Steering tumor progression through the transcriptional response to growth factors and stroma

FEBS Letters xxx (2014) xxx–xxx

journal homepage: www.FEBSLetters.org

Review

Steering tumor progression through the transcriptional response to growth factors and stroma Morris E. Feldman, Yosef Yarden ⇑ Department of Biological Regulation, Weizmann Institute of Science, Rehovot 76100, Israel

a r t i c l e

i n f o

Article history: Received 24 April 2014 Revised 19 May 2014 Accepted 19 May 2014 Available online xxxx Edited by Wilhelm Just Keywords: Growth factors EGFR Transcription Cancer immediate early genes Receptor tyrosine kinase

a b s t r a c t Tumor progression can be understood as a collaborative effort of mutations and growth factors, which propels cell proliferation and matrix invasion, and also enables evasion of drug-induced apoptosis. Concentrating on EGFR, we discuss downstream signaling and the initiation of transcriptional events in response to growth factors. Specifically, we portray a wave-like program, which initiates by rapid disappearance of two-dozen microRNAs, followed by an abrupt rise of immediate early genes (IEGs), relatively short transcripts encoding transcriptional regulators. Concurrent with the fall of IEGs, some 30–60 min after stimulation, a larger group, the delayed early genes, is upregulated and its own fall overlaps the rise of the final wave of late response genes. This late wave persists and determines long-term phenotype acquisition, such as invasiveness. Key regulatory steps in the orderly response to growth factors provide a trove of potential oncogenes and tumor suppressors. Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Introduction Beyond deep understanding of mechanisms underlying tumor initiation and progression, contemporary cancer research ultimately strives to develop more efficacious and selective anti-tumor drugs. In the last decade, this avenue of intense research has been inspired by the ‘oncogene addiction’ theory [1]. Accordingly, cancers that contain multiple genetic and chromosomal abnormalities are dependent on or ‘addicted’ to one or a few genes for maintenance of the malignant phenotype. Thus, reversal of only one or a few of these abnormalities can inhibit cancer cell growth and translate to improved survival rates. An example is provided by the multiple mutant forms of the epidermal growth factor receptor (EGFR) in lung tumors. Low molecular weight inhibitors of the EGFR’s kinase effectively inhibit lung tumors when they express one of the mutant, constitutively active forms of EGFR [2]. Another ‘addiction’ might be exemplified by breast cancers that overexpress HER2, a kin of EGFR, which are effectively controlled by a monoclonal anti-HER2 antibody [3]. While cancer genome sequencing initiatives continue to identify more mutant forms and candidates for targeted therapies, the remarkable multiplicity of mutations in solid tumors [4], along with inherent adaptive mechanisms that lead to patient resistance [5], set formidable limits to the ‘oncogene addiction’ strategy. ⇑ Corresponding author. Fax: +972 8 934 2488. E-mail address: [email protected] (Y. Yarden).

It is worthwhile noting that several, rather effective cancer drugs target non-mutated cellular components. They include the estrogen receptor in breast cancer, the microtubule network in a variety of tumors, and the vascular endothelial growth factor (VEGF) in colorectal and other cancers. Moreover, anti-EGFR antibodies effectively inhibit colorectal tumors despite the fact that EGFR in these malignancies is not usually amplified or mutated. Interestingly, the presence of amphiregulin and epiregulin, two ligands of EGFR, in colorectal tumors predicts response to anti-EGFR antibodies [6]. This suggests that the therapeutic antibody achieves impact by blocking autocrine or paracrine, stroma-mediated loops involving EGFR and one of its seven ligands. Consistent with this interpretation, growth factors play essential roles in most phases of tumor progression, including clonal fixation of oncogenic mutations, recruitment of blood and lymph vessels to the growing tumor and enhancing dissemination of tumor cells, leading to colonization of distant organs (metastasis) [7]. For example, an in vivo genetic screen for genes that enhance metastasis of breast cancer to lungs identified two ligands of EGFR [8]. Another, very important contribution of growth factors to tumor progression entails chemotherapy- and radiotherapyinduced autocrine loops that permit drug resistance. For example, by generating cisplatin-resistant MCF-7 cells, it was found that drug resistance associates with increased EGFR phosphorylation, high levels of AKT1 activity, inactivation of the p53 pathway and up-regulation of amphiregulin expression [9]. Moreover, knockdown of amphiregulin expression by specific short interfering

http://dx.doi.org/10.1016/j.febslet.2014.05.036 0014-5793/Ó 2014 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

2

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

RNA resulted in a nearly complete reversion of the resistant phenotype. In conclusion, in-depth understanding of the roles played by the stroma and growth factors in tumor progression might identify novel candidates for therapy and for combination treatments aimed at delaying the onset of resistance to drugs. Driven by this motivation, our mini-review highlights cytoplasmic and nuclear actions of growth factors, with an emphasis on transcriptional regulation by EGFR, a relatively well-understood growth factor receptor system. 2. Cytoplasmic signaling pathways activated by growth factors (see Fig. 1) Growth factors bind to and activate receptors at the plasma membrane [31]. In the case of EGFR, receptor activation upon ligand binding involves formation of homo- and hetero- dimers between EGFR and the other members of the ERBB family. Following ligand-induced dimerization, EGFR family members phosphorylate each other to induce full activation of their kinase domains. Other receptor tyrosine kinases (RTKs) have slightly different mechanisms of activation, but the end result is the activation of their intracellular tyrosine kinase domain. The main substrate of the EGFR kinase is itself; it phosphorylates a series of tyrosines in a long and probably unstructured carboxyl-terminal tail extending past the kinase domain. Other RTKs, however, such as the insulin and the insulin-like growth factor 1 (IGF1) receptors recruit and

extensively phosphorylate adapter molecules such as IRS, GAB and FRS. In either case the phosphotyrosines serve as a platform for the recruitment of downstream signaling complexes [11]. Major pathways downstream of EGFR and RTK activation in general include the mitogen-activated protein kinase (MAPK) and the phosphatidylinositol 3-kinase-protein kinase B (PI3K-AKT) routes [12]. EGFR carboxyl-terminal phosphotyrosines recruit upstream components of the MAPK/ERK pathway, such as Grb2 and Shc [13]. These adaptors bind to the phosphorylated EGFR and recruit SOS. SOS exchanges GDP for GTP in Ras to activate Ras, and active Ras then binds to and allosterically activates the Raf kinase. Raf is the first member of a cascade of three kinases in the MAPK pathway, with Raf activating MEK and MEK activating the terminal MAPK, Erk [14]. RTK activation of the PI3K pathway begins by recruitment of a class I PI3K to phosphotyrosines generated by the activated receptor [15]. PI3K phosphorylates the relatively abundant phosphoinositide, phosphatidylinositol 4,5-bisphosphate (also known as PI(4,5)P2 or simply PIP2) to generate PI(3,4,5)P3 (also referred to as PIP3). PIP3 recruits the Akt kinase to the membrane, where it is activated by phosphorylation by PDK1 and the mTOR complex 2, mTORC2 [16]. Akt itself phosphorylates several substrates that inhibit apoptosis, including BAD, the p53 regulator MDM2 and members of the FoxO family of transcription factors. The other major role of Akt is to activate mTORC1 through a series of biochemical steps beginning with the phosphorylation of TSC2 in the TSC1/TSC2 complex. Phosphorylation

Fig. 1. Receptor-mediated signaling and the initiation of transcriptional regulation. Receptor tyrosine kinases (RTKs), activated by growth factors (GFs) signal through several pathways to drive changes in the transcription of specific RNAs. Adapters (yellow), kinases (green) and other signaling proteins (gray) bind to phosphorylated tyrosines (encircled P letters) on the activated RTK, and then stimulate particular downstream pathways. For instance, upon binding to phosphotyrosine, phospholipase C, PLC, is activated to cleave the membrane lipid phosphatidylinositol 4,5-bisphosphate, PIP2, into inositol 1,4,5-trisphosphate, IP3, and diacylglycerol, DAG. Pathway activation culminates in changes in transcription factor (TF) activity, leading to changes in RNA expression. The extent of activation of each of these pathways will vary depending on the identity of the RTK. Also shown is the negative regulator CBL, which binds to phosphorylated tyrosines and ubiquitinates the RTK driving its internalization. Other RTK specific negative regulators such as MIG6 are not shown.

