An intrinsic timer specifies distal structures of the vertebrate limb

Abstract

How the positional values along the proximo-distal axis (stylopod-zeugopod-autopod) of the limb are specified is intensely debated. Early work suggested that cells intrinsically change their proximo-distal positional values by measuring time. Recently, however, it is suggested that instructive extrinsic signals from the trunk and apical ectodermal ridge specify the stylopod and zeugopod/autopod, respectively. Here, we show that the zeugopod and autopod are specified by an intrinsic timing mechanism. By grafting green fluorescent protein-expressing cells from early to late chick wing buds, we demonstrate that distal mesenchyme cells intrinsically time Hoxa13 expression, cell cycle parameters and the duration of the overlying apical ectodermal ridge. In addition, we reveal that cell affinities intrinsically change in the distal mesenchyme, which we suggest results in a gradient of positional values along the proximo-distal axis. We propose a complete model in which a switch from extrinsic signalling to intrinsic timing patterns the vertebrate limb.

Figure 1. The environment appears to determine distal graft fate.

Blocks (150 μm) of GFP-expressing (depicted as green in this and all schematic representations) HH20 chick wing distal mesenchyme denuded of ectoderm and grafted under the AER of wild-type HH20 buds (a) give rise to structures distal to the stylopod (n=5/5, b,c, blue asterisks—j). GFP-expressing HH24 distal mesenchyme tissue grafted beneath the AER of wild-type HH24 buds (d) give rise to structures distal to the zeugopod (n=7/7, e,f, red asterisks—j). GFP-expressing HH20 distal mesenchyme tissue grafted beneath the AER of wild-type HH24 buds (g) give rise to structures distal to the autopod (n=8/8, h,i, black asterisks—j). Each asterisk represents the proximal boundary of the grafted tissue for each experiment. Note, h-humerus; u-ulna; r-radius; 1, 2 and 3 are the digits in anterior to posterior sequence. Scale bars: 1 mm.

Figure 2. Hoxa13 expression is intrinsically timed.

In grafts of HH20 distal mesenchyme cells made to the same stage wing buds (a), the presence of the grafted tissue does not perturb the establishment of a normal domain of Hoxa13 expression. Note that the graft cannot be distinguished by Hoxa13 expression at 24 h. (b,c) n=2/2—note area of grafted tissue is shown by Gfp expression in a serial section (b) and dashed lines. In grafts of GFP-expressing HH20 distal mesenchyme tissue grafted beneath the AER of wild-type HH24 buds (d) Hoxa13 is expressed in an intrinsically timed manner shown 24 h (e,f—n=3/3) and 48 h after grafting (g,h—n=3/3). Note that the arrows in (e,f) indicate Hoxa13 expression in distal part of graft and asterisk absence in proximal part of the graft. Scale bars: 100 μm.

Figure 3. AER duration is locally controlled by the distal mesenchyme.

In grafts of HH24 distal mesenchyme cells made to the same stage wing buds (a), the pattern of Fgf8 expression in the AER is indistinguishable between left and right buds after 48 h (b–d, n=2/2) and 72 h (e–g n=2/2). Note area of grafted tissue is shown by GFP expression (b,e) and lower panels are higher magnifications of areas marked with asterisks (c,f). In grafts of GFP-expressing HH20 distal mesenchyme tissue grafted beneath the AER of wild-type HH24 buds (h) Fgf8 is stronger in manipulated right buds compared with equivalent region of contralateral left buds after 48 h (i–k, n=2/2) and 72 h (l–n, n=2/2). Note area of grafted tissue is shown by GFP expression (i,l) and lower panels are higher magnifications of areas marked with asterisks (j,m). Left limb photos flipped horizontally for comparison in (c,f,j,m). Scale bars in (c,f,j,m)—100 μm and in their enlarged panels beneath—25 μm.

Figure 4. Cell cycle parameters are intrinsically timed.

