Plant Molecular Biology 20: 1089-1096, 1992. © 1992 Kluwer Academic Publishers. Printed in Belgium.
1089
Nucleosomal structure and histone H1 subfractional composition of pea (Pisum sativum) root nodules, radicles and callus chromatin E.P. Bers 2, N.P. Singh 1 V.A. Pardonen 2, L.A. Lutova 2 and A.O. Zalensky 1,3,,
~Institute of Agricultural Microbiology, St. Petersburg, and 2Biological Institute of St. Petersburg State University, Russia; 3Present address: Department of Biological Chemistry, School of Medicine, University of California, Davis, CA 95616, USA (*author for correspondence) Received 30 August 1991; accepted in revised form 4 August 1992
Key words: nodule development, plant nuclei, chromatin, nucleosomal repeat, histone H1
Abstract
Higher-order packaging of D N A in chromatin structures could be an essential step in the complex chain of events leading to activation/repression of eukaryotic gene expression. With the goal to investigate this aspect of transcriptional regulation of plant genes involved in symbiotic interactions between legumes and rhizobia we analyze here the molecular parameters of chromatin structure in functioning root nodules, callus and radicles of pea. Morphological intactness and the typical nucleosomal organization are preserved in purified nuclei isolated from all three sources. The calculated values of nucleosomal repeat changed from 185 + 5 bp in the nuclei of radicles to 168 + 5 bp and 195 + 6 bp in nodules and callus respectively. The observed changes are due to alterations in linker D N A lengths. The core histones are identical in all cases, but the subfractional composition of H1 linker histone is subjected to quantitative alterations. The most pronounced is the several-fold increase in content of the lowest-molecularweight subfraction H 1-6 which takes place during nodule development.
Introduction
Cell-specific gene expression which is the basis of differentiation in multicellular eukaryotes is established by interactions of D N A regulatory cis elements with trans-acfing protein factors. However, the realization of these interactions during gene activation is linked with the rearrangement of nucleosomal and higher order structures of chromatin in regions flanking transcription unit [9]. Such connections were shown for a variety of animal and lower eukaryote genes. Thus, the region with an altered nucleosomal structure bounded by DNAse I hypersensitive sites was
described which corresponds to cis-regulatory sequences of mouse fi-globin gene [2]. Transient local perturbation of the chromatin during transcription were shown for genes such as Hsp82 gene of yeast [20] and chicken embryonic erythrocyte fi-globin gene [22] (for recent reviews see also [8, 12, 13]). Much less is known about participation of chromatin structure in plant gene expression. The local chromatin structures of sucrose synthase [29] and AdhI [27] genes of Zea mays were analyzed for DNAse I hypersensitive sites. Increased DNAse I sensitivity was shown to be associated with expression of T-DNA isopentyl
1090 transferase gene after its incorporation into Nicotiana tabacum chromosome [23]. These and other existing data suggest that the relation between gene expression and chromatin conformation in plants resemble that of animal genes. The progress in this field is slowed down by difficulties connected with isolation of purified and intact nuclei from plant tissues. One of the best studied cases of plant differentiation is legume root nodule formation. It is initiated by molecular signals from rhizobial bacteria and accompanied by sequential expression of sets of nodule-specific plant genes (nodulins) [21]. For several nodulins the complete D N A sequences of 5'-flanking regions were determined and within them a number of expression regulatory elements were localized [7, 16]. Our longterm aim is to study the chromatin organization of individual nodulin genes. Here we have studied the nucleosomal structure and histone composition of bulk chromatin in nuclei of pea root nodules. A hormone-dependent callus line was included for comparison, nuclei from radicles of aseptically grown plants were used as a standard. Typical nucleosomal organization was shown in all cases studied, but D N A repeat lengths and chromatin compactness varied. Nucleosomal D N A repeat value is minimal in nodule, intermediate in root and maximal in callus nuclei. These differences were attributed to changing lengths of linker D N A in nucleosome. Histone H 1 participates in formation of higherorder chromatin structure [1, 15]. Developmentally expressed H1 variants were thought to influence linker D N A length ofnucleosomes [6, 30, 31]. Therefore, we analyzed the subfractional composition of H I as a possible factor affecting the observed variation in linker size. Western blot analysis showed quantitative differences in histone H 1 subfractional composition while the core histones remained identical. The most pronounced change was the several-fold increase in content of low-molecular-weight subfraction H1-6 in nodules. H1-6 accumulation in the infected zone of roots began on days 5-7 after inoculation. We suggest the participation of H 1 proteins in modulation of overall chromatin
structure during pea tissues differentiation induced by interactions with rhizobia.
Materials and methods
Biological material Pea (Pisum sativum L.) cv. Viola, callus derived from pea line L-23/2 [5] and Rhizobium leguminosarum bv. vicae effective strain 250 a were from the collection of the Institute of Agricultural Microbiology. Inoculation and plant cultivation were performed as described [18]. Three-day-old radicles, root nodules collected 4 weeks after inoculation and callus collected one month after transfer were used for nuclei isolation. Root segments corresponding to regions of nodule formation (plants one to four weeks after inoculation) served as a source for chromatin isolation to study histone H I changes in the course of symbiosis development.
Isolation of plant nuclei The procedure was as described earlier in detail [5]; all steps were carried out at 4 °C. Briefly, radicle tissue was gently squashed between two layers of polyethylene film and the homogenate washed out with excess of M N E isolation buffer (10 mM MES pH 6.0, 100 mM NaC1, 0.25 M sucrose, 0.5 mM spermidine, 0.15 mM spermine, 25 mM 2-mercaptoethanol, 1 m M PMSF, 3 ~ BSA). After filtration, the suspension was pelleted by centrifugation for 15 min at 1000 x g. Crude nuclei pellet was removed from the starch granules with a spatula, resuspended in M N E containing 0.7 ~o Triton x 100 and pelleted again. Washing with detergent was repeated once more. Pellet containing crude nuclei was purified further by centrifugation in Percoll for 30 min at 3000 x g [28]. The floating material was collected, resuspended in M N E and washed twice with centrifugation for 15 rain at 1000 x g. Isolation procedures for callus and nodules were essentially the same except for the initial
1091 steps. Callus samples were immersed in cold 96 ~o ethanol for 1-3 min and then incubated for 1 h in MNE, mixed with glass beads (Ballotini N 14) and homogenized in a blender at maximal speed for 15 s. Nodule tissue was crushed in M N E with mortar and pestle. Crude nodule nuclei after centrifugation at 1000 x g were washed several times with fresh portions of M N E to abolish bacteroid contamination. Nuclei purity was checked by UV fluorescent microscopy after staining with 0.01 ~o acridine orange.
Nuclease treatment and chromatin structure analysis
Purified nuclei were resuspended in nuclease buffer (10 mM Tris-HC1 pH 7.5, 0.3 M sucrose, 1 mM CaC12, 0.1 m M PMSF) at D N A concentration 1 mg/ml. Micrococcal nuclease (Sigma) was added to 0.5 units per mg of D N A and digestion was carried out at 37 °C for 2-30 min. The reaction was stopped by chilling on ice and the samples were centrifuged at 4000 x g for 10 min. The nuclear pellet was lysed in (5 m M Tris-HC1 pH 7.5, 2 mM EDTA) for 15-30 min and spun down to recover supernatant fraction which contains solubilized chromatin fragments. Supernatant was used for electrophoretic analysis of nucleosomes and isolation of DNA. The chromatin particles were separated on 5 ~o P A G E as described [30] and the D N A from nucleosomal fraction on 1.8~o agarose gel. Sizes of D N A fragments were estimated from the calibration curve obtained from Pst I digest of 2 phage DNA. The values for nucleosomal D N A repeat length were determined from the slope of the regression line from plot of fragment size versus band number. Mean values and standard deviations were from not less than six independent experiments.
Protein analysis
Total nuclear proteins were separated in 15~o P A G E according to Laemmli [19]. To prepare the samples purified nuclei were briefly washed
with 10 mM Tris-HC1 pH 6.8 and lysed in Laemmli sample buffer by boiling for 3 min. After electrophoresis part of a gel was stained with Coomassie Blue R 250, the remaining unstained part was immediately used for western blotting. Proteins were transferred to nitrocellulose membranes (0.45 #m pore size) by electroblotting for 2 h at 0.25 mA/cm 2 (BioRad apparatus). Transfer buffer was the same as Laemmli electrophoretic buffer but with the SDS concentration reduced to 0.01 ~o. Nitrocellulose filters were blocked and treated with primary and secondary antibodies according to standard protocol (BioRad). Primary antibodies against chromatographically purified H 1 fraction and total histone of pea radicles were as described earlier [4], secondary antibodies were goat anti-rabbit antibodies conjugated with HRP. Blots were stained with 4-chloro-l-naphtol. Alternatively, 135I-labelled protein A was used as a secondary antibodies and blots were subjected to autoradiography. To analyze histone H 1 changes during nodule development proteins were extracted with 0.3 M HC1 from chromatin which was isolated from root segments according to Belyaev and Berdnikov [3]. Acid-soluble proteins were precipitated by 6 volumes of acetone and separated by twodimensional electrophoresis [18] (lst dimension: acetic acid/urea gel [3]; 2nd dimension: SDS gel [ 19]). Western blotting of two-dimensional gels was performed as described above.
Results and discussion
Nucleus characteristics and nucleosomal organization of chromatin
The described experimental procedures produce fairly pure preparations of undamaged, partially demembranized, plant nuclei, as judged from fluorescence microscopy (data not shown). Nuclei preserve their spherical shape and often nucleoli are visible. Intactness of histone proteins and chromatin nucleosomal structure were preserved (see results below). We had not been able to come
1092 up with a single method that was suitable for isolating nuclei from all the tissues. The homogenization step employed was essentially 'tissuespecific' - very mild crushing for radicles, pretreatment of callus with ethanol followed by preincubation in nuclei isolation buffer were essential. During isolation of plant nuclei from nodules several washes were necessary to remove bacteroids. This was achieved by repeated lowspeed pelleting before the final step of Percoll purification. Due to differences in water content, cell wall characteristics, nuclear membrane properties and other unknown factors, care is required when working with plants, even with established protocols. For example, we found that the commonly used step of tissue homogenization in liquid nitrogen leads to collapsing and crushing of nuclei. Size distribution of nuclei from different tissue sources was heterogeneous, reflecting the complex pattern of different cells forming nodules, radicle and callus. The most nonuniform were diameters of nodule nuclei. We were unable to link nuclei of a certain size to the known cell type (meristematic, infected, uninfected, etc). Nuclei from radicle and callus looked much more even in diameters. Thus in all cases under investigation we dealt with the mixture of nuclei originating from different cell types, so the characteristics obtained represent averaged features of a given tissue. To analyze the variation in chromatin structural organization we used the standard approach of nuclei treatment with micrococcal nuclease. Fractions of soluble nucleosomes obtained after nuclease digestion were analyzed on low-ionicstrength nucleoprotein gels (Fig. 1A). We observed a typical pattern of bands corresponding to core particles (or mononucleosomes lacking histone H1, MN-H1), complete mononucleosomes (MN + HI), dinucleosomes etc. The histone composition and approximate D N A lengths of these chromatin particles were verified in a second-dimension 15~o P A G E - S D S gels [25]. When the D N A was extracted from the soluble fraction and separated on native 1.8~o agarose gel, a characteristic nucleosomal ladder was
observed. Figure 1B shows the representative separation. To calculate nucleosomal repeat 6 to 8 runs of D N A from independent experiments were performed. The values of repeat lengths were determined for each run (see Materials and methods) and averaged afterwards. We obtained values of 185 + 5, 195 + 6 and 168 + 5 bp for radicle, callus and nodule nuclei respectively. Nucleosome repeats for pea radicle chromatin reported in the literature vary from 194 + 7 bp [11] to 185 + 5 bp [26], the discrepancies being attributed to different procedures of chromatin isolation and methods of calculations. We think that the value of 185 bp, also obtained by us, is more accurate since the nucleosome repeat length of the control chicken erythrocyte nuclei measured in a parallel experiment was 210 bp, which coincides with the value determined for these cells with high precision (see data summary in [15]). According to our two-dimensional electrophoresis data [25] the length of D N A in core particles was about 145 bp in all cases studied. The observed differences in repeat lengths are therefore determined by variations in linker D N A size. The length of linker D N A in nodule nuclei is at least 10 bp shorter than in radicles of uninfected plants or callus tissue. Such a difference is significant and can reflect the reverse correlation between linker D N A length and average cell transcriptional activity that was shown for some animal systems. An increase in linker D N A length correlates with transcriptional shut-down during terminal differentiation in sperm [1, 30, 31] and erythrocyte [ 15] cells. The nodule is composed of different cell types, many of them are active in nodulins gene expression [21], unfortunately reliable data on the total level of transcriptional activity are not available. Repeat lengths determined in this work represent values median for cell types and different parts of genomes and therefore should be relied on only as an indication of a trend in chromatin structure rearrangement. At the same time one could expect the more pronounced changes in chromatin characteristics in some nodule cell types and especially in D N A
1093
Fig. 1. Electrophoretic comparison of soluble chromatin fragments (A) and nucleosomal D N A (B) obtained from treatment of pea nuclei with micrococcal nuclease. A. Purified nuclei were digested with nuclease as outlined in Materials and methods during indicated times (minutes). Dry sucrose was added to the soluble nucleosomes fraction to 12% (w/v) and samples were loaded into slots of 5 % P A G E prepared on 10 m M Tris-borate buffer pH 8.3 containing 1 mM EDTA. After electrophoresis at 50 V for 2.5-3 h the gel was stained with ethidium bromide. Positions of core particle (mononucleosome lacking histone H 1 - M N - H 1), complete nucleosome (MN + H1) and dinucleosome (DN) were determined by two-dimensional electrophoresis [25]. B. The D N A fragments isolated from soluble chromatin fraction obtained by treatment of nuclei with micrococcal nuclease were separated on 1.8 % agarose gel in 20 mM Tris-acetate buffer pH 7.7 containing 2 mM EDTA. Lengths of the restriction fragments (bp) of 2 D N A digested with Pst I which served as markers for determination of nucleosome repeat lengths are shown at the left side of the panel.
