|Home | About | Journals | Submit | Contact Us | Français|
Human induced pluripotent stem cells (hiPSCs) derived from patient samples have tremendous potential for innovative approaches to disease pathology investigation and regenerative medicine therapies. However, most hiPSC derivation techniques utilize integrating viruses, which may leave residual transgene sequences as part of the host genome, thereby unpredictably altering cell phenotype in downstream applications. Here we describe a protocol for hiPSC derivation by transfection of a simple, nonviral minicircle DNA construct into human adipose stromal cells (hASCs). Minicircle DNA vectors are free of bacterial DNA and thereby capable of high expression in mammalian cells. Their repeated transfection into hASCs, an abundant somatic cell source that is amenable to efficient reprogramming, results in transgene-free hiPSCs. This protocol requires only readily available molecular biology reagents and expertise, and produces hiPSC colonies from an adipose tissue sample in ~4 weeks.
Human induced pluripotent stem cells (hiPSCs) can be derived from somatic cells through a reprogramming process driven by overexpression of a defined set of transcription factors [1, 2]. These hiPSCs share the properties of self-renewal and pluripotency with human embryonic stem cells (hESCs), and can therefore be used to generate unlimited quantities of differentiated cell types of all three germ layers, including cardiac cells, neural cells, and hepatic cells. hiPSCs can be generated from patients of virtually any genetic background, including those with disease-conferring genetic mutations [3–5]. In contrast, the derivation of hESCs from different genetic backgrounds is challenging because human embryo use is limited and ethically debated. The production of patient-specific and disease-specific hiPSCs enables a variety of downstream applications, including drug screening, disease modeling, pathogenesis studies, and regenerative medicine therapies.
However, traditional approaches to deriving hiPSCs require use of retroviruses or lentiviruses that integrate reprogramming genes into the host genome. Random integration of the reprogramming genes may result in insertional mutagenesis that causes malignant transformation of a clonal cell population . In addition, some of the genes used in the reprogramming process are known proto-oncogenes, and incomplete silencing of these transgenes may result in unknown adverse effects. These challenges have been partially circumvented in mice by the development of non-integrating viral , non-viral episomal , and excisional [9, 10] techniques for reprogramming. Yet clinical translation of these safer iPSC derivation techniques is challenging because human cells are relatively more resistant to non-viral transfection and are not immediately available in large quantities. Although hiPSCs may be generated by lentiviral transduction with subsequent Cre-loxP excision of reprogramming factors , residual vector sequences will be left behind in the genome. Transgene-free hiPSCs have been derived from neonatal foreskin fibroblasts using a combination of three episomal plasmids expressing seven reprogramming factors . Alternatively, transgene-free hiPSCs can be derived from fetal or neonatal cells by repeated transduction of proteins in the presence of chemical treatments (e.g., valproic acid) . However, none of the aforementioned techniques for transgene-free hiPSC derivation have been demonstrated using adult donors, a more clinically relevant population. Here, we describe in detail a protocol for the derivation of transgene-free hiPSCs with a non-viral minicircle DNA reprogramming construct used in conjunction with human adipose stromal cells (hASCs) . This technique is advantageous in translational studies because somatic cells from human adults can be reprogrammed in the absence of genomic modification, viral sequences, or proto-oncogenes (such as c-Myc), effectively mitigating safety concerns . This protocol can be used to derive hiPSCs from human samples in ~4 weeks using standard molecular biology reagents and cell culture expertise (Figure 1).
Cells are not transduced with infectious viral particles in this protocol, ensuring a high likelihood of generating transgene-free hiPSCs. However, reprogramming efficiency using this protocol is substantially lower (~0.005%) compared to lentiviral techniques for overexpression of the transcription factors OCT4, SOX2, NANOG, and LIN-28. Further improvement of reprogramming efficiency may be achieved by treatment with small molecules (e.g. valproic acid)  or cell signaling peptides (Wnt) . Also, users should be aware that the protocol as described here has not yet been successfully applied to the reprogramming of human dermal fibroblasts derived from adult sources. We have found that minicircle-based reprogramming of hASCs as described here is advantageous for deriving transgene-free hiPSCs from adult human donors, a clinically relevant cell source. Such methods offer the ability to develop patient-specific or disease-specific cell lines for exciting new translational and disease modeling studies.
hASCs are an attractive source for hiPSC derivation because they are available in large numbers and are amenable to reprogramming [17, 18]. While skin biopsies are necessarily minimal in size, adipose tissue can be harvested in very large quantities during lipoaspiration procedures, allowing quick expansion to a large starting population of hASCs. For example, reprogramming can start as early as 24 hours after the liposuction procedure with a large starting cell population of up to 10 million cells, whereas 3–4 weeks are required following punch skin biopsy for expansion of the dermal fibroblasts. hASCs are a heterogeneous group of multipotent progenitor cells that can differentiate into adipogenic, osteogenic, chondrogenic, and myogenic cell lineages [19, 20]. As such, hASCs express relatively high levels of genes such as c-Myc and Klf4, which may underlie the relative ease of their induction to pluripotency. Users should note that the protocol described here has been most successfully applied when using hASCs as the starting cell source, although neonatal fibroblasts (IMR90) can also be used but with lower efficiency.
