The expression of Rab proteins has been associated with cancer. However, few data are available on Rab5A expression in human breast cancer or its impact on disease progression. First, we examined the functional role of Rab5A in breast cancer cells. The expression of Rab5A in MDA-MB-231 cells can be stimulated by epidermal growth factor in a dose-dependent manner. The epidermal growth factor-induced increase of Rab5A expression correlated well with enhanced migration in wound healing migration assays in these cells. Furthermore, we evaluated the expression of Rab5A in breast cancer specimens using immunohistochemical staining, then analyzed the relationship between the expression of Rab5A and clinicopathological parameters. The increased expression of Rab5A protein in 123 breast cancer samples was associated with higher histological grade (P = 0.004), more lymphovascular invasion (P = 0.027), more axillary lymph node (LN) metastasis (P = 0.008), and a higher number of axillary LN metastases (P = 0.043). Among 218 axillary LNs of more than 10 breast cancer patients with node metastases, 167 metastatic LNs were found to have increased Rab5A expression. Rab5A is associated with axillary LN metastasis in breast cancer patients. (Cancer Sci 2011; 102: 2172–2178)
The majority of deaths from breast cancer are attributed to metastasis. Axillary LNs are often the first sites of metastasis in breast cancer patients.(1) The presence of axillary LN metastasis is a major criterion in the prognosis and in the decision-making for additional chemotherapy after primary tumor surgeries.(2) Only 20% of systemic metastases derived from tumor cells bypass the lymphatic route.(3) Therefore, axillary LN metastasis is one of the most important issues in breast cancer.
Rab GTPases, which are members of the Ras superfamily of small GTPases, are key regulators of membrane trafficking in both exocytic and endocytic pathways as well as receptor localization in eukaryotic cells.(4,5) Many studies have revealed the association between Rab GTPase dysfunction and human diseases, including cancer.(6) Rab5 is one of the most extensively studied members of Rab GTPases.(7) It plays important roles in a variety of cellular trafficking and signaling events, including receptor internalization, targeting and fusion of endocytic vesicles with early endosomes, fusion between early endosomes, actin remodeling, and signaling to the nucleus.(8) Recent studies have revealed that the overexpression of Rab5A is associated with the metastatic potential of lung and gastric cancer,(9,10) is involved in the migration of hepatocellular carcinomas,(11) and promotes ovary cancer cell proliferation.(12) Rab5A mediates the formation of EGFR-containing endosomes.(13,14) The EGFR signaling cascades can promote cell proliferation, angiogenesis, and invasion, and inhibit apoptosis, resulting in tumor growth and progression.(15) Poor prognosis was confirmed in patients of locally advanced breast cancer with overexpression of EGFR.(16) Whether Rab5A played any role in these is not clear; there are few data available on the expression of Rab5A in human breast cancer or its impact on disease progression.
In the present study, we showed that the Rab5A protein was involved in the migration of breast cancer cells. Furthermore, we evaluated the expression of Rab5A in human breast cancer specimens and correlated it with the patients’ clinicopathological parameters. Our results revealed that increased expression of Rab5A was significantly correlated with lymphatic dissemination in breast cancer patients.
Materials and Methods
Cell culture. Human MCF-7, MDA-MB-231, BT-474, and SK-BR-3 breast cancer cells were cultured in DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 2 mmol/L l-glutamine, 100 μmol/L non-essential amino acid, and 10 μg/mL gentamycin, at 37°C in a humidified atmosphere containing 5% CO2. Cells were subcultured by dispersal with TEG (0.125% trypsin, 0.25% EDTA, and 0.05% glucose in 1× PBS) and replated at 1:4 to 1:8 dilutions twice a week.
