Reaction Buffers & Hybridization Stringency¶
Every hybridization mix in this workflow is built on a ligase buffer — Ampligase buffer on the direct-RNA route, the ligase's own buffer on the cDNA/SNP routes — so that the downstream ligation shares one buffer system. That is convenient, but it means swapping the ligase silently changes the salt the probes anneal in. This page records what is in each buffer and how much that actually matters.
The short answer
Less than you would guess. At 10 mM Mg²⁺ — which every buffer here carries — the divalent term dominates ionic strength so completely that even a 2.7-fold change in monovalent salt moves arm Tm by only ≈2 °C. The stringency of this workflow is set by temperature and formamide, not by which ligase buffer you picked.
Buffer Compositions (at 1X)¶
| Buffer | Buffering / pH | Monovalent | Mg²⁺ | Cofactor | Other |
|---|---|---|---|---|---|
| Ampligase (Lucigen/Biosearch) | 20 mM Tris-HCl, pH 8.3 | 25 mM KCl | 10 mM MgCl₂ | 0.5 mM NAD⁺ | 0.01% Triton X-100 |
| SplintR (NEB M0375) | 50 mM Tris-HCl, pH 7.5 | — | 10 mM MgCl₂ | 1 mM ATP | 10 mM DTT |
| T4 DNA Ligase (NEB B0202) | 50 mM Tris-HCl, pH 7.5 | — | 10 mM MgCl₂ | 1 mM ATP | 10 mM DTT |
| HiFi Taq DNA Ligase (NEB M0647) | 20 mM Tris-HCl, pH 8.5 | 150 mM KCl | 10 mM MgCl₂ | 1 mM NAD⁺ | 10 mM DTT, 0.1% Triton X-100 |
| Taq DNA Ligase (NEB B0208) | 20 mM Tris-HCl, pH 7.3 | 25 mM KOAc | 10 mM Mg(OAc)₂ | 1 mM NAD⁺ | 10 mM DTT, 0.1% Triton X-100 |
| Tth DNA Ligase | 20 mM Tris-HCl, pH 7.6 | ~100 mM KCl | 10 mM MgCl₂ | 1 mM NAD⁺ | 10 mM DTT |
| 5X RT Buffer (Maxima H Minus) | 50 mM Tris-HCl, pH 8.3 | 75 mM KCl | 3 mM MgCl₂ | — | 10 mM DTT |
Tth row is literature-typical, not a vendor sheet
Thermo does not publish the 10X composition for its Tth DNA Ligase (EL07981); the row above is the standard Tth reaction condition from the literature. Verify against the lot datasheet before relying on it for a Tm calculation.
Two structural points worth noting:
- Cofactor splits the ligases cleanly. SplintR and T4 are ATP-dependent; Ampligase, Taq, HiFi Taq and Tth are NAD⁺-dependent. You cannot cross buffers between the two classes and expect ligation — the cofactor simply is not there.
- Ampligase buffer already carries NAD⁺ (0.5 mM at 1X). That is why the direct-RNA hybridization buffer is Ampligase buffer even though SplintR does the ligation — it is a historical carry-over from a thermostable-ligase protocol, not a requirement.
What Actually Moves Arm Tm¶
Four terms, in descending order of how much they matter here:
| Term | Effect | Notes |
|---|---|---|
| Mg²⁺ | Dominant | 10 mM Mg²⁺ ≈ 380 mM Na⁺-equivalent via von Ahsen's [Na⁺]eq = [Na⁺] + 120·√([Mg²⁺] − [dNTP]). This is why the monovalent differences above barely register |
| Temperature | Direct | The 45 °C anneal is the stringency-defining condition |
| Formamide | −0.5 °C per % | 20% formamide ≈ −10 °C. The designer uses 0.5 °C/% (literature range 0.6--0.72; 0.5 is the conservative in-house value) |
| Monovalent (K⁺/Na⁺/Tris) | Weak here | All monovalent cations are equivalent — only the total ionic strength matters, not which salt |
| Probe concentration | Few °C per 10x | Tm carries an R·ln(C_T/4) term, so lowering probe concentration lowers Tm |
Worked Comparison: Does Swapping the Ligase Buffer Matter?¶
The hybridization mix is 1X ligase buffer + 50 mM added KCl + 20% formamide:
| Hybridization built on | Monovalent total | Mg²⁺ | Na⁺-equivalent | ΔTm vs default |
|---|---|---|---|---|
| Ampligase — direct-RNA, FFPE, iLock (default) | 75 mM | 10 mM | ~454 mM | — (baseline) |
| HiFi Taq — SNP record | 200 mM | 10 mM | ~579 mM | ≈ +2 °C |
| Tth (literature-typical) | 150 mM | 10 mM | ~529 mM | ≈ +1 °C |
The trap: don't compute this from monovalent salt alone
Using the familiar monovalent-only rule ΔTm ≈ 16.6 × log₁₀([M⁺]₂/[M⁺]₁) gives
+7 °C for the Ampligase → HiFi Taq switch, which would be a serious change in
stringency. That number is wrong here, because it ignores the 10 mM Mg²⁺ common to
both buffers. Fold the Mg²⁺ in as Na⁺-equivalent first and the same switch is
≈ +2 °C — a shift small enough to ignore against a Tm model whose own error is a
few °C.
Practical consequence: the SNP route's use of HiFi Taq buffer during hybridization does not require re-tuning the formamide or the anneal temperature. If you ever want to compensate exactly, ~2 °C is about 4% formamide.
Post-Hybridization Wash: What the Field Actually Uses¶
Current condition, all routes: 10% formamide in 2X SSC, 3 × 10 min, at room temperature (RCA v7.0, RCA: cDNA v3.0). 20% is retained as a labelled high-stringency arm. This section records the published values behind that choice, so they do not have to be re-looked-up. Every entry was read from the source; the quotes are verbatim.
| Method | Hybridization | Post-hybridization wash | Wash vs hyb |
|---|---|---|---|
| Ours — current, and 2023–2024 records | 1X Ampligase + 50 mM KCl + 20% FA, 45 °C | 10% FA / 2X SSC, 3 × 10 min, RT | lower |
| Ours, 2026-07-29 → 2026-08-18 only | same | 20% FA / 2X SSC, 3 × 10 min, "45 °C optimal, or RT" | higher |
| SCRINSHOT1 | identical buffer to ours, 45 °C | 10% FA / 2X SSC | lower |
| HybISS2 | 1X Tth + 50 mM KCl + 20% FA, 37 → 45 °C, ligase in the mix | none — 2 × PBS, straight to RCA | n/a |
| Open-source ISS3 | 2X SSC + 10% FA, 37 °C overnight | 2 × 10% FA / 2X SSC, then 2 × 2X SSC | equal FA, short |
| Xenium4 | proprietary, 50 °C, 16–24 h | proprietary buffer, 37 °C, 30 min | 13 °C cooler |
The rule this encodes: wash below the stringency you hybridized at
Every published method washes below its own hybridization stringency, and there is a mechanical reason. A post-hybridization wash runs with zero probe in solution, so the same ΔG that was an equilibrium during hybridization now runs one way — the wash is strictly harsher than the hybridization even at identical buffer and temperature. Matching the hybridization's formamide is therefore already an escalation, not a like-for-like. Ours is additionally ~1 °C harsher on salt: 2X SSC is 390 mM Na⁺ with no Mg²⁺, against ~454 mM Na⁺-equivalent in the hybridization.
It also lands at the worst moment. The padlock is still linear — once ligated it is a closed circle catenated around the target and washing is nearly free, but here base pairing is all that holds it.
Temperature outweighs formamide. 10 versus 20 percentage points is worth ~5 °C on a DNA:RNA duplex; RT versus 45 °C is worth ~20 °C. The July-2026 page offered both and so specified neither — roughly 25 °C of spread inside one SOP. Whatever formamide is chosen, pin the temperature.
Full argument, arithmetic and the four-chamber ladder that would settle the formamide question empirically: the internal evidence review Stringency, Nonspecific Deposition and FFPE Background (lab site → Research).
Formamide destabilizes our duplex less than the textbook figure suggests
The commonly quoted 0.6–0.72 °C per percent is a DNA:DNA figure. At hybridization-relevant formamide concentrations the stability ordering is RNA:RNA > RNA:DNA > DNA:DNA, so a DNA-probe:mRNA duplex loses less per percent. The 0.5 °C/% used throughout this page is the appropriate value for the direct-RNA route and is what the designer already uses. On the cDNA route the padlock binds DNA, so the higher figure is the right one there.
The Ligation Step Is Not the Stringent One¶
It is natural to worry that the padlock arms fall off during ligation, since the SplintR and T4 buffers have no added monovalent salt at all. They do not, for two reasons that both point the same way:
- Ligation runs at 37 °C, well below the 45 °C anneal.
- The ligation mix has no formamide — removing 20% formamide is worth about +10 °C of duplex stability, which swamps the ~1 °C lost to the lower monovalent.
So arms are substantially more stable during ligation than during hybridization. Hybridization and the stringent wash are where specificity is won or lost; the ligation buffer is permissive by comparison. This is also why the ligase's own fidelity — not the buffer — is what determines whether single-nucleotide discrimination is possible.
Computing This Properly¶
Don't hand-calculate for real design work. The lab's probe designer models all of this
explicitly (probe_design/designer/probe_designer/chemistry.py, ReactionConditions):
monovalent, Mg²⁺, dNTP chelation, formamide, and strand concentration, with the
SantaLucia 1998 salt correction plus the von Ahsen Mg²⁺ conversion. It also picks the
right nearest-neighbour table per chemistry, which matters more than the buffer:
| Duplex | NN table | Used by |
|---|---|---|
| DNA probe : mRNA | Sugimoto 1995 (R_DNA_NN1) |
Direct RNA, iLock |
| DNA probe : cDNA | SantaLucia & Hicks 2004 (DNA_NN4) |
cDNA, SNP genotyping |
Full provenance for every parameter, including the sources behind the numbers on this page,
is in probe_design/designer/docs/thermodynamics_parameters.md.
Keep the designer's defaults in sync with these protocols
ReactionConditions defaults are pinned to a specific protocol version. When a
hybridization condition changes here — probe concentration, formamide, anneal
temperature, or the buffer — the designer's defaults need the same change, or probe
panels get designed against conditions the bench no longer runs.
How Many Washes Is Enough¶
Default: PBS-Tween 0.05%, 2 x 2 min. A wash is dilution, not a reaction. With a ~100 μl PDMS chamber and thorough aspiration the residual film is roughly 5–10 μl, so each change dilutes carry-over ~10–20-fold: two washes ≈ 10²–10³, three ≈ 10³–10⁴.
That extra decade only earns its place when what is being carried over is both concentrated and potently inhibitory to the next step. Three cases qualify:
| After | Why 3 x |
|---|---|
| The formamide stringency wash, before ligation | Two independent reasons, and the salt one survives the drop to 10% formamide. Formamide: 10% is ~2.5 M (20% is ~5 M), and two washes still leave low-mM at the ligase, which costs activity. Salt: the wash is 2X SSC 390 mM Na⁺ regardless of formamide, and SplintR is *"srongly inhibited by concentrations of NaCl > 100 mM"*, with the inhibited fraction diverted to dead-end 5′-adenylylated probe that can never be ligated.5 Three changes of PBS-T put the ligase at ~10 mM. This is the one wash where the count is load-bearing. |
| NaBH₄ quench | An actively gassing strong reductant; get it out before anything enzymatic. See Autofluorescence Quenching. |
| Pepsin / acid digestion on FFPE, which goes straight into probe hybridization | 0.1 M HCl carried into a ~10–20 mM buffered reaction can move its pH. FF is unaffected — it passes through an ethanol series and a rehydration before any enzyme. |
Everywhere else 2 x 2 min is sufficient, and adding washes costs section adhesion and handling time rather than buying anything. Aspirating properly matters more than adding a fourth change — the residual film volume is the term you actually control.
References¶
Each entry notes what it is cited for here, so the source of a number on this page can be found without re-reading the paper.
- von Ahsen N, Wittwer CT, Schütz E. Oligonucleotide melting temperatures under PCR
conditions. Clinical Chemistry 47, 1956--1961 (2001).
— The Mg²⁺ → Na⁺-equivalent conversion,
[Na⁺]eq = [Na⁺] + 120·√([Mg²⁺] − [dNTP]). Everything in What Actually Moves Arm Tm and the ~454 mM figure for our hybridization buffer comes from this. - SantaLucia J. A unified view of polymer, dumbbell, and oligonucleotide DNA
nearest-neighbor thermodynamics. PNAS 95, 1460--1465 (1998).
— The salt correction the designer applies, and the source of the
16.6 · log₁₀([M⁺]₂/[M⁺]₁)form used in the worked comparisons. - Owczarzy R, et al. Predicting stability of DNA duplexes in solutions containing magnesium and monovalent cations. Biochemistry 47, 5336--5353 (2008). — Why monovalent-only rules fail at 10 mM Mg²⁺. The basis for the trap warning: divalent and monovalent contributions are not additive in the naive way.
