We describe the production strategy of integrase-deficient lentiviral vectors (IDLVs) as vehicles for delivering CRISPR/Cas9 to cells. With an ability to mediate quick and robust gene editing in cells, IDLVs present a safer and equally effective vector platform for gene delivery compared to integrase-competent vectors.
Lentiviral vectors are an ideal choice for delivering gene-editing components to cells due to their capacity for stably transducing a broad range of cells and mediating high levels of gene expression. However, their ability to integrate into the host cell genome enhances the risk of insertional mutagenicity and thus raises safety concerns and limits their usage in clinical settings. Further, the persistent expression of gene-editing components delivered by these integration-competent lentiviral vectors (ICLVs) increases the probability of promiscuous gene targeting. As an alternative, a new generation of integrase-deficient lentiviral vectors (IDLVs) has been developed that addresses many of these concerns. Here the production protocol of a new and improved IDLV platform for CRISPR-mediated gene editing and list the steps involved in the purification and concentration of such vectors is described and their transduction and gene-editing efficiency using HEK-293T cells was demonstrated. This protocol is easily scalable and can be used to generate high titer IDLVs that are capable of transducing cells in vitro and in vivo. Moreover, this protocol can be easily adapted for the production of ICLVs.
Precise gene editing forms the cornerstone of major biomedical advances that involve the development of novel strategies to tackle genetic diseases. At the forefront of gene-editing technologies is the method relying on the usage of the clustered regularly-interspaced short palindromic repeats (CRISPR)/Cas9 system that was initially identified as a component of bacterial immunity against the invasion of viral genetic material (reviewed in references1,2). A major advantage of the CRISPR/Cas9 system over other gene-editing tools, such as zinc-finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs) (reviewed in reference3), is the relative simplicity of plasmid design and construction of CRISPR components — a feature that has powered the expansion of gene-editing from a few specialized laboratories to a much wider research community. Additionally, the simplicity of CRISPR/Cas9 programming and its capacity for multiplexed target recognition have further fueled its popularity as a cost-effective and easy-to-use technology. Among the various methods available to researchers to deliver such gene-editing components to cells, viral vectors remain by far the most popular and efficient system.
Lentiviral vectors (LVs) have emerged as the vehicle of choice to deliver the components of CRISPR/Cas9 system in vivo for diverse applications4,5,6,7. Several key features make LVs a popular choice for this process including their ability to infect both dividing and non-dividing cells, low immunogenicity, and minimal cellular toxicity (reviewed in reference8). As a result, LV-mediated gene therapy has been employed in treatments of infectious diseases, such as HIV-1, HBV, and HSV-1, as well as in the correction of defects underlying human hereditary diseases, such as cystic fibrosis and neo-vascular macular degeneration4,5,7,9,10,11. Moreover, LVs have been effectively modified to perform multiplex gene editing at distinct genomic loci using a single vector system12.
However, the inherent property of LVs to integrate into the host genome can be mutagenic and often handicaps their utility as transgene delivery vehicles, especially in clinical settings. Moreover, since stably-integrated LVs express their transgenes at sustainably high levels, this system is ill-suited for the delivery of gene-editing components such as CRISPR/Cas9; overexpression of Cas9-guide RNA (gRNA), and similar proteins such as ZFNs, are associated with elevated levels of off-target effects, which include undesirable mutations13,14,15,16,17 and can potentially enhance cytotoxicity18. Therefore, to achieve precise gene-editing with minimal off-target effects, it is imperative to design systems that allow for the transient expression of gene editing components.
