Efficient and versatile CRISPR engineering of human neurons in culture to model neurological disorders

Abstract

The recent identification of multiple new genetic causes of neurological disorders highlights the need for model systems that give experimental access to the underlying biology. In particular, the ability to couple disease-causing mutations with human neuronal differentiation systems would be beneficial. Gene targeting is a well-known approach for dissecting gene function, but low rates of homologous recombination in somatic cells (including neuronal cells) have traditionally impeded the development of robust cellular models of neurological disorders. Recently, however, CRISPR/Cas9 gene editing technologies have expanded the number of systems within which gene targeting is possible. Here we adopt as a model system LUHMES cells, a commercially available diploid human female mesencephalic cell line that differentiates into homogeneous mature neurons in 1-2 weeks. We describe optimised methods for transfection and selection of neuronal progenitor cells carrying targeted genomic alterations using CRISPR/Cas9 technology. By targeting the endogenous X-linked MECP2 locus, we introduced four independent missense mutations that cause the autism spectrum disorder Rett syndrome and observed the desired genetic structure in 3-26% of selected clones, including gene targeting of the inactive X chromosome. Similar efficiencies were achieved by introducing neurodevelopmental disorder-causing mutations at the autosomal EEF1A2 locus on chromosome 20. Our results indicate that efficiency of genetic "knock-in" is determined by the location of the mutation within the donor DNA molecule. Furthermore, we successfully introduced an mCherry tag at the MECP2 locus to yield a fusion protein, demonstrating that larger insertions are also straightforward in this system. We suggest that our optimised methods for altering the genome of LUHMES cells make them an attractive model for the study of neurogenetic disorders.

Keywords: CRISPR; Cas9; EEF1A2; LUHMES; MeCP2; human; neurological; neurons.

Figure 1.. Nucleofection for efficient transfection of LUHMES cells.

( A) The aims of this study are to develop methods for genetic manipulation of the LUHMES pre-neuronal cell line, which in itself requires methods of transient transfection to be developed. After successful genetic modifications have been confirmed, LUHMES cells can then be differentiated into mature neurons for study. ( B) Representative images of plasmid transfection of LUHMES cells using program D33 with a Nucleofector device. Pictures were taken 47 hours post-Nucleofection. Scale bar is 200 μm.

Figure 2.. Generation of MeCP2 knock-out LUHMES cell lines.

( A) Schematic representation of the MECP2 locus with the targeting sequence of sgRNA A and sgRNA B labelled. Arrowheads indicate sites of double-strand break. ( B) Schematic representation of the plasmid used to deliver Cas9 and sgRNA to LUHMES cells. ( C) T7E1 assay of sgRNA A and sgRNA B. WT: non-transfected wild-type cells. NDC: wild-type cells Nucleofected without any DNA. Asterisks indicate non-specific bands. Arrows indicate bands specific to samples containing Cas9 and sgRNA. ( D) Sequencing of genomic DNA from two LUHMES single-cell clones. Wild-type genomic DNA is in lower case at the top. ( E) Sequencing of cDNA from two LUHMES single-cell clones. Wild-type cDNA is in lower case at the top. ( F) Sequencing of genomic DNA from a single LUHMES unedited clone that was transfected with Cas9 and sgRNA B. Wild-type gDNA is in lower case at the top. ( G) Western blot of MeCP2 protein and Histone H3 loading control from wild-type cells (WT), wild-type cells that went through the Nucleofection and cloning process (WTC), KO1 and KO2 cell lines. ( H) Sequencing of sgRNA off-target sites in KO1 and KO2 cell lines. Numbers next to each locus name indicate the off-target score as determined by crispr.mit.edu.

Figure 3.. Generation of a human neuronal cell line containing a Rett syndrome-causing missense mutation in MECP2.

( A) Schematic representation of the MECP2 locus with the sgRNA 1 target sequence labelled and the ssODN 1 donor molecule with point mutation alterations indicated in upper case. The site of double strand break is labelled with two arrowheads and the distance between the point mutation of interest and the double-strand break site is indicated. ( B) Schematic representation of the RFLP screening assay used for identifying positive knock-in clones. Mutation of arginine at position 306 to cysteine results in the introduction of a novel target sequence for the restriction enzyme HpyCH4V. Primers used for PCR amplification are labelled. ( C) HpyCH4V digests of the PCR product ( Supplementary Figure 3A) to identify clones that have gained a novel HpyCH4V target sequence. A positive clone is identified with an asterisk. ( D) Sequencing of genomic DNA and cDNA from the RFLP-positive cell line confirms the cell line to be MECP2-R306C. ( E) Sequencing of the top two off-target sites for sgRNA 1 in the R306C cell line. Number next to the locus name is the off-target score as calculated by crispr.mit.edu.

