Inhibition of metastatic outgrowth from single dormant tumor cells by targeting the cytoskeleton

Abstract

Metastatic breast cancer may emerge from latent tumor cells that remain dormant at disseminated sites for many years. Identifying mechanisms regulating the switch from dormancy to proliferative metastatic growth has been elusive due to the lack of experimental models of tumor cell dormancy. We characterized the in vitro growth characteristics of cells that exhibit either dormant (D2.0R, MCF-7, and K7M2AS1.46) or proliferative (D2A1, MDA-MB-231, and K7M2) metastatic behavior in vivo. Although these cells proliferate readily in two-dimensional culture, we show that when grown in three-dimensional matrix, distinct growth properties of the cells were revealed that correlate to their dormant or proliferative behavior at metastatic sites in vivo. In three-dimensional culture, cells with dormant behavior in vivo remained cell cycle arrested with elevated nuclear expression of p16 and p27. The transition from quiescence to proliferation of D2A1 cells was dependent on fibronectin production and signaling through integrin beta1, leading to cytoskeletal reorganization with filamentous actin (F-actin) stress fiber formation. We show that phosphorylation of myosin light chain (MLC) by MLC kinase (MLCK) through integrin beta1 is required for actin stress fiber formation and proliferative growth. Inhibition of integrin beta1 or MLCK prevents transition from a quiescent to proliferative state in vitro. Inhibition of MLCK significantly reduces metastatic outgrowth in vivo. These studies show that the switch from dormancy to metastatic growth may be regulated, in part, through epigenetic signaling from the microenvironment, leading to changes in the cytoskeletal architecture of dormant cells. Targeting this process may provide therapeutic strategies for inhibition of the dormant-to-proliferative metastatic switch.

Figure 1. In vitro model for solitary tumor cell dormancy

A) Proliferation of D2.0R and D2A1 in 2-D culture and in 3-D Cultrex® (as described in Methods) n=8 (mean ± SE). Representative result of three experiments (* p≤0.05). B) D2.0R and D2A1 cells were cultured in 3-D Cultrex® (as described in Methods). Images of the cells were acquired at day 4, day 9 and day 12, magnification x20. Images in the upper right corner, day 12, magnification x40. C–D) Quiescence in the 3D is associated with elevated p16 and p27. Immunohistochemical staining for p16 and p27. C) Quiescent phase: D2.0R cells, culture days 4 and 10; D2A1 cells, day 4. D) Proliferative phase of D2A1 cells, day 10. Note loss of p16 expression and reduction of p27 nuclear localization during D2A1 proliferative phase, but continued high expression in quiescent D2.OR cells. Magnification x40.

Figure 2. Correlation of in vivo dormant or metastatic behavior of tumor cells with growth in 3-D culture

A) Proliferation of MDA-MB-231, 4T1, and MCF-7 cells in 3-D Cultrex® n=8 (mean ± SE). Representative result of three experiments. B) Morphology of MDA-MB-231, 4T1, MCF-7, K7M2 AS1.46 and K7M2 cells cultured in 3-D Cultrex®, magnification x40. C) Representative lungs from a mouse injected with GFP-expressing MDA-MB-231 (2 weeks post injection) or MCF-7 cells (9 weeks post injection). Left panels: SCOM images of MDA-MB-231 lung metastasis and dormant MCF-7 cell, magnification x100. Right panels: representative H&E staining of MDA-MB-231 lung metastasis and of lung with MCF-7 cells lacking proliferative metastases, magnification x20. D) Proliferation of K7M2 and K7M2 AS1.46 cells in 3-D Cultrex® n=8 (mean ± SE). Representative result of three experiments. (*** p≤0.001).

