Detection of dna in the nucleoids of chloroplasts and mitochondria in Euglena gracilis by immunoelectron microscopy

Chloroplasts and mitochondria contain DNA which appears to be combined with proteins to form nucleoids in the organelles (Kuroiwa, 1982; Nemoto et al., 1989). In general, the DNA content of these organelles is far lower than that of nuclei. Thus, it was previously difficult to detect the DNA in situ and to visualize the shapes and distribution patterns of the nucleoids in the organelles.

Use of the dye DAPI for detection of nucleoids by fluorescence microscopy has increased the information that is available about the shape and the distribution pattern of nucleoids. Coleman (1978) was the first to succeed in visualizing the nucleoids in chloroplasts, and later she classified the chloroplast nucleoids in various eucaryotic algae into 6 groups (Coleman, 1985). Kuroiwa et al. (1981) classified the nucleoids in various plants into 5 groups. Changes in the number, size, and distribution pattern of chloroplast nucleoids during the life cycle or development of several algae have been investigated in some detail (Nakamura et al., 1986; Hatano and Ueda, 1987). Relatively little has been reported about the behavior of mitochondrial nucleoids during the cell cycle except in Physarum (Kuroiwa, 1982), Saccharomyces (Miyakawa et al., 1984), and Euglena (Hayashi and Ueda, 1989, 1992).

Studies of chloroplast and mitochondrial nucleoids by fluorescence microscopy are subject to the limits of the resolving power of the light microscope. Detailed morphological studies should be done by electron microscopy. By conventional electron microscopy, Kellenberger’s (1958) or related fixatives have often been used for the visualization of nucleoids or ‘DNA fibrils’. In most published reports, nucleoids appear as electron-transparent regions containing fine fibrils (Nass and Nass, 1963a,b; Nass et al., 1965; Yokomura, 1967a,b). In most cases, small numbers of ‘DNA fibrils’ are found in relatively large electron-transparent regions. It has not yet been determined whether entire transparent regions, without clear boundaries, actually correspond to the nucleoids. Moreover, no evidence has been presented from which we can infer that only the fibrous regions correspond to the nucleoids in both chloroplasts and mitochondria. That is, even the outer boundaries of the nucleoids in chloroplasts and mitochondria have yet to be determined. Accordingly, detailed morphological studies on organellar nucleoids are necessary using methods other than conventional electron microscopy.

Immunoelectron microscopy provides reliable information about the distribution pattern of specific antigens at the ultrastructural level. Using immunogold electron microscopy, we have visualized the outer boundaries of the nucleoids and have detected the possible distribution of DNA in the nucleoids of chloroplasts and mitochondria in Euglena gracilis as described below.

Cells of Euglena gracilis Z strain were cultured at 25°C in Cramer-Myers medium (Cramer and Myers, 1952), and the culture was illuminated by fluorescent light with a photon flux density of 150 µmol m^-2 s^-1 for 12 h per day.

For visualization of nucleoids in the mitochondria and chloroplasts, cells were treated with 5 µg/ml DAPI (4’6-diamidmo-2phenylindole; Sigma, St. Louis, MO, USA) and were observed by fluorescence microscopy as described previously (Hayashi and Ueda, 1989, 1992).

For immunoelectron microscopy, cells were fixed for 1 h at 4°C in phosphate buffer (pH 7.4) that contained 3% paraformaldehyde and 2% glutaraldehyde. After washing with water, they were dehydrated with ethanol and embedded in Lowicryl resin at 0°C. Polymerization of the resin was carried out at -25°C under UV light.

Serial sections, each of 70 nm thickness, were attached on Formvar coated nickel glids. Sections were treated successively with 3% H₂O₂ for 20 min, 3% bovine serum albumin (BSA) for 30 min and 30 µg/ml anti-DNA antibodies (Boehringer Mannheim Biochemica, Mannheim, Germany) for 2 h at room temperature. After washing with Tween-PBS (20 mM PBS plus 0.5% Tween-20), sections were treated for 1 h at room temperature with colloidal gold (5 nm in diameter) conjugated anti-mouse IgM antibodies (Kirkegaard & Perry Laboratories Inc., Gaithersburg, MD, USA). The gold particles were conjugated with the antibodies by the method of De May and Moeremans (1986). Sections were stained with an aqueous solution of 1% uranyl acetate and with lead citrate. To examine the specificity of the binding between DNA and the anti-DNA antibodies, treatment of sections with the antibodies was omitted but the rest of the procedure for detection of DNA was performed. For digestion of DNA, sections were immersed for 1 h at 30°C in a buffered solution (pH 5.6) that contained 0.1% DNase I (Worthington Biochem. Corp, NJ, USA) and 1 mM MgSO₄. After washing with water, sections were treated for detection of DNA with the antibodies.

