To of the studied cell types, showing similar dependence

To reveal how the metastatic potential of
cancer cells, during brain metastases, expresses in their nanomechanical properties
SCFS were performed between three
different types of human melanoma cells, as indavers, and a confluent layer of human cerebral
endothelial cells (CEC), as a model of the inner side of brain
capillaries. The three melanoma cell types, with different metastatic characteristics were
represented by WM35, A2058 and A375 cell lines, while the CEC layer was formed by the hCMEC/D3
(shortly D3) cell line. In addition to measuring the intercellular interactions, in each case
the very same melanoma cells were tested against a cell free area of the Petri dish as well, by
using the same parameters
(load, force speed, dwell time in contact) for all cell types.

Figure 11. Relative elasticity of
the three studied cell types against bare Petri dish surface and confluent
endothelial (D3) layer, respectively. Significantly different (p < 0.05) groups are marked with black link bars. Figure 12. Apparent Young's modulus of the studied cell types against bare Petri dish surface and confluent endothelial layer (D3), respectively. Significantly different groups (p < 0.05) are marked with black link bars. Due to a more illustrative comparison of elastic behaviour we defined and used a relative dimensionless parameter, namely the relative elasticity. This, similarly to the elasticity index in case of polymers 90, was defined as the ratio between the remanent work (cyan area, Figure 2B) and the total work (cyan + yellow area, Figure 2B) having values between 0 and 1, where 0 corresponds to a perfectly plastic, while 1 to a perfectly elastic behaviour. Figure 11 represents the calculated relative elasticity of the studied cell types, showing similar dependence pattern, when the melanoma cells are pushed to bare Petri dish and to endothelial cells. This underlines that the measured parameters reflect mainly the characteristics of the melanoma cells, hence the endothelial layers can be considered practically invariable. Figure 11 also show a strong dependence between the relative elasticity and the melanoma cell type, where WM35 cells appears to have a more elastic, while A2058 and A375 cells a more plastic behaviour upon deformation. The slightly higher values when two cells are pushed together might be attributed to the presence of the endothelium. The most frequent parameter to characterize cellular elasticity is the Young's modulus. Figure 12 represents the apparent Young's modulus values obtained for the studied cell types when pressed against the bare Petri dish and to a confluent endothelial layer, where WM35 cells show the highest values, followed by A2058 and A375 cells being the softest. Interestingly, when these cells were pushed against a confluent layer of endothelial cells, the observed differences practically vanished. Figure 13. Maximal adhesion force between the studied cell types against bare Petri dish surface and confluent endothelial layer (D3), respectively. Significantly different groups (p < 0.05) are marked with black link bars. Figure 14. Number of rupture events per force curve for the three studied cell types against bare Petri surface and confluent endothelial layer. Significantly different groups (p < 0.05) are marked with black link bars. As a first parameter to characterize the linkage strength between the melanoma cells and the endothelium is maximal adhesion force represented in Figure 13. In spite of observing similar adhesion strength between the melanoma cells and the bare Petri dish, a clear dependence on cell type can be observed when melanoma cells were pushed against the endothelium (Figure 13). In addition, by comparing the results in Figure 11 with those in Figure 13, an inverse correlation can be observed between relative elasticity and maximal adhesion force. As the overall adhesion of the cells is build up by numerous discrete linkage, during the de-adhesion process individual rupture like events can be distinguished on the recorded force plots (see Fig. 2B). Figure 14 summarizes the number of these events for the studied melanoma cells, showing a similar dependence for both recorded sets, where more aggressive cells establish more active connections. Despite the weak negative correlation observed between the number of ruptures and relative elasticity, the exact reason why higher number of ruptures appears in the case of melanoma cells - bare Petri than melanoma cells - endothelium is still under debate. However, a clear dependence can be observed in both studied cases (Petri surface and endothelial layer), assuming that the highest number of connections are formed in case of A375 cells, while the least for the WM35 cells. In addition to the number of apparent bonds, step size and occurring place (length) carries important information as well. The apparent rupture size can be associated mostly with the molecular background of the studied cells. As Figure 15A shows, when the melanoma cells are pushed to Petri dish slightly shifted and more skewed distribution patterns were observed. However, in case of confluent endothelium the distribution patterns were similar (Figure 15B), having most frequent values around 30 pN. Figure 15. Rupture size probability distribution of the rupture size between the studied melanoma cells when pushed to Petri dish surface. An even more interesting feature is the occurrence distance of rupture events from the contact point. The length distribution of these distances are plotted in Figure 16 A and B.  Since cells suffer a slight shape deformation during their contact (mostly the melanoma, as the endothelial cell is adhered and flattened) the highest peaks can be observed below one micrometer. This zone is characterized by simultaneous events whose number is dependent on the contact area. Rupture length distribution of longer distances can be mainly associated with membrane dynamics, and spontaneous unlocking of cell adhesion molecules and other molecular cross talks. The higher frequencies of values below one micrometre can be associated with the apparent relative elasticity of the melanoma cells. As they appear more elastic, they regain their original size quicker, resulting most of de-adhesion events at close distances to contact point. Figure 16. Probability distribution of the rupture length for the de-adhesion events occurred during melanoma cell - Petri dish contact. De-adhesion events occurring at high distances (> 2 ?m from contact point) can be
predominantly associated with membrane tether formations, regarded as membrane
nanotubes extruded from a large membrane reservoir observed in both cases,
melanoma cell – Petri and melanoma cell – endothelial interactions.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

1.1. Discussions
and summary

Cell
adhesion depends on multiple and even multivariate individual molecular
connections, where the individual players are difficult to identify. Due to the
high force-resolution provided by the AFM besides maximal adhesion forces,
individual de-adhesion events can be identified and compared accurately, which
are the hallmark of the established connection 10,11. On single-cell force
spectroscopy based mechanical examination the interaction of three different
types of melanoma cells, having altered invasive characteristics, were
performed against brain endothelial cells. Our results show that nanomechanical
properties can be associated to higher metastatic potential and invasive characteristics
may rely on stronger adhesive properties mediated by altered tether formation
dynamics.

Arrest
of melanoma cells on the inner surface of the brain blood vessels is a crucial
but not a sufficient step in the process of brain metastasis formation. Nevertheless,
those blood-travelling melanoma cells which show higher adhesiveness to the
brain endothelium might have higher chance to successfully colonize the brain.
The mechanism of melanoma cell arrest and establishment of firm contact to
brain endothelial cells is still only partially described and understood. Here
we show a comparison of the dynamics of the first short term contact of three
types of melanoma cells (WM35, A2058 and A375) with brain endothelial cells.
The WM35 is a cutaneous, non-metastatic cell line, while A2058 and A375 are
highly metastatic cell lines 26. The difference in the metastatic potential
between the A2058 cells and A375 cells has not been clearly established so far.
Both cell lines are VCAM-1 negative 27, however, Rolland et al. found that
transmigration of A2058 cells across bovine brain capillary endothelial cells,
was twice as effective compared to A375 28. Both A2058 and A375 cells show
similar adhesive properties to brain endothelial cells with similar junction damaging
potential in static models 29,30. Pogoda et al. denotes A375 cells as highly
invasive 31 and compares its elasticity to WM35 emphasizing that the former
has lower elastic modulus (which correlates well with our results). The above
mentioned studies deal with the total transmigration process of the melanoma
cells across the endothelium, which includes but is not restricted to initial
affinity dynamics of firm adhesion establishment. Furthermore, these studies
were conducted on static conditions.

Intercellular
adhesion dynamics depends on multiple factors, amongst which we can find
apparent whole cell elasticity, visco-elastic properties, surface charge
density, surface linked adhesion molecule distribution and glycocalyx thickness
as well. The parameters investigated by us include relative elasticity, maximal
adhesion force, size and location of de-adhesive rupture events. Referring to
the first short term contact to brain endothelial cells our data suggests the
following order from lower to higher metastatic potential – WM35, A2058 and
A375 – based on apparent affinity to brain endothelial cells.

Unfortunately,
no proper model exists to obtain elastic or plastic properties when two cells
are pushed against each other. Therefore, in order to compare the elastic
properties of the studied cell types in situ. Similarly to the plasticity index
23 we have used the relative elasticity, or elastic index, as a dimensionless
comparing factor, which consists of the ratio between the remanent and total
work needed to obtain the pre-set load (for details see section of Materials
and methods). In our case the perfect elasticity would be 1 while total
plasticity 0. As presented in Fig. 3 the obtained relative elasticity values
are similar both in the case when melanoma cells were pushed against the Petri dish
surface and against the endothelial layer. This indicates that the calculated
relative elasticity is predominantly the property of the melanoma cells and the
endothelial cells have low contribution. The highest relative elasticity values
were shown by WM35 cells, followed by A2058 and A375 cells respectively. These
results are in line with the findings of Lekka et al., which demonstrate a
higher Young’s modulus for WM35 cells compared to A375 cells 32. Furthermore,
Xu et al. has reported the same relation between Young’s Modulus and metastatic
potential in case of ovarian cancer cells probed with a spherical indenter
33. In our case Fig. 4 depicts similar relation in case when the studied
melanoma cells were pushed to a Petri dish surface: WM35 has the highest
apparent Young’s modulus succeeded by A2058 and A375 cells respectively. This
relation vanishes when the same cells were pushed to endothelial cells. The
supporting connection between apparent Young’s modulus and invasiveness is an
important control, which in case of melanoma-endothelial interaction might be
hindered by several factors. In our opinion the calculated apparent Young’s modulus
(Fig. 4) is less sensitive to the characteristics of the probing melanoma cell,
while value of relative elasticity (Fig. 3) is more suitable for proper
comparison in our case.

Cell
elasticity is mainly determined by cytoskeletal structures and low elasticity
may reflect disorganization of the cytoskeletal characteristic to aggressive
cancer cells 34. Besides mechanical properties of a cell, physical aspects of
cell-cell interaction may influence cellular behaviour, too. The force
necessary to move apart two cells (adhesion force) reflects well the strength
of the connection. Maximal adhesion force is a well-established parameter to
characterize adhesion properties of biological samples, ranging from individual
molecules to living cells 10. Fig. 5 shows the comparison of the calculated
maximal adhesion forces between the three studied cell types contacting a bare Petri
dish and the endothelial layer. In case of the Petri dish, no differences can
be found between the investigated cells suggesting a nonspecific interaction
between the plastic surface of the Petri dish and the cells. In contrast, when
melanoma cells were pushed against the endothelium a clear difference can be
observed. WM35 cells show the lowest adhesion force, higher values can be
observed in the case of A2058 cells, whereas the A375 cells show the highest
adhesion forces from brain endothelial cells. This indicated that the more
aggressive melanoma cell types adhere stronger to the cerebral endothelium
possibly leading to an enhanced transmigratory and metastasis forming capacity.
Interestingly, – although independent of cell type – adhesion forces are higher
in case of Petri dish – melanoma cell contacts compared to melanoma cell –
endothelial cell contact. An explanation of this observation could be the
specially treated, cell culture grade plastic surface to which cultured cells
can easily adhere.

The
process of de-adhesion is not continuous; it can be decomposed into a series of
de-adhesion events. Intercellular adhesion is largely determined by specific
cell-cell adhesion molecules and non-specific interactions of the glycocalyx.
When pulling apart two adhering cells these interactions have to be released.
These bond ruptures can occur close to the contact point if the adhesion
molecules are well anchored to the underlying cytoskeleton. In case of non-anchored
adhesion molecules during the release these are either ripped out from the
membrane or form a tether (membrane nanotube), depending on the membrane properties
35. Dynamics of these tethers highly depends on the physical parameters
(pulling speed, temperature) as reported in case of rolling neutrophils 36
and adhesion of the monocytic cell line THP-1 to a surface coated with ICAM-1
37. In our case membrane nanotube formation can also be observed during the
release of the contact.

Comparing
the number of the observed rupture events (Fig. 6), which is directly related
to active contact points, the same relation can be observed as for the maximal
adhesion force. This implies the presence of a surface size related active
binding process, which has higher weighting in case of more inelastic cells
(see Fig. 3). Size distribution of the active de-adhesive events is depicted in
Fig. 8 and Fig. 9 for melanoma – Petri surface and melanoma – endothelial
contact respectively. The most abundant values are around 30 pN, presenting a
slight downshift with the cell type for both cases. According to literature,
this value is associated with de-coupling of membrane bound adhesion molecules
in case of membrane tether ruptures 38, although it is very close to the
adhesion forces reported in case of E-cadherin fragments 39. Additionally,
the occurrence place of these de-adhesion events is an important characteristic
of membrane dynamics, since in many cases they appear at several micrometre
distances from the contact point. Since physical parameters (pulling speed,
contact time, temperature) were not altered through the experiments, all
melanoma cell types were subjected to the same set of external parameters.
Hence, differences in release dynamics are associated to alterations in
membrane or cytoskeletal network properties. Based on our data, we can conclude
that the shift towards higher distances of occurrence places of detected
de-adhesion events (Figs. 10 and 11) might indicate that the role of tether
based adhesive properties of invading melanoma cells cannot be neglected in the
metastasis formation process. Tether formation and dynamics might contribute
considerably to site selection of melanoma cells ending in successful arrest on
the surface of brain endothelial layer. Although it is not an easy task to
quantify the weighting of tether based adhesive contribution within the full
detachment force, it might grant metastatic melanoma cells one step forward to
successful colonization.

Finally,
as a conclusion we can say that we have used successfully the AFM based single
cell spectroscopy for comparison and analysis of adhesion force dynamics
between a confluent brain endothelial layer and three different type of
melanoma cells presenting different invasive characteristics. Apparent
mechanical properties such as elasticity, maximal adhesion force, number, size
and distance of individual rupture events showed altered values pointing
towards cell type dependent aspects. Our results underline the importance of
mechanical details in case of intercellular interactions. Nevertheless, it
suggests that in adequate circumstances elastic and adhesive characterizations
might be used as biomarkers