A Cre-reporter transgenic mouse expressing the far-red fluorescent protein Katushka

Abstract

Cre/loxP-dependent expression of fluorescent proteins represents a powerful biological tool for cell lineage, fate-mapping, and genetic analysis. Live tissue imaging has significantly improved with the development of far-red fluorescent proteins, with optimized spectral characteristics for in vivo applications. Here, we report the generation of the first transgenic mouse line expressing the far-red fluorescent protein Katushka, driven by the hybrid CAG promoter upon Cre-mediated recombination. After germ line or tissue-specific Cre-driven reporter activation, Katushka expression is strong and ubiquitous, without toxic effects, allowing fluorescence detection in fresh and fixed samples from all tissues examined. Moreover, fluorescence can be detected by in vivo noninvasive whole-body imaging when Katuhska is expressed exclusively in a specific cell population deep within the animal body such as pancreatic beta cells. Thus, this reporter model enables early, widespread, and sensitive in vivo detection of Cre activity and should provide a versatile tool for a wide spectrum of fluorescence and live-imaging applications. genesis, 2011. © 2011 Wiley-Liss, Inc.

INTRODUCTION

The Cre-loxP recombination system in combination with the technology for targeted modification of the mouse genome has opened the door for a new era of genetic analysis in mammals (Nagy, 2000). In parallel, the development of fluorescent proteins as reporters in living cells has revolutionized in vivo imaging technologies allowing new insights into the processes that govern mammalian development and disease (Nowotschin et al.,2009). Thus, expression of fluorescent proteins has been thoroughly used for promoter activity characterization (Yoshimizu et al.,1999), cell-fate analysis (Lombardi et al.,2009) and lineage-tracing (Shenje et al.,2008). Furthermore, fluorescent proteins have been widely used as reporters of Cre-activity in genetic contexts in which their expression is dependent on Cre-mediated recombination of loxP-flanked (floxed) sequences (De Gasperi et al.,2008; Elghazi et al.,2008; Luche et al.,2007; Mao et al.,2001; Muzumdar et al.,2007; Srinivas et al.,2001; Stoller et al.,2008; Vintersten et al.,2004; Yamamoto et al.,2009). However, the use of most fluorescent proteins for whole-body noninvasive imaging in mammals is constrained due to strong light absorption by hemoglobin at wavelengths below 600 nm, where the excitation and emission spectra of most fluorescent proteins reach a maximum (350–550 nm). Between 600 and 1200 nm, where infrared absorption by water molecules begins, an “optical window” is defined in which tissues are relatively transparent to light (Lin et al.,2009). Red and far-red fluorescent proteins with excitation and emission peaks close to and above 600 nm, respectively, are therefore preferred for in vivo imaging applications.

Recently, a new generation of fluorescent protein markers has emerged with the description of infrared-fluorescent proteins (IFPs) with excitation and emission peaks at 684 and 708 nm, respectively (Shu et al., 2009). These bacteriophytochrome-based fluorescent proteins require incorporation of biliverdin as the chromophore. Although biliverdin is ubiquitous, its endogenous availability may be highly variable, because the enzyme responsible for its production (heme oxygenase-1) is stress-inducible (Ollinger et al.,2007). Moreover, exogenous administration of biliverdin, necessary to increase the infrared fluorescent signal in intact mice, has diverse anti-inflammatory effects (Overhaus et al.,2006). These issues may limit some in vivo applications of IFPs specially when used as promoter activity- or gene-expression- reporters.

The development of Katushka (Shcherbo et al., 2007), an autofluorescent protein characterized by a unique combination of high brightness, far-red emission, and fast rate of chromophore maturation, represents an important contribution to the list of fluorescent markers suitable for whole-body in vivo optical imaging. Excitation and emission spectra of Katushka peak at 588 and 635 nm, respectively. Moreover, Katushka is characterized by fast maturation at 37°C and is one of the brightest autofluorescent proteins with an emission maximum beyond 620 nm. All these properties make of Katushka one of the best far-red fluorescent proteins available for in vivo tissue imaging.

Here, we describe a novel Cre-reporter transgenic mouse line in which Katushka expression is dependent on Cre-mediated recombination. To our knowledge, this is the first transgenic mouse line expressing Katushka, since only transient expression of this fluorescent protein, after DNA electrotransfer into mouse skin and muscle, has been previously reported (Gothelf et al., 2010; Hojman et al., in press).

