Each week we showcase a hot new cell signaling article from a Nature Publishing Group journal. Free full text access to the paper will be maintained for three months, after which the research highlight will be accessible via the Updates page.
The sensitivity of breast cancer tumors to tamoxifen depends upon the inhibition and activation of ERBB2 gene expression, which is controlled by the competitive binding of PAX2 and AIB-1 to a cis-regulatory element.
Tamoxifen is an analogue of the hormone estrogen and is widely used in the treatment of estrogen receptor-positive breast cancer. Tamoxifen and estrogen both bind to estrogen receptor to repress the transcription of ERBB2 (epidermal growth factor (EGF)-receptor-related protein tyrosine kinase B2). Tamoxifen resistance is characterized by elevated levels of ERBB2 and is relatively common in some breast cancer tumors. However, a molecular mechanism linking estrogen receptor signaling, ERBB2 expression and tamoxifen resistance has remained elusive. Carroll and colleagues now report in Nature that ERBB2 expression is controlled by the balance between the estrogen receptor co-activator AIB-1 (amplified in breast cancer-1) and the transcription factor PAX2 (paired box 2 gene product), which both competitively bind to the same ERBB2 regulatory element.
A novel estrogen receptor-α binding site was identified in intron 4 of the longer ERBB2 isoform. PAX2 is known to be expressed in certain breast cancers and was shown to interact with estrogen receptor. The authors found that PAX2 and estrogen receptor both bound to the same site on ERBB2 following tamoxifen or estrogen treatment, and observed a reduction in ERBB2 mRNA levels upon PAX2 binding. Conversely, small interfering RNA (siRNA)-mediated knockdown of PAX2 derepressed tamoxifen- and estrogen-mediated inhibition of ERBB2 expression and allowed the proliferation of estrogen receptor-positive breast cancer cell lines. Thus, PAX2 appears to act as an ERBB2 transcriptional repressor.
The authors found that the response of breast cancer cells to tamoxifen is regulated by competition between AIB-1 and PAX2 binding to cis-regulatory elements in ERBB2. Indeed, a decrease in PAX2 expression in tamoxifen-resistant (Tam-R) cells correlated with an increase in ERBB2 expression. Estrogen receptor was recruited to ERBB2, whereas PAX2 binding decreased, in response to tamoxifen with a similar efficiency in Tam-R and tamoxifen-responsive cells. Conversely, AIB-1 binding increased in Tam-R cells in response to tamoxifen treatment. Overexpression of PAX2 reduced AIB-1 binding to ERBB2, restored tamoxifen-mediated repression of ERBB2 and inhibited cellular proliferation. These findings suggest that AIB-1-mediated ERBB2 expression is antagonized by competitive PAX2 binding.
Immunohistochemical staining of tamoxifen-treated estrogen receptor-positive breast cancer tumors demonstrated that PAX2 expression in the absence of AIB-1 correlated with cancer recurrence-free patient survival and a low rate of ERBB2 expression. Patients with PAX2/AIB-1-positive tumors had a higher rate of cancer recurrence. Thus, levels of PAX2 and AIB-1 expression can serve as prognostic indicators of patients' response to tamoxifen treatment. Further study is required in order to define the mechanism of PAX2 regulation in response to estrogen and tamoxifen.
Anna S. Kushnir Nature Publishing Group
Original References:
Hurtado, A. et al. Regulation of ERBB2 by oestrogen receptor–PAX2 determines response to tamoxifen Nature advance online publication, 12 November 2008 (DOI:10.1038/nature07483) Full text | PDF | Subscribe to Nature
Proteolytic cleavage of the Toll-like receptor TLR9 creates a functional receptor that is retained at intracellular endolysosome compartments.
Toll-like receptors (TLRs) sense microbial nucleic acids and stimulate an immune response. However, it is not known how TLRs distinguish between self and foreign genetic material. There is some evidence that their subcellular localization governs selectivity as ligand recognition appears to occur only at endolysosomes, but this hypothesis is complicated by an incomplete understanding of TLR trafficking. In Nature, Gregory Barton and colleagues now report that nascent TLR9 transits through the Golgi en route to endolysosomes, where it is cleaved. Surprisingly, only cleaved TLR9 binds to the adaptor MyD88 to affect an immune response.
