Perinuclear dense bodies: Characterization as DNA-containing structures using enzyme-linked gold granules

Interactions between the nucleus and the cytoplasm play a major role in gene expression and have been investigated in particular in the unicellular and uninucleate green alga Acetabularia (Schweiger & Berger, 1979).

The ultrastructure of the nucleus and the surrounding cytoplasm has been studied in detail (Franke et al. 1974; Berger & Schweiger, 1975; Liddle, Berger & Schweiger, 1976). Preliminary evidence suggested that the perinuclear region of Acetabularia, although it is extranuclear, contains DNA. The most striking features of this region are: (1) a 100nm thick perinuclear intermediate zone filled with filamentous material from which the characteristic cytoplasmic organelles and nuclear particles are excluded. (2) An adjacent lacunar labyrinth interlaced by many plasmatic junction channels connecting the intermediate zone and the cytoplasm. (3) Perinuclear dense bodies (PB), which are only found in close proximity to the intermediate zone and in particular to the junction channels (Fig. 1). The number of perinuclear dense bodies (PB) surrounding an Acetabularia primary nucleus is more than 20000 (Franke et al. 1974). Their structure is composed of fibrillar and granular material (Fig. 2). They are formed during the early vegetative phase of the life cycle and they disappear before the onset of nuclear division (Berger et al. 1975).

It has been suggested that the PB contain DNA since a characteristic bleaching of these bodies was observed with Bernhard’s staining technique (Bernhard, 1969; Franke et al. 1974; Spring, Franke, Falk & Berger, 1975; Liddle et al. 1976). Although this staining method cannot be considered as final proof of the extranuclear presence of DNA, it was nonetheless an interesting observation in that it was in contrast to the generally accepted dogma that DNA is located exclusively in nuclei and in cytoplasmic organelles. Nonorganelle cytoplasmic DNA has been described in different organisms by several authors (Kraszweska & Buchowicz, 1983; for references, see also Reid & Charlson, 1979) but it still represents an unorthodox view that needs further investigation.

Recently, specific enzyme-gold complexes have been used to localize substrates cytochemically (Bendayan, 1984). It was shown that nucleases attached to small colloidal gold beads are still active, undergo enzyme—substrate interactions and adhere to the substrate (Bendayan, 1981). This cytochemical approach has been applied to sections through the nuclear and perinuclear regions of Acetabularia cells.

Cells of Acetabularia mediterranea were grown in a 12h dark/12h light cycle in an artificial sea water as previously described (Schweiger, Dehm & Berger, 1977).

Cells in stages before cap formation were selected and fixed by immersion in 1% glutaraldehyde buffered with 0·1 M-phoβphate buffer (pH 7·4) at room temperature. After 15 min the cells were cut into 2-mm long pieces. Fixation was stopped after two more hours by rinsing the cell fragments six times in 0·1 M-phosphate buffer for 10 min each. The samples were not subjected to post-fixation with osmium tetroxide. They were dehydrated in graded series of ethanol and embedded in Epon. For labelling with antibodies the cells were fixed with 0·17% glutaraldehyde and 2% formaldehyde in 0·1 M-cacodylate buffer (pH7·4) for 90 min. The cells were cut into 2-mm long pieces after 15min fixation time and washed six times for 10min in 0·1% cacodylate buffer to remove the fixative. They were dehydrated and embedded in glycol methacrylate (GMA). Thin sections were mounted on 300-mesh nickel grids covered with carbon-coated collodion film.

For all steps where gold was involved plastic or siliconized glassware was used. The method for preparing gold granules and for loading the granules with protein followed that described by Frens (1973) and Horisberger (1981) in general.

For the preparation of colloidal gold 100 ml of 0·0l% tetrachloroauric acid (HAUC1₄) was heated. As soon as the solution started to boil 2 ml 1% Nascitrate was added. The solution was boiled for 5 min. During this period the gold chloride was reduced. The formation of colloidal gold was indicated by the appearance of a light red colour. Immediately after cooling down to room temperature the solution was adjusted to pH6·0 (for loading with DNase) or pH9·0 (for loading with RNase) or pH6·9 (for loading with protein A) with 0·2M-K₂CO₃.

For the preparation of the enzyme-gold complexes either 200µl DNase solution (10mgml⁻¹ DNase I) or RNase solution (5 mg ml⁻¹ bovine pancreas RNase A) was added to 1ml doubledistilled water and 50 ml of the pH-adjusted gold suspension was added dropwise with constant stirring. After 2 min 600 µl of the suspension was mixed with 100 µl 2·5 M-NaCl. If the colour of the enzyme-gold complex did not change from red to violet it was presumed that no gold particles were flocculated since all of them were coated with enzyme. After three more min of incubation 2·5 ml 1% polyethylene glycol (Carbowax 20 M) was added to stabilize the complexes.

Preparation of the protein A-gold complexes in general followed the same recipe. The minimal amount of protein A for loading of the gold granules was determined (Roth, Bendayan & Orci, 1978). A 10% excess of protein A was used in the preparation of the protein A-gold complexes.

The suspension containing the protein-gold complexes was centrifuged for 50 min at 40 000 g and 4°C to remove non-bound protein. The supernatant was carefully aspirated. The sediment was resuspended in 4ml PBS (phosphate-buffered saline; 0·02M-phosphate buffer, 0TS M-NaCl; pH 6·0 for the DNase-gold complex; pH 7·5 for the RNase-gold complex) or 0·0l M-Tris buffer; pH 8·0 (for the protein A-gold complex) containing 0·5 mgml⁻¹ polyethylene glycol. The suspension was gently shaken for 20 min at room temperature and centrifuged for 10 min at 1000 g to remove clumped granules. The centrifugation steps were repeated once more in order to wash off all non-bound molecules. The supernatant of the last 1000 g centrifugation step was used for the labelling experiments. The enzyme-gold complexes were stored at 4°C. They were freshly prepared every week.

For control experiments DNase inactivated by heating at 100°C for 10 min was bound to gold granules. Pilot and control experiments were performed on mouse liver sections that had been fixed and embedded under the same conditions as the Acetabularia cells.

For cytochemical labelling with enzymes the grids with the sections were floated for 5 min on PBS (pH6·0 for DNase labelling, pH7·5 for RNase labelling) containing 0·5mgml⁻¹ polyethylene glycol. The grids were transferred to the corresponding enzyme—gold complex and floated on the suspension for 45-55 min at either room temperature, S0°C, 34°C or 37°C. The grids were floated six times for 5 min on PBS plus polyethylene glycol, rinsed briefly with double-distilled water and stained with 5% uranyl acetate for 15 min.

For cytochemical labelling with antibodies the loaded grids were floated for 2 min on a solution containing inactivated (30 min 56°C) anti-rabbit immunoglobulin G (IgG) that had been diluted 1:10 with 0·01M-Tris buffer (pH8·0). The grids were then washed twice for 10 min with Tris buffer. Thereafter the grids were floated on a solution of anti-actin antibodies from rabbit that had been diluted 1:5 with Tris buffer for 2h at either room temperature or 34°C. Non-bound antibodies were removed by washing the grids five times for 10 min with Tris buffer. The grids were then transferred to gold-labelled protein A that had been diluted 1:5 with Tris buffer containing 0·5 mg ml⁻¹ polyethylene glycol. They were incubated for 60 min at room temperature. Non-bound protein A was removed by washing six times for 10 min with Tris buffer and polyethylene glycol. They were rinsed briefly with double-distilled water and stained with 5% uranyl acetate for 15 min.

The sections were examined with a Siemens I or Philips 400 T electron microscope.

Low concentrations of glutaraldehyde, short times for fixation and omitting osmification, as used in this study, affect the preservation of subcellular structures. But even under these conditions the major cellular structures of the Acetabularia cell can be identified unequivocally. No significant difference in labelling intensity was observed at different temperatures if RNase-coated gold granules were used.

In the case of DNase-gold particles the labelling was weak at room temperature and much more pronounced at higher temperatures. Optimum labelling with a low background was achieved at 34°C. At 37°C the labelling was as good but the background was increased.

Incubation with RNase-gold granules resulted in a distinct labelling of the nucleus (Fig. 3). Especially intense labelling was observed in the granular component of the nucleolus. The PB were also labelled but to a lesser extent (Fig. 4).