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

of TSC1/2 inhibits its GAP activity towards the GTPase Rheb. In its GTP bound form, Rheb is an activator of mTORC1. The net result is that mTORC1 is activated by Akt. mTORC1 regulates protein translation and cholesterol biosynthesis, thereby providing protein and lipid material to cells upon growth factor stimulation [17]. 3. Negative regulation of growth factor signaling Negative regulation of growth factor signaling begins immediately after receptor activation. In the case of EGFR the biochemical steps are especially well characterized and begin with the neddylation and subsequent ubiquitination of EGFR by c-Cbl [18]. Ubiquitinated EGFR is a substrate for internalization. Upon internalization, EGFR can be degraded by sorting to the multi-vesicular body, or deubiquitinated and then recycled back to the plasma membrane. Surprisingly the balance between degradation and recycling is influenced by EGFR ligands, with EGF primarily driving degradation and TGF-a primarily leading to recycling [19,20]. The weaker affinity and pH sensitive binding of TGF-a to EGFR allows it to unbind, as the early endocytic compartment is acidified. Unbinding then changes the conformation of EGFR, and routes it for recycling, presumably by encouraging its de-ubiquitination. In contrast, EGF binds very tightly to EGFR and is not released in the acidic environment of the endosome. Hence, EGFR bound to EGF is not de-ubiquitinated and therefore continues down the endocytic path to final degradation in the lysosome. The differential ability of TGF-a and EGF to drive EGFR degradation appears to account for the apparent paradox that the tighter binding ligand, EGF, is less powerful in inducing a transformed phenotype than the looser binding ligand, TGF-a. Another important translocation event in growth factor signaling is movement of proteins into or out of the nucleus. One of the major mechanisms to regulate transcription factors is by facilitating or allowing their access to the nucleus. For instance some MAPK substrates, including the ETS family member ERF, depart from the nucleus upon phosphorylation [21]. Similarly the FoxO transcription factors, which as mentioned above, are substrates for Akt, also leave the nucleus and therefore become inactive as transcription factors upon phosphorylation [22]. Even MAPK itself translocates into the nucleus upon activation, to gain better access to nuclear localized transcription factors [23]. In this way, MAPK enters the nucleus to phosphorylate and exclude some transcription factors from the nucleus, while at the same time activating other transcription factors, such as AP-1, which require phosphorylation to gain activity as transcription factors. 4. Transcriptional responses to growth factor signaling (see Fig. 2) A major output of activated MAPK signaling is changes in the activity of transcription factors. Most of the changes in transcription factor activity are probably due to phosphorylation, but other post-translational modifications such as ubiquitination, methylation and acetylation may also be involved. Some transcription factors, such Elk-1, a member of the ETS group of transcription factors, are directly phosphorylated by MAPK [24], whereas other transcription factors such as CREB are phosphorylated by kinases downstream of MAPK, in this case p90RSK [25]. In still other instances, parallel pathways downstream of receptor activation phosphorylate transcription factors, such as Stat [26]. The exact complement of transcription factors activated by post-translational modifications upon growth factor signaling is still an open research question, but the large changes in transcription that are observed upon growth factor activation, discussed below, suggest that multiple transcription factors are probably involved.

3

In the case of EGF, the transcriptional changes upon EGFR activation are particularly well understood. The expression of approximately 1000 genes is affected by EGF stimulation of epithelial cells, with activation more common than inhibition by approximately a 4:1 ratio [27]. Interestingly, however, transcriptional changes stimulated by steroid hormones such as the glucocorticoids, do not show such a strong bias towards activation over repression [28,29]. Because of the bias for gene activation by RTKs, much more research has focused on the mechanism and consequences of gene activation downstream of RTKs than on gene repression. About 5–10 min after the activation of an RTK, while signaling continues to evolve in the cytoplasm and at the plasma membrane, the first signals reach the nucleus and begin to affect transcription [30,31]. The first genes activated by a growth factor are typically seen to accumulate beginning approximately 20 min after the stimulus. These early genes, called immediate early genes or IEGs, usually rise rapidly in the levels of their mRNA and then shortly after rising they quickly fall, usually peaking within the first hour after the stimulus (see Fig. 2). Following the wave of IEGs, another set of genes, called the delayed early genes or DEGs, are activated and like the IEGs they also rise and fall, however the peak of DEG expression is usually seen approximately 2 h after the stimulus. Finally, approximately 2.5 h after stimulation, a third set of genes we will call late response genes, or LRGs, begins to rise. Unlike the IEGs and the DEGs, the LRGs do not drop in expression as long as the stimulus is maintained, but instead reach a steady state level of expression between 4 and 8 h after the stimulus [10]. The rise and fall of each wave intuitively suggests that each wave is inducing the next one, and until recently this has been the main mechanistic interpretation. Under this classical view the IEGs serve as transcription factors for the DEGs. The DEGs then serve to inactivate the IEGs while also signaling forward to the LRGs. In contrast to this conventional view of each wave activating the next, new evidence in a variety of systems suggests that some DEGs and LRGs are transcribed very soon after stimulation but remain unprocessed, in particular un-spliced, and unprepared for translation until much later times. In the following section we will discuss the waves of gene expression in more detail in the context of the conventional view that each wave of transcription triggers the next one. Then we will examine a new emerging model from several cellular systems suggesting that many LRGs are actually primary transcripts. We will close by exemplifying how a longterm phenotype, cell migration, is regulated by transcription. Note that some publications refer to LRGs as secondary response genes (or SRGs), but we will refer to the third wave of genes as LRGs, late response genes, since we wish to present evidence that many of these genes are in fact primary transcripts. 5. The fast rise and fall of IEGs Approximately 10–20 min after stimulation with a growth factor, IEGs begin to be transcribed. Because many IEGs, such as c-Fos, c-Jun and c-Myc, are proto-oncogenes, their transcription has been heavily studied [32]. Several unique features of IEGs facilitate their rapid increase in transcription. Beginning at the chromatin level, IEGs maintain an open permissive chromatin structure even in the absence of stimulation [31,33]. This open chromatin is reflected in the presence of CpG islands and strong TATA boxes at the 50 end of IEGs. The presence of strong CpG islands at the start of IEGs suggests that IEGs probably bind RNA Pol II constitutively but in a paused or stalled state and indeed this is often found to be the case experimentally [31]. At some IEGs, however, evidence suggests that Pol II is actively transcribing a low level of full-length transcripts in the absence of growth factor stimulation, but that

