sgRNA library construction using Molecular Chipper

The overall procedure of the Molecular Chipper procedure follows the scheme in Fig. 1b, using EcoP15I digestion to obtain random 19mers from input DNA.

As input DNA, genomic DNA fragments of mature miRNA and flanking sequences of 17 mouse miRNAs or miRNA clusters (Supplementary Fig. 1b) were prepared by PCR amplification from previously cloned miRNA expression constructs^19,32,33. The miRNA and flanking regions range from 362 to 1,026 bp (Supplementary Table 1). PCR was performed using primers gcctcgatcctccctttatc and aacgcgatcaccactttgta, which are located in the vector sequences outside the miRNA genomic DNA fragments. All PCR products were confirmed by their predicted sizes by running on agarose gels. The PCR products were purified using QIAquick PCR Purification Kit (Qiagen) and pooled together in the same molar ratio. To remove most of extra vector DNA sequences in the PCR products, 80 μg of pooled PCR products were digested with BsrGI, whose sites are immediately flanking the miRNA genomic sequences, followed by gel purification of the DNA fragments ranging 200–1,000 bp using QIAquick Gel Extraction Kit (Qiagen).

We next generated large randomly ligated products before random fragmentation. This step is optional and will not be required if large pieces of input DNA, such as bacterial artificial chromosome clones were used. This step was added to avoid biasing against regions located near the ends of PCR products (because we will perform a size selection after this fragmentation step, and sequences close to the PCR product ends will be represented by very small fragments after fragmentation, and thus will be under-represented in the final library). Specifically, 20 μg of the BsrG1-digested and purified DNA pool were then ligated using 80,000 units of T4 DNA ligase (NEB) in a 1,000-μl ligation reaction for 3 h at 37 °C, followed by ethanol precipitation (add 10% volume of sodium acetate, pH 5.2, and 2 volumes of 100% ethanol; precipitate in −20 °C for 1 h; spin down, wash with 70% ethanol and air dry) and resuspension in water. The sizes of ligation products were checked on agarose gel, which were >10 kb on average. To generate random DNA fragments, 14 μg of the ligated DNA in 120 μl of water were sonicated in a S220 Focused-ultrasonicator (Covaris) for 90 s to result in fragments peaking at sizes of 400–450 bp (peak power=140 V, duty factor=5, cycle/burst=200 and average power=7). Sonicated fragments were repaired by a 150-μl End Repair reaction with 15 μl of the End Repairing Enzyme Mix (NEB), followed by agarose gel purification of the 400–450-bp DNA fragments.

To obtain fragment ends from both the ends of the random DNA fragments, 12 μg of the DNA fragments were ligated with 20,000 units of T4 DNA ligase (NEB) in a 300-μl reaction for 3 h at 37°C, at a ∼1:10 molar ratio to 6.0 μg of an EcoP15I-adaptor that was prepared by annealing two oligonucleotides aaaactcgagcagcagtggatccG and/5phos/Cggatccactgctgctcgag (IDT). The annealed DNA adaptor contains an EcoP15I site (in bold) followed by a total 8-bp spacer, including a BamHI site (underlined) and a G (capitalized) at the end for later sgRNA cloning. The adaptor-ligated DNA fragments were purified from adaptor monomer and other non-specific bands by running on 1% agarose gels. An amount of 5 μg of the EcoP15I-adaptor-ligated gel-purified DNA was digested by 100 units of EcoP15I enzyme (NEB) in a 300-μl reaction for 1 h at 37 °C. After digestion, EcoP15I digestion reaction was cleaned by phenol/chloroform extraction and ethanol precipitation. Precipitated digestion products were gel-purified (on 4% low-melting-point agarose gel) to obtain a ∼38-bp DNA fragment pool (EcoP15I-adaptor+19/17 bases from ends of random DNA fragments). To ligate to the rest of sgRNA backbone, 280 ng of the purified 38-bp DNA pool was ligated in a 50-μl reaction with 4,000 units of T4 DNA ligase for 3 h at 37°C, at a 1:5 molar ratio to 2.75 μg of an sgRNA backbone adaptor. The sgRNA backbone adaptor contains two Ns for binding to overhangs from EcoP15I digestion products, the remaining sgRNA sequence (without the target recognition domain), a polyT stretch for polymerase III transcriptional termination and an HindIII site for cloning. This sgRNA backbone adaptor was prepared by annealing two oligonucleotides below (IDT), followed by gel purification on 4% low-melting-point agarose gels to eliminate improperly annealed products. /5phos/nngttttagagctagaaatagcaagttaaaataaggctagtccgttatcaacttgaaaaagtggcaccgagtcggtgc-tttttttaagctttat and ataaagcttaaaaaaagcaccgactcggtgccactttttcaagttgataac-ggactagccttattttaacttgctatttctagctctaaaac. The ligated sgRNA DNA pool was cleaned using QIAquick PCR Purification Kit (Qiagen), digested in 50 μl with 20 units each of BamHI and HindIII overnight at 37 °C, purified on 4% low-melting-point agarose gel to obtain a ∼115-bp sgRNA pool. This sgRNA pool was quantified by SYBR Safe Gel Stain (Invitrogen) on a fluorometer, and ligated into BamHI–HindIII sites of a retroviral vector pSUPER-CRISPR (see Constructs), which contains a U6 promoter and a puromycin selection marker. Ligation products were transformed by electroporation into competent NEB5 alpha cells (NEB). Several small fractions of transformation were plated, which led to an estimate of 1.54 million total transformed clones. Transformation culture was grown overnight in 100 ml of LB medium containing 100 μg ml⁻¹ of ampicillin for plasmid DNA preparation.