Hybrid Cells Derived from Human Breast Cancer Cells and Human Breast Epithelial Cells Exhibit Differential TLR4 and TLR9 Signaling

TLRs are important receptors of cells of the innate immune system since they recognize various structurally conserved molecular patterns of different pathogens as well as endogenous ligands. In cancer, the role of TLRs is still controversial due to findings that both regression and progression of tumors could depend on TLR signaling. In the present study, M13SV1-EGFP-Neo human breast epithelial cells, MDA-MB-435-Hyg human breast cancer cells and two hybrids M13MDA435-1 and -3 were investigated for TLR4 and TLR9 expression and signaling. RT-PCR data revealed that LPS and CpG-ODN induced the expression of pro-inflammatory cytokines, like IFN-β, TNF-α, IL-1β and IL-6 in hybrid cells, but not parental cells. Interestingly, validation of RT-PCR data by Western blot showed detectable protein levels solely after LPS stimulation, suggesting that regulatory mechanisms are also controlled by TLR signaling. Analysis of pAKT and pERK1/2 levels upon LPS and CpG-ODN stimulation revealed a differential phosphorylation pattern in all cells. Finally, the migratory behavior of the cells was investigated showing that both LPS and CpG-ODN potently blocked the locomotory activity of the hybrid cells in a dose-dependent manner. In summary, hybrid cells exhibit differential TLR4 and TLR9 signaling.


Introduction
TLRs are type I transmembrane receptors that belong to the innate immune system [1][2][3]. They are chiefly expressed by immune competent cells, like macrophages, dendritic cells, B-and T-lymphocytes, and they do recognize structurally conserved pathogen derived molecules, so-called PAMPs [1][2][3]. To date, 10 different TLRs have been identified in humans so far, each possessing a specificity for a certain ligand or ligands. For instance, TLR4 recognizes bacterial lipopolysaccharides and TLR5 bacterial flagella, whereas TLR9 binds unmethylated CpG-DNA of bacterial origin [1][2][3]. However, within the past few years it has become evident that TLRs also recognize endogenous ligands, also named DAMPs, such as extracellular matrix components (TLR4), heat shock proteins (TLR2/4), HMGB1 (TLR4) and self DNA (TLR9) [2]. Usually, the recognition of DAMPs plays a crucial role in identifying and removing cell debris and inducing wound healing in response to non-pathogen-induced tissue damage [4]. Due to recognizing intracellular proteins to be released and presented by necrotic, apoptotic and/or necroptotic cells, DAMPs and its receptors play a crucial role in chronic inflammatory diseases including rheumatoid arthritis [5] and even cancer [6]. It is well recognized that the tumor microenvironment resembles chronically inflamed tissue and thus cancer has been called "wounds that do not heal" [7].
However, the role of TLRs in cancer is much debated due to contradictory reports. Some studies provided evidence that TLR ligands, such as LPS from Gram negative bacteria or CpG-ODN, might be efficacious in the treatment of various cancer types, including colorectal cancer, glioblastoma, hepatocellular carcinoma and myeloma [8][9][10][11][12]. On the contrary, several reports demonstrated that TLR expression of cancer cells might be rather associated with tumor progression. For instance, LPS could induce epithelial-to-mesenchymal transition (EMT) in cancer cells and has been associated with cancer cell invasion and metastasis in a TLR4 dependent manner [13][14][15]. Moreover, TLR4 expression in breast cancer and ovarian cancer has been correlated to paclitaxel chemoresistance [16,17]. Likewise, TLR9 agonists like CpG-ODN or even DNA from dead cells could promote cancer cell invasion [18][19][20]. TLR9 has also been suggested as a prognostic factor in breast cancer, whereas Tuomela et al. demonstrated that rather low TLR9 levels define an aggressive subtype of triple-negative breast cancer [21].
Cell fusion has been suggested as a driving force in cancer progression because a plethora of data provided evidence that hybrid cells derived from tumor cells and tumor cells or tumor cells and normal cells, like macrophages [22,23] or epithelial cells [24][25][26], could exhibit novel properties, such as an enhanced drug resistance or an increased metastatic activity (for review see: [27][28][29][30][31]). We thus investigated M13MDA435-1 and -3 hybrid cells in comparison to their parental cells (human M13SV1-EGFP-Neo breast epithelial cells and human MDA-MB-435-Hyg breast cancer cells) [25,32] for TLR expression and signaling. We have recently demonstrated that LPS potently induced apoptosis in M13MDA435 hybrid cell clones, but not in parental cells [33]. Here, we additionally investigated the cells for TLR9 expression and signaling.

M13MDA435 Hybrid Cells Respond Differently to CpG-ODN and LPS Stimulation
In accordance to recently published data MDA-MB-435-Hyg human breast cancer cells and M13MDA435-1 and -3 hybrid cells exhibited comparable expression levels of TLR4, TLR9, TRIF, Myd88, and TRAF6 ( Figure 1A). In contrast to this, the expression of these proteins was rather moderate to low in M13SV1-EGFP-Neo human breast epithelial cells ( Figure 1A). However, all cell lines showed comparable IRAK1 expression levels ( Figure 1A). Stimulation of cells with either 100 ng/mL CpG-ODN and 100 ng/mL LPS, respectively, revealed a differential IRF7 and NF-κB activation. In accordance with recently published data [33], LPS treatment resulted in NF-κB activation in both hybrid cells, but not in parental cells ( Figure 1B). On the contrary, stimulation of cells with 100 ng/mL CpG-ODN did not activate NF-κB signaling in all cell lines ( Figure 1B). Analysis of IRF7 activation upon CpG-ODN and LPS stimulation indicated that the cells responded differently. While in M13SV1-EGFP-Neo cells and M13MDA435-3 hybrid cells both CpG-ODN and LPS stimulation resulted in IRF7 activation, no nuclear translocation of this transcription factor was detected in MDA-MB-435-Hyg breast cancer cells and M13MDA435-1 hybrid cells ( Figure 1B).

