Isolation of a residual protein structure from nuclei of the myxomycete Physarum polycephalum

A residual structural framework has been obtained when isolated nuclei of eukaryotic cells were treated with high salt concentrations and DNA-degrading enzymes (Berezney & Coffey, 1975, 1976; Comings & Okada, 1976). It was found to consist of the nuclear lamina and an internal nuclear matrix extending from the lamina inwards through the entire nuclear space. The proteins comprising the matrix were found by gel electrophoresis to be present in preparations of both the nuclear envelope and the nucleolus (Berezney & Coffey, 1976). This widespread distribution of structural proteins within the nuclear substructures may allow DNA to attach at specific sites throughout the nucleus such that the structural framework plays a primary role in the spatial organization and processing of both replicative and non-replicative DNA (Berezney & Coffey, 1976; Wanka et al. t<)fT, Mullenders, 1979).

The demonstration of a structural protein matrix in the ciliate Tetrahymena suggests a general occurrence, and possibly a general function, of the nuclear structural framework in eukaryotic cells (Wunderlich & Herlan, 1977; Herlan & Wunderlich, 1976). In contrast to the 3 high-molecular-weight proteins consistently found in preparations of the nuclear substructures from higher eukaryotes, the matrix of macronuclei from Tetrahymena is composed of 6 or more major polypeptides ranging from low to high molecular weight.

In this paper we extend these observations and demonstrate the existence of a structural protein framework in nuclei of the myxomycete Physarum polycephalum also, significantly, composed of predominantly low-molecular-weight polypeptides. These observations suggest there may be significant differences in the composition of the nuclear structural proteins between the higher and lower eukaryotes. The protein matrix from Physarum nuclei, similar to matrices from Tetrahymena and from higher eukaryotes, is associated with DNA and is structurally continuous both with the nuclear lamina and with a structural remnant of the nucleolus.

Microplasmodia of Physarum polycephalum, strain M₃c IV, were grown as described by Daniel & Baldwin (1964). Macroplasmodia were prepared and the stages of the nuclear cycle were determined as described previously (Schel & Wanka, 1973).

Nuclei and nucleoli were isolated, with slight modifications to the procedures, according to the methods of Mohberg & Rusch (1971). Briefly, the nuclei of Physarum were released into a medium containing 0·25 M sucrose, 5 mM MgCl₂, 10 Mm Tris-HCl, pH 7·2,0·1% Triton X-100 and 0·5 mM phenylmethylsulphonylfluoride (PMSF). After a brief centrifugation to remove cell debris the supernatant was filtered through 2 layers of milk filter (Brocades N.V., Nijmegen). Nuclei were collected from the filtrate by centrifugation at 1000 g for 15 min over an underlay of isolation medium containing 1 M sucrose. In order to increase the purity of the preparation and to stabilize the nuclei for the subsequent preparation of the nuclear protein matrix, the nuclei were washed once (and occasionally twice) in about 200 ml of isolation medium and were collected by centrifugation at 1000 g for 10 min.

Nucleoli were released from Physarum directly by homogenization into a medium with the same composition as described above, but containing 0·5 HIM MgCl₂ instead of 5 mM (Mohberg & Rusch, 197i). Contaminating nuclei were removed by centrifugation at 800 g for 5 min over an underlay of isolation medium containing 0·8 M sucrose. Nucleoli were pelleted from the supernatant by centrifugation at 1500 g for 15 min, and were washed once with isolation medium then pelleted at 1000 g for 10 min.

