Internalization of miPEP165a into Arabidopsis Roots Depends on Both Passive Diffusion and Endocytosis-Associated Processes.

MiPEPs are short natural peptides encoded by microRNAs in plants. Exogenous application of miPEPs increases the expression of their corresponding miRNA and, consequently, induces consistent phenotypical changes. Therefore, miPEPs carry huge potential in agronomy as gene regulators that do not require genome manipulation. However, to this end, it is necessary to know their mode of action, including where they act and how they enter the plants. Here, after analyzing the effect of Arabidopsis thaliana miPEP165a on root and aerial part development, we followed the internalization of fluorescent-labelled miPEP165a into roots and compared its uptake into endocytosis-altered mutants to that observed in wild-type plants treated or not with endocytosis inhibitors. The results show that entry of miPEP165a involves both a passive diffusion at the root apex and endocytosis-associated internalization in the differentiation and mature zones. Moreover, miPEP165a is unable to enter the central cylinder and does not migrate from the roots to the aerial part of the plant, suggesting that miPEPs have no systemic effect.


Introduction
Gene expression is the consequence of the transcription of an RNA molecule from a gene-modulated by transcription factors and modifications of the chromatin structure-and post-transcriptional mechanisms acting on the RNA stability of translation or on the protein it encodes. One of the best-known mechanisms of post-transcriptional regulation of gene expression is gene silencing induced by microRNAs (miRNAs). MiRNAs are small, regulatory RNA molecules (21-24 nucleotides) first discovered in the worm Caenorhabditis elegans and later in plants and humans [1][2][3][4]. Each miRNA regulates the expression of specific target gene(s) either by cleaving the mRNA transcribed from it or by inhibiting its translation. Target genes of miRNAs are often key regulatory genes encoding, for example, transcription factors or hormone receptors. MiRNAs are therefore required for the correct regulation of most developmental processes in plants and animals, and dysregulation of miRNA expression is a feature of many human pathologies.

MiPEP165a Promotes Cell Division in the Meristematic Zone to Increase Primary Root Length and Acts on Flowering Time in Arabidopsis
It has been previously shown that A. thaliana miPEP165a, as well as miR165a, is expressed in endodermis cells [5,23]. Exogenous treatment of A. thaliana seedlings with synthetic miPEP165a is sufficient to increase the primary root length [5]. However, the precise mechanisms (spatial and temporal) involved in the peptide uptake remained unknown. To study the entry of miPEPs, especially miPEP165a, we first defined the best experimental conditions to obtain a significant effect of miPEP165a on plant development. We first observed that watering plants with 100 µM of peptide was much more efficient at increasing the primary root length than treatments performed with only 10 µM of peptide, the concentration used in the previous study [5] ( Figure 1A). In addition, similar to the concentration of 10 µM previously used [5], applying miPEP165a at 100 µM also induced the activation of the pri-miRNA from which it originates ( Figure S1A). In addition, during the initial stages of the study, when the effect of miPEP165a on primary root length was studied, whatever the control used, i.e., scrambled miPEP165a, irrelevant peptides, or their corresponding solvents (acetonitrile or water), no response was observed compared to miPEP165a treatments ( Figure S1B). Similarly, water and scrambled miPEP165a had no effect on the expression of pri-miR165a compared to miPEP165a ( Figure  S1A). Finally, we observed that several freeze/thaw cycles of the peptide were detrimental to its activity on the length of primary roots ( Figure S1C). For these reasons, we used aliquots of unfrozen peptides only once and kept water as a reference in all the following experiments.
The increase in root length upon treatment may be a consequence of higher cell elongation or increased cell proliferation. To address this point, we analyzed the effect of miPEP165a at the cellular level on the meristematic zone since root growth was often determined by meristematic activity [24]. We revealed that more cells were present in the meristematic zone when roots were treated with miPEP165a ( Figure 1B-E). Therefore, these experiments suggest that the increase in root length induced by the miPEP165a treatment is likely due to the stimulation of cellular proliferation rather than an increase in cell length. Arabidopsis thaliana seedlings were treated with various concentrations of miPEP165a. Seedlings were treated daily with the peptide for 4 days, with the exception of those that received only one treatment. Peptides were thawed once, except those that underwent five freeze/thaw cycles. Two-way analysis of variance (ANOVA) significance levels were based on Tukey's post-test (1-way ANOVA), (a-c, p < 0.05, n = 70). At least three biological replicates were performed (B-E). Three-day-old seedlings were treated daily with water or 100 µM miPEP165a for a further 3 days and stained with 10 µg/mL propidium iodide for 20 min. (B) Confocal images showing the meristematic zone for the cortex cells, defined as the region between quiescent center cells and the first elongating cell that was twice the length compared to its distal neighbor (distance between white arrows) [25,26]. Meristematic cell number (C) and cell length (D) were determined with the software tool Cell-o-Tape, an open source ImageJ/Fiji macro [27][28][29]. (E) Quantification of root apical meristem length. (B-E) Four biological replicates were performed with at least 20 seedlings. Errors bars represent SEM. Asterisks indicate a significant difference at p < 0.01 (*) according to the t-test. Scale bar = 25 µm. Water was used as a control.
MiR165a and its target genes, REVOLUTA (REV), PHABULOSA (PHB), and PHAVOLUTA (PHV), are also known to be involved in flowering [30]. To investigate whether miPEP165a could have an effect on flowering, we treated the shoot apical meristem with a droplet of 100 µM miPEP165a three times a week during plant development. Treatments with miPEP165a accelerated plant development as illustrated by the decrease of the flowering day (Figure 2A,B) and the increase on primary root length. Arabidopsis thaliana seedlings were treated with various concentrations of miPEP165a. Seedlings were treated daily with the peptide for 4 days, with the exception of those that received only one treatment. Peptides were thawed once, except those that underwent five freeze/thaw cycles. Two-way analysis of variance (ANOVA) significance levels were based on Tukey's post-test (1-way ANOVA), (a-c, p < 0.05, n = 70). At least three biological replicates were performed (B-E). Three-day-old seedlings were treated daily with water or 100 µM miPEP165a for a further 3 days and stained with 10 µg/mL propidium iodide for 20 min. (B) Confocal images showing the meristematic zone for the cortex cells, defined as the region between quiescent center cells and the first elongating cell that was twice the length compared to its distal neighbor (distance between white arrows) [25,26]. Meristematic cell number (C) and cell length (D) were determined with the software tool Cell-o-Tape, an open source ImageJ/Fiji macro [27][28][29]. (E) Quantification of root apical meristem length. (B-E) Four biological replicates were performed with at least 20 seedlings. Errors bars represent SEM. Asterisks indicate a significant difference at p < 0.01 (*) according to the t-test. Scale bar = 25 µm. Water was used as a control.
MiR165a and its target genes, REVOLUTA (REV), PHABULOSA (PHB), and PHAVOLUTA (PHV), are also known to be involved in flowering [30]. To investigate whether miPEP165a could have an effect on flowering, we treated the shoot apical meristem with a droplet of 100 µM miPEP165a three times a week during plant development. Treatments with miPEP165a accelerated plant development as illustrated by the decrease of the flowering day (Figure 2A,B) and the increase of the length of the inflorescence stem ( Figure 2C,D). Interestingly, watering the roots with 10 µM peptide had no effect on the flowering, suggesting that peptides cannot migrate throughout the plant ( Figure 2). 5 of the length of the inflorescence stem ( Figure 2C,D). Interestingly, watering the roots with 10 µM peptide had no effect on the flowering, suggesting that peptides cannot migrate throughout the plant ( Figure 2).

MiPEP165a Entry Involves both Passive Diffusion at the Root Apex and Endocytic Pathways in the Differentiation and Mature Zones
To document this observation, we used the miPEP165a labelled with FAM, a fluorescent dye derived from fluorescein. As illustrated in Figure S2, the physiochemical properties of the miPEP165a-FAM are similar to those of the non-modified peptide. Although slightly less active, the labelled peptide was still able to increase the primary root length ( Figure 3).

MiPEP165a Entry Involves both Passive Diffusion at the Root Apex and Endocytic Pathways in the Differentiation and Mature Zones
To document this observation, we used the miPEP165a labelled with FAM, a fluorescent dye derived from fluorescein. As illustrated in Figure S2, the physiochemical properties of the miPEP165a-FAM are similar to those of the non-modified peptide. Although slightly less active, the labelled peptide was still able to increase the primary root length ( Figure 3).
Interestingly, while the labelled peptide penetrated rapidly (~2 h) into the root cap and the meristematic zone, it took longer to penetrate the other parts of the root ( Figure 4). Twenty-four hours after the application of the labelled peptide, the latter was present in most external parts of the roots. The central cylinder was never labelled by the peptide, which seemed to be blocked by the pericycle (Figure 4). Interestingly, while the labelled peptide penetrated rapidly (~2 h) into the root cap and the meristematic zone, it took longer to penetrate the other parts of the root ( Figure 4). Twenty-four hours after the application of the labelled peptide, the latter was present in most external parts of the roots. The central cylinder was never labelled by the peptide, which seemed to be blocked by the pericycle (Figure 4).  The entry of peptides into plants might occur passively, by diffusion, or actively, via specific transporters or by endocytosis. Because of the huge diversity of miPEPs in a plant and the lack of conservation between species [5], we hypothesized that specific transporters for each peptide are unlikely to exist and, more likely, the miPEPs might be internalized by generic internalization machinery or, more simply, by passive diffusion. To decipher the mechanisms involved in the entry of peptides into cells, we used A. thaliana mutants impaired on genes encoding proteins associated to the clathrin pathway (chc1-1, chc2-1, ap2σ2) [12,16,32] or to the membrane microdomain (rem1-2, rem1-3) [33][34][35]. Internalization of miPEP165a was not affected in most of the mutants tested, except in the root cap/meristematic zone of the chc1-1 mutant and in the differentiation zone of the chc1-1 and rem1-2 mutants, suggesting that uptake in these parts was mainly passive ( Figure 5, Figure S3). Conversely, the entry of the peptide into the mature zone of all mutants was strongly impaired ( Figure 5, Figure S3). These data suggest that peptide entry in plants involves, in addition to passive diffusion, both clathrin and membrane microdomain-mediated pathways.  In order to determine how and to what extent a defect in the peptide entry affects its biological effect on plant development, we treated the roots of chc1-1, rem1-2, and rem1-3 mutants with the peptide in parallel with the wild-type roots. While the mutants showed a longer primary root in the control conditions compared to the wild type plants, they were unable to respond to the peptide by increasing their primary root length ( Figure 6). Indeed, the rem1-2 mutant, which was strongly affected in the peptide uptake, was unable to respond to miPEP165a. 10 In order to determine how and to what extent a defect in the peptide entry affects its biological effect on plant development, we treated the roots of chc1-1, rem1-2, and rem1-3 mutants with the peptide in parallel with the wild-type roots. While the mutants showed a longer primary root in the control conditions compared to the wild type plants, they were unable to respond to the peptide by increasing their primary root length ( Figure 6). Indeed, the rem1-2 mutant, which was strongly affected in the peptide uptake, was unable to respond to miPEP165a. We next treated the aerial parts of the mutants with miPEP165a, and we observed similar results on the flowering time (Figure 7). These results suggest that the mechanisms of miPEP165a uptake into roots and aerial parts could be similar. We next treated the aerial parts of the mutants with miPEP165a, and we observed similar results on the flowering time (Figure 7). These results suggest that the mechanisms of miPEP165a uptake into roots and aerial parts could be similar.  Finally, we used TyrA23, a chemical inhibitor known to affect clathrin-mediated endocytosis [22,32,36], and MβCD, a cholesterol-depleting agent, which have been suggested to block microdomain-dependent endocytosis [17,18,22]. Interestingly, both molecules were able to inhibit the miPEP165a-activated root length phenotype, suggesting that peptide entry in plant involves clathrin-mediated endocytosis and membrane microdomain-dependent pathways (Figure 8).

12
Finally, we used TyrA23, a chemical inhibitor known to affect clathrin-mediated endocytosis [22,32,36], and MβCD, a cholesterol-depleting agent, which have been suggested to block microdomain-dependent endocytosis [17,18,22]. Interestingly, both molecules were able to inhibit the miPEP165a-activated root length phenotype, suggesting that peptide entry in plant involves clathrinmediated endocytosis and membrane microdomain-dependent pathways (Figure 8). Figure 8. Disruption of endocytic pathways prevents miPEP165a-induced root growth. Normalized primary root growth analysis after treatment with miPEP165a and TyrA23 (A) or miPEP165a and MβCD (B). Three biological replicates were performed by using at least 100 seedlings for each condition and root lengths were statistically analyzed using a t-test (p < 0.01, *). The data represent the mean value ± SEM. Water was used as a control for the miPEP165a treatment.
Altogether, our results showed that miPEP165a entry used passive diffusion at the root apex followed by endocytosis in the differentiation or mature zone of plant roots. All pathways are required to mediate full peptide uptake (and activity).

Discussion
MiRNAs have been considered for a long time as non-coding RNAs. However, a few years ago, it was shown that pri-miRNAs can encode regulatory peptides, which were named miPEPs. These miPEPs activate the transcription of their associated miRNA and thus downregulate the expression of their target genes [5]. Among miPEPs, miPEP165a induces the accumulation of mature miR165a, known to repress the expression of all five class III homeodomain-leucine zipper (HD-ZIP III) transcription factors, i.e., REV, PHB, PHV, CORONA (CAN/AtHB15), and AtHB8 [5,37]. In Arabidopsis, the overexpression of all HD-ZIP III results in plants with shorter roots whereas phb, phv double mutants and phv-11 mutants display longer roots as well as an increase in the number of meristem cells compared to wild-type plants [38,39]. Moreover, the overexpression of miR166, differing by only one nucleotide from miR165 and targeting the expression of three HD-ZIP III genes, also promotes Figure 8. Disruption of endocytic pathways prevents miPEP165a-induced root growth. Normalized primary root growth analysis after treatment with miPEP165a and TyrA23 (A) or miPEP165a and MβCD (B). Three biological replicates were performed by using at least 100 seedlings for each condition and root lengths were statistically analyzed using a t-test (p < 0.01, *). The data represent the mean value ± SEM. Water was used as a control for the miPEP165a treatment.
Altogether, our results showed that miPEP165a entry used passive diffusion at the root apex followed by endocytosis in the differentiation or mature zone of plant roots. All pathways are required to mediate full peptide uptake (and activity).

Discussion
MiRNAs have been considered for a long time as non-coding RNAs. However, a few years ago, it was shown that pri-miRNAs can encode regulatory peptides, which were named miPEPs. These miPEPs activate the transcription of their associated miRNA and thus downregulate the expression of their target genes [5]. Among miPEPs, miPEP165a induces the accumulation of mature miR165a, known to repress the expression of all five class III homeodomain-leucine zipper (HD-ZIP III) transcription factors, i.e., REV, PHB, PHV, CORONA (CAN/AtHB15), and AtHB8 [5,37]. In Arabidopsis, the overexpression of all HD-ZIP III results in plants with shorter roots whereas phb, phv double mutants and phv-11 mutants display longer roots as well as an increase in the number of meristem cells compared to wild-type plants [38,39]. Moreover, the overexpression of miR166, differing by only one nucleotide from miR165 and targeting the expression of three HD-ZIP III genes, also promotes primary root growth in Arabidopsis [39]. These results can be correlated with those of the present study, since we showed that miPEP165a promotes primary root growth by increasing cell division in the root apical meristem (Figure 1). Moreover, misexpression of the HD-ZIP III genes by making them resistant to miR165/166 and a reduction in the expression of HD-ZIP IIIs by overexpression of miR165/166 induces prolonged activity of floral stem cells [30]. Here, we observed that miPEP165a accelerates the appearance of the inflorescence stem and the flowering time of Arabidopsis wild-type plants (Figure 2).
Since some small peptides were considered as long-distance signaling molecules, we wondered whether miPEP165a was involved in root/shoot communication [40][41][42]. By tracking the FAM-labelled miPEP165a across all layers of Arabidopsis roots, we showed that the labelled peptide entered into the epidermis and migrated up to the pericycle but did not reach the root vessels ( Figure 4). Moreover, the acceleration of flowering observed in response to the miPEP165a treatment of the shoot apical meristem was not observed after watering Arabidopsis roots with miPEP165a ( Figure 2). Taken together, these results indicate that miPEP165a is not a root-to-shoot mobile signal molecule.
Consequently, in order to have a better understanding of miPEP uptake into plants, we investigated the mobility of FAM-labelled miPEP165a in Arabidopsis roots. Clathrin-mediated endocytosis is the major and the most studied route of entry in plants [8]. A recent study showed that this endocytic pathway is necessary for the internalization of the elicitor peptide Atpep1 and its receptor, leading to Atpep1-induced responses [16]. Here, we showed that the entry of miPEP165a could also be dependent on clathrin since miPEP165a uptake was significantly decreased in the primary roots of chc1-1 and strongly reduced in the mature zone in the three mutants chc1-1, chc2-1, and ap2σ2 ( Figure 5). These results were confirmed by the fact that the increase of the root length by miPEP165a was not observed in the chc1-1 mutant or after treatment with TyrA23 ( Figure 6A, Figure 8A), the most commonly used CME inhibitor [8,32,36]. Similarly, the acceleration of the flowering time induced by miPEP165a in wild-type plants was not observed in the chc1-1 mutant (Figure 7).
Besides clathrin-mediated endocytosis, membrane microdomain-associated endocytosis has been described in plants as an alternative route of entry pathway [8]. This endocytosis pathway is sensitive to sterol depletion and consequently to the sterol-depleting agent MβCD [8,17,18]. In the present study, we showed that MβCD prevented miPEP165a-FAM entry and correlatively the increase of root length induced by miPEP165a ( Figure 8B, Figure S4). Collectively, our results indicate that both clathrin-dependent pathways and microdomain-associated events may cooperate in peptide entry into Arabidopsis roots. Previous results have demonstrated that internalization of the aquaporin PIP2;1 and RbohD involved both dependent and independent clathrin-mediated endocytosis, the latter being stimulated in saline stress conditions [17,22]. Stimulation of the endocytic pathway under salt stress requires the simultaneous action of both clathrin-dependent and membrane microdomain-associated endocytosis [17,22]. In addition, Baral and his colleagues have shown that clathrin-mediated endocytosis allows the internalization of transmembrane proteins in all cell root layers whereas a sterol-sensitive clathrin-independent pathway internalizes lipid-anchored cargoes only in the epidermal cell layer [18]. Moreover, these authors showed that salt stress activates an additional clathrin-independent endocytosis pathway across all cell root layers that takes up both molecule types [18]. Considering membrane microdomain-associated endocytosis, it is known that proteins assemble into clusters in lipid rafts [8]. Among these proteins, remorins are considered as markers of membrane microdomains [35]. In Medicago truncatula, the symbiotic remorin 1 forms clusters and interacts with symbiotic receptors at the plasma membrane, playing a key role in bacterial signal perception [21]. Here, we showed that remorins 1-2 and 1-3, which are among the 10% of the most highly expressed genes in Arabidopsis [43], were also involved in miPEP165a entry into Arabidopsis roots ( Figure 6B, Figure 7). Indeed, miPEP165a-FAM failed to enter the differentiation zone of Arabidopsis roots in rem1-2 and rem1-3 mutants. Moreover, root length and flowering acceleration induced by miPEP165a were perturbed in both remorin mutants ( Figure 6B, Figure 7).
To conclude, we showed that endocytic pathways participate in miPEP uptake in plants. Thus, clathrin-mediated endocytosis as well as membrane microdomain-associated pathways seem to cooperate, allowing miPEPs to regulate their corresponding miRNAs and consequently modulate the plant phenotype, such as flowering and root development. Due to the simplicity of the mode of administration of miPEPs, a better understanding of miPEP uptake into plants is a first step towards the possible agronomic application of peptides.

Peptide Treatment of Arabidopsis Roots
Surface-sterilized Arabidopsis seeds were sown on the surface of cellophane membrane placed on 1 2 MS solid medium and stratified for one day at 4 • C in the dark. Seeds were vertically grown in controlled environmental chambers at 22/20 • C, with a photoperiod of 16h light/8h dark, an irradiance of~97.5 µmol photons.m −2 .s −1 , and a relative humidity of 40%. Three days after sowing, seedlings were treated daily for 4 days either with water, 2.5% acetonitrile, 100 µM scrambled miPEP165a, 100 µM irrelevant peptides (PEP1, PEP2, PEP3), or fluorescein (control conditions) or with 100 µM miPEP165a or miPEP165a-FAM (treated conditions). Twenty-four hours after the last treatment, seedlings were scanned in order to measure primary root lengths using NeuronJ plugin of ImageJ.

Peptide Uptake in Arabidopsis Roots
Surface-sterilized wild-type and mutant Arabidopsis seeds were grown onto 1 2 MS solid medium in the same conditions as those described in the previous section. Three days after germination, three seedlings were transferred to each well of a 48-well plate containing 200 µL of 1 2 MS liquid medium. One day later, medium was replaced by 10 µM miPEP165a-FAM diluted in 1 2 MS liquid medium until confocal microscopy observations. FAM fluorescence was analyzed with a confocal laser scanning microscope (Leica TCS SP2-AOBS using a 40 X water immersion objective lens (numerical aperture 0.80; HCX APO). FAM fluorescence was excited with the 488-nm ray line of the argon laser and recorded in the 511-551-nm emission range.
For quantification of miPEP165a-FAM entry into wild-type and mutant Arabidopsis roots, the fluorescence intensity was determined per surface unit in the different root zones using ImageJ software.

Inhibitor Treatment
TyrA23 was dissolved in dimethyl sulfoxide to yield a 50 mM stock solution and MβCD was prepared in deionized water at a final concentration of 38 mM. For each experiment, 3-day-old seedlings germinated on 1 2 MS solid medium + 1% sucrose (wt/vol) were pre-treated with 50 µM TyrA23 or 10 mM MβCD for 30 min [17]. Seedlings were then treated with the inhibitors supplemented with 100 µM miPEP165a. Treatments were performed daily for an additional 3 days and plates were scanned for analysis of the primary root length with NeuronJ, an Image J plugin [29,44].

Flowering Phenotype
Arabidopsis seeds were grown on Jiffy ® under a 16 h light/8 h dark cycle (22/20 • C), with a relative humidity of 80%. Fifteen days after seed sowing, either a 2-µL droplet of 100 µM miPEP165a was put on the shoot apical meristem or seedlings were watered with 500 µL of 10 µM miPEP165a three times a week. Analyses of the aerial parts were performed 24 days after sowing.

Propidium Iodide Staining
Wild-type seeds were grown for 3 days on 1 2 MS solid medium + 1% sucrose (wt/vol) in the same growth conditions as described above. Seedlings were then treated with water or 100 µM miPEP165a daily for 3 additional days and placed in the growth chamber at the same settings. Seedlings were then stained with 10 µg/mL propidium iodide for 20 min and Arabidopsis cell roots were analyzed with a laser scanning confocal microscope (Leica TCS SP8-AOBS) with a ×25 water immersion objective lens (numeral aperture 0.95; Fluotar Visir). The excitation and emission wavelengths of propidium iodide were 561 and 570-640 nm, respectively.
The meristematic zone for the cortex cells was defined as the region between quiescent center cells and the first elongating cell that was twice the length compared to its distal neighbor [20,26]. The meristematic cell length and cell number were determined with the software tool Cell-o-Tape, an open source ImageJ/Fiji macro [27][28][29]. At least 20 roots were analyzed for each treatment.

Immunoblots and RT-qPCR
Seven-week-old Arabidopsis seedlings were treated with 100 µM miPEP165a or its corresponding control for 24 h, and then the expression of pri-miR165a was evaluated by RT-qPCR according to Lauressergues et al. [5].
To evaluate miPEP165a stability, 5 nanomoles of miPEP165a were subjected to several freeze/thaw cycles and its degradation was detected by immunoblotting with an anti-miPEP165a antibody as previously described [5].