Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 18 October 2019

Role of Eukaryotic Initiation Factor eIF2B in Vanishing White Matter Disease

Abstract and Keywords

Vanishing white matter (VWM) disease is a recessive disorder characterized by gradual loss of white matter and of myelin. Its clinical severity is high variable. VWM is caused by mutations in any one of the five genes encoding subunits of eukaryotic initiation factor 2B (eIF2B), a ubiquitous, multimeric protein that plays crucial roles in protein synthesis and its control. There are now known to be at least 160 mutations in eIF2B genes that lead to VWM. Where tested, most mutations impair the activity or integrity of the eIF2B complex. However, it remains unclear how and why defects in eIF2B lead to VWM. This article discusses recent advances in understanding the structure and functions of eIF2B and the pathogenic basis of VWM.

Keywords: eIF2B, translation initiation, vanishing white matter, integrated stress response, ternary complex formation, three-dimensional structure model, mouse model, protein synthesis

Abbreviations

  • AGP,

    adenosine diphosphate-glucose pyrophosphorylase;

  • AHA,

    azidohomoalanine;

  • ATF4,

    activating transcription factor 4;

  • eIF,

    eukaryotic initiation factor;

  • GCN2,

    general control nonderepressible 2;

  • GDF,

    GDI-dissociation factor;

  • GDI,

    GDP-dissociation inhibitor;

  • GDP,

    guanosine diphosphate;

  • GEF,

    guanine nucleotide-exchange factor;

  • GFAPδ,

    glial fibrillary acidic protein δ;

  • GTP,

    guanosine triphosphate;

  • ISR,

    integrated stress response;

  • ISRIB,

    ISR inhibitor;

  • MBP,

    myelin basic protein;

  • MOG,

    myelin oligodendrocyte glycoprotein;

  • MRI,

    magnetic resonance imaging;

  • Met-tRNAMeti,

    initiator methionyl-transfer RNA;

  • NF,

    Asn-Phe;

  • OPC,

    oligodendrocyte precursor cell;

  • PERK,

    protein kinase RNA-like endoplasmic reticulum kinase;

  • uORF,

    upstream open reading frame;

  • VWM,

    vanishing white matter;

  • WT,

    wild-type.

Vanishing White Matter

Vanishing white matter (VWM; also termed childhood ataxia with diffuse central nervous system hypomyelination, CACH) is one of the most prevalent inherited childhood white matter disorders (van der Knaap, Breiter, Naidu, Hart, & Valk, 1999; van der Knaap et al., 1977). The diagnosis of VWM is based on magnetic resonance imaging (MRI) and genetic testing. The MRI pattern characteristic for VWM shows diffuse abnormality of the cerebral white matter. When the disease progresses, increasing amounts of white matter are replaced by fluid (van der Knaap et al., 1977). The brain consists of gray and white matter. The gray matter contains the neurons, while the white matter contains axons myelinated by oligodendrocytes. The myelin is high in fat and makes myelinated structures of the brain white. In VWM the myelin is not maintained properly and the white matter literally vanishes. Currently, there is no curative treatment for VWM.

The clinical severity of the disease is quite variable and correlates mostly with the age of onset (Fogli et al., 2004a; van der Knaap et al., 1998). Three groups have been discriminated (Fogli & Boespflug-Tanguy, 2006; van der Knaap, Pronk, & Scheper, 2006). The classical phenotype is associated with an early childhood onset at an age between 2 and 6 years (van der Knaap et al., 2006). A severe phenotype is observed in patients with a disease onset before the age of 2 years (antenatal or infantile onset) (van der Knaap et al., 2006). The third group comprises patients with a relatively mild phenotype, characterized by a late childhood or adult onset (van der Knaap et al., 2006).

The classical phenotype is most frequently observed. Patients display chronic as well as episodic neurological deterioration (van der Knaap et al., 1977, 1998). They develop ataxia and spasticity and become wheelchair-dependent (van der Knaap et al., 1977, 1998; van der Lei et al., 2010; Hanefeld et al., 1993). Eventually VWM patients die prematurely, usually a few years after diagnosis. Episodic deterioration is provoked by minor head trauma, infections with fever, or acute fright (van der Knaap et al., 2006; Vermeulen et al., 2005). These episodes may end in an unexplained coma, which in some cases results in death (van der Knaap et al., 2006). Alternatively, the episodes result in slow and partial recovery. In general, VWM patients have no or only mild epilepsy (van der Knaap et al., 2006). Patients receive interventions that prevent episodic deterioration (van der Knaap et al., 2006).

The brain is the most affected organ in all patients (van der Knaap et al., 2006). Severe cases of VWM with antenatal onset are affected also in other organs, including lens, liver, kidney, pancreas, and ovaries (Fogli et al., 2003; van der Knaap et al., 2003). In addition ovarian dysfunction has been observed in adolescent and adult female patients (Fogli et al., 2003; van der Knaap et al., 2003).

Postmortem neuropathological examination shows an abnormal morphology of astrocytes and oligodendrocytes in the white matter of the central nervous system. White matter oligodendrocytes look foamy, and white matter astrocytes have thick and blunted processes (Bugiani et al., 2011; Van Haren, van der Voorn, Peterson, van der Knaap, & Powers, 2004). The morphology of astrocytes in gray matter is normal (Bugiani et al., 2011). Interestingly, increased numbers of white matter astrocytes are positive for nestin, a marker for immature astrocytes. Reduced numbers of astrocytes express S100β, a marker for mature astrocytes. Combined these observations indicate that increased numbers of astrocytes are immature (Bugiani et al., 2011; Middeldorp et al., 2010; Raponi et al., 2007). In addition, high numbers of oligodendrocyte precursor cells (OPCs) are found in affected white matter structures (Bugiani et al., 2011; Van Haren et al., 2004). In general, astrocytes are important for blood–brain barrier formation and maintenance, uptake of neurotransmitters, regulation of oxygen, and glucose availability for neurons (Lundgaard, Osório, Kress, Sanggaard, & Nedergaard, 2014; Garden & Campbell 2016). Astrocytes have also been described to metabolically support oligodendrocytes (Hirrlinger & Nave, 2014; Kiray, Lindsay, Hosseinzadeh, & Barnett, 2016). Upon injury, astrocytes form scar tissue to help prevent further damage and maintain brain/tissue homeostasis in a process called “reactive astrogliosis”. Oligodendrocytes are the cells that produce the extensive myelin sheets around the axons of neurons. Functional myelin is essential for the normal conduction of action potentials in the brain (Garden & Campbell, 2016). It has been postulated that, in VWM, white matter astrocytes and oligodendrocytes do not mature properly and therefore cannot execute their normal function.

Mutations in eIF2B Cause VWM

VWM is a genetic disease with a recessive mode of inheritance. Due to a founder effect, the incidence of VWM in the Netherlands is relatively high. Together with the characteristic MRI pattern, this founder effect facilitated research into the underlying genetic cause. The first gene associated with VWM was found in a genetic linkage study of a group of Dutch families (Leegwater et al., 1999). This study localized the gene on chromosome 3q27 (Leegwater et al., 1999). Sequence analyses identified a T91A mutation in the gene EIF2B5 in VWM patients that segregated with disease (Leegwater et al., 2001). The mutation was found in a homozygous state. Some patients were compound heterozygous for a second mutation in EIF2B5 (Leegwater et al., 2001). A second group of patients from another geographical region with relatively high incidence was investigated. Sequence analysis of this group identified the E213G mutation in EIF2B2 (Leegwater et al., 2001). Patients without mutations in EIF2B2 or EIF2B5 are homozygous or compound heterozygous for mutations in the remaining EIF2B1, EIF2B3, or EIF2B4 genes (van der Knaap et al., 2002). Taken together, these findings lead to the conclusion that VWM is caused by mutations in any of the five genes encoding the subunits of eIF2B (α, β, γ, δ, and ε, in order of increasing size).

Founder mutations for VWM exist not only in the Dutch population; others have been described, such as in the Cree population in North America (R195H in eIF2Bε) and in China (I346T in eIF2Bγ) (Fogli et al., 2002; Zhang et al., 2015). In other countries patients with sporadic mutations in EIF2B1–5 are found in all five subunits (Pronk, van Kollenburg, Scheper, & van der Knaap, 2006). At the moment there are more than 100 different mutations known to be involved in VWM (Pronk et al., 2006; Bugiani, Boor, Powers, Scheper, & van der Knaap, 2010). Almost two thirds of the patients have mutations in EIF2B5, and only a few have mutations in EIF2B1 (Pronk et al., 2006).

