Histone H3 Methylated at Arginine 17 Is Essential for ... - Cell Press

Report

Histone H3 Methylated at Arginine 17 Is Essential for Reprogramming the Paternal Genome in Zygotes Graphical Abstract

Authors Yuki Hatanaka, Takeshi Tsusaka, Natsumi Shimizu, ..., Kazuya Matsumoto, Yoichi Shinkai, Atsuo Ogura

Correspondence [email protected] (Y.H.), [email protected] (K.M.), [email protected] (A.O.)

In Brief Maternal factors mediate nucleosome assembly and active DNA demethylation of the paternal genome during reprogramming in mouse oocytes; however, the underlying mechanisms are poorly described. Hatanaka et al. show that maternally provided H3R17me2a is responsible not only for H3.3 incorporation but also for active DNA demethylation in the paternal genome.

Highlights d

Mettl23 is an arginine methyltransferase that catalyzes H3R17me2a in mouse oocytes

d

H3R17me2a is responsible for H3.3 incorporation and active DNA demethylation

d

Mettl23 interacts with Tet3 via gonad-specific expression (GSE) protein

d

GSE and Mettl23 are indispensable for active DNA demethylation

Hatanaka et al., 2017, Cell Reports 20, 2756–2765 September 19, 2017 ª 2017 The Author(s). http://dx.doi.org/10.1016/j.celrep.2017.08.088

Cell Reports

Report Histone H3 Methylated at Arginine 17 Is Essential for Reprogramming the Paternal Genome in Zygotes Yuki Hatanaka,1,* Takeshi Tsusaka,2,3 Natsumi Shimizu,4 Kohtaro Morita,4 Takehiro Suzuki,5 Shinichi Machida,6 Manabu Satoh,4 Arata Honda,1,7 Michiko Hirose,1 Satoshi Kamimura,1,8 Narumi Ogonuki,1 Toshinobu Nakamura,9 Kimiko Inoue,1,8 Yoshihiko Hosoi,4 Naoshi Dohmae,5 Toru Nakano,10 Hitoshi Kurumizaka,6 Kazuya Matsumoto,4,* Yoichi Shinkai,2 and Atsuo Ogura1,8,11,12,* 1RIKEN

BioResource Center, Ibaraki 305-0074, Japan Memory Laboratory, RIKEN Wako, Saitama 351-0198, Japan 3Department of Diabetes, Endocrinology and Nutrition, Graduate School of Medicine, Kyoto University, Kyoto 606-8507, Japan 4Division of Biological Science, Graduate School of Biology-Oriented Science and Technology, Kindai University, Wakayama 649-6493, Japan 5RIKEN Center for Sustainable Resource Science, Wako, Saitama 351-0198, Japan 6Laboratory of Structural Biology, Graduate School of Advanced Science and Engineering, Waseda University, 2-2 Wakamatsu-cho, Shinjuku-ku, Tokyo 162-8480, Japan 7Organization for Promotion of Tenure Track, University of Miyazaki, Miyazaki 889-1692, Japan 8Graduate School of Life and Environmental Sciences, University of Tsukuba, Ibaraki 305-8572, Japan 9Department of Bio-Science, Nagahama Institute of Bio-Science and Technology, Shiga 526-0829, Japan 10Department of Pathology, Graduate School of Medicine, Osaka University, Osaka 565-0871, Japan 11Center for Disease Biology and Integrative Medicine, Faculty of Medicine, University of Tokyo, Tokyo 113-0033, Japan 12Lead Contact *Correspondence: [email protected] (Y.H.), [email protected] (K.M.), [email protected] (A.O.) http://dx.doi.org/10.1016/j.celrep.2017.08.088 2Cellular

SUMMARY

At fertilization, the paternal genome undergoes extensive reprogramming through protamine-histone exchange and active DNA demethylation, but only a few maternal factors have been defined in these processes. We identified maternal Mettl23 as a protein arginine methyltransferase (PRMT), which most likely catalyzes the asymmetric dimethylation of histone H3R17 (H3R17me2a), as indicated by in vitro assays and treatment with TBBD, an H3R17 PRMT inhibitor. Maternal histone H3.3, which is essential for paternal nucleosomal assembly, is unable to be incorporated into the male pronucleus when it lacks R17me2a. Mettl23 interacts with Tet3, a 5mC-oxidizing enzyme responsible for active DNA demethylation, by binding to another maternal factor, GSE (gonad-specific expression). Depletion of Mettl23 from oocytes resulted in impaired accumulation of GSE, Tet3, and 5hmC in the male pronucleus, suggesting that Mettl23 may recruit GSE-Tet3 to chromatin. Our findings establish H3R17me2a and its catalyzing enzyme Mettl23 as key regulators of paternal genome reprogramming. INTRODUCTION Mammalian development is regulated by the precise orchestration of a series of epigenetic events, which are heritable

changes in gene expression that do not involve changes to the underlying DNA sequence. Reprogramming of the paternal genome in mammals is essential for establishing the totipotent state of the zygotic genome at fertilization (Hackett et al., 2012; Sasaki and Matsui, 2008). This reprogramming initiates with the removal of protamine from the decondensing nucleus of the fertilizing sperm. The paternal genome then forms the male pronucleus (PN) with nucleosomes containing maternally derived core histones (Lin et al., 2014; Inoue and Zhang, 2014). At this stage, the methylation level of paternal DNA is higher than that of maternal DNA, but later, active DNA demethylation proceeds predominantly in the male PN, leading to an asymmetric epigenetic status that discriminates the parental genomes during the early cleavage stages (Inoue and Zhang, 2011). To date, several maternal factors responsible for these dynamic reprogramming processes have been identified, especially in mice. Maternal core histones are incorporated into the nucleosomal structure in the male PN. The major histone variants involved in this process are H2A.X and H3.3 (Nashun et al., 2010; Lin et al., 2014). As H2A.X is also incorporated into the maternal genome by replacing H2A.Z, it is assumed that this H2A variant exchange is responsible for establishment of the new epigenetic memory at fertilization (Nashun et al., 2010). By contrast, H3.3 is incorporated exclusively to the male PN, because the maternal genome already possesses H3.3 before fertilization. Therefore, depletion of the maternal HIRA, the histone H3.3 chaperone, resulted in the impairment of the male, but not female, PN formation in zygotes (Inoue and Zhang, 2014; Nashun et al., 2015). This finding indicates that HIRA-mediated H3.3 deposition plays a key role in paternal genomic reprogramming.

2756 Cell Reports 20, 2756–2765, September 19, 2017 ª 2017 The Author(s). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

A

C

B

D

E

F

G

Figure 1. GSE Protein Interacts with Tet3 and Mettl23 in Mouse Zygotes (A) Interaction of GSE with Tet3 analyzed using a yeast two-hybrid assay. The full-length and truncated Tet3 (471 aas; region C) interacted with GSE, as indicated by the blue colonies. (B) Interaction of GSE with Tet3 in PN3 zygotes, as demonstrated by coimmunoprecipitation using specific antibodies. A representative result from three replicate experiments each using 100–150 zygotes.IgG, immunoglobulin G. (C) The effect of deletion of the domain C in FLAG-Tet3 on Tet3 accumulation in male pronucleus of FLAG-Tet3 mRNA-injected PN3 zygotes. Left: DNA was stained with DAPI. PB, the second polar body. Scale bar, 50 mm. Right: the relative FLAG signal intensities in the male and female PNs. The level of DAPI was set as 1.0. Red bars indicate the mean ratios. *p < 0.01 (versus FLAG-Tet3-injected); ns, not significant. Ten zygotes were analyzed for each group.

(legend continued on next page)

Cell Reports 20, 2756–2765, September 19, 2017 2757

Previously, we have found that the gonadal-specific expression (GSE) protein, a maternal factor that is specifically expressed in germ cells, might be involved in active DNA demethylation during zygotic development (Hatanaka et al., 2013; Mizuno et al., 2006). Therefore, in this study, we examined whether GSE would interact with Tet3, an oxidative enzyme responsible for active DNA demethylation by conversion of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) (Gu et al., 2011), and whether other maternal factors might exist to recruit Tet3 to its target sites. We identified methyltransferase-like 23 (Mettl23) as a Tet3-related factor. Mettl23 is an arginine methyltransferase that catalyzes asymmetric dimethylation of histone H3 at arginine 17 (H3R17me2a). Importantly, we found that R17me2a was responsible for the incorporation of maternal H3.3 into the male PN. RESULTS GSE Protein Interacts with Tet3 in Zygotes As we have already identified GSE as a putative maternal factor involved in active DNA demethylation, we examined whether GSE interacted with Tet3 protein by a yeast two-hybrid system. As shown in Figure 1A, GSE interacted not only with the fulllength Tet3 but also with a part of its core catalytic region that has methylcytosine dioxygenase activity from amino acids (aas) 1012–1482 (region C in Figure 1A). This region is conserved in the Tet family of genes (Kohli and Zhang, 2013; Pastor et al., 2013; Tan and Shi, 2012). We next performed coimmunoprecipitation experiments using anti-GSE and anti-Tet3 antibodies and confirmed that GSE interacted with Tet3 in zygotes at the pronuclear 3 (PN3) stage (Figure 1B). Furthermore, we observed that the deletion of region C in Tet3 resulted in the impairment of Tet3 recruitment into the male PN, as indicated by the decrease of the signal intensity in Figure 1C. These results suggest that the interaction of GSE and Tet3 is responsible for the recruitment of Tet3 into the male PN in zygotes. GSE Interacts with Mettl23 in Zygotes GSE could be associated with core histones, as indicated by a yeast two-hybrid assay (Figure S1A), but it has no functional domains that could modify histone proteins (Mizuno et al., 2006). Therefore, we hypothesized that there might be another factor intercalating between GSE and core histones to recruit Tet3GSE to target sites. To test this, we first searched for interacting proteins in a yeast two-hybrid screening of a mouse ovary cDNA library using the GSE protein as bait and identified Mettl23, histone H3.3, and Rims3 as GSE-interacting proteins (Figure 1D). Rims3 is known to be involved in exocytosis and is expressed primarily in the nervous system. Therefore, we focused on Mettl23, a putative methyltransferase, as a candidate maternal

