Syntaxin 7 contributes to breast cancer cell invasion by promoting invadopodia formation

Tumor cells acquire invasive capability upon sensing their extracellular milieu. Invasion involves alteration of cell-to-cell and cell–extracellular matrix (ECM) contact and remodeling of the surrounding matrix to form micro-tracks that can be used by other cells during metastasis (Friedl et al., 1997; Wolf et al., 2007; Kawauchi, 2012). ECM remodeling is achieved through the secretion of various matrix-degrading proteases, such as soluble matrix metalloproteinases (MMPs) and membrane-type MMPs. These proteases help cleave the surrounding matrix proteins, enabling pathways for the trailing cells (Stetler-Stevenson, 2001; Chun et al., 2004; Even-Ram and Yamada, 2005; Wolf et al., 2007). Proteolytically active proteases with ECM degradation capability are accumulated and secreted from a specialized protrusive structures, formed at the ventral surface of the cells, termed invadopodia (Artym et al., 2006; Steffen et al., 2008; Poincloux et al., 2009; Kessenbrock et al., 2010; Frittoli et al., 2011; Itoh, 2015; Jacob and Prekeris, 2015; Linder, 2015; Castro-Castro et al., 2016; Cepeda et al., 2016; Jabłońska-Trypuć et al., 2016; Qiang et al., 2019). Invadopodium formation is a tightly regulated process involving initiation, assembly and maturation (Murphy and Courtneidge, 2011). Localized delivery of proteases to the invadopodia requires various intracellular trafficking machinery (Steffen et al., 2008; Frittoli et al., 2014; Kajiho et al., 2016; Sharma et al., 2020). Vesicles of endomembrane origin exploit soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) to target and fuse to the plasma membrane (PM) (Bonifacino and Glick, 2004; Hong, 2005; Jahn and Scheller, 2006; Cai et al., 2007; Südhof and Rothman, 2009). Several studies have shown that SNAREs are involved in transport of MT1-MMP (also known as MMP14) to the cell surface (Miyata et al., 2004; Steffen et al., 2008; Kean et al., 2009; Williams et al., 2014; Clancy et al., 2015; Sneeggen et al., 2019; West et al., 2021) although the role of SNAREs in regulating the invadopodia number in cancer cells is yet to be investigated.

Combining a reverse genetics approach with microscopic and biochemical studies, here we aimed at identifying SNAREs of the endocytic and recycling pathways with potential roles in invadopodia-mediated breast cancer cell invasion. The current study demonstrates that STX7 is crucial for the formation of invadopodia and the associated gelatin degradation and invasion activity. STX7 interacts with multiple SNARE proteins, forming multiple SNARE complexes. Depletion of these interacting partners phenocopies the effects of STX7 depletion.

To investigate the role of SNAREs in matrix remodeling and invasion, 13 SNAREs of the endocytic and recycling pathway were screened using a SMART pool siRNA-based reverse genetics approach. Depletion of six of 13 SNAREs (Fig. S1A) led to significant reduction in gelatin degradation activity in MDA-MB-231 cells (Fig. 1A; Fig. S1B). Subsequently, an Matrigel invasion assay showed a significant reduction in invasive potential upon depletion of STX2, STX4, STX7, VAMP3 and VAMP7 (Fig. 1B; Fig. S1C).

Previously, STX2 was found to be involved in invasion and metastasis by activating MMP-9 expression in hepatocellular carcinoma (Jia et al., 2011). Similarly, STX4, VAMP3 and VAMP7 are known to regulate MT1-MMP trafficking during invasion (Steffen et al., 2008; Kean et al., 2009; Williams et al., 2014; Brasher et al., 2017). Although the role of STX7 in endosome maturation (Mullock et al., 2000; Nakamura et al., 2000; Ward et al., 2000) is well known, its contribution to cancer cell invasion is unexplored. Therefore, we focused on STX7 in the current study.

To exclude off-target effects, an alternative siRNA (siRNA #2) was used, and a CRISPR-Cas9-mediated STX7 knockout clonal (STX7KO) as well as a non-clonal (STX7KO-NC) cell line was established (Fig. 1C′; Fig. S1D). Silencing of STX7 abrogated gelatin degradation (Fig. 1C). A similar phenotype was observed for cells with reduced expression of MT1-MMP (Fig. 1C). The loss of ECM degradation in STX7KO cells could be rescued by overexpressing wild-type GFP–STX7 (STX7FL), although the trans-membrane deleted mutant (STX7Cyto) failed to do so (Fig. 1D). The level of expression of the exogenous STX7 (GFP–STX7) was found to be 1.0±0.1, which was comparable to the endogenous STX7 with expression of 1.1±0.0 (mean±s.d., arbitrary units; Fig. S1E,E′). Expression of the other SNAREs identified from the initial screen in STX7-depleted cells was unaltered under these conditions indicating that the observed phenotype is specific to STX7 (Fig. S1F).

To establish that the involvement of STX7 is not limited to MDA-MB-231 cells, we extended the study to MDA-MB-468 and MCF10DCIS.com cell lines. Immunoblot analysis revealed a similar expression profile for STX7 in both the cell lines to that in MDA-MB-231 cells (Fig. 1E). STX7 depletion led to a significant loss in ECM degradation for both the cell lines, as revealed by gelatin degradation assay (Fig. 1F). Together, these observations implicate the involvement of STX7 in breast cancer cell invasion.

