Plant Cell EPICENTRE Biotechnologies
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online August 23, 2002; 10.1105/tpc.004028

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
14/9/2251    most recent
tpc.004028v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (14)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.
Agricola
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.
The Plant Cell, Vol. 14, 2251-2264, September 2002, Copyright © 2002,
American Society of Plant Biologists

Suppression of Transgene Silencing by Matrix Attachment Regions in Maize

A Dual Role for the Maize 5' ADH1 Matrix Attachment Region

Cory Brouwer1,a,b, Wesley Bruceb, Sheila Maddockb, Zoya Avramova2,c and Ben Bowen3,a,b

a Department of Zoology and Genetics, Iowa State University, Ames, Iowa 50011
b Pioneer Hi-Bred International, Inc., Johnston, Iowa 50131
c School of Biological Sciences, University of Nebraska, Lincoln, NE 68588

2 To whom correspondence should be addressed. E-mail zavramova2{at}unl.edu; fax 402-472-2083


   Abstract
TOP
 Abstract
INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
 References
 
Matrix attachment regions (MARs) are DNA sequences that bind an internal nuclear network of nonhistone proteins called the nuclear matrix. Thus, they may define discrete gene-containing chromatin loops in vivo. We have studied the effects of flanking transgenes with MARs on transgene expression levels in maize callus and in transformed maize plants. Three MAR elements, two from maize (Adh1 5' MAR and Mha1 5' MAR) and one from yeast (ARS1), had very different effects on transgene expression that bore no relation to their affinity for the nuclear matrix in vitro. In callus, two of the MAR elements (Adh1 5' MAR and ARS1) reduced transgene silencing but had no effect on the variability of expression. In transgenic plants, Adh1 5' MAR had the effect of localizing {beta}-glucuronidase expression to lateral root initiation sites. A possible model accounting for the function of Adh1 5' MAR is discussed.


   INTRODUCTION
TOP
 Abstract
 INTRODUCTION
RESULTS
 DISCUSSION
 METHODS
 References
 
Gene expression is influenced not only by the presence or absence of nearby transcription factors but also by the structure of surrounding chromatin. When chromatin competes with the transcriptional machinery, stochastic expression or epigenetic effects may result. To understand fully the regulation of gene expression inside the nucleus, it is necessary to consider the different levels of structural organization of DNA in chromatin and the factors involved in establishing, maintaining, and modifying these structures.

Three levels of chromatin compaction (the 11-, 30-, and 300-nm fibers, respectively) are thought to be required for packaging the genome inside the nucleus. The third level of compaction results from the folding of the 30-nm fiber into loops of various sizes attached to the nuclear matrix. In addition, dense masses of heterochromatin can be distinguished cytologically from euchromatin in interphase nuclei. Variable repression of transcription often is associated with heterochromatin and frequently manifests itself as position effect variegation (reviewed by Wakimoto, 1998Go; Wallrath, 1998Go). In plants, variable expression may be observed among independent transformation events expressing the same transgene from different positions within the genome.

Despite the fact that little is known about both the structure and the protein composition of heterochromatin in plants, gene-silencing phenomena often are attributed to heterochromatin effects (Avramova, 2002Go). The genomes of many commercially important crop plants contain large amounts of repetitive DNA. In maize, ~80% of the genome is composed of highly repetitive DNA that is organized either as morphologically visible knobs (Peacock et al., 1981Go; Ananiev et al., 1998Go; Fransz et al., 2000Go) or as large, densely methylated blocks in the intergenic space (SanMiguel et al., 1996Go).

In maize, genes may be separated from one another by highly repetitive DNA stretches of up to 80 kb (Avramova et al., 1996Go; Tikhonov et al., 1999Go). In closely related grass genomes, orthologous genes with high sequence homology are separated by intergenic regions that differ in sequence from those found in maize (Chen et al., 1997Go; Tikhonov et al., 1999Go). Given the importance of the genomic context for the function of a gene, these findings suggest an apparent paradox: orthologous genes in related species function within very different chromosomal settings.

One model accounting for this apparent paradox proposes that genes and blocks of repetitive DNAs might exist in different, structurally separated nuclear compartments (Lamond and Earnshaw, 1998Go; Cockell and Gasser, 1999Go) and that mislocation alters expression. Therefore, genes may have "anchors" that position them in the spatial architecture of the nucleus (Flavell, 1994Go). This anchoring function may be mediated by matrix-attachment regions (MARs). MARs are defined operationally as DNA sequences that bind preferentially to the proteins of the nuclear matrix. In eukaryotic genomes, MARs typically are localized at the borders of gene domains, implicating them in the formation of individual loops of higher order chromatin structure. In plants, these putative gene loops are relatively small (3 to 10 kb) (Breyne et al., 1992Go; van der Geest et al., 1994Go; Chinn et al., 1996Go; van Drunen et al., 1997Go; Avramova et al., 1998Go).

Previously, we reported that the intergenic stretches of repetitive DNA in the maize Adh1 region are segregated into topologically sequestered units and that the microcolinearity of gene composition between grass species is mirrored by the similar placement of orthologous genes between MARs (Avramova et al., 1995Go, 1998Go). Thus, the positioning of orthologous genes within their respective structural domains appears to be conserved in evolution (Avramova et al., 1998Go; Tikhonov et al., 2000Go). This is consistent with the ideas that endogenous genes within loops are in structural domains that separate them from neighboring nongenic or genic sequences and that transgene position may influence the availability of specific factors that affect transgene expression. Flanking transgenes with MARs could decrease the probability of silencing, reduce the variability of expression, or both.

Plant MARs have been implicated in a variety of gene expression phenomena (Breyne et al., 1992Go; Allen et al., 1993Go; Schöffl et al., 1993Go; Mlynarova et al., 1994Go, 1996Go; van der Geest et al., 1994Go; Iglesias et al., 1997Go; Matzke and Matzke, 1998Go; Ülker et al., 1999Go; Vain et al., 1999Go), suggesting that MARs are cis-regulatory elements. In other transgenic studies, MARs have failed to play a role as gene effectors (L. Sidorenko, W. Bruce, S. Maddock, and T. Peterson, unpublished data). This indicates that the role of MARs may be more elusive and complicated than initially expected. In particular, the effects of flanking a transgene with MARs appear to be highly variable in plants. One reason could be that in many cases, the samples studied have been too small to describe in full the expression level distributions among independent transformants. We suspected that a systematic study with larger samples might help elucidate the underlying basis of this reported variability.

In this study, we used maize cell cultures that could be transformed at high frequency to generate sufficient numbers of independent transformants to allow us to examine the underlying gene expression distributions without bias. Other factors that have complicated studies involving MARs are variability in transgene copy number and homology-dependent silencing effects observed in transformants with high transgene copy numbers. Here, transformations were performed under conditions in which predominantly low-copy inserts were introduced, minimizing the impact of these factors.

The effects on transgene expression of a heterologous MAR from yeast and of two endogenous maize MARs were studied. When enough transformants were analyzed, the distributions of transgene expression levels were found to be bimodal in every case, which complicated statistical analysis. We found that two of the MARs reduced the probability of transgene silencing, without affecting the variability, whereas the third had no effect on transgene expression. There was no relationship between the strength of binding to the nuclear matrix in vitro and the effects on transgene expression in vivo. Finally, the effect of the maize Adh1 5' MAR on transgene expression was examined in regenerated maize plants. The results suggested that this MAR may display both activating and repressing roles, functionally relating it to other recognized dual-function elements such as two immunoglobulin MARs ({kappa} and µ) and the polycomb response elements described in animal studies (reviewed by Lyko and Paro, 1999Go).


   RESULTS
TOP
 Abstract
 INTRODUCTION
 RESULTS
DISCUSSION
 METHODS
 References
 
Affinity of MARs for the Maize Nuclear Matrix in Vitro
Three MARs were used in this study: the yeast autonomously replicating sequence ARS1, which binds to the yeast nuclear matrix (Amati and Gasser, 1988Go) and affects transgene expression in tobacco (Allen et al., 1993Go), and two MARs associated with the 5' regions of two maize genes, Adh1 and Mha1, which encode alcohol dehydrogenase and a H+-ATPase, respectively.

