Precise large-fragment deletions in mammalian cells and mice generated by dCas9-controlled CRISPR/Cas3

Abstract

Currently, the Cas9 and Cas12a systems are widely used for genome editing, but their ability to precisely generate large chromosome fragment deletions is limited. Type I-E CRISPR mediates broad and unidirectional DNA degradation, but controlling the size of Cas3-mediated DNA deletions has proven elusive thus far. Here, we demonstrate that the endonuclease deactivation of Cas9 (dCas9) can precisely control Cas3-mediated large-fragment deletions in mammalian cells. In addition, we report the elimination of the Y chromosome and precise retention of the Sry gene in mice using CRISPR/Cas3 and dCas9-controlled CRISPR/Cas3, respectively. In conclusion, dCas9-controlled CRISPR/Cas3-mediated precise large-fragment deletion provides an approach for establishing animal models by chromosome elimination. This method also holds promise as a potential therapeutic strategy for treating fragment mutations or human aneuploidy diseases that involve additional chromosomes.

Deactivation Cas9 can control the size of Cas3-mediated deletions within the expected range in mammalian cells and mice.

INTRODUCTION

CRISPR/Cas9 systems, which have been widely adopted for gene inactivation, can generate desired DNA mutations with high precision and exhibit potential for treating genetic diseases and conducting high-throughput genetic screens (1–4). However, their ability to generate large genomic fragment deletions, especially those exceeding megabases or involving multiple genes, is limited. Although CRISPR/Cas9 systems have been previously reported to achieve large chromosome fragment deletion by simultaneously introducing double-strand breaks (DSBs) at two locations and using a synthetic single-stranded DNA (ssDNA) template (5, 6), they may induce undesirable off-target and unintended reverse mutations between the cleavage sites (7). In addition, while prime editing-Cas9–based deletion and repair (PEDAR) and PRIME-Del have been used for generating large-fragment deletions, ranging from 0.02 kb to approximately 10 kb in human embryonic kidney (HEK) 293T cells, they exhibit unsatisfactory efficiency, typically from 1% to 30% (8, 9).

A recent study demonstrated that the CRISPR/Cas3 system (type I-E CRISPR) can generate up to ~200-kb deletions in human cells with higher efficiencies than CRISPR/Cas9 (7, 10). Unlike CRISPR/Cas9, which induces large-fragment deletion by generating DSBs, CRISPR/Cas3 systems progressively shred DNA targets through a multistep process. The Cascade system uses CRISPR RNA (crRNA) to recognize a complementary target flanked by the 5′-ARG (where R is A or G) protospacer adjacent motif (PAM) (11, 12), which results in stable R-loop formation and triggers a large conformational change in the Cascade (11, 13, 14). Then, the helicase-nuclease fusion enzyme Cas3 is specifically recruited to the R-loop–forming Cascade, where the nontarget strand (NTS) DNA is repeatedly cleaved in cis and the target strand (TS) is cleaved by nonspecific ssDNA cleavage in trans, thus progressively degrading the upstream region and inducing large genomic fragment deletion (15). However, the unidirectional and unpredictable deletions induced by CRISPR/Cas3 may require caution in genome editing, especially in gene-dense regions or near regulatory elements. Therefore, controlling the size of Cas3-mediated DNA deletions is essential for precise genome editing in therapeutic applications.

Here, we hypothesize that site-specific DNA binding proteins may prevent Cas3 from reeling double-stranded DNA (dsDNA), potentially preventing Cas3-mediated end deletions. We designed four strategies [SpCas9, Cas9 nickase (nCas9), dCas9, and two crRNAs in different directions] to test this hypothesis. The results showed that dCas9 can control the size of Cas3-mediated deletions within the expected range in mammalian cells. Moreover, it could be used for targeted Y chromosome elimination and precise retention of Sry in mice. Thus, precise large-fragment deletions were generated in mammalian cells and mice by dCas9-controlled CRISPR/Cas3, which can broaden the applications of genome editing and potentially lead to future therapeutic applications in the clinic.

RESULTS

dCas9-controlled CRISPR/Cas3 can achieve precise large-fragment deletions in human cells

To assess the DNA cleavage activity of the CRISPR/Cas3 system, we transfected HEK293T cells with pre-crRNA, Cas3, and Cascade-coding plasmids. The results showed a broad pattern of Cas3-mediated DNA degradation, contrasting with the single band observed in wild-type (WT) controls, which is consistent with the findings of a previous study (7) (Fig. 1, A and B). Multiple primers were used to confirm the maximum fragment size that can be deleted by the Cas3 system (fig. S1A), and the results obtained by polymerase chain reaction (PCR) suggested that the large deletions were up to ~140 kb in length (fig. S1B). Then, we designed four strategies, including dCas9, nCas9, SpCas9, and two crRNAs in different directions, to control the size of Cas3-mediated large-fragment deletions in the following study (Fig. 1C).

