Analysis using a NanoSight LM10 module (Malvern, NanoSight Ltd., Amesbury, United Kingdom) equipped with a 532 nm laser. The camera and analysis settings were optimized to enable both size distribution and concentration determination, according to the manufacturer recommendations. All measurements were performed at room temperature and the samples were measured under flow using a Nanosight syringe pump module. The buffer viscosity was considered to be that of water. Each sample was recorded under flow


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 31

with collection of 5 movies of 1 minute each. Size distribution and total particle concentration were measured in scattering and fluorescence mode at 0.50 µg/mL total lipid concentration for liposomes and 0.65 µg/mL and 0.21 µg/mL total protein concentration for low and high density EV suspensions respectively. Fabrication of waveguide chips. A single-mode planar waveguide structure was fabricated on a Si(100) 4” wafer at the MC2 nanofabrication laboratory, Chalmers, Sweden (Fed. Std.209 E 10100). The three-layered waveguide consists of a 500 nm thick core layer made from spin-on-glass (IC1-200 from Futurrex Inc.), with a refractive index n=1.42 at 532 nm wavelength, which is embedded in a 7 micron thick cladding layer made of a fluorinated polymer CYTOP (CTX809AP2 from AGC Chemicals, ASAHI Glass Co., LTD.) with a refractive index of n=1.34 at 532 nm wavelength. Sample-wells were created by etching 2x2 mm2 areas through the top cladding using reactive ion etching. Finally, the wafer was diced into 1x1 cm2 chips, each containing a single sample well in the center. The chips were then stored in a cupboard until needed. Prior to an experiment, the chips were treated with low power oxygen plasma for 15 minutes, followed by a gently rinsing using MilliQ water (Millipore, France) and blow drying using nitrogen. The sample well was then exposed to 2 M sulfuric acid for 30 minutes followed by thorough rinsing with Milli-Q water. Finally, the remaining water was exchanged for 20 µL PBS and the waveguide chip was placed under the microscope equipped with standard x100, NA 0 1.0 water immersion objective. Transverse-electric polarized 532 nm light emanating from an optical fiber was coupled to the waveguide core layer by aligning the fiber to the facet of the waveguide chip, generating an evanescent electromagnetic field illumination on the surface in the sample well. The characteristic penetration depth (dp) of this evanescent light intensity, defined as I(dp) = I(z=0)/e, was approximately 130 nm.


Page 23 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Single particle measurements using waveguide microscopy. A typical experiment was performed as follows: polystyrene beads, unlabeled EVs and PKH26-labeled EVs were injected sequentially and visualized in scattering mode, or scattering and fluorescence modes for the PKH26-labelled objects (see Fig. 1b). To use the calibration beads as a reference for the signals generated by both unlabeled EVs and PKH26-labeled EVs, the scattering intensities of individual polystyrene beads, unlabeled EVs and PKH26-labeled EVs were extracted from the last frame after injection of all samples, with the position of the objects identified after each injection step used to determine the nature of the particles (bead or labeled/unlabeled EV) visualized in the last frame (Fig. 1b). In this way, the polystyrene beads (diameter ~63 nm, refractive index ~1.59) could be used as an internal standard to calibrate the measured scattering intensities both within a single experiment and between different chips. All sample suspensions were injected as a 20 µL volume into the sample well containing 20 µL of PBS. Unspecific adsorption of the particles to silica surface was followed in scattering mode until adequate surface coverage was obtained. The solution in the sample well was then exchanged with PBS to stop binding (average adsorption time: 1 minute). Each waveguide chip surface was dedicated for the observation of a subset of EVs i.e. low or high density EVs. Particles were adsorbed on the waveguide chip in sequential injection steps: i) polystyrene calibration beads observed in scattering mode, ii) unlabeled EVs particles observed in scattering mode and iii) PKH26 fluorescently labeled EVs particles observed first in fluorescence and then in scattering mode. Bindings of calibration beads and unlabeled EVs were performed under illumination with a green laser (wavelength: 532 nm) at 1 mW power while the binding of PKH26 fluorescently labeled EVs (~1 minute) was performed under illumination with a blue laser (wavelength: 488 nm) at 0.5 mW power to prevent photobleaching of the PKH26 fluorophores before recording. This was independently verified by monitoring over time the


