Proteasome inhibitors rescue hypomyelination in iPSC based model of HIKESHI-associated leukodystrophy

Abstract

Introduction

HIKESHI Associated Leukodystrophy (HAL) is a lethal hypomyelinating genetic condition with most patients dying in childhood. Patients with HAL have developmental delay, paraparesis, and high risk of death following fevers and infections. HAL is caused by homozygous loss-of-function mutations in the HIKESHI gene leading to loss of HIKESHI protein. The most seen missense mutation changes Valine to Leucine at residue 54. HIKESHI is a nuclear import carrier for chaperone HSP70 during heat shock response (HSR). As cellular proteins misfold when exposed to stressors, cells initiate HSR with chaperones refolding damaged proteins. As the primary abnormalities in HAL appear to be related to white matter of the brain, we hypothesized that lack of HIKESHI impairs HSR resulting in abnormal oligodendrocyte development and function. To test our hypothesis, iPSCs were produced from affected homozygous HAL patients, non-symptomatic heterozygous first-degree relatives, and healthy unrelated individuals.

Methods

iPSCs from patients with HAL were differentiated into oligodendrocytes, followed by immunofluorescent staining and various assays. RNASeq analyses were done with control, HIKESHI heterozygous, and HIKESHI homozygous cells at various stages of oligodendrocyte differentiation.

Results

Immunofluorescence showed HSP70 and other proteins had altered nuclear localization in HIKESHI homozygous cells upon heat shock. We used RNASeq during oligodendrocyte differentiation and found HIKESHI homozygous cells upregulated various heat shock genes and downregulated neurodevelopmental genes. Homozygous cells also had much fewer Olig2+ cells, aberrant myelination, and enlarged and less circular nuclei. Cells were treated with proteasome inhibitors (PI) in an attempt to rescue mutant HIKESHI protein and restore cellular function. PI treatment restored some HIKESHI protein and upon heat shock, cells displayed nuclear localization of HSP70 indicating restoration of some HIKESHI functionality. PI treatment also rescued Olig2+ cell numbers, myelination, and cell morphology abnormalities.

Conclusion

Loss of HIKESHI results in dysfunction of oligodendrocyte development and function. Repurposed PI treatment may be considered as a novel therapeutic for patients with HAL.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08708-8.

Background

HIKESHI Associated Leukodystrophy (HAL, also known as LEUKODYSTROPHY, HYPOMYELINATING, 13; HLD13, OMIM # 616881) is a rare pediatric neurological condition first reported in 2016 [1]. Neurological symptoms include developmental delay, spastic paraparesis, nystagmus, and acquired microcephaly [1–3]. Distinct from other leukodystrophies, patients with HAL can severely deteriorate when exposed to febrile illness or infection, resulting in hospitalization and sometimes death. Brain MRIs of afflicted individuals show extensive demyelination, thinning of the corpus callosum, and enlarged ventricular spaces. Neuropathological analyses indicate a near absence of myelin, decreased axon density, and decreased oligodendrocyte numbers, as well as increased presence of reactive microglia and astrocytes.

HAL is an autosomal recessive disorder caused by homozygous missense mutations in the HIKESHI gene. Heterozygous carriers for the HIKESHI missense mutation appear healthy with no neurological symptoms, however, detailed studies have not been conducted on this population. Most identified patients with HAL have a Val54Leu mutation in the HIKESHI gene, likely due to a founder effect, especially in those with Ashkenazi Jewish heritage. While not studied worldwide, the reported carrier frequency of HIKESHI Val54Leu mutation in the Ashkenazi Jewish population is 1:216 [3]. Other mutations in the HIKESHI gene also associated with leukodystrophy have been reported, including Cys4Ser, Phe2Val, and Pro78Ser, however these appear to be private mutations in each afflicted family with no known population frequencies [2, 4, 5].

HIKESHI was first identified in 2012 as a nuclear transport carrier for the chaperone protein HSP70 [6, 7]. When cells are exposed to physiological stressors such as thermal stress, oxidative stress, or heavy metals, proteins can misfold which may lead to apoptosis [8]. As a protective measure, cells have evolved a heat shock response (HSR)[9]. During HSR, a specific set of proteins called heat shock proteins (HSPs) are produced and activated to serve as molecular chaperones, helping refold misfolded proteins, and restoring proteostasis [10].

Under normal conditions, protein transport between the cytoplasm and nucleus is mediated by the importin β protein family that interact with the nuclear pore complex [11, 12]. However during thermal stress, importin-mediated nuclear transport is downregulated, yet the chaperone HSP70 still localizes to the nucleus [13–15]. HIKESHI was subsequently identified as a homodimer that interacts with the ATP-bound form of HSP70 in the cytoplasm. HIKESHI then shuttles HSP70 through the nuclear pore complex [16, 17]. In the nucleus, HSP70 is hydrolyzed to the ADP-bound form that results in dissociation from HIKESHI.

Although HIKESHI mutations are causative for HAL and HIKESHI protein normally plays a vital role in heat shock response, specific mechanisms and relevant cell types are poorly understood. Recent studies report that HIKESHI also plays a role in HSP70 nuclear transport and proteostasis during non-stressed conditions supporting a broader function for this protein [18]. Other studies have examined HIKESHI’s contributions to cancer and ischemic stroke [19–22].

Although HIKESHI protein is ubiquitously expressed in human cells, the most striking feature of HAL is hypomyelination. Oligodendrocytes are specialized glial cells in the brain responsible for producing the myelin that encloses neuronal dendrites and axons [23]. Myelinating oligodendrocytes differentiate from oligodendrocyte progenitor cells (OPCs) that migrate through the central nervous system [24]. Compared to other cells, oligodendrocytes have relatively high levels of HSP70 at baseline and following HSR suggesting that HIKESHI-deficient oligodendrocytes might be especially vulnerable to cellular dysfunction and death following stress [25–29]. In support of this model, recent reports using Hikeshi mutant rodents demonstrated altered oligodendrocyte differentiation with lack of nuclear Hsp70 and degradation of Sox10 via FBXW7-mediated ubiquitination [30–33]. As Sox10 is essential for oligodendrocyte development and maturation [34–36], this degradation could reasonably be thought to contribute to the HAL phenotype. Further analyses in complementary model systems including use of human cells with HIKESHI V54L mutations are needed to make fundamental advances leading to advanced therapeutics.

