Wireless Power Transfer via Strongly Coupled Magnetic Resonances Author(s): Andr Kurs, Aristeidis Karalis, Robert Moffatt,

J. D. Joannopoulos, Peter Fisher and Marin Soljai Source: Science, New Series, Vol. 317, No. 5834 (Jul. 6, 2007), pp. 83-86 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/20036648 . Accessed: 23/02/2014 21:27
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

http://www.jstor.org

This content downloaded from 129.8.242.67 on Sun, 23 Feb 2014 21:27:22 PM All use subject to JSTOR Terms and Conditions

is, 2Tw\aD(t)\2. Maximizing the efficiency r\ of

respect to the loading Tw, given an impedance to solving 1, is equivalent that the scheme finds One matching problem. the transfer with Eq. works best when in which the source case and the device is are resonant, the efficiency

Wireless Coupled

Power Magnetic

Transfer

via Strongly Resonances

Andr? Kurs,1* Aristeidis Karalis,2 Robert Moffatt,1 J. D. Joannopoulos,1 Peter Fisher,3 Marin Soljacic1 Using self-resonant coils in a strongly coupled regime, we experimentally demonstrated efficient
nonradiative power transfer over distances up to 8 times the radius of the coils. We were able

+ rw)k*D|2 rsH2 + (rD

_rp TsTp_ /~\

to transfer 60 watts with -40% efficiency over distances in excess of 2 meters. We present a quantitative model describing the power transfer,which matches the experimental results to within 5%. We discuss the practical applicability of this system and suggest directions for further study.
In the early 20th effort toward before the electrical not with interact with environmental were magnetic objects able to so interactions fields, even are suppressed the strongly identify two coupled nonradiative at

century,

The efficiency ismaximized when rw/TD = [1 + (K2/TsrD)]1/2.It is easy to show that the key to
efficient This transfer is to have K2/TsrD > 1. energy referred to as the strong coupl is commonly role in plays an essential ing regime. Resonance as the efficiency is this power transfer mechanism, relative to the case of parameters) nonresonant objects. tively coupled for self-resonant model Theoretical induc

wire grid, Nikola Tesla (1) devoted much

schemes typical lessly. However, The of past decade autonomous to transport power wire embodiments (e.g., Tesla a surge in the use cell

further. We coupled

coils) involved undesirably large electric fields.

has witnessed electronic devices (laptops,

magnetic

in the system of regime resonances by exploring resonant (near-field) magnetic At frequencies. transfer is reminiscent first of

induction

improved by approximately co2/TD2(~106 for

typical

megahertz power netic usual

phones, interest

robots, PDAs, etc.). As a consequence, in wireless (2-4). power has reemerged

induction nonresonant

Radiative transfer (5), although perfectly suitable

a number of for transferring information, poses The transfer applications: for power difficulties

(10); however, is very induction

such glance, the usual mag note that the inefficient

coils.

Our

for midrange

applications.

Overview of the formalism. Efficient mid

in particular regions transfer occurs range power resonant the parameter of space describing to one another. Using coupled strongly objects coupled-mode linear equations: theory to describe this physical

of the scheme con realization experimental coil (the coils. One sists of two self-resonant to an oscillat source coil) is coupled inductively

efficiency of power transfer is very low if the

and unidirectional is omnidirectional, line of sight radiation requires an uninterrupted A recent and sophisticated tracking mechanisms. radiation

ing circuit; the other (the device coil) is coupled inductively to a resistive load (12) (Fig. 1). Self
resonant distributed tance coils rely on the interplay and distributed inductance resonance. The coils between capaci are made

theoreticalpaper (6) presented a detailed analysis

of the feasibility of using resonant objects cou

system (77), we obtain the following set of = (i(?m Tm)am(t) +

to achieve

pled through the tails of theirnonradiative fields

for midrange energy transfer resonant objects of the same tend to exchange relatively objects. two (7). Intuitively, resonant frequency where

wire of total length of an electrically conducting into a helix radius a wound / and cross-sectional

?m(t)

^iKmnan(t)+Fm(t)
the indices denote the different

(1)

of n turns, radius r, and height h. To the best of for a our knowledge, there is no exact solution finite helix infinitely in the literature, and even in the case of the solutions rely on long coils, our are for that system inadequate assumptions that the simple (13). We have found, however, quasi-static model described -5%) with below is in good

sipating resonant (e.g., ar),

while dis energy efficiently, off in extraneous little energy In systems of coupled resonances magnetic, nucle

resonant

acoustic, electromagnetic, there is often a general of operation in a given to be

objects. The variables the energy contained

regime regime expected transfer

coupled" "strongly (8). If one can operate in that transfer is the energy system, power Midrange can be nearly in this way of the efficient, irrespective efficient

so that am(t) are defined in object m is \am(t)\2, com isolated ofthat is the resonant angular frequency rate its intrinsic is and (e.g., Tm decay object, In this and radiated due to absorption losses).

very

implemented and omnidirectional of

low in the surrounding space, with geometry terference and losses into environmental objects (6). The above considerations irrespective apply nature of the resonances. the physical Here, we focus on one particular physical embodiment: resonances resonances (9). Magnetic magnetic are particularly suitable for everyday applica of tions because most of the common materials do

an uncoupled oscilla and undriven framework, in evolve tor with parameters coo and T0 would = time as exp(/oV Knm are IV). The Kmn the resonant between coefficients coupling

(within agreement start by observing We be zero at the ends of the coil, and we make of that the resonant modes educated guess coil are well

experiment. that the current must the the

objects indicatedby the subscripts,andFm(t) are

driving We objects, en terms. limit the treatment by source at a constant a coupling to the case and device, of such two that

current by sinusoidal approximated wire. the of the conducting length along profiles so if we are interested in the lowest mode, We denote coordinate by s the parameterization that it such the of the conductor, length along runs from -111 to +1/2, then the time-dependent

denoted

the source (identified by the subscript S) is driv

externally two objects have of a frequency, coefficient and the k. Work

currentprofile has the form I0 cos(ns/l) exp(/co/). It follows from the continuity equation for charge that the linear charge density profile is
of the form Xq sin(jts//) contains q0 an exp(ior), so that one-half

is extracted from the device (subscriptD) by

of Physics, Massachusetts Institute of Technol department of Electrical ogy, Cambridge, MA 02139, USA.department Engineering and Computer Science, Massachusetts Instituteof of Technology, Cambridge, MA 02139, USA.department Physics and Laboratory for Nuclear Science, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. *To whom correspondence akurs@mit.edu should be addressed. E-mail: means circuit has Tw load resistance (subscript connected that acts W) to the device, an additional as a and term

of the coil (when sliced perpendicularly to its

axis) = amplitude but opposite half. As density total charge (of oscillating that is equal in magnitude in the other in sign to the charge XqI/k) the current is resonant, are n/2 out of phase and charge from each

the effect

of contributing to the unloaded device

TD. The fore FD

overall decay = TD + Tw. The work

rate decay object's rate at the device is there extracted is de

the coil profiles

terminedby the power dissipated in the load, that SCIENCE VOL 317

www.sciencemag.org

6 JULY 2007

83

This content downloaded from 129.8.242.67 on Sun, 23 Feb 2014 21:27:22 PM All use subject to JSTOR Terms and Conditions

ARTICLES RESEARCH
that the real part of one other, meaning mum when the real part of the other Equivalently, is at certain is maxi is zero. the first The term for our experimental coupled-mode the coil is therefore factor theory decay = + T Rr)/2L, (R0 = is Q (o/2r. parameters. constant for and its about 5% off from the measured resonance at 9.90 MHz. The Q for the loops is estimated a = 5.9 x 107 m/ohm), (assuming = 950 ? 50. We value is Q the measured theoretical of to the is mostly due the discrepancy the layer of poorly copper conducting on the surface to of the copper wire, the current the short We skin therefore = and Ts

in the coil the energy contained in time completely due to points

quality

to be -2500 but

due

it is completely the current, and at other points to the charge. electromagnetic Using an effective inductance theory, we can define an effective capacitance C for each coil

