Using the low-cost cationic polymer polyethylenimine (PEI), we produced lentiviral particles for stable expression of shRNAs in H9 human embryonic stem cells (hESCs) and transiently transduced H9-derived neural progenitor cells (NPCs) at high efficiency.
The current protocol describes the use of lentiviral particles for the delivery of short hairpin RNAs (shRNAs) to both human embryonic stem cells (hESCs) as well as neural progenitor cells (NPCs) derived from hESCs at high efficiency. Lentiviral particles were generated by co-transfecting HEK293T cells using entry vectors (carrying shRNAs) along with packaging plasmids (pAX and pMD2.G) using the low-cost cationic polymer polyethylenimine (PEI). Viral particles were concentrated using ultracentrifugation, which resulted in average titers above 5 x 107. Both hESCs and NPCs could be infected at high efficiencies using these lentiviral particles, as shown by puromycin selection and stable expression in hESCs, as well as transient GFP expression in NPCs. Furthermore, western blot analysis showed a significant reduction in the expression of genes targeted by shRNAs. In addition, the cells retained their pluripotency as well as differentiation potential, as evidenced by their subsequent differentiation into different lineages of CNS. The current protocol deals with the delivery of shRNAs; however, the same approach could be used for the ectopic expression of cDNAs for overexpression studies.
Human embryonic stem cells (hESCs) derived from the blastocyst inner cell mass are pluripotent and can be differentiated into different cell types depending upon external factors under in vitro conditions1,2. In order to fully harness the potential of hESCs, it is imperative to have rapid and reliable gene delivery methods for these cells. Conventionally, the techniques used can be broadly classified into two types: nonviral and viral gene delivery systems3,4. The more frequently used nonviral gene delivery systems are lipofection, electroporation, and nucleofection. Nonviral delivery systems are advantageous because of fewer insertion mutations and an overall decrease in immunogenicity5,6. However, these methods result in low transfection efficiency and a short duration of transient gene expression, which is a major limitation for long-term differentiation studies7. Electroporation results in better transfection efficiencies compared to lipofection; however, it results in more than 50% cell death8,9,10. Using nucleofection, the cell survival and transfection efficiency can be improved by combining lipofection and electroporation, but the approach needs cell-specific buffers and specialized equipment and, thus, becomes quite costly for scaled-up applications11,12.
In contrast, viral vectors have shown improved transfection efficiencies, as well as overall low cytotoxicity, following transduction. In addition, the genes delivered are stably expressed and, hence, make this method ideal for long-term studies13. Among the most commonly used viral vectors for gene delivery into hESCs are lentiviral vectors (LVS), which can give more than 80% transduction efficiency using high titer viral particles14,15. Lipofection and CaPO4 precipitation are amongst the most commonly used methods to transiently transfect HEK293T cells or its derivatives with gene transfer vectors along with packaging plasmids to yield lentiviral particles16. Although lipofection results in good transfection efficiency and low cytotoxicity, the technique is hampered by its cost, and scaling up to get high titer lentiviral particles would be very costly. CaPO4 precipitation results in relatively similar transfection efficiencies to those obtained using lipofection. Although cost-effective, CaPO4 precipitation results in significant cell death following transfections, which makes it difficult to standardize and to avoid batch-to-batch variations17. In this scenario, developing a method that gives high transfection efficiency, low cytotoxicity, and cost-effectiveness is crucial for the production of high titer lentiviral particles to be used in hESCs.
Polyethylenimine (PEI) is a cationic polymer that can transfect HEK293T cells at high efficiency without much cytotoxicity and has a negligible cost compared to lipofection-based methods18. In this situation, PEI can be used for scaled-up applications of high titer LVS production through the concentration of lentiviral particles from culture supernatants using various techniques. The presented article describes the use of PEI to transfect HEK293T cells and lentiviral vector concentration using ultracentrifugation through a sucrose cushion. Using this method, we regularly obtain titers well above 5 x 107 IU/mL with low batch-to-batch variations. The method is simple, straight-forward, and cost-effective for scaled-up applications for gene delivery to hESCs and hESCs-derived cells.
