利用CRISPR/Cas9在人类诱导多能干细胞中进行内源性蛋白标记

摘要

本文介绍了一种利用CRISPR/Cas9技术在人诱导性多能干细胞中对内源性表达的蛋白质进行荧光标记的实验方案。通过荧光激活细胞分选富集可能经过编辑的细胞，并建立单克隆细胞系。

摘要

一种用于生成表达内源性蛋白与同框融合标签的报告基因的人类诱导多能干细胞（hiPSCs）的实验方案 N端或C端荧光标签。原核生物CRISPR/Cas9系统（成簇规律间隔短回文重复序列/CRISPR相关蛋白9）可通过同源定向修复（HDR）将大型外源序列插入基因组位点。为实现目标基因敲入，本方案采用基于核糖核蛋白（RNP）的方法，使用野生型Cas9蛋白 Streptococcus pyogenes Cas9蛋白、合成的双组分向导RNA（gRNA）以及供体质粒通过电穿孔递送至细胞中。通过荧光激活细胞分选（FACS）富集表达荧光标记蛋白的潜在编辑细胞。随后建立单克隆细胞系，并可对其编辑结果进行精确分析。通过在目的基因的基因组位点引入荧光标签，可在内源性调控条件下研究融合蛋白的亚细胞定位与动态行为，这是相较于传统过表达系统的关键改进。利用人诱导多能干细胞（hiPSCs）作为基因标记的模型系统，可在二倍体、非转化细胞中研究标记蛋白。由于hiPSCs可分化为多种细胞类型，该策略提供了在多种同基因背景的细胞环境中构建并研究标记蛋白的机会。

引言

利用基因组编辑策略（尤其是CRISPR/Cas9）研究细胞过程正变得越来越便捷且具有重要价值^{1,2,3,4,5,6,7}。CRISPR/Cas9 的众多应用之一是通过同源定向修复（HDR）将GFP等大型外源序列插入特定基因组位点，从而作为基因或蛋白质产物活性的报告系统⁸。该技术可用于将荧光蛋白序列与内源性开放阅读框连接，使所生成的内源性调控融合蛋白可用于可视化目标蛋白的亚细胞定位及其动态变化^5,6,9,10,11。尽管与过表达系统相比，内源性标记蛋白具有诸多优势，但将大型序列插入人类基因组的效率较低，通常需要采用筛选或富集策略，才能获得可用于深入研究的细胞群体^5,12。

本方案描述了将编码荧光蛋白（FP）的DNA序列插入目标基因组位点的方法。该方案包括供体质粒的设计与递送，以及核糖核蛋白（RNP）复合物（野生型S. pyogenes Cas9蛋白与合成的CRISPR RNA（crRNA）和反式激活crRNA（tracrRNA）结合而成）的使用。此外，还介绍了通过荧光激活细胞分选（FACS）富集潜在编辑细胞，以及单克隆细胞系的建立过程。迄今为止，该方法已成功用于构建携带单等位基因或（罕见情况下）双等位基因绿色荧光蛋白（GFP）标签的人诱导多能干细胞（hiPSC）系，这些标签标记了代表主要细胞结构的25种蛋白质。经编辑的细胞已证实具有预期的基因插入，表达正确亚细胞定位的融合蛋白，并保持多能性及稳定的核型¹²（以及未发表数据）。该方法还被用于生成多种其他单基因和双基因（同一细胞中标记两种不同蛋白）编辑的hiPSC群体（未发表数据）。

选择来自健康供体的人iPSCs进行这些基因组编辑工作，因为与许多传统细胞系不同，它们是二倍体、核型稳定、未转化且具有增殖能力。这些特性为研究基础细胞生物学和疾病建模提供了理想的模型。此外，hiPSCs的分化潜能使得利用同基因型细胞（包括类器官、组织以及“"疾病模型"）平行研究多种发育阶段及不同谱系和细胞类型成为可能^13,14,15。尽管本实验方案是为hiPSCs（WTC细胞系）开发的，但其也可能为其他哺乳动物细胞系实验方案的建立提供参考。

