Genetically modified mouse with an inducible ACVR1 gene with a mutant R206H exon 5 that has ectopic bone formation

Abstract

A genetically modified mouse is provided that comprises a conditional Acvr1 allele that comprises a mutated exon that, upon induction, converts to a mutant exon phenotype, wherein the mutant exon phenotype includes ectopic bone formation. Mice comprising a mutant Acvr1 exon 5 in antisense orientation, flanked by site-specific recombinase recognition sites, are provided, wherein the mice further comprise a site-specific recombinase that recognizes the site-specific recombinase recognitions sites, wherein the recombinase is induced upon exposure of the mouse to tamoxifen. Upon exposure to tamoxifen, the recombinase is expressed and acts on the RRS-flanked mutant exon 5 and places the mutant exon 5 in sense orientation and deletes the wild-type exon.

Claims

1. A genetically modified mouse whose genome comprises a modified mouse Activin A receptor type 1 (Acvr1) gene comprising: i) a human .Iadd.Acvr1 gene .Iaddend.exon 5.[.of the Acvr1 gene.]..Iadd., .Iaddend.encoding glutamic acid at the first codon instead of aspartic acid.Iadd., .Iaddend.in sense orientation flanked upstream and downstream by first pair of recombination recognition sites; and ii) a mutant exon 5 of a mouse Acvr1 gene.[.comprising an R206H mutation.]..Iadd., .Iaddend.in antisense orientation .Iadd.and encoding an R206H variation, .Iaddend.flanked upstream and downstream by second pair of recombination recognition sites that are different than the first pair of recombination recognition sites; wherein the first and second recombination recognition sites are oriented so that a recombinase can invert the mutant exon 5 into sense .[.orientation,.]. .Iadd.orientation and .Iaddend.delete the human exon 5, .[.and allow.]. .Iadd.thereby producing .Iaddend.a mutant Acvr1 .Iadd.gene .Iaddend.comprising .[.the mutant.]. .Iadd.an .Iaddend.exon 5 .Iadd.encoding the R206H variation, and allowing the mutant Acvr1 gene .Iaddend.to be expressed resulting in ectopic bone formation.

2. A .[.transgenic.]. .Iadd.genetically modified .Iaddend.mouse derived from the mouse of claim 1, wherein .[.said transgenic.]. .Iadd.the genetically modified .Iaddend.mouse .Iadd.so derived .Iaddend.has a genome comprising .[.a nucleic acid sequence encoding the mutant.]. .Iadd.the modified mouse .Iaddend.Acvr1 .[.comprising the mutant exon 5, and wherein the mouse expresses the mutant Acvr1 comprising the mutant exon 5 resulting in ectopic bone formation.]. .Iadd.gene.Iaddend..

3. The genetically modified mouse of claim 2, wherein the first and second pairs of recombination recognition sites are Lox2372 and LoxP or vice versa.

4. The genetically modified mouse of claim 2, wherein the recombinase is a tamoxifen-inducible CreER.sup.T2.

.Iadd.5. The genetically modified mouse of claim 1, wherein the mouse further comprises a gene encoding an inducible recombinase. .Iaddend.

.Iadd.6. The genetically modified mouse of claim 5, wherein the inducible recombinase comprises inducible Cre recombinase. .Iaddend.

.Iadd.7. The genetically modified mouse of claim 6, wherein the inducible Cre recombinase comprises tamoxifen-inducible Cre-ER.sup.T or Cre-ER.sup.T2 recombinase. .Iaddend.

.Iadd.8. The genetically modified mouse of claim 5, wherein the inducible recombinase is expressed in a cell in the mouse comprising the modified mouse Acvr1 gene. .Iaddend.

.Iadd.9. The genetically modified mouse of claim 6, wherein the inducible recombinase is expressed in a cell in the mouse comprising the modified mouse Acvr1 gene. .Iaddend.

.Iadd.10. The genetically modified mouse of claim 7 wherein the inducible recombinase is expressed in a cell in the mouse comprising the modified mouse Acvr1 gene. .Iaddend.

.Iadd.11. The genetically modified mouse of claim 5, wherein the gene encoding an inducible recombinase is at the ROSA26 locus. .Iaddend.

.Iadd.12. The genetically modified mouse of claim 1, wherein the mouse is heterozygous for the modified mouse Acvr1 gene. .Iaddend.

