Transcript Slide 1

Cell and Tissue Engineering: 3D Effects
John Eichorst
BIOE 598
April 7, 2008
Outline of Talk: Review Article
Modeling Tissue Morphogenesis and Cancer in 3D: Cell 130, 601-610, (2007)
Yamada, Kenneth M. and Edna Cukierman
1) Discuss the motivation for examining cells in 3D systems
2) Look at some of the common types of 3D models currently being researched
3) Examine the effects that 3D ECMs have on morphology and behavior of cells
-polarity, migration, stress, local oxygen concentration, branching
morphogenesis
4) Finally, discuss the clinical applications of research using 3D ECM models
-Stem Cell
-Cancer Research
Why are we thinking about 3D environments?
Short Answer:
Cells cultured in 2D settings can differ greatly from those grown in 3D
environments:
-morphology
-cell-cell interactions
-cell-matrix interactions
-differentiation
Tissues can be harvested in vivo and later cultured in vitro. Often, they retain their
original 3D structure.
Branching Morphogenesis in salivary glands
In many organs, epithelial clefts
and resulting bud formations can
occur leading to interesting 3D
structures during branch
morphogenesis in many organs.
Cell 130, 601-610, (2007) modified from Nature 423, 876-881 (2003)
Initiation of Branching
morphogenesis
Shallow clefts
form in a
globular bud
As the clefts deepen
new buds are formed
Cells from a tissue culture can be placed into a 3D structure as single cells or as a
larger aggregate from a tissue.
WM793 Melanoma Spheroids
In Vitro Cell. Dev. Biol. Anim. 42, 242-247 (2006)
In Vitro Cell. Dev. Biol. Anim. 42, 242-247 (2006)
As shown in the procedure above, researchers in 2006 tried to simulate the environment by including other cell types
(keratinocytes and fibroblasts) in the 3D model.
In these spheroids, each cell inside the spheroid is subject to a different microenvironment within the sphere itself.
Interestingly, the authors noted that EMT6 tumors showed different drug(etoposide) resistances as function of the dimension of
the culture they were studied in.
General conclusions describing the differences between the cell’s behavior in 2D
and in 3D matrices are summarized below.
Human Fibroblasts in 3D vs. 2D Environment
In addition, the choice of the design of the 3D
model and the resulting unique
microenvironment that the cell experiences
can also introduce substantial variation in the
cell’s behavior.
Cell 130, 601-610, (2007)
Cell signaling and behavior in response to the extra-cellular environment can be
greatly affected by the 3D model being used.
HIF-α proteins are typically
translated and rapidly degraded by
the hydroxylation of two proline
residues by proline hydrolyase
enzymes.
Lack of oxygen stabilizes HIF-α
resulting in the expression of the
subunit HIF-β in the nucleus and
ultimately the expression of various
genes.
Cell 129, 465-472, (2007)
HIF proteins, by being able to promote the expression of the various proteins listed in the figure in
hypoxic locations in tumors has been considered as one of the potential causes of tumor progression.
There is also evidence that hypoxia can to some extent control the proliferation and differentiation of
certain stem cells including embryonic stem cells, neuronal stem cells, hematopoietic stem cells …
Various 2D and 3D models can be designed to contain a variety of stiffness in the ECM
that the cell is exposed to.
Actin Cytoskeleton Remodeling of Adult Human Dermal Fibroblasts
Single Cell Migration of Adult Human Dermal Fibroblasts
Although not necessarily
3D, the design of the 2D
construct was intended to
model the diverse
environment encountered
by cells from both a
mechanical and signaling
perspective.
Biomaterials 28, 671-379, (2007)
Cell and tissue polarity can also vary in 3D matrices depending on the cell type and the
microenvironment that they experience.
Epithelial cells are often
polarized and aggregate in
spherical 3D forms as shown,
to function properly and
secrete products based on
this organization.
This spherical form is not
possible on 2D tissue
cultures that tend to flatten
the cells.
Cell 130, 601-610, (2007)
However, as shown on the right, Mesenchymal cells lose their “dorsal-ventral”
polarity when placed in the 3D matrix and appear more as an elongated spindle
shape, similar to a fibroblast.
Branching morphogenesis has also been studied in 3D model systems in order to try to
understand the formation of glands and organs.
Nature 423, 876-881 (2003)
Nature 423, 876-881 (2003)
The hypothesis of the paper was that the local expression of fibronectin mRNA s in areas called
“presumptive cleft regions” ultimately left to the formation of a cleft in the tissue resulting in the
branching morphogenesis.