To further understand the role of STX7, we studied its subcellular localization in MDA-MB-231 cells. STX7 was observed to colocalize with EEA1 (45.0±4.0%), Rab11 (31.0±3.2%), Rab4 (28.0±2.6%), and CD63 (15.0±3.9%) (mean±s.d., Fig. S1G,G′). Next, we sought to investigate the association of STX7 with invadopodia. We observed (1) several STX7- and Tks5 (an invadopodium marker, also known as SH3PXD2A; Murphy and Courtneidge, 2011; Zagryazhskaya-Masson et al., 2020)-positive puncta in close vicinity, as revealed by total internal reflection fluorescence microscopy (TIRF-M) (Fig. 2A), (2) STX7 near sites of gelatin degradation along with Tks5 (Fig. 2B), and (3) STX7 puncta positive for Tks5 and F-actin (Fig. 2C). Similarly, STX7 was observed to be associated with another invadopodium marker, cortactin (Mader et al., 2011; Beaty et al., 2013; Hayes et al., 2013). That is, (1) several GFP–STX7 puncta were positive for cortactin and F-actin (Fig. S2A), and (2) STX7 was observed near sites of gelatin degradation along with cortactin (Fig. S2B). These observations suggest that STX7 is associated with endosomes and invadopodia.

Next, we investigated the effect of STX7 overexpression on invadopodium formation. It was observed that, compared to GFP-Vector, the overexpression of STX7 (GFP–STX7) did not stimulate invadopodia formation (Fig. S2C). However, the overexpression of STX7 did promote/increase STX7 colocalization with Tks5 in invadopodia (Fig. S2D) perhaps explaining the positive effect of enhanced gelatin degradation upon GFP-STX7 overexpression (Fig. 1D). Furthermore, to understand whether STX7 overexpression changes its localization, we compared the distribution of endogenous STX7 with that of EEA1 in the presence of GFP–STX7 by immunofluorescence. The percentage overlap of GFP–STX7 with STX7 and EEA1 was observed to be 56.0±6.4% and 52.4±9.3%, respectively (mean±s.d.; Fig. S2E). The localization of STX7 with EEA1 in the presence of GFP–STX7 was found to be 46.0±4.0% (Fig. S2E), which is similar to that found in the untreated condition (45.0±4.0%) (Fig. S1G′). Therefore, STX7 overexpression does not have any effect on its localization (Fig. S2E).

Remodeling of ECM requires targeted delivery of proteases to the invadosomes. Immunostaining revealed that 30.0±3.0% of vesicles positive for STX7 are positive for MT1-MMP and EEA1, whereas only 16.8±1.9% of vesicles positive for STX7 are positive for MT1-MMP and CD63 (mean±s.d.; Fig. 2D). Confocal and TIRF- based live-cell microscopy revealed that GFP–STX7 and Cherry–MT1-MMP co-traffic intracellularly and near the PM (Fig. 2E,F; Movies 1, 2). Thus, our data suggest that STX7 might be involved in the delivery of MT1-MMP to the cell surface from endosomes. This observation is in concurrence with earlier reports from multiple cell lines showing that STX7 undergoes recycling between endosomes and the PM (Heng Wong et al., 1998; Prekeris et al., 1999; Kasai and Akagawa, 2001; Bilan et al., 2008; Feldmann et al., 2009).

Since STX7 and MT1-MMP co-traffic to the PM, we hypothesized that STX7 depletion would impede MT1-MMP delivery. Surprisingly, TIRF-M imaging in STX7KO cells expressing pHluorinMT1-MMP (pHMT1-MMP) revealed a significant increase in the number of MT1-MMP vesicles near the PM (Fig. 3A). However, STX4 gene depletion, as expected, caused reduced protease delivery to the surface (Fig. 3A,A′; Movies 3, 4). We quantified the surface MT1-MMP level through biotinylation experiments and found a significant increase in the PM pool compared to the control, whereas levels of endogenous MT1-MMP remained unaltered in the STX7 KD cells (Fig. 3B). Surface biotinylated β1-integrin and EGFR levels remained unaltered (Fig. 3B). Next, we carried out an anti-MT1-MMP antibody uptake assay (Remacle et al., 2003) in STX7KO cells and observed significantly increased uptake of the antibodies, consistent with increased MT1-MMP availability at the PM (Fig. 3C). The above findings demonstrate that silencing of STX7 results in significantly higher MT1-MMP levels at the PM. To rule out the possibility that STX7 perturbation is altering endocytic pathways in the MDA-MB-231 cells, MT1-MMP and transferrin (Tfn) uptake were measured. There was no effect observed on the endocytosis of MT1-MMP or Tfn, as measured by percentage of endocytosed MT1-MMP (Fig. 3D) or internalized Tfn (Fig. S3A). However, compared to control cells, the increased intensity of accumulated Tfn at the end of 60-min of chase suggested that STX7 is involved in recycling of this cargo (Fig. S3B).

We also used a dominant-negative form of STX7 to study its role in MT1-MMP delivery. The overexpression of STX7Cyto reduced MT1-MMP antibody uptake (Fig. S3C) and surface MT1-MMP level (Fig. S3D), as revealed by confocal and TIRF microscopy, respectively. The opposing effect of STX7 depletion versus overexpression of STX7Cyto on MT1-MMP levels at the PM is possibly due to the dominant-negative effect caused by the mutant. The ability of the mutant to bind to its cognate SNAREs leading to the formation of inactive complexes might be a possible reason for the observed disparity (Suzuki et al., 2003).

