Interaction between an integral protein of the nuclear envelope inner membrane and human chromodomain proteins homologous to Drosophila HP1.

At the nuclear envelope in higher eukaryotic cells, the nuclear lamina and the heterochromatin are adjacent to the inner nuclear membrane, and their attachment is presumably mediated by integral membrane proteins. In a yeast two-hybrid screen, the nucleoplasmic domain of lamin B receptor (LBR), an integral protein of the inner nuclear membrane, associated with two human polypeptides homologous to Drosophila HP1, a heterochromatin protein involved in position-effect variegation. LBR fusion proteins bound to HP1 proteins synthesized by in vitro translation and present in cell lysates. Antibodies against LBR also co-immunoprecipitated HP1 proteins from cell extracts. LBR can interact with chromodomain proteins that are highly conserved in eukaryotic species and may function in the attachment of heterochromatin to the inner nuclear membrane in cells.

At the nuclear envelope in higher eukaryotic cells, the nuclear lamina and the heterochromatin are adjacent to the inner nuclear membrane, and their attachment is presumably mediated by integral membrane proteins. In a yeast two-hybrid screen, the nucleoplasmic domain of lamin B receptor (LBR), an integral protein of the inner nuclear membrane, associated with two human polypeptides homologous to Drosophila HP1, a heterochromatin protein involved in position-effect variegation. LBR fusion proteins bound to HP1 proteins synthesized by in vitro translation and present in cell lysates. Antibodies against LBR also co-immunoprecipitated HP1 proteins from cell extracts. LBR can interact with chromodomain proteins that are highly conserved in eukaryotic species and may function in the attachment of heterochromatin to the inner nuclear membrane in cells.
In higher eukaryotic cells, a portion of the transcriptionally inactive heterochromatin is adjacent to the inner nuclear membrane (1,2). The nuclear lamina is also adjacent to the inner nuclear membrane, but because the lamina is discontinuous (3), the heterochromatin can presumably interact directly with the membrane. During nuclear envelope reassembly at the end of mitosis, vesicles that derive from the inner nuclear membrane also bind to chromosomes in a lamin-independent fashion (4,5). In Saccharomyces cerevisiae, an organism in which a lamina has not been identified, transcriptionally silent regions of the genome are similarly localized to the nuclear envelope, further suggesting a direct interaction between the chromatin and the inner nuclear membrane (6,7).
Several integral proteins have been localized to the inner nuclear membrane that presumably mediate the attachment of the lamina and chromatin. One of these proteins is LBR, 1 or the lamin B receptor, a protein previously shown to bind to B-type lamins (8,9) and to double-stranded DNA (9). LBR was first characterized in birds (8,10) and subsequently in mammals (9,11,12). Human LBR has a nucleoplasmic, aminoterminal domain of 208 amino acids followed by eight putative transmembrane segments (9). We now demonstrate that the nucleoplasmic domain of LBR binds to human chromodomain proteins homologous to Drosophila HP1, a heterochromatin protein involved in position-effect variegation (13,14).

EXPERIMENTAL PROCEDURES
Yeast Two-hybrid Screening-Screening was performed as described (15,16). DNA encoding amino acids 1 to 208 of human LBR was amplified by the polymerase chain reaction using the Gene Amp Kit (Hoffmann-La Roche) and a PCR System 2400 thermocycler (Perkin-Elmer Corp.) with custom oligonucleotide primers (Genset, Paris, France) and a cDNA template (9). Amplified DNA was cloned into the GAL4 DNA binding domain fusion vector pGBT9 (provided by S. Fields, State University of New York at Stony Brook) to produce pGBT9-LBRAT. S. cerevisiae strain Y190 (provided by S. J. Elledge, Baylor College of Medicine, Houston, TX) was co-transformed with pGBT9-LBRAT, and approximately 10 6 recombinants of a HeLa cell cDNA library in the GAL4 activation domain fusion vector pGADGH (Clontech). The positive pGADGH-derived plasmids isolated in this screen were rescued and used to again co-transform yeast strain Y190 with pGBT9-LBRAT to confirm the interactions.
