Autosomal dominant axonal Charcot–Marie–Tooth disease type 2B (CMT2B) largely overlaps with hereditary sensory-autonomic neuropathies (HSANs) as it is characterized by predominant sensory loss, high frequency of ulcer formations and amputations, with variable motor deficits [1
]. Affected patients show severe distal sensory loss particularly for pain and touch, reduced-to-absent deep tendon reflexes, foot deformities, and sometimes distal wasting and weakness mainly in lower limbs [1
]. The disease typically starts during the second or third decade [10
] and runs a slowly progressive course [11
]. Males have a higher occurrence of ulcers [4
], suggesting a difference in disease severity according to gender as for HSAN I [13
CMT2B is caused by heterozygous mutations in RAB7A
, with five mutations (L129F, K157N, N161T/I, V162M) reported in patients from eleven families [1
], all displaying the characteristic ulcero-mutilating phenotype with variable motor involvement (Table 1
Pathomechanisms, whereby RAB7A
mutations lead to CMT2B, are a matter for debate and investigation. RAB7A, hereafter referred to as RAB7, is a member of the Rab family of small GTPases involved in the regulation of vesicular trafficking between early endosomes and lysosomes, controlling transport to the degradative compartments in the endocytic pathway and lysosome biogenesis [14
]. RAB7 modulates the Endoplasmic Reticulum (ER) morphology by controlling the ER homeostasis and ER stress [15
]. Crosstalk occurring at mitochondria-lysosome contact sites regulated by Rab7 has also been recently demonstrated [16
Although ubiquitously expressed, RAB7 has specific functions in neurons, where it regulates retrograde axonal trafficking and signaling of neurotrophin receptors, as well as neurite outgrowth [17
]. Furthermore, RAB7 regulates cell migration by influencing integrin trafficking and vimentin assembly [19
] and cortical neurons’ migration during development [20
Interestingly, RAB7 has specific effectors in neurons as co-immunoprecipitates with the neurotrophin receptor TrkA (Tropomyosin-receptor-kinase A) and interacts directly with the intermediate filament protein peripherin [18
]. Therefore, it is not surprising that mutations in RAB7
cause a disease restricted to neurons, although it is unclear why sensory neurons are so selectively vulnerable.
Previous biochemical characterization of four CMT2B-causative RAB7 mutants showed increased dissociation rate constant (Koff,
) for nucleotides and lower GTPase activity per binding event [22
]. Overexpression of these mutant proteins inhibits neurite outgrowth in several cell lines [25
]. Furthermore, these RAB7 mutant proteins display stronger interaction with some RAB7 effector proteins, including RILP (RAB-interacting lysosomal protein), required for lysosomal transport towards the microtubule organizing center (MTOC) by inducing dynein-dynactin motors recruitment [27
RAB7 and RILP control degradation of the epidermal-growth-factor receptor (EGFR), a member of the receptor tyrosine-kinase family involved in regulating cell proliferation, survival, differentiation and migration [28
]. Importantly, EGF seems also to have important neurotrophic functions [30
Previous experiments on EGFR degradation obtained on cells transiently or stably transfected with CMT2B-causing RAB7 mutants gave conflicting results: transient expression of these mutants demonstrated normal or increased EGFR degradation [22
] while stable transfection revealed inhibition [32
Here, we report a family with a novel CMT2B phenotype with motor predominance and absence of ulcers and mutilations, carrying a novel pathogenic RAB7 variant (c.377A>G, p.K126R) which is absent in global databases, affects a highly conserved amino-acid in the GTPase domain of Rab7, is predicted to be pathogenic by in silico analysis, and is transmitted as an autosomal dominant trait. We performed extensive biochemical and functional studies, which confirmed its pathogenic role.
2. Materials and Methods
We evaluated clinically and electrophysiologically (standard procedures) one healthy and two affected family members (Figure 1
A). Informed consent was obtained for all procedures from study participants.
