Skip to content

Sample Pre-treatment: Fresh Frozen

  • Applicable samples: Fresh Frozen (FF) cryosections
  • Version: v6.2
  • Last modified: 2026-09-08

What's next?

After completing pre-treatment, proceed to the unified RCA Protocol (Hybridization → Ligation → RCA → Post-fix).

1. Sample Preparation & Sectioning

Goal: Obtain high-quality sections free of ice crystals and cracks, with firm adhesion to slides.

  • Embedding & Freezing: Use isopentane + dry ice for rapid freezing.
  • Cryosectioning: Blade: -20°C, Head: -10°C, Thickness: 10 µm.
  • Section Placement: Pre-chilled slide pickup -> finger-back press for adhesion -> return to cold stage immediately.
  • Storage: Store at -80°C.

2. Fixation

Goal: Brief thermal retrieval for adhesion, then chemical fixation.

  • Retrieve: Transport slides on dry ice.
  • Thaw & Adhere: Place immediately on 37°C hot plate for 1 min.
  • Fixation: Immerse vertically in Slide Mailer with 4% PFA (Fresh), incubate at room temperature (RT) for 30 min.
  • Wash: PBS wash 2 x 2 min.

Assemble Chamber: Assemble pre-treatment chamber to limit liquid mount.

3. Permeabilization

Goal: Open the fixed protein mesh enough for probes and enzymes to reach the RNA — while leaving the RNA anchored where it is.

  • Incubate: 0.1 M HCl at room temperature for 5 min. No protease.
  • Wash: PBS-Tween 0.05% wash 2 x 2 min.

Not yet confirmed at our bench — and half the original reason for it has been withdrawn

The pepsin step was removed in v6.0 in favour of acid-only permeabilization. Two things were behind that decision, and as of 2026-09-08 only one of them stands.

  • Stands: four independent padlock/RCA systems permeabilize fresh-frozen tissue with no protease at all, two of them the closest published chemistry to ours. That is a good reason to test acid-only, and it is why this page still specifies it.
  • Withdrawn: the suspicion that our own protease step was over-digesting. See Was the protease ever the problem? below. 0.4 mg/mL pepsin in 0.1 M HCl for 2 min is not an over-digested condition, and restoring it is a legitimate ladder arm, not a regression.

The two have still never been run side by side. Run a control section, or the digestion ladder below, before committing a precious sample.

Previous condition, for reference: 0.4 mg/mL pepsin in 0.1 M HCl, 37 °C, 2 min.

Why acid-only is worth testing

An in situ readout only needs probes to reach RNA that stays inside the cell. A capture-based method like Visium needs the RNA to diffuse out to the slide, which is where the original 0.1% pepsin / 10–30 min recipes come from. Conditions inherited from that lineage are systematically too aggressive for padlock + RCA: past the point where RNA is still anchored, the molecule is simply lost during the ~16 h hybridization, the formamide stringency washes and the overnight RCA.

Four independent padlock/RCA systems permeabilize fresh-frozen tissue with no protease at all — SCRINSHOT (0.1 M HCl, 3 min; the same direct-RNA SplintR chemistry as ours), HybISS (0.1 M HCl, 5 min), the commercial CARTANA ISS kit, and Xenium FF (1% SDS + cold 70% methanol). The acid window is also wide: routine FISH slide pretreatment uses 0.2 N HCl for 20 min at room temperature.

Full evidence, citations and the exposure arithmetic are in the internal evidence review Permeabilization / Pepsin Digestion (lab site → Research).

Was the protease ever the problem?

Probably not. v6.0 was triggered by a fresh-frozen TNBC run accidentally digested at 0.5 mg/mL instead of 0.4 (in 0.01 M HCl, ~38 °C, 2 min) that produced markedly fewer signal dots; over-digestion was the reading at the time. Two things have since undercut it.

  • A BZ09 re-run with reduced digestion pointed away from digestion. What the low-yield SPRINTseq runs from BZ09CA1-ReCA2 onward had in common was an RCA primer synthesised without its 3' phosphorothioate. phi29's proofreading exonuclease acts preferentially on single-stranded DNA, and our RCA step leaves free primer in the enzyme for 12-16 h with no anneal-and-wash, so fewer circles get primed and fewer rolonies form — the same symptom, from a different step. See the internal note The RCA Primer Needs a 3' Phosphorothioate (lab site → Research).
  • The exposure arithmetic never supported it either. At 0.4 mg/mL P6887 (≥3,200 U/mg) for 2 min the old condition was ≈ 2,560 U·min/mL, against ≈ 1,250 for a published fresh-frozen HybISS/CARTANA condition (Lawrence 2024: 0.5 mg/mL P7012 in 0.1 M HCl, 37 °C, 1 min) and a shorter but stronger ISS original (Ke 2013: 2 mg/mL in 0.1 M HCl, 90 s). Within roughly 2x of published fresh-frozen practice is not an outlier.

One confound in that TNBC run is not resolved by any of this: it ran in 0.01 M HCl, which was the reagent page's recipe at the time and 10x weaker than every protocol page said. That acid discrepancy is independent of the protease question and was closed separately in Pepsin Solution Preparation v3.0.