.Iadd.13. The genetically modified mouse of claim 1, wherein the mouse is homozygous for the modified mouse Acvr1 gene. .Iaddend.

.Iadd.14. The genetically modified mouse of claim 2, wherein the genetically modified mouse has a genome comprising a nucleic acid sequence encoding the mutant Acvr1 gene, and wherein the mouse expresses the mutant Acvr1 gene resulting in ectopic bone formation. .Iaddend.

.Iadd.15. The genetically modified mouse of claim 14, wherein the recombinase is an inducible Cre recombinase. .Iaddend.

.Iadd.16. The genetically modified mouse of claim 15, wherein the inducible Cre recombinase is tamoxifen-inducible Cre-ER.sup.T or tamoxifen-inducible Cre-ER.sup.T2 recombinase. .Iaddend.

.Iadd.17. The genetically modified mouse of claim 2, wherein the genetically modified mouse has ectopic bone formation. .Iaddend.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 illustrates design of a conditional allele at an Acvr1 locus that converts, e.g., a mouse Acvr1 exon 5 to a human R206H exon using loxp and lox2372 sites.

(2) FIG. 2 illustrates design of a conditional allele of Acvr1 R206H classic FOP mutant receptor gene. Mouse Exon 5 (e5 in isoform 001) is replaced with human exon 5 (in human ACVR1 isoform 003); a mouse mutant exon is simultaneously introduced in the antisense strand, together with a FRT'ed selection cassette (hUB-Neo); human e5 is flanked with loxP and lox2372 pointing East, and another loxP and Lox2372 sites are placed downstream of mouse e5(R206H) and deletion of the human e5, upon exposure to Cre, as detailed schematically in FIG. 1.

(3) FIG. 3 illustrates activation of the Acvr1.sup.[R206H]COIN allele results in an FOP-like phenotype in mice genetically modified with the conditional allele.

(4) FIG. 4 illustrates ectopic bone formation in genetically modified mice comprising the conditional allele induced in mice administered tamoxifen; an example of ectopic bone formation at the sternum is indicated in the right panel with white arrows. In the absence of tamoxifen (left panel), no ectopic bone formation is detected.

(5) FIG. 5 provides another illustration of ectopic bone formation in genetically modified mice comprising the conditional allele induced in mice administered tamoxifen; an example of ectopic bone formation at the sternum is indicated in the right panel with white arrows. In the absence of tamoxifen (left panel), no ectopic bone formation is detected.

(6) FIG. 6 provides yet another illustration of ectopic bone formation in genetically modified mice comprising the conditional allele induced in mice administered tamoxifen; an example of ectopic bone formation at the sternum is indicated in the right panel with white arrows. In the absence of tamoxifen (left panel), no ectopic bone formation is detected.

(7) FIG. 7 illustrates control mice (left panels, ID 840095); and ectopic bone formation in genetically modified mice comprising the conditional allele induced in mice administered tamoxifen (Tamoxifen #2, ID:845202); top right panel shows ectopic bone formation at the sternebra; bottom right panel shows ectopic bone formation at the hip joint and the caudal vertebrae.

(8) FIG. 8 illustrates ectopic bone formation at the sternebra (left panel) and the caudal vertebrae (right panel) for genetically modified mice comprising the conditional allele induced in mice administered tamoxifen (Tamoxifen #3, ID:915546).

(9) FIG. 9 illustrates the absence of ectopic bone formation in genetically modified mouse comprising the conditional allele, induced with tamoxifen (Tamoxifen #4, ID:904067).

(10) FIG. 10 illustrates ectopic bone formation at the sternebra (left panel) in genetically modified mice comprising the conditional allele induced by administration of tamoxifen (Tamoxifen #5, ID:840098).

(11) FIG. 11 illustrates ectopic bone formation at the sternebra (left panel) and knee joint (right panel) in genetically modified mice comprising the conditional allele induced by administration of tamoxifen (Tamoxifen #6, ID:863713).

(12) FIG. 12 illustrates primers and probes used in a loss of allele assay to genotype genetically modified mice comprising the conditional mutation in the Acvr1 gene; SEQ ID NOs are, from top to bottom: for the forward primer from top to bottom SEQ ID NO:1, SEQ ID NO:2; for the reverse primer from top to bottom SEQ ID NO:3, SEQ ID NO:4; for the probe SEQ ID NO:5, SEQ ID NO:6.

DETAILED DESCRIPTION

(13) Fibrodysplasia ossificans progressiva (FOP) is an autosomal dominant disorder of ectopic bone formation. Linkage studies in affected families reveal that the FOP gene maps to chromosome 2q23-24 where a 617G-to-A mutation (R206to-H) in the activation domain of activin A type I receptor gene (Acvr1) was found on all affected individuals examined in the studies (Shore et al., (2006) A recurrent mutation in the BMP type I receptor Acvr1 causes inherited and sporadic fibrodysplasia ossificans progressiva, Nat. Genet. 38:525-527), consistent with FOP being caused by constitutive activation of Acvr1 (Id.).

(14) Genetically modified mice are provided that express an Acvr1 protein comprising a modification that results in a disorder characterized by ectopic bone formation, e.g., FOP. Mice expressing the modified Acvr1 protein include mice that are not chimeric, e.g., mice whose genomes carry a (conditional) modification of the Acvr1 protein that results in ectopic bone formation in a mouse that expresses the modified Acvr1 protein.

(15) Certain mutations in the Acvr1 protein, e.g., the FOP-associated R206H mutation, are difficult if not impossible to create in the germline of mice due to embryonic or perinatal fatality associated with the mutation. Genetically modified mice are provided that comprise an COnditional-by-INversion (COIN) design that provides for a conditional inversion and removal of a wild-type exon and replacement of the wild-type exon with a mutant exon. This COIN design allows for forming a conditional allele by placement of a nucleic acid sequence encoding an inverted mutant exon to be placed next to a wild-type exon to be deleted. Through selection of recombinase recognition sites (RRS's), the inverted mutant exon is reversed to place it in reading frame whereas the wild-type exon is deleted. This COIN approach relies on the placement of incompatible RSS's (e.g., lox2372 and loxp) surrounding the wild-type and mutant exons. This COIN approach thus does not allow for expression of the (perinatal/embryonic) lethal mutation unless the COIN allele is acted upon by the selected recombinase(s). Another advantage of this COIN approach is permanent removal of the wild-type exon upon exposure to the selected recombinase, and thus no inverted repeat remains in the genome post-inversion. This is advantageous because it eliminates the possibility of re-inversion, because the remaining recombinase sites are incompatible (e.g., lox2372 and loxP). In this instance, humanization of the wild-type mouse exon also minimizes inverted repeat sequence, thus facilitating cloning steps and alleviating concerns of rearrangements during and after targeting.

(16) If a mouse bearing the COIN allele is bred to a recombinase-containing mouse, the (perinatal/embryonic) lethal mutation will express in the progeny in utero, thus confounding the goal of making an animal that can be studied which expresses the allele. Therefore, the mouse bearing the COIN allele is not bred with an unregulated recombinase-containing mouse. Instead, the mouse is bred with a mouse that contains a Cre-ER protein that this modified with T2 mutations (a Cre-ER.sup.T2 mouse), or modified to contain a Cre-ER.sup.T2 allele. The Cre-ER.sup.T2 protein is a Cre protein modified with an estrogen receptor sequence that comprises T2 mutations that render the Cre protein inactive (see, Indra, A. et al. (1999) Temporally-controlled site-specific mutagenesis in the basal layer of the epidermis: comparison of the recombinase activity of the tamoxifen-inducible Cre-ER.sup.T and Cre-ER.sup.T2 recombinases, Nucleic Acids Res. 27(22): 4324-4327; Feil, R. et al. (1997) Regulation of Cre Recombinase Activity by Mutated Estrogen Receptor Ligand-Binding Domains, Biochem. Biophys. Res. Commun. 237: 752-757; U.S. Pat. No. 7,112,715). A mouse comprising a conditional allele constructed with Cre-responsive RSS's as described herein, and containing a Cre-ER.sup.T2 allele, would therefore express the wild-type allele unless and until the mouse was exposed to tamoxifen to induce Cre activity. In this way, mice are made that contain a mutant Acvr1 allele in their germline but that do not express a mutant Acvr1 protein unless and until the mice are exposed to tamoxifen. Following exposure to tamoxifen, the Cre-ER.sup.T2 fusion protein is activated and the conditional allele converts to a mutant allele and, in various embodiments, the conversion to the mutant allele is irreversible, with deletion of the wild-type allele. In this manner, a mouse line containing an otherwise lethal Acvr1 mutation can be maintained essentially indefinitely, producing the desired genetic lesion and accompanying phenotype whenever desired. In various embodiments, a genetically modified mouse comprising the Acvr1 COIN allele is made by modifying a mouse ES cell to contain the COIN allele, and modifying the same ES cell to contain a gene encoding the tamoxifen-inducible Cre-ER.sup.T or Cre-ER.sup.T2, and using the ES cell as a donor cell to make a mouse that contains the COIN allele and the modified Cre gene. All of the references cited herein are hereby incorporated by reference.

