A Novel Phloem-Specific HVA22-Like Protein Facilitates Protein Movement via Plasmodesmata in Potato

Abstract

Subcellular targeting and functionality of plant sucrose transporters (SUTs) are affected at the post-translational level by protein-protein interactions. A systematic screening for SUT-interacting proteins identified a novel plasmodesma (PD)-localized membrane protein belonging to the HVA22 family of stress-induced ER proteins. It carries three transmembrane domains and three zinc-finger domains, interacts with all three sucrose transporters from potato, and co-localizes with PD callose and PD marker proteins. Detailed analyses of transgenic potato plants with decreased expression of this PD protein displaying phenotypic alterations regarding trichome length, leaf expansion, root length, flowering time and tuberization helped to determine its physiological function. These effects are partially graft-transmissible indicating the participation of phloem-mobile signals. Changes in the levels of callose in RNAi plants suggested effects on PD permeability. Co-infiltration experiments revealed enhanced mobility of sucrose transporter—GFP fusion proteins via plasmodesmata in the presence of the HVA22-like protein. The analysis of transgenic plants further suggests that HVA22 protein is a general regulator of PD permeability in potato. Taken together, the HVA22-like protein is an ER protein localized close to phloem plasmodesmata and enhances mobility of GFP proteins of different sizes.

Introduction

Potato sucrose transporters are phloem-specifically expressed (Kühn et al., 1997; Barker et al., 2000; Weise et al., 2000), and the dynamics of their targeting depend on their oligomeric state and/or interacting proteins (Garg et al., 2020; Garg and Kühn, 2022). StSUT1 is the main phloem loading sucrose transporter from potato (Riesmeier et al., 1994), whereas StSUT2 and StSUT4 are less abundant low affinity transporters (Barker et al., 2000; Weise et al., 2000). All three sucrose transporters from potato have been localized to phloem sieve elements (Kühn et al., 1997; Barker et al., 2000; Weise et al. , 2000), whereas transcription takes place in neighboring companion cells (Kühn et al., 1996). The role of plasmodesmata herein remains a still fascinating question. All three sucrose transporters were shown to interact with a protein disulfide isomerase, StPDI1, a component of the detergent resistant membrane fraction (DRM) supposed to be involved in the escort of DRM-proteins from the ER to the plasma membrane (Krügel et al., 2012a) or in scaffolding them in lipid-raft like microdomains. StPDI1 also interacts with a stress-inducible member of the HVA22 family, StHVA22m (Eggert et al., 2016).

The plant HVA22 gene family is homologous to YOP (YIP one partner)/DP (defective in polyposis)/REEP (receptor enhanced expression protein) proteins, which are ubiquitous in other eukaryotes. The name is derived from the first identified aleurone protein from barley (Hordeum vulgare) as an abscisic acid (ABA) and stress-induced gene. HVA22 proteins can be involved in programmed cell death and vesicular trafficking (Brands and Ho, 2002; Guo and Ho, 2008). Most of them carry three N-terminal transmembrane spanning domains which are supposed to function in maintenance of membrane curvature and assumed to play a reticulon-like role in the ER (Lee et al., 2013; Meng et al., 2022). YOP1, the yeast homologous protein to HVA22 is able to interact with diverse ER resident proteins involved in vesicular trafficking and invertase secretion in yeast (Brands and Ho, 2002; Hu et al., 2008). It was suggested that reticulons and YOP1 oligomers are found in tubular domains in the ER and are important for their ability to form tubules (Shibata et al., 2008). Several HVA22 proteins are predicted to be phosphorylated and N-glycosylated (Wai et al., 2022).

Many HVA22 genes carry stress-related cis-regulatory elements in their promoter regions and show stress- and ABA-induced expression patterns (Chen et al., 2002). In Arabidopsis, AtHVA22d-RNAi plants show enhanced autophagy, and HVA22 homologs are discussed to act as suppressors of autophagy both in plants and in yeast (Chen et al., 2009). In tomato 15 non-redundant HVA22 genes are described (Wai et al., 2022), whereas in potato 16 non-redundant genes exist. Only one member of this large gene family, SlHVA22k, carries three zinc finger motifs at its C-terminus: the orthologous protein to the tomato SlHVA22k is StHVA22m in potato (Supplementary Figure S1).

Here, we describe the identification of the potato StHVA22m protein as a new direct interaction partner of all three potato sucrose transporters. As described for StPDI1, the StHVA22m gene is induced by salt stress and drought stress (Eggert et al., 2016). Interaction between the sucrose transporters of the SUT family with StHVA22m affects subcellular localization and targeting. RNA interference (RNAi) knock-down of StHVA22m produced strong developmental phenotypes that were partially graft-transmissible. Interestingly, the StHVA22m protein localized preferentially to plasmodesmata (PD) and we found that it influences PD trafficking. StHVA22m overexpression increased, whereas its knock-down decreased PD permeability for proteins, including the sucrose transporters. Plasmodesmata are important cell-cell–connections allowing the transport of solutes and macromolecules from one cell to another. It was postulated that either the SUT mRNA or the SUT protein or both are transported via PD from companion cells into sieve elements (Kühn, 1997 #339). PD may also represent a selectivity filter responsible for the specificity of the phloem sap composition (Garg and Kühn, 2020; Matilla, 2023). The phloem-specific expression pattern of StHVA22m suggests it to be potentially involved in phloem loading and unloading, and to be a component of the funnel-shaped PD of the unloading phloem or of the pore-plasmodesma-units (PPU), that are delta-shaped one-sided branched plasmodesmata specifically found between sieve elements and companion cells of the phloem (van Bel, 2003). Its potential role in sucrose transporter targeting to the plasma membrane of phloem sieve elements is discussed.

Material & Methods

Plant Cultivation

Cultivating Solanum tuberosum

S. tuberosum L var. Desiree plantlets were cultivated on 2-MS-media under sterile conditions. They were placed into a growth room with 22 °C, LD conditions (16 h of light, 8 h darkness). To keep a stock-population of selected plants, stem cuttings were transferred into fresh medium every 4–6 weeks. For experimental work, the rooted 4–6-week-old cuttings were transplanted into type-T soil and then placed into an air-conditioned (24°C at day, 20 °C at night) phytochamber with constant LD conditions. To prevent any infections, they were also inspected and treated with biological pesticides regularly. Phenotypes, flowering time and tuberization were documented throughout the cultivation process. Leaf samples were taken from source leaves exposed to light. The time when samples were taken was kept consistent throughout the entire research process (11 am – 1 pm). Plants were grown either at moderate temperature (20–24°C) or at cold temperature (≤ 18°C).

Stress Experiments

The drought stress experiment was conducted with four independent transgenic RNAi lines, representing four biological replicates. Three plants of each transgenic line were exposed to water stress for 14 days, whereas three others were watered continuously. The phenotype of both sets of plants was documented every 3 days. After 14 days the stressed plants were re-watered and recovery was documented.

Cultivating Nicotiana benthamiana

Seeds were directly sown onto type-T soil. After germination, single, similarly sized seedlings were transferred into individual pots and grown in the greenhouse under LD conditions until they had developed 4–6 leaves and used for infiltration experiments.

Potato Transformation

Potato transformation was performed as described earlier (Rocha-Sosa et al., 1989) using Agrobacterium tumefaciens strain pGV2260 (Deblaere et al., 1985). Potato leaf discs were co-cultivated with Agrobacteria for 4 weeks in the dark. After one month, first calli became visible and first sprouts appeared that were transferred onto root-inducing medium in the light. After three months, small plants were amplified and transferred to soil and analyzed further.

GATEWAY Cloning

All inserts were cloned full length into the ENTRY vector pDONR207 using a 2-step procedure following the Invitrogen protocol for GATEWAY cloning using gene-specific att-primers as given in Table S1.

Generation of the RNAi Construct

The LANCET RNAi construct was generated by inserting a 200 bp SpeI-HincII fragment from the LANCET cDNA into the pUC-RNAi vector (Eggert et al., 2016), linearized with BglII, and cut with SpeI after blunting. Similarly, the isolated fragment was inserted in opposite orientation (after BamHI restriction, blunting and XbaI digest) and verified by sequencing. The resulting stem-loop cassette including the 200 bp intron of GA20oxidase 1 was removed by PstI digest, and transferred into the binary pBinAR vector linearized with SpfI digest. This construct was used for transformation of the Agrobacterium strain pGV2260 and subsequently used for potato transformation.

Generation of LANCET-YFP Overexpression Construct

For LANCET overexpression, the full length 1647 bp LANCET cDNA was inserted into the binary vector pK7YWG2.0 (Karimi et al., 2005) using GATEWAY technology. For subcellular localization, the potential interaction partners of LANCET were cloned in either pK7YWG2.0 or pH7RWG2.0 (Karimi et al., 2005) using GATEWAY technology. The abbreviations in the figure captions are pK7 and pH7, respectively. The abbreviations in the figure captions are pK7 and pH7, respectively.

BiFC

Bimolecular fluorescence complementation was performed as described previously by cloning the putative interaction partners into the GATEWAY-compatible vectors VYNE and VYCE according to Gehl et al. (Gehl et al., 2009) where YFP moieties are translationally fused to the C-terminus of candidate genes.

Isolating Total RNA from Plant Tissue

RNA isolation was performed using TRIsure^TM (Bioline) according to the manufacturer's protocol. Samples from S. tuberosum plants were ground in liquid nitrogen using an electrical pestle (Heidolph) that was cooled using liquid nitrogen. After homogenization 1 ml of TRIsure reagent per 50–100 mg of ground sample was added. After mixing thoroughly, the samples were incubated at room temperature (RT) for 5 min. 200 μl of chloroform was added for every ml TRIsure. Samples were shaken vigorously for 15 s until no more layers were visible. Next, they were incubated at room temperature for 2–3 min and centrifuged at 11,000 rpm for 10 min at 4 °C. The aqueous phase was transferred into a new 1.5 ml tube. 500 μl of isopropanol per ml TRIsure was added. The tubes were inverted a few times. After incubation at RT for 10 min the samples were centrifuged at 11,000 rpm, inverted a few times after adding ethanol and then centrifuged at 8,500 rpm for 5 min at 4 °C. Washing was repeated once with 1.5 ml of ethanol and 1–2 times with 1 ml. The pellet was air-dried for 10–15 min before being resuspended in 30–40 μl of RNAse-free, autoclaved water. To help the pellet resuspend, the tubes were incubated in a thermal block at 65 °C for 10 min. Finally, the samples were centrifuged for 10 min at 11,000 rpm at RT. Extract concentrations were measured by Nanodrop and RNA integrity was checked on a 1.2% agarose gel (0.5× TBE).

Reverse Transcription (cDNA Synthesis)

The isolated RNA was reverse transcribed into cDNA. Prior to reverse transcription, the RNA was prepared as follows: 2 μg of total RNA were mixed with 1 μl 10× reaction buffer (containing MgCl₂), 1 μl RNAse free DNAse and 0.3 μl ribonuclease inhibitor to prevent the RNA from degradation. Autoclaved water was added to the mixture up to 10 μl. The mixture was then incubated at 37 °C for 30 min. Next, 0.5 μl of 25 mM EDTA were added, and the mixture was incubated again at 65 °C for 10 min. The pre-treated RNA was then used in reverse transcription.

1 μg of the RNA was mixed with 0.5 μl oligo(dT) primers or appropriate stem-loop-primers, 2 μl of 5×reverse transcriptase reaction buffer, 1 μl 10 mM dNTPs and 0.5 μL (100 U) of the reverse transcriptase. Sterile water was added to the mixture up to a volume of 20 μl. The mixture was gently mixed and briefly centrifuged. Subsequently, the mixture was incubated at 42 °C for 60 min. To terminate the reaction, the mixture was heated up to 70 °C for 10 min. The produced cDNA was used in qPCR experiments as a template.

Real Time Quantitative PCR

Real time qPCR (RT-qPCR) analysis was performed using the ChamQ Universal SYBR^® qPCR Master Mix. The SYBR Green dye included in this Master Mix binds amplified DNA strands enabling quantification via fluorescence measurements. The amount of PCR product was measured after every cycle. The cycles where the increase was exponential and above a given threshold were used for calculating the relative transcript amount via the 2^−ΔΔCt-method. RT-qPCR was performed using the C1000™ Thermocycler and the CFX96™ Real-Time System (BioRad). Each reaction mixture contained ddH2O, the cDNA template, ChamQ Universal SYBR^® qPCR Master Mix as well as gene specific forward and reverse primers as also listed in Table S1. To ensure more accurate results for each sample, three biological replicates for each transgenic line and four biological replicates for WT plants and two technical replicates were pipetted (n = 6 for transgenic lines, n = 8 for WT plants) and transferred into 96-well PCR-plates and sealed using MicroAmp™ foils. Significant differences in expression levels were evaluated by student's t-test with p < 0.05 (*), p < 0.01 (**), p < 0.001 (***).

Infiltration and Transient Expression

Transient transformation of Nicotiana benthamiana was performed by infiltration with Agrobacterium tumefaciens pGV2260 (Deblaere et al., 1985; Hellens et al., 2000) or Agrobacterium tumefaciens EHA105 (Hellens et al., 2000).

Cells from the Agrobacterium tumefaciens strain pGV2260 were used to induce a transient transformation in young Nicotiana benthamiana plants by performing so called agroinfiltration. Previously transformed A. tumefaciens cells were plated on YEB medium containing the appropriate selective antibiotics. After 2–3 days of incubation at 29 °C single colonies were used to inoculate 3 ml of liquid YEB medium. Those cultures were incubated at 29 °C and 250 rpm overnight. The next day, 1–2 ml of the cultures were used to start a 50 ml A. tumefaciens culture, to which 10 μl acetosyringone (in DMSO) were added. Those cultures were incubated at 250 rpm overnight at 29 °C. The following day, the cells were harvested in 50 ml falcon tubes by 15 min of centrifugation at 3500 rpm. The supernatant was carefully discarded, the pellet was washed with ddH2O and centrifuged an additional time. Cells were diluted using infiltration buffer until an OD_600nm of approx. 0.5 was reached. These cell suspensions were incubated at RT for 2–3 h.

To infiltrate young N. benthamiana plants, first the leaves were punctured using a small needle. A. tumefaciens cells were introduced into the leaf tissue by placing a syringe (without a needle) on the abaxial side of the leaves. The process was repeated until 3–4 leaves where fully saturated with the infiltrate. Expression of the desired genes was checked 3–5 days after infiltration with a confocal microscope (Zeiss LSM 800 with airyscan). GFP and YFP excitation was performed at 488 nm, aniline blue excitation at 405 nm, RFP and mCherry excitation at 555 nm and detection was performed using suitable emission spectra and, if necessary, signal amplification with a pinhole size of 1 Airy unit.

Yeast-two-Hybrid Split Ubiquitin Assay

Yeast transformation was performed using the strain Saccharomyces cerevisiae L40ccU A (Mat a, His3Δ200, trp1–901, leu2–3,112, lys2: (lexAop)4–HIS3, ura3::(lexAop)8–lacZ, ADE::(lexAop)8–URA, GAL4, gal80, can1, cyh2) (Goehler et al., 2004). Yeast cells were transformed first with the bait construct. 3-Aminotriazol (3-AT) is a competitive inhibitor of the his3 gene product and, for the LANCET bait the specific 3-AT concentration sufficient to prevent autoactivation was determined to range between 35 and 40 mM.

Large-scale yeast transformation was performed according to (Krügel et al., 2012a).

A serial dilution assay was performed with single colonies grown on liquid minimal medium and adjusted to an OD_600nm of 1 (corresponding to 3 × 10⁷ cells/ml). Equal amounts of cells were dropped on solid medium with 2% glucose or 2% sucrose. Cells were grown for at least 3 days at 30°C.

Callose Determination

Callose determination was performed by a modification of the aniline blue fluorochrome method (Köhle et al., 1985). 300 mg of leaf material were cleared overnight (or 1 h) in EtOH at RT, dried on sterile filter paper, and homogenized in 1N NaOH. Samples were heated to 80°C for 30 min, centrifuged for 15 min at 10,000 rpm. 200 μl of the supernatant correspond to 20 mg FW.

200 μl extract were mixed with 400 μl of 0.1% (w/v) aniline blue in water (Merck, Darmstadt). After addition of 210 μl of 1 N HCl the color changes to deep blue, indicating neutral to acidic pH values. The final pH value was adjusted by addition of 590 μl 1 M glycine/NaOH buffer (pH 9.5) and the tubes were mixed vigorously. During the following incubation for 30 min at 50°C and further 30 min at RT, the aniline blue becomes almost completely decolorized. Fluorescence was measured using 200 μl per sample in a dark microtiter plate with 405 nm excitation, 512 nm emission. Calibration curves were established using a freshly prepared solution of the 1,3-β-glucan in 1 N NaOH (1 mg/ml in 1 N NaOH in stepwise 1:5 dilutions).

Scanning Electron Microscopy

Leaf samples were fixed for 2 h in 2.5% glutaraldehyde and 4.0% formaldehyde in 0.05 M sodium cacodylate buffer, pH 7.2. Subsequently, the samples were rinsed in the same buffer and post-fixed for 1 h at room temperature with 1.0% osmium tetroxide in 0.05 M sodium cacodylate buffer, pH 7.2. The post-fixed samples were dehydrated in an ascending acetone series (1 h each step). Afterwards, the samples were submitted to critical point drying using CO₂ (CPD 030; Leica, Liechtenstein). Dried samples were placed on stubs, sputtered with 20 nm gold (K550× Sputter Coater; Quorum, Laughton, UK) and then observed with a digital scanning electron microscope (SEM) (EVO 40; Zeiss, Germany; Quanta 450FEG FEI, Czech Republic).

Graftings

Graftings were performed according to (Kollmann, 1992) with two-week-old plants by inserting a V-shaped graft scion in an appropriate root stock and stabilization with tape. Regeneration during the following days was guaranteed by covering the grafted plants carefully with moistened transparent bags. In total 75 plants were grafted and evaluated regarding tuber yield and leaf phenotype.

PD Permeability Assay

The PD permeability was quantified in N. benthamiana using GFP fusion constructs of different sizes agroinfiltrated at low optical density (10⁻³ for Solanum tuberosum and 10⁻⁴ for Nicotiana benthamiana as described by Barr & Tilsner (Barr and Tilsner, 2023) with small modifications using the vector pCK205 (Chincinska et al., 2008) encoding the 27 kDa GFP protein alone, or StSUT1 cloned in pCK205 coding for a fusion protein with the calculated mass of 82 kDa. In parallel, the LANCET protein was co-infiltrated in the vector pH7RWG2.0 (Karimi et al., 2005) encoding the red fluorescent protein RFP. As a negative control, the GFP constructs were co-infiltrated with the empty vector pH7RWG2.0. Statistical evaluation of the size of cell clusters showing GFP fluorescence was done by Student's t-test.

Alternatively, a vector system described by Ohtsu et al. 2024 was used (Ohtsu et al., 2024). The vector pICH4723.GFP.NLS-dTomato encodes the 27 kDa GFP protein, the vector pICH4723.2xGFP.NLS-dTomato encodes a 54 kDa GFP Homodimer. The NSL-red fluorescent protein allows the detection of the transformed cell, whereas GFP fluorescence allows to trace fluorescence movement due to diffusion through plasmodesmata. A. tumefaciens (GV2260) carrying binary plasmids was infiltrated into N. benthamiana leaves at OD_600nm = 1.0 × 10⁻⁵ to generate single-cell transformation events for the mobility assay. Samples were imaged 72h post-infiltration (Ohtsu et al., 2024). Statistical evaluation of cell-to-cell mobility was performed using a Student's t-test.

RNA in Situ Hybridization

RNA in situ hybridization was performed as described previously (Seibert et al., 2020) using LANCET-specific sense and antisense probes.

Primers are listed in supplementary Table S1

Results

The StHVA22m Protein Interacts with all Three Potato Sucrose Transporters via its N-Terminal YOP1 Domain

The StHVA22m/BG598159 protein was first identified via tandem affinity purification (TAP) using StPDI1 as a bait (Eggert et al., 2016). StPDI1 plays an important role in salt and drought stress tolerance in potato (Krügel et al., 2012b). Tandem-affinity purified proteins that interact with StPDI1 were identified by tandem mass spectrometry LC-MS/MS and among the 6 identified proteins, the StHVA22m protein (EST BG598159) was the most abundant showing 10fold enrichment in the elution fraction compared to the negative control (that is mCherry alone fused to the TAP-tag). HVA22 proteins belong to a large gene family of 16 HVA22-like genes in Solanum tuberosum and 15 in Solanum lycopersicum (Wai et al., 2022). Only one of the 15 HVA22-like proteins from tomato carries three conserved zinc finger domains with putative nucleic acid binding capacity at the C-terminus (SlHVA22k; accession Solyc10g082040). The orthologue in potato encodes 548 amino acids (StHVA22m, PGSC0003DMG400029015). Two of the three zinc finger motifs belong to a conserved neuromodulin-N domain present in gap junction proteins (Figure 1A). Interestingly, the longer form of HVA22 proteins containing three zinc finger domains does not occur in Brassicaceae. A phylogenetic analysis of HVA22-like proteins with extended C-terminus is given in Supplementary Figure S1.

Figure 1.

A. The full-length LANCET/StHVA22m protein has 548 amino acids, a HVA22-like N-Terminal domain with three transmembrane spans, and three conserved zinc finger domains at the C-Terminus with putative nucleic acid binding capacity. two of the three zinc finger motifs belong to a conserved neuromodulin-N domain present in gap junction proteins. B. Yeast two hybrid split ubiquitin assay show yeast growth in the absence of leu and trp (left) or in the absence of his, ura, leu, trp (right).The full-length LANCET/StHVA22m protein is able to directly bind to all three sucrose transporter from potato, StSUT1, StSUT2 and StSUT4 in the yeast two hybrid split ubiquitin system. C. Yeast growth in the absence of his, ura, leu, trp. Truncation of the C-Terminal domain containing all three zinc finger motifs (LANCETΔC) still retains the capacity to bind to StPDI1. Thus, the N-Terminal HVA22-like YOP1 domain of the protein is sufficient for protein-protein interactions.

Interaction between StHVa22m and StPDI1 was already shown in yeast using the yeast two hybrid split ubiquitin system. The full length cDNA sequence of StHVA22m expressed in the Nub or the Cub vector and co-expressed with StPDI1 enabled yeast growth even under stringent conditions with 25 mM 3-AT (Eggert et al., 2016), an inhibitor of his3 autoactivation. Here, we show that StHVA22m is able to interact directly with all three sucrose transporters of potato (Figure 1B) even in the presence of 15 mM 3-AT.

In order to test which part of the StHVA22m protein is responsible for protein-protein-interactions with plant sucrose transporters, the C-terminal domain with the three zinc finger motifs was deleted and the resulting truncated protein now only containing the YOP1 domain named LANCETΔC. The interaction with all 4 tested proteins was still possible even in the presence of 15 mM 3-AT (Figure 1C, Figure S2A) showing that the YOP1 domain is responsible for protein-protein interactions.

Interestingly, the interaction strength between LANCETΔC and the three SUT proteins is sucrose-dependent: adding 2% of sucrose, but not glucose to the medium greatly decreases the interaction in terms of yeast growth, whereas that between LANCETΔC and StPDI1 or StPDI1 and StSUT1 is not affected by sucrose supplementation (Supplementary Figure S2B).

StHVA22m Expression Pattern

StHVA22m expression is ubiquitous and ESTs are available from root, leaf, fruit and developing tuber tissue. The eFP potato browser suggests main expression in roots and fruits (https://bar.utoronto.ca/efp2/). qPCR analysis confirmed ubiquitous expression in all organs tested, and revealed highest expression in anthers and ovaries of mature flowers (Figure 2A). RNA in situ hybridization reveals LANCET/StHVA22m expression in anthers and gynoecium of young flowers (Figure 2 B,D), the vascular tissue of leaves (Figure 2C), and the stem, in what seems to be phloem companion cells (Figure 2E). LANCET expression is induced by drought and salt stress (Eggert et al., 2016) and expression decreases under low temperatures (4°C, Figure S9B, see below).

Figure 2.

A. qPCR analysis of tissue-specificity of LANCET/StHVA22m expression in potato plants. TEF was used as a reference gene. Three biological and two technical replicates (n = 6) were quantified for each primer set. B.-E. in situ hybridization using the LANCET antisense probe on various potato tissues. B. LANCET expression is detectable in anthers of flower buds (arrows). C. LANCET expression in a leaf cross-section reveals main expression in the vascular tissue. D. LANCET expression in the gynoecium of flower buds. E. phloem sieve elements showing LANCET expression in longitudinal stem sections. visualization occurred with DIG-labeled RNA probes. LANCET sense probes were used as negative control. scale bar corresponds to 100 µm.

Inhibition of StHVA22m expression Significantly Affects Growth Under low Temperature Conditions

An RNA interference (RNAi) approach was used for specific down-regulation of StHVA22m expression in transgenic potato plants. If grown under long day conditions at moderate temperature (24°C), the StHVA22m-RNAi plants do not show significant changes under standard conditions (Figure 3A). If shifted to cooler temperature (18°C), the newly developing leaves appear lancet-shaped with reduced leaf lamina expansion (Figure 3B). Since RNAi plants developed a lancet-shaped leaf phenotype, we named the protein LANCET in order to discriminate better from other HVA22-like proteins.

Figure 3.

The leaf phenotype of LANCET-RNAi plants is temperature-dependent. A. Leaves of LANCET-RNAi plants grown under moderate temperatures (20–24°C) are normally shaped or partially fused (arrow). B. the leaf phenotype of LANCET-RNAi plants under low temperature (<18°C) is lancet-shaped. the severity of the leaf phenotype correlates with the degree of inhibition of LANCET expression (Fig. 5A).

Not only the leaf phenotype is affected by cool temperature, but also flowering and tuberization is significantly changed: whereas tuber yield of LANCET-RNAi plants is significantly increased under SDs (Figure 4B) (although the assimilating leaf area is reduced under cold temperature), the flowering time is delayed (Figure 4E,F). The phenotype seems to be correlated more or less with the degree of LANCET-inhibition (Figure 3, 4A). Transgenic lines no #11 and #5.2 showing strong LANCET down regulation do not flower at all under LDs (Figure 4E,F).

Figure 4.

A. Real time qPCR analysis of LANCET expression in LANCET-RNAi plants. Source leaves from plants grown at moderate temperature under LD conditions were sampled. UBI was used as a reference. B. Tuber yield of LANCET-RNAi plants grown in a phytochamber under SD conditions at moderate temperature (with p < 0.05 *; p < 0.001 ***). C. the root length of LANCET-RNAi plants grown under sterile in vitro conditions is significantly reduced (with p < 0.001). D. the length of leaf trichomes of LANCET-RNAi plants grown under cold temperature in the phytochamber is doubled or even tripled compared to WT trichomes (stereo microscope with 4fold magnification). E. Flowering behavior of potato WT and LANCET –RNAi plants grown at moderate temperature under LD (16 h light/ 8 h dark) conditions in the green house. F. Flowering of potato WT and LANCET-RNAi plants grown at cold temperature under LD conditions in the green house. Note that flowering of WT potato plants under the same light conditions is strongly temperature-dependent. flowering is delayed in LANCET-RNAi plants under LD conditions regardless of the temperature. RNAi Lines #5.2 and #11 do not flower at all.

More detailed analysis of phenotypical modifications revealed elongated trichomes (2–3fold of the length of WT trichomes) under cold conditions (Figure 4D, 5A), a significant shortening of the roots if plants are grown on phyto-agar at moderate temperature (Figure 4C), increased leaf thickness (Figure S4A), and earlier tuber sprouting suggesting a shorter tuber dormancy period. Scanning electron microscopy revealed that not only the length of leaf trichomes is elongated at the abaxial as well as at the adaxial leaf epidermis (Figure 5A, S4B), but also the length of epidermis cells appears elongated two- to three-fold (Figure 5B).

Figure 5.

Scanning electron microscopic representation of upper and lower epidermis of LANCET-RNAi plants after fixation, critical point drying and gold sputtering. Note that the length of trichomes (A) and epidermis cells (B) on the abaxial as well as on the adaxial leaf side are more than doubled or even tripled in LANCET-RNAi plants compared to potato WT leaves. Scale bars represent 100 µm in A and 20 µm in B.

Interestingly, leaf cross-sections show the leaf morphology of LANCET-RNAi plants to be disturbed as well (Figure S5A) and this effect is even more pronounced under cold temperatures (Figure S5A). Whereas the potato WT source leaf is clearly organized with palisade parenchyma and spongy parenchyma, the morphological structure of LANCET-RNAi leaves is severely disordered (without clear palisade parenchyma organization), and the epidermis cells appeared to be prolonged (Figure S5A) as already seen in SEM images (Figure 5B). Surprisingly, whereas chloroplasts in WT palisade cells are arranged vertically, the chloroplasts in LANCET-RNAi plants are also found horizontally (Figure S5A, arrows), resembling leaves that were kept under low light. This chloroplast re-arrangement under low light is described to be blue-light dependent and triggered by phototropin photoreceptors (Suetsugu and Wada, 2007). Transmission electron microcopy was used to determine phloem ultrastructure and sieve element area, but no obvious differences were observed (Figure S5B).

LANCET is Associated to Plasmodesmata

Interaction between StHVA22m and all three sucrose transporters as well as with StPDI1 was not only confirmed in yeast, but also in planta using bimolecular fluorescence complementation (BiFC) in transient expression experiments in Nicotiana benthamiana leaves (Figure 6A, B). LANCET was also shown to self-interact in BiFC experiments (Figure 6B). Interestingly, LANCET interacts with StPDI1 in the ER (Figure 6C), whereas homotypic interactions and interactions with SUTs were observed at the cell periphery (Figure 6A) in a punctate manner. The different subcellular localizations suggest that the localization of the LANCET protein depends on the respective interaction partner.

Figure 6.

Confirmation of protein-protein interaction with the LANCET protein by bimolecular fluorescence complementation. YFP fluorescence is shown in green. A. LANCET interacts with StSUT1 at the cell periphery. Overlay picture with the transmission channel is given. B. LANCET forms homodimers in a punctate manner, most likely at plasmodesmata. C. LANCET interacts with StPDI1 in the ER; most likely at plasmodesmata. D. the LANCET protein fused to RFP (red) co-Localizes with the plasmodesmal marker protein TMV-MP-GFP (green) expressed under control of the CaMV35S promoter. Co-localization is indicated by white circles. E. the truncated LANCET protein LANCETΔC fused to YFP (yellow) co-localizes with plasmodesmal callose stained with aniline blue (blue). Co-localization is indicated by white arrow heads. Scale bars represent 10 µm (except in E where it is 30 µm).

To investigate LANCET localization independent of its interaction partners, fluorescent protein fusions were transiently expressed in N. benthamiana. Full-length LANCET-RFP localized to punctate and enlarged structures at the cell periphery (Figure 6D) that partially co-localized with the Tobacco mosaic virus movement protein (MP-GFP), a PD marker, and resembled the localization of the LANCET-SUT1 BiFC complex (Figure 6A). The truncated LANCETΔC protein fused to YFP fluorescent protein also appeared in plasmodesmata-like structures and co-localized with plasmodesmal callose stained with aniline blue (Figure 6E). Thus, LANCET is a plasmodesmal protein and its YOP1 transmembrane domain is sufficient for PD localization.

LANCET RNAi Phenotypes are Partially Graft-Transmissible

Graft experiments were performed in order to test whether phenotypic modifications are due to phloem-mobile effectors (Figure 7). The LANCET-shaped leaf phenotype does not occur in WT scions grafted on RNAi stocks after grafting, but the extent of lamina restriction under low temperature in the LANCET-RNAi graft partner is alleviated when RNAi scions were grafted on WT stocks. This is consistent with phloem-mobile signaling, where the WT provides an upwards moving phloem-mobile signal for leaf lamina expansion (Figure 7A). Regarding tuberization, increased tuber biomass production is observed in transgenic LANCET-silenced root stocks grafted with WT scions, in a dose-dependent manner with highest yields for the most reduced LANCET expression (Figure 7B). In reciprocal grafts of WT root stocks with LANCET-silenced scions, a trend towards increased tuber yield was observable, but the differences where not statistically significant, and tuberization does not show dose dependency.

Figure 7.

A. Graft experiments between LANCET-RNAi and WT plants. One representative example is given. The LANCET-shaped leaf phenotype of LANCET-RNAi plants was alleviated in RNAi scions by grafts with WT stocks. B. Tuber yield of grafted plants. WT scions grafted on WT stocks were taken as control (left column). Left side: LANCET-RNAi Scions were grafted on WT stocks; right side: WT scions were grafted on LANCET-RNAi stocks. Tuber yield was increased in almost all combinations under SD conditions suggesting that the tuber inducing signal is phloem mobile from LANCET-RNAi plants to WT plants regardless of the graft orientation. C. Real time qPCR of LANCET and potential downstream targets. RNA samples were taken from source leaves of adult potato plants grown at 24°C under greenhouse conditions. Reference genes were TEF for LANCET, CUC3 and NAM1, and 5SrRNA for the quantification of miRNAs.

LANCET-Inhibition Does not Affect Drought Stress Resistance

LANCET and its interacting StPDI1 interaction partner are both inducible under drought and salt stress (Eggert et al., 2016). Inhibition of StPDI1 expression in transgenic potato plants leads to severe problems under drought or salt stress conditions, with plants producing significantly more reactive oxygen species (ROS) under stress, and never recover when returned to non-stress conditions (Krügel et al., 2012a; Eggert et al., 2016).

In contrast to the behavior of StPDI1-RNAi plants, LANCET does not seem to play a role under drought stress, although induced in expression under these conditions ((Eggert et al., 2016) Figure S8). LANCET-RNAi plants behave very similar to potato WT plants if watering was stopped for 14 days. After re-watering after 2 weeks, the transgenic plants recover as well as WT (Figure S8) suggesting a very different role under abiotic stresses than postulated for StPDI1.

In an experiment analyzing the memory effect of cold temperatures on the sugar accumulation of potato plants, we realized that LANCET expression is significantly down-regulated in WT potato in both primed (previous exposition to 4°C for 24 h) and non-primed (first time exposure) plants exposed to 4°C for 24 h (Figure S9). In parallel, the amount of soluble sugars increased in leaves of the cold-treated plants, but hexose sugars increased to a higher level in non-primed compared to primed plants (Figure S9C).

Detailed Real-Time qPCR Analysis of Phloem-mobile Signals

Flowering and tuberization in potato plants are known to involve phloem mobile signaling molecules such as SP6A, miR172 and miR156 (Hannapel, 2013; Bhogale et al., 2014). Therefore, the expression level of phloem mobile miRNAs and transcription factors involved in leaf development and leaf lamina shape was quantified by qPCR (Figure 7C). Whereas the age-dependent miR156 was significantly up-regulated under moderate temperature (23°C), its antagonist miR172 tends to be decreased in LANCET-RNAi plants. miR156 overexpression in potato is described to affect potato development, especially leaf morphology and trichome length and is known to be graft-transmissible (Bhogale et al., 2014). miR164 is also increased in concordance with decreased expression of its target genes, transcription factors of the cup- shaped cotyledons (CUC) family, CUC3 and no apical meristem NAM1 (Figure 7C). The expression of CUC/NAM transcription factors plays a role in leaf lamina shape (Bar and Ori, 2014). These differences in the expression of phloem-mobile miRNAs are one possible explanation for the graft-transmissible phenotype.

LANCET Positively Affects PD Permeability

The impact of LANCET on plasmodesmal permeability was tested first in Nicotiana benthamiana using GFP fusion proteins of different sizes combined with or without the overexpressed LANCET protein in low OD infiltration experiments according to Barr & Tilsner (Barr and Tilsner, 2023). Experiments were conducted with either free GFP (27 kDa) or StSUT1 fused to GFP (82 kDa) co-expressed with LANCET-RFP or free RFP (Figure 8A) and cell clusters were quantified (Figure 8B). For free GFP, the cluster size significantly increased in the presence of LANCET-RFP, compared to the control experiment without LANCET (empty pH7 vector). Even when the integral plasma membrane protein StSUT1 was used in this permeability assay, it showed increased mobility in the presence of the LANCET-RFP fusion protein (Figure 8A, B). It is worth to mention, that integral PM proteins were only rarely found to move via PD (Tilsner et al., 2011).

Figure 8.

Co-infiltration experiments according to Barr & Tilsner (2023) revealed that the LANCET-RFP protein increases plasmodesmal mobility of co-infiltrated GFP fusion proteins in N. benthamiana. A. Images were taken 4 days after infiltration. Single cells are fluorescent in the absence of LANCET (empty vector pH7, left pannels), and cell clusters are detectable in the presence of LANCET-RFP (right pannels). B. Cell cluster size in the presence or absence of the LANCET protein 3 days after infiltration. Note that the soluble GFP protein is 27 kDa (green bars), the StSUT1-GFP is 82 kDa in size (red bars). The GFP vector pCK205 (Chincinska et al. 2008) was co-infiltrated with LANCET in pH7RWG2.0 (Karimi et al. 2005) or with the empty pH7RWG2.0 (pH7). C. Cell cluster size of the vectors pICH4723.GFP.NLS-dTomato (blue bars) or 217pICH4723.2xGFP.NLS-dTomato (Ohtsu et al. 2024). Note that the 2× eGFP protein is 54 kDa in size. Student‘s t-test was performed with p < 0.01 (**) and p < 0.001 (***), n = 200–213 foci for each combination in B, n = 149–225 foci for each combination in C).

In order to discriminate between initially transformed and recipient cells, 2× eGFP (54 kDa) was also co-expressed with LANCET from a vector simultaneously expressing non-mobile RFP with nuclear localization signal (NLS)(Ohtsu et al., 2024) (Figure 8C). As for free GFP and StSUT1-GFP, the diffusion rate of 2× eGFP was significantly increased in the presence of LANCET (Figure 8C). Thus, the permeability of plasmodesmata was tested with fusion proteins with three different molecular sizes and both cytoplasmic and plasma membrane-associated, and found to be increased in the presence of LANCET. It is concluded that LANCET increases the plasmodesmal permeability for GFP fusion proteins ranging in size between 27 kDa and 82 kDa.

Transgenic potato plants expressing a LANCET-YFP fusion construct under control of the CaMV 35S promoter do not show any relevant phenotypical modification. LANCET-YFP overexpressing and LANCET-RNAi plants were tested regarding their plasmodesmal permeability. Plasmodesmal permeability of transgenic potato plants with reduced or increased LANCET expression was assayed by a modified low OD infiltration of free GFP (27 kDa) and StSUT1-GFP (82 kDa). The OD_600nm of infiltrated Agrobacteria suspension was adopted to 0.001. Single fluorescent cells were observed in LANCET-RNAi plants as well as in the potato WT plants (Figure 9 A,B), whereas cell clusters were observed only in LANCET-overexpressing lines (Figure 9 C). Since the efficiency of transient transformation of potato leaves by Agrobacterium-infiltration is very low, the evaluation of diffusion was done in a qualitative manner (cell clusters detectable or not), but not evaluated quantitatively.

Figure 9.

Qualitative determination of GFP diffusion in transgenic potato plants by low OD infiltration experiments according to Barr & Tilsner (2023) in transgenic potato plants. Agrobacteria carrying the vector pCK encoding GFP were infiltrated at OD_600nm = 0.001. Images were taken 4 days after infiltration. Scale bars are 50 m in length. Single cells are fluorescent in LANCET-RNAi #5.2 plants (A) and in Solanum tuberosum Désirée WT plants (B). Cell clusters of GFP can only be observed in LANCET-overexpressing line #12.2 (C). D. Callose deposition in source leaves of LANCET-RNAi plants (red bars) grown under greenhouse conditions at moderate temperature is significantly increased. Callose deposition in source leaves of LANCET-overexpressing plants (green bars) is rather decreased. Callose was quantified photometrically after aniline blue staining according to Aidemark et al. (Aidemark et al., 2009) with 4 replicates per transgenic line. RNAi plants and overexpressing plants were harvested separately with appropriate WT controls. Student's t-test indicate significant differences with p < 0.001 (***), p < 0.01 (**).

Callose Determination in LANCET-RNAi and -Overexpressing Potato Plants

Plasmodesmal permeability strongly depends on callose depositions at the orifices of plasmodesmata (Zavaliev et al., 2011; Park et al., 2019). In order to learn more about the molecular mechanisms of size exclusion limit (SEL) control in relation to the LANCET protein, the callose level was quantified in bulk tissues (Figure 9D). Callose is significantly increased in the leaves of LANCET-RNAi plants (Figure 9D), whereas in LANCET-overexpressing plants, the amount of callose was decreased. Apparently, LANCET regulates plasmodesmal opening directly or indirectly via changes of callose deposition.

ER-Retention Inhibits PD Movement of Fluorescent Proteins

In order to answer the question whether spread of GFP fluorescence was due to protein movement via plasmodesmata, or increased movement of the corresponding mRNAs several control experiments were performed.Deletion of the putative RNA-binding zinc finger motifs of the LANCET C-terminus resulted in a more or less toxic variant of the HVA22 -like protein and evaluation of the cell number of fluorescent clusters after low OD infiltration was not possible.

In a second approach, the ER-marker HDEL-mCherry was co-infiltrated with the LANCET protein (Figure 10). In this case, no cell clusters were observed and mCherry fluorescence was retained in single fluorescent cells both in the absence and in the presence of the LANCET protein (Figure 10). The experiments have been repeated and quantified with a HDEL-YFP construct with similar results (Figure 10C). Thus, mRNA movement is unlikely also in analogous experiments with free GFP, 2xGFP or SUT1-GFP and cell clusters are assumed to be due to protein movement via plasmodesmata.

Figure 10.

Low OD infiltration experiments according to Barr & Tilsner (2023) in N. benthamiana leaves. A. control experiments with free GFP (27 kDa) show larger cell clusters in the presence of LANCET than in its absence. B. No larger cell clusters could be observed when mCherry (red) was retained in the ER by HDEL retention signal even in the presence of LANCET. Only single cells could be observed. Images were taken three days after infiltration. Experiments were repeated several times independently with similar results. C. Quantification of another experiment using ER-retained HDEL-YFP co-infiltrated or not with LANCET fused to RFP. Significantly more GFP diffusion in neighboring cells was only observed for free GFP, but not for the ER-retained YFP molecule. Total number of single cells or cell clusters was n = 249, 227, 121 and 107 respectively.

LANCET Affects the Expression of PD-Associated beta-1,3-Endoglucanases and Callose Synthases

Callose turnover depends on the activity of callose synthases and callose degrading beta-1,3-glucanases (BGs). Callose deposition occurs not only close to plasmodesmata, but is also involved in sieve plate plugging in phloem sieve elements following injury, and during plant defense responses. In potato the family of beta-1,3-glucanases comprises as many as 62 members (van den Herik et al., 2024). They fall into three phylogenetic clades: alpha, beta and gamma. In Arabidopsis, all BGs that are PD-related are found in the α-clade according to Levy et al. (Levy et al., 2007).

In order to identify those BGs involved in plasmodesmal opening and closure, we selected potato BGs that are most closely related to the Arabidopsis AtPdBG1, AtPdBG2 or AtBG_ppap (putative plasmodesmal associated protein) homologs according to Levy et al. (Levy et al., 2007). The following BGs were selected for further analysis based on homology to those three PD-glucanases from Arabidopsis: SoltuDM02g020260.1 (BG1); SoltuDM11g002490.1 (BG2-like); SoltuDM04g003300.1 (BG3-like); SoltuDM08g030220.1 (BG11-like); SoltuDM12g0008540.1 (BG11 precursor); SoltuDM01g001360.1 (BG21). Interestingly, all 6 genes show modified expression in LANCET-RNAi and/or LANCET-overexpressing plants (Figure S10): 4 out of 6 BGs were down-regulated in LANCET-RNAi plants, whereas 4 out of 6 BGs show up-regulation in LANCET-overexpressing plants.

Callose synthesis is mediated by callose synthases and only 2 out of 16 are assumed to be PD-associated according to van den Herik et al. (van den Herik et al., 2024): CalS3 (SoltuDM07g023050) and CalS8 (SoltuDM01g001920). Callose synthase 8 is up-regulated in LANCET-RNAi plants (Figure S10A).

Modified expression of PD-localized BGs and Callose synthases thus may explain the differences in callose accumulation in the leaves of transgenic LANCET-plants.

Discussion

LANCET Expression Pattern and Localization

The LANCET protein belongs to the large gene family of HVA22 integral membrane proteins with three transmembrane spanning domains but has an unusual C-terminus with three zinc finger domains that are not present in other members of the HVA22-like proteins (Wai et al., 2022). This extended C-terminus makes it unusual and orthologous proteins are not found in Brassicaceae (see phylogenetic analysis in Figure S2). LANCET is the putative orthologue of SlHVA22k, a single gene in tomato (Wai et al., 2022). Interaction studies in planta and in yeast showed that the N-terminal conserved YOP1-domain of LANCET is sufficient to maintain the interaction between LANCET and sucrose transporters (Figure 1B).

The expression of LANCET is ubiquitous, with highest levels in mature flowers, anthers and ovaries and in the vasculature (Figure 2). High expression in sink organs of the plant suggest a potential role in phloem unloading.

LANCET expression is increased under salt and drought stress (Eggert et al., 2016) and reduced under cold temperatures (Figure 9B). This agrees with published data regarding SlHVA22k, the only homologue also carrying three zinc finger domains at the C-terminus, which is mainly expressed in green tomato fruits, and significantly induced by salt, drought, heat and ABA, and inhibited by cold treatment (Wai et al., 2022). A role in tomato fruit development and ripening for SlHVA22k was suggested. A similar expression pattern was observed for the potato LANCET gene.

The members of the HVA22 family are ER-localized proteins and co-localization experiments showed a PD-specific localization of LANCET if fused to a fluorescent protein (Figure 6). Therefore, a localization in the ER close to PD, and close to the ER-derived desmotubule is suggested, that connects ER of neighboring cells. LANCET has three predicted transmembrane spanning domains with the N-terminus predicted to be extracellular (or in case of ER localized proteins luminal) and the C-terminus including the three potentially RNA-binding zinc finger motifs facing the cytoplasm according to the ARAMEMNON plant membrane protein database (https://aramemnon.botanik.uni-koeln.de/seq_viewBlast.ep). Deletion experiments revealed that the protein interaction domain allowing interaction with all three sucrose transporters and StPDI1 resides within the hydrophobic N-terminus of the protein (Figure S2).

Co-infiltration studies revealed a positive impact of the LANCET protein on the PD permeability of cytoplasmic and PM but not ER-luminal proteins (Figure 8–10). This increase in PD permeability is coupled to reduced total callose content in these plants (Figure 9D). Conversely, RNAi knock-down of LANCET reduced PD permeability and increased callose. Both effects appear to be related to differential expression of PD-specific callose synthases and PD-specific β−1,3-glucanases (Figure S10).

Temperature-Dependent Leaf Phenotype of LANCET-RNAi Plants

Leaf development and leaf lamina expansion involves interaction of WUSCHEL-like WOX1 transcription factors, auxin gradients and transcription factors of the CUC family (Vandenbussche et al., 2009; Vandenbussche, 2021). The expression of these transcription factors and the miR164 that targets CUC transcription factors were deregulated when LANCET expression was reduced (Figure 7C).

Inhibition of LANCET expression leads to severe defects of leaf development under cold temperatures (<18°C) and reduces expression of CUC family transcription factors, concomitant with increased mature miR164. LANCET-RNAi plants develop a lancet-shaped leaf lamina under cold conditions (Figure 3B), and LANCET expression was shown to be temperature responsive (Figure S9B). LANCET seems to play a role for cold acclimation of the plant, but the mechanism is still not understood at the molecular level.

This lancet-shaped leaf phenotype is alleviated if RNAi scions are grafted on WT stocks suggesting an upwards moving signal. An increased tuber yield in a LANCET dose-dependent manner is observed when WT scions are grafted in RNAi stocks (Figure 7). Both observations suggest, that phloem-mobile miRNAs that are deregulated in LANCET-RNAi plants might potentially be involved in the characteristic phenotypes.

Real-time qPCR revealed that indeed phloem-mobile miRNAs involved in the induction of flowering and tuberization in potato are differentially expressed in the leaves of LANCET-RNAi plants (Figure 7C) potentially explaining the impact on flowering and tuberization. Increased levels of miR156 in LANCET-RNAi plants agree with the phenotype of potato plants overexpressing miR156 (Bhogale et al., 2014). Increased levels of miR156 are also explaining some of the leaf-specific observations regarding length of epidermis cells, trichomes and effects on tuberization (Bhogale et al., 2014). miR156, miR172 and miR164 are known to be involved in leaf transition and leaf modification in Arabidopsis (Yang et al., 2018) and they are phloem-mobile explaining the graft-transmissible phenotype. It cannot be excluded that reduced mobility of miR156 in the phloem is leading to up-regulation of miR156 expression for compensation effects in these plants.

CUC transcription factors are involved in cell proliferation during leaf modification and negatively regulated by SPL transcription factors (targets of miR156) and miR164 (Yang et al., 2018). miR156 is also described to be involved in abiotic stresses like drought stress in Arabidopsis and apple via the transcription factor SPL13 (Feng et al., 2023; Huang et al., 2024).

Bioinformatic prediction tools suggest that the LANCET gene by itself is a potential target of miR172 (https://www.zhaolab.org/psRNATarget/home), and miR172 expression in potato is increased by sucrose (Garg et al., 2021). The question remains whether LANCET expression is repressed by sugars. Under cold conditions, the sugar accumulation was up to 6-fold increased with a shift of the hexose to sucrose ratio in favor of hexoses (Figure S9C). This agrees with observations on cold-stored potato tubers (Zrenner et al., 1996).

The LANCET promoter contains at least 4 CArG motifs (responsible for FLC binding), as well as the sugar responsive SURE motif AATAGAAAA (at position −100). In Arabidopsis, FLC is a cold-responsive transcriptional repressor, and the presence of a FLC binding domain in the LANCET promoter might explain the temperature dependent expression of the LANCET gene.

The temperature-dependent phenotype of LANCET-RNAi plants suggests an important role for adaptation to cold temperatures. It is assumed that the LANCET protein in the desmotubule of PDs might be involved in the transport of proteins (SUTs, transcription factors) or RNAs (including miRNAs) or phytohormones via plasmodesmata into the phloem or out of the phloem, thereby regulating leaf development in a temperature-dependent manner.

The fact that PD-associated beta-1,3-endoglucanases show decreased expression in LANCET-RNAi plants might be the reason for increased callose accumulation in those plants (Figure 9D). The increased expression of callose synthase 8 might be indicative for the involvement of the LANCET protein in phloem unloading. In Arabidopsis, Callose synthase 8 (CalS8) acts at the interface between sieve element and phloem pole pericycle, where phloem unloading occurs (Ross-Elliott et al., 2017). And callose synthases, as well as callose-induced plasmodesmata closure seems to be temperature dependent in Arabidopsis, maize and tomato (Bilska and Sowinski, 2010; Cui and Lee, 2016; Liu et al., 2022a; Wu et al., 2023). Callose Synthase CalS8 is involved in callose deposition under abiotic stresses such as low temperature, wounding, H₂O₂, leading to reduced PD permeability (Cui and Lee, 2016). Since LANCET expression is reduced under cold conditions, a lack of callose synthase inhibition might be the reason for increased callose deposition and reduced permeability.

Positive Impact of LANCET on the Permeability of Plasmodesmata

A recent manuscript describes an Arabidopsis HVA22 protein, AtHVA22a, that is highly enriched in the PD proteome and involved in the propagation of the turnip mosaic virus (Xue et al., 2024). Nothing is known about phloem-specifically localized HVA22 proteins.

The determinants of the composition of the phloem sap and the specificity of sieve element proteins have been discussed extensively (Garg and Kühn, 2020; Sanden and Schulz, 2021; van Bel, 2021; Sanden and Schulz, 2022). The specificity and selectivity of plasmodesmal macromolecular transport is supported e.g., by the absence or under-representation of certain tRNAs such as Isoleucine tRNA Ile (Zhang et al., 2009). It is the question how the transport of macromolecules occurs, and which plasmodesmal proteins are involved herein.

Sucrose transporters (SUTs) of higher plants are tightly controlled at the transcriptional, translational and post-translational level (Liesche et al., 2011). Sucrose transporters are phloem-localized membrane carriers whose targeting to the plasma membrane still needs to be understood in detail. Systematic screening for SUT-interacting proteins revealed specific interactomes for each of the transporters with only small overlaps (Krügel et al., 2012b; Bitterlich et al., 2014). Only few of the interaction partners are able to interact with all three Solanum sucrose transporters: StPDI1 and LANCET, two promiscuous interaction partners that both are inducible by abiotic stresses (Krügel et al., 2012b; Eggert et al., 2016).

Sucrose transporters of Solanaceous species have been localized at the plasma membrane of mature phloem sieve elements. In tomato and potato, the SUT1 protein was clearly localized unequivocally to phloem sieve elements (Kühn et al., 1997; Reinders et al., 2002). In tobacco plants, a recent proteomic study confirmed specific enrichment of the NtSUT1 and the NtSUT4 protein in phloem sieve elements as well (Liu et al., 2022b).

The expression of SUT1 was shown to occur in phloem companion cells (Kühn et al., 1996) and the mRNA is preferentially localized in proximity of the branched plasmodesmata connecting sieve elements and companion cells (Kühn et al., 1997). There are indications, that the StSUT1 mRNA might be able to cross plasmodesmata as shown by micro-injection experiments (Xoconostle-Cazares et al., 1999). In many plant species, the SUT1 mRNA was found in the phloem sap (Ruiz-Medrano et al., 1999; Knop et al., 2001; Doering-Saad et al., 2002; Deeken et al., 2008) and is shown to be phloem-mobile in grafted plants and between host plants and parasites (He et al., 2008). Also sucrose transporter mRNAs from various plants species are phloem mobile macromolecules according to the PlaMoM database (https://hdl.handle.net/11858/00-001M-0000-002B-B2AF-6)(Guan et al., 2017).

Regarding sieve element specific proteins two main strategies are discussed: either they are continuously synthesized in companion cells and transported into the mature sieve elements via plasmodesmata, or alternatively translated in immature sieve elements before differentiation and enucleation without undergoing continuous turnover (Sanden and Schulz, 2021). StSUT1 is a short-lived protein (Kühn et al., 1997), and translation in companion cells is therefore assumed.

Phloem sieve elements are certainly able to perform protein degradation (Ostendorp et al., 2017), but due to the lack of ribosomes, translation is very unlikely there (Zhang et al., 2009). It is therefore suggested that the translation of the StSUT1 mRNA takes place in companion cells, and the mature protein is transported via plasmodesmata from companion cells to the sieve elements (Garg and Kühn, 2020). Given that ectopic LANCET overexpression in N. benthamiana leaf epidermis permitted cell-to-cell transport of SUT1-GFP, it cannot be excluded that the LANCET protein is involved in delivery of SUTs to sieve elements.

Future Perspectives

A complex regulatory network can be imagined involving several feedback loops where many environmental factors play a role including low temperature, age, sugar status, and a hypothetical model explaining effectors and effects is shown in Figure S8. Further attempts are required to confirm these hypotheses.

Supplemental Material

sj-pptx-1-ctc-10.1177_25152564261418821 - Supplemental material for A Novel Phloem-Specific HVA22-Like Protein Facilitates Protein Movement via Plasmodesmata in Potato

Supplemental material, sj-pptx-1-ctc-10.1177_25152564261418821 for A Novel Phloem-Specific HVA22-Like Protein Facilitates Protein Movement via Plasmodesmata in Potato by Lara Lansky, Fabienne Drescher, Katarina Sugic, Karoline Diesing, Aleksandra Hackel, Vanessa Wahl, Jens Tilsner and Christina Kühn in Contact

Supplemental Material

sj-docx-2-ctc-10.1177_25152564261418821 - Supplemental material for A Novel Phloem-Specific HVA22-Like Protein Facilitates Protein Movement via Plasmodesmata in Potato

Supplemental material, sj-docx-2-ctc-10.1177_25152564261418821 for A Novel Phloem-Specific HVA22-Like Protein Facilitates Protein Movement via Plasmodesmata in Potato by Lara Lansky, Fabienne Drescher, Katarina Sugic, Karoline Diesing, Aleksandra Hackel, Vanessa Wahl, Jens Tilsner and Christina Kühn in Contact