Current Biology 19, 1677–1681, October 13, 2009 ª2009 Elsevier Ltd All rights reserved DOI 10.1016/j.cub.2009.08.053

ReportEpigenetic Resetting of a GeneImprinted in Plant Embryos

Stephanie Jahnke1 and Stefan Scholten1,*1Developmental Biology and Biotechnology, Biocenter Klein Flottbek, University of Hamburg, Ohnhorststraße 18, 22609 Hamburg, Germany

Summary

Genomic imprinting resulting in the differential expression

of maternal and paternal alleles in the fertilization productshas evolved independently in placental mammals and flow-

ering plants. In most cases, silenced alleles carry DNA meth-ylation [1]. Whereas these methylation marks of imprinted

genes are generally erased and reestablished in each gener-ation in mammals [2], imprinting marks persist in endo-

sperms [3], the sole tissue of reported imprinted geneexpression in plants. Here we show that the maternally

expressed in embryo 1 (mee1) gene of maize is imprintedin both the embryo and endosperm and that parent-of-

origin-specific expression correlates with differential allelicmethylation. This epigenetic asymmetry is maintained in

the endosperm, whereas the embryonic maternal allele is de-methylated on fertilization and remethylated later in embryo-

genesis. This report of imprinting in the plant embryoconfirms that, as in mammals, epigenetic mechanisms oper-

ate to regulate allelic gene expression in both embryonicand extraembryonic structures. The embryonic methylation

profile demonstrates that plants evolved a mechanism for

resetting parent-specific imprinting marks, a necessary pre-requisite for parent-of-origin-dependent gene expression in

consecutive generations. The striking difference betweenthe regulation of imprinting in the embryo and endosperm

suggests that imprinting mechanisms might have evolvedindependently in both fertilization products of flowering

plants.

Results and Discussion

Few imprinted genes are known in plants compared with mammals [1], so, in a strategy to discover imprinted genes ex­pressed during maize seed development, we used cDNA mi­croarrays and RNA from young microdissected embryos and endosperm to identify sequences that were expressed in reciprocal hybrids in a pattern resembling the expression level of either the maternal or the paternal parent (unpublished data). To date, all imprinted genes described in plants are either silent or show biallelic expression in embryos [4, 5]. We were thus surprised that the allele-specific expression analyses of one of our candidate sequences showed unambig­uous expression from only the maternal allele in both endo­sperm and embryos. We subsequently termed this gene maternally expressed in embryo 1 (mee1). mee1 was localized to chromosome 5 by sequence identity to the map-anchored maize bacterial artificial chromosome AC190547. The putative open reading frame of mee1 encodes a protein of 99 amino

*Correspondence: s.scholten@botanik.uni-hamburg.de

acids that shows significant homology to predicted or hypo­thetical proteins of mono- and dicotyledonous plants of similar length (see Figure S1 available online), indicating conservation of the protein among plant species.

Gene-Specific Imprinting in Plant Embryos

To test parent-specific expression, we identified polymor­phisms within mee1 transcripts in several inbred lines (see Table S1 for sequence details) and carried out cleaved ampli­fied polymorphic sequence analyses on material from hybrid embryos and endosperms 6 days after pollination (dap). Only maternal transcript sequences were detected in both fertiliza­tion products resulting from reciprocal crosses involving four independent genotypic combinations, confirming gene­specific imprinting of mee1 (Figure 1A).

To determine levels of mee1 expression throughout plant development, we carried out reverse transcriptase-poly­merase chain reaction (RT-PCR) analyses on RNA extracted from somatic and isolated reproductive cells. mee1 expres­sion is highly specific to central cells and to the fertilization products in the early seed. It is not expressed prior to fertiliza­tion in either egg or sperm cells (Figures 1B and 1C). In embryos, we detected expression after fertilization between 3 and 8 dap. In contrast, mee1 is active in central cells and continues expression until 10 dap with a peak at 6 dap in endo­sperms (Figure 1C; Figure S2). In situ hybridizations of 6 dap seeds confirmed that expression of mee1 is exclusive to the endosperm and embryo, with transcripts distributed through­out the embryo (Figure 1D).

Allelic expression pattern of mee1 during fertilization and subsequent development was further explored by using homogenous MassEXTEND (hME) chemistry to generate single base extensions, followed by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) spectros­copy. Transcripts were detected in the embryo starting at 3, 6, and 8 dap, but not at 14 dap. Over the entire time course of transcriptional activity, the maternal allele contributed 96.6% (mean value) of total mee1 expression (Figure 1E; Table S2). This expression pattern differs strikingly from 25 genes distributed throughout the maize genome with equivalent parental contributions in the same samples [6] and also from a mee1 neighboring gene as revealed by allele-specific sequencing (Figure S3). The highly specific maternal expres­sion pattern, combined with the fact that transcripts are not provided maternally by the egg cell, clearly identifies mee1as the first reported imprinted gene in plant embryos.

Differential Methylation Correlates with Allele-Specific

ExpressionMany imprinted genes are marked by differences in cytosine methylation at CpG dinucleotides located within differentially methylated regions (DMRs) [7]. In plants, cytosine methylation is required for epigenetic inheritance during gametogenesis [8] and is implicated in the control of imprinting in the endosperm [3, 9–11]. We cloned and sequenced the genomic region upstream of mee1 transcript and determined the transcrip­tional start site of the gene (Figure 2A). To investigate whether any allelic methylation differences were associated with the

1678 Current Biology Vol 19 No 19

Figure 1. Imprinted Expression of mee1 in Filial Tissue

(A) Cleaved amplified polymorphic sequence analyses of parental inbred lines and reciprocal hybrid cDNAs indicate maternal transcripts only in embryo

(Emb) and endosperm (End). All fragments were treated with HaeII. (+), restricted fragments (268 and 68 bp; note that the 68 bp fragment is not shown);

(2), unrestricted fragment (336 bp) of the inbred lines. Genotypes are as follows: B, B73; H, H99; T, Tx303; W, W22; M, Mo17; A, A188. In hybrid tissue

descriptions, the maternal parent is denoted first.

(B and C) RT-PCR analyses showing exclusive expression of mee1 in early embryo and endosperm development. actin expression was used as a positive

control. The following abbreviations are used: EC, egg cell; ZY, zygote; CC, central cell; PE, primary endosperm.

(D) In situ hybridizations showing transcripts of mee1 throughout the embryo with antisense probe. Sense probe was used as a negative control.

(E) Allele-specific expression analysis showing transient and exclusively maternal expression of mee1 in early maize development. Relative allelic expres­

sion levels of mee1 in the reciprocal F1 hybrid embryos and endosperm 0053301 and 3013005 (the maternal line is denoted first) at 1, 3, 6, 8, 14, and 6 days

after pollination (dap), respectively, are shown. Transcript or genomic DNA (gDNA) amounts of both alleles were measured and displayed as a percentage of

the total expression level. Black and white bar fractions indicate transcript abundance of the 005 and the 301 allele, respectively. The y axis scale refers to the

portion of the 005 allele. Mixes of gDNA (1:4 and 4:1) of the inbred lines (005:301) were used to control the assay performance concerning allelic ratio repro­

duction. gDNA from the inbred lines 005 (1:0) and 301 (0:1) was used to define the thresholds of each assay for both alleles. Hybrid gDNA (1:1 allelic ratio) was

used to normalize the data. The middle black line indicates 50% allelic proportion.

mee1 locus, we carried out an allele-specific methylation screen by bisulfite sequencing the genomic region spanning 21161 to +412 of mee1 relative to the transcriptional start in reciprocal hybrid endosperms 6 dap. Differential methyla-tion of the parental alleles was discovered throughout the genomic region analyzed. Methylation of active maternal alleles was low, whereas silent paternal alleles were highly methylated at CpG and CpNpG cytosines. Asymmetric CpNpN cytosines were unmethylated (Figure 2; Table S3). The greatest methylation differences were detected in the vicinity of the transcriptional start and within the transcribed region of mee1, with methylation of maternal alleles virtually absent (Figure 2B; Table S3). This region can thus be defined as a DMR correlated with allele-specific expression of mee1 in endosperm, as reported for other imprinted maize genes [12, 13].

Epigenetic Resetting in Plant EmbryosMethylation of imprinted genes in mammals is generally erased and reestablished in each generation [2], but imprinting marks persist in plant endosperms [3]. Such one-way control

of imprinting without resetting is feasible because the endo­sperm is terminally differentiated. Because mee1 also shows imprinting in cells of the embryo, which contribute to the next sporophytic generation (Figure 1D), some form of reset­ting of imprinting marks must occur. We therefore investigated allelic methylation of the mee1 DMR in gametes before fertil­ization, during embryo and endosperm development, and in seedlings postgermination.

Paternal alleles were found to be methylated in gametes, in both fertilization products, and at all stages analyzed, consis­tent with the exclusively maternal expression of mee1 (Figures 2C–2E). Single maternal DMRs were demethylated in central cells (Figures 2C and 2E) and collectively demethylated in the endosperm (Figures 2D and 2E), reflecting a current model for imprinting regulation in plants where silencing through methylation is relieved by a combination of targeted DNA gly­cosylase activity in the central cell [3, 14, 15] and suppression of methyltransferase 1 [11]. The persistence of this situation in later (16 dap) endosperms (Figures 2D and 2E) reflects the one-way control of imprinting [3]. In contrast, although DMRs of mee1 were methylated in egg cells, consistent with absence

1679 Imprinting in Plant Embryos

Figure 2. Methylation Profile of mee1

(A) Schematic representation of the mee1 locus. Numbers given are relative to the transcriptional start, indicated by the arrowhead. The 50 untranslated

region and the 50 portion of the coding region are shown by white and black boxes, respectively.

(B–E) Bisulfite sequencing analyses of genomic regions corresponding to (A). Relative positions of CpG and CpNpG sites are represented by circles and

triangles, respectively.

(B) Percentage of maternal and paternal allelic methylation of 6 dap hybrid endosperm. Gray signs indicate polymorphisms impeding analysis.

(C and D) Methylation states of individual sites of the differentially methylated region (+221 to +412) are represented by black (methylated) or white (unme­

thylated) signs; each line represents an individual clone.

(E) Aggregation of the data in (C) and (D). The full data set, including CpNpN sites, is given in Tables S3 and S4.

of mee1 transcripts (Figure 2C), they were rapidly demethy­lated after fertilization, with completely unmethylated DMRs detectable in zygotes 16–20 hr after fertilization (Figures 2D and 2E). This must result from active demethylation because unmethylated DMRs are present in zygotes prior to the first cell division. These findings reveal a new level of complexity of imprinting regulation in plants, because this demethylation in the embryo implies recognition of allele specificity by parental imprinting marks other than DNA methylation, which is not required for the demethylation of imprinted sequences prior to fertilization by DEMETER [14].

Importantly, maternal DMRs are then remethylated during embryo development, effectively resetting their epigenetic state to prefertilization levels of methylation. The decrease of the portion of demethylated DMRs between 6 and 16 dap is significant (c2 test, a % 0.025). In addition, fully demethylated DMRs were no longer detectable by 16 dap, and maternal DMRs were fully methylated in postgermination seedlings (Figures 2D and 2E; Table S4).

A comparison of the methylation and expression profiles indicates an additional level of developmental regulation of mee1 because demethylation does not necessarily lead to expression but seems to provide a permissive state for activity. Whereas the methylated state was strictly correlated with silence in all stages analyzed, mee1 was not expressed

in zygotes and late endosperm (Figures 1C and 1E), al­though the maternal alleles were demethylated (Figures 2D and 2E).

Recently, DNA demethylation relative to embryos and other tissues was shown to be a genome-wide phenomenon in Ara-bidopsis endosperm from torpedo-stage seeds [16, 17], and five new imprinted genes could be identified based on their reduced methylation and preferential expression in endo­sperm [16]. Reduced methylation levels in endosperm relative to embryos were also found in maize [18], signifying immense differences in the epigenetic landscape of the two fertilization products in higher plants.

The data reported here reveal that gene-specific imprinting occurs in plant embryos. The cycle of demethylation and re­methylation of maternal mee1 DMRs in the embryo represents a resetting of parent-specific imprinting marks, a necessary prerequisite for parent-of-origin-dependent gene expression in consecutive generations. The transient nature of the differ­ential methylation profile adds support to the view that the default state of some imprinted genes is to be methylated at key domains [3, 5]. It has been proposed for mammals that imprinting mechanisms might have evolved independently in embryonic and extraembryonic tissues [1, 19]; the striking difference between the regulation of imprinting in the embryo and endosperm suggests that imprinting mechanisms might

1680 Current Biology Vol 19 No 19

also have evolved independently in these corresponding structures of plants.

Experimental Procedures

Plant Material and Growth Conditions

The seeds of the inbred lines A188 (Ames22443), Mo17 (Pi558532), W22

(NSL30053), Tx303 (Ames19327), H99 (PI587129), and B73 (PI550473)

were received from the North Central Regional Plant Introduction Station

(Ames, IA). The inbred lines UH005 (national listing of plant varieties:

M9379, European flint) and UH301 (M8652, Iodent) were obtained from

A. Melchinger (University of Hohenheim, Stuttgart, Germany). Plant growth

conditions and crossbreeding for the production of hybrid and inbred mate-

rial as well as microdissection of 6 and 8 dap embryos and endosperm were

as described previously [20]. Microdissected embryo samples, used for

allele-specific expression and DNA methylation analyses, were shown to

be free of contamination with endosperm tissue by RT-PCR with primers

against the endosperm-specific gene Zmfie1 (Figure S4). Quantitative

RT-PCR analyses confirmed a comparable mee1 expression level in 6 dap

embryo and endosperm (Figure S2). Isolation of egg cells, zygotes, 3 dap

embryos, sperm cells, central cells, and primary endosperm was performed

by microdissection techniques as described by Kranz et al. [21, 22].

RT-PCR Analyses

For expression analysis of mee1, frozen material was homogenized in liquid

nitrogen, and mRNA isolation and cDNA synthesis were essentially con­

ducted as described by Meyer et al. [20]. Subsequent PCRs were performed

via standard procedures with 33 or 36 cycles. Actin primers served as a posi­

tive control. Primer sequences are available upon request.

Transcriptional Start Site Determination

We used 50 rapid amplification of cDNA ends (50RACE; Clontech, Takara) to

determine the transcriptional start site of mee1 according to the manufac­

turer’s protocol with amplified cDNA (SMART cDNA Synthesis Kit; Clontech,

Takara) as template. Determination of the transcriptional start site was per­

formed by alignment of the obtained cDNA sequences with the genomic

sequence of mee1.

Allele-Specific Expression Analysis of mee1

Relative gene expression analysis was performed on the MassARRAY

system (Sequenom) applying hME biochemistry and MALDI-TOF mass

spectrometry for analyte detection. Reactions were conducted according

to the standard protocol, generating the allele-specific analytes in a primer

extension reaction with a primer directly adjacent to the polymorphic site.

Assay design was carried out with platform-specific software for polymor­

phic sequences. We used the classical hME design because it suits the anal­

ysis of samples with one excessively expressed allele best. Primer

sequences of the assay are shown in Figure S5. All samples were analyzed

with three biological replicates. As ratio controls, genomic DNA (gDNA)

mixes (0:1, 1:4, 4:1, and 1:0) of the inbred lines (UH005 and UH301) and

gDNA of the hybrid 3013005 was used. gDNA isolation and further process­

ing of all samples were as described previously [6]. Four data sets (spectra)

were acquired for each sample on a MassARRAY Analyzer Compact fol­

lowed by automated data analysis with TYPER RT software version 3.4

(Sequenom). No hME primer extension occurred in all negative control reac­

tions with H2O instead of template, confirming contamination-free chem­

istry. mee1 was considered expressed if at least one-third of the spectra

of a given cDNA sample delivered clear results. Data were normalized by

hybrid gDNA ratios. Mean values of relative allelic expression and standard

deviations as well as the number of evaluable spectra are shown in Table S2.

Allele-Specific Methylation Analyses

Allele-specific methylation analyses were performed with samples of recip­

rocal hybrids of the inbred lines B73 and H99; for seedling samples, the

inbred lines UH301 and UH005 were used. Bisulfite conversion of isolated

gDNA from embryos and endosperm 6 dap or at later stages was performed

with the MethylCode Bisulfite Conversion Kit (Invitrogen). Bisulfite conver­

sion of gDNA from zygotes and gametes was performed with 20 to 100 cells

in each reaction with the EZ DNA Methylation-Direct Kit (Zymo). After bisul­

fite conversion, specific regions of mee1 and control genes were amplified

under standard PCR reactions with Taq Polymerase (Fermentas) or alterna­

tively with Advantage 2 Polymerase Mix (Takara) when the material was

limited. The PCR products were cloned in the pGEM-T Easy Vector (Prom­

ega) and transformed in XL1 blue cells. Inserts of individual colonies were

reamplified and sequenced. All sequences were analyzed with BiQ Analyzer

software [23] for quality control and removal of identical clones in a stan­

dardized manner. Further sequence details and the full methylation data

sets are given in Tables S3 and S4. Primer sequences and the polymorphism

used for allele discrimination are given in Table S5. A c2 test was performed

to revise the assumption that the methylation status of mee1 maternal

alleles in embryos is independent of the point of time after fertilization. We

considered DMRs with less than 50% methylation to be demethylated. A

total of 46 maternal DMRs were evaluated, 28 of 6 dap embryos and 18 of

16 dap embryos.

In Situ Hybridization

Nonradioactive in situ hybridizations were performed as described by Nard­

mann et al. [24].

Accession Numbers

The mee1 sequence reported in this paper has been deposited in GenBank

with the accession number FJ477242.

Supplemental Data

Supplemental Data include five tables and five figures and can be found with

this article online at http://www.cell.com/current-biology/supplemental/

S0960-9822(09)01630-3.

Acknowledgments

We thank M. Nissen, P. von Wiegen, A. Tran, and D. Rybka for excellent

technical assistance. We are grateful to H. Dickinson for critical reading of

the manuscript. This work was funded by the Deutsche Forschungsgemein­

schaft.

Received: January 7, 2009

Revised: August 11, 2009

Accepted: August 12, 2009

Published online: September 24, 2009

References

1. Feil, R., and Berger, F. (2007). Convergent evolution of genomic

imprinting in plants and mammals. Trends Genet. 23, 192–199.

2. Surani, M.A. (2001). Reprogramming of genome function through epige­

netic inheritance. Nature 414, 122–128.

3. Kinoshita, T., Miura, A., Choi, Y., Kinoshita, Y., Cao, Y., Jacobsen, S.E.,

Fischer, R.L., and Kakutani, T. (2004). One-way control of FWA

imprinting in Arabidopsis endosperm by DNA methylation. Science

303, 521–523.

4. Scott, R.J., and Spielman, M. (2006). Genomic imprinting in plants and

mammals: How life history constrains convergence. Cytogenet.

Genome Res. 113, 53–67.

5. Huh, J.H., Bauer, M.J., Hsieh, T.-F., and Fischer, R.L. (2008). Cellular

programming of plant gene imprinting. Cell 132, 735–744.

6. Meyer, S., and Scholten, S. (2007). Equivalent parental contributions to

early plant zygotic development. Curr. Biol. 17, 1686–1691.

7. Reik, W., and Walter, J. (2001). Genomic imprinting: parental influence

on the genome. Nat. Rev. Genet. 2, 21–32.

8. Saze, H., Scheid, O.M., and Paszkowski, J. (2003). Maintenance of CpG

methylation is essential for epigenetic inheritance during plant gameto­

genesis. Nat. Genet. 34, 65–69.

9. Vielle-Calzada, J.P., Thomas, J., Spillane, C., Coluccio, A., Hoeppner,

M.A., and Grossniklaus, U. (1999). Maintenance of genomic imprinting

at the Arabidopsis medea locus requires zygotic DDM1 activity. Genes

Dev. 13, 2971–2982.

10. Jullien, P.E., Kinoshita, T., Ohad, N., and Berger, F. (2006). Maintenance

of DNA methylation during the Arabidopsis life cycle is essential for

parental imprinting. Plant Cell 18, 1360–1372.

11. Jullien, P.E., Mosquna, A., Ingouff, M., Sakata, T., Ohad, N., and Berger,

F. (2008). Retinoblastoma and its binding partner MSI1 control

imprinting in Arabidopsis. PLoS Biol. 6, e194.

12. Gutie rrez-Marcos, J.F., Costa, L.M., Dal Pra, M., Scholten, S., Kranz, E.,

Perez, P., and Dickinson, H.G. (2006). Epigenetic asymmetrie of

imprinted genes in plant gametes. Nat. Genet. 38, 876–878.

Imprinting in Plant Embryos1681

13. Haun, W.J., Laoueille , S., O’Connell, M.O., Spillane, C., Grossniklaus, U.,

Phillips, A.R., Kaeppler, S.M., and Springer, N.M. (2007). Genomic

imprinting, methylation and molecular evolution of maize Enhancer of

zeste (Mez) homologs. Plant J. 49, 325–337.

14. Choi, Y., Gehring, M., Johnson, L., Hannon, M., Harada, J.J., Goldberg,

R.B., Jacobsen, S.E., and Fischer, R.L. (2002). DEMETER, a DNA glyco­

sylase domain protein, is required for endosperm gene imprinting and

seed viability in Arabidopsis. Cell 110, 33–42.

15. Xiao, W., Gehring, M., Choi, Y., Margossian, L., Pu, H., Harada, J.H.,

Goldberg, R.B., Pennell, R.I., and Fischer, R.L. (2003). Imprinting of

the MEA polycomb gene is controlled by antagonism between MET1

methyltransferase and DME glycosylase. Dev. Cell 5, 891–901.

16. Gehring, M., Bubb, K.L., and Henikoff, S. (2009). Extensive demethyla­

tion of repetitive elements during seed development underlies gene

imprinting. Science 324, 1447–1451.

17. Hsieh, T.-F., Ibarra, C.A., Silva, P., Zemach, A., Eshed-Williams, L.,

Fischer, R.L., and Zilberman, D. (2009). Genome-wide demethylation

of Arabidopsis endosperm. Science 324, 1451–1454.

18. Lauria, M., Rupe, M., Guo, M., Kranz, E., Pirona, R., Viotti, A., and Lund,

G. (2004). Extensive maternal DNA hypomethylation in the endosperm of

Zea mays. Plant Cell 16, 510–522.

19. Lewis, A., Mituya, K., Umlauf, D., Smith, P., Dean, W., Walter, J., Higgins,

M., Feil, R., and Reik, W. (2004). Imprinting on distal chromosome 7 in

the placenta involves repressive histone methylation independent of

DNA methylation. Nat. Genet. 36, 1291–1295.

20. Meyer, S., Pospisil, H., and Scholten, S. (2007). Heterosis associated

gene expression in maize embryos 6 days after fertilization exhibits

additive, dominant and overdominant pattern. Plant Mol. Biol. 63,

381–392.

21. Kranz, E., von Wiegen, P., Quader, H., and Lo rz, H. (1998). Endosperm

development after fusion of isolated, single maize sperm and central

cells in vitro. Plant Cell 10, 511–524.

22. Kranz, E., Bautor, J., and Lo rz, H. (1991). In vitro fertilization of single,

isolated gametes of maize mediated by electrofusion. Sex. Plant

Reprod. 4, 12–16.

23. Bock, C., Reither, S., Mikeska, T., Paulsen, M., Walter, J., and Lengauer,

T. (2005). BiQ Analyzer: Visualization and quality control for DNA meth­

ylation data from bisulfite sequencing. Bioinformatics 21, 4067–4068.

24. Nardmann, J., Zimmermann, R., Durantini, D., Kranz, E., and Werr, W.

(2007). WOX gene phylogeny in Poaceae: A comparative approach ad­

dressing leaf and embryo development. Mol. Biol. Evol. 24, 2474–2484.

Current Biology, Volume 19

Supplemental Data

Epigenetic Resetting of a Gene

Imprinted in Plant Embryos Stephanie Jahnke and Stefan Scholten

Table S1. Sequence Details of the Maize Inbred Lines Used for Cleaved Amplified Polymorphic Sequence Assays Inbred Abbreviation in Figure 1A Base at +237 HaeII restriction site H99 H A No B73 B G Yes W22 W A No Tx303 T G Yes A188 A A No Mo17 M G Yes Base position is given relative to the transcriptional start site of mee1 (GeneBank Accession Number: FJ477242).

Table S2. Relative Allelic Expression Data of mee1 Sample type Genotype Number of Mean value Mean value Standard deviation

evaluable allele 005 allele 301 spectra

1 dap EMB 005 x 301 0 NE NE 1 dap EMB 301 x 005 1 NE NE 3 dap EMB 005 x 301 8 1 0 0 3 dap EMB 301 x 005 4 0.098 0.902 0.019 6 dap EMB 005 x 301 12 0.996 0.004 0.006 6 dap EMB 301 x 005 12 0.034 0.966 0.05 8 dap EMB 005 x 301 12 0.984 0.016 0.022 8 dap EMB 301 x 005 12 0.113 0.887 0.138 14 dap EMB 005 x 301 3 NE NE 14 dap EMB 301 x 005 0 NE NE 6 dap END 005 x 301 12 1 0 0 6 dap END 301 x 005 12 0 1 0

Numbers of evaluable spectra as well as mean values of relative allelic expression and standard deviations are given for each sample type of both genotypes (005x301, 301x005). NE, not expressed.

Table S3. Percentage of Methylation at CpG, CpNpG, and CpNpN Sites of mee1 Genomic Regions in 6 dap Endosperm Analyzed by Bisulfite Sequencing Genomic Allele Number of % CpG % CpNpG % CpNpN region clones (Number of sites) (Number of sites) (Number of sites)

(BxH/HxB) -1161 to -748 paternal 6/5 78.6 (4) 77.3 (4) 1.2 (62)

maternal 5/8 23.9 (4) 11.5 (4) 1.7 (62) -789 to -251 paternal 8/4 93.9 (6) 94.5 (5) 1.4 (70)

maternal 15/10 37.0 (6) 39.9 (5) 1.8 (70) -471 to -163 paternal 4/5 92.1 (3) 92.5 (4) 2.0 (44)

maternal 5/5 40.0 (3) 27.9 (4) 1.9 (44) -222 to +113 paternal 7/1 78.6 (1) 54.2 (3) 1.6 (57)

maternal 9/9 2.8 (1) 1.9 (3) 1.5 (57) +62 to +412 paternal 4/2 87.4 (9) 78.0 (7) 1.7 (55)

maternal 6/2 2.5 (9) 2.7 (7) 1.6 (55) Sequence positions relative to the transcriptional start and numbers of clones from reciprocal hybrids are given.

Table S4. Percentage of Methylation at CpG, CpNpG, and CpNpN Sites of the mee1 DMR and Control Genes Analyzed by Bisulfite Sequencing Sample Allele Clonesa

(n) Mee1 DMR +221 to +412

180bp repeatc

248 to 362 Opaquec

-240 to -54 8 CpG

7 CpNpG

24 CpNpN

5 CpG

24 CpNpN

5 CpG

4 CpNpG

38 CpNpN

Egg cell 10 98.8 98.6 2.2 - - - - -Sperm cell 13 94.2 95.6 6.0 - - - - -Central cell 23 73.4 72.7 2.3 - - - - -Zygote p 5/3 92.2 98.2 4.9 - - - - -

m 9/4 33.3 32.1 2.2 Embryo 6 dap p 13/14 93.1 88.4 2.7 72.9 1.1 0.0 1.1 0.3

m 14/14 59.8 58.7 2.5 Embryo 16 dap p 9/7 87.5 89.3 2.5 73.2 1.5 0.0 1.1 0.5

m 8/10 84.0 81.0 2.7 Seedlingb p 6/8 77.7 51.0 1.9 - - - - -

m 6/4 92.5 74.3 2.6 Endosperm 6dap p 7/6 90.4 78.0 1.3 78.0 3.3 2.3 0.0 0.2

m 10/6 2.3 2.7 0.8 Endosperm 16dap p 5/17 89.8 61.7 1.6 78.9 2.4 0.0 1.0 0.5

m 9/9 2.1 2.4 1.7 aClone numbers of reciprocal hybrids (BxH/HxB) are given in hybrid samples. bHybrid seedlings of UH301/UH005 genotype combination were used. cA minimum of 8 clones of each reciprocal hybrid were analyzed. Accession numbers of the control genes are CC339714 for the 180 bp repeat and X15544 for opaque2. Position information for the 180bp repeat is relative to the plain sequence, for opaque relative to the translational start.

Table S5. Polymorphisms for Allelic Discrimination and Primer Sequences Used in Methylation Analyses Genomic region Forward primer 5’-3’ Reverse primer 5’-3’ Polymorphism

H99/B73 Position

mee1

-1161 to -748

mee1

-789 to -251

mee1

-471 to -163

mee1

-222 to +222a

mee1

+62 to +412

mee1

+221 to +481b

180bp repeat

248 to 362

opaque

-240 to -54

ttggtgggttaaggatattaggtta

ttttatgaattagtttaattaaaattttag

tgatagagataataggttttttttatgtat

ggtttaattttttgttattttgataaaaa

ttttttatttttaaaatttttaggtaat

ttatgtttagttatgagtttgtggaggaa

tatatgtggggtgagatgtatgagtttttg

ttgtagggaggtatagtaagagagagagt

aattcttattcaaaaaacccttacctt

ttatcaaaataacaaaaaattaacc

aaaaatttactaatacatcatcttaattt

cttccatcaataccaccattattac

aaaaactccttactcaccatacaatac

ttcacttactctaaccatacttctcttcaata

ttttccttaccataaatcacattcttaaattt

aattccaaacaaaaatcaaaaaaaa

G/T

G/T

G/A

A/T

CA/GG

CA/GG

G/A

A/G

--/AG

--/AG

G/T

G/T

A/Gc

A/Gc

-

-

-1124

-977

-912

-856

-786/5

-786/5

-720

-613

-316/7

-316/7

-166

-166

+238

+238

-

-

Sequence positions of mee1 are given relative to the transcriptional start. Position information for the 180bp repeat (CC339714) is relative to the plain sequence, for opaque (X15544) relative to the translational start. a -222 to-113 of this fragment was analysed. b +221 to +412 of this fragment was analysed. cThe same polymorphism is present in UH301/UH005 genotypes.

Figure S1. Alignment of Putative MEE1 Homolog Proteins Indicates Conservation of the Protein among Plant Species Based on BLASTP searches at GenBank predicted or hypothetical proteins of the following species were chosen and aligned with the putative MEE1 using Clustal W, DNASTAR MegAlign 5.05 software: Populus trichocarpa (XP_002313600, 106 aa, 8 x 10-5), Ricinus communis (EEF32221, 195 aa, 5 x 10-5), Arabidopsis thaliana (NP_566393, 117 aa, 1 x 10-4), Vitis vinifera (XP_002281204, 120 aa, 7 x 10-4), Oryza sativa (NP_001054311, 81 aa, 2 x 10-3). Accession numbers, protein length and expectation values are given in brackets.

Figure S2. Quantification of mee1 Transcript Levels in Embryo and Endosperm by Real-Time PCR

Each bar represents the mee1 expression level normalized against actin expression levels in the samples indicated. Isolation of mRNA and cDNA synthesis were conducted as described by Meyer et al. [17]. Subsequent real-time PCRs were performed using the primers 5’­AAGGAGCTCCCAATCTGCATCAACT and 5’-CTTGCTAGCCTCCCTTCACTTACTCT for mee1 and 5’-TCCTGACACTGAAGTACCCGATTGA and 5’­CGTTGTAGAAGGTGTGATGCCAGTT for actin at 58° annealing temperature. Each measurement was done in three replicates. For 3 dap embryos the mee1 expression level of the only one sample, which produced a signal, is shown.

Figure S3. Allele-Specific Sequencing Shows Biallelic Expression of a mee1 Neighboring Gene

The 40S ribosomal protein S24 gene (NM_001153496) showed biallelic expression in embryo and endosperm 6 days after pollination (dap). It is localized on the same maize genomic BAC (AC155410) as mee1, confirming a distance of less than 129 kb for the two genes. Two single nucleotide polymorphisms (SNP) between the inbred lines 005 (005 x 005) and 301 (301 x 301) within the transcript of the 40S ribosomal protein S24 gene (boxed) were identified and used to discriminate the alleles in embryo (EMB) and endosperm (END) of reciprocal crosses (301 x 005 and 005 x 301). In both tissues maternal and paternal alleles are represented in the sequence for both crosses, indicated by the double peaks. The same 6 dap cDNA samples as for allele-specific expression analyses of mee1 (shown in Figure 1e of the report) were used. PCRs were performed with the primers 5’-TTCATGACCAACCGCCTCCTCT and 5’-TTGCCGCCTCCGAAGTGG at 59°C annealing temperature.

Figure S4. Control Amplifications of the Endosperm-Specific fie1 Gene to Test for Contamination of Embryo Samples with Endosperm Tissue The intron spanning primers 5’-GTGCCAAAGTGTTCATTATGGTTTATG and 5’­TTGATCAGGTTTGGAGCTGGAAGC for fie1 [S1] and 5’-TCAACCCCAAGGCCAACAGAG and 5’-GAGGTCACGCCCCGCAAGAT for actin were used in standard PCR reactions with 2μl of the cDNA samples indicated and 50 ng gDNA as template. Actin was used as a positive control. The annealing temperature and cycle number was 60°C and 36 for fie1, 63°C and 33 for actin, respectively. All samples tested were used for allele­specific expression analyses. All samples of the H99/B73 genotype combination were used additionally for DNA methylation analyses by bisulfite sequencing. In these samples the gDNA was precipitated for bisulfite conversion after mRNA isolation for cDNA synthesis.

TCCCAATCTGCATCAACTGCAAGCAGCC[C/T]TGCTTTATCGGCTTCAACACTGAAGAG AAGCATGGCTAGAGTAAGTGAAGGGAGGCTAGCAAGCTAGAGCTTATTGGACTTGGTC TATTTGGTTCAATAGTAGGCTGGCTAGTAGAGTCTAAGCTTAGTTTCCTATGTGGAGTAA GTTTTGTTCTTTTATGGGGTATGGACTAGGTGGTTCTTAATGTTGTGG

Figure S5. The mee1 Sequenom Assay Design The mee1 genomic region +436 to +657, relative to the transcriptional start site, is shown. PCR primer regions are indicated bold, underlined, and italic; analyzed polymorphic positions are indicated in bold, red and brackets; the hME primer position is gray boxed and underlined.

Supplemental Reference S1. Danilevskaya, O. N., Hermon, P., Hantke, S., Muszynski, M. G., Kollipara, K. and Ananiev, E.V. (2003).

Duplicated fie genes in maize: Expression pattern and imprinting suggest distinct functions. Plant Cell 15, 425-438.