Mechanical control of notochord morphogenesis by extra-embryonic tissues in mouse embryos

MECHANISMS OF DEVELOPMENT

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

Available at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/modo

Mechanical control of notochord morphogenesis by extra-embryonic tissues in mouse embryos Yu Imuta a,b,c, Hiroshi Koyama d,e, Dongbo Shi Fujimori d,e, Hiroshi Sasaki a,h,*

d,f

, Mototsugu Eiraku g, Toshihiko

a

Department of Cell Fate Control, Institute for Molecular Embryology and Genetics, Kumamoto University, 2-2-1 Honjo, Chuo-ku, Kumamoto 860-0811, Japan b Laboratory for Cell Asymmetry, RIKEN Center for Developmental Biology, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan c Graduate School of Medicine, Osaka University, 2-2 Suita, Osaka 565-0871, Japan d Division of Embryology, National Institute for Basic Biology, 5-1 Higashiyama, Myodaiji, Okazaki 444-8787, Japan e Graduate University of Advanced Studies (Sokendai), 5-1 Higashiyama, Myodaiji, Okazaki 444-8787, Japan f Graduate School of Biostudies, Kyoto University, Yoshidakonoecho, Sakyo-ku, Kyoto 606-8501, Japan g 4-Dimensional Tissue Analysis Unit, RIKEN Center for Developmental Biology, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan h Program for Leading Graduate Schools ‘‘HIGO (Health life science: Interdisciplinary and Glocal Oriented) Program’’, Kumamoto University, 2-2-1 Honjo, Chuo-ku, Kumamoto 860-0811, Japan

A R T I C L E I N F O

A B S T R A C T

Article history:

Mammalian embryos develop in coordination with extraembryonic tissues, which support

Received 17 January 2014

embryonic development by implanting embryos into the uterus, supplying nutrition, pro-

Received in revised form

viding a confined niche, and also providing patterning signals to embryos. Here, we show

26 January 2014

that in mouse embryos, the expansion of the amniotic cavity (AC), which is formed

Accepted 29 January 2014

between embryonic and extraembryonic tissues, provides the mechanical forces required

Available online 7 February 2014

for a type of morphogenetic movement of the notochord known as convergent extension (CE) in which the cells converge to the midline and the tissue elongates along the ante-

Keywords: Mouse embryogenesis Notochord Convergent extension Live imaging Amniotic cavity Mechanical forces

1.

ro-posterior (AP) axis. The notochord is stretched along the AP axis, and the expansion of the AC is required for CE. Both mathematical modeling and physical simulation showed that a rectangular morphology of the early notochord caused the application of anisotropic force along the AP axis to the notochord through the isotropic expansion of the AC. AC expansion acts upstream of planar cell polarity (PCP) signaling, which regulates CE movement. Our results highlight the importance of extraembryonic tissues as a source of the forces that control the morphogenesis of embryos.

Introduction

Mammalian embryos develop inside of the liquid-filled cavities that are formed by extraembryonic tissues (Fig. 1A).

 2014 Elsevier Ireland Ltd. All rights reserved.

The development of embryos is coordinated with extraembryonic tissue formation, which support embryonic development by implanting embryos into the uterus (Cha et al., 2012), supplying nutrition (Bielinska et al., 1999), providing a confined

* Corresponding author at: Department of Cell Fate Control, Institute of Molecular Embryology and Genetics, Kumamoto University, 2-2-1 Honjo, Chuo-ku, Kumamoto 860-0811, Japan. Tel.: +81 96 373 6606; fax: +81 96 373 6609. E-mail address: [email protected] (H. Sasaki). http://dx.doi.org/10.1016/j.mod.2014.01.004 0925-4773/ 2014 Elsevier Ireland Ltd. All rights reserved.

MECHANISMS OF DEVELOPMENT

niche (Schmidt, 1992), and also providing patterning signals (Arnold and Robertson, 2009; Tam and Loebel, 2007) to embryos. Growing evidence suggests the importance of mechanical signals in the control of cellular responses and embryonic development (Halder et al., 2012; Hiramatsu et al., 2013; Orr et al., 2006; Vogel and Sheetz, 2009). Because embryos are connected with extraembryonic tissues, the forces generated by the embryonic/extraembryonic tissues also likely control the development of these tissues. To understand the mechanical roles of extraembryonic tissues on mouse embryonic development, we used mouse embryos and focused on the evolutionarily conserved morphogenesis of the notochord, specifically, convergent extension (CE) movement of the notochord (Keller, 2002; Keller et al., 1989; Sausedo and Schoenwolf, 1993, 1994; Wallingford et al., 2002; Wood and Thorogood, 1994) in which the cells converge to the midline and the tissue elongates along the antero-posterior (AP) axis. Studies in Xenopus showed that CE of the notochord relies on cell rearrangement (Keller, 2002; Wallingford et al., 2002). Wnt/Frizzled/planar cell polarity (PCP) signaling is required for the polarization of cells (Keller, 2002; Myers et al., 2002; Wallingford et al., 2002), and the AP patterning systems acting upstream or in parallel determine the direction of tissue elongation (Ninomiya et al., 2004). PCP signaling also plays important roles in CE of the mouse notochord (Wang et al., 2006; Ybot-Gonzalez et al., 2007; Yen et al., 2009), but cellular behaviors during CE in the mouse notochord has not been studied in detail. In this study, we performed time-lapse imaging of cellular behaviors during CE in the mouse notochord, and found that the cellular behaviors are different from those of Xenopus notochord. We also revealed that, in mouse embryos, the force generated by expansion of the amniotic cavity (AC) plays critical roles in CE of the notochord.

2.

Results

2.1.

Mouse notochord is stretched along the AP axis

To understand the cellular basis of notochord morphogenesis, we performed time-lapse imaging of the notochord during CE movements. To monitor the cell movements, we used Foxa2nEGFP–CreERT2 embryos, which express enhanced green fluorescent protein (EGFP) in the nuclei of the notochord and the endoderm (Imuta et al., 2013), and time-lapse movies were obtained between the early head fold (EHF) (Downs and Davies, 1993) and approximately 5-somite stages. A dissected EHF stage embryo was fixed in a modified culture dish, which was placed on the inverted confocal microscope, as to visualize the posterior notochord region (Yamanaka et al., 2007) (Fig. 1B), and were cultured with imaging for approximately 10 h. At this stage, the notochord is a one-cell-thick sheet continuous with the endoderm and is exposed to the visceral yolk sac cavity (Fig. 1A). During this period, the notochord undergoes CE morphogenesis (Sausedo and Schoenwolf, 1994), and our live-imaging analyses revealed dynamic cell rearrangements during the CE movements (Fig. 1C–F, Movie S1). Unexpectedly, the notochord cell nuclei were initially oriented toward the left–right (LR) axis, but gradually reoriented toward the AP axis

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

45

(Fig. 2A–F, Movie S2). Such reorientation of the notochord cells has not been previously reported in Xenopus notochord studies (Keller, 2002; Keller et al., 1989; Wallingford et al., 2002). The orientation of the cell division axis was also biased toward the AP and the LR axes in the notochord and endoderm, respectively (Fig. 2G, Movie S3). Such cellular behavior gave us an impression that the notochord is stretched along the AP axis during CE, and that the cells are rearranged and reoriented to support the rapid extension of this tissue. To examine whether the notochord is stretched along the AP axis, we ablated a notochord cell by using laser-pulse irradiation and examined the movement of the cells surrounding the ablated cell (Fig. 2H, Movie S4). Because ablation of a cell releases the tension from the surrounding cells, the released cells move toward the direction of the tension that the cells are receiving. Wound healing slowly occurred following the rapid relaxation behavior. Therefore, the distance of cell movement immediately after ablation mainly reflects the effects of relaxation behaviors and, thus, correlates with the strength of the tension. Upon ablation of a notochord cell, the cells located anterior (A) or posterior (P) to the ablated cell showed movements that were two to three times larger than those of the cells located at the left (L) or right (R) positions (Fig. 2H, and J–L). Ablation of a nearby endoderm cell caused smaller movements of the surrounding cells; the cells at the LR positions had a tendency to show larger movements (Fig. 2I, M and N). These results are consistent with the hypothesis that the notochord cells receive strong anisotropic tension along the AP axis, and that the neighboring endoderm cells experience weaker tension, which is slightly biased toward the LR axis.

