SEARCH

SEARCH BY CITATION

Keywords:

  • Sonic Hedgehog;
  • Gli;
  • heat shock;
  • transgenic;
  • zebrafish

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Background: Hedgehog (Hh) signaling is required for embryogenesis and continues to play key roles postembryonically in many tissues, influencing growth, stem cell proliferation, and tumorigenesis. Systems for conditional regulation of Hh signaling facilitate the study of these postembryonic Hh functions. Results: We used the hsp70l promoter to generated three heat-shock–inducible transgenic lines that activate Hh signaling and one line that represses Hh signaling. Heat-shock activation of these transgenes appropriately recapitulates early embryonic loss or gain of Hh function phenotypes. Hh signaling remains activated 24 hr after heat shock in the Tg(hsp70l:shha-EGFP) and Tg(hsp70l:dnPKA-BGFP) lines, while a single heat shock of the Tg(hsp70l:gli1-EGFP) or Tg(hsp70l:gli2aDR-EGFP) lines results in a 6- to 12-hr pulse of Hh signal activation or inactivation, respectively. Using both in situ hybridization and quantitative polymerase chain reaction, we show that these lines can be used to manipulate Hh signaling through larval and juvenile stages. A ptch2 promoter element was used to generate new reporter lines that allow clear visualization of Hh responding cells throughout the life cycle, including graded Hh responses in the embryonic central nervous system. Conclusions: These zebrafish transgenic lines provide important new experimental tools to study the embryonic and postembryonic roles of Hh signaling. Developmental Dynamics 242:529–539, 2013. © 2013 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Hedgehog (Hh)/Gli signaling is required for tissue patterning, cell differentiation, and growth during embryogenesis, with Hh ligands functioning as morphogens, mitogens, or cell survival factors to help induce cell fates and pattern the embryo (Hui and Angers, 2011; Ingham et al., 2011) These functions are maintained throughout the life cycle in select tissues and include the regulation of adult neurogenesis (Jacob and Briscoe, 2003; Tannahill et al., 2005). Mutations in Hh signaling underlie a range of human birth defects, including holoprosencephaly and polydactyly (Hui and Angers, 2011). Misregulation of Hh signaling at postembryonic stages and in the adult is associated with a large number of human cancers including medulloblastoma, glioma, basal cell carcinoma (BCC) (Stecca and Ruiz, 2010). Inappropriate activation of Hh signaling is involved in multiple tumorigenic processes including proliferation, metastasis, and relapse (Stecca and Ruiz, 2010). Hh signaling has now been implicated in neural regeneration following spinal cord injury in fish (Reimer et al., 2009; Kuscha et al., 2012), further underscoring the importance of understanding the mechanisms underlying these multiple postembryonic roles for Hh signaling.

A variety of methods have been developed that provide temporal and/or spatial control of gene expression. These experimental tools include the GAL4/UAS transcription regulatory system from yeast (Lewandoski, 2001; Halpern et al., 2008), site specific recombination-based systems for transgene insertion, deletion, or activation (e.g., Feil, 2007; Bischof and Basler, 2008; Hans et al., 2009) and tetracycline-inducible transcriptional activation systems (e.g., Huang et al., 2005). Heat-shock–inducible promoters that allow transgene activation following acute heat shock have been well characterized. The hsp70l gene encodes a chaperonin that functions to prevent the accumulation of misfolded proteins, and the hsp70l promoter has been used to manipulate gene expression throughout the zebrafish life cycle (Shin et al., 2007; Shoji and Sato-Maeda, 2008; Kaslin et al., 2009).

We have taken advantage of the hsp70l promoter to generate four transgenic zebrafish lines that allow the conditional activation or inactivation of Hh signaling at critical steps in the Hh signal transduction pathway. Hh signaling can be activated extracellularly by induction of a transgene encoding the Sonic Hh (Shh) ligand, intracellularly by induction of a transgene encoding a dominant-negative (dn) form of the inhibitory kinase protein kinase A (PKA) (dnPKA) (Hammerschmidt et al., 1996; Ungar and Moon, 1996), or directly at the transcriptional level by activation of a transgene encoding the Gli1 transcription factor (Karlstrom et al., 2003). Conversely, Hh signaling can be efficiently blocked at the transcriptional level by induction of a transgene encoding a truncated Gli2a transcription factor that acts as a potent dominant transcriptional repressor (Gli2DR) (Karlstrom et al., 2003). All four transgenic constructs also encode the green fluorescent protein (GFP) so that transgene activation can be easily visualized. Here we demonstrate that global heat shock can be used to effectively manipulate Hh/Gli signaling in transgenic zebrafish from embryonic to juvenile stages. We also describe two new lines that sensitively report graded Hh signaling in responding cells. These lines provide important new experimental tools for the study of the embryonic and postembryonic roles of Hh/Gli signaling, including the regulation of stem cell proliferation and differentiation, as well as tumorigenesis.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Heat-Shock–Regulated Hh/Gli Transgenic Lines

