Skip Navigation


JXB Advance Access originally published online on April 8, 2004
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
55/401/1307    most recent
erh133v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
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 Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Agricola
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Journal of Experimental Botany, Vol. 55, No. 401, pp. 1307-1313, June 1, 2004
© 2004 Oxford University Press

RESEARCH PAPER

Strategies for precise quantification of transgene expression levels over several generations in rice

Received 29 October 2003; Accepted 16 February 2004

Victoria A. James*, Barbara Worland, John W. Snape and Philippe Vain{dagger}

John Innes Centre, Norwich Research Park, Colney Lane, Norwich NR4 7UH, UK

* Present address: Agronomy Department, University of Florida, Gainesville, FL 32611-0300, USA.
{dagger} To whom correspondence should be addressed. Fax: +44 (0)1603 450023. E-mail: philippe.vain{at}bbsrc.ac.uk


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results and discussion
 Conclusions
 References
 
Variation in transgene expression levels can result from uncontrolled differences in experimental protocols. Studies conducted over generations could, by their design, generate additional unwanted variation. To study sources of spurious variation, transgene expression levels were quantified over five homozygous generations in two independent transgenic rice lines created by particle bombardment. Both lines contained the same gus expression unit and had been shown to exhibit stable inheritance of transgene structure and expression. All plants were cultured and sampled using previously developed standardized protocols. Plants representative of each generation (T2, T3, T4, T5, T6) were grown either all together or across several different growth periods. GUS activity in plants from different generations was quantified either in the same assay or over multiple independent assays. Strategies in which plants were grown and phenotyped independently, significantly increased (up to 3-fold) extraneous variation in transgene expression level quantification, thus reducing the precision of molecular genetic studies and generating artefactual results in transgenic studies conducted over generations. Identification of sources of unwanted variation and quantification of their effect allowed the development of new strategies designed to control spurious variation. Growth and phenotyping of all plants from all generations together, using standard operating procedures (SOP), led to a reduction in extraneous variation associated with transgene expression level quantification. Adoption of such strategies is key to improving the reproducibility of transgenic studies conducted over generations.

Key words: Generational study, matrix attachment regions (MARs), Oryza sativa L., reproducibility, spurious variation, transgenic plants.


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results and discussion
 Conclusions
 References
 
Studies conducted over generations are key to assessing transgene behaviour in plants. Generational studies of transgene expression have been conducted in cereal crops, including rice transformed by direct DNA transfer (Fujimoto et al., 1993; Goto et al., 1993; Peng et al., 1995; Duan et al., 1996; Pawlowski and Somers, 1996; Kumpatla and Hall, 1998; Vain et al., 1999, 2002) or by Agrobacterium-mediated transformation (Mohanty et al., 1999; Wu et al., 2002; Vain et al., 2003). The majority of generational studies, however, have focused on the qualitative evaluation of transgene expression levels when assessing the segregation of the transgene phenotype (Shimamoto et al., 1989) or gene silencing phenomenon (Iyer et al., 2000). More rarely have transgene expression levels been quantified over generations in plants (Mlynárová et al., 1996; Ülker et al., 1999). The precise quantification of transgene expression levels is important when comparing plants with different ploidy levels (Duan et al., 1996; James et al., 2002), or for the identification of stable transformation events (Vain et al., 2002) for crop improvement or molecular genetic studies. Precision in the quantification of transgene expression levels over generations could be affected by uncontrolled variation associated with experimental or environmental factors. The assessment of transgene performances in multiple environmental conditions can be desirable in some large-scale crop improvement programmes, but uncontrolled environmental variations are generally unwanted in most molecular genetics studies. It has previously been shown that spurious variation could be introduced as a result of differences in plant culture conditions, sampling or analysis strategy (James et al., 2004) and that the adoption of standard operating procedures (SOP) can minimize this spurious variation. In studies conducted over several generations, plants from each generation are commonly grown and analysed together, but independently from subsequent generations (Srivastava et al., 1996; Fearing et al., 1997; Bettany et al., 1998). Such an experimental design could serve to amplify extraneous variation on the quantification of transgene expression levels associated with multiple and independent growth periods and phenotypic analyses.

In previous studies of transgene expression levels in rice (Vain et al., 1999, 2002; James et al., 2002, 2004), it was observed that the coefficient of variation (standard deviation/mean) associated with populations of homozygous rice plants from the same generation and transformation event, grown and analysed independently, was approximately 35–45%. When all plants were grown together and phenotyped together, the coefficient of variation could be reduced to approximately 10–15% (James et al., 2004). This represented the lower limit to the measured variation in transgene expression levels, using a standard operating procedure (SOP, James et al., 2004), at a given generation in rice. Additional variation above this baseline of 10–15% (coefficient of variation) is likely to limit the precision of transgenic studies in rice. In this study, the aim was to develop strategies for the quantification of transgene expression levels across generations in rice with low baseline variation (CV=10–15%).

In order to investigate the impact of different strategies on quantifying transgene expression levels over several generations in rice, homozygous plants from two independent transformation events were studied. Transgene activity was quantified in plants grown either together (G) or at different times (g), phenotyped either together (A) or at different times (a). It was shown that strategies in which plants were grown and phenotyped independently could significantly increase unwanted variation in transgene expression level quantification, reduce the precision of molecular genetic studies, and generate artefactual results in transgenic studies conducted over generations. Identification of the sources of unwanted variation and quantification of their effect allowed the development of new strategies designed to control spurious variation. Adoption of such strategies was key to improving the reproducibility of transgenic studies conducted over generations.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results and discussion
 Conclusions
 References
 
Rice plant culture and sampling
African elite rice (Oryza sativa L.) variety ITA212 was co-transformed by particle gun bombardment with plasmid pJIC201 (ubi-5'region::aphIV::SoyT) and plasmid pGA984 (ARS1MAR:: CaMV35S::gus::nosT::ARS1MAR) as previously described (Vain et al., 1999; Allen et al., 1993). All plants were grown and sampled following a standardized protocol to minimize spurious variation associated with plant culture conditions or sampling procedures (James et al., 2004). Plant progenies were germinated from mature embryos on MSR6H50 medium (Vain et al., 2002), containing hygromycin selection. After 3 weeks in vitro, plants were transferred to soil (3 kg (pH 5.6) loam, 2 kg horticultural sand, and 0.5 kg pouzzolane) in a fully controlled environment growthroom with a 12 h photoperiod, 28/24 °C day/night, 90% humidity, 1000 µE s–1 at 1 m light intensity provided by Osram HQI 400WD lamps.

Six homozygous plants from two independent transformation events (H10 and H24) were grown at each of five generations (T2, T3, T4, T5, T6), giving a set of 60 plants (see Fig. 1 in the Results and discussion). For each transformation event, two homozygous T1 plants were identified, each derived from a different T0 plant. Three T2 plants were grown from each homozygous T1 plant and used to establish six homozygous lines (HLs) per transformation event. The six HLs from each transformation event were maintained by single seed descent, germinating one seed from each plant at each subsequent generation. At each generation, transgene inheritance patterns were confirmed as Mendelian (data not shown) and transgene expression levels were quantified by fluorometric GUS assay. Leaf samples (3 cm in length) were collected in the middle of the day, 3 cm from the tip of the newest leaf (fifth leaf, 15 cm size) of single-tiller plants at the five-leaf stage, just prior to emergence of the sixth leaf. Samples were stored at –70 °C until required for enzymatic assay (James et al., 2004).

Enzymatic assay of GUS activity
Quantitative fluorometric analysis for ß-glucuronidase activity was carried out according to Jefferson et al. (1987) on leaf tissue from rice plants at the five-leaf stage. Fluorescence was measured using a Titertek Fluoroskan II after 0, 30, and 60 min incubation. Each assay was performed in triplicate. Protein content was determined using a Bio-Rad protein assay kit. Data were expressed as pmoles of 4-methylumbelliferone (MU) min–1 mg–1 of extracted protein. A background fluorescence of 35±4 pmol MU min–1 mg–1 protein (mean ±interval of confidence, P <0.05, n=50), based on measurements from leaf samples not containing the gus transgene present in each assay, was subtracted from all fluorometric GUS activity values obtained from transformed plants.

Statistical analysis
Statistical analyses, following the requirements of each test, were performed using Minitab 13.1 and Genstat 6 software packages. Normality of data sets was evaluated using the Anderson Darling test. Since the variance of transgene expression level was strongly dependent upon the mean (Vain et al., 1999, 2002), the variation within each population was assessed using the coefficient of variation (CV=standard deviation/mean). CVs were compared using Levene’s test on data expressed as a percentage of the mean (Vain et al., 1999). Data were subjected to analysis of variance (ANOVA).


   Results and discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results and discussion
Conclusions
 References
 
Influence of different strategies on quantification of transgene expression levels in rice over generations
Transgene expression levels were quantified over five generations (T2 to T6) in homozygous plants from two independent transformation events (H10 and H24) containing the gus transgene flanked by ARS1 MARs. H10 and H24 were chosen as they exhibited different levels of gus expression (around 3-fold) from the same transgene construct (pGA984). Both lines contained a single transgenic locus and were considered stable with respect to transgene structure and expression for more than two generations (Vain et al., 2002). A set of 60 homozygous plants was studied (Fig. 1). All plants were grown and sampled following a standardized protocol to minimize spurious variation associated with plant culture conditions or sampling procedures (James et al., 2004). As in most studies conducted over generations, the plants from each independent generation (T2, T3, T4, T5, T6) were grown in different growth periods. Two leaf samples were collected from each plant, one to be assayed at the time of collection (independent assays, gan) and the other stored at –70 °C until all samples were available for assay together (gAn). Assuming that no extraneous variation in transgene expression levels is associated with independent growth periods and assays, no effect of generation, T0 plant or HL would be expected.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 1. Structure of each set of 60 rice plants studied. Arrows indicate the relationship between individual rice plants. HL, homozygous line. Transgene expression level was quantified in plants within the box. H10 and H24 are two stable independent transformation events.

 
Experimental strategies are described using the following codes to indicate differences in growth period and enzymatic assay: growth period, G (same growth period), g (different growth periods); fluorometric enzymatic assay, A (same assay), a (independent assays); normalization of fluorometric data, N (normalized data), n (raw data).

Quantification of transgene expression levels over several independent assays (gan) resulted in high levels of variation between plants from different generations of the same transformation event (Fig. 2). The average GUS activities at each generation for H10 ranged from 1236–2590 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 28% (Table 1). The average GUS activities at each generation for H24 ranged from 3176–7152 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 29% (Table 1). In this strategy of performing enzymatic assays at the time of sample collection (gan), any effect of generation on transgene expression levels could be due, in part, to differences in growth period and assay. The combined effect of generation, growth period and independent assay had a significant effect on GUS activity in both H10 (P <0.001, ANOVA) and H24 (P=0.002, ANOVA). For both independent transformation events (H10 and H24), it seemed that transgene expression levels increased over successive generations up to T4 before stabilizing. In fact, further studies on the same plants using alternative analysis protocols revealed this pattern to be an artefact due to the phenotyping strategy employed (see gAn and gaN below).



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Quantification of transgene expression levels over five homozygous generations in rice plants using three different strategies (gan, gAn and gaN): g, different growth periods; A, same enzymatic assay; a, independent enzymatic assays; N, normalized data; n, raw data. T2–T6: generations. H10 and H24 are two stable independent transformation events. GUS activity in pmol MU min–1 mg–1 extracted protein (data detailed in Table 1). Each curve represents the average GUS activity at different generations for the six HLs studied for H10 and H24 (Fig. 1). Bars represent the standard error of the mean calculated from the residual mean square in ANOVA (representing combined variability).

 

View this table:
[in this window]
[in a new window]
  		Table 1. Quantification of transgene expression levels over five homozygous generations in rice plants using three different strategies (gan, gAn and gaN) g, Different growth periods; A, same enzymatic assay; a, independent enzymatic assays; N, normalized data; n, raw data. GUS activity in pmol MU min–1 mg–1 extracted protein. HL, homozygous lines. H10 and H24 are two stable independent transformation events. CV, coefficient of variation (* among HL at a given generation; ** between generations).
 
It has previously been shown that, among rice plants of the same generation, multiple independent phenotypic analyses were a major source of spurious variation in transgene expression level quantification (James et al., 2004). To assess the effect of multiple phenotyping in transgenic studies across generations better, leaf samples from the H10 and H24 plants, previously grown in independent growth periods, were enzymatically assayed together (gAn). Variation in average GUS activity between generations was greatly reduced compared with measurements from independent assays (gan) (as illustrated by the flatter curves in Fig. 2). The average GUS activities at each generation for H10 (assayed together) ranged from 1893–2889 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 20% (Table 1). The average GUS activities at each generation for H24 (assayed together) ranged from 6541–8479 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 10% (Table 1). In this strategy (gAn), the effects of growth period and generation could not be distinguished, since each individual generation was grown in a different growth period. The combined effect of generation and growth period had a significant influence on transgene expression levels in transformation event H10 (P=0.02, ANOVA), but not H24 (P=0.535, ANOVA). By contrast with the previous strategy involving multiple phenotyping (gan), no increase in transgene expression levels across generations was observed (Fig. 2). Since the leaf samples assayed in both studies (gAn and gan) were collected from the same plants at the same time, it shows that multiple phenotypic analyses can introduce spurious variation to transgene expression level quantification over generations. The experimental strategy of performing all enzymatic assays together reduced the level of unwanted variation in the assessment of transgene expression levels over generations, as observed previously for large-scale transgenic studies at a single given generation (James et al., 2004). However, some spurious variation was still observed (line H10).

An alternative strategy to compensate for performing fluorometric assays at independent times was to normalize the data from independent assays to several internal control samples (gaN). Normalization of fluorometric data (gaN) appeared to offer some benefit over the use of raw data (gan), but did not reduce variation between generations to the extent achieved by phenotyping all samples together (gAn) (Fig. 2). The profile of the normalized expression level data (gaN) plotted over generations represented an intermediate result between the two previous strategies (gan and gAn). The average GUS activities at each generation for H10 ranged from 1657–2765 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 22% (Table 1). The average GUS activity for each generation for H24 ranged from 4744–6466 pmol MU min–1 mg–1 extracted protein and the coefficient of variation between generations was 12% (Table 1). Following normalization of fluorometric data, the combined effects of generation, growth period, and independent assay showed a significant effect on GUS activity in H10 (P=0.003, ANOVA), but not H24 (P=0.39, ANOVA).

Overall, there was a significant difference in the levels of GUS activity measured using the three strategies (H10 P=0.01, H24 P <0.001, ANOVA). Artefactual differences between generations were also reduced using the gAn (assay together) or gaN (normalization) approach when compared with gan (independent assays). The strategy for quantification of transgene expression levels is therefore critical when analysing the same set of plants. Independent phenotyping, in particular, can introduce spurious differences between generations. In the literature, most studies involving analysis of transgene expression levels over several generations used an experimental strategy whereby plants from each generation were grown and analysed at independent time points (Srivastava et al., 1996; Fearing et al., 1997; Bettany et al., 1998). The results of the present study show that such an experimental strategy could significantly influence transgenic studies. It is therefore important to quantify precisely and to minimize as much as possible the spurious variation associated with independent growth periods and phenotyping in order to attain precision in transgene expression level quantification.

Minimizing the effects of growth period and multiple phenotyping in transgenic studies over generations
The influence of growth period and independent phenotypic analysis on transgene expression level quantification in this study suggested that a strategy involving growth and analysis of all plants together (GAn) should provide the greatest precision. To test the robustness of this approach, four replicates of the complete set of the 60 plants from transformation events H10 and H24 were grown (each set as described in Fig. 1). Sets 1 and 2 were grown in growth period 1 and sets 3 and 4 were grown in growth period 2. Transgene expression levels were quantified in plants from each of the four complete sets together, but independently from the three replica sets. Within each set, plants from all generations were thus grown and assayed together (GAn). All plants were grown and sampled following a standardized protocol to minimize spurious variation associated with plant culture conditions or sampling procedures (James et al., 2004). Plants were evenly distributed in a controlled environment room and leaf samples were evenly positioned in each fluorometric GUS assay to minimize intra-assay variation. This experiment, comprising four sets of 60 homozygous plants, also allowed the effects of different growth periods (two) and independent assays (four) to be reassessed with increased precision.

Average GUS activities over generations are represented in Table 2 and Fig. 3. Within each set, where plants from all generations were grown and assayed together (GAn), there was no significant difference in GUS activities between plants of different generations, from each independent transformation event (P >0.05, ANOVA; Fig. 3). This showed that the GAn strategy could be reliably used to assess transgene expression levels across generations. Repetition of the GAn strategy across two independent growth periods and four independent phenotyping analyses did not generate significant levels of spurious variation in expression studies across generations. The coefficients of variation among generations within each set varied from 6% to 14% (GAn, Table 2) compared with 28% or 29% among generations assayed independently (gan, Table 1). The expected CV among homozygous plants of the same line and generation germinated and assayed at the same time was 10–15% (i.e. the lower limit to the measured variation in transgene expression levels, using a standard operating procedure, SOP, James et al., 2004). These data suggest that variation among homozygous plants representing different generations of the same transformation event is no greater than variation among plants of the same generation. It also demonstrates that spurious variation in transgene expression levels can be minimized by adopting strategies controlling both growth period and phenotypic analysis.


View this table:
[in this window]
[in a new window]
  		Table 2. Quantification of transgene expression levels over five homozygous generations in rice plants using a strategy designed to minimize spurious variation (GAn) GUS activity in pmol MU min–1 mg–1 extracted protein. CV, coefficient of variation within * and between ** generations. H10 and H24 are two stable independent transformation events.
 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. Quantification of transgene expression levels over five homozygous generations in rice plants using a strategy designed to minimize spurious variation (GAn). G, Same growth period; A, same enzymatic assay; n, raw data. GUS activity in pmol MU min–1 mg–1 extracted protein (data detailed in Table 2). Set 1: assay 1, growth period 1; Set 2: assay 2, growth period 1; Set 3: assay 3, growth period 2; Set 4: assay 4, growth period 2. Within each set, the plants from all generations were grown and assayed together (GAn). H10 and H24 are two stable independent transformation events. Bars represent the standard error of the mean calculated from the residual mean square in ANOVA (representing combined variability).

 
Analysis of all data from both transformation events (H10 and H24, growth periods 1 and 2, assays 1–4) revealed a significant effect of growth period (P=0.020, ANOVA) and independent assay (P=0.019, ANOVA) on the quantification of transgene expression levels as previously reported (James et al., 2004). In transgenic studies conducted across generations, such effects can accumulate with each generation, leading to misinterpretation of expression level data (see gan in the first section of this paper). In this study, growth periods 1 and 2 were only a few weeks apart and the two enzymatic assays of plants from each growth period were performed within a short period of time. It is likely that even greater effects would have been observed had the growth period and phenotypic analyses been separated by longer time intervals. The effect of growth period was difficult to determine since all samples from each independent assay had been grown in the same growth period. Analysis of factor interactions showed that growth period alone had a more limited effect than independent analyses in generating spurious variation in the quantification of transgene expression levels.


   Conclusions
Top
 Abstract
 Introduction
 Materials and methods
 Results and discussion
 Conclusions
References
 
Accurate measurement of expression levels over generations is important in identifying transformation events exhibiting stable transgene behaviour for molecular genetic studies or crop improvement. Any experimental strategy that increases spurious variation in transgene expression level quantification will consequently reduce the precision of transgenic studies. Artefacts resulting from differences in experimental protocols can be more easily introduced and more pronounced when plants from one generation are grown and analysed together, but independently from each subsequent generation (gan). Such an experimental strategy could potentially double or triple the variation in transgene expression level measurements compared with a strategy where plants are grown and analysed together (i.e. CVs from around 10% for GAn to around 30% for gan). In rice, differences in growth period and phenotypic analysis were shown to have significant effects on the quantification of transgene expression levels. In fact, the variation in transgene expression levels among plants of different generations was similar to the background level of variation expected among plants of the same generation, when both growth and analysis were controlled. The problems encountered in generational studies of transgene expression are similar, but more pronounced, than those previously observed in large-scale transgenic studies at a given generation (James et al., 2004). It is foreseeable that phenotypic analyses such as bioassays, which are more complex than simple enzymatic assays used in this study, could generate even more spurious variation when conducted independently across generations.

Normalization of fluorometric data to several controls can compensate for the effect of performing independent analyses. In this study, normalization of fluorometric data from plants grown and assayed independently (gaN) clearly reduced the variation in GUS activity between generations. Normalizing all values to a control population that itself has a degree of variation associated with it, however, may be disadvantageous when only limited uncontrolled variation exists between independent assays.

This work shows that studies of transgene expression levels over generations can be significantly influenced by unwanted and uncontrolled variation introduced by experimental protocols and strategies. Following a given protocol does not guarantee elimination of spurious variation and, therefore, misinterpretation in generational transgenic studies, if the major sources of unwanted variation have not been previously identified, their effect quantified, and new protocols designed to limit their effect. The best experimental strategy that has been identified to reduce spurious variation in transgene expression level quantification over generations involves the standardization of plant culture conditions and sampling methods (James et al., 2004) as well as growth of all plants together and phenotypic analysis of all samples together. Despite not fitting the natural time-course of plant studies over successive generations, this strategy is important for precision when comparing transgene expression levels. When growth and phenotypic analysis of all plants together is not feasible, normalization of the data set and statistical analysis taking into account growth period and phenotyping are required. It is suspected that the general rules established in this study for rice could be used as a guideline for transgenic studies across generations in other plant species. The development of standard operating procedures (SOP) for the quantification of transgene expression levels over generations is a key step to improving the reproducibility, and therefore the quality assurance, of transgenic studies.


   Acknowledgements
 
We gratefully acknowledge The John Innes Foundation for its support. This document is an output from a project (Plant Sciences Research Programme R8031) funded by the UK Department for International Development (DFID) and administered by the Centre for Arid Zone Studies (CAZS) for the benefit of developing countries. The views expressed are not necessarily those of DFID.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results and discussion
 Conclusions
 References
 
Allen GC, Hall GE, Childs LC, Weissinger AK, Spiker S, Thompson WF. 1993. Scaffold Attachment Regions increase reporter gene expression in stably transformed plant cells. The Plant Cell 5, 603–613.[Abstract/Free Full Text]

Bettany AJE, Dalton SJ, Timms E, Morris P. 1998. Stability of transgene expression during vegetative propagation of protoplast-derived tall fescue (Festuca arundinacea Schreb.) plants. Journal of Experimental Botany 49, 1797–1804.[Abstract/Free Full Text]

Duan XL, Li XG, Xue QZ, AboElSaad M, Xu DP, Wu R. 1996. Transgenic rice plants harboring an introduced potato proteinase inhibitor II gene are insect resistant. Nature Biotechnology 14, 494–498.[CrossRef][Web of Science][Medline]

Fearing PL, Brown D, Vlachos D, Meghji M, Privalle L. 1997. Quantitative analysis of CryIA(b) expression in Bt maize plants, tissues, and silage and stability of expression over successive generations. Molecular Breeding 3, 169–176.

Fujimoto H, Itoh K, Yamamoto M, Kyozuka J, Shimamoto K. 1993. Insect-resistant rice generated by introduction of a modified {delta}-endotoxin gene of Bacillus thuringiensis. Nature Biotechnology 11, 1151–1155.

Goto F, Toki S, Uchimiya H. 1993. Inheritance of a co-transferred foreign gene in the progenies of transgenic rice plants. Transgenic Research 2, 300–305.[CrossRef]

Iyer LM, Kumpatla SP, Chandrasekharan MB, Hall TC. 2000. Transgene silencing in monocots. Plant Molecular Biology 43, 323–346.[CrossRef][Web of Science][Medline]

James VA, Avart C, Worland B, Snape JW, Vain P. 2002. The relationship between homozygous and hemizygous transgene expression levels over generations in populations of transgenic rice plants. Theoretical and Applied Genetics 104, 553–561.[CrossRef][Web of Science][Medline]

James VA, Worland B, Snape JW, Vain P. 2004. Development of a standard operating procedure (SOP) for the precise quantification of transgene expression levels in rice. Physiologia Plantarum 120, 650–656.[CrossRef][Medline]

Jefferson RA, Kavanagh TA, Bevan MW. 1987. GUS fusions: ß-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO Journal 6, 3901–3907.[Web of Science][Medline]

Kumpatla SP, Hall TC. 1998. Recurrent onset of epigenetic silencing in rice harboring a multi-copy transgene. The Plant Journal 14, 129–135.[Medline]

Mlynárová L, Keizer LCP, Stiekema WJ, Nap JP. 1996. Approaching the lower limits of transgene variability. The Plant Cell 8, 1589–1599.[Abstract]

Mohanty A, Sarma NP, Tyagi AK. 1999. Agrobacterium-mediated high frequency transformation of an elite indica rice variety Pusa Basmati 1 and the transmission of the transgenes to R2 progeny. Plant Science 147, 127–137.

Pawlowski WP, Somers DA. 1996. Transgene inheritance in plants genetically engineered by microprojectile bombardment. Molecular Biotechnology 6, 17–30.[Web of Science][Medline]

Peng JY, Wen FJ, Lister RL, Hodges TK. 1995. Inheritance of gusA and neo genes in transgenic rice. Plant Molecular Biology 27, 91–104.[CrossRef][Web of Science][Medline]

Shimamoto K, Terada R, Izawa T, Fujimoto H. 1989. Fertile transgenic rice plants regenerated from transformed protoplasts. Nature 338, 274–276.

Srivastava V, Vasil V, Vasil IK. 1996. Molecular characterization of the fate of transgenes in transformed wheat (Triticum aestivum L). Theoretical and Applied Genetics 92, 1031–1037.[CrossRef]

Ülker B. Allen GC, Thompson WF, Spiker S, Weissinger AK. 1999. A tobacco matrix attachment region reduces the loss of transgene expression in the progeny of transgenic tobacco plants. The Plant Journal 18, 253–263.[CrossRef]

Vain P, Worland B, Kohli A, Snape JW, Christou P, Allen GC, Thompson WF. 1999. Matrix attachment regions increase transgene expression levels and stability in transgenic rice plants and their progeny. The Plant Journal 18, 233–242.[CrossRef]

Vain P, James VA, Worland B, Snape JW. 2002. Transgene behaviour across two generations in a large random population of transgenic rice plants produced by particle bombardment. Theoretical and Applied Genetics 105, 878–889.[CrossRef][Web of Science][Medline]

Vain P, Afolabi A, Worland B, Snape JW. 2003. Transgene behaviour in populations of rice plants transformed using a new dual binary vector system: pGreen/pSoup. Theoretical and Applied Genetics 107, 210–217.[CrossRef][Web of Science][Medline]

Wu G, Cui H, Ye G, Xia Y, Sardana R, Cheng X, Li Y, Altosaar I, Shu Q. 2002. Inheritance and expression of the cry1Ab gene in Bt (Bacillus thuringiensis) transgenic rice. Theoretical and Applied Genetics 104, 727–734.[CrossRef][Web of Science][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
55/401/1307    most recent
erh133v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
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 Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Agricola
Right arrow Articles by James, V. A.
Right arrow Articles by Vain, P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?