The relative affinity of each MAR for the maize nuclear matrix was tested by in vitro binding assay. As shown in Figure 1 , the MAR located at the 5' end of Mha1 displayed the highest affinity under the assay conditions, and the yeast ARS1 sequence bound less efficiently than either of the maize MARs (Figures 1A and 1B). In the presence of a large excess of Escherichia coli competitor DNA (150 µg/mL), ~10 times more Mha1 5' MAR remained bound compared with Adh1 5' MAR (Figure 1B).



View larger version (31K):
[in this window]
[in a new window]
  		Figure 1. Binding Assay Results for the Three MARs Used in this Study.

Vector and MAR fragments are labeled v and M, respectively. Labeled input DNA, without nuclear matrices, is shown in lane i. Adjacent lanes show the DNA fragments recovered from the matrix fraction in the presence of various concentrations of competitor DNA.

(A) Yeast ARS1 sequences (right) bound maize matrices with a lower affinity than the maize Adh1 5' MAR (left). In the presence of a 1000-fold molar excess of competitor DNA (100 µg/mL), 3% of yeast ARS1 and 10% of Adh1 5' MAR remained matrix bound.

(B) Mha1 5' MAR (left) displayed the highest affinity for binding to the matrix. In the presence of a 1500-fold molar excess of competitor (150 µg/mL), 30% of Mha1 5' MAR remained bound, whereas under the same conditions, only 3.5% of Adh1 5' MAR (right) remained bound.

(C) Binding of Mha1 5' MAR in the presence of different concentrations of E. coli DNA or 50 ng/mL unlabeled Adh1 5' MAR (left). Nine percent of Mha1 5' MAR remained bound to the matrix in the presence of 50 ng/mL unlabeled Adh1 5' MAR. When the binding of labeled Adh1 5' MAR was challenged with different concentrations of unlabeled Mha1 5' MAR (right), 25 ng/mL of the specific competitor left only 3% of Adh1 5' MAR bound to the matrix, and 50 ng/mL unlabeled Mha1 5' MAR completely abolished the matrix binding of Adh1 5' MAR.

 
Inhibition of Adh1 5' MAR binding by the addition of cold Mha1 5' MAR and vice versa (Figure 1C) confirmed the relative matrix affinity for each MAR: 50 ng/mL cold Mha1 5' MAR completely abolished Adh1 5' MAR binding, whereas 50 ng/mL cold Adh1 5' MAR left ~9% of labeled Mha1 5' MAR bound to the matrices (Figure 1C). This finding indicates that the highest affinity for the matrix is displayed by Mha1 5' MAR. Because Adh1 and Mha1 5' MARs can compete with each other, some of the components required for the matrix binding of Adh1 5' MAR also must be involved in binding to Mha1 5' MAR.

MAR Elements Do Not Affect Expression before Integration in Black Mexican Sweet Maize Cells
Before investigating the effects of MARs on transgene expression, we assayed the transient expression of a luciferase (LUC) reporter gene in Black Mexican Sweet (BMS) maize suspension cells. Because MARs often colocalize with gene regulatory elements (Gasser and Laemmli, 1987Go; Bonifer et al., 1994Go; Phi-Van and Strätling, 1996Go), it was important to separate effects caused by enhancers/silencers (eventually contained in the transformation constructs) from effects resulting from MAR activity. Effects on transgene expression displayed only after its stable integration into the genome might reflect a MAR role in affecting the chromatin structure at the integration site.

The plasmid constructs used in the study are shown in Figure 2A . Each LUC construct was introduced into BMS cells in combination with a fixed level of a 35S::{beta}-glucuronidase (GUS) plasmid (PHP264), and LUC expression was normalized to GUS expression. The expression levels of constructs in which 35S::LUC was flanked by each of the three MARs against a control vector lacking MARs (PHP1528) were compared.



View larger version (16K):
[in this window]
[in a new window]
  		Figure 2. Vectors Used in the Transformation and Transient Assay Studies.

(A) The vectors used for BMS transformations. In the first construct, for PHP264, the open reading frame (ORF) equals GUS; for PHP1528, ORF equals LUC; and for PHP3528, ORF equals BAR. In the second construct, for PHP5438, MAR equals ARS1 and ORF equals LUC; for PHP5456, MAR equals ARS1 and ORF equals BAR; for PHP6248, MAR equals Adh1 5' and ORF equals LUC; for PHP6344, MAR equals Adh1 5' and ORF equals BAR; for PHP6486, MAR equals Mha1 5' and ORF equals LUC; and for PHP6487, MAR equals Mha1 5' and ORF equals BAR. The third construct is PHP6086 containing the Rsyn7 promoter. The fourth construct is PHP7917 containing the Rsyn7 promoter and the Adh1 5' MAR elements.

(B) Effects of MARs on the transient expression of 35S::LUC. Five replicas for each treatment were assayed for LUC and GUS expression. The 35S::LUC vectors PHP1528, PHP5483 (ARS1), PHP6248 (Adh1), and PHP6486 (Mha1) were mixed 5:1 with the 35S::GUS vector PHP264. Relative levels of gene expression were calculated by normalizing LUC to the level of GUS expression. An F test indicated no significant difference between means (P = 0.21). LU, light units.

 
None of the three MAR elements tested had any significant effect on the expression level of LUC compared with the control vector without MARs (Figure 2B). Thus, none of the tested MAR-containing fragments appeared to carry significant enhancer/silencer activity when expressed transiently.

Expression Levels of Stably Integrated Transgenes in BMS
To assay the effects of each MAR on the expression of stably integrated transgenes, we used the BAR gene as a selectable marker to obtain a population of independently transformed BMS calli. 35S::BAR and 35S::LUC genes were introduced in trans on separate constructs. The transformation procedure generating independent transgenic lines has been used routinely (Grotewold et al., 1998Go; Bruce et al., 2000Go). In the present study, transformation under these conditions generated stably transformed lines representing individual insertion events. DNA gel blot hybridization analysis of a random sample of stably transformed lines displayed unique integration patterns for each event with more than one inserted copy (Figures 3A and 3B) .



View larger version (62K):
[in this window]
[in a new window]
  		Figure 3. DNA Gel Blot Analysis and Copy-Number Frequency Distribution for BMS Transgenic Lines.

(A) A random sample of stably transformed lines representing independent transformation events. The lines shown were from two filters from the same bombardment experiment. Lanes marked M contained molecular mass marker fragments. Lanes 1C and 5C show hybridization signals from one and five transgene copies, respectively.

(B) Copy numbers of random samples of BASTA-resistant BMS calli transformed with low-dose DNA. Lines with predominantly one or two copy numbers per genome are shown in lanes 1 to 6. High-dose DNA resulted in multiple insertions, which are shown in lanes 7 to 12. Lanes 1C and 5C are as described in (A).

 
Regulating the DNA dose, low-copy-number (lanes 1 to 6) or multiple-copy-number (lanes 7 to 12) transgene insertions could be achieved (Figure 3B). These experiments were performed under previously established conditions known to generate independently transformed lines with low-copy-number insertions (see Methods). In the control treatments, neither the reporter nor the selectable marker contained MARs, whereas in the test treatments, both the reporter and the selectable marker were flanked with one of the three MARs. LUC expression levels in BMS calli transformed with each of the four vector combinations are displayed in Figure 4A . In each case, >50 LUC-expressing events were analyzed per treatment. Reproducibly, the transgene expression levels were distributed bimodally. With smaller sample sizes (20 to 30), the expression level distributions were not significantly different from log normal (see Discussion).



View larger version (25K):
[in this window]
[in a new window]
  		Figure 4. LUC Expression in BMS Transformants.

(A) Transgene expression level distributions are shown in the histograms. The y axis corresponds to LUC expression on a logarithmic scale [ln(LU/µg protein)], where LU indicates light units. Above each histogram is an outlier box plot. The box represents the interquartile range, or the difference between the 25th and 75th percentiles. The "whiskers" (the horizontal lines outside of the box) represent the range (computed as 150% of the interquartile range). The vertical line inside the box represents the median, and the diamond represents the mean. The bracket underneath the box identifies the most dense 50% of all observations. All distributions were bimodal, and the shape of the distributions is similar for no-MARs and Mha1, indicating that Mha1 had no effect on expression. ARS1 and Adh1 both increased the expression level and shifted the majority of expressing events to the higher expressing peak of the bimodal distribution.

