Background
It is known for decades that tumor tissue resembles chronically inflamed tissue - a matter why tumors have been referred to as "wounds that do not heal" [
1,
2]. Since chronic inflammation is a strong stimulus for the recruitment of BMDCs [
3‐
5] it can be concluded that the migration of BMDCs into tumor tissues is a common process.
Several lines of evidence indicated that BMDCs, including macrophages and mesenchymal stem cells (MSCs), can trigger tumor growth and metastasis [
6‐
8]. It is assumed that BMDCs can promote a proglycolytic phenotype in tumor cells, thus giving them a survival advantage in hypoxic and inflammatory conditions [
9], or promote tumor cell survival through the activation of the integrin-linked kinase (ILK), thereby activating the prosurvival AKT signaling pathway [
10]. Another mechanism has been described by Karnoub and colleagues by demonstrating that breast cancer cells stimulated the de novo secretion of the chemokine CCL5 (also named RANTES) from MSCs, which then acted in a paracrine fashion on the cancer cells to enhance their motility, invasion and metastasis [
7].
In addition to these mechanisms, which are based on the intercellular communication between BMDCs and tumor cells, cell fusion and horizontal gene transfer (HGT) have also been associated with BMDC mediated promotion of tumor growth and metastasis. Data of Rizvi et al. indicated that BMDCs can fuse with neoplastic intestinal epithelial cells, thereby giving rise to stable tumor cell/BMDC hybrids [
4]. Similar data were recently provided by Powell and colleagues, whereby it was shown that not only macrophages, but also T- and B-cells can fuse with tumor cells [
11]. Moreover, gene expression analysis revealed a nuclear reprogramming in hybrid cells [
11]. Even though both studies did not investigate whether BMDC/tumor cell hybrids do contribute to tumor progression at all, several data of the past years indicated that cell fusion is a common phenomenon in cancer and that hybrid cells possess novel characteristics, which can be definitely linked to tumor progression (for review see [
12,
13]). These properties include tissue heterogeneity [
14,
15], an increased malignancy [
16,
17], drug resistance [
12,
16] and an enhanced resistance of tumor cells to apoptosis [
18]. Additionally, cell fusion has also been suggested as one mechanism how cancer stem cells might originate [
13,
19,
20].
In contrast to animal studies, here it was shown that BMDC × melanoma cell hybrids, either originated by artificial
in vitro cell fusion or by spontaneous
in vivo cell fusion, exhibited a markedly enhanced metastatic capacity [
21,
22], the available data records for "cell fusion in human tumors and the possible outcome" are controversial. Putative cell fusion events have been reported for renal cell carcinoma [
23,
24], breast cancer [
25,
26], rectal cancer [
27], and multiple myeloma [
28,
29]. While data of Shabo and colleagues revealed that expression of the macrophage receptor CD163 on breast cancer and rectal carcinoma cells was generally associated with a poorer outcome of afflicted patients, Larsson et al. reported that syncytin expression constitutes as a positive prognostic factor in breast cancer, possibly due to mediating fusions between breast cancer cells and endothelial cells [
30].
HGT has also been suggested as a mechanism how tumor cells could receive foreign DNA concomitant with rendering the cells malignancy (for review see: [
31]). Several lines of evidence indicated that HGT of chromosomal DNA between tumor cells and tumor cells or tumor cells and normal cells is efficiently facilitated through the uptake of apoptotic bodies [
32]. HGT of tumor DNA to endothelial cells
in vivo gave rise to endothelial cells maintaining the ability to form functional vessels and concurrently express tumor-encoded and endothelial-specific genes [
33]. Likewise,
in vivo gene transfer between interacting human osteosarcoma cell lines was associated with acquisition of an enhanced metastatic potential [
34].
In the present study we investigated the outcome of co-cultivation of murine mammary 67NR-Hyg breast cancer cells and murine BMDCs derived from Tg(GFPU)5Nagy/J mice [
35]. Our data show that mBMDC/67NR-Hyg clones possessed a marked up-regulation of Abcb1a/b ATP binding cassette (ABC) multidrug resistance transporters, which was correlated to an enhanced resistance of these cells towards chemotherapeutic drugs.
Discussion
In the present study we investigated the possible outcome of co-cultivation of murine mammary 67NR-Hyg breast cancer cells and murine BMDCs derived from Tg(GFPU)5Nagy/J mice [
35]. The rationale of this study was given by the fact that tumor tissue resembles chronically inflamed tissue (tumors are often described as wounds that do not heal) [
1,
2,
40] and that chronic inflammation is a strong stimulus for the recruitment of BMDCs [
3‐
5]. This correlation suggests that the recruitment of BMDCs into tumor tissues is a common process.
Here, co-cultivation of murine BMDCs with 67NR-Hyg mouse mammary carcinoma cells resulted in the origin of cells exhibiting markedly increased expression levels of the ABC multidrug resistance transporters Abcb1a and Abcb1b concomitant with an enhanced resistance towards chemotherapeutic drugs.
The finding that each of the analyzed mBMDC/67NR-Hyg clones was positive for EGFP suggests that the clones have originated by cell fusion. By contrast, both STR and SNP analyses revealed that only parental 67NR-Hyg alleles were present in mBMDC/67NR-Hyg clones, thus facing the question whether mBMDC/67NR-Hyg clones rather originated by horizontal gene transfer (HGT). 67NR-Hyg cells might have uptaken DNA containing microvesicles or apoptotic bodies shedded from mBMDCs. In a recent work Ehnfors and colleagues demonstrated in an
in vivo setting that endothelial cells can take up tumor cell DNA via HGT [
33] suggesting that a transfer of mBMDC DNA to 67NR-Hyg cells by HGT should be feasible. On the other hand, HGT is a random process. It can not be predicted which DNA fragments are embedded in microvesicles/apoptotic bodies, which of these vesicles/bodies will be uptaken by target cells and whether the uptaken DNA will be ultimately incorporated into the genome of the target cells. Because of that it remains ambiguous why all mBMDC/67NR-Hyg clones were positive for EGFP.
Another argument which would favor HGT as the mechanism of mBMDC/67NR-Hyg clone origin might be the fact that the mean chromosomal number of the clones was not the sum of the parental chromosomes. The mean chromosomal number of mBMDC/67NR-Hyg clones varied between 50 ± 8 to 66 ± 12 chromosomes, which was rather half of the sum of the mean chromosomal number of the parental cells. In fact, recent studies demonstrated that cell fusion commonly went along with a summation of parental chromosomes in hybrid cells [
16,
17,
41]. For instance, we have recently demonstrated that breast epithelial/breast cancer hybrid cells, which derived from spontaneous fusion events, possessed a mean chromosomal number of 78 ± 11 (M13MDA435 clone 1) to 88 ± 13 (M13MDA435 clone 2), which was nearly the sum of the parental mean chromosomal number [
20]. On the contrary, the transition of a hybrid cell from a heterokaryon (two nuclei) to a synkaryon (one nucleus) is generally associated with a loss of chromosomes [
19]. Likewise, the aneuploid karyotype of most hybrid cells is unstable and because of that chromosomes are unequally segregated or even lost during further cell divisions - a process, which has been termed autocatalytic karyotypic evolution [
42]. Another possibility that could explain the phenomenon of a reduced mean chromosomal number in mBMDC/67NR-Hyg clone might be ploidy reduction [
43]. This process has been recently reported for murine fusion-derived hepatocytes [
43], whereby the mechanisms that direct such a "meiosis-like" effect in somatic cells still remains elusive. In any case, the authors observed a high degree of marker loss in diploid daughter cells indicating that ploidy reductions lead to the generation of highly genetically diverse daughter cells with about 50% reduction in nuclear content [
43,
44]. This dynamic model of hepatocyte polyploidization, ploidy reversal and aneuploidy (which has been referred to as "ploidy conveyor") has likely evolved to generate genetic diversity, thereby permitting adaptation of hepatocytes to xenobiotic or nutritional injury [
44]. In context of our study, such a mechanism would be a suitable explanation for the finding that mBMDC/67NR-Hyg clones possessed a reduced mean chromosomal number concomitant with an increased drug resistance.
Thus, the question whether mBMDC/67NR-Hyg clones ultimately originated by cell fusion or by HGT can not be answered sufficiently. Both mechanisms are conceivable. Recent data of Duncan et al. let assume that the analyzed mBMDC/67NR-Hyg cells might have originated by cell fusion accompanied by subsequent ploidy reduction and autocatalytic karyotypic evolution. Nonetheless, HGT can not be ruled out completely.
