INTRODUCTION

The ubiquitin-proteasome proteolytic pathway is important for a wide range of cellular functions including cell cycle control, DNA repair, peroxisome biogenesis, and resistance to heavy metals. In addition to recognizing and eliminating unassembled proteins and disposing of damaged or misfolded proteins, it is also responsible for the degradation of short-lived proteins, a process that is complex and highly regulated (reviewed in Refs. 1-3). The first step in this process is the ligation of the C terminus of ubiquitin to -amino groups of lysine residues within the substrate. After ubiquitination, the substrates are targeted to the 26 S proteasome complex, which contains over 20 distinct subunit activities, including multiple peptidases, ATPases, and a deubiquitinating enzyme. The 26 S proteasome is composed of two functionally distinct protein components, a catalytic 20 S proteasome and a regulatory complex. Controlled proteolysis requires that a regulatory complex be attached to each end of the 20 S proteasome. The regulatory complex is composed of 15 different subunits named according to their size as measured by SDS-polyacrylamide electrophoresis, with S1 being the largest and S15 the smallest. The proteolysis of certain substrates is ATP-dependent and requires polyubiquitination of the substrate.

Assembly of the 26 S proteasome is ATP-dependent, and the regulatory subunits have been found to belong to an ATPase subfamily that includes TBP-1, MSS1, p45, subunit 4 (mts2), CIM5, CIM3, and SUG1 (4-9). MSS1, first identified as a co-activator of TAT-mediated transactivation (10, 11), was later found to be the seventh regulatory subunit of the 26 S proteasome complex (5). Temperature-sensitive mutants of CIM5, the homologue of human MSS1 in S. cerevisiae, progress through the cell cycle normally at permissive temperatures. At restrictive temperatures, they arrest at the G₂/M boundary coincident with the loss of 26 S proteasome activity (8). These observations linked 26 S proteasome activity directly to cell cycle control, specifically during the mitotic phase.

In metazoan cells, cyclin degradation by regulated proteolysis is crucial for M phase progression (reviewed in Refs. 1 and 12). In this stage of the cell cycle, a large protein complex called the anaphase-promoting complex (APC;¹ Ref. 13) starts chromosome segregation and exit from mitosis through targeting anaphase inhibitors and mitotic cyclins for degradation by way of the 26 S proteasome. Although interdependent, this pathway is distinct from the proteolytic CDC34 pathway that participates in the regulated proteolysis of the G₁ cyclins (12). Recognition of the APC substrates for ubiquitination is by way of a 9-amino acid motif referred to as the destruction (D) box. A key event in anaphase progression in yeast is the APC-mediated proteolysis of a non-cyclin substrate (CUT2, fission yeast; PSD1, budding yeast) that is an inhibitor of anaphase initiation (14, 15). Although much is known concerning the identity of the players in APC-mediated ubiquitination, little is understood regarding their regulation or the control of 26 S proteasomes. Nevertheless, it is clear that regulation of proteolysis in M phase is key to its orderly progression.

The retinoblastoma protein (Rb) has an important role in regulating the cell cycle as well as differentiation, development, and the suppression of malignant transformation (16). The isolation of cellular proteins that interact with Rb has provided important information regarding the molecular basis of its function (17-19). One of these interacting proteins (18), originally identified as a partial cDNA, C15, was later named HEC because of its high expression in cancers. During its characterization, HEC was found to be a centromere-associated protein with a critical role in chromosome segregation. When HEC is deregulated in cells by microinjection with specific antibodies, abnormal daughter cells are produced that fail to proceed to the next division and subsequently die (20).

To elucidate the potential role of HEC in M phase progression, the cellular proteins interacting with HEC through its long stretch of leucine heptad repeats were identified using yeast two-hybrid screening. Interestingly, one of these proteins had been previously identified as MSS1, the seventh regulatory subunit of the 26 S proteasome. The ATPase activity of MSS1 is down-regulated by HEC, and degradation of cyclin B by proteasomes is prevented by the addition of HEC in an in vitro assay. Furthermore, ectopic expression of a HEC mutant inhibits mitotic cyclin degradation during mitosis in vivo. These results suggests that the interaction between HEC and MSS1 is biologically significant.

EXPERIMENTAL PROCEDURES

The HEC protein has two major domains: an amino-terminal portion containing a region that interacts with the Rb protein and a carboxyl-terminal region that is enriched with heptad leucine repeats (20). The C-terminal half of the HEC cDNA encoding the leucine repeat-rich region (amino acids 251-618) was ligated into the modified yeast vector, pAS1, and used as bait in a yeast two-hybrid screen of a human lymphocyte cDNA library using previously published procedures (17, 18).

The full-length MSS1 cDNA was subcloned into the BamHI site of pGEX-3X (Pharmacia Biotech Inc.) for expression of GST-MSS1 fusion protein. The MSS1C construct was obtained by subcloning the 0.8-kilobase pair BamHI-XhoI fragment into a modified pGEX-3X vector. Expression of this construct in bacteria generates a GST-MSS1C fusion protein. HEC-GST fusion constructs were obtained by subcloning a 2.3-kilobase pair full-length cDNA fragment from pBKS-C15 (GST-HEC) and a 1.5-kilobase pair fragment BamHI-BglII (GST-15Pst, amino acids 251-618) into pGEX-3X.

Recombinant GST-MSS1, GST-MSS1C, GST-HEC, and GST-15Pst fusion polypetides were expressed in Escherichia coli. Six hours after induction with isopropyl--D-thiogalactopyranoside (0.1 mM) at room temperature, cells from 1 liter of culture were lysed with 10 ml of lysis-250 buffer (50 mM Tris-HCl, pH 7.4, 250 mM NaCl, 5 mM EDTA, 0.1% Nonidet P-40, 50 mM NaF, 1 mM phenylmethylsulfonyl fluoride). The soluble fraction was loaded onto a 0.5-ml glutathione-Sepharose (Pharmacia) column and eluted with 50 mM Tris-HCl, pH 7.5, 120 mM NaCl, 0.1% Nonidet P-40, and 5 mM glutathione (Sigma). The eluted proteins were analyzed by SDS-PAGE, and the gels were stained with Coomassie Blue to determine their concentration and purity.

Partially purified GST-MSS1 fusion protein (95% purity) was used as antigen to immunize mice. After three boosts with 100 µg of fusion protein, specific antiserum against MSS1 was obtained.

The enzymatic hydrolysis of ATP was done as described previously (21). Briefly, purified proteins were incubated for 30 min. at 30 °C in the presence of 1 µCi of [-³²P]ATP (3000 Ci/mmol; NEN Life Science Products) in a 20-µl reaction volume consisting of 20 mM Tris-HCl, pH 7.9, 4 mM MgCl₂, 1 mM dithiothreitol, 50 µg/ml bovine serum albumin, and 10 mM ATP. Reactions were stopped by the addition of EDTA to 50 mM. 2 µl of the reactions were spotted onto polyethyleneimine TLC plates (Sigma) and developed in 0.5 M LiCl, 1 M formic acid (22). The amount of released inorganic phosphate was quantified by PhosphorImager analysis. For the antibody blocking experiment, beads containing GST-MSS1 fusion protein were first incubated with anti-MSS1 antibody or anti-GST antibody for 30 min and then washed with 20 ml of 1 × PBS. The blocked beads were then used for the ATPase assay. In the competition experiments, unlabeled nucleoside triphosphates in the indicated concentrations (µM) were added to the basic ATPase assay.

Human 26 S proteasome preparation was kindly provided by M. Rechsteiner (University of Utah, Salt Lake City, UT). After separation on SDS-PAGE, the proteins were immunoblotted and developed with anti-MSS1 and anti-HEC antibodies as described previously (19).

HEC was translated in vitro using cDNAs and a TNT reticulocyte transcription translation-coupled system (Promega, Madison, WI) according to the supplier's instructions. GST pull-down assays were done as described previously (19).

Human bladder carcinoma cells, T24 (ATCC, Rockville, MD), grown in Dulbecco's modified Eagle's medium, 10% fetal calf serum, were synchronized at G₁ by density arrest in Dulbecco's modified Eagle's medium, 0.5% serum and then released at time 0 by replating in Dulbecco's modified Eagle's medium, 10% fetal calf serum at a density of 2 × 10⁶ cells/10-cm plate. At various time points thereafter (18 h for G₁/S, 22 h for S, 33 h for G₂), cells were harvested. To obtain cells in M phase, nocodazole (0.4 µg/ml) was added to the culture medium for 8 h prior to harvest as described previously (23).

Cells lysed in Lysis 250 buffer were subjected to three freeze/thaw cycles (liquid nitrogen/37 °C), and clarified by centrifugation (10,000 × g, 2 min at room temperature). The supernatants were used for immunoprecipitation as described (24). Briefly, 1 µl of mouse polyclonal anti-C15 antisera or 1 µg of anti-Myc 9E10 (25), were added to each clarified supernatant. After a 1-h incubation, protein A-Sepharose beads were added for another hour. The beads were then collected and washed five times with lysis buffer containing 250 mM NaCl and then boiled in SDS-loading buffer for immunoblotting analysis as described (23). Anti-green fluorescence protein (anti-GFP) (CLONTECH, Palo Alto, CA) was used at a 1:250 dilution for immunoblotting analysis.

The procedures to fractionate cell components were adapted from those previously published (26). Cell lysates prepared from differently synchronized populations of T24 cells were loaded on 5-25% preformed sucrose gradients and centrifuged in an SW41 rotor (Beckman, Fullerton, CA) at 28,000 rpm for 15 h at 4 °C. Aliquots of each fraction were then assayed for the presence of MSS1 and HEC by Western blotting analysis as described above.

Three constructs were used in transfection assays: 1) CHPL-GFP, a modified plasmid derived from a mammalian expression vector containing a Myc-tagged, mutant form of green fluorescence protein (S65T; Ref. 27) (CLONTECH); 2) CHPL-GFP-15PA, containing GFP fused to the N terminus of HEC (amino acids 1-250); and 3) CHPL-GFP-15Pst, containing GFP fused to the C-terminal portion of HEC (amino acids 251-618). Transfection of 1 × 10⁶ cells was done using a conventional calcium phosphate/DNA co-precipitation method (17). The precipitates were removed 12 h after transfection, and the cultures were refed with fresh medium. The cells were then observed under a fluorescence microscope (Axiophot Photomicroscope, Zeiss).

Cells grown on coverslips in tissue culture dishes were washed in phosphate-buffered saline (PBS) and fixed for 30 min in 4% formaldehyde in PBS with 0.5% Triton X-100. After treatment with 0.05% saponin in water for 30 min and extensive washing with PBS, cells were blocked in PBS containing 10% normal goat serum. An overnight incubation with antibody diluted in 10% goat serum at 4 °C was followed by three washes and then by a 1-h incubation with Texas Red-conjugated secondary antibody. Anti-cyclin A and B antibodies (Santa Cruz Biotechnology Inc., Santa Cruz, CA) were used at dilutions of 2 µg/ml. The respective antigens were visualized with goat anti-rabbit IgG conjugated to Texas Red (Amersham Corp.). After washing extensively in PBS with 0.5% Nonidet-P 40, cells were further stained with DAPI (Sigma) and mounted in Permafluor (Lipshaw-Immunonon, Inc., Pittsburgh, PA). Ektachrome P1600 film was used when pictures were taken from a standard fluorescence microscope (Axiophot Photomicroscope, Zeiss).

In vitro translated, [³⁵S]methionine-labeled cyclin B was incubated at 30 °C in 240 µl of degradation mix (final concentration, 33% (v/v) rabbit reticulocyte lysate, 50 mM Tris-HCl (pH 8.0), 5 mM MgCl₂, and 2 mM dithiothreitol) in the presence of 0.1 µM GST, GST-HEC, or GST-15Pst. At different times thereafter, aliquots (30 µl) of the reaction mix were withdrawn for SDS-PAGE and PhosphorImager analysis (Molecular Dynamics, Sunnyvale, CA).

RESULTS

To explore the potential biochemical basis for the biological function of HEC, we sought to obtain clues by searching for proteins with which it interacts. Since leucine heptad repeats are known to be important in protein-protein interactions (28), the C-terminal two-thirds of HEC containing the long stretches of these repeats was used as bait in a yeast two-hybrid screen of a human lymphocyte cDNA library. Among the 16 strongly interacting clones obtained, 12 were identified as cDNA fragments encoding MSS1, a component of subunit 7 of the 26 S proteasome complex (5, 10). Others obtained encoded subunit p45 of the 26 S proteasome (6); Sb1.8, the human homologue of yeast Smc1/Smc2 (29-31); and Nek2 (32), the human homologue of NimA (33), a kinase crucial for the progression of G₂/M phase in Aspergillus nidulans (Table I).

Table I. Cellular proteins that interact with HEC HEC-Aps clones Binding in yeasta In vitro binding Identity Homologue Mutant phenotype 1, 6, 7, 8...b 371.5  ± 19.3 + MSS1 CIM5 G2/M arrest 4 273.3  ± 10.0 + Sb1.8 Smc1/Smc2 G2/M arrest 14 239.4  ± 32.6 + Nek2 NimA G2/M arrest 24 1105.2  ± 159.3 + p45 subunit of 26 S proteasome CIM3 (SUG1) G2/M arrest a -Galactosidase activity. b Total number obtained was 12.

The importance of regulated proteolysis in the cell cycle (8) and, in particular, chromosome condensation during M phase (1, 12) is well established. To further assess the biological significance of the association between HEC and its interacting proteins, we chose to focus first on MSS1. To further confirm the yeast two-hybrid results, an in vitro binding experiment was performed using a GST pull-down assay. HEC was synthesized by a reticulocyte lysate transcription-translation system and tested for its ability to bind GST-MSS1 and GST-MSS1C (amino acids 328-434) fusion proteins. The results showed that the C-terminal portion of MSS1 is sufficient for binding to HEC (Fig. 1A). To determine the region(s) of HEC necessary for an interaction with MSS1, three HEC deletion constructs, each containing a variable number of the leucine heptad repeats, were fused in frame to the GAL4 DNA-binding domain. These constructs were then used in the yeast two-hybrid system to assay for their ability to interact with MSS1. As shown in (Fig. 1B), the second region of leucine heptad repeats located between amino acids 361 and 547 of HEC is sufficient for interactions with MSS1.

MSS1 is the seventh regulatory subunit of the 26 S proteasome complex (5). Since it binds to MSS1 in vitro, HEC could also be a constituent of the regulatory complex of the 26 S proteasome. To test this possibility, we obtained three active fractions of partially purified 26 S proteasomes (kindly provided by M. Rechsteiner; see Fig. 2, lanes 2-4) and assayed for the presence of MSS1 and HEC. Consistent with previous sequencing data from subunit 7 peptides isolated from 26 S proteasomes (5), MSS1 was detected in all three fractions containing 26 S proteasome activity (Fig. 2, lanes 6-8), while HEC was not detected in any of the fractions (Fig. 2, lanes 10-12). However, MSS1 and HEC are both detected in total cellular protein lysate (Fig. 2, lanes 5 and 9). This result suggests that HEC is not present in the 26 S proteasome complex.

Since HEC is not a component of the 26 S proteasome but does interact with the MSS1, a regulatory subunit of the proteasome, it is important to determine the temporal and spatial status of these two proteins in cells. To answer these questions, specific antibodies were used to examine expression during cell cycle progression. HEC is expressed most abundantly during late S to M phase (Fig. 3B, top panel), while MSS1 is expressed throughout the cell cycle (Fig. 3A, middle panel). Co-immunoprecipitation of HEC and MSS1 was detected only during M phase (Fig. 3B, middle panel). To further assess the potential of a HEC/MSS1 interaction, a sedimentation analysis of cellular lysates was performed. In cell lysates prepared from cells synchronized in S phase, the majority of MSS1 was found to be in fractions consistent with its association with the 26 S proteasome complex (Fig. 4A). This fraction is distinct from that containing HEC, which, during this phase, sediments at a lower S value (Fig. 4A). In M phase-synchronized cells, a significant portion of MSS1 co-sediments with HEC in the lower S value fractions (Fig. 4B). These data are consistent with the results in Fig. 3B, which showed that HEC/MSS1 co-immunoprecipitated only during M phase.

A potential function of the HEC/MSS1 interaction is to regulate proteasome activity by modulation of MSS1. MSS1, as well as several other cloned regulatory subunits of the 26 S proteasome (S4, TBP1, SUG1, CIM5, CIM3, and p45), contain ATPase domains (4). Presumably, the putative ATPase activity common to all of these regulatory subunits is an important aspect of their function. Alterations in this activity could be important for regulation of proteasome function. Although MSS1 is presumed to be an ATPase, this activity, to our knowledge, has never been tested. Therefore, to determine if HEC can modulate its ATPase activity, it was first necessary to assay MSS1 and HEC for their ability to hydrolyze ATP.

To test for ATPase activity, bacterially expressed MSS1 and HEC fusion proteins were prepared. For MSS1, a full-length cDNA was subcloned into a modified version of the pGEX-2T vector and expressed as a glutathione S-transferase fusion protein (GST-MSS1). The presence of a single 72-kDa band in an SDS-PAGE analysis (Fig. 5A, lane 1) demonstrates the purity of the fusion protein after glutathione-Sepharose affinity chromatography. To obtain HEC-GST fusion proteins, full-length (GST-HEC) and truncated (GST-15Pst, amino acids 251-618) cDNAs were translationally fused to GST in pGEX-2T, expressed, and purified (Fig. 5A, lanes 2 and 3, respectively). For controls in subsequent experiments, GST was also expressed and purified (data not shown).

As shown in Fig. 5B, GST-MSS1, but not GST-HEC, GST-15Pst, or GST, can hydrolyze ATP. Furthermore, the ATPase activity of GST-MSS1 is inhibited by incubating the purified protein with anti-MSS1 antibody but is not affected by anti-GST antibodies (Fig. 5C). To verify the nucleoside specificity of MSS1's ATPase activity, unlabeled nucleoside triphosphates at the indicated concentrations were added to the ATPase reaction (Fig. 5D). Significant inhibition of [-³²P]ATP hydrolysis by unlabeled ATP in a concentration-dependent manner, but not by GTP, CTP, or UTP (Fig. 5D), indicates that MSS1 predominantly hydrolyzes ATP.

Since MSS1, but not HEC, has ATPase activity, we next examined if HEC regulates this activity through its association with MSS1. The effect of HEC was examined by including varying amounts of GST, GST-HEC, or GST-15Pst in the ATPase reactions. As shown in Fig. 6, GST alone has very little effect on MSS1's ATPase activity, while GST-HEC and GST-15Pst showed dose-dependent inhibition of MSS1's ATPase activity. Since GST-15Pst contains the leucine heptad repeat domain with which MSS1 interacts (amino acids 251-618), the data suggest that this region of HEC is sufficient to inhibit the ATPase activity of MSS1.

Since HEC association with MSS1 is only detectable during M phase of the cell cycle and because it can negatively regulate the ATPase activity of MSS1, we next investigated the possibility that HEC can regulate the degradation of mitotic cyclins. As shown in Fig. 7, cyclin B synthesized by rabbit reticulocyte lysates can be degraded in vitro using a proteasome degradation assay. The addition of GST had no influence on cyclin B degradation, while the addition of GST-HEC or GST-15Pst significantly inhibited cyclin B degradation. These results suggest that HEC may down-regulate proteasome activity.

We next wished to address the possibility that HEC may regulate the degradation of mitotic cyclins in cells. This was hampered by the deleterious effects of HEC overexpression in cells. An alternative approach to circumvent this problem was to express a truncated form of HEC lacking an NLS. It was postulated that, under conditions of cytoplasmic accumulation of HEC, it should be possible to observe the effects after nuclear envelope breakdown. Since the heptad repeats in HEC that bind to MSS1 were sufficient for inhibition of MSS1 activity in vitro, we sought to determine whether expression of this region of HEC alone, which lacks the NLS, could influence cell division and/or mitotic cyclin degradation. Two constructs using two separate regions of HEC fused to GFP were used; GFP-15PA contains the N-terminal region only (amino acids 1-250), and GFP-15Pst contains the entire series of leucine heptad repeats (amino acids 251-618) (Fig. 8A). The GFP parent vector served as a control.

Transfection of these three constructs into Rb-negative Saos-2 cells resulted in expression of the corresponding proteins, which could be detected by Western blotting using anti-GFP and anti-Myc tag antibodies as probes (Fig. 8B, lanes 2-4). Twenty-four hours after transfection, cells were observed directly using fluorescence microscopy. Expression of GFP was detected in nuclei and cytoplasm (Fig. 8C, part b), while GFP-15PA was observed only in the nuclei (Fig. 8C, part d) and GFP-15Pst was found only in the cytoplasm (Fig. 8C, part f), consistent with loss of the potential NLS. In contrast to HEC inactivation with anti-HEC antibodies, no abnormal daughter cells were observed under conditions of HEC cytoplasmic accumulation. However, when these cells were immunostained with antibodies specifically recognizing cyclin A or cyclin B, neither cyclin A (Fig. 8D, part c, arrowhead) nor cyclin B (Fig. 8E, part c, arrowhead) could be detected in the untransfected cells after telophase. In contrast, in pre- or metaphase cells, both cyclin A and B were detectable (Fig. 8, D and E, labeled with asterisk). In cells expressing GFP-15Pst, cyclin A (Fig. 8D, part f, arrow) and cyclin B (Fig. 8E, part f, arrow) were detected even after telophase. These results strongly suggest that overexpression of the C-terminal region of HEC can inhibit the degradation of mitotic cyclins during M phase.

DISCUSSION

Previous studies suggest that HEC is essential for normal mitosis in mammalian cells (20). Cells injected with antibodies specific to HEC demonstrate multiple abnormalities with regard to mitosis: 1) chromosomes condense but fail to segregate properly; 2) no metaphase plates are observed; 3) spindles are disorganized in relation to the centromeres; 4) spindles fail to assume the proper orthogonal orientation to chromatids; and 5) cells are able to cytokinese, but chromosomes are separated haphazardly into abnormal, nonviable daughter cells. These studies point to an important role of HEC in cell division.

Inactivation of HEC apparently disrupts the delicate regulation of early M phase events. Since a common phenotype of HEC inactivation is a disorganized metaphase plate, HEC action probably begins early in mitosis, perhaps during prophase. The potential substrates for regulating this part of M phase are unknown at the present time. The mechanism by which HEC functions prior to and during mitosis remains to be elucidated. The localization of a portion of HEC at the centromere/kinetochore (20) indicates that the protein may be involved in spindle attachment to chromosomes during prophase and indirectly involved in subsequent chromosome movement (34). However, the lack of a signature tubulin-binding domain in the HEC molecule argues against direct microtubule attachment.

The association of HEC with a mitosis-specific kinase and with several subunits of the 26 S proteasome (Table I) suggests potential ways by which HEC may influence chromosome congression, separation, or segregation. In this role, HEC may function as an adaptor molecule through its long leucine heptad repeats, much like the Skp1 protein in budding yeast (35, 36). HEC may alter the conformation of multiple-subunit complexes and bring together a number of proteins, including components of the mitotic spindle or kinetochore, components of the 26 S proteasome, kinases or phosphatases, and checkpoint monitors. Regulatory events during chromosome alignment and separation are rapid and precisely timed, and they are likely to be profoundly disturbed without appropriate coordinating adaptor molecules. This concept is consistent with the recent finding that the APC, composed of at least seven distinct proteins, is required for both chromosome segregation and exit from mitosis (37, 38).

The dynamics of the association of MSS1 with different complexes are clear from the results presented here. HEC is not a component of the 26 S proteasome. It is, however, found in association with MSS1 as part of a smaller complex that appears in the late S to M phase. One explanation of how HEC might regulate mitotic cyclin degradation is that it binds to MSS1 and somehow inhibits its assembly into the 26 S proteasome. Alternatively, HEC may serve as a link between E3/APC/cyclosome (12) and the 26 S proteasome to mediate substrate specificity for controlled proteolysis.

In previous experiments, HEC inactivation by antibody injection leads to abnormal mitosis that includes many obvious defects discussed above. Overexpressing a truncated, cytoplasmic form of HEC does not produce exactly the same phenotype in cells. However, the common phenotype observed under conditions of antibody injection and mutant overexpression is that cells progress through cytokinesis, but daughter cells fail to enter the next division and die. Apparently, mutant HEC lacks a dominant negative effect and may fail to form dimers with the wild-type HEC. Since HEC is found in association with a portion of MSS1 separate from the 26 S proteasome specifically in M phase, it is likely that the overexpressed mutant HEC binds to MSS1 during the time when the nuclear membrane breaks down. Consistent with its negative influence on cyclin B degradation in vitro, overexpression of this mutant HEC did result in an abnormal nuclear accumulation of cyclin A and B in cells as late as telophase. Although the precise mechanism for this phenotype is yet to be elucidated, it is likely that mutated HEC disrupts the dynamics of MSS1 availability for 26 S proteasome assembly. In this regard, the N-terminal region of HEC may have important regulatory elements for HEC function. Indeed, it is known to contain at least three recognizable elements: a putative NLS, a Nek2 phosphorylation site (39), and a region for specific interaction with Rb.² These elements may provide an additional mechanism for regulating the quantity and/or quality of HEC during normal progression of the cell cycle.

Our working model envisions HEC as an adaptor that inhibits the degradation of mitotic cyclins, thus allowing their accumulation during early M phase. Upon receipt of an unidentified signal, HEC repression is relieved, M phase cyclins are degraded, sister chromatid separation ensues, and cell division is completed. This model as well as the consequence of other HEC interactions are currently under investigation to more fully understand the possible role of HEC in mitosis.