- McConaughy BL, Laird CD, McCarthy BJ. Nucleic acid reassociation in formamide. Biochemistry 8, 3289--3295 (1969). — The original linear Tm-versus-formamide relationship. See also Blake & Delcourt 1996 (DOI 10.1093/nar/24.11.2095) for the thermodynamic treatment, and Casey & Davidson, Nucleic Acids Res 4, 1539--1552 (1977) for the RNA:RNA > RNA:DNA > DNA:DNA stability ordering that justifies using 0.5 °C/% rather than 0.72 on the direct-RNA route.
- NEB — Taq DNA Ligase Reaction Buffer (B0208) · SplintR Ligase (M0375) · HiFi Taq DNA Ligase (M0647) — Vendor 1X buffer compositions in the table above, including which ligases carry ATP and which carry NAD⁺.
- Ampligase 10X Reaction Buffer — LGC Biosearch Technologies — Source of the 0.5 mM NAD⁺ and 0.01% Triton X-100 in Ampligase buffer. The Triton figure matters twice: it is why the hybridization mix is effectively detergent-free (0.001% at 1X, well below CMC), and it is the only detergent anywhere in the reaction.
- Thermo Fisher Scientific. Maxima H Minus Reverse Transcriptase user guide, Pub. No. MAN0012047 Rev. C.00 (2024). — The 5X RT buffer composition in the table, and the reason the cDNA route uses 20 µl rather than 10 µl per 100 µl reaction.
-
Sountoulidis A, et al. (Nilsson M, Samakovlis C). SCRINSHOT enables spatial mapping of cell states in tissue sections with single-cell resolution. PLoS Biol 2020;18(11):e3000675. DOI: 10.1371/journal.pbio.3000675. The source of our hybridization mix, component for component — 1X Ampligase buffer, 0.05 M KCl, 20% formamide, 0.2 µg/µl BSA, 0.2 µg/µl tRNA, 1 U/µl RiboLock, 55 °C 15 min → 45 °C. Its wash is "10% Formamide in 2X SSC", and its ligation is 0.5 U/µl SplintR in T4 RNA ligase buffer + 10 µM ATP, 25 °C overnight, chosen "to minimize the effect of the previously documented SplintR ligase nucleotide preferences". Specificity is attributed to "padlocks with high melting points (Tm around 70 °C) followed by stringent washes" on 20-nt arms, 40 nt total. ↩
-
Gyllborg D, Nilsson M. HybISS: Hybridization-based In Situ Sequencing. protocols.io (2020). DOI: 10.17504/protocols.io.xy4fpyw. Padlock hybridization and Tth ligation run in one mix, so there is no wash before RCA at all. Also the source of two details worth knowing: the RCA mix carries Exonuclease I at 0.4 U/µl to digest unligated linear probe, and the protocol deliberately "avoids the use of ethanol dehydration steps, and minimal detergent use in wash buffers". ↩
-
Lee H, et al. (Nilsson M). Open-source, high-throughput targeted in situ transcriptomics for developmental and tissue biology. Development 2024;151(16):dev202448. DOI: 10.1242/dev.202448. Direct-RNA padlock hybridization in "2× SSC, 10% formamide and 10 nm of each padlock probe" overnight at 37 °C, washed with "two washes of 10% formamide in 2× SSC, followed by two washes in 2× SSC". The mildest hybridization condition in the comparison, on the same chemistry as ours. ↩
-
10x Genomics. CG000582 — Xenium In Situ Gene Expression: Probe Hybridization, Ligation & Amplification, Rev H. PDF. Hybridization 50 °C, 16–24 h; post-hybridization wash 37 °C, 30 min — the cleanest demonstration that the wash is meant to sit below the hybridization, since the buffer is proprietary but the 13 °C drop is not. ↩
-
Lohman GJS, Zhang Y, Zhelkovsky AM, Cantor EJ, Evans TC Jr. Efficient DNA ligation in DNA–RNA hybrid helices by Chlorella virus DNA ligase. Nucleic Acids Res 2014;42(3):1831–1844. DOI: 10.1093/nar/gkt1032. The primary characterisation of SplintR, and the source of three numbers used on this site: ligation is "strongly inhibited by concentrations of NaCl > 100 mM" (why the pre-ligation PBS-T chase is load-bearing); rate rises monotonically up to 37 °C (why our 37 °C ligation is right and needs no change); and "at the lower (10−100 µM) ATP concentrations, a slight boost in the ligation rate and a notable reduction of AppDNA formation was observed" — though PBCV-1 "shows little inhibition at 1 mM ATP for most sequences", so the NEB buffer's 1 mM is acceptable rather than optimal. ↩