Background
The cancer stem cell theory proposes that a subpopulation of cells, the cancer stem cells, exist in solid tumors as well as cancers of hematopoietic origin and constitute a reservoir of self-sustaining cells with the exclusive ability to self-renew and maintain the tumor. These cancer stem cells have the capacity to both divide and expand the cancer stem cell pool and to differentiate into the heterogeneous non-tumorigenic cancer cell types that constitute the bulk of the cancer cells within the tumor [
1]. Cancer stem cells have been proposed to play a role in tumorigenesis and metastasis [
2‐
4] as well as in resistance to radiation and chemotherapy [
5‐
9]. Thus, it may be necessary to target and eliminate these cells to eradicate cancers.
Cancer stem cells have been identified in a mounting number of human malignancies. Using approaches developed in the hematopoietic malignancies, Clarke and colleagues demonstrated the existence of a subpopulation of tumorigenic cells (or tumor-initiating cells), isolated from breast cancer pleural effusions by limiting dilution transplantation of CD44+/CD24-/lineage- cells into the mammary fat pad of immunocompromised mice [
10]..Importantly, Dontu et al. adapted a sphere culture system described in the central nervous system [
11] by growing primary breast epithelial cells in serum-free medium containing EGF and FGF2 in suspension culture and demonstrated enrichment of the tumor-initiating population. This enriched population is suitable to demonstrate in vitro self-renewal and multi-lineage differentiation potential [
12]. This assay has been adapted to study cancer stem/progenitor cells in breast cancer cell lines as well as in colon, brain and pancreatic cancer tissues [
11,
13‐
18]. Another marker of the cancer stem cell population was recently established by Ginestier et al. who showed that aldehyde dehydrogenase 1 (ALDH1)-positive cells from human breast tumors were the tumorigenic cancer-initiating population [
19]. High ALDH activity has also been used to define cancer stem cell populations in other solid tumors including colon, liver and pancreatic cancers [
20‐
22]. Numerous studies suggest that cell lines can be useful in vitro tools to study these populations, and at this time the percentage of cancer stem cell markers has been characterized in numerous cancer cell lines [
15,
17]. These efforts have led to the use of appropriately selected cell lines rich in the surrogate phenotypes to screen for agents that target these cancer stem cell surrogates [
23]; however the number of lines expressing each of the surrogates reported thus far is few.
We observed several key biologic features of purported cancer stem cells in human embryonic kidney cells 293T (also designated HEK 293T), a widely used cell line derived through the sheared Adenovirus 5 (Ad5) DNA transformation [
24]. Most importantly, tumors from these cells self-renew as evidenced by serial xenograft passage through transplantation into the cleared mammary fat pad continued over eight generations. Guan et al previously showed that primary HEK 239T cells form tumors that could be inhibited by RbAP46 [
25], and Shen et al. demonstrated that tumor forming efficiency had increased with increasing passage of these cells [
26], but serial self-renewal has not been reported. We hypothesized that these embryonic, tumorigenic cells may be useful to study cancer stem cells and add to the arsenal of cell lines with significant cancer stem cell surrogate representation to facilitate the study of targeted agents against this population. Here we focused on the cancer stem cell surrogates described in breast tumor models and transplanted tumor cells into the cleared mammary fatpads to examine in vivo tumor pathology. We report that 293T cell readily form spheres that are enriched with cancer stem cell surrogates ALDH1 and CD44+/CD24- cells and express proteins associated with the epithelial-mesenchymal transition and stem cell developmental pathways. They are also depleted in the microRNA that has been implicated in cancer stem cell self-renewal. We propose that this is a valuable cell line to study cancer stem cell surrogates identified in human cancers as well as the plasticity of transformed embryonic cells.
Discussion
We propose 293T is a valuable tool to study cancer stem/progenitor features in that it rapidly generates both spheres and tumors that are readily serially passaged, and maintains significant populations of the tumor-initiating phenotypes described in the cancer stem cell literature. In this study, we observed that transplanting 293T cells into the cleared mammary fat pads of immunocompromised mice resulted in serially transplantable rapidly growing tumors. Propagation of spheres from tumor tissue or directly from the cell line enriched for ALDH activity and CD44+/CD24- markers and expression of EMT markers as well as genes that promote stem cell survival. Papers characterizing similar features in existing cancer cell lines have demonstrated a wide range in each of the purported cancer stem cell populations in these lines making the lines available to screen simultaneously against each surrogate limited.
The 293T cell line was developed through transformation of human embryonic kidney cells using the sheared Adenovirus 5 (Ad5) DNA [
24]. It expresses genes permitting the replication of E1-deficient, non-replicating adeno and adeno-associated viral vectors leading to their common use of generating viral vectors for laboratory and gene therapy use. From the initial report in 1977, the transformation was surprisingly inefficient. Only two of eight transformations were successful with only one of twenty cultures forming a single colony in each case. Attempts to isolate these were not successful, however and a single transformed colony was observed in one dish not discarded 75 days after the original experiment. Partial serum starvation (2%) was used to select for the transformed phenotype and eventually the isolated culture was expanded. The culture was reported to contain cells of several morphologies of generally epitheliod appearing cells. Injection of 2 × 10
7 cells subcutaneously into nude mice yielded no tumors after 6-8 weeks. Re-injection at that time resulted in tumors 15-20 weeks after the first injection in 3/20 animals. In stark contrast, Shen et al [
26] report that in cells passaged greater than 65 times, the tumorgenicity increases to 100% with tumors formed in two weeks. We have reproduced these findings with 100% tumor formation when injected into the cleared mammary fat pads of immunocompromised mice and further report that these tumors are serially transplantable with similarly aggressive biology and resembled poorly differentiated, high grade malignant primary tumor on histology. It's noteworthy that there is controversy on the origins of HEK 293T cells as it has been assumed that they were generated by transformation of either a fibroblastic, endothelial or epithelial cell all of which are abundant in kidney. However, the fact that the cells originated from cultured kidney cells does not say much about the exact cellular origin of the HEK 293T, as embryonic kidney cultures may contain small numbers of almost all cell types of the body. In fact, a study by Graham and colleagues provided evidence that these cells and several other human cell lines generated by adenovirus transformation of human embryonic kidney cells have many neuronal progenitor properties, suggesting that the adenovirus was taken up and transformed a neuronal lineage cell in the original kidney culture [
41]. Thus, the spheres generated from 293T cells under stem-cell promoting culture condition may have features observed in neurospheres.
Given the in vivo and histological findings, we explored known cancer stem cell surrogates and signaling pathways described in various cancer studies and confirm this cell line faithfully recapitulates the described biology of purported cancer stem cells. Previous studies showed that nuclear and cytoplasmic extracts from undifferentiated cells can reprogram 293T cells into a stem cell-like phenotype with activation of embryonic stem cell genes and outgrowth of spherical structures. As a result of treatment of 293T monolayer cells with embryonic carcinoma cell extract, spherical colonies of 293T cells develop and induce Oct4 gene expression which in turn induces the expression several Oct4-responsive genes like Sox2, Nanog and Rex1 [
42‐
44]. Consistent with these findings, our study has found activation of the pluripotency genes as well as formation of spheres in suspension culture suggesting that the stem cell promoting culture condition promotes dedifferentiation. The notion that cells with properties of cancer stem cells can be generated from differentiated somatic cells has been proposed in a recent study by Dean and colleagues [
45]. These authors demonstrated that culturing retinoblastoma-deficient mouse embryonic fibroblasts in suspension culture form spherical structures, re-express embryonic stem cell markers and at least a subset of this population adopts characteristics of cancer stem cells. Likewise, recent accounts from Weinberg's group [
46] and others [
47] suggest the possibility of reversion or reprogramming of differentiated cancer cells into cancer stem cells.
Several studies from our lab have shown that breast cancer stem/progenitor cells are resistant to radiation compared to bulk, differentiated cells [
9,
16,
48,
49]. Phillips and colleagues also demonstrated the radioresistance of putative breast cancer stem/progenitor cells by comparing the radiosensitivity of cells derived from the CD44+/CD24-subpopulation of MCF-7 cell line grown as spheres vs. monolayers [
8]. The resistance of cancer stem cells to radiation has been shown to be mediated by the β-catenin [
9,
49] or Notch signaling pathway [
8]. Similar findings were reported in brain tumors demonstrating that CD133-expressing glioma stem cells are relatively resistant to radiation in vivo and in culture compared to non-stem cells [
50]. Likewise, in our study the stem cell-enriched population from the 293T cells is much more resistant and with increased expression of β-catenin and Notch1 compared to the adherent population. Moreover, a study by Li et al demonstrated that the percentage of cancer stem cells, assayed by mammosphere forming efficiency, was significantly increased after chemotherapy highlighting the chemoresistant nature of cancer stem cells [
7].
The molecular basis underlying the EMT process involves multiple changes in expression and function of proteins that include vimentin and cadherins [
51‐
53]. Acquired expression of vimentin by carcinoma cells often symbolizes mesenchymal-like cell transformation [
52] while loss of E-cadherin or gain of N-cadherin on tumor cell surface is frequently observed in malignant carcinomas and also correlated with enhanced aggressiveness and dedifferentiation [
51]. Recent studies have demonstrated that induction of EMT using Twist and Snail in transformed mammary epithelial cells creates populations that are highly enriched for cancer stem cells as evidenced by increased CD44+/CD24- expression, mammosphere formation and tumor seeding ability [
18]. These cells showed attributes of mesenchymal phenotype including expression of vimentin, fibronectin and n-cadherin. Similarly, in a recent study in pancreatic cell lines the knockdown Zeb1, another EMT-inducer, affected not only the initial formation of spheres but also led to reduction of sphere numbers in subsequent generations indicating its role in self-renewal and maintenance of the stem cell phenotype[
13]. In our current study, we found higher expression of these EMT transcription factors in spheres from 293T cells vs. monolayer cells in line with the stem cell properties of 293T spheres.
MicroRNAs are emerging as important regulators of cellular differentiation and EMT. Recent findings have associated let-7 microRNAs with stem cell maintenance and loss of let-7 in cancer has been reported to cause dedifferentiation and self-renewal of breast cancer cells [
37]. Consistent with our finding that the 293T spheres have markedly reduced Let-7 expression and increased H-ras and HMGA2, two known let-7 targets these authors showed that cancer stem cell population have markedly reduced let-7 and increased H-ras and HMGA2 expression. By expressing let-7 in cancer stem cells, the authors also found that let-7 regulates self renewal, multipotent differentiation, and the ability to form tumors [
37]. Several recent studies have identified the miR-200 family and miR-205 as key regulators of EMT and enforcers of the epithelial phenotype of cancer cells [
38,
40]. MiR-200 was found to directly target the mRNA of the E-cadherin transcriptional repressors Zeb1 and Zeb2. While ectopic expression of miR-200 caused up-regulation of E-cadherin in cancer cell lines and reduced their motility, its inhibition reduced E-cadherin expression, increased expression of vimentin and induced EMT [
40]. Another microRNA, miR-335, has been identified as metastasis suppressor microRNA in human breast cancer [
39]. It inhibits metastatic cell invasion and migration through targeting of the progenitor cell transcription factor Sox4 and extracellular matrix component tenascin C. Moreover, the loss of expression of this microRNA is associated with poor distal metastasis-free survival [
39].
Methods
Cell culture
For sphere formation, 293T cells were placed in ultralow attachment dishes (Corning, NY) in mammosphere media as previously described by Dontu et al. [
12]. Briefly, single cells were seeded in 6-well ultralow attachment plates (Corning, NY) in serum free MEM media supplemented with 20 ng/ml bFGF, 20 ng/ml EGF, and B27 (Invitrogen). Adherent 293T cells were grown on standard tissue culture plate in growth media that contains DMEM with 10% fetal bovine serum and penicillin-streptomycin.
RT-PCR and miRNA analysis
Total RNA was isolated from cells 2D monolayer and 3D spheres with TRIzol reagent according to the manufacture's protocol. After treatment with DNase I (Ambion), two micrograms of the RNA samples were reverse-transcribed with random hexamers using Super Script III First-Strand Synthesis System from Invitrogen. Control reactions excluded reverse transcriptase. PCR was done with aliquots of the cDNA samples at the annealing temperature 60°C with the following primers: Oct4 forward 5'-ccagtatcgagaaccgagtgag-3', reverse 5'-gcatagtcgctgcttgatcg-3'; Nanog forward 5'-gatgcctcacacggagactg-3', reverse 5'-gctggggtaggtaggtgctg-3'; Sox2 forward 5'-tccacactcacgcaaaaacc-3', reverse 5'-ttgcaaacttcctgcaaagc-3'; Klf4 forward 5'-ttctcacctgtgtgggttcg-3', reverse 5'-cctgggtcttgaggaagtgc-3'; Snail forward 5'-cggaagcctaactacagcgag-3', reverse 5'-cttggcctcagagagctgg-3'; slug forward 5'-gatgcatattcggacccacac-3', reverse 5'-cctcatgtttgtgcaggagag-3'; Gapdh forward 5'-cacccagaagactgtggatgg-3', reverse 5'-ttctagacggcaggtcaggtc-3'. For microRNAs, small RNAs were extracted using a mirVana miRNA isolation kit (Ambion, Austin, TX) following the manufacture's protocol. Two micrograms of the total RNA sample was reverse-transcribed with RT miRNA First Strand Kit (SABiosciences, Frederick, MD). The cDNA was mixed with SYBR Green/ROX qPCR Master Mix and the mixture was added into a 96-well miRNA PCR Array (SABiosciences) that included primer pairs for 88 human miRNAs. PCR was performed on 7300 Real-Time PCR equipment (Applied Biosystems, Foster city, CA). Threshold cycle (Ct) for each miRNA was extracted using SDS2.3 software (Applied Biosystems) and data was analyzed using an Excel-based data analysis system (SA Bioscencies) that is based on the ΔΔCt method. Normalization of the raw data was made to one or more of the housekeeping genes in the Array.
Immunoblotting
Cell extracts were prepared from 2D monolayer and 3D sphere cultures of 293T cells using cell lysis buffer (Cell Signaling, MA), supplemented with protease inhibitor (1 mM phenylmethanesulfonyl fluoride, PMSF). Protein concentrations were determined using the Bio-Rad Protein Assay (Bio-rad, CA) and 40 ug of each sample was separated using 4-20% Tris-HCl polyacrylamide gel (Invitrogen, CA) and transferred onto a PVDF membrane from Bio-Rad. The membranes were blocked in 5% milk powder in TBS-T for ~1 hour and were then incubated with the following primary antibodies: anti-E-cadherin, anti-N-cadherin, and anti-Vimentin (BD Biosciences); anti-zeb1 and anti-Rex1 (Abcam); anti-β catenin and anti-survivin (Cell Signal); and anti-notch 1 (Santa Cruz). Membranes were washed three times and incubated with the corresponding secondary antibody conjugated with a horseradish peroxidase in 5% nonfat milk at room temperature. After incubation with secondary antibody, the membranes were washed three times and immunoreactivity was detected by enhanced chemiluminescence. Anti-actin and anti-vincullin (Sigma) antibodies were used as loading controls.
Flow Cytometry
For analyzing the CD44/CD24 subpopulation, 293T cells cultured under sphere and adherent conditions were harvested with 10 mM EDTA, centrifuged and resuspended in phosphate buffered saline (PBS) (105 cells/ml). They were then incubated with FITC-conjugated CD44 and PE-conjugated CD24 antibodies (BD Biosciences, San Diego, CA, USA) for 30 minutes at concentrations recommended by the manufacturer. Cells incubated in PBS, FITC or PE alone served as controls. Cell analysis for the expression of CD44 and CD24 was performed using a Beckman Coulter machine and the data files were analyzed using FlowJo software (Treestar, Ashland, OR). For aldehyde dehydrogenase (ALDH) activity assay, the Aldefluor kit (Lonza) was used according to the manufacturer's instructions. Briefly, about 5 × 105 2D monolayer and 3D sphere cultures of 293T cells were suspended in Aldefluor assay buffer containing ALDH substrate and incubated for 30 min at 37°C. As a negative control for each sample, a sample of cells was incubated with 50 mmol/L of the specific ALDH inhibitor diethylaminobenzaldehyde (DEAB). Aldefluor fluorescence was excited at 488 nm and fluorescence emission was detected using a Beckman Coulter machine and the data files were analyzed using FlowJo software (Treestar, Ashland, OR).
Label retention assay
Cells were trypsinized, resuspendend in PBS, and labeled with PKH26 (Sigma). Labeled cells were plated in suspension in ultralow attachment plates in mammosphere culture media [
12]. After 7 days, spheres were harvested, enzymatically dissociated and subjected to serial passages without further addition of PKH26. Intact spheres from the serial passages were subjected to PKH26 staining and were viewed by confocal microscopy.
Clonogenic assay/3D sphere formation assay
This was conducted as we previously described [
16]. Briefly, monolayer cultures of 293T cells were trypsinized into single cells and were seeded into individual wells of a 6-well tissue culture plate (for 2D) or ultralow attachment plates (for 3D). Both 2D and 3D 6-well plates containing seeded single cells were exposed to increasing doses of irradiation 4 hrs after plating. 2D plates were incubated for 14 days and colonies were stained with crystal violet while 3D cells were incubated in mammosphere media for 7 days, the spheres were stained with MTT and those with a size of 50 uM were counted using a Gelcount colony counter (Oxford Optronix, Oxford, UK). Survival curves were generated using Sigmaplot 8.0.
Transplantation
2×10
6 fresh cells from monolayer culture were mixed with matrigel (1:1), and injected directly into the epithelium-free cleared fat pads of three-week-old female SCID/Beige recipient mice [
54]. Mice were palpated weekly and tumor growth measured using calipers. When tumors reached 10 mm in diameter, they were collected for further analysis. Fresh tumor fragments were minced into 1 mm
3 fragments, and re-transplanted into new SCID/Beige hosts for subsequent transplant generations, and expanded as transplantable tumor tissue lines.
Immunostaining procedure
Immunostaining was performed in a autostainer (Vision Biosystems, Norwell, MA) using Bond max protocol. The primary antibodies that were used in the study included E-cadherin (HECD-1, 1:100, Zymed/Invitrogen), β-catenin (1:500, BD Biosciences), VEGF (A-20; 1:10, Santa Cruz Biotechnology) and Keratin 19 (RCK 108, 1:50, Dako). The slides were dewaxed, rinsed with alcohol for dehydration and antigen retrieval was performed using citrate buffer. The slides were then treated with 3% H2O2, and primary antibodies were applied on the tissue sections and incubated for 10 minutes. The slides were subsequently treated with Poly-horseradish peroxidase, anti-mouse/rabbit IgGs. For visualization of the signal, 3, 3'diaminobenzidine (DAB) was used as the chromogen. Hematoxylin counterstaining was used for visualization of the nucleus.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
BGD, XZ, AAR, MDL, WX carried out experiments; collection and/or assembly of data by BGD, WX, AAR, MDL, HG, EC, LLi; data analysis and interpretation by BGD, LL, WX, SK, FR, JMR, AL, NTU, MTL, WAW; Financial support by WAW, MC, TAB, MTL; Manuscript writing by BGD, WAW; Final approval of manuscript by all authors.