Plant Cell, Vol. 11, 2317-2330, December 1999, Copyright © 1999, American Society of Plant Physiologists
Linker Histones Play a Role in Male Meiosis and the Development of Pollen Grains in Tobacco
Marta Prymakowska-Bosak1,a,
Marcin R. Przewlokaa,
Joanna 
lusarczykb,
Mieczyslaw Kura
b,
Jacek Lichotaa,
Beata Kilia
czyka,c, and
Andrzej Jerzmanowskia,c
a Laboratory of Plant Molecular Biology, Warsaw University, Pawinskiego 5A, 02-106 Warsaw, Poland
b Laboratory of Plant Morphogenesis, Warsaw University, Pawinskiego 5A, 02-106 Warsaw, Poland
c Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Pawinskiego 5A, 02-106 Warsaw, Poland
Correspondence to:
Andrzej Jerzmanowski, andyj{at}ibb.waw.pl (E-mail), 4822-6584636 (fax)
 |
ABSTRACT |
|---|
To examine the function of linker histone variants, we produced transgenic tobacco plants in which major somatic histone variants H1A and H1B were present at ~25% of their usual amounts in tobacco chromatin. The decrease in these major variants was accompanied by a compensatory increase in the four minor variants, namely, H1C to H1F. These minor variants are smaller and less highly charged than the major variants. This change offered a unique opportunity to examine the consequences to a plant of major remodeling of its chromatin set of linker histones. Plants with markedly altered proportions of H1 variants retained normal nucleosome spacing, but their chromosomes were less tightly packed than those of control plants. The transgenic plants grew normally but showed characteristic aberrations in flower development and were almost completely male sterile. These features correlated with changes in the temporal but not the spatial pattern of expression of developmental genes that could be linked to the abnormal flower phenotypes. Preceding these changes in flower morphology were strong aberrations in male gametogenesis. The earliest symptoms may have resulted from disturbances in correct pairing or segregation of homologous chromosomes during meiosis. No aberrations were observed during mitosis. We conclude that in plants, the physiological stoichiometry and distribution of linker histone variants are crucial for directing male meiosis and the subsequent development of functional pollen grains.
|
|
INTRODUCTION |
|---|
Eukaryotic DNA is assembled into chromatin, in which it is packaged by histone proteins into nucleosomes. A complete nucleosome contains eight core histone molecules and one linker histone molecule bound to a variable-length linker DNA that extends between neighboring nucleosomes. Linker histones include histone H1 and its various somatic variants (e.g., H5 and H10 in animals and a drought-inducible H1 in plants) as well as gamete-specific variants. A typical metazoan linker histone has a three-domain structure with a central globular domain flanked primarily by unstructured and highly charged N- and C-terminal tails. Linker histones are by far the most evolutionarily variable of all histones. Most organisms express several different H1 variants in different cell types and during various developmental stages (Newrock et al. 1977 
).
This variability among the linker histones can be potentially important in determining the accessibility of chromatin DNA to the elements of the transcriptional apparatus. Thus far, however, the only well-documented case of such a function for H1 variants has been the differential transcriptional regulation of the 5S rRNA genes during early development in Xenopus (Bouvet et al. 1994 ; Tomaszewski and Jerzmanowski 1997 ). Thus, whereas the core histones have been shown to play an essential role in regulating access to gene promoters (Cairns 1998 ), the role of the different variants of linker histones in chromatin is far less clear.
Until recently, the presence of linker histones in chromatin was thought to be essential for the folding of chromatin fibers, and histone H1 was considered to be a general repressor of transcription. The current view, based on analysis of the condensation properties of H1-depleted chromatin, is that linker histones, although able to facilitate and guide the proper folding of nucleosomal fibers, do not cause the folding per se (van Holde and Zlatanova 1996 ). Some evidence suggests that H1 may be involved in selective regulation of specific classes of genes; however, the most recent data strongly support the notion that linker his-tones are not general repressors of transcription in vivo (Prymakowska-Bosak et al. 1996 ; Shen and Gorovsky 1996 ; Steinbach et al. 1997 ). For example, we have shown that overexpression of a somatic-type histone H1 in tobacco, which more than doubled the physiological H1/DNA ratio in chromatin and resulted in a marked increase in chromatin condensation, had little effect on basal cellular functions (Prymakowska-Bosak et al. 1996 ).
In this work, we reexamine the function of linker histones by analyzing the molecular and morphological phenotypic consequences of a marked change in the proportions of linker histone variants achieved by using an antisense approach with transgenic tobacco. We conclude that in plants, the native stoichiometry of linker histones is critical for the correct course of male meiosis and the subsequent development of functional pollen grains.
|
|
RESULTS |
|---|
Identification of Linker Histone Variants in Tobacco
Selective extractability with 5% perchloric acid (PCA) is a distinctive property of linker histones (Cole 1989 ). Using acetic acid–urea PAGE, we analyzed nuclei extracted with 5% PCA from various tobacco tissues and detected two major and four minor histone bands (Figure 1A). All six bands were also detected on protein gel blots with antibodies raised against plant histone H1; however, the reaction for the bands identified as H1C and H1D was weaker than for those identified as H1A, H1B, H1E, and H1F (Figure 1B). None of the six bands was recognized by antibodies raised against the tobacco high-mobility group HMG1 protein (results not shown). The sizes of the two major bands (H1A and H1B) corresponded well to that of the typical major somatic plant H1 (equivalent to 250 to 280 amino acid residues; Gantt and Lenvik 1991 ) and to the size predicted from the single histone H1 cDNA sequence that has been reported from tobacco, which encodes a protein of 282 amino acids (Szekeres et al. 1995 ). To determine whether this cDNA sequence codes for H1A or H1B, we used it as a template in an in vitro transcription/translation assay. In acetic acid–urea PAGE, the 35S-labeled product of the in vitro translation reaction comigrates with the faster migrating (H1B) of the two major tobacco H1 variants (Figure 1C).
View larger version (45K):
[in this window]
[in a new window]
|
Figure 1.
Identification of Linker Histone Variants in Tobacco.
(A) Analysis by acetic acid–urea PAGE of proteins extracted by 5% PCA from different tissues of tobacco plants. The amino acid sequences denote four of several peptide sequences obtained upon microsequencing of the proteins in the indicated bands that allowed us to identify them as linker histone variants. L, leaves; S, stems; F, flowers; FB, floral buds.
(B) Analysis by acetic acid–urea (AU) PAGE and immunodetection with anti-H1 antibodies of the proteins extracted with 5% PCA from flowers of control (lanes 1) and H1A- and H1B-deficient (lanes 2; see Figure 2) tobacco plants. H1A to H1F are as given in (A).
(C) Identification of the major variants of histone H1. The cDNA of histone H1 used in the construct shown in Figure 2 was used as a template for in vitro transcription and translation reactions in the presence of 35S-methionine, as described in Methods. The protein products of the reaction were analyzed on acetic acid–urea polyacrylamide gels. Lane 4, reaction products mixed with unlabeled tobacco H1A and H1B; lane 3, reaction performed without adding histone H1 cDNA or unlabeled tobacco H1A and H1B; lane 2, autoradiogram of lane 4; and lane 1, autoradiogram of lane 3. H1A and H1B denote major variants of tobacco histone H1.
|
|
To determine the identity of the four minor proteins, the corresponding bands were excised from the gel and subjected to proteolytic degradation. Several peptides from each band were separated by HPLC and microsequenced, and the sequences were used to search the databases with the BLAST program (Altschul et al. 1997 ). The peptides from all four bands showed the greatest similarity to those of the plant histone H1 family. The sequence of the peptide THPPYFQMIK from the H1C band is identical to residues 48 to 57 of tomato H1D (Wei and O'Connell 1996 ) and residues 23 to 32 of Arabidopsis H1-3 (Ascenzi and Gantt 1997 )—both proteins that have been described as being the specific drought-inducible variants of histone H1. A second peptide from the H1C band, HKDELPANFRK, is identical to residues 82 to 92 of tomato H1D and differs by only four amino acids from the corresponding sequence of Arabidopsis H1-3.
When aligned with the whole-protein sequences, the sequenced peptides mapped to the globular domain of tomato H1D and Arabidopsis H1-3. The sequence of the polymerase chain reaction (PCR) fragment obtained with tobacco cDNA and extending between the primers corresponding to the sequenced peptides showed extensive similarity to the corresponding region of tomato H1D and Arabidopsis H1-3; resemblance to the same region of the major somatic H1 from tobacco was less complete but still significant (results not shown). The peptide EALLALNEK from the H1D band on Figure 1A mapped to the globular region of H1. This peptide is identical to residues 58 to 66 of tomato H1-D, differing by three amino acids a corresponding region in Arabidopsis H1-3. Similarly, as with the peptides from the H1C band, the peptide EALLALNEK from the H1D band showed less similarity to the corresponding fragment of the major somatic H1 from tobacco than to corresponding fragments of tomato H1D and Arabidopsis H1-3. Proteins represented by the H1C and H1D bands are much smaller than the major plant H1 histones (Figure 1A), and their amino acid composition indicates a much lower net positive charge than that of a typical plant H1 (Table 1). These characteristics are very similar to those reported for drought-inducible variants of H1 described in tomato (H1D) and Arabidopsis (H1-3). On the basis of these similarities and the very high degree of sequence identity between peptides from the H1C and H1D bands and those from tomato H1D and Arabidopsis H1-3, we conclude that the proteins identified as H1C and H1D in Figure 1A are the tobacco homologs of the drought-inducible histone H1 variants.
The peptide sequences ERTGSSQVAIAK and ERTGSSQFAIAK, derived from bands H1E and H1F, respectively, correspond to the globular domain of H1 (residues 76 to 85 in the major somatic variant of tobacco H1) and show a similarly close resemblance to both the major tobacco H1 and the drought-inducible variants from tomato and Arabidopsis. However, the proteins represented by bands H1E and H1F are markedly smaller than the typical plant H1. Their net positive charge is less than that of major somatic H1 but more than that of the proteins in bands H1C and H1D (Table 1). From the above features, we conclude that proteins in bands H1E and H1F are low molecular mass histone H1 variants. Because the anti-H1 antibody detects variants H1E and H1F with a sensitivity that is similar to that for the major somatic variants H1A and H1B and greater than that for variants H1C and H1D (Figure 1B), histones H1E and H1F probably are not homologs of the drought-inducible variants.
Transformation of Tobacco with Antisense cDNA for Histone H1B Results in a Marked Decrease in Major Somatic H1 in the Chromatin of Transgenic Plants
In the chromatin of tobacco, the two variants of major somatic H1 (H1A and H1B) occur in a roughly 1:1 ratio (see Figure 1A). We adopted an antisense strategy to decrease the amount of major somatic H1 in vivo, using the pRok19F vector bearing the cDNA of the H1B variant placed in reverse orientation under the control of the constitutive cauliflower mosaic virus 35S promoter (Figure 2A). The vector was integrated into the tobacco genome by Agrobacterium-mediated transformation of young seedlings.
View larger version (47K):
[in this window]
[in a new window]
|
Figure 2.
Construction of Transgenic Tobacco Plants That Express the Antisense Histone H1B mRNA and Analysis of Their Histones.
(A) Construct used for transformation. cDNA (H1Tob) of tobacco histone H1 was placed in the reverse orientation between the enhanced cauliflower mosaic virus 35S promoter (F) and the cauliflower mosaic virus 35S terminator (t35S). KAN, kanamycin resistance cassette; RB and LB, right and left borders, respectively, of the T-DNA in the transformation vector.
(B) Analysis by acetic acid–urea PAGE of total histones extracted with 0.2 M H2SO4 from leaf chromatin. Lane 1, sample from a transgenic plant expressing antisense mRNA of tobacco histone H1B; lane 2, sample from a transgenic plant transformed with an empty vector (control). Samples loaded on each lane were of equal volume and were obtained from equal amounts of tissue. H1A and H1B denote major variants of histone H1; CORE denotes core histones.
(C) Analysis by acetic acid–urea PAGE of proteins extracted with 5% PCA from stems, leaves, and flowers of H1A- and H1B-deficient (lanes 1) and control (lanes 2) tobacco plants. For each of the analyzed organs, the samples loaded on each of two lanes were of equal volume and were obtained from equal amounts of tissue. H1A to H1F denote variants of histone H1.
|
|
Analysis of total chromatin histones from leaves of transgenic plants bearing the antisense H1B cDNA and expressing antisense H1B mRNA, as determined by hybridization (results not shown), showed a marked decrease in the amount of the H1B variant and a less-marked decrease in the H1A variant, whereas no change in the amount of core histone was observed (Figure 2B). The method used for extraction of total histones (with diluted sulfuric acid) was not as effective for the minor variants of H1 (H1C to H1F) as was extraction with 5% PCA. The extent of decrease of the major variants of H1 (H1A and H1B) in chromatin isolated from leaves, stems, or flowers varied in different plants. In >80% of the plants analyzed, H1B was decreased by 80 to 100% and H1A by 30 to 50%, compared with their amounts in untransformed plants. However, in the same plants, there was a marked increase of the minor variants of H1 (H1C to H1F) in all organs, most notably in flowers (Figure 2C). Quantification of the total amount of linker histones in plants deficient in major somatic variants of H1 showed that amounts of H1 averaged 85 to 90% of amounts in control plants. That is, reduction of the major variants was almost fully compensated by increases in the minor variants of H1. The net result was that in antisense plants, the proportions of the linker histone variants differed considerably from those in wild-type plants.
In H1A- and H1B-Deficient Plants, Nucleosome Spacing Is Not Affected, but the Chromatin Has a More Extended Organization than in Control Plants
Comparison of DNA products after micrococcal nuclease digestion of chromatin from flowers of the control and H1A- and H1B-deficient plants showed no meaningful difference in average nucleosomal spacing (Figure 3A). To compare the properties of chromatins with two different sets of linker histones (Figure 3B), we analyzed their sedimentation in linear sucrose gradients. The nucleoprotein complexes from the two chromatins that migrated at the same velocity within the sucrose gradient differed in size. This was seen most clearly for longer oligonucleosomes (Figure 3A), those from the H1A- and H1B-deficient chromatin being on average longer than those from control chromatin. Assuming that the DNA/protein ratio of the two chromatins is roughly the same (the loss of H1A and H1B being compensated by an increase in H1C to H1F; see Figure 3B), the results of this analysis indicate that the conformation of the H1A- and H1B-deficient chromatin is more extended than that of control chromatin.
View larger version (47K):
[in this window]
[in a new window]
|
Figure 3.
Sedimentation Analysis by Sucrose Gradient Centrifugation of Chromatin from H1A- and H1B-Deficient and Control Tobacco Plants.
(A) Distribution of chromatin fragments from H1A- and H1B-deficient and control plants in gradient fractions with the same sucrose concentration. Lanes 1, H1A- and H1B-deficient plants; and lanes 2, control. Chromatin obtained from flowers of H1A- and H1B-deficient and control plants was digested with micrococcal nuclease and sedimented through a 10 to 43% linear sucrose gradient. DNA isolated from individual fractions of the gradient was separated on a 1.5% agarose gel. Sucrose concentrations are denoted by roman numerals as follows: I, 31%; II, 25%; III, 22%; IV, 19%; and V, 17%. d, dinucleosomes; m, mononucleosomes; t, trinucleosomes.
(B) Linker histones in H1A- and H1B-deficient (lane 1) and control (lane 2) chromatins used for sedimentation analysis. H1A to H1F denote variants of histone H1.
|
|
Vegetative Growth of Plants Deficient in H1A and H1B Is Normal, but the Plants Produce No Seeds and Show Characteristic Changes in Flower Development
Transgenic T0 plants with changed stoichiometry of the linker histone variants had a basically normal appearance, except that at the flowering stage, they were 20 to 30% smaller and were bushier than the control plants, which had been regenerated from tissue transformed with an empty vector. The major differences between the transgenic plants and the control plants were related to flower morphology and the ability to produce seeds. All mature flowers of the H1A- and H1B-deficient plants had shorter petals and stamens than did control plants. Consequently, all of the transgenic flowers had a protruding pistil, which gave them the characteristic appearance shown in Figure 4A. These flowers did not form fruits and produced no seeds (Figure 4B), perhaps because of the differences in length between stamens and pistil, which presumably prevented self-pollination. However, we also found that >90% of the pollen grains of the H1A- and H1B-deficient plants were severely deformed and lacked the vegetative and generative cells typically seen in pollen from the control plants (Figure 4C). Moreover, >90% of the pollen grains failed to germinate; that is, they were not viable.
View larger version (59K):
[in this window]
[in a new window]
|
Figure 4.
Histone H1A and H1B Deficiency Is Linked with Changed Floral Morphology and Aberrant Development of Pollen Grains.
(A) Longitudinal sections of flowers of H1A- and H1B-deficient (H1A/B DEF.) and control (CONTR.) plants.
(B) Development of floral bud, flower, and fruit in H1A- and H1B-deficient (H1A/B-DEFICIENT) and control plants. Rulers on the left show measurements in centimeters.
(C) Scanning (top) and transmission (bottom) electron microscopy of mature pollen grains from H1A- and H1B-deficient (H1A/B-DEFICIENT; left) and control tobacco plants (right). Note that on the cross- sections (bottom), the vegetative (VC) and generative (GC) cells visible in pollen grain from the control plants are not formed in the pollen grain of the H1A- and H1B-deficient plants. Note also the difference in the thickness of intine and exine layers (arrow and arrowheads, respectively) between pollen grains from H1A- and H1B-deficient plants and from the control plants. Bars in top = 50 µm; bars in bottom = 3 µm.
|
|
We performed four independent transformation experiments with the antisense H1B cDNA, producing ~30 transgenic plants in each experiment. In each of the four experiments, >90% of the plants obtained had the same characteristic appearance of flowers, as shown in Figure 4A and Figure 4B. DNA gel blot analyses of individual regenerated plants revealed that the integration pattern of the transgene was not very complicated (from one to at most three or four loci). Because different integration patterns resulted in the same phenotype, we ruled out the possibility that the characteristic phenotypic changes were caused by a unique site-specific integration of the transgene. All aberrant plants expressed the antisense cDNA (results not shown) and were H1A and H1B deficient, as determined by gel electrophoresis. By performing self-crosses and outcrosses, we determined that the aberrant flower morphology could be linked to the presence of antisense H1B cDNA and to the H1A and H1B deficiency. In some of the T1 plants, a few flowers had a normal appearance, probably because of a local silencing of the transgene. Moreover, analysis of the histones extracted from these normal-looking flowers showed a wild-type profile of linker histone variants (results not shown), indicating that the abnormal flower phenotype is linked to the changed linker histone profile. The results of the self-crosses confirmed that a small fraction of pollen grains in H1-deficient plants remained viable.
In H1A- and H1B-Deficient Plants, the Transcription of Some Flower-Specific Genes Is Affected
To determine whether the changed stoichiometry of linker histones could be correlated with changes in the activity of genes during flower development, we hybridized total RNA from floral buds at four early stages of development and from organs of the mature flowers of the H1A- and H1B-deficient and control plants with cDNA probes that represented two classes of genes. We found that one class is temporally and spatially regulated and the other is expressed ubiquitously (Figure 5).