DNA Sequencing and Sequence Analysis-DNA sequencing was performed using the Sequenase Version II Kit (U. S. Biochemical Corp.). Sequences were analyzed using the Wisconsin Package (Genetics Computer Group, Inc., Madison, WI) and the computer facilities of the National Cancer Institute's Frederick Biomedical Supercomputing Center.
Binding Assays Using in Vitro Translated Proteins-For in vitro transcription-translation, cDNA inserts were cloned into pBFT4 (supplied by J. Licht, Mt. Sinai School of Medicine, New York). In vitro translation was performed using the TNT T3 Coupled Reticulate Lysate System (Promega) with [ 35 S]methionine (Amersham). In binding experiments, 20 l of in vitro translated lysate were added to 200 l of binding buffer (150 mM NaCl, 20 mM Na-Hepes (pH 7.4), 10% glycerol, 0.05% Nonidet P-40) with 20 l of glutathione-Sepharose (Pharmacia Biotech Inc.) coupled to equal amounts (3-5 g) of glutathione Stransferase (GST) or GST-LBR fusion protein (9). GST proteins were produced and purified as described (17). Suspensions were incubated at 4°C with rotation for 2 h. After incubation, the Sepharose was washed 5 times with binding buffer, and bound proteins were then eluted with 4% SDS and analyzed by autoradiography of 12.5% SDS-polyacrylamide slab gels (18). In assays using different salt or detergent concentrations, the binding buffers contained the concentrations indicated.
Binding of Proteins from Cell Lysates to GST Fusion Proteins-HeLa cell lysates were prepared as described below (see Immunoprecipitation), and 400 l of lysate were incubated with 5 g of the GST-LBR amino-terminal domain fusion protein or GST coupled to glutathione-Sepharose (20 l). Incubation was for 2 h at 4°C with rotation. After incubation, the Sepharose was washed 5 times with 150 mM NaCl, 20 mM Na-Hepes (pH 7.4), 10% glycerol, 0.05% Nonidet P-40. The bound proteins were then eluted with 4% SDS, separated by SDS-polyacrylamide slab gel electrophoresis (PAGE) and transferred to nitrocellulose sheets for immunoblotting as described (11). Autoantibodies from a patient with scleroderma that recognized HP1 Hs␣ and related polypeptides (19) (provided by W. C. Earnshaw, Johns Hopkins University School of Medicine, Baltimore, MD) were used to detect the proteins. For immunoblots to confirm that the autoantibodies recognized the polypeptides encoded by the clones isolated in the two-hybrid screen, * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM  cDNA inserts were cloned into plasmid QE31, and polyhistidine fusion proteins were expressed and purified according to the manufacturer's instructions (Qiagen, Chatsworth, CA).
Immunoprecipitation-Two 100-mm Petri dishes of 293 T cells were grown to 90% confluency in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 50 units/ml penicillin, 50 g/ml streptomycin, and 2 mM L-glutamine (Life Technologies, Inc.). Cells were washed 3 times with phosphate-buffered saline, harvested by scrapping with a rubber policeman, and collected by centrifugation at 500 ϫ g in 10 mM Tris-HCl (pH 7.4), 1 mM MgCl 2 , 1 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride (PMSF). The cells were then incubated on ice for 15 min and broken in a Dounce homogenizer (30 strokes with the B pestle). Broken cells were centrifuged at 6,000 ϫ g for 20 min at 4°C, and the pellets were resuspended in 400 l of 50 mM Tris-HCl (pH 8.0), 5 mM EDTA, 2% BSA, 1 mM dithiothreitol, 0.2 mM PMSF containing either 0.5% or 1% Nonidet P-40 and then disrupted with a tip sonicator. The suspension was centrifuged at 15,000 rpm for 30 min at 4°C in an Eppendorf 5415C microcentrifuge (Fisher Scientific), and the supernatant was incubated for 4 h at 4°C on a rotating mixer with the desired antibody that was coupled to protein A-Sepharose using the ImmunoPure IgG Orientation Kit (Pierce). Anti-LBR antibodies were autoantibodies from a patient with primary biliary cirrhosis previously shown to recognize this protein (9,20). Control antibodies were from another patient with primary biliary cirrhosis without LBR autoantibodies (20). Protein A-Sepharose-coupled antibodies were washed 5 times with the above buffer with 10% glycerol replacing BSA. Bound proteins were eluted with 4% SDS and divided into two portions. One portion was analyzed by immunoblotting with anti-LBR antibodies (provided by J.-C. Courvalin, Institut Jacques Monod, Paris, France) that were produced by immunizing rabbits with a GST-LBR aminoterminal domain fusion protein (9). The other portion was analyzed using the autoantibodies that recognized HP1 Hs␣ and related polypeptides (19). Immunoblotting was performed as described (11).
