Oxidative stress downstream of mTORC1 but not AKT causes a proliferative defect in cancer cells resistant to PI3K inhibition.

Compounds targeting phosphatidylinositol-3-kinase/mammalian target of rapamycin (PI3K/mTOR) signaling are being investigated in multiple clinical settings, but drug resistance may reduce their benefit. Compound rechallenge after drug holidays can overcome such resistance, yet little is known about the impact of drug holidays on cell biochemistry. We found that PI3K inhibitor (PI3Ki)-resistant cells cultured in the absence of PI3Ki developed a proliferative defect, increased oxygen consumption and accumulated reactive oxygen species (ROS), leading to lactate production through hypoxia-inducible factor-1α. This metabolic imbalance was reversed by mammalian target of rapamycin complex 1 (mTORC1) inhibitors. Interestingly, neither AKT nor c-MYC was involved in mediating the metabolic phenotype, despite the latter contributing to resistant cells' proliferation. These data suggest that an AKT-independent PI3K/mTORC1 axis operates in these cells. The excessive ROS hampered cell division, and the metabolic phenotype made resistant cells more sensitive to hydrogen peroxide and nutrient starvation. Thus, the proliferative defect of PI3Ki-resistant cells during drug holidays is caused by defective metabolic adaptation to chronic PI3K/mTOR pathway inhibition. This metabolic imbalance may open the therapeutic window for challenge with metabolic drugs during drug holidays.

Kaposi's sarcoma-associated herpesvirus: from cell biology to pathogenesis.

Kaposi's sarcoma-associated herpesvirus (KSHV or HHV-8) is linked to the etiopathogenesis of Kaposi's sarcoma, a plasma-blastic variant of Castleman's disease and primary effusion lymphoma. KSHV is related to a number of non-human primate viruses. Only a limited number of KSHV proteins are expressed in tumor cells. Here we discuss the putative role of these proteins in KSHV pathogenesis.

Activation of the interferon-inducible (2'-5') oligoadenylate synthetase by the Epstein-Barr virus RNA, EBER-1.

The 2'-5' oligoadenylate synthetases and the protein kinase PKR are both interferon-induced, double-stranded RNA-dependent proteins that play important roles in the antiviral effects of the interferons and in cellular growth control. Both enzymes are activated by natural or synthetic dsRNAs and by single-stranded RNAs that possess extensive secondary structure. This report describes the effects of the small Epstein-Barr virus-encoded RNA EBER-1 on the regulation of 2-5(A) synthetase activity. We demonstrate that EBER-1 RNA binds to and activates the human 40-kDa 2-5(A) synthetase in a dose-dependent manner. The efficiency of EBER-1 as an activator of 2-5(A) synthetase is approximately 25% of that of the synthetic double-stranded RNA poly(I)/poly(C), and poly(I)/poly(C) further stimulates enzyme activity even in the presence of a high concentration of EBER-1. Conversely, EBER-1 neither stimulates nor inhibits 2-5(A) synthetase that has been activated by a high concentration of poly(I)/poly(C). Competitive binding assays suggest that the relative affinity of the enzyme for poly(I)/poly(C) is considerably higher than that for EBER-1. Our data indicate that EBER-1, like VAI RNA of adenovirus, TAR RNA of HIV-1, and Rex-RE RNA of HTLV-1, is able to activate the 2-5(A) synthetases. The significance of why several viruses may activate the 2-5(A) synthetase/RNase L-mediated RNA degradation pathway is discussed.

The vaccinia virus E3L gene product interacts with both the regulatory and the substrate binding regions of PKR: implications for PKR autoregulation.

The vaccinia virus E3L gene product, pE3, is a dsRNA binding protein that prevents activation of the interferon-induced, dsRNA-activated protein kinase, PKR. Activation of PKR, which results in phosphorylation of the translation initiation factor, eIF2alpha, leads to the inhibition of protein synthesis, a process involved in defense against virus infection. The E3L gene product has a conserved dsRNA binding domain (DRBD) in its carboxyl-terminal region and has been shown to function in vitro by sequestration of dsRNA. We have utilized in vitro binding assays and the yeast two-hybrid system to demonstrate direct interactions of pE3 with PKR. By these methods, we demonstrate that pE3 interacts with two distinct regions in PKR, the amino-terminal (amino acids 1-99) located in the regulatory domain and the carboxyl-terminal (amino acids 367-523) located in the catalytic domain. The amino-terminal region of PKR that interacts with pE3 contains a conserved DRBD, suggesting that PKR can form nonfunctional heterodimers with pE3, analogous to those seen with other dsRNA binding proteins. Interaction of pE3 with the amino-terminal region of PKR is enhanced by dsRNA. In contrast, dsRNA reduces the interaction of pE3 with the carboxyl-terminal region of PKR. Competition experiments demonstrate that the carboxyl-terminal region of PKR, to which pE3 binds, overlaps the region with which eIF2alpha and the pseudosubstrate pK3 interact, suggesting that pE3 may also prevent PKR activation by masking the substrate binding domain. Like pE3, the amino-terminal region of PKR also interacts with the carboxyl-terminal domain of PKR. These interactions increase our understanding of the mechanisms by which pE3 downregulates PKR. In addition, the PKR-PKR interactions observed leads us to suggest a novel autoregulatory mechanism for activation of PKR in which dsRNA binding to the DRBD(s) induces a conformational change that results in release of the amino terminal region from the substrate binding domain, allowing access to eIF2alpha and its subsequent phosphorylation.

Homologous regions of the alpha subunit of eukaryotic translational initiation factor 2 (eIF2alpha) and the vaccinia virus K3L gene product interact with the same domain within the dsRNA-activated protein kinase (PKR).

The vaccinia virus K3L gene product, pK3, binds to the dsRNA-activated protein kinase, PKR, reducing its ability to interact with and phosphorylate eIF2alpha. On the basis of this characteristic and the homology of pK3 to the N-terminus of eIF2alpha, several laboratories have utilized pK3 to investigate the molecular determinants that specify substrate recognition by PKR. The data presented here demonstrate that the natural substrate, eIF2alpha, also binds to PKR in vitro and interacts with the same or an overlapping domain within PKR. A truncated form of eIF2alpha, representing the N-terminal 123 amino acids and containing the regions of homology to pK3, retains the ability to bind PKR. pK3, eIF2alpha, and the truncated form of eIF2alpha all bind to the C-terminus of PKR containing the catalytic domain, but not to the regulatory N-terminus. Variants of pK3 and eIF2alpha, des-(75-78)-K3L (pK3deltaGYID), and des-(80-83)-eIF2alpha (eIF2alphadeltaGYID), from which the conserved amino acids GYID have been deleted, exhibit a decreased ability to interact with PKR. Similarly, the in vitro binding of pK3, eIF2alpha, and the truncated form of eIF2alpha to PKR can be competed with purified pK3 but not with pK3deltaGYID. In addition, the deletion of GYID from eIF2alpha significantly reduces its ability to be phosphorylated by PKR, demonstrating that PKR recognizes its substrate, at least in part through interaction with sequences remote from the phosphorylation site. In summary, we have shown that the region within PKR that interacts with the pseudosubstrate, pK3, is the same region that interacts with the authentic substrate, eIF2alpha. In addition, we have shown that the N-terminal 123 amino acids of eIF2alpha contains structural elements necessary for recognition by PKR. The results pinpoint the GYID motif, shared between pK3 and eIF2alpha and distant from the phosphorylation site, as being important for the interaction of eIF2alpha with PKR, as well as its phosphorylation.

Regulation of the interferon-inducible protein kinase PKR and (2'-5')oligo(adenylate) synthetase by a catalytically inactive PKR mutant through competition for double-stranded RNA binding.

The interferon-inducible double-stranded RNA-dependent protein kinase PKR has been suggested to function as a tumour suppressor gene product. Catalytically inactive mutants of PKR give rise to a tumorigenic phenotype when overexpressed in NIH-3T3 fibroblasts and this has been attributed to a dominant negative effect on the activity of the wild-type enzyme. Here we show that the mutant with Lys296 replaced by Arg, [K296R]PKR, not only inhibits the protein kinase activity of wild-type PKR but is also inhibitory towards another double-stranded RNA-dependent enzyme, the 40-kDa form of (2'-5')oligo(adenylate) synthetase. Inhibition of both wild-type PKR and (2'-5')oligo(adenylate) synthetase is reversed by adding higher concentrations of double-stranded RNA. These results suggest competition between [K296R]PKR and wild-type PKR or (2'-5')oligo(adenylate) synthetase for limiting amounts of double-stranded RNA. Moreover, the data imply that the tumorigenic effect of this PKR mutant could be due to inhibition of additional pathways requiring low levels of double-stranded RNA for activation and cannot be unambiguously attributed to inhibition of endogenous PKR itself.

The La antigen inhibits the activation of the interferon-inducible protein kinase PKR by sequestering and unwinding double-stranded RNA.

The La (SS-B) autoimmune antigen is an RNA-binding protein that is present in both nucleus and cytoplasm of eukaryotic cells. The spectrum of RNAs that interact with the La antigen includes species which also bind to the interferon-inducible protein kinase PKR. We have investigated whether the La antigen can regulate the activity of PKR and have observed that both the autophosphorylation of the protein kinase that accompanies its activation by dsRNA and the dsRNA-dependent phosphorylation of the alpha subunit of polypeptide chain initiation factor eIF-2 by PKR are inhibited in the presence of recombinant La antigen. This inhibition is partially relieved at higher concentrations of dsRNA. Once activated by dsRNA the protein kinase activity of PKR is insensitive to the La antigen. We have demonstrated by a filter binding assay that La is a dsRNA binding protein. Furthermore, when recombinant La is incubated with a 900 bp synthetic dsRNA or with naturally occurring reovirus dsRNA it converts these substrates to single-stranded forms. We conclude that the La antigen inhibits the dsRNA-dependent activation of PKR by binding and unwinding dsRNA and that it may therefore play a role in the regulation of this protein kinase in interferon-treated or virus-infected cells.

Regulation of the interferon-inducible eIF-2 alpha protein kinase by small RNAs.

This review describes the structure and function of the double-stranded RNA-dependent protein kinase (PKR) and its interaction with RNA activators and inhibitors. The abilities of small virally-encoded RNAs such as VAI RNA of adenovirus, the Epstein-Barr virus encoded (EBER) RNAs and the Tat-responsive region RNA of HIV-1 to bind to and regulate PKR are reviewed, and the physiological implications of such regulation for the control of viral replication and cell growth are discussed. The potential effects on the activity of PKR of other proteins that bind double-stranded RNA and/or small viral and cellular RNAs are also considered.

Comparative analysis of the regulation of the interferon-inducible protein kinase PKR by Epstein-Barr virus RNAs EBER-1 and EBER-2 and adenovirus VAI RNA.

The interferon-inducible protein kinase PKR interacts with a number of small viral RNA species, including adenovirus VAI RNA and the Epstein-Barr virus-encoded RNA EBER-1. These RNAs bind to PKR and protect protein synthesis from inhibition by double-stranded RNA in the reticulocyte lysate system. Using a peptide phosphorylation assay we show here that EBER-1, like VAI, directly inhibits the activation of purified PKR. A second Epstein-Barr virus RNA, EBER-2, also regulates PKR. EBER-1, EBER-2 and VAI RNA exhibit mutually competitive binding to the native or recombinant enzyme, as assessed by U.V. crosslinking experiments and filter binding assays. The affinities of all three RNAs for PKR in vitro are similar (Kd = ca. 0.3 nM). Since this protein kinase has been proposed to exert a tumour suppressor function in vivo, the ability of EBER-1 to inhibit its activation suggests a role for this small RNA in cell transformation by Epstein-Barr virus.

Reversal of the double-stranded-RNA-induced inhibition of protein synthesis by a catalytically inactive mutant of the protein kinase PKR.

The interferon-inducible double-stranded-RNA(dsRNA)-dependent protein kinase PKR has been implicated in both the antiviral and cell growth-regulatory effects of the interferons. Over-expression of the wild-type form of this protein inhibits cell proliferation, whereas over-expression of inactive mutant forms transforms cells to a tumourigenic phenotype. It has been suggested that mutant PKR exerts a dominant negative effect on the activity of the wild-type protein kinase. We have investigated this possibility using the rabbit reticulocyte cell-free translation system in which protein synthesis is inhibited by dsRNA due to activation of PKR and phosphorylation of initiation factor eIF-2. Addition of a highly purified inactive PKR mutant, synthesised in a baculovirus-infected insect cell system, rescues protein synthesis from inhibition by low concentrations of dsRNA in a dose-dependent manner. The PKR mutant has no effect on protein synthesis in the absence of dsRNA or in the presence of another inhibitory protein kinase, the haem-controlled repressor. Inhibition of translation can be re-established in the presence of the mutant PKR by adding a higher concentration of dsRNA. These results suggest that inactive mutant PKR does exert a dominant negative effect on wild-type PKR and that this may be due to competition for dsRNA binding.