方案

1. In Silico Design of crRNA and Donor Template Plasmid for FP Knock-in

1. Obtain the annotated reference sequence from NCBI¹⁶ or the UCSC Genome Browser¹⁷ (e.g., GenBank format) of the gene of interest and import it into a bioinformatics software of choice. If the host genome sequence is known to contain variants relative to the reference, include those now by adjusting the reference sequence in the bioinformatics software (see Discussion).
2. Locate the desired FP insertion site. For C-terminal tags, the sequence for the FP tag will be introduced between the last base of the last codon and the first base of the stop codon. For N-terminal tagging, the sequence for the FP tag will typically be introduced between the last base of the start codon and the first base of the next codon. In some cases, such as when the start codon is a single codon exon, or where a signal sequence exists near the protein terminus, the desired FP insertion site may be situated in a more 3ʹ position, so long as it continues to be in frame.
3. Use 50 bp on each side of the desired insertion site as the input sequence for any publicly available crRNA design tool. Once 2-4 crRNA targets are identified near the insertion site, annotate the crRNA binding sites and protospacer-adjacent motif (PAM) sequences (NGG) in the bioinformatics software. These crRNAs will be used to induce double stranded breaks (see Discussion for more guidance on crRNA design).
   NOTE: Custom crRNA sequences can be submitted for synthesis with a commercial vendor (recommended), or the sequence can be used as a starting point to design a cloning or in vitro synthesis strategy, which is beyond the scope of this protocol (see Discussion).
4. To initiate the donor template plasmid, use 1 kb of sequence upstream of the desired insertion site as the 5ʹ homology arm (this should include the start codon for N-terminal insertions), and use 1 kb of sequence downstream of the desired insertion site as the 3ʹ homology arm (this should include the stop codon for C-terminal insertions). Bases between the two homology arms are typically not omitted. Including cell-line specific variants in the homology arms will preserve these genetic variants in the resulting edited cells.
5. Between the two homology arms, insert the sequence for the FP (or other knock-in sequence) and the linker sequence (see Discussion for more guidance on linkers). For N-terminal tags, the linker sequence should be directly 3ʹ of the FP; for C-terminal tags, the linker sequence should be directly 5ʹ of the FP.
6. Disrupt crRNA binding sites in the donor template plasmid to prevent Cas9 cutting of donor sequence (see Discussion for considerations when altering crRNA binding sites). If possible, disruption of the PAM to a sequence other than NGG or NAG is preferred. Alternatively, introducing point mutations to three bases in the seed region of the crRNA (10 bases proximal to the PAM) is predicted to sufficiently disrupt crRNA binding. Some crRNA binding sites are disrupted by introduction of the FP sequence in the donor template plasmid; ensure that no PAM, or intact binding region still exists in these cases.
   NOTE: In silico donor template plasmid can be submitted for gene synthesis by a commercial vendor, or it can be used as a starting point to design a cloning strategy, which is beyond the scope of this protocol. A simple backbone such as pUC19 or pUC57 is sufficient.

2. Ribonucleoprotein (RNP) Transfection for CRISPR/Cas9 Mediated Knock-in in hiPSCs

NOTE: In this protocol, the term 'gRNA' describes synthetic crRNA and tracrRNA properly re-suspended, quantified, and pre-complexed per manufacturer's instructions (see Table of Materials). Supplement all media with 1% Penicillin Streptomycin. General culturing guidelines of the WTC hiPSC line are described in more detail at the Allen Cell Explorer^18,19. WTC hiPSCs are used in this protocol, but with proper transfection optimization, electroporation of RNP and donor template plasmid may be successfully adapted to other cell types.

1. Prepare 10 µM working stocks of gRNA and wild type S. pyogenes Cas9 protein^2,20; keep on ice. Prepare 1 µg/µL working stock of donor template plasmid; keep at room temperature (RT). Use pH 8.0 TE buffer for all dilutions.
2. Prepare a matrix-coated 6-well tissue culture plate with 5 mL of fresh growth media supplemented with 10 µM ROCK inhibitor (Ri) per well. Keep plate with media in the incubator at 37 °C and 5% CO₂ until ready to plate cells after the transfection procedure (maximum 2 h).
   NOTE: All matrix-coated plates used in this protocol are made by adding a volume of ice-cold Matrigel diluted 1:30 in cold DMEM/F12 media according to the Allen Institute for Cell Science's protocol for culturing the WTC hiPSC line¹⁹.
3. Using a gentle single-cell dissociation reagent as recommended in the Table of Materials, passage hiPSCs into single-cell suspension and count cells using an automated cell counter, or hemocytometer.
   NOTE: A detailed protocol for the WTC hiPSC line used here can be found at the Allen Cell Explorer¹⁹. Briefly, wash cells once with RT DPBS and treat with dissociation reagent for 3-5 minutes. Then triturate cells into single-cell suspension by gentle pipetting and pellet by centrifugation. Resuspend the live cell pellet in growth media supplemented with 10 µM Ri.
   1. Prepare an aliquot of 1.84 x 10⁶ cells for each experimental condition to be transfected in separate 1.5 mL tubes.
      NOTE: All volumes are calculated for a 4.5 µL total reaction volume in a 100 µL electroporation volume, times 2.3 reactions. This accounts for duplicate transfection and excess for pipetting error.
4. Prepare ribonucleoprotein (RNP) complex tubes for each experimental condition by adding 2.88 µL of 10 µM gRNA and 2.88 µL of 10 µM Cas9 to a 1.5 mL tube. Incubate at RT for a minimum of 10 min (maximum 1 h).
5. Pellet one cell aliquot (prepared in step 2.3.1) at 211 x g for 3 min at RT. Aspirate supernatant and resuspend cell pellet in 220 µL of manufacturer's electroporation buffer.
6. Add 220 µL of resuspended cells from step 2.5 into the 1.5 mL RNP complex tube prepared in step 2.4.
7. Add 4.60 µL of 1 µg/µL donor template plasmid to the 1.5 mL tube prepared in step 2.4.
8. Use the nucleofection tip and pipette to mix the tube contents 2-3 times, then transfer 100 µL of suspension to the prepared electroporation device. Avoid introducing any bubbles in the tip. Apply 1300 V for 1 pulse of 30 ms.
9. Gently transfer the suspension into the prepared 6-well plate from step 2.2 with a swirling motion. Disperse cells by gently moving the plate side-to-side and front-to-back.
10. Using a new nucleofection tip, repeat steps 2.8-2.9 with the remaining 100 µL of suspension and transfer into a second well of the prepared 6-well plate.
    1. Repeat steps 2.4 through 2.10 for each gRNA and donor template plasmid combination, including non-targeting gRNA, donor template plasmid only, and buffer only controls. Take care to change pipette and nucleofection tips to avoid cross-contamination.
11. Incubate transfected cells at 37 °C and 5% CO₂. Change the media to regular growth media (no Ri) at 24 h, and continue feeding hiPSCs every 24 h for 72-96 h, monitoring confluency. When hiPSCs reach 60-80% confluence, proceed to step 3.
    NOTE: Heavy cell death (>70%, estimated) is normal 24-48 h after transfection.

