There’s no simple resource I can think of that covers what to use when, but here’s a basic overview that hopefully addresses some questions people might have if they’re thinking about how to study protein-protein and/or protein-nucleic acid interactions. 

note: this is a kinda weird post but it’s basically a response I wrote to someone studying, so thought I’d share in case anyone else might find it helpful. video added 1/27/22

Basically your strategy is going to depend on what your goals are and what your study system is – are you looking at what’s happening in cells or in vitro with purified components?

Your basic strategies include…

For studying protein/protein interactions…

  • in cells:
    • pull-down assays with western blot
    • mass spectrometry, often aided by biotinylation (i.e. BioID), cross-linking, or similar
    • yeast 2-hybrid or similar
  • in vitro w/purified components:
    • analytical size exclusion chromatography
    • native PAGE
    • depending on protein size & ease of purifying, something like SPR or ITC – these methods will allow you to measure affinity (both) and binding kinetics (SPR)

for studying nucleic acid/protein interactions

  • in cells: 
    • cross linking followed by sequencing (e.g. ChiP-seq or similar)
  • in vitro w/purified components
    • slot-blot filter binding assay
    • EMSA
    • SPR, MST, or other biophysical techniques
    • footprinting 

note that there are lots of variations of these basic strategies and there’s no one right way – so there won’t be one “right answer” 

Ideally, once you know 2 things interact, you have some idea of where they’re interacting – then you can mutate that place (on either or both thing, ideally you do each and show they have the same effect) and test the binding. If you were correct about where they were interacting, the affinity should now be greatly reduced. 

For example, if you think a protein is binding specifically to a DNA sequence, if you mutate that sequence, you shouldn’t see binding (or at least it should be greatly reduced). And if that binding really is specific, you shouldn’t be able to easily out-compete it if you toss in a bunch of random generic competitor. If you don’t know where a protein is binding, you can do things like ChIP-seq or similar to crosslink the protein to the DNA and sequence the DNA it’s bound too. If you have a region you know the protein binds but don’t know where exactly you can do things like DNAseI footprinting, where you look to see which areas of the DNA are protected from nuclease chewing, telling you where in the sequence the protein is bound. 

Do you have an idea where on the protein is doing the binding? (maybe from a structure or modeling or past research or something). If so, mutate those amino acids and see if it affects binding (but also make sure your protein still folds and functions normally in other aspects because if your mutations just make your protein a clump it might not bind the DNA but that might not tell you those amino acids were involved in the binding directly). If you don’t know where on the protein the binding is happening, you can try truncating the protein to find the minimal required section for binding (making sure to check for proper folding still, etc. – a good sign is a sharp peak in size exclusion chromatography).  If you’re in a high-throughput assay type lab you could do an alanine mutagenesis scan where you mutate the residues to alanine one-by-one to see what changes impact binding. You could also try cross-linking the DNA to the protein and using mass-spec to try to identify the interacting residues. 

note: if you think that charge is important, instead of mutating charged residues to alanine, which is also going to change the size a lot, you can mutate them to neutral but similar size alternatives (e.g. asparagine, glutamine) or swap the charge to the opposite charge. Also, if you think 2 oppositely-charged residues are forming a salt bridge (charge-charge attraction) that’s important, after swapping individually, you can try swapping the charge on both to see if that restores things.  

for getting and purifying the molecules

  • proteins
    • recombinant protein expression  
      • clone the cDNA for the protein into a plasmid and get into expression cells
        • can use site-directed mutagenesis methods (e.g. SLIC, QuikChange) to introduce amino acid substitutions
      • expression systems include bacteria (easiest but can struggle with big and/or post-translationally modified proteins), yeast, insect cells (e.g. Sf9, Hi5), mammalian cells (e.g. HEK cells)
      • typically express with a tag (N-or C-terminal, usually N) that will aid in purification and/or IP, Western, etc. – common tags include StrepTag, HisTag, GST, HA
        • sometimes also add a fusion partner such as SUMO or MBP to aid with expression and solubility
        • tags are often followed by a protease (e.g. TEV) cleavage site so you can remove the tag once you’re done with them
      • purification typically involves a series of column chromatography steps (e.g. affinity chromatography taking advantage of the tag -> ion exchange chromatography (separate by charge) -> size exclusion chromatography (SEC, aka gel filtration) (separate by size)
      • check for purity using SDS-PAGE – want to minimize extra bands showing other proteins still present
      • typically involves trial & error & lots of optimizing to find optimal expression & purification conditions
  • nucleic acids
    • for short ones, we typically order them synthesized (note that DNA is much cheaper than RNA!)
    • for longer ones, 
      • for DNA we use PCR or have bacteria make them for us once we’ve cloned the sequence we want copied into a plasmid
      • for RNA, we use in vitro transcription (typically w/T7 polymerase making RNA copies based on a cloned piece of DNA)
    • we can label the 5’ end with radioactive phosphorus (P-32) (or a fluorophore but that’s bulkier and less sensitive) to track it in binding assays

Hope that helps a bit! 

for more about these techniques, search the blog and/or go to the techniques page:

Leave a Reply

Your email address will not be published.