What this means at the bench: neither condition is ruled out and the deciding experiment has not been run. Follow the page as written, and settle it with the ladder when a spare block allows — with a 3'-protected RCA primer, or the comparison will be confounded the same way again.

If signal is low: titrate, don't just add pepsin back

Run the one-slide ladder in Permeabilization Optimization (Pilot) — 4–6 serial sections, one PDMS chamber each, every downstream step shared. That page also carries the in-process morphology criteria for judging digestion at the microscope, and the read-out metrics and decision rule for picking a winner.

4. Dehydration

Goal: Improve tissue permeability.

  • Pre-dehydrate: 80% Ethanol for 10 min.
  • Dehydrate: 100% Ethanol for 2 min.
  • Rehydrate: PBS-Tween 0.05% wash 2 x 2 min.

Next step: RCA Protocol (Hybridization → RCA)

Changelog

  • v6.2 -- 2026-09-08 -- The over-digestion attribution behind v6.0 is withdrawn. Nothing performed on this page changed — permeabilization is still 0.1 M HCl, 5 min, RT, no protease. MINOR, but read it before running a comparison: what changed is the reason the protease is absent, and therefore what a low-signal result means.

What was withdrawn. v6.0 rested on two arguments. The first — that four independent padlock/RCA systems permeabilize fresh-frozen tissue with no protease at all — is unaffected and is why acid-only remains the specified condition. The second, that our own 0.4 mg/mL / 2 min step was over-digesting, does not survive. A BZ09 re-run with reduced digestion pointed away from digestion entirely: the low-yield SPRINTseq runs from BZ09CA1-ReCA2 onward shared an RCA primer synthesised without its 3' phosphorothioate, which lets phi29's proofreading exonuclease consume free primer over a 12-16 h incubation and leaves fewer circles primed. Fewer rolonies, different step. See the internal note The RCA Primer Needs a 3' Phosphorothioate (lab site → Research).

And the arithmetic never supported it. 0.4 mg/mL P6887 for 2 min is ≈ 2,560 U·min/mL against ≈ 1,250 for a published fresh-frozen HybISS/CARTANA condition (Lawrence 2024) — within about 2x of the field, not an outlier. 0.4 mg/mL pepsin in 0.1 M HCl for 2 min is not an over-digested condition, and the page now says so, so that restoring it reads as a ladder arm rather than a regression.

One confound stays open, and it is not the protease: the TNBC run that started all of this used 0.01 M HCl, the reagent page's recipe at the time and 10x weaker than every protocol page. That was closed separately in Pepsin Solution Preparation v3.0.

Also. A garbled sentence in the acid-only rationale ("the lab-private note the internal evidence review …") is repaired, and the rationale admonition is retitled from "why the protease was the suspect" — which is exactly the claim being withdrawn.

  • v6.1 -- 2026-08-18 -- Follows RCA Protocol v6.0, which deleted the blocking step: this page now hands off to hybridization rather than to blocking. MINOR — nothing performed on this page changed, but the stated next step did, and a reader following the old pointer would look for a step that no longer exists.

  • v6.0 -- 2026-08-18 -- Pepsin removed; permeabilization is now 0.1 M HCl for 5 min at room temperature. MAJOR because data produced with and without a protease step are not directly comparable — detection efficiency, and which compartments survive, both change.

Why. A TNBC fresh-frozen run digested at 0.05% instead of the intended 0.04% produced markedly fewer rolonies, which prompted a review of where the step came from. Two findings drove this change. First, an in situ readout only requires that probes reach RNA that remains inside the cell, whereas the capture-based lineage our pepsin recipe descends from requires RNA to diffuse out of the tissue — so those conditions are systematically too aggressive here, and past the point where RNA is still anchored the molecule is lost across the ~16 h hybridization, the formamide washes and the overnight RCA. Second, four independent padlock/RCA systems permeabilize fresh-frozen tissue with no protease at all: SCRINSHOT (0.1 M HCl 3 min, the same direct-RNA SplintR chemistry as ours), HybISS (0.1 M HCl 5 min), the commercial CARTANA ISS kit, and Xenium FF (1% SDS + cold 70% methanol). 5 min was chosen over 3 as the safer starting point for dense tumour stroma: removing the protease moves the risk from over- to under-permeabilization.

Status. Adopted from the literature, not yet validated at our bench — the page carries a warning and a titration ladder to settle it. The v5.3 condition is recorded inline so it can be reinstated or run as a control arm.

Also. The acid is now stated as 0.1 M and the temperature as room temperature, both explicitly. reagents/pepsin.md previously specified 0.01 M HCl while every protocol page said 0.1 M; that conflict is resolved in favour of 0.1 M, which is what the entire padlock/RCA literature uses. Neither source paper states a temperature for the acid step, so RT is stated here rather than left to be inferred.

Evidence, citations and exposure arithmetic: lab-private note the internal evidence review Permeabilization / Pepsin Digestion (lab site → Research). - v5.3 -- 2026-01-29 -- Fresh-frozen pre-treatment split out of the former combined rca-ff protocol so that sectioning, fixation, permeabilization and dehydration are documented once and feed the sample-agnostic RCA Protocol. Steps are unchanged from the combined v5.3; earlier history is in git.

Comments

Ask / Feedback