Distinct functions of the unique C terminus of LAP2alpha in cell proliferation and nuclear assembly.

The non-membrane-bound lamina-associated polypeptide 2 isoform, LAP2alpha, forms nucleoskeletal structures with A-type lamins and interacts with chromosomes in a cell cycle-dependent manner. LAP2alpha contains a LEM (LAP2, emerin, and MAN1) domain in the constant N terminus that binds to chromosomal barrier-to-autointegration factor, and a C-terminal unique region that is essential for chromosome binding. Here we show that C-terminal LAP2alpha fragment efficiently bound to mitotic chromosomes and inhibited assembly of endogenous LAP2alpha, nuclear membranes, and lamins A/C in in vitro nuclear assembly assays. Full-length recombinant LAP2alpha, which bound to chromosomes, and N-terminal fragment, which did not bind, had no effect on assembly. This suggested an essential role for the LAP2alpha C terminus in chromosome association and for the N-terminal LEM domain in subsequent assembly stages. In vivo analysis upon transient expression of GFP-tagged LAP2alpha fragments confirmed that, unlike the N-terminal fragment, the C-terminal fragment was able to bind to chromosomes during mitosis, if expressed weakly. At higher expression levels, C-terminal LAP2alpha fragment and full-length protein led to cell cycle arrest in interphase and apoptosis, as shown by fluorescence-activated cell sorter analysis, time lapse microscopy, and BrdUrd incorporation assays. These data indicated distinct functions of LAP2alpha in cell cycle progression during interphase and in nuclear reassembly during mitosis.

In higher eukaryotic cells, lamins and lamin-binding proteins are major determinants of nuclear integrity and function. They are well known as major components of the nuclear envelope, but there is also increasing evidence for intranuclear lamin structures (1)(2)(3)(4). At the periphery lamins and lamin-binding inner nuclear membrane proteins form the lamina, a fibrous network underlying the inner nuclear membrane. In the nuclear interior lamin structures were detected on filaments and branched networks (5)(6)(7), as well as in speckles and DNA replication sites (8 -10), and were shown by fluorescence recovery after photobleaching (FRAP) analysis to include highly dynamic and stable complexes depending on the cell cycle stage (11).
Lamins are type V intermediate filament proteins and are grouped into constitutively expressed B-type lamins and developmentally regulated A-type lamins (12). Lamin-binding proteins in the nuclear lamina and the nuclear interior include several protein families and/or types of proteins in higher eukaryotes such as the inner nuclear membrane proteins, lamin B receptor, emerin, and MAN1, three isoforms of lamina-associated polypeptide 1 (LAP1), 1 and several isoforms of LAP2 (1,3). Up to six LAP2 isoforms derive from a single gene by alternative splicing in mammals (13,14) and various isoforms have been described in Xenopus (15,16). The best characterized LAP2 isoforms are the inner nuclear membrane protein LAP2␤ and the nucleoplasmic protein LAP2␣, which are identical in their N-terminal 187-amino acid constant region but differ in their C termini. While LAP2␤ binds to B-type lamins at the nuclear periphery (17,18) and was suggested to regulate nuclear lamina growth (19), LAP2␣ specifically interacts with A-type lamins within the nuclear interior as part of a detergent/salt-resistant nucleoskeletal structure (20,21).
The molecular and cellular functions of lamins and lamin complexes remain unclear, although functional disruption of lamins in Drosophila (22) and Caenorhabditis elegans (23) revealed that they are essential for viability. In mice, targeted disruption of A-type lamins caused muscular dystrophy, loss of adipose tissue, and early death (24), whereas mutations in the human genes for lamin A and emerin were linked to inherited forms of muscular and lipo-dystrophy (25)(26)(27)(28). Polymerized lamins have been suggested to serve as the structural backbone for the nucleus defining nuclear shape and allowing DNA replication (2,10,29,30). Furthermore, lamins and lamin-binding proteins are implicated in the structural organization of chromatin by binding to DNA and to chromosomal proteins (3). Nuclear structure is completely disassembled in higher eukaryotes during cell division, and lamins and lamin-binding proteins dissociate from chromosomes during metaphase and assemble around decondensing chromosomes during the formation of daughter nuclei in late anaphase/telophase (11,17,20,31,32). Therefore, the association of lamins and lamin-binding proteins with chromosomes has been implicated in targeting nuclear envelope components to newly forming nuclei and in the structural organization of chromatin following mitosis.
Studies by several groups have revealed various domains in LAP2 isoforms, which exhibit different chromosome binding properties (see Fig. 1). All LAP2 proteins share a highly con-served LEM domain (amino acid 111-152) (33) in their constant region with the inner nuclear membrane proteins emerin and MAN1. The LEM domain has been identified as a core binding region for the DNA cross-bridging protein, barrier-toautointegration factor (BAF) by yeast two hybrid and biochemical analysis (34,35) and by structural studies (36 -38). Moreover, the N-terminal 85 residues of the LAP2 constant region, which contains a LEM-like motif that was shown by structural analysis to bind DNA (38), was found to associate with chromosomes in vitro (18). In addition to the common chromosome binding domains in the LAP2 constant region, a DNA binding region was described in the LAP2␤-specific region (39), and we have recently identified a domain in the unique C terminus of LAP2␣ that is essential and sufficient for chromosome association of LAP2␣ during nuclear reassembly (40). We have also shown that LAP2␣ associates with chromosomes very early during nuclear reassembly after sister chromatid separation, clearly before LAP2␤-containing membranes and the bulk of lamins assemble around chromosomes (40). Interestingly, the early association of LAP2␣ with chromosomes did not require the N-terminal BAF and DNA interaction domains but was mediated by an ϳ350-amino acid-long region in the LAP2␣ unique C terminus (Fig. 1). Using LAP2␣ fragments containing either the N-terminal common or the C-terminal unique chromatin interaction domains in in vitro nuclear assembly assays, we show here that both regions are essential for nuclear assembly in a timely coordinated fashion, the C-terminal domain being important for initial association with chromosomes and the N-terminal LEM and LEM-like domains for further steps of nuclear reorganization. As a consequence, C-terminal LAP2␣ fragments, which are able to bind chromosomes but lack these LEM domains, inhibited assembly of membranes and lamins around chromosomes. Furthermore, we show that, unlike the N-terminal domain, the C-terminal LAP2␣ fragment was able to bind to chromosomes in vivo when expressed weakly but caused cell cycle arrest and apoptosis in interphase at higher expression levels without progression to mitosis. These data indicate an important additional function of LAP2␣ in cell cycle progression during interphase that is clearly distinct from its role in nuclear assembly.
Cell Culture and Synchronization-Normal rat kidney (NRK), Chinese hamster ovary (CHO), and HeLa cells were routinely maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 50 g/ml penicillin and streptomycin (all from Invitrogen) at 37°C in a humidified atmosphere containing 8.5% CO 2 . To obtain metaphasearrested cells, subconfluent NRK cell cultures were incubated for 10 -12 h in medium containing 2 mM thymidine (Sigma-Aldrich), released from the block in medium without thymidine for 4 h, and further incubated for 12-14 h in medium containing 0.2 g/ml nocodazole (Calbiochem).
Expression and Isolation of Recombinant LAP2 Proteins-Recombinant proteins were expressed in Escherichia coli BL21(DE3) using the inducible T7 RNA polymerase-dependent pET vector system (40). Protein expression was induced with 0.5 mM isopropyl-␤-D-thiogalactopyranoside for 3 h, and bacteria were harvested by centrifugation at 4000 rpm for 5 min (Heraeus Megafuge, 1.0R). Bacteria were frozen in one-tenth of the original culture volume of Tris buffer (20 mM Tris-HCl, pH 8, 500 mM NaCl, 5 mM imidazol, 1 mM dithiothreitol, protease inhibitors 1 mM phenylmethylsulfonyl fluoride, 3.3 g/ml aprotinin, leupeptin, and pepstatin (all from Sigma-Aldrich), thawed, lysed by adding 0.1 mg/ml lysozyme, 0.1% Triton X-100, 10 mM MgCl 2 , 50 g/ml DNase, and 20 g/ml RNase, left at 30°C for 30 min, and centrifuged for 10 min at 14,000 rpm. The pellet was resuspended in one-tenth of the original culture volume of Tris buffer, 7 M urea was added, and the samples were incubated for 40 min at room temperature and homogenized in a glass-glass homogenizer. Cell lysates were subsequently spun at 45,000 rpm (TY65 rotor, Beckman Instruments Inc., Palo Alto, CA) for 30 min, and supernatants were aliquoted and stored at Ϫ20°C. If fragments were soluble in Tris buffer, DNase and RNase digestion was left out, and urea was added directly to the cell extract prior to centrifugation at 14,000 rpm for 10 min.
Renaturation of recombinant proteins was achieved by dialyzing twice against KHM buffer (see below) containing 1 mM phenylmethylsulfonyl fluoride. Other protease inhibitors were added then, samples were incubated for 20 min at 37°C and centrifuged at 4000 rpm for 5 min, and supernatants were used for in vitro nuclear reassembly reactions.
In Vitro Nuclear Assembly-Metaphase chromosomes were isolated from nocodazole-arrested CHO cells as described previously (17,40). For in vitro nuclear assembly assays, nocodazole-arrested NRK cells were incubated in complete medium containing 20 M cytochalasin B (Sigma-Aldrich) and 0.2 g/ml nocodazole for 30 min at 37°C and were lysed in ϳ5 volumes of ice-cold KHM buffer (78 mM KCl, 50 mM HEPES, pH 7.4, 8.4 mM CaCl 2 , 10 mM EGTA, 4 mM MgCl 2 , 1 mM dithiothreitol) containing protease inhibitors and 20 M cytochalasin B using a metal ball cell cracker (EMBL, Heidelberg, Germany; 15 strokes with ball, r ϭ 8.008 or 8.006 mm). Cell lysates were either used directly for nuclear assembly or centrifuged at 2000 ϫ g for 5 min to remove chromosomes. For assembly, 200 l of chromosome-free cell lysates were mixed with 100 l of exogenous metaphase chromosome fraction (A 260 ϭ 3), and the mixture or total cell lysates containing endogenous chromosomes were incubated for 1 h at 37°C without or after the addition of dialyzed recombinant protein. Phosphatase inhibitors (0.1 M calyculin A, 0.1 M okadaic acid (Invitrogen), 1 mM orthovanadate, and 0.5 mM ␤-glycerophosphate (Sigma-Aldrich)) were added then, and samples were centrifuged at 2000 ϫ g for 10 min. Supernatant and pellet fractions were analyzed by immunoblotting.
Immunofluorescence Microscopy and Bromodeoxyuridine (BrdUrd) Incorporation Assay-For immunofluorescence microscopy, 3.7% formaldehyde was added to in vitro nuclear assembly extracts, and lysates were gently spun onto coverslips (500 rpm for 20 s) and fixed in 3.7% formaldehyde in PBS for 20 min at room temperature. HeLa cells were fixed on Petri dishes in 3.7% formaldehyde/PBS. Samples were incubated in 50 mM NH 4 Cl/PBS and in 0.1% Triton X-100/PBS for 5 min each. For BrdUrd labeling, cells were pulsed with 10 mol/liter BrdUrd (Roche Molecular Biochemicals) for 1 h and fixed with 70% ethanol, 50 mM glycine, pH 2. All samples were incubated in 0.2% gelatin/PBS for 30 min and with primary and secondary antibodies in PBS/gelatin for 1 h at room temperature. Antibodies used were antiserum to LAP2␣ (1:1000) generated against the recombinant LAP2␣ C terminus (amino acids 188 -693) or undiluted hybridoma supernatants containing antibodies to LAP2␣ or LAP2␤ (21), a monoclonal antibody to lamins A/C (diluted 1:50 (42), and secondary antibodies Alexa Fluor 488-coupled goat anti-mouse (1:50, Molecular Probes, Leiden, Netherlands) and Texas Red-conjugated goat anti-mouse and goat anti-rabbit (1:500, Jackson ImmunoResearch). Membranes were stained with 20 g/ml DHCC (Sigma-Aldrich) for 2 h at room temperature, and DNA was stained with 1 g/ml Hoechst dye 33258 for 10 min. Samples were mounted in Mowiol and viewed in a Zeiss Axiovert 100 M microscope equipped with a Zeiss LSM confocal laser scanning microscope.
Other Procedures-SDS-PAGE and immunoblotting were done as described previously (21). Primary antibodies used were anti-LAP2 antibodies (hybridoma supernatants against LAP2␣ and LAP2␤, undiluted) and monoclonal anti-lamin A/C antibody (1:1000). For immunological detection of the proteins, the Protoblot Immunoscreening System (Promega, Madison, WI) or the SuperSignal ECL system (Pierce) were used. Quantification of stained protein bands was done using NIH Image software. TUNEL assays were done using the apoptosis detection system (Promega) according to the manufacturer's instructions.

Recombinant Chromosome-binding Fragments of LAP2␣
Interfere with Nuclear Assembly in Vitro-In addition to the LEM-and LEM-like domains in the N-terminal constant region of LAP2, which mediate the binding to chromosomal protein BAF and to DNA, respectively (34,35,38), we have recently identified a region in the LAP2␣ unique C-terminal part (amino acids 270 -615, Fig. 1) that is essential and sufficient for binding of LAP2␣ to mitotic chromosomes in vitro (40). Because we found that LAP2␣ associates with chromosomes at early stages of nuclear assembly prior to the accumulation of lamins and LAP2␤ (20,21,40), we reasoned that the protein might have important functions in nuclear reassembly. To investigate these functions in more detail and to identify the specific contributions of N-and C-terminal chromosome-binding domains to this process, we performed in vitro nuclear assembly studies in the presence or absence of bacterially expressed LAP2␣ fragments. Incubation of nocodazole-arrested mitotic NRK cell lysates containing endogenous chromosomes or of chromosomedepleted mitotic cell fractions with exogenous metaphase chromosomes at 37°C caused the majority of LAP2␣ to shift from a soluble chromosome-free fraction to a sedimentable chromosome-containing fraction ( Fig. 2A, Control panel) (21), reflecting partial in vitro assembly of nuclear structures such as the nuclear membrane and nuclear lamina (43). For a more quantitative statistical analysis of protein assembly, we calculated the efficiencies of assembly (⌬P) defined by the relative amount of cellular LAP2␣, LAP2␤, and lamins A/C in the pellet fractions at 60 min of incubation minus the relative amounts in the pellet fractions at 0 min (Fig. 2B). Immunofluorescence microscopy of samples using antibodies to the proteins and a lipid dye to detect membranes showed that condensed chromosomes at the 0 min time point did not contain any LAP2␣, LAP2␤, lamins A/C, or membranes. However, after a 60-min incubation the chromosomes had visibly started to decondense, and LAP2␣ and lamins A/C were detected at the chromosomes, with LAP2␤ and the membrane detected mostly at the rim of the DNA structures (Fig. 3A). This reflects the distribution of the proteins in postmitotic stages in vivo (17,21).
To identify the potential effects of LAP2␣ fragments on nuclear assembly, we added recombinant proteins to the assembly mixture at metaphase and determined by statistical analyses the assembly efficiencies of endogenous LAP2␣, LAP2␤, and lamins relative to those in the control assemblies done in parallel (⌬P/⌬P control ). The addition of an ϳ5-fold molar excess of the constant N-terminal LAP2-(1-187) fragment over endogenous LAP2␣ had no significant effect on the assembly of LAP2␣, LAP2␤, and lamins A/C during incubation (Fig. 2C; for a representative experiment see Fig. 2A, left panels), whereas the recombinant LAP2 fragment clearly remained in the supernatant fraction after 60 min of incubation. In contrast, the addition of an ϳ5-fold molar excess of C-terminal LAP2␣-(188 -693) to the assembly extract caused a clear inhibition of LAP2␣ assembly to around 20% of the control assembly (Fig. 2, A and  C). This effect is most likely due to competition of the LAP2␣ fragment with endogenous protein for binding sites on chromosomes, as a significant fraction of recombinant LAP2␣-(188 -693) was found in the chromosome-containing pellet fraction at the 0 min time point (Fig. 2A), and the protein could also be detected on the chromosomal surface at 0 min by immunofluorescence microscopy (Fig. 3B). Intriguingly, we found that LAP2␣-(1-693) had also a dramatic effect on the assembly of other nuclear proteins. LAP2␤ assembly was inhibited to ϳ30% and lamins A/C assembly to ϳ15% of the controls in the sedimentation assay (Fig. 2). Immunofluorescence microscopy of these samples revealed that LAP2␤, membranes, and lamins A/C did not efficiently bind to and assemble around chromosomes (Fig. 3B), giving rise to only a few small spot-like structures at the chromosomes after 60 min of assembly.
This dominant negative effect of the LAP2␣ C-terminal fragment on in vitro nuclear assembly could be unspecific because of the assembly of a huge bulky aggregate of recombinant fragment around chromosomes, which prevents docking of other proteins to the chromosomal surface. Alternatively, one could imagine a more specific process in which the N-terminal LEM domain and/or LEM-like domain in the constant region is required for proper nuclear assembly. In the control assembly, where endogenous full-length protein binds to chromosomes, the N-terminal domain of the intact protein is brought in close proximity to potential interaction partners on chromosomes and may, by binding to those partners, favor nuclear assembly progression. In the presence of the dominant negative fragment, which binds metaphase chromosomes and thus prevents docking of full-length protein containing the N-terminal domain during the assembly reaction, this process might not occur. If this were the case, the addition of full-length recombinant LAP2␣, which was shown to bind to metaphase chromosomes (21,40) most likely because of lack of mitosis-specific phosphorylation, should not inhibit nuclear assembly. Therefore, we added full-length LAP2␣ to the cell extract and tested assembly by immunofluorescence microscopy (Fig. 3C). LAP2␣ bound to metaphase chromosomes, and the assembly of LAP2␤, nuclear membranes, and lamins A/C occurred. Therefore, we suggest that both the C-terminal and N-terminal LAP2␣ regions are essential for nuclear assembly and have to work in a

Functions of LAP2␣ in Proliferation and Nuclear Assembly
timely coordinated, and sequential manner.
C-terminal but Not N-terminal LAP2␣ Fragments Associate with Chromosomes in Vivo-Having identified a dominant negative effect of the C-terminal LAP2␣ fragment on assembly of LAP2␣, LAP2␤, and lamins around chromosomes in vitro, we sought to analyze the effects of LAP2␣ fragments in vivo. We transiently expressed N-terminal green fluorescence protein (GFP) fusion constructs of LAP2␣ fragments in HeLa cells and analyzed the localization of the ectopic proteins by confocal laser or time lapse microscopy during cell division. At low expression levels, ectopic C-terminal LAP2␣ fragment translocated from the cytoplasm in metaphase (Fig. 4A, 0 min) to separated chromosomes in telophase (18 min) and remained in the nucleoplasm of the newly formed daughter nuclei (30 min). At higher resolution in confocal microscopy of fixed samples, the LAP2␣ fragment was clearly detected in defined structures around the decondensing chromosomes at telophase (Fig. 5A) as shown previously for endogenous protein (21,40). In contrast, strongly expressed N-terminal LAP2␣ fragment did not associate with chromosomes at this particular cell cycle stage (Fig. 5C) and throughout cell division, as shown previously (40).
Expression of C-terminal LAP2␣ Fragments Induce Apoptosis-By generating stable cell clones expressing an N-terminal LAP2␣ fragment containing the LEM and LEM-like domains (amino acids 1-254), we have previously shown that the ectopic protein did not associate with chromosomes and did not interfere with cell cycle progression (40). However, stable lines expressing the C-terminal fragment-(188 -693) could not be generated, suggesting a toxic effect of this fragment. At low expression levels, slightly above the detection limit in fluorescence microscopy (roughly estimated to represent less than 30% of the endogenous protein), transiently expressed C-terminal fragment had no obvious drastic effect on cell division as shown above (Figs. 4A and 5A). However, cells expressing higher levels of the C-terminal LAP2␣ fragment, comparable with those of endogenous protein or of ectopically expressed N-terminal fragment, were never observed in time lapse microscopy to proceed through mitosis, although untransfected cells in the same sample divided normally. 100% of these cells could be observed in an interphase-like stage from 30 min up to 24 h, before they showed morphological features of apoptosis FIG. 2. Effects of LAP2␣ fragments on in vitro nuclear assembly. Mitotic NRK cell lysates containing endogenous or exogenous chromosomes were incubated for 60 min at 37°C without (Control) or after the addition of dialyzed recombinant LAP2-(1-187) or LAP2␣-(188 -693) fragment at an ϳ5-fold excess over endogenous LAP2␣. At 0 and 60 min, lysates were centrifuged, and supernatant and pellet fractions were analyzed by immunoblotting using antibodies to the indicated proteins. A, immunoblot samples representing an assembly with endogenous chromosomes. B, protein amounts in bands on immunoblots (assemblies with exogenous and endogenous chromosomes) were determined by densitometry, and the efficiency levels of protein assembly in control assays were calculated by subtracting the relative amount of the proteins in pellet fraction (%) at 0 min from that at 60 min (⌬P). C, efficiency levels of protein assembly in assays with recombinant proteins relative to efficiency levels in control assemblies done in parallel (⌬P/ ⌬Pcontrol) were determined as in B. The data represent the mean values of at least three independent experiments. Bars represent standard deviations.
such as membrane blebbing, cell shrinkage, and chromatin condensation (Fig. 4B). The detection of apoptosis-specific chromatin aggregates and chromatin-independent localization of LAP2␣ (44) by confocal microscopy (Fig. 5B) as well as TUNEL assays (data not shown) confirmed that cells entered apoptosis.

Ectopic Expression of Full-length LAP2␣ and of C-terminal LAP2␣ Fragments Inhibit Cell Proliferation and Progression to
S Phase-The observation that cells expressing significant levels of C-terminal LAP2␣ fragment entered apoptosis in interphase without proceeding to mitosis showed that apoptosis was not induced due to defects in nuclear reassembly after mitosis and suggested that the fragment interfered with a distinct function of LAP2␣ in interphase. To analyze the toxic effect of overexpressed LAP2␣ fragments in more detail and to correlate the toxic effects with a specific region, we transiently expressed GFP fusion constructs of various LAP2␣ fragments or GFP alone as a control in HeLa cells and followed the amount of GFP-positive cells over a time period of 4 days after transfection by FACS analysis (Fig. 6A). While the amount of cells expressing GFP alone or GFP-LAP2-(1-187) fragment containing the LEM domains but lacking the ␣-specific chromosome binding region showed only a minor decrease within 4 days, we detected a significant reduction of the GFP-positive cells expressing the constructs with a functional chromosome-binding region. The smallest fragment thus far reported with chromatin-binding activity (GFP-LAP2␣-(270 -615)) (40) had the most drastic effect, reducing the number of GFP-positive cells by 90% after 1 day in culture and to undetectable amounts after 4 days. GFP fusion constructs representing the entire LAP2␣specific region (GFP-LAP2␣-(188 -693)) or the full-length protein GFP-LAP2␣-(1-693) had a less severe effect but still caused a 50% reduction compared with the control after 1 day and a nearly 70 -80% decrease after 4 days. Different models could be used to explain these results. LAP2␣ fragments containing the C terminus could either interfere with cell proliferation or cell viability, or the stability of these fragments in the cells could be lower than those of the N-terminal fragments.
To distinguish between these different possibilities, we expressed LAP2␣ or LAP2␣ fragments together with GFP from a bicistronic expression vector (Fig. 6B). This ensures that all cells expressing GFP, which can be detected by FACS analysis, also express LAP2␣ fragments. If the effects of the GFP fusion proteins seen before were due to instability or aberrant behavior of the constructs, they would not be detected in this assay. However, we observed a similar effect of the untagged LAP2␣ fragments as with GFP fusion proteins. GFP alone or N-terminal LAP2-(1-187) caused only a slight decrease in the amount of GFP-positive cells, indicating that the transfected cells are viable. Expression of all other constructs containing the Cterminal region (amino acids 270 -615) led to a clear decrease in GFP-positive cells by at least 50% of the starting number after 1 day and a reduction of 10 -30% after 4 days. The less deleterious effect of LAP2␣-(270 -615) compared with the respective GFP fusion construct was most likely caused by a slightly lower stability of the untagged versus the GFP-tagged fragment. On the other hand, the less severe effect of GFP-LAP2␣-(1-693) compared with the untagged fragment may be caused by a slightly higher tendency of the GFP construct to form inactive aggregates in the nucleus and cytoplasm. Nevertheless, taking both approaches together, these results indicated that only LAP2␣ fragments containing the C-terminal chromosome-binding region have an inhibitory effect on cell proliferation and/or cell viability.
To confirm the arrest in cell cycle progression induced by the C-terminal LAP2␣ fragments, we transiently expressed the fragments in cells and tested incorporation of BrdUrd into DNA to detect specifically proliferating DNA-replicating cells in S phase. Whereas around 60% of cells expressing GFP alone or N-terminal GFP-LAP2-(1-187) fragment incorporated Brd-Urd, only 20 -25% of the BrdUrd-positive cells were detected in the cell fraction expressing the C-terminal constructs (amino acids 188 -693 or 270 -615) and 40% in cells expressing fulllength LAP2␣ (Fig. 7). Thus, it can be concluded that ectopically expressed LAP2␣ constructs containing the C-terminal chromosome-binding region inhibited progression into S phase, whereas an N-terminal fragment had no effect. Although DNA FACS analyses could not be done because of the low transfec-tion efficiencies in these experiments, the BrdUrd incorporation assays indicated that cell cycle arrest occurs in G 1 phase.

LAP2␣ Has Distinct Functions in Interphase and Mitosis-In
this study we provide evidence for two distinct functions of LAP2␣ in different stages of the cell cycle, in nuclear assembly during mitosis and in the progression of cells into S phase. We show that the addition of fragments containing the LAP2␣ unique C terminus blocks nuclear assembly and that overexpression of full-length protein or of C-terminal fragments causes cell cycle arrest and apoptosis. Although both the inhibitory effect on nuclear assembly and on cell cycle progression required the addition or overexpression of the C-terminal regions, the molecular mechanisms were clearly different. The inhibition of nuclear assembly was caused only by the addition of C-terminal LAP2␣ fragment, not by the addition of fulllength LAP2␣, suggesting that the fragment served as a dominant negative version, interfering with the function of endogenous protein (i.e. by competition with endogenous protein for binding partners). These findings not only revealed essential functions for the C terminus in chromosome binding, but they also implied that the N-terminal domain, which by itself is not able to bind to chromosomes, has important functions in nuclear assembly (see below).
In contrast, cell cycle arrest was caused by overexpression of both C-terminal fragments and full-length protein, indicating a specific function of the LAP2␣ C terminus in this process. These data also suggest that the tightly controlled expression level of the LAP2␣ C terminus is essential for cell growth, implying that the phenotype is caused by a gain of LAP2␣ function rather than a dominant negative effect. Although the LAP2␣ domain responsible for cell cycle arrest maps to the chromosome binding region (amino acids 270 -615) (40), it remains unclear whether chromosome binding of LAP2␣ is required for its growth-suppressive function. Various models of the potential functions of LAP2␣ consistent with our findings are described below.
Potential Functions of LAP2␣ in Progression into S Phase-Although we could not directly determine by DNA FACS analysis the cell cycle stage in which LAP2␣ overexpressing cells were arrested, due to extremely low transfection efficiencies, BrdUrd incorporation assays suggested an arrest in G 1 phase. In this case, one could explain the suppressive effect on cell proliferation of overexpressed LAP2␣ by a regulatory function of LAP2␣ or of nuclear LAP2␣ complexes on cellular components controlling G 1 -S phase progression. This hypothesis is supported by recent findings showing that lamina proteins might regulate E2F transcriptional activity, which is necessary for transcription of S phase-specific genes (45). The membranebound LAP2 isoform, LAP2␤, was shown by yeast two-hybrid analysis to bind to mouse germ cell-less (mGCL) (46), which in turn interacts with E2F-associated DP and regulates the cell cycle (47). As overexpressed LAP2␤ was found to reduce E2Fdependent reporter activity (46), it was suggested that LAP2␤ might negatively regulate E2F activity by tethering the transcription complex to the nuclear periphery, a mechanism known in other transcription factors (2). Because the mGCL interaction domain has been mapped to a LAP2␤-specific region, it is unlikely that LAP2␣ binds mGCL, but LAP2␣ might indirectly affect the structure and function of lamin-LAP2␤ complexes. In addition, lamin A/C was shown to associate directly with the hypophosphorylated active form of retinoblastoma protein (pRb) (48), which binds E2F and represses transcription of S phase-specific genes (49,50). LAP2␣, which has been identified as a direct binding partner of A-type lamins (20), might influence the lamin A-pRb interaction or might affect pRb function directly by binding to pRb, as suggested by our recent in vitro binding studies. 2 In addition to a direct effect of lamina proteins on S phasespecific transcription factors, lamins and lamin-LAP2 complexes could inhibit cell cycle progression more indirectly by changing the higher order chromatin structure, thus allowing or preventing DNA replication. In line with this model, formation of a lamina in Xenopus in vitro nuclear assembly systems has been found essential for the initiation of DNA replication (10,30,51,52), and lamin mutants causing nuclear lamina disassembly were shown to inhibit the elongation phase of DNA replication (10,30,52). Furthermore, lamin mutants causing a dramatic reorganization of the lamina and lobulated nuclei interfered with DNA replication and cell growth (53). Accordingly, ectopic expression of lamin-binding LAP2␤ fragments in mammalian cells inhibited progression into S phase (19), whereas LAP2␤ mutants added to Xenopus in vitro nuclear assembly reactions influenced DNA replication positively (15).
Potential Function of LAP2␣ in Apoptosis-We found that cells overexpressing LAP2␣ and LAP2␣ fragments entered apoptosis upon a G 1 arrest for up to 24 h. Apoptosis could have been induced indirectly by the misregulation of the cell cycle control machinery, but recent data indicate that lamina proteins might also be involved more directly in controlling apoptosis. In C. elegans, for instance, CED-4, a cell death activator, is translocated from mitochondria to the nuclear envelope before caspase activation (54), suggesting that the lamina provides an attachment site for the apoptotic signaling machinery (2). Lamins, LAP2␣, and LAP2␤ are early targets of apoptosis (44,55,56), and expression of uncleavable lamin mutants was shown to delay apoptosis for several hours (56). Furthermore, inhibition of lamin B assembly at the nuclear envelope upon preventing its postmitotic dephosphorylation induced apoptosis in human cells (57), suggesting that mislocalized lamins actively trigger apoptosis (58).
Functions and Interdependence of Various Chromosome Binding Regions of LAP2␣ During Nuclear Assembly-In this study we present a variety of evidence that LAP2␣ possesses several chromosome-binding domains showing different binding properties during nuclear assembly. In the N terminus all LAP2 isoforms as well as emerin and MAN1 contain a common structural motif, the LEM domain (33), which was found to interact with the chromosomal protein BAF (35). In addition, LAP2 isoforms contain a LEM-like motif at their extreme N termini, which was shown by structural studies to bind to DNA (38). In accordance with the proposed interaction of the LAP2 constant region with chromosomes, via BAF and/or via direct DNA binding, GST fusion proteins of the N-terminal 85 amino acids (18) or of the entire LAP2 constant region (amino acids 1-187) (40) were found to interact with chromosomes in vitro. On the other hand, unlike GST fusion proteins, His-tagged LAP2 N-terminal fragments did not interact with chromosomes, and ectopically expressed fragments in cells did not associate with chromosomes during nuclear reassembly as fulllength LAP2␣ (40). These observations suggested that additional domains downstream of the LEM domain or protein oligomerization (as achieved in GST fusion proteins) is required for establishing tight interactions between the LEM domain and BAF. This hypothesis is further supported by recent in vitro binding studies showing that various Xenopus LAP2␤-like isoforms, which are identical in their N-terminal part and contain the LEM domain but differ in their C-terminal region, varied 9-fold in their affinities for BAF. These data suggested that the C-terminal regions in LAP2 isoforms might regulate the activity of the LEM domain (35).
Furthermore, our previous studies (40) and the data presented here show that the C-terminal region of LAP2␣ contains a domain that is both essential and sufficient for binding to chromosomes in vivo and in vitro. Therefore, we speculate that LAP2␣ is targeted to chromosomes via its interaction with the C-terminal domain initially. This interaction may then promote the subsequent binding of the N-terminal LEM domain and/or LEM-like domain to BAF and/or DNA, whereas the N-terminal LEM domains are not capable of binding chromosomes on their own. Based on our in vitro nuclear assembly assays using the dominant negative LAP2␣ C-terminal domain, we propose that the subsequent activation of the LEM domains and their interaction with BAF and/or DNA are essential for the proper nuclear envelope assembly and chromatin organization. In the absence of this interaction, which is achieved by blocking all LAP2␣ binding sites on the chromosomal surface upon the addition of an excess of the chromosome-binding LAP2␣ fragments missing the LEM domains, nuclear reassembly is inhibited.
Implications of Our Findings for Laminopathies-The recent discovery that mutations in genes encoding nuclear lamina proteins are linked to inherited human diseases affecting the heart and skeletal muscle as well as adipose tissue (laminopathies) (25,27,28) has prompted a series of studies aimed at unraveling the molecular mechanisms that cause disease and the cellular functions of lamina proteins in general. The Xlinked form of Emery-Dreifuss muscular dystrophy is caused by mutations in the LEM-domain protein emerin, whereas autosomal-dominant Emery-Dreifuss muscular dystrophy is caused by mutations in lamins A/C (LMNA gene). Other mutations in LMNA cause dilated cardiomyopathy, Dunningantype partial lipodystrophy, and limb-girdle muscular dystrophy 1B. The reported direct interactions of emerin with A-type lamins (59,60) and the dependence of correct emerin localization in the nuclear membrane on lamin A expression (24,29) led to the model that function(s) of lamin A/emerin complexes are disturbed in laminopathic tissues. In view of the recently reported interaction of LAP2␣ with the C-terminal region of A-type lamins (20) containing lipodystrophy and Emery-Dreifuss muscular dystrophy mutations (61), it is intriguing to speculate that these mutations may affect LAP2␣-lamin A interactions and that the interference with LAP2␣ functions and/or localization might contribute to the specific disease phenotype. Lamin/emerin mutations would interfere mostly with peripheral lamina structure and function, whereas lamin mutations influencing lamin-LAP2␣ interactions may affect internal nuclear structures and may thus cause different disease phenotypes.
In view of the potential functions of LAP2␣ reported in this study, it is tempting to speculate that lamin A mutations may cause functional defects in lamin-LAP2␣ complexes in cell proliferation and/or apoptosis or cell division and chromatin organization and might thus contribute to the cellular phenotype of the disease.