DOI: 10.1126/science.1105458, 1728 (2005);307 Science

Bruce J. MacFaddenFossil Horses--Evidence for Evolution

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

): September 15, 2014 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/307/5716/1728.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/307/5716/1728.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/307/5716/1728.full.html#ref-list-1, 1 of which can be accessed free:cites 5 articlesThis article

8 article(s) on the ISI Web of Sciencecited by This article has been

http://www.sciencemag.org/content/307/5716/1728.full.html#related-urls6 articles hosted by HighWire Press; see:cited by This article has been

http://www.sciencemag.org/cgi/collection/evolutionEvolution

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2005 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Sep

tem

ber

15, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

15, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

15, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

15, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

http://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org/content/307/5716/1728.full.htmlhttp://www.sciencemag.org/content/307/5716/1728.full.html#relatedhttp://www.sciencemag.org/content/307/5716/1728.full.html#ref-list-1http://www.sciencemag.org/content/307/5716/1728.full.html#related-urlshttp://www.sciencemag.org/cgi/collection/evolutionhttp://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/http://www.sciencemag.org/

1728

from 12 patients with recessive dystrophicepidermolysis bullosa. They show convinc-ingly that susceptibility to developing inva-sive SCC, both clinically and experimen-tally, depends strictly on the retention ofpart of the collagen VII protein. Keratin-ocytes from patients carrying mutationsthat abrogate the deposition of collagen VIIdo not develop into invasive SCC, whereasthose from patients with mutations thatresult in deposition of a crucial fragment ofcollagen VII do become cancerous.

Collagen VII is produced primarily bykeratinocytes, with perhaps a small contri-bution from dermal fibroblasts. The colla-gen VII molecule has a characteristic cen-tral glycine-rich, triple-helical collagenousdomain, with noncollagenous domains at itsamino and carboxyl ends. Keratinocytesfrom patients with mutations that specifi-cally leave intact the amino-terminal non-collagenous domain (NC1) of collagen VII,and more specif ically the f ibronectinIIIlike repeats within the NC1 domain(FNC1) that bind to laminin 5, developedinto invasive SCC. Furthermore, introduc-tion of either the NC1 or FNC1 domainsinto patient keratinocytes deficient in colla-gen VII restored a predisposition to tumori-genesis, whereas introduction of NC1 with-out the f ibronectin repeats did not.Interestingly, antibodies that specificallyrecognized the FNC1 domain of collagenVII either prevented tumor development orsuppressed tumor invasion when adminis-tered to mice with SCC tumors caused byRas/IB-transformed keratinocytes fromnormal individuals. Invasion studies in vitroconfirmed the in vivo findings and furtherrevealed that interaction of FNC1 withlaminin 5 was required for the invasive phe-notype to develop.

What do these results tell us about epi-dermolysis bullosa and SCC? First, theysuggest an explanation for why chronicwounds seldom develop into SCC in patientswith mutations in adhesion complex pro-teins that are closer to the epidermis (forexample, laminin 5, hemidesmosomal pro-teins, and intermediate filament proteins).Keratinocytes harboring such mutationslack an intact adhesion complex between theNC1 domain of collagen VII and laminin 5and the hemidesmosomes. Hence, these ker-atinocytes are not tethered to the dermis andmay not receive the stromal signals that theywould need to migrate to and invade the der-mal layer. Laminin 5 is the ligand for 64integrin, a signaling receptor on the surfaceof basal keratinocytes. Hence, interactionsbetween collagen VII and laminin 5 may bethe conduit for stromal signals that direct themigratory and invasive behaviors of epider-mal tumors (6).

Ortiz-Urda et al. also show that boosting

production of NC1 enhances the invasive-ness of transformed keratinocytes from nor-mal individuals, and of keratinocytes frompatients with other skin diseases. A centralregulator of collagen VII expression istransforming growth factor (TGF-) (7),which enhances invasion and metastasis ofestablished squamous cell tumors and otherepithelial neoplasms (8). The new worksuggests that the relationship between col-lagen VII and TGF- is worth exploring fur-ther. There are also two possible clinicalapplications of the current study. Attemptsto restore collagen VII locally using genetherapy in patients with dystrophic epider-molysis bullosa are under active investiga-tion (9). The authors caution that for certainpatients, restoration of collagen VII con-taining the NC1 domain could increase theirrisk of developing SCC, particularly inthose who lack production of collagen VII.On the other hand, the good news is that theNC1 domain could be a therapeutic targetfor treating invasive SCC and other cancers.

However, a therapeutic molecule that bindsto the NC1 domain must block the molecu-lar interactions required for tumor invasionwhile leaving intact those required foranchoring the epidermis to the dermis. Weare faced with a possible Pyrrhic victory aswe contemplate the epithelial-stromal inter-face: perhaps winning the battle againstSCC but losing the battle against the disfig-uring skin defects of dystrophic epidermol-ysis bullosa.