The majority of mutations are of a missense character. Frameshift or nonsense mutations are always compound heterozygous with a missense mutation, and this missense mutation is associated with a mild phenotype (Pronk et al., 2006; Bugiani et al., 2010). The catalytic domain in eIF2Bε is relatively spared for pathogenic mutations (Pronk et al., 2006). The absence of homozygous frameshift or nonsense mutations in VWM patients likely reflects the function of eIF2B, which is essential for survival. The eIF2B protein is highly conserved in eukaryotes, and loss of either the eIF2Bβ, eIF2Bγ, eIF2Bδ, or eIF2Bε subunit makes yeast nonviable, indicating that—as expected for a translation factor—eIF2B knockout is not compatible with life (Hinnebusch, 1996). The genotype–phenotype correlation in VWM is difficult to determine due to the phenotypic variability between sibling patients (Bugiani et al., 2010). However, some mutations, such as R113H in EIF2B5 in a homozygous state, always correlate with a relatively late onset and slow disease progression. Moreover, this mutation is not conserved between mammals: for example, mice and rats have a histidine at the equivalent position. This lack of conservation may explain the relatively mild phenotype associated with the R113H mutation (van der Lei et al., 2010; Bugiani et al., 2010). The R195H mutation in EIF2B5 is associated with a severe phenotype (Black et al., 1988). An overview of the mutations referred to in this chapter is given in Table 1.

Table 1 Overview of eIF2B Mutations, the Effects on eIF2B Structure and Activity and Disease Severity

Mutation

Effect on eIF2B structure

Effect on eIF2B activitya

Disease severity

eIF2Bα V183F

Loss of eIF2Bα dimerization (Wortham et al., 2014; Wortham & Proud, 2015)

↓ (Wortham & Proud, 2015)

Mild (Ohlenbusch et al., 2005)

eIF2Bβ E213G

No effect (Li et al., 2004; Richardson, Mohammad, & Pavitt, 2004)

↓↓ (Li et al., 2004)

Classical (Fogli et al., 2004a)

eIF2Bβ P291S

Loss of complex integrity (Liu et al., 2011)

eIF2Bβ V316D

Loss of complex integrity (Li et al., 2004; Liu et al., 2011)

↓↓↓ (Li et al., 2004)

eIF2Bγ R225Q

No effect (Liu et al., 2011)

± (Liu et al., 2011)

Classical (Liu et al., 2011)

eIF2Bγ I346T

Classical (Wu et al., 2009)

eIF2Bδ R357W

Loss of complex integrity (Liu et al., 2011)

↓↓↓ (Liu et al., 2011)

Classical (Fogli et al., 2004a)

eIF2Bδ A391D

No effect (Liu et al., 2011)

± (Liu et al., 2011)

Severe (van der Knaap et al., 2003)

eIF2Bδ R483W

Loss of complex integrity (Liu et al., 2011)

↓↓↓ (Liu et al., 2011)

Severe (van der Knaap et al., 2003)

eIF2Bε T91A

Reduced binding to complex (Li et al., 2004)

↓↓ (Li et al., 2004)

Classical (Fogli et al., 2004a; van der Lei et al., 2010)

eIF2Bε R113H

No effect (Wang, Wortham, Liu, & Proud, 2012)

↓↓ (Li et al., 2004)

Mild (van der Lei et al., 2010)

eIF2Bε R136H

↓ (Geva et al., 2010)

Classical (Kantor et al., 2005)

eIF2Bε R195H

↓↓↓ (Li et al., 2004)

Severe (Fogli et al., 2004a)

eIF2Bε Y495C

No effect (Liu et al., 2011)

± (Liu et al., 2011)

Severe (van der Knaap et al., 2003)

(a) eIF2B activity (relative to control) ↓70–90%; ↓↓50–70%; ↓↓↓<50%; ±indicates no significant difference from controls.

eIF2B Function and Regulation

Regulation of mRNA translation plays a key role in cell and organismal physiology. This encompasses both the overall regulation of the rate of protein synthesis, to match the cell’s needs or resources, and the modulation of the translation of specific mRNAs and thus the production of specific proteins or protein isoforms. Since the translation of mRNAs for certain transcription factors is regulated, control of mRNA translation can affect the expression of many genes, including ones whose mRNAs are not themselves subject to translational control.

The regulatory network involving the translation initiation factors eIF2 and eIF2B (eukaryotic initiation factor 2 and 2B) provides an excellent and well-studied example of just this type of control system (Pavitt, 2005). eIF2 brings the initiator methionyl-tRNA (Met-tRNAMeti) to the ribosome to recognize the start codon in mRNAs and thus get translation underway (Figure 1). In order to bind Met-tRNAMeti, eIF2 must itself be associated with guanosine triphosphate (GTP). eIF2⋅GTP ⋅Met-tRNAMeti is known as a “ternary complex.” The GTP is hydrolyzed during translation initiation, and eIF2 leaves the ribosome as inactive eIF2⋅guanosine diphosphate (GDP). Since GDP dissociates only slowly from eIF2, an extra factor, eIF2B, is needed to maintain the required rate of translation initiation. eIF2B acts as a GDP-dissociation stimulating factor to promote release of GDP from eIF2 to allow its replacement by GTP and regenerate active eIF2⋅GTP (Williams, Price, Loughlin, & Proud, 2001) (Figure 1). eIF2B is generally referred to as a guanine nucleotide-exchange factor (GEF). In yeast, it has also been shown to promote the dissociation of a further factor, eIF5, from eIF2 to allow nucleotide exchange. eIF5 normally impairs the release of GDP from eIF2 and therefore has GDP-dissociation inhibitor (GDI) activity (Jennings et al., 2016; Jennings & Pavitt, 2010). eIF2B thus acts as a GDI-dissociation factor, or GDF (Jennings & Pavitt, 2010). eIF5 actually has dual roles, acting both as a GDI and a GTPase-activator protein for eIF2⋅GTP (Paulin, Campbell, O’Brien, Loughlin, & Proud, 2001) (Figure 1). This allows proper control of the initiation stage of protein synthesis under normal conditions and when eIF2 is phosphorylated in response to a range of stress conditions (Figure 2). It has not yet been tested whether eIF2B from other species, including mammals, also displays these additional roles; but given the high degree of similarity between eIF2, eIF2B, and eIF5 from diverse species, this seems likely.

Role of Eukaryotic Initiation Factor eIF2B in Vanishing White Matter DiseaseClick to view larger

Figure 1. Role of eIF2B.

eIF2B promotes the dissociation of GDP from eIF2, allowing GTP to bind to form active eIF2.GTP complexes. This figure also illustrates the function of eIF2, to recruit – in its GTP-bound form – the initiator methionyl-tRNA to the 40S ribosomal subunit and the dual role of eIF5 as a GAP for eIF2-bound GTP and a GDI for the resulting eIF2-GDP complex. Red and green indicate, respectively, the inactive and active forms of eIF2.

The eIF2/eIF2B system is subject to several types of regulation. The first to be discovered and the best understood involves the phosphorylation of eIF2 (Pavitt, 2005) (Figure 2). eIF2 comprises three different subunits, α–γ. eIF2α undergoes phosphorylation on Ser51, an event which can be catalyzed by any one of four different kinases in mammals (Figure 2). When phosphorylated in this way, eIF2 becomes a competitive inhibitor of its own GEF, eIF2B, thus slowing down the production of active eIF2⋅GTP (Rowlands, Panniers, & Henshaw, 1988). Overall protein synthesis is thus inhibited. Importantly, this mechanism can actually increase translation of the coding regions of some mRNAs.

Role of Eukaryotic Initiation Factor eIF2B in Vanishing White Matter DiseaseClick to view larger

Figure 2. Regulation of eIF2B by phosphorylation of eIF2.

Under stress conditions, such as endoplasmic reticulum stress or lack of essential amino acids, the eIF2α kinases PERK and GCN2, respectively, are activated. The resulting phosphorylation of eIF2 on its α subunit creates eIF2(αP) which competitively inhibits eIF2B, so that eIF2 builds up in its inactive GDP-bound form. This slows down general translation initiation but actually enhances the translation of a small subset of mRNAs which have upstream open reading frames (uORFs) in their 5’-UTRs. One such mRNA encodes ATF4, activating transcription factor 4, which promotes expression of a range of genes involved in coping with cell stress.

This increase is because those mRNAs contain short “open reading frames” upstream (uORF) of the main coding region. When eIF2⋅GTP⋅Met-tRNAMeti levels are high, at least some of these so-called uORFs are translated, preventing ribosomes from reaching the start of the main coding region. On the other hand, when eIF2B is inhibited and eIF2⋅GTP⋅Met-tRNAMeti levels are low, translation of the uORFs is impaired and the main coding region is translated more frequently (Pavitt, 2005). Thus, production of the corresponding proteins is actually enhanced when eIF2B activity is impaired, for example, when eIF2 is phosphorylated. In mammals, the most widely studied example of such an mRNA encodes activating transcription factor 4 (ATF4). This provides a clear example of the ability of translational control to regulate multiple genes; in the case of ATF4, the effects are widened by the fact that some genes controlled by ATF4 are themselves transcription factors—for instance, CHOP. Analyses of postmortem brain tissue from VWM patients shows increased expression of ATF4 and CHOP in white matter glia (van der Voorn et al., 2005; van Kollenburg et al., 2006).

