chromatin from the unicellular red alga porphyridium has a ...

J. Cell Set. 57, 151-160 (198a) Printed in Great Britain © Company of Biologists Limited 198a

CHROMATIN FROM THE UNICELLULAR RED ALGA PORPHYRIDIUM HAS A NUCLEOSOME STRUCTURE K. L. BARNES*, R. A. CRAIGIE, P. A. CATTINI AND T. CAVALIER-SMITHf Department of Biopkysics, King's College, London University, 26-29 Drvry Lone, London WC2B 5RL, U.K.

SUMMARY

We have isolated a crude nuclear preparation from the unicellular red alga Porphyridium aerugineum and investigated the structure of Porphyridium chromatin. Electrophoresis of deproteinized DNA fragments produced by micrococcal nuclease digestion of Porphyridium nuclei gives a typical ladder pattern, indicative of a repeating structure. The DNA repeat-length, calculated from plots of multimer length against multimer number, varies somewhat between different digestions, ranging from 160 to 180 base-pairs (average 173). We interpret this as evidence of heterogeneity in repeat-length; the calculated repeat-length depends on the extent of digestion because chromatin sub-populations with longer repeatlengths are on average digested earlier. Polyacrylamide/sodium dodecyl sulphate gel electrophoresis of basic proteins purified from Porphyridium nuclear preparations gives a pattern characteristic of core histones. Although our interpretation is complicated by some degradation, the result strongly suggests that Porphyridium chromatin contains each of the four core histones and that they are similar to those of higher eukaryotes. This, together with the micrococcal nuclease digestion results, demonstrates that Porphyridium chromatin is not fundamentally different from that of higher eukaryotes. INTRODUCTION The nuclear DNA of all eukaryotic cells, except for the Dinophyceae (Loeblich, 1976), is associated with histone proteins. The histone-DNA complex is organized as repeating units, nucleosomes, each containing approximately 200 base-pairs of DNA. Each nucleosome consists of an invariant length of 146 base-pairs of DNA wrapped around a core containing four species of histone, and a variable length of DNA between nucleosome cores (McGee & Felsenfeld, 1980). This linker DNA is usually associated with a fifth species of histone, usually Hi (Allan, Hartman, CraneRobinson & Avkes, 1980). Studies of nucleosome organization have been largely confined to animals in which each nucleosome is associated with approximately 200 base-pairs of DNA. However, the average repeat-length can vary greatly, even within the same tissue, e.g. from 197 base-pairs in rabbit non-astrocytic glial cells to 162 base-pairs in rabbit cortical neurons (Thomas & Thompson, 1977). There is also • Present address: Biophysics Laboratories, Portsmouth Polytechnic, White Swan Road, Portsmouth, U.K. f To whom correspondence should be addressed.

152

K. L. Barms, R. A. Craigie, P. A. Cattini and T. Cavalier-Smith

some heterogeneity in nucleosome spacing within single cells (Humpries, Young & Carroll, 1979); the figures obtained from micrococcal nuclease digestion are average repeat-lengths. Studies of fungi (Lohr & Van Holde, 1975; Noll, 1976; Morris, 1976; Thomas & Furber, 1976; Silver, 1979), protozoa (Gorovsky & Keevert, 1975; Johnson et al. 1976; Lipps & Morris, 1977) and one alga (Shupe, Rizzo & Johnson, 1980) have also revealed a large variation in repeat-lengths, varying from 154 basepairs in Aspergillus (Morris, 1976), to 220 base-pairs in Olisthodiscus (Shupe et al. 1980). The significance of these variations is not understood. It is possible that each major group of lower eukaryotes has a characteristic and distinctive nucleosome organization and a repeat-length that is highly conserved in evolution and would therefore be a valuable tool for unravelling protist phylogeny. Data from a wider variety of lower eukaryotes are needed to determine whether this is the case, or whether nucleosome repeat-length is rather variable within a group and therefore responsive to shorter-term evolutionary pressures. Studies of nucleosome organization in lower eukaryotes may also throw light on the origin of nuclei and the eukaryotic cell (Cavalier-Smith, 1975, 1981a). Red algae have often been argued to be the most primitive eukaryotes because of the close resemblance of their photosynthetic machinery to that of the prokaryotic cyanophytes (Allsopp, 1969; Klein, 1970; Cavalier-Smith, 1975, 1978; Chadefaud, 1976). However, acceptance of the symbiotic theory of plastid origin (Mereschowsky, 1905; Cavalier-Smith, 1982) makes this much less likely. Nevertheless, red algae are such a distinctive group, deserving classification with the Glaucophyceae as a distinct subkingdom, the Biliphyta (Cavalier-Smith, 19816, 1982), that it would be of considerable evolutionary interest to learn more about their nuclear organization. Such studies are limited by the absence of biochemical methods for isolating red algal nuclei. We have therefore devised a method for obtaining a crude nuclear fraction from the unicellular red alga Porphyridium aerugineum (which is actually blue-green in colour). We show by digestion of nuclei with micrococcal nuclease, followed by electrophoresis of the resulting DNA fragments on agarose gels, that Porphyridkan chromatin has a nucleosome structure with a repeat-length of 160-180 base-pairs. MATERIALS AND METHODS

