Well, since I mentioned just the other day (and not for the first time) that determining drug concentrations and localization in cells is a major unsolved problem, I should probably talk about this new paper (a collaboration between groups in Edinburgh and Glasgow, nice to see since their cities are not always collaborative in all fields!) It highlights the technique of stimulated Raman scattering (SRS) microscopy (here’s a review on the topic). In this case, they’re looking at the intracellular distribution of the tyrosine kinase inhibitor ponatinib, which has an internal alkyne that’s particularly useful for Raman work, since its vibrational frequency is out in a window where the cell background is low. Spontaneous Raman scattering is pretty weak, but the stimulated variety uses two wavelength (a “pump” beam and a Stokes beam), and you tune those based on the functional group you’re targeting to provide a much higher signal.
One complication (or feature, depending on your outlook) is that metabolic changes to a parent molecule may not have much of an impact on the Raman behavior of the alkyne functional group, so you may well be looking at more than one species by SRS imagining. In this case, the authors used DFT calculations on the known metabolite structures (N-oxide, N-desmethyl) as well as the plain protonated form of ponatinib and determined that they would likely be seeing all of them in their single-wavenumber experiment, but it has to be noted that LC/MS showed that only very small amounts of these species were produced in their cell culture experiments, anyway. A previous study of this sort on the kinase inhibitor neratinib had more complications with this sort of thing. That one was using spontaneous Raman, not stimulated as in this case, so the image acquisition time also changes from about a half hour to around 45 seconds – no bad thing.
At right are some of the results. The top two rows show SRS microscopy at a wavelength that images protein background versus the wavelength for the alkyne in ponatinib, and the lower two images are the same experiment done in a ponatinib-resistant cell line. And sure enough, there are concentrated spots of compound apparent: but what are they? Experiments with “Lysotracker” dyes indicate that the compound is (as one might predict) being sequestered in lysosomal compartments – or at least some sort of acidic organelle, and lysosomes are the best candidate. And it would appear that the resistant cells are dealing with the drug, at least partly, by increasing their lysosomal capacity. Proteomic analysis for lysosomal markers was consistent with that idea as well. Further experiments investigated treatment of the cells with chloroquine (a known inhibitor of lysosomal processing), and sure enough, this disrupted the pattern seen in the images shown here. Unfortunately (at least to my eyes), what it also did was decrease the SRS signal in the cells pretty drastically overall – apparently, once the compound was not piling up in the lysosomes, the signal/noise for plain ol’ cytoplasmic distribution is not very high. But that chloroquine treatment also increased the sensitivity of the cells to ponatinib in general, which is what you’d expect. There are quite a few drugs whose efficacy is thought to be impaired by such lysosomal partitioning, and it’s certainly nice to see it demonstrated so clearly.
So if the sensitivity of this technique can be further increased, there are some interesting possibilities. Now, not every drug has an alkyne handle on it that gives you instant Raman signals, but there are other possibilities, including deuteration. And odds are that adding an ethynyl or propynyl tag to an existing molecule is probably something one could get away with most of the time, after some experimentation. Worth keeping an eye on!