|Home | About | Journals | Submit | Contact Us | Français|
This study demonstrates a novel method by which multiple separate plasmids can be stably integrated into the genome using single antibiotic resistance for selection. This method was used to integrate three different cardiac-specific promoters driving different fluorophores into murine embryonic stem cells, allowing sequential visualization of various stages of cardiac differentiation. This method is broadly applicable to the study of cell lineage of different stem/progenitor cells.
There are many circumstances where it is desirable to insert multiple plasmids into the same cell and optimally, to be able to use different promoters to control expression of the different gene products. For example, embryonic stem (ES) with multiple stage-specific promoters driving different fluorophore markers would be an important tool for studying the molecular pathways mediating lineage commitment and differentiation. Such cells could also be used to isolate pure populations of cells at different defined stages of lineage commitment to compare their in vivo characteristics and use in cell-based therapies. Lineage analysis has classically been done with the insertion of a specific promoter driving a fluorophore (promoter-fluorophore element) or an antibiotic-resistance gene. This method requires the use of flow cytometry or antibiotic selection after transfection of the plasmid carrying the promoter-fluorophore combination. Engineering multiple tissue promoters on the same plasmid could possibly lead to mutual interference. For example, the proximal elements of the cardiac actin gene can be acted upon by distinct enhancer elements of the skeletal and cardiac myoblast program.1 Alternatively, multiple constructs with single promoter-fluorophore elements and different antibiotic-resistance genes in each construct can be introduced, but this requires the use of multiple antibiotics, which tend to differentiate ES cells. The efficiency of this approach is also too low to be practical routinely. Another approach is serial transfection with different promoter-fluorophore elements with one antibiotic selection at a time, but this method increases the passage number of the cells. ES cells at high passage number are known to accumulate genetic mutations and to differentiate poorly.2 This study describes a novel method by which a single antibiotic selection can be used to isolate cells in which multiple plasmids are stably integrated into the genome.
Protein heterodimerization was used in this method such that antibiotic resistance is dependent on the presence of three different plasmids. This method uses the dimerization properties of the modified ecdysone receptor (EcR) and the retinoic acid receptor (RXR). This complex binds to a glucocorticoid response element (GRE) and drives transcription in the presence of ponasterone.3,4 This system was modified by separating the expression of EcR, RXR, and the response element driving neomycin resistance into three different constructs, and their interdependency was tested with neomycin selection. The schematic of the system is shown in Figure 1a. Three fluorophore markers (RFP, CFP, and YFP) were cloned into these separate constructs, as shown in Figure 1b. The plasmid containing neomycin and the YFP fluorophore were linearized before electroporation by cutting between the YFP and GRE sequences (red bar in Fig. 1b) to prevent interactions between CMV-YFP and GRE-BP neomycin. The long-term expression of all three fluorophores was verified by electroporation of these constructs into murine ES cells and mesenchymal stem cells (MSCs). Resistant ES cell colonies that formed after selection were always positive for all three fluorophores, as shown in Figure 1c–h. These colonies were passaged to rule out episomal expression and for silencing of the promoters. All colonies continued to be positive for the three fluorophores after more than 10 passages. Plasmids were also electroporated individually and in various combinations to investigate for false-positive colonies as a result of leakage of the response element (Fig. 1i), and virtually no colonies were detected in the absence of all three constructs. MSCs had fewer resistant colonies as compared with ES cells (Fig. 2a–g), reflecting the lower transfection efficiency by electroporation5 of MSCs, but again, all three constructs were required for colony formation. A similar approach demonstrated previously with zinc finger proteins does not ensure antibiotic selection6 and was tested for transient transfection only.
