Immunobiology 219 (2014) 308–314
Contents lists available at ScienceDirect
Immunobiology journal homepage: www.elsevier.com/locate/imbio
Lung cancer neovascularisation: Cellular and molecular interaction between endothelial and lung cancer cells Sabine Kaessmeyer a,∗ , Kanti Bhoola b,1 , Svetlana Baltic b , Philip Thompson b,2 , Johanna Plendl a,2 a
Institute of Veterinary Anatomy, Department of Veterinary Medicine, Freie Universität Berlin, Koserstraße 20, 14195 Berlin, Germany Lung Institute of Western Australia and Centre for Asthma, Allergy and Respiratory Research, The University of Western Australia, Nedlands, WA 6009, Australia b
a r t i c l e
i n f o
Article history: Received 14 June 2013 Accepted 14 November 2013 Available online 22 November 2013 Keywords: CD31 Co-culture Endothelial cells Immunocytochemistry Immunohistochemistry Lung adenocarcinoma Lung squamous carcinoma
a b s t r a c t Background: Novel vascular-independent conduits have been observed in some cancers. These have been variously described as vasculogenic mimicry, mosaic vessel formation, vascular co-option and intratumour embryonic-like vasculogenesis. Despite lung cancer being the most common cancer worldwide, there is little information on its neovascularisation or the pathways involved. Methods: An in vitro model involving co-cultures of microvascular lung endothelial cells and squamous or adenocarcinoma lung cancer cells was developed to assess their angiogenic interaction. Cells were incubated and examined by phase contrast microscopy and by immunocytochemistry in both monoand co-cultures. Cultured cells and lung cancer tissue sections were assessed for new tumour vessel formation, expression of the endothelial marker CD31 and morphology. Results: Lung tumour cells and endothelial cells interacted morphologically via pseudopodia and used alternative pathways to generate new vessels. Co-culturing microvascular endothelial and squamous carcinoma cells led to endothelial cells surrounding tumour cells and the tumour cells being incorporated into vessel walls. Co-culturing endothelial and adenocarcinoma cells resulted in cellular contact and the formation of tumour cell bridges around clusters of endothelial cells. These adencocarcinoma cells became strongly positive for CD31. Tumour tissue section studies supported the in vitro findings. Conclusion: Lung carcinoma cells when co-cultured with lung endothelial cells modify their cellular and molecular features that encourage alternative means of providing blood supply. The mechanisms underpinning these non-angiogenic processes need to be further investigated and should be considered when anti-tumour therapeutic interventions are being considered. © 2013 Elsevier GmbH. All rights reserved.
Introduction
Abbreviations: CD, cluster of differentiation; EC, endothelial cells; NSCLC, nonsmall cell lung cancer; HMVEC-L, human microvascular endothelial cells-lung; rhFGF, recombinant human fibroblast growth factor; VEGF, vascular endothelial growth factor; GA, gentamicin/amphotericin; FBS, fetal bovine serum; H520, human lung squamous cell carcinoma cells 520; ATCC, American Type Culture Collection; H2126, human lung adenocarcinoma cells; DMEM, Dulbecco’s modified Eagle medium; PBS, phosphate buffered saline; PFA, paraformaldehyde; TBS, Tris-buffered saline. ∗ Corresponding author at: Institute of Veterinary Anatomy, Department of Veterinary Medicine, Freie Universität Berlin, Koserstraße 20, D-14195 Berlin, Germany. Tel.: +49 03083853558; fax: +49 03083853480. E-mail address: [email protected] (S. Kaessmeyer). 1 Institutes of Anatomy, Histology & Pathology, and Physiology, Universidad Austral de Chile, Valdivia Box 567, Chile 2 Equal last author. 0171-2985/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.imbio.2013.11.004
The formation of blood vessels is initiated by two distinct processes, namely vasculogenesis and angiogenesis. Vasculogenesis is the in situ differentiation of endothelial precursor cells (angioblasts) into endothelial cells (EC), which then re-assemble into primitive vascular plexuses. Angiogenesis generally involves the expansion of a primitive vascular network into a complex one (Hillen and Griffioen, 2007; Patan, 2000; Makanya et al., 2009) by growth of endothelial sprouts from pre-existing blood vessels through migration, proliferation, three-dimensional organisation and tube formation of ECs. Whereas formation of new blood vessels plays a major role during prenatal development, in adults, blood vessel growth is linked primarily to tissue repair and clinical disorders such as tumour growth (Vacca and Ribatti, 2011). The growth of solid tumours beyond 1–2 mm in diameter requires the induction of new blood vessel formation (Bergers and Benjamin, 2003). It has
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
been assumed that tumour vascularisation can be explained by angiogenesis. However, in the last decade, novel angiogenesisindependent pathways have been observed in the blood supply of certain tumours. Such alternative processes, found particularly in aggressive tumours, are designated variously as vasculogenic mimicry, mosaic vessel formation, vascular co-option and intratumour embryonic-like vasculogenesis (Hillen and Griffioen 2007; Bussolati et al., 2011; Chang et al., 2000; Ackermann et al., 2012). Vasculogenic mimicry, first described by Maniotis et al. (1999) in uveal melanoma, is the formation of a complete capillary-like network that is comprised only of tumour cells instead of vascular ECs. In tumours such as colon carcinoma, mosaic vessels have been described, in which, attached to the luminal surface of the vascular channels are both ECs as well as non-ECs; the latter, lacking expression of specific endothelial cell markers (Chang et al., 2000; Ackermann et al., 2012). Tumour cells that grow along pre-existing vessels but without evoking an angiogenic response, referred to as vessel co-option, have been found in cerebral glioblastoma, breast adenocarcinoma, and in melanomas (Bartha and Rieger 2006; Holash et al., 1999). Finally, de novo generation of tumour vessels can arise from the differentiation of stem and progenitor cells of hematopoietic origin or from those resident in tissues and participate in tumour progression (Bussolati et al., 2011). Lung cancer is the leading cause of cancer worldwide (Kimman et al., 2012) and non-small cell lung cancers (NSCLC; subtypes squamous cell carcinoma and adenocarcinoma), account for approximately 80–90% of lung cancers (Yano et al., 2011). Despite this, there is a lack of information on their neovascularisation, especially with respect to alternative cellular processes for vessel formation (McClelland et al., 2007; Passalidou et al., 2002). Tumour neovascularisation is a complex process based upon a sequence of interactions between tumour cells and ECs (Levine et al., 2001; Witz, 2009). In vitro culture models allow analysis of each step involved in vascular growth including the interaction between endothelial and tumour cells. As such an in vitro lung cancer vascular model should help identify the specific steps and mechanisms involved (Auerbach et al., 2003; Kaessmeyer and Plendl, 2009; Kassmeyer et al., 2009; De Spiegelaere et al., 2012; Sievers et al., 2011; Bahramsoltani et al., 2009). The aim of the current study was to develop an in vitro model to investigate the vascular and molecular interactions that occur between lung endothelial and lung carcinoma cells thereby providing evidence for recruitment of ECs and the induction of new blood vessel formation by lung carcinomas. A morphological study was designed to monitor co-cultures of microvascular lung ECs with two types of NSCLC. Materials and methods The study was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital Nedlands Western Australia, Australia. Cell culture Cells and culture media Lung derived normal human microvascular ECs (HMVEC-L, Lonza, Walkersville Inc., Walkersville, USA) were incubated in basic EC culture medium (EGM-2-MV; Lonza) supplemented according to the supplier’s instructions with rhFGF, VEGF, vitamin C, GA-1000, hydrocortisone and FBS. Human lung squamous cell carcinoma cells (H520, ATCC) and human lung adenocarcinoma cells (H2126, ATCC) were cultured in DMEM with 10% FBS, 1% l-glutamine and 1% penicillin/streptomycin (10,000 U/ml) (all from Sigma–Aldrich, Taufkirchen, Germany; and referred subsequently as ‘tumour cell
309
Table 1 Experimental design for culture and co-culture time-points: tumour cells, adapted to culture medium of endothelial cells, were added at a concentration of 2 × 104 cells to the lung endothelial cells (4 × 104 cells seeded). Human lung microvascular endothelial cells (HMVEC-LV) Time in culture
Duration of co-culture of H520 tumour cells and HMVEC-LV
Duration of co-culture of AC H2126 tumour cells and HMVEC-LV
28 Days 28 Days 28 Days 28 Days
24 h 7 Days 14 Days 32 Days
24 h 7 Days 14 Days 32 Days
medium’). Cells were incubated in a humidified atmosphere (37 ◦ C, 5% CO2 ) and the medium was replaced every 2–3 days. Priming of lung tumour cells Adaptation (priming) of the tumour mono-cell cultures to the HMVEC-L cell medium was carried out over a 48 hour period in steps that sequentially involved mixing fresh tumour cell medium with 25%, 50%, 75% and finally 100% of HMVEC-L cell medium. Endothelial cell culture HMVEC-L cells were seeded in 24-well plates from Iwaki (Tokyo, Japan) onto round glass cover slips (12 mm diameter, Roth, Karlsruhe, Germany) which were gelatine coated (Difco Laboratories, Detroit, USA, 1.5% in PBS). Optimal seeding density of 4 × 104 cells per well resulted in their adherence and formation of an even monolayer. Direct co-culture design (Fig. 1) The primary co-culture design allowed HMVEC-L cells to be incubated with the primed lung H520 or H2126 tumour cells (each trypsinised, seeding density: 2 × 104 ) for varying times. On the day of co-culturing, the HMVEC-L cells had already been in culture for 28 days (Stage 3, Table 2). The precise time schedules for the co-cultures are shown in Table 1. At the end of each co-culture incubation period, cells were rinsed, fixed and labelled with CD31 for subsequent immunocytochemistry. Phase contrast microscopy of cultured cells Light micrographs were taken with a Zeiss Axiovert 25 inverted microscope (Zeiss MicroImaging GmbH, Jena, Germany) and a 000610 video camera INTEQ (INTEQ, Berlin, Germany) using Zeiss Axiovision image-processing software. Lung tissue samples Archival lung carcinoma specimens were obtained from the Lung Institute of Western Australia and PathWest Inc. The tissue samples were fixed in 4% PFA, paraffin embedded, and cut into 5 m sections. Multiple tissue sections of resected normal lung, adenocarcinoma (1 female, 1 male) and squamous cell carcinoma (2 males) were stained by immunohistochemistry as described below. Immunohistochemistry Antigen retrieval was achieved by heating tissue slides in citrate buffer at 95 ◦ C in a water bath and blocking of endogenous peroxidase activity with TBS-buffer with 0.6% (v/v) H2 O2 , followed by washing in TBS-buffer plus 0.01% polysorbate 20 (Tween20, Sigma–Aldrich). The slides were pre-incubated with 20% (v/v) donkey IgG serum in buffer A, and finally incubated with CD31 overnight at 4 ◦ C (see Table 3).
