K+-dependent paradoxical membrane depolarization and Na+ overload, major and reversible contributors to weakness by ion channel leaks

Abstract

Normal resting potential (P1) of myofibers follows the Nernst equation, exhibiting about -85 mV at a normal extracellular K(+) concentration ([K(+)](o)) of 4 mM. Hyperpolarization occurs with decreased [K(+)](o), although at [K(+)](o) < 1.0 mM, myofibers paradoxically depolarize to a second stable potential of -60 mV (P2). In rat myofiber bundles, P2 also was found at more physiological [K(+)](o) and was associated with inexcitability. To increase the relative frequency of P2 to 50%, [K(+)](o) needed to be lowered to 1.5 mM. In the presence of the ionophore gramicidin, [K(+)](o) reduction to only 2.5 mM yielded the same effect. Acetazolamide normalized this increased frequency of P2 fibers. The findings mimic hypokalemic periodic paralysis (HypoPP), a channelopathy characterized by hypokalemia-induced weakness. Of myofibers from 7 HypoPP patients, up to 25% were in P2 at a [K(+)](o) of 4 mM, in accordance with their permanent weakness, and up to 99% were in P2 at a [K(+)](o) of 1.5 mM, in accordance with their paralytic attacks. Of 36 HypoPP patients, 25 had permanent weakness and myoplasmic intracellular Na(+) ([Na(+)](i)) overload (up to 24 mM) as shown by in vivo (23)Na-MRI. Acetazolamide normalized [Na(+)](i) and increased muscle strength. HypoPP myofibers showed a nonselective cation leak of 12-19.5 microS/cm(2), which may explain the Na(+) overload. The leak sensitizes myofibers to reduced serum K(+), and the resulting membrane depolarization causes the weakness. We postulate that the principle of paradoxical depolarization and loss of function upon [K(+)](o) reduction may apply to other tissues, such as heart or brain, when they become leaky (e.g., because of ischemia).

Fig. 1.

¹H and ²³Na measurements in the calf muscles of HypoPP patients. (A–C) Axial T1-weighted MR images from 3 related Cav1.1-R1239H patients: a 35-year-old woman (A), her 55-year-old uncle (B), and her 80-year-old grandmother whose limb muscles were predominantly replaced with fat (C). (D–I) T2-weighted STIR ¹H (Left) and ²³Na-MR images (Right) from a healthy control (D and E) and a 37-year-old patient (F–I), the sister of the patient in A. The images in F and G were taken before treatment, and the images in H and I were taken after treatment with 250 mg/d AZ for 4 weeks. Note the very high hydrogen intensities in the STIR sequence (F) and the elevated Na⁺ concentration before treatment (G, arrows indicate highest Na signal intensities) and their improvement after treatment. The central reference contains 0.3% NaCl solution; occasional side tubes containing 0.3% NaCl in 1% agarose (Left) and 0.6% NaCl in H₂O (Right) were additional standards.

Fig. 2.

Correlation of strength and [Na⁺]_i in 36 HypoPP patients. The [Na⁺]_i values obtained with a ²³Na-MRI of the lower legs and the values of plantar flexor muscle strength measured according to the MRC grading scale (closed symbols) show a clear correlation. Values from 12 healthy controls are shown as open circles.

Fig. 3.

Resting (E_m) and action potentials of normal and HypoPP fibers. (A) Bimodal distribution of E_m of human control fibers (white, n = 369) and R1239H-HypoPP muscle fibers (gray, n = 128) in 4-mM K⁺. HypoPP fibers have an increased probability for the second stable membrane potential of −60 mV (P2); the first stable membrane potential (P1) in HypoPP fibers is depolarized by ≈ 10 mV compared with the controls. (B) Action potentials (AP) in the endplate region elicited from various holding potentials of −82 mV (a), −75 mV (b), −62 mV (c), and −59 mV (d). At c there is a minimal AP beginning at the arrow; at d there is no AP, only the endplate potential.

Fig. 4.

Kir conductance and cation leak in human myofibers. Steady-state current density–voltage relationships were determined in 1.0-mM K⁺ for 7 normal myofibers (1K control) and 4 (1K R528H) and 5 myofibers (1K R1239H) from patients. The difference between the curves for patients and controls corresponds to the mutation-related leak current. In addition, 8 normal myofibers (4K control) were measured in 4.0-mM K⁺. Note that the 4K curve is much steeper (g_max = 261 μS/cm² at hyperpolarization) than the 1K curve (g_max = 26.5 μS/cm²) of the control fibers according to the higher Kir conductance in 4-mM K⁺. To avoid superimposing currents, the solutions contained TTX, not Cl⁻.

Fig. 5.

Resting potentials (E_m) of rat muscle strips treated with GD. (A) Bimodal distribution of E_m in control (white, n = 257 fibers) and rat diaphragm strips treated with 0.1 μM GD (gray, n = 265 fibers) to create leaks. As in HypoPP, the relative frequency of the second stable potential (P2) at −60 mV increased, and the first stable potential (P1) was depolarized by ≈ 5 mV for the leaky fibers. (B) Relative P1 frequency dependent on external K⁺. Fitting the data points (6–7 strips each) to a multimodal log-normal probability density function enabled comparison. The turning point is at 2.5 mM K⁺ in GD-treated leaky fibers, at 2.9 mM in AmB-treated fibers, and at 1.5 mM for controls, suggesting that the leaky fibers have a higher probability for the second stable potential (P2 = −60 mV) when [K⁺]_o decreases. The addition of 100 μM AZ to the leaky fibers resulted in the restoration of the fraction of polarized fibers, particularly in low K⁺. (C) Dependence of twitch force on [K⁺]_o (5 strips at each data point). The force reduction in GD-treated muscles and control muscles upon [K⁺]_o reduction reflects the loss of polarized fibers illustrated in B.