- Email: [email protected]
Membrane proteins significantly restrict exosome mobility
Biochemical and Biophysical Research Communications xxx (2018) 1e5
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Membrane proteins significantly restrict exosome mobility Mikhail Skliar a, b, *, Vasiliy S. Chernyshev c, d, David M. Belnap e, German V. Sergey d, Samer M. Al-Hakami a, Philip S. Bernard f, g, Inge J. Stijleman f, Rakesh Rachamadugu f a
Chemical Engineering, University of Utah, 50 S. Central Campus Dr, Salt Lake City, UT, 84112, USA The Nano Institute of Utah, University of Utah, 36 S. Wasatch Dr, Salt Lake City, UT, 84112, USA c Center for Translational Biomedicine, Skolkovo Institute of Science and Technology, Skolkovo Innovation Center, Building 3, Moscow, 143026, Russia d Biopharmaceutical Cluster ‘Northern’, Moscow Institute of Physics and Technology, Institutsky per. 9/7, Dolgoprudny, Moscow Region, 141700, Russia e Biochemistry and Biology Departments, University of Utah, 15 N Medical Dr, Salt Lake City, UT, 84112, USA f Huntsman Cancer Institute, University of Utah, 2000 Circle of Hope, Salt Lake City, UT, 84112, USA g Department of Pathology, University of Utah, 15 North Medical Dr, Salt Lake City, UT, 84112, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 May 2018 Accepted 15 May 2018 Available online xxx
Exosomes are membrane nanovesicles implicated in cell-to-cell signaling in which they transfer their molecular cargo from the parent to the recipient cells. This role essentially depends on the exosomes' small size, which is the prerequisite for their rapid migration through the crowded extracellular matrix and into and out of circulation. Here we report much lower exosome mobility than expected from the size of their vesicles, implicate membrane proteins in a substantially impeded rate of migration, and suggest an approach to quantifying the impact. The broadly distributed excess hydrodynamic resistance provided by surface proteins produces a highly heterogeneous and microenvironment-dependent hindrance to exosome mobility. The implications of the findings on exosome-mediated signaling are discussed. © 2018 Elsevier Inc. All rights reserved.
Keywords: Exosomes and extracellular vesicles Membrane proteins Proteolysis Vesicle trafficking
Exosomes are actively secreted by cells and found in all biological fluids, including blood, urine, and saliva. Compared to other extracellular vesicles (EVs), exosomes are distinguished by biomarkers of the late endocytic biogenesis and the smallest size, typically reported to be 30e150 nm in diameter. The uptake of an exosome by a local or distant cell transfers the molecular cargo derived from a secreting parent to a recipient cell. Biologically active molecules transferred in health and disease by this mechanism include surface and luminal proteins, membrane-bound microRNAs, and other compounds. A growing number of studies implicate exosomal signaling in tumor metastasis [1], drug resistance [2], and the modulation of immune response [3]. For signaling to occur it is necessary, but not sufficient, for a secreted EV to migrate from a parent to a recipient cell. In paracrine signaling, the transport barrier is overcome by diffusion through extracellular space, while to reach distal targets an EV may need to enter and exit a biofluid circulation. The smallest among EVs, the exosomes have the highest mobility which favors overcoming
* Corresponding author. Department of Chemical Engineering, University of Utah, 50 S. Central Campus Dr, Salt Lake City, UT, 84112, USA. E-mail address: [email protected] (M. Skliar).
transport resistance for exosome-mediated signaling to take place [4]. The mobility of the exosomes may be quantified by their mean squared displacements (MSD) with time, which proportionally depends on their (self-) diffusivity. The oft-reported hydrodynamic diameter of exosomes is the measure of the resistance to the migration and is inversely proportional to their diffusivity. Transport barriers to exosome dissemination are poorly understood. Here, we examine the impact of surface proteins and macromolecules anchored by them on the mobility of exosomes isolated by dissimilar methods (precipitation and size exclusion) from two biological fluids (growth medium of MCF-7 breast cancer cells and human sera). The excess resistance to the migration imposed by surface proteins is quantified as the thickness of a coronal layer formed by surface-conjugated macromolecules, which we measure as a difference between hydrodynamic and vesicle diameters of the exosomes. The exosomes released by MCF-7 cells were isolated by precipitation and their enrichment in the isolate was confirmed by positive expression of surface and luminal biomarkers, the globular shape of vesicles with bilayer membrane morphology, and the hydrodynamic diameters in the exosomal range (Fig. 1a and S1-S2 in Supporting Information, SI). The mean and mode hydrodynamic
https://doi.org/10.1016/j.bbrc.2018.05.107 0006-291X/© 2018 Elsevier Inc. All rights reserved.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
2
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
