Shh influences cell number and the distribution of neuronal subtypes in dorsal root ganglia

Abstract

The molecular mechanisms responsible for specifying the dorsal-ventral pattern of neuronal identities in dorsal root ganglia (DRG) are unclear. Here we demonstrate that Sonic hedgehog (Shh) contributes to patterning early DRG cells. In vitro, Shh increases both proliferation and programmed cell death (PCD). Increasing Shh in vivo enhances PCD in dorsal DRG, while inducing greater proliferation ventrally. In such animals, markers characteristic of ventral sensory neurons are expanded to more dorsal positions. Conversely, reducing Shh function results in decreased proliferation of progenitors in the ventral region and decreased expression of the ventral marker trkC. Later arising trkA(+) afferents make significant pathfinding errors in animals with reduced Shh function, suggesting that accurate navigation of later arising growth cones requires either Shh itself or early arising, Shh-dependent afferents. These results indicate that Shh can regulate both cell number and the distribution of cell types in DRG, thereby playing an important role in the specification, patterning and pathfinding of sensory neurons.

Figure 1

Programmed cell death and proliferation of early DRG cells are stimulated by Shh in vitro. (A) Top panel: Double labeling of cells with DAPI (blue) and TUNEL (green) in cultures derived from stage 23 dissociated DRG. Bottom panel: The percentage of TUNEL⁺ cells out of total DAPI⁺ DRG cells was determined from at least 100 DAPI⁺ cells in each of 7 independent experiments. (B) Top panel: Double labeling of cells with DAPI (blue) and BrdU (red) in cultures derived from stage 23 dissociated DRG. Bottom panel: The percentage of BrdU⁺ cells out of total DAPI⁺ DRG cells was determined from at least 100 DAPI⁺ cells in each of 5 experiments. *Shh significantly different from control (t-test, p<0.01). Arrowheads indicate doubly labeled cells. Arrows indicate cells unlabeled by TUNEL or BrdU.

Figure 2

Exogenous Shh selectively stimulates programmed cell death of DM DRG cells in vivo. (A) At stage 23, no TUNEL⁺ cells were observed in DRG for any condition (not shown). However, numbers of TUNEL⁺ cells in the apical ectodermal ridge and distal mesenchyme of the limb were decreased by Shh and increased by 5E1 at this stage, indicating that these treatments affect cell death as expected (Chiang et al., 1996). (B) At stage 25, TUNEL staining was observed in DRG of control, Shh and 5E1 treated embryos. Cartoon illustrates the division of DRG into DM vs. VL regions, as described in MATERIALS AND METHODS. (C) TUNEL⁺ cells in DM and VL regions of DRG LS1, LS2 and LS3, with data from at least three independent experiments shown. *Significantly different from DM controls (t-test, p<0.01). #Significantly different from VL Shh treated (t-test, p<0.01).

Figure 3

Exogenous Shh predominantly stimulates the proliferation of VL DRG cells. (A) Examples of BrdU labeling of DRG from control, Shh and 5E1 treated embryos at stage 25. Cartoon illustrates the division of DRG into DM vs. VL regions, as described in MATERIALS AND METHODS. (B) The number of BrdU⁺ cells in stage 23 LS2 ganglia is increased equally in both the DM and VL regions following Shh treatment at stage 21, while 5E1 treatment results in a reduction of BrdU⁺ cells throughout the ganglia. *Significantly different (p<0.01; t-test). Both treatments are also different from control ganglia (p<0.01; t-test). (C) At stage 25, Shh alters proliferation in a D/V pattern. BrdU⁺ cells in DM and VL regions of LS1, LS2 and LS3 DRG from control, Shh, 5E1 treated embryos, with data from at least three independent experiments shown. *Significantly different from 5E1 treated (t-test, p<0.02). #Significantly different from DM Shh treated (t-test, p<0.02).

Figure 4

Expansion of trkC expressing cells in animals with altered Shh function. At stage 23, LS2 DRG stained for trkC show comparable numbers of labeled cells and axons in control and Shh treated animals, while animals treated with 5E1, a Shh function-blocking antibody, have reduced numbers of trkC⁺ cells. By stage 25, in Shh treated animals there is a notable increase in the number of trkC⁺ cells, and a corresponding decrease in 5E1-treated animals. All images are mid-ganglion sections of the LS2 DRG. White lines enclose trkC⁺ cells, to illustrate approximate distribution of these cells within the ganglion.

Figure 5

Altered Shh function in vivo alters the normal D-V pattern of cUnc-5H3 expression. In situ hybridization on cryostat sections of stage 25 lumbar DRG from control and Shh treated embryos. Shh treatment increased the expression of cUnc-5H3 and expanded the domain of expression to more dorsal positions, an effect that appeared somewhat more prominent in more posterior ganglia (cf. LS2 to LS4). Lines outline the ganglion and show the DV midpoint, as in Figs. 2B and 3A. cUnc-5 expressing cells are concentrated near the outer margin of the ganglia and images reflect the maximum dorsal extent of cUnc-5 expressing cells for each condition. Because sections are not taken from the middle of the ganglia, the lines do not accurately reflect maximum DRG dimensions at the mid-point.

Figure 6

Pathfinding of late arising trkA⁺ cells is disrupted in animals with reduced Shh function. (A) A control embryo treated with vehicle at stage 21 and stained for trkA at stage 25. Axons are tightly fasciculated (bracket) and extend along preexisting pioneers (Guan et al., 2003). (B-D) Animals treated with 5E1 to reduce Shh function. (B) TrkA⁺ fibers extend in an aberrant dorsal-lateral direction (arrowheads). (C) TrkA⁺ fibers misroute laterally (arrowhead), the ventral nerve is broad and bifurcated (bracket) with fibers extending in ectopic medial positions (arrow). Similar errors were seen in 29 out of 60 5E1-treated ganglia (see Table 3). (D) Ventral route is broad and bifurcated (bracket) with fibers extending in aberrant medial positions (arrow). Medially projecting axons were seen in 41 out of 60 5E1-treated ganglia (see Table 3). Scale bar in (A) is 50 μm.

Figure 7

Summary of findings. In control conditions, DM and VL DRG neurons express distinct markers (Guan and Condic, 2003; Guan et al., 2003) and predominantly extend to different targets (skin and muscle, respectively). For example, trkA⁺ neurons are present throughout the DRG at this stage, while trkC⁺ neurons are found predominantly in the VL regions, with trkB⁺ neurons in the middle third of the D-V extend of the ganglia (see Supplemental Figure). In Shh gain of function conditions, DRG size increases concomitant with an increase in proliferation throughout the DRG at stage 23 that is maintained in ventral regions until at least stage 25. Shh treatment induces an expansion of ventral markers, and causes a significant increase in dorsal cell death at stage 25. When Shh function is reduced, proliferation of early progenitors (at stage 23) is reduced and the expression of the ventral marker trkC is suppressed.