single molecule FISH (smFISH) is a great way to detect single RNA molecules in fixed cells. The “traditional” FISH uses fluorescently labeled oligonucleotides which directly hybridize with the target RNA sequence. The two most common approaches are the use of 1-5 50-mer oligos, that are labeled with 5 fluorophores, or the use of more than 20 20-mer oligos labeled at one or both ends.
A more recent approach is nicknamed “Christmas tree”, in which 2-3 rounds of hybridizations are done, with only the last round uses fluorescently labeled probed. I’ve expanded on that when I discussed RNAScope. Other people have independently developed similar protocols (e.g. branched DNA FISH (bDNA), which was recently used to get smFISH transcriptomics in human cells).
Now, a new paper came out with a very similar method whish they call FISH-STIC (Fluorescence In Situ Hybridization with Sequential Tethered and Intertwined ODN Complexes).
However, their images do not compare to the quality shown by the RNAScope or bDNA papers. Maybe its just bad imaging and not the FISH itself, but their images look blurry and doesn’t seem to be up to modern FISH standards. I used 20-mer FISH and can get nicer images.
They also get a few high-intensity spots which the call “mRNA-independent” (since they can also see it outside the cell). To me it means that their protocol was just not optimized – either the washes, or the oligo sequences which should not make complexes like that.
They also seem to have higher background fluorescence compared to RNAScope.
Simultaneous detection of Actb and Actg mRNAs with FISH STIC probes. Source: Sinnamon & Czaplinski (2014) RNA 20:260-266.
On the other hand, their beta-actin FISH-STIC looks better than the RNAScope-FISH. beta-actin mRNA is found in 500-2000 copies per cell, so it is not easy to get smFISH. Still, 20-mer FISH that I do looks better than both.
20-mer FISH for beta-actin mRNA in immortalized MEFs. Source: myself.
Notice also in my image the transcription sites (TS) are clrearly visible, whereas I did not see any TS in their images, which is weird (unless they chose cells without TS on purpose. Don’t know why, since smFISH is a powerful tool to quantify transcription 1, 2).
They did not test FISH-STIC on a less prevalent mRNA, so I cannot comment on how it would look like compared to RNAScope or 20-mer FISH.
Anyway, the methods seem similar, but I think that the RNAScope demonstrates better results than the FISH-STIC. However, I haven’t tried this approach myself. I continue to do FISH with 20-mer probes and so far pleased with my results.
Sinnamon JR, & Czaplinski K (2014). RNA detection in situ with FISH-STICs. RNA (New York, N.Y.), 20, 260-266 PMID: 24345395
Battich N, Stoeger T, & Pelkmans L (2013). Image-based transcriptomics in thousands of single human cells at single-molecule resolution. Nature methods, 10 (11), 1127-33 PMID: 24097269