• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of genesdevCSHL PressJournal HomeSubscriptionseTOC AlertsBioSupplyNetGenes & Development
Genes Dev. Aug 1, 2002; 16(15): 1906–1918.
PMCID: PMC186425

Differential chromatin structure within a tandem array 100 kb upstream of the maize b1 locus is associated with paramutation


Recombination mapping defined a 6-kb region, 100 kb upstream of the transcription start site, that is required for B-I enhancer activity and paramutation—a stable, heritable change in transcription caused by allele interactions in maize (Zea mays). In this region, B-I and B‘ (the only b1 alleles that participate in paramutation) have seven tandem repeats of an 853-bp sequence otherwise unique in the genome; other alleles have one. Examination of recombinant alleles with different numbers of tandem repeats indicates that the repeats are required for both paramutation and enhancer function. The 6-kb region is identical in B-I and B‘, showing that epigenetic mechanisms mediate the stable silencing associated with paramutation. This is the first endogenous gene for which sequences required for paramutation have been defined and examined for methylation and chromatin structure. The tandem repeat sequences are more methylated in B-I (high expressing) relative to B‘ (low expressing), opposite of the typical correlation. Furthermore, the change in repeat methylation follows establishment of the B‘ epigenetic state. B-I has a more open chromatin structure in the repeats relative to B‘. The nuclease hypersensitivity differences developmentally precede transcription, suggesting that the repeat chromatin structure could be the heritable imprint distinguishing the two transcription states.

Keywords: Epigenetics, chromatin, DNA methylation, transcriptional enhancer

Paramutation is an interaction between specific alleles that causes a meiotically heritable change in one allele’s expression. Paramutation was originally defined as an interaction between two alleles (Brink 1956), but paramutation has also been shown between a transgene and a homologous endogenous gene in nonallelic positions (Sidorenko and Peterson 2001). Features are shared with trans-silencing phenomena in Drosophila (Henikoff and Comai 1998) and fungi (Colot et al. 1996; van West et al. 1999) and with transcriptional gene silencing in plants (Chandler et al. 2000; Vaucheret and Fagard 2001) and animals (Garrick et al. 1998; Henikoff 1998). However, with few exceptions (Grewal and Klar 1996; Cavalli and Paro 1998), silencing is not meiotically heritable in fungi and animals. Expression of transposable elements, or genes with transposable elements, can be heritably changed by interactions with other elements (McClintock 1965; Krebbers et al. 1987; Martienssen 1996). Also, allele interactions between transposons alter the excision-repair mechanism (van Houwelingen et al. 1999). An eventual understanding of mechanisms underlying paramutation should reveal how alleles and other homologous sequences interact in the nucleus to influence the regulation of each other, and how heritable expression states are established, maintained through numerous cell divisions, and transmitted to the next generation.

Paramutation has been described for four maize (Zea mays) genes (b1, r1, pl1, and p1), all of which encode transcription factors that activate the biosynthesis of flavonoid pigments (Brink 1956; Coe 1959; Hollick et al. 1995; Sidorenko and Peterson 2001), and for genes in other plant species (for review, see Brink 1973). Paramutation at b1 is extremely stable and the most penetrant of all systems described to date. The paramutagenic B‘ allele, which causes a lightly pigmented phenotype, arises spontaneously from the paramutable B-I allele, which confers a darkly pigmented phenotype, at a frequency of ~1%–10% (Coe 1966). Once formed, B‘ is extremely stable. In B-I/B‘ plants, B-I is always changed to B‘ (Coe 1966; Patterson et al. 1995). The new B‘ allele (which had been B-I in the previous generation) is indistinguishable from the parental B‘ allele in its ability to paramutate B-I in subsequent generations.

Paramutation is associated with a 10- to 20-fold reduction in transcription of B‘ relative to B-I (Patterson et al. 1993). No differences in DNA sequences, rearrangements, or methylation patterns have been detected within 10 kb spanning the transcribed region of B-I and B‘ (Patterson et al. 1993, 1995). However, a quantitatively distinct DNase I hypersensitive site was observed in B-I and B‘ chromatin near the transcription start site (Chandler et al. 2000). Recombination experiments showed that sequences required for paramutation are tightly linked to B‘ (≤0.1 cM) and map upstream of the transcribed region (Patterson et al. 1995). Recent fine-structure recombination experiments mapped the B‘ sequences required for paramutation to a 13-kb region located ~100 kb upstream of the transcription start site (Stam et al. 2002). Herein we describe fine-structure recombination experiments that map B-I enhancer and paramutation activities to a 6-kb region within the 13-kb B‘ region. We compare this region in B‘ and B-I with two alleles that do not participate in paramutation and discuss models consistent with our findings.


Isolation of recombinant B-I alleles delimiting the sequences required for high expression

Previously, the sequences required for B-I transcription were mapped upstream of the transcribed region (Patterson et al. 1995). To precisely map these sequences and determine their location relative to sequences required for paramutation, recombination experiments were performed using B-I and B-Peru. For simplicity, throughout the manuscript we refer to the region required for high expression as the enhancer. B-I plants are dark purple, whereas B-Peru plants are essentially green (Radicella et al. 1992). B-I participates in paramutation (it changes to B‘ when heterozygous with B‘), whereas B-Peru does not (it is unchanged when heterozygous with B‘; i.e., it is neutral). The two alleles are also distinguished by differences in seed expression: B-Peru produces purple seeds, whereas B-I produces colorless seeds. A 2.5-kb sequence in the promoter-proximal region of B-Peru confers seed expression (Radicella et al. 1992; Selinger et al. 1998). The seed pigment and unique 2.5 kb in B-Peru provide phenotypic and molecular markers for the promoter-proximal regions of the two alleles.

We hypothesized that a regulatory region containing a transcriptional enhancer, symbolized by E in Figure Figure1,1, functions in B-I and is either not present or not functional in B-Peru. Our strategy to map E is diagramed in Figure Figure1.1. Recombination events between E and the B-I transcribed region (Interval I) would transfer E to B-Peru, leading to an allele with high plant expression and the loss of E from B-I. The Interval I breakpoint occurring farthest upstream would define a 3′ boundary for E. Recombinants in Interval II should produce the same pigment phenotypes as the parental alleles, darkly pigmented B-I plants and nearly green B-Peru plants. The Interval II breakpoint located farthest downstream would define the 5′ boundary for E.

Figure 1
Recombination experiments with B-I and B-Peru (B-P) to map enhancer sequences. The gray oval (E) contains the hypothetical enhancer sequences. The phenotypic marker glossy2 (gl2) and the molecular marker tmp were used to identify recombinants; polymorphisms ...

To identify recombinants in Intervals I and II, we used the phenotypic glossy2 (gl2) and molecular tmp markers. Plants containing gl2 B-I were crossed with Gl2 B-Peru plants, and the F1 plants were crossed with gl2 b-K55 plants, resulting in 50% colorless seeds (B-I/b-K55) and 50% purple seeds (B-Peru/b-K55). To first test if we could transfer the B-I enhancer to B-Peru, we planted 8910 purple seeds and scored the resulting plants for pigment levels. One plant with vegetative pigment much darker than B-Peru was identified (Fig. (Fig.1),1), confirmed to carry a recombinant allele using tmp, and designated BI||BP-D1 (D refers to dark). The || symbol indicates a recombinant allele with the parental allele contributing sequences upstream of the breakpoint indicated left of the vertical bars and the parental allele contributing the promoter-proximal and coding regions indicated to the right of the vertical bars.

