Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Similar Intracellular Location and Stimulus Reactivity, but Differential Mobility of Tailless (Vicia faba) and Tailed Forisomes (Phaseolus vulgaris) in Intact Sieve Tubes

  • Alexandra C. U. Furch ,

    Contributed equally to this work with: Alexandra C. U. Furch, Stefanie V. Buxa

    Affiliation Institute of General Botany and Plant Physiology, Friedrich-Schiller-University, Jena, Germany

  • Stefanie V. Buxa ,

    Contributed equally to this work with: Alexandra C. U. Furch, Stefanie V. Buxa

    Affiliation Research Centre for BioSystems, Land Use and Nutrition, Institute of Phytopathology and Applied Zoology, Justus-Liebig University, Giessen, Germany

  • Aart J. E. van Bel

    Affiliation Plant Cell Biology Research Group, Institute of General Botany, Justus-Liebig University, Giessen, Germany

Similar Intracellular Location and Stimulus Reactivity, but Differential Mobility of Tailless (Vicia faba) and Tailed Forisomes (Phaseolus vulgaris) in Intact Sieve Tubes

  • Alexandra C. U. Furch, 
  • Stefanie V. Buxa, 
  • Aart J. E. van Bel


Sieve elements of legumes contain forisomes—fusiform protein bodies that are responsible for sieve-tube occlusion in response to damage or wound signals. Earlier work described the existence of tailless and tailed forisomes. This study intended to quantify and compare location and position of tailless (in Vicia faba) and tailed (in Phaseolus vulgaris) forisomes inside sieve elements and to assess their reactivity and potential mobility in response to a remote stimulus. Location (distribution within sieve elements) and position (forisome tip contacts) of more than altogether 2000 forisomes were screened in 500 intact plants by laser scanning confocal microscopy in the transmission mode. Furthermore, we studied the dispersion of forisomes at different locations in different positions and their positional behaviour in response to distant heat shocks. Forisome distribution turned out to be species-specific, whereas forisome positions at various locations were largely similar in bushbean (Phaseolus) and broadbean (Vicia). In general, the tailless forisomes had higher dispersion rates in response to heat shocks than the tailed forisomes and forisomes at the downstream (basal) end dispersed more frequently than those at the upstream end (apical). In contrast to the tailless forisomes that only oscillate in response to heat shocks, downstream-located tailed forisomes can cover considerable distances within sieve elements. This displacement was prevented by gentle rubbing of the leaf (priming) before the heat shock. Movement of these forisomes was also prohibited by Latrunculin A, an inhibitor of actin polymerization. The apparently active mobility of tailed forisomes gives credence to the idea that at least the latter forisomes are not free-floating, but connected to other sieve-element structures.


Sieve elements of legumes contain forisomes [1], fusiform giant macromolecular conglomerates up to a 100 microns in length [2] without—tailless—or with “tails” at the tips—tailed—[24]. Forisomes seem to be composed of spindle-like subunits [5] designated as forisomettes [6]. Their substructure exhibits a matrix of regularly ordered building blocks e.g. [1, 5, 7] which are largely of a proteinaceous nature. These proteins belong to the SEO family [8] which is presumably involved in sieve-element occlusion (SEO) and seems widespread among dicotyledons [913].

The capacity to sieve-element occlusion is reflected by a striking property of forisomes: they are able to disperse and re-condense in response to changes in the sieve-element internal milieu [1]. Studies using isolated forisomes obtained via sophisticated isolation methods [14] visualized the dimensions of these conformation changes [15, 16]. Ca2+ induced a 3 to 6-fold volume increase within 10 to 15 seconds [14, 16] (S4 Fig). Isolation of forisomes further enabled in vitro testing of their biotechnological potential for microfluidics [16] and other applications [17]. In this way, the occlusion ability of dispersed forisomes has been demonstrated using silicon-based tubular systems with sieve-element diameters [16].

Both in situ [1] as well as in vitro [14, 16] forisomes disperse in response to Ca2+ supply. After withdrawal of Ca2+-ions, forisomes resume the original, condensed conformation [1, 14, 16, 18]. The tails do not react to Ca2+ supply and, therefore, may be of a different composition or architecture [8, 13, 19]. The reversible conformation changes reflect forisome behaviour in response to sieve-element injury [20, 21]. In case of damage, apoplasmic Ca2+-ions are flooding the sieve-element lumen, which results in sieve-plate occlusion by forisome dispersion probably in combination with protein agglutination [20]. Abrupt and massive Ca2+ influx into lumina of intact sieve elements can also be achieved by vigorous stimuli [2022]. Local cold shocks [22] and remote heat shocks [20, 21] associated with abrupt Ca2+ influx bring about forisome dispersion.

Forisome dispersion is taken as an indication that Ca2+ has surpassed an activation threshold inside sieve elements after wounding or during passage of electropotential waves [21, 23, 24]. It appears that forisome only disperse if the Ca2+-concentration inside the sieve element rises above 50 μmol [21]. In intact sieve elements, such high Ca2+-concentrations may only be reached at certain hotspots. Calcium hotspots may be created by a concerted gating of voltage-dependent and mechano-sensitive Ca2+-permeable channels in the sieve-element plasma membrane and Ca2+-dependent Ca2+-channels at the membranes of the ER-stacks [23, 25]. This proposal is in agreement with the observation that forisomes are most reactive to distant stimuli in sieve-element areas where Ca2+ channels are densely aggregated e.g. in the vicinity of the sieve plates.

