Functional and Structural Characterization of ClC-1 and Nav1.4 Channels Resulting from CLCN1 and SCN4A Mutations Identified Alone and Coexisting in Myotonic Patients

Non-dystrophic myotonias have been linked to loss-of-function mutations in the ClC-1 chloride channel or gain-of-function mutations in the Nav1.4 sodium channel. Here, we describe a family with members diagnosed with Thomsen’s disease. One novel mutation (p.W322*) in CLCN1 and one undescribed mutation (p.R1463H) in SCN4A are segregating in this family. The CLCN1-p.W322* was also found in an unrelated family, in compound heterozygosity with the known CLCN1-p.G355R mutation. One reported mutation, SCN4A-p.T1313M, was found in a third family. Both CLCN1 mutations exhibited loss-of-function: CLCN1-p.W322* probably leads to a non-viable truncated protein; for CLCN1-p.G355R, we predict structural damage, triggering important steric clashes. The SCN4A-p.R1463H produced a positive shift in the steady-state inactivation increasing window currents and a faster recovery from inactivation. These gain-of-function effects are probably due to a disruption of interaction R1463-D1356, which destabilizes the voltage sensor domain (VSD) IV and increases the flexibility of the S4-S5 linker. Finally, modelling suggested that the p.T1313M induces a strong decrease in protein flexibility on the III-IV linker. This study demonstrates that CLCN1-p.W322* and SCN4A-p.R1463H mutations can act alone or in combination as inducers of myotonia. Their co-segregation highlights the necessity for carrying out deep genetic analysis to provide accurate genetic counseling and management of patients.


Clinical picture of Family 1 (Figure 1) [1]
Since this family was presumptively affected with myotonic dystrophy type 1, we studied several features typical for DM1, such as cataracts. The ophthalmological examination was performed on four individuals, NDM1, 2, 6 and 15, who resulted cataract-free and none reported visual disturbances. Therefore, we did not perform this study in the remaining relatives.
None of the family members evaluated reported muscle pain. Only the proband (NDM15) had walking problems characterized by frequent falls and resulted to be the most affected individual in the family. NMD15 presented motor problems after resting for a long time. However, when he made repeated movements, these problems gradually disappeared (warm up phenomenon). Individuals NDM6 and NDM15, showed adiadochokinesia. NDM6 and NDM8 showed difficulties to manipulate objects with their hands. Individual NDM5 had muscle cramps.
None of the family members reported digestive issues or sleep disturbances. Almost all individuals showed normal muscle strength. The proband (NDM15) showed weakness in the scapular waist muscles and in the neck flexors.
All the individuals were positive for generalized and percussion myotonia. Myotonia was more evident in arms than in hands, and in the proximal than in the distal muscles. Individuals NDM6 and NDM15 also showed ocular myotonia and hypertrophy of calves. The proband NDM15 also presented the Gower's sign. Very poor myotatic reflexes were found in individuals NDM8 and NDM15, and fragmented eye movements were also found in patient NDM8. All individuals had normal sensitivity. Due to lumbar radiculopathy, patient NDM2 showed reflex asymmetry in the lower limbs.
Thus, clinical signs of all evaluated patients were compatible with a myotonic disease, more specifically to dominant myotonia congenita.

RFLP and MLPA assays
By using the restriction fragment length polymorphism (RFLP) assay, we were able to confirm/identify the genetic variants in the probands, relatives or NDM unaffected individuals. The The multiplex ligation-dependent probe amplification (MLPA) assay was performed on all 17 samples involved in this study. For this, we used the MRC-Holland kits: 1-SALSA MLPA 350 CLCN-1 -KCNJ2 for CLCN1 gene; and 2-SALSA MLPA P397 SCN4A -CACNA15 for SCN4A gene, following the manufacturer's instructions. Approximately 30ng of DNA was used to perform the MLPA assay. To perform the hybridization of the probes, DNA was incubated for 5 min at 98 °C and subsequently the probe mix and the MLPA buffer were added. The reaction was incubated for 1 min at 95 °C and then kept at 60 °C for 18 hours. For the ligation, the enzyme ligase, the ligase buffers A and B and the DNA (from previous step) were used. The reaction was incubated at 54 °C for 15 min, 98 °C for 5 minutes and then kept at 20 °C. For the amplification of the probes, 8 µ l of the previous reaction, water, the primers and the DNA polymerase were added and amplified in the thermal cycler with the following conditions: 95 °C for 30 s, 60 °C for 30 s and 72 °C for one minute (x33), 72 °C for 20 minutes, and then kept 15 °C. MLPA products were sent to the Macrogen (Korea) for sequencing. The analyzes of the results were carried out with the Coffalyser program of MRC Holland.

Heterologous expression and electrophysiological recordings
Xenopus oocytes were obtained from 2-5 years old frogs and defolliculated by digestion with collagenase from Clostridium histolyticum (Sigma-Aldrich, USA). Oocytes were injected/co-injected with ClC-1-cRNA (WT/mutant) or were co-injected with SCN4A-cRNA (WT or mutant) and beta1-cRNA and incubated 48-72 h at 18 ºC in a maintenance solution containing (in mM) 90 NaCl, 2 KCl, 1 MgCl2, 1 CaCl2, and 10 Hepes (pH 7.5). In order to measure chloride currents, the membrane potential was kept at the holding potential of -30 mV and subjected to voltages stimuli consisting of a prepulse to +60 mV for 50 ms followed by different voltage steps from -140 to +100 mV in 20 mV increment for 100 ms, and a final tail pulse to -100 mV for 100 ms. The composition of the extracellular solution (in mM) was: 100 NaCl, 10 HEPES-Na, 5 MgSO4 (pH 7.3) and the electrodes were filled with 3 M KCl.
The initial tail currents obtained by back-extrapolating single exponential functions fitted after the decay of the capacitive current to the onset of the voltage step were fitted with a modified Boltzmann function of the form: where Imax is the maximal current extrapolated by the fitting, pmin is the residual open probability at negative voltages, V1/2 is the voltage of half-maximal activation, and k the slope factor of the where I(V) are the currents fitted at each voltage step. Sodium currents were measured with different protocols and the extracellular solution had a reduced sodium concentration to reduce the magnitude of inward currents and the reliability of the voltage-clamp (in mM): 15 NaCl, 100 NMDG-Cl, 10 Hepes, 1.8 CaCl2, 1 MgCl2 (pH 7.3).
The membrane was kept at a holding potential of -90 mV. The standard I-V protocol was performed by applying 20 ms long pulses at voltages ranging from -50 to +20 in 5 mV increment. The steady-state activation was determined by fitting the peak current-voltage (I-V) relationship with the equation: where Vrev is the reversal potential, gmax the maximum conductance, zact the apparent gating valence, F the Faraday constant, R the gas constant and T the temperature. The parameter gmax is a measure of the functional expression level.
Steady-state inactivation (channel availability) was measured after 100-ms conditioning pulses to various voltages followed by a test pulse to -20 mV for 40 ms, and currents were fitted with the equation: where V represents the pre-pulse potential, is the maximal current, 1/2 the voltage of halfmaximal inactivation, and zinact is apparent valence of the inactivation gate.
The voltage-dependent activation and the steady state inactivation were plotted as a function of voltage and the area subtended by the intersection of the curves was calculated and defined as window current.
The time course of recovery from inactivation at -90 mV was measured by repolarizing the cell to -90 mV for a variable time from 0.5 to 100 ms after a 100 ms pulse to 0 mV and assessing channel availability by a final test-pulse to -10 mV. Peak-currents at the final test pulse were analyzed by fitting an exponential function: is time constant of recovery. Linear capacitive and leak currents were online subtracted using a standard P/4 protocol.

Additional clinical information for subject NDM17
NDM17 reported progressive muscle weakness and stiffness in his hands, showed myotonia, which got worse with exercise. The motor nerve conduction study resulted normal, with mild decreases in the CMAP of the right ulnar, while the sensory nerve conduction study showed a slight decrease in the amplitude of the right ulnar potential and a severe decrease in the amplitude of the right sural potential.

Polymorphisms NP_000074.3:p.Gly118Trp and NP_000074.3:p.Pro727Leu
Structural analysis of these two genetic variants indicated that: 1-G118W had a low disease-propensity score (40/100) and missense3D did not predict any structural damage. In addition, the ΔΔG is predicted to be stabilizing (0.526 kcal/mol). Although it might seem that the impact of changing a glycine to a tryptophan can be severe, there is a neighboring helix that also has a tryptophan (W303) in a similar spatial location, suggesting that there is a kind of structural accommodation. This variant is only reported once in ClinVar and it is classified as "benign".
2-P727L had a very low disease-propensity score (11/100) and missense3D did not predict any structural damage. Although predicted stabilizing (ΔΔG: 0.399 kcal/mol), the mutation is located downstream of the first cystathionine β synthase (CBS) domain, in a disordered region. This variant is also classified as "benign" in ClinVar. Interestingly, proline at position 727 is highly conserved (in a random coil), therefore, a change in this position could generate an important modification in the channel's function. However, the structural analysis performed does not suggest that.
3-Accordingly to these results, it is not expected that these variants can severely affect the structure and function of the ClC-1 channel. Table S1. SCN4A primers and PCR conditions used in this study.  Table S2. CLCN1 primers and PCR conditions used in this study, which were reported previously [2].