key: cord-0955050-ftx9kwmf
authors: Piazzi, Manuela; Bavelloni, Alberto; Faenza, Irene; Blalock, William
title: Glycogen synthase kinase (GSK)-3 and the double-strand RNA-dependent kinase, PKR: When two kinases for the common good turn bad()
date: 2020-06-05
journal: Biochim Biophys Acta Mol Cell Res
DOI: 10.1016/j.bbamcr.2020.118769
sha: fe627e8022097b69a48140e799516064b4390c57
doc_id: 955050
cord_uid: ftx9kwmf

Glycogen synthase kinase (GSK)-3α/β and the double-stranded RNA-dependent kinase PKR are two sentinel kinases that carry-out multiple similar yet distinct functions in both the cytosol and the nucleus. While these kinases belong to separate signal transduction cascades, they demonstrate an uncanny propensity to regulate many of the same proteins either through direct phosphorylation or by altering transcription/translation, including: c-MYC, NF-κB, p53 and TAU, as well as each another. A significant number of studies centered on the GSK3 kinases have led to the identification of the GSK3 interactome and a number of substrates, which link GSK3 activity to metabolic control, translation, RNA splicing, ribosome biogenesis, cellular division, DNA repair and stress/inflammatory signaling. Interestingly, many of these same pathways and processes are controlled by PKR, but unlike the GSK3 kinases, a clear picture of proteins interacting with PKR and a complete listing of its substrates is still missing. In this review, we take a detailed look at what is known about the PKR and GSK3 kinases, how these kinases interact to influence common cellular processes (innate immunity, alternative splicing, translation, glucose metabolism) and how aberrant activation of these kinases leads to diseases such as Alzheimer's disease (AD), diabetes mellitus (DM) and cancer.

J o u r n a l P r e -p r o o f redundant network [21] [22] [23] [24] [25] [26] [27] [28] [29] . Interestingly, while all of these kinases are present in the cytoplasm, PKR is the only eIF2α kinase that is also present in the nucleolus and nucleoplasm [25, 30] . In humans, the PKR kinase is encoded by the EIF2AK2 gene located on Ch.2p22.2. The gene encodes a 68 kDa protein which is unusual in that the amino terminal end or regulatory region contains two double-strand RNA binding domains (dsRBDs) while the carboxyl terminus contains a protein kinase domain. The EIF2AK2 gene is ubiquitously and constitutively expressed with the highest levels of protein expression being observed in hematopoietic tissue (bone marrow, spleen and thymus) and the brain ( [31] [32] [33] ). Moreover, the promoter of EIF2AK2 contains interferon (IFN)stimulated response elements (ISREs), allowing for enhanced transcription of the gene in cells exposed to Type I (IFNα/β) interferons [34] .

The regulation of PKR kinase activity requires the phosphorylation of two primary threonine residues in the catalytic domain. Minimal activation of PKR requires the phosphorylation of T451, while subsequent autophosphorylation of T446 significantly enhances kinase activity. Additional sites of autophosphorylation are observed predominantly in two clusters: S83, T88, T89, T90 and Y101 between dsRBD I and dsRBD II, and Y162, S242, T255, T258 and Y293 situated between dsRBD II and the kinase domain; each of these enhancing PKR enzymatic activity [35, 36] . Other than the afore mentioned sites, which have all been biochemically verified, additional sites of phosphorylation have been identified in multiple studies using mass spectrometry: S33, S92, T115, S167, S179, S181, S456 and S542. The exact consequence of phosphorylating these residues though is not known (http://www.phosphosite.org/uniprotAccAction?id=P19525; [12] ). PKR is also highly ubiquitinated and sumoylated with a large number of sites within the carboxyl half of the protein. Ubiquitination is predominantly carried-out by the SCF E2 ubiquitinase FBXWII E3 ligase, which targets PKR for proteosomal degradation [37] . In contrast, SUMO1 and SUMO3 sumoylate PKR on K60, K150 and K440 in an enzyme specific manner altering PKR activation, localization and stability [38] .

Several mechanisms have been proposed to explain PKR activation. Initially, PKR activation was thought to require only dsRNA, a typical product of viral infection. The binding of PKR to dsRNA through the dsRBDs would facilitate PKR homodimerization and autophosphorylation of T451 followed by T446. Several lines of evidence have suggested that this model was not completely correct:; (1) the endogenous PKR activator, PACT/RAX, was demonstrated to promote PKR activation in the absence of dsRNA in in vitro studies [39] ; (2) PKR activation following vesicular stomatitis virus (VSV) infection was inhibited in the absence of PACT [40] ; (3) T451 phosphorylation is often induced following treatment of cells with the J o u r n a l P r e -p r o o f 8 commercial PKR inhibitor [41] ; (4) dsRNAs readily available in the cell do not activate PKR, in contrast, the cellular RNA non-coding 886 (nc886) binds to and inhibits PKR activation; and (5) a diverse number of miRNAs bind to the dsRBDs of PKR ( [42] ). These findings also suggested that PKR was not the only kinase capable of phosphorylating T451. Zykova et al. demonstrated that T451 could be phosphorylated by ERK2 and RSK2, likely establishing them as primers of PKR activation [43] . This would also explain PKR activation following Toll-like receptor (TLR) stimulation. During apoptosis, PKR may also be activated through caspase-dependent cleavage at D251, thus removing the regulatory dsRBDs and releasing an active kinase domain [44] . For years, PKR was studied for its ability to phosphorylate the eukaryotic initiation factor (eIF)-2α subunit in the presence of dsRNA (either during viral infection or treatment with poly I:C, a synthetic dsRNA) and was thus analyzed for its ability to block viral replication and/or induce cell death following infection [19] . It is now known that PKR has a much larger role in cell growth and homeostasis than previously thought.

Several seminal studies over the years have begun to shed more light on just how entwined PKR is in normal cellular homeostasis. The simultaneous reporting of PACT and its mouse orthologue RAX as endogenous activators of PKR demonstrated that PKR regulation was more complicated than just the presence or absence of dsRNA [39, 45] . The fact that multiple cellular stresses [DNA damage, reactive oxygen species (ROS), and cytotoxic cytokines (IFNγ, TNFα, IL1)] in addition to viral infection could result in PKR activation began to paint a picture of PKR as a sentinel for the cellular stress response. PKR was found to regulate p53, NF-κB, c-MYC, protein phosphatase 2A (PP2A), RNAase A (DHX9), GSK3 and be a component of several cytoplasmic (inflammasome and stress bodies) and nuclear complexes (splicesome, ribosome assembly and DNA repair) [46] . The role of PKR in disease has been controversial, but it is known that PKR is overexpressed and/or activity elevated in multiple solid tumors (breast carcinoma, colon carcinoma, melanoma, osteosarcoma, etc.), hematopoietic malignancies (acute leukemias and myelodysplastic syndromes) and neurodegenerative diseases (Alzheimer's disease, Huntington's corea and Creutzfeldt-Jakob disease) [25, 47, 48] . In most of these pathologies an increased nuclear presence of PKR has been observed, but while nuclear PKR is known to be associated with splicing factors and components of the assembling ribosome, the significance of PKR nuclear localization in these pathologies remains unclear.

J o u r n a l P r e -p r o o f Regulation of GSK3 by PKR appears to occur at multiple levels. As stated previously, activation of the PI3K-AKT-mTOR pathway leads to AKT-dependent phosphorylation of GSK3α/β on S21/S9 resulting in the inactivation of GSK3 kinase activity. The link between PKR activation and AKT was originally observed in PKR-null MEFs where AKT activation was altered [49] .

Baltzis et al. then later reported that both PKR and PERK could lead to the activation of nuclear GSK3β in a manner independent of eIF2α phosphorylation ( [50] ). Blalock et al. subsequently demonstrated that PKR could regulate GSK3α/β through the activation of a phosphatase, but this study stopped short of identifying the phosphatase [41] . The authors demonstrated that inhibition PKR with a small molecule inhibitor resulted in an increase in p-S21/S9 GSK3α/β while at the same time resulting in a slight mobility shift to a faster migrating form of GSK3α/β. In contrast, while use of the general phosphatase inhibitor okadaic acid (OA) resulted in an increase of p-S21/S9 GSK3α/β, it did not result in any mobility shift, suggesting a site specific regulation of GSK3 phosphorylation by PKR [41] . A more recent analysis of the PKR nuclear interactome brought forward the finding that the phosphatase PP1A is associated with active PKR in the nucleus, but the PP1A catalytic subunit is replaced with that of PP1B in the inactive PKR complex [46] . As dephosphorylation of S21/S9 is known to occur through PP1A, it might be presumed that PKRdependent regulation of GSK3 activity is mediated through PP1A dephosphorylation of S21/S9, which would account for the increase in levels of p-S21/S9 GSK3α/β in the presence of the PKR inhibitor. In the presence of active PKR, PP1A-mediated dephosphorylation at S21/S9 would increase GSK3 activity, thus enhancing Y279/Y216 phosphorylation, increasing GSK3 activity further ( Figure 1A ). As PKR also stimulates the p38 MAPK , PKR activity may favor additional phosphorylations at S389 and T390 in GSK3β, thus accounting for the significant shift or collapse in bands that occurs in the presence of the PKR inhibitor [18, 41] . To date, no clear mechanism for GSK3α/β to regulate PKR activity has been reported, although control at the level of PKR expression would be reasonable to assume taking into consideration the effects that GSK3α/β has on transcription factors, such as MYC and p53, and translation (see below).