After incubation of sections with DNase-gold complexes labelling was observed over the whole nucleus. The labelling was denser in the vicinity of the nuclear membrane. It was rather weak in the nucleolus (Figs 5, 6). No labelling was found in the nucleolar vacuoles. The most striking feature of sections incubated with DNase-gold complexes was the labelling of the PB (Figs 5, 6, 7). The label was not evenly distributed but was preferentially observed close to or in the vicinity of the fibrillar component.

Ina control experiment sections were incubated with gold granules that had been loaded with inactivated DNase. In this case the DNase-gold complexes had lost their labelling capability and no labelling was found throughout the sections (Fig. 8).

Using the same RNase-gold or DNase-gold complexes on mouse liver sections the labelling occurred at the expected sites. Since actin is known to be a naturally occurring inhibitor of DNase I (Lazarides & Lindberg, 1974; Mannherz, Barrington Leigh, Leberman & Pfrang, 1975; Hitchcock, Carlsson & Lindberg, 1976; Zechel, 1980) it had to be determined whether gold-labelling of the sections after incubation with DNase is due to complexes formed between DNase and actin. Therefore, sections through the nuclear region were incubated with anti-actin antibodies and thereafter with gold-labelled protein A. The labelling of the sections was rather high whether the incubation was performed at room temperature or at 34°C.

A comparison of the labelling of different areas in the cell showed that there is no preferential binding of anti-actin antibodies in the PB (Table 1).

The technique of incubating sections with enzyme-gold complexes has been used to localize nucleic acids in the nuclear region of Acetabularia cells. Nucleases lose their capability of digesting nucleic acids after fixation with aldehydes and embedding in Epon (Monneron & Bernhard, 1966). However, the enzymes retain their ability to establish stable binding to their substrates, presumably at their active sites (Bendayan, 1981, 1982).

Studies on mouse liver sections by Bendayan (1981) and in our laboratory have shown that the enzyme-gold technique gives reliable results. The activity of the enzyme-gold complex strongly depends on the kind and duration of fixation as well as on the pH at which the sections are incubated.

The results presented in this paper indicate the presence of RNA in the PB of Acetabularia although the labelling is substantially weaker than in the nucleolus and resembles that of the nucleoplasm. This observation suggests that there is RNA in the PB but that the packing of the RNA is not very dense. On the other hand it could mean that the RNA in the PB is highly protected against the attachment of RNase.

Attachment of DNase-coated gold particles is mainly observed in the nucleoplasm and in the PB. The nucleoplasm is more or less evenly labelled with a slight preference for the vicinity of the nuclear membrane. This might mean that in the primary nucleus the DNA is evenly distributed throughout the nucleus and that its location is not restricted to a certain part of the giant primary nucleus.

The distinct labelling of the PB by the DNase-coated gold particles represents additional evidence for the occurrence of DNA in the perinuclear region, i.e. outside the nuclear membrane. Although it cannot be completely ruled out that the binding of the DNase-gold particles is non-specific, the failure of heat-inactivated DNase-gold granules to label indicates that the binding sites are specific for active DNase. Similar conclusions can be drawn from studies on mouse liver sections. Furthermore, the results obtained with DNase-gold particles are in good agreement with studies using Bernhard’s (1969) cytochemical method, which indicated the presence of DNA in the PB.

The question arises as to what is the role and the function of DNA in the PB. Incubation of sections with antibodies against the small subunit of cytoplasmic ribosomes (60 S) resulted in intense labelling of the granular region of the nucleolus and distinct labelling of the PB (unpublished results). These results indicate that amplified rRNA genes occur outside the primary nucleus of Acetabularia.

If in Acetabularia nuclear DNA occurs outside the nucleus the question may be asked whether the morphogenetic capability of anucleate cells might be due to extra-nuclear DNA. This, however, is highly improbable since the PB are found exclusively in close proximity to the nucleus. They are removed together with the nucleus when the rhizoid is removed.

The characterization of the structure and the function of the perinuclear dense bodies demands detailed studies on isolated preparations of these particles. Unfortunately, to date attempts to isolate the PB have been unsuccessful.

The authors thank Mrs Ingrid Bõnig and Miss Inge Pfrang for their skilful technical assistance. The photographic assistance of Mrs Renate Fischer as well as the secretarial assistance of Miss Christiane Schardt are highly appreciated. The authors acknowledge the English revision of the manuscript by Dr Mark Wiser.