A Genetically Encoded Probe for Live-Cell Imaging of H4K20 Monomethylation

    A genetically encoded probe for live-cell imaging of H4K20 monomethylation Yuko Sato, Tomoya Kujirai, Ritsuko Arai, Haruhiko Asakawa, Chizuru Ohtsuki, Naoki Horikoshi, Kazuo Yamagata, Jun Ueda, Takahiro Nagase, Tokuko Haraguchi, Yasushi Hiraoka, Akatsuki Kimura, Hitoshi Kurumizaka, Hiroshi Kimura PII: DOI: Reference:

S0022-2836(16)30305-9 doi: 10.1016/j.jmb.2016.08.010 YJMBI 65172

To appear in:

Journal of Molecular Biology

Received date: Revised date: Accepted date:

13 April 2016 4 August 2016 5 August 2016

Please cite this article as: Sato, Y., Kujirai, T., Arai, R., Asakawa, H., Ohtsuki, C., Horikoshi, N., Yamagata, K., Ueda, J., Nagase, T., Haraguchi, T., Hiraoka, Y., Kimura, A., Kurumizaka, H. & Kimura, H., A genetically encoded probe for livecell imaging of H4K20 monomethylation, Journal of Molecular Biology (2016), doi: 10.1016/j.jmb.2016.08.010

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

A genetically encoded probe for live-cell imaging of H4K20

RI

PT

monomethylation

Yuko Sato1,*, Tomoya Kujirai2, Ritsuko Arai3,#, Haruhiko Asakawa4, Chizuru Ohtsuki4,

SC

Naoki Horikoshi2, Kazuo Yamagata5, Jun Ueda6, Takahiro Nagase7, Tokuko Haraguchi4,8, Yasushi Hiraoka4,8, Akatsuki Kimura3, Hitoshi Kurumizaka2, and Hiroshi

NU

Kimura1,* 1

#

AC CE P

TE

D

MA

Cell Biology Unit, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, 226-8501, Japan; 2Laboratory of Structural Biology, Graduate School of Advanced Science and Engineering, Waseda University, Shinjuku-ku, Tokyo 162-8480, Japan; 3Cell Architecture Laboratory, Structural Biology Center, National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan; 4Graduate School of Frontier Biosciences, Osaka University, Suita, 565-0871, Japan; 5Faculty of Biology-Oriented Science and Technology, Kindai University, Kinokawa City, Wakayama 649-6493, Japan; 6Center for Education in Laboratory Animal Research, Chubu University, Kasugai, Aichi 487-8501, Japan; 7Public Relations Team, Kazusa DNA Research Institute, Chiba, 292-0818, Japan; 8Advanced ICT Research Institute, National Institute of Information and Communications Technology (NICT), Kobe 651-2492, Japan Present address: Department of Anatomy and Histology, Fukushima Medical University, School of Medicine, Hikarigaoka, Fukushima 960-1295, Japan

*Correspondence:

Yuko Sato; [email protected] Hiroshi Kimura; [email protected] Phone: +81-45-924-5742 1

ACCEPTED MANUSCRIPT

Abstract

PT

Eukaryotic gene expression is regulated in the context of chromatin. Dynamic changes

RI

in post-translational histone modification are thought to play key roles in fundamental

SC

cellular functions such as regulation of the cell cycle, development, and differentiation.

NU

To elucidate the relationship between histone modifications and cellular functions, it is

MA

important to monitor the dynamics of modifications in single living cells. A genetically encoded probe called mintbody (modification-specific intracellular antibody), which is a

D

single-chain variable fragment tagged with a fluorescent protein, has been proposed as a

TE

useful visualization tool. However, the efficacy of intracellular expression of antibody

AC CE P

fragments has been limited, in part due to different environmental conditions in the cytoplasm compared to the endoplasmic reticulum where secreted proteins such as antibodies are folded. In this study, we have developed a new mintbody specific for histone H4 Lys20 monomethylation (H4K20me1). The specificity of the H4K20me1-mintbody in living cells was verified using yeast mutants and mammalian cells in which this target modification was diminished. Expression of the H4K20me1-mintbody allowed us to monitor the oscillation of H4K20me1 levels during the cell cycle. Moreover, dosage-compensated X chromosomes were visualized using the H4K20me1-mintbody in mouse and nematode cells. Using X-ray crystallography

2

ACCEPTED MANUSCRIPT

and mutational analyses, we identified critical amino acids that contributed to

PT

stabilization and/or proper folding of the mintbody. Taken together, these data provide

RI

important implications for future studies aimed at developing functional intracellular

SC

antibodies. Specifically, the H4K20me1-mintbody provides a powerful tool to track this

MA

NU

particular histone modification in living cells and organisms.

TE

AC CE P

live-cell imaging

D

Key words: histone modification; inactive X chromosome; intracellular antibody;

Abbreviations used: H4K20me1, histone H4 Lys20 monomethylation; mintbody, modification-specific intracellular antibody; scFv, single-chain variable fragment; PCNA, proliferating cell nuclear antigen; FR, framework region; CDR, complementarity determining region; HuCal, human combinatorial antibody libraries; Xi, inactive X chromosome

3

ACCEPTED MANUSCRIPT

Introduction

PT

Histone proteins are subjected to a variety of post-translational modifications

RI

that are thought to play crucial roles in gene regulation and genome integrity [1, 2].

SC

However, the precise role of specific histone modifications in transcription, replication,

NU

repair, and chromosome segregation remains to be elucidated. In addition, it is unclear

MA

how histone modifications are inherited or altered over time. Addressing the dynamic regulation of histone modifications has been limited because most studies conducted

D

thus far have utilized fixed cells, thus preventing single-cell assessment of modification

TE

changes. To overcome the limitation of using fixed cells, we have developed two

AC CE P

live-cell imaging systems based on the use of modification-specific antibodies [3]. These approaches include Fab-based live endogenous modification labeling (FabLEM [4]) and a genetically encoded system using modification specific intracellular antibody

probes (mintbodies [5-7]). The latter approach is particularly useful for investigations aimed at monitoring histone modifications in living animals. Histone H4 Lys 20 monomethylation (H4K20me1) is a key histone modification that regulates DNA replication [8-10] and repair [11] as well as chromosome compaction [12]. The level of H4K20me1 is regulated by three methyltransferases, PR-Set7 [13-15], Suv420H1, and Suv420H2 [16], and a

4

ACCEPTED MANUSCRIPT

demethylase, PHF8 [17, 18]. Chromatin immunoprecipitation combined with deep

PT

sequencing (ChIP-seq) and immunofluorescence analyses have shown that H4K20me1

RI

is generally distributed in euchromatin [16]. In addition, H4K20me1 enrichment in

SC

inactivated X chromosomes (Xi) and centromeres have also been reported [16, 19-21].

NU

However, the spatiotemporal regulation of H4K20me1 and its impact on DNA

MA

metabolism remain largely uncharacterized.

In this study, we report the generation of H4K20me1-specific mintbody. X-ray

D

crystallography and mutation analyses identified key amino acids that improved

TE

mintbody stability and/or folding. We demonstrate that the H4K20me1-mintbody

AC CE P

specifically recognizes the target histone modification and is useful to monitor the oscillations of H4K20me1 during the cell cycle and to visualize dosage-compensated X chromosomes in living mammalian cells and during Caenorhabditis elegans development.

5

ACCEPTED MANUSCRIPT

PT

Results

RI

Generation and characterization of the H4K20me1-specific mintbody

SC

To generate a new mintbody specific for histone H4 Lys 20 monomethylation

NU

(H4K20me1), we cloned cDNA encoding variable regions of heavy and light chains

MA

from 15F11/CMA421 hybridoma cell line [22], and constructed mintbody-expression vectors by fusing the single-chain variable fragment (scFv) to EGFP at the C-terminus

D

(Fig. 1a). When expressed in HeLa cells, H4K20me1-mintbody was diffuse in the

TE

cytoplasm but concentrated in nuclei and chromosomes, where it co-localized with a

AC CE P

DNA staining Hoechst33342 dye (Fig. 1b). This suggests that the probe binds to nuclear targets in living cells, as designed. To determine if the H4K20me1-mintbody specifically recognizes H4K20me1,

like its parental IgG [22], we performed both in vitro and in vivo analyses. We assessed binding in vitro using recombinant histone H4 proteins that harbor methylation-mimic modifications [22] as well as synthetic peptides arrays. When these were incubated with bacterially expressed and purified H4K20me1-scFv (Supplementary Fig. S1a), only H4 harboring monomethylation-mimic modifications and synthetic H4K20me1 peptides

6

ACCEPTED MANUSCRIPT

were bound by the scFv (Supplementary Fig. S1b-d). These results suggest that the

PT

specificity of the original IgG was well retained in the scFv format.

RI

We next evaluated the specificity of H4K20me1-mintbody in living cells using

SC

fission yeast Schizosaccharomyces pombe, which can be genetically manipulated to

NU

delete the target modification. Targeting of the H4K20me1-mintbody to the

MA

crescent-shaped chromatin compartment in the nucleus can also be assessed readily in S. pombe (Supplementary Fig. S2a) [23, 24]. We generated yeast strains expressing

D

H4K20me1-mintbody in the wild type and set9∆ background. As Set9 is the sole

TE

H4K20 methyltransferase in S. pombe [25], no H4K20me1 was expected in set9∆

AC CE P

strains. Indeed, immunofluorescence and immunoblotting using the specific antibody showed no H4K20me1 in set9∆ strains (Fig. 1c and Supplementary Fig. S2b). Note that similar results were obtained using H4K20me1-scFv, instead of the specific IgG, which further validates that the scFv retained the original specificity (Supplementary Fig. S2c,d). To assess H4K20me1-mintbody localization in the wild type and set9∆ background strains, growing yeast cells were incubated with Hoechst33342 for DNA labeling [26]. In the wild type background, the H4K20me1-mintbody was concentrated in the chromatin domain, as represented by Hoechst staining, and excluded from the

7

ACCEPTED MANUSCRIPT

nucleolar domain in the nucleus (Fig. 1d,e and Supplementary Fig. S2a). In contrast, in

PT

the set9∆ strain, the H4K20me1-mintbody diffused throughout the cell and the

RI

concentration in the chromatin domain was substantially reduced. Although

SC

H4K20me1-mintbody was still slightly enriched in the nucleus of the set9∆ strain, GFP

NU

alone also exhibited the same slight enrichment, probably due to their free diffusion

MA

throughout the cell.

To further assess the specificity of the mintbody, we generated a strain that

D

harbors a single H4 gene by deleting two out of three H4 genes in S. pombe (H4.2 wt),

TE

and another strain that harbors a single mutant H4 gene in which lysine 20 is substituted

AC CE P

with alanine (H4.2 K20A). As expected, H4K20me1-mintbody was enriched in the chromatin domain of the H4.2 wt, but not in the H4.2 K20A strain (Fig. 2d,e). These results indicate that H4K20 methylation is required for H4K20me1-mintbody’s chromatin targeting.

To examine if H4K20me1-mintbody specifically recognizes H4K20me1 over H4K20me2 and H4K20me3, we used a mammalian cell line harboring multiple H4K20 methyl transferases with selective activities (Fig. 2a); PR-Set7/SET8/SETD8/KMT5A is known to monomethylate unmodified H4K20, and SUV420H1/KMT5B and SUV420H2/KMT5C mediate further methylation (to me2 and me3) of H4K20me1 [16].

8

ACCEPTED MANUSCRIPT

A demethylase PHF8 is also known to demethylate H4K20me1 [17, 18]. We anticipated

PT

that overexpression of these enzymes would alter the balance among different

RI

methylation states in living cells. To screen the effect of overexpression on H4K20

SC

methylation, HeLa cells were transfected with HaloTag-fusion enzymes and then fixed

NU

and stained with the specific antibodies directed against H4K20me1, H4K20me2, and

MA

H4K20me3. HaloTag-fusion enzymes were labeled with tetramethylrhodamin-conjugated HaloTag ligand (TMR Halo ligand), and DNA was

D

counterstained with Hoechst33342. After collecting fluorescence images, the intensities

TE

of antibody and TMR Halo ligand in single cell nuclei were measured and plotted.

AC CE P

Among the four methylation and one demethylation enzymes we examined, Halo-SUV420H1 affected H4K20me1 levels the most drastically (Fig. 2b-d and Supplementary Fig. S3b). In cells expressing Halo-SUV420H1, H4K20me1 levels were substantially decreased, and reciprocally H4K20me3 was increased, which is consistent with the action of SUV420H1 [16]; H4K20me2 remained unchanged except in the highest expressing cells (Supplementary Fig. S3b). We then overexpressed Halo-SUV420H1 in HeLa cells stably expressing H4K20me1-mintbody to evaluate if the probe can monitor the decrease in living cells. We had previously shown that changes in the global acetylation level can be monitored by the nuclear enrichment of

9

ACCEPTED MANUSCRIPT

H3K9ac-mintbody [5]; H3K9ac-mintbody becomes more concentrated in the nucleus

PT

when acetylation levels increase. If H4K20me1-mintbody selectively binds to

RI

H4K20me1, changes in H4K20me1 level should also affect the nuclear concentration of

SC

H4K20me1-mintbody.

NU

A day after transfection of the Halo-SUV420H1 expression vector, cells were

MA

incubated with Halo TMR ligand and Hoechst33342, and fluorescence images were collected for quantifying the nuclear and cytoplasmic intensities of

D

H4K20me1-mintbody and the nuclear intensity of Halo-SUV420H1 (Fig. 2e). The

TE

nucleus to cytoplasm intensity ratios of H4K20me1-mintbody in SUV420H1-expressing

AC CE P

cells were indeed lower compared to untransfected cells, and more effects were seen in higher expressing cells (Fig. 2e, right), similar to the H4K20me1 antibody staining (Fig. 2c, left). This result indicates that H4K20me1-mintbody can be used in living cells to selectively monitor H4K20me1, as opposed to H4K20me2 or H4K20me3. To further evaluate the specific binding of H4K20me1-mintbody to H4K20me1 in HeLa cells, a simple co-localization analysis was employed by comparing the distribution of H4K20me1-mintbody with those of different modifications, including H4K20me1, H4K20me2, H4K20me3, H3K4me1, H3K9me1, and H3K27me1 (as detected with specific antibodies). After collecting confocal fluorescence images of

10

ACCEPTED MANUSCRIPT

H4K20me1 and each of the antibodies, we performed a co-localization analysis to

PT

obtain co-localization plots and correlation coefficients. As expected,

RI

H4K20me1-mintbody was most highly correlated with H4K20me1, even though

SC

H4K20me1-mintbody signals had higher background than the antibody stains, probably

NU

due to the presence of unbound mintbody (Supplementary Fig. S4). Taken together,

MA

these data support the view that H4K20me1-mintbody specifically binds to H4K20me1 in living cells and is useful as a probe to monitor the level and localization of

TE

D

H4K20me1.

AC CE P

Monitoring changes in H4K20me1 levels during the cell cycle To investigate whether the H4K20me1-mintbody could be used as a live cell

probe without disturbing cell function, we compared cell growth and cell cycle progression in HeLa cells that stably expressed H4K20me1-mintbody versus wild type cells. Both the growth curve and cell cycle profile—based on ethynyl-deoxyuridine (EdU) incorporation and DNA content—were very similar between the mintbody-expressing and wild type cells (Fig. 3a, b). These data suggest that H4K20me1-mintbody does not affect cell cycle progression.

11

ACCEPTED MANUSCRIPT

We then investigated whether H4K20me1-mintbody can monitor the

PT

oscillations of H4K20me1 levels during the cell cycle in living cells. A number of

RI

studies have reported that H4K20me1 increases during late G2 and M, and decreases

SC

after cell division and throughout G1 and S [9, 15, 27, 28]. We confirmed this

NU

oscillation in HeLa cells expressing mCherry-tagged proliferating cell nuclear antigen

MA

(PCNA), a marker of DNA replication factories [29], by time-lapse imaging followed by immunofluorescence with the H4K20me1-specific antibody (Fig. 3c and

D

Supplementary Fig. S5a). This trend remained the same after normalization using DNA

TE

density (Hoechst33342 intensity) (Fig. 3d). We then used HeLa cells that stably express

AC CE P

both H4K20me1-mintbody and mCherry-PCNA. After time-lapse imaging to identify the cell cycle position, the nuclear and cytoplasmic intensities of H4K20me1-mintbody at the final image were measured to obtain the nuclear/cytoplasm ratio. This ratio is expected to reflect H4K20me1 levels, as shown before. The nuclear/cytoplasm ratio increased in G2 and M compared to G1 and S (Fig. 3e and Supplementary Fig. S5b). A few examples of cells imaged for a long period (over 45 hr) are shown in Fig. 3f and Supplementary Fig. S6. These live cell data are qualitatively in good agreement with the previous studies that utilized Western blotting and mass spectrometric approaches [27, 28]. This result demonstrates that the H4K20me1-mintbody enables monitoring of

12

ACCEPTED MANUSCRIPT

changes in H4K20me1 levels in living cells. We noticed, however, that the

PT

nuclear/cytoplasm ratio during interphase should not be directly compared with the

RI

chromosome/cytoplasm ratio during M phase. Although the nuclear intensity of

SC

H4K20me1-mintbody peaked in M phase, it dropped to a minimum when normalized

NU

with Hoechst33342 intensity due to chromosome condensation (Supplementary Fig. S7

MA

and S8). In fact, the nuclear intensity includes both the bound and free H4K20me1-mintbody molecules, and free molecules disperse into the cytoplasm after

D

the nuclear membrane breaks down; a similar minimization in Fab-probe

TE

nuclear/cytoplasm ration during M phase was shown previously [30]. Therefore, a

AC CE P

careful interpretation is needed when cells undergo dynamic changes in shape.

Visualization of inactive X chromosomes using the H4K20me1-mintbody in living cells

H4K20me1 has been described as one of the epigenetic marks associated with the inactive X chromosome (Xi) in cultured cells, in addition to H3 trimethylated at lysine 27 (H3K27me3) [16, 19, 20]. To investigate whether Xi can be visualized with the H4K20me1-mintbody, we used mouse embryonal carcinoma cell line MC12 that harbors Xi fused to an autosome fragment [31]. Immunofluorescence using fixed cells

13

ACCEPTED MANUSCRIPT

revealed that when the H4K20me1-mintbody was expressed in MC12 cells, the probe

PT

was enriched on Xi together with H3K27me3 and H4K20me1 (detected by the specific

RI

antibodies) (Fig. 4a). Using metaphase chromosome spreads, we also observed that the

SC

H4K20me1-mintbody was enriched on the arm of Xi, but not on Hoechst-dense

NU

pericentromeric major satellite regions (Fig. 4b). Time-lapse imaging of

MA

H4K20me1-mintbody indeed allowed us to visualize in single cells the dynamic behavior of Xi during the cell cycle, including mitosis (Fig. 4c). Enrichment of the

D

H4K20me1-mintbody on Xi was also observed in hTERT-RPE1 cells, a human

TE

pseudo-diploid female cell line immortalized with telomerase (Supplementary Fig. S9).

AC CE P

In addition, accumulation of H4K20me1 on centromere was visualized using the H4K20me1-mintbody (Supplementary Fig. S9), as previously reported [21]. Making use of the detectability of Xi by the H4K20me1-mintbody, we

monitored the exact timing of Xi replication by time-lapse imaging using MC12 cells that stably express both the H4K20me1-mintbody and mCherry-PCNA (Fig. 4d). In early-S phase, numerous PCNA foci were observed in euchromatic regions, and were excluded from Xi. As cells progressed through mid-S phase, PCNA foci appeared preferentially around the nuclear periphery and were also enriched on Xi. A few hours later, we observed the appearance of larger PCNA foci and their separation from Xi.

14

ACCEPTED MANUSCRIPT

Replication of Xi often began at precisely the middle of S phase and was complete in

RI

PT

2-3 hours, allowing ~2 additional hours to complete S phase (Fig. 4e).

SC

Visualization of inactive X chromosomes during early development in C. elegans

NU

Finally, to demonstrate that the H4K20me1-mintbody can also be utilized in

MA

living multicellular organisms, we used C. elegans as a model. In XX hermaphrodite worms, X-linked gene expression is two-fold repressed by a dosage compensation

D

complex to equalize gene dosage with XO males [32]. H4K20me1 has been reported to

TE

be enriched in X chromosomes in hermaphrodites and to participate in the repression

AC CE P

[33]. We generated a transgenic nematode line that expressed the H4K20me1-mintbody. The worms developed normally and were fertile, indicating that the mintbody expression did not have inhibitory effects. We found that the H4K20me1-mintbody was present in nuclei with subtle concentration patterns in the 120-cell and Bean stages (Fig. 5; 1 and 2). After the Comma stage, however, two bright foci were clearly observed in each nucleus (Fig. 5; 3-6). The transition in H4K20me1 localization during development has been demonstrated in a previous study using an H4K20me1-specific antibody [33]. It is therefore likely that the change in mintbody localization is not due to differences in accessibility of the mintbody to the chromosomes, but instead reflects

15

ACCEPTED MANUSCRIPT

true localization of H4K20me1. This result demonstrates the utility of our new

RI

SC

dosage-compensated X chromosomes in C. elegans.

PT

H4K20me1 probe as a tool to visualize changes in chromosome structure and

NU

Structural and mutational analysis of the H4K20me1-specific mintbody

MA

Previous studies [34-36] have demonstrated that small differences in amino acid sequence can affect the folding and/or stability of functional scFv expressed in

D

living cells. In addition to the clone 15F11/CMA421 used to generate the

TE

H4K20me1-mintbody, we also obtained from the same immunized mouse two other

AC CE P

hybridoma clones, 12C8 and 22G3. The deduced amino acid sequences of VH and VL were highly similar among these clones, and the diversity was probably due to somatic hypermutation (Supplementary Fig. S10), as previously reported [5]. When expressed in HeLa cells in scFv-GFP form, these proteins failed to accumulate in nuclei (Fig. 6a and Supplementary Fig. S12a), suggesting that these two scFv-GFP proteins are less stable and/or more prone to aggregate than the H4K20me1-mintbody derived from 15F11/CMA421. To identify structure-function relationships that impact intracellular scFv expression, we performed mutational and structural analyses. H4K20me1-mintbody and

16

ACCEPTED MANUSCRIPT

12C8 scFv-GFP differed in only five amino acids, four of which were located in the

PT

framework regions (FRs; dark blue and pink in Fig. 6b) and the remaining one in a

RI

complementarity determining region (CDR; light blue in Fig. 6b). We then investigated

SC

the functional impact of these amino acids in living cells by introducing point mutations

NU

and assessing effects on chromatin binding, or nucleoplasmic enrichment. We replaced

MA

each of the five 12C8-specific amino acids with the corresponding residues present in the 15F11-type. When expressed in HeLa cells, only the 12C8 G99A mutant exhibited

D

nucleoplasmic concentration in the transfected cells (Fig. 6c), pointing to a key role for

TE

A99 in the function of 15F11. However, a reciprocal 15F11 mutant (A99G) was still in

AC CE P

the nucleoplasm even though it diffused more into the cytoplasm (Fig. 6d). We then generated double mutants that combined A99G and I95M mutations (I95M-A99G), and found that this 15F11 mutant was no longer concentrated in the nucleoplasm (Fig. 6d). Thus, both I95 and A99 are critical for the intracellular function of the 15F11-derived H4K20me1-mintbody. While the I95 and A99 residues are clearly important for the robust stability and/or folding of the H4K20me1-specific mintbody in living cells, they may differentially affect binding to H4K20me1. Therefore, we measured the binding affinities of different H4K20me1-specific antibodies in their IgG and scFv forms, using

17

ACCEPTED MANUSCRIPT

surface plasmon resonance (SPR). In their native IgG form, 15F11 and 12C8 showed

PT

similar profiles and affinities (52 and 33 pM, respectively, by two-state fitting) whereas

RI

22G3 was much weaker (15 nM) (Supplementary Fig. S11). To purify scFv versions of

SC

each, His6-tagged constructs were expressed in E. coli. Although a substantial fraction

NU

of 15F11-scFv was recovered in the cytosolic extract, most 12C8-scFv and 22G3-scFv

MA

remained in the insoluble pellet (Supplementary Fig. S12b), suggesting that 12C8 and 22G3 are more prone to aggregate than 15F11 in bacteria, as in mammalian cells. We

D

then purified scFvs from pellets by denaturation and renaturation. 15F11 and 12C8 were

TE

successfully purified (Supplementary Fig. S12c), although 22G3 aggregated during

AC CE P

renaturation. According to SPR, the affinities of 15F11 and 12C8 in scFv forms were again similar to each other (250 nM and 100 nM, respectively, by two state fitting), even though they were more than three orders of magnitude lower than those of the counterpart IgG forms. These results are consistent with the view that the difference between 15F11 and 12C8 is their stability and/or folding inside cells rather than their affinity to target modifications. To investigate the role of these amino acids in the structure of H4K20me1-mintbody, we determined the crystal structure of recombinant 15F11 scFv, which were purified from the soluble fraction of E. coli lysate (Supplementary Fig. S1a).

18

ACCEPTED MANUSCRIPT

The crystal structure of the scFv was solved at a resolution of 1.94 Å (Fig. 6e).

PT

Structural analysis revealed that both I95 and A99 are located inside the

RI

immunoglobulin fold. I95 is in hydrophobic contact with Y97, and is located near Q41,

SC

G116, T117, and T118 (Fig. 6e). Interestingly, the side chain of I95 is exposed to the

NU

solvent. In contrast, A99 is buried in the hydrophobic core composed of V39, F110, and

MA

W113 (Fig. 6e). When A99 is replaced with glycine, this hydrophobic interaction is likely loosened due to absence of the side-chain methyl-group (Supplementary Fig.

TE

AC CE P

H4K20me1-mintbody.

D

S13). These observations suggest that A99 play a key role in stabilizing the

Discussion

In this study, we developed an H4K20me1-mintbody, a new genetically encoded antibody-based probe specific for the detection of H4K20me1. From various in vitro and in vivo analyses, we concluded that the H4K20me1-scFv and -mintbody retain the original IgG’s specificity to H4K20me1. The H4K20me1-mintbody allowed us to monitor changes in the level of H4K20me1 during the cell cycle and localization of inactive X chromosomes. Furthermore, dosage-compensated X chromosomes were observed in living transgenic nematodes that express the H4K20me1-mintbody.

19

The structure of H4K20me1-specific variable fragments

PT

ACCEPTED MANUSCRIPT

RI

Cytoplasmic expression of antibody fragments has been challenging because

SC

the cytoplasmic environment differs from the endoplasmic reticulum, where antibodies

NU

are naturally folded and processed. This difference can cause low solubility and

MA

instability, often leading to aggregation [37]. In the present study, H4K20me1-mintbody (derived from 15F11) functioned properly in the cytoplasm, but 12C8-scFv-GFP and

D

22G3-scFv-GFP did not. The heavy-chain variable regions of the H4K20me1-specific

TE

antibodies were very similar to the human VH3-type shown to be thermodynamically

AC CE P

stable and less prone to aggregation by the systematic studies of human combinatorial antibody libraries (HuCal) [34, 38, 39]. Importantly, key amino acids present in the framework regions were conserved, including E8 (corresponding to amino acid H6 in the Kabat numbering system and H6 in the AHo numbering system [34, 40], written as “E8(H6/6)”), F31(H29/31), V39(H37/44), R69(H66/77), D92(H85/100), and T117(H107/143) (Supplementary Fig. S10). In addition, R100(H94/108) and D111(H101/137) at the edges of CDR3, which can make a salt bridge [34, 37], were conserved. These features were common to all three H4K20me1-specific antibody clones, suggesting that these proteins had high potential as functional intracellular

20

ACCEPTED MANUSCRIPT

antibodies [36]. However, conservation of framework regions was not sufficient, as only

PT

15F11 was functional in living cells. Crystallographic and mutational analyses revealed

RI

that A99(H93/97) is critical in the formation of a hydrophobic core with V39(H37/44),

SC

F110(H100e/119), and W113(H103/139). Replacement of G99(H93/97) with alanine in

NU

12C8 likely strengthened hydrophobic interactions and greatly improved functionality,

MA

as judged by the nucleoplasmic accumulation of the G99A mutant. A structure model of A99G reveals that hydrophobic contacts with the residue at position 99(H93/97) are

D

indeed greatly reduced (Supplementary Fig. S13). Alanine at the H93/97 position is

TE

conserved in human VH3, and the hydrophobic core involving this alanine may be

AC CE P

important for the general stability of the scFv structure. However, substitution of A99(H93/97) with glycine in 15F11 did not completely prevent nucleoplasmic accumulation, although the signal was relatively weak and diffused. Hence, in addition to A99(H93/97), 15F11 functionality was assisted by I95(H89/103), which associates with Q41(H39/45), Y97(H91/105), G116(H106/142), T117(H107/143), and T118(H108/144). Although Q41(H39/45), Y97(H91/105), G116(H106/142), and T117(H107/143) are conserved amino acids in human VH3, I95(H89/103) and T118(H108/144) are not. The I95(H89/103)-mediated intramolecular hydrophobic interaction may have a supporting function in the mintbody stability in living cells. It

21

ACCEPTED MANUSCRIPT

would be interesting to investigate how these variations affect the stability and folding

PT

of VH3 type scFv.

RI

An overall negative charge at cytoplasmic pH and low hydrophilicity has been

SC

proposed to improve the solubility of intracellular antibodies [41]. However, the net

NU

charge and hydropathicity (GRAVY score [41]) of H4K20me1-mintbodies are 0, -0.389,

MA

and -0.1 and -0.413, -3.1, and -0.386 for 15F11, 12C8, and 22G3, respectively. Thus, variations in charge and hydrophilicity do not account for differences in the solubility

D

we observed in the three H4K20me1-scFv-GFP. Instead, we found that the appropriate

TE

formation of hydrophobic cores involving A99(H93/97) plays a pivotal role in mintbody

AC CE P

folding and proper cellular function. The structure-function relationships hereby analyzed provide useful information on further development of intracellular antibodies.

Live-cell imaging using the H4K20me1-mintbody In this study, we demonstrated that Xi can be visualized in living cells using our newly developed H4K20me1-mintbody. This strategy, combined with mCherry-PCNA imaging, allowed precise determination of Xi replication timing at the single-cell level. In mouse MC12 cells, Xi replication took place in mid-S phase, a period during which facultative heterochromatin was replicated, in agreement with a

22

ACCEPTED MANUSCRIPT

previous study in mouse C2C12 cells [42]. It has recently been suggested that

PT

replication of Xi is rapid, random, and less organized than that of actively transcribed

RI

genome regions [43], and that this feature may account for increased mutation rates.

SC

Enrichment of H4K20me1 on Xi may represent involvement of this modification in

NU

DNA damage repair. Future studies using H4K20me1-mintbody combined with probes

MA

targeting DNA damage repair proteins and their modifications (e.g., gamma-H2A.X) may reveal the relationship between H4K20me1 and DNA damage repair on Xi during

D

DNA replication. Utilization of the H4K20me1-mintbody in differentiating cells is also

TE

anticipated to help visualize the process of Xi formation.

AC CE P

Two mintbodies have been developed to date, including the

H4K20me1-mintbody described here and a previously reported H3K9ac-speific mintbody [5, 6]. A recent in vivo study in transgenic frogs using the H3K9ac-mintbody has detected changes in acetylation during notochord regeneration [7], thus demonstrating the high potential of this genetically encoded system. Further development of additional mintbodies directed against diverse histone post-translational modifications will facilitate the identification of regulatory mechanisms that control epigenetic modifications in vivo.

23

ACCEPTED MANUSCRIPT

RI

PT

Materials and methods

SC

Plasmid construction

NU

To generate the H4K20me1-mintbody expression vectors, cDNAs encoding

MA

variable regions of heavy and light chains (VH and VL) of specific antibodies against H4K20me1 (CMA421/15F11, 12C8, and 22G3) were cloned as previously described [5].

D

VH and VL cDNAs were connected into a single chain through a linker encoding 3×

TE

GGGGS. The resulting DNA was subcloned into the pEGFP-N2 vector (Clontech).

AC CE P

Nucleotide sequence data of scFvs (CMA421/15F11, 12C8, and 22G3) are available in several public databases (DDBJ/EMBL/GenBank) under the accession numbers LC129890, LC129891, and LC129892, respectively. Amino acid substitution mutants were generated using a Quick Change kit (Stratagene). For expression of enzymes that control H4K20 methylation, Halo-PHF8, Halo-PR-Set7 (SETD8), Halo-SUV420H1, and Halo-SUV420H2 (Kazusa DNA Research Institute; FlexiHaloTag clone FHC01149, FHC06789, FHC01413, and FHC26822) were used. For Fig. 3 and 4, a pMX-puro-based expression vector carrying mCherry-PCNA [44] was co-transfected with the H4K20me1-mintbody expression vector. For analyzing enrichment of

24

ACCEPTED MANUSCRIPT

H4K20me1-mintbody on centromeres, hTERT-RPE1 cells were co-transfected with

RI

PT

expression vectors for CENP-A-TagRFP [45] and H4K20me1-mintbody.

SC

Histone modification antibodies

NU

Mouse monoclonal antibodies: anti-H4K20me1 (CMA421), anti-H4K20me2

MA

(CMA422), anti-H4K20me3 (CMA423) [22], anti-H3K4me1 (CMA302) [46], anti-H3K9me1 (CMA316) [11], anti-H3K27me1 (4C4), anti-H3K27me3 (CMA323) [4]

D

were used. For immunofluorescence, antibodies were labeled with Alexa Fluor 488

TE

(Thermo Fisher Scientific), Cy3 (GE Healthcare), or Cy5 (GE Healthcare). For

AC CE P

H3K9me1 (CMA316), unlabeled IgG and Cy5-conjugated goat anti-mouse Fc secondary antibody (Jackson Immunoresearch) were used.

Protein purification and biochemical analysis The DNA fragments encoding 15F11-scFv, 12C8-scFv, and 22G3-scFv were cloned into the pET15b vector (Novagen), which includes a His6 tag at the N-terminus. His6-15F11-scFv was expressed in Escherichia coli BL21 (DE3) at 25ºC. Bacterial cells producing His6-scFvs were harvested by centrifugation and then disrupted by sonication in 50 mM Tris-HCl (pH7.5) buffer, containing 500 mM NaCl, 1 mM

25

ACCEPTED MANUSCRIPT

phenylmethylsulfonyl fluoride (PMSF), and 10% glycerol. After centrifugation,

PT

precipitates were denatured in 50 mM Tris-HCl (pH8.0) buffer, containing 500 mM

RI

NaCl, 7 M Gdn-HCl, and 5% glycerol. After centrifugation, the supernatants were

SC

mixed with the Ni-NTA agarose (Qiagen) by rotating at 4ºC for 1 hr. The beads were

NU

packed into an Econo-column (Bio-Rad) and washed with a 50-column volume of 50

MA

mM Tris-HCl (pH8.0) buffer, containing 500 mM NaCl, 5 mM imidazole, 6 M Urea, 10% glycerol, and 2 mM 2-mercaptoethanol. His6-scFvs were eluted with a 50-column

D

volume liner gradient of 5-500 mM imidazole in 50 mM Tris-HCl (pH7.5) buffer,

TE

containing 500 mM NaCl, 6 M Urea, 10% glycerol, and 2 mM 2-mercaptoethanol. The

AC CE P

His6-scFvs were refolded during the dialysis against 20 mM Tris-HCl (pH7.5) buffer, containing 1 mM EDTA, 200 mM NaCl, 2 mM 2-mercaptoethanol, and 10% glycerol at 4ºC. The resulting His6-scFvs protein were further purified by gel filtration column chromatography (Superdex 75 HiLoad 16/60 prep grade; GE Healthcare) and eluted with phosphate-buffered saline (PBS). The purity was checked by 16% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (Supplementary Fig. S12c). For preparation of 15F11-scFv for Western blot analysis, histone modification arrays, immunofluorescence of yeast cells, and crystallization, the E. coli cells producing His6-15F11-scFv were disrupted by sonication in 50 mM Tris-HCl (pH7.5)

26

ACCEPTED MANUSCRIPT

buffer, containing 500 mM NaCl, 1 mM phenylmethylsulfonyl fluoride (PMSF), and

PT

10% glycerol. After removing cell debris by centrifugation, the supernatant was

RI

collected and mixed with Ni-NTA agarose by rotating at 4ºC for 30 min. The beads were

SC

washed with 50-column volume of 50 mM Tris-HCl (pH7.5) buffer, containing 500 mM

NU

NaCl, 5 mM imidazole, 1 mM PMSF, 10% glycerol, 0.01% Triton X-100, and 2 mM

MA

2-mercaptoethanol. His6-15F11-scFv was eluted with a 50-column volume liner gradient of 5-500 mM imidazole in 50 mM Tris-HCl (pH7.5) buffer, containing 500

D

mM NaCl, 1 mM PMSF, 10% glycerol, 0.01% Triton X-100, and 2 mM

TE

2-mercaptoethanol. The His6 tag peptide was cleaved by thrombin protease (GE

AC CE P

Healthcare) during the dialysis against 20 mM Tris-HCl (pH7.5) buffer, containing 1 mM EDTA, 200 mM NaCl, 2 mM 2-mercaptoethanol, and 10% glycerol at 4ºC. The resulting 15F11-scFv protein was further purified by gel filtration column chromatography (Superdex 75 HiLoad 26/60 prep grade; GE Healthcare), and eluted with phosphate-buffered saline (PBS). The purity was checked by 16% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (Supplementary Fig. S1a). For Western blot analysis (Supplementary Fig. S1b), recombinant histone H4 proteins harboring monomethylation-, dimethylation-, and trimethylation-mimic K20 (Active motif; 10-20 ng/lane for antibodies, 100 ng/lane for scFv) were separated on

27

ACCEPTED MANUSCRIPT

15% SDS-polyacrylamide gels. Proteins were transferred onto a polyvinylidene

PT

difluoride membranes in 100 mM Tris-HCl, 192 mM glycine, and 5% methanol, using a

RI

semidry transfer system. After blocking with Blocking One (Nacalai Tesque) for 20 min,

SC

membranes were incubated in H4K20me1-specific (15F11/CMA421; 1 µg/ml),

NU

H4K20me2-specific (CMA422; 1 µg/ml), and H4K20me3-specific (CMA423; 0.2

MA

µg/ml) mouse monoclonal antibodies [22] for 1 hr at room temperature, and then in HRP-conjugated goat anti-mouse IgG (Jackson) for 1 hr at room temperature. When

D

probed with the 15F11-scFv, membranes were incubated in 5 µg/ml scFv at 4ºC

TE

overnight, and then in HRP-conjugated goat anti-mouse IgG. For Supplementary Fig.

AC CE P

S1c, d, a peptide array (AbSurance Complete Core Histone Antibody Specificity Array; Millipore) was incubated in 15F11-scFv (2 µg/ml) overnight and then in HRP-conjugated goat anti-mouse IgG. Signals were detected using an enhanced chemiluminescent system (Immunostar LD; Wako Purechemicals). Surface plasmon resonance (SPR) signals (Supplementary Fig. S11, S12d) were measured using a Biacore X100 (GE Healthcare), as described previously [5]. H4K20me1 synthetic peptide (GKGGAKRHR(me1-K)VLRDNIQGI) (Sigma-Aldrich) was conjugated to a CM5 sensor chip (GE Healthcare) (for scFv, 87.1 RU; for IgG, 16.9 RU) by amine coupling in phosphate-buffered saline (PBS) containing 0.005% Tween20

28

ACCEPTED MANUSCRIPT

as a running buffer. For the binding assay, single cycle kinetics was performed, and

PT

different concentrations of antibodies (for scFv, 15.625, 31.25, 62.5, 125, and 250 nM;

RI

for IgG, 3.125, 6.25, 12.5, 25, and 50 nM) were successively injected at 30 µl/min. The

NU

SC

resulting curves were analyzed with a two state reaction model.

MA

Crystallization and structure determination

Purified 15F11-scFv was concentrated to 30 mg/ml, dialyzed against 20 mM

D

Bis-Tris-HCl (pH6.0) buffer, containing 100 mM NaCl, and crystallized using the

TE

hanging drop vapor diffusion method. One microliter of purified 15F11-scFv was mixed

AC CE P

with an equal volume of reservoir buffer (100 mM Sodium acetate [pH5.2], 7% poly-γ-glutamic acid polymer, and 12% PEG2000) at 20ºC and crystal growth was observed in 2 weeks. Crystals were soaked into the cryoprotectant solution (100 mM sodium acetate [pH5.2] and 30% PEG300) and flash-cooled in a stream of N2 gas (100K). The dataset was collected at the BL-1A, BL-5A, and BL-17A beamlines in the Photon Factory (KEK) and the BL41XU beamline in SPring-8. Diffraction data were indexed, integrated, and scaled using the HKL2000 program [47]. The structure of 15F11-scFv was solved by the molecular replacement method, using the Phaser program [48] and the PDB entry 2GHW as the search model [49]. Manual model

29

ACCEPTED MANUSCRIPT

building was performed using the COOT program [50]. Refinements of the models were

PT

performed using the PHENIX program [51]. The final structure contains no

RI

Ramachandran outliers, assessed by the MolProbity program (Supplementary Table I)

SC

[52]. All protein structure figures were prepared with PyMOL (Schrödinger;

NU

http://www.pymol.org). The structure of 15F11-scFv was deposited into the Protein data

MA

bank (PBD ID: 5B3N).

D

Cell culture

TE

Cells were routinely grown in DMEM (Dulbecco’s modified Eagle’s medium;

AC CE P

Nacalai Tesque) with supplements (L-glutamine/penicillin/streptomycin; Sigma-Aldrich; and 10% fetal calf serum) at 37 ºC under 5% CO2 atmosphere. Mouse embryonal carcinoma cell line MC12 was kindly provided by Dr. Nobuo Takagi [31]. Transfection of plasmid DNA was performed using Fugene HD (Promega) and cells stably expressing mintbody were selected in 1 mg/ml G418 (Nacalai Tesque). For live-cell imaging, cells were grown on a 35 mm glass bottom dish (Mat-Tek) or a 24-well glass bottom plate (Iwaki) and the medium was replaced with FluoroBrite DMEM (Thermo Fisher Scientific) containing supplements.

30

ACCEPTED MANUSCRIPT

Growth curve and cell cycle profile

PT

HeLa cells stably expressing H4K20me1-mintbody and the parental cells were

RI

plated on a six-well dish (3 wells each) at a concentration of 2×105 cells/ml (2 ml). Cells

SC

were harvested in the original volume every 12 hrs and cell concentration was counted

NU

using an EVE automatic cell counter (Nano EnTek). Cell concentrations from 3 wells

MA

were averaged to obtain the result of a single experiment. Three independent experiment were performed to obtain the mean and the standard deviation. For cell cycle profiling,

D

cells grown on a glass bottom dish were incubated in 10 µM 5-ethynyl-2'-deoxyuridine

TE

(EdU; Thermo Fisher Scientific) for 30 min, fixed with 4% formaldehyde (Electron

AC CE P

Microscopy Sciences) in 250 mM HEPES-NaOH (pH 7.4; Wako Purechemicals) containing 0.1% Triton X-100 for 5 min, and permeabilized with 1% Triton X-100 in PBS for 20 min at room temperature. EdU was labeled with Alexa Fluor 647 azide for 30 min, using Click-iT EdU Imaging Kit (Thermo Fisher Scientific), and DNA was stained with 1 µg/ml Hoschst33342 in PBS for 20 min, before imaging using a wide-field fluorescence microscope (Ti-E; Nikon; operated by NIS-elements software), equipped with an EM-CCD (iXON+ DU897 X-3019, Andor), a 40× Plan Apo (NA 0.95) lens. Hoechst33342 and Alexa Fluor 647 were detected using DAPI-1160A (Semrock; Ex FF01-387/11; DM FF02-447/60; Em FF409-Di02) and Cy5-4040A

31

ACCEPTED MANUSCRIPT

(Semrock; Ex FF02-628/40; DM FF660-Di02; Em FF01-692/40) filter sets. The nuclear

PT

intensities of fluorescence signals were automatically quantified using NIS-elements

RI

analysis software (over 1,000 cells per experiment). DNA content in each cell was

SC

obtained by multiplying mean Hoechst intensity with ROI area. Cell cycle phases were

NU

defined according to the DNA content and Cy5 intensity (G1, Cy5<200 and

MA

DNA<250,000; S phase, Cy5>200; G2/M, Cy5<200 and DNA>250,000). The average cell cycle fractions (with the standard deviations) were obtained from three independent

AC CE P

Immunofluorescence

TE

D

experiments.

Cells were fixed with 4% formaldehyde (Electron Microscopy Sciences) in 250

mM HEPES-NaOH (pH 7.4; Wako Purechemicals) containing 0.1% Triton X-100 for 5 min at room temperature, permeabilized with 1% Triton X-100 in PBS for 20 min at room temperature. After blocking with Blocking One-P (Nacalai Tesque) for 20 min at room temperature, cells were incubated with 2 µg/ml Fluorescence dye (Alexa Fluor 488, Cy3, or Cy5)-conjugated mouse monoclonal antibody for 2 hr at room temperature. After washing with PBS three times, cells were incubated with 0.25-1 µg/ml Hoechst33342 for 20min. After washing with PBS three times, fluorescence images

32

ACCEPTED MANUSCRIPT

were collected. For Supplementary Fig. S5a, a glass bottom dish with a grid

PT

(Matsunami) was used to match cells in time-lapse and immunofluorescence images.

RI

For Fig. 2b-d and Supplementary Fig. S3, 24 hrs after transfection with Halo Tag

SC

expression vectors, cells were stained with 2-4 µg/ml antibodies, 0.1 µM HaloTag

NU

Tetramethylrhodamine (TMR) ligand (Promega), and 0.25 µg/ml Hoechst33342 for 2 hr

D

Metaphase chromosome spreads

MA

at room temperature.

TE

MC12 cells stably expressing the H4K20me1-mintbody were incubated with

AC CE P

colcemid (50 ng/ml; Nacalai Tesque) for 1.5 hr, and mitotic cells were collected by pipetting and brief centrifugation. Mitotic cells were swelled using hypotonic buffer (75 mM KCl) for 10 min at room temperature. HEPES-NaOH (Wako Purechemicals; 250 mM final concentration, pH7.4) and formaldehyde (Electron Microscopy Sciences; 0.4% final concentration) were added just before the cytospin (WESCOR; 1,000 rpm, 2 min). Chromosome spreads were fixed for 5 min at room temperature, and then stained with 20 ng/ml Hoechst33342 in PBS for 10 min at room temperature.

Microscopy for mammalian cells

33

ACCEPTED MANUSCRIPT

Fluorescence images were collected using confocal microscopes: an FV1000

PT

(Olympus; operated by the built-in FV1000 software FLUOVIEW ver.4.2), a Ti-E

RI

(Nikon; operated by NIS-elements software) with a spinning disc (CSU-W1;

SC

Yokogawa) and an EM-CCD camera (iXON3 DU888 X-8465, Andor), or an IX-71

NU

(Olympus) equipped with a spinning disc (CSU-X1; Yokogawa) and an EM-CCD

MA

camera (iXON3 DU897E-CS0-#BV-Y, Andor). Live-cell imaging was performed with a heated stage (Tokai Hit) and a CO2-control system (Tokken). For Fig. 1b, 2e, and

D

Supplementary Fig. S5b, S7b, 1 µg/ml Hoechst 33342 was added to cells 1 hr before

TE

imaging. Acquisition parameters are shown in Supplementary Table II. Quantification of

AC CE P

fluorescence intensities was performed using NIS-elements analysis software (Nikon). For immunofluorescence images, nuclei were automatically defined by thresholding and fluorescence intensities in individual nuclei were quantified. For images of H4K20me1-mintbody, the ROI areas of the nuclei and cytoplasm were manually defined and the fluorescence intensities were quantified. The net intensity was obtained by background subtraction. To compare the nuclear enrichment of H4K20me1-mintbody in different cells, the Nuclear/Cytoplasm ratio was frequently used. The Nuclear/Cytoplasm ratio could be over influenced by a cytoplasmic intensity that is too low and variable, but the actual intensity counts a few hundred after

34

ACCEPTED MANUSCRIPT

background subtraction (in a 14-bit scale) and the deviation is comparable to or lower

PT

than the nuclear intensity. The statistics of the nuclear and cytoplasmic intensities used

RI

for Fig. 3e are shown in Supplementary Fig. S14, as an example. For Supplementary

SC

Fig. S4, co-localization analysis was performed using NIS-elements analysis software

MA

NU

(Nikon).

Fission yeast experiments

D

An S. pombe strain (HA1039-3D, h- lys1-131) was used as wild type. For set9+

TE

gene disruption, a gene disruption fragment was amplified by PCR using a

AC CE P

G418-resistant gene cassette, and used for transformation of the wild type strain [53]. An H4K20A mutant strain was constructed according to a previous study with some modifications [54]. Briefly, a pair of histone H4.2/H3.2 genes (hhf2+-hht2+) was cloned into a plasmid DNA, and an H4.2 K20A mutation was introduced by PCR-mediated site-directed mutagenesis to convert the 20th codon AAG to GCT. The genomic hhf2+-hht2+ gene was disrupted by replacement with the ura4+ marker gene, and the ura4+ marker gene was further replaced with the mutant gene. H4.2 K20A mutant cells were selected in a medium containing 5-fluoroorotic acid, which is toxic for ura4+ cells. To disrupt the other two histone H4/H3 genes, H4.1/H3.1 and H4.3/H3.3 were replaced

35

ACCEPTED MANUSCRIPT

with G418-resistant and Nourseothricin-resistant genes [55]. YES and EMM2 were used

PT

as complete and minimal media, respectively [56]. To express GFP-fused 15F11-scFv in

RI

S. pombe cells, the scFv-coding sequence was inserted downstream of the β-tubulin

SC

(nda3+) gene promoter and fused to GFPs65t. The resulting fusion gene was then

NU

integrated into the genomic lys1 locus. For Western blot analyses, whole cell extracts

MA

were prepared according to the following procedure. S. pombe cells were grown until they reached the log phase and further incubated in the presence of 1 mM PMSF for 10

D

min. 5×107 cells were then harvested and resuspended in 800 µl of ice-cold water. The

TE

cell suspension was mixed with 150 µl of 2 M NaOH, incubated for 10 min on ice,

AC CE P

added with 150 µl of 55% (w/v) trichloroacetic acid, mixed, and incubated for a further 10 min on ice. After centrifugation at 10,000 rpm for 2 min at 4ºC using a microcentrifuge, the pellet was washed with cold acetone and then resuspended in SDS-polyacrylamide gel sample buffer (60 mM Tris-HCl [pH 6.8], 2% SDS, 100 mM DTT, 10% glycerol and bromophenol blue). A volume equivalent to 2×106 cells from each sample was separated on 15% SDS-polyacrylamide gels. Proteins were transferred onto polyvinylidene difluoride membranes in 100 mM Tris-HCl, 192 mM glycine, and 5% methanol, using a semidry transfer system. Membranes were incubated in H4K20me1-specific (15F11/CMA421; 1 µg/ml) and pan-H4 (CMA400; 0.1 µg/ml)

36

ACCEPTED MANUSCRIPT

mouse monoclonal antibodies [22], and a GFP-specific rabbit polyclonal antibody

PT

(Rockland Immunochemicals Inc.; 0.25 µg/ml), and then in HRP-conjugated goat

RI

anti-mouse or anti-rabbit IgG (GE Healthcare). For immunoblot experiments with the

SC

15F11-scFv, membranes were incubated in 20µg/ml scFv overnight at 4ºC, and then in

NU

HRP-conjugated goat anti-mouse IgG.

MA

Live-cell imaging was performed at 26ºC using a DeltaVision fluorescence microscopy system (Applied Precision), which is built on a wide-field fluorescence

D

microscope IX71 (Olympus), equipped with a 60× PlanApoN (NA 1.40) oil-immersion

TE

lens and a CoolSNAP HQ2 CCD (Photometrics). Cells were immobilized as described

AC CE P

previously [57]. Images were deconvolved using SoftWoRx software. Fluorescence microscopy using anti-H4K20me1 antibody (CMA421) and

15F11-scFv was done as follows. Cells grown in complete medium were harvested and fixed using 4% formaldehyde in 0.1 M phosphate buffer (pH7.2) for 10 min at room temperature. After cell wall digestion by 0.6mg/ml zymolyase 100T (Nacalai tesque) in PEMS buffer (100mM PIPES, 1mM EGTA, 1mM magnesium chloride and 1.2M sorbitol) for 70 min at 36˚C, cells were permeabilized by 1% Triton X-100 in PEMS for 2 min at room temperature. Blocking was done with 1% BSA for 30 min at room temperature. For staining with CMA421, cells were incubated with 1 µg/ml antibody for

37

ACCEPTED MANUSCRIPT

1 hr at room temperature. For staining with 15F11-scFv, cells were incubated with 100

PT

µg/ml scFv overnight at 4˚C although the same result was obtained by incubation with

RI

20µg/ml scFv. An Alexa488-conjugated goat anti-mouse antibody (GE healthcare) was

SC

used as a secondary antibody. Chromatin was stained with a 0.2 µg/ml

NU

4',6-diamidino-2-phenylindole (DAPI). Cells were observed using DeltaVision

MA

fluorescence microscopy system (Applied Precision).

D

Expression and imaging of theH4K20me1-mintbody in the C. elegans embryos

TE

The DNA sequence of 15F11-scFv was synthesized, taking into consideration

AC CE P

codon usage in C. elegans (GeneArt Gene Synthesis; Thermo Fisher Scientific). To generate an expression vector compatible with Mos1-mediated single copy insertion (MosSCI), the following four segments were joined using a Gateway Compatible Plasmid Toolkit and Multisite Gateway Technology (Thermo Fisher Scientific), as previously described [58]. The segments joined were: (i) mex-5 5’UTR, (ii) codon-optimized 15F11-scFv sequence, (iii) mex-5 3’UTR following to gfp F64LS65T , and (iv) MosSCI integration site followed by unc-119+ rescue marker. Using the MosSCI technology, the resulting vector was integrated into the unc-119 (ed3) mutant genome. The integrated lines were screened according to previously described protocols

38

ACCEPTED MANUSCRIPT

[59]. Strains were maintained at 22°C or 26°C. For microscopic imaging, adult

PT

hermaphrodites were dissected in M9 buffer and released embryos were transferred onto

RI

a 2% agarose pad mounted on a glass slide and covered with a coverslip. Fluorescence

SC

signals of H4K20me1-mintbody were collected at room temperature using a

NU

microlens-enhanced spinning-disk confocal unit (CSU-X1; Yokogawa) mounted on an

MA

inverted microscope (IX71; Olympus) equipped with a 100× UPlanSApo (NA 1.40) lens and an EM-CCD camera (iXon; Andor), and controlled by MetaMorph imaging

D

software (ver 7.7.10.0, Molecular Devices). z-stacks of 12 planes with 2 µm intervals

TE

were collected using a 488-nm laser line (100 ms exposure). Stacked images, which

AC CE P

were composed by maximum projection of seven to nine serial z-planes, and extracted single section images (Fig. 5) were generated using MetaMorph imaging software (ver 7.7.10.0).

Acknowledgments We thank M.C. Cardoso (Technische Universität Darmstadt) for providing the PCNA-GFP construct, Nobuo Takagi for supplying MC12 cells, Naohito Nozaki (MAB Institute Inc) for generating hybridomas, and T.J. Stasevich (Colorado State University)

39

ACCEPTED MANUSCRIPT

for reading the manuscript. We are grateful to the beamline scientists of the BL41XU

PT

beamline of Spring-8 and the BL-1A, BL-5A, and BL-17A beamlines of the Photon

RI

Factory. TK was supported by Research Fellowships for Young Scientists from Japan

SC

Society for the Promotion of Science (JSPS). This work was supported by JSPS

NU

KAKENHI Grant Number JP25118714 (to YS), JP26440098 (to HA), JP26116511 (to

MA

YH), JP25116006 (to TH), JP20114008 (to AK), JP25116002 (to H.Kurumizaka), and JP25116005 and JP26291071 (to H.Kimura), by the Japan Synchrotron Radiation

D

Research Institute (JASRI) (proposal nos. 2012B1048, 2013A1036, 2013B1060,

TE

2014B1125, 2015A1020, and 2015B2020), and the Photon Factory Program Advisory

AC CE P

Committee (proposal no. 2012G569 and 2014G556).

Author Contributions

YS and H.Kimura conceived the study, designed the experiments, and wrote the manuscript. YS performed most experiments and analyzed data. TK and NH performed structural and biochemical analyses under the supervision of H.Kurumizaka. RA performed nematode experiments under the supervision of AK. HA and CO performed yeast experiments under the supervision of TH and YH. KY and JU collected images using a spinning disc confocal microscope. TN provided Halo-tag expression vectors.

40

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

41

ACCEPTED MANUSCRIPT

References

PT

[1] Bannister AJ, Kouzarides T. Regulation of chromatin by histone modifications. Cell

RI

Res. 2011;21:381-95.

NU

the human genome. Nature. 2012;489:57-74.

SC

[2] The ENCODE Project Consortium. An integrated encyclopedia of DNA elements in

MA

[3] Kimura H, Hayashi-Takanaka Y, Stasevich TJ, Sato Y. Visualizing posttranslational and epigenetic modifications of endogenous proteins in vivo. Histochem Cell Biol.

D

2015;144:101-9.

TE

[4] Hayashi-Takanaka Y, Yamagata K, Wakayama T, Stasevich TJ, Kainuma T,

AC CE P

Tsurimoto T, et al. Tracking epigenetic histone modifications in single cells using Fab-based live endogenous modification labeling. Nucleic Acids Res. 2011;39:6475-88. [5] Sato Y, Mukai M, Ueda J, Muraki M, Stasevich TJ, Horikoshi N, et al. Genetically encoded system to track histone modification in vivo. Sci Rep. 2013;3:2436. [6] Sato Y, Mukai M, Kimura H. Technical Note: Histone Acetylation on Drosophila Polytene Chromosomes Visualized by Mintbody. Cytologia. 2015;80:383-. [7] Suzuki M, Takagi C, Miura S, Sakane Y, Suzuki M, Sakuma T, et al. In vivo tracking of histone H3 lysine 9 acetylation in Xenopus laevis during tail regeneration. Genes to Cells. 2016;21:358-69.

42

ACCEPTED MANUSCRIPT

[8] Tardat M, Brustel J, Kirsh O, Lefevbre C, Callanan M, Sardet C, et al. The histone

PT

H4 Lys 20 methyltransferase PR-Set7 regulates replication origins in mammalian cells.

RI

Nat Cell Biol. 2010;12:1086-93.

SC

[9] Abbas T, Shibata E, Park J, Jha S, Karnani N, Dutta A. CRL4(Cdt2) regulates cell

NU

proliferation and histone gene expression by targeting PR-Set7/Set8 for degradation.

MA

Mol Cell. 2010;40:9-21.

[10] Beck DB, Burton A, Oda H, Ziegler-Birling C, Torres-Padilla ME, Reinberg D. The

TE

2012;26:2580-9.

D

role of PR-Set7 in replication licensing depends on Suv4-20h. Genes Dev.

AC CE P

[11] Tuzon CT, Spektor T, Kong X, Congdon LM, Wu S, Schotta G, et al. Concerted activities of distinct H4K20 methyltransferases at DNA double-strand breaks regulate 53BP1 nucleation and NHEJ-directed repair. Cell Rep. 2014;8:430-8. [12] Trojer P, Li G, Sims RJ, 3rd, Vaquero A, Kalakonda N, Boccuni P, et al. L3MBTL1, a histone-methylation-dependent chromatin lock. Cell. 2007;129:915-28. [13] Nishioka K, Rice JC, Sarma K, Erdjument-Bromage H, Werner J, Wang Y, et al. PR-Set7 is a nucleosome-specific methyltransferase that modifies lysine 20 of histone H4 and is associated with silent chromatin. Mol Cell. 2002;9:1201-13.

43

ACCEPTED MANUSCRIPT

[14] Xiao B, Jing C, Kelly G, Walker PA, Muskett FW, Frenkiel TA, et al. Specificity

PT

and mechanism of the histone methyltransferase Pr-Set7. Genes Dev. 2005;19:1444-54.

RI

[15] Oda H, Okamoto I, Murphy N, Chu J, Price SM, Shen MM, et al.

SC

Monomethylation of histone H4-lysine 20 is involved in chromosome structure and

NU

stability and is essential for mouse development. Mol Cell Biol. 2009;29:2278-95.

MA

[16] Schotta G, Lachner M, Sarma K, Ebert A, Sengupta R, Reuter G, et al. A silencing pathway to induce H3-K9 and H4-K20 trimethylation at constitutive heterochromatin.

D

Genes Dev. 2004;18:1251-62.

TE

[17] Qi HH, Sarkissian M, Hu GQ, Wang Z, Bhattacharjee A, Gordon DB, et al. Histone

AC CE P

H4K20/H3K9 demethylase PHF8 regulates zebrafish brain and craniofacial development. Nature. 2010;466:503-7. [18] Liu W, Tanasa B, Tyurina OV, Zhou TY, Gassmann R, Liu WT, et al. PHF8 mediates histone H4 lysine 20 demethylation events involved in cell cycle progression. Nature. 2010;466:508-12. [19] Kohlmaier A, Savarese F, Lachner M, Martens J, Jenuwein T, Wutz A. A chromosomal memory triggered by Xist regulates histone methylation in X inactivation. PLoS Biol. 2004;2:E171.

44

ACCEPTED MANUSCRIPT

[20] Schotta G, Sengupta R, Kubicek S, Malin S, Kauer M, Callen E, et al. A

PT

chromatin-wide transition to H4K20 monomethylation impairs genome integrity and

RI

programmed DNA rearrangements in the mouse. Genes Dev. 2008;22:2048-61.

SC

[21] Hori T, Shang WH, Toyoda A, Misu S, Monma N, Ikeo K, et al. Histone H4 Lys 20

NU

monomethylation of the CENP-A nucleosome is essential for kinetochore assembly.

MA

Dev Cell. 2014;29:740-9.

[22] Hayashi-Takanaka Y, Maehara K, Harada A, Umehara T, Yokoyama S, Obuse C, et

D

al. Distribution of histone H4 modifications as revealed by a panel of specific

TE

monoclonal antibodies. Chromosome Res. 2015.

AC CE P

[23] Toda T, Yamamoto M, Yanagida M. Sequential alterations in the nuclear chromatin region during mitosis of the fission yeast Schizosaccharomyces pombe: video fluorescence microscopy of synchronously growing wild-type and cold-sensitive cdc mutants by using a DNA-binding fluorescent probe. J Cell Sci. 1981;52:271-87. [24] Uzawa S, Yanagida M. Visualization of centromeric and nucleolar DNA in fission yeast by fluorescence in situ hybridization. J Cell Sci. 1992;101 ( Pt 2):267-75. [25] Sanders SL, Portoso M, Mata J, Bahler J, Allshire RC, Kouzarides T. Methylation of histone H4 lysine 20 controls recruitment of Crb2 to sites of DNA damage. Cell. 2004;119:603-14.

45

ACCEPTED MANUSCRIPT

[26] Haraguchi T, Ding DQ, Yamamoto A, Kaneda T, Koujin T, Hiraoka Y.

PT

Multiple-color fluorescence imaging of chromosomes and microtubules in living cells.

RI

Cell Struct Funct. 1999;24:291-8.

SC

[27] Rice JC, Nishioka K, Sarma K, Steward R, Reinberg D, Allis CD. Mitotic-specific

NU

methylation of histone H4 Lys 20 follows increased PR-Set7 expression and its

MA

localization to mitotic chromosomes. Genes Dev. 2002;16:2225-30. [28] Pesavento JJ, Yang H, Kelleher NL, Mizzen CA. Certain and progressive

TE

2008;28:468-86.

D

methylation of histone H4 at lysine 20 during the cell cycle. Mol Cell Biol.

AC CE P

[29] Leonhardt H, Rahn HP, Weinzierl P, Sporbert A, Cremer T, Zink D, et al. Dynamics of DNA replication factories in living cells. J Cell Biol. 2000;149:271-80. [30] Stasevich TJ, Sato Y, Nozaki N, Kimura H. Quantifying histone and RNA polymerase II post-translational modification dynamics in mother and daughter cells. Methods. 2014;70:77-88. [31] Abe K, Takagi N, Sasaki M. Nonhistone nuclear proteins specific to certain mouse embryonal carcinoma clones having an inactive X chromosome. Exp Cell Res. 1988;179:590-4.

46

ACCEPTED MANUSCRIPT

[32] Meyer BJ. Targeting X chromosomes for repression. Curr Opin Genet Dev.

PT

2010;20:179-89.

RI

[33] Vielle A, Lang J, Dong Y, Ercan S, Kotwaliwale C, Rechtsteiner A, et al.

NU

compensation. PLoS Genet. 2012;8:e1002933.

SC

H4K20me1 contributes to downregulation of X-linked genes for C. elegans dosage

MA

[34] Ewert S, Honegger A, Pluckthun A. Stability improvement of antibodies for extracellular and intracellular applications: CDR grafting to stable frameworks and

D

structure-based framework engineering. Methods. 2004;34:184-99.

TE

[35] Philibert P, Stoessel A, Wang W, Sibler AP, Bec N, Larroque C, et al. A focused

AC CE P

antibody library for selecting scFvs expressed at high levels in the cytoplasm. BMC Biotechnol. 2007;7:81.

[36] Lo AS, Zhu Q, Marasco WA. Intracellular antibodies (intrabodies) and their therapeutic potential. Handb Exp Pharmacol. 2008:343-73. [37] Cattaneo A, Biocca S. The selection of intracellular antibodies. Trends Biotechnol. 1999;17:115-21. [38] Ewert S, Cambillau C, Conrath K, Pluckthun A. Biophysical properties of camelid V(HH) domains compared to those of human V(H)3 domains. Biochemistry. 2002;41:3628-36.

47

ACCEPTED MANUSCRIPT

[39] Honegger A, Malebranche AD, Rothlisberger D, Pluckthun A. The influence of the

RI

variable domains. Protein Eng Des Sel. 2009;22:121-34.

PT

framework core residues on the biophysical properties of immunoglobulin heavy chain

SC

[40] Kabat EA, Wu TT, Perry HM, Gottesmann KS, Foeller C. Sequences of proteins of

NU

immunological interest. 5th ed: NIH publication; 1991.

MA

[41] Kvam E, Sierks MR, Shoemaker CB, Messer A. Physico-chemical determinants of soluble intrabody expression in mammalian cell cytoplasm. Protein Eng Des Sel.

D

2010;23:489-98.

TE

[42] Casas-Delucchi CS, Brero A, Rahn HP, Solovei I, Wutz A, Cremer T, et al. Histone

AC CE P

acetylation controls the inactive X chromosome replication dynamics. Nat Commun. 2011;2:222.

[43] Koren A, McCarroll SA. Random replication of the inactive X chromosome. Genome Res. 2014;24:64-9.

[44] Sato K, Ishiai M, Toda K, Furukoshi S, Osakabe A, Tachiwana H, et al. Histone chaperone activity of Fanconi anemia proteins, FANCD2 and FANCI, is required for DNA crosslink repair. EMBO J. 2012;31:3524-36.

48

ACCEPTED MANUSCRIPT

[45] Tachiwana H, Kagawa W, Shiga T, Osakabe A, Miya Y, Saito K, et al. Crystal

PT

structure of the human centromeric nucleosome containing CENP-A. Nature.

RI

2011;476:232-5.

SC

[46] Kimura H, Hayashi-Takanaka Y, Goto Y, Takizawa N, Nozaki N. The Organization

NU

of Histone H3 Modifications as Revealed by a Panel of Specific Monoclonal Antibodies.

MA

Cell Structure and Function. 2008;33:61-73.

[47] Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in

D

oscillation mode. Method Enzymol. 1997;276:307-26.

TE

[48] McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ.

AC CE P

Phaser crystallographic software. J Appl Crystallogr. 2007;40:658-74. [49] Hwang WC, Lin Y, Santelli E, Sui J, Jaroszewski L, Stec B, et al. Structural basis of neutralization by a human anti-severe acute respiratory syndrome spike protein antibody, 80R. J Biol Chem. 2006;281:34610-6. [50] Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60:2126-32. [51] Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr. 2010;66:213-21.

49

ACCEPTED MANUSCRIPT

[52] Chen VB, Arendall WB, 3rd, Headd JJ, Keedy DA, Immormino RM, Kapral GJ, et

PT

al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta

RI

Crystallogr D Biol Crystallogr. 2010;66:12-21.

SC

[53] Wach A. PCR-synthesis of marker cassettes with long flanking homology regions

NU

for gene disruptions in S. cerevisiae. Yeast. 1996;12:259-65.

MA

[54] Mellone BG, Ball L, Suka N, Grunstein MR, Partridge JF, Allshire RC. Centromere silencing and function in fission yeast is governed by the amino terminus of histone H3.

D

Curr Biol. 2003;13:1748-57.

TE

[55] Takayama Y, Takahashi K. Differential regulation of repeated histone genes during

AC CE P

the fission yeast cell cycle. Nucleic Acids Res. 2007;35:3223-37. [56] Moreno S, Klar A, Nurse P. Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Methods Enzymol. 1991;194:795-823. [57] Asakawa H, Hiraoka Y. Live-cell fluorescence imaging of meiotic chromosome dynamics in Schizosaccharomyces pombe. Methods Mol Biol. 2009;558:53-64. [58] Zeiser E, Frokjaer-Jensen C, Jorgensen E, Ahringer J. MosSCI and gateway compatible plasmid toolkit for constitutive and inducible expression of transgenes in the C. elegans germline. PLoS One. 2011;6:e20082.

50

ACCEPTED MANUSCRIPT

[59] Frokjaer-Jensen C, Davis MW, Sarov M, Taylor J, Flibotte S, LaBella M, et al.

PT

Random and targeted transgene insertion in Caenorhabditis elegans using a modified

AC CE P

TE

D

MA

NU

SC

RI

Mos1 transposon. Nat Methods. 2014;11:529-34.

51

ACCEPTED MANUSCRIPT

RI

Figure 1. Generation of H4K20me1 specific mintbody

PT

Figure legends

SC

(a) Schematic structure of mintbody. Antibody variable regions of heavy and light

NU

chains (VH and VL, respectively) were linked with a flexible linker and fused to EGFP

MA

at the C-terminus. (b) The localization of H4K20me1-mintbodies in living cells. H4K20me1-mintbody was stably expressed in HeLa cells, and DNA was counterstained

D

with Hoechst33342. H4K20me1-mintbody was concentrated in nuclei during interphase

TE

and bound to chromosomes during mitosis. (c-e) H4K20me1-mintbody in

AC CE P

Schizosaccharomyces pombe. Yeast strains expressing H4K20me1-mintbody (c and e) or GFP alone were generated. WT, the wild type; set9∆, set9-deleted strain; H4.2 wt, a strain that harbors a single H4 gene; H4.2 K20A, a strain that harbors K20A-substitution in the single H4.2 gene. (c) The levels of H4K20me1, histone H4, and H4K20me1-mintbody (GFP) were assessed by Western blotting. H4K20me1 was depleted in set9∆ and H4.2 K20A strains. (d and e) Living yeast cells expressing H4K20me1-mintbody (d) and GFP (e) were stained with Hoechst33342 and imaged using a fluorescence microscope. (d) The H4K20me1-mintbody was concentrated in Hoechst-positive chromatin regions in WT and H4.2 strains, and diffuse in the nucleus

52

ACCEPTED MANUSCRIPT

and cytoplasm in set9∆ and H4.2 K20A strains. (e) GFP distributed in the nucleus and

RI

PT

cytoplasm in all strains. Bars, 10 µm.

SC

Figure 2. The nuclear concentration of H4K20me1-mintbody depends on H4K20

NU

monomethylation in living cells

MA

(a) H4K20 methylation states and responsible enzymes. The level of H4K20 monomethylation can be controlled by PR-Set7, SUV420H1/2, and PHF8. (b-d)

D

Overexpression of SUV420H1 decreased H4K20me1 level. HeLa cells transiently

TE

expressing Halo-SUV420H1 were fixed and stained with Alexa Fluor 488-conjugated

AC CE P

mouse antibodies directed against H4K20me1, H4K20me2, or H4K20me3, Tetramethylrhodamine (TMR) Halo ligand, and Hoechst33342. Optical sections were collected using a confocal microscope, and fluorescent intensities of H4K20 mono-, di-, or tri-methylation and Halo-SUV420H1 in single cells were quantified. (b) Example images. H4K20me1 signal was weaker in a cell expressing SUV420H1 (arrowhead). (c and d) Intensity plots. Each dot represents the intensities of Halo-SUV420H1 (x-axis) and H4K20 methylation (y-axis) in a single cell nucleus. (c) Halo-SUV420H1 transfected cells. (d) Untransfected cells. See Supplementary Fig. S3 for the effects of other enzymes. (e) H4K20me1 levels in living cells monitored with

53

ACCEPTED MANUSCRIPT

H4K20me1-mintbody. Halo-SUV420H1 was transiently expressed in HeLa cells that

PT

stably expressed H4K20me1-mintbody. The mintbody was more diffuse in the

RI

cytoplasm in the cell that expressed SUV420H1 (arrowhead). Nuclear/Cytoplasm

SC

intensity ratios in single cells were plotted (right graph). Nuclear/Cytoplasm ratios

MA

NU

decreased depending on Halo-SUV420H1 expression. Bars, 10 µm.

Figure 3. Tracking the level of H4K20me1 during the cell cycle

D

(a) Growth curves of HeLa cells expressing H4K20me1-mintbody and the parental cells.

TE

Cells were plated on a dish (2×105 cells/ml; 2 ml) at time 0, and the cell number was

AC CE P

counted by resuspending into the original volume every 12 hr. Mean cell concentrations at each time point from three independent experiments is plotted with the standard deviation. Cells expressing H4K20me1-mintbody proliferated similarly to control HeLa cells. (b) Cell cycle analysis. Cell cycle phase was defined based on the intensity of incorporated 5-Ethynyl-2'-deoxyuridine (EdU) and DNA content (Hoechst33342 staining). Example profiles and the fractions of different phases (means with the standard deviations; n = 3) are shown. No significant difference in each cell cycle phase was observed between the control and mintbody-expressing cells (p-values 0.227, 0.256, and 0.392, for G1, S, and G2/M, respectively, by Student’s t-test). (c) H4K20me1 levels

54

ACCEPTED MANUSCRIPT

during the cell cycle. HeLa cells expressing mCherry-PCNA were imaged overnight to

PT

identify the cell cycle position, fixed, and stained with anti-H4K20me1 antibody

RI

(CMA421). Intensities of the antibody in single cells at different cell cycle positions are

SC

shown (means with the standard deviations, n = 10-15). See Supplementary Fig. S5a for

NU

details. (d) H4K20me1 levels normalized with intensities of Hoechst33342 during the

MA

cell cycle. The antibody intensity was divided by Hoechst intensity to compensate for chromatin condensation. (e) Nuclear/Cytoplasm ratio of H4K20me1-mintbody during

D

the cell cycle. HeLa cells that stably express H4K20me1-mintbody and mCherry-PCNA

TE

were imaged overnight to identify the cell cycle position, and Nuclear/Cytoplasm ratio

AC CE P

was measured. Means with the standard deviations are plotted (n = 9-13). See Supplementary Fig. S5b for details. (f) Representative images of H4K20me1-mintbody and mCherry-PCNA. Elapsed times are indicated above the images (hh:mm). The cell cycle phase, judged by the nuclear distribution of PCNA, are shown below. Scale bar, 10 µm. (g) The fluorescent intensities in the nucleus and cytoplasm were measured and the Nuclear/Cytoplasm ratios are plotted. Intensities during mitotic phase are indicated as dashed gray line as the values cannot be directly compared with those during interphase (See Supplementary Figs. S7 and S8).

55

ACCEPTED MANUSCRIPT

Figure 4. Visualization of inactive X chromosomes in living cells

PT

(a) H4K20me1-mintbody was expressed in mouse embryonal carcinoma MC12 cells,

RI

which harbor one or two Xi. The cells were fixed and stained with antibodies against

SC

H3K27me3 (Cy3) and H4K20me1 (Cy5), and with Hoechst33342. IF,

NU

immunofluorescence. (b) Chromosome spreads. Metaphase chromosomes were

MA

prepared by a cytospin and stained with Hoechst33342. (c) Time-lapse images of mitosis. Elapsed times are indicated (hh:mm). (d) Representative images of MC12 cells

D

expressing both H4K20me1-mintbody and mCherry-PCNA. The timing of Xi

TE

replication was judged by the co-localization of H4K20me1-mintbody and

AC CE P

mCherry-PCNA. Arrowheads indicate Xi, which replicated in the middle of S phase (Mid-S). (e) The timing and length of Xi replication in S-phase. The period of PCNA foci appearance (“S-phase”, red bars) and the timing of co-localization with Xi foci (“Xi replication”, yellow bars) are shown (9 individual cells and the averages with the standard deviations). Bars, 10 µm.

Figure 5. Visualization of dosage-compensated X chromosomes in Caenorhabditis elegans with H4K20me1-mintbody.

56

ACCEPTED MANUSCRIPT

H4K20me1-mintbody was stably expressed in Caenorhabditis elegans. Representative

PT

images at various developmental stages, from Bean to 3-fold, are shown (maximum

RI

projection images of twelve sections and single sections). Scale bar, 10 µm. Magnified

SC

images of six different positions (numbered boxes) are also shown. Scale bar in

MA

NU

magnified images, 5 µm.

Figure 6. Mutational and structural analysis of H4K20me1-mintbody

D

(a) Intracellular localization of H4K20me1-mintbody. H4K20me1-mintbody (derived

TE

from 15F11) and 12C8-scFv-EGFP were expressed in HeLa cells. 12C8-scFv-EGFP

AC CE P

was not functional in the cells. (b) The position of amino acid substitutions between 15F11 and 12C8. (c) The localization of 12C8-scFv-EGFP mutants. Each of the five amino acids in 12C8 was changed to the corresponding amino acid in 15F11 (R32S, M95I, G99A, D105H, and A150T) and expressed in HeLa cells. The nuclear localization of G99A mutant suggests this mutation significantly improved the stability and/or folding in cells. (d) The localization of H4K20me1-mintbody mutants. Either I95M or A99G mutation alone did not abolish the nuclear localization, although the nuclear signal was relatively weakened. The combinatorial mutation was not concentrated in nuclei. (e) The crystal structure of 15F11-scFv at 1.94 Å resolution. VH

57

ACCEPTED MANUSCRIPT

and VL are shown in light blue and pink, respectively. The amino acids that differ in

PT

12C8 are indicated in yellow. Space filling models of two important residues (I95 and

RI

A99; letters shown in red) are blown-up to the side. The oxygen and nitrogen atoms are

AC CE P

TE

D

MA

NU

SC

colored in red and dark blue. Bars, 10 µm.

58

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

59

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

60

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

61

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

62

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

63

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

64

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Graphical abstract

65

ACCEPTED MANUSCRIPT

Highlights

PT

▪A histone H4K20me1-specific live-cell probe, H4K20me1-mintbody, was developed.

RI

▪Dynamics and replication timing of inactive X chromosomes were visualized.

AC CE P

TE

D

MA

NU

SC

▪Critical amino acids for the stability and/or folding of the mintbody were revealed.

66