Supplemental Information

Chromatin-to-nucleoprotamine transition is controlled by the histone H2B variant TH2B

Emilie Montellier1, Fayçal Boussouar1, Sophie Rousseaux1, Kai Zhang2, Thierry Buchou1, François Fenaille3, Hitoshi Shiota1, Alexandra Debernardi1, Patrick Héry4, Sandrine Curtet1, Mahya Jamshidikia1, Sophie Barral1, Hélène Holota5, Aurélie Bergon5, Fabrice Lopez5, Philippe Guardiola6, Karin Pernet7, Jean Imbert5, Carlo Petosa8, Minjia Tan9,10, Yingming Zhao9,10, Matthieu Gérard4, Saadi Khochbin1*

1 - INSERM, U823; Université Joseph Fourier - Grenoble 1; Institut Albert Bonniot, Grenoble, F-38700 France 2 - State Key Laboratory of Medicinal Chemical Biology & Department of Chemistry, Nankai University, Tianjin 300071, China 3 – Laboratoire d'Etude du Métabolisme des Médicaments, DSV / iBiTec-S / SPI, CEA Saclay, 91191 Gif sur Yvette, Cedex, France 4 - CEA, iBiTec-S, Gif-sur-Yvette, F-91191 France 5 - INSERM UMR_S 1090; TGML/TAGC, Aix-Marseille Université, Marseille, France 6 - INSERM, U892; Centre de Recherche sur le Cancer Nantes Angers et UMR_S 892; Université d’Angers; Plateforme SNP, Transcriptome & Epigénomique; Centre Hospitalier Universitaire d’Angers; Angers, F-49000 France 7 - INSERM U836; Université Joseph Fourier, Grenoble Institute of Neuroscience, Grenoble, F-38700 France, 8 - Institut de Biologie Structurale Jean-Pierre Ebel, UMR 5075 CEA-CNRS-Université Joseph Fourier, 41 Jules Horowitz, 38027 Grenoble cedex 1. 9 - Ben May Department of Cancer Research, The University of Chicago, Chicago, IL 60637, USA 10 - Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zuchongzhi Road, Shanghai 201203, P.R. China

*Correspondance : Saadi Khochbin ([email protected]).

1 SUPPLEMENTAL DATA

SUPPLEMENTAL FIGURE LEGENDS

Figure S1. Generation of the Th2b tag knock-in and knock-down mouse models. A. Schematic representation of Th2b and Th2a genes. A “His-Flag-Ha” tag and a LoxP flanked Neo cassette were introduced by homologous recombination in fusion in frame with Th2b coding sequence replacing the STOP codon of the gene. The presence of the Neo cassette interferes with Th2b gene expression leading to the absence of TH2B expression in the spermatogenic cells of Th2bNeo/Neo mice. Crossing the Th2bNeo/Neo mice with transgenic CMV-Cre mice leads to the deletion of the Neo cassette and the expression of TH2B-tag. B. The shared regulatory region of Th2b and Th2a genes is shown in comparison with the corresponding rat sequence (in blue), where transcriptional regulatory elements have been characterized (indicated) (Huh et al., 1991). C. DNA prepared from Neo-resistant ES clones was subjected to southern blot analysis (left panel) using a 3’ external probe (indicated in A). Lane 1: heterozygous; lane 2 and 3: wild type. Heterozygous Th2b +/Neo mice were crossed to CMV-Cre mice to delete the Neo cassette. PCR (right panel) analysis distinguishes the wild type allele from the recombined allele (Th2bdelta-neo). Lanes 1 and 2: Th2bdelta-neo mice; lane 3: Th2b+/+ mouse.

Figure S2 (related to Figure 4A and B). Mass spectrometry analysis of histones within TH2B-containing nucleosomes. Protein ID, Mascot scores and identified sequences coverage were obtained from Mascot search of high-resolution MS/MS spectra against the IPI mouse v3.74 protein sequence database. Searching parameters were set as described in the Experimental Procedures. All the identified tryptic peptides with Mascot scores above 20.0 are shown in red color.

Figure S3. A. H2A.Z but not TH2B at gene TSS is associated with higher gene transcriptional activity. The overall expression of H2A.Z+ and H2A.Z-less gene categories (shown in Figure 4D) in both spermatocytes (spc) and round spermatids (R-Spt) are shown as box-plots. The expression values obtained from our transcriptomic experiments were normalized using the mean expression value of all genes in spermatocytes. * p value (t-test). B. FRAP experiment comparing the mobility of tagged and untagged GFP-TH2B. H1299 lung carcinoma cells stably expressing GFP-TH2B and GFP-TH2B-tag were established (GFP fluorescence in the corresponding cell lines is shown, upper panels).

2 The middle panel shows the time course of background-corrected fluorescence in bleached area, showing mean intensity and standard deviation from ten independent experiments for each protein (middle panel). The lower panel shows the FRAP profile analysis. FRAP profiles were compatible with the presence of two populations of mobile species, exhibiting high and low mobility. Individual

-k t -k t FRAP curves were fit to the equation F(t) = A(1 – e 1 ) + B(1 – e 2 ), where F(t) is the background-corrected fluorescence at time t, A and B are the fraction of high- and low-mobility species, respectively, and k1 and k2 are the corresponding first-order rate constants. The half- lives (t1/2) of the high- and low-mobility species are given by ln2/k1 and ln2/k2, respectively, while the fraction of molecules that are immobile (on the time scale of the experiment) is given by 1-A-B. The table shows the summary of mobility parameters derived from FRAP analysis. The values represent the mean and standard deviation from ten independent experiments.

Figure S4 (related to Figure 5). TH2B resists replacement by transition proteins and protamines in TH2B-tag expressing elongating spermatids. Co-detections of TH2B either with Transition Protein 2 (TP2, left panels) or with Protamine 1 (Prm1, right panels) (in the indicated colors) were performed on spermatogenic cell preparations from wild-type or Th2b+/tag testes by immunofluorescence using the corresponding antibodies. The squares indicate portions showed at higher magnification in Figure 5B.

Figure S5. A (related to Figure 6B). Characterization of H2B isoforms accumulation in response to the absence of TH2B. Deconvoluted electrospray mass spectra of oxidized mouse histones H4 and H2B are shown. Histone identification was realized by matching experimentally determined masses with theoretical ones obtained from known amino acid sequences and post-translational modifications (Table S3). Heights of peaks corresponding to the different H2B variants were expressed relative to the heights of the major H4 peaks (unacetylated H4 + monoacetylated H4), and H2B/H4 ratios calculated for wild-type and TH2B-less mice. Numbers in parenthesis represent the variations of these H2B/H4 ratios between wild-type and TH2B-less mice, and are the mean values obtained from three independent biological replicates (see Figure S5B for more details). B (related to Figure 6B). H2B isoform levels in the presence (wild type) or absence (TH2B-less) of TH2B in spermatogenic cells. Heights of peaks on Figure S5A corresponding to the different H2B isoforms were expressed relative to those of the major H4 forms (unacetylated H4 +

3 monoacetylated H4). H2A/H4 ratio was also calculated for TH2B wild-type and TH2B-less mice as a control. Each bar is the average from three distinct histone extracts (from three distinct mice). Bars indicate standard deviations. The numbers on top of the red bars represent the fold changes of the H2B/H4 ratios between wild-type and TH2B-less mice. C. Equal sensitivity of dissociating nucleosomes from wild-type and TH2B-less elongating/condensing spermatids to MNase digestion. Nuclei from wild-type elongated/condensing spermatids or from the corresponding cells isolated from TH2B-less cells were extensively digested by MNase as described in the legend of Figure 5A. The agarose gel shows nucleosomal and sub- nucleosomal DNA fragments released by MNase digestion from the two cell types during the indicated times. D. (related to Figure 6C). MS/MS spectra of the modified peptides in histones in wild-type or TH2B-less spermatogenic cells. All the identified peptides with Mascot score above 20 were manually verified according to the rules described previously (Chen et al., 2005). All the identified MS/MS fragment ions (“b” and “y” ions) are shown in red color. All the identified PTM and propionylated sites were annotated. “Pr’’ indicates a propionylated N-terminal residue or lysine residue; “Kac”, “Kme”, “Kme2” and “Kcr” respectively indicate acetylation, mono-methylation, di-methylation and crotonylation at a lysine residue; “Rme” indicates mono- methylation at an arginine residue.

Figure S6. TH2B-less males and females as well as Th2b+/tag female mice are fertile. A. No significant differences in the litter sizes could be observed when TH2B-less male and female mice were crossed compared to wild-type animals (n = 5 couples of each genotype). B. TH2B- tag expression in oocytes and eggs of Th2b+/tag female mice has no effect on the litter size when these mice are crossed with wild-type males compared to the litter size obtained for the corresponding wild-type animals (n = 6 couples of each genotype).

Figure S7 (related to Figure 7A). TH2B disappears from embryonic chromatin during early development. Th2b+/tag embryo cryosections obtained at 13.5 days after birth were stained with the indicated antibodies.

4 SUPPLEMENTARY TABLES

Table S1. List of histone variants identified in TH2B-containing nucleosomes. Mascot score for identified protein * (for detailed information, see Figure S2); The number of matched peptides for every identified protein (all identified tryptic peptides with Mascot scores above 20.0)** ; N/A Not detected.

Table S2. List of specific peptides identified in every H1-isoform. * The specific peptide sequence was identified in every H1-isoform. ** Mascot score for identified peptides (all identified tryptic peptides with Mascot scores above 20.0).

Table S3. Masses and accuracies of the main mouse histone species observed by UHPLC- MS and depicted on Figure 6B and Figure S5A. Theoretical and observed masses are monoisotopic masses. Theoretical masses were calculated based on primary sequences and modifications (post-translational modifications along with oxidized of methionine and cysteine residues).

Table S4. List of identified tryptic PTM peptides in histones (related to Figure 6C). All the identified peptides with Mascot score above 20 were manually verified according to the rules described previously (Chen et al., 2005). The relative ratios of abundance of histone peptides in wild-type and TH2B-less histone samples were calculated as described in the Experimental Procedures. The PTM with the highest variations are indicated in red.

5 Montellier et al., Table S1 (related to Fig 4 A and B)

Molecular ChIP Ha (Spc) Input (Spc) ChIP Ha (R-Spt) Input (R-Spt) Histone Prot_acc Proteion description Weight (Da) Matched Matched Matched Matched Score* #** Score # Score # Score #

H1 IPI00228616 Tax_Id=10090 Gene_Symbol=Hist1h1a Histone H1.1 21772 5728 532 N/A*** N/A 3814 275 N/A N/A

IPI00223713 Tax_Id=10090 Gene_Symbol=Hist1h1c Histone H1.2 21254 3378 425 362 24 2248 213 N/A N/A

IPI00331597 Tax_Id=10090 Gene_Symbol=Hist1h1d Histone H1.3 22086 2852 397 362 23 2065 210 N/A N/A

IPI00223714 Tax_Id=10090 Gene_Symbol=Hist1h1e Histone H1.4 21964 2235 325 N/A N/A 1877 169 N/A N/A

IPI00319556 Tax_Id=10090 Gene_Symbol=Hist1h1t histone H1t 21656 1582 212 N/A N/A 1407 156 N/A N/A

IPI00467914 Tax_Id=10090 Gene_Symbol=H1f0 Histone H1.0 20848 640 86 N/A N/A 379 19 N/A N/A

IPI00230133 Tax_Id=10090 Gene_Symbol=Hist1h1b Histone H1.5 22562 493 87 N/A N/A 468 65 N/A N/A H2A IPI00555055 Tax_Id=10090 Gene_Symbol=H2afv 14 kDa protein 13600 2135 672 1395 337 2309 526 N/A N/A

IPI00331734 Tax_Id=10090 Gene_Symbol=H2afz Histone H2A.Z 13545 2185 680 1417 339 2341 537 N/A N/A

IPI00623951 Tax_Id=10090 Gene_Symbol=Hist2h2ac;Hist2h2ab Histone H2A type 2-B 14005 4296 566 1429 244 2850 536 1264 337

IPI00230264 Tax_Id=10090 Gene_Symbol=H2afx Histone H2A.x 15133 4317 810 1521 347 2901 604 1341 363

IPI00221463 Tax_Id=10090 Gene_Symbol=Hist3h2a Histone H2A type 3 14113 7527 912 N/A N/A 5825 708 N/A N/A

IPI00229542 Tax_Id=10090 Gene_Symbol=Hist1h2aa Histone H2A 14048 7532 908 3752 553 5825 708 3074 555

IPI00330000 Tax_Id=10090 Gene_Symbol=Hist2h2aa2;Hist2h2aa1 Histone H2A type 2-A 14087 7650 927 N/A N/A 6019 745 N/A N/A

IPI00272033 Tax_Id=10090 Gene_Symbol=Hist2h2ac;Hist2h2ab Histone H2A type 2-C 13980 7650 926 3778 561 N/A N/A 3208 575 330 IPI00378480 Tax_Id=10090 Gene_Symbol=H2afy Isoform 2 of Core histone macro-H2A.1 39710 8832 480 N/A N/A 7614 N/A N/A H2B IPI00111957 Tax_Id=10090 Gene_Symbol=Hist1h2ba Histone H2B type 1-A 14228 10733 699 1882 486 10961 576 1306 300 Tax_Id=10090 Gene_Symbol=Hist1h2bf;LOC100046213;Hist1h2bj;

IPI00114642 Hist1h2bl;Hist1h2bn Histone H2B type 1-F/ 13928 2628 520 N/A N/A 1749 377 720 217 Tax_Id=10090 Gene_Symbol=Gm9998;Hist1h2be;Hist1h2bg;

IPI00874654 Hist1h2bc Histone H2B 14787 N/A N/A 1952 441 1475 376 N/A N/A

IPI00282269 Tax_Id=10090 Gene_Symbol=Hist1h2bm Histone H2B type 1-M 13928 2576 615 1954 441 1794 384 705 219

IPI00227930 Tax_Id=10090 Gene_Symbol=Hist1h2bh Histone H2B type 1-H 13912 2648 511 1958 321 1736 340 684 213

IPI00554853 Tax_Id=10090 Gene_Symbol=Hist1h2bb Histone H2B type 1-B 13944 2584 510 1949 422 1725 368 660 216

IPI00648991 Tax_Id=10090 Gene_Symbol=Hist1h2bp Isoform 1 of Histone H2B type 1-P 13984 2570 514 N/A N/A 1716 377 693 215

IPI00459318 Tax_Id=10090 Gene_Symbol=Hist1h2bp Putative uncharacterized protein 16297 902 678 843 219 853 157 N/A N/A H3 IPI00785343 Tax_Id=10090 Gene_Symbol=H3f3b;H3f3a;LOC100045490 Histone H3.3 15319 529 849 585 502 1080 625 N/A N/A

IPI00277753 Tax_Id=10090 Gene_Symbol=Gm12260 similar to histone H3 15419 1077 1100 1015 591 1362 622 794 284 Tax_Id=10090 Gene_Symbol=Hist1h3d;Hist2h3c2;Hist2h3b;Hist1h3f;

IPI00230730 Hist1h3c;Hist1h3e;Hist1h3b;Hist2h3c1 15379 784 929 937 582 1375 676 810 289 H4 Tax_Id=10090 Gene_Symbol=Hist1h4b;Hist4h4;Hist1h4k;Hist1h4h; IPI00407339 Hist2h4;Hist1h4c;Hist1h4a;Hist1h4i;Hist 11360 13730 751 6283 395 6053 378 2042 162

6 Montellier et al., Table S2 (related to Fig 4 A and B)

Histone Prot_acc Proteion description Specific peptide sequence* Peptide score**

H1 IPI00228616 Tax_Id=10090 Gene_Symbol=Hist1h1a Histone H1.1 SETAPVAQAASTATEKPAAAK 109.57

IPI00223713 Tax_Id=10090 Gene_Symbol=Hist1h1c Histone H1.2 KPAAAAVTKK 77.7

IPI00331597 Tax_Id=10090 Gene_Symbol=Hist1h1d Histone H1.3 SETAPAAPAAPAPVEK 71.93

IPI00223714 Tax_Id=10090 Gene_Symbol=Hist1h1e Histone H1.4 KAAGTATAK 20.96

IPI00319556 Tax_Id=10090 Gene_Symbol=Hist1h1t histone H1t GAKGVQQR 55.56

IPI00467914 Tax_Id=10090 Gene_Symbol=H1f0 Histone H1.0 TENSTSAPAAKPK 52.03

IPI00230133 Tax_Id=10090 Gene_Symbol=Hist1h1b Histone H1.5 KATGPPVSELITK 71.56

7 Montellier et al., Table S3 (related to Fig 6B)

Accession Number of Theoretical Observed Error Histones PTMsa ∆m number Cys/Metb masses (Da) masses (Da) (ppm) N-terminal acetylation H4c P62806 0/1 11,331.37 11,331.39 0.02 1.77 + di-methylation H2B type 1- Q6ZWY9 None 0/2 13,830.50 13,830.54 0.0 2.89 C/E/G H2B type 1-H Q64478 None 0/2

H2B type 1-K Q8CGP1 None 0/2 13,844.51 13,844.54 0.03 2.17

H2B type 2-B Q64525 None 0/2 H2B type 1- P10853 None 0/2 F/J/L 13,860.51 13,860.44 -0.07 -5.05 H2B type 1-M P10854 None 0/2

H2B type 1-B Q64475 None 0/2 13,876.50 13,876.60 0.10 7.21

H2B type 1-P Q8CGP2 None 0/2 13,916.57 13,916.50 -0.07 -5.03 tH2B (H2B type P70696 None 1/2 14,208.68 14,208.75 0.07 4.93 1-A)

a Acetylation and di-methylation lead to mass shifts of +42.01 Da and +28.03 Da, respectively. a Oxidized cysteine and methionine residues lead to mass shifts of +47.99 Da and +31.99 Da, respectively. c Most abundant form.

8 Montellier et al., Table S4 (related to Fig 6C)

Histone No variant Modifications Position Sequence Score TH2B -less/ Th2b +/+

1 H4 K5ac 4 – 17 Pr- 91 0.60 GKacGGKacGLGKacGGAKac 2 H4 K8ac 4 – 17 RPr - 40 0.71 GKprGGKacGLGKacGGAKac 3 H4 K12ac 9 – 17 RPr -GLGKacGGAKacR 40 0.94 4 H4 K16ac 9 – 17 Pr-GLGKacGGAKacR 40 0.94 5 H4 R35me 24 – 35 Pr-DNIQGITKprPAIRme 71 3.33 6 H4 R55me 46 – 55 Pr-ISGLIYEETRme 57 2.06 7 H4 K59cr 56 - 67 Pr-GVLKcrVFLENVIR 73 0.87 8 H4 R67me 60 – 67 Pr-VFLENVIRme 49 2.85 9 H4 K77cr 68 – 78 Pr-DAVTYTEHAKcrR 41 2.64 10 H4 K77me2 68 – 78 Pr-DAVTYTEHAKme2R 57 0.94

11 H4 K79ac 79 – 92 Pr-KacTVTAMDVVYALKprR 45 0.89 12 H3 K14ac 9 – 17 Pr-KprSTGGKacAPR 53 1.01 13 H3 K18ac 18 – 26 Pr-KacQLATKacAAR 46 1.16 14 H3 K23ac 18 – 26 Pr-KprQLATKacAAR 74 1.14 15 H3 K23cr 18 – 26 Pr-KprQLATKcrAAR 48 1.37

16 H3 K27me2 27 – 36 Pr-Kme2SAPSTGGVKpr 46 0.86 17 H3 K27ac 27 – 36 Pr-KacSAPSTGGVKpr 46 1.14 18 H3 K122cr 117 - 128 Pr-VTIMPKcrDIQLAR 49 1.94 19 H2B K46ac 44 – 57 Pr-VLKacQVHPDTGISSKpr 59 1.09 20 H2B R72me 58 – 72 Pr-AMGIMNSFVNDIFERme 72 4.16

21 H2B K108ac 100 - 116 Pr- 58 0.78 LLLPGELAKacHAVSEGTKpr 22 H2B K108cr 100 - 116 Pr- 73 0.27 LLLPGELAKcrHAVSEGTKpr 23 H2B K116cr 109 - 120 Pr-HAVSEGTKcrAVTKpr 46 1.12 24 H2B K116ac 109 - 120 Pr-HAVSEGTKacAVTKpr 79 0.75 25 H2A R96me 90 – 96 Pr-HLQLAIRme 38 1.35 26 H2A K133cr 127 - 135 Pr-KprTESHHKcrAKpr 60 0.93

9 EXTENDED EXPERIMENTAL PROCEDURES

Proteomic-based approaches

Materials for proteomics

Water and acetonitrile were from Fisher Scientific (Pittsburgh, PA). Trifluoroacetic acid (TFA)

12 and light propionic anhydride ( C6) were from Sigma-Aldrich (St. Louis, MO). Sequencing- grade trypsin was from Promega (Madison, WI). C18 ZipTips were from Millipore (Bedford,

13 MA). Heavy propionic anhydride ( C6) was from Cambridge Isotope Laboratories, Inc. (Andover, MA).

In-gel digestion of core histone proteins Gel bands of histones were excised and subjected to in-gel digestion as described previously (Chen et al., 2005).

In-solution proteolytic digestion and chemical derivatization of histone proteins In-solution tryptic digestion of histone samples and in vitro lysine propionylation of tryptic peptides were performed as previously described (Garcia et al., 2007; Tan et al., 2011). Tryptic peptides from histones extracted from TH2B-less and wild type testes were labeled by in vitro

12 13 reaction with light ( C6) and heavy propionic anhydride ( C6), respectively. Equal amounts of light and heavy labeled peptides were mixed (normalized based on the protein amount), and were then resolved into 12 fractions using an Agilent 3100 OFFGEL Fractionator (Agilent, Santa Clara, CA). The peptides in each fraction were desalted using a µ-C18 Ziptip before HPLC/MS/MS analysis.

Nano-HPLC/mass spectrometric analysis The peptide fraction of interest was dissolved in 10 µL of HPLC buffer A (formic acid 0.1% (v/v) in water), and 2 µL was injected into a Nano-HPLC system (Eksigent Technologies, Dublin, CA). The peptides were separated on a homemade capillary HPLC column (100-mm length×75-µm inner diameter) containing Jupiter C12 resin (4-µm particle size, 90-Å pore diameter, Phenomenex, St. Torrance, CA) with a 120 min HPLC-gradient from 5 to 90% HPLC buffer B (formic acid 0.1% in acetonitrile) at a flow rate of 200 nL/min. The HPLC elute was electrosprayed directly into an LTQ-Orbitrap Velos mass spectrometer (Thermo Fisher Scientific, Waltham, MA) using a nanospray source. The LTQ-orbitrap Velos mass spectrometer was operated in a data-dependent mode with resolution R = 60,000 at m/z 400.

10 Full scan MS spectra from m/z 350 – 2000 were acquired in the Orbitrap. The twenty most intense ions were sequentially isolated in the linear ion trap and subjected to collision-activated dissociation (CAD) with a normalized energy of 35% and an isolation width of 4 Da. The exclusion duration for the data-dependent scan was 36 sec, the repeat count was 2, and the exclusion window was set at ±2 Da. AGC settings were 1E6 for full scan Orbitrap analysis, 1E4 for MSn scan in the ion trap.

Protein sequencing alignment Peak lists were first generated by the extract_msn.exe software (v5.0, Thermo Scientific). All MS/MS spectra were searched against the NCBI IPI_mouse_v3_74 protein sequence database using the Mascot search engine (version 2.1.0, Matrix Science, London). Trypsin was specified as digesting enzyme. A maximum of 5 missing cleavages were allowed. Mass tolerances for precursor ions were set at ±10 ppm for precursor ions and ±0.5 Da for MS/MS. For in-gel digested samples, we included lysine acetylation, lysine mono and di-methylation, lysine crotonylation, and arginine methylation as variable modifications. For in-solution samples, we

13 also included lysine propionylation, lysine propionylation ( C3), N-terminal propionylation,

13 and N-terminal propionylation ( C3) as additional variable modifications. All the identified peptides with Mascot score above 20 and E-value below 0.05 were manually verified according to the rules described previously (Chen et al., 2005). Quantification of PTMs was based on the precursor ion intensities of peptides labeled by light and heavy stable isotopes. Briefly, for every PTM site, we first validated all the identified peptides containing the targeted PTM type and site. Then the highest intensity of the peptide in the mass spectra was selected manually. The intensity ratio of the light and heavy isotope peaks was used for quantifying the relative abundance of the PTM between the two samples.

Profiling of core histones The profiling of core histone variants was performed by ultra-high performance liquid chromatography coupled to mass spectrometry (UHPLC-MS) after mild performic oxidation of acid-extracted histones, as previously described (Contrepois et al., 2010). Briefly, oxidized histones were loaded and separated on a C18 Hypersil GOLD column (2.1 mm x 150 mm, 175 Å, 1.9 µm, Thermo) at a flow rate of 300 µL/min with a linear gradient of 0 to 80% B in 13.5 min (with solvent A: H2O containing formic acid 0.1% and solvent B: ACN containing formic acid 0.1%). MS acquisition was performed on an LTQ-Orbitrap Discovery mass spectrometer (Thermo, San Jose, CA) operating in the positive ion mode (acquisition from m/z 500 to 2000)

11 using a resolution set at 30,000 (at m/z 400). The resulting mass spectra were deconvoluted using the Xtract software included in the Xcalibur package (Thermo).

Chromatin preparation and ChIP Germ cell nuclei were prepared by incubation with lysis buffer (KCl 60mM, NaCl 15mM, Tris HCl pH 7.4 15mM, Saccharose 0.34M, EDTA 2mM, EGTA 0.5mM, Spermidine 0.65mM, DTT 1mM, Triton 0.03%, Glycerol 1%, Complete Protease Inhibitor Cocktail Tablets Roche) followed by centrifugation during 10 min at 1500 rpm 4°C. Nuclei were resuspended in wash buffer (KCl 60 mM, NaCl 15 mM, Tris HCl pH 7.4 15 mM, saccharose 0.34M, spermidine 0,65mM, DTT 1mM, PMSF 0,5mM, Complete Protease Inhibitor Cocktail Tablets Roche) and centrifuged again. Nuclei were incubated in MNAse buffer (Tris HCl pH7.5 10mM, KCl 10mM, CaCl2 2mM) with micrococcal nuclease (S7 nuclease, Roche) at 37°C for the desired time, to obtain mononucleosomes. The digestion reaction was stopped by adding EDTA at 5 mM final concentration. The nucleosome fraction was isolated by centrifugation at 2500 rpm 10min 4°C. ChIP were carried in LSDB 250 (glycerol 20%, Hepes 50mM, MgCl2 3mM, KCl 250mM, Complete Protease Inhibitor Cocktail Tablets Roche) overnight at 4°C. For each ChIP about 100µg of mononucleosomes and 5µg of antibodies (rabbit anti-TH2B antibody, Abcam ab23913 or rabbit polyclonal anti-Ha, Abcam, ab9110) previously coupled to magnetic beads (Dynabeads protein G, Invitrogen) for 4 hours in PBS 1X, BSA 5µg/µl, were used. Beads were washed five times in LSDB 250 before nucleosome elution with elution buffer (TE SDS 1%) and DNA purification. The immunoprecipitated DNA was analyzed as described in (Gaucher et al., 2012).

Spermatogenic cells preparation and immunofluorescence Tubules sections were used for all immunofluorescence staining except for the analysis of transition protein 2 (TP2) and protamine 1 (Prm1) staining in condensed spermatids where testis imprints were used for better antibody accessibility. Protein visualization by immunofluorescence was carried out as previously described in (Govin et al., 2007). The antibodies used were as follows: rabbit polyclonal anti-TH2B 1/1000 (Abcam, ab23913), rabbit polyclonal anti-Ha 1/1000 (Abcam, ab9110), mouse monoclonal anti-Sycp3 antibody 1/500 (Abcam, ab97672), rabbit monoclonal anti-H2A.X 1/500 (Abcam 81299), mouse monoclonal anti-H3.3 1/200 (Abnova clone 2D7-H1), goat polyclonal anti-TP2 1/100 (Santa Cruz) and mouse monoclonal anti-Prm1 antibodies 1/100 (Briar). The secondary antibodies were Alexa 488 or Alexa 546 fluor conjugates at 1/500 (Molecular Probes, Invitrogen). Elongated

12 spermatids isolated after sedimentation on a BSA gradient were used for morphological analysis by electronic microscopy, as described in (Govin et al., 2007).

Step 12-16 condensed spermatids isolation and chromatin preparation Step 12–16 spermatids from 4 testes were prepared and lysed in 75µl of lysis buffer (Tris pH 7.4 50 mM, NaCl 300 mM, NP-40 0.1%, DOC 0.1%, DTT 1 mM, Complete Protease Inhibitor Cocktail Tablets EDTA-free Roche) during 15min at 4°C. Spermatids were pelleted at 20000g, 4°C, 10min. The supernatant was kept and the pellet was re-suspended in 75µl of lysis buffer followed by sonication at 80 J to allow the suspension of large chromatin fragments. Unlysed spermatids were pelleted at 20000g, 4°C, 10min, and the supernatant was pooled with the first supernatant. 75µl of MNase buffer (Tris, pH 7.5 10 mM, KCl 10 mM, and CaCl2 1 mM) were added and MNase digestion of chromatin was performed using 7.5U of micrococcal nuclease S7 (Roche). Digestion was performed at 37°C for the indicated times, and EDTA 5mM final was used to stopped the digestion. The MNase released fragments were purified by phenol chloroform extractions and separated on a agarose gel 2%.

13 SUPPLEMENTARY REFERENCE

Chen, Y., Kwon, S.W., Kim, S.C., and Zhao, Y. (2005). Integrated approach for manual evaluation of peptides identified by searching protein sequence databases with tandem mass spectra. J Proteome Res 4, 998-1005. Contrepois, K., Ezan, E., Mann, C., and Fenaille, F. (2010). Ultra-high performance liquid chromatography-mass spectrometry for the fast profiling of histone post-translational modifications. J Proteome Res 9, 5501-5509. Garcia, B.A., Mollah, S., Ueberheide, B.M., Busby, S.A., Muratore, T.L., Shabanowitz, J., and Hunt, D.F. (2007). Chemical derivatization of histones for facilitated analysis by mass spectrometry. Nat Protoc 2, 933-938. Gaucher, J., Boussouar, F., Montellier, E., Curtet, S., Buchou, T., Bertrand, S., Hery, P., Jounier, S., Depaux, A., Vitte, A.L., et al. (2012). Bromodomain-dependent stage-specific male genome programming by Brdt. EMBO J 31, 3809-3820. Govin, J., Escoffier, E., Rousseaux, S., Kuhn, L., Ferro, M., Thevenon, J., Catena, R., Davidson, I., Garin, J., Khochbin, S., et al. (2007). Pericentric heterochromatin reprogramming by new histone variants during mouse spermiogenesis. J Cell Biol 176, 283-294. Huh, N.E., Hwang, I.W., Lim, K., You, K.H., and Chae, C.B. (1991). Presence of a bi- directional S phase-specific transcription regulatory element in the promoter shared by testis- specific TH2A and TH2B histone genes. Nucleic Acids Res 19, 93-98. Tan, M., Luo, H., Lee, S., Jin, F., Yang, J.S., Montellier, E., Buchou, T., Cheng, Z., Rousseaux, S., Rajagopal, N., et al. (2011). Identification of 67 histone marks and histone lysine crotonylation as a new type of histone modification. Cell 146, 1016-1028.

14