Small GTPase Rab37 targets tissue inhibitor of metalloproteinase 1 for exocytosis and thus suppresses tumour metastasis

Tsai, Chung-Han; Cheng, Hung-Chi; Wang, Yu-Shiuan; Lin, Pinpin; Jen, Jayu; Kuo, I-Ying; Chang, Ying-Hua; Liao, Pao-Chi; Chen, Ruey-Hwa; Yuan, Wei-Chien; Hsu, Han-Shui; Yang, Muh-Hwa; Hsu, Ming-Ta; Wu, Chu-Yi; Wang, Yi-Ching

doi:10.1038/ncomms5804

Article
Published: 03 September 2014

Small GTPase Rab37 targets tissue inhibitor of metalloproteinase 1 for exocytosis and thus suppresses tumour metastasis

Chung-Han Tsai¹,
Hung-Chi Cheng^1,2,
Yu-Shiuan Wang¹,
Pinpin Lin³,
Jayu Jen¹,
I-Ying Kuo¹,
Ying-Hua Chang⁴,
Pao-Chi Liao⁴,
Ruey-Hwa Chen^5,6,
Wei-Chien Yuan^5,6,
Han-Shui Hsu⁷,
Muh-Hwa Yang⁸,
Ming-Ta Hsu⁹,
Chu-Yi Wu¹⁰ &
…
Yi-Ching Wang^1,10

Nature Communications volume 5, Article number: 4804 (2014) Cite this article

6506 Accesses
41 Citations
1 Altmetric
Metrics details

Subjects

Abstract

Rab small GTPases are master regulators of membrane trafficking and guide vesicle targeting. Recent publications show that Rab-controlled trafficking pathways are altered during tumorigenesis. However, whether any of the Rabs plays a metastasis suppressor role is least explored. Here we address the metastasis suppressive function of human Rab37 (hRAB37) using secretomics, cell, animal and clinical analyses. We show that tissue inhibitor of metalloproteinase 1 (TIMP1), a secreted glycoprotein that inhibits extracellular matrix turnover, is a novel cargo of hRAB37. hRAB37 regulates the exocytosis of TIMP1 in a nucleotide-dependent manner to inactivate matrix metalloproteinase 9 (MMP9) migration axis in vitro and in vivo. Dysfunction of hRAB37 or TIMP1 abrogates metastasis suppression. Lung cancer patients with metastasis and poor survival show low hRAB37 protein expression coinciding with low TIMP1 in tumours. Our findings identify hRAB37 as a novel metastasis suppressor Rab that functions through the TIMP1-MMP9 pathway and has significant prognostic power.

You have full access to this article via your institution.

Download PDF

Rab27a plays a dual role in metastatic propensity of pancreatic cancer

Article Open access 30 April 2020

Deubiquitylation of Rab35 by USP32 promotes the transmission of imatinib resistance by enhancing exosome secretion in gastrointestinal stromal tumours

Article 01 February 2023

Rab13 regulates sEV secretion in mutant KRAS colorectal cancer cells

Article Open access 25 September 2020

Introduction

Rab proteins are members of the Ras superfamily of GTPases that switch between GDP-bound (inactive) and GTP-bound (active) forms. They transport cargo proteins to target membrane, organelle and cellular compartment to coordinate the cell membrane dynamics, cytoskeleton organization, signal transduction and membrane trafficking^1,2. Recent publications show that Rab-controlled trafficking pathways are altered during tumour development and progression. Overexpression of Rab8, Rab11, Rab21, Rab23, Rab25 and Rab27B (refs 3, 4, 5, 6, 7, 8, 9) promotes cancer cell migration and tumour invasion. For example, oncogenic Rab8 transports exocytic vesicles carrying MT1-matrix metalloproteinase (MMP) to the matrix degradation sites at the plasma membrane (PM) of migrating cancer cells cultured in collagen gel³. An increased expression of Rab27B is associated with the poor prognosis of oestrogen receptor-positive breast cancer patients. Through mass spectrometric (MS) analysis, heat-shock protein 90α has been identified as a component of Rab27B-regulated vesicles and it acts as a pro-invasive growth regulator required for MMP2 activation⁶. Previous studies indicate that Rab25 functions as a tumour suppressor or oncogene depending on the cancer type^10,11,12. In addition, loss of Rab21 correlates with increased malignancy in prostate and ovarian cancer¹³. Although there are >60 members in the Rab family, whether any of the Rabs plays a role in metastasis suppression is least explored.

We previously cloned the human Rab37 (hRAB37) residing in chromosome 17q25 (accession number NM_001006638) and demonstrated that hRAB37 was downregulated mainly by promoter hypermethylation in lung cancer cells and patients, and low mRNA expression of hRAB37 gene was associated with tumour metastasis in lung cancer patients¹⁴. We speculated that the hRAB37 may act as a metastasis suppressor in lung cancer. The murine Rab37 (mRab37) was found in secretory granules of bone marrow mast cell¹⁵. Rab37 together with Rab3, Rab26, Rab27A and Rab27B are referred to as secretory Rabs owing to their localization in secretory granules¹. Therefore, hRAB37 may function as a regulator in exocytotic pathway. Notably, mRab37 co-localizes with TNF-α-containing vesicles in activated macrophages¹⁶. In addition, mRab37 is involved in the control of insulin exocytosis in murine pancreatic β-cells^17,18. However, more investigation is required to identify the cargos of hRAB37 and to dissect the defect mechanism of hRAB37 in human disease such as cancer. Here we undertook several evidence-based studies to answer our four main questions. Is hRAB37 involved in transportation of exocytotic vesicles or recycling vesicles? What consequences can hRAB37 alteration have on cell migration and metastasis? Are any of the trafficking cargos important for hRAB37-mediated motility control? If so, what are the underlying mechanisms of metastasis suppression by hRAB37 and its cargo?

Tissue inhibitor of metalloproteinase 1 (TIMP1) is a secreted protein that inhibits MMPs and extracellular matrix turnover thereby decreasing cell motility^19,20,21 but very little is known²² about its mode of secretion. Our results provide first compelling evidence from lung cell, animal and clinical studies that hRAB37 is a metastasis suppressor. It suppresses metastasis through regulating TIMP1 for exocytosis to extracellular compartment leading to inactivation of MMP9 invasion signalling. The results of the present study have clinical implication for cancer prognosis prediction.

Results

hRAB37 is involved in transportation of exocytotic vesicles

Secretory pathways include tranportation of secretory granules and endocytic recycling vesicles^1,2. To dissect the trafficking pathway regulated by hRAB37 protein, we first compared the cellular distribution of endogenous hRAB37 with several organelle markers in CL1-0 lung cancer cell with low motility and high endogenous hRAB37 (ref. 14) using immunofluorescence confocal analyses. The markers included secretory granule marker VAMP2 and early endosome marker EEA1. Merged confocal images demonstrated the co-localization of endogenous hRAB37 with VAMP2, but not with endosome marker EEA1 (Fig. 1a). Ectopically expressed wild-type (WT) hRAB37 protein in CL1-5, an isogenic cell line derived from CL1-0 with high motility and low hRAB37 (ref. 14), showed cellular distributions similar to endogenous hRAB37 (Fig. 1b) where it co-localized more with secretory granule marker than with endosome marker.

**Figure 1: hRAB37 mediates trafficking of exocytotic vesicles and not recycling vesicles.**

We further performed quantitative analyses of vesicular stomatitis virus ts045 G protein fused to green fluorescent protein (VSVG-ts-GFP) exocytosis from endoplasmic reticulum through the Golgi to the PM^22,23 in Rab37 knockdown (Rab37KD) or control COS1 cells (Fig. 1c). As shown in epifluorescence microscopy images at 15 min after shifting to a permissive temperature, the arrival of VSVG-ts-GFP to the Golgi was observed in both control and Rab37KD cells. However, the subsequent transport of VSVG-ts-GFP to the PM was impaired by Rab37KD. VSVG signals were mainly retained in the perinuclear region and intracellular vesicles (comparing the 120-min images in control and Rab37KD groups; Fig. 1c). Quantification was performed in 150 cells transfected with control or Rab37KD oligos in three independent experiments (Fig. 1d). To discriminate whether hRAB37 was involved in secretion or endocytic recycling, we also performed quantitative analyses for the recycling of endocytosed transferrin receptor (TfnR)^24,25 (Fig. 1e). Of note, the recycling of internalized TfnR did not show difference between control and Rab37KD COS1 cells at any of the time points examined (Fig. 1e). KD efficiency was checked by immunoblotting (Fig. 1f). All together, these results indicated that hRAB37 is involved in the exocytotic pathway of nascent secretory protein from the Golgi to the PM but not in the recycling endosome-mediated secretory pathway.