The extent of interaction between GSK3α/β and PKR signaling goes beyond that of phosphatase regulation; these kinases influence many of the same downstream intermediates. In some cases, these intermediates are directly phosphorylated by both PKR and GSK3; in others, regulation of the downstream protein occurs through phosphorylation by one of the kinases and an alternate mechanism with the other. In the following sections, we discuss some of the common J o u r n a l P r e -p r o o f targets of GSK3 and PKR and the influence that these kinases have on these downstream targets (Table 1 ).

Protein synthesis is the most energy consuming process of the cell, requiring both ribosome biogenesis and the subsequent translation of mRNA. From prokaryotes to mammals, these processes are highly regulated in response to the surrounding environment to limit energy expenditure under conditions that are unfavorable for growth, as well as limit the possibility of producing mutant proteins. The most efficient and rapid point to regulate protein synthesis under cell stressing conditions is during the initiation phase of translation, thereby limiting the amount of cellular energy and resources that are fruitlessly utilized. This point also allows for the rapid synthesis of proteins once the stress is alleviated. Many upstream kinases converge on translation initiation, which is the rate limiting step. Recruitment of the mRNA to the 40S ribosomal subunit represents one level of control and is carried-out by the eIF4F complex, which requires the eIF4E subunit to directly bind the 5' CAP of the mRNA [51] . Under resting conditions, eIF4E is bound by the eIF4E-binding protein, 4E-BP, which sequesters it and inhibits its association with the mRNA.

Activation of the PI3K-AKT-mTOR axis promotes metabolism and protein synthesis through mTORC2-mediated phosphorylation of 4E-BP and other substrates [51] . As previously stated, active AKT phosphorylates GSK3α/β on S21/S9, inhibiting GSK3 catalytic activity. Under stress conditions where reduced flux through PI3K-AKT-mTOR pathway no longer leads to mTORmediated phosphorylation 4E-BP, active GSK3β is able to maintain CAP-dependent translation initiation via the eIF4F complex translation initiation but alter general translation ( Figure 1A and B, a) [51] .

In addition to the eIF4F translation initiation complex, two additional initiation factors represent major points of translation control, eIF2 and eIF2B ( Figure 1B ). These complexes bare both the GTP and the initiator Met-tRNA necessary for the formation of the pre-initiation complex (PIC) and translation initiation, as well as the proteins for the GDP to GTP exchange required to initiate the next round of translation ( Figure 1B,b) [52] . Regulation of the eIF2-GDP/GTP exchange is via phosphorylation of the α-subunit (eIF2α) by one of four different kinases (PKR, PERK, GCN2 or HRI) and is probably the best understood mechanism regulating translation initiation. Phosphorylation of eIF2α on S51 results in eIF2 being locked in the GDP bound state with the GDP/GTP exchange complex ,eIF2B, unable to catalyze the initiation of protein synthesis.

As the eIF2 complex is limited compared to eIF2B, it does not take much phosphorylated eIF2α to soon result in a complete block of general translation ( Figure 1B ,c) [53] .

Phosphorylation of eIF2α, which was originally thought to be strictly pro-apoptotic, actually results in a shut-down of general CAP-dependent translation while at the same time, allowing for efficient translation of upstream open reading frames (uORFs) in mRNAs that contain complex secondary structure at the 5' end and an internal ribosome entry site (IRES) element upstream [54] [55] [56] . Short-term inhibition of general translation through eIF2 phosphorylation establishes a prosurvival state by allowing for cellular repair and time for the cell to adjust following a particular stress [57] ; although, if this stress cannot be resolved and general translation remains inhibited the cell will likely die through apoptotic means. In contrast, under other conditions phosphorylation of eIF2α has been shown to inhibit IRES-mediated translation [58] . These differences may be due in part to specific regulator proteins that differ between IRES elements as well as additional elements inherent to the mRNA ( Figure 1B , c). Many of the mRNAs translated under conditions where eIF2 is phosphorylated encode inflammatory cytokines such as TNF, IL-1, FGF, VEGF, IL-6 and ATF4 [59] [60] [61] [62] [63] [64] .

In contrast to PKR which targets eIF2, GSK3α/β targets eIF2B. The function of eIF2B is to exchange GTP for the eIF2 complex bound GDP [65] . The eIF2B GTP exchange factor is composed of , , ,  and  subunits, of which the 82 kDa ε-subunit is the most critical to eIF2B regulation and is enzymatically responsible for the GDP/GTP exchange. The α-, β-, and δ-subunits associate with Ser51 phosphorylated eIF2α and inhibit eIF2Bε activity, locking eIF2B with eIF2 in the GDP-bound state. The γ-subunit, which is phosphorylated and regulated by casein kinase (CK)-II, promotes the activity of the ε-subunit. Phosphorylation of eIF2B on Ser540 by GSK-3 following amino acid starvation, inhibits eIF2B activity [25, 65, 66] . As GSK3 requires a priming phosphorylation, eIF2Bε must first be phosphorylated on Ser544 by one of the DYRK family kinases ( Figure 1B , d) [67] . The phosphorylation of these residues results in translation of a set of mRNAs that is different from those translated when only eIF2α is phosphorylated; thus regulation of translation can give rise to proteins that are most efficiently translated under one of three (or more) different conditions ( Figure 1A and B) [25, 66, 68] . Additional information on alternate translation can be found in the following reviews ([59, 69-71].

J o u r n a l P r e -p r o o f Glycogen synthase kinase-3 and PKR have been shown to influence the cell cycle both positively and negatively depending on the context. Among the cell cycle targets that have been demonstrated to be directly phosphorylated by the GSK3 kinases are cyclins D and E (GSK3α and GSK3β) and the dual cyclin phosphatase, CDC25A (GSK3β); each of these phosphorylations favoring ubiquitination and degradation of the target protein, thus resulting in cell cycle arrest [72] .

In contrast, while PKR was originally thought of as a tumor suppressor, it was observed that PKRnull mouse embryo fibroblasts (MEFs) grew slower than normal MEFs and expression of dominantnegative PKR (K296R) in the human glioblastoma cell line T98G resulted in a longer G1 phase, suggesting some role for PKR activity in normal cell proliferation, but direct targets were missing [73] . As will be discussed later both GSK3 and PKR influence the activity of diverse transcription factors which promote transcription of both pro-growth and cell cycle inhibiting factors. Among the common targets of GSK3 and PKR is the MDM2-p53 regulatory complex.

Most studies examining GSK3 and p53 have centered on GSK3β, likely a result of the fact that it is this GSK3 isoform that is prominent in the nucleus. Both positive and negative p53 regulation have been associated with GSK3β, and these discrepancies may be related to the cellular localization of the GSK3β in question. In ovarian cancer, GSK3β represses let-7 miRNA through activating p53 [74] . It is interesting that active GSK3β would target the repression of let-7, a miRNA that targets the AKT kinase family members among other proteins. This might be suggestive of a possible feedback regulation: low AKT leads to increased active GSK3β which subsequently stimulates p53 and the repression of let-7, allowing for increased expression of AKT.

Enhancement of p53 through GSK3β activation has been shown to occur through acetylation of p53 on K120, resulting in stabilization of p53 [75] . The reported mechanism was dependent on GSK3dependent phosphorylation of the histone acetylase Tip60 on S86 and increased acetylase activity toward p53. In addition, GSK3β-dependent phosphorylation of S376 leads to enhanced p53 activity ( Figure 2A , a) [76] . In contrast, GSK3β has also been shown to result in p53 degradation. The E3 ligase MDM2 (HDM2) is responsible for the ubiquitination of p53, resulting in its subsequent degradation by the proteosome. Phosphorylation of S240 and S254 in murine cell lines was reported to lead to MDM2 stabilization and p53 degradation, a process that was instrumental for promoting apoptosis following ionizing radiation exposure [77] . Loss of GSK3β or use of an isoform specific inhibitor was shown to result in p53 accumulation and protection from radiationinduced damage; thus loss of GSK3 activity is protective against radiation-induced DNA damage [77, 78] . Under physiological conditions it is usually AKT activation that results in the inhibition of GSK3 following ionizing radiation exposure (Figure 2A Originally, activation of PKR was shown to be associated with p53 expression and transcriptional transactivation of p53 associated genes [50, 79] . Studies in a PKR-null background also demonstrated that S18 phosphorylation was diminished [79] . Later studies by Cuddihy et al. indicated that PKR could directly phosphorylate p53 on S392 resulting in its stability [80] . Most recently, Bennett et al. reported that activation of the PACT/PKR pathway resulted in the association of the SUMO E2 ligase Ubc9 with p53 and the subsequent sumoylation of lysine 386 [81] . Sumoylation at this site appeared to be a prerequisite for S392 phosphorylation, as this site was not phosphorylated in the sumoylation deficient p53 mutant (K386R) ( Figure 2B from the nucleus, followed by its degradation in the cytosol. This effect was not dependent on the phosphorylation of eIF2α, suggesting a mechanism independent of translational regulation and believed to involve the phosphorylation of Y216 ( Figure 2B , b) [50] . This may indeed be the phosphatase-dependent effect that was observed in acute leukemia cells [41] . A critical factor that remains to be determined is what dictates whether these kinases promote or oppose p53 activation; this may center on the status AKT activity. PKR activation in the presence of AKT activity may allow PKR to enhance p53 stability and transactivation, while PKR activation in the absence of AKT activity may promote GSK3-mediated p53 decay ( Figure 2C ). The potential of this interplay and the influence that cellular localization of the associated protein complexes plays remains to be well defined.

