1389-2029/02 $35.00+.00 © 2002 Bentham Science Publishers Ltd.
Nucleolar Dominance: A ‘David and Goliath’ Chromatin Imprinting
Process
W. Viegas
1,*, N. Neves
1,2, M. Silva
1, A. Caperta
1,2, and L. Morais-Cecílio
1 1Centro de Botânica Aplicada à Agricultura, Secção de Genética/DBEB, Instituto Superior de Agronomia, 1349-017 Lisboa, 2Departamento de Ciências Biológicas e Naturais, Universidade Lusófona de Humanidades e Tecnologias, Campo Grande 376, 1749-042 Lisboa, Portugal
Abstract: Nucleolar dominance is an enigma. The puzzle of differential amphiplasty has remained unresolved since it was first recognised and described in Crepis hybrids by Navashin in 1934. Here we review the body of knowledge that has grown out of the many models that have tried to find the genetic basis for differential rRNA gene expression in hybrids, and present a new interpretation. We propose and discuss a chromatin imprinting model which re-interprets differential amphiplasty in terms of two genomes of differing size occupying a common space within the nucleus, and with heterochromatin as a key player in the scenario. Difference in size between two parental genomes induces an inherited epigenetic mark in the hybrid that allows patterns of chromatin organization to have positional effects on the neighbouring domains. This chromatin imprinting model can be also used to explain complex genomic interactions which transcend nucleolar dominance and which can account for the overall characteristics of hybrids. Gene expression in hybrids, relative to parentage, is seen as being based on the nuclear location of the sequences concerned within their genomic environment, and where the presence of particular repetitive DNA sequences are ‘sensed’, and render silent the adjacent information.
THE EARLY DAYS
A differential gene expression phenomenon that results in nucleolus formation by only one parental set of rRNA gene clusters is widely found in hybrids. The process was first observed about 50 years ago by Navashin [1], after crossing several different species from the genus Crepis, and later termed ‘Nucleolar Dominance’ [2]. Navashin clearly demonstrated that after hybridization the chromosomes from parental species undergo morphological modifications in the hybrid. Changes that affected all of the chromosomes from one parent were termed “amphiplasty” or “neutral amphiplasty”, whereas those that targeted only particular chromosomes he named “differential amphiplasty”. This latter case resulted from Navashin’s observations on a particular chromosome group of the genus Crepis - the D chromosomes - that usually display a distal small segment (the satellite) connected to the remainder of the chromosome by a thin strand of chromatin (the secondary constriction) at metaphase. In interspecific hybrids, only one D-chromosome showed this characteristic morphology, while the other failed to display its secondary constriction. Furthermore this differential behaviour always belonged to the chromosome from the same parental species in reciprocal crosses (Fig. 1). In backcrosses of the hybrid to its appropriate under-dominant parental species, Navashin detected the restoration of the secondary constriction and the satellite structures, thereby showing that these changes are reversible and not due to any permanent loss and/or damage of these chromatin regions. In parallel, Heitz [3] proved that the satellite and
*Address correspondence to this author at the Centro de Botânica Aplicada à Agricultura, Secção de Genética/DBEB, Instituto Superior de Agronomia, Tapada da Ajuda, 1349-017 Lisboa, Portugal; Tel: +351 21 365 32 81; Fax: +351 21 363 50 31; E-mail: [email protected]
the secondary constrictions observed in metaphase cells are clearly associated with the formation of nucleoli during interphase, since nucleoli were shown to form at or near these constrictions.
Another important contribution to the understanding of this phenomenon came from the work of Barbara McClintock. Linking her own results on differential rRNA gene expression patterns in maize with Navashin’s results, McClintock reasoned that species can be arranged in a hierarchy of dominance in their capacity to produce nucleoli, in hybrids (e.g. Crepis) or in chromosome re-arranged lines (e.g. maize). Moreover, McClintock was the first geneticist to use the term “nucleolar organiser” (or Nucleolar Organiser Region, NOR) to describe the chromosomal loci at which nucleoli form [4]. In maize chromosome six the NOR includes a centromere-proximal heterochromatic knob (a large chromomere) followed by the secondary constriction and a distal satellite region. By analysing nucleolus forma-tion in wild-type and translocaforma-tion lines, McClintock suggested that the NOR must bear redundant genetic infor-mation [4]. Later it was shown that the NORs correspond to clusters of multiple copies of rRNA genes [5, 6].
At this point, it became clear that nucleolar dominance was a process observed in hybrids which leads to nucleolus formation mediated by the NORs of only one parental species. Discrimination between transcriptional and post-transcriptional levels of regulation of NOR activity was addressed shortly afterwards through studies in the frog
Xenopus by Honjo and Reeder [2]. They used a biochemical
approach to determine the presence and origin of rRNA transcripts during the embryogenesis of X. laevis x X.
borealis hybrids; and showed that the only nucleoli present
originated from the NORs and transcripts of X. laevis, indicating that the process occurs at the level of gene transcription.
The next logical steps were straight-forward: (i) to determine if the process results from selective activation of rRNA genes or from silencing of the under-dominant rDNA; (ii) to find out whether the process is achieved ‘gene-by-gene” or through a single large-scale NOR-targeted mechanism; and (iii), to evaluate the molecular mechanisms involved in establishing dominance and enforcing the process throughout cell cycle. Although considerable efforts have been made to answer these fundamental questions, several possibilities remain, and no general rule has yet been discovered. As we will see, the most likely answer seems to be firmly based on the original amphiplasty phenomenon first described by Navashin.
MECHANISMS FOR ESTABLISHMENT AND ENFORCEMENT: RULES AND EXCEPTIONS
The Enhancer Imbalance Model
The first clue about how nucleolar dominance operates came from the early studies of McClintock [4], when she launched the enhancer imbalance hypothesis. In maize translocation lines under-dominant DNA loci in somatic cells produce large (normal) nucleoli when they are the only ones present (as in haploid pollen grains), and McClintock suggested that this may imply a competition for a factor (“positive substance”) essential for nucleolus formation that is present in limiting amounts, more efficiently tritated by the dominant rRNA genes. This popular hypothesis was settled by molecular studies using Xenopus as a model. In the hybrid X. laevis x X. borealis, the dominant X. laevis rRNA genes have larger intergenic spacer sequences upstream of the gene promoter and are composed of higher number of
repetitive DNA elements [7]. These repeat-DNA spacer sequences were proven to act as pol I transcription enhancers, since transcription was stimulated strongly when they were cloned and located adjacent to a ribosomal gene promoter, and injected into oocytes or young embryos [8, 9]. Nevertheless, when these same sequences were co-injected with a second plasmid transporting a promoter, transcription was severely inhibited by the original promoter [9], suggesting that enhancers bind to only one transcription factor that interacts with the gene promoter. The importance of the DNA repetitive sequences within the rDNA intergenic spacer was further supported by the observation that mini-genes, with a X. laevis promoter attached to X. laevis spacer sequences, are preferentially transcribed over identical constructs with a borealis rRNA gene spacer [10]. These and other findings prompted the hypothesis that differences in sequence or number of regulatory elements in the intergenic spacers create the conditions to sequester transcription factors in limiting concentration.
Further support favouring this model was gained by the analysis of nucleolar dominance in bread wheat. Bread wheat (Triticum aestivum) is an allohexaploid originating from three different ancestral diploid species with multiple NOR loci [11]. Of these, the NORs located on chromosomes 1B and 6B are responsible for 90% of the total nucleolar volume in wheat cells, and hence are called the wheat major NORs [12]. Several studies indicate that nucleolar volumes (as a sensitive measure of nucleolar activity) have more to do with the number of active rRNA genes at a NOR than with the total number of ribosomal genes per NOR locus [12-15]. The 1B loci account for 60% of nucleolar mass followed by 6B NORs (30%) and wheat minor NORs (10%); however, 1B NORs bear half the number of rRNA genes present in 6B loci [16, 17]. These authors clearly demonstrated not only that rRNA gene number is not the primary determinant responsible for NOR activity, but also that 1B NOR units have longer intergenic spacers than the 6B rDNA units [14, Fig. (1). Diagram representing one pair of chromosomes carrying NORs (secondary constrictions) and terminal satellites in two parental species and their F1 hybrid, as first described in Crepis spp by Navashin [1]. The dominant species retains its visible NOR in the hybrid, and the rRNA gene cluster in the under-dominant species becomes ‘silenced’. Silencing is accompanied by condensed chromatin and the NOR is no longer seen. In backcrosses of the hybrid to its appropriate under-dominant parental species the secondary constriction and the satellite structures are restored showing that these changes are epigenetic and reversible.