Culture methods P. aerugineum (no. I38o/Starr), obtained from the Culture Collection of Algae and Protozoa, Cambridge, was cultured in artificial fresh water medium (Gantt, Edwards & Conti, 1968) with minor modifications (Barnes, 1979). Batch cultures were grown in 1-5 1 vol. in 2 1 Erlenmeyer flasks at 20-25 °C in continuous light supplied by fluorescent tubes and aerated with filtered air at a rate sufficient to keep the cells in suspension.

Isolation of nuclei Harvesting the cells and isolation of the nuclei was carried out at 4 °C. Cells were harvested by centrifugation at 7500 g for 20 min in a Sorvall RC2B centrifuge, using a GSA rotor. Nuclei were isolated by the following modification of the method of Walmsley & Davies (1975) for chicken erythrocyte nuclei. The pelleted cells were resuspended in 10 ml of lysis buffer (15 % sucrose, 6 mM-MgClt, a mM-CaCl,, 100 mM-NaCl, 50 rnM-TrisHCl, pH 7-6). For nuclear preparations that were

Red algal nucleosomes

153

to be used for histone purification o-i mM-phenylmethylsulphonyl fluoride (PMSF) was included in the lysis buffer. After filtration through gauze to remove cell clumps, the cell suspension was added dropwise to 800 ml of lysis buffer containing 1 % Triton X-100 which was being rapidly stirred. After 20 min the cells had lysed and the nuclei were pelleted by centrifugation at 7500 g for 10 min using a Sorvall GSA rotor. The nuclear pellet was washed by resuspending it in 30 ml lysis buffer and pelleted at 9000 g for 10 min in a Sorvall HB4 rotor. In an attempt to purify the nuclei further some preparations were centrifuged through a sucrose cushion, as follows: the nuclei were resuspended in 5 ml lysis buffer and mixed with 55 ml 2 M-sucrose, 6 mM-MgCl,, 2 mM-CaCl|, 50 mM-Tris• HC1 (pH 7-6), overlayed onto an equal volume of 2 M-sucrose, 6 mM-MgCl,, 2 mM-CaClt, 50 mM-Tris-HCl (pH 7 6 ) in 30-ml tubes, and centrifuged at 9000 g in a Sorvall HB4 rotor for 1 h to form a pellet of the partially purified nuclei. A sucrose cushion was not used for later preparations since it reduced the yield of nuclei and resulted in little additional purification as observed by light microscopy. This lysis procedure normally worked well, but occasionally no breakage, or breakage of only a small percentage of the cells, occurred. This failure to lyse was most common in older cultures approaching stationary phase; we suspect that it w u due to variations in the extracellular layer of polysaccharide. The rate of production of this encapsulating polysaccharide is greatest in stationary phase cultures (Ramus, 1972).