3. FACS-Enrichment of Putatively Edited hiPSCs

Note: When sorting stem cells, adapt instrument settings to promote cell survival as suggested in the Discussion. Briefly, use the largest nozzle possible (130 µm), a low flow-rate (≤ 24 µL/min), preservative-free sheath fluid (such as saline, see Table of Materials), and low sample pressure (10 psi).

1. Prior to beginning a FACS experiment, change media to growth media supplemented with 10 µM Ri and incubate cells at 37 °C and 5% CO₂ for 2-4 h to promote survival after FACS.
2. Using a gentle single-cell dissociation reagent as recommended in the Table of Materials, passage hiPSCs into single-cell suspension in growth media supplemented with 10 µM Ri¹⁹.
3. Filter hiPSC suspension through 35 µm mesh filter into polystyrene round-bottomed tubes.
4. Sort cells using forward scatter and side scatter (including height vs. width) to exclude debris and doublets. Use live, buffer only control cells to set the FP-positive gate, such that <0.1% of buffer only cells fall within the gate.
5. Sort the entire population of FP-positive cells into a 1.5-15 mL polypropylene tube containing 0.5-2 mL of RT growth media supplemented with 10 µM Ri.
   NOTE: Polypropylene reduces the potential for cell adhesion to the plastic.
6. Centrifuge collected cells at 211 x g for 3 min at RT.
7. Carefully aspirate supernatant and resuspend cell pellet in 200 µL of growth media supplemented with 10 µM Ri. Transfer up to 3,000 sorted cells to a single well of a fresh matrix-coated 96-well plate¹⁹.
   NOTE: With appropriate instrument setup, cells can also be sorted (in bulk) directly into a single well of a matrix-coated 96-well tissue culture plate containing 200 µL of growth media supplemented with 10 µM Ri at a recommended density of 1,000-3,000 cells per well for hiPSC.
8. Incubate sorted cells at 37 °C and 5% CO₂. Change the media to growth media supplemented with 5 µM Ri at 24 h. At 48 h begin feeding cells regular growth media (no Ri) every 24 h for 72-96 h, monitoring confluency. Survival after FACS is estimated to be greater than 50% if a minimum of 500 cells are seeded in one well of a 96-well plate.
   1. When hiPSCs reach 60-80% confluence and show mature morphology (smooth, well-packed colony centers), passage into a larger format plate such as a 24-well plate, then from a 24-well plate into a 6-well plate.
   2. When the hiPSCs in a 6-well plate reach 60-80% confluence and show mature morphology, expand to a 100 mm plate, re-plate for imaging, cryopreserve, or seed at clone picking density (step 4)¹⁹.

