Transition from natal downs to juvenile feathers: conserved regulatory switches in Neoaves

doi:10.21203/rs.3.rs-3382427/v1

Download PDF

Article

Transition from natal downs to juvenile feathers: conserved regulatory switches in Neoaves

https://doi.org/10.21203/rs.3.rs-3382427/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 16 May, 2024

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

The transition from natal downs for heat conservation to juvenile feathers for simple flight is a remarkable environmental adaptation process in avian evolution. However, the underlying epigenetic mechanism for this primary feather transition is mostly unknown. Here we conducted time-ordered gene co-expression network construction, epigenetic analysis, and functional perturbations in developing feather follicles to elucidate four downy-juvenile feather transition events. We discovered that LEF1 works as a key hub of Wnt signaling to build rachis and converts radial downy to bilateral symmetry. Extracellular matrix reorganization leads to peripheral pulp formation, which mediates epithelial - mesenchymal interactions for branching morphogenesis. ACTA2 compartments dermal papilla stem cells for feather cycling. Novel usage of scale keratins strengthens feather sheath with SOX14 as the epigenetic regulator. We found this primary feather transition largely conserved in chicken (precocious) and zebra finch (altricial) and discussed the possibility that this evolutionary adaptation process started in feathered dinosaurs.

Biological sciences/Evolution/Evolutionary developmental biology

Biological sciences/Computational biology and bioinformatics/Gene regulatory networks

Biological sciences/Developmental biology/Differentiation

Evolutionary innovations have enabled birds to occupy different ecological niches. In particular, feathers show a very high degree of diversity, providing an excellent model for studying how animals adapt to different environments ^{1, 2, 3, 4}. Different developmental stages of a bird exhibit different types of feathers. Most hatchling plumages are either naked or have natal downs, which confer heat conservation for the hatchlings. When juvenile birds are ready to leave the nest, most radially symmetric downs are replaced by bilaterally symmetric juvenile feathers in the same follicle to form basic plumage. After several rounds of molting, the adult feathers that can respond to environmental changes (hormones, seasons, etc.) are eventually formed (Fig. 1a) ^{5, 6, 7, 8}. These remarkable morphological transitions are based on the stem cells and dermal niches from the same feather follicles, and can give plumages different morphologies and functions at different times of a bird’s life for optimized function, sometimes giving birds totally different appearances; this process is known as “organ level metamorphosis” ^{9, 10}.

The natal down to juvenile (rachidial) feather conversion is here referred to as the primary feather transition process, while the formations of diverse adult feathers after several rounds of molting is referred to as the secondary feather transition process. These two processes set up the foundation to generate regional and timing differences in feather cycling, and to achieve optimal functions required at the different stages of adult bird’s life ^8,10. The topological changes of feather follicle architectures in natal and adult feathers have been well-characterized ^{11, 12}. The roles of morphogens in forming rachidial and barb branch in adult feathers have also been elucidated ^{13, 14, 15, 16, 17, 18, 19}. The mechanism of secondary transition of feather follicles have been explored in several birds ^{20, 21, 22}. Yet, the evolution and underlying mechanism of the primary transition, which occurs only once in life in most birds, has rarely been studied.

Based on previous studies and our findings, rachis, strong feather sheath, regeneration ability, and feather vane are major distinctions of juvenile feathers from natal downs (Fig. 1a) ^{5, 11, 12, 23}. The rachis serves as the backbone of the feather, providing support and structural integrity ^{1, 16, 17, 19, 24}. Wnt signaling pathways are the major regulators of both hooklet and rachis development ^{16, 17, 19}. Feather sheath is a structure that maintains feather’s cylindrical shape until it starts to disintegrate near the tip, allowing the mature part of the feather to unfurl ¹. In adult chickens, feathers need to undergo molting to maintain their normal functions and the molting process is known to be controlled by the dermal papilla and regulated by Wnt inhibitors ^{15, 25, 26}. Feather hooklets are found on the barbules of bird feathers, which interlock onto the proximal barbules of the immediately adjacent barb and transform the barb branches into a planar vane ^{19, 27}. Here we will focus on the epigenetic controls of the primary feather transition.

To identify signaling pathways and transcription factor (TF) genes involved in the primary feather transition, we obtained embryonic and juvenile time series transcriptomes of chicken posterior dorsal skins and then used two approaches to analyze the data. First, we applied the time-ordered gene co-expression network (TO-GCN), a transcription factor prediction method, ²⁸ to infer the TO-GCN of the TF genes expressed in the transcriptomes. Second, in birds, several β-keratin gene subfamilies have been classified by sequence similarities and tissue specific expression, but little has been done on the regulation of their expression. Whether diversified β-keratin genes are involved in the primary feather transition and how they are regulated are interesting questions. Here, we annotated β-keratin genes in the newly released chicken genome (GRCg6a) to analyze the transcriptomes and applied ATAC-seq to elucidate the epigenetic regulation of the specific β-keratin gene subfamilies during the primary feather transition ^{29, 30}. We then functionally validated the predicted molecules in chicken flight feathers.

To test the molecular regulatory conservation in birds, we compared the primary feather transitions in chicken (white leghorn) and zebra finch (Taeniopygia guttata). Chicken, a precocial bird, belongs to Galliformes, which is a basal lineage of Neoaves. Zebra finch belongs to Passeriformes, which is the most derived Neoaves and the largest avian order; all species in this order are altricial birds ^{31, 32, 33}. We chose chicken and zebra finch as two model birds for our study because they represent two distant species in Neoaves, so that a trait conserved in both species is likely conserved in most or all Neoaves.

The primary feather transition in chicken and zebra finch: from natal downs to juvenile feathers

To study the morphological changes during the primary feather transition in the posterior dorsal skin, we compared the histology of chicken feather follicles between embryos (E14) and hatchlings (posthatch day 3 to day 7, i.e., D3 to D7), and then compared the structures with that of the posterior dorsal feather follicles in D7 zebra finches. In this region, juvenile feathers grew out from the skin around D6 in both species (Fig. 1a and Supplementary Figs. 1 and 2), suggesting that the time point of the transition initiation has been conserved in both species.

To understand the morphogenesis of feather follicles during the primary transition, both cross and longitudinal paraffin sections with hematoxylin and eosin (H&E) staining were applied to the whole skins or feather follicles of both species (Fig. 1b-m); muscle bundles were used as the anchor for the comparisons among different stages and species. In cross sections, we found that the structures of the feather germ were similar while the sizes increased slowly from E14 to D4, suggesting a slow growing phase (Fig. 1b-d, 1b’-d’, and g). Interestingly, muscle development was relatively obvious in this period, and the shapes of muscle bundles between follicles changed from square in embryonic skins to diamond in juvenile skins (Fig. 1b’-d’). The juvenile feather follicles started to enlarge quickly and formed rachis after D5, showing a fast growth phase (Fig. 1d’-e’, 1k-l, and Supplementary Fig. 1). In D7, the juvenile feathers in chicken and zebra finch showed similar maturities: thickened feather sheathes, rachis and barb ridges were clearly visible (Fig. 1e’-f’ and l-m). The clear bard ridges suggest the formation of organized barbs and barbules (Fig. 1g).

To confirm the above findings, longitudinal sections were conducted in the central (spinal) feather follicles and the results supported the conclusion that the maturity of D7 feather follicles is similar between chicken and zebra finch: curved collars, enlarged pulps, and biconcave dermal papilla are clearly visible (Fig. 1h-m). The peripheral pulp regions show condensed cells, suggesting the formation of peripheral pulps (Fig. 1h). Based on these data and the literature ^{1, 3, 11, 12, 14, 15, 16, 25}, we propose five major evolutionary morphogenesis events during the primary feather transition in Neoaves: biconcave dermal papilla formation, peripheral pulp formation, rachis formation, vane formation, and feather sheath thickening.

