Root growth of Nicotiana attenuata is decreased immediately after simulated leaf herbivore attack

Image-based non-destructive methods were used to quantify root growth reactions happening within hours following simulated leaf herbivore attack.¹ The induction of wound reactions in leaves of seedlings of Nicotiana attenuata led to transiently decreased root growth rates: Upon application of the oral secretions and regurgitants of the specialist herbivore Manduca sexta, a transient decrease in root growth was observed that was more pronounced than if a mere mechanical wounding was imposed. Root growth reduction was more severe than leaf growth reduction and the timing of the transient growth reduction coincided with endogenous bursts of jasmonate (JA) and ethylene emissions reported in literature. The reaction of root growth was superimposed by a strong diel variation of root growth, which was caused by the fluctuating temperature to which the plants were exposed. Apart from the observed root growth reaction, other defense-related traits such as increased nicotine concentration, trichome length and density were activated within 72 h after wounding. Further experiments indicated that the response was elicited by fatty acid-amino acid conjugates that are contained in the oral secretions and that JA signalling is crucial for root-shoot communication here.Key words: image analysis, plant-insect interactions, Manduca sextaPlants constantly need to acclimate to a suite of fluctuating biotic and abiotic factors. Within the framework of their genetically determined response options, they react towards stress situations by maximizing protection against stress while minimizing deviations from optimal growth and development. Feeding of the larvae of the specialist lepidopteran herbivore Manduca sexta on N. attenuata has become a model case scenario for studying biotic stress.² It has been shown in great detail there, that a diverse set of plant hormones like JA, methyl jasmonate and ethylene are rapidly induced immediately following wounding or herbivore attack.^3,4,5 Acclimation occurs both on the biochemical and morphological level, e.g., by increasing defence compounds,⁶ by synthesizing defence related proteins, by emitting volatiles to attract predators and parasites of herbivores⁷ and by altering plant morphology via increased formation of trichomes, thorns or scleromorphy.^8,9 Many studies demonstrated resource based trade-offs between growth and defence on a large time scale (e.g. refs. ¹⁰ and ¹¹). Yet, it is unclear how fast a trade-off-linked reorganization of plant metabolism, diverting resources away from growth or development towards defence, can occur.Recently, development of growth imaging methods has allowed to study short-term growth responses of above-and belowground sink organs towards fluctuations of environmental factors, such as alterations of light climate¹² or responses towards gravitropic stimuli.^13–15 Clear responses are often seen best in young seedlings as they grow fast and as they can be cultivated in agar-filled, translucent Petri dishes allowing live imaging. Hence, it was the aim of our study to monitor defence reactions in the N. attenuata seedling system while at the same time assessing the dynamics of root growth reaction following such an attack.Herbivory-induced wound reactions were simulated by applying different substances on the mechanically wounded primary leaf of the seedling plant 16 d after germination.¹ Investigated substances were: (i) oral secretions and regurgitants of M. sexta larvae (OS), (ii) methyl jasmonate, (iii) fatty acid-amino acid conjugates, (iv) control solutions for the different test substances (water, buffer, lanolin). All wounding treatments were performed at the same time of the day to make sure that the reaction was not masked by fluctuating responses of the plant to temperature changes (plants were grown in a 14 h/10 h light/ dark regime with temperatures of 26°C and 22°C at day and night, respectively).Root growth was affected strongly by temperature (Fig. 1A) and increased linearly with temperature in the analyzed range. Root growth decreased markedly, if OS was added to the wounds. OS-treated and wounded control plants differed in growth reduction from about 2 h after wounding onwards. To uncouple this reduction from temperature-induced growth reductions, the experiment was repeated under continuous light and temperature, leading to a comparable result (Fig. 1B). When OS was added to wounds, nicotine concentration and trichome length and density increased within 72 h after the wounding reaction. Root growth reaction was comparable between OS-treatment and treatments, in which methyl jasmonate or fatty-acid amino acid conjugates were supplied to the wounds. The strongest difference concerning the distribution of relative elemental growth rate along the root growth zone between control and OS treatment was observed about 1 mm behind the root tip in the beginning of the central elongation zone. Leaf growth was monitored in an additional set of experiments. There, a transient decrease of growth was seen as well, but it was not as pronounced as for root growth.Open in a separate windowFigure 1Root growth in Nicotiana attenuata, grown on agar-filled Petri dishes. (A) Velocity of the root tip (V_Tip) versus temperature of the agar medium. (B) Velocity of the root tip of control plants and of plants with leaves that were wounded at t=0 and treated with oral secretions and regurgitants (OS) of Manduca sexta. Plants were grown in continuous light at T = 26°C (n = 3, mean values and SE).The kinetics of the root growth depression point on a superposition of two physiological effects: During the first hour after wounding, root growth decreases similarly in control and OS-treated plants, indicating a mere hydraulic response due to the loss of water and turgor pressure.¹² Thereafter, a specific response towards OS is observed, which coincides with the time frame of about 2–3 h that was reported for the systemic increase of JA transported from shoot to root. Baldwin et al¹⁶ reported that JA pools in roots are increased systemically (3.5-fold) within 180 minutes following mechanical wounding. Hence, JA might mediate the reduction in root growth and induce the defense machinery, which diverts resources from growth to defense.     

Brassinosteroid negatively regulates jasmonate inhibition of root growth in Arabidopsis

Jasmonate (JA) inhibits root growth of Arabidopsis thaliana seedlings. The mutation in COI1, that plays a central role in JA signaling, displays insensitivity to JA inhibition of root growth. To dissect JA signaling pathway, we recently isolated one mutant named psc1, which partially suppresses coi1 insensitivity to JA inhibition of root growth. As we identified the PSC1 gene as an allele of DWF4 that encodes a key enzyme in brassinosteroid (BR) biosynthesis, we hypothesized and demonstrated that BR is involved in JA signaling and negatively regulates JA inhibition of root growth. In our Plant Physiology paper, we analyzed effects of psc1 or exogenous BR on the inhibition of root growth by JA. Here we show that treatment with brassinazole (Brz), a BR biosynthesis inhibitor, increased JA sensitivity in both coi1-2 and wild type, which further confirms that BR negatively regulates JA inhibition of root growth. Since effects of psc1, Brz and exogenous BR on JA inhibition of root growth were mild, we suggests that BR negatively finely regulates JA inhibition of root growth in Arabidopsis.Key words: jasmonate signaling, root growth, brassinosteroid, brassinazole, arabidopsisJasmonate (JA) regulates many plant developmental processes and stress responses.^1,2 COI1 plays a central role in JA signaling and is required for all JA responses in Arabidopsis.^3,4 coi1-1, a strong mutation in COI1, is male sterile and exhibits loss of all JA responses tested to date, such as JA inhibition of root growth, the expression of JA-induced genes, and susceptibility to insect attack and pathogen infection, and coi1-2, a weak mutant of COI1, shows similar JA responses to coi1-1 except for partially fertile that makes it able to produce a small quantity of seeds.⁵To investigate COI1-mediated JA responses and dissect JA signaling pathway, we conducted genetic screens for suppressors of coi1-2. Previously, we identified cos1 that completely suppresses coil-2 insensitive to JA.⁶ Recently, we isolated the psc1 mutant that partially suppresses coi1-2 insensitivity to JA, and found that PSC1 is an allele of DWF4.⁷Since the DWF4 gene encodes a key enzyme in brassinosteroid (BR) biosynthesis,⁸ we hypothesized that BR is involved in JA signaling. By physiological analysis, we showed that psc1 partially restored JA inhibition of root growth in coi1-2 background and displayed JA hypersensitivity in wild-type COI1 background, the effects of psc1 were eliminated by exogenous BR, and that exogenous BR could attenuated JA inhibition of root growth in wild type. These findings demonstrated that BR is involved in JA signaling and indicated that BR negatively regulates JA inhibition of root growth.BR is a family of polyhydroxylated steroid hormones involved in many aspects of plant growth and development. The BR-deficient mutants exhibited severely retarded growth that was able to be rescued by exogenous BR.⁹ Brassinazole (Brz) is a BR biosynthesis inhibitor. The Arabidopsis seedlings treated with Brz displayed a BR deficient-mutant-like phenotype, which could be elimilated by exogenous BR.¹⁰To determine wether treatment with Brz affects JA inhibition of root growth, the seedlings of wild type and coi1-2 were grown in MS medium supplemented with MeJA and/or Brz. As shown in Figure 1, the relative root length was obviously reduced in both coi1-2 and wild type when treated with Brz relative to without Brz, indicating that the repression of BR biosynthesis by Brz could increase JA sensitivity. These results further confirm BR negatively regulates JA inhibition of root growth.Open in a separate windowFigure 1Effect of Brz on JA inhibition of root growth. Brz increased JA inhibition of root growth in both coi1-2 and wild type (WT). Root length of 7-day-old seedlings grown in MS medium containing 0, 5 and 10 μM MeJA without (−) or with (+) 0.5 μM Brz was expressed as a percentage of root length in MS without (−) or with (+) 0.5 µM Brz. Error bars represent SE (n > 30).It has been demonstrated that JA connects with other plant hormones including auxin, ethylene, abscisic acid, salicylic acid and gibberellin to form complex regulatory networks modulating plant developmental and stress responses.^11–15 We found that BR negatively regulates JA inhibition of root growth, suggesting that a cross talk between JA and BR exists in planta, which extends our understandings on the JA signal transduction.COI1 is a JA receptor¹⁶ and DWF4 catalyzes the rate-limiting step in BR-biosynthesis pathway.⁸ We found that JA inhibits DWF4 expression, this inhibition was dependent on COI1,⁷ indicating that DWF4 is downregulated by JA and is located downstream of COI1 in the JA signaling pathway.Since the effects of psc1, Brz, and exogenous BR on JA inhibition of root growth were mild, and the DWF4 expression was partially repressed by JA (Ren et al. 2009, Fig. 1), we suggest that BR negatively finely regulates JA inhibition of root growth, and propose a model for these regulations. As shown in Figure 2A, JA signal passes COI1 repressing substrates, such as JAZs,^17,18 i.e., JA activates degradation of substrates via SCF^COI1-26S proteasome,^16–18 whereas substrates positively regulate root growth through other regulators. JA also partially inhibits DWF4 expression through COI1, reducing BR that is required for root growth.^7,9 Mutation in COI1 interrupts JA signaling for failing in degradation of substrates and repression of DWF4 as well, resulting in JA-insensitivity (Fig. 2B). However, mutation in DWF4 or treatment with Brz causes a reduction in BR, which affects root growth, leading to JA-hypersensitivity in wild-type COI1 background (Fig. 2C and E) and partial restoration of JA sensitivity in coi1-2 background (Fig. 2D and F). Whereas, an application of exogenous BR could eliminate the effect of BR reduction resulted from repression of DWF4 by JA on root growth, attenuating JA sensitivity in wild type (Fig. 2G). Because the inhibition of DWF4 expression by JA is dependent on COI1, the coi1 mutant treated with exogenous BR do not show alteration in JA sensitivity (Fig. 2H).Open in a separate windowFigure 2A model for that BR negatively finely regulates JA inhibition of root growth in Arabidopsis. (A–D) Treatment with JA in wild type (A), coi1-2 (B), psc1 (C) and psc1coi1 (D). (E and F) Treatments with JA and Brz in wild type (E) and coi1-2 (F). (G and H) Treatments with JA and exogenous BR in wild type (G) and coi1-2 (H). Arrows indicate positive regulation or enhancement, whereas blunted lines indicate repression or negative regulation. Crosses indicate interruption or impairment. The letter “S” indicates substrates of SCF^COI1. Thicker arrows and blunted lines represent the central JA signaling pathway regulating JA inhibition of root growth. Broken arrows represent JA signaling pathway in which other regulators are involved. The intensity of gray boxes represents the degree of JA inhibition on root growth.     

Chemical Signal as a Rapid Long-Distance Information Messenger After Local Wounding of a Plant?

A series of works have described an important role of chemical signaling compounds in generation of the stress response of plants in both the wounded and distant undamaged plant tissues. However, pure chemical signals are often not considered in the fast (minutes) long-distance signaling (systemic response) because of their slow propagation speed. Physical signals (electrical and hydraulic) or a combination of the physical and chemical signals (hydraulic dispersal of solutes) have been proposed as possible linkers of the local wound and the rapid systemic response. We have recently demonstrated an evidence for involvement of chemical compounds (jasmonic and abscisic acids) in the rapid (within 1 hour) inhibition of photosynthetic rate and stomata conductance in distant undamaged tobacco leaves after local burning. The aim of this addendum is to discuss plausible mechanisms of a rapid long-distance chemical signaling and the putative interactions between the physical and chemical signals leading to the fast systemic response.Key Words: tobacco, local burning, systemic response, hydraulic surge, electrical signal, abscisic acid, jasmonic acidPlants have evolved an amazingly complex system of defence-related strategies to protect themselves upon local wounding.^1–7 Important characteristics of self-defence responses of plants are their velocity and ubiquity. Indeed, fast (minutes to hours) responses to injurious factors have been detected in the site of injury and in distant regions (systemic response) in various plants.^8–11 These findings suggest that a signal generated by an attack to one leaf is transmitted through a whole plant. Several kinds of chemical^3,6 and physical¹² signals induced by local wounding and even their combination¹³ have been implicated. However, a little is known about the interactions of these signals and about the mechanisms of initiation of the short-term systemic responses.We have used a model system—tobacco plants exposed to the local burning—to study the signals involved in rapid wound responses of photosynthetic apparatus.¹⁴ Local burning of an upper leaf of a tobacco plant induced rapid changes in surface electrical potential (within seconds) and a pronounced fast decline in the stomatal conductance, CO₂ assimilation and transpiration (within minutes) in the basipetal direction (Fig. 1). Moreover, we have detected a fast (within minutes) transient increase in levels of endogenous abscisic acid (ABA) followed by a huge rapid rise in endogenous jasmonic acid (JA) in the leaf below the burned one. ABA and jasmonates are known to be involved in signaling pathways leading to stomatal closure and downregulation of photosynthesis.^15,16 Increases in ABA and/or JA levels have only previously been detected in remote untreated tissues several hours after local wounding^8,9 suggesting that chemical signals are too slow to induce rapid systemic response. Previous works have reported that fast physical (electrical) signals play an essential role in short-term systemic photosynthetic responses.^11,17 However, a several-minutes delayed stomata closing response after the initiation of electrical potential changes has been reported in Mimosa¹⁸ and in our case in tobacco¹⁴ plants. Therefore, the guard cell deflation is most likely triggered not only by the electrical signal, but also by indirect factors. Based on close correlations, our results now provide a new evidence for the idea that chemical signals (ABA and mainly JA) participate in mediating the short-term systemic photosynthetic responses to local burning in tobacco plants.Open in a separate windowFigure 1The model of putative signalling pathways leading to the rapid systemic responses of tobacco plants to local burning. Hypothetical (dashed lines) local responses, generation of signals and transport processes and detected (full line) systemic responses are demonstrated. For details see the text.The question is how do the physical (electrical and/or hydraulic) and chemical signals act? They may independently induce specific elements of systemic responses. However, they are more likely to act in a coordinated, interactive fashion. In this scenario (see Fig. 1), within first minutes after the local burning, hydraulic surge transmitted basipetally and acropetally through the xylem would transport chemicals released at the wound site (hydraulic dispersal¹⁹) and evoke changes in the ion fluxes in surrounding living cells leading to the local electrical activity.^12,13 The hypothesis of hydraulic dispersal is supported by our preliminary experiments with the fluorescent dye Rhodamine B applied on cut petiole of the upper leaf of tobacco plants showing that solutes can be rapidly transported (within minutes) basipetally following wounding.The rapid kinetics and transient character of ABA accumulation¹⁴ suggest that the main transport mechanism is the hydraulic dispersal in xylem. The participation of ABA in the generation of systemic electrical activity and/or vice versa cannot be ruled out.^8,20A rapid hydraulically driven transport of chemicals in the xylem of wounded plant in a reversed (basipetal) direction^19,21 to transpiration stream is not generally accepted. Exposing of leaves of undamaged plants to radioactive labelled molecules to determine the speed of chemical signal transport could be misrepresent, because hydraulic signal is not generated in undamaged plants and then the detected transport speed is too slow. Moreover, previous work²² demonstrated that neither the mass flow itself, nor the associated pressure changes induce the systemic response (the proteinase inhibitor activity). Thus, the efficacy of chemical agents in rapid systemic signaling seems to depend on transport by the mass flows associated with hydraulic signals.¹⁹However, hydraulic dispersal acts only for minutes, until all water released at the wound site is exhausted.²¹ A requirement for hydraulic dispersal of any solute is its presence in the wounded tissue at the time of wounding.¹⁹ Detected slower kinetics of JA accumulation than in the case of ABA and the huge rise of JA levels¹⁴ indicate a systemic accumulation of JA also by some additional processes.Does additional JA accumulation result from de-novo synthesis in undamaged leaves as a response to physical signal or does it result from a JA transport from the wounded leaves? In the longer time-frame the phloem transport²³ should also be considered. Experiments with tomato plants have shown that de novo JA synthesis in distant leaves is not required for the systemic response and that biosynthesis of JA at the wound site is necessary for the generation of a systemic signal.⁷ Indeed, a short-term increase in endogenous concentrations of JA has been detected in wounded tissue in Nicotiana sylvestris⁹ and rice.¹⁰However, a rapid burst in the systemic JA accumulation found in our experiments¹⁴ would implicate an ultra-rapid and extreme JA accumulation at the wound site before its transport. The systemic JA accumulation (within 1 hour¹⁴) preceded the generation of enzymes involved in the JA biosynthesis in the wounded leaf.Thus, several processes are suggested to play a role in the ultra-rapid and huge JA accumulation:

1. initiation of JA accumulation by preexisting enzymes,²⁴
2. fast release of free JA from its storage pools in cells (e.g., JA-conjugates²⁵),
3. direct uptake of elicitors (JA) by the phloem of the wounded leaf and exchange between the xylem and phloem as a consequence of severe wounds,²⁶
4. the mass flow (containing remaining JA) driven mainly acropetally in the xylem by transpiration after damping the hydraulic surge,²¹
5. JA accumulation evoked by the fast transmitting physical (electrical or hydraulic) signal that leads to imbalances in ion fluxes,^8,12,27
6. JA accumulation (and subsequent transport) directly in the phloem, where JA biosynthetic enzymes are located (at least in tomato²⁴),
7. volatile chemical compounds (methylester of JA) spreading in the surrounding air of wounded leaf could serve as signaling molecules and sources of JA.^25,28

The relevance of the above mentioned mechanisms should be checked by further research. Complex quantitative and kinetic analysis of JA and ABA content, levels of its biosynthetic derivatives (also volatiles in the surrounding air) and simultaneous physical signal detection in wounded and distant unwounded tissues would fill the remaining void about their role and interactions in the wound signal transduction networks. In addition, a suppression of other signaling pathways with similar transport kinetics (e.g., volatile compounds transmission, systemin and oligosaccharides generation and/or transport, using mutant plants) would be useful.Substantial similarity between the rapid physical (electrical) signaling in animal nervous system compared with the physical (electrical) signaling in plants has already been reported.^29,30 Interaction of chemical and electrical signals is the process well documented for post-synaptic events in animals. Our data now strengthen the role of chemical signals next to the role of physical signals in plants in the rapid systemic wound response; such a role of chemicals in plants was often underestimated up to now.     

ARF-GTPase as a Molecular Switch for Polar Auxin Transport Mediated by Vesicle Trafficking in Root Development

Polar auxin transport (PAT), which is controlled precisely by both auxin efflux and influx facilitators and mediated by the cell trafficking system, modulates organogenesis, development and root gravitropism. ADP-ribosylation factor (ARF)-GTPase protein is catalyzed to switch to the GTP-bound type by a guanine nucleotide exchange factor (GEF) and promoted for hybridization to the GDP-bound type by a GTPase-activating protein (GAP). Previous studies showed that auxin efflux facilitators such as PIN1 are regulated by GNOM, an ARF-GEF, in Arabidopsis. In the November issue of The Plant Journal, we reported that the auxin influx facilitator AUX1 was regulated by ARF-GAP via the vesicle trafficking system.¹ In this addendum, we report that overexpression of OsAGAP leads to enhanced root gravitropism and propose a new model of PAT regulation: a loop mechanism between ARF-GAP and GEF mediated by vesicle trafficking to regulate PAT at influx and efflux facilitators, thus controlling root development in plants.Key Words: ADP-ribosylation factor (ARF), ARF-GAP, ARF-GEF, auxin, GNOM, polar transport of auxinPolar auxin transport (PAT) is a unique process in plants. It results in alteration of auxin level, which controls organogenesis and development and a series of physiological processes, such as vascular differentiation, apical dominance, and tropic growth.² Genetic and physiological studies identified that PAT depends on efflux facilitators such as PIN family proteins and influx facilitators such as AUX1 in Arabidopsis.Eight PIN family proteins, AtPIN1 to AtPIN8, exist in Arabidopsis. AtPIN1 is located at the basal side of the plasma membrane in vascular tissues but is weak in cortical tissues, which supports the hypothesis of chemical pervasion.³ AtPIN2 is localized at the apical side of epidermal cells and basally in cortical cells.^1,4 GNOM, an ARF GEF, modulates the localization of PIN1 and vesicle trafficking and affects root development.^5,6 The PIN auxin-efflux facilitator network controls root growth and patterning in Arabidopsis.⁴ As well, asymmetric localization of AUX1 occurs in the root cells of Arabidopsis plants,⁷ and overexpression of OsAGAP interferes with localization of AUX1.¹ Our data support that ARF-GAP mediates auxin influx and auxin-dependent root growth and patterning, which involves vesicle trafficking.¹ Here we show that OsAGAP overexpression leads to enhanced gravitropic response in transgenic rice plants. We propose a model whereby ARF GTPase is a molecular switch to control PAT and root growth and development.Overexpression of OsAGAP led to reduced growth in primary or adventitious roots of rice as compared with wild-type rice.¹ Gravitropism assay revealed transgenic rice overxpressing OsAGAP with a faster response to gravity than the wild type during 24-h treatment. However, 1-naphthyl acetic acid (NAA) treatment promoted the gravitropic response of the wild type, with no difference in response between the OsAGAP transgenic plants and the wild type plants (Fig. 1). The phenotype of enhanced gravitropic response in the transgenic plants was similar to that in the mutants atmdr1-100 and atmdr1-100/atpgp1-100 related to Arabidopsis ABC (ATP-binding cassette) transporter and defective in PAT.⁸ The physiological data, as well as data on localization of auxin transport facilitators, support ARF-GAP modulating PAT via regulating the location of the auxin influx facilitator AUX1.¹ So the alteration in gravitropic response in the OsAGAP transgenic plants was explained by a defect in PAT.Open in a separate windowFigure 1Gravitropism of OsAGAP overexpressing transgenic rice roots and response to 1-naphthyl acetic acid (NAA). (A) Gravitropism phenotype of wild type (WT) and OsAGAP overexpressing roots at 6 hr gravi-stimulation (top panel) and 0 hr as a treatment control (bottom panel). (B) Time course of gravitropic response in transgenic roots. (C and D) results correspond to those in (A and B), except for treatment with NAA (5 × 10⁻⁷ M).The polarity of auxin transport is controlled by the asymmetric distribution of auxin transport proteins, efflux facilitators and influx carriers. ARF GTPase is a key member in vesicle trafficking system and modulates cell polarity and PAT in plants. Thus, ARF-GDP or GTP bound with GEF or GAP determines the ARF function on auxin efflux facilitators (such as PIN1) or influx ones (such as AUX1).ARF1, targeting ROP2 and PIN2, affects epidermal cell polarity.⁹ GNOM is involved in the regulation of PIN1 asymmetric localization in cells and its related function in organogenesis and development.⁶ Although VAN3, an ARF-GAP in Arabidopsis, is located in a subpopulation of the trans-Golgi transport network (TGN), which is involved in leaf vascular network formation, it does not affect PAT.¹⁰ OsAGAP possesses an ARF GTPase-activating function in rice.¹¹ Specifically, our evidence supports that ARF-GAP bound with ARF-GTP modulates PAT and gravitropism via AUX1, mediated by vesicle trafficking, including the Golgi stack.¹Therefore, we propose a loop mechanism between ARF-GAP and GEF mediated by the vascular trafficking system in regulating PAT at influx and efflux facilitators, which controls root development and gravitropism in plants (Fig. 2). Here we emphasize that ARF-GEF catalyzes a conversion of ARF-bound GDP to GTP, which is necessary for the efficient delivery of the vesicle to the target membrane.¹² An opposite process of ARF-bound GDP to GTP is promoted by ARF-GTPase-activating protein via binding. A loop status of ARF-GTP and ARF-GDP bound with their appurtenances controls different auxin facilitators and regulates root development and gravitropism.Open in a separate windowFigure 2Model for ARF GTPase as a molecular switch for the polar auxin transport mediated by the vesicle traffic system.