Painting a specific chromosome with CRISPR/Cas9 for live-cell imaging

Visualization of chromosome shapes and dynamics in a live cell is highly desirable and necessary in many areas of cell biology. For example, the copy number of a particular chromosome in cancer cells is often abnormal (e.g., more than two), and therefore probing chromosome copy numbers can aid cancer diagnosis. During interphase, each chromosome exists in its own territory in the nucleus, which can be imaged by fluorescence in situ hybridization (FISH) using sequence-specific probes of different colors [1, 2]. However, such chromosome painting has only been possible in fixed cells, and is not suitable for dynamic monitoring of live cells. Therefore, it would be valuable to visualize DNA replication of one chromosome during interphase, and follow chromosome dynamics in the M phase. Recent development of clustered regularly interspaced short palindromic repeats/CRISPR-associated proteins (CRISPR/Cas) [3] has provided a powerful tool for live-cell imaging of genomic loci [4]. In particular, the nucle-ase defective Cas9 (dCas9) fused with enhanced green fluorescent protein (EGFP) is used to target a particular DNA sequence upstream of a protospacer adjacent motif (PAM) sequence. Such targeting is achieved through Watson-Crick base pairing of ~20-bp single-guide RNA (sgRNA) that is pre-complexed with the dCas9-EGFP protein. The targeted loci can thus be fluorescently labeled in live mammalian cells [5-8]. However, the labeling achieved by this method is usually restricted to the genomic loci that consist of repetitive sequences, and has not been attempted to track an entire chromosome in a live cell. Here we report the specific labeling of a large number of loci in the genome, which makes it possible to paint an entire chromosome in a live cell. To do so, we designed a new strategy using a large number of sgRNAs targeting mainly the non-repetitive regions of the chromosome (Figure 1A and Supplementary information, Figure S1A). To design sgRNAs, we scanned the sequence of the entire chromosome 9 on human reference genome hg19. Each 19-23 bp genome sequence upstream of a PAM sequence NGG was taken as a candidate target region. Because the efficiency of sgRNA binding is dependent on its GC content [9, 10], sgRNAs with GC content of 45%-65% were selected (Supplementary information, Figure S1B). The sgRNAs that could also bind to other chromosomes were removed in order to assure labeling specificity and reduce fluorescent background. Additionally, when multiple targeting sequences overlapped in a chromosome region, only one of them was selected. Among all …