Coordination of multiple signaling pathways during the primary feather transition

As we wanted to know the overall skin tissue changes during the primary feather transition, we sampled the whole posterior dorsal skins from day 3 to 7 hatchlings in which complex gene interactions from different cell populations and different timings are involved. We applied the time-ordered gene co-expression network (TO-GCN) method ²⁸, which was designed to decipher the molecular regulations from time (or developmental) course transcriptomes of complex tissues, to analyze the transcriptomes from embryonic and juvenile skins. We focused mainly on transcription factor (TF) genes, because they are the major players of gene regulation.

RNA-seq analysis of the transcriptomes from embryonic and juvenile samples was conducted to assess the quality of sequencing libraries and generate normalized read counts as the input for the TO-GCN analysis (Supplementary Fig. 3 and Table 1). A TO-GCN of 11 levels was constructed in both feather types based on the expression profiles of the TF genes. The lower (earlier) levels included the TF genes expressed at the earlier feather developmental stages, while the higher (later) levels included those expressed at the later developmental stages (Fig. 2a and b; Supplementary Table 2). The expression profiles of TF genes in each of the 11 levels were similar between the two feather types (as shown in the bar chart, Fig. 2a and b; Supplementary Table 2).

In early embryonic skin development (levels 1 to 5), Rho GTPases and cell cycle related pathways are enriched (Supplementary Table 3). Rho GTPases are best known for their roles in regulating cytoskeletal rearrangements, cell motility, cell polarity, axon guidance, vesicle trafficking and the cell cycle ^{34, 35}, suggesting that Rho GTPase-mediated cell proliferation and polarization are the major event at this stage. Unexpectedly, polycomb repressive complex 2 (PRC2) related pathways were also enriched. PRC2 is a multiunit epigenetic protein complex that silences gene expression by catalyzing trimethylation of histone H3 at lysine 27 (H3K27me3). How methylation is involved in natal down growth regulation is an interesting question. In late embryonic skin samples (levels 6 to 11), cornified envelope and keratin formation pathways were enriched (Supplementary Table 3).

In the juvenile TO-GCN, there are two distinct level groups each with a large gene member, suggesting two morphogenesis groups in post-hatched feathers. The first group is at levels 3 (110 TF genes) and 4 (102 TF genes) in which the major enriched pathways are extracellular matrix (ECM) organization and metabolism (Fig. 2b and e, Supplementary Tables 3 and 4). The second group is at levels 9 and 10 in which 72 and 138 TF genes were assigned, respectively. Many molecular pathways were enriched in this period but WNT/β-catenin related pathways were dominant (Fig. 2b and e, Supplementary Tables 3 and 4). ECM is vital for determining and controlling the most fundamental behaviors and characteristics of cells such as proliferation, adhesion, migration, polarity, differentiation, and apoptosis ^{36, 37}. The WNT/β-catenin related pathway is known to control rachis and hooklet formation ^{16, 17, 19}. Both of them are highly associated with the specific structures of juvenile feathers and are therefore the focus in this study.

In addition to ECM and the WNT/β-catenin related pathways, many other known molecular mechanisms were also enriched. In both early embryonic and juvenile skin development (levels 1 to 5), muscle formation genes were enriched, consistent with the previous finding that muscle development is important for feather positioning (Supplementary Tables 3 and 4) ³⁸. In late juvenile skin development (levels 6 to 11), planar cell polarity (PCP) signaling pathway genes were enriched. Coupling of apical-basal polarity and PCP was identified to interpret the WNT signaling gradient, which controls the bilateral symmetric feather formation (Supplementary Table 4) ¹⁸.

Next, we wanted to know specific TF genes that are involved in the primary feather transition. In this study, we used the following criterion to compare the embryonic and juvenile TO-GCNs: for a TF, if d (= embryonic level – juvenile level) ≥ 3, we say it is an embryo specific TF; if d ≤-3, we say it is a juvenile specific TF; if d ≤ |2|, we say it is a common TF (Fig. 2c). Among the 638 expressed TF genes, 443 showed d ≤ |2|, suggesting that most TF genes are commonly used in both feather types (Fig. 2c; Supplementary Table 2, 3, and 4). The embryonic or juvenile specific TF genes at levels 8 to 10 in both feather types were selected as the embryonic or juvenile keratinization regulators, respectively. We matched these TF genes with those in two previous studies in which keratin regulators were well characterized (Supplementary Table 5) ^{39, 40}. Interestingly, compared to our TF genes identified from embryonic TO-GCN, our TF genes identified from juvenile TO-GCN overlapped more with TF genes controlling scale or claw keratins (embryonic TFs: 0/9 = 0% vs juvenile TFs: 3/8 = 37.5%, Supplementary Table 5), suggesting that scale/claw keratins and their regulators may contribute more to juvenile feathers than to embryonic feathers.

Wnt signaling is the major regulator of rachis formation and LEF1 is a key molecular hub converting radial downy to bilaterally symmetric juvenile feathers

The molecular gradient in feather follicle from anterior to posterior end is crucial for regulating the angles of barb ridges for rachis formation ¹⁴. During rachis formation, the anterior to posterior Wnt3a gradient is known to convert radial to bilateral feather symmetry via convergence of barb ridges toward the rachis region ¹⁶. Moreover, multiple Wnt genes showed gradient expressions during flight feather regeneration ²⁶, suggesting that the Wnt-based regulation could be redundant or regional specificity.

In the RNA-seq analysis, the expression levels of Wnt2b, Wnt5a, Wnt5b, Wnt7a, Wnt9a, Wnt9b, and Wnt16 were increased with the juvenile dorsal feather development, corresponding to the time point of the rachis formation (Fig. 1b’-e’ and Fig. 3a). Wnt3a showed constant expression and might not be a rachis regulator of the juvenile dorsal feathers (Fig. 3a). In our TO-GCN analysis, the lymphoid enhancer binding factor 1 (LEF1) was identified as a level 10 TF in juvenile feather development and is known to be a key mediator of Wnt signaling in diverse biological processes (Fig. 3a and Supplementary Table 2). The section in situ hybridization (SISH) of LEF1 and CTNNB1 in D7 chicken posterior dorsal skins revealed that LEF1 was expressed in an anterior to posterior gradient in the feather epidermis but CTNNB1 was expressed evenly (Fig. 3b), suggesting that LEF1 is the key factor responding to Wnt signaling. In zebra finch, although LEF1 was also expressed in feather epidermis, the feather pigments prevented us from visualizing the gradient (Fig. 3b, lower panel). Interestingly, both CTNNB1 and LEF1 showed higher expression in the chicken central feather follicles than in the flanking follicles, suggesting that the embryonic feather lateral propagation factor β-catenin could also be the activator for the secondary feather lateral propagation and LEF1 is also involved (Fig. 3b, upper panel).

To validate the effect of LEF1 on rachis formation, a dominant negative form of LEF1 (dnLEF1) was cloned into RCAS virus and injected into cavities of the plucked flight feathers. Previous studies showed that in flight feathers, the misexpression of Wnt inhibitor DKK1 could slightly disturb the rachis formation, and the overexpression of Wnt3a could disrupt the rachis formation and also cause abnormal barbs ¹⁶. Here, all the flight feathers injected with RCAS-dnLEF1 showed defects in rachis (Supplementary Fig. 4a). Half of the feathers lost part of the rachis while the others lost the entire rachis without influencing the surrounding barbs (Fig. 3c, Supplementary Fig. 4b), suggesting that the function of LEF1 is specific in feather follicles. SHH is a key morphogen for barb formation via their expression in bard ridges and absence in rachis forming regions ^{14, 16, 41}. In rachidial feather, barb ridges insert into the rachidial ridge with the helical insertion angle (Fig. 3d, indicated by θ). However, in both natal down and RCAS-dnLEF1 misexpressed flight feather, all the barb ridges were formed in parallel (θ = 0, Fig. 3d), suggesting that barb ridges were not able to insert into rachidial ridge.