310
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314 Table 2 Staging of cellular morphology during capillary-like tube formation of human lung microvascular endothelial cells in vitro. Stages
Lung endothelial cells
Time points
Stage 0 Stage 1
Proliferation Non-confluent monolayer with cell-free areas of fusiform cells Development of long cellular extensions Connecting 3-dimensional extensions Formation of 3-dimensional network of capillary-like structures
Days 0–7 Days 7–14
Stage 2 Stage 3 Stage 4
Days 14–28 Days 28–42 Days 42–60
Germany), a digital camera (DS-Ri1, Nikon, Düsseldorf, Germany) and a PC-based laboratory imaging programme (NIS-Elements, Version 3.0, Nikon, Düsseldorf, Germany). Analysis of the cellular morphology observed during capillary-like tube formation of ECs in vitro was classified as shown in Table 2. Results Mono-cell cultures
Fig. 1. Experimental design of direct co-culture method. (a) Adherent endothelial cells (blue); medium (pink). (b) Tumour cells (green) in suspension adapted to medium of endothelial cells. (c) Direct co-cultures. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Immunocytochemistry Cultured cells were washed with PBS and fixed in methanol/ acetone (1:1) at 4 ◦ C, washed with PBS and immersed in protein block, serum-free (DAKO Diagnostika, Hamburg, Germany). After removal of the blocker, cells were incubated with CD31 overnight at 4 ◦ C.
Human lung microvascular endothelial cells (HMVEC-L) HMVEC-L cells formed a three-dimensional network of capillary-like structures over a period of 42 days. During the first 14 days of incubation cells grew as a sub-confluent monolayer with clear spaces between individual cells (Fig. 3a, phase contrast microscopy). During the next 14 days cells developed long cellular connecting extensions. Three-dimensional extensions were observed in the time frame of 28 until 42 days in culture. Further expansion and connection of three-dimensional extensions resulted in a cellular network which developed into tube-like structures up to day 60, while the underlying network of cells on the bottom of the culture plate disintegrated. At all monitored time points the HMVEC-L cells immunolabelled positively for CD31 (Fig. 2a – Stage 2 as described in Table 2 and Fig. 2b – Stage 4). Fig. 3a displays the HMVEC-L cells at stage 1.
Immunodetection Slides and cells were incubated with horseradish peroxidase labelled secondary antibody for 30 min at room temperature. Labelling was visualised by diaminobenzidine (Sigma–Aldrich). Nuclei were counterstained with hemalaun (Fluka-Chemie, Buchs, Switzerland). Negative controls were performed in which the primary antibody was replaced by (a) buffer and (b) non-immune serum (Table 3). Light microscopy of immunostained cells Analysis of the immunostained cells was digitally recorded employing a light microscope (Axioskop, Carl Zeiss, Oberkochen,
Human lung squamous cell carcinoma cells (H520) H520 cells grew in an adherent monolayer and were heterogeneous in appearance. Cells were round or polygonal in shape with varying cell diameter and nuclear-cytoplasmic ratio. Only a small number of tumour cells showed bipolar or spindle-shaped forms. The monolayer never developed into complete confluency as shown in Fig. 3b. Human lung adenocarcinoma cells (H2126) The cellular morphology was more homogenous in appearance than in H520 cells. The majority of cells were polygonal in shape, creating a cobblestone like monolayer (Fig. 3c).
Fig. 2. Light microscopy of human lung endothelial cells, immunostained with CD31. (a) Long cellular extensions (stage 2) light microscopy. (b) Network of capillary-like structures (stage 4) light microscopy.
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
311
Fig. 3. Phase contrast micrographs of human lung endothelial cells and carcinoma cells in vitro. (a) Human lung endothelial cells (HMVEC-L), large cells, cell-free areas (stage 1). (b) Human lung squamous cell carcinoma (H520). (c) Human lung adenocarcinoma (AC H2126).
Co-cultures HMVEC-L cells were co-cultured with either the squamous cell carcinoma or the adenocarcinoma cells as per the research plan (Table 1 and Fig. 1).
Co-cultures of HMVEC-L (cultured 28 days) with H520 cells for 24 h, 7 days, 14 days and 32 days Whereas the large HMVEC-L cells immunolabelled positively with CD31, the small and polygonal tumour cells showed no immunoreactivity. After 24 h, both cell types started proliferating and formed pseudopodia which made contact with each other (data not shown).
Fig. 4. Light microscopy of co-cultures, immunostained with CD31, HMVEC-L (28 days in vitro) co-cultured with H520 cells. (a) After 7 days of co-culturing, HMVEC-L form a fine network around H520 tumour cells. (b) After 14 days of co-culturing, H520 line along HMVEC-L. (c) After 32 days of co-culturing, CD31 positive network of co-cultured endothelial and tumour cells over a monolayer of CD31 negative tumour cells.
Table 3 Immunohistochemistry – antibodies, dilutions and buffer A ingredients. Primary antibody Secondary antibody Negative serum control Buffer A
Cell culture: 1:40 in 0.01 M PBS pH 7.4 + 2% donkey serum CD31 (PECAM-1) monoclonal mouse anti-human CD31, AB9498, Tissue sections: 1:40 in buffer A + 2% donkey serum Abcam, Cambridge, MA, USA Cell culture: 1:80 in 0.01 M PBS pH 7.4 + 2% donkey serum IgG-HRP-conjugated donkey anti-mouse, ZMH2162, Linaris Tissue sections: 1:40 in buffer A + 2% donkey serum Biologische Produkte, Wertheim/Germany Cell culture: 1:40 in 0.01 M PBS pH 7.4 + 2% donkey serum Mouse-IgG1 serum, Item X0931, DAKO Diagnostika, Hamburg, Tissue sections: 1:40 in buffer A + 2% donkey serum Germany 0.05 M Tris–HCl pH 7.6 + 0.9% w/v NaCL + 0.66 m M MgCl2 + 1% w/v BSA + 0.1% w/v gelatine + 0.01% v/v Tween20
312
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
Fig. 5. Light microscopy of co-cultures, immunostained with CD31, HMVEC-L (28 days in vitro) co-cultured with H2126 cells. (a) After 14 days of co-culturing, HMVEC-L bridge above H2126. (b) After 14 days of co-culturing, bridging H2126 cells are positive for CD31, non-bridging H2126 cells stay negative for CD31.