The contrasting outcomes of higher protease delivery to the surface (Fig. 3A–C) but a reduced invadopodia-associated protease activity (Fig. 1A–C) prompted us to measure the availability of MT1-MMP at the invadopodia. Results from TIRF-M showed that in MDA-MB-231 STX7KO cells there was reduced numbers of Tks5-positive MT1-MMP puncta at the surface compared to the control. However, Tks5-negative MT1-MMP puncta were abundant at the TIRF plane (Fig. 4A), thus corroborating our above observations (Fig. 3A–C). This prompted us to investigate whether reduced MT1-MMP availability at the invadopodia is due to mis-trafficking or reduced invadopodia formation itself. MDA-MB-231 STX7KO and STX7KO-NC cells immunostained for the invadopodia markers Tks5 and cortactin (Fig. 4B) showed reduced invadopodia formation under TIRF-M. Reduction in invadopodia formation was further confirmed by examination of Tks5 and F-actin in STX7KO cells (Fig. 4C). Similar observations showing reduced invadopodia formation and fewer MT1-MMP-positive invadosomes were obtained with confocal microscopy (Fig. 4D). Several reports have demonstrated the role of cortactin phosphorylation in promoting invadopodia formation or degradation depending upon the phosphorylated residue (Sangrar et al., 2007; Ayala et al., 2008; Mader et al., 2011; Mukherjee et al., 2012; Zhao et al., 2012; Beaty et al., 2013; Hayes et al., 2013; Rosenberg et al., 2017). We measured the level of tyrosine-phosphorylated cortactin (phospho-cortactin) and found it to be slightly higher in the STX7KO cells (Fig. 4E).

Overall, we conclude that the depletion of STX7 results in targeting more MT1-MMP vesicles to the surface but reduces invadopodia formation, leading to compromised ECM degradation and Matrigel invasion.

STX7 mediates homotypic fusion of late endosomes (LEs) and heterotypic fusion of LE–lysosomes by forming trans-SNARE complexes consisting of STX7–STX8–Vti1b–VAMP7 or STX7–STX8–Vti1b–VAMP8, respectively (Pryor et al., 2004; Achuthan et al., 2008). Here, we sought to identify cognate SNARE partner(s) of STX7 that might be involved in fusion of MT1-MMP vesicles at invadopodia. We analyzed the endosomal and PM-residing SNAREs VAMP2, VAMP3, VAMP7, STX2, STX4 and SNAP23, which have been shown to have roles in ECM degradation either from our screen (Fig. 1A) or as reported previously (Miyagawa et al., 2019). A modified nanobody-based pulldown approach was employed (Rothbauer et al., 2006, 2008; Tang et al., 2013). GFP–SNAP23, GFP–VAMP2 and GFP–VAMP3 were able to pull down STX7 from cell lysates (Fig. 5A; Fig. S4A,B). The interaction of GFP–STX4 and GFP–VAMP7 with the STX7 was weak compared with the interaction seen for the other proteins (Fig. 5B; Fig. S4C,D). No interaction between endogenous STX2 and STX7 was detected (Fig. 5C). In concordance with these results, VAMP3 and VAMP7 have been previously shown to immunoprecipitate STX7 from HeLa cells and rat liver membrane (Pryor et al., 2004; Gordon et al., 2009).

Since STX7 interacted with multiple cognate SNARE partners, confocal microscopy was used to confirm the overlap of compartments containing endogenous STX7 and its partners in MDA-MB-231. STX7 colocalized abundantly with VAMP2 (43.6±2.2%) followed by VAMP3 (30.5±2.8%), VAMP7 (22.0±3.7%), SNAP23 (14.9±2.4%) and STX4 (14.0±0.3%) (mean±s.d.; Fig. 5D,D′). Furthermore, the colocalization of STX7 was assessed with HA–VAMP3 and GFP–VAMP2, HA–VAMP3 and GFP–VAMP7, and HA–VAMP3 and GFP–STX4 to measure the overlap of STX7 with double-positive interacting partners (Fig. S4E,E′). TIRF-M live-cell imaging also revealed that these binding partners co-traffic to the cell surface (Movies 5–7).

Based on these observations, we hypothesized that a lack of the above interacting SNAREs might phenocopy what was seen with a lack of STX7 regarding invadopodia-associated activity in MDA-MB-231 cells. Among the SNAREs, SNAP23 and VAMP7 were previously reported to contribute to trafficking of MT1-MMP to invadopodia (Williams et al., 2014). Moreover, and consistent with earlier reports and our present study, STX4 also plays a role in MT1-MMP trafficking (Brasher et al., 2017; Fig. 3A). Here, we observed that depletion of VAMP2 (Fig. S4F) led to a significant reduction in gelatin degradation (Fig. 6A). Depletion of VAMP2, VAMP3 and STX4 (Fig. S4F; Fig. 3A′) led to a significant reduction in invadopodia formation (Fig. 6B–D) in MDA-MB-231 cells. Moreover, STX4 depletion resulted in reduced numbers of MT1-MMP vesicles near the PM that contrasted to the STX7 phenotype, where a subtle but significant increase in the MT1-MMP vesicle numbers was observed (Fig. 3A). This prompted us to assess the effects of silencing both the SNAREs. Contrary to depletion of STX7, co-depletion of STX4 and STX7 resulted in a reduction of MT1-MMP numbers vesicles at the PM, similar to what was seen with STX4 depletion (Fig. 6E; Movie 8). These observations indicate that both SNAREs are important for invadopodia formation and function. STX4, however, appears to have a more generalized function for MT1-MMP trafficking to the cell surface, as it abruptly causes a reduction in surface localization. STX7, in contrast, has a more-specific role in delivery of the protease to invadopodia in support of their degradative activity.

