CRISPR-Cas9 strategies to insert MS2 stem-loops into endogenous loci in Drosophila embryos

Summary CRISPR-Cas9 genome editing has transformed biology by enabling site-specific genome modifications to be simply engineered. Here, we describe two CRISPR-Cas9 approaches to introduce MS2 stem-loop sequences into endogenous gene loci in Drosophila. This can facilitate live imaging of nascent transcription in Drosophila. For complete details on the use and execution of this protocol, please refer to Hoppe et al. (2020).

HIGHLIGHTS CRISPR-Cas9 editing allows insertion of MS2 sequences into the Drosophila genome A one-step approach enables rapid insertion of MS2 loops or other sequences MS2 insertion may show reduced viability. Insertion of stem-loops into a large intron is not predicted to greatly affect splicing, although if splicing is efficient the loops will be very short lived decreasing the fluorescent signal.
Note: For a protocol outlining how to image and quantitate transcription in embryos based on the MS2 system, please refer to . Hoppe and Ashe (2021) CRISPR design outline Timing: one-step CRISPR 10 weeks total Timing: two-step CRISPR 10 weeks for CRISPR-Cas9 mediated targeting of an endogenous locus and 11 weeks total to reintegrate MS2 stem-loops into the endogenous locus We outline two different CRISPR approaches, a one-step method using one cleavage site and a twostep approach that uses two DNA cleavage sites ( Figure 2). While MS2 sequences are used as an example here, other sequences can be inserted using the methods described.
In both approaches, CRISPR-Cas9-mediated HDR is performed using a dsDNA donor plasmid. The donor contains a gene encoding a visible transformation marker (for example DsRed or mini-white) flanked by loxP sites for Cre-recombinase-mediated removal and a multiple cloning site (MCS) for homology arm (HA) insertion ( Figures 3A and 3B). The donor plasmid for the twostep CRISPR approach additionally contains an attP FC31 phage recombination site ( Figure 3B), which can be used to reintegrate specific sequences including MS2 stem-loops in the second step.
In short, the one-step CRISPR strategy uses a single cut site to introduce MS2 stem-loop sequences directly adjacent to the cleavage site into the endogenous locus together with a selectable marker that is subsequently removed ( Figure 3A) (Fukaya, 2020;Gratz et al., 2015;Lim et al., 2018bLim et al., , 2018a. The two-step CRISPR approach uses two cut sites to first insert an attP FC31 phage recombination site into the gene region of interest. In the second step, the attP FC31 recombination site is used to integrate MS2 stem-loops and the DNA sequences that were initially removed in step 1 ( Figure 3B) (Baena-Lopez et al., 2013;Hoppe et al., 2020).
Note: If the protocols described here are to be used to insert a protein tag, the exact position of the tag can be chosen by removing a region from the genome using two gRNAs, whereas a single cut site will place the tag sequences directly adjacent to the cut site and therefore be limited by gRNA availability. The chances for off-target effects are increased when using two gRNAs versus one.
Note: Different CRISPR-Cas9 approaches haven been documented in Drosophila and mainly differ in the delivery of gRNAs and Cas9. These components can be injected in the form of expression plasmids or in vitro transcribed RNA together with the donor plasmid (Bassett et al., 2013;Gratz et al., 2013;Yu et al., 2013). Expressing Cas9 (Gratz et al., 2014;Ren et al., 2013) or Cas9 and gRNAs (Kondo et al., 2013;Port et al., 2014) stably as a transgene(s) was shown to result in increased CRISPR efficiency and consistency. The efficiency of different Cas9 transgenes and gRNA expression systems have been compared by Port et al. (2014Port et al. ( , 2015. A summary of CRISPR-Cas9 approaches, such as generation of loss-of-function alleles or complex gene modifications and their experimental considerations can be found in Gratz et al. (2015). Experimental designs to insert tags proximity to the target region. The TTAA site is targeted by a PiggyBac transposon to introduce a screenable marker (http://flycrispr.molbio.wisc.edu/scarless).