Fig. 1. dCas9-controlled CRISPR/Cas3 can achieve precise large-fragment deletions in HEK293T cells.

(A) The type I-E CRISPR effector was composed of crRNA, Cas3, and the Cascade complex, which contains Cas5, Cas6, multiple Cas7, Cas8 (Cse1), and two Cas11 (Cse2). (B) Electrophoresis of the PCR products. CRISPR/Cas3 targeting of the AAG PAM mediated deletions at four sites. (C) Schematic depiction of the four strategies for precise deletion via CRISPR/Cas3. (D) Electrophoresis of the PCR products of the precise 1-kb deletion generated by dCas9-controlled CRISPR/Cas3. The designed crRNAs of Cas3 and sgRNAs of SpCas9 and its variants were offset by 1 kb at four gene loci. The expected bands, which were cloned and inserted into pGM-T for Sanger sequencing, are marked with rectangular boxes. The sgRNAs of Cas9s and the crRNAs of Cas3 are highlighted in pink and blue, respectively. Cells transfected with CRISPR/Cas3 alone served as a negative control. (E) WGS data analysis workflow for determining the efficiency of Cas3-mediated deletions and the proportion of dCas9-controlled precision deletions. (F) Efficiency of Cas3-mediated deletions (middle) and proportion of 1-kb precision deletions (right) generated by dCas9-controlled CRISPR/Cas3 at the DMD and ERCC4 gene loci. The average reads in the A, B, and C regions are shown (left). (G) Electrophoresis of the PCR products of the precise 2-kb deletion by dCas9-controlled CRISPR/Cas3. The sgRNAs of Cas9s and the crRNAs of Cas3 were offset by 2 kb at four sites. The expected bands, which were cloned and inserted into pGM-T for Sanger sequencing, are marked with rectangular boxes. (H) Efficiency of Cas3-mediated deletions (middle) and proportion of 2-kb precision deletions (right) generated by dCas9-controlled CRISPR/Cas3 at the DMD and ERCC4 gene loci. The average reads in the A, B, and C regions are shown (left).

To evaluate the validity of the precise large-fragment deletions of the Cas9-controlled CRISPR/Cas3 system, three stable cell lines were generated, each overexpressing dCas9, nCas9, or SpCas9. Subsequently, crRNAs for Cas3 and single-guide RNAs (sgRNAs) for dCas9, nCas9, and SpCas9, each with a 1-kb offset [distance between the “AAG” (Cas3) and “NGG” (SpCas9) PAM sequences], were transfected into HEK293T cells at four gene loci (EMX1, CCR5, DMD, and ERCC4). The results showed that dCas9-controlled CRISPR/Cas3 generated distinct single bands of controlled size at all four gene loci, while broader patterns were observed in the other three groups (Fig. 1D). The 5′ deletion boundaries were consistently found within the ~0.5-kb downstream region of the dCas9 sgRNAs (0.204 kb for EMX1, 0.334 kb for CCR5, 0.245 kb for DMD, and 0.133 kb for ERCC4), as determined by T-clone and Sanger sequencing (Table 1).

Table 1. Summary of deletions using four strategies by Sanger sequencing.

The designed crRNAs of Cas3 and sgRNAs of SpCas9 and its variants are an offset of 1 kb [distance between the “AAG" (Cas3) and “NGG” (SpCas9) PAM sequences] to program a 1-kb deletion at the four sites. The 5′ deletion boundaries are distributed across the upstream or downstream region of sgRNAs (dCas9, nCas9, SpCas9). +, upstream; −, downstream; No, no obvious PCR bands were near the designed deletion boundaries.

Site	dCas9	nCas9	SpCas9	Two crRNAs in different directions
EMX1	−204 bp	+124 bp	+243 bp	No
CCR5	−334 bp	+515 bp	No	No
DMD	−245 bp	No	No	No
ERCC4	−133 bp	No	+233 bp	No

