Reporter cell lines offer a means to visualize, track and isolate cells of interest from heterogeneous populations. However, gene-targeting using conventional homologous recombination in human embryonic stem cells is extremely inefficient. Herein, we describe targeting CNS midbrain specific transcription factor PITX3 locus with EGFP using zinc-finger nuclease enhanced homologous recombination.
One major limitation with current human embryonic stem cell (ESC) differentiation protocols is the generation of heterogeneous cell populations. These cultures contain the cells of interest, but are also contaminated with undifferentiated ESCs, non-neural derivatives and other neuronal subtypes. This limits their use in in vitro and in vivo applications, such as in vitro modeling for drug discovery or cell replacement therapy. To help overcome this, reporter cell lines, which offer a means to visualize, track and isolate cells of interest, can be engineered. However, to achieve this in human embryonic stem cells via conventional homologous recombination is extremely inefficient. This protocol describes targeting of the Pituitary homeobox 3 (PITX3) locus in human embryonic stem cells using custom designed zinc-finger nucleases, which introduce site-specific double-strand DNA breaks, together with a PITX3–EGFP-specific DNA donor vector. Following the generation of the PITX3 reporter cell line, it can then be differentiated using published protocols for use in studies such as in vitro Parkinson’s disease modeling or cell replacement therapy.
Current embryonic stem cell (ESC) differentiation protocols result in heterogeneous cell populations, particularly with respect to the derivation of specific neuronal phenotypes. Thus, although cultures contain the desired cellular phenotype, other neuronal, non-neuronal and un-differentiated cell types are often present1. These characteristics limit the application of ESC derived cell sources for use in cell replacement therapy and in vitro disease modeling.
Genetic reporter cell lines offer a means to visualize, track and isolate cells of interest, provided that the expression of the reporter protein (RP) reproduces endogenous expression. Use of promoters immediately upstream of a gene coding region can be used to derive crude genetic reporters but such constructs lack the precise regulatory elements that control endogenous gene expression. In contrast, homologous recombination offers the opportunity to ensure high-fidelity expression of the RP. In the past, targeting vectors designed for homologous recombination at specific loci of interest have been used to target RPs in mouse ESCs (mESCs) using electroporation as a means of DNA delivery1,2. However the generation of reporter cell lines via conventional homologous recombination is extremely inefficient for human ESCs (hESCs), and thus has only been documented in a handful of cases (reviewed in3). By using an engineered chimeric protein containing a fusion of a FokI nuclease with site-specific zinc-finger motifs, known collectively as zinc-finger nucleases (ZFNs), DNA double-strand breaks can be introduced at pre-determined genomic loci. When a DNA vector with homology to both sides of the DNA double strand break is added, the genomic site can be repaired by homologous recombination, allowing the incorporation of the DNA donor sequence. This technique has proved useful for genomic editing in both human primary cells4,5 and hESCs6,7. More recent work has utilized transcription activator-like effector nucleases (TALENs), transcription factors used by plant pathogens8, to aid in the design of site-specific nucleases9.
In the following protocol we demonstrate the generation of a hESC reporter cell line by electroporation of an EGFP containing homologous targeting vector6 together with ZFNs for the human Pituitary homeobox 3 (PITX3) locus. Following antibiotic selection for 2-3 weeks, hESCs with correctly integrated DNA can be manually picked, expanded and screened initially via PCR, and subsequently validated by Southern blotting.
All procedures are carried out in a sterile laminar flow hood. All media and solutions are equilibrated to 37 °C unless otherwise specified.
1. Expansion of Human PSCs (hESCs and hiPSCs, WA-09 used in this Protocol) for Transfection
Note: It is not necessary to expand hESCs on Matrigel or Geltrex. hESCs expanded on mouse embryonic fibroblasts (MEFs) can be used for electroporation or nucleofection following a MEF removal step (see section 2.8).
