The role of talin2 in breast cancer tumorigenesis and metastasis

Recent studies show that talin2 has a higher affinity to β-integrin tails and is indispensable for traction force generation and cell invasion. However, its roles in cell migration, cancer cell metastasis and tumorigenesis remain to be determined. Here, we used MDA-MB-231 human breast cancer cells as a model to define the roles of talin2 in cell migration, invasion, metastasis and tumorigenesis. We show here that talin2 knockdown (KD) inhibited cell migration and focal adhesion dynamics, a key step in cell migration, and that talin2 knockout (KO) inhibited cell invasion and traction force generation, the latter is crucial for cell invasion. Re-expression of talin2WT in talin2-KO cells restored traction force generation and cell invasion, but that of talin2S339C, a β-integrin-binding deficient mutant, did not. Moreover, talin2 KO (or KD) suppressed tumorigenesis and metastasis in mouse xenograft models. However, surprisingly, re-expression of talin2WT in talin2-KO cells did not rescue tumorigenesis. Thus, talin2 is required for breast cancer cell migration, invasion, metastasis and tumorigenesis, although exogenous expression of high levels of talin2 could inhibit tumorigenesis.

However, we demonstrated that talin2 had a stronger binding to β integrin tails than talin1 [25,26]. Talin2 Ser339 is largely responsible for this affinity difference and substitution of Ser339 with Leu disrupted the binding of talin2 to β1-integrin tails. We also demonstrated that talin2 is localized at invadopodia and that a strong talin2-integrins interaction is required for traction force generation and invadopodium formation and cell invasion [25].
In the present study, we used highly aggressive breast cancer cell line MDA-MB-231 cells as a model to investigate the role of talin2 in cell migration, invasion, tumor growth and metastasis. We found that talin2 plays a Research Paper crucial role in cell migration, invasion, tumorigenesis and metastasis, although exogenous expression of high levels of talin2 could suppress tumorigenesis.

RESULTS
To examine the role of talin2 in cell migration, endogenous talin2 in MDA-MB-231 cells were depleted by expressing talin2 shRNA ( Figure 1A). Cell migration was determined by time-lapse cell migration assays, as described previously [27]. Depletion of talin2 significantly suppressed the migration of MDA-MB-231 cells ( Figure 1B), by inhibiting directionality and velocity ( Figure 1C). Depletion of talin1 also inhibited cell migration. However, talin1 affected velocity more than directionality, whereas talin2 functioned oppositely (Data not shown). To further examine the biological function of talin2, we used CRISPR/Cas9 to knockout talin2 from MDA-MB-231 cells. Ablation of talin2 was achieved by infecting the cells with lentiviruses that express Cas9 and talin2 gRNAs. Talin2 was completely ablated in clones #1, #3 and #4, while was partially in clone #2 ( Figure 2A). Thus, we chose clones #1, #3, and #4 for further experiments. We examined the role of talin2 in cell migration using transwell migration assays. Ablation of talin2 significantly inhibited cell migration in the absence of a growth factor chemoattractant, whereas had no effect on EGF-induced cell migration (20 ng/ml EGF in lower chambers; Supplementary Figure 1). These results suggest that talin2 has distinct roles in cell migration under different conditions.
Because FA dynamics is crucial for cell migration, we examined the role of talin2 in FA dynamics. MDA-MB-231 cells that stably express DsRed-paxillin were infected with lentiviruses that express talin2 shRNA. FA assembly and disassembly were determined as we described previously [13,27]. As shown in Figure 1D and 1E, depletion of talin2 significantly inhibited both FA assembly/disassembly rates in MDA-MB-231 cells, suggesting that talin2 regulates cell migration by modulating FA dynamics.
To ascertain the role of talin2 in breast cancer cell invasion, the invasion of talin2 knockout (KO) cells was measured by examining the functional capacities of the cells penetrating through transwell filters coated with 0.35 mg/ml Matrigel, using cells infected with empty LentiCrispr vector as a control. Ablation of endogenous talin2 significantly inhibited the basal (without growth factors) as well as HGF-stimulated invasion of MDA-MB-231 cells ( Figure 2B, 2C). Depletion of talin2 by using shRNAs also inhibited the invasion of MDA-MB-468 and MDA-MB-435S cells (Supplementary Figure 2).
Previously, we demonstrated that substitution of S339 with Cys caused significant reduction in the binding of the talin2 head domain to β-integrin tails [25]. We examined whether the mutation also inhibits the binding of the full-length talin2 to β-integrins. As shown in Figure 2D, substitution of S339 with Cys also resulted in significant reduction in the binding of the full-length talin2 to β1A-integrin tail.
To determine the essential role of the talin2-βintegrin interaction in cell invasion, EGFP-talin2 WT and -talin2 S339C were re-expressed in talin2-null MDA-MB-231 cells, respectively ( Figure 2E). The invasive capacities of these cells toward Matrigel were tested, using talin2-null cells and CRISPR vector cells as controls. Expression of EGFP-talin2 WT in talin2-null cells significantly rescued the basal and HGF-stimulated cell invasion ( Figure 2F, 2G). Expression of EGFP-talin2 S339C had no effect on basal invasion, but partially rescued HGF-stimulated invasion (P < 0.05 KO vs S339C), suggesting that a strong binding of talin2 to integrins is required for cell invasion.
Because traction force plays an important role in cell migration and invasion [23,25,28], we set out to examine the role of talin2 in traction force generation. To this end, talin2-null MDA-MB-231 cells were plated on the fibronectin-conjugated polyacrylamide gels containing Red Fluospheres (Life Technologies), using cells carrying empty CRISPR vector as a control. Traction force was measured by a Nikon A1 confocal microscope equipped with a CO 2 incubator system, and was analyzed using the method of Butler et al. Ablation of talin2 almost abolished the traction force generation in MDA-MB-231 cells ( Figure 3A, 3B). To know whether a strong talin2-β-integrin interaction is required for traction force generation, talin2-null MDA-MB-231 cells that express EGFP-talin2 WT or -talin2 S339C were plated on the gels containing Red FluoSpheres, and traction force was measured using talin2-null cells and CRISPR vector cells as controls. Expression of EGFP-talin2WT in talin2-null MDA-MB-231 cells restored more than 85% of traction force, whereas that of EGFP-talin2 S339C , which has reduced affinity to β-integrins, had little effect ( Figure 3C, 3D). This result suggests that a strong binding of talin2 to β-integrins is required for traction force generation.
To study the role of talin2 in tumorigenesis, talin2null MDA-MB-231 cells were injected into female NCRNU mice (Taconic) by subcutaneous injection, using cells that were infected with empty LentiCrispr vector as a control. Tumor volumes were measured at 3-7 day intervals. As shown in Figure 4A, tumor volumes from mice injected with talin2-null cells were significantly smaller than those from mice injected with the cells transfected with empty vector. At 36 days after injection, mice were euthanized, and tumors were removed, photographed and weighted. Talin2 KO significantly inhibited tumor growth ( Figure 4B, 4C). The role of talin2 in tumorigenesis was further demonstrated by using talin2 KD cells. To compare the roles of talin1 and talin2 in tumorigenesis, talin1 and talin2 were depleted with talin1 and talin2 shRNAs, respectively ( Figure 4D). MDA-MB-231 cells that express talin2 shRNAs were injected into female NCRNU mice by subcutaneous injection, using cells expressing talin1 shRNA and cells transfected with empty pLKO.1 vector as controls. Tumor volumes were measured at 3 day intervals. As shown in Figure 4E, tumor volumes from mice injected with the cells expressing talin2 shRNA were significantly smaller than those injected with the cells transfected with empty vector, and were similar to those injected with the cells expressing talin1 shRNA. At 30 days after injection, mice were euthanized, and tumors were removed, photographed and weighted. Either talin1 or talin2 knockdown (KD) significantly inhibited tumor growth ( Figure 4F, 4G). Depletion of talin1 or talin2 also suppressed cell proliferation in vitro ( Figure 4H). These data suggest that, like talin1, talin2 is also indispensable for breast cancer cell tumorigenesis.
To determine whether a strong talin2-β-integrin interaction is essential for tumorigenesis, talin2-null MDA-MB-231 cells that re-express EGFP-talin2 WT or -talin2 S339L were injected into female NCRNU mice (Taconic) by subcutaneous injection, using talin2 null cells as a control. Surprisingly, re-expression of talin2 WT in talin2-null cells completely abolished tumor growth, whereas re-expression of talin2 S339L , an integrin-binding deficient mutant, partially inhibited tumor growth ( Figure  5B-5D). This inhibition is probably caused by the high levels of re-expressed talin2 WT , which are three fold higher than the endogenous talin2 ( Figure 5A). These results suggest that exogenous expression of high levels of talin2 can inhibit tumorigenesis more efficiently than talin2 KD (or KO).
To know the role of talin2 in breast cancer cell metastasis, MDA-MB-231 cells that express talin2 shRNA or a control shRNA were injected into the tail vein of female NCRNU mice (Taconic), comparing to cells expressing talin1 shRNA. After six weeks, mice were euthanized and tumor nodules on the surface of the lungs were examined microscopically. The lungs from the mice injected with the cells expressing talin1 or talin2 shRNAs had significantly less tumor nodules than those injected with the cells infected with the control shRNA ( Figure 6A). Also, hematoxylin and eosin staining showed that the lung sections from the mice injected with cells expressing the control shRNA demonstrated numerous tumor nodules, whereas those from mice injected with cells expressing talin1 or talin2 shRNAs had fewer or no tumor nodules ( Figure 6B). These results indicate that both talin1 and talin2 are indispensable for the metastasis of MDA-MB-231 cells.
To examine the possible association of talin2 with breast cancer progression, human breast cancer tissue array slides, including 17 cases of ductal carcinoma in situ (DCIS), 32 cases of invasive ductal carcinoma (IDC) and 30 cases of tumor adjacent tissues, were stained for talin2. Immunohistochemical staining was performed as we described previously [29]. Talin2 staining was significantly higher in DCIS and IDC tissues than the staining in the adjacent tissues (P < 0.001), and talin2 staining was also higher in IDC tissues than DCIS tissues (P < 0.001; Figure 6C). These data suggest that talin2 expression positively correlates with human breast cancer progression.

DISCUSSION
In breast cancer cells, talin2 mediates traction force generation through its high affinity to β-integrins, thus regulating cell migration, invasion, tumor growth and metastasis. Talin2 is indispensable for tumor growth, although exogenous expression of excessive talin2 can inhibit tumor growth.
We found that depletion of talin2 inhibited the directionality of cell migration in time-lapse migration assays, and that ablation of talin2 suppressed cell migration in transwell migration assays in the absence of a growth factor as a chemoattractant (Figure 1, Supplementary  Figure 1). Although the time-lapse migration assays were performed in the presence of EGF, there was no EGF gradient in the assays. Thus, the common factor in these two different assays is lack of EGF gradient, suggesting that talin2 is indispensable for cell migration when there is no signal cue for the directionality. In contrast, talin2 KO had no significant effect on cell migration in the presence of EGF in transwell assays, where the EGF gradient determined the directionality of migration. This result is consistent with previous findings that talin2 deleted mouse embryo fibroblasts showed no obvious defects in the velocity of migration [24]. These results suggest that talin2 regulates cell migration by governing the directionality.
The role of talin2 in determining the directionality of cell migration could be attributed to its effect in traction  (Figure 3), which can regulate the directionality of cell migration [30]. Traction force also regulates focal adhesion dynamics [31][32][33][34], a key step in cell migration [20]. This is supported by our findings that talin2 is required for focal adhesion dynamics ( Figure 1D, 1E). Moreover, focal adhesions are implicated in determining the directionality of cell migration [35,36]. Thus, talin2 regulates the directionality of migration by modulating traction force and focal adhesion generation.
We recently reported that talin2 is required for EGFstimulated invasion of U2 OS human osteosarcoma cells [25]. We show here that ablation of talin2 also inhibited  (Figures 2, 3). Furthermore, re-expression of talin2 WT in talin2-null MDA-MB-231 cells rescued cell invasion and traction force generation, whereas reexpression of talin2 S339C , a β-integrin-binding deficient mutant, did not. These results indicate that a strong interaction between talin2 and β-integrins is essential for traction force generation and cell invasion. Because traction force plays an important role in invadopodia [25], a key cellular structure for invasion, we conclude that talin2-mediated traction force regulates breast cancer cell invasion.
Talin1 knockdown in MDA-MB-231 cells inhibited their tumorigenesis (Figure 4). This is consistent with a previous report in human hepatocellular carcinoma cells [16]. Similar to talin1, talin2 is also indispensable for tumorigenesis in MDA-MB-231 cells. Ablation of talin2 in MDA-MB-231 cells by CRISPR inhibited their tumorigenesis ( Figure 4). Similar result was observed in mice injected with talin2 knockdown MDA-MB-231 cells. These results suggest that talin2 is essential for tumor growth.
Surprisingly, re-expression of talin2 WT in talin2-null MDA-MB-231 cells did not rescue their tumorigenesis; instead, it completely abolished tumor growth ( Figure 5).
The inhibition of talin2 is probably associated with the abnormal expression levels of the exogenous talin2, which is much higher than the endogenous. The inhibition is dependent of the interaction of talin2 with β-integrins, because re-expression of talin2 S339L , a β-integrin-binding deficient mutant, only partially inhibited tumor growth. Thus, raising talin2 expression could be a better strategy to inhibit breast cancer cell tumorigenesis than depleting it, and it is likely that an activator (but not an inhibitor) of the talin2-β-integrin interaction should be developed for cancer treatment.
Nevertheless, the inhibition of talin2 on tumorigenesis seems to be inconsistent with our findings that talin2 staining in DCIS and IDC tissues was significantly higher than the staining in the adjacent tissues ( Figure 6). The reason could be that the exogenously expressed talin2 is much higher than the talin2 expression in DCIS and IDC. Because a gene has to collaborate with many other genes in the gene regulatory network to regulate breast cancer growth [37,38], it is likely that exogenous over-expression of talin2 alone will disrupt the gene regulatory network. Thus, to exogenously express talin2 or its mutants in breast cancer cells for studying tumor growth, a weaker promoter (such as EF-1α) should be selected. that re-express EGFP-talin2 WT or -talin2 S339L were injected subcutaneously to female NCR-NU mice, using talin2-KO cells as a control. Tumor size was measured when indicated, and tumor volumes were calculated and plotted (B). After 39 days, mice were euthanized. Tumors were removed, and photographed (C). Tumor weights were measured (D). www.impactjournals.com/oncotarget

