Cooperative requirement of the Gli proteins in neurogenesis

Abstract

The Gli proteins are critical components of multiple processes in development, homeostasis and disease, including neurogenesis and tumorigenesis. However, it is unclear how the Gli code, the sum of their combinatorial positive and negative functions, dictates cell fate and behavior. Using an antisense approach to knockdown gene function in vivo, we find that each of the three Gli proteins is required for the induction of all primary neurons in the amphibian neural plate and regulates the bHLH/Notch neurogenic cascade. Analyses of endogenous Gli function in Gli-mediated neurogenesis and tumorigenesis, and in animal cap assays, reveal specific requirements that are context specific. Nuclear colocalization and binding studies suggest the formation of complexes, with the first two zinc fingers of the Gli five zinc-finger domain acting as a protein-protein interaction site. The Gli proteins therefore appear to form a dynamic physical network that underlies cooperative function, greatly extending the combinatorial possibilities of the Gli code, which may be further fine-tuned in cell fate specification by co-factor function.

Fig. 1

Primary neurogenesis requires Gli1, Gli2 and Gli3. (A) PAGE analyses of in vitro transcription/translation of Gli1, Gli2 and Gli3 cDNAs mixed with 1 μM of each MO. (B) Percentage of reduction of N-tubulin⁺ primary neurons: motoneurons (M), interneurons (I), sensory neurons (S) and total neurons (T) in stage 14 embryos injected with each MO. Neurons were counted in the embryos with reduced N-tubulin expression. The percentage of each cell type was calculated by dividing the number of neurons on the injected side by the number of neurons on the control side. The statistical analysis was performed using Student’s t-test. Error bars are s.e.m. (C-H) Examples of N-tubulin expression on stage 14 embryos injected (inj.) with MO-1 (C), MO-2 (D), MO-3 (E,G), MO-C (F) and MO-3 plus GLI3 RNA (H), the latter showing phenotypic rescue. The β-gal tracer is red.

Fig. 2

Effects of Gli3 in neural plate patterning. Whole-mount in situ hybridizations on stage 14 embryos injected (inj.) with MO-C or MO-3 with probes as indicated on the left side of each panel. The β-gal tracer is red except in Pintallavis-MO-3 and the BrdU panels, in which it is blue. Arrows indicate the inhibition of Gli1, Xash3, XMyt1, Xaml, Slug and Snail expressions. Double-headed arrows indicate the size of the neural plate assessed by Sox3 expression. BrdU incorporation in stage 14 embryos injected with MO-3 is shown in brown. A high magnification panel (bottom right) shows BrdU incorporation in the neural plate of a unlilaterally injected MO-3+lacZ embryo.

Fig. 3

Gli1, Gli2 and Gli3 are required for ectopic neurogenesis by activated neurogenic pathways. (Left panels) N-tubulin expression (purple) on stage 14–15 embryos co-injected (inj.) with Ngn1, NeuroD or Delta1^stu and MO-C or MO-3, as indicated. Arrows indicate ectopic neuronal differentiation caused by Ngn1, NeuroD or Delta1^sm. Panels show dorsal views and the β-gal injection tracer is in red/pink for Delta1^Sm+MO-3. (Right panel) Lateral views of injected embryos at stage 13 showing the β-gal injection tracer in light blue. Ectopic neurogenesis as determined by N-tubulin expression (purple, arrow) is only detected after injection of 1 ng Ngn1a RNA plus MO-C, but not after co-injection of Ngn1a plus MO-1 or MO-2. Dorsal side is upwards, anterior towards the right.

Fig. 4

Gli-induced neurogenesis requires endogenous Gli function. (A,C) N-tubulin expression on stage 14 embryos co-injected (inj.) with Gli1 or human GLI1 and each MO. Arrows indicate ectopic neuronal differentiation. The β-gal tracer is red. (B,D) Percentage change of primary neurons: motoneurons (M), interneurons (I), sensory neurons (S) and total neurons (T) in stage 14 embryos injected with Gli1 (B) or human GLI1 (D) and each MO. Neurons were counted in the embryos that showed a phenotype. The percentage of each cell type was calculated by dividing the number of neurons on the injected side by the number of neurons on the control side. The statistical analysis was performed using Student’s t-test. Error bars are s.e.m.

Fig. 5

GLI1-induced tumorigenesis requires endogenous Gli1 and Gli3 function. (A) Tadpoles injected with human GLI1 and each MO. Arrows show tumors in tadpoles injected with GLI1 and MO-C or MO-2. (B) Sets of tadpoles injected with GLI1 and MO-C or MO-3 showing varying overall deformities present in tumor-bearing GLI1+MO-C-injected embryos but absent in GLI1+MO-3-injected siblings. The β-gal tracer is blue.

Fig. 6

Diagrammatic summary of Gli function. Frog proteins are in lower case whereas human proteins are in upper case. Green and red indicate activator or repressor activity. Arrows indicate induction and T bars repression. Thick lines refer to changes of fivefold or higher. The numbers next to the arrows or T bars indicate the requirement of the different endogenous Gli genes. The data derive from Table 1.

Fig. 7

Cooperative effects and physical interactions within the Gli superfamily. (A) (Top left) Human GLI1 is nuclear in transfected COS cells, here detected with specific polyclonal anti-human GLI antibodies (Lee et al., 1997). (Top right) Myc-tagged Gli2 is cytoplasmic. (Bottom left) Co-expression of GLI1 and Gli2 results in nuclear localization of Gli2 (bottom right). (B) (Top panels) Expression of Myc-tagged frog Gli1, Myc-tagged-Gli2, Myc-tagged GLI3 and FLAG-tagged Zic2 in COS cells. (Bottom panels) Co-expression with Zic2 forces the Gli1, Gli2 and GLI3 cytoplasmic proteins into the nucleus. DAPI staining highlights nuclei (blue). The color of the font indicates the protein visualized by immunostaining. (C,D) Quantification of the subcellular compartmentalization of Gli proteins when transfected alone versus co-transfection with human GLI1 (C), and when transfected alone versus co-transfection with Zic2 (D). (C) Only nuclear values are given for co-transfection of Gli2 and GLI1 and for GLI3 and GLI1 (E–H) (Top panels) Immunoprecipitation of Flag-tagged proteins with anti-FLAG Ab as labeled, followed by western blotting with anti-Myc Ab. Arrowheads indicate the immunoprecipitated species. (Bottom panels) Corresponding western analysis of total Myc-protein input from transfected COS cells as labeled. The horizontal alignment of bottom panels corresponds to the molecular weight markers on the left.