Extracellular vesicles (EVs) are lipid-bilayer nanovesicles commonly isolated from both blood plasma and serum [1
], produced by nearly all cell types [4
], and secreted into their surrounding extracellular fluid (ECF). Common EV populations found in various biofluids range in size from <50 nanometers (nm; 10−9
m) to larger than two microns (>2000 nm) [3
]. Exosomes, a specific subpopulation of EVs, are produced via an alternative cellular mechanism than other EVs, have a relatively narrow size range (70–120 nm), and contain unique intrinsic cargo molecules (e.g., soluble or membrane bound proteins, lipid species, metabolites, and various nucleic acids) [6
]. Although all EV subtypes carry distinct cargo molecules consequent to their diverse cells of origin, EVs are also thought to carry an intrinsic capacity to influence both local and/or distant cellular targets [5
] through their uptake from the ECF, and providing autocrine-, paracrine-, and endocrine-like influence on target cell biology.
Various cell types (e.g., primary, cancerous, and induced-pluripotent stem cells [iPSCs]) grown in vitro require a complete (or serum supplemented) culture medium (CM) to optimize growth and viability. This specific supplementation is distinguished from that found in a defined basal medium, such as Dulbecco’s Modified Eagle’s Medium (DMEM), which features basic nutrients for cellular metabolism, proliferation, and homeostasis. A common supplement found in complete CM for optimized cultured cell growth is provided by the addition of heat-inactivated fetal bovine serum (FBS). Differing from what is included in DMEM, the addition of FBS to CM provides additional growth factors and carrier proteins, promotes cell adhesion, and protects cells from stresses associated with the in vitro environment [8
]. However, serum in general, and FBS specifically, contains high concentrations of EVs that may also directly influence the cultured cells being nourished in vitro [9
]. In addition, these FBS-derived EVs “contaminate” any quantitative and/or qualitative assessments of EVs derived from the cultured cells. Under such circumstances, therefore, in vitro experiments designed to explore the biology and characteristics of in vitro-derived EVs necessitates the attempted removal of exogenous EVs from the complete CM, or by utilizing serum-free CM that also supports cultured cell growth and viability. Studies of cultured cell-derived EVs, therefore, have utilized EV-depletion methods in an attempt to eliminate the significant number of FBS-derived EVs found in complete CM, while maintaining other serum factors that support in vitro growth and viability.
Two major types of FBS EV-depletion approaches have been commonly used, including either (1) high-speed ultracentrifugation (UC), or (2) a proprietary chemical precipitation, commercially available as Exo-FBSTM
(System Biosciences, Inc., SBI; Palo Alto, CA). Reports using either of these methods have focused on their respective negative influences on in vitro cellular growth and viability in a variety of cell types [11
]. An additional issue, less commonly addressed, is whether the “depleted” supplement added to the CM contaminates any quantitative and/or qualitative analyses of EVs derived from such cultured cells.
In this manuscript, our aim is to present the quantitative assessment of the EV depletion provided via two commonly used FBS EV-depletion methods, using nanoparticle tracking analysis (NTA). In addition, we also present the differential effects of FBS EV-depletion approaches on the growth and viability of cultured primary astrocytes. It is anticipated that the information presented, and options discussed, will be informative to those planning similar analyses and may ultimately catalyze investigations of relevant cell type-specific EVs in vitro, that may have applicability towards clinical translation.
This study provides quantitative evidence of the incomplete depletion of FBS-derived EVs within the complete CM that has undergone presumed “depletion” using two common FBS-EV depletion methodologies. In addition, both depletion methods investigated provide a negative influence on primary astrocyte growth and viability in vitro. A failure in FBS EV-depletion is especially relevant to those investigating endogenously produced EVs (e.g., exosomes) in vitro by specific cell types. Although prior publications have also described the effects of FBS EV-depletion methods regarding cellular growth and viability, organelle stress, and altered gene regulation [12
], an additional concern advancing our work was regarding whether the depletion of exogenous FBS-derived EVs was adequate to prevent interference of in vitro analyses of EVs produced by primary cells. We believe that exogenous EVs present within the FBS supplement contaminate results provided for any bioengineering, therapeutic, and/or diagnostic investigations focused only on endogenously derived EVs produced by the cultured cells [15
We currently lack a full appreciation for the final makeup of the FBS pellet produced by either the 18-h UC or Exo-FBSTM
precipitation procedures. We infer that they likely include an incomplete population of FBS-derived EVs, in addition to a variety of proteins and other precipitant species. The incomplete removal of lipid bilayer nanovesicles from FBS makes it extremely difficult to claim any in vitro cell-specific biomolecular profiling of EVs, especially when utilizing complete CM, even if derived using the two described FBS EV-depletion methods. This latter point was recently illustrated in an article [16
] that found many of the FBS-derived RNA transcripts, both free-floating species and within EVs, map to both the human and mouse genomes, leading prior investigations towards the inadvertent false-positive reporting of extracellular RNA release from cultured cells.