17]. These results suggested a process similar to that enunciated for the Xenopus model, where the largest intergenic spacers are the more effective in attracting factors in limiting concentration which are essential for rRNA gene transcription [16, 17]. Although this case cannot be accommodated within a strict view of nucleolar dominance (as both NORs belong to the same ancestral B genome donator), the enhancer imbalance model still seems plausible in interspecific genomic interactions in cereal hybrids. In this case, a preferential suppression of wheat rDNA loci occurs in plants resulting from the cross of Aegilops umbellulata (a wheat wild relative) and wheat [15]. These results can be explained under this hypothesis, as A. umbellulata rDNA units bear longer spacers than the largest wheat rRNA gene spacers [16, 17].
Nucleolar dominance was also observed in wheat x rye hybrids, and in the artificial amphiploid triticale. In this case, wheat dominates over rye in the expression of rDNA loci, resulting in an almost total inactivation of rRNA genes of rye origin (Fig. 2); [18-22]. Comparison of rDNA spacers from wheat and rye showed that rye spacers are shorter, displaying several deletions amongst the repetitive DNA sequences [23]. Once more it is tempting to reason that rye rRNA genes are less effective in attracting transcription factors leading to the preferential expression of wheat NORs.
As previously pointed out [24, 25], the attractiveness of the enhancer imbalance hypothesis resides in its simplicity: rRNA genes from distinct genomic origins being differentially expressed on a biochemical basis. As presented, this mechanism could clearly explain the establishment of McClintock’s hierarchical NOR expression relationships, regardless of maternal or paternal effects and
independent of the number of rRNA genes located at a NOR. However, this simple and apparently effective rule cannot be extended to other systems, namely Brassica and
Arabidopsis. By detecting the genomic origin of rRNA
transcripts in Brassica hybrids it was possible to ascribe a nucleolar dominance relationship between several Brassica species, where the sequence is B. nigra > B. rapa > B.
oleracea [26]. Sequencing of intergenic rDNA spacers from
the three species allowed a direct comparison of the length of the repetitive regions and the number of repetitive elements. Unexpectedly, B. oleracea has the longest intergenic rRNA gene spacers, yet these genes are inactivated in the presence of B. napus or B. nigra NORs. Another clear-cut exception to the enhancer imbalance model has been reported in
Arabidopsis, based on ploidy modifications. In A. suecica,
the natural amphiploid between A. thaliana and A. arenosa , the A. arenosa NORs are dominant, and the A. thaliana rRNA genes are transcriptionally inactive [27]; even though the A. arenosa spacer is shorter than that of A. thaliana. Furthermore, this dominance can be altered following genome ratio modifications. By appropriate backcrosses, where recombinant lines were produced with a 3:1 A.
thaliana:A. arenosa ratio, the A. arenosa rDNA becomes
transcriptionally silent. This intriguing result also argues against the hypothesis of straightforward competition between rDNA intergenic spacers for limiting transcription factors: if less dominant rRNA genes are present, then more transcription factors may become available to allow under-dominant rDNA transcription, presumably leading to a co-dominance situation. However, the observation of silencing of the normally dominant genes (dominance reversal) cannot be predicted or explained by the enhancer imbalance hypothesis [24, 25, 28].
Fig. (2). Nucleolar dominance in triticale (wheat-rye amphiploid), as shown by sequential silver staining and in situ hybridization techniques. The silver staining procedure (Ag-staining) reveals only the metaphase rDNA loci that were transcribed during the preceeding interphase, and DNA:DNA in situ hybridization reveals DNA sequences in the genome regardless of their expression patterns. Most triticale metaphase cells show only two pairs of silver-stained NORs (2a, black arrows), and in situ detection of the rye chromosomes (2b, in red) and rDNA sites (2b, in green) reveals the two pairs of NORs previously detected by Ag-staining on the wheat genome and two additional rDNA loci on rye chromosomes that were not detected by Ag-staining (2b,white arrows).
The wheat-rye system, already referred to as an example supporting the enhancer imbalance hypothesis, can also be recruited to disprove the simple involvement of rDNA intergenic size in attracting positive transcription factors, or other factors that prevent binding of silencing mediators. In wheat addition lines for the rye nucleolar chromosomes, rye NORs are expressed (Fig. 3); if not less, then at least the same amount of transcription factors is present as in triticale (the artificial wheat-rye amphiploid), and both plant materials have the same complement of wheat chromosomes [29]. Even higher levels of rye NOR expression are moreover observed in wheat lines with the translocation of the short arm of rye nucleolar chromosomes (Fig. 4). If the nucleolar dominance process was based only in the enhancer imbalance hypothesis then rye NORs should remain silent in the addition and translocation wheat lines. It seems clear therefore that the process of nucleolar dominance must be established by other mechanisms.
Fig. (3). Rye NORs are usually silent in wheat x rye hybrids and triticale, but can be reactivated in other wheat-rye genomic combinations. In wheat lines carrying rye nucleolar chromosomes as the only source of alien chromatin, wheat and rye rDNA loci are expressed in the majority of cells. Hence, in most metaphase cells, two pairs of rRNA gene clusters from wheat-origin ( ) and one pair of rye-origin rDNA loci ( ) are revealed by silver-staining, indicating their previous transcription.
The Species-Specific Transcription Factor Model
An alternative hypothesis emerged from the fact that although the DNA sequences specifying the 18S, 5.8S and 25/26S rRNAs are highly conserved in higher eukaryotes [30-32], intergenic rDNA sequences have diverged substantially during evolution [23-25]. In parallel, RNA pol I transcription factors would have also co-evolved bringing rRNA gene transcription to a species-specific condition, where the rRNA genes of one species could only be transcribed if the appropriate species-specific transcription factors were present. Evidence for this hypothesis came from the use of a cell-free transcription system in mouse-human hybrid cell lines. It was shown that a mouse extract can be re-programmed to promote a human rRNA gene transcription event adding a specific human transcription factor [33-39]. Therefore, the absence of an adequate
(species-specific) human transcription factor could be the explanation for the dominance of mouse rDNA over human rRNA genes. In plants several results account to this hypothesis: a tomato rRNA gene promoter is not recognized when transfected into Arabidopsis protoplasts [40], and the same happens when a tobacco promoter is introduced into a bean cell-free transcription extract [41].
As pointed out by Reeder [42] the results achieved in evolutionary wide crosses may not be relevant in hybrids formed by normal reproductive processes. Several observations support this view, tending to rule out this hypothesis as the major mechanism responsible for establishing nucleolar dominance. Clear examples of this are found in interspecific hybrids in Brassica and Arabidopsis. In Brassica species, rRNA gene promoters were found to be functional across species limits in transient expression assays [43], although these species can be arranged in a nucleolar dominance relationship [26, 44]. Moreover, in these species both dominant and under-dominant rRNA genes were show to be equally transcribed in vivo after transfection into protoplasts of either the dominant or amphiploid species [45]. It seems clear that dominant and under-dominant rRNA genes of closely related species or genera may use the same transcription factors when together in the same nucleus [24, 25, 27]. In fact, an Arabidopsis rRNA gene promoter was found to be active in a cell-free system of a Brassica species [46]. Differential rRNA gene expression was also observed within a species, as in maize translocation and recombinant lines [4, 47], and between homologous rDNA loci in diploid rye [48]. Thus, the discrimination between active and silenced rDNA must be attributable to other mechanisms.
The Chromatin Modification Model: DNA Methylation and Histone Acetylation
During the last ten years another model for nucleolar dominance has been proposed based on chromatin modification related mechanisms, such as DNA methylation and histone acetylation which are responsible for alternative chromatin states of transcriptionally inactive and active rRNA gene loci. Nucleolar dominance, as with other epigenetic phenomena (such as the X chromosome inactivation in mammalian females, gametic imprinting, paramutation events, homology-dependent gene-silencing, and transposon activity), has been shown to be modulated by DNA methylation [49]. Several reports in cereals, such as hexaploid wheat, have shown that transcribed NOR loci are less methylated than ones that are untranscribed [16, 17]. In wheat x rye F1 hybrids, and in triticale, the under-dominant rye-origin NORs are more methylated within the rDNA intergenic spacer sequences [50, 51]. In addition, after treatment with the hypomethylating agent 5-aza-2’-deoxycytosine (5-AC) [52], strong reactivation of rye NORs in the wheat-rye system was observed [50, 53]. The same result was reported for under-dominant rRNA gene sets in
Arabidopsis [27] and Brassica [44]. The direct involvement
of methylation in the nucleolar dominance process has been primarily inferred from other studies that show a relation between methylated bases and point changes in DNA topology that either prevent the binding of transcription
activators or allow the binding of methyl-chromoproteins that inhibits gene transcription.
