To explore the adhesive properties of c-Kit we used MC/9 cells. Mast cells are of hematopoietic origin (Kitamura & Ito, 2005). The mast cell-committed precursors arise in the bone marrow and are recruited through the circulation to peripheral tissues where they mature under cytokine or growth factor stimulation (Jamur & Oliver, 2011). In normal BMMCs, the central adhesion molecules that have been identified are the integrins β1, α4 and, α5 (Morgado et al., 2014).
By using MC/9 cells, we have shown that KitL stimulates c-Kit dependent adhesion and spreading on fibronectin. This effect appeared to be differentially regulated by KitL presentation. In a soluble form, KitL stimulated cell spreading on rich-fibronectin surfaces. In contrast, ligand immobilization induced cell spreading on suboptimal fibronectin concentrations in collaboration with integrins. Also, we showed that the spreading response on immobilized-KitL and the FN-low matrix is a c-Kit dependent but an independent kinase function. However, s-KitL-induced spreading requires an active c-Kit kinase. All these data were published at the beginning of my thesis by our group in (S. Tabone-Eglinger et al., 2014).
These observations underlined the importance of KitL/c-Kit and ECM/integrins in mediating cell-matrix interactions and validated our model for the study of one aspect of cell-niche interactions. Following these results, we proposed that the c-Kit and integrins pathways might converge into a common pathway leading to cell adhesion and spreading towards immobilized-KitL. Thus, it is of importance to comprehend how this cooperative interaction take place during cell spreading.
After being in contact with fibronectin and KitL-coated surfaces MC/9 cells adhere, increase their surface of contact with the matrix, and the cell body flatten (Figure 25). These morphological changes have already been described for other cell types (Gauthier et al., 2009) and occurs in three phases (Dobereiner et al., 2004; Dubin-Thaler et al., 2008). During the
“basal” stage, the first cell-matrix contacts are formed. The second “continuous” phase is related to the formation of lamellipodia and an increase of the adhesive surface. Finally, the third “contractile” phase involves periodic contractions that require the establishment of a mechanical link between integrins and the actin cytoskeleton that reinforces the cell contacts and tensile forces with the environment. Moreover, cell polarization can occur depending on the organization of microtubules and the contractility of the actomyosin network (Giannone
et al., 2004; Giannone et al., 2007). The last step of spreading appears to depend on talin, which activates integrins and establishes the connection between integrins and the actin cytoskeleton. In the absence of talin, cells retract on themselves after the first two stages of spreading (X. Zhang et al., 2008).
Figure 25. Temporal course of mast cell spreading. The figure shows mast cells spreading on poor-fibronectin coated surfaces in the presence of immobilized-KitL. Upon contact with the coated surface, mast cells adhere, increase their surface of contact, and the cell body flatten resulting in the characteristic “fried-egg” phenotype indicative of cell spreading. Later, cells polarize and after two hours retract and detach from the surface.
Although MC/9 cells follow this classic pattern of cell spreading, they spread transiently, and after 2 hours they detach from the coated surface. We also observed that under i-KitL stimulation the rate and extent (spread surface) of the spreading response is increased and cells are more polarized (S. Tabone-Eglinger et al., 2014). Therefore, we can conclude that ligand immobilization mechanically activates the receptor c-Kit that in turns induces integrin-dependent adhesion on fibronectin. This activation might favor a more spread phenotype by reducing cell contractility at least during the early phase of spreading.
Cell adhesion, spreading and migration result from the cooperation between the ECM, integrins and their partners and the cytoskeleton. Binding of integrins to fibronectin occurs at the level of the RGD motif (Garcia et al., 2002). It causes a conformational change and engages the integrins in signaling pathways by an “outside-in” signaling (Friedland et al., 2009). We showed here that MC/9 cells did not spread in the complete absence of fibronectin.
Therefore, KitL/c-Kit alone are not sufficient to induce spreading, and it requires the support of integrins. Moreover, we also observed that integrins are directly involved in cell spreading as this response was blocked by adding RGD peptides to the cell suspension (not shown).
Therefore, activation of c-Kit by KitL might activate integrins to induce cell-matrix spreading on fibronectin surfaces.
The interaction of integrins which ECM molecules leads first to the formation of adhesive structures called focal adhesions (FA) (Campbell & Humphries, 2011; Y. Zhang & Wang, 2012).