Micrococcal nucleate digestion After removal of the supernatant the nuclear pellet was mixed and the nuclear concentration determined by fluorescence microscopy after staining with euchrysine 3R (Young & Smith, 1964; Bard, Dickens, Edwards & Smith, 1974). Nudei were then resuspended in 0-25 Msucrose, o-i mM-CaCl,, 1 mM-TrisHC1 (pH 8) at approx. 10* nuclei/ml (io/*g/ml DNA) and incubated at 37 °C for several minutes. Micrococcal nuclease (Worthington) was added to a concentration of 10 units/ml. After times varying from 30 s to 30 min digestion was terminated by the addition of ethylenediaminetetraacetic acid (EDTA) to a final concentration of io mM. One digestion only was done at 25 °C.

DNA purification DNA was purified using a procedure based on the UP extraction method (Britten, Pavich & Smith, 1970). After addition of EDTA, ribonuclease A (Sigma) was added to a concentration of ioo p%/m\ and the incubation was continued at 37 °C for 30 min. The mixture was then made to a final concentration of 8 M-urea, 0-5 % (w/v) sodium dodecyl sulphate (SDS), o-i M-sodium phosphate buffer (pH 7) and extracted twice with a 1:1 mixture of phenol saturated with o-i M-Tris-HCl (pH 8) containing o-6% (w/v) 8-hydroxyquinoline, and chloroform/isoamyl alcohol (24:1). After two further extractions with chloroform/isoamyl alcohol the aqueous layer was loaded onto a 3 ml column of hydroxyapatite (type HF), prepared as by Spencer (1978). The column was washed with 100 ml 8 M-urea, 0-24 M-sodium phosphate (pH 7), followed by 100 ml o-i M-sodium phosphate (pH 7). The DNA was eluted with 0-5 M-sodium phosphate (pH 7), dialysed against 10 mM-Tris-HC1 (pH 8), 1 mM-EDTA, precipitated with ethanol, washed with 70 % ethanol and dried. In early preparations the DNA was precipitated immediately after the chloroform/isoamyl alcohol extraction, and the ribonuclease treatment was omitted for some preparations. The redissolved precipitate from these preparations was extremely viscous, due to contamination with polysaccharide that co-purifies during the precipitation with ethanol. This problem was completely overcome by washing the DNA extensively on hydroxyapatite.

Sizing of DNA fragments The DNA, dissolved in 36 mM-Tris, 10 mM-NaHjPO,, 1 mM-EDTA (pH 7 8 ) (E buffer; Loening, 1969) containing 0-2 % (w/v) SDS and 2 % (w/v) Ficoll was loaded onto 1 % agarose vertical gels in E buffer. After electrophoresis at 4 V/cm for approximately 4 h gels were stained with 0-5 /*g/ml ethidium bromide in E buffer and photographed under short-wave ultraviolet light. 6 CEI 57

154

K. L. Barnes, R. A. Craigie, P. A. Cattini and T. Cavalier-Smith

Sizes were calculated relative either to a Haelll restriction digest of bacteriophage PM2 DNA (Martin et al. 1977) or to a micrococcal nuclease digest of chicken erythrocyte chromatin that had been standardized against an Haelll digest of PM2. Repeat-lengths were calculated from the slope of plots of multimer length against the corresponding multimer number. Purification of histones Nuclei were lysed in 2 M-NaCl, 5 M-urea, 10 mM-sodium phosphate (pH 7-0), 0-25 raMPMSF, 5 mM-/?-mercaptoethanol, and applied to a 2 ml column of hydroxyapatite (equilibrated in the same buffer) at 4 °C. After washing with several column volumes of the same buffer, the eluant containing the histones, which, unlike non-histone proteins, do not bind under these conditions (Gould et al. 1978) was dialysed overnight against two changes of 80 mM-NaCl, 10 miu-Tris-HCl (pH 75), 0-25 mM-PMSF, o-i mM-EDTA. At this stage the solution was still quite viscous because of the high concentration of polysaccharide. After dialysis the solution was recycled for 24 h at 4 CC through a 0-7 ml column of DNA-cellulose (Allan, Staynov & Gould, 1980) equilibrated with 80 mM-NaCl, 10 mM-Tris-HCl (pH 7-5), 0-25 mM-PMSF, o-i mM-EDTA. After washing with several column volumes of the same buffer to remove residual polysaccharide, the bound histones were eluted with 2 M-NaCl, 10 mM-Tris-HCl (pH 75), 025 mM-PMSF, 01 mM-EDTA, dialysed overnight against two changes of 0-25 mM-PMSF and lyophilized. Electrophoretic analysis of histones The histone sample was dissolved in 1 % SDS, 1 % /?-mercaptoethanol, 5 % glycerol, 0-0025% bromophenol blue, 62-5 mM-Tris-HCl (pH 6-8) and electrophoresed on a 15% polyacrylamide/SDS gel (Laemmli, 1970). As staining with Coomassie Blue was not sufficiently sensitive to visualize the small quantity of algal protein, the more sensitive silver stain (Morrissey, 1981) was used. Chicken erythrocyte histones were also run for comparison.

RESULTS Micrococcal nuclease digestion of Porphyridium chromatin

Though the ease of breakage of cells and the purity of the nuclei varied somewhat from experiment to experiment, we were able to obtain sufficiently pure and intact nuclear preparations for the study of chromatin structure by micrococcal nuclease digestion. Digestion with micrococcal nuclease produces DNA fragments, each a multiple of a basic monomer length (Fig. 1) as in other eukaryotes, showing that Porphyridium chromatin has a typical eukaryotic nucleosome structure. We measured the repeat-length of the DNA associated with Porphyridium nucleosomes in several separate preparations, as described in Materials and Methods. The results (Table 1) are somewhat variable. This variability results primarily from variability from one digestion to another, and not from inaccurate measurements or calibration, or the trimming effects that are observed after extensive digestion (Shaw et al. 1976). When several separate gels were run from the same digest the calculated repeat-lengths were in close agreement. A careful study of repeat-lengths in the slime mould Physarum (Johnson et al. 1976) showed a similar degree of variability, which was traced to a systematic reduction in calculated repeat-lengths with the degree of digestion of the chromatin. We suspect that this may also be the case in Porphyridium. We were unable to do a systematic study of repeat-length as a function of digestion time, since the quantity of chromatin


l

SS

a b e d

Fig. i. Electrophoresis on a I % agarose gel of DNA fragments produced by digestion of Porphyridium nuclei with micrococcal nuclease. Lanes a and d, micrococcal nuclease digest of chicken erythrocyte nuclei. Lane b, Haelll restriction digest of bacteriophage PM2 DNA. Lane c, micrococcal nuclease digest of Porphyridium nuclei.

Table i. Measurements of DNA repeat-lengths in several preparations Digest

Gel

DNA repeat-length (nucleotide pairs)

Mean repeat-length (nucleotide pairs)

172

172

178

178

a b

i8o\ 181/

181

a

1631 164 \ 160J 74\

b M

a

162

169/

Overall mean: 173

i^fl


obtained from each preparation was insufficient for a large range of digestion times and the rate of digestion varied considerably between different preparations of nuclei. This variability was probably due to variation in the purity and condition of different nuclear preparations; lysis of some of the nuclei was evident in many preparations. We noticed under the phase and fluorescence microscope that some preparations had considerable amounts of cytoplasm adhering to the nuclei, whereas others did not. This is supported by the presence in some, but not all, early preparations that were not ribonuclease-treated prior to electrophoresis of intense bands that migrated close to the hexamer to octamer position (not shown). These were shown to be ribosomal RNA and all digests used for measurements were ribonuclease-treated. Evidence for histones in Porphyridium The micrococcal nuclease-digestion studies clearly show that Porphyridium chromatin has a repeating structure, with a DNA repeat-length similar to that of chromatin from higher eukaryotes. However, it is possible that this repeat could be produced by a fundamentally different structure that, by coincidence, shares the same periodicity in its sensitivity to micrococcal nuclease as the nucleosome structure of chromatin from higher eukaryotes. It is especially important to consider this possibility, since

Fig. 2. Electrophoresis of basic proteins on a 15 % polyacrylamide/SDS gel, stained with silver. Lanes a, c, chicken erythrocyte histones. Lane b, Porphyridium histones. The positions of the chicken erythrocyte histones are labelled.