Minicircle vectors are supercoiled DNA molecules free of bacterial plasmid backbone elements, such as an origin of replication and antibiotic resistance gene. They primarily consist of a eukaryotic expression cassette, and therefore do not activate exogenous silencing mechanisms to the same extent as plasmids. Therefore, minicircle vectors benefit from higher transfection efficiences and more stable ectopic transgene expression than plasmid DNA (Figure 2) . The pMC.LGNSO plasmid is the parental DNA construct that is used to produce minicircle reprogramming vector in the initial steps of this protocol (Steps 1–13). Parental plasmid DNA (e.g. pMC.LGNSO) that produces minicircle vectors may be conceptually divided into two parts separated by attB and attP recognition sequences. Intermolecular recombination between the attB and attP sequences catalyzed by the ΦC31 integrase (Step 9) yields a minicircle vector separate from the remainder of the plasmid (Figure 3). The minicircle vector contains the reprogramming genes OCT4, SOX2, NANOG and LIN-28. This expression cassette is isolated and purified from the bacterial suspension (Steps 1–12) before being transfected into somatic cells to induce reprogramming (Step 13–47). The other part of the pMC.LGNSO parental plasmid contains bacterial plasmid backbone elements (e.g., the origin of replication and antibiotic resistance cassette), as well as the ΦC31 integrase and I-SceI restriction enzyme expression cassettes under the control of an L-arabinose inducible promoter. This part of the plasmid is linearized by I-SceI endonucleolytic cleavage and subsequently degraded (Step 9), allowing isolation of pure minicircle DNA by common plasmid purification procedures (e.g., Qiagen Plasmid Purification Kits). Yield and purity of the minicircle DNA preparation may be optimized by use of plasmids encoding multiple copies of ΦC31 integrase , or using E. coli strains that encode L-arabinose-inducible I-SceI as part of the bacterial genome (e.g., E. coli strain ZYCY10P3S2T). The purified minicircle expression cassette contains the four reprogramming genes plus GFP separated by 2A self-cleaving peptide sequences, thereby allowing for the equimolar expression of all five proteins from a single RNA transcript (Fig. 3). This protocol requires transfection of hASCs with the minicircle preparation three times; first via electroporation to optimize efficiency, followed by flow cytometry sorting to enrich for successfully transfected cells, followed by two rounds of transfection mediated by cationic lipids (e.g., using Lipofectamine) to optimize cell survival.
Detailed description of the flow cytometric enrichment of GFP+ hASCs after the first electroporation (Step 37) will vary by institution and is beyond the scope of this protocol. Fortunately, well-established protocols for sorting of GFP+ cells are readily available [23, 24]. For the initial sorting procedure, 5 × 105 untransfected hASCs should be trypsinized and resuspended in FACS buffer alongside the transfected hASCs in order to set up the gating parameters. Once set, the gating parameters can be saved and untransfected hASCs are no longer needed. We used a FACSAria equipped with FACSDiva software for cell analysis and sorting (http://facs.stanford.edu). Forward scatter (~82V) and side scatter (~176 V) were adjusted to exclude cell aggregates and debris. The band pass filter for FITC detection was used (~537 V), and cells were analyzed at a flow rate between 200 and 1,000 events per second in a stream formed from a nozzle with a 70 micron orifice.
hiPSC-like colonies with a tightly-packed, dome-like structure first appear at approximately day 18 (Steps 53–54; Figure 4). Colonies become large enough to be manually picked up and transferred to a separate irradiated mouse embryonic fibroblast (iMEF) feeder layer at approximately day 28 (Steps 55–56). At early stages, colonies remain GFP+, although with continued culture and passaging, bona fide hiPSC clones will become GFP-.
Expression of the minicircle-derived reprogramming cassette can be easily monitored throughout the procedure by GFP fluorescence. At the conclusion of the procedure, however, bona fide hiPSC colonies will no longer be dependent on exogenous reprogramming factor expression, and will thereby be GFP negative. The pluripotency of the derived hiPSC lines can be verified using standard protocols, including RT-PCR and immunocytochemistry for pluripotent markers, embryoid body differentiation, and teratoma formation, that are described in further detail elsewhere . In addition to the standard protocols above, the resulting hiPSCs can be screened for rare integration events by a simple PCR-based assay for the presence of the transgene in genomic DNA. The minicircle-derived reprogramming cassette contains the LIN-28, NANOG, SOX2, and OCT4 cDNA sequences in succession (Figure 3), while endogenous genomic DNA loci expressing these genes contain introns and are present on disparate chromosomes. Therefore, PCR of extracted genomic DNA using primers annealing to two contiguous minicircle reprogramming genes (e.g. SOX2 and OCT4) can readily identify rare integration events of the transgene into genomic DNA (Steps 57–60).