Exogenous expression of TLR9 in a variety of macrophage and dendritic cell lines resulted in the phagosomal accumulation of an 80-kDa TLR9 fragment comprised of the cytoplasmic and transmembrane domains and half of the ectodomain. The protease cathepsin produced an 80-kDa TLR9 fragment in vitro; however, neither specific cathepsin inhibitors nor broad-scale protease inhibitors prevented in vivo processing, suggesting that TLR9 is the target of other or multiple proteases. TLR7 was also cleaved, suggesting that receptor cleavage may be conserved amongst many nucleic-acid-sensing receptors.
Previous studies had indicated that full-length TLRs transit from the endoplasmic reticulum (ER) to endolysosomes without passing through the Golgi. However, the authors found that the TLR9 fragments had acquired Golgi-specific glycosyl modifications and confirmed that a small pool of full-length TLR9 trafficked through the Golgi and was then cleaved. Depletion of the membrane protein UNC93B1 — which delivers nucleic-acid-sensing TLRs to endolysosomes — inhibited TLR9 cleavage and phagosome accumulation, and blocked TLR9-mediated TNF-α production. However, production of full-length TLR9 was unchanged in the absence UNC93B1. Furthermore, although both full-length and cleaved TLR9 were able to bind ligand, only the processed form interacted with the MyD88 adaptor protein, suggesting that it is the functional form of the receptor.
Interestingly, the authors were unable to force plasma membrane localization of the TLR9 fragment. A TLR9 construct with a plasma-membrane-targeting signal passed through the Golgi but was not cleaved and had no response to ligand. Point mutations in the membrane-targeting sequence restored processing and ligand sensitivity, but destroyed plasma membrane localization. Thus, the stringent endolysosomal localization of cleaved TLR9 greatly reduces the possibility of mature TLR9 trafficking to the plasma membrane and encountering 'self' genetic material. These data identify a novel mechanism by which proteolytic processing and subcellular localization tightly regulate the activation of nucleic-acid-sensing TLRs.
Emily J. Chenette Signaling Gateway
Original References:
Ewald, S. E. et al. The ectodomain of Toll-like receptor 9 is cleaved to generate a functional receptor Nature advance online publication, 28 September 2008, Full text | PDF | Subscribe to Nature
Stress signaling: Stress granules RACK1 up MTK1
Stress granule formation inhibits apoptosis by suppressing stress-responsive MAPK pathways and sequestering the scaffold protein RACK1.
A stressed cell may either activate cell cycle checkpoints, and repair and protective pathways, depending on the type of stress and the cellular context. The formation of cytoplasmic stress granules prevents mis-folded protein aggregation to enable cell survival, whereas the activation of p38 and c-Jun N-terminal protein kinases (JNKs) leads to apoptosis. Crosstalk between stress-induced pro-survival and pro-apoptotic pathways has not yet been described. In Nature Cell Biology, Arimoto et al. report that stress granules inhibit apoptosis by sequestering RACK1 and preventing p38 and JNK activation.
The MAPK kinase kinase MTK1 mediates apoptosis in response to stress by activating p38 and JNK. Mass spectrometry of proteins that co-immunoprecipitated with tagged MTK1 revealed an interaction with the signaling scaffold protein RACK1. Knockdown of RACK1 expression by short hairpin RNA inhibition (shRNAi) reduced MTK1 activation in response to cellular stress, whereas RACK1 overexpression led to moderate increases in MTK1 activation at lower RACK1 concentrations. Thus, RACK1 is necessary, but not sufficient, for MTK1 activation.
The growth arrest and DNA damage-inducible (GADD45) family proteins bind to and activate MTK1 by relieving its autoinhibition, and allowing dimerization and autophosphorylation. Unlike GADD45, RACK1 was unable to relieve MTK1 autoinhibition, but it did promote MTK1 dimerization, suggesting that RACK1 acts as a bridge between two MTK1 molecules. Although RACK1 and GADD45 bind to the same region of MTK1, the binding of the two proteins appears to be independent. Taken together, these data suggest that RACK1 maintains MTK1 in an inactive but primed state in the absence of cellular stress.
MTK1 was found to colocalize with RACK1 in unstressed cells. However, inducing stress granule formation led to the relocalization of RACK1 to the stress granules and a decrease in MTK1 association. Furthermore, cellular stresses that induced stress granule formation inhibited GADD45-mediated MTK1 activation and induction of apoptosis, suggesting that stresses that induce the pro-survival stress granule response also suppress pro-apoptotic pathway activation.
Based on these findings, the authors propose a model in which RACK1 is sequestered to stress granules, preventing the activation of MTK1 by GADD45 and inhibiting apoptosis. Stresses that do not induce stress granule formation lead to GADD45-mediated MTK1 activation, p38/JNK phosphorylation and apoptosis. This suggests a potential mechanism for hypoxia-induced resistance to chemotherapy, as cells subjected to hypoxic shock form stress granules and are resistant to etoposide-induced apoptosis.
Anna S. Kushnir Nature Publishing Group
Original Reference:
Arimoto, K. et al. Formation of stress granules inhibits apoptosis by suppressing stress-responsive MAPK pathways Nature Cell Biology10, 1324-1332 (2008) Full text | PDF | Subscribe to Nature Cell Biology
Small molecule inhibitors: Double trouble
A rational design-based approach to developing targeted therapeutics has generated a small-molecule inhibitor that blocks both PI(3)K and tyrosine kinase activity.
Tyrosine kinases stimulate the Raf-MEK-Erk and phosphatidylinositol-3-OH kinase (PI(3)K) pathways and are mutationally activated in many human cancers. Small-molecule tyrosine-kinase inhibitors such as imatinib — which targets the oncogenic Bcr-Abl fusion protein and select receptor tyrosine kinases — have been used with clinical success. However, some cancers possess or acquire resistance to tyrosine kinase inhibition through drug-resistant mutations or activating mutations in PI(3)K pathway members. Thus, there is considerable interest in developing multi-kinase inhibitors that target the PI(3)K-mTOR (mammalian target of rapamycin) and tyrosine kinase pathways. In Nature Chemical Biology, Apsel et al. now report the generation of a novel pyrazolopyrimidine compound that inhibits tyrosine and PI(3) kinases and blocks the proliferation of many tumor cell lines, including imatinib-resistant chronic myelogenous leukemia (CML) cells.
Tyrosine kinases and PI(3)Ks lack primary sequence homology, but share a similar kinase domain architecture, suggesting that a single small molecule could target both families. To generate novel compounds with dual specificity, the authors screened a tyrosine kinase inhibitor library for compounds that blocked PI(3)K activity, and then performed structure-activity relationship (SAR) studies to increase potency and selectivity. One of the resulting novel compounds, PP121, exhibited nanomolar efficacy against both PI(3)K and tyrosine kinases. Importantly, PP121 did not inhibit serine/threonine kinase activity, an effect that was ascribed to a bulky 'gatekeeper' amino acid in the kinases that blocked inhibitor access to the ATP-binding pocket.
PP121 treatment inhibited the phosphorylation of PI(3)K-mTOR pathway members and induced G0/G1 cell cycle arrest of cancer cell lines containing mutations in PTEN (phosphatase and tensin homolog), the p110α PI(3)K catalytic subunit and Ras. However, PP121 also exhibited potent activity against tyrosine kinases. It blocked tyrosine phosphorylation and reversed morphological transformation induced by v-Src, as well as inhibiting Ret and VEGFR2 autophosphorylation and pathway activation.
Given that PP121 can inhibit both tyrosine kinases and PI(3)Ks, the authors next evaluated whether PP121 could block the growth of imatinib-resistant CML. Drug resistance in CML is achieved through a T315I mutation in Bcr-Abl (Bcr-AblT315I). Remarkably, even though PP121 did not inhibit Bcr-AblT315I-mediated tyrosine phosphorylation, it nevertheless blocked cellular proliferation by inducing G0/G1 cell cycle arrest. As the mTOR substrate S6 was not phosphorylated in these cells, the authors attributed this effect to PI(3)K-mTOR pathway inhibition.
This study provides compelling evidence that a rational design-based approach to targeting two signaling pathways can generate promising new anti-cancer drugs. It will be interesting to determine whether this technique could be applied to selectively inhibit other combinations of kinases.
Emily J. Chenette Signaling Gateway
Original References:
Apsel, B. et al. Targeted polypharmacology: discovery of dual inhibitors of tyrosine and phosphoinositide kinases Nature Chemical Biology4, 691-699 (2008) Full text | PDF | Subscribe to Nature Chemical Biology
Bilanges, B., Torbett, N. & Vanhaesebroeck, B. Killing two kinase families with one stone Nature Chemical Biology4, 648-649 (2008) Full text | PDF | Subscribe to Nature Chemical Biology
