Sonic hedgehog specifies flight feather positional information in avian wings

Abstract

Classical tissue recombination experiments performed in the chick embryo provide evidence that signals operating during early limb development specify the position and identity of feathers. Here, we show that Sonic hedgehog (Shh) signalling in the embryonic chick wing bud specifies positional information required for the formation of adult flight feathers in a defined spatial and temporal sequence that reflects their different identities. We also reveal that Shh signalling is interpreted into specific patterns of Sim1 and Zic transcription factor expression, providing evidence of a putative gene regulatory network operating in flight feather patterning. Our data suggest that flight feather specification involved the co-option of the pre-existing digit patterning mechanism and therefore uncovers an embryonic process that played a fundamental step in the evolution of avian flight.

Keywords: Avian; Chick; Embryo; Flight feather; Positional information; Shh.

Fig. 1.

Shh is required for flight feather bud formation. (A) Polarising region grafts made to the anterior margin of host chick wing buds at HH20 duplicate all tissues across the antero-posterior axis at day 13, including the digits and feather buds (asterisks show duplicated tissues; Roman numerals; primary flight feather buds; Arabic numerals; secondary flight feather buds; black is feather pigmentation). (B) Schematic showing general bird wing feather pattern, including the three types of flight feathers: primaries along the posterior margin of digits 2 and 3; secondaries along the posterior margin of the ulna; and alulars along the posterior margin of digit 1 (Lucas and Stettenheim, 1972). The chicken has 10 primary and 18 secondary flight feathers, a row of primary and secondary major coverts, and many rows of marginal coverts of different identities, including median and minor coverts. (C,D) Developing flight feathers, as shown by Ptch1 expression in all buds in untreated day 13 wings (C, arrow) but not in wings of embryos treated with cyclopamine (cyc) at HH19 (D, arrow; n=4/4). Treatments at HH19 often result in loss of digit 3, but not the radius (r) or ulna (u) (Towers et al., 2011). (E) Schematic depicting the flight feather bud pattern shown in C. (F) Schematic depicting posterior feather bud pattern shown in D. (G,H) Lmx1b expression in the dorsal mesendoderm of untreated (G; n=4/4) and HH19 cyclopamine-treated (H; n=4/4) wing buds at day 9. Scale bars: 2 mm in A,C,D; 1 mm in G,H.

Fig. 2.

Shh is required for flight feather development during late embryogenesis. (A) Schematic showing relationship between flight feathers and the skeleton to which they connect using ligaments. (B-D) Hematoxylin and Eosin staining on transverse sections of day 12 to day 15 forewings showing developing flight feathers in untreated embryos and their connections to the ulna (n=15/15). (E) Asymmetric expression of Shh in flight feather buds on forewings of untreated embryos (n=6/6 wings). (F-H) Absence of flight feathers making ligamentous connections with the ulna in cyclopamine-treated embryos (n=18/23). (I) Asymmetric expression of Shh is not observed in feather buds on the forewings of cyclopamine-treated embryos (n=6/6 wings). Scale bars: 100 μm in B,F; 125 μm in C,G; 150 μm in D,H; 50 μm in E,I.

Fig. 3.

RNA-sequencing reveals a flight feather bud transcriptome. (A-C) Schematics showing regions of day 10 limbs that were used to make RNA and the pairwise contrasts made: HH19 cyclopamine-treated Bovans brown wings versus control Bovans brown wings (A), Bovans brown wings versus Bovans brown legs (B) and Pekin bantam legs versus Bovans brown wings (C). The top ten genes up- and downregulated more than fivefold are shown for each comparison (P<0.005). (D) Cluster of genes downregulated in wings by earlier Shh signalling inhibition, upregulated in wings versus legs, and upregulated in Pekin bantam legs versus Bovans brown legs (P<0.005 and a greater than twofold change in at least one contrast; red, upregulated; blue, downregulated. r, radius; u, ulna; t, tibia; f, fibula; mt, metatarsals.

Fig. 4.

Shh signalling is required for flight feather bud-associated gene expression. (A,F,K,P) Box and whisker plots showing relative expression levels of Sim1 (A), Zic1 (F), Zic3 (K) and Zic4 (P) as normalised log₂ values of RNA sequencing read-count intensities. (B,G,L,Q) Expression of Sim1 (B, n=22/22), Zic1 (G, n=4/4), Zic3 (L, n=4/4) and Zic4 (Q, n=4/4) in flight feather-forming regions of day 10 wings. (C,H,M,R) Downregulation of Sim1 (C, n=12/14), Zic1 (H, n=2/2), Zic3 (M, n=2/2) and Zic4 (R, n=2/2) in forewing regions following Shh signalling inhibition. (D,I,N,S) Undetectable/weak expression of Sim1 (D, n=2/2), Zic1 (I, n=2/2), Zic3 (N, n=4/4) and Zic4 (S, n=2/2) in Bovans brown legs. (E,J,O,T) Upregulation of Sim1 (E, n=2/2), Zic1 (J, n=2/2), Zic3 (O, n=2/2) and Zic4 (T, n=2/2) along the posterior margins of Pekin bantam legs. For box and whisker plots, centre mark is median, whiskers are minimum/maximum. Arrows indicate ectopic gene expression. Scale bars: 1 mm. t, tibia; f, fibula; mt, metatarsals.

Fig. 5.