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 31

fluorescence intensity of single PKH26 fluorescently labeled EVs, which was found to decrease by less than 5% after 10 minutes of illumination with blue laser at 0.5 mW. After binding of the later sample, the blue laser was switched off, the optical fiber was uncoupled from the waveguide chip and the green laser was set to 1 mW for 4 minutes to stabilize its output power before recoupling the optical fiber to the waveguide chip for fluorescence and scattering read-out. All images were captured with 500 ms exposure time under illumination with a green laser (wavelength: 532 nm) at 1 mW power. Scattering and fluorescence intensities were recorded using band-pass 515-555 nm and band-pass 565-605 nm filters for scattering and fluorescence read-out, respectively. The scattering intensity distribution of calibration beads, unlabeled EVs and PKH26 labeled EVs were extracted from the same frame recorded after binding all three samples. Calibration beads were diluted to 8×10-6 w/v% in Milli-Q water prior to injection. Prior to injection, unlabeled EVs were first diluted in PBS to 6.5 µg/mL and 2.1 µg/mL total protein content for low and high density EV suspensions respectively and filtrated using size exclusion MicroSpin S-200 HR (GE Healthcare) column at 740 x g force for 2 minutes at 4 degrees. PKH26 labeling of EVs was performed as indicated in previous section. Filtrated unlabeled and labeled EVs suspensions were added directly from the column filtrate with a volume of 20 µL into the sample well. Waveguide microscopy image analysis. The fluorescent and/or scattered light from adsorbed particles was collected using a NA=1.0 water immersion objective and projected onto a 2048×2048 pixels CMOS sensor (Hamamatsu ORCA-Flash4.0 V2 Scientific CMOS camera). Prior to adsorbing particles to the waveguide’s surface, a reference image of the background was acquired. From that image the median background intensity value was evaluated from all image pixels. After particle adsorption and subsequent rinsing, the surface was again imaged and particles were then


Page 25 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

detected considering that any area consisting of at least 8 connected pixels passing a certain threshold criteria represents an adsorbed particle core location. This threshold could be anything between 20 and 300% higher than the background for the scattering images and 6 to 10% for the fluorescence one. After applying the threshold, each particle area was unilaterally dilated by a few pixels to ensure total engulfment of the detected particle. The scattering and fluorescence intensities of the bound particles were then extracted by first subtracting each pixel intensity by its local background intensity, determined as the average intensity of at least 300 neighboring background pixels surrounding of the particle area, and then integrating the intensities of all pixels present within the area for a given particle. The difference in surface illumination between chips was normalized using as reference the scattering intensity distribution from monodispersed polystyrene calibration beads that were adsorbed to the waveguide’s surface prior to EV each measurement. The correction factor required to overlay the peak of the fitted scattering intensity distribution of the beads bound to different chips was applied to all particles detected within the same frame. ACKNOWLEDGMENT The authors would like to thank Anders Lundgren for fruitful discussions and expertise, Tuan Phan Xuan together with Aleksandar Matic for their support in initial size characterization of EV suspensions and Agnieszka Siupa for her support and expertise with NTA measurements.This project was funded by the Swedish Research Council (VR), the Swedish Governmental Agency for Innovations Systems (VINNOVA) and the Göran Gustafsson foundation. REFERENCES [1]

M. Colombo, G. Raposo, and C. Thery. Biogenesis, secretion, and intercellular interactions of exosomes and other extracellular vesicles, Annu Rev Cell Dev Biol, 2014, 30, 255-89.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

