Search

AtTRB1, 2 and 3 are members of the SMH (single Myb histone) protein family, which comprises double-stranded DNA-binding proteins that are specific to higher plants. They are structurally conserved, containing a Myb domain at the N-terminus, a central H1/H5-like domain and a C-terminally located coiled-coil domain. AtTRB1, 2 and 3 interact through their Myb domain specifically with telomeric double-stranded DNA in vitro, while the central H1/H5-like domain interacts non-specifically with DNA sequences and mediates protein-protein interactions. Here we show that AtTRB1, 2 and 3 preferentially localize to the nucleus and nucleolus during interphase. Both the central H1/H5-like domain and the Myb domain from AtTRB1 can direct a GFP fusion protein to the nucleus and nucleolus. AtTRB1-GFP localization is cell cycle-regulated, as the level of nuclear-associated GFP diminishes during mitotic entry and GFP progressively re-associates with chromatin during anaphase/telophase. Using fluorescence recovery after photobleaching and fluorescence loss in photobleaching, we determined the dynamics of AtTRB1 interactions in vivo. The results reveal that AtTRB1 interaction with chromatin is regulated at two levels at least, one of which is coupled with cell-cycle progression, with the other involving rapid exchange.

Cyclin-dependent kinases (CDKs) play key regulatory roles in diverse cellular functions, including cell-cycle progression, transcription and translation. In plants, CDKs have been classified into several groups, named A through to G, but the functions of most are poorly characterized. CDKCs are known to phosphorylate the C-terminal domain (CTD) of RNA polymerase II (RNAP II), and therefore the CDKC-cyclinT (CycT) complex may have a role similar to the animal CDK9-CycT complex of the positive transcription elongation factor b (P-TEFb). However, we found that the predicted structure of the Arabidopsis CDKC2 protein is more similar to the mammalian cdc2-related kinase, CRK7, than to CDK9. CRK7 is proposed to link transcription with splicing, and CDKC2 contains all the structural features of CRK7 that make the latter distinct from CDK9. Consistent with this, we show that GFP-CDKC2 fusion proteins co-localize with spliceosomal components, that the expression of CDKC2 modifies the location of these components, and that co-localization was dependent on the transcriptional status of the cells and on CDKC2-kinase activity. We propose, therefore, that the Arabidopsis CDKC2 combines the functions of both CRK7 and CDK9, and could also couple splicing with transcription.

We describe a streamlined and systematic method for cloning green fluorescent protein (GFP)-open reading frame (ORF) fusions and assessing their subcellular localization in Arabidopsis thaliana cells. The sequencing of the Arabidopsis genome has made it feasible to undertake genome-based approaches to determine the function of each protein and define its subcellular localization. This is an essential step towards full functional analysis. The approach described here allows the economical handling of hundreds of expressed plant proteins in a timely fashion. We have integrated recombinational cloning of full-length trimmed ORF clones (available from the SSP consortium) with high-efficiency transient transformation of Arabidopsis cell cultures by a hypervirulent strain of Agrobacterium. To demonstrate its utility, we have used a selection of trimmed ORFs, representing a variety of key cellular processes and have defined the localization patterns of 155 fusion proteins. These patterns have been classified into five main categories, including cytoplasmic, nuclear, nucleolar, organellar and endomembrane compartments. Several genes annotated in GenBank as unknown have been ascribed a protein localization pattern. We also demonstrate the application of flow cytometry to estimate the transformation efficiency and cell cycle phase of the GFP-positive cells. This approach can be extended to functional studies, including the precise cellular localization and the prediction of the role of unknown proteins, the confirmation of bioinformatic predictions and proteomic experiments, such as the determination of protein interactions in vivo, and therefore has numerous applications in the post-genomic analysis of protein function.

We have developed a novel pre-embedding in situ hybridization labelling method for electron microscopy which has given much greater sensitivity and higher labelling levels than have been achieved previously, together with good ultrastructural preservation. Vibratome sections of plant tissue were labelled throughout their thickness with 1 nm gold antibodies and then silver enhanced, embedded in resin and sectioned for electron microscopy. Because the labelling extends throughout the depth of the specimen, this method permits the study of the 3D arrangement of the labelling at the electron microscope level by either stereo-pair recording, tomographic reconstruction or 3D reconstruction from serial sections. In this paper we describe the application of this method to study the organization of rDNA in pea root tissue.