M13MDA435 Hybrid Cells and Parental Cells Respond Differently to LPS and CpG-ODN Stimulation
Next, the expression of NF-κB and IRF7 target genes was investigated by RT-PCR and Western blot in MDA-MB-435-Hyg human breast cancer cells, M13SV1-EGFP-Neo human breast epithelial cells and M13MDA435-1 and -3 hybrid cells that were stimulated with LPS and CpG-ODN for 2, 6, 12, 24 and 48 h.
We recently demonstrated that LPS stimulation lead to the induction of TNF-α and IFN-β in M13MDA435-1 and -3 hybrid cells, but not parental cells [33], which could be reproduced in this study ( Figure 2). Even though both hybrid cell lines responded similarly in the overall expression of target genes in response to LPS stimulation, they differed markedly in the kinetics of gene/protein expression ( Figure 2). For instance, a permanent LPS induced expression of IL-1β was detected in M13MDA435-1 hybrid cells on both mRNA and protein, whereas in M13MDA435-3 hybrid cells, IL-1β expression was solely detected after 2 to 6 h of LPS treatment ( Figure 2). A marked expression of TNF-α was detected in both hybrid cell lines after 2 h of LPS stimulation ( Figure 2). On the contrary, RT-PCR data showed a sustained of IFN-β expression in both hybrid cell lines after LPS stimulation with a first peak maximum after 2 h and a second one after 12 h (Figure 2A). However, IFN-β protein expression in dependence of LPS was solely detected after 2 h in the hybrid cells ( Figure 2B). Differential mRNA and protein levels were further observed for IL-6, MCP-1 and TRAIL in both hybrid cell lines. For instance, rather sustained IL-6 and TRAIL mRNA levels in response to LPS stimulation were detected in M13MDA435-1 hybrid cells, whereas the peak maximum of IL-6 protein expression was observed between 2 h and 6 h ( Figure 2B). Prolonged stimulation of M13MDA435-1 hybrid cells with LPS was correlated with a decreasing IL-6 protein expression ( Figure 2B). Likewise, TRAIL protein expression was solely observed between 6 h to 12 h in LPS treated M13MDA453-1 hybrid cells ( Figure 2B). contrary, RT-PCR data showed a sustained of IFN-β expression in both hybrid cell lines after LPS stimulation with a first peak maximum after 2 h and a second one after 12 h ( Figure 2A). However, IFN-β protein expression in dependence of LPS was solely detected after 2 h in the hybrid cells ( Figure 2B). Differential mRNA and protein levels were further observed for IL-6, MCP-1 and TRAIL in both hybrid cell lines. For instance, rather sustained IL-6 and TRAIL mRNA levels in response to LPS stimulation were detected in M13MDA435-1 hybrid cells, whereas the peak maximum of IL-6 protein expression was observed between 2 h and 6 h ( Figure 2B). Prolonged stimulation of M13MDA435-1 hybrid cells with LPS was correlated with a decreasing IL-6 protein expression ( Figure 2B). Likewise, TRAIL protein expression was solely observed between 6 h to 12 h in LPS treated M13MDA453-1 hybrid cells ( Figure 2B). RT-PCR analysis revealed that both parental cells responded differentially to LPS stimulation with the expression of appropriate target genes. In MDA-MB-435-Hyg breast cancer cells increased mRNA levels of MCP-1 and TRAIL were observed after LPS stimulation (Figure 2A). Increased mRNA levels of IL-6, MCP-1 and TRAIL were observed in LPS treated M13SV1-EGFP-Neo breast epithelial cells (Figure 2A) indicating that these cells exhibit a functional TLR4 signaling despite rather low expression levels of TLR4 and components of the TLR4 signal transduction cascade ( Figure 1A). In accordance to previously published data [33], neither TNF-α nor IFN-β expression was induced in LPS treated parental cell lines. However, in contrast to M13MDA435-1 and -3 hybrid cells, virtually no target gene protein expression was observed in LPS stimulated parental cells ( Figure 2B). Only a slight TRAIL expression was detected in M13SV1-EGFP-Neo breast epithelial cells treated for 48 h with LPS ( Figure 2B). RT-PCR analysis revealed that both parental cells responded differentially to LPS stimulation with the expression of appropriate target genes. In MDA-MB-435-Hyg breast cancer cells increased mRNA levels of MCP-1 and TRAIL were observed after LPS stimulation (Figure 2A). Increased mRNA levels of IL-6, MCP-1 and TRAIL were observed in LPS treated M13SV1-EGFP-Neo breast epithelial cells ( Figure 2A) indicating that these cells exhibit a functional TLR4 signaling despite rather low expression levels of TLR4 and components of the TLR4 signal transduction cascade ( Figure 1A). In accordance to previously published data [33], neither TNF-α nor IFN-β expression was induced in LPS treated parental cell lines. However, in contrast to M13MDA435-1 and -3 hybrid cells, virtually no target gene protein expression was observed in LPS stimulated parental cells ( Figure 2B). Only a slight TRAIL expression was detected in M13SV1-EGFP-Neo breast epithelial cells treated for 48 h with LPS ( Figure 2B). It is well recognized that the expression of proteins is also controlled by miRNAs or lncRNAs, which either block expression by degrading the mRNA or by blocking translation at the ribosome [34]. We thus conclude that the non-expression of target genes despite detectable mRNA levels might be attributed to such as a regulatory mechanism. This may particularly apply for MCP-1. Increased MCP-1 mRNA levels in response to LPS stimulation were found in all investigated cell lines, whereas no MCP-1 protein expression was observed ( Figure 2). Treatment of cells with CpG-ODN was markedly different from LPS stimulation. Albeit CpG-ODN stimulation resulted in increased target gene mRNA levels in both hybrid cell lines and the M13SV1-EGFP-Neo breast epithelial cell line ( Figure 3A) all analyzed target genes, with exception of TRAIL, were not expressed on a protein level ( Figure 3B). On the one hand, these data indicate that the cells exhibit functional TLR9 signaling, but at present it remains unknown why a protein translation did not occur in CpG-ODN-treated cells. As mentioned above, we conclude that most likely miRNA or lncRNA based regulatory processes are responsible for this observation. Results show that the cell lines responded uniquely to LPS stimulation, which particularly applies for protein translation. Whereas in M13MDA435-1 and -3 hybrid cells most of the transcribed genes are expressed as proteins, protein expression is lacking in parental cells, which is putatively attributed to miRNA or lncRNA dependent mechanisms. Shown are representative RT-PCR and Western Blot data of three independent experiments.
It is well recognized that the expression of proteins is also controlled by miRNAs or lncRNAs, which either block expression by degrading the mRNA or by blocking translation at the ribosome [34]. We thus conclude that the non-expression of target genes despite detectable mRNA levels might be attributed to such as a regulatory mechanism. This may particularly apply for MCP-1. Increased MCP-1 mRNA levels in response to LPS stimulation were found in all investigated cell lines, whereas no MCP-1 protein expression was observed ( Figure 2).
Treatment of cells with CpG-ODN was markedly different from LPS stimulation. Albeit CpG-ODN stimulation resulted in increased target gene mRNA levels in both hybrid cell lines and the M13SV1-EGFP-Neo breast epithelial cell line ( Figure 3A) all analyzed target genes, with exception of TRAIL, were not expressed on a protein level ( Figure 3B). On the one hand, these data indicate that the cells exhibit functional TLR9 signaling, but at present it remains unknown why a protein translation did not occur in CpG-ODN-treated cells. As mentioned above, we conclude that most likely miRNA or lncRNA based regulatory processes are responsible for this observation.

