"In situ" translation: use of the cytoskeletal framework to direct cell-free protein synthesis

Heterogeneous nuclear ribonucleoprotein (hnRNP) E1 binds to hnRNP A2 and inhibits translation of A2 response element mRNAs

Heterogeneous nuclear ribonucleoprotein (hnRNP) A2 is a trans-acting RNA-binding protein that mediates trafficking of RNAs containing the cis-acting A2 response element (A2RE). Previous work has shown that A2RE RNAs are transported to myelin in oligodendrocytes and to dendrites in neurons. hnRNP E1 is an RNA-binding protein that regulates translation of specific mRNAs. Here, we show by yeast two-hybrid analysis, in vivo and in vitro coimmunoprecipitation, in vitro cross-linking, and fluorescence correlation spectroscopy that hnRNP E1 binds to hnRNP A2 and is recruited to A2RE RNA in an hnRNP A2-dependent manner. hnRNP E1 is colocalized with hnRNP A2 and A2RE mRNA in granules in dendrites of oligodendrocytes. Overexpression of hnRNP E1 or microinjection of exogenous hnRNP E1 in neural cells inhibits translation of A2RE mRNA, but not of non-A2RE RNA. Excess hnRNP E1 added to an in vitro translation system reduces translation efficiency of A2RE mRNA, but not of nonA2RE RNA, in an hnRNP A2-dependent manner. These results are consistent with a model where hnRNP E1 recruited to A2RE RNA granules by binding to hnRNP A2 inhibits translation of A2RE RNA during granule transport.

Compartmentalisation and localisation of the translation initiation factor (eIF) 4F complex in normally growing fibroblasts

Previous observations of association of mRNAs and ribosomes with subcellular structures highlight the importance of localised translation. However, little is known regarding associations between eukaryotic translation initiation factors and cellular structures within the cytoplasm of normally growing cells. We have used detergent-based cellular fractionation coupled with immunofluorescence microscopy to investigate the subcellular localisation in NIH3T3 fibroblasts of the initiation factors involved in recruitment of mRNA for translation, focussing on eIF4E, the mRNA cap-binding protein, the scaffold protein eIF4GI and poly(A) binding protein (PABP). We find that these proteins exist mainly in a soluble cytosolic pool, with only a subfraction tightly associated with cellular structures. However, this "associated" fraction was enriched in active "eIF4F" complexes (eIF4E.eIF4G.eIF4A.PABP). Immunofluorescence analysis reveals both a diffuse and a perinuclear distribution of eIF4G, with the perinuclear staining pattern similar to that of the endoplasmic reticulum. eIF4E also shows both a diffuse staining pattern and a tighter perinuclear stain, partly coincident with vimentin intermediate filaments. All three proteins localise to the lamellipodia of migrating cells in close proximity to ribosomes, microtubules, microfilaments and focal adhesions, with eIF4G and eIF4E at the periphery showing a similar staining pattern to the focal adhesion protein vinculin.

Differential translation and fragile X syndrome

Fragile X syndrome (FXS) is caused by the transcriptional silencing of the Fmr1 gene, which encodes a protein (FMRP) that can act as a translational suppressor in dendrites, and is characterized by a preponderance of abnormally long, thin and tortuous dendritic spines. According to a current theory of FXS, the loss of FMRP expression leads to an exaggeration of translation responses linked to group I metabotropic glutamate receptors. Such responses are involved in the consolidation of a form of long-term depression that is enhanced in Fmr1 knockout mice and in the elongation of dendritic spines, resembling synaptic phenotypes over-represented in fragile X brain. These observations place fragile X research at the heart of a long-standing issue in neuroscience. The consolidation of memory, and several distinct forms of synaptic plasticity considered to be substrates of memory, requires mRNA translation and is associated with changes in spine morphology. A recent convergence of research on FXS and on the involvement of translation in various forms of synaptic plasticity has been very informative on this issue and on mechanisms underlying FXS. Evidence suggests a general relationship in which the receptors that induce distinct forms of efficacy change differentially regulate translation to produce unique spine shapes involved in their consolidation. We discuss several potential mechanisms for differential translation and the notion that FXS represents an exaggeration of one 'channel' in a set of translation-dependent consolidation responses.

The microtubule-associated protein tumor overexpressed gene binds to the RNA trafficking protein heterogeneous nuclear ribonucleoprotein A2

In neural cells, such as oligodendrocytes and neurons, transport of certain RNAs along microtubules is mediated by the cis-acting heterogeneous nuclear ribonucleoprotein A2 response element (A2RE) trafficking element and the cognate trans-acting heterogeneous nuclear ribonucleoprotein (hnRNP) A2 trafficking factor. Using a yeast two-hybrid screen, we have identified a microtubule-associated protein, tumor overexpressed gene (TOG)2, as an hnRNP A2 binding partner. The C-terminal third of TOG2 is sufficient for hnRNP A2 binding. TOG2, the large protein isoform of TOG, is the only isoform detected in oligodendrocytes in culture. TOG coimmunoprecipitates with hnRNP A2 present in the cytoskeleton (CSK) fraction of neural cells, and both coprecipitate with microtubule stabilized pellets. Staining with anti-TOG reveals puncta that are localized in proximity to microtubules, often at the plus ends. TOG is colocalized with hnRNP A2 and A2RE-mRNA in trafficking granules that remain associated with CSK-insoluble tissue. These data suggest that TOG mediates the association of hnRNP A2-positive granules with microtubules during transport and/or localization.

Actin depolymerization affects stress-induced translational activity of potato tuber tissue

Changes in polymerized actin during stress conditions were correlated with potato (Solanum tuberosum L.) tuber protein synthesis. Fluorescence microscopy and immunoblot analyses indicated that filamentous actin was nearly undetectable in mature, quiescent aerobic tubers. Mechanical wounding of postharvest tubers resulted in a localized increase of polymerized actin, and microfilament bundles were visible in cells of the wounded periderm within 12 h after wounding. During this same period translational activity increased 8-fold. By contrast, low-oxygen stress caused rapid reduction of polymerized actin coincident with acute inhibition of protein synthesis. Treatment of aerobic tubers with cytochalasin D, an agent that disrupts actin filaments, reduced wound-induced protein synthesis in vivo. This effect was not observed when colchicine, an agent that depolymerizes microtubules, was used. Neither of these drugs had a significant effect in vitro on run-off translation of isolated polysomes. However, cytochalasin D did reduce translational competence in vitro of a crude cellular fraction containing both polysomes and cytoskeletal elements. These results demonstrate the dependence of wound-induced protein synthesis on the integrity of microfilaments and suggest that the dynamics of the actin cytoskeleton may affect translational activity during stress conditions.

Visualization of prosomes (MCP-proteasomes), intermediate filament and actin networks by "instantaneous fixation" preserving the cytoskeleton

A new "instantaneous" fixation/extraction procedure, yielding good preservation of intermediate filaments (IFs) and actin filaments when applied at 37 degrees C, has been explored to reexamine the relationships of the prosomes to the cytoskeleton. Prosomes are protein complexes of variable subunit composition, including occasionally a small RNA, which were originally observed as trans-acting factors in untranslated mRNPs. Constituting also the proteolytic core of the 26S proteasomes, they are also called "multicatalytic proteinase (MCP) complexes" or "20S-Proteasomes." In Triton X-100-extracted epithelial, fibroblastic, and muscle cells, prosome particles were found associated primarily with the IFs (Olink-Coux et al., 1994). Application of "instantaneous fixation" has now led to the new observation that a major fraction of prosome particles, composed of specific sets of subunits, is distributed in variable proportions between the IFs and the microfilament/ stress fiber system in PtK1 epithelial cells and human fibroblasts. Electron microscopy using gold-labeled antibodies confirms this dual localization on classical whole mounts and on cells exposed to instantaneous fixation. In contrast to the resistance of the prosome-IF association, a variable fraction of the prosome particles is released from the actin cytoskeleton by Triton X-100 when applied prior to fixation. Moreover, in vitro copolymerization of prosomes with G-actin made it possible to observe "ladder-like" filamentous structures in the electron microscope, in which the prosome particles, like the "rungs of a ladder," laterally crosslink two or more actin filaments in a regular pattern. These results demonstrate that prosomes are bound in the cell not only to IFs but also to the actin cytoskeleton and, furthermore, not only within large M(r) complexes (possibly mRNPs and/or 26S proteasomes), but also directly, as individual prosome particles.

