Published 2006 Wiley-Liss, Inc.
genesis 44:601–610 (2006)
TECHNOLOGY REPORT
Targeting iCre Expression to Murine Progesterone Receptor Cell-Lineages Using Bacterial Artificial Chromosome Transgenesis Atish Mukherjee,1 Selma M. Soyal,2 David A. Wheeler,3,4 Rodrigo Fernandez-Valdivia,1 Jonathan Nguyen,1 Francesco J. DeMayo,1 and John P. Lydon1* 1
Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas Department of Internal Medicine, Krankenhaus Hallein, Hallein, Austria 3 Human Genome Sequencing Center, Baylor College of Medicine, Houston, Texas 4 Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 2
Received 28 August 2006; Revised 2 October 2006; Accepted 15 October 2006
Summary: Gene-targeting in embryonic stem cells has been the dominant genetic approach when engineering mouse models to query the physiologic importance of the progesterone receptor (PR). Although these models have been instrumental in disclosing the in vivo significance of the progesterone signaling pathway, generation of such mice exacts considerable expenditure of time, effort, and expense. Considering the growing list of new PR mouse models that are urgently required to address the next questions in progestin biology, bacterial artificial chromosome (BAC) recombineering in conjunction with transgenesis was evaluated as an alternative method to accelerate the creation of these models in the future. Using this approach, we describe the generation of three PR-BACiCre transgenic lines in which improved Cre recombinase (iCre) was targeted in-frame, downstream, and under the control of the PR promoter contained within a BAC transgene. Crossing with the ROSA26R revealed that the PR-BACiCre transgenic expresses active iCre only in cell-lineages that express the PR. The specificity of the PR-BACiCre transgene not only underscores the importance of BAC-mediated transgenesis as a quick, easy, and affordable method by which to engineer the next generation of PR mouse models, but also provides a unique opportunity to investigate transcriptional control of PR expression as well as PR structure-function relationships in vivo. genesis 44:601–610, (2006). Published 2006 Wiley-Liss, Inc.y Key words: progesterone receptor; bacterial artificial chromosome; iCre; transgenic
INTRODUCTION For over a decade, homologous recombination (or genetargeting) in murine embryonic stem (ES) cells has been the method of choice by which to generate mouse models as tools to understand the physiologic importance of the
progesterone receptor (PR) (Fernandez-Valdivia et al., 2005). Using this approach, a PR knockout (PRKO) mouse was generated (Lydon et al., 1995), in which both isoforms of PR (PR-A and PR-B) were simultaneously ablated. Subsequent use of cre-loxP engineering strategies allowed for the creation of PR isoform specific knockouts, in which PR-A and PR-B were selectively abrogated to generate PR-AKO and PR-BKO models, respectively (Mulac-Jericevic et al., 2000, 2003). More recent ‘‘knockin’’ approaches enabled the insertion of heterologous genes (i.e., the LacZ reporter or Cre recombinase) downstream and under the tight control of the endogenous murine PR promoter (Ismail et al., 2002; Soyal et al., 2005). Recently, gene-targeting approaches have also been employed to construct a mouse model that harbors a floxed (or conditional) PR allele to facilitate selective knockout of PR function in a tissue or celltype specific manner (Hashimoto-Partyka et al., 2006). Collectively, the above-mentioned mouse models have furnished unprecedented insights into the pleiotropic effects of the PR on the hypothalamic-pituitary-ovarian axis, mammary morphogenesis/tumorigenesis, behavior, glucose homeostasis, parity-induced thymic involution, and the vascular system (Fernandez-Valdivia et al.,
* Correspondence to: John P. Lydon, Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030. E-mail:
[email protected] Contract grant sponsors: National Institute for Child Health and Disease (NICHD), National Institutes for Health (NIH); Contract grant number: HD42311; Contract grant sponsor: National Cancer Institute; Contract grant number: CA077530; Contract grant sponsor: Susan G. Komen Breast Cancer Program; Contract grant number: BCTR-0503763. y This article is a US government work and, as such, is in the public domain in the United States of America. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/dvg.20257
602
MUKHERJEE ET AL.
2005). Although these mouse models have been essential in establishing the in vivo significance of the PR in tissue homeostasis and tumorigenesis, the cost, effort, and length of time required to generate such models represent a significant drawback as a routine genetic approach. Because bacterial artificial chromosomes (BACs) can (in most cases) accommodate intact mammalian genes with their full repertoire of controlling elements, can be modified with facility and precision at base-pair resolution by homologous recombination in bacteria, and can be isolated with common DNA purification procedures, transgenics based on BAC technology offer enormous potential as an alternative avenue to conventional geneknockout/knockin approaches (Copeland et al., 2001). With their capacity for large genomic inserts (up to 300 kb), clonal stability, and lack of chimerism, BAC-mediated transgenesis can, in many cases, accurately recapitulate the endogenous pattern of gene expression in an integration site-independent manner. We have isolated a murine BAC, which contains the complete PR transcriptional unit with associated control elements. Using a BAC recombineering strategy (Yu et al., 2000), Cre recombinase was knocked into Exon 1 of the Pgr gene to generate a modified PR BAC, in which Cre was inserted downstream of the PR promoter. Through BAC transgenesis, we demonstrate that Cre activity is tightly modulated by the PR-BAC promoter in progestin target tissues of the PR-BAC transgenic. These results not only provide proof of principle that BAC recombineering is a feasible alternative approach by which heterologous genes can be efficaciously targeted to the murine Pgr locus, but also suggest that this affordable method could easily be applied to quickly introduce deletions, insertions, and/or point mutations in the Pgr locus to study Pgr gene expression control and structure/function relationships in vivo. Of the six murine BACs that were identified to contain the Pgr gene, the BAC clone, RP23-422-I-15 (PR-BAC from hereon), was chosen due to the central position of the Pgr gene within the BAC, which allows for significant stretches of 50 and 30 flanking regions to be accommodated within this construct; importantly, these regions do not contain additional genes (Fig. 1a). To reduce epigenetic silencing, the improved Cre (iCre) was used for PR-BAC modification (Shimshek et al., 2002); the conventional Cre gene was used to generate our previously reported PRCre knockin model (Soyal et al., 2005). Standard recombineering techniques in bacterial hosts were employed to insert the iCre gene into Exon 1 of the Pgr gene contained within the BAC to generate the PR-BACiCre construct (Fig. 1b–c; Materials and Methods). The iCre insertion strategy was based on the genetargeting design used to generate our previously described PRLacZ knockin mouse (Ismail et al., 2002), in which the LacZ reporter gene-encoding b-galactosidase (b-gal)-was ‘‘knocked-into’’ Exon 1 of the endogenous Pgr gene. Importantly, this targeting strategy maintains the complete complement of transcriptional control elements (including cis regulatory sequences for the estrogenesis DOI 10.1002/dvg
gen receptor-a (ER)) within the PR promoter (Hagihara et al., 1994; Kraus et al., 1994). Using standard BAC transgenic approaches (Marshall et al., 2004), three independent transgenic lines were generated, which carried the PR-BAC iCre transgene (Fig. 2). Southern analysis revealed multiple copies of the PR-BACiCre transgene in two of these transgenic lines (1163 and 1150), whereas one copy of the transgene was carried by the transgenic line: 1153 (Fig. 2b). To evaluate iCre activity in situ, progeny from the three BAC transgenic lines were crossed with the ROSA26 reporter (R26R) (Soriano, 1999), in which iCre-mediated recombination is predicted to irreversibly activate the transcription of the LacZ reporter gene within the ROSA26 locus. Although all transgenics demonstrated a similar spatial activity profile for iCre (data not shown), iCre activity appeared moderately stronger in lines 1163 and 1150 (presumably due to a higher copy number of the integrated BAC in these lines). Importantly, all three BAC transgenics, which were maintained as heterozygotes for the BAC transgene insertion, were functionally normal. For the remainder of this report, our characterization of the transgenic line 1150 (referred to as PR-BACiCre/R26R from hereon) is described. It should be noted that because Cre-loxP excision is irreversible in the PR-BACiCre/R26R transgenic, descendents of PR positive cells, in which the iCre recombinase is functional, will inherit and constantly express the R26R LacZ gene. To test the specificity of the BAC transgenic model, the PR-BACiCre/R26R mammary gland, female reproductive tract, and pituitary gland (archetypal progestin target sites (Fernandez-Valdivia et al., 2005)) were examined for PR-BAC promoter dependent iCre activity. Although previous PRKO studies highlighted the critical importance of the P proliferative signal to mammary morphogenesis and tumorigenesis (Lydon et al., 1999), the PRLacZ and PRCre knockin models spatiotemporally traced the activity profile of the mammary PR promoter in situ (Ismail et al., 2002; Soyal et al., 2005). Both the PRLacZ and PRCre models detected robust mammary PR promoter activity only in the epithelial compartment of the mammary gland, observations that concur with previous immunohistochemical studies (Ismail et al., 2004). As expected, an identical b-gal activity profile was observed in the mammary gland of the adult (10-week-old) nulliparous PR-BACiCre/R26R bigenic (Fig. 3a,b). Close examination of X-gal-stained sections revealed that the majority of luminal epithelial cells scored positive for b-gal activity in the PR-BACiCre/R26R mammary gland (Fig. 3c). When compared with mammary PR expression, X-gal-stained mammary sections from the PR-BACiCre/ R26R transgenic reveal an identical spatial expression pattern, in which only luminal epithelial cells score positive for b-gal activity (Fig. 3c–h). Closer scrutiny shows that (like the regional expression pattern for mammary PR) a subset of luminal epithelial cells score negative for PR-BAC promoter driven iCre activity (Fig. 3e,f). These results demonstrate that the mammary expression pro-
iCRE EXPRESSION TO MURINE PR CELL-LINEAGES
603
FIG. 1. Recombineering of the PR-BAC to insert iCre into Exon 1 of the Pgr gene. (a) Diagram of the PR-BAC clone (RPCI-23-422I 15). Although not to scale, the eight exons (listed 1–8) of the murine Pgr gene are indicated by black boxes; the coding region encompasses 65293 bp. The 50 and 30 regions outside the Pgr locus span 65900 and 66443 bp, respectively. The PR-BAC clone was propagated in the pBACe3.6 vector; pertinent restriction sites are included. The PCR result below shows the presence of amplicons A-I, which arise from PCR primers that span the exon/intron boundaries of Exon 1–8 of the mouse Pgr gene, respectively; the production of these amplicons supports the presence of all eight exons of the Pgr gene in the PR-BAC clone (M indicates a marker lane; PCR sequences and amplification conditions are described in the Materials and Methods section). (b) Recombineering strategy to target the iCre cassette into Exon 1 of the Pgr gene contained within the PR-BAC (Step 1). Note that the shuttle vector, containing the iCre and FRT-KanR-FRT cassette (flanked by 60 bp arms of homology to the Pgr gene (black boxes)), inserts the iCre and FRT-KanR-FRT genes 354 bp downstream of the initiating ATGB codon; the ATGA codon (with 370 bp of sequence) is replaced by this insertion (Step 2). The FRT-KanR-FRT cassette is selectively removed by flp recombinase when the PR-BACiCre-FRT-KanR-FRT is introduced into the 294-FLP deleter bacterial strain (Step 3). (c) Schematic shows the PCR strategy to confirm that the iCre-FRT-KanR-FRT cassette was inserted into the PR BAC in DY380 cells. Note the PCR result in the lower left panel reveals that the recombineered PR-BAC clone produces the expected PCR amplicons: PCR-A, PCR-B, and PCR-C (Lanes 1, 2, and 3, respectively; lane M denotes a marker lane), indicating that the iCre-FRT-KanR-FRT cassette was inserted in the correct location in the PR-BAC. The lower right panel shows the PCR result after the PR-BACiCre-FRT-KanR-FRT is transferred to the 294-FLP deleter bacterial strain. Lanes 1 and 3 are positive controls for PCR amplification of PCR-B and PCR-C products, respectively (positive controls consisted of PR-BACiCre-FRT-KanR-FRT DNA in DY380 cells). Lanes 2 and 4 show the inability to produce PCR-B and PCR-C amplicons from the modified PR BAC (PR-BACiCre) following flp recombinase in the 294-FLP deleter host. This result demonstrates that the FRT-KanR-FRT cassette was successfully removed from the modified PR-BAC in the 294-FLP deleter strain.
genesis DOI 10.1002/dvg
604
MUKHERJEE ET AL.
FIG. 2. Generation of PR-BACiCre transgenic mouse lines. (a) Schematic outlining the Southern strategy to screen for founder mice (and their progeny) carrying the PR-BACiCre transgene. For mice positive for the BAC insertion, genomic DNA digested with HindIII and hybridized with a 50 Southern probe (located in Exon 1 (immediately upstream of the iCre insertion)) is expected to yield a 6.0 kb band attributable to the endogenous Pgr gene as well as a 3.3 kb band arising from the PR-BACiCre transgene. The shorter hybridizing band from the integrated BAC is due to a novel Hindlll site located in the iCre cassette. Wild type (WT) siblings should only display the 6.0 kb hybridizing band. (b) A Southern shows the expected genotypic results for WT and transgenic progeny from three separate PR-BACiCre transgenic founders: 1163; 1150; and 1153. When compared with the intensity of the WT (6.0 kb) hybridizing band in all three samples, transgenic lines 1163 and 1150 contain at least three and two copies of the PR-BACiCre transgene, respectively; the 1153 line contains one copy of the BAC transgene.
file for iCre from the PR-BACiCre transgene is not only identical to endogenous PR but also active in this tissue. In the case of the PR-BACiCre/R26R female reproductive tract, strong iCre activity was observed uniformly throughout the uterine horn as well as in the oviduct
and preovulatory follicle (Fig. 4a,b). PR-BACiCre/R26R uterine sections stained for b-gal activity clearly show strong X-gal staining in both the luminal epithelial and subepithelial stromal compartments (Fig. 4c,d), cellular compartments known to express uterine PR (Tibbetts et al., 1998). Our group and others demonstrated that ovarian PR expression is rapidly and transiently induced in granulosa cells of the preovulatory follicle by human chorionic gonadotropin (hCG) in pubescent mice that were previously treated with pregnant mare serum gonadotropin (PMSG) (Ismail et al., 2002; Robker et al., 2000); hCG and PMSG are functionally equivalent to pituitary-derived luteinizing hormone (LH) and follicle stimulating hormone (FSH), respectively. As reported for the PRLacZ and PRCre knockins (Ismail et al., 2002; Soyal et al., 2005), iCre activity was not detected in the ovary of the PMSGprimed 21-day-old PR-BACiCre/R26R mouse (Fig. 4e); however, strong X-gal staining was evident in the oviduct (black arrow) at this time. After 8 h post-hCG injection, iCre activity becomes visible in X-gal-stained ovarian whole-mounts (Fig. 4f, black arrows). After ovulation (14–16 h after hCG treatment in the mouse) and subsequent luteinization, iCre activity is present in resultant corpora lutea (Fig. 4g,h). As previously reported for the PRCre knockin (Soyal et al., 2005), this observation is significant in that PR expression is not detected in the mouse/rat corpus luteum and therefore highlights the PR-BACiCre/R26R transgenic (as with the PRCre/R26R bigenic (Soyal et al., 2005)) as an approach to lineally trace the developmental fate of mural granulosa cells as they progress from a PR positive granulosa cell to a PR negative luteal cell. Whole-mount X-gal staining of pituitary glands from adult PR-BACiCre/R26R females clearly reveals iCre activity is regionally restricted to the anterior lobe (Fig. 4i,j); pituitaries from age-matched PR-BACiCre/R26R males did not exhibit X-gal staining (data not shown). The PR-BACiCre/R26R pituitary staining pattern is identical to pituitary X-gal staining profile previously reported for the PRLacZ and PRCre knockin models (Ismail et al., 2002; Soyal et al., 2005). Importantly, tissues known to be negative for PR expression (spleen, liver, kidney, and heart) were shown not to stain for b-gal activity in
FIG. 3. Confinement of iCre activity to the luminal epithelial compartment of the PR-BACiCre/R26R mammary gland. (a) X-gal-stained whole-mount of an inguinal mammary gland from a 10-week-old nulliparous PR-BACiCre/R26R mouse reveals b-gal activity throughout the arborized ductal epithelial network; LN indicates the position of the lymph node. A higher magnification of a region denoted by the black arrow is shown in (b). Note: b-gal activity is regionally confined to the epithelial ductal system as indicated by the black arrowhead; b-gal activity is not detected in the surrounding stromal compartment. (c) An X-gal-stained transverse section of a medial duct clearly reveals bgal activity is restricted to the luminal epithelial compartment (blue arrow); b-gal activity was not detected in the stroma (red arrow) or in the periductal myoepithelial compartment (black arrow). (d) A serial section of the same tissue shown in (c) stained for PR immunoreactivity discloses an identical spatial expression pattern as observed for iCre (compare (d) with (c)); PR is only detected in the luminal epithelial compartment (brown arrow) and not in the myoepithelial or stromal compartments (black and red arrows, respectively). (e) and (f) are higher magnification images of longitudinal sections of similar ductal structures shown in (c) and (d), respectively. Again note: b-gal activity ((e) (blue arrow)) and PR expression ((f) (brown arrow)) are localized to the luminal epithelial compartment; the myoepithelium scores negative for b-gal activity and PR expression (black arrow). The asterisk in (e) and (f) indicates a luminal epithelial cell that is negative for b-gal activity and PR expression, respectively. (g, h) Sections of distal tips of epithelial ducts located at the periphery of the fat-pad show b-gal activity and PR expression localized to the luminal epithelial compartment (blue and brown arrows, respectively) and not to the periductal fibroblasts or stroma (black and red arrows, respectively). Scale bar in (c) applies to (d) whereas the scale bar in (g) applies to (h). genesis DOI 10.1002/dvg
iCRE EXPRESSION TO MURINE PR CELL-LINEAGES iCre
the PR-BAC /R26R, confirming the specificity of this animal model. Collectively, these results demonstrate that iCre expression in the PR-BACiCre transgenic is as tightly controlled by the PR-BAC promoter as previously reported
605
for heterologous genes regulated by the endogenous PR promoter. The observation that the PR-BACiCre transgenic performs faithfully supports the conclusion that most of the requisite transcriptional regulatory elements are contained within this single BAC clone. Conversely, this fea-
FIG. 3 genesis DOI 10.1002/dvg
606
MUKHERJEE ET AL.
FIG. 4. iCre is active in a subset of celllineages of the PR-BACiCre/R26R ovary, oviduct, uterus, and pituitary gland. (a, b) X-gal-stained whole-mounts of the ovary (O), oviduct (*), and uterus (U) are shown for the R26R and PR-BACiCre/R26R mouse, respectively. Although the R26R scores negative for iCre activity (a), note strong iCre activity in the preovulatory follicle of the ovary (black arrowhead), oviduct, and uterus of the PR-BACiCre/ R26R bigenic. (c, d) X-gal-stained uterine sections from the R26R and PR-BACiCre/R26R respectively. As expected, the R26R uterus is negative for iCre activity (c) whereas the PRBACiCre/R26R uterus scores positive for iCre activity in luminal epithelial (LE) and stromal (S) compartments (d). (e) Whole-mount X-galstained PR-BACiCre/R26R ovary (48 h after PMSG administration). Note: iCre activity in the oviduct (black arrowhead) but not in the ovary (red arrowhead). (f) Eight hours after hCG administration, the PMSG-treated PRBACiCre/R26R ovary exhibits iCre activity in the preovulatory follicle (black arrowheads); white asterisk denotes iCre activity in the oviduct. (g) Twenty-four hours after hCG administration, the PMSG treated PR-BACiCre/ R26R ovary shows robust expression in the corpora lutea (black arrowheads); again, white asterisk indicates iCre activity in the oviduct. (h) X-gal-stained section of a corpus luteum (CL) shown in (g); note most luteal cells score positive for iCre activity. (i, j) X-gal-stained whole-mounts of the pituitary gland obtained from the R26R and PR-BACCre/R26R female mouse respectively. Although the R26R pituitary registers negative for iCre activity (i), the PR-BACiCre/R26R pituitary gland (j) exhibits iCre activity only in the anterior lobe (a) but not in the neural lobe (n). Scale bars in (c) and (e) apply to (d) and to (f–g), respectively; scale bar in (i) also applies to (j).
genesis DOI 10.1002/dvg
iCRE EXPRESSION TO MURINE PR CELL-LINEAGES
607
FIG. 5. Genomic organization and comparative analysis of the murine Pgr locus. The coordinates (nucleotide numbers: 8860593-9063389) on mouse chromosome 9 (chr. 9) denote the position of the PR BAC (RPC123-422I 15). The location of the murine Pgr gene within the PR BAC (RPC123-422I 15) is diagrammed in light blue; direction of transcription is from left to right. Four custom tracks at the top show the positions of putative full-length (not half sites) response elements for ER, PR, AP-1, and Sp-1 detected computationally in this study (vertical bars within each track). The bars are rendered in gray scale corresponding to the relative score of the site given by the detection program (methods); darker shades represent higher scores. The track labeled ‘‘Conservation’’ is a histogram on a vertical scale of 0 to 1 indicating the probability of phylogenetic conservation with the mouse at each base position from the nine species listed. Gaps in the conservation track for each species represent the location of repeat sequences as indicated by RepeatMasker. The conservation track for each species displays conservation in a gray scale (white is 0; black is 1) for the pairwise comparison with the mouse sequence.
