Phenotypically silent Cre recombination within the postnatal ventricular conduction system
Nikhil Vilas Munshi
The cardiac conduction system (CCS) is composed of specialized cardiomyocytes that initiate and maintain cardiac rhythm. Any perturbation to the normal sequence of electrical events within the heart can result in cardiac arrhythmias. To understand how cardiac rhythm is established at the molecular level, several genetically modified mouse lines expressing Cre recombinase within specific CCS compartments have been created. In general, Cre driver lines have been generated either by homologous recombination of Cre into an endogenous locus or Cre expression driven by a randomly inserted transgene. However, haploinsufficiency of the endogenous gene compromises the former approach, while position effects negatively impact the latter. To address these limitations, we generated a Cre driver line for the ventricular conduction system (VCS) that preserves endogenous gene expression by targeting the Contactin2 (Cntn2) 3’ untranslated region (3’UTR). Here we show that Cntn23’UTR-IRES-Cre-EGFP/+ mice recombine floxed alleles within the VCS and that Cre expression faithfully recapitulates the spatial distribution of Cntn2 within the heart. We further demonstrate that Cre expression initiates after birth with preservation of native Cntn2 protein. Finally, we show that Cntn23’UTR-IRES-Cre-EGFP/+ mice maintain normal cardiac mechanical and electrical function. Taken together, our results establish a novel VCS-specific Cre driver line without the adverse consequences of haploinsufficiency or position effects. We expect that our new mouse line will add to the accumulating toolkit of CCS-specific mouse reagents and aid characterization of the cell-autonomous molecular circuitry that drives VCS maintenance and function.We constructed a targeting vector containing an IRES-Cre-EGFP-FRT-neo-PGK-FRT KI cassette flanked by Homology Arms (HA) to ensure efficient Homologous Recombination (HR) within the 3’ UTR of the endogenous mouse Cntn2 locus (details of the strategy shown in Fig 1A). The fidelity of the targeting construct was verified prior to introduction into Embryonic Stem (ES) cells by direct Sanger sequencing. Genetically modified mice were generated by standard methodology (Cyagen Biosiences Inc., Santa Clara, CA). Briefly, the targeting vector was electroporated into ES cells. The ES cell clones with correct HR were selected for by Neomycin (G418) resistance (Neor) and screened by Polymerase Chain Reaction (PCR). ES cells were expanded, and injected blastocysts were implanted into pseudo-pregnant mice. In order to identify F1 mice with the recombined allele, standard PCR screening was used with F1-R1 and F2-R2 primer sets designed to identify the constitutive KI allele as shown in Fig 1A. The final PCR screening to verify Neor deletion was performed using primers F3-R3 designed as shown in Fig 1A. Following generation of genetically-modified mice, the 5’ and 3’ targeting construct insertion sites were re-verified by direct Sanger sequencing (S1 Fig). Sequences of all genotyping primers are provided in S1 Table.The Cntn2 gene (OMIM ID: 190197) contains 23 exons spanning approximately 40 kilobases (kb) with the ATG start codon in exon 2 and the TGA stop codon in exon 23 . The mCntn2 gene (GenBank accession number NM_177129.5; Ensembl ID ENSMUSG00000053024) is located on mouse chromosome 1. In order to generate a highly specific and sensitive mouse model to facilitate studies of the VCS, the Cntn2 locus was targeted for Cre insertion. The targeting vector was designed to have 5’ and 3’ Homology Arms (HA) with an IRES-Cre-EGFP-FRT-neo-PGK-FRT KI cassette to ensure proper HR into the 3’ UTR of the Cntn2 locus. This targeting strategy ensures unperturbed bi-cistronic expression of endogenous Cntn2 protein and a Cre-EGFP fusion protein under the control of the endogenous regulatory elements upon FLP recombination (Fig 1A). We used standard Polymerase Chain Reaction (PCR) screening to verify the following: 1) 5’ recombination using primer set F1-R1 designed to amplify a 272 bp fragment from the Cntn2 locus and the 5’ end of the KI cassette, 2) 3’ recombination using primer set F2-R2 designed to amplify a 535 bp product from the Cntn2 locus and the 3’ end of the KI cassette, 3) neo-PGK cassette removal with primer sets F3-R3 designed such that a band of 361bp is detected only upon neo deletion by FLP-mediated recombination, and 4) Cre sequence with primer sets F4-R4 that amplify a 408bp fragment within the coding sequence (Fig 1B and 1C). Furthermore, appropriate genomic targeting and fidelity of the recombined 5’ and 3’ junctions were confirmed by high-fidelity PCR coupled with direct Sanger sequencing of the KI cassette boundaries as shown in S1 Fig.To characterize the expression pattern of the KI-Cre allele, Cntn23’UTR-IRES-Cre-EGFP/+ mice were bred with the R26RtdTomato/tdTomato reporter strain. Cre dependent tdTomato expression was monitored at different stages of murine development. Consistent with Cntn2 expression in the central nervous system [27, 28], we observed robust fluorescent reporter expression in the brain at P28 (Fig 2A). In addition, we noted native tdTomato fluorescence in the heart at P28. Using epifluorescence imaging on freshly isolated P28 Cntn23’UTR-IRES-Cre-EGFP/+; R26RtdTomato/+ hearts (n = 15), we visualized reporter expression on the dorsal surface of the heart at the Atrio-Ventricular Junction (AVJ) upon Cre-mediated recombination of the R26R locus (Fig 2B). The same heart was further sectioned grossly to permit detailed visualization of tdTomato in specific regions and structures within the heart. The cells of the VCS exhibited bright and specific fluorescent labelling (Fig 2C and 2C’). Higher magnification imaging of the four chambers of an adult Cntn23’UTR-IRES-Cre-EGFP/+; R26RtdTomato/+ heart allowed appreciation of even finer and more complex structures of the AVB, right and left BB, and PF network (Fig 2D and 2E). To ensure that reporter expression was Cre-dependent, we sectioned multiple adult Cntn2+/+; R26RtdTomato/+ (hereafter referred to as wild-type [WT]) animals and found no evidence of tdTomato expression (see S4, S6C and S6D and S7 Figs). The litter size and sex-distribution also followed the expected Mendelian ratio (data not shown), suggesting that Cntn2 KI-allele does not have any major phenotypic consequences. Overall, we established a highly specific KI reporter mouse that labels key constituents of the VCS, namely the AVB-BB-PF system.Representative Sanger sequencing results confirm appropriate targeting of the Cntn2 locus and the fidelity of the (a) 5’ and (b) 3’ KI cassette boundaries(TIF)We thank John Shelton in the UT Southwestern Medical Center Histology Core Facility for expert technical assistance, Lin Wang for help with cryosections and mouse functional studies, and Antonio Fernandez-Perez for assistance with confocal imaging. N.V.M. was funded by the NIH (K08-HL094699), Burroughs Wellcome Fund (#1009838), and the March of Dimes Foundation (#5-FY14-203).