Arabidopsis HAK5 under low K+ availability operates as PMF powered high-affinity K+ transporter

Abstract

Plants can survive in soils of low micromolar potassium (K⁺) concentrations. Root K⁺ intake is accomplished by the K⁺ channel AKT1 and KUP/HAK/KT type high-affinity K⁺ transporters. Arabidopsis HAK5 mutants impaired in low K⁺ acquisition have been identified already more than two decades ago, the molecular mechanism, however, is still a matter of debate also because of lack of direct measurements of HAK5-mediated K⁺ currents. When we expressed AtHAK5 in Xenopus oocytes together with CBL1/CIPK23, no inward currents were elicited in sufficient K⁺ media. Under low K⁺ and inward-directed proton motive force (PMF), the inward K⁺ current increased indicating that HAK5 energetically couples the uphill transport of K⁺ to the downhill flux of H⁺. At extracellular K⁺ concentrations above 25 μM, the initial rise in current was followed by a concentration-graded inactivation. When we replaced Tyr450 in AtHAK5 to Ala the K⁺ affinity strongly decreased, indicating that AtHAK5 position Y450 holds a key for K⁺ sensing and transport. When the soil K⁺ concentration drops toward the range that thermodynamically cannot be covered by AKT1, the AtHAK5 K⁺/H⁺ symporter progressively takes over K⁺ nutrition. Therefore, optimizing K⁺ use efficiency of crops, HAK5 could be key for low K⁺ tolerant agriculture.

Fig. 1. Potassium dependent activation of AtHAK5.

A Representative original currents (black trace) at −120 mV of an oocyte co-expressing AtAKT1 and CIPK23/CBL1 at pH 4.5 in the presence of different potassium concentrations (2 mM, 20 µM or nominally 0 µM K⁺). Simultaneously, the bath K⁺ concentration was recorded via K⁺-selective electrodes (representative trace in red). B Representative K⁺-induced currents at −120 mV of either control oocytes or oocytes expressing either AtHAK5 alone or co-expressing AtHAK5 together with CIPK23/CBL1. C Quantification of K⁺-induced peak currents at −120 mV of either control oocytes (n = 4 experiments) or oocytes expressing AtHAK5 in the presence or absence of CIPK23/CBL1 (n = 5 experiments). D Measurement of original current traces (in black) at −120 mV in oocytes co-expressing AtHAK5 and CIPK23/CBL1 at pH 4.5 in the presence of different potassium concentrations (2 mM, 20 µM or nominally 0 µM K⁺). In the bath a K⁺-selective electrode simultaneously recorded the apparent K⁺ concentration (red trace; Representative traces are shown). E Normalized whole-oocyte K⁺-induced peak currents (ΔI_peak) (n = 3 experiments, mean ± SD) or steady-state currents (ΔI_SS) (n = 4 experiments, mean ± SEM) at −120 mV at pH4.5 are plotted against the applied K⁺-concentration. K_m (K⁺) was calculated by fitting ΔI_peak with a Michaelis-Menten equation. The modified Michaelis-Menten function used to fit ΔI_SS is described in the methods section.

Fig. 2. Cation dependency of AtHAK5.

A Box plot of cation-induced peak currents (ΔI_peak) at −120 mV of oocytes co-expressing AtHAK5 and CIPK23/CBL1 in response to either 2 mM Li⁺, Na⁺, K⁺, Rb⁺, Cs⁺ or NH₄⁺ (n = 5 experiments). B Representative current response of AtHAK5/CIPK23/CBL1 co-expressing oocytes upon application of different NH₄⁺ concentrations. C Normalized whole-oocyte NH₄⁺-induced peak currents (ΔI_peak) at −120 mV (pH4.5) plotted against the applied NH₄⁺-concentration. K_m (NH₄⁺) was calculated by fitting ΔI_peak with a Michaelis-Menten equation. (n = 6 experiments, mean ± SD). D Box plot of peak current response (ΔI_peak) of AtHAK5/CIPK23/CBL1 co-expressing oocytes in the presence of either 10 µM K⁺ or different NH₄⁺-concentrations as indicated in the figure (n = 4 experiments for 20, 50, 100 µM NH₄⁺ and n = 5 experiments for 10 µM K⁺ and NH₄⁺, mean ± SD).

Fig. 3. pH dependency of AtHAK5.

A Simultaneous recording of current response (black trace) and cytosolic pH changes (red trace) in oocytes expressing pHluorin:HAK5 with CIPK23/CBL1. Oocytes were clamped to −120 mV and currents were triggered by perfusion with 20 µM K⁺. A drop in 405 nm/470 nm ratio represents cytosolic acidification. Representative measurement from 4 independent experiments is shown. B Representative current response of AtHAK5/CIPK23/CBL1 co-expressing oocytes upon application of 200 µM K⁺ at different pH (as indicated in the figure). C Whole-oocyte K⁺-induced peak currents (ΔI_peak) at −120 mV plotted against the applied H⁺-concentration. K_m (H⁺) was calculated by fitting ΔI_peak with a Hill equation. (n = 4 experiments for pH 8.5, 7.5, 6 and 4, n = 8 experiments for pH 6.5, 5.5 and 4.5, mean ± SD). D Left panel: Representative current response of AtHAK5/CIPK23/CBL1 co-expressing oocytes at pH_ext 4.5 upon application of 2 mM K⁺ in the presence (red) or absence (black) of cytosolic acidification via sodium acetate treatment. pH_cyt was recorded via H⁺ selective electrodes. Right panel: Box plot of peak current responses upon application of 2 mM K⁺ at either pH_cyt 5.62 ± 0.1 (n = 4 experiments, mean ± SD) or pH_cyt 7.09 ± 0.08 (n = 5 experiments, mean ± SD).

Fig. 4. Voltage dependency of AtHAK5.

A Normalized whole-oocyte currents of AtHAK5 and CIPK23/CBL1 co-expressing oocytes. K⁺-induced peak currents (ΔI_peak) at different voltages (as indicated in the figure) (pH4.5) were plotted against the applied K⁺-concentration. Data points were fitted with a Michaelis-Menten equation. (n = 4 experiments, mean ± SEM). B Maximum currents (I_max) derived from the Michaelis-Menten fits shown in A) were plotted against the applied voltage (n = 4 experiments, mean ± SEM). C) K_m (K⁺) values derived from the Michaelis-Menten fits shown in A) were plotted as a function of the applied voltage (n = 4 experiments, mean ± SEM). D Normalized whole-oocyte currents of AtHAK5 and CIPK23/CBL1 co-expressing oocytes. Peak currents (ΔIpeak) were induced with 200 µM K⁺ at different voltages (as indicated in the figure) and plotted against the applied H⁺-concentration. Data points were fitted with a Hill equation (n = 5 experiments for −150 mV and n = 7 experiments for −60, −90, and −120 mV, mean ± SD). E) K_m [H⁺] values derived from the data shown in D) were plotted against the applied voltage (mean ± SD).

Fig. 5. Inactivation kinetics of AtHAK5.

A Representative current traces from oocytes co-expressing AtHAK5 and CIPK23/CBL1 at −120 mV when challenged with different K⁺ concentrations (as indicated in the figure) for 120 s. B Degree of inactivation (in %) derived from similar experiments as shown in A) were plotted against the applied K⁺ concentration (n = 3 experiments, mean ± SD). C Representative current response of AtHAK5/CIPK23/CBL1 co-expressing oocytes at pH4.5, pH5.5 or pH6.5 upon application of either 20, 200 or 2000 µM K⁺. D Box plot of the percentage of inactivation derived from similar experiments as shown in C) (number of experiments is indicated in the figure).

Fig. 6. Molecular nature of K⁺ sensing.

A Homology model of a single subunit of AtHAK5 based on the cryo-EM structure of KimA (PDB entry 6S3K). Only the integral transmembrane part comprising residues Q64 to R541 was modeled. The helices identifiable in the chosen orientation are labeled, the indicated box shows the region magnified and shown with more details in panel B. The two magenta spheres represent the two potassium ions located in the upper and lower coordination site, residue Y450 is indicated in stick representation with carbon atoms colored in green. B Magnification of the upper ion binding site around potassium ion K1. The stringent coordination of potassium ion K1 likely facilitates dehydration of the incoming cation. C Normalized whole-oocyte K⁺-induced peak currents (ΔI_peak) at −120 mV and pH4.5 plotted against the applied K⁺-concentration. Currents from oocytes expressing the mutant Y450A with CIPK23/CBL1 (red squares) are compared with WT HAK5/CIPK23/CBL1 expressing oocytes (black squares, cf. Fig. 1E). K_m (K⁺) was calculated by fitting ΔI_peak with a Michaelis-Menten equation. (n = 5 experiments, mean ± SD). D Currents from oocytes expressing the mutant Y450A with CIPK23/CBL1 (red squares) are compared with WT HAK5/CIPK23/CBL1 expressing oocytes (black squares, cf. Fig. 3C). Normalized whole-oocyte peak currents (ΔI_peak) were induced either by 10 mM (Y450A) or 200 µM (WT) K⁺ at −120 mV at different pH values and plotted against the applied H⁺-concentration. K_m (H⁺) was calculated by fitting ΔI_peak with a Hill equation (Hill coefficient = 2.17 ± 0.5 (Y450A) and 1.67 ± 0.3 (WT)) (n = 5 experiments, mean ± SD).