(17) Engineering a Conditional ACVR1 Allele that is Germline Transmissible

(18) In order to engineer a mouse model of Fibrodysplasia Ossificans Progressiva (FOP), the R206H “classic FOP” mutation of human Acvr1 (Shore et al. (2006)) was engineered into the corresponding mouse gene, Acvr1. This mutation has already been modeled non-conditionally in the mouse, but the resulting chimeric mice (arising from blastocyst microinjection of the targeted ES cells) were unable to transmit the mutation through the germline, presumably due to embryonic or perinatal lethality (Chakkalakal, S. A. et al. (2012) An Acvr1 R206H knock-in mouse had fibrodysplasia ossificans progressiva, J. Bone and Mineral Res. 27:1746-1756). Prior to knowledge of this phenotype, and based on the phenotype of Acvr1 homozygous-null mice, which reveals a profound role of Acvr1 during development (Mishina et al. (1999) Multiple roles for activin-like kinase-2 signaling during mouse embryogenesis, Dev. Biol. 212:314-326), it was decided to engineer the Acvr1.sup.[R206H] mutation in a conditional manner in the mouse, utilizing a variation on FIEx (Schnutgen, F. et al. (2003) A directional strategy for monitoring Cre-mediated recombination at the cellular level in the mouse, Nat. Biotech. 21:562-565) and COIN (U.S. Pat. No. 7,205,148) methodologies.

(19) FIEx employs a pair of mutant Lox sites—referred to as a FIEx array—that are recognized by the same recombinase—Cre—but which to do not react with one another, and laid out in an A-B/[A-B] configuration, where the “[A-B]” is in the opposite strand with respect to “A-B”, to enable inversion of the DNA sequence flanked by the arrays. In its published embodiment, FIEx utilized sites LoxP and Lox511. Less known, however, is that in the presence of Cre a low level of recombination takes place between LoxP and Lox511. Therefore, different combinations of Lox site variants were tested, and the LoxP-Lox2372 combination were selected for the conditional allele described herein, because these two sites did not exhibit any cross-reactivity. An additional feature of FIEx is that the sequence that is contained within each array—i.e., between the LoxP and Lox2372 sites of each array—will be deleted upon action by Cre. The engineering of the allele of the invention (Acvr1.sup.[R206H]COIN allele) takes into account these two properties of FIEx. One embodiment of an conditional allele is illustrated in FIG. 2.

(20) Mouse Acvr1 displays a variety of splice variants (e.g., 201, 202, 001, 003, 004). exon 5, which is mutated in FOP, is shared by all protein-coding splice variants of Acvr1. In one embodiment, the genetically modified mouse comprises a modification of exon 5 of an isoform selected from the group consisting of 201, 202, 001, 003, and 004.

(21) The Acvr1.sup.[R206H]COIN allele was engineered by placing the mutant version of the R206-encoding exon of mouse Acvr1 (ENSMUSE00001021301) in the antisense strand, so that it is not incorporated into Acvr1's transcript. As the sequence encoded by exon 5 is required for Acvr1 function, this necessitated that an exon encoding for the wild type exon 5's sequence is also incorporated into the design (exon 5 is shared by all protein-coding splice variants of Acvr1). Furthermore, since exons are not recognized as such without accessory intronic sequences, both upstream and downstream of the exon had to be incorporated into both mutant and wild type R206-encoding exon. However, doing so would generate a large inverted repeat, and such DNA structures are inherently prone to recombination both during the genetic engineering steps required to build the targeting vector as well as post-targeting, in vivo (Holkers, M. et al. (2012) Nonspaced inverted DNA repeats are potential targets for homology-directed gene repair in mammalian cells, Nucleic Acids Res. 40:1984-1999). Furthermore, if the wild type mouse sequence of the R206-encoding exon and the upstream and downstream intronic sequence associated with it were retained intact, and precede the mutant exon, then this wild type region could act as a homology arm and be utilized during targeting in the mouse ES cells, thereby resulting in exclusion of the mutated exon from the targeted allele. Therefore, in order to address all these concerns the Acvr1.sup.[R206H]COIN allele was designed in a manner such that: (a) A large inverted repeat is avoided. To accomplish this, the R206-encoding exon (ENSMUSE00001021301) as well associated upstream and downstream intronic sequences were replaced with the corresponding region from human ACVR1. (b) The wild type mouse sequence of the R206-encoding exon (ENSMUSE00001021301) is preserved at the protein level. Given that the mouse and human protein sequence respectively encoded by exons ENSMUSE00001021301 and ENSE00001009618 differ by one amino acid, the human ENSE00001009618 exon sequence was altered so as to match the mouse protein sequence of exon ENSMUSE00001021301. (c) The introduced human sequence is removed in its entirety upon action with Cre. Therefore, in the “conditional-on” state—where the Acvr1.sup.[R206H] mutant gene is transcribed—no human sequences remain and hence any resulting phenotype cannot be attributed to the presence of extraneous sequence.