Protein translation components are colocalized in granules in oligodendrocytes

The intracellular distribution of various components of the protein translational machinery was visualized in mouse oligodendrocytes in culture using high resolution fluorescence in situ hybridization and immunofluorescence in conjunction with dual channel confocal laser scanning microscopy. Arginyl-tRNA synthetase, elongation factor 1a, ribosomal RNA, and myelin basic protein mRNA were all co-localized in granules in the processes, veins and membrane sheets of the cell. Colocalization was evaluated by dual channel cross correlation analysis to determine the correlation index (% colocalization) and correlation distance (granule radius), and by single granule ratiometric analysis to determine the distribution of the different components in individual granules. Most granules contained synthetase, elongation factor, ribosomal RNA and myelin basic protein mRNA. These results indicate that several different components of the protein synthetic machinery, including aminoacyl-tRNA synthetases, elongation factors, ribosomes and mRNAs, are colocalized in granules in oligodendrocytes. We propose that these granules are supramolecular complexes containing all of the necessary macromolecular components for protein translation and that they represent a heretofore undescribed subcellular organization of the protein synthetic machinery. This spatial organization may increase the efficiency of protein synthesis and may also provide a vehicle for transport and localization of specific mRNAs within the cell.

A channeled tRNA cycle during mammalian protein synthesis

In earlier studies it was shown that the mammalian translation system is highly organized in vivo and that the intermediates in the process, aminoacyl-tRNAs, are channeled--i.e., they are directly transferred from the aminoacyl-tRNA synthetases to the elongation factor to the ribosomes without dissociating into the cellular fluid. Here, we examine whether spent tRNAs leaving the ribosome enter the fluid phase or are transferred directly to their cognate aminoacyl-tRNA synthetases to complete a channeled tRNA cycle. Using a permeabilized CHO cell system that closely mimics living cells, we find that there is no leakage of endogenous tRNA during many cycles of translation, and protein synthesis remains linear during this period, even though free aminoacyl-tRNA is known to rapidly equilibrate between the inside and outside of these cells. We also find that exogenous tRNA and periodate-oxidized tRNA have no effect on protein synthesis in this system, indicating that they do not enter the translation machinery, despite the fact that exogenous tRNA rapidly distributes throughout the cells. Furthermore, most of the cellular aminoacyl-tRNA synthetases function only with endogenous tRNAs, although a portion can use exogenous tRNA molecules. However, aminoacylation of these exogenous tRNAs is strongly inhibited by oxidized tRNA; this inhibitor has no effect on endogenous aminoacylation. On the basis of these and the earlier observations, we conclude that endogenous tRNA is never free of the protein synthetic machinery at any stage of the translation process and, consequently, that there is a channeled tRNA cycle during protein synthesis in mammalian cells.

Supramolecular organization of the mammalian translation system

Although evidence suggests that the protein synthetic machinery is organized within cells, this point has been difficult to prove because any organization that might exist is lost upon preparation of the cell-free systems usually used to study translation in vitro. To examine this process under conditions more representative of the intact cell, we have developed an active protein-synthesizing system using Chinese hamster ovary (CHO) cells permeabilized with the plant glycoside saponin. This procedure renders cells permeable to trypan blue and exogenous tRNA, but there is little release of endogenous macromolecules. Protein synthesis in this system proceeds at the same rate as that in intact cells and is about 40-fold faster than that in a cell-free system prepared from the same cells. Active protein synthesis in this system requires the addition of only Mg2+, K+, and creatine phosphate, with a small further stimulation by ATP and an amino acid mixture; no exogenous macromolecules are necessary. The proteins synthesized in this system are indistinguishable from those made by the intact cell, and the channeling of aminoacyl-tRNA observed in vivo is maintained. Our data suggest that the permeabilized cell system retains the protein-synthesizing capabilities of the intact cell and presumably its internal structure as well. Studies with this system demonstrate that the protein-synthesizing apparatus is highly organized and that its macromolecular components are not freely diffusible in mammalian cells.

The morphological substrate of autonomic regulation of the bronchial epithelium

Observations of explanted bronchial mucosa show that ciliary function is maintained for 7 days subsequent to explanation. This finding demonstrates that non-neural mechanisms exist which regulate ciliary function. Ultrastructural and immunohistochemical studies both for light and electron microscopy were performed on human bronchial biopsy material and lung resection specimens in order to recognize the morphological substrate of this regulatory mechanism. A complex system of cytokeratin filaments and microtubules radiate through the whole cytoplasm of ciliated cells with direct contact to the nucleus, cilia, microvilli, desmosomes and to the apical terminal adhesive complex. Between the basal bodies and the apical terminal adhesive complex microfilaments can be found. In the apical cytoplasm a dense filamentary network is seen in association with the adhesive complex. These morphological findings indicate that the cytoskeleton of the bronchial epithelium plays a key role in the co-ordination of ciliary function.

Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes

We have studied transport and localization of MBP mRNA in oligodendrocytes in culture by microinjecting labeled mRNA into living cells and analyzing the intracellular distribution of the injected RNA by confocal microscopy. Injected mRNA initially appears dispersed in the perikaryon. Within minutes, the RNA forms granules which, in the case of MBP mRNA, are transported down the processes to the periphery of the cell where the distribution again becomes dispersed. In situ hybridization shows that endogenous MBP mRNA in oligodendrocytes also appears as granules in the perikaryon and processes and dispersed in the peripheral membranes. The granules are not released by extraction with non-ionic detergent, indicating that they are associated with the cytoskeletal matrix. Three dimensional visualization indicates that MBP mRNA granules are often aligned in tracks along microtubules traversing the cytoplasm and processes. Several distinct patterns of granule movement are observed. Granules in the processes undergo sustained directional movement with a velocity of approximately 0.2 micron/s. Granules at branch points undergo oscillatory motion with a mean displacement of 0.1 micron/s. Granules in the periphery of the cell circulate randomly with a mean displacement of approximately 1 micron/s. The results are discussed in terms of a multi-step pathway for transport and localization of MBP mRNA in oligodendrocytes. This work represents the first characterization of intracellular movement of mRNA in living cells, and the first description of the role of RNA granules in transport and localization of mRNA in cells.

Beta-actin mRNA-binding proteins associated with the cytoskeletal framework

Association of mRNA with the cytoskeletal framework (CSK) is thought to play a strategic role in the placement of mRNA in the cytoplasm. However, the molecular determinants underlying mRNA/CSK association are completely unknown. To begin addressing this issue, we have employed a binding assay to identify proteins of the CSK compartment of NIH 3T3 cells that bind in-vitro-transcribed 32P-labelled beta-actin mRNA with high affinity. Three proteins, of approximate molecular masses 27, 50 and 97 kDa, were observed to exhibit strong binding. Binding to these proteins took place at physiological salt concentration and withstood washing in 0.5 M salt. Furthermore, binding was unaffected by heparin but was inhibited by unlabelled beta-actin mRNA. Treatment of isolated CSKs with the microfilament-severing agent DNase I abolished all beta-actin mRNA-binding activities, thus suggesting a possible association of beta-actin mRNA with the microfilament network in situ. Removal of the 3' untranslated region (UTR) significantly reduced beta-actin mRNA binding to all three CSK proteins but removal of the 5' UTR mainly affected binding to the 97-kDa species and that to a lesser extent. beta-Tubulin mRNA bound to the same three CSK proteins as did beta-actin mRNA, but with considerably less avidity. In contrast, vimentin mRNA strongly recognized these CSK proteins, and further bound to a group of smaller proteins (< 29 kDa). As beta-actin mRNA, beta-tubulin mRNA and vimentin mRNA have been observed to occupy separate cytoplasmic locales, the proteins detected here may be operative both in binding mRNAs to the CSK in situ, as well as in localizing mRNA in the cytoplasm.

mRNA association with the cytoskeletal framework likely represents a physiological binding event

A multitude of studies has indicated that the vast majority of mRNA and polyribosomes is associated with the detergent-resistant cytoskeletal framework (CSK). However, the nature and purpose of this association remain unclear. To begin unraveling the factors which may mediate this phenomenon, we examined the extent of association of four mRNAs (tubulin, vimentin, actin, and histone mRNA) with the CSKs of NIH 3T3 cells over a wide range of salt concentrations. Results indicate that the vast majority (greater than 90%) of each of these mRNAs remains associated with the CSK after detergent extraction of cells in low ionic strength buffer (25 mM NaCl). This association is manifest under conditions that cause the complete depolymerization of microtubules but that leave microfilaments and intermediate filaments intact. Even after extensive washing in buffer of approximately physiological ionic strength (150 mM NaCl), 75-85% of these mRNAs still remain associated with the CSK. However, at least 50% of each of these mRNAs can be eluted from the CSK by washing with buffer containing 250 mM NaCl. Not all the mRNAs, though, display the same elution profile. This suggests that different binding sites and/or different binding affinities may exist for different mRNAs. Surprisingly, close to 50% of the polyribosome population remains bound to the CSK despite washing in as much as 1.0 M NaCl. These adherent polyribosomes appear to be of the same size as those that are eluted, allaying the possibility that they are retained by the CSK simply due to size exclusion. Collectively, these data strongly imply that mRNAs are neither weakly adsorbed to the CSK nor physically trapped within the meshwork of cytoskeletal filaments.(ABSTRACT TRUNCATED AT 250 WORDS)