Target site identification
Timing: 2 h Figure 3. Overview of CRISPR-Cas9 genome engineering approaches (A) The one-step CRISPR approach uses a single site for gRNA directed, Cas9 mediated cleavage. A dsDNA donor plasmid is used for HDR and inserts 24xMS2 stem-loops into the endogenous gene locus. The donor plasmid also contains loxP sites and a marker gene. (B) For the two-step CRISPR approach, the genome is cut at two cleavage sites by gRNA directed, Cas9 mediated cleavage. The dsDNA donor plasmid contains loxP sites, an attP recombination site and a marker gene (1). In the second step, the previously removed DNA sequence together with MS2 stem-loops is reintegrated into the genome using attP/B recombination and the combined marker cassette is subsequently removed by Cre recombination between the two outermost LoxP sites (2). This section describes the identification of DNA cleavage sites and the design of gRNAs. gRNAs are generated to direct the Cas9 endonuclease to genomic regions of interest to introduce doublestranded DNA breaks. The gRNA is directed to the genomic region through base pairing with the 20 nucleotide (nt) genomic target sequence ( Figures 3A and 3B).
One-step CRISPR 1. Identify potential target sequences for DNA cleavage within the genomic region of interest by using the publicly available flyCRISPR target finder (https://flycrispr.org/target-finder/). a. Enter the genomic DNA sequence to find CRISPR target sites and, where appropriate, select the reference genome based on the Drosophila Cas9 line that will be used. b. Optimal target sequences are 20 nt in length and located next to a 3 nt protospacer adjacent motif (PAM), which is required for DNA cleavage (Jinek et al., 2012). c. Select CRISPR targets with a 5 0 G for optimized U6-driven transcription or add a 5 0 G during the oligonucleotide design. d. On the next page select high stringency. To reduce the probability for off-target effects the option ''NGG PAM sites only'' can be selected. e. The website suggests cleavage target sites and highlights possible off-target sites. Carefully choose the target site best suited for the experiment (see Notes). 2. Select ''design experiment'' to view oligonucleotide sequences for the selected target. Order two oligonucleotides suggested by the website to generate one gRNA. a. Oligonucleotides contain CTTC 5 0 and CAAA 3 0 overhangs respectively, which are compatible with overhangs obtained in the pU6-BbsI-gRNA plasmid after BbsI digestion. b. Oligonucleotides should be ordered with the 5 0 end phosphorylated or should be phosphorylated using T4 polynucleotide kinase before use.

Two-step CRISPR
3. Identify one target sequence for DNA cleavage on either side of the genomic region of interest and follow the steps outlined above.
Note: If the target site search returned no results, consider lower stringency, shorter or alternative target sequences or selecting targets with possible off-target sites (see troubleshooting 1).
Note: The DNA cleavage site is located 3 nt away from the PAM sequence within the target sequence ( Figures 3A and 3B).
Note: Some genomic regions exhibit reduced CRISPR efficiency. The reasons for locus-specific effects on CRISPR efficiency are poorly understood. The cleavage efficiency for Drosophila regions can be predicted using an online tool developed by the Perrimon group (Harvard Medical School). This prediction tool uses data from high throughput experiments in S2 cells (https://www.flyrnai.org/evaluateCrispr/).
CRITICAL: When selecting gRNAs, avoid those recognizing sequences overlapping or close to known regulatory motifs in the DNA or mRNA such as enhancers, binding sites for transcription factors or RNA binding proteins, or splice sites. The integration of DNA sequences at these sites may disrupt their regulatory function, as the methods described leave a small scar in the genome.
CRITICAL: Single nucleotide polymorphisms (SNPs) can occur naturally and could lower CRISPR efficiency when present in the targeting sequence. To verify that no SNP is present in the target sequences or PAM sites, sequence the genomic region in the lab stock of the ll OPEN ACCESS fly line that will be edited before cloning gRNAs or check the genome sequence if available.

Homology arm design
Timing: 2 h This section describes the design of HAs that contain upstream/downstream genomic regions bordering the target cleavage site. HAs will be inserted into a dsDNA donor plasmid for HDR (see description in the step-by-step method details section).
CRITICAL: Check the sequence of the genomic regions around the target sites in the fly line chosen for editing before cloning, either using the genome sequence if available or by sequencing the specific regions.
Note: A homology length of $1 kb was found to be efficient for integration. The efficiency of different homology length (100 bp to 7.5 kb) in dsDNA donors was investigated in detail by Beumer et al. (2013) and Kanca et al. (2019).

One-step CRISPR
4. Design the HAs to be approximately 1 kb in length and directly adjacent to the genomic cleavage site for efficient HDR. 5. Design primers to amplify HAs from genomic DNA.
a. To the primer ends add restriction enzyme target sites that are present in the multiple cloning site (MCS) in the donor plasmid but absent from the HAs (for details see Figure 5).
Note: Under certain circumstances the PAM sequence needs to be mutated, for example if a one-step CRISPR approach with two guide RNAs is chosen, as if the endogenous genomic DNA region is inserted directly next to the HA in the donor plasmid, the full guide and PAM sequences can be reformed. To prevent targeting of the donor plasmid by Cas9 and repeated targeting of the endogenous locus, a point mutation is necessary to mutate the ''NGG'' PAM (Gratz et al., 2015). In the example described here using one gRNA, the full target sequence will be disrupted by the MS2 stem-loop insertion at the cut site and therefore, no point mutation is necessary.

Two-step CRISPR
6. Design HAs to be approximately 1 kb in length. If gRNA/PAM sequences are in the PAM-in configuration as highlighted in this example ( Figure 3B), HAs should not contain the 6 bps between the cleavage site and the end of the PAM sequence. Inclusion of the full target sequence will make the donor plasmid a target for Cas9 cleavage (Gratz et al., 2015). 7. Design primers to amplify the HAs from genomic DNA.
a. Add restriction enzyme target sites, which are present in the donor plasmid multi-cloning sites (MCS) but absent from the HAs, to the primer ends (for details see Figure 6).
Alternatives: If the restriction sites to be used are present in the HA sequences, use In-Fusion or Gibson cloning, or commercially synthesize the plasmid DNA. The second step of the two-step CRISPR engineering approach uses fC31 integrase-mediated, sitespecific transgenesis to introduce the MS2 stem-loops and the genomic region, which was removed as part of step 1, back into the genome ( Figure 3B). This section outlines the primer design to generate inserts and the cloning procedure will be outlined in the step-by-step method details section.

Two-step CRISPR
8. Design primers to amplify the genomic region, which was removed in CRISPR step 1. In this example, the genomic DNA region will be inserted upstream of MS2 stem-loops into the RIV cherry +24xMS2-stem-loop plasmid (available upon request) using EcoRI and NotI restriction sites respectively (for details see Figure 9). Use different restriction sites, if these sites are present in the genomic region used for reintegration. 9. Order PCR primers to amplify the genomic region, containing sequences to insert an EcoRI (forward primer) and NotI (reverse primer) restriction site.
Alternatives: Instead of a RIV cherry backbone, the RIV white plasmid can be used which contains a mini-white marker gene (Baena-Lopez et al., 2013). In order to use this plasmid backbone, the mini-white marker, inserted by the HDR donor plasmid, needs to be removed by Cremediated recombination before reintegration or a donor plasmid with a different marker gene needs to be used for HDR.
Alternatives: Instead of using classical cloning techniques to insert DNA sequences using restriction sites, other methods such as In-Fusion or Gibson cloning methods can be used to reduce the insertion of ectopic nucleotides into the endogenous genome. Alternatively, plasmids can be commercially synthesized that lack any additional nucleotides that would be inserted into the genome. Scarless CRISPR methods have also been described (Kanca et al., 2019;Lamb et al., 2017) but are generally more complicated, whereas the addition of a scar here is not viewed as particularly problematic given that additional MS2 sequences are being inserted. This section describes how to insert gRNA oligonucleotides into the pU6-BbsI-gRNA (Addgene, Cat# 45946) plasmid, which will be used to deliver gRNAs ( Figure 4A). The gRNA target sequences start with a 5 0 G, which allows for efficient U6 promoter-driven expression. Oligonucleotide sequences were determined in the before you begin section. This protocol was established by (Gratz et al., 2015) and is part of a cloning protocol selection by the FlyCRISPR website https://flycrispr.org/ protocols/grna/.

KEY RESOURCES
One-step CRISPR and two-step CRISPR 1. Day 1. Reconstitute and anneal the designed phosphorylated oligonucleotides ( Figure 4B). a. Set up the following annealing reaction for each oligonucleotide pair:

REAGENT or RESOURCE SOURCE IDENTIFIER
b. Anneal oligonucleotides in a thermocycler at 95 C for 5 min, then ramp the temperature down to 25 C at a rate of À0.1 C/s ( Figure 4B, step 1).
Pause point: The plasmids can be stored at À20 C for years.

Introduction of homology arms into the dsDNA donor plasmid
Timing: $11 days This section outlines how to generate a dsDNA donor plasmid that is used for HDR after Cas9 mediated cleavage.

Reagent Amount Final concentration
Oligonucleotide forward (100 mM) 1 mL 1 0 mM Oligonucleotide reverse (100 mM) 1 mL 1 0 mM T4 103 ligation buffer 1 mL 1 3 For the one-step CRISPR approach, HAs are located directly upstream and downstream of the cut site where the MS2 loops will be inserted ( Figure 5A). The donor plasmid contains a pBlueScript backbone, a DsRed maker gene placed between two loxP sites, 24xMS2 stem-loops and MCSs for HA insertion ( Figure 5B) (Lim et al., 2018a). Other sequences can be inserted instead of 24xMS2 stem-loops.
For the two-step CRISPR approach, we use the pTV Cherry plasmid (DGRC, Cat# 1338), which contains an attP reintegration site, a mini-white marker gene, loxP sites and is flanked by MCSs into which HA sequences are introduced ( Figure 6A) (Baena-Lopez et al., 2013). The HAs are positioned upstream of the 5 0 cut site ( Figure 6B) and downstream of the 3 0 cut site.
Alternatives: Plasmids can be synthesized commercially.
Note: The cloning to insert HAs into the dsDNA donor plasmid can be performed concurrently with the gRNA cloning to shorten the time needed to prepare plasmids for microinjection.
One-step CRISPR 5. Day 1. PCR amplify HA1 (5 0 ) and HA2 (3 0 ) from genomic DNA from the fly line that will be used for injection and digest with restriction enzymes ( Figure 5C). Use primers that were designed in the before you begin section. a. Use a high-fidelity DNA polymerase, e.g., Q5 (NEB, Cat# M0491) or Phusion (Thermo Fisher Scientific, Cat# F530) for PCR ( Figure 5C, step 1). b. Digest the purified PCR inserts, with the restriction enzymes whose recognition sequences were added to the HA primers, for insertion into the donor plasmid. (C) HA1 is PCR amplified from genomic DNA (1). The plasmid backbone is linearized using restriction sites in the 5 0 MCS (2) and ligated with the HA1 fragment (3). The resulting pDsRed-24MS2+HA1 plasmid is linearized by restriction digest (4) and the HA2 fragment is amplified from genomic DNA (5). HA2 is ligated into the linearized vector (6) and purified for microinjection (7). (D) The finished plasmid pDsRed-24MS2+HA1+HA2 is used as a dsDNA donor for CRISPR engineering.

OPEN ACCESS
6. Insert HA1 into the linearized pBS-24xMS2-loxP-DsRed-loxP (hereafter referred to as pDsRed-24MS2) backbone plasmid ( Figure 5C). a. Digest the donor plasmid backbone with the same restriction enzymes as the HA1 insert (Figure 5C, step 2). b. Dephosphorylate the vector with Calf Intestinal Alkaline Phosphatase. c. Gel purify the linearized vector. Figure 6. Cloning strategy to insert HAs into the pTV cherry plasmid for two-step CRISPR (A) The pTV cherry plasmid contains two MCSs to insert HA sequences for HDR, a mini-white marker gene, loxP and FRT sites, and an attP site. It also contains a mCherry gene (not shown). (B) Detailed view of the target sequence, starting with a 5 0 G and ending with the NGG PAM site. The Cas9 cleavage site is located 3 nt away from the PAM sequence and the 5 0 HA ends upstream of the cleavage site. (C) HA1 is amplified from genomic DNA (1), the pTV cherry plasmid is linearized using restriction sites in the 5 0 MCS (2) and both are ligated (3). The pTV cherry +HA1 plasmid is linearized using restriction sites present in the 3 0 MCS (4) and the HA2 insert is PCR amplified from genomic DNA (5) before it is ligated into the plasmid backbone (6). The plasmid containing both HAs is purified for microinjection (7). (D) The resulting pTV cherry +HA1+HA2 plasmid can be used as a dsDNA donor plasmid during CRISPR engineering.  Figure 5C, step 3). e. Transform the ligated plasmid into bacterial competent cells. 7. Days 2-5. Verify successful HA1 insertion by diagnostic restriction digests and sequencing.
a. Day 2. Pick bacterial colonies and set up liquid cultures for approximately14 h. b. Day 3. Use liquid cultures to generate plasmid minipreps. Set up a diagnostic restriction digest https://www.addgene.org/protocols/diagnostic-digest/ using, for example, the restriction enzymes that were used to insert the HA1 insert. Check the DNA fragments using gel electrophoresis. c. Sequence minipreps, which show successful HA1 insertion by diagnostic digest, using primers positioned upstream and downstream of the insert. d. Day 5. Verify successful HA1 insertion from the sequencing reads (see troubleshooting 2). 8. Days 6-10. Clone the HA2 insert into the pDsRed-24MS2+HA1 plasmid ( Figure 5C, steps 4-6).
Note: To generate the pDsRed-24MS2 plasmid, the 24xMS2 stem-loops were extracted from the pCR4-24xMS2SL-stable plasmid (Addgene, Cat# 31865) by BglII and BamHI digestion and inserted into the BamHI restriction site of pBluescript together with two loxP sites and a DsRed marker gene (Lim et al., 2018a).

Two-step CRISPR
10. Cloning steps are similar to the one-step CRISPR approach outlined in detail above, but a different starting plasmid that contains an attP site is used ( Figure 6). In short: a. Day 1: HA1 and 2 are amplified using a high-fidelity DNA polymerase and digested with restriction enzymes that recognisethe sequences that were added to the HA primers (Figure 6C, steps 1 and 5). b. HA1 is inserted into the linearized pTV Cherry vector ( Figure 6C, steps 2 and 3). c. Days 2-5. Verify successful HA1 insertion by diagnostic restriction digest and sequencing (see troubleshooting 2). d. Day 6. Ligate HA2 into the linearized pTV Cherry +HA1 plasmid ( Figure 6C, steps 4 and 6). e. Days 7-10. Use diagnostic digests and sequencing to verify the successful insertion of HA2 in the plasmid. 11. Day 11. Generate a maxiprep of the pTV Cherry +HA1+HA2 plasmid for microinjection (see additional considerations in protocol step 4) ( Figure 6C, step7). The finished pTV Cherry +HA1+HA2 plasmid is shown in Figure 6D.
Pause point: The plasmids can be stored at À20 C for years.

Microinjection and selection of CRISPR edited flies
Timing: One-step CRISPR $8 weeks; two-step CRISPR $8 weeks to insert attP and 3 weeks to amplify for microinjection One-step CRISPR and two-step CRISPR Note: Initially inject 200-400 Drosophila embryos. More embryos can be injected if low recovery of edited flies is expected, for example due to targeting genomic regions on the X chromosome that may cause male lethality. Many commercial injection services are available that have optimized the injection protocol.
CRITICAL: When using the DsRed marker gene or another fluorescent marker, make sure to use a Cas9 transgenic line without fluorescent markers of similar wavelength.
Here we use a line with nos-Cas9 on the X-chromosome and no additional markers. This line has the advantage that it is compatible with CRISPR schemes based on selection with a fluorescent protein or mini-white, but relies on loss of nos-Cas9 over time. Note: The plasmid concentration to be injected varies. We inject 100 ng/ml of each gRNA plasmid and 500 ng/ml of the donor plasmid. Other concentrations have been tested (Gratz et al., 2015). Figure 7 and 8A show the crossing schemes for a targeted gene locus on the 2 nd chromosome.

Note: The examples in
Note: The example crossing scheme for the one-step CRISPR approach outlines a time efficient way to identify the CRISPR-Cas9 genome modification and remove the Cre-recombinase from the genome. Crossing the injection survivors to wildtype flies may increase the number of progeny, increasing the chances of finding a fly with a modified genome, but will then require the extra step of balancing the edited locus so it can be selected when the visible marker is removed by Cre.
One-step CRISPR  ii. After the cross has produced a sufficient amount of eggs and larvae are visible, verify the correct genome integration by purifying genomic DNA from the DsRed positive CRISPR edited fly and sequencing PCR amplified regions across the editing site troubleshooting 4. iii. Keep additional CRISPR positive flies as a backup. 16. Week 7. Identify flies that have undergone recombination (loss of DsRed expression) and carry the CyO, P[w + , Cre] expressing chromosome. Generate a homozygous stock by crossing the flies to each other (Figure 7, step 5). 17. End of week 8. In flies emerging from the cross, selecting flies homozygous for the insertion will allow the loss of P[w + , Cre], located on the CyO balancer, from the genome. The homozygous flies allow the stock to be maintained and are ready to be used for live imaging experiments.
Note: In the absence of selection, nos-Cas9 will be lost from the genome over time. Other lines with Cas9 linked to a fluorescent protein or white marker allow for it to be selected against in subsequent generations.
Note: The one-step CRISPR approach will leave a loxP (34 bps) site and a few nucleotides (MCS) behind as a scar in the genome. The scarring can minimized by using restriction sites closest to the MS2 stem-loops/loxP in the pDsRed-24MS2 plasmid.

Two-step CRISPR
18. Week 3. Cross all surviving flies to a whitefly stock (F0 generation) in individual crosses (Figure 8A, step 2). a. The example in Figure 8A, step 2, shows a surviving female crossed to y 1 w 67c23 males [RRID:BDSC_6599]. 19. Week 5. Identify F1 CRISPR edited flies by the presence of red eyes due to the mini-white marker gene ( Figure 8A, step 3) (see troubleshooting 3). Cross single transformants to a 2 nd chromosome balancer fly line (for example y 1 w 67c23 ; sna Sco /CyO (Ashe lab)) ( Figure 8A, step 4). a. Set up individual crosses with a number of potential CRISPR positives in case of off-target effects. b. After the cross has produced a sufficient amount of eggs and larvae are visible, verify correct genome integration by PCR and sequencing the targeted region in the adult fly troubleshooting 4. 20. Week 7. Select w + CyO flies and cross them to each other to generate a stock that is homozygous for the attP insertion (lacking CyO) ( Figure 8A, step 5). Alternatively, this will allow maintenance of a balanced stock if the CRISPR edit causes lethality (see Note section below). 21. Weeks 9-11. Amplify homozygous (or balanced where necessary) flies to obtain enough adults for a second round of microinjections.
Note: attP containing flies will not be homozygous viable if the deleted genome region contains important regulatory elements such as promoters or enhancers. In this case the balanced stock can be used for injection of the reintegration plasmid.
Note: The visible marker gene is not removed during this step, as the plasmid that will be used for reintegration through the attP site, contains a different marker. Both marker genes can be removed at the same time using Cre-recombinase after the reintegration step (Baena-Lopez et al., 2013).
fC31 integrase-mediated site-specific transgenesis Timing: Cloning, 6 days; transgenesis, $8 weeks ll OPEN ACCESS Two-step CRISPR In this step, site-directed reintegration utilizing the attB-attP system is used to insert MS2 stem-loops and the sequences that were removed in the first CRISPR step into the endogenous gene locus. To this end, the previously deleted genomic sequences are inserted into the RIV cherry +24xMS2-stemloop plasmid ( Figure 9A), upstream of the MS2 stem-loops. Other sequences can be introduced into the genome instead of the 24xMS2 stem-loops by using a similar approach.

Cloning
Timing: 6 days Note: It is most time efficient to perform the molecular cloning to insert the genomic region into the reintegration plasmid concurrently with the fly crosses described above that generate the homozygous CRISPR edited flies carrying an attP site. . Cloning strategy for the reintegration plasmid used for two-step CRISPR (A) The RIV cherry +24xMS2-stem-loop plasmid contains an attB site for fC31 integrase-mediated recombination and the MS2 stem-loops. (B) The plasmid is linearized using EcoRI and NotI restriction sites that are part of the 5 0 MCS (1). The genomic region, removed in the first CRISPR step, is amplified from genomic DNA and digested with EcoRI and NotI (2) followed by ligation into the linearized vector (3). After successful insertion, the plasmid is purified for microinjection (4). (C) The reintegration plasmid also contains a pax-mCherry marker gene to identify positives.
22. Day 1. Digest the RIV cherry +24xMS2-stem-loop plasmid with EcoRI and NotI restriction enzymes ( Figure 9B, step 1). Cloning steps are very similar to cloning the HA sequences (outlined in detail above, see steps 5-9), so are summarized in this section (see troubleshooting 2). 23. PCR amplify the genomic region, which was removed during the first CRISPR step and sits inside of the HAs. This includes the cleavage site to PAM site sequence shown in black and orange, Figure 3B. Use primers designed in the before you begin section that insert an EcoRI and NotI restriction site at either end of the insert (Figure 9, step 2). Ligate the insert into the linearized plasmid ( Figure 9B, step 3). 24. Days 3-5. Confirm the insertion by diagnostic digests and sequencing. 25. Day 6. Generate a maxiprep of the finished RIV cherry +24xMS2-stem-loop+genomic region plasmid ( Figure 9C) for microinjection ( Figure 9B, step 4) . Ensure the DNA is of high quality (see additional considerations in protocol step 4).
Alternatives: The genomic DNA can be inserted downstream of the MS2 stem-loops using the 3 0 MCS. If the MS2 stem-loops need to be placed within the genomic region that is reintegrated, two inserts can be generated and inserted upstream and downstream of the stemloops respectively or alternative cloning strategies can be used.
Alternatives: Instead of supplying a fC31 integrase plasmid, flies can first be crossed to a fly line that expresses fC31 integrase (for example RRID:BDSC_34771).
27. Week 3. Cross surviving F0 flies to 2 nd chromosome balancer flies ( Figure 8B, step 2). a. The example in Figure 8B, step 2, shows a surviving female fly crossed to males of the y 1 w 67c23 ; sna Sco /CyO genotype (Ashe lab) . b. Individual crosses should be set up for a number of transformants in case of incorrect integration events. 28. Week 5. Select transformants based on pax promoter-driven mCherry fluorescence in the eyes of F1 flies ( Figure 8B, step 3) (see troubleshooting 5). 29. Week 5. Remove the combined marker region, now containing 3 loxP sites, the mini-white marker and the mCherry marker by crossing F1 transformants to a Cre-recombinase expressing fly line ( Figure 8B a. Cross females and males that carry the modified locus (are mCherry negative) and the CyO balancer ( Figure 8B, step 5). 31. End of week 8. In the F3 generation select against the CyO balancer and the white marker (P[w + , Cre]) to remove Cre from the genome. These flies are now homozygous for the genome modification, which allows for the stock to be maintained and it is ready to be used for live imaging experiments.
Note: The two-step CRISPR approach will leave small scars in the genome in the form of the attR (attP/B) site at the 5 0 end and one LoxP site (34 bps) at the 3 0 end of the targeted locus. The plasmids used in this protocol contain minimal attP (50 bps) and attB (51 bps) sites (Baena-Lopez et al., 2013) that were established by Huang et al., (2009), reducing the attR sequence length contained in the final engineered fly line to 48 bps. The small attR size minimizes the amount of sequence to be added in addition to the MS2 stem-loop cassette, but adding additional sequences will be more problematic for other CRISPR applications that target the coding sequence. Additionally, a few nucleotides will have been inserted depending on the restriction sites that were used for cloning. Most of these additional nucleotides can be avoided by using In-Fusion or Gibson cloning or commercially synthesizing the plasmid.

EXPECTED OUTCOMES
The expected outcome is the insertion of 24xMS2 stem-loops (or alternative donor sequence) into a specific position within the gene locus. Using the two-step CRISPR approach, the gene locus is marked with an attP site as an intermediate step, which can be used for further modification of the gene region. Tagging an endogenous locus with MS2 sequences will enable live imaging studies to investigate and quantitate nascent transcription dynamics. For our recent study investigating the transcriptional regulation of endogenous Bone Morphogenetic Protein target genes see Hoppe et al., 2020.

LIMITATIONS
While inserting MS2 sequences into the genome using CRISPR gives the advantage of allowing the study of endogenous transcriptional activity, it is also time consuming. CRISPR modification takes more time than generating a transgene and inserting it into a targeted landing site within the genome. Therefore, the advantages of being able to investigate endogenous transcription dynamics versus studying a reporter transgene have to be balanced with the time it takes to generate the fly lines. In addition, once the attP line is made using the two-step approach, test sequences can be rapidly targeted to the gene locus.
Another limitation is that the addition of MS2 stem-loops could alter gene expression. Even though transcription is not expected to be altered, it is possible that the addition of MS2 stem-loops to the 5 0 or 3 0 UTR leads to changes in translation efficiency or mRNA regulation (Mayr, 2017;Palam et al., 2011;Vattem and Wek, 2004). These potential disruptions to gene expression could lead to a reduced viability of fly stocks.
Some regions of the genome appear resistant to targeting. However, as there is flexibility in the placement of the MS2 sequences and therefore the gRNAs that can be used, this is not a major issue here. Potential solution If no target sites are found, consider lowering the stringency of the search. With a lower stringency, more target sites will be identified, as gRNAs often tolerate several mismatches in their seed sequence. The danger is that a lower stringency also increases the potential for off-target events.

TROUBLESHOOTING
The length of the seed sequence can be shortened to 16 nt to increase target site identification and reduce off-target probability. While sequences shorter than 16 nt can be used to guide Cas9, they are insufficient to promote endonuclease activity (Dahlman et al., 2015;Fu et al., 2014). It was shown that seed sequence lengths of 17 and 18 nt reduce the potential for off-target effects while at the same time retain normal efficiency in mammalian cells (Fu et al., 2014).
If no target sites are identified at the region of interest when using a single gRNA approach or only target sites with a high number of possible off-target binding sites are available, consider using two gRNAs. Choosing two target sites that are located upstream and downstream of the region of interest will allow modification of the region of interest but will require insertion of the DNA sequence between the cleavage sites into the donor plasmid to avoid a deletion.
Theoretically, other Cas proteins can be used that favor different PAM sequences and therefore make other target sites available. For example the Cas12a protein is available for use in Drosophila (Port et al., 2020).
Try to avoid off-target locations on the same chromosome arm or even the whole chromosome as the region of interest. Potential off-target events on other chromosomes can be crossed out.

Problem 2
One of the plasmids cannot be generated.

Potential solution
Generally, oligonucleotides are easy to clone so problems generating the plasmids to express the gRNAs seem unlikely. It may be more difficult to clone a large sequence in the reintegration plasmid so if this is an issue, try to use gRNAs closer together to reduce the length of sequence needing to be replaced. Additionally, HAs can be shortened (Beumer et al., 2013;Kanca et al., 2019). If cloning is problematic, the DNA can be commercially synthesized.

Potential solution
There are many possibilities for why CRISPR-Cas9 genome engineering is unsuccessful that relate to low viability of the flies or poor efficiency of editing. The DNA quality of plasmids for injection is crucial to ensure good viability so the DNA should be repurified to ensure it is clean. More embryos can be injected to increase the chances of an editing event and there are a range of commercial injection services that have experience in generating CRISPR edited flies. Post injection, dehydration of the vial containing survivors must be avoided as it can kill the larvae. Incubate the fly food vial containing injection survivors in a humidified incubator and add saturated filter paper to the fly food if it appears to be drying out. Adding liquid yeast paste to the food also increases survival of larvae. Finally, survival of adults is enhanced by removing them from the tube as soon as they eclose. The choice of line for injection also influences survival. Many researchers favor lines with nos-Cas9 instead of vas-Cas9, as nos expression is more tightly confined to the germline (Port et al., 2015). Weak somatic vas-Cas9 expression may lead to some lethality caused by somatic CRISPR events. In addition, if the locus being targeted is located on the X chromosome, it may be necessary to inject a greater number of embryos to achieve a successful modification. This is particularly relevant to the two-step protocol as it involves deletion of some sequences, which may cause lethality in males.

OPEN ACCESS
For the two-step CRISPR approach, try to reduce the amount of sequence deleted, as the efficiency of CRISPR-Cas9 events may depend on the size of the deletion. Even though large deletions of up to 30 kb have been successfully replaced with an attP site, smaller deletions can be obtained with a higher efficiency and limit rearrangements at the edited region (Poernbacher et al., 2019).
The one-step protocol describes crossing the injected flies to a stock carrying balancers, as this saves time. As balancer stocks can be less healthy, to maximize progeny and increase the chances of identifying an edited fly it may be better in some cases to first cross injection survivors to wildtype flies.

Problem 4
No transformants are obtained for reintegration.

Potential solution
Transformation efficiency with the fC31 system is usually very high. However, if no transformants are identified, make sure that the attP site was inserted into the genome correctly by sequencing, and that the attB site in the reintegration plasmid is intact. Sequence the fC31 integrase containing plasmid and ensure good DNA quality of all plasmids before injecting more embryos. The efficiency of reintegration is reduced when longer sequences are reintegrated.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Hilary L. Ashe (hilary.ashe@manchester.ac.uk).

Materials availability
Plasmids used in this study are available at Addgene or the Drosophila Genomics Resource Center. The RIV cherry +24xMS2-stem-loop plasmid and the y 1 w 67c23 ; sna Sco /CyO and the yw, nos-Cas9 fly lines are available upon request from the lead contact. Other fly stocks are available from the Bloomington stock center.

Data and code availability
This study did not generate or analyze datasets or code.