Applying consecutive or non-consecutive sequential electroporations to generate novel conditional alleles: a roadmap
To date, the Easi-CRISPR method, employing a combination of two guide RNAs complexed with the Cas9 protein along with a lssDNA repair template, seems to be the most widely adapted method for generating conditional alleles. The method has been proven to be robust and reliable for generating conditional alleles for most genes and was successfully applied for two previously generated models in our laboratory (data not shown) . However, during the course of our work, we realized some limitations of this approach that precluded its applicability to all loci (details see Table 1). Specifically, the Easi-CRISPR method was unusable for four out of eight projects (details see Table 1). The projects involving Icam1 and Clcf1 were incompatible with the Easi-CRISPR approach due to the distance between both loxP sites, requiring a targeting construct greater than 2 kb. Moreover, Easi-CRISPR could not be successfully applied to the projects involving Lox and Pard6g due to the high sequence complexity surrounding the targeting region. To circumvent these limitations, we used a modified version of the electroporation conditions reported by Troder et al. along with the sequential electroporation method reported by Horii et al. (details see “Methods” section) [10, 11]. The rationale that we have used for each project is summarized in Fig. 1. In short, we employed a strategy to generate conditional alleles according to two possible scenarios; (1) by consecutive sequential electroporation (strategy A); or (2) by non-consecutive sequential electroporation (strategy B). The first attempt for each project was via consecutive sequential electroporation (Strategy A, blue rectangle Box, Fig. 1). We rationalized that this approach was the shortest path to success if it worked. If it failed, we investigated whether or not any pups resulting from the initial consecutive sequential electroporation session could be usable for the non-consecutive sequential electroporation approach (Strategy B, grey rectangle Box, Fig. 1).
This modified procedure was extended to the four remaining projects highlighted in Table 1. These included two projects where the Easi-CRISPR method failed to produce animals containing the desired alleles (Sar1b and Loxl1, Table 1 and Additional file 1). In this case, three positive animals were obtained that contained partial construct integrations at the targeted site (one for Sar1b and two for Loxl1, Additional file 1) and five animals were obtained that contained random construct integrations (three for Sar1b and two for Loxl1, Additional file 1). Random and partial integration screening strategies for Sar1b and Loxl1 are detailed in Additional files 2 and 3. The Sar1b partially integrated construct contained a properly targeted loxP site on one side of the desired exon and a 14 base pairs deletion on the opposite side. The Loxl1 partial integration consisted of a sequence inversion in the 3′homology arm of the repair template along with a 40 base pairs deletion for one animal and a properly targeted loxP site on one side of the desired exon with no indels on the opposite side. This latter observation suggested a difference in guide cleaving efficiency for this project. Interestingly, these two phenomena of random and partial integrations have been previously reported in the literature for projects using lssDNA [7, 12]. We extended our consecutive and non-consecutive electroporation strategies to the two remaining projects that had no limitation for using lssDNA as a repair template (Pard6a and Mapkapk5, Table 1). For these two projects, the consideration of cost and synthesis turnover time for the generation of a lssDNA construct weighed against the possibility of obtaining a partial integration model, prompted us to instead invest in short ssODNs. Our rationale also took into consideration the fact that in most cases, a single properly targeted loxP site animal with wild-type sequence on the opposite site could be obtained leading us to a strategy B alternative (non-consecutive sequential electroporation). Essentially, in this case, we reasoned that one properly targeted loxP site was better than none.
These limitations and challenges prompted us to try the sequential electroporation approach reported by Horii et al. . We initially applied this strategy using the electroporation conditions described by Troder et al., which used 4 μM Cas9: 4 μM Guide: 10 μM DNA repair template concentration . This method was used for the Icam1 project with a resulting final concentration of 8 μM Cas9: 8 μM Guide: 20 μM DNA repair template over two electroporation sessions (4 μM: 4 μM: 10 μM on each day). The embryo survival rate using these conditions was 86%, where 86 2-cells stage embryos out of 100 were implanted in four pseudopregnant females. However, the percentage of live born animals using this procedure was low, with only three pups born out of three gestations and none of them surviving past the first week of birth (details see Table 2). These results prompted us to consider modifying the reagent concentrations used in our electroporation procedure. Considering that the Cas9 protein remains active for more than 24 h after electroporation in embryos, and the fact that 2-cells stage embryos have a similar volume as 1-cell stage embryos, we rationalized that keeping a final concentration at 4 μM Cas9: 4 μM Guide: 10 μM DNA repair template would be optimal for cleavage efficiency and pups viability (i.e. 2 μM: 2 μM; 5 μM on each day). We performed a second round of electroporation for the Icam1 project, with the reagent concentrations mentioned above. This resulted in an embryo survival rate similar to the one obtained with the initial concentration previously used (81 2-cells embryos out of 100 electroporated). However, in this case, the percentage of live born animals was higher, with 18 pups born out of four gestations (Table 2). Hence, these results prompted us to apply the same reagent concentrations for each of our projects going forward.
Applying consecutive or non-consecutive sequential electroporation strategies to generate novel conditional allele: project design
For each project, the design relied on the selection and use of two annealed RNA guides, referred here as pgRNA (crRNA-tracrRNA formulation) and symmetric short single strand oligonucleotides (ssODNs) as repair templates that contained 60 base pairs homology arms on each side, and a loxP sequence in between (repair template details, see Additional file 5). Sequence length between both homology arms varied depending on whether a single loxP site (34 base pairs) or an associated adjacent EcoRI or NheI restriction sites (40 base pairs) was incorporated along the loxP sequence. The repair templates were designed to correspond to the targeting strand, complementary to the Cas9 selected guide and its associated PAM site sequence, with an exception for the Loxl1 project, where repair templates of both orientations in the Dn position (3′ of the targeted exon) were used to complete the project (Additional file 5).
Literature review and gene structure analyses were performed for each individual project to select exons that were predicted to have the most detrimental effect on the protein product when deleted. In silico guide cutting surveys were performed for each candidate exon using three different softwares (CRISPOR (http://crispor.tefor.net/crispor.py), CHOPCHOP (https://chopchop.cbu.uib.no/) and Breaking-Cas (https://bioinfogp.cnb.csic.es/tools/breakingcas/)) on selected genomic DNA regions as described in the methods section [13,14,15]. A total of three guide cutting pairs were selected for each individual project. crRNAs corresponding to the top-ranking guide pair, cutting on each side of the candidate(s) exon(s), were ordered from IDT along with the two corresponding ssODN repair templates. The remaining two pairs for each projects were kept in proviso. Complete lists of the different crRNA and corresponding repair templates are highlighted in Additional files 5 and 6. In some instances, an additional crRNA pair was ordered and used in the initial sequential electroporation procedure (Pard6a and Pard6g, Additional file 6). Consecutive sequential electroporation sessions were performed for each selected guide pairs as described in the methods section. Briefly, RNP complexes formed by the association of one of the two selected pgRNA with the purified Cas9 protein were electroporated in 1-cell stage embryos along with the corresponding repair template (Strategy A, Fig. 1). Electroporated embryos were recovered and left to develop to the 2-cells stage overnight at 37 °C under 5% CO2. 2-cells stage embryos were electroporated with the second RNP complex along with the corresponding repair template before being implanted in pseudopregnant females (0.5 dpc) (Strategy A, Fig. 1).
Applying consecutive or non-consecutive sequential electroporations to generate novel conditional allele: properly targeted pups characterization
The resulting pups were characterized using a genotyping approach previously described in the literature with primer series exemplified in Fig. 2 . Briefly, six primers were routinely designed for each project. These comprised two pairs, mapping outside the ssODN homology arms used to insert the loxP site either in the Up (5′ of the targeted exon(s)) or Dn (3′ of the targeted exon(s)) positions (Fig. 2, primers 1–3, Up; primers 4–6, Dn). Two additional primers were designed with overlaps between the genomic DNA sequence adjacent to the loxP insertion site (20 base pairs) and a portion (15 base pairs) of the loxP site itself (Fig. 2, primers 2 and 5). These last primers were designed to be used as a pair with one primer pointing in the forward and the other in the reverse orientation. A complete primer list for each project is found in Additional file 7.
Our standard genotyping strategy consisted of using the long loxP site overlapping primers 2 and 5 (Fig. 2, upper panel) as an initial screening step to identify any positive animals containing both loxP sites in cis (on the same allele). Animals were also investigated by using primers 2–6 and 1–5 combinations in separate PCR reactions (Fig. 2, lower panels). Positive PCR products from these last reactions were then send for sequencing using either primer 1 or 6 depending on the initial primer pairs used (red primers Fig. 2, lower panels). In some instance, primers 3 and 4 were used for further validation. This latest screening strategy was applied to all of the described projects except for the one involving Pard6g that required a different approach since it was impossible to obtain a full-length PCR product between the targeted exon due to high sequence complexity (genotyping strategy, see Additional file 4). Using this screening method, we were able to recover pups with both loxP sites in cis for a total of three projects (details see Table 3) with an average integration rate of 8% (range from 5 to 13%). Germline transmission was confirmed for two of these projects using the same genotyping strategy.
The remaining five projects were completed using the non-consecutive sequential electroporation strategy. In this case, we focused our investigation on finding positive pups with a single loxP site integration on one side and wild-type sequence on the other side (Fig. 3A). This was achieved by using positive PCR products from the same 2–6 or 1–5 primer pairs described previously and sequencing these PCR products with either primer 1 or 6 depending on the initial primer pairs used (Fig. 3A). In this case, the sequencing results informed us as to whether or not the insertion site that failed to incorporate the loxP site was exempt of indels. If this was the case, an additional primer, overlapping the genomic DNA sequence adjacent to the loxP insertion site and a portion of the loxP site itself in opposite direction to the one initially designed was used to confirm the integrity of the inserted loxP site (primer 7, Fig. 3A). In this case, PCR products from primers pairs 1–7 and 2–3 were sent for sequencing using primers 1 and 3 respectively. Pups that were exempt of indels in the site that failed to incorporate a loxP site on one side and had proper integration of the loxP site on the other side were bred for germline transmission. The resulting N1 animals were sequence verified as described above and bred to N2 before being intercrossed to produce embryos that were used to incorporate the missing loxP site (Strategy B, Fig. 1). We reasoned that using this strategy would increase the likelihood to obtain the properly targeted allele as 25% of the embryos would be homozygotes with a single loxP site on both allele, 50% would be heterozygotes with a single loxP site on one out of two alleles, and 25% would be wild-type. For each project, electroporation on 1-cell stage embryos was performed using the material to incorporate the missing loxP site as described above before being implanted in pseudopregnant females (0.5 dpc). The resulting pups were investigated for proper loxP targeting as described previously, with priority given to pups showing positive bands using the 2–5 primer pairs (Fig. 3B). Using this strategy, we were able to obtain properly targeted pups for the remaining five projects (details see Table 4), with a targeting efficiency averaging 11% (ranging from 3 to 25%). Data from the Loxl1 project were used to compare the targeting efficiency when using ssODNs corresponding to the targeting versus non-targeting strand for insertion of the second loxP site (details see Table 4). Interestingly, in this case, the ssODN corresponding to the non-targeting strand gave us a greater efficiency, with a value of 6%, when compared to the ssODN corresponding to the targeting strand that only resulted in 3% efficiency. Hence, these results suggest that, for conditional allele model generation using two ssODNs, the choice between using the targeting versus non-targeting strand as a repair template should be determined empirically as repair efficiency using either one of these strands appear to be context dependent. Germline transmission was confirmed in all five projects using the same genotyping strategy as the one described in Fig. 3B.
The PCR products from primer pairs 2–6 and 1–5 were also used to assess independent loxP site targeting efficiency between the electroporations performed at either the 1-cell or 2-cells stage in the resulting pups from the initial consecutive sequential electroporation procedure for each project (Table 5). Interestingly, the loxP targeting efficiency in pups for the projects completed using Strategy A showed no statistical differences, with an average of 27 ± 7% at the 1-cell stage and 32 ± 7% at the 2-cells stage (T-Test, P = 0.42). Whereas, the loxP targeting efficiency in pups for the projects completed using Strategy B showed significant statistical differences, with an average of 24 ± 7% at the 1-cell stage and 7 ± 4% at the 2-cells stage (T-Test, P < 0.05). These results raise the possibility that for all the projects completed using Strategy B, improving the targeting efficiency at the 2-cells stage may have increased the likelihood of completing these projects using consecutive sequential electroporation.
Furthermore, chromosomal deletions that are caused by simultaneous guide cleavage activity inducing double strand DNA breaks at two different positions on a chromosome have been reported using sequential electroporation, with an incidence varying between 9 and 38% . We did not systematically investigate this incidence during the course of our work as we mainly focused on identifying cis targeted animals. However, we were able to observe this phenomenon in some instances at a rate varying between 11 and 24%, which is similar to what has been reported previously (Table 5).
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.