2.2. AC expansion is required for proper notochord morphogenesis Notochord extension progresses with the growth of the embryonic and extraembryonic tissues, including the expansion of the AC (Fig. 1A). We hypothesized that the expansion of the AC could be a source of the force required for CE. To test this hypothesis, we inserted an open-tip fine glass capillary into the amniotic cavity by piercing through the visceral yolk sac and the amnion (Fig. 3A), which suppressed the accumulation of the inner pressure and the expansion of the AC (Fig. 3A, B, E and N). The capillary remained in situ throughout the culture period, and another capillary was inserted after 4 h to overcome clogging of the first capillary. In pierced embryos, the elongation of the body axis and CE of the notochord was reduced, and the notochord cells remained oriented toward the LR axis at the end of in vitro culture (Fig. 3C, D, and F–I). The oriented division of the notochord cells along the AP axis was also reduced (Fig. 3J). Despite the morphological changes in the pierced embryos, the expression of Foxa2 and T genes in the notochord (Ang and Rossant, 1994; Herrmann, 1991; Sasaki and Hogan, 1993), which was monitored by the knock-in of H2B-EGFP into the respective loci (Imuta et al., 2013), were comparable between the pierced and control embryos (Fig. 3K and L). Furthermore, the rate of somite formation at this stage is faster than later stages (about one pair in every hour) (Tam, 1981), and this fast rate was not

46

MECHANISMS OF DEVELOPMENT

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

amniotic cavity

A

Anterior decidua

chorion

Posterior ectoplacental cone exocoelomic cavity

ectoplacental cavity

neuroectoderm Reichert's membrane

amnion allantois trophoblast

parietal endoderm

visceral endoderm amniotic cavity

definitive endoderm

visceral yolk sac cavity

mesoderm

notochord definitive endoderm

node notochord

0 min

B

200 min

400 min

600 min

C

D

E

F

C’

D’

E’

F’

hair

Objective lens

A

P

Fig. 1 – Schematic presentation of organization of an E8.0 embryo in decidua, live-imaging set up, and a representative imaging result showing CE movement of the mouse notochord. (A) The left panel shows a sagittal section (anterior is left). The embryo is shown in pink and red. The notochord and the node are shown in red. Extraembryonic parts are shown in dark brown or grey. Note that the amniotic cavity is surrounded by both embryonic and extraembryonic tissues. The middle panel shows an overhead view of a cross section of an embryo. The right shows an enlargement around the notochord. Note that the notochord is continuous with the endoderm and is exposed on the outside of the embryo at this stage. (B) Liveimaging set up. Embryo was placed in a glass-bottom culture dish filled with a culture medium as to face the posterior notochord to the bottom. To achieve this, embryo was hanged to a handmade holder with a hair piercing through the ectoplacental cone. (C–F) Snapshot images of a time-lapse movie of a Foxa2nEGFP–CreERT2 mouse notochord between EHF and approximately 5-somite stages, which show CE movements. Cell positions are marked by dots with color codes. (C 0 –F 0 ) Cell positions marked in C–F. Positions of representative cells that show convergent extension movements are shown with larger dots. The nearby notochord cells were intercalated with each other and were rearranged from LR positions to AP positions, after which the notochord became narrower.

affected (Fig. 3M). These results suggest that the piercing did not significantly compromise the health of the embryos. Therefore, the most likely interpretation of these results is that AC expansion is required for CE movement and reorientation of the notochord cells, although we cannot completely rule out other unknown effects of embryo manipulation. To further examine the importance of the AC expansion for the reorientation of the notochord cells, we next analyzed

freshly dissected embryos, which showed moderate AC expansion. The dissected embryos showed enhanced AC expansion during in vitro culture (Beddington, 1981), likely because the removal of the surrounding visceral yolk sac cavity and decidua (Fig. 1A) eliminated the forces that antagonized AC expansion (Hiramatsu et al., 2013). In correlation with the weaker AC expansion, the freshly dissected embryos showed slightly reduced CE and reorientation of the notochord cells (Fig. 4). Taken

MECHANISMS OF DEVELOPMENT

A

B

47

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

C

D

A R

L P 180 min

E

F orientation (deg.)

A θ θ notochord cell

360 min

540 min

**** **** **** **** **** **** 147 171 156 90 162

G **** cell division (deg.)

0 min

264

noto

endo

0

0 0

P

before ablation

43

90

180 360 540 time (min)

after ablation

H

H’

J

notochord

anterior 512 (pixel) A

L

R

I

0

I’

(LΔ[, LΔy)

endoderm

P 1 1 posterior

notochord

K

shift (pixel)

8

endoderm

L

8 13

A

6 4 9

N

M

15

2

L

R P

shift (pixel)

**** **** **** **** 17

512 (pixel)

A

6 4 2

R

**** 12

12

7

7

L P

0

0 LΔx RΔx AΔy PΔy

LΔx RΔx AΔy PΔy

Fig. 2 – The notochord is stretched along the AP axis. (A–D) Representative snapshot images of a time-lapse movie between EHF and approximately 5-somite stages of an embryo around the notochord. Orientations (long axes) of notochord cells are indicated with yellow lines. (E) Schematic representation for measurements of orientations (angles) of cell (nucleus) and cell division. The orientations are indicated by the angles from the AP-axis. (F) Change of orientation of notochord cells during in vitro development. Orientation of the cells outside the area shown in panels A–D was also included in the graph. Sample numbers are shown above the columns. Data were obtained from time-lapse movies of three embryos. (G) Distribution of cell division orientation revealed by time-lapse movies. noto: notochord, endo: endoderm. Sample number is shown above each column. Data were obtained from two (notochord) and three (endoderm) embryos. (H–N) Laser ablation experiments showing the presence of anisotropic tension along the AP axis in the notochord. Snap shot images of a representative time-lapse movie of laser ablation of a notochord (H, H 0 ) and an endoderm (I, I 0 ) cell. The arrow in the left panels indicates the ablated cell. Positions of the nuclei of the cells surrounding the ablated cell before and after ablation are shown in green and red, respectively. (J) Schematic representation of the quantification method of shift distances of cell position. Surrounding cells were classified to L (left), R (right), A (anterior), and P (posterior) depending on their positions to the ablated cells, and their movement along the x-axis (for L and R) and y-axis (for A and P) are used for analyses shown in panels K–N. A graph summarizing the shift of cell positions by laser ablation of a notochord (K) or endoderm (M) cells. Number of nuclei examined is indicated above each bar. Data were obtained from 8 (K) or 7 (M) embryos. ****: p < 0.0001. (L and N) Vector representations of the data shown in K and M, respectively.

48

MECHANISMS OF DEVELOPMENT

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

A control embryo

immunostain with anti-Foxa2 antibody

in vitro culture for 8 hrs E7.5 glass capillary

B

analyze width of notochord (H) and orientation of notochord cell (I)

analyze orientaion of cell division (J)

D

C

control

H

width of notochord

**

50

width (mm)

G

pierced

F

orientation of cell

5

0

0 ctrl

division (deg.)

