DOI: 10.1126/science.1167094 , 1546 (2008); 322Science

et al.Amy McMahon,Provide Insights into Collective Cell Migration

GastrulationDrosophilaDynamic Analyses of

www.sciencemag.org (this information is current as of April 1, 2009 ):The following resources related to this article are available online at

http://www.sciencemag.org/cgi/content/full/322/5907/1546version of this article at:

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

http://www.sciencemag.org/cgi/content/full/322/5907/1546/DC1 can be found at: Supporting Online Material

http://www.sciencemag.org/cgi/content/full/322/5907/1546#otherarticles, 8 of which can be accessed for free: cites 23 articlesThis article

1 article(s) on the ISI Web of Science. cited byThis article has been

http://www.sciencemag.org/cgi/content/full/322/5907/1546#otherarticles 1 articles hosted by HighWire Press; see: cited byThis article has been

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

: subject collectionsThis article appears in the following

http://www.sciencemag.org/about/permissions.dtl in whole or in part can be found at: this article

permission to reproduce of this article or about obtaining reprintsInformation about obtaining

registered trademark of AAAS. is aScience2008 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

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org/cgi/content/full/322/5907/1546http://www.sciencemag.org/cgi/content/full/322/5907/1546/DC1http://www.sciencemag.org/cgi/content/full/322/5907/1546#otherarticleshttp://www.sciencemag.org/cgi/content/full/322/5907/1546#otherarticleshttp://www.sciencemag.org/cgi/collection/developmenthttp://www.sciencemag.org/about/permissions.dtlhttp://www.sciencemag.org

We have uncovered a negative regulationcascade that is essential for successful cyto-kinesis. Although negative regulation has beenproposed to be important during cytokinesis,previous models have emphasized inhibition ofcortical contractility by astral microtubules thatcontact the polar regions of the cell (29, 30). Arequirement for negative regulation of an inhib-itory pathway at the cell equator has not beenwidely considered. Our findings lead to a modelin which inactivation of Rac by CYK-4 GAPfunctions in parallel with activation of RhoA todrive contractile ring constriction during cyto-kinesis (fig. S7).

References and Notes1. M. Glotzer, Science 307, 1735 (2005).2. W. M. Bement, H. A. Benink, G. von Dassow, J. Cell Biol.

170, 91 (2005).3. Y. Nishimura, S. Yonemura, J. Cell Sci. 119, 104 (2006).4. O. Yuce, A. Piekny, M. Glotzer, J. Cell Biol. 170, 571 (2005).5. K. Kamijo et al., Mol. Biol. Cell 17, 43 (2006).6. M. Mishima, S. Kaitna, M. Glotzer, Dev. Cell 2, 41 (2002).7. W. G. Somers, R. Saint, Dev. Cell 4, 29 (2003).8. V. Pavicic-Kaltenbrunner, M. Mishima, M. Glotzer, Mol.

Biol. Cell 18, 4992 (2007).9. W. B. Raich, A. N. Moran, J. H. Rothman, J. Hardin, Mol.

Biol. Cell 9, 2037 (1998).10. J. Powers, O. Bossinger, D. Rose, S. Strome, W. Saxton,

Curr. Biol. 8, 1133 (1998).11. V. Jantsch-Plunger et al., J. Cell Biol. 149, 1391 (2000).12. A. F. Severson, D. R. Hamill, J. C. Carter, J. Schumacher,

B. Bowerman, Curr. Biol. 10, 1162 (2000).13. Y. Minoshima et al., Dev. Cell 4, 549 (2003).14. P. P. D’Avino, M. S. Savoian, D. M. Glover, J. Cell Biol.

166, 61 (2004).