Page 26 of 31

C. Lasser, V. S. Alikhani, K. Ekstrom, M. Eldh, P. T. Paredes, A. Bossios, et al., Human saliva, plasma and breast milk exosomes contain RNA: uptake by macrophages, J Transl Med, 2011, 9, 9. S. Kourembanas, Exosomes: vehicles of intercellular signaling, biomarkers, and vectors of cell therapy Annu Rev Physiol, 2015, 77, 13-27. A. V. Vlassov, S. Magdaleno, R. Setterquist, and R. Conrad, Exosomes: current knowledge of their composition, biological functions, and diagnostic and therapeutic potentials, Biochim Biophys Acta, 2012, 1820, 940-8. R. M. Johnstone, M. Adam, J. R. Hammond, L. Orr, and C. Turbide, Vesicle formation during reticulocyte maturation. Association of plasma membrane activities with released vesicles (exosomes) J Biol Chem, 1987, 262, 9412-20. M. Yanez-Mo, P. R. Siljander, Z. Andreu, A. B. Zavec, F. E. Borras, E. I. Buzas, et al., Biological properties of extracellular vesicles and their physiological functions, J Extracell Vesicles, 2015, 4, 27066. N. Chaput and C. Thery, Exosomes: immune properties and potential clinical implementations Semin Immunopathol, 2011, 33, 419-40. C. Admyre, E. Telemo, N. Almqvist, J. Lotvall, R. Lahesmaa, A. Scheynius, et al., Exosomes - nanovesicles with possible roles in allergic inflammation Allergy, 2008, 63, 404-8. D. Skokos, S. Le Panse, I. Villa, J. C. Rousselle, R. Peronet, A. Namane, et al., "Nonspecific B and T cell-stimulatory activity mediated by mast cells is associated with exosomes," Int Arch Allergy Immunol, 2001, 124, 133-6. A. G. Ibrahim, K. Cheng, and E. Marban, Exosomes as critical agents of cardiac regeneration triggered by cell therapy, Stem Cell Reports, 2014, 2, 606-19. C. Corrado, S. Raimondo, A. Chiesi, F. Ciccia, G. De Leo, and R. Alessandro, Exosomes as Intercellular Signaling Organelles Involved in Health and Disease: Basic Science and Clinical Applications, International Journal of Molecular Sciences, 2013, 14, 5338-5366. E. I. Buzas, B. Gyorgy, G. Nagy, A. Falus, and S. Gay, "Emerging role of extracellular vesicles in inflammatory diseases," Nature Reviews Rheumatology, 2014, 10, 356-364. P. Vader, X. O. Breakefield, and M. J. A. Wood, "Extracellular vesicles: emerging targets for cancer therapy," Trends in Molecular Medicine, 2014, 20, 385-393. C. Lasser, Exosomes in diagnostic and therapeutic applications: biomarker, vaccine and RNA interference delivery vehicle Expert Opin Biol Ther, 2015, 15, 103-17. S. A. Melo, L. B. Luecke, C. Kahlert, A. F. Fernandez, S. T. Gammon, J. Kaye, et al., Glypican-1 identifies cancer exosomes and detects early pancreatic cancer, Nature, 2015, 523, 177-82. K. B. Johnsen, J. M. Gudbergsson, M. N. Skov, L. Pilgaard, T. Moos, and M. Duroux, A comprehensive overview of exosomes as drug delivery vehicles - endogenous nanocarriers for targeted cancer therapy, Biochim Biophys Acta, 2014, 1846, 75-87. A. Tan, J. Rajadas, and A. M. Seifalian, Exosomes as nano-theranostic delivery platforms for gene therapy Adv Drug Deliv Rev, 2013, 65, 357-67. C. Thery, S. Amigorena, G. Raposo, and A. Clayton, Isolation and characterization of exosomes from cell culture supernatants and biological fluids Curr Protoc Cell Biol, 2006, 3, 3. E. Cocucci and J. Meldolesi, Ectosomes and exosomes: shedding the confusion between extracellular vesicles, Trends Cell Biol, 2015, 25, 364-72.


