Human embryonic kidney (HEK) 293T cells were obtained from American Type Culture Collection and grown (at 5% CO₂ and 37°C) in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (GE Healthcare Hyclone, Logan, UT) and 1% anti-biotics/anti-mycotics.

FLAG-HP1γ expression plasmids were used, as previously described [10]. Transient transfection was performed using FuGENE 6 reagent (Roche Applied Science, Indianapolis, IN).

Anti-FLAG antibody was purchased from Sigma (St. Louis, MO), and γH2AX antibody (EMD Millipore, Billerica, MA) was used as previously described [11]. The antibodies used for immunoblotting were as follows: anti-phospho-H3 Ser 10 (EMD Millipore), anti-Bloom syndrome protein (BLM) (Abcam, Cambridge, MA), and anti-proliferating cell nuclear antigen (PCNA) (Abcam).

Cells were synchronized at the late G₁ phase using a double thymidine block method [12]. Briefly, the cells were plated in 150-mm diameter Petri dishes, and thymidine was added to a final concentration of 2 mM after cell adherence. The cells were cultured for 16 hours. After removal of the thymidine and incubation for 10 hours in fresh medium, thymidine was again added to a final concentration of 2 mM and the cells were cultured for an additional 16 hours. After removal of the thymidine, synchronized cells were cultured in fresh medium and collected at different times for cell cycle analysis and immunoblotting. The cells were synchronized in the prometaphase with 17 hours of nocodazole treatment and then released into fresh medium for further incubation. Immunoblotting analysis was performed to confirm the indicated phases of the cell cycle using antibodies against phospho-H3 Ser10, a mitotic marker.

The shRNA for HP1γ was generated using a pSUPER. retro.puro, an H1 promoter-driven RNAi retroviral vector (Oligoengine, Seattle, WA). The shRNA primers were designed to target HP1γ (5´-GATCGACGTGTAGTGAATG-3´).

The cells were grown at 80%-90% confluence in 150-mm culture dishes, then washed with ice-cold phosphate-buffered saline. The harvested cells were lysed with ice-cold Tris lysis buffer (175 mM Tris, pH 8.0, 150 mM NaCl, 1% NP-40, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and complete protease/phosphatase-inhibitor cocktail, according to Thermo Scientifics’ instructions). The cellular extract (10 mg of total lysates for each sample) was mixed with 30 μL of EZview Red ANTI-FLAG M2 Affinity Gel (Sigma) and incubated for 2 hours at 4°C, with rotation. The bound proteins were eluted from the beads by boiling for 5 minutes in 2× sodium dodecyl sulfate sample buffer containing 5% (v/v) β-mercaptoethanol, and resolved by a one-dimensional polyacrylamide gel electrophoresis. Coomassie-stained HP1γ-interacting proteins were in-gel digested with trypsin (Promega, Madison, WI) and analyzed by capillary column liquid chromatography-tandem mass spectrometry (LC-MS/MS), using LTQ-Orbitrap mass spectrometry systems (Thermo Finnigan, San Jose, CA) equipped with nanospray ionization sources. Tandem mass spectra were interpreted by the Sorcerer program using the SEQUEST algorithm, and subsequently by the Scaffold program (version Scaffold_4.0.5, Proteome Software Inc., Portland, OR). Peptide and protein identifications were accepted if they could be established with greater than 95% probability and contained at least two identified unique peptides. All searches were performed against the human protein sequence database (IPI human DB v3.87 fasta).

For coimmunoprecipitation, the cell lysates were incubated with anti-HP1γ antibodies (EMD Millipore) and the immunoprecipitated complex was analyzed by immunoblotting using anti-BLM or anti-PCNA antibody.

The HeLa cells were strained with appropriate primary antibodies, followed by incubation with Alexa Fluor 488- or Alexa Fluor 546-conjugated secondary antibodies (Invitrogen, Carlsbad, CA). The cell nuclei were stained with 4´,6-diamidino-2-phenylindole. The samples were analyzed under an LSM700 confocal laser scanning microscope (Carl Zeiss, Thornwood, NY).

To identify new HP1γ-binding proteins, we performed affinity purification, using G₁/S phase or prometaphase HEK293T cell lysates that transiently express mock or FLAG-HP1γ. Coomassie staining was performed for HP1γ-binding complexes using cell lysates prepared by affinity chromatography FLAG-agarose beads (Fig. 1). We detected several bands that indicated elution from each transfected cell line. The bands were digested and mass spectrometry analysis was performed. A comparison of the binding proteins found in cell lysates eluted from each transfected cell line is shown in Supplementary Table 1. Most of the identified proteins were novel interacting proteins, which had not been previously reported. Subcellular distribution analysis showed that the proteins with HP1γ interaction were present in all major cellular compartments, including the nucleus, cytoplasm, and plasma membrane (Fig. 2A, Supplementary Table 2). These proteins were found to be involved in various biological processes, such as splicing, regulation of the actin cytoskeleton, tight junctions, pathogenic Escherichia coli infection, the mammalian target of rapamycin signaling pathway, nucleotide excision repair, DNA replication, homologous recombination, and mismatch repair (Fig. 2B, Supplementary Table 3). We also analyzed the domains in the HP1γ-binding proteins and found that HP1γ associates with many proteins via several functional domains (Fig. 2C, Supplementary Table 4). Therefore, our interactome analysis demonstrated that HP1γ has several potential roles in various biological processes.

By bioinformatic analysis, we identified 13 proteins implicated in the DNA damage response pathway, which comprise about 13% of potential candidates that interact with HP1γ (Table 1, Fig. 2D and E, Supplementary Table 5). First, we subtracted 31 proteins only in mock lysates, from a total of 130 proteins identified in our proteomic study, and then 99 proteins were considered as HP1γ-interacting partner candidates (Fig. 2D, Supplementary Table 1). Among these 99 HP1γ-interacting candidate proteins, we accepted 13 proteins if their unique spectral counts could be detected at greater than or equal to approximately two-fold ratio compared to the control (fold ratio of G₁-S or mitosis to mock control) (Fig. 2D and E, Supplementary Table 1), and functional relationship with DNA damage response pathways were validated by gene ontology analysis (Table 1, Supplementary Table 5) or literature survey (Table 1). Among the proteins identified as HP1γ-interacting partners, this data set, following validation analyses in Table 1, could postulate a functional connection between HP1γ and DNA damage response.

Therefore, we investigated the potential role of HP1γ in the DNA damage response pathway. Next, we tested whether HP1γ was indeed interacting with these identified proteins related to DNA damage response pathway. Interaction of HP1γ with BLM, the putative interacting partner identified by mass spectrometry (No. 3 in Supplementary Table 1), was detected by coimmunoprecipitation analysis (Fig. 3A). One of the surrogate proteins implicated in DNA damage response, PCNA, was not identified as an HP1γ-interacting protein by our mass spectrometric analysis (Fig. 2D and E, Supplementary Table 1), yet it was capable to interact with HP1γ in a coimmunoprecipitation experiment (Fig. 3A). One possible explanation for the difference in the interaction results of tandem mass spectrometry and coimmunoprecipitation experiments is that the antibody used affects the efficacy of protein-protein interaction in such a way that the specificity of anti-HP1γ antibody is higher for endogenous HP1γ proteins, but lower for anti-FLAG antibody for the recombinant FLAG-HP1γ proteins is lower. Another possible explanation is the discrepancy between two separate experiments. Without knowing the specificity of these antibodies or the difference between experiments, such issues are difficult to address.

We hypothesized that HP1γ translocates to DNA damage sites, following DNA damage with γH2AX. The HeLa cells were treated with etoposide and stained for HP1γ and γH2AX (damaged DNA sites or repair foci marker). HP1γ co-localized with γH2AX 6 hours after the treatment with etoposide (Fig. 3B), and this enabled us to determine whether HP1γ was a positive or negative regulator in the DNA damage response pathway. Notably, we discovered the accumulation of γH2AX at DNA damage sites in HP1γ-depleted cells (Fig. 3C), suggesting that HP1γ is required for the maintenance of genome integrity, as the deficiency of HP1γ may induce DNA damage. These data indicated that HP1γ functions as a positive regulator in the DNA damage response pathway.

Protein-protein interaction and post-translational modification play key roles in various biological processes. The co-immunoprecipitation affinity purification technique is widely used to investigate protein functions in these processes. In this study, we constructed an interactome network for HP1γ in order to investigate the protein’s novel functions. We identified several proteins as novel binding partners of HP1γ using an affinity purification system. Our results suggest that HP1γ functions in various biological processes (as listed in “Establishment of the interactome of HP1γ” section). BLM (No. 3 in Supplementary Table 1) and XRCC5 (No. 126 in Supplementary Table 1), essential proteins in the DNA damage response pathway, were identified in our purification analysis as proteins that interact with HP1γ (Table 1, Fig. 3D); this implies that HP1γ played a role in regulating the DNA damage response pathway. In the presence of a DNA damage signal, HP1γ can be translocated to DNA damage sites (Fig. 3A). Furthermore, phosphorylation of H2AX was induced in HP1γ-knockdown cells (Fig. 3B), suggesting that HP1γ is involved in the maintanence of genomic integrity by preventing DNA damage, and that HP1γ deficiency could lead to DNA damage. In summary, our interactome analysis demonstrated that HP1γ potentially plays a role in DNA damage response among various biological processes (Table 1, Fig. 3D).