Cell‐free expressed uniporter and symporter systems from the plant <scp>HKT</scp> transporter family display channel‐like gating and unitary conductances

Many areas of cultivated land are affected by salinity, including c. 30% of irrigated soils due to the use of poor-quality water (Hopmans et al., 2021). Since a large majority of food crops are highly sensitive to salt (Zörb et al., 2019), salinity is considered as one of the most limiting factors in crop production. High-affinity K+ transporter (HKT) genes have been identified as corresponding to the main salt tolerance QTLs in plants of many species, particularly cereals (Horie et al., 2009; Dave et al., 2022). HKTs include both active and passive transport systems. The name ‘HKT’ was first attributed to a wheat HKT (now named TaHKT2;1) originally thought to function as H+-K+ symporter (Schachtman & Schroeder, 1994), but which was shown in subsequent studies to be endowed with Na+-driven K+ transport activity (Rubio et al., 1995). The choice of HKT as the name for the transporter family to which TaHKT2;1 belongs is, however, misleading as some homologous systems to TaHKT2;1 in wheat and other plant species were later shown to selectively transport Na+ (Uozumi et al., 2000; Jabnoune et al., 2009; Munns et al., 2012). Thus, despite their name, only some HKTs are permeable to K+. These systems, which all belong to HKT class II (Platten et al., 2006), are also permeable to Na+. K+ transport in these systems can be achieved together with Na+ transport in a coupled manner (Rubio et al., 1995; Oomen et al., 2012), resulting then in K+-Na+ symport activity able to allow active K+ transport (i.e. transport against the transmembrane K+ electrochemical gradient). The other HKTs, which belong to HKT class I, are poorly permeable to K+ and behave like Na+-selective passive uniporters (Jabnoune et al., 2009; Böhm et al., 2016). This knowledge of HKT functional features has been obtained from the analysis of ionic fluxes or macroscopic currents measured on HKT-expressing cells, essentially yeast and Xenopus oocytes (Rubio et al., 1995; Horie et al., 2001; Jabnoune et al., 2009; Munns et al., 2012; Oomen et al., 2012; Sassi et al., 2012; Böhm et al., 2016). No information is available on their properties at the single-protein level. We focused our work on one HKT of each class: TmHKT1;4-A2 (from einkorn, Triticum monococcum), a Na+ uniporter responsible for salt tolerance in the Nax1 QTL (Tounsi et al., 2016); and OsHKT2;2 (from rice, Oryza sativa), a Na+-K+ symporter (Horie et al., 2001; Oomen et al., 2012). The activity at the single-protein level of both transporters, produced in vitro, was assessed in HKT-enriched artificial lipid bilayer membranes (BLMs). By demonstrating a channel-like gating mechanism and unitary conductances in uniporter and symporter HKTs, we show that both types of transporters display these channel properties and can act as coordinated units. TmHKT1;4-A2 and OsHKT2;2 are HKT members from einkorn (T. monococcum) and rice (O. sativa), respectively. The TmHKT1;4-A2 ORF was amplified by polymerase chain reaction on a TmHKT1;4-A2 cDNA clone (Tounsi et al., 2016) using the forward and reverse primers TK7X-L (5′-CGACTAGTTGCTCGAGATGGCCGGAGCTCATCATAAGGTCGGC-3′) and TK7N-R (5′-CCGAAGATTGCGGCCGCTAACTAAGTTTCCAGGCTTTGCCTCCTTTC-3′). The polymerase chain reaction fragment was digested with Xho I and Not I and ligated into the pEU-E01-MCS expression vector (CellFree Sciences Co., Ehime, Japan) digested with the same enzymes, leading to the pEU-HKT1.4 construct. The OsHKT2;2 ORF was amplified by polymerase chain reaction on the OsSLF1 cDNA clone (Oomen et al., 2012) using the primers Fus2-L (5′-GATATCACTAGTTCTCGAGATGACGAGCATTTACCAAGAATTC-3′) and Fus2-R (5′-CCAATTTATGGCGGCCGCTACCATAGCCTCCAATATTCAC-3′). The polymerase chain reaction fragment was integrated into the Xho I/Not I digested pEU-E01-MCS vector using InFusion cloning according to the manufacturer's instructions (Clontech, Takara Bio Inc., Shiga, Japan), leading to the pEU-HKT2.2 construct. The pEU-HKT1.4 and pEU-HKT2.2 constructs were designed to enable cell-free expression of the native tagless TmHKT1;4-A2 and OsHKT2;2 proteins, respectively, devoid of any additional amino acids. Liposomes of homogeneous size were obtained by extruding 5 ml of a solution of 50 mg ml−1 dimyristoylphosphatidylcholine (DMPC; Avanti Polar Lipids Inc., Alabaster, AL, USA) through a 100-nm filter using a 10-ml LIPEX Extruder (Northern Lipids Inc., Burnaby, BC, Canada does not exist anymore but a newer version of this device is available at Evonik Industries AG, Essen, Germany) following the manufacturer's instructions. High-affinity K+ transporters were produced in vitro by wheat germ cell-free protein synthesis (WG-CFPS; Fogeron et al., 2021). Transcription and translation were performed using the WEPRO7240 Dialysis Kit (CellFree Sciences) according to the manufacturer's instructions. Briefly, the empty vector pEU-E01-MCS (CellFree Sciences) and HKT constructs (pEU-HKT1.4 or pEU-HKT2.2) were transcribed in vitro with the SP6 RNA Polymerase for 6 h at 37°C and the whole transcription mixture was used for in vitro translation without RNA purification. In vitro translation was performed at 20°C for 40 h in a Slide-A-Lyzer MINI dialysis device 10 K MWCO (Thermo Fisher Scientific Corp., Waltham, MA, USA) in the presence of 37.5 μl of wheat germ extract and 10 mg ml−1 of extruded DMPC liposomes. Partial purification of HKT proteoliposomes (HKT-PLs) from total proteins was achieved by recovering the pellet (P fraction) obtained after a low-speed (4000 g, 1 h) centrifugation step followed by two washes with a phosphate-buffered saline solution (10 mM Na2HPO4, 10 mM NaH2PO4, 2.7 mM KCl, and 137 mM NaCl, pH 7.4). Protein fractions were analyzed by SDS-PAGE on a 12% polyacrylamide gel with a protein ladder (PageRuler Prestained Protein Ladder; Thermo Fisher Scientific Corp.). Quantity ratios of 1 : 235 of total proteins (T), 1 : 235 of the supernatant fraction (SN), and 1 : 50 of the P fraction were loaded on the gel. Mass spectrometry analysis of the major band in the P fraction of each transporter was performed according to Erde et al. (2014) with modifications (Supporting Information Methods S1). Electrophysiological activity of TmHKT1;4-A2 and OsHKT2;2 at the single-protein level was assessed using a horizontal BLM device (Orbit mini; Nanion Technologies GmbH, Munich, Germany) according to the manufacturer's instructions and Zaitseva et al. (2021). This system enables the simultaneous monitoring of four BLM systems. In each of them, two compartments are separated by a partition with a 50-μm hole over which the lipid bilayer was formed by painting with a solution of 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC; Avanti Lipids, 10 mg ml−1 in nonane). The two compartments were filled with the same solution (e.g. 100 mM KCl + 100 mM NaCl, symmetrical condition) or different salt solutions (e.g. 200 mM KCl vs 200 mM NaCl, asymmetric conditions). In order to set up asymmetric ionic conditions, the solution initially present in the ‘out’ compartment was carefully exchanged with the desired solution using a peristaltic pump. The resistance of BLMs was between 2 and 35 GΩ. To increase PL density and facilitate their fusion to BLMs, 200 mM sucrose was added to HKT-PLs (P fraction). HKT-PL samples (0.5–1 μl) were gently pipetted close to BLMs. After a short pulse at 200 mV, voltage was set at 100 mV until the first fusion events occurred. Application of various potentials to the BLMs and recording of the generated currents were performed using the software Elements Data Reader v.3.8.18 (Elements SRL, Cesena, Italy). Signals were acquired at a sampling rate of 1.25 kHz and were further analyzed using the software Axon Clampfit v.10.7 (Molecular Devices LLC, San José, CA, USA). Frequency histograms of current amplitudes were fitted to Gaussian functions to identify the values of current corresponding to the closed and open states of the HKT pores. Cation selectivity of TmHKT1;4-A2 was analyzed using a Goldman–Hodgkin–Katz equation (Hille, 2001). Coupled ion transport formalism was used to determine the theoretical reversal potential of currents in OsHKT2;2 Na+-K+ symporter, as in Rubio et al. (1995). TmHKT1;4-A2 and OsHKT2;2 transporters were produced by WG-CFPS in the presence of DMPC liposomes to generate HKT proteoliposomes (HKT-PLs). Both proteins were efficiently expressed in the WG-CFPS system (Fig. 1a) since they were detectable by SDS-PAGE in the total protein fraction (T), TmHKT1;4-A2 being even the major protein. Following low-speed centrifugation of total proteins, HKT-PLs were essentially recovered and partially purified in the pellet (P) fraction. The identity of the synthesized HKTs was confirmed by mass spectrometry analysis of the major band in the P fraction for each transporter (Fig. S1). Trypsin digestion of native HKT-PLs and mass spectrometry analysis indicated that both HKT proteins could be inserted in liposome membrane in either orientation (Fig. S2). HKT-PLs from the P fraction were integrated into BLMs in a horizontal BLM device. Since TmHKT1;4-A2 and OsHKT2;2 characterized at the macromolecular level upon heterologous expression were found to selectively transport Na+ (TmHKT1;4-A2; Tounsi et al., 2016) or co-transport Na+ and K+ (OsHKT2;2; Horie et al., 2001; Oomen et al., 2012), their unitary properties upon integration into BLMs were first investigated under ionic conditions containing both cations: 100 mM NaCl + 100 mM KCl on either side of the membrane (Fig. 1b–d). Upon electrical polarization of the membrane, repeated current fluctuations of varying durations were observed with both transporters (Fig. 1b), reminiscent of channel oscillations between closed and open states, that is of gating events such as those typically exhibited by various ion channels when their activity is analyzed at the level of a single protein (Sakmann & Neher, 1984; Hille, 2001). Such current fluctuations were not observed in control membranes (in the absence of PL addition to the mixture or in the presence of control empty PLs, Fig. S3). The unitary conductance of the transporters determined from the smallest amplitude of current fluctuation observed at each voltage, assumed to reflect the pore opening of a single protein, was 3.5 pS for TmHKT1;4-A2 and 8 pS for OsHKT2;2 (Fig. 1b–d). Thus, this whole set of recordings revealed that both transporters displayed channel-type unitary conductances and gating events. Ionic conditions were then changed to examine the ionic selectivity of the two transporters in BLMs. In the first series of experiments, the large weakly permeant cation N-methyl-d-glucamine (NMDG) was brought as chloride salt on either side of the BLM (Fig. S4). No channel-type activity was then recorded with either HKT (Fig. S4a,e). Under asymmetric conditions with NMDG-Cl 200 mM on the outer side (out) and NaCl 200 mM on the inner side (in), no channel-type activity was observed with OsHKT2;2 either (Fig. S4f). With TmHKT1;4-A2, instead, channel-type activity was recorded (Fig. S4b). The currents reversed under these conditions to a very negative value, c. −100 mV (Fig. S4b,c), in contrast to 0 mV under the symmetrical 100 mM NaCl + 100 mM KCl conditions (Fig. 1c). Such very negative reversal potential (Erev) of the unitary currents under the asymmetric presence of NMDG corresponded to a permeability ratio PNa : PNMDG of 50. Thus, the experiments using NMDG confirmed that the channel-type currents recorded in the presence of Na+ and K+ (Fig. 1) actually corresponded to the activity of the reconstituted HKTs and were not due to artifactual cationic interactions with the BLMs. TmHKT1;4-A2 showed a very low permeability to NMDG, while OsHKT2;2 activity was abolished when NMDG was present as the sole cation on one side of the membrane. The latter observation is in agreement with the absence of current reported in oocytes expressing OsHKT2;2 or TaHKT2;1 in the presence of K+ as the only monovalent cation in the external medium, which may indicate the need for Na+ on both sides of the membrane to enable the activity of these symporters (Gassmann et al., 1996; Horie et al., 2001). A second series of ionic selectivity experiments was carried out with asymmetric Na+ and K+ conditions (200 mM NaCl on one side and 200 mM KCl on the other side of the BLM) in order to discriminate between activities of the two transporters endowed with different Na+ vs K+ permeabilities (Fig. 1c,d). In experiments using OsHKT2;2, previously described as a Na+-K+ symporter, a small concentration of the co-transported ion (2 mM) was added to both compartments to ensure that transport activity was not impeded. In these asymmetric conditions, for TmHKT1;4-A2 with Na+ out and K+ in, the Erev of the unitary currents was shifted positively by 56 mV (Fig. 1c; it was 0 mV in symmetric conditions, Fig. 1c), in agreement with the Na+ to K+ permeability ratio of c. 10 determined at macroscopic level in oocytes (Tounsi et al., 2016). Also, when external Na+ concentration was increased to 200 mM compared with the 100 mM condition, the conductance of TmHKT1;4-A2 increased (Fig. 1c), as expected from oocyte data indicating that saturation of the conductance of this system was not reached at 100 mM Na+ (Tounsi et al., 2016). In OsHKT2;2, Erev of unitary currents under symmetric and asymmetric conditions remained close to 0 mV (Fig. 1d), as expected for a Na+-K+ coupled system displaying similar permeabilities to both ions (Rubio et al., 1995; Oomen et al., 2012) when the product of Na+ and K+ concentration values is identical on either face of the membrane. It should also be noted that the conductance of OsHKT2;2, unlike that of TmHKT1;4-A2, was much lower in the asymmetric than in the symmetric conditions (Fig. 1d), which can be ascribed to the very low concentration of the co-transported ion (Rubio et al., 1995; Horie et al., 2001; Oomen et al., 2012). Overall, these data indicated that the two cell-free expressed transporters exhibited the same ion transport properties as when expressed in oocytes or yeast. Most bilayers displayed multiprotein activity after HKT-PL fusion, as shown by multiple levels of current fluctuations at a given membrane voltage (Fig. 2a,b,e,f). Multiple current levels were observed, which may be interpreted as revealing several substates of a single conductance or the simultaneous opening or closure of several protein pores. An analysis of the main amplitudes of current fluctuations, not all of which were observed in each bilayer, was carried out using 14 and 17 bilayers in which the TmHKT1;4-A2 and OsHKT2;2 proteins, respectively, were reconstituted and placed under symmetrical bi-ionic conditions (100 mM Na+ + 100 mM K+; Fig. 2c,d). The different amplitudes of current fluctuation defined 6 (for TmHKT1;4-A2) and 7 (for OsHKT2;2) levels of conductance, a basal value (taken as the unitary conductance; Fig. 1b–d) and multiples of this value (2-, 3-, 4-, 6-, and 9-fold for TmHKT1;4-A2; 2-, 3-, 5-, 10-, 15-, and 30-fold for OsHKT2;2; Figs 2c,d, S5). The large number of conductance levels (≥ 6) in both HKTs renders the hypothesis that they correspond to different states of pore opening in a single protein (i.e. pore fully open with several substates) unlikely. The fact that the different levels of conductances were multiples of a basal value rather supports the hypothesis of the simultaneous activity (synchronized opening and closure events) of several pores of a single nature in each bilayer type. Recordings in asymmetric conditions of Na+ and K+ (Fig. 2e,f) or NMDG (Fig. S4d) revealed that the different levels of conductance displayed the same ionic selectivity, supporting the hypothesis of the single nature of the pores. Regarding the observed different levels of conductances, it is important to note that large stretches of repeated current fluctuations corresponding to simultaneous opening and closure of a fixed number of protein pores were often observed: for example the top two traces in Fig. 2(a) showing c. 30 events of simultaneous opening and closure of 6 TmHKT1;4-A2 pores over 7 s, and the top trace in Fig. 2(b) showing nine repetitions of simultaneous opening and closure of 30 OsHKT2;2 pores over 9 s. Altogether, these results support the hypothesis that both transporters can synchronize their activity within protein clusters. This study was carried out entirely in vitro, from DNA to functional analysis in artificial membranes, allowing the biophysical properties of individual HKT proteins to be deciphered in an environment free of other plasma membrane (PM) transporters. These conditions (e.g. lipid composition of synthetic membranes) and the fact that HKT reconstitution in BLM can very likely occur in either orientation may affect some aspects of HKT activity (e.g. synchronized gating). Nevertheless, the functional results obtained here at the single channel level are in perfect agreement with the selectivity and affinity data determined at the macroscopic level in heterologous expression systems, suggesting that the activity of the tag-free HKT proteins obtained by WG-CFPS was similar to that of the corresponding proteins when expressed in the PM of heterologous expression systems. The plant HKT transporters and their fungal and bacterial relatives Trk/Ktr (Corratgé-Faillie et al., 2010) have been proposed to have evolved from K+ channels based on sequence analysis and modeling (Durell et al., 1999). Furthermore, bacterial Trk/Ktr K+ transporters have been shown to share structural similarities with channel proteins, including the presence of a channel-type pore (Vieira-Pires et al., 2013). Thus, class I HKTs which have been shown to passively transport Na+ when expressed in Xenopus oocytes (Jabnoune et al., 2009; Böhm et al., 2016; Tounsi et al., 2016) have sometimes been proposed to be Na+-selective ion channels (Böhm et al., 2016, 2018). Recently, structural resolution of an Arabidopsis class I HKT and a wheat class II HKT confirmed broad similarity with the channel structure for both types of HKT (Wang et al., 2024). The electrophysiological recordings obtained here at the single-protein level provide direct evidence of an ion transport mechanism closely related to that of channels both in the case of the Na+ uniporter TmHKT1;4-A2 and the Na+-K+ symporter OsHKT2;2. The above conclusion raises the question of how the channel-like transport mechanism of OsHKT2;2 can underlie Na+-K+ symport activity. Over the past two decades, the boundary between channels and transporters has blurred, with the increasing availability of structural information on ion transport proteins and some reports of channel-like behavior among transporters (Ashcroft et al., 2009). In some cases, this has led to reconsideration, for the transporters studied, of the dominant model of large conformational in of channel-like with interactions with and gating of & In the case of OsHKT2;2, the conductance values and open can 100 in this symporter not support the hypothesis that the transport large conformational and gating closely to the present data and the structural and functional similarities between the OsHKT2;2 symporter and the TmHKT1;4-A2 the hypothesis to the of OsHKT2;2 to function as a symporter is that the pore of this protein enables Na+ and K+ by as a pore within which ions in single (Gassmann et al., the transporter pore be to allow the ions to or each other (Wang et al., Na+ and K+ the step and the fluxes of these two ions within the pore not be of each In this the of one ion electrochemical can the of the other ion by it against modeling has that interactions are likely to occur between HKT proteins and to in of as (Durell & 1999). This was confirmed by the structure of HKTs, achieved on (Wang et al., 2024). functional at the single transporter level support such and due to the observation of synchronized transporter activity within that these interactions have a functional and 2 HKT transporters within each BLM experiments were performed at in the and Mass spectrometry experiments were carried out using the of the and designed the and performed the and analyzed the and the DNA constructs are available from data for the of the are available as data for the of the Fig. Mass of native TmHKT1;4-A2 and OsHKT2;2 proteins expressed in Fig. the orientation of HKT proteins in proteoliposomes by mass Fig. for current in HKT-enriched BLMs. Fig. No or low permeability of TmHKT1;4-A2 and OsHKT2;2 BLMs to Fig. of current traces and histograms the multiple levels of conductances displayed by TmHKT1;4-A2 and OsHKT2;2 reconstituted in BLMs. Methods Mass spectrometry is not responsible for the or of any Information by the than should be to the The is not responsible for the or of any supporting information by the than should be to the corresponding for the

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.