Fig. 1. (a) The size of exosome vesicles in the solution (red curve, grey fill) was obtained from the AFM measurements and confirmed by cryo-TEM imaging. Cryo-TEM shows the expected globular shape and double-layer morphology of vesicles (inset). The vesicles are substantially smaller than their hydrodynamic diameters measured by NTA (blue curve, blue fill). The average thickness of the exosomal corona (inset summary) is estimated as the difference between the average hydrodynamic and vesicle sizes. (b) Trypsinization of surface proteins substantially reduces hydrodynamic diameters. (c) The shift in mobility sizes is even more pronounced after a less specific protease K digestion. PK proteolysis narrows the distribution of hydrodynamic sizes, pointing to the heterogeneity in the macromolecular surface decoration as the source of widely different mobilities of untreated exosomes. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
diameters measured by nanoparticle tracking analysis (NTA) were 111 ± 12 nm and 80 ± 9 nm, respectively. The volume encapsulated by hydrated vesicles was measured by atomic force microscopy (AFM; Fig. S5-S6 and Table S4, SI). The diameter of hydrated vesicles in their innate globular shape was estimated by mapping the AFM volume measurements, obtained for exosomes electrostatically immobilized on a surface, into volume-equivalent spheres in the solution. Fig. 1a shows the obtained probability density function of vesicle sizes, which has 30.1 ± 6.9 nm ensemble average. A much smaller size of vesicles compared with their hydrodynamic diameters was independently confirmed by cryo-TEM imaging which showed MCF-7 vesicles with an average diameter equal to 34.2 nm (Fig. S7a, SI). The macromolecular corona is indistinguishable in cryo-TEM images because of its low excess density [5,6]. The NTA measurements of hydrodynamic sizes were repeated after enzymatic proteolysis of MCF-7 exosomes. The four digestion protocols (Table S5, SI) differed in enzymes used and the treatment duration. A statistically significant (p < 0.01) increase in exosome mobility after the proteolysis by as much as 50% was observed (Fig. 1b and c; Fig. S3 and Table S2, SI). The duration of enzymatic treatments had little effect on hydrodynamic sizes, indicating a sufficiently long incubation to complete the digestion. The following factors were ruled out from contributing to the observed reduction in hydrodynamic sizes: a) Enzymatic treatments did not change the size of vesicles which remained in the same range before and after digestion (Fig. S8, SI and b) The protein concentration in the solution was too low (less than 3.5 mg/mL) to cause changes in the viscosity after the proteolysis [7]. The mobility of exosomes after proteolysis was enzyme dependent (Fig. 1b and c). Trypsin preferentially cleaves at lysine and arginine locations but does not affect surface-anchored segments of membrane proteins void of these a-amino acids. Trypsinization thinned the coronal layer, reducing the hydrodynamic size of exosomes to having the mode diameter equal to 64 nm, down from the original 80 nm (Fig. 1b and Table S2, SI). The reduction was even more significant (down to the 56 nm mode diameter; Fig. 1c) when we used a less specific proteinase K (PK), known to cleave peptide bonds of hydrophobic, aliphatic and aromatic amino acids. Compared to trypsinization, the broader PK activity leaves shorter undigested fragments of surface proteins and, thus, a thinner coronal layer, as reflected by a higher mobility of PK-treated exosomes. The indiscriminate PK proteolysis shifts the hydrodynamic diameters of the exosomes into the range of sizes partially overlapping with vesicle sizes. The remaining difference may be
explained by short fragments of surface proteins that survived the digestion and the presence of pericellular coats formed by glycosaminoglycans, and hyaluronan specifically [8], which are unaffected by trypsin and PK. The described mechanism of enzyme-dependent thinning of exosome corona is summarized in Fig. 2, where we compare the distributions of the vesicle and hydrodynamic diameters before and after the proteolysis. After the proteolysis, the hydrodynamic sizes are distributed more uniformly. The less specific PK treatment, which closely “shaves” the membrane surface, leads to the especially narrow range of hydrodynamic diameters with a closer resemblance to the distribution of vesicle sizes. The broadly distributed hydrodynamic diameters of vesicles with narrowly distributed membrane envelopes reveal the heterogeneity in the macromolecular surface decoration of exosomes and the widely varying thickness of their coronal layer from one exosome to another. The direct visualization of diverse migration rates of MCF-7 exosomes was obtained by tracking the light scattered by a single exosome as it moves in PBS. The three trajectories in Fig. 3a were
Fig. 2. The reduction in the coronal layers is enzyme dependent. Non-specific protein digestion by proteinase K shifts the hydrodynamic diameters closer to the range of vesicle sizes.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
3
Fig. 3. (a) Single-particle tracking of migrating exosomes having the mode, 10th, and 90th percentile hydrodynamic diameters. (b) The squared particle displacements corresponding to the tracks in (a) and the theoretical predictions of the MSD extended beyond the observation period show the long-term trend in size-dependent reach of migrating exosomes. (c) The extracellular matrix exaggerates the size dependence of exosome mobility. The exosomes become immobilized as their hydrodynamic diameter approaches the size of ECM pores.
