Association of NASP with HSP90 in Mouse Spermatogenic Cells STIMULATION OF ATPase ACTIVITY AND TRANSPORT OF LINKER HISTONES INTO NUCLEI*

NASP (nuclear autoantigenic sperm protein) is a linker histone-binding protein found in all dividing cells that is by the cell cycle and in the nucleus linker histones not bound to DNA are bound to In mouse spermatogenic cells tNASP binds the testis-specific linker histone H1t. Utilizing a cross-linker, 3,3 (cid:1) -dithio-bissulfosuccinimidyl propionate, and mass spectrometry, we have identified HSP90 as a testis/embryo form of NASP (tNASP)-binding partner. In vitro assays demonstrate that the association of tNASP with HSP90 stimulated the ATPase activity of HSP90 and increased the binding of H1t to tNASP. HSP90 and tNASP are present in both nuclear and cytoplasmic fractions of mouse spermatogenic cells; however, HSP90 bound 96–1099) were prepared as described previously (9). Rabbit antisera to full-length recom- binant NASP (nucleotides 92–2405) were made by Bethyl Laboratories (Montgomery, TX). Somatic linker histones from calf thymus were purchased from Roche Applied Science. Construction of Expression Vectors— The entire coding sequence of mouse tNASP (nucleotides 92–2406, GenBank TM accession number AF034610) was amplified from mouse testis Quick-clone cDNA (Clon-*

to DNA are bound to the histone-binding protein NASP (8). NASP is a histone-binding protein found in all dividing cells in either a somatic/embryo or testis/embryo (tNASP) form (9 -11). Overexpression of tNASP affects progression through the cell cycle, and Alekseev et al. (8) recently postulated that a dynamic equilibrium exists between H1-NASP complexes and H1 bound to DNA. Previous studies on NASP (9) demonstrated that in vivo the somatic linker histones H1a-H1e (see Ref. 12 for mouse histone H1 nomenclature) were co-precipitated with NASP from somatic cell lysates, indicating that NASP is not selective for H1 subtypes.
Of the several different subtypes of H1 histones, H1a, b, c, d, and e are found in most cells, whereas H1t is restricted to the testis. H1t is expressed during spermatogenesis, representing as much as 50% of the total H1 histones in pachytene spermatocytes (13)(14)(15), and is preceded by tNASP expression in preleptotene spermatocytes. Although their exact function during spermatogenesis is unknown, gene-targeting experiments (16) have shown that the absence of H1t and H1a in double null mice does not compromise spermatogenesis because compensation by H1c, H1d, and H1e most likely maintained normal spermatogenesis despite a reduced H1/nucleosome ratio (16). Under these conditions the major H1 histone binding activity of tNASP could have shifted from H1t to H1c, H1d, and H1e in double null mice. Unfortunately neither the relationship between tNASP, H1t, and other proteins nor the ability of tNASP to transport H1s into the mammalian cell nucleus has been characterized. Therefore we have studied tNASP in spermatogenic cells to understand more about the role of NASP-H1t complexes and now report that H1t binds tNASP in mouse germ cells in a complex in the cytoplasm that contains HSP90, which facilitates the initial loading of linker histones to tNASP. Moreover, tNASP can transport both H1t and somatic linker histones into nuclei.

EXPERIMENTAL PROCEDURES
All of the chemicals and reagents used in this study were molecular biology grade. The restriction enzymes were purchased from Roche Applied Science. Purification of plasmid DNA and PCR products were carried out using QIAprep Miniprep and QIAquick PCR purification kits (Qiagen), and sequencing was performed at the University of North Carolina at Chapel Hill automated sequencing facility. Rabbit antihistone H1 (FL-219) polyclonal antiserum and mouse monoclonal IgG 2a anti-HSP90 (F-8) antiserum were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-H1t antibody was a gift from Dr. M. Meistrich (Department of Experimental Radiation Oncology, M. D. Anderson Cancer Center, University of Texas, Houston, TX). Rabbit antisera to the N terminus of NASP (nucleotides 96 -1099) were prepared as described previously (9). Rabbit antisera to full-length recombinant NASP (nucleotides 92-2405) were made by Bethyl Laboratories (Montgomery, TX). Somatic linker histones from calf thymus were purchased from Roche Applied Science.
Construction of Expression Vectors-The entire coding sequence of mouse tNASP (nucleotides 92-2406, GenBank TM accession number AF034610) was amplified from mouse testis Quick-clone cDNA (Clon-* This work was supported by NICHD, National Institutes of Health through Cooperative Agreement U54HD35041 as part of the Specialized Cooperative Centers Program in Reproductive Research. 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. ‡ To whom correspondence should be addressed: tech, Palo Alto, CA) as described previously (8). A tNASP deletion mutant (NASP-⌬TPR; nucleotides 92-1590), that lacked all tetratricopeptide repeats (nucleotides 1591-1891) was PCR-amplified, cloned into pCR® T7/CT-TOPO vector (Invitrogen), expressed, and purified identically to the full-length construct (8). Preparation of the NLS deletion mutant (NASP-⌬NLS) has been described previously (8).
Isolation, Separation, and Short Term Culture of Spermatogenic Cells-Isolation, separation, and short term culture of spermatogenic cells were carried out as described previously (17,18). The cells were grown in plastic flasks for 16 -18 h to allow Sertoli cells and other somatic cells to adhere to the bottom of the flask, and spermatogenic cells were collected from the supernatant.
Harvested mouse spermatogenic cells were washed in phosphatebuffered saline and Earle's balanced salt solution (5.3 mM KCl, 117 mM NaCl, 26 mM NaHCO 3 , 1 mM NaH 2 PO 4 , and 5.6 mM glucose) and resuspended in 10 volumes of RBS buffer (10 mM Tris-HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl 2 ) containing 1 mM phenylmethylsulfonyl fluoride and protease inhibitor mixture (catalog number 8340; Sigma). The nuclei were prepared essentially as described (19). After incubation on ice for 10 min, the cells were transferred to a prechilled glass Douncetype homogenizer and subjected to 10 -12 quick strokes of the pestle. The nuclei were removed from the cytoplasmic fraction by centrifugation at 1000 ϫ g for 3 min, washed with phosphate-buffered saline, and sonicated for 20 s. The supernatant (nuclear fraction) was cleared by centrifugation for 10 min at 1000 ϫ g. Microscopic examination of the nuclei and Western blotting (staining for histones) of nuclear and cytoplasmic fractions confirmed their separation.
In Vitro Nuclear Transport Assay-The nuclear import assay was performed on permeabilized HeLa cells as described previously (20). The cells were grown on 22 ϫ 22-mm glass coverslips in six-well plates (Corning, Corning, NY) and washed in ice-cold import buffer (20 mM HEPES, pH 7.3, 110 mM potassium acetate, 2 mM magnesium acetate, 1 mM EGTA, 2 mM dithiothreitol, 1 g/ml each of aprotinin, leupeptin, and pepstatin). The cells were permeabilized in ice-cold import buffer containing 40 g/ml digitonin (40 mg/ml stock solution in dimethyl sulfoxide; Calbiochem, San Diego, CA) for 8 min. Digitonin containing buffer was replaced by ice-cold import buffer, and the cells were washed in several changes of import buffer for 30 min with gentle rocking. The complete import mixture contained 50% (v/v) cytosol (the cytoplasmic fraction of HeLa cells centrifuged at 60 000 ϫ g; total protein concentration was 10 mg/ml), 1 mM adenosine-5Ј-triphosphate, 5 mM creatine phosphate, 20 units/ml creatine phosphokinase (Calbiochem), and 10 M biotin-labeled histone H1t. Mouse tNASP or NASP-⌬NLS was added to a final concentration of 5 M. The coverslips were inverted over 150 l of complete import mixture on a sheet of parafilm in a humidified plastic box and incubated at 37°C for 30 min. The transport reaction was stopped by fixation in 3% formaldehyde for 15 min. The coverslips were incubated in Texas Red Avidin D (Vector Laboratories, Burlingame, CA) for 1 h, briefly washed in phosphate-buffered saline, and incubated in 4Ј,6Ј-diamino-2-phenylindole nucleic acid stain (Molecular Probes, Eugene, OR) for 2 min. The coverslips were mounted in ProLong Antifade (Molecular Probes) and examined with a Zeiss Axiophot as described previously (21).
Chromatography-Immunoaffinity chromatography was carried out as described previously (8). For the preparation of NASP-enriched fractions from testis, lysates of 50 mouse testes (Pel-Freez® Biologicals, Rogers, AR) were prepared in 2 ml of MPER reagent (Pierce) with Protease (catalog number 8340; Sigma) and Proteasome (catalog number 539160; Calbiochem) inhibitors added. The cells were vortexed thoroughly and rocked for 10 min at room temperature. Debris was removed by centrifugation (10 min, 12,000 rpm), and the supernatant was applied to a 21 ϫ 600-mm Bio-Sep SEC-S 3000 size exclusion HPLC column equipped with a 75 ϫ 21.2-mm precolumn (Phenomenex, Torrance, CA) in 10 mM phosphate buffer, pH 7.0, 252 mM NaCl. 1650 l were injected per run at a flow rate of 7 ml/min, and fractions were collected every 0.25 min for 27 min. 500-l aliquots of the collected fractions were concentrated in Amicon (Beverly, MA) microcentrifuge filters (molecular weight cut-off ϭ 10,000 Daltons), boiled with SDS sample buffer containing ␤-mercaptoethanol, and analyzed by Western blotting. NASP-positive fractions were pooled (see Fig. 2A).
Mass Spectrometry Identification of tNASP-binding Partners-Water-soluble cross-linker 3,3Ј-dithiobis-(sulfosuccinimidylpropionate (DTSSP; 2 mM; Pierce) with a spacer arm length of 12 Å was added to the Bio-Sep SEC-S 3000 column NASP pooled fraction (described above), incubated (at room temperature for 30 min), and analyzed by nonreducing SDS-PAGE. Western blots probed with anti-NASP antibody identified a high molecular weight NASP-positive band. A strip of  ATPase Assay-The amount of ATP in solution was detected by the ENLITEN® ATP Assay System Bioluminescence Detection Kit (Promega). Bioluminescence was measured on a TD-20/20 Luminometer (Turner BioSystems, Sunnyvale, CA). The assays were performed in 40 mM HEPES, pH 7.5, 5 mM MgCl 2 , and 10 Ϫ8 M ATP (Calbiochem). Typical protein concentrations were 0.2 M for HSP90 and albumin and 0.2 M for tNASP unless otherwise indicated. The temperature was set to 37°C. Background (nonspecific) readings were subtracted from the total activity. Relative light units were normalized for comparing the results from different experiments. As ATPase activity increases the relative light units decrease. Significance was determined by the Student's t test.
Co-immunoprecipitation-For binding experiments reverse phase HPLC purified H1t was biotin-labeled and mixed with recombinant tNASP (10 g) in a 500-l volume of 40 mM HEPES, pH 7.7, 5 mM MgCl 2 and incubated for 1 h at 37°C with gentle shaking. As required by the experiment, HSP90 (2 g) and/or ATP (1 mM) were added. The precipitations were carried out with affinity-purified rabbit anti-recombinant NASP antibodies. Antibody-protein complexes were collected by Ultra Link® immobilized protein A/G (Pierce) and eluted by boiling in SDS-PAGE sample buffer containing ␤-mercaptoethanol. The samples were separated in 10 -20% SDS-polyacrylamide gels (Bio-Rad) by electrophoresis, and the amount of co-immunoprecipitated H1t was calculated from background-subtracted images using Gel Expert software (Nucleotech, San Carlos, CA).

Association of H1t
Histones with tNASP-Isolation of NASP from mouse testis lysates by affinity chromatography with anti-NASP antibodies resulted in co-purification of H1 histones. Previous studies on NASP (9) demonstrated that the somatic linker histones were co-precipitated with NASP. The presence of linker histone H1t was identified by probing Western blots with specific H1t antibody (Fig. 1, lane 3). Western blot analysis of the H1t co-purified with tNASP and H1t histones extracted from the mouse testis and isolated by reverse phase HPLC indicated that they had identical SDS-PAGE profiles (Fig. 1, lane 4).
Identification of tNASP-binding Partners-To identify additional binding partners that may be complexed with tNASP in mouse germ cells, we used size exclusion chromatography and gel electrophoresis to isolate tNASP complexes. tNASP complexes were identified by SDS-PAGE and Western blot analysis with anti-NASP antibodies following size exclusion HPLC separation of a mouse testis lysate on a Bio-Sep SEC-S 3000 column ( Fig. 2A). NASP complexes were pooled and crosslinked with DTSSP. Nonreducing SDS-PAGE analysis of the cross-linked complexes indicated the appearance of an ϳ200-kDa NASP-positive band (Fig. 2B, asterisk). This cross-linked complex was cleaved by incubation of the gel in dithiothreitol, and the gel lane was subsequently placed horizontally and electrophoresed in the second dimension. Immediately under the reduced NASP complex two major protein-staining bands were revealed (Fig. 2C, bands 1 and 2).
Initial identification of band 1 by MALDI-TOF (Fig. 3A) and additional confirming MALDI-TOF/TOF analysis of its peaks 1485.82 and 1187.67 (Fig. 3, B and C) identified this band as an endoplasmic reticulum protein 99 precursor (accession number A29317; tumor rejection antigen gp96). This protein belongs to the heat shock protein 90 family and has a histidine kinase-like ATPase domain. Band 2 (Fig. 2C) was identified as mouse heat shock protein 90␣ (accession number P070901; HSP86-1, tumorspecific transplantation 86-kDa antigen) with MALDI-TOF/TOF analysis of peaks 1264.59 and 1513.77 (Fig. 4). Heat shock protein 90 is a molecular chaperone with ATPase activity.
Affinity chromatography with anti-NASP antibodies confirmed the interaction between tNASP and HSP90 in mouse testis lysates (Fig. 5A). Moreover, as expected from the crosslinking studies of the HPLC (Bio-Sep SEC-S 3000 column) tNASP complexes, HSP90 was present in significant amounts (Fig. 5A). The distribution of tNASP and HSP90 in nuclear and cytoplasmic compartments was also studied to confirm the interaction between tNASP and HSP90 in mouse germ cells. As

tNASP-H1t in Germ Cells
shown in Fig. 5B, tNASP and HSP90 appear unequally distributed between nuclear and cytoplasmic fractions of mouse spermatogenic cells. Somatic NASP was present in small amounts in both nuclear and cytoplasmic fractions (Fig. 5B) and probably represents a contaminant from nonspermatogenic cells in testis. The quality of the nuclear and cytoplasmic preparations was monitored by the presence of histone H1 staining (Fig. 5C). Affinity chromatography with anti-NASP antibodies confirmed that the nuclear and cytoplasmic fractions of mouse germ cells contain tNASP (Fig. 5D, lanes 1 and 2). However, HSP90 co-purified with tNASP only from the cytoplasmic fraction (Fig.  5D, lane 4) and HSP90 alone did not bind to the anti-NASP affinity column (Fig. 5D, lane 5).
ATPase Activity of HSP90 Is Stimulated by tNASP-Nonactivated mammalian HSP90 has very low ATPase activity (23). The HSP90 ATPase activity, in the presence of ATP, was not significantly different from ATP alone or ATP ϩ tNASP alone
The presence of a TPR domain has been reported from a variety of proteins to interact with HSP90 (24). NASP contains TPR domains; therefore we tested a mutant tNASP lacking the TPR sites (NASP-⌬TPR). When the HSP90-ATP reaction mixture was incubated with NASP-⌬TPR, the HSP90 ATPase activity was strongly activated, although somewhat lower in activity than with intact tNASP (Fig. 6C). The tNASP TPR domains are apparently not required for the activation of HSP90 ATPase activity.
H1t Binding in the Presence of HSP90 -Previous studies demonstrated that NASP binds somatic linker histones (9), and in the present study we have demonstrated that tNASP binds mouse testis-specific histone H1t. Because NASP stimulates HSP90 ATPase activity, we tested whether or not NASP binding H1t would increase in the presence of HSP90 and ATP. Fig.  7 demonstrates that the binding of H1t to NASP significantly increases in the presence of HSP90 and ATP (p Ͻ 0.04).
Nuclear Transport of Linker Histones by tNASP-Protein nuclear import requires soluble cytoplasmic factors (20), and linker histones may use importin ␤ and importin 7 heterodimers for nuclear transport (25). However, tNASP has both a functional nuclear localization signal (10) and functional linker histone-binding sites (9). Therefore we tested the ability of tNASP to transport biotin-labeled H1t and somatic H1 histones into the nucleus utilizing an in vitro nuclear transport assay (20). As shown in Fig. 8 (A, B, and G) and Table I and as described previously (25), permeabilized HeLa cell nuclei do not transport H1 histones into the nucleus without the presence of cytoplasmic factors. Strikingly, tNASP (Table I) can completely replace the cytoplasmic factors and transport both H1t (Fig. 8C) and somatic H1 histones (Fig. 8H) into the nucleus. Transport by tNASP does not proceed at 4°C (Fig. 8D), in the absence of ATP (Fig. 8E), or in the absence of the nuclear localization signal (Fig. 8F), indicating that transport is an energy-dependent process that requires the NLS. DISCUSSION In this study we have demonstrated that the histone-binding protein tNASP binds the testis-specific linker histone H1t and that the association of tNASP with HSP90 (identical to HSP86-1; accession number P07901) stimulated the ATPase activity of HSP90 and the binding of H1t to tNASP. We have shown that tNASP and HSP90 are present in both nuclear and cytoplasmic fractions of mouse germ cells; significantly HSP90 is bound to NASP only in the cytoplasm, leading to the conclusion that one function of HSP90 may be to chaperone the proper folding of NASP to bind linker histones. The timing of the stage-specific expression of HSP90, tNASP, and H1t supports the conclusion that the tNASP-H1t-HSP90 complex forms in the cytoplasm because HSP90 is expressed beginning on day 10 postpartum and approximately 2 days before the rapid increase in tNASP expression (Ref. 26; Mammalian Reproductive Genetics Data Base, mrg.genetics.washington. edu/). Expression of HSP90 and tNASP precede the expression of H1t in pachytene spermatocytes (13)(14)(15).
The activation of HSP90 ATPase activity by tNASP was similar to that reported for the specific activator Aha1 in yeast (27) and C143 in human cells (28). Surprisingly tNASP activation of HSP90 did not depend upon binding to any of the three TPR sites found in the C terminus of NASP. TPR sites have previously been reported (29) as protein-protein interaction sites between heat shock proteins and a number of functionally unrelated proteins. Our observation that NASP-HSP90 complexes occur only in the cytoplasm implies that there is a mechanism to release NASP from the complex. HSP90 can autophosphorylate serine (30) or threonine (23), and this has been proposed to control chaperone binding (31)(32)(33)(34); therefore phosphorylation of HSP90 may regulate tNASP binding (35).
In addition to HSP90 ATPase activation and increased H1t binding, this study demonstrated that tNASP transports H1t as well as somatic H1 histones into the nucleus. Linker histones have not been reported to be free in the cytoplasm and with the help of HSP90 ATPase activity are likely to be immediately bound by tNASP upon completion of their synthesis in the cytoplasm. Released from the HSP90 complex, tNASP-H1t would then translocate to the nucleus using the NASP NLS, which has been shown to be necessary and sufficient for NASP transport into the nuclei of Xenopus oocytes (10). Linker histones do not have an NLS; however H1 histones can translocate into the nucleus in the presence of a cytoplasmic factor(s) (36) or importin ␤ and importin 7 heterodimers (25), processes that are both temperature-and ATP-dependent. Similarly tNASP, in the absence of any other cytoplasmic factors, transports linker histones into the nucleus in an energy-and NLS-dependent manor. Consequently we hypothesize that after the synthesis of linker histones in the cytoplasm, they are bound to a complex containing NASP and HSP90, whose ATPase activity is stimulated by binding NASP. NASP-H1 is subsequently released from the complex and translocates to the nucleus where the H1 is released for binding to the DNA (8).