The present study highlights that STX7 is crucial for invadopodia formation and associated activity. There could be several mechanisms by which STX7 contributes to this process. First, STX7 might be involved in the trafficking of signaling cargo crucial for the initiation stage of invadopodia formation (Mori et al., 2021; West et al., 2021). Integrins and EGFR have been shown to play an important role in this stage (Destaing et al., 2010; Mader et al., 2011). However, we observed surface β1-integrin and EGFR levels to be unaltered upon STX7 depletion (Fig. 3B). Second, there could be perturbed trafficking of MT1-MMP to the invadopodia. Previous reports indicate that invadopodia are formed at specialized domains of the PM containing cholesterol-rich lipid rafts. MT1-MMP has been observed to be enriched in lipid raft membrane fractions along with the other invadopodia-associated proteins in MDA-MB-231 cells (Yamaguchi et al., 2009). It is possible that STX7 mediates specific delivery of MT1-MMP to lipid raft-enriched domains at the cell surface, initiating the assembly of a MT1-MMP–p130^CAS–SHIP2 complex at nascent invadopodia, preceding its expansion (Ferrari et al., 2019; Zagryazhskaya-Masson et al., 2020). This is impeded when STX7 is depleted, hence causing reduced invadopodia formation. Third, numerous studies have demonstrated that, although phosphorylation of cortactin on Tyr-421 and Tyr-466 by Src and Arg (also known as Abl2) is crucial for the invadopodia formation (Oser et al., 2009; Mader et al., 2011; Weidmann et al., 2016) and efficient ECM degradation (Clark et al., 2007; Ayala et al., 2008; Clark and Weaver, 2008), the phosphorylation of cortactin at Ser-405 and Ser-418 by ERK proteins leads to its E3 ubiquitin-ligase mediated degradation (Zhao et al., 2012). Hakai, an E3 ubiquitin-ligase, recognizes cortactin phosphorylated at Tyr-482. However, it still remains unclear whether Hakai targets proteasome-mediated cortactin degradation in cancer cells (Mukherjee et al., 2012; Jeannot and Besson, 2020). Our study suggests that STX7 silencing led to increased phospho-tyrosine cortactin levels (Fig. 4E), but decreased invadopodia formation (Fig. 4B,C), which might be explained by hyper-phosphorylation of Tyr-482 leading to its degradation and causing invadopodial disassembly. Fourth and finally, in an alternative hypothesis regarding the role of STX7 in invadopodia formation, a report suggests that, in comparison to soft polyacrylamide gels, breast carcinoma cells cultured over hard polyacrylamide gels produce more invadopodia contributing to more ECM degradation (Alexander et al., 2008). This observation is further supported by a recent computational study demonstrating enhanced MT1-MMP secretion at invadopodia in stiff ECM (Kim et al., 2022). It was also found that, during membrane docking and fusion, the tension generated at the membrane due to the stiff ECM environment regulates the fusion pore dynamics and release of the vesicular content (Wang and Galli, 2018). It has been observed that substrate rigidity from 1.5 kPa to 28 kPa promotes the exocytosis of VAMP7 vesicles in fibroblasts, whereas VAMP2 exocytosis was insensitive to rigidity (Wang et al., 2018). Based on these studies, we hypothesize that increased matrix stiffness might trigger a mechanical response that allows STX7 to support transport of cargo molecules involved in initiation of invadopodia formation. The cargo molecules delivered to the surface initiate the formation of contact sites that contribute to MT1-MMP-dependent invadopodia formation (Zagryazhskaya-Masson et al., 2020). This can not be achieved when STX7-mediated cargo trafficking is impaired.

Our data also suggest that STX7 and its cognate partners VAMP2, VAMP3 or SNAP23, have stronger interaction than STX4 and VAMP7 (Fig. 5A,B; Fig. S4A–C). Of note, this is the first study to demonstrate the interaction between STX7 and VAMP2 or SNAP23 in a cellular context. VAMP3 depletion led to perturbed invadopodia formation (Fig. 5F) and cell invasion (Fig. 1B), consistent with previous studies showing the role of VAMP3 in MT1-MMP transport (Kean et al., 2009; Clancy et al., 2015; Sneeggen et al., 2019). We also observed the interaction of STX7 with VAMP3 in MDA-MB-231 cells, in agreement with a previous study performed in HeLa cells to investigate the trafficking of R-SNAREs (Gordon et al., 2009). Furthermore, VAMP2 has been reported as having a role in the fusion of recycling vesicles carrying Tfn–Tfn receptor (Kubo et al., 2015) and α5β1 integrin (Hasan and Hu, 2010) to the PM, which is important for cell survival and migration in HeLa cells. Collectively, we propose that STX7 might be forming a functional SNARE complex with VAMP2 (Movie 5) or VAMP3 (Movie 6) that contributes to MT1-MMP-associated invadopodial activity. In contrast, STX4 exhibited weak interaction with STX7 and showed 15% colocalization. Accordingly, vesicles bearing both the SNAREs were also less abundant near the cell surface (Movie 7). Moreover, whereas STX4 depletion led to reduced MT1-MMP transport to the surface, the depletion of STX7 caused an enhanced MT1-MMP pool at the cell surface (Fig. 3A). Although depletion of both the molecules resulted in reduced invadopodia formation, and hence abrogated ECM degradation (Figs 5F and 1A), the role of these two SNAREs in MT1-MMP trafficking to cell surface appears to be distinct. The enhanced MT1-MMP trafficking upon STX7 silencing suggests a possible inhibitory role of STX7 in regulating the surface pool of the protease. Similar inhibitory role of t-SNAREs have been reported in an earlier study, where authors have demonstrated STX2 as an inhibitory SNARE in insulin granule secretion in pancreatic β-cells (Zhu et al., 2017).

Taken together, we conclude that STX7 forms multiple SNARE complexes by interacting with STX4, VAMP2, VAMP3 and VAMP7. Depletion of any of these SNAREs causes reduced invadopodia formation and breast cancer cell invasion (Fig. 7).

The anti-rabbit EEA1 antibody was a kind gift from Professor Marino Zerial (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany) and was used at 1:1000 [for immunofluorescence (IF)]. The following antibodies were purchased commercially: mouse anti-MT1-MMP (Millipore, MAB3328), 1:1000 [immunoblotting (IB)], 1:500 (IF); mouse anti-MT1-MMP (R& D Systems, MAB9181-SP), 1:200 (IF); mouse anti-vinculin (Sigma, V9131), 1:1000 (IB); mouse anti-cortactin (Millipore, 05-180), 1:1000 (IB), 1:300 (IF); mouse anti-transferrin receptor (Invitrogen, 136800), 1:500 (IF); rabbit anti-actin (Sigma, A2066), 1:2000 (IB); mouse anti-GFP (Roche, 11814460001), 1:2000 (IB); mouse anti-γ-tubulin (Sigma, T6557), 1:3000 (IB); mouse anti-CD63 (DSHB, H5C6), 1:500 (IF); rabbit anti- STX2 (Proteintech, 55033-1-AP), 1:1000 (IB), mouse anti- STX7 (Santa Cruz Biotechnology, sc-514017), 1:1000 (IB), 1:300 (IF); mouse anti-β1-integrin (P5D2, DSHB), 1:1000 (IB), mouse anti-phospho-tyrosine (Millipore, 05-1050), 1:1000 (IB); and rabbit anti-Tks5 (Santa Cruz Biotechnology, sc-30122), 1:500 (IF).

The pEGFPN1-STX7 (full length) and pmCherryN1-STX7 (full length) construct were generated by amplifying STX7 from cDNA obtained from MDA-MB-231 cells with enzyme sites SacI/SacII using the following primers: Fwd, 5′-AAGAGCTCATGTCTTACACTCCAGGAGT-3′ and Rev, 5′-ATCCGCGGGTGGTTCAATCCCCATA-3′ and then cloning into pEGFPN1 and pmCherryN1 (kind gifts from Professor Marino Zerial, Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany). The Rat STX7cyto gene was amplified from pET28a-C(+) vectors (Professor Reinhard Jahn's laboratory, Max Planck Institute for Biophysical Chemistry, Göttingen, Germany) and cloned into pEGFPN1 and pmCherryN1 mammalian expression vectors with enzyme sites XhoI/HindIII using the following primers: Fwd, 5′-CTCGAGATGTCTTACACTCCGGGGATT-3′ and Rev, 5′-AAGCTTCTTTTCTGGATTTGCGCTGATAATTTGCC-3′.

MDA-MB-231 and MDA-MB-468 cell lines were purchased from the ATCC and authenticated for contamination. Cells were maintained in L-15 (Invitrogen) medium with 10% fetal bovine serum (FBS), 100 μg ml⁻¹ penicillin, and 100 μg ml⁻¹ streptomycin (Invitrogen), hereafter L-15 complete medium, at 37°C in CO₂ free conditions. The MCF10DCIS.com cell line (kind gift from Professor Phillipe Chavrier, Centre National de la Recherche Scientifique, Paris, France) was maintained in DMEM-F12 medium (Invitrogen) supplemented with 2 mM glutamine with 5% horse serum (Invitrogen) and maintained at 37°C in 5% CO₂ conditions. HeLa and HEK293T (ATCC) cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% FBS, 100 μg/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine (all from Invitrogen) at 37°C with 5% CO₂. For transient transfection of expression plasmids, cells were grown on coverslips in 24-well plates 1 day before the experiment. Cells were transfected at 50–60% confluency with 0.5 μg of plasmid DNA using LTX/Plus transfection reagent (Life Technologies).

Cells were seeded 24 h before performing siRNA transfection. A standard protocol, available from Dharmacon, was used for transfection using Dharmafect as the transfection reagent. SMARTpool siRNAs targeting the gene of interest were purchased from Dharmacon. All siRNA were used at the working concentration of 10–20 nM. The sequences of SMARTpool siRNAs are provided in Table S1. Cells were seeded 24 h before performing shRNA transfection. Plasmids for expression of shRNAs targeting the gene of interest were MISSION shRNA plasmids (Sigma). Cells were transfected at 50–60% confluency with 0.5 μg of plasmid DNA using LTX/Plus transfection reagent (Life Technologies). The sequences for MISSION plasmid shRNAs are provided in Table S2.

Total RNA was extracted from the cells using RNA easy kit (Qiagen, Cat. 74104), and cDNA was prepared using the High-Capacity RNA-to-cDNA kit (Life Technologies, Cat. 4387406). Real-time quantitative (q)PCR reactions were performed using the SYBR Green Kit and corresponding primers on Applied Biosystems 7300 real-time PCR system. Primer sequences are available upon request.

Cells were fixed in 4% paraformaldehyde (PFA) for 15 min at room temperature. Indirect immunofluorescence was performed on fixed cells permeabilized with 0.1% Triton X-100 for 10 min at room temperature, followed by blocking with 5% FBS in 1× PBS before staining with respective antibodies. The coverslips were mounted using Mowiol (Calbiochem, Cat. 475904) on glass slides and imaged using a Zeiss LSM 780 laser-scanning confocal microscope with the Zen 2010 software.

Data from three independent experiments were subjected to analysis by the automated image analysis program Motion Tracking (http://motiontracking.mpi-cbg.de). The cells were randomly selected for imaging. At least 15 images were acquired for each condition in a given experimental setup, and all the images were processed together for quantification. The objects were identified as vesicles in each channel based on their size, fluorescence intensity and other parameters by the automated Motion Tracking image analysis program. Objects detected in two different channels were considered to be colocalized if the relative overlap of respective areas was >35% (Rink et al., 2005; Collinet et al., 2010; Priya et al., 2015). The apparent colocalization value was calculated as the ratio of integral intensities of colocalized objects to the integral intensities of all objects carrying a given marker, and varied from 0.0 to 1.0. The colocalization-by-chance (random colocalization) was estimated by random permutation of objects localization in different channels. The apparent colocalization was corrected for random colocalization. Based on this method of calculating colocalization, we thereafter defined multicolor objects as as those with a colocalization value greater than 0.35.

Cells were co-transfected with GFP- and mCherry-fused proteins for 12 h, followed by trypsinization, and were seeded on glass-bottom dishes coated with gelatin. Cells in L-15 complete medium were incubated for 4 h at 37°C in an incubator and further imaged with an Olympus FV3000 confocal laser-scanning microscope with a 60× Plan Apo N objective (oil, 1.42 NA) on an inverted stage. Dual-color sequential imaging was performed using 488-nm and 561-nm lasers to excite GFP and mCherry fusion proteins, respectively. Images were acquired and processed using FV31S-SW software.

To capture events specifically near the cell surface, images were captured with a Nikon Eclipse Ti2-E inverted microscope system equipped with H-TIRF setup and an on-stage incubation system with temperature and CO₂ controller (Okolab). Cells co-transfected with GFP and mCherry fusion proteins were plated on gelatin-coated, glass-bottom dishes filled with L-15 complete medium. Simultaneous dual-color imaging was performed using an Apo-TIRF 60× oil DIC N2 objective (oil, 1.49 NA), and a Nikon laser combiner LU-N4 (Nikon Corporation) with two Orca-Flash4.0 V3 sCMOS cameras (Hamamatsu Photonics). Image acquisition was performed with NIS-Elements AR 5.11.00. To excite GFP and mCherry, 488-nm and 561-nm lasers were used, respectively. Bandpass filters were selected to control fluorescence emission detection. Cells were transfected with siRNA or sgRNA, followed by pHluorin-fused MT1-MMP (Professor Phillipe Chavrier) and images were acquired at an exposure time of 300–500 ms.

The number of MT1-MMP vesicles that are available in the TIRF plane and could form a track spanning a minimum of four consecutive frames in live-cell TIRF video captured over 1 min duration was quantified.

Detailed steps involved in quantification are as follows:

(1) Object identification. Object-based analysis was performed by the automated image analysis program Motion Tracking (http://motiontracking.mpi-cbg.de; Rink et al., 2005; Collinet et al., 2010). The core of Motion Tracking is finding the position of objects in the acquired frames of the movie, such as a fluorescent molecule. Once found, the software assigns an x-y position and quantifies the intensity profiles of the objects. Each object is basically a group of pixels. The number of pixels combined with the scale will lead to ‘object size’ and ‘area’ determination. Each pixel has its own brightness called ‘intensity’. The vesicles are identified as objects by setting parameters.

(2) Identifying the tracks. Once Motion Tracking has successfully localized every object in the movie, the objects in consecutive frames can be linked into trajectories, or tracks. A track is the location of an object over time. Tracks have properties such as speed, direction, maximum displacement, intensity and area that are adjusted to identify tracks. The relative weights can take any positive values from 0 to 1.

(3) Setting track break thresholds. To adjust the track break thresholds, the same dialogue box is used. Parameters are: total score threshold, area score threshold, intensity score threshold, and integral score threshold. These thresholds are critical as they will command the program to select a range of the variation in properties of the object that it can tolerate while building a track.

(4) Searching of tracks in the entire movie.

(5) Quantifying the number of vesicles near the cell surface. This is the number of tracks detected in a 1 min movie. Each of the tracks identified using above method are considered as an MT1-MMP vesicle near the TIRF surface.

Fixed-cell TIRF-M was performed by plating the siRNA- or shRNA-treated cells over a gelatin-coated glass-bottom dish for 8 h or 4 h, respectively, followed by fixation using 4% PFA, permeabilization with 0.1% Triton X-100 for 10 min at room temperature, and blocking with 5% FBS in 1× PBS before staining with respective antibodies. Finally, after staining, the cells were kept in 1× PBS for performing TIRF imaging. Z-stack images of the fixed samples were acquired with z-step of 0.02 µm, with total step count of 51. Deconvolution was performed using the Richardson–Lucy method with 20 iterations with NIS-Element AR 5.11.00 software. Acquired images were searched for multicolor objects (as described above) using the automated image analysis program Motion Tracking.

MDA-MB-231 cells were trypsinized and counted, and 30,000 cells per well were seeded on the gelatin-coated coverslips for 24 h and transfected with siRNA. After 72 h of transfection, antibody uptake was performed using anti-MT1-MMP antibody (R&D Systems, MAB9181-SP). Cells were incubated at 4°C with 10 μg ml⁻¹ anti-MT1-MMP antibody diluted in serum-free L-15 medium for 1 h. The unbound antibody was removed by washing with 1× PBS. L-15 complete medium was added, and cells were shifted to 37°C to allow internalization of the surface-bound antibody. At the 15-min time point, cells were fixed by adding 4% PFA, followed by staining with DAPI.

MDA-MB-231 cells were trypsinized and counted, and 30,000 cells per well were seeded for 24 h and transfected with siRNA. Cells were incubated at 4°C with 10 μg ml⁻¹ Alexa Fluor 568-labeled transferrin diluted in L-15 serum-free medium for 1 h. The unbound antibody was removed after washing with 1× PBS. L-15 complete medium was added, and cells were shifted to 37°C to allow internalization for 5 and 10 min. Cells were fixed by adding 4% PFA, followed by staining with DAPI.

Similarly, HeLa cells were transfected with siRNA. After 70 h of knockdown, cells were serum starved for 2 h and pulse labeled with Alexa Fluor-568-conjugated transferrin (20 μg/ml) diluted in serum-free DMEM (SFM) for 1 h at 4°C. Cells were washed with 1× PBS to remove unbound transferrin. Alexa Fluor 568–transferrin-labeled surface receptors were chased in the presence of holotransferrin (100 μg/ml) diluted in L-15 SFM for different time points. At each time point, cells were quickly washed and fixed using 4% PFA and further proceeded for immunofluorescence.

MDA-MB-231, MDA-MB-468 and MCF10DCIS.com cells after 60 h of transfection were trypsinized and 50,000 cells were plated on Alexa Fluor 568-labeled fluorescent gelatin for 12 h. Cells were then fixed and stained with DAPI. Images were captured using a Zeiss LSM780 confocal laser scanning microscope with a 40×1.4 NA oil immersion objective. Images were processed in the Motion Tracking software, where degradation area was identified as an object and the degradation index was calculated as described in Sharma et al. (2020).

The Matrigel invasion assay was performed by transfecting MDAMB-231 cells with 20 nM siRNA targeting the gene of interest. After 60 h of transfection, cells were trypsinized and counted, and 50,000 cells were resuspended in serum-free medium and seeded on BD Transwell cell inserts (8-μm pore size) coated with 0.5 mg ml⁻¹ of Matrigel (BD Biosciences, Cat. 354277). Complete medium with 10% FBS was added to the lower chamber. Transwell plates were kept at 37°C for 16–18 h. Cells on the inserts were fixed with 4% PFA for 20 min, and washed three times with 1× PBS. Non-invaded cells were removed using a cotton swab. Inserts were dried for 6 h at room temperature and imaged using an inverted epifluorescence microscope using a 10× objective in bright field mode. Cells were counted using ImageJ software.

Cells were treated with a pervanadate/H₂O₂ mixture; for this, 1 ml of a 20 mM sodium orthovanadate stock solution was mixed with 330 µl of 30% H₂O₂ and incubated for 5 min at room temperature, yielding a solution of 6 mM pervanadate plus remaining H₂O₂. To stimulate cells with 40 µM pervanadate/H₂O₂, the prepared solution was diluted 1:5 with H₂O, and 16 µI of this solution was diluted in 2 ml of serum-free L-15 medium, and 500 µl was added to each well containing 3×10⁵ cells. As pervanadate/H₂O₂ is unstable, a fresh solution was prepared each time. Cells were treated with pervanadate for 30 min at 37°C. Approximately 20 s before the end of the desired time, cells were pelleted (12,000 g for 4 min) and lysed using lysis buffer consisting of 137 mM NaCl, 50 mM Tris-HCl (pH 7.8), 10% (v/v) glycerol, 1 mM sodium orthovanadate, 2 mM EDTA (pH 8.0), 1% vol/vol Nonidet P-40, plus protease inhibitors leupeptin and aprotinin (10 µg/ml each; Sigma), and 1 mM PMSF. Cell debris was spun down for 15 min at 19,000 g at 4°C and cleared cellular lysates were analyzed by anti-phosphotyrosine immunoblotting (Wienands et al., 1996; Cory et al., 2002).

To measure the surface population, MDA-MB-231 cells (9×10⁴ cells per ml) were grown in a 35-mm plate for 24 h. Cells were then transfected with siRNA. At the 72-h post-transfection time point, biotinylation was performed. The culture medium was removed and the cells were washed carefully with 2 ml PBS (6.7 mM NaHPO₄, 3.3 mM NaH2PO₄, and 140 mM NaCl, pH 7.2) and incubated with EZ-link Sulfo-NHS-Biotin (Invitrogen, Cat. 21217) at 4°C for 45 min. Further, to quench unreacted biotin, cells were treated with a quenching solution (50 mM Tris-HCl, pH 8.0). After quenching, cells were washed with 1× PBS and solubilized in lysis buffer. The lysate was incubated with Neutravidin beads (Pierce, Cat. 29200) for 30 min at room temperature. Biotin-labeled proteins were eluted from the beads and subjected to immunoblotting. The intensity of the biotinylated MT1-MMP was quantified for Scrambled (Scr) and siRNA-treated samples for STX7 using ImageJ software. The scanned western blot image was converted into a grayscale image for densiometric analysis (Sharma et al., 2020).

To measure endocytosis, cells were transfected with siRNA as mentioned above, and surface proteins were labelled with cleavable biotin and EZ-Link Sulfo-NHS-SS-Biotin (0.2 mgml⁻¹; Invitrogen, 21331) for 45 min at 4°C. Unreacted biotin was quenched with the quenching solution. After quenching, cells were washed with 1× PBS and solubilized in lysis buffer (0 min MeSNa−). In parallel, one set was proceed for MeSNa treatment i.e. 50 mM MeSNa in MeSNa buffer [50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA and 0.2 (w/v) BSA, pH 8.6] to remove biotin-labeled proteins, followed by quenching (0 min MeSNa+) after which cells were solubilized. To measure internalization, cells were allowed to endocytose biotinylated proteins at 37°C for 15 min and then shifted to 4°C. After 15 min of endocytosis (15 min, MeSNa−), cells were immediately lysed to measure the total biotinylated population. However, other cells were washed and given MeSNa treatment followed by quenching to remove non-internalized population (15 min, MeSNa+). Cells were solubilized in the lysis buffer to measure the endocytosed population. The biotin-labeled proteins (lysate) were incubated with Neutravidin beads (Pierce, Cat. 29200) for 30 min at room temperature and eluted by boiling the beads in 2× SDS loading dye. Immunoblotting was performed to detect protein levels. The percentage of endocytosis was calculated as: amount endocytosed at time 15 min (t₁₅)= B/A, where A=intensity at t₁₅ (MeSNa−) after background subtraction, and B=intensity at t₁₅ (MeSNa+) after background subtraction.