(22) More specifically, the region bounded by nucleotides 58474046 to 58474368 in mmuAcvr1 (i.e., nucleotides 58474046 to 58474368 of mouse Chromosome 2) where replaced with nucleotides 15863048 to 158630803 of hsaACVR1 (i.e., nucleotides 15863048 to 158630803 of human Chromosome 2), in a manner such that the introduced sequence, which includes hsaACVR1 exon ENSE00001009618 is transcribed as part of the resulting modified Acvr1.sup.[R206H]COIN locus. In addition, the coding sequence of the first amino acid of human exon ENSE00001009618 was replaced from aspartic acid (D) to glutamic acid (E) to correspond at the protein level to the exactly the same protein sequence as that encoded by mouse exon ENSMUSE00001021301. (This introduced human sequence is referred to hereafter as hsa_e5+.) Therefore, prior to inversion of the COIN element (mutated exon ENSMUSE00001021301 and associated upstream and downstream intronic sequences—see below), the resulting locus, Acvr1.sup.[R206H]COIN, should function as wild type.

(23) The R206H mutation was modeled by mutating exon ENSMUSE00001021301 in the corresponding position, by altering the codon defined by nucleotides 5847419 to 58474200 from CGC (coding for arginine) to CAC (coding for histidine). The resulting mutant exon, along with flanking intronic sequences upstream and downstream were placed 3′ to hsa_e5+ and in the antisense strand of mmuAcvr1, replacing nucleotides 58473775 to 58473879 of mmuAcvr1 in order to also create a small deletion and accommodate LOA probes (Gomez-Rodriguez, J. et al. (2008) Advantages of q-PCR as a method of screening for gene targeting in mammalian cells using conventional and whole BAC-based constructs, Nucleic Acids Res. 36:e117; Valenzuela, D. et al. (2003) High-throughput engineering of the mouse genome coupled with high-resolution expression analysis, Nat. Biotech. 21:652-659). (This introduced mutated mouse sequence is hereafter referred to as mmu_e5R206H+.)

(24) In order to enable Cre-dependent inversion of the mmu_e5R206H+ and simultaneous deletion of hsa_e5+, a combination of FIEx like Lox arrays where used such that: (a) hsa_e5+ is preceded by a LoxP site, and followed by a Lox2372 site. In this respect, hsa_e5+ is contained with the 5′ LoxP-Lox2372 FIEx-like array. (b) mmu_e5R206H+ is followed by the 3′ LoxP-Lox2372 FIEx-like array, but this array is engineered such that it is in a mirror image configuration to 5′ LoxP-Lox2372 FIEx-like array. This enables permanent inversion of mmu_e5R206H+ into the sense strand by Cre.

(25) When the resulting allele, Acvr1.sup.[R206H]COIN is exposed to Cre, the hsa_e5+ will be deleted and the mmu_e5R206H+ will be inverted into the sense strand. As a result, Acvr1.sup.[R206H] will be expressed in place of Acvr1.

(26) Genetically modified mice were genotypes employing a loss of allele assay (see, e.g., Valenzuela et al., (2003), supra). Primers and probes were as shown in FIG. 12 (Table 5).

(27) Phenotype of Acvr1.sup.R206HCOIN/+ Mice

(28) Acvr1.sup.R206HCOIN/+ mice are phenotypically normal but develop FOP after activation of the R206H conditional mutation.