2. Preparation of Human ESCs for Transfection
3. Electroporation of hESCs
4. Picking Transgenic hESC Clones
5. Screening and Expansion of Positive Clones
Following co-electroporation of the custom designed PITX3 zinc-finger pair together with the PITX3–EGFP-specific DNA donor vector, and subsequent puromycin selection, positive hESC clones were initially detected via genomic PCR screening (Figure 1). Southern blot hybridization of these PCR positive clones with 5’ and 3’ specific probes confirmed correct targeting to exon 1 of the PITX3 locus (Figure 2), with an efficiency of 19%. Immunofluorescent images shown in Figure 3 demonstrate EGFP expression driven by the PITX3 promoter following differentiation using a published PA6 stromal cell differentiation protocol12,13. Co-localization of EGFP could be observed with markers of midbrain dopaminergic neurons such as FOXA2 (red) (Figure 3A) and tyrosine hydroxylase (TH; red) (Figure 3B), while importantly no co-localization could be observed with the GABAergic marker, gamma-aminobutyric acid (GABA; red) (Figure 3C).
Figure 1. Preliminary genomic PCR screening. Following colony picking and expansion, preliminary genomic PCR screening detected numerous positive clones as shown by the PCR product (band at 984 bp). (Clones were screened using primers hPITX3 L arm gen. F, which lies just outside the left targeting arm, and hPITX3 L arm GFP R which is located within the EGFP gene.) Right-hand and left-hand side lanes are a 1 kb ladder. Please click here to view a larger version of this figure.
Figure 2. Validation of EGFP targeting to the hPITX3 locus using zinc finger nucleases in hESCs. (A) Schematic of the Southern blot strategy employed to validate targeting. The top panel schematic illustrates the native hPITX3 locus before targeting. The bottom panel schematic depicts the hPITX3 locus following correct insertion of the EGFP transgene. The red lines under each panel illustrate the sites at which the 5’ and 3’ Southern blot probes will bind, and correspond to the bands in Figure 2B. (B) Representative Southern blot of correctly targeted colonies selected from PCR pre-screening. Abbreviations: EGFP, enhanced green fluorescent protein; PGK, human phosphoglycerol kinase promoter; PURO, puromycin resistance gene. Other features: loxP sites, purple; polyadenylation sequence, blue, hPITX3 exons, orange. Please click here to view a larger version of this figure.
Figure 3. hPITX3-EGFP hESCs differentiated under PA6/LSB/SAG/FGF8 conditions. Immunofluorescent images of EGFP labeled with (A) FOXA2 (red), (B) TH (red) and (C) GABA (red). Counterstained with DAPI (blue). Scales bars, 50 μm. Please click here to view a larger version of this figure.
Following co-electroporation of the custom designed PITX3 zinc-finger pair together with the PITX3–EGFP-specific DNA donor vector, and subsequent puromycin selection, positive hESC clones were initially detected via genomic PCR screening (Figure 1). Southern blot hybridization of these PCR positive clones with 5’ and 3’ specific probes confirmed correct targeting to exon 1 of the PITX3 locus (Figure 2), with an efficiency of 19%. Immunofluorescent images shown in Figure 3 demonstrate EGFP expression driven by the PITX3 promoter following differentiation using a published PA6 stromal cell differentiation protocol12,13. Co-localization of EGFP could be observed with markers of midbrain dopaminergic neurons such as FOXA2 (red) (Figure 3A) and tyrosine hydroxylase (TH; red) (Figure 3B), while importantly no co-localization could be observed with the GABAergic marker, gamma-aminobutyric acid (GABA; red)(Figure 3C).
Figure 1. Preliminary genomic PCR screening. Following colony picking and expansion, preliminary genomic PCR screening detected numerous positive clones as shown by the PCR product (band at 984 bp). (Clones were screened using primers hPITX3 L arm gen. F, which lies just outside the left targeting arm, and hPITX3 L arm GFP R which is located within the EGFP gene.) Right-hand and left-hand side lanes are a 1 kb ladder. Please click here to view a larger version of this figure.
Figure 2. Validation of EGFP targeting to the hPITX3 locus using zinc finger nucleases in hESCs. (A) Schematic of the Southern blot strategy employed to validate targeting. The top panel schematic illustrates the native hPITX3 locus before targeting. The bottom panel schematic depicts the hPITX3 locus following correct insertion of the EGFP transgene. The red lines under each panel illustrate the sites at which the 5’ and 3’ Southern blot probes will bind, and correspond to the bands in Figure 2B. (B) Representative Southern blot of correctly targeted colonies selected from PCR pre-screening. Abbreviations: EGFP, enhanced green fluorescent protein; PGK, human phosphoglycerol kinase promoter; PURO, puromycin resistance gene. Other features: loxP sites, purple; polyadenylation sequence, blue, hPITX3 exons, orange. Please click here to view a larger version of this figure.
Figure 3. hPITX3-EGFP hESCs differentiated under PA6/LSB/SAG/FGF8 conditions. Immunofluorescent images of EGFP labeled with (A) FOXA2 (red), (B) TH (red) and (C) GABA (red). Counterstained with DAPI (blue). Scales bars, 50 μm. Please click here to view a larger version of this figure.
Substance | ml/100 ml | Catalogue No. (Supplier) |
DMEM | 89 | 10566-016 (Life Technologies) |
Fetal Bovine Serum | 10 | 16140-071 (Life Technologies) |
Pen/Strep | 1 | 15070-063 (Life Technologies) |
Table 1.
Component | 100 mm dish | 60 mm dish | 6-well | 48-well | 96-well | 75cm2 flask |
HBSS | 5 ml | 2 ml | 1 ml | 0.5 ml | 0.1 ml | 8 ml |
Accutase | 5 ml | 2 ml | 1 ml | 0.5 ml | 0.1 ml | 8 ml |
Dispase | 5 ml | 2 ml | 1 ml | – | – | – |
hESC media | 10 ml | 6 ml | 3 ml | 0.5 ml | 0.2 ml | 15 ml |
Matrigel | – | – | – | – | 0.1ml | – |
Table 2.
Substance | ml/100 ml | Catalogue No. (Supplier) |
DMEM/F12 | 80 | 11320-033 (Life Technologies) |
KSR | 20 | 10828-028 (Life Technologies) |
NEAA | 1 | 11140-050 (Life Technologies) |
GlutaMAX | 1 | 35050-061 (Life Technologies) |
Pen/Strep | 1 | 15070-063 (Life Technologies) |
ß-mercaptoethanol | 0.1 | 21985-023 (Sigma-Aldrich) |
rhFGF2 | 7 ng/ml [final] | 233-FB-025/CF (R&D Systems) |
Table 3.
Substance | Concentration | Catalogue No. (Supplier) |
N-Lauroylsarcosine sodium salt solution (Sarcosyl) | 0.5% (w/v) | 61747 (Sigma-Aldrich) |
EDTA | 10.0 mM | E9884 (Sigma-Aldrich) |
NaCl | 10.0 mM | S3014 (Sigma-Aldrich) |
Tris-Cl (pH7.5) | 10.0 mM | T3253 (Sigma-Aldrich) |
Proteinase K | 1.0 mg/ml | BIO-37084 (Bioline Australia) |
Table 4.
Substance | Concentration | Catalogue No. (Supplier) |
Tris-Cl (pH8.0) | 10.0 mM | T3253 (Sigma-Aldrich) |
EDTA (pH8.0) | 0.1 mM | E9884 (Sigma-Aldrich) |
Table 5.
The generation of reporter cell lines offers a powerful means to track, visualize and isolate cells of interest from a heterogeneous population derived from hESCs. However, gene targeting via conventional homologous recombination has proven to be extremely inefficient for human ESCs3. In this protocol we describe a relatively simple technique for introducing a RP into exon one of the PITX3 locus, in hESCs using a widely available targeting vector together with ZFNs. In our hands the gene targeting system is efficient, with 19% of puromycin-resistant clones containing a correctly targeted PITX3 locus (similar to that previously published6).
A limitation with using ZFNs to facilitate gene targeting is the difficulty in designing ZFNs from scratch in house. Thus, in this protocol we used custom designed commercially available ZFNs, which can be expensive to acquire. More recently, the TALEN system has been used for gene targeting using homologous recombination9. The use of TALENs greatly reduces the cost of gene editing and can be designed in-house, expediting the process. Although not explicitly demonstrated in this study, this protocol can be easily modified to incorporate the use of TALENs as the methodology is the same, except TALEN plasmids are used in Step 3.5 instead of ZFN mRNA, as shown in 9.
When using this technique it is imperative to ensure that hPSCs are in log phase growth, (e.g., cultures are not 70% or more confluent). We believe that this is critical as the plasmid gains access to the nucleus following nuclear envelope breakdown during cell division. Moreover, the DNA donor vector will only incorporate into DNA following DNA synthesis as this process exploits homology-directed repair during division to incorporate the donor DNA into the genome.