Stable clones of Rab5A-overexpressing MDA-MB-231 breast cancer cells. The TransIT-LT1 Transfection Kit (MoBiTec, Göttingen, Germany) was used for transfection of Rab5A plasmid. The day before transfection, the MDA-MB-231 breast cancer cells were seeded into a 6-well plate and incubated at 37°C for 16 h to 70% confluence without antibiotics in a humidified 5% CO2 incubator. Prior to transfection, 7.5 μL TransIT-LT1 transfection reagent (Mirus Bio, Madison, WI, USA) was mixed gently with 250 μL Opti-MEMI (Invitrogen), then 2.5 μg Rab5A plasmid was added, mixed gently, and kept at 25°C for 15–30 min. The medium of cells was replaced with fresh medium without serum and the transfection complexes were added drop-wise to the cells. After 48 h, cells were selected by incubating with the medium containing G418 and cells stably expressing Rab5A-mCherry were subcloned into a 96-well plate.
Western blot analysis. Whole cell lysates were prepared by first resuspending the cells in M-PER protein extraction reagent (Pierce, Rockford, IL, USA) supplemented with protease inhibitor cocktail (Pierce), then incubating the cells on ice for 30 min. The cell lysates were centrifuged at 14 000g for 10 min to collect the supernatant. The protein concentration was measured using the Bradford method (Bio-Rad, Berkeley, CA, USA). An aliquot of protein lysate (30 μg) from each sample was separated on a 10% SDS-polyacrylamide gel for 1 h. After separation, the proteins were transferred to a nitrocellulose membrane and the membrane was then blocked with 5% milk in 1× TTBS buffer (20 mmol/L Tris [pH 7.5], 150 mmol/L NaCl, 0.1% Tween-20 [pH 7.4]) for 1 h at room temperature. The proteins were probed with anti-Rab5A (1:200; an affinity-purified rabbit polyclonal antibody raised against a peptide mapping with the C-terminus of Rab5A of human origin) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or anti-β-actin (1:50 000) antibodies at 4°C overnight. This incubation was followed by incubation with HRP-conjugated secondary antibodies (Sigma, St Louis, MO, USA). Protein visualization was carried out using an enhanced chemiluminescence kit (Pierce) according to the manufacturer’s protocol.
Wound healing migration assay. MDA-MB-231 (8 × 105 cells per well), MCF-7 (1 × 106 cells per well), BT-474 (1 × 106 cells per well), and SK-BR-3 (6 × 105 cells per well) breast cancer cells were seeded in 6-well plates in growth medium. Confluent monolayers were incubated with serum-free DMEM overnight. A culture insert (ibidi, Martinsried, Germany) was used to generate a 500 ± 50 μm gap. The cells were washed with PBS to remove cell debris and left untreated or treated with EGF (10 or 100 ng/mL). Images of all movement in the wound area were captured with a light microscope using a 5× objective at 0 and 18 h post-wounding.
Transwell migration assay. The control MDA-MB-231 breast cancer cells or Rab5A-overexpressing MDA-MB-231 breast cancer cells (1 × 105 cells) were added to the upper chambers of 24-well Transwell plates (8 μm pore size; Corning Costar, Corning, NY, USA). The upper chambers contained cells with FBS-free DMEM medium, and the lower chambers contained DMEM containing 10% FBS as a chemoattractant. The cells were incubated for 4 h at 37°C in 5% CO2/95% air. Non-migrating cells were scraped off the upper surface of the membrane with a cotton swab. Migrating cells that remained on the bottom surface were fixed with 100% methanol, then stained with 0.4% Giemsa solution (Sigma-Aldrich, Steinheim, Germany). Migrating cells were counted using a light microscope (Nikon, Tokyo, Japan). At least three separate microscopic fields were counted per membrane.
Tissue samples. Paraffin-embedded breast cancer samples (n = 123) were randomly selected and provided by the Surgery Residual Tissue Bank (Taipei Veterans General Hospital, Taiwan). Institutional review board approval had been obtained for this retrospective study from Taipei Veterans General Hospital. The morphologic classification of breast cancer was carried out according to World Health Organization specifications.
Immunohistochemical staining of Rab5A. The IHC stain was carried out for both the primary breast cancer site and metastatic axillary LNs on representative slides. Rab5A expression in the tumor was determined using the Dako LSAB2 system kit (Dako, Carpinteria, CA, USA). Briefly, the rehydrated tissue sections were first microwaved in a sodium citrate buffer (10 mM, pH 6.0) then treated with 3.0% H2O2 for 10 min and soaked with blocking solution for an additional 10 min. Next, the tissue sections were incubated with anti-Rab5A antibody (1:100; Santa Cruz Biotechnology) at room temperature overnight in a moist chamber. Then tissue section slides were washed in PBS and incubated with biotin-labeled secondary antibody for 10 min, followed by treatment with streptavidin HRP conjugate for 10 min. After incubation with 3,3′-diaminobenzidine-tetrachloride substrate chromogen for 10 min, Mayer’s hematoxylin counterstain was applied for 10 min (Muto Pure Chemicals, Tokyo, Japan). Finally, the mounting solution (Kaiser’s glycerin gelatine; Merck, Darmstadt, Germany) was added to the sections before they were covered with coverslips for histological examination.
Evaluation of IHC staining for Rab5A. The intensity, staining percentage, and pattern of staining (nuclear and cytoplasmic) of the Rab5A protein were examined by two independent pathologists (Dr CY Hsu and Dr MY Lee) who were unaware of the patients’ clinicopathological findings. Nuclear versus cytoplasmic location of expression was recorded in each sample. The cytoplasmic stain was categorized based on the intensity of the staining as 0 (negative), 1 (low), 2 (moderate), or 3 (strong) compared to background staining. We set the IHC stain on a scale from 0 to 3 according to the expression of Rab5A in the cytoplasm and nucleus separately. Negative or low cytoplasmic staining was considered weak cytoplasmic Rab5A expression, whereas moderate or strong staining was considered strong cytoplasmic Rab5A expression. The percentage of positive cells was determined by calculating the ratio between the positively stained tumor cells and the total breast cancer cells examined. The nuclear stain was defined using a scale of 0 to 3 to represent the percentage of cells stained positive: 0, 0–5% nuclear stain; 1, 5–30% nuclear stain; 2, 30–60% nuclear stain; and 3, >60% nuclear stain. The nuclear stain values of 0 or 1 were considered weak nuclear Rab5A expression; values of 2 or 3 were considered strong nuclear Rab5A expression. To evaluate the overall expression of Rab5A in tumor cells, we added the cytoplasmic score and nuclear score together. A total score equal or >3 represented “strong” overall Rab5A expression. Everything <3 was defined as weak overall Rab5A expression.
Statistical analyses. Statistical analyses were carried out using SAS (version 8; SAS Institute, Cary, NC, USA). Data are presented as the mean ± SD except where indicated. The univariate analysis was used to determine the relationship between Rab5A and the clinicopathological parameters in breast cancer patients. The difference between the means is considered significant when P < 0.05.
Rab5A expression and the migration ability of different breast cancer cells with or without EGF stimulation. We examined Rab5A expression in the four breast cancer cell lines stimulated by EGF (Fig. 1). The Rab5A level of MCF-7 cells had no change after treatment with EGF. Rab5A expression mildly increased in both BT-474 and SK-BR-3 cells after treatment with 10 ng/mL EGF, but no further increase was observed after treatment with 100 ng/mL EGF. For MDA-MB-231 cells, EGF treatment resulted in a dose-dependent increase of Rab5A expression. Figure 2 shows a comparison of the migration ability of MCF-7, MDA-MB-231, BT 474, and SK-BR-3 breast cancer cells obtained using the wound healing method. The MDA-MB-231 cells have the highest migration ability, compared to the other three breast cancer cell lines, at 18 h after wounding. The EGF (10 and 100 ng/mL) treatment of MDA-MB-231 cells increased the migration ability of these cells, but not of the other three breast cancer cell lines. Similar results were observed in the Transwell migration assay (data not shown).
Migration ability increased in Rab5A-overexpressing MDA-MB-231 stable clones. To further understand the relationship between Rab5A and migration in MDA-MB-231 breast cancer cells, we overexpressed Rab5A in MDA-MB-231 cells. Figure 3 shows that the protein level of Rab5A was increased in Rab5A-overexpressing clone 1 and 2 MDA-MB-231 cells. The migration abilities of the control and Rab5A-overexpressing stable clone 1 and 2 MDA-MB-231 breast cancer cells were compared using the wound healing method at 4 and 18 h (Fig. 4). The migration ability increased in Rab5A-overexpressing stable clone 1 and 2 MDA-MB-231 breast cancer cells as compared to the control cells. Furthermore, we also examined the migration ability of Rab5A-overexpressing stable clone 1 and 2 MDA-MB-231 breast cancer cells using the Transwell method. The migration ability increased in Rab5A-overexpressing stable clone 1 and 2 MDA-MB-231 breast cancer cells compared to the control cells (Fig. 5).
Characteristics of patients. There were 123 breast cancer patients included in this study. All patients were operated at the Taipei Veterans General Hospital between September 2001 and October 2005. The mean age was 53.4 years (range, 25–88 years). The median follow-up lasted 34 months (range, 4–71 months). Eighteen out of 123 (14.6%) patients died during the follow-up period. The majority of lesions presented with T2 (78/123, 63.4%) (tumor size 2–5 cm), followed by Tis-1 (30/123, 24.4%) (ductal carcinoma in situ or tumor size <2 cm), then T3 (15/123, 12.2%) (tumor size >5 cm). All patients received complete axillary LN dissection. The mean dissected axillary LN number was 17 (range, 4–41). In 65 of 123 patients (52.8%), axillary LN metastasis was found. Among them, 26 patients were noted with 1–3 metastatic axillary LNs, 22 patients with 4–9 metastatic nodes, and 17 with 10 or more metastatic nodes.
Immunohistochemical staining pattern of Rab5A in breast cancer patients. Sections from the 123 paraffin-embedded breast cancer samples were used for the analysis of Rab5A expression levels by IHC staining. The specificity of the Rab5A antibody was confirmed using a blocking peptide. There was granular staining of the tumor cell cytoplasm in positive cases (Fig. 6A,B). Some tumor cells also showed nuclear staining (Fig. 6A,C). Based on our previously mentioned criteria, 92 of the 123 cases (74.8%) were categorized as having strong overall expression of Rab5A by IHC staining, 63 (51.2%) had strong cytoplasmic Rab5A staining, and 71 (57.7%) had strong nuclear Rab5A staining.
Overexpression of Rab5A associated with axillary LN metastasis. The relationship between the expression of Rab5A and the clinicopathological findings in these 123 breast cancer cases is shown in Table 1. The histological grade was the only association found with cytoplasmic Rab5A expression (P = 0.033). Axillary LN metastasis was found to be associated with nuclear Rab5A expression (P = 0.045). The patients with strong overall Rab5A expression in their tumor cells had significantly higher histological grade (P = 0.004). Strong overall expression of Rab5A was also found to be associated with axillary LN metastasis (P = 0.008) and lymphovascular invasion (P = 0.027). Interestingly, we found that the higher the number of axillary LN metastases, the higher the percentage of strong overall Rab5A expression (the percentages of strong Rab5A expression in 0, 1–3, 4–9, and ≥10 metastatic nodes groups were 63.8, 80.8, 81.8, and 94.1%, respectively) (P = 0.043). The expression of Rab5A was not associated with ER, PR, or HER2/neu status. We further analyzed the clinicopathological factors against axillary LN status in our breast cancer patients and found that lymphovascular invasion (P < 0.001) and the expression of Rab5A (P = 0.008) were associated with axillary LN metastasis. The histological grade was significantly correlated with axillary LN metastasis (P = 0.024) as well. In contrast, tumor size, ER, PR, and HER2/neu were not linked to nodal metastasis in our series.
Table 1. Correlation between Rab5A immunohistochemical staining and clinicopathological variables in 123 breast cancer patients
Expression of Rab5A in metastatic axillary LNs of breast cancer patients. We randomly chose 10 of the 17 breast cancer patients with more than 10 axillary LN metastases to analyze Rab5A expression in the axillary LNs. A total of 218 axillary LNs were used for analysis by IHC staining; of these, 167 (76.6%) were noted with metastatic breast cancer cells. Surprisingly, Rab5A was found to be expressed in all of these 167 metastatic LNs. Further analysis showed that tumor cells showed diffuse and strong granular immunoreactivity for Rab5A in the cytoplasm of all metastatic cells, but less so in the nucleus of metastatic breast cancer cells in the LNs (Fig. 6D). No immunoreactivity of Rab5A was observed in the remaining 51 LNs free of metastatic breast cancer cells. The expression of Rab5A in the cytoplasm or nucleus of the primary breast cancer was not associated with the expression in these metastatic LNs (data not shown).
The main prognostic factors associated with breast cancer are the number of LNs involved, tumor size, histological grade, and hormone receptor status.(17) Our results clearly showed that Rab5A expression was associated with axillary LNs metastasis, lymphovascular invasion, and histological grading in breast cancer patients. In addition, the expression of Rab5A is involved in the ability of breast cancer cells to migrate.
It is well known that higher histological grading is associated with an increasing risk of distant recurrence, especially in node-negative breast cancer.(18,19) In this study, we observed that Rab5A expression in our breast cancer patients was associated with histological grading (Table 1). The histological grade of an invasive carcinoma of the breast is calculated by tubular formation, nuclear pleomorphism, and mitotic frequency, according to the Bloom–Richardson method.(18) The histological grade could reflect the proliferative capacity and differentiation of the tumor cells. Rab5 was reported to be responsible for structuring the endoplasmic reticulum during mitosis.(20) Because endoplasmic reticulum structure is closely related to other organelles, including the Golgi apparatus, plasma membrane, lysosomes, and late endosomes,(21) it is very likely that increased Rab5A expression modulated endoplasmic reticulum structure, which resulted in changes in the histological grade or differentiation of breast cancer cells.
The relationship between the distribution of Rab5A in the cytoplasm or nucleus and its functional role in breast cancer cells are still unknown. It is interesting to note that the expression of Rab5A in the cytoplasm is associated with histological grade. In addition, the expression of Rab5A in the nucleus is associated with axillary LN metastasis (Table 1). Rab5A plays a key role in controlling protein (such as receptors) traffic through the early stages of the endocytic pathway. The modulators and effectors of Rab5A have been reported to be involved in the regulation of several signal transduction pathways involved in the movement of protein through endocytic compartments.(22) Activation of EGFR leads to a redistribution of APPL1, a Rab5 effector that propagates signals from early endosomal structures directly to the nucleus. APPL1 protein may play a role in modulating cell proliferation through interactions with a number of components of the nucleosome remodeling and histone deacetylase NuRD/MeCP1 complex.(23) Zhao and colleagues(12) reported that Rab5A overexpression that promotes ovarian cancer cell proliferation may be associated with the APPL1-related EGF signaling pathway. The nuclear transport of the membrane protein EGFR functions in transcriptional regulation and protein kinase signaling.(24) Additionally, there is a report that EGFR expression in the nucleus, but not in the non-nuclear compartment, correlated significantly with overall survival in a cohort of breast cancer patients.(25) Activation of EGFRs can result in the net activation of Rab5 and an increase in EGFR internalization and EGF signal attenuation.(22) The function of Rab5A in the cytoplasm and in the nucleus needs further study.
Identification of a “lymphatic metastasis signature” in the primary tumor is important. The nodal metastases could represent a selected subpopulation of breast cancer cells;(26) however, to accurately identify these cells is not an easy task. In this study, we found that Rab5A expression was heterogeneous in some primary tumors but immunoreactivity was absent in others. Some primary breast cancers had dominant cytoplasmic Rab5A staining, whereas others had dominant nuclear Rab5A staining. The expression level of Rab5A in the cytoplasm or nucleus of primary breast tumors was not associated with their expression in metastatic LNs (data not shown). Nevertheless, an important finding was that Rab5A was expressed in all of the 167 metastatic LNs in our series. It may be that only some selected subpopulations of primary breast cancer cells can metastasize to axillary LNs. We further compared ER expression by IHC staining in these 167 metastatic axillary LN samples and their primary tumors. Six of 10 primary breast cancer samples were found to be ER− and all of their metastatic axillary nodes were also ER−. The other four primary breast cancer samples were ER+, and some metastatic nodes were ER− (59 of 82 metastatic nodes presented with negative ER) (data not shown). These results support our hypothesis that only a selected set of breast cancer cells metastasize to the LN. These results, together with our observation of the significant relations between Rab5A expression and LN metastasis (Table 1), strongly suggest that Rab5A may be a potential marker related to metastasis to LNs.
It has been reported that EGF stimulation can activate Rab5 functions, which is consistent with the notion that a relationship exists between receptor activation and internalization.(13,14,27) Once activated by a tyrosine kinase receptor, such as the one for EGF, Rab5 stimulates PI3K, whose p85 regulatory subunit acts as a GTPase-activating protein (GAP) on Rab4 and Rab5.(28) At the same time, the EGFR controls the activity of Rab5 and the rate of its own endocytosis.(13,14,29) We observed that EGF treatment resulted in the concomitant increase of Rab5A expression and the migration ability of MDA-MB-231 cells, but not MCF-7 cells (Figs 1,2). The detailed mechanism behind this difference is not clear. One possibility is that EGFR levels in these cells play an important role. The expression of EGFR was reported to be the highest in MDA-MB-231, followed by SK-BR-3, and the lowest level was found in MCF-7 breast cancer cell lines.(30) The expression of EGFR was tightly coupled to migration in MCF-7 cells; Kruger and Reddy(31) clearly showed that forced overexpression of EGFR in MCF-7 cells resulted in increased migration. From our data, Rab5A expression increased more in MDA-MB-231 cells (with higher EGFR expression) than in other breast cancer cell lines (with lower EGFR expression) under the stimulation of EGF. Moreover, overexpression of Rab5A also increased the migration ability of MDA-MB-231 cells (Figs 4,5). Together, these results suggest that the function of Rab5A is tightly coupled to EGF and EGFR signaling and linked to breast cancer cell migration.
Rab5 was reported to regulate cell motility by Rab5-dependent endocytic control over Rac activation and spatially restricted actin dynamics.(32) Caspase-8 was shown to influence cell adhesion and migration through an interaction with PI3K in a Rac-dependent manner, and Rab5 can be activated by caspase-8.(33,34) Some intimate relationships between Rab5 and integrin trafficking are thought to coordinate signaling among endocytic, cytoskeletal, and apoptotic pathways during cell migration.(34,35) Our finding, that cell migration increased under the stimulation of EGF in MDA-MB-231 breast cancer cells, is consistent with these observations. The ability of Rab5-based circuitries to control the cellular modes of motility has important implications for understanding the molecular mechanism of metastasis and for therapeutic intervention as well.(32)
This study is the first to analyze the expression of Rab5A in a series of human breast cancer tissues and cell lines. We found that Rab5A is an important factor, but not the only factor, in regulating the migration of breast cancer cells, and that Rab5A was associated with axillary LN metastasis in breast cancer patients. Rab5A has the potential to become a therapeutic target in breast cancer patients in the future.
This work was supported by grants NSC 95-2314-B-075-060-MY2 and NSC 97-2314-B-075-007-MY3 from the National Science Council of Taiwan, VGH 99C1-059, VGH 98C1-033, and VGH 98C1-090 from Taipei Veterans General Hospital, and Department of Health (DOH100-TD-C111-007) (Taipei, Taiwan). We thank Dr. Yueh-Hsin Ping for providing the Rab5A-mCherry plasmid and Dr. Ming-Ta Sung for technical support and critical review of the manuscript.
The authors have no conflict of interest.
adaptor protein containing Ph domain, PTB domain, and leucine zipper motif