Rare Variants in MME, Encoding Metalloprotease Neprilysin, Are Linked to Late-Onset Autosomal-Dominant Axonal Polyneuropathies

Abstract

Axonal polyneuropathies are a frequent cause of progressive disability in the elderly. Common etiologies comprise diabetes mellitus, paraproteinaemia, and inflammatory disorders, but often the underlying causes remain elusive. Late-onset axonal Charcot-Marie-Tooth neuropathy (CMT2) is an autosomal-dominantly inherited condition that manifests in the second half of life and is genetically largely unexplained. We assumed age-dependent penetrance of mutations in a so far unknown gene causing late-onset CMT2. We screened 51 index case subjects with late-onset CMT2 for mutations by whole-exome (WES) and Sanger sequencing and subsequently queried WES repositories for further case subjects carrying mutations in the identified candidate gene. We studied nerve pathology and tissue levels and function of the abnormal protein in order to explore consequences of the mutations. Altogether, we observed heterozygous rare loss-of-function and missense mutations in MME encoding the metalloprotease neprilysin in 19 index case subjects diagnosed with axonal polyneuropathies or neurodegenerative conditions involving the peripheral nervous system. MME mutations segregated in an autosomal-dominant fashion with age-related incomplete penetrance and some affected individuals were isolated case subjects. We also found that MME mutations resulted in strongly decreased tissue availability of neprilysin and impaired enzymatic activity. Although neprilysin is known to degrade β-amyloid, we observed no increased amyloid deposition or increased incidence of dementia in individuals with MME mutations. Detection of MME mutations is expected to increase the diagnostic yield in late-onset polyneuropathies, and it will be tempting to explore whether substances that can elevate neprilysin activity could be a rational option for treatment.

Figure 1

MME Mutations Associated with Autosomal-Dominant Late-Onset Peripheral Neuropathies (A) Families with MME mutations identified by WES. Women are represented by circles, men by squares. Individuals with anonymized gender data are represented by diamonds. The symbols of affected individuals are filled (black, clinically affected; gray, subclinically affected), and those of unaffected individuals are empty. Members with unknown clinical status contain a question mark. There was no history of disease in obligate mutation carrier AT3/III-6 until her death at age 80 but neurological examination has not been documented. Individuals’ MME genotypes are shown below the pedigree symbols. (B) Schematic representation of neprilysin and distribution of mutations identified in this study and in the recent study of Higuchi et al. Functionally relevant protein domains are shown. Exon numbering is given below the diagram. Variants shown in red color have been associated with late-onset peripheral neuropathies in this study. Variants shown in green color have been reported in individuals with autosomal-recessive late-onset CMT2.

Figure 2

Clinical and Neuropathological Findings in Individuals with MME Mutations (A) Distal muscle wasting and foot drop in individuals carrying heterozygous mutations in MME. Shown are individuals AT1/III-5 and AT2/III-6 (c.466delC [p.Pro156Leufs^∗14]), individual AT3/IV-2 (c.71G>A [p.Trp24^∗]), and individual AT8/III-2 (c.1265C>A [p.Ala422Asp]). Sequencing electropherograms displaying genotypes are shown above the clinical photographs. Note that reverse complement sequences are provided for the c.466delC (p.Pro156Leufs^∗14) single-nucleotide deletion. (B–D) Severe axonal neuropathy in the sural nerve biopsy of individual US3/IV-2. (B and C) The density of large-caliber myelinated axons is severely reduced while small myelinated fibers (Aδ) look largely normal in number. Semithin sections, toluidine blue staining. (D) Electron microscopy revealed denervation of non-myelinating Schwann cells with collagen pockets and empty Schwann cell units.

Figure 3

Neprilysin Levels in Biological Samples and Measurement of Neprilysin Activity In Vitro (A–C) Neprilysin levels in affected individuals with MME mutations and control subjects. (A) Dot plots comparing neprilysin levels in subcutaneous adipose tissue obtained from individuals with MME mutations (n = 6) and from healthy age-matched control subjects (n = 6). The symbols represent neprilysin concentrations for the individuals in each group and the bars indicate the respective mean values. In the cases group, dots represent individuals with missense mutations and diamonds represent individuals with loss-of-function mutations. After verification of normal distribution (Shapiro-Wilk test), the data were analyzed with an unpaired two-sided t test. The exact p value is given. (B and C) Dot plots comparing neprilysin levels in blood plasma obtained from affected individuals with MME mutations and from healthy age-matched control subjects. (B) EDTA plasma. Case subjects, n = 7; control subjects, n = 24. (C) Citrate plasma. Case subjects, n = 7; control subjects, n = 25. The symbols represent neprilysin concentrations for the individuals in each group and the bars indicate the respective median values. Empty dots were considered outliers and were not used for statistical analysis. In the cases group, dots represent individuals with missense mutations and diamonds represent individuals with loss-of-function mutations. Since measured values were not normally distributed (Shapiro-Wilk tests), the data were analyzed with a Mann-Whitney U test. The exact p values are given. (D) Measurements of in vitro neprilysin activity using HEK293 cells transiently transfected with plasmids encoding human wild-type neprilysin, the neuropathy-related mutants c.1040A>G (p.Tyr347Cys) and c.1265C>A (p.Ala422Asp), and the artificial catalytically inactive neprilysin mutant c.1754A>T (p.Glu585Val). Fluorescence intensity was measured at 420 nm (λexc = 320 nm) after cells were incubated with 50 mM Tris buffer (pH 7.4) containing 5 μM of the neprilysin substrate Abz-GGfL-Agp-RV-EDA-Dnp. HEK293 cells transfected with an empty pcDNA3.1 vector served as control. Measurements were performed twice in triplicate and results are presented relatively to control cells as mean ± standard deviation (^∗p < 0.05, ^∗∗∗p < 0.001).

Figure 4

Electron Microscopy of Femoral Nerves of 13- to 14-Month-Old Wild-Type and Mme^−/− Mice (A) Some large-caliber myelinated axons in Mme^−/− mice (right, asterisks) show a thicker myelin sheath than usually found in wild-type mice (left). Scale bars represent 5 μm. (B) Axons of Remak fibers in wild-type mice (left) are usually separated by Schwann cell processes (arrows). Such separating Schwann cell protrusions are rare in Mme^−/− mice (right) and axons often show direct contacts with neighboring axons (arrows). Scale bars represent 0.5 μm.