Page 27 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31]

[32] [33] [34] [35] [36]

Y. Yin, X. Huang, K. D. Lynn, and P. E. Thorpe, Phosphatidylserine-targeting antibody induces M1 macrophage polarization and promotes myeloid-derived suppressor cell differentiation Cancer Immunol Res 2013, 1, 256-68. L. Ayers, R. Nieuwland, M. Kohler, N. Kraenkel, B. Ferry, and P. Leeson. Dynamic microvesicle release and clearance within the cardiovascular system: triggers and mechanisms Clin Sci (Lond), 2015, 129, 915-31. S. J. Gould and G. Raposo, As we wait: coping with an imperfect nomenclature for extracellular vesicles, J Extracell Vesicles, 2013, 2, 20389 R. Crescitelli, C. Lasser, T. G. Szabo, A. Kittel, M. Eldh, I. Dianzani, et al., Distinct RNA profiles in subpopulations of extracellular vesicles: apoptotic bodies, microvesicles and exosomes J Extracell Vesicles, 2013, 2, 20677. H. Valadi, K. Ekstrom, A. Bossios, M. Sjostrand, J. J. Lee, and J. O. Lotvall, Exosomemediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells Nat Cell Biol, 2007, 9, 654-9. B. K. Thakur, H. Zhang, A. Becker, I. Matei, Y. Huang, B. Costa-Silva, et al., Doublestranded DNA in exosomes: a novel biomarker in cancer detection Cell Res, 2014, 24, 7669. Y. Lee, S. El Andaloussi, and M. J. Wood, Exosomes and microvesicles: extracellular vesicles for genetic information transfer and gene therapy Hum Mol Genet, 2012, 21, R12534. V. R. Minciacchi, S. You, C. Spinelli, S. Morley, M. Zandian, P. J. Aspuria, et al., Large oncosomes contain distinct protein cargo and represent a separate functional class of tumor-derived extracellular vesicles Oncotarget, 2015, 6, 11327-41. D. S. Choi, D. K. Kim, Y. K. Kim, and Y. S. Gho, "Proteomics, transcriptomics and lipidomics of exosomes and ectosomes," Proteomics, 2013, 13, 1554-71. R. J. Simpson, S. S. Jensen, and J. W. Lim, Proteomic profiling of exosomes: current perspectives Proteomics, 2008, 8, 4083-99. W. H. James, Sex ratio, testosterone and postpartum thyroid dysfunction Br J Obstet Gynaecol, 1992, 99, 698-9. J. Lötvall, A. F. Hill, F. Hochberg, E. I. Buzás, D. Di Vizio, C. Gardiner, et al., Minimal experimental requirements for definition of extracellular vesicles and their functions: a position statement from the International Society for Extracellular Vesicles, J Extracell Vesicles 2014, 3, 26913. H. G. Lamparski, A. Metha-Damani, J. Y. Yao, S. Patel, D. H. Hsu, C. Ruegg, et al., Production and characterization of clinical grade exosomes derived from dendritic cells J Immunol Methods, 2002, 270, 211-26. C. S. Hong, L. Muller, T. L. Whiteside, and M. Boyiadzis, Plasma exosomes as markers of therapeutic response in patients with acute myeloid leukemia Front Immunol, 2014, 5, 160. H. Rabesandratana, J. P. Toutant, H. Reggio, and M. Vidal, "Decay-accelerating factor (CD55) and membrane inhibitor of reactive lysis (CD59) are released within exosomes during In vitro maturation of reticulocytes," Blood, 1998, 91, 2573-80. M. Logozzi, A. De Milito, L. Lugini, M. Borghi, L. Calabro, M. Spada, et al., High Levels of Exosomes Expressing CD63 and Caveolin-1 in Plasma of Melanoma Patients Plos One, 2009, 4, e5219. C. Lasser, Exosomal RNA as biomarkers and the therapeutic potential of exosome vectors, Expert Opin Biol Ther, 2012, 12, S189-97.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[37] [38] [39]

[40] [41] [42] [43] [44] [45]

[46] [47] [48] [49] [50] [51]