Recent studies on yeast by Pavitt and colleagues have revealed several unusual features of the eIF2B/eIF2 GEF system (Jennings, Kershaw, Adomavicius, & Pavitt, 2017). For example, eIF2B has similar affinity for eIF2 whether bound to GDP or GTP; one implication of this is that eIF2B and Met-tRNAMeti compete with each other for binding to eIF2⋅GTP. Formation of ternary complexes depends on Met-tRNAMeti “winning” this competition. In contrast, by binding eIF2, eIF5 can bind and stabilize ternary complexes. When eIF2 is phosphorylated on its α subunit, its increased affinity for eIF2B means that binding of eIF2 to eIF5 or Met-tRNAMeti is outcompeted by eIF2B, which will impair eIF2’s ability to bind Met-tRNAMeti and form ternary complexes. This provides an additional “brake” mechanism to slow down translation initiation (Hronova & Valasek, 2017). Again, it will be important to investigate whether this also applies to the mammalian factors.

It is possible that some VWM mutations affect this delicate balance of interaction between eIF2B and eIF2; indeed, some mutations are already known to do so (Li, Wang, van der Knaap, & Proud, 2004), but effects of VWM mutations on these “new” functions of eIF2B have yet to be tested.

eIF2B Structure in Relation to Function

eIF2B is a multimeric protein; it has five nonidentical types of subunit, α–ε (Cigan, Bushman, Boal, & Hinnebusch, 1993). Two αβγδε heteropentamers associate to form a heterodecamer (Gordiyenko et al., 2014; Wortham, Martinez, Gordiyenko, Robinson, & Proud, 2014) (Figure 3). However, the data for the yeast and human complexes show considerable disparities. In particular, the data for the human complex indicate an important role for α–α dimers in stabilizing the decameric complex (Wortham et al., 2014), while in contrast this subunit appears to be dispensable for formation of the yeast (Saccharomyces cerevisiae) protein complex (Gordiyenko et al., 2014). Subsequent determination of the crystal structure of eIF2B from a quite distinct yeast species, Schizosaccharomyces pombe, revealed an α–α dimer (Kashiwagi, Ito, & Yokoyama, 2017; Kashiwagi et al., 2016), similar to the situation for human eIF2B.

Role of Eukaryotic Initiation Factor eIF2B in Vanishing White Matter DiseaseClick to view larger

Figure 3. Three-dimensional structure of eIF2B.

Overall structure of eIF2B. The β, γ, δ and ε subunits are shown, respectively, in pink, green, orange, and cyan. The two α subunits are shown in yellow and olive. The ‘NF’ motifs (which augment eIF2B’s GEF function) and the visible C termini of the ε subunits are shown in red and blue, respectively. The substrate, eIF2, docks into the ‘central cavity’ with GEF function being associated with the distal face (dotted circle), with which the GTP-binding γ-subunit of eIF2 associates. This figure was adapted,with permission, from an illustration in Kashiwagi, Ito, & Yokoyama, 2017.

eIF2Bε contains the catalytic GEF domain toward its C terminus (Gomez, Mohammad, & Pavitt, 2002) (Figure 4). Its N-terminal region shows sequence similarity to eIF2Bγ, with which it associates to form a “catalytic subcomplex,” and to nucleotidyl transferases and acyltransferases (Pavitt, Ramaiah, Kimball, & Hinnebusch, 1998). The three-dimensional structures of these subunits resemble that of adenosine diphosphate-glucose pyrophosphorylase (AGP). The similarity of eIF2B subunits to these types of proteins may have significance for the function, or perhaps more likely the regulation, of the eIF2B complex; but this remains to be investigated. However, the structure of the potential nucleotide binding pocket in eIF2Bγ differs from that in AGP and is likely unable to bind ligand. eIF2Bε is extended at its C terminus compared to eIF2Bγ and contains additional regions that include a HEAT domain and an Asn-Phe (NF) pair that is important for GEF activity (Gomez et al., 2002; Gomez & Pavitt, 2000). This region of eIF2Bε extends out from the main 3D structure of the decameric complex.

Role of Eukaryotic Initiation Factor eIF2B in Vanishing White Matter DiseaseClick to view larger

Figure 4. Layout of individual eIF2B subunits (domains) and the catalytic and regulatory subcomplexes of eIF2B.

Homologous regions are shown in matching colours, and are labelled to show the nucleotidyltransferase and acyltransferase regions of eIF2Bγ and eIF2Bε. The latter contain the so-called I-patch repeats. The catalytic (GEF) domain of eIF2Bε is shown, as are the mutually homologous C-terminal regions of eIF2Bα, βand δ. eIF2Bγ and eIF2Bε together form the catalytic subcomplex, which displays some GEF activity; full activity of mammalian eIF2B requires all five types of subunit. The other three subunits form the regulatory subcomplex.

The α, β, and δ subunits also show mutual homology and similarity to ribose-1,5-bisphosphate isomerase. At least in yeast, they can form a stable “regulatory complex” (Pavitt et al., 1998) (Figure 4). This term reflects the facts that mutations in these subunits can make eIF2B insensitive to phosphorylation of eIF2 and that eIF2Bα is absolutely required for this control. The actual roles of the noncatalytic subunits of eIF2B are not fully understood. Some are reported to bind nucleotides, which could play roles in the catalytic activity or the control of eIF2B (Dholakia, Francis, Haley, & Wahba, 1989; Kuhle, Eulig, & Ficner, 2015), while others (e.g., eIF2Bγ [Jennings & Pavitt, 2014; Jennings, Zhou, Mohammad-Qureshi, Bennett, & Pavitt, 2013]) likely contribute to eIF2B’s GDF function.

The effects of VWM mutations on eIF2B’s GEF activity have been tested in a number of studies. Two approaches have been used. In the first, wild-type (WT) or mutated eIF2B subunits were overexpressed in human cell lines, and their activity and certain other properties were then studied (Li et al., 2004; Liu et al., 2011; Matsukawa et al., 2011; Wortham & Proud, 2015). In the second, GEF activity was measured in lysates of cells from VWM patients (Fogli et al., 2004b; Horzinski et al., 2009; Liu et al., 2011). In many cases, eIF2B GEF activity was found to be decreased in both types of test systems. However, in some cases, it was unchanged or actually higher than for WT eIF2B. For example, the first kind of study showed that the A391D mutation in eIF2Bδ and the Y495C change in eIF2Bε had little or no effect on GEF activity or the integrity of eIF2B α–ε pentamers, even though they are associated with very severe disease (Liu et al., 2011). This result was confirmed with experiments that assessed GEF activity of endogenous eIF2B in lysates from lymphoblasts from patients homozygous for these mutations. The authors of this study concluded that such mutations may affect (at that time, unknown) additional functions of eIF2B. The exciting discovery that eIF2B exerts GDF function raises the possibility that some VWM mutations may affect this property of eIF2B. A391 lies at the β/δ interface (Kashiwagi et al., 2017); mutating it to Asp does not impair the overall interaction between the subunits (Liu et al., 2011) but might affect their geometry, thereby affecting the function of eIF2B. It should be noted that another study, where the A391 mutation was “knocked in” to the chromosomal copy of the EIF2B4 gene in Chinese hamster ovary cells, reported diminished GEF activity (Sekine et al., 2016). These authors introduced discriminatory silent mutations (introduced to check the efficacy of their CRISPR/Cas9 genome editing procedure), which in a human setting are predicted to affect mRNA splicing. It is possible that levels of eIF2Bδ mRNA and protein were reduced. This could well contribute to the observed phenotype but was not investigated.

As noted, the initial studies of the effects of VWM mutations in EIF2B2, EIF2B3, and EIF2B5 on the function of eIF2B indicated that all the mutations tested reduced GEF activity, albeit to differing extents (Li et al., 2004; Liu et al., 2011; Matsukawa et al., 2011). Some mutations also affected the association between eIF2B subunits or increased association with phosphorylated eIF2. Two implications of this were that (1) restoring eIF2B GEF activity might provide potential therapies and (2) measuring GEF activity in lysates of cells from potential VWM patients could provide a diagnostic tool (Fogli et al., 2004b; Horzinski et al., 2009). The subsequent discovery that some mutations do not cause reduced GEF activity (or may even increase it [Liu et al., 2011; Wortham & Proud, 2015]) promoted a revision of the earlier conclusions for the diagnosis or treatment of VWM.

Given that the effects of VWM mutations on the GEF activity of eIF2B differ so markedly (changes are not even evident for some mutations), a key question is, do all VWM mutations affect the formation of ternary complexes (which is the key parameter as far as translation initiation is concerned)? As eIF2B exerts other effects on eIF2 in addition to acting as its GEF (i.e., acting as a GDI to displace eIF5 and competing with Met-tRNAMeti), it is indeed possible that even mutations that do not reduce GEF activity still impair ternary complex levels by affecting other properties of eIF2B. It is important to devise ways to test this possibility, especially in cells and tissues.