After 7 days of co-culture both cell types continued to proliferate and could be clearly distinguished by their morphology. Tumour cells did not overgrow HMVEC-L cells. Three-dimensional and finely branching long pseudopodia of the HMVEC-L cells established contact with the H520 tumour cells. The HMVEC-L cells and their pseudopodia which immunostained positively with CD31 seemed to invade groups of adjacent tumour cells and to form a fine network around single tumour cells (Fig. 4a). After 14 days in co-culture HMVEC-L cells which continued to stain positively for CD31, were lined by finger-like formations of elongated H520 tumour cells (Fig. 4b). These tumour cells which had positioned themselves along the HMVEC-L cells now stained positively for CD31 (Fig. 4b). By day 32 a CD31 positive network of co-cultured endothelial and tumour cells had grown over a monolayer of CD31 negative tumour cells (Fig. 4c). In mono-cell cultures of tumour cells, which had been already adapted to HMVEC-L cell medium, no such growth pattern was observed.
Co-cultures of HMVEC-L (cultured 28 days) with H2126 cells for 24 h, 7 days, 14 days and 32 days Whereas HMVEC-L cells immunostained positively, the smaller, roundish tumour cells showed no immunoreactivity with CD31. Both cell types developed cellular extensions, which made contact with each other after 24 h (data not shown). After 7 days in co-culture tumour cells continued to proliferate but they did not overgrow HMVEC-L cells (data not shown). Narrow strands of tumour cells formed bridges over islands of HMVEC-L cells. After 14 days many more strand-like bridges of tumour cells were observed which were orientated perpendicularly to the axis of the HMVEC-L cells. Only bridging tumour cells, tumour cells in close proximity to HMVEC-L cells, and HMVEC-L cells themselves immunostained positively with CD31 (Fig. 5a, b). By the 32nd day of co-culturing, a CD31 positive capillary-like network of endothelial and tumour cells had developed over a monolayer of CD31 negative tumour cells (data not shown). In mono-cell cultures of tumour cells, which had been adapted to HMVEC-L cell medium, no differentiation as described above was observed.
Squamous cell carcinoma tissue Most ECs of blood vessels within the tumour were spread out and were CD31 immuno-positive. Tumour cells seemed to line along vascular channels. However, in specific areas of blood vessels single or even a few (5–7) CD31 negative cells seemed to have been incorporated into the lining of the vessel wall (Fig. 7a, b). Adenocarcinoma tissue In most areas of the tumour, ECs immunostained positively with CD31 (Fig. 8a, b). However, in some areas clusters of tumour cells also immunolabelled positively for CD31. Controls No immunolabelling was observed in the absence of the primary antibody or its replacement by non-immune serum. Discussion Tumour cells act as angiogenic inducers by secreting vascular endothelial growth factors. They are able to transform the normally quiescent vascular ECs through the so called “angiogenic switch” into a morphologically altered and activated state (Bergers and Benjamin, 2003; Bach et al., 2007; Rocha and Adams, 2009). A somewhat unexpected but important finding was the development of endothelial pseudopodia that appeared to invade groups of tumour cells and form networks around tumour cells particular in the case of the squamous carcinoma. This was the opposite cell behaviour to what was expected. Although it is possible this is an in vitro phenomenon, this finding was however supported
CD31 immuno-reactive cells in normal and carcinoma lung tissue sections Normal human lung tissue ECs of small capillaries near alveolar spaces as well as larger blood vessels in stromal connective tissue immunostained positively for CD31 (Fig. 6).
Fig. 6. Light microscopy of healthy human lung tissue sections immunostained with CD31, endothelium of large and small vessels is CD31 positive.
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
313
Fig. 7. Light microscopy of human lung tissue sections with lung squamous cell carcinoma immunostained with CD31. (a) Endothelium is CD31 positive, tumour cells line along vessels, intermittently tumour cells instead of ECs as inner vascular layer (arrow). (b) Endothelium is CD31 positive, tumour cells line along vessels, intermittently tumour cells instead of ECs as inner vascular layer (arrows).
by Elzarrad et al., (2009) who has shown that pulmonary ECs play an active role in tumour metastasis by building intercellular connections with tumour cells and growing with them to create new vessels in a tumour environment of oxygen deficiency indicating that our finding is not exclusively an in vitro phenomenon. Furthermore, endothelial cells also play an active integral role in lung repair and display various mechanisms for airway regeneration (Toya and Malik, 2012). The lung ECs investigated in our study were able to proliferate, migrate and form vascular networks without the presence of tumour cells emphasising they functioned normally. In the cocultures of lung ECs and H520 tumour cells a capillary-like network was formed that was very similar to that generated in endothelial monocultures. However the tumour cells closely nestled up and lined themselves along the endothelial tubular structures. Lining of tumour cells along blood vessels is well described in in situ tissue studies of many tumours and is said to be a prerequisite for metastasis (Qin et al., 2012). Such a phenomenon in vivo is known as co-option of existing host vessels, whereby malignant cells initially form a vascularised tumour mass. Holash et al. (1999) have described tumour cells inducing apoptosis of ECs during vascular co-option. However in our study an overgrowth or replacement of the ECs was not observed. Rather it seemed that the tumour cells used the lung ECs as a framework/navigation system for their own proliferation and structure. The ECs appear to be spared by tumour cells in order to create new vessels in communication/cooperation with them. Our in vitro and ex vivo results indicate that tumour vessels may exhibit characteristics of mosaicism, as the vessel walls appeared to be composed of both, tumour cells and ECs. Tumour cells which had
positioned themselves in proximity to the ECs and incorporated around the vascular channels expressed the endothelial marker CD31. In tissue sections of squamous cell carcinoma, some tumour cells in the vessel walls lacked expression of the EC marker CD31. Mosaic vessels have not been described in lung tumours but have been seen in melanoma and carcinomas of the prostate, ovary, breast and colon (Chang et al., 2000; di Tomaso et al., 2005). Chang et al. (2000) found that 15% of blood vessels in a human colon carcinoma grown in mice were intermingled with tumour cells in the vessel wall. The most prominent result was the finding that ECs were bridged over by narrow strands of adenocarcinoma tumour cells. Both the bridging as well as those tumour cells in direct contact with ECs were CD31 positive. Thus, surprisingly these tumour cells showed a “CD31 staining switch” on contacting ECs, changing from a negative to a positive CD31 expression pattern, thereby mimicking ECs. Tumour cells not in contact with ECs and in more distant areas were CD31 negative. It is unlikely that the tumour cells appeared CD31 positive because of positive staining of closely co-located ECs especially since tumour cells located adjacent to ECs and definitely not covered by an EC were also seen to be immuno-positive. Tumour cells and stem cells have many similar properties including their intrinsic plasticity. Stem cells have an almost unlimited potential to proliferate, accompanied by an ability to differentiate. Similarly, tumour cells show uncontrolled proliferation, cellular plasticity and expression of genes involved in maintaining pluripotency and plasticity (Clevers, 2011; Ridolfi et al., 2009). This plasticity of tumour cells is suggested to be one element underlying the generation of channels via vasculogenic mimicry.
Fig. 8. Light microscopy of human lung tissue sections with adenocarcinoma immunostained with CD31. (a) Endothelium is CD31 positive. (b) Vessel like channel which contains CD31 positive tumour cells as well as CD31 negative ECs.