TIMP1 localizes in hRAB37-containing vesicles

Thus far, the putative cargos identified for mRab37 are TNF-α in murine macrophage and insulin in murine pancreatic cell lines^16,18. To identify the cargos of hRAB37, we performed an unbiased proteomics screen of conditioned medium (CM) from control and ectopically expressed hRAB37-WT CL1-5 cells. Among the 185 proteins that were identified with high confidence in two separate experiments, secreted TIMP1 level was increased in CM from WT compared with that from control (Supplementary Data 1 and Supplementary Fig. 1). We focused on TIMP1 because it is a secreted glycoprotein that inhibits extracellular matrix turnover thereby decreasing cell motility, but very little is known about its mode of secretion.

To validate our proteomics results, protein levels of TIMP1 and TNF-α in the CM were examined in CL1-5 cells treated with exocytosis inhibitor EXO1. hRAB37-mediated TIMP1 secretion was blocked by EXO1 in a dose-dependent manner and TNF-α secretion was also inhibited, although to a lesser extend owing to low basal secretion level in CL1-5 cells (Supplementary Fig. 2). To further determine whether TIMP1 was actually present in hRAB37-containing vesicles in various cell types, we performed biochemical and microscopic analyses. We found that TIMP1 was associated with hRAB37 in the same vesicle lysate fractions using immunoprecipitation western assay (Fig. 2a). We then performed immunogold electron microscopy (immuno-EM) assay for ultrastructural localization of hRAB37 and TIMP1, and found that the endogenous Rab37 and TIMP1 localized on the same vesicles in COS1 cell and mouse lung tissue (Fig. 2b).

**Figure 2: hRAB37 co-localizes with TIMP1 in the same vesicular structures.**

Next, we determined whether TIMP1 localized in hRAB37-containing vesicles in a nucleotide-dependent manner. To this end, we established four stable clones in CL1-5 cells, including empty vector control (control), Flag-hRAB37-WT, Flag-hRAB37-Q89L (Q89L) constitutively active GTP-bound mutant and Flag-hRAB37-T43N (T43N) dominant-negative GDP-bound mutant cells. We measured TIMP1 and hRAB37 co-localization proportion of 100 TIMP1-containing vesicles each in control, WT, Q89L and T43N cells by immuno-EM in two independent experiments. Strikingly, WT and Q89L cells showed significantly higher co-localization proportion (WT: 46%, Q89L: 50%) than those in control cells (12%). Conversely, there was a marked reduction of co-localization proportion (13%) in T43N cells (Fig. 2c,d). Our quantitative immuno-EM studies also revealed that T43N-hRAB37 displayed 20% more distribution at the nuclear periphery region and lost membrane binding to the vesicles compared with WT- and Q89L-hRAB37 (Supplementary Fig. 3). Collectively, these results demonstrate that TIMP1 localizes in hRAB37-containing vesicle in a nucleotide-dependent manner.

TIMP1 is a trafficking cargo of hRAB37

Furthermore, we determined whether hRAB37 could mediate TIMP1 secretion in a nucleotide-dependent manner. Confocal immunofluorescence microscopy of four established stable clones of CL1-5 cells demonstrated that TIMP1 and hRAB37 co-localized in WT and Q89L cells, whereas TIMP1 and hRAB37 were distributed to distinct regions in the T43N inactive hRAB37 cells (Fig. 3a). Importantly, higher concentrations of TIMP1 in CM from WT and Q89L cells compared with control and T43N cells were found (Fig. 3b). These data indicated that hRAB37 regulates TIMP1 secretion in a nucleotide-dependent manner. Notably, although RAB8a transports exocytic vesicles, it did not regulate TIMP1 secretion (Supplementary Fig. 4). These results further support the notion that hRAB37 specifically regulated TIMP1 secretion.

**Figure 3: hRAB37 regulates TIMP1 trafficking in a nucleotide-dependent manner.**

For real-time visualization of the effect of hRAB37 on TIMP1 trafficking, CL1-5 cells were analysed using two-colour total internal reflection fluorescence (TIRF) microscopy, which permits selective visualization of fluorescently labelled vesicles located in close proximity to the PM. The penetration depth of our TIRF evanescence field was measured to be ~100 nm. The TIRF images of cells co-transfected with red fluorescent protein (RFP)-tagged hRAB37 and GFP-tagged TIMP1 illustrated the co-localized yellow fluorescent vesicles, which were defined as TIMP1-containing vesicles trafficked by hRAB37. These co-localized vesicles appeared quickly and diffused laterally from the point of insertion within the PM in the TIRF evanescence field (Fig. 3c and Supplementary Movie 1), while the fluorescence signals remained steady in cells co-transfected with empty vectors of RFP and GFP (Fig. 3d, Supplementary Fig. 5 and Supplementary Movie 2). To obtain a quantitative measurement of vesicle trafficking event, we tracked the position of the yellow vesicles and measured the ratio between the moving yellow vesicles and the total yellow vesicles in the cell. While vesicle trafficking events were low in vector control cells, they were markedly enhanced in cells co-transfected with RFP-tagged hRAB37 and GFP-tagged TIMP1 (Fig. 3e). To the best of our knowledge, this is the first report on the dynamics of hRAB37-mediated trafficking and docking of TIMP1-containing vesicles to the PM.

hRAB37 suppresses metastasis in a nucleotide-dependent manner

We previously showed that promoter hypermethylation and low mRNA expression of hRAB37 gene correlated with increased tumour metastasis potential in lung cancer patients¹⁴. Therefore, we hypothesized that hRAB37 may be a metastasis suppressor. To test this hypothesis, we performed migration and invasion assays on established stable cells of control, WT, Q89L (the GTP-bound active mutant) and T43N (the GDP-bound inactive mutant) derived from CL1-5. The CL1-5 high-invasive lung cancer cell line with low expression of endogenous hRAB37 provides a good model to study hRAB37-mediated metastasis suppression. The results showed that WT and Q89L cells reduced migration and invasion abilities compared with the control and dominant-negative T43N mutant cells (Fig. 4a,b). To confirm that this is not unique to CL1-5 cells, we also performed invasion assays with other human lung cancer cells including A549 lung adenocarcinoma cell line and H460 large-cell lung cancer cell line. Similarly, transient overexpression of WT-hRAB37 exerted an invasion suppressive function in multiple lung cell models (Supplementary Fig. 6).

Figure 4: hRAB37 suppresses lung cancer invasion and tumour metastasis in a nucleotide-dependent manner *in vitro* and *in vivo.*

We have demonstrated that hRAB37 is important in the control of TIMP1 exocytosis to CM, therefore, we proposed that hRAB37 functions in the context of TIMP1 exocytosis to regulate cancer motility. To verify this, we first treated highly metastatic CL1-5 cells with CM from cells expressing control, WT, Q89L and T43N, and then performed invasion assays. Inhibition of invasion was observed by treatment of CL1-5 cells with CM collected from WT and Q89L expressed cells but not T43N inactive hRAB37 cells (Fig. 4c). Notably, the concentrations of TIMP1 measured in CM were entirely consistent with the scenario that hRAB37 transported TIMP1 to inhibit cell migration and invasion (Fig. 4c). Importantly, control and T43N cells, which secreted TIMP1 at low level, reduced invasion abilities by treatment with TIMP1 recombinant protein (Fig. 4d). TIMP1 is known to inactivate MMP9 activity^19,20,26. Accordingly, gelatin-zymography assay revealed that MMP9 activity was inhibited in WT and Q89L cells compared with control and T43N cells (Fig. 4e). In addition, MMP9 activity inversely correlated with TIMP1 level in CM from the other hRAB37 reconstituted lung cancer cells (Supplementary Fig. 6). Migration signalling proteins such as p-FAK and GTP-RhoA were downregulated in WT and Q89L cells compared with the control (Supplementary Fig. 7a,b). These data suggested that the suppression of migration and invasion by hRAB37 is mediated via TIMP1 secretion to inactivate MMP9 activity in a nucleotide-dependent manner.

The strong inhibition of cell migration and invasion in cell model prompted us to determine whether hRAB37 acted as a tumour metastasis suppressor in animals. Control, WT, Q89L and T43N cells were tail-vein injected into BALB/c nude mice. Eight weeks post injection, the mice were killed, and their lungs were collected for analysis of tumour extravasation and colonization. We found that WT and Q89L markedly inhibited metastases to mouse lungs compared with control and T43N groups (Fig. 4f,g). Importantly, the concordant expression between hRAB37 and TIMP1 was confirmed in vivo except for the tumours derived from T43N inactive hRAB37 group that showed high hRAB37 expression and low TIMP1 staining (Fig. 4h).

KD of hRAB37 promotes metastasis in vitro and in vivo

Next, we undertook a KD approach to further verify that hRAB37 was a metastasis suppressor. Importantly, stable hRAB37KD H460 lung cancer cells showed increased migration and invasive potential compared with control (Fig. 5a,b). Note that hRAB37KD significantly reduced TIMP1 secretion in CM leading to MMP9 activation (Fig. 5c and Supplementary Fig. 8c). Similar results were obtained in reconstitution experiments after KD of ectopically expressed WT-hRAB37 in CL1-5 cells (WT-hRAB37KD; Supplementary Fig. 8a,b). Migration signalling proteins such as p-FAK and GTP-RhoA were upregulated in hRAB37KD cells compared with the control (Supplementary Fig. 7c).

Figure 5: KD of hRAB37 promotes lung cancer invasion and tumour metastasis *in vitro* and *in vivo.*

Lung-to-lung metastasis in animal model was performed using stable hRAB37KD H460 cells. Cancer cells were inoculated percutaneously into the left lateral thorax, and the tumour nodules at the right lungs of injected mice were determined by haematoxylin and eosin (H&E) staining. Strikingly, hRAB37KD significantly promoted lung-to-lung tumour metastases by sixfold compared with the control (Fig. 5d,e). Of note, concordantly low expression of hRAB37 and TIMP1 was confirmed in metastatic lung nodules in vivo (Fig. 5f). All these data corroborated with the hRAB37 overexpression results, indicating that hRAB37 is a bona fide metastasis suppressor.

TIMP1 is essential to hRAB37-mediated metastasis suppression

To further verify the role of TIMP1 in hRAB37-mediated metastasis suppression, migration and invasion assays using WT-hRAB37 (WT-Con), TIMP1KD in the WT-hRAB37 background (WT-TIMP1KD) and TIMP1 reconstitution by addition of TIMP1 recombinant protein (WT-TIMP1KD+TIMP1) were performed. WT-TIMP1KD showed an increase of cell migration and invasion, whereas WT-TIMP1KD+TIMP1 abolished the migration and invasion ability (Fig. 6a,b). Similar results were obtained with the TIMP1 reconstituted Q89L-hRAB37 cells (Supplementary Fig. 9). Treatment of T43N cell, which secreted TIMP1 at low level, with TIMP1 recombinant protein also reduced the migration and invasion ability (Supplementary Fig. 10a,b). In addition, MMP9 activity inversely correlated with TIMP1 level in CM from the reconstituted cells (Fig. 6c). These data suggested that TIMP1 secretion to inactivate the MMP9 activity attributes to metastasis suppression by hRAB37.

**Figure 6: hRAB37 inhibits lung cancer metastasis *in vitro* and *in vivo* via TIMP1.**

We further investigated the interplay between hRAB37 and TIMP1 on anti-metastasis effects in animals. WT-Con, WT-TIMP1KD and WT-TIMP1KD+TIMP1 cells were tail-vein injected into BALB/c nude mice, and the mice were observed for tumour extravasation and colonization to lungs. Notably, WT-TIMP1KD mice showed more tumour nodules than WT-Con mice, while such aggressive tumour metastasis phenotype was attenuated in WT-TIMP1KD+TIMP1 mice (Fig. 6d,e), suggesting that the metastasis inhibition of hRAB37-WT GTPase is mediated by TIMP1. In addition, TIMP1 treatment dramatically inhibited lung cancer metastasis of the animal group tail-vein injected with high metastatic T43N cells (Supplementary Fig. 10c,d). Together, these data indicated that TIMP1 secretion is crucial to hRAB37-mediated tumour metastasis suppression.

Clinical impacts of concordantly low hRAB37/TIMP1 expression

Since hRAB37 promotes TIMP1 secretion to inhibit migration and invasion in cell model and tumour metastases in animal model, we further investigated whether low hRAB37 protein expression correlated with low TIMP1 level in cancer patients with cancer metastasis and poor prognosis. For this, we examined the expression of hRAB37 and TIMP1 in lung tumour specimens from 165 lung cancer patients by immunohistochemical analysis (Fig. 7a). Importantly, a total of 47.3% (78/165) patients showed low hRAB37 protein expression (Table 1 and Supplementary Table 1). Of note, patients with low hRAB37 expression were significantly associated with advanced disease (stages III and IV; P=0.005, Pearson χ²-test) and with lymph node metastasis (P=0.013; Fig. 7b and Table 1). Importantly, low expression of hRAB37 correlated with worse prognosis in overall survival (P<0.001, log-rank test; Fig. 7c) or progression-free survival (P=0.016; Fig. 7d). These clinical results confirmed that hRAB37 is a metastasis suppressor with prognostic implication.

**Figure 7: Low hRAB37 protein coincides with low TIMP1 level and poor survival of lung cancer patients.**

Table 1 Alteration of hRAB37 and TIMP1 protein expression levels in relation to clinicopathological parameters in 165 lung cancer patients*.

Full size table

The relationship between hRAB37 and TIMP1 expression has never been examined in human cancer patients; we therefore performed double fluorescence and confocal immunohistochemistry on lung tumour specimens. Strikingly, confocal immunohistochemistry demonstrated that patients with low hRAB37 showed concordantly reduced TIMP1 (P<0.0001; Fig. 7e,f and Table 1; and Supplementary Fig. 11). Patients with low TIMP1 expression were significantly associated with advanced disease (stages III and IV) and with lymph node metastasis (Fig. 7g and Table 1). Indeed, low expression of TIMP1 correlated with poor prognosis in overall survival (Fig. 7h). Importantly, patients with low expression of hRAB37 and TIMP1, singly or together, showed significantly worse survival compared with those patients with expression of both proteins (Fig. 7i). Altogether, these clinical data demonstrated that low hRAB37 protein coincided with low TIMP1 level in tumour tissues, and reduced expression of both hRAB37 and TIMP1 in surgical specimens are prognostic biomarkers for lung cancer.

We conducted an analysis of The Cancer Genome Atlas (TCGA) data set for lung squamous cell carcinoma and lung adenocarcinoma^27,28 looking at the mRNA expression levels of hRAB37 and TIMP1 (Fig. 8a and Supplementary Data 2). There was a significant concordance of hRAB37 and TIMP1 expression level in the data set of 450 lung cancer patients (P<0.001; Fig. 8b and Supplementary Data 3). Importantly, patients with lower mRNA level of hRAB37 correlated with advanced tumour stages and had a worse survival (Fig. 8c,d and Supplementary Data 3) than patients with higher level of hRAB37, indicating that the relevance of our finding was also seen in a large cohort of patients as part of TCGA project.

**Figure 8: Low *hRAB37* coincides with low *TIMP1* level, leading to poor survival of lung cancer patients in the TCGA data set.**

Discussion

Here we show that TIMP1 is a novel cargo of hRAB37. We provide compelling evidence from lung cells, animal and clinical studies that hRAB37 is a metastasis suppressor via regulating TIMP1 for exocytosis leading to inactivation of MMP9 signalling with clinical implication of prognosis prediction (Fig. 8e,f). Together, our findings that hRAB37 acts as a metastasis suppressor by regulating the exocytotic transport of TIMP1 in a nucleotide-dependent manner to inhibit cancer cell migration and invasion reveal the involvement of a novel component of the vesicular exocytotic machinery in tumour progression.

