Plasmodesmata transport of GFP and GFP fusions requires little energy and transitions during leaf expansion

Abstract

Plasmodesmata (Pd) are symplastic channels between neighboring plant cells and are key in plant cell-cell signaling. Viruses of proteins, nucleic acids, and a wide range of signaling macromolecules move across Pd. Protein transport Pd is regulated by development and biotic signals. Recent investigations utilizing the Arrhenius equation or Coefficient of conductivity showed that fundamental energetic measurements used to describe transport of proteins across membrane pores or the nuclear pore can also apply to protein movement across Pd. As leaves continue to expand, Pd transport of proteins declines which may result from changes in cell volume, Pd density, or Pd structure.

Addendum to: Schoenknecht G, Brown JE, Verchot-Lubicz J. Plasmodesmata transport of GFP alone or fused to potato virus X TGBp1 is diffusion driven Protoplasma 2008; 232:143-52.

Researchers have argued for the last decade that movement of proteins and other macromolecules across Pd is regulated by development, stress and biotic signals. There are four current models describing different mechanisms of Pd transport. First is the non cell autonomous protein (NCAP) pathway that carries ribonucleoprotein complexes across Pd. NCAPs often carry RNA in a ribonucleoprotein complex to the Pd.1–4 This mode of transport involved targeted movement, meaning that a set of proteins must dock within the Pd to gate it open to enable transfer between cells. Proteins which are normally too large to move across Pd can gate open Pd to enable its own transfer into neighboring cells. This is contrasted by nontargetted movement, which is passive movement of proteins that are sufficiently small enough to pass between cells.5,6 The green fluorescent protein (GFP) has been described as a protein whose movement is non-targeted, meaning that it can diffuse across Pd. Reasons that we do not see continuous movement of small proteins between cells include protein compartmentalization or subcellular targeting signals. For example proteins may be synthesized and modified via the ER and Golgi networks and then transferred into vesicles and transported within cells to their destination. These proteins would not be free in the cytosol for diffusion across Pd. Alternatively, proteins which have dominant subcellular targeting signals which direct them to certain organelles such as the nucleus, peroxisome, or other destination would not be free to move across Pd.5,6 A third model represents proteins in the ER that move laterally along the membrane or through the ER lumen into neighboring cells. This transport is quite rapid and investigations are ongoing to determine how this is regulated.7–11 Finally, there is vesicle transport which deliver cargo to Pd.12,13 The origin of these vesicles is still under investigation. Much more research has been accomplished toward defining non-targeted movement and the NCAP pathway while the ER and vesicle transport pathways are only recently described and very little is known about the regulatory mechanisms underlying these pathways.

Pd permeability is governed in part by architecture, but also by key regulatory factors that determine Pd conductivity. Factors such as mysoin VIII, actin and calreticulin were identified in Pd which likely regulate expansion and contraction.14–19 In addition calcium, ATP and plant hormones can downregulate Pd permeability during development and stress.20,21 The tools for measuring Pd permeability has been to study the transport of fluorescently tagged proteins, fluorescent dextran beads, GFP or GFP fusions following microinjection or biolistic delivery to the cytoplasm of one cell.22–26 Then video imaging or captured still images at select time intervals are used to characterize Pd transport. Until recently researchers quantified movement by the frequency they observed a certain type of movement. Therefore our ability to describe Pd permeability has been limited.

Evidence that ATP impacts Pd conductivity has led investigations to explore the energy requirements for macromolecular transport across Pd. By understanding the energy requirements for transport of various proteins and nucleic acids we can better characterize passive or active transport processes. Toward this end two recent studies detailed quantitative approaches that can be employed to describe the developmental and energy requirements cell-to-cell transport of cytosolic proteins. Both papers used biolistic bombardment to deliver plasmids expressing GFP or GFP fusions to tobacco leaf epidermal cells and then captured still images of GFP fluorescence in neighboring cells. We employed the Arrhenius equation to characterize transport of GFP or GFP fused to the Potato virus X (PVX) TGBp1 movement protein. PVX TGBp1 was selected to compare with GFP alone since it is known to gate open Pd and has ATPase activity.45 We predicted that the abilities of GFP alone and GFP-TGBp1 to move across Pd might be different and were surprised to learn that the energy for transport of both proteins was similar. This project established the principle that GFP and GFP-TGBp1 transport is temperature dependent showing a linear relationship between protein movement and the temperatures at which leaves were incubated.

Green fluorescent sites on bombarded leaves were scored for the movement or no movement. Movement is defined as evidence of fluorescence in 2 or more cells at 24 h and no movement is when fluorescence is in single cells. These were then presented as a percentage of the total. So by digitizing the representation of movement we were able to represent a linear relationship between movement and temperature. Representing movement in this way also enabled us to represent movement values on a logarithm scale necessary for a classic Arrhenius plot. The activation energy (Ea) was calculated by fitting the data to the Arrhenius equation:% movement = A exp(-Ea/RT); and the Ea for GFP and GFP-TGBp1 was approximately 38 kJ/mol and 29 kJ/mol. These low activation energies are comparable to the reported 30 kJ/mol calculated for temperature dependence of protein transport through the cytosol. Evidence that GFP movement across Pd requires slightly more energy than through the cytoplasm suggests there may be some resistance within the pore. The lower energy for GFP-TGBp1 suggests that movement is facilitated, which likely reflects Pd gating by TGBp1, enabling greater transfer between cells.

Liarzi and Epel define a new coefficient of conductivity of Pd.42 This study also concluded that cell-to-cell transport of GFP in nontransgenic or transgenic N. benthamiana plants expressing the Tobacco mosaic virus (TMV) movement protein (MP) is temperature dependent. The method was to measure the exponential decay, which is a measure of the impedance to diffusion driven cell-to-cell movement of fluorescence. The exponential decay factor? was determined by calculating the ratio of GFP fluorescence in bombarded cell 0 and neighboring cells. This was presented as a measure of fluorescence transfer from cell 0 to cell 1 to cell 2. A coefficient for conductivity C(Pd), 1/? for GFP was reflective of diffusion. Interestingly the (TMV) MP did not increase conductivity of GFP between neighboring leaf epidermal cells indicating that movement was already maximal. Considering prior reports that the TMV MP shows preferential spread into mesophyll rather than epidermal tissues during virus infection, it is possible that preferential spread into mesophyll cells would prevent experimental efforts to achieve improved conductivity of GFP between epidermal cells.27,28 In which case the absence of a trans effect of TMV MP on GFP conductivity in the epidermis may not be surprising. In fact, prior investigations of TMV MP gating activities were conducted in mesophyll cells.29,30 The best explanation for the combined studies is that cytosolic GFP can diffuse across Pd , however viral proteins which gate Pd enable their own low energy transfer into neighboring cells without allowing other proteins to flood into neighboring cells. Therefore viral movement proteins, such as PVX TGBp1 and TMV MP, which gate Pd provide themselves with an energetic advantage for transport into neighboring cells which is essential for rapid dissemination of virus into further tissues.

These studies provide an interesting contrast between PVX TGBp1 and TMV MP. Both proteins gate open Pd for virus cell-to-cell transport, but there seems to be differences in how these proteins function in epidermal cells. This is likely due to their different roles in promoting virus cell-to-cell movement. PVX TGBp1 protein is also a suppressor of RNA silencing. We recently proposed a model in which TGBp1 rapidly moves from cell-to-cell ahead of virus infection, to suppress the cell's RNA degradation machinery, as a means to promote infection.31 The TMV MP on the other hand is reported to bind viral RNA for transfer into neighboring cells.32,33 Therefore, the different observations of PVX TGBp1 and TMV MP transport between epidermal cells likely reflect their functional differences. Both proteins are required for virus cell-to-cell movement, but their exact roles in virus movement are not identical.

As mentioned earlier, Pd permeability is downregulated during plant development. Research tracking GFP diffusion through Pd in embryonic cells, in young emerging leaves, and in fully expanded leaves showed that fluorescence is highly mobile between cells in young tissues but is restricted during maturation. Viral movement proteins such as Cucumber mosaic virus 3a, and PVX TGBp1 remain highly mobile in mature leaves because they gate open Pd under conditions that normally restrict movement of GFP.34,35 Schoenknecht et al., undertook a straightforward investigation of leaf maturation describing Pd transport in relationship to leaf area expansion. The outcome of this study was evidence that GFP movement between cells declines as leaves expand.

It is reasonable to consider that simultaneous changes in gene expression and physiology is reflected in a downward trend in Pd conductivity and an increased requirement for Pd gating to enable selected transport of macromolecules between cells. In Arabidopsis embryos there is an obvious transition between developmental stages which are also represented by a decline in the ability for GFP to diffuse across Pd.36,37 A detailed analysis of Pd structure in source and sink tissues revealed that Pd are simple single channeled structures in sink tissues while source tissues contain predominantly “H” shaped branching Pd structures. The change in Pd structure has been correlated with changes in conductivity and is often correlated with changes in sink to source metabolism.38,39 The sink-to-source transition in leaf development is typically monitored using phloem loading of carboxyfluorescein diacetate. Leaves where CF dye unloads are defined as sink leaves and leaves that were restricted in dye unloading were defined as source leaves. Then biolistic bombardment of GFP expressing plasmids to sink and source leaves revealed that GFP readily diffuses across Pd in sink leaves but is more often restricted in source leaves.26,34,40–42

Leaf development is typically defined as a transition from juvenile to adult which is represented by homeotic transformations as well as vegetative phase changes.43,44 Source and sink regions of a leaf have been shown to correlate with changes in Pd structure and conductivity during leaf expansion. However, in our study we found that N. tabacum leaves identified as source during week 2 or 3 would continue to expand over an 8 week period to twice or three times the leaf area which provides a real indication that the source designation may not entirely reflect final leaf maturation or completion of leaf development.45> For example, as cells transition from sink to source physiology it is suggested that the frequency of single channeled Pd declines while the frequency of branched Pd increases.39 It is possible that even after leaves transition into photosynthetic sources that Pd architecture continues to change and there is a further decline in the proportion of single channel to branched channels. Therefore either the change in cell volume or Pd architecture or both can slow-down diffusion of GFP between cells.

Researchers often point to the ER continuity between cells as a driving force for Pd formation and function. During cell division the cell wall is laid down and forms around the ER creating Pd channels.46 However, it is also worth noting that the actin cytoskeleton is also present in Pd and is central to organ and reproductive development.19,47 Actin and actin binding proteins are necessary for a number of plant processes determining the cell division plane, cell polarity, cell elongation, cytoplasmic streaming, transporting mRNAs and proteins, and defense.48–51 Overexpression of ACT1 in Arabidopsis leaves can lead to changes in epidermal leaf shape and cause dwarfism in plants.52 Actin binding proteins are also necessary for organizing and remodeling the F-actin network which drives normal development of specific cell types and organs.53 Actin filament bundling and remodeling are also seen in nonhost defense responses.54 We do not know the effects of overexpressing certain actin homologues or actin remodeling on Pd formation or conductivity. Because the F-actin network is also central to Pd trafficking of proteins and macromolecules between cells it is worth considering F-actin as an early factor contributing to Pd formation which may be necessary to ensure cell-to-cell communication when cell polarity and elongation as well as defense machinery are being established.

In summary, the novel quantitative tools developed for measuring protein movement across Pd reveal the temperature dependence of protein trafficking. Both the use of Arrhenius equation and C(Pd) provide new opportunities to measure the energy requirements for protein transport. These tools will enable researchers to quantify effects of environmental and developmental conditions on Pd conductivity, as well as comparing differences in Pd conductivity between plant species or induced by genetic mutations.

Acknowledgements

I thank D. Meinke and S. Poethig for relevant discussions of leaf development across plant species. This research was supported by a research grant from the National Science Foundation IBM-9982552, LS-OKAMP, and the Native Americans in Biological Sciences (NABS) program.

Addendum to: