How does freeze-thaw impact EV characterization?

Freezing is the most common way of storing extracellular vesicles (EVs) and the biofluids they are isolated from. However, there is now good evidence that freezing and thawing can change the vesicles that are measured afterwards. This article looks at what freezing does to EVs, how each of the common characterisation methods is affected, how the type of sample influences the result, and what can be done to reduce these effects.

What happens to an EV when it is frozen?

EVs are small vesicles surrounded by a lipid bilayer and suspended in an aqueous buffer. As the sample cools, ice crystals form and grow. Salts, proteins and vesicles are concentrated into the liquid that remains between the crystals, and the vesicle membranes are exposed to ice surfaces and to osmotic stress as water moves out of the vesicle. During thawing the process is reversed, and slow warming allows small ice crystals to grow into larger ones.

This has three main consequences for characterisation. Membranes can be damaged, allowing the contents of the vesicle to leak out. Vesicles that are forced close together can aggregate or fuse. Proteins that are loosely associated with the vesicle surface can also be lost. All three effects were seen in airway exosomes that were frozen at −80 °C for four days. These showed multilamellar and aggregated structures by transmission electron microscopy (TEM), an increase in size by dynamic light scattering (DLS), and loss of proteins into the storage buffer.¹ Similar changes are well known to occur when liposomes are frozen.

The storage buffer can also contribute. When phosphate-buffered saline (PBS) freezes, disodium hydrogen phosphate crystallises and the pH of the remaining liquid can shift by around four units.² It is not yet known whether this pH change is large enough to damage EVs. However, exosomes stored at pH 4 for 24 hours showed a much greater loss in concentration than those stored at pH 7,³ and PBS remains the most widely used storage buffer for EVs.

Fusion is more difficult to demonstrate than aggregation. In one study, EVs labelled with GFP and EVs labelled with mCherry were mixed before storage. After storage, a significant population of particles carried both labels, suggesting that fusion had occurred.⁴

How are EV characterisation methods affected?

The table below summarises the changes that have been reported for the methods most widely used to characterise EVs. Most studies agree on the direction of change, but the size of the effect varies with the source of the EVs, their purity and the storage buffer.

Particle concentration. In a systematic study, plasma EVs were taken through three freeze-thaw cycles, each separated by a week at −80 °C. Particle concentration, measured by tunable resistive pulse sensing (TRPS), decreased significantly after the first cycle but did not decrease further after the second or third.⁴ In contrast, exosomes from HEK293T cells showed a progressive loss of particles over five cycles, together with a decrease in the exosome markers ALIX, HSP70 and TSG101.³ Other studies have found little or no effect. Neutrophil-derived EVs, counted by calibrated flow cytometry, showed no significant change in number or size after storage at −80 °C,⁵ and repeated freezing to −20 °C did not change the size of exosomes from three human cell types measured by nanoparticle tracking analysis (NTA).⁶

As these results vary between studies, no single correction can be applied. It is important that samples which are to be compared have been through the same number of freeze-thaw cycles, and that fresh samples are measured where possible.

Size and aggregation. NTA, TRPS and DLS all measure a diameter that approximates the hydrodynamic size of a particle. If four 100 nm vesicles aggregate, they are measured as one larger particle and the particle count decreases. An increase in mean size and a decrease in concentration may therefore both be caused by aggregation. In the airway exosome study, the mean diameter measured by DLS increased from 94.5 to 125 nm after a single freeze, with aggregates of up to 400 nm. The same change was seen whether samples were thawed slowly on ice or rapidly at 37 °C.¹ DLS is weighted by intensity and so overestimates the contribution of large particles, and a smaller shift would be expected by NTA. Neither method can distinguish aggregation from fusion. This requires single-vesicle analysis or an approach such as the dual-label study described above.

Zeta potential. Freshly isolated airway exosomes had a zeta potential of −34.8 to −32.4 mV. After freezing at −80 °C and thawing, this fell to between −16.5 and −9.88 mV.¹ At these values there is little electrostatic repulsion to prevent particles aggregating. Plasma EVs stored as purified isolates also showed a shift in their size-charge profiles.⁴ Where zeta potential is used as a quality attribute, for example in the specification of a therapeutic EV product, reference and test samples should have the same freeze history.

Protein markers and the proteome. Total protein concentration may not detect changes in EV protein content. The protein concentration of plasma EVs did not change after two freeze-thaw cycles.⁴ However, in the airway exosome study, 315 proteins were depleted from exosomes after storage at −80 °C. Of the 31 proteins that were lost specifically as a result of freezing (compared with storage at 4 °C), 29 were found in the storage supernatant, suggesting that they had leaked from damaged or fused vesicles.¹ Luminal markers such as ALIX and TSG101 are therefore more likely to be affected than total protein, and surface markers may also be altered. Using a multiplexed antibody array, EV marker signals in heparin plasma and serum were found to differ significantly between samples frozen and thawed once and those frozen and thawed twice.¹¹ When EVs are quantified using surface marker ELISAs such as ExoLISA™, standards, controls and samples should all have been through the same number of cycles.

RNA cargo. The EV membrane provides some protection for RNA. In pooled plasma, total RNA decreased markedly after a single freeze-thaw cycle, whereas RNA from the exosome fraction was less affected. Exosomal RNA was, however, significantly lower after two cycles.⁷ This is important for studies that compare RNA from whole plasma with RNA from EVs isolated from the same samples.

Function. Changes in EV function may occur without any change in physical properties. Neutrophil-derived EVs stored at −80 °C showed no change in number or size, but their antibacterial activity was only partly retained over 28 days. The authors recommended that EVs should be stored at −80 °C for no more than seven days where storage is necessary, and that freshly prepared EVs should be used for functional studies.⁵ Freezing has also been shown to change the uptake of HEK293T exosomes by cells, with more single vesicles taken up after storage at −80 °C than after storage at 4 °C.³ This suggests that freezing can alter the properties of the EV membrane.

Does it matter whether the biofluid or the isolated EVs are frozen?

Studies of EV storage are easier to compare when they are grouped according to what was frozen: whole biofluid, partly processed biofluid, or purified EVs in buffer. Each is affected in different ways.

Plasma and serum. In blood samples, residual platelets are the main source of freeze-thaw artefacts. When plasma was frozen without first removing platelets, a single freeze-thaw cycle resulted in an approximately 20-fold increase in platelet-derived (CD41+) particles compared with platelet-poor plasma that had not been frozen. Centrifugation after thawing removed most of the remaining platelets, but around a third of the newly formed platelet particles and their associated miRNA remained.⁸ Another study found that the more centrifugation steps a sample had been through before freezing, the smaller the increase in particle counts after thawing. Freezing platelet-free plasma had no effect on platelet microparticle counts, although annexin V positive particles increased slightly.⁹ More recently, calibrated flow cytometry has confirmed that freezing plasma containing residual platelets generates EV subpopulations and cell-free mRNA that cannot be removed by processing after thawing.¹⁰

The quality of archived plasma for EV analysis therefore depends on how it was centrifuged before freezing. For many biobanked samples, this was done some years ago using protocols that were not designed for EV research.

Urine. Urinary exosomes are particularly sensitive to storage temperature. After one week, recovery of exosome-associated protein was 27% at −20 °C and 86% at −80 °C, compared with freshly processed urine. Vortexing for 90 seconds after thawing increased recovery to between 87 and 100%, and urine stored at −80 °C for seven months gave results comparable to fresh urine when vortexed.¹² This suggests that some of the apparent loss is due to vesicles being trapped in material that precipitates during freezing.

Purified EVs. Purified EVs appear to be more sensitive to freezing than EVs in biofluids. Plasma EVs stored as purified isolates showed a greater increase in particle size than EVs stored in plasma and isolated after thawing.⁴ A 2024 systematic review of 50 studies also concluded that storage in the native biofluid generally provided better stability than storage of purified EVs in buffer.¹³ It has been suggested that proteins present in the biofluid may act as cryoprotectants.¹

Can storage buffers and cryoprotectants help?

The strongest evidence for an improved storage method comes from a study that compared a number of storage buffers at different temperatures for up to two years, using EVs from several cell sources. Storage in PBS led to a large loss of EVs at all temperatures within days, particularly for highly purified samples, and dilution of EVs in PBS reduced recovery within minutes. PBS supplemented with human albumin and trehalose (PBS-HAT) improved both short-term and long-term storage at −80 °C, maintained EV stability over several freeze-thaw cycles, and improved recovery when used as a diluent.¹⁴ Part of this improvement is likely to be due to albumin reducing the binding of EVs to tube surfaces, which would otherwise appear as a loss of particles.