In recent years, a variety of delivery platforms have been developed to transiently express CRISPR/Cas9 in cells16,19,20,21 (reviewed in reference22). These include methods that rely on directly introducing purified Cas9 along with the appropriate guide RNAs into cells, which was shown to be more effective at targeted gene-editing in comparison to plasmid-mediated transfection16. Studies have demonstrated that ribonucleoprotein (RNP) complexes consisting of guide RNA/Cas9 particles are rapidly turned over after mediating DNA cleavage at their targets, suggesting that short-term expression of these components is sufficient to achieve robust gene editing16. Conceivably, non-integrating viral vector platforms such as adeno-associated viral vectors (AAVs) can provide a viable alternative to deliver gene-editing machinery to cells. Unfortunately, AAV capsids possess significantly lower packaging capability than LVs (<5kb), which severely impedes their ability to package the multi-component CRISPR toolkit within a single vector (reviewed in reference8). It is worth noting that addition of compounds that inhibit histone deacetylases (e.g., sodium butyrate23) or impede the cell cycle (e.g., caffeine24) have been shown to increase lentiviral titers. Despite the recent progress, the transient expression systems developed so far are still impeded by several shortcomings, such as lower production efficiency, which lead to reduced viral titers, and low transduction efficiency of the viruses generated through such approaches25.
Integrase-deficient lentiviral vectors (IDLVs) represent a major advancement in the development of gene-delivery vehicles, as they combine the packaging capability of LVs with the added benefit of AAV-like episomal maintenance in cells. These features help IDLVs largely circumvent the major issues associated with integrating vectors, vis-à-vis continuous overexpression of potentially genotoxic elements and integration-mediated mutagenicity. It was previously demonstrated that IDLVs can be successfully modified to enhance episomal gene expression26,27. With regards to IDLV-mediated CRISPR/Cas9 delivery, low production titers and lower expression of episome-borne genomes relative to integrase-proficient lentiviral systems limits their utility as bona fide tools for delivering genome-editing transgenic constructs. We recently demonstrated that both transgene expression and viral titers associated with IDLV production are significantly enhanced by the inclusion of binding sites for the transcription factor Sp1 within the viral expression cassette28. The modified IDLVs robustly supported CRISPR-mediated gene editing both in vitro (in HEK-293T cells) and in vivo (in post-mitotic brain neurons), while inducing minimal off-target mutations compared to the corresponding ICLV-mediated systems28. Overall, we developed a novel, compact, all-in-one CRISPR toolkit carried on an IDLV platform and outlined the various advantages of using such a delivery vehicle for enhanced gene editing.
Here, the production protocol of the IDLV-CRISPR/Cas9 system is described, including the various steps involved in the assembly, purification, concentration, and titration of IDLVs, as well as strategies to validate the gene-editing efficacy of these vectors. This protocol is easily scalable to meet the needs of different investigators and is designed to successfully generate LV vectors with titers in the range of 1 x 1010 transducing units (TU)/mL. The vectors generated through this protocol can be utilized to efficiently infect several different cell types, including difficult-to-transduce embryonic stem cells, hematopoietic cells (T-cells and macrophages), and cultured and in vivo-injected neurons. Furthermore, the protocol is equally well-suited for the production of integrase-competent lentiviral vectors in similar quantities.
Figure 1: IDLV packaging. (a) Schematic of the wild type integrase protein (b) The modified plasmid was derived from psPAX2 (see Methods, plasmid construction for details). Representative agarose gel image of clones screened for mutated integrase clones. DNA samples prepared using a standard plasmid DNA isolation mini-kit were analyzed by digestion with EcoRV and SphI. The correctly-digested clone (number 5, dashed red box) was further verified by direct (Sanger) sequencing for the D64E substitution in INT. The integrase-deficient packaging cassette was named pBK43. (c) Schematic of the transient transfection protocol employed to generate IDLV-CRISPR/Cas9 vectors, showing 293T cells transfected with VSV-G, packaging, and transgene cassettes (Sp1-CRISPR/Cas9 all-in-one plasmid). Viral particles that bud out from the cell membrane contain the full-length RNA of the vector (expressed from the transgene cassette). The second generation of the IDLV-packaging system was used, which includes the regulatory proteins Tat and Rev. Rev expression is further supplemented from a separate cassette (RSV-REV-plasmid). Abbrev: LTR-Long-terminal repeat, VSV-G, vesicular stomatitis virus G-protein, pCMV-cytomegalovirus promoter; Rous sarcoma virus (RSV) promoter; RRE- (Rev Response Element). Other regulatory elements on the expression cassette include Sp1-binding sites, Rev Response element (RRE), Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element (WPRE), a core-elongation factor 1α promoter (EFS-NC), the vector packaging element ψ (psi), human Cytomegalovirus (hCMV) promoter, and human U6 promoter. Please click here to view a larger version of this figure.