600

L ctrl pierce

ctrl pierce

control

11

11

200

ctrl

2

size of cavity

17

1000

**** 10

1.0

500

0

4

1.5

ns 15

15

ctrl pierce

N

1500

12

0

pierce

pierced

pierce

ns 6

ns

400

0

Foxa2

0

800

intensity (/pixel)

90

34

M

pierced

somite number

* 43

endoderm ns 133 102

control

ctrl

pierce

area (cm2)

notochord

K

intensity (/pixel)

orientation of cell division

T

J

**** 240

234

90 5

E

I

orientation (deg.)

pierced embryo

11

0.5

ctrl pierce

0 ctrl pierce

Fig. 3 – Expansion of AC is required for proper notochord morphogenesis. (A) Schematic representation of experimental design. (B, C, E and F) External views of control (B, C) and AC-pierced (E, F) embryos at the end of in vitro culture. Side views of conceptus (B, E), and dorsal views of dissected embryos (C, F). (D, G) Representative images of the notochord in control (D) and pierced (G) embryos. (H–J) Graphs showing distribution of the width of the notochord (H), orientation of the notochord cells (I), and orientation of cell division (J) in control (ctrl) and pierced (pierce) embryos. The number above each column indicates the number of embryos (H) or the number of the cells (I, J) examined. Data for panels I and J was obtained from 8 (notochord) and 7 (endoderm) embryos, respectively. *: p < 0.05, **: p < 0.01, ****: p < 0.0001, ns: not significant. (K, L) Comparison of gene expression in the notochord between control and pierced embryos. Expression of T (K) and Foxa2 (L) genes visualized by histone H2B-EGFP fusion proteins expressed in respective knock-in mice. Representative images (left panels) and quantification of signal intensities (right graphs). (M) Comparison of somite numbers between control and pierced embryos at the end of the culture. ns: not significant. (N) Comparison of cavity sizes between control and pierced embryos at the end of the culture. ****: p < 0.0001.

MECHANISMS OF DEVELOPMENT

together, these results support the hypothesis that AC expansion controls CE and reorientation of the notochord cells.

2.3. CE movement is controlled by the morphology of the notochord Expansion of the AC should stretch the overlying embryonic tissues. Although expansion of the AC is likely unbiased, our results are consistent with the hypothesis that AC expansion provides AP-biased force to the notochord cells. Therefore, we next examined whether unbiased AC expansion causes AP-biased application of the stretching forces to the notochord cells by using mathematical modeling (Fig. 5). AC expansion is a three-dimensional (3D) morphogenesis that can be approximated by the expansion of a spherical substrate (Fig. 5A). To simplify the model, however, we generated a twodimensional (2D) model in which AC expansion was represented by the 2D expansion of a flat substrate (Fig. 5B), and the notochord was represented as a rectangle object placed on the substrate (Fig. 5C). The notochord was defined as a grid, which mimicked the cells, and each vertex of the grid received forces derived from the surface tension (Ft, constant force) or the elastic spring (Fs, variable force that correlates with side length) defined for each side of the grid (Fig. 5D-i) and from the area elasticity (Fa, the force that maintained the area of each grid) defined for each grid (Fig. 5D-ii). Each vertex also received a friction force from the expanding substrate (Fe) (Fig. 5D-iii). To clarify the effects of notochord morphology

EHF

B

in vivo

on the forces that the notochord cells receive with AC expansion, we did not consider cell division or cell rearrangement. Although these assumptions clearly differ from the actual CE properties in the notochord, under these conditions, changes in notochord morphology reflect the forces that the entire notochord cell population receives. Thus, the forces can be estimated from the morphological changes. If no force was considered within the cells (the vertex only received the friction force, Fe), the notochord was enlarged in a homothetic shape (Fig. 6A, A 0 ). When tension and area elasticity were considered (Model 1, Fig. 5D-iv), the notochord was extended in the direction of the long side (AP axis) but was shortened in the direction of the short side (LR axis) (Fig. 6B, B 0 , Movie S5). Essentially the same morphological change was observed with broad range of the parameter values. When extreme parameter values, i.e. very strong area elasticity or surface tension, were used, the notochord either did not alter the morphology or simply became smaller [cells were compressed], respectively. When the original notochord shape was defined as a square, it was extended in the diagonal directions and was shortened in the other directions (Fig. 6C, C 0 ). Because diagonal axes are longer than each side, this result also indicates that the notochord is stretched in the direction of the longer axis. Similar results were obtained when the elastic force, instead of the surface tension, and the area elasticity were considered (Fig. 6D, D 0, Model 2, Fig. 5Div). These morphological changes indicate that the notochord cells receive biased stretching forces along the longer axis.

LHF

A

49

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

2-somite

5-somite

C

D

e

F

**** **** cell orientation (deg.)

**** 90

237

197

**** 216

**** 182

width of notochord 100

5

80 width (μm)

E

5 60 5 40

9

20 0

EHF

LHF 2-somite 5-somite

0

EHF

LHF 2-somite5-somite

Fig. 4 – Changes in orientations of notochord cells during CE movements in freshly dissected embryos (A–D) Representative images of the notochord of freshly dissected Foxa2nEGFP–CreERT2 mouse embryos. EHF: early head fold stage, LHF: late head fold stage. The stages shown in panels A–D approximately correspond to the stages shown in Figs. 1C–D and 2A–D. (E) Distribution of the orientation of the notochord cells in freshly dissected embryos. Sample numbers are shown above the columns. Data were obtained from two or three embryos for each stage. ****: p < 0.0001. (F) Distribution of width of the notochord in freshly dissected embryos. 5-somite stage corresponds to the end of in vitro culture (compared with Fig. 3H). Sample numbers are shown above the columns.

50

MECHANISMS OF DEVELOPMENT

A

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

3D AC expansion

increase of arc length

Vs

AC

arc origin expansion

B

2D AC expansion expanding field origin expansion Vs

C

Notochord modeling expanding field notochord expansion

original position of notochord

notochord friction

Force definition

ii Ft ,Fs

iii Fa

+

iv gin

i

ori

D

Fe

+

Fsum

=

Vv =

Fsum

µ

Model1: Fe

Fsum = Ft + Fa + Fe Model2:

surface tension (Ft) or area elasticity (Fa) elastic spring (Fs)

frictional force accompanying expansion of the substrate (Fe)

Fsum = Fs + Fa + Fe

Fig. 5 – Schematic presentation of the mathematical model The morphological change in the notochord was mathematically simulated under a 2-dimensional expanding field that mimicked AC expansion. (A) In the 3-dimensional situation of AC expansion, the arc on the spherical AC is elongated during AC expansion with a rate proportional to the arc length. The velocity vector Vs of a point (solid blue point) on the sphere is shown in the case that an origin is set at the open blue point. (B) 3-dimensional AC expansion was reduced to a 2-dimensional situation in which an expanding field (light blue disc) was considered. All of the points on the field move away from the origin (open blue point) with a rate proportional to the distance from the origin. For instance, the velocity vectors Vs of four points are shown (solid blue points). Similarly, the distance between arbitrary two points expands with a rate proportional to the distance. (C) The notochord was placed on the 2dimensional expanding field. The notochord was modeled as grids, which mimicked cells (black). In our model, the notochord was assumed to receive frictional forces from the expanding field. Lower panels represent the side-views. The original position of the notochord after the expansion is also presented (gray). (D) Mechanical elements are shown. (i) A surface tension or an elastic spring was defined on each side of the grids, one of which was introduced in a model named Model 1 or Model 2, respectively. The forces derived from the surface tensions or the elastic springs are presented as Ft or Fs, respectively. If a side was the boundary of two grids, two surface tensions or elastic springs were defined. For instance, the red vertex receives forces from the four sides (green lines) with eight surface tensions or elastic springs (black arrows). When a surface tension is defined to be larger than 0 or the length of a side is larger than the natural length of the spring, Ft or Fs becomes negative, respectively, resulting in the red vertex being pulled by the side. Note that the surface tensions were set to be constant independently of the length of the sides, leading to Ft also being constant. (ii) Area elasticity, which exerts forces Fa to the vertices of the grid to maintain the area of the grid, was defined on each grid. For instance, the red vertex receives forces from the four grids (green squares and black arrows). In the case that the area of the grid is smaller than the natural area of the grid, Fa is positive, resulting in the vertices being pushed outwardly. Due to the elasticity area, both the area of the grid and the total area of the notochord were maintained nearly constant during our simulations. (iii) The expanding field was defined to exert the frictional forces Fe to the vertices. By assuming that the Fe were proportional to Vs and inertial effect was negligible, the free vertices, which corresponded to the absence of Ft, Fs, and Fa, moved with Vs, resulting in the gray position in C. The direction and the magnitude of the forces at three points are intuitively presented (gray arrows). A lower panel represents the side-view. (iv) The velocity of the vertex movement Vv was assumed to be proportional to the total force that the vertex received. The total force was the sum of the forces described in (i)–(iii), and Ft or Fs was considered in Model 1 or Model 2, respectively. The inertial effect was neglected and the friction dominated. l is the coefficient of friction. The definitions of the parameters and the details of the model are described in Section 4.

MECHANISMS OF DEVELOPMENT

Extending Model 1, we next added the surrounding endoderm cells, which have a much weaker tension and area elasticity. Consistent with the experimental results, the notochord was extended along the AP axis (long side) and shortened along the LR axis (short side), and extension of the neighboring endoderm was larger along the LR axis than that along the AP axis (Fig. 6E, E 0 , Movie S6). These results suggest that the rectangular shape of the original notochord causes its biased stretching, i.e. biased application of the stretching forces to the cells, along the AP axis with unbiased AC expansion. The mathematical modeling revealed application of APbiased stretching forces in the rectangular objects, similar to the notochord. However, cellular behaviors in these models appear to differ from those occurring in the actual notochord. In the mathematical model, no cellular rearrangement takes place, and morphological changes in the individual cells support alteration of tissue morphology. By contrast, in the notochord, the contribution to morphological changes of the cells is limited, and cell rearrangements mostly support extensive tissue elongation (Figs. 1 and 2). Therefore, to further refine the hypothesis, we next performed physical simulations that considered cell rearrangements. Small metal beads representing notochord cells were placed on a latex membrane, and a small amount of water was added to connect the beads with the surface tension of the water. Then, the latex membrane was uniformly stretched in all directions mimicking unbiased AC expansion (Fig. 6F). In this simulation, no cell division or morphological change of the individual cells were considered. When the beads were initially placed in a rectangular shape, the beads rearranged and extended in the direction of the long axis and shortened along the short axis (Fig. 6G, G 0 , Movie S7). When the beads were placed in a hexagonal shape, no rearrangement was observed (Fig. 6H, H 0 ). Therefore, both mathematical modeling and physical simulations indicated that a combination of isotropic expansion of a substrate and an anisotropic shape of a material on the substrate is sufficient to cause application of the biased stretching forces along the longer axis and anisotropic extension of the material. Because the adherence of cells to their neighbors is a prerequisite for both models, cell–cell adhesion in the notochord is essential to operate this mechanism. This conclusion supports the hypothesis that isotropic expansion of the AC applies AP-biased anisotropic stretching forces to the notochord to support its extension along the AP axis.

2.4. AC expansion regulates notochord morphogenesis upstream of PCP signaling PCP signaling is important for CE movement (Keller, 2002; Wallingford et al., 2002). To understand the relationship between expansion force and PCP signaling, we next analyzed embryos that had a homozygous mutation in Vangl2, a PCP gene (Montcouquiol et al., 2003). When Vangl2Lp/Lp embryos were cultured in vitro, the notochord of the Vangl2Lp/Lp embryos remained short and wide, as reported for Vangl2Lp/Lp and Dvl1/;Dvl2/embryos (Wang et al., 2006; Ybot-Gonzalez et al., 2007) (Fig. 7A, C and E). In addition, the orientation of the notochord cells remained oriented in the LR direction (Fig. 7C and F), suggesting that PCP signaling is also required for the orientation change of the notochord cells during CE.

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

51

The orientation of the notochord cells of the Vangl2Lp/Lp embryos showed a stronger bias toward the LR axis than toward those of the AC-pierced embryos (Fig. 7A, B, C and F), and suppression of the AC expansion in the Vangl2Lp/Lp embryos did not enhance the LR orientation bias of the notochord cells (Fig. 7D and F). The development of the Vangl2Lp/Lp embryos was slightly delayed (Fig. 7I), but the comparison of stagematched embryos (4–5 somite stages) revealed that AC expansion was comparable between the normal and Vangl2Lp/Lp embryos (Fig. 7J). These results suggest that the AC expansion force acts either upstream of or in parallel to PCP signaling. To further clarify the relationship between AC expansion and PCP signaling, we next examined the distribution of Vangl2 proteins in AC-pierced embryos. In the AC-pierced embryos, the membrane accumulation of Vangl2 protein was significantly reduced (Fig. 7K–M). The reduction of Vangl2 protein was specific because the expression of Foxa2, T, and cortical F-actin in the notochord were not significantly altered in the pierced embryos (Fig. 3K, L, 7K 0 , L 0 and N). These results support the hypothesis that the AC expansion force acts upstream of PCP signaling by sustaining the membrane accumulation of Vangl2.

3.

Discussion

Mouse embryogenesis is regulated by several physical constraints or mechanical forces. The establishment of the embryonic-abembryonic axis of a blastocyst is regulated by the physical constraints conferred by the elliptical shape of the zona pellucida that surrounds the embryo (Kurotaki et al., 2007; Motosugi et al., 2005). At 5.5 dpc, the formation of the distal visceral endoderm, which controls anterior development, requires the external force from the decidua to constrain the lateral expansion of the egg cylinder (Hiramatsu et al., 2013). Our current study suggests the importance of another force for notochord morphogenesis that is generated by AC expansion. Based on the results, we propose a model of mechanical control of notochord morphogenesis via AC expansion. According to this model, AC expansion provides the force that stretches the notochord along the AP axis (Fig. 7O). The rectangular morphology of the notochord is the likely mechanism that applies anisotropic forces to the notochord. The CE of the mouse notochord consists of two types of cellular behaviors. One is lateral intercalation (Sausedo and Schoenwolf, 1994), which is evolutionary conserved and is regulated by PCP signaling (Wang et al., 2006; Ybot-Gonzalez et al., 2007; Yen et al., 2009), and the other is cell reorientation. The mechanical forces generated by AC expansion likely control both mechanisms. To our knowledge, this is the first report of the mechanical role that extraembryonic tissues play in the development of mouse embryos. Although we only examined the roles of expansion of the AC, the adjacent exocoelomic cavity, which is separated from the AC by a very thin amnion, also expands, and coordinated expansion of both cavities are probably important for production of the forces generated by AC expansion. Our results also suggest that the AC expansion force acts upstream of PCP signaling. Because mechanical stress controls PCP signaling in Drosophila tissues (Aigouy et al., 2010; Olguin et al., 2011),

52

MECHANISMS OF DEVELOPMENT

tension+area elasticity (Model 1)

no force

A

A’

start

B

end

B’

start

D’

start

C’

C

end

elastic force + area elasticity (Model 2)

D

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

start

end

tension+area elasticity (Model 1)

E’

E

end

end

start before extension

F

after extension

G

G’

H

H’

metal beads water

latex membrane

Fig. 6 – Morphology of the notochord controls its CE movement. (A–E) Mathematical modeling of the notochord morphogenesis triggered by uniform expansion of AC. Notochord received frictional forces from the expanding substrate. (A) Cell movements in the absence of forces in the cells. Only frictional force from the substrate was considered. (B, C) Cell movements in the presence of tension and area elasticity of the cells (Model 1). Tension is constant regardless of the side length. (D) Cell movements in the presence of elastic force and area elasticity of the cells (Model 2). Elastic force is proportional to the changes of the side length. (E) Cell movements in the presence of endoderm cells. Tension and area elasticity of the cells were considered (Model 1). (F–H) Physical simulations of notochord morphogenesis triggered by uniform expansion of AC. (F) Schematic representation of experimental design. (G, H) Snapshot images of two physical simulation experiments before (G, H) and after (G 0 , H 0 ) extension of substrate.

the application of anisotropic tension by AC expansion likely contributes the directional cue for PCP signaling in CE together with the AP patterning system (Ninomiya et al., 2004) and Wnt signaling (Gao et al., 2011) found in Xenopus embryos. The AC expansion force is applied not only to the notochord/endoderm layer but also to the overlying epiblast/neural plate and mesoderm. Indeed, the pierced embryos had a

shortened body axis (Fig. 3), suggesting that the AC expansion force is important for the global morphogenesis of embryos and notochord morphogenesis. AC expansion regulates PCP protein, and defects in PCP signaling (Vangl2Lp/Lp mutants) compromised both the lateral intercalation and reorientation of the notochord cells (Fig. 7). While the role of PCP signaling in CE and lateral intercalation has been well established (Keller, 2002; Myers et al., 2002; Wallingford et al., 2002; Wang

