STRC Research

STRC PE Phase1 pegRNA E1659A

7 min read

STRC PE Phase 1 pegRNA E1659A

Complete PE3 pegRNA design for Misha’s maternal allele c.4976A>C (p.E1659A) at chr15:43600551 (GRCh38). Spacer 5′-GCCCAGCTCCCCACCTGCTA-3′ matches the + strand protospacer immediately 5′ of the confirmed TGG PAM at chr15:43600540. The nick falls between chr15:43600536 and 43600537 on the + strand. Recommended 3′ extension: RT template (20 nt) 5′-CACTGAAATTGGCACCATAG-3′ + PBS (13 nt) 5′-CAGGTGGGGAGCT-3′. The edit (G→T on + strand, reverting MUT→WT) is installed at position 6 of the RT template reading 5′→3′ = 14 nt from the 3′ end of RT = 14 nt from the nick site. This is beyond Anzalone 2019’s 3–10 nt optimal window — efficiency penalty is structural and unavoidable without a closer PAM. No PE3b nicking sgRNA found: zero − strand NGG PAMs place a 20-nt protospacer across the variant within 100 nt, so we fall back to classic PE3 with nicking sgRNA 5′-TTGGTTTGGTTTCTATACCA-3′ (− strand, 63 nt from edit, GC 35%).

Method

  1. Fetched chr15:43600400-43600700 from Ensembl REST (GRCh38 + strand, 301 nt).
  2. Verified reference at variant: + strand chr15:43600551 = T (WT); Misha MUT = G. Confirmed TGG PAM at chr15:43600540-42.
  3. pegRNA spacer: 20 nt of + strand immediately 5′ of PAM → chr15:43600520-43600539.
  4. Nick site: SpCas9 H840A nickase cuts the non-target strand (+ strand, PAM-containing) between protospacer positions 17 and 18 = between chr15:43600536 and 43600537.
  5. RT template: reverse-complement of + strand WT sequence from nick 3′ end (chr15:43600537) downstream. Swept lengths 10–24 nt (min 16 to reach variant at nick+14); recommended 20 nt for 5-nt 3′-side homology arm past the edit.
  6. PBS: reverse-complement of + strand WT sequence immediately 5′ of nick. Swept 11–15 nt; recommended 13 (Chen 2021 default).
  7. PE3b nicking sgRNA scan: − strand NGG PAMs (CCN on + strand) in the 30–100 nt window from edit; protospacer 20 nt; PE3b requires the protospacer to span the variant so the nicker discriminates edited from unedited.
  8. PE3 (classic) fallback: − strand NGG PAMs in the same distance window, no spanning requirement.

Deterministic; no stochastic components.

Results

pegRNA design (+ strand, TGG PAM)

ComponentSequence (5′→3′)LengthGC%Notes
SpacerGCCCAGCTCCCCACCTGCTA20 nt70.0Matches + strand protospacer
PAM (genomic, not in pegRNA)TGG3 nt—chr15:43600540-42
RT templateCACTGAAATTGGCACCATAG20 nt50.0edit at RT pos 6 (5′→3′) = 14 nt from 3′
PBSCAGGTGGGGAGCT13 nt69.2RC of nick-5′ + strand
Full 3′ extensionCACTGAAATTGGCACCATAGCAGGTGGGGAGCT33 nt54.5RT + PBS

The RT template encodes the WT nucleotide A at the edit position (reading RT 5′→3′ on the pegRNA; this installs T at chr15:43600551 on the + strand during RT-driven synthesis = A on coding/− strand = reverts p.E1659A back to E1659).

Nick-to-edit geometry — unavoidable efficiency penalty

  • Nick position (+ strand): between chr15:43600536 and 43600537
  • Edit position (+ strand): chr15:43600551
  • Distance from nick to edit: 14 nt (downstream on + strand)
  • Anzalone 2019 optimal nick-to-edit distance for PE3: 3–10 nt
  • Expected efficiency penalty: ~50% of geometry-optimal (based on Anzalone distance-response curves)

The prior PAM survey (Prime Editing for STRC) showed this TGG PAM is already the closest SpCas9 NGG option. The next-closest PAMs (− strand, 22–25 nt from variant) make the geometry worse, not better. To access 3–10 nt nick distance would require an engineered Cas9 variant with relaxed PAM (SpRY NGN, SpG NGN, or SpCas9-NG) or an orthogonal nickase (SaCas9 NNGRRT). These are Phase 2 alternatives.

PE3b search — corrected in Phase 2

Phase 1 scan used a flawed filter (required both protospacer-spans-variant AND nick-to-edit distance 30–100 nt; these constraints are geometrically incompatible because spanning implies nick distance ≤ 17 nt). STRC PE Phase2 PAM Expansion reruns with the correct filter and recovers 5 PE3b-eligible nickers even at baseline SpCas9 NGG, including the lead − strand nicker ACTGAAATTGGCACCATAGC (PAM AGG, 12.5 nt from edit, GC 45%). PE3b IS available for this variant. The classic-PE3 fallback below was recommended in Phase 1 due to the bug; the PE3b nicker is strictly better.

PE3b nicking sgRNA — revised recommendation (from Phase 2)

Lead PE3b nicker: 5′-ACTGAAATTGGCACCATAGC-3′ (− strand, PAM AGG at + strand coords, nick 12.5 nt from edit, GC 45%). Protospacer spans the variant at chr15:43600551 → allele-discriminating; only nicks after successful edit installation.

Phase 1 originally recommended the non-discriminating classic-PE3 nicker TTGGTTTGGTTTCTATACCA (− strand, 63 nt from edit) as a fallback; this is superseded by the PE3b option above. Full alternative PE3b list in pe_phase2_pam_expansion.json.

Interpretation for Misha

  • Viable path, single-edit efficiency will be below published cortex/cardiomyocyte benchmarks (42% / 35%) due to the forced 14-nt nick-to-edit distance. PE3b IS available (corrected in Phase 2) so allele discrimination applies. Realistic OHC PE efficiency estimate with PE3b: 8–25% (Phase 2 revised up from Phase 1’s 5–20%).
  • The maternal allele carries E1659A on top of otherwise-functional STRC; PE correction at Misha-attainable efficiencies (say 10%) would recover a small additional fraction of functional stereocilin. Paired with mini-STRC gene replacement (covering the paternal deletion regardless), this produces an additive benefit.
  • Size: PE3 system (SpCas9-H840A–RT fusion) is ~6.3 kb — too large for single AAV. Requires dual-AAV split-intein delivery (Chen 2024 architecture; Prime Editing for STRC).
  • Decision boundary: if OHC-specific PE3 efficiency is demonstrated ≥20% in a mouse DFNB16 surrogate, PE becomes clinically competitive. If <10%, it’s academically interesting but inferior to mini-STRC replacement.

Limitations

  • Off-target scan for spacer is not included in Phase 1 — requires transcriptome-wide alignment (Phase 2). Expect the 70% GC spacer to have some off-targets; a CRISPROff / Cas-OFFinder run is mandatory before lab.
  • No biochemical folding simulation of the 53-nt pegRNA (spacer + scaffold + 3′ extension). RNA-fold secondary structure around the PBS can reduce priming efficiency if it forms stable hairpins (ViennaRNA check deferred to Phase 2).
  • Empirical RT length / PBS length are a combinatorial surface (3–4 RT × 5 PBS = 15–20 variants) — lab optimization needed, not something to call from computation alone.
  • PAM scan was limited to SpCas9 NGG; engineered variants (SpRY, SaCas9, enCas9) were not surveyed. A single parallel pass with SpRY NGN could restore PE3b eligibility and shorten nick-to-edit distance. Worth Phase 2.
  • Efficiency extrapolation (5–20%) is an order-of-magnitude guess calibrated to Anzalone distance-response curves, not a measurement. Real OHC PE efficiency for this variant is unknown.

Next steps

  1. Off-target scan: Cas-OFFinder against human genome for spacer + nicker, allow 0–3 mismatches.
  2. ViennaRNA fold of full pegRNA (spacer + tracr scaffold + RT + PBS) to check for PBS-region hairpins.
  3. SpRY / SaCas9 PAM expansion sweep — find a PAM placing nick 3–10 nt from edit; restores PE3b + geometry optimum.
  4. Ligand-assisted PE (PEn/PE5max) chemistry decision: newer PE architectures (PE5max with epegRNA scaffolds) yield 2–3× efficiency in post-mitotic cells; worth building into a parallel design.
  5. HEK293T rescue assay at a PE-competent partner lab: HDR-matched minigene + lentiviral PE delivery + deep sequencing of edited site. ~$8–12k, 6–10 weeks.

Replication

cd ~/STRC/models
/opt/miniconda3/bin/python3 pe_phase1_pegrna_design.py
# outputs: pe_phase1_pegrna_design.json

Files / Models

  • ~/STRC/models/pe_phase1_pegrna_design.py — full pegRNA + PBS + RT + nicker sweep
  • ~/STRC/models/pe_phase1_pegrna_design.json — all candidate sets and the recommended combination

Connections

Graph View

Backlinks