Traditional cDNA-based overexpression techniques have a limited applicability for the overexpression of long noncoding RNAs due to their multiple splice forms with potential functionality. This review reports a protocol using CRISPR technology to overexpress multiple splice variants of a long noncoding RNA.
Long noncoding RNA (lncRNA) biology is a new and exciting field of research, with the number of publications from this field growing exponentially since 2007. These studies have confirmed that lncRNAs are altered in almost all diseases. However, studying the functional roles for lncRNAs in the context of disease remains difficult due to the lack of protein products, tissue-specific expression, low expression levels, complexities in splice forms, and lack of conservation among species. Given the species-specific expression, lncRNA studies are often restricted to human research contexts when studying disease processes. Since lncRNAs function at the molecular level, one way to dissect lncRNA biology is to either remove the lncRNA or overexpress the lncRNA and measure cellular effects. In this article, a written and visualized protocol to overexpress lncRNAs in vitro is presented. As a representative experiment, an lncRNA associated with inflammatory bowel disease, Interferon Gamma Antisense 1 (IFNG-AS1), is shown to be overexpressed in a Jurkat T-cell model. To accomplish this, the activating clustered regularly interspaced short palindromic repeats (CRISPR) technique is used to enable overexpression at the endogenous genomic loci. The activating CRISPR technique targets a set of transcription factors to the transcriptional start site of a gene, enabling a robust overexpression of multiple lncRNA splice forms. This procedure will be broken down into three steps, namely (i) guide RNA (gRNA) design and vector construction, (ii) virus generation and transduction, and (iii) colony screening for overexpression. For this representative experiment, a greater than 20-fold enhancement in IFNG-AS1 in Jurkat T cells was observed.
While most biomedical research has focused on protein-coding transcripts, the majority of transcribed genes actually consists of noncoding RNAs (Ensembl release 93).Current research is beginning to explore this field, with the number of publications on lncRNAs in disease processes rising exponentially between 2007 and 20171. These publications demonstrate that many lncRNAs are associated with disease. However, the molecular mechanisms of these lncRNAs are difficult to study due to their diverse functions as compared to mRNAs. Compounding the problem of understanding the role of lncRNAs in disease, lncRNAs are often expressed at lower levels than coding RNAs2. Additionally, lncRNAs are poorly conserved, which limits the functional studies in human-cell-line-based techniques3. One method to study the mechanism of these novel genes is to endogenously overexpress them in cultured cells. Overexpression studies can provide key information as to the function of specific genes and enable researchers to dissect key molecular pathways.
A new method to activate the transcription of lncRNA genes is based on CRISPR technologies developed initially by the Gersbach laboratory4. This protocol has been adapted for use in lncRNA biology and for the expression of these genes in other model systems. In the CRISPR overexpressing technique, a protein called CRISPR-associated protein 9 (Cas9) can be directed toward a DNA sequence via an antisense gRNA that is recognized by Cas9. Normally, Cas9 will induce DNA cleavage; however, mutations in Cas9, previously developed for the overexpression technique, deactivate this step5. When a transcriptional activator is fused to a "dead" Cas9 (dCas9) and transduced into cell lines, the endogenous overexpression of lncRNAs can be achieved4,6. The addition of additional modifications to the gRNA, enabling additional transcriptional factors to bind to the gRNA, increased the efficacy of the dCas9 gene activation system 10-fold7. Importantly, it was demonstrated that transcriptional activation requires close proximity (<200 bp) to the transcriptional start site genes, enabling a specific upregulation in gene-rich areas7. Unlike CRISPR knockout technologies, dCas9 and gRNA cassettes need to be integrated into the genome to allow for cells to retain an overexpression over multiple generations. One method to achieve this is to use lentiviruses to integrate the dCas9- and gRNA-containing cassettes. After integration, the lncRNA gene expression can be determined.
In this article, a protocol for lncRNA overexpression in a Jurkat T-cell model will be demonstrated. A step-by-step procedure is shown and can also be adapted to adherent cells.
NOTE: It is important to note that this protocol uses replication-deficient lentiviruses. Perform viral handling only after appropriate lab safety training. Bleach all items and surfaces that come in contact with live viruses for a minimum of 10 min after handling. Use a disposable lab coat and face/eye protection, as well as double gloves, at all times. Perform virus work in biosafety level 2+ labs with viral certification. Dedicate tissue culture hoods and incubators to viral work.
1. Vector Design and Generation
NOTE: The best way to identify gRNA sequences is to use online design tools (e.g., http://crispr-era.stanford.edu) (Supplemental Figure 1: gRNA design). For the accompanying representative experiment, a company designed and created the gRNA vectors used in the representative experiment. The dCas9 plasmid was also purchased online.
2. Virus Generation and Particle Count
3. dCas9-VP64 Transduction
4. Selection and Clone Creation
5. gRNA Transduction, Selection, Clone Creation, and Screening
A dual vector system to overexpress the lncRNA IFNG-AS1
The example experiment in this manuscript is the overexpression of a Jurkat T-cell model system expressing the lncRNA IFNG-AS19. IFNG-AS1 is an lncRNA associated with inflammatory bowel disease, that has been seen to regulate Interferon Gamma10. The IFNG-AS1 gene contains three splice variants that all use the same transcriptional start site (Figure 1A). Therefore, gRNA sequences that were 10 and 100 bp away from the transcriptional start site and had an "NGG" sequence upstream were used for directing the Cas9-activating complex to the transcriptional locus of IFNG-AS1 (Figure 1A). A two-plasmid system was used to transduce either dCas9 or gRNAs/enhancers into cells as a single plasmid alone makes viral particle generation difficult due to plasmid size (Figure 1B). To enable the selection of double-transduced cells, the dCas9 vector contained a puromycin (aminonucleoside) resistance gene, and the gRNA-containing plasmid contained a hygromycin (aminoglycoside) resistance gene. gRNAs were fused to an MS2 scaffold sequence that enabled the MS2 scaffold protein to bind to the gRNAs. Fused to the MS2 protein are additional transcriptional activators to enhance the overexpression of IFNG-AS17. Using these vectors, viral particles can be created to transduce this overexpression system into most human cell lines.
Viral titering and colony screening
After generating the dCas9- and gRNA-containing plasmids, plasmid purifications were performed and lentiviruses were created. As lentiviruses randomly integrate their cassettes into the genome, a quantification of the number of viral particles enables the least number of integrations possible. To accomplish this, conditioned-media-containing viruses were measured with a p24 ELISA kit, allowing for the calculation of the number of virions per milliliter. After measuring, both viral supernatants had nearly 1 µg/mL of p24, which allowed the use of 100 µL of virus to transduce the Jurkat cells. After transduction, antibiotic-selected cells were serially diluted and clones were expanded. Cas9 expression was then analyzed by real-time PCR and agarose gel electrophoresis (Figure 2B). Both clones selected were positive for Cas9 expression. To confirm mRNA expression, reverse transcriptase was omitted from the cDNA reaction (Figure 2B). Primers against HPRT1 were used to confirm the presence of RNA in the nontransduced cells.
Measuring IFNG-AS1 gene expression
Using the dCas9 clones as a parental cell line, cells were transduced with either a nontargeting gRNA-containing virus or a virus containing gRNAs against IFNG-AS1. After gRNA transduction, selection, and clone creation analogous to dCas9 cell line creation, the IFNG-AS1 expression was measured to verify overexpression. IFNG-AS1 produces three splice variants, each of which can be detected individually with transcript-specific primers (red) or against all known IFNG-AS1 transcripts (blue) (Figure 3A). All fluorescence curves were exponential with HPRT1 reaching the exponential phase (or, Ct) within a half cycle between control and IFNG-AS1-gRNA-expressing cells (Figure 3B,C). Primers against all known IFNG-AS1 transcripts were the most detectable between experiments with measurements of 20-fold increases in IFNG-AS1 (Figure 3B). While primers against transcripts against 1 and 2 successfully amplified in peripheral blood mononuclear cells (PBMCs), these transcripts were not detectable in Jurkat cells (data not shown). However, when the third transcript of IFNG-AS1 was detectable in concentrated RNA, a five- to tenfold significant increase in IFNG-AS1 levels was seen. These data suggest either alternative splicing of IFNG-AS1 in Jurkat cells exist compared to primary cells. Primers against IFNG-AS1 detected large increases in this gene, thereby validating the activating CRISPR overexpression system.
