- Research article
- Open Access
Methylome reorganization during in vitro dedifferentiation and regeneration of Populus trichocarpa
© Vining et al.; licensee BioMed Central Ltd. 2013
Received: 18 January 2013
Accepted: 12 June 2013
Published: 25 June 2013
Cytosine DNA methylation (5mC) is an epigenetic modification that is important to genome stability and regulation of gene expression. Perturbations of 5mC have been implicated as a cause of phenotypic variation among plants regenerated through in vitro culture systems. However, the pattern of change in 5mC and its functional role with respect to gene expression, are poorly understood at the genome scale. A fuller understanding of how 5mC changes during in vitro manipulation may aid the development of methods for reducing or amplifying the mutagenic and epigenetic effects of in vitro culture and plant transformation.
We investigated the in vitro methylome of the model tree species Populus trichocarpa in a system that mimics routine methods for regeneration and plant transformation in the genus Populus (poplar). Using methylated DNA immunoprecipitation followed by high-throughput sequencing (MeDIP-seq), we compared the methylomes of internode stem segments from micropropagated explants, dedifferentiated calli, and internodes from regenerated plants. We found that more than half (56%) of the methylated portion of the genome appeared to be differentially methylated among the three tissue types. Surprisingly, gene promoter methylation varied little among tissues, however, the percentage of body-methylated genes increased from 9% to 14% between explants and callus tissue, then decreased to 8% in regenerated internodes. Forty-five percent of differentially-methylated genes underwent transient methylation, becoming methylated in calli, and demethylated in regenerants. These genes were more frequent in chromosomal regions with higher gene density. Comparisons with an expression microarray dataset showed that genes methylated at both promoters and gene bodies had lower expression than genes that were unmethylated or only promoter-methylated in all three tissues. Four types of abundant transposable elements showed their highest levels of 5mC in regenerated internodes.
DNA methylation varies in a highly gene- and chromosome-differential manner during in vitro differentiation and regeneration. 5mC in redifferentiated tissues was not reset to that in original explants during the study period. Hypermethylation of gene bodies in dedifferentiated cells did not interfere with transcription, and may serve a protective role against activation of abundant transposable elements.
A growing body of evidence documents extensive epigenetic changes as a result of in vitro plant tissue culture . The genetic and epigenetic mutations induced can be a detriment to clonal propagation but they can also provide a tool for producing stress-tolerant and/or disease resistant plants by in vitro selection of somaclonal variants [2, 3]. However, the nature of the epigenetic changes produced by in vitro regeneration are poorly known, particularly on a genome scale.
The process of eukaryotic cellular dedifferentiation is often referred to as a return to a ‘stem-cell like’ state, as cells first must re-enter the cell cycle. This developmental shift involves large-scale chromatin reorganization, leading to acquisition of pluripotency (reviewed by ). Cellular differentiation, in contrast, occurs in response to the balance of growth regulators in the culture medium and ultimately leads to organogenesis. This transition involves cell fate decisions and eventual exit from the cell cycle. Both dedifferentiation and differentiation involve changes in expression of key genes. Epigenomic reprogramming underlying these large developmental shifts is thought to be a major cause of somaclonal variation.
Somaclonal variation can be a serious problem in commercial nurseries, occurring both in the field and during in vitro propagation. While the intention of clonal propagation is regeneration of phenotypically identical individuals, it is often not the case in practice. This is illustrated by the mantled floral phenotype in oil palm (Elaeisguineensis Jacq.), which affects ~5% of regenerated palms . The mantled mutation results in abnormal flowers, fruits, and ultimately decreased oil yield. Several studies have revealed genome-wide DNA hypomethylation in mantled somaclones compared to normal counterparts [6–8], demonstrating that changes in DNA methylation can be associated with phenotypes occurring after in vitro propagation.
DNA methylation is known to vary as a result of in vitro culture in other species as well. Methylation-sensitive amplified polymorphism (MSAP) has been widely usedto study epigenetic instability during regeneration [2, 9–11]. Comparisons of banding patterns between regenerants and donor plants in Freesia, oil palm and hops (Humulus lupulus L.) generally show both gains and losses of polymorphic bands, but substantially higher proportions of polymorphisms showing demethylation at the target site [9, 11, 12]). Using the MSAP method, hop plants from repeated rounds of regeneration from callus tissue were found to have increased 5mC differentiation compared to donor tissue . In cocoa (Theobroma cacao L.), MSAP variability in regenerants from leaves and staminodes increased over time in culture .
Changes in chromatin structure have been observed during in vitro dedifferentiation, and have mainly been studied in Arabidopsis protoplast systems. On Arabidopsis chromosome I, changes included condensation of 18 s ribosomal DNA and decondensation of telomeric and pericentromeric regions, but no changes in centromeric repeats . In long-term Arabidopsis suspension cultures, examination of chromosome IV showed that methylation of euchromatin increased and methylation of heterochromatin decreased . Studies of genomic 5mC content of Arabidopsis dedifferentiating cells have reported mixed results. Elhiti et al.  observed global DNA hypomethylation during the induction phase of somatic embryogenesis in Arabidopsis, but Tessadori et al.  reported no difference in genomic 5mC during chromatin decondensation in Arabidopsis mesophyll cells dedifferentiating into protoplasts. It appears that epigenomic changes vary widely among chromatin domains, and vary widely among the cell types and culture systems employed.
In general, plant transposable elements are transcriptionally inactivated by cytosine methylation . Under in vitro culture conditions, methylation of specific families of transposable elements is altered and demethylation can result in transcriptional reactivation [19, 20]). In oil palm, transposable elements are generally hypomethylated at the dedifferentiated stage then remethylated in regenerated plants, but to a lower level than donor plants .The 5’ LTR portion of an LTR retroelement of the Gypsy superfamily LORE1 was demethylated and in some cases transcriptionally activated in regenerated plants of Lotus japonicus. In addition, this LTR-Gypsy retroelement was transpositionally active in pollen of regenerated plants. Comparisons of active and inactive copies of the MITE element mPing in rice (Oryza sativa L.) calli and regenerated plants showed hypermethyation of flanking regions of the immobile copy . Comparisons of 5mC in Arabidopsis long-term suspension cultures relative to leaf cells showed that transposable elements in heterochromatin were hypomethylated, including Athila and LTR-Copia retroelements and AtMu DNA transposons. However, LTR-Gypsy class retrotransposons had become hypermethylated . Variation among transposon classes in their epigenetic responses to in vitro culture may be an important source of the variation observed among chromatin domains.
Methylation of particular genes has been shown to change during in vitro plant dedifferentiation and redifferentiation. In Arabidopsis, cellular dedifferentiation was accompanied by MET1- and DRM2-mediated promoter hypermethylation of MAPK12, GSTU10 and BXL1 in callus cells, and of TTG1, GSTF5, SUVH8, fimbrin and CCD7 in cell suspension cultures . Transcription of genes involved in DNA methylation, including met1, cmt3, drm1 and drm2 were upregulated in Arabidopsis cell suspension culture, while the DNA demethylase ros1 was downregulated .
There have been few studies of genome-scale reorganization during in vitro organogenesis, and most have focused on the transcriptome. In rice (Oryza sativa), only one to three percent of genes showed differential regulation during organogenesis of shoot, roots, and somatic embryos . Bao et al.  studied transcriptome changes in a Populus tremulax P. alba hybrid during in vitro dedifferentiation and organogenic regeneration using a commercial (Affymetrix) oligonucleotide genome-scale microarray. In sequential time-point comparisons, the greatest number of differentially-regulated genes (~9,000) was detected at the early callogenesis stage, with roughly equal numbers of upregulated and downregulated genes. They identified a large number of overrepresented Gene Ontology (GO) categories associated with the transition, including those related to protein metabolism, stress response, cellular signaling, and transcription. Surprisingly, there were far fewer genes that were differentially regulated during shoot regeneration from callus. In the same poplar system, Bao et al.  showed that regulatory genes in meristem development, including members of the WUSCHEL (WUS) and SHOOTMERISTEMLESS (STM) families, were among the genes downregulated at early callus initiation, and that one of the two poplar WUS paralogs was upregulated during shoot initiation, though the other was not differentially regulated.These studies provide further evidence of the highly complex and variable nature of gene expression programs during in vitro regeneration.
To further understanding of epigenetic modification and its relationship to transcription in poplar, we compared genome-wide 5mC profiles in stem internode tissues at three stages of in vitro plant regeneration: the micropropagated internode explant, undifferentiated callus, and internodes from regenerated shoots while still in vitro. To facilitate genome sequence mapping, we used genotype ‘Nisqually-1’ of Populus trichocarpa, from which the Populus reference genome sequence was derived . We analyzed changes as 1 kb tiled genome windows throughout the genome, as well as specific genome features including gene promoters, gene bodies, and transposable elements. In addition, we compared 5mC profiles among the three tissue culture stages to the Populus Affymetrix microarray dataset  to ascertain if any broad relationships between cytosine methylation and gene expression exist. We report substantial changes in DNA methylation during regeneration, changes in methylation that varied widely among chromosomal and genic regions, and relationships of methylation in specific genic regions to gene expression.
Sequencing samples and MeDIP-seq read mapping
Summary of sequencing samples and results
204, 768, 821
214, 470, 025
48, 813, 758
147, 824, 747
35, 116, 532
61, 453, 962
628, 517, 555
Reads per 1 kb window per million mapped reads (RPKM) values were calculated for each biological replicate within tissues (Figure 2). Mean RPKM variance among biological replicate samples per window ranged from 4.5-9.3 for explants, 5.1-8.6 for calli, and 6.1-34.9 for regenerated internodes. When biological replicates for all samples were subjected to hierarchical clustering analysis using RPKM calculations over all annotated gene bodies, non-immunoprecipitated control samples formed a distinct cluster, and explant and callus tissues formed a separate cluster (Additional file 1). One of the three regenerated internode replicates clustered with explant and callus replicates, while the other two regenerated internode replicates clustered separately.
Decrease in overall genome 5mC during in vitro culture
Genome methylation tallied over 1 kb tiled genome windows, with tissues pooled
No. methylated 1 kb windows
239, 225 (82.5%)
Genome methylation tallied over 1 kb tiled genome windows, by tissue
No. methylated 1 kb windows
52, 566 (55.4%)
46, 863 (54.2%)
39, 384 (58.9%)
Gene body 5mC changes were different from those of gene promoters and intergenic space
Gene bodies – which consist of annotated gene transcripts – had a very different pattern of methylation than promoters or intergenic DNA. Over 60% (3,573 of 5,906) of body-methylated genes showed tissue differentiation (Figure 7), mainly because gene body 5mC was markedly higher in callus (Additional file 2). Of the 3,573 body-methylated genes with tissue differentiation, 1,729 (48%) were transiently methylated (methylated during tissue dedifferentiation, then demethylated during organogenic regeneration). In contrast, only 61 genes (1%) showed the opposite pattern of demethylation during dedifferentiation and remethylation during regeneration. Relatively few untranslated regions were 5mC-enriched. Only 0.3% to 1% of 5’UTRs were enriched, with none significantly different in pairwise comparisons.Only 1% to 3% 3’of UTRs were 5mC-enriched, with less than 10 significantly different in pairwise comparisons.
Transposable elements show 5mC increase during in vitro culture
Transposable element methylation increases through dedifferentiation and organogenesis for abundant elements
Gene body methylation associated with GO enrichmentand gene density
Overrepresented Gene Ontology (GO) categories of genes with transient body methylation
Number in input list
Number in BG/Ref
Intracellular protein transport
Cellular protein localization
protein cataolic process
ubiquitin-dependent protein catabolic process modification-dependent protein catabolic
Nucleic acid metabolism
nucleobase, nucleoside and nucleotide
DNA metabolic process
RNA metabloic process
nCRNA metablic process
tRNA metabloic process
regulation of small GTPase mediated signal
regulation of Ras protein signal transduction
Ras protein signal transduction
regulation of signal transduction
regulation of signaling process
regulation of cell communication
ARF protein signal transduction
Association of methylation and gene expression
Genic 5mC enrichment was compared to levels of gene expression using previously-obtained Affymetrix microarray data. When gene expression was divided into deciles from low to high, genes with intermediate expression exhibited the highest median 5mC levels, and this trend was consistent for all tissues (Additional file 4). However, when mean 5mC levels were considered, the tissues had distinct expression/5mC profiles (Additional file 5). In all tissues, a near bell-shaped curve was observed, where genes with intermediate expression levels were higher in 5mC than weakly or highly expressed genes. At nearly all expression levels, calli had a higher median RPKM and greater RPKM range than did either explants or regenerated internodes (Additional file 4).
We compared expression levels of sets of genes with different 5mC status. Expression of genes with 5mC-enriched promoters did not differ significantly from unmethylated genes (Additional file 6). However, median expression of genes with 5mC-enrichment in both promoters and gene bodies was substantially lower than that of non-5mC-enriched genes in all tissues. In explants and regenerated internodes, genes that were 5mC-enriched in gene bodies and not in promoters had significantly lower expression than genes 5mC-enriched in both promoters and gene bodies (p < 0.05). In contrast, genes in this category in calli had the highest median expression of genes in any category in any tissue.
We examined genes related to hormone signaling including auxin, abscisic acid (ABA) and gibberellic acid (GA) signaling, as well as six categories of transcription factors related to stress response, many of which were differentially regulated during in vitro culture . In all cases, 5mC increased in callus tissue relative to explants, then decreased in regenerated internodes, but there was no clear relationship between RPKM and gene expression (data not shown).
Expression of DNA methyltransferases
To probe the causes of changes in 5mC among tissues, we examined the expression of annotated methyltransferase homologs. Of nine homologs of Arabidopsis DNA methyltransferase (MET1/2), CHROMOMETHYLASE 3 (CMT3), DECREASE IN DNA METHYLATION 1 (DDM1), and DOMAINS REARRANGED METHYLTRANSFERASE 1 and 2 (DRM1/2), three showed >1.5-fold expression changes during dedifferentiation and redifferentiation: POPTR_0019s00240 (MET1), POPTR_0004s14140 (MET2), and POPTR_0007s12710 (DDM1). Each showed an upward trend through both callus and regenerated internode stages (Additional file 7). Expression of one of the three DRM1/2 homologs (POPTR_0010s16200) showed a slight downward trend. Expression of the second DDM1 homolog (POPTR_0019s15030) showed a slight upward trend. Expression of the two CMT3 homologs (POPTR_0003s21520, POPTR_0003s21510), and two of the three DRM1/2 homologs (POPTR_0001s35160, POPTR_0014s04840), did not change appreciably (<50%).
Summary of changes in methylation during in vitro development for genome features
Explant to callus
Callus to regen
Whole genome (tiled windows)
Although we found some strong associations between expression and 5mC, we believe that these are minimal estimates of the strength and diversity of these associations. First, our work was done with a P. trichocarpa genotype that was not the same species as the P. tremula x tremuloides genotype that had been used for expression profiling. Second, only 27,753 of the genes represented on the expression array, which was based on the v.1.1 poplar genome assembly, could be cross-referenced to the 39,009 genes in the v2.2 poplar genome assembly that was used for short read mapping. Third, the short sizes of the 25 bp Affymetryx oligos used on the array, and sequence differentiation and high heterozygosity within and among species, create further obstacles to making direct comparisons and inferences. Fourth, the in vitro culture systems and hormones employed in the current study vs. the transcription profiling study differed in some important respects, including use of distinct basal media, and types and/or concentrations of plant hormones used during dedifferentiation and shoot differentiation steps. The extent of callus growth prior to visible shoot regeneration was also, on average, two months longer for P. trichocarpa Nisqually-1 compared to P. tremula x alba 717-1B4, and the frequency of shoot initiation per callus much less. Finally, these in vitro systems represent only a subset of the diversity of callogenic-organogenic systems used in poplar and related species. Commonly used systems employ distinct explant tissue types and diverse combinations of growth media, hormones, and environmental conditions. Thus, we believe that expression-5mC associations are likely to be both much stronger and highly diverse, and thus warrant further exploration.
Variation in genome methylation appeared to increase during in vitro culture. We found relatively low variance in 5mC among biological replicates from explant and callus tissues. In contrast, 5mC showed extremely high variance in biological replicates of regenerated internodes. This was obvious in hierearchical clustering where biological replicates from regenerated internodes did not cluster together as did replicates from explants and calli. It was also obvious in the much higher variances among biological replications observed for regenerated explants vs. the other tissue types. While some of this variability may be due to the particular tissue culture system employed for this work (with its long period of callus growth prior to shoot regeneration, as discussed above), this finding is consistent with several studies that have reported 5mC polymorphisms among plants regenerated from calli [9–11]. In Arabidopsis, clonal regenerants showed a range of mutations including base substitutions, insertions, and deletions that resulted in variant phenotypes . We did not maintain clonal regenerants over a long enough period to be able to observe mutant phenotypes. In addition, each biological replicate was composed of internodes from several regenerated plantlets, and a few individuals may have accounted disproportionately for the variation seen in the sample. Nonetheless, our results support the hypothesis that increased variation in 5mC contributes to the epigenetic instability often observed after in vitro regeneration.
Plant cell dedifferentiation involves two distinct processes: acquisition of pluripotency and reentry into the mitotic S phase. These processes are characterized by large changes in gene expression, which in turn are enabled by major changes in chromatin structure, especially chromatin decondensation . Histone methylation may be required for establishment and maintenance of a dedifferentiated cell state via upregulation of genes involved in the ubiquitin proteolytic pathway . Our results support this hypothesis, as analysis of 5mC-enriched genes specific to callus included GO categories related to intracellular protein transport, ubiquitin proteolysis, and protein catabolism.
We found an overall decrease in genome 5mC through callogenesis and redifferentiation. Both promoter and intergenic regions followed this trend. Previous studies of genomic 5mC content in dedifferentiating cells in Arabidopsis have reported mixed results [16, 17]. In rice, bisulfite genome sequencing showed that DNA methylation was modified in a wide variety of genome regions in both transgenic and in vitro regenerated plants . The overall level of 5mC in plant genomes is determined, in large part, by methyltransferase activity. In comparisons of 5mC with expression microarray data, we found that expression of three methyltransferase homologs was higher in callus tissue relative to explants, and still higher in regenerated tissue. Increases in methyltransferase expression coincident with genome-wide hypomethylation have been reported in oil palm calli , as well as in many different types of human cancers [33, 34], and have been described as an apparent paradox. However, this general trend may belie a great deal of genomic complexity in methylation patterns; as particular genomic features are becoming hypomethylated, others may be hypermethylated. The increased methyltransferase activity may also represent an adaptive response to hypomethylation, even if inadequate to fully recover methylation levels.
Transposable elements (TEs) are abundant features in plant genomes that are frequently marked by cytosine methylation. We examined four major categories of transposable elements from the RepBase transposable element annotation database. Copia, Gypsy and Ogre are long terminal repeat (LTR) retroelements, the predominant type of transposable element sequences in plant genomes . En/Spm (CACTA) is one of 19 families classified as eukaryotic “cut-and-paste” DNA TEs . Our results showed that all four TE categories had their highest 5mC content in regenerated tissue. The three retroelement categories increased in 5mC through dedifferentiation and redifferentiation; the EnSpm category did not have significantly increased 5mC in explant and callus stages, but did have a significant increase in 5mC in regenerated internodes. Previous studies using methylation-deficient mutants have demonstrated that hypomethylation can result in transposable element mobilization (Fujomoto et al., 2008). The increased 5mC in TEs in the present study suggests the operation of adaptive genomic mechanisms that might help protect genomes against transposable element spread. It would be of interest to measure whether there is indeed TE element mobilization during in vitro culture in future studies.
We found that chromosomal regions with higher gene density underwent gene body hypermethylation in callus tissue that was then released during regeneration. A similar result has been reported for Arabidopsis, in which cell suspension cultures exhibited hypermethylation of euchromatin . Bao et al.  reported large scale transcriptome reorganization during early callus induction; they catalogued >3,000 poplar genes with at least 5-fold changes in expression during callus induction, with roughly equal numbers of up-regulated and down-regulated genes. When we examined genes in terms of expression deciles, gene body 5mC in callus was higher than in the differentiated tissues in every decile. While gene body 5mC appeared to have a repressive effect on expression in explants and regenerated internodes, the transient body 5mC observed in calli did not appear to repress transcription (Additional file 6). Genes with moderate transcription levels had the highest overall 5mC, similar to what has been observed in Arabidopsis. Methylation patterns along plant genes showed minima around transcription start and termination sites, with higher levels over transcribed gene bodies [28, 37]. We found sharp increases in gene body methylation in callus tissue, followed by even sharper decreases in redifferentiated internode tissue. This transient gene body methylation was also correlated with higher chromosomal gene density (Figure 8 and 9). Hypermethylation of gene bodies may protect these genes against potential transposable element reactivation from TEs that reside in introns, or possibly from TE insertion into coding regions. In addition, gene body methylation may serve to prevent aberrant transcriptional initiation . In conditions of physiological stress such as during in vitro cellular dedifferentiation, gene body methylation may serve as a protective mechanism in chromosomal regions with higher gene density; as euchromatin decondenses, it may become more transcriptionally active and thus more prone to ectopic gene expression.
Cytosine DNA methylation has long been recognized as a mechanism for genome stabilization, playing a role in control of chromatin dynamics during development, suppressing transposable element activity, and repressing expression of particular genes at specific times during developmental or under specific environmental conditions. Here, we show that the protective role of 5mC is not necessarily at odds with gene expression, and in fact increases in regions of chromosomes where large numbers of genes are undergoing transcriptional upregulation during a developmental fate switch. Our results demonstrate great complexity in epigenomic processes when cells are subject to conditions that induce them to acquire pluripotency and redifferentiate. Given this complexity, it is not surprising that plants regenerated from dedifferentiated cells exhibit a wide variety of epigenomes and attendant phenotypic diversity.
Callus induction and shoot regeneration from Nisqually-1
Micro-cuttings of genotype Nisqually-1 were initially cultured on hormone-free WPM media (Lloyd and McCown 1981). Shoot cultures were maintained on this medium at 25°C under a 16-h photoperiod. Light was provided by fluorescent tubes (TL70, F25T8/TL735, Philips) at a photon flux density of 45 μE · m-2 s-1.
Three- to five- cm internode explants were collected from in vitro-grown, 40- to 50-day old poplar plantlets and cultured on callus induction medium (CIM-NB):MS + F vitamin + 1 mg/L NAA + 1 mg/L BA + Phytagel 1 g/L + Phytablend 1.5 g/L, pH 5.8). These were incubated in the dark for 30 days, and subcultured once on day 14. Resulting calli were then transferred to shoot induction medium (SIM-BN): MS + F vitamin + 0.1 mg/L NAA + 1 mg/L BA + Phytagel 1 g/L + Phytablend 1.5 g/L, pH 5.8) under light and subcultured three times at 3–4 week intervals, until shoots regenerated, ~ 4 months. Regenerated shoots were propagated on WPM until roots grew and shoots were substantial enough for sample collection, ~ 2 months. Regenerated internodes were collected at a single time point. Each biological replicate for internode explants and regenerated internodes consisted of tissue from 5–6 magenta boxes (25–30 shoots).
Total genomic DNA was isolated from three biological replicates of each tissue stage using a CTAB-based method, and the methylated DNA fraction was immunoprecipitated with an antibody to 5-methylcytidine, as described previously . A non-immunoprecipitated sample was sequenced as a control.
Illumina read trimming, filtering, normalization, and alignment to the P. trichocarpa V2.2 reference genome sequence were conducted as described previously . Reads per 1 kb target sequence per million reads mapped (RPKM) were calculated for 1 kb tiled genome windows and for genomic features (gene promoters, gene bodies, 5’ and 3’ UTRs, intergenic regions). To determine whether genome windows and genomic features were 5mC-enriched relative to the non-immunoprecipitated control, we modeled the counts of the sequencing reads using negative binomial distributions. A negative binomial distribution uses a dispersion parameter to capture the extra-Poisson variation that is often observed in sequencing read counts from independent biological samples. It has been observed in this study and in earlier RNA-Seq studies that the amount of dispersion often depends on 5mC or expression level, [38, 39]. We modeled the dispersion parameter as a smooth function of the mean relative frequency of the sequencing counts and fit a separate dispersion model to each group of plants. For comparing 5mC levels of features between two groups, we fit a negative binomial regression model to each feature using the indicator of group membership as a predicting variable, and identified differentially-5mC-enriched features by testing the corresponding regression coefficient. We computed test p-values using a likelihood ratio test with high-order asymptotic adjustment.
Basic data manipulations and statistical tests were performed using the R base package (http://www.r-project.org/). A Pearson’s product–moment correlation test was run to determine overall correlation between gene density and 5mC within each tissue, and over all tissues, using all biological replicates. To further examine the relationship between gene density and gene body 5mC, genes were binned in five gene density groups from low to high, and paired t-tests with average MeDIP-seq read countsover genes within each binwere used to make pairwise comparisons within each bin (explant vs. callus, explant vs. regen, callus vs. regen). Gene expression-5mC relationships were examined by filtering the set of genes with > =2-fold expression differences, and the 27,753 that could be cross-referenced, from available Affymetrix microarray data (Bao et al., 2009), and running pairwise t-tests on gene RPKM for each pair of tissues, assuming independence of gene values within biological replicates.Gene Ontology (GO) category enrichment for selected gene sets was determined using the AgriGO Singular Enrichment Analysis (SEA) tool (http://bioinfo.cau.edu.cn/agriGO/analysis.php), using the Fisher’s Exact Test statistical method with Benjamini and Hochberg multi-test correction option. RPKM means and variances for the sets of 1 kb tiled genome windows were calculated among biological replicates within each tissue.
This work was supported by the U.S. Department of Energy Plant Feedstock Genomics Program (# ER64665).
- Miguel C, Marum L: An epigenetic view of plant cells cultured in vitro: somaclonal variation and beyond. J Exp Bot. 2011, 62: 3713-3725. 10.1093/jxb/err155.PubMedView ArticleGoogle Scholar
- Dann AL, Wilson CR: Comparative assessment of genetic and epigenetic variation among regenerants of potato (Solanum tuberosum) derived from long-term nodal tissue-culture and cell selection. Plant Cell Rep. 2011, 30: 631-639. 10.1007/s00299-010-0983-9.PubMedView ArticleGoogle Scholar
- Rai MK, Kalia RK, Singh R, Gangola MP, Dhawan AK: Developing stress tolerant plants through in vitro selection—an overview of the recent progress. Environ Exp Bot. 2011, 71: 89-98. 10.1016/j.envexpbot.2010.10.021.View ArticleGoogle Scholar
- Grafi G: How cells dedifferentiate: a lesson from plants. Dev Biol. 2004, 268: 1-6. 10.1016/j.ydbio.2003.12.027.PubMedView ArticleGoogle Scholar
- Jaligot E, Adler S, Debladis É, Beulé T, Richaud F, Ilbert P, Finnegan EJ, Rival A: Epigenetic imbalance and the floral developmental abnormality of the in vitro-regenerated oil palm Elaeis guineensis. Ann Bot. 2011, 108: 1453-1462. 10.1093/aob/mcq266.PubMedPubMed CentralView ArticleGoogle Scholar
- Jaligot E, Rival A, Beule T, Dussert S, Verdeil JL: Somaclonal variation in oil palm (Elaeis guineensis Jacq.): the DNA methylation hypothesis. Plant Cell Rep. 2000, 19: 684-690. 10.1007/s002999900177.View ArticleGoogle Scholar
- Jaligot E, Beulé T, Rival A: Methylation-sensitive RFLPs: characterisation of two oil palm markers showing somaclonal variation-associated polymorphism. Theor Appl Genet. 2002, 104: 1263-1269. 10.1007/s00122-002-0906-4.PubMedView ArticleGoogle Scholar
- Matthes M, Singh R, Cheah S-C, Karp A: Variation in oil palm ( Elaeis guineensis Jacq.) tissue culture-derived regenerants revealed by AFLPs with methylation-sensitive enzymes. TAG Theor Appl Genet. 2001, 102: 971-979. 10.1007/s001220000491.View ArticleGoogle Scholar
- Gao X, Yang D, Cao D, Ao M, Sui X, Wang Q, Kimatu JN, Wang L: In vitro micropropagation of Freesia hybrida and the assessment of genetic and epigenetic stability in regenerated plantlets. J Plant Growth Regul. 2009, 29: 257-267.View ArticleGoogle Scholar
- Li X, Yu X, Wang N, Feng Q, Dong Z, Liu L, Shen J, Liu B: Genetic and epigenetic instabilities induced by tissue culture in wild barley (Hordeum brevisubulatum (Trin.) Link). Plant Cell, Tissue Org Cult. 2007, 90: 153-168. 10.1007/s11240-007-9224-5.View ArticleGoogle Scholar
- Peredo EL, Revilla MA, Arroyo-Garcia R: Assessment of genetic and epigenetic variation in hop plants regenerated from sequential subcultures of organogenic calli. J Plant Physiol. 2006, 163: 9-1071.View ArticleGoogle Scholar
- Kaeppler SM, Phillips RL: Tissue culture-induced DNA methylation variation in maize. Proc Natl Acad Sci USA. 1993, 90: 8773-8776. 10.1073/pnas.90.19.8773.PubMedPubMed CentralView ArticleGoogle Scholar
- Rodríguez López CM, Wetten AC, Wilkinson MJ: Progressive erosion of genetic and epigenetic variation in callus-derived cocoa ( Theobroma cacao) plants. New Phytol. 2010, 186: 856-868. 10.1111/j.1469-8137.2010.03242.x.PubMedView ArticleGoogle Scholar
- Avivi Y, Morad V, Ben-Meir H, Zhao J, Kashkush K, Tzfira T, Citovsky V, Grafi G: Reorganization of specific chromosomal domains and activation of silent genes in plant cells acquiring pluripotentiality. Dev Dyn. 2004, 230: 12-22. 10.1002/dvdy.20006.PubMedView ArticleGoogle Scholar
- Tanurdzic M, Vaughn MW, Jiang H, Lee T-J, Slotkin RK, Sosinski B, Thompson WF, Doerge RW, Martienssen RA: Epigenomic consequences of immortalized plant cell suspension culture. PLoS Biol. 2008, 6: 2880-2895.PubMedView ArticleGoogle Scholar
- Elhiti M, Tahir M, Gulden RH, Khamiss K, Stasolla C: Modulation of embryo-forming capacity in culture through the expression of Brassica genes involved in the regulation of the shoot apical meristem. J Exp Bot. 2010, 61: 4069-4085. 10.1093/jxb/erq222.PubMedPubMed CentralView ArticleGoogle Scholar
- Tessadori F, Chupeau M-C, Chupeau Y, Knip M, Germann S, Van Driel R, Fransz P, Gaudin V: Large-scale dissociation and sequential reassembly of pericentric heterochromatin in dedifferentiated Arabidopsis cells. J Cell Sci. 2007, 120: 1200-1208. 10.1242/jcs.000026.PubMedView ArticleGoogle Scholar
- Lisch D: Epigenetic regulation of transposable elements in plants. Annu Rev Plant Biol. 2009, 60: 43-66. 10.1146/annurev.arplant.59.032607.092744.PubMedView ArticleGoogle Scholar
- Fukai E, Umehara Y, Sato S, Endo M, Kouchi H, Hayashi M, Stougaard J, Hirochika H: Derepression of the plant chromovirus LORE1 induces germline transposition in regenerated plants. PLoS Genet. 2010, 6: e1000868. 10.1371/journal.pgen.1000868.PubMedPubMed CentralView ArticleGoogle Scholar
- Peschke VM, Phillips RL: Activation of the maize transposable element Suppressor-mutator (Spm) in tissue culture. Theoret. Appl. Genetics. 1991, 81: 90-97.View ArticleGoogle Scholar
- Kubis SE, Heslop-Harrison JS, Vershinin AV, Castilho AMMF: Retroelements, transposons and methylation status in the genome of oil palm (Elaeis guineensis) and the relationship to somaclonal variation. Plant Mol Biol. 2003, 52: 69-79. 10.1023/A:1023942309092.PubMedView ArticleGoogle Scholar
- Ngezahayo F, Xu C, Wang H, Jiang L, Pang J, Liu B: Tissue culture-induced transpositional activity of mPing is correlated with cytosine methylation in rice. BMC Plant Biol. 2009, 9: 91-91. 10.1186/1471-2229-9-91.PubMedPubMed CentralView ArticleGoogle Scholar
- Berdasco M, Alcázar R, García-Ortiz MV, Ballestar E, Fernández AF, Roldán-Arjona T, Tiburcio AF, Altabella T, Buisine N, Quesneville H, Baudry A, Lepiniec L, Alaminos M, Rodríguez R, Lloyd A, Colot V, Bender J, Canal MJ, Esteller M, Fraga MF: Promoter DNA hypermethylation and hene repression in undifferentiated Arabidopsis cells. PLoS One. 2008, 3: e3306. 10.1371/journal.pone.0003306.PubMedPubMed CentralView ArticleGoogle Scholar
- Su N, He K, Jiao Y, Chen C, Zhou J, Li L, Bai S, Li X, Deng XW: Distinct reorganization of the genome transcription associates with organogenesis of somatic embryo, shoots, and roots in rice. Plant Mol Biol. 2007, 63: 337-349. 10.1007/s11103-006-9092-0.PubMedView ArticleGoogle Scholar
- Bao Y, Dharmawardhana P, Mockler TC, Strauss SH: Genome scale transcriptome analysis of shoot organogenesis in Populus. BMC Plant Biol. 2009, 9: 132. 10.1186/1471-2229-9-132.PubMedPubMed CentralView ArticleGoogle Scholar
- Bao Y, Dharmawardhana P, Arias R, Allen MB, Ma C, Strauss SH: WUS and STM-based reporter genes for studying meristem development in poplar. Plant Cell Rep. 2009, 28: 947-962. 10.1007/s00299-009-0685-3.PubMedView ArticleGoogle Scholar
- Tuskan GA, DiFazio S, Jansson S, Bohlmann J, Grigoriev I, Hellsten U, Putnam N, Ralph S, Rombauts S, Salamov A, Schein J, Sterck L, Aerts A, Bhalerao RR, Bhalerao RP, Blaudez D, Boerjan W, Brun A, Brunner A, Busov V, Campbell M, Carlson J, Chalot M, Chapman J, Chen G-L, Cooper D, Coutinho PM, Couturier J, Covert S, Cronk Q, et al: The Genome of Black Cottonwood, Populus trichocarpa (Torr. & Gray). Science. 2006, 313: 1596-1604. 10.1126/science.1128691.PubMedView ArticleGoogle Scholar
- Vining KJ, Pomraning KR, Wilhelm LJ, Priest HD, Pellegrini M, Mockler TC, Freitag M, Strauss SH: Dynamic DNA cytosine methylation in the Populus trichocarpa genome: tissue-level variation and relationship to gene expression. BMC Genomics. 2012, 13: 27. 10.1186/1471-2164-13-27.PubMedPubMed CentralView ArticleGoogle Scholar
- Jiang C, Mithani A, Gan X, Belfield EJ, Klingler JP, Zhu J-K, Ragoussis J, Mott R, Harberd NP: Regenerant Arabidopsis Lineages Display a Distinct Genome-Wide Spectrum of Mutations Conferring Variant Phenotypes. Curr Biol. 2011, 21: 1385-1390. 10.1016/j.cub.2011.07.002.PubMedPubMed CentralView ArticleGoogle Scholar
- Grafi G, Ben-Meir H, Avivi Y, Moshe M, Dahan Y, Zemach A: Histone methylation controls telomerase-independent telomere lengthening in cells undergoing dedifferentiation. Dev Biol. 2007, 306: 838-846. 10.1016/j.ydbio.2007.03.023.PubMedView ArticleGoogle Scholar
- Stroud H, Ding B, Simon SA, Feng S, Bellizzi M, Pellegrini M, Wang G-L, Meyers BC, Jacoben SE: Plants regenerated from tissue culture contain stable epigenome changes in rice. eLife. 2013, 2: e00354. 10.7554/eLife.00354.PubMedPubMed CentralView ArticleGoogle Scholar
- Rival A, Jaligot E, Beulé T, Finnegan EJ: Isolation and expression analysis of genes encoding MET, CMT, and DRM methyltransferases in oil palm (Elaeis guineensis Jacq.) in relation to the “mantled” somaclonal variation. J Exp Bot. 2008, 59: 3271-3281. 10.1093/jxb/ern178.PubMedView ArticleGoogle Scholar
- Peedicayil J: The Role of DNA Methylation in the Pathogenesis and Treatment of Cancer. Curr Clin Pharmacol. 2012, 74: 333-340.View ArticleGoogle Scholar
- Woo HD, Kim J: Global DNA hypomethylation in peripheral blood leukocytes as a biomarker for cancer risk: a meta-analysis. PLoS One. 2012, 7: e34615. 10.1371/journal.pone.0034615.PubMedPubMed CentralView ArticleGoogle Scholar
- Kumar A, Bennetzen JL: Plant retrotransposons. Annu Rev Genet. 1999, 33: 479-532. 10.1146/annurev.genet.33.1.479.PubMedView ArticleGoogle Scholar
- Yuan Y-W, Wessler SR: The catalytic domain of all eukaryotic cut-and-paste transposase superfamilies. Proc Natl Acad Sci. 2011, 108: 7884-7889. 10.1073/pnas.1104208108.PubMedPubMed CentralView ArticleGoogle Scholar
- Zilberman D, Gehring M, Tran RK, Ballinger T, Henikoff S: Genome-wide analysis of Arabidopsis thaliana DNA methylation uncovers an interdependence between methylation and transcription. Nat Genet. 2007, 39: 61-69. 10.1038/ng1929.PubMedView ArticleGoogle Scholar
- Anders S, Huber W: Differential expression analysis for sequence count data. Genome Biol. 2010, 11: R106. 10.1186/gb-2010-11-10-r106.PubMedPubMed CentralView ArticleGoogle Scholar
- Di Y, Schafer DW, Cumbie JS, Chang JH: The NBP negative binomial model for assessing differential gene expression from RNA-seq. Stat Appl Genet Mol Biol. 2011, 10: 1-28.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.