I will start my “Protein crystallography rescue strategies” series with a rational biochemical approach to determine the boundaries of domains.
Knife & needle (subcloning)
Most try to crystallize their protein of interest as is, sometimes with an affinity tag fused to it, and as full length wild type protein. While this is a biologically sensible thing to do (as long as we assume the protein’s coding region dosen’t undergo post-translation modifications), this may not be the most crystallography sensible thing to do. How come?
Proteins are modular entities, which may contain several domains (large proteins) or parts which are flexible, both may impede crystal growth. This is one reason why crystallographers will usually clone the full length protein and its subclones in parallel. However, how can one know where one domain starts and when it ends? Under most circumstances, a priori knowledge of the protein’s domains boundaries is limited to secondary structure predictions, homology modeling and other bioinformatics tools discussed in earlier post. Utilizing a biochemical approach, such as limited proteolysis coupled to N-terminal sequencing or mass spectrometry, can yield empirical information from the yet unknown protein structure. Briefly, the protocols follows these steps:
- Mix full length protein (10-20 ul, 50-100 ug) with different ratios of proteolytic enzymes (trypsin, chemotyrpsin, subtilisin etc.). Recommended starting concentration can be 1:500 molar ratio and then serially dilute six or more dilutions.
- Incubate at room temperature for 60 minutes and then quench the reaction by addition of sample buffer.
- Resolve products on a SDS-PAGE. Since the full length protein will not be the focus of this analysis, prepare a SDS-PAGE which aims at 50% of the protein’s full length molecular weight. For example, if you’re protein is 70 kDa, use a gel density fitting 35 kDa, which is 15%. Keep ~10 ul of each sample for later runs (see below).
★ Note that you can use a single protease concentration and every 5-10 minutes remove a sample and quench it for later analysis.
What should you expect to see in the gel?
You should be looking for a strong band developing over time – there might be several of does, hinting toward several stable domains. Something like this:
You can see here that one major domains is detectable, at ~34 kDa.
If the protein doesn’t have a core domain or it is very dynamic it will be exposed to the enzyme throughout the proteolytic process and thus the band pattern and strength will be even more or less, not revealing a core domain.
Once you determine the stable domain(s), you can repeat run the sample again (have you kept some leftovers) and then blot the gel onto a PVDF membrane. Once you stain the blot, you can mark the band(s) and subject them to N-terminal sequencing or Edman degradation. N-terminal sequencing can determine the 5-6 first residues of the fragment and thus you determine the starting residue of the fragment. But what about the end of the fragment? Here you should run the leftover sample on a gel, remove the relevant band and subject it to peptide fingerprinting via tandem mass spec (MS/MS) analysis. Through this analysis you will obtain a list of peptides that originate from that specific band so now you can build back the segment and determine the last residue.
How do you make rational biochemical decisions?
★ When was the last time you organized your fridge or freezer? Wouldn’t it be nice if you could find things quickly? Try the brand new Labguru v2.0 lab management online system – it will increase your working efficiency!