Trehalose alone has been shown to reduce aggregation of beta cell-derived EVs and to protect them during freezing and thawing.¹⁵ Trehalose also prevents the crystallisation of buffer salts in frozen PBS,² which may contribute to its protective effect.

However, not all studies have found cryoprotectants to be beneficial. In the systematic study of plasma EVs, none of the additives tested, which included trehalose, DMSO, glycerol and protease inhibitors, prevented the loss of particles or the increase in size.⁴ In the neutrophil EV study, commonly used cryoprotectants caused EV lysis.⁵ It is not yet clear whether these differences are due to the concentration of cryoprotectant, the source of the EVs or the methods used to measure them.

The choice of storage buffer can also affect downstream analysis. Albumin will interfere with total protein assays, increase background in proteomic analysis and may contribute to particle counts on sensitive instruments. A stabilising buffer is therefore better suited to EVs intended for functional studies or particle analysis than to those intended for mass spectrometry. Whichever buffer is used, it should be the same for all samples in a study.

Do freezing rate, thawing rate and storage temperature matter?

Storage at −80 °C is consistently better than storage at −20 °C. As well as the results for urinary exosomes described above, the exosome markers ALIX, HSP70 and TSG101 were found to degrade more slowly over 30 days at −80 °C than at −20 °C or 4 °C.³ Frost-free −20 °C freezers, which warm periodically during defrost cycles, may also expose samples to partial thawing.

The evidence on the rate of freezing and thawing is less consistent. Snap and slow freeze-thaw cycles produced no difference in plasma EV size or concentration,⁴ snap-freezing did not improve the preservation of neutrophil EVs,⁵ and the size of airway exosomes changed to the same extent whether they were thawed on ice or at 37 °C.¹ The 2024 systematic review, however, concluded that rapid freezing and constant storage at −80 °C gave the best preservation of EV quantity and cargo.¹³ Until this is resolved, all samples in a study should be frozen and thawed in the same way and kept on ice after thawing.

Good practices for storing EVs

Where samples must be stored, they should be characterised after storage as well as before. Particle concentration measured on the day of isolation may not reflect the material that is thawed months later for an experiment. Analysis of matched fresh and thawed aliquots using an NTA size profiling service provides a simple way to measure how much a preparation has changed.

Is freeze-drying an alternative?

Freeze-drying (lyophilisation) avoids repeated freezing and thawing, although the drying process introduces its own stresses and a lyoprotectant is required. Melanoma-derived exosomes freeze-dried without a cryoprotectant formed aggregates. When trehalose was added, aggregation was prevented, protein and RNA were protected, and the activity of loaded cargo was retained for about four weeks at 25 °C.¹⁶ For EVs that need to be shipped, or stored for long periods without reliable cold storage, a validated freeze-drying process can provide more consistent material than repeated freezing and thawing. Cell Guidance Systems offers a freeze drying service as well as custom freeze dried EVs.

Reporting storage conditions

The MISEV2023 guidelines include post-isolation handling and storage as part of the experimental method. They provide recommendations for EVs from conditioned medium, blood, urine and other sources, while recognising that there is no universal storage protocol.¹⁷ Researchers should therefore report the storage temperature and duration, buffer, tube type, number of freeze-thaw cycles and thawing method alongside characterisation data.

Future directions

As EV research moves towards clinical biomarkers and therapeutics, so does the need to understand the effect of storage. For biomarker studies based on biobanked plasma, frozen samples are the only material available, and consistent handling may be more important than matching the circulating EV population. For mechanistic and therapeutic studies, the aim is to work with EVs as they are released by cells. It is not yet known whether some EV subpopulations are more resistant to freezing than others, or whether these are the populations of greatest biological interest.

Frequently asked questions

Does a single freeze-thaw cycle change EV characterisation results?
It can. In a systematic study of plasma EVs, most of the particle loss occurred after the first cycle, while particle size continued to increase with each further cycle.⁴ Other studies have found smaller effects. Samples that are to be compared should therefore have the same freeze history.