There is some evidence that the forisome tips are in contact with or even attached to sieve-element membrane structures [21], which opens the interesting perspective that they may be kept in place near Ca2+-hotspots for optimal functioning This idea concurs with the observation that the forisome position is an important determinant of reactivity. Forisomes with both tips in contact with the plasma membrane region dispersed to distant heat shocks to a greater extent than those with one tip or neither of them against the plasma membrane [21]. Supportive to potential forisome anchoring is that several forisomes are located at the sieve plate at the upstream, apical end of the sieve element [15] and not at the downstream, basal end as to be expected for free-floating forisomes dragged along with mass flow.

In the first place, therefore, the reactivity of upstream Vicia forisomes in different positions to distant stimuli needed further verification by situ examination in intact plants. Furthermore, location, position, and reactivity to distant heat shocks (expressed as dispersion) of the tailless forisomes in Vicia faba (broadbean) and the tailed forisomes in Phaseolus vulgaris (bushbean) were compared using water immersion objectives in confocal laser scanning microscopy in the transmission mode. Finally, we studied positional changes and eventual mobility of both forisome types in response to distant heat shocks.

Material and Methods

Plant Material

Plants of Vicia faba–broadbean—cv Witkiem major (Nunhems Zaden BV, Haelen, The Netherlands) and Phaseolus vulgaris–bushbean—cv Hildora (Hild, Marbach, Germany) were cultivated in pots in a greenhouse at temperature varying between 20–30°C, 60–70% relative humidity, and a 14/10 h light/dark regime. Supplementary lighting (model SONT Agro 400 W; Philips Eindhoven, the Netherlands) was used to give an irradiance level of 200–250 μmol-2 sec-1 at the plant apex. Plants were used 4–7 weeks after germination, in the vegetative phase just before flowering. The plants were transferred in protective boxes from the greenhouse to the laboratory the day before the experiments and stayed there overnight at room temperature of about 20°C to standardize the conditions.

Preparation of intact plants for observation by confocal laser scanning microscopy

For in vivo observation of sieve tubes, cortical cell layers were removed to create an observation window on the phloem at the abaxial side of the main vein of a mature leaf still attached to the intact plant. The cortical tissue was locally removed by manual paradermal slicing with a fresh razor blade, while avoiding damage to the phloem [26]. The bare-lying tissue was immediately submersed in bathing medium containing 2 mol m-3 KCl, 1 mol m-3 CaCl2, 1 mol m-3 MgCl2, 50 mol m-3 mannitol and 2.5 mol m-3 MES/NaOH buffer, pH 5.7. The leaf was fixed onto a microscope slide with two strips of double-sided adhesive tape and mounted on the stage of a confocal laser scanning microscope. After manipulation of the plants, a resting period of at least one hour was taken to restore the ground situation in order to avoid visual misinterpretations. Intactness of sieve elements was checked microscopically via a water immersion objective in the dipping mode by verifying the presence of condensed forisomes. After verification of phloem viability, the leaf tip was burnt carefully by a focused lighter flame, which caused a heat shock. The distance between leaf tip and observation window was 3–4 cm. So-called “priming” of Phaseolus vulgaris plants was elicited by delicate rubbing of the leaf tip to be burnt between two finger tips without causing damage.

The actin-polymerization inhibitor latrunculin A (LatA) was prepared as a stock solution dissolved in DMSO and diluted in bathing medium to give a final working concentrations of 500 nM LatA in 0.5mM (0.01% [v/v]) DMSO. LatA was purchased from Invitrogen (Carlsbad, California, USA).

Confocal laser scanning microscopy

Forisomes and phloem tissue were imaged by confocal laser scanning microscopy using a Leica TCS 4D and SP2 (Leica Microsystems, Heidelberg, Germany) in the black-and-white mode. The phloem tissues were observed using a water immersion objective (HCX APO L40x0.80 W U-V-l objective, Leica, Heidelberg, Germany) in the dipping mode. Digital images were processed with Adobe® PhotoShop to optimize brightness and contrast.

Statistical analyses

Statistical analyses were carried out using SPSS® (IBM® SPSS® Statistics 22). Statistical significance of the differences in forisome location and position were examined by Chi-Square tests with Bonferroni correction. Significance level is defined as the two-sided asympototic significance of chi-square statistics (p<0.05). Forisome reactivity was examined by one-way analysis of variance (ANOVA) with a Bonferroni correction to increase the rigorosity of significance [27] and given as the Welch’s F ratio (p<0.05). The correlation between forisome dispersion and re-condensation times was evaluated by use of Pearson product-moment correlation coefficient, which is considered significant if p<0.05.


Location and position of tailless and tailed forisomes in sieve elements

In this study, location and position of more than 2000 forisomes in at least 500 intact plants were screened. According to the nomenclature adopted here, forisomes are located in basal, central and apical parts of sieve elements (Fig 1A) in the 4 following positions (Fig 1B): one forisome tip is in contact with the plasma membrane section lining the sieve plate, the other with the parietal plasma membrane of the sieve element (position 1), one forisome tip is in contact with the plasma membrane section lining the sieve plate, the other tip is free-floating (position 2), one tip is in contact with the parietal plasma membrane, the other is free-floating (position 3), or no apparent contacts between forisome tips and plasma membrane (position 4).

Fig 1. Forisome locations and positions inside sieve elements.

A. Basal (downstream), central and apical (upstream) locations. B. Position1, Forisome tips in contact with the plasma membrane lining the sieve plate and the parietal plasma membrane, position 2 one tip in contact with the plasma membrane lining the sieve plate, the other is free-floating in the sieve-element lumen, position 3 one tip in contact with the parietal plasma membrane, position, the other is free-floating in the sieve-element lumen, position 4 no apparent tip contacts.