tracked for ~58 s and show the Brownian displacement of exosomes with 88.5, 70.5, and 142.5 nm hydrodynamic diameters, which correspond to the mode, 10th, and 90th percentiles values of the hydrodynamic size distribution before proteolysis. The squared displacements along these trajectories and the theoretical predictions of the MSD beyond the observation period illustrate strong size-dependence of the exosomal reach and a higher probability of smaller exosomes overcoming the transport barrier to reach a target. Little is known about the migration of exosomes through the extracellular space (ECS). An early insight into the size-dependence to the hindered transit of exosomes through the ECS and the extracellular matrix (ECM) that occupies it may be gained from modeling. Following the approach used to describe the ECM migration of molecules and nanoparticles [9], we related the effective diffusivity of the exosomes in the ECS, De , to their unimpeded diffusivity in the solution, D0 , as D0 ¼ ðtg tv Þ2 De , where the geometric tortuosity, tg , describes the reduction in the migration rate due to a longer path an exosome must traverse through a meandering extracellular pore. We used tg ¼ 1:67 which provides the best fit to the experimental data on the tortuosity in the cortex of rats [10]. The “viscous” tortuosity tv describes the contribution of steric hindrance to the exosome migration, their specific and nonspecific interactions with ECM, and all other factors impeding the movement of the exosomes through the ECS. Its value depends on the hydrodynamic diameter of the exosomes relative to the size of an ECS pore through which the diffusion occurs, and was expressed using a steric partition coefficient and the enhanced pore-to-solution friction ratio [11]. Fig. 3c compares the ECM diffusivity in a 150 nm ECS pore (equal to mean confining ECS dimension inside a brain of a living rat [12]) with unimpeded diffusivity in the solution. Average distance traveled in the ECS by an exosome of a small hydrodynamic size (in the range of membrane vesicles) is less than one fifth the distance translocated in the solution, as indicated by the ratio of the MSDs during hindered and free diffusion (Fig. 3c). The distance drops to ~1% for 80 nm particles and becomes negligible as the hydrodynamic size of the exosomes approaches the pore size. The exaggerated influence of sizes on the ECS mobility increases the dwell time of hydrodynamically larger exosomes in the proximity of a parent cell and favors their sequestration. The routine trapping of EVs in urinary bladder matrix and other ECM scaffolds was
reported recently [13] without concluding the process is size dependent. The developed model predicts cessation in mobility for exosomes with hydrodynamic sizes smaller than pores. According to this estimate, approximately half of analyzed MCF-7 exosomes (a subpopulation with hydrodynamic diameters above 100 nm) would be unable to transit through the ECM with 150 nm maximum pore size. This prediction is conservative and does not account for conformational flexing of macromolecular corona which may overcome steric hindrance, just as flexible molecules were shown to translate through the ECM while the diffusion of rigid molecules of similar molecular weight was restricted [14]. An alternative estimate that uses vesicle sizes to predict the mobility cut-off is less restrictive and predicts that all MCF-7 exosomes should be able to migrate in the ECM with 150 nm pores, though with an increased hindrance compared to the unimpeded diffusion. Exosome-microenvironment interactions influence the mobility cut-off, shifting it towards one of the described theoretical bounds defined by either hydrodynamic or vesicle sizes of exosomes. For example, the reduction of the coronal layer in vivo by extracellular enzymes, such as metalloproteinases, will increase the mobility towards a more optimistic prediction. Interestingly, the proteolytic activity may be bidirectional, with tumor exosomes reportedly degrading the extracellular matrix [15], thus reducing the hindrance to their ECM mobility. A shift towards lower mobility will occur with the increase in the coronal layer as a result of specific and non-specific adsorption and glycosylation of surface proteins. The influence of the microenvironment on the exosome mobility suggests that their hydrodynamic size should be viewed as a dynamic property that changes with time and spatial translocation. The common practice of reporting only the distribution of hydrodynamic diameters provides an insufficient characterization of complex and conditional migration properties of the exosomes and should be supplemented, at a minimum, by the routine characterization of their vesicle sizes. The thickness of the coronal layer in sera exosomes was examined next. To that end, the hydrodynamic and vesicle sizes of exosomes isolated from the blood of two healthy individuals (samples H1 and H2 in Fig. 4) were analyzed by two independent laboratories. Multiple methods of exosome enrichment are in current use, each with its benefits and shortcomings. We ruled out the influence of an isolation technique on our conclusions by employing
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
4
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
exosomes treated with proteinase K which were less likely to be internalized by cells [19] despite their higher mobility. Further investigations are needed to understand the multifaceted roles of surface proteins in controlling the exosome trafficking, docking, and cellular internalization. Experimental methods The details on experimental methods are given in Supporting Information. Associated content Supporting Information contains the description of the samples, the experimental procedures, supporting results, figures, tables, and references. Author information
Fig. 4. Vesicles of sera exosomes, characterized by cryo-TEM and SEM, are much smaller and more narrowly distributed than their hydrodynamic sizes measured by the NTA.
dissimilar enrichment techniques. The vesicles in sample H1 were isolated by precipitation. Their sizing in cryo-TEM and SEM images (Fig. S4 and S7, SI) showed substantially smaller (p < 0.01) vesicles than the corresponding mobility sizes determined by the NTA. The exosomes in sample H2 were isolated by size exclusion chromatography and their vesicles sized in SEM images (Fig. S4, SI). As with H1 sample, the difference between the vesicle and hydrodynamic sizes was statistically significant. Consistent with MCF-7 case, the hydrodynamic diameters of sera exosomes in the two characterized samples are much larger than their vesicles, indicating the presence a substantial macromolecular corona restricting the exosome mobility. While the difference is surprisingly large, it is not without a precedent as the addition of the PEGylation layer to gold nanoparticles has been previously shown to cause a similarly large increase in hydrodynamic diameters [16,17]. In summary, we report a substantially impeded migration of exosomes relative to the expectations based on the size of their vesicles and suggest quantifying the excesses hindrance as a thickness of the coronal layer formed by macromolecular surface decorations of the exosomes. We find the size of exosomal vesicles to be narrowly distributed compared to their broadly distributed hydrodynamic sizes, indicating a widely varying surface decoration and thickness of the coronal layer. The significant and heterogeneous impact of the coronal layer on the reduction in the exosome mobility is reduced by proteolysis of surface proteins, which increases the mobility substantially and shifts it closer to the range expected for the size of membrane envelopes. Exosomes with smaller hydrodynamic sizes are more likely to overcome transport barriers, reach cellular targets, and participate in signaling. The revealed wide variations in coronal thickness imply that the exosomes migrate with dissimilar rates even when their vesicles are of the same size. These conclusions are invariant to the types of biofluids from which we isolated the exosomes, the enrichment techniques used, and the analysis performed by two independent laboratories. The importance of coronal layer extends beyond its influence on the mobility. Cleavage of surface proteins was shown to reduce binding and cellular internalization [18], as in the case of the
Corresponding authors: Mikhail Skliar. The authors declare no competing financial interests. Acknowledgment The authors acknowledge financial support from the National Science Foundation (award number IGERT-0903715), the University of Utah (Department of Chemical Engineering Seed Grant and the Graduate Research Fellowship Award), Skolkovo Institute of Physics and Technology (Skoltech Fellowship) and Moscow Institute of Science and Technology (5top100 Fellowship). Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.05.107. References €fer, E. Beerling, [1] A. Zomer, C. Maynard, F.J. Verweij, A. Kamermans, R. Scha R.M. Schiffelers, E. de Wit, J. Berenguer, S.I.J. Ellenbroek, T. Wurdinger, D.M. Pegtel, J. van Rheenen, In vivo imaging reveals extracellular vesiclemediated phenocopying of metastatic behavior, Cell 161 (2015) 1046e1057. [2] R. Safaei, B.J. Larson, T.C. Cheng, M.A. Gibson, S. Otani, W. Naerdemann, S.B. Howell, Abnormal lysosomal trafficking and enhanced exosomal export of cisplatin in drug-resistant human ovarian carcinoma cells, Mol. Canc. Therapeut. 4 (2005) 1595e1604. [3] D.W. Greening, S.K. Gopal, R. Xu, R.J. Simpson, W. Chen, Exosomes and their roles in immune regulation and cancer, Semin. Cell Dev. Biol. 40 (2015) 72e81. [4] F. Caponnetto, I. Manini, M. Skrap, T. Palmai-Pallag, C. Di Loreto, A.P. Beltrami, D. Cesselli, E. Ferrari, Size-dependent cellular uptake of exosomes, Nanomed. Nanotechnol. Biol. Med. 13 (2017) 1011e1020. [5] V.S. Chernyshev, R. Rachamadugu, Y.H. Tseng, D.M. Belnap, Y. Jia, K.J. Branch, A.E. Butterfield, L.F. Pease, P.S. Bernard, M. Skliar, Size and shape characterization of hydrated and desiccated exosomes, Anal. Bioanal. Chem. 407 (2015) 3285e3301. [6] M. Hansen, M.C. Smith, R.M. Crist, J.D. Clogston, S.E. McNeil, Analyzing the influence of PEG molecular weight on the separation of PEGylated gold nanoparticles by asymmetric-flow field-flow fractionation, Anal. Bioanal. Chem. 407 (2015) 8661e8672. [7] P.S. Sarangapani, S.D. Hudson, K.B. Migler, J.A. Pathak, The limitations of an exclusively colloidal view of protein solution hydrodynamics and rheology, Biophys. J. 105 (2013) 2418e2426. [8] K. Rilla, H. Siiskonen, M. Tammi, R. Tammi, Hyaluronan-coated extracellular vesiclesda novel link between hyaluronan and cancer, Adv. Canc. Res. (2014) 121e148. [9] D.A. Rusakov, D.M. Kullmann, Geometric and viscous components of the tortuosity of the extracellular space in the brain, Proc. Natl. Acad. Sci. Unit. States Am. 95 (1998) 8975e8980. [10] M. Mota, J.A. Teixeira, J.B. Keating, A. Yelshin, Changes in diffusion through the brain extracellular space, Biotechnol. Appl. Biochem. 39 (2004) 223e232. [11] J.L. Anderson, J.A. Quinn, Restricted transport in small pores, Biophys. J. 14 (1974) 130e150.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5 , J.P. Dupuis, B. Lounis, L. Groc, [12] A.G. Godin, J.A. Varela, Z. Gao, N. Danne L. Cognet, Single-nanotube tracking reveals the nanoscale organization of the extracellular space in the live brain, Nat. Nanotechnol. 12 (2017) 238e243. [13] L. Huleihel, G.S. Hussey, J.D. Naranjo, L. Zhang, J.L. Dziki, N.J. Turner, D.B. Stolz, S.F. Badylak, Matrix-bound nanovesicles within ECM bioscaffolds, Sci. Adv 2 (2016) e1600502. , C. Nicholson, Diffusion in brain extracellular space, Physiol. Rev. 88 [14] E. Sykova (2008) 1277e1340. €ller, Host matrix modulation by tumor exosomes pro[15] W. Mu, S. Rana, M. Zo motes motility and invasiveness, Neoplasia 15 (2013) 875e887. IN4. [16] T. Ishii, K. Miyata, Y. Anraku, M. Naito, Y. Yi, T. Jinbo, S. Takae, Y. Fukusato, M. Hori, K. Osada, K. Kataoka, Enhanced target recognition of nanoparticles by
