Modeling the parameters for plasmodesmal sugar filtering in active symplasmic phloem loaders

Abstract

Plasmodesmata (PD) play a key role in loading of sugars into the phloem. In plant species that employ the so-called active symplasmic loading strategy, sucrose that diffuses into their unique intermediary cells (ICs) is converted into sugar oligomers. According to the prevalent hypothesis, the oligomers are too large to pass back through PD on the bundle sheath side, but can pass on into the sieve element to be transported in the phloem. Here, we investigate if the PD at the bundle sheath-IC interface can indeed fulfill the function of blocking transport of sugar oligomers while still enabling efficient diffusion of sucrose. Hindrance factors are derived via theoretical modeling for different PD substructure configurations: sub-nano channels, slit, and hydrogel. The results suggest that a strong discrimination could only be realized when the PD opening is almost as small as the sugar oligomers. In order to find model parameters that match the in vivo situation, we measured the effective diffusion coefficient across the interface in question in Cucurbita pepo with 3D-photoactivation microscopy. Calculations indicate that a PD substructure of several sub-nano channels with a radius around 7 Å, a 10.4 Å-wide slit or a hydrogel with 49% polymer fraction would be compatible with the effective diffusion coefficient. If these configurations can accommodate sufficient flux of sucrose into the IC, while blocking raffinose and stachyose movement was assessed using literature data. While the slit-configuration would efficiently prevent the sugar oligomers from "leaking" from the IC, none of the configurations could enable a diffusion-driven sucrose flux that matches the reported rates at a physiologically relevant concentration potential. The presented data provides a first insight on how the substructure of PD could enable selective transport, but indicates that additional factors are involved in efficient phloem loading in active symplasmic loading species.

Keywords: carbon allocation; hindered diffusion; phloem loading; plasmodesmata; polymer trap.

Figure 1

Schematic representations of hypothetical plasmodesmata substructure. (A) Nano channel model proposed by Ding et al. (1992). (B) Cytoplasmic sleeve model proposed by Waigmann et al. (1997). (C,D) Cytoplasmic sleeve filled by polymer-meshwork hydrogel. Light gray—cell wall, black—plasma membrane, dark gray—proteins, green—desmotuble.

Figure 2

Logarithmic scale plots of diffusion hindrance in relation to hypothetical plasmodesmata substructure configurations. Hindrance factor for sucrose, raffinose, and stachyose increases faster with smaller dimensions in single plasmodesmal channels (A) than in a cytoplasmic sleeve (B). Hindrance in hydrogels (C) is moderate, even at high volume fractions.

Figure 3

Plot of effective diffusion coefficient for diffusion of fluorescein through plasmodesmata with sub-nano channel configuration. The curve shows at which combination of number of channels and channel radius the experimentally determined value is matched.

Figure 4

Logarithmic scale plot of sugar flux and concentration potential between bundle sheath cells and intermediary cells for different plasmodesmata configurations. (A) Sub-nano channel configuration assuming 9 channels with 6.5 Å radius. (B) Slit configuration assuming half slit width of 5.2 Å. (C) Hydrogel configuration assuming a polymer volume fraction of 0.49. Vertical lines indicate sugar concentration potentials as provided by Haritatos et al. (1996). Horizontal lines indicate sucrose flux values provided by Schmitz et al. (, gray) and a flux range based on values listed in Table 2 (blue). Raffinose flux is strongly hindered in the slit model in contrast to the other configurations, which would enable considerable “leakage” out of the intermediary cell. Hindrance of sucrose flux at concentration potentials observed by Haritatos et al. (1996) would in all models be orders of magnitude too high to realize flux rates reported by Schmitz et al. (1987) and other authors.

Figure A1

Confocal image of a photoactivation experiment in a Cucurbita pepo leaf. The image shows a single optical section of a 30 μm thick z-stack that was recorded at one time point during the 10 s-long photoactivation phase. The fluorescence of the activated tracer is displayed as extended focus projection in green, overlaid over the brightfield image. In image analysis, three-dimensional regions of interest are drawn adjacent to the cell wall between the bundle sheath target cell (T) and the intermediary cell (*) in order to quantify diffusion of activated tracer. Scale bar: 10 μm.