The index case underwent a biopsy of sural nerve biopsy which was processed for histological and ultrastructural examination [33
]. 3-mm skin biopsies were taken (from shoulder and the lateral aspect of the proximal phalanx of the index finger) for fibroblast culture, immunohistochemistry, and intraepidermal nerve fiber (IENF) count [34
2.2. Gene Sequence Analysis
The proband’s DNA was analyzed by Next Generation Sequencing (NGS) technology with a probe-based customized panel for CMT and related disorders (Illumina Nextera Rapid Capture Custom kit, Illumina Inc., San Diego, CA, USA), (Tables S1–S3
). Sequencing was performed using the NGS MiSeq sequencer (Illumina Inc.). The entire RAB7
gene-targeted region (6 coding exons and 25 bp of flanking introns) was sequenced by NGS with a depth of coverage >20×. The sequence variant was confirmed in proband and her father by the Sanger method (Figure 1
2.3. Mutagenesis and Plasmid Construction
Most constructs used in this study have been described previously [14
mutant was constructed using the QuickChange XL Site-Directed Mutagenesis Kit (Stratagene, San Diego, CA, USA). The oligonucleotides used to generate the mutant were 5′-GTGTTGGGAAACAGGATTGACCTCG-3′ and 5′-CGAGGTCAATCCTGTTTCCCAACAC-3′. The mutant RAB7K126R
plasmids were obtained using RAB7wt
cDNA previously cloned in pcDNA3-2xHA or pET-16b vector in frame with DNA coding for hemagglutinin (HA) or poly-His tag, respectively.
Western blotting analysis (WB): anti-HA (1:500, sc-805), anti-peripherin (1:500, sc-28539) and anti-vinculin (1:10000, sc-25336) were from Santa Cruz Biotechnology, Santa Cruz, CA, USA; anti-actin (1:5000, ab-8224) and anti-tubulin (1:6000, clone B512), were from Sigma-Aldrich, St-Louis, MO, USA, while anti-EGFR (1:2000, 20-ES04) was from Fitzgerald, Concord, MA, USA and anti-RILP (1:400, 13574–1-AP) from Proteintech, Rosemont, IL, USA. Immunofluorescence analysis: anti-early endosome antigen 1 (EEA1, 1:1000, ab70521, Abcam), anti-HA (1:500, ab9110, Abcam), anti-EGFR (1:100, 20-ES04, Fitzgerald). Secondary antibodies conjugated to fluorochromes for immunofluorescence or horseradish peroxidase (HRP) were from Invitrogen (Carlsbad, CA, USA), Santa Cruz Biotechnology or Fitzgerald. Immunohistochemistry: myelin basic protein (MBP, 1:100, ab7349, Abcam), anti-peripherin (1:1000, ab4666, Abcam), anti-vasoactive intestinal peptide (VIP, 1:400, ab22736, Abcam), anti-neurofilament 200 (NF-H, 1:400, N0142, Sigma-Aldrich), anti-PGP9.5 (1:500, MCA4750GA, Bio-Rad, Hercules, CA, USA), anti-EGFR (1:100, 20-ES04, Fitzgerald). Secondary antibodies Alexa Fluor-conjugated (Jackson ImmunoResearch, Cambridge, UK) were employed.
2.5. Cells and Transfection
Neuro2A and NCI H1299 cells, fibroblasts from CMT2B patients and control were grown in DMEM supplemented with 10% or 15% FBS, 2 mM L-glutamine, 100 U/mL penicillin and 10 mg/mL streptomycin.
Transfection was performed using Metafectene PRO reagent (Biontex, München, DE) according to manufacturer’s instructions. Cells were analyzed 24 h after transfection.
2.6. Western Blotting and Co-Immunoprecipitation
Cells were processed for SDS-PAGE and WB as previously described [22
]. Frozen sural nerves were pulverized and sonicated in lysis buffer (95 mM NaCl, 25 mMTris-HCl, pH 7.4, 10 mM EDTA, 2% SDS, and protease inhibitors) and lysates were subjected to WB [20
Densitometry analysis was performed using the NIH ImageJ program or Gene Tools from Syngene and normalized to appropriate loading controls signal intensity. For immunoprecipitation, we used the anti-HA affinity gel (Ezview Red Anti-HA E6779, Sigma-Aldrich) according to the manufacturer’s instruction.
2.7. Confocal Immunofluorescence Microscopy
Cells grown on 12-mm round glass coverslips were fixed, permeabilized with 0.1% Triton X-100 for 5 min at room temperature and incubated with primary and secondary antibodies as previously described [21
]. Cells were viewed with a Zeiss LSM700 confocal microscope. Zen 2011 software (Carl Zeiss, Oberkochen, Germany) was used for image capture and to calculate the weighted colocalization coefficient of EGFR and EEA1.
Skin biopsies were sectioned (50 μm) using a standard cryostat (Bio-Optica, Milan, Italy). Immunohistochemistry assays were performed following a standard free-floating protocol. Fluorescent images were acquired with Leica TSC SP8 confocal microscope (Leica Microsystems, Wetzlar, Germany).
2.8. Nucleotide Dissociation and GTPase Assay
Nucleotide exchange assay and determination of Koff
were performed as previously reported [36
]. The GTPase assay was performed as previously described [37
]. GTP hydrolysis was quantified as a GDP signal relative to GDP+GTP signal [37
2.9. EGFR Degradation Assay
To measure EGFR degradation cells were treated and processed as previously described [22
2.10. Neurite Outgrowth Assay
Differentiation of Neuro2A was triggered by serum withdrawal as previously described [25
]. Cells were co-transfected with the pEGFPC1 vector (to express GFP and visualize the entire cell) and with a plasmid for expression of HA-tagged RAB7wt
or mutant proteins, fixed, permeabilized and stained [19
]. For each experiment, to determine the percentage of cells bearing neurites (>50 μm), approximately 100 transfected cells were counted in at least 10 randomly chosen visual fields. We compared the number of cells bearing neurites longer than 50 μm between control cells and cells expressing HA-tagged RAB7wt
and mutant proteins.
2.11. Real-Time PCR
Total RNA was isolated using the RNeasy Micro kit (Qiagen, Hilden, Germany) and retrotranscription was made using SuperScript II Reverse Transcriptase (Invitrogen) according to the manufacturer’s instructions.
Quantitative real-time PCR was carried out with Power SYBR Green (Applied Biosystems, Foster City, USA) using Applied Biosystems 7900HT Fast Real-time PCR System. The primers used were GAPDH forward 5′-GGTGGTCTCCTCTGACTTCAACA-3′ and reverse 5′-GTTGCTGTAGCCAAATTCGTTGT-3′; EGFR forward 5′-GGCAGGAGTCATGGGAGAA-3′ and reverse: 5′-GCGATGGACGGGATCTTAG-3′ from Eurofins Genomics (Ebersberg, Germany). The thermal profile used: 1 cycle of 2 min at 50 °C; 1 cycle of 10 min at 95 °C; 40 cycles of 15 s at 95 °C, 1 min at 55 °C; 1 cycle of 15 s at 95 °C and 15 s at 60 °C. The relative expression level was calculated using the comparative CT method and expressed as “fold change.” The relative quantification was considered significant when there was a minimum of two-fold change.
2.12. Molecular Modeling Studies
Wild type RAB7 structure was retrieved from the Protein Data Bank (pdb code 1T91) and used to generate the model of RAB7K126R protein (Pymol, The PyMOL Molecular Graphics System, Version 1.2r3pre, Schrödinger, LLC). Molecular dynamics simulations were performed within MOE (MOE, Chemical Computing Group Inc., Montreal H3A 2R7 Canada).
2.13. Statistical Analysis
Data were statistically analyzed using Student’s t-test or χ2 test (* p < 0.05, ** p < 0.01 and *** p < 0.001). Experiments were performed at least in triplicate.
CMT2B has long been considered mainly a sensory neuropathy with a large overlap with the HSANs. Only five RAB7 mutations have been reported, all involving the GTP binding domain and associated with a mainly sensory phenotype.
We identified and characterized a novel missense variant (p.K126R) in RAB7
associated with predominantly motor CMT2 in the two affected family members. Notably, both subjects had remarkable muscle wasting and weakness with only minor sensory signs, in keeping with CMT2 with motor predominance, never linked previously to RAB7
mutations. All previous families showed ulcers and amputations, with variable motor involvement (Table 1
). Lower occurrence of ulcers is reported in CMT2B females [4
], but the motor predominant phenotype was shown also by the father, ruling out a gender effect.
The p.K126R variant affects, like the other CMT2B mutations, the RAB7 GTP-binding domain and its pathogenicity is supported by the involved site, segregation analysis, and in silico
predictions. The lysine 126 is very conserved not only in RAB7 during evolution (Figure 1
G) but also in many other GTPases as this amino-acid is part of the switch II region; it is one of the amino-acids forming the guanine binding pocket and it is a crucial amino-acid residue involved in the hydrogen bond with the nucleotide ribose oxygen [38
]. Importantly, mutations in this amino-acid residue are predicted to have an increased rate of dissociation of the nucleotides favoring the active GTP-bound state [38
]. Despite arginine shares with lysine basic and charged features, it impairs the GTP pocket architecture. The bulkier side chain of arginine exceeds lysine volume and is thus unable to perform the same specific interactions with GTP and some amino acid residues (water bridging with glycine 28) in the surrounding pocket (Figure 6
We carried out a series of functional experiments to demonstrate the RAB7K126R pathogenic role and look for differences with other RAB7 mutants, to explain the distinctive predominantly motor phenotype. Our studies confirmed that the mutation affects a series of RAB7 properties, very similar to previously characterized mutants, but with at least one important difference.
First, biochemical data indicate that the mutant protein has an increased Koff
for both nucleotides and particularly high for GDP (Figure 2
A), inhibiting also GTPase activity per binding event (Figure 2
C), similarly to previous uncovered CMT2B-causing RAB7 mutant proteins [22
]. Interestingly, the corresponding mutation in the HRAS
gene, a gene encoding a member of the RAS superfamily of small GTPases as the RAB7
gene, is associated with the Costello syndrome, a disease causing an increased risk of developing tumors in different organs further supporting pathogenicity of this variant [39
Second, expression of RAB7K126R
strongly inhibits neurite outgrowth, compared to the other CMT2B-causing RAB7 mutants (Figure 2
]. As the onset of CMT2B occurs in the second or third decade, similarly to many other CMT types, development appears to be unaffected. However, the mutant proteins can rather impact on axonal regeneration, considering that this process recapitulates all the stages of neuronal differentiation [40
]. Indeed, the detrimental action of RAB7 mutants could be counteracted by other factors during development, but with age, these factors could become less effective and a decrease in regeneration capabilities may contribute to the progressive axonal loss underlying clinical manifestations.
Third, the RAB7K126R
mutant protein interacts more strongly with the intermediate filament peripherin (Figure 3
B–D), as for already reported RAB7
mutations. Peripherin is important for neuronal morphology, maturation, and differentiation, but also axonal regeneration [41
]. RAB7 affects peripherin assembly [21
] and altered RAB7-peripherin interaction could impair axonal regeneration thus contributing to nerve degeneration. Notably, we found increased labeling of peripherin in skin biopsy where numerous peripherin-positive filamentous structures were seen in contrast to the few seen in controls (Figure 3
With respect to both findings, it is noteworthy that the nerve biopsy showed—at least for this single sensory nerve—no evidence of regenerating clusters despite the moderate loss of myelinated fibers (Figure 1
mutant protein strongly reduces RILP downstream effector levels (Figure 3
E), similarly to RAB7N161T
]. Notably, in previous functional studies, siRNA-mediated RILP depletion was sufficient to cause strong effects on EGFR degradation and the biogenesis of late endosomes [27
Fifth, our data demonstrate also that expression of the RAB7K126R
mutant protein causes inhibition of EGFR degradation in contrast to previously studied CMT2B-RAB7 mutants [22
]. Indeed, expression of the RAB7L129F
, and RAB7V162M
mutants did not reveal any inhibition of EGFR degradation but rather a small and reproducible, though not statistically significant, increase [22
]. Furthermore, analysis of EGFR degradation in RAB7V162M
fibroblasts revealed an increase of EGFR degradation compared to control fibroblasts (Figure 4
C), while, expression of RAB7K126R
caused EGFR degradation inhibition (Figure 4
A), which was further confirmed in patient fibroblasts (Figure 4
C). In addition, the total EGFR amount was strongly increased in the RAB7K126R
patient compared to control fibroblasts (Figure 4
B). These findings were confirmed by WB on the sural nerve (Figure 5
A) and immunohistochemistry on skin biopsy which showed increased EGFR staining in the proband compared to the control although primarily in the surround of nerve fibers (Figure 5
B). Three hours after internalization most of EGFR was still in early endosomes in patient cells, demonstrating that impaired EGFR trafficking to late endosomes and lysosomes causes inhibition of degradation and accumulation of EGFR (Figure 4
It is tempting to speculate that the prominent motor symptoms observed in our patients might be related to the differences in EGFR amount and degradation rate in contrast with previously studied CMT2B mutants associated with the typical sensory-predominant phenotype. EGF signaling mediated by EGFR is important for development and maintenance of various tissues including the nervous system, and EGFR is expressed in differentiated post-mitotic neurons suggesting pleiotropic functions, although the possibility for this receptor to exert distinct actions on different kind of neurons (e.g., motor versus sensory) is still unclear. Besides its key role in regulating neural stem cell proliferation, self-renewal, differentiation, and migration, EGF has neurotrophic and neuromodulatory functions on different kinds of neurons, increasing neuronal survival [31
]. EGFR signaling is also required for proper cutaneous innervation during development and it seems to limit axonal outgrowth and branching [44
]. Interestingly, after central nervous system (CNS) injuries and diseases, activation of the EGFR pathway triggers quiescent astrocytes to become reactive and destructive to neurons [45
]. After spinal cord injury, EGFR activates astrocytes that represent a physical barrier to axon regeneration, expressing and secreting molecules that inhibit nerve growth [47
]. Consistently, rats subjected to weight-drop spinal cord injury treated with a potent EGFR inhibitor to the injured area show motor and sensory function improvement [48
]. Lack of EGFR expression is associated with progressive neurodegenerative disorders in mice [49
]. Overexpression of EGFR is associated with cancer (high-grade glioma) [51
]. Moreover, EGFR has a crucial role in the neurometabolic axis seen in the aging process [52
]. Glia and endothelial cells demonstrate induced expression of EGFR after an acute injury or chronic neurodegeneration and there are multiple and mechanistically diverse routes by which the EGFR may be involved in the neuronal aspect of aging-related disorders and neurodegeneration. Consistently, it has been demonstrated that inhibition of EGFR enhances peripheral nerve regeneration after injury [30
Thus, the inhibited degradation and the consequent accumulation of EGFR observed in RAB7K126R cells could cause an increase in EGFR signaling leading to inhibition of axonal regeneration. The above-mentioned absence of nerve regeneration at nerve biopsy in our patient is of interest in this context.
EGFR has also been implicated in the expansion of the peripheral nervous system (PNS) progenitor cells, including Schwann cell precursors [55
]. After peripheral nerve injury, EGFR expression is upregulated and promotes proliferation and migration of Schwann cells [56
]. Importantly, EGFR hyperactivation induces a peripheral neurodegenerative disease as transgenic mice overexpressing the EGFR ligand epigen show a phenotype similar to CMT1A and CMT4F animal models [57
Although there are at present no data regarding a specific role of EGFR in motoneurons, literature data on EGF signaling and EGFR in both CNS and PNS, suggest that the inhibition of EGFR degradation could be a possible explanation for the different clinical presentation in our family compared to the previously studied CMT2B families.
Recently, Wong et al. showed a role for RAB7 in driving the interactions between lysosomes and mitochondria, opening a new field of investigations for unraveling RAB7 functions [16
]. We did not find any evidence of mitochondrial abnormalities in the proband’s sural nerve, but it will be interesting to study the role of mitochondria in CMT2B in the future and to verify whether these organelles play a key role in the phenotype and its variants.