5
cocktail PEGylation with chains of varying lengths, Chem. Commun. 52 (2016) 1517e1519. [17] E. Oh, J.B. Delehanty, K.E. Sapsford, K. Susumu, R. Goswami, J.B. Blanco-Canosa, P.E. Dawson, J. Granek, M. Shoff, Q. Zhang, P.L. Goering, A. Huston, I.L. Medintz, Cellular uptake and fate of PEGylated gold nanoparticles is dependent on both cell-penetration peptides and particle size, ACS Nano 5 (2011) 6434e6448. [18] T.J. Smyth, J.S. Redzic, M.W. Graner, T.J. Anchordoquy, Examination of the specificity of tumor cell derived exosomes with tumor cells in vitro, Biochim. Biophys. Acta Biomembr. 1838 (2014) 2954e2965. [19] C. Escrevente, S. Keller, P. Altevogt, J. Costa, Interaction and uptake of exosomes by ovarian cancer cells, BMC Canc. 11 (2011) 108.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Membrane proteins significantly restrict exosome mobility Mikhail Skliar a, b, *, Vasiliy S. Chernyshev c, d, David M. Belnap e, German V. Sergey d, Samer M. Al-Hakami a, Philip S. Bernard f, g, Inge J. Stijleman f, Rakesh Rachamadugu f a
Chemical Engineering, University of Utah, 50 S. Central Campus Dr, Salt Lake City, UT, 84112, USA The Nano Institute of Utah, University of Utah, 36 S. Wasatch Dr, Salt Lake City, UT, 84112, USA c Center for Translational Biomedicine, Skolkovo Institute of Science and Technology, Skolkovo Innovation Center, Building 3, Moscow, 143026, Russia d Biopharmaceutical Cluster ‘Northern’, Moscow Institute of Physics and Technology, Institutsky per. 9/7, Dolgoprudny, Moscow Region, 141700, Russia e Biochemistry and Biology Departments, University of Utah, 15 N Medical Dr, Salt Lake City, UT, 84112, USA f Huntsman Cancer Institute, University of Utah, 2000 Circle of Hope, Salt Lake City, UT, 84112, USA g Department of Pathology, University of Utah, 15 North Medical Dr, Salt Lake City, UT, 84112, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 May 2018 Accepted 15 May 2018 Available online xxx
Exosomes are membrane nanovesicles implicated in cell-to-cell signaling in which they transfer their molecular cargo from the parent to the recipient cells. This role essentially depends on the exosomes' small size, which is the prerequisite for their rapid migration through the crowded extracellular matrix and into and out of circulation. Here we report much lower exosome mobility than expected from the size of their vesicles, implicate membrane proteins in a substantially impeded rate of migration, and suggest an approach to quantifying the impact. The broadly distributed excess hydrodynamic resistance provided by surface proteins produces a highly heterogeneous and microenvironment-dependent hindrance to exosome mobility. The implications of the findings on exosome-mediated signaling are discussed. © 2018 Elsevier Inc. All rights reserved.
Keywords: Exosomes and extracellular vesicles Membrane proteins Proteolysis Vesicle trafficking
Exosomes are actively secreted by cells and found in all biological fluids, including blood, urine, and saliva. Compared to other extracellular vesicles (EVs), exosomes are distinguished by biomarkers of the late endocytic biogenesis and the smallest size, typically reported to be 30e150 nm in diameter. The uptake of an exosome by a local or distant cell transfers the molecular cargo derived from a secreting parent to a recipient cell. Biologically active molecules transferred in health and disease by this mechanism include surface and luminal proteins, membrane-bound microRNAs, and other compounds. A growing number of studies implicate exosomal signaling in tumor metastasis [1], drug resistance [2], and the modulation of immune response [3]. For signaling to occur it is necessary, but not sufficient, for a secreted EV to migrate from a parent to a recipient cell. In paracrine signaling, the transport barrier is overcome by diffusion through extracellular space, while to reach distal targets an EV may need to enter and exit a biofluid circulation. The smallest among EVs, the exosomes have the highest mobility which favors overcoming
* Corresponding author. Department of Chemical Engineering, University of Utah, 50 S. Central Campus Dr, Salt Lake City, UT, 84112, USA. E-mail address: [email protected] (M. Skliar).
transport resistance for exosome-mediated signaling to take place [4]. The mobility of the exosomes may be quantified by their mean squared displacements (MSD) with time, which proportionally depends on their (self-) diffusivity. The oft-reported hydrodynamic diameter of exosomes is the measure of the resistance to the migration and is inversely proportional to their diffusivity. Transport barriers to exosome dissemination are poorly understood. Here, we examine the impact of surface proteins and macromolecules anchored by them on the mobility of exosomes isolated by dissimilar methods (precipitation and size exclusion) from two biological fluids (growth medium of MCF-7 breast cancer cells and human sera). The excess resistance to the migration imposed by surface proteins is quantified as the thickness of a coronal layer formed by surface-conjugated macromolecules, which we measure as a difference between hydrodynamic and vesicle diameters of the exosomes. The exosomes released by MCF-7 cells were isolated by precipitation and their enrichment in the isolate was confirmed by positive expression of surface and luminal biomarkers, the globular shape of vesicles with bilayer membrane morphology, and the hydrodynamic diameters in the exosomal range (Fig. 1a and S1-S2 in Supporting Information, SI). The mean and mode hydrodynamic