MECHANISMS OF DEVELOPMENT

et al., 2006; Ybot-Gonzalez et al., 2007; Yen et al., 2009), the mechanism by which PCP signaling controls cell reorientation remains unknown. It is possible that increased cell motility by PCP/Wnt-Ror2 signaling (Gao et al., 2011; Nomachi et al., 2008) is required for reorientation by AP stretching. Alternatively, it is also possible that defects in the CE of neighboring tissues, e.g., neural plate and somites, in the Vangl2Lp/Lp mutants (Wang et al., 2006; Yen et al., 2009) indirectly affected AP

wild type

pierced embryo

B

Lp/Lp

C

stretching of the notochord by AC expansion, which in turn suppressed the reorientation of the cells. In contrast to mouse embryos, many other mammals form a flat blastoderm that is either sandwiched by an AC and a visceral yolk sac cavity or located on one side of a visceral yolk sac cavity (Eakin and Behringer, 2004a,b). We speculate that the coordinated expansion of these cavities would also provide the forces that expand flat blastoderms. It is of

pierced + Vangl2

Lp/Lp

E

D

width of notochord **** 40 6 30

width (μm)

A

Vangl2

53

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

20

6

10 0 ctrl

Vangl2

G

Lp/Lp

I

somite

H

6

42

0 Vangl2 +/+ +/+ -/+ pierce -

+

Vangl2

F-actin

size of cavity 1.5

8

2

ns 8

ctrl Vangl2Lp/Lp

4

1.0 0.5 0

K’’

O Posterior

control

Anterior

embryo

L’

L

amnion 1 Amniotic cavity

L’’

pierced

E7.75-E8.25 notochord

2

notochord Anterior 3 Vangl2

mem/cyto ratio

**** 4

60

2.0

1.0

0

N

7

ctrl pierce

intensity (/pixel)

M

4 7

40

20

0

convergent extension & cell reorientation

F-actin ns

ctrl pierce

Posterior

Lp/Lp

ctrl Vangl2

merged

K’

Kf

*

4

0

-/-

13

J

Lp/Lp

2

90

wild type

area (cm )

**** **** **** **** 200 101 89

somite number

cell orientation (deg.)

F

Vangl2

54

MECHANISMS OF DEVELOPMENT

interest to know whether a similar anisotropic tensiondependent mechanism also operates in the CE of the Xenopus notochord. Isotropic tension caused by archenteron expansion is required for the proper assembly of fibronectin, which is controlled by Wnt/PCP signaling (Dzamba et al., 2009), and fibronectin is required for polarized cell movement (Goto et al., 2005). However, in the explants of dorsal tissues of Xenopus embryos, which are unconstrained by tension from other tissues, extend well, and within them, cells continue to intercalate while oriented mediolaterally (Keller et al., 1989). Therefore, while local (subcellular/local ECM) tension might be required for fibronectin assembly, Xenopus notochord extension does not require the type of global tension that has been documented here. In conclusion, our study identifies a new role of extraembryonic tissues: mechanical control of embryo morphogenesis. Such communications between the embryonic and extraembryonic tissues through mechanical signals are also likely to play important roles in the morphogenesis of other tissues and/or the coordinated development of the embryonic and extraembryonic tissues.

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

in the RIKEN CDB and Kumamoto University and the laws and notifications of the Japanese government, and were approved by the institutional committees for animal and recombinant DNA experiments of RIKEN CDB and Kumamoto University.

4.2.

Statistical analyses

4.

Experimental Procedures

Data were statistically analyzed with Prism 6 statistical software (GraphPad). For box-whisker plots, analyses were performed by applying the Shapiro–Wilks test for normality; the null hypothesis was rejected (i.e., data were not normally distributed) for all cases except for one experiment (laser ablation). For the laser ablation experiment, Bartlett’s test showed that the variance was not equal. Therefore, non-parametric tests were used to analyze all data. The Wilcoxonbased Mann–Whitney U-test was used for comparisons of two groups. The Kruskal–Wallis test, followed by Dunn’s multiple comparison test, was used for comparisons of multiple groups. Sample numbers are shown above columns. Boxwhisker plots are presented with the box containing 50% of the data around the median and whiskers encompassing 100% of the data values. The horizontal line in the box indicates the median.

4.1.

Mouse lines

4.3.

Foxa2nEGFP–CreERT2 (Imuta et al., 2013) and R26-H2B-EGFP (Abe et al., 2011) mouse lines (Acc. Nos. CDB0601K and CDB0238K, respectively; http://www.cdb.riken.jp/arg/mutant%20 mice%20 list.html) were used for live imaging and laser ablation experiments, respectively. TnEGFP–CreERT2 mice (Imuta et al., 2013) (Acc. No. CDB0604K; http://www.cdb.riken.jp/arg/mutant%20 mice%20 list.html) were used for monitoring the expression of T in the notochord. Mice were housed in environmentally controlled rooms of the Laboratory Animal Housing Facility of RIKEN CDB and the Center for Animal Resources and Development (CARD) of Kumamoto University. All experiments were carried out following the regulations for animal and recombinant DNA experiments

Time-lapse imaging of mouse embryos

Time-lapse confocal images of live embryos were acquired by using an inverted microscope (Olympus IX81N-ZDC) equipped with an incubator unit CU109 (Live Cell Instrument), a 488-nm diode laser (Olympus), a spinning disk confocal unit (CSU-X1; Yokogawa), and a 512 · 512 pixel EMCCD camera (iXon + DU897E; Andor) as described previously (Imuta et al., 2013). The stage and the optics carrier of the microscope were enclosed by using a custom-made environmental chamber. The entire imaging system was controlled using MetaMorph software (Molecular Devices). Time-lapse images of the notochord cells near the node of the mouse embryos between early head fold (EHF) (Downs and Davies, 1993) and early somite stages were acquired. Embryos were immobilized

b Fig. 7 – Expansion of AC regulates notochord morphogenesis upstream of PCP signaling. (A–D) Representative images of the notochord at the end of in vitro culture. Yellow lines outline the notochord. (E) Distribution of the width of the notochord. The numbers above each column indicate the number of embryos examined. ****: p < 0.0001. (F) Distribution of orientations of the notochord cells. The numbers above each column indicate the number of cells examined. Data were obtained from 2 to 4 embryos for each group. ****: p < 0.0001. (G, H) Left-side views of the representative 5-somite stage embryos showing comparable expansion of extraembryonic cavities (amniotic and exocoelomic cavities). (I) Distribution of the somite numbers in control and Vangl2Lp/Lp embryos at the end of in vitro culture. The numbers above each column indicate the number of embryos examined. *: p < 0.05. (J) Distribution of cavity sizes. Cavity size was represented by the area of extraembryonic cavities in the left-side view of the embryo. The numbers above each column indicate the number of embryos examined. ns: not significant. (K, L) Distribution of Vangl2 and F-actin proteins in the notochord and surrounding endoderm in control (K) and pierced (L) embryos after 4 h of in vitro culture. (M) Ratios of membrane (mem) Vangl2 to cytoplasmic (cyto) Vangl2 signal intensities in the notochord cells. ****: p < 0.0001. (N) Intensities of F-actin signals in the notochord. ns: not significant. (O) Schematic diagram illustrating that expansion of amniotic cavity promotes notochord morphogenesis. In E7.75 to E8.25 mouse embryos, expansion of the amniotic cavity (1) causes stretching of the notochord along the antero-posterior axis (2). This stretching is required for convergent-extension movement of the notochord (3), which involves reorientation of the notochord cells. In the conceptus, embryonic tissues are shown in pink and red (notochord is shown in red), while extraembryonic tissues are shown in grey and dark brown.