314
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
In de novo tumour vessel channels non-endothelial cells have been found to express typical endothelial markers as seen in uveal melanoma cells which express the EC markers CD31 and CD34 (Folberg and Maniotis, 2004). Other research groups have reported CD31 immunoreactivity in a range of sarcomas and carcinomas (Lutzky et al., 2006; Nicholson et al., 2000). Importantly the co-culture observations of our study were confirmed in tissue sections from resected lung squamous cell carcinoma and adenocarcinoma. In the adenocarcinoma sections a staining switch of tumour cells had occurred in some areas with clusters of tumour cells immunolabelling positively for CD31. These results suggest that the tumour plasticity has been influenced by the endothelial microenvironment and by direct interactions with ECs. Fan and Sun (2010) have highlighted the important influence of microenvironment on tumour cells and on their long-term proliferation and their formation of vasculogenic-like networks. At the end-points of our co-cultures it was impossible to differentiate whether the vascular-like networks observed were formed by the tumour or by the ECs. However, the adenocarcinoma cells seemed to demonstrate plasticity as evident by their morphology and molecular characteristics mimicking ECs. It is also likely that lung tumour cells (by acquiring the morphology and molecular status of ECs) escape apoptosis and possibly therapeutic interventions. In conclusion, lung carcinoma cells modify their cellular and molecular features when co-cultured with lung ECs. Various interactions between lung endothelial and lung tumour cells were detected that mirrored the formation of alternative blood supply in lung tumours. Those non-angiogenic processes should be considered when anti-tumour therapeutic interventions are planned. Funding This project was funded under the auspices of the “Go8 Australia-Germany Joint Research Co-operation” of the German Academic Exchange Service (DAAD). Competing interests The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported. Ethical standards The study was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital Nedlands Western Australia, Australia. Acknowledgements The authors would like to thank Tania Fuhrmann-Selter, Barbara Drewes, Faang Cheah, and Karin Briest-Forch for technical support. References Ackermann, M., Morse, B.A., Delventhal, V., Carvajal, I.M., Konerding, M.A., 2012. Anti-VEGFR2 and anti-IGF-1R-adnectins inhibit Ewing’s sarcoma A673xenograft growth and normalize tumor vascular architecture. Angiogenesis 15, 685. Auerbach, R., Lewis, R., Shinners, B., Kubai, L., Akhtar, N., 2003. Angiogenesis assays: a critical overview. Clin. Chem. 49, 32. Bach, F., Uddin, F.J., Burke, D., 2007. Angiopoietins in malignancy. Eur. J. Surg. Oncol. 33, 7. Bahramsoltani, M., Plendl, J., Janczyk, P., Custodis, P., Kaessmeyer, S., 2009. Quantitation of angiogenesis and antiangiogenesis in vivo, ex vivo and in vitro – an overview. ALTEX 26, 95.
Bartha, K., Rieger, H., 2006. Vascular network remodeling via vessel cooption, regression and growth in tumors. J. Theor. Biol. 241, 903. Bergers, G., Benjamin, L.E., 2003. Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3, 401. Bussolati, B., Grange, C., Camussi, G., 2011. Tumor exploits alternative strategies to achieve vascularization. FASEB J. 25, 2874. Chang, Y.S., di Tomaso, E., McDonald, D.M., Jones, R., Jain, R.K., Munn, L.L., 2000. Mosaic blood vessels in tumors: frequency of cancer cells in contact with flowing blood. Proc. Natl. Acad. Sci. U. S. A. 97, 14608. Clevers, H., 2011. The cancer stem cell: premises, promises and challenges. Nat. Med. 17, 313. De Spiegelaere, W., Casteleyn, C., Van den Broeck, W., Plendl, J., Bahramsoltani, M., Simoens, P., Djonov, V., Cornillie, P., 2012. Intussusceptive angiogenesis: a biologically relevant form of angiogenesis. J. Vasc. Res. 49, 390. di Tomaso, E., Capen, D., Haskell, A., Hart, J., Logie, J.J., Jain, R.K., McDonald, D.M., Jones, R., Munn, L.L., 2005. Mosaic tumor vessels: cellular basis and ultrastructure of focal regions lacking endothelial cell markers. Cancer Res. 65, 5740. Elzarrad, K., Haroon, A., Reed, D., Al-Mehdi, A.B., 2009. Early incorporated endothelial cells as origin of metastatic tumor vasculogenesis. Clin. Exp. Metastasis 26, 589. Fan, Y.Z., Sun, W., 2010. Molecular regulation of vasculogenic mimicry in tumors and potential tumor-target therapy. World J. Gastrointest. Surg. 2, 117. Folberg, R., Maniotis, A.J., 2004. Vasculogenic mimicry. Acta Pathol. Microbiol. Immunol. Scand. 112, 508. Hillen, F., Griffioen, A.W., 2007. Tumour vascularization: sprouting angiogenesis and beyond. Cancer Metastasis Rev. 26, 489. Holash, J., Maisonpierre, P.C., Compton, D., Boland, P., Alexander, C.R., Zagzag, D., Yancopoulos, G.D., Wiegand, S.J., 1999. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 284, 1994. Kaessmeyer, S., Plendl, J., 2009. Angiogenesis and vasculogenesis in the corpus luteum in vitro. Clin. Hemorheol. Microcirc. 41, 83. Kassmeyer, S., Plendl, J., Custodis, P., Bahramsoltani, M., 2009. New insights in vascular development: vasculogenesis and endothelial progenitor cells. Anat. Histol. Embryol. 38, 1. Kimman, M., Norman, R., Jan, S., Kingston, D., Woodward, M., 2012. The burden of cancer in member countries of the Association of Southeast Asian Nations (ASEAN). Asian Pac J. Cancer Prev. 13, 411. Levine, H.A., Pamuk, S., Sleeman, B.D., Nilsen-Hamilton, M., 2001. Mathematical modeling of capillary formation and development in tumor angiogenesis: penetration into the stroma. Bull. Math. Biol. 63, 801. Lutzky, V.P., Carnevale, R.P., Alvarez, M.J., Maffia, P.C., Zittermann, S.I., Podhajcer, O.L., Issekutz, A.C., Chuluyan, H.E., 2006. Platelet-endothelial cell adhesion molecule1 (CD31) recycles and induces cell growth inhibition on human tumor cell lines. J. Cell Biochem. 98, 1334. Makanya, A.N., Hlushchuk, R., Djonov, V.G., 2009. Intussusceptive angiogenesis and its role in vascular morphogenesis, patterning, and remodeling. Angiogenesis 12, 113. Maniotis, A.J., Folberg, R., Hess, A., Seftor, E.A., Gardner, L.M., Pe’er, J., Trent, J.M., Meltzer, P.S., Hendrix, M.J., 1999. Vascular channel formation by human melanoma cells in vivo and in vitro: vasculogenic mimicry. Am. J. Pathol. 155, 739. McClelland, M.R., Carskadon, S.L., Zhao, L., White, E.S., Beer, D.G., Orringer, M.B., Pickens, A., Chang, A.C., Arenberg, D.A., 2007. Diversity of the angiogenic phenotype in non-small cell lung cancer. Am. J. Respir. Cell Mol. Biol. 36, 343–350. Nicholson, S.A., McDermott, M.B., DeYoung, B.R., Swanson, P.E., 2000. CD31 immunoreactivity in small round cell tumors. Appl. Immunohistochem. Mol. Morphol. 8, 19. Passalidou, E., Trivella, M., Singh, N., Ferguson, M., Hu, J., Cesario, A., Granone, P., Nicholson, A.G., Goldstraw, P., Ratcliffe, C., Tetlow, M., Leigh, I., Harris, A.L., Gatter, K.C., Pezzella, F., 2002. Vascular phenotype in angiogenic and non-angiogenic lung non-small cell carcinomas. Br. J. Cancer 86, 244. Patan, S., 2000. Vasculogenesis and angiogenesis as mechanisms of vascular network formation, growth and remodeling. J. Neurooncol. 50, 1. Qin, L., Bromberg-White, J.L., Qian, C.N., 2012. Opportunities and challenges in tumor angiogenesis research: back and forth between bench and bed. Adv. Cancer Res. 113, 191. Ridolfi, L., Petrini, M., Fiammenghi, L., Riccobon, A., Ridolfi, R., 2009. Human embryo immune escape mechanisms rediscovered by the tumor. Immunobiology 214, 61. Rocha, S.F., Adams, R.H., 2009. Molecular differentiation and specialization of vascular beds. Angiogenesis 12, 139. Sievers, H., Bahramsoltani, M., Kassmeyer, S., Plendl, J., 2011. In vitro angiogenic potency in human microvascular endothelial cells derived from myocardium, lung and skin. Clin. Hemorheol. Microcirc. 49, 473. Toya, S.P., Malik, A.B., 2012. Role of endothelial injury in disease mechanisms and contribution of progenitor cells in mediating endothelial repair. Immunobiology 217, 569. Vacca, A., Ribatti, D., 2011. Angiogenesis and vasculogenesis in multiple myeloma: role of inflammatory cells. Recent Results Cancer Res. 183, 87–95. Witz, I.P., 2009. The tumor microenvironment: the making of a paradigm. Cancer Microenviron. 2 (1), 9. Yano, T., Haro, A., Shikada, Y., Maruyama, R., Maehara, Y., 2011. Non-small cell lung cancer in never smokers as a representative ‘non-smoking-associated lung cancer’: epidemiology and clinical features. Int. J. Clin. Oncol. 16, 287.