We

find the coupling coefficient kds by

L and

as follows:

looking at the power transferred from the source a steady-state to the device solu coil, assuming currents and charge densities tion in which vary in time as exp(zcor):

believe effect oxide which

JJ 47:|7o|2

|r-r'|

PDS= ??rEs(r).JD(r)
= -Jjr[?s(r)+V<|)s(r)].JD(r)

is confined by at this frequency. depth (-20 urn) use the experimentally observed = r = it in rD co/2g derived from computations. We find

Q all subsequent

= i C ?^
where p(r) and charge L U

" (4) ?\?rdr>>p^ k r'| 47ceoko|2JJ

current J(r) respectively along and charge density the current from the isolated coil, in

k experi the coupling coefficient two coils the self-resonant mentally by placing h, to the same (fine-tuned, by slightly adjusting resonant D a distance when isolated) frequency in the fre the splitting and measuring Accord of the two resonant modes. this splitting In the present the two coils similar (figs. maxi on is the

the spatial are obtained

= -i\Urdr'
m Js(0

densities with

JJ<
, Ps(r')
?o

apart

conjunction defined, energy

the geometry of the object. As the property that the have in the coil is given by contained and C

r' ~r
|r' r|" Jo(r')

quencies ing to should

= -/coM7s7D (8)
Where M scalar the is the effective mutual is inductance, <J) is the vector potential,

work, are aligned results

theory, coupled-mode = be Aco 2[(k2 T2)172]. we focus on the case where coaxially are obtained

(Fig. 2), although for other orientations

SI and S2).
Measurement mum theoretical of the efficiency. The only depends = the parameter k/T, which k/[(LsLd)1/2] = 2.4 m (8 times than 1 even for D greater efficiency radius

= ?M2
Given tinuity, this relation and the resulting can now We \/[2n(LC)V2]. oscillator the equation resonant frequency of treat this coil

the

(5)
con = is^o as a by

and field from

A potential, that the electric S indicates subscript then conclude is due to the source. We

standard coupled-mode theory arguments = = K = When that kds KSD c?A//[2(?sLd)1/2]. the centers of the coils the distance D between is much

standard

defining a(t)

in coupled-mode

theory

[(L/2)m]I0(t).

can estimate We the power dissipated by that the sinusoidal noting profile of the current of distribution that the spatial average implies a the peak current is For coil squared \Io\2/2. with n turns and made with con of a material ductivity for ohmic accordingly: o, we (R0) modify and radiation the standard (Rr) formulas resistance

size, k larger than their characteristic characteristic with the D~3 dependence k and V are Both of dipole-dipole coupling. and k/T and the of the frequency, functions scales efficiency are maximized for a particular value

of the coils) (Fig. 3). Thus, we operate the in the strongly throughout coupled regime entire range of distances probed. As our driving circuit, we use a standard con element inductive oscillator whose Colpitts 25 cm in sists of a single loop of copper wire in radius (Fig. 1); this loop of wire couples to the source coil and drives the entire ductively wireless power to its own transfer apparatus. The load

of/

which

is in the range 1 to 50 MHz for

of interest. Thus, parameters picking typical an appropriate for a given coil size, frequency as we do in this experimental demonstration, a major the power role in optimizing plays transfer.

consists of a calibrated light bulb (14) and is

attached loop of insulated wire, which in proximity coil and of the device to the dis it. By varying coupled inductively tance between the light bulb and the device coil, is placed are able itmatches so that to adjust the parameter rw/T its optimal value, given theoretically

Comparisonwith experimentally determined

parameters. validation The parameters for the two iden = transfer scheme are h of the power = = = 5.25. Both 20 cm, a 3 mm, r 30 cm, and n coils loops The first term term in Eq. (assuming 7 radiation is a magnetic dipole r? c 2nc/(?, where is smaller than are made of the of copper. The spacing helix is not uniform, about between and we

we

tical helical coils built for the experimental

(The loop connected to the by [1 + (k2/T2)]172.

encapsulate The

the uncertainty

their uniform

ity by attributinga 10% (2 cm) uncertainty to h.

resonant these frequency given expected = which is 10.56 ? 0.3 MHz, dimensions is^?

is the speed of light); the second term is due to

the electric dipole of the coil and

Fig.

1.

Schematic

of the exper

imental setup. A is a single 25 cm that copper loopof radius o is part of the driving circuit, which outputs a sine wave with
frequency respectively 9.9 MHz. S and D are the source and de

/\y "y,"~uu,u

175 125 150 Distance (cm) and theoret of experimental Fig. 2. Comparison ical values for k as a function of the separation source and device coils between coaxially aligned (the wireless power transfer distance).

to in the text. B is a loopof wire attachedto the load (lightbulb).Thevariousks represent vice coils referred direct couplingsbetween the objects indicatedby the arrows.The angle between coil D and the loopA is The direct couplings adjustedto ensure that theirdirect coupling is zero. CoilsS and D are aligned coaxially. between B and A and between B and S are negligible. 84 6 JULY 2007 VOL 317 SCIENCE

www.sciencemag.org

This content downloaded from 129.8.242.67 on Sun, 23 Feb 2014 21:27:22 PM All use subject to JSTOR Terms and Conditions

ARTICLES RESEARCH I
light bulb Tw, which adds a small reactive is compensated ing the coil.) We measure to component for by slightly retim the work extracted by fer at that distance the Colpitts oscillator. and the low efficiency of It is essential that the remarks. Concluding for the power transfer to be coils be on resonance the maxi coil, we calculate = 1 -4 kV/m, values for the fields to be ?rms = = The and Srms 0.2 W/cm2. 8 A/m, #rms is -5 these radiated for W, power parameters an order of magnitude which is roughly higher surface of the device mum we that In the particular geometry than cell phones. contribution the overwhelming studied, (by one to two orders of magnitude) to the electric to the near-field and hence near-field, the electric dipole vector, comes from of the coils. If instead one uses Poynting moment

adjusting the power going into the Colpitts oscillator until the light bulb at the load glows at its full nominal brightness.
We determine taking place of the current at the midpoint load by measuring a current each of the self-resonant coils with of the transfer the efficiency between the source coil and the

practical (6). We find experimentally that the

to the load drops transmitted power either one of the coils is detuned from current sharply as resonance.

For a fractional detuning A/% of a few times the

inverse loaded Q, detailed the induced in the device

from noise. coil is indistinguishable

A effect of external and quantitative analysis on our scheme objects of the is be

probe (which does not lower theQ of the coils

noticeably). current parameters oretical dissipated + model. in each This gives Is coil a measurement and then ID used = of in our the the

a capacitively

loaded single-turn loop design (6)?which

the advantage electric field the system sible nearly confining inside the capacitor?and of all of

has
the

We

compute from PSD that

the power

and obtain the efficiency from rj

^d setup pled mode To ensure ^w)described is well theory

Pw/(Ps +

rZ,|/s,D|2,

but we note here the scope of this work, yond as affected transfer is not visibly that the power as such and various humans objects, everyday devices and electronic wood, metals, large and small, are placed in cases where line of S3 sight between the two coils?even obstruct the

tailors our

to operate

at lower frequencies, cited above for

calculations show (77) that it should be pos

to reduce the values the the Poynting and magnetic vector, fields, so that they fall below radiated and the power thresholds specified by general safety regulations electric [e.g., the IEEE safety standards for general public

the experimental cou a by two-object we position the de model,

they between

completely source and device

vice coil such that its direct coupling to the copper loop attached to theColpitts oscillator is
zero. Fig. maximum able results experimental the theoretical 4, along with The are shown in for prediction efficiency, given by Eq. 2. We were to transfer several tens of watts with the than 2 m

to S5). External have objects effect they are within only when either one of the timeters from (such as aluminum foil,

(figs. a noticeable a few coils. cen Some

exposure (18)].
the two coils are currently of iden Although to make the device it is possible tical dimensions,

materials

Styrofoam,

and humans) mostly just shift the resonant fre

can in principle which quency, a feedback rected with circuit; and polyvinyl (cardboard, wood, than closer Q when placed from the coil, thereby lowering the transfer. When culate coils, cor be easily other materials

coil small enough to fit into portable devices

without the efficiency. One could, for decreasing of the charac the product instance, maintain coils con teristic sizes of the source and device in (6). transfer distances could be appre the coils, which by silver-plating their Q, or by working with more

use of this setup, fully lighting up a 60-W

light bulb from As power distances we more away

(figs. S3 and S4).

a cross-check, also measured the total going from the wall power outlet into the circuit. The efficiency of the wireless driving in this way, transfer itself is hard to estimate as the efficiency os of the Colpitts however, cillator itself is not precisely extracted gives known, although it

lower chloride) a few centimeters the efficiency of

stant, as argued and the power ciably improved should increase elaborate Nonetheless,

We believe that the efficiency of the scheme

cal the of

is expected to be far from 100% (75). Still, the

ratio of power circuit driving to power the entering a lower bound on the ef

the electric field is?rms = 210 V/m, thatof the = HRMS 1A/m, and thatof the magnetic field is
= vector is Srms Poynting increase closer These values the fields at source (16). to the coils, where are comparable. and device 20 cm away from the 3.2 mW/cm2

60 W across 2 m, we transferring between that at the point halfway the root mean square (RMS) magnitude

for the resonant objects (79). geometries characteristics of the the performance

system presented here are already at levels where in practical applications. they could be useful References and Notes 1. N. Tesla, U.S. patent 1,119,732 (1914). 2. ].M. Fernandez, ].A. Borras, U.S. patent 6,184,651 (2001). Trans. Ind.Appl. 27, 872 3. A. Esser, H.-C Skudelny, IEEE (1991). Trans. Power 4. J. Hirai, T.-W. Kim, A. Kawamura, IEEE Electron. 15, 21 (2000). 5. T. A. Vanderelli, ]. G. Shearer, ]. R. Shearer, U.S. patent 7,027,311 (2006). 6. A. Karalis, J. D. Joannopoulos, M. Soljacic, Ann. Phys., (2007). 10.1016/j.aop.2007.04.017 7. Here, by midrange, we mean that the sizes of the devices that participate in the power transfer are smaller than the distance between devices by a factor of at least 2 to 3. For example, if the device being powered is a laptop (size -50 cm) and the power source (size -50 cm) is in the same room as the laptop, the distance of power transfer could be within a room or a factory pavilion (size on the order of a few meters). 8. T. Aoki et ai, Nature 443, 671 (2006). 9. K. O'Brien, G. Scheible, H. Gueldner, in IECON '03, 29th Annual Conference of the IEEE(http://ieeexplore.ieee.org/ xpl/RecentCon.jsp?punumber=9011) (2003). W. Hay, P. G.W. Beart, U.S. patent 7,042,196 10. L.Ka-Lai, J. (2006). 11. H. Haus, Waves and Fields in Optoelectronics (Prentice Hall, Englewood Cliffs, N], 1984). 12. The couplings to the driving circuit and the load do not have to be inductive. They may also be connected by a wire, for example. We have chosen inductive coupling in the present work because of its easier implementation. 13. S. Sensiper, thesis, Massachusetts Institute of Technology (1951).

ficiency. When
over a distance

transferring 60 W
of 2 m, for example,

to the load
the power

For example,

at distances

flowing into the driving circuit is 400 W. This

yields which an overall wall-to-load is reasonable to 50% of 15%, efficiency efficien the given expected trans for the wireless power

cy of 40

75

TOO 125

?5?

175 200
(cm)

225

Distance

75

100

125

150

175 200
(cm)

225

Fig. 4. Comparison of experimental and the oretical efficiencies as functions of the wireless
power transfer distance. The shaded area rep

Distance

Fig. 3. Comparisonof experimental and theoret

ical values the wireless for the parameter power transfer measured k/T as a function The of distance. r. theory area

resents the theoretical prediction for maximum efficiency and is obtained by inserting the the oreticalvalues from Fig. 3 into Eq.2, with rVTD =
[1 + (k^/T2)]172. The black squares are the max

values are obtained by using the theoreticalk and

the experimentally The shaded

representsthe spread in the theoreticalk/T due to the 5% uncertainty inQ.

imum efficiency obtained from Eq. 2 and the experimental values of k/T from Fig. 3. The red dots present the directlymeasured efficiency, as described in the text.