S339L
. The full-length pAAVS1-EGFP-talin2 S339L was created by digesting full-length pAAVS1-EGFP-talin2 WT with BsrG1/EcoRV and ligating the resulting larger fragment with the smaller fragment from pEGFP-talin2 1- shRNA knockdown shRNA plasmids were co-transfected with packaging plasmids (pMDLg/pRRE, pRSV-Rev, and CMV-VSVG) into 293T cells using Safectine RU50 transfection reagent according to the manufacturer's protocol. The virus particles were applied to overnight cultures of breast cancer cells for infection. Cells that stably express shRNAs were obtained by selecting the infected cells with 1 μg/ml puromycin for 10 days.

Cell migration assays
Cells were treated with trypsin and resuspended in DMEM medium containing 1% FBS and 10 ng/ml EGF, plated at low densities on glass-bottomed dishes (MatTek) coated with 5 µg/ml fibronectin and cultured for three hours in a CO 2 incubator. Cell motility was measured with a Nikon Biostation IMQ. Cell migration was tracked for six hours; images were recorded every 10 minutes. The movement of individual cell was analyzed with NIS-Elements AR (Nikon) as described previously [27,29].
To perform Transwell migration assays, Transwell polycarbonate filters were coated with fibronectin (20 µg/ ml) for one hour. Cells were trypsinized and washed 3 times with DMEM containing 1% FBS. The cells were resuspended in DMEM containing 1% FBS at a density of 5 × 10 5 cells/ml. The cell suspensions (100 µl) were seeded into the upper chambers, and 600 µl of DMEM medium containing 1% FBS, and 10 µg/ml fibronectin (and 20 ng/ml EGF where indicated) were added to the lower chambers. The cells were allowed to migrate for 3.5 hours (or as indicated) in a CO2 incubator, fixed, stained and quantitated as described previously [40].

