CRISPR-Cas9 editing at the HBB sickle cell locus: MMEJ-driven large deletions, micronuclei-mediated chromosomal loss, and LOA-induced ineffective erythropoiesis

Cas9 editing at HBB triggers MMEJ-driven kilobase deletions, micronuclei-mediated chromosomal loss, and LOA disrupting fetal hemoglobin compensation.


Molecular Pathway Insights

CRISPR-Cas9 editing at the HBB sickle cell locus: MMEJ-driven large deletions, micronuclei-mediated chromosomal loss, and LOA-induced ineffective erythropoiesis

Cas9-mediated correction of the sickle cell mutation in HBB exon 1 can achieve precise single-nucleotide repair, but a substantial fraction of edited hematopoietic stem cells instead undergo catastrophic loss of allele (LOA), where an entire gene copy is destroyed rather than corrected. The molecular determinants that decide whether a double-strand break resolves as a benign small indel or a kilobase-scale chromosomal deletion converge on repair pathway kinetics, micronucleus-driven replication failure, and the physical architecture of the beta-globin locus on chromosome 11p15.4.

Pathogenic origin of CRISPR-induced loss of allele at the HBB sickle cell locus

Loss of allele at the HBB locus arises from the convergence of three mechanistic forces: DNA repair pathway competition at Cas9-induced double-strand breaks, mitotic segregation failure of acentric chromosome fragments, and the structural vulnerability of the beta-globin gene cluster on chromosome 11p15.4. When Cas9 ribonucleoprotein (RNP) complexes generate blunt-ended DSBs at the sickle mutation site in HBB exon 1, the dominant NHEJ pathway competes with slower MMEJ and HDR pathways for break resolution, and the kinetic outcome of this competition determines whether the cell acquires a benign small indel or a catastrophic multi-kilobase deletion (PMID: 36269834, 41736887). The high sequence homology between HBB and the adjacent delta-globin gene HBD, combined with off-target Cas9 activity at sites such as OT18 on chromosome 11q, creates additional substrates for interhomologue recombination and megabase-scale chromosomal rearrangements that amplify allelic loss beyond the immediate cut site (PMID: 31147717, 37069266).

Molecular mechanism of DSB misresolution, MMEJ-driven deletion, and chromosomal missegregation

Non-homologous end joining (NHEJ) is the first-responder pathway at Cas9-induced DSBs, ligating broken ends through Ku70/Ku80 recruitment and DNA-PKcs-mediated synapsis to produce small insertions and deletions under 50 base pairs (PMID: 36269834). NHEJ-mediated repair typically saturates within 24 hours of Cas9 delivery. However, when NHEJ fails to resolve the break promptly, 5'-to-3' end resection by the MRN complex (MRE11/RAD50/NBS1) and CtIP generates single-stranded DNA overhangs that expose microhomology tracts flanking the cut site. These resected ends are channeled into microhomology-mediated end joining (MMEJ), which aligns short homologous sequences (5 to 25 bp) across the break junction and deletes the intervening DNA, producing kilobase-scale deletions that continue to accumulate for up to 72 hours post-editing (PMID: 36269834, 41736887). The guide RNA sequence determines the balance between these outcomes: the R-02 gRNA predominantly generates a 9 bp in-frame deletion via a local MMEJ microhomology, yielding lower rates of large deletions, whereas the R-66S gRNA produces a more diverse, frameshift-skewed indel spectrum with higher large-deletion frequency (PMID: 41736887, 36269834).

Pharmacological inhibition of NHEJ using the DNA-PKcs inhibitor M3814 was designed to shift repair toward HDR and improve therapeutic knock-in rates, but paradoxically increases the formation of large genomic modifications and loss of allele events by extending the window of unresolved DSBs available for MMEJ processing and chromosomal missegregation (PMID: 41736887). The REC3 domain of Cas9 functions as a conformational checkpoint sensor that governs the allosteric transition of the HNH nuclease domain to its catalytically active state upon RNA/DNA heteroduplex recognition, and variations in this checkpoint stringency across engineered high-fidelity Cas9 variants influence the kinetics of DSB formation and consequently the repair pathway balance at the HBB locus (PMID: 28931002).

When DSBs persist unrepaired through S phase and into mitosis, the Cas9 cut site divides chromosome 11 into a centric fragment (retaining the centromere) and an acentric fragment (lacking centromeric attachment). The acentric fragment, unable to attach to the mitotic spindle, missegregates during anaphase and is partitioned into a micronucleus (PMID: 34615869). Within micronuclei, the encapsulated chromosome undergoes severely defective DNA replication, detectable as reduced EdU incorporation, causing progressive degradation and permanent loss of the genomic material distal to the cut site. Concurrently, fusion of the broken centric ends of sister chromatids generates dicentric chromosomes that form anaphase bridges during subsequent divisions; resolution of these bridges can result in co-segregation of both chromosome 11 homologs to a single daughter cell, leaving the other daughter with whole-chromosome monosomy (PMID: 34615869). Interhomologue recombination (IHR) adds a further layer of complexity: Cas9-induced DSBs at HBB can be repaired using the highly homologous HBD gene as an endogenous template, producing gene conversion events that correct the sickle mutation but simultaneously generate 7.4 kb deletions spanning the HBB-to-HBD interval, and concurrent off-target DSBs at the OT18 site on chromosome 11q can produce 54 Mb pericentric deletions or inversions (PMID: 31147717, 37069266, 25894090).

Cellular and molecular damage in hematopoietic stem cells and erythroid differentiation

The consequences of MMEJ-driven large deletions and micronuclei-mediated chromosomal loss are disproportionately concentrated in the clinically critical long-term repopulating HSC compartment. CD34+CD38-CD45RA-CD90+ HSCs exhibit significantly higher rates of large deletions and correspondingly lower rates of homology-directed repair compared to more differentiated hematopoietic progenitor cells (HPCs), indicating that the slow-cycling, quiescent stem cell state favors delayed repair kinetics that predispose to MMEJ and chromosomal missegregation (PMID: 36269834). TREX1 exonuclease activity further compounds the problem in primary cells by degrading single-stranded oligodeoxynucleotide (ssODN) HDR templates before they can engage the homologous recombination machinery, reducing the fraction of precisely corrected alleles and increasing the relative proportion of LOA outcomes (PMID: 39569586). Standard short-read next-generation sequencing and digital droplet PCR significantly underestimate the true frequency of these large genomic modifications; clonal SMRT-seq analysis with dual unique molecular identifiers reveals that approximately 40% of edited colonies harbor large-deletion-containing alleles missed by population-level short-read assays (PMID: 36269834).

LOA and large-deletion alleles persist through erythroid differentiation and produce qualitatively distinct pathological consequences compared to small NHEJ indels. Small indels confined to HBB exon 1 disrupt adult beta-globin production but preserve the upstream HBG1 and HBG2 promoter architecture, enabling compensatory gamma-globin reactivation through reduced promoter competition within the locus control region (LCR) (PMID: 36269834). By contrast, LOA lesions extend beyond HBB into the centromeric region encompassing HBG1/HBG2 promoter elements and their associated chromatin regulatory domains, destroying the capacity for fetal hemoglobin (HbF) induction despite biallelic HBB disruption (PMID: 41736887). This creates a thalassemia-like imbalance where neither adult nor fetal beta-like globin chains are produced from the affected allele, resulting in unpaired alpha-globin precipitation, oxidative membrane damage, and markedly elevated Annexin V positivity (a marker of phosphatidylserine externalization and early apoptosis) in LOA-bearing erythroid precursors (PMID: 41736887). Clinical trial data reinforce this distinction: in an HBB gene correction trial, precisely corrected alleles declined from 33% to 1.3% of the graft post-infusion, replaced by indel-bearing clones that induced therapeutic HbF levels through an as-yet-unresolved compensatory mechanism, suggesting strong negative selection against LOA-carrying clones in vivo (PMID: 41736887).

Downstream pathophysiological outcome: a self-amplifying genotoxic cascade

The pathophysiology of CRISPR-induced LOA at the HBB locus operates as a self-amplifying genotoxic cascade in which unresolved Cas9 DSBs generate acentric fragments that missegregate into micronuclei, where defective replication produces additional DNA damage and chromosomal instability that propagates through subsequent cell divisions. This feed-forward loop is compounded by pharmacological NHEJ inhibition (M3814, DNA-PKcs blockade) that extends the window of unresolved breaks, by TREX1-mediated degradation of HDR donor templates that reduces precise repair and increases the relative LOA fraction, and by the physical architecture of the beta-globin locus where LOA lesions extending into centromeric HBG1/HBG2 territory destroy fetal hemoglobin compensation and create an irreversible thalassemia-like state (PMID: 41736887, 34615869, 36269834). The convergence of HSC-intrinsic repair vulnerability, detection bias from short-read sequencing that underestimates LOA frequency, and the apoptotic elimination of LOA-bearing erythroid precursors collectively define a genotoxic threshold that determines whether a Cas9-edited graft achieves durable therapeutic correction or undergoes progressive clonal attrition driven by unintended chromosomal damage.

Frequently asked questions

Why does CRISPR-Cas9 editing at the HBB locus cause loss of allele?

Cas9-induced double-strand breaks at HBB that are not rapidly resolved by NHEJ undergo end resection and shift toward MMEJ, generating kilobase-scale deletions. When DSBs persist into mitosis, acentric chromosome fragments missegregate into micronuclei where defective DNA replication permanently destroys the targeted allele.

What is the difference between NHEJ and MMEJ repair outcomes at the HBB locus?

NHEJ is the predominant repair pathway and typically produces small insertions and deletions under 50 base pairs within 24 hours. MMEJ operates on a slower timescale, generating large deletions ranging from 200 bp to several kilobases that continue accumulating for up to 72 hours after the initial double-strand break.

How do micronuclei contribute to chromosomal loss after Cas9 editing?

A Cas9 DSB divides the chromosome into a centric fragment (with centromere) and an acentric fragment (without centromere). The acentric fragment missegregates during mitosis and is partitioned into a micronucleus, where it undergoes severely defective DNA replication, leading to permanent loss of the genomic material distal to the cut site.

Why does loss of allele at HBB fail to induce compensatory fetal hemoglobin?

Unlike small indels confined to HBB exon 1, LOA lesions extend beyond the HBB coding region and disrupt the upstream HBG1 and HBG2 promoter elements or the local chromatin architecture required for gamma-globin activation. This prevents the compensatory fetal hemoglobin switch that normally mitigates beta-globin deficiency.

Are hematopoietic stem cells more susceptible to large deletions than progenitor cells?

Yes. Long-term repopulating HSCs (CD34+CD38-CD45RA-CD90+) exhibit higher rates of large deletions and lower rates of homology-directed repair compared to more differentiated hematopoietic progenitor cells, making the clinically relevant stem cell compartment disproportionately vulnerable to LOA.

How does interhomologue recombination cause unintended rearrangements at HBB?

The high sequence homology between HBB and the nearby delta-globin gene HBD enables Cas9-induced DSBs at HBB to be repaired using HBD as an endogenous template, producing gene conversion events. Concurrent on-target and off-target DSBs (such as at OT18 on chromosome 11q) can generate 7.4 kb deletions between HBB and HBD or 54 Mb pericentric inversions.

Can base editing avoid the loss of allele problem at the HBB sickle cell locus?

Adenine base editors (ABE8e) convert the sickle HbS codon to the non-pathogenic Makassar variant (HbG) with approximately 80% efficiency without creating double-strand breaks, thereby eliminating the MMEJ-driven large deletions and micronuclei-mediated chromosomal loss that cause LOA in nuclease-based approaches.

Explore CRISPR genotoxicity pathways on BioSkepsis

Run your own literature-grounded research threads on gene editing safety, DNA repair dynamics, or hematopoietic stem cell biology. Every claim is citation-verified against PubMed.

Start free

Sources and further reading

  1. Samata B et al. High-resolution genotyping of CRISPR-Cas9 edited hematopoietic stem cells reveals large genomic modifications and loss of allele at HBB. PMID: 41736887
  2. Leibowitz ML et al. Chromothripsis as an on-target consequence of CRISPR-Cas9 genome editing. Nat Genet. 2021;53(6):895-905. PMID: 34615869
  3. Lattanzi A et al. Large genomic deletions and other unintended outcomes after CRISPR-Cas9 editing of human hematopoietic stem cells. Nat Med. 2022. PMID: 36269834
  4. Pattabhi S et al. In vivo outcome of homology-directed repair at the HBB gene in HSC using alternative donor template delivery methods. Mol Ther Nucleic Acids. 2019;17:291-302. PMID: 31147717
  5. Liang P et al. CRISPR/Cas9-mediated gene editing in human tripronuclear zygotes. Protein Cell. 2015;6(5):363-372. PMID: 25894090
  6. Papathanasiou S et al. Whole chromosome loss and genomic instability in mouse embryos after CRISPR-Cas9 genome editing. Nat Genet. 2021;53(6):906-913. PMID: 34615869
  7. Cancellieri S et al. Off-target effects and large structural variants from CRISPR-Cas9 editing in HSPCs. Blood Adv. 2023. PMID: 37069266
  8. Chen F et al. REC3 domain checkpoint control of Cas9 HNH nuclease activation. Cell. 2017;170(5):845-855. PMID: 28931002
  9. Newby GA et al. Base editing of haematopoietic stem cells rescues sickle cell disease in mice. Nature. 2021;595(7866):295-302. PMID: 34079130
  10. Jinek M et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012;337(6096):816-821. PMID: 22745249
  11. Cong L et al. Multiplex genome engineering using CRISPR/Cas systems. Science. 2013;339(6121):819-823. PMID: 23287718
  12. Dever DP et al. CRISPR/Cas9 beta-globin gene targeting in human haematopoietic stem cells. Nature. 2016;539(7629):384-389. PMID: 27820943
  13. TREX1 exonuclease degradation of ssDNA donor templates in primary hematopoietic cells. PMID: 39569586
  14. Park SH et al. Low-density culture priming of long-term repopulating HSCs for CRISPR targeting. PMID: 30195800
  15. Walters MC et al. OTQ923 HBG promoter editing: pancellular HbF induction in sickle cell disease. PMID: 37646679
  16. Hendel A et al. Chemically modified guide RNAs enhance CRISPR-Cas genome editing in human primary cells. Nat Biotechnol. 2015;33(9):985-989. PMID: 26121415
  17. Wilkinson AC et al. Cas9-AAV6 gene correction in human HSCs. PMID: 34135108