4. Generating Putatively Edited Clonal hiPSC Lines

1. Using a gentle single-cell dissociation reagent as recommended in the Table of Materials, passage hiPSCs into single-cell suspension and determine the number of cells per mL¹⁹.
2. Seed 10,000 cells of the edited population of hiPSCs onto a fresh matrix-coated 100 mm tissue culture dish using growth media supplemented with 10 µM Ri¹⁹. Change the media to growth media without Ri 24 h after seeding, and feed the hiPSCs with fresh growth media every 24 h for 5-7 days.
3. When hiPSCs have formed colonies that are visible macroscopically (approximately 500 µm) they are large enough to be isolated. Prepare a matrix-coated 96-well plate by aspirating excess matrix and adding 100 µL of growth media supplemented with 10 µM Ri per well¹⁹.
4. On a dissecting microscope use a P-200 pipette, or similar, to gently scrape and aspirate individual colonies from the plate surface. Transfer volume (~20-100 µL) containing the colony to a single well of the 96-well plate prepared in step 4.3.
   1. After all colonies have been transferred, incubate the plate in a tissue culture incubator at 37˚C and 5% CO₂. Change the media to regular growth media (no Ri) at 24 h, and continue feeding cells every 24 h for 72-96 h until colonies have approximately tripled in size (approximately 1500 µm).
      NOTE: Picking 24-96 colonies per crRNA used in the transfection is recommended. Survival of isolated clones is typically greater than 95%.
5. Using a gentle single-cell dissociation reagent as recommended in the Table of Materials, passage hiPSC clones into a new matrix-coated 96-well plate as follows¹⁹.
   1. Using an 8-channel aspirator, remove and discard media from the first column of the 96-well plate.
   2. Using a P-200 multichannel pipette, add ~200 µL of DPBS to the first column of the 96-well plate to wash the cells. Using an 8-channel aspirator, remove and discard DPBS wash from the first column of the 96-well plate.
   3. Using a P-200 multichannel pipette, add 40 µL of dissociation reagent to the first column of the 96-well plate.
   4. Repeat steps 4.5.1-4.5.3 for up to a total of six columns of the 96-well plate, changing tips to be sure to not cross-contaminate wells. Place the plate in the 37 °C incubator for 3-5 minutes from the time the dissociation reagent was added to the first column (step 4.5.3).
      NOTE: It is recommended to only passage a maximum of six columns (48 wells) at a time due to the time it takes to perform these steps. When performing this protocol for the first time, start with only passaging one or two columns at a time. Limiting the number of columns passaged at one time ensures that cells are not left in the dissociation reagent for too long, which could be harmful to hiPSCs.
   5. When the cells in the first column of the plate have begun to lift off the plate bottom, use a P-200 multichannel pipette to add 160 µL of DPBS to the first column of the 96-well plate and gently triturate the cells at the "12:00", "3:00", "6:00", and "9:00" positions of each well. Transfer the entire volume of cell suspension (200 µL) to a V-bottom 96-well plate.
   6. Repeat step 4.5.5 for the remaining columns of cells that have dissociation reagent in them; change tips as to not cross-contaminate wells.
   7. Spin the V-bottom plate in a centrifuge at 385 x g for 3 min at RT.
   8. Using a P-200 multichannel pipette, gently remove the supernatant and resuspend cells in 200 µL of fresh growth media supplemented with 10 µM Ri per well. Repeat for all wells, changing tips as to not cross-contaminate.
   9. Transfer all of the cell suspension to a fresh matrix-coated 96-well plate¹⁹. Incubate the plate in a tissue culture incubator at 37 °C and 5% CO₂. Change the media to regular growth media (no Ri) at 24 h, and continue feeding cells every 24 h for 72-96 h, until the majority of clones reach 60-80% confluence.
      NOTE: This passage helps to spread out the cells and allow for more growth over the entire area of the 96-well plate.
6. Observe clones and identify an appropriate split ratio for each individual clone in the 96-well plate (e.g., 1:10, or 1:8). Using a gentle single-cell dissociation reagent as suggested in the Table of Materials, passage hiPSC clones into a new matrix-coated 96-well plate (steps 4.5.1-4.5.8) transferring a ratio of the cell suspension appropriate for each clone. Incubate the plate in a tissue culture incubator at 37˚C and 5% CO₂. Change the media to regular growth media (no Ri) at 24 h, and continue feeding cells every 24 h for 72-96 h, until the majority of clones reach 60-80% confluence, and show mature morphology.
   NOTE: Each clone may have a different split ratio because of slightly different growth rate or survival from the previous passage, so this passage helps to normalize the number of cells per well of each clone for the freezing step to follow. Due to the varying rates of survival and growth, some clones may overgrow or fail to grow during these passaging steps.
   1. Save the remainder of the cell suspension and pellet for gDNA isolation by centrifuging cells in a V-bottom 96-well plate at 385 x g for 3 min at RT. Remove supernatant and proceed to gDNA isolation using a 96-well kit, or store plate of pelleted cells at -20˚C for up to three weeks.

5. Cryopreservation of clonal cell lines in 96-well plate format

1. Using a single-cell dissociation reagent, passage hiPSC clones as previously described (steps 4.5.1-4.5.7). Aspirate the supernatant using a P-200 multichannel pipette and re-suspend in 60 µL of growth media supplemented with 10 µM Ri. Repeat for all wells; change tips as to not cross-contaminate.
2. Transfer 30 µL of cell suspension to a non-matrix coated 96-well tissue culture plate. Then quickly add 170 µL freezing buffer (see Table of Materials) to each well, without mixing. Repeat by transferring the remaining 30 µL of suspension into a sister plate and adding 170 µL freezing buffer.
   NOTE: This process is done in duplicate sister plates so that a back-up population of cells exists after thawing one of the individual plates. Putting cryopreserved cells into only every other column of a 96-well plate allows for faster thawing (step 5.6).
3. Wrap plate with parafilm and place in a RT Styrofoam box with lid. Place the whole box in a -80 °C freezer.
4. After 24 h plates can be transferred out of the Styrofoam box and stored at -80˚C for up to four weeks.
   NOTE: While the cells are temporarily stored at -80 °C, genetic quality control assays can be performed with the gDNA harvested from cells obtained in step 4.6.1 in order to identify the clones to thaw and propagate further, as discussed in previously published work¹². Briefly, a copy number droplet digital PCR assay can be used to identify clones that contain one or two copies of GFP and no donor template plasmid backbone integration. A combination of end-point PCR assays and Sanger sequencing can then identify clones that contain a precise insert.
5. To thaw, bring the entire plate to 37 °C in a tissue culture incubator, watching carefully for the ice pellets to melt. Wells at the edge of the plate tend to thaw first.
   1. When the ice pellet of desired clone melts, gently transfer entire 200 µL to a 15 mL conical tube containing 3 mL of RT growth media supplemented with 10 µM Ri and centrifuge at 211 x g for 3 min at RT.
   2. Aspirate supernatant and resuspend pelleted cells in 1 mL of RT growth media supplemented with 10 µM Ri. Transfer to a fresh matrix-coated 24-well plate and incubate at 37˚C and 5% CO₂. Change the media to regular growth media (no Ri) at 24 h, and continue feeding cells every 24 h for 72-96 h until the clone reaches 60-80% confluency and has mature morphology¹⁹.

结果

本实验的目标是通过将mEGFP（单体增强型绿色荧光蛋白）序列插入LMNB1基因的5ʹ端（对应蛋白的N端），实现mEGFP与核纤层蛋白B1的融合表达。所选用的连接肽（氨基酸序列为SGLRSRAQAS）基于Michael Davidson荧光蛋白文库中先前的cDNA构建体²¹。由于在供体质粒的候选crRNA结合区域中，经过in silico（计算机模拟）插入mEGFP和连接肽序列后，该结合区已被破坏，因此无需引入点突变来阻断Cas9对供体序列中潜在crRNA识别与切割位点的活性（图1）。供体序列在mEGFP-连接肽序列的两侧各包含1 kb的同源臂。最终获得的2,734 bp DNA片段被克隆至pUC57载体骨架中，经测序验证后，使用无内毒素大提试剂盒纯化得到供体模板质粒。将供体模板质粒与RNP复合物共转染细胞，富集潜在编辑成功的细胞，并通过荧光显微镜观察确认mEGFP-核纤层蛋白B1融合蛋白的定位（图2）。本文仅描述crRNA1转染的结果，尽管两种crRNA序列均获得了潜在编辑的细胞群体¹²。

与阴性对照相比，阴性对照的电穿孔反应中未添加gRNA、Cas9蛋白或供体质粒（仅含缓冲液），转染LMNB1 crRNA1的细胞中mEGFP阳性细胞占0.95%，代表了可能被编辑的mEGFP-核纤层蛋白B1细胞群体（图3a）。该结果处于此前报道中使用此方法在多个基因组位点实现的敲入效率范围内¹²。

通过流式细胞术（FACS）分离出mEGFP阳性细胞，并利用活细胞显微镜观察，以确认mEGFP-核纤层蛋白B1融合蛋白定位于核膜的预期定位。FACS富集后，LMNB1 crRNA1群体中约90%的分选细胞为mEGFP阳性（由显微镜观察确定），表明在分选过程中部分mEGFP阴性细胞与GFP阳性细胞共同被纯化。该富集水平可接受，足以用于挑选96个克隆，进而进行遗传学筛选以鉴定编辑成功的克隆。通常，富集成功的判定标准为至少50%的细胞呈GFP阳性。

在非分裂细胞中，大部分分选后的细胞群体在核膜（核周区域）显示出荧光信号，而在有丝分裂期间，荧光信号则延伸至细胞质中的核纤层结构，这表明LMNB1基因座的基因组编辑正确无误。富集后的细胞群体中包含荧光信号强和弱的细胞。信号强度的差异可能反映了正确与错误编辑结果的混合，凸显了构建经遗传学验证的单克隆细胞系以用于后续研究的重要性（见讨论部分）（图3b）¹²。在获得单克隆细胞系后，经遗传学验证的细胞在显微镜实验中表现出均一的GFP荧光强度（图3c）。

图1。LMNB1基因N端GFP标记的设计策略。 GFP标签设计为插入位于5号染色体上的LMNB1基因第一个外显子的5ʹ端起始密码子（ATG）上游。5ʹ和3ʹ同源臂各长1 kb，在起始密码子（ATG）与第二个密码子之间汇合（图中仅部分显示同源臂）。设计了两个候选crRNA，用于引导Cas9在尽可能靠近目标插入位点的位置进行切割，同时确保其在基因组中的唯一性。mEGFP序列及一段氨基酸连接序列被插入至起始密码子的3ʹ端（mEGFP与连接序列未按比例绘制）。请点击此处查看该图的放大版本。

图2．生成内源性标记hiPSC单克隆细胞系的工作流程。 将转染组分（包括Cas9/crRNA/tracrRNA核糖核蛋白复合物（图示为红色Cas9蛋白、金色crRNA和紫色tracrRNA）、含有同源臂（HAs，图示为金色）的供体质粒以及荧光蛋白+连接肽序列（FP+linker，图示为绿色））通过电穿孔导入细胞。转染4天后，通过FACS富集表达荧光蛋白（FP）的潜在编辑细胞，并将所有分选后的细胞接种至96孔板的一个孔中（约1 000个细胞），扩增形成细胞群体，继续培养至获得数百万细胞的工作群体，用于后续检测，称为“富集群体”（见方案步骤3.8）。由于不同实验中同源定向修复（HDR）效率存在差异，FP阳性细胞的产量也有所不同¹²；在转染约1.6 × 10⁶个hiPS细胞后，成功的富集通常可获得约300–5 000个FP阳性细胞。初步成像研究证实了富集群体中融合蛋白的信号及其亚细胞定位。随后通过人工挑取单克隆，接种至96孔板中进行扩增和冷冻保存。进一步采用微滴数字PCR（ddPCR）及其他基于PCR的检测方法进行基因组质量控制筛选，以鉴定正确编辑的克隆，方法如前所述¹²。请点击此处查看该图的放大版本。

图 3。潜在编辑细胞群体的富集。 （A）转染后第四天 LMNB1 编辑细胞的流式细胞术图谱。纵轴显示 GFP 荧光强度，横轴显示前向散射。分选门控基于仅含缓冲液的对照组设定。由于人诱导多能干细胞（hiPSCs）对扰动敏感，未使用活/死细胞染色，而是采用非常保守的 FSC/SSC 门控策略。（B） 富集后，LMNB1 Cr1 编辑细胞群体中约 90% 的细胞显示 GFP 定位于核膜（符合预期的 LMNB1 定位）。该群体包含 GFP 荧光强度不同的细胞，以及部分 GFP 阴性细胞。比例尺为 10 微米。（C） 在获得单克隆细胞系后，细胞表现出均一的 GFP 荧光强度，仅存在部分与细胞周期相关的差异。比例尺为 20 微米。

讨论

本文介绍的在hiPSCs中生成内源性调控的荧光蛋白融合体的方法，是一种多功能且强大的基因编辑细胞系构建策略，其应用范围广泛，包括活细胞成像、多种功能研究，以及利用患者来源的hiPSC系建立"疾病在培养皿中"模型^13,14,15。尽管该方法已被用于在内源性蛋白的N端或C端引入较大的荧光蛋白标签，但其也有潜力用于引入其他标签或小的遗传改变，以模拟或纠正致病突变^22,23。对于较小的插入片段，同源臂的长度可适当缩短，但本方法所描述的基因编辑总体策略仍然适用^24,25。尽管强烈推荐使用hiPSCs以充分发挥其广泛的应用价值，但通过谨慎优化，本实验方案也可适用于其他哺乳动物细胞系的基因组编辑。

在确定用于荧光蛋白（FP）标记的目的基因时，转录本丰度估计值（来自芯片或RNA-Seq数据）是评估该基因或特定亚型是否表达的良好起点，尽管转录本水平并不总是与蛋白水平相关。本文所述的FACS富集策略最适合在目标细胞类型中至少中等表达水平的基因。该策略在筛选具有点状和/或明确定位模式的融合蛋白（如centrin、desmoplakin和paxillin）时也已取得成功，这些蛋白的信号与背景比非常低^12,19。对于表达水平较低或仅在衍生细胞类型中表达的基因，可能需要额外的筛选策略。

在人类细胞系中设计crRNA和供体质粒时，应以人类参考基因组（GRCh38）为起点。由于同一种类的不同细胞系基因组可能存在差异，而CRISPR/Cas9具有序列特异性，因此明确与参考基因组不同的细胞系特异性变异（单核苷酸多态性或插入/缺失（indels））并将其纳入设计中极为重要。这可确保crRNA与宿主基因组兼容，并使供体质粒的同源臂保留细胞系特异性的变异。建议在设计过程中将纯合变异整合到crRNA和供体质粒的同源臂中；杂合变异的整合则为可选。下文将讨论大片段敲入实验所用的具体试剂及其他关键注意事项。

Cas9 蛋白

使用Cas9蛋白的主要优势在于，将Cas9和gRNA以核糖核蛋白（RNP）复合物形式导入细胞，其核酸酶活性持续时间较短，相比之下，基于质粒的方法中Cas9和gRNA的表达可能持续数天，从而导致更高的靶上和脱靶效应^26,27。使用Cas9蛋白的另一个优势是，一旦进入细胞，该蛋白可立即发挥切割作用，而传统的Cas9 mRNA或Cas9/gRNA质粒方法则需要经过转录、翻译和蛋白质加工等过程才能产生成熟的功能蛋白^26,28。目前，野生型S. pyogenes Cas9蛋白已可从多家商业来源获得。

向导RNA

有许多公开可用的工具可用于在目标荧光蛋白（FP）插入位点附近寻找 crRNA 靶点，这些靶点在宿主基因组中预测的脱靶效应为零或极少^29,30,31,32。在特定基因位点使用不同 crRNA 靶点时，同源定向修复（HDR）的效率及修复结果的精确性存在广泛差异¹²。因此，建议每个位点测试多个 crRNA（2–4 个，最好位于目标插入位点 50 bp 范围内），以提高基因编辑实验成功的可能性。目前 gRNA 的递送方式包括：合成的双组分 crRNA 与 tracrRNA、合成的单体 gRNA（sgRNA）、in vitro 转录的 sgRNA，或通过质粒递送在细胞中由 U6 启动子驱动表达 sgRNA。本实验方案未针对高切割活性进行优化。本研究采用未经修饰的双组分 crRNA 与 tracrRNA（见材料表），旨在生成单等位基因荧光蛋白标记的细胞系，同时尽可能减少对细胞的潜在干扰。

供体模板质粒

由于供体质粒中提供的部分同源臂序列会在HDR过程中被整合到宿主基因组中，因此应针对crRNA识别位点引入点突变，以防止HDR完成后Cas9对修复位点的再次切割。通常最简单的破坏方法是突变PAM序列。由于某些非经典PAM序列仍可被野生型S. pyogenes Cas9识别，因此最好避免使用NGG、NAG或NGA³³。在突变同源臂时，应避免引入非同义突变和稀有密码子。如果无法通过同义突变改变PAM序列，可考虑在crRNA结合位点的种子区域（靠近PAM的10 bp区域）引入三个同义点突变。在5ʹ非翻译区（UTR）进行此类修改时需格外谨慎，因为这些区域可能含有重要的调控序列。在此类情况下，可参考UCSC基因组浏览器的比较基因组学轨迹等遗传保守性数据库，以指导突变位点的选择，因为对非保守碱基的修改通常比对高度保守碱基的修改更易被耐受¹⁷。有时仅插入荧光蛋白（FP）序列就足以破坏crRNA结合位点（如图1所示）；然而，仍应检查新插入的序列是否仍保留有crRNA结合位点或PAM序列。

建议在荧光蛋白（FP）与天然蛋白之间使用氨基酸连接序列，以保持融合蛋白的功能³⁴。通常可根据连接序列的特定电荷或大小来选择合适的氨基酸连接序列。如果已有研究较为深入的cDNA融合设计，且其结构类似于目标内源性融合蛋白，则可采用相同的连接序列用于CRISPR/Cas9基因敲入实验^12,19。若缺乏此类信息，也可成功使用如GTSGGS等较短的连接序列¹²。其他研究还表明，对于多种不同靶标，通用的三氨基酸小分子连接序列同样可实现有效融合³⁵。

转染与FACS富集

许多市售转染试剂专为向细胞递送特定类型的分子而设计，而电穿孔系统则可用于递送在大小、电荷和组成上具有广泛差异的试剂。除了是转染难度较高的细胞（如人诱导多能干细胞，hiPSCs）常用的转染方法外，电穿孔还可同时递送本方法中所述CRISPR/Cas9介导的定点插入（FP knock-in）所需的全部三种组分。在本方法开发过程中，与其他市售试剂相比，电穿孔获得了最佳效果（数据未显示），且已有其他研究将其用于核糖核蛋白（RNP）的递送^26,28,36。

在对hiPSCs进行基因编辑操作时，应特别注意在基因编辑前后轻柔处理细胞，以确保细胞存活率最佳，并最大限度减少自发性分化。具体而言，进行FACS富集时应针对干细胞分选进行优化：使用尽可能大的喷嘴（130 µm）、低流速（≤24 µL/min）、无防腐剂的鞘液（如生理盐水，参见材料表）以及低样本压力（10 psi）。为避免单细胞分选导致干细胞存活率不理想，建议将FACS富集后的hiPSCs以群体形式分选并扩增培养，以优化细胞存活率和干细胞完整性。然而，对于敏感性较低的细胞类型，单细胞分选可能是合适的。为促进细胞存活，细胞在用于FACS富集而收获后，应在一小时内重新接种回培养体系，并在整个分选过程中保持室温。对于某些细胞类型，在分选过程中将细胞置于冰上（4°C）孵育也可能有助于提高细胞存活率。

大量扩增FP阳性细胞可为在建立单克隆细胞系之前通过成像分析评估融合蛋白定位提供机会。尽管所获得的富集细胞群体可能已满足某些研究的需求，但这些群体通常表现出强度不一的FP信号。分离得到的单克隆细胞系则具有均一的信号（图3），因此更适用于功能实验¹²。

克隆细胞系的建立

在整个编辑和克隆系生成过程中，监测细胞形态非常重要。在无饲养层条件下培养的hiPSC集落应具有光滑的边缘以及均匀、紧密排列的中心^12,18,19。分化细胞的比例应低于培养物的5%。在挑选单个集落时，应选择形态良好的集落。在96孔板传代过程中，需检查各克隆的形态，及时终止那些过度生长的克隆，因为这可能导致细胞分化或提示存在遗传不稳定性。

获得单克隆细胞系可用于精确编辑的遗传学验证，这一点非常重要，因为尽管在目标位点成功插入了标签，Cas9诱导的基因组双链断裂仍常常被非精确修复。先前报道的基于PCR的检测方法显示，在十个不同的基因组位点中，许多表达荧光蛋白（FP）的克隆（45%）均存在供体质粒骨架在靶向位点处的整合，或（较少见地）在基因组中随机整合¹²。此外，在十个不同位点中，23%的GFP阳性克隆（n=177）被发现其未标记等位基因的预期crRNA切割位点附近存在突变，这极有可能是由于NHEJ修复机制所致¹²。对大量单克隆细胞系（约100个克隆/编辑）的遗传学分析凸显了进行遗传学验证的重要性，而这种验证在混合细胞群体中无法实现，因为仅凭荧光蛋白表达和预期融合蛋白的定位并不能保证编辑的精确性¹²。此外，这些基于PCR的检测方法无法在富集的细胞群体中可靠地进行，因此在开展有意义的分析之前，必须首先建立单克隆细胞系。为了确保在靶向位点的精确编辑，不仅需要对插入的荧光蛋白标签进行遗传学确认，还需验证未编辑等位基因的遗传完整性（在单等位基因编辑的克隆中）。

迄今为止，使用该方法已观察到较低的双等位基因编辑率，并且未发现脱靶突变（通过Sanger测序和外显子组测序检测）（未发表数据）¹²。这与先前研究中描述的在CRISPR/Cas9实验中使用短暂存在的RNP的结果一致^26,27。缺乏具有双等位基因编辑的克隆细胞系也可能具有位点特异性，或由于细胞无法耐受两个拷贝的标记必需蛋白所致；此前已发表的实验表明，对于某一基因位点（LMNB1）可观察到推测的双等位基因编辑细胞，而在另一基因位点（TUBA1B）则未能观察到¹²。利用该方法已成功生成经过完全验证的双等位基因编辑克隆细胞系，用于将ST6 β-半乳糖苷α-2,6-唾液酸转移酶1（ST6GAL1）和RAS癌基因家族成员RAB5A（RAB5A）与mEGFP进行标记¹⁹。