1. Culturing HEK-293T Cells and Seeding Cells for Transfection
NOTE: Human Embryonic Kidney 293T (HEK-293T) cells are grown in DMEM, high glucose media supplemented with 10% bovine calf serum supplemented with iron and growth promoters, and 1x antibiotic-antimycotic solution (100x solution contains 10,000 units penicillin, 10 mg streptomycin and 25 µg amphotericin B per mL). The media is also supplemented with 1x sodium pyruvate, 1x non-essential amino acid mix, and 2 mM L-Glutamine (stock 200 mM L-alanyl-L-glutamine dipeptide in 0.85% NaCl). Cells are cultured in 100 mm tissue culture plates (approximate growth surface area is 55 cm²). A sub-cultivation ratio of 1:10 is used with sub-culturing every 2 – 3 days. Trypsin-EDTA 0.05% is used for dissociation of cells between passages. To maintain consistency between experiments, we recommend testing calf sera when switching to a different lot/batch and monitor any changes in cell growth, transfection efficiency, and vector production.
2. Transfecting HEK-293T Cells Using a Calcium Phosphate-based Protocol
Figure 2: Concentration of the viral particles using double-sucrose gradient protocol. Viral particles collected from the supernatant (SN) are loaded onto gradient sucrose gradient. 70%, 60%, 30%, and 20% sucrose solutions are used to create the gradient. Following centrifugation, the particles collected from the 30 – 60% sucrose fractions are further loaded onto a 20% sucrose cushion and precipitated. The final pellet containing purified viral particles is resuspended in 1x PBS for further usage (see text for further details). Please click here to view a larger version of this figure.
3. Day After Transfection
4. Harvesting Virus
5. Concentration of Viral Particles by Ultracentrifugation
NOTE: We utilize a double-sucrose method of purification that involves two steps: a sucrose gradient step and a sucrose cushion step (Figure 2).
6. Estimation of Viral Titers
Validation of the knockout-efficiency of IDLV-CRISPR/Cas9 vectors
We used GFP-expressing 293T cells as a model to validate the efficiency of CRISPR/Cas9-mediated gene knockout. GFP+ cells were generated by transduction of HEK-293T cells with pLenti-GFP (vBK201a) at an MOI of 0.5 (Figure 3b, "no-virus" panel). The sgRNA-to-GFP/Cas9 all-in-one vector cassette was packaged into IDLV or ICLV particles and the production efficiency was assessed by p24 ELISA (see above; and Figure 3a). The titers of the vectors containing Sp1-binding sites (following concentration) were found to be in the range of 1010 TU/mL. We then assessed the efficiency of GFP knockout using fluorescent microscopy (Figure 3b). To this end, 5 x 105 GFP-expressing 293T cells were seeded into a 6-well plate and we transduced them 24 h later with IDLV- or ICLV-CRISPR/Cas9 at MOIs of 1 and 5 (Figure 3b). The cells were incubated for 24 h, after which the culture medium was replaced and cells were incubated for an additional 48 h, before reseeding the plates for subsequent days of the experiment.