Page 28 of 31

N. Arraud, R. Linares, S. Tan, C. Gounou, J. M. Pasquet, S. Mornet, et al., Extracellular vesicles from blood plasma: determination of their morphology, size, phenotype and concentration Journal of Thrombosis and Haemostasis, 2014, 12, 614-627. R. Wyss, L. Grasso, C. Wolf, W. Grosse, D. Demurtas, and H. Vogel, Molecular and Dimensional Profiling of Highly Purified Extracellular Vesicles by Fluorescence Fluctuation Spectroscopy Analytical Chemistry, 2014, 86, 7229-7233. D. W. Greening, R. Xu, H. Ji, B. J. Tauro, and R. J. Simpson, "A protocol for exosome isolation and characterization: evaluation of ultracentrifugation, density-gradient separation, and immunoaffinity capture methods," Methods Mol Biol, 2015, 1295, 179209. V. Palmieri, D. Lucchetti, I. Gatto, A. Maiorana, M. Marcantoni, G. Maulucci, et al., Dynamic light scattering for the characterization and counting of extracellular vesicles: a powerful noninvasive tool Journal of Nanoparticle Research 2014, 16, 2583. B. J. Berne and R. Pecora, Dynamic light scattering : with applications to chemistry, biology, and physics. Mineola, N.Y. 2000, Dover Publications. C. Gardiner, Y. J. Ferreira, R. A. Dragovic, C. W. Redman, and I. L. Sargent, "Extracellular vesicle sizing and enumeration by nanoparticle tracking analysis," J Extracell Vesicles, 2013, 2. E. van der Pol, F. A. W. Coumans, A. Sturk, R. Nieuwland, and T. G. van Leeuwen, Refractive Index Determination of Nanoparticles in Suspension Using Nanoparticle Tracking Analysis Nano Letters, 2014, 14, 6195-6201. V. S. Chernyshev, R. Rachamadugu, Y. H. Tseng, D. M. Belnap, Y. L. Jia, K. J. Branch, et al., Size and shape characterization of hydrated and desiccated exosomes Analytical and Bioanalytical Chemistry, 2015, 407, 3285-3301. V. Sokolova, A. K. Ludwig, S. Hornung, O. Rotan, P. A. Horn, M. Epple, et al., Characterisation of exosomes derived from human cells by nanoparticle tracking analysis and scanning electron microscopy Colloids and Surfaces B-Biointerfaces 2011, 87, 146150. S. Sharma, B. M. Gillespie, V. Palanisamy, and J. K. Gimzewski, Quantitative Nanostructural and Single-Molecule Force Spectroscopy Biomolecular Analysis of Human-Saliva-Derived Exosomes, Langmuir, 2011, 27, 14394-14400. Y. Yuana, T. H. Oosterkamp, S. Bahatyrova, B. Ashcroft, P. G. Rodriguez, R. M. Bertina, et al., Atomic force microscopy: a novel approach to the detection of nanosized blood microparticles Journal of Thrombosis and Haemostasis, 2010, 8, 315-323. L. Grasso, R. Wyss, L. Weidenauer, A. Thampi, D. Demurtas, M. Prudent, et al., Molecular screening of cancer-derived exosomes by surface plasmon resonance spectroscopy Analytical and Bioanalytical Chemistry, 2015, 407, 5425-5432. H. Im, H. L. Shao, Y. I. Park, V. M. Peterson, C. M. Castro, R. Weissleder, et al., Labelfree detection and molecular profiling of exosomes with a nano-plasmonic sensor Nature Biotechnology, 2014, 32, 490. D. L. M. Rupert, C. Lasser, M. Eldh, S. Block, V. P. Zhdanov, J. O. Lotvall, et al., Determination of Exosome Concentration in Solution Using Surface Plasmon Resonance Spectroscopy Analytical Chemistry 2014, 86, 5929-5936. I. Reviakine, F. F. Rossetti, A. N. Morozov, and M. Textor, Investigating the properties of supported vesicular layers on titanium dioxide by quartz crystal microbalance with dissipation measurements Journal of Chemical Physics, 2005, 122, 204711.