157

the only published attempt to isolate histone-like proteins from a red alga, Rhodymenia palmata (Duffus, Penman & Webb, 1973), reports that the proteins obtained showed very little chemical similarity to histones, were present in an 11:1 ratio of 'histone' to DNA (w/w) and would not enter polyacrylamide gels under conditions commonly used for histones. Studies of nuclear proteins in Porphyridium are hindered by the heavy contamination of our nuclear preparations with polysaccharide, the small quantity of material obtainable due to the small nuclear size (Barnes, 1979), and the poor yields of nuclei obtained from cultures grown to a high cell density. We have successfully purified basic proteins from these crude nuclear preparations by hydroxyapatite followed by DNA-cellulose chromatography. Polyacrylamide/SDS gel electrophoresis of the Porphyridium DNA-binding proteins eluted from the DNA-cellulose shows a characteristic histone-like pattern (Fig. 2); a Porphyridium band migrates quite closely with each of the chicken erythrocyte core histones used as standards. The main differences are the reduced intensity of the Porphyridium band that migrates with the chicken erythrocyte H3, and the presence in the Porphyridium lane of an intense band that migrates just ahead of chicken erythrocyte H2A. A weak band is just visible at this position in the chicken erythrocyte lane and is known to be an endogenous degradation product of H3 (Bohm, Briand, Sautiere & Crane-Robinson, 1981). We suspect that degradation of Porphyridium H3 accounts for the low intensity of the band that migrates with chicken erythrocyte H3 and the presence of a band that migrates with the endogenous degradation product of chicken erythrocyte H3. If this interpretation is correct, the absence of strong bands in the region where we would expect to see linker histones should not be taken as evidence against their existence in Porphyridium, because linker histones, like H3, degrade more readily than H2A, H2B or H4 (Weintraub & Van Lente, 1974; Sollner-Webb, Camerini-Otero & Felsenfeld, 1976). Without further characterization, which would require more material than we can readily obtain, it is not possible to establish a correspondence between any of the Porphyridium bands and histones from higher eukaryotes. However, our results clearly show that Porphyridium contains DNA-binding proteins with electrophoretic mobilities characteristic of core histones.

DISCUSSION

We have established that the red alga P. aerugineum has a typical eukaryotic chromatin structure consisting of nucleosomes. We interpret the observed variation in repeat-length between different digests as evidence for heterogeneity of nucleosome spacing in Porphyridium. Chromatin containing longer linkers presents more potential target sites for micrococcal nuclease and may therefore be cleaved earlier during digestion; this has also been suggested for similar results with Physarum (Johnson et al. 1976). Alternatively, this variation might reflect differences in accessibility to micrococcal nuclease, perhaps caused by differences in higher-order structure of chromatin sub-populations differing in repeat-length; the more accessible fractions would be cleaved earlier. In either case the measured repeat-length will depend on the extent

