Posts Tagged ‘cell & molecular biology and cancer’

A Mechanism of Cancer Metastasis

Larry H. Bernstein, MD, FCAP, Curator



A Protein that Spreads Cancer

Nils Halberg at the University of Bergen has identified a protein that makes it possible for cancer cells to spread

The cells inside a tumour differ a lot. While some remains “good” and do not cause trouble, others become aggressive and starts to spread to other organ sites. It is very hard to predict which cells become aggressive or not.

Nevertheless, by isolating these aggressive cancer cells in in vivotests on animals, Nils Halberg at the Department of Biomedicinet the University of Bergen (UiB) and the researchers Dr. Sohail Tavazoie and Dr. Caitlin Sengelaub at The Rockefeller University have discovered a certain protein (PITPNC1) that characterise aggressive cancer cells.

“We discovered that the aggressive cancer cells that are spreading in colon, breast, and skin cancer contained a much higher portion of the protein PITPNC1, than the non-aggressive cancer cells,” says researcher Nils Halberg of the CELLNET Group at the Department of Biomedicine at UiB.

“This means we can predict which of the cancer cells are getting aggressive and spread, at a much earlier stage than today.”

How cells penetrate tissue

The researcher also discovered that this protein, that characterizes the aggressive cancer cells, has got a very specific function in the process of spreading cancer.

The cancer cells spread from one place in the body to another, through the blood vessel. To get into the blood vessels, the cell needs to penetrate tissue, both when it leaves the tumour and when it is attaching to a new organ.

“The protein PITPNC1 regulates a process whereby the cancer cells are secreting molecules, which cut through a network of proteins outside the cells, like scissors. The cancer cell is then able to penetrate the tissue and set up a colonies at new organ sites,” Halberg explains.

Custom-made therapy

A tumour that is not spreading, is usually not dangerous for the patient if it is removed. The hard part in cancer therapy is when the tumour starts to spread. Guided by the new discoveries, supported by the Bergen Research Foundation´s (BFS) Recruitment Programme, Halberg hopes to contribute to a better treatment of cancer patients.

“If we get to the point where we can offer a custom-made therapy that targets the function of this protein, we might be able to stop it spreading,” says Nils Halberg.



This gene encodes a member of the phosphatidylinositol transfer protein family. The encoded cytoplasmic protein plays a role in multiple processes including cell signaling and lipid metabolism by facilitating the transfer of phosphatidylinositol between membrane compartments. Alternatively spliced transcript variants encoding multiple isoforms have been observed for this gene, and a pseudogene of this gene is located on the long arm of chromosome 1. [provided by RefSeq, May 2012]


Phosphatidylinositol Transfer Protein, Cytoplasmic 1 (PITPNC1) Binds and Transfers Phosphatidic Acid*

Kathryn Garner‡ , Alan N. Hunt§ , Grielof Koster§ , Pentti Somerharju¶ , Emily Groves‡1, Michelle Li‡ , Padinjat Raghu , Roman Holic‡ , and Shamshad Cockcroft‡2
JBC Papers in Press, July 21, 2012,

Background: Phosphatidylinositol transfer protein, cytoplasmic 1 (PITPNC1) (alternative name, RdgB) promotes metastatic colonization and angiogenesis in humans.

Results: We demonstrate that RdgB is a phosphatidic acid (PA)- and phosphatidylinositol-binding protein and binds PA derived from the phospholipase D pathway.

Conclusion: RdgB is the first lipid-binding protein identified that can bind and transfer PA.

Significance: PA bound to RdgB is a likely effector downstream of phospholipase D


PITPNC1 Recruits RAB1B to the Golgi Network to Drive Malignant Secretion

Nils Halberg3,4,, Caitlin A. Sengelaub3, Kristina Navrazhina, Henrik Molina, Kunihiro Uryu, Sohail F. Tavazoie
Cancer Cell 14 Mar 2016; Volume 29, Issue 3:339–353
  • PITPNC1 promotes metastasis by melanoma, breast cancer, and colon cancer cells
  • PITPNC1 recruits RAB1B to the Golgi compartment of the cell
  • Golgi localization of RAB1B enhances vesicular secretion via GOLPH3 recruitment


Enhanced secretion of tumorigenic effector proteins is a feature of malignant cells. The molecular mechanisms underlying this feature are poorly defined. We identify PITPNC1 as a gene amplified in a large fraction of human breast cancer and overexpressed in metastatic breast, melanoma, and colon cancers. Biochemical, molecular, and cell-biological studies reveal that PITPNC1 promotes malignant secretion by binding Golgi-resident PI4P and localizing RAB1B to the Golgi. RAB1B localization to the Golgi allows for the recruitment of GOLPH3, which facilitates Golgi extension and enhanced vesicular release. PITPNC1-mediated vesicular release drives metastasis by increasing the secretion of pro-invasive and pro-angiogenic mediators HTRA1, MMP1, FAM3C, PDGFA, and ADAM10. We establish PITPNC1 as a PI4P-binding protein that enhances vesicular secretion capacity in malignancy.



Cancerous Conduits

Metastatic cancer cells use nanotubes to manipulate blood vessels.

By Amanda B. Keener | April 1, 2016

MAKING CONTACT: Breast cancer cells (white arrows) in culture deliver microRNAs to endothelial cells through filamentous nanotubes (yellow arrow).



The paper
Y. Conner et al., “Physical nanoscale conduit-mediated communication between tumour cells and the endothelium modulates endothelial phenotype,” Nat Commun, 6:8671, 2015.

Branching Out
Harvard bioengineer Shiladitya Sengupta and his team were establishing a culture system to model the matrix and blood vessel networks that surround tumors when they found that human breast cancer cells spread out along blood vessel endothelial cells rather than form spheroid tumors as expected. Taking a closer look using scanning electron microscopy, they spied nanoscale filaments consisting of membrane and cytoskeletal components linking the two cell types.

Manipulative Metastases
These cancer cell–spawned nanotubes, the team discovered, could transfer a dye from cancer cells to endothelial cells both in culture and in a mouse model of breast cancer metastasis to the lungs.The cells also transferred microRNAs known to regulate endothelial cell adhesion and disassociation of tight junctions, which Sengupta speculates may help cancer cells slip in and out of blood vessels. This study is the first to suggest a role for nanotubes in metastasis.

Breaking the Chain
Sengupta’s team then used low doses of cytoskeleton-disrupting drugs to block nanotube formation. Emil Lou, an oncologist at the University of Minnesota who studies nanotubes in cancer and was not involved in the study, says this approach is a “good start,” though such drugs would not be used in human patients because they are not specific to nanotubes.

In the Details
Lou says the study emphasizes the importance of understanding interactions between tumors and their surrounding tissues on a molecular level. Going forward, Sengupta plans to study how the tubes are formed in melanoma as well as breast and ovarian cancers to try to identify other drug targets.


Physical nanoscale conduit-mediated communication between tumour cells and the endothelium modulates endothelial phenotype

Yamicia ConnorSarah TekleabShyama NandakumarCherelle Walls,….., Bruce ZetterElazer R. Edelman & Shiladitya Sengupta
Nature Communications6,Article number:8671

Metastasis is a major cause of mortality and remains a hurdle in the search for a cure for cancer. Not much is known about metastatic cancer cells and endothelial cross-talk, which occurs at multiple stages during metastasis. Here we report a dynamic regulation of the endothelium by cancer cells through the formation of nanoscale intercellular membrane bridges, which act as physical conduits for transfer of microRNAs. The communication between the tumour cell and the endothelium upregulates markers associated with pathological endothelium, which is reversed by pharmacological inhibition of these nanoscale conduits. These results lead us to define the notion of ‘metastatic hijack’: cancer cell-induced transformation of healthy endothelium into pathological endothelium via horizontal communication through the nanoscale conduits. Pharmacological perturbation of these nanoscale membrane bridges decreases metastatic foci in vivo. Targeting these nanoscale membrane bridges may potentially emerge as a new therapeutic opportunity in the management of metastatic cancer.


Metastasis is the culmination of a cascade of events, including invasion and intravasation of tumour cells, survival in circulation, extravasation and metastatic colonization4. Multiple studies have reported a dynamic interaction between the metastatic tumour cell and the target organ, mediated by cytokines4, 12 or by exosomes that can prime metastasis by creating a pre-metastatic niche13. Interestingly, the interactions between cancer cells and endothelium in the context of metastasis, which occurs during intravasation, circulation and extravasation, remains less studied. Cancer cell-secreted soluble factors can induce retraction of endothelial cells and the subsequent attachment and transmigration of tumour cells through the endothelial monolayers14, 15. Recently, studies indicate a more intricate communication between cancer cells and the endothelium. For example, a miRNA regulon was found to mediate endothelial recruitment and metastasis by cancer cells16. Similarly, exosome-mediated transfer of cancer-secreted miR-105 was recently reported to disrupt the endothelial barrier and promote metastasis17. We rationalized that a better understanding of cancer–endothelial intercellular communication, primarily during extravasation, could lead to novel strategies for inhibiting metastasis18.

Recently, nanoscale membrane bridges, such as tunnelling nanotubes (TNTs) and filopodias, have emerged as a novel mechanism of intercellular communication19. For example, specialized signalling filopodia or cytonemes were recently shown to transport morphogens during development20. Similarly, TNTs, which unlike filopodia have no contact with the substratum21, were shown to facilitate HIV-1 transmission between T cells, enable the spread of calcium-mediated signal between cells and transfer p-glycoproteins conferring multi-drug resistance between cancer cells22, 23, 24, 25. TNTs were also recently implicated in trafficking of mitochondria from endothelial to cancer cells and transfer miRNA between osteosarcoma cells and stromal murine osteoblast cells, and between smooth muscle cells and the endothelium26, 27, 28. However, whether similar intercellular nanostructure-mediated communication can be harnessed by cancer cells to modulate the endothelium is not known.

Here we report that metastatic cancer cells preferentially form nanoscale intercellular membrane bridges with endothelial cells. These nanoscale bridges act as physical conduits through which the cancer cells can horizontally transfer miRNA to the endothelium. We observe that the recipient endothelial cells present an miRNA profile that is distinct from non-recipient endothelial cells isolated from the same microenvironment. Furthermore, the co-cultures of cancer and endothelial cells upregulate markers associated with pathological endothelium, which is inhibited by pharmacological disruption of the nanoscale conduits. Additionally, the pharmacological inhibitors of these nanoscale conduits can decrease metastatic foci in vivo, which suggests that these nanoscale conduits may potentially emerge as new targets in the management of metastatic cancer.


Figure 1: Nanoscale structures physically connect metastatic cells and the endothelium

Nanoscale structures physically connect metastatic cells and the endothelium.

(a) Representative image of MDA-MB-231 cancer cells exhibiting an invasive phenotype in the presence of preformed endothelial tubes in co-culture. (b) Representative image of a mammosphere typically formed by MDA-MB-231 cells when cultured on 3D tumour matrix in the absence of endothelial cells. MDA-MB-231 cells were loaded with CFSE. Actin was labelled with rhodamine phalloidin and nuclei were counterstained with DAPI. (c) A representative SEM of epithelial (EPI) MDA-MB-231 cells aligning on HUVEC (ENDO) tubules in the co-culture. Lower panel shows higher magnification. (d) SEM image reveals nanoscale membrane bridges connecting (nCs) metastatic breast cancer (EPI) cells and endothelial vessels (arrows). (e) A representative transmission electron micrograph shows intercellular connectivity through the nanoscale membrane bridge between MDA-MB-231 and an endothelial cell. (f) A cartoon represents the types of homotypic and heterotypic intercellular nanoscale connections that an epithelial cell may form in the presence of endothelial tubules. Highly metastatic (MDA-MB-468, MDA-MB-231 or MDA-MB-435) or low metastatic (MCF7 and SkBr3) cancer cells were co-cultured with the endothelial tubes. Normal HMECs were used as control. Graphs show percentage of total population of epithelial cells that exhibit either homotyptic (Epi–Epi) or heterotypic (Epi–Endo) nanoscale connections and (g) average number of nanoscale connections formed per cell. Quantification analysis was done on >300 cells of each cell type. Data shown are mean±s.e.m. (n=6 replicates per study, with 2–3 independent experiments). **P<0.01, ***P<0.001 (analysis of variance followed by Bonferroni’s post-hoctest).

The nanoscale bridges are composed of cytoskeletal elements

Figure 3: Structure and function of the heterotypic intercellular nanoscale membrane bridges.

a) Representative images show the heterotypic nanoscale membrane bridges are composed of both F-actin and α/β-tubulin cytoskeletal components. Co-cultures were stained with α/β-tubulin antibody (green) and phalloidin (purple) to label actin, and counterstained with DAPI (nuclear)+WGA (plasma membrane) (blue). Endothelial cells were labelled with DiL-Ac-LDL (red). (b) Mathematical modelling of the structure of the nanoscale connections. The physical properties of actin filaments necessitate microtubules for projections of certain length scales. The maximum projection length for a given minimum diameter at the buckling limit is plotted for actin-only nanoscale structures (purple line). This curve is overlaid with the experimental length and diameter measurements (red dots) from the observed thin projections measured in these studies. Projections containing only actin or projections containing both actin and tubulin can exist to the right of the curve (purple line). However, actin-only projections cannot exist to the left of the curve (green region). (c) The effect of incorporating tubulin in these projections. The maximum length is plotted against the minimum diameter for varying fractions of tubulin incorporated in the nanoscale projection. Addition of microtubules to the projections increases the overall flexural rigidity, shifting the curves left of the actin-only limit (purple line), thus allowing for longer and thinner nanoscale connections. However, owing to the larger radius of microtubules (4 × radius of actin filaments), there is an optimal fraction of tubulin (green line) that can be incorporated into the projection before the effect is reversed. (d) The optimal fraction of microtubules is about 6.6% (red dashed line) to maximize nanostructure flexural strength, while minimizing thickness. (e) Representative confocal image shows the presence of myosin V motor proteins within the intercellular nanostructure (inset shows higher magnification). Scale bar 10μm.

Nanoscale bridges act as conduits for communication

Figure 4: The nanoscale membrane bridges act as conduits for intercellular communication between cancer and endothelial cells.

(a) Confocal image of nanoscale membrane bridge-mediated transfer of cytoplasmic contents. CFSE (green)-loaded MDA-MB-231 cells were co-cultured with the Dil-Ac-LDL (red)-labelled HUVECs. Transfer of the CFSE dye was observed after 24-h co-culture. CFSE dye can be seen within HUVEC cells (yellow arrow). Tumour cells can form a nanobridge with a distal endothelial cell (EC1) than an endothelial cell (EC2) in close proximity. (b,c) Cartoon shows the experimental design, where dual cultures control for vesicle-mediated intercellular transfer. FACS plot show gating for sorting endothelial cells from the co-cultures using dual staining for DiI-Ac-LDL and PECAM-1, and then quantification for CFSE transfer in the isolated endothelial cells. (d) Graph shows quantification of FACS analysis, highlighting increased transfer of CFSE to endothelial cells in the co-culture. (N>100,000 events, n=36 replicates, 3 replicates per study). (e) Graph shows the temporal kinetics of nanoscale connection-mediated intercellular transfer of CFSE from MDA-MB-231 cells to the endothelium (n=2 studies, 3 replicates per study). (f) Effect of small molecule inhibitors of cytoskeletal components on membrane nanobridges. (g) Graphs show treatment with vehicle (control) or a low-dose combination of docetaxel and cytochalasin do not affect the exosome shedding (n=2 independent studies). (h,i) Graphs show the effect of pharmacological inhibitors on the formation of heterotypic and homotypic nanoscale bridges (arrows). (n=2 studies, 6 replicates per study). (j) Graph shows the effect of pharmacological inhibitors on intercellular transfer of CFSE to endothelial cells from cancer cells (n=10 studies, 3 replicates per study). Data shown are mean±s.e.m. (*P<0.05,**P<0.01, ****P<0.001, analysis of variance followed by Bonferroni’s post-hoc test).

Effect of pharmacological inhibition of nanoscale bridges

Nanobridges transfer miRNA from cancer cells to endothelium

Figure 5: The nanoscale membrane bridges act as conduits for intercellular transfer of miRNA between cancer and endothelial cells.