https://doi.org/10.1016/j.bbrc.2018.05.107 0006-291X/© 2018 Elsevier Inc. All rights reserved.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
2
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
Fig. 1. (a) The size of exosome vesicles in the solution (red curve, grey fill) was obtained from the AFM measurements and confirmed by cryo-TEM imaging. Cryo-TEM shows the expected globular shape and double-layer morphology of vesicles (inset). The vesicles are substantially smaller than their hydrodynamic diameters measured by NTA (blue curve, blue fill). The average thickness of the exosomal corona (inset summary) is estimated as the difference between the average hydrodynamic and vesicle sizes. (b) Trypsinization of surface proteins substantially reduces hydrodynamic diameters. (c) The shift in mobility sizes is even more pronounced after a less specific protease K digestion. PK proteolysis narrows the distribution of hydrodynamic sizes, pointing to the heterogeneity in the macromolecular surface decoration as the source of widely different mobilities of untreated exosomes. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
diameters measured by nanoparticle tracking analysis (NTA) were 111 ± 12 nm and 80 ± 9 nm, respectively. The volume encapsulated by hydrated vesicles was measured by atomic force microscopy (AFM; Fig. S5-S6 and Table S4, SI). The diameter of hydrated vesicles in their innate globular shape was estimated by mapping the AFM volume measurements, obtained for exosomes electrostatically immobilized on a surface, into volume-equivalent spheres in the solution. Fig. 1a shows the obtained probability density function of vesicle sizes, which has 30.1 ± 6.9 nm ensemble average. A much smaller size of vesicles compared with their hydrodynamic diameters was independently confirmed by cryo-TEM imaging which showed MCF-7 vesicles with an average diameter equal to 34.2 nm (Fig. S7a, SI). The macromolecular corona is indistinguishable in cryo-TEM images because of its low excess density [5,6]. The NTA measurements of hydrodynamic sizes were repeated after enzymatic proteolysis of MCF-7 exosomes. The four digestion protocols (Table S5, SI) differed in enzymes used and the treatment duration. A statistically significant (p < 0.01) increase in exosome mobility after the proteolysis by as much as 50% was observed (Fig. 1b and c; Fig. S3 and Table S2, SI). The duration of enzymatic treatments had little effect on hydrodynamic sizes, indicating a sufficiently long incubation to complete the digestion. The following factors were ruled out from contributing to the observed reduction in hydrodynamic sizes: a) Enzymatic treatments did not change the size of vesicles which remained in the same range before and after digestion (Fig. S8, SI and b) The protein concentration in the solution was too low (less than 3.5 mg/mL) to cause changes in the viscosity after the proteolysis [7]. The mobility of exosomes after proteolysis was enzyme dependent (Fig. 1b and c). Trypsin preferentially cleaves at lysine and arginine locations but does not affect surface-anchored segments of membrane proteins void of these a-amino acids. Trypsinization thinned the coronal layer, reducing the hydrodynamic size of exosomes to having the mode diameter equal to 64 nm, down from the original 80 nm (Fig. 1b and Table S2, SI). The reduction was even more significant (down to the 56 nm mode diameter; Fig. 1c) when we used a less specific proteinase K (PK), known to cleave peptide bonds of hydrophobic, aliphatic and aromatic amino acids. Compared to trypsinization, the broader PK activity leaves shorter undigested fragments of surface proteins and, thus, a thinner coronal layer, as reflected by a higher mobility of PK-treated exosomes. The indiscriminate PK proteolysis shifts the hydrodynamic diameters of the exosomes into the range of sizes partially overlapping with vesicle sizes. The remaining difference may be
explained by short fragments of surface proteins that survived the digestion and the presence of pericellular coats formed by glycosaminoglycans, and hyaluronan specifically [8], which are unaffected by trypsin and PK. The described mechanism of enzyme-dependent thinning of exosome corona is summarized in Fig. 2, where we compare the distributions of the vesicle and hydrodynamic diameters before and after the proteolysis. After the proteolysis, the hydrodynamic sizes are distributed more uniformly. The less specific PK treatment, which closely “shaves” the membrane surface, leads to the especially narrow range of hydrodynamic diameters with a closer resemblance to the distribution of vesicle sizes. The broadly distributed hydrodynamic diameters of vesicles with narrowly distributed membrane envelopes reveal the heterogeneity in the macromolecular surface decoration of exosomes and the widely varying thickness of their coronal layer from one exosome to another. The direct visualization of diverse migration rates of MCF-7 exosomes was obtained by tracking the light scattered by a single exosome as it moves in PBS. The three trajectories in Fig. 3a were
Fig. 2. The reduction in the coronal layers is enzyme dependent. Non-specific protein digestion by proteinase K shifts the hydrodynamic diameters closer to the range of vesicle sizes.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
3
Fig. 3. (a) Single-particle tracking of migrating exosomes having the mode, 10th, and 90th percentile hydrodynamic diameters. (b) The squared particle displacements corresponding to the tracks in (a) and the theoretical predictions of the MSD extended beyond the observation period show the long-term trend in size-dependent reach of migrating exosomes. (c) The extracellular matrix exaggerates the size dependence of exosome mobility. The exosomes become immobilized as their hydrodynamic diameter approaches the size of ECM pores.