Cell cycle parameters of HH20 (three repeat experiments to account for variation in staging), HH24 (6), HH27 (6), HH29 (2) and HH30 (5) distal mesenchyme tissue—bars indicate standard error of the mean between the repeat experiments (a). In each experiment between 8,000 and 10,000 cells were counted from 10 to 12 pooled strips of distal tissue each obtained from separate buds. Blocks (150 μm) of GFP-expressing HH20 distal mesenchyme tissue grafted beneath the AER of wild-type HH24 buds (b), after 48 h (c), G1=64.2%, S=13.4%, G2/M=22.4% phase cells in grafted right wing distal mesenchyme blocks (150 μm) compared with G1=70.4%, S=8.6%, G2/M=21.0% phase cells in equal numbers of contralateral left wing bud distal mesenchyme blocks (one pool of n=10 cubes of tissue, each taken from separate left control and experimental right buds). After 72 h (d), G1=74.0%, S=9.5%, G2/M=16.5% of cells in grafted right wing distal mesenchyme blocks (150 μm) compared with G1=81.3%, S=6.9%, G2/M=11.8% of cells in equal numbers of contralateral left wing bud distal mesenchyme blocks (one pool of n=10 cubes of tissue in both cases). In both experiments, more than 10,000 cells were counted and there is a significant difference in G1, S and G2/M numbers between left and right distal mesenchyme (Pearson's χ² test—P<0.0001) consistent with graft behaving intrinsically.

Figure 5. An intrinsic timer specifies distal positional values.

Disaggregated GFP-expressing HH24 and wild-type HH24 autopod progenitor cells re-aggregated into pellets and grafted to HH24 buds (a) show an approximately equal distribution in the inner and outer regions of grafts after 24 (b,c, n=2/2) and 48 h (d,e, n=3/3, green shows GFP immunofluorescence and blue DAPI staining—the inner/outer regions defined as half the distance from the centre of the graft to the periphery as indicated by yellow rings). Disaggregated HH24 GFP-expressing autopod progenitor cells and wild-type HH20 zeugopod progenitor cells re-aggregated into pellets and grafted to HH24 buds (f) sort out after 24 h (g,h, n=3/3) and 48 h (i,j, n=3/3). Note the difference in cell distribution between the homochronic and heterochronic mixed grafts was very significantly different (Mann–Whitney–Wilcoxon U-test—P-value <0.0001). Autopod progenitor cells (then at HH27—g or HH29—i) predominantly localize to the periphery of the grafts (between the yellow rings) to aggregate with host cells of the same age and zeugopod progenitor cells (then at HH24—g or HH27—i) are confined to the centre of the grafts. Development of grafts comprising of HH24 GFP-expressing autopod progenitor cells and wild-type HH20 zeugopod progenitor cells made to HH24 buds (k). Autopod progenitor cells (then at day 12) localize to edge of grafted tissue to contact host cells of same age and zeugopod progenitor cells (then at day 11) are confined to the centre of the graft (l,m n=4/5). Such grafts contribute to the autopod (n,o –purple asterisks n=5/5; compare with HH20–20 grafts, blue asterisks; HH24-HH24 grafts, red asterisks and HH20–24 grafts black asterisks). Note in (n) u-ulna; r-radius; 1, 2 and 3 are the digits in anterior to posterior sequence). Scale bars—100 μm in (b,d,g,i; 1 mm in l,n).

Figure 6. Model for chick wing proximo-distal patterning.

At early limb initiation stages (HH18/19) trunk-derived retinoic acid (RA) specifies the positional value of the stylopod (humerus, a,d—orange) and then (HH20–22) intrinsic timing specifies the positional values of the zeugopod (forewing, b,d -blue) and later still (HH23–24) the autopod (wrist/digits, c,d—green). AER-derived FGFs are permissive factors that sustain growth of the limb bud and suppress the proximal programme including Meis1/2 expression by inducing Cyp26b1 that degrades retinoic acid (red line). Elimination of retinoic acid from the distal mesenchyme of the wing bud triggers the switch to intrinsic timing and mesenchyme cells express 5'Hoxa/d genes and maintain AER-derived FGFs. Following the phase of proximal specification (a), cell adhesion properties and hence positional values intrinsically change over time (b,c, greater adhesion of cells shown by additional+symbols in lower insets) and this results in a spatial gradient of positional values along the proximo-distal axis as cells are displaced from the distal mesenchyme by an intrinsic programme of proliferation (arrows—lower insets a–c—graded blue/green on skeleton, d).