Extracellular vesicles (EVs) are nanosized lipid bilayer vesicles which contain an assortment of proteins, nucleic acids and small molecules (Figure 1). As the contents of EVs are impacted by the internal (e.g., genetic) and external (e.g., disease) environment of a cell, they have great diagnostic potential, offering clinicians a window into cells and tissues. In the search for EV diagnostics, size exclusion chromatography (SEC) and polymer precipitation are two common EV isolation methods. But which is better? To help you navigate this choice in a data-driven way, we break down the data. Turning to the literature to inspect EV isolate yield and purity, we also look at the all-important factors of cost and time – practical considerations which have particular importance when scaling up a diagnostic test.
How does size exclusion chromatography enable extracellular vesicle isolation?
As the name suggests, SEC isolates EVs based upon the principles of size. Whilst SEC columns generally look like a clear column partially filled with a solid white-ish resin, the resin is actually made up of many, many tiny beads. After flushing out the storage buffer, a crude sample containing EVs and proteins (among other things) is loaded into a SEC column. The sample then passes into the resin and meets the resin beads. These beads are honeycombed with pores that are – as long as you pick a resin suitable for EV isolation – intentionally too small for EVs (at least the EVs we currently know about!), but plenty big enough for the vast majority of proteins. As your sample runs through the columns, proteins enter into the beads and become delayed in the labyrinth of pores. The EVs, on the other hand, can race ahead as they have a straight route for the exit, meaning that they leave the column before proteins do. In this way, EVs can be collected separately from proteins, meaning that EVs can be isolated in a theoretically pure manner.
EVs isolated using SEC are suitable for any downstream application as they are isolated whole under the force of gravity (rather than the extreme forces of ultracentrifugation) and, unlike with precipitation, are free from contaminants introduced in the isolation process.
How does polymer precipitation for extracellular vesicle isolation work?
Polymer precipitation can be done in a ‘do it yourself’ manner with a protocol utilising polyethylene glycol (PEG) or by using commercial kits. Two of the most popular commercial polymer precipitation brands are ExoQuick (System Biosciences) and Total Exosome Isolation Reagent (Invitrogen), which have been used in several comparison studies with SEC (and, specifically, our qEV columns). Polymer precipitation works by forming a hydrophobic polymer surrounding the EVs, excluding water from the tangle of polymers, and allowing for the pelleting of EVs and proteins caught within. As this process is not size-dependent or dependent on any other EV parameter, how will this affect yield, protein contamination and purity? We turn to the literature for answers down below.
Importantly, Gámez-Valero et al (2016) and others have found that EVs are altered by the precipitation process.1 For example, they have been shown to be cytotoxic, making them unsuitable for functional studies.1 Additionally, precipitation may be ill-advised for diagnostic studies, as Yang et al (2021) found the RNA from precipitation EV isolates contain high levels of non-EV RNA contamination, likely from protein-bound RNA.2 As such, protein and RNA contamination from non-EV sources may be an issue for precipitation EV isolates.
How does EV yield and purity compare for EVs isolated using SEC versus polymer precipitation?
The best way to see how both methodology types match up was to see how they perform in the field. For this, we used open access studies which directly compared SEC and precipitation for plasma EV isolation. Where numerical data was not available, this was extracted from figures using PlotDigitiser.
Firstly, we looked at whether there were any statistical differences between the different brands or ‘do it yourself’ versions of each method. The reasoning here was to give each method the best chance of succeeding by removing any poorly performing variations. When it came to SEC, qEV columns were ~14 times purer than other SEC columns (p=0.02). Due to this, we used only qEV columns to represent SEC for the rest of the analyses. However, whilst ExoQuick appeared to result in slightly purer isolates than PEG and other precipitation brands, there was no significant difference. Therefore, all precipitation brands were kept in the analysis. This left us with 6 studies to compare.1-6
How does EV isolation method impact upon yield?
Next, we dove into the data and took a look at yield. As you can see from Figure 3, the yield of precipitation appears higher, however this is not statistically significantly different (p=0.7). If yield is the most important parameter for you, then this might (initially) make precipitation a more favourable method. Before making that decision, though, let’s take a look at other important factors in picking an EV isolation method.
How does EV isolation method impact upon protein contamination?
Next, we assessed the protein content of isolates. This was in non-lysed isolates, so the protein signal here is overwhelmingly from contaminating protein, not from EV-associated protein. As you can see in Figure 4, there is a significantly higher protein concentration in isolates from precipitation methods than from SEC.
How does EV isolation method impact upon EV isolate purity?
By minimising protein contamination, the use of SEC results in a significantly purer EV isolate (Figure 5). In fact, the data here shows that SEC using qEV columns results in EV isolates which are around 30 times purer than isolates from precipitation methods.
If purity of your EV isolate matters, which in essentially every circumstance it will, then methods favouring purity should be considered above those resulting in less pure isolates.
Time comparison of SEC and polymer precipitation for EV isolation
In diagnostics especially, time is important. The faster EVs can be isolated, the more diagnostic tests can be performed in a given time period. To test the time commitment, we measured how long it takes to run our most used column for plasma (i.e., qEVoriginal / 70 nm) both manually and using the Automatic Fraction Collector (AFC) over 4 repeats. It took a median time of 36 minutes 12 seconds for manual collection (without the AFC) and 18 minutes 25 seconds when using the AFC. This includes all the pre-flushing of columns and the cleanup at the end, as well as the actual EV isolation (note that this reflects the current qEV isolation platform; in future, high-throughput clinical settings, columns would be single use and multiple samples would be isolated in parallel – drastically reducing the time commit even further).
To calculate how long precipitation takes, we looked to the most popular plasma EV precipitation kit: the ExoQuick Ultra for Serum/Plasma. We estimated from the manual that this would take around 80 minutes, taking into account the steps with timings and allowing 1 minute per step where pipetting was needed or the user needed to set up a centrifuge.
Taking this into account, SEC using qEV columns is around 4.5 times faster than using polymer precipitation.
Cost comparison of SEC and polymer precipitation for EV isolation
Two of the most popular commercial polymer precipitation brands are ExoQuick (System Biosciences) and Total Exosome Isolation Reagent (Invitrogen). Respectively, their plasma kits cost ~US$35 and ~US$25 per sample, with the Invitrogen kit being calculated based on a 500 μL sample. For reference, using the qEVoriginal Gen 2, which is optimised for a sample loading volume of 500 μL, costs less than US$10 per sample, or even less for bulk orders.
This alone makes SEC using qEV columns currently about 3 times cheaper than using precipitation – and doesn’t take into account the loss of time and resources that could result from basing your findings on impure EV isolates (both for your own studies and any subsequent studies from others).
Should you use size exclusion chromatography or polymer precipitation for extracellular vesicle Isolation?
In this article we have taken an unbiased look at how SEC and polymer precipitation compare for isolating EVs. Precipitation did win on the basis of yield, so if that is your only consideration, then precipitation might be the way to go. However, with SEC using qEV columns being faster, cheaper and isolating purer EVs than polymer precipitation, we think that SEC should be your go-to choice for EV isolation for diagnostics. Ultimately though, you have to decide which method is best suited for your application and circumstances. Hopefully, the information presented here will help you to make the decision that is right for you.