Materials

The lipids POPC, β-sitosterol and stigmasterol were supplied from Avanti Polar Lipids (USA). Campesterol and 4-kDa fluorescein isothiocyanate (FITC)–dextran (FD) were purchased from Sigma-Aldrich Inc. (USA) The GUV membrane stain Lissamine™ Rhodamine B 1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt (rhodamine-DHPE), Sytox Green nucleic acid stain and the fluorescent probes Alexa Fluor 488 NHS Ester (succinimidyl ester) and Alexa Fluor 647 NHS Ester were purchased from Thermo Fischer Scientific (USA). Dimethyl sulphoxide (DMSO) was from Merck KGaA (Germany)and 2-(N-morpholino)ethanesulfonic acid (MES) from Sigma-Aldrich Inc. (USA). The protoplasting enzymes Cellulase Onozuka R-10 and 0.1% Macerozyme R-10 were from Duchefa (Netherlands) and D-mannitol from Sigma-Aldrich Inc. (USA).

Expression and Purification of NLP Pya

The NLPPya protein lacking its secretory signal sequence was overexpressed and purified as previously described (Luberacki et al. 2008; Lenarčič et al. 2017). Briefly, Escherichia coli BL21 (DE3) cells were transformed with the expression vector pET-21c containing NLPPya sequence fused to the hexahistidine tag at the C-terminus. The cells were grown to an OD600 of 0.8 and induced with 0.5 mM IPTG overnight at 20 °C. The sedimented cells were resuspended in 8 M urea, 0.1 M NaH2PO4, 0.01 M Tris–HCl pH 8.0 and disrupted by sonication with 1-s pulses alternating with 2-s pauses for 30 min. After centrifugation at 50,000 × g for 1 h, the protein was purified from the bacterial supernatant by nickel-chelate chromatography (Ni-nitrilotriacetic acid (NTA) Superflow column, Qiagen, Germany) under denaturing conditions. The column-bound protein was washed three times with 0.1 M NaH2PO4, 0.02 M imidazole, 0.01 M Tris–HCl pH 6.3 with decreasing urea concentrations of 2, 0.5 and 0.1 M. Following the final wash step with 0.05 M Na2HPO4/NaH2PO4 at pH 8.0, 0.3 M NaCl and 0.3 M imidazole, it was eluted with 6 M guanidinium hydrochloride and 0.2 M acetic acid. The protein was then dialyzed in four steps against distilled water and finally against 20 mM MES pH 5.8 and 150 mM NaCl to ensure refolding. The final yield of soluble NLPPya was high, ~ 20 mg/L. The pure monomeric fraction of NLPPya was obtained by gel filtration on a HiLoad Superdex 75 26/600 column (GE Healthcare, UK), equilibrated in protein buffer. The purity of NLPPya was verified by SDS–polyacrylamide gel electrophoresis (SDS-PAGE), its ability to bind GIPCs by a liposome sedimentation assay and its toxicity by infiltration into tobacco leaves as previously described (Pirc et al. 2022) (data not shown).

Labelling of NLP Pya with Alexa Fluor Dyes

The purified NLPPya was labelled with Alexa Fluor 488 (A488) or Alexa Fluor 647 (A647) NHS esters according to the manufacturer's instructions. First, 0.5 mL of 1 M sodium bicarbonate buffer, pH 8.3, was added to 2 mL of NLPPya at a concentration of 3.2 mg/mL to achieve a slightly more basic pH of the reaction buffer. Alexa Fluor 488 and Alexa Fluor 647 were freshly prepared in DMSO at a concentration of 15.5 mM immediately before starting the reaction. 50 µL of the reactive dye solution was added dropwise to the protein while stirring. The reaction was incubated in the dark for 1 h at 22 °C with constant stirring. The ratio of dye to protein in the reaction was 3:1 (mol:mol). The conjugate was separated from the unreacted dye by gel filtration on PD-10 desalting columns packed with Sephadex G-25 resin (Cytiva, USA). The concentration of the labelled protein and the degree of labelling were determined spectrophotometrically from the absorbance at 280 nm, and at 495 nm for Alexa Fluor 488 or at 651 nm for Alexa Fluor 647, using molar absorption coefficients of 48 025 M−1 cm−1 for the NLPPya, and 71 000 M−1 cm−1 for Alexa Fluor 488 or 239 000 M−1 cm−1 for Alexa Fluor 647. A correction factors of 0.11 and 0.03 were used for the absorbance of Alexa Fluor 488 and Alexa Fluor 647, respectively, at 280 nm. The degree of labelling was determined to be ~ 70%. for NLPPya-A488 and ~ 60% for NLPPya-A647. The homogeneity of the labelled proteins was verified by SDS-PAGE followed by ProBlue Safe (GiottoBiotech, Italy) staining.

Infiltration Assay

The infiltration assay was performed on tobacco plants (N. tabacum “White Burley”). 50 µL of proteins, diluted in deionized distilled water to a final concentration of 500 nM, were infiltrated abaxially into tobacco leaves using a syringe without a needle. The formation of lesions was analysed 2 days post-infiltration. The experiment was repeated three times with three leaves from two different plants.

Preparation of Glycosylinositol Phosphorylceramides

GIPCs were extracted and purified as previously described (Buré et al. 2011; Lenarčič et al. 2017). In brief, tobacco (Nicotiana tabacum) leaves were blended with cold 0.1 N aqueous acetic acid and filtered through miracloth. The slurry was washed, until green filtrate became colourless and was then extracted with hot 70% ethanol/0.1 N HCl overnight. The pellet was washed with cold acetone and diethyl ether to yield a whitish precipitate, which was dissolved in tetrahydrofuran (THF):methanol:water (4:4:1; v:v:v) containing 0.1% formic acid. Dried precipitate was submitted to a butan-1-ol:water (1:1; v:v) phase partition. The upper, GIPC-containing butanol phase was dried and the residue was dissolved in THF:methanol:water (4:4:1; v:v:v) containing 0.1% formic acid. GIPC was characterized by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) (Lenarčič et al. 2017) and their mass was estimated from dry weight.

Giant Unilamellar Vesicles Preparation by Electroformation

GUVs were prepared with a modified electroformation protocol (Aden et al. 2021) on a commercially available vesicle preparation station (Vesicle Prep Pro station, Nanion Technologies, Germany). Lipid mixtures of POPC alone, POPC with GIPCs 1:1.2 (n:n) or POPC with GIPCs and 30% plant sterols 1:3.7:2 (n:n:n) were dissolved in chloroform:methanol 9:1 (v:v) to a final concentration of 2.5 mM. The fluorescent probe rhodamine-DHPE was added to the lipid mixtures at the final lipid concentration of 0.5% for membrane labelling. The lipid solution (10 µL) was placed on a commercially available glass slide coated with conductive indium tin oxide and dried under reduced pressure for at least 90 min. The lipid-covered slide was then placed onto the vesicle preparation station and a rubber O-ring coated with silicone grease was placed around the lipids. The lipid film was then rehydrated in 750 µL of ultrapure water preheated to 55 °C and covered with another conductive slide. Electroformation was carried out inside the station with an amplitude of 3.5 V and a frequency of 10 Hz, which was passed over the slides at 65 °C for 165 min. This was continued for a further 75 min at 22 °C and 0.5 V amplitude. Once the protocol was completed, the solution with vesicles was carefully removed from the slide with a pipette and slowly deposited on the wall of the conical Falcon tube, where GUVs spontaneously settled to the bottom within ~ 4 days.

Isolation of Protoplasts

First, the washing solution was prepared by dissolving saccharose and macro-salt mixture Nitsch in ultrapure water at a final concentration of 0.3 M and 0.3 mg/mL, respectively. After autoclaving and cooling, MES, pH 5.6, was added to a final concentration of 5 mM. The protoplasting enzyme solution was prepared by dissolving 0.5% Cellulase Onozuka R-10 (Duchefa), 0.1% Macerozyme R-10 (Duchefa), and 0.1 M D-mannitol in the washing solution and sterilized using a Minisart cellulose acetate syringe filter (Sartorius). The protoplasting enzyme solution was freshly prepared for each use. Healthy tobacco leaves (1.5 g) approximately 10 cm in length were cut into strips less than 1 mm wide perpendicular to the midrib using a sterile scalpel. The leaf strips were placed in an aluminium foil-wrapped Erlenmeyer flask containing 50 mL of protoplasting enzyme solution. The flask was placed in a shaker overnight, where it was gently agitated at 50 revolutions per minute (rpm) and 20 °C. After 16–20 h of incubation, the flask was transferred to a laminarium and the leaf debris was gently mixed with the glass rod for 3–5 min to facilitate the release of protoplasts until the solution turned dark green in colour. The digestion mixture was filtered through the Cell strainer (Falcon) with a pore size of 70 µm, placed on a sterile glass funnel and collected in 50-mL Falcon conical polystyrene tube. The remaining leaf debris in the Erlenmeyer flask was washed twice with 4 mL of washing solution, mixed with the glass rod and added to the original protoplast suspension through the Cell strainer. The filtered protoplast solution was centrifuged at 100×g, 20 °C for 14 min in a fixed bucket rotor. The acceleration and deceleration of the centrifuge were reduced to the minimum possible option (level 1, Sorvall Lynx 4000 centrifuge, ThermoScientific, ZDA). Approximately 10 min after the end of centrifugation, a green circle of floating, intact protoplasts appeared on the surface of the solution, while debris had settled to the bottom of the tube. The floating protoplasts were transferred to a fresh 50-mL Falcon tube containing 4 mL wash solution using a wide tip (cut to approximately 1.5 mm). This was repeated two more times. After the last rinse, the protoplasts were carefully collected and transferred to a 2-mL tube. Protoplasts were stored in the dark at 4 °C and used within two days.

Confocal Laser Scanning Microscopy

All studies with GUVs were performed on an inverted DMI 6000 CS Leica TCS SP5 laser scanning microscope (Leica Microsystems GmbH, Germany) with a 63 × oil immersion objective (numerical aperture = 1.4). Microscope slides were coated with bovine serum albumin (BSA) to prevent bursting of the GUVs upon contact with glass. BSA solution in ultrapure water (10 mg/mL) was spread on the slides, allowed to stand for 5 min, then rinsed with ultrapure water and dried under the air stream. Aluminium foil squares of 0.5 × 0.5 mm were used as spacers between the slide and the cover glass (Aden et al. 2021). Leica LAS AF software (Leica Microsystems GmbH, Germany) was used for image acquisition.

For the NLPPya permeabilization studies, freshly prepared GUV suspension of POPC:GIPC 1:1.2 (n:n) was mixed with 4-kDA FD and (i) 500 nM NLPPya and incubated for different time intervals (from 15 min to 2 h) or (ii) with different concentrations of NLPPya but fixed duration of incubation (30 min). Since GUVs were prepared, stored and imaged in ultrapure water, the later was also used to highly dilute NLPPya, originally in the protein buffer (20 mM MES, pH 5.8, 150 mM NaCl). Content of the protein buffer in the final mixture was low (0.05 mM MES, 0.4 mM NaCl) to prevent changes in GUVs due to osmotic shock. 4-kDa FD was always freshly prepared in ultrapure water and completely dissolved by ultrasonication. The final concentration in the mixture was 1 mg/mL. The mixtures were incubated for 30 min in the dark at 22 °C, followed by imaging of samples on confocal microscope. Rhodamine-containing GUV membranes were excited with 543-nm HeNe laser and emission was detected between 590 and 650 nm. FD 4-kDa was excited with 488-nm argon laser and emission was detected between 500 and 535 nm. As a negative control, protein buffer alone, diluted as NLPPya in ultrapure water, was added to GUVs instead of the protein. The permeabilisation of GUVs in the recorded images was analysed using ImageJ software (National Institute of Mental Health, Bethesda, USA). First, the percentage of permeabilization of individual vesicle was calculated by measuring the intensity of FD fluorescence in the centre of the vesicle and dividing it by the average value of FD fluorescence intensity of 5 different points outside the GUVs on the same image. Based on the percentage of permeabilization, GUVs were sorted into bins with a range of 10% and the relative frequency in each bin was calculated based on the number of all analysed GUVs. The relative frequency of GUVs that were at least 50% full was calculated by summing the relative frequencies in bins from 50 to 100%. Finally, the relative frequencies of GUVs that were at least 50% full were averaged from three independent experiments and plotted.

For the NLPPya binding studies, GUV suspension of POPC:GIPC 1:1.2 (n:n) or POPC:GIPC:plant sterols 1:3.7:2 (n:n) was mixed with different amounts of NLPPya labelled with Alexa Fluor 488 to reach final concentrations of 0.1, 0.25, 0.5, 1 or 2 µM. The mixtures were incubated for 10 min in the dark at 22 °C followed by imaging of samples on confocal microscope. Rhodamine in GUV membranes was imaged as described above. Alexa Fluor 488 was excited at 488 nm and emission was detected between 500 and 530 nm. As a negative control, NLPPya-A488 was added to GUVs from pure POPC lipids.

GUVs of POPC:GIPC 1:1.2 (n:n) were used for fluorescence resonance energy transfer (FRET) imaging on Leica TCS SP5 confocal microscope. We imaged three different samples, namely a donor-only (NLPPya-A488), an acceptor-only (NLPPya-A647), and the combination of both (equimolar mixture of NLPPya-A488 and NLPPya-A647). The same protein concentration (1 µM) was used in all three samples, whereby in the first two cases half of it was replaced by the unlabelled NLPPya. All three combinations were imaged in 4 separate channels, namely one for rhodamine in GUV membranes, which was excited at 543 nm and detected from 570 to 590 nm, second for A488, excited at 488 nm and detected from 500 to 530 nm, third for A647 excited at 633 nm and detected from 720 to 790 nm and the last for detection of FRET, where A488 was excited with 488-nm laser and emission of A647 detected from 720 to 790 nm.

Protoplasts were imaged using a Leica TCS SP5 laser scanning microscope with 20 × and 63 × oil immersion objectives (numerical apertures of 0.8 and 1.4, respectively). They were incubated for 30 min with 2 µM NLPPya and 1 µM Sytox Green to detect the pore-forming activity of NLPPya on natural membranes, or for 15 min with 2 µM NLPPya-A488 to observe membrane binding and morphological changes. Aluminium foil squares of 0.5 × 0.5 mm were used as spacers between the slide and the cover glass. Autofluorescence of chlorophyll was detected in the range of 650–730 nm and the emission of Alexa 488 or Sytox Green at 500–530 nm and 510–540 nm, respectively, all three being excited with an argon 488-nm laser line.