We have generated four transgenic constructs to allow the conditional manipulation of Hh/Gli signaling at different levels of the Shh signaling cascade (Fig. 1A). All transgenic constructs contain the zebrafish hsp70l promoter (Halloran et al., 2000) as well as sequences needed for Tol2-mediated transgenesis (Kawakami and Shima, 1999). For extracellular activation of the pathway, we generated an hsp70l:shha-EGFP transgene construct. We fused EGFP to the C-terminus of the full-length shha coding region. Shh proteins are post-translationally cleaved and modified to create a secreted N-terminal fragment (Shh-N) and a C-terminal fragment that remains in the cell (reviewed in Ingham et al., 2011). This fusion construct is thus designed to allow proper processing and secretion of an unlabeled Shh-N fragment, with the GFP containing C-terminal fragment reporting transgene induction in Shh producing cells. Similar heat-shock–mediated activation of Shh signaling was recently reported using a hsp:GAL4/UAS-Shh system (Balczerski et al., 2012). To cell-autonomously activate the Shh signaling pathway, we generated two additional lines. One uses the dominant-negative form of protein kinase A (dnPKA), which prevents the PKA-dependent processing of Gli transcription factors into transcriptional repressor forms (Epstein et al., 1996). dnPKA is an effective activator of Hh signaling, with overexpression in zebrafish embryos reproducing Shh over-expression defects (Concordet et al., 1996; Barresi et al., 2000). dnPKA could potentially affect other signaling systems, including the Wnt signaling pathway (Hino et al., 2005), although one study showed no effect of dnPKA on Wnt signaling in vitro (Liu and Habener, 2008). The second encodes full-length Gli1, which functions as a transcriptional activator and is similar to the line reported by Huang and Schier (2009). Finally, to inactivate Hh signaling at the transcriptional level, we took advantage of the dominant repressor form of Gli2a encoded by the zebrafish you-too (yot) mutants (Karlstrom et al., 1999, 2003). We used a luciferase reporter assay to verify the repressor activity of the Gli2DR-GFP fusion protein (Fig. 1B). Both the Gli2DR protein and the Gli2DR-GFP fusion protein efficiently blocked Gli1-induced luciferase activity of an 8x Gli Binding Site (GBS) luciferase reporter construct (Sasaki et al., 1997).

image

Figure 1. Heat-shock–inducible activation and repression of Hedgehog (Hh) signaling. A: Four transgene constructs designed to allow heat-shock manipulation of Hh/Gli signaling. (i) hsp:shh activator line [Tg(hsp70l:shha-EGFP)]: (ii) hsp:dnPKA activator line [Tg(hsp70l:dnPKA-BGFP)]: (iii) hsp:gli1 activator line [Tg(hsp70l:gli1-EGFP)]. (iv) hsp:gli2DR repressor line [Tg(hsp70l:gli2DR-EGFP)]. B: Luciferase reporter assay. The modified Gli2DR protein and the Gli2DR-EGFP fusion protein caused an 80% reduction in luciferase activity in the presence of Gli1. C–F: Live images of 25 hours postfertilization (hpf) transgenic embryos, 4 hr after the completion of a 1-hr heat shock. Insets show green fluorescent protein (GFP) expression in the somites of the trunk. C: In the hsp:shh line, diffuse GFP expression is consistent with cytoplasmic localization of the Shhc-term-EGFP fusion protein. D,E: Cytoplasmic expression of the dnPKA-BGFP fusion and Gli1-EGFP fusion proteins. F: The Gli2DR-EGFP fusion protein is localized to nuclei. G–K: 48 hpf embryos that were heat shocked at 10 hpf. G: Wild-type embryo, inset shows V-shaped somites at 24 hpf. H: hsp:dnPKA embryo displaying Hh over expression defects including ventrally reduced eyes (arrow), dorsal brain defects (arrowhead), and slightly flattened somites at 24 hpf typical of Hh over-expressing embryos (Koudijs et al., 2008). I: hps:gli1 embryo with no visible Hh over-expression defects. Somites appear morphologically normal at 24 hpf (inset). J: hsp:shh embryo showing more severe Hh gain of function defects including flattened somites (arrowhead and inset) and reduced eyes (arrow). K: hsp:gli2DR embryo showing defects associated with a loss of Hh signaling, including U-shaped somites (arrowhead and right inset) and ventrally positioned but well formed eyes. lim2.1 labeling reveals an ectopic midline lens (arrow, left inset) in a 30 hpf hsp-gli2DR embryo that was heat shocked at the 8-somite stage. L: Forty-eight hpf yot(gli2DR) mutant embryo with Hh loss of function defects including u-shaped somites (arrowhead and right inset) and a lim2.1-labeled ectopic midline lens (left inset, arrow). M: Hh overexpression defects in the somites (arrowhead) and forebrain (arrow) caused by injection of 100 pg of shh mRNA at the 2-cell stage. fb, forebrain; hb, hindbrain; mb, midbrain; op, otic placode; s, somite.