www.sciencemag.org

SCIENCE

VOL 317

6 JULY 2007

85

This content downloaded from 129.8.242.67 on Sun, 23 Feb 2014 21:27:22 PM All use subject to JSTOR Terms and Conditions

RESEARCH ARTICLES
14. We experimented with various power ratings from 5 to 75 W. 15. W. A. Edson, Vacuum-Tube Oscillators (Wiley, New York, 1953). 16. Note that E * c\iqH, and that the fields are out of phase and not necessarily perpendicular because we are not in a radiative regime. 17. See supporting material on Science Online. 18. IEEE IEEE Std C95.1?2005 Standard for Safety Levels with Respect to Human Exposure to Radio Frequency Electromagnetic Fields, 3 kHz to 300 GHz (IEEE, Piscataway, N), 2006). 19. ]. B. Pendry, Science 306, 1353 (2004). 20. We thank ]. Pendry for suggesting the use of magnetic resonances and M. Grossman and I.Celanovic for technical assistance. Supported by NSFMaterials Research Science and Engineering Center grant DMR02-13282, U.S. Department of Energygrant DE-FG02-99ER45778, and the Army ResearchOffice through the Institutefor Soldier Nanotechnologies under contract DAAD-19-02-D0002. Supporting Online Material www.sciencemag.org/cgi/content/full/1143254/DCl S0M Text Figs. SI to S5

30 March 2007; accepted 21 May 2007 Published online 7 June 2007; 10.1126/science.ll43254 Include this information when citing this paper.

Sea Anemone Ancestral Repertoire

Genome

Reveals Gene Organization

Anthozoan Among anemone Nematostella model

cnidarians, vectensis

the starlet

sea

Eumetazoan and Genomic

anemone coasts

This system (12,13). is found on the Atlantic America, as well

is an emerging estuarine burrowing and Pacific of as the coast

of North

Nicholas H. Putnam/ Mansi Srivastava,2 Uffe Hellsten,1 BUI Dirks,2 Jarrod Chapman,1 Asaf Salamov,1 Astrid Terry,1 Harris Shapiro,1 Erika Lindquist,1 Vladimir V. Kapitonov,3 Jerzy ]urka,3 Grigory Genikhovich,4 IgorV. Grigoriev,1 Susan M. Lucas,1 Robert E. Steele,5 ]ohn R. Finnerty,6 Ulrich Technau,4 Mark Q. Martindale,7 Daniel S. Rokhsar1'2* Sea anemones are seemingly primitive animals that, along with corals, jellyfish, and hydras,
of the oldest eumetazoan the Cnidaria. Here, we report a comparative analysis phylum, cnidarian of an emerging the starlet sea anemone Nematostella model, genome vectensis. The sea anemone is complex, with a gene repertoire, and exon-intron structure, genome that the than to flies or nematodes, gene implying large-scale linkage more similar to vertebrates ancestor was similarly of the eumetazoan of the inferred genes genome complex. Nearly one-fifth constitute the draft of the ancestor signaling, are eumetazoan and synaptic novelties, which are enriched Analysis for animal functions suggests like cell that these adhesion, transmission. of diverse

cultures (14). Nematostella England are easily maintained in the laboratory; with inducible and external separate sexes, spawning, southeast fertilization throughout are available (12, 15), embryos the year. are often character cnidarians study of revealed

Although or primitive, closer ized as simple and its relatives has Nematostella considerable molecular (16-19)

and morpho

logical complexity (75). Based on expressed sequence tag (EST) analyses (17, 18) and the
targeted ways study of specific gene families

[reviewed in (13, 16, 20-22)], signaling path

in the and transcription factors involved and development of bilaterians early patterning are present in cnidarian and are ac genomes

pathways

gene "inventions" along the lineage leading to animals were likely already well integrated with
preexisting All eukaryotic genes in the eumetazoan or the eu last progenitor. conserved identifying an intrinsically slow preserved by (4-6) between selective ancestral rate of features change that have and/or are

tive in development (13, 23-28), indicating that

these predate and pathways the eumetazoan regulatory radiation. mechanisms Perhaps the main also most body

living

metazoans, common worms, ones, jellies) snails,

animals, tissue-grade are descended from ancestor of bilaterians

(flies,

pressures.