The transgenic construct (Fig. 1a,b) in this line (Fig. 1a,b) includes the ubiquitous promoter CAG, composed of the cytomegalovirus (CMV) immediate early enhancer and the modified chicken β-actin promoter (Niwa et al., 1991), a loxP-flanked transcription STOP cassette (LSL) (Tsien et al.,1996) followed by the cDNA of Katushka fluorescent protein (KFP), and a rabbit β-globin polyadenylation signal. In this construct, CAG-LSL-KFP, Katushka expression is driven by the CAG promoter upon Cre recombinase-mediated excision of the floxed STOP cassette.

Figure 1. Transgenic construct and founder characterization. (a) Schematic structure of the transgenic construct. The transgene contains the CMV early enhancer/chicken β-actin promoter (CAG), a transcription STOP cassette (STOP) flanked by loxP sites (black triangles), the Katushka cDNA, and the rabbit β-globin polyadenylation signal (pA). (b) Cre-mediated transgene recombination in a head-to-tail multicopy integration event (a two-tandem copy integration is depicted as an example). The position of the relevant restriction sites EcoRI (E), SacI (S), and XbaI (X), the size of the corresponding restriction fragments, and the Katushka cDNA fragment used as a probe for Southern blot (small solid black box) are also indicated. (c) Relative transgene copy-number analysis. Southern blot analysis of tail DNA from founders 3, 5, 10, 13, and 14, double-digested with E and S and hybridized with a probe from Katushka cDNA. The probe recognizes a 2.2 kb internal fragment from the nonrecombined transgenic insert, and the intensity of the signal is proportional to the number of integrated copies. A probe corresponding to the Tie2 locus was used as loading control. (d) Multiple integration site analysis. Southern blot analysis of tail DNA from founders 3, 10, and 13 digested with X. Multiple copy integrations in a head-to-tail orientation are identified by a band of 4.7 kb. Multiple integration sites are recognized by the appearance of more than one additional band (additional bands marked with arrowheads). Founders 3 and 13 have a single integration site. (e) Cre-dependent transgene recombination analysis. Southern blot analysis of tail DNA from progeny of founders 3 and 13 before and after germ-line Cre-mediated recombination. The 2.2 kb and 0.8 kb bands correspond to the floxed and excised transgenes, respectively.

Katushka expression from the transgenic construct was first tested in vitro by cotransfection of murine embryonic stem (ES) cells and H460 lung carcinoma cells with the CAG-LSL-KFP-bearing plasmid, alone or in combination with a Cre expressing vector (CMV-NLSCre). In the absence of Cre, none of the CAG-LSL-KFP transfected clones were Katushka-positive, while ES and H460 cells transfected with both the transgenic reporter and CMV-NLSCre exhibited far-red fluorescence (data not shown).

Having validated the construct in vitro, transgenic mice were generated by pronuclear injection of the transgene in fertilized oocytes. Five founder mice carrying the Katushka reporter transgene, Tg(CAG-LSL-KFP), were generated, three of which transmitted the transgene to the next generation. Relative transgene copy-number and number of integration events were estimated in these three lines by Southern blot analysis of tail genomic DNA (Fig. 1c,d). Two lines, derived from founders 3 and 13, which contained single integration event, were further expanded and characterized. No significant differences were observed between them. Most of the studies were performed with line 3.

To assess the in vivo Cre-dependent control of Katushka expression and its compatibility with normal embryonic development and postnatal life, reporter mice were mated with animals carrying the universal deleter CMV-Cre transgene (Zinyk et al., 1998). The N1 progeny were analyzed for Cre-mediated STOP cassette excision by Southern blot of tail biopsy DNA (Fig. 1b,e) and for Katushka expression by direct fluorescence detection in newborn pups (Fig. 2a). Double transgenic N1[Tg(CMV-Cre);Tg(CAG-LSL-KFP)] pups were born at the expected Mendelian ratio, developed normally and were fertile. They exhibited nearly ubiquitous and intense far-red fluorescence in all the organs examined. N1 animals were further mated with wild-type mice, and the hemizygous offspring carrying the recombined transgenic allele were also born at normal Mendelian ratio (Fig. 2b). These results indicate that there is no toxicity associated with ubiquitous expression of Katushka in mice. Furthermore, multiple generations carrying the germ-line excised and therefore constitutively activated reporter allele have not exhibited any symptoms of Katushka-related toxicity up to 12 months of age, and both males and females are fertile. Katushka fluorescence is robust in all tissues and organs examined from germ-line Cre-activated adult mice, while those of animals with the nonrecombined reporter transgene show a level of tissue autofluorescence indistinguishable from that of control wild-type mice (Fig. 2c,d).

Figure 2. Detection of Katushka fluorescence in reporter mice after Cre-mediated recombination. (a) In vivo whole-body direct fluorescence analysis of N1 progeny (newborns) from crosses of Tg(CAG-LSL-KFP) with Tg(CMV-Cre) mice. Top: bright field image of a double transgenic littermate above a single transgenic pup. Bottom: corresponding direct red fluorescence image. Only the double transgenic pup is visible and exhibits ubiquitous and strong expression of Katushka. (b) Fluorescence image acquired with an IVIS Spectrum (Xenogen Co.) of seven littermates obtained after mating a germ-line Cre-recombined reporter male with a wild-type female. The four pups with fluorescent signal have inherited the recombined reporter allele while the three without signal have not, as confirmed by PCR genotyping (data not shown). (c) Katushka fluorescence in isolated organs of germ-line recombined reporter adult mice (center panel). The two panels at the left show the bright field (BF) and red fluorescence (RF) images of a wild-type mouse as control of tissue autofluorescence, while the two panels at the right show those of a Tg(CAG-LSL-KFP) mouse carrying the intact floxed reporter transgene. (d) Direct fluorescence images from liver cryosections of germ-line Cre-excised and original unexcised Tg(CAG-LSL-KFP) reporter mice.

Expression of Katushka was also investigated in hematopoietic lineages of hemizygous mice with a germ-line Cre-recombined reporter allele. Fluorescence-assisted cell sorter analysis of single-cell suspensions from bone marrow and spleen revealed bright far-red fluorescence in the majority of nucleated cells (Fig. 3a). Importantly, no differences regarding fluorescence emission were detected between wild-type controls and animals that had not undergone Cre-mediated recombination. The heterogeneity in fluorescence intensity, especially evident in bone marrow single-cell suspensions, is probably due to the cellular heterogeneity of the analyzed samples. Further analysis of hematopoietic lineages was conducted through antibody-based fractionation of erythrocyte-depleted single-cell suspensions. Cells of hematopoietic origin (CD45⁺), B lymphocytes (CD19⁺), T lymphocytes (CD4⁺ or CD8⁺), granulocytes, and monocytes (Gr-1⁺) from bone marrow and spleen express Katushka to an extent that makes them clearly distinguishable from wild-type or unrecombined control cells (Fig. 3b). Furthermore, constitutive expression of the reporter does not affect the relative abundance of these hematopoietic cell subpopulations with respect to those of wild-type mice (data not shown). Given that germ-line excised mice are not chimeric with regard to the recombination status of the Katushka reporter allele, the level of red fluorescence in different cell populations is mostly determined by the activity of the CAG promoter in each particular cell type.

Figure 3. Flow cytometry analysis of Katushka expression in hematopoietic lineages from spleen and bone marrow of mice that inherited one copy of the Cre-recombined reporter allele and lack the Cre transgene. (a) Katushka expression in erythrocyte-depleted, unfractioned, hematopoietic cells. Note that Fluorescence-assisted cell sorter-profiles of wild-type (black line) and floxed Tg(CAG-LSL-KFP) mice (red line) are undistinguishable. (b) Analysis of Katushka expression in different hematopoietic cell lineages. The dot plots show the fluorescence intensities of different CD4-, CD8-, CD19-, Gr-1-, and CD45-positive cell populations from spleen and bone marrow.