tracked for ~58 s and show the Brownian displacement of exosomes with 88.5, 70.5, and 142.5 nm hydrodynamic diameters, which correspond to the mode, 10th, and 90th percentiles values of the hydrodynamic size distribution before proteolysis. The squared displacements along these trajectories and the theoretical predictions of the MSD beyond the observation period illustrate strong size-dependence of the exosomal reach and a higher probability of smaller exosomes overcoming the transport barrier to reach a target. Little is known about the migration of exosomes through the extracellular space (ECS). An early insight into the size-dependence to the hindered transit of exosomes through the ECS and the extracellular matrix (ECM) that occupies it may be gained from modeling. Following the approach used to describe the ECM migration of molecules and nanoparticles [9], we related the effective diffusivity of the exosomes in the ECS, De , to their unimpeded diffusivity in the solution, D0 , as D0 ¼ ðtg tv Þ2 De , where the geometric tortuosity, tg , describes the reduction in the migration rate due to a longer path an exosome must traverse through a meandering extracellular pore. We used tg ¼ 1:67 which provides the best fit to the experimental data on the tortuosity in the cortex of rats [10]. The “viscous” tortuosity tv describes the contribution of steric hindrance to the exosome migration, their specific and nonspecific interactions with ECM, and all other factors impeding the movement of the exosomes through the ECS. Its value depends on the hydrodynamic diameter of the exosomes relative to the size of an ECS pore through which the diffusion occurs, and was expressed using a steric partition coefficient and the enhanced pore-to-solution friction ratio [11]. Fig. 3c compares the ECM diffusivity in a 150 nm ECS pore (equal to mean confining ECS dimension inside a brain of a living rat [12]) with unimpeded diffusivity in the solution. Average distance traveled in the ECS by an exosome of a small hydrodynamic size (in the range of membrane vesicles) is less than one fifth the distance translocated in the solution, as indicated by the ratio of the MSDs during hindered and free diffusion (Fig. 3c). The distance drops to ~1% for 80 nm particles and becomes negligible as the hydrodynamic size of the exosomes approaches the pore size. The exaggerated influence of sizes on the ECS mobility increases the dwell time of hydrodynamically larger exosomes in the proximity of a parent cell and favors their sequestration. The routine trapping of EVs in urinary bladder matrix and other ECM scaffolds was
reported recently [13] without concluding the process is size dependent. The developed model predicts cessation in mobility for exosomes with hydrodynamic sizes smaller than pores. According to this estimate, approximately half of analyzed MCF-7 exosomes (a subpopulation with hydrodynamic diameters above 100 nm) would be unable to transit through the ECM with 150 nm maximum pore size. This prediction is conservative and does not account for conformational flexing of macromolecular corona which may overcome steric hindrance, just as flexible molecules were shown to translate through the ECM while the diffusion of rigid molecules of similar molecular weight was restricted [14]. An alternative estimate that uses vesicle sizes to predict the mobility cut-off is less restrictive and predicts that all MCF-7 exosomes should be able to migrate in the ECM with 150 nm pores, though with an increased hindrance compared to the unimpeded diffusion. Exosome-microenvironment interactions influence the mobility cut-off, shifting it towards one of the described theoretical bounds defined by either hydrodynamic or vesicle sizes of exosomes. For example, the reduction of the coronal layer in vivo by extracellular enzymes, such as metalloproteinases, will increase the mobility towards a more optimistic prediction. Interestingly, the proteolytic activity may be bidirectional, with tumor exosomes reportedly degrading the extracellular matrix [15], thus reducing the hindrance to their ECM mobility. A shift towards lower mobility will occur with the increase in the coronal layer as a result of specific and non-specific adsorption and glycosylation of surface proteins. The influence of the microenvironment on the exosome mobility suggests that their hydrodynamic size should be viewed as a dynamic property that changes with time and spatial translocation. The common practice of reporting only the distribution of hydrodynamic diameters provides an insufficient characterization of complex and conditional migration properties of the exosomes and should be supplemented, at a minimum, by the routine characterization of their vesicle sizes. The thickness of the coronal layer in sera exosomes was examined next. To that end, the hydrodynamic and vesicle sizes of exosomes isolated from the blood of two healthy individuals (samples H1 and H2 in Fig. 4) were analyzed by two independent laboratories. Multiple methods of exosome enrichment are in current use, each with its benefits and shortcomings. We ruled out the influence of an isolation technique on our conclusions by employing
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
4
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5
exosomes treated with proteinase K which were less likely to be internalized by cells [19] despite their higher mobility. Further investigations are needed to understand the multifaceted roles of surface proteins in controlling the exosome trafficking, docking, and cellular internalization. Experimental methods The details on experimental methods are given in Supporting Information. Associated content Supporting Information contains the description of the samples, the experimental procedures, supporting results, figures, tables, and references. Author information
Fig. 4. Vesicles of sera exosomes, characterized by cryo-TEM and SEM, are much smaller and more narrowly distributed than their hydrodynamic sizes measured by the NTA.