Data from structural studies (Figure 3), first on the β and δ subunits of eIF2B (Kuhle et al., 2015) and then on the entire complex (Kashiwagi et al., 2016, 2017) have provided crucial insights into both the mechanism of action and regulation of eIF2B and the potential basis for the effects of a number of VWM mutations. Here, we will focus primarily on the implications of these studies for our understanding of VWM disease.

The crystal structure of the eIF2B complex confirms it is a heterodecamer (Kashiwagi et al., 2016, 2017) and shows that the α2β2δ2 hexamer is flanked by the two γε dimeric catalytic subcomplexes (Figure 3). Many VWM mutations are located at the subunit interfaces, suggesting that VWM frequently involves defects in the proper assembly of the decameric complex and/or correct orientation of its component subunits.

Cross-linking data suggest that eIF2 (specifically its GTP-binding γ subunit) binds to the distal surface of the complex, where eIF2Bγ and eIF2Bε are located (Kashiwagi et al., 2016, 2017). Some cross-links are affected by phosphorylation of eIF2, notably those involving the NF-containing region of eIF2Bε. The NF domain is important for catalysis of nucleotide exchange, and a number of VWM mutations are located in this region.

In contrast, other cross-links which do not involve the NF motif, are not affected by the phosphorylation status of eIF2. Indeed, eIF2 also binds to a second site in eIF2B, the so-called central cavity within the α2β2δ2 hexamer. This is true for both phosphorylated and nonphosphorylated eIF2, but phosphorylated eIF2 interacts with higher affinity. Docking studies (Kashiwagi et al., 2016) indicate that eIF2 cannot readily bind at the same time both to the central cavity and to the NF domain. This provides a potential explanation for the inhibitory effect of eIF2 phosphorylation on the GEF function of eIF2B; because phosphorylated eIF2 binds better to the central cavity, it is less able to access the NF domain, which is important for catalysis of the GEF reaction (Figure 3). Interestingly, however, few VWM mutations map to the central cavity region. This indicates that effects on the nonproductive interaction between this feature and eIF2 are not a major cause of the defects caused by VWM mutations (Kashiwagi et al., 2016, 2017). Unexpectedly, the cavity is not positively charged, as one might have expected given its greater affinity for phosphorylated eIF2. The basis of the higher affinity of P-eIF2 for the cavity, and thus the ability of P-eIF2 to inhibit eIF2B function, therefore remains obscure. The high affinity of P-eIF2 for eIF2B means that, in cells where eIF2 levels substantially exceed those of eIF2B, even low levels of P-eIF2 cause substantial inhibition of eIF2B activity by preventing eIF2B from acting on nonphosphorylated eIF2.

However, several VWM mutations map to the same regions of eIF2B as the cross-links that are insensitive to eIF2 phosphorylation. This may indicate that such mutations affect the GDF function of eIF2B, which is also not affected by phosphorylation of eIF2. Mutational studies show that distinct residues in eIF2B (e.g., in eIF2Bγ) are involved in its GEF and GDF functions (Jennings & Pavitt, 2014).

The new structural models help us rationalize the effects of a number of VWM mutations. As Kuhle et al. (2015) point out, many lie close to the interface between the β and δ subunits in their mutual dimer. Furthermore, the correct association between β and δ is needed to create the binding site for the ε subunit (Kashiwagi et al., 2016). Indeed, other VWM mutations are at exposed residues in a hydrophobic surface of eIF2Bδ and are likely involved in protein–protein interactions (Bogorad et al., 2014).

Some mutations that affect residues at or close to the β/δ interface have been tested and shown to impair binding between the subunits (e.g., R357W in eIF2Bδ and V316D and P291S in eIF2Bβ [Li et al., 2004; Liu et al., 2011]). However, another mutation at the same interface did not impair association between subunits (A391D [Liu et al., 2011]). The V183F mutation in eIF2Bα lies at the α–α dimerization interface and impairs the integrity of the whole complex (Wortham & Proud, 2015). R225Q in eIF2Bγ lies on the distal surface of the complex, which interacts with eIF2; and, consistent with this, eIF2 binds less well to complexes containing this mutation (Liu et al., 2011).

Thus, functional consequences of mutations at residues involved in the interactions between subunits in GEF activity and GDF function all seem likely to play roles in the pathogenesis of VWM. However, some VWM mutations are in residues that are exposed on the surface of the protein that are not associated with known functions of eIF2B (Kashiwagi et al., 2016). This raises the possibility that these mutations affect roles of eIF2B that are not yet understood.

Mouse Models for VWM

In 2010, the first transgenic mouse model for VWM was described (Geva et al., 2010). This mouse model is homozygous for the R132H (R136H in human) mutation in eIF2Bε, which is associated with a classical phenotype in patients (Kantor et al., 2005). The mutation affected eIF2 GDP release in brain lysates by approximately 20% (Kantor et al., 2005). The effects of this mutation on the mouse phenotype are subtle and do not include a reduction in life span. The mice show some motor problems on rotarod and open field tests. Their body weight and body fat content are reduced. Brain analyses indicated a high proportion of small-caliber axons, reduced myelin, an increased number of OPCs, but reduced numbers of astrocytes at 3 weeks of age in the internal capsule white matter structure (Geva et al., 2010). At an age of 4 months, the OPC and astrocyte numbers had normalized, although the mice still performed less well at the rotarod than WT mice. The authors concluded that the R132H mutation negatively affected normal brain white matter development. Myelin recovery in the eIF2Bε[R132H] mice was impaired after cuprizone-induced demyelination as well, indicating that remyelination was also affected by the mutation. Follow-up studies with the R132H mouse model showed time-dependent changes in expression of mRNAs with various functions (Marom, Ulitsky, Cabilly, Shamir, & Elroy-Stein, 2011). This genome-wide analysis was not informative on the mechanism by which that R132H mutation altered the levels of expression of these mRNAs.

A recent proteomic study on cerebrum brain samples of the eIF2Bε[R132H] mice revealed a difference in expression of mitochondrial proteins as well as proteins important for protein synthesis and degradation via he proteasome (Gat-Viks, Geiger, Barbi, Raini, & Elroy-Stein, 2015). Indeed, a subtle reduction in 20S proteasome activity was reported in brain lysates from 21-day-old eIF2Bε[R132H] mice. At this age, the white matter may not have become fully myelinated (Baumann & Pham-Dinh, 2001). It would be informative to investigate if the reduction in 20S proteasome activity is due to the developmental difference or more directly to differences in eIF2Bε activity. The 20S proteasome function would have to be measured at different ages and compared to the gene expression differences at these ages.

Most recently, Raini et al. (2017) reported that astrocytes from the R132H VWM knock-in mouse model show an imbalance in mitochondrial proteins involved in oxidative phosphorylation, impaired mitochondrial function (i.e., decreased oxygen consumption), and (presumably as a compensatory effect) enhanced mitochondrial biogenesis. Since almost all mitochondrial proteins are actually made in the cytoplasm, these effects may be a consequence of dysregulation of protein synthesis due to altered eIF2B function.

Another approach for generating a mouse model for VWM has been reported, in which eIF2B activity was inhibited in oligodendrocytes by activation of the eIF2α kinase protein kinase RNA-like endoplasmic reticulum kinase (PERK) (Lin et al., 2008, 2014). Mice were made in which PERK was activated in oligodendrocytes in a cell-autonomous manner. PERK was activated in young and adult animals; especially oligodendrocytes in young mice appeared vulnerable to PERK activation. When activated at postnatal day 10, PERK in oligodendrocytes induced a tremoring phenotype as early as 4 days; and the mice died approximately 14 days later (Lin et al., 2014). When PERK was activated at postnatal day 28, tremor and death were not observed for a further 2 weeks. The young mice showed modest hypomyelination in the brain, whereas the adult mice did not. The astrocytes were not investigated in this model, so at the moment it is unclear if these cells functioned normally or not. It is difficult to interpret the results of the study: the data show that activation of PERK signaling is detrimental for young, but not adult, oligodendrocytes. It remains unclear whether eIF2B inhibition is the sole factor for determining tremor and hypomyelination. Also, the approach for activating PERK in the oligodendrocytes did not activate the stress sensors ATF6 and IRE1α in the endoplasmic reticulum, which are usually activated all together in response to endoplasmic reticulum stress (Walter & Ron, 2011). The chaperone protein BIP was not induced by PERK activation (Lin et al., 2008, 2014). Increased BIP expression is a result of ATF6 activation and constitutes an important negative feedback loop, which restores the stress sensors PERK, ATF6, and IRE1a to normal levels (Walter & Ron, 2011). As such, the model is perhaps best considered as a PERK “out-of-control” model rather than a model for VWM, in which diverse and often subtle mutations in eIF2B result in white matter pathology in young and adult patients.

Recently, additional transgenic knock-in mouse models have been described for VWM: the Eif2b4R484W/R484W (2b4ho) and the Eif2b5R191H/R191H (2b5ho) lines (Dooves et al., 2016a). The 2b4ho mice are homozygous for R484W in eIF2Bδ and the 2b5ho mice, for R191H in eIF2Bε. The equivalent mutations in human eIF2Bδ and eIF2Bε (R483W and R195H, respectively) are associated with a severe phenotype (Fogli et al., 2004a; van der Knaap et al., 2003). A variety of analyses show that the mouse models are representative of the human disease. For example, the 2b4ho as well as the 2b5ho mice develop ataxia as visualized by a footprint assay with a slight variation in age at onset (7 and 5 months, respectively) (Dooves et al., 2016a). Both mouse models have a reduced life span, and again the humane endpoint varies between the two genotypes (19 months for 2b4ho and 8 months for 2b5ho mice) (Dooves et al., 2016a). Crossbreeding of the 2b4ho and 2b5ho mice resulted in more severely affected genotypes: the average survival of double homozygous animals (2b4ho2b5ho) was 3 weeks, and that of double transgenic animals heterozygous for either mutation and homozygous for the other (2b42b5he/ho) was 4 months.

The white matter in all these mutant mice shows perturbed myelination and progressive myelin vacuolation (Dooves et al., 2016a). Myelin vacuolation was reported to correlate with disease severity, genotype, and disease stage. The progressive nature of the vacuolation indicates deficient myelin formation, maturation, and maintenance. Although white matter damage was pronounced, formation of scar tissue (reactive gliosis) was not observed. This suggests that astrocytes in the white matter are unable to fulfill their normal reactive function to tissue damage. As in patients, increased numbers of astrocytes double-positive for nestin and glial fibrillary acidic protein δ (GFAPδ) were found in white matter of 2b5ho mice, without increased GFAPα expression, indicating an immature state (Dooves et al., 2016a). These double-positive astrocytes have an abnormal morphology with thick, coarse processes (Dooves et al., 2016a). The astrocyte morphology in the gray matter appears normal. The number of mature oligodendrocytes was also decreased, and the number of OPCs was increased in 2b4ho2b5ho and 2b42b5he/ho mice (Dooves et al., 2016a).

A co-culture assay of astrocytes and OPCs was developed in which OPC maturation was measured by counting cells that express mature markers (myelin basic protein [MBP] and myelin oligodendrocyte glycoprotein [MOG]). Co-cultures with mutant (2b4ho or 2b5ho) cells yielded reduced MBP- and MOG-positive cells when compared to WT co-cultures, indicative of impaired OPC maturation in mutant cells (Dooves et al., 2016a). This block in OPC maturation was not observed when mutant OPCs were co-cultured with WT astrocytes, indicating that the OPC maturation program itself is not directly affected by the mutation in eIF2B. Also, co-cultures with 2b4ho astrocytes and WT OPCs showed that 2b4ho astrocytes reduced the number of MBP- and MOG-positive cells (Dooves et al., 2016a). This finding strongly suggests that mutant astrocytes are compromised in supporting OPC maturation and substantially affected by the mutation in eIF2B.

The block in OPC maturation in vitro appears to be due to one or more inhibitory factors secreted by 2b4ho astrocytes as cell-free, conditioned medium from 2b4ho astrocyte cultures blocked OPC maturation in co-cultures of WT astrocytes and WT OPCs (Dooves et al., 2016a). This observation suggests that mutant astrocytes produce a factor that inhibits OPC maturation rather than being deficient in secreting one or more factors that promote OPC maturation. The secreted inhibitory factor(s) has not yet been identified, and further investigation is obviously needed to identify this factor.

Future Perspective

VWM is a devastating disease for which there is so far no cure or even treatment. Recent research into the structure and functions of eIF2B as well as studies using VWM disease models and patient materials have provided significant new insights into some aspects of eIF2B in relation to the disease.

A major remaining question concerns why mutations in genes for a protein which is crucial for protein synthesis in all cells should cause a neurological disease. Some patients do show defects in other tissues, but their disease is still primarily neurological. At this stage, one can only speculate about the possible reasons: perhaps the eIF2/eIF2B ratio or the levels of eIF5 or Met-tRNA are particularly low (or high) in glial cells, perhaps eIF2B exerts a novel and cell-specific unknown function in glial cells, or eIF2 and levels of ternary complexes affect the translation of a glia-specific mRNA(s) important for glial function. It is probably not due to a high rate of protein synthesis in glial cells since other cells, such as fast dividing cells (e.g., immune cells or cells in the jejunum) or pancreatic cells, producing many secreted factors with high protein synthetic rates are not affected.

For mechanistically unknown reasons, astrocytes appear to be most affected by the pathogenic mutations in eIF2B. Targets that are potentially interesting for therapy development may be involved in astrocyte proteasome or mitochondrial function (Gat-Viks et al., 2015; Raini et al., 2017). Targeting proteasome or mitochondria, however, have the risk of being too general and affecting other cell types as well; thus, drugs lacking astrocyte specificity may ultimately result in unwanted side effects. Alternatively, it may be of interest to identify the factor(s) secreted by mutant astrocytes that affects OPC maturation in co-culture test systems. Perhaps these factors are specifically secreted by astrocytes, providing a potentially elegant target for cell-specific treatment. Alternatively, diseased astrocytes could be replaced by healthy astrocytes lacking eIF2B mutations (Dooves, van der Knaap, & Heine, 2016b). As the inhibitory factor may persist in the extracellular matrix of the affected brain (Dooves et al., 2016b), its removal is likely to be necessary for successful cell replacement therapy.

Identifying the Secreted Factors that Inhibit OPC Maturation Is Not an Easy Task

Currently we do not know what the relevant factors secreted from astrocytes are. We may assume that they are proteins as eIF2B mutations likely affect translation of specific mRNAs. However, we cannot rule out that the factors are of a different nature. Their abnormal synthesis and secretion may, for instance, be the result of abnormal expression of the enzymes involved in their metabolism. Insight into the mRNA and protein expression profile of WT and mutant astrocytes is therefore needed first, and using an open screen (“omics”) approach is likely to be most insightful. Polysomal profiling or proteomics screening of mouse brain tissue or astrocyte cultures from WT and mutant mice would be insightful. Obtaining astrocyte-specific protein expression profiling can be done by creating mutant mouse models in which the ribosomes are tagged with, for instance, a hemagglutinin tag (Sanz et al., 2009).

Subsequently, the secreted factors in astrocyte-conditioned medium can be further investigated to identify the OPC maturation inhibitory factor. To first gain insight into the nature of the inhibitory factor, the astrocyte-conditioned medium can be subjected to proteinase K treatment. If MBP and MOG expression is restored upon protein K treatment, the factor is likely a polypeptide. Proteomics may identify the protein composition of the inhibitory medium. Secretome composition of other cell types has successfully been determined with azidohomoalanine (AHA) labeling, which allows purification of the secreted proteins from culture medium factor (e.g., bovine serum albumin and growth factors in the medium) (Eichelbaum, Winter, Berriel Diaz, Herzig, & Krijgsveld, 2012). Proteins synthesized by WT and mutant astrocytes can be labeled with AHA in combination with pulsed stable isotope labeling of new proteins to strengthen quantification (Howden et al., 2013; Huo et al., 2012). Of course, it should first be determined whether the labeling technique does not interfere with the OPC inhibition function. The labeled samples can be fractionated on the basis of molecular weight, as has been reported for other test systems (Coyle, 1995; Polazzi et al., 2009; Sugino et al., 2016). The fractions that retain the inhibitory capacity can be subjected to mass spectrometry. If the factor is not a protein but is, for example, a metabolite, exosome, or fatty acid derivative, a metabolomics, rather than a proteomic, type of screening may be warranted. Eventually compounds can be searched for that interfere with the synthesis or activity of the inhibitory factor. One should keep in mind that successful targeting of the inhibitory factor may only provide relief of the abnormal OPC maturation. Other affected astrocytic functions, for example, reactive astrogliosis, may not be repaired by this treatment. VWM may therefore not be completely solved by such a specific therapy.

Discovery of a Small Molecule that Enhances eIF2B Activity

An approach that possibly targets all disease aspects lies in direct targeting of eIF2B enzyme activity. Recently, the small molecule integrated stress response (ISR) inhibitor was identified in a phenotypic, cell-based screen that scored absence of ATF4 induction after ISR activation by thapsigargin and tunicamycin (Sidrauski et al., 2013). In a recent genetic screen, mutations in eIF2Bδ were identified that made cells unresponsive to ISR inhibitor (ISRIB), indicating that eIF2Bδ is the target of ISRIB (Sekine et al., 2015). ISRIB enhanced eIF2B GEF activity in vitro and stabilized the eIF2B complex in biochemical analyses (Sidrauski et al., 2015). ISRIB prevents and resolves an already activated stress response. ISRIB has been tested in mice: it possesses good pharmacological properties and crosses the blood–brain barrier, although insolubility of the substance may be a problem to reach an efficacious dose.

