Optimizing Nervous System-Specific Gene Targeting with Cre Driver Lines: Prevalence of Germline Recombination and Influencing Factors

Abstract

The Cre-loxP system is invaluable for spatial and temporal control of gene knockout, knockin, and reporter expression in the mouse nervous system. However, we report varying probabilities of unexpected germline recombination in distinct Cre driver lines designed for nervous system-specific recombination. Selective maternal or paternal germline recombination is showcased with sample Cre lines. Collated data reveal germline recombination in over half of 64 commonly used Cre driver lines, in most cases with a parental sex bias related to Cre expression in sperm or oocytes. Slight differences among Cre driver lines utilizing common transcriptional control elements affect germline recombination rates. Specific target loci demonstrated differential recombination; thus, reporters are not reliable proxies for another locus of interest. Similar principles apply to other recombinase systems and other genetically targeted organisms. We hereby draw attention to the prevalence of germline recombination and provide guidelines to inform future research for the neuroscience and broader molecular genetics communities.

Keywords: Cre-lox; conditional gene targeting; conditional knockin; conditional knockout; conditional reporter; germline recombination; molecular genetics; mosaic recombination; parental sex bias; site-specific recombinase.

Figure 1.. Dlx5/6-Cre Mice Show Maternal Germline Recombination

(A) Pedigree and sample genotyping results of Clstn3^f/f crosses with Dlx5/6-Cre transmitted from either the male or female parent. F/F Cre⁺ indicates Clstn3^f/f; Dlx5/6-Cre. F/F Cre^– indicates Clstn3^f/f without Cre. F/–Cre⁺ indicates Clstn3^f/–; Dlx5/6-Cre in which recombination has occurred (this labeling is used for simplicity, but some of these mice may be F/–Cre⁺ and some may be F/F Cre⁺ genotypes because mosaic recombination was observed in tail tissue used for genotyping). F/–Cre^– indicates Clstn3^f/– without Cre in which recombination has occurred. P and N indicate controls (multiple mice were used for P). The numbers below genotypes indicate the total number of offspring obtained with that genotype. (B) Representative genotyping results and numbers of offspring from Ai32 Gt(ROSA)26Sortm32(CAG-COP4*H134R/EYFP)Hze/J reporter and Dlx5/6-Cre crosses. LSL-tg indicates the CAG promoter and lox-stop-lox sequences before the transgene channelrhodopsin-2(H134R)-EYFP (CHR2-EYFP) on the Rosa locus. L-tg indicates the transgene after germline recombination by Cre resulting in global transgene expression. Genotypes of the animals are 1, +/+ and Cre^–; 2 and 3, LSL-tg/+ and Cre^–; 4, L-tg/+ and Cre⁺; 5, LSL-tg/+ and Cre⁺ (showing some mosaic recombination); 6, L-tg/+ and Cre^–; 7, +/+ and Cre⁺. (C–E) Tiled images of the hippocampus (C) and sample CA1 and CA3 regions (D) showing CHR2-EYFP transgene expression and DAPI nuclear stain for selected offspring from (B). (E) Higher-magnification images are shown for hippocampal CA1 stratum pyramidale (s.p.), CA3 stratum lucidum (s.l.), dentate gyrus molecular layer (DG m.l.), and dentate gyrus hilus regions. Note that the laser power used for the EYFP channel in the far right panel for images from Rosa26^L-tg/+;Dlx5/6-Cremice was only 15% of that for the rest. Scale bars, 500 mm (C), 100 mm (D), and 20 mm (E).

