Immobilization secondary to cell death of muscle precursors with a dual transcriptional signature contributes to the emu wing skeletal pattern

Abstract

Limb reduction has occurred multiple times in tetrapod history. Among ratites, wing reductions range from mild vestigialization to complete loss, with emus (Dromaius novaehollandiae) serving as a model for studying the genetic mechanisms behind limb reduction. Here, we explore the developmental mechanisms underlying wing reduction in emu. Our analyses reveal that immobilization resulting from the absence of distal muscles contributes to skeletal shortening, fusion and left-right intraindividual variation. Expression analysis and single cell-RNA sequencing identify muscle progenitors displaying a dual lateral plate mesodermal and myogenic signature. These cells aggregate at the proximal region of wing buds and undergo cell death. We propose that this cell death, linked to the lack of distal muscle masses, underlines the morphological features and variability in skeletal elements due to reduced mechanical loading. Our results demonstrate that differential mobility during embryonic development may drive morphological diversification in vestigial structures.

Fig. 1. Skeletal and muscular patterns of the forelimb of emu and chicken.

a Three-dimensional renderings from CT images of the distal part of adult emu forelimb skeletons and transverse sections taken from the emu limb at the level of the dashed line. The left and right limbs of two specimens are shown. Arrowheads indicate the fusions. 3-4, digits 3-4; 4*, rudiment of digit 4; d3, metacarpal of digit 3; r, radiale; R, radius; U, ulna. Scale bars, 1 cm. b Ratio of left to right bone length (coefficient of variation: 2.23%, 3.60 %, 1.93%, 4.24%, 0.59%, 3.54% for chicken humerus, emu humerus, chicken ulna, emu ulna, chicken metacarpus, emu metacarpus, respectively). Mean ± SEM. n = 6 (chicken), n = 8 (emu). c Immunostaining with MF20 in forelimbs of chicken and emu embryos. Although the formation of muscles was recognized in autopodial regions of stage 35 (n = 4) and 37 (n = 5) chicken embryos, no autopodial muscles and only a few or no zeugopodial muscles were observed in the forelimbs of emu embryos at the same stages (EMR, EIL, Anc on the dorsal side and the flexor carpi ulnaris (FCU) on the ventral side at stage 35 (n = 3); EMR, Anc on the dorsal side and FCU on the ventral side (n = 1), Anc only (n = 1), no muscles (n = 1)). Brackets indicate autopodial regions. 2-4, digits 2-4; Anc, anconeus; EDC, extensor digitorum communis; EIL, extensor indicis longus; EML, extensor medius longus; EMR, extensor metacarpi radialis. Scale bars, 1 mm. d Rate of distal movements (see Materials and Methods for details). n = 5 (chicken), n = 4 (emu). Mean ± SEM. Welch’s two tailed t-test. **p = 0.0023. e Ratio of left to right bone length of chicken embryos treated with PBS (n = 7) or DMB (n = 13) from E10 to E18 (coefficient of variation: 1.39%, 7.08%, 1.17%, 4.66%, 1.37%, 5.93% for control humerus, immobilized humerus, control ulna, immobilized ulna, control metacarpus, immobilized metacarpus, respectively). f Chicken embryos were treated with PBS or DMB from E6 (stage 28) to E14 (stage 39). Safranin O staining of the wrist joints of chicken and emu embryos and Alcian blue staining of forelimbs of emu embryos. The control panel is flipped horizontally. The distance between the ulna and the distal carpal/metacarpal of digit 3 (yellow lines) was measured in control chickens (n = 4), immobilized chickens (n = 4) and emu embryos (n = 3). Mean ± SEM. Welch’s two tailed t-test. ***p = 0.0003. dc, distal carpal; d3, metacarpal of digit 3; U, ulna. Scale bars, 500 μm.

Fig. 2. Expression of Lbx1, cMet, Nkx2.5 and MyoD in developing emu forelimb buds.

a Expression of Lbx1, cMet and Nkx2.5 (arrowheads) in forelimb fields (FL) of emu embryos (stage 18 Lbx1 (n = 6), cMet (n = 2), Nkx2.5 (n = 4); stage 19 Nkx2.5 (n = 2)). Scale bars, 500 μm. b Expression of Lbx1, cMet, Nkx2.5 and MyoD (arrowheads) in serial sections of emu forelimb buds at mid- (mid. FL) or posterior level (post. FL). A dotted circle indicates Nkx2.5 negative region. Panels of stages 19 and 23 sections were flipped horizontally. (stage 19 Lbx1 (n = 6), cMet (n = 4), Nkx2.5 (n = 6); stage 23 cMet (n = 2), Nkx2.5 (n = 2); stage 25 Lbx1 (n = 3), Nkx2.5 (n = 3), MyoD (n = 2)). dm, dermomyotome; D, dorsal; V, ventral. Scale bars, 50 μm. c, c’, c” Distribution of cMet and Nkx2.5 transcripts in developing emu forelimb buds at stage 23 (n = 4). Arrows indicate the cMet transcripts in muscle precursors derived from the dermomyotome. Arrowheads indicate co-locaized transcripts of cMet and Nkx2.5. d, d’ Distribution of cMet, Nkx2.5 and MyoD transcripts in emu forelimb buds at stage 25 (n = 3). Note that transcripts of cMet and Nkx2.5 (arrowheads) are present in cell clusters, in which MyoD expression is also detected (arrows). White cells are blood cells (asterisks), not stained cells.

