Detailed Legend for Fig. 1

SUPPLEMENTARY MATERIAL

Detailed legend for Fig. 1

(A) An improved donor vector was co-injected with the Cre expression vector pCAG/NCre (Sato et al. 2000) into the pronuclei of eggs that were fertilized in vitro with epididymal spermatozoa carrying the targeted allele. In this study, three vectors—pAWK, pAWV and pAXV—were used, all of which contained the FLPe expression unit and GOI expression cassette (wherein the artificial miRNA targeted to the EGFP gene, driven by the CAG promoter, was included for pAWK and pAWV and the Dre reporter construct for pAXV). RMCE is expected to occur in these microinjected eggs to generate the RMCEex allele (Ohtsuka et al. 2010). After RMCE, self-removal of the extra sequence such as the FLPe expression unit is also expected to occur to generate the RMCEΔex allele. CAG, cytomegalovirus enhancer + chicken beta-actin promoter; pA, poly(A) addition sites; neo, neomycin resistant gene; IRES, internal ribosomal entry site; CFP, ECFP-Nuc (Clontech); JTZ17 and JT15, inverted repeat variants of loxP (Thomson et al. 2003); lox2272, spacer variant of loxP (Lee and Saito 1998); FRT, FLP recognition target; GOI, gene of interest; FLPe, enhanced FLP recombinase (Buchholz et al. 1998); a–f, primers used to detect PITT. In this schematic diagram, a simple illustration of the improved donor vector was provided for easy understanding of this system. See Supplementary Materials and Methods for the details of the GOI used in this study and direction of DNA fragments in the donor vectors. (B)The plasmid mixture at a final concentration of 15 ng/µl (10 ng/µl of improved donor vector [pAWK, pAWV or pAXV] and 5 ng/µl of pCAG/NCre) in injection buffer (5 mM Tris-Cl/0.1 mM EDTA, pH 7.5) was microinjected according to the standard protocol (Nagy et al. 2003). The following day, embryos that developed into the two-cell stage were transferred to the oviducts of a foster mother. GenomicDNA extracted from the ears of the offsprings resulting 3 weeks after birth (the time of weaning) was then subjected to PCR to analyse the presence of integrated transgenes. Founder mice identified as positive after PCR using all three primer sets (‘a and b’, ‘c and d’ and ‘e and f’ shown in [A]) were considered to possess the RMCEΔex allele. Mice that were identified as positive after PCR using two primer sets (‘a and b’ and ‘c and d’), but negative using another primer set (‘e and f’) were also considered to possess the RMCEex allele. The mice that were identified as positive after PCR using only the primer set (‘a and b’) were regarded as randomly integrated transgenics. The 95% confidence interval (CI) of success rate was calculated with the binomial distribution by using Stata version 11.0. N.D., not determined.

Supplementary materials and methods

Donor vector construction

The sequences and maps of the plasmids and primers used in the present study are available upon request. pAWK, pAWV and pAXV were obtained through ligation-based multi-step cloning. Site-specific recombination sites such as FRT, JTZ17 (Thomson et al. 2003) and lox2272(Lee and Saito 1998) were the same as those in the previous donor vector used by Ohtsuka et al. (2010). The CAG promoter and poly(A) addition site and the FLPe coding region included in the FLPe expression cassette were derived from pAOK (Ohtsuka et al. 2010) and pCAGGS-FLPe (Schaft et al. 2001) (Gene Bridges), respectively. The vector backbone of both donor vectors was derived from a 3.4 kbClaI/EagI fragment of pBR322. In the GOI region, the following cassettes were included in each donor vector: ‘CAG promoter–artificial miRNA targeted to EGFP gene–poly(A)’ in pAWK, ‘CAG promoter–artificial miRNA targeted to EGFP gene–tdTomato–poly(A)’ in pAWV and ‘CAG promoter–rox-EGFP–poly(A)–CAT (chrolamphenicol acetyltransferase)–poly(A)–rox–tdTomato–poly(A)’ in pAXV. The nucleotide sequences of junctional portions generated after cloning and PCR-amplified regions were confirmed by sequencing. The pAWK vector had the following components in the 5′to 3′direction: JTZ17, lox2272, CAG promoter, artificial miRNA, poly(A), FRT, poly(A), FLPe, CAG promoter and pBR322-based backbone. The pAWV vector had the following components in the 5′to 3′direction: FRT, poly(A), tdTomato, artificial miRNA, CAG promoter, lox2272, JTZ17, poly(A), FLPe, CAG promoter and pBR322-based backbone. The pAXV vector had the following components in the 5′to 3′direction: FRT, poly(A), tdTomato, rox, poly(A), CAT, poly(A), EGFP, rox, CAG promoter, lox2272, JTZ17, poly(A), FLPe, CAG promoter and pBR322-based backbone.