In addition, previous reports providing FBS EV-depletion analyses, using either dynamic light scattering (DLS) and/or NTA, have lacked protocol standardization and reproducibility across instruments and laboratories. Specifically, such analytic methods pose distinct issues, such as: (1) DLS, NTA, and resistive pulse sensing (RPS), hold a limit of detection to be at 70–80 nm [17
], thereby questioning the ability to define and quantify EV populations at the sub-50 nm range of accuracy; and (2) standardization of DLS, NTA, and RPS instrumentation requires reporting of not only dilution factors, but also specific instrument settings (for consistency across multiple sample replicates) that aim to accurately define and replicate the total number of EV-like particles found in a suspension [18
]. In our study, all quantitative analyses of EVs using our NTA device were limited to the 75–165 nm range, while maintaining consistent and reproducible instrument settings between sample runs and adjusting only the sample dilution prior to analyses for optimization of instrument readings and reliability.
Moreover, defining “pure” EV fractions has become an important topic in the field and has led to the establishment of recommended guidelines from the International Society for Extracellular Vesicles (ISEV) for characterizing the final EV isolate from cultured cells [19
]. However, the guidelines have lacked the inclusion of criteria to confirm the lack of FBS EV cross contamination within the complete CM used for isolation of cultured cell-derived EVs. In fact, many publications [20
] detailing in vitro exosome release characteristics have yet to quantify the residual EVs within the complete CM used for exosome (and EV) collection and subsequent analyses. The first step in adequately isolating “pure” cultured cell-derived EV fractions should be the inclusion of media that is completely devoid of FBS (or other extraneous) EV populations.
We believe, therefore, that it remains relevant for investigators to consider alternative complete CM formulations, beyond using the two common FBS depletion protocols we investigated, for any in vitro studies featuring cell-derived EV analyses. A novel methodology, ultrafiltration [22
], may provide such an alternative, through the use of a 100 kDa molecular weight membrane filter that removes FBS-derived EVs from complete CM used for in vitro EV studies. More functional assays (e.g., nucleic acid sequencing, protein quantification, and robust nanoparticle tracking analytics) using these and other methodologies, however, need to be thoroughly vetted prior to achieving widespread adoption. Finally, an evolving, but seemingly viable option utilizes serum-free CM, providing no exogenous EVs to the in vitro environment [23
]. The latter method has yet to define the ideal additives for a complete CM recipe necessary for cell growth and viability approaching that of FBS-supplemented complete CM for use with all cell types in vitro. The benefits of such an engineered serum-free CM may include a better appreciation of individual cultured cell physiology and pathobiology, and how the specific addition of external influences (e.g., EVs) might influence cell biology. Understanding the quantities (concentrations) of such “specifically-engineered supplements” required for proper cell growth and phenotypic function are essential to defining a future serum-free CM option that is devoid of exogenous EVs. Such an option is likely to yield an improved understanding of the genesis of EVs directly from cultured cells, with delineation of their cell- and phenotype-specific cargos, and eventually lead to an advanced diagnostic [24
] as well as novel therapeutic approaches for maintaining human health.
The current study reveals that available FBS EV-depletion methods fail to provide adequate reductions in FBS-derived EV populations within “depleted” complete CM. In addition, there are apparent negative effects on cultured primary astrocyte growth and viability using both tested “depletion” methods. We conclude, therefore, that the commonly used methods (ultracentrifugation and Exo-FBS™) used to produce EV-depleted complete CM remain inadequate for quantitative assessments of EVs produced in vitro and for providing optimal primary cultured cell growth and viability. With the latter result, our study suggests that other soluble factors, precipitated and/or removed from the FBS during the two tested depletion processes, influence primary astrocytes’ ability to thrive in culture. There remains a great need for an engineered CM, completely devoid of exogenous EVs, that supports cell growth and viability while eliminating the likelihood of exogenous EVs influencing the phenotype and EV production of a primary cultured cell. The development of such a complete CM is likely to catalyze investigations of specific cell–derived EVs (and their cargos) derived from primary cell cultures. Such details will advance our understanding of cell biology and ultimately EV-associated functions within complex biofluids.