Insights into vertebrate development:

rucksackbulgeAI and Robotics

Dec 1, 2013 (3 years and 6 months ago)

61 views

Paul Kulesa

Stowers Institute for Medical Research

Insights into vertebrate development:

merging bioimaging and computational modeling

Paul Kulesa

Stowers Institute for Medical Research

Insights into vertebrate development:

merging bioimaging and computational modeling

chick

alligator

duck

quail

From
www.saviorfare.wa

& B.S. Arnold et al., 2001

We have developed culture and imaging techniques

to analyze avian development



Up to 1 day of imaging



Upright or inverted imaging



Video and confocal time
-
lapse microscopy

Intravital Imaging of Chick Embryos

Whole Embryo Explant



Up to 5 days of imaging



Embryo in natural setting



Neural crest

(from origin to destination)



Up to 1 day of imaging



Upright or inverted imaging



Video and confocal time
-
lapse microscopy

Intravital Imaging of Chick Embryos

Whole Embryo Explant

In ovo

Craniofacial Patterning
:
Cell migration and guidance

Model system: The Neural Crest

Cutis, 1999

Incorrect migration can lead to birth defects:



Frontonasal dysplasia



Waardenburg’s syndrome (pigment)



Neurofibromas (peripheral nerve tumors)

Craniofacial Patterning
:
Cell migration and guidance

Model system: The Neural Crest

Cutis, 1999

Incorrect migration can lead to birth defects:



Frontonasal dysplasia



Waardenburg’s syndrome (pigment)



Neurofibromas (peripheral nerve tumors)

How do cells sort into and maintain migrating streams?

Highlights of Cranial Neural Crest Cell Patterning

Cells emigrate from all rhombomeres

Previous model hypotheses

1) Diffusion


Cells diffuse from

specific segments (rhombomeres)

(Le Douarin, 1995)

PK & S. Fraser Dev. Biol., 1998

Highlights of Cranial Neural Crest Cell Patterning

Cells emigrate from all rhombomeres

Previous model hypotheses

1) Diffusion


Cells diffuse from

specific segments (rhombomeres)

(Le Douarin, 1995)

PK & S. Fraser Dev. Biol., 1998

r3

r5

but avoid some areas

Highlights of Cranial Neural Crest Cell Patterning

Cells can reroute their migratory paths

Previous model hypotheses

2) Genetic


Cells are endowed

with migration/destination instructions

(Lumsden et al., 1991)

1) Diffusion


Cells diffuse from

specific segments (rhombomeres)

(Le Douarin, 1995)

Premigratory neural crest cells ablated in r5
-
r6

wt

Cell trajectories

are disrupted

PK, Bronner
-
Fraser, S. Fraser, Dev., 2000

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

?

?

?

?

?

?

?

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

(cells proliferate during migration)

(some cells follow one another after contact)

Highlights of Cranial Neural Crest Cell Patterning

Our working model

contact guidance

proliferation

+

(cells follow one another, but can become leaders)

?

?

?

?

?

Rate of change in

neural crest cells

=

chemotaxis

+

N(x,y,t)

Lu, Fraser, & PK, Dev Dyn. 2003

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

Average cell speed =

49 +
-

9 um/h


Average directionality =

0.29 +
-

0.1

Cells at the stream fronts:


higher directionality (+28%)


slower avg speed


directed filopodia

Cell tracking w/J. Solomon & S. Speicher/Caltech

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

?

?

?

?

?

Long range chemoattractant

Areas of inhibition

(cell
-
contact mediated)

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

(N(x,y,t))

Rate of change in

chemical attractant

=

diffusion

+

production

degradation

+

(C(x,y,t))

Source

of

cells

(midline)

t = 0

L(t)

0 < x < L(t)

Boundary moving at speed = s1 (um/hr)

L(t) = L(0) +s1*t

Long range chemoattractant

at destination

site

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

(N(x,y,t))

Rate of change in

chemical attractant

=

(C(x,y,t))

Source

of

cells

(midline)

t = 0

L(t)

0 < x < L(t)

Boundary moving at speed = s1 (um/hr)

L(t) = L(0) +s1*t

f (Diffusion, degradation, production,s1)

Long range chemoattarctant

at destination

site

Assume that C may be ~netrin

(long range chemoattractant evidence from

axon guidance studies

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

(some cells are repelled from an area after contact)

(some cells are attracted to other cells to form a chain like array)

r6

r7

Highlights of Cranial Neural Crest Cell Patterning

Our working model

Rate of change in

neural crest cells

=

chemotaxis

+

contact guidance

proliferation

+

N(x,y,t)

(some cells are repelled from an area after contact)

(some cells are attracted to other cells to form a chain like array)

r6

r7

Highlights of chains
:



Neural crest chains are made up of 5
-
10 cells




May be a general mechanism of cell migration




Chains form in neuronal precursors migrating to the

olfactory bulb (Alvarez
-
Buylla, 2002)




Tumor cells form chains in 3D collagen gels

(Friedl, 2002)




Dictyostelium (slime mold) form chains to assemble

a multicellular organism


Highlights of Cranial Neural Crest Cell Patterning

Our working model hypotheses
(Discrete model for contact guidance term)

1) Cells in the chain are linked together by filopodia

2) A cell within a chain emits a chemoattractant at its posterior end

3) A cell links with another cell after contacting posterior end

(evidence from dictyostelium (cAMP))

Highlights of Cranial Neural Crest Cell Patterning

Our working model hypotheses
(Discrete model for contact guidance term)

1) Cells in the chain are linked together by filopodia

2) A cell within a chain emits a chemoattractant at its posterior end

3) A cell links with another cell after contacting posterior end

Main assumption for all 3 hypotheses:

Either lead cell chews a hole in the extracellular matrix (ECM) or ECM is permissive and

lead cell lays down a trail for others to follow.

Simple model (cellular automata)



Define a lead cell



Lead cell moves mostly in lateral direction



Leaves open spaces behind which other cells may move into



Gives clues as to how close the lead cell must stay to attract followers



Can leave behind clues instead of open spaces, such as chemoattractant



short or long range interactions?

(evidence from dictyostelium (cAMP))

Cellular structure of the chains

Our working model hypotheses
(Discrete model for contact guidance term)

1) Cells in the chain are linked together by filopodia

DiI

Cellular structure of the chains

Our working model hypotheses
(Discrete model for contact guidance term)

1) Cells in the chain are linked together by filopodia

DiI

Gfp via electroporation

Cellular structure of the chains

Our working model hypotheses
(Discrete model for contact guidance term)

1) Cells in the chain are linked together by filopodia

DiI

Projection of 30 um confocal sections

Direction of motion

r4

r5

Do cranial neural crest cells in mouse migrate with a rich set of behaviors?

Challenges



3D embryo



Gas exchange important



Finer temperature control

than in chick

Benefits to Mouse culture and imaging



Genetics (target mutations of genes related to craniofacial patterning)



Several mutant mouse models available with craniofacial defects


Do cranial neural crest cells in mouse migrate with a rich set of behaviors?

Challenges



3D embryo



Gas exchange important



Finer temperature control

than in chick

Jones et al., Genesis 2002

Somites form slightly slower in whole embryo culture

Gfp labeled blood cells in early circulation

GFP transgenic mouse line from M. Baron/Mt. Sinai

It is important to maintain the embryo in one place

P. Trainor

Acknowledgements

Caltech



Scott Fraser



Marianne Bronner
-
Fraser



Mary Dickinson



Dave Crotty

Stowers Institute for Medical Research



Paul Trainor