ECM re-organization generates peripheral pulp for feather branching morphogenesis

ECM organization was specifically enriched in three levels of the juvenile feather TO-GCN (L2 to L4, Fig. 2e). Components of ECM link together to form a structurally stable composite, contributing to the mechanical properties of tissues. ECM is also a reservoir of growth factors and bioactive molecules ^{36, 37}. In chicken embryonic feathers, ECM regulates mesenchymal mechanics which can spontaneously break skin symmetry ⁴². In chicken adult feathers, ECM reorganization enables peripheral pulp formation ²⁵. Therefore, we asked what is the function of ECM in primary feather transition and which molecules control and maintain the ECM reorganization process?

In the juvenile feather TO-GCN, levels 2 to 4 basically correspond to the D3 to D5 stages in feather development (Fig. 2b). A novel morphogenesis in this stage is the generation of peripheral pulp, which has about five layers of mesenchymal cells closely attached to the feather filament and basement membrane. The peripheral pulp expands the epithelial-mesenchymal interactive interface for barb patterning ^{25, 43}. The gain of ECM mediated pulp differentiation is therefore our hypothesis for the primary feather transition. Tenascin C (TNC) is frequently used as ECM and differentiated pulp markers ^{11, 12, 15, 25, 26}. The immunochemistry (IHC) signals of TNC reveal that the peripheral pulp was gradually differentiated from embryonic pulp along with the growth of embryonic to juvenile feathers in both chicken and zebra finch (Fig. 4a’-e’ and 4a”-e”). We further picked up two ECM reorganization related transcription factors, TWIST2 and ZEB2, from levels 2 to 4 of TO-GCN to examine their expressions. The SISH of TWIST2 showed the initial expression in the collar of embryonic feathers, the induction in whole pulp in early juvenile feathers, and the restricted expression in apical dermal papilla and peripheral pulp in late juvenile feathers in both chicken and zebra finch (Fig. a1-e1), suggesting its role in peripheral pulp formation. Interestingly, the IHC of ZEB2 showed an almost opposite pattern in that the signals were enriched in the central pup in both species (Fig. a2-e2), suggesting that it may maintain the mature tissues.

ACTA2 shapes adult dermal papilla to compartment dermal papilla stem cells for cyclic renewal

ACTA2 (encoding α-SMA) is a major feather dermal papilla marker and our TO-GCN analysis showed that ACTA2 co-expressed with levels 2 to 4 TF genes (Supplementary Table 6) ^{15, 25}. We therefore paid additional attention to the role of α-SMA in primary feather transition. Dermal papilla of downy feather is long and slender, while dermal papilla in the juvenile and adult feathers are biconvex-shaped (Fig. 4 left side schemes) ^{11, 12, 15, 23, 25}. We found that, in the IHC on skin longitudinal sections, the expression of α-SMA in dermal papilla increased with the juvenile follicle development and eventually filled up the biconvex-shaped dermal papilla in D7 juvenile feathers in both chicken and zebra finch (Fig. 4a-e). Most interesting, the study of feather cycling showed that in resting phase, dermal papilla stem cells are located in the apical part of the bi-concaved dermal papilla²¹. In the growth phase, the activation of apical part dermal papilla generates pulp progenitors, giving rise to both central pulp for nutrition purpose and peripheral pulp for continuous epithelial - mesenchymal interactions required for feather branching morphogenesis.

We then tested the importance of ACTA2 in dermal papilla formation. Since the protein sequence of ACTA2 is identical between human and chicken, we cloned a published human dominant negative ACTA2 form into RCAS virus (RCAS-dnACTA2) and injected it into the cavities of the plucked flight feather in chicken ⁴⁴. The knock-down virus caused such severe effects that most (7/8) of the injected follicles could only regenerate tiny or no visible feathers two weeks after injection (Supplementary Fig. 4d). All the eight injected follicles could not be regenerated after the second plucking. Histologically, the dermal papilla and part of the collar structures were also disrupted (Fig. 4f and f’). To further understand the molecular changes in those abnormal follicles, immunostaining of several known ECM factors as well as ACTA2 were applied to feather sections (Fig. 4g-j’) ²⁵. Dermal papilla markers α-SMA and vimentin (VIM) were enriched in whole dermal papilla in the control feather follicles but attenuated and dispersed in the virus injected follicles (Fig. 4g and g’, i and i’). Unexpectedly, TNC and Neural Cell Adhesion Molecule (NCAM), which were found to be located in papilla ectoderm in the control feather follicles, were attenuated and dispersed in the virus injected follicles (Fig. 4h and h’, j and j’), suggesting the simultaneous disruption of dermal papilla and pulp structures. It could be that the dermal papilla of juvenile feathers contains myofibroblast cells and knock-down of α-SMA may then affect the integrity of these cells. These results suggest that α-SMA not only is a structural protein but also establishes an environment for the primary feather transition in Neoaves.

Many scale keratins are specifically upregulated in juvenile feather sheath by SOX14

β-keratin genes are mainly classified into feather, scale, claw keratins, and keratinocytes ^{45, 46, 47}. A previous study revealed that chicken dorsal natal downs mainly express feather-β-keratin genes on Chr1, Chr10, Chr25 and also some members on Chr27 during keratinization, whereas feather-β-keratin genes on Chr2 and Chr6 are exclusively enriched in adult wing feathers ⁴⁵. However, subsequent studies revealed that β-keratin genes are differentially regulated in different skin regions ^{39, 40, 47}. Since our TO-GCN analysis along with the previous studies suggested the importance of scale keratins in the juvenile feather formation ^{39, 40}, we employed several modifications in the transcriptomic analysis to decipher the β-keratin gene regulation: (1) Our embryonic and juvenile feather tissues were only from chicken posterior dorsal skins. (2) In addition to β-keratin genes, we analyzed the whole epidermal differentiation complex (EDC), which was known to participate in feather functional evolution ^{48, 49, 50, 51}. (3) We used the newly published chicken genome (GRCg6a), which has a much-improved assembly in micro-chromosomes, where EDC gene clusters are located. (4) We manually annotated the EDC genes, especially the β-keratin genes, which are still poorly annotated in GRCg6a. (5) We conducted an ATAC-seq analysis of embryonic and adult feather tissues to look for epigenetic regulators during the primary feather transition.

We analyzed the co-expression profile of EDC genes in both feather types during their development. Our data were partially inconsistent with the previous finding, showing that, regardless of chromosomal locations, most of the β-keratin genes were expressed during both embryonic and juvenile feather keratinization (Clusters 7 and 11, Fig. 5a). Genes in Cluster 11were highly expressed at the keratinizing stage in both feather types, which include most feather β-keratin genes located on Chr1, 2, 7, 10, 25, and 27 (Fig. 5a; Supplementary Table 7). Genes in Cluster 7 are highly expressed at both the keratinizing stage and a stage ahead of it in both feather types, which include other β-keratin genes, such as scale, claw, feather-like β-keratin, and feather-β-keratin genes on Chr10 (Fig. 5a; Supplementary Table 7). In addition to β-keratin genes, most other EDC genes also were in cluster 7 or 11 (Fig. 5a; Supplementary Table 7).

Only small co-expression clusters and individual genes showed distinct profiles between the two feather types. Two β-keratin genes in Clusters 8 and 9 (GG6AChr25Ktn11 and GG6AChr25FK5) were highly expressed in keratinizing juvenile feathers but not in natal downs (Fig. 5a; Supplementary Table 7). In Cluster 7, CRNN was highly expressed in embryonic but not in juvenile feathers (Fig. 5a; Supplementary Table 7). Interestingly, when we analyzed the β-keratin genes based on their subfamilies (β-keratin related proteins, feather keratins, feather keratin like proteins, scale keratins, claw keratins, and keratinocytes), β-keratin gene expression differences between the two feather types could be found. Chr25 feather and feather-like keratin genes expressed higher in natal downs than in juvenile feathers, while Chr25 scale keratin genes expressed higher in juvenile feathers (Fig. 5a and b, Supplementary Fig. 6, Supplementary Table 7), suggesting that: (1) the ratio differences of Chr25 β-keratin gene expression in different chromosomes may contribute to feather type differences, and (2) some scale keratin genes could be important specifically for juvenile feather formation. To understand the expression differences of the scale keratin genes in the two feather types, we conducted SISH with GG6AChr25Scale2 and GG6AChr25Scale10 in cross sections of embryonic and juvenile feather follicles (Fig. 5c-f and Supplementary Table 7). Although the two keratin genes were faintly expressed in the barbs of both feather types, they were highly enriched in the feather sheath in juvenile feathers but not in natal downs.

