Developmental Changes Due to Long-Distance Movement of a Homeobox Fusion Transcript in Tomato

See allHide authors and affiliations

Science  13 Jul 2001:
Vol. 293, Issue 5528, pp. 287-289
DOI: 10.1126/science.1059805


Long-distance movement of RNA through the phloem is known to occur, but the functional importance of these transported RNAs has remained unclear. Grafting experiments with a naturally occurring dominant gain-of-function leaf mutation in tomato were used to demonstrate long-distance movement of mutant messenger RNA (mRNA) into wild-type scions. The stock-specific pattern of mRNA expression was graft transmissible, indicating that the mRNA accumulation pattern is inherent to the transcript and not attributable to the promoter. The translocated mRNA caused changes in leaf morphology of the wild-type scions, suggesting that the translocated RNA is functional.

Increased plant size and multicellularity require that plant cells and organs communicate with each other so that the organism can develop as a coordinated whole and adapt to the changing environment. Short-distance communication occurs through plasmodesmata. Regulatory proteins such as KNOTTED-1, DEFICIENS, GLOBOSA, and LEAFY may move from cell to cell in a developmentally significant manner (1–3). Specific and spatially restricted short-distance mRNA movement establishes the postanterior and dorsal-ventral polarity of the early oocyte in Drosophila (4, 5). Long-distance movement of water, plant hormones, minerals, sugars, and amino acids occurs through phloem and xylem. Long-distance movement of nucleic acids was first observed in plant viral infections. Viral movement proteins (MP) facilitate the cell-to-cell movement of viral nucleic acids through plasmodesmata by forming MP–nucleic acid complexes (6). Similarly CmPP16, a Cucurbita maxima paralog to viral movement protein, carries various mRNA molecules from cell to cell (7). Other examples of intercellular mRNA movement include SUCROSE TRANSPORTER1(8), the small RNAs that mediate cosuppression (9), and CmNACP (10). However, in the absence of a phenotypic effect caused by the translocated RNA, the functional importance of long-distance mRNA movement in regulating morphological development in plants remained unclear.

Tomato normally produces unipinnate compound leaves (Fig. 1A), whereas the dominant mutantMouse ears (Me) has up to octapinnate compound leaves (Fig. 1D) (11). Compared with wild-type leaflets with pinnate venation and acute lobes (Fig. 1B), Me leaflets are rounded and unlobed and have palmate venation at the base of the leaflets (Fig. 1E). This phenotype of the Me mutant is caused by a gene fusion between PYROPHOSPHATE-DEPENDENT PHOSPHOFRUCTOKINASE (PFP), an enzyme in the glycolytic pathway, and LeT6, a tomato KNOTTED-1–like homeobox (KNOX) gene. The PFP-LeT6 fusion gene includes about 10 kb of native PFP upstream sequence, allowing for a high-level expression pattern of the functional homeobox fusion transcript in the Me plants, leading to extra orders of leaf compounding (11–13). Wild-type plants [carrying the semidominant Xanthophyllic (Xa) mutation causing yellow normal leaves] were grafted onto Me plants (Table 1). In 11 out of 13 grafts of Xa scions on Me stocks (Xaheterografts), leaves had higher orders of pinnation than normal (Fig. 1G) and rounded lobes that were reduced in number (Fig. 1H). These leaves resembled those produced on plants carrying theMe mutation (Fig. 1E) as well as plants that overexpressLeT6 under control of the CaMV 35S promoter (11,12). Delayed trichome development at the tip of initiating leaf primordia was seen in Me and in Xa scions grafted onto Me stocks, indicating an early and sustained change in leaf phenotype in the Xa scions (Fig. 1, C, F, and I; Table 2).

Figure 1

Grafting on mutant stocks induces leaf shape changes in wild-type scions in tomato. (A) Wild-type leaf in theXanthophyllic mutation. (B) Wild-type leaflet. (C) Xa scanning electron micrograph (SEM) shows trichomes (arrow) forming at the developing leaf tip in P2 leaves. (D) Mouse ears (Me) mutant leaf. (E) Me leaflet. (F) SEM of theMe shoot apex shows trichome formation (arrow) in the middle region of the P2 developing leaf. (G) Xaheterografted scion with altered leaf shape and increased leaf subdivision. (H) Xa heterografted leaflets. (I) SEM of the Xa heterografted apex showing trichome formation (arrow) in the middle region of the developing P2 leaf. Scale bars, 50 μm in (C), (F), and (I).

Table 1

Phenotypic changes observed after grafting.

View this table:
Table 2

Summary of trichome patterns on leaf primordia.

View this table:

In the Me mutation, differential splicing creates two kinds of fusion RNA, but only one is in frame with the LeT6homeodomain (11). Reverse transcription polymerase chain reaction (RT-PCR) with PFP- and LeT6-specific primers (Fig. 2A) detected two products with the expected sizes of 518 (in frame) and 357 (out of frame) base pairs in RNA extracted from mature Me leaves (Fig. 2B) and only the in-frame product in RNA extracted from phenotypically altered Xa scions (Fig. 2B). The PFP-LeT6 fusion DNA could only be detected in the Me stocks (where the gene fusion is present) but not in DNA from Xa or Xaheterografted scions (Fig. 2C). Because both the endogenousPFP and LeT6 mRNAs could be detected in the scion (14), the observed phenotypic alterations are not caused by cosuppression of these genes. These results indicate that the chimeric PFP-LeT6 fusion RNA is transported from theMe stocks into Xa heterografted scions and likely causes the phenotypic changes seen in the scions.

Figure 2

A fusion RNA is transported from the stock into the scion. (A) The primers used in the RT-PCR experiments. PFP2 and LeT6-2 lead to an RT-PCR product in Me but not inXa. (B) RT-PCR experiments (20) with primers PFP2 and LeT6-2 performed on Me mature leaf RNA (lane 1), blank lane (lane 2), and RNA from shoot apices ofXa heterografted scions (lanes 3 and 4). The PCR products were blotted and hybridized to a LeT6 cDNA probe. Exposing the film for greater lengths of time did not reveal the presence of any 344–base pair product in lanes 3 and 4. (C) ThePFP-T6 fusion DNA was detected with primers PFP4 (to exon 13 of PFP) and LeT6-3 (to exon 1 of LeT6) in Meleaves (lane 1), but not in Xa (lane 2) or scion (Xa heterographs) (lane 3) leaves showing the phenotypic change (20). (D) PFP from exons 13 to 15 was amplified from the same genomic DNA samples as in (C) with primers PFP4 and PFP5. The PCR reactions were blotted and probed with exons 13 to 15 of the PFP cDNA.

LeT6 RNA was most strongly expressed in the central zone of the wild-type shoot apical meristem and in the vascular traces of developing leaf primordia (Fig. 3A), as reported in earlier studies (11). Nongrafted wild-type plants (as well as homografted Xa plants) did not show any signal for the presence of the PFP-LeT6 fusion mRNA (Fig. 3C). PFP-LeT6 fusion mRNA signals were seen in the periphery of the shoot apical meristem and tips of developing leaf and leaflet primordia in Me plants (Fig. 3B) and in the Xaheterografted scions (Fig. 3D). However, this expression pattern is different from the distribution pattern of PFP mRNA in the wild-type shoot apices (14). The fusion RNA appears in the Me-specific pattern in the heterografted scions independent of the presence of the PFP-LeT6 fusion promoter. Taken together, these in situ RT-PCR results show that thePFP-LeT6 fusion transcript moves from the Mestock to the Xa scion and the accumulation pattern of the fusion mRNA mimics that seen in Me plants but is not promoter driven.

Figure 3

In situ RT-PCR experiments (21) with Oregon Green–labeled uridine triphosphate show that translocated RNA accumulates in grafted shoot apices of scions in a pattern specific for the stock. The red signal indicates tissue autofluorescence, whereas the green signal represents the Oregon Green–labeled RNA detected by RT-PCR. When the two signals overlap, yellow fluorescence is detected. (A) Confocal laser scanning image of the wild-type (andXa) shoot apical meristem showing the T6 transcript detected with primers LeT6-1 and LeT6-2 in the apical dome, leaf primordia, and vascular traces. A leaf primordium is shown in the inset. (Bto D) Confocal laser scanning images of in situ RT-PCR experiments to specifically detect the PFP-LeT6 fusion transcript with the primers PFP2 and LeT6-2. (B) The fusion transcript was detected in the shoot apical meristem and leaf primordium in theMe apex. (C) Wild-type meristem showing the absence of fusion RNA. (D) The Xa heterografted shoot apex showed fusion transcript in the apical dome and leaf primordia. (Eto I) The PFP-LeT6 fusion transcript is transported in the phloem. Transverse (E and F) and longitudinal (G and H) sections of Xa heterografted scion stems were imaged. (E) Signal was seen in the region of the vascular bundle. (F and G) The fusion transcript is specifically localized to the sieve tubes (ST), sieve plates (SP), and associated companion cells (CC). (H) The signal was absent in the vessel elements of the heterografted scion stem. (I) No fusion transcript was detected in the nongrafted wild-type phloem. Nonspecific accumulation of the fluorescent labeled nucleotide was seen in large parenchyma cells in both control and experimental tissues [(E), (G), (H), and (I)]. Scale bars: (A) to (D), 50 μm; (E), 10 μm; (F), (G), and (I), 2 μm; and (H), 5 μm.

Chimeric PFP-LeT6 transcript was seen in the vascular bundles (Fig. 3E) and was specifically localized in the phloem sieve tubes and associated companion cells (Fig. 3, F and G) but not in the xylem (Fig. 3H). No accumulation of fusion transcript was seen in the nongrafted wild-type phloem (Fig. 3I). Translocation of the chimeric transcript through the phloem could be due to signals contained within either the PFP or LeT6 mRNAs. Analysis of interspecific grafts showed a small but detectable amount of stock-specific PFP mRNA in the scion (14). PFP and LeT6 mRNAs were detected in the phloem of Xa plants (14).LeT6 overexpression phenotypes are graft transmissible, suggesting that wild-type LeT6 mRNA is also translocated (15).

In the absence of an active circulatory system, plants may have evolved a system to transport RNAs in order to communicate signaling events throughout the organism and to thereby coordinate development. Our reciprocal grafting data suggest an acropetal direction of RNA movement (Table 1). At grafting, all scions had immature leaves, whereas the stocks had a few mature leaves on them, representing the source sink relations seen during normal development. Thus, mature leaves would perceive environmental or other signals and transport specific RNA molecules acropetally to allow the shoot apex to respond to these signals.

The accumulation of PFP-LeT6 fusion mRNA in the scion shoot apices and leaf primordia requires loading of fusion mRNA onto the sieve elements and its transport to the meristem. Similar transport of unloaded CmNACP mRNA to the meristematic regions was reported (10). Phloem-transported viruses rarely localize in the shoot apical meristem (16). Thus, the transport mechanism seems to be selective for certain mRNAs. Temporally and spatially regulated symplastic domains have been shown to exist in theArabidopsis shoot apical meristem (17) and could allow selective transport of various leaf-derived signals. Short-distance mRNA movement in Drosophila oocyte development requires the RNA binding protein STAUFEN (18).PFP-LeT6 mRNA loading into the phloem and unloading and transport to the meristematic regions might be facilitated by chaperones that recognize specific signals on the PFP orLeT6 mRNA. The intracellular distribution of specific RNAs has been shown to depend on both a start codon and specific 3′ regulatory sequences (19). Similar regulatory signals may also be involved in intercellular transport of RNAs.

RNA localization may not always be a property inherent in the promoter of the gene, but rather certain specific RNAs might exhibit the capacity to move and accumulate in regions of the shoot with distinct developmental and phenotypic consequences. This suggests a new paradigm for gene expression patterns. We show that mRNAs transported through the phloem are functional and can accumulate in patterns specific to that observed by in situ hybridization, suggesting that perhaps the native pattern of accumulation of the transcript is also due to transport and not always attributable to the promoter.

  • * These authors contributed equally to the work.

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


View Abstract

Navigate This Article