Figure 4.. Generation of human neuronal cell lines containing missense mutations that cause neurological disorders.

( A) Schematic representation of the MECP2 locus with the sgRNA 2 target sequence labelled. Site of double-strand break is indicated with arrowheads. ssODN molecule is shown with point mutation changes highlighted in upper case and the DSB to point mutation distance labelled. ( B) Genomic DNA sequencing of two R306C positive clones. Both clones have the correctly inserted R306C point mutation as well as downstream deletions induced by sgRNA 2. The site of DSB is indicated by an arrowhead. ( C) Schematic representation of a plasmid containing Cas9, sgRNA and EGFP expression constructs. ( D) Genomic DNA sequencing of two R111G positive cell lines. ( E) Genomic DNA sequencing of two R133C positive cell lines. ( F) Genomic DNA sequencing of two T158M positive cell lines. ( G) Schematic representation of part of the EEF1A2 locus, with sgRNA target sequences labelled and the ssODN donor molecule with point mutation alterations indicated in upper case. Sites of single-strand nicks are indicated with an arrowhead. Distances between each nick and the point mutation of interest are labelled. ( H) Genomic DNA sequencing of the EEF1A2-D252H positive cell line.

Figure 5.. Endogenous knock-in of an mCherry tag into the MECP2 locus in LUHMES cells.

( A) Schematic representation of the MECP2 locus with the sgRNA C target region labelled, the targeting vector, and the recombined genomic DNA allele. Positions of primers used for screening purposes in ( C) are indicated. ( B) Flow cytometry analysis of WT cells and cells that were transfected with Cas9, sgRNA and targeting plasmid. ( C) PCR screening of genomic DNA from single-cell clones that were identified as being mCherry positive by flow cytometry. ( D) Immunofluorescence imaging of WT cells and two single-cell clones using DAPI and antibodies probing for MeCP2 and MAP2. Images are slices through a z-stack. Scale bar represents 50 μm. ( E) Immunoblot analysis of WT cells and four MECP2-mCherry positive clones. ( F) Immunoblot analysis of WT cells, two MECP2-mCherry positive clones and two MECP2-mCherry negative clones. ( E + F) Top panel probe: MeCP2. Middle panel probe: mCherry. Bottom panel probe: Histone H3 as a loading control.

Figure 6.. Analysis of range of mutations induced by CRISPR/Cas9 in LUHMES cells.

( A+ B) Partial recombination can occur with the ssODN molecule. ( A) Three single cell clones from the R111G targeting experiment, two of which contain a partially recombined allele. ( B) Two single cell clones from the R133C targeting experiment; clone 1 has an allele with 2/3 of the KI residues, clone 2 has an allele with only 1/3 of the KI residues. ( C) Genomic DNA sequencing of a single clone reveals one allele to contain two point mutations and the other allele to contain a 6bp deletion, yet cDNA sequencing identifies the HDR to have occurred with the allele on the inactive X chromosome. ( D) cDNA sequencing identifies the allele on the active X chromosome having undergone recombination, with all three point mutations integrated plus a single base pair insertion. ( E) Sequencing of single-cell clones reveals duplication events occurring as a result of multiple recombination events with the ssODN. ( A, B, C, D, E) mPAM - silent point mutation that abolishes the PAM site for each sgRNA. Rett-syndrome causing point mutations of interest and silent restriction enzyme target sequence point mutations are highlighted with arrows.

Figure 7.. Simple targeting pipeline for generation of genetically modified LUHMES cell lines using CRISPR technology.

Cells are Nucleofected with the plasmids and ssODNs necessary for the specific targeting experiment and taken for FACS sorting two days later. After approximately 10 days single cell colonies will have expanded enough to require dissociation and transfer to 24-well plates. From 24-well plates cells can be frozen down for storage in liquid nitrogen and some cells harvested for genomic DNA. The first step of the screening strategy is to perform an RFLP assay to identify a subset of clones that will be taken for genomic DNA sequencing which will identify cell lines that are positive for a clean KI. For more detailed information see methods and materials.