What TFs control the scale keratins and whether this regulation is conserved in Neoaves were the next questions. All the scale keratins are on chromosome 25 for both chicken and zebra finch. We first compared the synteny of chromosome 25 between the two species and found that they are basically conserved along the whole chromosome. Next, we conducted ATAC-seq to compare the differential genomic accessible regions between embryonic feather filaments and the epidermis of adult flight feathers. Figure 5g shows the ATAC signals of the two tissues surrounding scale keratin gene cluster. The differential accessible regions overlapped with the conserved chromosome 25 sequences were extracted for footprint analysis to detect the enriched binding motifs and TFs. The results revealed that TFs SOX14, ESRRB, ESRRG, PRDM4, SREBF2, and SMAD5 were enriched in juvenile feathers; the corresponding binding motifs are shown in Fig. 5g’. Related studies are limited but SOX14 is essential for the initiation of neuronal differentiation in the chick spinal cord ⁵².

Since SOX14 is the most significant TF and more highly expressed in juvenile feather than in embryonic feather transcriptomes (Fig. 5g and Supplementary Table 1), we used SISH to detect its expression in the natal and juvenile feather follicles of chickens and zebra finches. The results showed that, the same as the scale keratins, SOX14 was detected in barbs of both feather types in both species, but the expression was only enriched in the feather sheath of juvenile follicles in both species (Fig. 5h-k). The overlapped expression of scale keratins and SOX14 suggests that SOX14 is an upstream transcription factor of scale keratins. To functionally validate the prediction, we overexpressed SOX14 in embryonic feather using the RCAS-virus system and 1/3 of the embryos (N = 12) generated abnormal feather follicles with either round or thickened and shortened phenotypes (Fig. 5l). Since the skin regions included both abnormal and normal feather follicles, SISH was applied to detect the expression of SOX14 and scale keratins in the same skin sections. Follicles with higher SOX14 expressions showed abnormal phenotypes and enhanced expressions of GG6AChr25Scale2 and GG6AChr25Scale10 (Fig. 5l), suggesting that SOX14 is the regulator of the scale keratins in feather sheath.

Feather diversity appears in different body regions and at different developmental stages of a bird ^{5, 23}. Previous studies revealed regulatory differences among different body feathers or among different parts of a chicken feather ^{3, 47, 53, 54, 55}. Feather transition represents a novel type of timing control evolved for better adaptation, depending on different needs at different life times, but the underlying molecular controls have not been well characterized. Feather transition is at least achieved by five tissue reorganizations and our study revealed four of them: 1. Wnt is the common signaling for rachis formation and LEF1 is the downstream key hub for diverse Wnt proteins. 2. ECM reorganization is essential for peripheral pulp formation for further branching morphogenesis. 3. ACTA2 is a key factor to compartment dermal papilla formation for feather cycling. 4. Scale keratin is recruited from scale to strengthen the sheath of juvenile feathers and SOX14 appears to be a major activator for this co-option process. Feather vane formation achieved by hooklets is the primary feather transition property that was not included in this study. Although our transcriptomes included the initial time points for hooklet development (D7), we did not identify regulators of the known morphogen WNT2B ¹⁹. Since hooklets only contribute to a feather fraction, the finer analysis may be worth pursuing in the future. The evolutionary morphogenesis events and the underlying molecular regulators during the primary feather transition are summarized in Fig. 6. All the chicken primary feather transition factors showed the same expression patterns in zebra finch feather follicles, suggesting that the primary feather transition has been conserved in Neoaves.

Multiple Wnt genes seem to participate in the primary feather transition. Whether it is simply functional redundancy or regional specificity is an interesting question. Hox genes represent the best example that elicit distinct developmental programs along the head-to-tail axis of animals and are the upstream signals of many Wnt proteins ^{56, 57, 58, 59, 60, 61}. Whether distinct HOXs activate different WNTs and generate regional specific feather rachis is another interesting question. In feather regeneration, our previous study has uncovered a cell movement from apical dermal papilla to peripheral pulp along the collar bulge ²⁵. This remarkable feature allows prolonged interactions between the dermal niche and epidermal stem cells, providing a tunable interface that is essential for barb patterning ²⁵. The path of the cell movement corresponds to the position of ECM molecule expression and so ECM reorganization could be a key feather regeneration factor. ACTA2 is a dermal papilla marker but is not known to regulate ECM molecules like TNC and NCAM. While the knock-down of ACTA2 disrupted their expressions, we proposed that ACTA2 could be used to establish a competent environment for molecular interaction during peripheral pulp and apical dermal papilla formation. Interestingly, we identified TWIST2, whose expression during feather transition corresponds to the position of apical dermal papilla and peripheral pulp, suggesting its possible role in feather regeneration. TWIST2 is known to regulate ECM in some cases ^{62, 63, 64}.

Previous studies have indicated that SOX genes are important for scale keratin regulation and overexpression of SOX18 enhanced the expression of a scale keratin in embryonic feather ^{39, 40, 65}. However, different SOX genes were identified as β-keratin regulators in different skin appendages (SOX10 in scale, SOX18 in embryonic feather sheath, and SOX14 in juvenile feather sheath in this study) ^{39, 40, 65}. It implies that SOX genes perform regional or temporal specific regulations. Moreover, based on the differential SOX14 regulation and ATAC-seq analysis, we hypothesize that the embryonic feather sheath can be suppressed epigenetically (Fig. 5). The primary feather transition activates the juvenile feather sheath cells to recruit SOX14 to trigger the expression of β-keratins to enhance keratinization (Fig. 6). Notably, a juvenile bird shows various feather types ²³. Although we focus on the major one, i.e., the contour feathers with vane structure, whether other feather types, like bristles and filoplumes, are achieved by the same molecular mechanisms is another interesting question.

Precocial chickens are covered with natal downs, while altricial zebra finches have only limited natal downs in hatchlings ^{31, 32, 33} (Supplementary Figs. 1 and 2). Despite these differences between chicken and zebra finch hatchlings, their juvenile feather developmental profiles are quite similar ³¹ (Fig. 1). After several rounds of molting, diverse feathers are derived from the juvenile feather follicles in the secondary transition process, as seen in the sail-feathers of the mandarin duck (Aix galericulata) or colorful contour feather of the pheasants ^{8, 20, 66, 67}. The conserved primary feather transition suggests its functional importance for survival while the following diverse feather transitions are mainly for mating choices. Interestingly, although egg incubation in chickens is around 21 days while that in zebra finches is around only 14 days, their juvenile feathers are visible at similar post-hatch stages (Fig. 1e’ and f’), suggesting that the primary feather transition is sensitive to the post hatch stimulus, but are regardless of nesting habitats and parental cares.

The conserved primary feather transition among Neoaves implies conserved regulatory sequences in their genomes. How to identify the regulatory regions and how the regulations are controlled are important questions. Although many chicken mutants have been created, the phenotype without the primary feather transition has never been found ⁶⁸, suggesting that the loss of feather transition is highly detrimental. The avian conserved non-exonic elements (CNEEs) database could be used to screen the regulatory sequences and the combination of epigenetic marks can be used to understand the epigenetic changes of the regulatory sequences ^{69, 70}. On the other hand, when the mechanism evolved is a fascinating question. The evolution of feather structures in Dinosauria was proposed in the previous studies in which the basal family Compsognathidae (such as Sinosaurpteryx) was covered with unbranched monofilaments (down feather without barbules) while the derived family Caudipteridae (such as Caudipteryx) showed feather types similar to current birds ^{71, 72, 73, 74}. Whether Dinosauria encompassed feather transitions during their development is not clear, but both the natal and juvenile feather forms had been discovered, suggesting that the primary feather transition in current birds could either be inherited or modified from their dinosaur ancestry.

The success of the Aves to venture into new eco-spaces relied much on the flexibility to convert their feather phenotypes during the lifetime of an individual, which is based on the module-based complexity formation ⁹. These modules are feather follicles consisting of stem cells and their niches. In this study, we tried to decipher the strategies used to evolve these regulatory switches in these modules. We found no new molecules needed to be created. Instead, the transcription regulation redeployed existing molecular module and cellular components, which include the topological positioning of Wnt signaling center, the tissue remodeling to extend the interface for epidermal-dermal interactions, the re-shaping dermal papilla to create dermal papilla stem cell compartment and the co-optional use of scale keratins in feather sheath. Conversions of integument phenotypes from the young to the adult are also seen in some mammals and other animals. These transitions may have been driven by environmental adaptation ⁷⁵. Some strategies we learn here may help understand this fundamental mechanism in Evo-Devo.

Ethics statement

All the animals used in this study were processed following the approved protocols of the Institutional Animal Care and Use Committees of the University of Southern California (USC; Los Angeles, CA, USA) and Institutional Animal Care and Use Committees of National Chung Hsing University (NCHU, Taichung, Taiwan).

Sample collection

Fertilized pathogen-free White Leghorn chicken eggs were purchased from Charles River Laboratories. Eggs were incubated at100℉ and 65% humidity until embryos reached the desired developmental stages. Chicken embryos at14 days (E14), posthatched 3 days (D3), posthatched 4 days (D4), posthatched 5 days (D5), posthatched 6 days (D6), and posthatched 7 days (D7) were used. Chicken hatchlings are mostly covered by natal downs. To visualize the juvenile feather growth, we plucked posterior dorsal natal downs of the hatched chicken and observed the growth patterns of juvenile feathers (Fig. 1). Although a previous mammalian study found that hair plucking can stimulate early regeneration of underneath hairs ⁷⁶, this phenomenon is not observed in the primary feather transition because the plucked and un-plucked region showed similar juvenile feather growth in both species (Fig. 1 and supplementary Figs. 1 and 2).The zebra finch eggs were hatched and the hatchlings were raised in the bird breeding room at NCHU. E12 and D7 Zebra finches were used.

Quantitative PCR

To quantify the candidate gene expressions, the cDNAs were synthesized from the total RNA by QuaniTect Reverse Transcription kit (Qiagen). Each cDNA sample containing SYBR green (KAPA SYBR FAST qPCR kit) was run on LightCycler 480 (Roche) under the appropriate conditions. Quantification of the TATA box binding protein (TBP) RNA was used to normalize target gene expression levels. SHH forward primer: CTGGTGAAGGACCTGAGCCCT; reverse primer: GCCCAACTGTGCTCCTCGAT, TBP forward primer: CACAGCAAGCGACACAGGGA; reverse primer: AGGTGTGGTTCCCGGCAAAG.

Collection and construction of the transcriptomic libraries

To choose the appropriate time points representing juvenile feather development, we screened several feather morphogens and found that SHH is a good marker gene because its expression corresponded to the phenotypic changes of the juvenile feather follicles (Fig. 1 and Supplementary Fig. 3a). Based on the expression profile of SHH, we selected posthatch day 3 to 7 posterior dorsal skins of the chickens to represent the primary feather transition. All the chicken transcriptomic libraries from the posterior dorsal area used in this study and their sources are listed below:

Library ID	Description	Source	Replicates
E7E	Embryonic day 7 epidermis	Made in this study	2
E7D	Embryonic day 7 dermis	Made in this study	2
E9E	Embryonic day 9 epidermis	Made in this study	2
E9D	Embryonic day 9 dermis	Made in this study	2
E12F	Embryonic day 12 feather filaments	From ^{40, 49}	2
E14F	Embryonic day 14 feather filaments	Made in this study	2
E16F	Embryonic day 16 feather filaments	From ^{40, 49}	2
D3	Posthatch day 3 skin	Made in this study	3
D4	Posthatch day 4 skin	Made in this study	3
D5	Posthatch day 5 skin	Made in this study	3
D6	Posthatch day 6 skin	Made in this study	3
D7	Posthatch day 7 skin	Made in this study	3

For posthatch samples we collected the whole skin because the feather follicles cannot be quickly dissected to ensure the RNA quality at these stages. The dissected tissues were preserved in RNALater solution (Invitrogen), incubated at 4°C overnight, and then transferred to -20°C until processing for isolation of total RNA. Total RNAs were isolated using the RNeasy Fibrous Tissue Mini Kit (Qiagen). The 30 min DNaseI treatment was carried out at room temperature. The RNA quantities and qualities of each individual were analyzed by NanoDrop 1000 (Thermo Scientific) and BioAnalyzer II (Agilent Technologies). The RNA samples from the same litter that passed the quality control (RNA integrity number [RIN] > 8.0, A260/280 and A260/230 > 1.9) were used for sequencing library constructions. Paired-end 2 × 101 nt sequencing for juvenile samples and paired-end 2 × 150 nt sequencing for embryonic samples were conducted by the High Throughput Genomics Core Facility, Biodiversity Research Center, Academia Sinica, Taiwan, and Novogen, United States, respectively.

Manual annotation of keratin genes

The annotations of the latest chicken genome assembly (GG6a) were downloaded from Ensembl ⁷⁷ and RefSeq ⁷⁸. The Hidden Markov Models (HMMs) that represent alpha-keratins and beta-keratins were downloaded from Pfam 33.1⁷⁹. The putative alpha-keratins and beta-keratins in Ensembl and RefSeq annotations was predicted using the hmmsearch function in HMMER 3.1 (https://hmmer.org)⁸⁰. For alpha-keratins, we further eliminated the genes that were not located in the typical type I and type II alpha-keratin gene clusters. Finally, the predicted alpha-keratin genes and beta-keratin genes in Ensembl and RefSeq annotations were further compared to the annotation in Ng, et al. (2014) ⁴⁶ and the final annotation of alpha-keratins and beta-keratins was obtained by manual judgement for the most appropriate gene models.

RNA-seq analysis

For each transcriptome, low-quality reads were trimmed and adapters were removed using Trimmomatic ⁸¹. The processed paired-end reads were mapped to the reference genomes of chicken (GRCg7w for TO-GCN and manually-annotated GRCg6a for keratin analysis) using Hisat2 with default settings ⁸². Two genomes were used because the manually annotated GRCg6a was made before GRCg7w was released. From the aligned reads, the transcripts were assembled using StringTie with annotation-based settings ⁸³. The gene expression level was calculated in terms of Transcripts Per Kilobase Million (TPM). For TPM calculations, uniquely mapped reads were first assigned to the genes. Multiple-hit reads were then redistributed to genes based on their relative abundances of uniquely mapped reads. A gene is considered expressed if its TPM is ≥ 1 in at least one of the transcriptomes. To compare expression profiles of the selected genes across the conditions, we applied the upper quartile normalization procedure ⁸⁴. Differential expression analysis between different samples was calculated using the R package DESeq2 ⁸⁵.

Construction of TO-GCNs.