dissimilar enrichment techniques. The vesicles in sample H1 were isolated by precipitation. Their sizing in cryo-TEM and SEM images (Fig. S4 and S7, SI) showed substantially smaller (p < 0.01) vesicles than the corresponding mobility sizes determined by the NTA. The exosomes in sample H2 were isolated by size exclusion chromatography and their vesicles sized in SEM images (Fig. S4, SI). As with H1 sample, the difference between the vesicle and hydrodynamic sizes was statistically significant. Consistent with MCF-7 case, the hydrodynamic diameters of sera exosomes in the two characterized samples are much larger than their vesicles, indicating the presence a substantial macromolecular corona restricting the exosome mobility. While the difference is surprisingly large, it is not without a precedent as the addition of the PEGylation layer to gold nanoparticles has been previously shown to cause a similarly large increase in hydrodynamic diameters [16,17]. In summary, we report a substantially impeded migration of exosomes relative to the expectations based on the size of their vesicles and suggest quantifying the excesses hindrance as a thickness of the coronal layer formed by macromolecular surface decorations of the exosomes. We find the size of exosomal vesicles to be narrowly distributed compared to their broadly distributed hydrodynamic sizes, indicating a widely varying surface decoration and thickness of the coronal layer. The significant and heterogeneous impact of the coronal layer on the reduction in the exosome mobility is reduced by proteolysis of surface proteins, which increases the mobility substantially and shifts it closer to the range expected for the size of membrane envelopes. Exosomes with smaller hydrodynamic sizes are more likely to overcome transport barriers, reach cellular targets, and participate in signaling. The revealed wide variations in coronal thickness imply that the exosomes migrate with dissimilar rates even when their vesicles are of the same size. These conclusions are invariant to the types of biofluids from which we isolated the exosomes, the enrichment techniques used, and the analysis performed by two independent laboratories. The importance of coronal layer extends beyond its influence on the mobility. Cleavage of surface proteins was shown to reduce binding and cellular internalization [18], as in the case of the
Corresponding authors: Mikhail Skliar. The authors declare no competing financial interests. Acknowledgment The authors acknowledge financial support from the National Science Foundation (award number IGERT-0903715), the University of Utah (Department of Chemical Engineering Seed Grant and the Graduate Research Fellowship Award), Skolkovo Institute of Physics and Technology (Skoltech Fellowship) and Moscow Institute of Science and Technology (5top100 Fellowship). Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.bbrc.2018.05.107. References €fer, E. Beerling, [1] A. Zomer, C. Maynard, F.J. Verweij, A. Kamermans, R. Scha R.M. Schiffelers, E. de Wit, J. Berenguer, S.I.J. Ellenbroek, T. Wurdinger, D.M. Pegtel, J. van Rheenen, In vivo imaging reveals extracellular vesiclemediated phenocopying of metastatic behavior, Cell 161 (2015) 1046e1057. [2] R. Safaei, B.J. Larson, T.C. Cheng, M.A. Gibson, S. Otani, W. Naerdemann, S.B. Howell, Abnormal lysosomal trafficking and enhanced exosomal export of cisplatin in drug-resistant human ovarian carcinoma cells, Mol. Canc. Therapeut. 4 (2005) 1595e1604. [3] D.W. Greening, S.K. Gopal, R. Xu, R.J. Simpson, W. Chen, Exosomes and their roles in immune regulation and cancer, Semin. Cell Dev. Biol. 40 (2015) 72e81. [4] F. Caponnetto, I. Manini, M. Skrap, T. Palmai-Pallag, C. Di Loreto, A.P. Beltrami, D. Cesselli, E. Ferrari, Size-dependent cellular uptake of exosomes, Nanomed. Nanotechnol. Biol. Med. 13 (2017) 1011e1020. [5] V.S. Chernyshev, R. Rachamadugu, Y.H. Tseng, D.M. Belnap, Y. Jia, K.J. Branch, A.E. Butterfield, L.F. Pease, P.S. Bernard, M. Skliar, Size and shape characterization of hydrated and desiccated exosomes, Anal. Bioanal. Chem. 407 (2015) 3285e3301. [6] M. Hansen, M.C. Smith, R.M. Crist, J.D. Clogston, S.E. McNeil, Analyzing the influence of PEG molecular weight on the separation of PEGylated gold nanoparticles by asymmetric-flow field-flow fractionation, Anal. Bioanal. Chem. 407 (2015) 8661e8672. [7] P.S. Sarangapani, S.D. Hudson, K.B. Migler, J.A. Pathak, The limitations of an exclusively colloidal view of protein solution hydrodynamics and rheology, Biophys. J. 105 (2013) 2418e2426. [8] K. Rilla, H. Siiskonen, M. Tammi, R. Tammi, Hyaluronan-coated extracellular vesiclesda novel link between hyaluronan and cancer, Adv. Canc. Res. (2014) 121e148. [9] D.A. Rusakov, D.M. Kullmann, Geometric and viscous components of the tortuosity of the extracellular space in the brain, Proc. Natl. Acad. Sci. Unit. States Am. 95 (1998) 8975e8980. [10] M. Mota, J.A. Teixeira, J.B. Keating, A. Yelshin, Changes in diffusion through the brain extracellular space, Biotechnol. Appl. Biochem. 39 (2004) 223e232. [11] J.L. Anderson, J.A. Quinn, Restricted transport in small pores, Biophys. J. 14 (1974) 130e150.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107
M. Skliar et al. / Biochemical and Biophysical Research Communications xxx (2018) 1e5 , J.P. Dupuis, B. Lounis, L. Groc, [12] A.G. Godin, J.A. Varela, Z. Gao, N. Danne L. Cognet, Single-nanotube tracking reveals the nanoscale organization of the extracellular space in the live brain, Nat. Nanotechnol. 12 (2017) 238e243. [13] L. Huleihel, G.S. Hussey, J.D. Naranjo, L. Zhang, J.L. Dziki, N.J. Turner, D.B. Stolz, S.F. Badylak, Matrix-bound nanovesicles within ECM bioscaffolds, Sci. Adv 2 (2016) e1600502. , C. Nicholson, Diffusion in brain extracellular space, Physiol. Rev. 88 [14] E. Sykova (2008) 1277e1340. €ller, Host matrix modulation by tumor exosomes pro[15] W. Mu, S. Rana, M. Zo motes motility and invasiveness, Neoplasia 15 (2013) 875e887. IN4. [16] T. Ishii, K. Miyata, Y. Anraku, M. Naito, Y. Yi, T. Jinbo, S. Takae, Y. Fukusato, M. Hori, K. Osada, K. Kataoka, Enhanced target recognition of nanoparticles by
5
cocktail PEGylation with chains of varying lengths, Chem. Commun. 52 (2016) 1517e1519. [17] E. Oh, J.B. Delehanty, K.E. Sapsford, K. Susumu, R. Goswami, J.B. Blanco-Canosa, P.E. Dawson, J. Granek, M. Shoff, Q. Zhang, P.L. Goering, A. Huston, I.L. Medintz, Cellular uptake and fate of PEGylated gold nanoparticles is dependent on both cell-penetration peptides and particle size, ACS Nano 5 (2011) 6434e6448. [18] T.J. Smyth, J.S. Redzic, M.W. Graner, T.J. Anchordoquy, Examination of the specificity of tumor cell derived exosomes with tumor cells in vitro, Biochim. Biophys. Acta Biomembr. 1838 (2014) 2954e2965. [19] C. Escrevente, S. Keller, P. Altevogt, J. Costa, Interaction and uptake of exosomes by ovarian cancer cells, BMC Canc. 11 (2011) 108.
Please cite this article in press as: M. Skliar, et al., Membrane proteins significantly restrict exosome mobility, Biochemical and Biophysical Research Communications (2018), https://doi.org/10.1016/j.bbrc.2018.05.107