Low or inactive hRAB37 expression promoted metastasis of lung cancer but altered hRAB37 expression had no effect on lung tumour xenograft growth (Supplementary Fig. 12). In the lung-to-lung metastasis animal model, the primary growth judging by the estimated overall tumour sizes of the implanted tumours within the left lateral pleural cavity appeared to be similar among sh-control, sh1 and sh2 groups. Notably, the softer tumours in sh1 and sh2 groups were substantially more dispersed and easily detachable with irregular surfaces compared with the tumours in the control group in which the solid mass was better encapsulated. In line with these observations, the tumours in sh-RAB37 groups showed more parietal and visceral pleura invasion and even invaded the lung parenchyma, consequently leading to higher metastases in the contralateral lungs. Our study demonstrates that hRAB37 is a bona fide metastasis suppressor Rab. According to the multilevel information provided in TCGA, homozygous deletion of hRAB37 has been found in patients with breast invasive carcinoma and ovarian serous cystadenocarcinoma, as well as cell lines from intestine and ovary cancers. Note that G52W and R91W missense mutations, and R93* nonsense mutation were reported for TCGA lung adenocarcinoma as well as uterine corpus endometrioid carcinoma tumours, respectively. In addition, we showed that hRAB37 mRNA was downregulated in some of the liver and colon tumour samples (Supplementary Fig. 13). These findings suggested that hRAB37 may also act a suppressor in liver, colon and other cancers, which warrants further investigations.

Most of the previously reported Rab proteins, such as Rab8, Rab11, Rab21, Rab23, Rab25 and Rab27B (refs 3, 4, 5, 6, 7, 8, 9), function as metastasis promoters, although Rab25 acts as a tumour growth suppressor in intestinal epithelial cells from the Apc^Min/−;Rab25^−/− mice background¹¹. These results indicated that Rab25 is an oncogene-like protein that promotes tumour growth in ovarian and breast cancer models^4,5, but acts as a tumour growth suppressor in intestinal epithelial cells, which raises the question of why the same Rab protein functions differently in different cancers. Such discrepancy was also found for hRAB37 in lung cancer and renal cell carcinoma where KD of hRAB37 reduced cell growth of renal cell carcinoma cell lines²⁹. Vesicle isolation assays indicated that TIMP1 was not present in the hRAB37-containing vesicles of renal cell carcinoma cells in which hRAB37 functions as an oncogene (Supplementary Fig. 14). The context-dependent role of Rabs to serve either as an oncoprotein or tumour suppressor may be owing in part to the interaction with different effectors in various tumours³⁰.

Our proteomics screens and Gene Ontology (GO) analysis revealed that hRAB37 trafficking components were involved in multiple cell biological functions such as cell adhesion, migration, ECM remodelling, actin-mediated cell contraction and cell morphogenesis processes, further supporting the roles of hRAB37 in metastatic control (Supplementary Fig. 1a). Moreover, disease biomarker analysis showed that alteration of these putative secreted cargos are mainly involved in tumour progression, including lung cancer (Supplementary Fig. 1b). The range of trafficking cargos or effector proteins of hRAB37 involved in the control of metastasis will be further analysed. In addition, whether the function of other secretory Rabs such as Rab3, Rab26 and Rab27A/B are altered in hRAB37-inhibited cells is worthy of further studies. Our immuno-EM observation that dominant-negative T43N-hRAB37 lost membrane binding to the TIMP1-containing vesicles extends the notion that hRAB37 facilitates vesicle docking and fusion but does not affect TIMP1 vesicle formation. Notably, T43N-hRAB37 displayed a perturbed distribution profile with increased perinuclear localization. It raises intriguing questions of how other proteins such as effector and adaptor proteins or other Rabs contribute to regulate TIMP1 sequestration and vesicle formation. Munc13-1 has been shown to function as an effector of mRab37-mediated TNF-α secretion in activated macrophages¹⁶. Investigations of effectors of hRAB37-mediated trafficking and migration inhibition under stimulated conditions will shed light on the physiological role of hRAB37 (ref. 31).

Although TIMP1 shows paradoxical effects on tumour growth and metastasis^32,33, and very little is known about its mode of secretion, we demonstrate that hRAB37 directs TIMP1 secretion in a nucleotide-dependent manner to inhibit tumour metastasis attributing to its MMP inhibitor activity. In agreement, TIMP1 inhibits the invasion and metastasis of malignant tumour cells mainly owing to its role as an inhibitor of MMPs^34,35. Of note, our immunohistochemistry in patient specimen and animal xenograft demonstrated that reduced TIMP1 often occurred in tissue specimens with low hRAB37. Reduced TIMP1 expression may result from additional alterations such as promoter hypermethylation of TIMP1 in resected lung tumour³⁶. In stratification analyses, low expression of hRAB37 and TIMP1, singly or together, correlated with poor overall survival. From a clinical point of view, it is therefore of particluar interest that low expression of TIMP1 in resected lung tumour tissue could be also an effective prognosis marker. However, other reports indicate that high TIMP1 levels in blood are associated with poor prognosis in some cancers^37,38. A murine model of experimental liver metastasis suggests that the metastasis-mediated by TIMP1 may depend on the specific microenvironmental context³⁹. Similarly, stromal caveolin-1, which is transported from the Golgi complex by the exocytotic pathway, also modulates the microenvironment during tumour invasion⁴⁰. The roles of TIMP1 in local and distant extracellular environments are worthy of further interrogations.

Our previous DNA methylation and deletion studies have suggested that hRAB37 may be a tumour suppressor, but the mechanism of its anti-metastatic property is not known. The present results identify a novel link that hRAB37 is a metastasis suppressor via regulating TIMP1 for exocytosis to inactive MMP9 invasion signalling. Therapeutic strategies such as DNA demethylation of hRAB37 gene (ref. 14) and increased stability of hRAB37 protein⁴¹ could facilitate the development of cancer therapy.

Methods

Cell culture and transfection reagents

Human lung cancer cell lines CL1-0 and CL1-5 were derived from Taiwanese patients with lung adenocarcinoma⁴². A549, H460 and monkey kidney cells COS1 were purchased from the American Type Culture Collection. All cells were maintained in DMEM (Gibco), with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Gibco), and cultured at 37 °C with 5% CO₂ in air. To generate stable WT, Q89L and T43N cells, cDNAs of hRAB37 were cloned into the pLP’s-Flag vector (Clontech) and transfected into CL1-5 cells by electroporation (Digital Bio Technology, Seoul, Korea). G418 (1,000 μg ml⁻¹, Invitrogen) was used to select the stable cells. To KD hRAB37 (L-008933-01-0020) and TIMP1 (L-011792-00-0005), siRNA oligos (Thermo Fisher Scientific Dharmacon RNAi Technologies) were transfected into COS1 or CL1-5 cells at a final concentration of 100 nmol using Lipofectamine 2000 (Invitrogen). To stably KD hRAB37 in the H460 cell line, cells were infected with control and sh-hRAB37 lentiviral particles produced in packaging vector pLKO_TRC005. The viruses were obtained from the RNAi core (Academia Sinica, Taiwan). The KD sequences for hRAB37 sh1; 5′- CCGGAGCTTCCAGATCCGAGACTATCTCGAGATAGTCTCGGATCTGGAAGCTTTTTTG -3′, and for hRAB37 sh2; 5′- CCGGAGCGTCACCCATGCTTATTACCTCGAGGTAATAAGCATGGGTGACGCTTTTTTTG -3′ were used. Cells were selected with puromycin (2.5 μg ml⁻¹) for 1 week after 24 h of infection with control or sh-hRAB37 lentiviral particles with polybrene (4 μg ml⁻¹, Sigma-Aldrich). hRAB37 KD efficiency was checked by western blotting.

Generation of hRAB37 antibody

hRAB37-specific peptide QADEP SFQIR DYVES QKKRS SC (0.1 mg in 200 μl) was injected subcutaneously (s.c.) into the flanks of mice once every 2 weeks. Mice were killed to collect serum containing antibody after 3 months post injection. Western blots were performed to detect the specificity of hRAB37 antibody (Supplementary Fig. 15). The antibody was generated by LTK BioLaboratories, Taiwan.

VSVG transport and Tfn recycling assays

Cos1 cells were co-transfected with VSVG-ts045-GFP (kindly provided by Dr Jennifer Lippincott-Schwartz at Cell Biology and Metabolism Program, National Institutes of Health, USA), control or hRAB37KD siRNA, and mcherry (to mark the transfected cells) at a ratio of (2.5:5: 1; wt/wt) and were incubated at 40 °C for 6 h post transfection. After 20 h of incubation, cycloheximide (150 ng ml⁻¹) was added to the culture medium and the cells were incubated at 32 °C to allow VSVG transport. Cells were fixed with 4% formaldehyde and stained with 4',6-diamidino-2-phenylindole at various time points. VSVG arrival at the PM was analysed by epifluoresrescence microscopy and the percentage of cells displaying VSVG-tsO45-GFP at the PM was scored in 150 transfected cells in each experiment.

For quantifying the release of internalized Tfn, COS1 cells were transfected with control or hRAB37KD siRNA. At 48 h post transfection, cells were serum-starved for 3 h and incubated with AlexaFluor488-conjugated Tfn (Molecular Probes) at 37 °C for 60 min. After three washes with ice-cold PBS containing 0.1% bovine serum albumin (BSA), cells were stripped of surface-bound Tfn with ice-cold acid buffer (150 mM NaCl and 100 mM glycine, pH 3.0). Cells were incubated in medium containing unlabelled holo-Tfn at 37 °C for various time points, washed with PBS and lysed using lysis buffer (0.5% Triton X-100 in PBS). AlexaFluor488-conjugated Tfn was quantified by enzyme-linked immunosorbent assay (ELISA) and normalized with the total protein concentration in cell lysates at each time point compared with time point zero.

CM preparation

Serum-free CM was prepared from cells (1 × 10⁷) cultured with 20 ml serum-free medium for 24 h. The cell viability was ascertained using the Trypan blue dye exclusion assay and was >98%. The media were collected following the protocol of Amicon Ultra centrifugal filter units (UFC900396, Millipore) and the Bradford assay (Bio-Rad Laboratories) was used to determine the total protein concentration in the CM samples and cell motility analysis.

Proteomics analysis and GO analysis

A total of 100 μg of CM was mixed with 4 × SDS sample buffer and 0.5 M DTT and then boiled for 10 min at 95 °C. Samples were run on NuPAGE 4–20% Bis-Tris gradient gels (Invitrogen) in denaturating SDS buffer, then stained using 0.5% Coomassie Brilliant Blue (Bio-Rad Laboratories). The gel pieces corresponding to the differential proteins (control versus WT) were diced into ~1 mm³ pieces. For in-gel digestion, peptides were extracted twice with 100 μl of 50% acetonitrile in 5% formic acid. The extracted peptides were enriched using OMIX C18 pipet tips (Varian) to remove the effects of any remaining reagents on the iTRAQ labelling⁴³. The enriched peptides were then labelled with the isobaric tags for relative and absolute quantification using iTRAQ reagent (Applied Biosystems). Before labelling, iTRAQ-labelled samples were reconstituted and analysed by liquid chromatography-MS/MS. Peptide fragmentation by collision-induced dissociation was performed automatically using the information-dependent acquisition in Analyst QS v1.1 (Applied Biosystems). For identification of proteins, raw MS/MS files were batch-searched against the Swiss-Prot human sequence database (version 20090616; 468851 sequences) using the MASCOT algorithm (v2.1.0, Matrix Science, London, UK). For quantification of proteins, data analysis for the iTRAQ (Isobaric Tags for Relative and Absolute Quantization) experiments was performed with the software Multi-Q⁴⁴. The identified proteins were analysed using the SignalP⁴⁵, SecretomeP⁴⁶ and ExoCarta⁴⁷ programs to predict the possibility of protein secretion through classic or nonclassic secretion pathways and the presence of transmembrane domains in the protein sequence. The molecular functions of the identified proteins were determined based on a search against the Human Protein Reference Database (http://www.hprd.org/). For pathway analysis, proteins were uploaded and analysed for evidence of an interaction network using MetaCore from GeneGo Inc. (http://www.genego.com/metacore.php). The GO analysis was performed by home-made script. In this script, GO terms were assessed by Fisher’s exact test with Benjamini–Hochberg procedure. The enriched GO terms were visualized using Enrichment Map⁴⁸, a Cytoscape plugin.

Vesicle isolation and immunoprecipitation

Vesicle isolation procedures are modified from Hendrix’s report⁹. In brief, cells (2 × 10⁸) were sonicated on ice, vesicles were then isolated by centrifugation at low speed (3,000 g for 10 min at 4 °C) and high speed (30,000 g for 60 min at 4 °C) using 40-Ti rotor (Beckman). Supernatant containing vesicles (500 μg) was incubated with anti-Flag antibody to isolate hRAB37-specific vesicles (Supplementary Table 2).

Confocal and EM imaging

Transfected cells (1 × 10⁵) were fixed and then incubated with primary and secondary fluorescent antibodies (Supplementary Table 2), and analysed by Olympus FV1000 confocal microscope. For immuno-EM analysis, cells were fixed, sectioned and immersed in H₂O₂ for 10 min, blocked with 1% BSA for 1 h and then incubated with indicated antibodies and protein A coupled to 10 or 15 nm gold particles (Supplementary Table 2). Sections were post stained with uranyl acetate and lead citrate, and then examined by Hitachi electron microscope (H7650). To measure the co-localization of hRAB37 and TIMP1 in the same vesicles, we obtained >60 immuno-EM images each from control, WT, Q89L and T43N groups. The representative images are shown in Fig. 2c. We then quantified co-localization of both TIMP1 and hRAB37 in the same vesicle of 100 TIMP1-containing vesicles each in control, WT, Q89L and T43N cells. These quantitative procedures were done by two independent immuno-EM experiments, which are shown in Fig. 2d.

TIRF imaging

CL1-5 cells were seeded on 29-mm glass bottom dish. After 24 h, cells were transiently co-transfected with RFP-hRAB37/GFP-TIMP1 or RFP-vector/GFP-vector for 24 h before image analysis. TIRF system was built on an inverted microscope (Olympus IX81, Japan). The system was equipped with a high-sensitivity EMCCD camera (iXOn3897, Andor Technology) and a UPON × 100 OTIRF objective lens (numerical aperture=1.49; Olympus) to achieve TIRF. GFP and RFP were excited with 491 and 561 nm solid lasers, respectively (100 mW, 10–30% power) and driven by Xcellence software (Olympus imaging software). To capture the hRAB37 trafficking events, images sequences were recorded in stream mode and the frame rate was 30 frames for 1 min. For quantification of vesicle trafficking event, we defined each yellow fluorescence spot as a TIMP1-containing vesicle and trafficking by hRAB37 in cells, and then tracked each yellow vesicle trafficking distance with trackIT software (Olympus). The cutoff length of moving vesicle was 3 μm. The vesicle trafficking event was measured as the ratio of moving yellow vesicles to the total yellow vesicles in cells. For each experimental condition, a total of at least 20 co-localized yellow vesicles per cell were tracked and six cells were scored.

Immunoblotting

Cells were lysed in 1 × RIPA lysis buffer with a cocktail of antiproteases (Sigma-Aldrich), equal amounts of total cellular extracts were electrophoresed by SDS–PAGE and transferred on to a polyvinylidene difluoride membrane. Primary antibodies used for immunoblotting are list in Supplementary Table 2. Full-length images of immunoblots are shown in Supplementary Fig. 16.

Enzyme-linked immunosorbent assay

The TIMP1 (Invitrogen) and TNF-α (R&D Systems) ELISA kits were used to detect TIMP1 and TNF-α concentration in CM according to the manufacturer’s protocol. The standards and CM samples were added into 96-well plate that contains antibody. The biotin conjugate solution was added into the wells and incubated at room temperature for 2 h. After incubation, wells were washed with wash buffer and incubated with streptavidin-horseradish peroxidase working solution. Stabilized chromogen and stop solution were then added. The absorbance is measured at 450 nm using an ELISA plate reader.

Gelatin-zymography assay

CM were collected and analysed by gelatin-zymography in 0.1% gelatin-8% acrylamide gels. After electrophoresis, gel was washed with 2.5% Triton X-100 to remove SDS and renature MMP9. Gel was then incubated overnight in the developing buffer to induce gelatin lysis by renatured MMP9 and stained with 0.5% Coomassie blue G for 1 h. The proteolytic activities were identified as clear bands.

Migration and invasion assays

For transwell migration assay, cells (1 × 10⁵) were placed in the upper chamber of transwell (Falcon) with serum-free medium. For transwell invasion assay, the transwell membranes were pre-coated with Matrigel (2.5 mg ml⁻¹, Sigma-Aldrich). Medium containing 10% fetal bovine serum was added to the lower chamber as chemoattractants and cells were incubated at 37 °C for 16 h. Cells attached to the reverse side of the membrane were stained with crystal violet, and 10 randomly selected fields were counted under inverted microscope (Nikon TS100) for each experiment.

Experimental metastasis assays in vivo

The use and care of the BALB/c nude mice were approved by the Institutional Animal Care and Use Committee of National Cheng Kung University. Three groups of animals were used; the first group included control, WT, Q89L and T43N stable expression CL1-5 cells. In the second group, WT-Con, WT-TIMP1KD and WT-TIMP1KD+TIMP1 CL1-5 cells were used. In the third group, T43N stable expression CL1-5 cells treated with (+TIMP1) or without (T43N) recombinant TIMP1 protein (410-01, Peprotech) for 24 h. For these three groups, cells were resuspended in serum-free medium (2 × 10⁶ cells per 200 μl) and intravenously injected via the tail vein into BALB/c nude mice. Mice were killed after 6–8 weeks and lung tissues were resected, fixed and embedded in paraffin for histological H&E and immunohistochemistry staining.

Lung-to-lung metastasis assay in vivo

Six-week-old male BALB/c nude mice were purchased from BioLASCO (Taipei, Taiwan) and acclimated for 2 weeks in the animal facilities at the National Health Research Institutes (NHRI), Taiwan. All animal treatments and experimental protocols for this study were reviewed and approved by the Institutional Animal Care and Use Committee at NHRI. Mice were divided into eight mice per group. Stable hRAB37 KD (sh1 and sh2) and sh-control H460 cells were used. A total of 5 × 10⁴ cells were suspended in 100 μl Matrigel (Becton Dickinson) in PBS. Mice anaesthetized with isoflurane were placed in the right lateral decubitus position. Tuberculin syringes (1 ml) with 26-gauge hypodermic needles were used to inject the cell inoculum percutaneously into the left lateral thorax, at the lateral dorsal axillary line, 1.5 cm above the lower rib line just below the inferior border of the scapula. The needle was quickly advanced 5–7 mm into the thorax and was quickly removed after the injection of cell suspension. After tumour injection, the mouse was turned to the left lateral decubitus position. Animals were observed for 45–60 min until fully recovered. Four weeks after cell inoculum, animals were killed. The solitary lesions located in the injected left lateral lung and within the pleural thoraxes were removed. The contralateral lungs were isolated and prepared for pathological examination. Lung tissues were fixed with 10% neutral buffered formalin for 48 h before tissue processing, which included dehydration, and embedding in paraffin. H&E stain was used for general histopathological examinations. The outward appearance of the lungs and the solitary nodules were recorded.

Tumour growth assay in vivo

Six-week-old BALB/c nude mice, approved by the Institutional Animal Care and Use Committee of National Cheng Kung University, were s.c. inoculated with vector control and hRAB37-WT stable expression CL1-5 cells (1 × 10⁷ cells per 100 μl) and allowed to grow for 21 days. The animals were maintained under sterile conditions in laminar flow rooms and provided with sterilized food and water. After 3 weeks, the mice were killed and the tumour xenografts were removed, photographed and weighed.

Patient samples and clinical information

A total of 165 lung cancer patients were recruited from Taipei Veterans General Hospital between 1999 and 2010 after obtaining appropriate institutional review board permission and written informed consent from patients. Paraffin blocks of tumours were collected. The detailed clinicopathological characteristics of the enrolled patients are listed in Supplementary Table 1.

Immunohistochemistry and fluorescence immunohistochemistry

Immunohistochemistry was performed to detect protein expression of hRAB37 and TIMP1 in tumour sections from lung cancer patients. Staining for expression region was scored as follows: absent (0–25%), decrease (25–50%) and preserve (>50%), while absent and decrease was defined as protein low expression. The surrounding normal tissue served as an internal positive control for each slide. For fluorescence immunohistochemistry and confocal microscopy of hRAB37 and TIMP1 localization, primary antibody followed by incubation with the fluorescent secondary antibodies and 4',6-diamidino-2-phenylindole (Supplementary Table 2) for 1 h, after which images were obtained with confocal microscope.

mRNA expression analysis

For hRAB37 mRNA expression assay, a semi-quantitative reverse transcription PCR assay using the GAPDH gene as an internal control to indicate the amount of RNA in each reaction was performed. The primer nucleotide sequences are as follows: for hRAB37 gene, sense, 5′- ACGGGCAAGGTGATGCTTCTGG -3′, antisense, 5′- CATCTCTGTAATAAGCATGGGTGACGCT -3′; for the GAPDH gene, sense, 5′- AATCCCATCACCATCTTCCA -3′, antisense, 5′- CCTGCTTCACCACCTTCTTG -3′. PCR was performed for 34 cycles with annealing temperature of 68 °C using cDNA synthesized from total RNA.

TCGA analysis

The analysis of the TCGA data set for lung cancer was conducted through the cBio Cancer Genomics Portal²⁸ based on the 24 October, 2012 Broad Firehose run of TCGA data. Expression and survival data of patients are presented in Supplementary Data 2 and 3. The data were normalized and log2 was transformed.

Statistical analysis

Pearson χ²-test was used to compare the altered frequency of hRAB37 and TIMP1 between lung cancer patients with different clinicopathological parameters. Overall survival and disease-free survival curves were calculated according to the Kaplan–Meier method, and comparison was performed using the log-rank test. Two-tailed Student’s t-test was used in cell and animal studies. Three independent experiments for cell studies and eight mice per group for animal studies were analysed unless indicated otherwise. Data represent mean+s.e.m. P<0.05 was considered to be statistically significant.

Additional information

How to cite this article: Tsai, C.-H. et al. Small GTPase Rab37 targets tissue inhibitor of metalloproteinase 1 for exocytosis and thus suppresses tumour metastasis. Nat. Commun. 5:4804 doi: 10.1038/ncomms5804 (2014).

References

Stenmark, H. Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol. 10, 513–525 (2009).
Article CAS PubMed Google Scholar
Santiago-Tirado, F. H. & Bretscher, A. Membrane-trafficking sorting hubs: cooperation between PI4P and small GTPases at the trans-Golgi network. Trends Cell Biol. 21, 515–525 (2011).
Article CAS PubMed PubMed Central Google Scholar
Bravo-Cordero, J. J. et al. MT1-MMP proinvasive activity is regulated by a novel Rab8-dependent exocytic pathway. EMBO J. 26, 1499–1510 (2007).
Article CAS PubMed PubMed Central Google Scholar
Caswell, P. T. et al. Rab25 associates with alpha5beta1 integrin to promote invasive migration in 3D microenvironments. Dev. Cell 13, 496–510 (2007).
Article CAS PubMed Google Scholar
Cheng, K. W. et al. The RAB25 small GTPase determines aggressiveness of ovarian and breast cancers. Nat. Med. 10, 1251–1256 (2004).
Article CAS PubMed Google Scholar
Hendrix, A. et al. Effect of the secretory small GTPase Rab27B on breast cancer growth, invasion, and metastasis. J. Natl. Cancer Inst. 102, 866–880 (2010).
Article CAS PubMed PubMed Central Google Scholar
Hou, Q. et al. Integrative genomics identifies RAB23 as an invasion mediator gene in diffuse-type gastric cancer. Cancer Res. 68, 4623–4630 (2008).
Article CAS PubMed Google Scholar
Pellinen, T. et al. Small GTPase Rab21 regulates cell adhesion and controls endosomal traffic of beta1-integrins. J. Cell Biol. 173, 767–780 (2006).
Article CAS PubMed PubMed Central Google Scholar
Yoon, S. O., Shin, S. & Mercurio, A. M. Hypoxia stimulates carcinoma invasion by stabilizing microtubules and promoting the Rab11 trafficking of the alpha6beta4 integrin. Cancer Res. 65, 2761–2769 (2005).
Article CAS PubMed Google Scholar
Liu, Y. et al. Knockdown of RAB25 promotes autophagy and inhibits cell growth in ovarian cancer cells. Mol. Med. Rep. 6, 1006–1012 (2012).
Article CAS PubMed Google Scholar
Nam, K. T. et al. Loss of Rab25 promotes the development of intestinal neoplasia in mice and is associated with human colorectal adenocarcinomas. J. Clin. Invest. 120, 840–849 (2010).
Article CAS PubMed PubMed Central Google Scholar
Tong, M. et al. Rab25 is a tumor suppressor gene with antiangiogenic and anti-invasive activities in esophageal squamous cell carcinoma. Cancer Res. 72, 6024–6035 (2012).
Article CAS PubMed Google Scholar
Hognas, G. et al. Cytokinesis failure due to derailed integrin traffic induces aneuploidy and oncogenic transformation in vitro and in vivo. Oncogene 31, 3597–3606 (2012).
Article CAS PubMed Google Scholar
Wu, C. Y., Tseng, R. C., Hsu, H. S., Wang, Y. C. & Hsu, M. T. Frequent down-regulation of hRAB37 in metastatic tumor by genetic and epigenetic mechanisms in lung cancer. Lung Cancer 63, 360–367 (2009).
Article PubMed Google Scholar
Masuda, E. S. et al. Rab37 is a novel mast cell specific GTPase localized to secretory granules. FEBS Lett. 470, 61–64 (2000).
Article CAS PubMed Google Scholar
Mori, R. et al. Release of TNF-alpha from macrophages is mediated by small GTPase Rab37. Eur. J. Immunol. 41, 3230–3239 (2011).
Article CAS PubMed Google Scholar
Brunner, Y. et al. Proteomics analysis of insulin secretory granules. Mol. Cell Proteom. 6, 1007–1017 (2007).
Article CAS Google Scholar
Ljubicic, S. et al. The GTPase Rab37 participates in the control of insulin exocytosis. PLoS ONE 8, e68255 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Coussens, L. M., Fingleton, B. & Matrisian, L. M. Matrix metalloproteinase inhibitors and cancer: trials and tribulations. Science 295, 2387–2392 (2002).
Article ADS CAS PubMed Google Scholar
Hiratsuka, S. et al. MMP9 induction by vascular endothelial growth factor receptor-1 is involved in lung-specific metastasis. Cancer Cell 2, 289–300 (2002).
Article CAS PubMed Google Scholar
Sato, T. et al. Inhibition of activator protein-1 binding activity and phosphatidylinositol 3-kinase pathway by nobiletin, a polymethoxy flavonoid, results in augmentation of tissue inhibitor of metalloproteinases-1 production and suppression of production of matrix metalloproteinases-1 and -9 in human fibrosarcoma HT-1080 cells. Cancer Res. 62, 1025–1029 (2002).
CAS PubMed Google Scholar
Naslavsky, N., McKenzie, J., Altan-Bonnet, N., Sheff, D. & Caplan, S. EHD3 regulates early-endosome-to-Golgi transport and preserves Golgi morphology. J. Cell Sci. 122, 389–400 (2009).
Article CAS PubMed PubMed Central Google Scholar
Presley, J. F. et al. ER-to-Golgi transport visualized in living cells. Nature 389, 81–85 (1997).
Article ADS CAS PubMed Google Scholar
Sakane, H. et al. α-Taxilin interacts with sorting nexin 4 and participates in the recycling pathway of transferrin receptor. PLoS ONE 9, e93509 (2014).
Article ADS PubMed PubMed Central Google Scholar
Shin, H., Morinaga, N., Noda, M. & Nakayama, K. BIG2, a guanine nucleotide exchange factor for ADP-ribosylation factors: its localization to recycling endosomes and implication in the endosome integrity. Mol. Biol. Cell 15, 5283–5294 (2004).
Article CAS PubMed PubMed Central Google Scholar
Xu, D. et al. Matrix metalloproteinase-9 regulates tumor cell invasion through cleavage of protease nexin-1. Cancer Res. 70, 6988–6998 (2010).
Article CAS PubMed PubMed Central Google Scholar
Hammerman, P. S. et al. Comprehensive genomic characterization of squamous cell lung cancers. Cancer Genome Atlas Research Network. Nature 489, 519–525 (2012).
Article ADS CAS Google Scholar
Cerami, E. et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2, 401–404 (2012).
Article PubMed Google Scholar
Dobashi, S. et al. Involvement of TMEM22 overexpression in the growth of renal cell carcinoma cells. Oncol. Rep. 21, 305–312 (2009).
CAS PubMed Google Scholar
Mitra, S., Cheng, K. W. & Mills, G. B. Rab25 in cancer: a brief update. Biochem. Soc. Trans. 40, 1404–1408 (2012).
Article CAS PubMed PubMed Central Google Scholar
Christoforides, C., Rainero, E., Brown, K. K., Norman, J. C. & Toker, A. PKD controls alphavbeta3 integrin recycling and tumor cell invasive migration through its substrate Rabaptin-5. Dev. Cell 23, 560–572 (2012).
Article CAS PubMed PubMed Central Google Scholar
Jiang, Y., Goldberg, I. D. & Shi, Y. E. Complex roles of tissue inhibitors of metalloproteinases in cancer. Oncogene 21, 2245–2252 (2002).
Article CAS PubMed Google Scholar
Xia, Y. et al. Reduced cell proliferation by IKK2 depletion in a mouse lung-cancer model. Nat. Cell Biol. 14, 257–265 (2012).
Article CAS PubMed PubMed Central Google Scholar
Chen, Y. et al. Runx3 suppresses gastric cancer metastasis through inactivation of MMP9 by upregulation of TIMP-1. Int. J. Cancer 129, 1586–1598 (2011).
Article CAS PubMed Google Scholar
Lopez-Otin, C. & Matrisian, L. M. Emerging roles of proteases in tumour suppression. Nat. Rev. Cancer 7, 800–808 (2007).
Article CAS PubMed Google Scholar
Wistuba, I. I., Gazdar, A. F. & Minna, J. D. Molecular genetics of small cell lung carcinoma. Semin. Oncol. 28, 3–13 (2001).
Article CAS PubMed Google Scholar
Holten-Andersen, M. N. et al. High preoperative plasma tissue inhibitor of metalloproteinase-1 levels are associated with short survival of patients with colorectal cancer. Clin. Cancer Res. 6, 4292–4299 (2000).
CAS PubMed Google Scholar
Lipton, A. et al. Serum TIMP-1 and response to the aromatase inhibitor letrozole versus tamoxifen in metastatic breast cancer. J. Clin. Oncol. 26, 2653–2658 (2008).
Article CAS PubMed Google Scholar
Schelter, F. et al. Tumor cell-derived Timp-1 is necessary for maintaining metastasis-promoting Met-signaling via inhibition of Adam-10. Clin. Exp. Metastasis 28, 793–802 (2011).
Article CAS PubMed Google Scholar
Goetz, J. G. et al. Biomechanical remodeling of the microenvironment by stromal caveolin-1 favors tumor invasion and metastasis. Cell 146, 148–163 (2011).
Article CAS PubMed PubMed Central Google Scholar
Sundberg, T. B. et al. Disruption of Wnt planar cell polarity signaling by aberrant accumulation of the MetAP-2 substrate Rab37. Chem. Biol. 18, 1300–1311 (2011).
Article CAS PubMed PubMed Central Google Scholar
Chu, Y. W. et al. Selection of invasive and metastatic subpopulations from a human lung adenocarcinoma cell line. Am. J. Respir. Cell Mol. Biol. 17, 353–360 (1997).
Article CAS PubMed Google Scholar
Chiu, K. H., Chang, Y. H., Wu, Y. S., Lee, S. H. & Liao, P. C. Quantitative secretome analysis reveals that COL6A1 is a metastasis-associated protein using stacking gel-aided purification combined with iTRAQ labeling. J. Proteome Res. 10, 1110–1125 (2011).
Article CAS PubMed Google Scholar
Lin, W. T. et al. Multi-Q: a fully automated tool for multiplexed protein quantitation. J. Proteome Res. 5, 2328–2338 (2006).
Article CAS PubMed Google Scholar
Bendtsen, J. D., Nielsen, H., von Heijne, G. & Brunak, S. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340, 783–795 (2004).
Article PubMed Google Scholar
Bendtsen, J. D., Jensen, L. J., Blom, N., Von Heijne, G. & Brunak, S. Feature-based prediction of non-classical and leaderless protein secretion. Protein Eng. Des. Sel. 17, 349–356 (2004).
Article CAS PubMed Google Scholar
Mathivanan, S. & Simpson, R. J. ExoCarta: A compendium of exosomal proteins and RNA. Proteomics 9, 4997–5000 (2009).
Article CAS PubMed Google Scholar
Merico, D., Isserlin, R., Stueker, O., Emili, A. & Bader, G. D. Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS ONE 5, e13984 (2010).
Article ADS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank the transmission electron microscope core, Department of Pathology, National Cheng Kung University Hospital for providing technical support and Taiwan Bioinformatics Core at the National Cheng Kung University, supported by a National Science Council for assisting with MetaCore analyses. We acknowledge Dr Mien-Chie Hung and Dr Yi-Hsin Hsu, Department of Molecular and Cellular Oncology at MD Anderson Cancer Center, Dr Nan-Shan Chang, Institute of Molecular Medicine at National Cheng Kung University and Dr Chia-Jung Yu, Department of Biochemistry and Molecular Biology at Chang-Gung University for discussion and comments on the manuscript, Dr Hsueh-Fen Juan, Department of Molecular and Cellular Biology at National Taiwan University for GO analysis of secretomic proteins and Mr Tseng-Wei Wang and Mr Ming-Hsien Tsai at NHRI for their technical support. This work was supported by Taiwan Ministry of Science and Technology (103-2627-B-006-007), Taiwan National Health Research Institutes (EX103-10202BI) and Taiwan Ministry of Health and Welfare (103-TDU-PB-211-133005).

Author information

Authors and Affiliations

Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, No. 1, University Road, Tainan, 701, Taiwan
Chung-Han Tsai, Hung-Chi Cheng, Yu-Shiuan Wang, Jayu Jen, I-Ying Kuo & Yi-Ching Wang
Department of Biochemistry and Molecular Biology, College of Medicine, National Cheng Kung University, No. 1, University Road, Tainan, 701, Taiwan
Hung-Chi Cheng
Division of Environmental Health and Occupational Medicine, National Health Research Institutes, No. 35, Keyan Road, Zhunan, 350, Taiwan
Pinpin Lin
Department of Environmental and Occupational Health, College of Medicine, National Cheng Kung University, No. 1, University Road, Tainan, 701, Taiwan
Ying-Hua Chang & Pao-Chi Liao
Institute of Biological Chemistry, Academia Sinica, No. 128, Section 2, Academia Road, Taipei, 115, Taiwan
Ruey-Hwa Chen & Wei-Chien Yuan
Institute of Biochemical Sciences, College of Life Science, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei, 100, Taiwan
Ruey-Hwa Chen & Wei-Chien Yuan
Division of Thoracic Surgery, Taipei Veterans General Hospital, Institute of Emergency and Critical Care Medicine, National Yang-Ming University, No. 155, Section 2, Linong Street, Taipei, 112, Taiwan
Han-Shui Hsu
Institute of Clinical Medicine, College of Medicine, National Yang-Ming University, No. 155, Section 2, Linong Street, Taipei, 112, Taiwan
Muh-Hwa Yang
Institute of Biochemistry and Molecular Biology and Genome Center, College of Medicine, National Yang-Ming University, No. 155, Section 2, Linong Street, Taipei, 112, Taiwan
Ming-Ta Hsu
Department of Pharmacology, College of Medicine, National Cheng Kung University, No. 1, University Road, Tainan, 701, Taiwan
Chu-Yi Wu & Yi-Ching Wang

Authors

Chung-Han Tsai
View author publications
You can also search for this author in PubMed Google Scholar
Hung-Chi Cheng
View author publications
You can also search for this author in PubMed Google Scholar
Yu-Shiuan Wang
View author publications
You can also search for this author in PubMed Google Scholar
Pinpin Lin
View author publications
You can also search for this author in PubMed Google Scholar
Jayu Jen
View author publications
You can also search for this author in PubMed Google Scholar
I-Ying Kuo
View author publications
You can also search for this author in PubMed Google Scholar
Ying-Hua Chang
View author publications
You can also search for this author in PubMed Google Scholar
Pao-Chi Liao
View author publications
You can also search for this author in PubMed Google Scholar
Ruey-Hwa Chen
View author publications
You can also search for this author in PubMed Google Scholar
Wei-Chien Yuan
View author publications
You can also search for this author in PubMed Google Scholar
Han-Shui Hsu
View author publications
You can also search for this author in PubMed Google Scholar
Muh-Hwa Yang
View author publications
You can also search for this author in PubMed Google Scholar
Ming-Ta Hsu
View author publications
You can also search for this author in PubMed Google Scholar
Chu-Yi Wu
View author publications
You can also search for this author in PubMed Google Scholar
Yi-Ching Wang
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Y.-C.W. conceived the project; C.-H.T. and Y.-C.W. designed the experiments; C.-H.T., Y.-S.W., J.J., I.-Y.K., C.-Y.W. and M.-T.H. performed and analysed experiments; H.-C.C., P.L and M.-H.Y. helped C.-H.T. to perform the animal experiments; Y.-H.C. and P.-C.L. contributed to proteomics assay; R.-H.C. and W.-C.Y contributed to VSVG/TfnR trafficking assays; H.-S.H. provided clinical samples; Y.-C.W. and C.-H.T. wrote the manuscript with edits from all authors; and Y.-C.W. obtained funding.

Corresponding authors

Correspondence to Chu-Yi Wu or Yi-Ching Wang.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Supplementary information

Supplementary Figures and Tables

Supplementary Figures 1-16 and Supplementary Tables 1-2 (PDF 2312 kb)

Supplementary Movie 1

Time-lapse movie of TIRF images of CL1-5 cells expressing RFP-tagged hRAB37 and GFP-tagged TIMP1. CL1-5 cells were co-transfected with RFP-tagged hRAB37 (hRAB37-RFP) and GFP-tagged TIMP1 (TIMP1-GFP). Images were captured with TIRF microscope at 491 and 561 nm laser every 2 second over a period of 3 min. Stills corresponding to frames from 00:00 to 00:20 of this movie are presented in Fig. 3c. Time intervals in minutes and seconds are shown (bottom). Scale bars: 10 μm. (MP4 9148 kb)

Supplementary Movie 2

Time-lapse movie of TIRF images of CL1-5 cells expressing RFP and GFP vector control. CL1-5 cells were co-transfected with RFP-tagged and GFP-tagged empty vector. Images were captured with TIRF microscope at 491 and 561 nm laser every 2 second over a period of 3 min. Stills corresponding to frames from this movie are presented in Fig. 3d. Time intervals in minutes and seconds are shown (bottom). Scale bars: 10 μm. (MP4 8973 kb)

Supplementary Data 1

The 185 proteins, identified with high confidence in two independent experiments (experiment 1 and 2), by proteomic analysis of CM from control and WT-hRAB37 expressing CL1-5 cells. (XLSX 46 kb)

Supplementary Data 2

TCGA raw data of mRNA expression of hRAB37 and TIMP1 genes. (XLS 127 kb)

Supplementary Data 3

TCGA data of hRAB37 mRNA expression and clinical follow-up. (XLS 77 kb)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Tsai, CH., Cheng, HC., Wang, YS. et al. Small GTPase Rab37 targets tissue inhibitor of metalloproteinase 1 for exocytosis and thus suppresses tumour metastasis. Nat Commun 5, 4804 (2014). https://doi.org/10.1038/ncomms5804

Download citation

Received: 17 January 2014
Accepted: 25 July 2014
Published: 03 September 2014
DOI: https://doi.org/10.1038/ncomms5804

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.