First prepare a stock of 20% L-arabinose by dissolving 2 g L- arabinose in 100 ml of ddH2O. Freeze stock at -20°C for up to 6 months. To prepare induction broth, mix 384 ml LB, 16 ml 1N NaOH, and 0.4 ml 20% L-arabinose (final concentration of L- arabinose is 0.01%). Make fresh just prior to use. hASC medium: DMEM High Glucose with L-Glutamine containing 10% (vol/vol) FBS, 0.5% (vol/vol) penicillin/streptomycin (final concentration 50 units/mL penicillin and 50 ug/mL streptomycin). Store at 4 °C for ~1 month.
Knockout DMEM, 20% (vol/vol) ES Cell FBS, 0.1 mM nonessential amino acids, 0.1 mM 2-mercaptoethanol, and bFGF to final concentration of 8 ng/mL. Store at 4 °C for ~2 weeks.
Add 75 mg of Type II Collagenase to 100 ml of ddH2O (0.075% wt/vol) and filter sterilize using a 0.44μm vacuum-driven filter system. Make 10 ml aliquots of solution and store at −20 °C for up to 2 months.
Dissolve 0.5 g gelatin powder in 500 ml of distilled water, autoclave and store at 4 °C for up to 2 months.
Add a sufficient volume of gelatin solution to cover the surface of the plate or well. Incubate the dish for at least 30 minutes at 37 °C. Prior to use, aspirate excess gelatin solution.
|pMC.LGNSO plasmid||1 μg|
|NEB Buffer||2 μl|
|10 × BSA||2 μl|
|restriction endonuclease (HindIII/SphI/NdeI/XhoI)||0.5 μl|
|ddH2O||up to 20 μl|
|Restriction endonuclease||Fragment size (bp)|
|37||Low yield of GFP+ cells||Low transfection efficiency||Use high quality DNA preparations. Consider using an endotoxin removal kit to further purify the minicircle DNA preparation of contaminants that may limit efficient transfection.|
|Begin with higher numbers of starting cell source, up to 1 × 107 cells can be used during the initial transfection.|
|52||Cells do not attach||Trypsin cells damaged||Do not incubate hASCs in TrypLE Express for longer than 5 minutes. Cells can be gently scraped from the surface of the plate to aid detachment.|
|Cells experience excessive shear force||Pipette gently|
|56||Unstable hiPSC clones||Plasmid integration into genome||Use high quality DNA preparations. Store DNA at −20 °C and avoid repeated freeze- thaw cycles as this may cause strand nicking and degradation.|
|Partial reprogramming||Choose high quality clones. Colonies with good hESC morphology are tightly compacted with sharp borders and contain cells containing prominent, large, round nucleoli and scant cytoplasm. Use freshly prepared medium and ensure that the appropriate amount of bFGF has been added.|
Transfection efficiency can be monitored by GFP fluorescence under an inverted epifluorescence microscope or by flow cytometry. During the plasmid transfections, floating cells will appear because of the cytotoxicity of electroporation and Lipofectamine®. The initial nucleofection procedure has a transfection efficiency of 2-10%, and GFP fluorescence of the sorted population declines over time as shown in Figure 4C. Picked hiPSC clones will stain positively for pluripotency markers (Tra-1-60, SSEA-4), express high transcript levels of pluripotency genes (Oct4, Sox2, Nanog, Crypto), and display hESC-like morphology, such as large, round nucleoli and scant cytoplasm (Figure 4). We have successfully derived 22 hiPSC lines using this protocol from three adult donors. We calculate our reprogramming efficiency using this technique as ~0.005%, although any such calculation may be obviated by the splitting of cells during the procedure.
We are grateful to Ning Sun for expert assistance with cell culture techniques. We thank Zhi Ying Cheng for help with minicircle production techniques. We thank funding support from Howard Hughes Medical Institute (K.H.N.); Mallinckrodt Foundation, NIH DP2OD004437, RC1HL100490, Burroughs Wellcome Foundation, and American Heart Association 0970394N (J.C.W.); NIH R90 DK 07010301, NIH R21 DE018727, NIH R21 DE019274, NIH RC2DE020771, the Oak Foundation and the Hagey Laboratory for Pediatric Regenerative Medicine (M.T.L.); NIH RC1HL100490-02 (J.C.W. & M.T.L.); U01HL099776 (R.C.R.).
AUTHOR CONTRIBUTIONSK.N., F.J., and J.C.W. prepared most of the paper. R.C.R., M.A.K., and M.T.L. provided advice and proofread the paper.
The authors declare that they have no competing financial interests.