References1. S. Ortiz-Urda et al., Science 307, 1773 (2005).2. L. Pulkkinen, J. Uitto, Matrix Biol. 18, 29 (1999).3. J. L.Arbiser et al., Mol. Med. 4, 191 (1998).4. J. L.Arbiser et al., J. Invest. Dermatol. 123, 788 (2004).5. M. Dajee et al., Nature 421, 639 (2003).6. M. M. Mueller, N. E. Fusenig, Nature Rev. Cancer 4, 839

(2004).7. M. J. Calonge, J. Seoane, J. Massague, J. Biol. Chem. 279,

23759 (2004).8. A. B. Glick, Cancer Biol. Ther. 3, 276 (2004).9. D.T.Woodley et al., Nature Med. 10, 693 (2004).

10. J. R. McMillan, M.Akiyama, H. Shimizu, J. Dermatol. Sci.31, 169 (2003).

10.1126/science.1110346

18 MARCH 2005 VOL 307 SCIENCE www.sciencemag.org

P E R S P E C T I V E S

Thomas Huxley, an early advocate ofDarwinian evolution, visited theUnited States in 1876 on a lecture tour.

Huxley had planned to talk about evidencefor evolution based on a fragmentarysequence of fossil horses from Europe. Oneof Huxleys first stops was at Yale, where hestudied the fossil horse collection assembledby the paleontologist O. C. Marsh duringexpeditions to the western territories. Huxleywas so taken with the definitive evidenceprovided by Marshs fossil horse collectionthat he used this evolutionary sequence as thefocal point for his subsequent talk to the NewYork Academy of Sciences (1).

Since the late 19th century, the 55-million-year (My) phylogeny of horses (FamilyEquidae)particularly from North Americahas been cited as definitive evidence of long-term quantum evolution (2), now calledmacroevolution. Macroevolution is the studyof higher level (species, genera, and above)evolutionary patterns that occur on timescales ranging from thousands to millions ofyears. The speciation, diversification, adap-tations, rates of change, trends, and extinc-

tion evidenced by fossil horses exemplifymacroevolution.

The sequence from the Eocene dawnhorse eohippus to modern-day Equus hasbeen depicted in innumerable textbooks andnatural history museum exhibits. In Marshstime, horse phylogeny was thought to be lin-ear (orthogenetic), implying a teleologicaldestiny for descendant species to progres-sively improve, culminating in modern-dayEquus. Since the early 20th century, however,paleontologists have understood that the pat-tern of horse evolution is a more complex treewith numerous side branches, some leadingto extinct species and others leading tospecies closely related to Equus. Thisbranched family tree (see the figure) is nolonger explained in terms of predestinedimprovements, but rather in terms of randomgenomic variations, natural selection, andlong-term phenotypic changes (3).

The Equidae, a family within the odd-toed ungulate Order Perissodactyla (whichincludes rhinoceroses, tapirs, and otherclosely related extinct groups), consists ofthe single extant genus Equus. Dependingupon interpretation, it also includes severalsubgenera, 8 to 10 species, and numeroussubspecies (4). On the basis of morphologicaldifferences, Equus is separated into two or

E V O L U T I O N

Fossil HorsesEvidence for Evolution

Bruce J. MacFadden

The author is in the Florida Museum of NaturalHistory, University of Florida, Gainesville, FL 32611,USA. E-mail: bmacfadd@flmnh.ufl.edu

Published by AAAS

1729

three deep clades withinthe genus. These includecaballines (domesticatedhorse, E. caballus); zebras(three species recognized);and asses, donkeys, andrelated species. Recentstudies of mitochondrialDNA indicate two deepclades within Equus,namely, the caballines andthe zebras/asses (5). Thesedeep clades split ~3 mil-lion years ago (Ma) inNorth America and subse-quently dispersed into theOld World. Equus becameextinct in the New World~10,000 years ago, proba-bly as a result of multiplefactors including climatechange and hunting byearly humans. In the OldWorld, although its rangecontracted, Equus persistedand was then domesticatedin central Asia about 6000years ago from a stock sim-ilar to Przewalskis wildhorse, E. caballus (some-times considered its ownspecies, E. przewalskii) (4).

The single moderngenus Equus stands inmarked contrast to ahighly diverse adaptiveradiation of the FamilyEquidae over the past 55My that resulted in somethree dozen extinct generaand a few hundred extinctspecies (3). Although theoverall branched patternof horse phylogeny (seethe figure) has remainedsimilar for almost a cen-tury, new discoveries andreinterpretation of exist-ing museum fossil horsecollections have added tothe known diversity ofextinct forms. Recentwork reveals that Eocenehyracothere horses,previously known aseohippus or Hyraco-therium, include an earlydiversification of a half-dozen genera that existed between 55 and52 Ma in North America and Europe (6).New genera have recently been proposedfor the complex middle Miocene radiation(7), although the validity of these genera isstill debated.

