Skip to main content

Identification of repurposable cytoprotective drugs in vanishing white matter disease patient-derived cells



Vanishing white matter disease (VWMD) is a rare disease caused by mutations of the guanine exchange factor eIF2B. VWMD typically presents with juvenile onset, and there are few treatments for the disease. Recent progress in the field has established mitochondrial dysfunction and endoplasmic reticulum (ER) stress to be strongly implicated in observed glial cell pathology. Drug repurposing offers a rapid approach toward translation of therapeutics using already-licensed drugs.


The aim of this study was to use fibroblasts and induced pluripotent stem cell (iPSC)-derived astrocytes from patients bearing the EIF2B5R113H/A403V or EIF2B2G200V/E213G VWMD mutations to identify potential repurposable FDA-approved drugs based on in vitro assays.


Cell viability in the presence or absence of stress was assessed by resazurin reduction activity and assays for mitochondrial membrane potential and oxidative stress by tetramethylrhodamine and dichlorofluorescein-based assays, respectively. Integrated stress response markers, including eIF2α phosphorylation, GADD34 and CHOP were quantified by fluorescent western blot.


Dysregulated GADD34 and CHOP were identified in patient fibroblasts and iPSC-derived astrocytes under induced stress conditions. A drug screen from a 2400 FDA-approved drug library with EIF2B5R113H/A403V VWMD patient fibroblasts identified 113 anti-inflammatory drugs as a major class of hits with cytoprotective effects. A panel of candidate drugs, including berberine, deflazacort, ursodiol, zileuton, guanabenz and Anavex 2–73, and preclinical ISRIB, increased cell survival of EIF2B5R113H/A403V or EIF2B2G200V/E213G VWMD astrocytes, and were further investigated for their effect on the integrated stress response and mitochondrial stress. Ursodiol demonstrated capacity to ameliorate oxidative stress and loss of mitochondrial membrane potential in VWMD patient iPSC-derived astrocytes in the presence or absence of stress conditions.


Patient-derived cells can be used to identify cellular phenotypes and for large-scale drug screening. Anti-inflammatory compounds, such as berberine, deflazacort, ursodiol and zileuton are potentially repurposable drug candidates for VWMD that should be further investigated for translation in vivo.


Vanishing white matter disease (VWMD) is a rare, autosomal recessive disease, caused by mutations in the genes EIF2B1–5, encoding the eukaryotic initiation factor eIF2B [1]. The eIF2B protein is a guanine nucleotide exchange factor that is involved in the integrated stress response (ISR). VWMD mutations can cause alterations in the activity of the wild-type eIF2B protein and the disease has a variable onset, progression and severity, which can be exacerbated by environmental stress factors [2]. Although the disease course varies widely, earlier disease onset is associated with a more rapid progression and patients developing symptoms as young children typically survive few years beyond diagnosis [3].

The eIF2B proteins control mRNA translation, converting the inactive eIF2-GDP to the active eIF2-GTP form. Activation of the cytoprotective ISR leads to phosphorylation of eIF2α (p-eIF2α), which binds to eIF2B to mediate translational repression, and upregulation of stress-induced genes. These genes include PPP1R15A that encodes growth arrest and DNA damage-inducible protein (GADD34). GADD34 facilitates dephosphorylation of p-eIF2α to promote recovery from stress and resumption of normal protein translation in a negative feedback loop, and transcription factors ATF4 and CHOP [4]. Mutations in eIF2B can lead to delayed translation of stress-induced genes and dysregulated ISR [5]. Although eIF2B is ubiquitously expressed and plays a role in multiple cell types, the disease manifests most significantly in the loss of white matter of the brain [6].

VWMD mouse models and induced pluripotent stem cell (iPSC) models [7, 8] have identified a central role for dysfunctional astrocytes in the development of VWMD, with evidence for astrocytic apoptosis [8] and an inability to promote oligodendrocyte maturation [6]. A key driver of cellular pathogenesis in VWMD involves the ISR, with alterations in responses to endoplasmic reticulum stress and oxidative stress [9, 10]. Mitochondrial dysfunction and increased accumulation of reactive oxygen species have been identified in VWMD murine fibroblasts, astrocytes and oligodendrocyte precursor cells [11, 12]. Currently there are no approved treatments for VWMD, hence the aim of this research was to identify candidates from an FDA-approved drug library that could protect VWMD patient cells against cellular stress. This pre-clinical research can be used to inform future in vivo studies. For phenotypically diverse diseases, such as VWMD, patient-derived cells provide a useful model of disease mutations that can aid translational pathways.

Materials and methods

Cell culture

All experimental protocols were approved by the University of Wollongong Human Research Ethics Committee (HE17/522). Primary human dermal fibroblasts were collected from two VWMD patients that were diagnosed in early childhood or their control family members. The VWMD1 iPSC line was generated from a patient bearing mutations in EIF2B5R113H/A403V, whilst the VWMD6 iPSC line was generated from a patient bearing mutations in EIF2B2G200V/E213G. The VWMD1 control line is a healthy control (no disease mutations, EIF2B5R113/A403), whilst the VWMD6 control bears one VWMD allele and one wild-type allele (non-disease carrier, EIF2B2G200V/E213). Cells were maintained in DMEM/F12 (Gibco, Ireland, 21331020), 10% FBS (Bovogen Biologicals, Australia, SFBS-AU), 2 mM L-glutamine (Gibco, Ireland, 25030081) and 1% penicillin/streptomycin (Gibco, 15140122). Fibroblasts were reprogrammed with the StemRNA (ReproCell, Beltsville, Maryland, 00-0076) non-integrating method, as previously described [13, 14]. The mRNA-based reprogramming of patient fibroblasts into iPSCs, pluripotency marker immunofluorescence and RT-qPCR, karyotyping and mutation genotype characterisation are described in Supplementary Information (Figure S1). The iPSCs were maintained in mTESR1 (Stem Cell Technologies, Australia, 85850) and astrocytes in Astrocyte Growth Supplement Medium (AGS; ScienCell Research Laboratories, San Diego, California, USA, 1852). All cells were maintained in humidified incubators at 37 °C supplemented with 5% CO2 for fibroblasts or hypoxic 3% O2 conditions for iPSCs and astrocytes. Neural inductions were carried out as previously described [15]. Astrocyte differentiations were performed in ciliary neurotrophic factor (CNTF), epidermal growth factor (EGF) and basic fibroblast growth factor (FGF-2)-based medium and transitioned to AGS, prior to characterisation by immunofluorescence and inflammatory activation, to confirm the production of functional astrocytes from iPSCs. Details of neural inductions and astrocyte characterisations are in Supplementary Information (Figure S2). Genotyping and characterisation RT-qPCR primers are listed in Table S1.

Cell stress and viability assays

Fibroblasts or astrocytes were seeded in 96 well plates (5000 cells/well) and incubated with a range of concentrations of hydrogen peroxide (H2O2; 0–2 mM, Sigma-Aldrich, 16911), MG132 (0–20 μM; Focus Bioscience, HY-13259) or thapsigargin (0–20 μM; Abcam, Australia, ab120286) overnight. Viability was assessed via resazurin reduction assay (STEMCELL Technologies, 75005), in which cells received 15 μM resazurin for 1 h prior to measurements by fluorescence plate spectroscopy (excitation 544/emission 590 nm). For coincubation assays, cells were similarly prepared and incubated with candidate drugs and, based on the IC60 for each cell type, MG132 at either 2.5 μM (fibroblasts) or 0.1 μM (astrocytes) for 48 h. The coincubation drug screen was carried out in VWMD1 patient fibroblasts with the MicroSource Spectrum FDA collection (Compounds Australia), a library of 2400 FDA-approved drugs. Cell seeding and drug reagent preparation were performed by a robotic liquid handler (Hamilton Microlab STAR, Reno, Nevada). First-pass screening of the drug library was carried out at 20 μM; drug candidates selected for further testing underwent dose-response curves to identify appropriate concentrations. Single drug concentrations were selected based on protective efficacy in dose-response coincubation assays and used in downstream assays at concentrations of: Anavex 2–73 (AVex-73, 5 μM; Focus Bioscience, Australia, HY-101864), berberine (1.25 μM; Sigma-Aldrich, Australia, 14050), guanabenz (5 μM; Focus Bioscience, HY-B0566), ISRIB (1.25 μM; Focus Bioscience, HY-12495), deflazacort (5 μM; Sigma-Aldrich, SML0123), ursodiol (40 μM; Sigma-Aldrich, PHR1579) and zileuton (1.25 μM; Focus Bioscience, HY-14164). A DMSO (Sigma-Aldrich, D-4540) vehicle (solvent) concentration of 0.2% was included in all experiments.

Integrated stress response protein quantification

Fibroblasts or astrocytes were seeded in 6 well plates (200,000 cells/well) and incubated with MG132 for 0, 24 or 48 h, or coincubated with candidate compound and MG132 overnight. Cultures were lysed in RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS and protease inhibitors cOmplete Protease Inhibitor (Roche, 4693116001) and phosphatase inhibitors PhosSTOP EASY (Roche, 4906837001)) before sonication and quantification by a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, 23225). Protein lysates were heated for 5 min at 95 °C in loading buffer (2% SDS, 10% 2-mercaptoethanol, 10% glycerol, 0.001% bromophenol blue, 0.0625 M Tris-HCl, pH 6.8). Denatured protein lysates (5 μg) were separated on 4–20% Criterion TGX Precast Gels (Bio-Rad, Australia, 5678095) and transferred to Immobilon-FL PVDF membrane (0.45 μm pore; Merck, Darmstadt, Germany, IPFL00010). Membranes were blocked with 0.5% casein (Bio-Rad, 1610782) for 1 h (RT), primary antibody at 1:1000 dilution for 16 h (4 °C), and secondary antibody at 1:5000 dilution for 1 h (4 °C), facilitated by an automated western blot processor (Cytoskeleton GOBlot, Denver, Colorado, USA). Antibodies utilised included anti-eIF2α (Abcam, ab5369), anti-eIF2α-phopsho-S51 (Abcam, ab32157), anti-DDIT3 (DNA damage inducible transcript 3, also known as C/EBP homologous protein (CHOP; Abcam, ab179823), anti-GADD34 (Sigma-Aldrich, SAB2701237), donkey anti-mouse Alexa Fluor Plus 488 (Thermo Fisher Scientific A32766) and donkey anti-rabbit Alexa Fluor Plus 647 (Thermo Fisher Scientific A32795). Blots were imaged with a Bio-Rad ChemiDoc MP. Band intensities were normalised to stain-free total protein. Relative eIF2α phosphorylation was normalised to total eIF2α. Dynamic range of signal was confirmed by linear signal to loading ratio (Figure S3).

Oxidative stress, mitochondrial membrane potential and membrane integrity microscopy assays

Astrocytes were seeded in 96 well plates (10,000 cells per well) and coincubated overnight with MG132, as described above, and candidate drugs, or the positive controls for oxidative stress: H2O2 (100 μM), or the proton uncoupler carbonyl cyanide-4-phenylhydrazone (FCCP; 10 μM; Focus Bioscience, HY-100410). Cells were loaded with 20 μM H2DCFDA (DCF; Thermo Fisher Scientific, D399) or 0.5 μM tetramethylrhodamine ethyl ester (TMRE; Thermo Fisher Scientific, T669) and 10 μM Hoechst 33342 (Sigma-Aldrich, B2261) for 0.5 h before replacement of medium with phenol red-free DMEM/F12 medium (Gibco, 21,041,025) and further incubation for 1 h. Relative fold change in DCF signal was acquired by fluorescence spectroscopy and normalised to cell density quantified by 0.004% sulforhodamine B (Sigma-Aldrich, 230162) solubilised in 10 mM Trizma (Sigma-Aldrich, T1503) after fixation in 4% trichloroacetic acid (Sigma-Aldrich, T9159) [16]. TMRE images were acquired by confocal microscopy, TMRE (excitation 522/emission 590–620 nm) and analysed with ImageJ to determine fluorescence normalised to cell density.

Statistical analyses

Data are presented as the mean and standard error of the mean of at least 3 independent experiments; statistical significance was assessed by two-way analysis of variance followed by Holm-Sidak posthoc test for multiple comparisons unless otherwise stated. Significance was accepted where p < 0.05. Charts and statistical analyses were prepared in Prism GraphPad 8.0.


VWMD patient fibroblasts and astrocytes exhibit dysregulated ISR marker expression

Mutations in eIF2B suppress both global and stress-induced protein translation, in response to ER stress [5, 17]. Given that eIF2B is ubiquitously expressed, we examined patient fibroblasts and iPSC-derived astrocytes for evidence of a dysfunctional ISR. Fibroblasts were reprogrammed into iPSCs from two VWMD patients, along with gender-matched relatives as non-disease controls (Figure S1). A previous study identified that white matter-derived astrocytes, generated using a ciliary neurotrophic factor (CNTF)-based differentiation protocol, showed a more vulnerable phenotype to stress, compared to grey matter astrocytes generated using fetal bovine serum [7]. Consequently, we generated astrocytes from iPSCs using a CNTF-based method (Figure S2).

Responses to oxidative and endoplasmic reticulum (ER) stresses are all proposed to be affected in VWMD cell and animal-based models [17]. Studies employing cells bearing different VWMD mutations have shown variability in the ISR [5], and primary astrocytes from mice bearing a homozygous R191H mutation in eIF2Bε did not exhibit the ISR in vitro [18]. Thus, first we established stress conditions that could generate an exacerbated cellular phenotype in the VWMD patient cells compared to controls. We evaluated the dose-dependent effect of H2O2, as a mediator of oxidative stress, thapsigargin, as a mediator of ER stress, and MG132. MG132 is commonly utilised as a proteasomal inhibitor that also induces ER and oxidative stress via the unfolded protein response, all of which trigger the ISR [19, 20]. There was a small but significant reduction in cell survival for VWMD fibroblasts and iPSC-derived astrocytes, compared to non-disease controls, under all three stressors (Figure S4A).

Disrupted ISR homeostasis can be assessed by measuring the expression levels of ISR-relevant markers. The control of eIF2α phosphorylation or dephosphorylation acts as a pivotal mechanism that regulates global protein synthesis in response to cell stress [21]; GADD34 dephosphorylates eIF2α [22] while CHOP is activated by the ISR to promote apoptosis [23]. The expression levels of these ISR-relevant proteins, p-eIF2α (normalised to eIF2α), GADD34, and CHOP, were compared in VWMD and control fibroblasts, and iPSC-derived astrocytes, under MG132 stress (Fig. 1; antibody characterisation and representative blots shown in Figure S34). VWMD lymphoblasts have previously been reported to show reduced GADD34 expression and increased levels of p-eIF2α, following thapsigargin-induced acute ER stress [5]. Under MG132-induced stress, VWMD fibroblasts showed reduced upregulation of eIF2α phosphorylation and reduced GADD34 expression, and increases in CHOP expression, compared to controls (Fig. 1a-c). MG132-stressed VWMD astrocytes also showed reduced upregulation of eIF2α phosphorylation and increased CHOP, compared to controls (Fig. 1d-f). VWMD astrocytes exhibited increased GADD34 (Fig. 1e), consistent with glia in animal model studies [17] and the hypophosphorylation of eIF2 observed in white matter patient tissue [24].

Fig. 1
figure 1

ISR marker protein expression is affected in VWMD fibroblasts and astrocytes. a-f Effect of MG132 stress on protein expression of the ISR markers, p-eIF2α (normalised to eIF2α), GADD34 or CHOP, at 0, 24 or 48 h incubation period. Protein levels were quantified by western blot, normalised to total protein, and are shown relative to control at time 0. Individual data points are shown from VWMD1 patient EIF2B5R113H/A403V; VWMD6 patient EIF2B2G200V/E213G or their non-disease controls VWMD1 Control EIF2B5R113/A403; VWMD6 Control EIF2B2G200V/E213, with mean ± SEM, n = 3. Significant differences between pooled disease and control groups were identified by two-way ANOVA followed by Holm-Sidak post hoc test, * p < 0.05, ** p < 0.01, *** p < 0.001. Representative blots shown in (Figure S4B)

Cytoprotective drug screen in VWMD patient cultures

Based on the established capacity of MG132 to induce oxidative stress, and exacerbate ISR disease phenotypes, we performed a first-pass drug screen for candidates able to protect against the effect of low dose MG132 in VWMD1 EIF2B5R113H/A403V patient fibroblasts, with cell viability assessed by resazurin reduction activity (Fig. 2).

Fig. 2
figure 2

Drug screen of cytoprotective candidates. a VWMD1 EIF2B5R113H/A403V disease fibroblasts were coincubated with MG132 and each of 2400 drugs from an FDA-approved drug library. Cell viability was measured and normalised to cell viability with MG132 stressor in the absence of drug. Drugs were scored based on increase in cell viability with threshold set to > 1.5× standard deviation (n = 3). Cytoprotection ranged from 2-fold increase in viability to 0 (100% cell death for cytotoxic drugs). Horizontal dotted red line indicates 1.5 × standard deviation of MG132 stressed controls. b Hit candidates were taken forward to assess their dose-response effect on cell viability of VWMD1 EIF2B5R113H/A403V fibroblasts and iPSC-derived astrocytes, relative to MG132 stress (n = 5–6). Heat map represents cytoprotection (blue) versus cytotoxicity (red) of individual drugs

Following the initial drug screen, a panel of 20 compounds (Table S2) was selected for downstream evaluation in VWMD1 EIF2B5R113H/A403V fibroblasts and in both VWMD1 EIF2B5R113H/A403V and VWMD6 EIF2B2G200V/E213G patient-derived astrocyte cultures. The compounds selected for downstream assays were chosen based on their potential for clinical translation, including considerations of bioavailability, route of administration and potential side effects. The drug panel included 14 protective compounds from the screen (> 1.5 × standard deviation of MG132-stressed controls), and a further six compounds with relevant modes of action for VWMD. The sigma-1 receptor pathway was recently identified in a drug screen to protect against mitochondrial dysfunction in a murine EIF2B5R132H/R132H model [11]. Nominated drugs on the basis of relevant mode of action to VWMD, included guanabenz [25], ISRIB [26], sigma-1 receptor agonists, AVex-73 and amitriptyline [11], as well as the central nervous system (CNS) cytoprotective compounds tauroursodeoxycholic acid and alkaloid berberine [27]. Overall, the panel of compounds for further evaluation included glucocorticosteroids, bile acids, iron chelators, antioxidants, ISR modulators and sigma-1 receptor agonists. All candidates elicited cytoprotective effects against MG132-induced stress at varying concentrations in VWMD1 fibroblasts, with the exception of ISRIB. Ursodiol and its taurine derivative, tauroursodeoxycholic acid, showed similar cytoprotective efficacies. Anti-inflammatories were amongst the largest category of compounds that improved cell viability under proteasomal stress, with a high proportion of these being glucocorticosteroids. The recent demonstration of mitochondrial dysfunction and inefficient respiration in murine models of VWMD [11] has expanded the search for possible therapeutics to include mitochondrial protective compounds and antioxidants. The sigma-1 receptor is a chaperone protein in ER membranes that governs a range of cellular processes, including calcium homeostasis and reactive oxygen species accumulation [28]. However, in our study, the cytoprotective effect of the sigma-1 receptor agonists, AVex-73 and amitriptyline, was limited at higher concentrations (≥5 μM; Fig. 2, S5). Drugs assessed in iPSC-derived astrocytes from VWMD1 EIF2B5R113H/A403V and VWMD6 EIF2B2G200V/E213G yielded comparable results (Figure S5). The majority of the drugs that were protective in VWMD1 fibroblasts were also protective in VWMD1 and VMWD6 astrocytes, with the exception of amitriptyline, curcumin and budesonide (Fig. 2, S5).

Further studies were conducted to gain insight into the mode of action of drugs with therapeutic potential for VWMD (Fig. 2, S5), including AVex-73, berberine, deflazacort, guanabenz, ISRIB, ursodiol and zileuton. Assays to investigate drug mechanisms were performed with VWMD1 EIF2B5R113H/A403V patient astrocytes, given that the majority of VWMD-causing mutations affect the eIF2Bε subunit [3].

Effect of drug candidates on cell death in VWMD iPSC-astrocytes

To determine whether candidate drugs reduced the proportion of bulk cell death, the membrane permeabilisation of treated astrocyte cultures was evaluated. In MG132-stressed VWMD1 astrocytes there were no significant changes in the percentage of membrane-permeabilised cells, with the exception of ISRIB increasing the proportion of membrane-permeabilised cells (Figure S6). To test whether the drugs caused a reduction in mitochondrial apoptosis, the BAX:BCL2 ratio was evaluated. The ratio of BAX:BCL2 is used to assess the levels of Bax (proapoptotic) and Bcl-2 (antiapoptotic) expression. There was no detectable change in BAX:BCL2 ratio caused by any of the drug candidates, with the exception of an increase in BAX:BCL2 caused by ISRIB (Figure S6).

Effect of candidate drugs on ISR-relevant proteins p-eIF2α, GADD34 and CHOP

Candidate drugs were assessed for their effect on the ISR under MG132 stress in astrocytes. The expression levels of the ISR-relevant proteins, p-eIF2α (normalised to eIF2α), GADD34 and CHOP, were evaluated following candidate drug treatment (Fig. 3a-c). ISRIB markedly increased p-eIF2α and decreased GADD34 and CHOP expression under MG132 stress conditions. AVex-73, berberine and deflazacort significantly decreased CHOP expression in the presence of MG132.

Fig. 3
figure 3

Effect of candidate drugs on ISR markers in vehicle control (blue circle) or 0.1 μM MG132-treated (orange triangle) VWMD1 EIF2B5R113H/A403V patient iPSC-derived astrocytes. Cells were treated with candidate drugs, with or without MG132 for 24 h. Protein expression was measured by western blot and normalised to total protein for the ISR markers: a phosphorylated eIF2α, normalised to eIF2α; b GADD34; c CHOP. Individual data points are shown with mean ± SEM (n = 4). Significant differences were identified by two-way ANOVA, followed by Holm-Sidak posthoc test * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Representative western blots shown in (Figure S3)

Previous cell stress and neurodegeneration studies have established ISR-modulating roles for guanabenz and ISRIB [29, 30]. Guanabenz is considered to exert cytoprotective effects by inhibiting the activity of GADD34 to recruit eIF2 phosphatases, thus prolonging translation inhibition and avoiding the added stress of resuming protein synthesis [29]. However, at concentrations of guanabenz that induced a cytoprotective effect (5 μM), we did not observe a significant impact on expression of any ISR markers in astrocytes. Conversely, ISRIB-mediated cytoprotection of astrocytes (1.25 μM) was associated with increased eIF2α phosphorylation, and downregulation of GADD34 and CHOP.

Effect of candidate drugs on indicators of mitochondrial function

The eIF2B mutations in murine models have been shown to decrease mitochondrial membrane potential and impair mitochondrial complex I function, resulting in a compensatory increase in mitochondrial abundance [9]. Furthermore, mutations in eIF2B genes impair mitochondrial function during oxidative stress conditions in VWMD murine fibroblasts and astrocytes [11]. VWMD1 and VWMD6 patient iPSC-derived astrocytes demonstrated reduced mitochondrial membrane potential in both the presence and absence of MG132, suggesting reduced mitochondrial membrane potential is a cellular phenotype of VWMD patient cells, even under basal conditions (Figure S8).

The candidate drugs were investigated for their ability to protect against oxidative stress and improve mitochondrial membrane potential. VWMD1 EIF2B5R113H/A403V patient astrocytes treated with ursodiol or zileuton led to decreased generation of reactive oxygen species, in both the presence and absence of MG132 stress (Fig. 4a). This correlates with the purported radical scavenging activity of the 5-lipoxygenase antagonist zileuton [31]. Ursodiol also increased the relative mitochondrial membrane potential of VWMD1 EIF2B5R113H/A403V astrocytes under both unstressed and stressed conditions (Fig. 4b), consistent with a loss of mitochondrial membrane potential as a cellular phenotype of VWMD astrocytes. Furthermore ursodiol reduced oxidative stress in both control and VWMD astrocytes (Fig. 4c) and improved the mitochondrial membrane potential phenotype of VWMD astrocytes, bringing the levels of TMRE to those of control cells (Fig. 4d-g). Together, these data suggest that ursodiol may promote mitochondrial function and reduce oxidative stress in VWMD astrocytes.

Fig. 4
figure 4

Effect of candidate drugs on oxidative stress and mitochondrial membrane potential in VWMD astrocyte cell culture. VWMD1 iPSC-derived astrocytes were treated with vehicle control (DMSO; blue circle) or 0.1 μM MG132 (orange triangle) and candidate drugs for 24 h. (a) Reactive oxygen species measured via DCF fluorescence and (b) TMRE fluorescence as an indicator of mitochondrial membrane potential. Effect of ursodiol on (c) ROS generation via DCF fluorescence and (d) mitochondrial membrane potential via TMRE fluorescence in disease and control astrocyte cultures. Representative confocal fluorescence microscopy images of: (e) FCCP (protonophore, mitochondrial uncoupler, loss of TMRE fluorescence), (f) TMRE assay solvent, (g) ursodiol in VWMD1 EIF2B5R113H/A403V astrocytes (scale bar = 30 μm) (orange = TMRE; blue = Hoechst 33342 nuclear stain). For A-D individual data points represent mean ± SEM (n = 3–6); significant differences were identified by two-way ANOVA followed by Holm-Sidak posthoc test * p < 0.05, ** p < 0.01, *** p < 0.001


This study screened a panel of 2400 FDA-approved drugs, of varying classes, in an aim to identify repurposable drugs to protect against VWMD. A phenotype of VWMD cells, under basal conditions or in the presence of stress, highlighted dysfunction in the mitochondrial membrane potential of VWMD cells. Protective drugs targeted common pathways, and included compounds with anti-inflammatory activity and/or drugs purported to improve mitochondrial function or reduce oxidative stress. Steroids can regulate inflammation, mitochondrial function and apoptosis with neuroprotective effects in brain injury, Alzheimer’s disease, Parkinson’s disease, multiple sclerosis, and stroke [32]. Although glucocorticosteroids are a commonly administered class of drugs, an anecdotal study of corticosteroids on three VWMD patients did not identify benefits and the patients were removed from this treatment due to potential clinical complications [33]. Nonetheless, deflazacort was included in our study as a representative glucocorticosteroid on the basis of fewer reports of adverse effects in the literature and its existing use in the clinic, including in children with muscular dystrophy [34, 35]. Reduced CHOP expression in stress-induced cells following deflazacort treatment may reflect an ability of this drug to improve mitochondrial function [36].

The protective effect of the antioxidant edaravone was evident in astrocytes at a higher efficacy than any other candidate, potentially due to its well established radical scavenging activity [37]. Edaravone was recently approved for amyotrophic lateral sclerosis and while its administration is currently limited to intravenous injection, oral and mucosal formulations are in development [38,39,40].

Zileuton is an oral anti-asthmatic leukotriene inhibitor with reactive oxygen species scavenging activity [31]. In this study an increase in cell viability was correlated with a significant decrease in ROS generation but no significant effects on integrated stress response modulation. Its neuroprotective effects have been correlated to anti-inflammatory activity in cerebral ischemia, possibly as a result of reducing ROS-activated inflammation, and preventing glutamate oxidative toxicity via ferroptosis [31, 41].

ISRIB is thought to allow eIF2B to escape inhibitory complex formation with p-eIF2α, thus limiting repression of protein synthesis while under ER stress [17, 26]. Treatment of EIF2B5 VWMD mice with ISRIB, or a derivative, 2BAct, ameliorated myelin loss, and improved ISR signature and motor function [17, 24, 26]. It is also worth noting that the stabilising effect of ISRIB may be dependent on the precise mutations and the cellular concentration of p-eIF2α, whereby it has been found to inhibit low level but not high level ISR activation [24, 42]. As ISRIB inhibits the integrated stress response, it is evident that enabling protein synthesis to resume under acute proteotoxic stress may have an adverse effect on cell survival, and the enhanced effect of cytotoxic drugs by ISRIB have been recently demonstrated in prostate and pancreatic cancer studies [43, 44].

The phytochemical, berberine, was included in the panel as a well-tolerated natural compound [27]. In MG132-stressed astrocytes berberine treatment increased cell viability, potentially via decreasing pro-apoptotic CHOP. The antioxidant, anti-inflammatory, ER and mitochondrial protective effects of berberine have been noted in numerous studies, and include increased levels of superoxide dismutase and glutathione, inhibition of apoptosis and decreased cytochrome C and BAX:BCL2 ratio in ischemic injury and diabetic animal models [45, 46].

Overall, one of the most promising drugs identified with translational potential was ursodiol, a bile acid naturally formed in the liver and administered for gallstones. Ursodiol has a demonstrated capacity to cross the blood brain barrier, based on clinical trials for motor neurone disease, reaching levels in cerebral spinal fluid that could be protective [47]. Additionally, ursodiol has been identified as protective in other neurodegenerative and optical atrophy research, where its anti-apoptotic and neuroprotective effects were observed, although the underlying molecular mechanism of ursodiol-mediated protection was not identified [48, 49]. Although ursodiol is considered to be generally well tolerated, fatigue, nausea, weight gain, and depression are possible symptoms [50]. Progression of VWMD includes glial cell death, developing towards neuronal cell death, and leading to paralysis and neuropathy [51]. We observed ursodiol to decrease oxidative stress and increase mitochondrial membrane potential, under stressed and non-stressed conditions, and improve cell viability in VWMD astrocytes under proteasomal stress. The increase of mitochondrial membrane potential by ursodiol identified in our study is consistent with reports for Alzheimer’s disease patient fibroblasts [52]. Considering disrupted mitochondrial function has been extensively implicated in VWMD and a wide range of neurodegenerative disorders, further investigation of the neuroprotective effects of ursodiol are warranted. Future studies should confirm these findings in additional VWMD patient cell lines and in vivo studies.


The hypothetical premise of cytoprotective drugs for VWMD are those that can reverse cellular phenotypes and limit disease-related degeneration. Patient-derived cells are a useful resource to support drug repurposing studies for rare diseases. Together with information from previous studies on other CNS diseases [47, 52] our preliminary findings are supportive toward ongoing in vivo investigation of ursodiol, zileuton and other anti-inflammatory drugs as cytoprotective agents for VWMD.

Availability of data and materials

The authors confirm that the data supporting the findings of this study are available within the article. Data and materials may be available on request if the requests conform to the Human Research Ethics Committee approval requirements.



Anavex 2–73


C/EBP homologous protein


Central nervous system


Ciliary neurotrophic factor




Epidermal growth factor


Endoplasmic reticulum


Food and Drug Administration


Carbonyl cyanide-4-phenylhydrazone


Basic fibroblast growth factor


Growth arrest and DNA damage-inducible protein


induced pluripotent stem cell


Integrated stress response


phosphorylated eIF2α


Tetramethylrhodamine ethyl ester


Vanishing white matter disease


  1. Bugiani M, Vuong C, Breur M, van der Knaap MS. Vanishing white matter: a leukodystrophy due to astrocytic dysfunction. Brain Pathol. 2018;28(3):408–21.

    PubMed  Article  PubMed Central  Google Scholar 

  2. Proud CG. Vanishing white matter: the next 10 years. Future Neurol. 2012;7(1):81–92.

    CAS  Article  Google Scholar 

  3. Hamilton EM, van der Lei HD, Vermeulen G, Gerver JA, Lourenço CM, Naidu S, et al. Natural history of vanishing white matter. J Ann Neurol. 2018;84(2):274–88.

    Article  Google Scholar 

  4. Brush MH, Shenolikar S. Control of cellular GADD34 levels by the 26S proteasome. Mol Cell Biol. 2008;28(23):6989–7000.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  5. Moon SL, Parker R. EIF2B2 mutations in vanishing white matter disease hypersuppress translation and delay recovery during the integrated stress response. RNA. 2018;24(6):841–52.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  6. Dooves S, Bugiani M, Postma NL, Polder E, Land N, Horan ST, et al. Astrocytes are central in the pathomechanisms of vanishing white matter. J Clin Invest. 2016;126(4):1512–24.

    PubMed  PubMed Central  Article  Google Scholar 

  7. Leferink PS, Dooves S, Hillen AEJ, Watanabe K, Jacobs G, Gasparotto L, et al. Astrocyte subtype vulnerability in stem cell models of vanishing white matter. Ann Neurol. 2019;86(5):780–92.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  8. Zhou L, Li P, Chen N, Dai L-F, Gao K, Liu Y-N, et al. Modeling vanishing white matter disease with patient-derived induced pluripotent stem cells reveals astrocytic dysfunction. CNS Neurosci Ther. 2019;25(6):759–71.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  9. Raini G, Sharet R, Herrero M, Atzmon A, Shenoy A, Geiger T, et al. Mutant eIF2B leads to impaired mitochondrial oxidative phosphorylation in vanishing white matter disease. J Neurochem. 2017;141(5):694–707.

    CAS  PubMed  Article  Google Scholar 

  10. van der Voorn JP, van Kollenburg B, Bertrand G, Van Haren K, Scheper GC, Powers JM, et al. The unfolded protein response in vanishing white matter disease. J Neuropathol. 2005;64(9):770–5.

    Article  Google Scholar 

  11. Atzmon A, Herrero M, Sharet-Eshed R, Gilad Y, Senderowitz H, Elroy-Stein O. Drug screening identifies sigma-1-receptor as a target for the therapy of VWM leukodystrophy. Front Mol Neurosci. 2018;11:336.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  12. Herrero M, Mandelboum S, Elroy-Stein O. eIF2B mutations cause mitochondrial malfunction in Oligodendrocytes. NeuroMolecular Med. 2019;21(3):303–13.

    CAS  PubMed  Article  Google Scholar 

  13. Bax M, Balez R, Muñoz SS, Do-Ha D, Stevens CH, Berg T, et al. Generation and characterization of a human induced pluripotent stem cell line UOWi005-a from dermal fibroblasts derived from a CCNFS621G familial amyotrophic lateral sclerosis patient using mRNA reprogramming. Stem Cell Res. 2019;40:101530.

    CAS  PubMed  Article  Google Scholar 

  14. Balez R, Berg T, Bax M, Muñoz SS, Cabral-da-Silva MC, Engel M, et al. The mRNA-based reprogramming of fibroblasts from a SOD1E101G familial amyotrophic lateral sclerosis patient to induced pluripotent stem cell line UOWi007. Stem Cell Res. 2020;42:101701.

    CAS  PubMed  Article  Google Scholar 

  15. Denham M, Dottori M. Neural differentiation of induced pluripotent stem cells. Neurodegeneration. New York: Springer; 2011. p. 99–110.

  16. Vichai V, Kirtikara K. Sulforhodamine B colorimetric assay for cytotoxicity screening. Nat Protoc. 2006;1(3):1112.

    CAS  PubMed  Article  Google Scholar 

  17. Wong YL, LeBon L, Basso AM, Kohlhaas KL, Nikkel AL, Robb HM, et al. eIF2B activator prevents neurological defects caused by a chronic integrated stress response. ELife. 2019;8:e42940.

    PubMed  PubMed Central  Article  Google Scholar 

  18. Wisse LE, ter Braak TJ, van de Beek M-C, van Berkel CGM, Wortel J, Heine VM, et al. Adult mouse eIF2Bε Arg191His astrocytes display a normal integrated stress response in vitro. Sci Rep. 2018;8(1):3773.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  19. Park HS, Jun DY, Han CR, Woo HJ, Kim YH. Proteasome inhibitor MG132-induced apoptosis via ER stress-mediated apoptotic pathway and its potentiation by protein tyrosine kinase p56lck in human Jurkat T cells. Biochem Pharmacol. 2011;82(9):1110–25.

    CAS  PubMed  Article  Google Scholar 

  20. Maharjan S, Oku M, Tsuda M, Hoseki J, Sakai Y. Mitochondrial impairment triggers cytosolic oxidative stress and cell death following proteasome inhibition. Sci Rep. 2014;4(1):5896.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  21. Jiang H-Y, Wek RC. Phosphorylation of the α-subunit of the eukaryotic initiation factor-2 (eIF2α) reduces protein synthesis and enhances apoptosis in response to proteasome inhibition. J Biol Chem. 2005;280(14):14189–202.

    CAS  PubMed  Article  Google Scholar 

  22. Brush MH, Weiser DC, Shenolikar S. Growth arrest and DNA damage-inducible protein GADD34 targets protein phosphatase 1 alpha to the endoplasmic reticulum and promotes dephosphorylation of the alpha subunit of eukaryotic translation initiation factor 2. Mol Cell Biol. 2003;23(4):1292–303.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  23. Teske BF, Fusakio ME, Zhou D, Shan J, McClintick JN, Kilberg MS, et al. CHOP induces activating transcription factor 5 (ATF5) to trigger apoptosis in response to perturbations in protein homeostasis. Mol Biol Cell. 2013;24(15):2477–90.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  24. Abbink TEM, Wisse LE, Jaku E, Thiecke MJ, Voltolini-González D, Fritsen H, et al. Vanishing white matter: deregulated integrated stress response as therapy target. Ann Clin Transl Neurol. 2019;6(8):1407–22.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  25. Dooves S, Bugiani M, Wisse LE, Abbink TEM, van der Knaap MS, Heine VM. Bergmann glia translocation: a new disease marker for vanishing white matter identifies therapeutic effects of Guanabenz treatment. Neuropathol Appl Neurobiol. 2018;44(4):391–403.

    CAS  PubMed  Article  Google Scholar 

  26. Wong YL, LeBon L, Edalji R, Lim HB, Sun C, Sidrauski C. The small molecule ISRIB rescues the stability and activity of vanishing white matter disease eIF2B mutant complexes. Elife. 2018;7:e32733.

    PubMed  PubMed Central  Article  Google Scholar 

  27. Vuddanda PR, Chakraborty S, Singh S. Berberine: a potential phytochemical with multispectrum therapeutic activities. Expert Opin Investig Drugs. 2010;19(10):1297–307.

    CAS  PubMed  Article  Google Scholar 

  28. Weng T-Y, Tsai S-YA ST-P. Roles of sigma-1 receptors on mitochondrial functions relevant to neurodegenerative diseases. J Biomed Sci. 2017;24(1):74.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  29. Wang L, Popko B, Tixier E, Roos RP. Guanabenz, which enhances the unfolded protein response, ameliorates mutant SOD1-induced amyotrophic lateral sclerosis. Neurobiol Dis. 2014;71:317–24.

    CAS  PubMed  Article  Google Scholar 

  30. Halliday M, Radford H, Sekine Y, Moreno J, Verity N, Le Quesne J, et al. Partial restoration of protein synthesis rates by the small molecule ISRIB prevents neurodegeneration without pancreatic toxicity. Cell Death Dis. 2015;6(3):e1672-e.

    Article  CAS  Google Scholar 

  31. Liu Y, Wang W, Li Y, Xiao Y, Cheng J, Jia J. The 5-lipoxygenase inhibitor zileuton confers neuroprotection against glutamate oxidative damage by inhibiting ferroptosis. Biol Pharm Bull. 2015;38(8):1234–9.

    CAS  PubMed  Article  Google Scholar 

  32. Garcia-Segura LM. Hormones and brain plasticity. Oxford: Oxford University press; 2009.

  33. Turón-Viñas E, Pineda M, Cusí V, López-Laso E, Del Pozo RL, Gutiérrez-Solana LG, et al. Vanishing white matter disease in a spanish population. Journal of central nervous system disease. J Cent Nerv Syst Dis. 2014;6:JCNSD. S13540.

  34. Petnikota H, Madhuri V, Gangadharan S, Agarwal I, Antonisamy B. Retrospective cohort study comparing the efficacy of prednisolone and deflazacort in children with muscular dystrophy: a 6 years’ experience in a south Indian teaching hospital. Indian J Orthop. 2016;50(5):551–7.

    PubMed  PubMed Central  Article  Google Scholar 

  35. McAdam LC, Mayo AL, Alman BA, Biggar WD. The Canadian experience with long-term deflazacort treatment in Duchenne muscular dystrophy. Acta Myol. 2012;31(1):16–20.

    PubMed  PubMed Central  Google Scholar 

  36. Guiraud S, Davies KE. Pharmacological advances for treatment in Duchenne muscular dystrophy. Curr Opin Pharmacol. 2017;34:36–48.

    CAS  PubMed  Article  Google Scholar 

  37. Kawasaki T, Kitao T, Nakagawa K, Fujisaki H, Takegawa Y, Koda K, et al. Nitric oxide-induced apoptosis in cultured rat astrocytes: protection by edaravone, a radical scavenger. Glia. 2007;55(13):1325–33.

    PubMed  Article  Google Scholar 

  38. Sato T, Mizuno K, Ishii F. A novel administration route of edaravone–II: mucosal absorption of edaravone from edaravone/hydroxypropyl-beta-cyclodextrin complex solution including L-cysteine and sodium hydrogen sulfite. Pharmacology. 2010;85(2):88–94.

    CAS  PubMed  Article  Google Scholar 

  39. Parikh A, Kathawala K, Tan CC, Garg S, Zhou X-F. Development of a novel oral delivery system of edaravone for enhancing bioavailability. Int J Pharm. 2016;515(1–2):490–500.

    CAS  PubMed  Article  Google Scholar 

  40. Rothstein JD. Edaravone: a new drug approved for ALS. Cell. 2017;171(4):725.

    CAS  PubMed  Article  Google Scholar 

  41. X-k T, Yang W-z, Wang C-h, S-s S, Y-l Z, C-m C, et al. Zileuton reduces inflammatory reaction and brain damage following permanent cerebral ischemia in rats. J Inflammation. 2010;33(5):344–52.

    Article  CAS  Google Scholar 

  42. Rabouw HH, Langereis MA, Anand AA, Visser LJ, de Groot RJ, Walter P, et al. Small molecule ISRIB suppresses the integrated stress response within a defined window of activation. Proc Natl Acad Sci. 2019;116(6):201815767.

    Article  CAS  Google Scholar 

  43. Palam L, Gore J, Craven K, Wilson J, Korc M. Integrated stress response is critical for gemcitabine resistance in pancreatic ductal adenocarcinoma. Cell Death Dis. 2015;6(10):e1913-e.

    Article  CAS  Google Scholar 

  44. Nguyen HG, Conn CS, Kye Y, Xue L, Forester CM, Cowan JE, et al. Development of a stress response therapy targeting aggressive prostate cancer. Sci Transl Med. 2018;10(439):eaar2036.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  45. Yu W, Sheng M, Xu R, Yu J, Cui K, Tong J, et al. Berberine protects human renal proximal tubular cells from hypoxia/reoxygenation injury via inhibiting endoplasmic reticulum and mitochondrial stress pathways. J Transl Med. 2013;11(1):24.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  46. Li Z, Geng Y-N, Jiang J-D, Kong W-J. Antioxidant and anti-inflammatory activities of berberine in the treatment of diabetes mellitus. Evid Based Complement Altern Med. 2014;2014:289264.

  47. Parry GJ, Rodrigues CM, Aranha MM, Hilbert SJ, Davey C, Kelkar P, et al. Safety, tolerability, and cerebrospinal fluid penetration of ursodeoxycholic acid in patients with amyotrophic lateral sclerosis. J Clin Neuropharmacol. 2010;33(1):17–21.

    CAS  Article  Google Scholar 

  48. Boatright JH, Nickerson JM, Moring AG, Pardue MT. Bile acids in treatment of ocular disease. J Ocul Biol. 2009;2(3):149.

    Google Scholar 

  49. Foster SL, Kendall C, Lindsay AK, Ziesel AC, Allen RS, Mosley SS, et al. Development of Bile Acids as Anti-Apoptotic and Neuroprotective Agents in Treatment of Ocular Disease. Drug Product Dev for the Back of the Eye. Boston: Springer; 2011. p. 565–76.

  50. National Institute of Health. LiverTox: Clinical and Research Information on Drug-Induced Liver Injury. Ursodiol (Ursodeoxycholic Acid). 2012. Available from: Accessed 15 Sept 2020.

  51. Barros SR, Parreira SC, Miranda AF, Pereira AM, Campos NM. New insights in vanishing white matter disease: isolated bilateral optic neuropathy in adult onset disease. J Neuroophthalmol. 2018;38(1):42–6.

    PubMed  Article  Google Scholar 

  52. Bell SM, Barnes K, Clemmens H, Al-Rafiah AR, Al-ofi EA, Leech V, et al. Ursodeoxycholic acid improves mitochondrial function and redistributes Drp1 in fibroblasts from patients with either sporadic or familial Alzheimer's disease. J Mol Biol. 2018;430(21):3942–53.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

Download references


The authors thank Dr. Gil Stynes (IHMRI, and Department of Surgery, Wollongong Hospital) for fibroblast collection and Prof Justin Yerbury, A/Prof Mirella Dottori and Prof Heath Ecroyd for valuable discussions, and Dr. Reece Gately and Rachelle Balez for technical assistance. The authors wish to thank all of the donors, the patients and their families that have made this research possible.


This research was funded by philanthropic donations and a grant from the Illawarra Health and Medical Research Institute (IHMRI). L.O. is supported by a National Health and Medical Research Council (NHMRC) of Australia Boosting Dementia Research Leadership Fellowship (APP1135720).

Author information

Authors and Affiliations



NN – data generation, data analysis, manuscript writing; MC – data generation, data analysis; TB – data generation, data analysis; SM – data generation, data analysis; ME - data generation; DS - data generation; DDH – data analysis; JL – data generation; SSM – data analysis; NSB – data generation; CS – data generation; LO - intellectual input, data analysis, manuscript writing, supervision, funding. All authors reviewed the manuscript. The author(s) read and approved the final manuscript.

Corresponding author

Correspondence to Lezanne Ooi.

Ethics declarations

Ethics approval and consent to participate

All experimental protocols were approved by the University of Wollongong Human Research Ethics Committee (HE17/522). All participants provided consent to participate.

Consent for publication

Participants provided consent for publication.

Competing interests

The authors declare they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ng, N., Cabral-da-Silva, M.C., Maksour, S. et al. Identification of repurposable cytoprotective drugs in vanishing white matter disease patient-derived cells. transl med commun 5, 18 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • Vanishing white matter disease
  • Leukodystrophy
  • Cell stress
  • Drug repurposing
  • Induced pluripotent stem cells
  • Astrocytes