Considering the limitations of PCR analysis, evaluating gene editing efficiency via whole-genome sequencing (WGS) was essential. We calculated the deletion efficiency as the number of reads aligning to a reference sequence of the deletion, out of the total number of reads aligning to reference sequences either with or without the deletion according to a previous study (8). The WGS results showed that the editing efficiency of Cas3-mediated large deletions was 67.23% for DMD and 25.74% for ERCC4 (Fig. 1, E and F). Additionally, we calculated the proportion of precise deletions as the number of reads aligning to a reference sequence of the intended deletion, out of the total number of reads aligning to reference sequences of the deletion (8). The proportion of precise dCas9-controlled CRISPR/Cas3-mediated 1-kb deletions in DMD and ERCC4 were 94.52% and 75.89%, respectively (Fig. 1, E and F). In addition, we managed to precisely generate 2-kb deletion fragments using these systems. As shown in Fig. 1G, single bands of controlled sizes were detected via PCR for dCas9-controlled CRISPR/Cas3, while broad patterns were observed for the other three groups at all four gene loci. The 5′ deletion boundaries were located within the ~0.5-kb downstream region of dCas9 sgRNAs (0.124 kb for EMX1, 0.407 kb for CCR5, 0.316 kb for DMD, and 0.187 kb for ERCC4), as determined by T-clone and Sanger sequencing (Table 2), consistent with the results of precise 1-kb deletions using dCas9-controlled CRISPR/Cas3. The WGS results showed that the editing efficiency of Cas3-mediated large-fragment deletions was 55.61% for DMD and 10.69% for ERCC4 (Fig. 1, E and H). Additionally, the proportion for dCas9-controlled CRISPR/Cas3-mediated precise 2-kb deletion in DMD and ERCC4 was 98.49% and 73.91%, respectively (Fig. 1, E and H). Furthermore, we recapitulated and confirmed these results in LO2 cells, A549 cells, mouse embryonic fibroblasts (MEFs), and porcine embryonic fibroblasts (PEFs) (fig. S2 and tables S1 to S4). To determine the effectiveness of dCas9 blockade at various distances, dCas9 at distances of 0.01, 0.1, 0.5, 1.5, 2.5, and 3 kb from the positions of crRNAs were used in this study. As shown in fig. S3 and table S5, precise large-fragment deletions can be achieved by dCas9-controlled CRISPR/Cas3 at 0.5- to 3-kb distances, whereas no notable editing events were observed at 0.01- or 0.1-kb distances.

Table 2. Summary of deletions using four strategies by Sanger sequencing.

The designed crRNAs of Cas3 and sgRNAs of SpCas9 and its variants are an offset of 1 kb [distance between the “AAG” (Cas3) and “NGG” (SpCas9) PAM sequences] to program a 2-kb deletion at the four sites. The 5′ deletion boundaries are distributed across the upstream or downstream region of sgRNAs (dCas9, nCas9, SpCas9). +, upstream; −, downstream; in left, the 5′ deletion boundaries are distributed across the upstream of left crRNA; in right, the 5′ deletion boundaries are distributed across the upstream of right crRNA. No, no obvious PCR bands were near the designed deletion boundaries.

Site	dCas9	nCas9	SpCas9	Two crRNAs in different directions
EMX1	−124 bp	+285 bp	+243 bp	+966 bp in left, +118 bp in right
CCR5	−407 bp	+237 bp	No	No
DMD	−316 bp	+155 bp	No	No
ERCC4	−187 bp	+330 bp	+165 bp	+724 bp in left, +259 bp in right

To further examine whether other dCas9 homologs can also be used to prevent Cas3 from reeling and looping the target dsDNA, two inactivation proteins, Cje3Cas9 (16) and SpaCas9 (17), were used in this study. Our analysis revealed that 5′ deletion boundaries were present within the downstream regions of the dCje3Cas9 sgRNAs (0.356 kb for EMX1 and 0.145 kb for ERCC4) and dSpaCas9 sgRNAs (0.427 kb for EMX1 and 0.47 kb for ERCC4) for the 2-kb deletion programmed by dCas9-controlled CRISPR/Cas3 (fig. S4 and table S6). These results demonstrate that the dCas9 homologs of Cje3Cas9 and SpaCas9 can also achieve accurate large-fragment deletions.

dCas9-controlled CRISPR/Cas3 can achieve precise large genomic deletions of XIST and 10q26 in HEK293T cells

The long noncoding RNA (lncRNA) X-inactive specific transcript, an important lncRNA derived from the XIST gene in mammals, is approximately 32 kb in length (18, 19). This gene plays a crucial role in silencing the transcription of a multitude of genes across an entire chromosome (18, 19). Here, deleting the whole locus of the XIST gene was attempted by using dCas9-controlled CRISPR/Cas3 systems with a single crRNA in HEK293T cells (Fig. 2A). PCR amplicons of approximately 1.2 kb in length were detected via gel electrophoresis in the dCas9-controlled CRISPR/Cas3 group (Fig. 2B), but these amplicons were not detected in the WT group. Thus, precise genomic deletions of XIST were successfully achieved by using dCas9-controlled CRISPR/Cas3 in human cells (Fig. 2C), as confirmed by T-clone and Sanger sequencing.

Fig. 2. Precise large-fragment deletions were achieved in XIST and chromosome 10q26 by dCas9-controlled CRISPR/Cas3 in HEK293T cells.