However, the association of DNA methylation patterns with broader differential ribosomal gene expression processes has several unclear and some clear but contradictory aspects in its general appliance. Firstly, it is difficult to reason how DNA methylation may account for early discrimination between rRNA gene sets in hybrids, leading to the preferential silencing of one parental set. In this context, this phenomenon could be due to differences in rDNA intergenic length and/or spacer sequences, in terms of competition for either methylation preventing molecules (that would preferentially attach to the dominant NORs) or pro-methylating agents (that would selectively bind to the under-dominant rDNA loci). However, mechanisms that are only based in the intergenic rRNA gene spacers seem unlikely to account for such a discriminatory process, as no clear rule concerning size and/or type of intergenic sequences seem applicable. Aside from these theoretical considerations, the experimental results are not in agreement with the general validity of this relationship. For example, in
Brassica hybrids, rRNA transcripts from normally
under-dominant rDNA loci in untreated plants are clearly detected following seed germination in the presence of 5-AC [44]; and both normal and methylated under-dominant rRNA minigenes were shown to have equal abilities for in vitro transcription [45]. Several hypothesis can be discussed, such as: (i) minor changes of methylation density in particular sequences at the rDNA, (ii) other (regulatory) loci may be responsible for the activation of previously silent ribosomal genes [45], or (iii) methylation might help in establishing a repressive chromatin state preventing gene promoter accessibility to the transcription machinery [24, 25]. Another intriguing result came from studies in Xenopus hybrids where there was no apparent difference in the methylation state of dominant and under-dominant rRNA genes [54]. In the wheat-rye system, it was already shown that rye rRNA
genes, usually silent in the amphiploid triticale [21, 22, 53, 55], are more heavily methylated [50, 51]. However, in the wheat addition line for the rye nucleolar chromosomes a significant increase in rye rRNA gene expression is observed [29], the rye rDNA being methylated at levels equivalent to the ones observed in triticale where the same genes are highly suppressed [51]. Furthermore, where nucleolar dominance has been described in interspecific hybrids in
Drosophila [56], only very diminutive amounts of
methyla-ted cytosine residues have been recently discovered in these species, showing no particular distribution patterns through-out the genome, nor any variation during development [57].
Another mechanism usually related to chromatin alterations and gene activity is histone acetylation [58]. To our knowledge, there are few studies proving the involvement of histone acetylation in nucleolar dominance processes, and only in Brassica hybrids it was proven that treatments with hyper-acetylating agents cause activation of normally non-transcribed rRNA genes [45]. However, the direct relation found between histone acetylation at the gene level and transcriptional activity mediated by RNA pol II and III [59, 60] may not extend to all rRNA genes. In the field bean (Vicia faba) it seems that the NOR transcribed by the RNA pol I complex does not follow this general rule [61]. In this species, as in barley [62], direct immunolabelling of acetylated and non-acetylated histones showed that the overall acetylation patterns of the rDNA are not strictly correlated with their transcriptional activity. These patterns are inversely correlated with their replication patterns, as the strongest histone acetylation at the NOR occurs during mitosis when the rRNA genes are being neither transcribed nor replicated. The absence of both acetylated and non-acetylated histone forms within the nucleolus may be related with the fact that transcriptional activation of the rDNA requires disruption of preformed nucleosomes [63]. In fact, active rRNA genes within nucleoli have been shown to be devoid of complete nucleosome core particles [61]. Fig. (4). In the wheat-rye system, rye-origin NORs can be expressed at various levels, according to the wheat-rye genomic combination. Whereas in the wheat-rye amphiploid triticale the rye NORs are clearly under-dominant (revealed by silver staining in only 26% of metaphase cells), their expression increases in the addition line (61%), and increases even more in the wheat-rye translocation lines (88%). These results, obtained in the same system, show that the enhancer-imbalance hypothesis along with the potential need of species-specific transcription factors are not applicable to this case.
Nevertheless, this feature does not seem to be a particular characteristic of active rRNA genes, as transcriptionally silent rDNA copies at condensed perinucleolar knobs are also free of acetylated histones [61]. Interestingly, induced DNA hypomethylation and histone hyperacetylation do not have an additive effect on reactivating usually silent rDNA loci [45]. This suggests that both elements may play a part in chromatin remodelling mechanisms that may be responsible for chromatin topological states that can either promote or prevent transcription [24, 25].
Collectively these models for explaining nucleolar dominance suggest that at the level of chromatin organiza-tion is where most of the establishment and enforcement mechanisms reside. Hence the necessity of finding a broader and more unifying explanation for this process, probably associated with chromatin topology.
A CHROMATIN IMPRINTING MODEL: A RE-INTERPRETATION OF NAVASHIN’S WORK
Dominance and Codominance in Hybrids: States of Chromatin Organization?
We have alluded to, and ruled out, certain contenders for a role in provisioning nucleolar dominance, namely gene promoter sequences, species-specific transcription factors, DNA methylation and histone acetylation. A consideration of these options has led us to the view that DNA methylation and histone acetylation are more likely to be involved in maintaining a transcriptionally silent state for the under-dominant rRNA genes; and that the actual process of discriminating between parental rDNA sets occupying the same nucleus is a matter of understanding the topology of the rDNA chromatin.
Wallace and Langridge [64] stated that the centromere-proximal portion of the NOR chromatin (the chromomere) is composed of rRNA genes that are in a highly condensed state (hence not being transcribed), while the secondary constriction is composed of a subset of rRNA genes that, by virtue of having been previously transcribed remain decondensed even at metaphase. These authors also noted, by analyzing nucleoli formation in maize translocation lines, that genes packaged in the chromomere can be reactivated following translocation events. Furthermore, several studies based on DNase I accessibility showed that the more active rRNA gene loci have more DNase I hypersensitive sites. These sites are also less methylated in wheat and maize [47, 65]; and in Xenopus hybrids dominant rRNA genes are also more sensitive to DNase I treatments, but without detectable changes in DNA methylation [54]. These results also indicate that actively transcribed ribosomal loci are in a more decondensed chromatin state. In situ hybridization with rRNA gene probes showed that rDNA chromatin within the nucleolus appears in general as diffuse chromatin fibers. The non-transcribed fraction of the NOR is visualized as a condensed chromatin block that adopts a perinucleolar position [61, 66-69]. Based on cytogenetic evidences it has also been shown that rDNA decondensation is strictly necessary for rRNA gene transcription [70]. Thus, it seems that the state of organization of rDNA chromatin, and the
transcription of rRNA genes, determine the form and function of NORs. The key to our understanding of nucleolar dominance may therefore reside in the chromatin structure of rRNA gene loci.
We also note that this type of genomic interaction, causing preferential expression of genes according to their parental origin, does not seem to be of general appliance. There are only a few examples of redundant gene silencing in hybrids and polyploids [71-73]. Several studies based on isozyme analysis clearly suggest that most gene loci inherited from two or more progenitor species are codominant [74-76] (Pontes and Viegas, unpublished data). Using AFLP-cDNA to perform a genome screen for orthologous genes silenced in Arabidopsis polyploids it was shown that only 2,5% of genes are subjected to differential expression [77]. Thus, these results indicate that nucleolar dominance may not be part of a broader genome silencing process as a result of a large-scale genomic interaction [78]. One apparent difference between the rRNA genes exhibiting differential amphiplasty and other genes that display codominance is the chromatin structure that underlies the two types of genetic information. While most genes exist in the genome as single sequences in euchromatic domains, rRNA genes are present as large tandem arrays of repetitive rDNA units arranged in clusters that clearly resemble the heterochromatic fraction. In fact, since McClintock’s observation of the chromocentre associated with maize rDNA loci, it was shown that heterochromatin domains are frequently present in NORs [79].
In passing, we may also note that when genomes are challenged by hybridity there are sometimes other responses of a more ‘violent’ nature, involving the exclusion of an entire genome as in haploidisation following wide species crosses [80], or even more significant events where somatic recombination/translocation accompanies the genome exclu-sion [81, 82]. These matters, however, are beyond the scope of this review.
Size Matters: Smaller Genomes have More Penetrance Heterochromatin, a term introduced by Heitz [83], was initially used to describe that part of a genome which is more compact than the euchromatin. This chromatin fraction is composed of highly repetitive DNA sequences which establish close relations with several heterochromatin-specific proteins [79, 84, 85]. Heterochromatin was initially distinguished by cytogenetics analysis, and was described as the fraction of the genome that maintained its overall condensation throughout the cell cycle. However, repetitive DNA sequences are not only present in large/conspicuous heterochromatic clusters. They can be dispersed in the genome, organized into smaller heterochromatic islands which may pass the cytogeneticist analysis, even during nucleus division [86, 87]. In fact, recent results have specified the minimum size/mass for C-band positive staining [88, 89].