The microtubule-associated protein TAU, the product of mapt gene which is expressed almost exclusively in neurons, serves the purpose of facilitating microtubule assembly and links axonal microtubule components to those in the plasma membrane, thus defining the polarity of the neuron [83] . At least nine different isoforms of TAU, which are produced by alternative splicing, have been reported, with some isoforms restricted to development [84] . The largest isoform, PNS-J o u r n a l P r e -p r o o f 14 TAU, is expressed in the peripheral nervous system, whereas the other isoforms are found in the central nervous system. TAU is a highly modified protein that is a target of at least twelve different kinases, including GSK3α/β and PKR [32, [83] [84] [85] [86] . Analysis of the Phosphosite Plus database indicates that while GSK3α has been shown to phosphorylate a number of TAU isoforms (isoforms 2, 5, 6 and 8) both GSK3β and PKR have been demonstrated to phosphorylate namely isoform 8 (Table 2 ) [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] . With the exception of two sites phosphorylated by GSK3β, the majority of the sites phosphorylated by the GSK3 and PKR kinases lie near to or within the C-ter tubulin binding repeat (aa 560-690). In general, phosphorylation of these residues results in the detachment of TAU from the microtubule complex and disassembly. Normal neuronal development will see the interplay between O-GlcNacylation, phosphorylation and dephosphorylation by phosphatases. As phosphorylation and O-GlcNacylation are mutually exclusive in this regulation, the interplay between these mechanisms in simple terms either leads to O-GlcNacylation and microtubule assembly or phosphorylation and microtubule disassembly. In multiple degenerative diseases such as Alzheimer's and Parkinson's a loss of O-GlcNacylation with an associated hyperphosphorylation is observed with TAU [83, 84] . These hyperphosphorylated TAU molecules become seeds for the development of neural filamentous tangles which lead to progressive neuronal cell death and disease. As will be discussed later, chronic inflammatory signaling leading to the constitutive elevated activation of both GSK3β and PKR has been demonstrated to be intricately associated with neurodegeneration.

The MYC transcription factor is a critical proto-oncogene for cellular growth and differentiation. In addition, to being indispensible for the first steps in ribosome biogenesis, the synthesis of ribosomal RNAs, and protein synthesis, MYC is also critical to cellular differentiation and the regulation of apoptosis [102] . While MYC is influenced by the PI3K-AKT-mTOR pathway through multiple mechanisms, it is also subject to direct control by GSK3 through phosphorylation.

Phosphorylation of MYC on T58, by GSK3α/β, and on S62, by GSK3α/β and other kinases (DYRK2, CDK2 and CDK5) enhances the transcriptional transactivating activity of MYC but at the same time functions to prime a negative regulatory mechanism [103] [104] [105] . Phosphorylation of T58 in particular allows for the Fbw7 F-box proteins of the SCF-type ubiquitin ligase to associate with MYC. Association of MYC with Fbw7α and the ubiquitin protease Ubc28 in the nucleus favor MYC stability and transcriptional activity. In contrast, in the nucleolus, MYC phosphorylated on T58 associates with Fbw7γ and is degraded, thus providing a control mechanism over MYC-J o u r n a l P r e -p r o o f induced transcription, which is dependent on Fbw7 isoform expression [104] . As is observed with TAU, the interplay between phosphatase-dependent dephosphorylation (T58 and S62), O-GlcNacylation (T58) and phosphorylation (T58 and S62) plays a major role in MYC activity ( Figure 3A ) [103, 104, 106, 107] .

Similarly, PKR also regulates the expression of MYC, but not through direct phosphorylation. Diverse groups have demonstrated that PKR influences the expression of c-myc through the stimulation of nuclear factor (NF)-κB and signal transducer and activator of transcription (STAT) transcription factors ( Figure 3B , a) [49, 108] . Previous work examining the effects inhibiting PKR kinase activity in acute leukemia cells with a small molecule inhibitor demonstrated that loss of PKR activity could both enhanced MYC expression and influence the isoform of MYC expressed. Overexpression of PKR was shown to favor p64 MYC expression while siRNA-mediated knock-down of PKR favored p67 MYC expression ( Figure 3B , b and c) [46] . This is an interesting point as the p64 isoform of MYC is initiated from a standard AUG start codon; while in contrast, the p67 isoform is produced from a non-canonical CUG start codon and encodes an additional 15 amino acids at the amino terminus. Both MYC isoforms, p64 and p67, target E-box sites in MYC responsive promoters, but only p67 can also target C/EBP elements, thus leading to the transcription of an additional set of responsive genes ( Figure 3B , c) [46, 109] . It has been suggested that the ratio p64/p67 dictates whether MYC expression favors growth and proliferation or stimulates the expression of pro-apoptotic factors, with p64 favoring proliferation and p67 favoring growth arrest [46, 109] . As the difference is these isoforms appears to be an effect of altered translation initiation, it might be assumed that PKR activity can influence additional aspects of translation initiation beyond that of simply eIF2α phosphorylation. Moreover, this might also suggest that the loss of PKR may stimulate a feedback control through MYC isoform expression to limit growth and proliferation of cells that do not have the necessary safeguards in place to monitor translation initiation.

Nuclear factor-B (NF-κB) consists of hetero-or homodimers derived from five different subunits: p65/RelA, p50 (processed from p105), p52 (processed from p100), RelB and c-Rel. The genes preferentially induced differ according to the hetero-or homodimer present. The NF-B family regulates the expression of numerous genes involved in cell cycle progression (cyclin D1 and c-MYC), apoptosis inhibition (c-IAP1/2, BCL-X L and A20) oxidation-reduction regulation and inflammation (COX2, GADD45β, HO-1, IL-6. iNOS, MMP-9, MnSOD and RANTES) [110] .

Under most situations the regulation of NF-B relies on subunit binding to the inhibitory proteins IB- and -. These proteins maintain NF-B localized to the cytoplasm. Following stimulation, IB proteins are phosphorylated by the IκB kinase or IKK complex. This phosphorylation leads to dissociation between IB and NF-B, the targeting of IB to the proteosome for degradation, and translocation of the NF-B dimer to the nucleus ( Figure 4 ) [110] . In addition to this level of regulation, the NF-κB proteins can also be post-translationally modified bringing about changes in their activity and stability.

NF-B homo-and heterodimer activation by PI3K-AKT-mTOR signaling occurs on many levels, including AKT-dependent modification of IKK on Thr23, which, rather than stimulating IB degradation, seems to stimulate the transactivating domain of p65/RelA [111, 112] . The effect of AKT on IKK also regulates the processing of NF-B p100 to p52 [113, 114] . Downstream targets of AKT can also influence NF-B activity ( Figure 4 ). The mTOR complex can stimulate IKK activity; IKK in-turn has been demonstrated to feedback on mTOR activity through the phosphorylation of the mTOR inhibitor proteins TSC1/2 [115] [116] [117] [118] [119] . Additionally, AKT also has the ability to inhibit the formation and activation of certain NF-κB transcriptional complexes through its ability to modify GSK3/. As mentioned previously, GSK3-α and -β are often activated in response to inflammatory/stress signaling. GSK3 activity has not only been shown to be required for processing of p105 NF-B, but also for NF-B transcriptional activity in the nucleus; the main reason GSK3-/-mice are embryonic lethal and have a phenotype similar to IKK-/-mice ( Figure 4 ) [120] [121] [122] [123] [124] [125] [126] [127] [128] . This regulation again appears to occur at two levels, phosphorylation of the NEMO subunit of the IKK complex and phosphorylation of p65 or the NF-κB precursor proteins p100 and p105. Phosphorylation of S8, S17, S31 and S43 of NEMO results in protein stabilization favoring kinase activity and downstream activation of NF-κB [129] . However, GSK3β-dependent phosphorylations of NF-κB p65, p100 and p105 have very differing effects on the transcriptional outcome. GSK3β-dependent phosphorylation of p65/RelA on S468 primarily occurs in the nucleus and alters the transcriptional activity, favoring the expression of some responsive genes while suppressing others [130] . In contrast, phosphorylation of NF-κBp100 on S222 suppresses homodimerization of p52 favoring p52/c-REL heterodimers, while phosphorylation of S707 and 711 leads to p100 degradation by the Fbxw7α E3 ubiquitin ligase, librating non-canonical NF-κBmediated transcription [131, 132] . Phosphorylation of NF-κBp105 on S903 and S907 inhibits the proteolysis of p105 to p50 and primes p105 for degradation, thus reducing p50-mediated transcription ( Figure 4) [120, 133] . Although data involving the activation status of NF-B in J o u r n a l P r e -p r o o f GSK3-/-mice have not been reported nor has GSK3α-dependent phosphorylation of NF-κB transcription factors, some reports have suggested GSK3 activity may also play a role in NF-B nuclear functions, with its loss having severe consequences on survival and proliferation ( Figure 4) [134].

As mentioned above, PKR activation often leads to increased nuclear GSK3 activity.

Although the outcome of PKR-mediated modification of GSK3 on NF-B associated transcription has not been determined, it is interesting to note that many of the same NF-B transcribed genes that are repressed in PKR-null fibroblasts or by overexpression of catalytically-dead PKR are also inhibited in GSK3-null fibroblasts and cells overexpressing kinase defective GSK3 mutants ( Figure 4 and Table 3 ) [57, 121, 124, 127, [135] [136] [137] [138] .

The role of PKR in regulating NF-B-dependent transcription was first suggested when it was discovered that dsRNA and IFN activate NF-B DNA binding [139] [140] [141] . Later, studies reported that expression of DN-PKR blocked dsRNA and IFN activation of NF- [142, 143] .

Conclusive evidence was finally demonstrated by the inability of TNF to stimulate NF-B translocation and DNA binding activity in PKR-/-cells [144, 145] . Exactly how PKR induces the activation of NF-B is somewhat still debated. Originally, PKR was thought to directly phosphorylate inhibitor B (IB)- resulting in its release of NF-B and subsequent proteosomal degradation [142, 146] . This idea was later challenged by the discovery that PKR associated with the IKK complex in order to activate NF-B transcription [147, 148] . Other data challenged whether PKR catalytic activity was even need at all, as truncated forms of PKR consisting of the amino terminus were shown to associate with the IKK complex and stimulate IB phosphorylation [149, 150] . Similarly, kinase-deficient mutants of PKR were reported to associate with the IKK complex and induce NF-B; while other reports suggested that kinase activity was required [151, 152] . Much of this discrepancy was solved when Donze et al. published a seminal paper demonstrating that PKR, irregardless of catalytic activity, could induce the transcriptional activation of NF-B and the synthesis of some NF-B-dependent transcripts, but the transcriptional activity of NF-B and the transcription of additional NF--dependent genes was greatly potentiated in the presence of active PKR kinase [57] . This suggested that both PKR association with the IKK complex and PKR catalytic activity are important for PKR-mediated effects on NF-B. It should be stated at this point that PKR and its associated catalytic activity mainly favor p65 NF-κB-containing transcription complexes. To this end the current understanding is that PKR activity is required for J o u r n a l P r e -p r o o f the full effects of PKR on NF-B-mediated transcription, by both influencing IB expression and phosphorylation/release as well as later points, such as enhancing the nuclear activity of GSK-3β ( Figure 4 ) [50] (166) . A list of NF-B-induced genes/proteins dependent on PKR and GSK3β is presented in Table 3 .

Similar to MYC, both GSK3α/β and PKR influence CCAAT-enhancer binding transcription factor proteins, C/EBPα and C/EBPβ through phosphorylation and alternative translation, respectively. These transcription factors are essential for the differentiation of hepatocytes, adipocytes, and myeloid progenitors as well as tissues of the lung and placenta [153, 154] . Through association with IRE elements, C/EBP transcription factors can regulate gluconeogenesis and lipogenesis programs in the liver to balance metabolic energy demand and storage. Of significant importance is the activation of these transcription factors under stress and inflammation, resulting in cellular differentiation and energy homeostasis. C/EBP is expressed as either a full-length (p42) or truncated (p30) protein [153] . Although both forms contain sites for interaction with additional transcription factors, only p42 contains the transactivation domain [153] . In the hematopoietic compartment, C/EBP is required for normal granulocyte differentiation and C/EBP -/-mice have a loss of myeloid maturation accompanied by blast accumulation [153, 155] . Additionally, C/EBP is mutated or repressed in many other hematologic malignancies [153, 156, 157] . Interestingly, expression of p30 C/EBP increases the expression and activity of the Ubc9 ubiquitin ligase toward p42 C/EBP resulting in p42 degradation; thus blocking differentiation in CD34+ hematopoietic stem cells, favoring self-renewal [25, 158] .

In many ways similar to C/EBPC/EBP is expressed as one of three forms: p38, p33 and p20, and shares many overlapping functions with C/EBPα; but unlike C/EBPα, it promotes the proliferation of a number of cell types, with the noted exception of T-lymphocytes whose proliferation is blocked by C/EBPβ-mediated MYC repression [153, 159] . The full-length form is most associated with regulating the expression of genes involved in the immunologic response and inflammation while the p33 isoform is associated with gene transcription in activated macrophages, the CD4+ T-cell response, and diverse aspects of intracellular innate immunity. The shortest form, p20, functions as a dominant-negative factor when bound to p33 and has been best associated with osteoblast differentiation and osteoclastogenesis. Interestingly, whereas the loss of C/EBP transcriptional activity is associated with tumorigenesis, loss of C/EBP is not [25, 153, 160, 161] .

In humans, the GSK3 kinases phosphorylate T226, T230 and S234 in C/EBPα and T226 and S231 in C/EBPβ, which is proposed to enhance their transcriptional activity favoring among other things adipogenesis and hematopoietic differentiation [162, 163] . In contrast, stimulation of the PI3K-AKT-mTOR pathway following stimulation of insulin-related growth factor receptor (IGF-1R), or other growth factor receptors, leads to GSK3 inhibition through AKT-dependent phosphorylation of S21/S9 and the suppression of differentiation programming with a concatenate increase in glucose uptake/usage and proliferation. Zeng et al. reported that C/EBPα activation can contrast this activity through the induction of miR-122 expression, which represses IGF-1R expression [164] . In contrast to the control mechanisms mediated by GSK3α/β, phosphorylation of eIF2 by PKR in response to inflammatory/stress signaling favors the translation of full-length C/EBP and  [154] .

In the past several years proteomic studies examining the interactome and phosphoproteome have begun to associate a diverse number of kinases to cellular processes to which they were previously not associated; this is no truer than for GSK3α/β and PKR. As stated above, both PKR and GSK3α/β are considered major regulators of ribosome biogenesis and translation by directly or indirectly influencing these processes at multiple levels (for more detail information see Ref. [51] ).

More recently a role for both these kinases in alternative splicing has been documented. Blalock et al. previously reported that multiple proteins involved in RNA processing and alternative splicing were associated with PKR in the nucleus [46] . The study examined which proteins were complexed with nuclear PKR when active and when inactive (in the presence of the PKR inhibitor). Proteins associated with RNA processing/splicing that interacted only with the active PKR complex included the EIF4A3 (or DDX48), HNRNPM, SNRPA1, SON, SRSF2, THOC2, U2AF1, U2AF2, and YBX1. Those associated with the inactive PKR complex included COIL, SNRPD2, SNRPD3, SRRM1 and TRA2B (Table 4 ). Interestingly, while several RNA binding proteins were found to be associated with both active and inactive PKR complexes, nucleophosmin (NPM1) and the adenosine deaminase acting on dsRNA (ADAR)-1 were the only proteins with a known role in splicing/alternative splicing that were identified in both active and inactive PKR complexes [46] . This suggested that PKR activity could significantly affect assembly and composition of the splicing complex, thereby changing the alternatively spliced RNA landscape ( Figure 5) . A shortcoming of this study was the lack of identifying direct targets of PKR phosphorylation in these complexes. More recently, Shinde et al. reported in a phosphoproteome study examining J o u r n a l P r e -p r o o f 20 GSK3α/β-dependent signaling in wild-type and gsk3α/β double-knockout mouse embryonic stem cells (ESCs) that GSK3α/β regulates the phosphorylation of a number of proteins involved in RNA splicing, going on to both define the sites of GSK3α/β-dependent phosphorylation as well as a number of transcripts that are alternatively spliced in the presence of GSK3α/β activity [165] . The proteins whose phosphorylation was reduced in the gsk3α/β double-knockout ESCs and the sites of phosphorylation that were identified are as follows: BCLAF1, CELF1, NPM1, PPP4R2, RBM8A (S166 and/or S168), RBM39 (S129), SRRM1, SRRM2 (S1864), TRA2B (S39 and S83 and/or S85 and/or S87), SRSF9 (S190), and WBP11. While this data does not necessarily indicate direct phosphorylation of all these proteins by GSK3α/β, it does indicate that phosphorylation of these proteins is regulated by GSK3α/β. Additionally, GSK3α and GSK3β have also been demonstrated to interact with U2AF1 (α and β), U2AF2 (α only) and YBX1 (α and β), but phosphorylation of these proteins by GSK3α/β has not been observed ( Figure 5 and Table 4 ) [165] .

Interestingly, activity of both PKR and GSK3α/β has been associated with hyperphosphorylated TAU, neural filamentous tangles and Alzheimer's disease. In addition, the inclusion or exclusion of TAU exon 10 through alternate splicing results in TAU isoforms containing either 3 or 4 microtubule repeat domains (3R or 4R). In TAU-related pathologies, or TAUopathies, the 3R/4R ratio is disturbed with an increase in the 3R:4R ratio [166] . Of the proteins mentioned above that have a role in TAU exon 10 splicing, U2AF2, which represses exon 10 inclusion, interacts with GSK3α and was found in association with active nuclear PKR [46, 165] .

In contrast, TRA2B which promotes the inclusion of TAU exon 10, as mention above, is phosphorylated in response to active GSK3α/β and was found to interact with inactive PKR in the nucleus (Table 4 ) [46, 165] . Also associated with inactive nuclear PKR was the serine/arginine-rich splicing factor (SRSF)-6, which also represses the inclusion of TAU exon 10 [46] . Previous studies have demonstrated that GSK3β could phosphorylate SC35, a serine/arginine splicing factor that favors TAU exon 10 inclusion [166] . The GSK3β phosphorylated SC35 is redistributed in the nucleus and TAU exon 10 inclusion is repressed. These data seem to support a combined role for PKR and GSKα/β in the regulation of alternate splicing of TAU transcripts and the production of isoforms that have been linked to neurodegenerative pathologies.

Recent data also suggest that mRNA transcripts of both kinases undergo alternative splicing.

In several organisms, GSK3β has been found to be expressed in multiple alternatively spliced forms [167, 168] . In humans, two alternatively spliced forms have been observed. The primary form consists of 420 amino acids and has a predicted molecular weight of 46.7 kDa, while the alternative spliced form contains an additional 13 amino acids and has a predicted molecular weight of 48 kDa.

Isoform 2, which shows reduced binding to AXIN1 and reduced kinase activity toward TAU, has been predicted to have a role in axon and neurite growth [168] . Similarly, PKR, which previously was not thought to undergo significant alternative splicing, has recently been found to produce more than 10 different alternately spliced forms in osteosarcoma (unpublished data). The significance of these isoforms in the associated pathology is currently under investigation.

Recent data also indicate that the adenosine deaminase acting on double-strand RNA (ADAR) family proteins are significant players in an alternative splicing regulatory network consisting of ADAR1/2, AKT, GSK3α/β and PKR. The ADAR proteins are a family of RNA editing enzymes that carry-out the deamination of specific adenosine (A) residues in areas of complex RNA structure, converting them into inosine (I). As inosine is interpreted by the cells molecular machinery as guanine, the A-to-I conversion can alter amino acid coding, change mRNA stability, and alter splice acceptor/donor sites, favoring alternative splicing of transcripts. The ADAR family, which consists of ADAR1 (DSRAD), ADAR2 (ADARB1 or RED1) and ADAR3 (ADARB2 or RED2), are responsible by far for the majority of A-to-I editing events in the cell.

While ADAR1 and ADAR2 are rather ubiquitously expressed, ADAR3 expression is more restricted to the brain. Of the family members, only ADAR1 and ADAR2 demonstrate deaminase activity [169, 170] . As these enzymes operate as homo-and heterodimers, ADAR3 may serve to regulate activity of the other two family members. Likewise, the dimer composition has a great influence on substrate selection and rate of catalysis. Moreover, multiple forms of ADAR1 and ADAR2 resulting both from alternate transcriptional initiation and alternative splicing have been documented [170] [171] [172] [173] . It was previously established in diverse studies that ADAR1 interacts with PKR and has the ability to suppress PKR activity [174, 175] . Our group has also found that nuclear PKR associates with ADAR2. Whether this interaction is direct or mediated through ADAR1 is currently not known, and the full significance of the PKR-ADAR1/2 interaction in the nucleus has not been determined. What is known about this interaction is that ADAR1 and ADAR2 in the nucleus are found either in AKT containing complexes or PKR containing complexes, as no duel AKT-PKR complexes with ADAR1/2 are observed. Recently, Bavelloni et al. reported that AKT could regulate ADAR1 and ADAR2 through the phosphorylation of a conserved threonine residue in the catalytic domain [176, 177] . Mimicking phosphorylation by site-directed mutagenesis of this site in either ADAR1 or ADAR2 resulted in a differential decrease in editase activity toward tested substrates ( Figure 5 ). Conversely, it was observed that phosphorylation of the ADARs also had an effect on AKT expression. To date whether, PKR can phosphorylate ADAR1 or ADAR2 has not been determined; a result of few studies examining the substrate network of this kinase and the lack J o u r n a l P r e -p r o o f 22 of a true consensus phosphorylation site. What is now evident is that a regulatory balance is apparently at play whereby AKT-and PKR-dependent signaling influence ADAR-dependent editing and GSK3α/β activity. Both ADAR and GSK3α/β then have a tremendous influence on spliceosome composition and alternative splicing; ADAR though direct editing of splice donor/acceptor sites or alteration of transcripts encoding additional splice regulators and GSK3α/β through the modification of spliceosome components and splicing regulatory proteins. Similarly, both AKT and PKR are known to alter the composition of the spliceosome. Figure 5 depicts the balance that must be maintained between PKR and AKT to achieve proper mRNA splicing in the cell and how metabolic stimulation of AKT or inflammatory activation of PKR can have a significant influence on alternative splicing.

Chronic inflammation and stress have increasingly become associated with disease. Most degenerative pathologies and cancer each have a significant inflammatory component. What has remained elusive is, much like the chicken or the egg, which comes first. Is chronic inflammation a result of the disease or is the disease a result of chronic inflammation. Insulin/insulin-related growth factor signaling may prove the best example. Normal insulin, signaling through the insulin receptor (IR) or insulin-like growth factor (IGF)-1 signaling through the IGF1 receptor (IGF-1R) results in PI3K-AKT-mTOR pathway activation among others. Stimulation of this pathway and other secondary pathways lead to increased protein synthesis, enhanced glucose uptake and metabolism, glycogen storage, and mitochondrial respiration; thus producing the biomolecules necessary for cell growth and proliferation (amino acids, lipids, nucleic acids, ATP). At the same time, increased metabolism generates reactive oxygen species (ROS) and toxic metabolic byproducts that result in the stimulation of stress/inflammatory pathways to protect the cell from subsequent damage. In addition, transient imbalances in synthetic pathways, such as ribosome biogenesis and translation, result in the stimulation of stress response pathways as a mode to reestablish equilibrium [51] . The kinases PKR and GSK3α/β each become activated in response to insulin/IGF1-induced metabolic stress and either directly phosphorylate and/or reduce the synthesis of the insulin receptor substrates (IRS)-1 or IRS2, inhibiting IR and IGF-1R ligand-mediated stimulation of PI3K, thereby making the cells resistant to insulin and IGF1 [178] [179] [180] . As insulin resistance is the main feature defining diabetes mellitus, it is not surprising that constitutive activation of PKR and GSK3α/β is observed in obese/diabetic patients [181, 182] . In fact, the feeding of high fat diets in mice has been observed to result in constitutively elevated levels of J o u r n a l P r e -p r o o f 23 active PKR and GSK3α/β, favoring insulin resistance [180, 181] . In addition to IRS1 and IRS2, the active PKR and GSK3α/β kinases target other downstream proteins that tweak additional pathways regulating transcription, cell division, ribosome biogenesis and translation ( Figure 6A ).

In the case of Alzheimer's disease, a chronic inflammatory state leads to improper protein folding, activation of the unfolded protein response (UPR), and continuous stimulation of GSK3α/β and PKR kinases, resulting among other things in alternate splicing and phosphorylation of TAU, the formation of fibrogenous neural tangles and neuronal cell death, favoring disease progression [32, 183] . Both PKR and GSK3 have become prime candidates for the design of targeted therapies to AD. Those directed toward inhibiting GSK3 have shown some promise in Phase II trials but only at early stages of the disease [184] . In contrast, therapies targeting PKR have thus far shown promising results in in vitro and animal studies, but have yet to enter clinical trials mainly as a result of the lack of critical attention given to PKR in the past [85, 185] . This stress signaling circuit has made it clear that AD, like many other inflammatory related diseases may result from chronic systemic stress and not necessarily local stresses [186] . Empirical evidence supports this conclusion as metabolic disorders like diabetes correlate with enhanced risk of Alzheimer's and various cancers ( Figure 6A ) [187] .

proposed that the relation between cancer and inflammation/stress is that chronic inflammation and stress lead to a selective pressure on the tissues of an organism with survival dependent on random mutations. This is no more clearly evident than in bone marrow failure disorders, where genetic or idiopathic stresses contribute to the loss of bone marrow progenitors in the early stages. In the later stages this is usually replaced by accumulation of blasts in the periphery and full blown leukemia [25, 188] . In many cases, alterations in apoptotic proteins or altered splicing are observed. In cancer, PKR has been suggested to have both a positive and negative role in the disease [25, 30, 48, 189] . What seems evident is that the influence of PKR on diverse cancers is closely associated with its primary cellular location; cytosolic (suppresses tumorigenesis/pro-apoptotic) or nuclear (favors tumorigenesis/progression). Similarly, GSK3α/β have been found to display both tumor promoting and tumor suppressor functions. Enhanced activity, mainly that of GSK3β, has been documented in pancreatic cancer, colorectal cancer and renal cancer while alternative splicing has been reported in leukemia [5] . In contrast, loss of GSK3α/β activity have been associated with skin, lung and breast cancers [5] . What is not clear from many of these studies is the major cellular localization of the GSK3 kinase in question. The importance of this is critical as enhanced nuclear GSK3β activity would be expected to enhance alternative splicing and lead to the degradation of p53, thus favoring J o u r n a l P r e -p r o o f genome instability and a pro-tumorigenic climate. On the other hand, active GSK3α/β activity in the cytosol would favor antiproliferative/homeostasis oriented cellular programming ( Figure 6A ).

Under the current global circumstances, it would be a failure to ignore the role of PKR and GSK3α/β in viral infection. As modulators of the inflammatory response, both PKR and GSK3α/β might be considered innate immune kinases. PKR has long held that role, possibly too well, as the majority of studies over the years have left it labeled simply as an antiviral kinase that autophosphorylates in the presence of dsRNA, thereby lagging far behind the number of studies involving AKT and GSK3. Both kinases are quick to respond to cellular stresses which include among other things viral infection. PKR is the quintessential antiviral kinase. As stated previously, PKR is both a pathogen recognition receptor (PRR) and interferon response gene (ISG); as such it can be activated by viral dsRNAs in a PACT-dependent manner and its expression induced by interferons [19, 25, 40] . Activation of PKR under these conditions leads to the phosphorylation of eIF2α, the accumulation of stress granules, the suppression of general protein synthesis, and the induction of interferon response factor (IRF) stimulated genes as well as ifnα/β transcription and mRNA stability; thus, in theory, having a negative effect on viral protein synthesis [19, 25] . But a number of viruses encode or have found a means to by-pass this checkpoint while still allowing other arms of PKR-mediated signal transduction to continue, thereby turning the activation and synthesis of PKR into a facilitator of the pathology associated with viral infection, as well as in some cases the viral infection itself (see the following Reviews [19, 190, 191] ). Similarly, GSK3α/β are known mediators of the inflammatory response, and GSK3 activity is modulated by certain viruses. In human immunodeficiency virus (HIV) infection, activation of GSK3β is largely responsible for viral associated neuronal cell death and accounts for the main reason why infection with neurotropic viruses such as HIV and herpesviruses are linked to increased risk of developing neurodegenerative disease [192] [193] [194] . For other viruses such as influenza A, GSK3 activation has antiviral activity, while in contrast GSK3β activity promotes hepatitis C viral replication and viral particle production [195, 196] .

Coronaviruses such as severe acute respiratory syndrome coronavirus (SARS-CoV), Mediterranean emerging respiratory syndrome coronavirus (MERS-CoV) and SARS-CoV-2 (COVID-19) are positive-strand RNA viruses that as a family are known for producing a number of unique proteins that repress the cellular antiviral response [197] . In fact, the major pathological manifestations of an enhanced inflammatory response with these viruses result from a delay in IFN induction following infection; this results in an over abundance of proinflammatory monocytes and macrophages and the release of proinflammatory cytokines in the lung, causing tissue damage [198, Journal Pre-proof 199] . Previous, investigations have shown that PKR becomes activated in SARS-CoV and MERS-CoV infections, but this activation was reported to have little antiviral effect while it was still able to contribute to apoptosis of infected cells ( Figure 6B ) [200] [201] [202] . In MERS-CoV infections, the absence of PKR antiviral effects were found to be mediated by the viral accessory protein p4a, which binds to PKR and inhibits its ability to phosphorylate eIF2α, while at the same time leaving other aspects of PKR signaling largely intact [202] . In fact, the targeting of eIF2α phosphorylation more than PKR activity seems to be paramount to coronavirus infection and replication as other non-human coronaviruses both inhibit PKR-dependent phosphorylation of eIF2α as well as enhance the dephosphorylation of eIF2α S51 through a viral-mediated enhancement of GADD34, a component of the PP1 phosphatase [203] . The exact role of PKR in coronavirus infection and the associated pathology is complex, requiring additional clarifying studies. In addition, the current lack of promising activators or inhibitors of PKR-mediated signaling places specific therapeutic interventions along this line out of immediate reach.

In contrast to PKR, the few studies examining the role of GSK3α/β in coronavirus infection have demonstrated that GSK3 activity has a proviral effect. In SARS-CoV infection, GSK3 activity is required to phosphorylate the viral nucleocapsid protein (N), thus allowing the interaction of the host cell RNA helicase DDX1 with phosphorylated nucleocapsid complexes, which alters the viral polymerase in a manner that favors viral genome replication over viral mRNA transcription [204, 205] . Inhibition of GSK3 with a specific inhibitor was shown to block viral replication and reduce the viral titer in vitro [205] . Compounds, targeting GSK3β, such as the orphan drug Tideglusib, which was given orphan status after phase II trials did not reach the set end points in treating Alzheimer's disease, may be promising therapeutics to repurpose as antiviral therapies to combat emerging coronavirus diseases like SARS-CoV-2 ( Figure 6B) [206].

The GSK3 and PKR kinases control responses from ribosome biogenesis, translation, cellular division and metabolic regulation placing them at key positions in normal cell homeostasis and pathologies such as cancer, neurodegenerative disease, diabetes mellitus and viral infection.

While these kinase share much in common (substrates and signal transduction pathways) as well as have the capability to regulate one another or facilitate/reinforce the signal transduction mediated by the other, there is still a tremendous amount of work to be done to understand the regulatory effects that are specific to each kinase and what upstream signaling dictates the intersection of their [20] B.R. Williams, PKR; a sentinel kinase for cellular stress, Oncogene, 18 (1999) 6112-6120. Other PKR-dependent signaling is often left intact, enhancing inflammation. While active PKR is denoted as phosphorylated on T446 and T451, only T451 is required for activity. Phosphorylation of T446 as well as numerous additional sites influences the degree of activation. Phosphorylation sites labeled in green (ex., Tyr216-GSK3β) indicate activating or stabilizing modifications; phosphorylation sites labeled in red (ex., Ser51-eIF2α) indicate modifications that are inhibitory to the substrates activity. Lines with arrowhead termini represent promoting effects on the downstream signaling; lines with boxed termini represent inhibitory effects on downstream BCL-X L is the major hematopoetic antiapoptotic BCL-2 family member. BCL-X L is best known for its ability to bind BAX blocking the homodimerization of BAX and the initiation of the caspase-9 mediated apoptotic pathway. [49, 135, 136] Cyclin D1

TNFα stimulation in PKR+/+ and -/-MEFs 

Highlights:

 The GSK3α/β and PKR kinases have major roles in maintaining cellular homeostasis and regulating the cells response to stress/inflammation and infection.  Not only do GSK3α/β and PKR interact with and/or modify many of the same target proteins, they also affect the expression of a number of genes similarly.  Evidence supports the hypothesis that a balance between AKT and PKR signaling in the nucleus may be responsible for regulating the activation of nuclear GSK3β.  GSK3α/β-and PKR-dependent signaling pathways both influence major molecular mechanisms of the cell (transcription, translation, ribosome biogenesis and alternative splicing) through similar intermediates.  Aberrant activation of GSK3α/β, PKR and their associated pathways resulting from chronic inflammatory signaling is intimately involved in cancer, diabetes/metabolic disorders, and neurodegenerative diseases.