(B) Cumulative distribution function (CDF) graphs. Adh1 5' MAR and ARS1 both have distributions significantly different from that of the no-MAR control. Mha1 5' MAR has the same distribution as the no-MAR control.

 
Adh1 5' MAR and ARS1 Reduce Transgene Silencing in BMS
Because LUC expression levels were not distributed normally among transformants, even after log transformation, the data could not be analyzed by analysis of variance. Instead, we used a nonparametric approach that compares cumulative distributions.

Although differences in the LUC expression level distributions between the various MAR elements and the control can be seen by visual inspection, the histograms shown in Figure 4A include data only from LUC-expressing events. By contrast, the cumulative distribution function (see Methods) effectively summarizes the data from both expressing and nonexpressing events. This allows quick determination of the percentile expressing at a certain level. However, instead of plotting against expression level on the x axis, we plotted against the quantile so that we could make comparisons across treatments. The cumulative distribution function graphs (Figure 4B) reveal that flanking LUC with either Adh1 5' MAR or ARS1 increased the proportion of events that express the LUC transgene, whereas flanking LUC with Mha1 5' MAR had no effect on the expression level distribution.

To examine these effects further, the LUC-expressing events were subdivided into high and low expressors by selecting a threshold between the two peaks of each bimodal distribution shown in Figure 4A. A single expression value [4 ln(light units/µg protein)] was chosen arbitrarily and used as the cutoff between high- and low-expressing events. The same cutoff value was applied consistently for all treatments and for all data sets. The graph in Figure 5 clearly indicates that flanking the reporter with ARS1 or Adh1 5' MAR increased the frequency of higher expressing cells. By contrast, flanking the reporter with Mha1 5' MAR had no detectable effect.



View larger version (13K):
[in this window]
[in a new window]
  		Figure 5. N-fold difference graph of LUC expression in BMS.

BASTA-resistant transformants were classified as nonexpressors (black bars), low expressors (gray bars), or high expressors (white bars), as described in Results. A single expression value [4 ln(LU/µg protein)] was used as the cutoff for all data sets. Values indicate the n-fold difference from the control without MARs in numbers of transformants in each category. The main effect of ARS1 and Adh1 MARs was to increase the number of high-expressing events at the expense of nonexpressors and low expressors. Mha1 had no effect on expression levels.

 
ARS1 increased the number of high-expressing events primarily at the expense of nonexpressing events, whereas Adh1 5' MAR nearly tripled the number of high-expressing events at the expense of both nonexpressing and low-expressing events (Figure 5). Both ARS1 and Adh1 5' MAR decreased the number of nonexpressors approximately twofold. This finding indicates that these two MARs increased average expression by converting a population of silent cells into cells expressing the reporter gene (i.e., they reduced transgene silencing).

MARs May Increase Transgene Expression Levels but Do Not Influence the Variability of Expression
The net result of reducing transgene silencing is an increase in the average expression of LUC (Table 1). When the LUC transgene was flanked with ARS1, Adh1 5' MAR, or Mha1 5' MAR, the average level of LUC expression was increased 5.8-, 26.8-, or 1.5-fold, respectively. Although Mha1 5' MAR had a greater affinity for the nuclear matrix than Adh1 5' MAR, its effect on transgene expression was significantly lower than that of Adh1 5' MAR. In addition, yeast ARS1, the MAR with the lowest binding affinity for the maize nuclear matrix proteins in vitro, increased the average LUC expression levels more than Mha1 5' MAR. Thus, the affinity for the matrix per se cannot serve as a reliable indicator for predicting the effects of MARs on transgene expression in vivo.


View this table:
[in this window]
[in a new window]
  		Table 1. Summary of Results in BMS Callus

 
The increase in average LUC expression associated with ARS1 and Adh1 5' MAR resulted from a significant increase in the number of high expressors (Table 1, Figure 5). However, neither MAR affected significantly the range of expression seen among the transformants (Table 1), indicating that the presence of MARs flanking the transgene did not decrease the variability of expression. Thus, neither MAR influenced the level of variation of transgene expression to any measurable extent.

Effects of Maize Adh1 5' MAR on GUS Expression in Transgenic Maize Plants
The effect of Adh1 5' MAR was explored further in regenerated transgenic maize plants using the GUS gene as a reporter. The reporter gene was cloned in a vector under the control of a synthetic promoter, Rsyn7 (see Methods), flanked by Adh1 5' MAR (Figure 2) and introduced with Ubi::PAT, also flanked by Adh1 5' MAR. The control constructs did not have MARs flanking Rsyn7::GUS or Ubi::PAT. Mature T1 plants grown in the greenhouse were analyzed for GUS expression enzymatically and by histochemical staining.

Eight Adh1 5' MAR transformants and eight non-MAR control transformants were analyzed. In seven of eight test plants transformed independently with Rsyn7::GUS flanked by Adh1 5' MAR, GUS expression in all tissues was lower than that in control plants transformed with Rsyn7::GUS without MARs. In root tissue of constructs without MAR, the relative activity (GUS assay units normalized to total soluble protein) ranged from 1223 to 2,086,225, with an average of 329,561 units. For Adh1 5' MAR transformants, the respective units were 339 to 7237, with an average of 3789. The probability for the means of the non-MAR controls and Adh1 5' MAR transformants by t test was P > 0.05. This finding indicates that GUS activity in the two types of transformants was significantly different, being much lower in plants expressing GUS from the construct flanked by the MARs.

Histochemical analysis of transformed roots provided an even more striking picture. Rsyn7 functions as a constitutive promoter (B.A. Bowen, W.B. Bruce, G. Lu, L.E. Sims, and L.A. Tagliani [2000], U.S. patent 6,072,050). Accordingly, most of the GUS staining in control plants was observed throughout the plant, including all cell types in the roots (Figure 6A) . By contrast, plants transformed with Rsyn7::GUS flanked by Adh1 5' MAR exhibited GUS expression only in patches of epidermal and cortical cells localized at lateral root-emerging sites (Figures 6B and 6C). The restriction of GUS expression to small regions of root cells would explain the lower level of overall GUS expression activity established above using the biochemical approach.



View larger version (52K):
[in this window]
[in a new window]
  		Figure 6. Plants Stably Transformed with the Rsyn7-Driven GUS Gene.

Histochemical analysis of GUS expression in roots of transgenic maize plants transformed with Rsyn7::GUS unflanked (A) or flanked ([B] and [C]) with Adh1 5' MAR. The expression patterns in (B) and (C) show GUS activity primarily at the sites of lateral root emergence (arrowheads).

 
The timing of expression coincided with the earliest stages of emergence of the lateral roots and was restricted exclusively to cells in the parent root. No significant expression was observed in the meristems of growing lateral roots. Low levels of GUS expression also were seen in ear tissue, confined primarily to the inner and outer glumes (data not shown).


   DISCUSSION
TOP
 Abstract
 INTRODUCTION
 RESULTS
 DISCUSSION
METHODS
 References
 
DNA Elements Involved in Chromatin-Dependent Gene Regulation
The complex and inconsistent behavior of MARs on transgene expression suggests that our understanding of the nature of MAR effects must be incomplete. In addition to enhancers and promoters, various other classes of cis-DNA elements may affect a nearby gene's expression by influencing how a given chromosomal environment is established, maintained, and modified. Such DNA sequences may direct nuclear sublocalization, may nucleate or propagate specific forms of higher order chromatin structures, and may define the boundaries of "open" and "closed" domains. Presumably, a single segment of DNA containing all of these elements would be unaffected by chromosomal position, shield a promoter from nearby regulatory elements, and promote copy number–dependent gene expression.

Some reports have suggested elements that could mediate any or all of these effects in any context. These include special chromatin structures (scs and scs'), insulators, locus control regions, polycomb response elements, and MARs (reviewed by West et al., 2002Go). Other studies have indicated that the functions of these elements are elusive, more complicated than initially expected, and sometimes even antagonistic (Cai and Levine, 1995Go; Avramova and Tikhonov, 1999Go).

Because the majority of transgenic events involve insertions into uncharacterized chromosomal locations, the role of the transgenic DNA must be evaluated carefully. The modular structure and frequently overlapping cis-elements of promoter regions suggest that the effects observed (or not observed) at their native locations are a result of their combined influences (Nabirochkin et al., 1998Go). Evidently, MARs may exhibit interactions with neighboring sequences. MARs collaborate with enhancers for the transcriptional activation of immunoglobulin and {beta}-globin genes to generate an extended domain of accessible chromatin (Forrester et al., 1994Go; Jenuwein et al., 1997Go). Likewise, the 5' and 3' MARs from the tomato Heat Shock Cognate 80 (HSC80) gene are essential for regulated expression, but only when combined with transgenes that harbor introns of HSC80. The absence of either MAR or any of the introns reduces or eliminates HSC80 expression (Chinn and Comai, 1996Go; Chinn et al., 1996Go). Finally, the only study to explore the relationship between the stability of promoter-driven transgene expression and its integration site in plants established that some stably expressing lines were correlated with integrations at natural matrix binding sites (Iglesias et al., 1997Go).

These facts provide a context for interpreting the differences observed with the Adh1 and Mha1 MARs in this work. All constructs used in the study contained the first intron of maize Adh1, which is known to enhance maize transgene expression (Callis et al., 1987Go). Therefore, cooperation between Adh1 5' MAR and the first intron of Adh1 could be important for the Adh1 5' MAR effect in vivo and less so for the Mha1 5' MAR effect. At any rate, binding by proteins that compete for these two MARs in vitro (Figure 1) must be insufficient to mediate transgene silencing in vivo.

The nature of the Adh1 5' MAR DNA fragment used to flank the reporter gene constructs in these studies has been reported (Avramova and Bennetzen, 1993Go; Tikhonov et al., 2000Go). The entire fragment displayed matrix binding activity, and splitting the region significantly diminished its affinity for the matrix. The adjacent downstream region that did not display any matrix binding capacity defined the 3' boundary of the region that we called MAR.

The non-MAR region, which contains recognized Adh1 gene regulatory elements (Walker et al., 1987Go; Paul and Ferl, 1991Go), was not included in the fragment used to flank the constructs in the experiments described here. However, the presence of cryptic, cis-regulatory elements within the MAR region are not excluded; in fact, they could explain the effects observed (as discussed above and below). The fact that MAR fragments affected LUC expression in BSM cells only after stable integration in the genome suggests that MARs function as regulatory elements only after they bind to the nuclear matrix.

Does a Higher Affinity for the Matrix in Vitro Predict a Higher Effect in Vivo?
If MARs exert their effect through binding to the nuclear matrix, then MARs with a higher matrix affinity might be expected to have more pronounced effects on transgene expression. Allen et al. (1996)Go found that a tobacco MAR displaying greater affinity for tobacco matrices than the yeast ARS1 MAR increased reporter gene expression 140-fold compared with a 24-fold increase with ARS1. They concluded that there is a correlation between MAR binding strength and the effects of MARs on transgene expression.

In this study, we found no correlation between MAR binding strength and MAR effects on transgene expression in vivo: the maize Adh1 5' MAR and the yeast ARS1 decreased transgene silencing, whereas the maize Mha1 5' MAR had no apparent effect on average expression levels, despite binding maize nuclear matrices more strongly than Adh1 5' MAR. This result suggests that binding strength in vitro does not necessarily correlate with effects on transgene expression in vivo. One possible explanation for the discrepancy between the two studies could be that, in the former case, the conclusion was made after comparing a homologous and a heterologous MAR. We would have drawn a similar conclusion had we compared the homologous MAR from maize (Adh1 5' MAR) with the heterologous MAR from yeast (ARS1).

If the basis of the interactions between a MAR and the matrix is competition, it is possible that a higher affinity of Mha1 5' MAR for the matrix makes it less available for the binding of transcriptional regulators that, under similar circumstances, interact with other available MARs (e.g., Adh1 5' MAR). In other words, Mha1 5' MAR might be involved more in structural than in regulatory function. Defining MARs as "strong" or "weak" on the basis of in vitro binding affinity may reflect a different nature of the DNA-protein interactions (Tikhonov et al., 2000Go).

Adh1 5' MAR as a Positive/Negative Effector of Gene Expression: The Binary Model
An unexpected result of this study was the observation that Adh1 5' MAR may have a dual function, both activating (in callus) and repressing (in transgenic plants) transcription. The postulated mechanism behind a dual-function capacity is the ability of a DNA element to assemble both activating and repressive complexes. These may result from changes in their chromatin structure or in protein–protein interactions that affect DNA binding activity (Diamond et al., 1990Go). A binary model, termed binary because of its on/off nature, had been suggested to explain a positive and a negative regulation from a single DNA element (Walters et al., 1995Go, 1996Go). An important factor in the development of this model was the distinction made between expressing and nonexpressing cell populations (Walters et al., 1995Go).

Several aspects of this model agree well with our results. MAR-flanked, stably integrated constructs increased the number of expressors at the expense of nonexpressors (Figure 5) but were not able to reduce expression variability (Figure 4A). Therefore, MARs appear to protect transgenes from repression by counteracting the silencing effect. Thus, a function of a MAR might be to stabilize transcription, not modulate its level. The shift toward expressing cells observed here, without an increase in transcription levels within the population, suggests that MARs increased the chance of creating active templates. The frequency of such occurrences would depend on the availability of factors involved in the formation of regulatory complexes.

Bimodal Distribution Pattern
Large-sample analysis of transformed cells revealed a bimodal distribution pattern for the expression level data (Figure 4A). Some transformation studies have reported expression level data that are distributed normally after log transformation (Mlynarova et al., 1994Go; Allen et al., 1996Go), whereas others have reported a non-log-normal distribution (Breyne et al., 1992Go). In our analysis, graphing of the findings from 20 to 30 events did not result in a bimodal profile. Analysis of 30 to 50 events sometimes resulted in a bimodal distribution, but most of the time the pattern was not clear. However, samples of >50 events always resulted in a bimodal distribution curve. One possible explanation for this effect could be the wide range in transgene expression levels. When small sample numbers are analyzed, the events might be scattered across the range, obscuring any pattern.

A bimodal expression pattern was related to gene silencing caused by high copy numbers of integrated transgenes (Hobbs et al., 1990Go). Because the transformants in our case contain predominantly low copy numbers, we favor the possibility that the bimodal distribution resulted from a combination of silenced and expressing events. Probably, MARs increase the likelihood that an integrated transgene will be expressed, not the level of its expression. If MARs make a construct more efficient by creating an individual region, they would tend to increase the number of sites at which activity could occur after integration, and therefore the number of expressing cells. The probability of establishing expression would depend on the interplay between the chromatin at the integration site and the control elements of the construct. The action of the MARs is consistent with the view that they counteract the silencing.

How Do MARs Affect Expression in Vivo?
One possibility is that MARs act as structural elements that may create autonomously regulated chromatin loops. In addition to such a structural role, MARs may function as regulatory elements responsible for the state of chromatin, either through their inherent unwinding capability (Bode et al., 1992Go; Benham et al., 1997Go) or by binding regulatory proteins. SATB1 and Bright are two proteins that bind specifically the µ-immunoglobulin MAR (Dickinson et al., 1997Go; Dillon et al., 1997Go; Kohwi et al., 1997Go; Liu et al., 1997Go), and they illustrate how the same MAR may act as a positive regulator in B-cells and as a negative regulator in the non-B lineage (Scheuermann and Chen, 1989Go; Cunningham et al., 1994Go).

We suggest that the effects of the maize Adh1 5' MAR in the roots might be explained by a similar mechanism. Because in control (without MAR) plants GUS is expressed in all cell types, the loss of GUS expression in the majority of the root cells in constructs with MAR indicated a silencing potential for the MAR fragment. The specificity of this silencing function (except in nonrandom cell clusters where a transcriptionally active GUS state was preserved) suggested that root-specific protein factors might be involved.

Therefore, the different effects displayed by the same MAR in BMS cells and in whole plants could be attributable to different trans-acting factors present in BMS cells versus plants, to different interactions between the MAR and the two promoters, to different interactions between the MAR and the two transgene coding sequences, or to any combination of these causes.

Conclusion
The establishment and maintenance of appropriate expression patterns requires a balance between positive and negative regulatory mechanisms. The decision to establish an activating or a repressive complex may be the result of a stochastic binding of transcription factors, of the site of integration of transfected DNA into transcriptionally active or inactive regions of the nucleus, of the cell cycle state, of the age of the transformed cell, or of the cell's epigenetic status. The decision may be influenced by modulating the concentrations of the participating proteins or by modulating the protein binding capacity of DNA. The differentiation between a silenced and an expressing state may occur at the time of its establishment or during the process of its maintenance. The prevention of silencing by MARs could operate at either level.

Two of the MARs suppressed silencing of a LUC transgene in BMS callus, but none of the MARs used in this study displayed any effect on the variability of transgene expression among transformants. Contrary to a general belief that a MAR with the highest affinity for the nuclear matrix in vitro would have higher effects in vivo, we found that the two characteristics did not define each other. Thus, MAR binding affinity alone is insufficient to predict MAR effects in vivo.

Sample size, nature of MAR, proximity to promoter, and whether cultured cells or whole plants were analyzed may account, at least partly, for the variance in the behavior of these elements, explaining some reported controversies. Finally, the effect of Adh1 5' MAR on LUC expression in BMS callus was markedly different from the effect on GUS expression in transgenic maize plants, revealing a dual nature of Adh1 5' MAR. These results support the idea that MARs are not only structural elements but may be involved actively in maintaining a state that controls promoter activity (C. Brouwer and B. Bowen, unpublished data).


   METHODS
TOP
 Abstract
 INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
References
 
Nuclear Matrix Binding Assays
Nuclear matrix binding assays (MARs) were performed essentially as described by Avramova and Bennetzen (1993)Go. Matrices were isolated from 0.5 A260 units of leaf nuclei. A total of 50 ng of MAR containing plasmid (pUC19 containing ARS1, Adh1 5' MAR, or Mha1 5' MAR) was cut with HindIII-EcoRI, HindIII-NotI, or HindIII, respectively, to separate each MAR insert from the plasmid backbone. The DNA was end-labeled with {gamma}-32P-ATP and incubated with the isolated nuclear matrices corresponding to the residual protein obtained after extraction of 0.3 to 0.5 A260 units of starting nuclei.

Varying concentrations of Escherichia coli DNA (100- to 2500-fold molar excess) were used as competitors to prevent nonspecific binding (the A-T content of E. coli DNA is ~50%) and to compare the affinity of different DNA fragments for proteins in the matrix preparation. After the incubation, the 100-µL reaction was placed in a centrifuge, and the nuclear matrices were pelleted. The supernatant was discarded, and DNA that remained bound to the pelleted nuclear matrices was run on a gel, transferred to a nylon membrane, and exposed to film.

Vector Construction
Plasmid maps for vectors used in this study are shown in Figure 2A. PHP264 consists of an enhanced 35S promoter of Cauliflower mosaic virus (bases -421 to -90 and -421 to +2; Gardner et al., 1981Go), a 79-bp fragment from the 5' leader sequence of Tobacco mosaic virus (Gallie et al., 1987Go), the first intron of the maize (Zea mays) Adh1-S gene (Dennis et al., 1984Go), the coding sequence of the {beta}-glucuronidase (GUS) gene (Jefferson et al., 1987Go), and the potato proteinase II gene (bases 2 to 310; An et al., 1989Go). PHP264 and all other plasmids in this study have pUC-derived backbones. The vectors PHP1528 and PHP3528 are similar to PHP264 except for the coding sequence. PHP1528 contains the firefly luciferase (LUC) gene (de Wet et al., 1987Go), and PHP3528 contains the BAR gene from Streptomyces hygroscopicus (Thompson et al., 1987Go).

An 839-bp EcoRI-HindIII fragment region of ARS1 (Struhl et al., 1979Go) was modified by site-specific mutagenesis (Sambrook et al., 1989Go) so that it contained a BamHI site on the 3' end and a NotI site on the 5' end. This modified ARS1 fragment was inserted into PHP1528 at the 5' and 3' ends of the 35S::LUC cassette, to create PHP5438 (Figure 2). PHP5456 (Figure 2) was made by replacing the LUC coding sequence with the BAR coding sequence, using unique sites in the promoter and the terminator.

The 948-bp BamHI-PstI fragment of the maize Adh1 gene (Dennis et al., 1984Go) containing Adh1 5' MAR (Avramova and Bennetzen, 1993Go) was cloned into pBluescript SK+ (Stratagene). Mha1 5' MAR (-2562 to -1442 of GenBank entry U09989; Jin and Bennetzen, 1994Go) also was subcloned into pUC19. Vectors containing these MARs flanking LUC or BAR (PHP6248, PHP6344, PHP6486, and PHP6487) (Figure 2A) were constructed by inserting the MARs into sites at the 5' or 3' ends of PHP1528 or PHP3528, as described for PHP5438 and PHP5456.

PHP6086, PHP6608, PHP7819, and PHP7917 were constructed in a similar manner. The enhanced 35S promoter of Cauliflower mosaic virus was replaced with the Rsyn7 promoter, and GUS was used as the reporter. Rsyn7 is a root-specific, constitutively expressed promoter consisting of a synthetic sequence of ~140 bp containing three TGACG core motifs and a TATA box. This synthetic promoter has been described in detail elsewhere (B.A. Bowen, W.B. Bruce, G. Lu, L.E. Sims, and L.A. Tagliani [2000], U.S. patent 6,072,050; Lu and Bruce, 2000Go).

Transformation Methods
The Black Mexican Sweet cell line (kindly provided by Dave Somers, University of Minnesota, Minneapolis-St. Paul) was subcultured twice per week using 586 medium (modified MS2-D medium; Bittel et al., 1996Go). Two days before transformation, cells were subcultured and prepared for bombardment essentially as described previously (Bruce et al., 2000Go). One day before bombardment, cells were placed in osmoticum (586 medium plus 3% polyethylene glycol 8000) at a density of 200 mg/mL. Ten nanograms of reporter plasmid (PHP1528, PHP5438, PHP6248, or PHP6486) or 10 ng of selectable marker plasmid (PHP3528, PHP5456, PHP6344, or PHP6487) was precipitated onto 750 µg of 1.8-µm tungsten beads (General Electric, Fairfield, CT), and each preparation was divided into six aliquots for bombardment. Twenty nanograms of total DNA was used in all experiments.

Under the established transformation conditions, the DNA dose used yielded a majority of events that contain two or fewer copies of unselected transgenes. This fact has been determined and confirmed by numerous quantitative DNA gel blot analyses. An example of the copy-number distribution under the standard protocol conditions is shown for the control lines used in this study (Figure 3B). Based on this distribution and on massive previous experience, we assume that the MAR-flanked lines would have similar copy-number insertions. In the absence of the respective blots, however, the possibility that these particular MAR transformants might carry a different copy-number distribution cannot be excluded.

A total of 0.5 mL of cells was pipetted in a 2-cm circle onto sterile filter sets (consisting of a grade 391 filter on top and a grade 363 filter [both Whatman] beneath) premoistened with 750 µL of the same medium used for osmoticum. Cells were bombarded with a PDS1000 helium gun (Bio-Rad) using a 1100-p.s.i. rupture disk. Immediately after bombardment, cells were removed by placing the top filter on solid 586 medium containing 3% Gel-Rite (Merck & Co., Rahway, NJ) without selection. Three days after bombardment, the cells were scraped off the filter, suspended in 4 mL of 586 liquid medium, and divided into aliquots (1 mL/plate) on four plates of 586 solid medium containing the herbicide BASTA (AgrEvo, Wilmington, DE) at 3 mg/L. Stable transformants were recovered at 4 to 8 weeks after bombardment and were subcultured twice to confirm BASTA resistance. Events were assayed for LUC expression at 7 days after the second subculture. Expression levels were normalized to total protein levels (Bradford, 1976Go) and reagents from Bio-Rad.

For transient gene expression assays, Black Mexican Sweet cell lines were subcultured at 24 h before bombardment and placed in osmoticum at 4 h before shooting. A total of 100 mg of cells was plated onto filters. Ten micrograms of the test LUC constructs was mixed with 2 µg of PHP264 and precipitated onto 1.0-µm tungsten beads (General Electric or Bio-Rad). Bombardments were performed using a 600-p.s.i. rupture disc. At 20 h after bombardment, cells were harvested and assayed for LUC and GUS expression.

Stably transformed plants were produced from embryogenic callus induced from immature embryos of Hi-Type II maize (Armstrong et al., 1991Go). Particle bombardments of DNA constructs and subsequent tissue culture steps were conducted essentially as described (Klein et al., 1989Go; Bowen, 1992Go). PCR and DNA gel blot analyses were conducted on resistant calli and subsequently on selected transformed plants to confirm transgene incorporation.

Regenerated transgenic plants (T0) were grown to maturity in a greenhouse and crossed to a proprietary Pioneer inbred line to produce T1 plants. These were grown in a greenhouse until flowering (5 days after silking). Tissues, including leaf, tassel, ear, stem, and whole roots, were collected at or near flowering and incubated in a solution of 1 mg/mL 5-bromo-4-chloro-3-indolyl-{beta}-glucuronic acid cyclohexylamine (Rose Scientific, Edmonton, Alberta, Canada) at 37°C for 20 h essentially as described by Capellades et al. (1996)Go and then cleared in 70% ethanol.

Quantitative GUS assays were conducted by homogenizing fresh root tissue in Eppendorf tubes using a Kontes pestle (Vineland, NJ) in a solution of 50 mM sodium phosphate, pH 7.0, 10 mM EDTA, and 1 mM DTT. The tubes were centrifuged for 5 min at 4°C, and the supernatant was recovered and subjected to Gus Light assay from Tropix, Inc. (Bedford, MA), according to manufacturer's protocol. Concentrations of total soluble protein were determined using the method of Bradford (1976)Go.

Gene Expression Assays and Data Analysis
Callus tissue (100 mg) was ground in 300 µL of 0.1 M phosphate buffer, pH 7.8, and 1 mM DTT, and 10% of the cleared extract was diluted with 200 µL of buffer. A total of 100 µL of 1 mM luciferin was added, and light units (LU) were measured using a 10-s integration time on a single-well luminometer (model 2010; Analytical Luminescence Laboratories, San Diego, CA). For stable transformants, LU were normalized to total soluble protein. Transient gene expression levels were normalized by dividing LUC expression measurements (in LU/µL) by GUS expression levels (in LU/µL) measured in an equivalent volume of extract using the Gus Light kit from Tropix.

Statistical Analysis
Graphic and statistical analyses of log-transformed LUC expression level data were performed using JMP version 3.1.5 for Macintosh (SAS Institute, Cary, NC). To test for normality, we used the Shapiro-Wilk test. The cumulative distribution function was calculated with the formula

where n is the total number of transformation events and k = 1, 2, 3, ... n.

Upon request, all materials described in this article and owned by us will be made available in a timely manner for noncommercial research purposes. No restrictions or conditions will be placed on the use of any materials described in this article that would limit their use for noncommercial research purposes.

Accession Number
The GenBank accession number for Mha1 5' MAR is U09989.


   Acknowledgments
 
We are grateful to Jeff Bennetzen (Purdue University) for his interest and helpful discussions. We also thank Bruce Drummond, Margit Ross, Grace St. Clair, and Diane Bond-Nutter for providing their unpublished data on transgene copy number and DNA dose dependence, Loralee Logan for statistical advice, and Laura Tagliani for guidance on plasmid construction. This work was partially supported by U.S. Department of Agriculture National Research Initiative Competitive Grants Program Grant 98-353000-6167 to Z.A.


   Footnotes
 
Article, publication date, and citation information can be found at www.plantcell.org/cgi/doi/10.1105/tpc.004028.

1 Current address: CuraGen Corp, 555 Long Wharf Drive, New Haven, CT 06511. Back

3 Current address: Lynx Therapeutics, Inc., 25861 Industrial Boulevard, Hayward, CA 94545. Back

Received April 18, 2002; accepted May 29, 2002.


   References
TOP
 Abstract
 INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
 References
 
Allen, G.C., Hall, G.E., Jr., Childs, L.C., Weissinger, A.K., Spiker, S., and Thompson, W.F. (1993). Scaffold attachment regions increase reporter gene expression in stably transformed plant cells. Plant Cell 5, 603–613.[Abstract/Free Full Text]

Allen, G.C., Hall, G.E., Jr., Michalowski, S., Newman, W., Weissinger, A.K., Spiker, S., and Thompson, W.F. (1996). High level transgene expression in plant cells: Effects of a strong scaffold attachment region from tobacco. Plant Cell 8, 899–913.[Abstract]

Amati, B.B., and Gasser, S.M. (1988). Chromosomal ARS and CEN elements bind specifically to the yeast nuclear scaffold. Cell 54, 967–978.[CrossRef][Web of Science][Medline]

An, C., Mitra, A., Choi, H.K., Costa, M.A., An, K., Thornburg, R.W., and Ryan, C.A. (1989). Functional analysis of the 3' control region of the potato wound-inducible proteinase inhibitor II gene. Plant Cell 1, 115–122.[Abstract/Free Full Text]

Ananiev, E.V., Phillips, R.L., and Rines, H.W. (1998). Complex structure of knob DNA on maize chromosome 9: Retrotransposon invasion into heterochromatin. Genetics 149, 2025–2037.[Abstract/Free Full Text]

Armstrong, C.L., Green, C.E., and Phillips, R.L. (1991). Development and availability of germplasm with high type II culture formation response. Maize Genet. Coop. Newsl. 65, 92–93.

Avramova, Z. (2002). Heterochromatin in animals and in plants: Similarities and differences. Plant Physiol. 129, 40–49.[Free Full Text]

Avramova, Z., and Bennetzen, J.L. (1993). Isolation of matrices from maize leaf nuclei: Identification of a matrix-binding site adjacent to the Adh1 gene. Plant Mol. Biol. 22, 1135–1143.[Medline]

Avramova, Z., SanMiguel, P., Georgieva, E., and Bennetzen, J.L. (1995). Matrix attachment regions and transcribed sequences within a long chromosomal continuum containing maize Adh1. Plant Cell 7, 1667–1680.[Abstract]

Avramova, Z., and Tikhonov, A. (1999). Are scs and scs' ‘neutral’ chromatin domain boundaries of the locus? Trends Genet. 15, 138–139.[CrossRef][Web of Science][Medline]

Avramova, Z., Tikhonov, A., Chen, M., and Bennetzen, J.L. (1998). Matrix attachment regions and structural colinearity in the genomes of two grass species. Nucleic Acids Res. 26, 761–767.[Abstract/Free Full Text]

Avramova, Z., Tikhonov, A., SanMiguel, P., Jin, Y.K., Liu, C., Woo, S.S., Wing, R.A., and Bennetzen, J.L. (1996). Gene identification in a complex chromosomal continuum by local genomic cross-referencing. Plant J. 10, 1163–1168.[CrossRef][Web of Science][Medline]

Benham, C., Kohwi-Shigematsu, T., and Bode, J. (1997). Stress-induced duplex DNA destabilization in scaffold/matrix attachment regions. J. Mol. Biol. 274, 181–196.[CrossRef][Web of Science][Medline]

Bittel, D.C., Shaver, J.M., Somers, D.A., and Gengenbach, B.G. (1996). Lysine accumulation in maize cell cultures transformed with a lysine-insensitive form of maize dihydrodipicolinate synthase. Theor. Appl. Genet. 90, 70–77.[CrossRef]

Bode, J., Kohwi, Y., Dickinson, L., Joh, T., Klehr, D., Mielke, C., and Kohwi-Shigematsu, T. (1992). Biological significance of unwinding capability of nuclear matrix associating DNAs. Science 255, 195–197.[Abstract/Free Full Text]

Bonifer, C., Yannoutsos, N., Krüger, G., Grosveld, F., and Sippel, A.E. (1994). Dissection of the locus control function located on the chicken lysozyme gene domain in transgenic mice. Nucleic Acids Res. 22, 4202–4210.[Abstract/Free Full Text]

Bowen, B. (1992). Anthocyanin genes as visual markers in transformed maize tissues. In GUS Protocols: Using the GUS Gene as a Reporter of Gene Expression, S.R. Gallagher, ed (San Diego, CA: Academic Press), pp. 163–177.

Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254.[CrossRef][Web of Science][Medline]

Breyne, P., van Montagu, M., Depicker, N., and Gheysen, G. (1992). Characterization of a plant scaffold attachment region in a DNA fragment that normalizes transgene expression in tobacco. Plant Cell 4, 463–471.[Abstract/Free Full Text]

Bruce, W., Folkerts, O., Garnaat, C., Crasta, O., Roth, B., and Bowen, B. (2000). Expression profiling of the maize flavonoid pathway genes controlled by estradiol-inducible transcription factors CRC and P. Plant Cell 12, 65–79.[Abstract/Free Full Text]

Cai, H., and Levine, M. (1995). Modulation of enhancer-promoter interactions by insulators in the Drosophila embryo. Nature 376, 533–536.[CrossRef][Web of Science][Medline]

Callis, J., Fromm, M., and Walbot, V. (1987). Introns increase gene expression in cultured maize cells. Genes Dev. 1, 1183–1200.[Abstract/Free Full Text]

Capellades, M., et al. (1996). The maize caffeic acid O-methyltransferase gene promoter is active in transgenic tobacco and maize plant tissues. Plant Mol. Biol. 31, 307–322.[CrossRef][Web of Science][Medline]

Chen, M., SanMiguel, P., de Olivieira, A., Woo, S.S., Zhang, H., Wing, R.A., and Bennetzen, J.L. (1997). Microcolinearity in sh2-homologous regions of the maize, rice, and sorghum genomes. Proc. Natl. Acad. Sci. USA 94, 3431–3435.[Abstract/Free Full Text]

Chinn, A.M., and Comai, L. (1996). The heat shock cognate 80 gene of tomato is flanked by matrix attachment regions. Plant Mol. Biol. 32, 959–968.[CrossRef][Web of Science][Medline]

Chinn, A.M., Payne, S.R., and Comai, L. (1996). Variegation and silencing of the heat shock cognate 80 gene are relieved by a bipartite downstream regulatory element. Plant J. 9, 325–339.[Medline]

Cockell, M., and Gasser, S.M. (1999). Nuclear compartments and gene regulation. Curr. Opin. Genet. Dev. 9, 199–205.[CrossRef][Web of Science][Medline]

Cunningham, J.M., Purucker, M.E., Jane, S.M., Safer, B., Vanin, E.F., Ney, P.A., Lowrey, C.H., and Nienhuis, A.W. (1994). The regulatory element 3' to the A gamma-globin gene binds to the nuclear matrix and interacts with special A-T-rich binding protein 1 (SATB1), an SAR/MAR-associating region DNA binding protein. Blood 94, 1298–1308.

Dennis, E.S., Gerlach, W.L., Pryor, A.J., Bennetzen, J.L., Inglis, A., Llewellyn, D., Sachs, M.M., Ferl, R.J., and Peacock, W.J. (1984). Molecular analysis of the alcohol dehydrogenase (Adh1) gene of maize. Nucleic Acids Res. 12, 3983–4000.[Abstract/Free Full Text]

de Wet, J.R., Wood, K.V., DeLuca, M., Helinski, D.R., and Subramani, S. (1987). Firefly luciferase gene: Structure and expression in mammalian cells. Mol. Cell. Biol. 7, 725–737.[Abstract/Free Full Text]

Diamond, M.I., Miner, J.N., Yoshinaga, S.K., and Yamamoto, K.R. (1990). Transcription factor interactions: Selectors of positive or negative regulation from a single DNA element. Science 249, 1266–1272.[Abstract/Free Full Text]

Dickinson, L.A., Dickinson, C.D., and Kohwi, S.T. (1997). An atypical homeodomain in SATB1 promotes specific recognition of the key structural element in a matrix attachment region. J. Biol. Chem. 272, 11463–11470.[Abstract/Free Full Text]

Dillon, N., Trimborn, T., Strouboulis, J., Fraser, P., and Grosveld, F. (1997). The effect of distance on long-range chromatin interactions. Mol. Cell 1, 131–139.[CrossRef][Web of Science][Medline]

Flavell, R.B. (1994). Inactivation of gene expression in plants as a consequence of specific sequence duplication. Proc. Natl. Acad. Sci. USA 91, 3490–3496.[Abstract/Free Full Text]

Forrester, W.C., van Genderen, C., Jenuwein, T., and Grosschedl, R. (1994). Dependence of enhancer-mediated transcription of the immunoglobulin M gene on nuclear matrix attachment regions. Science 265, 1221–1225.[Abstract/Free Full Text]

Fransz, P.F., Armstrong, S., de Jong, J.H., Parnell, L.D., van Drunen, C., Dean, C., Zabel, P., Bisseling, T., and Jones, G.H. (2000). Integrated cytogenetic map of chromosome arm 4S of A. thaliana: Structural organization of heterochromatic knob and centromere region. Cell 100, 367–376.[CrossRef][Web of Science][Medline]

Gallie, D.R., Sleat, D.E., Watts, J.W., Turner, P.C., and Wilson, T.M. (1987). The 5'-leader sequence of tobacco mosaic virus RNA enhances the expression of foreign gene transcripts in vitro and in vivo. Nucleic Acids Res. 15, 3257–3273.[Abstract/Free Full Text]

Gardner, R.C., Howarth, A.J., Hahn, P., Brown-Luedi, M., Shepherd, R.J., and Messing, J. (1981). The complete nucleotide sequence of an infectious clone of cauliflower mosaic virus by m13mp7 shotgun sequencing. Nucleic Acids Res. 9, 2871–2888.[Abstract/Free Full Text]

Gasser, S.M., and Laemmli, U.K. (1987). A glimpse at chromosomal order. Trends Genet. 3, 16–22.

Grotewold, E., Chamberlain, M., Snook, M., Siame, B., Butler, L., Swenson, J., Maddock, S., Clair, G.S., and Bowen, B. (1998). Engineering secondary metabolism in maize cells by ectopic expression of transcription factors. Plant Cell 10, 721–740.[Abstract/Free Full Text]

Hobbs, S.L.A., Kpodar, P., and DeLong, C.M.O. (1990). The effect of T-DNA copy number, position and methylation on reporter gene expression in tobacco transformants. Plant Mol. Biol. 15, 851–864.[CrossRef][Web of Science][Medline]

Iglesias, V.A., et al. (1997). Molecular and cytogenetic analyses of stably and unstably expressed transgene loci in tobacco. Plant Cell 9, 1251–1264.[Abstract]

Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUS fusions: {beta}-Glucuronidase as a sensitive and versatile gene fusion marker in higher plants. Eur. Mol. Biol. J. 6, 3901–3907.[Web of Science][Medline]

Jenuwein, T., Forrester, W.C., Fernandez, H.L., Laible, G., Dull, M., and Grosschedl, R. (1997). Extension of chromatin accessibility by nuclear matrix attachment regions. Nature 385, 269–272.[CrossRef][Medline]

Jin, Y.-K., and Bennetzen, J.L. (1994). Integration and nonrandom mutation of a plasma membrane proton Atpase gene fragment within the bs1 retroelement of maize. Plant Cell 6, 1177–1186.[Abstract]

Klein, T.M., Kornstein, L., Sanford, J.C., and Fromm, M.E. (1989). Genetic transformation of maize cells by particle bombardment. Plant Physiol. 91, 440–444.[Abstract/Free Full Text]

Kohwi, S.T., Maass, K., and Bode, J. (1997). A thymocyte factor SATB1 suppresses transcription of stably integrated matrix-attachment region-linked reporter genes. Biochemistry 36, 12005–12010.[CrossRef][Medline]

Lamond, A.I., and Earnshaw, W.C. (1998). Structure and function of the nucleus. Science 280, 547–553.[Abstract/Free Full Text]

Liu, J., Bramblett, D., Zhu, Q., Lozano, M., Kobayashi, R., Ross, S.R., and Dudley, J.P. (1997). The matrix attachment region-binding protein SATB1 participates in negative regulation of tissue-specific gene expression. Mol. Cell. Biol. 17, 5275–5287.[Abstract]

Lu, G., and Bruce, W.B. (2000). A novel cis-acting element conferring root-preferred gene expression in maize. J. Plant Physiol. 156, 277–283.

Lyko, F., and Paro, R. (1999). Chromosomal elements conferring epigenetic inheritance. Bioessays 21, 824–832.[CrossRef][Web of Science][Medline]

Matzke, A.J.M., and Matzke, M.A. (1998). Position effects and epigenetic silencing of plant transgenes. Curr. Opin. Plant Biol. 1, 142–148.[CrossRef][Web of Science][Medline]

Mlynarova, L., Keizer, L.C.P., Stiekema, W.J., and Nap, J.P. (1996). Approaching the lower limits of transgene variability. Plant Cell 8, 1589–1599.[Abstract]

Mlynarova, L., Loonen, A., Heldens, J., Jansen, R.C., Keizer, P., Stiekema, W.J., and Nap, J.-P. (1994). Reduced position effect in mature transgenic plants conferred by the chicken lysozyme matrix-associated region. Plant Cell 6, 417–426.[Abstract]

Nabirochkin, S., Ossokina, M., and Heidmann, T. (1998). A nuclear matrix/scaffold attachment region co-localizes with the gypsy retrotransposon insulator sequence. J. Biol. Chem. 273, 2473–2479.[Abstract/Free Full Text]

Paul, A.-L., and Ferl, R.J. (1991). In vivo footprinting reveals unique cis-elements and different modes of hypoxic induction in maize Adh1 and Adh2. Plant Cell 3, 159–168.[Abstract/Free Full Text]

Peacock, W.J., Dennis, E.S., Rhoades, M.M., and Pryor, J. (1981). Highly repeated DNA sequence limited to knob heterochromatin in maize. Proc. Natl. Acad. Sci. USA 78, 4490–4494.[Abstract/Free Full Text]

Phi-Van, L., and Strätling, W.H. (1996). Dissection of the ability of the chicken lysozyme gene 5' matrix attachment region to stimulate transgene expression and to dampen position effects. Biochemistry 35, 10735–10742.[CrossRef][Medline]

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).

SanMiguel, P., Tikhonov, A., Jin, Y.-K., Motchoulskaya, N., Zakharov, D., Berhan, A., Springer, P., Edwards, K., Lee, M., Avramova, Z., and Bennetzen, J. (1996). Nested retrotransposons in the intergenic regions of the maize genome. Science 274, 765–768.[Abstract/Free Full Text]

Scheuermann, R.H., and Chen, U. (1989). A developmental-specific factor binds to suppressor sites flanking the immunoglobulin heavy-chain enhancer. Genes Dev. 3, 1255–1266.[Abstract/Free Full Text]

Schöffl, F., Schroder, G., Kliem, M., and Rieping, M. (1993). An SAR sequence containing 395 bp DNA fragment mediates enhanced, gene-dosage-correlated expression of a chimaeric heat shock gene in transgenic tobacco plants. Transgenic Res. 2, 93–100.[CrossRef][Medline]

Struhl, K., Stinchcomb, D.T., Scherer, S., and Davis, R.W. (1979). High-frequency transformation of yeast: Autonomous replication of hybrid DNA molecules. Proc. Natl. Acad. Sci. USA 76, 1035–1039.[Abstract/Free Full Text]

Thompson, C.J., Movva, N.R., Tizard, R., Crameri, R., Davies, J.E., Lauwereys, M., and Botterman, J. (1987). Characterization of the herbicide-resistance gene bar from Streptomyces hygroscopicus. Eur. Mol. Biol. J. 6, 2519–2523.[Web of Science][Medline]

Tikhonov, A.P., Bennetzen, J.L., and Avramova, Z.V. (2000). Structural domains and matrix attachment regions along colinear chromosomal segments of maize and sorghum. Plant Cell 12, 249–264.[Abstract/Free Full Text]

Tikhonov, A.P., SanMiguel, P.J., Nakajima, Y., Gorenstein, N.M., Bennetzen, J.L., and Avramova, Z. (1999). Colinearity and its exceptions in orthologous adh regions of maize and sorghum. Proc. Natl. Acad. Sci. USA 96, 7409–7414.[Abstract/Free Full Text]

Ülker, B., Allen, G.C., Thompson, W.F., Spiker, S., and Weissinger, A.K. (1999). A tobacco matrix attachment region reduces the loss of transgene expression in the progeny of transgenic tobacco plants. Plant J. 18, 253–264.[CrossRef][Web of Science]

Vain, P., Worland, B., Kohli, A., Snape, J.W., Christou, P., Allen, G.C., and Thompson, W.F. (1999). Matrix attachment regions increase transgene expression levels and stability in transgenic rice plants and their progeny. Plant J. 18, 233–242.[CrossRef]

van der Geest, A.H.M., Hall, G.E.J., Spiker, S., and Hall, T.C. (1994). The {beta}-phaseolin gene is flanked by matrix attachment regions. Plant J. 6, 413–423.[CrossRef]

van Drunen, C.M., Oosterling, R.W., Keultjes, G.M., Weisbeek, P.J., van Driel, R., and Smeekens, S.C.M. (1997). Analysis of the chromatin domain organisation around the plastocyanin gene reveals an MAR-specific sequence element in Arabidopsis thaliana. Nucleic Acids Res. 25, 3904–3911.[Abstract/Free Full Text]

Wakimoto, B.T. (1998). Beyond the nucleosome: Epigenetic aspects of position-effect variegation in Drosophila. Cell 93, 321–324.[CrossRef][Web of Science][Medline]

Walker, J.C., Howard, E.A., Dennis, E.S., and Peacock, W.J. (1987). DNA sequences required for anaerobic expression of the maize alcohol dehydrogenase 1 gene. Proc. Natl. Acad. Sci. USA 84, 6624–6628.[Abstract/Free Full Text]

Wallrath, L.L. (1998). Unfolding the mysteries of heterochromatin. Curr. Opin. Genet. Dev. 8, 147–153.[CrossRef][Web of Science][Medline]

Walters, M.C., Fiering, S., Eidemiller, J., Magis, W., Groudine, M., and Martin, D.I. (1995). Enhancers increase the probability but not the level of gene expression. Proc. Natl. Acad. Sci. USA 92, 7125–7129.[Abstract/Free Full Text]

Walters, M.C., Magis, W., Fiering, S., Eidemiller, J., Scalzo, D., Groudine, M., and Martin, D.I. (1996). Transcriptional enhancers act in cis to suppress position-effect variegation. Genes Dev. 10, 185–195.[Abstract/Free Full Text]

West, A.G., Gaszner, M., and Felsenfeld, G. (2002). Insulators: Many functions, many mechanisms. Genes Dev. 16, 271–288.[Free Full Text]




This article has been cited by other articles:


Home page
 Plant CellHome page
C. Halweg, W. F. Thompson, and S. Spiker
The Rb7 Matrix Attachment Region Increases the Likelihood and Magnitude of Transgene Expression in Tobacco Cells: A Flow Cytometric Study
PLANT CELL, February 1, 2005; 17(2): 418 - 429.
[Abstract] [Full Text] [PDF]

Home page
 Plant CellHome page
L. Mlynarova, A. Hricova, A. Loonen, and J.-P. Nap
The Presence of a Chromatin Boundary Appears to Shield a Transgene in Tobacco from RNA Silencing
PLANT CELL, September 1, 2003; 15(9): 2203 - 2217.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
14/9/2251    most recent
tpc.004028v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (14)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.
Agricola
Right arrow Articles by Brouwer, C.
Right arrow Articles by Bowen, B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
ASPB Publications THE PLANT CELL PLANT PHYSIOLOGY
Copyright © 2002 by the American Society of Plant Biologists