Rice and Maize Prolamine RNAs are Localized by an RNADependent Mechanism

The high-resolution in situ hybridization approach, although sensitive, was extremely laborious and alternative more facile approaches were clearly required to study how these RNAs were sorted to the PB-ER and cisternal ER. Choi et al. (2000) developed a light microscopy-based approach using an in situ reverse transcriptase-mediated polymerase chain reaction (RT-PCR) with fluorescently labeled nucleotides to study RNA localization in developing rice endosperm cells. Figure 2 shows that fluorescence patterns obtained by in situ RT-PCR and their spatial relationships to the PB-ER are markedly distinct for prolamine and glutelin RNAs. Prolamine RNAs are distributed as small, spherical, and sometimes ringlike structures of 1-2 |m diameter which are identical to the structures labeled by BiP antibodies and visualized by indirect immunofluorescence (Choi et al. 2000). Consistent with this view is that these structures were also stained by DiOC6 and rhodamine B hexyl ester, two ER vital stains which specifically label PB-I because they also react with the hydrophobic prolamine polypeptides as well as the ER membrane (Choi et al. 2000) (Fig. 2). In contrast, glutelin RNAs are seen as much larger irregularly shaped patches located adjacent to PB-I, a pattern indicative of their localization on cisternal membranes (Fig. 2).

in situ RT-PCR Rhodamine stained Merged (mRNA) Prolamine PB image in situ RT-PCR Rhodamine stained Merged (mRNA) Prolamine PB image

Fig. 2 Prolamine and glutelin mRNA localization as viewed by in situ RT-PCR. Prolamine and glutelin RNAs were detected by in situ RT-PCR using gene-specific primers and Oregon Green 488 d-UTP. The spherical PB-ERs were visualized by staining the sections with Rhodamine B hexyl ester (middle images). Confocal microscopic analysis showed that prolamine RNAs were localized to spherical PB-ERs. In contrast, glutelin RNAs were distributed as large patches indicative of their location on cisternal ER located near PB-ER. Scale bar = 10 |im

Fig. 2 Prolamine and glutelin mRNA localization as viewed by in situ RT-PCR. Prolamine and glutelin RNAs were detected by in situ RT-PCR using gene-specific primers and Oregon Green 488 d-UTP. The spherical PB-ERs were visualized by staining the sections with Rhodamine B hexyl ester (middle images). Confocal microscopic analysis showed that prolamine RNAs were localized to spherical PB-ERs. In contrast, glutelin RNAs were distributed as large patches indicative of their location on cisternal ER located near PB-ER. Scale bar = 10 |im

The development of facile methods to assess RNA localization enabled a systematic study to determine whether this phenomenon was dependent on either an RNA or peptide-based elements. A series of transgenic plants expressing synthetic prolamine RNAs were generated and analyzed for RNA localization. They included a wild-type prolamine gene and three prolamine variants containing a substituted glutelin 3'UTR in place of the normal prolamine 3'UTR, a signal peptide deleted version, and an AUG translational knockout. These various prolamine transgenes were also tagged with unique DNA sequences to distinguish the transcripts made from the transgene from those derived from endogenous genes. The signal peptide deleted prolamine RNA was faithfully localized to the PB-ER, indicating that the signal peptide and the corresponding coding sequences were not required. In contrast, the prolamine glutelin 3'UTR RNA was mistargeted to the cisternal ER. As the epitope-tagged prolamine coded by this prolamine RNA variant was readily detected by immunoblot analysis, these results indicate that prolamine RNA localization is not dependent on the prolamine primary sequence. Interestingly, although the prolamine primary amino acid sequence is not essential for RNA targeting to the PB-ER, an intact AUG translation initiation codon is required for correct prolamine RNA localization to the PB-ER (Choi et al. 2000; Hamada et al. 2003b). This result indicates that the regulated prolamine RNA pathway requires the participation of the translational machinery. On the surface, this latter conclusion would appear to conflict with our earlier interpretation of RNA-directed localization. However, the requirement for translation and RNA targeting are not strictly coupled and can be contributed by different sequences, as demonstrated by the analysis of RNA reporters containing translatable green fluorescent protein (GFP) or ^-glucuronidase (GUS) sequences. Placement of prolamine sequences in the 3'UTR was able to direct the reporter RNAs to the PB-ER. Consequently, the requirement for translation is provided by the GUS or GFP reporter, with the RNA cis elements conferring PB-ER targeting located in the 3'UTR (Choi et al. 2000).

The dependence on RNA localization for protein localization appeared to be a unique feature of the rice system, even though other cereals, e.g., maize and sorghum, accumulate storage proteins in similar compartments to those used by rice. Indeed, based on in situ hybridization analysis using endosperm thin sections, RNAs for the maize zeins were reported to be essentially randomly distributed on the PB-ER and cisternal-ER membranes in developing maize endosperm (Kim et al. 2002). Protein body formation in maize was suggested to occur by diffusion of the zeins to the intracisternal inclusion granule followed by self- and interassembly of the various zein polypep-tide classes. Curiously, heterologous expression of the 10-kDa 8-zein RNA in developing rice endosperm showed a striking localization pattern. Unlike the random distribution of the 10-kDa 8-zein RNA reported in maize endosperm, the localization of this RNA was clearly restricted to the PB-ER (Hamada et al. 2003b).

To resolve this apparent discrepancy, Washida et al. (2004) assessed the distribution of the storage protein RNAs in developing maize endosperm using the in situ RT-PCR technique mentioned earlier. In developing maize endosperm cells, the PB-ER was visualized and the localization was confirmed by both double immunofluorescence using 10-kDa 8-zein antibody and rhodamine B hexyl ester staining. Analysis of the mRNA localization using in situ RT-PCR revealed that maize zein RNAs coding for 22-kDa a-zein, 15-kDa P-zein, 27-kDa y-zein, and 10-kDa 8-zein were localized to ER-bounded zein protein bodies (Washida et al. 2004). In contrast, RNAs coding for 51-kDa legumin-1, a storage protein sharing homology to the 11S globulins and rice glutelin, were distributed on adjacent cisternal ER. These results indicate that the same RNA targeting mechanism is at work in rice and maize, and that this mechanism has been conserved since the divergence of these species 50 million years ago.

The apparent differences in RNA distribution patterns depending on the technique used deserve comment. The in situ RT-PCR technique is significantly more sensitive as the tissue sample is much thicker, resulting in a larger target size as compared to the ultrathin sections of tissue analyzed by in situ hybridization at the electron microscopy level. The basis for the apparent random localization of zein RNAs when viewed by electron microscopy (Kim et al. 2002) is not known. However, the measured gold particle densities in this study were about an order of magnitude lower than those measured in rice (Li et al. 1993). One possibility that may account for the observed random distribution of RNAs was a preferential loss of RNAs associated with the PB-ER during the in situ hybridization procedure, as suggested by Kim et al. (2002).

Was this article helpful?

0 0

Post a comment