Comparison and optimization of CRISPR/dCas9/gRNA genome-labeling systems for live cell imaging
- 3.7k Downloads
CRISPR/dCas9 binds precisely to defined genomic sequences through targeting of guide RNA (gRNA) sequences. In vivo imaging of genomic loci can be achieved by recruiting fluorescent proteins using either dCas9 or gRNA. We thoroughly validate and compare the effectiveness and specificity of several dCas9/gRNA genome labeling systems. Surprisingly, we discover that in the gRNA-labeling strategies, accumulation of tagged gRNA transcripts leads to non-specific labeling foci. Furthermore, we develop novel bimolecular fluorescence complementation (BIFC) methods that combine the advantages of both dCas9-labeling and gRNA-labeling strategies. The BIFC-dCas9/gRNA methods demonstrate high signal-to-noise ratios and have no non-specific foci.
KeywordsGenome labeling CRISPR/dCas9 Bimolecular fluorescence complementation (BIFC)
The dynamic localization of a particular genomic locus in a three-dimensional (3D) genome has been proposed to regulate various genome functions including gene transcription, DNA recombination, DNA replication, and DNA repair [1, 2]. Until recently, several strategies have been developed to trace the dynamic movement of genomic loci in living cells . Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9), an RNA-guided endonuclease that mediates highly sequence-specific binding and efficient cleavage on genomic DNA, has been extensively developed recently for genome editing [4, 5, 6, 7]. On the other hand, a nuclease-deficient Cas9 (dCas9) could bind to a guide RNA (gRNA)-specific genomic locus, where by recruiting various effectors it could achieve precise and programmed transcription activation and repression, epigenetic remodulations of local histone and DNA modifications, labeling and visualization of the genomic locus, and single base genome mutagenesis [8, 9, 10]. Various dCas9/gRNA systems have been designed to label genomic loci in live cells. In its first version, direct fusion of fluorescent proteins such as green fluorescent protein (GFP) with dCas9 protein was used by Huang’s laboratory . To increase the signals, a SunTag that contains multiple copies (24X) of GCN4 peptide epitopes has been added to the C-terminal dCas9 . Fusion with single-chain fragment variable (scFv) antibody against GCN4 peptide allows more copies of fluorescent proteins to be recruited to a single tethered dCas9/gRNA complex. Recently, tandem FP11-tags were also fused to dCas9 to allow proportional enhancement of the fluorescence signal . To achieve simultaneous labeling of several genomic loci at the same time, two approaches have been developed. First, several CRISPR/Cas9 orthologous proteins from distinct bacterial species that have different gRNA-binding specificities could be fused to different fluorescent proteins [14, 15]. On the other hand, both RNA aptamer binding effectors [16, 17, 18, 19] and Pumilio/FBF (PUF) RNA-binding proteins  have been utilized to label the different gRNAs, which could work with the same dCas9 protein. In addition, multiple copies of RNA motifs could be fused to the gRNA to greatly amplify the signals. Here we compare several of the latest gRNA labeling and dCas9 labeling systems in the same experimental settings such as the cell type, transfection method, and gRNA expression cassette, as well as genomic targets. We have identified and solved the intrinsic nonspecific labeling issue associated with the gRNA labeling methods. We also developed novel bimolecular fluorescence complementation (BIFC)-dCas9/gRNA methods that combine the specificities of both dCas9 and gRNAs. The BIFC-dCas9/gRNA methods have high signal-to-noise ratio and no nonspecific foci.
Comparison of dCas9/gRNA genome-labeling systems for chromosome imaging
Dissection of the nonspecific labeling foci associated with gRNA labeling for chromosome imaging
BIFC-dCas9/gRNA strategies for optimal chromosome imaging
Bimolecular fluorescence complementation (BIFC) was originally developed to validate protein-protein interactions through detection of fluorescence from the assembly of fluorescent protein fragments tethered to interacting proteins [25, 26]. BIFC measures the spatial and temporal changes for specific protein interactions but not for their noninteracting subunits. This property has been recently used to reduce the background fluorescence generated from free unbound fluorescent proteins, which could increase the signal-to-noise ratio and labeling efficiency for both RNA and protein labeling [27, 28].
The recently developed CRISPR/Cas9 system provides a simple way to efficiently recognize and manipulate the targeted genome sequences in organisms [4, 5, 6, 7]. In particular, several dCas9/gRNA genome-labeling methods that differentially tether fluorescent proteins with dCas9/gRNA complexes have been developed recently to track genome dynamics in living cells [11, 12, 13, 14, 15, 16, 17, 18, 19, 20]. Here, we compared representative dCas9 and gRNA labeling strategies in the same experimental setting and cellular context. Our results have shown that, although they are able to label multiple foci at the same time, gRNA labeling strategies have intrinsic nonspecific labeling foci from the accumulation of gRNA transcripts, and this is more severe when more RNA-binding motifs are used. Finally, we developed several BIFC-dCas9/gRNA labeling methods that have a higher signal-to-noise ratio and also no nonspecific foci.
Several key issues have been mainly considered for optimizing the dCas9/gRNA genome labeling. First, careful validation of the observed fluorescent signals is necessary. For example, fluorescent in situ hybridization (FISH) has only been used in few studies , likely due to its technical difficulties, especially in combination with live cell imaging. Therefore, co-localization with other known labeling markers has been frequently used for telomere and centromere labeling [18, 19, 20]. In addition, co-labeling pairs of two nearby and distant foci could be performed [11, 14, 15, 16]. More essentially, adequate controls, especially negative controls, should be included to clarify potential nonspecific and specific artifacts. Moreover, simultaneously labeling several genome loci with different colors in the same cells is necessary to visualize dynamic interactions of genomic regions (e.g., the interaction between enhancer and promoter). Although direct labeling of different dCas9 orthologs with distinct fluorescent proteins has been developed [14, 15], this requires co-transfections of multiple dCas9 orthologs with their corresponding gRNAs into the same cell. Labeling gRNAs with different RNA motifs seems to be an easier way to image multiple genomic loci [16, 17, 18, 19, 20]. However, here we observed that the binding of corresponding proteins on such RNA motifs could lead to stabilization and accumulation of gRNA transcripts surrounding their transcription cassettes and formation of nonspecific labeling foci. Finally, the ultimate goal of dCas9/gRNA imaging is to label genomic loci with low- or nonrepetitive regions with a minimal number of targeted dCas9/gRNA complexes. Increasing the numbers of fluorescent proteins tethered to each dCas9/gRNA complex could lead to amplification of fluorescent signals. On the other hand, reducing the background noise, such as background fluorescence generated from free unbound fluorescent proteins, could also greatly enhance the signal-to-noise ratio. The BIFC-dCas9/gRNA methods developed here showed no nonspecific foci, especially those artifacts originating from gRNA transcripts. More importantly, the BIFC-dCas9/gRNA methods have a higher signal-to-noise ratio and could be the best choices for low-repeat-containing genome regions. The multiple-color choices for BIFC-dCas9/gRNA could be further extended to different bimolecular fluorescent complexes that have distinct spectrums  and could also be further combined with the CRISPRainbow system .
We carefully compared current major CRISPR/dCas9/gRNA methodologies for genome labeling and provided the community with sets of validated reagents and protocols. In addition, we surprisingly discovered that in the gRNA-labeling strategies, accumulation of tagged gRNA transcripts could lead to significant nonspecific labeling foci in the absence and presence of dCas9. More importantly, we developed novel bimolecular fluorescence complementation (BIFC) methods that combine the advantages of current dCas9 labeling and gRNA labeling strategies. The BIFC-dCas9/gRNA methods demonstrate a higher signal-to-noise ratio compared to other existing dCas9/gRNA labeling systems and have absolutely no nonspecific foci.
The pHRdSV40-NLS-dCas9-24xGCN4_v4-NLS-P2A-BFP-dWPRE (Addgene #60910), pHR-scFv-GCN4-sfGFP-GB1-dWPRE (Addgene #60907), pHAGE-EFS-MCP-3XBFPnls (Addgene #75384), pHAGE-EFS-PCP-3XGFPnls (Addgene #75385), pLH-sgRNA1-2XMS2 (Addgene #75389), pAC1373-pX-sgRNA-25xPBSa (Addgene #71890), pAC1399-pX-sgRNA-20xPBSc (Addgene #71899), pAC1380-pX-sgRNA-5xPBSc (Addgene #71895), pAC1404-pCR8-mRuby2_NLSPUFa (Addgene #71902), pAC1403-pCR8-Clover_NLSPUFc (Addgene #71901), and pAC1358-pmax-NLSPUFc_VP64 (Addgene #71884) plasmids were obtained from Addgene, Cambridge, MA, USA. Their detailed information is listed in Ref. , Additional file 1: Table S1. The pcDNA3.1-dCas9 plasmid was described previously . A list of the new plasmids generated in this work is provided in Ref. , Additional file 1: Table S2. Schemes and nucleotide sequences for those plasmids are also listed in Ref. . A list of gRNAs used in this work is presented in Additional file 1: Table S3 (Ref. ).
