Hypoxic exosomes facilitate angiogenesis and metastasis in esophageal squamous cell carcinoma through altering the phenotype and transcriptome of endothelial cells
In cancer progression, hypoxia, or low oxygen tension, is a major regulator of tumor aggressiveness and metastasis. However, how cancer cells adapt to the hypoxia and communicate with other mesenchymal cells in microenvironment during tumor development remains to be elucidated. Here, we investigated the involvement of exosomes in modulating angiogenesis and enhancing metastasis in esophageal squamous cell carcinoma (ESCC).
Differential centrifugation, transmission electron microscopy and nanoparticle tracking analysis were used to isolate and characterize exosomes. Colony formation and transwell assay were performed to assess the proliferation, migration and invasion of human umbilical vein endothelial cells (HUVECs). The tube formation assay and matrigel plug assay were used to evaluate the vascular formation ability of HUVECs in vitro and in vivo respectively. An in vivo nude mice model was established to detect the regulatory role of exosomes in ESCC progression. Microarray analysis was performed to analyze the transcriptome profiles in HUVECs.
Exosomes derived from ESCC cells cultured under hypoxia played a better role in promoting proliferation, migration, invasion and tube formation of HUVECs in vitro and in vivo than exosomes from ESCC cells cultured under normoxia. Moreover, hypoxic exosomes significantly enhanced the tumor growth and lung metastasis compared with normoxic exosomes in nude mice models. Interestingly, endothelial cells were programmed by hypoxic and normoxic exosomes from ESCC cells which altered the transcriptome profile of HUVECs.
Taken together, our data identified an angiogenic role of exosomes from ESCC cells which shed light on the further application of exosomes as valuable therapeutic target for ESCC.
KeywordsESCC Angiogensis Exosomes Metastasis
Esophageal squamous cell carcinoma
Fetal bovine serum
Human umbilical vein endothelial cells
Kyoto Encyclopedia of Genes and Genomes
Long non-coding RNA
Search Tool for the Retrieval of Interacting Genes
Esophageal cancer is one of the most common malignant tumors worldwide which seriously threatens human health . Esophageal squamous cell carcinoma (ESCC) is the predominant form of esophageal cancer worldwide . Conventional therapeutic strategies such as radiotherapy, chemotherapy and surgery show limited effect for ESCC treatment, especially for advanced ESCC with metastasis . During cancer progression, tumors cannot grow beyond 1–2 mm without a vascular supply, due to an insufficient supply of oxygen, hypoxia, appears [4, 5]. Hypoxia is a major driver of tumor angiogenesis [5, 6]. As a multi-step physiological process, angiogenesis comprises several sequential steps involving the proliferation, migration, and morphogenesis of endothelial cells [4, 7, 8] . In tumor microenvironment (TEM), such coordination is partially accomplished by the transfer of angiogenic factors from tumor cells to endothelial cells [4, 9].
Recent studies has verified the potential role of exosomes as important signaling entities in the cross-talk between various cell types. Exosomes are vesicles of endocytic origin released by many cells [10, 11]. The involvement of exosomes in esophageal cancer proliferation, metastasis, and drug resistance is becoming increasingly apparent [12, 13, 14, 15, 16, 17, 18]. Besides, exosomes can reflect the hypoxic status of cancer cells and further help them adapt to hypoxia through communicating with their surrounding microenvironment during tumor development . However, whether exosomes play a role in modulating angiogenesis and, hence, helping the ESCC cells conquer the hypoxic microenvironment remains to be elucidated.
In the present study, the potential effect of exosomes from ESCC on endothelial cells in TEM was firstly explored. Moreover, by culturing the ESCC in hypoxic condition, we demonstrated that hypoxia might alter the content of exosomes secreted by ESCC cells and further modulate the activation of endothelial cells which internalized these exosomes.
Material and methods
Cell lines and cell culture
In order to obtain the exosomes-depleted fetal bovine serum (FBS), FBS was firstly centrifuged at 100,000×g for 12 h in an ultracentrifuge (Beckman Coulter, Optima L-100XP, USA) at 4 °C and then discarded the pellet.
ECA109, KYSE410 and HET-1A cell lines were purchased from American Type Culture Collection. HET-1A cell line was cultured in Epithelial Cell Growth Basal Medium (Lonza, Switzerland). ECA109 and KYSE410 cell lines were cultured in RPMI 1640 (Life Technologies/Invitrogen) supplemented with 10% FBS (Gibco, New Zealand) and 1% penicillin/streptomycin (Gibco, New Zealand).
Primary human umbilical vein endothelial cells (HUVECs) were also purchased from American Type Culture Collection and maintained in endothelial cell medium (ECM) (Science cell, USA). All cells were cultivated in a humidified incubator at 37 °C with 5% CO2. For hypoxia experiments, cells were incubated in a humidified 5% CO2 and 0.5% O2 for 5 days at 37 °C.
Exosomes were extracted from ESCC cell culture medium using differential centrifugation. To this end, culture medium (9 ml) were collected from ESCC cells (1 × 107) cultured at normoxic or hypoxic conditions in a 10 cm diameter petri dish. Culture medium were centrifuged at 300 g - 5 min to eliminate cell debris. Supernatant were further centrifuged at 16,500 g-30 min and 100,000 g-2 h. Finally, the exosome pellets were washed once in a large volume of phosphate buffer saline (PBS) and then followed by centrifugation (100,000 g-2 h). The exosomes were quantified by measuring the exosomal protein with BCA™ Protein Assay Kit (Pierce, USA).
Transmission electron microscopy
The exosome-enriched suspension were suspended in 50 μL of PBS, fixed with paraformaldehyde and glutaraldehyde. Exosome sample was adsorbed onto a carbon-coated copper grid and immersed in phosphotungstic acid solution for 30 s. Then the samples were observed in a Zeiss transmission electron microscope (Zeiss, Germany).
Nanoparticle tracking analysis
The number of exosomes and the size distribution were analyzed using the Nanosight (Malvern, UK) and NTA analytical software (version 2.3, Nanosight).
Western blotting analysis
The purified exosomes and ECA109 or KYSE410 cells were lysed using RIPA buffer (Roche). The protein concentration of lysates was quantified with BCA™ Protein Assay Kit (Pierce, USA). Then proteins were separated by SDS-PAGE and transferred onto a PVDF membrane. The PVDF membrane was blocked with 5% non-fat milk and incubated with the primary antibodies (CD9 and TSG101) (Cell Signaling Technology, USA) overnight at 4 °C. The bands were probed with secondary antibody (Icllab, USA) and visualized by chemiluminescence (Millipore, MA, USA). The intensity of the protein bands was quantified by densitometry using Image J Software (National Institutes of Health). Each assay was repeated at least three times. One representative of three independent experiments was shown.
Colony formation assay
As we previously depicted, HUVECs were seeded into 6-well plates (500 cells/well) and incubated with medium containing exosomes (25 μg /mL) or not for 2 weeks. The colonies were stained with crystal violet for 15 min and then counted [20, 21]. Each assay was repeated at least three times. One representative of three independent experiments was shown.
Cell cycle analysis
As we have previously depicted, HUVECs were fixed with ice-cold ethanol for 24 h and then dyed with propidium iodide/RNase buffer (BD Biosciences, USA) for 30 min in a dark place. Samples were analyzed by BD FACS Calibur flow cytometer. Data was analyzed using Modfit . Each assay was repeated at least three times. One representative of three independent experiments was shown.
Invasion and migration assay
As we previously depicted, 0.1 mL FBS-free ECM containing 2 × 104 HUVECs, in the presence of exosomes (25 μg /mL) or not, were seeded into the upper chamber of transwell chambers (24 wells) (Merck, Germany) coated with or without Matrigel (BD Biosciences, USA). 0.6 mL of the ECM containing 10% FBS was placed into the lower chamber. 48 h later, the bottom surface of the membranes were stained with crystal violet and photographed. The number of invasive cells were counted in three random microscopic fields under a light microscope  . Each assay was repeated at least three times. One representative of three independent experiments was shown.
Exosomes labeling, internalization and confocal microscopy
Exosomes were labeled with PKH26 Red Fluorescent marker (PKH26GL, Sigma-Aldrich, Germany) as recommended by the manufacturer. HUVECs were incubated with labeled exosomes (25 μg /mL) for varying times. Then HUVECs were fixed and stained with Phalloidin-iFluor 488 Reagent (Abcam, UK) and DAPI (Solarbio, USA) according to the manufacturer’s instruction. The images were acquired with a Zeiss Laser Scanning Confocal Microscope (Zeiss, Germany).
Matrigel tube formation assay
HUVECs were seeded into a Matrigel Basement Membrane Matrix (BD, New Jersey, USA) precoated 96-well plate at 2 × 104 cells per well and cultured in ECM, in the presence of exosomes (25 μg /mL) or not. Calcein-AM (Sigma-Aldrich, Germany) was used to stain the HUVECs after seeding for 8 h. The tube-like structures were imaged using the Zeiss fluorescence inverted microscope (Zeiss, Germany). The angiogenic property was assessed by measuring the total branching length from three random microscopic fields using Image J software (National Institutes of Health). Each assay was repeated at least three times. One representative of three independent experiments was shown.
In vivo matrigel plug assay
Five hundred ul of Matrigel (BD Biosciences) mixed with (25 μg /mL) normoxic or hypoxic exosomes was injected subcutaneously into the ventral region of BALB/c nude mice (Institute of Laboratory Animal Sciences, Chinese Academy of Medical Sciences (CAMS)). After inoculation for 7 days, the matrigel was excised and then fixed with formalin overnight, embedded in paraffin, and sectioned into slides. The plugs were stained with hematoxylin and eosin (H&E) and visualized using the Zeiss inverted microscope (Zeiss, Germany) . Three mice were used for each group. One representative of three independent experiments was shown.
Expression profile analysis of RNAs in exosomes
HUVECs were divided into three groups: HUVECs incubated without exosomes (control group), HUVECs incubated with exosomes (25 μg /mL) from KYSE410 which cultured in hypoxic environment (hypo-Exo group) and HUVECs incubated with exosomes (25 μg /mL) from KYSE410 which cultured in normoxic environment (norm-Exo group). After incubation for 12 h, the whole RNA of HUVECs was extracted using trizol (Thermo, USA). Gene expression profiling was performed by Beijing CNKINGBIO Biotechnology Company Limited using Clariom™ D Pico Assay, human (Affymetrix, USA) according to user guide.
Quantile normalization and subsequent data processing were performed using Applied Biosystems™ Transcriptome Analysis Console (TAC) Software (Affymetrix, USA). Heat maps representing differentially regulated genes were generated using R software (Vienna University of Economics and Business, Austria). According to the results of microarray, RNAs with fold change > 1.5 were marked as significantly differentially expressed genes.
Gene annotation and pathway enrichment analysis
Differentially regulated mRNAs were put into gene ontology (GO) biological process enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathways analysis as previously depicted [20, 21, 24]. The results of bioinformatics analysis were plotted as bubble chart using R packaging ‘ggplot2’ [25, 26]. The interaction relationship between these mRNAs were evaluated by applying Search Tool for the Retrieval of Interacting Genes (STRING) which is an online tool designed to evaluate the protein–protein interaction (PPI) information . Node connectivity (degree) were calculated using contextual hub analysis tool in Cytoscape software to identify hub genes [28, 29]. Nodes with degree > 10 were considered as highly connected nodes (hub genes).
In vivo nude mice xenograft model
As we previously depicted, 2 × 106 ECA109 or KYSE410 cells were inoculated subcutaneously to the hind limb of BALB/c nude mice (CAMS) to construct the ESCC xenograft model. Animals injected with ECA109 or KYSE410 cells were randomly divided into five groups. When the nude mice generate tumors with a size of 100 mm3, PBS, 10 μg exosomes from normal squamous esophageal epithelial cell line (HET-1A), exosome release inhibitor (GW4869, 1 mg/kg), 10 μg exosomes from normoxic ECA109 or KYSE410 cells (norm-Exo), or 10 μg exosomes from hypoxic ECA109 or KYSE410 cells (hypo-Exo) were then injected into the center of tumor sites every 2 days. The tumor size was measured every 5 days. After 40 days, the nude mice were sacrificed and tumors were taken out to measure the weight. The proliferation and angiogenesis status of tumor tissues and lung metastasis were determined using histological examination. Tumor volume was calculated with the formula: length × width2 × 0.5 [22, 30]. Three mice were used for each group. One representative of three independent experiments was shown.
Primary tumor tissues and lungs were firstly fixed in formalin and embedded in paraffin, and then cutted into sections. The lung sections were then stained by hematoxylin and eosin. For immunohistochemistry staining, sections were incubated with primary antibodies (Ki-67, CD31) (Servicbio, China) at 4 °C overnight, followed by secondary antibody. The stained tumor sections were visualized using the Zeiss inverted microscope (Zeiss, Germany). The number of positive staining tumor cells were counted in three random microscopic fields.
Statistical analysis was conducted using GraphPad Prism 6 software. Unpaired t-test was used to analyze the differences between two groups. Comparisons among more than two groups were performed using one-way ANOVA followed by Holm-Sidak’s multiple comparison tests. P value < 0.05 was considered significant. All data are expressed as mean ± standard deviation (SD).
Characterization of exosomes from ESCC cells
Western blotting analysis demonstrated that specific exosome markers (CD9 and TSG101) were enriched in purified extracellular vesicles from the supernatant of both ECA109 and KYSE410 (Fig. 1c). All together, these results confirmed that exosomes were extracted from the ESCC cell supernatant.
ESCC cells derived exosomes were internalized by endothelial cells
Hypoxic exosomes promoted endothelial cell proliferation, cell cycle progression and migration
Hypoxic exosomes accelerated vascular formation in vitro and angiogenesis in vivo
The angiogenic potential of exosomes from ESCC cells was further evaluated in vivo by examining the recruitment of vasculature into subcutaneously implanted Matrigel plugs containing exosomes. Figure 4c showed that the plugs containing norm-Exo or hypo-Exo became more vascularized than implants without exosomes (Exosome (−)).
Histological examination showed that matrigel plug containing norm-Exo or hypo-Exo showed much more micro-vessels. The vascular density of Matrigel plugs containing hypo-Exo were significantly higher than those containing norm-Exo. (Figure 4d-e) These data suggested that hypo-Exo played a better role for endothelial cells recruitment and vascular organization than norm-Exo both in vitro and in vivo.
Hypoxic exosomes facilitated ESCC tumor growth, enhanced angiogenesis and promoted metastasis
Also, a significant larger number of lung metastasis nodules was observed in hypo-Exo group than norm-Exo and control group. Immunohistochemistry stain of the lungs showed that the metastasis nodules in hypo-Exo group were more extensive which was consistent with the general observation results (Fig. 5o-r).
Besides, exosomes from normal squamous esophageal epithelial cell line (HET-1A) had no significant effects on ESCC progression. However, exosome release inhibitor, GW4869, suppressed ESCC tumor proliferation or angiogenesis in situ or distant lung metastasis notably in both ECA109 and KYSE410 mice xenograft model.
Microarray analysis of transcriptome in HUVECs treated with hypoxic and normoxic exosomes
Results aforementioned showed that hypoxia facilitate the angiogenic effects of exosomes, hence, we focused on the RNAs which experienced an up-or down-regulation in both norm-Exo group and hypo-Exo group. According to the intersection of transcriptome, we identified 839 mRNAs which down-regulated in both norm-Exo and hypo-Exp groups (Fig. 6c). In this way, 113 up-regulated mRNAs (Fig. 6d), 232 down-regulated lncRNAs (Fig. 6e), 99 up-regulated lncRNAs (Fig. 6f), 692 down-regulated circular RNAs (Fig. 6g) and 86 up-regulated circular RNAs (Fig. 6h) were identified in both norm-Exo and hypo-Exp groups.
Function annotation and signaling pathways analysis of differentially expressed genes
Hypoxia, or low oxygen tension, has emerged as a specific and general feature of the tumor microenvironment which contribute to cancer development and aggressiveness . In oncology, it still remains to be elucidated that how cancer cells adapt to the hypoxia environment. Actually, hypoxic microenvironment induces a series of adaptive mechanisms including phenotypic modulation of stromal cells in the tumor microenvironment that can prolong the survival and mediate the dissemination of malignant cells [31, 32]. Importantly, hypoxia activates the angiogenic signaling pathway and results in the sprouting of blood vessels from the surrounding tissues into the tumor, during which the intercellular communication between cancer cells and endothelial cells is indispensable .
Exosomes are small vesicles of endocytic origin released by most cell types. Exosomes can facilitate eukaryotic intercellular communication under a wide range of normal physiological contexts. In malignancies, this regulatory circuit is co-opted to promote cancer cell survival and outgrowth .
Here, we provide evidence that exosomes from ESCC cells constitute a potent mediator of intercellular communication between cancer and vascular endothelial cells. We firstly characterized the exosomes from ESCC cells and visualized the internalization of these exosomes by HUVECs. Subsequently, we verified that exosomes derived from ESCC cells enhanced HUVECs proliferation through regulating its cell cycle. The invasive ability of HUVECs was also activated by exosomes from ESCC. In vitro and in vivo assay showed that exosomes from ESCC cells significantly promoted the formation of capillary-like structures of HUVECs and improved the microvessel density in transplanted gel plugs from nude mice. Based on these results, we concluded that exosomes from ESCC cells enhanced angiogenesis in TEM.
Moreover, previous studies have demonstrated that hypoxia promotes the release of exosomes by cancer cells and the secreted proteome from hypoxic carcinoma cells are closely associated with exosomes, indicating a potential role of exosomes in regulating the hypoxic response of tumor cells [35, 36]. In the present study, we compared the biological modulatory role of exosomes from ESCC cells cultured in normoxic and hypoxic condition by performing a series function experiments aforementioned. Hypoxia augmented the angiogenic effects of exosomes derived from ESCC and resulted in the enhancement of vascular formation. These results suggested that exosomes might act as a potential regulator which participate in the hypoxia-driven, phenotypic alteration of endothelial cells in TEM and, as a result, affect angiogenesis.
It has been widely accepted that angiogenesis is closely related with tumor growth metastasis . Hence, we further explored the contribution of hypoxic exosomes in ESCC angiogenesis, tumor growth and metastasis. In tumor-bearing mice model, hypoxic exosomes significantly enhanced the ESCC progression by promoting the proliferation of cancer cells, vascular formation and metastasis. Based on the results we mentioned above, we supposed that hypoxia altered the content of exosome secreted by ESCC and enhanced the vascular formation by endothelial cells after internalized these exosomes in TEM. The newly-formed vascular alleviated the hypoxia condition in TEM of ESCC and in turn contribute to the tumor growth. Moreover, increased micro vessel density in TEM facilitated the escape of cancer cells into the bloodstream. Disseminated tumor cells in the blood paved the way for follow-up establishment of metastatic colonies in secondary sites.
On the other side, tumor-derived exosomes are demonstrated to be the major drivers of the pre-metastatic niche. Previous studies demonstrated that exosomes destroyed endothelial barriers and increased vascular permeability which provide an escape route for the tumor cells to enter the circulation . Besides, in the target organ, a microenvironment suitable for tumor metastasis has been created by exosomes for tumor metastasis before cancer cells transfection [39, 40]. Hence, whether hypoxia take an effect on the cancer cells secreted exosomes which change the vascular permeability and result in the pre-metastatic niche formation in ESCC deserves further study.
Furthermore, exosomes carry genetic messages, such as mRNA, miRNA, lncRNA and circular RNA, and can be internalized by the recipient cells . The mRNAs presented in exosomes has been demonstrated to be functional and translatable, which means they are capable of encoding polypeptides and supporting protein synthesis in host cells. Specific protein production may provide the necessary signal(s) to modulate the function of the recipient cells [41, 42]. The regulatory capacity of noncoding RNAs, such as lncRNA and miRNA, in the exosomes have also been found as extensive. Noncoding RNAs in the exosomes may act as ceRNAs and interfere with mRNAs . Hence, in order to explore the mechanisms underlying the phenotypic modulation effects of exosomes in angiogenesis, we performed the comprehensive analyses of the transcriptome in HUVECs after the internalization of hypoxic and normoxic ESCC exosomes.
As we have mentioned above, hypoxia facilitate the angiogenic effects of exosomes, we focused on the RNAs which experienced an up-or down regulation in both of the two groups. We applied gene annotation and pathway enrichment analysis on the mRNAs to identify their functions. Results showed that these mRNAs were significantly enriched in biological processes such as cell proliferation and migration, and pathways such as cell cycle. Hub genes were also identified which were considered as topologically important to the structure of the network and played crucial roles in the biological process. In cell cycle term (GO: 0007049), PLK1, BUB1 and AURKA were identified as hub genes in cell cycle regulation. Previous studies demonstrate that PLK1-mediate the activation of phosphorylates glucose-6-phosphate dehydrogenase which is critical for the promoting the cell cycle progression and tumor growth in liver cancer and cervical cancer . Moreover, BUB1–PLK1 complex mediate the phosphorylation of Cdc20 and inhibit the anaphase-promoting complex or cyclosome (APC/C) which result in the promotion of spindle checkpoint signaling in cervical cancer and osteosarcoma . Silence of AURKA inhibits tumor growth by inducing apoptosis and G2/M cell cycle arrest in human osteosarcoma and breast cancer [46, 47].
VEGFA, CXCL8 and CCL2 were identified as the hub genes in cell migration process (GO: 0016477). It is generally accepted that VEGFA play key roles in angiogenesis . The epithelial-mesenchymal transition effect of VEGFA is also verified in cancer and retinal pigment epithelial cells [49, 50]. CCL2 and CXCL8 induces epithelial-mesenchymal transition in colon cancer and bladder cancer [51, 52]. In view of this, we hypothesized that hypoxia might alter the transcriptome of exosomes from cancer cells and mediated the dysregulation of effector molecules, which triggered a series of cascade reaction, in host cells-HUVECs after internalization. Hence, we next planned to identify the hub genes which induced the angiogenesis-related signaling pathway activation and resulted in phenotype alteration of endothelial cells in the hypoxic TEM of ESCC.
In summary, our data suggests that exosomes in ESCC represents a potentially targetable regulating factor in the hypoxia-driven tumor development. This study sheds light on the possible application of exosomes to fight against cancer progression and metastasis.
YM and ZF contributed to the study design, experiment performance and drafting the article. LD, YZ, YZ, CW, QZ, SY, LC, XZ and XL contributed to the experiment performance and prepared the figures. YW contributed to the experiment performance and revision of the manuscript. All authors read and approved the final manuscript.
This work was supported by Scientific Research Fund of Hebei Health and Family Planning Commission (20181192).
Ethics approval and consent to participate
The Medical Ethics Committees of First Hospital of Qinhuangdao approved this study.
Consent for publication
The authors declare that they have no competing interests.
- 3.Predescu D, Gheorghe M, Boeriu M, Constantin A, Iosif C, Anghel R, et al. Molecular factors and criteria for predicting the response to neoadjuvant treatment in patients with esophageal squamous cell carcinoma (ESCC) -- responder/non-responder. Chirurgia (Bucharest, Romania : 1990). 2012;107(5):583–90.Google Scholar
- 19.Kucharzewska P, Christianson HC, Welch JE, Svensson KJ, Fredlund E, Ringner M, et al. Exosomes reflect the hypoxic status of glioma cells and mediate hypoxia-dependent activation of vascular cells during tumor development. Proc Natl Acad Sci U S A. 2013;110(18):7312–7.CrossRefPubMedPubMedCentralGoogle Scholar
- 24.Mao Y, Fu Z, Dong L, Zheng Y, Dong J, Li X. Identification of a 26-lncRNAs risk model for predicting overall survival of cervical squamous cell carcinoma based on integrated bioinformatics analysis. DNA Cell Biol. 2019.Google Scholar
- 27.Damian S, Andrea F, Stefan W, Kristoffer F, Davide H, Jaime HC, et al. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015;43(Database issue):D447.Google Scholar
- 40.Pasquier J, Thawadi HA, Ghiabi P, Abu-Kaoud N, Maleki M, Guerrouahen BS, et al. Microparticles mediated cross-talk between tumoral and endothelial cells promote the constitution of a pro-metastatic vascular niche through Arf6 up regulation. Cancer Microenvironment. 2014;7(1–2):41–59.CrossRefPubMedPubMedCentralGoogle 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.