Precise somatic genome editing for treatment of inborn errors of immunity

Abstract

Rapid advances in high throughput sequencing have substantially expedited the identification and diagnosis of inborn errors of immunity (IEI). Correction of faulty genes in the hematopoietic stem cells can potentially provide cures for the majority of these monogenic immune disorders. Given the clinical efficacies of vector-based gene therapies already established for certain groups of IEI, the recently emerged genome editing technologies promise to bring safer and more versatile treatment options. Here, we review the latest development in genome editing technologies, focusing on the state-of-the-art tools with improved precision and safety profiles. We subsequently summarize the recent preclinical applications of genome editing tools in IEI models, and discuss the major challenges and future perspectives of such treatment modalities. Continued explorations of precise genome editing for IEI treatment shall move us closer toward curing these unfortunate rare diseases.

Keywords: base editing; gene therapy; genome editing; hematopoietic stem cell (HSC); inborn errors of immunity (IEI); prime editing.

Figure 1

The use of programmable nucleases for genome editing. Three major programmable nucleases, i.e., ZFNs, TALENs and CRISPR/Cas9 are schematically illustrated on the left part. ZFNs and TALENs are respectively composed of sequence-specific DNA-binding protein modules linked to a nonspecific DNA cleavage domain (derived from FokI endonuclease). In ZFNs, an array of zinc figures (small blocks in different shades of purple), each specifically recognizing a 3-bp of DNA sequence, are assembled in tandem to program a specific DNA-binding event. In TALENs, variants of TALE repeats correspond to each of the four single bases. Therefore, programmed DNA binding by TALEN is enabled by an assembled array of TALE repeats (dense purple bands). Two other constant domains in TALEN are also depicted in dark and light blue, respectively. For DNA cleavage to occur, the FokI domain requires dimerization. As a result, only two adjacent targeting events by a pair of targeting ZFN or TALEN monomers in correct orientations would trigger double-strand break (DSB) at a specific locus. The mechanistic principle for CRISPR/Cas9 (a prototype for many CRISPR/Cas tools) is different from those for ZFNs and TALENs. It employs a Cas9 protein and an engineered guide RNA. The Cas9 protein features two nuclease domains, each responsible for cleaving one strand of DNA. The engineered guide RNA (in a form of single guide RNA, sgRNA [purple]) directs Cas9 to the DNA target by base-pairing, which results in generation of a targeted DSB. The target sequence for CRISPR/Cas9 also requires an immediately adjacent PAM sequence (shown in red). Regardless of the nuclease tools used, the DSBs generated would stimulate cellular DNA repair mechanisms (right part), in the form of the error-prone nonhomologous end joining (NHEJ) or homology-directed repair (HDR). The former pathway often leads to uncontrolled insertions or deletions of bases (indels, shown in yellow) around the break site, which tends to cause gene disruptions. The latter pathway engages precise repair directed by homologous donor DNA. In practice, when the cells are supplied with a donor DNA template (either as dsDNA or single strand oligodeoxynucleotides [ssODN] as indicated) containing the desired edits (green) and flanking homologous regions, the HDR pathway can lead to precise genome editing outcomes.

Figure 2

Distribution of IEI-associated variants. A survey of ClinVar database was conducted (June, 2022) for the distribution of the pathogenic variants associated with a keyword of “immunodeficiency”. The numbers of different variant types belonging to each of the indicated categories were filtered. Subsequently, the percentages for each category of variants within all different variants are marked on the pie graph.

Figure 3

Base editing tools. The two major classes of base editors are illustrated on the left and right parts, respectively. CBE is constructed by fusing the Cas9 nickase (nCas9) with a cytidine deaminase enzyme, which targets the Cs within the exposed R-loop for conversion into U, the latter serving as a temporary surrogate for thymidine in base pairing properties. The uracil glycosylase inhibitor domains (2xUGI, in yellow) engineered to prevent base excision of the U are also shown. The activity window of the initially developed CBE often covers positions 4-8 (schematically indicated with light blue). The adoption of nCas9 in CBE is by design. Its activity would nick the unedited strand, thereby signaling the cellular repair to install the corresponding As on the complementary strand. The design of ABE follows similar principles, except for the employment of an evolved, DNA-targeting adenosine deaminase. Deamination of adenosine produces inosine, which serves as a temporary surrogate for guanosine. Note that no inhibitory domains against base excision of I are required for optimal ABE activities. Some key considerations for the current developments of base editors are summarized in the middle part.

Figure 4

Prime editing. (A) The principle underlying the prime editor (PE) is illustrated. PE employs nCas9 (H840A) fused with a reverse transcriptase (RT) domain. PE is directed by pegRNA whose structure is schematically shown in a grey box. The pegRNA is composed of a targeting sgRNA scaffold and an extended 3′ region. This region can base pair with the nicked R-loop (via its terminal portion named primer binding sequence, PBS), and directs the ensuing reverse transcription (via its edits-containing segment named RT-template). The potential mechanistic stages underlying PE are depicted in order. (B) Major focuses for improving the PE efficiencies are summarized. (C) The advantages of PE with paired pegRNAs are summarized. Note that each pegRNA would enable reverse transcription of a segment of ssDNA (indicated in red). Programming the paired pegRNA sequences would enable increased PE activities, or empower precise editing of larger scales.