The isolation of BI||BP-D1 showed that recombination mapping could be used to localize sequences required for high expression in B-I. Subsequently, additional seeds were planted in the greenhouse, with 5677 colorless (B-I coding and promoter-proximal sequences) and 10,106 purple (B-Peru coding and promoter-proximal sequences) seeds germinating. Seedlings with a recombination event between gl2 and b1 (~20%) were transferred to the field, and leaf samples were screened via PCR for molecular markers (see Materials and Methods). Two Interval I recombinants were identified: the original allele isolated by screening for increased pigment levels (BI||BP-D1) and a second allele (BI||BP-D2). Both recombinants produced dark B-Peru plants, showing that enhancer sequences were transferred from the high-expressing B-I allele to the weakly expressing B-Peru allele. In all, 29 Interval II recombination events were identified: 23 BP||BI recombinants were darkly pigmented (BP||BI-D); 3 BP||BI recombinants had a light, B‘-like phenotype consistent with spontaneous paramutation of B-I to B‘; and 3 BI||BP recombinants had the B-Peru pigment phenotype, 2 of which set seed (BI||BP-L; L for light). The Interval III recombinants identified (~4000) were not analyzed further.

Sequences required for high expression of B-I are between 99 and 105 kb upstream of the transcription start site

We used B‘ sequence information (Stam et al. 2002) to identify the most informative recombinants. Recombination breakpoints were mapped using DNA blot analyses with unique copy probes distributed throughout the 110-kb region upstream of the B-I transcription start site (Stam et al. 2002), and then designing primers to amplify DNA surrounding the breakpoints from the recombinant alleles. All fragments were cloned, sequenced, and compared with parental B-I and B-Peru sequences (Table (Table1).1).

Table 1
Recombinants defining B-I sequences required for enhancer and paramutation activities

The Interval I recombinant allele with a breakpoint at 99 kb (BI||BP-D1) upstream of the B-I transcription start site defined the 3′ boundary. The darkly pigmented phenotype of this recombinant indicated the enhancer sequences had been transferred to B-Peru and were therefore mapped upstream of the breakpoint. The 5′ boundary was defined by the BP||BI-D1 and BP||BI-D2 alleles, which fully retained the enhancer function and have a recombination breakpoint 105 kb upstream of the B-I transcription start site.

Enhancer and paramutation sequences are contained in the same DNA region

Sequences required for paramutation in B‘ mapped between 93 kb and 106 kb upstream of the B‘ transcription start site (Stam et al. 2002), overlapping the region required for enhancer activity. We therefore asked which recombinant alleles retained B-I sequences required for paramutation (Table (Table1).1). The BI||BP-D1 and BI||BP-D2 alleles were not paramutagenic; they did not decrease B-I expression when heterozygous with B-I. To test if these alleles were paramutable (if their expression was reduced by B‘), they were crossed to B‘. To test if the alleles could become paramutagenic, the resulting F1 plants were crossed to B-I. When heterozygous with B‘, the BI||BP-D1 and BI||BP-D2 alleles became lightly pigmented and paramutagenic (Table (Table1).1). A similar result was obtained with the BP||BI-D alleles. All of these alleles contained not only the B-I enhancer sequences, but also the sequences required for B-I to participate in paramutation. The two BI||BP-L alleles that lacked the enhancer were not paramutagenic in crosses with B-I and did not become paramutagenic when crossed to B‘ and subsequently B-I (Table (Table1).1). Thus, the B-I enhancer and paramutation sequences map to the same 6-kb region, within the 13-kb region previously shown to be required for B‘ paramutagenicity (Stam et al. 2002).

B‘ and B-I are epialleles

Comparison of the B‘ and B-I regions required for paramutation should show whether paramutation is associated with DNA sequence changes or not. We used information from B‘ (between 94 and 110 kb upstream of the transcription start site; Stam et al. 2002) to isolate the same region from B-I and two neutral alleles, B-Peru and b-K55. The resulting fragments were cloned and sequenced (see Materials and Methods). Extensive DNA blot analyses were done on leaf DNA from B‘, B-I, B-Peru, and b-K55 plants to verify the sequence contigs (data not shown).

B‘ and B-I sequences between 94 and 110 kb upstream of the transcription start site are identical (accession nos. AY078063, AF475145, AF483657). DNA sequence comparisons of the promoter-proximal and coding regions of B-I and B‘ also revealed no differences between these two alleles (Patterson et al. 1993). Furthermore, extensive physical mapping using pulsed field gel electrophoresis (PFGE) and regular gel electrophoresis with DNA blot analyses revealed no differences between B‘ and B-I within a 220-kb region upstream of the transcribed region (Stam et al. 2002). These data indicate that B-I and B‘ are epialleles and paramutation is mediated by epigenetic mechanisms.

Sequences involved in paramutation contain tandem repeats

To investigate potential epigenetic mechanisms, we compared the 6-kb region of B-I and B‘ with two neutral alleles, B-Peru and b-K55. B-I and B‘ contain an 853-bp fragment that is repeated seven times in tandem, whereas both B-Peru and b-K55 contain a single copy of this fragment (white and shaded arrows in Fig. Fig.2A).2A). The sequence is not repeated elsewhere in the genome. The 3.69-kb fragment in B-Peru that spans the single-copy 853-bp sequence is nearly identical to the same region in B-I and B‘ (Fig. (Fig.2A),2A), suggesting that the repeated nature of this sequence in B-I and B‘, rather than sequence polymorphisms, is mediating paramutation. We examined three other neutral alleles using DNA blots, and all three also had one copy of this region (data not shown).

Figure 2
Sequence and structure of B‘, B-I, B-Peru, b-K55, and recombinants. (A) Percentage identity between B-I, B-Peru (B-P), and b-K55 as compared with B‘ is indicated by bar color. (B) Restriction maps of recombinant alleles defining 5′ ...

