A novel p.T139M mutation in HSPB1 highlighting the phenotypic spectrum in a family

Abstract Introduction Mutations in the HSPB1 gene encoding the small heat shock protein B1 are associated with an autosomal dominant, axonal form of Charcot–Marie–Tooth disease 2F (CMT2F) and distal hereditary motor neuropathy. Recently, distal myopathy had been described in a patient carrying HSPB1 mutation adding to the complexity of phenotypes resulting from HSPB1 mutations. Methods Five patients in a family with concerns of hereditary neuropathy were included. Detailed clinical examinations, including assessments of motor and sensory function, and electrophysiological data were obtained. Genetic analysis was requested through a commercial laboratory. In vitro studies were carried out to assess the pathogenicity of the novel mutation found in this family studies. Results All patients carried a novel mutation, c.146 C>T (p.T139M), substitution in the α‐crystallin domain of HSPB1 causing a clinical phenotype with hyperreflexia and intrafamilial variability, from muscle cramps as the only presenting symptom to a classic CMT phenotype. In vitro studies showed that cells expressing HSPB1‐T139M displayed decreased cell viability with increased expression of apoptosis markers. Moreover, overexpression of the mutant, not the wild‐type HSPB1, caused formation of congophilic aggregates. Conclusions In vitro findings strongly support the pathogenicity of this novel mutation. We propose that Congo red histochemical stain may serve as a simple screening tool for investigating if the aggregates in mutant cells have misfolded β‐pleated sheet secondary structures.


| INTRODUCTION
Charcot-Marie-Tooth disease (CMT) is the most common genetic neuromuscular disease, comprising a heterogeneous group of neuropathic disorders for which more than 80 causative genes have now been described (Rossor, Evans, & Reilly, 2015). Mutations in the small heat shock protein, HSPB1, have been found to cause degeneration of peripheral nerves, presenting with two major phenotypes. One is CMT type 2F (CMT2F; OMIM 606595), which is an axonal form of hereditary neuropathy characterized by progressive degeneration of motor and sensory nerves in a length-dependent manner, causing weakness in the lower limbs affecting distal muscles, and mild to moderate sensory impairment in feet and hands. The second phenotype is distal hereditary motor neuronopathy 2B (OMIM 608634), which resembles CMT2F with progressive, symmetrical, and predominantly distal weakness, but differs having minimal or lack of sensory symptoms (Rossor, Kalmar, Greensmith, & Reilly, 2012). Recently, distal myopathy had been described in a patient carrying HSPB1 mutation (Lewis-Smith et al., 2016) adding to the complexity of phenotypes resulting from HSPB1 mutations. HSPB1 (also named HSP27) belongs to a group of ubiquitous molecular chaperone proteins, which are composed of a highly conserved α-crystallin domain, a β-sandwich structure, that is the hallmark of all proteins belonging to this family (Basha, O'Neill, & Vierling, 2012;Kappe, Boelens, & de Jong, 2010;Mymrikov, Seit-Nebi, & Gusev, 2011). Small heat shock proteins are increased following exposure of cells to the environmental or physiological stressors. They act as molecular chaperones that counteract the formation of aberrantly folded polypeptides and allowing correct refolding during stress recovery (Mymrikov et al., 2011). HSPB1 has been thought to play an important role in maintaining neuronal cell homeostasis through stabilizing cytoskeleton, facilitating the clearance of abnormal protein accumulation through its ability to modulate the ubiquitin-proteasome pathway (Elliott et al., 2013;Parcellier et al., 2003;Treweek, Meehan, Ecroyd, & Carver, 2015). Decreased expression of HSPB1 was shown to cause impairments in growth and cytoskeletal organization (Mairesse, Horman, Mosselmans, & Galand, 1996). HSPB1 is also known to have the ability to bind and stabilize microtubules (Hino, Kurogi, Okubo, Murata-Hori, & Hosoya, 2000;Launay, Goudeau, Kato, Vicart, & Lilienbaum, 2006); in fact, treatment with a selective HDAC6 inhibitor successfully reversed the clinical phenotype of both S135F and P182L transgenic mice (d' Ydewalle et al., 2011). In addition, HSPB1 protected cells against several proapoptotic agents when expressed at high level, while the inhibition of its expression was found to predispose cells to apoptosis (Kamada et al., 2007;Rocchi et al., 2006), presumably by triggering the release of cytochrome c (Paul et al., 2002). Similar to αBcrystallin, HSPB1 is constitutively expressed in tissues with high rates of oxidative metabolism, including, the heart, type I and type IIa skeletal muscle fibers, brain, and spinal cord neurons (Arrigo et al., 2007;Maatkamp et al., 2004;Srivastava et al., 2012).
In the present report, we described a novel mutation, c.146 C>T (p.T139M), substitution in the α-crystallin domain of HSPB1 causing a clinical phenotype with hyperreflexia and intrafamilial variability and muscle cramps as the only presenting symptom. Furthermore, we carried out in vitro studies and showed that cells transfected with the mutant construct displayed decreased cell viability with increased expression of apoptosis markers and congophilic aggregate formations. These studies strongly support the pathogenicity of this novel mutation.

| Patients
Five patients in a family were included in this study. The proband was initially referred to the Neuromuscular Clinic at Nationwide Children's Hospital, Columbus, Ohio for an evaluation of muscle cramps.

| Expression of wild-type and T139M HSPB1 protein in SHSY-5Y and HeLa cells
Plasmid vector carrying full length of wild-type HSPB1 cDNA, driven by an elongation factor-1 alpha promoter, was designed in our laboratory. T139M mutation was induced by using site-directed mutagenesis kit (Agilent Technologies, USA) to obtain mutant HSPB1 vector with same backbone. All vectors were sequenced and confirmed to have right nucleotide order.
Human HeLa and SHSY-5Y cells were grown in culture medium containing Dulbecco's modified Eagle's medium with 10% fetal calf serum and 1% penicillin and streptomycin, as a monolayer at 37°C in a humidified incubator with 5% CO 2 . SHSY-5Y cells were transfected using electroporation, a well-established method in our laboratory (Feng et al., 2000). Briefly, 2 × 10 6 cells were electroporated with 10 μg of vector DNA at 200 V and 100 μF using Gene Pulser II (Bio-Rad Laboratories, Hercules, CA, USA) were plated in six-well plate for further analyses. HeLa cells were transfected using TransIT ® -LT1 (Mirus Bio LLC, Madison, WI, USA) according to the manufacturer's protocol.
Transfection efficiency was determined by transfecting the cells with the same amount of control vector carrying green fluorescent protein cDNA sequence and was around 70% and 50% in SHSY-5Y cells and in HeLa cells, respectively.

| Cell viability analysis
Cell viability after transfection was measured using CellTiter 96 MTS assay (Promega, Fitchburg, WI, USA) at 24, 48, and 72 hr after transfection. Briefly, the cells were incubated with MTS solution for half an hour, subsequently the absorbance at 490 nm was read from each well by using Biotek ELISA plate Reader (BioTek Instruments, Winooski, VT, USA). The assays were performed three times in triplicates for each group.
Three independent assays were done in triplicates by using ABI 7500 real-time PCR machine and the results computed and analyzed using DataAssist ™ Software (RRID:SCR_014969; DataAssist).

| Statistics
Data are presented by GraphPad Prism 7 (RRID:SCR_002798; Graphpad Prism) as mean values and standard error of mean bars. The number of significance is calculated by unpaired two-tailed Student's t-test with p values that are categorized for different significance levels: *p < .05, **p < .01, and ***p < .001.

| Transfection and in situ Congo red staining for morphological evaluation
HeLa cells were selected for these studies due to their large cytoplasmic volume. Cells transfected with mutant or wild-type HSPB1 constructs were grown on coverslips for 24 hr and then were stained with Congo red to determine if the aggregates display amyloidogenic properties using our previously published protocol (Feng et al., 2000).
Briefly, cells were first fixed with 10% formalin for 10 min and stained with 1% Congo red (Sigma-Aldrich, St. Louis, MO, USA) for 5 min, followed by destaining with 0.01% potassium hydroxide in 50% ethanol.
Coverslips were then passed through graded ethanol concentrations for dehydration and mounted in a mounting medium and examined by fluorescent microscopy under rhodamine filter.

| Clinical presentation
Patients' age ranged from 35 to 55 years. Progressive ankle instability and difficulty walking were the presenting symptoms in all except the proband who complained of muscle cramps. Disease onset was in the second and fourth decade in two sisters who were more severely affected with gait difficulties and foot drop. Another sister in her fifth decade had a 5-year history of ankle instability, mild gait difficulty, and weakness in her ankle stabilizer muscles. One male sibling had disease onset late in the third decade and an 8-year history of ankle instability and sensory symptoms The pedigree of the family is presented in Figure 1 and the relevant clinical examination findings are detailed in

| Electrophysiological findings
Motor and sensory nerve conduction studies were carried out in four patients (II-2, II-3, II

| Functional analysis of the mutant HSPB1
SH-SY5Y is a human-derived neuroblast-like cell line, generally used as in vitro model for studying neuronal function and differentiation.
F I G U R E 1 (a) Pedigree of the family. Filled square and circles indicate affected males and females, respectively. Proband II-3 is indicated with an arrow. (b) The sequences of the HSPB1 protein from a range of divergent species, compared with multiple sequence alignment tool. The mutated region of the threonine amino acid at position 139 (highlighted with dash box) is highly conserved To study the pathogenicity of HSPB1-T139M mutation, we first evaluated the impact of this novel mutation on the overall biological fitness of this stable neuroblastoma cells using MTS cell viability assay (Cory, Owen, Barltrop, & Cory, 1991). Evgrafov et al. (2004) previously showed that N2a cells overexpressing S135F-HSBP1 were less viable than the N2a overexpressing the wild-type HSBP1. Similarly, we found that the expression of mutant HSBP1 lowered the cell viability significantly (Figure 2a) (Acunzo, Katsogiannou, & Rocchi, 2012;Concannon, Gorman, & Samali, 2003). The major pathways mediated apoptosis regulations by HSPB1 are through inactivation of the caspase cascade and caspase-independent apoptosis pathways via Daxx (death domain-associated protein; Acunzo et al., 2012). In this study, we in- We also assessed if this novel mutation impairs protein degradation function, resulting in congophilic aggregate formations of misfolded proteins, similar to mutations in another heat shock protein, HSPB5 (αB-crystallin) in muscle (Andley, Hamilton, Ravi, & Weihl, 2011;Selcen, 2011;Selcen & Engel, 2003). For these experiments, HeLa cells transfected with mutant or wild-type HSPB1 constructs were grown on coverslips for 24 hr and then were stained by Congo red to determine if the aggregates display amyloidogenic properties.
Cells examined under rhodamine optics revealed varying size of intracytoplasmic aggregates with congophilic properties in the HSPB1-T139M-expressing cells (Figure 3). Wild-type HSPB1 overexpressing cells did not display congophilia.

| DISCUSSION
We report five siblings of a family with distal, predominantly motor neuropathy due to the c.416C>T transition leading to a T139M missense mutation in HSPB1. Although the severity of the weakness in ankle stabilizers were highly variable, hyperreflexia was invariably present in all patients and the sensory involvement were confined to a mild decrease of pinprick appreciation and vibration. The proband in this family had a very mild phenotype, presented with occasional leg cramps and mild decrease in pinprick appreciation without detectable motor deficit, which was corroborated by the presence of normal or borderline normal motor and sensory nerve conduction studies.
However, easily elicitable brisk reflexes and high-arched feet with absent toe spreading prompted us to obtain a detailed family history Pyramidal signs (brisk reflexes and Hoffman sign) were present in several members of a multigenerational family with A136L mutation (Stancanelli et al., 2015). E41K mutation in the N-terminal domain of HSPB1 was also reported to present with muscle atrophy and weakness of distal extremities as well as hyperreflexia and upward plantar responses in a 39-year-old woman (Capponi et al., 2011). These reports are in support of our view that the disease process should be recognized as a central peripheral distal axonopathy according to the site of anatomical involvement, a classification proposed several decades ago (Schaumburg, Wisniewski, & Spencer, 1974;Spencer & Schaumburg, 1977a,b). This is reinforced by the fact that HSPB1 is ubiquitously expressed in both central and peripheral motor neurons.
Moreover, we provided evidence from a transgenic animal model with R136W mutation in HSPB1, which showed neuropathological changes consistent with distal axonal pathology not only in peripheral nerves but also in the central nervous system (Srivastava et al., 2012). Axonal pathology, predominantly found in the anterior and lateral funiculus of the lumbar and sacral cords from these mice, was consistent with a length-dependent involvement of the corticospinal tracts (Srivastava et al., 2012). In addition, other heat shock proteins, such as chaperonin 60 (heat shock protein 60) have been reported to cause prominent upper motor neuron degeneration presenting with hereditary spastic paraplegia (Fink, 2013). Mutations in other axonal CMT-associated genes such as mitofusin-2 (MFN2) and alanyl-tRNA synthetase expressed in the both central and peripheral neurons, although reported rarely were associated with pyramidal signs (Motley et al., 2015).
We found that p.T139M mutation in HSPB1 is highly conserved across all different species and the variant was predicted as disease causing with high probability. We further investigated the pathogenicity of this mutation with in vitro studies, which showed that the expression levels of apoptotic genes were dramatically increased in SH-SY5Y cells transfected with the mutant suggesting that T139M substitution induced the cell death via apoptosis. In addition, we have shown that overexpression of the HSPB1-T139M mutant in HeLa cells causes formation of congophilic aggregates. Congo red histochemical dye has the ability to bind specifically to crossed β-pleated sheet structures. Wild-type HSPB1 should maintain protein homeostasis by binding proteins in non-native conformations, thereby preventing substrate aggregation. The T139M mutant, however, failed in this function and resulted in an accumulation of misfolded proteins, which were targeted by Congo red for intercalating between the β-pleated sheet structures.
In conclusion, our study showed that the novel mutation c.416C>T; p.T139M in HSPB gene is pathogenic and caused phenotypic variability among the siblings in a family. The clinical examination findings are compatible with a disease process presenting as centralperipheral distal axonal disease. Our study also shows that Congo red histochemical stain may serve as a simple tool to investigate if the aggregates in mutant cells have misfolded β-pleated sheet secondary structures.