RCA Protocol: FFPE Tissue¶
- Applicable samples: Formalin-Fixed Paraffin-Embedded (FFPE)
- Version: v3.0
- Last modified: 2026-08-18
This page is the FFPE delta, not a standalone protocol
Run the two canonical pages in order:
- Pre-treatment: FFPE — baking, deparaffinization, RNA retrieval with RVC, rehydration, re-fixation, digestion
- RCA Protocol — hybridization, ligation, RCA, post-fix
Everything below is the short list of what FFPE does differently. Until v2.0 this page carried its own full copy of both, which drifted from them; that copy is gone and this page now holds only what is genuinely FFPE-specific.
There is no longer an FFPE-specific reagent change
Until v3.0 this page's first item was RNase H added to the blocking mix. Both halves of that instruction are gone: the blocking step was deleted from RCA Protocol v6.0, and RNase H has no role on the direct-RNA route. See Where RNase H belongs. What remains below are two judgement calls, not recipe changes.
1. Prefer the overnight, low-concentration hybridization¶
Both conditions in Hybridization work, but the choice is less free on FFPE:
- 10 nM per probe, ~16 h at 45 °C — strongly preferred for FFPE. The low probe concentration is what suppresses nonspecific deposition, and that matters more here than on fresh frozen, where autofluorescence and background already cost signal-to-noise before the probes are added.
- 50 nM / 120 min remains available when the run has to finish in one day, but expect the background penalty to be more visible on FFPE than on FF.
2. Glycerol: keep the 10% default¶
RCA notes that fresh-frozen sections have historically run ~5% glycerol. FFPE should stay at 10% — FFPE yields fewer circles than FF, so the confinement that keeps rolonies compact and even is worth more, and the smaller rolonies are an acceptable trade.
Where RNase H belongs¶
RNase H is a cDNA-route reagent, and on that route it is a deliberate reaction step rather than an additive: it digests the RNA strand of the RNA:cDNA hybrid to expose single-stranded cDNA for the padlock probes, in the same mix as the probe hybridization at 37 °C — see RCA: cDNA §3 (0.4 U/μl, 30 min at 37 °C before the denaturation ramp) and the SNP variant built on it. FFPE run through the cDNA route gets its RNase H there, at full working concentration, and needs nothing added on this page.
Never put RNase H in a direct-RNA hybridization mix
On the direct-RNA route the padlock probe is hybridized to the mRNA itself, so the probe:target complex is an RNA:DNA heteroduplex — exactly RNase H's substrate. Adding RNase H to that mix cuts the template out from under the probe and destroys the nick SplintR is supposed to seal. This is why RNase H can live in the cDNA route's hybridization mix and nowhere on this one.
The 0.1 U/μl RNase H that FFPE carried in the old blocking mix (v1.x--v2.0) is therefore dropped rather than relocated. Its stated purpose — clearing RNA that is already in an RNA:DNA hybrid so probes compete against less background — does not survive inspection either: at room temperature, with no DNA probe present, there is almost no heteroduplex substrate in an FFPE section for it to act on. If FFPE background is later shown to need an enzymatic pre-clear, that is a new, testable step with its own controls, not a line item restored on the strength of the old one.
Next step: Signal readout -- PRISM Imaging or SPRINTseq Sequencing
Changelog¶
- v3.0 -- 2026-08-18 -- RNase H removed from the FFPE direct-RNA workflow. MAJOR: a reagent leaves the reaction, and anyone trained on v2.0 would still be adding it. Sections renumbered 1--2.
Two things forced the decision at once. RCA Protocol v6.0 deleted the blocking step that RNase H was riding in, and the question of where to put it instead answered itself: on the direct-RNA route there is nowhere it can go. The padlock probe is hybridized to the mRNA, so the probe:target complex is an RNA:DNA heteroduplex and RNase H would degrade the template SplintR needs — see Where RNase H belongs. The cDNA route already runs it properly, as a reaction step inside the hybridization mix at 0.4 U/μl (RCA: cDNA §3), which is 4x the concentration this page used and in the one place where its substrate actually exists.
The old rationale — that RNase H at room temperature clears pre-existing RNA:DNA hybrid so probes compete against less background — was never tested and does not hold up: with no DNA probe present there is almost no substrate. Removing it is therefore expected to be a no-op at the bench; it is filed as MAJOR because that expectation has not been measured. The evidence behind the parent change is summarised in RCA Protocol v6.0.
- v2.0 -- 2026-08-18 -- Reduced from a full standalone protocol to the FFPE delta. MAJOR: the page no longer contains runnable steps, so anyone following it has to be redirected.
This page carried its own copy of the FFPE pre-treatment and of the RCA workflow. Both had drifted from the canonical pages: the pre-treatment predated Pre-treatment: FFPE v2.0 (RVC, no NaBH₄ pre-quench, added re-fixation), and a diff of the RCA half against RCA Protocol v5.6 showed ~90 differing lines of which only two were substantive — RNase H in the blocking mix, and the FFPE-specific reason to prefer the overnight hybridization. Everything else was duplicated prose that had aged apart.
Keeping two copies in sync was the failure mode, not the fix. Both substantive
deltas are preserved here (plus the glycerol note, which was previously implicit),
the duplicated text is deleted, and status: historical from v1.3 is dropped — the
page is current again, just much shorter. Full previous text is in git.
- v1.2 -- 2026-08-10 -- In line with the unified RCA protocol v5.5: post-amplification fix (Section 11) pinned to 30 min, from the previous 15--30 min. Formaldehyde's methylene-bridge crosslinking takes hours to complete, so 15 and 30 min both sit on the rising part of that curve and anchor measurably differently; the fix has to hold the rolonies through the 65% formamide strip that follows, so err long.
- v1.1 -- 2026-07-29 -- Brought in line with the unified RCA protocol v5.4. Hybridization (Section 8): 10 nM per probe overnight is now preferred (lower nonspecific background), 50 nM / 2 h kept as the same-day option. Stringent wash: 20% formamide in 2X SSC is the new default, 10% retained as the milder fallback. RCA (Section 10): primer lowered 0.3 → 0.1 μM, RiboLock removed (the template is DNA by this point), glycerol rationale documented, and a sample-dependent option to double dNTPs + amino-dUTP (e.g. TNBC) added. All PBS-T washes standardised to 2 min per change (previously 1 min). Corrected the RCA table, which previously summed to 99.5 μl instead of 100 μl. The NaBH₄ pretreatment in Section 5 is unchanged; the former post-RCA quench section now lives in Autofluorescence Quenching.
- v1.0 -- 2026-01-29 -- First FFPE release.