Figure 1: A dual vector system to overexpress the lncRNA IFNG-AS1. (A) A schematic of the IFNG-AS1 gene structure and the relationship between guide RNA (gRNA) binding sites and the transcriptional start site (TSS). (B) The features of the gRNA and dCAS9 vectors. Please click here to view a larger version of this figure.
Figure 2: Viral titering and colony screening. (A) An example p24 ELISA standard curve for lentivirus-containing, conditioned media. The black dots represent standard samples and the red dots unknown samples. (B) After transducing, selecting, and generating colonies, real-time PCR and gel electrophoresis against dCas9 were performed on the RNA from either nontransduced Jurkat cells (parental) or dCas9-transduced clones. No reverse transcriptase (No RT) controls were performed. Please click here to view a larger version of this figure.
Figure 3: Measuring IFNG-AS1 gene expression. (A) A diagram of the relationship between primer sets and transcript variants. The red arrows represent transcript-specific primers, while the blue arrows indicate primers against all known transcripts. (B and C) Representative average PCR curves and fold-change quantifications for control and IFNG-AS1-overexpressing cells. RFU = relative fluorescence units. N = 3 samples per group. Mean ± SD. *p < 0.05, ***p < 0.001. A Student's t-test was used to calculate the p-values. Please click here to view a larger version of this figure.
Supplemental Figure 1: gRNA design Please click here to download this file.
Supplemental Figure 2: BLAT Please click here to download this file.
Supplemental Table 1: Solutions Please click here to download this file.
Supplemental Table 2: ELISA dilutions Please click here to download this file.
This manuscript presents a protocol for using activating CRISPR to overexpress lncRNAs in vitro. This is an especially important technique when studying long noncoding RNAs as the transcriptomic product is the functional unit. After overexpression, the researcher can, then, use these cells to increase the signal-to-noise ratio when studying binding partners and even measure the cellular consequences of increased levels of lncRNAs. Additionally, as lncRNAs frequently act on cis-genes, this endogenous overexpression technique enables these events to be studied11,12.
This manuscript highlights a generalized protocol for gRNA creation and lentivirus generation, transduction, and selection, and a representative example of lncRNA overexpression in a peripheral blood T cell line was outlined. Additionally, this technique has already been shown to be successful in bone-marrow-derived immune cells13, neurons7, mouse embryonic stem cells14, and kidney epithelial cells4. While this protocol is generally applicable for most cell lines, cells that are hard to transduce might require individual titers as to enable higher infection rates. Additionally, it is important to perform antibiotic kill curves when using new cell lines, as this protocol utilizes a positive selection strategy.
Of note, this protocol has several technical aspects that require special attention. Reproducible and consistent pipetting for real-time PCR is critical to establish reproducible results. Even small differences in volumes will cause highly variable values, thereby making interpretation difficult. In addition to a proper quantitative PCR primer design, controls, including the no-reverse-transcriptase control for the real-time PCR primers, are critical as they confirm that the RNA being quantified is not genomic DNA. The selection of housekeeping genes for real-time PCR is also important in order to adequately perform these experiments. While HPRT1 was chosen in this protocol, other genes of interests targeted by this technique might alter HPRT115. A careful selection of housekeeping genes based on these factors should be taken into consideration.