The duration of Shh signalling is interpreted into the later spatial pattern of Sim1 expression. (A-D) Application of cyclopamine at HH18, HH19, HH21 and HH22 reduces Sim1 expression in digit 1 at day 10 and causes loss of expression in the ulna and digit 2 (A, n=3/3), reduces Sim1 expression in the ulna and the proximal region of digit 2 (B, n=12/14), (C) reduces Sim1 expression in digit 3 (C, n=10/15) and does not affect Sim1 expression (D, n=5/5). (E) HH20 polarising region grafts made to the anterior margin fully duplicate the pattern of Sim1 expression (n=5/5). (F) Smaller HH20 polarising region grafts made to the anterior margin duplicate Sim1 expression in the additional digit 1* (n=2/2). Scale bars: 1 mm.

Fig. 6.

Sim1 and flight feather buds are adjacent to the polarising region lineage. (A) GFP-expressing polarising regions transplanted in place of normal polarising regions at HH20 contribute to posterior soft tissues of day 10 wings (green labelling, n=2/2; r, radius; u, ulna; 1-3, digits 1, 2 and 3). (B,C) Sim1 expression along posterior margin of the ulna and digit 3 (n=2/2; the same limb as shown in A). (D,D′) Transverse section through forewing shown in A and B reveals adjacent expression of GFP and Sim1 in dermis (n=2/2; GFP protein and Sim1 mRNA are detected on same section). Arrowheads indicate the domain of Sim1 expression. (E,G) Transverse sections through a day 11 wing (E, n=2/2, experiment performed as in A) and a day 12 wing (G, n=2/2) showing GFP expression ventral to emerging flight feather (ff; bv, blood vessel; u, ulna; blue shows DAPI staining). (F,H) Hematoxylin and Eosin staining on serial sections to those in E,G show tissue anatomy. (I) HH20 GFP-expressing polarising region grafts made to the anterior margin of a host wing bud duplicate the distal structures of day 10 wings (n=5/5). (J,K) GFP-expressing polarising region grafts duplicate the pattern of Sim1 expression (n=5/5; there is a second line of Sim1 expression, n=3/5; this is the same limb as in I). (L,L′) Sim1 expression is found adjacent to the polarising region lineage (n=5/5). Unlabelled arrowheads indicate the domain of Sim1 expression. Asterisks indicate duplicated skeletal elements. Scale bars: 1 mm in A,B,I,J; 150 μm in E-H; 75 μm in C,D,K,L.

Fig. 7.

Shh signalling is required for flight feather formation in the wings of hatchlings. (A) An example of an untreated chicken wing at hatching (incubation day 21) showing the normal pattern of primary flight feathers and primary major covert feathers (n=10/10). (A′) Enlarged area of the wing shown in A. (B) Schematic of normal flight feather pattern at hatching. (C) Hematoxylin and Eosin staining on a transverse section through the wing in A showing ligaments connecting the flight feathers to the ulna. (C′) Enlarged area of section shown in C. (D) Example of a HH19 cyclopamine-treated wing at hatching showing loss of both primary flight feathers and primary major covert feathers (n=13/16). (D′) Enlarged area of the image shown in D. (E) Schematic of flight feather pattern in the cyclopamine-treated wing at hatching. Malformed flight feathers often form in distal regions, but can only be seen when natal down is fully removed (see Fig. S4 for examples). (F) Hematoxylin and Eosin staining on a transverse section through the wing in D shows that down feathers are still present at the posterior margin of the wing where flight feathers would normally develop. (F′) Enlarged part of section shown in F. Scale bars: 8 mm in A,D; 1 mm in C,F.

Fig. 8.

Shh signalling is required for flight feather formation in the wings of mature birds. (A) Experimental procedure in which an embryo was treated with cyclopamine at HH19 for 10 h and then its right-hand wing was grafted in place of a stage-matched wing bud of an untreated embryo at HH20/21. (B) Example of a hatched chick (p3, postnatal day 3) that underwent the procedure described in A (see Table S2 for details of the other five chickens that hatched). (C) Same chicken as shown in B at p22. Primary flight feathers are absent in the distal regions of the cyclopamine-treated wing. (D,E) Dorsal views of untreated (D) and cyclopamine-treated (E) wings at p66 (same chicken as in B,C) showing alular flight feathers (red asterisks, digit 1), distal primary flight feathers (blue asterisks, digit 2), proximal primary flight feathers (orange asterisks, digit 3), secondary flight feathers (green asterisks, ulna), dorsal major covert feathers (purple asterisks) and dorsal median covert feathers (light-blue asterisks). Proximal primary flight feathers and overlying primary major covert feathers are absent in the cyclopamine-treated wing (E). (F,G) Ventral views of untreated (F) and cyclopamine-treated (G) wings at p66 showing alular flight feathers (red asterisks, digit 1), distal primary flight feathers (blue asterisks, digit 2), proximal primary flight feathers (orange asterisks, digit 3), secondary flight feathers (green asterisks, ulna) and ventral major covert feathers (purple asterisks). Proximal primary flight feathers are absent in the cyclopamine-treated wing (G). Scale bars: 3 cm in C; 5 cm in D-G.

Fig. 9.

Positional information model of flight feather specification. Predicted temporal gradient of Shh from the polarising region between HH18 and HH22 (blue shading, day 3-3.5) is interpreted into a spatiotemporal pattern of Sim1 expression in flight feather-forming regions of the wing at day 9. The order in which the pattern of flight feathers is specified across the antero-posterior axis is the same as the skeletal elements: alulars, digit 1 (red); distal primaries, digit 2 (blue); secondaries, ulna (green); proximal primaries, digit 3 (orange). The first generation of flight feathers form during late embryogenesis (not shown) and the second generation of flight feathers can be seen in mature wings. The experiments in which embryos were systemically treated with cyclopamine were used to define the temporal requirement of Shh signalling for Sim1 expression and flight feather development. In addition, chicks treated at HH18 failed to hatch; thus, we predict that alulars would be specified at this stage based on Sim1 expression.