To test the hypothesis that loss of HIKESHI protein impairs HSR leading to abnormal oligodendrocyte development and function, we generated human oligodendrocytes with HIKESHI V54L mutations. We demonstrate that loss of HIKESHI protein results in aberrant oligodendrocyte development in patient-derived oligodendrocytes, resulting in reduced numbers of oligodendrocytes, atypical morphology, gene expression, and decreased myelination. We further demonstrate the therapeutic potential of repurposed proteasome inhibitors that rescued oligodendrocyte numbers and myelination.

Methods

Patient selection

Skin biopsies were conducted under the approval of Health Sciences Review Committee #2 of the Institutional Review Board (IRB) of Vanderbilt University Medical Center, IRB #080369 or Ben Gurion University IRB #0464 − 20 TLV and #0586-20-TLV. All research was performed according to the relevant regulations, including those in accordance with the Declaration of Helsinki. Skin punch biopsies were collected from HAL-affected patients and designated as HIKHOM (HIKESHI homozygous or HIK-/-). Skin biopsies were also obtained from healthy, asymptomatic first-degree relatives who had heterozygous HIKESHI mutations and are designated as HIKHet (HIKESHI heterozygous or HIK+/-). Samples were also collected from unrelated, healthy individuals with a normal HIKESHI genotype and designated as CON (unrelated control or HIK+/+) cells.

iPSC production and validation

iPSCs were reprogrammed from patient derived primary fibroblast cultures utilizing Sendai viral reprogramming and non-integrating episomal vector reprogramming [5, 37]. iPSCs were generated from multiple patients with homozygous V54L mutations in the HIKESHI gene (HIKHOM) as well as multiple first-degree relatives with heterozygous V54L mutations (HIKHet) and unrelated controls (CON). iPSCs were validated with a pluripotency assay and a trilineage embryoid body assay [38–41].

Neuron differentiation

iPSCs were differentiated into neurons using a previously utilized protocol [42, 43].

Oligodendrocyte differentiation

Cells were grown in accordance with a published protocol [44]. iPSCs were grown in 6-well plates in mTeSR1 media (Stem Cell Technologies #85857) until 80% confluent. Cells were disassociated with Accutase (Innovative Cell Technologies AT104-500) at 37 °C for 5 minutes then neutralized with mTeSR1. Cells were centrifuged at 200G for 5 minutes, media was aspirated, and pellet resuspended in mTeSR1. iPSCs were counted on a Countess Cell Counter and plated at 1 × 10⁵ cells per well in a Matrigel-coated sterile 6-well plates. iPSCs were fed daily with mTeSR1 media for 2–3 days, until colonies were ~ 100 μm wide. Neural induction media was then added and changed daily from Day 0–8, then moved to N2 media, feeding daily. On Day 12, cells were disassociated with Accutase and resuspended in N2/B27 media in 6-well low-adherence plates. N2/B27 media was changed every other day from Day 12–20. On Day 20, cells were fed with PDGF media every other day until Day 30. Cells were then plated on nanofiber plates or on poly-L-ornithine and laminin-coated plates. From Day 30–75, media was changed every 2–3 days with PDGF media. From Day 75–90, media was changed every 2–3 days with Glial media.

Nanofiber plating

Oligodendrocytes were grown on randomized nanofibers as they provide a complex 3D matrix that allows cells to spread across the well and prevents clumping. 6-well plates with randomized nanofibers (Mimetix AMS.TECL-002-1X) were washed for 5 min with 70% ethanol. Ethanol was aspirated and 200 µL Poly-L-Lysine (Sigma Aldrich P4707) was added to each well and incubated overnight at 37 °C.

Plates were washed twice with molecular biology grade water (Corning MT46000CM). 200 µL of PDGF media was added and 5–7 Day 30 OPC clusters were transferred into each well and fed every 48 h with PDGF media from Day 30–75. From Day 75–90, cells were fed every 48 h with Glial media. On Day 90, media was aspirated from all wells and 100 µL of 4% PFA added as a fixative. Cells were incubated at room temperature for 30 min before PFA was aspirated and cells were washed 3 times with 1X PBS prior to being used for immunofluorescence. PBS was aspirated and ProLong Glass Antifade Mountant (Invitrogen P36890) was added to Superfrost Plus Microscope slides (Fisher Scientific 22-037-246). Nanofiber discs were removed from the plate using forceps and placed in mounting media and cover-slipped, then allowed to cure for 24 h at room temperature. Nanofiber discs were scanned on Olympus SLIDEVIEW VS200 Scanner at 10x magnification and imaged on 3i MARIANAS Spinning Disk Confocal at various magnifications.

Proteasome inhibitor treatments

Carfilzomib (CAS 868540-17-4, SelleckChem S2853) or marizomib (CAS 437742-34-2, MedChem HY-10985) stock solutions were made at 1 mM in DMSO (dimethyl sulfoxide) and further diluted as needed. Carfilzomib or marizomib were added to media and added to the cells every 48 h at the time points listed in Table 1.

Table 1.

Proteasome inhibitor treatments

Condition	Time Points	Carfilzomib Concentration	Marizomib Concentration
Vehicle	Day 30 – Day 90 Vehicle	Vehicle	Vehicle
30 day treatment	Day 30-Day 60 Drug + Day 60 -Day 90 Vehicle	100 nM	100 nM
2nd 30 day treatment	Day 30-Day 60 Vehicle + Day 60 -Day 90 Drug	100 nM	100 nM
60 day treatment	Day 30 – Day 90 Drug	100 nM	100 nM

Immunofluorescence

Slides were incubated with blocking buffer [1X PBS, 5% Normal Donkey Serum (Sigma-Aldrich D9663), 0.3% Triton X-100] for 1 h at room temperature. Primary antibodies (Table 2) were diluted in the blocking buffer and 100 µL added to each well in a humidified chamber at 4 °C overnight. Cells were washed 3 times with 1X PBS. Secondary antibodies were diluted in blocking buffer and incubated at room temperature for 2 h then washed 3 times with 1X PBS. Hoescht 33,342 (Invitrogen H3570) was added for 5 min at room temperature then washed 3 times with 1X PBS.

Table 2.

Antibodies utilized for immunofluorescence and immunoblotting

Antibody	Catalog Number	Antibody Registry Number	Usage
MBP	Abcam ab7349	AB_305869	1:1000
Olig2	Cell Signaling cst65915	AB_2936997	1:500
β-III-Tubulin	Cell Signaling cst4466	AB_1904176	1:500
MAP2	Cell Signaling cst4542	AB_10693782	1:500
HIKESHI	Proteintech 14808-1-AP	AB_10644295	1:250
HSP40	Cell Signaling cst4868	AB_2094249	1:1000
HSP70	Cell Signaling cst46477	AB_2687559	1:1000
HSP70	Invitrogen MA3-007	AB_325455	1:500
HSP90	Cell Signaling cst4877	AB_2233307	1:1000
GAPDH	Cell Signaling cst97166	AB_2756824	1:2000
Grp78	ThermoFisher PA1-014 A	AB_559362	1:1000

Immunofluorescence images were acquired on 3i MARIANAS Spinning Disk Confocal and EVOS FL Auto microscopes.

Immunoblotting

Cells were washed with cold 1X PBS and scraped off plates, then spun down and resuspended in RIPA Lysis buffer prior to sonication at 10% amplitude for 5–10 s using a QSonica Q55 Sonicator and then incubated on ice for 20 min. Extracts were centrifuged to pellet debris and supernatant removed. Protein concentration of the supernatant was measured using a Bradford microassay with BioRad DC Protein Assay Kit (BioRad #5000111) measured on an Eppendorf Biophotometer. Protein samples were prepared mixing 3X βME RIPA+++ buffer with supernatant and denaturing at 80 °C for 10 min.

Samples and loading dye (BioRad #1610374) were run on Bolt 4–12% Bis-Tris gel (Invitrogen NW04127) with 1X MES Running Buffer (Invitrogen NP0002). Gels were run for 65 min at 165 V and 400 mA then transferred to PVDF membranes (Millipore IPFL00010) using transfer buffer (192 mM Glycine, 75 mM Tris-base, 20% methanol, 80% ddiH2O). Transfer was run for 65 min at 95 V/400 mA.

The membrane was removed and blocked with blocking buffer (LiCor Intercept 927-70001) for 1 h at room temperature. Primary antibodies (Table 1) were diluted in staining buffer (1X TBS, 0.1% Tween20, 5% Bovine Serum Albumin) and incubated overnight at 4 °C. Membranes were washed 3 times with 1X TBS-T. Secondary antibodies were diluted in staining buffer and incubated for 1 h at room temperature. They were then washed 3 times with 1X TBS-T. Blots were scanned on the LICORbio Odyssey F and analyzed and quantified utilizing LICORbio Image Studio 6.0 and GraphPad Prism.

RNA sequencing

Oligodendrocyte cultures were differentiated as previously mentioned. Samples included 3 biological replicates of HIKESHI heterozygous (HIKHet) and homozygous (HIKHOM) lines collected at differentiation days 8, 30, 60, and 90 of the 90-day oligodendrocyte differentiation protocol. These time points demonstrate various stages of development. Cells were harvested from 6 well plates and suspended in 1X PBS. RNA was extracted using the Qiagen RNeasy Kit (Qiagen 74104), Qiagen RNA-free DNase kit (Qiagen 79254), and Qiagen RNeasy Plus Kit (Qiagen 74134). Samples were quantified using a NanoDrop and submitted to the CU Anschutz Genomics Shared Resource. Paired-end reads were obtained with Illumina NovaSeqX PE150 Sequencing with 40M reads per sample.

Raw paired-end RNA-seq reads in FASTQ format were first assessed for overall quality using FastQC [45]. Low-quality bases and adapter sequences were then removed using Trimmomatic [46]. High-quality trimmed reads were subsequently aligned to the human reference genome (GRCh38) using HISAT2 [47], with splice-aware alignment optimized for paired-end data. This process generated sorted and indexed BAM files, followed by the extraction of alignment statistics. Gene-level quantification was performed using featureCounts from the Subread package [48], guided by GTF annotation files from Ensembl (release 106). Counts across all samples were compiled into a matrix for downstream differential expression analysis.

Raw gene-level read counts were imported into R, and genes with very low counts (≤ 50 total across all samples) were removed. Samples were grouped by developmental time point (e.g., d8) and separated by condition, HIKHet or HIKHOM. Differential expression analysis was conducted with DESeq2 [49]. Wald tests generated log₂ fold changes and p-values, which were adjusted using the Benjamini–Hochberg method [50]. For visualization, DESeq2-normalized gene expression values were summarized by computing the mean expression within the HIKHet (control) and HIKHOM (mutant) groups. The mean differential expression of these groups was evaluated, and log₂ fold changes greater than 1/-1 and p < 0.001 were deemed biologically and statistically significant and used to construct heatmaps and volcano plots.

To identify differentially expressed pathways, Gene Ontology (GO) enrichment analysis was performed using clusterProfiler [51]. For each time point, significantly up- and down-regulated genes (p ≤ 0.001 and |log₂FC| > 1) were analyzed separately using enrichGO with ENSEMBL gene identifiers mapped through org.Hs.eg.db.

Image and statistical analysis

Immunofluorescence images were analyzed on ImageJ and Cellprofiler [52–54]. All statistical analysis and graphs were produced using GraphPad Prism version 10.5.0 for Windows (GraphPad Software, Boston, Massachusetts USA). The conducted tests, number of replicates, and p values are detailed in the Figure legends.

Results

Patient-derived cells demonstrate lack of HIKESHI and HSP70 nuclear localization

iPSCs generated from HIKHOM (homozygous HAL patients), HIKHet (first-degree heterozygous relatives), and CON (unrelated controls) were analyzed. Immunoblotting for HIKESHI and HSPs (Fig. 1A) demonstrated a lack of detectable HIKESHI protein in HIKHOM extracts from iPSCs when compared to HIKHet and CON extracts (Fig. 1B). HIKHet cells also appeared to have significantly less HIKESHI protein when compared to CON cells. As the primary function of HIKESHI appears to be the transport of HSP70, we determined the levels of HSP70 from patient-derived iPSCs. We also measured levels of other heat shock proteins including HSP90, HSP40, and Grp78 that interact with HSP70 [55–57]. We found no significant changes across all measured heat shock proteins (Fig. 1C).

Fig. 1.

Characterization of HIKHOM cells phenotype and successful production of MBP+ oligodendrocytes. A. Immunoblotting for HIKESHI and various heat shock proteins in patient-derived iPSCs (n = 4 biological replicates per genotype). B. Immunoblotting quantification demonstrates a lack of detectable HIKESHI protein in HIKHOM iPSC cells and a significant decrease in HIKHet cells (n = 4 biological replicates per genotype, one-way ANOVA analysis with Tukey multiple comparisons, * = p < 0.05, **** = p < 0.0001). C. Immunoblotting and quantification demonstrates no significant changes in heat shock protein amounts between CON, HIKHet, and HIKHOM iPSC cells (n = 4 biological replicates per genotype, one-way ANOVA analysis with Tukey multiple comparisons, ns = p > 0.05). D. Fluorescence staining indicates mislocalization of HSP70 after heat shock in HIKHOM fibroblasts compared to HIKHet and CON fibroblasts (Scale bar 25 μm, representative image per genotype). E. Fluorescence staining shows production of Olig2+ and MBP+ oligodendrocytes (Scale bar 25 μm, representative image per genotype)

Heat shock was induced in patient-derived fibroblasts at 42 °C for an hour [58–61]. Immunofluorescence showed heat shock causing nuclear localization of HSP70 in both CON and HIKHet cells (Fig. 1D). However, there is a lack of nuclear HSP70 in the HIKHOM cells, confirming that lack of HIKESHI protein is preventing nuclear localization of HSP70 during heat shock response in cells. This has also previously been shown in other cell types [1, 2, 6, 7].

As patients with HAL have prominent hypomyelination, we generated oligodendrocytes from CON, HIKHet, and HIKHOM iPSC lines [44]. Oligodendrocyte differentiation was validated with immunofluorescence for the oligodendrocyte lineage marker Olig2 and the myelin marker MBP [62, 63] (Fig. 1E). As this differentiation protocol produces a mixed neural culture [64, 65], immunofluorescence staining also demonstrated some GFAP+ astrocytes and MAP2 + neurons in the cultures.

RNA sequencing demonstrates changes across oligodendrocyte development

To further understand oligodendrocyte development across our genotypes, we performed bulk RNA-Seq across oligodendrocyte differentiation. The most highly upregulated and downregulated differentially expressed genes in HIKHOM versus HIKHet cells were plotted on heat maps and volcano plots (Supplementary Fig. 1). Across all time points, multiple pseudogenes were highly upregulated and downregulated.

On Day 8, we noted an upregulation in multiple protein encoding genes including RTL9, SPTLC3, FSTL4, and CHRNB4 (Fig. 2A). There was also a downregulation in multiple histone encoding genes including H4C2, H4C4, and H2BC7. By Day 30, there is an upregulation in protein encoding genes including UPK1B, AGR2, TFF2, and PAX8 (Fig. 2B). There is also a downregulation in genes AQP1, POTEF, and PRLHR. On Day 60, upregulated protein coding genes include VIP, PTHLH, and HS3ST3A1, and downregulated genes include HCRT, CST1, and PAX2 (Fig. 2C). At the final time point, Day 90, upregulated protein coding genes include SHOX2, P2RY6, and CSPG4 (Fig. 2D). Genes downregulated in HIKHOM cells include FSIP2 and UBB as well as the neurodevelopmental genes NEUROD2 and NEUROD6. On both Day 60 and Day 90, there was an upregulation of genes associated with the heat shock pathway including HSP90AA1 and HSPH1. Both HSP90AA1 and HSPH1 interact with HSP70, and upregulation can be induced by cellular stress. GSTM1 is consistently downregulated across differentiation.

Fig. 2.

HIKHOM oligodendrocyte differentiation shows expression changes over development when compared to HIKHet. RNA was extracted across oligodendrocyte differentiation and sequenced on the NovaSeq 6000 platform. Heat maps were generated after analysis with transcripts with 1/-1 Log₂FC and p < 0.001 values. (A) Expression changes between HIKHet and HIKHOM at Day 8 of differentiation. (B) Expression changes between HIKHet and HIKHOM at Day 30 of differentiation. (C) Expression changes between HIKHet and HIKHOM at Day 60 of differentiation. (D) Expression changes between HIKHet and HIKHOM at Day 90 of differentiation

Gene Ontology analysis was conducted to observe pathway changes (Supplementary Table 1). We found significant upregulation of pathways related to heat shock proteins including protein folding, cellular response to heat, and response to unfolded proteins. GO analysis also demonstrated downregulation of various pathways including nucleosome and telomere organization at Day 8, development and maintenance of the blood-brain barrier at Day 60, and various metabolic processes and synapse maturation at Day 90.

A previous study reported that lack of HSP70 nuclear localization leads to increased Sox10 degradation possibly underlying reduced oligodendrogenesis and myelination [31]. In the current differentiation model, Sox10 protein is detectable around Day 60 of differentiation [44]. Immunofluorescence staining demonstrated Sox10 + Olig2+ cells across all genotypes (Supplementary Fig. 2A). Quantification of Sox10 + Olig2+ cells shows wide variability between all genotypes with no significant changes (Supplementary Fig. 2B). Relative gene expression of SOX10 displays no significant changes between HIKHet and HIKHOM cells across oligodendrocyte differentiation (Supplementary Fig. 2C).

Differentiated oligodendrocytes demonstrate significant phenotypic changes

HIKESHI genotypes were differentiated for 90 days on randomized nanofibers [66–68]. Immunofluorescence of fixed HIKHOM cells revealed much fewer Olig2+ cells and barely detectable MBP+ protein compared to HIKHet and CON cells (Fig. 3A). Total myelin was compared between CON, HIKHet, and HIKHOM oligodendrocytes through analysis of total intensity of MBP+ protein. HIKHOM cells had significantly less detectable MBP+ protein compared to HIKHet and CON cells (Fig. 3B) suggesting a reduction in myelination, similar to brain images seen in HAL patients [1, 3]. HIKHet cells also had significantly less MBP+ protein compared to CON cells, suggesting loss of myelin in the heterozygous phenotype but less pronounced than that seen in HIKHOM cells. Quantification of cell numbers displayed similar changes with HIKHOM cells having significantly fewer Olig2+ cells when compared to HIKHet and CON cells (Fig. 3C). HIKHet cells also have significantly fewer Olig2+ cells when compared to CON cells suggesting a mild heterozygous phenotype.

Fig. 3.

HIKHOM oligodendrocytes display abnormalities in differentiation and physiology. A. Fluorescence staining indicates decreased Olig2+ oligodendrocytes and MBP+ myelin in HIKHOM cells compared to HIKHet and CON cells (Magnification 20x, scale bar 100 μm, representative image per genotype, n = 3–4 cell lines per genotype). B. Total integrated pixel density of MBP+ protein was quantified and analyzed demonstrating significant decrease in MBP+ protein in HIKHOM cells compared to both CON and HIKHet cells with HIKHet cells having less than CON cells. (n = 3 cell lines per genotype, 13–25 images per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, * = p < 0.05, ** = p < 0.01, **** = p < 0.0001). C. Cell count indicates significantly less Olig2+ cells in HIKHOM differentiation (n = 3 cell lines per genotype, 13–25 images per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). D. HIKHOM and HIKHet oligodendrocyte nuclei appear larger than CON oligodendrocytes stained for Olig2 (Magnification 63x, scale bar: 25 μm, representative image per genotype, n = 3 cell lines per genotype). E. HIKHOM oligodendrocyte nuclei are significantly larger than both HIKHet and CON oligodendrocytes with HIKHet oligodendrocytes being larger than CON oligodendrocytes (n = 3 cell lines per genotype, 50–250 cells per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, * = p < 0.05, **** = p < 0.0001). F. HIKHOM oligodendrocyte nuclei are significantly less circular than both HIKHet and CON oligodendrocytes with HIKHet oligodendrocytes being less circular than CON oligodendrocytes (n = 3 cell lines per genotype, 50–250 cells per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, * = p < 0.05, **** = p < 0.0001)

When present, the HIKHOM oligodendrocytes appeared larger across multiple independent differentiations (Fig. 3D). To assess cell nuclei size and morphology, CON, HIKHet, and HIKHOM oligodendrocytes were imaged and size and circularity measured utilizing Olig2+ fluorescence staining. HIKHOM oligodendrocytes were significantly larger and less circular than both HIKHet and CON oligodendrocytes (Fig. 3E, F). HIKHet oligodendrocytes were also significantly larger and less circular than CON oligodendrocytes, indicating morphological changes but still less than the HIKHOM phenotype (Fig. 3E, F).

Carfilzomib treatment improves hyomyelination phenotype

HIKESHI protein is almost undetectable in HIKHOM cells (Fig. 1B) with decreased protein also noted in HIKHet cells when compared to CON. It was previously reported that the relative expression level of HIKESHI mRNA is unchanged across CON, HIKHet, and HIKHOM genotypes [5]. These results suggest the V54L mutation results in protein instability and may result in rapid degradation, possibly by the ubiquitin-mediated proteasome [69]. To test this hypothesis, we treated the cells with proteasome inhibitors to see if HIKESHI protein levels could be increased. We initially used carfilzomib, a FDA-approved selective proteasome inhibitor targeting the 20 S proteasome [70]. Patient-derived neural cells were treated with carfilzomib at various concentrations and time points to determine optimal dosage. Preliminary data treating HIKHOM cells at 24, 48, and 72 h across concentrations from 1 nM to 10 µM was conducted (Supplementary Fig. 3A). Immunoblotting found restoration of HIKESHI protein to around 20% of CON cell amount at 100 nM carfilzomib treatment for 48 h. This value was chosen for further treatments. Immunoblotting also indicated that HSP70 protein steadily increased with increasing carfilzomib dosage, especially at shorter time points. Previous research has indicated that proteasome inhibition induces HSR and increases HSP70 values [71–74]. Therefore, this increase could be a result of the normal cell response rather than restoration of HIKESHI. Carfilzomib-treated cells were then exposed to heat shock and stained to analyze nuclear localization of HSP70 (Supplementary Fig. 3B). Vehicle-treated HIKHOM cells displayed lack of HSP70 nuclear localization, as previously shown. However, carfilzomib treatment restored nuclear localization of HSP70. CON and HIKHet cells demonstrated nuclear localization of HSP70 in both vehicle and carfilzomib-treated conditions.

We treated differentiated oligodendrocytes every 48 h with carfilzomib from Day 30–60 during the 90-day differentiation and saw significant improvement across various metrics. Immunofluorescence staining indicates an increase in Olig2+ cells and MBP+ protein in carfilzomib-treated HIKHOM cells (Fig. 4A). Carfilzomib-treated HIKHOM cells had a significant increase in MBP+ protein compared to Vehicle-treated HIKHOM cells (Fig. 4B). The increased MBP+ protein of carfilzomib-treated HIKHOM cells had no significant difference when compared to vehicle and carfilzomib-treated CON cells implying restoration of myelin production. Both HIKHet and CON cells had no significant changes between vehicle and carfilzomib-treated cells. In contrast, carfilzomib-treated HIKHOM cells had a significant increase in the percentage of Olig2+ cells compared to vehicle-treated HIKHOM cells (Fig. 4C). The percentage of Olig2+ carfilzomib-treated HIKHOM cells had no significant difference when compared to vehicle and carfilzomib-treated CON cells implying restoration of oligodendrocyte numbers. Carfilzomib treatment did not appear to affect the percentage of Olig2+ cells in both CON and HIKHet genotypes.

Fig. 4.

30-day Carfilzomib Treatment results in improvement of physiology and myelination of HIKHOM oligodendrocytes. A. Immunofluorescence indicates carfilzomib-treated HIKHOM cells have increased Olig2+ cells and MBP+ protein (Magnification 20x, scale bar 100 μm, representative image per genotype, n = 3–4 cell lines per genotype). B. Carfilzomib-treated HIKHOM cells show significant improvement in MBP+ protein levels (n = 3 cell lines per genotype, 13–25 images per cell line per treatment, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). C. Cell count indicates significantly more Olig2+ cells in HIKHOM differentiation after carfilzomib treatment (n = 3 cell lines per genotype, 13–25 images per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). D. Carfilzomib-treated HIKHOM and HIKHet oligodendrocyte nuclei are significantly smaller than Vehicle-treated oligodendrocyte nuclei with no difference in the CON oligodendrocytes (n = 3 cell lines per genotype, 50–250 cells per cell line per treatment, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). E. Carfilzomib-treated CON oligodendrocyte nuclei are significantly more circular than vehicle-treated CON oligodendrocytes with no difference in HIKHet and HIKHOM oligodendrocytes (n = 3 cell lines per genotype, 50–250 cells per cell line per treatment, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, *** = p < 0.001, **** = p < 0.0001)

We then analyzed cell morphology in HIKHOM Olig2+ cells (Fig. 4D). Carfilzomib-treated HIKHOM Olig2+ cellular nuclei were significantly smaller than vehicle-treated HIKHOM cells. Carfilzomib-treated HIKHOM cells also appeared to be significantly smaller than vehicle-treated CON cells, but no significant difference was seen when compared to carfilzomib-treated CON cells. Carfilzomib-treated HIKHet cells were also significantly smaller than vehicle-treated HIKHet cells with the treated HIKHet cells having no significant difference when compared to vehicle and carfilzomib-treated CON cells. CON Olig2+ cells showed no significant change in nuclear size with carfilzomib treatment. However, carfilzomib increased circularity in CON cells whereas both HIKHet and HIKHOM cells had no significant circularity changes seen (Fig. 4E).

All HIKESHI genotypes were also treated with 100 nM carfilzomib every 48 h from Day 60–90 or from Day 30–90 of the 90-day oligodendrocyte differentiation protocol (Supplementary Fig. 4). Day 60–90 treatment resulted in reduced numbers of Olig2+ cells and MBP+ protein across all HIKESHI genotypes suggesting that later treatment might result in increased cell death. However, there were DAPI+ cells present indicating other neural cells were present. 60 days of treatment from Day 30–90 appeared toxic and resulted in lack of detectable Olig2+ cells and MBP+ cells. There were very few DAPI+ cells in all 3 genotypes indicating that long-term carfilzomib treatment highly impacted cell differentiation and may cause cell death in multiple lineages.

Marizomib treatment improves hypomyelination phenotype

As carfilzomib appears to minimally cross the blood brain barrier, it may have limited therapeutic potential in patients with neurological disorders [74, 75]. Therefore, we then used marizomib, another proteasome inhibitor. Marizomib is a naturally derived proteasome inhibitor targeting the 20 S proteasome [76] that is currently going through clinical trials for cancer and critically for translational purposes, is blood brain barrier penetrant [77, 78].

All HIKESHI genotypes were treated with 100 nM marizomib with the media changed every 48 h during the 90-day oligodendrocyte differentiation protocol. Immunofluorescence indicates that treating cells from Day 30–60 of differentiation appears to increase the number of Olig2+ cells and amount of MBP+ protein in the HIKHOM cells (Fig. 5A). Marizomib-treated HIKHOM cells had significantly higher levels of MBP+ protein compared to vehicle-treated HIKHOM cells (Fig. 5B). MBP+ levels of marizomib-treated HIKHOM cells had no significant difference when compared to both vehicle-treated and marizomib-treated CON cells, indicating restoration of myelin production to normal amounts. Marizomib-treated HIKHet cells had no significant difference when compared to vehicle-treated HIKHet cells. While vehicle-treated HIKHet cells had significantly less MBP+ protein than vehicle-treated CON cells, marizomib-treated HIKHet cells show no significant difference with vehicle-treated or marizomib-treated CON cells, indicating an improvement in MBP+ production. Marizomib-treated CON cells had no significant difference when compared to vehicle-treated CON cells.

Fig. 5.

30-day Marizomib Treatment results in improvement of physiology, cell count, and myelination of HIKHOM oligodendrocytes. A. Immunofluorescence indicates marizomib-treated HIKHOM cells have increased Olig2+ cells and MBP+ protein (Magnification 20x, scale bar 100 μm, representative image per genotype, n = 3–4 cell lines per genotype). B. Marizomib-treated HIKHOM cells show significant improvement in MBP+ protein levels (n = 3 cell lines per genotype, 13–25 images per cell line per treatment, mean with SEM, one-way ANOVA analysis, Tukey multiple comparisons test, ns = not included, * = p < 0.05, ** = p < 0.01, **** = p < 0.0001). C. Cell count indicates significantly more Olig2+ cells in HIKHOM differentiation after marizomib treatment (n = 3 cell lines per genotype, 13–25 images per cell line, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001). D. Marizomib-treated HIKHOM and HIKHet oligodendrocyte nuclei are significantly smaller than Vehicle-treated oligodendrocyte nuclei with no difference in the CON oligodendrocytes (n = 3 cell lines per genotype, 50–250 cells per cell line per treatment, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, * = p < 0.05, *** = p < 0.001, **** = p < 0.0001). E. Marizomib-treated CON oligodendrocyte nuclei are significantly more circular than vehicle-treated CON oligodendrocyte nuclei with no difference in HIKHet and HIKHOM oligodendrocyte nuclei (n = 3 cell lines per genotype, 50–250 cells per cell line per treatment, mean with SEM, one-way ANOVA analysis with Tukey multiple comparisons, ns = not included, * = p < 0.05, *** = p < 0.001, **** = p < 0.0001)

Similar to carfilzomib, marizomib-treated HIKHOM cells also had a significant increase in the percentage of Olig2+ cells compared to vehicle-treated HIKHOM cells (Fig. 5C). The percentage of Olig2+ marizomib-treated HIKHOM cells were not statistically different from vehicle-treated CON cells implying restoration of oligodendrocyte numbers. Marizomib treatment did not appear to affect the percentage of Olig2+ cells in both CON and HIKHOM genotypes, although there is a trending decrease in the CON cells.

Marizomib treatment for 30 days also changed the morphology of Olig2+ cells. Marizomib-treated HIKHOM Olig2+ cellular nuclei were significantly smaller than vehicle-treated HIKHOM cells (Fig. 5D). Marizomib-treated HIKHOM cells were significantly smaller than vehicle-treated CON cells but had no significant difference with the marizomib-treated CON cells. Marizomib-treated HIKHet cells were significantly smaller than the vehicle-treated HIKHet cells and had no significant difference with the vehicle and marizomib-treated CON cells indicating that the morphology has been restored to normal values. Marizomib-treated CON cells had no significant changes when compared to vehicle-treated CON cells. Marizomib treatment appeared to result in varying results in nuclear circularity, with the CON cells showing no significant difference, the HIKHet cells showing a significant decrease in circularity, and the HIKHOM cells showing a significant increase (Fig. 5E). Marizomib-treated HIKHOM cells had no significant difference with the vehicle and marizomib-treated CON cells indicating that the marizomib treatment restored both nuclear area and circularity to CON cell levels.

Finally, all HIKESHI genotypes were treated every 48 h with 100 nM of marizomib from Day 60–90 or from Day 30–90 of the 90-day oligodendrocyte differentiation (Supplementary Fig. 5). The 60-day treatment appeared toxic and resulted in decrease of detectable Olig2+ and MBP+ cells with reduced DAPI+ cells across all 3 genotypes. The 60-day marizomib treatment appeared to be less toxic than the 60-day carfilzomib treatment. Day 60–90 treatment displayed a reduced amount of Olig2+ cells and MBP+ protein across all HIKESHI genotypes indicating that later treatment might result in increased oligodendrocyte cell death. However, there were abundant DAPI+ cells suggesting the ongoing presence of other neural cells.

Discussion

HIKESHI Associated leukodystrophy is a devastating and progressive neurogenetic disorder. Afflicted patients have markedly decreased white matter, and autopsy studies show decreased myelin as well as oligodendrocytes. An additional feature is decompensation or even death following a febrile illness or infection. This may be due to the fundamental role of HIKESHI during HSR. This study further elucidates the phenotype of HAL by detailing patient-derived oligodendrocyte and neural models. Our analyses indicates that patient-derived cells do not have global changes in mRNA expression and display expected markers across oligodendrocyte development. However, HIKHOM oligodendrocytes are produced in fewer numbers, have decreased myelination, and aberrant morphology. As part of our study, we tested the hypothesis that HIKESHI V54L protein is unstable and may be rapidly degraded by the proteasome. By partially rescuing HIKESHI levels, we also show that proteasome inhibitors may be potential therapeutics with restoration of oligodendrocyte numbers, myelination, and cell morphology.

Although HIKESHI protein is barely detectable in HIKHOM patient samples and also decreased in HIKHet samples, HSP70 and other interacting heat shock proteins had no changes implying that although HIKESHI is critical for transporting HSP70 to the nucleus, lack of HIKESHI does not affect HSP70 stability. Additional feedback mechanisms impacting HSP gene regulation are implied by the changes we see in oligodendrocyte mRNA expression.

We expected HIKHet samples to have 50–100% of wild-type HIKESHI protein levels but we only measured 25–30% of the levels seen in control cells. This suggests the postulated degradation of mutant HIKESHI V54L protein in HIKHet cells may also lead to loss of some wild-type protein as well. As HIKESHI normally functions as a homodimer, the HIKESHI wildtype/V54L heterodimer might also be unstable and thus targeted for degradation. This may explain the observed 25% expression levels seen in HIKHet cells if it consists only of HIKESHI wildtype/wildtype homodimers. This observation needs further replication as it has therapeutic implications since heterozygous V54L relatives appear asymptomatic and healthy. Future precision therapies may then only require partial restoration of functional HIKESHI to restore homeostasis.

To expand our results, we are developing mouse models of HAL with Hikeshi knockout, Hikeshi floxed, and V54L encoding knockin alleles. We plan to use these preclinical models to further examine the phenotype of HAL including lineage-specific transgenic mice to assess the impact and timing of Hikeshi loss in various cell types. Analysis of HIKHOM cells during oligodendrocyte development indicates several important changes including alterations in heat shock protein related pathways related to protein folding and response to heat. However, comparison to HIKHet oligodendrocytes resulted in a modest list of differentially expressed genes across all time points. Genes relating to heat shock such as HSP90 and HSPH1 were upregulated at later time points. HSP90 is highly regulated and can be upregulated by a variety of processes, including increased stress or injury mediated by the NF-κB family of transcription factors [79]. This suggests that HIKHOM oligodendrocytes even in the absence of heat shock, are experiencing cellular stressors that result in alteration of gene expression. For example, neurodevelopmental genes such as NEUROD2, NEUROD6, and PAX2 were downregulated. Although these genes are best understood to regulate neuronal differentiation, this could suggest an impact to oligodendrocyte differentiation and function [80, 81]. There appear to be multiple genes differentially expressed that relate to other systems including FSIP2, CST1, and VIP. These variations might be a result of the differentiation protocol or might indicate an unknown role for these genes in oligodendrocyte and HIKESHI-related pathways. As the differentiation protocol generates a mixed neural culture, some changes in gene expression might be tied to changes in the neurons or astrocytes. Future studies would have to analyze isolated oligodendrocytes at specific time points though they do not differentiate appropriately when isolated over extended periods of time. Single-cell RNA sequencing could also be used in the future to more precisely define transcriptional changes at both baseline, following heat shock, as well as following proteasome inhibitor or gene therapy treatments.

Previous studies have demonstrated lack of Hikeshi in a rodent model results in decreased oligodendrogenesis and myelination due to increased Sox10 degradation [31]. In contrast to these results, we did not observe changes in Sox10 expression during oligodendrocyte development and found Sox10 + Olig2+ cells across all genotypes. This indicates that Sox10 is still present in the HIKHOM cells although in fewer cells. However, increased Sox10 degradation at an earlier time point could cause a smaller percentage of oligodendrocytes developing, resulting in the decreased cell numbers we observe. This would have to be further analyzed at different time points and determine the role of Sox10 in the HAL human iPSC models.

HIKHOM oligodendrocytes were positive for expected developmental markers including Pax6 at Day 8, O4 at Day 75, and MBP at Day 90 (data not shown). However, at the end of differentiation at Day 90, we observed significant changes in oligodendrocyte number and myelination. This implies that oligodendrocytes are appropriately developing, however there might be higher levels of oligodendrocyte progenitor cell death during development and the resulting decrease in cell number could underlie the decrease in myelination.

Prior studies have reported that heat stress in rodent models, especially during juvenile ages, resulted in decreased myelination [82, 83]. Aberrant stress responses have also been implicated in other demyelinating conditions such as Pelizaeus-Merzbacher disease and Vanishing White Matter disease with changes in ER stress resulting in myelin loss [84–85]. Heat stress also results in increased cell death in Schwann cells that myelinate the peripheral nervous system [86]. The noted changes in cell morphology such as increased nuclear size and decreased circularity could result from various causes including hypoxia, senescence, apoptosis, and changes to the mTOR signaling pathway [87–89]. These possibilities are being investigated in our cell-based model as well as mouse models of HAL.

Although we observed a decrease in total presence of MBP+ protein, we were unable to assess the quality of the myelin being produced. The oligodendrocytes produced in the nanofiber and 2-D culture system were unable to be successfully processed for electron microscopy to assess the quality of the myelin sheath [90]. Further assays of the length of the sheath, number of axons wrapped, and Sholl analysis are also required. This can also be further studied in the future utilizing our newly developed mouse models, either through analysis of mouse oligodendrocytes or from transplantation of human-derived oligodendrocytes into the mouse brain to assess intergration and function [91–93].

Treatment with either carfilzomib and marizomib resulted in significant rescue of the phenotype with increased oligodendrocyte numbers, increased myelination, and decreased nuclear size, with most metrics returning to CON levels. This implies that inhibiting the proteasome prevents degradation of HIKESHI, restoring it to a critical threshold with even low levels sufficient to improve oligodendrocyte development and function. Proteasome inhibitors would be expected to alter levels of many proteins and specificity of these drugs to only Hikeshi rescue seems unlikely. Further experiments to assess HIKESHI ubiquitination and degradation will be needed in both iPSC and mouse based systems. Although the tested dosage of 100 nM every 48 h did prove effective, further studies need to be conducted on optimizing the dosage for different cell types and analyzing the effect of long-term cellular exposure. While the tested dosage did restore HSP70 nuclear localization in almost all cells, the percentage of cells demonstrating proper localization and the duration of effect could also be analyzed. Proteasome inhibitors are a drug category targeting the core 20 S proteasome [94]. Currently there are multiple drugs in this pathway including bortezomib, carfilzomib, marizomib, ixazomib, and oprozomib [95]. While currently no proteasome inhibitors have been FDA-approved for pediatric usage, older generation proteasome inhibitors such as bortezomib have been utilized in off-label compassionate use and demonstrated an acceptable safety profile, with pediatric patients tolerating the medication better than adults [96]. There are also ongoing clinical trials evaluating its potential as pediatric cancer therapeutics [97, 98].

Carfilzomib has multiple side effects including fatigue, headache, and nausea as well as more severe side effects including peripheral neuropathy, cardiotoxicity, and respiratory adverse events [99–102]. Trials are ongoing for use in pediatric cancer with potential use in relapsed acute lymphoblastic leukemia [103]. Side effects noted for marizomib also include fatigue, nausea, and vomiting though other side effects such as peripheral neuropathy and hematologic toxicity are decreased compared to other proteasome inhibitors [78, 104]. Clinical studies demonstrated tolerance for treatment times of over 4 months, however, the long-term side effects on pediatric patients are unknown. As newer proteasome inhibitors are currently being tested and approved, including the orally administered ixazomib and oprozomib, additional clinical trials will need to be conducted on determining safety and efficacy [105, 106]. Due to the lack of long-term side effect information and the possible cytotoxicity due to high levels of proteasome inhibition, we speculate that any treatment for HAL may be most efficacious and well tolerated as a short-term intervention during febrile illness or infections to prevent significant clinical deterioration. We will utilize mouse models to further test potential interventions with proteasome inhibitors as well as other genetic approaches. Mouse models can also be utilized to determine whether there is an appropriate therapeutic window for proteasome inhibition during neural development or whether such inhibition could better serve as a rescue medication.

Additional structural studies are required to assess how the V54L mutation affects the HIKESHI protein. The relatively small change in the HIKESHI protein structure due to the single amino acid change from valine to lysine appears to drastically alter function. We hypothesized this could be related to ubiquitin-mediated protein degradation [107]. This mechanism could explain the lack of detectable protein in the HIKHOM cells and how proteasome inhibition partially restores function.

Although HIKESHI heterozygous carriers appear healthy, their cells had a possible mild phenotype in our findings across various metrics. This may be due to the various additional stressors induced during in vitro cell culture when compared to embryonic development [108]. This could also be due to the isolation of the neural cells in our cell culture and lack of other interacting factors including microglia and vascularization. This also implies the need for more complex and physiologically relevant model systems such as organoids and pre-clinical animal models. Alternatively, the HIKHet results compared to unrelated CON cells may also be due to differing genomic backgrounds. We mitigated this possibility by using iPSCs derived from multiple patients with HAL and their first-degree relatives with heterozygous mutations in the HIKESHI gene. We attempted to generate isogenic iPSC to further circumvent this issue but the presence of several HIKESHI pseudogenes has thwarted specific CRISPR based targeting.

Further research into the iPSC models of HAL could also open new therapeutic approaches including cell-based therapies such as stem cell transplants or oligodendrocyte transplants. Currently, genetically corrected stem cell transplants are being utilized in other leukodystrophies such as Metachromatic Leukodystrophy with remyelination seen [109–111]. Other leukodystrophies have also seen advances in gene therapy and stem cell transplant-based approaches [112]. Recent reports demonstrated that iPSC-derived glial progenitor cell transplants improved oligodendrocyte numbers and remyelination in white matter stroke models [113, 114]. Although HAL is an ultra-rare condition, further research into the role of HIKESHI in cell biology and development should provide valuable insights into oligodendrocyte development and function as well as a wide range of diseases. HSP70 for example is a highly abundant protein and has been implicated in a variety of other conditions including cancer, neurodegenerative diseases, diabetes, and cardiovascular conditions [115–117]. There is also ongoing research on producing HSP70 targeting drugs for a variety of conditions [118]. Other interacting chaperones have also been implicated in a range of diseases termed ‘chaperonopathies’ which include protein misfolding neurodegenerative disorders such as Alzheimer’s and Parkinson’s disease [119, 120].

In conclusion, we produced an oligodendrocyte model system to study HAL. In our system, we found marked changes in oligodendrocyte development and physiology and successfully rescued this phenotype through repurposing of proteasome inhibitors. Due to the high morbidity and mortality of HAL, future optimization of this therapeutic approach as well as the development of other pharmaceutical and gene therapies are necessary to hopefully improve the quality and lifespan of patients with HAL. Use of stem cell and other therapeutics that initiate brain composition and increase remyelination may also be beneficial to HAL patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HAL

HIKESHI Associated Leukodystrophy

HSR

Heat Shock Response

PI

Proteasome Inhibitors

HSPs

Heat Shock Proteins

OPC

Oligodendrocyte Progenitor Cell

CON

Unrelated control, HIK+/+ genotype

HIKHet

HIKESHI Heterozygous, HIK+/–genotype

HIKHOM

HIKESHI HOMozygous, HIK–/–genotype

iPSCs

induced Pluripotent Stem Cells

PFA

Paraformaldehyde

GEO

Gene Expression Omnibus

Author contributions

SRK and KCE conceived and designed the study. SRK performed experiments as well as data collection. TR and GDV generated and characterized iPSCs. SRK, MMW, and KCE analyzed the data and interpreted the results. SRK and KCE wrote the initial manuscript draft. All authors reviewed the results and approved the final version of the manuscript.

Funding

We thank the generous donors who supported this project. They wish to remain anonymous. Reprogramming at BGU was supported by the Israel Science Foundation (ISF) grant 923/22 and the European Leukodystrophy Association (ELA) grant 2023-025I3.

Data availability

RNA Sequencing data was deposited in Gene Expression Omnibus (GEO) under accession GSE313678.

Declarations

Ethics approval and consent to participate

Approval of Health Sciences Review Committee #2 of the Institutional Review Board of Vanderbilt University Medical Center, IRB #080369 or Ben Gurion University IRB #0464 − 20 TLV and #0586-20-TLV. Consent to participate was given by subjects and their families.

Consent for publication

Not applicable.

Competing interests

Not applicable.