MECHANISMS OF DEVELOPMENT

using modified culture dishes as described previously (Yamanaka et al., 2007). Embryos were cultured in 50% rat serum and 50% Dulbecco’s modified Eagle medium (DMEM) under 5% CO2. Images were acquired every 10 min, with 12–15 Z-positions at each time point, and the section interval was 8 lm. Drift in embryo positioning during imaging was corrected using the iSEMS program (Kato and Hayashi, 2008). Z-stack images were generated by using MetaMorph, and analysis of cell movements was performed using Image J software.

4.4.

Analysis of orientations of cells and cell divisions

Orientations of notochord cells were determined by evaluating the long axes of the nuclei. Orientations of cells and cell divisions were indicated by the angles from the AP-axis (Fig. 1E).

4.5.

Analysis of cell movements

Analyses of cell movements from time-lapse movies were performed by manually tracking each nucleus using the Manual Tracking plugin for Image J.

4.6.

Whole-mount immunostaining of embryos

Whole-mount immunostaining of embryos was performed with a standard procedure. Briefly, embryos were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 1 h at 4 C. Fixed embryos were washed briefly with PBS, permeabilized with PBS containing 0.3% Triton-X 100 for 10 min, and then washed three times in PBS containing 0.1% Tween20 (PBT) for 10 min at room temperature. The embryos were incubated in blocking buffer (PBT containing 10% goat serum) for 1 h at room temperature and then incubated with the blocking buffer containing primary antibodies at 4 C overnight. The embryos were washed six times with PBT, followed by overnight incubation with secondary antibody in blocking buffer at 4 C. Confocal images of immunostained embryos were obtained with an inverted laser confocal microscope (A1; Nikon or FV1000; Olympus), or a macro-zoom laser confocal microscope (AZ-C1; Nikon). Rat anti-GFP (1:500; 04404-84; Nacalai Tesque), mouse anti-b-catenin (1:2000; 610154; BD Biosciences), and rabbit anti-Vangl2 (Montcouquiol et al., 2006) (1:1000; gift from M. Montcouquiol) antibodies were used as primary antibodies. Anti-rat IgG labeled with Dylight488 (1:1000; 112-485167; Jackson ImmunoResearch), anti-mouse IgG labeled with Alexa Fluor 555 (1:2000; A21424; Invitrogen), and anti-rabbit IgG labeled with Alexa Fluor 488 (1:500; A11008; Invitrogen) were used as secondary antibodies. F-actin and nuclei were stained with Texas Red-X phalloidin (1:100; T7471; Invitrogen) and Hoechst 33258 (1:200,000; Invitrogen), respectively.

4.7.

Laser ablation of cells in embryos

Laser ablation of a single cell in LHF to 1-somite stage embryos was performed using a specially assembled inverted multi-photon microscope combined with a full-sized CO2/O2 incubator as described previously (Eiraku et al., 2011). For high-resolution multi-photon live imaging during laser ablation, an optical section image (512 · 512 pixels) was taken at each time point using a ·25 water-immersion lens (N.A.

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

55

1.05; Olympus) and multi-photon femtosecond laser (900 nm; Mai-Tai DeepSee eHP; Spectra-Physics) with group velocity dispersion auto-compensation. Laser ablation was performed using the same multi-photon optical system for 3D live imaging. A 900-nm laser beam from Mai-Tai Deep-See eHP at full power was condensed at the target cells via a ·25 water-immersion lens (N.A. 1.05) using a spot illumination of three times of 100 microseconds. The first image after laser ablation was acquired 6.47 s after ablation. After laser ablation, the surrounding cells first moved away from the ablated cells, but then approached the ablated cells to close the wound. Therefore, the distance of cellular movement before and after laser ablation was quantified at the time when the distance reached the maximum (9–16 s after ablation), and the position of the cells was quantified by the centroids of the nuclei. Cells were classified into four groups (anterior (A), posterior (P), left (L), and right (R)) depending on the position of the cells relative to the ablated cell (Fig. 1J). Their movement along the x-axis (for L and R) and y-axis (for A and P) were used for analyses. For the cells in the A or P positions, absolute values of the movement along the A–P axis are shown as ADy or PDy, respectively. For the cells in the L or R positions, the absolute values of the movement along the L–R axis are shown as LDx or RDx, respectively. The number of nuclei examined is indicated above each bar (Fig. 1K and M).

4.8.

Suppression of AC expansion

LB-stage embryos (Downs and Davies, 1993) were cultured in vitro for 8 h (approximately 5-somite stage) with 50% rat serum and 50% DMEM under 5% CO2 with rotation. To suppress AC expansion, a fine glass capillary that has an open tip was inserted into the AC of LB-stage embryos by piercing through the yolk sac and the amnion. Because the glass capillary tends to clog during the in vitro culture of embryos, another glass capillary was inserted after 4 h of culture.

4.9.

Comparison of cavity size

The size of extraembryonic cavities, i.e., the sum of amniotic and exocoelomic cavities, was represented by the 2D area of extraembryonic cavities in the left-side view of the embryo. Essentially, the same results were obtained when the area of cavities in the frontal view was used for comparison.

4.10.

Mathematical modeling

4.10.1. Modeling of notochord and surrounding endoderm 4.10.1.1 notochord. We considered a 2-dimensional notochord. The notochord was assumed to be an elastic architecture composed of grids that mimicked cells, and was placed on the expanding field (Fig. 5C). The frictional force between the notochord and the expanding field will be defined in the part 3 ‘‘Modeling of frictional force from AC expansion’’. Each grid was defined to include surface tension or an elastic spring on each side (Fig. 5D-i) and area elasticity (Fig. 5D-ii). Surface tension or an elastic spring exerted a force on the two vertices of the side. The surface tension was assumed to be constant (Fig. 5D-i).

56

Ft ¼ T ðT ¼ const:Þ

MECHANISMS OF DEVELOPMENT

ð1Þ

Here, Ft is the force exerted on a vertex, and T is the surface tension. The force provided by the elastic spring is defined as follows (Fig. 5D-i): Fs ¼ ks ðL  L0 Þ

ð2Þ

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

4.10.3. Modeling of frictional force from AC expansion We defined a frictional force exerted between the expanding field and an object placed on the field. We assumed that the motion of the object was dominated by the friction rather than inertial effect, and that the frictional force was defined to be proportional to the velocity of the expanding field: Fe ¼ lVs

Here, Fs is the force exerted on a vertex of the side, ks is the elastic constant of the springs, and L or L0 is the length of the side or the natural length of the spring, respectively. The potential energy conferred from the area elasticity was defined as follows (Fig. 5D-ii): 1 Ua ¼ ka ðA  A0 Þ2 2

Here, Fe is the frictional force applied to the object on the expanding field, and l is the coefficient of friction. Because Vs is proportional to p, Fe is also proportional to p as shown below, Fe ¼ lVs / p

ð3Þ

ð7Þ

Thus, we can rewrite the equation as follows (Fig. 5D-iii): Fe ¼ lke p

ð8Þ

Here, Ua is the potential energy, ka is the area elastic constant, and A or A0 is the area of each grid or the natural area of each grid, respectively. The potential energy generates forces on the four vertices of each grid, and the forces were calculated by numerical differentiations of the potential with respect to the positional vector,

Here, ke is the proportionality coefficient between Vs and p. Therefore, Fe is a function of p. Objects on this force field are dragged by Fe. In the notochord, each vertex was assumed to receive the frictional force Fe from the expanding substrate.

Fa ¼ rUa ;

4.10.4. Movement of notochord and endoderm on AC expansion

ð4Þ

where Fa is the force exerted on a vertex. Finally, the total force applied on a vertex is the sum of the three forces, and Ft or Fs was considered in Model 1 or Model 2, respectively, as follows: Model1 : Ftot ¼ Ft þ Fa

ð5-1Þ

Model2 : Ftot ¼ Fs þ Fa

ð5-2Þ

4.10.1.2 Surrounding endoderm.

When we took the endoderm into consideration (Fig. 6E), the endoderm was placed to surround the notochord. We modeled the endoderm similarly to the notochord: surface tension, elastic springs, and area elasticity were included, but with different values. The frictional force between the endoderm and the expanding field was also considered in the similar way to the notochord.

4.10.2. Modeling of AC expansion To model AC expansion, we first considered a 3D situation. AC expansion was mimicked by using a uniformly expanding sphere, such as a rubber balloon. The arc length between two points on the spherical surface is increased through the sphere expansion, and the increasing rate is proportional to the arc length (Fig. 5A). Thus, if an origin is defined on the surface, all points on the surface will move away from the origin with speeds proportional to the arc lengths from the origin. To simplify the model, we reduced the 3D situation to a 2D situation. On the 2D expanding field, all points should move away from an origin with speeds proportional to the distance from the origin, as shown below (Fig. 5B): Vs / p

ð6Þ

Here, Vs is the velocity of the points, and p is the positional vector representing the distance from the origin. This formulation results in that the distance between arbitrary two points expands with a rate proportional to the distance.

Movement of the vertices comprising the notochord or the endoderm might be determined by the sum of the forces and friction as follows: m

dVv ¼ Ftot  lðVv  Vs Þ dt

ð9Þ

Here, m is the weight of the vertex, Vv is the velocity of the vertices, t is time, and l is the coefficient of friction. We can rewrite the equation as follows by using equation 7: m

dVv ¼ Fsum  lVv dt

ð10Þ

Here, Fsum is the sum of the forces: Fsum ¼ Ftot þ Fe

ð11Þ

We assumed that the inertial effect was negligible. Thus, the left-hand of equation 11, in which dVv/dt means acceleration, should be = 0.0. Consequently, the velocity of a vertex was written as (Fig. 5D-iv): Vv ¼

Fsum Ftot or Vv ¼ þ Vs l l

ð12Þ

When Ftot is zero, Vv becomes Vs. In other words, a vertex moves with the same speed as the expanding field, if there are no other forces except the frictional force (Figs. 5B and 6A).

4.10.5. Parameter values used in our simulation The simulations were run for 100,000 iterations with each step representing 0.01 time units [A.U.]. The AC was expanded 3-fold during the simulations. The origin of the coordinates was put at the centroid of the notochord in Fig. 6, but the position did not affect the results. The coefficient of friction, l, was set as 1.0. The grid size was set as 1.0 · 1.0. In the notochord, the surface tension, T, was set as 0.0, 0.005, 0.005, or 0.005 in Fig. 6A, B, C, or E, respectively. The elastic area constant, ka, was 0.0, 0.1, 0.1, or 0.1 in Fig. 6A, B, C, or E, respectively. When the endoderm was considered (Fig. 6E), T and ka in the endoderm were set as 0.0005 and 0.01, respectively.

MECHANISMS OF DEVELOPMENT

In both the notochord and the endoderm, the natural area of the grids, A0, and the natural length of the sides, L0, were set as 1.0 and 1.0, respectively. In Fig. 6D, T, ka, and ks in the notochord were set as 0.0, 0.1, and 0.01, respectively.

4.11.

Physical simulation

Latex membrane was clipped from latex groves. The latex membrane was set in an embroidery hoop (120 mm diameter) and put on a cylinder (90 mm diameter). Metal beads (3 mm diameter) were placed on the latex membrane, and the proper amount of water was added so that the metal beads were assembled. The latex membrane was stretched by drawing the embroidery hoop downward.

4.12.

Quantification of Vangl2 signals in the notochord

Notochord regions were isolated from the entire images and were used for the analyses. The intensity of Vangl2 signals in the plasma membrane was quantified by identifying F-actin in the cortex of the plasma membrane. F-actin bundles were segmented from projected confocal images and their density was quantified using the KBI ImageJ plugins (http://hasezawa.ib.k.u-tokyo.ac.jp/zp/Kbi/ImageJKbiPlugins) (Ueda et al., 2010). The image was filtered by using a gradient inverse-weighted smoothing filter (Wang et al., 1981) to reduce noise. Fibrous structures in the image were enhanced with a multiple directional non-maximum suppression algorithm (Sun and Vallotton, 2009). The F-actin regions were determined as high-signal regions of the enhanced image. Finally, the F-actin regions were converted to 1-pixel thick segments by using a thinning algorithm (Wu and Tsai, 1992). The total intensity of Vangl2 signals overlapping with F-actin regions was quantified with MetaMorph (Molecular Devices) as total Vangl2 signals in the plasma membranes, and divided by the area of F-actin to obtain the average intensity. Total signal intensity in the notochord cells was also quantified with MetaMorph and divided by the area of the notochord to obtain the average intensity. The total intensity of cytoplasmic signals was obtained by subtracting the total intensity of membrane signals from the total signal intensity of the notochord cells. The average intensity of the cytoplasmic signals was obtained by dividing the total intensity of the cytoplasmic signals by the area of cytoplasm.

Author contributions Y.I. and H.S. conceived this project and designed all of the experiments using the mouse embryos. Y.I. performed most of the biological experiments. D.S. and T.F. performed the immunostaining of Vangl2. H.K. and T.F. performed the computer simulation and physical modeling. M.E. assisted with the laser ablation experiments. Y.I., H.K. and H.S. wrote the paper.

Acknowledgements We thank the Laboratory for Animal Resources and Genetic Engineering of RIKEN CDB for housing the mice, providing

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

57

the mutant mice, and assisting with the laser ablation experiments, the CARD of Kumamoto University for housing the mice. We also thank Dr. Y. Sasai for help with the laser ablation experiments, Drs. K. Kato and S. Hayashi for help with the iSEMS program, and Dr. M. Montcouquiol for the antibody. These studies were supported by grants from RIKEN (H.S. is a former member of RIKEN CDB), and by Grants-in-Aid for Scientific Research (KAKENHI) from MEXT (21116003) and JSPS (23247036) to H.S.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.mod.2014.01.004.

R E F E R E N C E S

Abe, T., Kiyonari, H., Shioi, G., Inoue, K., Nakao, K., Aizawa, S., Fujimori, T., 2011. Establishment of conditional reporter mouse lines at ROSA26 locus for live cell imaging. Genesis 49, 579–590. Aigouy, B., Farhadifar, R., Staple, D.B., Sagner, A., Roper, J.C., Julicher, F., Eaton, S., 2010. Cell flow reorients the axis of planar polarity in the wing epithelium of Drosophila. Cell 142, 773– 786. Ang, S.L., Rossant, J., 1994. HNF-3 b is essential for node and notochord formation in mouse development. Cell 78, 561–574. Arnold, S.J., Robertson, E.J., 2009. Making a commitment: cell lineage allocation and axis patterning in the early mouse embryo. Nat. Rev. Mol. Cell Biol. 10, 91–103. Beddington, R.S., 1981. An autoradiographic analysis of the potency of embryonic ectoderm in the 8th day postimplantation mouse embryo. J. Embryol. Exp. Morphol. 64, 87–104. Bielinska, M., Narita, N., Wilson, D.B., 1999. Distinct roles for visceral endoderm during embryonic mouse development. Int. J. Dev. Biol. 43, 183–205. Cha, J., Sun, X., Dey, S.K., 2012. Mechanisms of implantation: strategies for successful pregnancy. Nat. Med. 18, 1754–1767. Downs, K.M., Davies, T., 1993. Staging of gastrulating mouse embryos by morphological landmarks in the dissecting microscope. Development 118, 1255–1266. Dzamba, B.J., Jakab, K.R., Marsden, M., Schwartz, M.A., DeSimone, D.W., 2009. Cadherin adhesion, tissue tension, and noncanonical Wnt signaling regulate fibronectin matrix organization. Dev. Cell 16, 421–432. Eakin, G.S., Behringer, R.R., 2004a. Diversity of germ layer and axis formation among mammals. Semin. Cell Dev. Biol. 15, 619–629. Eakin, G.S., Behringer, R.R., 2004b. Gastrulation in other mammals and humans. In: Stern, C.D. (Ed.), Gastrulation: From Cells to Embryo. Cold Spring Harbor Laboratory Press, New York, pp. 275–287. Eiraku, M., Takata, N., Ishibashi, H., Kawada, M., Sakakura, E., Okuda, S., Sekiguchi, K., Adachi, T., Sasai, Y., 2011. Selforganizing optic-cup morphogenesis in three-dimensional culture. Nature 472, 51–56. Gao, B., Song, H., Bishop, K., Elliot, G., Garrett, L., English, M.A., Andre, P., Robinson, J., Sood, R., Minami, Y., Economides, A.N., Yang, Y., 2011. Wnt signaling gradients establish planar cell polarity by inducing Vangl2 phosphorylation through Ror2. Dev. Cell 20, 163–176.

