Spotlight Lipid Peroxides and Signaling The activity of receptor protein tyrosine kinases is controlled by protein tyrosine phosphatases (PTPs). The activity of many PTPs can be suppressed by oxidation of an active site cysteine thiolate anion. The importance of glutathione peroxidase-4 (GPx4) in protecting PTPs from oxidative inactivation is the subject of a recent report by Conrad et al., using the platelet-derived growth factor β-receptor as a model system (PDGF β-R) [(2010) Proc. Natl. Acad. Sci. U.S.A., 107, 15774]. Conrad et al. used tamoxifen-inducible disruption of Gpx4 in mouse embryo fibroblasts to show that loss of GPx4 activity led to increased lipid peroxidation and increased PTP oxidation accompanied by decreased PTP activity. Higher levels of PDGF β-R phosphorylation, particularly on Y1009 and Y1021, correlated with elevated activity of phospholipase Cγ1 (a phosphoY1021-dependent enzyme) and increased lamellipodia
High Potency Protection The Keap1/Nrf2/ARE signaling pathway controls a network of over 100 genes that protect the cell from oxidative and electrophilic stress. Initiation of this pathway occurs through the reaction of an inducer molecule with cysteine sulfhydryl groups of Keap1, preventing it from targeting Nrf2 for proteasomal degradation. Nrf2 then translocates to the nucleus where it activates genes bearing one or more ARE consensus sequences in their promoters. The resultant transcriptional response protects the organism against conditions that contribute to the pathogenesis of chronic and degenerative diseases, leading to the quest for highly potent pathway inducers for clinical applications. One outcome of this quest is the semisynthetic triterpenoid CDDO, which has shown promise in a number of preclinical animal models. Now, Dinkova-Kostova et al. [(2010) J. Biol. Chem., published online August 26, DOI: 10.1074/jbc. M110.163485] report an even more potent acetylenic tricyclic bis(cyano enone) TBE 31 and explore the basis for its potency. TBE 31 induced the expression of ARE-dependent NAD(P)H:quinone oxidoreductase-1 (NQO1) in a cell-based high-throughput screen with an EC50 of 1.1 nM as compared to 2.5 nM for CDDO. Topical application of TBE 31 resulted in increased NQO1 and glutathione S-transferase (GST, another ARE-dependent gene) activity in the skin of mice. Oral administration resulted in a dose-dependent induction of both enzymes in stomach, skin, and liver, indicating excellent oral bioavailability. Failure of enzyme Published online 11/15/2010 • DOI: 10.1021/tx100330z © 2010 American Chemical Society
formation (a PDGF β-R-dependent cellular function). Increased PDGF β-R phosphorylation was attributable to a reduced rate of receptor dephosphorylation, consistent with low PTP activity. A specific inhibitor of 12- and 15-lipoxygenases AA861 reduced lipid peroxide and PDGF β-R phosphorylation levels in GPx4-depleted cells. This finding correlated with the observation that nanomolar concentrations of the 15-lipoxygenase product 15-hydroperoxyeicosatetraenoic acid increased the oxidation of three PTP enzymes. Inhibitors of other lipid oxygenases had no effect. These findings suggest that 12-/ 15-lipoxygenases are primarily responsible for formation of the lipid peroxides that accumulate in the absence of GPx4 activity and that these lipid peroxides, in turn, are primarily responsible for the inactivation of PTPs that regulate PDGF β-R signaling. • Carol A. Rouzer
induction in the brain suggested that TBE 31 does not pass the blood-brain barrier. An ARE-dependent luciferase reporter assay established that TBE 31 activates the Keap1/ Nrf2/ARE pathway, and ultraviolet spectroscopy confirmed that TBE 31 directly reacts with cysteine residues of Keap1.
Figure illustrating mechanism of Nrf2/Keap1/ARE-dependent signaling kindly provided by A. Dinkova-Kostova. Copyright 2010 A. Dinkova-Kostova.
Synthesis of a series of monocyclic cyano enones provided the foundation of structure-activity relationship studies. MCE 1, based on the structure of the acetylenecontaining C ring of TBE 31, exhibited relatively high potency (EC50 ) 34 nM) as compared to MCE 5 (EC50 ) 1.1 µM), based on the structure of the A ring of TBE 31. Removal of the acetylene group from MCE 1 reduced the potency by a factor of 5 (MCE 2, EC50 ) 150 nM), a finding consistent with the 20-fold increase in EC50 observed upon removal of the acetylene function of TBE 31. The results indicate that the high potency of TBE 31 derives primarily from its C ring, but that the presence of two Michael acceptors on the tricyclic scaffold and the acetylene Vol. 23,
No. 11, •
CHEMICAL RESEARCH IN TOXICOLOGY 1631
Spotlight functional group between the B and the C rings is also an important contributor. • Carol A. Rouzer
Key to RNAi Toxicity The great excitement over the therapeutic potential of RNA interference approaches to reduce the expression of pathogenic proteins in a number of different diseases has been dampened by the appearance of toxicity in preclinical and clinical trials. The observation of a dosedependent but sequence-independent toxicity associated with the hepatic expression of shRNA in vivo led Grimm et al. [(2010) J. Clin. Invest. 120, 3106] to propose that the mechanism of the cellular damage is saturation of the endogenous microRNA (miRNA) processing machinery. They now report evidence in support of that hypothesis. Using transgenic mice that express human R1-antitrypsin (hAAT) and an adeno-associated viral vector carrying shRNA against hAAT, Grimm et al. showed that high level shRNA expression caused a nearly 100% reduction in hepatic hAAT expression in vivo. However, toxicity rapidly led to cell death and regeneration resulting in loss of the shRNA vector and return of hAAT expression. Coexpression of the anti-hAAT shRNA with the karyopherin exportin-5 (Xpo-5), which transports miRNAs across the nuclear membrane, or Argonaute-2 (Ago-2), which is involved in miRNA processing, reduced liver toxicity and extended the period of shRNA blockade of hAAT expression. Coexpression of other Argonaute proteins (Ago-1, -3, and -4) reduced shRNA effectiveness, while coexpression of other miRNA processing proteins had no effect. The results suggested that the anti-hAAT shRNA interacts with all Argonaute proteins but that it likely saturates Ago2, leading to cellular toxicity. Grimm et al. showed that directing the shRNA to the 3′-UTR of the hAAT gene or reducing the homology between the shRNA and the target sequence reduced the interaction between the shRNA and Ago-2. However, designing shRNAs for clinical use based on this approach would likely be an impractical way to reduce toxicity. In contrast, when animals were pretreated with vectors leading to increased expression of both Xpo-5 and Ago-2, followed by the anti-hAAT shRNA, they exhibited markedly prolonged suppression of hAAT expression and minimal hepatotoxicity. Similarly, when the expression of the anti-hAAT shRNA was placed under the control of a weaker promoter, prolonged suppression of hAAT was achieved, despite a delayed onset. Similar results were obtained in the suppression of virus-related proteins in a clinically relevant mouse model of chronic hepatitis B virus infection, suggesting that lower level shRNA expression provides an adequate therapeutic window for many clinical applications. Together, the results support the hypothesis that high levels of shRNA expression overwhelm key miRNA processing proteins, leading to cellular toxicity. These findings have important implications for the design of shRNAs for clinical use, emphasizing the need for limited shRNA levels or
1632
Vol. 23,
No. 11, •
CHEMICAL RESEARCH IN TOXICOLOGY
bolstering the expression of proteins such as Ago-2 and Xpo-5. • Carol A. Rouzer
Selective Nanoparticle Toxicity The synthesis, analysis, and application of nanomaterials continues to generate excitement in every discipline of chemistry. One possible use for these tiny particles is as multivalent drug delivery systems. The ability to coat a nanoparticle (NP) with a large number of ligand molecules increases the likelihood of binding and uptake via target receptors on a cell’s surface. Recently, Wang et al. [(2010) J. Am. Chem. Soc. 132, 11306] designed NPs targeting the human transferrin receptor hTfR. High levels of this receptor on rapidly growing and malignant cells suggested that these NPs should be valuable in targeting chemotherapeutic agents to tumors.
Reprinted from Wang et al. [(2010) J. Am. Chem. Soc. 132, 11306]. Copyright 2010 American Chemical Society.
Wang et al. used PRINT technology to produce highly uniform cylindrical NPs of 200 nm height and diameter. Coating the NPs with human transferrin (hTf) or a monoclonal antibody directed against hTfR (OKT9) facilitated specific binding to the hTfR. Negative control NPs were coated with bovine transferrin (bTf) or mouse immunoglobulin (IgG1). NP-hTf and NP-OKT9 were efficiently taken up by six cancer cell lines (HeLa, Ramos, H460, SK-OV-3, HepG2, and LNCaP), which express high levels of hTfR. Moderate uptake was also observed by low hTfR-expressing HEK293 cells. Uptake of control NPs was insignificant in all cell lines. Kinetic studies showed that the rate of uptake of NP-hTf and NP-OKT9 was directly related to the density of TfR on the cell surface, with Ramos B lymphoma cells exhibiting the greatest rate. Uptake of NP-hTf and NP-OKT9 by Ramos cells was blocked by soluble hTf and OKT9, respectively, confirming the specificity of the receptor-ligand interaction. Electron microscopy using fluorescently labeled particles revealed that they were sequestered in acidic endosomes but did not reach the lysosomal compartment. A surprising finding was that NP-hTf and NP-OKT9 caused selective cytotoxicity to Ramos cells. Cell death was ligand density-dependent and was associated with caspase 3/7 activation. Soluble hTf and OKT9 were not toxic to the cells. Supplementation with ferrous ammonium sulfate rescued cells in the case of NP-OKT9 but not NP-hTf. Together, the results confirm the value of hTfR-directed particles in targeting tumor cells, but they also reveal the potential for selective cytotoxicity of multivalent NPs. • Carol A. Rouzer TX100330Z
Published online 11/15/2010 • DOI: 10.1021/tx100330z © 2010 American Chemical Society