15. T. Yamada, M. Hikida, T. Kurosaki, Exp. Cell Res. 312,3517 (2006).

16. A. Toure et al., J. Biol. Chem. 273, 6019 (1998).

17. T. Kawashima et al., Blood 96, 2116 (2000).18. W. M. Bement, A. L. Miller, G. von Dassow, Bioessays 28,

983 (2006).19. See supporting material on Science Online.20. S. E. Encalada et al., Dev. Biol. 228, 225 (2000).21. S. Ahmed et al., J. Biol. Chem. 269, 17642 (1994).22. K. Rittinger, P. A. Walker, J. F. Eccleston, S. J. Smerdon,

S. J. Gamblin, Nature 389, 758 (1997).23. K. Oegema, A. Desai, S. Rybina, M. Kirkham,

A. A. Hyman, J. Cell Biol. 153, 1209 (2001).24. E. A. Lundquist, in WormBook (www.wormbook.org/

chapters/www_smallGTPases/smallGTPases.pdf).25. H. Yoshizaki et al., J. Biol. Chem. 279, 44756 (2004).26. T. D. Pollard, Annu. Rev. Biophys. Biomol. Struct. 36, 451

(2007).27. A. F. Severson, D. L. Baillie, B. Bowerman, Curr. Biol. 12,

2066 (2002).28. J. Withee, B. Galligan, N. Hawkins, G. Garriga, Genetics

167, 1165 (2004).29. R. Rappaport, Cytokinesis in Animal Cells (Cambridge

Univ. Press, Cambridge, 1997).30. R. Dechant, M. Glotzer, Dev. Cell 4, 333 (2003).31. We thank all members of the Oegema, Desai, and

Bowerman labs; A. Maddox, J. Dumont, and R. Greenfor reading this manuscript; J. Dumont for makingdouble-stranded RNAs; Y. Kohara for cDNA clones; andthe Caenorhabditis Genetics Center for strains. Supportedby the Jane Coffin Childs Memorial Fund for MedicalResearch and the Leukemia and Lymphoma Society( J.C.C.), National Institute of General Medical Sciencesgrant T32 GM008666 and National Cancer Institutegrant T32 CA067754 (L.L.), the Ludwig Institute forCancer Research (K.O. and A.D.), and NIH grantGM058017 (B.B.).

Supporting Online Materialwww.sciencemag.org/cgi/content/full/322/5907/1543/DC1Materials and MethodsFigs. S1 to S7Movies S1 and S2References

10 July 2008; accepted 24 October 200810.1126/science.1163086

Dynamic Analyses of DrosophilaGastrulation Provide Insights intoCollective Cell MigrationAmy McMahon,1* Willy Supatto,2* Scott E. Fraser,2 Angelike Stathopoulos1†

The concerted movement of cells from different germ layers contributes to morphogenesisduring early embryonic development. Using an optimized imaging approach and quantitativemethods, we analyzed the trajectories of hundreds of ectodermal cells and internalized mesodermalcells within Drosophila embryos over 2 hours during gastrulation. We found a high level of cellularorganization, with mesoderm cell movements correlating with some but not all ectodermmovements. During migration, the mesoderm population underwent two ordered waves of celldivision and synchronous cell intercalation, and cells at the leading edge stably maintainedposition. Fibroblast growth factor (FGF) signaling guides mesodermal cell migration; however,we found some directed dorsal migration in an FGF receptor mutant, which suggests thatadditional signals are involved. Thus, decomposing complex cellular movements can providedetailed insights into collective cell migration.

Anembryo is shaped by a complex combi-nation of collective cell movements thatresult in cell diversification and tissue for-mation (1–4). Themajority of thesemorphogeneticevents are dynamic and involve the simultaneousexecution of different movements, with large pop-

ulations of cells moving in three-dimensional (3D)space deep inside the embryo (4, 5). Gastrulationis the earliest morphogenetic event involving mas-sive cellular movements of the germ layers (6).Because it is technically challenging to image in-dividual cellmovements inside an embryowithout

B

An=6wve-1(RNAi)

100%0%0%0%0%

n=5wsp-1(RNAi)100%0%0%0%0%

n=9arx-2(RNAi)100%0%0%0%0%

n=22wve-1(RNAi)CYK-4 GAP(E448K);

0%0%0%100%0%

n=33CYK-4 GAP(E448K);wsp-1(RNAi)

6%33%33%61%0%

n=26CYK-4 GAP(E448K);wsp-1;wve-1(RNAi)

