Protein aggregates are nonvesicular assemblies that can overlap EVs in size, density, sedimentation, and light-scattering behavior. They may arise from source material, cell death, high protein concentration, freeze-thaw stress, agitation, pH changes, or processing. Aggregates can inflate particle counts, contribute protein signal, bind dyes or antibodies, and confound functional assays. Their presence should be evaluated with orthogonal methods rather than inferred from one particle-counting assay.
Figure 1 placeholder — Protein aggregates
Lipoproteins
Lipoproteins are lipid-protein particles that transport lipids and can be major contaminants in biofluid-derived EV preparations. Chylomicrons, very-low-density lipoproteins, low-density lipoproteins, and high-density lipoproteins overlap EVs in important physical properties and can co-isolate during common workflows. Because they carry abundant proteins, lipids, and nucleic-acid-associated material, inadequate separation can distort EV composition and functional interpretation.
Figure 2 placeholder — Lipoproteins
Cell debris
Cell debris includes membrane fragments, organelle fragments, cytoskeletal material, apoptotic material, and other particles released by damaged or dying cells. Clarification steps reduce this material, but incomplete removal can contaminate downstream EV fractions. Cell viability, harvest technique, processing delay, centrifugation, and filtration influence debris burden. Debris can contribute proteins, nucleic acids, lipids, enzymes, and biological activity incorrectly attributed to EVs.
Figure 3 placeholder — Cell debris
Media-derived particles
Media-derived particles originate from culture components rather than the intended producer cells. Protein supplements, lipid particles, reagent complexes, growth-factor formulations, and other constituents can enter an EV workflow and create background in particle counting, proteomics, lipidomics, RNA analysis, or functional assays. A media-only control processed through the complete workflow helps distinguish producer-cell-derived material from process background.
Figure 4 placeholder — Media-derived particles
Serum-derived EVs
Serum-derived EVs are naturally present in animal serum supplements such as fetal bovine serum. When serum-containing media are used, these vesicles and other serum nanoparticles can be difficult to distinguish from producer-cell EVs. EV-depleted serum can reduce but does not necessarily eliminate this background, and depletion itself can alter serum composition. Serum-free or chemically defined media can reduce selected confounders but introduce different biological and manufacturing considerations.
Figure 5 placeholder — Serum-derived EVs
Viruses
Viruses and virus-like particles can overlap EVs in size, density, membrane composition, and aspects of biogenesis. In infected or virus-producing systems, physical separation can be especially difficult. Even in nonviral manufacturing, adventitious-virus control remains part of source-cell qualification, raw-material control, and safety strategy. Morphology or size alone cannot distinguish EVs from viruses; appropriate virological testing and risk-based adventitious-agent testing are required.
Figure 6 placeholder — Viruses
Residual processing materials
Residual processing materials are substances introduced during collection, separation, purification, formulation, or storage that remain in the final preparation. Examples include polyethylene glycol, density-gradient medium, detergents, enzymes, nucleases, affinity reagents, membrane extractables or leachables, buffer components, antibiotics, and cryoprotectants. Residuals can alter measurements, interfere with functional assays, or create safety concerns. Manufacturing specifications should define relevant limits and methods for control.
Figure 7 placeholder — Residual processing materials
Clinical takeaway
EV purity is not established by particle count, size, or a few positive markers. Protein aggregates, lipoproteins, cell debris, media-derived particles, serum-derived EVs, viruses, and residual processing materials can contribute signals or biological effects mistakenly attributed to EVs.
Purity is method-dependent
Purity depends on source material, separation workflow, analytical method, and the contaminant being considered. A preparation can be relatively free of one contaminant while remaining enriched in another.
Particle-to-protein ratios
A particle-to-protein ratio has sometimes been used as a rough contextual indicator, but it is not a universal purity metric because both particle counts and protein measurements can include non-EV material.
Media and process controls
A media-only control or process blank should be carried through the same collection, concentration, separation, and analytical steps when feasible to identify background introduced by media, reagents, equipment, or workflow.
Cell health and functional confounding
Cell death releases intracellular proteins, nucleic acids, organelles, membrane fragments, and apoptotic particles. Biological effects can also arise from soluble proteins, lipoproteins, residual reagents, cytokines, viruses, or other co-isolated material.
Manufacturing relevance
For a clinical product, contaminant control should be incorporated into raw-material qualification, in-process testing, release specifications, stability studies, and comparability assessments.
Clinical interpretation
Claims that an EV or exosome product is pure should identify the contaminants tested, methods, acceptance criteria, and results. Absence of one contaminant marker does not establish comprehensive purity.
Knowledge gaps
Reference standards and universally accepted purity thresholds remain limited. Complex matrices contain overlapping particle populations that current methods cannot always completely resolve.
Bottom-line evidence statement
EV purity is a multidimensional assessment of what is present in addition to intended vesicles. Strong evidence measures relevant contaminants directly, uses process-matched controls, and avoids attributing molecular or biological signals to EVs until plausible non-EV sources are addressed.
Additional controlled terminology
A chylomicron, very-low-density lipoprotein, low-density lipoprotein, or high-density lipoprotein can contaminate EV preparations. Cell debris may include a membrane fragment, organelle fragment, or apoptotic material, making clarification important. A media-only control helps identify culture-derived background. Fetal bovine serum can contribute particles; EV-depleted serum reduces but may not eliminate them, while serum-free or chemically defined media change the background differently. A virus-like particle can overlap EV properties, and an adventitious virus requires an appropriate adventitious-agent testing strategy. Residuals may include polyethylene glycol or density-gradient medium; manufacturing materials can contribute an extractable or leachable. A process blank can identify workflow background. A particle-to-protein ratio provides limited context but is not a universal measure of EV purity.