Our comparative transcriptomics method ²⁸ was designed to analyze time-course or developmental-stage transcriptomes that have five or more different conditions between the two feather types: E7 epidermis, E7 dermis, E9 epidermis, E9 dermis, and E14 feather filaments in posterior dorsal regions for embryonic samples; D3, D4, D5, D6, and D7 posterior dorsal skins for juvenile samples. The method consists of three steps: determining co-expression cutoffs, constructing GCNs, and determining the time order of TF gene expression to transform a GCN into a TO-GCN. First, the Pearson correlation coefficient (PCC) values of all TF–TF gene pairs were calculated and used to determine the cutoffs of co-expression (for p < 0.05, PCC > 0.92 and 0.9 for natal down and juvenile feather, respectively). Second, using the co-expression relationships between TF genes, we determined the GCN for connected TF genes. Third, the time order of TF genes in the GCN was assigned by the breadth-first search (BFS) algorithm ⁸⁶ initiated from the selected node that should be the first upregulated TF in the GCN. BFS is an algorithm for searching a network graph. It starts with an initial seed and searches all its neighbors (nodes with connecting edges) to form a set of nodes (level 1). Then, the process proceeds from all nodes in level 1 and searches their neighbors (excluding level 1 nodes) to form the second set of nodes (level 2) and so on, until all nodes in the network are assigned. The computer programs for the method are available at https://github.com/petitmingchang/TO-GCN (31). SP5 and MYOD1 TF genes were selected as the initial nodes in E7 embryonic samples and D3 juvenile samples, respectively.

Co-expressed gene sets and overrepresented functions at each TO-GCN level

For the TF genes at each level of a TO-GCN, a corresponding set of co-expressed genes (usually non-TF genes) can be identified with the same co-expression relationship for adding the genes to the TO-GCN. Since a gene may be co-expressed with TF genes in multiple levels, two neighboring gene sets may have some overlapping genes. For each set of genes corresponding to a level in a TO-GCN, a functional enrichment analysis was conducted with the background set of all expressed genes in this study. Fisher’s exact test with the false discovery rate (FDR) < 0.05 was applied with functional annotations from Reactome (https://reactome.org/).

Paraffin sections and immunostaining

The chicken and zebra finch posterior dorsal skins were dissected and fixed in 4% paraformaldehyde (PFA) at 4 ℃ overnight, 7 µm paraffin sections were obtained. Hematoxylin and eosin (H&E) staining and immunostaining were performed following a previous procedure ^{47, 87}. The following antibodies were used in the immunostaining: α-SMA (Invitrogen, MA1-06110, 1:50), vimentin (Developmental Studies Hybridoma Bank, H5, 1:30) ), ZEB2 (Proteintech, 14026-1-AP, 1:50). NCAM and TNC were from the Chuong lab ( 1:100) ⁸⁸.

ATAC-seq libraries construction and analysis

The epidermis tissues of adult flight feather follicles were dissected using 2X CMF solution, following a previous method ⁴⁰. The dissected tissues were dissociated and the ATAC-seq libraries were constructed following the same study. Cutadapt was used to remove adapter sequences with the following parameters: -q 20 -cut 1 -length 75 -minimum-length 36 ⁸⁹. Hisat2 with the following parameters was applied to reads mapping: --no-temp-splicesite –no-spliced-alignment -X 2000 ⁸². Samtools was used to remove low quality and non-unique hits with the parameters: view -b -f 3 -F 4 -F 8 -F 256 -F 2048 -q 30 ⁹⁰. After removing the mitochondria alignments, Picard toolkit was used to remove the PCR duplicates (http://broadinstitute.github.io/picard/). MACS2 was used to call peaks with the following parameters: --nomodel –shift − 37 –extsize 73 –keep-dup all –SPMR -f BAMPE ⁹¹. The footprint analysis was performed using HINT-ATAC with default settings ⁹². Embryonic ATAC-seq data were from the previous study ⁴⁰.

Section in situ hybridization (SISH)

To generate RNA probes, PCR was performed using cDNA from E14 feather filaments or adult flight feathers. The following prime pairs were used: CTNNB1 forward: GCAACTCGTGCAATCCCAGA; reverse: CAAAGGCCAGTGTGAGGGTG, LEF1 forward: TCAAGTCCTCGCTGGTCAAC; reverse: GGACATGGAAGGGTCGACTG, GG6AChr25Scale2 forward: CAGTGCCCCGACTCAACG; reverse: AGCAACTAAGAAGACAGGGACT, GG6AChr25Scale10 forward: ATGTCTTGCTCCGACCTGT; reverse: GGGATGTGAAGCTGATAGCATTG, BMP4 forward: TTCCACCATGAAGAGCACCTG; reverse: CAACCCACGTCGCTGAAATC, SOX14 forward: CTCCTTACTTGACCCCAGCCA; reverse: CGACCAAGCGGTACAGTTACAC, TWIST2 forward: TTTCCAAAGGATCTGTCTCAGGA; reverse: TAGTGCGAGGCTGACATGGA, SHH forward: CTGGTGAAGGACCTGAGCCCT; reverse: GCCCAACTGTGCTCCTCGAT. The two scale probes can hit around three scale keratin genes, but two hit groups were not overlapped.

Construction and misexpression of dominant negative form of the candidate genes

To perform tissue specific gene knock down, the misexpression of sequence modified genes were used to achieve the dominant negative (dn) effect. The RCAS-dnLEF1 has been applied in chicken studies and was obtained from Addgene (Plasmid #14022) ⁹³. The dnACTA2 function was published previously ⁴⁴, and the sequence was cloned into RCAS-cherry plasmid in this study. Virus was made according to Jiang ⁸⁷ and concentrated by ultracentrifugation. The flight feathers of SPF chickens were plucked and around 50ul of concentrated RCAS virus was injected into each follicle cavity. All the experiments were conducted in both sexes of chickens to eliminate the probable sexual specific effects. The flight feathers that were plucked at the same time without virus injection were used as the control. Feather follicles including dermal papilla were dissected after two weeks of the virus injection and fixed immediately in 4% PFA.

Acknowledgments

We thank Prof. Gunter Wagner for helpful suggestions. We thank Gee-Way Lin for providing zebra finch scheme templates. We also thank the National Center for High-performance Computing (URL: https://iservice.nchc.org.tw) and the Center for Advanced Research Computing (CARC, URL: https://carc.usc.edu.) at the University of Southern California for providing computing resources that have contributed to the research results in this study. RCAS(A) mdnLef1en was a gift from Cliff Tabin (Addgene plasmid # 14022). The Taiwan team was supported by National Science and Technology Council, Taiwan (NSTC 112-2311-B-045-). The USC team was supported by NIH grants RO1 AR 047364, R37 AR 060306, and the China Medical University (in Taiwan) – University of Southern California collaborative grant 005884-00001. C.-K.C. is supported by the iEGG and Animal Biotechnology Center from the Feature Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE-112-S-0023-A) in Taiwan.

Competing interests

The authors declare no competing or financial interests.

Author contributions

Conceptualization: C.-K.C., W.-H.L., C.-M.C.; Methodology: C.-K.C., Y.-M. C., T.-X.J., P.W.; Histology: C.-K.C., T.-X.J., T.-Y. L., T.-D. V., T.-Y. H., P. W., H. I-C. H.; Bioinformatic analysis: C.-K.C., Y.-M. C., J.-J. L.; Experimental validations: C.-K.C., T.-X.J., Z.C. Y., P. W., T.-Y. L., J. L., Z. Y., H. I-C. H.; Resources: C.-K.C. and C. S. N.; Writing and editing: C.-K.C., Y.-M. C., P. W., W.-H.L., C.-M.C.; Supervision and funding acquisition: W.-H.L., C.-M.C.

Prum RO, Brush AH. The evolutionary origin and diversification of feathers. The Quarterly review of biology 77, 261–295 (2002).
Prum RO. Evolution of the morphological innovations of feathers. Journal of experimental zoology Part B, Molecular and developmental evolution 304, 570–579 (2005).
Chen CF, et al. Development, regeneration, and evolution of feathers. Annual review of animal biosciences 3, 169–195 (2015).
Strasser B, Mlitz V, Hermann M, Tschachler E, Eckhart L. Convergent evolution of cysteine-rich proteins in feathers and hair. BMC evolutionary biology 15, 82 (2015).
Podulka S, Rohrbaugh RW, Bonney R. Handbook of bird biology, 2nd edn. Cornell Lab of Ornithology in association with Princeton University Press (2004).
Lovette IJ, Fitzpatrick JW. Handbook of Bird Biology. Wiley (2016).
Humphrey PS, Parkes KC. An Approach to the Study of Molts and Plumages. The Auk 76, 1–31 (1959).
Terres JK, National Audubon Society. The Audubon Society encyclopedia of North American birds. Wings Books: Distributed by Outlet Book Co. (1991).
Chuong CM, Yeh CY, Jiang TX, Widelitz R. Module-based complexity formation: periodic patterning in feathers and hairs. Wiley Interdiscip Rev Dev Biol 2, 97–112 (2013).
Chuong CM, Randall VA, Widelitz RB, Wu P, Jiang TX. Physiological regeneration of skin appendages and implications for regenerative medicine. Physiology (Bethesda) 27, 61–72 (2012).
Chuong CM, Edelman GM. Expression of cell-adhesion molecules in embryonic induction. I. Morphogenesis of nestling feathers. J Cell Biol 101, 1009–1026 (1985).
Chuong CM, Edelman GM. Expression of cell-adhesion molecules in embryonic induction. II. Morphogenesis of adult feathers. J Cell Biol 101, 1027–1043 (1985).
Yu M, Wu P, Widelitz RB, Chuong CM. The morphogenesis of feathers. Nature 420, 308–312 (2002).
Harris MP, Williamson S, Fallon JF, Meinhardt H, Prum RO. Molecular evidence for an activator-inhibitor mechanism in development of embryonic feather branching. Proc Natl Acad Sci U S A 102, 11734–11739 (2005).
Yue Z, Jiang TX, Widelitz RB, Chuong CM. Mapping stem cell activities in the feather follicle. Nature 438, 1026–1029 (2005).
Yue Z, Jiang TX, Widelitz RB, Chuong CM. Wnt3a gradient converts radial to bilateral feather symmetry via topological arrangement of epithelia. Proc Natl Acad Sci U S A 103, 951–955 (2006).
Li A, et al. Diverse feather shape evolution enabled by coupling anisotropic signalling modules with self-organizing branching programme. Nat Commun 8, ncomms14139 (2017).
Lin J, Yue Z. Coupling of apical-basal polarity and planar cell polarity to interpret the Wnt signaling gradient in feather development. Development 145, (2018).
Chang WL, et al. The Making of a Flight Feather: Bio-architectural Principles and Adaptation. Cell 179, 1409–1423 e1417 (2019).
Widelitz RB, et al. Morpho-regulation in diverse chicken feather formation: Integrating branching modules and sex hormone-dependent morpho-regulatory modules. Dev Growth Differ 61, 124–138 (2019).
Dunn PO, Armenta JK, Whittingham LA. Natural and sexual selection act on different axes of variation in avian plumage color. Sci Adv 1, e1400155 (2015).
Gazda MA, et al. A genetic mechanism for sexual dichromatism in birds. Science 368, 1270–1274 (2020).
Lucas AM, Agriculture USDo, Stettenheim PR, Science US, Administration E. Avian Anatomy Integument. Avian Anatomy Project, Poultry Research Branch, Animal Science Research Division, Agricultural Research Service, U.S. Department of Agriculture (1972).
Lin PY, Huang PY, Lee YC, Ng CS. Analysis and comparison of protein secondary structures in the rachis of avian flight feathers. PeerJ 10, e12919 (2022).
Wu P, et al. Cyclic growth of dermal papilla and regeneration of follicular mesenchymal components during feather cycling. Development 148, (2021).
Chu Q, et al. Dkk2/Frzb in the dermal papillae regulates feather regeneration. Developmental biology 387, 167–178 (2014).
Zhang F, Jiang L, Wang S. Repairable cascaded slide-lock system endows bird feathers with tear-resistance and superdurability. Proc Natl Acad Sci U S A 115, 10046–10051 (2018).
Chang YM, et al. Comparative transcriptomics method to infer gene coexpression networks and its applications to maize and rice leaf transcriptomes. Proc Natl Acad Sci U S A 116, 3091–3099 (2019).
Buenrostro JD, Wu B, Chang HY, Greenleaf WJ. ATAC-seq: A Method for Assaying Chromatin Accessibility Genome-Wide. Curr Protoc Mol Biol 109, 21 29 21–21 29 29 (2015).
Corces MR, et al. An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues. Nature methods 14, 959–962 (2017).
Chen CK, et al. Regulatory Differences in Natal Down Development between Altricial Zebra Finch and Precocial Chicken. Molecular biology and evolution, (2016).
Starck JM, Ricklefs RE. Avian growth and development: evolution within the altricial-precocial spectrum. Oxford University Press (1998).
Chen CK, Chuang HF, Wu SM, Li WH. Feather Evolution from Precocial to Altricial Birds. Zool Stud 58, e24 (2019).
Hodge RG, Ridley AJ. Regulating Rho GTPases and their regulators. Nat Rev Mol Cell Biol 17, 496–510 (2016).
Heasman SJ, Ridley AJ. Mammalian Rho GTPases: new insights into their functions from in vivo studies. Nat Rev Mol Cell Biol 9, 690–701 (2008).
Lu P, Takai K, Weaver VM, Werb Z. Extracellular matrix degradation and remodeling in development and disease. Cold Spring Harb Perspect Biol 3, (2011).
Mecham RP. Overview of extracellular matrix. Curr Protoc Cell Biol Chap. 10, 10 11 11 – 10 11 16 (2012).
Wu XS, et al. Self-assembly of biological networks via adaptive patterning revealed by avian intradermal muscle network formation. Proc Natl Acad Sci U S A 116, 10858–10867 (2019).
Bhattacharjee MJ, et al. Regulatory Divergence among Beta-Keratin Genes during Bird Evolution. Molecular biology and evolution 33, 2769–2780 (2016).
Liang YC, et al. Folding Keratin Gene Clusters during Skin Regional Specification. Dev Cell 53, 561–576 e569 (2020).
Harris MP, Fallon JF, Prum RO. Shh-Bmp2 signaling module and the evolutionary origin and diversification of feathers. J Exp Zool 294, 160–176 (2002).
Palmquist KH, et al. Reciprocal cell-ECM dynamics generate supracellular fluidity underlying spontaneous follicle patterning. Cell 185, 1960–1973 e1911 (2022).
Lin SJ, et al. Topology of feather melanocyte progenitor niche allows complex pigment patterns to emerge. Science 340, 1442–1445 (2013).
Guo DC, et al. Mutations in smooth muscle alpha-actin (ACTA2) cause coronary artery disease, stroke, and Moyamoya disease, along with thoracic aortic disease. Am J Hum Genet 84, 617–627 (2009).
Greenwold MJ, et al. Dynamic evolution of the alpha (alpha) and beta (beta) keratins has accompanied integument diversification and the adaptation of birds into novel lifestyles. BMC evolutionary biology 14, 249 (2014).
Ng CS, et al. Genomic organization, transcriptomic analysis, and functional characterization of avian α-and β-keratins in diverse feather forms. Genome biology and evolution 6, 2258–2273 (2014).
Wu P, et al. Topographical mapping of alpha- and beta-keratins on developing chicken skin integuments: Functional interaction and evolutionary perspectives. Proc Natl Acad Sci U S A 112, E6770-6779 (2015).
Davis AC, Greenwold MJ, Sawyer RH. Complex Gene Loss and Duplication Events Have Facilitated the Evolution of Multiple Loricrin Genes in Diverse Bird Species. Genome biology and evolution 11, 984–1001 (2019).
Lin GW, Lai YC, Liang YC, Widelitz RB, Wu P, Chuong CM. Regional Specific Differentiation of Integumentary Organs: Regulation of Gene Clusters within the Avian Epidermal Differentiation Complex and Impacts of SATB2 Overexpression. Genes (Basel) 12, (2021).
Holthaus KB, et al. Comparative Analysis of Epidermal Differentiation Genes of Crocodilians Suggests New Models for the Evolutionary Origin of Avian Feather Proteins. Genome biology and evolution 10, 694–704 (2018).
Davis A, Greenwold MJ. Evolution of an Epidermal Differentiation Complex (EDC) Gene Family in Birds. Genes (Basel) 12, (2021).
Katsuyama T, Kadoya M, Shirai M, Sasai N. Sox14 is essential for initiation of neuronal differentiation in the chick spinal cord. Developmental dynamics: an official publication of the American Association of Anatomists 251, 350–361 (2022).
Ng CS, et al. Genomic organization, transcriptomic analysis, and functional characterization of avian alpha- and beta-keratins in diverse feather forms. Genome biology and evolution 6, 2258–2273 (2014).
Ng CS, et al. Transcriptomic analyses of regenerating adult feathers in chicken. BMC genomics 16, 756 (2015).
Chen CK, Juan WT, Liang YC, Wu P, Chuong CM. Making region-specific integumentary organs in birds: evolution and modifications. Curr Opin Genet Dev 69, 103–111 (2021).
Mallo M, Alonso CR. The regulation of Hox gene expression during animal development. Development 140, 3951–3963 (2013).
Deschamps J, Duboule D. Embryonic timing, axial stem cells, chromatin dynamics, and the Hox clock. Genes Dev 31, 1406–1416 (2017).
Noordermeer D, Leleu M, Schorderet P, Joye E, Chabaud F, Duboule D. Temporal dynamics and developmental memory of 3D chromatin architecture at Hox gene loci. eLife 3, e02557 (2014).
Li J, et al. The crest phenotype in domestic chicken is caused by a 197 bp duplication in the intron of HOXC10. G3 (Bethesda) 11, (2021).
Hidalgo M, et al. A conserved molecular template underlies color pattern diversity in estrildid finches. Sci Adv 8, eabm5800 (2022).
Tsai DY, et al. Chicken HOXC8 and HOXC10 genes may play a role in the altered skull morphology associated with the Crest phenotype. Journal of experimental zoology Part B, Molecular and developmental evolution, (2023).
Marchegiani S, et al. Recurrent Mutations in the Basic Domain of TWIST2 Cause Ablepharon Macrostomia and Barber-Say Syndromes. Am J Hum Genet 97, 99–110 (2015).
Zhang HJ, et al. Twist2 promotes kidney cancer cell proliferation and invasion by regulating ITGA6 and CD44 expression in the ECM-receptor interaction pathway. Onco Targets Ther 9, 1801–1812 (2016).
Li S, et al. Twist2 amplification in rhabdomyosarcoma represses myogenesis and promotes oncogenesis by redirecting MyoD DNA binding. Genes Dev 33, 626–640 (2019).
Wu P, et al. Multiple Regulatory Modules Are Required for Scale-to-Feather Conversion. Molecular biology and evolution 35, 417–430 (2018).
Gao G, et al. Comparative genomics and transcriptomics of Chrysolophus provide insights into the evolution of complex plumage coloration. Gigascience 7, (2018).
Ng CS, et al. Genome Assembly and Evolutionary Analysis of the Mandarin Duck Aix galericulata Reveal Strong Genome Conservation among Ducks. Genome biology and evolution 14, (2022).
Dorshorst B, Okimoto R, Ashwell C. Genomic regions associated with dermal hyperpigmentation, polydactyly and other morphological traits in the Silkie chicken. J Hered 101, 339–350 (2010).
Edwards SV, Cloutier A, Baker AJ. Conserved Nonexonic Elements: A Novel Class of Marker for Phylogenomics. Syst Biol 66, 1028–1044 (2017).
Sackton TB, et al. Convergent regulatory evolution and loss of flight in paleognathous birds. Science 364, 74–78 (2019).
Ksepka DT. Feathered dinosaurs. Curr Biol 30, R1347-R1353 (2020).
Xu X, Zheng X, You H. A new feather type in a nonavian theropod and the early evolution of feathers. Proc Natl Acad Sci U S A 106, 832–834 (2009).
Xu X, Zheng X, You H. Exceptional dinosaur fossils show ontogenetic development of early feathers. Nature 464, 1338–1341 (2010).
Xu X, Zhou Z, Prum RO. Branched integumental structures in Sinornithosaurus and the origin of feathers. Nature 410, 200–204 (2001).
Love AC, Wagner GP. Co-option of stress mechanisms in the origin of evolutionary novelties. Evolution 76, 394–413 (2022).
Plikus MV, et al. Self-organizing and stochastic behaviors during the regeneration of hair stem cells. Science 332, 586–589 (2011).
Yates AD, et al. Ensembl 2020. Nucleic Acids Res 48, D682-D688 (2020).
O'Leary NA, et al. Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic Acids Res 44, D733-745 (2016).
El-Gebali S, et al. The Pfam protein families database in 2019. Nucleic Acids Res 47, D427-D432 (2019).
Eddy SR. Accelerated Profile HMM Searches. PLoS computational biology 7, e1002195 (2011).
Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014).
Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nature methods 12, 357–360 (2015).
Pertea M, Pertea GM, Antonescu CM, Chang TC, Mendell JT, Salzberg SL. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nature biotechnology 33, 290–295 (2015).
Bullard JH, Purdom E, Hansen KD, Dudoit S. Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments. BMC Bioinformatics 11, 94 (2010).
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome biology 15, 550 (2014).
Knuth DE. The art of computer programming, volume 2 (3rd ed.): seminumerical algorithms. Addison-Wesley Longman Publishing Co., Inc. (1997).
Jiang T. Current methods in the study of avian skin appendages. Molecular basis of epithelial appendage morphogenesis, 395–408 (1998).
Chuong CM, Chen HM. Enhanced expression of neural cell adhesion molecules and tenascin (cytotactin) during wound healing. Am J Pathol 138, 427–440 (1991).
Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. 2011 17, 3 (2011).
Li H, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Zhang Y, et al. Model-based analysis of ChIP-Seq (MACS). Genome biology 9, R137 (2008).
Li Z, Schulz MH, Look T, Begemann M, Zenke M, Costa IG. Identification of transcription factor binding sites using ATAC-seq. Genome biology 20, 45 (2019).
Kengaku M, et al. Distinct WNT pathways regulating AER formation and dorsoventral polarity in the chick limb bud. Science 280, 1274–1277 (1998).