The cytotoxicity data presented here nicely fitted to the ABC multidrug transporter expression levels of the analyzed cells. Drug efflux is chiefly mediated by ABC multidrug transporter(s), whereby a single drug can be exported by several ABC multidrug transporters, and each ABC multidrug transporter can confer characteristics resistance pattern to cells [
45]. By using the power of combination knockout mice, being deficient for distinct murine ABC multidrug transporters, Lagas et al. provided an update for murine ABC multidrug transporter and substrate specificities [
46,
47]. Thereby, etoposide resistance is mediated by Abcb1a/b, Abcc1, Abcc2, and Abcc3, whereas resistance to both doxorubicin and paclitaxel is facilitated by Abcb1a/b and Abcc2 [
46,
47]. Thus the increased Abcb1a/b expression levels of mBMDC/67NR-Hyg clones correlate well to the cells resistance towards doxorubicin, which is further strengthen by reverting the cells resistance with the ABC1B (PGP/MDR1) inhibitor verapamil. Data showing that resistance towards 17-DMAG (clone 1 and clone 3), etoposide and paclitaxel was solely partially blocked by verapamil suggests that further ABC multidrug transporter(s) or other mechanism do contribute to the cells resistance towards these chemotherapeutic compounds. Paclitaxel resistance has been associated with Abcb1a/b and Abcc2 [
47]. Since all mBMDC/67NR-hyg clones possessed slightly higher Abcc2 expression levels and Abcc2 has been reported to be the main transporter for biliary excretion [
47] it can be assumed that Abcc2 contribute to paclitaxel resistance in the presence of verapamil. By contrast, Abcb1a/b are the main transporters for biliary excretion of doxorubicin, and Abcc2 only partly compensates for the absence of Abcb1a/b [
47], which nicely correlates with the finding that doxorubicin resistance of mBMDC/67NR-Hyg clones was completely abrogated by verapamil. Likewise, etoposide resistance has been correlated to Abcb1a/b, Abcc1, Abcc2, and Abcc3 [
46,
47]. Since only Abcc2 was slightly higher expressed in mBMDC/67NR-Hyg clones as compared to parental murine 67NR-Hyg carcinoma cells, which further correlates with the relative survival rates of mBMDC/67NR-Hyg clones in comparison to 67NR-Hyg cells, we conclude that etoposide is also effluxed by this ABC multidrug transporter. Transfection of mouse fibroblasts with murine Abcg2 resulted in an increased etoposide resistance of the cells [
48] suggesting that also the increased Abcg2 levels in mBMDC/67NR-Hyg clones might have contributed to the enhanced etoposide resistance.
In addition to the basal ABC multidrug resistance transporter expression levels further mechanisms might also contribute to the enhanced drug resistance of mBMDC/67NR-Hyg clones. RealTime-PCR-array data showed that several cytochrome p450 family members were slightly higher expressed in mBMDC/67NR-Hyg clones than in parental cells suggesting that drug inactivation might be another mechanism. Likewise, due to the aneuploid karyotype mBMDC/67NR-Hyg cells will be able to adapt to the cytotoxic conditions. This assumption is in view with preliminary data showing that mBMDC/67NR-Hyg clones growing in 10 μM doxorubicin exhibited an altered phenotype, such as an increased cell diameter.
Materials and methods
Cell culture and transfection
The murine mammary carcinoma cell line 67NR was purchased from the American Tissue Culture Collection (ATCC, LGC Standards GmbH, Wesel, Germany) and was maintained in Dulbecco's Modified Eagle's Medium (DMEM, PAA, Linz, Austria) supplemented with 10% fetal calf serum (FCS; PAA, Linz, Austria) and 1% Penicillin/Streptomycin (100 U/ml Penicillin, 0.1 mg/ml Streptomycin; PAA Laboratories, Linz, Austria) at 37°C and 5% CO2 in a humidified atmosphere. Stable transfection of 67NR cells with the pKS-Hyg plasmid was done by electroporation using the AMAXA Nucleofector Technology (Lonza Cologne AG, Cologne, Germany) in accordance to the manufacturers' instructions. Hygromycin resistant cells were selected by addition of 200 μg/ml Hygromycin B (PAA Laboratories, Linz, Austria) to the media.
Murine bone marrow-derived cells (mBMDCs) were obtained from the femurs of 8 - 9 week-old female transgenic Tg(GFPU)5Nagy/J mice (Jackson Laboratories, Bar Habor, Maine, USA) expressing the enhanced green fluorescent protein (EGFP) and the puromycin-resistance gene [
35]. Bone marrow cells were collected by flushing the bone shaft with Iscove's Modified Dulbecco's Medium (IMDM; PAA, Linz, Austria) using a syringe and a 26G needle. The cell suspension was seeded into a 75 cm
2 tissue culture flask and cultivated for 24h in DMEM (PAA, Linz, Austria) supplemented with 10% FCS (PAA, Linz, Austria) and 1% Penicillin/Streptomycin (100 U/ml Penicillin, 0.1 mg/ml Streptomycin; PAA Laboratories, Linz, Austria) at 37°C and 5% CO
2 in a humidified atmosphere. Subsequently, non-adherent cells were removed and the remained mBMDCs were cultivated for additional two weeks. Cells were passaged once prior for their use in experiments.
Co-cultivation of mBMDC and 67NR-Hyg mammary carcinoma cell
Murine BMDCs (1 × 106) and 67NR-Hyg mouse mammary cancer cells (1 × 106) were co-cultivated for 24h in DMEM (PAA, Linz, Austria) supplemented with 10% FCS (PAA, Linz, Austria) and 1% Penicillin/Streptomycin (100 U/ml Penicillin, 0.1 mg/ml Streptomycin; PAA Laboratories, Linz, Austria) without Hygromycin and Puromycin at 37°C and 5% CO2 in a humidified atmosphere. After 24h both antibiotics were added to the media (Hygromycin B: 200 μg/ml; Puromycin: 5 μg/ml; Puromycin was purchased from Sigma Aldrich, Taufkirchen, Germany). Double resistant clones were first analyzed for a green fluorescence, then isolated and cultivated separately from each other. Isolated clones were named as mBMDC/67NR-clone-X, whereby "X" marks the clone number.
RT-PCR
RNA was isolated from 1 × 10
6 cells by using the NucleoSpin
® RNA II Kit (Macherey-Nagel GmbH, Düren, Germany) in accordance to the manufacturers' instructions. Reverse Transcription of RNA into cDNA was performed using the RevertAid
TM First Strand cDNA Synthesis Kit (Fermentas, St. Leon-Rot, Germany) as referred to the instruction manual. PCR was performed in a 25 μl reaction mixture containing 1.25U Taq Polymerase, 1 × reaction buffer, 2 mM MgCl2, 200 μM of each dNTP (all reagents were purchased from Fermentas, St. Leon-Rot, Germany) and 100 pM primers (Invitrogen, Karlsruhe, Germany). The cycling conditions comprised of an initial denaturation of 5 min at 94°C and 30 cycles of 0.5 min at 94°C, 0.5 min at the appropriate annealing temperature and 0.5 min at 72°C followed by a final elongation for 10 min at 72°C. The used primer pairs concomitant with their specific annealing temperature and product lengths are summarized in Table
3.
Table 3
Summary of primer pairs for PCR
Ccr7 | 61°C | 208 bp | forward | AGCACCATGGACCCAGGGA |
| | | reverse | CTGCCTCTCATGTATTCTGT |
Cxcr4 | 57°C | 390 bp | forward | GGCTGTAGAGCGAGTATTGC |
| | | reverse | GTAGAGGTTGGTGACAGTGTAGAT |
Cxcl12 | 61°C | 512 bp | forward | ACACTCCGCCATAGCATATGGT |
| | | reverse | TGAAGCATGCGTTTGGAGG |
EGFP | 61.5°C | 220 bp | forward | GACAAGCAGAAGAACGGCATCAAG |
| | | reverse | CGGCGGCGGTCACGAACT |
HYG | 60.5°C | 500 bp | forward | AGCTGCGCCGATGGTTTCTACAA |
| | | reverse | ATCGCCTCGCTCCAGTCAATG |
Syn A | 60°C | 281 bp | forward | TACCTGATGCGCCTGGAGCT |
| | | reverse | AAGCTTTGCAGGAACTGGAGAA |
Syn B | 60°C | 201 bp | forward | CCACCACCCATACGTTCAAA |
| | | reverse | GGTTATAGCAGGTGCCGAAG |
Slc1a5 | 59°C | 585 bp | forward | CTGGATTATGTGGTACGCCAC |
| | | reverse | GACCTGTCCACTAGCCAGTC |
Dhfr | 59°C | 159 bp | forward | CCACAACCTCTTCAGTGGAAGGTAAACAGA |
| | | reverse | TTGGCAAGAAAATGAGCTCCTCGTGG |
Gapdh | 68°C | 980 bp | forward | TGAAGGTCGGTGTGAACGGATTTGGC |
| | | reverse | CATGTAGGCCATGAGGTCCACCAC |
Short tandem repeat (STR) and Single nucleotide polymorphism (SNP) analysis
Genomic DNA was extracted by using the NucleoSpin
® Tissue Kit (Macherey-Nagel GmbH, Düren, Germany) in accordance to the manufacturers' instructions. Amplification of DNA fragments for STR and SNP analysis was performed by conventional PCR according to above-mentioned protocol, whereby 25 ng genomic DNA were used as a template. Suitable STRs and SNPs for analysis were determined in accordance to the mouse genome information strains, SNPs & polymorphisms database (
http://www.informatics.jax.org/strains_SNPs.shtml) and the genetic quality control annual report (
http://jaxmice.jax.org/geneticquality/gqcreport.pdf). The used primer pairs and product lengths are summarized in additional file
3 and additional file
4. For STR analysis PCR products (1 μl) were mixed with 9.5 μl Hi-Di
TM Formamide and 0.5 μl GeneScan™-500 LIZ
® size standard (Applied Biosystems, Darmstadt, Germany) in a 96-well microtiter plate. After heating the loading cocktail for 3 min at 95°C samples were immediately chilled on ice. Samples were analyzed by capillary electrophoresis for 2,500s by using a 3130 × l ABI PRISM Genetic Analyzer (Applied Biosystems, Darmstadt, Germany). Data analysis was performed with the GeneMapper (v 4.0) software (Applied Biosystems, Darmstadt, Germany). For SNP analysis PCR products were purified with the QIAquick
®PCR purification kit (QIAGEN, Hilden, Germany) as recommended. PCR products (1-10 ng) were sequenced with the BigDye
®Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems, Darmstadt, Germany) in accordance to the manufacturer's instructions. Not incorporated labeled ddNTPs were removed with the DyeEx
® 2.0 spin kit (QIAGEN, Hilden, Germany) as described in the manual. Purified sequencing products were analyzed by using a 3130 × l ABI PRISM Genetic Analyzer (Applied Biosystems, Darmstadt, Germany). Data analysis was performed with the ABI PRISM 3130 data collection software (Applied Biosystems, Darmstadt, Germany) and the ClustalW2-Multiple Sequencing Alignment tool (European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, UK).
Chromosome spreading
Cells (1 × 106) were cultivated with 0.2 μg/ml colcemid (Sigma Aldrich, Taufkirchen, Germany) for 4 - 6h. Subsequently, cells were harvested, washed once with PBS, and were carefully resuspended in 10 ml 75 mM KCl. After 30 min the cells were sedimented (160 × g, 10 min) and the supernatant was discarded. Cells were carefully resuspended in the remaining KCl solution. Hereafter, 10 ml methanol/acetic acid solution (3:1; both chemicals were purchased from Sigma Aldrich, Taufkirchen, Germany) was added dropwise under continuous stirring to the cells. Cells were then washed at least twice in methanol/acetic acid solution. Finally, the methanol/acetic acid fixed cells were dropped onto a H2O wetted cover slip. To visualize spread chromosomal DNA cover slips were air-dried and subsequently stained with Sytox Green (Invitrogen, Karlsruhe, Germany) as recommended to the manufacturers' instructions combined with confocal laser scanning microscopy (Leica TCS SP5; Leica, Bensheim, Germany).
RealTime-PCR
RNA was isolated from 2-4 × 106 cells by using the NucleoSpin® RNA II Kit from Macherey-Nagel (Macherey-Nagel GmbH, Düren, Germany) in accordance to the manufacturers' instructions. Reverse Transcription of RNA into cDNA was performed using the RT2 First Strand Kit (QIAGEN GmbH, Hilden, Germany) as referred to the instruction manual. In this study, the RT2ProfilerTM PCR Array "Mouse Cancer Drug Resistance and Metabolism" (QIAGEN GmbH, Hilden, Germany) covering 84 genes was applied by using an Applied Biosystems 7700 RealTime-PCR cycler (Applied Biosystems, Darmstadt, Germany). RealTime-PCR was performed using the RT2 SYBR Green Master Mix (QIAGEN GmbH, Hilden, Germany) according to the manufacturers' protocol under the following cycler conditions: 95°C: 10 min; 40 Cycles (95°C: 15s; 60°C: 60s). RealTime-PCR data were analyzed by the 2-ΔCT method on a MicrosoftTM Excel® template provided by the manufacturer (QIAGEN GmbH, Hilden, Germany). Significant changes in gene expression among the analyzed cell lines were defined as at least 2-fold up- or downregulation of genes.
Western Blot
Western Blot analysis
To verify RealTime-PCR data for Abcb1a and Abcb1b expression cells (5 × 10
5) were lysed in SDS sample buffer for 1h at 36°C. Subsequently, samples were separated by SDS-PAGE on a 8% SDS polyacrylamide gel and transferred to PVDF membranes (Millipore, Schwalbach, Germany) under semi-dry conditions. Membranes were blocked overnight with 10% (w/v) non-fat dry milk in TBS-T. Abcb1a/Abcb1b and β-actin were detected by using the following primary antibodies: MDR (Abcb1a/Abcb1b; clone C-19; Santa Cruz Biotechnology, Heidelberg, Germany) and β-actin (clone 13E5; rabbit monoclonal, Cell Signaling, New England Biolabs, Frankfurt am Main, Germany). For detection of primary antibodies the HRP-conjugated secondary anti-rabbit IgG (Cell Signaling, New England Biolabs, Frankfurt am Main, Germany) was used. Bands were visualized using the LumiGLO
® Reagent (Cell Signaling, New England Biolabs, Frankfurt am Main, Germany) in accordance to the manufacturers' instructions and were detected with the Aequoria Macroscopic Imaging system (Hamamatsu Photonics Germany, Herrsching am Ammersee, Germany). Relative densities of Abcb1a/b expression in relation to β-actin was determined by using the ImageJ software (
http://rsb.info.nih.gov/ij/).
Flow cytometry
The efflux of Rhodamine 123 (Sigma Aldrich Taufkirchen, Germany) was analyzed by using a FACScalibur flow cytometer (Becton Dickenson, Heidelberg, Germany). Cells (5 × 105) were incubated in PBS containing 200 μg/ml Rhodamine 123, 50 μM verapamil (Sigma Aldrich Taufkirchen, Germany), or a combination of both for 20 min at 37°C. Thereafter, cells were washed once with PBS and incubated for additional 20 min at 37°C. Rhodamine 123 fluorescence was detected using the FL1-H channel.
XTT cytotoxicity assay
The XTT cytotoxicity assay was performed as described [
49] with slight modifications. In brief, cells (1 × 10
3/well) were seeded in triplicates in a 96-well flat-bottom microtiter plate in 0.25 ml of the appropriate culture medium. After 24 h media was replaced by culture media containing different concentrations of either 5-Fluorouracil (5-FU), 17-DMAG, doxorubicin, etoposide or paclitaxel (all drugs were purchased from Sigma Aldrich, Taufkirchen, Germany). In case of verapamil, 1 μM of the appropriate chemotherapeutic drug was used alone or in combination with 50 μM verapamil (Sigma Aldrich, Taufkirchen, Germany). After 48h media was removed and the plates were analyzed with XTT reagent (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions. The absorption of the formed XTT-formazan derivative was measured using a BioTek EL800 microplate reader (BioTek, Bad Friedrichshall, Germany). The EGFP fluorescence of mBMDCs and mBMDC/67NR-Hyg hybrids did not interfere with the XTT-formazan formed derivative. Statistical significance was calculated using Student's
t-test: n.s. = not significant; * =
P < 0.05; ** =
P < 0.01; *** =
P < 0.001;
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
CN performed the experiments. CH performed the STR and SNP analyses and corrected the manuscript. KSZ wrote and corrected the manuscript. TD designed the experiments, wrote and corrected the manuscript. All authors have read and approved the final manuscript.