Initiation of Branching
morphogenesis
Shallow clefts
form in a
globular bud
As the clefts deepen
new buds are formed
Guiding the differentiation of stem cells into specific lineages by manipulating their
extra-cellular environment could have definite clinical applications as shown by the
summary of some of the current research below.
-Controlling the differentiation of stem cells into different lineages by examining their
interaction with various 3D ECMs .
-Embryonic stem cells have been lead to differentiate into epithelial cells as a
function of the 3D environment (collagen matrix)
-Culturing embryonic stems cells with fibroblasts was noted as generating a
neural lineage
-Culturing with keratinocytes lead to epithelial differentiation.
-Growing stem cells in 2D culture has been shown to promote differentiation into
blood vessels
-Matrix stiffness as well as been shown to be able to select for different lineages
in in stem cell differentiation.
Cell 130, 601-610, (2007)
3D ECM models can also be applied to study the growth of cancer and
proliferation of tumor cells.
MT1-MMP, which typically degrades the ECM proteins to encourage the invasion and proliferation of
tumor cells.
-However, in a 3D collagen gel model, researchers noted that the elimination of all
extracellular proteases did not affect cell invasion.
Gene expression to some extent as hinted before, can be a function of a tumor cell’s microenvironment
as implied by the previous description of hypoxic cells.
The microenvironment of a tumor with respect to the stiffness of the ECM it encounters as well as the
features of the stroma have been implicated as possible factors determining the tumor’s progression.
Cell 130, 601-610, (2007)
Conclusions:
3D cell cultures can begin to more accurately model the environment
experienced by cells in vivo. The striking differences between the behavior of
cells in 2D and in 3D environments imply appealing questions about how these
differences can be quantified and related to the knowledge of in vivo systems for
scientific and clinical applications.
Areas of Research with Clinical Applications:
-Controlling or directly modulating the differentiation of stem cells
-Engineering cells to be able to be introduced into tissues, having them
respond to their local environment in a way beneficial to the tissue itself
-Replacing damaged tissue
-Develop high throughput methods for drug screening for in vitro settings
-development of “therapeutic targets” and methods for cancer
treatment
Research Article:
Science 294, 1708-1712, (2001)
Outline of Paper
The purpose of the paper is to study the structure and function of 3D matrix
adhesions formed from the interaction of fibroblastic cells with a 3D extracellular matrix.
The experiments described in the paper examine the following:
1) Physiological responses of the cell to the 3D environment
2) Determine the composition of the 3D matrix adhesions
3) Study the cell’s response to rigidity and lack of de novo protein synthesis
in a 3D model
The definition of a 3D matrix adhesion is related to the co-localization of the α5
integrin, Paxillin and Fibronectin.
Direct
immunofluorescence
staining from 2D
Culture of mouse
fibroblast
α5 integrin - important
fibronectin receptor
Focal Adhesion Kinase
Paxillin
Focal Adhesions
Fibrillar Adhesions
Tensin
Vinculin
αvβ3 integrin
There are more components of these adhesions. However, those listed were the subject of the initial
experiments.
The next set of experiments explored in more depth how the 3D matrix adhesions
were associated with the physiological responses of the cell in 2D and 3D
environments using human fibroblasts.
The 3D set of experiments:
The 2D set of experiments:
-“tissue derived 3D matrices”
derived from detergent extracted
mouse embryo sections
-fibronectin
-“cell derived 3D matrices” from
naturally deposited 3D ECMs of
NIH-3T3 fibroblasts
-collagen 1
-laminin
-2D matrix -> 3D matrix that was
mechanically compressed.
-3D collagen gel lattices
-2D mix -> 2D coatings of
solubilized, cell derived 3D matrix
Below are the results presented for the 10 min cell attachment assay for the
experimental setup described previously
Attachment (a.u.) = the relative number of
attached cells compared to fibronectin
sample after plating
= condition where α5 integrin
was inhibited
Cell attachment was determined by staining the cells with bisenzimide
H3342 fluorochrome, for nuclear staining?
In addition to cell attachment, the cell morphology also changed depending on the
conditions of the various substrates used
The fibroblasts, themselves, were
stained with Dil to generate the
images.
The images were then thresholded for
the representation in the figure.
Interestingly, the fibroblasts in the 3D matrix (3D mat, control) achieved the
elongated morphology after 5 hours.
