Arabidopsis (Arabidopsis thaliana) seeds were surface sterilized for 10 min in 80% ethanol, 5% w/v calcium hypochlorite, and 0.1% Triton X-100. After three washes in 80% ethanol and one in 100% ethanol, seeds were left to dry under sterile conditions. Seeds were sown on plates containing 1% w/v sucrose, half-strength MS salts (Duchefa), and 15 g l⁻¹ agar-agar (Merck) (pH 5.7). After stratification overnight at 4°C in darkness, plates were transferred to a growth chamber (16 h light/8 h darkness, 21°C, 100 μM m⁻² s⁻¹) for seed germination and maintained in a vertical position. In experiments with BFA treatments, 4-day old seedlings were incubated with 25 μM BFA (in the experiment with abp1 mutants) for 90 min or with 50 μM BFA (in the experiments with aux1 and tir1/afb mutants) dissolved in liquid 0.5 MS medium for 45 min or pre-treated with 10 μM 1-NAA (dissolved in liquid 0.5 MS medium) for 30 min followed by incubation with respective concentration of BFA and 10 μM 1-NAA. BFA stock solutions were made in DMSO up to a concentration of 50 mM. Control treatments contained an equal amount of DMSO. For electrophysiological experiments, Arabidopsis seedlings were grown sterile on 0.8–1% agarose supplemented with ¼-strength MS under controlled environmental conditions (12 h day vs. 12 h night; 21°C at day vs. 16°C at night; 120 μmol photons m⁻² s⁻¹) for 5 days. The following previously described lines of Arabidopsis thaliana Col-0, aux1-7, aux1-22, eir1-1 (pin2), tir1afb2afb3, abp1-c1, and abp1-TD1 lines have been used in this study.

Sterile grown seedlings were exposed to standard bath solution (0.1 mM KCl, 1 mM CaCl₂, 5 mM MES/BTP pH 5.5). Microelectrodes for impalement and preparation of application pipettes were pulled from borosilicate glass capillaries (Øout 1 mm, Øin 0.58 mm, w/filament, Hilgenberg, Germany) on a horizontal light amplification by stimulated emission of radiation puller (P2000, Sutter Instruments Co, USA). Microelectrodes were back-filled with 300 mM KCl and connected via an Ag/AgCl half-cell to a headstage (1 GΩ, HS-2A, Axon Inst., USA). The reference electrode was filled with 300 mM KCl as well. An IPA-2 amplifier (Applicable Electronics Inc., USA) and an NI USB-6259 interface (National Instruments, USA) were used for data collection. For application pipettes, the tips of microelectrodes were manually broken off to a 20–40 μm wide opening and back-filled with auxin-containing bath solution. Root hair cells of sterile grown A. thaliana seedlings were impaled under microscopic inspection (Axioskop, Zeiss, Germany) by using electronic micromanipulators (MM3A-LMP, Kleindiek Nanotechnik, Germany or Triple Axis Micromanipulator, Sensapex Oy, Finland). Application pipettes were also mounted on a micromanipulator (Triple Axis Micromanipulator, Sensapex Oy) and connected to a Picospritzer II microinjection dispense system (General Valve, USA) operating at 20 kPa to apply 1 s back pressure pulses. In cases where auxins were applied for longer periods of time, seedling roots were constantly perfused with bath solution (1 ml/min). In cases where application pipettes were applied, no perfusion was used.

For whole-mount immunolocalization of PIN1 and PIN2 in roots, 4-day-old Arabidopsis seedlings were fixed with 2% (w/v) paraformaldehyde (Fluka) and 0.4% Triton X-100 in microtubule-stabilizing buffer (MTSB) (pH 7.0) for 1 h and rinsed five times with MTSB/0.1% Triton X-100. Labeling was performed with rabbit anti-PIN1, guinea pig anti-PIN2 both at 1:500. Alexa Fluor 488-conjugated goat anti-rabbit and Alexa Fluor 555-anti-guinea pig secondary antibodies were used at 1:400 dilution. Solutions during the immunolocalization procedure were changed using a pipetting robot (Insitu Pro; Intavis).

