Organoid Easytag: an efficient workflow for gene targeting in human organoids

Human organoid systems recapitulate key features of organs offering platforms for modelling human developmental biology and disease. However, tissuederived organoids suffer from low efficiency of genetic manipulations, especially CRISPR/Cas9-mediated knock-ins. We have systematically optimised and developed an “Organoid Easytag” pipeline for efficient (40-65%) and accurate gene targeting to facilitate generation of reporter lines and gene knock-outs in organoid based research.

In order to achieve efficient gene targeting, we first sought to maximise: (1) the efficiency of DNA delivery into organoid cells, which impacts repair template presentation; (2) the efficiency of site-specific DNA cleavage by the Cas9-gRNA complex, which influences the likelihood of homology directed repair (HDR) being triggered. Commonly used methods for mammalian cell DNA delivery, including lipofectamine, nucleofection and lentivirus, were tested.
Nucleofection consistently achieved up to 70% transfection efficiency across different organoid lines ( Supplementary Fig. 1a, 1b). To optimize site-specific  Fig. 1c, 1d). Remarkably, the ssRNP generated almost twice the amount of indels compared to the cr/tr RNP. Thus, we adopted nucleofection and ssRNP for downstream experiments. This strategy has the advantage that the RNP is rapidly degraded and should produce minimal off-target effects.
To optimize our workflow (Fig. 1a), we first focused on generating an ACTBfusion protein, taking advantage of the abundance of ACTB protein in human foetal lung organoids and a previously-published ACTB targeting strategy 7 .
Having an efficient gRNA for the ACTB locus ( Supplementary Fig. 1c), we designed a repair template to generate an N terminal monomeric (m)EGFP-ACTB fusion (Fig. 1b). We set the following rules for repair template design to facilitate efficient and consistent gene targeting: (1) protospacer adjacent motif (PAM) sequence mutated to prevent editing by ssRNP 8 ; (2) 700 to 1000 nucleotide length of each homologous arm 9 ; (3) minimal plasmid size to maximise delivery into organoid cells; (4) monomeric forms of fluorescent protein to avoid undesirable fusion protein aggregates. As expected, 72 hours after nucleofection of the ssRNP and repair template, mEGFP + organoid cells could be enriched by flow cytometry (Fig. 1c). These cells were collected and pooled together, but seeded sparsely, and were successfully expanded into organoid colonies (Fig. 1d). The mEGFP-ACTB fusion protein localized to cell-cell junctions, consistent with previous reports 7 . These small colonies could be further expanded into new organoid lines and 59% of lines (n = 17/29 lines, from N = 2 parental organoid lines) were correctly targeted. Targeted organoids continued to express the multipotent lung progenitor marker, SOX9 ( Fig. 1e).
We sought to further increase targeting efficiency ( Supplementary Fig. 2a).
Previous research reported that various drugs including a RAD51 agonist (RS-1), a β3-adrenergic receptor agonist (L755507) and a DNA ligase IV inhibitor (SCR-7) enhance gene targeting efficiency [10][11][12][13] . However, using flow cytometry as a simple assay, none of the drugs tested increased the rate of gene targeting in the organoids ( Supplementary Fig. 2b).
Synthetic single-stranded donor oligonucleotide (ssODN) repair templates offer a 'cloning free' workflow for targeting short peptide tags into different loci 14-16 , although they are limited by oligonucleotide length. We explored ssODN performance in our Easytag workflow using the split GFP system, by tagging the N terminal of ACTB with GFP11 ( Supplementary Fig. 3a). We provided GFP11-ACTB ssODN, or a positive-control plasmid, repair template together with a small transient GFP(1-10) expressing vector. Targeted cells would become GFP + in the presence of the GFP(1-10) vector. We obtained GFP + cells by flow cytometry in both the plasmid and ssODN repair template groups ( Supplementary Fig. 3b). However, when ssODN was used no organoid colonies formed after 10 days of culture, whereas the plasmid group started to show numerous colonies ( Supplementary Fig. 3c). We reasoned that ssODN could be error prone and generate indels in the ACTB locus, which would be detrimental to cells. We tested this hypothesis by tagging SOX2, a protein that cells are less sensitive to, with a V5 tag using an ssODN repair template ( Supplementary Fig. 3d). In this experiment, we obtained V5 + colonies with low efficiency (1/14 colonies with V5 integration in SOX2) and the one tagged allele obtained had a random insertion near the gRNA cutting site ( Supplementary Fig.   3e, 3e'). We conclude that synthetic ssODN templates are error prone and not optimal for the organoid Easytag workflow.
To expand our pipeline to target other loci, we targeted SOX9, a transcription factor, to represent genes expressed in a less abundant manner. SOX9 is a tip progenitor cell marker for developing lungs 2 . Thus, incorrect targeting may have a detrimental impact and SOX9 reporters are useful for monitoring progenitor state in human foetal lung organoids. In order to overcome the low expression level of SOX9, we used a Histone H2B-EGFP fusion (H2B-EGFP hereafter) to concentrate the EGFP signal in nucleus (Fig. 2a). A T2A sequence, a self-cleavage peptide, was also inserted between SOX9 and H2B-EGFP, to ensure that SOX9 protein was minimally influenced. This strategy allowed us to enrich correctly targeted cells. Colonies could be expanded and maintained normal SOX2, SOX9 and NKX2.1 expression (Fig. 2b, Supplementary Fig. 4a).
Importantly, we noted that although we were only able to generate SOX9 reporter lines as heterozygotes ( Supplementary Fig 4b, 4c), the gRNA sites in the wildtype alleles were intact (6/6 lines tested, N=3 parental organoid lines) ( Supplementary Fig. 4d). This offers the opportunity of retargeting the second allele if desired.
The AAVS1 locus has been generally considered to be 'safe harbour locus' for expressing exogenous genes in a controllable manner in human cells without silencing 17 . We sought to target the AAVS1 locus in organoids. As a proof of concept, we targeted a membrane tagged TagRFP-T (mTagRFP-T) to visualise cell shapes under the control of an EF1α promoter to the AAVS1 locus (Fig. 2c).
Generation of straightforward gene knockouts using the CRISPR-Cas9 system can suffer from translation retention and exon skipping 18,19 . Moreover, in the absence of a strong, immediate phenotype the knockout cells cannot readily be identified. We sought to solve these problems by generating a gene knockout in a more controlled manner using the Organoid Easytag workflow. We focused on the SOX2 gene as its function remains to be elucidated in human foetal lung progenitors. We swapped the SOX2 exon with T2A-H2B-EGFP to generate SOX2 knockout organoids. Using two gRNAs targeting the N and C terminal of the SOX2 coding sequence respectively, we sequentially replaced both copies of the SOX2 coding sequence (CDS) (Figure 2f, 2g; Supplementary Fig. 5). This again illustrated the power of re-targeting the second allele using the Organoid Easytag workflow. SOX2 knockout colonies can proliferate and grow normally, suggesting that SOX2 is not crucial for human foetal lung tip progenitor cell self-renewal. Thus, we have generalised our Organoid Easytag pipeline to target various loci, including highly abundant genes, transcription factors, the human safe harbour locus, and to generate knockouts.
The Organoid Easytag workflow provides a robust and versatile pipeline to perform gene targeting in the human foetal lung organoid system. Heterozygous knock-ins can be efficiently generated in organoids to produce reporter lines.
Moreover, the WT allele typically remains intact, providing the option of targeting the second allele which enables generation of knockouts. We envision that the Organoid Easytag workflow can be easily adapted for gene targeting in other systems, including organoids from other tissues, human pluripotent stem cells, cancer cell lines and cancer organoids.

Derivation and maintenance of human foetal lung organoid culture
Human foetal lung organoids were derived and maintained as previously

Lentiviral production
We grew HEK293T cells in 10-cm dishes to a confluence of 80% before

Lentivirus infection of organoids
Organoid single cell suspension was prepared as for nucleofection. 5 µl lentivirus (CMV-myrAKT-IRES-GFP) suspension was applied to 2 × 10 5 organoid single cells suspended in 500 µl self-renewing medium with 10 µM ROCKi (without Matrigel) in one well of 24-well plate and incubated at 37 °C overnight. The following day, cells were harvested and washed twice with PBS before pelleting and seeding, in Matrigel, in two wells of 24-well plate. Cells were grown in self-renewing medium with ROCKi (10 µM) for 72 hrs before flow cytometry. CMV-myrAKT-IRES-GFP was used for checking lentiviral transduction efficiency.

Lipofectamine transfection of organoids
Organoid single cells were prepared the same way as for nucleofection. For

T7 Endonuclease Assay
To test for site specific DNA cleavage using the T7 endonuclease assay, organoid cells were harvested 48 hrs after nucleofection of ssRNP, tr/cr RNP, plasmid encoding Cas9 and gRNA or WT control organoids. Genomic DNA was extracted using QIAamp Fast DNA Tissue Kit (51404, Qiagen). PCR was performed using PrimeSTAR® GXL DNA Polymerase (R050A, Takara) with 20 ng of genomic DNA as template according to manufacturer's protocol.

ICE analysis for Indel production
Genomic DNA was extracted from organoid cells were harvested 48 hrs after nucleofection of ssRNP, tr/cr RNP, plasmid encoding Cas9 and gRNA or WT control organoids using QIAamp Fast DNA Tissue Kit (51404, Qiagen). PCR was performed using PrimeSTAR® GXL DNA Polymerase (R050A, Takara) with 20 ng of genomic DNA as template according to manufacturer's protocol.