Cancer Metastasis, Motility and Matrix Mechanics
1 Video View·Sep 5, 2023
"Cancer Metastasis, Motility and Matrix Mechanics"
Muhammad Yakut Ali, Mechanical Science and Engineering
(Advisors: Taher Saif and Mark Kuhlenschmidt)
CNST Nanotechnology Workshop 2013
And our lab is primarily interested in working with colon cancer metastasis
because colon cancer is one of the major cause of death in males and being ranked
as third in 2012 statistics. So our work, our lab work with colon cancer cell
originates from an interesting observation with the human colon
carcinoma cell line HCT8 30 years back. So in this work what they did, they
actually plated this HCT8 cell line on hard polystyrene dish and what happens
that these cells form an epithelial monolayer on this polystyrene dishes and
eventually very few percentage of them go through a spontaneous differentiation
process which are completely morphologically distinct and they call
it like HCTR rounded variant of these original epithelial cells. So as you can
see in details like these R cells are formed on the epithelial monolayer of
this HCT8 cells and interestingly enough these R cells downregulate,
significantly downregulate their e-cadherin expression and alpha
catenin expression and they were also found to be far more invasive compared
to their parent HCT8 e-cells. So by keeping this observation in mind, one of
graduate student in our lab, he wants to explore the role of mechanical
microenvironment on this HCT8 cells. So what he do, he take different stiffness
polyacrylamide hydrogels and also polystyrene dishes and the plate HCT8
cells on different stiffness gels which are functionalized with
extracellular matrix proteins and also on polystyrene or glass substrate and as
a control he also plated regular endothelial and epithelial cells. So what
happens that these HCT8 cells on gels, they form cell cluster consisting of
numerous cells depending on initial cell plating density and by two days and as
time progresses, if you notice carefully the periphery, cell proliferation mainly
happens in the periphery after two days and the thickening or some sort of
three-dimensional structure occurs in this periphery only. In contrast on the
polystyrene dishes, they actually form monolayer and they don't show this kind
of 3D structure in polystyrene dishes. So what happens after a week or so, some
rounded single cells start to dissociate from the periphery of these cell
islands as you can see in this phase contrast images and if you keep longer,
some of these islands completely dissociate into numerous single motile
live cells. So this is like sort of in vitro metastasis type of thing and in
contrast on the hard polystyrene substrate, you don't see any dissociative
morphological phenotype transition. So what happens with the control healthy
cells, epithelial cells and endothelial cells, so they also form island and
confluent layer as the HCT8 cells on gels and dish respectively but they
neither on gels nor on hard substrate they form, show these like phenotype
transition or rounded variants. So now as we know that these HCT8 cells
significantly change their morphological phenotype on extracellular
matrix polyacrylamide gels, we ask the question that whether this
morphological transition depends on the cell-to-gel adhesion. So if we
replace this cell-to-gel adhesion with e-cadherin which is a cell-cell
adhesion molecule instead of extracellular matrix fibrinogen or
collagen, whether we still see this morphological transition on gels. So the
results come out pretty interesting. On e-cadherin coated gels they also form
islands and eventually more or less at the same time they do undergo this
morphological transition on e-cadherin coated gels as well on the control ECM
fibrinogen coated gels. So far our findings is that by using PA gels
instead of hard polystyrene dish or glass substrate, we can produce large
scale epithelial to rounded morphological transition of HCT8 cells
and this morphological transition is independent of the cell-to-gel adhesion.
With that in mind that we do see the significant change in morphology but we
want to investigate whether these cells are really more invasive compared to
their parent cells. So to address that we have performed a basement membrane
invasion assay in Professor Kulinsky's lab and the results indeed show that these
R-cells are by far more invasive compared to their parent e-cells and we
have also performed the RNA sequencing analysis and the gene analysis results
actually come out pretty interesting. We found that several oncogenes are
actually turned on in these R-cells which are not present in the e-cells.
These genes include the anti-apoptotic ons and cancerous stencil marker.
These results suggest that these R-cells not only change their phenotype, they
also change their genotype and they're by far more invasive compared to their
parent e-cells and they're also different mechanistically. So if we do
the traction force microscopy of these E and R-cells on the same stiffness
polyacrylamide gel, the traction force microscopy results show that these R-cells
really generate very weak traction force on the similar stiffness
substrate compared to their parent e-cells. So these results kind of give us
an idea that these R-cells might lack mechanosensitivity on this 2D
substrate. So to address that we have performed a motility assay with these
e-cells and R-cells. So the materials and methods for this motility assay is as
follows. What I have done, I have taken these e-cells and R-cells and plated
them on different stiffness polyacrylamide gels ranging from 1 kPa to
40 kPa and we have performed a time-lapse microscopy over 12 hours. So
starting from 6 hours of plating, so we tracked them basically between 6 hours
to 18 hours and then we calculate the average migration speed or motility
speed. And if you see the results, the e-cells show a functional dependence on
average migration speed with the substrate stiffness. So their average
migration speed basically decreases with the increase in stiffness. Whereas the
R-cells, they don't show any trend in the average migration speed by changing
substrate stiffness. So this also in line with their morphological transition and
traction force results that they really do lack mechanosensitivity compared to
their counterparent e-cells. So then we think closely that what is so common in
both cases in PHL, whether it's functionalized with ECM or e-cadherin
molecule. So we notice that two things are common. One is that you have
always three-dimensional structure at the periphery, which appears to be
necessary if not sufficient condition for this morphological transition. And
another thing, you do need cell-cell contact for this morphological transition
to occur. So if that is true, I come up with a hypothesis that if we can
encapsulate these cells inside a 3D hydrogel rather than a 2D, plating them
on a 2D gel, so and if they can form a multicellular spheroid, they get the
cell-cell contact and at the same time you can provide the mechanical matrix Q
by the three-dimensionality of the biophysical environment of the matrix.
And I collaborate with Professor Brendan Harley in chemical and
biomolecular engineering, and I use gelatin methacrylate hydrogels as the
3D matrix. So the justification of using 3D gelatin methacrylate is that it's
basically hydrolyzed collagen, but yet it retains the natural binding sites of
collagen, and collagen is the major ECM in vivo column. So that's why we use like
four different stiffness gelatin methacrylate hydrogels. And we perform
live-dead assay for imaging, and we found that by day four, they indeed
actually form multicellular spheroid inside this 3D hydrogels. So these
are like inside the hydrogels. So if we closely look on of these islands, we can
actually see that they do form multicellular spheroid, and you have
very good cell-cell contact here, so cell junction established here by day four.
And as this is very preliminary data, we do see significant epithelial to rounded
morphological transition inside these hydrogels when we do this live-dead
assay again one day 10. And this morphological transition occurs on all
the stiffnesses that we have tested, ranging from 5 kPa to 20 kPa. So these
results suggest that like the E2R transition was not only dependent on the
two-dimensional environment. It's kind of dependent on, independent of matrix
dimensionality as well. So to summarize the work, I have shown that these human
colon carcinoma cells, they can show this metastatic-like transition on and in
gels, and the results that I'm not showing for the interest of time, that
this transition is actually irreversible. So by that, I mean that if we replate
these RCS on any stiffness gel or hard glass, they retain their morphology, so
it's stable and irreversible. So with that, I will take any question.
It must be very, very clear.
Oh no, it's like gelatin-methacrylate hydrogels. So gelatin is like hydrolyzed
collagen. So it contains the natural binding sites. Some of them are RGD, some
are like other sites. So you don't need to functionalize them with any additional
like extracellular matrix or anything, any RGD or anything.
Exactly, exactly. So you can see like very clearly, these are some of the optical
stack from the confocal. So if you see the 3D, you can see even clear like some
of the cells come off from the spheroids, and some like completely transformed
their morphology from epithelial to rounded type.
So far, we don't find any stiffness dependent in the range we have tested,
like 5 to 21 kPa. So this also goes with 2D, because in 2D also, we see on a wide
range of stiffness on PA gel. So we don't see on hard substrate like glass or
petri dish. So it's difficult to say that what the matrix mechanics in a bulk
sense is playing a role, but we don't know about the local mechanical
microenvironment, because the local environment may be completely different.
This is actually the compressive bulk modulus of those 3D gels. So it's
difficult to tell that what the cells are exactly experiencing the local
mechanical microenvironment. Is the red supposed to be the dead ones, and so
there's so few red, that means they're not dying? For sure not, exactly, exactly.
Do that affect the porosity you mean? For sure, yeah, it does affect the porosity. So
if you increase the stiffness, the pore becomes smaller. But as I said, that if
you talk about the overall or bulk stiffness for the gels, we don't see
like the stiffness role, overall stiffness role that much. We don't see on
glass or hard substrate, but what the PA gels or these 3D gels is playing the
role, what we believe, like they do some sort of confinement, as you notice from
the phase contrast images on gels, whether it's functionalized with
fibronic tin or ECAD here encoded, they do form an island. They never form a kind
of monolayer. And then they keep proliferating on the periphery and sort
of kind of thickening at the periphery and kind of form three-dimensional
structure. So, so far we believe that these are the cues that are playing the
role, like the sort of three-dimensionality and the cell-cell contact. But like if
you talk about the only the gel stiffness, so far the bulk stiffness, we
don't think that it's playing that of a role, except that it's providing the
confinement. For 3D you mean? No, it's it's basically very difficult to make
softer gels with gelma rather than 5 kPa, because the structure of gelma is
like that you have gelatin and you use methacrylate group and use some
photo radical and then you UV polymerize it. So because of its structural
nature, you are photopolymerizing it, you cannot go less than 5 kPa for this
material system. But if you use like other 3D matrixes, for example collagen,
then you can go much softer. You can go like a couple of hundred pascal. So, but
we have not done any experiment with collagen though. But you can go much
softer. Okay, would you like to extend your degree?