38%31%31%31%0%

n=23CYK-4 GAP(E448K);arx-2(RNAi)

52%22%22%26%0%

30 s % of Embryos0 50 100

Complete CytokinesisFull Ingression Followed by Furrow Regression

Partial Ingression Followed by Furrow RegressionNo Ingression

100%0%0%0%0%

wsp-1;wve-1(RNAi) n=9

Fig. 4. CYK-4 GAP inactivates Rac and its effectors, WASp/WAVE and the Arp2/3 complex, topromote cytokinesis. Cytokinesis in CYK-4GAP(E448K) embryos is rescued by (A) co-depletion ofWASpWSP-1 and WAVEWVE-1 or (B) depletion of Arp2ARX-2. Scale bar, 20 mm.

5 DECEMBER 2008 VOL 322 SCIENCE www.sciencemag.org1546

REPORTS

on

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org

compromising its viability, studies of mesodermcell migration during gastrulation in Drosophilahave relied on the extrapolation of dynamical eventsfrom observations of fixed embryos (Fig. 1, A andB) or from in vivo descriptions of small numbersof cells (7–9).

We used optimized two-photon excited fluo-rescence (2PEF) (10, 11) to image large domainsof Drosophila embryos ubiquitously expressingnuclear green fluorescent protein (GFP) (Fig. 1,C andD) (12) with sufficient spatial and temporalresolution to examine mesoderm spreading non-invasively over 2 hours (Fig. 1E and movie S1)(13). We extracted the complex cell movementsof the mesoderm and ectoderm cells from eachlarge imaging data set (~3 billion voxels) by using3D segmentation of cell positions and 3D trackingover time (Fig. 1, F to H, and movie S2). Thisinvolved the analysis of over 100,000 cell positionsper embryo (movieS3) (13).Weused computationalanalysis to capture the three main morphogeneticevents of the mesoderm (Fig. 1F) and confirmed

that the ectoderm cell layer, upon which meso-derm cells are migrating, undergoes germ-bandelongation by means of convergent extensionmovements (Fig. 1, I and J) (14, 15).

We developed custom software tools to extractquantitative information from the cell trajectoriesand to describe the dynamic behavior in detail (movieS3) (13). First, we redefined the positions of cellsin accordance with a cylindrical coordinate system[radial (r), angular (q), and longitudinal (L)] byfitting a cylinder on the average position of ecto-derm cells. This coordinate system, unlike thestandard Cartesian system (x, y, and z), is moreappropriate for the body plan of Drosophila em-bryos and the geometry of their morphogeneticevents (Fig. 2, A to E, fig. S1, and movie S4)(14, 15).

We determined the influence of ectoderm cellmovements on the migratory path of the over-lying mesoderm by investigating the coupling be-tween the motions of these two cell populations.The ectoderm is in close physical contact with themesoderm: The mesoderm invaginates from theectoderm, and the ectoderm serves as the sub-stratum onwhich themesoderm cells spread duringgerm-band elongation (15, 16). Previous qualita-tive studies suggested a coupling of their move-ments; in mutants that fail to form ectoderm,mesoderm cells are specified but fail to move

(14). Statistical analysis of our data revealed thatthe trajectories of mesoderm and ectoderm cellscorrelate highly in the anterior-posterior (AP) di-rection (the L axis) (Fig. 2H). However, in theother directions (the r and q axes), little to nocorrelation was found (Fig. 2, F and G). Sub-tracting axial motions of the local ectoderm cellsfrom the motion of each mesoderm cell resultedin no residual movement of the mesoderm in theL direction (Fig. 2I andmovies S5 and S6), whichsuggests that the mesoderm cells are carried bythe strongmovement of the ectoderm during germ-band elongation in this direction. The lack of cor-relation in the radial and angular directions suggeststhat mesoderm cells undergo active movement,distinct from that of the ectoderm.

