Proteins as Signal Molecules

Proteins synthesized in companion cells can clearly enter the sieve elements through the plasmodesmata that connect the two cell types. As noted earlier, the P-proteins in cucurbit sap, PP1 and PP2, appear to be synthesized in companion cells. The plasmodesmata connecting the companion cells and sieve elements must thus allow these macromolecules to move across them. Furthermore, subunits of PP1 and PP2 move with the translocation stream to sink tissues.

Subunits of P-proteins from pumpkin (Cucurbita maxima) can move across graft unions from a pumpkin stock (basal graft partner) to a cucumber (Cucumis sativus) scion (upper graft partner), as well as from a pumpkin scion to a cucumber stock. Both proteins are able to move from the sieve elements to companion cells; neither protein is able to move beyond the sieve element–companion cell complex (la Cour Petersen et al. 2005). Furthermore, no definite signaling function has yet been established for either protein.

Other proteins move in the translocation stream from sources to sinks. For example, passive movement of proteins from companion cells to sieve elements has been demonstrated in Arabidopsis and tobacco plants. These plants were transformed with the gene for green fluorescent protein (GFP) from jellyfish, under control of the SUC2 promoter from Arabidopsis. The SUC2 sucrose–H⁺ symporter is synthesized within the companion cells, so proteins expressed under the control of its promoter, including GFP, are also synthesized in the companion cells. GFP, which is localized by its fluorescence after excitation with blue light, moves through plasmodesmata from companion cells into sieve elements of source leaves (see textbook Figure 11.23A) and migrates within the phloem to sink tissues.

Finally, the jellyfish green fluorescent protein is unloaded symplastically through the plasmodesmata into sink tissues, such as seed coats, anthers, root tips, and mesophyll cells in importing leaves (see textbook Figure 11.23B) (Imlau et al. 1999, Stadler et al. 2005). Because jellyfish GFP is unlikely to possess specific sequences for interaction with plasmodesmatal structures, its movement into and out of sieve elements is likely to occur by passive diffusion. Some soluble fusion proteins between GFP and other proteins (ubiquitin, for example), expressed in Arabidopsis companion cells, show similar transport patterns. However, only free GFP is able to move symplastically into sink tissues of the root (Stadler et al. 2005).

Phloem transport of proteins that modify cellular functions has also been demonstrated (Web Table 11.12.A below). For example, FLOWERING LOCUS T (FT) protein appears to be a significant component of the floral stimulus that moves from leaf to apex in response to conditions inductive for flowering (see textbook Chapter 25). The PRms protein is a pathogenesis-related protein that is involved in the defense response of maize against fungal pathogens. PRms protein moves into and through the phloem, as shown by its presence in tobacco phloem exudate and its movement across graft unions. PRms can move across graft unions from a transgenic stock to a nontransgenic control (Bortolotti et al. 2005). Unloading of the protein has not yet been demonstrated, however, nor is its exact mechanism of action known.

WEB TABLE 11.12.A Functions of Some Soluble Proteins Found in Sieve Element Sap
FUNCTION	EXAMPLES
Sugar metabolism	Sucrose synthase, glycolytic enzymes
Protein degradation	Ubiquitin
Redox regulation	Thioredoxin h
Protein phosphorylation	Protein kinases
Protein folding	Chaperones
Antioxidant defense	A suite of enzymes and radical scavengers
Protection of phloem proteins from degradation plus defense against pathogens and phloem-feeding insects	Protease inhibitors
RNA-binding	CmPP16, PP2, glycine-rich RNA-binding protein
Flowering signal	FLOWERING LOCUS T protein
(Source: Hayashi et al. 2000; Walz et al. 2004, Gaupels et al. 2008, Aki et al. 2008)

Clearly, proteins can be transported from the companion cells in the source through the intervening sieve elements to sink companion cells. However, little evidence exists for a similar movement of proteins synthesized outside the companion cells. Other signals from outside the sieve element–companion cell complex may give rise to the production of mobile proteins in the companion cells.

RNAs as Signal Molecules

RNAs transported in the phloem consist of endogenous mRNAs, pathogenic RNAs, and small RNAs associated with gene silencing (see textbook Chapter 2). Most of these RNAs appear to travel in the phloem as complexes of RNA and protein (ribonucleoproteins [RNPs]) (Gomez et al. 2005, Kehr and Buhtz 2008). RNPs in the phloem resemble the complexes formed by viral RNAs with their movement proteins (MPs). Viral "movement proteins" interact directly with plasmodesmata to allow the passage of viral nucleic acids between cells.

The mechanism of cell-to-cell movement of viruses may differ from the mechanism of their import into the phloem and long-distance transport. While the role of MPs has been established in cell-to-cell movement through plasmodesmata, viral coat proteins may or may not be required for cell-to-cell movement. While viral coat proteins are widely required for viral translocation in the phloem, MPs may not be required (Kehr and Buhtz 2008). In addition, endogenous phloem proteins could be involved in the phloem transport of viruses. For example, PP2, a phloem structural protein well known for its ability to seal off damaged sieve elements, is found in the phloem exudate of cucumber in a complex with viroid (pathogenic) RNA. Further, PP2 and the viroid RNA are simultaneously translocated from a cucumber stock to a pumpkin scion. Following transport, the viroid is capable of causing disease symptoms in the pumpkin scion (Gomez and Pallas 2004).

Posttranscriptional gene silencing (see textbook Chapter 2) is one system that clearly demonstrates the role of RNAs in long-distance signaling in the phloem (van Bel 2003). RNA silencing of gene expression causes degradation of target mRNAs, for example, excess or foreign mRNA transcribed from transgenes or viral nucleic acid. When the transgenes are homologous to host genes, the host genes are also silenced. RNA silencing is triggered by the presence of double-stranded RNAs, induced by the virus or transgene. The double-stranded RNAs are processed by proteins called Dicer-like proteins to form small RNAs called short interfering RNAs (siRNAs). These siRNAs are incorporated into complexes that carry out the degradation of the target mRNAs. In this way, the gene is "silenced" (Eamens et al. 2008).

RNA silencing plays a role not only in defense processes against foreign mRNAs but also in developmental processes. Small RNAs called microRNAs (miRNAs) are formed similarly to siRNAs and likewise are incorporated into complexes that mediate degradation of mRNA (plants) or inhibit translation of mRNA (animals and insects) (Eamens et al. 2008) (see Chapter 17).

Small RNAs appear to meet all the requirements for signaling molecules: entry into sieve elements and transport in the phloem to sink tissues, where functional changes are triggered. Small RNAs shown to be authentic regulatory RNAs (siRNA and miRNA) are found in phloem sap, and their levels respond to growth conditions and viral infection (Yoo et al. 2004). Furthermore, phloem transport of small RNAs has been demonstrated, at least for siRNAs. An siRNA complementary to a viral coat-protein gene sequence was found in the sap of a wild-type cucumber scion that had been grafted onto a squash stock silenced for transgene coat-protein expression. Silencing occurred in the tissues of a scion from a nonsilenced line that had been grafted onto a stock from a silenced squash line; that is, the levels of coat protein mRNA were reduced in the nonsilenced line to the same low levels as found in the silenced line (Yoo et al. 2004).

Transmission of a silencing signal in the opposite direction, from a silenced scion to a nonsilenced rootstock, has also been demonstrated in Nicotiana benthamiana, a close relative of tobacco native to Australia. Radioactive phosphate was utilized as a phloem tracer, and source-sink relationships were manipulated by removing source leaves from the rootstock. The transport of the silencing signal followed the pattern of source-to-sink phloem transport, from the scion to the rootstock (Tournier et al. 2006).

Finally, the RNA transported in the phloem can cause visible changes in the sink. Changes in phenotype (appearance) due to the transport of RNA to a scion has been shown in a number of grafting experiments: the changes include chlorosis in tobacco due to posttranscriptional gene silencing and changes in leaf morphology in tomato due to transport of a mutant homeobox mRNA. In both cases, RNA movement occurred in an acropetal (toward the apex) direction, but movement in the phloem was not confirmed. However, mRNA for a regulator of gibberellic acid responses (called GAI) was localized to sieve elements and companion cells of pumpkin and was found in pumpkin phloem sap.

Transgenic tomato plants expressing a mutant version of the regulator gene were dwarf and dark green. The mRNA for the mutant regulator was localized to sieve elements, was able to be transported across graft unions into wild-type scions, and was unloaded into apical tissues. As a result, the mutant phenotype developed in new growth on the wild-type scion (Haywood et al. 2005). In a final example, a transcription factor which regulates tuber formation in potato accumulates in leaves and stolons in response to inductive short days and moves basipetally across a graft union from the upper graft partner (scion, overexpressing the transcription factor) to the stolon tips (wild-type stock), the site of tuber induction. Increased tuber production results (Banerjee et al. 2006).

Plasmodesmata Function in Signaling

The endogenous proteins commandeered by viruses to facilitate their specific movement to and through plasmodesmata are thought to carry out similar functions for endogenous macromolecules. Proteins have been identified from phloem exudate that bind RNA and increase the SEL of plasmodesmata between mesophyll cells (for example, PP2 and CmPP16); in addition, proteins have been identified in plasmodesmata that interact with the exudate proteins (for example, NtNCAPP1). The control over plasmodesmatal transport exerted by NtNCAP1 appears to be specific and selective (Ding et al. 2003).

It is believed that proteins such as CmPP16 are involved in macromolecular trafficking into sieve elements and long-distance transport in the phloem. Recently, CmPP16 from pumpkin was introduced into rice sieve elements through aphid stylets; the transport of CmPP16 to roots appeared to be selective (Aoki et al. 2005). Although it has not been shown to increase plasmodesmatal permeability, the maize pathogenesis-related protein (PRms) described earlier has been located in plasmodesmata between companion cells and sieve elements in transgenic tobacco plants. Localization was accomplished using an anti-PRms antibody. PRms increases allocation into sucrose, decreases allocation into starch, and thus increases sucrose efflux from source leaves of transgenic tobacco plants (Bortolotti et al. 2005),