• Q: What are the compound selection criteria (potency, logD, etc) and ideal compound concentrations for CETSA® experiments?
    A:The CETSA® method can resolve potencies from mM to fM. In order to establish a melt and shift curve with a saturating concentration of a low potency compound, it need to have a very good solubility in saline buffers. As a rule of thumb we suggest to test at least 100x the expected XC50 concentration. For lysate experiments there is no particular demand on logD or permeability, only on solubility. For intact cell incubations compounds need to be permeable and stable in the intracellular matrix for the length of incubation.
  • Q: What compound properties are crucial to enable CETSA® experiments?
    A: For lysate experiments there is no particular demand on logD or permeability, only on solubility. For intact cell incubations compounds need to be permeable and stable in the intracellular matrix for the length of incubation.
  • Q: What would be the ideal target to run through CETSA®?
    A: The most readily ‘CETSA®-able’ proteins are cytosolic proteins that are less than 100 kDa in size. 
  • Q: Can you run CETSA® on compartmentalized- or membrane proteins?
    A: Yes. Nuclear or other compartmentalized proteins may need additional steps of specific extraction from the relevant compartments. Membrane proteins, protein complexes or multi-domain proteins may pose problems and need development of special buffers/conditions. These types of proteins are attainable for CETSA® experiments as long as it exists specific high affinity tools (i.e. antibodies). 
  • Q: Do certain target classes benefit from pre-fractionation (membrane or nuclear targets, for instance)?  If so, what data across target classes exists? 
    A: Proteins in organelles could either be fractionated prior to or after heating and this could be favourable if these are "hard to get" targets by normal lysis procedure. It is important that the fractionation does not affect the solubility of the target either in its folded or precipitated form.
  • Q: What types of plates are compatible or incompatible with CETSA® based on prior experience?
    A: We perform most of our experiments in PCR-strips (from Thermo Fischer), this is true also when we handle many samples as this is compatible with high speed centrifugation.
  • Q: What types of plates are compatible with the plate-based CETSA® assays?
    A: During the development of plate-based assays with dual epitope detection (for example AlphaLisa) we can use antibodies targeting natively folded protein and that way skip the centrifugation step. We can therefore perform all the "CETSA®-steps" in the same plate (i.e. compound incubation, heat challenge, lysis and detection). The caveat here is that if we include freeze-thawing in the lysis procedure then that treatment may skew the plate (something we are currently investigating). For our AlphaLisa assay we use Perkin Elmers ProxiPlate (#6008289) both for lysate experiments but also for intact cell experiments, where lysis through freeze-thaw is omitted (in which case, the lysisbuffer contains detergents and low salt).
  • Q: For intact cell experiments, is the suspension or cell pellet heated? Are there different results depending on the procedure?
    A: We heat suspensions. At the drug incubation step, we apply continuous rocking of the samples to keep the cells from pelleting on the bottom of the tube. This is mainly in order to make sure that the cells are better mixed with compound. In reality it shouldn’t affect the experiment. 
  • Q: How is the lysis buffer chosen for target proteins? For membrane proteins, what lysis buffer is recommended?
    A: When we lyse cells after heating we usually do it in the media in which we incubate the cells, normally HBSS. For proteins that precipitate at high temperatures we let the cells pellet prior to heating (for example by skipping the mixing during the last part of incubation, as discussed above). After heating we then remove most of the "incubation buffer " and switch to a harsher buffer (for example with less salt and maybe with the addition of mild, non-denaturing detergents) that makes the protein unfold more easily. For the membrane proteins, the latest paper by Savitski and colleagues at Cellzome is a good summary (Nature Methods 12, 2015). When working with kinases it could be a good idea to think about buffer choice, for example if using PBS, could the extra amount of phosphate be bound to the ATP binding site and hinder binding by compound? If so, then use for example kinase buffer (from Cell signalling) or a Hepes buffered saline.
  • Q: Are certain proteins/cell lines/tissues CETSA®-friendly and/or do not perform well in CETSA® experiments?
    A: CETSA® is in principle about; incubate (by different means) your cells with compound, aliquot, heat, separate and detect the remaining soluble protein. If any of these steps are made difficult, due to the choice of cell matrix, they will affect the experiment. We work with both suspension- and adherent cells and these work nicely most of the time. A potential problem could be if one needs to incubate adherent cells in a suspension so that the cells are affected by being detached from the plastic. One way to solve that problem is to treat the cells while they are still attached, and heat the vessel to a certain temperature in a water bath with the cells on the plastic. Alternatively, one can detach the cells chemically and run the risk of affecting the equilibrium that formed during incubation. For tissue samples the procedure is more challenging. Incubation of a large chunk of material could lead to insufficient exposure in parts of the sample. Too small samples could make it hard to fractionate it uniformly to provide enough material for detection. The heating step is normally not a problem, but when working with blood or blood-rich tissue it is advisable to dilute the sample or get rid of most/all IgGs and Albumin etc, these fall out above 55°C and cannot be pipetted. These proteins affect the gel running and they make western blot detection problematic, by attracting the secondary antibody used for detection. A solution to the latter problem is using a secondary antibody that only recognizes folded IgGs. 
  • Q: Have custom proteomes been generated for CETSA® experiments; have mixtures of cell-lysates been used for target identification?
    A: No. We have an idea of pooling the knowledge from all "control samples" ran on CETSA® MS, in order to sum up and get statistics on this data. In general, regarding proteomes, is that a full proteome covered in a database can be used to search for target engagement (TE) hits. I.E. one can work with any organism that has a sequenced proteome.
  • Q: Are there successful examples using CETSA® to investigate target engagement for biologics?
    A: There is unpublished data from academic sources that suggest CETSA® to work nicely with biologics. The exact way forward is not known but there are several ways possible.  
  • Q: How do you treat fragment spectra with missing reporter ion ratios (i.e. not all 10 reporter ions are present) in CETSA MS?
    A: We routinely run label checks on all samples prior to the actual analysis-run. That way we catch insufficiently labeled samples before they are injected.