Download figure to PowerPoint

Transgenesis was done as previously described (Kikuta and Kawakami, 2009; Suster et al., 2009) and founders were identified by heat-shocking progeny and screening for GFP fluorescence. Lines were validated by examining the effects of heat shock on ptch2 expression by in situ hybridization and real-time PCR (Figs. 2, 3). All four lines have been through at least five generations of outcrossing without observable changes in transgene efficacy. For convenience, these four lines will subsequently be referred to as hsp:shh, hsp:dnPKA, hsp:gli1, and hsp:gli2DR.

image

Figure 2. Timing and relative magnitude of Hedgehog (Hh) signal manipulation. A: Wild-type ptch2 expression in the trunk at 32 hours postfertilization (hpf). Cross-section shows spinal cord and somite ptch2 expression. B: Dramatically expanded ptch2 expression in the hsp:shh line 7 hr after heat shock. Cross-section shows dorsal expansion in the spinal cord and lateral expansion throughout the somites. C: ptch2 expression is dramatically reduced 5 hr following heat-shock–induced activation of the gli2DR-GFP transgene in the trunk and head (inset). D,E: ptch2 expression is expanded in both the somites and spinal cord in the hsp:dnPKA (D) and hsp:gli1 (E) transgenic lines 7 hr after heat shock. F: Nine hours following heat-shock treatment of the hsp:gli2DR line ptch2 expression reappears, with expression expanded into inappropriate dorsal domains of the spinal cord (arrow). G: Normal ptch2 expression in the head of a wild-type embryo 7 hr after heat shock. H–J: Dorsally expanded ptch2 expression in hsp:shh, hsp:dnPKA, and hsp:gli1 activator lines,7 and 9 hr after heat shock, respectively. J: Ectopic and mosaic ptch2 expression was consistently seen in the hsp:gli1 line (arrowheads). K: F59 antibody labeling reveals the normal slow muscle fiber distribution in a 38 hpf wild-type embryo that was heat shocked at the 10-somite stage (14 hpf). Cross-section of somite 17 shows the lateral position of these slow muscle fibers. L: Heat-shock activation of the hsp:shh transgene at the 10-somite stage led to ectopic slow muscle fiber differentiation starting at somite 16–17. Left panel shows normal slow muscle fibers in somite 7, which had differentiated before heat shock. Middle panel shows lateral view of trunk, with disrupted fiber visible starting around somite 17. Right panel shows a cross-section of somite 17 with increased numbers of slow muscle fibers. M,N: hsp:gli1 and hsp:dnPKA lines show subtle somite morphology defects (arrows). O: Activation of the hsp:gli2DR transgene at the 10-somite stage led to a loss of slow fibers starting at somite 16. Slow fiber differentiation appears relatively normal in somite 23, indicating Hh signaling was no longer blocked when this somite was forming. P: Time course of ptch2 mRNA expression in all four lines following a 1 hr heat shock starting at 24 hpf. Symbols report the mean value (± SEM) for each treatment at a given time point, error bars are visible only if larger than symbols. fb, forebrain; hb, hindbrain; mb, midbrain; nc, notochord; s, somite; sc, spinal cord.

Download figure to PowerPoint

image

Figure 3. Heat-shock–induced manipulation of Hedgehog (Hh) signaling at larval and juvenile stages. A: Reduced ptch2 expression in the ventricular regions of the brain 6 hr post heat shock in hsp:gli2DR transgenic 1-week-old larvae (compare top two panels) and 1-month-old juveniles (bottom two panels), as visualized by in in situ hybridization (ISH). Quantitative polymerase-chain reaction (qPCR) analysis revealed that ptch2 mRNA levels were reduced approximately 70% and 30% relative to non–heat-shocked 1-week and 1-month transgenics, respectively (right). EGFP mRNA levels were increased 4.8- and 2.2-fold in larvae and juveniles, respectively. B: Regionally increased Shh signaling in double heat-shocked hsp:shh transgenic fish visualized by ISH. In larvae, ptch2 expression was similar to non–heat-shocked transgenic and wild-type fish when examined by ISH, with ectopic expression seen in the anterior tectum (arrows). In juveniles, ptch2 expression was increased in its normal expression domain. No changes in ptch2 gene expression were seen in non–heat-shocked transgenic fish (right insets). EGFP mRNA was induced throughout the central nervous system in juveniles (inset). qPCR analysis revealed 1.8- and 1.6-fold increases in ptch2 expression in larvae and juveniles, respectively. EGFP mRNA levels were increased 15- and 25-fold, indicating effective heat-shock activation of the transgene at both ages. cb, cerebellum; di, diencephalon; hb, hindbrain; tec, tectum; tel, telencephalon. ***P < 0.001, **P < 0.01, *P < 0.05.

Download figure to PowerPoint

To characterize the timing and location of transgene activation following heat-shock embryos derived from crosses between identified transgenic carriers were heat shocked (hs) at 20 hours postfertilization (hpf) for 1 hr at 37°C and monitored for GFP fluorescence (Fig. 1C–F). In all lines, GFP fluorescence was visible 2 to 3 hr after heat shock when observed using a fluorescent dissecting microscope, and reached maximum intensity within approximately 4 hr. GFP fluorescence was detectable 30 min to 1 hr earlier using a confocal microscope. Confocal analysis revealed GFP fluorescence predominantly in the cytoplasm of cells in the hsp:shh, hsp:dnPKA, and hsp:gli1 lines (Fig. 1C–E), while GFP labeling was restricted to nuclei in hsp:gli2DR transgenic fish (Fig. 1F), consistent with this truncated gli protein acting as dominant repressor of transcription in the nucleus.