Horse teeth frequently preserve as fos-

sils and are readily identifiable taxonomi-cally. They serve as objective evidence ofthe macroevolution of the Equidae. Horseteeth have undergone considerable changesover the past 55 My. The tempo of this mor-phological evolution has sometimes beenslow and at other times rapid (2, 3).

Primitive Eocene through early Miocene(between 55 and 20 My) horses had short-crowned teeth adapted for browsing on soft,leafy vegetation. During the later Miocene(between 20 and 15 Ma), horses underwentexplosive adaptive diversification in toothmorphology. Shorter crowned browsers,CR

EDIT

:PRE

STO

N H

UEY

/SC

IEN

CE

Plio

.Q

uat

.M

ioce

ne

Olig

oce

ne

Mill

ion

yea

rs a

go

Eo

cen

e

5

S. America

Ono

hipp

idiu

m

Hip

pidi

on

Ast

rohi

ppus

Pro

tohi

ppus

Old

Wor

ld H

ippa

rion

Cla

des

Sin

ohip

pus

Anc

hith

eriu

m

Kal

obat

ippu

s

Mio

hipp

us

Mer

ychi

ppus

II

Mes

ohip

pus

Hap

lohi

ppus

Epi

hipp

us

Oro

hipp

us

Hyr

acot

here

Cla

des

Din

ohip

pus

Mer

ychi

ppus

I

Arc

haeo

hipp

us

Par

ahip

pus

Cal

ippu

s

Plio

hipp

us Pse

udhi

ppar

ion

Equ

us

Neo

hipp

ario

n

Hip

pario

n

Nan

nipp

us

Cor

moh

ippa

rion

Hyp

ohip

pus

Meg

ahip

pus

N. America Old World

10

15

20

25

30

35

40

45

Mostly grazers

Mixed feeders

Mostly browsers

50

55

Adaptive radiation of a beloved icon. Phylogeny, geographic distribution, diet, and body sizes of the Family Equidae overthe past 55 My.The vertical lines represent the actual time ranges of equid genera or clades.The first ~35 My (Eocene toearly Miocene) of horse phylogeny are characterized by browsing species of relatively small body size.The remaining ~20My (middle Miocene until the present day) are characterized by genera that are either primarily browsing/grazing or aremixed feeders, exhibiting a large diversification in body size. Horses became extinct in North America about 10,000 yearsago, and were subsequently reintroduced by humans during the 16th century.Yet the principal diversification of this fam-ily occurred in North America.Although the phylogenetic tree of the Equidae has retained its bushy form since the 19thcentury [for example, see (2, 3)], advances in knowledge from fossils have refined the taxonomy, phylogenetic interrela-tionships, chronology, and interpretations of the ancient ecology of fossil horses.

www.sciencemag.org SCIENCE VOL 307 18 MARCH 2005

P E R S P E C T I V E S

Published by AAAS

1730

which inhabited forests and open-countrywoodlands, declined in diversity during thistime (8). In contrast, many other clades ofhorses evolved high-crowned teeth adaptedfor grazing on the extensive grasslands ofmore open-country biomes, which spreadduring the Miocene (25 to 15 Ma). Oncehigh-crowned teeth evolved, some cladesunderwent a secondary adaptation, that is,they went from being grazers to beingmixed feeders with diets consisting of grassand some leafy plants (9). Studies of car-bon isotopes preserved in fossil horse teethindicate that before ~7 Ma, early tropicaland temperate grasslands of the world con-sisted primarily of grasses that used the C3photosynthetic pathway, whereas todaythese grasslands consist mostly of C4grasses (10).

In many fossil groups, the trend towardlarger body size in ancestral-descendentsequences has been termed Copes rule.Early Eocene hyracothere horses classicallyhave been compared in size to a small dog(~10 to 20 kg), although house-catsized

species have been discovered more recently.At the other end of the evolutionary spec-trum, wild modern Equus attains a bodysize of ~500 kg (3, 4). Although the 55-My-old fossil horse sequence has been used as aclassic example of Copes rule, this notionis now known to be incorrect. Rather than alinear progression toward larger body size,fossil horse macroevolution is characterizedby two distinctly different phases. From 55to 20 Ma, primitive horses had estimatedbody sizes between ~10 and 50 kg. In con-trast, from 20 Ma until the present, fossilhorses were more diverse in their bodysizes. Some clades became larger (likethose that gave rise to Equus), othersremained relatively static in body size, andothers became smaller over time (3).