pGEX6P1-GFPNanobody was purchased from Addgene (Addgene #61838). The GFP-binding protein (GBP) clone was further subcloned into a pGEX-6P1 vector for improved expression of GST-tagged GBP. GFP-Vector or GFP fusion proteins were overexpressed in a six-well format in the cell line. 16 h after transfection, cells were treated with N-ethylmaleimide (NEM; 1 mM) diluted in 1× PBS for 15 min at 4°C, followed by 1 mM NEM diluted in complete medium at 37°C for 30 min. Cells were lysed in 20 mM Tris-HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 2 mM EDTA, and 1% NP-40, and cleared by centrifuging at 19,000 g at 4°C for 15 min. Cleared lysates were mixed with 25 μg of GST–GBP and bound to the glutathione sepharose beads. The beads bound with GBP–GFP protein complexes were then washed three times with PBS, followed by elution using 2× SDS sample loading buffer. Samples were then separated over 12% SDS-PAGE followed by western blotting. Protein transferred over nitrocellulose membrane was sequentially probed with anti-STX7 or anti-STX2, and anti-GFP antibodies.

Cells were harvested and lysed in lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 2 mM DTT, 1% Triton X-100, and 1 mM EDTA) supplemented with 10 μg ml⁻¹ protease inhibitor cocktail (PIC, Sigma). Cell debris was removed by centrifuging at 19,000 g for 15 min at 4°C. Protein concentration was estimated using the Bradford assay. Samples were prepared by adding 1× SDS loading dye and separated by running on SDS-PAGE, then transfer to a nitrocellulose membrane (0.45 μm; GE Healthcare, Cat. 10600002). The membrane was blocked with 5% BSA, then incubated for 1 h at room temperature with the respective primary antibody. To detect the bound antibody signal, the membrane was incubated with fluorescently labelled secondary antibody, and protein bands were detected using the Li-Cor Odyssey Infrared Scanning System. The uncropped blots of main western blots are provided in Fig. S5.

Using an online CRISPR design tool (https://www.thermofisher.com/in/en/home/life-science/genome-editing/geneart-crispr/crispr-libraries/trueguide-grnas.html) from Thermo Fisher Scientific, a 20-nucleotide guide sequence was designed against human STX7. The targeted DNA sequence was the G*G*A*UGUUAGAAGAGAUCCUC+ modified scaffold. A negative control non-targeting single gRNA (sgRNA) 5′-AAAUGUGAGAUCAGAGUAAU-3′ that does not recognize any specific sequence was also used. To generate the clonal cell line, 1000 cells per well were seeded for 24 h before transfection in a 96-well plate. Cells were transfected with Cas9–sgRNA complex at a ratio of 250 ng per 1.5 pmol per protein using Lipofectamine Cas9 plus (Invitrogen, Cat. A36496) and Lipofectamine CRISPRMAX (Invitrogen, Cat. CMAX00003) reagents. At 48 h after transfection, cells were serially diluted to single cells and grown until they reached 80% confluency. CRISPR knockout was confirmed by western blotting. To generate the non-clonal cell line, 1000 cells per well were seeded for 24 h before transfection in a 96-well plate. Cells were transfected as described above, and at 48 h after transfection, transfection reagent was removed and cells were grown until they reached 80% confluency. CRISPR knockout was confirmed by western blotting.

For the statistical evaluation of the datasets from quantitative image analysis, paired or unpaired two-tailed Student's t-test or one-way ANOVA with a Dunnett's multiple comparisons test was performed using GraphPad Prism 6 and 8 software version 6.01 and 8.0.2, respectively. Datasets are presented as mean±s.d. For the data that are displayed using SuperPlots, each biological replicate is distinctly color-coded and each dot represents percentage colocalization for identified multicolor objects or number of identified multicolor objects in a field of view (frame). One frame comprises 1–3 cells, contributing to a total number of cells (n) across the total number of frames acquired. The frame-wise values were separately pooled for each biological replicate to calculate the mean in the SuperPlots and the s.d. represents the deviation of the data from these three means (Lord et al., 2020). The significance is as follows: P>0.05, ns (non-significant), *P≤0.05, **P≤0.01, ***P≤0.001 and ****P≤0.0001. The number of samples, images and experiments used for quantification are mentioned in the respective figure legends or in figures. In all figures, n=number of cells and N=number of experiments.

We thank Professor Philippe Chavrier [Centre National de la Recherche Scientifique (CNRS), Paris, France] for sharing pCDNA3.1 MT1-MMP pHluorin construct and the MCF10DCIS.com line; Professor Marino Zerial (Max Planck Institute, Dresden, Germany) for sharing EEA1 antibody, the pEGFP-C1 Rab4 construct, the pEGFP-C3 Rab11 construct; Professor Reinhard Jahn (Max Planck Institute for Biophysical Chemistry, Göttingen, Germany) for sharing the rat STX7Cyto construct; Professor Andrew Peden (University of Sheffield, South Yorkshire, England) for sharing the protocol for SNAREs interaction and the HA-VAMP3 construct; and Professor Ludger Johannes (Institut Curie, Paris, France) for sharing the pEGFP-C2-VAMP2-myc construct. We thank Professor Sanjeev Shukla (IISERB) for providing us the MDA-MB-468 cell line. We are thankful to the Central Instrumentation Facility at IISER Bhopal that was used to perform confocal and TIRF microscopy-based experiments. We also acknowledge the FIST facility at IISER Bhopal by DST for providing facility to conduct live- cell imaging-based experiments. We sincerely thank Dr Dipanjana Ghosh (IISERB) Dr Amulya Priya (France), Dr Priyanka Sharma (IISERB) for critically analyzing the manuscript and insightful suggestions. We also thank Dr Ebsy Jaimon (IISERB), Ms Achala Apte (IISERB) for suggestions in the manuscript and Rabiya Naaz for lab management.

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.259576.