The use of this system was tested further by cloning three promoters specific to the cardiac lineage into three constructs that use the protein dimerization for antibiotic resistance. The schematic of these plasmids is shown in Figure 3a. Nkx2.5 is a transcription factor that marks the earliest known cardiac progenitors in heart development.7 A 2.5-kb cardiac-specific enhancer of Nkx2.5 with a basal promoter8 driving CFP was cloned into the first construct that constitutively expresses EcR, driven by a CMV promoter. Mef2c is a transcription factor that is expressed at a later stage in the anterior heart field.9 The 4-kb Mef2c enhancer, with its endogenous promoter driving RFP, was cloned into the second construct, constitutively expressing RXR driven by a CMV promoter. α-MHC is a structural protein that is expressed in progenitor cardiomyocytes.10 The 5-kb α-MHC promoter driving GFP was cloned into the third construct, which also has the response element driving neomycin resistance. The functionality and tissue specificity of these constructs were verified in mouse neonatal cardiomyocytes and 293 cells as shown in Figure 4. All of the constructs were expressed abundantly by transfected cardiomyocytes (Fig. 4a–d) but were not active in 293 cells (Fig. 4e–h). ES cells were then electroporated with these constructs to create reporter lines that would glow with different fluorophores as the cells reached different stages of cardiac differentiation. All resistant clones were screened for the complete insertion of the constructs by PCR, and approximately 35% of all of the colonies had complete integration of all three constructs. Clones that were verified for genomic integration of these constructs were investigated for concatenated insertion by PCR, and none of the clones was found to have concatenated insertions.
Mouse ES cells were cultured on gelatin-coated dishes in a medium with DMEM, 20% FBS, leukemia inhibitory factor (LIF), 2-ME, nonessential amino acids, L-glutamate, and antibiotics. MSCs were isolated from mouse femur and grown in tissue-culture dishes in isolation medium consisting of RPMI, 9% FBS, 9% equine serum, L-glutamate, and antibiotics. Adherent cells were cultured continuously for 2 weeks in isolation medium and monitored for cell proliferation. The medium was changed to expansion medium consisting of IMDM, 9% FBS, 9% equine serum, L-glutamate, and antibiotics. Differentiation of ES cells was achieved by withdrawal of LIF.
Plasmid DNA was linearized and purified in sterile conditions. The plasmid (25 μg) was electroporated using a BioRad GenePulser II machine (BioRad, Hercules, CA, USA). The voltage of the pulse was set at 250 volts and the capacitance at 330 μF. Cells were incubated in ice for 20 min, immediately after administering the pulse. Cells were transferred to gelatin-coated tissue-culture dishes with ES medium, and selection was started after 48 h. Ponasterone was added in the first 12 h to enable the production of neomycin phosphotransferase and subsequently, included in all media preparations involving the three element system.
RXR and VgEcR were cloned from pERV3 (Stratagene, La Jolla, CA, USA). The GRE was cloned from pEGSH. RFP was cloned from monomeric RFP vector. CFP and YFP were cloned from pECFP and pEYFP, respectively. Neomycin was cloned from pcDNA 3.1. The cardiac-specific enhancer of Nkx2.5 with a Hsp68 basal promoter was cloned into the first plasmid expressing VgEcR. Mef2c promoter was cloned into the second plasmid expressing RXR. The α-MHC promoter was cloned into the third plasmid with the response element driving neomycin resistance. The MLC2v promoter was cloned from genomic DNA and was ligated adjacent to the α-MHC promoter.
Total cellular RNA was extracted using Trizol (Invitrogen, Carlsbad, CA, USA) reagent, according to the protocol provided by the company. Trace amounts of genomic DNA were removed by using the Turbo-free DNAase kit (Ambion, Austin, TX, USA). cDNA was made using oligo dT, and RT reaction was done with the Thermoscript kit (Invitrogen). GAPDH was used as an internal control, and PCR products were run on an agarose gel.
All images were acquired on a Zeus Axiovert fluorescence microscope.
All samples were suspended as single cells and sorted at the FACS facility on a MoFlo machine (Dako Cytomation, Carpinteria, CA, USA). Two hundred ninety-three cells, transiently transfected with constructs expressing single fluorophores, were used as controls to set the gates for fluorescence intensity.