The rDNA transcription units are enormous macromolecular structures located in the nucleolus and containing 50-100 RNA polymerases together with the nascent pre-rRNA attached to the rDNA. It has not previously been possible to visualize nucleolar transcription units directly in intact nucleoli, although highly spread preparations in the electron microscope have been imaged as "Christmas trees" 2-3 microm long. Here we determine the relative conformation of individual transcription units in Pisum sativum plant nucleoli using a novel labelling technique. Nascent transcripts were detected by a highly sensitive silver-enhanced 1 nm gold procedure, followed by 3D electron microscopy of entire nucleoli. Individual transcription units are seen as conical, elongated clusters approximately 300 nm in length and 130 nm in width at the thickest end. We further show that there were approximately 300 active ribosomal genes in the nucleoli examined. The underlying chromatin structure of the transcribing rDNA was directly visualized by applying a novel limited extraction procedure to fixed specimens in order to wash out the proteins and RNA, thus specifically revealing DNA strands after uranyl acetate staining. Using this technique, followed by post-embedding in situ hybridization, we observed that the nucleolar rDNA fibres are not extended but show a coiled, thread-like appearance. Our results show for the first time that native rDNA transcription units are linear, compacted Christmas trees.

Summary We have developed a method using vibratome sectioning coupled with confocal microscopy to analyse pre-meiotic and meiotic events in developing wheat anthers. This report describes the application of fluorescence in situ hybridization to three-dimensionally preserved tissue sections from intact florets. We show that excellent imaging of all the stages of meiosis up to telophase I can be achieved, and show centromere positions and pre-meiotic homologous pairing both in meiocytes and tapetum cells. These clear results using intact tissue sections are contrasted with conventional spreading methods for analysis of meiosis in plants, which we show can produce artefactual or uninterpretable results.

Incorporation by RNA polymerases of BrUTP into both plant root tissue and isolated plant nuclei as a method for localization of the sites of transcription has been used. In this paper pea root tissue was used, and under the conditions employed, nearly all the incorporation occurs in the nucleolus, and thus must be catalysed by RNA polymerase I. Immunofluorescence and confocal microscopy shows that incorporation occurs in a pattern consisting of many small foci distributed widely through the dense fibrillar component of the nucleoli. Immunogold labelling using silver-enhanced Nanogold probe at the electron microscopic level confirms the sites of transcription as small foci approximately 200nm in diameter. Simultaneous fluorescence in situ hybridization with a probe to the external transcribed spacer (ETS) region of the pre-rRNA shows that the structures revealed by this probe and the BrUTP immunofluorescence labelling are very similar. A probe to the transcribed portion of the rDNA (18S) also shows a good correlation to the sites of BrUTP incorporation within the nucleolus. On the other hand a probe to the non-transcribed intergenic spacer region (NTS) shows very little coincidence with the sites of BrUTP incorporation, and double fluorescence in situ labelling with both 18S and NTS probes confirms this difference in localization. These results suggest that most BrUTP foci correspond to single transcribed genes.

Summary By microinjecting rhodamine-labelled tubulin into living plant cells, it is possible to observe microtubules (MTs) directly and to see how the cortical array reorganizes itself. The validity of the conclusions drawn from such observations depends upon the assumption that most, if not all, of the native MTs are dynamic and incorporate labelled tubulin. However, if arrays also contain MTs that are not exchanging tubulin subunits, such MTs will remain unlabelled, and the labelled MT population will be under-representative of the whole array. To address this potential problem, we microinjected pea epidermal cells with rhodamine-labelled tubulin, then fixed the cells and used fluorescein-conjugated antibodies against tubulin to detect the entire MT array. The two fluorescent patterns corresponded well, confirming that the MTs labelled with exogenous tubulin were evenly distributed throughout the entire array. Also, by comparing the MT image before and after aldehyde fixation, we observed that, although some of the MTs were lost in the procedure, the fixation was able to preserve the arrangement of MTs seen in the living cell. We conclude that fluorescence analogue cytochemistry provides a valid representation of the entire cortical MT array.