Nuclei were suspended at a concentration of about 3 × 10⁷ nuclei/ml in an ice-cold solution of 10 mM EDTA, pH 7·2, by gentle passage through a stainless steel needle (0·8 × 120 mm), and then left on ice for 15 min with occasional mixing by inversion. Nuclei were pelleted at 1500 g for 15 min and the extraction with EDTA was repeated. The EDTA-treated nuclei were then resuspended at a concentration of about 6 × 10⁷ nuclei/ml in a solution of 0·1 mM Tris-HCl, pH 6·8. After incubation on ice for 10 min in silicone-coated centrifuge tubes, a solution of buffered NaCl was slowly added to the suspension to a final concentration of 2·5 M NaCl and 50 mM Tris-HCl, pH 7·2. Extensive clumping of the nuclear material was avoided by slow addition of the salt solution.

This suspension was then passed 4 times through a 0·8-mm needle and left on ice for 15 min. A pellet of high salt-treated nuclear material was collected by centrifugation for 15 min at 5000 g. The pellet was resuspended into a solution of 20 mM Tris-HCl, pH 7·2, 1 M NaCl, 7·5 mM MgCl ₂ and 0·5 mM PMSF, firstly by swirling gently, then by careful passage several times through a needle. Gentle procedures were necessary when resuspending the high salt-treated nuclear pellet, as disruption or mechanical distortion readily occurred. DNase I and RNase A (both electrophoretically pure, Sigma Corp.) were added to a concentration of 20 and 10/4g/ml, respectively, and the suspension incubated at 30 °C for 2 h. The final residual nuclear material was pelleted at 16 000 g for 15 min.

Pellet samples, suspended in 2–3 ml of distilled water, were precipitated by addition of TCA to 25 % and were then washed twice with 5 % TCA, once with acidified acetone and finally with acetone before solubilization in an electrophoresis sample buffer containing 6 M urea (Laemmli, 197o). Slab gel electrophoresis was carried out in the presence of SDS according to Laemmli (1970), but contained a linear gradient of 6 to 18% polyacrylamide. Molecular weight was determined in either 10 or 15 % SDS-polyacrylamide gels using appropriate protein standards (Weber & Osborn, 1969). Physarum nuclear histones were isolated as described by Mohberg & Rusch (1971) and were prepared for electrophoresis by dialysis for 24 h against 2 changes of electrophoresis sample buffer.

Samples for chemical analysis were precipitated with 0·1N PCA containing 25 % ethanol, extracted with ethanol: ether (3:1) at 70 °C for 5 min, and then with 0·5 N PCA at 70 ^º C for 70 min. The remaining pellet was washed with PCA, then extracted with 1 M NaOH at too °C for 10 min. Insoluble material was hydrolysed in 30% KOH at 100 °C for 30 min. The 0·5 N PCA and alkaline extracts were analysed for hexose with anthrone (Weeks, 1954). The combined 0·5 N PCA extracts were analysed for total nucleotides by u.v. spectrophotometry and for DNA with diphenylamine (Burton & Peterson, 1957). Protein in the 1 M NaOH extract was determined by the method of Lowry, Rosebrough, Farr & Randall (1951). Because the nucleotide content of the nuclear residual material was barely at the level of chemical determination, the DNA and RNA contents were confirmed by isolating residual nuclear material from microplasmodia grown continuously in the presence of [5-³H]uridine and [2-¹⁴ C]thymidine (both products of NEN). Corrections for the incorporation of tritium label into DNA and protein were made assuming the values given by Hall & Turnock (1976).

For whole-mount electron microscopy, 2o-μl samples of high salt-treated nuclei suspended either in 10 mM Tris-HCl, pH 7·2 or in 40 % glycerol, Tris-HCl, pH 7·2 were allowed to settle directly onto copper grids covered with carbon-coated Formvar, or material was picked up by such grids from a 50-μl droplet of the same sample placed on a flat piece of Teflon. The adherent nuclear material was then depleted of DNA in situ by floating the grids on a 50-μl droplet of water containing 20 μg/ml of DNase I and7·5 mM MgCl₂for periods of up to 30 min. Excess fluid was removed from the samples with filter paper and the preparation was dehydrated in a graded series of ethanol washes. Positive staining was carried out with 1 % phosphotungstic acid in the 50 % ethanol step for 5 min. The samples were air-dried after a final step in pentane. The grids were examined with a Philips EM 201 electron microscope operating at 60 kV.

Interphase nuclei from the Physarum plasmodia were treated with EDTA and NaCl to remove both nucleoplasmic and chromatin proteins and the resulting high salt-treated nuclear structures (termed high salt-treated nuclei throughout the paper) were then incubated in the presence of DNase, RNase and 1 M NaCl to release remaining non-structural material. The final preparation, when observed under the phase-contrast microscope, consisted of faint ghost-like structures with diameters slightly larger than those of isolated nuclei. Omission of the 1 M NaCl from the nuclease digestion resulted in highly condensed spherical structures unsuitable for further structural studies.

The residual material was composed of 92% protein, 6% carbohydrate, about 2% DNA and 0·1% RNA (Table 1). The recovery of nuclear proteins from the final fraction varied between 17 and 22% over 8 preparations.

The apparent fragility of the residual structure from the Physarum nuclei prevented the satisfactory spreading of material for whole-mount electron microscopy, and most attempts resulted in highly condensed and collapsed structures. As a consequence, we resorted to in situ DNase digestion of high salt-treated nuclei adhered to the Formvar support to obtain well-spread samples of the residual nuclear structures. The features seen in relatively well-spread residual structures and in in situ DNase-treated high salt nuclei were similar.

In general the electron micrographs of whole-mount preparations showed coherent frameworks of 3 prominent structural components known from mammalian cell nuclei (Comings & Okada, 1976; Mullenders, 1979): a dense fibrous-globular matrix extending throughout the internal space of the nucleus and continuous with this, a more dense area at the site of the nucleolus (Fig. 1 A, B). The internal matrix was surrounded by a thin lamina containing residual nuclear pore structures. Occasionally large areas of the fibrous-globular lamina with associated annuli of the residual nuclear pores were exposed in disrupted nuclei (Fig. 1 c). The annuli showed outer diameters of about 100 nm and frequently contained central granules attached by fibrous stalks (Schel & Wanka, 1973)· Nuclear membranes were not visible in these preparations, presumably due to the removal of lipid and soluble protein components during the isolation of the nuclei in the presence of Triton.

Preparations of high salt-treated nuclei, not digested with DNase, showed large masses of thin threads associated with the residual nuclear structures. They probably consist of single and aggregated DNA molecules. Many of these DNase-sensitive strands emerged from the annular pore remnants (Fig. 1 D), supporting the supposition that nuclear pores provide binding sites for DNA (Schel & Wanka, 1973; Schel, Steenbergen, Bekers & Wanka, 1979).

The protein composition of nuclei and the residual nuclear material was analysed by SDS polyacrylamide gel electrophoresis. Fig. 2 shows that histones constitute the major class of proteins in isolated Physarum nuclei, although 2 of them have been largely lost to the Mg²⁺ -containing isolation medium during the preparation (Mohberg & Rusch, 1970).

The most prominent polypeptides of the residual nuclear structure were 2 bands of molecular weights 3 6 5 00 and 2 3 000 Daltons, respectively, and a minor band at about 52 000 Daltons was also present (Fig. 2, slot B). The significance of the 2 minor polypeptides with similarity in mobility to the histones is not yet clear. They might represent either marginal amounts of tightly bound histones or non-histone proteins whose electrophoretic mobilities fortuitously coincide with those of the histones.

Between 30 and 40 less-significant polypeptides, the majority of which fall into various molecular weight classes of between 23 000 and 115 000 Daltons, were also visible. Repeated extraction of the residual structure with either 2 M NaCl or 0·3 M MgCl₂, or repetition of the isolation procedure of EDTA and NaCl washes had little effect on the polypeptide patterns observed and, significantly, there was little loss of the 2 major proteins during any of the steps of the isolation of residual structural material from the nuclei. Occasionally minor variations between preparations were observed (Mohberg & Rusch, 1971).

