The Tg(ccnb1:EGFP) transgenic zebrafish line labels proliferating cells during retinal development and regeneration

Zebrafish maintenance and light lesion protocol

Adult zebrafish (Danio rerio) were raised under normal facility lights (300 lux; 14 h light:10 h dark) in the Center for Zebrafish Research at the University of Notre Dame according to established protocols. Adult albino zebrafish were dark-treated for 14 days and then subjected to constant intense light (18,000 lux) using four halogen lamps [19]. Retinas were examined at 0 h, 48 h, and 72 h of light exposure. Control animals (0 h intense light exposure) of the same age and genotype were raised under normal facility light conditions. The adult transgenic fish used for immunohistochemical analyses were seven to nine months post-fertilization. Prior to enucleation, fish were deeply anesthetized in 0.2% 2-phenoxyethanol and euthanized by anesthetic overdose. All experimental protocols were approved by the animal use committee at the University of Notre Dame and were in compliance with the US Public Health Service and ARVO statement for the use of animals in vision research.

Transgenic line generation

A 1.5 kb region of the zebrafish cyclin B1 promoter was PCR-amplified with primers containing an XhoI restriction site on the forward primer and BamHI restriction sequence on the reverse primer (Forward: 5′-CTC GAG CCT TAA CTG AAG ACC GCA CCT GC-3′; Reverse: 5′-GGA TCC CTG CTT TCT TAG TTT AGA GTA GG-3′). The product was cloned upstream of the EGFP reporter gene in a plasmid containing the zebrafish Tol2 transposable element (pT2KXIG) [50]. The plasmid was purified using the Qiagen Maxi Prep Kit (Qiagen, Valencia, CA) and extracted by phenol/chloroform. This construct along with in vitro transcribed Tol2 transposase mRNA was injected into wild-type (AB strain) embryos [2] that were between one and four cells. The EGFP-positive F0 fish were outcrossed to AB fish. The F1 generation carriers were identified by EGFP expression and used to generate stable and independent cyclin B1:EGFP transgenic lines. Two stable lines were examined and found to generate indistinguishable EGFP expression patterns in the developing and adult retina. One of these lines, Tg(ccnb1:EGFP)^nt18, yielded ~50% EGFP-positive progeny when mated to AB fish, which suggests it has a single Tol2 insertion. This Tg(ccnb1:EGFP)^nt18 line was used in all of the data presented.

Immunohistochemistry

Embryos and adult eyes were fixed in 9:1 ethanolic formaldehyde overnight at 4 °C. Embryos were cryoprotected in 5% sucrose/1X PBS overnight at 4 °C then in 30% sucrose/1X PBS overnight at 4 °C and finally in 30% sucrose/1X PBS:Tissue Freezing Media (1:1; Triangle Biomedical Sciences, Durham, NC) for 4 h at room temperature. Tissue was then embedded and frozen in 100% Tissue Freezing Medium and sectioned at 18 μm. Slides were dried at 50 °C for 2 h. Tissue sections were rehydrated with 1X PBS, blocked with 1X PBS/2.5% normal goat serum/0.3% Triton X-100/1% DMSO, and incubated overnight at 4 °C with a 1:1500 dilution of rabbit anti-EGFP polyclonal antiserum (Abcam, Cambridge, MA), a 1:1000 dilution of mouse anti-PCNA polyclonal antiserum (Sigma, St. Louis, MO), a 1:1000 dilution of mouse anti-glutamine synthetase monoclonal antiserum (Chemicon, Temecula, CA), a 1:1500 dilution of mouse anti-EGFP monoclonal antiserum (Abcam), or a 1:250 dilution of rabbit anti-Pax6 polyclonal antiserum (Covance, Berkely, CA). Sections were washed three times in 1X PBS/0.1% Tween-20. Goat anti-rabbit Alexa Fluor 488 or 594 and goat anti-mouse Alexa Fluor 488- or 594-conjugated secondary antibodies (Molecular Probes, Carlsbad, CA) were diluted 1:500 in the blocking buffer, and sections were incubated with the secondary antibody for 1 h at room temperature. Sections were washed three times in 1X PBS/0.1% Tween-20, and then mounted using Vectashield (Vector Laboratories, Burlingham, CA). The Bio-Rad 1024 confocal microscope (Bio-Rad, Hercules, CA) was used to take 10 micron Z-stacked images.

For Pax6 labeling, retinal sections were rehydrated in 1X PBS and then incubated in preheated 1X AntigenPlus buffer (pH 10; Novagen, Madison, WI) for 25 min at 95 °C. Slides were cooled for 45 min and immunolabeled as described above.

In situ hybridization

A 550-bp segment of the zebrafish cyclin B1 open reading frame was PCR-amplified (Forward: 5′-CGA GTC ACA GCA ATA AAC CAC GAG-3′ and Reverse: 5′-TTC CCA GTA ACT TCC TTT CCT GC-3′) and ligated into a pCR 4-TOPO vector (Invitrogen, Carlsbad, CA). The construct was linearized then in vitro transcribed with either a T3 or T7 RNA polymerase (Ambion, Austin, TX) and labeled with digoxigenin UTP (Ambion). Whole-mount in situ hybridization with sense or antisense digoxigenin-labeled RNA probe was performed as previously described [51] with minor modifications. Embryos were fixed in 4% paraformaldehyde/5% sucrose overnight at 4 °C. Following dehydration in methanol and a 20 min digestion with Proteinase K (10 μg/ml in 1X PBS/0.1% Tween-20), embryos were fixed in 4% paraformaldehyde/1X PBS and then incubated for 2 h at 55 °C in hybridization buffer (50% formamide, 5X SSC, 500 μg/ml yeast tRNA, 0.1% Tween-20, 1% CHAPS, and 100 μg/ml heparin). Cyclin B1 digoxigenin-labeled sense and antisense RNA probes were incubated with embryos for three days at 55–60 °C. Embryos were washed, incubated for 2 h at room temperature in blocking solution, and then incubated overnight at 4 °C in blocking solution containing alkaline phosphatase-conjugated anti-digoxigenin Fab fragments (Roche, Indianapolis, IN) that had been pre-absorbed with embryos from the same stage. Following washes in staining buffer (100 mM Tris [pH 9.5], 50 mM MgCl₂, 100 mM NaCl, 0.1% Tween-20, 1 mM levamisole), embryos were incubated in BM purple (Roche) substrate for 1.5 h at 37 °C for whole-mount imaging. For embryonic retinal sections, embryos were incubated in BM purple substrate for two days at 4 °C then fixed in 4% paraformaldehyde/1X PBS overnight at 4 °C. Embryos were embedded in JB-4 plastic resin (Polysciences, Warrington, PA) for two days at room temperature. Two and a half μm sections were cut and mounted on slides, and pictures were taken on a Nikon Microphot-FXA upright microscope with a RTKE Spot camera (Diagnostic Instruments Inc., Sterling Heights, MI).

Establishing a cyclin B1:EGFP transgenic zebrafish line

A 1.5 kb fragment of the cyclin B1 promoter that is upstream of the translation initiation codon was PCR-amplified and cloned upstream of the EGFP reporter gene within the zebrafish Tol2 transposable element pT2KXIG plasmid (Figure 1). This construct was co-injected with in vitro transcribed Tol2 transposase mRNA into one to four-cell stage AB strain embryos. The EGFP-positive F0 fish were outcrossed to AB fish. The resulting EGFP-positive F1 carriers were identified and used to generate two independent Tg(ccnb1:EGFP)^nt18 transgenic lines. Both transgenic lines had the same EGFP expression pattern in the retina during development, as an adult, and during retinal regeneration.

At 48 h post-fertilization (hpf), the Tg(ccnb1:EGFP)^nt18 zebrafish head and eye exhibited strong EGFP expression, and the midsection and posterior of the fish expressed EGFP at a reduced intensity compared to the anterior of the fish (Figure 2A). This EGFP expression pattern was confirmed by whole-mount in situ hybridization with an in vitro transcribed cyclin B1 antisense mRNA probe. At 48 hpf, the cyclin B1 mRNA was strongly expressed throughout the head including the retina (Figure 2B). We also detected cyclin B1 mRNA throughout the body and tail of the fish but at a reduced level relative to the anterior portion of the fish (Figure 2B). Thus, the EGFP expression throughout the 48 hpf Tg(ccnb1:EGFP)^nt18 embryo appeared to mimic the expression of the endogenous cyclin B1 mRNA.

