Loss of Oxidation Resistance 1, OXR1, Is Associated with an Autosomal-Recessive Neurological Disease with Cerebellar Atrophy and Lysosomal Dysfunction

Abstract

We report an early-onset autosomal-recessive neurological disease with cerebellar atrophy and lysosomal dysfunction. We identified bi-allelic loss-of-function (LoF) variants in Oxidative Resistance 1 (OXR1) in five individuals from three families; these individuals presented with a history of severe global developmental delay, current intellectual disability, language delay, cerebellar atrophy, and seizures. While OXR1 is known to play a role in oxidative stress resistance, its molecular functions are not well established. OXR1 contains three conserved domains: LysM, GRAM, and TLDc. The gene encodes at least six transcripts, including some that only consist of the C-terminal TLDc domain. We utilized Drosophila to assess the phenotypes associated with loss of mustard (mtd), the fly homolog of OXR1. Strong LoF mutants exhibit late pupal lethality or pupal eclosion defects. Interestingly, although mtd encodes 26 transcripts, severe LoF and null mutations can be rescued by a single short human OXR1 cDNA that only contains the TLDc domain. Similar rescue is observed with the TLDc domain of NCOA7, another human homolog of mtd. Loss of mtd in neurons leads to massive cell loss, early death, and an accumulation of aberrant lysosomal structures, similar to what we observe in fibroblasts of affected individuals. Our data indicate that mtd and OXR1 are required for proper lysosomal function; this is consistent with observations that NCOA7 is required for lysosomal acidification.

Keywords: Drosophila; MiMIC; NCOA7; T2A-GAL4; TLDc; V-ATPase; mustard; oxidative stress; seizures; speech delay.

Figure 1

Five Affected Individuals Have Bi-Allelic Loss-of-Function Variants in OXR1 (A) Pedigrees from three families show affected individuals and the OXR1 variants inherited. In Family 1, the parents are each carriers for a single copy of a different variant (one splice site and one nonsense variant) in OXR1, and they have one affected daughter who inherited both variants. In Family 2, the parents are consanguineous and are carriers of the same variant. Family 2 has two unaffected males and one affected male, indicated as Individual 2. In Family 3, the parents are also consanguineous and are carriers of the same variant. Family 3 has one affected female (Individual 3.1) and two affected males (Individuals 3.2 and 3.3). Individual 3.4 is an unaffected female sibling who is a carrier of the variant. (B) Magnetic resonance imaging (MRI) images from Individual 1 when she was 15 years old. Three perspectives are shown (from left to right): coronal, sagittal, and transverse. On coronal and sagittal views, we note the marked cerebellar atrophy as can be evidenced by the increased cerebrospinal fluid space above and below the cerebellum (arrowheads), and the prominent cerebellar folia. On the transverse view, we see wide cerebellar fissures. (C) OXR1 encodes a protein with 846 amino acids (RefSeq NM_018002.3). It has three protein domains: LysM (aa 97 to 142), GRAM (aa 207 to 274), and TLDc (aa 684 to 846). The four variants found in affected individuals are mapped to their locations on the protein. The green bars indicate the antigens that the two antibodies used in Figure 1D immunoblots were raised against. (D) OXR1 has six validated RefSeq mRNA transcripts. Three transcripts (in black) are not expressed in the central nervous system (CNS), and the other three transcripts (in magenta) are expressed in the CNS. Each of the OXR1 variants is mapped onto the transcripts where all three transcripts that are expressed in the CNS contain all four variants. The protein of the second variant, RefSeq NM_0.18002.3 or RefSeq NP_060472.2, is shown in Figure 1D. (E) Two antibodies were used to detect OXR1 protein levels in fibroblasts from four control individuals, including an unaffected sibling from Family 3. The antibodies used are: Bethyl A302-035A (epitope 175- 225) and Sigma HPA027395 (epitope 267- 359). GAPDH is used as loading control. No truncated proteins can be detected on immunoblots with either antibody. At least three biological replicates were tested, and representative images are shown.

Figure 2

Loss-of-Function Alleles of mtd Are Lethal and Can Be Rescued by a Short Form of Human OXR1 (A) mtd fly gene structure and mtd-SA-T2A-Gal4-polyA allele. Twenty-three isoforms are annotated in FlyBase. A Minos-mediated integration cassette (MiMIC) inserted into an intron provided a landing site for the SA-T2A-Gal4-polyA-containing cassette. Recombinase-mediated cassette exchange (RMCE) was performed with PhiC31 integrase to exchange the cassettes to generate Mi{Trojan-GAL4.0}mtd^{MI02920-TG4.0} (mtd-T2A-Gal4). (pA, poly(A) signal; SA, splice-acceptor; 2A, viral T2A peptide.) (B) The Mi{Trojan-GAL4.0}mtd^{MI02920-TG4.0}/ Df(3R)3-4 (abbreviated mtd-T2A-Gal4 / Df(3R)3-4) flies mostly die as pupae, and some escapers fail to eclose. Those that did escape their pupa display wrinkled wings. (C) Quantification of phenotypes shown in (B). Progeny from three independent crosses were counted. Escapers that reach adulthood die within the first week of life. (D) mtd mutant phenotypes (mtd-T2A-Gal4 / Df) can be fully rescued by reintroducing 80kb fly genomic construct (CH321-04E08), fly mtd-RH cDNA, human OXR1.HA cDNA, or human NCOA7.HA cDNA. Each data point is an independent cross with at least 100 progeny. Percent of expected eclosion is calculated based on the number of observed over expected flies with the genotype on the x axis. The expected number of flies is calculated based on Mendelian ratios multiplied by the total number flies and/or progeny. A one-way ANOVA statistical test was conducted, and the p value is <0.01. (E) Bang sensitivity is measured by quantifying the percentage of flies able to return to an upright position after 10 s of vigorous shaking on a vortex instrument. 10 flies per data point with at least 10 data points per group were tested. One-way ANOVA statistical test was conducted and the p value is < 0.0001. (F) Climbing is measured as the percentage of flies able to climb 15 cm within 20 s. Each data point is the average of three trials with 10 flies. A one-way ANOVA statistical test was conducted, and the p value is <0.0001. For (D), (E), and (F), two-tailed Student’s t tests were conducted for data analysis. ns indicates p value >0.05, ^∗ indicates p value ≤ 0.05, ^∗∗ indicates p value ≤ 0.01, ^∗∗∗ indicates p value ≤ 0.001, ^∗∗∗∗ indicates p value ≤ 0.0001. Error bars represent the standard error of the mean.

Figure 3

Neuronal Knockdown of mustard Leads to Selective R7 Defects and Cell Loss (A) The lifespan of flies that express two different RNAis against mtd. The neuronal drivers used are elav-GAL4 and Nsyb-GAL4. The RNAi lines used are RNAi-1 (HMS01666) and RNAi-2 (GL00665). n > 100 per group from three independent fly crosses. Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests were conducted for each experimental group compared to controls (Elav > Empty) and the p values are <0.0001 for all three with both tests. The error bars represent standard errors. (B) Adult fly heads at day 9 were sectioned and stained with hematoxylin and eosin (H&E). Flies with neuronal knockdown of mustard exhibit severe vacuolization. At least three biological replicates were examined, and representative images are shown. (C) Transmission electron microscopy (TEM) of the retina of Elav > empty or Elav > mtd RNAi adult fly heads at days 5 and 8. Photoreceptors are labeled on the left, and R7 is labeled is red. The images on the bottom show details of Elav > empty or Elav > mtd RNAi R7 photoreceptors. The representative images shown are consistent with results from three biological replicates. (D) Quantification of the percentage of ommatidia with abnormal R7 cells. At least three biological replicates per group were quantified. (E) Abnormal autophagosomes in R7 photoreceptors were quantified, and an increased number of autophagosomes were found in neuronal mtd knockdown conditions. At least 14 R7 photoreceptors per genotype were quantified for each group. Three biological replicates per group were processed in parallel with the same conditions and imaging settings. Two-tailed Student’s t tests were conducted for data analysis. ^∗∗∗∗ indicates p value ≤ 0.0001. A one-way ANOVA statistical test was conducted, and the p value is <0.0001.

Figure 4

mtd Mutant Brains Show Mild Elevation of Oxidative Stress and Antioxidant Feeding Does Not Increase Lifespan of Neuronal mtd Knockdown Flies (A) Elav (neuronal) > mtd RNAi knockdown and Elav > luciferase RNAi (control) flies were fed with oxidants that are known to induce reactive oxidative species. Rotenone concentrations are 100μM. Each group has >100 flies from three independent crosses. The error bars represent standard errors. Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests were conducted for the survival curves. Comparing mutant (elav > mtd RNAi) and control (elav > luciferase RNAi) flies fed with normal food, the p values are <0.0001 for both tests. Comparing mutant (elav > mtd RNAi) and control (elav > luciferase RNAi) flies fed with 100μM Rotenone, the p values are <0.0001 for both tests. Comparing flies fed with 5% H₂O_2, the p values are >0.05 for both tests. Comparing flies fed with 15% H₂O₂, the p values are >0.05 for both tests. Comparing control flies fed with normal food to control flies fed with oxidants, the p values are <0.0001 for both tests. Comparing mutant flies fed with normal food to mutant flies fed with oxidants, the p values are <0.0001 for both tests. Luc, luciferase. (B) To test the susceptibility of OXR1-deficient fibroblasts to oxidative stress, fibroblasts were treated with 500μM H₂O₂. Dead fibroblasts’ nuclei are marked by propidium iodide (PI), and total number of fibroblasts are counted by number of Hoechst positive nuclei. At least three fields are quantified from each of three biological replicates for each group. The experiments were conducted under the same conditions with the same confocal microscope settings. Two-tailed Student’s t tests were conducted for data analysis. ns indicates p value > 0.05, ^∗ indicates p value ≤ 0.05. Error bars represent the standard error of the mean. A one-way ANOVA statistical test was conducted and the p value is <0.0001. (C) The top row shows wild-type whole adult fly brains that were either fully oxidized with 2mM diamide (DA) or fully reduced with 20mM dithiothreitol (DTT) (dark purple = reduced state, yellow = oxidized state). The second and third rows show 2–3 day old mtd-T2A-GAL4 / + and mtd-T2A-GAL4 / Df(3R)3-4 fly brains with four different genetically encoded oxidative stress sensors. The experiments were conducted under the same conditions with the same confocal microscope settings, and representative images are shown. (D) Quantification of 405nm to 488nm ratio of each in vivo encoded reportes for either control (mtd-t2a-gal4 / +) or mutant (mtd-t2a-gal4 / Df(3R)3-4). At least five flies per group were measured. Two-tailed Student’s t tests were conducted for data analysis. ns indicates p value > 0.05, ^∗ indicates p value ≤ 0.05, ^∗∗ indicates p value ≤ 0.01. Error bars represent the standard error of the mean. (E) Adult head extracts from day 5 elav > luciferase RNAi and elav > mtd RNAi flies were quantified for aconitase activity. Three biological replicates were measured. Two-tailed Student’s t tests were conducted for data analysis. ns indicates p value > 0.05. Error bars represent the standard error of the mean. (F) Elav (neuronal) > mtd RNAi knockdown flies were fed with DMSO and four anti-oxidation drugs. Each group has > 100 flies from three independent crosses. NAC (n-acetyl cysteine); tBHQ (tert-Butylhydroquinone); D3T (3-H-1,2-dithiole-2-thione). Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests were conducted for the survival curves. Compared with DMSO (control), the Log-rank (Mantel-Cox) p values for NAC and tBHQ are <0.0001 for NAC, and the Log-rank (Mantel-Cox) p values for Oltipraz and D3T are <0.05. The Gehan-Breslow-Wilcoxon p values for NAC and tBHQ are <0.0001 for NAC, the p value for D3T is ≤0.001, and the p value for Oltipraz is >0.05. The error bars represent standard errors.

Figure 5

Human Fibroblasts that lack OXR1 Accumulate Lysosomes (A) Human fibroblasts were stained with Lysotracker staining (red) and DAPI (blue). Yellow arrowheads indicate lysosomes. Scale bar = 10μm. The representative images shown are consistent with results from three independently performed experiments. The images were obtained from samples processed in parallel in the same experiment and acquired with identical microscopy settings. (B) LysoTracker staining of fibroblasts was quantified by number of LysoTracker punctae in a field divided by number of nuclei. At least 10 fields from three biological replicates were counted for each group. The experiments were conducted under the same conditions with the same confocal microscope settings. A one-way ANOVA statistical test was conducted, and the p value is <0.0001. In addition, two-tailed Student t tests were conducted where ns indicates ^∗∗ indicates P value ≤ 0.01. Error bars represent the standard error of the mean. (C) Transmission electron microscopy (TEM) images of control (C’) and OXR1-deficient (C’’ and C’’’) fibroblasts. Scale bar = 600nm. (D) Quantification of autolysosomes/ endolysosomes in C’-C’’’ and unaffected sibling from family 3 (Individual 3.4). One-way ANOVA statistical test was conducted and the p value is < 0.001. In addition, two-tailed Student t tests were conducted where ns indicates p value > 0.05, and ^∗∗ indicates p value ≤ 0.01. Error bars represent the standard error of the mean.