Prime Editing in vivo: Correcting the Leptin Receptor of db/db Mice

Kyung Eun Lee^1,†, Yanping Xu^1,†, Bingchuan Geng¹, Jongsoo Kim¹, Natalie Kellon¹, Michele He¹, Zhentao Zhang¹, Deqiang Li², Doug A. Gouchoe³, Hua Zhu^1,*

Abstract: Genetic diseases can be caused by monogenic diseases, which result from a single gene mutation in the DNA sequence. Many innovative approaches have been developed to cure monogenic genetic diseases, namely by genome editing. A specific type of genomic editing, prime editing, has the potential advantage to edit the human genome without requiring double-strand breaks or donor DNA templates for editing. Additionally, prime editing does not require a precisely positioned protospacer adjacent motif (PAM) sequence, which offers flexible target and more precise genomic editing. Here we detail a novel construction of a prime editing extended guide RNA (pegRNA) to target mutated leptin receptors in B6.BKS(D)-Leprdb/J mice (db/db mice). The pegRNA was then injected into the flexor digitorum brevis (FDB) muscle of db/db mice to demonstrate in vivo efficacy, which resulted in pegRNA mediated base transversion at endogenous base transversion. Genomic DNA sequencing confirmed that prime editing could correct the mutation of leptin receptor gene in db/db mice. Furthermore, prime editing treated skeletal muscle exhibited enhanced leptin receptor signals. Thus, the current study showed in vivo efficacy of prime editing to correct mutant protein and rescue the physiology associated with functional protein.

Keywords: Prime Editing; Leptin Receptor; db/db mouse

1. Introduction

Single genetic mutations in the DNA sequence can sometimes result in monogenic diseases. Most pathogenic alleles arise from specific insertions, deletions, or base substitutions ^[1]. The CRISPR-Cas9 system is currently one of the most well-known gene editing tools that scientists use, which requires the need of double-strand breaks (DSB) or donor DNA (dDNA) template^[2]. However, this may not be the most ideal way to engage in gene editing. A modified version of the CRISPR-Cas9, Cas9 nickase, can offer higher editing efficiency without DSBs and fewer indel byproducts ^[3,4]. However, Cas9 nickase still has several limitations. It requires a very well-positioned target base, in which a protospacer adjacent motif (PAM) sequence should be in the approximately 15 bases from the target site. Additionally, it is limited in the point mutations it can edit, namely a C:G base pair into a T:A base pair and an A:T base pair into a G:C base pair^[5]. This is one of the main drawbacks of Cas9 nickase, as there is a litany of different classes of mutations, signaling the need for expanded approaches for single mutation editing. Recently, Andrew V. Anzalone et al. established a powerful genome-editing tool known as prime editing (PE), which includes a Cas9n-H840A and C-terminal reverse transcriptase (M-MLV RTase)^[6]. The RNA component of PE, prime editing extended guide RNA or pegRNA, could specify target locuses and accurately transmit genomic information from the pegRNA to complete the desired edit ^[6]. PE also offers major advantages since it does not require a PAM sequence, which ultimately offers more targeting flexibility and greater editing precision. After its invention, this technology has been applied in many in vitro studies^[7]. However, while it has been used in vivo^[8-10], its application in animal is limited due to the large size of PE constructs.

B6.BKS(D)-Leprdb/J (db/db) mice offer an ideal in vivo model in order to study the efficacy of PE. The db/db mouse shows the phenotype of obesity caused by G to T point mutation in intron 18 of the leptin receptor gene, which leads to alternative splicing resulting in the lack of C-terminal of leptin receptor. Ultimately, this results in the inability to initiate intracellular signal transduction, causing severe obesity in db/db mice ^[11]. In this manuscript, we describe our design of a guide RNA (gRNA) to target leptin receptor intron 18, around the position of G to T point mutation. The gene editing efficacy was confirmed by Sanger sequencing. Then, a pegRNA was constructed based on the gRNA sequence, which we observed to mediate successful base transversion in vitro . The pegRNA and PE constructs were delivered to the skeletal muscle of db/db mice via established electroporation. Genomic sequencing confirmed the constructs could effectively and precisely correct leptin receptor gene mutation in db/db mice. Furthermore, leptin receptor signaling was significantly enhanced in prime editing treated skeletal muscle, indicating the corrected leptin receptor protein is functional. Taken together, our study provided solid evidence for in vivo application of prime editing to precisely correct disease-causing mutations.