Pronuclear injection-based targeted transgenesis (PITT)

Donor plasmids and pCAG/NCre were prepared using a HiSpeed Plasmid Midi Kit (Qiagen, Hilden, Germany). To perform PITT, the plasmid mixture was first prepared, as per the detailed legend of Fig. 1. Unfertilized oocytes isolated from super-ovulated BDF1 female mice were subjected to in vitro fertilization (IVF) with spermatozoa obtained from a male carrying a homozygote-targeted allele at the Rosa26 locus. The plasmid mixture was introduced into these in vitro fertilized eggs, as per the detailed legend of Fig. 1. After transfer of DNA-injected embryos to the recipient females, the resulting offsprings were inspected for the presence of transgenes, as per the detailed legend of Fig. 1.

Detection of successful RMCE and elimination of the extra sequence

To detect correct recombinants, the primer sets were designed to amplify the junctional region generated by recombination, as shown in Fig. 1. RMCE-mediated integration of donor vectors into the Rosa26 locus was assessed by PCR using the two primer sets, ‘a and b’ and ‘c and d’. The removal of the extra sequence was confirmed by PCR using the primer set ‘e and f’. Mice that were identified as positive after PCR using the primer set ‘a and b’ alone were regarded as randomly integrated transgenic mice (Fig. 1).

Supplementary comments

We have recently developed a targeted transgenesis strategy termed PITT through pronuclear injection (Ohtsuka et al. 2010). PITT is based on recombinase-mediated cassette exchange (RMCE) using the Cre-loxP system and generates clean integration of GOI into the predetermined locus (e.g. Rosa26). This study includes three technological improvements for PITT. Supplementary comments regarding each improvement are listed below.

1. Self-removal of the extra sequence

In our previous PITT strategy, two steps (RMCE-mediated integration and removal of the extra sequence) were required to obtain a cleanly integrated RMCEΔex allele. Removal of the extra sequence was time-consuming and often laborious because it was achieved by an additional FLPe-administration step such as pronuclear injection of the FLPe plasmid or crossing the RMCEex mice with FLPe mice. In this study, an FLPe expression cassette was introduced into the donor vector to remove the extra sequence automatically (direct creation of RMCEΔex allele). As a result, three out of five founder mice exhibited self-removal of the extra sequence, enabling us to analyse these transgenic mice in the next (F1) generation. EGFP knockdown was observed in F1 offsprings that were obtained by crossing a founder F0 mouse (pAWV line) with a Rosa26CAG::EGFP mouse (Ohtsuka et al. 2010) (unpublished data). When the previous donor vector was used, an additional generation (to remove the extra sequence) was required to perform the above analysis. This result indicated the usefulness of new donor vectors as a time-saving tool.

