RESEARCH ARTICLE SUMMARY◥

CELL FATE DECISIONS

Spatiotemporal antagonism inmesenchymal-epithelial signaling insweat versus hair fate decisionCatherine P. Lu, Lisa Polak, Brice E. Keyes, Elaine Fuchs*

INTRODUCTION: Across the vertebrate king-dom, epithelial appendages—including mam-mary, sweat, and salivary glands; hair follicles(HFs); teeth; scales; and feathers—begin to formduring embryogenesis whenWNT signaling in-structs progenitorswithin the epithelial sheet toorganize into morphologically similar placodes.Most animals restrict these epidermal append-ages to distinct body regions. This paradigmshifted late inmammalian evo-lution; the dual presence of HFsandeccrinesweatglands (SwGs)in the skin is a recent acquisi-tion of primates.Classical tissue recombina-

tion experiments from the 1960srevealed that mesenchyme di-rects the divergent downstreamevents that determine append-age specification and pattern-ing. Relatively little is knownabout the specific spatiotemporalcross-talk and molecular mech-anisms that underlie epithelialfate specification in response tomesenchymal signals. Elucidat-ing the epithelial-mesenchymalcross-talk involved in regionalskin appendage specification isintegral to understanding howhumans have adjusted thesemechanisms to endow themwith a greater capacity thanthat of their hairy cousins tolive in diverse environments.

RATIONALE: The acquisition of SwGs andtheir importance in thermoregulation and waterbalance are underscored by human patients whosuffer from the life-threatening condition oflacking SwGs, either from loss in severe burnsor from genetic disorders. Conversely, a gain-of-function variant that elevates SwG numbershas been expanding among the Southeast Asianpopulation, where excessive SwGs are desirable.By elucidating the underlying mechanisms thatdistinguish humans from other mammals intheir ability to make both glands and follicles

over their body skin, our findings could pavethe way for future therapeutic advances in skinregeneration with dual appendages.

RESULTS: Using mouse as a model, we ex-plored the differences between the back skinmesenchyme, which is only competent to spec-ify HFs, and the foot skin mesenchyme, whichcan onlymake SwGs.Using genome-wide analy-

ses and functional studies, we discovered thatjust after the formation ofmorphologically sim-ilar epidermal buds, appendage choice is de-termined through regional skin differences inmesenchymal expressionofbonemorphogeneticproteins (BMPs). Probing into mechanisms, weshowed thatwhenBMPsare elevated in foot skinmesenchyme, BMP signaling is activated in bothdermis and epidermis. This triggers a cascade ofdownstream signaling events. WNT signaling iselevated in thedermis and reduced in the epider-mis.Mesenchymal fibroblastgrowthfactors (FGFs)

appear and affect the overlying epithelium.These converging pathways lead to suppressionof sonic hedgehog (SHH) in the epithelium.This BMP:SHH antagonism within the ep-

ithelial bud specifies SwG fate and preventsHF fate. Moreover, by manipulating gene ex-pression in vivo at specific developmentalstages, we demonstrated that this signalingcircuitry acts only within a narrow window

of time during mouse em-bryogenesis. Thus, whenSHH is ectopically expres-sed in foot skin epitheli-um during the permissivephase, HF-specific gene ex-pression is up-regulated

in the epithelial bud, whereas SHH signaling inthe mesenchyme stimulates expression of BMPantagonists, further suppressing local BMPsignaling and blocking SwG fate. In humanskins, this antagonistic interplay of BMP:SHHsignaling occurs temporally, in addition tospatially. The first bud waves are specified asHFs, and then a tipping in the balance ofBMP:SHH signaling results in the last waves

of buds becoming SwGs.

CONCLUSION: Our findings re-vealed a differential impact ofBMP signaling on appendage fatespecification that has ancient rootsand occurs repeatedly throughoutvertebrate evolution. In the evolu-tionary developmental biology viewof BMP signaling and fate specifi-cation of integument, chicken scalesand mammalian SwGs require BMPsignaling to specify their fate, whereasfeathers and HFs must suppress it.Our discovery of BMP-SHH antag-onism in bud fate choice uncoveredadditional evolutionary parallels,this time between two even moredistantly related epidermal append-ages, the mammalian SwG andthe fly wing. Our studies providenew insights into how elevatedmesenchymal BMP signaling trig-gers a self-perpetuating molecularcascade that culminates in silenc-ing of SHH signaling to suppressone appendage fate and specify

another. In most mammals, the BMP:SHHantagonism is regulated spatially. Humans,however, have evolved to regulate it tem-porally, endowing them with greater abilityto run marathons and survive in extremeclimates.▪

RESEARCH

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 1551

BMP-SHH antagonism specifies SwG versus HF fate. To specify SwGs,mesenchymal-derived BMPs and FGFs signal to epithelial buds and suppressepithelially derived SHH production. Conversely, hair follicles are specifiedwhen mesenchymal BMP signaling is inhibited, permitting SHH production.This antagonism is spatially restricted in most mammals but temporallyregulated in humans, permitting the presence of HFs (pink), SwGs (purple),or both HFs and SwGs (pink with purple droplet) throughout our body skin.

The list of author affiliations is available in the full article online.*Corresponding author. Email: fuchslb@rockefeller.eduCite this article as C. P. Lu et al., Science 354, aah6102 (2016).DOI: 10.1126/science.aah6102

ON OUR WEBSITE◥

Read the full articleat http://dx.doi.org/10.1126/science.aah6102..................................................

on March 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

RESEARCH ARTICLE◥

CELL FATE DECISIONS

Spatiotemporal antagonism inmesenchymal-epithelial signaling insweat versus hair fate decisionCatherine P. Lu,1 Lisa Polak,1 Brice E. Keyes,1* Elaine Fuchs1,2†

The gain of eccrine sweat glands in hairy body skin has empowered humans to runmarathons and tolerate temperature extremes. Epithelial-mesenchymal cross-talk isintegral to the diverse patterning of skin appendages, but the molecular events underlyingtheir specification remain largely unknown. Using genome-wide analyses and functionalstudies, we show that sweat glands are specified by mesenchymal-derived bonemorphogenetic proteins (BMPs) and fibroblast growth factors that signal to epithelial budsand suppress epithelial-derived sonic hedgehog (SHH) production. Conversely, hair folliclesare specified when mesenchymal BMP signaling is blocked, permitting SHH production. Fatedetermination is confined to a critical developmental window and is regionally specified inmice. In contrast, a shift from hair to gland fates is achieved in humans when a spike in BMPsilences SHH during the final embryonic wave(s) of bud morphogenesis.

Epithelial appendages—including hair fol-licles (HFs) and teeth as well as mammary,sweat, and salivary glands—begin to formduring embryogenesis when Wnt signalingtriggers progenitors within the epithelial

sheet to organize spatially into morphologicallysimilar placodes. Whereas most mammals restrictthe specification of epidermal appendages re-gionally, HFs and sweat glands (SwGs) coexistthroughout much of the skin of primates. Theacquisition of SwGs and their importance in ther-moregulation are underscored by humans whosuffer from a life-threatening condition whenSwGs are eliminated, either from loss in severeburns or from the genetic disorder hypohidroticectodermal dysplasia (HED).Patients with HED display mutations in genes

encoding proteins such as ectodermal dysplasiaantigen (EDA), the EDA receptor (EDAR), andWNT10a, a ligand for Wnt signaling. These find-ings have illuminated a critical role for thesepathways in controlling the development of anumber of epidermal appendages, including SwGsand coarse hairs (1, 2). Conversely, an EDAR gain-of-function variant has been expanding amongthe Southeast Asian population, where excessiveSwGs are desirable because of the hot and humidclimate of that region (3). In this regard, both Wntand EDA/EDAR pathways appear to function inpromoting placode formation, increasing the den-sity of SwGs and several other appendage types (4).Classical tissue recombination experiments

have revealed that mesenchyme plays a critical

role in dictating the divergent downstream eventsthat determine appendage selection (5, 6). Forexample, salivary mesenchyme combined withmammary epithelium generates an epithelialmorphology resembling salivary glands (7), andregardless of its regional origin, embryonic chickepidermis develops feathers if combined withfeather-forming dermis but develops scales ifthe epidermis is exposed to scale-forming dermis(8). Similarly, rabbit corneal epithelium gener-ates either HFs or SwGs depending on whetherit is exposed to the mesenchyme of dorsal backskin or ventral foot skin, respectively (9). Despitethe existence of many such examples, relativelylittle is known about the specific spatiotemporalcross-talk and molecular mechanisms that under-lie epithelial fate specification in response tomesenchymal exposure.

Regional differences in BMP pathwaygenes in skin mesenchymes

To understand how epithelial fate is specifiedby mesenchymal signals, we began by exploit-ing the fact that in mice, dorsal back skin sup-ports only HF morphogenesis, whereas eccrineSwG morphogenesis is restricted to ventral footskin. We first pinpointed the appearance ofWNT-induced epidermal placodes at these bodysites, reasoning that the underlying mesenchymesbecome competent to specify the differential fatesof these placodes. We then carried out genome-wide, high-throughput RNA sequencing (RNA-seq) on body site–specific dermal mesenchymesat the embryonic ages when their respectiveplacodes emerge [embryonic day 14.5 (E14.5),dorsal back skin; E17.5, ventral foot skin] (Fig. 1A).Comparative bioinformatics revealed that bone

morphogenetic protein (BMP) signaling path-way genes were significantly enriched [≥2×, P <

0.05, false-discovery rate (q value) < 0.05] amongthe transcripts [with fragments per kilobase permillionmapped reads (FPKM) > 1] in ventral footskin versus dorsal back skin dermis (Fig. 1B). Ofthese, the BMP5 ligand stood out for its regional-specific expression, confirmed with quantitativepolymerase chain reaction (PCR) (Fig. 1C). Alsosurfacing in our differential transcriptome anal-ysis were Bmpr1a, encoding a key component ofthe BMP receptor kinase, and Id1 and Msx1, twodownstream targets of its activated transcrip-tional effectors, phospho-SMAD1/5/8 (Fig. 1B).Anti-phospho-SMAD1/5/8 immunofluorescence

confirmed the marked differential elevation inBMP signaling within ventral foot skin versusdorsal back skin during the fate-permissive stateof their epithelial buds (Fig. 1D). However, therewas also strong nuclear immunolabeling in theoverlying SwGbuds, not seen inHFbuds.Bmpr1atranscripts were also markedly elevated in SwG-permissive epidermis, relative to HF-permissiveepidermis (Fig. 1E).We therefore examinedBMPsand BMP inhibitors in both mesenchyme andepidermis. Despite expression of other BMPsand BMP inhibitors in both mesenchyme andepidermis, only mesenchymal BMP5 exhibiteda regional-specific expression that might ac-count for the striking difference in BMP sig-naling that we observed between SwG and HFbuds (Fig. 1C).

Mesenchymal-epithelial BMP signalingin specifying SwGs

If paracrine mesenchymal-epithelial BMP sig-naling in foot skin occurs, as our data suggest,and if this cross-talk is physiologically impor-tant, then loss of mesenchymal Bmp5 shouldaffect overlying SwG development. Ablating Bmp5significantly reduced SwG numbers in ventralfoot skin and resulted in decreased epithelialBMP signaling (pSMAD1) (Fig. 1F and fig. S1A).Although the effect was robust, SwGs still de-veloped and BMP signaling was still seen inboth mesenchymal and epithelial compartmentsof ventral foot skin (fig. S1A), suggesting thatother mesenchymal BMPs may partially com-pensate for the loss of BMP5. To test this pos-sibility and assess the full functional relevanceof the elevated BMP signaling in the footpads,we used K14-Cre mice to conditionally ablateBmpr1a in embryonic skin epidermis. Immuno-pSMAD1/5/8 labeling confirmed that BMP sig-naling was completely abrogated in the epithe-lium, whereas it was still sustained as expectedin the underlying mesenchyme (fig. S1B). Un-der these conditions, no SwGs were generated(fig. S1C).Instead of forming a coiled gland, the Bmpr1a-

null epithelial down-growths in the ventral footpads enveloped the associated dermal mesen-chyme and assumed amorphology, biochemistry,and proliferative status more typical of HF bulbs(Fig. 1G and fig. S1D). By postnatal day 5 (P5),K17, LHX2, and P-cadherin—hallmarks of HFs(10) and not expressed by SwGs—were all ectopi-cally expressed in Bmpr1a-null foot skin epithe-lial down-growths. In contrast, smooth muscle

RESEARCH

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-1

1Laboratory of Mammalian Cell Biology and Development,The Rockefeller University, New York, NY 10065, USA.2Howard Hughes Medical Institute, Chevy Chase, MD, 20815-6789, USA.*Present address: Calico Life Sciences, South San Francisco, CA94080, USA. †Corresponding author. Email: fuchslb@rockefeller.edu

on March 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

actin, featured by the myoepithelial cells of thegland, was absent (Fig. 1G). Further reflectivethat an epithelial fate switch had occurred, weobserved condensation of alkaline phosphatase–

positive mesenchymal cells into a structure re-sembling a dermal papilla (DP), the cluster ofmesenchymal cells associated with the HF thatare essential for its growth and maintenance

(11). These data not only document the physi-ological relevance of elevating epithelial BMPsignaling in developing SwGs but also suggestthat paracrine cross-talk between mesenchymal

aah6102-2 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

*** **

****

***ns ns

E14.5 dorsal back derE17.5 ventral foot derE14.5 dorsal back epiE17.5 ventral foot epi

Rel

ativ

e ex

pres

sion

0

0.5

1

1.5

2

2.5

3

E14 Back epi

E17 Foot epi

E18 Foot epi

Rel

ativ

e ex

pres

sion

EpidermalBmpr1a

********

Bmp5 +/-

Bmp5 -/-

dap toof lartnev 5P

Epidermalplacode

E17.5Ventral foot skin

E14.5 Dorsal

back skin

Hair Follicle Sweat gland

Ventral foot skin

8/5/1D

AM

Sp

Gβ

TI4

Dorsal back skin

?? ?

E17.5 ventral foot dermis (FPKM) 1e+01 1e+051e+03

simred kcab lasrod 5.41

E)

MK

PF(

1e

+01

1e+0

31e

+05

1e+0

7

Transcriptome analysis

1 +01 0Msx1Bmp5

Id1

Wild type E17.5

K18+SMA+

SwG fate

K17 hiLHX2+

HF fate

WT

Bmpr1a cKO

Tota

l no.

of s

wea

t duc

ts

Foot skin

Foot skin

K5 AP DAPI LHX2 PCAD DAPIK17 DAPISMA ITG β4

(P= 5.3E-05)****

Bmp5+/-

Bmp5-/-

Fig. 1. Mesenchymal-epithelial BMP signaling is required for sweat glandfate. (A) Schematics of spatiotemporally distinct development of mouse sweatglands and hair follicles. (B) Scatter plot of RNA-seq analysis of E14.5 dorsalback skin and E17.5 ventral foot skin dermis. Bmpr1a, Bmp5, Id1, and Msx1(targets of BMP signaling) are enriched in ventral foot skin dermis (blue dots).(C) Quantitative PCR for various BMP ligands and inhibitors in E14.5 dorsal backskin (HF-permissive) and E17.5 ventral foot skin (SwG-permissive) epidermis anddermis. Shown is a strong up-regulation of Bmp5 in mouse foot skin dermis.(D) Immunofluorescence of phospho-SMAD1/5/8 in dorsal back and ventralfoot skins. Arrows denote hair and sweat placodes, respectively. Scale bars,50 mm. (E) Quantitative PCR for epidermalBmpr1amRNAs at indicated embryonic

stages and regional skins. (F) (Left) Quantification of sweat ducts in adult Bmp5heterozygous and null mice. n = 10 and 16 foot skin samples, respectively. (Right)Representative images of adult foot pad epidermis stained with Nile blue A in orderto visualize sweat ducts. (G) Immunofluorescence images of foot pads of wild type(WT) and Krt14-Cre; Bmpr1a conditional knockout (cKO) at P5. SMA, smoothmuscle actin, a myoepithelial glandular marker; K5, keratin 5, which marks thebasal layer of epidermal progenitors; ITGb4, integrinb4, which marks the basalsurface of the epidermis; AP, alkaline phosphatase, a marker for HF dermalpapillae; K17, LHX2 and PCAD, markers of the HF. Gray solid lines denote skinsurface. Dashed lines indicate dermo-epidermal border. Scale bars, 50 mm.(Right) Schematics of SwG fate and HF fate in the foot skin.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

BMPs and BMP receptorhigh epithelial buds iscritical for SwG development.

BMP signaling can cause a hair follicleto glandular fate switch

To further test whether these regional differencesin BMP signaling are key to SwG versus HF ap-pendage fate, we generated lentiviruses harbor-ingaPGK-H2BGFPmarkeranddoxycycline-inducibletransgenes encoding BMPs and BMP inhibitor.We then used an ultrasound-guided in utero in-jection system (12) to introduce lentivirus into theamniotic sacs of living E9.5 K14rtTA mouse em-bryos (Fig. 2A).Within 24hours, the transgene andgreen fluorescent protein (GFP) marker selectivelyand stably integrate into the host genome ofepidermal progenitor cells and their progeny (12).

Transgene expression can then be controlled bythe timing of doxycycline administration.We first tested the consequences of elevating

BMPs in dorsal hairy skin. After transducing withTRE-Bmp5, we administered doxycycline at E16.5of gestation. Transgene expression was robust byE17.5 during permissive phases of sweat buds andfinal waves of HF buds (Fig. 2, B and C). By ana-lyzing at P5, we learned that BMP5 dramaticallyperturbed back skin HF morphogenesis, and HFmarkers were lost (Fig. 2C). A similar block inHF morphogenesis was seen upon activation ofTRE-Bmp4 in embryonic back skin epidermis(Fig. 2D). Elevating BMP levels not only blockedHF morphogenesis in dorsal back skin but alsoectopically induced glandular and ductal-likestructures, accompanied by expression of glan-

dular keratin, K18 (Fig. 2D). Together with ourloss-of-function data, these results suggest thatcollective BMP levels, rather than specific ligand,affect fate choice.

BMP inhibition acts within a criticalwindow to cause SwG to HF fate switch

We next examined the consequences of inhibit-ing BMP signaling throughout the ventral footskin by ectopically expressing the secreted BMPinhibitor NOGGIN. Prior studies on back skinrevealed a role forNOGGIN in promotingHF fate(13), and it was previously postulated that ectopicNOGGIN might trigger “transdifferentiation” ofSwGs into hairs (14). Our findings raised theadditional possibility that by intercepting mes-enchymal BMPs, NOGGINmight act by blocking

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-3

GOITRE H2BGFPPGK

K14rtTA+E9.5 E16.5

(or postnatal)

+ Doxy

Analyze dorsal or foot skin at various

time points

Ultrasound-guidedlentiviral (LV) injection

doxy-inducible lentiviral construct

E9.5 E16.5 P5

TRE-Bmp5 + Doxy

K14rtTA- (control)K14rtTA+ (experiment)

P5 Control P5 (BMP5 a/E16).

niks rae lasroD

WT Ventral foot skinSweat gland fate

WT Dorsal back skinHair follicle fate

LHX2

Epi

Der

Placode

Peg

K18+AQP5+NKA+

Epi

K17 hiLHX2+ LEF1+

E14.5 - E17.5

E17.5 - P0

E17.5 - P0

P4

Epi

Der

Epi

Dorsal skin

HF fate

K18+Glandular

fate

P5 (BMP5 a/E16.5)niks daeh lasroD

TRE-Bmp5-H2BGFP K18

E9.5 E16.5 P5

TRE-Bmp4 + Doxy

TRE-Bmp4-H2BGFP LHX2 ITG β4

niks kcab lasroD

Dorsal skin

HF fate

K18+Glandular

fate

P5 (BMP4 a/E16.5)

P5 (BMP4 a/E16.5)

DAPI

K14rtTA+

TRE-Bmp5-H2BGFP K17 ITG β4 TRE-Bmp4-H2BGFP K17 DAPI TRE-Bmp4 K18 ITG β4

P5 (BMP4 a/E16.5)niks kcab lasroD

Fig. 2. Ectopic elevation of BMP signaling in dorsal skin elicits a HF:glandular fate shift. (A) Schematic of ultrasound-guided in utero lentiviralinjections and doxycycline (Doxy)–inducible ectopic expression system. K14-rtTAembryos express the Doxy-inducible transactivator beginning at ~E13.5.(B) Schematic of wild-type hair and sweat placode and peg stages. HF andSwG markers are shown in pink and blue, respectively. (C and D) Timeline oflentiviral injection, BMP5 (C) and BMP4 (D) induction and analyses. Im-

munofluorescence images of dorsal skin show loss of HFmarkers and ectopicexpression of glandular luminal cell marker K18. Lentiviral injections wereperformed at least two times, and two to four animals were collected andanalyzed at each time point. Boxed areas are shown in higher magnificationat right. Gray solid lines mark the skin surface. Dashed lines indicate dermo-epidermal border and arrows indicate ductal structures associated with BMP4-expressing epidermal cells. Scale bars, 50 mm; except (C) lower right, 25 mm.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

signaling directly to the epithelium, redirectingfoot skin buds to a HF fate. To test this possibility,it was necessary to control NOGGIN expression.We first induced a TRE-Noggin transgene at

E16.5 of embryogenesis just before sweat budformation. Soon thereafter, the epithelial footskin buds showed signs of HF morphogenesis,including LIM/homeobox protein (LHX2) expres-sion (Fig. 3A). Quantification revealed that evenat these early times after Noggin activation, ~50%of transduced (GFP+) cells were already LHX2+,an early HF fate indicator. These changes oc-curred before signs of dermal condensation.A short pulse of NOGGIN expression at this

critical placode stage was sufficient to triggerthese events, which then lasted postnatally, asevidenced by the LHX2-positive hair pegs with-in P5 foot skin (Fig. 3B). In contrast, when wewaited until P2 to initiate NOGGIN expression,no adverse effects were seen in the foot skin(Fig. 3C). Glands were still formed and expressedonly SwG and not HF markers, without anytrace of early HF marker LHX2 in NOGGIN-expressing cells. Together, these results extendearlier findings of Plikus et al. (14) but add threekey new points: First, BMP activation in backskin or BMP inhibition in foot skin potently af-fects SwG versus HF fate determination. Sec-ond, the impact of BMP signaling on SwG fatechoice occurs exclusively within a narrow tem-poral window when ventral foot skin buds appear.Third, the BMP-mediated changes in the epithe-lium precede those in mesenchyme, underscor-ing the importance of mesenchymal-epithelialcross-talk and offering molecular insights intothe classical tissue recombination experiments.

BMP signaling enhances mesenchymalbut dampens epithelial Wnt signaling

Canonical Wnt signaling, mediated through itseffector b-catenin and its transcriptional coac-tivator lymphoid enhancer–binding factor 1 (LEF1),is essential for placode formation irrespectiveof appendage type (15–18). Indeed, epidermal-specific loss of b-catenin abrogated sweat budformation (fig. S2A) (18). Cross-talkat theepidermal-dermal border involving Wnt signaling is alsowell established (19). However, we were curiousto know how the subsequent elevation in BMPsignaling throughout the ventral foot skin wouldaffect this cross-talk; although BMP inhibition isknown to elevateWnt signaling and stabilize LEF1in hair buds (20, 21), back skin explant studieshave suggested a positive role for BMP signalinginmesenchymalLef1 expression (22). If these priorstudies are functionally relevant, the heightenedBMP signaling throughout the ventral foot skinshould skew theWnt cross-talk at the epidermal-dermal border, elevating mesenchymal Wnt sig-naling and dampening epithelial Wnt signalingdownstream from placode formation.Indeed, during the bud-permissive phase of

ventral foot skin, WNT reporter (Axin2-LacZ) ac-tivity was markedly stronger in the mesenchymethan in epithelium (Fig. 4A). Closer inspectionrevealed its presence in the epithelial placodes,but as buds developed, it became even weaker

aah6102-4 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

0%

25%

50%

75%

100%

Neg Pos

GFP+ Lhx2+

GFP- Lhx2+

GFP+ Lhx2-

GFP- Lhx2- SG fate

TRE-Nog

)%( egatnecre

P

100

75

50

25

0

0%

25%

50%

75%

100%

Neg Pos

GFP+ Lhx2+

GFP- Lhx2+

GFP+ Lhx2-

GFP- Lhx2-

TRE-Nog

HF fate

SG fate

)%( egatnecre

P

100

75

50

25

0

E9.5 E16.5 E18.5 P0

TRE-Nog + Doxy

(placode)

E9.5 E18.5 P2 P4P0

+ Doxy

P0 (NOG 3d a/E16.5)P0 WT Epi

K18

DerEpi

Der

dap toof lartneV

SwG fate unaltered

K18+

Foot skin

P4 (NOG 2d a/P2)P4 WT

K14rtTA+

TRE-Nog

TRE-Nog-H2BGFP K18 ITG β4

K14rtTA+ (placode)

dap toof lartneV

P4 (NOG 2d a/P2)P4 WT

E9.5 E16.5 E18.5 P5

+ Doxy

(placode)

LHX2

P5 (NOG 1d a/E16.5)P5 WT

TRE-Nog

LHX2DerHF fate

Foot skin

K14rtTA+

dap toof lartneV

dap toof lartneV

TRE-Nog-H2BGFP LHX2 ITG β4

ITG β4TRE-Nog-H2BGFP LHX2

ITG β4TRE-Nog-H2BGFP LHX2

LHX2DerHF fate

Foot skin

n = 197

n = 252

Fig. 3. Ectopic inhibition of BMP acts during a critical window to suppress SwG fate in the footskin. (A to C) Timeline of LV injection, NOGGIN induction, and analyses. (A) Immunofluorescenceimages and quantification of LHX2+ and LHX2– cells in the LV-infected and uninfected epidermis fromthree different foot pads (cell number as indicated). When BMP is inhibited during early stages ofepithelial bud formation, a switch from SwG to HF fate occurs, reflected by the appearance of LHX2 inthe epithelium. (B) Immunofluorescence images of LHX2. A brief pulse of NOGGIN expression at theplacode stage is sufficient to cause a fate change that lasts postnatally. (C) Immunofluorescence imagesand quantification of LHX2+ and LHX2– cells in the LV-infected and uninfected epidermis from threedifferent foot pads (cell number as indicated). Inhibiting BMP after the placode stage fails to elicit a fatechange, as reflected by the sustained expression of SwG luminal cell marker K18 in the NOGGIN+ ventralfoot skin. Lentiviral injections were performed at least two times, and two to four animals were collectedand analyzed at each time point. Gray solid lines mark the skin surface. Dashed lines indicate dermo-epidermal border. Boxed area is shown in higher magnification at right. Scale bars, 50 mm and (bottomfar right) 25 mm.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-5

0

0.5

1

1.5

2

2.5

Bmpr

1a

Fgfr2

OE-Fgf18 neg

OE-Fgf18 pos

P0 dorsal head epi

Epidermis

***

*****

*****

Ventralfoot

Dorsalback

Up in Axin2+

2939 219 1419

Down in Axin2+

4044 264 862

)eulav p( 01gol -Transcriptome analyses

Ventralfoot

Dorsalback

WNT-mediated deviations

-10 0 10log2 fold change

0

3

2

1

4

P = 2.75e-16

P = 2.56e-37Dorsalback

Ventralfoot

K

5 β

Gal

Ventral foot skinE17.5

Axin2-LacZ

0

5

10

15

20

25

30

Fgf1

Fgf2

Fgf4

Fgf7

Fgf10

Fgf11

Fgf12

Fgf13

Fgf18

Fgf20

Fgf21

Fgf22

E14.5 back_Axin2 -

E14.5 back_Axin2+

E17.5 foot_Axin2 -

E17.5 foot_Axin2+

All genesUp in Axin2+

Dermis

Rela

tive

expr

essio

n

** *****

+ FGFs

E17.5 Foot skin

Epi

Der

RNA extraction

and qPCR

ex vivo culture

Sweat bud

E14.5Dorsal back skin

Hair bud

βGal βGal

K5

β G

al

Der

Epi

E17.5

Der

Epi

Axin2-LacZ

Dermal Expression

0

4

8

12

16

E17 paw epi

E17 paw der

E14 back epi

E14 back der

Axin2 Expression

Rel

ativ

e ex

pres

sion

E ED D

E14.5back

E17.5foot

FP

KM

Rela

tive

expr

essio

n

E9.5 E16.5 P28

TRE-Fgf18 + Doxy

P0

Fo

ld c

hang

e

K14rtTA neg

K14rtTA pos K18

P28 (FGF18 a/E16.5)

K17

Dorsal skin

K17 lossHF fate

K18+

Glandular fate

niks kcab lasroD

K14rtTA+

TRE-Fgf18-H2BGFP DAPI

*******

Fig. 4. WNT responsiveness is elevated in mesenchyme and damp-ened in epithelium of BMP-high ventral foot skin, leading to dermaland epidermal changes in fate determinants of SwG-permissive skin.(A and B) Axin2-LacZ (WNT) reporter expression reveals a sharpenedepithelial-mesenchymal WNTdifferential in ventral foot skin (E17.5) (A), com-pared with dorsal back skin (E14.5 and E17.5) (B). Boxed areas highlightmarked differences in Wnt signaling between sweat and hair buds (fromE17.5 foot skin and back skin, respectively). Epi, epidermis; Der, dermis. Dashedlines indicate dermo-epidermal border. Scale bars, 50 mm. (C) QuantitativePCR for Axin2 expression in the epidermis (E) and dermis (D) of E17.5 footskin and E14.5 back skin, confirming the sharpened epithelial-mesenchymalWNT differential in ventral foot skin (E17.5), compared with dorsal backskin (E14.5). Dermal values are normalized to the corresponding epider-mis. (D) Venn diagrams summarizing transcriptome analyses of genes up-regulated or down-regulated in Axin2-LacZ–positive ventral compared withdorsal dermis. (E) Volcano plot showing all transcripts that are log2-foldchanged in dorsal versus ventral dermis. Transcripts greater than fivefoldchanged between Axin2-positive and Axin2-negative cells of each dermal

population are labeled in red. WNT-responsive genes are generally morehighly expressed in ventral foot skin relative to dorsal back skin dermis.(F) Expression of FGF ligand family member transcripts in WNT-respondingand -nonresponding fibroblasts from dorsal and ventral dermis at their respec-tive embryonic stages when epidermal placodes emerge. Shown are strong up-regulation of Fgf2, Fgf7, and Fgf18 in Axin2-LacZ–positive versus –negative footdermis. (G) Ex vivo effects of FGFs on E17.5 foot skin. (Left) After 16 to 20hours of treatment, dermis and epidermis were manually separated andanalyzed by means of quantitative PCR (middle and right, respectively) fortranscripts indicated. Shown is a FGF-mediated Bmp elevation in dermis;Bmpr and Fgfr2 in epidermis. n = 6 to 8 ventral foot skin explants from threeembryos. (H) Timeline of lentiviral injection,TRE-FGF18 induction, and analysesof dorsal skin. Quantitative PCR shows FGF18-induced up-regulation of Bmpr1aand Fgfr2 in vivo. Hair loss is visible in the highly transduced head skin as aresult of ectopic FGF18 expression. (Right) Immunofluorescence images anddiagram showing loss of HF marker K17 and gain of glandular marker K18upon FGF18 induction. Dashed lines indicate dermo-epidermal border. Scalebars, 50 mm.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

(Fig. 4A) and only reappeared in sweat epithe-lium when sweat ducts were formed (fig. S2B).In contrast, during the bud-permissive phaseof dorsal back skin, WNT reporter activity waspresent in both mesenchyme and epithelium(Fig. 4B) (23). Similarly, nuclear LEF1, a keycanonical Wnt effector, showed nuclear LEF1transiently in SwG buds and diminished LEF1in down-growths of ventral foot skin, whereasin dorsal back skin, LEF1 was high in the epi-thelial buds and sustained at the leading frontof the down-growths, where it was seen in boththe epithelium and the DP (fig. S3A). Quantifi-cation by means of quantitative PCR confirmedthat the Wnt gradient was sharper in the ven-tral foot skin than in the dorsal back skin duringthe bud-permissive phase (Fig. 4C).During the time when ventral foot skin was

permissive for sweat bud development, the ele-vation in BMP and Wnt signaling was accom-panied by other features that distinguished thismesenchyme from its dorsal counterpart. Reflect-ing the elevation in BMP signaling at this time,alkaline phosphatase exhibited a broad mesen-chymal pattern similar to LEF1 (fig. S3B). LHX2,which in back skin is restricted to the WNT-high

cells of the hair bud (10), was also mesenchy-mally expressed in permissive ventral foot skindermis (fig. S3C). Postnatally, as the window ofsweat bud fate specification waned, so too didall of these features (Fig. 1G and fig. S3).

Elevated Wnt signaling throughBMP signaling contributes todivergent mesenchymes

Probing further into how differences in mes-enchymal Wnt signaling at the epidermal-dermalborder regionally influence the transcriptionallandscape of body-site mesenchymes that sup-port distinct bud fates, we transcriptionally pro-filed the Axin2-positive and Axin2-negative dermalcells after their fluorescence-activated cell sort-ing (FACS) purification from E17.5 ventral footskin and E14.5 dorsal back skin (fig. S4). RNA-seq analyses revealed a small percentage (≤10%)of WNT-responsive genes that are shared be-tween glandular and follicular-competent mes-enchyme (Fig. 4D). Volcano plot analysis furtherrevealed that transcripts greater than fivefoldup-regulated in the Axin2-positive versus Axin2-negative dermal subfractions were distributedaway from the midline (Fig. 4E), suggesting that

elevated WNT activation through BMP signalingcontributes to divergence between glandular andfollicular-competent mesenchymes.Axin2-reporter and Gene Ontology (GO)–term

analyses revealed that although Wnt pathwaygenes were enriched in WNT-receiving cells ofboth foot and back skin dermis, a number ofdifferences emerged (fig. S5A).Wnt5a andWnt10atranscriptsmorehighly expressed in foot skin thanback skin dermal cells (fig. S5B):WNT5A is knownto induce Krt9 (24), encoding a signature keratinof palmoplantar epidermis, and WNT10A is theonly known Wnt ligand whose mutations havebeen implicated in humanHED (2, 25). However,among the family of BMP ligands, none showedchanges between the Axin2+ and Axin2− popu-lation (fig. S5C), which is consistent with thenotion that BMP signaling is upstream of Wntsignaling in skin mesenchyme (22).

Wnt signaling elicits elevated FGFsignaling in foot skin mesenchyme

There was a difference in mRNAs encoding fibro-blast growth factor (FGF) ligands and other Ras–mitogen-activatedproteinkinase (MAPK) pathwaymembers. In particular, Fgf2, Fgf7, and Fgf18 were

aah6102-6 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

+ Doxy

Dorsal back Epidermal En1

0

2

4

6

8

10

E14 E15 E16 E17

Dorsal back

Ventral trunk

Ventral foot

E9.5

TRE-En1

P0

K17

K18

niks kcab lasroD

P21 (En1 a/E16.5)

E9.5 E16.5 P21

TRE-En1 + Doxy

P0niks kcab lasro

D

E18.5 (En1 a/E9.5)

TRE-En1-H2BGFP AP DAPI TRE-En1-H2BGFP DAPI

K14rtTA+ K14rtTA+

AP

4β

Rel

ativ

e ex

pres

sion

Rel

ativ

e ex

pres

sion

Epidermal En1

ND ND

Dorsal skin

Defective DP

K18+Glandular

fate

***

****

***

Fig. 5. Engrailed-1 is expressed during SwG fate specification in ventralfoot skin and interferes with DP formation when ectopically expressedin dorsal back skin. (A) En1 quantitative PCR of dorsal back, ventral trunk,and ventral foot epidermal mRNAs during placode-permissive stages. (B) En1is elevated in dorsal back skin epidermis of ectopic Fgf18-, Bmp5-, and Bmp4-expressing embryos after 3 days of doxy-induction. (C) Timeline of lentiviral

injection, TRE-En1 induction, and analyses of dorsal back skin after ectopicexpression. Immunofluorescence images for AP (alkaline phosphatase, amarker for DP) and K18 (a marker for glandular fate).There was diminishedDP and ectopic induction of K18 in En1-expressing epidermal down-growths ofdorsal back skin. Dashed lines indicate dermo-epidermal border. Scale bars,50 mm.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

morehighly transcribed inWnt-responsive ventralfoot skin dermis than in any other dermal fractionexamined (Fig. 4F and fig. S5A). We thereforeturned to addressing whether FGFs might con-tribute to sweat fate specification.To assess FGF functional relevance, we ex-

posed E17.5 foot skin explants for 16 to 20 hourswith each of these FGFs and then separateddermis and epithelium and analyzed the tran-scriptional consequences. FGFs showed little or

minor effect on the dermal expression of BMPligands or BMP receptor (Fig. 4G, middle). Incontrast, their effects on epidermal expressionof Bmpr1a and Fgfr2 were appreciable (Fig. 4G,right). Although these FGFs are known to ac-tivate different receptors and elicit distinct mo-lecular effects in different contexts, their actionon these epidermal receptors was similar. To-gether, these findings suggest that FGFs areselectively up-regulated in BMP-hi and Wnt-

receiving foot skin mesenchyme at the dermo-epidermal border and act directly on the overlyingepithelium to sensitize it to paracrine-generatedBMPs and FGFs. Indeed, when we transducedembryos with TRE-Fgf18 and activated FGF18expression during the HF-permissive phase ofdorsal skin,Bmpr1a and Fgfr2were up-regulated(Fig. 4H). In addition, like BMP5 and BMP4,ectopic FGF18 caused hair loss in dorsal headskin, suppressed expression of HF marker K17,

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-7

K18

TRE-Shh-H2BGFP LHX2

P0 (SHH 2d a/E16.5) P15 WT P15 (SHH 10d a/P5)

TRE-Shh-H2BGFP K18 DAPI

E9.5 E16.5 E18.5 P0

LV-Shh + Doxy

ShhTRE H2BGFPPGK

(placode)E9.5 E16.5 P15P5

LV-Shh

P0(placode)

+ Doxy

K14rtTA+

DAPI

P0 WT

WT Bmpr1a cKO

Shh in situ hybridization

+NOG + BMPs + FGFs

E17.5 Foot skin

Epi

Der

P1 ventral foot pad

Rel

ativ

e ex

pres

sion

ex vivo culture

qPCRanalysis

dap toof lartneV

Epi

Der

Epi

Der

LHX2

SwG fateunaltered

K18, AQP5, NKA+LHX2--

dap toof lartneV

P84 WT P84 (SHH 2d a/E16.5)

LHX2

E16.5 E18.5 P84LV-Shh

+ Doxy

LHX2+ K17 hi

Der

HF fate

β4

ITG β4 ITG β4TRE-Shh-H2BGFP K17 TRE-Shh-H2BGFP LHX2

short exposure to SHH

β4 β4

Foot skin

Foot skin

Foot skin

E17.5 ventral foot dermis (FPKM) 1e+01 1e+051e+03

simred kcab lasrod 5.41

E)

MK

PF(

1e

+01

1e+0

31e

+05

1e+0

7Transcriptome analysis

Ptch1Gli1

(placode)E9.5K14rtTA+

K14rtTA+

P84 (SHH 2d a/E16.5)

Epidermal Shhafter treatment

dap toof lartneV

HF fate

Fig. 6. SHH signaling is suppressed in SwG buds and upon ectopic in-duction and is only effective within a critical window in switching sweatto hair fate. (A) Scatter plot of RNA-seq transcriptome analyses from E14.5dorsal back skin and E17.5 ventral foot skin dermis. Targets of SHH signaling,Gli1 and Ptch1, are enriched in dorsal versus ventral dermis (pink dots). (B) Shhin situ hybridizations of ventral foot pads of WT and Bmpr1a cKO P1 pups,revealing the role of BMP in suppressing Shh in SwG-buds. (C) Ex vivo cultureof E17.5 foot skin with addition of BMP-inhibitor, NOGGIN, BMPs, or FGF18.Epidermal Shh expression was analyzed by means of quantitative PCR. n = 6to 8 foot skin samples analyzed from three embryos per experiment. (D) Timelineof lentiviral injection, Shh induction, and analyses. Immunofluorescence imagesshowing that when induction occurs at the epithelial bud stage, hair bud marker

LHX2 is ectopically induced in the foot pad epidermis, which is reflective ofa switch from SwG to HF fate. Dashed line indicates dermo-epidermalborder. Scale bar, 50 mm. (E) Timeline of lentiviral injection, Shh induction,and analyses. Immunofluorescence images showing that when inductionoccurs after the epithelial bud stage, appendage fate is unaltered in theventral foot skin. Boxed area is shown in higher magnification at right.Scale bar, 50 mm and (far right) 25 mm. (F) Timeline of lentiviral injection,Shh induction, and analyses. Immunofluorescence images showing thatshort exposure to SHH during foot skin epithelial bud formation is suf-ficient to cause long-lasting effects on fate change and expression of HFmarkers K17 and LHX2. Dashed lines indicate dermo-epidermal border. Scalebars, 50 mm.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

and induced expression of glandularmarker K18,altogether acting toward glandular fate.

Expression of Engrailed during SwGfate specification: Parallels to flywing development

The BMP-induced sharpened differential thatwe observed in Wnt signaling at the dermo-epidermal border of permissive ventral foot skinwas reminiscent of Drosophila wing develop-ment, in which strong Wnt signaling cells arejuxtapositioned with low-Wnt signaling cells, acondition that in turn was necessary to sustainexpression of the transcription repressor Engrailedspecifically in the low-Wnt-signaling cells (26).We found that En1was highly expressed in mouse

ventral epithelium during the SwG fate–permissivestage of foot skin but not in the dorsal epitheliumduring the HF fate–permissive stage of back skin(Fig. 5A). En1 was also expressed in early ventraltrunk (belly) skin, and its expression waned af-ter E14.5 (Fig. 5A). This timing coincided withthe completion of mammary gland fate specifi-cation (E13.5) and was just before hair folliclefate specification (E15.5) in the belly skin.These data point to a complex circuitry trig-

gered by BMP elevation in ventral foot skin mes-enchyme that enhances differential Wnt signalingto reach a permissive threshold to result in En1expression. In agreement with this hypothesis,TRE-Bmp5 and TRE-Bmp4 activation resultedin a dramatic up-regulation of En1 in dorsal skin

epidermis, as did TRE-Fgf18 (Fig. 5B). Moreover,when En1 was ectopically induced, it resulted intransformations consistentwith amore glandularfate, including diminished DP and K18 induction(Fig. 5C). Together, these data suggest a rolefor epidermal ENGRAILED-1 in promoting theSwG fate; we found this noteworthy becausemice lacking Engrailed-1 dorsalize the ventralskin (27), and Engrailed-1 heterozygote mutantmice display fewer SwGs in their ventral footskin (28).

All roads lead to suppression of SHHsignaling in SwG-permissive skin

Studies on the HF have emphasized the role ofepithelial sonic hedgehog (SHH), downstream

aah6102-8 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

20wk

LHX2 ITG α6 DAPI K18

*

0

1

2

3

4

5

BMP2 BMP4 BMP5 NOG FGF2 FGF7 FGF18

15wk 17wk

*

****

***

ns

***

***

*

0

5

10

15

20

25

BMP2 BMP4 BMP5 NOG FGF2 FGF7 FGF18

15wk scalp

15wk palm

15wk vs 17wk scalp dermis

15wk

LHX2 K5 DAPI

17wk

LHX2 K5 DAPI

*

LHX2

SHH BMP

NOG FGF FGFFGF

E17.5 Ventral foot skin

E14.5 Dorsal

back skin

Hair Follicle Sweat gland

0

2

4

6

8

SHH ID1 NKA EN1

15wk

17wk

20wk

15wk - 17wk - 20wk scalp epidermal buds

***

***

****

****

****

****

ns ns

**

****

***ns

ns**

**

15wk dermis

Re

lative

exp

ressio

n

Re

lative

exp

ressio

n

Rela

tive e

xpre

ssio

n

20w

k sc

alp

20w

k palm

Alkaline Phosphatase

Rel

ativ

e ex

pres

sion

15wk epidermal buds

***

** ***********

****

Fig. 7. Human dorsal skin temporally shifts from BMPloFGFloSHHhi toBMPhiFGFhiSHHlo during embryogenesis, yielding HF and SwG coexis-tence. (A) Model showing differences in the balance between BMP/FGF andSHH signaling specify the fate of HFs or SwGs. (B) Immunofluorescenceimages of human fetal dorsal scalp skin at 15, 17, and 20 weeks of gestation.Boxed area shows emerging LHX2+ HF buds. Stars indicate LHX2neg placodeand bud. HFs and SwG buds are distinguished by hair bud–specific markerLHX2. (Far right) Appearance ofK18-positivematuring sweat gland by 20weeks.Scale bars, 100 mm. (C) Histology of dorsal scalp skin and ventral palm skin,costained for alkaline phosphatase activity to identify dermal papillae (pinkarrows), absent in sweat buds and glands (blue arrows).Visible is coexistence

of glands and HFs in dorsal scalp skin but only SwGs in palm skin. Scale bars,200 mm. (D to G) Quantitative PCR of transcripts isolated from dermis andepithelial buds of scalp skin and palm skin, at ages indicated. (D) Increase inBMP and FGF transcripts in scalp dermis between 15 and 17 weeks. (E) Highlevel of BMP5 expression in 15-week ventral palm (SwG only) versus 15-weekdorsal scalp (HF only) dermis. (F) Early decline of SHH production and laterincrease in BMP signaling (ID1) and SwGmarkers (NKA) in emerging epithelialbuds between 15 and 20 weeks. Also shown is the emergence of EN1 in dorsalscalp skin, which contrasts with mice, where EN1 is confined to ventral footskin. (G) At 15 weeks, SwGmarkers are already high in palm but not in scalpepidermis.

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

of Wnt signaling, in stimulating mesenchymalclustering to form the Noggin-expressing DP(29, 30). Because sweat buds display reducedepithelial Wnt signaling and also lack this mes-enchymal cluster, we posited that in sweat buds,Shh is silenced, which in turn should be reflectedin the failure of foot skin dermis to activatedownstream SHH-signaling target genes. OurRNA-seq analyses showed that indeed, Gli1 andPtch1, the classic downstream target genes ofSHH signaling, are highly enriched in dorsalback skin (HF-competent) versus ventral footskin (SwG-competent) dermis (Fig. 6A). Quan-titative PCR confirmed the strong expressionof Shh in E14.5 dorsal epidermis relative to ven-tral foot skin epidermis (fig. S6A).Because Bmpr1a-null foot skin buds displayed

signs ofmesenchymal cell clustering (Fig. 1G), wehypothesized that Shh expression must rely notonly on epithelial Wnt signaling (21, 30) but alsoon inhibition of mesenchymal BMP signaling tothe epithelial buds. In other words, when mes-enchymal BMPs actively signal to the epithelialbuds, as in the case of ventral foot skin, Shh ex-pression must be suppressed. In situ hybrid-izations and quantitative PCR showed that Shh,although not detected in wild-type ventral footskin, was indeed activated in Bmpr1a-null ven-tral foot skin buds in vivo, as well as in E17.5 footskin explants exposed to NOGGIN (Fig. 6, B andC). In contrast, exposure of foot skin explants toadditional BMPs further suppressed already lowexpression of Shh (Fig. 6C). Furthermore, elevat-ing NOGGIN in dorsal back skin resulted inhigher activation of Shh, as well as HF markersLhx2 and K17 (fig. S6B).Shh repression in ventral foot skin buds seemed

also to be influenced by FGF-MAPK signaling,perhaps through its ability to sensitize buds toBMP signaling (Figs. 4G and 6C). In particular,FGF18 seemed to have the strongest effect on Shhrepression as shown in explants and in vivowhenactivated in dorsal back skin (Fig. 6C and fig. S6,C andD). The combinatorial effect of FGFs in Shhsuppression could be partially rescued by MAPKinhibitor SB203580 (fig. S6D). These results cor-relatedwith GO-term analyses ofWNT-responsivechanges, revealing that theMAPK pathway is pos-itively enriched in ventral foot skin but negativelyenriched in dorsal back skin (fig. S5A).Our data on FGFs further bolstered the im-

portance of mesenchymal-to-epithelial signalingcircuits for Shh suppression in foot skin. Thatsaid, Wnt signaling is required and upstreamof Shh expression in hair buds (20, 21) and isrepressed in SwG buds (Fig. 4A and fig. S3A).Additionally, ENGRAILED has been previouslyreported to act directly to transcriptionally repressShh and/or the genes encoding its downstreamtranscriptional effectors (31–34). All of thesepathways appeared to converge on repression ofShh and SHH signaling in SwG-permissive skin.

Regional BMP:SHH antagonism dictatesSwG versus HF fates

To evaluatewhether Shh up-regulation in ventralfoot skin epithelium can prevent SwG fate, we

again turned to the inducible lentiviral system.Analogous to our experimental design forNoggin,we induced Shh for several days at E16.5. By P0,foot skin epithelial buds ectopically inducedLHX2 and other HF markers and failed to pro-gress to SwGs (Fig. 6D and fig. S6E, left), whereasin the underlying mesenchyme, signs of SHHsignaling were seen, including a rise in BMPinhibitors (fig. S6E, right). Similar to our findingswith Noggin induction, when Shh was inducedafter sweat buds were already established, glandformation progressed, as evidenced by morphol-ogy, luminal marker K18, and SwGmarkers NKAand AQUAPORIN-5 (Fig. 6E and fig. S6F). Thiswas true even after 10 days of a high level of Shhexpression (Fig. 6E), suggesting that once thenarrowwindowof permissiveness is passed, glan-dular fate is no longer sensitive to SHH.Last, we tested whether ectopic SHH adminis-

tered during the permissive window has long-term consequences on appendage choice. Weactivated epithelial Shh in embryos for only 2 daysat the sweat bud–formation stage and then stoppedDoxy treatment and examined the foot pads ofadult mice ~3 months later. Although morphol-ogy was clearly abnormal, the activation of hair-specific markers and absence of gland-specificmarkers provided compelling evidence that briefexposure to SHH during the critical window ofgland bud morphogenesis was sufficient to causeirreversible changes in developing buds, prevent-ing SwG fate and promoting features typicallyassociated with HF fate (Fig. 6F). Additionally,when we generated mouse embryos whose skincontained mosaic patches of a constitutively ac-tive SHH receptor (SmoM2) (35) in the epithelium,only wild-type foot skin patches displayed K18+

SwGs, whereas SmoM2 patches expressed LHX2+

budlike structures (fig. S7). This suggested thatselective activation of SHH signaling within theventral foot skin epithelium is sufficient to trig-ger the fate switch, regardless of dermal signal-ing feedback.

Temporal regulation of BMP:SHHantagonism endows human skin withboth SwGs and HFs

Our data are summarized in the model shownin Fig. 7A. We tested this model by addressingwhether it can be used to explain how humanskin is able to accommodate both SwGs andHFs within a common dermis. The organizedpatterning of glands and HFs within humanskin argues against the notion that its dermissimply maintains intermediate levels of FGFsand BMPs, permissive for both types of epider-mal appendages. Rather, our mouse data predictthat within embryonic human dermis, there mustbe a temporal switch in dermal permissiveness,which allows for the sequential emergence ofeach appendage type. Indeed, upon analyzinghuman fetal scalp skin at gestational ages 15,17, and 20 weeks with our markers establishedfrom mouse studies, we observed temporal seg-regation of waves of HF and SwG formation,with hair bud morphogenesis beginning to phaseout between 15 and 17 weeks and sweat buds

emerging at 17 weeks (Fig. 7, B and C). By20 weeks, sweat buds were still forming, inthe absence of further signs of de novo HFmorphogenesis.To test whether a temporal change in BMP,

FGF, and SHH signaling drives the intriguingdevelopmental switch from hair to sweat fatewithin human scalp skin, we FACS-purified epi-thelial bud and mesenchymal cells (fig. S8, Aand B) and analyzed their relative gene expres-sion at these key windows of development. With-in scalp dermis, expression of BMP2, BMP4, BMP5,FGF2, and FGF7 were all significantly up-regulatedat week 17 relative to week 15, which is coincidentwith the shift from hair to sweat bud formation(Fig. 7D). Thus, in contrast to mice, in which thesedermal BMPs and FGFs were regionally restrictedto the ventral foot skin, in humans the samegenes were now temporally restricted to estab-lish discrete hair versus sweat bud windows ofpermissiveness during embryogenesis of dorsalscalp skin.To determine whether humans maintained

evolutionary vestiges of regional skin differences,we also compared human scalp and palm skinsat these gestational stages. As shown in Fig. 7E,the sweat bud–promoting mesenchymal signalsthat were elevated temporally by week 17 in hu-man scalp skin were already present by week15 in human palm skin, which was exclusivefor SwG appendages (fig. S8C). Like in mousefoot skin, BMP5 was the most prominent BMPexpressed in palmar dermis (Fig. 7E), and LHX2was exclusively expressed in palmar dermis andscalp HF epithelium (fig. S8C).Consistent with the antagonistic regulatory

circuitry that we unveiled in mouse, SHH ex-pression decreased dramatically within the scalpepithelial buds as mesenchymal BMPs and FGFs,and epithelial ENGRAILED-1, were elevated (Fig.7F). Additional signs of the developmental switchin appendage specification were revealed by tem-poral up-regulation of BMP target gene ID1 andSwG marker NKA (Na/K-ATPase) (Fig. 7F). Thesequential appearance of the appendages wassimilar but temporally delayed in dorsal limbskin, which like scalp is typified by both hairand SwGs (fig. S8D). In contrast in human palmskin, which (like its mouse counterpart) developsexclusively SwGs, SHH was already markedlysuppressed regionally by week 15, which is co-incident with the low level of WNT-(LEF1)–signaling and high level of ENGRAILED-1 at thisearly time (Fig. 7G). Overall, our findings under-score the distinct temporal flip in BMP:SHHsignaling that occurs in human development andendows human skin with the ability to permitsweat bud fate commitment as the final wave(s)of bud development in otherwise hairy skin.

Discussion

In the evolutionary developmental biology viewof BMP signaling and fate specification of in-tegument, both chicken scales (36) and mouseand human SwGs (our study) require BMP signal-ing to specify their fate, whereas feathers (37) andHFs (13) must suppress it. Thus, the differential

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-9

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

impact of BMP signaling on appendage fate spec-ification has ancient roots and occurs repeatedlythroughout vertebrate evolution in making epi-dermal bud fate decisions. In addition to its rolein ectodermal fate specification, BMP-SHH antag-onism is also critical for the patterning of chickenendodermal digestive epithelia (38, 39), especiallyin the glandular region of the stomach. Togetherwith other findings on BMP requirement forvarious ectodermal glands (mammary and mei-bomian) (40, 41), a particular role begins to emergefor BMP signaling in promoting glandular fate,typically occurring on the ventral side of thebody. Our studies provide insights into howelevated mesenchymal BMP signaling drivesa thereafter self-perpetuating molecular cas-cade that culminates in silencing SHH signalingto suppress one appendage fate and specify an-other. By spatially regulating BMP:SHH antag-onism, mammals can specify HF and glandularfates in distinct body sites. With additional tem-poral regulation of BMP:SHH antagonism, hu-man skin has the special capacity first to generatewaves of HF fates and then to switch midstreamto specify glandular fates. Our findings pavethe way for future therapeutic advances in skinregeneration with dual epithelial appendages.

Materials and methodsMice and in utero lentiviral injection

All animals used for the experiments in thismanuscript were generated previously: K14Cre(Fuchs lab) (42), Bmpr1a fl/fl (kindly from Y.Mishina)(43), SEA/GnJ (The Jackson Laboratory) (44),K14-rtTA (Fuchs lab) (45), Smofl/fl (The JacksonLaboratory) (46), R26SmoM2 (The Jackson Lab-oratory) (47), R26 Flox-Stop-Flox-YFP (The JacksonLaboratory) (48), Ctnnbl fl/fl (kindly fromR. Kemler)(49), Axin2-LacZ (The Jackson Laboratory) (50).In utero ultrasound-guided lentiviral injectionprocedures and preparation of high-titer lenti-virus have been described (12). K14rtTA trans-genic male (CD1 strain) were mated with CD1wild type female, and their embryos were in-jected with high-titer lentivirus at E9.5. Eachlentiviral injection experiments were performedat least two times and all foot skins from 2-4animals were collected and analyzed at eachtime points. Animals were maintained in anAssociation for Assessment and Accreditation ofLaboratory Animal Care (AALAC)-accredited ani-mal facility, and procedures were performedwith Institutional Animal Care and Use Com-mittee (IACUC)-approved protocols. rtTA tran-scription factor was activated by feeding micewith doxycycline (2 mg/kg) chow at times spec-ified. Lentiviral doxycycline-inducible Shh expres-sion construct (pLKO-TRE-Shh-PGK-H2BGFP)has been previously described (30). Constructfor the lentiviral Cre transductions has been pre-viously described (12). Mouse Noggin was clonedby PCR from pMgB950 (kindly from R. Harland)to replace Shh in the pLKO-TRE-Shh-PGK-H2BGFPto generate pLKO-TRE-Nog-PGK-H2BGFP. MouseBmp5 and Fgf18 cDNAs were purchased fromSino Biological Inc. and were cloned into pLKO-TRE-PGK-H2BGFP vector.

Human tissue procurementFetal skin samples (scalp and palm) of 15, 17, 20wks (2-3 samples for each stage) were procuredby Novogenix Laboratories (Los Angeles, CA) andAdvanced Bioscience Resources (Alameda, CA)in compliance with federal and state laws, andNational Institute of Health guidelines. Tissueswereprocessedandparaffin-embeddedbyHistoWiz,Inc. Standard histology and Alkaline Phospha-tase immunohistochemistry staining were alsoperformed at HistoWiz (Brooklyn, NY).

Quantification of sweat ducts

Experimental details have been described pre-viously (3). Briefly, mouse hind foot skins wereremoved by dissection and incubated in EDTA(50 mM) 37°C for 30 min to separate epidermisand dermis. Under these conditions, sweatducts remain associated with the epidermalcell sheet. Whole mount histochemical stain-ing was with 0/1% Nile Blue A (Sigma). Afterwashing, sweat ducts showed intense blue stain-ing and were scored and imaged using a LeicaM165 FC stereomicroscope.

In situ hybridization

In situ hybridization for Shh was performed asdescribed previously (19).

Immunofluorescence and imaging

Skin (dorsal back and ventral foot) samplesfrom 3–4 animals for each experiment weredissected and embedded in OCT (Tissue Tek)and cryo-sectioned at a thickness of 10–12 mm.Tissue sections were fixed in 4% PFA and thenblocked in gelatin block (1% gelatin, 2.5% nor-mal donkey serum, 1% BSA, 0.3% Triton in PBS)1 hour at RT. Primary antibodies were incu-bated at 4°C overnight. After washing with PBS,samples were incubated for 2 hours at roomtemperature with secondary antibodies conjugatedwith Alexa488, RRX, or Alexa647 (1:800, 1:400,and 1:400, respectively, Life Technologies) and4′6-diamidino-2-phenylindole (DAPI). Sampleswere washed again with PBS and then mountedin Prolong Gold (Invitrogen). EdU incorporationwas detected by Click-It EdU AlexaFluor594Imaging Kit (Invitrogen) per manufacturer’sinstructions. The following primary antibodieswere used: P-Cadherin (Rat, 1:50, Fuchs lab),E-Cadherin (Rat, 1:100, Fuchs lab), phospho-Smad1/5/8 (rabbit, 1:800, Cell signaling), LEF1(rabbit, 1:300, Fuchs Lab), K5 (rabbit, 1:600,Fuchs Lab), K5 (guinea pig, 1:500, Fuchs Lab), K17(rabbit, 1:600, Fuchs Lab), K18 (rabbit, 1:400,Fuchs Lab), LHX2 (rabbit, 1:2000, Fuchs Lab),anti-GFP/YFP (chicken, 1:2000, Abcam), ITGa6(Rat, 1:100, BD), ITGb4 (Rat, 1:100, BD), Alkalinephosphatase (Goat, 1:100, R&D), b-Galactosidase(chicken, 1:2000, Abcam), NKA (Rabbit, 1:8000,Abcam), AQP5 (rabbit, 1:2000, Abcam), EpCam-Alexa488 (mouse, 1:100, Cell signaling). Epi-fluorescence images were acquired with anAxioOberver.Z1 microscope equipped with aHamamatsu ORCA-ER camera (HamamatsuPhotonics), and with an ApoTome.2 (Carl Zeiss)slider that reduces the light scatter in the flu-

orescent samples. 20× and 40× objectives wereused. Data collection was controlled by Zensoftware (Carl Zeiss).

Tissue preparation and fluorescence-activated cell sorting

Mouse E14.5 dorsal back skin and E17.5 ventralfoot skin of Axin2-LacZ reporter mice were dis-sected and incubated in EDTA (50mM), 37°C for30 min to allow separation of epidermis and der-mis. Note that for collecting E17.5 foot skin, onlyfront feet were used, to avoid potential contami-nation of HF-competent dermal cells in theinter-pad region of the hind feet. Since these areskin tissues harvested at the placode formationstage, epidermal placodes were mostly removedwith epidermal fraction and minimum epider-mal cells remained with dermal fraction. Thedermal tissues were then digested in HanksBalanced Salt Solution (Lonza) with 1000 U/mlcollagenase and 300U/ml hyaluronadase (Sigma)and washed and re-suspended in PBS with 4%FBS tomake single cell suspension. The followingantibodies were used for fluorescence-activatedcell sorting (FACS): a6-PE (CD49f, 1:1000, eBio-science), CD140a-APC (1:100, eBioscience). Stain-ing of intracellular b-galactosidasewas performedusing The FluoReporter lacZ Flow Cytometry Kit(Molecular Probes) as permanufacturer’s instruc-tions. DAPI was used to exclude dead cells. Cellisolations were performed on FACSAria sortersequipped with DIVA software (BD Biosciences).Human skin tissuesweredigested inLiberaseTL

(Roche), 37°C for 1 hour 20 min. Digested tissueswere resuspended andwashed inPBSwith4%FBSto make single cell suspensions. The followingantibodies were used for FACS: CD140a-PE (1:50,BioLegnad), EpCAM-Alexa488 (1:100, Cell signal-ing), CD49f-PECy7(1:1000, eBioscience), CD133-APC (1:100, Miltenyi). DAPI was used to excludedead cells. Cell isolations were performed onFACSAria sorters equipped with DIVA software(BD Biosciences).

Explant culture and RNA extraction andRT-qPCR

Six to eight pieces of E17.5 foot skin from >3different mouse embryos per condition wereplaced as explants into culture medium. Theculture conditions were modified from thehanging-drop method (51), and 100–120 ml ofculture medium were used for each drop. Thefinal concentrations of recombinant proteinsand inhibitors used were: bFGF (50 ng/ml), FGF7(50 ng/ml), FGF18 (50 ng/ml), BMP4 (100 ng/ml),BMP5 (100 ng/ml), Noggin (4 mg/ml), MEKi(U0126, 10 mM), MAPKi (SB203580, 1 mM). Afterovernight culture (16–18 hour), skin tissues wereincubated in EDTA (50 mM), 37°C for 20 min,and epidermis and dermis were separated andimmediately frozen in liquid nitrogen. For em-bryonic foot skin tissues from lentiviral injec-tion experiments to be analyzed by RT-qPCR,note that only front foot skins were collectedbecause of higher infection efficiency comparedto hind foot skin, and at least 6–8 pieces from3–4 embryos were collected for each experiment.

aah6102-10 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

For embryonic dorsal skin from lentiviral in-jection experiments, samples from head region(highest infection efficiency) were collected fromat least two different embryos. All these sampleswere similarly incubated in EDTA (50 mM), 37°Cfor 20 min, to separate epidermis/dermis andwere immediately frozen. The frozen tissues werehomogenized using Bessman Tissue Pulverizer(SpectrumTM) and collected in Trizol (Invitro-gen). RNA was extracted with a Direct-zol RNAMiniPrep kit (Zymo Research) as per manufac-turer’s instructions. Equivalent amounts of RNAwere reverse-transcribed by SuperScript VILOcDNA Synthesis Kit (Invitrogen). cDNAs weremixed with indicated primers and Power SYBRGreen PCR Master Mix (AppliedBiosystems), andquantitative PCR (qPCR) was performed on aApplied Biosystems 7900HT Fast Real-Time PCRsystem. cDNAs were normalized to equal amountsusing primers against Hprt or Ppib. Primer se-quences for RT-PCR were obtained using NCBIPrimer-BLAST webtool.

RNA-seq analyses and statistics

For RNA-seq, RNA quality was determined usingAgilent 2100 Bioanalyzer. Samples used for se-quencing had RNA integrity numbers (RIN) > 8.Library preparation using Illumina TrueSeqmRNA sample preparation kit was performedat the Weill Cornell Medical College GenomicCore facility, andRNAswere sequencedon IlluminaHiSeq 2000machines. RNA reads were alignedto the mm9 build of the mouse genome usingTophat. Transcript assembly and differential ex-pression was done using Cufflinks with RefseqmRNAs to guide assembly (52). Analysis of RNA-seq data was performed using the cummeRbundpackage in R (53). DAVID bioinformatic databasewas used to find enriched functional GO-termannotations in RNA-seq data (https://david.ncifcrf.gov/content.jsp?file=citation.htm). Box-and-whisker plots were generated using Prism5 soft-ware (GraphPad) to determine the significancebetween two groups. Error bars plotted on graphsdenote SD. Unpaired student’s t test was used todetermine the significance between two groups:*P< 0.05; **P< 0.01; ***P< 0.001; ****P<0.0001;ns, not significant.

REFERENCES AND NOTES

1. J. Kere et al., X-linked anhidrotic (hypohidrotic) ectodermaldysplasia is caused by mutation in a novel transmembraneprotein. Nat. Genet. 13, 409–416 (1996). doi: 10.1038/ng0895-409; pmid: 8696334

2. C. Cluzeau et al., Only four genes (EDA1, EDAR, EDARADD, andWNT10A) account for 90% of hypohidrotic/anhidroticectodermal dysplasia cases. Hum. Mutat. 32, 70–72 (2011).doi: 10.1002/humu.21384; pmid: 20979233

3. Y. G. Kamberov et al., Modeling recent human evolution in miceby expression of a selected EDAR variant. Cell 152, 691–702(2013). doi: 10.1016/j.cell.2013.01.016; pmid: 23415220

4. T. Mustonen et al., Stimulation of ectodermal organdevelopment by Ectodysplasin-A1. Dev. Biol. 259, 123–136(2003). doi: 10.1016/S0012-1606(03)00157-X; pmid: 12812793

5. P. Sengel, Pattern formation in skin development. Int. J. Dev.Biol. 34, 33–50 (1990). pmid: 2203463

6. C. M. Chuong, R. Chodankar, R. B. Widelitz, T. X. Jiang,Evo-devo of feathers and scales: Building complex epithelialappendages. Curr. Opin. Genet. Dev. 10, 449–456 (2000).doi: 10.1016/S0959-437X(00)00111-8; pmid: 11023302

7. K. Kratochwil, Organ specificity in mesenchymal inductiondemonstrated in the embryonic development of the mammarygland of the mouse. Dev. Biol. 20, 46–71 (1969). doi: 10.1016/0012-1606(69)90004-9; pmid: 5795848

8. D. Dhouailly, G. E. Rogers, P. Sengel, The specification offeather and scale protein synthesis in epidermal-dermalrecombinations. Dev. Biol. 65, 58–68 (1978). doi: 10.1016/0012-1606(78)90179-3; pmid: 680361

9. C. Ferraris, G. Chevalier, B. Favier, C. A. Jahoda, D. Dhouailly,Adult corneal epithelium basal cells possess the capacity toactivate epidermal, pilosebaceous and sweat gland geneticprograms in response to embryonic dermal stimuli.Development 127, 5487–5495 (2000). pmid: 11076768

10. H. Rhee, L. Polak, E. Fuchs, Lhx2 maintains stem cell characterin hair follicles. Science 312, 1946–1949 (2006). doi: 10.1126/science.1128004; pmid: 16809539

11. M. Rendl, L. Polak, E. Fuchs, BMP signaling in dermal papillacells is required for their hair follicle-inductive properties.Genes Dev. 22, 543–557 (2008). doi: 10.1101/gad.1614408;pmid: 18281466

12. S. Beronja, G. Livshits, S. Williams, E. Fuchs, Rapid functionaldissection of genetic networks via tissue-specific transductionand RNAi in mouse embryos. Nat. Med. 16, 821–827 (2010).doi: 10.1038/nm.2167; pmid: 20526348

13. V. A. Botchkarev et al., Noggin is a mesenchymally derivedstimulator of hair-follicle induction. Nat. Cell Biol. 1, 158–164(1999). doi: 10.1038/11078; pmid: 10559902

14. M. Plikus et al., Morpho-regulation of ectodermal organs:Integument pathology and phenotypic variations in K14-Nogginengineered mice through modulation of bone morphogenicprotein pathway. Am. J. Pathol. 164, 1099–1114 (2004).doi: 10.1016/S0002-9440(10)63197-5; pmid: 14982863

15. C. van Genderen et al., Development of several organs thatrequire inductive epithelial-mesenchymal interactions isimpaired in LEF-1-deficient mice. Genes Dev. 8, 2691–2703(1994). doi: 10.1101/gad.8.22.2691; pmid: 7958926

16. T. Andl, S. T. Reddy, T. Gaddapara, S. E. Millar, WNT signals arerequired for the initiation of hair follicle development. Dev. Cell2, 643–653 (2002). doi: 10.1016/S1534-5807(02)00167-3;pmid: 12015971

17. L. Ahtiainen et al., Directional cell migration, but notproliferation, drives hair placode morphogenesis. Dev. Cell 28,588–602 (2014). doi: 10.1016/j.devcel.2014.02.003;pmid: 24636260

18. C.-Y. Cui et al., Involvement of Wnt, Eda and Shh at definedstages of sweat gland development. Development 141,3752–3760 (2014). doi: 10.1242/dev.109231;pmid: 25249463

19. R. DasGupta, E. Fuchs, Multiple roles for activated LEF/TCFtranscription complexes during hair follicle development anddifferentiation. Development 126, 4557–4568 (1999).pmid: 10498690

20. C. Jamora, R. DasGupta, P. Kocieniewski, E. Fuchs, Linksbetween signal transduction, transcription and adhesion inepithelial bud development. Nature 422, 317–322 (2003).doi: 10.1038/nature01458; pmid: 12646922

21. T. Ouspenskaia, I. Matos, A. F. Mertz, V. F. Fiore, E. Fuchs,WNT-SHH antagonism specifies and expands stem cells priorto niche formation. Cell 164, 156–169 (2016). doi: 10.1016/j.cell.2015.11.058; pmid: 26771489

22. K. Kratochwil, M. Dull, I. Farinas, J. Galceran, R. Grosschedl,Lef1 expression is activated by BMP-4 and regulates inductivetissue interactions in tooth and hair development. Genes Dev.10, 1382–1394 (1996). doi: 10.1101/gad.10.11.1382;pmid: 8647435

23. Y. Zhang et al., Reciprocal requirements for EDA/EDAR/NF-kappaB and Wnt/b-catenin signaling pathways in hair follicleinduction. Dev. Cell 17, 49–61 (2009). doi: 10.1016/j.devcel.2009.05.011; pmid: 19619491

24. J. L. Rinn et al., A dermal HOX transcriptional programregulates site-specific epidermal fate. Genes Dev. 22, 303–307(2008). doi: 10.1101/gad.1610508; pmid: 18245445

25. J. Plaisancié et al., Mutations in WNT10A are frequentlyinvolved in oligodontia associated with minor signs ofectodermal dysplasia. Am. J. Med. Genet. A. 161A, 671–678(2013). doi: 10.1002/ajmg.a.35747; pmid: 23401279

26. J. P. Vincent, P. A. Lawrence, Drosophila wingless sustainsengrailed expression only in adjoining cells: Evidence frommosaic embryos. Cell 77, 909–915 (1994). doi: 10.1016/0092-8674(94)90139-2; pmid: 8004677

27. C. A. Loomis et al., The mouse Engrailed-1 gene and ventrallimb patterning. Nature 382, 360–363 (1996). doi: 10.1038/382360a0; pmid: 8684466

28. Y. G. Kamberov et al., A genetic basis of variation in eccrinesweat gland and hair follicle density. Proc. Natl. Acad. Sci. U.S.A.112, 9932–9937 (2015). doi: 10.1073/pnas.1511680112;pmid: 26195765

29. W.-M. Woo, H. H. Zhen, A. E. Oro, Shh maintains dermal papillaidentity and hair morphogenesis via a Noggin-Shh regulatoryloop. Genes Dev. 26, 1235–1246 (2012). doi: 10.1101/gad.187401.112; pmid: 22661232

30. Y.-C. Hsu, L. Li, E. Fuchs, Transit-amplifying cells orchestratestem cell activity and tissue regeneration. Cell 157, 935–949(2014). doi: 10.1016/j.cell.2014.02.057; pmid: 24813615

31. C. Dahmann, K. Basler, Opposing transcriptional outputs ofHedgehog signaling and engrailed control compartmentalcell sorting at the Drosophila A/P boundary. Cell 100,411–422 (2000). doi: 10.1016/S0092-8674(00)80677-7;pmid: 10693758

32. F. Prin, D. Dhouailly, How and when the regional competenceof chick epidermis is established: Feathers vs. scutate andreticulate scales, a problem en route to a solution. Int. J.Dev. Biol. 48, 137–148 (2004). doi: 10.1387/ijdb.15272378;pmid: 15272378

33. K. Basler, G. Struhl, Compartment boundaries and the controlof Drosophila limb pattern by hedgehog protein. Nature 368,208–214 (1994). doi: 10.1038/368208a0; pmid: 8145818

34. C. Alexandre, J.-P. Vincent, Requirements for transcriptionalrepression and activation by Engrailed in Drosophila embryos.Development 130, 729–739 (2003). doi: 10.1242/dev.00286;pmid: 12506003

35. K. K. Youssef et al., Adult interfollicular tumour-initiating cellsare reprogrammed into an embryonic hair follicle progenitor-like fate during basal cell carcinoma initiation. Nat. Cell Biol. 14,1282–1294 (2012). doi: 10.1038/ncb2628; pmid: 23178882

36. H. Zou, L. Niswander, Requirement for BMP signaling ininterdigital apoptosis and scale formation. Science 272,738–741 (1996). doi: 10.1126/science.272.5262.738;pmid: 8614838

37. S. Noramly, B. A. Morgan, BMPs mediate lateral inhibition atsuccessive stages in feather tract development. Development125, 3775–3787 (1998). pmid: 9729486

38. D. J. Roberts, D. M. Smith, D. J. Goff, C. J. Tabin, Epithelial-mesenchymal signaling during the regionalization ofthe chick gut. Development 125, 2791–2801 (1998).pmid: 9655802

39. T. Narita et al., BMPs are necessary for stomach glandformation in the chicken embryo: A study using virally inducedBMP-2 and Noggin expression. Development 127, 981–988(2000). pmid: 10662637

40. J. A. Mayer, J. Foley, D. De La Cruz, C.-M. Chuong, R. Widelitz,Conversion of the nipple to hair-bearing epithelia bylowering bone morphogenetic protein pathway activity atthe dermal-epidermal interface. Am. J. Pathol. 173,1339–1348 (2008). doi: 10.2353/ajpath.2008.070920;pmid: 18832580

41. J. Huang et al., FGF-regulated BMP signaling is required foreyelid closure and to specify conjunctival epithelial cell fate.Development 136, 1741–1750 (2009). doi: 10.1242/dev.034082; pmid: 19369394

42. V. Vasioukhin, L. Degenstein, B. Wise, E. Fuchs, The magicaltouch: Genome targeting in epidermal stem cells induced bytamoxifen application to mouse skin. Proc. Natl. Acad. Sci. U.S.A.96, 8551–8556 (1999). doi: 10.1073/pnas.96.15.8551;pmid: 10411913

43. Y. Mishina, M. C. Hanks, S. Miura, M. D. Tallquist,R. R. Behringer, Generation of Bmpr/Alk3 conditional knockoutmice. Genesis 32, 69–72 (2002). doi: 10.1002/gene.10038;pmid: 11857780

44. J. A. King, P. C. Marker, K. J. Seung, D. M. Kingsley, BMP5 andthe molecular, skeletal, and soft-tissue alterations in short earmice. Dev. Biol. 166, 112–122 (1994). doi: 10.1006/dbio.1994.1300; pmid: 7958439

45. H. Nguyen, M. Rendl, E. Fuchs, Tcf3 governs stem cellfeatures and represses cell fate determination in skin. Cell127, 171–183 (2006). doi: 10.1016/j.cell.2006.07.036;pmid: 17018284

46. F. Long, X. M. Zhang, S. Karp, Y. Yang, A. P. McMahon, Geneticmanipulation of hedgehog signaling in the endochondralskeleton reveals a direct role in the regulation of chondrocyteproliferation. Development 128, 5099–5108 (2001).pmid: 11748145

47. J. Mao et al., A novel somatic mouse model to surveytumorigenic potential applied to the Hedgehog pathway.Cancer Res. 66, 10171–10178 (2006). doi: 10.1158/0008-5472.CAN-06-0657; pmid: 17047082

SCIENCE sciencemag.org 23 DECEMBER 2016 • VOL 354 ISSUE 6319 aah6102-11

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

48. S. Srinivas et al., Cre reporter strains produced by targetedinsertion of EYFP and ECFP into the ROSA26 locus. BMC Dev.Biol. 1, 4 (2001). doi: 10.1186/1471-213X-1-4;pmid: 11299042

49. H.-Y. Lee et al., Instructive role of Wnt/b-catenin in sensoryfate specification in neural crest stem cells. Science 303,1020–1023 (2004). doi: 10.1126/science.1091611;pmid: 14716020

50. B. Lustig et al., Negative feedback loop of Wnt signalingthrough upregulation of conductin/axin2 in colorectal and livertumors. Mol. Cell. Biol. 22, 1184–1193 (2002). doi: 10.1128/MCB.22.4.1184-1193.2002; pmid: 11809809

51. M. Pummila et al., Ectodysplasin has a dual role in ectodermalorganogenesis: Inhibition of Bmp activity and induction of Shhexpression. Development 134, 117–125 (2007). doi: 10.1242/dev.02708; pmid: 17164417

52. C. Trapnell et al., Transcript assembly and quantification byRNA-Seq reveals unannotated transcripts and isoformswitching during cell differentiation. Nat. Biotechnol. 28,511–515 (2010). doi: 10.1038/nbt.1621; pmid: 20436464

53. C. Trapnell et al., Differential gene and transcript expressionanalysis of RNA-seq experiments with TopHat and Cufflinks.Nat. Protoc. 7, 562–578 (2012). doi: 10.1038/nprot.2012.016;pmid: 22383036

ACKNOWLEDGMENTS

We thank Fuchs laboratory members, particularly M. Genander,T. Ouspenskaia, H.A. Pasolli, H. Yang, V. Fiore, S. Liu, and Y. Ge forreagents and helpful discussions, and J. Levorse, J. Racelis, andS. Chai for technical assistance. We thank Rockefeller ComparativeBioscience Center (Association for Assessment and Accreditationof Laboratory Animal Care–accredited) for care of mice inaccordance with National Institutes of Health (NIH) guidelines,Bioimaging Resource Center (A. North, director), and FlowCytometry Resource Center (S. Mazel, director, and S. Semova,S. Tadesse, and S. Han). E.F. is an investigator of the HowardHughes Medical Institute. C.P.L. is a recipient of fellowships fromthe National Institutes of Health (National Institute of Arthritis andMusculoskeletal and Skin Diseases grant K01-AR066073) andRobertson Therapeutic Development Fund. This work was

supported by NIH grant R01-AR050452 (E.F.). All reagentsengineered in the course of this study are available fromjdelacruz@rockefeller.edu under a materials transfer agreementwith the Rockefeller University. C.P.L. and E.F. conceived theproject, designed the experiments, and wrote the manuscript. C.P.L.performed the experiments. C.P.L. and L.P. performed mouse surgicalprocedures. B.E.K. and C.P.L. performed RNA-seq data analyses. RNA-seq data are deposited in the Gene Expression Omnibus underaccession number GSE85249 (www.ncbi.nlm.nih.gov/geo). The authorsdeclare no competing financial interests. Correspondence andrequests for materials should be addressed to E.F. (fuchslb@rockefeller.edu).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/354/6319/aah6102/suppl/DC1Figs. S1 to S8

21 July 2016; resubmitted 16 September 2016Accepted 18 November 201610.1126/science.aah6102

aah6102-12 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE

RESEARCH | RESEARCH ARTICLEon M

arch 14, 2020

http://science.sciencemag.org/

Dow

nloaded from

decisionSpatiotemporal antagonism in mesenchymal-epithelial signaling in sweat versus hair fate

Catherine P. Lu, Lisa Polak, Brice E. Keyes and Elaine Fuchs

DOI: 10.1126/science.aah6102 (6319), aah6102.354Science 

, this issue p. 10.1126/science.aah6102; see also p. 1533Sciencespecific times in development, are key to producing different skin appendages for adaptation to the environment.glands in the same area of the body. Epithelialmesenchymal interactions, with varied signaling pathways that act atPerspective by Lai and Chuong). They also examined human skin, which is capable of making both hairs and sweat

investigated skin appendage diversity during development of the furry backs and sweaty paws of mice (see theet al.Lu feat.tolerate extreme heat. Sweat glands, which enable humans to run in marathons, are instrumental for this remarkable

Unlike other mammals that must pant or seek shade or water when overheated, humans are able to self-cool toHow to grow hair or sweat glands

ARTICLE TOOLS http://science.sciencemag.org/content/354/6319/aah6102

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2016/12/21/354.6319.aah6102.DC1

CONTENTRELATED

http://science.sciencemag.org/content/sci/354/6319/1533.fullhttp://stke.sciencemag.org/content/sigtrans/6/261/ra8.full

REFERENCES

http://science.sciencemag.org/content/354/6319/aah6102#BIBLThis article cites 53 articles, 22 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2016, American Association for the Advancement of Science

on March 14, 2020

http://science.sciencem

ag.org/D

ownloaded from