The fibroblasts cultured in the 2D matrices were noted as achieving the same
morphology after 18 hours.
This experiment also described the cell motility as a function of the dimension of
the substrate as well as α5 integrin.
Summary of the data in previous 3 plots
Sixteen paths were positioned at the same
starting point to generate the figures.
The motility of the cells was tracked for 5 hours.
The next question asked was about how the previously described fibrillar and
focal adhesions respond to a similar set of ECM substrates.
3 substrates were designed to study the 3D matrix adhesions
A cell-derived 3D matrix
was also used to culture
the fibroblast cells and
later flattened to form a 2D
matrix
A 3D matrix was also
made containing solely
fibronectin fibrils to
study the fibroblasts
Human fibroblasts were
cultured overnight on a
tissue-derived 3D
matrix. Mouse tissue
was used for this matrix
As a result of the integrin dependent behavior noticed in the previous
experiments, a substrate of entirely fibronectin was used to determine to some
extent, the interplay between the matrix composition and the dimension of the
environment.
Only the tissue derived 3D matrix showed indication of the localization
consistent with 3D matrix adhesion.
Insets of Images:
Red/Purple indicate focal
adhesions.
Turqoise indicates fibrillar
adhesions.
White describes the
combination of the adhesion
structures indicating the overall
3D matrix adhesion
As mentioned before, the definition of a 3D matrix adhesion is related to the co-localization of the α5
integrin, Paxillin and Fibronectin.
Other observations mentioned in paper included that rigidifying the cell-derived (although tissue-derived is
shown above for 3D) caused a lack of 3D matrix adhesions.
The effects of both de novo protein synthesis and rigidity were studied to
determine their effects on the formation of 3D matrix adhesions.
Human fibroblasts were plated in cell-derived 3D matrices.
Inhibitor of
protein
synthesis
To obtain the rigidified cell condition, the cell was treated by the covalent
fixation with glutaraldehyde prior to plating.
As mentioned before, cell rigidity in the 3D matrix substantially reduced the
formation of the 3D matrix adhesion. However, the inhibition of the de novo
protein synthesis has very little effect in this regard.
The spatial localization of other proteins and signaling molecules related to cell
adhesion were also studied to determine composition of the 3D matrix adhesions.
Human fibroblasts were again
cultured on a 2D substrate
(fibronectin), a mechanically
compressed cell-derived matrix and a
3D matrix.
The data in the table describe the
results of the immunofluorescence
staining examined at the adhesion
sites in the 3D matrix.
Green = α5 integrin
Red = Paxillin
Fluorescently labeled antibodies also pointed to other components involved in the
formation of 3D matrix adhesions.
Planar
fibronectin
surface
As before, FAK and paxillin co-localized with the α5 integrin in the 3D matrix adhesions.
Only phospho-paxillin31 co-localized with the α5 integrin. FAK[pY397] was absent from the 3D matrix adhesions as shown by
the lack of co-localization with the α5 integrin.
Paxililin [pY31] = phospho-paxillin31
FAK [pY397] = tyrosine phosphorylated FAK
In a similar experiment, confocal images were also acquired from the
immunofluorescence staining of a section of a E13.5 mouse embryo
As also shown in the
previous images,
phospho-FAK397 does
not co-localize well
with the α5 integrin
while the blue and
green image in B infer
that FAK and α5
integrin do co-localize.
The last question discussed concerning the final experiment asked if it was
possible that the lack of FAK[pY397] could be the result of the inactivity of FAK
at stable adhesion sites.
Phosphorylation Levels of FAK, Paxillin, and ERK in 3D Matrix and in 2D fibronectin
FAK phosphorylation on tyrosine 397 generated by
the signaling interaction with integrins is required in
many signaling pathways associated with focal
contacts.
Therefore, is it possible that there is independent
regulation of FAK and paxillin phosphorylation
throughout the 3D matrix adhesion?
The western blot analysis was performed to show
that the lack of FAK phosphorylation can occur in 3D
matrices with relatively little effect on paxilin or
ERK1/2.
Questions:
1) The definition of the 3D matrix adhesion given in the initial section of the
paper does not seem consistent with the other molecules found in the 3D
matrix adhesions. Why did they concentrate so much on α5 integrin and
paxillin?
2) How are the 3D matrix adhesions organized to allow for motility and is there
coordination specific for random or directed movement?
3) Generally, are there multi-faceted functions of the 3D matrix adhesions that
may distinguish from those in 2D cultures.