1. Transfection of HEK293T cells using either PEI or Lipofectamine 3000 reagent
2. LVS collection and ultracentrifugation
3. LVS titer measurement for pLKO.1-based vectors
4. LVS titer measurement for pll3.7-based vectors
5. Infection of hESCs and stable selection
6. Transduction of H9-derived neural progenitor cells (NPCs)
NOTE: NPCs were derived from H9 cells using a dual-Smad inhibition protocol, as described previously19.
Following gene transfer, high viability of hESCs is inevitably required. Despite the efforts and optimization of protocols to reduce cell death following electroporation of hESCs, more than 50% cell death is still observed after electroporation of these cells, along with low transfection efficiency20. Lentiviral mediated gene transfer not only results in high efficiency of gene transfer but also demonstrates high levels of cell viability following transduction. The results below present two approaches: one using GFP as reporter and transient expression in hESCs-derived NPCs and the other one using puromycin for the stable selection of hESCs for long-term differentiation experiments.
In the first set of experiments, lentiviral particles were produced by co-transfecting shRNAs carrying the pll3.7 vector with GFP as reporter along with second-generation lentiviral packaging plasmids psPAX2.0 and pMD2.G. Lentiviral particles were collected following 48 h and 72 h post transfection and concentrated using ultracentrifugation. Figure 1A shows the abundant GFP expression in HEK293T cells after 72 h of transfection. The expression of viral proteins also resulted in syncytia formation of HEK293T cells where single cells are fused together, as shown by arrows in Figure 1A. The titers were determined using FACS analysis. We also compared the lentiviral titers using low-cost PEI and Lipofectamine 3000 reagent and found no significant difference between the two in terms of viral titers (Figure 2A). Next, H9 hESCs-derived NPCs were infected twice with these GFP-containing lentiviral particles at an MOI of 6 and harvested for analysis post 72 h. Figure 1B shows marked expression of GFP in NPCs following infection. As discussed above, for long-term differentiation experiments, it is imperative to have hESCs that stably express shRNAs. In an attempt to create stable cell lines expressing shRNA for various genes, we used plKO.1-based vectors and cloned shRNAs according to the manufacturer's recommendations. Next, lentiviral particles were produced as described above, and titers were determined using a commercial lentiviral titer measurement kit using qPCR-based methods (Figure 2B). hESCs were infected with lentiviral particles at an MOI of 10 and selected using puromycin selection. Figure 3 shows the results of stable selection of hESCs following transduction. After 5 days of selection with 0.8 µg/mL puromycin, lentiviral transfected cells showed more than 80% cell viability as compared to non-transduced cells, where no cells survived the treatment. After the selection, the stably transduced H9 cells were further expanded, cryopreserved, and analyzed using western blot to assay for efficient gene knockdown.
In our laboratory, we have created multiple stable cell lines of H9 hESCs expressing shRNAs for different genes using the above approach. Here, we have presented the results of one of those, which is the L-2-hydroxyglutarate dehydrogenase (L2HGDH) gene using western blot analysis. Abundant expression of L2HGDH is observed in cells transduced with empty vector backbone. However, no expression of L2HGDH is observed in cells transduced with a vector carrying shRNAs for L2HGDH, showing the efficacy of the stable generation of an H9 hESCs cell line with reduced expression of L2HGDH, which can be used for further differentiation studies (Figure 4).
Figure 1: Expression of GFP in HEK293T cells and viral-infected NPCs. (A) shRNAs were cloned into a pll3.7 vector carrying GFP as reporter and co-transfected with packaging plasmids in HEK293T cells using PEI. The high GFP expression shows efficient transfection efficiency post 72 h of transfection. Arrows indicate syncytia formation in the HEK293T cells, where groups of individual HEK293T cells have fused together due to the expression of viral proteins. (B) H9-derived NPCs were infected twice at an MOI of 6 with lentiviral particles containing GFP as reporter, and GFP expression was observed at 72 h post infection. Scale bar = 50 µm. Please click here to view a larger version of this figure.