1094 regions to correspond to genes specifically induced to expression as a result of symbiosis development. Our preliminary data [ 32] show such rearrangements in chromatin of the promoter region of pea leghemoglobin gene. The difference in callus versus radicle linker D N A lengths is less pronounced, and is within experimental error. Nevertheless, we can propose the tendency for linker length to increase in callus. From the data in Fig. 1 it is clear that digestion of nodule nuclei chromatin occurs more easily. (Compare the amounts of core to higher-order nucleosomes (Fig. 1A) and the abundance of shorter D N A fragments in nodules compared to radicles at 20 min digestion (Fig. 1B). The same is supported by our data on the kinetic of acidsoluble products accumulation upon digestion of nodules and radicle nuclei with DNase ! and micrococcal nuclease [25]. There was no significant endogenous nuclease activity in nodule nuclei since majority of D N A had remained in highmolecular-weight band after 20 min incubation in absence of exogenous enzyme (Fig. 1B). Taken into consideration all obstac'les mentioned above we can assume the D N A in nodule nuclei is more accessible to micrococcal nuclease action due to a more 'open' chromatin conformation. It is believed that local decondensation of chromatin is a prerequisite for active transcription [9, 12, 13]. Additional experiments are needed to establish whether the observed alterations in total chromatin from nodules promotes an increased level of gene expression in some cells.
Histone composition
One of the possible molecular mechanisms to induce the observed changes in average size of linker D N A in pea radicle, callus and nodule tissues could be an alteration in histone composition. Comparison of total proteins from purified nuclei (Fig. 2A) showed that the core histone fractions H4, H3, H2A and H2B are indistinguishable in SDS gel in all three cases. Visual analysis of the stained gels was supported by the
Fig. 2. SDS-PAGE of total nuclear proteins from pea radicle (R), callus (C) and nodules (N) and Western-blotting with antibodies against chromatographically purified pea radical histone H 1. A. Nuclei at a D N A concentration of 2 mg/ml were mixed with an equal volume of 2 x Laemmli sample buffer, lysed by boifing and loaded on 15% acrylamide gel. Staining with Coomassie Blue R 250. B. Identical part of unstained gel was blotted onto nitrocellulose and treated with antibodies against pea H 1 as described in Materials and methods. The figure shows only part of the blot corresponding to H 1 histones, the position of subfraction H 1-6 which increases specifically in nodules is indicated.
immunoblotting data (not shown) with antibodies against core histones. To identify the H 1 fraction we used polyclonal monospecific antibodies raised against purified pea radicle HI [4]. The western blot analysis (Fig. 2B) reveals that pea H1 has a complex subfractional composition. In the radicle this protein is represented by 6 major polypeptide variants 1-6 which differ in molecular weights. Subfraction H 1-1 with the lowest electrophoretic mobility is the most abundant. This result agrees with the data by Berdnikov et al. [3] which was obtained on isolated pea H1 using the electrophoretic separation on 40 cm long acetic acid/ urea gels. In callus the same composition of H1 is preserved, with minor but consistent variations in amounts of the different subfractions between the callus and radicle. Sets of H1 subfractions in nodule nuclei differed from that characteristic of radicles or callus. The content of HI-1 decreased while the amount
1095 of H 1-6 significantly increased, so that these subfractions become almost equal in quantity (Fig. 2 right panel). Fraction H1-6 is not a rhizobial bacteroid polypeptide since the proteins isolated from purified bacteroids and moving within molecular weights range of interest do not crossreact with anti-H1 serum ([32], Zalensky et al., manuscript in preparation). To follow the kinetics of accumulation of nodule enriched subfraction H1-6 we isolated total acid-soluble chromatin proteins from the root segments corresponding to the zone of nodule formation from plants collected at different times after inoculation. Proteins were separated by twodimensional electrophoresis and analyzed by western blotting with anti-Hi antibodies. This procedure is less time- and material-consuming than that utilizing isolation of nuclei. Figure 3 shows that fraction H 1-6 appears very early after plant inoculation with Rhizobia. Approximately at the same time proliferation of cells in nodule primordium take place [21], so we can speculate about relationship of H1 changes with this event. Tentative connection between H1-6 variant and high rate of mitotic activity can be checked by studying histone composition of root tip cells. Participation of H 1-6 in any particular
step of symbiosis could be proved only by more elaborated experiments utilizing cell separation or immunolocalization with specific antibodies. H1 is known as a linker histone interacting with the stretches of D N A connecting adjacent nucleosome core particles [1]. Tissue specific molecular variants of H1 in some mammalian systems were shown to influence linker D N A length [1, 6, 15, 30]. From this, we can speculate that the decrease of linker D N A observed in nodule nuclei is mediated by the lower-molecularweight H 1- 6. H 1 also participates in maintaining higher order structure of chromatin and is involved in chromatin condensation [10, 15, 30]. Non-random distribution of different H1 variants (changing during pea nodule development) along regulatory sequences of genes involved in symbiotic process could be one of the molecular mechanisms modulating their specific expression via local modifications of supranucleosomal structure. Alternatively, the same result may be achieved by differential competition between H 1 variants and transcriptional regulatory proteins [22] or R N A polymerase [ 13]. Based on experimental procedures set up in this work we are planning to study the role of
anti HI
Fig. 3. Western blot (autoradiogram) of the two-dimensional separation of acid-soluble chromatin proteins isolated from root segments (see Materials and methods) at different times after inoculation. Primary antibodies as in Fig. 2, secondary antibodies _ I 3 5 [ protein A. Part of the blot corresponding to position of H1 histones is shown.