Journal Pre-proof 

Activation of the dsRNA-Activated Protein Kinase PKR in Mitochondrial Dysfunction and Inflammatory Stress in Metabolic Syndrome

Glycogen synthase kinase-3 (GSK3): regulation, actions, and diseases

Targeting GSK3 signaling as a potential therapy of neurodegenerative diseases and aging

The impact of PKR activation: from neurodegeneration to cancer

Glycogen synthase kinases: Moonlighting proteins with theranostic potential in cancer

Multifaceted Roles of GSK-3 in Cancer and Autophagy-Related Diseases

The PKR kinase promoter binds both Sp1 and Sp3, but only Sp3 functions as part of the interferon-inducible complex with ISGF-3 proteins

Modulation of tau phosphorylation by the kinase PKR: implications in Alzheimer's disease

PKR regulates proliferation, differentiation, and survival of murine hematopoietic stem/progenitor cells

Isolation of the interferon-inducible RNA-dependent protein kinase Pkr promoter and identification of a novel DNA element within the 5'-flanking region of human and mouse Pkr genes

Tyrosine phosphorylation acts as a molecular switch to full-scale activation of the eIF2alpha RNA-dependent protein kinase

Autophosphorylation in the activation loop is required for full kinase activity in vivo of human and yeast eukaryotic initiation factor 2alpha kinases PKR and GCN2

Protein Kinase R Degradation Is Essential for Rift Valley Fever Virus Infection and Is Regulated by SKP1-CUL1-Fbox (SCF)FBXW11-NSs E3 Ligase

Differential effects of SUMO1 and SUMO3 on PKR activation and stability

RAX, a cellular activator for double-stranded RNA-dependent protein kinase during stress signaling

RAX, the PKR activator, sensitizes cells to inflammatory cytokines, serum withdrawal, chemotherapy, and viral infection

PKR activity is required for acute leukemic cell maintenance and growth: a role for PKR-mediated phosphatase activity to regulate GSK-3 phosphorylation

A Compendium of RNA-Binding Proteins that Regulate MicroRNA Biogenesis

Involvement of ERKs, RSK2 and PKR in UVA-induced signal transduction toward phosphorylation of eIF2alpha (Ser(51))

Translation inhibition in apoptosis: caspase-dependent PKR activation and eIF2-alpha phosphorylation

PACT, a protein activator of the interferon-induced protein kinase, PKR, The EMBO journal

Identification of the PKR nuclear interactome reveals roles in ribosome biogenesis, mRNA processing and cell division

PKR: A Kinase to Remember

Protein kinase R and its cellular regulators in cancer: An active player or a surveillant?

Genetic deletion of PKR abrogates TNF-induced activation of IkappaBalpha kinase, JNK, Akt and cell proliferation but potentiates p44/p42 MAPK and p38 MAPK activation

The eIF2alpha kinases PERK and PKR activate glycogen synthase kinase 3 to promote the proteasomal degradation of p53

Signal Transduction in Ribosome Biogenesis: A Recipe to Avoid Disaster

Regulation of guanine nucleotide exchange through phosphorylation of eukaryotic initiation factor eIF2alpha. Role of the alpha-and delta-subunits of eiF2b

Tight binding of the initiation

Regulation of internal ribosome entry site-mediated translation by eukaryotic initiation factor-2alpha phosphorylation and translation of a small upstream open reading frame

The zipper model of translational control: a small upstream ORF is the switch that controls structural remodeling of an mRNA leader

Phosphorylation of initiation factor-2 alpha is required for activation of internal translation initiation during cell differentiation

The protein kinase PKR: a molecular clock that sequentially activates survival and death programs

Phosphorylation of eIF2alpha is responsible for the failure of the picornavirus internal ribosome entry site to direct translation from Sindbis virus replicons

Cellular internal ribosome entry segments: structures, trans-acting factors and regulation of gene expression

TNF blockade: an inflammatory issue

Inflammation and cancer: how hot is the link?

CDDO induces granulocytic differentiation of myeloid leukemic blasts through translational up-regulation of p42 CCAAT enhancer binding protein alpha

An upstream open reading frame regulates translation of GADD34 during cellular stresses that induce eIF2alpha phosphorylation

Regulation of Cited2 expression provides a functional link between translational and transcriptional responses during hypoxia

Proud, eIF2 and the control of cell physiology

Eukaryotic initiation factor 2B: identification of multiple phosphorylation sites in the epsilonsubunit and their functions in vivo

The kinase DYRK phosphorylates protein-synthesis initiation factor eIF2Bepsilon at Ser539 and the microtubule-associated protein tau at Thr212: potential role for DYRK as a glycogen synthase kinase 3-priming kinase

Eukaryotic initiation factor 2B and its role in alterations in mRNA translation that occur under a number of pathophysiological and physiological conditions

Translational control of growth factor and proto-oncogene expression

Translational control of gene expression during hypoxia

Internal ribosome entry segmentmediated translation during apoptosis: the role of IRES-trans-acting factors

The interaction of glycogen synthase kinase-3 (GSK-3) with the cell cycle

Cell cycle regulation of the double stranded RNA activated protein kinase

Novel MicroRNA Reporter Uncovers Repression of Let-7 by GSK-3beta

Phosphorylation of Tip60 by GSK-3 determines the induction of PUMA and apoptosis by p53

Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3beta

Glycogen synthase kinase 3-dependent phosphorylation of Mdm2 regulates p53 abundance

Ionizing radiation can induce GSK-3beta phosphorylation and NF-kappaB transcriptional transactivation in ATM-deficient fibroblasts

Double-stranded-RNA-activated protein kinase PKR enhances transcriptional activation by tumor suppressor p53

The double-stranded RNA activated protein kinase PKR physically associates with the tumor suppressor p53 protein and phosphorylates human p53 on serine 392 in vitro

The RAX/PACT-PKR stress response pathway promotes p53 sumoylation and activation, leading to G(1) arrest

Hepatitis C Virus Indirectly Disrupts DNA Damage-Induced p53 Responses by Activating Protein Kinase R, mBio

Pathological missorting of endogenous MAPT/Tau in neurons caused by failure of protein degradation systems

Tau mis-splicing in the pathogenesis of neurodegenerative disorders

Alzheimer's research & therapy

GSK-3beta, a pivotal kinase in Alzheimer disease

Protein kinase C and calcium/calmodulin-dependent protein kinase II phosphorylate three-repeat and four-repeat tau isoforms at different rates

Tau is phosphorylated by GSK-3 at several sites found in Alzheimer disease and its biological activity markedly inhibited only after it is prephosphorylated by A-kinase

Rapid tau protein dephosphorylation and differential rephosphorylation during cardiac arrestinduced cerebral ischemia and reperfusion

The endogenous and cell cycledependent phosphorylation of tau protein in living cells: implications for Alzheimer's disease

Phosphorylation that detaches tau protein from microtubules (Ser262, Ser214) also protects it against aggregation into Alzheimer paired helical filaments

Site-specific effects of tau phosphorylation on its microtubule assembly activity and self-aggregation, The European journal of neuroscience

PKA modulates GSK-3beta-and cdk5-catalyzed phosphorylation of tau in site-and kinase-specific manners

Involvement of aberrant glycosylation in phosphorylation of tau by cdk5 and GSK-3beta

Spectroscopic studies of GSK3{beta} phosphorylation of the neuronal tau protein and its interaction with the N-terminal domain of apolipoprotein E

PP2A regulates tau phosphorylation directly and also indirectly via activating GSK-3beta

Regulation of phosphorylation of tau by cyclin-dependent kinase 5 and glycogen synthase kinase-3 at substrate level

Phosphorylation of tau at both Thr 231 and Ser 262 is required for maximal inhibition of its binding to microtubules

Thr175-phosphorylated tau induces pathologic fibril formation via GSK3beta-mediated phosphorylation of Thr231 in vitro

c-jun N-terminal kinase hyperphosphorylates R406W tau at the PHF-1 site during mitosis

High-content siRNA screening of the kinome identifies kinases involved in Alzheimer's disease-related tau hyperphosphorylation

Unmasking the Mysteries of MYC

Phosphorylation by glycogen synthase kinase-3 controls cmyc proteolysis and subnuclear localization

The Fbw7 tumor suppressor regulates glycogen synthase kinase 3 phosphorylation-dependent c-Myc protein degradation

Protein phosphatase 2A regulatory subunit B56alpha associates with c-myc and negatively regulates c-myc accumulation

Regulation of c-myc expression by IFN-gamma through Stat1-dependent and -independent pathways

The alternatively initiated c-Myc proteins differentially regulate transcription through a noncanonical DNA-binding site

Signaling via the NFkappaB system

NF-kappaB activation by tumour necrosis factor requires the Akt serine-threonine kinase

Distinct roles of the Ikappa B kinase alpha and beta subunits in liberating nuclear factor kappa B (NF-kappa B) from Ikappa B and in phosphorylating the p65 subunit of NF-kappa B

Akt regulates basal and induced processing of NF-kappaB2 (p100) to p52

A cell cycle regulatory network controlling NF-kappaB subunit activity and function

Regulation of mammalian target of rapamycin activity in PTEN-inactive prostate cancer cells by I kappa B kinase alpha

IKK beta suppression of TSC1 links inflammation and tumor angiogenesis via the mTOR pathway

Bile acid exposure up-regulates tuberous sclerosis complex 1/mammalian target of rapamycin pathway in Barrett'sassociated esophageal adenocarcinoma

Akt-dependent regulation of NF-{kappa}B is controlled by mTOR and Raptor in association with IKK

Differential involvement of IkappaB kinases alpha and beta in cytokine-and insulin-induced mammalian target of rapamycin activation determined by Akt

Glycogen synthase kinase-3 beta regulates NF-kappa B1/p105 stability

Glycogen synthase kinase 3beta functions to specify gene-specific, NF-kappaB-dependent transcription

Role of glycogen synthase kinase-3 in TNF-alpha-induced NF-kappaB activation and apoptosis in hepatocytes

A model for NF-kappa B regulation by GSK-3 beta

Glycogen synthase kinase-3beta participates in nuclear factor kappaB-mediated gene transcription and cell survival in pancreatic cancer cells

Toll-like receptor-mediated cytokine production is differentially regulated by glycogen synthase kinase 3