Besides being present at the NORs, some protein coding genes are also found in the heterochromatin, as for example in Drosophila [90-92]. But in general heterochromatin
domains are a gene-poor fraction of the genome [93]. Hence, the initial status attributed to heterochromatin classed this fraction as “ignorant”, “selfish” or “junk” DNA [93], a dispensable or parasitic portion of most genomes [94]. John [95] pointed to heterochromatin as an “evolutionary relic”, maintained in the genome only as a “by-product of molecular events”. Lately however, studies on evolutionary genetics along with modern molecular approaches has led to a conceptual shift towards heterochromatin as a source of increasing variability and function [96-99]. It has been shown that this genome fraction can play important roles in the modulation of gene expression, in chromosome structure, in speciation and in evolution (review in [79]). Particularly interesting are the studies that reveal a strong contribution of heterochromatin to the broad range of genome sizes throughout biological taxa, that is to variation in nuclear DNA amounts which are not correlated with organism complexity (review in [79]). Several approaches indicate that the variability in genome size in eukaryotes is consistently achieved through a quantitative variability in copy numbers of repetitive DNA sequences, usually organized in heterochromatic domains [79, 100-102]. Studies in mammals, cryptonomads and Drosophila (amongst others)
prove that differences in heterochromatin contents between closely related species, or even individuals of the same species, may account for several phenotypic effects [103, 104], clearly bringing heterochromatin into the field of functional genomics. This view is strengthened by Brenner’s studies on vertebrate genome evolution, which suggest that even the smallest ones have some repetitive (heterochro-matic) sequences, sufficient for them to function properly. The basic vertebrate genome, may then have accumulated other repetitive sequences during evolution leading to high order speciation [105, 106]. Hence, the generic rule that can be drawn from these findings is that larger genomes usually reflect higher amounts of heterochromatin [79].
To return to differential amphiplasty, and to join up the thinking, we will now argue the case that the differential expression of rRNA loci is related to genome size, as the compilation of data in Table 1, from a number of well established cases, clearly indicates. While various mechanisms have been found as contributing factors for establishing nucleolar dominance, a unifying feature of all the examples given is the difference in DNA amount between the parental species concerned. It seems clear that Table 1. The Relation Between Nucleolar Dominance in Hybrids and Genome Size of Parental Species
Genome size (Mpb/1C)
Hybrid Dominant sp. References
Dominant sp. Under-dominant sp.
Crepis capillaris x C. alpina C. capillaris [1] 2058 [110] 2940 [110] C. capillaris x C. dioscoridis C. capillaris [1] 2058[110] 6003 [110]
C. capillaris x C. neglecta C. capillaris [1] 2058[110] 2842[110]
C. capillaris x C. tectorum C. capillaris [1] 2058[110] 3308 [110]
C. setosa x C. tectorum C. setosa [1] 2279 [110] 3308 [110]
Xenopus leavis x X. borealis X. leavis [2] 3100[111] 3480[111]
D. melanogaster x D. virilis D. melanogaster [56] 116[113] 340[112]
D. melanogaster x D. simulans D. melanogaster [56] 116[113] 116[113]
Triticum aestivum x Aegilops umbellulata A. umbellulata [15] 4949[110] 16979*[110]
Hordeum vulgare x Secale africanum H. vulgare [107] 4873[110] 7277[110]
H. bulbosum x H. vulgare H. vulgare [108] 5500[110] 5500[110]
T. aestivum x Secale cereale T. aestivum [19] 16979*[110] 8110 [110]
T. durum x Secale cereale T. durum [19] 11785*[110] 8110[110]
Aegilops squarrosa x S. cereale Ae. squarrosa [109] 4998[110] 8110[110]
Brassica nigra x B. rapa B. nigra [26] [44] 473[114] 512[114]
B. nigra x B. oleracea B. nigra [26] [44] 473[114] 633[114]
B. rapa x B. oleracea B. rapa [26] [44] 512[114] 633[114]
the under-dominant rRNA genes in the hybrid always reside in the parental species with larger genome (Fig. 5). The exceptions to this rule are the hybrids Drosophila
melanogaster x D. simulans and Hordeum bulbosum x H. vulgare, where the parental species have similar nuclear
DNA contents. Interestingly, where both parental species do have equivalent amounts of DNA, a small amount of sequence elimination occurs after hybridization [115, 116], and this may impose an imbalance in genome size which is sufficient to induce nucleolar dominance in the hybrid. It should also be stressed that the differences in DNA contents listed in the table are much larger than any potential DNA loss observed after hybridization. In hybrids involving polyploids, such as Triticum aestivum x Secale cereale and
Triticum durum x Secale cereale the species with larger
genome dominates over the smaller one. This apparent exception is accounted for if DNA contents are compared per haploid genome.
Completing the circle, by drawing in the relationship previously established between DNA content and the amount of heterochromatin, it is plausible to argue that preferential rDNA transcription in relation to nucleolar dominance in hybrids occurs from the parental genome with the least amount of heterochromatin. Besides the examples presented in Table 1, further support for the case can be gained by looking to other cases of differential rRNA gene expression. For example, in maize translocation lines larger nucleoli form from recombinant NORs displaying the smaller rDNA heterochromatic knobs [4]; and in wheat 1B NORs, with a lower level of associated heterochromatin, organize nucleoli double the size of the 6B rDNA loci [11, 117].
Functional Genomics of Heterochromatin: the Basis for Nucleolar Dominance
To sustain the argument it now has to be reasoned how differential rRNA gene expression is related to
heterochromatin content, and what is the rational underlying the gene silencing process that it mediates. The configuration of the repetitive DNA in heterochromatin, i.e. its dense chromatin packaging, appears to correlate strongly with the potent transcription silencing of this section of the genome (reviews in [79, 118-120]). Several examples in Drosophila and yeast [120] show that by juxtaposing an euchromatic gene with heterochromatin (by translocation or transgene insertion) a mosaic pattern of expression of the gene is observed. This position effect variegation (PEV) is extensive in mammals, where transgene silencing occurs at a high level if the transgene is integrated as a high copy-number array, but normal expression is allowed for low copy number insertions [121]. Moreover, this cis-acting mode for heterochromatin mediated gene silencing is not the only way for its suppressive effects on gene expression, since nuclear compartmentalization and chromosome pairing are also influenced by long-range silencing mechanisms [122, 123]. Both the cis- and trans-inactivation modes were clearly described for different genes in Drosophila [124-126]. Inactivation of gene expression following specific sequence duplication has been also observed in plants [127]. An interesting result was obtained in Arabidopsis hybrids, where parental dominance, examined for 20 candidate genes, was only found at three loci that are composed of repetitive DNA, the same sequences that build up heterochromatin [128]. Other studies in Drosophila also provide evidence that heterochromatic regions can play a role in interspecific nucleolar dominance; such as D. melanogaster NORs (located on chromosomes X and Y) which dominate over the
D. simulans NOR (located on chromosome X) [56].
However, rearrangement of the heterochromatin that flanks
D. melanogaster NORs causes activation of the D. simulans
rDNA locus, as well as allowing normal transcription of the
D. melanogaster rRNA genes [129]. Furthermore, the role of
heterochromatin domains in establishing repressive effects on gene transcription has been clarified by studies proving the nuclear co-localization of the silenced genes and the heterochromatic domains [79, 118]. It is intellectually
Fig. (5). Diagram illustrating the relationship between genome size of parental species and nucleolar dominance in Crepis hybrids. The under-dominant rDNA loci always belong to the larger genome (data taken from Table 1 in the text).
inevitable to propose that the same mechanism underlies nucleolar dominance.
In support of the argument one may look into the parallel behavior of heterochromatin dynamics and rDNA expression patterns throughout development, and examples of the role of heterochromatin in modulating rDNA transcription. In
Drosophila the strength of heterochromatin mediated gene
silencing decreases during development [130]. In addition, C-bands cannot be distinguished in metaphase chromosomes during early embryogenesis, only acquiring this chromatin state of C-band manifestation from the blastoderm onwards [131]. More recently the action of some protein groups has been related to pattern formation during Drosophila embryo-genesis, through the establishment of repressive chromatin states [132, 133]. In plants, the number and size of interphase heterochromatin domains is known to be related to the functional activity of the nucleus and to tissue differentiation [134], and nucleolar activity seems to be correlated with cell type and metabolic rates [135]. Nucleolar dominance also appears to be developmentally regulated: the usually silent under-dominant rRNA genes becoming active in the transition to the floral meristem (e.g. Brassica hybrids, [26]), or following meiosis (e.g. wheat x rye amphiploid, [22]). In wheat, the largest nucleoli are formed in the haploid micro-spore [12]. Interestingly, modification in other epigenetic markers, such as DNA methylation patterns, have also been observed during these reproductive related events [136]. Seventy Years of Nucleolar Dominance: Closing the Circle
Notwithstanding the commentary given above, the fundamental question of how discrimination operates to
recognise and regulate parental gene sets, and to provide for amphiplasty, remains unanswered. Old theories and new results converge and integrate to finally close the circle. Bringing together two entire or partial genomes in the same nucleus can be regarded as adjusting each parental genotype to a common space, the hybrid nucleus. Nuclear volumes of polyploids do not usually reflect the addition of the parental nuclear volumes. By determining the nuclear volumes of meristematic root-tip cells (by confocal microscopy) in rye, wheat and its amphiploid triticale, this lack of additivity can be easily observed: the amphiploid volume is significantly smaller than the added rye and wheat volumes (Silva and Viegas, unpublished results). It is also known that rye and wheat genomes do not intermix throughout the cell cycle and that at interphase both genomes are organized as discrete chromatin domains in triticale, revealing a closer association of parental-origin chromatin [55, 137]. This non-random distribution finds parallel in the heterochromatin disposition patterns, since heterochromatin domains are not randomized throughout the nuclear territory [79]. If a genome has less available nuclear space in the hybrid, relative to its diploid condition, chromatin association is more likely to occur, and to a greater extent in the larger genome, establishing a higher level of het-het interactions (Fig. 6). Considering the co-localization of enriched heterochromatin domains, as observed for the rDNA, telomere and silent-mating type repetitive DNA regions in yeast [135, 138], and the proven capacity of heterochromatin to propagate silent conformations in neighboring chromatin domains, we can begin to develop a model for how parental DNA sets are discriminated and nucleolar dominance established. The under-dominant genes will belong to the largest genome, and hence to the genome enriched in repetitive DNA organized in heterochromatin domains. Intuitively we can now imagine
Fig. (6). Nucleolar dominance: a chromatin imprinting model. Parental genomes are represented as nuclei with large (top left) and with small genomes (top, right). Differences in genome sizes are shown as being due to distinct amounts of repetitive DNA sequences that build up heterochromatin domains. The amphiploid nucleus (below) has a smaller nuclear volume than the sum of the nuclear volumes of the parental species. In this hybrid nucleus, parental chromatin behaves differently in comparison to its behaviour in its own genomic/nuclear environment, where the larger genome adjusts to a smaller volume, triggering a more intimate association between heterochromatic domains. Genes or gene clusters in the under-dominat DNA will undergo a chromatin imprinting process responsible for their silencing, by virtue of being positioned adjacent to repetitive DNA sequences.
how the occupancy of a common space will cause the heterochromatin-rich under-dominant NORs to be silenced, and to assume a quiescent state that is maintained throughout successive cell divisions. Chromatin modeling mechanisms, such as DNA methylation, histone hypoacetylation, and histone methylation (reviewed in [139]) can be enlisted to police this silent mode. This model can also explain why both in Brassica and Arabidopsis hybrids, under-dominant rRNA minigenes (cloned in plasmid vectors) are expressed in hybrid protoplasts whereas the chromosomal copies of the under-dominant rDNA units remain silent in the same protoplasts [27, 45]. Some hitherto unclear results on nucleolar dominance in the wheat-rye system can also be sustained by the current hypothesis. In wheat addition lines of rye nucleolar chromosomes, the larger expression of the rDNA loci of rye origin, relative to their behaviour in the wheat-rye amphiploid [29], can be accredited to a diminution of the rye-heterochromatin content which leads to less severe het-het interactions in the addition line. The observation of high expression of both rye and wheat NORs in substituted triticales, where the 2R rye chromosomes are substituted by the 2D wheat chromosomes [29], can be also related to heterochromatin content. In this case, the missing rye chromosome pair is the larger rye chromosome with large telomeric heterochromatic blocks [140, 141]. Furthermore, in wheat-rye translocation lines where the only rye-origin chromatin present is the short arm of rye nucleolar chromosomes translocated onto wheat chromosomes, rye rDNA is transcribed at even higher levels than in the addition line.
The model we present here is essentially one of ‘position effect’ extended to hybrids which differ in their amounts of nuclear DNA, and it can be used to explain the complex genomic interactions that lead to co-dominance or parental dominance of genes from either genome progenitor in determining the overall characteristics in these hybrids. Arguing that large-scale chromatin re-organization processes of parental genomes may take place in hybrids, mediated by heterochromatin association and leading to preferential gene silencing, may also explain why different parents may contribute unequally to different hybrid traits. Occupancy of a single nuclear space by unfamiliar/alien genomes estab-lishes the arena of conflict be it at the level of the whole genome or of the particular parts which are the NORs. We close the circle with Navashin in the realm of the visible manifestation of amphiplasty, the heterochromatin which we can observe with the light microscope. We meet on the common ground of epigenetics, where we share the understanding but the world is new. The mechanism can now be seen to involve the well known processes of DNA methylation and histone acetylation which are related to states of differential chromatin organization. The repetitive nature of the heterochromatin, and the quantitative difference distinguishing the hybrid partners provide the environment for the mediation of chromatin conformation and its control of gene expression. Goliath is encumbered with heaviness, and the pervasiveness and positional aspects of the weight silence his actions. David needs less space, and can expresses himself unhindered and to greater effect in a competitive situation. Further experimental validation of this model will be gained through genome sequencing of parental species coupled with gene expression pattern analysis in
hybrids, allowing for the identification of repetitive DNA sequences linked to parental silenced genes.
ACKNOWLEDGEMENTS
The authors are most grateful to Prof. Neil Jones for the enlightening discussions and critical review of the manu-script. This work results from research projects supported by Fundação para a Ciência e Tecnologia, Portugal (PRAXIS/ C/AGR/11171/1998; POCTI/AGR/34000/2000; POCTI/BCI/ 38557/2001).
REFERENCES
[1] Navashin, M. Chromosomal alterations caused by hybridization and their bearing upon certain general genetic problems. Cytologia, 1934, 5: 169-203.
[2] Honjo, T.; Reeder, R.H. Preferential transcription of
Xenopus laevis ribosomal RNA in interspecies hybrids
between Xenopus laevis and Xenopus nulleri. J. Mol. Biol., 1973, 80: 217-228.
[3] Heitz, E. Die ursache der gasetzmmassigen zahl lage, form und grosse pflanzlicher nukleolen. Planta (Berl.), 1931, 12: 775-844.
[4] McClintock, B. The relationship of a particular chromosomal element to the develpment of the nucleoli in
Zea mays. Z. Zellforch. Mikrosk. Anat., 1934, 21: 294-328.
[5] Wallace, H.; Birnstiel, M.L. Ribosomal cistrons and the nucleolar organizer. Biochem. Biophys. Acta., 1966, 114: 296-310.
[6] Phillips, R.L.; Kleese, R.A.; Wang, S.S. The nucleolus organizer region of maize (Zea mays L.), chromosomal site of DNA complementary to ribosomal RNA. Chromosoma , 1971, 36: 79-88.
[7] Bach, R.; Allet, B.; Crippa, M. Sequence organization of the spacer in the ribosomal genes of Xenopus clivii and
Xenopus borealis. Nucleic Acids Res., 1981, 9: 5311-5330.
[8] Busby, S.J.; Reeder, R.H. Spacer sequences regulate transcription of ribosomal gene plasmids infected in Xenopus embryos. Cell, 1983, 34: 989-996.
[9] Labhart, P.; Reeder, R.H. Enhencer-like properties of the 60/81bp elements in the ribosomal gene spacer of Xenopus
laevis. Cell, 1984, 37: 285-289.
[10] Reeder, R.H.; Roan, J.G. The mechanism of nucleolar dominance in Xenopus hybrids. Cell, 1984, 38: 39-44. [11] Mukai, Y.; Endo, T.R.; Gill, B.S. Physical mapping of the
18S.26S rRNA multigene family in common wheat: identification of a new locus. Chromosoma, 1991, 100: 71-78.
[12] Martini, G.; Flavell, R. The control of the nucleolus volume in wheat, a genetic study of three developmental stages.
Heredity, 1985, 54: 111-120.
[13] Flavell, R.B.; O`Dell, M. Ribosomal RNA genes on homoeologous chromosomes of groups 5 and 6 in hexaploid wheat. Heredity, 1976, 37: 377-385.
[14] Flavell, R.B.; O`Dell, M. The genetic control of nucleolus formation in wheat. Chromosoma, 1979, 71: 135-152. [15] Martini, G.; O'Dell, M.; Flavell, R.B. Partial inactivation of
wheat nucleolus organizers by the nucleolus organizer chromosomes from Aegilops umbellulata. Chromosoma, 1982, 84: 687-700.