Figure 2: LVS titer measurements. (A) Viral titers for pll3.7-based vectors were measured using FACS analysis by quantifying the %GFP positive cells for those viral dilutions that gave 2%-20% GFP positive cells. (B) Viral titers for pLKO.1-based vectors were determined by using a commercial qPCR lentivirus titer kit, following the manufacturer's recommendations. Please click here to view a larger version of this figure.
Figure 3: Establishment of stable shRNA-expressing H9 hESC lines. shRNAs were cloned into pLKO.1-based lentiviral vectors carrying the puromycin selection gene. H9 hESCs were infected with lentiviral particles at an MOI of 10 and selected using puromycin. Using 0.8 µg/mL puromycin for 5 days, 100% of the non-infected cells were killed, while most of the cells survived the treatment in the viral-infected group. These cells were expanded without clonal selection for cryopreservation and further analysis. Scale bar = 50 µm. Please click here to view a larger version of this figure.
Figure 4: A representative western blot analysis for stable knockdown of the L2HGDH gene in H9 hESCs. Total cell lysates from negative control (H9, puro uncut) cells, as well as cells stably expressing shRNAs against L2HGDH (H9, sh-L2HGDH) were subjected to western blot using the anti-L2HGDH antibody. As shown in the figure, no expression of L2HGDH was observed in H9 hESCs stably expressing shRNAs for L2HGDH. A non-specific band was observed just above the specific band corresponding to the correct molecular weight of 46-48 kDa (indicated by a triangle) for the L2HGDH antibody. Anti-GAPDH was used as a loading control. Please click here to view a larger version of this figure.
The ability to genetically modify stem cells for study or clinical purposes is limited both by technology and the basic understanding of the biology of hESCs. Techniques that have shown significant potential in mouse ESCs, like lipofection and electroporation, are not highly efficient for hESCs, which are notoriously difficult for gene delivery by conventional methods20. This notion has led to not only the optimization of existing techniques but also the development of novel methods for increased transfection efficiency in hESCs. The focus of these new developments has been efficient and stable gene delivery to hESCs while maintaining minimal effect on hESCs' growth, pluripotency, and differentiation potential over prolonged periods. Among the various new developments that meet these strict criteria is lentiviral mediated gene transfer to hESCs21,22. For the production of lentiviral particles expressing a certain gene, the desired gene is cloned in one of the transfer vectors and is transfected in HEK293T cells along with packaging plasmids to harvest a cell culture supernatant containing lentiviral particles. To be used in hESCs, it is essential to concentrate these viral particles using various techniques to yield high titers of viruses at least in the range of 1 x 107-1 x 108 IU/mL. In general, large volumes of LVS are concentrated from 200x-500x the original volume to achieve these titers. This means using a highly cost-effective method without compromising the transfection efficiency in HEK293T cells is a prerequisite for these methods to be routinely used in the lab. The present method describes the use of the cationic polymer PEI as a cost-effective method to produce large volumes of LVS with a transfection efficiency similar to lipofection-based methods, which are considered gold standard. Using the presented method, we can get average LVS titers well above 5 x 107 IU/mL after a concentration factor of 200x the original volume. Using these LVS particles, we have been able to establish several stable cell lines of H9 hESCs expressing shRNAs for several genes by infecting the cells at high MOIs. The method presented here bypasses the use of tedious and time-consuming methods of clonal selection and expansion while still getting knockdown efficiencies close to 100%, as shown by one of the representative western blot results for L2HGDH gene knockdown in Figure 4. Following are some of the recommendations for a successful approach.