Inhibition of GSK-3beta decreases NF-kappaB-dependent gene expression and impairs the rat liver regeneration

Inhibition of glycogen synthase kinase-3 activity leads to epigenetic silencing of nuclear factor kappaB target genes and induction of apoptosis in chronic lymphocytic leukemia B cells

Requirement for glycogen synthase kinase-3beta in cell survival and NF-kappaB activation

GSK-3beta controls NF-kappaB activity via IKKgamma/NEMO

Phosphorylation of serine 468 by GSK-3beta negatively regulates basal p65 NF-kappaB activity

Proteomic screen reveals Fbw7 as a modulator of the NF-kappaB pathway

Fbxw7alpha-and GSK3-mediated degradation of p100 is a pro-survival mechanism in multiple myeloma

A role for phosphorylation in the proteolytic processing of the human NF-kappa B1 precursor

Maintenance of constitutive IkappaB kinase activity by glycogen synthase kinase-3alpha/beta in pancreatic cancer

Genetic deletion of glycogen synthase kinase-3beta abrogates activation of IkappaBalpha kinase, JNK, Akt, and p44/p42 MAPK but potentiates apoptosis induced by tumor necrosis factor

Glycogen synthase kinase-3 inhibition disrupts nuclear factor-kappaB activity in pancreatic cancer, but fails to sensitize to gemcitabine chemotherapy

Hepatocyte growth factor exerts its anti-inflammatory action by disrupting nuclear factor-kappaB signaling

Glycogen synthase kinase 3beta: a novel marker and modulator of inflammatory injury in chronic renal allograft disease

Double-stranded RNA activates binding of NF-kappa B to an inducible element in the human beta-interferon promoter

The involvement of NF-kappa B in betainterferon gene regulation reveals its role as widely inducible mediator of signal transduction

Double-stranded RNA and bacterial lipopolysaccharide enhance sensitivity to TNF-alpha-mediated cell death

Double-stranded RNA-dependent protein kinase activates transcription factor NF-kappa B by phosphorylating I kappa B

The interferoninducible protein kinase PKR modulates the transcriptional activation of immunoglobulin kappa gene

Deficient signaling in mice devoid of double-stranded RNA-dependent protein kinase

Deficient cytokine signaling in mouse embryo fibroblasts with a targeted deletion in the PKR gene: role of IRF-1 and NF-kappaB

Chronic human immunodeficiency virus type 1 infection of myeloid cells disrupts the autoregulatory control of the NF-kappaB/Rel pathway via enhanced IkappaBalpha degradation

NF-kappaB activation by double-stranded-RNA-activated protein kinase (PKR) is mediated through NF-kappaB-inducing kinase and IkappaB kinase

Activation of NF-kappa B by the dsRNA-dependent protein kinase, PKR involves the I kappa B kinase complex

PKR stimulates NF-kappaB irrespective of its kinase function by interacting with the IkappaB kinase complex

The N-terminus of PKR is responsible for the activation of the NF-kappaB signaling pathway by interacting with the IKK complex

Activation of the I kappa B alpha kinase (IKK) complex by double-stranded RNA-binding defective and catalytic inactive mutants of the interferon-inducible protein kinase PKR

The catalytic activity of dsRNAdependent protein kinase, PKR, is required for NF-kappaB activation

The C/EBP family of transcription factors: a paradigm for interaction between gene expression and proliferation control

Translational control of C/EBPalpha and C/EBPbeta isoform expression

Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein alpha-deficient mice

Five members of the CEBP transcription factor family are targeted by recurrent IGH translocations in B-cell precursor acute lymphoblastic leukemia (BCP-ALL)

Overexpression of CEBPA resulting from the translocation t(14lymphoblastic leukemia

Target proteins of C/EBPalphap30 in AML: C/EBPalphap30 enhances sumoylation of C/EBPalphap42 via up-regulation of Ubc9

SUMOylation interferes with CCAAT/enhancer-binding protein beta-mediated c-myc repression, but not IL-4 activation in T cells

CCAAT/enhancer binding proteinbeta is a mediator of keratinocyte survival and skin tumorigenesis involving oncogenic Ras signaling

Conditional ablation of C/EBP beta demonstrates its keratinocyte-specific requirement for cell survival and mouse skin tumorigenesis

Functional characterisation of the regulation of CAAT enhancer binding protein alpha by GSK-3 phosphorylation of Threonines 222/226

Sequential phosphorylation of CCAAT enhancer-binding protein beta by MAPK and glycogen synthase kinase 3beta is required for adipogenesis

A novel GSK-3 beta-C/EBP alpha-miR-122-insulin-like growth factor 1 receptor regulatory circuitry in human hepatocellular carcinoma

Phosphoproteomics reveals that glycogen synthase kinase-3 phosphorylates multiple splicing factors and is associated with alternative splicing

Amyloid-beta peptide alteration of tau exon-10 splicing via the GSK3beta-SC35 pathway

Gene structure and alternative splicing of glycogen synthase kinase 3 beta (GSK-3beta) in neural and non-neural tissues

An alternatively spliced form of glycogen synthase kinase-3beta is targeted to growing neurites and growth cones

ADAR RNA editing in human disease; more to it than meets the I

Expression of interferon-inducible RNA adenosine deaminase ADAR1 during pathogen infection and mouse embryo development involves tissueselective promoter utilization and alternative splicing

Human RNA-specific adenosine deaminase ADAR1 transcripts possess alternative exon 1 structures that initiate from different promoters, one constitutively active and the other interferon inducible

Differential Enzymatic Activity of Rat ADAR2 Splicing Variants Is Due to Altered Capability to Interact with RNA in the Deaminase Domain

Novel splice variants of human ADAR2 mRNA: skipping of the exon encoding the dsRNA-binding domains, and multiple C-terminal splice sites

Human ADAR1 Prevents Endogenous RNA from Triggering Translational Shutdown

Stress granule formation induced by measles virus is protein kinase PKR dependent and impaired by RNA adenosine deaminase ADAR1

AKT-dependent phosphorylation of the adenosine deaminases ADAR-1 and -2 inhibits deaminase activity

AKT-Dependent Phosphorylation of ADAR1p110 and ADAR2 Represents a New and Important Link Between Cell Signaling and RNA Editing

Serine 332 phosphorylation of insulin receptor substrate-1 by glycogen synthase kinase-3 attenuates insulin signaling

Sequential phosphorylation of insulin receptor substrate-2 by glycogen synthase kinase-3 and c-Jun NH2-terminal kinase plays a role in hepatic insulin signaling

Double-stranded RNA-dependent protein kinase links pathogen sensing with stress and metabolic homeostasis

Dysregulation of glycogen synthase kinase-3 in skeletal muscle and the etiology of insulin resistance and type 2 diabetes

Modulation of double-stranded RNA-activated protein kinase in insulin sensitive tissues of obese humans

Effect of amyloid-beta (Abeta) immunization on hyperphosphorylated tau: a potential role for glycogen synthase kinase (GSK)-3beta

A. investigators, A phase II trial of tideglusib in Alzheimer's disease

Restoring synaptic plasticity and memory in mouse models of Alzheimer's disease by PKR inhibition

Could Alzheimer's Disease Originate in the Periphery and If So How So?

Diabetes mellitus and Alzheimer's disease: GSK-3beta as a potential link

RNA processing and RNA ribosome biogenesis in bone marrow failure disorders

PKR is activated in MDS patients and its subcellular localization depends on disease severity

Protein kinase R and the inflammasome

Protein kinase PKR and RNA adenosine deaminase ADAR1: new roles for old players as modulators of the interferon response

Glycogen synthase kinase-3beta (GSK-3beta) inhibitors AR-A014418 and B6B3O prevent human immunodeficiency virus-mediated neurotoxicity in primary human neurons

Herpes Simplex Virus type-1 infection induces synaptic dysfunction in cultured cortical neurons via GSK-3 activation and intraneuronal amyloid-beta protein accumulation

Glycogen synthase kinase 3 beta (GSK-3 beta) as a therapeutic target in neuroAIDS

Inhibition of Akt kinase activity suppresses entry and replication of influenza virus

Glycogen Synthase Kinase 3beta Enhances Hepatitis C Virus Replication by Supporting miR-122

Innate Immune Evasion by Human Respiratory RNA Viruses

Active replication of Middle East respiratory syndrome coronavirus and aberrant induction of inflammatory cytokines and chemokines in human macrophages: implications for pathogenesis

SARS-CoV pathogenesis is regulated by a STAT1 dependent but a type I, II and III interferon receptor independent mechanism

Antagonism of dsRNA-Induced Innate Immune Pathways by NS4a and NS4b Accessory Proteins during MERS Coronavirus Infection

Proteins are identified as to whether they were found in association with GSK3α, GSK3β, active PKR (PKR+), inactive PKR (PKR-) or with PKR in an unknown state of activation (PKR)

Information was obtained from the following: 1. UniProtKB database

PhosphoSite Plus (www.phosphosite.org)

National Institute for Biotechnological Information (NCBI) National Institutes of Health (NIH) Gene Database