Expression of Pro-Inflammatory and Apoptosis-Inducing Cytokines in Hybrid Cells Does Not Correlate with BAX and BCL-2 Expression Levels
We have recently demonstrated that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells in an IFN-β dependent manner [33]. Because of that BAX and BCL-2 protein levels were determined by Western Blot of both CpG-ODN and LPS treated cells ( Figure 4). The pro-apoptotic acting protein BAX was induced by both CpG-ODN and LPS in MDA-MB-435-Hyg breast cancer cells, whereas only LPS stimulation caused increased BAX levels in M13SV1-EGFP-Neo breast epithelial cells ( Figure 4A,B). Interestingly, M13SV1-EGFP-Neo breast epithelial cells lack BCL-2 expression ( Figure 4A,B). In contrast to parental cells BAX expression levels remained unaltered in both CpG-ODN and LPS treated M13MDA435-1 and -3 hybrid cell clones ( Figure 4A,B). Likewise, a slight to moderate up-regulation of BCL-2 was observed in LPS treated hybrid cells ( Figure 4A,B). This, however, is contrary to our previously published data showing that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells, but not parental cells [33]. It is well recognized that BCL-2 belongs to the group of anti-apoptotic proteins [35].

Expression of Pro-Inflammatory and Apoptosis-Inducing Cytokines in Hybrid Cells Does Not Correlate with BAX and BCL-2 Expression Levels
We have recently demonstrated that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells in an IFN-β dependent manner [33]. Because of that BAX and BCL-2 protein levels were determined by Western Blot of both CpG-ODN and LPS treated cells ( Figure 4). The pro-apoptotic acting protein BAX was induced by both CpG-ODN and LPS in MDA-MB-435-Hyg breast cancer cells, whereas only LPS stimulation caused increased BAX levels in M13SV1-EGFP-Neo breast epithelial cells ( Figure 4A,B). Interestingly, M13SV1-EGFP-Neo breast epithelial cells lack BCL-2 expression ( Figure 4A,B). In contrast to parental cells BAX expression levels remained unaltered in both CpG-ODN and LPS treated M13MDA435-1 and -3 hybrid cell clones ( Figure 4A,B). Likewise, a slight to moderate up-regulation of BCL-2 was observed in LPS treated hybrid cells ( Figure 4A,B). This, however, is contrary to our previously published data showing that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells, but not parental cells [33]. It is well recognized that BCL-2 belongs to the group of anti-apoptotic proteins [35].

CpG-ODN and LPS Both Induce AKT and ERK1/2 Signaling in Parental Cells and Hybrid Cells
It is well recognized that stimulation of TLR4 and TLR9 also result in the engagement of MAPK signaling via TRAF6 mediated activation on TAB2/TAB3/TAK1 [3]. Conjointly, Bauerfeld et al. demonstrated that TLR4 signaling could also activate AKT via a Myd88/TRIF dependent mechanism [36]. Similar findings were reported for TLR9 signaling [37]. We thus investigated whether stimulation of MDA-MB-435-Hyg human breast cancer cells, M13SV1-EGFP-Neo human breast epithelial cells and their hybrids resulted in activation of AKT and ERK1/2 signaling. In fact, increased pAKT (S473 and T308) and pERK1/2 levels were detected in all cell lines upon CpG-ODN and LPS stimulation ( Figure 5A,B), whereby the analyzed cell lines differed markedly in the kinetics of AKT and ERK1/2 phosphorylation. In parental cells, both CpG and LPS resulted in increased pAKT S473 levels, but not pAKT T308 levels ( Figure 5A,B). On the contrary, increased pAKT S473 and pAKT T308 levels were detected in both hybrid cell lines stimulated with CpG and LPS ( Figure 5A,B). Phosphorylation of AKT at positions T308 and S473 is regulated via different signal transduction pathways [38]. While phosphorylation of the activation loop of AKT at T308 is facilitated by PDK1 in a PI3K dependent manner, phosphorylation of AKT on the hydrophobic motif S473 is mediated by mTORC2 [38]. Thus, two different AKT activating pathways are engaged in hybrid cells by TLR4 and TLR9 signaling. Interestingly, pAKT S473 levels were rather low to moderate in MDA-MB-435-Hyg human breast cancer cells, whereas in M13SV1-EGFP-Neo human breast epithelial cells higher pAKT S473 levels were detected with a maximum after 6 and 12 h ( Figure 5A,B). On the contrary, markedly higher pAKT S473 levels were detected in M13MDA435-1 and -3 hybrid cells. Comparison of the kinetics of T308 and S473 AKT phosphorylation in M13MDA435-1 and -3 hybrid cells revealed a rather identical phosphorylation pattern in M13MDA435-3 hybrid cells ( Figure 5A,B). Here, markedly increased pAKT T308 and S473 levels were detected after 6 and 24 h of CpG-ODN and LPS stimulation ( Figure 5A,B). On the contrary, in CpG treated M13MDA435-1 hybrid cells highest pAKT S473 were observed after 2 to 6 h, whereas the maximum peak level of pAKT T308 was observed after 24 h ( Figure 5A). Both parental cell lines showed a similar ERK1/2 phosphorylation in response to both CpG-ODN and LPS. Increased pERK1/2 were observed after 2, 6 and 48 h of stimulation with either CpG-ODN and LPS, respectively ( Figure 5A,B). On the contrary, both hybrid cell clones exhibited a unique ERK1/2 activation profile in response to CpG-ODN and LPS stimulation. For instance, a CpG-ODN dependent ERK1/2 phosphorylation was first detected after 48 h in M13MDA435-1 hybrids, whereas in LPS treated cells pERK1/2 were observed after 24 h ( Figure 5A,B). On the

CpG-ODN and LPS Both Induce AKT and ERK1/2 Signaling in Parental Cells and Hybrid Cells
It is well recognized that stimulation of TLR4 and TLR9 also result in the engagement of MAPK signaling via TRAF6 mediated activation on TAB2/TAB3/TAK1 [3]. Conjointly, Bauerfeld et al. demonstrated that TLR4 signaling could also activate AKT via a Myd88/TRIF dependent mechanism [36]. Similar findings were reported for TLR9 signaling [37]. We thus investigated whether stimulation of MDA-MB-435-Hyg human breast cancer cells, M13SV1-EGFP-Neo human breast epithelial cells and their hybrids resulted in activation of AKT and ERK1/2 signaling. In fact, increased pAKT (S473 and T308) and pERK1/2 levels were detected in all cell lines upon CpG-ODN and LPS stimulation ( Figure 5A,B), whereby the analyzed cell lines differed markedly in the kinetics of AKT and ERK1/2 phosphorylation. In parental cells, both CpG and LPS resulted in increased pAKT S473 levels, but not pAKT T308 levels ( Figure 5A,B). On the contrary, increased pAKT S473 and pAKT T308 levels were detected in both hybrid cell lines stimulated with CpG and LPS ( Figure 5A,B). Phosphorylation of AKT at positions T308 and S473 is regulated via different signal transduction pathways [38]. While phosphorylation of the activation loop of AKT at T308 is facilitated by PDK1 in a PI3K dependent manner, phosphorylation of AKT on the hydrophobic motif S473 is mediated by mTORC2 [38]. Thus, two different AKT activating pathways are engaged in hybrid cells by TLR4 and TLR9 signaling. Interestingly, pAKT S473 levels were rather low to moderate in MDA-MB-435-Hyg human breast cancer cells, whereas in M13SV1-EGFP-Neo human breast epithelial cells higher pAKT S473 levels were detected with a maximum after 6 and 12 h ( Figure 5A,B). On the contrary, markedly higher pAKT S473 levels were detected in M13MDA435-1 and -3 hybrid cells. Comparison of the kinetics of T308 and S473 AKT phosphorylation in M13MDA435-1 and -3 hybrid cells revealed a rather identical phosphorylation pattern in M13MDA435-3 hybrid cells ( Figure 5A,B). Here, markedly increased pAKT T308 and S473 levels were detected after 6 and 24 h of CpG-ODN and LPS stimulation ( Figure 5A,B). On the contrary, in CpG treated M13MDA435-1 hybrid cells highest pAKT S473 were observed after 2 to 6 h, whereas the maximum peak level of pAKT T308 was observed after 24 h ( Figure 5A). Both parental cell lines showed a similar ERK1/2 phosphorylation in response to both CpG-ODN and LPS. Increased pERK1/2 were observed after 2, 6 and 48 h of stimulation with either CpG-ODN and LPS, respectively ( Figure 5A,B). On the contrary, both hybrid cell clones exhibited a unique ERK1/2 activation profile in response to CpG-ODN and LPS stimulation. For instance, a CpG-ODN dependent ERK1/2 phosphorylation was first detected after 48 h in M13MDA435-1 hybrids, whereas in LPS treated cells pERK1/2 were observed after 24 h ( Figure 5A,B). On the contrary, ERK1/2 phosphorylation in response to CpG-ODN stimulation was detected after 6, 12, 24 and 48 h in M13MDA435-3 hybrid cells ( Figure 5A). Likewise, increased LPS-mediated pERK1/2 levels were found after 6 and 48 h ( Figure 5B). These data show that hybrid cell clones exhibit unique kinetics of AKT and ERK1/2 activation in response to CpG-ODN and LPS stimulation. contrary, ERK1/2 phosphorylation in response to CpG-ODN stimulation was detected after 6, 12, 24 and 48 h in M13MDA435-3 hybrid cells ( Figure 5A). Likewise, increased LPS-mediated pERK1/2 levels were found after 6 and 48 h ( Figure 5B). These data show that hybrid cell clones exhibit unique kinetics of AKT and ERK1/2 activation in response to CpG-ODN and LPS stimulation.

The Migratory Activity of Parental Cells and Hybrid Cells is Impaired by CpG-ODN in a Dose-Dependent Manner
TLR4 and TLR9 signaling have been associated with breast cancer progression due to findings revealing that both LPS and CpG-ODN could promote breast cancer cell migration, invasion and metastatic spreading [13,15,20]. We thus analyzed the cells locomotory activity within a 3D collagen matrix in dependence of different CpG-ODN and LPS concentrations. However, in contrast to studies providing evidence that LPS and CpG-ODN could foster breast cancer cell migration our migration data rather indicated an inhibitory effect of both compounds on the cells motility. The migratory activities of MDA-MB-435-Hyg human breast cancer cells and M13MDA435-1 and -3 hybrid cells were impaired by LPS in a dose dependent manner, whereas the migratory activity of M13SV1-EGFP-Neo breast epithelial cells remained unaffected in the presence of LPS ( Figure 6). On the contrary, CpG-ODN inhibited the migratory activity of all investigated cell lines in a dose-dependent manner ( Figure 6). As shown recently, LPS potently induced apoptosis in M13MDA4351 and -3 hybrid cells, but not parental cells [33], suggesting that the decreased migratory activity of hybrid cells within the presence of LPS might be attributed to an increased number of apoptotic cells. However, the viability of MDA-MB-435-Hyg cells was not affected by LPS, which also applies to CpG-ODN for all cell lines. Thus, the means by which LPS and CpG-ODN impairs the migration of the cells remains to be elucidated.

The Migratory Activity of Parental Cells and Hybrid Cells is Impaired by CpG-ODN in a Dose-Dependent Manner
TLR4 and TLR9 signaling have been associated with breast cancer progression due to findings revealing that both LPS and CpG-ODN could promote breast cancer cell migration, invasion and metastatic spreading [13,15,20]. We thus analyzed the cells locomotory activity within a 3D collagen matrix in dependence of different CpG-ODN and LPS concentrations. However, in contrast to studies providing evidence that LPS and CpG-ODN could foster breast cancer cell migration our migration data rather indicated an inhibitory effect of both compounds on the cells motility. The migratory activities of MDA-MB-435-Hyg human breast cancer cells and M13MDA435-1 and -3 hybrid cells were impaired by LPS in a dose dependent manner, whereas the migratory activity of M13SV1-EGFP-Neo breast epithelial cells remained unaffected in the presence of LPS ( Figure 6). On the contrary, CpG-ODN inhibited the migratory activity of all investigated cell lines in a dose-dependent manner ( Figure 6). As shown recently, LPS potently induced apoptosis in M13MDA4351 and -3 hybrid cells, but not parental cells [33], suggesting that the decreased migratory activity of hybrid cells within the presence of LPS might be attributed to an increased number of apoptotic cells. However, the viability of MDA-MB-435-Hyg cells was not affected by LPS, which also applies to CpG-ODN for all cell lines. Thus, the means by which LPS and CpG-ODN impairs the migration of the cells remains to be elucidated.

Discussion
In the present study, human M13SV1-EGFP-Neo breast epithelial cells, human MDA-MB-435-Hyg and two of their hybrids M13MDA435-1 and -3 [25,32] were analyzed for TLR4 and TLR9 expression and signaling. Our data show that although all cell lines express both receptors, they differed markedly in the kinetics of TLR specific signal transduction cascades and target gene expression. On the contrary, stimulation of TLR4 or TLR9 signaling with LPS or CpG-ODN, respectively, resulted in a decreased locomotory activity of all investigated cells.
The finding that LPS stimulation resulted in a time-dependent expression of LPS target genes, like IL-1β, IL-6 and TRAIL in M13MDA435 hybrid cells, but not parental cells, is in line with previously published data [33]. Of interest in this context is the correlation of RT-PCR data and Western Blot data revealing marked differences. For instance, LPS-induced and sustained TRAIL mRNA levels were detectable in M13MDA435-1 hybrid cells for up to 48 h, whereas a maximum of TRAIL expression on a protein level was observed between 6 and 12 h (Figures 2B and 3B). Similar findings were found for LPS-induced TRAIL expression in M13SV1-EGFP-Neo cells. Here, an increased TRAIL expression was found upon 48 h of LPS stimulation while TRAIL mRNA levels were already detected after 2 h of LPS stimulation (Figures 2B and 3B). Likewise, MCP-1 expression was solely detected by RT-PCR, but not Western blot analysis (Figures 2B and 3B), which

Discussion
In the present study, human M13SV1-EGFP-Neo breast epithelial cells, human MDA-MB-435-Hyg and two of their hybrids M13MDA435-1 and -3 [25,32] were analyzed for TLR4 and TLR9 expression and signaling. Our data show that although all cell lines express both receptors, they differed markedly in the kinetics of TLR specific signal transduction cascades and target gene expression. On the contrary, stimulation of TLR4 or TLR9 signaling with LPS or CpG-ODN, respectively, resulted in a decreased locomotory activity of all investigated cells.
The finding that LPS stimulation resulted in a time-dependent expression of LPS target genes, like IL-1β, IL-6 and TRAIL in M13MDA435 hybrid cells, but not parental cells, is in line with previously published data [33]. Of interest in this context is the correlation of RT-PCR data and Western Blot data revealing marked differences. For instance, LPS-induced and sustained TRAIL mRNA levels were detectable in M13MDA435-1 hybrid cells for up to 48 h, whereas a maximum of TRAIL expression on a protein level was observed between 6 and 12 h (Figures 2B and 3B). Similar findings were found for LPS-induced TRAIL expression in M13SV1-EGFP-Neo cells. Here, an increased TRAIL expression was found upon 48 h of LPS stimulation while TRAIL mRNA levels were already detected after 2 h of LPS stimulation (Figures 2B and 3B). Likewise, MCP-1 expression was solely detected by RT-PCR, but not Western blot analysis (Figures 2B and 3B), which particularly applies for IFN-β, IL-1β, and IL-6 expression in CpG-ODN treated hybrid cells (Figures 2B and 3B). These findings indicate that LPS and CpG-ODN not only induce target gene expression by activation of transcription factors, but most likely also by modulating the expression or the activation state of regulating molecules, like miRNA or lncRNA. Both, miRNAs and lncRNAs are well-known modulators of gene expression [39,40].
The finding that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells via an IFN-β dependent mechanism [33] still remains ambiguous, which also applies for the role of LPS and TLR4 signaling in breast cancer. As mentioned in the introduction, TLRs do not only recognize PAMPs, but also DAMPs, which are derived from necrotic, apoptotic and necroptotic cells [4]. It would thus be of interest to study the impact of endogenous TLR4 ligands, like extracellular matrix components or heat shock proteins, on parental cells and hybrid cells and whether they do also induce a differential TLR4 signaling concomitant with a differential expression of specific target genes and induction of apoptosis in hybrid cells as compared to parental cells. In this context, it would be of interest to investigate whether necrotic, apoptotic and necroptotic cell-derived DAMPs could stimulate cells in a paracrine manner.
Stimulation of M13MDA435-1 and -3 hybrid cells with LPS resulted in the expression of pro-inflammatory cytokines including IL-1β, TNF-α, and IL-6. Whether the expression levels of these cytokines would be sufficient to activate macrophages or to direct macrophage differentiation towards a "M1-macrophage" or "classical activated macrophage" phenotype [4,41] remains to be elucidated. Tumor-associated macrophages have been identified as a double-edged sword in cancer progression [42]. "M1-macrophage" or "classical activated macrophage", which are commonly present in acute inflammatory conditions [4,41,43,44], have been generally associated with "tumor rejection" and thus a better prognosis [41,45]. On the contrary, "M2-macrophages" or "wound-healing/resolution macrophages" secrete a variety of growth factors, like EGF, FGF and VEGF, and immunosuppressive factors including IL-10, PGE 2 and TGF-β, thus providing a tumor-friendly micromilieu concomitant with a much worse prognosis for the afflicted patients [41,[43][44][45]. However, whether LPS (or more likely endogenous TLR4 ligands) will cause a rather acute inflammatory tumor micromilieu remains unclear. As shown here, induction of pro-inflammatory cytokines was solely induced in hybrid cells, but not parental cells. Thus, more breast cancer cells as well as breast cancerˆnormal cell hybrids have to be analyzed first. Conjointly, the role of IL-6 in breast cancer has to be further clarified since recent studies revealed that IL-6 appears to play a critical role in the growth and metastasis of breast cancer cells, renewal of breast cancer stem cells and drug resistance of breast cancer stem cells [46]. Thus, activation of TLR4 signaling by LPS or endogenous ligands may foster the renewal and expansion of breast cancer stem cells. It would also be worthwhile to speculate about the role of TNF-α, which is not only a pro-inflammatory cytokine, but which has also been determined as a pro-fusogenic factor [47][48][49]. For instance, we have recently demonstrated that the fusion of M13SV1 breast epithelial cells and MDA-MB-231 and MDA-MB-435 breast cancer cells is promoted by TNF-α [47]. Thus, secretion of TNF-α either induced by LPS or endogenous ligands could promote the fusion of tumor cells and other cells.
