Netrin-1 acts as a survival factor via its receptors UNC5H and DCC

Abstract

The membrane receptors DCC and UNC5H have been shown to be crucial for axon guidance and neuronal migration by acting as receptors for netrin-1. DCC has also been proposed as a dependence receptor inducing apoptosis in cells that are beyond netrin-1 availability. Here we show that the netrin-1 receptors UNC5H (UNC5H1, UNC5H2, UNC5H3) also act as dependence receptors. UNC5H receptors induce apoptosis, but this effect is blocked in the presence of netrin-1. Moreover, we demonstrate that UNC5H receptors are cleaved in vitro by caspase in their intracellular domains. This cleavage may lead to the exposure of a fragment encompassing a death domain required for cell death induction in vivo. Finally, we present evidence that during development of the nervous system, the presence of netrin-1 is crucial to maintain survival of UNC5H- and DCC-expressing neurons, especially in the ventricular zone of the brainstem. Altogether, these results argue for a role of netrin-1 during the development of the nervous system, not only as a guidance cue but as a survival factor via its receptors DCC and UNC5H.

Fig. 1. UNC5H receptors induce apoptosis in 293T or 13.S.24 cells. HEK 293T cells or rat olfactory neuroblasts 13.S.24 were transiently transfected as described previously (Bordeaux et al., 2000) with either the pCMV control plasmid (cont.) or the UNC5H1-3 expression plasmids and harvested 48 h after transfection. (A) UNC5H protein expression induces cell death. Cell death was monitored by Trypan blue exclusion as described in Materials and methods. (B) UNC5H2 protein expression induces caspase activation in 293T cells. Caspase-3 activation was measured as described in Materials and methods. Index of relative caspase activity is presented as the ratio between the caspase activity of the sample and that measured in 293T cells transfected with pCMV. Inset: UNC5H2 protein expression was checked by western blot analysis using an anti-HA antibody. (C) UNC5H2 expression induces DNA fragmentation in 13.S.24 cells. TUNEL assays were performed on cytospun 13.S.24 cells as described in Materials and methods. Phase contrast of control cells and TUNEL staining of either control (cont.) or UNC5H2-transfected cells is shown. Scale bar, 10 µm.

Fig. 2. Netrin-1 blocks UNC5H2-induced cell death. 293T cells were transiently transfected with either the pCMV control plasmid (cont.) or the UNC5H2 expression construct. Twenty-four hours after transfection 250 ng/ml of purified netrin was added or not to UNC5H2-transfected cells. Twenty-four hours later, UNC5H2 expression was analysed by western blot using an anti-HA antibody (A), and cell death (B) was monitored by measuring caspase activity as described for Figure 1.

Fig. 3. UNC5H proteins are cleaved by caspase. (A) 293T cells were transfected with the UNC5H2 expression plasmid and incubated for 24 h in the presence or not of 20 µM zVAD-fmk. Cell death was then analysed by measuring caspase activity as described for Figure 1. (B) UNC5H proteins are caspase substrates. In vitro translated full-length UNC5H1, UNC5H2 or UNC5H3 were incubated without caspase (–) or with purified caspase-3 (0.3 µM) or caspase-7 (1.4 µM) for 2 h as indicated. An autoradiograph is shown. (C) Mapping of the caspase cleavage site. Similar experiment to that in (B) except that the UNC5H2-IC and mutant UNC5H2-IC D412N were in vitro translated and incubated with caspase-3. (D) Diagram of UNC5H.

Fig. 4. Caspase cleavage and the presence of death domain are required for UNC5H2-induced cell death. (A) 293T cells were transfected with the UNC5H2 or UNC5H2 D412N construct and cell death was then analysed by measuring caspase activity as described for Figure 1. Inset: expression of wild-type UNC5H2 or mutant UNC5H2 D412N was checked by western blot analysis using an anti-HA antibody. (B) 293T cells were transfected with the UNC5H2 expression plasmid, the constructs allowing expression of full-length UNC5H2 deleted from amino acids 413 to 946 (UNC5H2-δ413–946) or from amino acids 857 to 946 (UNC5H2-δDD), or constructs allowing expression of the UNC5H2 fragment from the caspase cleavage (amino acids 413–946) with or without addition of a myristoylation signal. Cell death was then analysed by measuring caspase activity as described for Figure 1.

Fig. 5. Netrin-1 as a regulator of survival of UNC5H- or DCC-expressing precerebellar neurons. (A and B) TUNEL-positive cells on sections treated with the Apoptag kit and revealed with the substrate solutions NBT/BCIP (nitroblue tetrazolium/5-bromo-4-chloro-3-indoylphosphate; Roche Diagnostics). At E12, the TUNEL reactivity was higher in homozygous netrin-1 mutant mice (B) than in wild-type embryos (A). (C and D) In situ hybridization at E12 using either a UNC5H2 probe (C) or a DCC probe (D). Note that TUNEL-positive reactivity is increased in the area expressing UNC5H2 (arrowhead in C) and DCC transcripts (arrow in D), which contains, among others, precursors of precerebellar neurons. Sensory ganglions (empty arrows) display numerous TUNEL-positive neurons in both control wild-type (A) and homozygous mutant (B) embryos.