We next examined whether the Katushka reporter allele was accessible to Cre-mediated excision during early embryogenesis. We crossed our reporter strain with the transgenic line Tg(Sox2-Cre) (Hayashi et al., 2002), in which Cre recombinase expression is driven by the regulatory sequences of the epiblast-specific gene Sox2. It has been previously described that Sox2 mRNA is first detected in some cells of the preimplantation embryo at 2.5 days postcoitum (dpc), at morula stage. In the blastocyst, at 3.5 dpc, Sox2 is expressed specifically within the inner cell mass (ICM), which later differentiates into two lineages, epiblast and hypoblast (Avilion et al.,2003). In double transgenic Tg(CAG-LSL-KFP);Tg(Sox2-Cre) mice, Katushka expression was first seen at 3.5 dpc in early blastocysts where, as expected, it is restricted to the ICM without residual expression in trophectoderm cells (Fig. 4a). Earlier expression at morula stage was not detected, probably due to incomplete or delayed Cre-recombination. Red fluorescence gains intensity as the blastocyst expands, so that, at 4.5 dpc, the ICM is tightly packed at one pole of the developing embryo and exhibits strong far-red fluorescence (Fig. 4b). Further differentiation of the epiblast gives rise to all embryonic tissues, while the majority of extraembryonic components come from the trophectoderm and the hypoblast. Therefore, after implantation and differentiation of embryonic and extraembryonic tissues, Katushka expression is restricted to the developing fetus while the placenta shows no signal over background autofluorescence (Fig. 4c). All these results demonstrate an efficient induction of the reporter by Cre-mediated recombination and robust Katushka expression from a very early stage in mouse embryogenesis.

Figure 4. Epiblast-specific expression of Katushka in double transgenic Tg(CAG-LSL-KFP); Tg(Sox2-Cre) embryos. (a and b) Overlay of single confocal section images of red fluorescence and bright field channels of early blastocysts (3.5 dpc) (a) and late blastocysts (4.5 dpc) (b) obtained from matings of Tg(Sox2-Cre) males and Tg(CAG-LSL-KFP) females. (c) IVIS Spectrum image of developing fetuses (15.5 dpc) from the same type of mating. The embryo at the bottom of the image lacks the Sox2-Cre transgene while the one at the top is double transgenic for Sox2-Cre and the Katushka reporter.

One of the main goals for generating a Cre-reporter transgenic line expressing Katushka was to enable monitoring Cre activity in vivo by direct fluorescence detection using noninvasive whole-body optical imaging techniques, even in minor cell populations within the animal body. This has been accomplished before using bioluminescent reporters such as firefly luciferase (Ishikawa and Herschman, 2010; Liao et al., 2007; Lyons et al.,2003; Safran et al.,2003; Woolfenden et al.,2009). The use of fluorescent reporters has the additional advantage of allowing histological analysis through direct imaging using fluorescence- and confocal-microscopy. In order to test if Katushka signal coming from an internal organ can be monitored noninvasively, we crossed our reporter line with the transgenic strain Tg(RIP-Cre) (Gannon et al.,2000) in which the Cre recombinase is expressed specifically in pancreatic β-cells under the control of the rat insulin-2 promoter. Katushka-specific florescence is readily detected by noninvasive whole-body optical imaging without ectopic expression (Fig. 5a). Pancreatic origin of the fluorescent signal was confirmed after necropsy and pancreas isolation (Fig. 5b), and further specificity of reporter expression to β cells was verified by confocal microscopy imaging of paraformaldehyde-fixed tissue sections (Fig. 5c,d). These data show that tissue-specific inducibility of the reporter is reliable in adult mice and that the spectral characteristics of Katushka fluorescence enable noninvasive whole-body imaging.

Figure 5. In vivo assessment of pancreatic-β-cell-specific, Cre-mediated expression of Katushka in adult mice. (a) IVIS Spectrum whole body fluorescence images of a single transgenic Tg(CAG-LSL-KFP) mouse (left) and a double transgenic Tg(RIP-Cre); Tg(CAG-LSL-KFP) mouse (right). Katushka fluorescence is observed only in the region where the pancreas is positioned. The abdominal region has been previously depilated to avoid signal attenuation by the hair. (b) IVIS fluorescence image of the isolated pancreas from mice shown in panel a. (c) Confocal images of cryosections of the pancreas shown in panel b after 4% paraformaldehyde fixation and OCT embedding. A white-dashed line has been drawn around the pancreatic islets. Blue channel shows nuclei stained with DAPI. Katushka fluorescence (red channel) is detected only in the double transgenic islet. Background autofluorescence is indistinguishable between single and double transgenic mice. (d) Confocal image of a cryosection of the pancreas from the double transgenic mouse shown in panel c (top). Katushka expression is restricted to the β-cells of the pancreatic islets. The zoomed image at the right shows β-cells that strongly express red fluorescence (arrows), while other cells within the pancreatic islet do not express Katushka (arrowheads).

In summary, we have described a Cre-reporter transgenic line that allows early, widespread, and sensitive detection of Cre-mediated recombination. Unlike previous fluorescence-based Cre-reporter models, this line allows noninvasive whole-body detection of recombination events even in tissues located deep within the animal. Additionally, the possibility to detect Katushka-expressing cells not only cytofluorometrically, but also in paraformaldehyde-fixed tissue sections is another advantage of this reporter line. These characteristics make this line a versatile tool for studies requiring noninvasive, efficient, and sensitive detection of Cre activity.

METHODS

Transgenic Construct and Generation of Transgenic Mice

The transgene was constructed by inserting a PCR-amplified fragment containing Katushka cDNA from pTurboFP635-C (Evrogen, Moscow, Russia) into the NotI site of the pVL1 plasmid (a gift from Angel Nebreda, CNIO). The resulting plasmid contains the CAG promoter followed by a transcription/translation STOP sequence flanked by directly repeated loxP sites, the Katushka cDNA, and the rabbit β-globin polyadenylation signal. For transgenic mice generation, a 4.7 kb PvuII-SalI fragment containing the reporter transgene was microinjected into the pronuclei of zygotes obtained from crosses of F1[(C57BL/6J.OlaHsd)(CBA/J)] animals following standard protocols (Nagy, 2003). Potential founders were identified by PCR and Southern blot analyses of tail DNA. PCR primers for transgene genotyping were as follows: Katushka forward primer, 5′-AACGACCACCACTTCAAGTGC-3′; Katushka reverse primer, 5′-TAGCCAGAAGTCAGATGCTCAAGG-3′; Cre forward primer, 5′-CCGGTTATTCAACTTGCACC-3′; Cre reverse primer, 5′-CTGCATTACCGGTCGATGCAAC-3′. All experiments with mice were performed in accordance with protocols revised and approved by the institutional ethics committee of the Spanish National Cancer Research Centre (CNIO).

Southern Blot Analysis of Transgene Copy Number, Integration Events, and Cre-Mediated Recombination

Tail tips were incubated in lysis buffer containing 20 mM Tris–HCl (pH 8.0), 0.5 mg/ml proteinase K, 100 mM NaCl, 0.5% sodium dodecyl sulfate, and 10 mM EDTA at 55°C overnight. After precipitation of proteins and cell debris with saturated NaCl, DNA was precipitated from the supernatant with isopropanol. DNA samples were double digested with EcoRI and SacI for copy-number and recombination efficiency analysis and with XbaI for integration analysis and subjected to electrophoresis on a 0.8% agarose gel. DNA was transferred to a nylon membrane (Amersham Hybond-N+, GE Healthcare, Buckinghamshire, UK). The membrane was hybridized with a 331-bp Katushka probe amplified by PCR using the following primers: forward, 5′-AACGACCACCACTTCAAGTGC-3′; reverse, 5′-GGTGAACTTCCCATCCAACG-3′. The probe was labeled by random priming with Amersham Ready-To-Go DNA labeling beads (GE Healthcare, Buckinghamshire, UK) and dCTP 5′-triphosphate [α-³²P] (Perkin Elmer, Boston, MA). After hybridization and washing, the blot was exposed on a phosphorimager screen and revealed with a Typhoon 9400 scanner (Amersham Biosciences, New Jersey).

Whole-Body and Isolated Organ Fluorescence Detection

Fluorescence from newborn pups and adult organs was analyzed in a Leica MZ16F (Leica, Wetzlar, Germany) fluorescence stereomicroscope equipped with a Texas Red filter set. For epifluorescence pictures, exposure time was optimized based on the level of fluorescence emitted from each organ: testis 300 ms, spleen 660 ms, pancreas 70 ms, lung 660 ms, liver 300 ms, kidney 660 ms, heart 600 ms, and brain 600 ms. Pictures were captured with a DFC 350 FX (Leica, Wetzlar, Germany) camera. Noninvasive whole-body imaging of fluorescence was performed with the IVIS Spectrum system (Xenogen Co., Alameda, CA). Fluorescence signal was acquired with a band-pass excitation filter of 570 nm (30 nm bandwidth) and a band-pass emission filter of 640 nm (20 nm bandwidth), with a photon acquisition time of 1 s, medium binning and F/Stop 1 or 2. Adult animals were sedated by continuous inhalation of 3% isofluorane during image acquisition. The area of focus was depilated using commercial depilatory creams. For organ fluorescence signal analysis, the various organs were harvested and briefly stored at 4°C in phosphate-buffered saline (PBS) with 5% fetal bovine serum before imaging acquisition.

Cytofluorometry

Single-cell suspensions were obtained from bone marrow and spleen using standard procedures. Briefly, mice were sacrificed by CO₂ inhalation and spleen and hind leg long bones dissected. Bone marrow was flushed out with Dulbecco's modified Eagle medium (DMEM, Sigma, St. Louis, MO), and splenocytes were obtained by mechanical dissociation of the organ between two glass slides in DMEM. Erythrocytes were lysed with ammonium chloride and cells resuspended in PBS containing 0.5% bovine serum albumin (Sigma, St. Louis, MO) and 2 mM EDTA (Sigma, St. Louis, MO). Fc receptors were blocked by incubation with a purified anti-mouse CD16/CD32 antibody (BD Pharmingen, San Jose, CA) for 15 min on ice. Single-cell suspensions were stained with the following monoclonal antibodies from BD Pharmingen, San Jose, CA: phycoerythrin (PE) anti-CD4, fluorescein isothiocyanate anti-CD8, allophycocyanin (APC) anti-CD19, PE-cyanine 7 (PE-Cy7) anti-Gr-1, and APC anti-CD45. All stainings were performed on 10⁶ cells in 200 μL for 30 min on ice in the dark. Samples were filtered through a 70-μm filter mesh and processed on a BD ARIA II SORP Cell Sorter (BD Pharmingen, San Jose, CA) equipped with 355, 488, 561, and 640 nm lines. Pulse processing was used to exclude cell aggregates and 4′,6-diamidino-2-phenylindole (DAPI) to discriminate dead cells. At least 50,000 single alive events were collected. Data were analyzed using FlowJo version 9.0.2 (Treestar, Ahsland, OR).

Confocal Microscopy

For analysis of reporter performance in preimplantation embryos, matings were set up between Tg(CAG-LSL-KFP) hemizygous females and Tg(Sox2-Cre) hemizygous males. Mating time was determined by observation of the female vaginal plug (0.5 dpc). Embryos were harvested from pregnant females at 2.5 or 3.5 dpc by flushing oviducts and/or uterine horns with M2 medium (Sigma, St. Louis, MO) and observed immediately after harvesting or incubated in KSOM medium (Specialty Media, Phillipsburg, NJ) at 37°C and 5% CO₂ for 24 h under oil. Pancreas were dissected and fixed by immersion in 25 ml of 4% paraformaldehyde in PBS, pH 7.4 for 2 h at room temperature, and rinsed in three changes of PBS over 30 min and a final rinsing step on a rotary wheel overnight at room temperature. Fixed tissues were embedded in Tissue-Tek OCT (Sakura, Zoeterwoude, NL) on dry ice and stored frozen at −20°C. Ten micrometer-wide sections were collected on Superfrost glass slides (Thermo Scientific, Braunschweig, Germany) and mounted in Vectashield with DAPI (Vector Laboratories, Burlingame, CA). Images were acquired with a Leica TCS-SP5 (AOBS) Confocal Microscope with 63× HCX PL APO 1.30 N.A glycerol immersion objective and 20× HCX PL APO CS 0.7 N.A glycerol immersion objective. The software for capture was LAS AF v2.1.

Acknowledgements

We thank Francisca Mulero and the Molecular Imaging Unit at the CNIO for providing the pTurboFP635-C plasmid and for the support with the use of the IVIS Spectrum, Vanesa Lafarga and Angel Nebreda for the pVL-1 plasmid and Jaime Muñoz and Miriam García for their excellent assistance with mouse colony management.