Representative confocal images show the transfer of Cy3-labelled miRNA from MDA-MB-231 cells (EPI) to endothelial cells (ENDO) at (a) 24h and (b) 36h of co-culture. Alexa Fluor 488-Ac-LDL (green)-labelled endothelial cells were co-cultured with Cy3-labelled miRNA-transfected MDA-MB-231. Co-cultures were counterstained with phalloidin (purple) and DAPI+WGA (blue). A 3D visualization shows the localization of miRNA within the nanoscale connections (white arrows), which act as conduits for horizontal transfer of miRNAs to endothelial cells. (c) Schema shows quantification of Cy3-labelled control miRNA and Cy3-labelled miR132 transfer between cancer cell and endothelium using flow cytometry. Endothelial cell populations were isolated from the co-cultures and percentage of miRNA+ve cells was determined. Dual cultures in Boyden chambers were included as controls. (d) Graph shows the effect of pharmacological disruption of nanoscale conduits on miRNA transfer. (e) Schema shows experimental design for reverse transcriptase–PCR-based detection of transferred miR-132 in endothelial cells under different experimental conditions. MDA-MB-231 cells transfected with miR-132 and α-miR-132 were co-cultured with endothelial tubes. FACS-isolated endothelial cell populations were analysed for the expression of miR-132. (f) Graph shows miR-132+ve cell populations (solid red) show 5 × increase compared with miR-132−ve populations (solid blue) (P<0.0001), whereas anti-miR-132+ve cells (striped red) show 26 × decrease in miR-132 expression (P<0.0001) compared with α-miR-132−ve cells (striped blue). Direct transfection of miR-132 (black) and α-miR-132 (light blue) in endothelial cells is used as positive and negative controls, respectively. Upregulation of miR-132 from baseline was observed in dual culture (solid green), which could be inhibited with anti-miR-132 (striped green). MiR-132 levels are increased compared with dual only in those cells that are positive for intercellular transfer. Fold change was determined compared with endothelial cell transfection with control miRNA (grey). (g) FACS analysis shows nanoscale bridges-mediated transfer of miRNAs leads to changes in p120RasGAP and pAkt (S473) expression downstream of the miR-132 pathway in endothelial cell populations isolated from co-cultures. (h) Graphs show p120RasGAP expression is decreased in the miR-132+ve cell populations and increased in the α-miR-132+ve cell populations, while further downstream miR-132 positively regulates pAkt expression. Data shown are mean±s.e.m. (N=2–5 independent studies, with 3 replicates per study,*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, analysis of variance followed by Bonferroni’s post-hoctest).

Nanobridge-mediated transfer alters endogenous miRNA profile

Figure 6: Cancer cell–endothelial intercellular transfer alters the endogenous miRNA profile and phenotype of recipient endothelial cells.


The complexity of regulatory tumour parenchyma–endothelial communication is increasingly being unravelled7, 50. The altered phenotypic behaviour of the metastatic cancer cells in the presence of endothelial cells observed in this study, instead of forming classical mammospheres, is consistent with the emerging paradigm of modulatory tumour parenchyma–stroma communication and the creation of a pre-metastatic niche. Indeed, a recent study proposed the concept of the formation of a pre-metastatic niche mediated via metastatic cell-secreted exosomes, leading to vascular leakiness at the pre-metastatic sites13. Here we demonstrate that cancer cells form nanoscale membrane bridges, which can act as conduits for horizontal transfer of miRNA from the cancer cells to the endothelium, switching the latter to a pathological phenotype. Our findings reveal that the ability to form the nanoscale conduits with endothelial cells correlates with the metastatic potential of the cancer cell, and that the pharmacological perturbation of these nanoscale connections can lead to a reduction in the metastatic burden in experimental metastasis models. Together, our studies shed new insights into the tumour parenchyma–endothelial communication, adding depth to the emerging paradigm of the ability of a cancer cell to ‘hijack’ a physiological stromal cell for self-gain13.

Indeed, exosomes have emerged as an extensively studied mechanism of horizontal intercellular transfer of information51. However, a key distinction exists between the exosome-mediated versus the nanoscale membrane bridge-mediated intercellular communication. Although the former is stochastic, that is, it is unlikely the cancer cell has control over which cell will be targeted by a secreted exosome, the communication via nanoscale membrane bridges is deterministic, that is, the cancer cell can connect to a specific endothelial cell, which could be further away than the most proximal endothelial cell.

Although the aim of this study was to study the nanoscale membrane bridges as a mode of horizontal transfer of miRNAs from the metastatic cancer cells to the endothelium, and not to characterize a specific miRNA that are implicated in metastasis, many of the miRNAs, which were differentially regulated in the recipient endothelial cells, have previously been shown to regulate metastasis (Supplementary Discussion).


Read Full Post »