ture is not shared by the murine ER (Esr1) locus, which spans a number of BAC clones (Swope et al., 2002), thereby precluding ER BAC transgenesis. Although in vitro studies have identified a number of cis regulatory regions in the Pgr locus (i.e., response elements for ER, PR, AP-1, and SP-1; Hagihara et al., 1994; Kraus et al., 1994; Petz et al., 2002; Sriraman et al., 2003) that may control PR expression in vivo, these sites have yet to be validated in a physiological context. Examination of the PR-BAC in silico clearly reveals that many more sites for these transcription factors could also be implicated in PR transcriptional control (Fig. 5). Demonstrating that this BAC region is sufficient to faithfully control PR promoter activity in vivo provides confidence that these and additional regulatory elements can now be functionally evaluated for physiological significance, possibly by first the judicious deletion of selected conserved regions within the BAC (Fig. 5), followed further by point mutation analysis. In conclusion, our first line of studies demonstrates that the PRiCre BAC transgenic operates as efficiently and specifically as our recently described PRCre knockin model (Soyal et al., 2005). However, while the knockin and BAC transgenic both target Cre activity to cell-lineages that score positive for PR expression, the following features distinguish the PRiCre BAC transgenic from the PRCre knockin: (1) instead of Cre, iCre was targeted to the Pgr locus; (2) rather than one copy, multiple copies of iCre (driven by the PR-BAC promoter) were inserted into the murine genome; (3) modification of the endogenous Pgr locus was not required to generate the PRiCre BAC transgenic (monoallelic inactivation of the Pgr locus was essential to generate the PRCre knockin); and im-
portantly, (4) instead of 2 years, less than 6 months were required to generate the first PR BACiCre transgenic founders at a fraction of the cost and effort that was expended to create the PRCre knockin. Although the PRCre knockin continues to provide essential insights into PR’s action in vivo (Lee et al., 2006; Mukherjee et al., 2006), the studies described herein support BAC-mediated transgenesis as an important genetic approach to consider when designing the next generation of mouse models required to address the next questions raised by studies on progestin biology. For example, targeting genes that encode the enhanced green fluorescent protein ((EGFP) and spectral variants thereof) or the reverse tetracycline transactivator (rtTA) to the Pgr locus could easily be accomplished using BAC transgenic approaches described herein. A PR-EGFP BAC transgenic would be useful not only for cell-lineage tracing in vivo, but also in combination with fluorescence activated cell sorting (FACS), this model could be used to isolate viable PR positive cells for studies that may require microarray or transplantation approaches. On the other hand, the PR-rtTA BAC transgenic (in conjunction with the TET-ON system (Lewandoski, 2001)) would allow the turning on of a target gene of choice specifically in PR positive or PRKO cell-lineages in the mouse. We have demonstrated that all transcriptional control elements as well as the complete coding region of the Pgr gene are contained in a single BAC clone. This finding suggests that transgenesis using this PR BAC clone, containing a deletion, insertion, or point mutation could be exploited to identify and characterize critical distal and proximal enhancer elements required to control genesis DOI 10.1002/dvg
608
MUKHERJEE ET AL.
PR expression in a specific target tissue. Similar BAC transgenic approaches could also be used for structurefunction analyses of the PR protein; an area of investigation that has relied heavily on in vitro studies, which in many cases suggest no clear physiological relevance. In addition to the murine PR BAC, we recently isolated a human (h) BAC clone, which by sequencing and in silico analyses contains the entire human PR (PGR) gene including requisite control regions. The similar structural organization of the murine and human BACsin terms of the central positioning of the PGR gene, which ensures the inclusion of all relevant (50 and 30 ) regulatory sequences in the BAC transgene for correct spatiotemporal expression in vivo-suggests that similar transgenic approaches could be applied to the hPR BAC to generate a number of important mouse models for studying hPR regulation and function in vivo. These include the generation of a ‘‘humanized’’ PR mouse, in which hPR is expressed in the PRKO mouse as well as mouse models to address the physiological relevance of epigenetic modifications of the PGR locus (Sasaki et al., 2001; Xiong et al., 2005) as well as the functional significance of specific single nucleotide polymorphisms in the PGR gene that have recently been associated with endometrial and/or breast cancer predisposition in the human population (De Vivo et al., 2002, 2003). In summary, the PR BACiCre transgenic provides proofof-principle that BAC recombineering in combination with transgenesis can easily be applied to effectively target heterologous genes to the Pgr locus with a minimum of expended time, effort, and expense. Moreover, the PR BAC represents an important resource for the future study of Pgr gene expression and receptor structure/function relationships in vivo.
MATERIALS AND METHODS BAC Recombineering and Transgenesis Cloned into the pBACe3.6 vector (GenBank Accession # U80929 (Frengen et al., 1999)), the murine BAC clone, RP23-422-I-15 (RPCI-23 female mouse (C57BL/6J) BAC Library, BAC PAC Resource Center at Children’s Hospital, Oakland Research Institute, Oakland, CA), contains an insert size of 197.6 kb of genomic DNA from the 9qA1 region of chromosome 9; a region syntenic to the human PGR locus located on chromosome 11q22.1–22.3 (AC020735; www.ensemble.org (Soyal et al., 2002)). Sequence, Southern, and PCR analysis revealed that the RP23-422-I-15 BAC (termed PR BAC hereon) contains the entire coding region of the murine Pgr gene, including its transcriptional control elements. The following PCR primers were used to confirm the presence of all eight exons in the progesterone receptor bacterial artificial chromosome (PR-BAC) clone: 50 UTR/Exon 1 junction: forward: 5 0 -CTCTGCCCCTATCACCGGC-3 0 ; reverse: 50 -GGGACCTGAGTCCAAGCGTG-30 (Amplicon A (198 bp)); Exon 1/Intron 1 junction: forward: 50 -CGGgenesis DOI 10.1002/dvg
CCTCAATGGGCTCCCGC-30 ; reverse: 50 -CAATGTTTGGGAGAATGGTACAC-3 0 (Amplicon B (300 bp)); Intron 1/Exon 2 junction: forward: 50 -GGGAATTCATGAGTTCAAGG-3 0 ; reverse: 5 0 -CTTCCATTGCCCTCTTAAAG-3 0 (Amplicon C (374 bp)); Intron 2/Exon 3 junction: forward: 5-GACCCTGAACTCAAAGGTGAG-30 ; reverse: 50 CAAGGAGGACTGCCCCTTCTC-30 (Amplicon D (511 bp)); Intron 3/Exon 4 junction: forward: 50 -GATGATGAGTGACAGCACAGTG-30 ; reverse: 50 -GGAGAGCAACACCGTCAAGG-30 (Amplicon E (171 bp)); Intron 4/Exon 5 junction: forward: 50 -GTTAACATGGTTCATG-30 ; reverse: 50 -CATTTAGGATTAGATCAGG-30 (Amplicon F (197 bp)); Intron 5/Exon 6 junction: forward: 50 -CAACAGGAAAGAGAGTTCC-30 ; reverse: 50 -GTAAGGTGCCAAGTGTCTTTAC-30 (Amplicon G (288 bp)); Intron 6/Exon 7 junction: forward: 50 -GTTAGTTGCCTAGCTCAGG-30 ; reverse: 50 -GATAATGGACTGAACCTGTG-30 (Amplicon H (403 bp)); and Intron 7 and 30 UTR (primers flank Exon 8): forward: 50 -GCAGTTCATTCAAGGGATGC-30 ; reverse: 50 -GACACATGACCTGACCATC-30 (Amplicon I 487 bp)); PCR was performed using pfu Turbo DNA polymerase1 (Stratagene, La Jolla, CA). In accordance with the previously described redmediated recombineering methods (Lee et al., 2001; Yu et al., 2000), the iCre recombinase cassette was inserted into Exon 1 of the PR BAC gene. The shuttle (or targeting vector) contained the iCre-FRT-Kanamycin resistance (KanR) gene-FRT fragment; the drug selection cassette was flanked by two FRT sites in the same orientation. The iCre-FRT-KanR-FRT fragment was flanked by the following PR homologous arms: 50 arm: 50 -GATCCAATTCCAGACCCCCGGAGAACAGCAGACTCTTAGACAGTGTCTTAGACTCGTTGTTA-30 and 30 arm: 50 -GTCGACGAGCACTGGAAGGCACCGGCCAGGGAGGAGGAGTCGCAGCCAACGCGCCGTCAGCGGCCGCTAGC-30 ); the sequences of these flanking arms reside on either side of the insertion site in Exon 1 of the mouse Pgr gene. The location of the iCre insertion site is identical to that used to generate our previously described PRLacZ knockin reporter mouse (Ismail et al., 2002). Following redmediated homologous recombination in the recombinogenic bacterial strain: DY380, the KanR cassette within the modified PR BAC (PR-BACiCre-FRT-KanR-FRT) was removed in the 294-FLP deleter bacterial host (Buchholz et al., 1996) to generate the PR-BACiCre in which Exon 1 of the PR BAC gene only retains the iCre insertion. The PCR-A, -B, and -C amplicons, used to confirm the insertion of the iCre-FRT-KanR-FRT cassette as well as subsequent removal of the FRT-KanR-FRT gene, were generated using the following primers: PCR-A (forward: 50 -GGAGGGAGCTTTCTCTGG-30 (located 248 bp downstream from the ATGB) and reverse: 50 -CAGATCTCCTGTGCAGCATG-30 (located in the 50 region of the iCre gene)); PCR-B (forward: 50 -CATAGTGATGAACTACATCAG-30 (located in the 30 region of the iCre gene) and reverse: 50 -CACTGAGCGTCAGACCAAGTT-30 (located in the 50 region of the KanR gene)); and PCR-C (forward: 50 GACTTTCCACACCCTAACTGAC-30 (located in the 30 region of the KanR gene) and reverse: 50 -CCTCCAGCAG-
iCRE EXPRESSION TO MURINE PR CELL-LINEAGES 0
CTGCCGGGTGCG-3 (located 973 bp downstream of ATGB)). PR-BACiCre DNA was purified using the NucleoBond BAC Maxi kit (BD Biosciences Clontech, Mountain View, CA) and linearized with the homing endonuclease PI-SceI (New England Biolabs, Ipswich, MA) prior to zygote microinjection. To generate PR-BACiCre transgenic mouse lines, the linearized PR-BACiCre construct (1 ng/ml) was microinjected into the male pronucleus of FVB/N inbred embryos, which were subsequently implanted into pseudopregnant (C57BL6) foster mothers using standard approaches (Marshall et al., 2004). Transgenic founder mice and their progeny were identified by Southern blotting and PCR analysis with PR and Cre specific probes and primers, respectively. Transgenics, in a mixed strain background of FVB/N and C57BL6, were maintained as heterozygotes for the BAC insertion. A basic intercross with the R26R mouse generated the PR-BACiCre/R26R bigenic. The iCre gene contains its own initiating ATG and nuclear localization signal (NLS) (Shimshek et al., 2002); however, the LacZ reporter within the R26R locus does not contain a NLS (Soriano, 1999). Staining for b-Gal Activity and PR Expression Detection of b-gal activity and PR expression by X-gal staining and immunohistochemistry, respectively, is described elsewhere (Ismail et al., 2002). Using an AxioCam MRc5 camera, digital images were obtained of X-gal-stained whole-mounts and sections thereof using Axioplan 2 and Stemi 2000-C microscopes (Carl Zeiss, Jena, Germany), respectively. Captured digital images were initially processed using Metavue Software 4.6r9 (Universal Imaging, Downington, PA); final image montages were assembled using Photoshop1 CS (Adobe Systems, San Jose, CA). Mice and Hormone Treatments Animals were maintained in a temperature controlled (228C 6 28C) room, with a 12-h light, 12-h dark photocycle, and provided rodent chow meal (Purina Mills., St. Louis, MO) with fresh water, ad libitum. A superovulation hormonal regimen was followed according to our previous report (Ismail et al., 2002). Animal care and manipulations were approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine and were in accordance with practices and procedures detailed in the Guide for Care and Use of Laboratory Animals (NIH publication 85-23). Computational Techniques The estrogen response element (ERE) set was obtained from transcription factor binding sites made available in the Dragon ERE Finder version 1.0 (http//research.i2r. a-star.edu.sg/DRAGON/TFAM/) (Pan et al., 2004). Progesterone response elements (PREs) were collected from the literature (DeMayo and Wheeler, unpublished data). Activator Protein-1 (AP-1) and Sp1 binding sites
609
were obtained from Quality 4 level entries in the Transcription Factor Database: TRANSFAC1 (Matys et al., 2006). Each set of sites, ERE, PRE, AP-1, and SP-1, were used to generate position-specific scoring matrices (PSSM) using the program Consensus (Hertz and Stormo, 1999); a PSSM encodes the frequency of each base at each position of a given DNA element. Comparing the PSSM to a target sequence of interest enables the identification of matching elements and assigns a probability score to each match. The program Patser was used to search chromosome 9 sequences (mouse genome March 2006 build) for full-length binding sites for ER, PR, AP-1, and SP-1; sites that matched the PSSM with a log probability of 8 or less were saved. A profile hidden Markov model (Durbin et al., 1998, 1999) was constructed using HMMer version 2.3.2 (Eddy, 2003) and used to scan chromosome 9 for matching transcription factor binding sites; all sites with an expected value of 0.01 or less were saved. Binding sites for all four transcription factors were displayed at the Pgr locus on chromosome 9 (one track for each factor) using the custom tracks feature provided by the University of California Santa Cruz (UCSC) Genome Browser (http://www.genome. ucsc.edu). ACKNOWLEDGMENTS The technical assistance of Jie Li, Yan Ying, Jie Han, Jessica Li, and Jinghua Li is gratefully acknowledged. We thank Dr. Donald L. Court, Dr. A. Francis Stewart, Dr. Rolf Sprengel, and Dr. Philip Soriano for providing the bacterial recombinogenic strain (DY380), the 294Flp deleter bacterial strain, the iCre recombinase, and the R26R mouse, respectively. LITERATURE CITED Buchholz F, Angrand PO, Stewart AF. 1996. A simple assay to determine the functionality of Cre or FLP recombination targets in genomic manipulation constructs. Nucleic Acids Res 24:3118–3119. Copeland NG, Jenkins NA, Court DL. 2001. Recombineering: A powerful new tool for mouse functional genomics. Nat Rev Genet 2:769– 779. De Vivo I, Hankinson SE, Colditz GA, Hunter DJ. 2003. A functional polymorphism in the progesterone receptor gene is associated with an increase in breast cancer risk. Cancer Res 63:5236–5238. De Vivo I, Huggins GS, Hankinson SE, Lescault PJ, Boezen M, Colditz GA, Hunter DJ. 2002. A functional polymorphism in the promoter of the progesterone receptor gene associated with endometrial cancer risk. Proc Natl Acad Sci USA 99:12263–12268. Durbin R, Eddy S, Krogh A, Mitchison G. 1998. The theory behind profile HMMs: Biological sequence analysis: Probabilistic models of proteins and nucleic acids. Cambridge: Cambrige University Press. Durbin R, Eddy S, Krogh A, Mitchison G. 1999. Profile HMMs for sequence families in biological sequences analysis: Probabilistic models of protein and nucleic acids. Cambridge: Cambridge University Press. pp 102–132. Eddy S. 2003.http://hmmer.wustl.edu. Fernandez-Valdivia R, Mukherjee A, Mulac-Jericevic B, Conneely OM, DeMayo FJ, Amato P, Lydon JP. 2005. Revealing progesterone’s role in uterine and mammary gland biology: Insights from the mouse. Semin Reprod Med 23:22–37. Frengen E, Weichenhan D, Zhao B, Osoegawa K, van Geel M, de Jong PJ. 1999. A modular, positive selection bacterial artificial chromosome vector with multiple cloning sites. Genomics 58:250–253. genesis DOI 10.1002/dvg
610
MUKHERJEE ET AL.
Hagihara K, Wu-Peng XS, Funabashi T, Kato J, Pfaff DW. 1994. Nucleic acid sequence and DNase hypersensitive sites of the 50 region of the mouse progesterone receptor gene. Biochem Biophys Res Commun 205:1093–1101. Hashimoto-Partyka MK, Lydon JP, Iruela-Arispe ML. 2006. Generation of a mouse for conditional excision of progesterone receptor. Genesis 44:391–395. Hertz GZ, Stormo GD. 1999. Identifying DNA and protein patterns with statistically significant alignments of multiple sequences. Bioinformatics 15:563–577. Ismail PM, DeMayo FJ, Amato P, Lydon JP. 2004. Progesterone induction of calcitonin expression in the murine mammary gland. J Endocrinol 180:287–295. Ismail PM, Li J, DeMayo FJ, O’Malley BW, Lydon JP. 2002. A novel lacZ reporter mouse reveals complex regulation of the progesterone receptor promoter during mammary gland development. Mol Endocrinol 16:2475–2489. Kraus WL, Montano MM, Katzenellenbogen BS. 1994. Identification of multiple, widely spaced estrogen-responsive regions in the rat progesterone receptor gene. Mol Endocrinol 8:952–969. Lee EC, Yu D, Martinez de Velasco J, Tessarollo L, Swing DA, Court DL, Jenkins NA, Copeland NG. 2001. A highly efficient Escherichia colibased chromosome engineering system adapted for recombinogenic targeting and subcloning of BAC DNA. Genomics 73:56–65. Lee K, Jeong J, Kwak I, Yu CT, Lanske B, Soegiarto DW, Toftgard R, Tsai MJ, Tsai S, Lydon JP, Demayo FJ. 2006. Indian hedgehog is a major mediator of progesterone signaling in the mouse uterus. Nat Genet 38: 1204–1209. Lewandoski M. 2001. Conditional control of gene expression in the mouse. Nat Rev Genet 2:743–755. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA Jr, Shyamala G, Conneely OM, O’Malley BW. 1995. Mice lacking progesterone receptors exhibit pleiotropic reproductive abnormalities. Genes Dev 9:2266–2278. Lydon JP, Ge G, Kittrell FS, Medina D, O’Malley BW. 1999. Murine mammary gland carcinogenesis is critically dependent on progesterone receptor function. Cancer Res 59:4276–4284. Marshall VM, Allison J, Templeton T, Foote SJ. 2004. Generation of BAC transgenic mice. In: Zhao S, Stodolsky M, editors. Methods in molecular biology: Bacterial artificial chromosomes-functional studies. Totowa, NJ: Humana Press. pp 159–182. Matys V, Kel-Margoulis OV, Fricke E, Liebich I, Land S, Barre-Dirrie A, Reuter I, Chekmenev D, Krull M, Hornischer K, Voss N, Stegmaier P, Lewicki-Potapov B, Saxel H, Kel AE, Wingender E. 2006. TRANSFAC and its module TRANSCompel: Transcriptional gene regulation in eukaryotes. Nucleic Acids Res 34:D108–D110. Mukherjee A, Soyal SM, Fernandez-Valdivia R, Gehin M, Chambon P, Demayo FJ, Lydon JP, O’Malley BW. 2006. Steroid receptor coactivator 2 is critical for progesterone-dependent uterine function and mammary morphogenesis in the mouse. Mol Cell Biol 26:6571– 6583. Mulac-Jericevic B, Lydon JP, DeMayo FJ, Conneely OM. 2003. Defective mammary gland morphogenesis in mice lacking the proges-
genesis DOI 10.1002/dvg
terone receptor B isoform. Proc Natl Acad Sci USA 100:9744– 9749. Mulac-Jericevic B, Mullinax RA, DeMayo FJ, Lydon JP, Conneely OM. 2000. Subgroup of reproductive functions of progesterone mediated by progesterone receptor-B isoform. Science 289:1751–1754. Pan H, Zuo L, Choudhary V, Zhang Z, Leow SH, Chong FT, Huang Y, Ong VW, Mohanty B, Tan SL, Krishnan SP, Bajic VB, Dragon TF. 2004. Association miner: A system for exploring transcription factor associations through text-mining. Nucleic Acids Res 32: W230–W234. Petz LN, Ziegler YS, Loven MA, Nardulli AM. 2002. Estrogen receptor a and activating protein-1 mediate estrogen responsiveness of the progesterone receptor gene in MCF- 7 breast cancer cells. Endocrinology 143:4583–4591. Robker RL, Russell DL, Espey LL, Lydon JP, O’Malley BW, Richards JS. 2000. Progesterone-regulated genes in the ovulation process: ADAMTS-1 and cathepsin L proteases. Proc Natl Acad Sci USA 97:4689–4694. Sasaki M, Dharia A, Oh BR, Tanaka Y, Fujimoto S, Dahiya R. 2001. Progesterone receptor B gene inactivation and CpG hypermethylation in human uterine endometrial cancer. Cancer Res 61:97– 102. Shimshek DR, Kim J, Hubner MR, Spergel DJ, Buchholz F, Casanova E, Stewart AF, Seeburg PH, Sprengel R. 2002. Codon-improved Cre recombinase (iCre) expression in the mouse. Genesis 32:19–26. Soriano P. 1999. Generalized lacZ expression with the ROSA 26 Cre reporter strain. Nat Genet 21:70–71. Soyal SM, Ismail PM, Li J, DeMayo FJ, Lydon JP. 2002. Genomic organization and comparative analysis of the murine progesterone receptor gene. In: Proceedings of the 84th meeting of The Endocrine Society, San Francisco, CA. p255. Soyal SM, Mukherjee A, Lee KY, Li J, Li H, DeMayo FJ, Lydon JP. 2005. Cre-mediated recombination in cell lineages that express the progesterone receptor. Genesis 41:58–66. Sriraman V, Sharma SC, Richards JS. 2003. Transactivation of the progesterone receptor gene in granulosa cells: Evidence that Sp 1/Sp 3 binding sites in the proximal promoter play a key role in luteinizing hormone inducibility. Mol Endocrinol 17:436–449. Swope DL, Harrell JC, Mahato D, Korach KS. 2002. Genomic structure and identification of a truncated variant message of the mouse estrogen receptor a gene. Gene 294:239–247. Tibbetts TA, Mendoza-Meneses M, O’Malley BW, Conneely OM. 1998. Mutual and intercompartmental regulation of estrogen receptor and progesterone receptor expression in the mouse uterus. Biol Reprod 59:1143–1152. Xiong Y, Dowdy SC, Gonzalez Bosquet J, Zhao Y, Eberhardt NL, Podratz KC, Jiang SW. 2005. Epigenetic-mediated upregulation of progesterone receptor B gene in endometrial cancer cell lines. Gynecol Oncol 99:135–141. Yu D, Ellis HM, Lee EC, Jenkins NA, Copeland NG, Court DL. 2000. An efficient recombination system for chromosome engineering in Escherichia coli. Proc Natl Acad Sci USA 97:5978–5983.