In the angular direction (q), mesoderm cellmovement was symmetrical with respect to theventral midline of the embryo, as demonstratedby a q mean value of 0 (Fig. 2D). Using a colorcode to identify each cell track by its position oforigin in the furrow (Fig. 3A), we revealed a sta-ble chromatic pattern of the trajectories in the qdirection, highlighting the fact that the spatial or-ganization of cells in this direction is preservedover time. The straightness of the trajectories andthe limited intermixing of cells support the viewthat cell movements are directed. The cell trajec-tories revealed that a group of cells originating

1Division of Biology, California Institute of Technology, 1200East California Boulevard, Pasadena, CA 91125,USA. 2BeckmanInstitute, California Institute of Technology, 1200 East CaliforniaBoulevard, Pasadena, CA 91125, USA.

*These authors contributed equally to this work.†To whom correspondence should be addressed. E-mail:angelike@caltech.edu

Fig. 1. Two-photon mi-croscopy and analysis ofhistone2A (H2A)–GFP ex-pressing embryos captureskey events in gastrula-tion. (A and B) Cross-sections of wild-type (A)and htl mutant (B) em-bryos stained with anti-body to Twist. (C and D)Confocal 1PEF (C) fails toimage internalized meso-derm cells, whereas 2PEF(D) captures the positionsof the internalized cells.(E) A 50-mm-deep and10-mm-thick lateral slicethrough an H2A-GFP em-bryo demonstrates thesignal-to-noise ratio (an-terior, left). (F) Segmenta-tion of mesoderm nuclei(orange spheres) by theuse of Imaris software(BitplaneAG,Zurich, Switz-erland). Each sphere wasdefined by the fluores-cent intensity of H2A-GFP.Furrow formation, furrowcollapse as a result of anEMT, and spreading ofthe mesoderm to form amonolayer are illustrated from top to bottom, respectively. (G to J)Tracking cell positions in three dimensions over time. Shown are dorsal (G)and posterior (H) views of mesoderm tracks (blue and yellow indicate early

and late time points, respectively) and dorsal (I) and posterior (J) views ofmesoderm (orange) and ectoderm (gray) net displacement vectors. Scalebars, 20 mm.

www.sciencemag.org SCIENCE VOL 322 5 DECEMBER 2008 1547

REPORTS

on

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org

from the upper lateral parts of the furrow (Fig. 3A)becomes positioned at each leading edge of themesoderm cell population, which was maintainedfor the entire course of theirmigration (movie S7).These leading cells were neither the first nor thelast to invaginate; instead, their locationwithin thefurrow positioned them to land in the leading po-sition as the furrow collapsed after the epithelial-to-mesenchymal transition (EMT).

We explored other morphogenetic events thatmight contribute to mesoderm spreading, such ascell division pattern and cell intercalation, basedon our cell-tracking data. Each mesoderm celldivided twice (7, 8, 17, 18), and these divisionswere ordered in space and time (Fig. 3B). Cellsnearest the ectoderm divided first, followed bycells nearer to the top of the ventral furrow. Thisorder was maintained during the second divisioncycle. Analysis of the cell division mutants didnot uncover any of the characteristic mesodermmigration defects observable in fixed sections(fig. S6) (18). Our tracking data revealed that theorientation of cell divisions within the mesodermis random and that altering the organization ofcell divisions had no effect on mesoderm spread-ing or embryo viability (fig. S7, A to C). Thus, itis unlikely that these organized cell divisions playa role in mesoderm spreading. The radial cell in-tercalation events (19) were synchronous with thesecondwave of cell division (Fig. 3, C andD), butthe orientation of the cell divisions did not seem toplay a causal role in the intercalation motions.Mesoderm cell intercalation contributes to mono-layer formation and spreading (Fig. 3, C and D).

To facilitate comparisons between embryos,we developed a statistical analysis characterizingthe spreading behavior of the mesoderm cells. Assuggested by the spatial organization of the spread-ing (Fig. 3A), the angular positions of each cell atthe onset (qstart) and at the end (qend) of the pro-cess were highly correlated. A plot of starting andending positions revealed a linear relationship(fig. S4, A to C). Given this, linear regressionsthat were applied to the qend(qstart) valuesprovided a measure of both the strength of thespreading (as the slope of the line, A) (fig. S4,D and E) and a quantitative measure of col-lective behavior (the degree of correlation, R)(13).Wild-type cells followed an ordered spread-ing behavior [qend ≈ 2(qstart)], which is sharedby the majority of cells (R > 0.9) (fig. S5). Com-parison of the regression analysis from five wild-type embryos showed the consistency of cellbehaviors (n = 5 embryos and n = 596 cells)(fig. S5).

Previous studies of fixed embryos (8, 9, 20, 21)have suggested that fibroblast growth factor (FGF)signaling is involved in regulating mesoderm cellmigration, but its exact function has remainedelusive. We used our methodology to study thefunction of the FGF signaling pathway on theregulation of gastrulation by analyzing embryosof the FGF receptor mutant heartless (htl) in thesame way as wild-type embryos (figs. S2 and S3and movie S9). We decomposed the cell move-

ments within htl mutant embryos into their com-ponents in r, q, and L (fig. S3, A to C), permittingdirect comparisonswithwild-type embryos (Fig. 2,

C to E). The ectoderm-coupled movements ofmesoderm cells in the L direction were unaffectedin htlmutants (fig. S3F), and we obtained no evi-

Fig. 2. Decomposition and correlative analysis of cell movements with the use of cylindrical coordinates.(A and B) The use of cylindrical coordinates allows the positioning of cells according to the body plan ofthe embryo at stage 6. (C to E) Cell trajectories (blue lines) reveal that each axis corresponds to a mor-phogenetic movement. (C) r is the radial position over time (for example, furrow collapse and intercalation;0 indicates the center of the embryo). (D) q is the angular movement (for example, mesoderm spreadingand ectoderm convergence; 0 indicates the position of the ventral midline). (E) L corresponds to themovement of cells along the length of the embryo (for example, germ-band elongation). In (C) to (E),time (t) = 0 is set as the point when AP movement begins. (F to H) Correlation of the velocity (v) of eachmesoderm cell with its six nearest ectodermal neighbors along the (F) radial, (G) angular, and (H) AP axes,with correlation values of 0.21 T 0.43, 0.08 T 0.18, and 0.90 T 0.06, respectively (n = 3 embryos). (I)Dorsal view of mesoderm cell displacement before (orange) and after (blue) subtraction of local ectodermcell movements.

5 DECEMBER 2008 VOL 322 SCIENCE www.sciencemag.org1548

REPORTS

on

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org

dence for defects in cell-division events (fig. S7D).However, htl mutant embryos displayed meso-derm cell defects that affected both collapse ofthe furrow (r axis) and spreading in the angulardirection (q axis) (fig. S3, A and B). A statisticalanalysis of cellmovement conducted on htlmutant–tracking data showed a scattered distribution ofqend(qstart) values (figs. S4I and S5), resulting inlow spreading and correlation values (A < 1 and

R < 0.5 to 0.7, respectively) (fig. S5C). Values ob-tained with analysis of individual htl embryos orby pooling the cells frommultiple htl embryos (n =3 embryos and n = 284 cells) (fig. S5, B and C)quantitatively demonstrated that a similar dis-ruption of spreading is present in all htl embryos.

Cell tracking analysis revealed that loss ofFGF signaling affected the mesoderm cells non-homogenously (movie S10). In the radial direc-

tion, cells originating from the upper half of thefurrow (“upper-furrow” cells) in general did notcollapse, remaining far from the ectoderm duringthe entire acquisition time (Fig. 4A, fig. 3SA, andmovies S11 and S8). The angular movement ofupper-furrow cells was strongly affected in htlmutants (Fig. 4, B to G). In contrast, the last cellsto invaginate in htl mutants, which make up thelower furrow, behaved in amanner similar towild-type mesoderm cells and could achieve the samedorsal position as the wild type (Fig. 4G). Ourstatistical analysis of cell movements of upper-and lower-furrow cells confirmed the presence oftwo distinct cell behaviors in htl embryos (fig. S5,D and E). Other cell labeling approaches, such asphotoactivatable GFP, can be used to characterizemutant phenotypes, but the limited number of cellsthey can follow (7) make interpretation difficult,especially when there are multiple behaviors, suchas in htl mutant embryos.

