Background
Teratocarcinomas are highly malignant tumours, cells that have a big potential to differentiate in other cell types. They are considered to recapitulate many events occurring in early embryogenesis, but with a lesser degree of organization and regulation, being also called embryonal carcinomas (EC)
The NT2/D1 cell line is a human EC that does not require any feeder layer to preserve its undifferentiated potential and it may form embryoid body-like structures
When NT2/D1 cells give rise to neurons less than 10% of the cells adopt this phenotype, while the rest of the cells in the nonneural populations remain to be further analyzed and identified. Thus the hypothesis of mesodermal derivatives from NT2/D1 cannot be excluded
We believe that NT2/D1 might be the human counterpart of P19, a mouse teratocarcinoma cell line with cardiac differentiation potential. One of the genes highly expressed in P19, that disappears after the differentiation is finished, is
To find a human embryonic cell line that could differentiate with high frequency into myocardium and to identify the factor(s) that would lead the cells through that path, would be a powerful tool to study cardiomyogenesis,
Here, we have studied the cell morphology and the transcription profile of the NT2/D1 cells cultured for 30 days with and without BMP2. The gene expression analyzed was particularly focused on EC, endothelial, neural progenitors and early and late cardiac differentiation genes. An angiomyogenic medium was tested as differentiation culture medium.
Results
Morphological analysis
The typical NT2/D1 morphological characteristics, as the increase in nuclear/cytoplasm volume ratio, the flattening of the cells, the prominent nucleoli (Figure
NT2/D1 cells in expansion medium
NT2/D1 cells in expansion medium. Phase contrast images of three typical morphologies of NT2/D1 cells in expansion medium, (RPMI, FBS and antibiotics) analyzed in a CK2 Olympus microscope: A. Embryoid body-like structures, magnification 40×; B. epithelioid-like phenotype, magnification 100×; C. neuronal-like phenotype, magnification 200×.
As can be seen in figure
NT2/D1 cells in IAM medium
NT2/D1 cells in IAM medium. Cell culture images of NT2/D1 cells after 20 days growing in control, IAM and IAM+BMP2 medium. All the non fluorescent and fluorescent images are 100× magnified and analyzed in a CK2 and BX50 Olympus microscope, respectively. Ultrastructural images have a scale bar in the bottom right corner. NT2/D1 cells formed many embryonic body-like structures (white arrows), surrounded by an epithelioid squamous layer of adherent flattened cells. Desmosomes can be seen between adjacent cells (TEM black arrow). To see non fluorescent and fluorescent images of days 1, 5, 10 and 30, please [see Additional file
Ultrastructural analysis
As mentioned above, some villosities could be detected in TEM (figure
Human pluripotent embryonal carcinoma NT2/D1 in an angiomyogenic medium. TEM images at differentiation day 20.
Click here for file
Transcriptional analysis
To see if would be any tendency of NT2/D1 to differentiate in a more muscle related or endothelial related fashion, genes specific for that tissues were chosen, ([see Additional file
Human pluripotent embryonal carcinoma NT2/D1 in an angiomyogenic medium. RT-PCR Qualitative analysis. This is not a quantitative analysis, what is the best choice to compare transcripts of different genes. However, a qualitative analysis can be done, comparing the intensity of the gel bands with the intensity of a house keeping gene, in our case the β actin. The scores we have chosen to construct these graphics were: 0 – PCR product not detected; 1 – PCR product with much less intensity of the β actin product; 2 – PCR product with less intensity of the β actin product; 3 – PCR product with the intensity of the β actin product; 4 – PCR product with more intensity of the β actin product; 5 – PCR product with much more intensity of the β actin product. Of course, that with this method the conclusions one can take are very limited, and only big differences have caught our attention. The main observations were: i) contrarily to what we have expected, in general, all genes were equally expressed in the different time points analyzed, with the exception of the
Click here for file
Expression studies
Expression studies. Polaroid images of 0,8% agarose gels with Ethidium Bromide of polymerase amplified gene products of NT2/D1 cells cDNA. In the first and last lanes one can see the 1 Kb ladder (Invitrogen). To see gel images of the days 1, 5, 10 and 30, or other details related to the amplified products and primers, please [see Additional file
Muscle related genes
At all time points analyzed, NT2/D1 cells expressed undifferentiated cardiac specific genes, like
Human pluripotent embryonal carcinoma NT2/D1 in an angiomyogenic medium. The case of the
Click here for file
Endothelial progenitor related genes
Transcripts of
Other progenitor related genes
Knowing that NT2/D1 expresses the
Bone morphogenetic protein receptors (BMPRs)
Because one of our main hypothesis was that BMP2 was an inducer of mesodermal differentiation in NT2/D1 cultures, we analyzed whether the main BMP receptor genes, namely
NT2/D1 related genes
Human pluripotent embryonal carcinoma NT2/D1 in an angiomyogenic medium. 100× images. Phase contrast, hystochemical and fluorescent images of NT2/D1 cells at time points 1, 5, 10, 20 and 30 days in Control, IAM and IAM+BMP2 culture mediums. Transcriptional analysis of 16 genes at the same time points is also shown.
Click here for file
As mentioned before, one of the genes highly expressed in P19 cells, a murine EC, with cardiac differentiation potential is
Discussion
Despite the difference of human and mouse EC at the level of cell surface markers, they share some common characteristics, such as, the morphology of little cytoplasm, prominent nucleoli and the preferential clustering growth pattern. In this line of thought, one can say that the mouse teratocarcinoma cell line, P19, has a relevant similar behavior to NT2/D1, also, regarding the potential of both to originate neuroectodermal derivatives in the presence of retinoic acid.
However, in the presence of RA P19 cells can still give rise to mesodermal derivatives, such as, smooth muscle, skeletal muscle and cardiac muscle
BMP2, belonging to the transforming growth factor superfamily, is involved in many important processes in early human and mouse development
For this, BMP2 has been considered to be an inhibitor of the neural differentiation program and an embryonic ectoderm
We also believe that BMP2 can indirectly induce NT2/D1 cells into mesodermal derivatives like cardiac muscle, because: 1) observations from amphibians, chicks and more recently from mice, have proposed that the factors produced by embryonic endoderm are good candidates to be cardiogenic inducers
Hence, one can say that the main potential of P19 cells to spontaneously differentiate into cardiac derivatives may come from the factors produced by the cells that undergone the visceral endoderm-like differentiation, thus, inducing other cells to the mesodermal lineage. So, we speculate that when the BMP2 is present it may contribute indirectly to cardiomyogenesis by increasing the number of visceral endoderm like cells in culture.
Then, if BMP2 can inhibit the neural fate and induce the mesodermal and the visceral endoderm-like phenotype in NT2/D1 cells it would possibly induce some cells to the cardiac lineage too. Unfortunately, morphological signals of muscle differentiation as myotube formation and sarcomeric intracellular actin organization were not detected, in an angiomyogenic inductive medium, in the presence of BMP2 or in 5-Aza treated cells, as we did not see any muscle beating areas at all the time points of the experiment, indicating that the NT2/D1 were not committed to mesodermal lineages like striated muscle, but instead, they differentiate mainly into the expected epithelioid morphologies. There were some cultures that exhibit some angiomyogenic like structures, but further studies are needed to confirm that.
However, transcripts from the cardiomyogenic (
Conclusion
We believe that the apparent blockage of the differentiation pathways
Another possibility is that the aneuploid state of NT2/D1 cells inhibits the key differentiating genes, and that what was once an advantage, regarding the tumor clonal selection of differentiation-resistant cells, is now a relevant obstacle in differentiation studies like this one.
Still, identifying the factors capable of regulating differentiation of pluripotent human EC cell lines, will give us clues for as to how manipulate human stem cells, an important raw material for
Methods
Cell culture
NT2/D1 cells (Stratagene) were cultured and expanded in 80 cm2 flasks in RPMI 1640 medium without Glutamine (Invitrogen), with 10 mM HEPES (Invitrogen), 20 μM 2-β-Mercaptoetanol (Sigma), 2 mM Glutamine (Invitrogen) 10% FBS (Invitrogen) and 1% (v/v) Penicillin 10000 IU/ml + Streptomycin 10 mg/ml (Invitrogen) and left for growth to 80% confluence, harvested by scraping and reseeded 1:4, continuously until they were needed.
In angiomyogenic induction, NT2/D1 cells were suspended in inductive angiomyogenic medium (IAM) with endothelial cell growth supplement (ECGS) as described by Shmelkov
Reverse Transcription Polymerase Chain Reaction (RT-PCR) Analysis
Total RNA was extracted according to Trizol total RNA extraction protocol (Invitrogen). cDNA first strand synthesis was performed on 1 μg total RNA, by using the ThermoScript RT-PCR System (Invitrogen) with an oligo dT primer. As a second step, cDNA samples were subjected to PCR amplification, in a separate tube, using primers specific for the gene of interest, [see Additional files
Diagram of the NT2/D1 Transcripts
Click here for file
Primer data and sizes of the PCR amplified products
Click here for file
Histochemistry
Glass coverslips, containing the NT2/D1 cells, were fixed for 30 min with the addition of buffered formaldehyde pH 7.0 into the cell culture wells at a 2% final concentration, washed in D-PBS pH 7,2–7,4 (Invitrogen) and stored at 4°C until processed. When processed, coverslips were washed in water, immersed in Harris's hematoxylin (Merck) for 10 minutes, washed in water, differentiated in 70% EtOH with 1% HCl for 10 seconds, immersed in eosin Y (Merck) for 10 seconds and washed again. Stained cells in coverslips were dehydrated in graded series of ethanol, followed by xylol (Merck), and mounted on glass microscope slides, cell side down, using entellan (Merck).
Phalloidin staining
Dilution of phalloidin-TRITC stock solution (Sigma) and washing of the glass coverslips were performed in D-PBS pH 7,2–7,4 containing 0,05% saponin (Sigma). Coverslips were fixed for 30 min with the addition of buffered formaldehyde pH 7.0 into the cell culture wells at a 2% final concentration, incubated with 1 μg/ml phalloidin-TRITC solution for 90 min at RT and washed 3 times. The coverslips were then mounted on microscope slides, cell side down, using Glicerol:D-PBS (1:1) and visualized in a Olympus BX50 microscope.
Ultrastructural analysis
Thermanox coverslips (Nunc) from day 20, were fixed in 0,1 M sodium cacodylate buffer (EMS) pH 7,3 with 3% gluteraldeyde (EMS) for 1 h, washed in 0,1 M sodium cacodylate, followed by 1% osmium tetroxide (EMS) in the same buffer for 1 h and washed again, first in cacodylate buffer and then in 0,1 M sodium acetate buffer (Merck) pH 5,0. Fixed cells were incubated with 1% uranyl acetate (EMS) in 0,1 M sodium acetate buffer pH 5,0 for 1 h, washed in the same buffer, dehydrated in graded series of ethanol, followed by propylene oxide (EMS), embedded in epon (EMS) and left for 3 days at 70°C. Thin sections cut on a LKB III microtome with a diamond knife for transmission electron microscopy (TEM) were stained on 300 mesh square copper grids with 2% uranyl acetate and lead citrate (Fluka) and observed in a Philips CN10 transmission electron microscope. All the steps were made at room temperature, except when otherwise documented.
Sequencing
The PCR products were purified using Qiaquick PCR purification kit (Quiagen) as described by the manufacturer and visualized by ethidium bromide after electrophoresis in a 1% agarose gel. A molecular weight standard a 1 kb DNA ladder (Invitrogen,) was used. All sequences were determined in an automated DNA capillary sequencer CEQ 2000-XL (Beckman Coulter, USA, in ICAT – Lisbon Faculty of Sciences Sequencing Services) by a dye-labeled dideoxy termination method (DTCS, Dye Terminator Cycle sequencer start kit, Beckman Coulter). Two sequencing reactions were performed using the two primers used for PCR amplification. The sequences obtained, were compared with existing gene data from the National Centre for Biotechnology Information (GenBank) for all the amplified genes. All the primers have given rise to the expected amplified fragments.
Authors' contributions
PDS conceived the study, carried out all the experimental procedures of the current work and drafted the manuscript. TR participated in the elaboration of the design of the study and has made substantial contributions to the analysis, interpretation of data and critically revising of the manuscript. All the authors have read and approved the final manuscript.