Confocal images were taken using a Zeiss LSM 510 NLO confocal laser scanning microscope. Alexa Fluor 488 was excited by applying a 488 nm argon laser line in combination with a 500–550 band-pass filter. Alexa Fluor 555 was excited by applying a helium-neon 543 nm laser (HeNe laser) in conjunction with a 575-long pass filter. Quantitative analysis of confocal microscopic images was performed using Imaris 7.5.6 software (Bitplane AG). The fluorescence signal was detected using the “create surface” tool, and the fluorescence signal at the plasma membrane and in the BFA bodies was calculated. The level of signal internalization (the signal in the BFA bodies) was calculated as the ratio between an intensity of intracellular fluorescence signal and the intensity of total fluorescence signal expressed as a percentage. For every root, the estimation of the level of PIN internalization was based on 20–32 and 10–18 cells for PIN1 and PIN2, respectively. We used 5–10 roots for every treatment. Averages for every root were used for the calculation of standard deviation. Student’s t test was used for the evaluation of statistical significance between the experimental groups.

The lack of visible phenotypes in qualified null abp1 alleles led us to revisit the question of whether depolarization of the plasma membrane and endocytosis are initiated by the binding of auxin to ABP1. Using voltage-recording single-barreled microelectrodes, we measured plasma membrane potential in Arabidopsis thaliana accession Col-0 root cells before and after the external application of indole-3-acetic acid (IAA) and selected synthetic auxins (Figures 1A,B). In line with Dindas et al. (2018), resting membrane potentials were in the range of −150 mV. Plasma membrane potential depolarization was pH dependent after application of the active auxins IAA and 5F-IAA with depolarization being observed at acidic external pH (Figure 1B). In the case of IAA, strong depolarization of epidermal cell membranes was seen from 100 nM with an EC₅₀ of 200 nM (Figure 1B). In contrast, the synthetic auxin analog 1-NAA and its non-auxinic isomer 2-NAA did not induce comparable depolarizations at concentrations <1 μM (Dindas et al., 2018). In order to test whether ABP1-mediated auxin-induced depolarization, we compared the plasma membrane potential of wild type and recently generated abp1 knockout mutants (Gao et al., 2015). Impalement of root hair cells with voltage-recording microelectrodes revealed that locally applied short pulses of IAA caused an instant membrane depolarization of up to 76 mV in wild-type root hairs (Figure 1C). Membrane depolarization induced by pulses of IAA in plants homozygous for the abp1-c1 mutant allele (a full knock-out of the ABP1 gene) was indistinguishable from wild type. This result was corroborated with an independent null mutant allele abp1-TD (Gao et al., 2015). Figure 1D shows no significant changes in the absolute values of auxin-induced plasma membrane depolarization in all ABP1 knockout mutants tested. We therefore conclude that ABP1 does not mediate auxin-induced membrane potential depolarization.

Several families of plasma membrane localized proteins also specifically bind and transport auxin, with at least one, AUX1, mediating auxin-induced depolarization. In the current study, we investigated the additional auxin transport mutant, pin2 to understand whether members of the PIN family of auxin transporters are also able to affect auxin-induced plasma membrane depolarization. As we reported previously (Dindas et al., 2018), auxin-dependent depolarization was strongly altered in the two tested aux1 mutants, as well as in the tir1afb2afb3 auxin-receptor mutant (Figures 1D,E). We therefore conclude that AUX1, and not ABP1, acts as an auxin-dependent signaling platform for plasma membrane depolarization. Auxin-induced plasma membrane depolarization was also decreased in pin2 seedlings, which indicates that this response might be related to the regulation of auxin concentration in the cell (Figure 1E).

In order to test the scope of AUX1 and ABP1-related influences over auxin-dependent physiological processes at the plasma membrane, we applied an established assay for testing the effects of synthetic auxin analogs 1-NAA on endocytosis (Figures 2, 3). After exposure to BFA (a reversible inhibitor of exocytosis) treatments, PIN1 and PIN2 (markers of plasma membrane protein internalization) accumulated in large intracellular compartments in wild-type seedlings (Figures 2E,M, 3E,M), indicating endocytic internalization of PIN1 and PIN2. Pretreatment with 10 μM 1-NAA significantly inhibited BFA-induced PIN1 and PIN2 internalization (Figures 2H,I,L,O,R, 3H,I,L,O,R). As AUX1 was involved in rapid membrane depolarization, we next tested the effects of 1-NAA in aux1 and found that 1-NAA decreased BFA-induced PIN internalization to a similar extent as was observed in wild-type roots (Figures 2O,P,R, 3O,P,R). We therefore conclude that the inhibitory effect of 1-NAA on PIN internalization is not mediated by AUX1. After analysis of the 1-NAA-induced PIN internalization response in abp1-c1 and abp1-TD null mutant alleles, we observed similar degrees of PIN internalization in roots of both genotypes and wild type (Figures 2J–L, 3J–L), showing that BFA-induced PIN internalization is also independent of ABP1.

It has long been hypothesized that ABP1 is an extracellular auxin receptor, which mediates electric responses at the plasma membrane (Barbier-Brygoo et al., 1989; Leblanc et al., 1999). However, this function has never been directly shown. Characterization of two new lines of abp1 mutants has recently downgraded the significance of ABP1 for plant growth and development. Despite this, the possibility that ABP1 is involved in electric responses at the plasma membrane has never been ruled out. Using abp1 loss-of function lines, we can now do just this, by showing that ABP1 is not involved in auxin-dependent plasma membrane depolarization (Figures 1C,D).

Auxin-induced depolarization is dependent on auxin transport as auxin-dependent membrane depolarization is reduced in both influx and efflux carrier mutants, aux1 and pin2 (Figures 1C–E). The impact of the influx carrier was much higher, as depolarization was almost completely absent in aux1 mutants compared to a reduction of 20% in the pin2 mutant when compared to wild type. Both auxin transporters contribute to auxin accumulation in epidermis cells; the mechanism of action is through auxin-induced cytoplasmic Ca²⁺ transient concentration maxima, which require SCF^TIR1/AFB-based auxin signaling (Dindas et al., 2018).

The depolarization induced by IAA is likely to be auxin specific and physiologically relevant. This conclusion is supported by the correlation between the activity of auxins on membrane depolarization and their activity on inhibition of root growth: the most potent auxin IAA with respect to membrane depolarization caused the strongest inhibition of root growth (Rahman et al., 2007). Second, the dose responses with respect to membrane depolarization and root growth are similar: the minimal concentration sufficient to affect both processes is about 0.1 μM with saturation of response for both processes at 10 μM.

Endocytosis was not affected in aux1 (Figures 2, 3), indicating that its regulation is independent of the auxin-related regulation of plasma membrane potential. The hypothesis that an auxin-ABP1 interaction initiates a signal, which regulates endocytosis, is based on an analysis of conditional immune-modulation and antisense abp1 knockdown lines (Robert et al., 2010). However, the phenotypes of these lines have been attributed to off-target effects (Dai et al., 2015). The re-evaluation of this effect on qualified null abp1 allele lines was therefore necessary. Our measurements show that ABP1 does not regulate endocytosis (Figures 2, 3).

Both membrane depolarization and the inhibition of endocytosis are rapid auxin-induced responses. However, pharmacological and genetic experiments have shown that they are independent. The conclusion can be made as 1) auxin-induced membrane depolarization has a higher sensitivity to IAA than 1-NAA, whereas endocytosis is more sensitive to 1-NAA than to IAA (Paciorek et al., 2005); 2) membrane depolarization and the inhibition of endocytosis are also different in their speed of response: after the application of auxin, the plasma membrane depolarizes instantaneously, but endocytosis is only halted after 5 min (Dindas et al., 2018; Robert et al., 2010); and 3) auxin-induced depolarization was strongly reduced in the aux1 mutant, whereas the inhibition of endocytosis was unaffected. It is important to mention that to test the dependence of endocytosis inhibition by auxin from AUX1, we used 1-NAA because this is one of the most active auxins at inhibiting endocytosis (Paciorek et al., 2005; Robert et al., 2010). However, 1-NAA is able to diffuse through membranes in the absence of the AUX1 transporter (Yamamoto and Yamamoto, 1998); thus, we expected that 1-NAA-induced inhibition of endocytosis would be independent of AUX1 function. In contrast to 1-NAA, the natural auxin IAA without the addition of antioxidants has very low activity in terms of inhibition of endocytosis in the wild type, as the minimal concentration sufficient to inhibit endocytosis is about 1,000 times higher than the IAA concentration sufficient to modulate membrane depolarization (100 and 0.1 μM for inhibition of endocytosis and membrane depolarization, respectively) (Paciorek et al., 2005 and our data). This, in itself, indicates different mechanisms of regulation of endocytosis and membrane depolarization by IAA.

The data reported here clarify an important and intensively discussed point regarding the mechanisms of a plant’s non-transcriptional responses to auxin. We show that the long observed but poorly understood IAA-induced plasma membrane depolarization does not depend on ABP1 but instead is due to the co-transport of H⁺ and IAA into the cell by plasma membrane localized AUX1 rather than ABP1. Furthermore, our results show that neither AUX1 nor ABP1 plays roles in mediating endocytosis, another non-transcriptional auxin response. Given that ABP1 seems not to fulfill the functions, it was assigned in auxin action initially, and future studies need to find the real function of ER auxin-binding protein in plant growth and development.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.