In accordance to the differential expression of LPS and CpG-ODN target genes both compounds also induced a differential AKT and ERK1/2 signaling in the cells. Interestingly, both parental cells showed a similar ERK1/2 activation pattern, with a peak maximum after 2, 6 and 48 h of CpG-ODN and LPS stimulation, whereas in M13MDA435-3 hybrid cells the peak maximum of ERK1/2 phosphorylation was detected after 6 h and 48 h. Conjointly, AKT S473 and AKT T308 phosphorylation induced by CpG and LPS stimulation was solely detected in M13MDA435-1 and -3 hybrid cells, but not parental cells. It is well recognized that AKT is a central signaling node being involved in various cellular functions including metabolism, proliferation and survival [38]. Western blot data demonstrated that LPS induced AKT phosphorylation both at position S473 and T308 ( Figure 5B), representing the two activation sites of AKT [38]. Thus, stimulation of M13MDA435-1 and -3 hybrid cells with LPS should result in full AKT activation concomitant with the induction of AKT mediated cellular processes including survival [38]. However, as shown recently, LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells [33]. It can thus be concluded that two opposing signal transduction pathways regulating cell survival are engaged by LPS in M13MDA435-1 and -3 hybrid cells. This would also apply to the transcription factor NF-κB. On the one hand, LPS induced TLR4 signaling leads to nuclear translocation of NF-κB concomitant with IFN-β expression and induction of apoptosis [33]. These data are in view with findings of Jung and colleagues demonstrating that the LPS induced apoptosis of cultured microglia cells was not only dependent on IFN-β, but was also dependent on NF-κB activation mediating NO synthesis (via inducible NO synthase (iNOS) induction, caspase-11 induction and its subsequent activation [50]. On the contrary, survival of endothelial cells is mediated by TNF-α in a NF-κB dependent mechanism, whereby constitutive or inducible NF-κB-independent pathway(s) protects HUVECs from cell death [51]. Likewise, a link between Bcl-2 and NF-κB signaling and suppression of apoptosis in ventricular myocytes has been reported [52], suggesting that NF-κB plays a dual role in both inducing and preventing apoptosis. The finding that LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid cells albeit pro-survival pathways are simultaneously engaged might be thus attributed to a differential kinetics and strength of the different TLR4 induced pathways. In M13MDA435-1 and -3 hybrid cells activation of AKT (and possibly NF-κB) may thus not be strong enough to counteract LPS induced pro-apoptotic pathways, thus preventing cell death.
Whether the differential AKT activation in M13MDA435-1 and -3 hybrids in comparison to the parental cells might also be attributed to differential receptor expression levels remains to be elucidated. As shown in Figure 1A M13SV1-EGFP-Neo human breast epithelial cells expressed markedly lower levels of TLR4, TLR9, TRIF, Myd88 and TRAF6 suggesting that rather weak AKT T308 levels as well as the lower induction of target gene expression in response to LPS and CpG (Figures 2 and 3) might be attributed to the cells overall lower receptor and signaling protein expression. However, markedly enhanced pAKT S473 levels were detected in LPS and CpG treated M13SV1-EGFP-Neo cells indicating that despite lower TLR4 and TLR9 expression levels activation of AKT at position S473 is potently induced. Moreover, TLR4 and TLR9 as well as TRIF, Myd88 and TRAF6 expression levels of MDA-MB-435-Hyg breast cancer cells were comparable to M13MDA435-1 and -3 hybrid cells. Nonetheless, rather weak AKT S473 to rather low AKT T308 levels were observed in LPS and CpG treated MDA-MB-435-Hyg cells, revealing that the activation of signal transduction cascades does not only depend on the expression levels of receptors and signal transduction proteins, but most likely also to other regulatory mechanisms.
Cell migration studies revealed that the locomotory activities of the parental cell lines and hybrid cell clones were significantly blocked by LPS and CpG-ODN, which is contrary to findings revealing that both LPS and CpG-ODN could promote tumor cell invasion and even metastasis formation [13][14][15][18][19][20]53]. On the contrary, other studies pointed out that e.g., silencing of TLR4 increased tumor progression and lung metastases in a murine model of breast cancer [54] and that TLR9 agonists could induce apoptosis in A20 lymphoma cells [55] and neuroblastoma cells [56]. Because LPS potently induced apoptosis in M13MDA435-1 and -3 hybrid, we assume that the LPS-impaired migratory behavior of the cells was rather attributed to an increased number of apoptotic cells in the assay. However, it cannot be ruled out completely that LPS might induce signal transduction pathways and/or might cause an altered gene expression profile that ultimately impairs the migratory activity of the cells. This assumption would be in line with the finding that the locomotory activity of MDA-MB-435-Hyg breast cancer cells was significantly impaired within the presence of 150 ng/mL LPS. As published recently, the viability of MDA-MB-435-Hyg was not affected by LPS [33]. Because the cell viability was also not affected by CpG-ODN, we conclude that the TLR9 dependent inhibition of cell migration was as well attributed to intracellular signaling and/or an altered gene expression profile caused by CpG-ODN induced TLR9 signaling. In fact, various data revealed that TLR agonists could inhibit chemotaxis of monocytes through inhibition of receptor expression [57] or endocytic clearance of receptors from the plasma membrane [58]. Additionally, TLR signaling could impair monocyte chemotaxis independent of receptor downmodulation through synergized effects of p38 MAPK and global Rap-1 activation, which ultimately abolished actin reorganization [59]. Whether a similar mechanism might be responsible for the CpG-ODN dependent inhibition of the migration of the investigated cells remains to be elucidated.

RT-PCR
Total RNA was isolated from cells (1ˆ10 6 ) by using the NucleoSpinR ® RNA Kit II (Macherey-Nagel GmbH, Düren, Germany) in accordance to the manufacturer's instructions. cDNA was reverse transcribed from mRNA using the RevertAid™ First Strand cDNA Synthesis Kit (VWR International, Darmstadt, Germany) are referred to instruction manual. For PCR (total volume 25 µL per reaction) 1.25 U Taq Polymerase, 1ˆreaction buffer, 2 mM MgCl 2 , 2 µM dNTPs (all reagents from VWR International, Darmstadt, Germany) and 100 µM primers (Life Technologies, Darmstadt, Germany) were used. The cycling conditions comprised of an initial denaturation step (5 min at 95˝C), 35 cycles of amplification (30 s 94˝C; 30 s appropriate annealing temperature; 30 s 72˝C) and final elongation (10 min 72˝C). PCR products were separated on a 1% TAE agarose gel. PCR bands were visualized by GelRed™ staining (VWR International GmbH, Darmstadt, Germany) and the GelDoc TM EZ Imager System (Bio-Rad, Munich, Germany). Primer pairs used in this study are summarized in Table 1.

Extraction of Nuclear Proteins
Harvested cells (2ˆ10 6 ) were resuspended in culture media and were treated with 100 ng/mL LPS (Sigma-Aldrich, Taufkirchen, Germany) or 100 ng/mL CpG-ODN (Invivogen, San Diego, CA, USA), respectively, for 2 h at 37˝C and 5% CO 2 in a humidified atmosphere. Non-stimulated cells served as a control. The NE-PER Nuclear and Cytoplasmic Extraction Reagent Kit (Thermo Fischer Scientific, Bonn, Germany) was used for purification of NF-κB and IRF-7 from nuclear extracts in accordance to the manufacturer's instructions. Nuclear extracts were boiled in 3ˆLaemmli Sample Buffer (5 min, 95˝C) [61] and stored at´80˝C prior to SDS-PAGE and Western Blot analysis.

Western Blot Analysis
Cells (2ˆ10 5 ) were seeded in 6-well plates and were stimulated with either 100 ng/mL LPS (Sigma-Aldrich, Taufkirchen, Germany) or 100 ng/mL CpG-ODN (Invivogen, San Diego, CA, USA), respectively, for 2, 4, 6, 12, 24, and 48 h at 37˝C and 5% CO 2 in a humidified atmosphere. Subsequently, cells were harvested and lysed in 3ˆLaemmli Sample Buffer (5 min, 95˝C) [61]. Samples were separated on a 10% or 12%, respectively, SDS-PAGE and transferred to an Immobilon PVDF nitrocellulose membrane (EMD Millipore, Darmstadt, Germany) under semi-dry conditions. Membranes were blocked with 10% (w/v) not-fat milk powder or 5% BSA in TBS-T. Bands were visualized using the Pierce ECL Western Blotting Substrate (Thermo Fischer Scientific, Bonn, Germany) in accordance to the manufacturer's instruction and the Aequoria Macroscopic Imaging System (Hamamatsu Photonics Germany, Herrsching am Ammersee, Germany). Antibodies used for Western blot analysis are listed in Table 2.

3D-Collagen Matrix Migration Assay
The migratory behavior of the cells in response to LPS and CpG-ODN was performed by using the 3D-collagen matrix migration assay as described recently [24][25][26][62][63][64]. Briefly, cells (4ˆ10 4 to 6ˆ10 4 ) were mixed with liquid collagen solution (Purecol; Nutacon BV, Leimuiden, The Netherlands) was mixed with 10ˆMEM (Sigma-Aldrich, Taufkirchen, Germany), and 7.5% sodium bicarbonate solution (Sigma-Aldrich, Taufkirchen, Germany). In dependence of the experimental setting different concentrations of either LPS or CpG-ODN were added. The cellˆcollagen mixture was filled in self-constructed cell migration chambers and the collagen was allowed to polymerize. Subsequently, cell migration chambers were placed under a microscope in a compartment tempered to 37˝C.
The migration of the cells was recorded by time-lapse video microscopy overnight (at least 15 h). To analyze the migration of the cells, in each experiment 30 cells were randomly selected and the paths of the cells were tracked in 15 min real-time intervals using a manual cell tracking software application.

Statistical Analysis
Statistical significance of the cell migration data was calculated using the Mann-Whitney U-test.

Conclusions
In summary, here we have shown that M13MDA435-1 and -3 hybrid cells derived from MDA-MB-435-Hyg human breast cancer cells and M13SV1-EGFP-Neo breast epithelial cells exhibit a differential TLR4 and TLR9 signaling, which is in view with the cell fusion hypothesis that hybrid cells could exhibit novel properties. Cell fusion is a random and unpredictable process that is chiefly attributed to HST representing the merging of the parental nuclei-a process which is characterized by loss of whole chromosomes, unequal and random segregation of chromosomes to the daughter cells and even chromothripsis [28,29,[65][66][67]. The random distribution of parental chromosomes to daughter cells was recently visualized by Zhou and colleagues, which further showed that hybrid cells exhibited a greater extent of DNA double strand breaks [66]. Thus, cell fusion and HST might be associated with chromothripsis specifying the shattering and random rearrangement of one or more chromosomes [67]. Chromothripsis is a common phenomenon in many (if not all) cancers and has been associated with the loss of tumor suppressors, dysregulation of genes with known cancer links and oncogene amplification [67]. Moreover, chromosome segments that fail to get reincorporated can circularize to become double minutes, which are frequently amplified [67]. Karyotypes of hybrid cells derived from the fusion of hamster cells and human tumor cells revealed the existence of a series of small unidentifiable chromosomes/chromosomal structures [68] that may have originated from chromothripsis.
Consequently, cell fusion is a potent mechanism that gives rise to unique hybrid cells that, concomitant with their progenies, will increase the heterogeneity of the tumor mass. How these cells ultimately behave in the tumor microenvironment and react to surrounding stimuli strongly depends on the cells receptor repertoire and kinetics of signal transduction cascade controlled by the cells' genetic and epigenetic background concomitant.