Fig. 3. A subpopulation of muscle progenitors exhibits a dual somite-derived myogenic cell/LPM cell signature in emu forelimb buds.

a, d, tSNE plots of body trunk at the forelimb level of stage 20/21 emu embryo data (a) and stage 25 emu forelimb buds (d), respectively. b, e, Dot plots of subcluster marker gene expression of stage 20/21 data (b) and stage 25 data (e), respectively. Dot color represents the average expression level, and dot size represents the percentage of cells expressing marker genes. c, f, Venn diagram showing number of cells expressing Pax3 and/or Hand2 in the muscle cluster of stage 20/21 data (c) and stage 25 data (f), respectively. g, Violin plots showing the expression levels of Pax3, Lbx1, MyoD1, Tnnt3, Hand2, Prrx1, and Tbx5 in Pax3 + /Hand2- cells, and Pax3 + /Hand2+ cells in the muscle cluster of stage 25 emu forelimb data. The two-sided Wilcoxon rank sum-test was used for statistical test. The exact p-values are indicated.

Fig. 4. A subpopulation of muscle progenitors co-expressing Pax3 and Hand2 aggregates at the proximal region of emu forelimb buds.

a–e HCR for Pax3 and Hand2 and immunostaining for MF20 in emu forelimb buds at stages 19 (a, n = 2), 23 (b, n = 3) and 25 (c–e, n = 3). c’ Enlarged images indicated in (c). Arrowheads indicate fragmented nuclei. Note that transcripts of both Pax3 and Hand2 were observed in aggregated cells (arrows). d, Note that MF20 signals (arrows) were not detected in the aggregated cell population (a dotted circle). e HCR for Pax3 and Hand2 in the aggregated cell population at the proximal region of stage 25 emu forelimb buds (arrows). Note that transcripts of Pax3 and Hand2 are co-localized. Arrowheads indicate fragmented nuclei. f–h HCR for Pax3 and Hand2 in chicken forelimb buds at stages 19 (f, n = 3), 23 (g, n = 4) and 25 (h, n = 2). h’ Enlarged images indicated in (h). Unlike emu forelimb buds, no aggregated muscle progenitors are observed at the proximal part of chicken forelimb buds at stages 23 or 25. Scale bars, 100 μm (a–d, c’, h’), 20 μm (e), 200 μm (f–h). Panels of (a), (c, c’) and (g) are flipped horizontally.

Fig. 5. A subpopulation of muscle progenitors aggregated at the proximal region of emu forelimb buds undergoes cell death.

a Violin plots showing the expression levels of Bak1, Casp10 and Apaf1 in Pax3 + /Hand2- cells, or Pax3 + /Hand2+ cells in the muscle cluster of stage 25 emu forelimb data. b–d TUNEL staining (b, n = 3), immunostaining for active caspase-3 (c, n = 2) and immunostaining for 8-oxoguanine (d, n = 2) in the aggregated cell population at the proximal region of stage 25 emu forelimb buds. b’–d’ Enlarged images indicated in (b–d). Panels of (b, b’) and (c, c’) are flipped horizontally. Scale bars, 50 μm. e Schematic model of forelimb development in emu embryos. Migratory muscle precursors (Pax3 + , Lbx1 + , cMet + ) delaminated from the ventral edge of the dermomyotome and begin to migrate into the forelimb mesenchyme (Hand2 + , Prrx1 + , Tbx5 + ). Subsequently, a subpopulation of muscle precursors with a dual somite-derived myogenic cell (Pax3 + , Lbx1 + , cMet + ) /LPM cell (Hand2 + , Prrx1 + , Tbx5 + ) appears and aggregates at the proximal part of forelimb buds. This aggregated cell population undergoes cell death and thereby failing to form majority of muscles. Impaired formation of limb muscles seems to be at least partially responsible for the asymmetric reduction or fusion of distal skeletal elements. Our results and those of others^,– suggest that multiple mechanisms contribute to the unique emu wing morphology. See text for details.