(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Ras/ERK and PI3K/AKT signaling differentially regulate oncogenic ERG mediated transcription in prostate cells

['Brady G. Strittmatter', 'Department Of Molecular', 'Cellular Biochemistry', 'Indiana University', 'Bloomington', 'Indiana', 'United States Of America', 'Travis J. Jerde', 'Department Of Pharmacology', 'Toxicology']

Date: 2021-10

The TMPRSS2/ERG gene rearrangement occurs in 50% of prostate tumors and results in expression of the transcription factor ERG, which is normally silent in prostate cells. ERG expression promotes prostate tumor formation and luminal epithelial cell fates when combined with PI3K/AKT pathway activation, however the mechanism of synergy is not known. In contrast to luminal fates, expression of ERG alone in immortalized normal prostate epithelial cells promotes cell migration and epithelial to mesenchymal transition (EMT). Migration requires ERG serine 96 phosphorylation via endogenous Ras/ERK signaling. We found that a phosphomimetic mutant, S96E ERG, drove tumor formation and clonogenic survival without activated AKT. S96 was only phosphorylated on nuclear ERG, and differential recruitment of ERK to a subset of ERG-bound chromatin associated with ERG-activated, but not ERG-repressed genes. S96E did not alter ERG genomic binding, but caused a loss of ERG-mediated repression, EZH2 binding and H3K27 methylation. In contrast, AKT activation altered the ERG cistrome and promoted expression of luminal cell fate genes. These data suggest that, depending on AKT status, ERG can promote either luminal or EMT transcription programs, but ERG can promote tumorigenesis independent of these cell fates and tumorigenesis requires only the transcriptional activation function.

ERG is the most common oncogene in prostate cancer. The ERG protein can bind DNA and can activate some genes and repress others. Previous studies indicated that ERG cannot promote cancer by itself, but that ERG works together with mutations that activate the protein AKT. In this study we found that activation of AKT changes the genes that ERG regulates, leading to luminal epithelial differentiation, which is a hallmark of most prostate tumors. However, we also found that a mutant version of ERG that can activate, but cannot repress genes, can drive prostate tumorigenesis without activation of AKT, but this mutant ERG cannot promote luminal differentiation. Our findings suggest that ERG mediated tumorigenesis only requires ERG’s activation function and can occur independent of luminal cell differentiation.

Funding: This work was supported by the National Cancer Institute ( www.cancer.gov ) of the National Institutes of Health under Award Number R01CA204121 (P.C.H.), and with support from the National Institute of Health (T32 GM131994) the College of Arts and Sciences of Indiana University – Bloomington and the University Graduate School (B.G.S.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2021 Strittmatter et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Here, we demonstrate that a phospho-mimetic mutation of ERG at S96 (S96E) can promote clonogenic growth and tumorigenesis in a mouse xenograft model independent of AKT activation. Further, we show that only a portion of cellular ERG is phosphorylated in prostate cells and that this phosphorylation takes places in the nucleus at sites of ERK recruitment. ERK binding to chromatin correlated with ETS/AP-1 motifs and ERG-activated genes. ERG S96E counteracted ERG-mediated repression at sites not bound by ERK by reducing EZH2 binding and depleting H3K27me3. Activation of AKT redistributed ERG to new binding sites and allowed ERG to promote a luminal gene expression program. Consistent with these findings, tumors driven by ERG/AKT expressed AR, while tumors caused by ERG S96E did not. Tumors with both ERG S96E and activated AKT did not express AR, indicating that ERG-mediated repression is required for luminal cell fates, but not for tumorigenesis.

Both the Ras/ERK and PI3K/AKT signaling pathways are critical regulators of ERG function in prostate [ 29 ]. ERK2 can phosphorylate ERG at S96 and S215 in prostate and endothelial cells as well as at S276 in leukemic cells [ 21 , 30 , 31 ]. ERG gene rearrangements can result in expression of either full-length ERG, or N-terminal deletions of 32 or 92 amino acids (numbering is Uniprot isoform 1, isoform 2 adds seven amino acids) and S96 is phosphorylated in all three of these proteins [ 21 ]. Phosphorylation of ERG S215 by ERK results in a conformational change in the ERG protein that allows subsequent ERK phosphorylation at ERG S96. S96 phosphorylation allows for transcriptional activation by disrupting the interaction between ERG and the Polycomb Repressive Complex 2 (PRC2) [ 21 ]. The role of the PI3K/AKT pathway in ERG function is less well understood mechanistically. AKT has not been demonstrated to directly phosphorylate ERG and the canonical kinase complexes downstream of AKT, mTORC1, and mTORC2 are not involved in ERG transcriptional activity [ 32 ]. GSK3B, a downstream target of AKT, has been shown to phosphorylate ERG and alter tumorigenic capabilities, but only when DNA damage is present and WEE1 signaling is active [ 33 ].

ERG can either activate or repress target genes. Studies that have measured expression changes of direct ERG target genes upon addition or depletion of ERG in prostate cells have identified approximately equal numbers of activated and repressed genes [ 10 , 20 , 21 ]. A recent study [ 8 ] indicates that direct ERG repression of TP63, a master regulator of basal cell fate, is important for ERG to promote luminal fates. ERG has been demonstrated to downregulate VCL, an epithelial related cytoskeleton protein [ 22 ] as well as regulators of the PI3K/AKT pathway PTEN and IRS2 [ 23 , 24 ]. ERG interacts with various co-repressors including EZH2, HDAC1, HDAC2, and KDM4A [ 22 , 25 ]. Transcriptional activation by ERG is thought to be important for oncogenic function. ERG can activate CTGF, ENC1, ETS1, PLAU, and VIM which are all thought to be involved promoting oncogenic phenotypes [ 15 , 26 , 27 ] and has been shown to interact with co-activators P300, BRD4, and the BAF Complex [ 9 , 28 ]. We have also identified an interaction with the co-activator EWS that is necessary for ERG to promote tumorigenesis in the RWPE1 xenograft system [ 19 ].

ERG expression can influence prostate cell fate decisions. The epithelium of the normal prostate gland is composed primarily of luminal secretory cells separated from the stroma by a layer of supportive basal epithelial cells. Most prostate tumors are characterized by an expansion of luminal epithelial cells and a relative absence of basal epithelial cells [ 8 ]. ERG expression in prostate tumor models can promote luminal epithelial cell fates, indicating that this could be part of the oncogenic function of ERG [ 8 , 9 , 10 ]. The androgen receptor (AR) is a key transcription factor that promotes luminal cell fates and AR activity is necessary for the growth of all early stage prostate tumors. Several studies indicate that ERG can cooperate with AR in promoting luminal cell fate [ 11 , 12 , 13 ]. In contrast to these findings, we and others have shown that expression of ERG in the immortalized-normal prostate cell line RWPE1 does not promote luminal epithelial differentiation, but rather promotes the seemingly opposite phenotypes of migration, invasion, and epithelial to mesenchymal transition (EMT) [ 14 , 15 , 16 , 17 , 18 ]. Despite this difference regarding cell fate, we have found that ERG expression and PI3K/AKT activation cooperate to promote RWPE1 xenograft tumor growth [ 19 ]. Therefore, with regard to tumor formation, ERG expression has a similar role in RWPE1 xenografts and transgenic mouse models.

Prostate cancer is the most common, and second deadliest, malignancy amongst American men [ 1 ]. In ~50% of prostate cancers a chromosomal rearrangement results in the fusion of the androgen-regulated promoter of TMPRSS2 to the open reading frame of ERG and results in aberrant expression of either full-length, or N-terminally truncated ERG protein in prostate epithelium [ 2 ]. ERG is an ETS family transcription factor that is not expressed in normal prostate epithelial cells [ 3 ]. In mouse models, ERG expression cooperates with mutations that activate the PI3K/AKT pathway, such as PTEN deletion, to drive prostate adenocarcinoma [ 4 , 5 , 6 ]. Consistent with this two-hit model, PTEN deletion positively correlates with TMRPSS2/ERG rearrangement in patient tumors [ 7 ]. However, the mechanism of cooperation between ERG and the PI3K/AKT pathway is not known. It is also unclear which transcriptional targets of ERG are necessary to promote tumor formation. ERG can both activate and repress the expression of direct target genes, but the relative role of these activities in tumor formation is also in question.