View larger version (66K):
[in this window]
[in a new window]
|
Figure 5.
Linker Histones Content and the Transcription Pattern of Developmentally Regulated and Ubiquitously Expressed Genes during Flower Bud Development and in Organs of the Mature Flower in H1A- and H1B-Deficient and Control Tobacco Plants.
(A) and (B) Linker histones in developing buds and mature flower organs of H1A- and H1B-deficient and control plants analyzed by acetic acid–urea PAGE. Lanes 1 to 4, stages in the development of flower bud, as described in Results. S, sepals; P, petals; St, stamens; Pi, pistils.
(C) and (D) Transcription profiles of developmentally regulated (Nap3, Ta29, and Tac25) and ubiquitous (Pbf46) genes in developing buds and in mature flower organs in H1A- and H1B-deficient and control plants. The quantitative results, expressed in arbitrary units per unit amount of RNA, are from RNA gel blot hybridization analysis (shown below the diagrams) of total RNA isolated from buds and flower organs (RNA shown at the bottom) using the radioactive cDNA probes. The hybridization signals and the amounts of RNA on the gels were quantified by using the ImageQuant program (Molecular Dynamics). For RNA isolation, the developing buds and flower organs were pooled from many individual plants. The results thus represent an arithmetic average for the populations of H1A- and H1B-deficient and control plants. Similar results were obtained for a second independent series of experiments (not shown).
|
|
The flower buds studied were from stages 1 to 4, according to the description proposed by Koltunov et al. 1990 . Stages 2 to 4 correspond to the postmeiotic period of development. Stage 1 (8- to 9-mm floral bud) is the period immediately after meiosis in which anthers and pistil are fully differentiated but not yet mature. At stage 4 (16-mm flower bud), sepals are completely separated at the top of the calyx. The last (12th) stage in this classification corresponds to the fully open mature flower, with a length of ~46 mm. Note that in the control plant, during the later stages of bud development, the proportion of H1C and H1D variants decreased in comparison with the H1E and H1F variants (Figure 5B); there was no similar decrease of H1C and H1D in the H1A- and H1B-deficient plant (Figure 5A). Note also the high concentrations of minor variants H1C to H1F in the flower organs of the H1A- and H1B-deficient plant (Figure 5A), which is in keeping with the results of Figure 2. Because the RNA was isolated from pooled material consisting of flower buds and flower organs from many individual plants, the hybridization results represent an arithmetic average for populations of the H1A- and H1B-deficient plants and the control plants.
Strikingly, the H1A and H1B deficiency could be correlated with opposite effects on transcription of the two developmentally regulated genes Nap3 and Ta29. Nap3 is a tobacco homolog of the Arabidopsis APETALA3 (AP3) gene (C.D. Day and V.F. Irish, personal communication), which plays an important role in floral development in that species. Ta29 encodes a tapetum-specific glycine-rich protein (Goldberg et al. 1993 ). Interestingly, in the H1A- and H1B-deficient plants, there was no change in the spatial pattern of expression of the two genes in the mature flower. Similarly, as is also the case in the control plants, the expression of Nap3 was restricted to stamens and petals, and that of Ta29 was not detected in mature flowers (Figure 5C). However, the H1A and H1B deficiency correlated with selective abolishment of Nap3 expression in stages 3 and 4 of flower bud development (cf. Figure 5C and Figure 5D). In H1A- and H1B-deficient plants, the Ta29 gene was induced already at stage 1, whereas in the control plants, the induction was not until stage 3 (cf. Figure 5C and Figure 5D). This pattern of induction of Ta29 in control plants has been reported for this gene by others (Goldberg et al. 1993 ).
In Arabidopsis, the AP3 gene is required for correct development of stamens and petals. Downregulation at the early stages of flower development of the related Nap3 gene thus could be a direct cause of the aberrant flowers on the H1A- and H1B-deficient plants. Regarding a third selectively expressed gene, Tac25, which encodes a pollen-specific actin (Thangavelu et al. 1993 ), H1A and H1B deficiency did not markedly disturb its temporal and spatial pattern of expression. The general pattern of expression for the ubiquitously expressed genes Pbf46 (encoding the translation elongation factor EF2) and Ubp1 (encoding a protein involved in splicing; results for this gene are not shown) showed no change in the H1A- and H1B-deficient plants, although there were some characteristic differences in the abundance of the two transcripts compared with those in the control plants. Both genes were upregulated until stage 3 and then downregulated during stage 4, and the amounts of both transcripts were also decreased in the mature flower parts, except in the pistil (cf. Figure 5C and Figure 5D).
Taken together, these data indicate that the H1A and H1B deficiency, although not correlating with changes in the spatial specificity of gene expression, does correlate with changes in the temporal pattern of expression of certain developmentally regulated genes.
Male Gametogenesis Is Strongly Disturbed in H1A- and H1B-Deficient Plants
In tobacco plants, male gametogenesis begins with division of the diploid sporophytic cell into the tapetal initial and the sporogenous initial (pollen mother cell). The pollen mother cell undergoes meiosis, resulting in a tetrad of haploid cells. Individual tetrads are released as free microspores. The uninucleate minispores undergo an asymmetric mitotic division (pollen mitosis I), resulting in a pollen grain containing two cells: a larger vegetative cell and a smaller generative cell. The pollen grain is released from the anther at this stage. The second mitotic division of the generative cell, resulting in two sperm cells, occurs when the pollen tube grows through the female pistil (McCormick 1993 ).
In the H1A- and H1B-deficient plants, most of the pollen grains present in the mature anthers had an abnormal appearance. Cross-sections of these defective pollen grains showed no vegetative or generative cells (Figure 4C). To determine the earliest perturbations in male spermatogenesis, we examined microscopically the initial stages of this process. Although there was no difference in the appearance of nondividing pollen mother cells in control and H1A- and H1B-deficient plants, a considerable fraction of pollen mother cells of the deficient plants showed serious aberrations at the stage of meiotic division (Figure 6C to 6F, and Figure 6I to 6L). In contrast to a regular first meiotic division in the control plants (Figure 6A and Figure 6B), many cells in the H1A- and H1B-deficient plants lost numerous "minichromosomes" that failed to assemble at the equatorial plate (e.g., Figure 6C and Figure 6E) at metaphase and during early anaphase of the first meiotic division. At anaphase and telophase of the first meiotic division, some of the cells had aberrations in cytokinesis (e.g., Figure 6D and Figure 6F). Cells with micronuclei were also frequent at the second meiotic division (Figure 6K), which, unlike the second meiotic division in the control plants (Figure 6G and Figure 6H), was very asynchronous and often resulted in aberrant and irregular tetrads with unusually arranged cells (Figure 6I, Figure 6J, and Figure 6L). The quantitative analysis of these observations is given in Table 2 and Table 3.
View larger version (69K):
[in this window]
[in a new window]
|
Figure 6.
Male Meiosis Is Impaired in Histone H1A1- and H1B-Deficient Plants.
Male meiotic divisions I and II in control and H1A- and H1B-deficient (H1A/B DEFICIENT) plants was analyzed by light microscopy.
(A) to (F) Light microscopy of male meiotic division I. (A) Metaphase I, control plants. (B) Telophase I, control plants. (C) and (E) Metaphase I to early anaphase I, H1A- and H1B-deficient plants. (D) and (F) Anaphase I to telophase I, H1A- and H1B-deficient plants.
(G) to (L) Light microscopy of male meiotic division II. (G) Post-telophase II, control plants. (H) Microspore tetrads, control plants. (I) and (K) Post-telophase II, H1A- and H1B-deficients plants. (J) and (L) Microspore tetrads, H1A- and H1B-deficient plants.
Arrows in (C), (E), (F), and (K) show examples of chromosomal abnormalities (broken chromosomes and minichromosomes). Note the regular pattern and full synchrony of meiotic events in cells from the control plants in (A), (B), (G), and (H) in contrast to the aberrations and high degree of asynchrony in cells of corresponding stages from the H1A- and H1B-deficient plants. In (C) to (F) and (I) to (L), the microscopic analyses were performed on plants from three independent transformation experiments with the H1B cDNA. The results were similar for each of the analyzed lines. Bars in (A) to (L) = 10 µm.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2.
Type and Frequency of Meiotic Aberrations during Male Meiotic Divisions in 8-mm Flower Buds in Control and H1A- and H1B-Deficient Tobacco Plants
|
|
In summary, these data indicate that the first structural defects in male spermatogenesis characteristic for H1A- and H1B-deficient plants could be the aberrant meiotic division of pollen mother cells. We did not see any noticeable defects in the mitotic chromosomes or in the course of mitosis in H1A- and H1B-deficient plants (results not shown).
We also examined microscopically the postmeiotic stages of microsporogenesis and gametogenesis in anthers at different stages of development in the H1A- and H1B-deficient plants and the control plants. We found that in the H1A- and H1B-deficient plants, 90 to 95% of the microspores failed to undergo asymmetric mitotic division (pollen mitosis I) but remained in the form of uninucleate cells.
We have not analyzed microscopically the appearance and development of female germ cells. However, in the outcrosses of H1A- and H1B-deficient plants, in which pollen grains were from the control plant, the amount of seeds produced was consistently less than that produced following pollination of control plants. Whether this indicates some disturbances also in the female gametogenesis remains to be tested.
|
|
DISCUSSION |
|---|
We have shown that the linker histone complement in tobacco consists of six variants, two (H1A and H1B) typical major somatic variants and four (H1C, H1D, H1E, and H1F) minor variants. All of the minor variants are considerably smaller and carry less charge than the major variants. H1C and H1D are most probably the tobacco homologs of the specific drought-inducible linker histone variants described in tomato (Wei and O'Connell 1996 ) and Arabidopsis (Ascenzi and Gantt 1997 ). Using an antisense approach with a histone H1B cDNA, we constructed transgenic plants in which the major linker histone variants H1A and H1B had ~25% of the normal amounts of chromatin. This reduction, however, was almost completely compensated by the simultaneous increase in the chromatin amounts of the minor linker histone variants H1C to H1F. The final result was transgenic tobacco plants in which the stoichiometry of linker histone variants was markedly different from that characteristically found in wild-type plants, offering a unique opportunity to examine the consequences of such a major remodeling of the chromatin set of linker histones.
The plants with changed stoichiometry of linker histone variants had normal vegetative growth but showed distinctive morphological aberrations during flower development. The most characteristic of these aberrations were abnormally developed stamens and corollas. The stamens and petals were shortened, which resulted in styles protruding from corollas. Because this appearance was highly reproducible and always linked to perturbed stoichiometry of H1 variants, we conclude that the change in proportions of linker histones is causal to the observed aberrations in flower development.
Strikingly, the major reconfiguration of the quantitative proportions of linker histone variants in chromatin had practically no effects on morphological development until flowering. Indeed, the changed stoichiometry of linker histone variants correlated with the premature cessation of transcription of the Nap3 gene during flower bud development (Figure 5). Because Nap3 is a homolog of the Arabidopsis AP3 gene, which is a major participant in the regulation of flower development and a gene transcribed only in stamens and petals (Hill et al. 1998 ), perturbation in Nap3 transcription could well be linked to the underdevelopment of these organs in H1A- and H1B-deficient plants. On the other hand, the H1A and H1B deficiency also correlated with an accelerated induction during flower development of the strictly regulated Ta29 gene encoding a tapetum-specific protein. Remarkably, the perturbations of the two different developmental genes in the H1A- and H1B-deficient plants concern the temporal pattern but not the spatial pattern of transcription. This observation should be interpreted carefully. If linker histone variants play an important role in determining the overall accessibility of DNA to transcription factors, a marked change in the stoichiometry of H1 variants would be expected to affect the spatial specificity of transcription. The fact that only the temporal specificity has been affected could indicate that this result is not a direct effect of changes in the H1 variants in chromatin but rather is a generalized regulatory response to perturbations occurring earlier in flower development.
Histone H1 has long been associated with the maintenance of higher order structure in chromosomes. In tobacco, most of the somatic H1 is retained in chromosomes during male meiosis (Oakeley et al. 1997 ). In this work, we noticed that H1A and H1B deficiency is linked with characteristic chromosomal aberrations in male meiosis, that is, at the stage preceding the maturation of the flower organs. The aberrations are first visible at metaphase, anaphase, and telophase of the first meiotic division—the appearance of micronuclei, which could result from broken chromosomes, chromosomes, or parts of chromosomes scattered throughout the cytoplasm and from chromosomes that are not completely separated to the poles. A characteristic feature of H1A- and H1B-deficient plants is the loss of synchrony of the meiotic divisions of pollen mother cells during spermatogenesis (Table 3). The data in Table 2 make clear that observable large-scale meiotic defects do not involve all of the chromosomes. Although many of the apparently normal chromosomes can have smaller structural defects that are not manifested as visible chromosomal aberrations, the data do point out the somewhat random distribution of the defects in chromosomal structure. (A very small but reproducible amount of chromosomal aberrations also occurs in control plants, as documented in Table 2.) The randomness of the structural effects in chromosomes of the H1A- and H1B-deficient plants may reflect a random distribution of the compensatory histones (H1C to H1F) among the binding sites of the major H1A and H1B histones both within and between the chromosomes. This could explain the normal occurrence of meiosis in some cells of the H1A- and H1B-deficient plants and could account for the production in these plants of a small number of functional pollen grains.
The chromosomal defects observed resemble those in the male synaptic mutants that have been described for other plant species (Peirson et al. 1997 ). These mutants have been characterized by the presence of unpaired chromosomes, of chromosomes not associated with the telophase I poles, of small and large groupings of DNA scattered throughout the cytoplasm, and of abnormally distributed microtubules. All of the microspores from these mutants were found to collapse soon after the degradation of the surrounding wall, leading to complete male sterility. Female fertility in these mutants also was affected. The underlying cause of the above abnormalities may be a defect in the formation of synaptonemal complexes or some cohesion between sister chromatids of homologous chromosomes, or both. If a similar defect caused the aberrations observed during meiotic segregation in the H1A- and H1B-deficient plants, the reason could be either deregulation attributable to the change in H1 variants of specific genes (e.g., those encoding the components of the synaptonemal complex) or impairment of the gross chromatin structure essential for proper recognition and separation of homologous chromosomes. Although we cannot exclude that some genes encoding specific synaptonemal components could have been affected, several lines of evidence point to the possibility of a direct link between the observed aberrations and the changes in chromatin structure caused by perturbed proportions of H1 variants.
Evidence is strong that the heterochromatin regions play a key role in establishing and maintaining the alignment of homologous chromosomes during meiosis (Derenburg et al. 1996 ; Karpen et al. 1996 ; Renauld and Gasser 1997 ). Because histone H1 is enriched in heterochromatin and is thought to stabilize the heterochromatic regions in chromosomes, the overall distribution of variants can greatly affect the global heterochromatin pattern of the chromosomes. In previous work, we showed that overexpression of major somatic H1 in tobacco indeed led to a considerable increase in the amount of heterochromatin in nuclei (Prymakowska-Bosak et al. 1996 ). Although the lack, or aberrant distribution, of H1 variants may not affect the processes of chromatin condensation and dynamics necessary to maintain and segregate chromosomes at mitosis, it obviously might influence the recognition and pairing of similar chromosomes in meiosis, based on interaction of specifically positioned heterochromatin regions. Intriguingly, H1 and proteins of the HMG1 and HMG2 groups promote pairing between homologous (but not heterologous) DNA fragments containing the tandemly repeated sequences often found in heterochromatic regions (Mishima et al. 1997 ). Results of the sedimentation analysis confirm that the H1A- and H1B-deficient chromatin has a more extended conformation than does normal chromatin (Figure 3).
The fact that the marked change in the proportions of H1 variants had a weak effect on most of the observable phenotypic traits except the course of male meiosis and the subsequent development of flower organs raises the possibility that the native distribution of linker histone variants in chromatin is critical for proper behavior of chromosomes during meiosis. The fact that female fertility in the H1A- and H1B-deficient plants was considerably less affected implies that the defects of meiotic chromosomes can be less harmful for female than for male meiosis. If this is so, it would be reminiscent of the situation in Drosophila in which sites for achiasmatic pairing in both sex chromosomes and autosomes are heterochromatic in females but euchromatic (except for rDNA) in males (McKee 1996 ). It also would agree with the finding in many animals (including humans) that during female meiosis (but not during male meiosis) there is no chromosome misalignment checkpoint. As a result, female meiosis proceeds with any number of univalent chromosomes or disorganization of the meiotic spindle (McKim and Hawley 1995 ). Also, in many animals, female and male germ cells differ considerably in the type and content of linker histones (Clarke et al. 1992 ). Whereas data show that in tobacco a typical somatic H1 occurs in the pollen nuclei (Oakeley et al. 1997 ), no data are available to confirm its presence in the oocytes.
An alternative explanation of the different effects of the H1A and H1B deficiency on male and female germ cells is based on the fact that in angiosperms, the development of pollen is thought to be associated with subtle changes in the composition of linker histones between generative and vegetative nuclei (Tanaka et al. 1998 ). Thus, the cause underlying the phenotype of the H1A- and H1B-deficient plants could be a defect in pollen differentiation immediately after meiosis, a result of the distorted proportions of linker histones. In that case, the defects occurring at the stage of meiosis, although clearly documented, would be less important for the overall phenotype.
The results of this work do not rule out the role of H1 variants in gross regulation of transcriptional activity, for example, through influencing the extent of chromatin compactness. On the contrary, the finding that the so-called drought-inducible variants of H1 undergo characteristic changes during flower bud development (Figure 5) points to an intriguing possibility—the involvement of these variants in the specific regulatory changes that occur during development. However, our results show that the predominant phenotypic dysfunction resulting from the change of the physiological proportions of linker histone variants in chromatin is linked to structural aberrations of the meiotic (but not mitotic) chromosomes and to the subsequent failure of functional pollen grains to develop. This observation can be important in determining the primary function of histone H1 in plants and other organisms.
|
|
METHODS |
|---|
Generation of Tobacco Plants Deficient in Histone H1A and H1B
Tobacco histone H1 cDNA (1104 bp) was derived from the pBluescript SK+ (Stratagene, La Jolla, CA) plasmid-bearing H1c12 clone (Szekeres et al. 1995 ) kindly provided by F. Nagy (Institute of Plant Biology, Hungarian Academy of Sciences, Szeged, Hungary). The BamHI-KpnI fragment containing the H1 cDNA was inserted into pRokF19 transformation vector described earlier (Prymakowska-Bosak et al. 1996 ). Localization of the restriction sites in the polylinker assured that the gene was inserted in the 3' to 5' orientation with respect to the promoter. The final construct was introduced into whole 2-week-old seedlings of tobacco (Nicotiana tabacum cv SR1) as described earlier (Prymakowska-Bosak et al. 1996 ) by using the Agrobacterium tumefaciens–mediated procedure. Whole plants were grown directly from the transformed kanamycin-resistant tissue by a standard regeneration procedure.
In Vitro Transcription and Translation of Tobacco Histone H1 cDNA
Tobacco H1 cDNA in pBluescript SK+ was transcribed by using T3 RNA polymerase (Life Technologies, Rockville, MD). The mRNA was translated in a rabbit reticulocyte lysate (Boehringer Mannheim) in the presence of 35S-methionine (Amersham Pharmacia Biotech, Uppsala, Sweden). Aliquots of the 35S-methionine–labeled translation products were mixed with unlabeled total tobacco histone H1, separated electrophoretically, and detected by autoradiography.
Isolation and Analysis of Chromatin
Nuclei and chromatin were isolated from tobacco tissues as described previously (Moehs et al. 1988 ). Chromatin was analyzed after digestion of nuclei isolated from leaf or flower tissues with micrococcal nuclease as previously described (Prymakowska-Bosak et al. 1996 ). To obtain soluble chromatin for sedimentation analysis, we centrifuged the digestion mixture to remove nuclear debris.
Protein Isolation and Characterization Procedures, Gel Electrophoresis, and Protein Gel Blot Analysis
The total tobacco chromatin histones preparations were obtained by extraction with 0.2 M H2SO4 as described previously (Moehs et al. 1988 ). Linker histones were prepared by extracting the total tissue, isolated nuclei, or chromatin with 5% perchloric acid (PCA) and analyzed by using acetic acid–urea 15% PAGE, as described earlier (Prymakowska-Bosak et al. 1996 ). Highly sensitive amino acid analysis of proteins extracted after SDS-PAGE was performed at Eurosequence (Groningen, The Netherlands). For microsequencing, proteins were digested in the gels by LycC protease. After protein cleavage, peptides were separated by reverse-phase HPLC on an octadecyl (C18) silica gel column. Masses and purity of the peptides were checked by mass spectrometry. Selected peptides were sequenced by standard Edman degradation and using an ABI 477 (Perkin Elmer Biosystems, Foster City, CA) sequencer. Microsequencing analysis was performed at the Protein and Nucleic Acid facility in the Beckman Center, Stanford University Medical Center (Palo Alto, CA). Detection with antibodies of proteins separated on gels was as described earlier (Prymakowska-Bosak et al. 1996 ). Quantitative analyses were performed with the ImageQuant program (Molecular Dynamics, Sunnyvale, CA).
Sucrose Density Gradient Centrifugation
A linear (43 to 10%) sucrose density gradient in 1 mM Tris-HCl, 0.2 mM EDTA, pH 7.6, was prepared in 12-mL tubes. The supernatant (0.5 mL) obtained after pelleting the nuclei digested with micrococcal nuclease was layered on top of 10.5 mL of the gradient, and the gradient was ultracentrifuged at 40,000 rpm (197,000g) for 22.5 hr in a Beckman L-60 ultracentrifuge equipped with an SW-41 rotor at 4°C. The density of the 0.5-mL fractions collected was determined by using a refractometer. The DNA in the fractions was precipitated with isopropanol, washed with 70% ethanol, and analyzed on 1.5% agarose gels.
Isolation of Nucleic Acids and RNA Gel Blot Analysis
Plant DNA and RNA were isolated and RNA gel blot analyses were performed as described earlier (Prymakowska-Bosak et al. 1996 ). Tobacco Ta29- and Tac25-specific probes were obtained by polymerase chain reaction (PCR) amplification from genomic DNA of ~300-bp fragments of the coding regions by using the appropriate primers. The fragments obtained by PCR were ligated into the EcoRI and BamHI sites of the pBluescript SK+ plasmid and sequenced on both strands by the dideoxy method with use of an automated laser fluorescent DNA sequencer (Pharmacia, Uppsala, Sweden). The plasmids containing cloned fragments of the Ta29 (Goldberg et al. 1993 ) and Tac25 (Thangavelu et al. 1993 ) genes and the plasmid containing the tobacco H1B gene in reverse orientation were used as templates for generating 32P-labeled or digoxigenin-labeled (for tobacco H1B gene) antisense riboprobes with the aid of T7 RNA polymerase (Promega). The probes for tobacco Nap3 and Ubp1 genes consisted of the entire coding regions cut from the plasmids kindly provided by V. Irish (Yale University, New Haven, CT) and W. Filipowicz (Friedrich Miescher Institute, Basel, Switzerland), respectively. The probe for the tobacco Pbf46 gene was a 1300-bp partial cDNA cut from the plasmid kindly provided by N. Chaubet (Institut de Biologie Moleculaire des Plantes, Centre National de la Recherche Scientifique, Strasbourg, France). For hybridization, 
–32P-dATP–labeled Nap3, Ubp1, and Pbf46 probes were generated by random priming (Fermentas, Vilnius, Lithuania), according to the manufacturer's instructions.
Microscopic Analyses
For examination of microsporogenesis and gametogenesis by light microscopy, anthers isolated from flower buds at different stages of development were fixed in 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2, for 2 hr and then were fixed in 1% OsO4 for 12 hr, dehydrated in a graded ethanol series, and embedded in an Epon/Spurr mixture (Serva, Heidelberg, Germany). Semithin (1- to 2-µm-thick) sections were cut with a microtome (LKB, Bromma, Sweden), stained with 0.1% toluidine blue, and analyzed with a Nikon (Tokyo, Japan) microscope. To examine meiosis, we stained squashed preparations of isolated anthers with orcein and analyzed the results with a Nikon microscope. For transmission electron microscopy, ultrathin (60- to 80-nm-thick) sections cut from Epon/Spurr–embedded anthers were stained with uranyl acetate and lead citrate, as described previously (Prymakowska-Bosak et al. 1996 ), and were viewed with a transmission electron microscope (model JEM-1200Ex; JEOL Ltd., Tokyo, Japan). For scanning electron microscopy of mature pollen grains, the material was fixed in 2% glutaraldehyde and dehydrated in a graded acetone series. The specimens were critical point–dried by using liquid CO2, coated with gold in a sputter-coater (model JEE-4B; JEOL), and examined with a transmission and scanning electron microscope (model JEM-1200Ex; JEOL).
|
|
FOOTNOTES |
|---|
1 Current address: Laboratory of Molecular Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892. 
|
|
ACKNOWLEDGMENTS |
|---|
We are grateful to Ferenc Nagy, Vivian Irish, Nicole Chaubet, and Witold Filipowicz for making available the cDNA clones (as indicated in Methods) and to Steven Spiker (North Carolina State University, Raleigh) and Klaus Grasser (Freiburg University, Germany) for generously providing the antibodies. This research was supported by Howard Hughes Medical Institute Grant No. 75195-543403 (A.J.), Polish Committee of Scientific Research Grant No. 6PO4A 02913 (A.J.), and in part by a grant from the Polish–French Plant Biotechnology Center.
Received July 26, 1999; accepted September 20, 1999.
|
|
REFERENCES |
|---|
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zheng, Z., Miller, W., and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402[Abstract/Free Full Text].
Ascenzi, R., and Gantt, J.S. (1997) A drought-stress-inducible histone gene in Arabidopsis thaliana is a member of a distinct class of a plant linker histone variants. Plant Mol. Biol. 34:629-641[CrossRef][ISI][Medline].
Bouvet, P., Dimitrov, S., and Wolffe, A.P. (1994) Specific regulation of Xenopus chromosomal 5S rRNA gene transcription in vivo by histone H1. Genes Dev. 8:1147-1159[Abstract/Free Full Text].
Cairns, B.R. (1998) Chromatin remodeling machines: Similar motors, ulterior motives. Trends Biochem. Sci. 23:20-25[CrossRef][ISI][Medline].
Clarke, H.J., Oblin, C., and Bustin, M. (1992) Developmental regulation of chromatin composition during mouse embryogenesis: Somatic histone H1 is first detectable at the 4-cell stage. Development 115:791-799[Abstract].
Cole, R.D. (1989) Purification and analysis of H1 histones. Methods Enzymol. 170:524-532[Medline].
Derenburg, A.F., Sedat, J.W., and Hawley, R.S. (1996) Direct evidence of a role for heterochromatin in meiotic chromosome segregation. Cell 86:136-146.
Gantt, J.S., and Lenvik, T.R. (1991) Arabidopsis thaliana H1 histones. Analysis of two members of a small gene family. Eur. J. Biochem. 202:1029-1039[Medline].
Goldberg, R.B., Beals, T.P., and Sanders, P.M. (1993) Anther development: Basic principles and practical applications. Plant Cell 5:1217-1229[Free Full Text].
Hill, T.A., Day, C.D., Zondlo, S.C., Thackeray, A.G., and Irish, V.F. (1998) Discrete spatial and temporal cis-acting elements regulate transcription of the Arabidopsis floral homeotic gene APETALA3.. Development 125:1711-1721[Abstract].
Karpen, G.H., Le, M.-H., and Le, H. (1996) Centric heterochromatin and the efficiency of achiasmate disjunction in Drosophila female meiosis. Science 273:118-122[Abstract].
Koltunov, A.M., Truettner, J., Cox, K.H., Wallroth, M., and Goldberg, R.B. (1990) Different temporal and spatial gene expression patterns occur during anther development. Plant Cell 2:1201-1224[Abstract/Free Full Text].
McCormick, S. (1993) Male gametophyte development. Plant Cell 5:1265-1275[Free Full Text].
McKee, B.D. (1996) The licence to pair: Identification of meiotic pairing sites in Drosophila.. Chromosoma 105:135-141[ISI][Medline].
McKim, K.S., and Hawley, R.S. (1995) Chromosomal control of meiotic cell division. Science 270:1595-1601[Abstract/Free Full Text].
Mishima, Y., Kaizu, H., and Kominami, R. (1997) Pairing of DNA fragments containing (CGA:TCC)n repeats and promotion by high mobility group protein 1 and histone H1. J. Biol. Chem. 272:26578-26584[Abstract/Free Full Text].
Moehs, C.P., McElwain, E.F., and Spiker, S. (1988) Chromosomal proteins of Arabidopsis thaliana.. Plant Mol. Biol. 11:507-515[CrossRef][ISI].
Newrock, K.M., Alfageme, C.R., Nardi, C.V., and Cohen, L.H. (1977) Histone changes during chromatin remodeling in embryogenesis. Cold Spring Harb. Symp. Quant. Biol. 42:421-431[ISI].
Oakeley, E.J., Podesta, A., and Jost, J.-P. (1997) Developmental changes in DNA methylation of the two tobacco pollen nuclei during maturation. Proc. Natl. Acad. Sci. USA 94:11721-11725[Abstract/Free Full Text].
Peirson, B.N., Bowling, S.E., and Makaroff, C.A. (1997) A defect in synapsis causes male sterility in a T-DNA–tagged Arabidopsis thaliana mutant. Plant J. 11:659-669[CrossRef][ISI][Medline].
Prymakowska-Bosak, M., Przewloka, M.R., Iwkiewicz, J., Egierszdorff, S., Kuras, M., Chaubet, N., Gigot, C., Spiker, S., and Jerzmanowski, A. (1996) Histone H1 overexpressed to a high level in tobacco affects certain developmental programs but has limited effect on basal cellular functions. Proc. Natl. Acad. Sci. USA 93:10250-10255[Abstract/Free Full Text].
Renauld, H., and Gasser, S.M. (1997) Heterochromatin: A meiotic matchmaker? Trends Cell Biol. 7:201-205.
Shen, X., and Gorovsky, M.A. (1996) Linker histone H1 regulates specific gene expression but not global transcription in vivo. Cell 86:475-483[CrossRef][ISI][Medline].
Steinbach, O.C., Wolffe, A.P., and Rupp, R.A.W. (1997) Somatic linker histones cause loss of mesodermal competence in Xenopus.. Nature 389:395-399[CrossRef][Medline].
Szekeres, M., Haizel, T., Adam, E., and Nagy, F. (1995) Molecular characterization and expression of a tobacco histone H1 cDNA. Plant Mol. Biol. 27:597-605[Medline].
Tanaka, I., Ono, K., and Fukuda, T. (1998) The developmental fate of angiosperm pollen is associated with a preferential decrease in the level of histone H1 in the vegetative nucleus. Planta 206:561-569[CrossRef].
Thangavelu, M., Belostotsky, D., Bevan, M.W., Flavell, R.B., Rogers, H.J., and Lonsdale, D.M. (1993) Partial characterization of the Nicotiana tabacum actin gene family: Evidence for pollen-specific expression of one of the gene family members. Mol. Gen. Genet. 240:290-295[CrossRef][ISI][Medline].
Tomaszewski, R., and Jerzmanowski, A. (1997) The AT-rich flanks of the oocyte-type 5S RNA gene of Xenopus laevis act as a strong local signal for histone H1-mediated chromatin reorganization in vitro.. Nucleic Acids Res. 25:458-465[Abstract/Free Full Text].
van Holde, K., and Zlatanova, J. (1996) What determines the folding of the chromatin fiber? Proc. Natl. Acad. Sci. USA 93:10548-10555[Abstract/Free Full Text].
Wei, T., and O'Connell, M.A. (1996) Structure and characterization of a putative drought-inducible H1 histone gene. Plant Mol. Biol. 30:255-268[CrossRef][ISI][Medline].
ABSTRACT
INTRODUCTION