The location of tailless (Vicia faba) and tailed (Phaseolus vulgaris) forisomes inside the sieve elements was seemingly similar at first sight (Fig 2A and 2B). After statistical treatment, however, the distribution of forisomes was significantly different between the two species (Fig 2A and 2B). In Vicia faba, 18% of the forisomes was not located at the basal, downstream end of the sieve element (Fig 2A), in Phaseolus vulgaris 27.7% (Fig 2B). Within the species forisome distribution was non-random (p<0.05) with the downstream location strongly favoured both in broadbean and in bushbean (Fig 2A and 2B) which confirms an earlier report that forisomes are mostly located at the basal end of Vicia sieve elements [15].

Fig 2. Absolute numbers and percentage of forisome locations (basal, central, apical) in sieve elements of A. Vicia faba and B. Phaseolus vulgaris.

Different letters indicate significant differences (p < 0.05) between locations and plant species.

Positions of forisomes at various locations were determined using Z-stacks of transmission pictures (S1 Fig and Fig 2). The respective positions of forisomes at diverse locations were almost identical in Vicia (Fig 3A–3D) and Phaseolus (Fig 3E–3H). Statistical analysis showed no significant differences of forisome position between Vicia and Phaseolus (p<0.05). Most of the forisomes, both at the downstream and upstream side in Vicia (Fig 3A–3C) and Phaseolus (Fig 3E–3G), were in position 1 followed by the positions 2, 4, and 3, respectively. About 30% of the forisomes, both at the downstream and upstream ends, were in contact with the sieve-element membrane facing the companion cell (Fig 3D and 3H), significantly favouring contact to plasma membrane facing parenchyma cells (p < 0.05).

Fig 3. Forisome positions at diverse sieve-element locations in Vicia faba (A-D) and Phaseolus vulgaris (E-H).

A. and E. basal (downstream); B. and F. central; C. and G. apical (upstream). D. and H. Percentage of forisomes at each location and position. The black areas indicate forisomes in contact with the sieve-element side facing the companion cell. Different letters indicate significant differences between the positions of forisomes.

Dispersion of tailless and tailed forisomes in response to distant heat shocks in intact plants

After application of a distant heat shock, we examined forisome reactivity (expressed as dispersion) at the respective locations and positions (Fig 4). For reasons of brevity, we have omitted images of forisome dispersion and recondensation, which have been amply documented in the literature. In vitro reversibility of forisomes of the present lot was also tested and the volumes of the dispersed foriomes were in the order of the values reported. While the forisomes in Vicia readily dispersed in response to a heat shock after recovery from tissue preparation, forisome dispersion in Phaseolus often required a slight touch of the leaf tip cf. [28] prior to the burning stimulus. Without this mechanical stimulus only 26% of the Phaseolus forisomes dispersed. After the manual “priming touch”, forisomes dispersed dependent on location (20 to 54%; Fig 4B) or position at the downstream end (44 to 64%; Fig 4F). Basal and apical Phaseolus forisomes were significantly (p < 0.05) less dispersive than Vicia forisomes at identical locations (Fig 4A and 4B).

Fig 4. Forisome reactivity (dispersion and re-condensation) in response to a remote heat stimulus at different locations in A. and C. Vicia faba and B. and D. primed Phaseolus vulgaris and different positions at the basal (downstream) location (E. and G. Vicia faba and F. and H. Phaseolus vulgaris.

The stippled line indicates the relationship between dispersion and re-condensation time.

The forisomes at the downstream end showed significantly higher reactivity rates than those at other locations (p<0.05) both in Vicia (Fig 4A) and in Phaseolus (Fig 4B). The average reaction times (time lapse between stimulus and dispersion) were the lowest for basal forisomes in Vicia (Fig 4A and 4B) and apical forisomes in Phaseolus (Fig 4C and 4D), albeit not to a significant extent (p>0.05). A significant correlation neither exists between the dispersion and re-condensation times both in Vicia and in Phaseolus forisomes at all locations (Fig 4C and 4D; correlation coefficient R < 0,005).

Taking the positions of solely the basal forisomes into account (Fig 4E and 4F), Vicia forisomes in the positions 1 and 2 showed significantly higher reactivity rates to heat shocks (p<0.05), followed by those in position 3 and 4, in this order (Fig 4E). This contrasted the situation in Phaseolus, where the forisomes in position 2 were significantly most reactive, followed by those in position 3, 1 and 4 (Fig 4F). A significant effect of the forisome position on the reaction times was not found in both plants (p<0.05). Again, a statistically significant correlation between reaction times and re-condensation times could not be shown (Fig 4G and 4H; correlation coefficient R< 0,005). In conclusion, locations, positions and reactivity patterns of tailed and tailless forisomes were largely similar in Vicia and “primed” Phaseolus plants,

Differential mobility of tailless and tailed forisomes

As for the forisome mobility, we observed a spectacular difference between Vicia and primed Phaseolus plants on the one hand and non-primed Phaseolus plants on the other. In the latter group, forisomes moved through the sieve elements. Another set of plants was screened for further studies on forisome mobility (Fig 5). After dispersion, Vicia forisomes stayed in place and took their initial position after re-condensation despite slight local oscillations (Fig 5B–5E). The same held for dispersed forisomes in “primed” Phaseolus plants (Fig 5M–5O). However, when Phaseolus plants were exposed to a heat shock without prior mechanical “priming”, 40% of the basal forisomes (37 from 95, Table 1) did not disperse, but travelled longer distances to find a new position (Fig 5G–5I) or moved shorter distances before they returned to their initial position (Fig 5J–5L). Apical forisomes never moved (n = 48, Table 1); movement of central forisomes was rarely observed (2 from 34, Table 1).

Fig 5. Reactions and change of forisome position in Vicia faba (A-E) and forisome location in Phaseolus vulgaris sieve elements (F-O) in response to remote heating (forisomes are marked by asterisks).

A. Vicia faba forisome (upstream position 1). B-E Vicia faba forisome (downstream position, position1). B. Initial position at an angle of about 10° to the longitudinal axis, C. Dispersion in response to a remote heat shock, D. Re-condensation, 4 min after the stimulus, position parallel to the longitudinal axis, E. 5 min after stimulus, position at an angle of about 10° to the longitudinal axis. F. Phaseolus vulgaris forisome (upstream position). G-I Long-distance ovement of a non-primed condensed Phaseolus vulgaris forisome from the downstream position 1 to the upstream position 4, J-L Short-distance movement of a non-primed condensed Phaseolus vulgaris forisome from the downstream position 1 (J.), via the central position 4 (K) back to the original downstream position 1 (L). M-O No position change of a dispersed Phaseolus vulgaris forisome. Direction of flow in G to O from right to left.

Table 1. Forisome reactivity in response to remote burning of intact non-primed Phaseolus vulgaris plants as related to its location.

“Reaction” includes both dispersion or movement.

Breaking down the data for the basally located forisomes in Phaseolus (n = 95, Table 1) according to their position demonstrates that almost exclusively those in position 1 reacted to burning by movement (Table 2). The forisomes in other positions (2–4) reacted by dispersion or did not react at all (Table 2). In case of displacement, basal Phaseolus forisomes moved in acropetal direction at an average speed of 24.5 μm/s and returned in basipetal direction at an average speed of 5 μm/s (Fig 6A). As only the slower movements were more suitable for photographic registration, the pictures here present forisomes in slower movement range with a speed of about 2 μm/s (Fig 5F–5I). The reaction times widely varied between the forisomes (Fig 6B). There is a positive correlation between the onset of the movement and the distance covered within the sieve element i.e. longer reaction times are correlated with farther movements (Fig 6B; correlation coefficient R = 0.677, p = 0.01).

Fig 6. A. Average acropetal and subsequent basipetal movement velocities of originally downstream located condensed Phaseolus vulgaris forisomes after a distant heat stimulus in non-primed plants. B. Response lag times of location changes of forisomes in Phaseolus vulgaris and the distances of movement.

The stippled line indicates the relationship between dispersion and re-condensation time.

Table 2. Reactivity of basal forisomes in response to remote burning of intact Phaseolus vulgaris plants as related to its position.

“Reaction” includes both dispersion and movement.

Effects of Latrunculin A on forisome conformation and movement

Mobility of forisomes may be associated with the actin network identified in SEs [29]. To investigate the involvement of actin in forisome displacement, its dispersion and movement were investigated after application of 500 nM Latrunculin A (LatA), an inhibitor of actin filament polymerization (see [29] for interference with sieve-element cytoskeleton action). LatA pretreatment had no effect on the rate of forisome dispersion in response to distant burning in Vicia faba (n = 84, results not shown).

By contrast, LatA pre-treatment considerably changed the reaction pattern of basal forisomes in response to burning in non-primed Phaseolus vulgaris plants (Tables 2 and 3). LatA pretreatment had a severe impact on the capacity of movement. None of the basal forisomes (n = 74) did move after burning (Table 3), but dispersed instead (Fig 7A–7C). The dispersion percentages reached even higher values than those obtained with primed control plants (cf. Fig 4F). As with other dispersion reactions, there was no correlation between reaction and re-condensation times of these forisomes (Fig 7D; correlation coefficient R< 0,005).

Fig 7. A-C Dispersion and re-condensation of a forisome in a LatA pretreated non-primed Phaseolus vulgarisplant in response to a remote heat stimulus. D. Relationship between dispersion and re-condensation times.

The stippled line indicates the relationship between dispersion and re-condensation time.

Table 3. Reactivity of basal forisomes pretreated with LatA in response to remote burning of intact non-primed Phaseolus vulgaris plants as related to its position, without prior priming.

“Reaction” includes both dispersion and movement.

In control experiments, dimethyl sulfoxide (DMSO) at the concentrations used here to dissolve LatA had no effect on forisome conformation changes (Vicia faba n = 6; Phaseolus vulgaris n = 12, results not shown).


Relationship between location or position and forisome dispersion responsiveness

The distribution (i.e. the locations) of forisomes is quantitatively similar, but significantly different in sieve elements of Vicia faba and Phaseolus vulgaris (Fig 3). The positions of tailless and tailed forisomes at the respective locations are almost identical (Fig 3). About 30% of the forisomes are in contact with the sieve-element side facing the companion cell (Fig 3). Since the common interface between sieve element and companion cell amounts about 30% of the total sieve-element inner surface in transport phloem [30, 31], there seems no preference for forisome contacts with the sieve-element companion cell interface.

Visual inaccuracies due to using the confocal technique in the transmission mode prevented us to precisely observe if the forisome was docked onto a membrane structure. The significantly differential reactivity of forisomes in various positions (Fig 4) however demonstrates that the position assessment was largely correct. We refrained from employing stained forisomes (a CMEDA/CMFDA mixture; [20, 21]) which might have provided sharper images, because they displayed an aberrant dispersion behaviour in preliminary experiments. Thus, attachment or anchoring to membranes could not be established with certainty by microscopy. Therefore, we avoided the term “attachment” and preferred the vaguer term “contact”.

Reactivity of forisomes to remote stimuli has been related to the amount of Ca2+ released into the sieve-element lumen via Ca2+-permeable channels during passage of an electropotential wave [23]. The reactivity (expressed as the dispersion rate) of the basally located forisomes exceeds that of the others (Fig 4E and 4F) which concurs with a clustering of Ca2+-permeable channels in the sieve-plate area [21, 23]. However, it does not explain why the forisomes located at the upstream end are much less reactive (Fig 4E and 4F), although forisomes at the sieve-element ends are both positioned near a sieve plate. It infers that the Ca2+-permeable channels are more densely aggregated at the downstream side of the sieve plate which finds confirmation in pictures, where Ca2+ channels inside the sieve tubes are stained by DM-BODIPY-DHP (lower inset Fig 4J in [21]). As exemplified by the basally located forisomes (Fig 4E and 4F), contacts with sieve-element membrane structures are decisive for the degree of forisome reactivity. This renders credence to the view that Ca2+ hotspots in the vicinity of forisome tips are required for dispersion [23].

Dispersion and re-condensation times do not appear correlated as observed in a less detailed study [21]. This is logical given the fact that dispersion and re-condensation depend on different, only weakly interdependent processes [24]. Dispersion is due to gating of Ca2+-permeable channels [21, 23], whereas re-condensation presumably relies on the activity of Ca2+-ATPases [32, 33].

Arguments in favour of actin-mediated mobility of bushbean forisomes

About 20 to 30% of the forisomes are not located at the downstream end of sieve elements (Figs 2 and 3). This seems inconsistent with the view that forisomes are free-floating and, hence, would be pressed against the downstream sieve plate by mass flow. One may argue, however, that free-floating forisomes dragged along by mass flow get stuck behind obstacles in the mictoplasm [15] which might explain the forisome presence throughout the sieve element. This interpretation may hold for a few forisomes—the anchoring does not seem excessively tight—but does not explain the uneven forisome distribution over the sieve element [15] (Figs 2 and 3). The preference for the upstream location in comparison with the paucity of forisomes in the central region (Fig 2) speaks for some mode of attachment, unless preferential forisome residence near sieve plates results from opposite fluxes in sieve tubes. However, phloem mass flow in apical direction is unlikely in petioles of source leaves.

All in all, upstream localization seems to indicate that forisomes are “anchored” in some way, preferably in the vicinity of Ca2+ hotspots [24]. Some form of forisome attachment is corroborated by forisome displacement in Phaseolus in response to remote heating (Fig 5). Forisome movement vice versa points to a mechanism for active forisome movement which requires anchoring to structural sieve-element components. As forisomes themselves do not seem to have the instruments for mobility, movement may depend on an interaction between forisomes and sieve-element components via structural links. These connections are likely coupled to the sieve-element cytoskeleton as shown by the full inhibition of forisome displacement by LatA (Tables 13).

One could argue that the reversible mobility of bushbean forisomes is purely passive due to reverse fluid movements—pressure followed by retraction—induced by distant heating. Purely passive fluid-propelled movement of bushbean forisomes in response to heat stimuli, however, is incompatible with the immobility of broadbean forisomes in response to heat pulses and the elimination of the mobility of bushbean forisomes in the presence of LatA. In conclusion, the differential mobility of bushbean and broadbean forismes is inconsistent with passive massive-flow driven movement.

Admittedly, the presence of a cytoskeleton in SEs is a matter of debate. Electron microscopic studies in the 20th century could not identify a cytoskeletal system in sieve elements [34]. It is hard to conceive, however, how the intense macromolecular exchange between SEs and CCs [35] can take place without a cytoskeleton crossing the Plasmodesm-pore units (PPUs) s like in other plasmodesmata [36]. A cytoskeleton extending from CCs into SEs would also provide a conveyer belt for sorting macromolecules inside the SE for local and distant destinations [37]. An actin network in SEs was visualized by fluorescent phalloidin and immunocytochemistry [29]. Conversely, Cayla et al. [38] could not find actin filaments in SEs of Arabidopsis, but they do not exclude that the fluorescent fABD2:GFP constructs expressed in CCs are too large to pass PPUs or that the density of actin filaments in SEs is low. A coarse-meshed network of thin actin filaments in SEs might indeed explain their apparent absence in EM pictures.

Questions regarding presumptive forisome anchoring and mobility

Despite the first glimpses on their remarkable behaviour, the nature of dispersion and mobility of Phaseolus forisomes is fraught with questions that remain to be answered:

  1. 1. It is unclear as how forisomes find their original position after a stimulus. Tailless forisomes that are stimulated below a certain Ca2+ threshold do not disperse, but start wiggling for a while [23]. A small change of orientation happens to dispersing Vicia forisomes (Fig 5A–5E). Both movements of broadbean forisomes may be interpreted as a release of anchors which are not connected with a frame for movement. Subsequent to their reaction, Vicia forisomes are able to regain their original position (Fig 5A–5E). The same applies to Phaseolus forisomes, the return of which to the original position must be more complex than in Vicia given the extensive longitudinal movement (Fig 5F–5O). Phaseolus forisomes may be anchored to a motive apparatus, perhaps via the tails, which have a SEO composition distinct from the rest of the forisome body [8, 13]. Forisome tails do not disperse in response to Ca2+ supply (Fig 1) [19] and are only weakly fluorescent in forisomes in which certain SEO-proteins are GFP-tagged [8].
  2. 2. Once Vicia forisomes have reached the condensed state following the preparation procedure, they are ready to disperse in response to remote heat shocks. By contrast, Phaseolus forisomes require a precedent slight mechanical shock to become much more reactive in terms of dispersion. On the other hand, heat shocks induce forisome movement in non-primed Phaseolus plants. It appears that both dispersion and movement require input of Ca2+ ions. The crux for the differential behaviour of bushbean forisomes may lie in the availability of Ca2+ and competition for Ca2+ by respective binding sites. Application of LatA fully impairs longitudinal movement of tailed condensed forisomes, which disperse at the rates usual for non-moving forisomes when the capacity of movement is blocked. This behaviour may be interpreted in terms of competition. If Ca2+ cannot be invested into movement due to the LatA-induced inactivity of actin, it is available for alternative binding sites needed for dispersion. The interpretation also implies that the Ca2+ threshold for activation of forisome movement exceeds that needed for triggering of dispersion.
  3. 3. The manual touch as an inductor of “priming” may provoke electropotential waves akin to touch-triggered [28, 39, 40] and vibration-generated [41] action potentials, which are associated with Ca2+ influx into sieve tubes along the pathway. They likely prepare the plant to potential attacks of herbivores [41].

The “priming cue” may actually suppress Ca2+ influx after a heat shock. The slight touch may evoke an electropotential wave causing a suboptimal Ca2+ influx. A second electropotential wave triggered by the heat shock during the refractive period of the touch-induced action potential [42, 43] may cause less Ca2+ influx than in “non-primed” plants. As an effect of “priming”, the Ca2+ level may become sufficient for dispersion, but falls short to reach the threshold for forisome movement. A critical Ca2+ level for forisome movement also explains why merely tailed down-stream forisomes (39% of 95 forisomes) move in upstream direction, while a few forisomes located at the centre move acropetally (6% of 34 forisomes) and none of the apical forisomes did (0% of 48 forisomes). As argued above, there may be an base-to-apex Ca2+ gradient in sieve elements [21] as exemplified by forisome reactivity (Fig 4) after passage of an electropotential wave.

  1. 4. At first glance, it is puzzling, why the effect of LatA application on forisome dispersion differs between distant heat shocks (control level of dispersion with LatA, this paper) or local cold shocks (suppression of dispersion with LatA, [29]). On second thoughts, the reason may be obvious: in cold shocks gating of Ca2+ channels is strongly modulated by actin [29], whereas actin is less involved in the long-distance propagation of electropotential waves [24].
  2. 5. Temporary occlusion of the sieve tube turned out to be a major function of forisome dispersion in intact sieve elements [16, 22, 24]. In this frame, it is hard to conceive what the function of non-dispersed forisome movement could be.

Is forisome mobility related to the plant species or to presence of tails?

Since the present experiments have been carried out with just two species for reasons of work-load reduction, the question arises as whether forisome mobility is a species-specific trait or is rather associated with the presence of tails. In contrast to the forisome body that contains SEO subgroup 1 proteins [44], forisome tails are composed of proteins of the SEO subgroup 2 [13, 44] and do not disperse in response to Ca2+ [8, 19]. They might react to Ca2+ supply in another manner. It has been speculated that forisomes detach from anchoring sites under the control of Ca2+ during dispersion [23, 24]. On top of that, Ca2+ might be involved in forisome movement along actin filaments. On the basis of provisional circumstantial evidence, we believe that tails are decisive for forisome movement.

Supporting Information

S1 Fig. A-I. Z-stack illustration of Phaseolus vulgaris phloem tissue using the CLSM transmission mode shows a series of 2μm sections.

The sieve tube diameter is approximately 18μm. Direction of flow is from right to left. A. SE = sieve element; SP = sieve plate; CC = companion cell. A,B. The forisome is marked by an asterisk, C. The tails are marked by arrowheads. Scale bar = 10μm.


S2 Fig. A-E. Z-stack illustration of Vicia faba phloem tissue using CLSM the transmission mode shows a series of 4μm sections.

The sieve tube diameter is approximately 20μm. Direction of flow is from right to left. A. SE = sieve element; SP = sieve plate; CC = companion cell. B. The forisome is marked by an asterisk. Scale bar = 10μm.


S3 Fig. Movie showing movement of a tailed forisome in Phaseolus vulgaris in the middle of the screen after a distant heat shock.


S4 Fig. Isolated tailless (A,B) and tailed (C,D) forisomes (without Ca2+ A and C; with Ca2+ supply B and D).

A,B Vicia faba, C,D Phaseolus vulgaris. Isolation of forisomes: Forisomes were isolated from Vicia faba and Phaseolus vulgaris phloem tissue according to Knoblauch et al. 2003. Isolation media containing 10 mM Tris (pH 7.3), 50 mM potassium chloride and ethylene-diaminetetraacetic acid (EDTA) concentrations of 2 or 10 mM were previously outgassed and covered with argon gas. 1 mM sodium sulfite was added to the isolation medium to suppress oxygen effects on forisomes. The cortex of the stems of 4- to 7-week-old Vicia faba or Phaseolus vulgaris plants was carefully pulled off, and phloem was scraped off with a scalpel. The phloem shreds were transferred to 2 ml of forisome isolation medium containing 10 mM EDTA. After 30 min of incubation, the phloem material was homogenized in liquid nitrogen and transferred to 4 ml of 2 mM EDTA solution. After filtration of solubilized plant material through a 60 μm mesh filter freshly isolated forisomes were used for in vitro studies. To observe forisome reactions, a drop of 10 μl 2mM EDTA solution containing forisomes was transferred to a microscope slide. After having a spindle shape forisome in focus this drop was exchanged successively against 10 mM calcium chloride solution to induce forisome dispersion.



The authors thank Dietmar Haffer (JLU Gießen) and Daniel Dittrich for technical assistance. This work was supported by the Justus Liebig University by a ‘Just’us postdoc grant’ and a ‘Margarete Bieber grant’.

Author Contributions

Conceived and designed the experiments: AJEvB ACUF SVB. Performed the experiments: SVB ACUF. Contributed reagents/materials/analysis tools: AJEvB ACUF. Wrote the paper: AJEvB ACUF and SVB.


  1. 1. Knoblauch M, Peters WS, Ehlers K, van Bel AJE (2001) Reversible calcium-regulated stopcocks in legume sieve tubes. Plant Cell 13: 1221–1230. pmid:11340193
  2. 2. Peters WS, Haffer D, Hanakam CB, van Bel AJE, Knoblauch M (2010) Legume phylogeny and the evolution of a unique contractile apparatus that regulates phloem transport. American Journal of Botany 97: 797–808.
  3. 3. Lawton DM (1978a) Ultrastructural comparison of the tailed and tailless P-protein crystals respectively of runner bean (Phaseolus multiflorus) and garden pea (Pisum sativum) with tilting stage electron microscopy. Protoplasma 97: 1–11.
  4. 4. Lawton DM (1978b) P-protein crystals do not disperse in uninjured sieve elements in roots of runner bean (Phaseolus multiflorus) fixed with glutaraldehyde. Annals of Botany 42: 353–361.
  5. 5. Schwan S, Menzel M, Fritzsche M, Heilmann A, Spohn U (2009) Micromechanical measurements of P-protein aggregates (forisomes) from Vicia faba plants. Biophysical Biochemistry 139: 99–105.
  6. 6. Tuteja N, Umate P, van Bel AJE (2010) Forisomes: calcium-powered protein complexes with potential as ‘smart’ biomaterials. Trends in Biotechnology 28: 102–110. pmid:20004992
  7. 7. Müller B, Groscurth S, Menzel M, Rüping BA, Twyman RM, Prüfer D, et al. (2014) Molecular and ultrastructural analysis of forisome subunits reveals the principles of forisome assembly. Annals of Botany 113: 1121–1137. pmid:24694827
  8. 8. Pélissier HC, Peters WS, Collier R, van Bel AJE, Knoblauch M (2008) GFP tagging of sieve element occlusion (SEO) proteins in green fluorescent forisomes. Plant and Cell Physiology 49: 1699–1710. pmid:18784195
  9. 9. Rüping B, Ernst AM, Jekat SB, Nordzieke S, Reineke AR, Müller B, et al. (2010) Molecular and phylogenetic characterization of the sieve element occlusion gene family in Fabaceae and non-Fabaceae plants. BMC Plant Biology 10: 219. pmid:20932300
  10. 10. Anstead JA, Froehlich DR, Knoblauch M, Thompson GA (2012) Arabidopsis P-protein filament formation requires both AtSEO1 and AtSEO2. Plant and Cell Physiology 53: 1033–1042. pmid:22470058
  11. 11. Ernst AM, Jekat SB, Zielonka S, Müller B, Neumann U, Rüping B, et al. (2012) Sieve element occlusion (SEO) genes encode structural phloem proteins involved in wound sealing of phloem. Proceedings of the National Academy of Sciences USA 109: E1980–E1989.
  12. 12. Jekat SB, Ernst A, Zielonka S, Noll G, Prüfer D (2012) Interactions among tobacco sieve element occlusion (SEO) proteins. Plant Signaling Behavior 7: 1918–1920.
  13. 13. Zielonka S, Ernst AM, Hawat S, Twyman RM, Prüfer D, Noll GA (2014) Characterization of five subgroups of the sieve element occlusion gene family in Glycine max reveals genes encoding non-forisome P-proteins, forisomes and forisome tails. Plant Molecular Biology 86: 51–67. pmid:24928491
  14. 14. Knoblauch M, Noll G, Müller T, Prüfer D, Schneider-Hüther I, Scharner D, et al. (2003) ATP-independent contractile proteins from plants. Nature Materials 2: 600–603. pmid:12942070
  15. 15. Peters WS, van Bel AJE, Knoblauch M (2006) The geometry of the forisome-sieve element-sieve plate complex in the phloem of Vicia faba L. leaflets. Journal of Experimental Botany 57: 3091–3098. pmid:16882644
  16. 16. Knoblauch M, Stubenrauch M, van Bel AJE, Peters WS (2012) Forisome performance in artificial sieve tubes. Plant Cell & Environment 8: 1419–1427.
  17. 17. Groscurth S, Müller B, Schwan S, Menzel M, Diekstall F, Senft M, et al. (2012) Artificial forisomes are ideal models of forisome assembly and activity that allow the development of technical devices. Biomacromolecules 13: 3076–3086. pmid:22963540
  18. 18. Will T, Tjallingii WF, Thönnessen A, van Bel AJE (2007) Molecular sabotage of plant defense by aphid saliva. Proceedings of the National Academy of Sciences USA 104: 10536–10541.
  19. 19. Peters WS, Knoblauch M, Warmann SA, Schnetter R, Shen AQ, Pickard WF (2007) Tailed forisomes of Canavalia gladiata: a new model to study Ca2+-driven protein contractility. Annals of Botany 100: 101–109. pmid:17584851
  20. 20. Furch ACU, Hafke JB, Schulz A, van Bel AJE (2007) Ca2+-mediated remote control of reversible sieve tube occlusion in Vicia faba. Journal of Experimental Botany 58: 2827–2838. pmid:17615409
  21. 21. Furch ACU, van Bel AJE, Fricker MD, Felle HH, Fuchs M, Hafke JB (2009). Sieve element Ca2+ channels as relay stations between remote stimulus and sieve tube occlusion. Plant Cell 21: 2118–2132. pmid:19602624
  22. 22. Thorpe MR, Furch ACU, Minchin PEH, Föller J, van Bel AJE, Hafke JB (2010) Rapid cooling triggers forisome dispersion just before phloem transport stops. Plant Cell & Environment 33: 259–271. pmid:19930129
  23. 23. Hafke JB, Furch ACU, Fricker MD, van Bel AJE (2009) Forisome dispersion in Vicia faba is triggered by Ca2+ hotspots created by concerted action of diverse Ca2+ channels in sieve elements. Plant Signaling Behavior 4: 968–972. pmid:19826217
  24. 24. van Bel AJE, Furch ACU, Will T, Buxa SV, Musetti R, Hafke JB (2014) Spread the news: systemic dissemination and local impact of Ca2+ signals along the phloem pathway. Journal of Experimental Botany 65: 1761–1787. pmid:24482370
  25. 25. Hafke JB, van Bel AJE (2013) Cellular basis of electrical potential waves along the phloem and impact of coincident Ca2+ fluxes. In: Thompson GA and van Bel AJE, editors. Phloem: molecular cell biology, systemic communication, biotic interactions. Wiley-Blackwell, Cichester. pp. 122–140.
  26. 26. Knoblauch M, van Bel AJE (1998) Sieve tubes in action. Plant Cell 10: 35–50.
  27. 27. Cabin RJ, Mitchell RJ (2000) To Bonferroni or not to Bonferroni: when and how are the questions. Bulletin of the Ecological Society of America 81, 246–248.
  28. 28. Degli Acosta R (2014) Touch-induced action potentials in Arabidopsis thaliana. Archives des Sciences 67: 125–138.
  29. 29. Hafke JB, Ehlers K, Föller J, Höll SR, Becker S, van Bel AJE (2013) Involvement of the sieve-element cytoskeleton in electrical responses to cold shocks. Plant Physiology 162: 707–719. pmid:23624858
  30. 30. Van Bel AJE (1996) Interaction between sieve element and companion cell and the consequences for photoassimilate distribution. Two structural hardware frames with associated physiological software packages in dicotyledons? Journal of Experimental Botany 47: 1129–1140. pmid:21245242
  31. 31. Van Bel AJE (2003) The phloem, a miracle of ingenuity. Plant Cell & Environment 26: 125–150.
  32. 32. Kudla J, Batistic O, Hashimoto K (2010) Calcium signals: the lead currency of plant information processing. Plant Cell 22: 541–563. pmid:20354197
  33. 33. Huda KMK, Banu MSA, Tuteja R, Tuteja N (2013) Global calcium transducer P-type Ca2+-ATPases open new avenues for agriculture by regulating stress signalling. Journal of Experimental Botany 64: 3099–3109. pmid:23918957
  34. 34. Behnke HD, Sjolund RD (1990) Sieve elements. Comparative structure, induction and development. Berlin: Springer Verlag. 305 p.
  35. 35. Lucas WJ, Ham BK, Kim JY (2009) Plasmodesmata–bridging the gap between neighbouring plant cells. Trends in Cell Biology 19: 495–503. pmid:19748270
  36. 36. Kumar E, Kumar D, Hyun TK, Kim JY (2015) Players at plasmodesmal nanochannels. Journal of Plant Biology 58: 75–86.
  37. 37. Van Bel AJE, Furch ACU, Hafke JB, Knoblauch M, Patrick JW (2011) (Questions)n on phloem biology. 2. Mass flow, molecular hopping, distribution patterns and macromolecular signalling. Plant Science 181: 325–330. pmid:21889037
  38. 38. Cayla T, Batailler B, Le Hir R, Revers F, Anstead JA, Thompson GA, et al. (2015) Live imaging of companion cells and sieve elements in Arabidopsis leaves. PloS ONE 10(2): e0118122. pmid:25714357
  39. 39. Braam JB (2005) In touch: plant responses to mechanical stimuli. New Phytologist 165: 373–389. pmid:15720650
  40. 40. Monshausen GB, Gilroy S (2009) Feeling green: mechanosensing in plants. Trends in Plant Science 19: 228–235.
  41. 41. Appel HM, Cocroft RB (2014) Plants respond to leaf vibrations caused by insect herbivore chewing. Oecologia 175: 1257–1266. pmid:24985883
  42. 42. Zawadski T, Davies E, Dziubinska H, Trebacz K (1991) Characteristics of action potentials in Helianthus annuus. Physiologia Plantarum 83: 601–604.
  43. 43. Fromm J, Lautner S (2007) Electrical signals and their physiological significance in plants. Plant Cell & Environment 30: 249–257. pmid:17263772
  44. 44. Ernst AM, Rüping B, Jekat SB, Nordzieke S, Reineke AR, Müller B, et al. (2011) The sieve element occlusion gene family in dicotyledonous plants. Plant Signaling Behavior. 6: 151–153. pmid:21422825