WO2020152163A1 - Improved cre/lox dna construct - Google Patents

Abstract

The present invention refers to an improved DNA construct, based on Cre/lox technology, which confers an improved way of studying gene functions by means deletions, inversions and/or translocations generated by this DNA construct at specific lox sites inserted in the genome of any type of cell.

Description

IMPROVED CRE/LOX DNA CONSTRUCT

FIELD OF THE INVENTION

The present invention pertains to the field of biomedical science, particularly to the genetic engineering field. Specifically, it refers to an improved DNA construct, based on Cre/Lox technology, which confers a more efficient and reliable way of studying gene function by means deletions, inversions and/or translocations generated by this improved DNA construct at specific sites of the genome.

PRIOR ART

Cre Lox system enables the recombination of a pair of short (34bp) DNA sequences called Lox sequences by the recombinase encoded by the bacteriophage PI gene Cre. This technology has revolutionized biomedical science since it can be used to carry out genetic deletions, inversions or translocations at specific engineered sites in the genome. Today, most mouse genes have been flanked by LoxP sites (floxed), and mouse lines are available expressing constitutively active Cre or tamoxifen-inducible CreERT2 in almost any cell type. The availability of these genetic resources enables the precise conditional deletion of almost any mouse gene in any cell type and at a defined time-point, which is crucial for understanding the function of genes during organ development and disease. However, numerous studies demonstrate the need for caution in the use of this powerful technology. Many transgenic mouse strains expressing Cre or CreERT2, when interbred with other strains containing floxed alleles, produce highly variable expression patterns and phenotypes that are not apparent when the mice were first created. This is partly due to the inconsistent nature or epigenetic silencing of the promoters used in Cre or CreERT2 transgenic mouse lines. Two other important issues with Cre/Lox technology are the variable recombination or gene deletion efficiency and the methods used for its detection. The efficiency of Cre recombination depends on the location of the LoxP sites in the genome and the distance between them. Thus, one should expect different a recombination efficiency for every floxed allele, even when it is deleted with the same Cre- or CreERT2-expressing mouse line. It is therefore critical to have a reliable method to confirm that a given gene is properly deleted, and only in the desired cell type. Fast PCR-based methods are frequently used to confirm genetic deletions in whole tissues or groups of isolated cells, but these methods are insufficient because they do not provide in situ and single-cell resolution. Moreover, these methods only indicate the average gene-deletion efficiency, and cannot quantify the heterogeneity in genetic deletion efficiency among cells. The safest method for confirming inducible and specific gene deletion is to co-immunostain for the encoded protein and a tissue or cell marker. However, for many proteins there are no antibodies able to distinguish between the morphology of cells with and without protein expression in the tissue. Another issue is that gene transcription and protein stability oscillate in a cell, and thus a cell with no detectable expression of a given protein at a given moment may still be wildtype for the coding gene.

To overcome some of these technical problems with the Cre/Lox system, scientists have generated reporters of Cre/LoxP recombination, and these have become widespread and essential genetic tools in any laboratory performing genetic studies. These reporters are usually alleles targeted to the ubiquitous mouse ROSA26 locus and are activated or expressed only after the cell expresses Cre or has induced CreERT2 activity. However, since there is no genetic linkage between the reporter allele and other floxed alleles in the cell, it is unsafe to assume correlation between recombination of the reporter and the target allele. Indeed, several studies have highlighted the unreliability of Cre/LoxP recombination reporters, reporting discrepancies between multiple reporter alleles and target allele recombination.

Being able to rely on results arising from the use of Cre/LoxP technology is essential for the progress of biomedical science and to effectively reduce the number of animals and experiments required in the biomedical and biotechnology sectors. On the other hand, understanding gene function in physiology and pathology is a key to understand basic mechanisms of disease and to target them with existent or new pharmacological compounds. Consequently, the present invention is focused on solving the above cited problems and it is directed to the generation of an improved inducible reporter-Cre mouse line, called iSuRe- Cre. This mouse line is compatible with all existent Cre/CreERT2 and floxed mouse lines and significantly increases the efficiency and reliability of inducible and Cre-dependent gene function analysis.

DESCRIPTION OF THE INVENTION

Brief description of the invention The inventors of the present invention have developed an improved DNA construct (hereinafter iSuRe-Cre or the construct of the invention) based on Cre/Lox technology, with the goal of overcoming most of the drawbacks associated to above cited Cre/Lox technology. iSuRe-Cre is a new genetic tool to reliably induce and report Cre-dependent genetic modifications (Figure 1A).

The first embodiment of the present invention refers to a DNA construct (Figure 2A) characterized by comprising in the following order: a) a promoter, b) a LoxP-flanked DNA sequence, this DNA sequence coding for a first reporter of transgene promoter activity and a transcription stop poly(A) signal, c) a DNA sequence coding for a second reporter of transgene recombination and Cre expression, d) a DNA sequence coding for the viral 2A peptide, and e) a DNA sequence coding for Cre protein comprising an intron within its sequence.

In a preferred embodiment, Cre protein is constitutively active and permanently expressed once a pre-induced Cre-activity removes the LoxP-flanked DNA sequence.

In a preferred embodiment, Cre protein has an addition of the aminoacids proline, glutamic acid and phenylalanine in the N-terminal region.

In a preferred embodiment, the first reporter is the non-fluorescent reporter N-PhiM.

In a preferred embodiment, the second reporter is the fluorescent reporter MbTomato.

In a preferred embodiment, the promoter is the synthetic CAG promoter.

In a preferred embodiment the LoxP nucleotide sequence is the SEQ ID NO: 1 :

AT A AC T T C GT AT AGC AT ACATTATAC GA AGTT AT

In a preferred embodiment the N-PhiM nucleotide sequence is the SEQ ID NO: 2: AT GGCTCCTAAGAAGAAGAGGAAGGT GAT GAGC AGCGGCGCCCT GCT GTT CC AC

GGCAAGATCCCCTACGTGGTGGAGATGGAGGGCAATGTGGATGGCCACACCTTC

AGCATCCGCGGCAAGGGCTACGGCGATGCCAGCGTGGGCAAGGTGGATGCCCAG

TTCATCTGCACCACAGGCGATGTGCCCGTGCCCTGGAGCACCCTGGTGACCACCC

TGACCGCAGGCGCCCAGTGCTTCGCCAAGTACGGCCCCGAGCTGAAGGATTTCT

ACAAGAGCTGCATGCCCGATGGCTACGTGCAGGAGCGCACCATCACCTTCGAGG

GTGATGGCAATTTCAAGACCCGCGCCGAGGTGACCTTCGAGAATGGCAGCGTGT

ACAATCGCGTGAAGCTGAATGGCCAGGGCTTCAAGAAGGATGGCCACGTGCTGG

GCAAGAATCTGGAGTTCAATTTCACCCCCCACTGCCTGTACATCTGGGGCGATCA

GGCCAATCACGGCCTGAAGAGCGCCTTCAAGATCTGCCACGAGATCACCGGCAG

CAAGGGCGATTTCATCGTGGCCGATCACACCCAGATGAATACCCCCATCGGCGG

CGGCCCCGTGCACGTGCCCGAGTACCACCACATGAGCTACCACGTGAAGCTGAG

CAAGGATGTGACCGATCACCGCGATAATATGAGCCTGAAGGAGACCGTGCGCGC

CGTGGATTGCCGCAAGACCTACCTGTGA

In a preferred embodiment pA nucleotide sequence is Sv40pA of sequence is the SEQ ID

NO: 3 :

AGATCATAATCAGCCATACCACATTTGTAGAGGTTTTACTTGCTTTAAAAAACCT

CCCACACCTCCCCCTGAACCTGAAACATAAAATGAATGCAATTGTTGTTGTTAAC

TTGTTT ATT GC AGCTTATAATGGTTAC AAAT AAAGC AAT AGCATCAC AAATTT CA

CAAATAAAGCATTTTTTTCACTGCATTCTAGTTGTGGTTTGTCCAAACTCATCAAT

GTATCTTAAGGCG

In a preferred embodiment the nucleotide sequence of MbTomato reporter is the SEQ ID NO: 4:

ATGGGTTGCTGTTTCTCCAAGACCATGGTGAGCAAGGGAGAGGAGGTCATCAAA

GAGTTCATGCGCTTCAAGGTGCGCATGGAGGGCTCCATGAACGGCCACGAGTTC

GAGATCGAGGGCGAGGGCGAGGGCCGCCCCTACGAGGGCACCCAGACCGCCAA

GCTGAAGGTGACCAAGGGCGGCCCCCTGCCCTTCGCCTGGGACATCCTGTCCCCC

CAGTTCATGTACGGCTCCAAGGCGTACGTGAAGCACCCCGCCGACATCCCCGATT

ACAAGAAGCTGTCCTTCCCCGAGGGCTTCAAGTGGGAGCGCGTGATGAACTTCG

AGGACGGCGGTCTGGTGACCGTGACCCAGGACTCCTCCCTGCAGGACGGCACGC

TGATCTACAAGGTGAAGATGCGCGGCACCAACTTCCCCCCCGACGGCCCCGTAA

TGCAGAAGAAGACCATGGGCTGGGAGGCCTCCACCGAGCGCCTGTACCCCCGCG ACGGCGTGCTGAAGGGCGAGATCCACCAGGCCCTGAAGCTGAAGGACGGCGGCC

ACTACCTGGTGGAGTTCAAGACCATCTACATGGCCAAGAAGCCCGTGCAACTGC

CCGGCTACTACTACGTGGACACCAAGCTGGACATCACCTCCCACAACGAGGACT

ACACCATCGTGGAACAGTACGAGCGCTCCGAGGGCCGCCACCACCTGTTCCTGG

GGCATGGCACCGGCAGCACCGGCAGCGGCAGCTCCGGCACCGCCTCCTCCGAGG

ACAACAACATGGCCGTCATCAAAGAGTTCATGCGCTTCAAGGTGCGCATGGAGG

GCTCCATGAACGGCCACGAGTTCGAGATCGAGGGCGAGGGCGAGGGCCGCCCCT

ACGAGGGCACCCAGACCGCCAAGCTGAAGGTGACCAAGGGCGGCCCCCTGCCCT

TCGCCTGGGACATCCTGTCCCCCCAGTTCATGTACGGCTCCAAGGCGTACGTGAA

GCACCCCGCCGACATCCCCGATTACAAGAAGCTGTCCTTCCCCGAGGGCTTCAAG

TGGGAGCGCGTGATGAACTTCGAGGACGGCGGTCTGGTGACCGTGACCCAGGAC

TCCTCCCTGCAGGACGGCACGCTGATCTACAAGGTGAAGATGCGCGGCACCAAC

TTCCCCCCCGACGGCCCCGTAATGCAGAAGAAGACCATGGGCTGGGAGGCCTCC

ACCGAGCGCCTGTACCCCCGCGACGGCGTGCTGAAGGGCGAGATCCACCAGGCC

CTGAAGCTGAAGGACGGCGGCCGCTACCTGGTGGAGTTCAAGACCATCTACATG

GCCAAGAAGCCCGTGCAACTGCCCGGCTACTACTACGTGGACACCAAGCTGGAC

ATCACCTCCCACAACGAGGACTACACCATCGTGGAACAGTACGAGCGCTCCGAG

GGCCGCCACCACCTGTTCCTGTACGGCATGGACGAGCTGTACAAG

In a preferred embodiment the nucleotide sequence coding for the Viral 2A peptide is the SEQ ID NO: 5:

GGCAGTGGAGAGGGCAGAGGAAGTCTGCTAACATGCGGTGACGTCGAGGAGAA

TCCTGGCCCA

In a preferred embodiment the DNA sequence coding for Cre protein comprising an intron within its sequence is the SEQ ID NO: 6 wherein the intron is underlined:

ATGTCCAATTTACTGACGGTGCACCAAAACTTGCCTGCATTACCGGTCGATGCAA

CGAGTGATGAGGTTCGCAAGAACCTGATGGACATGTTCAGGGATCGCCAGGCGT

TTTCTGAGCATACCTGGAAAATGCTTCTGTCCGTTTGCCGGTCGTGGGCGGCATG

GTGCAAGTTGAATAACCGGAAATGGTTTCCCGCAGAACCTGAAGATGTTCGCGA

TTATCTTCTATATCTGCAGGTAAGTATCAAGGTTACAAGACAGGTTTAAGGAGAC

C AAT AGAAAC TGGGCTT GTCGAGAC AGAGAAGACTC TT GCGTTT CT GAT AGGC A

CCTATTGGTCTTACTGACATCCACTTTGCCTTTCTCTCCACAGGCGCGCGGTCTGG CAGTAAAAACTATCCAGCAACATTTGGGCCAGCTAAACATGCTTCATCGTCGGTC

CGGGCTGCCACGACCAAGTGACAGCAATGCTGTTTCACTGGTTATGCGGCGGATC

CGAAAAGAAAACGTTGATGCCGGTGAACGTGCAAAACAGGCTCTAGCGTTCGAA

CGCACTGATTTCGACCAGGTTCGTTCACTCATGGAAAATAGCGATCGCTGCCAGG

ATATACGTAATCTGGCATTTCTGGGGATTGCTTATAACACCCTGTTACGTATAGC

CGAAATTGCCAGGATCAGGGTTAAAGATATCTCACGTACTGACGGTGGGAGAAT

GTTAATCCATATTGGCAGAACGAAAACGCTGGTTAGCACCGCAGGTGTAGAGAA

GGCACTTAGCCTGGGGGTAACTAAACTGGTCGAGCGATGGATTTCCGTCTCTGGT

GTAGCTGATGATCCGAATAACTACCTGTTTTGCCGGGTCAGAAAAAATGGTGTTG

CCGCGCCATCTGCCACCAGCCAGCTATCAACTCGCGCCCTGGAAGGGATTTTTGA

AGCAACTCATCGATTGATTTACGGCGCTAAGGATGACTCTGGTCAGAGATACCTG

GCCTGGTCTGGACACAGTGCCCGTGTCGGAGCCGCGCGAGATATGGCCCGCGCT

GGAGTTT C AAT ACC GGAGAT CAT GC AAGC T GGT GGC T GGAC C AAT GT A AAT ATT

GTCATGAACTATATCCGTAACCTGGATAGTGAAACAGGGGCAATGGTGCGCCTG

CTGGAAGATGGCGATTAG

The second embodiment of the present invention refers to a DNA vector comprising the above defined DNA construct.

The third embodiment of the present invention refers to cells comprising in their genome the above defined DNA construct or to cells transfected with the above mentioned DNA vector. In a preferred embodiment, the cells are embryonic stem cells.

The fourth embodiment of the present invention refers to a transgenic non-human organism comprising in their genome the above defined DNA construct or cells. In a preferred embodiment, the transgenic non-human organism is a mouse.

A preferred version of the DNA construct of the invention is shown in Figure 2A.

Particularly, the DNA construct of the invention comprises the ubiquitous and strong CAG promoter (Figure IB) followed by a very short (l . lkb), and therefore easy to recombine LoxP-flanked DNA sequence containing the N-PhiM reporter gene and a transcription stop signal (pA). The PhiM reporter allows confirming that the transgene promoter is active and expressed in all cells in the non-induced state. After recombination/deletion by Cre or CreERT2 of the LoxP-PhiM-pA-LoxP cassette, this transgenic construct enables the strong co-expression of a bright membrane-localized fluorescent reporter (MbTomato; Figure IF) and a constitutively active and permanently expressed Cre protein. These two proteins are separated by the viral 2 A peptide to guarantee equimolar expression and 100% correlation between the MbTomato reporter expression and high Cre activity. Cre protein (after 2A peptide and splicing) may have an addition of three amino acids (Pro, Glu and Phe) in the N- terminal region (due to the way the 2A peptide works and the way Cre was cloned downstream of it with an EcoRI site, GAATTC). The self-cleaving 2A peptide between MbTomato and Cre generates two independent proteins: MbTomato-2A + Pro-Glu-Phe - Cre.

It is important to note that, a construct like this with a normal Cre (instead of an intron- containing Cre sequence), cannot be generated in E. Coli (bacteria) due to self-recombination (Figure 1C, D). So, an important part of the invention was to clone an intron-containing Cre sequence downstream of a floxed cassette. Since splicing does not occur in prokaryotes, this intron-containing Cre sequence avoids the correct translation and activity of the Cre protein, circumventing thus the self-recombination of the DNA construct (Figure 1C, E).

With this strategy, the inventors of the present invention have overcome the limitations mentioned above for the Cre/Lox system and of all already existing Cre recombination reporters. Thus, the main advantages of the DNA construct of the invention are as follows:

• It is the first Cre-inducible, dual reporter-Cre expressing mouse line where it is possible to achieve a very high (in most cases close to 100%) correlation between reporter expression and single or multiple genes and floxed reporters deletion (Figure 3E-H and Fig. 4B-I and Fig. 5)

• It allows a significant increase in the deletion efficiency of any floxed gene, since it increases the efficiency of Cre-dependent genetic deletions. Most scientists in the biomedical field are moving towards tissue-specific CreERT2 inducible systems, in order to have temporal control in the genetic deletion/phenotype. The DNA construct of the invention is especially relevant in this situation, given that CreERT2 activity is usually weak, due to the transient and incomplete nature of the tamoxifen induction, and the consequent poor correlation between standard Cre-reporter expression and gene deletion (Figure 4).

• Simultaneous loss-of-function of multiple genes in the same cell or tissue is an essential technique to determine genetic redundancy, or if two or more genes functionally interact. The DNA construct of the invention is the first genetic tool that enables reliable multiple gene loss-of-function. Simultaneous deletion of at least 3 genes (6 alleles) has been demonstrated. Any combination of tested floxed genes has been efficiently deleted with the DNA construct of the invention (Figure 5).

• The iSuRe-Cre mouse line is compatible with all existent Cre/CreERT2 and gene floxed mouse lines.

• It provides high cellular resolution and allows direct imaging of mutant cells on live or fixed tissues.

• Given its insulator and strong promoter regulatory elements, the iSuRe-Cre transgene is ubiquitously expressed, being possible to induce it in any cell type or tissue of the mouse.

• The integration of this transgene in a defined position of Chromossome 17 (Fig. 6F) allows its combination with the wide range of tools that already exist targeted to the ROSA26 locus in chromosome 6.