Some cells from the upper furrow in htl mu-tants displayed normal positions in the qend(qstart)graph, similar to those of wild-type embryos.These cells were positioned close to the ectodermat the end of spreading (Fig. 4, I and J, and fig. S4J).This suggested that the distance from the ecto-derm might have a major influence on spreadingbehavior. Indeed, the distinction between the twomigratory behaviors observed was more clearwhen analyzing cells that were close to or far fromthe ectoderm (Fig. S5, D and E). We confirmedthis by plotting a qend(qstart) graph using a colorcode for the radial position of the cells at the endof the spreading process (Fig. 4, H and I): The htlcells that followed wild-type behavior [qend ≈2(qstart) such that A = 2] ended up close to theectoderm (Fig. 4I, green), whereas the cells thatstayed far from the ectoderm (Fig. 4I, red) hadclearly disrupted behaviors, with several cellscrossing the midline and migrating in the wrongdirection (A < 0). All wild-type cells ended upclose to the ectoderm (Fig. 4H).

Our analysis provides several insights into thehtlmutant phenotype. First, the primary functionof FGF signaling must be to help all cells withinthe furrow to collapse, directing them toward theectoderm (Fig. 4K). Second, another as-yet un-identified signal must guide the migration of thecells in the angular direction toward the dorsalectoderm, becausemovement is observed even inthe absence of FGF signaling. Third, contact withthe ectoderm is key for the mesoderm to respond tothis guidance cue, because the distance of the meso-dermcells from the ectodermdefines theirmigratorycompetence. Any cell that encounters the ectodermis capable of directed movement in the angulardirection in response to a cue that cannot be solelyFGF-dependent. Movement of the mesoderm cellsmight require contact with the ectoderm to makethem competent to respond to a directional signal, asevidenced in other systems (22–24).

This study demonstrates that stereotypicalmorphogenetic events during embryo develop-ment can be systematically quantified, analyzed,and compared between wild-type and mutant em-

Fig. 3. Quantitative analysis of morphogenetic events reveals a high level of organization in wild-typeembryos. (A) A color code marks the angular position of cells in the furrow at stage 7 and shows thespatial organization as cells move over time. rad, radians. Each line represents the trajectory of one cell.(B) Position and timing of each cell division (colored circle). The color code represents the radialposition in the furrow at stage 7. DNA morphology during cell division in H2A-GFP embryos is shown(top left). (C) Analysis of intercalation events within the mesoderm over time shown as a percentage ofmesoderm cells intercalating (n = 3 embryos). (D) The position of mesoderm cells before and afterintercalation events.

www.sciencemag.org SCIENCE VOL 322 5 DECEMBER 2008 1549

REPORTS

on

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org

bryos by means of the live imaging of largegroups of cells. Complex cell movements are de-composed into particular cell behaviors, reveal-ing a high level of organization and permittingthe interpretation of subtle mutant phenotypes inDrosophila. Future developments in imaging andcell tracking will facilitate this quantitative ap-proach, enabling its application at a larger scaleand in other model systems, to expand the un-derstanding of collective cell migration and em-bryonic development from the molecular level tothat of the entire organism (25).

References and Notes1. P. Rorth, Trends Cell Biol. 17, 575 (2007).2. D. J. Montell, Curr. Opin. Genet. Dev. 16, 374

(2006).3. C. D. Stern, Gastrulation: From Cells to Embryo.

(Cold Spring HarborLaboratory Press, Cold Spring Harbor,NY, 2004).