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

4

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

Fig. 2. The timeline of transcription regulation by growth factors. Transcriptional waves downstream of an RTK stimulus, like EGFR ligands, alternate between putative protooncogenes and candidate tumor suppressors [10]. Immediately after the stimulus a group of microRNAs (miRs) are down-regulated. These immediately down-regulated miRs (ID-miRs) primarily target the immediate-early genes (IEGs) and are enriched for miRs of tumor suppressor function [36]. Examples include mir-101 [74], miR-320 [75], miR370 [76] and miR-663 [77]. Notably, however, some ID-miRs seem to play dual roles as both tumor suppressors and oncogenes depending on the cancer context. Released from inhibition by the destruction of the ID-miRs, the IEGs, encoding oncogenic transcription factors like EGR1 [78], rise and fall within the first hour post stimulus [32]. Trailing the IEGs, the delayed-early genes (DEGs) encode many candidate tumor suppressors, whose role in the transcriptional program is to inhibit growth factor signaling [30]. For example, the group of DEGs includes DUSPs [79], KLF6 [80], ZFP36 [40] and MIG-6 [81,82]. Culminating the RTK stimulated gene expression program, late-response genes are up- and down-regulated, driving phenotypic changes due to growth factor signaling [51,83]. Up-regulated late response genes (UR-LRGs), are enriched for putative proto-oncogenes. For example, some highly mitogenic and motogenic RTK ligands and inflammatory cytokines like HB-EGF [84] and IL8 [85], respectively, rise soon after the stimulus, but unlike the IEGs, they persist. Other positive regulators of tumor progression, like matrix metalloproteinases, MMPs [86], and CTEN [68], are also included in the UR-LRGs group. Reciprocally, the stably down-regulated late response genes (DR-LRGs) might act as tumor suppressors in specific cellular contexts. This group includes cell cycle regulators like p27-Kip1 and p53, as well as EHF/ESE3, a putative suppressor of prostate cancer [87].

transcription of these transcripts is uncoupled from RNA splicing, and the transcripts are quickly degraded [34]. The binding of Pol II, promoted by the strong TATA boxes, prevents CpG DNA methylation. Since DNA methylation is mutagenic on the evolutionary time scale, CpGs are only allowed to accumulate at sites in the genome where methylation is prevented, such as the 50 end of genes that constitutively bind RNA Pol II [35]. Genes that constitutively bind RNA Pol II can be roughly divided into house keeping genes that are constitutively transcribed in most cells, and gene such as IEGs. Although IEGs are not constitutively transcribed, due to their strong TATA boxes and other motifs important for recruiting the transcription initiation machinery, they bind Pol II all the time in a poised state ready for rapid transcription. IEGs are poised for production but because of their power as oncogenes, there needs to be a mechanism to suppress spurious transcripts. In the case of signaling by EGFR and probably by other RTKs, a class of miRs suppresses the IEGs prior to growth factor stimulus [36]. Upon stimulation however, to facilitate the rapid rise of the IEGs, the concentration of the IEG-targeting miRs (called immediately down-regulated miRs, or ID-miRs) drops quickly, within the first few minutes after stimulation. Because the drop in the ID-miRs is so quick, their fall is likely due to degradation rather than reduced transcription, although the exact mechanism degrading the ID-miRs is still unclear.

Also helping to maintain low levels of IEG mRNA prior to RTK stimulation is the inherent instability of IEG mRNA [37]. Any IEG mRNA that is accidently produced will be quickly degraded. The instability of IEG mRNA also facilitates their rapid rise because it allows a high rate of mRNA production to be coupled to rapidly reaching a moderate steady state level of mRNA. Without rapid degradation, a high production rate of mRNA would instead lead to the slow accumulation of a massive amount of mRNA. To minimize the waste of resources associated with rapidly synthesizing and degrading IEG mRNA, most IEGs are very short, only a few kb long, and unlike most genes, many have few or no introns. Being short also minimizes the time required to transcribe the IEGs, allowing their mRNA to be accessible for translation with a minimal delay following the induction of their transcription. Within their rise, the IEGs encode their own demise; the transient rise and fall of many IEGs is nearly complete within the first hour after stimulation [30]. IEGs rapidly fall because their transcription is stopped and their mRNA is further destabilized [32]. Some IEGs negatively regulate their own promoters forming a simple network motif termed negative autoregulation, or NAR. In the absence of other influences however, NAR is predicted to lead to a rapid rise to a steady state [38]. Other factors beyond simple NAR must therefore be required to shape the IEG mRNA profile into a short pulse. One such factor is the mRNA binding protein ZFP36,

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

which is up-regulated in response to growth factors. ZFP36 binds to and destabilizes mRNAs containing AU-rich elements (AREs) such as the IEG mRNAs [39]. Recently, consistent with its role in destabilizing potentially oncogenic IEGs, ZFP36 has emerged as a potential tumor suppressor in a diverse set of cancers, including glioma [40–43]. 6. DEGs and LRGs: calming down and looking forward Many of the IEGs are transcription factors, and according to the classical model of growth factor-induced transcription, their activity is responsible for much of the induction of DEGs. Unlike the IEGs, which are mainly transcriptional activators, the DEGs contain a preponderance of negative regulators, of both transcription and signaling, such as MAPK phosphatases, whose induction after the IEGs further shapes the IEG profile into a tight pulse. Although DEGs share many characteristics with IEGs, they are not tuned for the extremely rapid rise and fall of the IEGs. Following RTK stimulation they rise more slowly than IEGs, and often reach a broad peak in expression around 2 h post stimulation, before slowly falling between 2 and 3 h after the stimulus. DEG mRNA is more stable than IEG, but less stable than other mRNAs [37]. DEGs mostly possess CpG islands at their 50 end and, like IEGs, are occupied by poised RNA polymerase II, but at generally lower levels than the IEGs. Unlike the IEGs however, DEGs show no enrichment for strong TATA boxes and have a similar length and quantity of introns to the genome at large [31]. In summary DEGs have genomic characteristics consistent with their ability to rise and fall after RTK stimulus without the level of instability and other properties required by the IEGs for extremely rapid rise and fall. Many DEGs negatively regulate the growth factor response [10]. Some DEGs directly inhibit the transcription factors that activate the IEGs. For instance, the Id subfamily of helix-loop-helix proteins inhibits TCF transcription factors such as Elk-1, which are phosphorylated by MAPK and activate IEGs. Another level of transcriptional repression is exemplified by the DEGs FOSL1 and JUNB, which inhibit the activity of the IEGs FOS and JUN, respectively, in AP-1 complexes. In these ways, DEG transcription factors repress the transcription of IEGs, as well as directly antagonize the IEGs themselves. The other major mode of negative regulation by the DEGs is to inhibit the signaling cascades, especially MAPK, downstream of RTK activation. Examples of DEG negative regulators of cell signaling include Mig6/ERFFI1, and members of the dual specificity phosphatase family, DUSPs. Mig6 is a direct inhibitor of EGFR itself [44,45], while DUSPs are specialized phosphatases that can dephosphorylate and thereby inactivate MAPKs on both threonine and tyrosine phosphorylation sites [46]. Consistent with their role in antagonizing IEG signaling downstream of RTKs, many DEGs, such as Mig-6 are tumor suppressors (see Fig. 2). After the IEGs and cell signaling have been repressed, several DEGs continue to persist for an extended period of time, suggesting that they impose a refractory period during which subsequent RTK activation and IEG transcription will be impossible. The presence of a refractory period after growth factor signaling indicates that oscillations might occur in the presence of a sustained source of growth factor or in the case of an autocrine loops. 7. Primary late response genes (PLRGs) The final wave of induced transcriptional changes, the lateresponse genes (LRGs), follows the DEG wave. LRGs rise slowly beginning around 2 h after the stimulus and unlike the IEGs and DEGs, LRGs continue to rise until they reach a steady state plateau level of mRNA. Also unlike IEGs and DEGs, LRG mRNAs are very

5

stable [37] with a stability similar to that of un-induced mRNAs. As we discuss in the end of this mini-review, the stable expression of LRGs is thought to be responsible for phenotypic changes such as cell migration [47]. To further stabilize the post-stimulus phenotype, some LRGs are RTK ligands that will continue to activate the RTK through an autocrine mechanism, even if the original stimulus wanes. Because their appearance is delayed by several hours following RTK stimulation, it was initially assumed that late response genes, LRGs, would all be secondary response genes, SRGs. The delayed appearance of LRGs was assumed to be due to the time required for a transcriptional network downstream of the initial RTK stimulus to produce new activating transcription factors to initiate transcription of the LRGs. Supporting this intuition, when the first LRGs were discovered they were all found to be SRGs, meaning that their transcription was strictly inhibited by CHX, a protein synthesis inhibitor [48] (see Box 1). Subsequent genome wide analysis of transcription in the presence of CHX, however, found that many LRGs are actually PRGs, meaning that they are relatively insensitive to CHX [31]. Understanding the delayed appearance of these genes presents a conundrum because it is unclear how a cell can delay the appearance of an mRNA without making the RNA dependent on preceding slow transcriptional steps. Stated another way, there is no clear way to explain the fact that transcription initiation of some LRGs apparently begins shortly after the RTK stimulation, at the same time as the IEGs, yet the mRNA for these genes does not appear for several hours. Box 1. Cycloheximide differentiates primary from secondary inducible transcripts. Cycloheximide (CHX), which blocks protein synthesis from mRNA, has been an essential tool for understanding stimulus-induced changes in gene expression. A characteristic feature of the IEGs, which was recognized early on [32,70], is that in the presence of CHX, rather than rising and quickly falling after the stimulus, IEG mRNAs accumulate to higher than normal levels and remain high without falling. Degradation of IEG mRNAs requires active protein synthesis because the act of translation shortens the polyA tail of IEG mRNAs, eventually leading to their degradation [71]. The translation dependent removal of IEG polyA requires AU-rich elements present in the IEGs [72], probably through a mechanism involving ZFP36 [73]. Since the degradation of IEGs requires active translation, in the presence of CHX, IEGs accumulate to high levels and persist, rather than showing a short pulse. In addition to uncovering the translation-dependent degradation of IEG mRNAs, CHX is also valuable because it can divide stimulus-responsive genes into primary responsive genes (PRGs) and secondary responsive genes (SRGs) [48]. IEGs are a subset of PRGs, whose mRNA accumulation is not blocked by CHX, because they do not require the synthesis of new proteins to induce their expression. Activation of pre-existing transcription factors downstream of RTK activation is all that is required for PRG activation. Unlike PRGs, SRGs require new protein synthesis for their induction following RTK stimulation. The conventional interpretation for their CHX sensitivity is that SRGs are activated by transcription factors, which do not exist in the cell prior to stimulation and these transcription factors must be translated and probably transcribed in order to activate the SRGs. Notably, other interpretations of CHX sensitivity are possible because the factor(s) that must be translated to allow SRG expression could be involved in another step of mRNA production than transcription initiation. As discussed in the main text, for some LRGs, the splicing step in RNA processing seems to be a regulated step and this step could theoretically be CHX sensitive if it relies on a newly translated factor.

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

6

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

New data from several systems suggest that two non-mutually exclusive mechanisms may be capable of explaining the late appearance of some PRGs. Danko et al. measured the transcription rate in estrogen-stimulated cells using global run-on sequencing, GRO-seq, and found that transcription rates vary by almost 10-fold between different genes, ranging from 0.4 to 4 kb/min [49]. In addition, the authors showed that the transcription rate at individual genes can vary by as much as 4-fold between different cells and different stimuli. Since the median length of a human gene is about 20 kb [50], for most genes any transcription rate faster than about 1 kb/min should be fast enough to allow for fairly rapid accumulation of mRNA following a stimulus, but for the minority of human genes longer than 100 kb, including JARID2 and NFKB1, a transcription rate of 1 kb/min or less would severely limit the speed of their transcriptional response. Interestingly, many genes, such as EGFR, harbor alternate promoters such that a cell can choose to produce a transcript with a greatly extended first intron from a promoter far upstream, or to produce a short transcript coding for a very similar protein product, differing perhaps only in the first exon. One attractive explanation for the common genomic feature of alternative promoters differing only by a long first intron is that the length RNA polymerase II must travel when transcribing a gene can be varied by the cell to change the timing of a transcriptional response. Although some PLRGs like vinculin are quite long at 125 kb, other PLRGs such as laminin gamma 2 (60 kb), Dickkopf-related protein 1 (3.4 kb), and cyclin D1 (13 kb) are not especially long, suggesting that gene length is not the primary factor leading to delayed expression of PLRGs. RNA splicing has recently emerged as a regulated step in the production of stimulus responsive genes that can help explain the origin of PLRGs. Hao and Baltimore [37] measured the kinetics of pre-mRNA and mRNA production in response to TNF-induced transcription and found that the transcription of the LRGs they examined began immediately following the stimulus, at the same time as IEG transcription in response to TNF. Although TNF-induced transcription of LRGs began a few minutes after the stimulus, unlike the IEGs, whose transcription began at the same time, spliced mRNA for the LRGs did not appear till much later. Like Hao and Baltimore, Zeisel et al. also examined stimulus-induced changes in mRNA and pre-mRNA following EGF stimulation [51], however instead of PCR, their primary assay was based on microarray measurements of intronic and exonic sequences, allowing genome wide estimation of mRNA and premRNA levels. Looking genome wide, Zeisel et al. found many instances where un-spliced pre-mRNA accumulated long before the appearance of the respective spliced mRNAs. The magnitude of time delay seen by Zeisel et al. between pre-mRNA and mRNA accumulation was quite surprising. For instance, laminin gamma 2 pre-mRNA was seen to accumulate in the first hour following EGF stimulation, while the corresponding spliced mRNA did not rise until 4–8 h after the stimulus. Using LPS and Lipid-A as stimuli, Bhatt et al. [52] isolated chromatin associated RNA, to find that immediately following stimulation, un-spliced transcripts accumulate and remain associated with the chromatin. Subsequently after a delay of between 10 and 30 min, spliced mRNAs appeared in the cytoplasm. Altogether, these studies suggest that regulation of RNA splicing may be an important control point in determining the time course of gene expression. 8. Transcription-mediated regulation of cell migration by growth factors To exemplify the ability of cytoplasmic and nuclear events to collaboratively initiate long lasting, delayed responses to growth factors, we focus below on EGF-regulated migration of epithelial

cells. To sustain a migratory phenotype, cells apply both early transcription-independent events, and transcriptional switches. Several cytoplasmic pathways are rapidly engaged by growth factorstimulated migrating cells. For example, phospholipase C (PLC) gamma, the enzyme that hydrolyzes PIP2, is rapidly phosphorylated and activated by EGFR, and this is essential for the induction of cellular motility [53]. The underlying mechanism entails activation of PIP2-inhibited actin-modifying proteins, such as gelsolin [54] and cofilin [55]. Likewise, phosphoinisitides generated by PI3K, specifically PI(3,4,5)P3 [56] and its dephosphorylated product PI(3,4)P2 [57], establish signposts essential for invadopodia formation at the ventral side of migrating cells [58]. In addition, activation of PI3K and Src by EGF leads to stimulation of RAC1, which enhances cell migration [59]. In parallel, active ERK regulates membrane protrusions [60] and localizes to the leading edge, where it phosphorylates cortactin and activates the ARP2/3 actin nucleator [61]. In addition, ERK activates the myosin light chain kinase (MLCK), which induces phosphorylation of myosin and leads to actin polymerization and membrane protrusion [62]. These and additional cytoplasmic processes precede transcription-controlled molecular switches, which convert a polarized epithelium to a collection of motile cells [63,64]. This process of epithelial-to-mesenchymal transition (EMT) involves loss of epithelial markers like E-cadherin, and gain of mesenchymal markers, such as N-cadherin, fibronectin and several transcription factors, including Ets-1 and snail [64]. Loss of E-cadherin, the major hallmark of EMT, is often coupled to up-regulation of another calcium-dependent cell adhesion molecule, N-cadherin [65]. For example, cells stimulated with TGF-beta show increased N-cadherin expression and loss of junctional E-cadherin [66]. This switch is directly regulated by a group of repressors, like snail and the ZEB proteins, and indirectly by twist, goosecoid, E2.2 and FOXC2 [67]. Stimulation of mammary cells with EGF activates another switch, namely transcriptional up-regulation of CTEN (also called tensin4) and a concomitant down-regulation of tensin 3 [68]. This reciprocal switch likely enables CTEN to displace tensins from the cytoplasmic tail of integrins thereby disassembling focal adhesions and promoting cell migration. By contrasting a signaling cascade leading to cell migration (stimulated by EGF) and a distinct pathway culminating in cell proliferation (stimulated by serum), we uncovered another transcriptional switch, which involves Erk-mediated phosphorylation of ERF, an ETS family member, translocation of the latter out of the nucleus and transcription of Egr1, an IEG that regulates several target genes essential for cell migration (e.g., PLAUR, SerpinB2 and the parathyroid hormonerelated protein, PTHrP) [21]. In addition, Egr1 also regulates several inhibitors of cell migration, such as the MAPK phosphatase DUSP4, EGR3 and the repressor BHLHB2, which likely represents homeostatic regulation of cell migration by IEGs like Egr1. 9. Concluding remarks Although acquired somatic aberrations and multiple growth factors propel tumor progression, the latter, along with other tumor–stroma interactions offer genetically more stable targets for cancer therapy. Mounting a strategy able to intercept growth factor action entails in-depth understanding of molecular mechanisms underlying not only signal transduction but also organization of cellular networks and control circuitries. By highlighting time series analyses and focusing on the EGFR pathway, we portrayed a highly dynamic order of biochemical events, which initiates at the plasma membrane and culminates in the nucleus. These transcriptional steps are very logical, highly reproducible and common to multiple growth factors, but they appear to be distorted in tumors. For example, the well-characterized IEG group of

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

inducible transcription factors has served as a rich source of protooncogenes, counterparts of retrovirus-encoded cancer promoting genes [69]. Conversely, the newly discovered group of ID-miRs is commonly down-regulated in mammary tumors, as compared to the surrounding normal epithelium [36]. Similarly, the group of DEGs, which trails the IEGs, is significantly down-regulated in different tumor types [30], as if some human tumors are locked in a growth factor stimulated ‘active’ state, which would normally return to a basal level soon after stimulation. The final and most persistent wave of growth factor inducible genes, might be also the most divergent one, as it includes both oncogenic and tumor suppressor genes and reflects long term phenotypic alterations in cell proliferation, migration and differentiation. Studying each of these genes and understanding their group dynamics might spearhead the identification of new biomarkers and novel targets for drug intervention. Acknowledgments Our research is supported by the U.S. National Cancer Institute, the European Research Council, the Seventh Framework Program of the European Commission, the German-Israeli Project Cooperation (DIP), the Israel Cancer Research Fund and the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation. Y.Y. is the incumbent of the Harold and Zelda Goldenberg Professorial Chair. Our experiments are routinely performed in the Marvin Tanner Laboratory for Cancer Research. References [1] Weinstein, I.B. and Joe, A.K. (2006) Mechanisms of disease: oncogene addiction – a rationale for molecular targeting in cancer therapy. Nat. Clin. Pract. Oncol. 3, 448–457. [2] Perez-Soler, R. (2004) The role of erlotinib (Tarceva, OSI 774) in the treatment of non-small cell lung cancer. Clin. Cancer Res. 10, 4238s–4240s. [3] Slamon, D.J., Leyland-Jones, B., Shak, S., Fuchs, H., Paton, V., Bajamonde, A., Fleming, T., Eiermann, W., Wolter, J., Pegram, M., Baselga, J. and Norton, L. (2001) Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344, 783– 792. [4] Stephens, P.J., Tarpey, P.S., Davies, H., Van Loo, P., Greenman, C., Wedge, D.C., Nik-Zainal, S., Martin, S., Varela, I., Bignell, G.R., Yates, L.R., Papaemmanuil, E., Beare, D., Butler, A., Cheverton, A., Gamble, J., Hinton, J., Jia, M., Jayakumar, A., Jones, D., Latimer, C., Lau, K.W., McLaren, S., McBride, D.J., Menzies, A., Mudie, L., Raine, K., Rad, R., Chapman, M.S., Teague, J., Easton, D., Langerod, A., Lee, M.T., Shen, C.Y., Tee, B.T., Huimin, B.W., Broeks, A., Vargas, A.C., Turashvili, G., Martens, J., Fatima, A., Miron, P., Chin, S.F., Thomas, G., Boyault, S., Mariani, O., Lakhani, S.R., van de Vijver, M., van ‘t Veer, L., Foekens, J., Desmedt, C., Sotiriou, C., Tutt, A., Caldas, C., Reis-Filho, J.S., Aparicio, S.A., Salomon, A.V., BorresenDale, A.L., Richardson, A.L., Campbell, P.J., Futreal, P.A. and Stratton, M.R. (2012) The landscape of cancer genes and mutational processes in breast cancer. Nature 486, 400–404. [5] Engelman, J.A. and Settleman, J. (2008) Acquired resistance to tyrosine kinase inhibitors during cancer therapy. Curr. Opin. Genet. Dev. 18, 73–79. [6] Khambata-Ford, S., Garrett, C.R., Meropol, N.J., Basik, M., Harbison, C.T., Wu, S., Wong, T.W., Huang, X., Takimoto, C.H., Godwin, A.K., Tan, B.R., Krishnamurthi, S.S., Burris 3rd, H.A., Poplin, E.A., Hidalgo, M., Baselga, J., Clark, E.A. and Mauro, D.J. (2007) Expression of epiregulin and amphiregulin and K-ras mutation status predict disease control in metastatic colorectal cancer patients treated with cetuximab. J. Clin. Oncol. 25, 3230–3237. [7] Witsch, E., Sela, M. and Yarden, Y. (2010) Roles for growth factors in cancer progression. Physiology (Bethesda) 25, 85–101. [8] Minn, A.J., Gupta, G.P., Siegel, P.M., Bos, P.D., Shu, W., Giri, D.D., Viale, A., Olshen, A.B., Gerald, W.L. and Massague, J. (2005) Genes that mediate breast cancer metastasis to lung. Nature 436, 518–524. [9] Eckstein, N., Servan, K., Girard, L., Cai, D., von Jonquieres, G., Jaehde, U., Kassack, M.U., Gazdar, A.F., Minna, J.D. and Royer, H.D. (2008) Epidermal growth factor receptor pathway analysis identifies amphiregulin as a key factor for cisplatin resistance of human breast cancer cells. J. Biol. Chem. 283, 739–750. [10] Avraham, R. and Yarden, Y. (2011) Feedback regulation of EGFR signalling: decision making by early and delayed loops. Nat. Rev. Mol. Cell Biol. 12, 104– 117. [11] Deribe, Y.L., Pawson, T. and Dikic, I. (2010) Post-translational modifications in signal integration. Nat. Struct. Mol. Biol. 17, 666–672. [12] Hynes, N.E. and MacDonald, G. (2009) ErbB receptors and signaling pathways in cancer. Curr. Opin. Cell Biol. 21, 177–184.

7

[13] Shaul, Y.D. and Seger, R. (2007) The MEK/ERK cascade: from signaling specificity to diverse functions. Biochim. Biophys. Acta 1773, 1213–1226. [14] Katz, M., Amit, I. and Yarden, Y. (2007) Regulation of MAPKs by growth factors and receptor tyrosine kinases. Biochim. Biophys. Acta 1773, 1161–1176. [15] Hay, N. and Sonenberg, N. (2004) Upstream and downstream of mTOR. Genes Dev. 18, 1926–1945. [16] Engelman, J.A., Luo, J. and Cantley, L.C. (2006) The evolution of phosphatidylinositol 3-kinases as regulators of growth and metabolism. Nat. Rev. Genet. 7, 606–619. [17] Laplante, M. and Sabatini, D.M. (2012) MTOR signaling in growth control and disease. Cell 149, 274–293. [18] Mosesson, Y., Mills, G.B. and Yarden, Y. (2008) Derailed endocytosis: an emerging feature of cancer. Nat. Rev. Cancer 8, 835–850. [19] Ebner, R. and Derynck, R. (1991) Epidermal growth factor and transforming growth factor-alpha: differential intracellular routing and processing of ligand-receptor complexes. Cell Regul. 2, 599–612. [20] Lenferink, A.E., Pinkas-Kramarski, R., van de Poll, M.L., van Vugt, M.J., Klapper, L.N., Tzahar, E., Waterman, H., Sela, M., van Zoelen, E.J. and Yarden, Y. (1998) Differential endocytic routing of homo- and hetero-dimeric ErbB tyrosine kinases confers signaling superiority to receptor heterodimers. EMBO J. 17, 3385–3397. [21] Tarcic, G., Avraham, R., Pines, G., Amit, I., Shay, T., Lu, Y., Zwang, Y., Katz, M., Ben-Chetrit, N., Jacob-Hirsch, J., Virgilio, L., Rechavi, G., Mavrothalassitis, G., Mills, G.B., Domany, E. and Yarden, Y. (2012) EGR1 and the ERK-ERF axis drive mammary cell migration in response to EGF. FASEB J. 26, 1582–1592. [22] Huang, H. and Tindall, D.J. (2007) Dynamic FoxO transcription factors. J. Cell Sci. 120, 2479–2487. [23] Plotnikov, A., Zehorai, E., Procaccia, S. and Seger, R. (2011) The MAPK cascades: signaling components, nuclear roles and mechanisms of nuclear translocation. Biochim. Biophys. Acta 1813, 1619–1633. [24] Buchwalter, G., Gross, C. and Wasylyk, B. (2004) Ets ternary complex transcription factors. Gene 324, 1–14. [25] Xing, J., Ginty, D.D. and Greenberg, M.E. (1996) Coupling of the RAS-MAPK pathway to gene activation by RSK2, a growth factor-regulated CREB kinase. Science 273, 959–963. [26] Quesnelle, K.M., Boehm, A.L. and Grandis, J.R. (2007) STAT-mediated EGFR signaling in cancer. J. Cell. Biochem. 102, 311–319. [27] Köstler, W.J., Zeisel, A., Körner, C., Tsai, J.M., Jacob-Hirsch, J., Ben-Chetrit, N., Sharma, K., Cohen-Dvashi, H., Yitzhaky, A., Lader, E., Tschulena, U., Rechavi, G., Domany, E., Wiemann, S. and Yarden, Y. (2013) Epidermal growth-factor – induced transcript isoform variation drives mammary cell migration. PLoS One 8, e80566. [28] Hah, N., Danko, C.G., Core, L., Waterfall, J.J., Siepel, A., Lis, J.T. and Kraus, W.L. (2011) A rapid, extensive, and transient transcriptional response to estrogen signaling in breast cancer cells. Cell 145, 622–634. [29] Reddy, T.E., Pauli, F., Sprouse, R.O., Neff, N.F., Newberry, K.M., Garabedian, M.J. and Myers, R.M. (2009) Genomic determination of the glucocorticoid response reveals unexpected mechanisms of gene regulation. Genome Res. 19, 2163–2171. [30] Amit, I., Citri, A., Shay, T., Lu, Y., Katz, M., Zhang, F., Tarcic, G., Siwak, D., Lahad, J., Jacob-Hirsch, J., Amariglio, N., Vaisman, N., Segal, E., Rechavi, G., Alon, U., Mills, G.B., Domany, E. and Yarden, Y. (2007) A module of negative feedback regulators defines growth factor signaling. Nat. Genet. 39, 503–512. [31] Tullai, J.W., Schaffer, M.E., Mullenbrock, S., Sholder, G., Kasif, S. and Cooper, G.M. (2007) Immediate-early and delayed primary response genes are distinct in function and genomic architecture. J. Biol. Chem. 282, 23981–23995. [32] Lau, L.F. and Nathans, D. (1987) Expression of a set of growth-related immediate early genes in BALB/c 3T3 cells: coordinate regulation with c-fos or c-myc. Proc. Natl. Acad. Sci. USA 84, 1182–1186. [33] Ramirez-Carrozzi, V.R., Braas, D., Bhatt, D.M., Cheng, C.S., Hong, C., Doty, K.R., Black, J.C., Hoffmann, A., Carey, M. and Smale, S.T. (2009) A unifying model for the selective regulation of inducible transcription by CpG islands and nucleosome remodeling. Cell 138, 114–128. [34] Hargreaves, D.C., Horng, T. and Medzhitov, R. (2009) Control of inducible gene expression by signal-dependent transcriptional elongation. Cell 138, 129–145. [35] Cohen, N.M., Kenigsberg, E. and Tanay, A. (2011) Primate CpG islands are maintained by heterogeneous evolutionary regimes involving minimal selection. Cell 145, 773–786. [36] Avraham, R., Sas-Chen, A., Manor, O., Steinfeld, I., Shalgi, R., Tarcic, G., Bossel, N., Zeisel, A., Amit, I., Zwang, Y., Enerly, E., Russnes, H.G., Biagioni, F., Mottolese, M., Strano, S., Blandino, G., Borresen-Dale, A.L., Pilpel, Y., Yakhini, Z., Segal, E. and Yarden, Y. (2010) EGF decreases the abundance of microRNAs that restrain oncogenic transcription factors. Sci. Signal. 3, ra43. [37] Hao, S. and Baltimore, D. (2009) The stability of mRNA influences the temporal order of the induction of genes encoding inflammatory molecules. Nat. Immunol. 10, 281–288. [38] Alon, U. (2006) An Introduction to Systems Biology, CRC Press. [39] Chen, C.Y. and Shyu, A.B. (1994) Selective degradation of early-response-gene mRNAs: functional analyses of sequence features of the AU-rich elements. Mol. Cell. Biol. 14, 8471–8482. [40] Brennan, S.E., Kuwano, Y., Alkharouf, N., Blackshear, P.J., Gorospe, M. and Wilson, G.M. (2009) The mRNA-destabilizing protein tristetraprolin is suppressed in many cancers, altering tumorigenic phenotypes and patient prognosis. Cancer Res. 69, 5168–5176.

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036

8

M.E. Feldman, Y. Yarden / FEBS Letters xxx (2014) xxx–xxx

[41] Suswam, E.A., Shacka, J.J., Walker, K., Lu, L., Li, X., Si, Y., Zhang, X., Zheng, L., Nabors, L.B., Cao, H. and King, P.H. (2013) Mutant tristetraprolin: a potent inhibitor of malignant glioma cell growth. J. Neurooncol. 113, 195–205. [42] Mahat, D.B., Brennan-Laun, S.E., Fialcowitz-White, E.J., Kishor, A., Ross, C.R., Pozharskaya, T., Rawn, J.D., Blackshear, P.J., Hassel, B.A. and Wilson, G.M. (2012) Coordinated expression of tristetraprolin post-transcriptionally attenuates mitogenic induction of the oncogenic Ser/Thr kinase Pim-1. PLoS One 7, e33194. [43] Selmi, T., Martello, A., Vignudelli, T., Ferrari, E., Grande, A., Gemelli, C., Salomoni, P., Ferrari, S. and Zanocco-Marani, T. (2012) ZFP36 expression impairs glioblastoma cell lines viability and invasiveness by targeting multiple signal transduction pathways. Cell Cycle 11, 1977–1987. [44] Frosi, Y., Anastasi, S., Ballaro, C., Varsano, G., Castellani, L., Maspero, E., Polo, S., Alema, S. and Segatto, O. (2010) A two-tiered mechanism of EGFR inhibition by RALT/MIG6 via kinase suppression and receptor degradation. J. Cell Biol. 189, 557–571. [45] Segatto, O., Anastasi, S. and Alema, S. (2011) Regulation of epidermal growth factor receptor signalling by inducible feedback inhibitors. J. Cell Sci. 124, 1785–1793. [46] Caunt, C.J. and Keyse, S.M. (2013) Dual-specificity MAP kinase phosphatases (MKPs): shaping the outcome of MAP kinase signalling. FEBS J. 280, 489–504. [47] Ben-Chetrit, N., Tarcic, G. and Yarden, Y. (2013) ERK-ERF-EGR1, a novel switch underlying acquisition of a motile phenotype. Cell Adhes. Migration 7, 33–37. [48] Lanahan, A., Williams, J.B., Sanders, L.K. and Nathans, D. (1992) Growth factorinduced delayed early response genes. Mol. Cell. Biol. 12, 3919–3929. [49] Danko, C.G., Hah, N., Luo, X., Martins, A.L., Core, L., Lis, J.T., Siepel, A. and Kraus, W.L. (2013) Signaling pathways differentially affect RNA polymerase II initiation, pausing, and elongation rate in cells. Mol. Cell 50, 212–222. [50] S. Scherer, A Short Guide to the Human, Genome, (2008) 173. [51] Zeisel, A., Kostler, W.J., Molotski, N., Tsai, J.M., Krauthgamer, R., Jacob-Hirsch, J., Rechavi, G., Soen, Y., Jung, S., Yarden, Y. and Domany, E. (2011) Coupled premRNA and mRNA dynamics unveil operational strategies underlying transcriptional responses to stimuli. Mol. Syst. Biol. 7, 529. [52] Bhatt, D.M., Pandya-Jones, A., Tong, A.J., Barozzi, I., Lissner, M.M., Natoli, G., Black, D.L. and Smale, S.T. (2012) Transcript dynamics of proinflammatory genes revealed by sequence analysis of subcellular RNA fractions. Cell 150, 279–290. [53] Chen, P., Xie, H., Sekar, M.C., Gupta, K. and Wells, A. (1994) Epidermal growth factor receptor-mediated cell motility: phospholipase C activity is required, but mitogen-activated protein kinase activity is not sufficient for induced cell movement. J. Cell Biol. 127, 847–857. [54] Chen, P., Murphy-Ullrich, J.E. and Wells, A. (1996) A role for gelsolin in actuating epidermal growth factor receptor-mediated cell motility. J. Cell Biol. 134, 689–698. [55] van Rheenen, J., Song, X., van Roosmalen, W., Cammer, M., Chen, X., Desmarais, V., Yip, S.C., Backer, J.M., Eddy, R.J. and Condeelis, J.S. (2007) EGF-induced PIP2 hydrolysis releases and activates cofilin locally in carcinoma cells. J. Cell Biol. 179, 1247–1259. [56] Yamaguchi, H., Yoshida, S., Muroi, E., Yoshida, N., Kawamura, M., Kouchi, Z., Nakamura, Y., Sakai, R. and Fukami, K. (2011) Phosphoinositide 3-kinase signaling pathway mediated by p110alpha regulates invadopodia formation. J. Cell Biol. 193, 1275–1288. [57] Oikawa, T., Itoh, T. and Takenawa, T. (2008) Sequential signals toward podosome formation in NIH-src cells. J. Cell Biol. 182, 157–169. [58] Courtneidge, S.A., Azucena, E.F., Pass, I., Seals, D.F. and Tesfay, L. (2005) The SRC substrate Tks5, podosomes (invadopodia), and cancer cell invasion. Cold Spring Harb. Symp. Quant. Biol. 70, 167–171. [59] Dise, R.S., Frey, M.R., Whitehead, R.H. and Polk, D.B. (2008) Epidermal growth factor stimulates Rac activation through Src and phosphatidylinositol 3-kinase to promote colonic epithelial cell migration. Am. J. Physiol. Gastrointest. Liver Physiol. 294, G276–G285. [60] Mendoza, M.C., Er, E.E., Zhang, W., Ballif, B.A., Elliott, H.L., Danuser, G. and Blenis, J. (2011) ERK-MAPK drives lamellipodia protrusion by activating the WAVE2 regulatory complex. Mol. Cell 41, 661–671. [61] Kelley, L.C., Hayes, K.E., Ammer, A.G., Martin, K.H. and Weed, S.A. (2010) Cortactin phosphorylated by ERK1/2 localizes to sites of dynamic actin regulation and is required for carcinoma lamellipodia persistence. PLoS One 5, e13847. [62] Totsukawa, G., Wu, Y., Sasaki, Y., Hartshorne, D.J., Yamakita, Y., Yamashiro, S. and Matsumura, F. (2004) Distinct roles of MLCK and ROCK in the regulation of membrane protrusions and focal adhesion dynamics during cell migration of fibroblasts. J. Cell Biol. 164, 427–439. [63] Kalluri, R. and Weinberg, R.A. (2009) The basics of epithelial-mesenchymal transition. J. Clin. Invest. 119, 1420–1428. [64] Thiery, J.P., Acloque, H., Huang, R.Y. and Nieto, M.A. (2009) Epithelialmesenchymal transitions in development and disease. Cell 139, 871–890.

[65] Hazan, R.B., Phillips, G.R., Qiao, R.F., Norton, L. and Aaronson, S.A. (2000) Exogenous expression of N-cadherin in breast cancer cells induces cell migration, invasion, and metastasis. J. Cell Biol. 148, 779–790. [66] Bhowmick, N.A., Ghiassi, M., Bakin, A., Aakre, M., Lundquist, C.A., Engel, M.E., Arteaga, C.L. and Moses, H.L. (2001) Transforming growth factor-beta1 mediates epithelial to mesenchymal transdifferentiation through a RhoAdependent mechanism. Mol. Biol. Cell 12, 27–36. [67] Peinado, H., Olmeda, D. and Cano, A. (2007) Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat. Rev. Cancer 7, 415–428. [68] Katz, M., Amit, I., Citri, A., Shay, T., Carvalho, S., Lavi, S., Milanezi, F., Lyass, L., Amariglio, N., Jacob-Hirsch, J., Ben-Chetrit, N., Tarcic, G., Lindzen, M., Avraham, R., Liao, Y.C., Trusk, P., Lyass, A., Rechavi, G., Spector, N.L., Lo, S.H., Schmitt, F., Bacus, S.S. and Yarden, Y. (2007) A reciprocal tensin-3-cten switch mediates EGF-driven mammary cell migration. Nat. Cell Biol. 9, 961–969. [69] Ransone, L.J. and Verma, I.M. (1990) Nuclear proto-oncogenes fos and jun. Annu. Rev. Cell Biol. 6, 539–557. [70] Greenberg, M.E., Hermanowski, A.L. and Ziff, E.B. (1986) Effect of protein synthesis inhibitors on growth factor activation of c-fos, c-myc, and actin gene transcription. Mol. Cell. Biol. 6, 1050–1057. [71] Wilson, T. and Treisman, R. (1988) Removal of poly(A) and consequent degradation of c-fos mRNA facilitated by 3' AU-rich sequences. Nature 336, 396–399. [72] Mukherjee, N., Jacobs, N.C., Hafner, M., Kennington, E.A., Nusbaum, J.D., Tuschl, T., Blackshear, P.J. and Ohler, U. (2014) Global target mRNA specification and regulation by the RNA-binding protein ZFP36. Genome Biol. 15, R12. [73] Baou, M., Jewell, A. and Murphy, J.J. (2009) TIS11 family proteins and their roles in posttranscriptional gene regulation. J. Biomed. Biotechnol. 2009, 634520. [74] Li, S., Fu, H., Wang, Y., Tie, Y., Xing, R., Zhu, J., Sun, Z., Wei, L. and Zheng, X. (2009) MicroRNA-101 regulates expression of the v-fos FBJ murine osteosarcoma viral oncogene homolog (FOS) oncogene in human hepatocellular carcinoma. Hepatology 49, 1194–1202. [75] Zhang, L., Huang, J., Yang, N., Greshock, J., Megraw, M.S., Giannakakis, A., Liang, S., Naylor, T.L., Barchetti, A., Ward, M.R., Yao, G., Medina, A., O’brien-Jenkins, A., Katsaros, D., Hatzigeorgiou, A., Gimotty, P.A., Weber, B.L. and Coukos, G. (2006) MicroRNAs exhibit high frequency genomic alterations in human cancer. Proc. Natl. Acad. Sci. USA 103, 9136–9141. [76] Meng, F., Wehbe-Janek, H., Henson, R., Smith, H. and Patel, T. (2008) Epigenetic regulation of microRNA-370 by interleukin-6 in malignant human cholangiocytes. Oncogene 27, 378–386. [77] Lehmann, U., Hasemeier, B., Christgen, M., Muller, M., Römermann, D., Länger, F. and Kreipe, H. (2008) Epigenetic inactivation of microRNA gene hsa-mir-9-1 in human breast cancer. J. Pathol. 214, 17–24. [78] Baron, V., Adamson, E.D., Calogero, A., Ragona, G. and Mercola, D. (2006) The transcription factor Egr1 is a direct regulator of multiple tumor suppressors including TGFb1, PTEN, p53, and fibronectin. Cancer Gene Ther. 13, 115–124. [79] Keyse, S.M. (2008) Dual-specificity MAP kinase phosphatases (MKPs) and cancer. Cancer Metastasis Rev. 27, 253–261. [80] Narla, G., Heath, K.E., Reeves, H.L., Li, D., Giono, L.E., Kimmelman, A.C., Glucksman, M.J., Narla, J., Eng, F.J., Chan, A.M., Ferrari, A.C., Martignetti, J.A. and Friedman, S.L. (2001) KLF6, a candidate tumor suppressor gene mutated in prostate cancer. Science 294, 2563–2566. [81] Zhang, Y.W. and Vande Woude, G.F. (2007) Mig-6, signal transduction, stress response and cancer. Cell Cycle 6, 507–513. [82] Ying, H., Zheng, H., Scott, K., Wiedemeyer, R., Yan, H., Lim, C., Huang, J., Dhakal, S., Ivanova, E., Xiao, Y., Zhang, H., Hu, J., Stommel, J.M., Lee, M.A., Chen, A.J., Paik, J.H., Segatto, O., Brennan, C., Elferink, L.A., Wang, Y.A., Chin, L. and DePinho, R.A. (2010) Mig-6 controls EGFR trafficking and suppresses gliomagenesis. Proc. Natl. Acad. Sci. USA 107, 6912–6917. [83] Amit, I., Garber, M., Chevrier, N., Leite, A.P., Donner, Y., Eisenhaure, T., Guttman, M., Grenier, J.K., Li, W., Zuk, O., Schubert, L.A., Birditt, B., Shay, T., Goren, A., Zhang, X., Smith, Z., Deering, R., McDonald, R.C., Cabili, M., Bernstein, B.E., Rinn, J.L., Meissner, A., Root, D.E., Hacohen, N. and Regev, A. (2009) Unbiased reconstruction of a mammalian transcriptional network mediating pathogen responses. Science 326, 257–263. [84] Schulze, A., Lehmann, K., Jefferies, H.B., McMahon, M. and Downward, J. (2001) Analysis of the transcriptional program induced by Raf in epithelial cells. Genes Dev. 15, 981–994. [85] Sparmann, A. and Bar-Sagi, D. (2004) Ras-induced interleukin-8 expression plays a critical role in tumor growth and angiogenesis. Cancer Cell 6, 447–458. [86] Egeblad, M. and Werb, Z. (2002) New functions for the matrix metalloproteinases in cancer progression. Nat. Rev. Cancer 2, 161–174. [87] Albino, D., Longoni, N., Curti, L., Mello-Grand, M., Pinton, S., Civenni, G., Thalmann, G., D’Ambrosio, G., Sarti, M., Sessa, F., Chiorino, G., Catapano, C.V. and Carbone, G.M. (2012) ESE3/EHF controls epithelial cell differentiation and its loss leads to prostate tumors with mesenchymal and stem-like features. Cancer Res. 72, 2889–2900.

Please cite this article in press as: Feldman, M.E. and Yarden, Y. Steering tumor progression through the transcriptional response to growth factors and stroma. FEBS Lett. (2014), http://dx.doi.org/10.1016/j.febslet.2014.05.036