Contents lists available at ScienceDirect
Immunobiology journal homepage: www.elsevier.com/locate/imbio
Lung cancer neovascularisation: Cellular and molecular interaction between endothelial and lung cancer cells Sabine Kaessmeyer a,∗ , Kanti Bhoola b,1 , Svetlana Baltic b , Philip Thompson b,2 , Johanna Plendl a,2 a
Institute of Veterinary Anatomy, Department of Veterinary Medicine, Freie Universität Berlin, Koserstraße 20, 14195 Berlin, Germany Lung Institute of Western Australia and Centre for Asthma, Allergy and Respiratory Research, The University of Western Australia, Nedlands, WA 6009, Australia b
a r t i c l e
i n f o
Article history: Received 14 June 2013 Accepted 14 November 2013 Available online 22 November 2013 Keywords: CD31 Co-culture Endothelial cells Immunocytochemistry Immunohistochemistry Lung adenocarcinoma Lung squamous carcinoma
a b s t r a c t Background: Novel vascular-independent conduits have been observed in some cancers. These have been variously described as vasculogenic mimicry, mosaic vessel formation, vascular co-option and intratumour embryonic-like vasculogenesis. Despite lung cancer being the most common cancer worldwide, there is little information on its neovascularisation or the pathways involved. Methods: An in vitro model involving co-cultures of microvascular lung endothelial cells and squamous or adenocarcinoma lung cancer cells was developed to assess their angiogenic interaction. Cells were incubated and examined by phase contrast microscopy and by immunocytochemistry in both monoand co-cultures. Cultured cells and lung cancer tissue sections were assessed for new tumour vessel formation, expression of the endothelial marker CD31 and morphology. Results: Lung tumour cells and endothelial cells interacted morphologically via pseudopodia and used alternative pathways to generate new vessels. Co-culturing microvascular endothelial and squamous carcinoma cells led to endothelial cells surrounding tumour cells and the tumour cells being incorporated into vessel walls. Co-culturing endothelial and adenocarcinoma cells resulted in cellular contact and the formation of tumour cell bridges around clusters of endothelial cells. These adencocarcinoma cells became strongly positive for CD31. Tumour tissue section studies supported the in vitro findings. Conclusion: Lung carcinoma cells when co-cultured with lung endothelial cells modify their cellular and molecular features that encourage alternative means of providing blood supply. The mechanisms underpinning these non-angiogenic processes need to be further investigated and should be considered when anti-tumour therapeutic interventions are being considered. © 2013 Elsevier GmbH. All rights reserved.
Introduction
Abbreviations: CD, cluster of differentiation; EC, endothelial cells; NSCLC, nonsmall cell lung cancer; HMVEC-L, human microvascular endothelial cells-lung; rhFGF, recombinant human fibroblast growth factor; VEGF, vascular endothelial growth factor; GA, gentamicin/amphotericin; FBS, fetal bovine serum; H520, human lung squamous cell carcinoma cells 520; ATCC, American Type Culture Collection; H2126, human lung adenocarcinoma cells; DMEM, Dulbecco’s modified Eagle medium; PBS, phosphate buffered saline; PFA, paraformaldehyde; TBS, Tris-buffered saline. ∗ Corresponding author at: Institute of Veterinary Anatomy, Department of Veterinary Medicine, Freie Universität Berlin, Koserstraße 20, D-14195 Berlin, Germany. Tel.: +49 03083853558; fax: +49 03083853480. E-mail address: [email protected] (S. Kaessmeyer). 1 Institutes of Anatomy, Histology & Pathology, and Physiology, Universidad Austral de Chile, Valdivia Box 567, Chile 2 Equal last author. 0171-2985/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.imbio.2013.11.004
The formation of blood vessels is initiated by two distinct processes, namely vasculogenesis and angiogenesis. Vasculogenesis is the in situ differentiation of endothelial precursor cells (angioblasts) into endothelial cells (EC), which then re-assemble into primitive vascular plexuses. Angiogenesis generally involves the expansion of a primitive vascular network into a complex one (Hillen and Griffioen, 2007; Patan, 2000; Makanya et al., 2009) by growth of endothelial sprouts from pre-existing blood vessels through migration, proliferation, three-dimensional organisation and tube formation of ECs. Whereas formation of new blood vessels plays a major role during prenatal development, in adults, blood vessel growth is linked primarily to tissue repair and clinical disorders such as tumour growth (Vacca and Ribatti, 2011). The growth of solid tumours beyond 1–2 mm in diameter requires the induction of new blood vessel formation (Bergers and Benjamin, 2003). It has
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
been assumed that tumour vascularisation can be explained by angiogenesis. However, in the last decade, novel angiogenesisindependent pathways have been observed in the blood supply of certain tumours. Such alternative processes, found particularly in aggressive tumours, are designated variously as vasculogenic mimicry, mosaic vessel formation, vascular co-option and intratumour embryonic-like vasculogenesis (Hillen and Griffioen 2007; Bussolati et al., 2011; Chang et al., 2000; Ackermann et al., 2012). Vasculogenic mimicry, first described by Maniotis et al. (1999) in uveal melanoma, is the formation of a complete capillary-like network that is comprised only of tumour cells instead of vascular ECs. In tumours such as colon carcinoma, mosaic vessels have been described, in which, attached to the luminal surface of the vascular channels are both ECs as well as non-ECs; the latter, lacking expression of specific endothelial cell markers (Chang et al., 2000; Ackermann et al., 2012). Tumour cells that grow along pre-existing vessels but without evoking an angiogenic response, referred to as vessel co-option, have been found in cerebral glioblastoma, breast adenocarcinoma, and in melanomas (Bartha and Rieger 2006; Holash et al., 1999). Finally, de novo generation of tumour vessels can arise from the differentiation of stem and progenitor cells of hematopoietic origin or from those resident in tissues and participate in tumour progression (Bussolati et al., 2011). Lung cancer is the leading cause of cancer worldwide (Kimman et al., 2012) and non-small cell lung cancers (NSCLC; subtypes squamous cell carcinoma and adenocarcinoma), account for approximately 80–90% of lung cancers (Yano et al., 2011). Despite this, there is a lack of information on their neovascularisation, especially with respect to alternative cellular processes for vessel formation (McClelland et al., 2007; Passalidou et al., 2002). Tumour neovascularisation is a complex process based upon a sequence of interactions between tumour cells and ECs (Levine et al., 2001; Witz, 2009). In vitro culture models allow analysis of each step involved in vascular growth including the interaction between endothelial and tumour cells. As such an in vitro lung cancer vascular model should help identify the specific steps and mechanisms involved (Auerbach et al., 2003; Kaessmeyer and Plendl, 2009; Kassmeyer et al., 2009; De Spiegelaere et al., 2012; Sievers et al., 2011; Bahramsoltani et al., 2009). The aim of the current study was to develop an in vitro model to investigate the vascular and molecular interactions that occur between lung endothelial and lung carcinoma cells thereby providing evidence for recruitment of ECs and the induction of new blood vessel formation by lung carcinomas. A morphological study was designed to monitor co-cultures of microvascular lung ECs with two types of NSCLC. Materials and methods The study was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital Nedlands Western Australia, Australia. Cell culture Cells and culture media Lung derived normal human microvascular ECs (HMVEC-L, Lonza, Walkersville Inc., Walkersville, USA) were incubated in basic EC culture medium (EGM-2-MV; Lonza) supplemented according to the supplier’s instructions with rhFGF, VEGF, vitamin C, GA-1000, hydrocortisone and FBS. Human lung squamous cell carcinoma cells (H520, ATCC) and human lung adenocarcinoma cells (H2126, ATCC) were cultured in DMEM with 10% FBS, 1% l-glutamine and 1% penicillin/streptomycin (10,000 U/ml) (all from Sigma–Aldrich, Taufkirchen, Germany; and referred subsequently as ‘tumour cell
309
Table 1 Experimental design for culture and co-culture time-points: tumour cells, adapted to culture medium of endothelial cells, were added at a concentration of 2 × 104 cells to the lung endothelial cells (4 × 104 cells seeded). Human lung microvascular endothelial cells (HMVEC-LV) Time in culture
Duration of co-culture of H520 tumour cells and HMVEC-LV
Duration of co-culture of AC H2126 tumour cells and HMVEC-LV
28 Days 28 Days 28 Days 28 Days
24 h 7 Days 14 Days 32 Days
24 h 7 Days 14 Days 32 Days
medium’). Cells were incubated in a humidified atmosphere (37 ◦ C, 5% CO2 ) and the medium was replaced every 2–3 days. Priming of lung tumour cells Adaptation (priming) of the tumour mono-cell cultures to the HMVEC-L cell medium was carried out over a 48 hour period in steps that sequentially involved mixing fresh tumour cell medium with 25%, 50%, 75% and finally 100% of HMVEC-L cell medium. Endothelial cell culture HMVEC-L cells were seeded in 24-well plates from Iwaki (Tokyo, Japan) onto round glass cover slips (12 mm diameter, Roth, Karlsruhe, Germany) which were gelatine coated (Difco Laboratories, Detroit, USA, 1.5% in PBS). Optimal seeding density of 4 × 104 cells per well resulted in their adherence and formation of an even monolayer. Direct co-culture design (Fig. 1) The primary co-culture design allowed HMVEC-L cells to be incubated with the primed lung H520 or H2126 tumour cells (each trypsinised, seeding density: 2 × 104 ) for varying times. On the day of co-culturing, the HMVEC-L cells had already been in culture for 28 days (Stage 3, Table 2). The precise time schedules for the co-cultures are shown in Table 1. At the end of each co-culture incubation period, cells were rinsed, fixed and labelled with CD31 for subsequent immunocytochemistry. Phase contrast microscopy of cultured cells Light micrographs were taken with a Zeiss Axiovert 25 inverted microscope (Zeiss MicroImaging GmbH, Jena, Germany) and a 000610 video camera INTEQ (INTEQ, Berlin, Germany) using Zeiss Axiovision image-processing software. Lung tissue samples Archival lung carcinoma specimens were obtained from the Lung Institute of Western Australia and PathWest Inc. The tissue samples were fixed in 4% PFA, paraffin embedded, and cut into 5 m sections. Multiple tissue sections of resected normal lung, adenocarcinoma (1 female, 1 male) and squamous cell carcinoma (2 males) were stained by immunohistochemistry as described below. Immunohistochemistry Antigen retrieval was achieved by heating tissue slides in citrate buffer at 95 ◦ C in a water bath and blocking of endogenous peroxidase activity with TBS-buffer with 0.6% (v/v) H2 O2 , followed by washing in TBS-buffer plus 0.01% polysorbate 20 (Tween20, Sigma–Aldrich). The slides were pre-incubated with 20% (v/v) donkey IgG serum in buffer A, and finally incubated with CD31 overnight at 4 ◦ C (see Table 3).
310
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314 Table 2 Staging of cellular morphology during capillary-like tube formation of human lung microvascular endothelial cells in vitro. Stages
Lung endothelial cells
Time points
Stage 0 Stage 1
Proliferation Non-confluent monolayer with cell-free areas of fusiform cells Development of long cellular extensions Connecting 3-dimensional extensions Formation of 3-dimensional network of capillary-like structures
Days 0–7 Days 7–14
Stage 2 Stage 3 Stage 4
Days 14–28 Days 28–42 Days 42–60
Germany), a digital camera (DS-Ri1, Nikon, Düsseldorf, Germany) and a PC-based laboratory imaging programme (NIS-Elements, Version 3.0, Nikon, Düsseldorf, Germany). Analysis of the cellular morphology observed during capillary-like tube formation of ECs in vitro was classified as shown in Table 2. Results Mono-cell cultures
Fig. 1. Experimental design of direct co-culture method. (a) Adherent endothelial cells (blue); medium (pink). (b) Tumour cells (green) in suspension adapted to medium of endothelial cells. (c) Direct co-cultures. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Immunocytochemistry Cultured cells were washed with PBS and fixed in methanol/ acetone (1:1) at 4 ◦ C, washed with PBS and immersed in protein block, serum-free (DAKO Diagnostika, Hamburg, Germany). After removal of the blocker, cells were incubated with CD31 overnight at 4 ◦ C.
Human lung microvascular endothelial cells (HMVEC-L) HMVEC-L cells formed a three-dimensional network of capillary-like structures over a period of 42 days. During the first 14 days of incubation cells grew as a sub-confluent monolayer with clear spaces between individual cells (Fig. 3a, phase contrast microscopy). During the next 14 days cells developed long cellular connecting extensions. Three-dimensional extensions were observed in the time frame of 28 until 42 days in culture. Further expansion and connection of three-dimensional extensions resulted in a cellular network which developed into tube-like structures up to day 60, while the underlying network of cells on the bottom of the culture plate disintegrated. At all monitored time points the HMVEC-L cells immunolabelled positively for CD31 (Fig. 2a – Stage 2 as described in Table 2 and Fig. 2b – Stage 4). Fig. 3a displays the HMVEC-L cells at stage 1.
Immunodetection Slides and cells were incubated with horseradish peroxidase labelled secondary antibody for 30 min at room temperature. Labelling was visualised by diaminobenzidine (Sigma–Aldrich). Nuclei were counterstained with hemalaun (Fluka-Chemie, Buchs, Switzerland). Negative controls were performed in which the primary antibody was replaced by (a) buffer and (b) non-immune serum (Table 3). Light microscopy of immunostained cells Analysis of the immunostained cells was digitally recorded employing a light microscope (Axioskop, Carl Zeiss, Oberkochen,
Human lung squamous cell carcinoma cells (H520) H520 cells grew in an adherent monolayer and were heterogeneous in appearance. Cells were round or polygonal in shape with varying cell diameter and nuclear-cytoplasmic ratio. Only a small number of tumour cells showed bipolar or spindle-shaped forms. The monolayer never developed into complete confluency as shown in Fig. 3b. Human lung adenocarcinoma cells (H2126) The cellular morphology was more homogenous in appearance than in H520 cells. The majority of cells were polygonal in shape, creating a cobblestone like monolayer (Fig. 3c).
Fig. 2. Light microscopy of human lung endothelial cells, immunostained with CD31. (a) Long cellular extensions (stage 2) light microscopy. (b) Network of capillary-like structures (stage 4) light microscopy.
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
311
Fig. 3. Phase contrast micrographs of human lung endothelial cells and carcinoma cells in vitro. (a) Human lung endothelial cells (HMVEC-L), large cells, cell-free areas (stage 1). (b) Human lung squamous cell carcinoma (H520). (c) Human lung adenocarcinoma (AC H2126).
Co-cultures HMVEC-L cells were co-cultured with either the squamous cell carcinoma or the adenocarcinoma cells as per the research plan (Table 1 and Fig. 1).
Co-cultures of HMVEC-L (cultured 28 days) with H520 cells for 24 h, 7 days, 14 days and 32 days Whereas the large HMVEC-L cells immunolabelled positively with CD31, the small and polygonal tumour cells showed no immunoreactivity. After 24 h, both cell types started proliferating and formed pseudopodia which made contact with each other (data not shown).
Fig. 4. Light microscopy of co-cultures, immunostained with CD31, HMVEC-L (28 days in vitro) co-cultured with H520 cells. (a) After 7 days of co-culturing, HMVEC-L form a fine network around H520 tumour cells. (b) After 14 days of co-culturing, H520 line along HMVEC-L. (c) After 32 days of co-culturing, CD31 positive network of co-cultured endothelial and tumour cells over a monolayer of CD31 negative tumour cells.
Table 3 Immunohistochemistry – antibodies, dilutions and buffer A ingredients. Primary antibody Secondary antibody Negative serum control Buffer A
Cell culture: 1:40 in 0.01 M PBS pH 7.4 + 2% donkey serum CD31 (PECAM-1) monoclonal mouse anti-human CD31, AB9498, Tissue sections: 1:40 in buffer A + 2% donkey serum Abcam, Cambridge, MA, USA Cell culture: 1:80 in 0.01 M PBS pH 7.4 + 2% donkey serum IgG-HRP-conjugated donkey anti-mouse, ZMH2162, Linaris Tissue sections: 1:40 in buffer A + 2% donkey serum Biologische Produkte, Wertheim/Germany Cell culture: 1:40 in 0.01 M PBS pH 7.4 + 2% donkey serum Mouse-IgG1 serum, Item X0931, DAKO Diagnostika, Hamburg, Tissue sections: 1:40 in buffer A + 2% donkey serum Germany 0.05 M Tris–HCl pH 7.6 + 0.9% w/v NaCL + 0.66 m M MgCl2 + 1% w/v BSA + 0.1% w/v gelatine + 0.01% v/v Tween20
312
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
Fig. 5. Light microscopy of co-cultures, immunostained with CD31, HMVEC-L (28 days in vitro) co-cultured with H2126 cells. (a) After 14 days of co-culturing, HMVEC-L bridge above H2126. (b) After 14 days of co-culturing, bridging H2126 cells are positive for CD31, non-bridging H2126 cells stay negative for CD31.
After 7 days of co-culture both cell types continued to proliferate and could be clearly distinguished by their morphology. Tumour cells did not overgrow HMVEC-L cells. Three-dimensional and finely branching long pseudopodia of the HMVEC-L cells established contact with the H520 tumour cells. The HMVEC-L cells and their pseudopodia which immunostained positively with CD31 seemed to invade groups of adjacent tumour cells and to form a fine network around single tumour cells (Fig. 4a). After 14 days in co-culture HMVEC-L cells which continued to stain positively for CD31, were lined by finger-like formations of elongated H520 tumour cells (Fig. 4b). These tumour cells which had positioned themselves along the HMVEC-L cells now stained positively for CD31 (Fig. 4b). By day 32 a CD31 positive network of co-cultured endothelial and tumour cells had grown over a monolayer of CD31 negative tumour cells (Fig. 4c). In mono-cell cultures of tumour cells, which had been already adapted to HMVEC-L cell medium, no such growth pattern was observed.
Co-cultures of HMVEC-L (cultured 28 days) with H2126 cells for 24 h, 7 days, 14 days and 32 days Whereas HMVEC-L cells immunostained positively, the smaller, roundish tumour cells showed no immunoreactivity with CD31. Both cell types developed cellular extensions, which made contact with each other after 24 h (data not shown). After 7 days in co-culture tumour cells continued to proliferate but they did not overgrow HMVEC-L cells (data not shown). Narrow strands of tumour cells formed bridges over islands of HMVEC-L cells. After 14 days many more strand-like bridges of tumour cells were observed which were orientated perpendicularly to the axis of the HMVEC-L cells. Only bridging tumour cells, tumour cells in close proximity to HMVEC-L cells, and HMVEC-L cells themselves immunostained positively with CD31 (Fig. 5a, b). By the 32nd day of co-culturing, a CD31 positive capillary-like network of endothelial and tumour cells had developed over a monolayer of CD31 negative tumour cells (data not shown). In mono-cell cultures of tumour cells, which had been adapted to HMVEC-L cell medium, no differentiation as described above was observed.
Squamous cell carcinoma tissue Most ECs of blood vessels within the tumour were spread out and were CD31 immuno-positive. Tumour cells seemed to line along vascular channels. However, in specific areas of blood vessels single or even a few (5–7) CD31 negative cells seemed to have been incorporated into the lining of the vessel wall (Fig. 7a, b). Adenocarcinoma tissue In most areas of the tumour, ECs immunostained positively with CD31 (Fig. 8a, b). However, in some areas clusters of tumour cells also immunolabelled positively for CD31. Controls No immunolabelling was observed in the absence of the primary antibody or its replacement by non-immune serum. Discussion Tumour cells act as angiogenic inducers by secreting vascular endothelial growth factors. They are able to transform the normally quiescent vascular ECs through the so called “angiogenic switch” into a morphologically altered and activated state (Bergers and Benjamin, 2003; Bach et al., 2007; Rocha and Adams, 2009). A somewhat unexpected but important finding was the development of endothelial pseudopodia that appeared to invade groups of tumour cells and form networks around tumour cells particular in the case of the squamous carcinoma. This was the opposite cell behaviour to what was expected. Although it is possible this is an in vitro phenomenon, this finding was however supported
CD31 immuno-reactive cells in normal and carcinoma lung tissue sections Normal human lung tissue ECs of small capillaries near alveolar spaces as well as larger blood vessels in stromal connective tissue immunostained positively for CD31 (Fig. 6).
Fig. 6. Light microscopy of healthy human lung tissue sections immunostained with CD31, endothelium of large and small vessels is CD31 positive.
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
313
Fig. 7. Light microscopy of human lung tissue sections with lung squamous cell carcinoma immunostained with CD31. (a) Endothelium is CD31 positive, tumour cells line along vessels, intermittently tumour cells instead of ECs as inner vascular layer (arrow). (b) Endothelium is CD31 positive, tumour cells line along vessels, intermittently tumour cells instead of ECs as inner vascular layer (arrows).
by Elzarrad et al., (2009) who has shown that pulmonary ECs play an active role in tumour metastasis by building intercellular connections with tumour cells and growing with them to create new vessels in a tumour environment of oxygen deficiency indicating that our finding is not exclusively an in vitro phenomenon. Furthermore, endothelial cells also play an active integral role in lung repair and display various mechanisms for airway regeneration (Toya and Malik, 2012). The lung ECs investigated in our study were able to proliferate, migrate and form vascular networks without the presence of tumour cells emphasising they functioned normally. In the cocultures of lung ECs and H520 tumour cells a capillary-like network was formed that was very similar to that generated in endothelial monocultures. However the tumour cells closely nestled up and lined themselves along the endothelial tubular structures. Lining of tumour cells along blood vessels is well described in in situ tissue studies of many tumours and is said to be a prerequisite for metastasis (Qin et al., 2012). Such a phenomenon in vivo is known as co-option of existing host vessels, whereby malignant cells initially form a vascularised tumour mass. Holash et al. (1999) have described tumour cells inducing apoptosis of ECs during vascular co-option. However in our study an overgrowth or replacement of the ECs was not observed. Rather it seemed that the tumour cells used the lung ECs as a framework/navigation system for their own proliferation and structure. The ECs appear to be spared by tumour cells in order to create new vessels in communication/cooperation with them. Our in vitro and ex vivo results indicate that tumour vessels may exhibit characteristics of mosaicism, as the vessel walls appeared to be composed of both, tumour cells and ECs. Tumour cells which had
positioned themselves in proximity to the ECs and incorporated around the vascular channels expressed the endothelial marker CD31. In tissue sections of squamous cell carcinoma, some tumour cells in the vessel walls lacked expression of the EC marker CD31. Mosaic vessels have not been described in lung tumours but have been seen in melanoma and carcinomas of the prostate, ovary, breast and colon (Chang et al., 2000; di Tomaso et al., 2005). Chang et al. (2000) found that 15% of blood vessels in a human colon carcinoma grown in mice were intermingled with tumour cells in the vessel wall. The most prominent result was the finding that ECs were bridged over by narrow strands of adenocarcinoma tumour cells. Both the bridging as well as those tumour cells in direct contact with ECs were CD31 positive. Thus, surprisingly these tumour cells showed a “CD31 staining switch” on contacting ECs, changing from a negative to a positive CD31 expression pattern, thereby mimicking ECs. Tumour cells not in contact with ECs and in more distant areas were CD31 negative. It is unlikely that the tumour cells appeared CD31 positive because of positive staining of closely co-located ECs especially since tumour cells located adjacent to ECs and definitely not covered by an EC were also seen to be immuno-positive. Tumour cells and stem cells have many similar properties including their intrinsic plasticity. Stem cells have an almost unlimited potential to proliferate, accompanied by an ability to differentiate. Similarly, tumour cells show uncontrolled proliferation, cellular plasticity and expression of genes involved in maintaining pluripotency and plasticity (Clevers, 2011; Ridolfi et al., 2009). This plasticity of tumour cells is suggested to be one element underlying the generation of channels via vasculogenic mimicry.
Fig. 8. Light microscopy of human lung tissue sections with adenocarcinoma immunostained with CD31. (a) Endothelium is CD31 positive. (b) Vessel like channel which contains CD31 positive tumour cells as well as CD31 negative ECs.
314
S. Kaessmeyer et al. / Immunobiology 219 (2014) 308–314
In de novo tumour vessel channels non-endothelial cells have been found to express typical endothelial markers as seen in uveal melanoma cells which express the EC markers CD31 and CD34 (Folberg and Maniotis, 2004). Other research groups have reported CD31 immunoreactivity in a range of sarcomas and carcinomas (Lutzky et al., 2006; Nicholson et al., 2000). Importantly the co-culture observations of our study were confirmed in tissue sections from resected lung squamous cell carcinoma and adenocarcinoma. In the adenocarcinoma sections a staining switch of tumour cells had occurred in some areas with clusters of tumour cells immunolabelling positively for CD31. These results suggest that the tumour plasticity has been influenced by the endothelial microenvironment and by direct interactions with ECs. Fan and Sun (2010) have highlighted the important influence of microenvironment on tumour cells and on their long-term proliferation and their formation of vasculogenic-like networks. At the end-points of our co-cultures it was impossible to differentiate whether the vascular-like networks observed were formed by the tumour or by the ECs. However, the adenocarcinoma cells seemed to demonstrate plasticity as evident by their morphology and molecular characteristics mimicking ECs. It is also likely that lung tumour cells (by acquiring the morphology and molecular status of ECs) escape apoptosis and possibly therapeutic interventions. In conclusion, lung carcinoma cells modify their cellular and molecular features when co-cultured with lung ECs. Various interactions between lung endothelial and lung tumour cells were detected that mirrored the formation of alternative blood supply in lung tumours. Those non-angiogenic processes should be considered when anti-tumour therapeutic interventions are planned. Funding This project was funded under the auspices of the “Go8 Australia-Germany Joint Research Co-operation” of the German Academic Exchange Service (DAAD). Competing interests The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported. Ethical standards The study was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital Nedlands Western Australia, Australia. Acknowledgements The authors would like to thank Tania Fuhrmann-Selter, Barbara Drewes, Faang Cheah, and Karin Briest-Forch for technical support. References Ackermann, M., Morse, B.A., Delventhal, V., Carvajal, I.M., Konerding, M.A., 2012. Anti-VEGFR2 and anti-IGF-1R-adnectins inhibit Ewing’s sarcoma A673xenograft growth and normalize tumor vascular architecture. Angiogenesis 15, 685. Auerbach, R., Lewis, R., Shinners, B., Kubai, L., Akhtar, N., 2003. Angiogenesis assays: a critical overview. Clin. Chem. 49, 32. Bach, F., Uddin, F.J., Burke, D., 2007. Angiopoietins in malignancy. Eur. J. Surg. Oncol. 33, 7. Bahramsoltani, M., Plendl, J., Janczyk, P., Custodis, P., Kaessmeyer, S., 2009. Quantitation of angiogenesis and antiangiogenesis in vivo, ex vivo and in vitro – an overview. ALTEX 26, 95.
Bartha, K., Rieger, H., 2006. Vascular network remodeling via vessel cooption, regression and growth in tumors. J. Theor. Biol. 241, 903. Bergers, G., Benjamin, L.E., 2003. Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3, 401. Bussolati, B., Grange, C., Camussi, G., 2011. Tumor exploits alternative strategies to achieve vascularization. FASEB J. 25, 2874. Chang, Y.S., di Tomaso, E., McDonald, D.M., Jones, R., Jain, R.K., Munn, L.L., 2000. Mosaic blood vessels in tumors: frequency of cancer cells in contact with flowing blood. Proc. Natl. Acad. Sci. U. S. A. 97, 14608. Clevers, H., 2011. The cancer stem cell: premises, promises and challenges. Nat. Med. 17, 313. De Spiegelaere, W., Casteleyn, C., Van den Broeck, W., Plendl, J., Bahramsoltani, M., Simoens, P., Djonov, V., Cornillie, P., 2012. Intussusceptive angiogenesis: a biologically relevant form of angiogenesis. J. Vasc. Res. 49, 390. di Tomaso, E., Capen, D., Haskell, A., Hart, J., Logie, J.J., Jain, R.K., McDonald, D.M., Jones, R., Munn, L.L., 2005. Mosaic tumor vessels: cellular basis and ultrastructure of focal regions lacking endothelial cell markers. Cancer Res. 65, 5740. Elzarrad, K., Haroon, A., Reed, D., Al-Mehdi, A.B., 2009. Early incorporated endothelial cells as origin of metastatic tumor vasculogenesis. Clin. Exp. Metastasis 26, 589. Fan, Y.Z., Sun, W., 2010. Molecular regulation of vasculogenic mimicry in tumors and potential tumor-target therapy. World J. Gastrointest. Surg. 2, 117. Folberg, R., Maniotis, A.J., 2004. Vasculogenic mimicry. Acta Pathol. Microbiol. Immunol. Scand. 112, 508. Hillen, F., Griffioen, A.W., 2007. Tumour vascularization: sprouting angiogenesis and beyond. Cancer Metastasis Rev. 26, 489. Holash, J., Maisonpierre, P.C., Compton, D., Boland, P., Alexander, C.R., Zagzag, D., Yancopoulos, G.D., Wiegand, S.J., 1999. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 284, 1994. Kaessmeyer, S., Plendl, J., 2009. Angiogenesis and vasculogenesis in the corpus luteum in vitro. Clin. Hemorheol. Microcirc. 41, 83. Kassmeyer, S., Plendl, J., Custodis, P., Bahramsoltani, M., 2009. New insights in vascular development: vasculogenesis and endothelial progenitor cells. Anat. Histol. Embryol. 38, 1. Kimman, M., Norman, R., Jan, S., Kingston, D., Woodward, M., 2012. The burden of cancer in member countries of the Association of Southeast Asian Nations (ASEAN). Asian Pac J. Cancer Prev. 13, 411. Levine, H.A., Pamuk, S., Sleeman, B.D., Nilsen-Hamilton, M., 2001. Mathematical modeling of capillary formation and development in tumor angiogenesis: penetration into the stroma. Bull. Math. Biol. 63, 801. Lutzky, V.P., Carnevale, R.P., Alvarez, M.J., Maffia, P.C., Zittermann, S.I., Podhajcer, O.L., Issekutz, A.C., Chuluyan, H.E., 2006. Platelet-endothelial cell adhesion molecule1 (CD31) recycles and induces cell growth inhibition on human tumor cell lines. J. Cell Biochem. 98, 1334. Makanya, A.N., Hlushchuk, R., Djonov, V.G., 2009. Intussusceptive angiogenesis and its role in vascular morphogenesis, patterning, and remodeling. Angiogenesis 12, 113. Maniotis, A.J., Folberg, R., Hess, A., Seftor, E.A., Gardner, L.M., Pe’er, J., Trent, J.M., Meltzer, P.S., Hendrix, M.J., 1999. Vascular channel formation by human melanoma cells in vivo and in vitro: vasculogenic mimicry. Am. J. Pathol. 155, 739. McClelland, M.R., Carskadon, S.L., Zhao, L., White, E.S., Beer, D.G., Orringer, M.B., Pickens, A., Chang, A.C., Arenberg, D.A., 2007. Diversity of the angiogenic phenotype in non-small cell lung cancer. Am. J. Respir. Cell Mol. Biol. 36, 343–350. Nicholson, S.A., McDermott, M.B., DeYoung, B.R., Swanson, P.E., 2000. CD31 immunoreactivity in small round cell tumors. Appl. Immunohistochem. Mol. Morphol. 8, 19. Passalidou, E., Trivella, M., Singh, N., Ferguson, M., Hu, J., Cesario, A., Granone, P., Nicholson, A.G., Goldstraw, P., Ratcliffe, C., Tetlow, M., Leigh, I., Harris, A.L., Gatter, K.C., Pezzella, F., 2002. Vascular phenotype in angiogenic and non-angiogenic lung non-small cell carcinomas. Br. J. Cancer 86, 244. Patan, S., 2000. Vasculogenesis and angiogenesis as mechanisms of vascular network formation, growth and remodeling. J. Neurooncol. 50, 1. Qin, L., Bromberg-White, J.L., Qian, C.N., 2012. Opportunities and challenges in tumor angiogenesis research: back and forth between bench and bed. Adv. Cancer Res. 113, 191. Ridolfi, L., Petrini, M., Fiammenghi, L., Riccobon, A., Ridolfi, R., 2009. Human embryo immune escape mechanisms rediscovered by the tumor. Immunobiology 214, 61. Rocha, S.F., Adams, R.H., 2009. Molecular differentiation and specialization of vascular beds. Angiogenesis 12, 139. Sievers, H., Bahramsoltani, M., Kassmeyer, S., Plendl, J., 2011. In vitro angiogenic potency in human microvascular endothelial cells derived from myocardium, lung and skin. Clin. Hemorheol. Microcirc. 49, 473. Toya, S.P., Malik, A.B., 2012. Role of endothelial injury in disease mechanisms and contribution of progenitor cells in mediating endothelial repair. Immunobiology 217, 569. Vacca, A., Ribatti, D., 2011. Angiogenesis and vasculogenesis in multiple myeloma: role of inflammatory cells. Recent Results Cancer Res. 183, 87–95. Witz, I.P., 2009. The tumor microenvironment: the making of a paradigm. Cancer Microenviron. 2 (1), 9. Yano, T., Haro, A., Shikada, Y., Maruyama, R., Maehara, Y., 2011. Non-small cell lung cancer in never smokers as a representative ‘non-smoking-associated lung cancer’: epidemiology and clinical features. Int. J. Clin. Oncol. 16, 287.