Other Materials-Routine chemicals were from Fisher Scientific or Sigma. Enzymes were from New England Biolabs or Promega. 125 I-Protein A was from DuPont NEN Products and nitrocellulose from Schleicher and Schuell.

RESULTS
Using the nucleoplasmic, amino-terminal domain of LBR as bait, we performed a yeast two-hybrid screen of a HeLa cell cDNA library. Screening of approximately 10 6 recombinant clones led to the isolation of 10 that grew on the appropriate selection medium and gave detectable ␤-galactosidase activity. Plasmids isolated from these clones were used to again transform yeast. Only two, termed ATBP115 and ATBP8132, remained positive for ␤-galactosidase activity when co-transformed with the plasmid that encoded the LBR fusion protein but not the DNA binding domain of GAL4 alone.
Sequencing of the cDNA insert of ATBP115 showed that it encoded HP1 Hs␣ (19). Sequencing of the cDNA insert of clone ATBP8132 showed that it encoded the majority of a novel human protein with 65% sequence similarity to HP1 Hs␣ . The ATBP8132 cDNA overlapped with two nucleotide sequences of unknown function in GenBank TM and alignment of the translated sequences with those of homologous polypeptides in Drosophila (13,14,21), mouse (22), and humans (19,22) allowed us to deduce the protein's likely amino terminus. This protein, which we termed HP1 Hs␥ (see below), and HP1 Hs␣ were highly homologous to Drosophila melanogaster HP1 (Fig. 1), a chromodomain protein localized to heterochromatin (13).
The human protein encoded by ATBP8132 is 98% identical in amino acid sequence with mouse chromodomain protein M32 (22). A human protein 100% identical with M31, another mouse chromodomain protein, has been termed HP1 Hs␤ (19,22). HP1 Hs␣ is 80% and 76% similar to M31 and M32, respectively. In keeping with the previous nomenclature used for human HP1 homologues (19), we have termed the protein encoded by ATPB8132 HP1 Hs␥ . HP1-like chromodomain proteins are conserved in a wide range of eukaryotic species (23). In particular, HP1 Hs␥ is 38% identical and HP1 Hs␣ is 29% identical with SWI6 of Schizosaccharomyces pombe (24). Drosophila polycomb, which differs from HP1 for most of its sequence, also contains the chromodomain (25), which spans from amino acids 24 to 60 in HP1 and 20 to 56 in HP1 Hs␣ and HP1 Hs␥ (Fig. 1).
To confirm the interactions detected in the yeast two-hybrid assay, we showed that the amino-terminal domain of LBR bound to HP1 Hs␣ and HP1 Hs␥ in vitro (Fig. 2a). Purified GST-LBR amino-terminal domain fusion protein was coupled to glutathione-Sepharose. HP1 Hs␣ (Fig. 2a, lanes 1-3) and HP1 Hs␥ lacking its first 17 amino acids (Fig. 2a, lanes 4 -6), synthesized in reticulocyte lysates, bound to the LBR fusion protein but not to GST coupled to glutathione-Sepharose in buffer containing 150 mM NaCl. The amount of HP1 Hs␣ that bound to the aminoterminal domain of LBR started to decrease slightly at NaCl concentrations greater than 0.5 M, but significant binding was still observed in buffers containing 1.0 M (Fig. 2b). Binding of HP1 Hs␣ to the amino-terminal domain of LBR was also observed in the presence of up to 1% of the nonionic detergent Nonidet P-40 but the denaturing detergent SDS essentially abolished the interaction (Fig. 2c). Binding of HP1 Hs␣ to the LBR fusion protein was also not influenced by excessive amounts of BSA, non-fat milk, and/or DNA (data not shown).
The LBR amino-terminal domain GST fusion protein also bound to HP1 proteins in extracts of HeLa cells (Fig. 3). Two proteins of approximately 30 kilodaltons, recognized by scleroderma autoantibodies against HP1 Hs␣ and related chromodomain proteins (19), were present in HeLa cell extracts (Fig. 3a,  lane 1). These two polypeptides were retained on glutathione-Sepharose coupled to a fusion protein of GST and the aminoterminal domain of LBR (Fig. 3a, lane 3) but not GST (Fig. 3a,  lane 2). The autoantibodies used to detect the cellular HP1 FIG. 1. Primary structures of HP1 Hs␥ , HP1 Hs␣ , and D. melanogaster (Dm) HP1. Identical or conservative amino acids are shown as white on black. Conservative amino acid substitutions used are: aliphatic hydrophobic, L/I/V/M/A; aromatic hydrophobic, F/Y; basic, R/K/H; acidic, E/D; aliphatic alcoholic, S/T; polar amides, Q/N. The cDNA insert of clone ATBP115 starts at the G of the initiation codon of HP1 Hs␣ and contains the remainder of the downstream sequence including a polyadenylate tail. Clone ATBP8132 starts at the 18th putative amino acid codon of HP1 Hs␥ and extends in the 3Ј direction to a polyadenylate tail. ATBP8132 is 98% identical, over a stretch of 209 nucleotides, with a random overlapping sequence of 345 nucleotides in GenBank TM (accession number Z15820) that extends 136 nucleotides in the 5Ј direction. The first 17 putative amino acids of HP1 Hs␥ are deduced from this sequence, and the position of the first methionine is based on strong homology to other HP1 proteins. The amino acid at position 11 (X) cannot be deduced from the available nucleotide sequence data.
Anti-LBR antibodies also co-immunoprecipitated HP1 proteins from cell extracts (Fig. 4). Antibodies from serum of a patient with primary biliary cirrhosis that recognized LBR (9, 20) immunoprecipitated LBR from extracts of 293 T cells (Fig.  4a, lanes 1 and 3). No detectable LBR was present in immuno-precipitates when control antibodies from another patient with primary biliary cirrhosis were used (Fig. 4a, lane 2). Two polypeptides of approximately 30 kilodaltons that were recog-  3 and 6) coupled to glutathione-Sepharose. Glutathione-Sepharose was then pelleted by centrifugation and washed, and the bound proteins were eluted with SDS, separated by SDS-PAGE, and examined by autoradiography. HP1 Hs␣ and HP1 Hs␥ bound to the LBR fusion protein (lanes 3 and 6) but not to GST (lanes 2 and 5). b, a standard amount of 35 S-labeled HP1 Hs␣ , 10% of which is shown (lane 1), was used in each binding assay. HP1 Hs␣ was incubated with glutathione-Sepharose alone (lane 2) or 3-5 g of GST coupled to glutathione-Sepharose (lane 3) in binding buffer containing 150 mM NaCl. HP1 Hs␣ was also incubated with 3-5 g of LBR amino-terminal domain GST fusion protein coupled to glutathione-Sepharose (lanes 4 -8) in buffers containing the NaCl concentrations indicated above each lane. Glutathione-Sepharose was then washed with buffer containing the indicated NaCl concentration, and the bound proteins were eluted with SDS and examined as above. c, a standard amount of 35 S-labeled HP1 Hs␣ , 10% of which is shown (lane 1), was used in each binding assay. HP1 Hs␣ was incubated with glutathione-Sepharose alone (lane 2) or 3-5 g of GST coupled to glutathione-Sepharose (lane 3) in binding buffer containing 0.05% Nonidet P-40. HP1 Hs␣ was also incubated with 3-5 g of LBR amino-terminal domain GST fusion protein coupled to glutathione-Sepharose in buffers containing detergents at the concentrations indicated above each lane (lanes 4 -8).