jellyfish,

and humans), cnidarians (anem and hydra), and ctenophores (comb eumetazoan ancestor lived (1, 2). This in the fossil record (3), we can infer

perhaps 700 million years ago. Although it is not

preserved many of its characteristics?flagellated sperm, de velopment through a process of gastrulation, mul tiple germ layers, basement membrane, muscular fixed true epithelia lying a lined gut (enteron), upon a neuro

fruit fly, nematode, in reveal greater genomic genomes complexity the vertebrates other deuterostomes (7, [and 8)] as measured by gene content and structure, but at the same time show are shared that many and net genes across bilaterians. the Probing

Comparisons and vertebrate

that establish genes notably, axes are in bilaterian embryos in Nematostella asymmetrically even viewed cussion, cnidarians though as radial animals see are

expressed

development,

conventionally dis [for a critical

works ancestral a from

eumetazoan

even

sensory system, multiple systems, these conserved body axes?because tures are retained by its modem descendants.

and fea

comparing In comparison

genome requires sequences of the animal tree, deeper branches bilaterian and nonbilaterian phyla. with bilaterians, simple. The cnidarians phylum ap is de

(29)]. the draft genome of the Here, we report starlet sea anemone and use its gene repertoire to reconstruct features and genome organization of of the ancestral the Nematostella eumetazoan genome genome. Analysis in the context of

pear morphologically

fined (2) by a sac-like body plan with a single

two epithelial tissue layers, the pres oral opening, ence of numerous tentacles, a nerve net, and the characteristic cells (cnidocytes, stinging literally its translated as "nettle cells") that give the phylum name (fig. S1.1G). includes The diverse class Anthozoa anemones, lack a medusa corals, ("flower and

from other eukaryotes reveals the ge sequences nomic complexity cnidarian of this last common ancestor. The bilaterian picture from emerging

the genome of Similarly, we can characterize this long-extinct eumetazoan progenitor by com DNA and protein and sequences paring modem

the genome and EST studies (17, 18) is one of

extensive in gene content, structure, conservation and organization between Nematostella and ver tebrates. We show that even chromosome-scale linkage has been preserved between Nematostella are the most and vertebrates. These ancient con of prokaryotic served known outside linkages In contrast, the fruit fly and nematode op?rons. extensive model gene systems have experienced loss (18), intron loss (30), and genome rearrange ment. Thus, eumetazoan from ancestor a genomic the perspective, more resembled closely and sea anemones.

department CA 94598,

of Energy Joint Genome Institute, Walnut Creek, USA. 2Center for Integrative Genomics and Department of Molecular and Cell Biology, University of California, Berkeley, CA 94720, USA. 3Genetic Information Research Institute, 1925 Landings Drive, Mountain View, CA 94043, USA. 4Sars International Centre forMarine Molecular Biology, University of Bergen, Thormohlensgt 55, 5008, of Biological Chemistry and Bergen, Norway, department the Developmental Biology Center, University of California, of Biology, Boston Irvine, CA 92697, USA. department University, Boston, MA 02215, USA. 7KewaloMarine Labora tory, University of Hawaii, Honolulu, HI 96813, USA. *To whom correspondence dsrokhsar@lbl.gov should be addressed. E-mail:

animals") sea pens, other medusae

all of which classes

stage. The

cnidarian and

are united

characteristically

by their pelagic linear mitochon

drial genomes (9) into the Medusozoa, including Hydra and related hydroids, jellyfish, and box jellies. The disparate bilaterianphyla of the early
Cambrian the cnidarian indeed some suggest a Precambrian lineage from of the oldest are plausibly of divergence the bilaterian stem, and animal body and em

modem

vertebrates

Nematostella Genome Assembly and Gene Set

The was sequence produced with draft of the Nematostella genome shotgun the use of a random

bryo fossils

relics of stem cnidarians

[reviewed in (10,11)]. 86 6 JULY 2007 VOL 317 SCIENCE

www.sciencemag.org

This content downloaded from 129.8.242.67 on Sun, 23 Feb 2014 21:27:22 PM All use subject to JSTOR Terms and Conditions