(A) Schematic of the CRISPR/Cas3 and dCas9-controlled CRISPR/Cas3 systems targeting XIST in human. The sgRNAs used for dCas9 and the crRNAs used for Cas3 are highlighted in pink and blue, respectively. F and R represent the PCR primers (red arrows) used for genotyping the targeted deletions. (B) Electrophoresis of the PCR products of the deletion and precise deletion by the CRISPR/Cas3 and dCas9-controlled CRISPR/Cas3 systems, respectively. The expected bands, which were cloned and inserted into pGM-T for Sanger sequencing, are marked with rectangular boxes. Cells transfected with CRISPR/Cas3 alone served as a negative control. (C) T-clone and Sanger sequencing results of the PCR products in (B). The sgRNAs of dCas9 and the crRNAs of Cas3 are highlighted in pink and blue, respectively. (D) Schematic of the CRISPR/Cas3 and dCas9-controlled CRISPR/Cas3 systems targeting fragment deletion of human chromosome 10q26. (E) Schematic for crRNAs of Cas3 and sgRNAs of dCas9 designed for fragment deletion of human chromosome 10q26. The sgRNAs of dCas9 and the crRNAs of Cas3 are highlighted in pink and blue, respectively. F and R represent the PCR primers (red arrows) used for genotyping the targeted deletions. (F) Electrophoresis of the PCR products. The expected bands, which were cloned and inserted into pGM-T for Sanger sequencing, are marked with rectangular boxes. The bars indicate the boundaries of deletions. The sgRNAs of dCas9 and the crRNAs of Cas3 are highlighted in pink and blue, respectively. Cells transfected with CRISPR/Cas3 alone served as a negative control.

To date, chromosomal deletions have emerged primarily from clinical observations (20–22). For instance, 10q26 deletion syndrome is a cytogenetic abnormality caused by the interstitial or terminal deletion of the long arm of chromosome 10 (23, 24). Approximately 100 cases of terminal 10qter deletions have been reported worldwide (25). Here, the 10q26 deletion mutations were mimicked using dCas9-controlled CRISPR/Cas3 systems in HEK293T cells (Fig. 2, D and E). The 860-kb deletions were induced by dCas9-controlled CRISPR/Cas3 via two crRNAs, and the 5′ deletion boundary was observed at the 0.06-kb downstream region of dCas9 (Fig. 2F), as confirmed by T-clone and Sanger sequencing. These results collectively indicated that the 10q26 deletion can be precisely mimicked by dCas9-controlled CRISPR/Cas3.

Establishment of a Turner syndrome mouse model by Y chromosome elimination

To determine whether CRISPR/Cas3 systems can be used to establish a Turner syndrome mouse model by eliminating the Y chromosome (26), three crRNAs targeting the Y chromosome were designed (Fig. 3, A and B). The elimination of Y chromosome–specific genes located in both the short arm and long arm of the Y chromosome was confirmed through PCR genotyping using genomic DNA isolated from each pup (Fig. 3C and fig. S5). Moreover, the expression levels of the Zfy1, Usp9y, Rbmy, and Ddx3y genes significantly decreased, while the expression of the autosome genes Selenos, Padi2, and Derl1 did not significantly differ between knockout (KO) and WT male mice (Fig. 3D). As anticipated, all the gene-edited mice were female (six of six), characterized by the presence of female genitals and nipples. Then, XO karyotypes with 39 chromosomes were determined in 58% of the tested cells by karyotyping (Fig. 3, E and F). Moreover, WGS revealed 65% and 18% efficiency of elimination of the Y chromosome in two KO mice (Fig. 3, G to I), and compared to the WT mice, the KO mice showed normal body weights and an expected survival rate (Fig. 3, J and K). Overall, these results showed that complete elimination of the Y chromosome can be achieved by using CRISPR/Cas3 in mice.

Fig. 3. Establishment of a Turner syndrome mouse model by Y chromosome elimination.

(A) Schematic of Cas3 mRNA, Cascade mRNA, and specific crRNAs injected into individual mouse zygotes. (B) Schematic diagram of three crRNAs of CRISPR/Cas3 located on the Y chromosome. (C) Genotyping analysis of gene-edited mice. Sry and Mecp2 are located on the Y chromosome and X chromosome, respectively. XY, male mice. XX, female mice. (D) The qRT-PCR results showed a significant down-regulation of Y chromosome genes (Zfy1, Usp9y, Rbmy, and Dxd3y) but no difference in autosome genes (Selenos, Padi2, and Derl) in KO mice compared with WT mice (n = 3 per group). The data are presented as the mean ± SD. **P < 0.01; ***P < 0.001; ns (not significant), P > 0.05 (Student’s two-tailed unpaired t test). (E) Representative karyotype image of a KO mouse and a WT mouse. KO, knockout mice generated using CRISPR/Cas3. (F) Karyological characteristics of the KO mice. (G) WGS data analysis workflow for determining the editing efficiency of CRISPR/Cas3 in eliminating the Y chromosome. (H) Schematic of Y chromosome elimination in #Y6 and #Y10 KO mice by WGS. The efficiency of Y chromosome elimination in mice was calculated as (W − K)/K, where K is the average number of reads on the Y chromosome and W is the average reads of the whole genome. (I) Editing efficiency of eliminating the Y chromosome in #Y6 and #Y10 KO mice. (J) Body weight comparison of KO mice and WT mice from 1 to 7 weeks. The data are presented as the mean ± SD. (K) Survival curves of KO and WT mice.