The structure of three recombinant alleles provided a test of whether the repeats are involved in paramutation. BI||BP-L1, BI||BP-D1, and BI||BP-D2 contain 1, 3, and 5 repeats, respectively, whereas B‘ and B-I contain seven repeats (Fig. (Fig.2).2). A comparison of the paramutation strength of each of these alleles shows that the BI||BP-L1 allele with one repeat is neutral. In contrast, both alleles with multiple repeats, BI||BP-D1 and BI||BP-D2, become paramutagenic when crossed to B‘. The BI||BP-D1 allele with three repeats is less paramutagenic (4 exceptions in 118 plants; Table Table1)1) than the BI||BP-D2 allele with five repeats (0 exceptions in 589 plants). When each of the four dark exceptions was crossed again to B-I and purple seeds (BI||BP-D1/B-I) were planted, both B‘ and B-I plants were observed. The presence of B‘ plants indicated that the BI||BP-D1 allele retained paramutagenic activity. The occurrence of even more B-I plants than what was observed in the first generation (2/18, 3/19, 12/104, 30/41) suggests that the paramutagenic state of BI||BP-D1 is unstable. The BI||BP-D1 allele was unstable even after multiple crosses to B‘. When BI||BP-D1 was crossed to B‘ for two or three sequential generations and then crossed with B-I, 3/200 and 4/498 B-I plants were observed out of the total examined, respectively. In sharp contrast, no exceptions have ever been observed for the B‘ allele; when it is crossed to B-I only B‘ plants are observed with more than 100,000 plants tested in 50 years of study (Coe 1966; Patterson et al. 1993, 1995; E.H. Coe, Jr. and V.L. Chandler, unpubl.). The facts that the allele with one repeat is not paramutagenic and the allele with three repeats (two B-I and one hybrid repeat) has a reduced stability suggest that multiple repeats are required for paramutation.

Although all the repeats are nearly identical, the six sequence polymorphisms between B-Peru and the seven repeats in B-I and B‘ (Fig. (Fig.3)3) allowed us to compare where recombination occurred relative to these nucleotide polymorphisms in the alleles with differing numbers of repeats (Fig. (Fig.2B).2B). In BI||BP-L1 the breakpoint is upstream of 220 bp (Fig. (Fig.3),3), resulting in retention of the six B-Peru polymorphisms. These same polymorphisms are retained in the hybrid fifth repeat in BI||BP-D2, as its breakpoint was between 424 and 470 bp. The fact that BI||BP-D2 is fully paramutagenic supports the argument that the B-Peru repeat sequence polymorphisms alone are not what prevents paramutation activity.

Figure 3
Sequence features of the tandem repeat in B‘ and B-I. Alignment of the B-Peru sequence with the consensus of the seven repeats in B‘ and B-I. Polymorphisms among the seven repeats are indicated below the consensus, the number in parentheses ...

Comparison of the structure of BI||BP-L1, with no enhancer activity and containing one repeat and B-I sequences upstream (the B-I and B-Peru sequences in the first 220 bp of the first repeat and 1292 bp upstream are identical), with that of BI||BP-D1, BI||BP-D2, BP||BI-D1, and BP||BI-D2, which have multiple repeats and retain enhancer activity (Fig. (Fig.2B),2B), indicates that multiple repeats are also needed for enhancer activity. Three repeats can mediate enhancer function, as there was no obvious pigment difference between BI||BP-D1 and BI||BP-D2 with three and five repeats, respectively. Both BI||BP-D1 and BI||BP-D2 have slightly less pigment than B-I, which could mean that seven repeats are needed for full enhancer activity. However, there could be additional sequences in B-I between the enhancer and promoter-proximal regions that contribute to the higher expression level; such a sequence was identified in B‘ (Stam et al. 2002). Also, the enhancer may not act as efficiently with the distinct promoter-proximal sequences in B-Peru relative to B-I.

Sequence characteristics of the tandem repeat in B-I and B‘

The above studies show that multiple repeats are involved in paramutation and enhancer activity; therefore, we were interested in the tandem repeat’s sequence properties. Sequencing identified a few sequence polymorphisms among the repeats in B‘, but B-I had exactly the same polymorphisms. Figure Figure33 shows an alignment of the sequence found in B-Peru with a consensus of the seven repeats found in B‘ and B-I. The 60% AT-rich sequence with blocks of AT nucleotides interspersed with short blocks of high GC content appears to be noncoding and showed no significant identity with any known gene. Two sequences associated with epigenetic properties, yeast autonomously replicating sequences (ARS; Maundrell et al. 1988) and matrix attachment regions (MARs; Michalowski et al. 1999), are also AT-rich. We therefore examined the tandem repeat for the yeast ARS consensus and for sequence properties typical of MARs; numerous putative ARS and a putative MAR were identified (Fig. (Fig.3).3). Repeat finder programs identified short (~10–20 bp) inverted and direct repeats too numerous to display.

B-I shows more DNA methylation within the repeated region than B‘

In higher plants and animals, differences in transcription states frequently correlate with cytosine methylation. Previous analyses did not detect differences in DNA methylation within B‘ and B-I in an ~10-kb region (the first 941 bp of the transcribed region plus 9 kb upstream; Patterson et al. 1993). To compare the methylation pattern in the sequences required for paramutation and enhancer activity in B‘ and B-I, genomic DNA was isolated from B‘, B-I, and B-Peru leaves at three different developmental stages (two-week-old seedlings, just prior to and after tassel emergence) and from husk tissue on the first day that silks were visible. Each DNA sample was cut with EcoRI and Sau3AI (which does not cut its recognition site, GATC, when the C is methylated) or with EcoRI and MboI (a methylation-insensitive isoschizomer of Sau3AI). We examined 24 samples from 23 B‘ individuals, 30 samples from 26 B-I individuals, and 9 samples from 8 B-Peru individuals. No differences in methylation patterns were detected between the different developmental stages or tissues.

The Sau3AI sites flanking the fragments recognized by probe A served as a control for complete digestion as these sites were unmethylated in all genotypes (Fig. (Fig.4,4, probe A). There are three GATC sites within the 853-bp repeat (Fig. (Fig.4,4, labeled 1, 2 and 3), and when all sites are cut in B‘ and B-I DNA, fragments of 74, 296, and 483 bp are expected, but only the 296- and 483-bp fragments are seen along with the 189- and 397-bp fragments flanking the repeats (Fig. (Fig.4,4, MboI digests hybridized with the repeat probe C). When all sites are cut in B-Peru DNA, fragments of 397, 296, and 189 bp are observed (Fig. (Fig.4,4, probe C); eight of the nine B-Peru samples had unmethylated Sau3AI sites (Fig. (Fig.4,4, the Sau3AI and MboI patterns were identical). The one exception showed methylation at the third Sau3AI site (data not shown). Both B‘ and B-I samples have methylated Sau3AI sites in the tandem repeat region (Fig. (Fig.4,4, probe C). All B‘ and B-I samples showed methylation of the symmetric C in the third Sau3AI site (GATCTG), resulting in the 779-bp fragment. In 90% of the B-I samples (27/30) the asymmetric C in the second Sau3AI site (GATCCA) was methylated, generating the 376- and 853-bp fragments. The presence of the 296- and 779-bp fragments in B-I shows that some repeats are unmethylated at site 2. In contrast, none of the B‘ samples (0/24) showed detectable methylation of this site. The asymmetric site 1 (GATCCC) was generally cut in all repeats in both genotypes, but we occasionally saw a weak 557-bp band, representing methylation at this site.

Figure 4
B-I DNA is more methylated within the tandem repeats than B‘ DNA. DNA blot analyses were done on B‘, new B‘ (nB‘), B-I, and B-Peru genomic DNA cut with EcoRI and Sau3AI (methylation-sensitive) or MboI (methylation-insensitive). ...

To determine if the differential B‘ and B-I repeat methylation pattern extended to other regions, two other probes were used on the same DNA blots. All other sites showed equivalent methylation in B-I and B‘. The Sau3AI sites monitored by probe B (Fig. (Fig.4,4, upstream of the repeats) were always unmethylated in B-Peru (9/9) and occasionally methylated in B‘ (6/24) and B-I (9/30) samples. A similar result was obtained with probe D located immediately downstream of the repeats. Most of the B‘ (22/24), B-I (25/30), and B-Peru (9/9) samples had a low level of methylation in this region; 2/24 B‘ and 5/30 B-I showed additional methylation. The only reproducible difference in methylation pattern between B‘ and B-I was at the second Sau3AI site in the repeats, suggesting it could be used as a molecular marker for the B‘ versus B-I epigenetic state.

The B‘ methylation pattern follows the establishment of the B‘ epigenetic state

To investigate the dynamics of the methylation pattern difference in the B‘ and B-I repeats, we next examined B‘ alleles that formed spontaneously. B-I spontaneously changes to B‘ at a relatively high frequency (Coe 1966; Patterson et al. 1995), and these events are seen as B‘ progeny arising from B-I parents. To determine whether a newly formed B‘ allele had the B-I (parental) or B‘ (present epigenetic state) methylation pattern, DNA was isolated from leaves (just before and after tassel emergence) and husks of plants with new B‘ alleles. The DNA was digested with EcoRI plus Sau3AI or MboI, and DNA blot analyses were performed using all four probes shown in Figure Figure4.4. In the regions flanking the repeats, the new B‘ samples (9 samples from 7 individuals) showed the same methylation patterns as those observed with B‘ and B-I samples (0/9 were methylated at probe A, 3/9 were methylated at probe B, 1/9 had additional methylation at probe D). All nine new B‘ samples showed methylation of the third Sau3AI site in the repeats as seen with both B‘ and B-I. Interestingly, seven out of the nine new B‘ DNA samples showed methylation of the second Sau3AI site in the repeat, characteristic of the B-I pattern (Fig. (Fig.4,4, 376- and 853-bp fragments). There was variation in how many of the repeats were methylated at the second Sau3AI site in both new B‘ and B-I, but the variation was similar. We calculated the intensity of the 853-bp fragment and divided it by the total intensity (853 + 779 bp) in the 27 B-I and 7 new B‘ samples with substantial methylation at the second Sau3AI site. The ratios were 0.29 ± 0.08 and 0.3 ± 0.19 for B-I and new B‘ alleles, respectively. The remaining two new B‘ samples were much less methylated, but a slight hint of the 853-bp fragment characteristic of B-I was seen (the ratio of 853/(853 + 779) was 0.09 for both, whereas the ratios of the three B-I samples with low methylation were 0.04, 0.04, and 0.06). All plants had B‘ pigment levels throughout, suggesting that both alleles in most cells were B‘. Thus, the B‘ epigenetic state can be established before the distinct methylation pattern, which indicates that the methylation reflects the epigenetic state rather than causes it.

Repeat methylation pattern in recombinants

To determine if the number of repeats or the nature of the flanking DNA affects the methylation pattern within the repeats, the same DNA methylation analyses were performed on DNA samples from tissues of 44 plants homozygous for eight different recombinants with recombination breakpoints within the repeats, upstream or downstream of the repeats. We tested the paramutable and paramutagenic alleles of BI||BP-D1 and BI||BP-D2 and four recombinants from previous experiments with B-Peru and B‘ (Stam et al. 2002). These were paramutagenic alleles with seven repeats, B‘||BP-pg1, B‘||BP-pg2, BP||B‘-pg1, and BP||B‘-pg2. With the probes flanking the repeats (Fig. (Fig.4,4, probes A, B, and D), all alleles showed the same methylation pattern previously observed. With the exception of BI||BP-D1, the correlation between the repeat methylation pattern and the epigenetic state was maintained in the recombinants: if the allele was paramutable it had the B-I repeat methylation pattern, if it was paramutagenic it had the B‘ repeat methylation pattern.

The exceptional BI||BP-D1 allele has three tandem repeats and becomes paramutagenic (BI||BP-D1‘) when crossed to B‘. However, the paramutagenic state is unstable as compared with alleles with five (BI||BP-D2‘) or seven (B‘) repeats. We tested 14 samples from 11 BI||BP-D1 individuals in the B-I paramutable state: all showed the B‘ methylation pattern. We tested 11 samples from 8 BI||BP-D1‘ paramutagenic plants, which had been carried with B‘ for two generations: 9 samples showed the B‘ methylation pattern; however, 2 samples showed the B-I methylation pattern. These individuals were also an exception to the rule that there were no methylation differences between leaves and husks: all 8 young leaf samples had the B‘ pattern, whereas 2 of the 3 husk samples showed the B-I pattern. Thus, in two out of three plants examined, the methylation pattern changed during development. One possibility is that fewer repeats reduces the stability of the methylation state. Alternatively, this allele may be in an epigenetic state distinct from B‘ or B-I.

B-I has a more open chromatin structure than B‘

Altered chromatin structure frequently correlates with different epigenetic expression states. Previously we used a DNase I hypersensitivity assay on B‘ and B-I nuclei to compare the chromatin structure around the transcription start site (Chandler et al. 2000). In that experiment, nuclei were isolated from mature sheath tissue of B‘ and B-I plants, which show clear pigment differences. We observed increased hypersensitivity at a site near the transcription start in B-I nuclei relative to B‘ nuclei.

To determine if there were chromatin differences within the region required for paramutation, the same DNA blot was hybridized with several probes within and flanking the repeated region (Fig. (Fig.5).5). The 8.3-kb fragment spanning the repeat (Fig. (Fig.5,5, probe C) was more sensitive to DNase I in B-I nuclei than in B‘ nuclei, suggesting that B-I has a more open chromatin structure in this region compared with B‘ (Fig. (Fig.5).5). To determine if this difference was present outside the repeat region, the blot was stripped and rehybridized with probes B and E, which recognize regions upstream and downstream of the repeated sequences (Fig. (Fig.5).5). Both regions were equivalently resistant to DNase I digestion in B‘ and B-I nuclei. Thus, differential sensitivity between B‘ and B-I was only detected in the region containing the repeats.

Figure 5
The tandem repeat region shows DNase I hypersensitivity in B-I chromatin. Nuclei derived from B‘ and B-I sheaths were treated with increasing amounts of DNase I (0.05, 0.2, 0.5, and 8 units of DNase I/mL; indicated by wedge); naked DNA was treated ...

Newly formed B‘ alleles have a chromatin structure intermediate between B‘ and B-I

New B‘ alleles often retain the B-I methylation pattern, suggesting that the B‘ epigenetic state is established earlier than the methylation changes. To address whether the chromatin structure of new B‘ alleles reflected the parental B-I or the B‘ pattern, another DNase I assay was performed. This experiment used nuclei isolated from young sheaths and leaves surrounding the shoot meristem of B‘, B-I, and new B‘ plants. These young tissues were green because B‘ and B-I were not yet expressed, allowing us to determine the chromatin state prior to transcription of B‘ and B-I.

To map hypersensitive sites within the repeats, we hybridized with probe D1, a single-copy probe located just downstream of the repeats and just upstream of the EcoRI site (Fig. (Fig.6).6). The DNase I sensitivity of the chromatin in the new B‘ repeats was intermediate between that of B‘ and B-I (Fig. (Fig.6).6). The non-DNase I controls indicated that primarily endogenous nucleases rather than the exogenously added DNase I are responsible for the digestions. Nonetheless, comparison of the hybridization pattern of probe B1 hybridizing upstream of the repeats with that of the probe monitoring the repeats makes clear that the endogenous nucleases preferentially act on the repeated region in both B-I and new B‘ nuclei. B‘ nuclei also contained endogenous nucleases because probe A (Fig. (Fig.4,4, unmethylated, gene-like sequence upstream of the repeats) showed similar digestion in both B‘ and new B‘ samples (Fig. (Fig.6),6), yet the repeats in B‘ are not cut by the endogenous nucleases. In B-I nuclei the gene-like region monitored with probe A is slightly more digested with endogenous nucleases. This could mean that in B-I nuclei this region has a more sensitive chromatin structure, or that there were slightly more endogenous nucleases.

Figure 6
The tandem repeats in new B‘ chromatin show an intermediate chromatin structure relative to B-I and B‘. Nuclei from B‘, new B‘, and B-I tissue were treated with increasing amounts of DNase I (0, 0.05, 0.2, 0.4, 0.8, 1.6, ...

The digestion pattern reflecting multiples of ~850 bp seen in B-I and new B‘ chromatin is consistent with nucleases cutting in each repeat (indicated with arrows in Fig. Fig.6).6). In the naked DNA control, the repeats were not cut, but at the concentrations used the DNase I had little activity. If the pattern of cutting within each repeat is caused by the repeat DNA structure, we should see the pattern in naked DNA. We did extensive DNase I titrations on naked genomic and plasmid DNA containing the repeats. We only observed general degradation in naked DNA; no repeat ladder was observed (data not shown), indicating that the pattern of hypersensitive sites in the repeated region is mediated through chromatin structure.

We next tested for hypersensitive sites around the transcription start site by hybridizing with the F probe (Fig. (Fig.6);6); this region was sensitive in B-I and equivalently resistant in B‘ and new B‘ nuclei. Thus, in young, unpigmented tissue, the chromatin structure spanning the transcription start site predicts the transcription state that will occur: B-I will be more transcribed and is more sensitive than B‘ and new B‘.


Using fine-structure recombination mapping, we localized the region within B-I required for enhancer activity, paramutability, and paramutagenicity to a 6-kb region, located ~100 kb upstream of the transcription start site. This region has an identical DNA sequence in B-I and B‘, and consists mostly of seven tandem repeats of a 853-bp sequence unique to this region of the genome. The fact that B-I and B‘ have identical sequences within the region required for paramutation establishes that B‘ paramutation is an epigenetic phenomenon, in spite of the fact that B‘ is extremely stable. Comparison of recombinant alleles with B-I sequences upstream, B-Peru sequences downstream, and one, three, or five repeats, each allele possessing all or most of the B-Peru sequence polymorphisms in the repeat, shows that multiple repeats are required for paramutation and enhancer activity. The recombinant with three repeats showed a reduced stability of the paramutagenic state, whereas the recombinant with five repeats was fully paramutagenic. In contrast, the recombinant with one repeat could not participate in paramutation and had no enhancer activity. Further experiments will be required to determine if multiple repeats are sufficient for both activities, or if the repeats are interacting with other sequences shared between B-I and B-Peru.

The region required for b1 paramutation is unique in the maize genome, which suggests that some structural feature may be shared with other maize loci that undergo paramutation. Inverted and direct repeats of transgenes are often, but not always, correlated with silencing (for reviews, see Henikoff 1998; Wolffe and Matzke 1999; Muskens et al. 2000). In many cases the silencing is posttranscriptional and associated with inverted repeats that produce double-stranded RNA homologous to the transcribed region. Transcriptional gene silencing is usually associated with repeated promoter sequences, but the repeats often span the transcribed region as well. Simply the presence of repeats is not necessary for paramutation, because a single-copy transgene insert can efficiently mediate paramutation in Petunia (Meyer et al. 1993). Although the strength of paramutation at r1 correlates with the number of repeats (Kermicle et al. 1995), the r1 repeats are very different from what we observe at b1. In two paramutagenic r1 alleles, each repeat is at least 13 kb (Eggleston et al. 1995) or 26 kb (Panavas et al. 1999) long, spanning the transcribed region and thousands of base pairs of flanking DNA. In contrast to the repeats correlated with transgene silencing and r1 paramutation, the b1 tandem repeats are 100 kb upstream and share no homology with the promoter-proximal or transcribed regions.

Specific sequences required for paramutation have been identified for paramutation mediated by transgenes containing p1 regulatory sequences. A 1.2-kb p1 fragment, containing an enhancer, with ~78 bp 100% identical to the p1 transcript, and present in a multicopy transgene array, paramutates the endogenous allele (Sidorenko and Peterson 2001). Another enhancer from the p1 promoter in a multicopy transgene array has no paramutagenic activity (Sidorenko and Peterson 2001). It is not known whether the enhancer or transcript homology regions are important contributors nor whether repeats are required. Also, the sequences required for paramutation within the endogenous allele are not known. Paramutation mechanisms are shared at several maize genes because a mutation that prevents paramutation at b1 also prevents paramutation at two other loci (Dorweiler et al. 2000). We suggest that the common denominator is heritable alterations in chromatin structure; there are no common sequences and no clear correlation with the types of repeats among these maize genes.

Comparison of the DNA methylation pattern and chromatin structure of the B-I and B‘ repeat region indicates that in the high-expressing B-I allele the repeats are more hypersensitive to nucleases, but they are more methylated relative to the repeats in the low-expressing B‘ allele. Correlations between r1 paramutation and methylation have been seen, but the correlation is in the opposite direction (the more silenced state is more methylated), and the sequences being monitored have not been shown to be required for paramutagenicity (for review, see Chandler et al. 2000). Possibly at b1, a repressor protein binds unmethylated sites in B‘ repeats. Alternatively, the binding of an activator could cause the strong hypersensitivity of the repeats in B-I. Potentially, the more open chromatin structure of B-I enables the methylation machinery better access, resulting in the methylation increases relative to B‘. Intriguingly, in the three-repeat allele the repeat methylation pattern is unstable, as is the paramutagenic ability of this allele.

Our studies with new B‘ alleles deriving spontaneously from B-I parents show it often takes more than one generation to go from the B-I to the B‘ methylation pattern. This indicates that the repeat methylation pattern at the second Sau3AI site follows the establishment of the epigenetic state rather than causes it. This finding also suggests the increased methylation in B-I does not directly affect the binding of regulatory proteins required for establishment and maintenance of the paramutagenic state. The repeat chromatin structure of new B‘ alleles, assayed with endogenous nucleases, had a sensitivity intermediate to that of B-I and B‘, consistent with the observation that, on average, new B‘ alleles are slightly more pigmented relative to alleles that have been B‘ for multiple generations (Patterson and Chandler 1995). The differing sensitivity between B‘ and B-I is observed in tissues developmentally earlier than the onset of transcription, suggesting that chromatin structure could be the heritable imprint that distinguishes the distinct transcription states. Further studies will be required to determine whether the imprint is mediated by DNA methylation, histone modification, or other chromatin modifications.

This is the first example in plants of regulatory sequences located far upstream of the coding region of a gene. Based on paradigms that derive from studies in mammals, we hypothesize that the enhancer exists in two states, functional (B-I) and nonfunctional (B‘), possibly because it is part of or adjacent to boundary/insulator sequences that exist in distinct states. The LCR (locus control region) of the β-globin loci in mammals contains hypersensitive sites far upstream from the globin genes, some of which function as insulator/boundary elements, others as enhancers (Bell et al. 2001). In mice, an enhancer ~10 kb from H19 and ~100 kb from Igf2 functions on one of the two genes depending on the chromatin structure and methylation status of the ICR region between the two genes (Bell and Felsenfeld 2000; Hark et al. 2000). The CTCF protein binds to the unmethylated and nuclease-hypersensitive ICR, which then functions as a boundary, preventing the enhancer from acting on Igf2. When the ICR is methylated and nuclease-insensitive, the enhancer acts on Igf2. Decreased DNA methylation in the ICR correlates with a more open chromatin structure, opposite of what is observed at B‘ and B-I. A striking similarity between our results and imprinted genes is that imprinted genes are associated with tandem repeat sequences in the differentially methylated ICR. These are generally short (25–120 bp; Barlow 1995), but at the human H19 locus there are several copies of two different ~400-bp repeats (Jinno et al. 1996). The repeat sequences associated with imprinted genes are also unique to one region in the maize genome. Further dissection of the 6-kb B-I fragment using transgenic approaches will address if the enhancer activity and postulated boundary or insulator elements are distinct or overlap. Identification of the proteins encoded by genes required for paramutation (Dorweiler et al. 2000; Hollick and Chandler 2001) may also help to support or refute this model.

The most intriguing aspect of paramutation is that trans interactions alter the transcription state of one allele. Mitotically heritable trans-sensing phenomena in Drosophila provide paradigms (Henikoff and Comai 1998; Wu and Morris 1999), because in plants, mitotically heritable changes in expression are often transmitted to progeny as germ cells differentiate from somatic cells late in development. Many animals and insects set germ cells aside early in development. Pairing may mediate allele interactions, as observed with many trans-sensing phenomena, in which translocations that disrupt somatic pairing disrupt these phenomena. Translocation of B-I or B‘ to other chromosome arms does not disrupt paramutation (Coe 1966), but the translocated regions are large such that meiotic pairing (and potentially somatic pairing) is not disrupted. Although there is no evidence for homolog synapsis in somatic plant cells, transient interactions are all that are needed, because once the B‘ state is established, it is maintained independent of what other allele is on the homologous chromosome. There are also examples of transvection in Drosophila that do not require stable synapsis in somatic tissues (Hopmann et al. 1995). Subnuclear localization could also be involved in paramutation. The heterochromatic insertion at brownD interacts with centromeric heterochromatin, sequestering the wild-type gene into a specific heterochromatic compartment and thus affecting its expression (Csink and Henikoff 1996). Possibly B‘ and B-I are usually in distinct subnuclear domains and allele interactions result in similar compartmentalization.

An alternative model to DNA interactions is DNA/RNA interactions. X-Chromosome inactivation involves the stabilization of an untranslated RNA, which appears to be both necessary and sufficient to confer a chromatin-based mechanism of inactivation on adjacent sequences (Boumil and Lee 2001). Although usually not meiotically heritable, RNA silencing provides another paradigm because it occurs in trans, is mediated by small RNAs, and is often associated with methylation of homologous DNA (for review, see Wolffe and Matzke 1999; Jones et al. 2001). Possibly RNA molecules mediate trans interactions, which establish a chromatin state that is epigenetically maintained.

Paramutation and other examples of homology-dependent gene silencing have been postulated to represent cellular mechanisms for protection against invasive DNA (Matzke et al. 1996; Yoder et al. 1997). The machinery that mediates paramutation may function to identify and maintain chromatin boundaries between genes and nearby repetitive sequences, or to provide an adaptive mechanism for transmitting favorable expression states to progeny (Chandler et al. 2000). Our identification of the sequences required for paramutation, coupled with trans-acting mutants altered in paramutation, provide the tools to fully understand the underlying mechanisms and cellular function in the near future.

Materials and methods

Maize nomenclature

In maize nomenclature, a gene is designated with lowercase italics, and specific alleles are indicated by an allele designation separated from the gene designation with a hyphen. Dominant alleles are indicated by an uppercase gene designation (B-I) and recessive alleles by a lowercase gene designation (b-K55; http://www.agron.missouri.edu/maize_nomenclature.html).

Plant stocks

All plant stocks had dominant functional alleles for all genes encoding anthocyanin biosynthetic enzymes. Stocks containing various b1 alleles maintained in the Chandler laboratory were originally obtained from various sources: B-I (W23 background), B‘, and b-K55 (K55 background) were from E.H. Coe, Jr. (University of Missouri, Columbia); B-Peru (W22 background) was from M.G. Neuffer (University of Missouri); and bm-1 contains a Ds transposon in the B-Peru coding region (Clark et al. 1990).

Recombinant screen

When indicating genotypes, a single allele indicates a homozygote, and in heterozygotes the two alleles are separated by a slash (/). To identify Interval I and II recombinant alleles between B-I and B-Peru, we made use of the phenotypic marker gl2 (glossy2), 19 cM upstream of the b1 transcription start site (http://www.agron.missouri.edu/maps.html), and the tightly linked gene tmp, 0.18 cM upstream of the b1 transcription start site (Stam et al. 2000). Wild-type and mutant glossy2 phenotypes result in different juvenile waxes, visually scored in young seedlings. The gl2 B-I and Gl2 B-Peru stocks were polymorphic within the third intron of the tmp gene, tmp-I and tmp-P, respectively.

To isolate recombinants between B-I and B-Peru, gl2 B-I/Gl2 B-Peru plants were crossed with gl2 b-K55 (colorless seeds, tmp-I) or bm-1 (colorless seeds, tmp-P). To isolate B-I recombinants, colorless seeds from the gl2 b-K55 cross were planted in trays, Gl2 seedlings were transplanted, and the resulting plants were pooled and screened for tmp markers as described (Stam et al. 2002). To isolate B-Peru recombinants that gained enhancer sequences, dark seeds from the gl2 b-K55 cross were planted either directly in the field or in trays, and resulting gl2 seedlings were then transplanted. When mature, plants were scored for their pigment phenotype, and plants with vegetative pigment darker than B-Peru were individually tested for loss of tmp-P. To isolate B-Peru recombinants with breakpoints upstream of the enhancer sequences, dark seeds from the cross with bm-1 were planted in trays. Then gl2 seedlings were transplanted and screened in pools using PCR to specifically amplify a 543-bp B-I fragment between 108.5 and 109 kb upstream of the B-I transcription start site (VC90–VC57). As an internal control, a 963-bp upstream B-Peru fragment (VC81–VC80) was amplified; VC90, 5′-GGATATTCATGCATAACCATTGCTTC-3′; VC57, 5′-GTTGCTACTGTTACACCAGTTTGC-3′; VC81, 5′-CGGCTCCCCTGCGGATCCTC-3′; VC80, 5′-CCATATTCA CGCTAGTCCAAATCG-3′. The PCR conditions were 1× PCR and 1× Q buffer (both QIAGEN), 0.1 mM dNTPs, 0.8 mM of each primer, 1 unit of Taq DNA polymerase, reaction volume of 25 μL; 2 min at 94°C, 35 times (1 min at 94°C, 1 min at 61°C, 1 min 20 sec at 72°C), and then 10 min at 72°C, up to 1 h at 4°C.

Cloning and sequence analyses of b alleles

The comparable region of the B‘ sequence 94–110 kb upstream of the transcription start was PCR-amplified from B-I, B-P, and b-K55 DNA using primers designed from the previously determined B‘ sequence (Stam et al. 2002). Resulting fragments were cloned into pGEM-T Easy (Promega) using standard molecular techniques. For each fragment, plasmid DNA from five to eight clones was pooled and sequenced. To verify that the amplified fragments were physically linked in the genome, fragments overlapping two contiguous fragments were amplified, cloned, and sequenced. All sequences of a specific allele were assembled with FAKtory, a sequence assembly program (http://bcf.arl.arizona.edu/faktory). Accession numbers are AY078064 (b-K55), AF483658 and AF484216 (B-Peru), and AF483657 (B-I).

The repeat sequence was analyzed with a variety of computer programs. No significant similarity to the nr database using TBLASTX (Altschul et al. 1997) was found. For gene annotation we used: GENESCAN using Maize matrix at http://genes.mit.edu/GENSCAN.html; FGENESH using plant model at http://genomic.sanger.ac.uk/gf/gf.shtml; and GeneMark Eukaryotic model at http://opal.biology.gatech.edu/GeneMark/eukhmm.cgi. Two programs, REPuter (http://bibiserv.techfak.uni-bielefeld.de/reputer) and Tandem Repeats Finder (http://c3.biomath.mssm.edu), were used to search for small inverted and direct repeats. To search for ARS we used the FindPatterns program from GCG (University of Wisconsin). To search for MARs we used SMARTest (http://www.genomatix.de/cgi-bin/smartest_pd/smartest.pl).

Genomic DNA analyses

To confirm sequence contigs and examine DNA methylation patterns, ~4 μg of genomic DNA from B‘, B-I, B-Peru, and b-K55 leaves (Dellaporta et al. 1983) was digested with several enzymes and enzyme combinations according to manufacturers’ specifications, size-fractionated by electrophoresis in 0.5× TBE agarose gels, followed by DNA blot analyses (Stam et al. 2002).

DNase I hypersensitivity assay

Nuclei were isolated from ~7-week-old plants, from the most mature, pigmented, outer sheaths (Fig. (Fig.5);5); and from green, inner sheaths and associated leaves surrounding the shoot meristem (Fig. (Fig.6).6). Nuclei were prepared (Steinmuller and Apel 1986) with the buffer modified to contain 5 mM EDTA, 0.5 mM EGTA, and 0.1 mM PMSF. Aliquots of 1 million nuclei from each genotype were incubated with increasing concentrations of DNase I from Roche (Chandler et al. 2000).


We greatly appreciate the technical assistance of Pascal Lennertz and Zahra Mobasher. This research was funded by postdoctoral fellowships from the Human Frontier Science Program to M.S. and the National Science Foundation to J.E.D. (BIR-9626082); by a graduate fellowship from CNPq-Brazil to C.B.; and by grants from the National Science Foundation to V.L.C. (MCB99-82447, MCB96-03638). DNA sequencing was performed by the Arizona GATC.

 The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC section 1734 solely to indicate this fact.


E-MAIL ude.anozira.ga@reldnahc; FAX (520) 621-7186.

Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/gad.1006702.


  • Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 1997;25:3389–3402. [PMC free article] [PubMed]
  • Barlow DP. Gametic imprinting in mammals. Science. 1995;270:1610–1613. [PubMed]
  • Bell AC, Felsenfeld G. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature. 2000;405:482–485. [PubMed]
  • Bell AC, West AG, Felsenfeld G. Insulators and boundaries: Versatile regulatory elements in the eukaryotic genome. Science. 2001;291:447–449. [PubMed]
  • Boumil RM, Lee JT. Forty years of decoding the silence in X-chromosome inactivation. Hum Mol Genet. 2001;10:2225–2232. [PubMed]
  • Brink RA. A genetic change associated with the R locus in maize which is directed and potentially reversible. Genetics. 1956;41:872–890. [PMC free article] [PubMed]
  • ————— Paramutation. Annu Rev Genet. 1973;7:129–152. [PubMed]
  • Cavalli G, Paro R. The Drosophila Fab-7 chromosomal element conveys epigenetic inheritance during mitosis and meiosis. Cell. 1998;93:505–518. [PubMed]
  • Chandler VL, Eggleston WB, Dorweiler JE. Paramutation in maize. Plant Mol Biol. 2000;43:121–145. [PubMed]
  • Clark JK, Chandler VL, Neuffer MG. Characterization of two Ds mutants of B-Peru. Maize Genet Coop Newsletter. 1990;64:59–60.
  • Coe EH., Jr A regular and continuing conversion-type phenomenon at b locus in maize. Maydica. 1959;24:49–58. [PMC free article] [PubMed]
  • ————— The properties, origin and mechanism of conversion-type inheritance at the b locus in maize. Genetics. 1966;53:1035–1063. [PMC free article] [PubMed]
  • Colot V, Maloisel L, Rossignol JL. Interchromosomal transfer of epigenetic states in Ascobolus: Transfer of DNA methylation is mechanistically related to homologous recombination. Cell. 1996;86:855–864. [PubMed]
  • Csink AK, Henikoff S. Genetic modification of heterochromatic association and nuclear organization in Drosophila. Nature. 1996;381:529–531. [PubMed]
  • Dellaporta SL, Wood J, Hicks JB. A plant DNA mini preparation. Plant Mol Biol Rep. 1983;1:19–21.
  • Dorweiler JE, Carey CC, Kubo KM, Hollick JB, Kermicle JL, Chandler VL. Mediator of paramutation1 is required for establishment and maintenance of paramutation at multiple maize loci. Plant Cell. 2000;12:2101–2118. [PMC free article] [PubMed]
  • Eggleston W, Alleman M, Kermicle J. Molecular organization and germinal instability of R-stippled maize. Genetics. 1995;141:347–360. [PMC free article] [PubMed]
  • Garrick D, Fiering S, Martin DI, Whitelaw E. Repeat-induced gene silencing in mammals. Nat Genet. 1998;18:56–59. [PubMed]
  • Grewal SI, Klar AJ. Cell 86: 95–101. 1996. Chromosomal inheritance of epigenetic states in fission yeast during mitosis and meiosis. [PubMed]
  • Hark AT, Schoenherr CJ, Katz DJ, Ingram RS, Levorse JM, Tilghman SM. CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature. 2000;405:486–489. [PubMed]
  • Henikoff S. Conspiracy of silence among repeated transgenes. Bioessays. 1998;20:532–535. [PubMed]
  • Henikoff S, Comai L. Trans-sensing effects: The ups and downs of being together. Cell. 1998;93:329–332. [PubMed]
  • Hollick JB, Chandler VL. Genetic factors required to maintain repression of a paramutagenic maize pl1 allele. Genetics. 2001;157:369–378. [PMC free article] [PubMed]
  • Hollick JB, Patterson GI, Coe EH, Jr, Cone KC, Chandler VL. Allelic interactions heritably alter the activity of a metastable maize p1 allele. Genetics. 1995;141:709–719. [PMC free article] [PubMed]
  • Hopmann R, Duncan D, Duncan I. Transvection in the iab-5,6,7 region of the bithorax complex of Drosophila: Homology independent interactions in trans. Genetics. 1995;139:815–833. [PMC free article] [PubMed]
  • Jinno Y, Sengoku K, Nakao M, Tamate K, Miyamoto T, Matsuzaka T, Sutcliffe JS, Anan T, Takuma N, Nishiwaki K, et al. Mouse/human sequence divergence in a region with a paternal-specific methylation imprint at the human H19 locus. Hum Mol Genet. 1996;5:1155–1161. [PubMed]
  • Jones L, Ratcliff F, Baulcombe DC. RNA-directed transcriptional gene silencing in plants can be inherited independently of the RNA trigger and requires Met1 for maintenance. Curr Biol. 2001;11:747–757. [PubMed]
  • Kermicle JL, Eggleston WB, Alleman M. Organization of paramutagenicity in R-stippled maize. Genetics. 1995;141:361–372. [PMC free article] [PubMed]
  • Krebbers E, Hehl R, Piotrowiak R, Lonning W, Sommer H, Saedler H. Molecular analysis of paramutant plants of Antirrhinum majus and the involvement of transposable elements. Mol Gen Genet. 1987;214:325–327. [PubMed]
  • Martienssen R. Epigenetic phenomena: Paramutation and gene silencing in plants. Curr Biol. 1996;6:810–813. [PubMed]
  • Matzke M, Matzke A, Eggleston WB. Paramutation and transgene silencing: A common response to invasive DNA? Trends Plant Sci. 1996;1:382–388.
  • Maundrell K, Hutchison A, Shall S. Sequence analysis of ARS elements in fission yeast. EMBO J. 1988;7:2203–2209. [PMC free article] [PubMed]
  • McClintock B. The control of gene action in maize. Brookhaven Symp Biol. 1965;18:162–184.
  • Meyer P, Heidmann I, Niedenhof I. Differences in DNA-methylation are associated with a paramutation phenomenon in transgenic petunia. Plant J. 1993;4:89–100. [PubMed]
  • Michalowski SM, Allen GC, Hall GE, Jr, Thompson WF, Spiker S. Characterization of randomly-obtained matrix attachment regions (MARs) from higher plants. Biochemistry. 1999;38:12795–12804. [PubMed]
  • Muskens MW, Vissers AP, Mol JN, Kooter JM. Role of inverted DNA repeats in transcriptional and post-transcriptional gene silencing. Plant Mol Biol. 2000;43:243–260. [PubMed]
  • Panavas T, Weir J, Walker EL. The structure and paramutagenicity of the R-marbled haplotype of Zea mays. Genetics. 1999;153:979–991. [PMC free article] [PubMed]
  • Patterson GI, Chandler VL. Timing of b locus paramutation. Maydica. 1995;40:35–41.
  • Patterson GI, Thorpe CJ, Chandler VL. Paramutation, an allelic interaction, is associated with a stable and heritable reduction of transcription of the maize b regulatory gene. Genetics. 1993;135:881–894. [PMC free article] [PubMed]
  • Patterson GI, Kubo KM, Shroyer T, Chandler VL. Sequences required for paramutation of the maize b gene map to a region containing the promoter and upstream sequences. Genetics. 1995;140:1389–1406. [PMC free article] [PubMed]
  • Radicella JP, Brown D, Tolar LA, Chandler VL. Allelic diversity of the maize B regulatory gene: Different leader and promoter sequences of two B alleles determine distinct tissue specificities of anthocyanin production. Genes & Dev. 1992;6:2152–2164. [PubMed]
  • Selinger DA, Lisch D, Chandler VL. The maize regulatory gene B-Peru contains a DNA rearrangement that specifies tissue-specific expression through both positive and negative promoter elements. Genetics. 1998;149:1125–1138. [PMC free article] [PubMed]
  • Sidorenko LV, Peterson T. Transgene-induced silencing identifies sequences involved in the establishment of paramutation of the maize p1 gene. Plant Cell. 2001;13:319–335. [PMC free article] [PubMed]
  • Stam M, Lavin T, Chandler V. Npi402 and ncsu1 are identical; inra1 (tmp) maps upstream of the b promoter. Maize Genet Coop Newsletter. 2000;74:66–67.
  • Stam, M., Belele, C.L., Ramakrishna, W., Dorweiler, J.E., Bennetzen, J.L., and Chandler, V.L. 2002. The regulatory region required for B′ paramutation and expression are located far upstream of the maize b1 transcribed sequences. Genetics (In press). [PMC free article] [PubMed]
  • Steinmuller K, Apel K. A simple and efficient procedure for isolating plant chromatin which is suitable for studies of DNase I-sensitive domains and hypersensitive sites. Plant Mol Biol. 1986;7:87–94. [PubMed]
  • van Houwelingen A, Souer E, Mol J, Koes R. Epigenetic interactions among three dTph1 transposons in two homologous chromosomes activate a new excision-repair mechanism in petunia. Plant Cell. 1999;11:1319–1336. [PMC free article] [PubMed]
  • van West P, Kamoun S, van 't Klooster JW, Govers F. Internuclear gene silencing in Phytophthora infestans. Mol Cell. 1999;3:339–348. [PubMed]
  • Vaucheret H, Fagard M. Transcriptional gene silencing in plants: Targets, inducers and regulators. Trends Genet. 2001;17:29–35. [PubMed]
  • Wolffe AP, Matzke MA. Epigenetics: Regulation through repression. Science. 1999;286:481–486. [PubMed]
  • Wu CT, Morris JR. Transvection and other homology effects. Curr Opin Genet Dev. 1999;9:237–246. [PubMed]
  • Yoder JA, Walsh CP, Bestor TH. Cytosine methylation and the ecology of intragenomic parasites. Trends Genet. 1997;13:335–340. [PubMed]

Articles from Genes & Development are provided here courtesy of Cold Spring Harbor Laboratory Press
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...