There is NO Competing Interest.

SupplementaryFigures.docx
SupplementaryTable1TPMtableallsamples.xlsx
Supplementary Table 1
SupplementaryTable2differenceofTOGCNlevelbetweenembryonicandposthatch.xlsx
Supplementary Table 2
SupplementaryTable3EnrichedReactomepathwaysforembryonicsamples.xlsx
Supplementary Table 3
SupplementaryTable4EnrichedReactomepathwaysforposthatchsamples.xlsx
Supplementary Table 4
SupplementaryTable5TOGCNrefoverlappedTFslevel8to10diff3.xlsx
Supplementary Table 5
SupplementaryTable6TFcoexpressedgeneswithTOGCNlevel.xlsx
Supplementary Table 6
SupplementaryTable7TPMmt1uniquecodingkeratinocyteAVGstringtieegg60129FT.xlsx
Supplementary Table 7

Download PDF

Journal Publication

published 16 May, 2024

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Transition from natal downs to juvenile feathers: conserved regulatory switches in Neoaves

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

The primary feather transition in chicken and zebra finch: from natal downs to juvenile feathers

Coordination of multiple signaling pathways during the primary feather transition

ECM re-organization generates peripheral pulp for feather branching morphogenesis

ACTA2 shapes adult dermal papilla to compartment dermal papilla stem cells for cyclic renewal

Many scale keratins are specifically upregulated in juvenile feather sheath by SOX14

Discussion

Materials and methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1