Establishment of a precise Sry retention mouse model by dCas9-controlled CRISPR/Cas3

To investigate the precise large-fragment deletions induced by dCas9-controlled CRISPR/Cas3 in mouse zygotes, we investigated three sites in this study. The results showed that dCas9-controlled CRISPR/Cas3 can generate precise large deletions in mouse zygotes (fig. S6 and table S7).

To examine whether the precise retention of the Sry gene, which is responsible for initiating male sex determination (27), could be achieved by dCas9-controlled CRISPR/Cas3 in mice, we designed sgRNAs for dCas9 located at the 3′ end of the Sry gene (Fig. 4A). The toe tissues of the mouse pups were isolated for PCR genotyping (Fig. 4A). The PCR and Sanger sequencing results showed single bands (Fig. 4B), and 5′ deletion boundaries were observed within the 32- to 35-bp (base pair) downstream region of the dCas9 sgRNA in dCas9-controlled CRISPR/Cas3 mice (Fig. 4C). The retention of Sry and large-fragment deletions of the Y chromosome were detected in five of eight (62.5%) pups, which were further identified as male mice by gel electrophoresis and the presence of male genitals (Fig. 4, D and E); however, the external genitalia of these pups were smaller than those of WT male mice (Fig. 4, F and G). In addition, typical reproductive dysfunction was evidenced by increased numbers of shrunken spermatogonia with pyknotic nuclei in dCas9-controlled CRISPR/Cas3 mice compared to WT male mice, as determined by hematoxylin and eosin (H&E) staining (Fig. 4H). Although the typical phenotype of short stature is found in Turner syndrome (28), the KO and dCas9-controlled CRISPR/Cas3 mice exhibited normal body weights and survival rates (Fig. 4, I and J). In summary, precise Y chromosome elimination was achieved by dCas9-controlled CRISPR/Cas3 in mice.

Fig. 4. Establishment of a precise Sry retention mouse model by dCas9-controlled CRISPR/Cas3.

(A) Schematic of the strategy for precisely deleting the mouse Y chromosome using dCas9-controlled CRISPR/Cas3. The sgRNAs of dCas9 and the crRNAs of Cas3 are highlighted in pink and blue, respectively. The targeted deletion is shown as a black line. F and R represent the PCR primers (red arrows) used for genotyping. (B) PCR-based genotyping of precise deletions. Genomic DNA isolated from dCas9-controlled CRISPR/Cas3 pups was subjected to PCR. The arrowhead indicates the region containing the exact deletion. (C) Sanger sequencing results of mice generated by using dCas9-controlled CRISPR/Cas3. (D) Genotyping analysis of dCas9-controlled CRISPR/Cas3 mice via PCR. Sry and Mecp2 are located on the Y chromosome and X chromosome, respectively. 1 to 8, gene editing mice; XY, male mice; XX, female mice. (E) Sex ratio of the mice generated by using dCas9-controlled CRISPR/Cas3. (F) Twelve-week-old dCas9-controlled CRISPR/Cas3 mice had smaller male genitals than WT mice (red arrows). (G) The testes, penis, and whole male reproductive system were smaller in the dCas9-controlled CRISPR/Cas3 mouse than in the WT mouse. (H) H&E-stained testes tissue from dCas9-controlled CRISPR/Cas3 and WT mice. An increased number of shrunken spermatogonia with pyknotic nuclei (red arrows) were observed in the dCas9-controlled CRISPR/Cas3 mice. Scale bar, 100 μm. (I) Body weight comparison of dCas9-controlled CRISPR/Cas3 mice and WT mice from 1 to 7 weeks. The data are presented as the mean ± SD. (J) Survival curves of dCas9-controlled CRISPR/Cas3 and WT mice.