In a recent study, we measured the mean fluorescence intensity of GFP+ cells transduced with ICLV/CRISPR or IDLV/CRISPR components via flow cytometry28. We saw comparable GFP-depletion in both samples 14 days post-transduction (pt) (2 – 4% signal depletion), and nearly-identical GFP depletion 21 days pt (>99% signal depletion)28. In concordance with previous observations, we observed a ~five-fold reduction in the number of GFP-positive cells as early as 7 days pt (Figure 3), with a nearly complete signal depletion observed by 14 days pt (data not shown) following transduction with both ICLV- and IDLV-CRISPR/Cas9 systems. The signal loss was evaluated as the ratio of the cells that remained GFP-positive after treatment and the total number of naïve GFP-positive cells. These results clearly demonstrate that CRISPR/Cas9 constructs delivered by IDLVs are comparable with those delivered by their integrating counterparts in their ability to mediate rapid, robust, and sustained gene editing in dividing cells.
Figure 3: Evaluation of viral titers. (a) by p24-ELISA assay. Titers for the concentrated Sp1-IDLV-CRISPR/Cas9 (black bar) and Sp1-ICLV-CRISPR/Cas9 (white bar) were evaluated. The results are recorded in copy numbers per mL, where 1 ng p24-gag = 1 x 104 viral particles. Bar graph data represents mean ± SD from triplicate experiments. (b) Evaluation of CRISPR-mediated GFP-knockout efficiency. Depletion of GFP signal was compared between ICLV-CRISPR/Cas9- and IDLV-CRISPR/Cas9-transduced 293T GFP+ cells at MOIs of 1 and 5. Un-transduced ("no virus" panel in figure) GFP-positive cells were used as controls. Images were acquired using a fluorescence microscope at 40X magnification at 7 days post transduction. Abbrev: RRE: Rev Response Element, EFS-NC: core-elongation factor 1α promoter, ψ (psi): vector packaging element, hU6: human U6 promoter, Puro: Puromycin-resistance cassette, WPRE: Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element, LTR: Long-terminal repeat. Please click here to view a larger version of this figure.
Supplementary File 1: Plasmids Please click here to download this file.
IDLVs have begun to emerge as the vehicle of choice for in vivo gene-editing, especially in the context of genetic diseases, owing largely to the low risk of mutagenesis associated with these vectors compared to integrating delivery platforms22,28. In the current manuscript, we sought to detail the protocol associated with production of the improved all-in-one IDLV-CRISPR/Cas9 system that was recently developed in our laboratory28.
Modifications to existing platforms
With this method, we were able to generate IDLV-CRISPR/Cas9 and ICLV-CRISPR/Cas9 vectors at titers in the range of 1 x 1010 TU/mL (Figure 3a). This enhancement in production efficiency can be attributed to the addition of Sp1-binding site into all-in-one CRISPR/Cas9 vector cassette. Indeed, we recently reported that inclusion of Sp1 results in a ~2.5-fold increase in the packaging efficiency of IDLV- and ICLV-CRISPR/Cas9 vectors and a ~7-fold increase in the overall functional titers. These results are in agreement with earlier work from various groups highlighting Sp1 as a key regulator of wild-type HIV-130,31,32,33,34,35.
Critical steps and troubleshooting
As pointed out above, Sp1-IDLVs carrying CRISPR/Cas9 transgenes were capable of generating titers in the vicinity of 1 x 1010 TU/mL per ~5 x 107 producer cells. Titers lower than these would be indicative of errors in the production process, in which case the following crucial points should be considered for titer improvement: 1) It is recommended that the producer cells preferably be of low passage number, with replacement after ≥15 passages and/or when slower growth is observed. 2) The choice of cell media components is important in terms of production efficiency. For instance, in place of the commonly-used fetal bovine serum, we find that the usage of Cosmic Calf Serum consistently improved cell growth and viral production, while being cost-effective at the same time. 3) The production fitness of the different HEK lines should be carefully evaluated. For example, we found a ~threefold difference in viral production yield between 293T cells and 293-FT (see Table of Materials) cells (data not shown). 4) It is recommended that cells be transfected when they are roughly 70 - 80% confluent, with lower cell densities resulting in premature cell death due to viral toxicity, and higher densities resulting in a marked drop in production efficiency. As a rule of thumb, we suggest accounting for a cell density that allows cells to undergo one additional round of cell division post-transfection. 5) Lastly, the efficiency of transfection is highly dependent on the pH of 2x BBS, which is to be maintained at exactly 6.95 for ideal transfection. It is therefore highly recommended that each new batch of 2x BBS be checked on a pilot-transfection scale.