A recent study assessed ISRIB’s impact in prion-infected mice (Halliday et al., 2015). In this neurodegenerative disease model, eIF2B in the brain is inhibited by infectious prion particles. ISRIB prevented prion-related disease symptoms and neurodegeneration. ISRIB did not reduce the prion accumulation or eIF2 phosphorylation; however, prion-related disease symptoms were greatly diminished, and ATF4 expression in brain was reduced. This result is expected considering ISRIB’s action on eIF2B (downstream of eIF2 phosphorylation and upstream of ATF4 induction). Prolonged treatment resulted in weight loss in the treated mice (Halliday et al., 2015). It is not clear if this weight loss was due to the ongoing prion infection, to ISRIB, or to their combination. Overall, it is reasonable to test ISRIB in VWM mice. One caveat is the possibility that (some) mutations in eIF2B that cause VWM may interfere with ISRIB’s action. Furthermore, given the very different biochemical consequences of VWM mutations on eIF2B function, ISRIB may not prevent the consequences of all such mutations. If all mutations do affect ternary complex levels, trying to restore such levels may represent a possible therapeutic strategy (although it not clear how to achieve this in vivo). Also, if mutations in eIF2B interfere in a yet undiscovered astrocyte- or oligodendrocyte-specific function that is unrelated to the GEF activity, ISRIB may not be useful. Further research is needed to investigate ISRIB’s therapeutic potential in VWM. In the case that ISRIB is not suitable for therapy in patients, other ISR modulators may be of interest for therapy development (Smith & Mallucci, 2016; Sundaram, Lee, & Shenolikar, 2017).

Acknowledgments

Work on eIF2B in C.G.P.’s lab was supported by the UK Biotechnology & Biological Sciences Research Council.

References

Baumann, N., & Pham-Dinh, D. (2001). Biology of oligodendrocyte and myelin in the mammalian central nervous system. Physiological Reviews, 81(2), 871–927.Find this resource:

Black, D. N., Booth, F., Watters, G. V., Andeermann, E., Dumont, C., Halliday, W. C., . . . O’gorman, G. (1988). Leukoencephalopathy among native Indian infants in northern Quebec and Manitoba. Annals of Neurology, 24(4), 490–496.Find this resource:

Bogorad, A. M., Xia, B., Sandor, D. G., Mamonov, A. B., Cafarella, T. R., Jehle, S., . . . Marintchev, A. (2014). Insights into the architecture of the eIF2Balpha/beta/delta regulatory subcomplex. Biochemistry, 53(21), 3432–3445.Find this resource:

Bugiani, M., Boor, I., Powers, J. M., Scheper, G. C., & van der Knaap, M. S. (2010). Leukoencephalopathy with vanishing white matter: A review. Journal of Neuropathology and Experimental Neurology, 69(10), 987–996.Find this resource:

Bugiani, M., Boor, I., van Kollenburg, B., Postma, N., Polder, E., van Berkel, C., . . . van der Knaap, M. S. (2011). Defective glial maturation in vanishing white matter disease. Journal of Neuropathology and Experimental Neurology, 70(1), 69–82.Find this resource:

Cigan, A. M., Bushman, J. L., Boal, T. R., & Hinnebusch, A. G. (1993). A protein complex of translational regulators of GCN4 mRNA is the guanine nucleotide-exchange factor for translation initiation factor 2 in yeast. Proceedings of the National Academy of Sciences of the United States of America, 90(11), 5350–5354.Find this resource:

Coyle, D. E. (1995). Adaptation of C6 glioma cells to serum-free conditions leads to the expression of a mixed astrocyte–oligodendrocyte phenotype and increased production of neurite-promoting activity. Journal of Neuroscience Research, 41(3), 374–385.Find this resource:

Dholakia, J. N., Francis, B. R., Haley, B. E., & Wahba, A. J. (1989). Photoaffinity labeling of the rabbit reticulocyte guanine nucleotide exchange factor and eukaryotic initiation factor 2 with 8-azidopurine nucleotides. Identification of GTP- and ATP-binding domains. Journal of Biological Chemistry, 264, 20638–20642.Find this resource:

Dooves, S., Bugiani, M., Postman, N. L., Polder, E., Land, N., Horan, S. T., . . . van der Knaap, M. S. (2016a). Astrocytes are central in the pathomechanisms of vanishing white matter. Journal of Clinical Investigation, 126(4), 1512–1524.Find this resource:

Dooves, S., van der Knaap, M. S., & Heine, V. M. (2016b). Stem cell therapy for white matter disorders: don't forget the microenvironment! Journal of Inherited Metabolic Disease, 39(4), 513–518.Find this resource:

Eichelbaum, K., Winter, M., Berriel Diaz, M., Herzig, S., & Krijgsveld, J. (2012). Selective enrichment of newly synthesized proteins for quantitative secretome analysis. Nature Biotechnology, 30(10), 984–990.Find this resource:

Fogli, A., & Boespflug-Tanguy, O. (2006). The large spectrum of eIF2B-related diseases. Biochemical Society Transactions, 34(Pt 1), 22–29.Find this resource:

Fogli, A., Rodriguez, D., Eymard-Pierre, E., Bouhour, F., Labauge, P., Meaney, B.F., . . . Boespflug-Tanguy, O. (2003). Ovarian failure related to eukaryotic initiation factor 2B mutations. American Journal of Human Genetics, 72(6), 1544–1550.Find this resource:

Fogli, A., Schiffmann, R., Bertini, E., Ughetto, S., Combes, P., Eymard-Pierre, E., . . . Boespflug-Tanguy, O. (2004a). The effect of genotype on the natural history of eIF2B-related leukodystrophies. Neurology, 62(9), 1509–1517.Find this resource:

Fogli, A., Schiffmann, R., Hugendubler, L., Combes, P., Bertini, E., Rodriguez, D., . . . Boespflug-Tanguy, O. (2004b). Decreased guanine nucleotide exchange factor activity in eIF2B-mutated patients. European Journal of Human Genetics, 12, 561–566.Find this resource:

Fogli, A., Wong, K., Eymard-Pierre, E., Wenger, J., Bouffard, J. P., Goldin, E., . . . Schiffmann, R. (2002). Cree leukoencephalopathy and CACH/VWM disease are allelic at the EIF2B5 locus. Annals of Neurology, 52(4), 506–510.Find this resource:

Garden, G. A., & Campbell, B. M. (2016). Glial biomarkers in human central nervous system disease. Glia, 64(10), 1755–1771.Find this resource:

Gat-Viks, I., Geiger, T., Barbi, M., Raini, G., & Elroy-Stein, O. (2015). Proteomics-level analysis of myelin formation and regeneration in a mouse model for vanishing white matter disease. Journal of Neurochemistry, 134(3), 513–526.Find this resource:

Geva, M., Cabilly, Y., Assaf, Y., Mindroul, N., Marom, L., Raini, G., . . . Elroy-Stein, O. (2010). A mouse model for eukaryotic translation initiation factor 2B-leucodystrophy reveals abnormal development of brain white matter. Brain, 133, 2448–2461.Find this resource:

Gomez, E., Mohammad, S. S., & Pavitt, G. D. (2002). Characterization of the minimal catalytic domain within eIF2B: The guanine-nucleotide exchange factor for translation initiation. EMBO Journal, 21(19), 5292–5301.Find this resource:

Gomez, E., & Pavitt, G. D. (2000). Identification of domains and residues within translation initiation factor eIF2Bepsilon required for guanine nucleotide-exchange reveals a novel activation function promoted by eIF2B complex formation. Molecular and Cellular Biology, 20, 3965–3976.Find this resource:

Gordiyenko, Y., Schmidt, C., Jennings, M. D., Matak-Vinkovic, D., Pavitt, G. D., & Robinson, C. V. (2014). eIF2B is a decameric guanine nucleotide exchange factor with a γ2ε2 tetrameric core. Nature Communications, 5, 3902.Find this resource:

Halliday, M., Radford, H., Sekine, Y., Moreno, J., Verity, N., le Quesne, J., . . . Mallucci, G. R. (2015). Partial restoration of protein synthesis rates by the small molecule ISRIB prevents neurodegeneration without pancreatic toxicity. Cell Death & Disease, 6, e1672.Find this resource:

Hanefeld, F., Holzbach, U., Kruse, B., Wilichowski, E., Christen, H. J., & Frahm, J. (1993). Diffuse white matter disease in three children: An encephalopathy with unique features on magnetic resonance imaging and proton magnetic resonance spectroscopy. Neuropediatrics, 24(5), 244–248.Find this resource:

Hinnebusch, A. G. (1996). Translational control of GCN4: Gene-specific regulation by phosphorylation of elF2. In J. W. B. Hershey, M. B. Matthews, & N. Sonenberg (Eds.), Translational Control (pp. 199–244). CSH Monographs 30. Cold Spring Harbor, NY: Cold Spring Harbor Laboratories.Find this resource:

Hirrlinger, J., & Nave, K. A. (2014). Adapting brain metabolism to myelination and long-range signal transduction. Glia, 62(11), 1749–1761.Find this resource:

Horzinski, L., Huyghe, A., Cardoso, M. C., Gonthier, C., Ouchchane, L., Schiffmann, R., . . . Fogli, A. (2009). Eukaryotic initiation factor 2B (eIF2B) GEF activity as a diagnostic tool for EIF2B-related disorders. PLoS One, 4(12), e8318.Find this resource:

Howden, A. J. M., Geoghegan, V., Katsch, K., Efstathiou, G., Bhushan, B., Boutureira, O., . . . Acuto, O. (2013). QuaNCAT: Quantitating proteome dynamics in primary cells. Nature Methods, 10(4), 343–346.Find this resource:

Hronova, V., & Valasek, L. S. (2017). An emergency brake for protein synthesis. Elife, 6, e27085.Find this resource:

Huo, Y., Iadevaia, V., Yao, Z., Kelly, I., Cosulich, S., Guichard, S., . . . Proud, C. G. (2012). Stable isotope-labelling analysis of the impact of inhibition of the mammalian target of rapamycin on protein synthesis. Biochemical Journal, 444(1), 141–151.Find this resource:

Jennings, M. D., Kershaw, C. J., Adomavicius, T., & Pavitt, G. D. (2017). Fail-safe control of translation initiation by dissociation of eIF2alpha phosphorylated ternary complexes. Elife, 6, e24542.Find this resource:

Jennings, M. D., Kershaw, C. J., White, C., Hoyle, D., Richardson, J. P., Costello, J. L., . . . Pavitt, G. D. (2016). eIF2beta is critical for eIF5-mediated GDP-dissociation inhibitor activity and translational control. Nucleic Acids Research, 44(20), 9698–9709.Find this resource:

Jennings, M. D., & Pavitt, G. D. (2010). eIF5 has GDI activity necessary for translational control by eIF2 phosphorylation. Nature, 465(7296), 378–381.Find this resource:

Jennings, M. D., & Pavitt, G. D. (2014). A new function and complexity for protein translation initiation factor eIF2B. Cell Cycle, 13(17), 2660–2665.Find this resource:

Jennings, M. D., Zhou, Y., Mohammad-Qureshi, S. S., Bennett, D., & Pavitt, G. D. (2013). eIF2B promotes eIF5 dissociation from eIF2*GDP to facilitate guanine nucleotide exchange for translation initiation. Genes and Development, 27(24), 2696–2707.Find this resource:

Kantor, L., Harding, H. P., Ron, D., Shiffmann, R., Kaneski, C. R., Kimball, S. R., & Elroy-Stein, O. (2005). Heightened stress response in primary fibroblasts expressing mutant eIF2B genes from CACH/VWM leukodystrophy patients. Human Genetics, 118(1), 99–106.Find this resource:

Kashiwagi, K., Ito, T., & Yokoyama, S. (2017). Crystal structure of eIF2B and insights into eIF2–eIF2B interactions. FEBS Journal, 284, 868–874.Find this resource:

Kashiwagi, K., Takahashi, M., Nishimoto, M., Hiyama, T. B., Higo, T., Umehara, T., . . . Yokoyama, S. (2016). Crystal structure of eukaryotic translation initiation factor 2B. Nature, 531(7592), 122–125.Find this resource:

Kiray, H., Lindsay, S. L., Hosseinzadeh, S., & Barnett, S. C. (2016). The multifaceted role of astrocytes in regulating myelination. Experimental Neurology, 283(Pt. B), 541–549.Find this resource:

Kuhle, B., Eulig, N. K., & Ficner, R. (2015). Architecture of the eIF2B regulatory subcomplex and its implications for the regulation of guanine nucleotide exchange on eIF2. Nucleic Acids Research, 43, 9994–10014.Find this resource:

Leegwater, P. A., Könst, A. A., Kuyt, B., Sandkuijl, L. A., Naidu, S., Oudejans, C. B., . . . van der Knapp, M. S. (1999). The gene for leukoencephalopathy with vanishing white matter is located on chromosome 3q27. American Journal of Human Genetics, 65(3), 728–734.Find this resource:

Leegwater, P. A., Vermeulen, G., Könst, A. A., Naidu, S., Mulders, J., Visser, A., . . . van der Knapp, M. S. (2001). Subunits of the translation initiation factor eIF2B are mutant in leukoencephalopathy with vanishing white matter. Nature Genetics, 29(4), 383–388.Find this resource:

Li, W., Wang, X., van der Knaap, M. S., & Proud, C. G. (2004). Mutations linked to leukoencephalopathy with vanishing white matter impair the function of the eukaryotic initiation factor 2B complex in diverse ways. Molecular and Cellular Biology, 24(8), 3295–3306.Find this resource:

Lin, W., Kunkler, P. E., Harding, H. P., Ron, D., Kraig, R. P., & Popko, B. (2008). Enhanced integrated stress response promotes myelinating oligodendrocyte survival in response to interferon-gamma. American Journal of Pathology, 173(5), 1508–1517.Find this resource:

Lin, Y., Pang, X., Huang, G., Jamison, S., Fang, J., Harding, H. P., . . . Lin, W. (2014). Impaired eukaryotic translation initiation factor 2B activity specifically in oligodendrocytes reproduces the pathology of vanishing white matter disease in mice. Journal of Neuroscience, 34(36), 12182–12191.Find this resource:

Liu, A. R., van der Lei, H. D., Wang, X., Wortham, N. C., Tang, H., van Berkel, C. G., . . . Proud, C. G. (2011). Severity of vanishing white matter disease does not correlate with deficits in eIF2B activity or the integrity of eIF2B complexes. Human Mutation, 32, 1036–1045.Find this resource:

Liu, R., van der Lei, H. D., Wang, X., Wortham, N. C., Tang, H., van Berkel, C. G., . . . Proud, C. G. (2011). Severity of vanishing white matter disease does not correlate with deficits in eIF2B activity or the integrity of eIF2B complexes. Human Mutation, 32(9), 1036–1045.Find this resource:

Lundgaard, I., Osório, M. J., Kress, B. T., Sanggaard, S., & Nedergaard, M. (2014). White matter astrocytes in health and disease. Neuroscience, 276, 161–173.Find this resource:

Marom, L., Ulitsky, I., Cabilly, Y., Shamir, R., & Elroy-Stein, O. (2011). A point mutation in translation initiation factor eIF2B leads to function—and time-specific changes in brain gene expression. PLoS One, 6(10), e26992.Find this resource:

Matsukawa, T., Wang, X., Liu, R., Wortham, N. C., Onuki, Y., Kubota, A., . . . Tsuji, S. (2011). Adult-onset leukoencephalopathies with vanishing white matter with novel missense mutations in EIF2B2, EIF2B3, and EIF2B5. Neurogenetics, 12, 259–261.Find this resource:

Middeldorp, J., Boer, K., Sluijs, J. A., De Fillippis, L., Encha-Razavi, F., Vescovi, A. L., . . . Hol, E. M. (2010). GFAPdelta in radial glia and subventricular zone progenitors in the developing human cortex. Development, 137(2), 313–321.Find this resource:

Ohlenbusch, A., Henneke, M., Brockmann, K., Goerg, M., Hanefeld, F., Kohlschütter, A., & Gärtner, J. (2005). Identification of ten novel mutations in patients with eIF2B-related disorders. Human Mutation, 25(4), 411.Find this resource:

Paulin, F. E., Campbell, L. E., O’Brien, K., Loughlin, J., & Proud, C. G. (2001). Eukaryotic translation initiation factor 5 (eIF5) acts as a classical GTPase-activator protein. Current Biology, 11(1), 55–59.Find this resource:

Pavitt, G. D. (2005). eIF2B, a mediator of general and gene-specific translational control. Biochemical Society Transactions, 33(Pt. 6), 1487–1492.Find this resource:

Pavitt, G. D., Ramaiah, K. V., Kimball, S. R., & Hinnebusch, A. G. (1998). eIF2 independently binds two distinct eIF2B subcomplexes that catalyse and regulate guanine-nucleotide exchange. Genes & Development, 12, 514–526.Find this resource:

Polazzi, E., Altamira, L. E., Eleuteri, S., Barbaro, R., Casadio, C., Contestabile, A., & Monti, B. (2009). Neuroprotection of microglial conditioned medium on 6-hydroxydopamine-induced neuronal death: Role of transforming growth factor beta-2. Journal of Neurochemistry, 110(2), 545–556.Find this resource:

Pronk, J.C., van Kollenburg, B., Scheper, G. C., & van der Knaap, M. S. (2006). Vanishing white matter disease: A review with focus on its genetics. Mental Retardation and Developmental Disabilities Research Reviews, 12(2), 123–128.Find this resource:

Raini, G., Sharet, R., Herrero, M., Atzmon, A., Shenoy, A., Geiger, T., & Elroy-Stein, O. (2017). Mutant eIF2B leads to impaired mitochondrial oxidative phosphorylation in vanishing white matter disease. Journal of Neurochemistry, 141, 694–707.Find this resource:

Raponi, E., Agenes, F., Delphin, C., Assard, N., Baudier, J., Legraverend, C., & Deloulme, J. C. (2007). S100B expression defines a state in which GFAP-expressing cells lose their neural stem cell potential and acquire a more mature developmental stage. Glia, 55(2), 165–177.Find this resource:

Richardson, J. P., Mohammad, S. S., & Pavitt, G. D. (2004). Mutations causing childhood ataxia with central nervous system hypomyelination reduce eukaryotic initiation factor 2B complex formation and activity. Molecular and Cellular Biology, 24(6), 2352–2363.Find this resource:

Rowlands, A. G., Panniers, R., & Henshaw, E. C. (1988). The catalytic mechanism of guanine nucleotide exchange factor action and competitive inhibition by phosphorylated eukaryotic initiation factor 2. Journal of Biological Chemistry, 263(12), 5526–5533.Find this resource:

Sanz, E., Yang, L., Su, T., Morris, D. R., McKnight, G. S., & Amieux, P. S. (2009). Cell-type-specific isolation of ribosome-associated mRNA from complex tissues. Proceedings of the National Academy of Sciences of the United States of America, 106(33), 13939–13944.Find this resource:

Sekine, Y., Zyryanova, A., Crespillo-Casado, A., Amin-Wetzel, N., Harding, H. P., & Ron, D. (2016). Paradoxical sensitivity to an integrated stress response blocking mutation in vanishing white matter cells. PLoS One, 11(11), e0166278.Find this resource:

Sekine, Y., Zyryanova, A., Crespillo-Casado, A., Fischer, P. M., Harding, H. P., & Ron, D. (2015). Stress responses. Mutations in a translation initiation factor identify the target of a memory-enhancing compound. Science, 348(6238), 1027–1030.Find this resource:

Sidrauski, C., Acosta-Alvear, D., Khoutorsky, A., Vedantham, P., Hearn, B. R., Li, H., . . . Walter, P. (2013). Pharmacological brake-release of mRNA translation enhances cognitive memory. Elife, 2, e00498.Find this resource:

Sidrauski, C., Tsai, J. C., Kampmann, M., Hearn, B. R., Vedantham, P., Jaishankar, P., . . . Walter, P. (2015). Pharmacological dimerization and activation of the exchange factor eIF2B antagonizes the integrated stress response. Elife, 4, e07314.Find this resource:

Smith, H. L., & Mallucci, G. R. (2016). The unfolded protein response: Mechanisms and therapy of neurodegeneration. Brain, 139(Pt. 8), 2113–2121.Find this resource:

Sugino, I. K., Sun, Q., Springer, C., Cheewatrakoolpong, N., Liu, T., Li, H., & Zarbin, M. A. (2016). Two bioactive molecular weight fractions of a conditioned medium enhance RPE cell survival on age-related macular degeneration and aged Bruch's Membrane. Translational Vision Science & Technology, 5(1), 8.Find this resource:

Sundaram, J. R., Lee, I. C., & Shenolikar, S. (2017). Translating protein phosphatase research into treatments for neurodegenerative diseases. Biochemical Society Transactions, 45, 101–112.Find this resource:

van der Knaap, M. S., Barth, P. G., Gabreëls, F. J., Franzoni, E., Begeer, J. H., Stroink, H., . . . Valk, J. (1977). A new leukoencephalopathy with vanishing white matter. Neurology, 48(4), 845–855.Find this resource:

van der Knaap, M. S., Breiter, S. N., Naidu, S., Hart, A. A., & Valk, J. (1999). Defining and categorizing leukoencephalopathies of unknown origin: MR imaging approach. Radiology, 213(1), 121–133.Find this resource:

van der Knaap, M. S., Kamphorst, W., Barth, P. G., Kraaijeeveld, C. L., Gut, E., & Valk, J. (1998). Phenotypic variation in leukoencephalopathy with vanishing white matter. Neurology, 51(2), 540–547.Find this resource:

van der Knaap, M. S., Leegwater, P. A., Könst, A. A., Visser, A., Naidu, S., Oudejans, C. B., . . . Pronk, J. C. (2002). Mutations in each of the five subunits of translation initiation factor eIF2B can cause leukoencephalopathy with vanishing white matter. Annals of Neurology, 51(2), 264–270.Find this resource:

van der Knaap, M. S., Pronk, J. C., & Scheper, G. C. (2006). Vanishing white matter disease. Lancet Neurology, 5(5), 413–423.Find this resource:

van der Knaap, M. S., van Berkel, C. G., Herms, J., van Coster, R., Baethmann, M., Naidu, S., . . . Pronk, J. C. (2003). eIF2B-related disorders: Antenatal onset and involvement of multiple organs. American Journal of Human Genetics, 73(5), 1199–1207.Find this resource:

van der Lei, H. D., van Berkel, C. G., van Wieringen, W. N., Brenner, C., Feigenbaum, A., Mercimek-Mahmutoglu, S., . . . van der Knaap, M. S. (2010). Genotype–phenotype correlation in vanishing white matter disease. Neurology, 75(17), 1555–1559.Find this resource:

van der Voorn, J. P., van Kollenburg, B., Bertrand, G., Van Haren, K., Scheper, G. C., Powers, J. M., & van der Knaap, M. S. (2005). The unfolded protein response in vanishing white matter disease. Journal of Neuropathology and Experimental Neurology, 64(9), 770–775.Find this resource:

Van Haren, K., van der Voorn, J. P., Peterson, D. R., van der Knaap, M. S., & Powers, J. M. (2004). The life and death of oligodendrocytes in vanishing white matter disease. Journal of Neuropathology and Experimental Neurology, 63(6), 618–630.Find this resource:

van Kollenburg, B., van Dijk, J., Garbern, J., Thomas, A. A., Scheper, G. C., Powers, J. M., & van der Knaap, M. S. (2006). Glia-specific activation of all pathways of the unfolded protein response in vanishing white matter disease. Journal of Neuropathology and Experimental Neurology, 65(7), 707–715.Find this resource:

Vermeulen, G., Seidl, R., Mercimek-Mahmutoglu, S., Rotteveel, J. J., Scheper, G. C., & van der Knaap, M. S. (2005). Fright is a provoking factor in vanishing white matter disease. Annals of Neurology, 57(4), 560–563.Find this resource:

Walter, P., & Ron, D. (2011). The unfolded protein response: From stress pathway to homeostatic regulation. Science, 334(6059), 1081–1086.Find this resource:

Wang, X. M., Wortham, N. C., Liu, R., & Proud, C. G. (2012) Identification of residues that underpin interactions within the eukaryotic initiation factor (eIF2) 2B complex. Journal of Biological Chemistry, 287(11), 8263–8274.Find this resource:

Williams, D. D., Price, N. T., Loughlin, A. J., & Proud, C. G. (2001). Characterisation of the mammalian initiation factor eIF2B complex as a GDP-dissociation stimulator protein. Journal of Biological Chemistry, 276(27), 24697–24703.Find this resource:

Wortham, N. C., & Proud, C. G. (2015). Biochemical effects of mutations in the gene encoding the alpha subunit of eukaryotic initiation factor (eIF) 2B associated with vanishing white matter disease. BMC Medical Genetics, 16, 64.Find this resource:

Wortham, N. C., Martinez, M., Gordiyenko, Y., Robinson, C. V., & Proud, C. G. (2014). Analysis of the subunit organization of the eIF2B complex reveals new insights into its structure and regulation. FASEB Journal, 28(5), 2225–2237.Find this resource:

Wu, Y., Pan, Y., Du, L., Wang, J., Gu, Q., Gao, Z., . . . Jiang, Y. (2009). Identification of novel EIF2B mutations in Chinese patients with vanishing white matter disease. Journal of Human Genetics, 54(2), 74–77.Find this resource:

Zhang, H. H., Dai, L., Chen, N., Zang, L., Leng, X., Du, L., . . . Wu, Y. (2015). Fifteen novel EIF2B1-5 mutations identified in Chinese children with leukoencephalopathy with vanishing white matter and a long term follow-up. PLoS One, 10(3), e0118001.Find this resource: