Ultrastructural Localization of β-Actin and Amphoterin mRNA in Cultured Cells: Application of Tyramide Signal Amplification and Comparison of Detection Methods

Abstract

We describe a nonradioactive preembedding in situ hybridization protocol using digoxigenin-labeled RNA probes and tyramide signal amplification to increase the sensitivity of detection. The protocol is sensitive enough for electron microscopic localization of endogenous messenger RNAs encoding β-actin and amphoterin. Three visualization methods were compared: diaminobenzidine enhanced by nickel, Nanogold enhanced by silver and gold toning, and fluorescently labeled tyramides. Diaminobenzidine and Nanogold can be used in both light and electron microscopy. The nickel-enhanced diaminobenzidine was the most sensitive visualization method. It is easy to accomplish but a drawback is poor spatial resolution, which restricts its use at high magnifications. Nanogold visualization has considerably better spatial resolution and is therefore recommended for electron microscopy. Fluorescent tyramides, especially TRITC-tyramide, offer a good detection method for fluorescence and confocal microscopy. The methods were used to localize amphoterin and β-actin mRNAs in motile cells. Both mRNAs were found in the soma and cell processes. In double labeling experiments, β-actin mRNA localized to filamentous structures that also contained ribosomal proteins. Especially in the cortical cytoplasm, β-actin mRNA was associated with actin filaments. Direct localization to microtubules was only rarely seen.

Keywords

in situ hybridization electron microscopy tyramide signal amplification mRNA localization β-actin amphoterin

Subcellular compartmentalization of mRNA has been recognized as a mechanism for regulation of gene expression and protein sorting (Hesketh 1996; Nasmyth and Jansen 1997; Steward 1997; Gao 1998; Nung and Jan 1998). β-Actin mRNA is one of the best known examples. It locates in the leading edge of migrating cells (Lawrence and Singer 1986; Hoock et al. 1991) and in the growth cones of neurites in cultured neurons (Bassell et al. 1998). Targeting of the mRNA to the leading edge augments cell motility (Kislauskis et al. 1997), suggesting that localized protein synthesis is important for this cell function. Amphoterin was identified and cloned from rat brain as a molecule that binds heparin and promotes neurite outgrowth (Rauvala and Pihlaskari 1987). Amphoterin lacks a signal sequence, which suggests that it is synthesized in the cytoplasm outside the endoplasmic reticulum. The protein localizes to the cytoplasm and is particularly enriched at the leading edges of motile cells (Parkkinen et al. 1993) and at the filopodia and neurites in neuronal cells (Merenmies et al. 1991). The localization of amphoterin and β-actin proteins is therefore strikingly similar (Fages et al. 1997). Moreover, both mRNAs are found in the cell soma and growing cell processes in cultured neuroblastoma cells (Fages et al. 1997). It is possible that these cell motility-associated molecules are transported to the cell processes as mRNA and translated locally at the site at which the proteins are needed.

These recent findings have greatly added to the need for sophisticated electron microscopic localization methods for mRNA, which have not been widely available thus far. The ultrastructural localization of mRNA has mainly been limited by the low sensitivity of the methods described for in situ hybridization at the electron microscopic level. The specificity of mRNA localization is based on specific hybridization between the target and the complementary probe molecule (Gall and Pardue 1969). The sensitivity of detection and preservation of cell structure are dependent on several factors, including probe construction and labeling, conditions of hybridization, tissue fixation, pretreatments, embedding, and signal detection method. Le Guellec et al. (1992) compared ultrastructural in situ hybridization methods using preembedding and postembedding protocols as well as ultrathin frozen sections. They concluded that preembedding conserved the ultrastructure well but had poor resolution, whereas postembedding gave medium preservation but the lowest sensitivity. Frozen sections offered the highest sensitivity but suffered from poor ultrastructural preservation. Another drawback of the cryosection technique is that the orientation of cultured cells cannot be retained because no good methods exist for flat-embedding. This is particularly detrimental if cell processes and leading edges are to be studied, as is the case in the present study.

Here we describe a sensitive preembedding in situ hybridization protocol for β-actin and amphoterin mRNAs, both encoding cytoplasmic proteins that are synthesized outside the endoplasmic reticulum. β-Actin mRNA is abundant and expressed in practically all cell types and is widely used as a control in Northern blot hybridization. Amphoterin mRNA is mainly expressed in the developing nervous system and motile cells in culture (Merenmies et al. 1991; Parkkinen et al. 1993). The amount of amphoterin mRNA is about one fifth of the β-actin mRNA level in cultured cells (Fages et al. 1997).

Tyramide signal amplification (Bobrow et al. 1989) was recently introduced as a method to amplify signals in immunoelectron microscopy and in situ hybridization (Schöfer et al. 1997). It is based on the use of peroxidase activity to introduce labeled tyramide molecules to the reaction site. We have applied this amplification to enhance the sensitivity of detection. We used three signal detection methods: diaminobenzidine enhanced by nickel, ultrasmall gold enhanced by silver and gold toning, and fluorescent detection. Our method produced good sensitivity, spatial resolution, and ultrastructural preservation. We show localization of β-actin and amphoterin mRNA in the cell soma and processes and describe association of β-actin mRNAs with ribosomes and cortical actin filaments.

Materials and Methods

Cell Culture

Rat C6 glioma cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum at 37C in 5% CO₂. For the experiments, the cells were plated on glass (light microscopy) or plastic (electron microscopy) Lab-Tek chamber slides (Naperville, IL) or on plastic Lab-Tek coverslips coated with laminin (Sigma; St Louis, MO) (10 μg/ml in PBS, overnight at 4C). The cells were plated in DMEM containing 1% bovine serum albumin 3-4 hours before fixation. These conditions were used to produce well-spread, motile cells having many processes. In some experiments, the cells were extracted with 0.1% Triton X-100 in extraction buffer (10 mM Pipes, pH 6.9, 0.3 M sucrose, 0.1 M NaCl, 3 mM MgCl₂, 1 mM EGTA) for 60 sec at room temperature (RT) before fixation. This protocol removes the plasma membrane and most of the soluble cytoplasm, leaving the cytoskeleton intact, including the material that is associated with it (Lenk et al. 1977).

Probes

Linearized Bluescript plasmid containing the full-length coding region (650 BP) of rat amphoterin (Merenmies et al. 1991) or the Hinc II fragment (850 BP) of rat β-actin was transcribed in vitro using a DIG RNA labeling kit (Boehringer Mannheim; Mannheim, Germany). Incorporation of digoxigenin into the probes in the sense or anti-sense orientation was confirmed as instructed by the manufacturer. Some of the probes were partially hydrolyzed with sodium carbonate buffer to reduce the probe size.

In Situ Hybridization

The cell monolayers were fixed in 4% paraformaldehyde and 0.05% glutaraldehyde in 0.1 M Hepes buffer, pH 7.4, for 30 min at RT. In situ hybridization protocol was modified from Macville et al. (1995). After fixation, the cells were permeabilized in 0.1% saponin in PBS for 30 min at RT, then washed in PBS, and treated with 50 mM NH₄Cl in PBS for 15 min. The hybridization solution contained 60% deionized formamide, 2 x SSPE, and 500 μg/ml yeast total RNA (Boehringer) or 200 μg/ml salmon sperm DNA (Boehringer). The probes were used at a concentration of 20 ng/ml and hybridization was carried out at 42C for 16 hr. The cells were then washed in 60% formamide, 2 x SSC three times for 10 min at 42C, in 1 x SSC twice for 15 min at RT, and finally in PBS twice for 15 min. Nonspecific protein binding sites were blocked with 10% skimmed milk powder or 0.5% blocking reagent (NEN Life Science Products; Boston, MA) in 0.1 M Tris, pH 7.4, 170 mM NaCl, 0.1% saponin (TBS-saponin) at RT for 60 min. Digoxigenin epitopes were detected with sheep anti-digoxigenin F(ab) fragments coupled to horseradish peroxidase (Boehringer), diluted 1:50-1:400 in the blocking solution for 60 min. The cells were then washed in TBS-saponin five times for 5 min. Visualization of the peroxidase activity was performed with one of the three alternative methods.

Detection Methods

Direct Visualization of Peroxidase with Diaminobenzidine. DAB (Sigma) was diluted to 150 μg/ml in 50 mM Tris, pH 7.5. Nickel ions were added to the solution as diammonium nickel(II)sulfate 6-hydrate (BDH; Poole, UK) at 6 mg/ml to increase sensitivity (Tago et al. 1986). Finally, H₂O₂ was added to the mixture at 0.0003%. The mixture was kept on the cells for 20 min to 16 hr. Color development was followed under an inverted light microscope and the reaction was stopped by rinsing with 50 mM Tris. The nickel-enhanced DAB signal was unstable in glutaraldehyde and Mowiol. Therefore, the samples were brought directly to osmium tetroxide postfixation (electron microscopy) or to dehydration in a graded series of ethanol, air-drying, and DPX-embedding (light microscopy).

Tyramide Signal Amplification. The cells were incubated in biotinylated tyramide (NEN Life Science Products) diluted 1:50 in amplification diluent (NEN Life Science Products) at RT for 4-10 min. For fluorescence microscopy, tyramide coupled to tetramethylrhodamine (TRITC) or fluorescein (FITC) was used. The cells were then washed in TBS-saponin three times for 5 min. After biotinyl-tyramide, the biotin epitopes were detected with (a) streptavidin-peroxidase (NEN Life Science Products) diluted 1:100, (b) streptavidin-Nanogold (1.4 nm gold) (Nanoprobes; New York, NY) diluted 1:300-1:400, or (c) streptavidin-fluorescein (NEN Life Science Products) diluted 1:400 in TBS-saponin containing 1% bovine serum albumin or 0.5% blocking reagent (NEN Life Science Products), at RT for 1 hr. The cells were then washed in TBS-saponin four times for 5 min. Peroxidase activity (a) was visualized with DAB as described above. After Nanogold incubation (b), the samples were fixed in 1% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4 for 10 min. Nanogold was silver-enhanced with the HQ Silver kit (Nanoprobes) for 5-9 min once or twice and then toned with gold chloride (Liposits et al. 1982; Trembleau et al. 1994) to make the signal more stable for osmium tetroxide and irradiation. The cells were treated with 2% sodium acetate three times for 5 min, then with 0.05% gold chloride (Sigma) on ice in the dark for 10 min, and finally with a fresh solution of 0.3% sodium thiosulfate on ice in the dark twice for 10 min. The samples were then rinsed in distilled water and PBS and postfixed in osmium tetroxide. Samples for light microscopy were dehydrated, air-dried, and embedded in DPX. Samples for fluorescence microscopy (c) were embedded in Mowiol containing DABCO (Sigma) as an antifading agent.