Results

The Ras/ERK and PI3K/AKT signaling pathways differentially regulate ERG mediated phenotypes To determine the roles of the Ras/ERK and PI3K/AKT pathways in ERG-mediated phenotypes, FLAG-tagged WT ERG and ERG S96 phospho-mutants were stably expressed with or without constitutively activate myristoylated-AKT (mAKT) in RWPE1 normal prostate epithelial cells. ERG Isoform 1 was used (Uniprot) in this study. Isoform 1 differs from isoform 2 by the N-terminal amino acids (MAST in isoform 1 and MIQTVPDPAAH in isoform 2). ERG S96 (S103 in isoform 2) is phosphorylated by ERK1/2 and S96E functions as a phosphomimetic, with S96A a phosphonull [21]. Immunoblotting of whole cell lysates confirmed ERG expression and showed an increased level of S473 phosphorylated AKT in the lines expressing mAKT (Fig 1A). Similar to our previous findings [21] ERG and ERG-S96E significantly increased RWPE1 cell migration while ERG-S96A had no effect. AKT activation alone increased RWPE1 migration, but the expression of mAKT in combination with ERG or ERG S96E did not further increase RWPE1 migration (Figs 1B and S1A). The same lines were tested in a clonogenic survival assay (Figs 1C and S1B). ERG promoted a moderate increase in clonogenic survival, and this function was not lost in the ERG S96A mutant. Expression of mAKT alone also moderately increased colony formation, however, the combination of ERG and mAKT significantly and synergistically increased colony formation. Further, expression of ERG-S96E alone promoted colony formation to a similar level as the ERG/mAKT combination, but neither ERG S96E nor ERG S96A synergized with mAKT. Together, these data suggest that S96 phosphorylation, but not AKT activation, is required for ERG to promote cell migration, and either S96 phosphorylation or AKT activation allows ERG to promote clonogenic growth. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Ras/ERK and PI3K/AKT signaling differentially regulate ERG mediated phenotypes. (A) Immunoblot of RWPE1 cells expressing FLAG-tagged ERG or mutant, and myristoylated AKT as indicated. (B) Transwell cell migration of RWPE1 cells expressing indicated constructs. Shown are the mean and SEM normalized to vector of 4 biological replicates, each the mean of 2 technical replicates. (C) Clonogenic growth of RWPE1 cells sparsely plated and allowed to form colonies over 10 days followed by fixing, staining, and quantification via Sygene GeneTools colony counting software. (D) Mouse xenograft experiments measured by calipers of RWPE1 cell lines grown (n = 6) in the flanks of immunocompromised mice for 8 weeks with patient derived CAFs or (E) RWPE1 cells grown (n = 4 or 6) with DLP-derived stromal cells from INK4A null mice. https://doi.org/10.1371/journal.pgen.1009708.g001 The cell lines were then tested in two different mouse xenograft models to assay tumor formation ability. First, RWPE1 cell lines, primary patient derived cancer associated fibroblasts (CAFs), and matrigel were co-injected into the flanks of nude mice and allowed to form tumors for eight weeks (Fig 1D). Consistent with previous findings using this system [19], cells expressing ERG in combination with mAKT formed significantly larger tumors than those expressing ERG or mAKT alone. Similar to the clonogenic growth assay however, the S96E mutation allowed ERG to promote tumor formation in the absence of mAKT. Overall, the sizes of the tumors grown with CAFs were much smaller than in previous work [19]. We attributed this change to differences in the primary CAF cells. To confirm these results in a more defined system, the same RWPE1 cell lines, and matrigel were co-injected with DLP derived stromal cells from INK4a null mice as alternative support cells [34]. This system revealed a similar trend in relative tumor growth, but with larger tumors (Fig 1E). ERG synergized with mAKT to promote tumor formation. Expression of ERG-S96E alone formed the largest tumors and grew statistically significant tumors by week four (S1C Fig). Similar to the clonogenic growth assay, the addition of mAKT actually decreased tumor size when combined with ERG S96E. Overall, our data indicates that the ERG-S96E phosphomimetic mutation abrogates the need for AKT activation in ERG-mediated tumor formation.