What temperature should extracellular vesicles be stored at?
−80 °C is recommended for both biofluids and isolated EVs. Storage at −20 °C results in much greater losses, particularly of urinary exosomes.¹² Even at −80 °C, EVs stored in PBS are lost over time, and buffers containing albumin and trehalose have been shown to perform better.¹⁴

Should plasma be frozen before or after EV isolation?
EVs are generally more stable when stored in their native biofluid than as purified isolates in buffer.^4,13 For plasma, residual platelets should be removed before freezing, as a freeze-thaw cycle of plasma containing platelets generates platelet-derived particles that cannot be fully removed afterwards.^8,10

Can EVs be refrozen after thawing?
Repeated freeze-thaw cycles are associated with lower particle counts, larger and more aggregated particles, loss of luminal proteins and reduced activity.¹³ Freezing EVs in single-use aliquots avoids the need to refreeze, and the number of cycles each sample has undergone should be recorded.

References

1. Maroto R, Zhao Y, Jamaluddin M, et al. Effects of storage temperature on airway exosome integrity for diagnostic and functional analyses. J Extracell Vesicles. 2017;6(1):1359478.
2. Thorat AA, Suryanarayanan R. Characterization of phosphate buffered saline (PBS) in frozen state and after freeze-drying. Pharm Res. 2019;36(7):98.
3. Cheng Y, Zeng Q, Han Q, Xia W. Effect of pH, temperature and freezing-thawing on quantity changes and cellular uptake of exosomes. Protein Cell. 2019;10(4):295-299.
4. Gelibter S, Marostica G, Mandelli A, et al. The impact of storage on extracellular vesicles: a systematic study. J Extracell Vesicles. 2022;11(2):e12162.
5. LÅ‘rincz ÁM, Timár CI, Marosvári KA, et al. Effect of storage on physical and functional properties of extracellular vesicles derived from neutrophilic granulocytes. J Extracell Vesicles. 2014;3:25465.
6. Sokolova V, Ludwig AK, Hornung S, et al. Characterisation of exosomes derived from human cells by nanoparticle tracking analysis and scanning electron microscopy. Colloids Surf B Biointerfaces. 2011;87(1):146-150.
7. Ge Q, Zhou Y, Lu J, et al. miRNA in plasma exosome is stable under different storage conditions. Molecules. 2014;19(2):1568-1575.
8. Mitchell AJ, Gray WD, Hayek SS, et al. Platelets confound the measurement of extracellular miRNA in archived plasma. Sci Rep. 2016;6:32651.
9. Chandler WL. Microparticle counts in platelet-rich and platelet-free plasma, effect of centrifugation and sample-processing protocols. Blood Coagul Fibrinolysis. 2013;24(2):125-132.
10. Kim HJ, Rames MJ, Tassi Yunga S, et al. Irreversible alteration of extracellular vesicle and cell-free messenger RNA profiles in human plasma associated with blood processing and storage. Sci Rep. 2022;12. doi:10.1038/s41598-022-06088-9.
11. Bæk R, Søndergaard EKL, Varming K, Jørgensen MM. The impact of various preanalytical treatments on the phenotype of small extracellular vesicles in blood analyzed by protein microarray. J Immunol Methods. 2016;438:11-20.
12. Zhou H, Yuen PS, Pisitkun T, et al. Collection, storage, preservation, and normalization of human urinary exosomes for biomarker discovery. Kidney Int. 2006;69(8):1471-1476.
13. Ahmadian S, Jafari N, Tamadon A, et al. Different storage and freezing protocols for extracellular vesicles: a systematic review. Stem Cell Res Ther. 2024;15. doi:10.1186/s13287-024-04005-7.
14. Görgens A, Corso G, Hagey DW, et al. Identification of storage conditions stabilizing extracellular vesicles preparations. J Extracell Vesicles. 2022;11(6):e12238.
15. Bosch S, de Beaurepaire L, Allard M, et al. Trehalose prevents aggregation of exosomes and cryodamage. Sci Rep. 2016;6:36162.
16. Charoenviriyakul C, Takahashi Y, Nishikawa M, Takakura Y. Preservation of exosomes at room temperature using lyophilization. Int J Pharm. 2018;553(1-2):1-7.
17. Welsh JA, Goberdhan DCI, O'Driscoll L, et al. Minimal information for studies of extracellular vesicles (MISEV2023): from basic to advanced approaches. J Extracell Vesicles. 2024;13(2):e12404.

Cell Guidance Systems helps researchers isolate and characterise EVs with Exo-spin™ purification, EVlution™ TFF for concentrating large volumes, ExoLISA™ quantification and a range of EV and exosome services, including NTA size profiling and freeze drying, that can help accelerate your research with confidence.

Main Image: Ice crystals  Credit: Maxim Bilovitskiy - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=45343144

Learn more about powerful technologies that are enabling research: