This article reviews research in the Rodriquez-Perales and Torres Ruiz laboratory, and their publication: Torres-Ruiz R, Martinez-Lage M, et al. (2017). Efficient recreation of t(11:22) EWSR1-FLl1 in human stem cells using CRISPR/Cas9. Stem Cell Reports, 8(5):1408–1420 .
Creating biological models for translocations
Genetically modified cell and animal models that reproduce chromosomal translocations have been critical for understanding how these genomic rearrangements initiate cancers (see the sidebar, Chromosomal 3D structure and genomic rearrangements, below). While scientists have successfully used traditional gene-targeting strategies (such as P1 Cre-loxP [2,3], zinc-finger nucleases , and TALENs ) to generate translocations, CRISPR-Cas9 genome editing approaches have proven more efficient in some cell types, reducing the arduous selection procedures needed with these earlier methods. Indeed, researchers have employed CRISPR-Cas9 reagents to produce specific translocations, both in vitro  and in vivo [7,8]. These groups used 2 guide RNAs (gRNAs) to target the double-strand breaks (DSBs) in the 2 genes involved in a particular translocation. The researchers relied on the endogenous DNA repair machinery to complete the chromosomal joining.
Though CRISPR-Cas9 technologies have greatly facilitated the creation of cell lines with specific translocations, isolating the specific clones with the desired genomic rearrangements still requires time-consuming screening, especially when working with stem cells. As part of the Molecular Cytogenetics Group that studies chromosomal aberrations typical of different human cancers at Centro Nacional de Investigaciones Oncologicas (CNIO; Madrid, Spain), Dr Sandra Rodriguez-Perales and Dr Raúl Torres Ruiz have been designing CRISPR genome editing strategies to generate cellular models for acute lymphocytic leukemia (ALL) and Ewing sarcoma chromosomal translocations (see the sidebar, Chromosomal 3D structure and genomic rearrangements, for a depiction of the translocation attributed to Ewing sarcoma). While their initial experimental designs successfully recreated these specific genomic rearrangements in HEK-293, these methods were not as efficient in reproducing the translocations in the primary cells in which the scientists hoped to develop model cell lines, specifically, in human mesenchymal stem cells (hMSC), hematopoietic progenitor CD34 stem cells (HPSCs), and human-induced pluripotent stem cells (hiPSC).
Co-delivery of CRISPR components increases editing efficiency
Subsequent testing focused on improving efficiency of CRISPR-Cas9–mediated chromosomal translocations in human stem cells. These experiments convinced Rodriguez-Perales and Torres Ruiz that the most successful approach involved concurrent cellular delivery of CRISPR components. The scientists designed and compared 2 methods for CRISPR component co-delivery:
- RNP complexes—Purified, recombinant Cas9 enzyme (Alt-R S.p. Cas9 Nuclease 3NLS, IDT) bound together as a ribonucleoprotein (RNP) complex prior to delivery, with in vitro transcribed gRNAs targeting each of the desired breakpoints.
- An all-in-one expression plasmid—Transcription of both gRNAs driven by distinct RNA polymerase III promoters (U6 and H1), and expression of Cas9 with GFP 2A self-cleaving peptide driven by CMV promoter. Two nuclear location signal (NLS) sequences, which facilitate protein transport into the cell nucleus, were included in the plasmid, resulting in a 12X increase in genomic rearrangements.
Both of these strategies additionally included concurrent delivery of 2 single-stranded oligodeoxynucleotides (ssODN; 150 nt Ultramer® Oligonucleotides, IDT), one targeting each derivative chromosome, to serve as splints for guiding ligation of the distinct-gene DSB ends. The ssODNs thus promote translocation formation over the otherwise more efficient rejoining of same-gene DSBs through non-homologous end-joining (NHEJ) activity (Figure 1).
Figure 1. Single-stranded oligodeoxynucleotides (ssODNs) facilitate translocation through CRISPR genome editing. Translocation takes place between the EWSR1 gene of chromosome 22 and the FLI1 gene of chromosome 11. Locations of gRNA binding and Cas9 cutting are shown. The respective ssODNs (Ultramer Oligonucleotides, IDT) facilitate alignment of EWSR1 and FLI1 cut ends. The resulting derivative chromosome EWSR1/FLI1 der(22) contains the upstream portion of EWSR1 (e1) joined to the downstream portion (f2) of FLI1. The resulting derivative chromosome FLI1/ EWSR1 der(11) contains the upstream portion of FLI1 (f1) joined to the downstream portion of EWSR1 (e2). [Figure courtesy of Rodriguez-Perales and Torres Ruiz, Centro Nacional de Investigaciones Oncologicas (CNIO; Madrid, Spain).]
Increased editing efficiency when CRISPR components are delivered as RNPs instead of expression plasmids
The CNIO researchers compared the efficiency of these 2 strategies to reproduce the t(11;22) Ewing sarcoma chromosomal translocation, thought to occasionally occur naturally in hMSCs . They also used HEK-293 cells in these experiments for optimization and control purposes. Delivery testing suggested that electroporation was the most efficient delivery method in hMSCs. Thus, the 2 sets of assembled RNPs ± ssODNs, targeting each of the desired breakpoints, or the all-in-one expression plasmid were transfected into hMSCs and HEK-293 cells using a Neon® transfection system (Thermo Fisher Scientific) (Figure 2).
Figure 2. Delivery of RNP complexes + ssODNs or an “all-in-one” expression plasmid + ssODNs. CRISPR reagents and ssODNs were nucleofected into hMSCs and HEK-293 cells in one of two formats: (A) as an RNP containing purified Cas9 enzyme complexed with in vitro transcribed gRNAs, or (B) as an all-in-one plasmid designed to express gRNAs and Cas9 nuclease. (Figure courtesy of Rodriguez-Perales and Torres Ruiz, Centro Nacional de Investigaciones Oncologicas (CNIO; Madrid, Spain.)
Cell colonies were screened for the desired translocation 48–72 hr after transfection, using independent methods. Multi-color fluorescence in situ hybridization (FISH) was used to verify translocation events (Figure 3). PCR and Sanger sequencing confirmed translocation at the genome level. RT-PCR was used to detect expression of derivative fusion genes, while western blotting was used to identify fusion protein.
Figure 3. FISH detection of EWSR1 rearrangement. A “break-apart” FISH probe that detects the sequences 3′ to the EWSR1 gene (red) and 5′ to the EWSR1 gene (green), respectively, was used to detect the EWSR1 rearrangement. The red and green signals are separated when a chromosomal break and translocation event occur at the breakpoint within the EWSR1 gene. Left interphase nucleus: negative cell showing 2 fusion signals from normal EWSR1 loci that have not been rearranged. Right interphase nucleus: positive cell showing both a fusion signal, and green and red split signals for a rearranged EWSR1 gene. [Figure courtesy of Rodriguez-Perales and Torres Ruiz, Centro Nacional de Investigaciones Oncologicas (CNIO; Madrid, Spain).]
The RNP format (using in vitro gRNA or Alt-R CRISPR-Cas9 System reagents), was 11X more efficient at successfully recreating the t(11;22) translocation in hMSCs than when the all-in-one plasmid was delivered. Co-delivery of ssODNs further enhanced translocation . The CNIO scientists also noted that off-target effects at known off-target sites in t(11;22) cells generated by RNP nucleofection were undetectable .
Subsequent experiments in hiPSCs likewise showed that co-nucleofection of RNP complexes and ssODNs gave 3.2X higher translocation frequency compared to co-electroporation of the expression plasmid and ssODNs. Unlike hMSCs, where the fusion gene is incompatible with viability, and thus, the genetically edited cells are only short lived, expanded t(11;22)+ hiPSC clones maintained the genomic rearrangement through the 8 weeks tested.
“It is hard to identify 2 gRNAs that produce the same cutting efficiency, likely due to location and chromatin structure. While we consistently see differences in cutting efficiency between both sites regardless of reagent delivery format, there is consistently higher cutting efficiency when Alt-R CRISPR reagents are delivered as RNP complexes vs. expression plasmid, and also a higher efficiency of translocations.” – Sandra Rodriguez-Perales, PhD, Centro Nacional de Investigaciones Oncologicas (CNIO), Madrid, Spain
The value of stem cell model systems and optimized CRISPR reagents
Traditional animal cancer models are important to biological and biomedical studies. However, it is difficult to simulate the real context of a human body due to species differences. hiPSCs overcome this limitation. CRISPR-Cas9 can be used to easily and rapidly recreate a chromosomal translocation in these human stem cells to model human cancer. The use of the CRISPR-Cas9 system can precisely introduce point mutations or gross genome rearrangements (translocation, insertion, or deletion) into human PSCs to provide an array of genetically edited human disease models.
- IDT provides optimized genome editing reagents and protocols (RNP delivery) that outperform other CRISPR approaches for producing on-target, double-stranded DNA breaks, which can be repaired to create substitutions, insertions, deletions, or in this case, translocations.
- The Ultramer DNA Oligonucleotides used in this study are available up to 200 bp from IDT. IDT also offers longer Megamer® Single-Stranded DNA Fragments between 200–2000 bp in length that can be used as donor DNA for homology directed repair (HDR) experiments.
- Using electroporation, the recommended RNP delivery format used by these researchers was shown to outperform delivery of a Cas9 and gRNA expression plasmid. IDT also offers an electroporation enhancer that further boosts electroporation of Cas9 RNPs and that is compatible with many electroporation instruments (enhancer not used in this study).