Page 29 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67]

A. H. Kunding, M. W. Mortensen, S. M. Christensen, and D. Stamou, "A fluorescencebased technique to construct size distributions from single-object measurements: Application to the extrusion of lipid vesicles," Biophysical Journal, 2008, 95, 1176-1188. R. A. Dragovic, C. Gardiner, A. S. Brooks, D. S. Tannetta, D. J. P. Ferguson, P. Hole, et al., Sizing and phenotyping of cellular vesicles using Nanoparticle Tracking Analysis, Nanomedicine-Nanotechnology Biology and Medicine, 2011, 7, 780-788. S. S. Kanwar, C. J. Dunlay, D. M. Simeone, and S. Nagrath, Microfluidic device (ExoChip) for on-chip isolation, quantification and characterization of circulating exosomes Lab Chip, 2014, 14, 1891-900. E. Marx and G. W. Mulholland, Size and refractive-index determination of single polystyrene spheres Journal of Research of the National Bureau of Standards, 1983, 88, 321-338. B. Agnarsson, A. Lundgren, A. Gunnarsson, M. Rabe, A. Kunze, M. Mapar, et al., Evanescent Light Scattering Microscopy for Label-Free Interfacial Imaging: From Single Sub-100 nm Vesicles to Live Cells ACS Nano, 2016, 10, 9974-9982. B. Agnarsson, S. Ingthorsson, T. Gudjonsson, and K. Leosson, Evanescent-wave fluorescence microscopy using symmetric planar waveguides Optics Express, 2009, 17, 5075-5082. X. Y. Ma, J. Q. Lu, R. S. Brock, K. M. Jacobs, P. Yang, and X. H. Hu, Determination of complex refractive index of polystyrene microspheres from 370 to 1610 nm Physics in Medicine and Biology, 2003, 48, 4165-4172. H. Fischer, I. Polikarpov, and A. F. Craievich, Average protein density is a molecularweight-dependent function Protein Science, 2004, 13, 2825-2828. P. K. Horan, M. J. Melnicoff, B. D. Jensen, and S. E. Slezak, Fluorescent cell labeling for in vivo and in vitro cell tracking Methods Cell Biol, 1990, 33, 469-90. A. Medina and A. Ghahary, Transdifferentiated circulating monocytes release exosomes containing 14-3-3 proteins with matrix metalloproteinase-1 stimulating effect for dermal fibroblasts Wound Repair Regen, 2010, 18, 245-53. E. J. van der Vlist, E. N. Nolte-'t Hoen, W. Stoorvogel, G. J. Arkesteijn, and M. H. Wauben, Fluorescent labeling of nano-sized vesicles released by cells and subsequent quantitative and qualitative analysis by high-resolution flow cytometry Nat Protoc, 2012, 7, 1311-26. U. K. Chettiar and N. Engheta, Internal homogenization: effective permittivity of a coated sphere. Opt. Express, 2012, 20, 22976–22986. D. L. M. Rupert, G. V. Shelke, G. Emilsson, V. Claudio, S. Block, A. Dahlin, et al. DualWavelength Surface Plasmon Resonance for Determining the Size and Concentration of Sub-Populations of Extracellular Vesicles. Analytical Chemistry 2016, 88, 9980-9988. R. C. Macdonald, R. I. Macdonald, B. P. M. Menco, K. Takeshita, N. K. Subbarao, and L. R. Hu, Small-Volume Extrusion Apparatus for Preparation of Large, Unilamellar Vesicles, Biochimica Et Biophysica Acta, 1991, 1061, 297-303. G. V. Shelke, C. Lasser, Y. S. Gho, and J. Lotvall, Importance of exosome depletion protocols to eliminate functional and RNA-containing extracellular vesicles from fetal bovine serum J Extracell Vesicles 2014, 3, 24783. Lasser, C.; Shelke, G. V.; Yeri, A.; Kim, D. K.; Crescitelli, R.; Raimondo, S.; Sjostrand, M.; Gho, Y. S.; Jensen, K. V.; Lotvall, J., Two distinct extracellular RNA signatures released by a single cell type identified by microarray and next-generation sequencing. Rna Biology 2017, 14, 58-72.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[68]

Page 30 of 31

S. Block, G. Glöckl, W. Weitschies, and C. A. Helm, Direct Visualization and Identification of Biofunctionalized Nanoparticles using a Magnetic Atomic Force Microscope Nano Letters 2011, 11, 3587-3592.

ASSOCIATED CONTENT Supporting Information. AUTHOR INFORMATION Corresponding Author *e-mail: [email protected] Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest.


Page 31 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

TOC Graphics


Effective Refractive Index and Lipid Content of Extracellular Vesicles

Recommend Documents