Vector handling and safety
There are several important safety considerations during the production of IDLV and ICLV vectors with this protocol. First, working with lentiviruses requires Bio-Safety Level II containment. Despite the safety features afforded by SIN vectors, residual transcriptional activity from SIN vectors has been reported36. Furthermore, previous work has demonstrated that IDLV- and ICLV-genomes can be productively rescued by HIV-127. Therefore, it is strongly recommended that replication competence assays (RCA) be performed, especially when concentrated lentiviruses are being used37. For safety procedures regarding handling of lentiviral vector preparations, see Biosafety in Microbiological and Biomedical Laboratories, 4th edition, published by the Centers for Disease Control (CDC), which can be found online38. On the same note, third generation packaging systems39 with enhanced biosafety features can be used to package IDLVs and ICLVs, albeit with lower efficiency than that of the second-generation system.
Significance and future directions
Overall, the production protocol for IDLVs for CRISPR/Cas9- mediated gene editing described here represents the first evaluation of the efficiency of this system in rapidly-dividing cells. Using GFP-positive HEK-293T cells, we demonstrated that Sp1-CRISPR/Cas9 delivered by IDLVs can efficiently and quickly edit GFP in these cells, with similar kinetics as ICLVs. These observations are broadly consistent with the results from previous work using ribonucleoprotein complexes (Cas9 RNPs) for gene editing through non-viral transfection methods. This novel platform further enriches the ever-expanding toolbox for the delivery of gene-editing components and other molecular cargos to cells.
As discussed earlier, despite the recent progress in developing lentiviral-based gene-delivery systems, there are few to no options for platforms that allow the sustainable production of high-titer viral vectors for rapid, transient, and targeted gene manipulation. Through the incorporation of binding motifs for a highly-expressed transcription factor in the transgene expression cassette, we were able to simultaneously address several of these issues. This simple yet critical manipulation opens up a number of avenues for developing similar delivery platforms. It would be particularly useful to test if transgene expression can be enhanced through either the addition of numerous copies of the same binding motif, or by multiplexing the Sp1 motif with binding sites for other transcription factors. With such tools at our disposal, it is tempting to speculate that expression from relatively weak tissue-specific promoters, such as those for human Synapsin I (hSyn) and mouse calcium/calmodulin-dependent protein kinase II (CaMKII), could be significantly improved both in vitro and in vivo.
While the current study was limited to testing the gene knockdown capability of IDLV-delivered CRISPR/Cas9, in principle, the versatility of our system would be amenable to diverse gene-editing applications, such as those relying on the catalytically-inactive dCas940. Finally, combined with smaller endonucleases, such as SaCas941 and Cpf142, the platform described in this study can be adopted toward establishing highly-efficient AAV-based gene-delivery systems in a relatively short span of time, which would provide the added advantage of low immunogenicity compared to lentiviral vectors. Needless to say, such approaches would be a positive step towards the development of novel, efficient, and clinically safe viral vectors.
The authors have nothing to disclose.
We would like to thank the Department of Neurobiology, Duke University School of Medicine and Dean's Office for Basic Science, Duke University. We also thank members of the Duke Viral Vector Core for comments on the manuscript. Plasmid pLenti CRISPRv2 was gift from Feng Zhang (Broad Institute). The LV-packaging system including the plasmids psPAX2, VSV-G, pMD2.G and pRSV-Rev was a kind gift from Didier Trono (EPFL, Switzerland). Financial support for this work was provided by the University Of South Carolina School Of Medicine, grant RDF18080-E202 (B.K).