1096 chromosomal proteins and chromatin structure in regulation of expression of nodulin genes with established primary structures.
Acknowledgements We would like to thank Dr T. Bisseling and Dr J. Th'ng for critically reading this manuscript. N.P.S. was a Ph.D research fellow supported by the Ministry of Education of India.
References 1. Bawkin SG, Usachenko SJ, Zalensky AO, Mirzabekov AD: Structure of nucleosomes and organization of internucleosomal DNA in chromatin. J Mol Biol 212: 495511 (1990). 2. Behezra R, Cantor CR, Axel R: Nucleosomes are phased along the b-globin gene in erythroid and nonerythroid ceils. Cell 44:697-704 (1986). 3. Belyaev AI, Berdnikov VA: Polymorphism and localization of I-I1 histone genes in Pisum sativurn L. Genetics (USSR) 17:498-504 (1981). 4. Bers EP, Ivanova SV: Application of immunoblotting technique to study Chlamydomonas reinhardtii histones. Vestnik LGU (Leningrad Univ Herald) 1:102-110 (1988). 5. Bers EP, Pardonen VA: Quick method of nuclei isolation from some plant tissues. Tsitologiia (Cytology, Aead Sci USSR) 32:899-902 (1990). 6. Clark DJ, Thomas JO: Differences in the binding of H1 variants to DNA: Cooperativity mad linker length related distribution. Eur J Biochem 178: 225-233, 1988. 7. de Bruijn FJ, Szabados L, Shell J: Chimeric genes and transgenic plants are used to study the regulation of genes involved in symbiotic plant-microbe interactions (Nodulin genes). Devel Genet 11:182-196 (1990). 8. Elgin SCR: The formation and function of DNAse I hypersensitive sites in the process of gene activation. J Biol Chem 263:19259-19262 (1988). 9. Felsenfeld G: Chromatin as an essential part of the transcriptional mechanism. Nature 355:219-224 (1992). 10. Finch JT, Klug A: A solenoidal model for superstructure of chromatin. Proc Natl Acad Sci USSR 73:1897-1901 (1976). 11. Grellet F, Pennon P, Cooke R: Analysis of DNA associated with nueleosomes in pea ehromatin. Planta 148: 346-353 (1980). 12. Gross RD, Garrard WT: Poising chromatin for transcription. Trends Bioehem Sei 12:293-297 (1987). 13. Grustein M: Histone function in transcription. Annu Rev Cell Biol 6:643-678 (1990). 14. Harmon R, Bateman E, Allan J, Harborne N, Gould H: Control of RNA polymerase binding to ehromatin by variations of linker histone composition. J Mol Biol 180: 131-150 (1985). 15. van Holde KE: Chromatin. Springer-Verlag (1988).
16. Jensen EO: Regulation of nodule-expressed soybean genes. In: Hennecke H, Verma DPS (eds) Advances in Molecular Genetics and Plant-Microbe Interactions, Vol. 1, pp. 310-316. Kluwer Academic Publishers, Dordrecht (1991). 17. Kahl G, Weising K, Gorrz A, Schafer W, Hirasawa E: Chromatin structure and plant gene expression. Devel Genet 8:405-434 (1987). 18. Kulikova OA, Muravieva MYu, Zalensky AO, Filatov AA, Tikhonovich IA: Identification of proteins related to symbiotic nitrogen fixation in the pea. Appl Biochem Microbiol (USSR) 26:506-513 (1990). 19. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 (1970). 20. Lee M-S, Garrard WT: Transcription induced 'splitting' and underlying structure for DNAse I sensitive chromatin. EMBO J 10: 607-615, 1991. 21. Nap J-P, Bisseling T: Developmental biology of plantprokaryote symbiosis: The legume root nodule. Science 250:948-954 (1990). 22. Postnikov YV, Shick VV, Belyavsky AV, Khrapko KR, Borodin KL, Nikolskaya TA, Mirzabekov AD: Distribution of high mobility group proteins 1/2, E and 14/17 and linker histones HI and H5 on transcribed and nontranscribed regions of chicken erythrocyte chromatin. Nucl Acids Res 19:717-725 (1991). 23. Reid RA, John MC, Amasino RM: Deoxyribonuclease I sensitivity of the T-DNA ipt gene is associated with gene expression. Biochemistry 27:5748-5754 (1988). 24. Shlissel MS, Brown DD: The transcriptional regulation of Xenopus 5S RNA genes in chromatin: the roles of active stable transcription complexes and histone H 1. Cell 37:903-913 (1984). 25. Singh NP, Zalensky AO: Nuclear proteins and chromatin structure of pea (Pisum sativum) root nodule. Nucleus (in press). 26. Ull MA, Franco L: The nucleosomal repeat length of pea (Pisum sativum) chromatin changes during germination. Plant Mol Biol 7:25-31 (1986). 27. Vayda ME, Freeling M: Insertion of the Mut transposable element into the first intron of maize Adh 1 interferes with transcription elongation but does not disrupt chromatin structure. Plant Mol Biol 6:441-454 (1986). 28. Willmitzer L, Wagner KG: The isolation of nuclei from tissue-cultured plant cells. Exp Cell Res 135:67-77 (1981). 29. Wurtzel ET, Burr FA, Burr B: DNAase I hypersensitivity and expression of the Shrunken- 1 gene of maize. Plant Mol Biol 8:251-264 (1987). 30. Zalenskaya IA, Pospelov VA, Zalensky AO, Vorob'ev VI: Nucleosomal structure of sea urchin and starfish chromatin. Nucl Acids Res 9:473-487 (1981). 31. Zalenskaya IA, Odintsova NA, Zalensky AO, Vorob'ev VI: Nucleosome organization of chromatin from sperm of the bivalvia mollusc. Mol Biol (USSR) 16:335-344 (1982). 32. Zalensky AO, Singh NP, Pardonen VA, Bers EP, Aronstam AA: Changes in structural organization of plant and bacterial genome during Pisurn sativum-Rhizobiurn legurninosarum symbiosis. Int. Symposium on Molecular Genetics and Plant-Microbe Interactions (Interlaken), p. 238 (1990).
1089
Nucleosomal structure and histone H1 subfractional composition of pea (Pisum sativum) root nodules, radicles and callus chromatin E.P. Bers 2, N.P. Singh 1 V.A. Pardonen 2, L.A. Lutova 2 and A.O. Zalensky 1,3,,
~Institute of Agricultural Microbiology, St. Petersburg, and 2Biological Institute of St. Petersburg State University, Russia; 3Present address: Department of Biological Chemistry, School of Medicine, University of California, Davis, CA 95616, USA (*author for correspondence) Received 30 August 1991; accepted in revised form 4 August 1992
Key words: nodule development, plant nuclei, chromatin, nucleosomal repeat, histone H1
Abstract
Higher-order packaging of D N A in chromatin structures could be an essential step in the complex chain of events leading to activation/repression of eukaryotic gene expression. With the goal to investigate this aspect of transcriptional regulation of plant genes involved in symbiotic interactions between legumes and rhizobia we analyze here the molecular parameters of chromatin structure in functioning root nodules, callus and radicles of pea. Morphological intactness and the typical nucleosomal organization are preserved in purified nuclei isolated from all three sources. The calculated values of nucleosomal repeat changed from 185 + 5 bp in the nuclei of radicles to 168 + 5 bp and 195 + 6 bp in nodules and callus respectively. The observed changes are due to alterations in linker D N A lengths. The core histones are identical in all cases, but the subfractional composition of H1 linker histone is subjected to quantitative alterations. The most pronounced is the several-fold increase in content of the lowest-molecularweight subfraction H 1-6 which takes place during nodule development.
Introduction
Cell-specific gene expression which is the basis of differentiation in multicellular eukaryotes is established by interactions of D N A regulatory cis elements with trans-acfing protein factors. However, the realization of these interactions during gene activation is linked with the rearrangement of nucleosomal and higher order structures of chromatin in regions flanking transcription unit [9]. Such connections were shown for a variety of animal and lower eukaryote genes. Thus, the region with an altered nucleosomal structure bounded by DNAse I hypersensitive sites was
described which corresponds to cis-regulatory sequences of mouse fi-globin gene [2]. Transient local perturbation of the chromatin during transcription were shown for genes such as Hsp82 gene of yeast [20] and chicken embryonic erythrocyte fi-globin gene [22] (for recent reviews see also [8, 12, 13]). Much less is known about participation of chromatin structure in plant gene expression. The local chromatin structures of sucrose synthase [29] and AdhI [27] genes of Zea mays were analyzed for DNAse I hypersensitive sites. Increased DNAse I sensitivity was shown to be associated with expression of T-DNA isopentyl
1090 transferase gene after its incorporation into Nicotiana tabacum chromosome [23]. These and other existing data suggest that the relation between gene expression and chromatin conformation in plants resemble that of animal genes. The progress in this field is slowed down by difficulties connected with isolation of purified and intact nuclei from plant tissues. One of the best studied cases of plant differentiation is legume root nodule formation. It is initiated by molecular signals from rhizobial bacteria and accompanied by sequential expression of sets of nodule-specific plant genes (nodulins) [21]. For several nodulins the complete D N A sequences of 5'-flanking regions were determined and within them a number of expression regulatory elements were localized [7, 16]. Our longterm aim is to study the chromatin organization of individual nodulin genes. Here we have studied the nucleosomal structure and histone composition of bulk chromatin in nuclei of pea root nodules. A hormone-dependent callus line was included for comparison, nuclei from radicles of aseptically grown plants were used as a standard. Typical nucleosomal organization was shown in all cases studied, but D N A repeat lengths and chromatin compactness varied. Nucleosomal D N A repeat value is minimal in nodule, intermediate in root and maximal in callus nuclei. These differences were attributed to changing lengths of linker D N A in nucleosome. Histone H 1 participates in formation of higherorder chromatin structure [1, 15]. Developmentally expressed H1 variants were thought to influence linker D N A length ofnucleosomes [6, 30, 31]. Therefore, we analyzed the subfractional composition of H I as a possible factor affecting the observed variation in linker size. Western blot analysis showed quantitative differences in histone H 1 subfractional composition while the core histones remained identical. The most pronounced change was the several-fold increase in content of low-molecular-weight subfraction H1-6 in nodules. H1-6 accumulation in the infected zone of roots began on days 5-7 after inoculation. We suggest the participation of H 1 proteins in modulation of overall chromatin
structure during pea tissues differentiation induced by interactions with rhizobia.
Materials and methods
Biological material Pea (Pisum sativum L.) cv. Viola, callus derived from pea line L-23/2 [5] and Rhizobium leguminosarum bv. vicae effective strain 250 a were from the collection of the Institute of Agricultural Microbiology. Inoculation and plant cultivation were performed as described [18]. Three-day-old radicles, root nodules collected 4 weeks after inoculation and callus collected one month after transfer were used for nuclei isolation. Root segments corresponding to regions of nodule formation (plants one to four weeks after inoculation) served as a source for chromatin isolation to study histone H I changes in the course of symbiosis development.
Isolation of plant nuclei The procedure was as described earlier in detail [5]; all steps were carried out at 4 °C. Briefly, radicle tissue was gently squashed between two layers of polyethylene film and the homogenate washed out with excess of M N E isolation buffer (10 mM MES pH 6.0, 100 mM NaC1, 0.25 M sucrose, 0.5 mM spermidine, 0.15 mM spermine, 25 mM 2-mercaptoethanol, 1 m M PMSF, 3 ~ BSA). After filtration, the suspension was pelleted by centrifugation for 15 min at 1000 x g. Crude nuclei pellet was removed from the starch granules with a spatula, resuspended in M N E containing 0.7 ~o Triton x 100 and pelleted again. Washing with detergent was repeated once more. Pellet containing crude nuclei was purified further by centrifugation in Percoll for 30 min at 3000 x g [28]. The floating material was collected, resuspended in M N E and washed twice with centrifugation for 15 rain at 1000 x g. Isolation procedures for callus and nodules were essentially the same except for the initial
1091 steps. Callus samples were immersed in cold 96 ~o ethanol for 1-3 min and then incubated for 1 h in MNE, mixed with glass beads (Ballotini N 14) and homogenized in a blender at maximal speed for 15 s. Nodule tissue was crushed in M N E with mortar and pestle. Crude nodule nuclei after centrifugation at 1000 x g were washed several times with fresh portions of M N E to abolish bacteroid contamination. Nuclei purity was checked by UV fluorescent microscopy after staining with 0.01 ~o acridine orange.
Nuclease treatment and chromatin structure analysis
Purified nuclei were resuspended in nuclease buffer (10 mM Tris-HC1 pH 7.5, 0.3 M sucrose, 1 mM CaC12, 0.1 m M PMSF) at D N A concentration 1 mg/ml. Micrococcal nuclease (Sigma) was added to 0.5 units per mg of D N A and digestion was carried out at 37 °C for 2-30 min. The reaction was stopped by chilling on ice and the samples were centrifuged at 4000 x g for 10 min. The nuclear pellet was lysed in (5 m M Tris-HC1 pH 7.5, 2 mM EDTA) for 15-30 min and spun down to recover supernatant fraction which contains solubilized chromatin fragments. Supernatant was used for electrophoretic analysis of nucleosomes and isolation of DNA. The chromatin particles were separated on 5 ~o P A G E as described [30] and the D N A from nucleosomal fraction on 1.8~o agarose gel. Sizes of D N A fragments were estimated from the calibration curve obtained from Pst I digest of 2 phage DNA. The values for nucleosomal D N A repeat length were determined from the slope of the regression line from plot of fragment size versus band number. Mean values and standard deviations were from not less than six independent experiments.
Protein analysis
Total nuclear proteins were separated in 15~o P A G E according to Laemmli [19]. To prepare the samples purified nuclei were briefly washed
with 10 mM Tris-HC1 pH 6.8 and lysed in Laemmli sample buffer by boiling for 3 min. After electrophoresis part of a gel was stained with Coomassie Blue R 250, the remaining unstained part was immediately used for western blotting. Proteins were transferred to nitrocellulose membranes (0.45 #m pore size) by electroblotting for 2 h at 0.25 mA/cm 2 (BioRad apparatus). Transfer buffer was the same as Laemmli electrophoretic buffer but with the SDS concentration reduced to 0.01 ~o. Nitrocellulose filters were blocked and treated with primary and secondary antibodies according to standard protocol (BioRad). Primary antibodies against chromatographically purified H 1 fraction and total histone of pea radicles were as described earlier [4], secondary antibodies were goat anti-rabbit antibodies conjugated with HRP. Blots were stained with 4-chloro-l-naphtol. Alternatively, 135I-labelled protein A was used as a secondary antibodies and blots were subjected to autoradiography. To analyze histone H 1 changes during nodule development proteins were extracted with 0.3 M HC1 from chromatin which was isolated from root segments according to Belyaev and Berdnikov [3]. Acid-soluble proteins were precipitated by 6 volumes of acetone and separated by twodimensional electrophoresis [18] (lst dimension: acetic acid/urea gel [3]; 2nd dimension: SDS gel [ 19]). Western blotting of two-dimensional gels was performed as described above.
Results and discussion
Nucleus characteristics and nucleosomal organization of chromatin
The described experimental procedures produce fairly pure preparations of undamaged, partially demembranized, plant nuclei, as judged from fluorescence microscopy (data not shown). Nuclei preserve their spherical shape and often nucleoli are visible. Intactness of histone proteins and chromatin nucleosomal structure were preserved (see results below). We had not been able to come
1092 up with a single method that was suitable for isolating nuclei from all the tissues. The homogenization step employed was essentially 'tissuespecific' - very mild crushing for radicles, pretreatment of callus with ethanol followed by preincubation in nuclei isolation buffer were essential. During isolation of plant nuclei from nodules several washes were necessary to remove bacteroids. This was achieved by repeated lowspeed pelleting before the final step of Percoll purification. Due to differences in water content, cell wall characteristics, nuclear membrane properties and other unknown factors, care is required when working with plants, even with established protocols. For example, we found that the commonly used step of tissue homogenization in liquid nitrogen leads to collapsing and crushing of nuclei. Size distribution of nuclei from different tissue sources was heterogeneous, reflecting the complex pattern of different cells forming nodules, radicle and callus. The most nonuniform were diameters of nodule nuclei. We were unable to link nuclei of a certain size to the known cell type (meristematic, infected, uninfected, etc). Nuclei from radicle and callus looked much more even in diameters. Thus in all cases under investigation we dealt with the mixture of nuclei originating from different cell types, so the characteristics obtained represent averaged features of a given tissue. To analyze the variation in chromatin structural organization we used the standard approach of nuclei treatment with micrococcal nuclease. Fractions of soluble nucleosomes obtained after nuclease digestion were analyzed on low-ionicstrength nucleoprotein gels (Fig. 1A). We observed a typical pattern of bands corresponding to core particles (or mononucleosomes lacking histone H1, MN-H1), complete mononucleosomes (MN + HI), dinucleosomes etc. The histone composition and approximate D N A lengths of these chromatin particles were verified in a second-dimension 15~o P A G E - S D S gels [25]. When the D N A was extracted from the soluble fraction and separated on native 1.8~o agarose gel, a characteristic nucleosomal ladder was
observed. Figure 1B shows the representative separation. To calculate nucleosomal repeat 6 to 8 runs of D N A from independent experiments were performed. The values of repeat lengths were determined for each run (see Materials and methods) and averaged afterwards. We obtained values of 185 + 5, 195 + 6 and 168 + 5 bp for radicle, callus and nodule nuclei respectively. Nucleosome repeats for pea radicle chromatin reported in the literature vary from 194 + 7 bp [11] to 185 + 5 bp [26], the discrepancies being attributed to different procedures of chromatin isolation and methods of calculations. We think that the value of 185 bp, also obtained by us, is more accurate since the nucleosome repeat length of the control chicken erythrocyte nuclei measured in a parallel experiment was 210 bp, which coincides with the value determined for these cells with high precision (see data summary in [15]). According to our two-dimensional electrophoresis data [25] the length of D N A in core particles was about 145 bp in all cases studied. The observed differences in repeat lengths are therefore determined by variations in linker D N A size. The length of linker D N A in nodule nuclei is at least 10 bp shorter than in radicles of uninfected plants or callus tissue. Such a difference is significant and can reflect the reverse correlation between linker D N A length and average cell transcriptional activity that was shown for some animal systems. An increase in linker D N A length correlates with transcriptional shut-down during terminal differentiation in sperm [1, 30, 31] and erythrocyte [ 15] cells. The nodule is composed of different cell types, many of them are active in nodulins gene expression [21], unfortunately reliable data on the total level of transcriptional activity are not available. Repeat lengths determined in this work represent values median for cell types and different parts of genomes and therefore should be relied on only as an indication of a trend in chromatin structure rearrangement. At the same time one could expect the more pronounced changes in chromatin characteristics in some nodule cell types and especially in D N A
1093
Fig. 1. Electrophoretic comparison of soluble chromatin fragments (A) and nucleosomal D N A (B) obtained from treatment of pea nuclei with micrococcal nuclease. A. Purified nuclei were digested with nuclease as outlined in Materials and methods during indicated times (minutes). Dry sucrose was added to the soluble nucleosomes fraction to 12% (w/v) and samples were loaded into slots of 5 % P A G E prepared on 10 m M Tris-borate buffer pH 8.3 containing 1 mM EDTA. After electrophoresis at 50 V for 2.5-3 h the gel was stained with ethidium bromide. Positions of core particle (mononucleosome lacking histone H 1 - M N - H 1), complete nucleosome (MN + H1) and dinucleosome (DN) were determined by two-dimensional electrophoresis [25]. B. The D N A fragments isolated from soluble chromatin fraction obtained by treatment of nuclei with micrococcal nuclease were separated on 1.8 % agarose gel in 20 mM Tris-acetate buffer pH 7.7 containing 2 mM EDTA. Lengths of the restriction fragments (bp) of 2 D N A digested with Pst I which served as markers for determination of nucleosome repeat lengths are shown at the left side of the panel.
1094 regions to correspond to genes specifically induced to expression as a result of symbiosis development. Our preliminary data [ 32] show such rearrangements in chromatin of the promoter region of pea leghemoglobin gene. The difference in callus versus radicle linker D N A lengths is less pronounced, and is within experimental error. Nevertheless, we can propose the tendency for linker length to increase in callus. From the data in Fig. 1 it is clear that digestion of nodule nuclei chromatin occurs more easily. (Compare the amounts of core to higher-order nucleosomes (Fig. 1A) and the abundance of shorter D N A fragments in nodules compared to radicles at 20 min digestion (Fig. 1B). The same is supported by our data on the kinetic of acidsoluble products accumulation upon digestion of nodules and radicle nuclei with DNase ! and micrococcal nuclease [25]. There was no significant endogenous nuclease activity in nodule nuclei since majority of D N A had remained in highmolecular-weight band after 20 min incubation in absence of exogenous enzyme (Fig. 1B). Taken into consideration all obstac'les mentioned above we can assume the D N A in nodule nuclei is more accessible to micrococcal nuclease action due to a more 'open' chromatin conformation. It is believed that local decondensation of chromatin is a prerequisite for active transcription [9, 12, 13]. Additional experiments are needed to establish whether the observed alterations in total chromatin from nodules promotes an increased level of gene expression in some cells.
Histone composition
One of the possible molecular mechanisms to induce the observed changes in average size of linker D N A in pea radicle, callus and nodule tissues could be an alteration in histone composition. Comparison of total proteins from purified nuclei (Fig. 2A) showed that the core histone fractions H4, H3, H2A and H2B are indistinguishable in SDS gel in all three cases. Visual analysis of the stained gels was supported by the
Fig. 2. SDS-PAGE of total nuclear proteins from pea radicle (R), callus (C) and nodules (N) and Western-blotting with antibodies against chromatographically purified pea radical histone H 1. A. Nuclei at a D N A concentration of 2 mg/ml were mixed with an equal volume of 2 x Laemmli sample buffer, lysed by boifing and loaded on 15% acrylamide gel. Staining with Coomassie Blue R 250. B. Identical part of unstained gel was blotted onto nitrocellulose and treated with antibodies against pea H 1 as described in Materials and methods. The figure shows only part of the blot corresponding to H 1 histones, the position of subfraction H 1-6 which increases specifically in nodules is indicated.
immunoblotting data (not shown) with antibodies against core histones. To identify the H 1 fraction we used polyclonal monospecific antibodies raised against purified pea radicle HI [4]. The western blot analysis (Fig. 2B) reveals that pea H1 has a complex subfractional composition. In the radicle this protein is represented by 6 major polypeptide variants 1-6 which differ in molecular weights. Subfraction H 1-1 with the lowest electrophoretic mobility is the most abundant. This result agrees with the data by Berdnikov et al. [3] which was obtained on isolated pea H1 using the electrophoretic separation on 40 cm long acetic acid/ urea gels. In callus the same composition of H1 is preserved, with minor but consistent variations in amounts of the different subfractions between the callus and radicle. Sets of H1 subfractions in nodule nuclei differed from that characteristic of radicles or callus. The content of HI-1 decreased while the amount
1095 of H 1-6 significantly increased, so that these subfractions become almost equal in quantity (Fig. 2 right panel). Fraction H1-6 is not a rhizobial bacteroid polypeptide since the proteins isolated from purified bacteroids and moving within molecular weights range of interest do not crossreact with anti-H1 serum ([32], Zalensky et al., manuscript in preparation). To follow the kinetics of accumulation of nodule enriched subfraction H1-6 we isolated total acid-soluble chromatin proteins from the root segments corresponding to the zone of nodule formation from plants collected at different times after inoculation. Proteins were separated by twodimensional electrophoresis and analyzed by western blotting with anti-Hi antibodies. This procedure is less time- and material-consuming than that utilizing isolation of nuclei. Figure 3 shows that fraction H 1-6 appears very early after plant inoculation with Rhizobia. Approximately at the same time proliferation of cells in nodule primordium take place [21], so we can speculate about relationship of H1 changes with this event. Tentative connection between H1-6 variant and high rate of mitotic activity can be checked by studying histone composition of root tip cells. Participation of H 1-6 in any particular
step of symbiosis could be proved only by more elaborated experiments utilizing cell separation or immunolocalization with specific antibodies. H1 is known as a linker histone interacting with the stretches of D N A connecting adjacent nucleosome core particles [1]. Tissue specific molecular variants of H1 in some mammalian systems were shown to influence linker D N A length [1, 6, 15, 30]. From this, we can speculate that the decrease of linker D N A observed in nodule nuclei is mediated by the lower-molecularweight H 1- 6. H 1 also participates in maintaining higher order structure of chromatin and is involved in chromatin condensation [10, 15, 30]. Non-random distribution of different H1 variants (changing during pea nodule development) along regulatory sequences of genes involved in symbiotic process could be one of the molecular mechanisms modulating their specific expression via local modifications of supranucleosomal structure. Alternatively, the same result may be achieved by differential competition between H 1 variants and transcriptional regulatory proteins [22] or R N A polymerase [ 13]. Based on experimental procedures set up in this work we are planning to study the role of
anti HI
Fig. 3. Western blot (autoradiogram) of the two-dimensional separation of acid-soluble chromatin proteins isolated from root segments (see Materials and methods) at different times after inoculation. Primary antibodies as in Fig. 2, secondary antibodies _ I 3 5 [ protein A. Part of the blot corresponding to position of H1 histones is shown.
1096 chromosomal proteins and chromatin structure in regulation of expression of nodulin genes with established primary structures.
Acknowledgements We would like to thank Dr T. Bisseling and Dr J. Th'ng for critically reading this manuscript. N.P.S. was a Ph.D research fellow supported by the Ministry of Education of India.
References 1. Bawkin SG, Usachenko SJ, Zalensky AO, Mirzabekov AD: Structure of nucleosomes and organization of internucleosomal DNA in chromatin. J Mol Biol 212: 495511 (1990). 2. Behezra R, Cantor CR, Axel R: Nucleosomes are phased along the b-globin gene in erythroid and nonerythroid ceils. Cell 44:697-704 (1986). 3. Belyaev AI, Berdnikov VA: Polymorphism and localization of I-I1 histone genes in Pisum sativurn L. Genetics (USSR) 17:498-504 (1981). 4. Bers EP, Ivanova SV: Application of immunoblotting technique to study Chlamydomonas reinhardtii histones. Vestnik LGU (Leningrad Univ Herald) 1:102-110 (1988). 5. Bers EP, Pardonen VA: Quick method of nuclei isolation from some plant tissues. Tsitologiia (Cytology, Aead Sci USSR) 32:899-902 (1990). 6. Clark DJ, Thomas JO: Differences in the binding of H1 variants to DNA: Cooperativity mad linker length related distribution. Eur J Biochem 178: 225-233, 1988. 7. de Bruijn FJ, Szabados L, Shell J: Chimeric genes and transgenic plants are used to study the regulation of genes involved in symbiotic plant-microbe interactions (Nodulin genes). Devel Genet 11:182-196 (1990). 8. Elgin SCR: The formation and function of DNAse I hypersensitive sites in the process of gene activation. J Biol Chem 263:19259-19262 (1988). 9. Felsenfeld G: Chromatin as an essential part of the transcriptional mechanism. Nature 355:219-224 (1992). 10. Finch JT, Klug A: A solenoidal model for superstructure of chromatin. Proc Natl Acad Sci USSR 73:1897-1901 (1976). 11. Grellet F, Pennon P, Cooke R: Analysis of DNA associated with nueleosomes in pea ehromatin. Planta 148: 346-353 (1980). 12. Gross RD, Garrard WT: Poising chromatin for transcription. Trends Bioehem Sei 12:293-297 (1987). 13. Grustein M: Histone function in transcription. Annu Rev Cell Biol 6:643-678 (1990). 14. Harmon R, Bateman E, Allan J, Harborne N, Gould H: Control of RNA polymerase binding to ehromatin by variations of linker histone composition. J Mol Biol 180: 131-150 (1985). 15. van Holde KE: Chromatin. Springer-Verlag (1988).
16. Jensen EO: Regulation of nodule-expressed soybean genes. In: Hennecke H, Verma DPS (eds) Advances in Molecular Genetics and Plant-Microbe Interactions, Vol. 1, pp. 310-316. Kluwer Academic Publishers, Dordrecht (1991). 17. Kahl G, Weising K, Gorrz A, Schafer W, Hirasawa E: Chromatin structure and plant gene expression. Devel Genet 8:405-434 (1987). 18. Kulikova OA, Muravieva MYu, Zalensky AO, Filatov AA, Tikhonovich IA: Identification of proteins related to symbiotic nitrogen fixation in the pea. Appl Biochem Microbiol (USSR) 26:506-513 (1990). 19. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 (1970). 20. Lee M-S, Garrard WT: Transcription induced 'splitting' and underlying structure for DNAse I sensitive chromatin. EMBO J 10: 607-615, 1991. 21. Nap J-P, Bisseling T: Developmental biology of plantprokaryote symbiosis: The legume root nodule. Science 250:948-954 (1990). 22. Postnikov YV, Shick VV, Belyavsky AV, Khrapko KR, Borodin KL, Nikolskaya TA, Mirzabekov AD: Distribution of high mobility group proteins 1/2, E and 14/17 and linker histones HI and H5 on transcribed and nontranscribed regions of chicken erythrocyte chromatin. Nucl Acids Res 19:717-725 (1991). 23. Reid RA, John MC, Amasino RM: Deoxyribonuclease I sensitivity of the T-DNA ipt gene is associated with gene expression. Biochemistry 27:5748-5754 (1988). 24. Shlissel MS, Brown DD: The transcriptional regulation of Xenopus 5S RNA genes in chromatin: the roles of active stable transcription complexes and histone H 1. Cell 37:903-913 (1984). 25. Singh NP, Zalensky AO: Nuclear proteins and chromatin structure of pea (Pisum sativum) root nodule. Nucleus (in press). 26. Ull MA, Franco L: The nucleosomal repeat length of pea (Pisum sativum) chromatin changes during germination. Plant Mol Biol 7:25-31 (1986). 27. Vayda ME, Freeling M: Insertion of the Mut transposable element into the first intron of maize Adh 1 interferes with transcription elongation but does not disrupt chromatin structure. Plant Mol Biol 6:441-454 (1986). 28. Willmitzer L, Wagner KG: The isolation of nuclei from tissue-cultured plant cells. Exp Cell Res 135:67-77 (1981). 29. Wurtzel ET, Burr FA, Burr B: DNAase I hypersensitivity and expression of the Shrunken- 1 gene of maize. Plant Mol Biol 8:251-264 (1987). 30. Zalenskaya IA, Pospelov VA, Zalensky AO, Vorob'ev VI: Nucleosomal structure of sea urchin and starfish chromatin. Nucl Acids Res 9:473-487 (1981). 31. Zalenskaya IA, Odintsova NA, Zalensky AO, Vorob'ev VI: Nucleosome organization of chromatin from sperm of the bivalvia mollusc. Mol Biol (USSR) 16:335-344 (1982). 32. Zalensky AO, Singh NP, Pardonen VA, Bers EP, Aronstam AA: Changes in structural organization of plant and bacterial genome during Pisurn sativum-Rhizobiurn legurninosarum symbiosis. Int. Symposium on Molecular Genetics and Plant-Microbe Interactions (Interlaken), p. 238 (1990).