(29) Based on published results with a non-conditional, simple knock-in Acvr1 R206H chimeric mouse (Chakkalakal et al., 2012) as well as the fact that FOP is an autosomal-dominant disorder (for a review see (Pignolo et al., 2011)), it was hypothesized that: (a) Unlike the non-conditional Acvr1.sup.R206H allele (Chakkalakal et al., 2012), targeted ES cells for Acvr1.sup.[R206H]COIN will produce VELOCIMICE®, i.e., F0 mice that are entirely derived from the targeted ES cells (Poueymirou et al. (2007) F0 generation mice fully derived from gene-targeted embryonic stem cells allowing immediate phenotypic analyses, Nat. Biotech. 25:91-99). (b) Unlike the non-conditional Acvr1.sup.R206H/+ chimeric mice (Chakkalakal et al., 2012), F0 Acvr1.sup.[R206H]COIN/+ mice will be phenotypically normal, and will transmit the Acvr1.sup.[R206H]COIN allele to the next generation. (c) Upon inversion of mutant exon bearing the R206H mutation into the sense strand—an action mediated by Cre recombinase—cells that have been converted to the Acvr1.sup.[R206H]INV/+ genotype will express the mutant Acvr1.sup.[R206H] allele as well as the wild-type allele, mirroring the situation in FOP patients. Along the same lines, the resulting Acvr1.sup.[R206H]INV/+ mice should overtime develop FOP-like symptoms.

(30) All of these hypotheses were born out. For example, ES cell clone 1649C-A2 gave rise to 16 VELOCIMICE® out of 19 mice generated using that clone (Table 1).

(31) TABLE-US-00001 TABLE 1 Acvr1.sup.[R206H]COIN/+ ES Cells Give Rise Mainly to Male F0 Mice Wholly Derived from Donor ES Cells Mouse ID Chimerism (%) 1649C-A2/758470 100 1649C-A2/758471 100 1649C-A2/758472 100 1649C-A2/758473 100 1649C-A2/758474 100 1649C-A2/758475 100 1649C-A2/758476 100 1649C-A2/758477 100 1649C-A2/758478 100 1649C-A2/758479 100 1649C-A2/758480 100 1649C-A2/758481 100 1649C-A2/758482 100 1649C-A2/758483 100 1649C-A2/758484 100 1649C-A2/758485 100 1649C-A2/758486 80 1649C-A2/758487 70 1649C-A2/758488 30

(32) Furthermore, these mice had no discernible phenotype and were able to reproduce and father Acvr1.sup.[R206H]COIN/+ F1 generation mice (Table 2).

(33) TABLE-US-00002 TABLE 2 F1 Mice Born to Acvr1.sup.[R206H]COIN/+ F0 Fathers Clone Name/ID Genotype Gender 1649C-A2/2251A-C6/840095 1649 Het 2251 Het M 1649C-A2/2251A-C6/840098 1649 Het 2251 Het M 1649C-A2/2251A-C6/845202 1649 Het 2251 Het M 1649C-A2/2251A-C6/845203 1649 Het 2251 Het F 1649C-A2/2251A-C6/845204 1649 Het 2251 Het F 1649C-A2/2251A-C6/845205 1649 Het 2251 WT F 1649C-A2/2251A-C6/845809 1649 Het 2251 WT F 1649C-A2/2251A-C6/863706 1649 Het 2251 WT F 1649C-A2/2251A-C6/863707 1649 Het 2251 WT F 1649C-A2/2251A-C6/863713 1649 Het 2251 Het M 1649C-A2/2251A-C6/863714 1649 Het 2251 WT M 1649C-A2/2251A-C6/897113 1649 Het 2251 WT F 1649C-A2/2251A-C6/897115 1649 Het 2251 WT F 1649C-A2/2251A-C6/897117 1649 Het 2251 Het F 1649C-A2/2251A-C6/904065 1649 Het 2251 WT M 1649C-A2/2251A-C6/904067 1649 Het 2251 Het M 1649C-A2/2251A-C6/904069 1649 Het 2251 WT F 1649C-A2/2251A-C6/904783 1649 Het 2251 WT M 1649C-A2/2251A-C6/904785 1649 Het 2251 WT F 1649C-A2/2251A-C6/907167 1649 Het 2251 WT F 1649C-A2/2251A-C6/915545 1649 Het 2251 WT M 1649C-A2/2251A-C6/915546 1649 Het 2251 Het M 1649C-A2/2251A-C6/964988 1649 Het 2251 Het F 1649C-A2/2251A-C6/964989 1649 Het 2251 Het F F1 generation Acvr1.sup.[R206H]COIN/+; Gt(ROSA26)Sor.sup.CreERt2/+ mice born to Acv1.sup.[R206H]COIN/+ F0