DISCUSSION

Although the type I-E CRISPR system has been reported to induce phage plasmid degradation in Escherichia coli (29) and large-scale genome editing in human cells (7, 30), further investigation is needed to determine the optimal size of Cas3-mediated DNA deletions. Here, we exploited dCas9, a site-specific DNA binding protein serving as a roadblock, to efficiently prevent Cas3 from reeling and looping the target dsDNA and generate precise large-fragment deletions. We noticed that the 5′ deletion boundaries were consistently found within the ~0.5-kb downstream region of the dCas9 sgRNAs; this observation may explain the absence of notable editing events at 0.01 kb or 0.1 kb during dCas9 blockade (fig. S3). The above results proved that dCas9-controlled CRISPR/Cas3 systems can be adapted for precise genome editing applications in eukaryotic cells and mice. Additionally, we propose that this strategy can be extended to other class I CRISPR systems for precise deletion. The utilization of dCas9-controlled CRISPR/Cas3 could broaden the potential applications of other class I CRISPR systems as genome editing technologies in the future.

The performance of Cas3-mediated human genome editing across an extended distance differs substantially from that of by Cas9 and Cas12 during localized editing (31, 32). Heterogeneity manifests in the wide distribution of deletion sizes and distal endpoints. The onset of deletions was also nonuniform (7, 30), spreading within a predictable range of ~0.5 kb or ~0.4 kb downstream of the target site at four loci in human cells in this study (Fig. 1, D and G). Overall, type I-E CRISPR is unidirectional, accompanied by rare small indels and a predictable range of onset points.

Recently, three alternative methods have been developed, which also enable the modeling of large-fragment deletions. PEDAR and PRIME-Del have achieved high precision in programming deletions spanning from 0.02 to ~10 kb in length (8, 9) but with a lower editing efficiency than that of the dCas9-controlled CRISPR/Cas3 systems. Moreover, the process of cloning a pair of prime editing guide RNAs (pegRNAs) in tandem poses greater challenges than does the process of cloning sgRNA pairs. The length of each pegRNA ranges from 135 to 140 bp. Therefore, combining their unique component sequences into a single long oligonucleotide approaches the limits of traditional DNA synthesis techniques. CRISPR/Cas9 was used to engineer the Y chromosome in mice by using two sgRNA-targeted repeat sequences in another study (33), while the crRNAs used in this study were randomly selected from the Y chromosome. Our approach has the advantage that CRISPR/Cas3 systems are universal and are not restricted to gene loci with highly repetitive sequences. In addition, the CRISPR/Cas3-mediated large-fragment deletion protocol is less laborious. Y chromosome elimination was achieved by using three crRNAs in our study, whereas 14 sgRNAs were needed in previous study (33).

Here, gene-edited mice were generated by the class I CRISPR system. Compared with their WT counterparts, the dCas9-controlled CRISPR/Cas3 mice exhibited reproductive impairments (Fig. 4, F to H). These impairments mimicked the symptoms that are commonly observed in patients with Turner syndrome (34, 35). These results suggest that dCas9-controlled CRISPR/Cas3 systems hold potential as therapeutic interventions for aneuploid diseases, including Down syndrome (DS), Klinefelter syndrome, and XYY syndrome (33, 36–38). Furthermore, recent studies have indicated that loss of the Y chromosome (LOY) in aging men is associated with elevated mortality rates and an increased incidence of hematologic and bladder cancer, uveal melanoma, and Alzheimer’s disease (39–41). Consequently, the animal model of Cas3-mediated KO of the entire Y chromosome proves instrumental in elucidating the role of the Y chromosome in various physiological contexts.

Despite the promising aspects of the CRISPR/Cas3 system, several challenges still need to be addressed: (i) The delivery of multiple effector components to animal tissues using adeno-associated virus (AAV) systems is a considerable challenge. Genetic engineering by type I-E Cascade-Cas3 requires six Cas genes and a CRISPR array, with a 7- to 8-kb total gene size that is 60 to 80% larger than that of the commonly used ~4.2-kb Streptococcus pyogenes Cas9. Similarly, type I-B is approximately 6 to 7 kb (42), I-C is 5 to 7 kb (43, 44), and type I-D is 7 to 8 kb (45, 46). For AAV vectors, which have a 4.7-kb packing limit, type I systems cannot fit into a single AAV as compact as Cas9s, potentially requiring a dual AAV strategy (44). 2. The possibility of extensive chromosomal rearrangements must be carefully considered and monitored, similar to the challenges encountered with Cas9 (47, 48). Hence, further research and engineering of Cas3 or the Cascade are still needed.

In summary, this study describes the precise large-fragment deletions of dCas9-controlled CRISPR/Cas3, thereby expanding the potential applications of this technique as a genome editing technology, including epigenome editing and the tagging of multiple domains at specific genomic loci. Furthermore, this study highlights the validity of class I CRISPR systems as genome editing tools in mammals. Precise large-fragment deletions hold promise for combating infectious viruses or pathogenic bacteria, including herpes simplex virus and Epstein-Barr (EB) virus. Overall, CRISPR/Cas3 systems may exert substantial influence on genetic research by enabling programmable gene regulation and offering innovative strategies for genome-wide screenings and strain engineering.

MATERIALS AND METHODS

Ethics statement

Institute for Cancer Research (ICR) mice were obtained from the Laboratory Animal Center of Jilin University (Changchun, China). All animal studies were conducted under the guidance of the Animal Welfare and Research Ethics Committee at Jilin University (SY202301003).

Plasmid construction

The pCAG-All-in-one-hCascade (#134919), pPB-CAG-hCas3 (#134920), and pBS-U6-crRNA-empty (#134921) plasmids were obtained from Addgene. Site-directed mutagenesis of the nCas9, dCas9, dCje3Cas9, and dSpaCas9 plasmids was performed using a Fast Site-Directed Mutagenesis Kit (Tiangen, Beijing, China). The site-specific primers used for mutation are listed in table S8. All the target sequences are listed in table S9. All crRNA expression plasmids were created by inserting 32-bp double-stranded oligonucleotides at the Bbs I restriction site and subsequently transformed into E. coli (DH5α) (Sangon Biotech, Shanghai, China).

Cell lines

HEK293T cell lines stably expressing SpCas9, nCas9, and dCas9 were constructed by using a piggyBac transposase as previously reported (49). The primers used to identify the stably expressing cell lines are listed in table S10.

Cell culture, transfection, and large-fragment detection by PCR

The HEK293T, LO2, A549, MEF, and PEF cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (HyClone), 2 mM GlutaMAX (Life Technologies), penicillin (100 U/ml), and streptomycin (100 mg/ml) and incubated at 37°C in an incubator containing 5% CO₂. The cells were seeded into six-well plates (Jet, Guangzhou, China) at a density of 120,000 per well and transfected using Hieff Trans Liposome (Yeasen, Shanghai, China). A total of 3000 ng of sgRNA was transfected into the cell lines per well. After 24 hours, a total of 1000 ng of Cas3, 1000 ng of Cascade, and 1000 ng of crRNA were transfected into cells per well. Puromycin (3 μg/ml, Meilunbio, Dalian, China) was used to enrich the positive cells after 24 hours of transfection. The isolated DNA was PCR-amplified with the TIANamp Genomic DNA Kit (Tiangen, Beijing, China) according to the manufacturer’s instructions. The PCR products were gel-purified, cloned, and inserted into the pGM-T vector (Tiangen, Beijing, China). The target sites, sgRNAs, crRNAs, and primers used for genotyping are listed in tables S9 and S11.

Cell electroporation

Approximately 15 μg of sgRNA expression plasmids, 15 μg of crRNA expression plasmids, 20 μg of Cas3 expression plasmids, 20 μg of cascade expression plasmids, and 20 μg of dCas9 expression plasmids were cotransfected into 3 × 10⁶ MEFs or PEFs using the Neon transfection system. The electroporation conditions for MEFs or PEFs were as follows: 1360 V, 30 ms, and 1 pulse.

In vitro transcription

pCAG-All-in-one-hCascade, dCas9, and pPB-CAG-hCas3 mRNA was synthesized using an in vitro RNA transcription kit [HiScribe T7 ARCA mRNA kit (with tailing), NEB]. The sgRNAs were amplified and transcribed in vitro using the MAXIscript T7 kit (Ambion; Applied Biosystems, CA, USA) and purified with the miRNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The crRNA oligo sequences are listed in table S9.

Microinjection of mouse zygotes and embryo transfer

Briefly, a mixture of Cas3 mRNA (50 ng/μl), Cascade mRNA (50 ng/μl), dCas9 mRNA (50 ng/μl), and sgRNA/crRNA (50 ng/μl) was coinjected into the cytoplasm of pronuclear-stage mouse zygotes. The injected mouse zygotes were cultured at 37°C in an incubator with 5% CO₂ until the two-cell stage, after which approximately 30 to 50 injected embryos were transferred into the oviduct of the recipient mother.

Genotyping by PCR

At the blastocyst stage, DNA was extracted from each injected embryo with embryo lysis buffer (Beyotime Biotechnology, Shanghai, China) at 56°C for 60 min and 95°C for 10 min on a Bio-Rad PCR machine. Then, the extracted products were amplified via PCR under the following thermocycling conditions: 95°C initial denaturation for 5 min; 42 cycles of 95°C denaturation for 30 s, 58°C annealing for 30 s, and 72°C extension for 30 s; and a final extension at 72°C for 5 min. The PCR products were gel-purified with a TIANgel Midi Purification Kit (Tiangen, Beijing, China), cloned, and inserted into pGM-T (Tiangen, Beijing, China). Ten positive plasmid clones were sequenced and analyzed by SnapGene software. The PCR primers used for amplification and mutation detection are listed in table S11.