Immunolabeling of Tubulin, Actin, and Ribosomal 60S Subunit

For double localization, TRITC-tyramide or Nanogold enhanced by silver and gold was first used for detection of mRNA. Cells extracted with Triton X-100 before fixation were used for electron microscopic experiments. In this case, the cells were fixed in 0.1% glutaraldehyde in 0.1 M phosphate buffer for 10 min before the silver enhancement step. After in situ hybridization, cytoskeletal filaments were labeled with monoclonal antibodies. Anti-tubulin (clone E7) was from Developmental Studies Hybridoma Bank (Iowa City, IA) and anti-actin was from Boehringer (clone C4) or Sigma (AC-15, anti-β-actin). Proteins of the ribosomal 60S subunit were labeled with a rabbit antibody kindly provided by Dr. John Hesketh (Horne and Hesketh 1990). The antibodies were diluted in 3% bovine serum albumin, 0.1% saponin in PBS (PBS-saponin) and incubated at RT for 1 hr. After washing in PBS-saponin, the antibodies were visualized with F(ab)₂ goat anti-mouse-FITC or anti-rabbit-FITC (Jackson Immuno Research Laboratories; West Grove, PA). F(ab)₂ goat anti-mouse 5-nm colloidal gold, or anti-rabbit 10-nm colloidal gold, (British BioCell; Cardiff, UK) was used for electron microscopy. After washing in PBS-saponin, the fluorescent samples were embedded in Mowiol+DABCO and the electron microscopic samples were fixed in 1% glutaraldehyde in 0.1 M phosphate buffer for 10 min before Epon embedding.

Epon Embedding

The cells were postfixed in 1% osmium tetroxide in 0.1 M phosphate buffer, pH 7.4, dehydrated in a graded series of ethanol, and flat-embedded in Epon. After dehydration, the cells were covered with a thin layer of Epon. The samples were then examined under an inverted microscope, and gelatin capsules filled with Epon were inverted to the monolayers on sites containing an adequate number of labeled cells. Thin sections were cut parallel to the culture substrate, mounted on single-slot copper grids, and stained with uranyl acetate and lead citrate.

Microscopy

The samples were examined and photographed under an Olympus AX-70 light microscope, a Zeiss Axiovert 135M inverted fluorescence microscope with BioRad MRC1024 confocal equipment, or a Jeol JEM 1200EX transmission electron microscope.

Results

Evaluation of the Detection Methods with Light Microscopy

Detection with Diaminobenzidine. β-Actin mRNA levels in spreading C6 cells were high enough to be detected by direct labeling with anti-digoxigenin-peroxidase followed by DAB incubation (Figure 1A). It was essential to use nickel ions in the DAB solution to enhance the signal (Tago et al. 1986). Cells labeled with the sense probe were completely negative (Figure 1B). Amphoterin mRNA levels were too low to be reliably visualized with this direct protocol.

Tyramide Amplification Followed by Streptavidin-Peroxidase and DAB or by Streptavidin-Gold. Tyramide was used to amplify the signals after hybridization with digoxigenin-labeled probe and labeling with anti-digoxigenin-peroxidase. Peroxidase activity causes covalent incorporation of tyramide (in this case biotinylated tyramide) molecules to the reaction site (Bobrow et al. 1989). Visualization of the bound biotin was then performed using either streptavidin-peroxidase followed by DAB (Figures 1C, 2A, and 2B), or streptavidin-Nanogold followed by silver enhancement (Figures 1D, 2C, and 2D. The amplification step greatly increased the sensitivity. For detection of β-actin mRNA, the DAB incubation time was shortened from overnight in the direct visualization (Figure 1A) to 25 min after a 10-min amplification with biotinyl-tyramide (Figure 1C). No background was detected with the β-actin sense probes (not shown). Amphoterin mRNA was also visible after tyramide amplification and a 40-min DAB incubation (Figure 2A). The samples hybridized with the amphoterin sense probe (Figure 2B) and those hybridized without any probe (not shown) started to develop some background after a 40-min DAB incubation. In DAB visualization, background can be avoided by monitoring the reaction under a light microscope.

Figure 1

In situ hybridization of β-actin mRNA in C6 cells. The cells were hybridized with digoxigenin-labeled anti-sense (A,C-E) or sense (B,F) probes and incubated with anti-digoxigenin-peroxidase (diluted 1:50 in A,B, 1:200 in C,D, 1:100 in E,F). The signals were amplified with biotinyl-tyramide (10 min in C,D, 4 min in E,F) and then detected with streptavidin-peroxidase (C), streptavidin-Nanogold (D), or streptavidin-FITC (E,F). Peroxidase activity was visualized with DAB and nickel (overnight in A,B, 25 min in C) or with silver enhancement (once for 8 min and once for 6 min, D). β-Actin mRNA is found in the cell soma as well as in the processes (arrow in E). Bars = 50 μm.

Figure 2

In situ hybridization of amphoterin mRNA in C6 cells. The cells were hybridized with digoxigenin-labeled anti-sense (A,C,E) or sense (B,D,F) probes and incubated with anti-digoxigenin-peroxidase (diluted 1:200). The signals were amplified with botinyl-tyramide for 10 min and then detected with streptavidin-peroxidase (A,B), streptavidin-Nanogold (C,D), or streptavidin-FITC (E,F). Peroxidase activity was visualized with DAB and nickel (40 min in A,B) or with silver enhancement (once for 8 min and once for 6 min, C,D). The arrowheads in E indicate a long cell process extending from the cell soma, containing the nucleus (n), to the growth cone, containing amphoterin mRNA (arrow). Bars = 50 μm.

Visualization with Nanogold and silver enhancement, after tyramide amplification for 10 min, was not as sensitive as DAB and nickel. β-Actin mRNA was readily detectable (Figure 1D), but amphoterin mRNA was only weakly visible, even after repeated incubations with the silver-enhancement solutions (Figure 2C).

In conclusion, visualization with DAB and nickel can be recommended for light microscopic experiments. It is less expensive and simpler to perform than ultrasmall gold and silver enhancement. In addition, the development of color can be followed during DAB incubation and the reaction can therefore be stopped once the desired signal intensity has been reached or background starts to appear. In our experiments, the anti-digoxigenin-peroxidase conjugate was the most obvious source of background. This was evident after longer tyramide amplification (Figure 2B), both in cells hybridized with the sense probe and in cells treated identically but without any probe. No background developed if anti-digoxigenin-peroxidase was omitted.

Detection for Fluorescence Microscopy

Two protocols were tested: indirect labeling with biotinyl-tyramide followed by streptavidin-FITC, and direct labeling with FITC or TRITC-tyramide. Indirect labeling was the most sensitive method. Both β-actin (Figure 1E) and amphoterin (Figure 2E) mRNA were readily visible with this protocol, without background in cells hybridized with the sense probes (Figures 1F and 2F). However, a longer biotinyl-tyramide amplification was necessary to visualize amphoterin mRNA. In direct detection, only TRITC-tyramide gave signal for amphoterin mRNA. TRITC-tyramide is therefore more sensitive than FITC-tyramide. In addition, TRITC-tyramide gave a better spatial resolution. It produced well-structured granular label (Figure 5), whereas FITC-tyramide gave a more diffuse signal (not shown).

Electron Microscopy

Visualization of the hybridized probes with DAB and nickel ions produced electron-dense precipitates that were visible as spots in the cytoplasm (Figure 3A). β-Actin was readily detectable without tyramide amplification (not shown), whereas a 10-min biotinyl-tyramide amplification step, followed by incubation with streptavidin-peroxidase, was necessary to reliably visualize amphoterin mRNA (Figure 3A). The ultrastructure of the cells, especially endoplasmic reticulum, endosomal vesicles, and the Golgi apparatus, was good (Figure 3A). Some of the mitochondria were partially swollen (not shown).

The use of ultrasmall gold (1-1.4 nm in diameter) is essential for penetration of the conjugates into cells if the use of drastic permeabilization methods is to be avoided to preserve the ultrastructure of membrane-bound organelles. Ultrasmall gold has the additional advantage of high sensitivity. Compared to 5- or 10-nm colloidal gold particles, it has been shown to markedly increase the sensitivity of labeling (Takizawa and Robinson 1994). Visualization of the hybridization signals with streptavidin coupled to 1.4-nm gold after biotinyl-tyramide amplification gave a good spatial resolution compared to DAB and nickel (Figures 3B and 4A-4C). β-Actin mRNA signals in particular occasionally formed rows (Figures 4A and 4B), suggesting that the mRNA is aligned along cytoskeletal filaments.

Silver enhancement is necessary for visualization of the ultrasmall gold in heavy metal-stained samples. Humbel et al. (1995) tested all commercially available silver enhancement kits and found the HQ Silver kit (that was used in our experiments) to be best suitable for electron microscopy. This kit is light-sensitive, which has the drawback that the development of enhancement cannot be followed under a light microscope. Unfortunately, the silver produced on the gold particles may be destroyed or displaced by the electron beam (Humbel et al. 1995). In our hands, the silver was often lost from the Epon sections even before observation in the electron microscope, leaving empty holes in the thin sections (not shown). To avoid this, we replaced the silver with gold according to the protocol of Liposits et al. (1982) as modified by Trembleau et al. (1994). The appearance of the resulting label was dependent on the length of the silver enhancement step before gold toning. Repeated silver enhancement, which was necessary to make the signal visible in the light microscope, produced small (about 40-80 nm in diameter) clusters of gold grains (Figures 3B and 4A-4C). Shorter silver enhancement produced smaller clusters that were not visible in the light microscope (Figure 6B).

In addition to producing larger gold clusters after gold toning, repeated silver enhancement also produced more intense labeling. This is probably due to the nonhomogeneous growth of ultrasmall gold particles in the silver enhancement step (Humbel et al. 1995). Relatively long enhancement appears to be necessary to enlarge all 1.4-nm particles to a visible size. The attempts to produce light microscopically visible silver particles in one enhancement step produced unspecific precipitation. We therefore chose to use a shorter silver enhancement and repeat it when necessary. The ultrastructural preservation of the cells after silver enhancement and gold toning steps was comparable to the structure seen after DAB detection (Figures 3B and 4A-4C).

Figure 3

In situ hybridization of amphoterin mRNA (A) and β-actin mRNA (B) in C6 cells. The cells were hybridized with digoxigenin-labeled anti-sense probes and incubated with anti-digoxigenin-peroxidase (diluted 1:200). The signals were amplified with biotinyl-tyramide (for 10 min in A and 5 min in B), and the cells were then incubated with streptavidin-peroxidase (A) or with streptavidin-Nanogold (B). The signals in A were visualized using DAB and nickel (45 min). Nanogold on B was visualized using silver enhancement (twice for 7 min). e, endosomal vesicle; ER, endoplasmic reticulum; G, Golgi apparatus; n, nucleus. Bars: A = 1 μm; B = 0.5 μm.

Figure 4

Localization of β-actin (A,B) and amphoterin (C) mRNAs to the processes of C6 cells. In situ hybridization was done as in Figure 3. (A) A region close to the tip of a cell process. Bar = 1 μm. The area marked with brackets in A is enlarged in B. The arrows indicate signals arranged in rows. (C) A region of a thin cell process containing amphoterin mRNA signals. Bars = 0.5 μm.

Subcellular Localization of β-Actin and Amphoterin mRNAs

Both β-actin and amphoterin mRNAs were found in the cell body and cell processes (Figures 1E, 2E, 3A, 3B, and 4A-4C). This is compatible with the role of both proteins in cell motility and process growth (Merenmies et al. 1991; Kislauskis et al. 1997).

Figure 5

Double labeling of β-actin mRNA (red) and tubulin (green, A,B) or actin (green, C,D) protein in C6 cells. mRNA was detected with TRITC-tyramide (amplification for 5 min) after in situ hybridization with digoxigenin-labeled anti-sense probe and anti-digoxigenin-peroxidase (dilution 1:200) labeling. Proteins were labeled with indirect immunofluorescence using FITC-conjugated secondary antibodies. B and D show enlarged areas from A and C, respectively. Small arrows indicate yellow or orange color, showing that the mRNA is present in close proximity to the protein. Arrowheads indicate mRNA granules aligned along a microtubule in A and B. Large arrows indicate areas devoid of mRNA labeling on one side of the nucleus (n). All panels show one optical section of 0.5-μm thickness. Bars = 25 μm.