Fossil horses have held the limelight asevidence for evolution for several reasons.First, the familiar modern Equus is abeloved icon that provides a model forunderstanding its extinct relatives. Second,horses are represented by a relatively con-tinuous and widespread 55-My evolution-

ary sequence. And third, important fossilscontinue to be discovered and new tech-niques developed that advance our knowl-edge of the Family Equidae. The fossilhorse sequence is likely to remain a popularexample of a phylogenetic pattern resultingfrom the evolutionary process.

References1. C. Schuchert, C. M. LeVene, O. C. Marsh Pioneer in

Paleontology (Yale Univ. Press, New Haven, CT, 1940).2. G. G. Simpson, Major Features of Evolution (Columbia

Univ. Press, New York, 1953).3. B. J . MacFadden, Fossil Horses: Systematics,

Paleobiology, and Evolution of the Family Equidae(Cambridge Univ. Press, New York, 1992).

4. R. M. Nowak, Walkers Mammals of the World, 5.1Online (Johns Hopkins Univ. Press, Baltimore, 1997).

5. E. A. Oakenfull, H. N. Lim, O. A. Ryder, Conserv. Genet.1, 341 (2000).

6. D. J. Froehlich, Zool. J. Linn. Soc. 134, 141 (2002).7. T. S. Kelly, Contrib. Sci. Nat. Hist. Mus. Los Ang. Cty.

455, 1 (1995).8. C. M. Janis, J. Damuth, J. M. Theodor, Proc. Natl. Acad.

Sci. U.S.A. 97, 7899 (2000).9. B. J. MacFadden, N. Solounias,T. E. Cerling, Science 283,

824 (1999).10. T. E. Cerling et al., Nature 389, 153 (1997).

10.1126/science.1105458

CRE

DIT

:HEL

ENE

LETO

URN

EAU

/NAT

ION

AL

RESE

ARC

H C

OU

NC

IL O

F C

AN

AD

A A

ND

PRE

STO

N H

UEY

/ SC

IEN

CE

The early 20th century saw a rapid evo-lution in the concept of the atom (seethe f irst f igure). In 1911, Ernest

Rutherford proposed that the atom resem-bled a tiny solar system, with most of themass concentrated in a nucleus, and elec-trons revolving around it in planetary orbits(see the first figure, panel A). Although hismodel has been supplanted by quantummechanics, Maeda et al. show on page 1757of this issue (1) that it is possible to makeRutherford atoms in the laboratory.

Doubts about the Rutherford modelwere raised soon after he proposed it. Theemission spectrum of hydrogen was knownto have a regular progression of lines, andJohannes Rydberg had shown in 1888 thatthese lines could be fit to a simple algebraicformula. The Rutherford model could notexplain this behavior.

In 1913, Niels Bohr postulated that theangular momentum of the electrons must bequantized. This quantization led to discreteorbits (see the first figure, panel B), whichwere labeled by an integer quantum number

and explained Rydbergsformula. When Bohrsmodel was supplanted byquantum mechanics as weknow it today, the conceptof the electron as a planetwas replaced by a mathe-matical wave function thatwas not directly observ-able. The electron orbitsbecame fuzzy clouds (seethe first figure, panel C).

Quantum mechanics has been highlysuccessful in predicting the structure ofatoms and molecules, giving birth to notionssuch as quantum teleportation and quantumcomputing. It is so precise that experimentshave been proposed to look for tiny changesin the fundamental constants as the universeages over a period of years (2). Yet we knowthat objects in the macroscopic world that

we inhabit are not fuzzyclouds. Electrons are realparticles that travel throughwires and form images ontelevision screens. So wheredoes the quantum world endand our everyday classicalworld begin?

This question is beingaddressed by scientists whotry to construct atoms thatresemble classical objects.Maeda et al. (1) show thatthis approach can yield atomsthat behave like Rutherfordsminiature solar system.

To make a Rutherfordatom, one must first localizethe electron cloud, that is,confine it to a small volume.In quantum mechanics, thisis achieved by creating acoherent superposition ofstates, called a wave packet(see the second figure). Forexample, a femtosecondlaser pulse is a superposi-tion of many sine waves thatadd constructively only in

P H Y S I C S

Toward Creatinga Rutherford Atom

David M.Villeneuve

Evolution of the concept ofatomic structure. (A) In theRutherford model, an elec-tron orbits a massive nucleusin a planetary orbit. (B) In theBohr model, the electron ispartly a wave that can only goin discrete orbits. (C) In thequantum mechanical picture,the electron is spread outwithin an orbital. (D) In theRydberg atom, the electron islocalized into a small volumeand follows an almost classi-cal orbit, as demonstrated byMaeda et al. (1).

The author is with the National Research Council ofCanada, 100 Sussex Drive, Ottawa, Ontario, CanadaK1A 0R6. E-mail: david.villeneuve@nrc.ca

A

B

C

D

P E R S P E C T I V E S

18 MARCH 2005 VOL 307 SCIENCE www.sciencemag.orgPublished by AAAS