For scFv-VenusN 173 and scFv-VenusC 155 constructions, scFv-sfGFP-gb1-nls was digested with BamHI and NotI. The gb1-nls was amplified from scFv-sfGFP-gb1-nls and then fused with VenusN 173/VenusC 155 (amino acid residues 1–173, 155–238, respectively) by polymerase chain reaction (PCR). The gb1-nls-VenusN 173/VenusC 155 fragments were inserted back to scFv-sfGFP-gb1-nls by In-Fusion cloning.
For NLS-HA-MCP VenusN 173 construction, pHAGE-EFS-MCP-3XBFPnls was digested by NcoI and XbaI. MCP and VenusN 173 (amino acid residues 1–173) were amplified by oligos containing nuclear localization signal (NLS)-hemaglutinin (HA) sequences and were fused by overlap PCR to generate an NLS-HA-MCP-VenusN173 fragment, which was then inserted into the digested pHAGE-EFS-MCP-3XBFPnls vector by Gibson ligation.
Cell culture and transfection
HEK293T, B16, HeLa, and U2OS cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 100 unit/ml penicillin, 100 μg/ml streptomycin, and 10% fetal bovine serum. The cells were plated in 35-mm glass-bottom dishes the day before transfection, and were transfected with the indicated plasmids (see Ref. , Additional file 1: Table S4) by VigoFect (Vigorous Biotechnology, Beijing, China).
Image acquisition and analysis
All images were taken on an UltraVIEW VoX spinning disc microscope (PerkinElmer, Hopkinton, MA, USA). The microscope incubation chamber was maintained at 37 °C and 5% CO2 when we acquired the images. Z-stack images were taken with a step size of 500 nm and enough steps to cover the depth of each nucleus.
Co-localization analysis was carried out using the Volocity software’s “Co-localization” function. The number counting of foci was performed by the “Measurement, Find Objects” function in Volocity software.
To measure the signal-to-noise ratio, a line was drawn across the spot first, and the “Plot Profile” function in ImageJ was used to generate an intensity profile [16, 19, 29]. To calculate the signal-to-noise ratio, the intensity profile was subjected to Gaussian calibration with a baseline, and then the highest intensity value of each peak was divided by the baseline value. The baseline (background noise) was defined as the fluorescence intensity of the nucleus (i.e., unbound fluorescent protein).
The ChIP procedure was performed as described previously . Briefly, cells were crosslinked with 1% formaldehyde for 5 min at room temperature, and the formaldehyde was then inactivated by the addition of 125 mM glycine for 5 min at room temperature. After cell lysis and sonication, the chromatin extracts were incubated with GFP-binding protein (GBP) beads overnight at 4 °C. After washing and reverse crosslinking, DNA was purified for qPCR quantification with specific primers (see Ref. , Additional file 1: Table S3).
RNA extraction and RT-qPCR analysis
Cells were harvested 48 h post-transfection. Total cellular RNA was extracted using the Direct-zol™ RNA MiniPrep Kit (Zymo Research, Irvine, CA, USA). We used 100 ng total RNA to synthesize complementary DNA (cDNA) with the ImProm-II™ Reverse Transcriptase kit (Promega, Madison, WI, USA). The KAPA SYBR Universal 2× quantitative PCR kit (KAPA Biosystems, Wilmington, MA, USA) was used for qPCR reactions. Gene expression was normalized to the expression of the GAPDH gene. The qPCR primer sequences are included in Ref. , Additional file 1: Table S3.
We thank Drs. Haoyi Wang (Institute of Zoology, Chinese Academy of Sciences), Hanhui Ma (UMass Medical School), and Yujie Sun (Peking University) for reagents and members of Y.Z.’s laboratory for helpful discussions and support. We thank the municipal government of Beijing and the Ministry of Science and Technology of China for funds allocated to the National Institute of Biological Sciences, Beijing.
This research was supported by the “Hundred, Thousand and Ten Thousand Talent Project” by the Beijing municipal government (2017A02), the “National Thousand Young Talents Program” of China, and the National Natural Science Foundation of China (81572795 and 81773304) to Y.Z.
Availability of data and materials
Datasets generated and analyzed during the current study are available at https://figshare.com/articles/Comparison_and_optimization_of_CRISPR_dCas9_gRNA_genome_labeling_systems_for_live_cell_imaging_Additional_file_1/5914279/2 under an open source license (CC BY 4.0) .
YH and YZ conceived the study. YH, GL, JD, and WL performed experiments and analyzed data. YZ analyzed data and wrote the manuscript with support from all authors. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Ethics approval was not needed for the study.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 10.Nishida K, Arazoe T, Yachie N, Banno S, Kakimoto M, Tabata M, Mochizuki M, Miyabe A, Araki M, Hara KY, et al. Targeted nucleotide editing using hybrid prokaryotic and vertebrate adaptive immune systems. Science. 2016;353.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.