FA dynamics assays
MDA-MB-231 cells that stably express DsRedpaxillin [27] were infected with lentiviruses that express shRNA control, talin1 or talin2 shRNA and then selected with puromycin. The cells were trypsinized and plated on MatTek dishes (with a glass coverslip at the bottom) that had been precoated with fibronectin (5 µg/ml). The cells were cultured for three hours and TIRF images were taken using the Nikon Eclipse Ti TIRF microscope equipped with a 60×, 1.45 NA objective, CoolSNAP HQ2 CCD camera (Roper Scientific). The temperature, CO 2 and humidity were maintained by using INU-TIZ-F1 microscope incubation system (Tokai Hit). Images were recorded at 1-minute intervals for a 60-minute period. FA assembly and disassembly rate constants were analyzed as described previously [13,27,40].

Cell invasion assays
Cell invasion was performed as described previously [29,40]. Briefly, 100 µl of Matrigel (1:30 dilution in serum-free DMEM medium) was added to each Transwell polycarbonate filter and incubated at 37°C for five hours. Cells were trypsinized, washed (three times), and resuspended in DMEM containing 1% FBS at a density of 5 × 10 5 cells/ml. The cell suspensions (100 µl) were seeded into the upper chambers, and 600 µl of DMEM medium containing 1% FBS, 20 ng/ml HGF and 10 µg/ml fibronectin were added to the lower chambers. The cells were allowed to invade for 10 hours (or as indicated) in a CO2 incubator, fixed, stained and quantitated as described previously [40].

Traction force measurement
Traction force was measured as described previously [23,25]. Briefly, glass-bottom dishes were silanized using silane, and activated using glutaraldehyde. Forty µL of gel containing acrylamide (6%), bis-acrylamide (0.75%), ammonium persulfate, TEMED, and FluoSpheres ® carboxylate-modified beads (diameter 0.2 μm, 1:85 dilution by volume) was added to the dishes and covered by a coverslip. The coverslip was removed, and gels were activated with sulfo-SANPAH under UVA exposure and then conjugated with fibronectin (0.1 mg/ml). Cells were plated on the gels and traction force was measured as described previously [41], using an A1 confocal microscope in Lexington VA Medical Center.

Tumor growth assays
Female NCR-NU mice (Taconic, six-seven weeks old) were maintained and treated under pathogen-free conditions. MDA-MB-231 cells (1 × 10 6 cells/mouse) were injected subcutaneously to female NCR-NU mice. The injected mice were maintained under standard humane conditions. Tumor volume (V) was measured with calipers periodically and calculated by using formula V = 0.5 × LW2 (L, Length; and W, width). Mice were euthanized after 30-40 days. Tumors were removed, weighed and photographed. www.impactjournals.com/oncotarget

Tumor metastasis assays
Tumor metastasis was performed as described previously [27]. Briefly, female ICR-SCID mice (Taconic) were maintained and treated under pathogenfree conditions. Talin1 or talin2 KD MDA-MB-231 cells (1 × 10 6 cells/mouse) were injected into the tail vein of mice (six-seven weeks old), using shRNA vector cells as a control. After six weeks, mice were euthanized and lungs were removed and photographed. Tumor nodules present on the surface of lungs were examined under a dissection microscope or detected in paraffin-embedded sections stained with hematoxylin and eosin.

Immunohistochemical staining
Breast cancer tissue arrays, with stage, grade and normal breast tissue, were purchased from US BioMax. The slides with tissue arrays were deparaffinized with Xylene, dehydrated through three alcohol changes, and performed antigen retrieval in 0.1 M citrate buffer (pH 6.0). The slides were treated with hydrogen peroxide and then blocked with goat serum. The slides were incubated with an affinity purified anti-talin2 rabbit polyclonal antibody and then VECTASTAIN ABC kit (Vector Laboratories, Inc). The slides were visualized by incubating with a DAB substrate kit from the same company. The correlation between talin2 expression and tumor progression was scored and statistically analyzed.