The use of a low passage number of HEK293T cells is required (ideally below 30) for high viral titers. The confluency of HEK293T cells is another important factor in this regard. HEK293T cells must be at least 80% confluent before transfection as cell-cell contact greatly reduces the cytotoxicity and improves virus production. The vectors used must be prepared using endotoxin-free plasmid isolation kits followed by ethanol precipitation and reconstituted at a minimum concentration of 1mg/mL before usage for transfection. There could be batch-to-batch variations of PEI solutions prepared at different dates. For that reason, each batch of PEI solution must be first tested for performance by using GFP vectors, and the solution must be stored at −20 °C in single-use aliquots to avoid repeated freeze-thaw cycles. For high titers, the collection of LVS supernatant post 72 h of transfection of HEK293T cells is only recommended provided the cells are healthy and attached. Filtration of the LVS-containing supernatant greatly increases the solubility of LVS post ultracentrifugation. It is not advised to store the LVS supernatant at 4 °C for more than 5 days as it will result in a significant reduction of virus titers. Sucrose gradient during ultracentrifugation is required for optimal titers. Temperatures must be maintained close to 4 °C during all the steps of ultracentrifugation. Overnight incubation of concentrated LVS particles in DPBS is required for complete resuspension. Centrifugation following resuspension of LVS removes cellular debris (if any) as it can cause cytotoxicity when used on target cells. After the infection of H9 cells, it is important to replace the virus-containing media with fresh growth media after 8 h as prolonged incubation with the virus would result in significant cell death of the H9 hESCs. The puromycin selection should only be started when the confluency is more than 80%.
The method described above was originally adapted for the expression of shRNAs. A similar approach could be used for the ectopic expression of cDNAs to be cloned into expression vectors. However, a major limitation of lentiviral mediated gene delivery is the size constraint. As size increases, the packaging of lentiviral vectors is not optimal, which results in reduced viral titers23. Future research should be directed to optimize the protocol to overcome these constraints.
The authors have nothing to disclose.
This work was supported by research grants from the United Arab Emirates University (UAEU), grant # 31R170 (Zayed Center for Health Sciences) and # 12R010 (UAEU-AUA grant). We thank Prof Randall Morse (Wadsworth Center, Albany, NY) for helping us to edit the manuscript for style and grammar.
All data are available upon request.
2-Mercaptoethanol | Invitrogen | 31350010 | |
38.5 mL, Sterile + Certified Free Open-Top Thinwall Ultra-Clear Tubes | Beckman Coulter | C14292 | |
Accutase | Stem Cell Technologies | 7920 | |
bFGF Recombinant human | Invitrogen | PHG0261 | |
Bovine serum albumin FRAC V | Invitrogen | 15260037 | |
Corning Matrigel Basement Membrane Matrix, LDEV-free | Corning | 354234 | |
Cyclopamine | Stem Cell Technologies | 72074 | |
DMEM media | Invitrogen | 11995073 | |
DMEM Nutrient mix F12 | Invitrogen | 11320033 | |
DPBS w/o: Ca and Mg | PAN Biotech | P04-36500 | |
Fetal bovie serum | Invitrogen | 10270106 | |
GAPDH (14C10) Rabbit mAb Antibody | CST | 2118S | |
Gentle Cell Dissociation Reagent | Stem Cell Technologies | 7174 | |
HyClone Non Essential Amino Acids (NEAA) 100X Solution | GE healthcare | SH30238.01 | |
L Glutamine, 100X | Invitrogen | 2924190090 | |
L2HGDH Polyclonal antibody | Proteintech | 15707-1-AP | |
L2HGDH shRNA | Macrogen | Seq: CGCATTCTTCATGTGAGAAAT | |
Lipofectamine 3000 kit | Thermo Fisher | L3000001 | |
mTesR1 complete media | Stem Cell Technologies | 85850 | |
Neurobasal medium 1X CTS | Invitrogen | A1371201 | |
Neuropan 2 Supplement 100x | PAN Biotech | P07-11050 | |
Neuropan 27 Supplement 50x | PAN Biotech | P07-07200 | |
Penicillin streptomycin SOL | Invitrogen | 15140122 | |
pLKO.1 TRC vector | Addgene | 10878 | |
pLL3.7 vector | Addgene | 11795 | |
pMD2.G | Addgene | 12259 | |
Polybrene infection reagent | Sigma | TR1003- G | |
Polyethylenimine, branched | Sigma | 408727 | |
psPAX2.0 | Addgene | 12260 | |
Purmorphamine | Tocris | 4551/10 | |
Puromycin | Invitrogen | A1113802 | |
ROCK inhibitor Y-27632 dihydrochloride |
Tocris | 1254 | |
SB 431542 | Tocris | 1614/10 | |
Trypsin .05% EDTA | Invitrogen | 25300062 | |
XAV 939 | Tocris | 3748/10 |