The present invention reinforces the notion that the commonly used methods to induce and report gene loss-of-function in mice are often inefficient and unreliable. This is especially problematic for genes that are difficult to delete, or when they encode for proteins whose absence cannot be clearly detected in multicellular tissues. In conditional genetics, fluorescent Cre-reporter-expressing cells are usually considered mutant cells, even when a variable fraction of them may escape the desired gene deletion, especially if that gives them a competitive advantage (Figure 4e, 4f). Many Cre transgenes are sensitive to epigenetic regulation, which causes variegation or weak expression in a fraction of cells. The present invention shows several examples in which a weaker Cre or CreERT2 expression/activity may be enough to recombine a floxed reporter allele localized in the permissive Rosa26 locus, but not other genes, that need higher or long-lasting Cre expression levels, which can be provided by the Tg(iSuRe-Cre) allele (Figure 9 c,d). Multiple gene deletion in the same cell or tissue is frequently required to determine epistasis or functional genetic interactions. The use of the new Tg(iSuRe-Cre) allele will allow the analysis of phenotypes produced by multiple gene deletions, with high cellular resolution and genetic reliability in live or fixed tissues. Efficient genetic deletions can also be achieved in a mosaic fashion in single cells, by regulating the dose of the CreERT2 ligand tamoxifen, since gene deletion per MbTomato-2A-Int-Cre+ cell is highly efficient, even if the initial CreERT2-dependent recombination induction frequency is very low.

In summary, it is herein proposed that the Tg(iSuRe-Cre) mouse line should be used instead of other conventional Cre activity reporters, which the inventors of the present invention (in figure 3 and 4) have shown not to be bona fide reporters of the recombination/deletion of other genes. Moreover, unlike other binary FlpO/FlpOERT2-inducible Cre-expressing systems, the Tg(iSuRe-Cre) allele is not leaky in the male germline (Figure 3A-3C) and is fully compatible with the numerous existent LoxP and Cre/CreERT2 alleles. This new genetic tool will significantly increase the ease, efficiency and reliability of conditional genetic modifications in the mouse, the most widely used model organism in biomedical research.

With the purpose of the present invention the following terms are defined:

• The term "comprising" it is meant including, but not limited to, whatever follows the word "comprising". Thus, use of the term "comprising" indicates that the listed elements are required or mandatory, but that other elements are optional and may or may not be present.

• By "consisting of’ is meant including, and limited to, whatever follows the phrase “consisting of’. Thus, the phrase "consisting of’ indicates that the listed elements are required or mandatory, and that no other elements may be present.

Moreover, with the purpose of the present invention the following abbreviations are defined:

• CAG prom: enhancer/promoter driving construct expression.

• LoxP: short DNA sequences recognized and recombined by Cre.

• N-PhiM: nuclear localized and mutated PhiYFP.

• pA, Sv40: polyadenylation sequence. • Ascl: restriction enzyme site used to clone the inserts.

• MbTomato: fluorescent protein reporter.

• 2A: self-cleaved viral 2A peptide to guarantee equimolar expression but separate localization of the upstream and downstream proteins.

• Int-Cre: Intron-containing Cre sequence.

• WPRE: Woodchuck hepatitis vims posttranscriptional regulatory element.

• ROSA26 hom: ROSA26 locus homology arms.

• INS: chicken b-globin HS4 insulator sequence.

• FRT: short DNA sequences recognized by the recombinase Flp allowing deletion of the PGK-Neo selection cassette.

• P1-P4: primers used to screen or genotype the ES cells or mice (Embryonic stem cells of mice).

Description of the figures

Figure 1. Design and pre-validation of iSuRe-Cre plasmids, a, b) Cre or CreERT2 expression from transgenes carrying gene-specific promoters (i.e. Tie2 or Cdh5) is variable and often weak. In addition, in the case of CreERT2 expressing alleles, tamoxifen needs to be provided to induce activity, of only a fraction of available CreERT2 proteins, and in a limited temporal window. Due to the variability in Cre expression and induced activity, commonly used Cre activity reporter alleles do not guarantee full recombination of other floxed alleles in the same cell. The new iSuRe-Cre construct enables stronger (driven by the CAG promoter) and constitutive co-expression of a reporter and Cre, which is expected to increase the correlation between reporter expression and gene deletion c) Vector and DNA inserts ligated in E. coli. d) Vector containing the normal Cre sequence always had deletion of the N-PhiM- pA cassette in E. coli , rendering them non-inducible e) The use of an intron-containing Cre sequence (Int-Cre) disrupts its expression and function in E. Coli, allowing the generation of a Cre-inducible, Cre-expressing vector (iSuRe-Cre). f) Confocal micrographs showing that cells transfected with iSuRe-Cre plasmids express the N-PhiM protein in the nuclei. If cells are co-transfected with iSuRe-Cre and Cre expressing plasmids, the LoxP-N-PhiM-pA-LoxP cassette is recombined/deleted and MbTomato-2A-Int-Cre expressed. Figure 2. iSuRe-Cre DNA construct, ES cells, and mice, a) iSuRe-Cre DNA construct used to produce gene-targeted or transgenic ES cells and mice b) PCR result with primers detecting the integration of the vector in the ROSA26 locus of ES cell clones 1-18. c) PCR result with primers detecting the presence of the iSure-Cre allele in the genome of ES cells 1- 18. d) Representative images of different ROSA26-targeted (Gt(ROSA)26Sor) ES cell clones at baseline. N-PhiM is expressed if the allele is non-recombined, and MbTomato shows cells that had recombination of the construct without induction. Clone #6 has a higher proportion of non-recombined N-PhiM+ cells e) Representative images of different Tg(iSuRe-Cre) ES cell clones at baseline. Clone #18 has a higher proportion of non-recombined N-PhiM+ cells f) Mouse chimeras generated with Gt(ROSA)26Sor-iSuRe-Cre ES cell clone #6 were interbred with Rhp ^{ox iIox} animals and the progeny genotyped for the Rbpj and iSuRe-Cre alleles. All animals containing the iSuRe-Cre allele in the ROSA26 locus had deletion of the Rbpj- floxed allele and expressed the MbTomato reporter in all cells g) Mouse chimeras generated with Tg(iSuRe-Cre) ES cell clone #18 were interbred with Rbpj ^^ox/^°^x animals and the progeny genotyped for the Rbpj and iSuRe-Cre alleles. Animals containing the Tg(iSuRe- Cre) allele did not have deletion of the Rbpj floxed allele and did not express the MbTomato reporter. Instead they expressed the N-PhiM reporter due to the absence of Cre activity.

Figure 3. The Tg(iSuRe-Cre) allele is ubiquitously expressed and is a reliable reporter of the recombination of other reporter alleles, a) Analysis of Tg(iSuRe-Cre) expression in the absence of Cre activity reveals its expression (N-PhiM+) in most cells of mouse embryos. The allele does not self-recombine in embryos (MbTomato-negative). b, c) FACS analysis reveals that the Tg(iSuRe-Cre) allele does not self-recombine in blood, unlike its ROSA26 gene-targeted version d) Confocal micrograph of postnatal day (P) 6 retina vessels from animals with the genotype indicated beneath panels d-f and induced with tamoxifen from PI to P3. All endothelial cells (ECs; nuclei, ERG+) expressing MbTomato-2A-Int-Cre also recombined the reporter allele ROSA26^{lsl yfp}. e) Quantification of the different recombination events/reporters in retinal vessels with high and low frequencies of tamoxifen- induced CreERT2 recombination f) FACS analysis of liver ECs from tamoxifen-induced adult animals with the genotype indicated beneath panels d-f. All induced MbTomato+ cells also recombined the reporter allele ROSA26^{LSL YFP} . The MbTomato reporter from the Tg(iSuRe-Cre) allele is easier to separate from baseline autofluorescence g) Confocal micrograph of P6 retina vessels from animals with the genotype indicated above panels g-g" and induced with tamoxifen at P3, which resulted in a low recombination frequency. All MbTomato+ cells also recombined the reporter allele ROSA26^{lsl yfp} . In contrast, a large fraction of YFP+ cells did not recombine the Tg(iSuRe-Cre ) allele h, i) 6-channel confocal micrograph of P6 retina vessels from animals with the genotype indicated above the panels and induced with tamoxifen at P3. All cells expressing the MbTomato reporter recombined the two reporter alleles ( ROSA26^{LSL rFP} and ROSA26^{lChr2 Mosaic} ), resulting in expression of EYFP and one of the three possible nuclear-localized proteins (H2B-Cherry, H2B-EGFP, or HA-H2B-Cerulean). In contrast, only a small fraction of EYFP+ cells recombined and expressed this reporter in their chromatin/nuclei. Data in 3c and 3e are presented as mean ± StDev (no variability in 3c). *** pO.OOOl. Data in 3i are the mean frequencies obtained in the indicated cell groups.

Figure 4. The Tg(iSuRe-Cre) allele significantly increases the efficiency of inducible genetic modifications, a-c) Representative confocal micrographs of P6 retina vessels labeled with IsolectinB4 (endothelial surface) and anti-ERG antibody (endothelial nuclei), obtained from animals with the genotype indicated to the left and induced with high-dose tamoxifen from PI to P3. Images 1-8 represent the phenotypic variability of animals analysed with the same genotype; EC number for each image is depicted in chart c (yellow dots) d) Linear regression showing the high correlation (r²) between total Erg+ EC number (phenotype) and the number of Erg+/MbTomato- cells e, f) Representative confocal micrographs of P6 retina vessels from animals with the genotype indicated to the left and induced with high-dose tamoxifen from PI to P3. Images 9-10 represent the phenotypic variability of animals with the same genotype. A comparison of EC number and reporter expression frequency is depicted in chart f (green and red dots) g) Genotyping PCR showing the efficiency of Rbpj gene inducible deletion in the MbTomato- and MbTomato+ cells of several distinct organs from mice with ubiquitous expression of CreERT2 (Rosa26-CreERT2) and induced with tamoxifen h) PCR for the Notchl floxed allele and a control genomic sequence showing the efficiency of Notchl gene inducible deletion in MbTomato+ and MbTomato- cells of the liver of mice with the indicated genotype i) Epasl mRNA relative levels (qRT-PCR) in LysM-Cre reporter expressing bone marrow-derived macrophages. Data in 4c and 4i are presented as mean ± StDev. T-test (4c) or ANOVA (4i). ** p<0.001; *** p<0.0001.

Figure 5. The Tg(iSuRe-Cre ) allele enables multiple gene deletions in single cells or tissues, a) The schematics illustrate the D114 and Kdr floxed alleles, showing inter-LoxP-site genetic distance, which is significantly larger in the Kdr allele. All 4 alleles must be deleted to achieve full dual gene loss-of-function. Kdr and D114 proteins are expressed in most liver ECs (ERG+, nuclei) of Dll4^^ox/^^ox/Kdr^^ox/^^ox animals injected with tamoxifen on 3 consecutive days b) Adult mice carrying in addition the Tg(Cdh5-CreERT2) and Tg(iSuRe-Cre) alleles and treated with the same high-dose tamoxifen for 3 consecutive days show very pronounced deletion of D114, but not Kdr, in liver MbTomato /ERG+ ECs. However, MbTomato⁺ cells have complete deletion of both genes c) Quantification of the immunostaining signals for ERG, D114, Kdr, and MbTomato in large liver sections of the indicated animals d) Illustration of the Myc , Mycn and Rbpj floxed alleles showing the genetic distances between the LoxP sites e) Genotypes of control and mutant adult mice injected once with lmg of tamoxifen and used for the gene deletion quantification by PCR. Control PCR band provides a DNA input quantitative control for the PCR, since it corresponds to a wildtype genomic sequence, present in all DNA samples (see Table 2 for primer sequences). Animals containing the Tg(Cdh5-CreERT2) and Tg(iSuRe-Cre) alleles have deletion of the 6 floxed alleles only in FACS sorted MbTomato⁺ cells, as detected by semi-quantitative competitive PCR. Weak Mycn and Rbpj floxed bands in MbTomato+ sample PCR may result from contamination of this sample with MbTomato negative cells or their DNA during the FACS protocol, or incomplete gene deletion f) Quantification with Image I of the relative intensity of the floxed and control PCR gel bands shown in e), which provides an estimation of the degree of the indicated floxed gene deletion. Error bars represent StDev. ANOVA and multiple comparisons ** P < 0,01. NS, not significant.

Figure 6. The Tg(iSuRe-Cre) allele is not leaky, a, b) Comparison of the frequency of cells with YFP or MbTomato-2A-Int-Cre expression. Both LysM-Cre and Vavl-Cre lines recombine the ROSA26^LSL~YFP reporter in non-desired cell types, but not the Tg(iSuRe-Cre) allele c, d, e) The Tg(iSuRe-Cre) allele is not leaky and is only induced in the desired cell types/organs when interbred with different tissue-specific Cre expressing lines f) Illustration showing the genes and regulatory regions surrounding the transgene integration site g) Genomic DNA PCR detection of homozygous animals, found at the expected mendelian ratios (n=79). Error bars represent StDev.

Figure 7. Comparison of Cre expression levels in different tissues and mouse lines, a)

Western of lung lysates from mice with the indicated genotypes, showing Cre (35kDA) and CreERT2 (70kDa) expression differences in Tie2-Cre and Cdh5-CreERT2 mouse lines b) Western of FAC sorted endothelial cells (CD31-APC+) from the indicated mice, showing Cre (35kDA) and CreERT2 (70kDa) expression in wildtype, MbTomato+ and MbTomato- cells c) Western of FAC sorted bone marrow derived macrophages (BMDMs) obtained from the indicated mice d, e) Western of lung and liver lysates from mice with the indicated genotypes, showing the increase in Cre protein levels in tissues of animals with expression of the Tg(iSuRe-Cre) allele f) Western of heart lysates from mice with the indicated genotypes, showing no significant increase in Cre protein levels in tissues of MyHC-Cre +Tg(iSuRe- Cre) animals g) Western of heart lysates showing that animals with recombination of the Gt(Rosa)26Sor-iSuRe-Cre allele in the germline, and with ubiquitous iSuRe-Cre expression in their hearts (not only MyHC-Cre+ cardiomyocytes), have significantly lower Cre levels than MyHC-Cre+ hearts h-j) Metabolic studies in adult littermate mice (9 months) having or not ubiquitous expression of the Gt(Rosa)26Sor-iSuRe-Cre allele. Values in red below the blots indicate the fold change in Cre protein related with the Tg(iSuRe-Cre) allele expression. Error bars in charts indicate standard deviation. T-Test, NS, not significant.

Figure 8. The Tg(iSuRe-Cre) allele is expressed in all organs and self-recombines only in some adult myocytes, a-e) Representative confocal micrographs of different mouse organs, showing that the Tg(iSuRe-Cre) allele is expressed in most cell types (N-PhiM-positive) and does not self-recombine (Mb Tomato-negative) in the cells of the indicated organs f) A fraction of adult skeletal muscle fibers self-recombine the Tg(iSuRe-Cre) allele and are MbTomato+ g, h) A fraction of adult, but not postnatal (P6), cardiomyocytes also recombine the allele (MbTomato+) in the absence of Cre/CreERT2 drivers i) Frequency of non-induced recombination in the embryo, the postnatal heart, and the indicated adult organs. MbTomato+ cells were counted on several immunostained sections and also by whole-organ FACS analysis (not shown). Self-recombination of the Tg(iSuRe-Cre) allele is found in only a fraction of quiescent muscle and heart myocytes j) Number and genotype of animals obtained with compound floxed alleles and the Tg(iSuRe-Cre) allele. The animals were bom at the expected Mendelian ratio and display normal reproduction rates and behavior. Error bars in (i) indicate StDev; each dot represents the average result obtained per animal and organ.

Figure 9. Specificity of iSuRe-Cre and its advantages, a) Representative confocal micrographs of retinas showing that in animals with the Tie2-Cre and Tg(iSuRe-Cre) alleles MbTomato is only expressed in endothelial (isolectinB4) and blood cells (macrophages, red), and not the many other cell types existent in the retina tissue (green channel) b) Chart showing that among all liver cells, only CD31+ endothelial and CD45+ blood cells express the MbTomato reporter c) Chart representing the theoretically predicted and experimentally observed correlation between Cre activity and efficiency of gene deletion. Constitutive Cre- expressing lines achieve, in general, higher gene deletion efficiencies than tamoxifen- inducible CreERT2 lines. Given the higher Cre expression provided by the iSuRe-Cre allele, the observed gene deletion efficiency is higher with this line than with most other Cre or CreERT2 lines d) Chart representing the theoretically predicted and experimentally observed correlation between Cre activity and the full recombination of reporter alleles or full deletion of genes. Like other classical Rosa26-reporter alleles, the iSuRe-Cre allele is efficiently induced/recombined at relatively low levels of Cre activity. With the exception of the gene D114, all the other experimentally tested genes (Kdr, Myc, Mycn, Rbpj, Notchl) are more difficult to delete/recombine than the iSuRe-Cre or Rosa26-reporter alleles, requiring the boosting in Cre activity provided after the pre-induction of the iSuRe-Cre, a dual reporter and cre expressing allele.

DETAILED DESCRIPTION OF THE INVENTION

Example 1. Material and Methods.

Example 1.1. Mice.

To generate mice (Mus musculus ) several mouse ES cell clones were generated and genotyped (Figure 2) to identify clones with ROSA26 gene-targeting or transgenesis of the iSuRe-Cre construct (see ES cell culture method below). With the selected ES cell clones, several mouse chimeras with high germline transmission rates were obtained, and used to produce mouse lines Gt(ROSA)26Sor-iSuRe-Cre and Tg(iSuRe-Cre) . In addition, we interbred Tg(iSuRe-Cre) mice with the following mouse lines: Tg(Tie2-Cre), Gt(Shh-GFP- Cre), Tg(Vavl-Cre), Tg(Alb-Cre), Tg(MyHC-Cre), Gt(LysM-Cre), Tg(Ubc-CreERT2), Gt(ROSA ) 26-CreERT2, Tg(Cdh5-CreERT2) , Gt(ROSA)26^LSL-^EYFP, Gt(ROSA)26^,Chr2-^Contro‘- ^Mosa Kάt^boc/boc , Dll4^lox/flox, Notchl^flox/flox , Rbpf^ox/fl^ox , Epasl^flox/flox , Myc^floxed or Mycrf^oxed mice. To activate recombination in animals containing CreERT2 alleles, tamoxifen (Sigma) was injected in pups or adult mice at the indicated stages and with a dose of 40pg/g. Genotyping primers are provided in Table 2.

Table 2. Primer sequences used to genotype or quantify genetic deletions.

Experiments involving animals were conducted in accordance with official guidelines and laws, following protocols approved by local animal ethics committees and authorities (Comunidad Autonoma de Madrid and Universidad Autonoma de Madrid - CAM-PROEX 177/14 and CAM-PROEX 167/17). Male and female mice were used for the analysis and maintained under specific pathogen-free conditions.

Example 1.2. Recombinant DNA, ES cell culture and genotyping. The basic elements of the iSuRe-Cre DNA constructs (Figure 1 and Figure 2) were obtained from different sources and assembled by standard DNA cloning as previously described. The Int-Cre sequence was obtained from plasmid p210 pCMV-CREM, a gift from Jeffrey Green (Addgene plasmid #8395). The plasmid used to generate the Gt(ROSA) 26Sor-iSuRe-Cre and Tg(iSuRe-Cre) mouse lines will be available at Addgene. Mouse ES cells with the G4 background, were cultured in standard ES cell media (DMEM containing Glutamax (31966- 047, Gibco), 15% FBS (tested for germline transmission), 1 x EAA (Hyclone, SH3023801), 0,1% b-mercaptoethanol (Sigma, M7522), 1 x Pen/Strep (Lonza, DE17-602E) and LIF) in dishes covered with a feeder layer of mouse embryonic fibroblasts (MEFs). For classical gene targeting of the large iSuRe-Cre plasmid, 25ug of linearized DNA was used to electroporate 5 million ES cells. Selection in 200ug/ml G418 (Geneticin) was performed for 6 days, after which individual colonies were picked for storage, PCR and Southern blot screening. Selected positive clones were expanded and used for microinjection in host blastocysts of the C57B1/6J strain. Chimeras with high percentage of agouti coat color were then crossed with mice to obtain germline transmission of the targeted insertion. PCR with ROSA26 5’homology arm flanking primers (Table 2), allowed us to identify ES cell clones with precise homologous recombination and insertion of the iSuRe-Cre construct in the ROSA locus. After identification of clones with precise gene targeting, we performed PCR with primers iSuRe-Cre F and iSuRe-Cre R (Table 2) to detect ES cell clones containing the transgene MbTomato-2A-Int-Cre sequence Selected ES cell clones were expanded for further analysis and the generation of mice.

Example 1.3. Immunostaining.

For immunostaining of ES cells (Figures 1 and 2), cells were fixed for 10 minutes in PBS containing PFA4% and Sucrose 4%. After a brief rinse in PBS, cells were permeabilized in 0.1% Triton for 10 minutes and then immersed in a blocking solution (10% Fetal bovine serum in PBS). Primary antibody (Rabbit Anti-PhiYFP, AB602, Evrogen) was diluted in blocking solution and incubated for 2 hours at room temperature or overnight, followed by three washes in PBS of 10 minutes each and incubation for 1 to 2 hours with conjugated secondary antibodies (Invitrogen or Biotium) at room temperature. After three washes in PBS, cells were mounted with Fluoromount-G (SouthernBiotech).

For immunostaining of mouse retinas (Figures 3, 4 and 9), eyes from mouse pups were dissected and fixed for 1 hour in a solution of PFA4% in PBS. After washing the tissue in PBS twice, retinas were microdissected and processed for immunostaining following a very similar protocol previously described above for the cells. The only difference is that the blocking/permeabilization buffer contains 0.3% Triton (Sigma), 3% FBS, 3% Donkey Serum (Millipore) and antibody washes were more extended in time; on average for 30 minutes each. Biotinylated IsolectinB4 (Vector Labs B-1205) and Streptavidin-405 (Invitrogen, S- 32351) were used to label the surface of the vessels. Conjugated rabbit anti-ERG-Alexa647 (Abeam, ab 196149) was used to detect the endothelial nuclei, and mouse anti-HA-647 (Cell signalling, #3444) was used to detect HA-H2B-Cerulean+ nuclei. The other endogenous fluorescent signals (H2B-EGFP, EYFP, MbTomato, H2B-Cherry) were excited and scanned with different compatible laser lines and defined detectors (488nm, 514nm, 546nm and 595nm respectively).

For immunostaining of mouse embryos and organ sections (Figures 3a, 5a, 5b and Figure 8) tissues were fixed for 2 hours in a solution of PFA 4% in PBS at 4°C. After washing the tissue in PBS three times, organs were stored overnight in 30% sucrose (Sigma) in PBS. Then, organs and embryos were embed in OCT™ (Sakura) and frozen at -80C. Cryosections of organs (35pm) and embryos (15pm), were cut on a cryostat (Leica). Sections were washed three times 10 minutes each in PBS and blocked/permeabilized in PBS with 10% Donkey Serum (Milipore) and 1% Triton (organs) or 0,5% Triton (embryos). Primary antibodies were diluted in blocking/permeabilization buffer and incubated overnight at 4°C. This step was followed by three washes in PBS of 10 minutes each and incubation for 2 hours with conjugated secondary antibodies (Jackson Laboratories) and DAPI at room temperature. After three washes in PBS, cells were mounted with Fluoromount-G (SouthemBiotech). The following primary antibodies were used: goat anti-D114 (R&D systems, AF1389); rat anti- VEGFR2 (BD Pharmingen, 550549), rabbit anti-Dsred (Clontech, 632496) and rabbit anti- ERG-Alexa647 (Abeam, Ab 110639). To detect MbTomato in the same section as ERG, endogenous signals were scanned or rabbit anti-Dsred plus a Fab fragment CY3 secondary antibody (711-167-003) was used, which is compatible with the use after of rabbit anti-ERG- Alexa647.

Example 1.4. Flow Cytometry and Fluorescence-Activated Cell Sorting (FACS).

Mice organs were minced and digested with 2.5 mg/ml of Collagenase (Thermofisher) type I, 2.5 mg/ml Dispase II (Thermofisher) and 50 ng/ml of DNAsel (Roche) at 37C for 30 minutes to create a single cell suspension. Cells were filtered through a 70 pm filter to remove non- dissociated tissue. Erythroid cells were removed from cells suspensions with a Blood Lysis Buffer (0.15 M NFLCl, 0.01M KHCO₃ and 0.01 M EDTA in distilled water) incubated for 10 minutes on ice. Cells suspensions were immediately analyzed or blocked with in DPBS no Ca2+ or Mg2+, containing 3% Dialyzed FBS (Termofisher) and incubated at 4°C for 30 minutes with APC conjugated rat anti-mouse CD31 (BD Pharmigen, 551262). DAPI was added prior to cell analysis. Anti-CD31-APC, MbTomato or EYFP signals were gated with a negative control tissue (Figure 3f and data supporting Figure 5e and Figure 8).

Blood for FACS analysis (Figure 3b, 3c and Figure 8) was collected by submandibular sampling and erythroid cells were also removed with the same blood lysis buffer described above. In the case of the LysM-Cre line analysis (Figure 6a), blood cells were pre-incubated on ice with rat anti-Mouse CD16/CD32 (BD Pharmingen, 553141) and then immunostained for the different blood cell types using rat anti-CD 1 lb-647 (BD Pharmingen, 557686), rat anti-CD45.2-APC-Cy7 (Tonbobio 25-0454), rat anti-Terl 19-biotin (BD Biosciences, 553672), rat anti-CD3 e-biotin (eBiosciences), rat anti-B220-biotin (BD Biosciences, 553085) and rat anti-Ly6G-PE-Cy7 (BD Biosciences, 560601). For the Vavl-Cre analysis, cells were incubated with APC rat anti mouse CD45 (BD Biosciences, 561018).

For the isolation of endothelial cells from dissociated tissues (Figure 3f), viable cells were selected by DAPI negative fluorescence. All viable cells were interrogated by examining FSC and SSC to select by size and complexity, and by comparing FSC-H and FSC-W repeated with SSC-H and SSC-W in order to discern single cells. An additional channel lacking any endogenous or fluorescent label was also acquired to detect and exclude autofluorescence. Cells were selected by their APC, endogenous EYFP and endogenous MbTomato positive signal. Flow cytometry analysis and FACs were performed on Fortessa or Aria Cell Sorter or Synergy4L machines. Experiments were analyzed using DIVA software.

Example 1.5. Differentiation of bone marrow derived macrophages (BMDMs) from LysM-Cre mice.

Bone marrow cells were flushed, passed through a 70um filtered and plated three days in RMPI 1640 medium (supplemented with 10% FCS, Glut 2mM, P/S, mouse lOng/uL M- CSF). Medium was renewed at Day3. At Day 5, cells were detached and sorted for YFP+ or MbTomato+ for downstream protein or RNA extraction and analysis. qRT-PCR was performed using specific primers for Epasl (SEQ ID NO: 7 TGAGTTGGCTCATGAGTTGC and SEQ ID NO: 8 TTGCTGATGTTTTCCGACAG) and controls 36b4 (SEQ ID NO: 9 AGAT GC AGC AGAT C CGC AT and SEQ ID NO: 10 GTTCTTGCCCATCAGCACC) and cyclophilin (SEQ ID NO: 11 ACAGGTCCTGGCATCTTGTC and SEQ ID NO: 12 CAT GGC TT CC AC AAT GTT C A) . Example 1.6. Western blot analysis.

For the analysis of protein expression, dissected organs were transferred to a reagent tube and frozen in liquid nitrogen. On the day of the immunoblotting the tissue was lysed with lysis buffer [(Tris-HCl pH=8 20mM, EDTA ImM, DTT ImM, Triton X-100 1% and NaCl 150mM, containing protease inhibitors (P-8340 Sigma) and phosphatase inhibitors (Calbiochem 524629) and orthovanadate-Na 1 mM)] and homogenized with a cylindrical glass pestle. Sorted CD31-APC+ endothelial cells (Figure 7b) or bone marrow derived macrophages samples (Figure 7c) were lysed with RIPA buffer (Sigma R0278) containing protease inhibitors (Sigma P-8340) and phosphatase inhibitors (Calbiochem 524629) and orthovanadate-Na 1 mM. Tissue/ cell debris was removed by centrifugation, and the supernatant was diluted in loading buffer and analysed by SDS-PAGE and immunoblotting. Membranes were blocked with BSA and incubated with primary antibodies diluted 1/1000 against Cre (Merck, 69050-3), Cdh5/VE-cadherin (BD Biosciences 555289) or b-Actin (Santa Cruz Biotechnologies, sc-47778).

Example 1.7. DNA extraction from FACS sorted cells and semi-quantitative PCR.

Blood, liver or endothelial cells from different mice were sorted based on their MbTomato or anti-CD31-APC fluorescent signals in DPBS no Ca2+ or Mg2+, containing 3% Dialyzed FBS (Termofisher). A minimum of 42.000 cells per sample, were sorted and spin down for 5 minutes at 5000g, and resuspended in 50 mΐ of lysis buffer prepared as follows: 25ul of DirectPCR (Cell) Lysis Reagent for PCR (VIAGEN Cat #301-C) and 25 pi of distilled water and a final concentration of 0,4mg/ml of proteinase K. Cells were incubated 55°C overnight and then the proteinase was inactivated at 85°C for 45 minutes. Semi-quantitative and competitive PCR was performed with lul of Tail DNA from control animals or lul of the MbTomato positive or negative cells using the primers mentioned in Table 2. Groups of 3 or 4 primers were used per PCR reaction. Some of them served as control of DNA input, and the others corresponded to the different floxed genes sequences.

Example 1.8. Image acquisition and analysis.

ES cells, flat-mounted retinas, organs and embryo sections with or without immunostaining were imaged at high resolution with a Zeiss LSM710 or Leica SP8/SP5 confocal microscopes. lOx, 20x or 40x objectives were used for confocal scanning. Individual fields or tiles of large areas were acquired. Fiji/ImageJ was used to threshold, select and quantifies obj ects in confocal micrographs. For chart in Fig. 3e, cell frequency was determined by counting the number of ERG+ ECs with MbTomato or EYFP or dual reporter expression (representative pictures Figure 3d and 3g). ERG labels the nuclei of individual ECs, allowing their accurate quantification. In Figure 3h, six laser scanning confocal channels were acquired and quantified. ECs are very elongated and both MbTomato (546nm) and EYFP (514nm) proteins accumulate more in the cytoplasm around the nucleus (Figure 3h”), being possible to count the total number of cells even in the absence of the incompatible ERG immunostaining channel. Besides this, we also counted on independent pictures the number of ERG+ nuclei/cells per MbTomato+ endothelial surface area, obtaining similar results. The ratio of ERG+ objects in MbTomato+ EC surface area can be used to estimate the total number of nuclei/cells in the EYFP+ endothelial surface area. The fluorescent proteins with nuclei localization (H2B-Cherry (595nm), H2B-EGFP (488nm) and HA-H2B-Cerulean (anti-HA-647 antibody signal, 647nm) were detected with separate and compatible laser excitation and detectors (Figure 3h””). ImageJ was used to detect the number of nuclei (H2B-Cherry, H2B-EGFP or HA-H2B-Cerulean) objects in MbTomato+ or EYFP+ endothelial surface area (Figure 3i). ImageJ was also used to detect the total ERG+ (EC nuclei) objects in IsolectinB4 or MbTomato+ or MbYFP+ surface area in the pictures supporting the data in Figures 4c, 4d and 4f. For the data shown in Figure 5c, images representing large fields were selected and used to quantify the number of cells with detectable expression of MbTomato, Kdr or D114 proteins. Quantification of PCR and Western gel bands intensity shown in Figure 5e and Figure 7 was done with Image J by calculating the mean grey value of each band and subtracting it from the background mean grey value. Indicated relative ratios in Figure 5e represent the ratio between control and floxed PCR bands when comparing control samples (with no inducible genetic deletion) with MbTomato negative and MbTomato positive samples. In Figure 7, fold change indicated represent the relative difference in Cre protein levels taking as a reference the protein loading controls (Cdh5 or B-actin).

Example 1.9. Statistical Analysis In Figure 3, each dot in the charts indicate the frequency obtained in each animal (c) or in large confocal micrographs taken from the retinas of different animals (e, i). In Figure 3c a t- test was performed. Each dot in charts Figure 4c, 4d and 4f corresponds to the quantification of an image acquired with the same settings and dimensions as the ones shown in Figures 4a, 4b and 4e. Yellow dots indicate the quantification numbers corresponding to Figures 4a, 4b and 4e with labels 1-10. A minimum of 20 images were quantified per group. An unpaired t- test was performed in Figure 4c. In Figure 4d a linear regression line and the obtained R square is shown. In Figure 5c each dot represents the average value obtained in large microscopic fields, with 485-654 ERG+ objects (endothelial nuclei) per field. ANOVA and multiple comparisons were used to calculate the indicated p values. For all other charts, not involving microscope pictures, T-test or ANOVA was used to calculate the p values indicated in figure legends. All calculations and charts were performed with GraphPad Prism software. No randomization or blinding was used and animals/tissues were selected for analysis based on the detected Cre-dependent recombination frequency and quality of multiplex immunostaining. The sample size was chosen according to the observed statistical variation and published protocols.

Example 2. Results.

Example 2.1. Design and test of iSuRe-Cre constructs.

The limitations of the Cre/LoxP technology outlined above are mainly related with the weak and often variable expression of promoters in transgenes containing the Cre or CreERT2 coding sequences that frequently do not induce the deletion of all different types of floxed alleles in the cells where they are expressed. To overcome this, we sought to develop a new DNA construct, easier to induce by Cre/CreERT2-mediated recombination, and that would subsequently enable strong and sustained co-expression of a fluorescent reporter and a constitutively active Cre (Figure la). We reasoned that this simple construct {iSuRe-Cre), when inserted in the genome of ES cells or mice, would significantly increase the efficiency and reliability of conditional genetic modifications. In the design of this construct, we considered three distinct requirements. First, recombination of this construct should be much easier to induce than most other floxed genes. Second, the construct must include a strong and ubiquitous promoter driving high expression of a bright fluorescent reporter of recombination, to clearly label live or fixed cells with induced Cre activity. Third, and more importantly, once construct recombination is induced, the expression of the reporter must match 100% the full recombination of any other floxed alleles in the same cell due to the high and permanent expression of Cre (Figure la, b). Our strategy was to assemble a construct (Figure lc), containing the ubiquitous and strong CAG promoter followed by a very short (l. lkb), and therefore easy to recombine, LoxP-flanked DNA sequence containing the N- PhiM reporter gene and a transcription stop signal (pA). After recombination/deletion by Cre or CreERT2, this construct would enable the strong co-expression of a bright membrane- localized fluorescent reporter (MbTomato) and a constitutively active and permanently expressed Cre protein; these two proteins would be separated by the viral 2A peptide to guarantee equimolar expression and complete correlation between the MbTomato reporter expression and Cre activity (Figure lc). With this strategy, we expected to significantly enhance the efficiency and reliability of inducible gene deletions.

Our first attempts to include these elements in the same plasmid failed. All plasmids containing the Cre-coding sequence had recombination/deletion of the LoxP-flanked N- PhiM-pA cassette (Figure Id). This undesired recombination in E. Coli of plasmids containing Cre and LoxP flanked cassettes in cis is due to the low-level expression of the Cre sequence even in the absence of adjacent prokaryotic promoters. This may explain why similar DNA constructs and mouse lines have not been produced or published in the past. This modification disrupts the correct expression of the Cre in splicing-incompetent prokaryotic cells but is compatible with expression in eukaryotic cells, which splice out the non-coding intron and therefore express Cre. When we inserted the intron-interrupted Cre sequence (Int-Cre), downstream of the short LoxP-flanked transcription stop cassette, we were able to recover iSuRe-Cre plasmids without any leaky recombination/deletion in E. Coli (Figure le). To monitor CAG promoter expression in the absence of Cre activity, we introduced the small N-PhiM gene, which encodes a non-cytotoxic and mutated PhiYFP protein (Evrogen) with a nuclear localization signal and lacking endogenous fluorescence. This protein can be easily detected by immunostaining, in cells expressing the iSure-Cre plasmid in the absence of induced Cre activity or recombination (Figure If). Expression of the MbTomato-2A-Int-Cre cassette allows us to detect cells with Cre-dependent recombination and equimolar high Int-Cre expression, by direct detection of the fluorescence emitted by the membrane-localized Tomato protein (MbTomato) or after immunostaining with an anti-Dsred antibody (Figure If). Example 2.2. Generation of transgenic and ROSA26 gene targeted iSuRe-Cre ES cells and mice.

The iSuRe-Cre DNA elements were finally inserted on a modified ROSA26 gene-targeting vector, containing four chicken b-globin HS4 insulator sequences (Figure 2a), to reduce genomic interference and silencing from neighbouring genomic regions. The assembled plasmid was used to generate several independent embryonic stem (ES) cell lines (Figure 2b, 2c). In some of the generated ES cell clones, the construct was inserted in the ROSA26 locus ( Gt(ROSA)26Sor-iSuRe-Cre ), while in others it was randomly inserted in the genome ( Tg(iSuRe-Cre )). If the constitutive CAG promoter of the iSure-Cre allele is expressed, N- PhiM expression is detectable, and if the allele is recombined, the N-PhiM-pA cassette is deleted and MbTomato expression induced. We observed a high degree of non-induced and sporadic recombination in most ES cell lines containing this construct in the genome (Figure 2d, 2e), even in the absence of exogenous CreERT2 or Cre expression. The penetrance of this undesired recombination or leakiness was variable among clones and ranged from 15% to 95%. Although these results were disappointing, we proceeded to generate mice with the clone having the lowest degree of uninduced recombination from the Tg(iSuRe-Cre) and Gt(ROSA)26Sor-iSuRe-Cre ES cell lines (Figure 2d, 2e, clones 6 and 18). The male chimeras generated with Gt(ROSA)26Sor-iSuRe-Cre ES cells carried the non-recombined allele in a large fraction of their cells, but the non-recombined allele was not passed on to any of their progeny. All progeny expressed the MbTomato-2A-Int-Cre cassette in all cells and had full deletion of the Rbpj floxed allele (Figure 2f), suggesting that when the iSuRe-Cre construct is targeted to the Rosa26 locus, the leakiness is high in the male germline, rendering the allele useless.

We next analyzed the progeny of chimeras derived from Tg(iSuRe-Cre) ES cells with the construct integrated outside the ROSA26 locus. From these chimeras, we were able to obtain progeny carrying the Tg(iSuRe-Cre) allele and without MbTomato-2A-Int-Cre expression or recombination of the Rbpj floxed allele (Figure 2g). Thus, the Tg(iSuRe-Cre) transgenic allele is not leaky in the male germline or in embryos, allowing us to generate, for the first time, a Cre-inducible or CreERT2-inducible constitutive dual Reporter-Cre expressing mouse line. Example 2.3. The Tg(iSuRe-Cre) allele is ubiquitously expressed and reliably reports cells with recombination of other alleles.

Transgenic alleles can be silenced in some cell types, reducing their utility as research tools. The iSuRe-Cre transgene contains four chicken b-globin HS4 insulator sequences (Figure 2a), to reduce genomic interference and silencing from neighbouring genomic regions. However, insulated transgenic alleles can still be affected by the surrounding genome and chromatin status, which varies between cell types and across different developmental stages. Therefore, we analyzed expression of N-PhiM (reporter of transgene promoter activity) and MbTomato (reporter of transgene recombination and Int-Cre expression) in most embryonic and adult tissues of Tg(iSuRe-Cre) mice. This analysis revealed that the Tg(iSuRe-Cre) allele was not silenced and did not self-recombine in the embryo or in most organs and cell types (Figure 3a-c and Figure 8). The exception was adult and quiescent myocytes that form the skeletal and cardiac striated muscle tissue; a fraction of these cells consistently had leaky recombination of the Tg(iSuRe-Cre) allele and MbTomato expression (Figure 8 f-i). Since the purpose of this line is the performance of tissue-specific and inducible gene loss-of- function experiments, which should yield a phenotype only after induction, the leaky expression and recombination in a fraction of quiescent adult myocytes will not affect the utility of the line for manipulating and understanding gene function during embryonic development or in other cell types in adults. Echocardiographic studies of adult mice having 3 months revealed no significant differences in heart function (Table 1).

Table 1. Echocardiographic parameters obtained in Tg(iSuRe-Cre) and control littermates.

All animals containing the Tg(iSuRe-Cre) allele and singly or compound homozygous for different floxed alleles (. D114 , Kdr, Rbpj, Fgfrl, Fgfr2, Myc, Mycn, Hifl, or Hif2) were born at Mendelian rates, healthy, and had normal behavior (Figure 81j), indicating that the spontaneous deletion of these genes in a fraction of quiescent adult myocytes is not deleterious.We next evaluated the relative inducibility rate of the Tg(iSuRe-Cre) allele by combining it with Cdh5-CreERT2 and other available Cre reporter alleles, such as the Gt(ROSA)26Sor-LoxP-STOP-LoxP-EYFP allele (abbreviated here as ROSA26^{LSL ErEP}). The Tg(iSuRe-Cre) allele has a high recombination frequency even if not targeted to the ROSA26 locus (Figure 3d, e), possibly due to the short genetic distance between the LoxP sites flanking the N-PhiM-pA cassette (1.1 kb). In addition, inclusion of the stronger CAG promoter, the brighter fluorescent protein (Tomato), and the WPRE element (Figure 2a) results in higher reporter expression in each recombined cell, making it easier to detect and distinguish from background tissue autofluorescence than other commonly used ROSA26 reporter alleles, as determined by FACS (Figure 3f). To determine if the Tg(iSuRe-Cre) allele was also a better reporter of the recombination of other floxed alleles, we compared it with the commonly used ROSA26^{LSL EYFP} allele in reporting recombination of the ROSA26^{lChr2 Mosa,c} allele. At both high and low recombination frequencies, all MbTomato+ cells were EYFP+ (Figure 3d-g), whereas a significant fraction of EYFP+ cells were MbTomato-. The ROSA26^{lChr2 Mosaic} reporter contains three distinct and mutually exclusive LoxP sites that compete for the recombination event, lowering the recombination efficiency of this allele. Once this allele is recombined, cells express chromatin/nucleus-localized proteins (H2B-Cherry, H2B-GFP, or HA-H2B-Cerulean) that can be clearly distinguished from MbTomato or EYFP (Figure 3h). Whereas only 23% of single EYFP+ cells had also recombined the ROSA26^,Chr2~Mosmc allele, all individual cells expressing the MbTomato-2A- Int-Cre cassette had recombined the ROSA26^{,Chr2 Mosa,c} allele (Figure 3h and 3i). All MbTomato+ cells recombined and expressed both the ROSA26^{LSL EYFP} and ROSA26^{lChr2 Mosa,c} reporters, demonstrating the high recombination efficiency in cells with activation of the Tg(iSuRe-Cre) allele (Figure 3h’). Collectively, these results show that the Tg(iSuRe-Cre) allele is expressed and inducible in most tissues and that it is also a much more reliable reporter of cells with recombination of other floxed alleles.

Example 2.4. The Tg(iSuRe-Cre) allele significantly increases the efficiency of inducible genetic modifications.

Having confirmed the Tg(iSuRe-Cre) allele as a bona fide gene-deletion reporter, we next sought to determine if it would also allow us to achieve higher efficiencies of CreERT2- inducible gene deletion. In contrast to the very short (l . lkb), inter-loxP genetic distance in the Tg(iSuRe-Cre) allele, the floxed alleles of many other genes have significantly larger genetic distances and require two recombination events for full gene loss-of-fimction. We generated mice containing the Cdh5-CreERT2 and Tg(iSuRe-Cre) transgenes and two Kdr floxed (Kdr ^Jlox) alleles, where the LoxP sites are 4.7kb apart. Kdr is an essential gene for EC differentiation, proliferation, migration, and survival The results show that when Kdi^JtnxJtnx Cdh5-CreERT2^Tg/Wt mice contain the Tg(iSuRe-Cre) allele, Kdr gene deletion is much more efficient (Figure 4a-c). They also show that full Kdr recombination and loss-of-fimction cannot be consistently achieved with tamoxifen-inducible conditional genetics unless the Tg(iSuRe-Cre) allele is present. As with any tamoxifen induction experiment, there was some variability in the degree of recombination of the Kdr^lo and Tg(iSuRe-Cre) alleles among littermates. However, the mutants with almost complete expression of the MbTomato-2A-Int- Cre cassette could be safely selected for phenotypic and statistical analysis (Figure 4d). In animals containing the ROSA26^LSL~YFP reporter allele, Kdr gene deletion was significantly more variable and less efficient, even in tissues with very high reporter recombination rates (Figure 4e, 4f). Importantly, the vascular phenotype obtained at high reporter recombination rates was not representative of the real full gene loss-of-fimction phenotype, reflecting the poor correlation between the recombination of a conventional Cre-reporter (i.e. ROSA26^{lsl~ YFP}) and deletion of the gene of interest. Since Kdr is an essential gene for EC proliferation and survival, the mutant cells are strongly outcompeted by the remaining non-deleted cells, making it difficult to achieve full deletion of this gene in most endothelial cells (ECs) during vascular development, unless the Tg(iSuRe-Cre) allele is used (Figure 4f).

To analyze the efficiency of the Tg(iSuRe-Cre) allele in the recombination of other genes, in other cell types, we interbred it with the ubiquitously expressed Rosa26-CreERT2 and Ubc- CreERT2 lines, and induced recombination of the genes Rbpj and Notchl with tamoxifen. In cells where the Tg(iSuRe-Cre) allele was induced (MbTomato+), the Rbpj and Notchl genetic deletion was highly efficient (Figure 4g, 4h). In addition to inducible CreERT2 mouse lines, we also compared gene deletion efficiency of the Tg(iSuRe-Cre) allele versus a conventional ROSA26^{LSL YFP} allele when interbred with the constitutively active LysM-Cre line, that is specifically expressed in granulocytes and macrophages. In contrast to inducible CreERT2, Cre usually recombines most floxed alleles in the cells where it is expressed; however this may vary with the levels of Cre expression per cell. We found that while deletion of the floxed gene Epasl is not complete in LysM-Cre+ YFP+ cells, it is in LysM-Cre+ MbTomato- 2A-Int-Cre+ cells (Figure 4i).

Example 2.5. The Tg(iSuRe-Cre) allele enables multiple gene deletions in single cells or tissues.

A common strategy for understanding and validating genetic redundancy or gene interaction networks is to induce the deletion of two or more genes in the same tissue and analyse if deletion of one gene aggravates or rescues the phenotype caused by the other. This is a fundamental method for understanding how genes interact to regulate a given biological process. Given the stronger Cre activity and recombination efficiency shown by the Tg(iSuRe-Cre) allele, we reasoned that it would be particularly useful in reporting cells with full deletion of multiple genes or floxed alleles. As an example, we present data for the simultaneous deletion of the genes DIM/Kdr or Myc/Mycn/Rbpj in ECs expressing Cdh5- CreERT2, after inducing animals with tamoxifen. This implies the deletion of 4 or 6 floxed alleles in each cell to achieve cell-autonomous multiple gene loss-of- function and in this way perform epistasis analysis. D114 and Kdr are expressed by most liver ECs, and their presence in ECs could be detected by immunostaining (Figure 5a). In animals containing the Cdh5- CreERT2 and Tg(iSuRe-Cre) alleles and injected three times with tamoxifen, we observed deletion of D114 in most liver ECs, but not of Kdr (Figure 5 b,c). This result shows how recombination efficiency differs among different floxed genes. Importantly, MbTomato+ ECs had complete deletion of both D114 and Kdr, whereas most MbTomato negative cells maintained Kdr expression, even after three inj ections of tamoxifen at high dose (Figure 5b, 5c).

We next analysed multiple gene deletion efficiency in Myc/Mycn/Rbpf^ox,^^ox , Tg(Cdh5- CreERT2), Tg(iSnRe-Cre) mice induced with a single injection of lmg of tamoxifen (Figure 5d, 5e). Technically, it is much more difficult to confirm the simultaneous deletion of more than 2 genes in single cells by direct immunostaining, due to the lack of compatible commercial antibodies to distinguish all epitopes of the encoded proteins in combination with other tissue or reporter markers. We therefore isolated DNA from FACS sorted MbTomato+ and MbTomato- ECs (CD31+), from the induced animals and assessed gene deletion by semi-quantitative PCR. The results show that only MbTomato-2A-Int-Cre+ cells had deletion of the 3 genes (Figure 5d-f). The Tg(iSuRe-Cre) allele enabled us to achieve high efficiency in multiple gene deletion, and safely correlate the expression of the single and easy to detect MbTomato reporter, with the deletion of 3 genes (6 floxed alleles), which encode for proteins that could not be detected by tissue immunostaining.

Example 2.6. The Tg(iSuRe-Cre) allele is not leaky when interbred with several other Cre lines.

Tissue-specific Cre mouse lines, normally express Cre at high levels in a given tissue where the transgene promoter is active, but may also express it at low levels in other undesired tissues, which can be enough to recombine the most sensitive floxed genes or Cre activity reporters localized in the Rosa26 locus. Given the high sensitivity of the Tg(iSuRe-Cre) allele to Cdh5-CreERT2 induced recombination (Figure 3e), and its location outside the Rosa26 locus, we next analysed if the combination of this allele with other constitutive Cre- expressing alleles could result in non-specific recombination, or leakiness. This analysis allowed us to see that the Tg(iSuRe-Cre) allele is not leaky in the different cell-types, organs and developmental contexts analysed. With LysM-Cre and Vavl-Cre, recombination of the Tg(iSuRe-Cre) allele was detected only in the expected fraction of hematopoietic lineages (Figure 6a, 6b). Interestingly, we detected non-specific or leaky recombination of the ROSA26^{LSL YFP} allele with these two Cre lines (Figure 6a, 6b), as also reported before. Embryos containing the Shh-GFP-Cre allele had expression of the transgene in GFP-Cre+ endoderm derived cells and limb buds (Figure 6c). Finally, when combined with the Tie2- Cre, Alb-Cre and MYHC-Cre alleles, the Tg(iSuRe-Cre) transgene was also specifically expressed in blood or endothelial cells (Tie2+), liver hepatocytes (Alb+), or heart cardiomyocytes (MyHC+), respectively (Figure 6d, 6e and Figure 9 a, b). Altogether, these results indicate that the genomic integration site of the Tg(iSuRe-Cre) allele is only leaky in a small fraction of adult myocytes (Figure 8), even when combined with other tissue-specific Cre expressing transgenes. Importantly, we were able to map the transgene integration site to a non-coding region between genes Tmem200c and Epb41I3 (Figure 6f) in chromossome 17 (Cergentis). This allowed us to design primers to detect homozygous animals, and confirm they are bom at the expected mendelian ratios (Figure 6g), indicating that the transgene integration did not disrupt any important regulatory elements. This genomic region will be of interest for gene targeting of other similar constructs, since it escapes the Rosa26-associated leakiness in the mouse germline (Figure 2f and 3b, 3c), while being permissive to the expression of the integrated transgenes in all the analyzed developmental stages and organs, similarly to the Rosa26 locus.

Example 2.7. Permanent expression of Cre after induction of the Tg(iSuRe-Cre) allele is not toxic.

High levels of Cre expression, or induction of CreERT2 with high doses of tamoxifen, has been shown to cause cellular toxicity and induce unexpected cellular phenotypes. Injection, transfection or infection of cells with Cre expressing DNA constructs, where multicopy epissomal or multicopy integration in the same locus of the genome, generates often high Cre expression levels and cellular toxicity. In vivo , Cre toxicity was reported for the Rosa26- CreERT2 and a-MHC-MerCreMer mouse lines, only when high doses of tamoxifen were used, and for the a-MyHC-Cre mouse line, which shows minor cardiac defects in adult mice with more than 3 months of age, that become more pronounced at 6 months of age. In contrast to these alarming reports, more than 300 Cre expressing mouse lines have been generated and characterized (Jackson laboratory), and for the large majority cellular toxicity was not reported. All these lines have tissue-specific promoters driving continuous expression of Cre, in the target tissue. This includes numerous mouse lines with multicopy integration of Cre expressing transgenes, such as the 10 copies of Alb-Cre mice and 20 copies of the Tie2- Cre mice. The Tg-iSuRe-Cre allele integrated in chromosome 17, and is expressed as unicopy, which may result in weaker expression in relation to other tissue-specific Cre mouse lines containing several copies of the Cre transgene. Although the expression of the Tg-iSuRe-Cre allele is driven by the strong CAG promoter, and enhanced by the WPRE element (Figure 2a), the 2A peptide included decreases overall translation efficiency. Therefore, instead of comparing RNA expression levels, we decided to compare Cre protein expression levels in cells expressing the Tg(iSuRe-Cre) allele and other existing Cre alleles, to assess potential cellular toxicity related to high Cre levels. This analysis revealed that the levels of Cre expression driven by the Cre-induced Tg(iSuRe-Cre) allele, are within the normal range, providing a moderate increase in Cre protein levels in the tissues expressing Cdh5-CreERT2 (endothelial cells), LysM-Cre (macrophages), Tie2-Cre (endothelial and blood cells) and Alb- Cre (hepatocytes) alleles (Figure 7a-e). Importantly, when we compared the expression of Cre in hearts of MyHC-Cre and MyHC-Cre Tg(iSuRe-Cre) mice, we did not detect any significant increment in Cre protein levels (Figure 7f). Analysis of Cre levels in hearts of animals with germline recombination of the Gt(iSuRe-Cre) allele (ubiquitous), that do not contain the MYHC-Cre allele, but have expression of Cre in all cardiomyocytes and all other heart cells (i.e. vascular cells, fibroblasts, blood cells), revealed that the Cre expression driven by the CAG promoter is much lower than the MyHC-Cre transgene (Figure 7g). This very high MyHC-Cre transgene (6 copies) expression levels, may explain the cardiotoxicity detected in MyHC-Cre mice with 3-9 months of age.

Further supporting the non-toxic effect of iSuRe-Cre expression, mice with germline recombination of the Gt(ROSA)26Sor-iSuRe-Cre allele (Figure 2f), are alive and fertile. Expression of MbTomato-2A-Int-Cre in all their cells since conception, throughout their embryonic, postnatal and adult life was not deleterious. In addition, metabolic studies of 7 months old mice did not reveal any significant difference among them (Figure 7h-7j). Therefore, we conclude that expression of MbTomato-2A-Int-Cre driven by the unicopy CAG promoter is not toxic. To have a complete control of conditional genetic experiments, the Tg-iSuRe-Cre allele can be induced in animals with and without other floxed genes, where the only difference between the cells under study will be the induced deletion of the floxed genes.

Claims

1. A DNA construct characterized by comprising in the following order: a. A promoter,

b. A LoxP-flanked DNA sequence, this sequence coding for a first reporter of transgene promoter activity and a transcription stop poly(A) signal, c. A second reporter of transgene recombination and Cre expression, d. A viral 2A peptide coding sequence, and

e. A Cre protein coding sequence comprising an intron within its sequence.

2. The DNA construct, according to claim 1, wherein Cre protein is constitutively active and permanently expressed once a pre-induced Cre-activity removes the LoxP- flanked DNA sequence.

3. The DNA construct, according to any of the claims 1 or 2, wherein Cre protein has an addition of the aminoacids proline, glutamic acid and phenylalanine at the N-terminal region.

4. A DNA construct, according to any of the claims 1 to 3, wherein the first reporter is the non-fluorescent reporter N-PhiM.

5. A DNA construct, according to any of the claims 1 to 4, wherein the second reporter is the fluorescent reporter MbTomato.

6. A DNA construct, according to any of the claims 1 to 5, wherein the promoter is the synthetic CAG promoter.

7. DNA vector comprising the construct according to any of the claims 1 to 6.

8. Cells comprising in their genome the DNA construct of any of the claims 1 to 6 or cells transfected with the vector of claim 7.

9. Cells, according to claim 8, characterized in that they are embryonic stem cells.

10. A transgenic non-human organism comprising in their genome the DNA construct of any of the claims 1 to 6 or comprising the cells of any of the claims 8 or 9.

11. The transgenic non-human organism, according to the claim 10, characterized in that it is a mouse.