Adhesion
(5 min) Spreading
(30 min) Polar/spread
(60 min) Retraction
(>120 min)
As discussed in section VII.2, focal adhesions are reinforced by the recruitment of cytoplasmic proteins (e.g., talin, kindling, paxillin) to the cytoplasmic tail of integrins. The engagement of integrins in cell adhesion activates signaling molecules such as the Focal Adhesion Kinase (FAK) (Harburger & Calderwood, 2009). FAK can be phosphorylated in response to growth factor stimulation. Indeed, c-kit activation induces cell adhesion and migration to fibronectin as well as an increase in β1 integrin-mediated phosphorylation of FAK and paxillin (Levesque et al., 1995; Takahira et al., 1997). Elevated FAK mRNA levels have been found in different malignancies (Sulzmaier et al., 2014) being the target of ATP-competitive kinase inhibitors including PF-573228 and PF-562271 (Dunn et al., 2010). In using these inhibitors, we assessed the effect of FAK inhibition in our in vitro model. As for most TKIs tested here, treatment of MC/9 cells with both FAK did not affect i-KitL induced spreading on low fibronectin concentrations. On the contrary, both inhibitors slightly reduce s-KitL induced cell spreading at high fibronectin concentrations. Like FAK, its hematopoietic homolog Pyk2 is critical for Rho-mediated cell adhesion by forming a Pyk2/Vav1 complex which is recruited to tyrosine phosphorylated β3-integrins (Gao & Blystone, 2009). Because PF-573228 and PF-562271 were not efficient in inhibiting cell spreading on FN-rich surfaces, even under s-KitL stimulation, we think that at the concentrations we used these drugs are unable to block Pyk2. In fact, they are about 10-fold less potent for Pyk2. However, treatment with higher levels of these inhibitors was toxic for our cells. Once activated, FAK supports the recruitment of Src and PI3K activating the PI3K/AKT (Xia et al., 2004) and the ERK (Bouchard et al., 2008) survival signaling pathways. It is, therefore, possible that in MC9 cells, targeting FAK alone is not enough to block Integrin-dependent spreading on fibronectin. In fact, active apoptosis of cancer cells was achieved by inhibiting both FAK and EGFR (V. Golubovskaya et al., 2002) or FAK and Src (V. M. Golubovskaya et al., 2003). Consequently, we suggest that the dual inhibition of c-Kit/FAK can be useful in inhibiting KitL-mediated cell spreading.
Therefore, we can conclude that immobilized-KitL mechanically activates the receptor c-Kit that in turns induces integrin-dependent adhesion on fibronectin. This activation might favor a more spread phenotype by reducing cell contractility at least during the early phase of spreading.
The migratory behaviors would depend on the shape that cells adopt when in contact with ECM. Hence, cells with a “fried-egg” phenotype will be more stationary, while those cells that have lamellipodia at the front and a retraction tail at the rear will present a more migratory
behavior. Lamellipodia formation was observed in growth factor-stimulated cells and is regulated by the small Rho-GTPase Rac1 (Ridley, 2015). Indeed, Rac was shown to be activated following integrin engagement during the initial phase of cell spreading (L. S. Price et al., 1998). Although we did not investigate this aspect, we propose that Rac might be activated in MC/9 cells under i-KitL stimulation while the contractility of the actomyosin module might be blocked. This mechanism might be prevented with imatinib as when MC/9 cells were treated with this drug cell spreading was delayed and polarity lost (S. Tabone-Eglinger et al., 2014). Thus, inhibition of the c-Kit kinase activity with imatinib might blocks cell contractility rather than spreading. In the process of contractility, talin (X. Zhang et al., 2008) and ROCK (Leung et al., 1996), which respectively connects the ECM to the cytoskeleton and phosphorylate myosin increasing the contractility. The ATPase activity of myosin can be blocked directly with blebbistatin (Kovacs et al., 2004). Also indirectly, by preventing the phosphorylation of the Myosin Light Chain (MLC) either through inhibiting ROCK with Y27632 (Ishizaki et al., 2000) or MLC Kinase (MLCK) with ML7 (Connell & Helfman, 2006). We explored this aspect by using ML7 which did not have any effect on cell spreading (data not shown) compared to the DMSO treated cells. Neither the percentage of spread cells nor the phenotype of cells changed. Further studies need to be done to unravel the mechanism involved in cell contractility during cell adhesion, spreading and migration.
To summarize, cells integrate signals coming from the matrix and respond according to the properties of the microenvironment. Adherence and spreading are the processes by which cells attach to their support. Following adhesion, the cellular responses are varied ranging from morphological changes, differentiation, survival, proliferation, and migration. Thus, our model provides a useful tool to understand the interaction between the hematopoietic cells and their niche as well as the consequences on physiological and pathological conditions.