There are a few drawbacks to activating CRISPR and to particular aspects of this protocol. One potential limitation is the expression of transcripts in highly gene-rich regions as other neighboring genes could be turned on. It is possible that neighboring genes in close proximity may be activated and controls should be performed to assess this. In addition, other limitations include the lack of binding for particular gRNAs to a given DNA sequence. The selection of multiple gRNAs may be required to identify the appropriate gRNA to drive expression. Another caveat to the system is that the cell of interest has to have the capacity to drive the promoters in the cassettes in order to drive the expression of the gRNAs and dCas9. For the studies presented here, the Jurkat T-cell model was able to drive the expression of these components for a successful overexpression system.
Ideally, the protocol described here should be paired with other corroborating information, such as single transcript overexpression data and functional effects from knockdown or knockout studies, to bolster the case for any of the subsequent findings. This technique offers a robust way of overexpressing any gene in a particular cell type. Given the limitations of lncRNA biology, such as poor species conservation, techniques such as the one described here are critical to exploring their function in relevant cell types. This study focused on one lncRNA that has been implicated in inflammatory bowel disease pathophysiology but serves as a model of studying the function of other lncRNAs in human disease biology.
The authors have nothing to disclose.
C.P. is supported by RO1 DK60729 and P30 DK 41301-26. D.P. is supported by a Crohn's & Colitis Foundation (CCFA) career development award, CURE: Digestive Diseases Research Center (DDRC) DK41301, and UCLA Clinical and Translational Science Institute (CTSI) UL1TR0001881. The UCLA virology core was funded by the Center of AIDS Research (CFAR) grant 5P30 AI028697. The UCLA Integrated Molecular Technologies core was supported by CURE/P30 grant DK041301. This work was also supported by the UCLA AIDS Institute.
Ampicillin | Fisher Scientific | BP1760-5 | |
CaCl2 | Sigma Aldrich | C1016 | |
cDNA synthesis kit (iScript) | BioRad | 1708891 | |
dCas9 forward primer | Integrated DNA Technologies | n/a | 5'-TCGCCACAGCATAAAGAAGA |
dCas9 reverse primer | Integrated DNA Technologies | n/a | 5'-CTTTTCATGGTACGCCACCT |
dCas9 vector | Addgene | 50918 | |
DMEM | Corning | 10013CM | |
Ethanol | Acros Organics | 61509-0010 | |
FBS | Sigma Aldrich | F2442 | |
gRNA plasmid | VectorBuilder | VB180119-1195qxv | |
HEK293T cells | ATCC | CRL-1573 | |
HEPES | Sigma Aldrich | H3375 | |
HPRT1 forward primer | Integrated DNA Technologies | n/a | GACCAGTCAACAGGGGACAT |
HPRT1 reverse primer | Integrated DNA Technologies | n/a | GCTTGCGACCTTGACCATCT |
Hygromycin B | Corning | MT3024CR | |
Isopropanol | Fisher Scientific | BP2618-500 | |
Jurkat cells | ATCC | TIB-152 | clone E6-1 |
L-glutamine | Corning | 25005Cl | |
LB agar | Fisher Scientific | BP1425-500 | |
LB broth | Fisher Scientific | BP1426-2 | |
Maxi-prep kit (Plasmid Purification Kit) | Qiagen | 12362 | |
Na2HPO4 | Sigma Aldrich | NIST2186II | |
Optimem I reduced serum media | Gibco | 31985070 | |
p24 elisa | Perkin Elmer | NEK050B | |
PBS | Corning | 21-040-CMR | |
Penicillin and Streptomycin | Corning | 30-002-Cl | |
pMDG2.G | Addgene | #12259 | |
pMDLg/pRRE | Addgene | #60488 | |
Poly-L-Lysine | Sigma Aldrich | P-4832 | |
Polybrene | EMD Millipore | TR-1003 | |
pRSV-REV | Addgene | #12253 | |
Puromycin dihydrochloride | Sigma Aldrich | P8833 | |
RNA purification kit (Aurum RNA mini) | BioRad | 7326820 | |
Sodium Butyrate | Sigma Aldrich | B5887 | |
SYBR green (iTaq universal) | BioRad | 1725122 | |
Triton X-100 | Sigma Aldrich | X100 |