Figure 2.. Gpr26-Cre Mice Show Paternal Germline Recombination

(A) Pedigree and sample genotyping results of Clstn3^f/f crosses with Gpr26-Cre from either the male or female parent. F/F Cre⁺ indicates Clstn3^f/f;Gpr26-Cre. F/F Cre^– indicates Clstn3^f/f without Cre. F/Cre⁺ indicates Clstn3^f/–;Gpr26-Cre in which recombination has occurred. This was confirmed to be germline deletion by the absence of a KO band in tail tissue from Clstn3^f/f;Gpr26-Cre mice (n = 46 mice generated from maternal Cre crosses) indicating the absence of local recombination in the tissue used for genotyping and by transmission of the KO allele to offspring. F/–Cre^– indicates Clstn3^f/– without Cre in which recombination has occurred. P and N indicate controls (multiple mice were used for P). (B) Representative genotyping results and numbers of offspring from Ai32 Gt(ROSA)26Sortm32(CAG-COP4*H134R/EYFP)Hze/J reporter and Gpr26-Cre crosses. LSL-tg indicates the CAG promoter and lox-stop-lox sequences before the transgene channelrhodopsin-2(H134R)-EYFP (CHR2-EYFP) on the Rosa locus. L-tg indicates the transgene after germline recombination by Cre resulting in global transgene expression. Genotypes of the animals are 1, LSL-tg/+ and Cre^–; 2, +/+ and Cre⁺; 3 and 4, L-tg/+ and Cre^–; 5, LSL-tg/+ and Cre⁺; 6, +/+ and Cre^–. (C–E) Tiled images of the hippocampus and cortex (C) and sample CA1 and CA3 regions (D) showing CHR2-EYFP transgene expression and DAPI nuclear stain for selected offspring from (B). Higher-magnification images are shown for hippocampal CA1 stratum pyramidale (s.p.) and CA3 stratum lucidum (s.l.) (E). Note that the laser power used for the EYFP channel in the far right panel for images from Rosa^tg/+ mice was only 15% of that for the rest. Scale bars, 500 mm (C), 100 mm (D), and 20 mm (E).

Figure 3.. Differential Recombination at Two Target Loci

Breeding scheme and genotyping result from tail tissue for a litter from crossing female Clstn3^f/+;Rosa^LSL-tg/+; Dlx5/6-Cre with male WT mice. In one offspring (*), ubiquitous recombination happened at the Clstn3^f/+ locus but not at the Ai32 Rosa^LSL-tg/+ locus implying differential activity of Cre at these target loci in the female germ cells. This mouse exhibited mosaic deletion in tail tissue at the Rosa locus as indicated by the presence of WT, LSL, and Rec PCR bands (see Figure 1B for a diagram).

Figure 4.. Breeding and Genotyping Strategies for Conditional KO/KI Mice

(A) A recommended breeding scheme is outlined. Target^Δ indicates a target allele that has undergone recombination in male or female germline cells (red) or more rarely in zygotes (brown); thus, Target^– or Target^KI. Target^f/+ instead of Target^f/f mice can be used for the F0 cross, reducing the frequency of generating Cre⁺; Target^f/+ mice for the F1 cross. Routine use of F2 crosses to generate experimental mice is recommended to minimize required animal numbers, but F1 crosses can also be used. It is recommended that F1 crosses using both male and female Cre⁺; Target^f/+ mice be established and the resultant germline recombination rates be tracked in offspring. Then male or female Cre⁺; Target^f/f mice can be used for the F2 crosses, depending on which sex gave the lowest germline recombination rate in the F1 crosses. It is important that Cre⁺; Target^f/f mice (green experimental mice) be validated by immunostaining or in situ hybridization for the target protein/RNA in the region of interest to confirm consistent recombination in the expected cell type. Cre^–; Target^f/f mice can be used as controls; separate breeding of congenic Cre⁺; Target^+/+ and WT controls is also recommended. “Cre⁺” refers to mice with one allele of the Cre transgene. In these recommended breeding schemes, Cre⁺ mice are not bred to Cre⁺ mice as this would result in a subset of offspring have 2 alleles of the Cre driver gene. This scenario can be problematic. For random insertion transgenic Cre drivers, it is typically not possible to differentiate among mice with one or two Cre driver alleles by PCR genotyping, leading to unknown variation in Cre expression levels upon subsequent breeding of these mice (which could result in further variability in germline recombination rates). For KI Cre driver lines, it is generally possible to differentiate among mice with one or two Cre driver alleles. However, homozygous insertion of the Cre driver may result in deleterious effects not seen with heterozygous Cre drivers, due to possible disruption of the native gene at the random or targeted insertion site. An exception may apply to targeted insertion Cre driver lines shown to have normal native gene expression; then, if one wanted to maximize Cre expression level, one might breed Cre⁺ with Cre⁺ mice and select those with 2 Cre alleles for further breeding.(B and C) Recommended genotyping strategies are diagrammed for conditional KO and KI mice, assuming a mini-gene strategy was used for conditional KI. Genotyping should also be done for the presence of the Cre driver gene (as in Figures 1 and 2, not shown here). The black PCR bands are diagnostic, and the gray bands are additional heteroduplexes that may appear. Potential PCR products that are too large to be generated under typical conditions are not diagrammed here, but these may be generated under some conditions (B with a+c primers for WT and Flox alleles, and C with a+b primers for Flox allele). For mice with one target allele, the presence of Flox, WT, and KO/KI bands indicates the occurrence of local recombination in the tissue used for genotyping rather than ubiquitous germline recombination (Target^f/+*). For mice with two target alleles, the presence of Flox and KO/KI bands indicates either germline recombination (Target^f/– or Target^f/KI) or local recombination in the tissue used for genotyping (Target^f/f*). The additional absence of a Cre driver identifies such mice to be Target^f/– or Target^f/KI, but such Cre⁺ mice would have to be bred further, or local recombination in genotyping tissue ruled out, to determine whether the recombination is germline.