Equipment | |||
Optima XPN-80 Ultracentrifuge | Beckman Coulter | A99839 | |
Allegra 25R tabletop centrifuge | Beckman Coulter | 369434 | |
xMark Microplate Absorbance plate reader | Bio-Rad | 1681150 | |
BD FACS | Becton Dickinson | 338960 | |
Inverted fluorescence microscope | Leica | DM IRB2 | |
0.45-μm filter unit, 500mL | Corning | 430773 | |
Conical bottom ultracentrifugation tubes | Seton Scientific | 5067 | |
Conical tube adapters | Seton Scientific | PN 4230 | |
SW32Ti swinging-bucket rotor | Beckman Coulter | 369650 | |
15 mL conical centrifuge tubes | Corning | 430791 | |
50mL conical centrifuge tubes | Corning | 430291 | |
High-binding 96-well plates | Corning | 3366 | |
150 mm TC-Treated Cell Culture dishes with 20 mm Grid | Corning | 353025 | |
100mm TC-Treated Culture Dish | Corning | 430167 | |
0.22 μM filter unit, 1L | Corning | 430513 | |
QIAprep Spin Miniprep Kit (50) | Qiagen | 27104 | |
Tissue culture pipettes, 5 mL | Corning | 4487 | |
Tissue culture pipettes, 10 mL | Corning | 4488 | |
Tissue culture pipettes, 25 mL | Corning | 4489 | |
Hemacytometer with cover slips | Cole-Parmer | UX-79001-00 | |
Name | Company | Catalog Number | Comments |
Cell culture reagents | |||
Human embryonic kidney 293T (HEK 293T) cells | ATCC | CRL-3216 | |
293FT cells | Thermo Fisher Scientific | R70007 | |
DMEM, high glucose media | Gibco | 11965 | |
Cosmic Calf Serum | Hyclone | SH30087.04 | |
Antibiotic-antimycotic solution, 100X | Sigma Aldrich | A5955-100ML | |
Sodium pyruvate | Sigma Aldrich | S8636-100ML | |
Non-Essential Amino Acid (NEAA) | Hyclone | SH30087.04 | |
RPMI 1640 media | Thermo Fisher Scientific | 11875-085 | |
GlutaMAX | Thermo Fisher Scientific | 35050061 | |
Trypsin-EDTA 0.05% | Gibco | 25300054 | |
BES (N, N-bis (2-hydroxyethyl)-2-amino-ethanesulfonic acid) | Sigma Aldrich | B9879 – BES | |
Gelatin | Sigma Aldrich | G1800-100G | |
Name | Company | Catalog Number | Comments |
p24 ELISA reagents | |||
Monoclonal anti-p24 antibody | NIH AIDS Research and Reference Reagent Program | 3537 | |
HIV-1 standards | NIH AIDS Research and Reference Reagent Program | SP968F | |
Polyclonal rabbit anti-p24 antibody | NIH AIDS Research and Reference Reagent Program | SP451T | |
Goat anti-rabbit horseradish peroxidase IgG | Sigma Aldrich | 12-348 | Working concentration 1:1500 |
Normal mouse serum, Sterile, 500mL | Equitech-Bio | SM30-0500 | |
Goat serum, Sterile, 10mL | Sigma | G9023 | Working concentration 1:1000 |
TMB peroxidase substrate | KPL | 5120-0076 | Working concentration 1:10,000 |
Name | Company | Catalog Number | Comments |
Plasmids | |||
psPAX2 | Addgene | 12260 | |
pMD2.G | Addgene | 12259 | |
pRSV-Rev | Addgene | 12253 | |
lentiCRISPR v2 | Addgene | 52961 | |
Name | Company | Catalog Number | Comments |
Restriction enzymes | |||
BsrGI | New England Biolabs | R0575S | |
BsmBI | New England Biolabs | R0580S | |
EcoRV | New England Biolabs | R0195S | |
KpnI | New England Biolabs | R0142S | |
PacI | New England Biolabs | R0547S | |
SphI | New England Biolabs | R0182S |