除了确认基因组编辑的精确性之外，还可采用多种质量控制检测方法，进一步表征单克隆细胞系，并筛选出符合未来研究所需的所有干细胞特性、基因组特征及细胞生物学标准的克隆。可通过细胞生物学和功能实验，验证融合蛋白的表达水平、亚细胞定位及功能是否正常¹²。与未经编辑的亲本对照细胞进行比较，有助于评估编辑过程对蛋白定位、动态行为及功能的影响。其他检测方法，如细胞生长分析和基因组稳定性检测，也有助于判断标记蛋白是否对细胞产生干扰。在本方案中使用人诱导多能干细胞（hiPSC）时，对多能性标志物的检测以及分化潜能的评估，对于筛选适用于后续研究的优质克隆至关重要¹²。由于长期培养hiPSC已被证实可能导致遗传不稳定性，因此监测单克隆细胞系的生长速率和核型同样重要^12,37。然而，编辑后细胞的最终应用目的将决定所需质量控制分析的深度和广度，并因具体应用而异。

材料

本文使用的材料清单

姓名	公司	目录编号	评论
Geneious R9	Biomatters，或类似产品		用于in silico供体质粒设计的生物信息学软件
TE缓冲液 pH 8.0	IDT，或类似产品	11-01-02-05
HERAcell VIOS 160i CO2培养箱，或类似设备	ThermoFisher Scientific，或类似产品	51030408
移液器（1000 µL、200 µL、20 µL、10 µL、2 µL）	Rainin，或类似产品
移液器吸头（1000 µL、200 µL、20 µL）	Rainin，或类似产品
多通道移液器（200 µL）	Rainin，或类似产品
血清移液管（25 mL、10 mL、5 mL）	Costar，或类似产品
BRAND 8通道手动分液器（Quiksip型，可高压灭菌）	Millipore Sigma，或类似产品	BR704526-1EA	需与非滤芯吸头配合使用，例如下方列出的Molecular BioProducts低吸附移液吸头Pure 10
Molecular BioProducts低吸附移液吸头，Pure 10	Thermo Fisher，或类似产品	3501-05
XP2移液控制器	Drummond，或类似产品	4-000-501
一次性巴斯德吸管	VWR，或类似产品	53300-567
II级A2型生物安全柜	CELLGARD，或类似产品	NU-481
Matrigel基质（生长因子减少型）	Corning	354230	使用前需对hiPSC进行批次检测
DMEM/F12（无酚红）	Gibco	11039-021	冷藏，用于以1:30比例稀释Matrigel
mTeSR1完全培养基	StemCell Technologies	85850	推荐用于WTC hiPSC细胞系的生长培养基
青霉素-链霉素	Gibco	15070-063
WTC hiPSC细胞系	Coriell	GM25256	本实验方案中使用的hiPSC细胞系可通过Coriell获取
组织培养皿（100 mm）	Falcon	353003
6孔细胞培养板	CELLSTAR	657160
StemPro Accutase	Gibco	A11105-01
杜氏磷酸盐缓冲液（DPBS，无钙、无镁）	Gibco	14190144
15 mL聚苯乙烯锥形离心管	Sarstedt	62.554.100
Y-27632（ROCK抑制剂）	StemCell Technologies	72308
Edit-R CRISPR-Cas9合成crRNA（未修饰，定制序列）	Dharmacon	Custom0247
Edit-R CRISPR-Cas9合成tracrRNA	Dharmacon	U-002005-05
重组野生型 Streptococcus pyogenes Cas9-NLS纯化蛋白，40 µM	加州大学伯克利分校QB3 Macrolab
定制供体质粒（PriorityGENE）	Genewiz		供体插入序列由Genewiz合成并克隆至pUC57骨架中
DNA LoBind管（1.5 mL）	Eppendorf	22431021
NucleoBond Xtra Maxi EF	Clontech	740424.50
Neon转染系统	ThermoFisher Scientific	MPK5000
Neon转染系统100 µL试剂盒	ThermoFisher Scientific	MPK10096
5 mL聚苯乙烯圆底试管（带细胞筛网盖）	Falcon	352235
15 mL高透明度聚丙烯锥形离心管	Falcon	352196
FACSAriaIII Fusion流式细胞仪	BD Biosciences	656700
FACSDiva软件	BD Biosciences
FlowJo 10.2版	TreeStar
NERL血库生理盐水	ThermoFisher Scientific	8504	用作无防腐剂的FACS缓冲液
Olympus SZX7体视显微镜，或类似设备	Olympus，或类似产品
组织培养板（96孔）	Falcon	353072
96孔细胞培养板（V型底）	CELLSTAR	351180
CryoStor CS10	Sigma	C2874-100ML	用作96孔板格式细胞的冷冻保存缓冲液
Parafilm封口膜	Bemis	PM-996
24孔细胞培养板	CELLSTAR	662160