Phosphorylated ERG is in the nucleus To further investigate how ERK phosphorylation regulates ERG-mediated transcription we conducted an immunoprecipitation (IP) coupled with mass spectrometry of total cellular ERG from VCaP prostate cancer cells which express ERG due to a TMPRSS2/ERG gene rearrangement. We identified seven high confidence phosphorylation sites on the ERG protein (Fig 2A) including residues previously identified [21,30] to be phosphorylated by the Ras/ERK pathway including S96, S215, and S276 (S283 in isoform 2). Acetylation of ERG at K89 and K92 was also detected (S2A Fig). These acetylation sites have been demonstrated to play a key role regulating ERG function in Acute Myeloid Leukemia [28]. Eight unique phospho-peptides and twelve unique apo-peptides were identified in the region spanning S96 (Fig 2A) indicating that a portion (~40%) of total cellular ERG is phosphorylated at S96 in VCaP cells. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Ras/ERK signaling occurs at ETS/AP1 genomic locations to regulate cell migration. (A) MS/MS analysis of ERG immunoprecipitated from VCaP and RWPE-ERG cells. Data tables show phospho-residue number of ERG NCBI Isoform 1 and compares number of apo/phospho peptides with coverage of S96. (B) Cytoplasmic/Nuclear fractionation of RWPE-ERG immunoblotted for indicated proteins. (C) Heatmap of ERK ChIP-Seq data centered on ERG bound regions [15] depicting motifs over-represented in the top 500 ERK enriched and bottom 500 ERK depleted genes and GSEA analysis of ERK enrichment across ERG bound regions run for motif analysis. (D) GSEA analysis of ERK ChIP-Seq data enrichment ranked on ERG bound regions. (E) ERK ChIP-Seq data centered on consensus ETS/AP1 motifs previously found to be ERG bound in RWPE1. (F) Log2(FC) of nearest genes bound by ERG in RWPE1 cells determined by ChIP-Seq in cells expressing ERG or ERG S96E. (G) Log2(FC) of nearest gene to ERG alone or ERG/ERK co-bound regions. https://doi.org/10.1371/journal.pgen.1009708.g002 IP-mass spectrometry of RWPE1 cells expressing ERG (RWPE-ERG) identified three high confidence phosphorylation sites (Figs 2A and S2B). Two of these sites, S96 and S215 were also identified in VCaP. Five apo-peptides and three phospho-peptides (Fig 2A) spanned the S96 region indicating that a similar portion of cellular ERG was phosphorylated at S96 in VCaP and RWPE-ERG cells. In order to identify the cellular location of phosphorylated ERG, nuclear/cytoplasmic fractionation was performed in RWPE1-ERG cells. ERG was found in both the cytoplasmic and nuclear fraction, however only nuclear ERG was phosphorylated at S215 (Fig 2B).

ERK binds chromatin at sites of ERG activation function We previously demonstrated [21] that S96E and S96A do not alter nuclear/cytoplasmic ratios of ERG, indicating that phosphorylation does not alter nuclear trafficking. Therefore, to explain the bias of phosphorylated ERG for the nucleus, we postulated that a non-diffusible fraction of ERG might be phosphorylated by ERK in the nucleus. The ETS transcription factor ELK1 is phosphorylated when bound to a subset of target genes via differential recruitment of ERK to chromatin [35]. To determine if ERK similarly associates with a subset of chromatin-bound ERG, we conducted chromatin immunoprecipitation-sequencing (ChIP-Seq) of ERK2 in RWPE-ERG cells. We found ERK2 was present at a portion of ERG binding sites (Fig 2C). To determine if specific sequence motifs were associated with ERK2 occupancy, the top 500 ERG-bound regions enriched by ERK2 ChIP and the bottom 500 ERG-bound regions with no ERK2 ChIP signal were analyzed by RSAT motif using each data set as a control for the other to allow identification of differential motifs. A non-canonical ETS motif and an AP1 motif were significantly enriched at ERG/ERK2 co-bound regions, while ERG-alone regions were associated with CpG and SMAD motifs (Fig 2C). Canonical ETS motifs were enriched in both datasets, and therefore cancelled out. Gene Set Enrichment Analysis (GSEA) was also used to identify enriched motifs by ranking ERK2 ChIP signal across ERG binding sites (Fig 2C). This analysis identified a sequence (MGGAAGTG) consistent with the ETS/AP-1 half site spacing (GGAAGTGA) we have previously reported is enriched in the regulatory regions of cell migration genes [36]. Consistent with a cell migration role for ERK/ERG target genes, GSEA of ERK ChIP signal across ERG-bound regions was enriched for the gene ontologies Ameboidal Type Cell Migration and Tumor Invasiveness (Fig 2D). To further investigate the recruitment of ERK to ETS/AP-1 motifs, we subjected wild-type RWPE1 to ERK2 ChIP-seq (Fig 2E). In the absence of ERG, ERK2 was enriched at the same ETS/AP-1 motifs that ERG binds, indicating that some factor other than ERG does recruit ERK2 to these sites. However, the ERK2 enrichment at ETS/AP-1 motifs was higher in RWPE-ERG cells (Fig 2E) and this enrichment was not observed at ERK binding sites that lack ERG (S2D Fig), indicating that ERG increases ERK2 recruitment. This is consistent with our previous finding that ERG directly binds ERK with relatively high affinity compared to other ETS factors [32]. To test how ERK binding and phosphorylation of ERG alters ERG mediated transcription, we analyzed previously published RNA-Seq data from RWPE-vector, RWPE-ERG and RWPE-ERG S96E cell lines (GSE86232) [21]. Previous studies indicate that ERG can activate some genes and repress others [10,20,22]. Consistent with this, we found that genes near ERG binding sites were both activated and repressed resulting in a mean expression change of zero when ERG is expressed in RWPE1 cells (Fig 2F). In contrast, expression of these genes in cells expressing ERG-S96E was significantly higher than in RWPE-vector cells (Fig 2F), suggesting S96 phosphorylation favors gene activation. We then tested the correlation with ERK1 binding. Genes near ERK-ERG co-bound sites were activated, while genes near ERG bound sites lacking ERK were repressed (Fig 2G), suggesting that ERK binding allows ERG to switch from a repressor to an activator.

Ras/ERK signaling opposes PRC2 repression at ERG binding sites Phosphorylation of ERG at S96 by ERK disrupts ERG’s interaction with polycomb repressive complex 2 (PRC2) including the enzymatic subunit EZH2 [21]. To further investigate the interplay between ERG, ERK, and EZH2 on chromatin we analyzed previously published ChIP-Seq data of ERG and EZH2 in RWPE1, RWPE1-ERG, and RWPE1-ERG S96E cells (GSE86232) and conducted ChIP-Seq of H3K27me3, the repressive mark that EZH2 deposits on chromatin. We examined occupancy at ERG-bound regions in promoters or enhancers (Fig 3A). Expression of ERG increased ERK binding, decreased EZH2 binding, and decreased H3K27me3. Loss of EZH2 binding occurred mostly at promoters, but loss of H3K27me3 was observed at both promoters and enhancers. Expression of ERG-S96E resulted in an even further depletion of EZH2 and H3K27me3 at promoters and enhancers compared to ERG alone (Fig 3A). Further, GSEA of ranked ERK binding across ERG-bound regions showed negative correlation with high density CpG promoters marked with H3K27me3 and positive correlation with genes that are upregulated upon knockdown of EZH2 (Fig 3B). These data suggest opposing functions of ERK and EZH2 at ERG binding sites. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Ras/ERK signaling opposes PRC2 repression at ERG binding sites. (A) Average plots of ERK, EZH2, and H3K27me3 ChIP-Seq conducted in RWPE1, RWPE1-ERG, or RWPE1-ERG S96E cell lines centered on ERG bound regions in RWPE1 cells [15] (B) GSEA analysis of ERK ChIP signal ranked across ERG bound regions in RWPE1 cells. (C) Volcano plot showing differentially regulated genes determined by RNA-Seq between RWPE1 cell lines expressing ERG S96E and ERG (red—significantly activated / blue significantly repressed). (D) Log2(Fold change vs. Vector) of genes nearest ERG-alone or ERG/ERK co-bound regions in RWPE1 cells expressing ERG or ERG S96E. (E) GSEA analysis of differentially regulated genes called by RNA-Seq comparing RWPE1 ERG S96E to RWPE1 ERG. https://doi.org/10.1371/journal.pgen.1009708.g003 ERG S96E can promote tumor formation in the absence of activated AKT signaling, whereas ERG cannot (Fig 1E). To determine how the S96E point mutation alters gene expression, we generated a volcano plot comparing differentially regulated genes between ERG and ERG S96E and found that, consistent with Fig 2F, the majority (77.3%) of the differentially regulated genes were activated in ERG S96E compared to ERG alone (Figs 3C and S3A). This ability of ERG S96E could be due to increased activation of activated targets, loss of repression of repressed targets, or both. Therefore, we compared differences in the ERG-activated targets co-bound by ERK2 and the ERG-repressed targets bound by ERG alone (Fig 3D). ERG S96E did not further activate ERG/ERK co-bound target genes, consistent with these genes already being associated with phosphorylated ERG. Instead the major effect of ERG S96E was loss of repression of genes bound by ERG alone. Further, GSEA analysis of the differentially regulated genes between ERG S96E and ERG found that ERG S96E was negatively correlated with targets of EED and SUZ12, two essential components of the PRC2 (Fig 3E). Together these data suggest that the major role of the S96E mutation is the loss of ERG-mediated repression via PRC2 recruitment.

Constitutive PI3K/AKT signaling co-operates with ERG to promote luminal fate associated gene programs To determine how ERG, mAKT, and the combination effect gene expression we conducted RNA-Sequencing of these cell lines in triplicate. Expression of mAKT alone resulted in a relatively small number (1,245) of significantly differentially regulated genes (S5 Fig). The addition of ERG resulted in a more robust change in differentially regulated genes (3,367) compared to RWPE1-vector. However, the largest number of differentially regulated genes (3684) was observed when RWPE-ERG/mAKT was compared to RWPE-ERG. This larger role for mAKT in the presence of ERG suggests synergy. Unsupervised clustering of RNA-Sequencing data resulted in two distinct clusters one containing RWPE and RWPE-ERG and the other RWPE-mAKT and RWPE-ERG/mAKT (Fig 5A). Of the genes which were differentially expressed upon addition of mAKT to RWPE-ERG cells, 1098 (30%) had neighboring ERG bound regions in the RWPE-ERG/mAKT ChIP-seq indicating that they are potential direct targets (Fig 5B). This was a statistically significant enrichment (p < 0.001). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Constitutive PI3K/AKT signaling co-operates with ERG to promote luminal fate associated gene programs. (A) Unsupervised clustering of RNA-Sequencing data from RWPE1 cells expressing Vector, ERG, myr-AKT, or ERG+mAKT (green-activated, red-repressed) (B) Overlap of genes nearest ERG bound regions in RWPE-ERG/mAKT and significantly differentially regulated genes in RWPE-ERG/mAKT compared to RWPE/ERG (C) Top GSEA enrichments when comparing gene expression in the RWPE1 cells expressing the indicated constructs (below). https://doi.org/10.1371/journal.pgen.1009708.g005 GSEA indicated that the addition of mAKT alone to RWPE1 cells resulted in an active AKT signature, inhibition of EMT, and epidermal differentiation (Fig 5C). In contrast, ERG expression alone in RWPE1 cells resulted in an active Ras/ERK signature and was correlated with positive regulation of chemotaxis (Fig 5C), consistent with the ability of ERG to promote cell migration (Fig 1B) and previous findings that ERG expression in RWPE1 can mimic Ras/ERK signaling [15]. Strikingly, the combination of ERG and mAKT expression in RWPE1 cells promoted a luminal epithelial differentiation program compared to expression of either ERG or mAKT alone. Comparing gene expression from RWPE-ERG/mAKT to RWPE-ERG revealed that the addition of mAKT promoted epithelial and luminal differentiation and inhibited EMT (Fig 5C). Similarly, when comparing RWPE1-mAKT cells to RWPE-ERG/mAKT, the addition of ERG activated programs associated with luminal differentiation and immune response (Fig 5C). Taken together, these data indicate that expression of ERG alone promotes an active Ras/ERK signature responsible for cell migration and EMT, but constitutive activation of the PI3K/AKT pathway reprograms the ERG cistrome and transcriptome to promote luminal epithelial differentiation.

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009708

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.