Arf6-driven endocytic recycling of CD147 determines HCC malignant phenotypes
- 215 Downloads
Adhesion molecules distributed on the cell-surface depends upon their dynamic trafficking that plays an important role during cancer progression. ADP-ribosylation factor 6 (Arf6) is a master regulator of membrane trafficking. CD147, a tumor-related adhesive protein, can promote the invasion of liver cancer. However, the role of Arf6 in CD147 trafficking and its contribution to liver cancer progression remain unclear.
Stable liver cancer cell lines with Arf6 silencing and over-expression were established. Confocal imaging, flow cytometry, biotinylation and endomembrane isolation were used to detect CD147 uptake and recycling. GST-pull down, gelatin zymography, immunofluorescence, cell adhesion, aggregation and tight junction formation, Transwell migration, and invasion assays were used to examine the cellular phenotypes. GEPIA bioinformatics, patient’s specimens and electronic records collection, and immunohistochemistry were performed to obtain the clinical relevance for Arf6-CD147 signaling.
We found that the endocytic recycling of CD147 in liver cancer cells was controlled by Arf6 through concurrent Rab5 and Rab22 activation. Disruption of Arf6-mediated CD147 trafficking reduced the cell-matrix and cell-cell adhesion, weakened cell aggregation and junction stability, attenuated MMPs secretion and cytoskeleton reorganization, impaired HGF-stimulated Rac1 activation, and markedly decreased the migration and invasion of liver cancer cells. Moreover, high-expression of the Arf6-CD147 signaling components in HCC (hepatocellular carcinoma) was closely correlated with poor clinical outcome of patients.
Our results revealed that Arf6-mediated CD147 endocytic recycling is required for the malignant phenotypes of liver cancer. The Arf6-driven signaling machinery provides excellent biomarkers or therapeutic targets for the prevention of liver cancer.
KeywordsArf6 CD147 Endocytic recycling Malignant phenotype Liver cancer
ADP-ribosylation factor 6
guanine nucleotide exchange factor
anti-CD147 monoclonal antibody 18
Dynamic endocytosis and recycling of membrane proteins control numerous pathophysiological functions including cell homeostasis, nutrient uptake, and oncogenic signaling. Membrane proteins containing the clathrin-adapter-bound sequence are internalized via clathrin-mediated endocytosis (CME). When they lack specific sorting sequence they enter the cell through clathrin-independent endocytosis (CIE). Two types of CIE cargo proteins join different trafficking itineraries within the cell . One (such as MHCI, CD59, and CD55) travels along a bulk route (B-cargo) that directs some cargos for slow recycling and some cargos toward lysosomes for degradation. Another (including CD147, CD98, CD44, and Glut1) travels an alternative route (A-cargo) in which nearly all of these cargos are rapidly recycled to the cell surface . As a consequence, CD147 and other ‘A-cargo’ proteins are long-lived.
CD147 (Basigin/EMMPRIN) is an adhesion molecule overexpressed in multiple tumors [3, 4]. Mature CD147 is an N-linked glycosylated protein and exists both in transmembrane and soluble forms. Depending on the glycosylation level, CD147 exists in two forms: high-glycosylated CD147 (HG-CD147) and low-glycosylated CD147 (LG-CD147) [5, 6]. Because it is activity involved in metabolic reprogramming, apoptosis inhibition, motile migration, and multidrug resistance, CD147 serves as a hub protein in hepatocarcinogenesis [5, 7, 8, 9]. Although previous studies revealed the role of CD147 trafficking in the progression of several types of cancers [10, 11, 12], the detailed regulation signal in liver cancer needs to be clarified.
Cell-surface protein abundance is controlled by G protein (guanine nucleotide-binding protein)-mediated transport through the regulation of endocytic recycling. Previous studies reported that CD147 is internalized through Rab5-associated CIE, recycled via Rab22-dependent endosomes, and by-passed merging with the EEA1-positive endosomes in cervical cancer cells [13, 14, 15]. Arf6 (ADP-ribosylation factor 6) is another GTPase that regulates the endocytic recycling process in concert with different Rab GTPases [2, 16, 17, 18, 19, 20, 21, 22]. Arf6 activation can further promote CD147 trafficking, especially to accelerate it entering in the fast recycling pathway [15, 23]. This might avoid its transport to the slow recycling route through Rab11-positive endosomes or the default degradation pathway. Although Arf6 and Rab GTPases are independently involved in the CD147 trafficking process [9, 10, 15, 24, 25, 26, 27], investigations on the precise mechanism governing CD147 distribution on the cancer cell surface are still required.
In this study, the role of Arf6 on CD147 trafficking in liver cancer cells and its contribution to the malignant behaviors of HCC (hepatocellular carcinoma) were examined. We showed that, by modulating the Rab5 and Rab22 co-activation, cell adhesion and junction formation, Arf6-driven CD147 endocytic recycling causes liver cancer cells to acquire migratory and invasive phenotypes. The Arf6-mediated CD147 signaling functions as a critical determinant for poor clinical outcome of HCC patients.
Material and methods
Cell cultures, plasmids, antibodies, and chemicals
7721, HepG2, and Huh7 cells were obtained from Type Culture Collection of the Chinese Academy of Sciences (China). Arf6(wt)- and Arf6(Q67L)-HA/pcDNA3 plasmids were kindly supplied by Dr. Shumei Wei (Zhejiang University) . EEA1- and Rabaptin-5/pGEX-4T-3 plasmids were kindly supplied by Prof. Byung-Ha Oh (Korea Advanced Institute of Science and Technology). The pGEX-GST-PAK (CRIB) plasmid was previously obtained from the Lab of Prof. Open image in new window (Memorial Sloan-Kettering Cancer Center).
Mouse anti-CD147 Ab (H18) was originally produced . Rabbit anti-Rab5 Ab (D160063), anti-Rab22 Ab (D160036), anti-Rac1 (SC-95), anti-Na/K ATPase Ab (sc-71,637) and mouse anti-Arf6 (sc-7971) were obtained from Santa Cruz. Rabbit anti-ZO-1 Ab (WL03419), anti-E-cadherin Ab (WL01482), anti-pan-cadherin Ab (WL03295) and anti-β-catenin Ab (WL0962a) were obtained from Wanleibio. Rabbit anti-CD147 PcAb (AB22048b), anti-HA Ab (D110004), mouse anti-β-actin Ab (D190606) and collagen (A001654) were obtained from Sangon Biotech. Mouse anti-ARNO Ab (AA 314–399) was obtained from 4A Biotech. Cell fractionation isolation kits (89881 and 78840), goat anti-mouse Ab-Alexa Fluor (AF) 488 (A-11001), anti-mouse AF647 (A-31571), anti-rabbit Ab-AF647 (A27040), NeutrAvidin agarose (29200), GSH Glutathione sepharose (G2879), Hochest33342 (62249), and Rhodamine-phalloidin (R415) were obtained from Life Technologies. Matrigel (356234), fibronectin (F2006), laminin (L2020), gelatin (G1890), polybrene (H9268), puromycin (P7255), NHS-SS-biotin (21328), HGF (H0536), and the remaining chemicals used in this study were from Sigma.
Gene stable knock-down and overexpression
The pLV-RNAi system (BIOSETTIA, SORT-B19) was utilized to produce Arf6-KD (knocked-down) stable cell lines. Three independent Arf6-targeting shRNA (A1: AAAAGGAAGGTGCTATCCAAAATTTGGATCCAAATTTTGGATAGCACCTTCC, A2: AAAACAACAATCCTGTACAAGTTGATTGGATCCAATCAACTTGTACAGGATTGTTG, and A3: AAAAGCTCACATGGTTAACCTCTAATTGGATCCAATTAGAGGTTAACCATGTGAGC) were generated as previously described . Arf6(wt) and Arf6(Q67L) plasmids were transfected in liver cancer cells for gene over-expression.
Biotinylation and subcellular fractionation
Cell-surface and endomembrane proteins were isolated by using fractionation kits. Briefly, dish-grown confluent cells were surface-labeled with NHS-SS-biotin (0.2 mg/ml in PBS) at 4 °C for 30 min, quenched with 0.1 M glycine, and collected and solubilized by sonication. Cell lysates were applied on a NeutrAvidin agarose-loaded column, and the labeled proteins were centrifuge-eluted after incubation with 50 mM DTT solution. For endomembrane isolation, detached cells were digested with Typsin again (to maximize the removal of the surface protein). The cell pellet was incubated with cold cytoplasmic extraction buffer for 10 min, and the resulting pellet was further treated with cold membrane extraction buffer for 10 min. The final supernatant was the extracted endomembrane fraction. The resulting samples were corrected to equivalent protein concentrations and levels of CD147 (cell surface vs endomembrane fraction) as determined by Western blot.
Antigen internalization and recycling
CD147 uptake was measured as previously described [9, 30]. Briefly, coverslip-grown cells were incubated with the H18Ab-AF488 complex (H18Ab, 1:500, anti-mouse AF488, 1:1000) for 15 min at 37 °C, rinsed with PBS/1 M NaCl, fixed with 4% PFA, permeabilized with 0.2% saponin, FBS-blocked and stained with anti-HA Ab plus anti-rabbit-AF647 at 4 °C overnight. Cells were imaged under a confocal microscope. Eight-bit maximal projections of the z-series were established using ImageJ software. To determine the uptake by flow cytometry, trypan blue was added to quench cell surface-associated fluorescence before cells were trypsinized for immediate analysis. Controls without H18Ab were included in all experiments. 80 cells were confocal-imaged in each group, and 10,000 cells per sample were counted by flow-cytometry.
CD147 recycling was performed as previous reported without biotinylation [9, 30]. Briefly, coverslip-grown cells were incubated at 4 °C with H18Ab (1:500) for 30 min binding, washed with serum-free medium, and transferred to 37 °C for 15 min uptake (to allow antigen uptake into early endosomes). After washing with cold PBS/1 M NaCI, the cells were transferred to 37 °C for a 30 min chase (to permit antigen trafficking into recycling compartments). Then, the cells were returned to the ice followed by washing with PBS/1 M NaCI (to remove the surface-recycled Ag-Ab complex) or not. After cells were fixed and permeabilized, the surface-recycled and intracellular non-recycled CD147-H18Ab complexes were stained with anti-mouse AF488 and confocal imaged as indicated above.
Rab and Rac GTPase activation
GTPase activation was performed as previously described [9, 31]. Briefly, serum-starved cells were stimulated with HGF (20 ng/mL)-contained medium (0.5% FBS), lysed and incubated with GST-EEA1 (for Rab22 activation)- and GST-Rabaptin-5 (for Rab5 activation)-immobilized beads, respectively. After subjecting the collected pellets to SDS-PAGE, the GTP-bound Rab5 and Rab22 in samples were determined by Western blot. The Rac1-GTP level was assessed using GST-PAK-CRIB immobilized beads as previously described [29, 32, 33].
Cell adhesion and aggregation
The processes were performed as previously described [8, 9, 34]. Cell adhesion: cells were detached with EDTA (0.02%), suspended in serum-free medium, added to Matrigel (5 mg/ml)-, collagen (10μg/mL)-, fibronectin (10μg/mL)- or laminin (10μg/mL)-coated 96-well plates (2 × 104/well) and incubated for 30 min and 2 h, respectively. After removing the medium, attached cells were stained with 0.2% crystal violet, lysed with 5% SDS, and the absorbance was read at 540 nm. Slow aggregation assay: single-cell suspensions (2 × 105/ml) seeded in agar-coated six-well plates were static incubated at 37 °C for 24 h. Photographs were taken under an inverted microscope. The degree of cell aggregation was scored as follows: solitary cells (≤2 cells), small and loose aggregates (3–20 cells), middle and compact aggregates (21–200 cells), and large compact aggregates (≥200 cells). In oder to block CD147 trafficking on the cell surface, anti-CD147 pcAb (1:1000) was added to Arf6(Q67L)-expressed cells.
Transwell cell migration and invasion
Cell migration and invasion assays were performed using 24-well Transwell units with an 8-mm pore size polycarbonate filter (Millipore) according to previous method . Briefly, 5 × 104 cells were seeded into Matrigel-coated (5 mg/ml) or -uncoated culture inserts with medium containing 0.5% FBS. The lower chamber was filled with 0.5% FBS medium containing 20 ng/ml HGF as a chemoattractant. After 24 h incubation, cells remaining in the upper compartment were completely removed, whereas cells that invaded into the Matrigel and/or migrated out onto the lower surface of the membrane were fixed with 4% PFA and stained with 1% crystal violet. Ten fields were photographed for each group. Data were collected from three independent experiments, each performed in duplicate.
The process was performed as previously described . The 24 h culture cell medium (serum-free) was 20 fold concentrated by using Amicon Ultra-4 10 k devices. Equivalent amounts of protein were separately loaded in SDS-PAGE gel containing 1% gelatin, washed by 2.5% Triton, activated by incubation buffer, and stained with 0.05% Coomassie blue.
Western blot and immunofluorescence
The samples were quantified by the BCA kit, resolved by SDS-PAGE, blotted with prime Abs diluted as follows: anti-CD147 (H18, 1:2000), anti-Arf6 (1:500), anti-HA (1:1000), anti-ZO-1 (1:500), anti-E-cadherin (1:500), anti-pan-cadherin (1:500) and anti-β-catenin (1:500), anti-Rac1 (1:200), anti-Rab5 (1:500), anti-Rab22 (1:500), and anti-β-actin (1:1000). Immunofluorescent assays were performed as previously described , with Ab dilution as follows: anti-HA (1:200), anti-CD147 (H18, 1:200), anti-ZO-1 (1:100), and anti-E-cadherin (1:100).
Patient samples and immunohistochemistry
Sixty HCC patient’s specimens were collected from the Third Central Hospital of Tianjin Medical University. All patients underwent successful hepatectomy and were not treated with radiotherapy or chemotherapy before operation. Overall survival was defined as the interval between surgery and death, or between surgery and the last observation point. Kaplan–Meier analysis was used for the survival data. Informed consents were obtained and this study was approved by the Medical Ethics Committee of the Hospital.
Immunohistological analysis was performed as previous reported with revision [36, 37]. Briefly, after hydrogen peroxide blocking, paraffin sections were microwave-heated for 15 min in Tris-EDTA buffer (10 mM Tris-HCI, 1 mM EDTA, pH 9.0), blocked with goat serum, and incubated with mouse anti-CD147 Ab (1:100), anti-Arf6 Ab (1:50), rabbit anti-Rac1 Ab(1:30), and anti-ARNO Ab (1:30) overnight at 4 °C. Non-immune mouse or rabbit IgG was used as the negative control. Immunohistochemistry was performed with the Envision™ two step system (Dako, USA). Sections were treated with 3, 3-diaminobenzidine and counterstained with haematoxylin. Immunopositivity was independently evaluated by two pathologists, who were blinded to the clinical data, and scored as follows: low staining (negligible, or 1+ positivity regardless of positive cell percentages, or 2+ positivity of < 30% of cells), high staining (2+ positivity of ≥30% of cells, or 3+ positivity of ≤50% of cells, or 3+ positivity of > 50% of cells).
Bioinformatics and statistical analysis
Microarray mining analysis as performed based on the GEPIA (Gene Expression Profiling Interactive Analysis) database (http://gepia.cancer-pku.cn/) . The ANOVA differential method was used for tumor (T) vs paired normal (N) samples. Confocal microscopy, flow cytometry, and western blot data were derived from three independent experiments. All data were analyzed using GraphPad Prism 5 software and are described as the mean ± SD values.
Arf6 up-regulates the endocytic recycling of CD147 by activating Rab5 and Rab22
Arf6-mediated CD147 recycling promotes cell adhesion, aggregation, and junction formation
To check whether Arf6-mediated CD147 recycling affected intercellular junctions, expression of key junction proteins were investigated. Western blot results showed that Arf6-KD significantly suppressed the levels of the tight junction marker ZO-1 and the adheren junction marker E-cadherin (Fig. 4g, h). Notably, extraneous over-expression of Arf6(wt), especially Arf6(Q67L), rescued the Arf6-KD-reducted ZO-1 and E-cadherin levels. In contrast, levels of pan-cadherin and β-catenin were not affected by Arf6 intervention. Correspondingly, ZO-1 staining markedly disappeared due to Arf6-KD-induced CD147 diminishing on the cell surface (Fig. 4i). Attenuated E-cadherin staining was also observed in Arf6-KD cells but not detected in those further overexpressed with Arf6(wt) or Arf6(Q67L). Seeing that CD147 is generally targeted to the basolateral membrane in epithelial cells , we next checked the apical-basal turnover of CD147. In Huh7 cells, CD147 was polarized to the basolateral membrane, and was translocated to the apical membrane when Arf6 was depleted (Fig. 4j). Interestingly, over-expression of Arf6(Q67L) but not Arf6(wt) restored the targeting of CD147 to the basolateral membrane. In addition, the apical recycling of CD147 in Arf6-KD cells correlated with an enhanced CD147 secretion level (Fig. 4k, l). These data revealed the tight-link between Arf6-mediated CD147 recycling and liver cancer cell adhesion, aggregation and junction stability.
Arf6-mediated CD147 recycling facilitates liver cancer cell migration and invasion
Arf6-CD147 signaling correlates with a poor clinical outcome of liver cancer patients
Clinicopathological features of HCC patients and association with Arf6-CD147 signaling components
All patients (n = 60)
Male (n = 45)
Female (n = 15)
Age at surgery (years)
<59 (n = 34)
>59 (n = 26)
1, 2 (n = 22)
3, 4 (n = 38)
Portal vein tumor thrombus
+(n = 6)
- (n = 54)
G1,2 (n = 53)
G3,4 (n = 7)
<400 (n = 46)
≥400 (n = 14)
Maximal tumor size (cm)
<5 (n = 42)
≥5 (n = 18)
Background liver status
With cirrhosis (n = 40)
Without cirrhosis (n = 20)
+(n = 35)
- (n = 25)
Compared with much research on Arf6-mediated clathrin-dependent trafficking [2, 19, 20, 22], Arf6-driven clathrin-independent trafficking events have been less studied. Previous studies using HeLa cell as the model reported that Arf6 does not contribute to the uptake of the CIE cargo, but its inactivation is required for cargo sorting soon after entry and Arf6 activation is essential for the recycling of the CIE cargo . CD147 is a typical ‘A-cargo’ protein that uses CIE to enter cells and directly recycles to the cell surface [9, 15]. Here, we found that Arf6 intervention slightly influenced CD147 uptake but markedly affected its recycling (Fig. 1a-c, Fig. 2a-c and Additional file 1: Figure S2), which resulted in CD147 being highly present on the surface of liver cancer cells. Further over-expression of the Arf6(Q67L) active-mutant completely reversed Arf6-KD-reduced CD147 endocytic recycling, highlighting that Arf6 activation can facilitate both the endocytosis and the recycling of CD147. Similar to the observation in HeLa cells [2, 18, 40], CD147 was accumulated in the endomembrane when Arf6 was depleted or further overexpression of Arf6(wt) or Arf6(Q67L) (Fig. 1d-f). This Arf6 mutant-induced endosome-trapping mirrors with its excessive reversion effect on CD147 uptake, strongly suggesting that cyclic activation and inactivation of Arf6 are required for the endocytic recycling of CD147.
Intracellular trafficking of ‘A-cargo’ CIE proteins is regulated by certain Rab GTPases [2, 18]. Generally, Rab5 activation boosts early steps of CD147 uptake, and Rab22 activation accelerates the direct recycling of CD147 to the cell surface [24, 25]. We found that Arf6-KD reduced Rab5 and Rab22 activation in liver cancer cells, and such reductions were recovered by Arf6(wt), especially Arf6(Q67L) over-expression (Fig. 3). To our knowledge, this is the first report on Arf6 expression acting on Rab activation. As Rab22 is responsible for sorting ‘A-cargo’ proteins away from the Rab5-associated endosomes and into tubular recycling endosomes [18, 41], the phenomenon that Arf6-KD reduced CD147 recycling is logical. On the other hand, because Rab5 is the central endosome Rab defining initial sorting events , Arf6(Q67L)-induced Rab5 over-activation that leads to CD147 trapped in the CIE endosomes is a legitimate inference.
Recycled endosomes return membrane proteins back to the cell surface which is important for cell adhesion . Previous studies revealed the contribution of CD147 to cell adhesion with a direct knock-down or over-expression strategy [42, 43, 44, 45, 46]. We found that Arf6-mediated CD147 recycling promotes liver cancer cells adhering to key ECM-components (Fig. 4a-c, and Additional file 1: Figure S3). CD147 decrease on the cell surface reduced cell adhesion to collagen and fibronectin but not to laminin, suggesting that the endocytic recycling of ECM-bound molecules (including but not limited to CD147) are differentially regulated by Arf6. In epithelial cells, Arf6 is an important regulator of intercellular adhesion and CD147 plays a significant role in adhesion modulation of liver cancer cells [47, 48]. It was found that Arf6-KD reduced CD147 recycling which impaired the cell-cell contact and prevented the aggregation of liver cancer cells (Fig. 4e, f and Additional file 1: Figure S4). Reduced cell-matrix adhesion and weakened cell-cell contact facilitate liver cancer cells acquiring the migration phenotype (Fig. 5).
To initiate local tissue invasion, epithelial cancer cells have to detach from the primary site by disassembling the cell-cell junction. Using polarized epithelial cells as the model, most studies showed that Arf6 activation promotes E-cadherin uptake from cell-cell contact sites to early endosomes, which leads to the disassembly of adherent junctions [20, 22, 47]. Here, we observed E-cadherin being significantly diminished or degraded in Arf6-KD liver cancer cells. In contrast, with further over-expression of Arf6(wt) or Arf6(Q67L), a recovered E-cadherin level was detected (Fig. 4g). This phenomenon implies that adequate activation of Arf6 is critical for stimulating the turnover of E-cadherin. ZO-1 is a key tight junction protein involved in the establishment of hepatic cell polarity . Although a study reported that CD147 depletion results in a ZO-1 increase in prostate cells , we found that Arf6-KD reduced CD147 recycling diminished ZO-1 in liver cancer cells, and Arf6 activation restored ZO-1 turnover (Fig. 4g-i). As Rab22 and Rac1 are activated by Arf6 (Fig. 3, Fig. 5g), and they direct the traffic of junction proteins to cell-cell contact sites via mutual signal crosstalk , it is most likely that Arf6 intervention changed the selective targeting of CD147 and the retention of ZO-1 and E-cadherin to lateral membranes (Fig. 4g-j), which cooperatively triggered the disassembly of functional junctions between adjacent liver cancer cells.
Although CD147 and Arf6 significant expression in liver cancer were already reported [5, 7, 8, 10, 48, 55, 56], we found for the first time that expression of the CD147-assocaited Arf6-ARNO-Rac1 signal axis in liver cancer tissues was significantly higher than in surrounding non-tumorous tissues (Fig. 6a-m, and Fig. 7a, b). Markedly, high expression of Arf6-CD147 signaling components was significantly correlated with more aggressive characters, in terms of advanced TNM stage, portal vein tumor thrombus, high AFP level, and short overall survival, which are putative clinicopathological markers for HCC development, invasiveness, and unfavorable prognosis (Fig. 7c-k, Table. 1, and Additional file 1: Table S1). These data strongly revealed that active Arf6-CD147 signaling occurs in HCC and is tightly associated with the HCC malignant phenotype and renders it as a potential excellent biomarker. Further studies are required to gain insight into the details of this signaling, and to develop new therapies targeting this Arf6-driven CD147 trafficking pathway.
In summary, our study demonstrates that Arf6 is essential for the endocytic recycling of CD147 and its mediated malignant phenotypes in liver cancer cells. Moreover, we provide evidence supporting that high expression of the Arf6-CD147 signaling components are tightly correlated with poor overall survival of HCC patients.
We are grateful to Prof. Zhinan Chen (Fourth Military Medical University) for supplying key materials. Thanks for Dr. Zhihui Gao (previous member of Lab) doing preliminary experiments at the early stage of this research. Thanks to Dr. Edward C. Mignot, Shandong University, for linguistic advice.
SZ initiated the idea, supervised the research, and wrote the manuscript. SQ, LS and JL performed most of the experiments. CZ, ZM and GL performed the immunohistochemical staining and clinic statistical analysis. GJ and QZ performed the gelatin zymography and adhesion experiments. YP performed the bioinformatics analysis. All authors read and approved the final manuscript.
This work was supported by the National Natural Science Foundation of China (No. 81373318, 30700829), Natural Science Foundation of Tianjin (No. 13JCYBJC21000, 16JCYBJC23900) and Fundamental Research Funds for the Central Universities, Nankai University (No. 63191172).
Ethics approval and consent to participate
This study was approved by the Medical Ethics Committee of the Third Central Hospital of Nankai University.
Consent for publication
The authors declare that they have no competing interests.
- 30.Cihil KM, Swiatecka-Urban A. The cell-based L-glutathione protection assays to study endocytosis and recycling of plasma membrane proteins. J Vis Exp. 2013;82:e50867.Google Scholar
- 43.Qian AR, Zhang W, Cao JP, Yang PF, Gao X, Wang Z, Xu HY, Weng YY, Shang P. Downregulation of CD147 expression alters cytoskeleton architecture and inhibits gelatinase production and SAPK pathway in human hepatocellular carcinoma cells. J Exp Clin Cancer Res. 2008;27:50.PubMedPubMedCentralCrossRefGoogle Scholar
- 46.Moreno V, Gonzalo P, Gómez-Escudero J, Pollán Á, Acín-Pérez R, Breckenridge M, Yáñez-Mó M, Barreiro O, Orsenigo F, Kadomatsu K, et al. An EMMPRIN-γ-catenin-Nm23 complex drives ATP production and actomyosin contractility at endothelial junctions. J Cell Sci. 2014;127:3768–81.PubMedPubMedCentralCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.