Enhanced green fluorescent protein expression during zebrafish retinal development

At 28 hpf, the entire zebrafish retina is composed of proliferating retinal neuroepithelial cells [52]. At this time, a group of cells in the ventronasal region of the retina exit the cell cycle and differentiate into ganglion cells [52]. Neurogenesis continues in a wave from the central retina to the periphery [28,53]. After development of the three retinal nuclear layers, proliferation continues in the circumferential germinal zone (CGZ; circumferential marginal zone [CMZ] in the adult) throughout the remaining life of the fish [23]. Tg(ccnb1:EGFP)^nt18 zebrafish were collected at 36, 48, 72, and 120 hpf to visualize cell proliferation during retinal development. At 36 hpf, EGFP expression was observed throughout the retina with stronger expression near the periphery of the sheet of undifferentiated retina (Figure 3A). Proliferating cell nuclear antigen (PCNA) is a well documented marker for cell proliferation in the zebrafish retina [19,24,44]. At 36 hpf, PCNA was expressed throughout the undifferentiated retina with a higher degree of expression in the periphery (Figure 3E), which nicely overlaid the EGFP expression in the Tg(ccnb1:EGFP)^nt18 embryonic retinas (Figure 3I). At 48 hpf, EGFP and PCNA expression decreased in the central retinal cells as these cells are the first to exit the cell cycle [28,52], but expression persisted in the CGZ (Figure 3B,F,J). These data confirmed that the cyclin B1:EGFP transgene was expressed in proliferating cells in the early stages of zebrafish retinal development.

EGFP was also expressed in the lens of Tg(ccnb1:EGFP)^nt18 embryos at 36, 48, and 72 hpf, as well as at 5 days post-fertilization (Figure 3A-D). While the lens contains a population of proliferating lens epithelial cells, the EGFP expression is found throughout the lens. This unexpected broad expression pattern, which was not observed with PCNA expression, is either due to unusually stable EGFP expression or ectopic expression in all lens cell types. We found that EGFP was also misexpressed in the lens of three other transgenic lines that our laboratory either generated or examined, the rod photoreceptor cell rho promoter Tg(rho:EGFP) transgenic line [34], the progenitor cell olig2 promoter Tg(olig2:EGFP)^vu12 transgenic line [45], and the Müller glial gfap promoter in Tg(gfap:EGFP)^nt11 [19,36]. The reason for the misexpression of several different transgenes in the lens remains unknown.

To further confirm that the transgene expression pattern recapitulated the endogenous cyclin B1 expression pattern, we performed whole-mount in situ hybridizations with an in vitro transcribed cyclin B1 mRNA probe and sectioned the labeled retinas. At 48 hpf, the cyclin B1 mRNA transcripts were primarily detected near the CGZ with decreasing levels visualized in the central portion of the retina (Figure 2E). This pattern was similar to the EGFP and PCNA expression patterns in the Tg(ccnb1:EGFP)^nt18 embryos at 48 hpf (Figure 2D,C, respectively). As retinal development proceeded to 72 hpf, EGFP and PCNA expressions became further restricted to the retinal margin (Figure 3C,G,K). By 5 dpf, EGFP expression was confined to the CMZ, which corresponded to the PCNA expression pattern (Figure 3D,H,L) where retinal cell proliferation persists throughout the lifetime of the zebrafish [23,35,54].

Enhanced green fluorescent protein expression in the undamaged adult Tg(ccnb1:EGFP)^nt18 zebrafish retina

The adult zebrafish retina contains three distinct sources of mitotically active cells, the CMZ, ONL rod precursor cells, and INL Müller glia [22,23,35,54]. We performed immunohistochemistry on the undamaged adult Tg(ccnb1:EGFP)^nt18 retinal sections to determine if EGFP was expressed in these three progenitor cell populations. Similar to the 5 dpf retina, EGFP was highly expressed in the adult retinal CMZ (Figure 4C), which co-labeled with PCNA expression (Figure 4A,B). EGFP-positive cells near the periphery of the CMZ coexpressed Pax6 (Figure 4D-F), an early marker for neuronal progenitor cells in the CMZ [23], while the EGFP-positive cells toward the central CMZ had minimal Pax6 expression (Figure 4D-F, arrow). The further centrally located Pax6-positive EGFP-negative signal corresponds to the newly differentiated amacrine and ganglion cells (Figure 4D-F, arrowheads). Rod precursor cells in the ONL are specialized progenitors that are committed to the rod photoreceptor cell lineage during the persistent adult retinal neurogenesis [18,35]. EGFP and PCNA were coexpressed in these proliferating rod precursor cells (Figure 4G-I, identified by arrow). These rod precursor cells arise from a subset of Müller glia in the undamaged adult zebrafish retina [55], where a limited number of Müller glia reenter the cell cycle and produce neuronal progenitor cells that migrate to the ONL and become rod precursor cells during persistent neurogenesis [22,36,55]. In the Tg(ccnb1:EGFP)^nt18 zebrafish retina, we identified a few glutamine synthetase (GS)-positive Müller glia that coexpressed PCNA and EGFP (Figure 4J-O). In the central retina, there was unexpected EGFP expression in PCNA-negative short single cone cells (Figure 4G,I). The EGFP expression in the short single cones may be either a failure to clone all the silencer elements associated with the ccnb1 promoter or the actual representation of cyclin B1 expression in the differentiated short single cones, similar to the expression of cyclins and cyclin-dependent kinases in other differentiated neurons [56]. Regardless, there was an excellent correspondence between EGFP expression and the three proliferating progenitor cell populations in the undamaged adult Tg(ccnb1:EGFP)^nt18 retina.

Enhanced green fluorescent protein expression in the regenerating Tg(ccnb1:EGFP)^nt18 zebrafish retina

Damage to any cell type of the adult zebrafish retina will initiate a robust regenerative response [19,24-26,54,57]. Previously, our laboratory characterized a light damage model that destroys both rod and cone photoreceptors [19,21]. Photoreceptor cell death initiates a regenerative response with the Müller glia exhibiting an increased amount of proliferation and a generation of large clusters of neuronal progenitor cells, which migrate to the ONL and replenish the lost rod and cone photoreceptors [19,21]. To determine if the proliferating Müller glia in the regenerating retina expressed the cyclinB1:EGFP transgene, Tg(ccnb1:EGFP)^nt18 zebrafish were exposed to constant light. After 48 h of constant light, EGFP was expressed in all the PCNA-expressing Müller glia (Figure 5A-C). Glutamine synthetase, a known marker for Müller glia [19,21], confirmed that the EGFP-positive cells in the light-damaged retina were Müller glial cells (Figure 5D-F). After 72 h of light damage, multiple progenitors were visualized migrating along the processes of the EGFP-positive Müller glia (Figure 5G-L). Therefore, EGFP expression in the Tg(ccnb1:EGFP)^nt18 zebrafish line detected the proliferating Müller glia and neuronal progenitor cells during zebrafish retinal regeneration.

Comparison of enhanced green fluorescent protein expression in the Tg(1016α1tubulin:EGFP) and Tg(ccnb1:EGFP)^nt18 transgenic fish lines

EGFP expression in the Tg(1016α1tubulin:EGFP) zebrafish line, which contains a fragment of the α1tubulin promoter, co-labeled with proliferating Müller glia and neuronal progenitor cells during zebrafish retinal regeneration [22], which is similar to the ccnb1:EGFP transgene expression. We compared the EGFP expression profiles in the Tg(ccnb1:EGFP)^nt18 transgenic line relative to the Tg(1016α1tubulin:EGFP) line during retinal development. At 54 hpf, EGFP in the Tg(1016α1tubulin:EGFP) zebrafish line was expressed throughout the retina while PCNA expression was mostly restricted to the retinal margin (Figure 6A-C). At 72 hpf, EGFP persisted in the central retina and PCNA was further restricted to the margin (Figure 6D-F). This differed from the early time points analyzed in the Tg(ccnb1:EGFP)^nt18 transgenic line where EGFP decreased in the central retina relative to the margin (Figure 3B,F,J,C,G,K compared to Figure 6A-F). The EGFP expression in the Tg(1016α1tubulin:EGFP) transgenic line did not decrease in the central retina and become more restricted to the margin until 4 dpf (Figure 6G-I).

To further demonstrate the differences between the EGFP expression patterns in the Tg(ccnb1:EGFP)^nt18 and Tg(1016α1tubulin:EGFP) lines, we examined EGFP and PCNA expression in the larval brain and retinal sections at 72 hpf. PCNA-expressing cells in the retinal margins, mandibular tissue, and dorsal epithelium and tectum did not co-label with EGFP in the Tg(1016α1tubulin:EGFP) transgenic line (Figure 7A-C). However, PCNA expression in similar regions of the Tg(ccnb1:EGFP)^nt18 transgenic zebrafish line co-labeled with EGFP (Figure 7D-F).