2. Materials and Methods

2.1. Design of gRNA to Target Leptin Receptor

The guide RNAs (gRNAs) were designed using the online predictor CCTop-CRISPR/Cas9 target (https://cctop.cos.uni-heidelberg.de)^[12,13] to target mouse leptin receptor intron 18, and three gRNAs were selected for the testing. The designs are as follows: gRNA1: 5’-CACC CCA TCA AAA TGT AAA CAT C-3’; 5’- AAAC GAT GTT TAC ATT TTG ATG G-3’, gRNA2: 5’- CACC TCA AAA TGT AAA CAT CTT C-3’; 5’ AAAC GAA GAT GTT TAC ATT TTG A-3’ gRNA3: 5’- CACC GTC ATT CAA ACC ATA GTT T-3’; 5’-AAAC AAA CTA TGG TTT GAA TGA-3’. 1 µl of each 100 µM oligos were put in annealing buffer (10 mM Tris-HCl pH 8.0, 50 mM NaCl), and were heated and cooled down by a thermocycler for complementary annealing. Double-strand oligos were ligated into px459 V2.0-eSpCas9 (1.1) plasmid, which was a gift from Yuichiro Miyaoka (Addgene plasmid # 108292; http://n2t.net/addgene:108292; RRID: Addgene_108292)^[14]. The gRNA clone was confirmed by Sanger sequencing.

2.2. Determining the Efficacy of gRNA in vitro by Using C2C12 Cells

First to demonstrate in vitro efficacy, we demonstrated proper transfection into C2C12 cells with gRNAs. To determine the concentration of puromycin to destroy more than 90 % of C2C12 cells, C2C12 cells were seeded on the 6-well plate and treated with puromycin at the concentration of 0, 1, 2.5, 5, 7.5, 10 µg/ml. It was determined that 10 µg/ml was the ideal concentration to destroy C2C12 cells and permit proper transfection. Next, C2C12 cells (1 X 10⁵) were again seeded in a 6-well plate. On day 2, 2 µg of either: constructed gRNA CRSIPR vector or green fluorescent protein (GFP) or vector only construct was transfected into the C2C12 cells by mixing 4 µl of Viafect transfection (Promega, # E4981) with them. On day 3, cells were treated with 10 µg/ml of puromycin for 24-hours followed by changing to growth medium containing high glucose DMEM (Gibco, # 11965092) supplemented with 10% Fetal Bovine Serum (FBS) (VWR, # 97068-091) and 1% penicillin–streptomycin (Sigma, #P0781). On day 6, cells were collected for the genomic DNA isolation. Cas9 sequence PCR primers were designed to amplify a 277 bp product evenly flanking the CRISPR/Cas9 site. The primers were as follows: forward primer, 5’- CAG AGA ACG GAC ACT CTT TG -3’reverse primer -3’, reverse primer, 5’- GCA GAG TCC ATG AAT ATC AAC -3’. Genomic DNA was extracted from a 6-well plate using the proteinase K (Sigma, #3115879001). PCRs were performed using Q5® High-Fidelity DNA Polymerase (NEB, #M0491). 100 ng of C2C12 genomic DNA was used as a template for PCR. Each 20 µl of PCR reaction consisted of 4 µl of 5 x buffer (NEB), 2 µl of 2.5 mM dNTPs, 0.5 µl of Q5 Taq DNA polymerase (NEB), 1 µl of forward primer and 1 µl of reverse primer at an initial concentration of 10 µM, with respective amounts of DNA templates and H2O to reach a reaction volume of 20 µl. The PCR conditions were set at 95 ℃, 5 min; 30 cycles of 95 ℃ for 3 min, 55 ℃ for 30 sec and 72 ℃ for 1 min; and a final extension at 72 ℃ for 5 min. PCR products were then extracted using 1.5 % agarose gel followed by isolation of DNA using a Qiagen kit (Qiagen, #12943). The purified DNA then underwent Sanger Sequencing to confirm.

2.3. Cloning of pegRNAs to Target the Mutant Sequence of Leptin Receptor in db/db Mice

Methods for cloning pegRNAs have been described previsouly^[6]. Briefly, three kinds of oligos were prepared as follows. The 1^st component oligo set is for the pegRNA spacer annealed oligos containing gRNA3 targeting the leptin receptor mutant sequence of db/db mice: 5’- caccg GGTCATTCAAACCATAGTTT gttttaga- 3’, 5’- tagctctaaaac AAACTATGGTTTGAATGACC c -3’. The 2^nd component oligo set is for the pegRNA 3’ extension annealed oligos: two kinds of pegRNAs were designed. One is pegRNA8 containing 8 nucleotides for primer binding site (PBS). The other is pegRNA10 which has 10 nucleotides for primer binding site (PBS). pegRNA8: 5’ -gtgc GGAAACAAACCTAAACTATGGTT -3’, 5’-aaaaAACCATAGTTTAGGTTTGTTTCC -3’, pegRNA10: 5’ -gtgc GGAAACAAACCTAAACTATGGTTTG -3’, 5’-aaaaCAAACCATAGTTTAGGTTTGTTTCC -3’. The 3^rd component oligo set is for the pegRNA scaffold annealed oligos: 5’- GCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCG -3’, 5’- GCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAACTTGCTATTTC -3’. For cloning of the pegRNA, 1 µl of each forward and reverse gRNA oligos at an initial concentration of 100 µM were mixed in the annealing buffer (10 mM Tris-HCl pH 8.5 and 50 mM NaCl). The oligos were annealed in a thermocycler, heated at 95 ℃ for 3 minutes, then cooled gradually (0.1℃/s) to 22℃. Then, the 1st and 2nd component oligos were diluted with annealed oligonucleotides 1:4 by adding 75 µl of H2O. To phosphorylate the 3rd component oligos 6.25 µl of 4 µM oligonucleotide duplex, 2.5 µl of 10x T4 DNA ligase buffer (NEB), 0.5 µl of T4 PNK (NEB), 15.75 µl of H2O were mixed and incubated at 37 ℃ for 60 min followed by 65 ℃ for 20 min for the heat inactivation. 2 µg of pU6-pegRNA-GG-Vector plasmid, which was a gift from David Liu) (Addgene plasmid # 132777; http://n2t.net/addgene:132777; RRID: Addgene_132777)^[6] was digested in a mixture of 1.0 µl of BsaI, 5.0 µl of 10x fast digestion buffer, and H2O to bring the volume to 50 µl and the mixture was subsequently incubated at 37℃ for 3 hours. Cut plasmids were then isolated by phenol extraction. The dephosphorylation reaction of the vector consisted of 1.5 µl of 10x Antartic phosphatse buffer, 1 µl Antartic phosphatase, and 12.5 µl of cut vectors was incubated at 37℃ for 1 hour followed by 70 ℃ for 10 minutes for heat inactivation. Next, the pegRNA assembly was performed using 1 µl of 10x ligation buffer, 1 µl of 10x PEG, 0.5 µl of T4 DNA ligase, 0.25 µl of BsaI, 1 µl at 30 ng of digested pU6-pegRNA-GG plasmid vector, 1 µl at 1 µM of 1st component oilgos, 1 µl at 1 µM of 2nd component oligos, 1 µl at 1 µM of 3rd component oligos, which was incubated in a thermocycler, cycle between 5 min at 16℃ and 5 min at 37 ℃ for 8 cycles followed by 15 min at 37 ℃, then 15 min at 80 ℃, then hold at 12 ℃. Transformation of competent E. coli cells (Promega, Bacterial Strain JM109, # P9751) with the ligation reaction was performed by the conventional calcium method. Three clones were picked for sequencing. The sequencing primer used was the human U6 promoter forward primer: 5’-GAC TAT CAT ATG CTT ACC GT-3’.

2.4. Flexor Digitorum Brevis Electroporation of pegRNA into db/db Mice

All in vivo experiments used B6.BKS(D)-Leprdb/J mice (db/db mice) (The Jackson Laboratory, Bar Harbor, ME) that were approximately 8 weeks old and weighed about 40g. All animals were housed under standard conditions and were provided access to water and standard diet. Animal experiments were carried out with the approval of the Institutional Animal Care and Use Committee (IACUC protocol #2016A00000017). All experiments were performed in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals^[15].

Exact details for flexor digitorum brevis (FDB) electroporation have been previously described ^[16]. Briefly, the constructed pegRNA plasmid and PE2-GFP plasmid (pCMV-PE2-P2A-GFP was a gift from David Liu (Addgene plasmid # 132776; http://n2t.net/addgene:132776; RRID: Addgene_132776)^[6] were purified from E. coli using the QIAGEN endotoxin free maxiprep kit (Qiagen, #12362) following the manufacturer’s instructions. The plasmids were dissolved in sterile H2O at 2 µg/µl, and 10 µg of constructed pegRNA and 30 µg of PE2-GFP were prepared by mixing in a final volume of 20 µl. The db/db mice were anesthetized and placed in the supine position. 10 µl of 1x hyaluronidase solution (8 unit) (Sigma, #H4272) was then injected into the footpad of mouse and allowed to absorb for 1 hour while the mouse was sedated. After this, the mixed plasmids were injected into the FDB followed by electroporating the muscle (20 pulses, 20 msec in duration/each, at 1Hz, at 100V/cm).

After 5 days, the mouse was then again anesthetized and the FDB muscle was isolated, and GFP positive single myofibers were collected under epifluorescence microscopy. Additionally, the ipsilateral extensor digitorum longus (EDL) muscle was harvested as a negative control. PCR was then performed by using of Q5 (NEB) Taq DNA polymerase. The primer set used as: forward primer, 5’- CAG AGA ACG GAC ACT CTT TG -3’reverse primer -3’, reverse primer, 5’- GCA GAG TCC ATG AAT ATC AAC -3’.

2.5. Protein Extraction and Western Blot Analysis

All muscle samples were washed with PBS and homogenized with RIPA lysis buffer (EMD Millipore, 20-188) containing protease inhibitors (Sigma, #P8340) and phosphatase inhibitors (Sigma, #P0044). The denatured proteins (20ug/well) were separated by 10% SDS/polyacrylamide gels and electrotransferred onto PVDF membranes (Merck Millipore, Berlin, Germany). Membranes were incubated in 5% w/v nonfat dry milk for 1 hour at room temperature, further probed with primary antibody and incubate at 4 °C with gentle shaking overnight. Then, they were washed with Tris-buffered saline with 0.1% Tween 20 detergent (TBST) and probed with secondary antibody. Immunodetection was performed by Bio-Rad imaging system using SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, #34580). Primary antibodies used were as follows: STAT3 (Cell Signaling, #9139), phospho-STAT3 (Tyr705) (Cell Signaling, #9145), both were diluted in 5% w/v BSA (Cell Signaling, #9998) at a ratio of 1:1000; GAPDH (ABclonal, #AC002) was diluted in 2.5% w/v nonfat dry milk at a ratio of 1:1000.

2.6 Statistical Analysis

Statistical analyses were conducted using GraphPad Prism 9.0 (GraphPad Software Inc., La Jolla, CA, USA). The data are presented as mean ± standard deviation, depicted by error bars for each group. Student’s t-test was employed for two-group comparisons. One-way ANOVA was utilized for comparisons involving more than two groups. The significance level was set at P < 0.05.

3. Results

3.1. The Screening of gRNA for Mouse Leptin Receptor

We tested three gRNAs to determine their efficiency to bind to the mouse leptin receptor (intron 18) in C2C12 cells. First, we determined the concentration of puromycin to kill more than 90 % of C2C12 cells and found the concentration of 10 µg/ml of puromycin to be suitable. Following transfection of each gRNA into C2C12 cells, subsequent treatment with 10 µg/ml of puromycin for 24 hours, genomic DNA was isolated from cells. PCR was performed to check the peak of gRNA target sequence (Figure1), and Sanger sequencing revealed the noise peaks around PAM sequence from each gRNA of interest. gRNA1 and gRNA2 were deemed to be insufficient matches. Ultimately, gRNA3 targeted the leptin receptor most efficiently, and therefore was chosen as the backbone of the pegRNA. According to previous studies ^[6,17], editing efficiency was dependent on primer binding site (PBS) length. We constructed pegRNAs using gRNA3 combined with reverse transcription (RT) template 15 nucleotides (nt) adding PBS with length of 8 nt (pegRNA8), 10 nt (pegRNA10) for verification.