[16] Flavell, R.B. The structure and control of expression of ribosomal RNA genes. Oxford Surv. Pl. Mol. Cell Biol., 1986, 3: 251-273.
[17] Sardana, R.; O'Dell, M.; Flavell, R.B. Correlation between the size of the intergenic regulatory region, the status of cytosine methylation of rRNA genes and nucleolar expression in wheat. Mol. Gene Genet., 1993, 236: 155-162.
[18] Thomas, J.; Kaltsikes P.J. Effect of chromosomes 1A and 6B on nucleolus formation in hexaploid triticales. Can. J.
Genet. Cytol., 1983, 25: 292-297.
[19] Cermeño, M.C.; Orellana, J.; Santos, J.L.; Lacadena, J.R. Nucleolar activity in wheat, rye and derivates analyzed by a silver staining procedure. Chromosoma, 1984, 89: 370-376. [20] Cermeño, M.C.; Friebe, B.; Zeller, F.J.; Krolow, K.D.
Nucleolar competition in different (A/B)(A/B)RR and DDRR tetraploid triticales. Heredity, 1987, 58: 1-4. [21] Lacadena, J.R.; Cermeño, M.C.; Orellana, J.; Santos, J.L.
Evidence for wheat-rye nucleolar competition (amphiplasty) in triticale by silver staining procedure.
Theor. Appl. Genet., 1984, 67: 207-213.
[22] Silva, M.; Queiroz, A.; Neves, N.; Barão, A.; Castilho, A.; Morais-Cecílio, L.; Viegas, W. Reprogramming of rye rDNA in triticale during microsporogenesis. Chrom. Res., 1995, 3: 492-496.
[23] Appels, R.; Dvorak, J. The wheat ribosomal DNA spacer region: its structure and variation in populations and amongst species. Theor. Appl. Genet., 1982, 63: 337-348. [24] Pikaard, C. Nucleolar dominance: uniparental gene
silencing on a multi-megabase scale in genetic hybrids.
Plant Mol. Biol., 2000a, 43: 163-177.
[25] Pikaard CS. The epigenetics of nucleolar dominance.
Trends Genet., 2000b, 16: 495-500.
[26] Chen, Z.J.; Pikaard, C.S. Transcriptional analysis of nucleolar dominance in polyploid plants: biased expression/silencing of progenitor rRNA genes is developmentally regulated in Brassica. Proc. Natl. Acad.
Sci. USA, 1997a, 94: 3442-3447.
[27] Chen, J.; Comai, L.; Pikaard, C. Gene dosage and stochastic effects determine the severity and direction of uniparental rRNA gene silencing (nucleolar dominance) in Arabidopsis allopolyploids. Proc. Natl. Acad. Sci. USA, 1998, 95: 14891-14896.
[28] Pikaard, C. Nucleolar dominance and silencing of transcription. Trends Plant Sci., 1999, 12: 478-483. [29] Neves, N.; Silva, M.; Heslop-Harrison, J.S.; Viegas, W.
Nucleolar dominance in triticale: control by unlinked genes.
Chrom. Res., 1997b, 5: 125-131.
[30] Messing, J.; Carlson, J.; Hagen, G.; Rubenstein, I.; Oleson, A. Cloning and sequencing of the ribosomal RNA genes in maize: the 17S region. DNA, 1984, 3: 31-40.
[31] Eckenrode, V.K.; Arnold, J.; Meagher, R.B. Comparison of the nucleotide sequence of soybean 18s rRNA with the sequences of other small subunit rRNAs. J. Molec. Evol., 1985, 21: 259-269.
[32] Flavell, R.B.; O´Dell, M.; Sharp, P.; Nevo, E.; Beiles, A. Variation in the intergenic spacers in the ribosomal DNA of wild wheat, Triticum dicocoides, in Israel. Mol. Biol. Evol. , 1986, 3: 547-558.
[33] Grummt, I.; Roth, E.; Paule, M.R. Ribosomal RNA transcription is species specific. Nature, 1982, 296: 173-174.
[34] Miesfeld, R.; Arnheim, N. Species-specific rDNA transcription is due to promoter-specific binding factors.
Mol. Cell Biol., 1984, 4: 221-227.
[35] Miesfeld, R.; Sollner-Webb, B.; Croce, C.; Arnheim, N. The absence of human-specific ribosomal DNA transcription factor leads to nucleolar dominance in mouse-human hybrid cells. Mol. Cell Biol., 1984, 4: 1306-1312. [36] Learned, R.M.; Cordes, S.; Tjian, R. Purification and
characterization of a transcription factor that confers specificity to human RNA polimerase I. Mol. Cell Biol., 1985, 5: 1358-1369.
[37] Bell, A.; Gaston, K.; Williams, R.; Chapman, K.; Kolb, A.; Buc, H.; Minchin, S.; Williams, J.; Busby, S. Mutations that alter the stability of the Escherichia coli cyclic AMP receptor protein to activate transcription. Nucl. Acids Res., 1990, 18: 7243-7250.
[38] Schnapp, A.; Rosenbauer, H.; Grumm, I. Trans-acting factors involved in species-specificity and control of mouse ribosomal gene transcription. Mol. Cell Biochem., 1991,
104: 137-147.
[39] Heix, J.; Grummt, I. Species specificity of transcription by RNA polymerase I. Curr. Opin. Genet. Dev., 1995, 5: 652-656.
[40] Doelling, J.H.; Pikaard, C.S. Species-specificity of rRNA gene transcription in plants manifested as a switch in polymerase specificity. Nucleic Acids Res., 1996, 24: 4725-4732.
[41] Fan, H.; Yakura, K.; Miyanishi, M.; Sugita, M.; Sugiura, M. Transcription of plant RNA polymerase I-dependent rRNA genes is species-specific. Plant J., 1995, 8: 295-298. [42] Reeder, H.R. Mechanisms of nucleolar dominance in
animals and plants. J. Cell Biol., 1985, 101: 2013-2016. [43] Pikaard, C.; Chen, J. Nucleolar dominance. 1998 In
Transcription of Ribosomal RNA Genes by Eukaryotic RNA Polimerase I (ed. M.R. Paule) Springer-Verlag and Landes, R. G. Company, pp. 277-294.
[44] Chen, Z.J.; Pikaard, C.S. Epigenetic silencing of RNA polymerase I transcription: a role for DNA methylation and histone modification in nucleolar dominance. Gene Dev., 1997b, 11: 2124-2136.
[45] Frieman, M.; Chen, Z.J.; Saez-Vasquez, J.; Shen, L.A.; Pikaard, C.S. RNA polymerase I transcription in a Brassica interspecific hybrid and its progenitors: Tests of
transcription factor involvement in nucleolar dominance.
Genetics, 1999, 152: 451-460.
[46] Saez-Vasquez, J.; Pikaard, C.S. Extensive purification of a putative RNA polymerase I holoenzyme from plants that accurately initiates rRNA gene transcription. Proc. Natl.
Acad. Sci. USA, 1997, 94: 11869-11874.
[47] Jupe, E.R.; Zimmer, E.A. DNase I sensitivity and undermethylated rDNA is preferentially expressed in a maize hybrid. Plant Mol. Biol., 1993, 21: 805-821. [48] Caperta, A.D.; Neves, N.; Morais-Cecilio, L.; Malho, R.;
Viegas, W. Genome restructuring in rye attects the expression, organization and disposition of homologous rDNA loci. J. Cell Sci., 2002, 115: 2839-2846.
[49] Wolffe, A.; Matzke M. Epigenetics: regulation through repression. Science, 1999, 286: 481-486.
[50] Neves, N.; Silva, M.; Heslop-Harrison, J.S.; Viegas, W. rRNA gene activity and control of expression mediated by methylation and imprinting during embryo development in wheat x rye hybrids. Theor. Appl. Genet., 1995, 91: 529-533.
[51] Houchins, K.; O`Dell, M.; Flavell, R.B.; Gustafson, J.P. Cytosine methylation and nucleolar dominance in cereal hybrids. Mol. Gen. Genet., 1997, 255: 294-301.
[52] Gabbara, S.; Bhagwat, A.S. The mechanism of inhibition of DNA (cytosine-5)-methyltransferases by 5-azacytosine is likely to involve methyl transfer to the inhibitor. Biochem.
J., 1995, 307: 87-92.
[53] Vieira, R.; Queiroz, Á.; Morais, L.; Barão, A.; Mello-Sampayo, T.; Viegas, W. 1R chromosome nucleolus organizer region activation by 5-azacytidine in wheat-rye hybrids. Genome, 1990, 33: 707-712.
[54] Macleod, D.; Bird, A. DNase I sensitivity and methylation of active versus inactive rRNA genes in Xenopus species hybrids. Cell, 1982, 29: 211-218.
[55] Neves, N.; Castilho, A.; Silva, M.; Heslop-Harrison, J.S.; Viegas, W. Genomic interactions: gene expression, DNA methylation and nuclear architecture. Chrom. Today, 1997a,
12: 182-200.
[56] Durica, D.S.; Krider H.M. Studies on the ribosomal RNA cistrons in interspecific Drosophila hybrids. I. Nucleolar dominance. Devel. Biol., 1977, 59: 62-74.
[57] Gowher, H.; Leismann, O.; Jeltsch, A. DNA of Drosophila
melanogaster contains 5-methylcytosine. EMBO. J., 2000, 19: 6918-6923.
[58] Cheung W.L.; Briggs S.D.; Allis C.D. Acetylation and chromosomal functions. Curr. Opin. Cell Biol., 2000, 12: 326-333.
[59] Lee, D.Y.; Hayes, J.J.; Pruss, D.; Wolffe, A.P. A positive role for histone acetylation in transcription factor access to nucleosomal DNA. Cell, 1993, 72: 73-84.
[60] Vettese-Dadey, M.; Grant, P.A.; Hebbes, T.R.; Crane-Robinson, C.; Allis, C.D.; Workman, J.L. Acetylation of histone H4 plays a primary role in enhancing transcription factor binding to nucleosomal DNA. EMBO J., 1996, 15: 2508-2518.
[61] Jasencakova, Z.; Meister, A.; Walter, J.; Turner, B.M.; Schubert, I. Histone H4 acetylation of euchromatin and heterochromatin is cell cycle dependent and correlated with replication rather than with transcription. Plant Cell , 2000,
12: 2087-2100.
[62] Idei, S.; Kondo, K.; Turner, B.M.; Fukui, K. Tomographic distribution of acetylated histone H4 in plant chromosomes, nuclei and nucleoli. Chromosoma, 1996, 105: 293-302. [63] Lucchini, R.; Sogo, J.M. Replication of transcriptionally
active chromatin. Nature, 1995, 374: 276-80.
[64] Wallace, H.; Langridge, W.H.R. Differential amphiplasty and the control of ribosomal RNA synthesis. Heredity, 1971, 27: 1-13.
[65] Thompson, W.F.; Flavell, R.B. DNaseI sensitivity of ribosomal RNA genes in chromatin and nucleolar dominance in wheat. J. Mol. Biol., 1988, 204: 535-548. [66] Leitch, A. R.; Mosgoller, W.; Shi, M.; Heslop-Harrison, J.
S. Different patterns of rDNA organization at interphase nuclei of wheat and rye. J. Cell Sci., 1992, 101: 751-757. [67] Shaw, P.; Rawlins, D.; Highett, M. Nuclear and nucleolar
structure in plants. 1993 In: The Chromosome. Heslop-Harrison, J.S.; Flavell, R.B. (eds.), Bios Scient. Publish. Oxford, pp. 161-171.
[68] Delgado, M.; Morais-Cecílio, L.; Neves, N.; Jones, R.N.; Viegas, W. The influence of B chromosomes on rDNA organization in rye interphase nuclei. Chrom. Res., 1995, 3: 487-491.
[69] Shaw, P.J.; Jordan, E.G. The nucleolus. Annu. Rev. Cell
Devel. Biol., 1995, 11: 93-121.
[70] Jimenez, R.; Burgos, M.; Diaz de el Guardia, R. A study of the Ag-staining significance in mitotic NORs. Heredity, 1988, 60: 125-127.
[71] Wilson, H.D.; Barber, S.C.; Walters, T. Loss of duplicate gene expression in tetraploid Chenopodium. Biochem. Syst.
Ecol., 1983, 11: 7-13.
[72] Heslop-Harrison, J.S. Gene expression and parental dominance in hybrid plants. Development Suppl., 1990, 21-28.
[73] Gastony, G.J. Gene silencing in a polyploid homosporous fern: paleopolyploidy revisited. Proc. Natl. Acad. Sci. USA, 1991, 88: 1602-1605.
[74] Coulthart, M.; Denford, K.E. Isozyme studies in the genus
Brassica. Can. J. Plant Sci., 1982, 62: 621-630.
[75] Gottlieb, L.D. Conservation and duplication of isozymes in plants. Science, 1982, 216: 373-380.
[76] Soltis, D.E.; Soltis, P.S. Isozimes in plant biology. 1990 Chapman and Hall, London.
[77] Lee, H.; Chen, Z.J. Protein-coding genes are epigenetically regulated in Arabidopsis polyploids. Proc. Natl. Acad. Sci.
USA, 2001, 98: 6753-6758.
[78] Lewis, M.S.; Pikaard, C.S. Restricted chromosomal silencing in nucleolar dominance. Proc. Natl. Acad. Sci.
[79] Redi, C.A.; Garagna, S.; Zacharias, H.; Zuccotti, M.; Capanna, E. The other chromatin. Chromosoma, 2001, 110: 136-147.
[80] Kasha K.J.; Yao, Q.; Simion, E.; Hu, T.; Oro, R. Production and applications of doubled haploids in crops. 1995 In: Induced Mutations and Molecular Genetics for Crop Improvement. IAEA, Vienna, pp.23-37.
[81] Pasakinskiene, I.; Anamthawat-Jónsson, K.; Humphries M.W.; Jones, R.N. Novel diploids following chromosome elimination and somatic recombination in Lolium
multiflorum x Festuca arundinacea hybrids. Heredity,
1997, 78: 464-469.
[82] Wilkinson, M.J.; Bennett, S.T.; Clulow, S.A.; Allainguillaume, J.; Harding, K.; Bennett, M.D. Evidence for somatic translocation during potato dihaploid induction.
Heredity, 1995, 74: 146-151.
[83] Heitz, E. Das heterochromatin der moose. I. Jahrb. Wiss.
Biol., 1928, 69: 762-818.
[84] Belyaeva, E.S.; Demakova, O.V.; Umbetova, G.H.; Zhymulev, I.F. Cytogenetic and molecular aspects of position-effect variegation in Drosophila melanogaster. V. Heterochromatin-associated protein HP1 appears in euchromatin chromosomal regions that are inactivated as a result of position-effect variegation. Chromosoma, 1993,
102: 583-590.
[85] Eissenberg, J.C.; Elgin, S.C. The HP1 protein family: getting a grip on chromatin. Curr. Opin. Genet. Dev. , 2000,
10: 204-210.
[86] Zhimulev, I.F.; Semeshin, V.F.; Kulichkov, V.A.; Belyaeva, E.S. Intercalary heterochromatin in Drosophila. I. Localization and general characteristics. Chromosoma, 1982, 87: 197-228.
[87] Bedo, D.G. Cytological characterization of heterochromatin in meiotic and mitotic chromosomes of the Old World screwworm fly, Chrysomya bezziana (Diptera:
Calliphoridae). Genome, 1991, 34: 631-637.
[88] Kunze, B.; Weichenhan, D.; Virks, P.; Traut, W.; Winking, H. Copy numbers of a clustered long-range repeat determine C-band staining. Cytogenet. Cell Genet., 1996,
73: 86-91.
[89] Weichenhan, D.; Kunze, B.; Traut, W.; Winking, H. Evolution by fusion and amplification: the murine Sp100-rs gene cluster. Cytogenet. Cell Genet., 1998, 80: 226-231. [90] Devlin, J.J.; Panganiban, L.C.; Devlin, P.E. Random
peptide libraries: a source of specific protein binding molecules. Science, 1990, 249: 404–406.
[91] Gatti, M.; Pimpinelli, S. Functional elements in Drosophila
melanogaster heterochromatin. Annu, Rev, Genet., 1992, 26: 239-275.
[92] Berghella, L.; Dimitri, P. The rolled gene of Drosophila
melanogaster is extensively polytenized and
transcriptionally active in the salivary gland chromocenter.
Genetics, 1996, 144: 117-125.
[93] Ohno, S. So much junk DNA in our genome. 1972 In: Smith, H.H. (ed), Brookhaven Symposia in Biology 23: Gordon & Breach, New York, pp. 366-370.
[94] Ohno, S. Dispensable genes. Trends in Genet., 1985, 1: 160-164.
[95] John, B. The biology of heterochromatin. 1988 In: Verma, S. (ed) Heterochromatin. Cambridge University Press, Cambridge, PP. 1-147.
[96] Henning, W. Heterochromatin and germ line-restricted DNA. Results Probl. Cell Differ., 1986, 13: 175-92. [97] Pardue, M.L.; Henning, W. Heterochromatin: junk or
collectors item? Chromosoma, 1990, 100: 3-7.
[98] Zuckerkandl, E.; Hennig, W. Tracking heterochromatin.
Chromosoma, 1995, 104: 75-83.
[99] Henning, W. Heterochromatin. Chromosoma, 1999, 108: 1-9.
[100] Tiersch, T.R.; Wachtel S.S. On the evolution of genome size of birds. J. Hered., 1991, 82: 363-8.
[101] Van den Busche, R.A.; Longmire, J.L.; Baker, R.J. How bats achieve a small C-value: frequency of repetitive DNA in Macrotus. Mamm. Genome, 1995, 6: 521-525.
[102] Garagna, S.; Perez-Zapata, A.; Zuccotti, M.; Mascheretti, S.; Marziliano, N.; Redi C.A.; Aguilera, M.; Capanna, E. Genome composition in Venezuelan spiny-rats of the genus
Proechimys (Rodentia, Echimyidae). I. Genome size,
C-heterochromatin and repetitive DNAs in situ hybridization patterns. Cytogenet. Cell Genet., 1997, 78: 36-43.
[103] Vinogradov, A.E. Nucleotypic effect in homeotherms: body mass-corrected basal metabolic rate of mammals is related to genome size. Evolution, 1995, 49: 1249-1259.
[104] Beaton, M.J.; Cavalier-Smith, T. Eukaryotic non-coding DNA is functional: evidence for different scaling of cryptomonal genomes. Proc. R. Soc. Lond. Biol. Sci., 1999,
266: 2053-2059.
[105] Brenner, S.; Elgar, G.; Sandford, R.; Macrae, A.; Venkatesh, B.; Aparicio, S. Characterization of the pufferfish (Fugu) genome as a compact model vertebrate genome. Nature, 1993, 366: 265-268.
[106] Bennetzen, J.L. The many hues of plant heterochromatin.
Genome Biology, 2000, 1: Reviews 107.
[107] Ramsay, G.; Dyer, A.F. Nucleolar organiser suppression in barley x rye hybrids. Kew Chromosome Conf. 2nd 1983. P. 361.
[108] Padilla, J.A.; Martin, A. Morphology and cytology of
Hordeum chilense x Hordeum bulbosum Hybrids. Theor. Appl. Genet., 1983, 65: 353-355.
[109] Bennett, M.D. Nuclear architecture and its manipulation. 1984. In: Gene Manipulation in Plant Improvement, Gustafson, J.P. (ed.), Plenum Publish. Corp., London, pp. 469-502.
[110] Bennett, M.D.; Smith, J.B. Nuclear DNA amounts in angiosperms. Philos. Tr. Roy. Soc., B 1976, 274: 227-274. [111] Thiebaud, C.H.; Fischberg, M. DNA Content in the Genus
Xenopus Chromosoma, 1977, 59: 253-257.
[112] Laird, C.D.; McCarthy, B.J. Molecular characterization of the Drosophila genome. Genetics, 1969, 63: 865-882.
[113] Rasch, E.M. DNA "standards" and the range of accurate DNA estimates by Feulgen absorption microspectrophoto-metry. 1985 In: Advances in Microscopy, edited by R. R. Cowden and S. H. Harrison. Alan R. Liss, New York, pp. 137-166.
[114] Arumuganathan, K.; Earle, E.D. Nuclear DNA content of some important plant species. Plant Mol. Biol. Rep., 1991,
9: 208-219.
[115] Shaked, H.; Kashkush, K.; Ozkan, H.; Feldman, M.; Levy, A.A. Sequence elimination and cytosine methylation are rapid and reproducible responses of the genome to wide hybridization and allopolyploidy in wheat. Plant Cell, 2001,
13: 1749-1759.
[116] Ozkan, H.; Levy, A.A.; Feldman, M. Allopolyploidy-induced rapid genome evolution in the wheat
(Aegilops-Triticum) group. Plant Cell, 2001, 13: 1735-1747.
[117] Morais-Cecilio, L.; Delgado, M.; Jones, R.N.; Viegas, W. Modification of wheat rDNA loci by rye B chromosomes: a chromatin organization model. Chrom. Res. , 2000, 8 : 341-351.
[118] Henikoff, S. Nuclear organization and gene expression: homologous pairing and long-range interactions. Curr.
Opin. Cell Biol., 1997, 9: 388-395.
[119] Wallrath, L.L. Unfolding the mysteries of heterochromatin.
Curr. Opin. Genet. Dev., 1998, 8: 147-153.
[120] Lyko, F.; Paro, R. Chromosomal elements conferring epigenetic inheritance. Bioessays, 1999, 21: 824-832. [121] Garrick, D.; Fiering, S.; Martin, D.I.; Whitelaw, E.
Repeat-induced gene silencing in mammals. Nat. Genet., 1998, 18: 56-59.
[122] Dernburg, A.F.; Sedat, J.W.; Hawley, R.S. Direct evidence of a role for heterochromatin in meiotic chromosome segregation. Cell, 1996, 86: 135-146.
[123] Renauld, H.; Gasser, S.M. Heterochromatin: a meiotic matchmaker. Trends Cell Biol., 1997, 7: 201-205.
[124] Csink, A.K.; Henikoff, S. Genetic modification of heterochromatic associations and nuclear organization in
Drosophila. Nature, 1996, 381: 529-531.
[125] Dorer, D.R.; Henikoff, S. Transgene repeat arrays interact with distant heterochromatin and cause silencing in cis and trans. Genetics, 1997, 147: 1181-1190.
[126] Martin-Morris, L.E.; Csink, A.K.; Dorer, D.R.; Talbert, P.B.; Henikoff. S. Heterochromatic trans-inactivation of
Drosophila white transgenes. Genetics, 1997, 147: 671-677.
[127] Flavell, R.B. Inactivation of gene expression in plants as a consequence of specific sequence duplication. Proc. Natl.
Acad. Sci. USA, 1994, 91: 3490-3496.
[128] Comai L. Genetic and epigenetic interactions in allopolyploid plants. Plant Mol. Biol., 2000, 43: 387-399. [129] Durica, D.S.; H.M. Krider. Studies on the ribosomal RNA
cistrons in interspecific Drosophila hybrids. II. Heterochromatic regions mediating nucleolar dominance.
Genetics, 1978, 89: 37-64.
[130] Lu, B.Y.; Ma, J.; Eissengerg, J.C. Developmental regulation of heterochromatin-mediated gene silencing in Drosophila.
Development, 1998, 125: 2223-2234.
[131] Vlassova, I.E.; Graphodatsky, A.S.; Belyaeva, E.S.; Zhimulev, IF. Constitutive heterochromatin in early embryogenesis of Drosophila melanogaster. Mol. Gen.,
Genet.,1991, 229: 316-318.
[132] Gould, A. Functions of mammalian Polycomb group and trithorax group related genes. Curr. Opin. Genet. Dev., 1997, 7: 488–494.
[133] Gerasimova, T.I.; Corces, V.G. Polycomb and trithorax group proteins mediate the function of a chromatin insulator. Cell, 1998, 92: 511-521.
[134] Ceccarelli, M.; Cionini, P.G. Tissue-specific nuclear repatterning in plant cells. Genome, 1993, 36: 1092-1098. [135] Carmo-Fonseca, M.; Mendes-Soares, L.; Campos, I. To be
or not to be in the nucleolus. Nature Cell Biol., 2000, 2: 107-112.
[136] Holliday, R. The biological significance of meiosis. In Controlling events in meiosis, 1984 Evans, C. and Dickenson, H. eds. Society for Experimental Biology, London, pp. 381-394.
[137] Castilho, A.; Neves, N.; Rufini-Castiglione, M.; Viegas, W.; Heslop-Harrison, J.S. 5-Methylcytosine distribution and genome organization in triticale before and after treatment with 5-azacytidine. J. Cell Sci., 1999, 112: 4397-4404.
[138] Cockell, M.; Gasser, S.M. Nuclear compartments and gene regulation. Curr. Opin. Genet. Dev., 1999, 9: 199-205. [139] Richards, E.J.; Elgin, S.C.R. Epigenetic codes for
heterochromatin formation and silencing, rounding up the usual suspects. Cell, 2002, 108: 489-500.
[140] Gustafson, J.P.; Bennett, M.D. Preferencial selection from wheat-rye substitutions in 42-chromosome triticale. Crop
Sci., 1976, 16: 688-693.
[141] Gustafson, J.P.; Dillé, J.E.; Skovmand, B. Wheat substitution in hexaploid triticale. Plant Breeding, 1989,