Glutathione-Sepharose was washed with buffer containing the indicated detergent type and concentration, and the bound proteins were eluted with 4% SDS and examined as above. Migrations of molecular mass standards are indicated in kilodaltons at the left of each panel.

LBR-HP1 Interaction 14655
nized by anti-HP1 antibodies in cell extracts (Fig. 4b, lane 1) were present in immunoprecipitates obtained with the experimental antibodies (Fig. 4b, lane 3) but not in immunoprecipitates obtained with the control antibodies (Fig. 4b, lane 2). The serum-containing anti-LBR antibodies did not recognize HP1 proteins on immunoblots (Fig. 4c), making it unlikely that the anti-LBR antibodies directly immunoprecipitated HP1 proteins. Serum from both the control and experimental patients also contained antimitochondrial antibodies (Fig. 4c and Ref. 20), a useful internal control for immunoprecipitation, and immunoblotting (data not shown) showed that both immunoprecipitates contained the major mitochondrial autoantigen pyruvate dehydrogenase E2 (26). These experiments strongly suggested that HP1 proteins and LBR were associated in human cells.

DISCUSSION
The present results show that human chromodomain proteins homologous to Drosophila HP1 interact with LBR, an integral membrane protein of the inner nuclear membrane. This interaction can contribute to the association of the heterochromatin with the inner nuclear membrane in higher eukaryotic cells. The specificity of this interaction is demonstrated by the fact that the amino-terminal domain of LBR identified only two clones from approximately 10 6 in a yeast two-hybrid screen, each of which respectively contained the near fulllength sequences of HP1 Hs␣ and HP1 Hs␥ . Direct interactions between HP1 Hs␣ and the amino-terminal domain of LBR also occur in vitro under stringent conditions utilizing various salt and detergent concentrations. A LBR fusion protein can also extract HP1 proteins from cell lysates, and antibodies that immunoprecipitate LBR can co-precipitate HP1 proteins.
In D. melanogaster, HP1 functions as a suppressor of position-effect variegation. If an active genetic locus in Drosophila euchromatin is translocated near constitutive heterochromatin, transcription from the locus is variably repressed leading to variegation of gene expression (27). Position-effect variegation is associated with altered chromatin structure, and Drosophila heterozygous HP1 mutants demonstrate suppression of position-effect variegation presumably because heterochromatin stabilization is defective (14,25,28). Drosophila HP1 has been directly localized to heterochromatin (13,29), as have been HP1 homologues from mammals (30) and fission yeast (31). The interaction of HP1-type chromodomain proteins with LBR may therefore explain, at least in part, the long-observed association of the heterochromatin with the inner nuclear membrane.
Chromatin is a dynamic structure, and heterochromatin packaging changes dramatically at different stages of the cell cycle. Similarly, the nuclear envelope undergoes profound structural changes during mitosis. LBR undergoes phosphorylation catalyzed by p34 cdc2 protein kinase in mitosis when the inner nuclear membrane breaks down into vesicles that dissociate from the lamina and the chromatin (32). It is phosphorylated by different protein kinases in interphase when the mem-brane is associated with these structures (33,34). Phosphorylation of LBR and phosphorylation of HP1 proteins (35) may therefore be responsible for some of the alterations in chromatin organization and nuclear structure which occur at various times during the cell cycle.