T-clone and Sanger sequencing

The PCR products were gel-purified using a TIANgel Midi Purification Kit (Tiangen, Beijing, China) and subsequently cloned and inserted into the pGM-T vector using a pGM-T Cloning Kit (Tiangen, Beijing, China). Ten single positive T-clones were subjected to Sanger sequencing performed by Sangon Biotech (Shanghai, China).

RNA extraction and qRT-PCR

Total RNA was isolated from the testes of WT and dCas9-controlled CRISPR/Cas3 mice with TRIgent (Mei5 Biotech, Beijing, China) and treated with deoxyribonuclease (DNase) I (Tiangen, Beijing, China). First-strand cDNA was synthesized using a cDNA first strand synthesis kit (TIANGEN, Beijing, China) and used for quantitative reverse transcription PCR (qRT-PCR) analyses to evaluate the expression of the genes on the Y chromosome. Target mRNA expression levels were normalized to those of Gapdh. The sequences of the primers used are listed in table S11. All the experiments were repeated three times for each gene. The data are expressed as the mean ± SD.

Histological analysis

Y chromosome KO and WT mice were euthanized at 12 weeks, and testes were fixed in 4% paraformaldehyde. Then, the tissue was dehydrated in increasing concentrations of ethanol (70% for 6 hours, 80% for 1 hour, 96% for 1 hour, and 100% for 3 hours), cleared in xylene, and embedded in paraffin for histological analysis. H&E staining was performed as previously described (50, 51), and the stained sections were analyzed under a light microscope (Nikon ts100) (Servicebio, Wuhan, China).

Karyotype analysis

Fibroblasts were derived from mouse skin and were cut into small pieces and cultured for 7 days. Then, the fibroblasts were incubated with demecolcine (200 ng/ml) (Sigma, USA) for 1 hour, resuspended in 0.075 M KCl at 37°C for 10 to 30 min, incubated with carboy’s fixative (25% acetic acid in methanol) for 30 min, and plated on precleaned slides. For chromosome counting, the slides were stained with Hoechst 33342. For G banding, the slides were incubated with 0.025% pepsin and then stained with Giemsa for 15 min. More than 10 metaphase spreads were analyzed in this study (Hangzhou Kayotapu Biological Technology Co. Ltd.).

Whole-genome sequencing

Genomic DNA was extracted from transfected HEK293T cells, and a DNA library was prepared with the HiSeq PE Cluster Kit v4-cBot-HS (Illumina). Genomic sequence analysis was performed with the NovaSeq 6000 S4 Reagent Kit V1.5 (BGI Americas Corporation, Cambridge, MA, USA). Raw reads from each sample were mapped with the human genome assembly GRCh38 by using Burrows-Wheeler Aligner software following the recommended best practices for variant analysis with the Genome Analysis Toolkit (GATK). Genomic variations were detected by the HaplotypeCaller tool of GATK, and the variant quality score recalibration method was applied to obtain high-confidence variant calls (52, 53). The efficiency (E) of Cas3-mediated deletions was calculated as follows: E = (A − B)/A. The proportion (F) of precision deletions by dCas9-controlled CRISPR/Cas3 was calculated as follows: F = [(A − B) − (A − C)]/(A − B), where A is the average reads 1 to 2 kb downstream of the crRNAs, B is the average reads of the crRNAs at the sites, and C is the average reads of 0.5 kb downstream of the sgRNAs. The analysis method is detailed in Fig. 1 (E, F, and H).

High-quality genomic DNA was extracted from KO mice according to a standard phenol:chloroform extraction and ethanol precipitation protocol. A genomic DNA library was prepared for sequencing using the standard PacBio Sequel protocol. Each accession was sequenced on a HiSeq 2500 Illumina machine using 150-base paired-end reads (Anoroad, Guangzhou, China). The aligned and cleaned data of each sample were mapped to the mouse reference genome (GRCm39) using Burrows-Wheeler Aligner software following the recommended best practices for variant analysis with the GATK to ensure accurate variant calling. The efficiency (P) of Y chromosome elimination in the KO mice was calculated as follows: P = (W − K)/K, where K is the average reads of the Y chromosome from KO mice and W is the average reads of the whole genome of KO mice. The analysis method is detailed in Fig. 3 (G and H).

Drawing tools

The diagrams of the gene-edited mice were generated using Figuredraw (https://www.figdraw.com).

Statistical analysis

The data are expressed as the mean ± SD of at least three individual experiments. The data were analyzed by Student’s two-tailed unpaired t test using GraphPad Prism software 8.0. A probability of P < 0.05 was considered statistically significant.

Supplementary Materials

This PDF file includes:

Figs. S1 to S6

Tables S1 to S11