Does media choice matter for peptide library purification using solid phase extraction?

Unbiased peptide screening technologies have evolved, improved and recently gained acceptance by the community as strategies for identifying new lead compounds against old targets with in novel chemical space.  As a result, the need for small scale, high throughput peptide synthesis has increased dramatically.  Post-synthesis handling of peptide libraries introduces a level of manual intervention distinct, and often much more annoying, from that encountered during larger batch syntheses.  Herein we demonstrate that media choice during solid phase extraction (SPE) work up for peptides greatly impacts success or failure for this purification strategy. Purification is generally perceived as the most significant bottleneck in the peptide synthesis production workflow.  Many technologies have been leveraged to reduce this bottleneck, but when synthesizers can produce hundreds of compounds in a single synthesis run, even the highest tech mass-directed, fully automated purification system-mediated process adds days to weeks to the production timeline.  The more relevant question, for peptide libraries in particular, is "how pure do the samples really need to be for reliable assay data?"

Many peptide libraries are screened as crude samples, suggesting that HPLC purification simply isn't worth the resource investment required to purify hundreds of compounds for a simple "yes" or "no" type screening assay.  However, there is significant anecdotal evidence that the ether precipitation work-up step is critical for the crude assay results to be reliable, suggesting that something should be done to the crude peptide samples before assays can begin. 

Ether precipitation, Figure 1, is a common workup step in the traditional peptide synthesis workflow, but it is not infallible.  There are some peptides, particularly very hydrophobic ones, that will not precipitate out of ether causing a headache, rather than a simple clean up step.  And when this occurs during processing of peptide libraries in a 96-well plate, those particular samples could be very difficult to identify, leading to sample loss.

Figure 1. Progression of crude peptide precipitation from ether after cleavage from the resin.  When performed in a conical tube as shown, the presence or absence of peptide precipitate is readily identifiable.

Recently, we have been working on a potential alternative to ether precipitation that delivers:

• similar (or higher) purity,
• can be performed in parallel,
• and ideally can also be automated
  
  

It's important to note - protecting groups, scavengers, and other impurities captured in an ether precipitation behave VERY differently than the majority of peptides during standard reversed phase chromatography.  With this in mind, it became clear that solid phase extraction (SPE) with a C18 functionalized media, Figure 2, could potentially meet all of the above criteria during our preliminary evaluations. 

Figure 2. Graphical representation of the steps performed during a solid phase extraction protocol.

While SPE is a commonly used strategy in other scientific fields, it's not commonly used in the peptide workflow, although there is some literature precedence.  It just gave me a chance to do some experiments though.  During those proof of concept experiments the following criteria were established

1. peptides must elute from the SPE media in a predictable manner
2. peptides must elute from the SPE media efficiently (high recovery)
3. SPE media must be readily available in amounts relevant to synthetic chemists (mg loading capacity rather than ng-μg loading capacity)

With these criteria in hand, a series of commercially available, C18-functionalized medias were tested with a variety of purified peptides, only one of which is highlighted here, Figure 3.  In these experiments, 2%, 5% or 10% (w/w) of an 18 amino acid peptide was loaded onto each of 3 different C18-functionalized SPE media, washed, and eluted using 1500 μL of 65% Acn - a concentration ~5% greater than the required elution concentration from analytical HPLC.  The media was then treated with an additional 1500 μL of 70% Acn to elute any residual peptide from the sorbent. 

Figure 3. Elution behavior of an 18 amino acid peptide when loaded onto three commercially available, C18-functionalized SPE media.  The peptide elutes from C18 media in analytical HPLC at approximately 60% Acn as is therefore expected to be contained in the 65% Acn fraction when eluted from the C18 SPE media.  Any peptide present in the 70% Acn fraction would be lost during this workup. 

What is clear with this data is that despite the fact that all the media types tested have the same functionalization (end-capped C18), very different elution and retention behavior was observed. This is likely due to the specific parameters, pore size and particle size, of the silica particles to which the C18 chains are added.  Importantly though, the PeptiRen C18 media met all the above criteria set before the evaluation began.  

The "Standard" SPE media (60 Å particles, 60 μm pores) still retained a signification portion of the total sample load.  Which was not surprising given that the peptide evaluated here contains 18 amino acids and is generally elongated in solution.  Other previous experiments have indicated that peptide is lost in other fractions with this media as well, making it less predictable and lower yielding (data not shown)

The wide pore media (30 Å particles, 300 μm) though, also held onto the peptide sample.  This is was less expected given it's improved performance for peptide purification in more traditional column chromatography purifications.

This data laid the groundwork for validating that SPE, specifically C18-mediated SPE, could be a viable strategy for post-cleavage workup for peptide libraries in parallel.  It has also clearly demonstrated that not all media behaves the same, so make sure you check the particle parameters before you get started exploring this new and exciting strategy.

Have you struggled with post-synthesis handling of your peptide libraries?  Follow the link below to learn more about this and other automatable strategies.