We next tested whether these transgenic lines appropriately alter Hh signaling in vivo by heat-shocking embryos from each transgenic line at the tail bud/one-somite stage (10 hpf), a time when Hh signaling is required for zebrafish embryonic patterning (Bergeron et al., 2011; Huang et al., 2012). Heat shock of hsp:shh, hsp:dnPKA, and hsp:gli1 transgenic embryos at 10 hpf led to variable defects associated with a gain of Hh function, including visible somite, brain, and eye defects (Fig. 1H–J) similar to those observed in embryos injected with shh mRNA (Fig. 1M and Hammerschmidt et al., 1996). Activation of the shh-GFP transgene gave the most prominent defects, with the hsp:gli1 line producing milder defects, similar to a previously described hsp:GFP-gli1 line (Huang and Schier, 2009). These data indicate that heat-shock–induced activation of all three transgenes results in up-regulation of Hh signaling, albeit to different degrees. Differences seen between lines could be due to the variations in levels of Hh signal activation, the differential timing required for each line to activate Hh signaling, the stability and the half-life of the fusion proteins, or the different functions of each protein in the Hh pathway.

Heat-shock–induced expression of the gli2DR-GFP transgene during early development (at 10 hpf) produced defects similar to those seen in the yot(gli2DR) mutant (Karlstrom et al., 1996). Visible defects include U-shaped somites, ventrally misplaced but well-formed eyes, and an ectopic lens at the midline (Fig. 1K,L). These embryos also had posterior tail defects not seen in yot mutants (see Fig. 2). These tail defects may be due to the broad expression of the Gli2DR protein in all cells of the embryo (as opposed to being restricted to the gli2 expression domains in yot), the timing of transgene activation, and/or because the gli2DR protein is expressed at higher levels than in mutant animals. Overall, these results suggest that the Gli2DR-EGFP fusion protein effectively blocks Gli-mediated Hh signaling in vivo as observed in vitro (Fig. 1K).

Duration and Relative Magnitude of Hh Signal Manipulation

To determine how activation of these transgenes affects Hh signaling, we heat-shocked embryos at 24 hpf and monitored expression of the Hh transcriptional target gene ptch2 (formerly ptc1 in zebrafish) by in situ hybridization (Fig. 2A–J) and real-time PCR (Fig. 2P). For all three activator lines (hsp:shh, hsp:dnPKA, and hsp:gli1), ptch2 expression was noticeably increased 3 hr after the onset of heat shock, with high levels of ptch2 gene expression seen by 5–12 hr post heat shock (Fig. 2B,D,E,P). ptch2 expression levels gradually decreased over the next 12 hr but some ectopic expression remained 19–24 hr post heat shock (data not shown). Robust ptch2 expression was seen throughout the somites of all lines (Fig. 2B,D,E). In the central nervous system (CNS), ptch2 expression was expanded dorsally in the spinal cord and brain, but was not seen in all cells (Fig. 2G–J), similar to what was reported using a hsp:GAL4/ UAS:shha Shh overexpression system (Balczerski et al., 2012). Because all three activator transgenes were expressed robustly throughout the CNS following heat shock at 20 (Fig. 1) and 24 hpf (data not shown), this may reflect either that activation levels are sub-threshold for ptch2 induction or that a subset of CNS cells are not competent to respond to Hh signaling after 24 hpf. Gli1–GFP activation resulted in more punctate labeling in the dorsal brain (Fig. 2J), which may indicate higher levels of Hh signal activation but in fewer cells.

Heat-shock–induced activation of the gli2DR-GFP transgene resulted in near complete loss of ptch2 expression 5 hr after heat shock in the spinal cord, somites, and brain (Fig. 2C). Nine hours after heat-shock treatment, ptch2 expression had recovered in the spinal cord, with ectopic dorsal expression in the spinal cord (Fig. 2F). This dorsal expression in the recovery period is likely due to the loss of the Ptch2 “boundary effect,” in which ventrally expressed Ptch proteins bind Shh and thus restrict its diffusion (reviewed in Goodrich et al., 1997; Ribes and Briscoe, 2009). The loss of Ptch2 protein following heat shock would allow the Shh protein to diffuse dorsally and activate ptch2 gene expression in more dorsal regions until high levels of Ptch2 are re-established ventrally.

To more precisely determine the time between heat-shock–induced transgene activation and changes in functional Hh signaling levels, we took advantage of the clocklike addition of new somites in the trunk during the segmentation period (Kimmel et al., 1995), and the fact that Hh signaling is necessary and sufficient to induce slow muscle fiber differentiation during this period (Karlstrom et al., 1999; Barresi et al., 2000). hsp:gli2DR and hsp:shh embryos were heat-shocked for 1 hr at the 10 somite (S) stage (14 hpf) and slow muscle fibers were examined at 38 hpf using the F59 antibody (Devoto et al., 1996). During this period, somites form at a rate of approximately two somites per hour. In both the hsp:shh and hsp:gli2DR lines, major slow muscle fiber defects were present starting at somite 16, representing a delay of approximately 3 hr from the onset of heat shock (Fig. 2L,O). Overexpression of Shh in the hsp:shh line resulted in increased slow muscle fiber numbers starting at somite 16–17 (Fig. 2L). Ectopic fibers were seen in all somites after somite 17, indicating that high levels of Hh signaling persisted until the end of somitogenesis at approximately 26 hpf, or for over 9 hr of developmental time. More subtle slow fiber and somite boundary defects were seen in both the hsp:gli1 and hsp:dnPKA lines with similar timing, starting at somite 16 and persisting through all remaining somites (Fig. 2M,N). In contrast, loss of Hh signaling was more temporally restricted in the hsp:gli2DR line, with slow muscle fibers being affected in somite 16 but beginning to reappear in somite 22 (Fig. 2O). This reveals that a single heat shock of the hsp:gli2DR line lowers Hh signaling to a level below the threshold needed for slow fiber differentiation for approximately 3 hr.