Nucleoids in the chloroplasts and in the mitochondria, visualized by fluorescence microscopy after staining with DAPI, are shown in Fig. 1. Nucleoids in chloroplasts appeared as rods of 0.5-2 µm in length. Their diameters cannot be accurately measured because they are so fine. Nucleoids in the mitochondria appeared as dots. Most nucleoids emitted fluorescence of similar intensity, but one third of the nucleoids emitted stronger fluorescence.

Fig. 2A shows part of a nucleus, a chloroplast and a mitochondrion that were treated by the immunocytochemical procedure for detection of DNA. As was anticipated, prominent deposits of gold particles could be seen on the chromosomes in the nucleus. Similar deposits of gold particles were also visible at places both in the chloroplast and the mitochondrion. Treatment of sections with DNase during the immunocytochemical procedure eliminated depositions of gold particles in defined groups in nuclei, chloroplasts (Fig. 2B), and mitochondria (Fig. 2C). In the absence of treatment with the anti-DNA antibodies during the immunocytochemical procedure, no gold particles were also deposited in groups in nuclei, chloroplasts, or mitochondria (Fig. 2D).

Seven sectional profiles of chloroplasts are seen in Fig. 3. In each chloroplast, 5-8 small electron-dense regions of irregular shape are visible, which are the nucleoids. The high electron-density results from the deposits of gold particles on the nucleoids. The total area of the 7 chloroplasts is 27.5 µm² and that of nucleoids 0.51 µm². Thus the nucleoids occupy about 1.9% of the cross-sectional area of the chloroplasts.

Three serial sections of a portion of two chloroplasts are shown in Fig. 4 (A, B, C). In each panel, there are 6 to 8 regions in which gold particles are distinctly localized. Deposits of gold particles are seen in similar regions in the chloroplasts in all three panels, strongly suggesting that the particles were not deposited randomly. We can conclude that DNA was present in the three serial sections and that the gold particles were deposited as a result of the immunological reaction. The sites of the gold particles in the specific regions of the chloroplasts in the three panels were superimposed on one another to give Fig. 5. The nucleoid regions a, b, and c overlap completely with the respective corresponding regions in the three panels. Region d shifts to the left gradually as the sections advance from Fig. 4A to C. There are three regions designated e in Fig. 4A, and they shift to combine finally as region e in Fig. 4C. Region f appears in two sections (Fig. 4A,B), while regions g and h appear in only one section each. Nucleoids a-h are not completely included within the three sections, and parts of them may be contained in the sections adjacent to those shown in Fig. 4.

Fig. 5 shows that the nucleoids are twisted threads of 5070 nm in diameter. The diameters of these threads or the nucleoids cannot be accurately measured due to undetermined factors concerning the binding pattern between antigens, antibodies, and gold particles at specific regions, so that the figures are put as approximations. The left halves of nucleoids a, b, d, and e in Fig. 5 seem to be twisted spirally, and the central portions seem to be wound more than twice into thicker nodes of 100-200 nm in thickness. Gold particles are not distributed randomly in the nucleoids but are arranged along wavy or irregularly folded lines. These particle-lines may reflect the arrangement of DNA fibrils in the nucleoids. Transverse sections of the nucleoids reveal DNA-free central cores and surrounding DNA areas (Fig. 4, arrows). The cores appeared to be slightly more electron-dense than the stroma of the chloroplast. Central cores were 35-50 nm in diameter and the total diameter of the nucleoids, including the DNA areas, was 60-100 nm. The presence of central cores could also be recognized in oblique sections (Fig. 4, arrowheads). The central cores seemed not to be completely covered by DNA.

Some profiles of mitochondria were partially covered with deposits of gold particles. As in the cases of chloroplasts, deposits appeared in the same regions of mitochondria in 2 or 3 successive sections (Figs 6-9). The regions corresponding to the nucleoids were generally spherical or ovoid, 70-130 nm in diameter; the nucleoid in Fig. 7 is 100 nm and the nucleoids in Figs 6 and 8 are 120 nm in diameter. Therefore, the regions corresponding to nucleoids were seen in two successive sections in Fig. 7, and in three successive sections in Figs 6 and 8. Sometimes, large regions covered with gold particles were seen (Fig. 9). When sites of gold particles in two successive sections (Fig. 9B,C) were superimposed, Fig. 9D was obtained. The nucleoid in Fig. 9D appears to contain DNA threads, 50-70 nm in diameter, that are twisted spirally. DNA threads surrounded cores in several profiles (Fig. 9B,C, arrows). Cores of mitochondrial nucleoids could also be seen in nucleoids of average size (Fig. 8, arrow). Nucleoids were slightly more electron-dense than other regions of the mitochondria (Fig. 10A,B).

Regions with groups of particles located side by side or end to end were sometimes observed (Fig. 11A,B). Dumbbell-shaped nucleoids and paired nucleoids have been observed by fluorescence microscopy in mitochondria, and they have been interpreted as nucleoids in the process of division (Hayashi and Ueda, 1989). The groups of particles located side by side or end to end may correspond to dividing nucleoids.

Several reports have described investigations of the distribution of DNA in nuclei and chromosomes with DNA-specific antibodies (Hansmann and Falk, 1986; Thiry and Thiry-Blaise, 1989; Martin et al., 1992; Thiry, 1992). The presence of regions in chloroplasts and mitochondria in which antibody-conjugated gold particles were precipitated has been reported as an incidental observation in research processes on the nuclei. For example, Hansmann and Falk (1986) reported, in addition to their observations of nuclei and nucleomorphs, that DNA-containing regions were vis-ible around the center of chloroplasts and that isolated DNA-containing regions were visible in the mitochondria of Cryptomonas. In Thiry’s paper (1992), a photograph showing the deposition of gold particles on chloroplast and a similar photograph on the mitochondria of Chlamy domonas are included among photographs concerning the main topic of the detection of DNA within the nucleolus.

One of the essential procedures for immunocytochemical detection of extremely small structures is the examination of serial sections, which is necessary to confirm that deposits of gold particles appear at the same positions in organelles in successive sections. This analysis eliminates the possibility that non-specific or artifactual deposits will be misinterpreted as specific deposits. Without an examination of serial sections, deposits of antibody-coated gold particles do not provide convincing evidence for the location of DNA. Previous studies on the detection of organellar DNA failed to include this procedure.

In the present study, gold particles in each serial section were examined and compared with those in adjacent sections. The deposits in one section matched those in the adjacent sections. Thus, the possibility that the particles were deposited artifactually is clearly excluded, and the distribution pattern of antigen DNA in the two kinds of organelles appears convincing.

The figure constructed from 3 serial sections (Fig. 5) shows that the nucleoids in chloroplasts are rather evenly distributed in the inter-thylakoid regions and are constructed with twisted threads of 50-70 nm in diameter much more twisted than can be recognized by fluorescence microscopy (Fig. 1). The twisted threads were entangled to form thicker nodes of 100-200 nm in diameter. The way in which the threads are entangled in the nodes could not be clarified. As judged from the reconstructed images, nucleoids in the chloroplasts appeared to have no standard shape.

Most nucleoids in the mitochondria were spherical or ovoid and they were smaller than those in the chloroplasts. However, large nucleoids (Fig. 9) appeared to be composed of twisted threads of 50-70 nm in diameter, being fundamentally similar in structure to the nucleoids in the chloroplasts.

Mitochondrial nucleoids have been isolated from the mitochondria of Physarum (Suzuki et al., 1982). They are 0.25-0.3 µm in diameter and 0.7-1.7 µm in length, and they are composed of DNA filaments that are packed compactly into three-dimensional rod-shaped structures. They are larger than those in Euglena and similar in size to the nucleoids in the chloroplasts of Euglena. The twisting or the folding of DNA-containing threads seems to be more complicated in the nucleoids of the chloroplasts of Euglena than in the isolated mitochondrial nucleoids of Physarum.

One of the important findings in the present study is the presence of cores in the nucleoids. These cores were found in the nucleoids of both chloroplasts and mitochondria. There are two possibilities about the nature of such core; they may be real cores with some specific chemical constituents or they may be spaces surrounded by DNA threads. Since most cores were more electron-dense than matrixsubstances, it seems likely that they were real cores. We plan to examine whether the cores contain histones, RNA, or enzymes, such as RNA polymerases.

Around the nucleoid regions, which were specified by the deposition of gold particles, no special regions could be recognized that correspond to the electron-transparent zones of nucleoids seen in organelles fixed by Kellenberger’s or related fixatives (Nass and Nass, 1963a,b; Yokomura, 1967a,b). Moreover, the distribution of gold particles in the present work was very different from profiles of ‘DNA fibrils’ seen in organelles by those fixations. ‘DNA fibrils’ appeared in some cases as thick fibrils and in other cases as fine networks without any fundamental organization by such fixations. Kellenberger’s fixative may introduce artifacts by leaching out certain substances from the nucleoids and causing aggregation that results in the formation of ‘DNA fibrils’ and the electron transparent zones around them.

This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.