Figure 5.. Breeding and Genotyping Strategy for Conditional Reporter Mice

(A) For conditional reporter mice, it is simplest to breed F0 mice and study F1 Cre⁺;Locus^LSL-tg/+ mice. Thus, the Cre driver gene and target locus are not together in the germline so unwanted global recombination could only occur by recombination in the zygote, which is not as common as in germline cells. It is important that Cre⁺;Locus^LSL-tg/+ mice (green, experimental mice) be validated by immunostaining or in situ hybridization for the transgene protein/RNA in the region of interest to confirm consistent recombination in the expected cell type. The optional F1 breeding scheme could be used to increase reporter expression level in Cre⁺; Locus ^{LSL-tg/LSL-tg} mice, but this also results in possible germline recombination. If F1 crosses are performed, both male and female Cre⁺; Locus^LSL-tg/+ mice should be used initially to track resultant germline recombination rates so that the sex resulting in the lowest germline recombination rate can be used in further F1 crosses. Locus^LSL-tg+ indicates a lox-stop-lox-transgene cassette that expresses the transgene upon Cre-mediated recombination, but our recommendation applies to other Cre-dependent loci such as those using a flip excision or double-inverted orientation mechanism. Locus^L-tg+ indicates a globally recombined locus resulting from recombination in male or female germline cells (red) or more rarely in the zygote (brown). “Cre⁺” refers to mice with one allele of the Cre transgene (see Figure 4 legend). (B) A recommended genotyping strategy is diagrammed for conditional reporter mice. Only the first four lanes depicting PCR bands are relevant to F1 mice in the above breeding scheme. Genotyping should also be done for the presence of the Cre driver gene (as in Figures 1 and 2, not shown here). Potential PCR products that are too large to be generated under typical conditions are not diagrammed here, but these may be generated under some conditions (with a+a’ primers for LSL-tg and L-tg alleles, and d+b’ primers for the LSL-tg allele). For mice with one target allele, the presence of WT, LSL, Tg, and Rec bands indicates the occurrence of local recombination in the tissue used for genotyping rather than ubiquitous germline recombination (Locus^LSL-tg/+*). For mice with two target alleles, the presence of LSL, Tg, and Rec bands indicates either germline recombination (Locus^LSL-tg/L-tg) or local recombination in the tissue used for genotyping (Locus^{LSL-tg/LSL-tg}*). The additional absence of a Cre driver identifies such mice to be Locus^LSL-tg/L-tg, but such Cre⁺ mice would have to be bred further or local recombination in genotyping tissue ruled out to determine whether the recombination is germline.