4. V. Lecaudey, D. Gilmour, Curr. Opin. Cell Biol. 18, 102(2006).

5. L. A. Rohde, C. P. Heisenberg, Int. Rev. Cytol. 261, 159(2007).

6. M. Leptin, Dev. Cell 8, 305 (2005).7. M. J. Murray, R. Saint, Development 134, 3975

(2007).8. R. Wilson, E. Vogelsang, M. Leptin, Development 132,

491 (2005).9. S. Schumacher, T. Gryzik, S. Tannebaum, H. A. Muller,

Development 131, 2631 (2004).10. F. Helmchen, W. Denk, Nat. Methods 2, 932 (2005).11. W. Supatto et al., Proc. Natl. Acad. Sci. U.S.A. 102, 1047

(2005).12. M. Clarkson, R. Saint, DNA Cell Biol. 18, 457 (1999).13. Materials and methods are available as supporting

material on Science Online.14. K. D. Irvine, E. Wieschaus, Development 120, 827

(1994).15. J. A. Zallen, J. T. Blankenship, Semin. Cell Dev. Biol. 19,

263 (2008).16. R. Wilson, M. Leptin, Philos. Trans. R. Soc. London Ser. B

355, 891 (2000).17. T. C. Seher, M. Leptin, Curr. Biol. 10, 623 (2000).18. J. Grosshans, E. Wieschaus, Cell 101, 523 (2000).19. O. Voiculescu, F. Bertocchini, L. Wolpert, R. E. Keller,

C. D. Stern, Nature 449, 1049 (2007).20. S. Gisselbrecht, J. B. Skeath, C. Q. Doe, A. M. Michelson,

Genes Dev. 10, 3003 (1996).21. M. Beiman, B. Z. Shilo, T. Volk, Genes Dev. 10, 2993

(1996).

22. M. Krieg et al., Nat. Cell Biol. 10, 429 (2008).23. M. Sato, T. B. Kornberg, Dev. Cell 3, 195 (2002).24. X. Yang, D. Dormann, A. E. Munsterberg, C. J. Weijer,

Dev. Cell 3, 425 (2002).25. S. G. Megason, S. E. Fraser, Cell 130, 784 (2007).26. We thank M. Levine and E. Meyerowiz for

comments on the manuscript, M. Liebling foradvice on Imaris and Matlab, and the CaltechBiological Imaging Center for sharing equipment.This work was supported by grants to A.S. from NIH(R01 GM078542), the Searle Scholars Program,and the March of Dimes (Basil O'Conner StarterScholar Award, 5-FY06-12), and grants to S.F. fromthe Caltech Beckman Institute and NIH (Centerfor Excellence in Genomic Science grant P50HG004071).

Supporting Online Materialwww.sciencemag.org/cgi/content/full/322/5907/[page]/DC1Materials and MethodsFigs. S1 to S7Movie S1 to S11References

16 May 2008; accepted 12 November 200810.1126/science.1167094

Fig. 4. Furrow collapse and spreading of mesoderm cells are disrupted in htlmutants. (A) Position of mesoderm cells (circles) at stage 7 and stage 10 in wild-type and htl embryos shown with a radial color code. (B toG) Angular movementof cells over time analyzed in wild-type [(B) to (D)] and htl mutant [(E) to (G)]embryos within the entire [(B) and (E)], upper [(C) and (F)], and lower [(D) and(G)] furrow (black line indicates the averagemesoderm displacement with respectto the midline). (H and I) Spreading profile of wild-type (H) and htl (I) embryos.The color code represents the distance from the ectoderm at the end of spreading(red indicates far from ectoderm and green indicates close to ectoderm). The grayline represents a spreading coefficient of A = 2, where qend = A(qstart) + B [B,

constant (13)]. Cells that do not spread within the collective are representedwithin gray regions of the graph (13). In general, cells located close to theectoderm fall along the gray line. (J) The radial position (r) of two particulargroups of mesoderm cells from the upper furrow of htl mutants is depictedover time. One group exhibits normal spreading behavior (light blue), and theother group exhibits aberrant spreading (dark blue). (K) The furrow collapse inhtl mutants is disrupted, resulting in cells falling randomly to one side of theembryo. Upper-furrow cells that reach the ectoderm (light blue) undergonormal spreading, whereas cells that remain far from the ectoderm spreadabnormally (dark blue). Red cells are Red indicates lower-furrow cells.

5 DECEMBER 2008 VOL 322 SCIENCE www.sciencemag.org1550

REPORTS

on

Apr

il 1,

200

9 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

http://www.sciencemag.org