(34) From a phenotypic standpoint, Acvr1.sup.[R206H]COIN/+ mice appear normal, and display no discernible phenotypes. The same applies to Acvr1.sup.[R206H]COIN/+; Gt(ROSA26)Sor.sup.CreERt2/+ mice, which in addition to the Acvr1.sup.[R206H]COIN/+ allele also carry a CreER.sup.T2 transgene knocked into the Gt(ROSA26)Sor locus. This allows ubiquitous expression of an inactive version of Cre, one that is dependent upon tamoxifen for activation (Feil et al. (1997) Regulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains, Biochem. Biophys. Res. Commun. 237:752-757). This enables the activation of Cre at a specific point in time, and hence not only allows bypassing the embryonic lethality experienced with the conventional Acvr1.sup.[R206H] knock-in of but also empowers the investigator to choose the time of activation of the Acvr1.sup.[R206H] expression in the corresponding mice.

(35) In order to investigate whether Acvr1.sup.[R206H]COIN/+; Gt(ROSA26)Sor.sup.CreERt2/+ mice develop FOP after exposure to tamoxifen, we generated a small cohort and treated it with tamoxifen starting at approximately one year of age (Table 3); it is notable that by this age mice have completed their development, and therefore no modeling or development-related mechanisms are at play and therefore cannot contribute to the pathological process. Delivery of tamoxifen was by injection into the peritoneum using a 10 mg/mL solution in corn oil. Injections were performed daily for 8 days. In three mice (Mice 1, 2, and 3 of Table 3), a small piece of muscle was resected to induce injury.

(36) TABLE-US-00003 TABLE 3 Protocol for Cre-Mediated Tamoxifen-Dependent Activation of Acvr1.sup.[R206H]COIN Allele in Acvr1.sup.[R206H]COIN/+; Gt(ROSA26.sup.)CreERt2/+ Mice Sac- Age at Sac- rifice Mouse Daily Start Start End rifice Age Mouse ID Injection Day (yrs) Day Day (yrs) 1 840095 corn oil 1 0.9 8 143 1.3 2 845202 TAM* 1 0.9 8 143 1.3 3 915546 TAM 1 0.56 8 143 1.0 4 904067 TAM 1 0.61 8 143 1.0 5 840098 TAM 1 0.90 8 143 1.3 6 863713 TAM 1 0.80 8 143 1.2 TAM: tamoxifen

(37) All but one of the tamoxifen-treated mice developed ectopic ossification, mirroring what has been observed in FOP (Table 4). Although the specific cell type(s) that might be contributing to the disease process were not determined in this experiment due to the fact that the expression of CreER.sup.t2 is ubiquitous (a property imparted by the fact that it is expressed from the Gt(ROSA26)Sor locus), one of the important aspects of this work is that it removes the developmental aspects of FOP (which are not those most important to FOP's pathology, as they do not contribute to the devastating loss in quality of life the FOP patients experience), and shows that the ectopic bone formation that is the major post-natal hallmark of FOP pathology is independent of developmental processes.

(38) TABLE-US-00004 TABLE 4 Four Acvr1.sup.[R206H]COIN/+; Gt(ROSA26).sup.CreERt2/+ Mice Exposed to Tamoxifen Develop FOP-Like Skeletal Pathology Mouse Mouse ID Ectopic Bone Formation 1 840095 None* 2 845202 sternebra, hip joint, caudal vertebrae 3 915546 sternebra, hip joint, caudal vertebrae 4 904067 none 5 840098 sternebra 6 863713 sternebra, knee joint *Treated with corn oil (vehicle) only, not tamoxifen

(39) Ectopic ossification is shown in images of genetically modified mice as described herein exposed to tamoxifen (which display ectopic ossification). Mice that are genetically modified as described herein but not exposed to tamoxifen do not display ectopic ossification See, e.g., FIG. 3, FIG. 4, FIG. 5, FIG. 6, FIG. 7, FIG. 8, FIG. 10, and FIG. 11. Ectopic ossification is demonstrated in a variety of body areas. As shown in FIG. 9, one mouse showed no apparent ectopic bone formation.