|Home | About | Journals | Submit | Contact Us | Français|
Understanding the cellular events that underlie epithelial morphogenesis is a key problem in developmental biology. Here we describe a new transgenic mouse line that makes it possible to visualize individual cells specifically in the Wolffian duct and ureteric bud, the epithelial structures that give rise to the collecting system of the kidney. myr-Venus, a membrane–associated form of the fluorescent protein Venus, was expressed in the ureteric bud lineage under the control of the Hoxb7 promoter. In Hoxb7/myr-Venus mice, the outlines of all Wolffian duct and ureteric bud epithelial cells are strongly labeled at all stages of urogenital development, allowing the shapes and arrangements of individual cells to be readily observed by confocal microscopy of freshly excised or cultured kidneys. This strain should be extremely useful for studies of cell behavior during ureteric bud branching morphogenesis in wild type and mutant mouse lines.
During the development of the excretory system, the Wolffian duct gives rise to an evagination, the ureteric bud (UB), which grows and branches extensively within the developing kidney to form the collecting duct system. Defects in either the initial outgrowth, or the subsequent branching and elongation, of the UB can result in severe renal developmental defects including agenesis or hypodysplasia, which are relatively common birth defects in humans (Pohl et al., 2002) (Airik and Kispert, 2007) (Dressler, 2006). The mouse has been used extensively as a model system to study the genetic control of renal branching morphogenesis, and many genes that participate in this process have been identified using knockout mice (Yu et al., 2004) (Dressler, 2006). However, relatively little is known about the cellular behaviors that are controlled by these genes, and that ultimately carry out the “work” of branching morphogenesis (Davies, 2002) (Costantini, 2006). Among the various processes that may participate in epithelial growth and branching are cell proliferation, cell movements and changes in cell shape (Ettensohn, 1985) (Meyer et al., 2004). As an example of the latter process, the conversion of cells from cuboidal to wedge-shaped can cause the epithelium to fold (Ettensohn, 1985). In this paper, we describe a new transgenic mouse line that was designed to aid in the analysis of cell shape and arrangement during UB morphogenesis.
We previously described a transgenic mouse strain, Hoxb7/eGFP, which expressed enhanced GFP under the control of the Hoxb7 promoter (Kress et al., 1990), and thus specifically in the Wolffian duct, the ureteric bud, and their derivatives during excretory system development, as well as in the collecting ducts of adult kidneys (Srinivas et al., 1999). This strain has been very useful for studies of renal development, both in vivo and in organ cultures, as it allows ureteric bud morphogenesis to be visualized by fluorescence microscopy in live tissues (Watanabe and Costantini, 2004) (Shakya et al., 2005) (Davies, 2005). However, one limitation of this line is that eGFP is a cytoplasmic protein which labels the ureteric bud cells rather uniformly (Srinivas et al., 1999). Thus, while Hoxb7/GFP mice are useful for studying morphogenesis at the tissue level, they are less useful for studies of individual cell behaviors or cell shapes. To overcome this limitation, we have generated a new transgenic strain that uses the same Hoxb7 promoter to drive expression of myr-Venus (Rhee et al., 2006), a membrane-targeted form of Venus, which is a variant of EYFP (enhanced Yellow Fluorescent Protein). While mice expressing this protein in a widespread pattern have been previously described (Rhee et al., 2006), the use of tissue-specific regulatory elements, such as the Hoxb7 promoter, makes it possible to focus on a specific structure of interest within a complex organ such as the kidney.
The myr-Venus DNA fragment, originally in pCX vector (Rhee et al., 2006) was recloned into the EcoRI site of pBluescript SK vector. The 756bp myr-Venus fragment was then excised with BamHI and EcoRV, and ligated upstream of a human ß-globin 3′ sequence (in pKS-hßglobin3′ vector) at the BamHI and XbaI sites (the XbaI site was filled in with Klenow DNA polymerase, to be compatible with the EcoRV blunt end). The ligated myr-Venus and ß-globin DNA was cut with NotI, Klenow-filled, then digested with EcoRV to release a 2.5kb fragment containing myr-Venus-ß-globin 3′, which was ligated downstream of the Hoxb7 promoter at the SmaI site of a Hoxb7 promoter construct (in pBluescript KS vector). The final plasmid is called pHoxb7/myr-Venus.