Table 1

Summary of the detection methods

Detection method	Sensitivity	Resolution	Comments
DAB + Ni	Good	Average	Best suitable for light microscopy
Nanogold + silver enhancement	Average	Good	Best suitable for electron microscopy^a
Fluorescent tyramides	Best with TRITC-tyramide or indirectly via biotinyl-tyramide	Best with TRICT-tyramide	Suitable for confocal microscopy

If silver enhancement is used for electron microscopy, it is recommended to use gold toning before osmification.

In the cell body, the mRNA was restricted to certain regions of cytoplasm. Typically, β-actin mRNA was localized in the region rich in cytoplasm surrounding the nucleus, not reaching the thin peripheral area (Figures 5A and 5C). However, the cytoplasm at one side of the nucleus, especially if the nucleus was asymmetrically located, was devoid of label, as was a small area in the perinuclear cytoplasm (Figures 5A and 5C). Electron microscopic images suggested that the Golgi apparatus is located in the area devoid of β-actin mRNA (Figure 3B).

Association of β-Actin mRNA with the Cytoskeleton

To increase the contrast of the cytoskeletal filaments, we extracted the cells with Triton X-100 before fixation to wash out the plasma membrane and soluble cytoplasm. This also enabled the use of colloidal gold particles for double labeling of mRNA and proteins. Labeling for β-actin mRNA was detected in the extracted cells (Figures 6A-6D), suggesting that the mRNA is associated with the insoluble cytoskeleton. The mRNA was associated with characteristic knobbly, filamentous structures that also showed labeling for the proteins of ribosomal 60S subunit (Figure 6A). Some of the ribosomes are known to be associated with the cytoskeleton (Lenk et al. 1977; Hesketh 1996) but, because formamide treatment during the hybridization destroys their ultrastructure (Racca et al. 1997; Bassell et al. 1998), they were not morphologically identifiable in our samples.

Previous studies have shown that many mRNAs, including β-actin mRNA, are associated with the cytoskeleton, especially with actin and/or tubulin filaments (Singer et al. 1989; Taneja et al. 1992; Bassell et al. 1994a,b, 1998). Therefore, we wanted to test whether our in situ hybridization protocol has enough spatial resolution to demonstrate close proximity of β-actin mRNA with tubulin or actin filaments. Confocal microscopy of double labeled cells revealed that the mRNA labeling was present as granules and that many of the granules were aligned on, or present in close proximity to, microtubules (Figures 5A and 5B). Close association of mRNA granules with actin filaments was also observed (Figures 5C and 5D).

Overlap of fluorochromes at the light microscopic level can be caused either by localization in close proximity, or by real colocalization, of the two labels (Griffiths et al. 1993). To distinguish between these alternatives, we performed electron microscopic double labeling of mRNA and the cytoskeletal proteins. Double labeling of β-actin mRNA and tubulin showed that, although the mRNA was frequently found in close proximity to microtubules, it was seldom directly associated with them (Figure 6B). Because only short stretches of microtubules were visible in thin sections, we also looked at semithick (200-nm) sections, with identical results. However, labeling for β-actin mRNA was frequently found associated with actin protein. This was especially prominent in the cortical cytoplasm (Figures 6C and 6D). The double labeling experiments were repeated three times with identical results.

Discussion

Electron microscopic in situ hybridization has earlier been used to localize abundant mRNAs, such as messages for cytoskeletal proteins (Singer et al. 1989), respiratory chain enzymes (Escaig-Haye et al. 1991), tyrosine hydroxylase (Liposits et al. 1991), myosin heavy chain (Pomeroy et al. 1991), elongation factor and immediate early gene products (Macville et al. 1995), EGF receptor (Sibon et al. 1996), and calcium/calmodulin-dependent protein kinase II (Martone et al. 1996).

Figure 6

Double labeling of β-actin mRNA and ribosomal 60S subunit proteins (A), tubulin protein (B), and actin protein (C,D). The cells were extracted with Triton X-100 before fixation to wash out soluble cytoplasm. In situ hybridization was carried out as in Figure 3B. Silver enhancement was performed twice for 6 min (A,C,D) or once for 9 min (B). β-Actin mRNA is labeled by larger, irregular gold grains (indicated by arrows), 60S proteins by regular 10 nm gold (arrowheads in A), and tubulin and actin by regular 5-nm gold (arrowheads in B-D). (A) mRNA label (arrows) is associated with filamentous structures having a characteristic knobbly, electron-dense morphology. These structures also show labeling for the ribosomal proteins (arrowheads). (B) mRNA labeling (arrows) is not directly associated with microtubules. Arrowheads indicate a microtubule parallel to the section. (C,D) mRNA labeling (arrows) is associated with the cortical actin, which is indicated by the parallel arrowheads (at left in C; at top in D). Bar = 0.4 μm.

Messenger RNAs concentrated in the endoplasmic reticulum for translation have also been localized successfully (Lamperth et al. 1990; Le Guellec et al. 1990, 1991; Dirks et al. 1992; Dörries et al. 1993; Mitchell et al. 1993; Trembleau et al. 1994; Matsuno et al., 1995; Yi et al. 1995; Trembleau and Bloom 1996; Chevalier et al. 1997; Racca et al. 1997). Because of the low sensitivity, postembedding protocols have been most successfully used to localize very abundant RNAs, such as viral RNA (Lin et al. 1993; Egger et al. 1994; Bienz and Egger 1995), ribosomal RNA (Binder et al. 1986; Medina et al. 1995; Shaw et al. 1996), and small cytoplasmic RNAs (Farris et al. 1997).

Singer et al. (1989) localized β-actin mRNA in cytoskeletal preparations of cultured cells. Biotin-labeled probes were detected with anti-biotin and a colloidal gold-labeled secondary antibody. This method has the advantage of high spatial resolution. However, the pictures reveal that its sensitivity is relatively low, because the labeling is limited to sparsely located sites. Recently developed fluorescent in situ hybridization protocols (Bassell et al. 1998; Femino et al. 1998; and the present study) have revealed the very abundant occurrence of β-actin mRNA in the cytoplasm of cultured cells. We have demonstrated here that a short amplification with biotinylated tyramide, in combination with ultrasmall gold- or peroxidase-coupled streptavidin, raises the sensitivity of electron microscopic localization to the same level as that of fluorescent in situ hybridization. Using this protocol, we were able to localize both β-actin mRNA and the much less abundant amphoterin mRNA. Furthermore, our protocol is gentle enough to preserve the fine structure of the membrane-bound organelles so that ultrastructural details of mRNA localization could be studied. Table 1 presents a summary of the detection methods used in our study.

Amplification of the signal always decreases the spatial resolution of detection. We used tyramide signal amplification to enhance the sensitivity. In a quantitative study, Schöfer et al. (1997) showed that tyramide amplification for 5 min did not significantly decrease spatial resolution. However, longer tyramide amplification times increased the labeling also at a distance of 100 nm from the target. In the present study, tyramide amplification for 4-5 min was sufficient for β-actin mRNA visualization with Nanogold. Therefore, β-actin mRNA could be localized with good spatial resolution. Detection of amphoterin mRNA was difficult without a longer tyramide amplification. The fine structural resolution of this mRNA localization is therefore not equally reliable.

We observed that β-actin mRNA was restricted to certain areas of the cytoplasm (Figures 3B, 5A, and 5B). The compartmentalization phenomenon was also seen in cells extracted with Triton X-100 before fixation, suggesting that it was not caused by hindered diffusion of the probe. This result may be an indication of compartmentalization of the cytoplasm into structural areas and nonstructural channels, which are used for intracellular transport and protein synthesis (Luby-Phelps 1993).

Both β-actin and amphoterin mRNA localized as separate spots in the cytoplasm. Granular localization of mRNAs encoding proteins synthesized on cytoskeleton-bound polysomes has also been reported in earlier light microscopic studies. (Taneja et al. 1992; Ainger et al. 1993; Knowles et al. 1996; Bassell et al. 1998; Femino et al. 1998). The postulated mRNA granules may contain only one, or several, mRNA molecule(s), in addition to some transport and/or translation factors (Barbarese et al. 1995; Knowles et al. 1996; Bassell and Singer 1997; Bassell et al. 1998). Most β-actin mRNA labeling inside the cortical cytoplasm was associated to knobbly filamentous structures. We also detected immunolabeling for the ribosomal 60S subunit associated with these structures. This suggests that β-actin mRNA and ribosomes are associated, either for translation or for transport. Bassell (1993) has suggested that filament-associated proteins, rather than the cytoskeletal filaments themselves, are the actual molecules associated with mRNA. Recently, the microtubule-associated protein, MAP 1A, was reported to be a component of a general RNA binding protein complex (DeFranco et al. 1998). The nature of the supramolecular RNA binding complex, or putative transport granule, remains a topic of future studies.

Our recent experiments using drugs that disrupt microtubules or microfilaments showed that both μ-actin and amphoterin mRNAs are transported along and anchored to the actin microfilament network (not to microtubules) in neuroblastoma cells (Fages et al. 1997) and in C6 cells (unpublished observations). Actin-dependent localization of mRNA has also been described in fibroblasts (Singer et al. 1989; Sundell and Singer 1991; Bassell et al. 1994a) and yeast cells (Takizawa et al. 1997). These results are in agreement with the present study showing that β-actin mRNA can associate, especially in the cortical cytoplasm, with actin filaments in C6 cells. However, in neurons, microtubules appear to be more involved in the transport and localization of mRNA (Bassell et al. 1994b,1998).

Footnotes

Acknowledgments

We thank Mervi Lindman for excellent technical assistance and Dr John Hesketh for the ribosome antibody.

References

Ainger

Avossa

Morgan

Hill

Barry

Barbarese

Carson

(1993) Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J Cell Biol 123:431–441.

Barbarese

Koppel

Deutscher

Smith

Ainger

Morgan

Carson

(1995) Protein translation components are colocalized in granules in oligodendrocytes. J Cell Sci 108:2781–2790.

Bassell

(1993) High resolution distribution of mRNA within the cytoskeleton. J Cell Biochem 52:127–133.

Bassell

Powers

Taneja

Singer

(1994a) Single mRNAs visualized by ultrastructural in situ hybridization are principally localized at actin filament intersections in fibroblasts. J Cell Biol 126:863–876.

Bassell

Singer

(1997) mRNA and cytoskeletal filaments. Curr Opin Cell Biol 9:109–115.

Bassell

Singer

Kosik

(1994b) Association of poly(A) mRNA with microtubules in cultured neurons. Neuron 12:571–582.

Bassell

Zhang

Byrd

Femino

Singer

Taneja

Lifshitz

Herman

Kosik

(1998) Sorting of β-actin mRNA and protein to neurites and growth cones in culture. J Neurosci 18:251–265.

Bienz

Egger

(1995) Immunocytochemistry and in situ hybridization in the electron microscope: combined application in the study of virus-infected cells. Histochemistry 103:325–338.

Binder

Tourmente

Roth

Renaud

Gehring

(1986) In situ hybridization at the electron microscope level: localization of transcripts on ultrathin sections of Lowicryl K4M-embedded tissue using biotinylated probes and protein A-gold complexes. J Cell Biol 102:1646–1653.

10.

Bobrow

Harris

Shaugnessy

Litt

(1989) Catalyzed reporter deposition, a novel method of signal amplification. Application to immunoassays. J Immunol Methods 125:279–285.

11.

Chevalier

Michel

Tang

X-M

(1997) Biotin and digoxigenin as labels for light and electron microscopy in situ hybridization probes: where do we stand?. J Histochem Cytochem 45:481–491.

12.

DeFranco

Chicurel

Potter

(1998) A general RNA-binding protein complex that includes the cytoskeleton-associated protein MAP 1A. Mol Biol Cell 9:1695–1708.

13.

Dirks

van Dorp

AGM

van Minnen

Fransen

JAM

van der Ploeg

Raap

(1992) Electron microscopic detection of RNA sequences by non-radioactive in situ hybridization in the mollusk Lymnea stagnalis . J Histochem Cytochem 40:1647–1657.

14.

Dörries

Bartsch

Nolte

Roth

Schachner

(1993) Adaptation of a non-radioactive in situ hybridization method to electron microscopy: detection of tenascin mRNAs in mouse cerebellum with digoxigenin-labeled probes and gold-labeled antibodies. Histochemistry 99:251–262.

15.

Egger

Troxler

Bienz

(1994) Light and electron microscopic in situ hybridization: non-radioactive labeling and detection, double hybridization, and combined hybridization-immunocytochemistry. J Histochem Cytochem 42:815–822.

16.

Escaig-Haye

Grigoriev

Peranzi

Lestienne

Fournier

J-G

(1991) Analysis of human mitochondrial transcripts using electron microscopic in situ hybridization. J Cell Sci 100:851–862.

17.

Fages

Huttunen

Nolo

Punnonen

E-L

Wartiovaara

Lahtinen

Rauvala

(1997) Amphoterin mRNA and protein in the processes of motile cells. Soc Neurosci Abst 23:601.

18.

Farris

Puvion-Dutilleul

Puvion

Harlev

Lee

(1997) The ultrastructural localization of 60-kDa Ro protein and human cytoplasmic RNAs: association with novel electron-dense bodies. Proc Natl Acad Sci USA 94:3040–3045.

19.

Femino

Fay

Fogarty

Singer

(1998) Visualization of single RNA transcripts in situ. Science 280:585–590.

20.

Gall

Pardue

(1969) Formation and detection of RNA-DNA hybrid molecules in cytological preparations. Proc Natl Acad Sci USA 63:378–383.

21.

Gao

F-B

(1998) Messenger RNAs in dendrites: localization, stability, and implications for neuronal function. Bioessays 20:70–78.

22.

Griffiths

Parton

Lucocq

van Deurs

Brown

Slot

Geuze

(1993) The immunofluorescent era of membrane traffic. Trend Cell Biol 3:214–219.

23.

Hesketh

(1996) Sorting of messenger RNAs in the cytoplasm: mRNA localization and the cytoskeleton. Exp Cell Res 225:219–236.

24.

Hoock

Newcomb

Herman

(1991) β-Actin and its mRNA are localized at the plasma membrane and the regions of moving cytoplasm during the cellular response to injury. J Cell Biol 112:653–664.

25.

Horne

Hesketh

(1990) Immunological localization of ribosomes in striated rat muscle. Evidence for myofibrillar association and ontological changes in the subsarcolemmal:myofibrillar distribution. Biochem J 268:231–236.

26.

Humbel

Sibon

OCM

Stierhof

Y-D

Schwartz

(1995) Ultra-small gold particles and silver enhancement as a detection system in immunolabeling and in situ hybridization. J Histochem Cytochem 43:735–737.

27.

Kislauskis

Zhu

Singer

(1997) β-Actin messenger RNA localization and protein synthesis augment cell motility. J Cell Biol 136:1263–1270.

28.

Knowles

Sabry

Martone

Deerick

Ellisman

Bassell

Kosik

(1996) Translocation of RNA granules in living neurons. J Neurosci 16:7812–7820.

29.

Lamperth

Manuelidis

Webster

(1990) P0 glycoprotein mRNA distribution in myelin-forming Schwann cells of the developing rat trigeminal ganglion. J Neurocytol 19:756–764.

30.

Lawrence

Singer

(1986) Intracellular localization in messenger RNAs for cytoskeletal proteins. Cell 45:407–415.

31.

Le Guellec

Frappart

Desprez

(1991) Ultrastructural localization of mRNA encoding for the EGF receptor in human breast cell cancer line BT20 by in situ hybridization. J Histochem Cytochem 39:1–6.

32.

Le Guellec

Frappart

Willems

(1990) Ultrastructural localization of fibronectin mRNA in chick embryo by in situ hybridization using ³⁵S or biotin labeled cDNA. Biol Cell 70: 159–165.

33.

Le Guellec

Trembleau

Pechoux

Gossard

Morel

(1992) Ultrastructural non-radioactive in situ hybridization of GH mRNA in rat pituitary gland: pre-embedding vs ultra-thin frozen sections vs post-embedding. J Histochem Cytochem 40: 979–986.

34.

Lenk

Ransom

Kaufmann

Penman

(1977) A cytoskeletal structure with associated polyribosomes obtained from HeLa cells. Cell 10:67–78.

35.

Lin

N-S

Chen

C-C

Hsu

Y-H

(1993) Post-embedding in situ hybridization for localization of viral nucleic acid in ultra-thin sections. J Histochem Cytochem 41:1513–1519.

36.

Liposits

Görcs

Gallyas

Kosaras

Sétáló

(1982) Improvement of the electron microscopic detection of peroxidase activity by means of the silver intensification of the diaminobenzidine reaction in the rat nervous system. Neurosci Lett 31:7–11.

37.

Liposits

Petersen

Paull

(1991) Amplification of the in situ hybridization signal by silver postintensification: the biotin-dUTP-streptavidin-peroxidasediaminobenzidine-silver-golddetection system. Histochemistry 96:339–342.

38.

Luby-Phelps

(1993) Effect of cytoarchitecture on the transport and localization of protein synthetic machinery. J Cell Biochem 52:140–147.

39.

Macville

MVE

Wiesmeijer

Dirks

Fransen

JAM

Raap

(1995) Saponin pre-treatment in pre-embedding electron microscopic in situ hybridization for detection of specific RNA sequences in cultured cells: A methodological study. J Histochem Cytochem 43:1005–1018.

40.

Martone

Pollock

Jones

Ellisman

(1996) Ultrastructural localization of dendritic messenger RNA in adult rat hippocampus. J Neurosci 16:7437–7446.

41.

Matsuno

Ohsugi

Utsunomiya

Takekoshi

Sanno

Osamura

Watanabe

Teramoto

Kirino

(1995) Changes in the ultrastructural distribution of prolactin and growth hormone mRNAs in pituitary cells of female rats after estrogen and bromocriptine treatment, studied using in situ hybridization with biotinylated oligonucleotide probes. Histochem Cell Biol 104: 37–45.

42.

Medina

Cerdido

Fernández-Gómez

(1995) Components of the nucleolar processing complex (pre rRNA, fibrillarin, and nucleolin) colocalize during mitosis and are incorporated to daughter cell nucleoli. Exp Cell Res 221:111–125.

43.

Merenmies

Pihlaskari

Laitinen

Wartiovaara

Rauvala

(1991) 30-kDaheparin-binding protein of brain (amphoterin) involved in neurite outgrowth. J Biol Chem 266:16722–16729.

44.

Mitchell

Gambiez

Beauvillain

(1993) Fine-structural localization of proenkephalin mRNAs in the hypothalamic magnocellular dorsal nucleus of the guinea pig: a comparison of radioisotopic and enzymatic in situ hybridization methods at the light- and electron-microscopic levels. Cell Tissue Res 274:219–228.

45.

Nasmyth

Jansen

R-P

(1997) The cytoskeleton in mRNA localization and cell differentiation. Curr Opin Cell Biol 9:396–400.

46.

Nung

Jan

(1998) Asymmetric cell division. Nature 392:775–778.

47.

Parkkinen

Raulo

Merenmies

Nolo

Kajander

Baumann

Rauvala

(1993) Amphoterin, the 30-kDa protein in a family of HMG1-type polypeptides. Enhanced expression in transformed cells, leading edge localization, and interactions with plasminogen activation. J Biol Chem 268:19726–19738.

48.

Pomeroy

Lawrence

Singer

Billings-Gagliardi

(1991) Distribution of myosin heavy chain mRNA in embryonic muscle tissue visualized by ultrastructural in situ hybridization. Dev Biol 143:58–67.

49.

Racca

Gardiol

Triller

(1997) Dendritic and postsynaptic localization of glycine receptor α subunit mRNA. J Neurosci 17: 1691–1700.

50.

Rauvala

Pihlaskari

(1987) Isolation and some characteristics of an adhesive factor of brain that enhances neurite outgrowth in central neurons. J Biol Chem 262:16625–16635.

51.

Schöfer

Weipoltshammer

Almeder

Wachtler

(1997) Signal amplification at the ultrastructural level using biotinylated tyramides and immunogold detection. Histochem Cell Biol 108: 313–319.

52.

Shaw

Thompson

Sinden

(1996) Localization of ribosomal RNA and Pbs21-mRNA in the sexual stages of Plasmodium berghei using electron microscope in situ hybridization. Eur J Cell Biol 71:270–276.

53.

Sibon

OCM

Wansink

Boonstra

Humbel

Verkleij

Cremers

FFM

(1996) EGF-receptor RNA metabolism in the nucleus of A431 cells. Eur J Cell Biol 71:45–52.

54.

Singer

Langevin

Lawrence

(1989) Ultrastructural visualization of cytoskeletal mRNAs and their associated proteins using double label in situ hybridization. J Cell Biol 108:2343–2353.

55.

Steward

(1997) mRNA localization in neurons: a multipurpose mechanism?. Neuron 18:9–12.

56.

Sundell

Singer

(1991) Requirement of microfilaments in sorting of actin messenger RNA. Science 253:1275–1277.

57.

Tago

Kimura

Maeda

(1986) Visualization of detailed acetyl-cholinesterase fiber and neuron staining in rat brain by a sensitive histochemical procedure. J Histochem Cytochem 34: 1431–1438.

58.

Takizawa

Robinson

(1994) Use of 1.4-nm immunogold particles for immunocytochemistry on ultra-thin cryosections. J Histochem Cytochem 42:1615–1623.

59.

Takizawa

Sil

Swedlow

Herskowitz

Vale

(1997) Actin-dependent localization of an RNA encoding a cell-fate determinant in yeast. Nature 389:90–93.

60.

Taneja

Lifshitz

Fay

Singer

(1992) Poly(A)RNA codistribution with microfilaments: evaluation by in situ hybridization and quantitative digital imaging microscopy. J Cell Biol 119: 1245–1260.

61.

Trembleau

Bloom

(1996) Spatial segregation of Gas mRNA and vasopressin mRNA to distinct domains of the rough endoplasmic reticulum within secretory neurons of the rat hypothalamus. Mol Cell Neurosci 7:17–28.

62.

Trembleau

Morales

Bloom

(1994) Aggregation of vasopressin mRNA in a subset of axonal swellings of the median eminence and posterior pituitary: light and electron microscopic evidence. J Neurosci 14:39–53.

63.

Michel

Sassy-Prigent

Chevalier

(1995) Electron microscopic location of mRNA in the rat kidney: improved post-embedding in situ hybridization. J Histochem Cytochem 43:801–809.