factor interacting with the Tet3-GSE complex. We have previously shown that transcription of the GSE gene involves splicing variants: a long form of the protein (27.6 kDa) involving all exons, and a short form (23.1 kDa) with exons 4–5 not transcribed. As the long form was expressed from the zygote-to-blastocyst stages (Hatanaka et al., 2013), we tested whether Mettl23 would interact with the long form of GSE using a yeast two-hybrid system. Mettl23 interacted with the product of exons 4–5 as well as the full-length form, indicating that it was associated with the long form in zygotes (Figure 1E). In agreement with this, Mettl23 was expressed from oocytes through the preimplantation stages, in addition to several other tissues (Figures S1B and S1C). We confirmed interaction of Mettl23 with GSE in PN3 zygotes by a coimmunoprecipitation assay using pooled zygotes (Figure 1F). Their direct binding was also confirmed by an assay using His-tagged Mettl23 and His-tagged GSE proteins purified from transfected HEK293 cells (Figure 1G). These results suggested that Mettl23 is a binding partner of GSE in zygotes. Mettl23 Is Responsible for the Accumulation of H3R17me2a into PNs during Zygotic Development Mettl23 belongs to a methyltransferase-like (Mettl) family that has distant homology with protein arginine methyltransferases (PRMTs), but its enzymatic activity has not yet been defined (Cloutier et al., 2013). We found that Mettl23 has homology with the conserved domains of PRMT activity in PRMT1–8 (Figure 2A). Next, we confirmed the PRMT activity of Mettl23 in metaphase II (MII) oocytes using Mettl23-knockdown (Mettl23-KD) oocytes, which had been injected with short interfering (si)RNA specific for Mettl23. Mettl23-KD specifically decreased the level of H3R17me2a but not the level of other arginine methylation residues in oocytes (Figure 2B; see also Figure S2A). Immunoblot analysis indicated the localization of H3R17me2a in germinal vesicle (GV)-stage and MII oocytes and preimplantation embryos until the blastocyst stage (Figure S2B). In PN3-stage zygotes, Mettl23 was localized in both PNs and cytoplasm in association with nuclear H3R17me2a (Figure S2C). We also noted that Mettl23, as well as H3R17me2a, were distributed symmetrically in both PNs (Figure S2C), as indicated by the M/F fluorescence ratio of approximately 1.0 (Figure S2D). This behavior was in contrast to that of GSE, which showed a predominant distribution in the male PN (Figure 3F). To demonstrate the PRMT activity of Mettl23, we performed in vitro assays in the presence of recombinant histone H3.1 by using 3xFLAG-His6-Mettl23 purified from transfected HEK293T cells. As shown in Figure 2C, the Mettl23 protein methylated histone H3.1; this H3.1 methylation was reduced by treatment with TBBD (ellagic acid), which specifically inhibits methylation at H3R17 by binding to KAPR17K aa motifs (Selvi et al., 2010). We

(D) Identification of GSE-interacting protein with a cDNA library of the mouse ovary using a yeast two-hybrid assay. Letters A to F indicate the positions in the dish. (E) Identification of a Mettl23-binding region of GSE using a yeast two-hybrid assay. The region derived from exons 4 and 5 was responsible for binding with Mettl23. (F) Interaction of GSE with Mettl23 in PN3 zygotes, as demonstrated by coimmunoprecipitation using specific antibodies. A representative result from three replicate experiments each using 100–150 zygotes. (G) Direct interaction of GSE and Mettl23 confirmed by immunoprecipitation followed by immunoblotting using His-GSE and His-Mettl23 purified from transfected HEK293 cells. A representative result from three replicate experiments. sup, supernatant.

2758 Cell Reports 20, 2756–2765, September 19, 2017

B siControl

PRMT1 PRMT2 PRMT3 PRMT4 (CARM1) PRMT5 PRMT6 PRMT7 PRMT8 Mettl23

Mettl23 H3R8 me2a H3R17 me2a H3R26 me2

IB

III

II


GSTG9a

C

Mettl23 GSTG9a

H3.1

IB

CBB staining

H3R17me2a

WT WT RA RK WT RA RK

H3.1

50 37

IgG

25 20

IgG

15 kDa

H3.1

75 50 37

GST- G9a IgG g

25 20

IgG

15 kDa

H3.1

3xFLAG-His-Mettl23

Mettl23

F

GST- G9a IgG

50 37

3xFLAG-His-Mettl23

25 20

IgG

15 kDa

H3.1

0.6 0.4 0.2 0 WT RA RK

100 100 80 80 60 60 40 40 20 20 00 100 100 80 80 60 60 40 40 20 20 00

Relative abundance

Mettl23

Relative abundance

0.8

100 100 80 80 60 60 40 40 20 20 00

Relative Abundance

G

1

Relative abundance

Methylated H3.1

Relative abundance

Relative signal level

0.2 0

3xFLAG-His-Mettl23 Histone H3.1 TBBD 14C-SAM

3xFLAG-His-Mettl23 Histone H3.1 TBBD 14C-SAM

Autoradiography

1.2

0.4

Mettl23

CBB staining

CBB staining

0.6

Autoradiography

D

Mettl23

1 0.8

Mettl23 R8 R17 R26 me2a me2a me2


3xFLAG-His-Mettl23 Histone H3.1 14C-SAM

1.2

Actin

THW LOOP

E

1.4

siControl Mettl23-KD siControl Mettl23-KD siControl Mettl23-KD siControl Mettl23-KD

Post I

MII oocytes

Relative signal level


Mettl23 -KD

A

100 100 80 80 60 60 40 40 20 20 00

b2

y3 y2 y1

++

y1 169.0982 y3 200.6381 147.1136 148.0613 z=?

104.0715 z=1

170.0822 z=1

235.1300 z=1

187.1089 z=1

86.0971 z=?

272.1728 z=1

129.1027 147.1132 200.6377 z=? 175.1195 z=2 z=1

b2 169.0976 y1161.1288

115.0871 z=1

z=1 130.0868 147.1131 z=1

86.0972 z=?

y1

215.1018 z=1

199.1318 z=2

175.1447 z=?

100 100

150 150

228.1462 254.1620 z=? z=1

347.2050 z=?

b3

325.1990 347.2045 z=?

303.2137 z=1

y3

+Mettl23 APRK

400.2675 400.2675 z=1 387.2360 z=?

414.2440 z=?

300.2024 z=1

237.1352 z=1

317.2298 z=1

458.2719 z=?

y3

343.2093 z=1 361.2194 z=?

331.2455 331.2455

250.1719 272.1725 z=2 z=1 230.1557 296.2080 314.2188 z=1 z=1 z=? 250 250

m/z

300 300

343.2095 z=1 350 350

+Mettl23 APRK + methyl

414.2839 414.2839 z=1 397.2571 z=1

456.2585 z=?

y2

214.6531 z=2

b1 b2 b3

A P R K b1 b2 b3 y3 y2

y2 317.2298

254.1618 z=1

214.6531

200 200

458.2718 z=?

A P R K y3 y2 y1

303.2137

y3++

188.1218 z=?

Control APRK

y3

400.2685 z=1 383.2411 400.2685 426.1908 z=1 z=?

272.1723 z=1

y3++

b2

147.1132

272.1724 z=1

207.6453 207.6453 z=2

169.0975

130.0868 z=? 109.1018 147.1132 z=? z=1

325.1988

y2

b2 200.6377 y1 169.0975 211.1194 z=?

98.0970 z=1

y2

303.2150

303.2150 z=1

y3++

112.0875 z=1

b3

y3

428.2993

A P R(me) K b1 b2

b3

y3 y2 +Mettl23 APRK + dimethyl

411.2726 428.2993 z=1 385.2570 456.2569 z=1 z=1 z=? 400 400

y1

450 450

y1

A P R(me2) K b1 b2

b3

Figure 2. Arginine Methylation Activity of Mettl23 Protein (A) Amino acid sequence alignment of the methyltransferase I region of mouse Mettl23, together with PRMT1–8. The same amino acids are shaded in black, and similar amino acids are shaded in gray. Conserved signature methyltransferase motifs are boxed in red. (B) The effect of Mettl23-KD on the methylation of different arginine residues of histone H3 in oocytes. The graph on the right indicates relative band intensities. The level of actin was set as 1.0. Results from three replicate experiments (about 30 oocytes each). *p < 0.001 (versus siControl). Error bars indicate SEM.

(legend continued on next page)


also found a reduction in the level of H3R17me2a in the presence of TBBD using anti-H3R17me2a antibody by immunoblot analysis (Figure 2D). We identified the localization of CARM1 (also known as PRMT4), another H3R17me2a PRMT. It was localized in the cytoplasm, but not in the PNs, of zygotes (Figure S2C) and in the nucleus and cytoplasm of GV oocytes (Figure S2E). Then, to clarify whether the methylation activity of Mettl23 was CARM1 dependent, we performed co-immunoprecipitation experiments using anti-FLAG and anti-CARM1 antibodies in the mixture of purified 3xFLAG-His6-Mettl23. We found that Mettl23 did not interact with CARM1 (Figure S2F), indicating that the H3.1R17me2a was methylated by Mettl23 but not by CARM1. To confirm that H3R17 was the target of Mettl23, we investigated the effect of unmethylatable R17 mutants, H3.1R17A and H3.1R17K, on the Mettl23 activity in vitro. These mutations resulted in the reduction of the levels of H3.1 methylation, as indicated by the alterations in signal intensity (H3.1R17A, 0.75; and H3.1R17K, 0.60 [WT = 1.0]) (Figure 2E). This result indicated that Mettl23 methylated H3R17, although it might also methylate other residues of H3 in vitro. In MII oocytes, however, Mettl23 did not methylate H3R8 or H3R26, at least at the immunoblot level (Figure 2B). To identify the accumulation of R17 methylation by Mettl23, we next analyzed the resulting H3.115–18 peptide containing R17, APRK peptide, by mass spectrometry. The mono- and dimethylated APRK peptides could be detected in the presence of 3xFLAG-His6-Mettl23 with recombinant H3.1 (Figure 2F). We also found a tandem mass spectroscopy spectrum of mono- and dimethylation on R17 (Figure 2G). These results collectively suggest that Mettl23 catalyzes dimethylation of H3R17. GSE and Mettl23 Are Indispensable for 5mC Oxidation by Recruiting Tet3 To determine the function of GSE and Mettl23 in 5mC oxidation during zygote development, we generated GSE gene knockout (GSE-KO) and Mettl23 gene knockout (Mettl23-KO) mice by the CRISPR/Cas9 system (Figure S3A). Mice carrying homozygous mutations were born alive in both KO lines. We confirmed the successful deletions of GSE and Mettl23 in the KO mice by western blot analysis using their testes (Figure S3B). Homozygous mutant mice from the GSE-KO line were indistinguishable from their heterozygous counterparts in terms of gross appear-

ance, growth rate (data not shown), and reproductive performance, such as litter sizes (Figure S3C). By contrast, the Mettl23-KO line gave birth to significantly smaller litter sizes following homozygous mating, compared with heterozygous mating (Figure S3C). Moreover, about 33% (11/33) of the homozygous Mettl23 newborn mice died before weaning from unknown causes, and only a few survived to adulthood. Because Mettl23 is ubiquitously expressed in many tissues (Figure S1C), and the PRMTs and Mettl families might catalyze nonhistone proteins (Blanc and Richard, 2017; Shimazu et al., 2014), these Mettl23-KO phenotypes might be attributable to the loss of systemic, multifunctional catalytic activity of this enzyme. In zygotes derived from GSE-KO or Mettl23-KO oocytes, the 5mC level remained high and the 5hmC level remained low in the male PN (Figures 3A, S4A, and S4B). The injection of Mettl23 mRNA into Mettl23-KO oocytes restored the conversion of 5mC to 5hmC in maternal Mettl23-KO zygotes (Figures 3A, S4A, and S4B). These findings indicated that GSE-KO or Mettl23-KO in oocytes resulted in the failure of the 5mC-to-5hmC conversion. This impaired 5mC oxidation was confirmed by methylated DNA immunoprecipitation (MeDIP) and hydroxymethylated DNA immunoprecipitation (hMeDIP) experiments directed to the Line1 retrotransposon regions, which are known to be actively demethylated in zygotes (Gu et al., 2011; Wossidlo et al., 2010). The 5mC level at Line1 was higher in maternal GSE-KO or Mettl23-KD zygotes than in siRNA-treated control (siControl) zygotes, while the 5hmC level was lower in these knockdown zygotes (Figure 3B). Furthermore, our chromatin immunoprecipitation (ChIP) analysis using about 1,400 intact zygotes revealed that H3R17me2a was enriched at the Line1 regions (Figure 3C). All these findings suggest that maternal GSE and Mettl23, in association with H3R17me2a, can play essential roles in the oxidation of 5mC to 5hmC in zygotes, although this is not a required event for viability. We next observed Tet3 localization in maternal GSE-KO and Mettl23-KO zygotes. In control zygotes, Tet3 was localized in both male and female PNs, with a stronger intensity in the male PN (Figure 3D), as reported previously (Guo et al., 2014; Shen et al., 2014). The Tet3 signal was significantly reduced in GSE- or Mettl23-KO zygotes compared with the wild-type (WT) (Figures 3D and S4C). The forced expression of Mettl23 in its KO zygotes restored the accumulation of Tet3 into the male PN (Figures 3D and S4C). These results indicated that Mettl23

(C) Mettl23 methylates histone H3.1 in vitro. 3xFLAG-His6-Mettl23 was tandem affinity purified from transfected HEK293T cells. GST-fused catalytic domain of G9a, which is known as H3K9 methyltransferase, was used as positive control for the methylation assay. The H3R17 methylation inhibitor TBBD (final, 100 mM) impaired methylation by Mettl23, but not by G9a. Top: autoradiography. Bottom: Coomassie brilliant blue (CBB) stain. CBB staining was performed as an equal loading control. (D) Asymmetric dimethylation at arginine 17 of histone H3 (H3R17me2a) induced by Mettl23 in vitro. The H3R17 methylation inhibitor TBBD (final, 100 mM) impaired the methylation by Mettl23. (E) Effect of unmethylatable mutations of H3.1R17 on the accumulation of H3.1 methylation in the presence of 14C-SAM in vitro. Top: autoradiography indicating the methylation signals of H3.1 induced by Mettl23. CBB staining was performed as an equal loading control. The mutations of H3.1R17 did not change the methylation levels induced by the GST-fused catalytic domain of G9a. Bottom: the relative methylated H3.1 signal intensities in the mutations by Mettl23. The level of CBB stain was set as 1.0. (F) Selected ion chromatograms of y3 ions (m/z = 400.267, 414.283, and 428.299; see Figure 2G) that from the parent ions of unmethylated (me0), monomethylated (me1) and dimethylated (me2) forms of APRK peptides obtained from digestion of recombinant H3.1 with or without Mettl23. Predicted methylation sites are indicated in red. RT indicates retention time of the displayed ion chromatograms. m/z, mass-to-charge ratio. (G) Tandem mass spectrometry (MS/MS) spectra of the three methylation states of APRK peptides with Mettl23. The indicated b (blue) and y (pink) ions indicate mono- and dimethylation on the arginine.


A

B

C

D

E

F

1.2 1 0.8 0.6 0.4 0.2 0

(legend on next page)


and GSE are involved in Tet3 localization in the PNs. Next, we examined whether Mettl23-KO might affect localization of the GSE protein in the PNs, because it was possible that Mettl23 might function as a recruiter of GSE-Tet3 to the target sites. As expected, only faint signals for GSE were found in the PNs of the Mettl23-KO zygotes, and the rescue experiment resulted in recruitment of GSE into the male PN (Figure 3E). Taken together with the results from the biochemical analyses discussed earlier, it is likely that maternal Mettl23 is responsible for 5mC oxidation by recruiting the GSE-Tet3 complex into the chromatin during zygote development. It also suggests that CARM1 is not able to compensate for the function of Mettl23 in active DNA demethylation. PGC7 Is Responsible for the Asymmetric Localization of GSE-Tet3 in PNs It is known that the asymmetrical pattern of active DNA demethylation between the paternal and maternal genomes is achieved by the presence of the PGC7 protein, which protects the female genome from Tet3 (Nakamura et al., 2007). Therefore, we examined whether the localization patterns of GSE and Mettl23 in zygotes might be altered by deletion of the maternal PGC7. In maternal PGC7 KO zygotes, the fluorescence intensity for GSE in the female PN increased approximately to the level of that in the male PN, the male/female (M/F) fluorescence ratio becoming 1 (Figure 3F). By contrast, the fluorescence levels of Mettl23 were unchanged and very similar between the male and female PNs in the same zygotes (Figure 3F). These findings indicate that PGC7 protects the maternal genome from GSE-Tet3, while the attachment of Mettl23 to both parental genomes is not influenced by the presence of PGC7. Methylation at R17 Is Essential for H3.3 Incorporation into the Male PN As both CARM1 and Mettl23 are present in the cytoplasm of unfertilized oocytes, it is possible that maternal histone H3.3 has been modified with R17me2a before fertilization. To assess this possibility, we injected oocytes with mRNA for EGFP -tagged H3.3 with amino acid substitutions. Normal H3.3, as well as R8- and R26-mutated H3.3, was incorporated into the male PN, whereas R17-mutated H3.3 failed to be localized into the male PN (Figure 4). This result indicated that methyl-

ation at R17 may be a prerequisite for H3.3 incorporation into the paternal nucleosomes. DISCUSSION Reprogramming of the paternal genome at fertilization is one of the most dynamic epigenetic changes during the life cycle in mammals. Here, we identified H3R17me2a as a histone modification that may regulate the reprogramming of the paternal genome through protamine-histone exchange and active DNA demethylation. A large part—if not all—of the R17me2a modification on H3.3 in the oocytes/zygotes may be attributable to the activity of Mettl23, here reported to have PRMT. A key question that remains unanswered is why H3R17me2a is the essential histone modification for reprogramming of the paternal genome. It is known that H3R17me2a is frequently associated with transcriptionally active chromatin (Chen et al., 1999; Di Lorenzo and Bedford, 2011; Wu et al., 2009). In addition, conversion of 5mC to 5hmC occurs on transcriptionally active euchromatin in mouse embryonic stem cells and human somatic cells (Ficz et al., 2011; Kubiura et al., 2012). Therefore, we postulate that Mettl23-catalyzed H3R17me2a might induce an open chromatin structure so that the subsequent molecular events can proceed normally until the full reprogramming of the paternal genome. Generally, information on reader proteins that recognize methylated arginine is limited, compared with that on lysine methylation (Musselman et al., 2012). A recent study suggested that the TDRD3 Tudor domain, which recognizes H3R17me2a and H4R3me2a, was tightly complexed with DNA topoisomerase IIIb (TOP3B) to induce relaxation of negative supercoiled DNA in cultured cells (Yang et al., 2014). It would be interesting to see whether a similar mechanism is active at the H3R17me2a sites in zygotes. When sperm enters an egg, it must undergo biochemical remodeling that is reliant on maternal factors. Decondensation of the sperm head takes place by maternal nucleoplasmin 2, releasing protamine into the oocyte cytoplasm, in preparation for PN formation. It is known that de novo nucleosome assembly is dependent on H3.3 and its chaperone Hira, which is essential for PN formation (Inoue and Zhang, 2014; Nashun et al., 2015). In this study, we found that H3.3 devoid of R17me2a was unable to participate in PN formation, suggesting that the R17me2a modification was a prerequisite for nucleosome assembly by

Figure 3. GSE and Mettl23 Are Indispensable for Active DNA Demethylation by Recruiting Tet3 to the PNs (A) Top: representative immunofluorescence images of 5mC and 5hmC in different groups of PN3 zygotes. PB, the second polar body. Scale bar, 50 mm. Bottom: the fluorescence M/F ratios for 5mC and 5hmC. Each dot represents a single zygote (11 zygotes for each group). Red bars indicate the mean ratios. *p < 0.001 (versus WT). (B) qPCR analysis of Line1 regions on immunoprecipitation using an anti-5mC antibody (for MeDIP) and anti-5hmC antibody (for hMeDIP). Results are from 30 oocytes (PN0) or 30 PN3–4 zygotes. *p < 0.05. Error bars indicate SEM. Results from three biological replicate experiments. The values for PN0 were calculated by adding the values of spermatozoa to those of MII oocytes. (C) ChIP analysis of H3R17me2a on Line1 regions. Results are from three biological replicate experiments. *p < 0.05. Error bars indicate SEM. (D) Top: representative immunofluorescence images of Tet3 in maternal GSE-KO or Mettl23-KO zygotes. H3 is indicated in green as a control for chromatinbound factors. DNA was stained with DAPI. PB, the second polar body. Scale bar, 50 mm. Bottom: the relative Tet3 signal intensities in the male and female PNs. The numbers of zygotes analyzed were 10 to 15 for each group. The level of H3 was set as 1.0. Red bars indicate the mean ratios. *p < 0.001 (versus WT). (E) Top: representative immunofluorescence images of GSE in Mettl23-KO zygotes. Bottom: the relative GSE signal intensities in the male and female PNs. For other information see the legend for Figure 3D. (F) Immunofluorescence images of GSE and Mettl23 in maternal PGC7-KO zygotes. The M/F ratio for the GSE level was significantly reduced (*p < 0.005), but the Mettl23 level was unchanged. Fifteen zygotes were analyzed for each group. For other information, see the legend for (D). ns, not significant.


Figure 4. H3R17me2a Is Essential for Maternal H3.3 Incorporation into the Male PN Effect of an unmethylatable mutation of H3.3R17 on deposition into paternal nucleus by injection of EGFP-fused histone mutants. Left: immunofluorescence images of EGFP in mRNA-injected zygotes. Right: the EGFP signal intensity in male PN. Each dot represents a single zygote. The numbers of zygotes analyzed were 11 to 20 for each group. Red bars indicate the mean ratios. Asterisks indicate significant differences compared with control embryos (p < 0.05). PB, the second polar body. Scale bar, 50 mm.

H3.3. Our experimental results indicate that the enzymatic activity of Mettl23 as R17me2a PRMT is exerted before fertilization (Figure 2B). This may result in accumulation of H3.3R17me2a in the MII ooplasm, which may induce paternal nucleosomal assembly upon fertilization. Because Mettl23 knockout oocytes normally form PNs, it is probable that the catalytic role of Mettl23 might be compensated for—at least in part—by CARM1, another R17me2a PRMT, during paternal nucleosome formation. The abundant localization of CARM1 in the immature oocytes may support this assumption (Torres-Padilla et al., 2007). Generation of oocytes lacking both Mettl23 and CARM1 would help solve these questions. Mettl23 has distant homology with other PRMTs (Figure 2A), and two point mutations of the putative catalytic sites (D113V and D163V) still retained the enzymatic activity (unpublished data). This may suggest that the active site or reaction mechanism of Mettl23 is different from that of other PRMTs. If another unidentified H3R17 PRMT exists in oocytes, Mettl23 might interact with it as a partner protein. In contrast to the obvious involvement of Mettl23 in H3R17me2a formation before PN formation, its involvement in active DNA demethylation remains to be studied. After PN formation, Mettl23, but not CARM1, was localized in the PNs and bound to GSE, a Tet3-interacting protein. Therefore, Mettl23 may have some non-PRMT role in Tet3-mediated active DNA demethylation, which CARM1 does not possess. Depletion of Mettl23 resulted in impaired pronuclear localization of GSE, Tet3, and 5hmC, indicating that Mettl23 functions as a recruiter of GSE and Tet3 to target sites. A structural analysis of Mettl23 could elucidate the mechanisms underlying its multifunction. Indeed, Mettl23 has low homology to all the PRMTs in their nonconserved domains (Figure 2A). CARM1 is known to be a multifunctional protein and is engaged in non-PRMT roles, such as mRNA processing (Cheng et al., 2007) and regulating protein stability (Feng et al., 2006). Thus, Mettl23 and CARM1 have the

PRMT activity in oocytes and zygotes by the common conserved domains, while Mettl23 might have another function as a recruiter of GSE-Tet3. One of the important features of active DNA demethylation in zygotes is its predominant contribution to reprogramming of the paternal genome (Mayer et al., 2000), although this event is not essential for subsequent embryonic development (Inoue et al., 2015). This parental-origin-dependent DNA demethylation is ensured by the preferential localization of maternal PGC7 to the female PN (Nakamura et al., 2007). The female PN is enriched with H3K9me2, which attracts the PGC7 protein to the maternal genome. It was shown experimentally that PGC7 protected 5mC from Tet3-mediated oxidation (Nakamura et al., 2012). Detailed biochemical analyses using isolated male and female PNs revealed that conversion of 5mC to 5hmC by Tet3 also occurs in the female PN, albeit on a smaller scale (Guo et al., 2014; Shen et al., 2014). These studies also identified target regions of active DNA demethylation, some of which were specific for either the maternal or the paternal genome. Our study provides another layer of information to explain this process. We found that GSE was preferentially localized to the male PN, while Mettl23H3R17me2a was evenly distributed in both PNs. Interestingly, maternal PGC7-KO resulted in elevation of the GSE level as well as the Tet3 level in the female PN. This finding supports the idea that while Mettl23-H3R17me2a can be localized to the maternal (oocyte) genome, H3K9me2-PGC7 inhibits the attachment of GSE-Tet3 to Mettl23, thus preventing active DNA demethylation in the female PN. Thus, active DNA demethylation in zygotes may be regulated by two different histone modifications, H3R17me2a and H3K9me2, which have opposite effects on Tet3 recruitment. EXPERIMENTAL PROCEDURES Detailed methods are available in the Supplemental Information.


Animals All animal experiments described here were approved by the Animal Experimentation Committees at the RIKEN Tsukuba Institute and Kinki University and were performed in accordance with the committees’ guiding principles. Animals were housed under controlled lighting conditions (daily light, 0700–2100 hr) and were maintained under specific pathogen-free conditions.

AUTHOR CONTRIBUTIONS

Immunocytochemistry and Microscopy The classification of PN stages was performed according to previous studies (Santos et al., 2002; Wossidlo et al., 2010), where the pronuclear morphology and hours postinsemination were taken into consideration. Oocytes or zygotes were treated with 0.2% Triton X-100 (Nacalai Tesque) in PBS for 30 s, washed with PBS three times, and fixed in 4% PFA (Nacalai Tesque) (Nakamura et al., 2012). In some experiments (Figure 4A), oocytes or embryos were fixed in 4% PFA first, followed by treatment with PBS containing 0.1%–0.2% Triton X-100 at room temperature (RT) for 1 hr. For 5mC, 5hmC, and DNA, the specimens were incubated in 4 N HCl at RT for 30 min and then incubated in 0.1 M EDTA at RT for 30 min. They were then incubated with the primary antibodies (Table S1) in PBS containing 30 mg/mL BSA at 4 C overnight. After washing, they were reacted with the secondary antibodies as appropriate (Table S1) at RT for 1 hr. Specimens were mounted on glass slides in Vectashield mounting medium (Vector Laboratories) containing 2–5 mg/mL DAPI (Invitrogen; D1306). Finally, the slides were imaged using a CellVoyager CV1000 confocal scanner system (Yokogawa Electric). ImageJ software (NIH, Bethesda, MD, USA; http://rsbweb.nih.gov/ij/) was used to quantify DAPI staining and antibody signals in the central region of each PN. At least three independent replicates were performed for each experiment.

ACKNOWLEDGMENTS

ChIP These procedures were performed essentially as described (Hatanaka et al., 2015). In brief, about 1,400 mouse zygotes were harvested for cross-linking. Fixation and isolation of the nuclei of the zygotes were performed using truChIP Chromatin Shearing Reagent Kits (Covaris). Chromatin shearing was performed using a Covaris S220 instrument. Immunoprecipitation was performed as described earlier. For binding to magnetic beads, each antibody was incubated with Dynabeads Protein G (Invitrogen) in a ChIP dilution buffer containing 0.01% SDS, 1.1% w/v Triton X-100, 1.2 mM EDTA, and 167 mM NaCl at 4 C for 1 hr with rotation. After washing, the beads were incubated with the chromatin shearing solution at 4 C for 6 hr with rotation. After immunoprecipitation, the beads were washed using a lowsalt wash buffer containing 20 mM Tris-HCl, 0.1% SDS, 1% w/v Triton X-100, 2 mM EDTA, and 150 mM NaCl and a high-salt wash buffer containing 20 mM Tris-HCl, 0.1% SDS, 1% w/v Triton X-100, 2 mM EDTA, and 500 mM NaCl. For elution of the DNA from the beads, they were incubated with ChIP direct elution buffer containing 10 mM Tris-HCl, 1% SDS, 5 mM EDTA, and 300 mM NaCl at 4 C for 15 min with rotation. The solution was incubated at 65 C overnight and then treated with proteinase K and RNase A and subjected to DNA purification. Prepared DNA samples were amplified and analyzed by qRT-PCR. The primers used are described in Table S2. Amplifications were run in a 7900HT Sequence Detector System (Applied Biosystems). Statistical Analysis For statistical analysis, we performed one-way ANOVA using StatView v.5.0 (SAS Institute) or Microsoft Excel. A post hoc procedure using Fisher Protected Least Significant Difference (PLSD) was adopted for multiple comparisons between the groups where appropriate. The p values < 0.05 were considered significant. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, four figures, and two tables and can be found with this article online at http://dx.doi.org/10.1016/j.celrep.2017.08.088.


Y. Hatanaka, K. Matsumoto, and A.O. conceived the project and designed the research. Y. Hatanaka, T.T., N.S., K. Morita, T.S., S.M., M.S., A.H., M.H., S.K., N.O., and K.I. performed the experiments. Y. Hatanaka, T. Nakamura, Y. Hosoi, N.D., T. Nakano, H.K., K. Matsumoto, Y.S., and A.O. analyzed the data and discussed the research. Y. Hatanaka and A.O. wrote the paper.

The authors wish to thank Dr. Guoliang Xu for the anti-Tet3 antibody and Dr. Tadahiro Shimazu for the suggestion of an in vitro assay of Mettl23 activity. This study was supported by JSPS KAKENHI grant numbers JP25112009 (to A.O.), JP23220011 (to A.O.), JP25292189 (K.M.), JP25116002 (to H.K.), and JP15K18545 (to Y. Hatanaka) and by the RIKEN Epigenetics Program (Strategic Programs for R&D) (to A.O. and Y.S.), the RIKEN Incentive Research Projects 2015 (to Y. Hatanaka) and 2016 (to Y. Hatanaka), and the RIKEN Junior Research Associate Program (to T.T.). Received: November 22, 2016 Revised: May 31, 2017 Accepted: August 25, 2017 Published: September 19, 2017 REFERENCES Blanc, R.S., and Richard, S. (2017). Arginine methylation: The coming of age. Mol. Cell 65, 8–24. Chen, D., Ma, H., Hong, H., Koh, S.S., Huang, S.M., Schurter, B.T., Aswad, D.W., and Stallcup, M.R. (1999). Regulation of transcription by a protein methyltransferase. Science 284, 2174–2177. ^ te´, J., Shaaban, S., and Bedford, M.T. (2007). The arginine methCheng, D., Co yltransferase CARM1 regulates the coupling of transcription and mRNA processing. Mol. Cell 25, 71–83. Cloutier, P., Lavalle´e-Adam, M., Faubert, D., Blanchette, M., and Coulombe, B. (2013). A newly uncovered group of distantly related lysine methyltransferases preferentially interact with molecular chaperones to regulate their activity. PLoS Genet. 9, e1003210. Di Lorenzo, A., and Bedford, M.T. (2011). Histone arginine methylation. FEBS Lett. 585, 2024–2031. Feng, Q., Yi, P., Wong, J., and O’Malley, B.W. (2006). Signaling within a coactivator complex: methylation of SRC-3/AIB1 is a molecular switch for complex disassembly. Mol. Cell. Biol. 26, 7846–7857. Ficz, G., Branco, M.R., Seisenberger, S., Santos, F., Krueger, F., Hore, T.A., Marques, C.J., Andrews, S., and Reik, W. (2011). Dynamic regulation of 5-hydroxymethylcytosine in mouse ES cells and during differentiation. Nature 473, 398–402. Gu, T.P., Guo, F., Yang, H., Wu, H.P., Xu, G.F., Liu, W., Xie, Z.G., Shi, L., He, X., Jin, S.G., et al. (2011). The role of Tet3 DNA dioxygenase in epigenetic reprogramming by oocytes. Nature 477, 606–610. Guo, F., Li, X., Liang, D., Li, T., Zhu, P., Guo, H., Wu, X., Wen, L., Gu, T.P., Hu, B., et al. (2014). Active and passive demethylation of male and female pronuclear DNA in the mammalian zygote. Cell Stem Cell 15, 447–458. Hackett, J.A., Zylicz, J.J., and Surani, M.A. (2012). Parallel mechanisms of epigenetic reprogramming in the germline. Trends Genet. 28, 164–174. Hatanaka, Y., Shimizu, N., Nishikawa, S., Tokoro, M., Shin, S.W., Nishihara, T., Amano, T., Anzai, M., Kato, H., Mitani, T., et al. (2013). GSE is a maternal factor involved in active DNA demethylation in zygotes. PLoS ONE 8, e60205. Hatanaka, Y., Inoue, K., Oikawa, M., Kamimura, S., Ogonuki, N., Kodama, E.N., Ohkawa, Y., Tsukada, Y., and Ogura, A. (2015). Histone chaperone CAF-1 mediates repressive histone modifications to protect preimplantation mouse embryos from endogenous retrotransposons. Proc. Natl. Acad. Sci. USA 112, 14641–14646.

Inoue, A., and Zhang, Y. (2011). Replication-dependent loss of 5-hydroxymethylcytosine in mouse preimplantation embryos. Science 334, 194. Inoue, A., and Zhang, Y. (2014). Nucleosome assembly is required for nuclear pore complex assembly in mouse zygotes. Nat. Struct. Mol. Biol. 21, 609–616. Inoue, A., Shen, L., Matoba, S., and Zhang, Y. (2015). Haploinsufficiency, but not defective paternal 5mC oxidation, accounts for the developmental defects of maternal Tet3 knockouts. Cell Rep. 10, 463–470. Kohli, R.M., and Zhang, Y. (2013). TET enzymes, TDG and the dynamics of DNA demethylation. Nature 502, 472–479. Kubiura, M., Okano, M., Kimura, H., Kawamura, F., and Tada, M. (2012). Chromosome-wide regulation of euchromatin-specific 5mC to 5hmC conversion in mouse ES cells and female human somatic cells. Chromosome Res. 20, 837–848. Lin, C.J., Koh, F.M., Wong, P., Conti, M., and Ramalho-Santos, M. (2014). Hiramediated H3.3 incorporation is required for DNA replication and ribosomal RNA transcription in the mouse zygote. Dev. Cell 30, 268–279. Mayer, W., Niveleau, A., Walter, J., Fundele, R., and Haaf, T. (2000). Demethylation of the zygotic paternal genome. Nature 403, 501–502. Mizuno, S., Sono, Y., Matsuoka, T., Matsumoto, K., Saeki, K., Hosoi, Y., Fukuda, A., Morimoto, Y., and Iritani, A. (2006). Expression and subcellular localization of GSE protein in germ cells and preimplantation embryos. J. Reprod. Dev. 52, 429–438. ^ te´, J., and Kutateladze, T.G. (2012). Musselman, C.A., Lalonde, M.E., Co Perceiving the epigenetic landscape through histone readers. Nat. Struct. Mol. Biol. 19, 1218–1227. Nakamura, T., Arai, Y., Umehara, H., Masuhara, M., Kimura, T., Taniguchi, H., Sekimoto, T., Ikawa, M., Yoneda, Y., Okabe, M., et al. (2007). PGC7/Stella protects against DNA demethylation in early embryogenesis. Nat. Cell Biol. 9, 64–71. Nakamura, T., Liu, Y.J., Nakashima, H., Umehara, H., Inoue, K., Matoba, S., Tachibana, M., Ogura, A., Shinkai, Y., and Nakano, T. (2012). PGC7 binds histone H3K9me2 to protect against conversion of 5mC to 5hmC in early embryos. Nature 486, 415–419. Nashun, B., Yukawa, M., Liu, H., Akiyama, T., and Aoki, F. (2010). Changes in the nuclear deposition of histone H2A variants during pre-implantation development in mice. Development 137, 3785–3794. Nashun, B., Hill, P.W., Smallwood, S.A., Dharmalingam, G., Amouroux, R., Clark, S.J., Sharma, V., Ndjetehe, E., Pelczar, P., Festenstein, R.J., et al.

(2015). Continuous histone replacement by Hira is essential for normal transcriptional regulation and de novo DNA methylation during mouse oogenesis. Mol. Cell 60, 611–625. Pastor, W.A., Aravind, L., and Rao, A. (2013). TETonic shift: biological roles of TET proteins in DNA demethylation and transcription. Nat. Rev. Mol. Cell Biol. 14, 341–356. Santos, F., Hendrich, B., Reik, W., and Dean, W. (2002). Dynamic reprogramming of DNA methylation in the early mouse embryo. Dev. Biol. 241, 172–182. Sasaki, H., and Matsui, Y. (2008). Epigenetic events in mammalian germ-cell development: reprogramming and beyond. Nat. Rev. Genet. 9, 129–140. Selvi, B.R., Batta, K., Kishore, A.H., Mantelingu, K., Varier, R.A., Balasubramanyam, K., Pradhan, S.K., Dasgupta, D., Sriram, S., Agrawal, S., and Kundu, T.K. (2010). Identification of a novel inhibitor of coactivator-associated arginine methyltransferase 1 (CARM1)-mediated methylation of histone H3 Arg-17. J. Biol. Chem. 285, 7143–7152. Shen, L., Inoue, A., He, J., Liu, Y., Lu, F., and Zhang, Y. (2014). Tet3 and DNA replication mediate demethylation of both the maternal and paternal genomes in mouse zygotes. Cell Stem Cell 15, 459–470. Shimazu, T., Barjau, J., Sohtome, Y., Sodeoka, M., and Shinkai, Y. (2014). Selenium-based S-adenosylmethionine analog reveals the mammalian seven-beta-strand methyltransferase METTL10 to be an EF1A1 lysine methyltransferase. PLoS ONE 9, e105394. Tan, L., and Shi, Y.G. (2012). Tet family proteins and 5-hydroxymethylcytosine in development and disease. Development 139, 1895–1902. Torres-Padilla, M.E., Parfitt, D.E., Kouzarides, T., and Zernicka-Goetz, M. (2007). Histone arginine methylation regulates pluripotency in the early mouse embryo. Nature 445, 214–218. Wossidlo, M., Arand, J., Sebastiano, V., Lepikhov, K., Boiani, M., Reinhardt, R., Scho¨ler, H., and Walter, J. (2010). Dynamic link of DNA demethylation, DNA strand breaks and repair in mouse zygotes. EMBO J. 29, 1877–1888. Wu, Q., Bruce, A.W., Jedrusik, A., Ellis, P.D., Andrews, R.M., Langford, C.F., Glover, D.M., and Zernicka-Goetz, M. (2009). CARM1 is required in embryonic stem cells to maintain pluripotency and resist differentiation. Stem Cells 27, 2637–2645. Yang, Y., McBride, K.M., Hensley, S., Lu, Y., Chedin, F., and Bedford, M.T. (2014). Arginine methylation facilitates the recruitment of TOP3B to chromatin to prevent R loop accumulation. Mol. Cell 53, 484–497.


Cell Reports, Volume 20

Supplemental Information

Histone H3 Methylated at Arginine 17 Is Essential for Reprogramming the Paternal Genome in Zygotes Yuki Hatanaka, Takeshi Tsusaka, Natsumi Shimizu, Kohtaro Morita, Takehiro Suzuki, Shinichi Machida, Manabu Satoh, Arata Honda, Michiko Hirose, Satoshi Kamimura, Narumi Ogonuki, Toshinobu Nakamura, Kimiko Inoue, Yoshihiko Hosoi, Naoshi Dohmae, Toru Nakano, Hitoshi Kurumizaka, Kazuya Matsumoto, Yoichi Shinkai, and Atsuo Ogura

Figure S1 A Linker histones

Core histones

H1 H1foo H2A H2A.Z H2B H3.1 H3.2 H3.3 H4 (213a.a) (305a.a) (373a.a) (127a.a) (127a.a) (137a.a) (137a.a) (137a.a) (104a.a)

GSE −

−

＋

＋

＋

＋

＋

＋

＋

Spleen

Ovary

Testis

Muscle

Brain

Lung

Kidney

Liver

C Heart

Positive control Ovary

Water N.C

Blast 96hpi

Morula 72hpi

8-cell 54hpi

4-cell 48hpi

36hpi

24hpi

1-cell 12hpi

6hpi

MⅡ

GV

Oocytes

B

2-cell

-Ura, -His, -Trp, -Leu

Mettl23 Actin

Mettl23 Actin

Figure S1. Information of interaction factors of GSE protein, related to Figure 1. (A) Interaction of GSE protein with different histones analyzed using a yeast two-hybrid assay. GSE was associated with core histones but not with linker histones, as demonstrated by the blue positive reactions. (B) Immunoblotting analysis for Mettl23 protein expression in mouse oocytes and early embryos. Actin was used as a loading control in immunoblotting analyses. About 50 oocytes/embryos were used for each blot. Image is representative of three independent replicate experiments. (C) Immunoblotting analysis for Mettl23 protein expression in various mouse tissues. Actin was used as a loading control. About 25 g protein was used for each blot. Image is representative of three independent replicate experiments.

15 kDa

kDa

H3R17 me2a

DAPI

Merge ♀ PB ♂

PN3 (6 hpi)

CARM1

H3R17 me2a

DAPI

1.02

H3R17 Mettl23 me2a

F

GV H3R17 me2a

1.13

PB

♂

E

PN3

Merge ♀

CARM1

72 hpi Morula

H3R17me2a

H3 15

Signal intensity (M/F ratio)

Mettl23

54 hpi 8-cell

6 hpi

H3R17 me2a Actin

D C

48 hpi 4-cell

25 20

36 hpi

25 20

24 hpi

50 37

12 hpi

75

50 37

1-cell

Oocyte GV

siControl Mettl23 -KD

75

2-cell

B

MII oocytes

siControl Mettl23 -KD

MII oocytes

M II

A

96 hpi Blastocyst

Figure S2

Input Purified

3xFLAG-His-Mettl23 －＋ DAPI

Merge

－＋

75

α-CARM1

37

α-FLAG

kDa

Figure S2. H3R17me2a enzyme Mettl23 and CARM1 in oocytes and early mouse embryos related to Figure 2. (A) The effects of Mettl23-KD on the levels of H3 and H3R17me2a in MII oocytes. About 30 oocytes were used for each experiment. Image is representative of three independent replicate experiments. (B) Immunoblot analysis for H3R17me2a in mouse oocytes and early embryos. About 50 oocytes/embryos were used for each blot. Image is representative of three independent replicate experiments. (C) Immunocytochemical analysis of Mettl23, CARM1, and H3R17me2a in PN3 zygotes. DNA was stained with DAPI. PB, the second polar body. Scale bars = 50 m. At least 10 zygotes were analyzed for each group. (D) Fluorescence intensity ratios of the male and female PNs (M/F ratios) for Mettl23 and H3R17me2a in PN3 zygotes. Each dot represents a single zygote. The intensities are similar between the male and female PNs (M/F ratio ~1). (E) Immunocytochemical analysis of CARM1 (green) and H3R17me2a (red) in GV oocytes, DNA was stained with DAPI (blue). PB, the second polar body, Scale bar = 50 m. At least 10 oocytes were analyzed for each group. (F) Purified 3xFLAG-His6-Mettl23 does not contain CARM1. The purified 3xFLAGHis6-Mettl23, which is same as in Fig. 2C and D, was subjected to immnoblotting with anti-CARM1. The immunoblot failed to detect CARM1 protein in the precipitates (lane 3-4). The input (lane 1-2) showed the CARM1 protein. One-tenth amount of the precipitates was used for the immnoblot with anti-FLAG antibody.

Figure S3 A

5’ AAGAGGAGGTCGCAATGGCCGAGCCAACCGTTGTCATAGCCCCCACCACA 3’ WT AAGAGGAGG - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - CCCCCACCACA Mut

GSE WT

Mutant 5’ TTTGCCATGGACAGTGTCCGGCCCCGCGCT 3’ WT Mut TTTGCCATGG - - - - - - - - - - - CCCCGCGCT

Mettl23

WT

Mutant

C +/-

Adult testis +/- -/-

-/-

GSE IB

IB Actin

Mettl23 Actin

Mettl23-KO 12

ns

10

Number of pups

Adult testis

GSE-KO 12

Number of pups

B

8 6 4 2 0

＊

10 8 6 4 2 0

Hetero KO

Hetero

KO

Figure S3. Production of GSE and Mettl23 knockout mice by genome editing using the CRISPR/Cas9 system, related to Figure 3. (A) The sequence of targeted regions in the genes encoding GSE and Mettl23. The START codon is boxed in solid blue. PAM sequence is marked with red characters. (B) Immunoblot analysis for the expression levels of GSE and Mettl23 proteins in the testes from heterozygous and homozygous male mice. Actin was used as a loading control. (C) The numbers of pups born after natural mating of KO mice. “Hetero” means mating of heterozygous pairs of mice and “KO” means mating of homozygous pairs. Each dot represents the number of pups per litter at the day of birth. Red bars represent the mean number. Asterisks indicate significant differences compared with control embryos (p < 0.0005); ns, not significant.

Figure. S4 A

PN3 siControl 5mC

Mat Mettl23-KO Mettl23-injected Untreated

Mat GSE-KO

5hmC Merge

5mC

5hmC Merge

PB

#1

5mC

5hmC Merge

PB

PB

♀

♀

♂

♀

♂ ♂

PB ♀

♀

♀ PB

♂

♀ ♂

♂

PB

♂

♂

♂

♀

♂

♀

♀

PB ♀

#5

PB

♂

♀ ♂

PB ♀

#4

PB

♀ ♂

PB

♂

♂

♀

♂

♂

#3

5hmC Merge

♂

♂

#2

5mC

♀

♀

♀

PB ♀

♂

♀ PB

♀

PB

♂ ♀

#6

♂

#9 #10

B

♂

♂

C

PN5 Mat Mettl23-KO Mat Mettl23 GSE-KO Untreated -injected

WT

Mat Mettl23 GSE-KO Untreated -injected

PB

♀

♀ ♂

Mat Mettl23-KO WT

♂

PB ♂

PN5

PB

♂

PB ♀

♀

♀

♀ ♀

♂

PB ♂

♂

PB ♀

♀

♀

♂ ♂ ♂

♀

♂

PB PB ♀ PB ♀

♀

PB

♂

♀ ♂

PB ♂

♂ ♀

♀

♀

♂

#8

PB

♂

PB

♀

#7

PB

Tet3 5mC H3 5hmC ♀

♀

♂

♂

PB

DAPI

♂ PB ♂

♀

Merge

TP male

♂

TP female

WT

＊＊

Mat GSE-KO Untreated Mettl23injected

Mettl23 -KO

Relative Tet3 signal intensity

WT

＊＊

Mettl23 -KO

♀

♂

5hmC


WT

＊＊


Signal intensity (M/F ratio)

5mC

♀ PB

♂ ♀

Mettl23 -KO

WT

Merge

♀


♀ ♂

Mettl23 -KO

4

Figure S4. GSE and Mettl23 are indispensable for active DNA demethylation by recruiting Tet3 to the PNs, related to Figure 4. (A) Representative images of 5mC (green) and 5hmC (red) in control WT, maternal Mettl23-KO, and maternal GSE-KO zygotes at PN3. PB, the second polar body. Scale bar = 50 m. (B) Upper panel: representative immunofluorescence images of 5mC and 5hmC in control WT, maternal Mettl23-KO, and maternal GSE-KO zygotes. PB, the second polar body. Scale bar = 50 m. Lower panel: the fluorescence M/F ratios for 5mC (left) and 5hmC (right) in PN5 zygotes. Each dot represents a single zygote. Red bars show the mean ratios. Asterisks indicate significant differences compared with WT zygotes (P < 0.05). Ten zygotes were analyzed for each group. (C) Upper panel: representative immunofluorescence images of Tet3 in control WT, maternal Mettl23-KO, and maternal GSE-KO zygotes at the PN5 stage. H3 is shown in green as a control for chromatin-bound factors. DNA was stained with DAPI (blue). PB, the second polar body. Scale bar = 50 m. Lower panel: the relative Tet3 signal intensities in the male and female PNs. The level of H3 was set as 1.0. Red bars show the mean ratios. Asterisks indicate significant differences compared with WT zygotes (P < 0.05); ns, not significant. Ten zygotes were analyzed for each group.

Table S1 Table S1| Information of antibodies, related to Experimental Procedures Name Dilution Application Anti-GSE (raised in our laboratory) 1:2000 WB Anti-GSE (raised in our laboratory) 1:2000 IF Anti-GSE (raised in our laboratory) 1µg IP Anti-GSE (raised in our laboratory) 1µg ChIP Anti-Flag M2 (SIGMA; F3165) 1:10000 WB Anti-ACTIN (Santa Cruz; sc-1616) 1:10000 WB Anti-Mettl23 (raised in our laboratory) 1:2000 WB Anti-Mettl23 (raised in our laboratory) 1:2000 IF Anti-Mettl23 (raised in our laboratory) 1µg IP Anti-Mettl23 (raised in our laboratory) 1µg ChIP Anti-H3R8me2a (Active Motif; 39651) 1:1000 WB Anti-H3R17me2a (Millipore Corp; 07-214) 1:1000 WB Anti-H3R17me2a (Millipore Corp; 07-214) 1:1000 IF Anti-H3R17me2a (Abcam; ab8284) 1µg ChIP Anti-H3R26me2 (Abcam; ab194679) 1:1000 WB Anti-H3 (MAB; MA301B) 1:1000 IF Anti-H3 (MAB; MA301B) 1:1000 WB Anti-CARM1 (Santa Cruz; sc-5418) 1:500 IF Anti-TET3 (a gift from G.L.Xu) 1:50000 IF Anti-TET3 (a gift from G.L.Xu) 1µg IP Anti-PGC7 (a gift from T. Nakamura) 1:500 IF Anti-5mC (Calbiochem; NA81) 1:2000 IF Anti-5mC (Calbiochem; NA81) 1µg MeDIP Anti-5hmC (Active motif; 39769) 1:2000 IF Anti-5hmC (Active motif; 39769) 1µg hMeDIP Anti-rabbit IgG-Alexa 555 (Invitrogen; A-21428) 1:2000 IF Anti-mouse IgG-Alexa 488 (Invitrogen; A-21202) 1:2000 IF Anti-goat IgG-Alexa 488 (Invitrogen; A-11078) 1:2000 IF Anti-rabbit IgG HRP conjugate (MILLIPORE; AP182P) 1:20000 WB Anti-goat IgG HRP conjugate (MILLIPORE; AP180P) 1:20000 WB Anti-mouse IgG HRP conjugate (MILLIPORE; AP192P) 1:20000 WB

Table S2

Table S2| Information of primers, related to Experimental Procedures Name Sequence (5' to 3') GSE_F1 TTGAATTCCACGGTTCCTAGCGGTTC GSE_R1 TTGGATCCGCTAAACATTCCTCATCCGATT Mettl23_F1 TTGAATTCAAAACTTCTGTGGGCCAAT Mettl23_R1 TTGGATCCTATTGGGCCAGGACTACAGC GSE_F2 CAGTCATGAGGAGCGGCTCG GSE_R2 GCAGGCTTGGAATCAATCGG Mettl23_F2 TGGCCTACGCAACACTTCCG Mettl23_R2 ACCAGACCTGGTGCTAATGA LINE1_F GTCTGTACCACCTGGGAACTG LINE1_R TGCGCAGATATCTTGTATTTGG LINE1_F AGTGCAGAGTTCTATCAGACCTTC LINE1_R AACCTACTTGGTCAGGATGGATG IAP_F CTCCATGTGCTCTGCCTTCC IAP_R CCCCGTCCCTTTTTTAGGAGA

Application SSA assay SSA assay SSA assay SSA assay Genotyping Genotyping Genotyping Genotyping ChIP ChIP MeDIP/hMeDIP MeDIP/hMeDIP ChIP ChIP

SUPPLEMENTAL EXPERIMENTAL PROCEDURES Collection of Oocytes, in vitro Fertilization (IVF), and Embryo Culture In this study, all zygotes observed were prepared by IVF. Collection of spermatozoa, oocytes, and fertilized embryos was performed as described (Mizuno et al., 2006; Hasegawa et al., 2014). Oocytes were collected from B6D2F1 (C57BL/6N  DBA/2 hybrid) or ICR strain female mice that had been superovulated with equine chorionic gonadotropin (Sankyo Yell Yakuhin) and human chorionic gonadotropin (hCG; Aska Pharmaceutical Co.) at 46- to 52-h intervals. Spermatozoa were collected from the cauda epididymides of ICR male mice. The sperm suspension was preincubated in human tubal fluid (HTF) medium for 1.5 h at 37 °C under 5% CO2 in humidified air. Cumulus– oocyte complexes were recovered into preequilibrated HTF medium. The sperm suspension was added to the oocyte cultures, and morphologically normally fertilized oocytes were collected 2 h after insemination. Fertilized embryos were cultured in potassium simplex optimized medium (KSOM) (Lawitts et al., 1993) at 37 °C under 5% CO2 in humidified air.

Yeast Two-hybrid Screening Yeast two-hybrid screening was performed as described (Shin et al., 2013). In brief, the GSE cDNA was amplified using polymerase chain reaction (PCR) and subcloned into the pGilda vector (TaKaRa Bio Inc.). A mouse ovarian cDNA library in the vector pB42AD was screened. The EGY48 yeast strain used for the screening assay contained both Leu2 and LacZ reporter genes under the control of LexA-responsive upstream activation sites. For the assay, bait and library plasmids were used simultaneously to transform yeast using the lithium acetate procedure. Double transformant cells grown on Ura-, His-, Trp-, and Leu- plates were incubated for 3 days at 30 °C. Positive colonies were selected and assayed for the LacZ phenotype. Putative positives were detected and then further tested by assaying the colonies for -galactosidase activity. Following confirmation of the specificity of the interaction, GSE-binding partners were identified by sequence analysis. To identify the interaction domain of GSE with Mettl23, partial GSE cDNA fragments of aa 1-134, aa 135-174, and aa 175-247 were PCR-amplified and subcloned into the pGilda LexA vector, and the full-length open reading frame (ORF) sequence of mouse Mettl23 was PCR-amplified and subcloned into the pB42AD vector.

Yeast Two-hybrid System Yeast two-hybrid system assays were performed as described above. To investigate the interaction of GSE with histones H1, H1foo, H2A, H2A.Z, H2B, H3.1, H3.2, H3.3, and H4, GSE cDNA sequences were PCR-amplified and subcloned into the pGilda LexA vector, and the full-length ORF sequences of mouse histones H1, H1foo, H2A, H2A.Z, H2B, H3.1, H3.2, H3.3, and H4 were PCRamplified and subcloned into the pB42AD vector. To identify the interaction of GSE with Tet3, GSE cDNA was PCR-amplified and subcloned into the pGilda LexA vector and the full-length ORF sequence of mouse Tet3 and partial Tet3 cDNA fragments of aa 1-583, aa 551-1106, aa 1012-1482, and aa 1460-1669 were PCR-amplified and subcloned into the pB42AD vector.

Coimmunoprecipitation This was performed as described (Hirano et al., 2006) using 300 embryos at the PN3 stage. We used antibodies against GSE, Mettl23, and Tet3 (1 μg). For immunoblot analysis, we used a donkey anti-rabbit IgG–horseradish peroxidase (HRP) conjugate (1:50,000; AP182P; Millipore Corp.) for anti-GSE and anti-Tet3, and a donkey anti-mouse IgG HRP conjugate (1:50,000; AP192P; Millipore Corp.) for anti-Mettl23, as secondary antibodies (Table S1).

Western Blot Analysis The procedures were essentially performed as described (Satoh et al., 2009). In brief, protein samples (30 cells for each embryonic stage) were extracted and then subjected to sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis. The proteins were resolved in 12% or 15% running gels and transferred electrophoretically to polyvinylidene difluoride membranes (GE Healthcare). The membranes were incubated in Block Ace (DS Pharma Biomedical) at room temperature (RT) for 1 h. They were washed with phosphate-buffered saline containing 0.2% Tween 20 (PBST) and incubated at 4 °C overnight with antibodies (Table S1). The membranes were washed in PBST, incubated with HRP-conjugated antibodies at RT for 1 h, washed with PBST, and developed using ECL Prime Western Blotting detection reagents (GE Healthcare). Densitometric quantification of the immunoblot bands was performed using a Molecular Imager FX with Quantity One software (BioRad).

Purification of 3xFLAG-His6-Mettl23 protein in HEK293 cells HEK293T cells were transfected with 3xFLAG-His6-tagged Mettl23 expression vector. At 48 h posttransfection, the transfected cells were harvested and lysed with His-Lysis buffer (20 mM Tris-HCl at pH 8.0, 500 mM NaCl, 0.5% NP-40, 1 mM phenylmethanesulfonyl fluoride (PMSF), and 1x protease inhibitor cocktail; Nacalai tesque, Japan). After centrifugation at 20,400 g (15,000 rpm) for 10 min, the supernatants were incubated with Ni-NTA Agarose (Qiagen) for 2 h at 4°C with gentle rotation. The Ni-NTA beads were washed 3 times with His-Lysis buffer and additional 3 times with Wash buffer (50 mM Tris-HCl, pH 8.0, 20 mM imidazole, and 0.1% NP-40) and then eluted with 500 µL Elution buffer (50 mM Tris-HCl pH 8.0, 250 mM imidazole, and 0.1% NP-40). The eluates were diluted with 1.5 mL His-Lysis buffer and incubated with anti-FLAG affinity gel (Sigma-Aldrich) for 2 h at 4°C. The FLAG M2 beads were washed twice with His-Lysis buffer and once with 50 mM TrisHCl at pH 8.0. The FLAG M2 beads were used for in vitro methylation assay. Un-transfected HEK293T cells were subjected to the same procedure described above and used as negative control.

In vitro methylation assay One and a half micrograms of recombinant Histone H3.1 protein (M2503S from NEB, or purified from E.coli) was incubated in reaction buffer (50 mM Tris-HCl pH 8.5, 5 mM DTT) with 3xFLAGHis6-Mettl23-immobilized FLAG M2 agarose beads (A2220 from Sigma-Aldorich) and 14C-labeled

SAM (0.01 µCi, Perkin Elmer) for at least 3 h at 30°C. The reaction was stopped by adding Laemmli SDS-sample buffer. Proteins were resolved on acrylamide SDS-PAGE gel, and the dry gel was exposed to an imaging plate (FUJI-FILM) for 48 h to 72 h, and the autoradiographic signal was detected using a BAS-5000 Image analyzer (FUJI-FILM).

Production of recombinant H3.1 mutations The H3.1 R17K and H3.1 R17A mutations were introduced in the DNA fragment encoding human histone H3.1 by site-directed mutagenesis. Recombinant human histone H3.1 and the H3.1 R17K and H3.1 R17A mutants were expressed in Escherichia coli cells, and were purified, as described previously (Tachiwana et al., 2010).

Mass spectrometry The recombinant H3.1 with or without treatment of Mettl23 were subjected to SDS-PAGE and the protein bands of the recombinant H3.1 were digested with trypsin in gel. The digestion mixture was separated on a nanoflow LC (Easy nLC; Thermo Fisher Scientific) using a nano-electrospray ionization spray column (self-packed Hypercarb column , φ100 μm × 100 mm, 5 μm, Thermo Fisher Scientific) equilibrated with buffer A (0.1% formic acid in water) and eluted with a linear gradient of 0-35% buffer B (0.1% formic acid in 100% ACN) at a flow rate of 300 nL/min over 10 min and subjected on-line to a Q-Exactive mass spectrometer (Thermo Fisher Scientific) that was equipped with a nanospray ion source. MS/MS chromatograms and spectra were acquired using targeted MS/MS method according to the inclusion list of three doubly protonated parent ions of unmethylated (me0, m/z= 236.1555), monomethylated (me1, m/z=243.1634) and dimethylated (me2, m/z=250.1712) forms of APRK peptides. And selected ion chromatograms of y3 ion (me0 m/z= 400.267± 20ppm, me1 m/z=414.283± 20ppm and me2 m/z=428.299 ± 20ppm) of these parent ions were drawn (Figure 2F).

DNA Immunoprecipitation with Anti-5mC and Anti-5hmC Antibodies These procedures were performed essentially as described (Gu et al., 2011). In brief, to extract the DNA, samples were incubated in a proteinase K (Roche) solution including 1  SSC buffer, proteinase K (50 ng), and 10% SDS for 1 h at 37 °C. Each sample was incubated with RNase for 1 h at 37 °C to remove total RNA. The samples were purified by phenol-chloroform treatment and ethanol precipitation with glycogen. The samples were fragmented by AluI digestion overnight at 37 °C and purified by phenol-chloroform treatment and ethanol precipitation with glycogen. After being denatured for 10 min at 95 °C, the immunoprecipitation of prepared DNA samples was performed as described above using 1 μg of anti-5mC (NA8; Calbiochem) or anti-5hmC (39769; Active Motif) antibodies, and 10 μl of Dynabeads (Life Technologies). Quantitative reverse transcription (RT)-PCR analyses were performed as described below. The primers used in this study are described in Table S2. Three independent experiments were performed.

Quantitative Real-time RT–PCR Analyses Quantitative RT-PCR analyses were performed as described (Hatanaka et al., 2013). Total RNA was isolated from 30 pooled oocytes or embryos using RNAqueous-Micro kits (Ambion). In brief, cDNA was synthesized from total RNA using High Capacity cDNA Reverse Transcription kits (Applied Biosystems). Prepared cDNA samples were amplified and analyzed by quantitative RTPCR. The primers used are described in Table S2. Amplifications were run in a 7900HT Sequence Detector system (Applied Biosystems). Microinjection of Mettl23 siRNA For knockdown of Mettl23, MII oocytes were injected with a specific siRNA using a Piezo-driven micromanipulator and cultured in KSOM until use. To determine the knockdown efficiency, proteins isolated from injected oocytes were used for immunoblotting. The injected oocytes were then fertilized by IVF as described above, except that a small incision was made in the zona pellucida using a Piezo-driven micromanipulator to facilitate sperm penetration.

Production of GSE- and Mettl23-KO Mice by the CRISPR/Cas9 System Production of GSE- and Mettl23-KO mice by the CRISPR/Cas9 system was performed as described (Honda et al., 2014; Mashiko et al., 2013). In brief, a pCAG-EGxxFP plasmid expressing a 5’ and 3’ enhanced green fluorescent protein (EGFP) sequence was used to perform the single-strand annealing (SSA) assay (Mashiko et al., 2013). The 511-bp or 595-bp genomic fragments of the genes encoding GSE or Mettl23 containing a single guide (sg) RNA target sequence were amplified by PCR and inserted between the EGFP fragments. Plasmids coexpressing hCas9 and sgRNA were prepared by ligating oligonucleotides into pX330 (http://www.addgene.org/42230/). The primers and oligonucleotides used are described in Table S2. The SSA assay was performed as described (Honda et al., 2014). IVF was performed as described above. Oocytes were collected from superovulated female B6D2F1 mice and spermatozoa were collected from male C57BL/6N mice. Fertilized embryos at the PN stage were injected with 5 ng/μl of pX330 plasmids. The embryos were cultured in KSOM until the 2-cell stage and transferred into the oviducts of pseudopregnant ICR female mice. For genotyping of the offspring, tail tips were collected from the pups at 1 week after birth, and genomic DNA was extracted and purified. Prepared genomic DNA samples were amplified by PCR and sequenced using a 3730 DNA analyzer (Applied Biosystems). The primers used are described in Table S2.

In Vitro mRNA Synthesis This was performed as described (Hatanaka et al., 2015). In brief, H3.3R8K, H3.3R17K, and H3.3R26K were produced using a PrimeSTAR Mutagenesis Basal Kit (TaKaRa). Each cassette of Flag-Tet3, Flag-Tet3mut, 3xFLAG-His6-Mettl23, EGFP-H3.1, EGFP-H3.3, EGFP-H3.3R8K, EGFPH3.3R17K, EGFP-H3.3R26K was cloned into the pcDNA4 vector (Invitrogen). Amplification and poly(A) tailing of these mRNAs were performed using mMESSAGE mMACHINE T7 Ultra (Ambion) for each construct. Microinjection of each mRNA was performed as described above.

Microinjection of Mettl23 siRNA For knockdown of Mettl23, MII oocytes were injected with a specific siRNA using a Piezo-driven micromanipulator and cultured in KSOM until use. To determine the knockdown efficiency, proteins isolated from injected oocytes were used for immunoblotting. The injected oocytes were then fertilized by IVF as described above, except that a small incision was made in the zona pellucida using a Piezo-driven micromanipulator to facilitate sperm penetration.

Production of GSE- and Mettl23-KO Mice by the CRISPR/Cas9 System Production of GSE- and Mettl23-KO mice by the CRISPR/Cas9 system was performed as described (Honda et al., 2014; Mashiko et al., 2013). In brief, a pCAG-EGxxFP plasmid expressing a 5’ and 3’ enhanced green fluorescent protein (EGFP) sequence was used to perform the single-strand annealing (SSA) assay (Mashiko et al., 2013). The 511-bp or 595-bp genomic fragments of the genes encoding GSE or Mettl23 containing a single guide (sg) RNA target sequence were amplified by PCR and inserted between the EGFP fragments. Plasmids coexpressing hCas9 and sgRNA were prepared by ligating oligonucleotides into pX330 (http://www.addgene.org/42230/). The primers and oligonucleotides used are described in Table S2. The SSA assay was performed as described (Honda et al., 2014). IVF was performed as described above. Oocytes were collected from superovulated female B6D2F1 mice and spermatozoa were collected from male C57BL/6N mice. Fertilized embryos at the PN stage were injected with 5 ng/μl of pX330 plasmids. The embryos were cultured in KSOM until the 2-cell stage and transferred into the oviducts of pseudopregnant ICR female mice. For genotyping of the offspring, tail tips were collected from the pups at 1 week after birth, and genomic DNA was extracted and purified. Prepared genomic DNA samples were amplified by PCR and sequenced using a 3730 DNA analyzer (Applied Biosystems). The primers used are described in Table S2.

In Vitro mRNA Synthesis This was performed as described (Hatanaka et al., 2015). In brief, H3.3R8K, H3.3R17K, and H3.3R26K were produced using a PrimeSTAR Mutagenesis Basal Kit (TaKaRa). Each cassette of Flag-Tet3, Flag-Tet3mut, 3xFLAG-His6-Mettl23, EGFP-H3.1, EGFP-H3.3, EGFP-H3.3R8K, EGFPH3.3R17K, EGFP-H3.3R26K was cloned into the pcDNA4 vector (Invitrogen). Amplification and poly(A) tailing of these mRNAs were performed using mMESSAGE mMACHINE T7 Ultra (Ambion) for each construct. Microinjection of each mRNA was performed as described above.

SUPPLEMENTAL REFERENCES Hasegawa, A., Mochida, K., Tomishima, T., Inoue, K., and Ogura, A. (2014). Microdroplet in vitro fertilization can reduce the number of spermatozoa necessary for fertilizing oocytes. J. Reprod. Dev. 60, 187–193.

Hirano, Y., Hayashi, H., Iemura, S., Hendil, K.B., Niwa, S., Kishimoto, T., Kasahara, M., Natsume, T., Tanaka, K., and Murata, S. (2006). Cooperation of multiple chaperones required for the assembly of mammalian 20S proteasomes. Mol. Cell 24, 977–984.

Honda, A., Hirose, M., Sankai, T., Yasmin, L., Yuzawa, K., Honsho, K., Izu, H., Iguchi, A., Ikawa, M., and Ogura, A. (2014). Single-step generation of rabbits carrying a targeted allele of the tyrosinase gene using CRISPR/Cas9. Exp. Anim. 1, 31–37.

Lawitts, J.A., and Biggers, J.D. (1993). Culture of preimplantation embryos. Methods Enzymol 225, 153–164.

Mashiko, D., Fujihara, Y., Satouh, Y., Miyata, H., Isotani, A., and Ikawa, M. (2013). Generation of mutant mice by pronuclear injection of circular plasmid expressing Cas9 and single guided RNA. Sci. Rep. 3, 3355.

Satoh, M., Tokoro, M., Ikegami, H., Nagai, K., Sono, Y., Shin, S.W., Nishikawa, S., Saeki, K., Hosoi, Y., Iritani, A., et al. (2009). Proteomic analysis of the mouse ovary in response to two gonadotropins, follicle-stimulating hormone and luteinizinghormone. J. Reprod. Dev. 55, 316–326.

Shin, S.W., Shimizu, N., Tokoro, M., Nishikawa, S., Hatanaka, Y., Anzai, M., Hamazaki, J., Kishigami, S., Saeki, K., Hosoi, Y., et al. (2013). Mouse zygote-specific proteasome assembly chaperone important for maternal-to-zygotic transition. Biol. Open 2, 170–182.

Tachiwana, H., Kagawa, W., Osakabe, A., Kawaguchi, K., Shiga, T., Hayashi-Takanaka, Y., Kimura, H., Kurumizaka, H. (2010). Structural basis of instability of the nucleosome containing a testisspecific histone variant, human H3T. Proc Natl Acad Sci U S A. 107, 10454-10459.