Figure 3. Actin stress fiber formation associated with myosin light chain phosphorylation occurs during the transition from quiescence to growth in 3-D culture

D2.0R, D2A1, MCF-7, MDA-MB-231, 4T1, K7M2 and K7M2 AS1.46 cells were cultured in 3-D Cultrex® on glass cover slips (as described in Methods). Cells were fixed and stained with DAPI (blue) for nuclear localization , phalloidin (green) for f-actin and with an antibody against the phosphorylated form of myosin light chain (red), as indicated at various time points; confocal microscopy, magnification x63. A) Proliferative growth is associated with actin stress fibers formation. Cortical f-actin staining was evident in D2.0R cells at day 1 (d1), day 4 (d4) and day 7 (d7) (arrowheads). D2A1 cells exhibited cortical f-actin at day 1(d1) (arrowhead), but actin stress fibers form at day 4 (d4) and day 7 (d7) (arrows). B) Actin stress fiber formation and proliferative growth is associated with MLC phosphorylation. D2A1 cells show localization of phosphorylated MLC with actin at day 1(d1), and with actin stress fibers by day 7 (d7). C) 4T1 and MDA-MB-231 (day 4) and K7M2 cells (day 6) displayed actin stress fibers (arrows), whereas non -proliferative MCF-7 (day 14) and K7M2-AS1.46 (day 10) displayed cortical f-actin staining (arrowheads). White bar equals 20 microns

Figure 4. Cytoskeletal reorganization and formation of actin stress fibers during the switch from dormancy to metastatic growth

A) SCOM images (magnification x100) of lungs from mice injected with either D2A1-GFP (removed 1 or 3 weeks post-injection), D2.0R GFP (removed 4 or 12 weeks post-injection) MDA-MB-231-GFP (removed 2 weeks post-injection) or MCF-7-GFP cells (removed 9 weeks post-injection). Time points varied according to the growth properties of the cells. B) Frozen sections of lungs from mice injected with the above cells at the indicated time points stained for f-actin (red); confocal microscopy, white bar equals 20 microns. White arrowheads depict cortical actin staining; white arrows indicate f-actin stress fibers.

Figure 5. Inhibition of MLCK-mediated actin stress fiber formation blocks proliferation of metastatic cells in 3-D culture and reduces metastatic outgrowth in mice

A) D2A1 cells were cultured in 3-D Cultrex on glass cover slips. Cells were untreated (control), or treated with ML-7 (5 μM), or with W13 (5μM) for 48 hr beginning on culture day 5, or treated with scrambled or MLCK shRNA as described in Methods, and stained for the phosphorylated form of myosin light chain (red), f- actin (green), and nuclei (blue). Merge of f-actin, and MLC-p staining (yellow). Light and confocal microscopy, magnification x40 and x63, respectively. White bar equals 20 microns B) Time course of D2A1 cell proliferation in 3-D Cultrex® in the absence or presence of ML-7 (5 μM), n=8 (mean ±SE). C) Inhibition of the metastatic outgrowth in lungs of mice treated with ML-7. Data presented as the percentage of single cells vs. proliferative metastatic lesions in mice that received either ML-7 (22mM) or vehicle (control group) (p≤0.0001 across all samples)(n=9). Single cells (<1000 pixel intensity) and clusters of cells (>1000 pixel intensity), as depicted in upper right panel.

Fig 6. Fibronectin activates MLCK leading phosphprylation of MLC, actin stress fibers formation and transition from quiescence to growth

D2.0R and D2A1 cells were cultured in 3-D Cultrex® (designated as C) on glass cover slips (as described in Methods). A) Staining for fibronectin and Intβ1 in D2.0R and D2A1 cells day 1 and day 7. B, C) D2A1 and D2.0R cells stained for MLC-p (red) f-actin (green) and nuclei (blue). Merge of f-actin, and MLC-p staining (yellow). B) D2A1 cells treated with either control IgG or neutralizing antibody against Int 1 for 6 days. C) D2.0R or D2A1 cells cultured on 3-D Cultrex® mixed with fibronectin (designated as C+F) (750μg/ml) for 6 days in the presence of either nonspecific IgG (150μg/ml), antibody against Integrin β1 (150μg/ml) or ML-7 (5μM). Confocal microscopy, magnification x63, white bar equals 20 microns