Initial attempts, using the methods of Dwyer & Blobel (1976), to separate the nuclear lamina from the intranuclear components resulted in condensed nuclear structures, which in electron-microscopic preparations retained the internal protein matrix. The preparations displayed polypeptide patterns similar to the pattern shown by high salt-treated nuclei.

The polypeptide patterns of high salt-treated nuclei and residual nuclear structures were similar, except that several minor polypeptides of 59000 and 55000 Daltons molecular weight and residual histones were solubilized during the digestion of the high salt nuclear material with DNase I and RNase A (Fig. 3A and B) The loss of histones from the high salt-treated nuclei correlates with the solubilization of most of the remaining DNA and RNA (Table 1). In a further attempt to subfractionate the residual nuclear structure, nucleoli were isolated from Physarum and were then treated in the same manner as for the preparation of residual nuclear structures. The 2 major polypeptides of the residual nuclear structures were also amongst the most prominent polypeptides in these preparations (Fig. 3C). These observations are consistent with the observations made on whole-mount electron-microscopic preparations, and strongly suggest that the major polypeptides of the nuclear structures may contribute to the composition of a fibrous structural material which is common both to the matrix of the intranuclear space and to a nucleolar matrix (Berezney & Coffey, 1976). The complex fibrous, globular and annular structures seen in the electron-microscopic structures of eukaryotes must be composed of arrays of the relatively few polypeptides.

Although we have not further identified the residual nuclear polypeptides, the electrophoretic mobilities of the 2 major proteins indicate that they do not correspond to any of the familiar proteins such as tubulin, actin, or myosin which have themselves been isolated from Physarum nuclei (Jockusch, Ryser & Behnke, 1973; Le Stourgeon, Totten & Forer, 1974; Le Stourgeon et al. 1975). Several of the minor proteins of the nuclear structures may correspond to such proteins; for example, β-tubulin (52 000 Daltons) may be present as a prominent minor component (Fig. 3B). However, we cannot exclude the possibility that the similarity in mobilities may be simply fortuitous.

Preliminary studies of the residual nuclear structures at times throughout the mitotic cycle show little quantitative difference between electrophoretic patterns of the polypeptides (manuscript in preparation). In addition, extensive studies by Le Stourgeon and colleagues (Le Stourgeon & Rusch, 1973; Le Stourgeon, Nations & Rusch, 1973; Le Stourgeon et al. 1974) on changes in the patterns of Physarum nuclear acidic proteins during conditions of chromatin quiescence and activation induced by starvation and refeeding demonstrate that 2 nuclear acidic proteins of 37000 and approximately 23 000 Daltons undergo little quantitative variation during changes in the cell state. These observations suggest that the major elements of Physarum residual nuclear structure may be conserved during greatly differing states of chromatin activity.

The nuclear protein matrix and lamina of higher eukaryotes are believed to maintain the structure and organization of the nucleus (Berezney & Coffey, 1975, 1976; Herlan & Wunderlich, 1976), and to mediate the spatial ordering and processing of replicative and non-replicative DNA during replication and mitosis (Berezney & Coffey, 1976; Wanka et al.1977; Mullenders, 1979; Cook, Brazell & Jost, 1976). Preliminary data on Physarum also indicate an association of the replicative DNA with components of the nuclear structural framework (Wanka & Mitchelson, 1979), suggesting that it plays a similar important role in all eukaryotes.

The great differences in composition and molecular weights of the structural proteins of Physarum, higher eukaryotes and macronuclei of Tetrahymena are unexpected if one considers the ultrastructural similarities. The question arises whether the differences are fortuitous or whether they are related to functional differences, such as the type of nuclear division which occurs by closed mitosis, open mitosis and amitosis in Physarum, higher eukaryotes, and Tetrahymena macronuclei, respectively.