Amyotrophic lateral sclerosis (ALS) linked mutation in Ubiquilin 2 affects stress granule assembly via TIA‐1

Abstract Aims The ubiquilin‐like protein ubiquilin 2 (UBQLN2) is associated with amyotrophic lateral sclerosis and frontotemporal degeneration (ALS/FTD). The biological function of UBQLN2 has previously been shown to be related to stress granules (SGs). In this study, we aimed to clarify the regulatory relationship between UBQLN2 and SGs. Methods In this study, we transfected UBQLN2‐WT or UBQLN2‐P497H plasmids into cell lines (HEK293T, HeLa), and observed the process of SG dynamics by immunofluorescence. Meanwhile, immunoblot analyses the protein changes of stress granules related components. Results We observed that ubiquilin 2 colocalizes with the SG component proteins G3BP1, TIA‐1, ATXN2, and PABPC1. In cells expressing WT UBQLN2 or P497H mutants, in the early stages of SG formation under oxidative stress, the percentage of cells with SGs and the number of SGs per cell decreased to varying degrees. Between WT and mutant, there was no significant difference in eIF2α activity after stress treatment. Interestingly, the UBQLN2 P497H mutant downregulates the level of TIA‐1. In addition, the overexpression of the UBQLN2 P497H mutant inhibited the phosphorylation of 4E‐BP1 and affected the nucleoplasmic distribution of TDP‐43. Conclusions Ubiquilin 2 colocalizes with the SG component proteins G3BP1, TIA‐1, ATXN2, and PABPC1. It participates in regulating SG dynamics. And UBQLN2 mutation affects the assembly of stress granules by regulating TIA‐1. In addition, the overexpression of the UBQLN2 P497H mutant inhibited the phosphorylation of 4E‐BP1 and affected the nuclear and cytoplasmic distribution of TDP‐43. These provide new insights into the role of UBQLN2 in oxidative stress and the pathogenesis of ALS.


| INTRODUC TI ON
Amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig's disease, is an adult-onset neurodegenerative disease caused by loss of motor neurons in the brain and spinal cord. 1,2 ALS is primarily characterized by progressive degeneration and atrophy of voluntary skeletal muscles, ultimately resulting in paralysis and death from respiratory failure. 3 The incidence of ALS is approximately 1-2.6 cases per 100,000 persons annually. 4 Only approximately 5%-10% of ALS cases are familial (fALS), with an autosomal dominant pattern of inheritance and at least 40 genes and loci that exert a significant effect known to contribute to familial pathology. 5,6 In 2011, missense mutations in the UBQLN2 gene were identified in large ALS and ALS/ FTD families. 7 Ubiquilin 2 is a member of the ubiquilin family (ubiquilins), which regulates the degradation of ubiquilinated proteins. 8 More recently, the P497H mutation in UBQLN2 was found to be associated with dysfunction of the ubiquilin proteasome system, neuro inflammation, 9 and the formation of stress granules (SGs). 10 Functional analysis showed that P497H mutation could impair protein degradation and cause abnormal accumulation of proteasomal subunits, such as PSMC2. 11 In addition, the P497H mutation can also affect autophagy pathway degradation by regulating liposomal acidity. 12,13 Stress granules are cytoplasmic RNA granules that form when untranslated mRNPs (ribonucleoprotein particles) accumulate in cells subjected to biotic stress or environmental stress. 14 Abnormal formation of stress granules has been linked to the pathogenesis of neurodegenerative disorders. Mutations in a series of RNA-binding protein (RBP) genes (TDP-43, FUS, hnRNPA1, ATXN2, TIA-1) have been confirmed to cause ALS and/or FTD. [15][16][17] Most of the diseasecausing mutations map to the low complexity domain regions of SG-related RBPs and disrupt their biophysical properties, leading to increased LLPS (liquid-liquid phase separation) into SGs and aggregate formation. 18 The LLPS of UBQLN2 under certain environmental conditions is similar to that of RNP particles. A group recently suggested that UBQLN2 modulates the state of the components to be recruited to SGs. It has recently been reported that overexpression of UBQLN2 mutants (P497H and P506T) in HeLa cells directly decreased the interaction between UBQLN2 and FUS, which resulted in loss of the ability of UBQLN2 to regulate FUS-RNA complexes and SG formation. 19 However, the effect of UBQLN2 on the dynamic process and the disturbance of stress granules is still unknown.

| Western blot analysis
Cells were lysed in RIPA buffer containing 50mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, sodium orthovanadate, sodium fluoride, EDTA, leupeptin, and protease inhibitor. According to the manufacturer's protocol, protein concentrations were determined using a Pierce BCA protein assay kit (Thermo scientific). Proteins were denatured at 99°C for 10 min with Laemmli sample buffer containing β-mercaptoethanol. Western blot analysis was performed as described. Western blots were quantified with the Image plug in Gel Analyzer.

| Data analysis
GraphPad Prism software (San Diego, CA, USA) was used. The data are presented as the mean ± standard error (SE) from more than three or four independent tests. The D'Agostino & Pearson test or Shapiro-Wilk test was used to test for normality. If the data passed the normality test, we use the Student's t-test or the One-way ANOVA to determine the statistical significance of the data. If not, we use the rank sum test. Statistically significant p values were lower than 0.05 and are marked with an asterisk.

| UBQLN2 is Localized to SGs
There are many components of SGs, including G3BP1, TIA-1, TIA-R, ATXN2, and PABPC1. [20][21][22] To investigate whether UBQLN2 associates with SG components, we performed an immunostaining experiment. Under sodium arsenite (SA) treatment, endogenous UBQLN2 was shown to colocalize with the SG components G3BP1, TIA-1, and ATXN2 in HEK293 cells ( Figure 1A). To further confirm the relationship between exogenous UBQLN2 and SGs, HEK293T cells were transfected with UBQLN2-WT or UBQLN2-P497H plasmid. Immunofluorescence results showed that overexpressed UBQLN2-WT and UBQLN2-P497H also colocalized with the SG component ATXN2 ( Figure 1B). As some SG components depend on cell lines and stress conditions, we then determined whether UBQLN2 localizes to SGs in HeLa cells. UBQLN2 was shown to colocalize with the SG components G3BP1, TIA-1, ATXN2, and PABPC1 in HeLa cells under conditions of both SA treatment and heat shock ( Figure S1). Based on these observations, UBQLN2 appears to generally localize to SGs.

| UBQLN2 affects SG dynamics
Given that UBQLN2 associates with SGs, we questioned whether UBQLN2-WT and the UBQLN2 mutant impact the process of SG dynamics. To measure the SG assembly kinetics, HEK293T cells were transfected with UBQLN2-WT-GFP or UBQLN2-P497H-GFP and treated with SA for 15 min, 20 min, and 30 min, followed by fixation and immunostaining for GFP and G3BP1 ( Figure S4). Thus, these results suggest that UBQLN2-P497H affects the SG assembly by regulating the level of the critical SG protein TIA-1 upon treatment with SA.

| UBQLN2 does not affect eIF2α phosphorylation but regulates 4E-BP1 phosphorylation
Phosphorylation at serine 51 of eIF2α is involved in the early initiation step in SGs assembly. Phospho-eIF2α (phosphorylation of eIF2α) inhibits protein synthesis, and a large amount of mRNA and protein are recruited into the stress granules. 24 To explore whether UBQLN2 can regulate the initial stage of SG formation via the cells was significantly reduced than UBQLN2-WT and control group after SA treatment. (Figure 4C, D). In addition,we found that 4E-BP1 accumulated in cytoplasmic puncta positive for SG marker G3BP1, and nearly all SGs contained 4E-BP1 ( Figure S5). This phenomenon suggests that overexpression of UBQLN2-P497H may modulate SG formation by regulating 4E-BP1 phosphorylation.

| UBQLN2 impacts the nucleoplasmic distribution of TDP-43
TDP-43, FUS, and C9ORF72 are pathogenic genes in ALS and significantly affect the dynamic process of SG formation. 14 Moreover, UBQLN2 has an interactive relationship with these genes. Therefore, to investigate whether overexpression of UBQLN2 affects the ex-

| DISCUSS ION
Here, we investigated the influence of overexpressed wild-type UBQLN2 and the P497H mutant on SG dynamics and found that they both inhibited the formation of SGs. It should be noted that the overexpression of UBQLN2 in animal models or cells reproduces protein aggregation. In this study, we found that UBQLN2 aggregation increased with culture time, and the aggregation reproduced by the P497H mutant ( >2μm2 ) was much greater than that reproduced by the wild-type strain ( Figure S6A, B). Overexpression of UBQLN2 causes neuronal death in rodent models. 9 Therefore, the impact of UBQLN2 on the formation of SGs may be related to the aggregates induced by overexpression of wild-type UBQLN2 and the P497H mutant. We also observed that the disassembly of SGs in UBQLN2-P497H was accelerated compared to that in the control at 150 min ( Figure 2B). Other groups have shown that the P497H mutation impacts the protein solubility and ubiquitylation and that the mutated protein can bind more ubiquilin proteins. 26 However, the specific binding between ubiquilin and UBQNL2 can pull UBQLN2 out of SGs, 10 which may direct ubiquilination protein to the proteasome for degradation. Therefore, we speculate that the P497H mutant promotes the disassembly of SGs because of the ability of UBQLN2 to combine with more ubiquilin proteins.
TIA-1, a cytotoxic T lymphocyte granule-associated RNA-binding protein, is a critical component of stress granules, which can promote the assembly of stress granules. 23 Similar reports revealed that knockout of TIA-1 can make cells more sensitive to stress. 27 Moreover, mutations in TIA-1 also affect the formation of SGs. 28 We reported that the level of TIA-1 in both UBQLN2-overexpressing wild-type and P497H mutant cells was decreased to different degrees by SA treatment (Figure 3A, B), which results the suppression of SGs assembly ( Figure 6). Meanwhile, in mutant SOD1-transfected cells, the formation of TIA-1-positive stress granules was delayed.
Mutation in TIA-1 slows SG disassembly following heat shock in HeLa cells and decreases TDP-43 mobility in SGs. 29 Interestingly, the assembly of SGs was also affected in the overexpression of wildtype UBQLN2, and there was no significant difference in the level of TIA-1. This indicates that there is another mechanism which independent of TIA-1, involved in the regulation of SGs with UBQLN2. in wild-type UBQLN2, surprisingly, we found that P497H UBQLN2 downregulated phospho-4E-BP1 after SA treatment ( Figure 4C, D), and the number of SGs was also decreased (Figure 2A-C). It indicates that mTOR may be involved in the inhibition of SGs formation by UBQLN2-P497H.
Many studies have shown that the broad prospects of mouse models in the mechanism and therapy research of ALS. [36][37][38][39][40][41] It is reported that UBQLN2-P497H overexpression mice have pathological phenomena of neuron loss, behavioral defects, and protein aggregation. [42][43][44] Another UBQLN2-P497H mice expressing about 20% endogenous level using NF-H neuron promoter was reported, and no obvious pathological phenomenon was found. 45 The UBQLN2-P506T knock in mice had cognitive defects and hippocampal endosomes, but no behavioral abnormalities were found. 46 These suggest that UBQLN2 mutation may lead to the disease in a harmful and expression level-dependent manner. Considering the difficulty of stressing in animal models, there are few in vivo studies focused SGs dynamic in ALS mouse model. Our finding that the role of UBQLN2 mutant in regulating SGs was different from wild-type in vitro needs further verification in animal models.

| CON CLUS IONS
In summary, ubiquilin 2 colocalizes with the SG component proteins   H) The nucleoplasmic distribution of TDP-43 was detected through a nucleo-cytoplasmic separation experiment, with GAPDH as the cytoplasmic reference and H3 as the intra-cellular reference. (I) Relative quantitative analysis of the cytoplasmic/nuclear TDP-43 ratio (C/N ratio) in UBQLN2-WT and UBQLN2-P497H cells in the presence and absence of SA treatment. One-way ANOVA was used to test the significance of Tukey's test results. *p < 0.05; ns, not significant F I G U R E 6 Model of TIA-1 affects stress granule assembly. (1) Sodium selenite activates mTOR, leading to increased phosphorylation of 4E-BP1 and blocking translation initiation. (2) Stresses activate eIF2α kinases that phosphorylate eIF2α, deplete the ternary complex, and promote the assembly of a noncanonical Pre-initation complex (PIC); these noncanonical PICs differ from canonical PICs in composition and exposure of the 40S subunit interface and mRNA. These interfaces recruit RNA-BPs such as TIA-1, increasing the local concentration of these proteins to promote LLPS and assembly of SG seeds. However, UBQLN2-P497H could repress the level of TIA-1, thereby inhibiting stress granule assembly These provide new insights into the role of UBQLN2 in oxidative stress and the pathogenesis of ALS.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
GP, AG, HN, LC, and YC performed the experiments. MZ, JL, and YZ analyzed the data. XL and GP wrote the manuscript with support from DL and ML. XL and ML designed and supervised the study. All authors contributed to the article and approved the submitted version.

I N S TITUTI O N A L R E V I E W B OA R D S TATE M E NT
Not applicable.

I N FO R M ED CO N S ENT S TATEM ENT
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.