We next used quantitative real-time polymerase chain reaction (qPCR) to measure the effects of transgene activation on Hh signaling levels. Heat shock of the three activator lines induced ptch2 expression more than three-fold, with distinct kinetics for each line (Fig. 2P). ptch2 expression steadily increased up to 24 hr after the completion of a 1-hr heat shock in the hsp:shh and hsp:dnPKA lines, while in the hsp:gli1 line ptch2 levels peaked ∼5-hr post heat shock and then returned to wild-type levels by 24 hr. In the hsp:gli2DR line, ptch2 levels were reduced to 30% of wild-type levels within 4 hr of heat shock, remained low through 13 hr, and then returned to wild-type levels by 24-hr post heat shock. The more rapid and transient effects seen in the hsp:gli1 and hsp:gli2DR lines may reflect the fact that these transgenes encode the Hh-responsive transcription factors themselves, which can immediately act on target gene transcription. Furthermore, a rapid turnover of the Gli-GFP fusion proteins might be expected for highly regulated transcription factors.

Conditional Hh Signal Manipulation at Postembryonic Stages

We next assessed the utility of our transgenic lines to manipulate Hh/Gli signaling at larval and juvenile stages. We heat shocked 1-week-old larvae or 1-month-old juvenile transgenic fish at 37°C for 1 hr and examined the expression of GFP and ptch2 using in situ hybridization (ISH) and quantitative real-time PCR (qPCR). As seen at embryonic stages, heat-shock activation of the gli2DR-GFP transgene in 7-dpf larvae resulted in ptch2 down regulation within 6 hr. ISH labeling revealed a clear reduction of expression throughout the fish and qPCR analysis indicated a 70% reduction in ptch2 mRNA (Fig. 3A). In 1-month-old juvenile fish, ptch2 expression was visibly reduced when assayed by ISH and was reduced by 30% when assayed by qPCR (Fig. 3A). GFP mRNA was induced approximately five- and two-fold in 1-week larvae and 1-month juveniles, respectively, relative to a non–heat-shocked transgenic fish.

The hsp:gli2DR repressor line remained effective in reducing ptch2 expression through larval and juvenile stages (Fig. 3A) while the activator lines became less effective in inducing ptch2 expression with increasing age. The hsp:dnPKA and hsp:gli1 lines only weakly affected Hh signaling as determine by ISH analyses starting at approximately 4 dpf (data not shown). The hsp:shh line continued to be effective at increasing Hh signaling through larval and adult stages, with a double heat-shock paradigm further increasing ptch2 expression (Fig. 3B, data not shown). We thus focused our analysis on the hsp:shh activator line at 1-week and later stages (Fig. 3B). Double heat shock of the hsp:shh line led to 1.8- and 1.6-fold increases in ptch2 expression in 7-dpf larvae and 1-month juveniles, respectively (Fig. 3B). ISH revealed increased ptch2 expression mainly in its normal expression domain around the ventricles, with a small region of ectopic ptch2 expression seen in the anterior tectum of 1-week larvae (Fig. 3B, arrows). Because transgene expression is clearly activated throughout the CNS at these ages (Fig. 3B inset), this suggests either that Hh levels are not elevated sufficiently to induce broad ectopic ptch2 expression at these ages, or that only a subset of CNS cells remain competent to express ptch2. For the hsp:shh line, qPCR analysis revealed that EGFP mRNA expression was induced 15- and 25-fold in larvae and juveniles, respectively (Fig 3B).

Several lines of evidence suggest that any “leakiness” of these HSP70l-driven transgenes is not sufficient to disrupt Hh signaling, as has been observed by others (Hans et al., 2009). First, both homozygous and heterozygous transgene carriers are viable and fertile, indicating any leaky transgene expression is insufficient to cause developmental defects associated with altered Hh signaling. Second, in the absence of heat shock, no GFP expression is seen in any tissue, with the exception of the lens in the hsp:dnPKA and hsp:gli2DR lines, a site of normal HSP70l gene activation (Blechinger et al., 2002). Third, ptch2 expression was unaltered in non–heat-shocked transgenic embryos derived from any of the lines at 24 hpf, 7 days, and at 1 month (Fig. 3, data not shown).