Another potential limitation with the ZFN approach is the introduction of DNA breaks elsewhere in the genome. Although Southern hybridization with external 5’ and 3’ probes detects correct integration it will not detect additional integration events or error-prone repair elsewhere in the genome. Extra integration events can be examined by Southern hybridization using vector-specific probes (e.g., EGFP) whereas off-target cleavage of the ZFN pairs can be detected using SELEX3,6.
Although we did not observe any obvious effects in the development of midbrain dopaminergic neurons that could be directly attributed to the inactivation of one PITX3 gene, researchers should bear in mind that any targeting event where one copy of a gene is disrupted may be detrimental to the development of that cell and should take this into account when analyzing results. Whether this is the case will depend largely on the gene being targeted and may only be ascertained by performing the experiment.
Using a PA6 based differentiation protocol12,13 we were able to confirm the fidelity of the reporter system by immunofluorescence (Figure 3) and qPCR analysis of EGFP positive vs. negative populations following fluorescence-activated cell sorting (data not shown). As PITX3 is a transcription factor specific to midbrain dopaminergic neurons14, the engineering of a human PITX3 reporter cell line will facilitate the use of PITX3+ midbrain dopaminergic neurons derived from hESCs in studies of either in vitro PD modeling or cell replacement therapy.
The authors have nothing to disclose.
Work described in this protocol was made possible by funding from the Victorian Government as part of an collaborative alliance with the Californian Institute for Regenerative Medicine (CIRM).
Mouse Embryonic Fibroblasts (DR4) | GlobalStem | GSC-6004G | |
Gelatin | Sigma | G1393-100ML | |
HBSS, no Calcium, no Magnesium, no Phenol Red (HBSS) | Life Technologies | 14175-095 | |
Dispase | StemCell Technologies | 7923 | |
Cell Scraper | Corning | 3008 | |
Accutase | Life Technologies | A11105-01 | |
Y27632 | Axon Medchem | Axon 1683 | |
Cell Strainer – 40um | BD Biosciences | 352340 | |
33mm – 0.22 um filter | Millipore | SLGP033RS | |
rhFGF2 | R&D Systems | 233-FB-025/CF | |
Electroporator | BioRad | 165-2661 | |
0.4cm electroporation cuvette | BioRad | 165-2088 | |
Human TV-hPITX3-forward targeting construct | Addgene | 31942 | |
ZFN Kit; Human PITX3 | Sigma-Aldrich | CKOZFN1050-1KT | |
Puromycin | Invivogen | ant-pr-1 | |
Matrigel | BD Biosciences | 354277 | |
Geltrex | Life Technologies | A1569601 | |
NaCl | Sigma-Aldrich | S3014 | |
Dulbecco's Modified Eagle Medium (DMEM) | Life Technologies | 10566-016 | |
Fetal Bovine Serum, Qualified, Heat Inactivated, US Origin (FBS) | Life Technologies | 16140-071 | |
Pen/Strep | Life Technologies | 15070-063 | |
Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12) | Life Technologies | 11320-033 | |
KnockOut Serum Replacement (KSR) | Life Technologies | 10828-028 | |
MEM Non-Essential Amino Acids (NEAA) | Life Technologies | 11140-050 | |
GlutaMAX | Life Technologies | 35050-061 | |
ß-mercaptoethanol | Life Technologies | 21985-023 | |
N-Lauroylsarcosine sodium salt solution (Sarcosyl) (30%) | Sigma-Aldrich | 61747 | |
EDTA | Sigma-Aldrich | E9884 | |
Trizma Hydrochloride | Sigma-Aldrich | T3253 | |
Proteinase K solution (20mg/ml) | Bioline Australia | BIO-37084 | |
Expand Long Template PCR System | Roche Australia | 11681834001 | |
hPITX3 L arm gen. F (primer): 5’ TGTCCTAAGGAGAATGGGTAACAGACA 3’ | GeneWorks, Australia | ||
hPITX3 L arm GFP R (primer): 5’ ACACGCTGAACTTGTGGCCGTTTA 3’ | GeneWorks, Australia | ||
HyperLadder 1kb | Bioline Australia | BIO-33025 | |
GelRed | Jomar Bioscience, Australia | 41003 |