Structure-function relation of phospholamban: modulation of channel activity as a potential regulator of SERCA activity

Abstract

Phospholamban (PLN) is a small integral membrane protein, which binds and inhibits in a yet unknown fashion the Ca(2+)-ATPase (SERCA) in the sarcoplasmic reticulum. When reconstituted in planar lipid bilayers PLN exhibits ion channel activity with a low unitary conductance. From the effect of non-electrolyte polymers on this unitary conductance we estimate a narrow pore with a diameter of ca. 2.2 Å for this channel. This value is similar to that reported for the central pore in the structure of the PLN pentamer. Hence the PLN pentamer, which is in equilibrium with the monomer, is the most likely channel forming structure. Reconstituted PLN mutants, which either stabilize (K27A and R9C) or destabilize (I47A) the PLN pentamer and also phosphorylated PLN still generate the same unitary conductance of the wt/non-phosphorylated PLN. However the open probability of the phosphorylated PLN and of the R9C mutant is significantly lower than that of the respective wt/non-phosphorylated control. In the context of data on PLN/SERCA interaction and on Ca(2+) accumulation in the sarcoplasmic reticulum the present results are consistent with the view that PLN channel activity could participate in the balancing of charge during Ca(2+) uptake. A reduced total conductance of the K(+) transporting PLN by phosphorylation or by the R9C mutation may stimulate Ca(2+) uptake in the same way as an inhibition of K(+) channels in the SR membrane. The R9C-PLN mutation, a putative cause of dilated cardiomyopathy, might hence affect SERCA activity also via its inherent low open probability.

Figure 1. Titration curve: unitary channel conductance vs KCl concentration.

(A) Example of current fluctuations measured at +80 mV in symmetrical solution with 100 mM or 750 mM KCl in 10 mM MOPS (pH = 7) buffer. (B) Titration of unitary channel conductance generated by PLN as a function of the KCl concentration. Experiments were done in symmetrical solution with KCl at increasing concentration in 10 mM MOPS (pH = 7) buffer. Fit of data with equation (1) yields concentration for half maximal conductance at 125 mM and a maximal conductance of 18 pS.

Figure 2. Estimation of pore size by means Polyethylene glycol method.

(A) Normalized unitary channel conductance of PLN channel as a function of the hydrodynamic radius of polymers in buffer solution. Error bars indicate the standard error of the mean considering 4 independent experiments. (B) Second derivative of the fit in A. Arrows indicate the minimum and the maximum pore radius respectively.

Figure 3. Effects of Phosphorylation on Phospholamban forming channel activity.

(A) Current/voltage relation of unitary channel currents generated by phosphorylated (open circles) and non-phosphorylated wt-PLN (full squares) in symmetrical solutions with 500 mM KCl in 10 mM MOPS (pH = 7) buffer. Data are mean ± standard deviation (SD) of 10 independent experiments for the wt-PLN and 4 independent experiments for the pPLN. (B) The mean open probability Po (±SD) of phosphorylated (open circles) and non-phosphorylated PLN (full squares) obtained in symmetrical solutions with 500 mM KCl in 10 mM MOPS (pH = 7) buffer. The mean was calculated on the basis of 8 and 5 independent experiments with the wtPLN and pPLN, respectively. (C) Relative distribution of measured open probabilities for wtPLN (black) phosphorylated wtPLN (gray) and for R9C mutant (white) of wtPLN. Data are normalized for the total number of recordings.

Figure 4. Effects of 3 selected PLN mutants (R9C, K27A, I47A) on channel activity.

(A) Current/voltage relation of unitary channel currents generated by wt-PLN (full squares), K27A (full triangles), I47A (open triangles) and R9C (open circles); all data recorded in symmetrical solutions with 500 mM KCl in 10 mM MOPS (pH = 7) buffer. Data are mean ± standard deviation (SD) of 10 independent experiments for the wt-PLN and a minimum of 9 recordings for each of the 3 mutants. (B) The mean open probability (±SD) of wt-PLN (full squares) and R9C mutant (open circles) obtained in symmetrical solutions with 500 mM KCl in 10 mM MOPS (pH = 7) buffer. The mean was calculated on the basis of 8 independent experiments for both the wtPLN and the R9C mutant.

Figure 5. Increase of open probability in R9C mutant in reducing/oxidizing conditions.

(A) Scatter plot of open probability for wtPLN (open symbols) and R9C mutant (filled symbols) before (x-axis) and after adding (y-axis) 80 mM DTT (circles) or 10 mM H₂O₂ (squares) to trans and cis chamber. The small symbols show the mean Po values from multiple clamp protocols in ≥4 individual bilayers; for each Po value data from clamp steps to +80, +60, +40, −40, −60 and −80 mV were pooled. In the case that more than one channel was active in a bilayer we estimated the number of channels from the maximum number of simultaneous openings observed in the absence and presence of the redox compounds. The corresponding large symbols represent the mean ±SD of the independent bilayer recordings given by the small symbols. (B) Example of current fluctuations measured in the same bilayer at −80 mV in the absence and in the presence of 80 mM DTT. (C) Example of current fluctuations measured in the same bilayer at −80 mV in the absence and in the presence of 10 mM H₂O₂. In B and C the solid lines represent the close state of the channel. Dashed lines represent the open level of the high conductance state.