The insert of plasmid pHoxb7/myr-Venus was excised with NotI and XhoI, gel-purified, and microinjected into B6CBAF1/J x B6CBAF1/J zygotes by standard procedures (Hogan et al., 1994). Transgenic lines were identified by PCR analysis of tail tip DNA, using a pair of Venus primers (5′-GCC CGC TAC CCC GAC CAC ATG A-3′ and 5′-CGG CGG CGG TCA CGA ACT CC-3′) that yielded a 469-bp product. The PCR conditions were: 2 min at 94 °C, then 32 cycles of 30 s at 94°C, 30 s at 55°C and 45s at 72°C. The founder transgenic mice were confirmed by Southern blot analysis using restriction enzyme SpeI and HindIII and a 469bp probe for Venus. Two independent transgenic mouse lines (#10 and #17) were evaluated, and both showed indistinguishable patterns of expression of myr-Venus, which were strong and heritable, in the Wolffian duct/ureteric bud lineage. The studies in this paper were carried out using line #17, which was propagated in the hemizygous state.
Embryos were dissected at various stages of development and transgenic embryos were identified by examination on a stereomicroscope equipped with epifluorescence (Leica MZ16FA). Renal organ cultures were carried out as previously described (Srinivas et al., 1999) (Watanabe and Costantini, 2004).
Kidney explant cultures were fixed in 4% PFA for 3 hours at 4 °C. Anti-phosphohistone H3 (Ser10) (6G3) (Cell Signaling, 9706S) staining of organ cultures was performed on Transwell filters, using a protocol described previously (Basson et al., 2006). Cy3-conjugated secondary antibodies were from Jackson ImmunoResearch.
Both live kidneys and fixed kidneys were imaged by confocal laser scanning microscopy. For live imaging experiments, E14.5 Hoxb7/myr-Venus kidneys were placed in cover slip-bottomed dishes (MatTek, Ashland, MA) in PBS. Confocal images were acquired as z-stacks of x,y images taken at 2 μm z-intervals, using a Zeiss LSM510 META on a Zeiss Axiovert 200M, with a Plan-Neofluar 40×/NA1.3 and a Plan-Apochromat 20×/NA0.75. Fluorophores were excited with a 488-nm argon laser (EGFP/Venus). E15.5 kidneys fixed overnight in 4% paraformaldehyde were imaged with a Bio-Rad laser scanning confocal microscope equipped with an OlympusUM PlanFl water immersion lens 10× /NA 0.3 and Olympus U PlanApo/IR water immersion lens 60× /NA1.2, at 0.5-5.3μm z-intervals. Volume-rendering of image stacks and rotations were performed using Volocity (Improvision, www.improvision.com) and Voxx (http://www.nephrology.iupui.edu/imaging/voxx/index.htm). Image segmentation and isosurface models were generated by Amira 3.1 by TGS Template Graphics Software, Inc. (www.amiravis.com).
Kidneys from Hoxb7/myr-Venus transgenic mice were examined at various stages of development to analyze the subcellular expression of the myr-Venus fluorescent protein, and to examine the organization and shapes of ureteric bud epithelial cells during kidney development. Figure 1a shows a low magnification image of the urogenital system of a Hoxb7/myr-Venus mouse at stage P0 (newborn), revealing the expression of myr-Venus in the ureteric bud tips throughout the periphery of the kidney, as well as in the ureters, but not in the bladder, reproductive organs, or adrenals. Figure 1 b and c show 3-D reconstructions of the ureteric tree of a Hoxb7/myr-Venus E15.5 kidney, generated from stacks of low-magnification (10×) confocal images, and Figure 1 d-g show z-projections at different levels of the kidney, revealing that myr-Venus expression is limited to the ureteric bud and its derivatives, the collecting ducts. Thus, the tissue-specificity of expression is identical to that previously observed for other Hoxb7-driven transgenes, and it can be used similarly to the Hoxb7/GFP mice to examine the development and overall organization of the collecting duct system. The faint honeycomb-like staining surrounding the UB tips visible in some images (e.g., Figure 1d) was the results of brief staining with Alexa488-conjugated collagen (type I) antibody (Sigma cat # C2456, Molecular Probes Zenon Conjugation kit, cat # Z-25002), which stains some mesenchymal/stromal cells.