158


of digestion. We were unable to control this reproducibly, probably because of variation in the purity of our nuclear preparations. The average value of 173 base-pairs is not quite as short as in ascomycete fungi, but comparable to that of slime moulds. The fact that the nucleosomal DNA repeatlength can vary in different mammalian tissues almost as much as it does between different protists cautions one against placing undue phylogenetic emphasis on such a simple character. Nonetheless the consistently low repeat-lengths of the ascomycete fungi, and the especially long repeats of the chrysophyte Olisthodiscus, which is best classified in a kingdom, Chromista (Cavalier-Smith, 19816), distinct from fungi, plants, animals and protozoa, suggests that in protists repeat-length may have some phylogenetic significance, and that study of additional species to test this possibility will be worthwhile. Our results show that chromatin structure in red algae is not fundamentally more primitive than that in other eukaryotes. In itself this does not rule out the idea that red algae are the most primitive eukaryotes (Klein, 1970; Chadefaud, 1976), but it is fully consistent with the idea that they evolved instead from a biciliated protozoan by the endosymbiosis of a cyanophyte to form a plastid (Cavalier-Smith, 1982). The isolation of histones has been attempted in a variety of other algae (Horgen & Silver, 1978) but histones like those of other eukaryotes have previously been clearly demonstrated only in Euglena (Jardine & Leaver, 1977). Nucleosome repeat-length is a relatively crude evolutionary marker, yet the observed variations suggest that underlying it there may be important evolutionary changes in histone structure, especially in Hi whose variation is most likely to affect repeatlengths, which if studied comparatively might profoundly illuminate eukaryote phylogeny. The use of recombinant DNA techniques should greatly facilitate the study of histone gene sequences for this purpose. We thank J. Allan for advice, discussion, and providing the DNA-cellulose, and M. Dickens for a gift of euchrysine 3R. Some of this work formed part of a thesis submitted in part fulfilment of the requirements for the Ph.D. degree of London University (Barnes, 1979). REFERENCES J., HARTMAN, P. G., CRANE-ROBINSON, C. & AVKES, F. X. (1980). The structure of histone Hi and its location in chromatin. Nature, Land. 288, 675-679. ALLAN, J., STAYNOV, D. Z. & GOULD, H. (1980). Reversible dissociation of linker histone from chromatin with preservation of internucleosomal repeat. Proc. natn. Acad. Set. U.S.A. 77, 885-889. ALLSOPP, A. (1969). Phylogenetic relationships of the Procaryota and the origin of the eukaryotic cell. New Phytol. 68, 591-612. BARD, D. R., DICKENS, M. J., EDWARDS, J. & SMITH, A. U. (1974). Studies on slices and isolated cells from fresh osteoarthritic human bone. J'. Bone Jt Surg. 56 B, 340-351. BARNES, K. L. (1979). Studies on the nuclei and chromatin of red algae. Ph.D. thesis, London University. BOHM, L., BRIAND, G., SAUTIERE, P. & CRANE-ROBINSON, C. (1981). Proteolytic digestion studies of chromatin core-histone structure. Identification of the limit peptides of histones H3 and H4. Eur.J. Biochem. 119, 67-74. BRITTEN, R. J., PAVICH, M. & SMITH, J. (1970). A new method for DNA purification. Carnegie Inttn Wash. Yb. 68, 400-402. ALLAN,

Red algal nucleosomes CAVALIER-SMITH,

159

T. (1975). The origin of nuclei and of eukaryotic cells. Nature, Lond. 256,

463-468. T. (1978). The evolutionary origin and phylogeny of microtubules, mitotic spindles and eukaryote flagella. BioSystems 10, 93-114. CAVALIER-SMITH, T. (1981 a). The origin and early evolution of the eukaryotic cell. Symp. Soc. gen. Microbiol. 32, 33-84. CAVALIER-SMITH, T. (1981&). Eukaryote kingdoms: seven or nine? BioSystems 14, 461-481. CAVALIER-SMITH, T. (1982). The origins of plastids. J. Limt. Soc. {Biol.) 17 (In Press). CHADEFAUD, M. (1976). Sur l'origine des plastes, les plastes ' Cyanelloides' et la 'Classe?' des Glaucophyc6es. C. r. hebd. Sianc. Acad. Set., Paris 283, 1029-1032. CAVALIER-SMITH,

DUFFUS> J- H., PENMAN, C. S. & WEBB, N . W. G. (1973). Histones from the red alga, Rhody-

menia palmata. Experientia 29, 632-633. E., EDWARDS, M. R. & CONTI, S. F. (1968). Ultrastructure of Porpkyridium aerugmeum, a blue-green colored rhodophytan. J. Pkycol. 4, 65-71. GOROVSKY, M. A. & KEEVERT, J. B. (1975). Subunit structure of a naturally occurring chromatin lacking histones Fi & F3. Proc. natn. Acad. Set. U.S.A. 72, 3536-3540. GOULD, H. J., MARYANKA, D., FEY, S. J., COWLING, G. J. & ALLAN, J. (1978). The assay of globin gene transcription in reconstituted chromatin. In Methods in Cell Biology, vol. 19 (ed. G. Stein, J. Stein & L. J. Kleinsmith), pp. 377-422. New York, San Francisco, London: Academic Press. HORGEN, P. A. & SILVER, J. C. (1978). Chromatin in eukaryotic microbes. A. Rev. Microbiol. 32, 240-284. HUMPHRIES, S. E., YOUNG, D. & CARROLL, D. (1979). Chromatin structure of the 5 S ribonucleic acid genes of Xenopus laevis. Biocltemistry 18, 3223-3231. JARDINE, N. J. & LEAVER, J. L. (1977). The isolation of nuclei and histones from Euglena gracilis. Expl Cell Res. 106, 423-426. JOHNSON, E. M., LITTAU, V. C , ALLFREY, V. G., BRADBURY, E. M. & MATTHEWS, H. R. (1976). The subunit structure of chromatin from Pkysarum polycephalum. Nucl. Acids Res. 3, 3313-3328. KLEIN, R. M. (1970). Relationships between blue-green and red algae. Ann. N.Y. Acad. Sci. 175. 623-632. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, Lond. 227, 680-685. LIPPS, H. J. & MORRIS, N. R. (1977). Chromatin structure in the nuclei of the ciliate Stylonychia mytilus. Biochem. biophys. Res. Commun. 74, 230-234. LOEBLICH, A. R. Ill (1976). Dionoflagellate evolution: speculation and evidence.,7. Protozool. 23, 13-28. LOENING, U. E. (1969). The determination of the molecular weight of ribonucleic acid by polyacrylamide-gel electrophoresis. Biochem. J. 113, 131-138. LOHR, D. & VAN HOLDE, K. E. (1975). Yeast chromatin subunit structure. Science, N.Y. 188, 165-166. MCGHEE, J. D. & FELSENFELD, G. (1980). Nucleoaome structure. A. Rev. Biochem. 49, 11151156. MARTIN, D. Z., TODD, R. D., LANG, D., PEI, P. N. & GARRARD, W. T. (1977). Heterogeneity in nucleosome spacing. J. biol. Chem. 252, 8269-8277. MERESCHKOWSKY, C. (1905). Ober Natur und Ursprung der Chromatophoren im Pflanzenreiche. Biol. Centralbl. 25, 593-604. MORRIS, N. R. (1976). Nucleosome structure in Aspergillus nidulans. Cell 8, 357-363. MORRISEY, J. H. (1981). Silver stain for proteins in polyacrylamide gels: A modified procedure with enhanced uniform sensitivity. Analyt. Biochem. 117, 307-310. NOLL, M. (1976). Differences and similarities in chromatin structure of Neurospora crassa and higher eukaryotes. Cell 8, 349-355. RAMUS, J. (1972). The production of extracellular polysaccharide by the unicellular red alga Porphyridium aerugmeum. J. Pkycol. 8, 97-111. SHAW, B. R., HERMAN, T. M., KOVACIC, R. T., BEAUDREU, G. S. & VAN HOLDE, K. E. (1976). Analysis of subunit organization of chicken erythrocyte chromatin. Proc. natn. Acad. Sci. U.S.A. 73, 5O5-5O9GANTT,

160


K.( RIZZO, P. J. & JOHNSON, J. R. (1980). DNA repeat length of chromatin from the unicellular alga Olisthodiscus luteus. FEBS Lett. 115, 221-224. SILVER, J. C. (1979). Chromatin organization in the oomycete Achlya ambisexualis. Biochim. biophys. Acta 561, 261-264. SOLLNER-WKBB, B., CAMERINI-OTKRO, R. D. & FELSENFELD, G. (1976). Chromatin structure as probed by nucleases and proteases: Evidence for the central role of histones H3 & H4. Cell 9, 179-193SPENCER, M. (1978). Hydroxyapatite for chromatography. II. Sources of variability and improved methods of preparation. J. Ckromat. 166, 435-446. THOMAS, J. O. & FURBKR, V. (1976). Yeast chromatin structure. FEBS Lett. 66, 274-280. THOMAS, J. O. & THOMPSON, R. J. (1977). Variation in chromatin structure in two cell types from the same tissue: a short DNA repeat length in cerebral cortical neurons. Cell 10, 633-640. WALMSLEY, M. E. & DAVIES, H. G. (1975). Ultrastructural and biochemical observations on interphase nuclei isolated from chicken erythrocytes. J. Cell Set. 17, 113-139. WEINTRAUB, H. & VAN LENTE, F. (1974). Dissection of chromosome structure with trypsin and nucleases. Proc. natn. Acad. Set. U.S.A. 71, 4249-4253. YOUNG, M. R. & SMITH, A. U. (1964). The use of euchrysine in staining cells and tissues for fluorescence microscopy. Jl R. rmcrosc. Soc. 82, 233-244. SHUPE,

(Received 9 March 1982)