🧬 Aging Wiki

cas12a2-trans-cleavage

Properties1
aliasesCas12a2, SuCas12a2, transcript-activated chromatin shredding, RNA-triggered chromatin shredding, trans-cleavage, collateral cleavage, RNA-guided cell ablation, type V CRISPR collateral nuclease

8 min read

⚠️ Auto-extracted by Claude on 2026-06-13. The mammalian/therapeutic claims attributed to 1 have been verified against the bioRxiv preprint full text (DOI 10.64898/2026.05.08.723607, verified 2026-06-13); see zeng-2026-cas12a2-chromatin-shredding for the verified study page. The foundational bacterial-mechanism claims (Dmytrenko 2023; Scholz 2026) are sourced from metadata/abstracts only — not verified against full text. This page remains verified: false pending full-text access to Dmytrenko 2023 and Scholz 2026. gap/no-fulltext-access

Cas12a2 trans-cleavage (transcript-activated chromatin shredding)

Cas12a2 is an RNA-guided nuclease from a bacterial CRISPR abortive-infection system. Unlike a precision editor, its defining property is collateral (trans) cleavage: once its guide RNA recognises a complementary target RNA, Cas12a2 becomes a non-specific nuclease that degrades single-stranded RNA, single-stranded DNA, and double-stranded DNA indiscriminately 2. In bacteria this is a suicide defence — a phage transcript activates Cas12a2, which destroys the cell’s nucleic acids to abort the infection and protect the colony.

Transcript-activated chromatin shredding is the repurposing of this property to kill a chosen eukaryotic cell: deliver Cas12a2 + a guide RNA against a transcript that marks the unwanted cell, and in any cell expressing that transcript the activated enzyme shreds chromatin, causing a genome-wide DNA-damage response, mitotic catastrophe, and death 3 1. The trigger transcript is the targeting signal; the killing itself is deliberately non-specific within the doomed cell.

This page covers the technique. For the primary cancer paper see zeng-2026-cas12a2-chromatin-shredding; for the therapeutic-modality framing (and the aging/senolytic relevance) see transcript-activated-cell-ablation. Contrast with in-vivo-base-editing, the precision end of the CRISPR-method spectrum (a single intended base change, no DSB) — Cas12a2 is the opposite design point (maximal, non-specific damage, gated by an RNA sensor).

Principle

A standard DNA-targeting CRISPR nuclease (Cas9, Cas12a) cuts one DNA locus the guide specifies. Cas12a2 is an RNA sensor coupled to a non-specific nuclease:

  1. Sensing. The guide RNA base-pairs with a complementary target RNA (e.g. a messenger RNA). Activation additionally requires an adenine-rich protospacer-flanking site (PFS) near the target (functionally analogous to the PAM of DNA-targeting CRISPR) — this is the recognition handle that single-nucleotide-variant targeting exploits 2 1.
  2. Conformational switch. Target-RNA binding opens the nuclease active site into a trans-active state.
  3. Collateral shredding. The activated enzyme degrades nearby nucleic acids non-specifically — ssRNA, ssDNA, and dsDNA — including, in a eukaryotic nucleus, chromatinised genomic DNA. It preferentially cuts the accessible internucleosomal linker DNA (producing a nucleosome-laddered digestion pattern in vitro) 1.
  4. Cell death. Genome-wide double-strand breaks trigger a DNA-damage response (γH2AX, phospho-KAP1 S824), G2 arrest, and mitotic catastrophe — enlarged/fragmented nuclei and loss of proliferation, progressing to propidium-iodide⁺/Annexin-V⁺ death 1.

Why this is a different design point from base editing

in-vivo-base-editing is engineered to minimise collateral damage (no double-strand break, a ~4–8 nt window, a single intended transition). Cas12a2 is engineered around the opposite: the payload IS catastrophic, genome-wide DNA damage, and all of the precision lives in the RNA-sensing trigger, not in the cut. The therapeutic question therefore inverts: for base editing you ask “how clean is the edit?”; for Cas12a2 you ask “how exclusive is the trigger transcript to the cells you want dead?”

Workflow (mammalian transcript-activated killing, after Zeng 2026)

  1. Pick a trigger transcript present in the target cells and absent (or below the killing threshold) in bystanders. Demonstrated classes: a point-mutant mRNA (mutant TP53), an indel-junction mRNA (EGFR exon-19 deletion), an over-expressed oncogene above an abundance threshold (CCNE1, MYC), or any cell-specific/viral transcript.
  2. Design the guide RNA complementary to the target, positioned so an adenine-rich PFS is present. For single-nucleotide-variant selectivity, place the guide so the mutation either creates the activating PFS (e.g. the R248Q G→A “mutA” PFS) or sits in the guide–target duplex where a single mismatch to wild-type abolishes activation.
  3. Deliver Cas12a2 + guide. Demonstrated routes: purified ribonucleoprotein by nucleofection (research); stable expression (engineered lines); and — for therapy — co-delivered Cas12a2 mRNA + guide RNA in lipid nanoparticles (LNPs), including tissue-tropic SORT LNPs for lung.
  4. Sensing → activation → killing proceeds only in cells expressing the trigger transcript with its PFS.
  5. Read-out: loss of proliferation, DNA-damage markers (γH2AX, phospho-KAP1), nuclear fragmentation/mitotic catastrophe, cell-death stains; in vivo, tumour burden by bioluminescence / surface area; in competition, an isogenic mutant-vs-wild-type ratio shift.

Output format

  • Selective-killing readout — growth/viability of trigger⁺ vs trigger⁻ cells; strongest as an isogenic co-culture ratio (the spared population outgrows the targeted one only with the on-target guide).
  • In-vitro trans-cleavage gel — target-RNA-dependent degradation of labelled collateral substrate (capability, not selectivity).
  • DNA-damage / death markers — γH2AX, phospho-KAP1 S824; FUCCI cell-cycle (G2 arrest); PI/Annexin-V.
  • Abundance–lethality relationship — killing vs transcripts/cell (single-molecule FISH), defining the targeting window.
  • In-vivo efficacy + delivery efficiency — tumour-burden change, plus the fraction of target cells actually transfected (a reporter such as Cre→tdTomato).

Key parameters

ParameterValue / behaviour (Zeng 2026 context)Effect
EffectorSulfuricurvum SuCas12a2 (NLS-tagged for nuclei)RNA-guided collateral nuclease; the killing enzyme
TriggerComplementary target RNA + adenine-rich PFSGates activation; the selectivity handle
Collateral substratesssRNA, ssDNA, dsDNA (incl. chromatin)Genome-wide damage; preferential linker-DNA cleavage
Selectivity leversSNV-created PFS · indel junction · abundance thresholdThree demonstrated ways to make the trigger cancer-exclusive
Mg²⁺ dependenceactive down to 0.1 mM (slower); killing Mg²⁺-independent 0.1–10 mMFunctions at mammalian free Mg²⁺
Abundance thresholdkilling rises with transcripts/cell (~50 vs ~560 CCNE1 discriminable)Enables high-vs-low expression targeting; limits low-abundance triggers
DeliveryRNP nucleofection (research) · mRNA+gRNA LNP (therapy); ~7–18% in-vivo tumour uptakeIn-vivo delivery is the limiting variable
Multiplexingself-processes CRISPR arraySeveral triggers from one construct

Validation and QC

  • Isogenic negative control sharing everything but the trigger. The decisive controls in Zeng 2026 are RPE1 wild-type vs RPE1 p53^R248Q (same background), and the p53^R175H line that is identical to wild-type at the targeted position — both spared by the R248Q guide.
  • Decouple selectivity from abundance. Selectivity was shown to track the mutation, not expression, because the spared R175H line expressed more TP53 transcript than the killed R248Q line.
  • Endogenous-locus confirmation. Beyond over-expression transgenes, endogenous mutant transcripts (PC9: endogenous p53^R248Q + EGFR E746_A750del) were targeted.
  • Damage-mechanism confirmation. γH2AX + phospho-KAP1 + FUCCI G2 arrest + nuclear fragmentation confirm death proceeds through the expected chromatin-damage/mitotic-catastrophe route.
  • Delivery accounting. A Cre→tdTomato reporter quantified the (modest) in-vivo transfected fraction — the appropriate honesty check on an in-vivo efficacy claim.

Limitations and failure modes

  1. Selectivity is only as good as the trigger transcript. The cut is non-specific by design; if any bystander expresses the trigger above threshold, it dies too. There is no second specificity checkpoint. gap/long-term-unknown
  2. Genome-wide DNA damage is the mechanism, not a side effect. Sub-lethal activation could in principle leave survivors with mutagenic damage; the long-term consequence of partial shredding is uncharacterised. gap/no-mechanism
  3. Delivery dominates in-vivo efficacy. ~7–18% tumour-cell uptake, repeat dosing required, and failure to clear established advanced tumour burden (only delayed metastasis) — reaching dispersed cells (e.g. senescent cells across many tissues) is harder than a tumour mass.
  4. Abundance threshold cuts both ways. Low-abundance trigger transcripts may never reach the lethal dose.
  5. Mismatch sensitivity. Single guide–target mismatches can blunt killing — exploited for variant discrimination, but a robustness risk for the intended on-target cells.
  6. Innate-immune activation. Both the DNA-damage payload and the LNP carrier engage innate-immune/inflammatory sensing; this is intrinsic to the modality.

Evidence-weight implications for this wiki

When a paper reports Cas12a2 (or any collateral-nuclease, incl. Cas13) cell-killing, weight it by:

  1. Control quality. Isogenic, same-position-control comparisons (and co-culture/competition) ≫ monoculture growth curves.
  2. Abundance decoupling. Selectivity is only meaningful if it is not just the on-target line expressing more trigger transcript.
  3. Endogenous vs over-expressed trigger. Endogenous-locus targeting ≫ a transfected over-expression transgene.
  4. Delivery realism. RNP nucleofection / stable expression overstate deliverable efficacy; LNP in-vivo data with an explicit transfected-fraction accounting is the honest tier.
  5. For any aging application, treat the senescence/senolytic use as hypothesis, not result — no collateral-nuclease senescence-clearance data exist (see transcript-activated-cell-ablation).
MethodRelationshipNotes
in-vivo-base-editingOpposite design pointPrecision, DSB-free, single-base; Cas12a2 is maximal non-specific damage gated by an RNA sensor. Shared LNP-mRNA delivery chassis
Cas13 collateral RNA cleavageSibling collateral nucleaseRNA-targeting, RNA-collateral (HEPN domains); also used for target-RNA-triggered cell elimination, but kills via RNA degradation/stress rather than chromatin shredding
single-cell-rna-seqTrigger discoveryDefines cell-state-specific transcripts that could serve as triggers (e.g. senotype markers)
Nuclease knockout (CRISPR-Cas9 DSB)ContrastOne programmed DNA cut; Cas12a2 makes genome-wide collateral cuts after RNA sensing

Pages citing this method

Footnotes

Footnotes

  1. zeng-2026-cas12a2-chromatin-shredding · doi:10.1038/s41586-026-10738-7 (Nature 2026; preprint doi:10.64898/2026.05.08.723607) · Zeng J, …, Doudna JA · mammalian transcript-activated chromatin shredding; SuCas12a2 trans-cleaves chromatin on RNA recognition; SNV/indel/abundance selectivity; LNP mRNA+gRNA delivery. Verified against CC-BY bioRxiv preprint full text (2026-06-13); Nature version paywalled. ↩ ↩2 ↩3 ↩4 ↩5

  2. doi:10.1038/s41586-022-05559-3 · Dmytrenko O, Neumann GC, Hallmark T, …, Jackson RN, Beisel CL · Nature 2023 · 613(7944):588-594 · origin paper — Cas12a2 abortive infection: RNA-triggered, PFS-dependent non-specific destruction of dsDNA/ssDNA/ssRNA, inducing an SOS DNA-damage response. Establishes the trans-cleavage mechanism. Closed-access — cited from metadata. gap/no-fulltext-access ↩ ↩2

  3. doi:10.1038/s41586-026-10466-y · Scholz N, Thompson J, Crosby KT, …, Jackson RN, Beisel CL, Liu Y · Nature 2026 (6 May 2026) · companion foundational paper — RNA-triggered, sequence-specific killing of eukaryotic cells by Cas12a2 via genome-wide DNA shredding (selective elimination by gene expression, incl. virus-infected / mutation-bearing cells). Closed-access — cited from metadata. gap/no-fulltext-access ↩

Graph View

Backlinks