Figure 2 shows somewhat higher-magnification optical sections of Hoxb7/myr-Venus kidneys explanted at E12.5 (a, b) or E14.5 (c, d) and cultured for 72 hours (a, b) or overnight (c, d). The kidney in c, d was fixed and stained with phosphohistone H3 (pH3) antibody to detect mitotic cells. Fluorescence of myr-Venus is preserved well after fixation in 4% PFA, and the overlap with anti-pH3 staining indicates the location of mitotic cells within the UB epithelium. The myr-Venus protein clearly demarcates the ureteric bud epithelium from the surrounding structures (e.g., metanephric mesenchyme, nephron epithelia and renal stroma), which derive from different cell lineages and do not express Hoxb7-transgenes (Kress et al., 1990) (Srinivas et al., 1999). Within the UB cells, myr-Venus fluorescence is present only at the cell membrane, as expected. It shows the organization of cells into a simple cuboidal epithelium in the tips (arrows) as well as the stalks (arrowheads), and also reveals the cross-sectional shapes of the individual cells.
Figure 3 a-o shows two series of image stacks through the outer nephrogenic zone of E14.5 kidneys that had developed in vivo. At these magnifications, the shapes of individual cells are clearly visible, as is the generally cuboidal nature of the UB epithelium, with a distinct lumen (Meyer et al., 2004). For comparison, Figure 3p shows an optical section thorough a Hoxb7/eGFP transgenic kidney, which shows two collecting ducts in cross section, but fails to reveal the shapes of individual cells. Thus, the Hoxb7/myr-Venus transgenic line represents a considerable advance for studies in which individual cell morphology and cell arrangement is important.
Figure 4 a-f shows several 3-dimensional reconstructions of the surface of an E15.5 kidney, which were rendered from multiple confocal images such as those shown in Figure 3. These reconstructions reveal several interesting features of the organization of ureteric bud tips, such as the highly regular arrangements and similar sizes of most individual tip cells (Figure 4b), the single-layered nature of the tip epithelium (Figure 4f), and the apparent “indentations” where the ureteric buds join the “connecting tubules” (asterisks in Figure 4d, e), a segment of the nephron that does not express Hoxb7 transgenes. For comparison, Figure 4g shows a volume rendered image of a Hoxb7/eGFP kidney, where the overall shape of the UB tips is visible, but the individual cells are not.
While the images in Figure 4 show the 2-D cell shapes at the surface of the UB epithelium, confocal image stacks of Hoxb7/myr-Venus kidneys can be further processed by segmentation to reveal the 3-dimensional shapes and the spatial arrangements of individual cells, as shown in Figure 5. This reveals that the cells closer to the UB tip (6 and 8) are somewhat larger than those in the cleft (1-4).
These images demonstrate that the Hoxb7/myr-Venus mice are a useful tool to examine epithelial cell organization and cell shape in the developing ureteric bud, in live tissue, and without the need for staining with antibodies or lectins. The ability to image live tissue, together with the established methods for renal organ cultures (Saxen, 1987) (Srinivas et al., 1999) (Watanabe and Costantini, 2004), should make it possible to examine, in real time, the changes in cell organization and shape that occur during branching morphogenesis. While methods are currently available for time-lapse epifluorescence imaging of kidneys growing in organ culture, and we have used them to examine the patterns of branching morphogenesis in Hoxb7/eGFP kidneys (Srinivas et al., 1999) (Watanabe and Costantini, 2004), the resolution in this culture/microscopy system is not sufficient to follow individual cells in developing Hoxb7/myr-Venus kidneys (data not shown). Therefore, it will be necessary to adapt the renal organ culture system to the confocal microscope to take full advantage of potential of this new transgenic strain.
We thank Tingting Tam and Jacob Kreugel for experimental assistance, and Zaiqi Wu for DNA microinjections. This work was supported by grants DK055388 and DK075578 from the NIDDK (to F.C.), HD052115 from the NICHD (to A-K. H.) and a fellowship from the American Heart Association (to X.C). Confocal laser scanning microscopy and volume rendering with Volocity and Amira was performed at the MSSM-Microscopy Shared Resource Facility, and was supported by NIH-NCI shared resources grant 5R24 CA095823-04, NSF Major Research Instrumentation grant DBI-9724504 and NIH shared instrumentation grant 1 S10 RR0 9145-01.