58

MECHANISMS OF DEVELOPMENT

Goto, T., Davidson, L., Asashima, M., Keller, R., 2005. Planar cell polarity genes regulate polarized extracellular matrix deposition during frog gastrulation. Curr. Biol. 15, 787–793. Halder, G., Dupont, S., Piccolo, S., 2012. Transduction of mechanical and cytoskeletal cues by YAP and TAZ. Nat. Rev. Mol. Cell Biol. 13, 591–600. Herrmann, B.G., 1991. Expression pattern of the Brachyury gene in whole-mount TWis/TWis mutant embryos. Development 113, 913–917. Hiramatsu, R., Matsuoka, T., Kimura-Yoshida, C., Han, S.-W., Mochida, K., Adachi, T., Takayama, S., Matsuo, I., 2013. External mechanical cues trigger the establishment of the anterior-posterior axis in early mouse embryos. Dev. Cell 27, 131–144. Imuta, Y., Kiyonari, H., Jang, C.W., Behringer, R., Sasaki, H., 2013. Generation of knock-in mice that express nuclear EGFP and tamoxifen-inducible Cre recombinase in the notochord from Foxa2 and T loci. Genesis 51, 210–218. Kato, K., Hayashi, S., 2008. Practical guide of live imaging for developmental biologists. Dev. Growth Differ. 50, 381–390. Keller, R., 2002. Shaping the vertebrate body plan by polarized embryonic cell movements. Science 298, 1950–1954. Keller, R., Cooper, M.S., Danilchik, M., Tibbetts, P., Wilson, P.A., 1989. Cell intercalation during notochord development in Xenopus laevis. J. Exp. Zool. 251, 134–154. Kurotaki, Y., Hatta, K., Nakao, K., Nabeshima, Y., Fujimori, T., 2007. Blastocyst axis is specified independently of early cell lineage but aligns with the ZP shape. Science 316, 719–723. Montcouquiol, M., Rachel, R.A., Lanford, P.J., Copeland, N.G., Jenkins, N.A., Kelley, M.W., 2003. Identification of Vangl2 and Scrb1 as planar polarity genes in mammals. Nature 423, 173– 177. Montcouquiol, M., Sans, N., Huss, D., Kach, J., Dickman, J.D., Forge, A., Rachel, R.A., Copeland, N.G., Jenkins, N.A., Bogani, D., Murdoch, J., Warchol, M.E., Wenthold, R.J., Kelley, M.W., 2006. Asymmetric localization of Vangl2 and Fz3 indicate novel mechanisms for planar cell polarity in mammals. J. Neurosci. 26, 5265–5275. Motosugi, N., Bauer, T., Polanski, Z., Solter, D., Hiiragi, T., 2005. Polarity of the mouse embryo is established at blastocyst and is not prepatterned. Genes Dev. 19, 1081–1092. Myers, D.C., Sepich, D.S., Solnica-Krezel, L., 2002. Convergence and extension in vertebrate gastrulae: cell movements according to or in search of identity? Trends Genet. 18, 447– 455. Ninomiya, H., Elinson, R.P., Winklbauer, R., 2004. Antero-posterior tissue polarity links mesoderm convergent extension to axial patterning. Nature 430, 364–367. Nomachi, A., Nishita, M., Inaba, D., Enomoto, M., Hamasaki, M., Minami, Y., 2008. Receptor tyrosine kinase Ror2 mediates Wnt5a-induced polarized cell migration by activating c-Jun Nterminal kinase via actin-binding protein filamin A. J. Biol. Chem. 283, 27973–27981. Olguin, P., Glavic, A., Mlodzik, M., 2011. Intertissue mechanical stress affects Frizzled-mediated planar cell polarity in the Drosophila notum epidermis. Curr. Biol. 21, 236–242. Orr, A.W., Helmke, B.P., Blackman, B.R., Schwartz, M.A., 2006. Mechanisms of mechanotransduction. Dev. Cell 10, 11–20.

1 3 2 ( 2 0 1 4 ) 4 4 –5 8

Sasaki, H., Hogan, B.L., 1993. Differential expression of multiple fork head related genes during gastrulation and axial pattern formation in the mouse embryo. Development 118, 47–59. Sausedo, R.A., Schoenwolf, G.C., 1993. Cell behaviors underlying notochord formation and extension in avian embryos: quantitative and immunocytochemical studies. Anat. Rec. 237, 58–70. Sausedo, R.A., Schoenwolf, G.C., 1994. Quantitative analyses of cell behaviors underlying notochord formation and extension in mouse embryos. Anat. Rec. 239, 103–112. Schmidt, W., 1992. The amniotic fluid compartment: the fetal habitat. Adv. Anat. Embryol. Cell Biol. 127, 1–100. Sun, C., Vallotton, P., 2009. Fast linear feature detection using multiple directional non-maximum suppression. J. Microsc. 234, 147–157. Tam, P.P., 1981. The control of somitogenesis in mouse embryos. J. Embryol. Exp. Morphol. 65 (Suppl.), 103–128. Tam, P.P., Loebel, D.A., 2007. Gene function in mouse embryogenesis: get set for gastrulation. Nat. Rev. Genet. 8, 368–381. Ueda, H., Yokota, E., Kutsuna, N., Shimada, T., Tamura, K., Shimmen, T., Hasezawa, S., Dolja, V.V., Hara-Nishimura, I., 2010. Myosin-dependent endoplasmic reticulum motility and F-actin organization in plant cells. Proc. Natl. Acad. Sci. U.S.A. 107, 6894–6899. Vogel, V., Sheetz, M.P., 2009. Cell fate regulation by coupling mechanical cycles to biochemical signaling pathways. Curr. Opin. Cell Biol. 21, 38–46. Wallingford, J.B., Fraser, S.E., Harland, R.M., 2002. Convergent extension: the molecular control of polarized cell movement during embryonic development. Dev. Cell 2, 695–706. Wang, D.C.C., Vagnucci, A.H., Li, C.C., 1981. Gradient inverse weighted smoothing scheme and the evaluation of its performance. Comp. Graph. Image. Proc. 15, 167–181. Wang, J., Hamblet, N.S., Mark, S., Dickinson, M.E., Brinkman, B.C., Segil, N., Fraser, S.E., Chen, P., Wallingford, J.B., WynshawBoris, A., 2006. Dishevelled genes mediate a conserved mammalian PCP pathway to regulate convergent extension during neurulation. Development 133, 1767–1778. Wood, A., Thorogood, P., 1994. Patterns of cell behaviour underlying somitogenesis and notochord formation in intact vertebrate embryos. Dev. Dyn. 201, 151–167. Wu, R.Y., Tsai, W.H., 1992. A new one-pass parallel thinning algorithm for binary images. Pattern Recogn. Lett. 13, 715–723. Yamanaka, Y., Tamplin, O.J., Beckers, A., Gossler, A., Rossant, J., 2007. Live imaging and genetic analysis of mouse notochord formation reveals regional morphogenetic mechanisms. Dev. Cell 13, 884–896. Ybot-Gonzalez, P., Savery, D., Gerrelli, D., Signore, M., Mitchell, C.E., Faux, C.H., Greene, N.D., Copp, A.J., 2007. Convergent extension, planar-cell-polarity signalling and initiation of mouse neural tube closure. Development 134, 789–799. Yen, W.W., Williams, M., Periasamy, A., Conaway, M., Burdsal, C., Keller, R., Lu, X., Sutherland, A., 2009. PTK7 is essential for polarized cell motility and convergent extension during mouse gastrulation. Development 136, 2039–2048.