Embryoid bodies were plated on fibronectin-coated coverslips and fixed with 4% paraformaldehyde. Free-floating embryoid bodies were also centrifuged and fixed. Coverslips were washed in PBS twice and incubated with PBS-Tween for 5 min. Goat serum was added to the solution to a total concentration of 5%. Primary antibodies for cardiac troponin T (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and α-MHC (MF20, Developmental Studies Hybridoma Bank, Iowa City, IA, USA) were added to the solution at a 1:500 dilution and incubated overnight. The next day, after three washes with PBS, secondary antibody conjugated to Alexafluor 647 was added for 1 h. The coverslips were then washed thrice with PBS, and DAPI was added for 10 min. After a set of three final washes with PBS, the coverslips were mounted with Fluormount G.
The sequence of fluorophore activation during differentiation of the cells in embryoid bodies is shown in Figure 3b–p. CFP expression driven by the Nkx2.5 became apparent at Day 3 and persisted through Day 6. RFP driven by the Mef2c promoter was expressed between Days 5 and 6 and colocalized as expected at this time with CFP/Nkx2.5. As the cells continued to differentiate, CFP and RFP expression became almost undetectable by Day 9 when GFP, driven by the α-MHC promoter, was expressed. Spontaneously contracting clusters seen upon differentiation had distinctly glowing fluorescent cell clusters in them (Supplemental Video 1). The cardiac nature of these cells was also verified by counterstaining on Day 9 with cardiac troponin T antibody (Fig. 3q and u). The fidelity of expression and potential use for enriching for specific subpopulations of lineage species were examined by comparing mRNA levels of the three genes in cells FACS sorted for the fluorophores compared with unsorted cells. (Fig. 4i). We engineered a fourth promoter, MLC2v driving YFP, into the three-vector system (Fig. 5a), which would mark ventricular cardiomyocytes. Quantification of the relative proportion of progenitor and differentiated cells at two different points of differentiation was possible by flow cytometry (Fig. 5b). Thus, this system is not only an ideal tool for studying the effects of various small molecules and differentiating agents on the differentiation patterns of ES cells or other cell populations but also has the added advantage of allowing FACS sorting of enriched, live subpopulations for further studies in vivo.
Green fluorescent cells inside a spontaneously contracting EB on day 9 of differentiation indicate that the plasmid is expressed by genuine cardiomyocytes.
One of the challenges in the cell-replacement therapy is to find the most suitable state of differentiation where cells are committed to a particular lineage but still retain properties of progenitor cells that enhance their use for transplantation. This system enables the isolation of subpopulations of cells at different stages of commitment and hence, will provide a basis for the measurement of their stage-specific properties. The schematic shown in Figure 6a allows for the overexpression of at least three genes, such that PCR verification of the insertion of the plasmids into the genome would not be required, and furthermore, the overexpression can also be fine-tuned to be stage-specific as shown in Figure 6b. Overlap of fluorophore spectra is unlikely to be a concern, as it is now possible to accurately distinguish as many as 12 spectrally distinct optical signatures during FACS sorting of ES cells.11 Multiple imaging modalities besides optical imaging, such as bioluminescence, positron emission tomography, and magnetic resonance imaging, have been shown to be useful in isolation of homogenous ES cell populations and tracking transplanted ES cells in vitro.12–14 The method described in this paper can draw on the strengths of these multiple imaging modalities and can be suitably modified to encompass them. Thus, this study demonstrates a novel method that may be useful for inserting multiple stage-specific promoter-fluorophore elements or promoter-gene elements into the genome of ES cells or other cell types. The method may be particularly useful for studying the lineage commitment of stem/progenitor cells in varied settings and to purify pure populations of stage-specific cells for cell-replacement therapy.
The Nkx2.5 cardiac enhancer was a kind gift of Dr. Eric N. Olson (Southwestern Medical Center, Dallas, TX, USA). The Mef2c promoter was shared by Dr. Brian L. Black (University of California, San Francisco, CA, USA). The α-MHC promoter was shared by Dr. James Gulick (University of Cincinnati, Cincinnati, OH, USA).
Conflict of interest statement: Both authors agree to the content and data presented in this manuscript. The authors have no competing interests. No human or animal subjects were used in this study.