As mentioned in the main manuscript, two other founders failed to show self-removal of the extra sequence. The reason why such mice were generated is unclear, but it may be due to the silencing of the FLPe expression cassette. This appears to be similar to the situation we experienced previously as an inconsistent expression of GOI in the RMCEex allele and is attributed to promoter interference and/or epigenetic silencing(Ohtsuka et al. 2010). Alternatively, the secondary structure of the DNA possibly formed by the two CAG promoters included in the FLPe expression cassette and GOI region may have interfered with the transcription of the FLPe gene. In the latter case, use of different promoters (for the FLPe gene or GOI) may overcome this problem, and this point will be assessed in the near future. In any case, we hypothesize that FLPe expression was transient; therefore, the removal efficiency was 60% (3/5), which was similar to the rate obtained by pronuclear injection of the FLPe plasmid (58%) (Ohtsuka et al. 2010). Since we plan to generate more PITT mice using these improved vectors, a more exact removal efficiency will be obtained in the near future.

The extra sequence in the mice that did not exhibit self-removal was successfully removed through additional FLPe treatment by crossing with an FLPe deleter mouse (Kanki et al. 2006). In pAWV transgenics, similar knockdown efficiencies were obtained among the three mouse lines tested (data not shown). This property enables us to reduce the number of mouse lines analysed. Once aRMCEΔex line is directly obtained using the improved donor vector (e.g. pAWV transgenic mice in this study), we can begin functional analysis using these F1 mice without additional FLPe administration. This system should shorten the time required for sampling. In other words, three months that correspond to one generation will be omitted with this system.

1. The pBR322-derived vector backbone

In the improved vector, the pBR322-vector backbone (low-copy) was used and not that based on pUC119 (high-copy), as used in the previous vector. This modification was aimed at stably maintaining the insert DNA (e.g. genomic DNA) in Escherichia coli. High-copy-number plasmids such as pBluescript have been frequently and routinely used for subcloning genomic DNA (e.g. targeting vector construction). However, in some cases, the high-copy-number of the vector failed to maintain stable clones carrying large fragments, regions with repetitive sequences or those with secondary structure and caused partial deletions or rearrangements (Feng et al. 2002; Godiska et al. 2010). We have experienced such instability problems during targeting vector construction using pBluescript-based vector (Ohtsuka et al. 2007) and successfully resolved them by using either a specific E. coli host strain (e.g. SURE2 [Stratagene]) or the pBR322-based retrieval vector recently developed by us (unpublished data). A similar modified pBR322 plasmid (PL611) has been reported as a retrieval vector for similar reasons (Chan et al. 2007).

Since GOI regions in pAWK, pAWV and pAXV do not contain genomic fragments, these plasmids are probably free from the above-mentioned instability problem. However, we recently identified that the low-copy donor vector is useful for cloning the genomic fragment. For example, when we cloned the mouse nephrin promoter in pBluescript II KS(+) using the DH5α strain, these recombinants were labile. However, when the improved donor vector was used for cloning, the recombinants could be successfully maintained without any instability problem (data not shown).

1. Use of in vitro fertilized eggs

In this study, we used in vitro fertilized eggs for pronuclear injection and demonstrated that IVF is a useful approach for generating PITT mice using our seed mouse strain. Compared with natural mating, IVF is beneficial because the number of male mice used can be greatly reduced. For example, only one male with the targeted allele was required for IVF in each injection, while ten to twenty male mice were always required to obtain fertilized eggs by natural mating. Therefore, use of in vitro fertilized eggs is beneficial in terms of cost performance and the reduction in space and effort required to maintain the mice.

Based on the results obtained in our previous study (natural mating) and current study (IVF), we compared the efficiency of both strategies, as shown in the following table. The 95% CI, calculated with the binomial distribution by using Stata version 11.0, is shown in parentheses.

Methods for obtaining fertilized eggs / No. embryos transferred/no. eggs injected (%) / No. pups obtained/no. embryos transferred (%) / No. PITT mice/no. pups obtained (%) / No. PITT mice/no. eggs injected (%)
Natural mating
(Ohtsuka et al. 2010) / 55.1
(53.3–56.8) / 27.4
(25.4–29.6) / 4.3
(2.7–6.5) / 0.6
(0.4–1.0)
IVF
(this study) / 58.7
(55.4–62.0) / 17.6
(14.5–21.1) / 5.4
(1.8–12.1) / 0.6
(0.2–1.3)

The rate of ‘no.pups obtained/no. embryos transferred’ appears to be low when in vitro fertilized eggs were used (statistically significant; p<0.01). This may reflect the property of in vitro fertilized eggs because they generally contain some portions that potentially exhibit developmental abnormalities. Because PITT efficiency (‘no. PITT mice/no. pups obtained’) and the rate of ‘no. PITT mice/no. eggs injected’ were comparable between both strategies, we concluded that PITT mice can be produced from in vitro fertilized eggs at a level similar to those from eggs derived by natural mating. Notably, in our laboratory, one person was able to perform PITT with nearly 300 in vitro fertilized eggs in a day.

In this study, we used live spermatozoa for IVF. In addition, the use of frozen-thawed spermatozoa to obtain in vitro fertilized eggs is promising because IVF can be performed in the absence of any available males. This strategy is also beneficial for researchers because the number of males to be maintained in good condition can be greatly reduced, and PITT mice can be produced in external laboratories if the frozen spermatozoa are transferred and thawed in that laboratory.

Supplementary references

Buchholz F, Angrand PO, Stewart AF (1998) Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat Biotechnol 16:657-662

Chan W, Costantino N, Li R, Lee SC, Su Q, Melvin D, Court DL, Liu P (2007) A recombineering based approach for high-throughput conditional knockout targeting vector construction. Nucleic Acids Res 35:e64

Feng T, Li Z, Jiang W, Breyer B, Zhou L, Cheng H, Haydon RC, Ishikawa A, Joudeh MA, He TC (2002) Increased efficiency of cloning large DNA fragments using a lower copy number plasmid. Biotechniques 32:992, 994, 996 passim

Godiska R, Mead D, Dhodda V, Wu C, Hochstein R, Karsi A, Usdin K, Entezam A, Ravin N (2010) Linear plasmid vector for cloning of repetitive or unstable sequences in Escherichia coli. Nucleic Acids Res 38:e88

Kanki H, Suzuki H, Itohara S (2006) High-efficiency CAG-FLPe deleter mice in C57BL/6J background. Exp Anim 55:137-141

Lee G, Saito I (1998) Role of nucleotide sequences of loxP spacer region in Cre-mediated recombination. Gene 216:55-65

Nagy A, Gertsenstein M, Vintersten K, Behringer R (2003) Manipulating the Mouse Embryo: A Laboratory Manual. 3rd edn. Cold Spring Harbor Laboratory Press, New York

Ohtsuka M, Mizutani A, Kikuti YY, Kulski JK, Sato M, Kimura M, Tanaka M, Inoko H (2007) One-step generation of recombineering constructs by asymmetric-end ligation and negative selection. Anal Biochem 360:306-308

Ohtsuka M, Ogiwara S, Miura H, Mizutani A, Warita T, Sato M, Imai K, Hozumi K, Sato T, Tanaka M, Kimura M, Inoko H (2010) Pronuclear injection-based mouse targeted transgenesis for reproducible and highly efficient transgene expression. Nucleic Acids Res 38:e198

Sato M, Yasuoka Y, Kodama H, Watanabe T, Miyazaki JI, Kimura M (2000) New approach to cell lineage analysis in mammals using the Cre-loxP system. Mol Reprod Dev 56:34-44

Schaft J, Ashery-Padan R, van der Hoeven F, Gruss P, Stewart AF (2001) Efficient FLP recombination in mouse ES cells and oocytes. Genesis 31:6-10

Thomson JG, Rucker EB, 3rd, Piedrahita JA (2003) Mutational analysis of loxP sites for efficient Cre-mediated insertion into genomic DNA. Genesis 36:162-167