Plant Paralog to Viral Movement Protein That Potentiates Transport of mRNA into the Phloem

See allHide authors and affiliations

Science  01 Jan 1999:
Vol. 283, Issue 5398, pp. 94-98
DOI: 10.1126/science.283.5398.94


CmPP16 from Cucurbita maxima was cloned and the protein was shown to possess properties similar to those of viral movement proteins. CmPP16 messenger RNA (mRNA) is present in phloem tissue, whereas protein appears confined to sieve elements (SE). Microinjection and grafting studies revealed that CmPP16 moves from cell to cell, mediates the transport of sense and antisense RNA, and moves together with its mRNA into the SE of scion tissue. CmPP16 possesses the characteristics that are likely required to mediate RNA delivery into the long-distance translocation stream. Thus, RNA may move within the phloem as a component of a plant information superhighway.

Phloem represents an advanced long-distance transport system that delivers nutrients and hormones to plant tissues and organs. Mature SE are enucleate (1) and thus must rely on their associated companion cells (CC) for maintenance of their physiological functions (2). To this end, SE are connected to CC through specialized, branched plasmodesmata (3) that mediate delivery of proteins into the long-distance translocation stream (4, 5).

The observation that specific mRNA molecules, such as sucrose transporter 1 (SUT1) RNA, are present within CC-SE plasmodesmata and in parietal locations in functional SE (6) suggests that RNA can similarly traffic through CC-SE plasmodesmata. These findings are consistent with experiments that probe the mechanism by which plant viruses establish a systemic infection. Genetic, molecular, and cellular approaches have established that plant viruses express movement proteins (MP) having the capacity to interact with plasmodesmata to mediate cell-to-cell transport of MP and viral nucleic acid–MP complexes (7–9). Thus, plant viruses likely have evolved the capacity to exploit the endogenous pathways utilized by the plant to traffic macromolecules from their sites of synthesis into surrounding cells (1).

Delivery of RNA to distant tissues and developing organs may reflect a mechanism used by plants to regulate translational events (11). Operation of this endogenous RNA translocation system could involve phloem-specific proteins whose functions would parallel those of plant viral MPs. To identify such proteins within the phloem sap (Fig. 1A), we used polyclonal antibodies raised against the 35-kD MP of red clover necrotic mosaic virus (RCNMV) (12) in immunoblot analyses (Fig. 1B). Here we report on the isolation and characterization of the 16-kD Cucurbita maxima (pumpkin) phloem protein (CmPP16) that displays functional similarity, and a limited degree of sequence identity, to the MP of RCNMV.

Figure 1

Immunodetection of pumpkin phloem proteins. Phloem proteins were extracted (4), resolved by SDS–polyacrylamide gel electrophoresis (SDS-PAGE), and either stained with Coomassie blue (A) or immunoblotted (14,24) to identify proteins that cross-react with antiserum to RCNMV MP (B) or polyclonal antibodies to R-CmPP16-1 (C). Samples tested include recombinant RCNMV MP, recombinant CmPP16-1, and phloem sap proteins (4). Because R-CmPP16-1 contains a histidine tag and a peptide linker, it is about 2 kD larger than the endogenous proteins, which differ in mass by 1 kD, in agreement with the deduced size of the CmPP16-1 and CmPP16-2 proteins (Fig. 2). We did not detect a reaction with either antiserum directed against a total protein preparation extracted from E. coli or with preimmune serum. Total phloem protein, 50 μg per lane; recombinant proteins (RCNMV MP and R-CmPP16-1), 5 μg per lane.

CmPP16 was purified, its NH2-terminus was sequenced, and reverse transcriptase–polymerase chain reaction (RT-PCR) was used to amplify a CmPP16 probe (13). We identified two cDNA clones from a pumpkin stem cDNA library constructed from stem mRNA (14) that encode proteins with predicted masses of 16.5 and 15.6 kD. These cDNA clones shared 82% identity at the amino acid level and were designated CmPP16-1 and CmPP16-2, respectively. We synthesized a fragment of CmPP16 by PCR and identified four introns within the CmPP16-1 genomic clone (Fig. 2A). This result established that these are endogenous plant genes, not derived from viral RNA, which would lack introns (and could have been present in the phloem of infected plants). Hybridization of 32P-labeledCmPP16-1 RNA to a genomic Southern blot identified two genes (15). Recombinant His6-CmPP16-1 (R-CmPP16-1) was produced in Escherichia coli and used to generate polyclonal antiserum. Immunoblot analysis with this R-CmPP16-1 antiserum recognized R- CmPP16-1, endogenous CmPP16, and the RCNMV MP (Fig. 1C). A BLAST search identified homologues in rice and maize (Fig. 2B); incomplete sequences for two additional genes were also present in theArabidopsis expressed sequence tag database.

Figure 2

Gene structure and comparison of CmPP16 and related proteins. (A) Schematic representation ofCmPP16-1 gene displaying its five exons. (B) TheCmPP16-1 (GenBank accession number AF079170) andCmPP16-2 (GenBank accession number AF079171) genes encode proteins of 16.5 and 15.6 kD, respectively (23). Homologous proteins were identified in rice (Os-FIERG2, accession no. U95136; Os-FIERG1, accession no. U95135; Os-1, accession no. D24373) and maize (Zm-1, accession no. U64437); the regions of greatest identity were present toward the NH2-terminus. A PKC-like C2 Ca2+/phospholipid-binding domain (16) is indicated by open rectangles. A comparison of blocks (25) of these proteins with RCNMV MP identified four domains having a high antigenic index (DNAstar, Madison, Wisconsin). Black boxes indicate identity, and shaded boxes indicate conservative changes. MegAlign software was used to align sequences (25).

Immunocytochemical experiments established that CmPP16 protein is confined to the SE periphery within the vascular tissue (Fig. 3, C and D; schematic transverse section of pumpkin stem), which suggests that it may be associated with the plasma membrane. Consistent with this observation, we found that CmPP16 contains a C2 domain (Fig. 2B) present in the Ca2+-binding protein kinase C (PKC) family (16). Northern blot (Fig. 3B) and immunocytochemical localization analyses demonstrated that CmPP16 mRNA and protein are present in the SE of leaves, stems, roots, and flowers. Finally, both high-resolution in situ hybridization and in situ RT-PCR (17) methods demonstrated that the CmPP16 mRNA is predominantly located within CC, and to a lesser extent in SE, of the functional phloem of petiole and stem tissues (Fig. 3, E to H).

Figure 3

Localization ofCmPP16 mRNA and protein in C. maxima vascular tissue. (A) Schematic transverse section of a pumpkin stem. Vascular bundles are composed of internal and external phloem (IP and EP, respectively), cambium (CA), and xylem (X) [see supplementary material (Fig. 2)]. Such bundles are distributed in a ring around the outer region of the stem (or petiole). (B) Northern blot analyses establish that theCmPP16 RNA is located in a variety of plant tissues. Total RNA (10 μg) from the indicated tissue was electrophoresed, blotted (14), probed with 32P-labeled CmPP16 open reading frame at 65°C, and washed with 0.1× SSC (0.015 M NaCl and 0.0015 M sodium citrate). (C) Cellular arrangement within the IP illustrating functional (arrowheads) and immature (asterisks) SE derived from the CA. Semithin section stained with toluidine blue. (D) Confocal laser scanning microscope (CLSM) image of a pumpkin semiserial section [see (C)] labeled with antiserum to R-CmPP16-1 reveals the presence of CmPP16 within the phloem (26). CmPP16 signal (green fluorescence) is present at the periphery of mature and immature SE. Tissue structure was observed (red fluorescence) with Safranin O used as a histochemical stain. Controls with preimmune sera yielded images devoid of fluorescent signal. (E to H) CLSM images of transverse pumpkin petiole sections processed for in situ RT-PCR (17). To identify cellular structure, we collected images (E and G) before removal of unincorporated CF-labeled deoxyuridine triphosphate. Asterisks identify immature SE. CmPP16 mRNA is detected (green fluorescent signal) within immature CC-SE complexes and, predominantly, in mature CC (F); note low but detectable amount of green fluorescent signal in SE. White asterisks in (E) and (F) facilitate identification of the same cell in the two images. In control experiments, tissues were treated in the same manner except that primers were omitted (H). Red fluorescence in (F) and (H) represents autofluorescence. Scale bars, 50 μm; bar in (D) is common to (C) and bar in (F) is common to (E), (G), and (H).

Sequence comparison of CmPP16 and RCNMV MP showed four motifs that display similarity (Fig. 2B) and likely account for the amount of immunological cross-reactivity detected between these proteins. Gel mobility-shift assays were next performed to test the capacity of these proteins to interact with RNA. R-CmPP16-1 was able to bind both sense and antisense CmPP16-1 RNA as well as RCNMV RNA2 (Fig. 4A). An equivalent capacity for RNA binding was displayed by the RCNMV MP (Fig. 4B). Similar to RCNMV MP (8), R-CmPP16-1 bound RNA but failed to interact with both single-stranded and double-stranded DNA (Fig. 4, D and E, respectively). Bovine serum albumin was used as a control, showing no interaction with DNA (Fig. 4C). To examine the functions of these two proteins, we carried out a series of microinjection experiments (18). As found for the RCNMV MP (8), R-CmPP16-1 has the capacity to interact with plasmodesmata to induce an increase in size-exclusion limit, potentiate its own cell-to-cell transport, and mediate the trafficking of RNA (Table 1).

Figure 4

R-CmPP16-1 binds various RNA molecules. Electromobility-shift assays were performed by mixing R-CmPP16-1 (A) or RCNMV MP (B) with 50 ng of the appropriate32P-labeled RNA in binding buffer (8). These mixtures were incubated for 1 hour at 4°C, resolved in 1% (w/v) agarose gels, and then processed for autoradiography. (C) Control RNA binding assays were performed with bovine serum albumin. The 32P-labeled RNA probe was synthesized in vitro (MAXIscript kit, Ambion, Austin, Texas). A similar series of electromobility-shift assays was performed with single-stranded (ss) (D) and double-stranded (ds) (E) DNA. DNA was electrophoresed as in (A) and visualized by ethidium bromide staining. The amounts of protein used in each experiment were as follows: 0.0, 0.1, 0.2, 0.4, and 0.8 μg. Both R-CmPP16-1 and RCNMV MP interact with RNA in a manner consistent with cooperative binding. However, R-CmPP16-1 did not interact with DNA. Both R-CmPP16-1 and RCNMV MP formed complexes with sense and antisense RNA, and similar threshold amounts of protein were required for RNA retardation. In contrast, bovine serum albumin failed to retard the RNA probes.

Table 1

R-CmPP16-1 interacts with mesophyll plasmodesmata to mediate its own cell-to-cell transport and potentiate the trafficking of sense and antisense RNA. Lucifer yellow CH and 11-kD FITC-dextran were from Molecular Probes. TOTO, TOTO-1 iodide; TRITC, tetramethylrhodamine-5-(and 6)-isothiocyanate.

View this table:

Consistent with our gel mobility-shift assays, R-CmPP16-1 mediated the cell-to-cell transport of both sense and antisense RNA of two different sequences but was unable to effect the movement of single- or double-stranded DNA. No cell-to-cell movement was detected when chromatide fluorescein (CF)-labeled RNA or DNA-TOTO (Molecular Probes) alone was introduced into a target mesophyll cell (Table 1). These results support the hypothesis that R-CmPP16-1 mediates the cell-to-cell transport of RNA through mesophyll plasmodesmata (19).

Grafting experiments between pumpkin and cucumber plants provided another test of the capacity of endogenous CmPP16 to move from the CC into the SE through plasmodesmata. Phloem sap was collected from the cucumber scion of 10-day-old grafted plants (5), and proteins were electrophoresed, blotted, and immunodetected with antiserum to R-CmPP16-1. Endogenous CmPP16 was pres-ent in the phloem sap of both pumpkin and the cucumber scion (Fig. 5A). Hence, the heterograft, immunolocalization, and microinjection experiments establish that CmPP16 protein has the capacity to move from the CC into the phloem long-distance translocation stream. Aliquots of these same phloem saps were used to extract RNA, which was then amplified by RT-PCR, which revealed the presence of CmPP16-1 mRNA in both pumpkin (control) and the cucumber scion but not in homografted cucumber (Fig. 5B). These results are consistent with our in situ RT-PCR, gel mobility-shift, and microinjection assays and provide support for the hypothesis that CmPP16 protein plays a role in mRNA delivery into the phloem translocation stream.

Figure 5

Endogenous CmPP16 and its RNA move in the phloem from pumpkin stock into a heterografted scion. Phloem sap was collected from pumpkin, cucumber, and cucumber scion heterografted onto pumpkin stock (5). (A) Proteins were resolved in a 4 to 20% SDS-PAGE gradient and then immunodetected with antiserum to R-CmPP16-1. CmPP16-1 and CmPP16-2 are present in phloem sap collected from pumpkin (control) and the heterografted cucumber scions, indicating that the two forms of CmPP16 move within the phloem long-distance translocation stream, albeit with seemingly different efficiencies. Two putative CmPP16 homologues are present in the phloem sap collected from cucumber but are absent or present at much reduced amounts in the sap of heterografted cucumber scions. Absence of the cucumber homologues from the scion phloem sap may reflect a block on their entry into the SE by the pumpkin proteins. Total phloem protein loaded was 50 μg per lane. (B) CmPP16 RNA is present in phloem sap collected from pumpkin and heterografted cucumber scion but absent from sap obtained from cucumber. CmPP16 RNA was analyzed by RT-PCR (17).

Our results add to the emerging picture of non–cell-autonomous regulation of gene expression in plants (20). The discovery that sequence-specific cosuppression can operate through an imposed graft union (21) likely reflects the involvement of RNA transport through the phloem (11). The ability of R-CmPP16-1 to mediate cell-to-cell transport of both sense and antisense transcripts of various sequences provides a possible molecular basis to explain how plants can translocate RNA present within the functional sieve tube system (17).

Insight into the possible involvement of supracellular control over developmental processes in animal systems has come from studies in which double-stranded RNA was microinjected into Caenorhabditis elegans (22). It will be intriguing to learn the extent to which the macromolecular trafficking systems used by plants and animals share common features.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: wjlucas{at}


View Abstract

Navigate This Article