New Hh Signaling Reporter Lines

To better visualize cells that respond to Hh signaling we generated Hh reporter lines that express GFP under the control of a 900 bp zebrafish ptch2 promoter fragment containing an additional Gli recognition sequence (Kinzler and Vogelstein, 1990). We generated two independent transgenic lines that express either cytoplasmically localized EGFP (Fig. 4A) or EGFP containing a nuclear localization signal (nlsEGFP) (Fig. 4B–D). Both lines express GFP in all tissues known to express ptch2 mRNA (Fig. 4A–H), including the floor plate of the spinal cord, brain, and somites. This labeling is similar to but more intense than that seen in a line generated using an eight Gli Binding Site-repeat/minimal lens crystallin promoter containing construct (Schwend et al., 2010). The line encoding cytoplasmic GFP (Tg(GBS-ptch2:EGFP)) showed inappropriate expression in the otic placode (Fig. 4A), possibly due to a transgene position effect. The nuclear GFP expressing line (Tg(GBS-ptch2:nlsEGFP)) had no inappropriate expression and was found to be significantly brighter than the EGFP line. This line was more sensitive in reporting Hh signaling levels in vivo and accurately reported the ventral to dorsal gradient of Shh signaling levels (Fig. 4D). Transgene expression from this promoter was drastically reduced when Hh signaling was blocked using the small molecule inhibitor cyclopamine (Fig. 4J–L) and was up-regulated when Hh signaling levels were artificially increased by the injection of shh mRNA (Fig. 4M), indicating these transgenes are appropriately regulated by changes in Hh signaling levels. Both lines continue to express GFP in known Hh responsive regions through larval (Fig. 4N) and adult stages (Fig. 4O).

image

Figure 4. Transgenic zebrafish lines that report Hedgehog (Hh) signaling. A,B: Live 24 hours postfertilization (hpf) embryos imaged using a fluorescent dissecting microscope. The Tg(GBS-ptch2:EGFP) (A) and Tg(GBS-ptch2:nlsEGFP) (B) lines accurately report Hh signaling in responding tissues of the floor plate, ventral brain, and somites. Only the (Tg(GBS-ptch2:EGFP) line has inappropriate green fluorescent protein (GFP) expression in the otic placode (A, arrowhead). C: Confocal image of the head of an nlsEGFP expressing embryo showing expression in Hh responsive regions of the brain (compare with ptch2 in situ labeling in F). D,E: Confocal images of the trunk region in live 24 hpf nlsEGFP (D) and enhanced GFP (EGFP) (E) transgenic embryos show appropriate nuclear or cytoplasmic GFP labeling (arrowheads). GFP intensity in the spinal cord allows visualization of the ventral-to-dorsal Shh signaling gradient (triangles) (Stamataki et al., 2005). More laterally, GFP is expressed in Hh responsive slow muscle fibers (arrowheads). F: Shh responsive regions of the brain visualized by ptch2 ISH. G,H: In situ labeling for GFP mRNA is the same as the known ptch2 expression pattern in the brain (G) and trunk (H). I: GFP expression in the spinal cord and somites in an uninjected Tg(GBS-ptch2:EGFP) embryo. J–L: ptch2 and GFP mRNA expression are eliminated when embryos are treated with the Hh signaling inhibitor cyclopamine (Incardona et al., 1998). M: Dorsally expanded GFP expression following injection of Shha encoding mRNA. N: GFP expression in a live 8-day-old Tg(GBS-ptch2:EGFP) larvae includes the floor plate (arrowhead) and retinal ganglion cell axons of the eye (arrow). O: Sagittal section through the brain of a 6-month-old Tg(GBS-ptch2:EGFP) adult co-labeled to show neurons (HuC/D antibody) and nuclei (dapi). Retinal ganglion cell axons express EGFP in the optic chiasm and where they terminate in the tectum (arrowheads). Ventricular regions of the brain continue to be Hh responsive in adults (arrow). cb, cerebellum; di, diencephalon; fp, floor plate; hb, hindbrain; nc, notochord; oc, optic chiasm; op, otic placode; r, rhombomere; s, somite; tec, tectum; tel, telencephalon; zli, zona limitans interthalamica.

Download figure to PowerPoint

In conclusion, we have characterized a new set of zebrafish transgenic lines that allow the temporal manipulation and visualization of Hh signaling at embryonic, larval, and juvenile stages. New reporter lines accurately and sensitively mark Hh responding cells at all stages of the life cycle. Activation of Hh pathway components by heat-shock recapitulates known loss or gain of function Hh/Gli signaling phenotypes and remains effective into early adult stages, but becomes less robust with age, perhaps due to changes in heat-shock promoter activation and/or changes in tissue responsiveness (Murtha and Keller, 2003). Region-specific manipulation of Hh signaling using these lines is now possible using a micro-scale heating device (Hardy et al., 2007; Placinta et al., 2009). Two new Hh reporter lines provide a highly sensitive read-out of Hh signal activity throughout the life cycle. Together, these transgenic lines provide important new tools for the study of Hh/Gli signaling throughout embryonic and postembryonic stages, periods associated with major tissue growth as well as continued cell differentiation and tissue patterning. Detailed analyses of the mechanisms by which Hh/Gli regulates postembryonic cell proliferation, differentiation, and survival are particularly important given the role this signaling system plays in somatic stem cell regulation, tissue regeneration, and cancer.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Transgene Constructs

The hsp70l:shha-EGFP and hsp70l:gli1-EGFP transgenes was constructed using multisite Gateway technology (Invitrogen) and Gateway compatible tol2 kits generously supplied by the Chien and Lawson Labs (Kwan et al., 2007; Villefranc et al., 2007). Full-length shha or gli1 encoding DNA was amplified with attB1 and attB2 site containing PCR primers (Table 1) and a middle entry clone (pMEshha or pMEgli1) was created by BP recombination of the PCR product with the pDONR221 vector. To make the expression vectors, three fragment LR recombination reactions were performed using the pMEshh or pMEgli1, p5Ehsp70l, p3EEGFP, and pDestTol2pA fragments. To generate the hedgehog signaling reporter lines, a 0.9 kb ptch2 promoter region was amplified from zebrafish genomic DNA using primers that included an additional Gli Binding Site (GBS, GACCACCCA): attB4.GBS.ptch2.Fw and attB1.ptch2.Rv. This amplicon was then used in a BP reaction with the pDONRP4-P1R plasmid to create the ptch2 entry clone. The final expression clone was made using the LR reaction with the ptch2 entry clone and either the pME-EGFP (cytoplasmic EGFP) or pME-nlsEGFP (nuclear EGFP) plasmids (Kwan et al., 2007) and the pTolDestR4R2pA plasmid (Villefranc et al., 2007). The hsp70l:gli2DR-EGFP and hsp70l:dnPKA-BGFP transgenes were constructed using restriction fragment-based sub-cloning methods starting with the truncated Gli2DR transcription factor sequences encoded by the yotty119 mutant allele (Karlstrom et al., 1999) or the dnPKA sequence from (Hammerschmidt et al., 1996). These sequences were then combined with hsp70l-EGFP (Halloran et al., 2000) into the pTol2000 vector (Kawakami, 2007; Kwan et al., 2007). Gli activator and repressor activity was assayed using a luciferase reporter assay as in Tyurina et al. (2005) using the C17 neural cell line (Snyder et al., 1992).

Table 1. PCR Primer Sets for Generating Transgenic Constructs and for Identifying Transgenic Fish
attB1.shha.FwGGGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAGGAGATAGAACCATGCGGCTTTTGACGAGAG
attB2.shha.RvGGGGACCACTTTGTACAAGAAAGCTGGGTAGCTTGAGTTTACTGACATCCC
attB1.gli1.FwGGGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAGGAGATAGAACCATGGCGATGCCAGT
attB2.gli1.RvGGGGACCACTTTGTACAAGAAAGCTGGGTAAGATAACGTGTTTAGATAC
attB4.GBS.ptch2.FwGGGGACAACTTTGTATAGAAAAGTTGGACCACCCATGTTCACCCATATTTTGATC
attB1.ptch2.RvGGGACTGCTTTTTTGTACAAACTTGTGGGTGGTCAGTCCAAACGGGAGGCAGAAG
EGPF.FwCCTGAAGTTCATCTGCACCA
EGFP.RvGGGTGCTCAGGTAGTGGTTG
BGFP.FwCCTGTCCATGACAACACTTG
BGFP.RvGGTAAAAGGACAGGGCCATC

Fish Strains and Generation of Transgenic Lines

Wild-type and transgenic fish are maintained in the Karlstrom Lab Zebrafish Facility at the University of Massachusetts, Amherst. The embryos are incubated at 28.5°C and staged as described (Kimmel et al., 1995). The wild-type lines used are TL/GOL and WIK. Transgenic lines were generated as in (Kikuta and Kawakami, 2009; Suster et al., 2009). Briefly, 25 pg of purified plasmid DNA was injected to one- to two-cell zebrafish embryos with 25 pg of transposase RNA. Injected fish (potentially mosaic carriers) were raised to adulthood and out-crossed to wild-type individuals to identify potential founder fish by examining 24 hpf embryos for GFP fluorescence 4 hr after a 1 hr heat shock at 37°C. Founders were then outcrossed again and the progeny were raised to adulthood. These adults (potential stable transgenic fish) were screened for the presence of the GFP(EGFP or BGFP) transgene using PCR-based genotyping and DNA isolated from fin clips (see Table 1 for primer sequences). Three founder fish (F0) were identified for the Tg(hsp70l:shha-EGFP) line, two for the Tg(hsp70l:gli2DR-EGFP) and Tg(hsp70l:dnPKA-BGFP) lines, and one for the Tg(hsp70l:gli1-EGFP) line. For convenience, these lines are abbreviated as hsp:shh, hsp:gli2DR, hsp:dnPKA, and hsp:gli1, respectively (see Fig. 1). Transgenic lines were compared for GFP fluorescence following heat shock. While the different founder lines behaved similarly, one founder from the hsp:shh line showed significantly brighter GFP fluorescence, and this line was used for larval and juvenile experiments. Hh reporter lines were generated similarly, with plasmid injected fish being grown to adulthood and out crossed to wild-type fish. The offspring were then screened under a fluorescent dissection microscope for GFP expression between 1 and 4 dpf. GFP expressing larvae were raised to establish stable transgenic lines. One founder was identified for the GBS-ptch2:EGFP line, and four founders were found for the GBS-ptch2:nlsEGFP line. The four different GBS-ptch2:nlsEGFP lines show same expression pattern, but with different GFP intensities. The brightest line was chosen for all further studies. Lines/alleles have been registered with zfin.org as follows: Tg(GBS-ptch2:EGFP)umz23, Tg(GBS-ptch2:nlsEGFP)umz24, Tg(hsp70l:shha-EGFP)umz30, Tg(hsp70l:gli1-EGFP)umz31, Tg(hsp70l:dnPKA-BGFP)umz32, Tg(hsp70l:gli2aDR-EGFP)umz33.

Heat Shock and Drug Treatment

Staged embryos and larvae (up to 7 dpf) were placed into a 50-ml conical tube containing 25 ml of embryo raising medium (Westerfield, 2000) and the tube was then placed in a 37°C water bath for 1 hr. After heat shock, fish were returned to Petri dishes and placed at 28.5°C. For 1-month juveniles, a beaker containing 500 ml of system water was preheated to 37°C in a water bath. Fish were netted and transferred directly into the beaker for 1 hr. Double heat shocks were for 1 hr (37°) at t = 0 and t = 13 hr. After heat shock, 100 ml of room temperature system water was added to the beaker and fish were placed at 28.5°C. GFP fluorescence was visualized using a Leica MZ 16FA fluorescence dissecting scope. To validate the appropriate response to Hh signal manipulation, Tg(GBS-ptch2:EGFP) or Tg(GBS-ptch2:nlsEGFP) embryos were (1) incubated in 100 μM cyclopamine (LC Labs) from 6 to 24 hpf to block Hh signaling as in (Sbrogna et al., 2003) or (2) injected with 100 pg shha mRNA at the one- to two-cell stage to activate Hh signaling. Embryos were assayed either for GFP fluorescence or for ptch2 expression by ISH at 24 hpf. Confocal imaging was done on a Zeiss LSM-700 confocal microscope.

In Situ Hybridization and Immunohistochemistry

Fish were fixed in 4% paraformaldehyde overnight at 4°C, dehydrated in 100% methanol, and stored at −20°C. Whole mount ISH was performed on fish up to 4 dpf as previously described (Karlstrom et al., 1999). For larval and juvenile stages, 2–10 fixed fish per experimental condition were rinsed with PBS, placed in embedding media (1.5% agar, 5% sucrose), and then immersed in 30% sucrose solution overnight at 4°C. Embedded samples were sectioned (20 μm) using a Leica CM 1950 cryostat. Immunohistochemistry was performed as previously described (Karlstrom et al., 1999). Primary antibodies used were rabbit anti-GFP (1:400, Invitrogen) and the F59 monoclonal antibody that recognizes the myosin heavy chain to label slow adaxial muscle cells (1:10, Developmental Studies Hybridoma Bank), and mouse IgG2b anti-HuC/D (1:200, Invitrogen). Secondary antibodies used were goat anti-rabbit alexa 488, goat anti-mouse alexa 546, and goat anti-mouse IgG2b Alexa 546 (1:800, Invitrogen).

Quantitative Real-Time PCR (qPCR) and Statistical Analysis

Six replicates were collected for each experimental condition. RNA was purified from homogenized whole fish using Trizol (Invitrogen). Five to 10 embryos, 3–6 larvae, or 1 juvenile were used per replicate. cDNA was synthesized from 50–100 ng total RNA using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). The MxPro3000P RT-PCR system (Stratagene) was used to determine relative gene expression levels. A total reaction volume of 10 μl contained 1 μl cDNA template, 5 μl FastStart Universal SYBR Green Master Kit (Roche), 400 nM forward and reverse primers (Table 2). RT-PCR conditions were as follows: 1 cycle of 95°C for 15 min; 40 cycles each of 95°C for 15 sec, 60°C for 30 sec, 72°C for 30 sec. All values were normalized to ef1a mRNA levels. qPCR data were analyzed using the ΔΔCt method (Livak and Schmittgen, 2001; Pfaffl, 2001). Relative gene expression levels are compared with heat-shocked wild-type (Fig. 2) or non–heat-shocked transgenic controls (Fig. 3). All analyses were conducted using GraphPad Prism software (San Diego, CA). Comparisons between non–heat-shocked and heat-shocked groups were analyzed by Student's t test.

Table 2. qPCR Primer Sets
ptch2.FwGCCGCATCCCAGGCCAACATBGFP.FwGTCAGTGGAGAGGGTGAAGG
ptch2.RvCGTCTCGCGAAGCCCGTTGABGFP.RvTACATAACCTTCGGGCATGG
EGFP.FwACGTAAACGGCCACAAGTTCef1α.FwCTGGTGTCCTCAAGCCTGGTA
EGFP.RvAAGTCGTGCTGCTTCATGTGef1α.RvACTTGACCTCAGTGGTTACATTGG

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

The authors thank Judy Bennett for fish care and the members of the Karlstrom lab for help proofing the manuscript. Special thanks go to Nathan Lawson and the late Chi-Bin Chien for their generosity in sharing tol2Kit cloning constructs and to Mary Halloran for sharing hsp70l promoter constructs. R.O.K. was funded by the NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES