Lamina‐associated polypeptide (LAP) 2 of the inner nuclear membrane (now LAP2β) and LAP2α are related proteins produced by alternative splicing, and contain a common 187 amino acid N‐terminal domain. We show here that, unlike LAP2β, LAP2α behaved like a nuclear non‐membrane protein in subcellular fractionation studies and was localized throughout the nuclear interior in interphase cells. It co‐fractionated with LAP2β in nuclear lamina/matrix‐enriched fractions upon extraction of nuclei with detergent, salt and nucleases. During metaphase LAP2α dissociated from chromosomes and became concentrated around the spindle poles. Furthermore, LAP2α was mitotically phosphorylated, and phosphorylation correlated with increased LAP2α solubility upon extraction of cells in physiological buffers. LAP2α relocated to distinct sites around chromosomes at early stages of nuclear reassembly and intermediarily co‐localized with peripheral lamin B and intranuclear lamin A structures at telophase. During in vitro nuclear assembly LAP2α was dephosphorylated and assembled into insoluble chromatin‐associated structures, and recombinant LAP2α was found to interact with chromosomes in vitro. Some LAP2α may also associate with membranes prior to chromatin attachment. Altogether the data suggest a role of LAP2α in post‐mitotic nuclear assembly and in the dynamic structural organization of the nucleus.
The nuclear envelope (NE) forms the boundary of the nucleus in eukaryotic cells and separates nuclear and cytoplasmic activities. It consists of a double membrane, an underlying filamentous meshwork (the nuclear lamina) and nuclear pore complexes which mediate nucleo‐cytoplasmic transport (for reviews see Gerace and Burke, 1988; Moir et al., 1995; Gant and Wilson, 1997). Although the outer and inner nuclear membranes are periodically joined at nuclear pore complexes, they are structurally and functionally distinct. While the outer nuclear membrane is continuous with the smooth and rough endoplasmic reticulum, the inner nuclear membrane contains specific integral membrane proteins which link the membrane to the underlying lamina and to other nuclear components (reviewed in Georgatos et al., 1994; Gerace and Foisner, 1994). The structural integrity of the NE is essential for nuclear functions such as DNA replication, RNA transcription and RNA processing. There is increasing evidence that the NE is involved in the organization of interphase chromatin, but the molecular mechanisms are not yet understood.
The lamins, intermediate filament‐type proteins, comprising constitutively expressed B‐type lamins and developmentally regulated A‐type lamins, form a two‐dimensional quasi‐tetragonal network‐like structure underneath the nuclear membrane and build the major structural framework of the NE (reviewed in Gerace and Burke, 1988). Previous studies, demonstrating in vitro interactions of lamins with chromosomes and chromatin (Glass and Gerace, 1990; Hoeger et al., 1991; Yuan et al., 1991; Glass et al., 1993; Taniura et al., 1995), the co‐localization of lamins with DNA at the nuclear periphery (Paddy et al., 1990; Belmont et al., 1993; Fricker et al., 1997) and within the nucleus (Bridger et al., 1993; Moir et al., 1994; Hozak et al., 1995), and the lack of DNA replication in nuclei without an intact lamina (Newport et al., 1990; Goldberg et al., 1995; Ellis et al., 1997; Spann et al., 1997) suggested a role of lamins in chromatin organization. More recently, lamina‐associated proteins have also been implicated in the structural organization of the nucleus. Several inner nuclear membrane proteins associated with the nuclear lamina have been characterized in higher eukaryotes: lamina‐associated polypeptide (LAP) 1 (Senior and Gerace, 1988; Martin et al., 1995); LAP2 (Foisner and Gerace, 1993; Furukawa et al., 1995); p58/lamin B receptor (LBR) (Worman et al., 1990); and otefin (Padan et al., 1990). These proteins have been found to co‐fractionate with detergent‐high‐salt‐insoluble nuclear lamina/nuclear matrix preparations (Senior and Gerace, 1988; Worman et al., 1988; Ashery‐Padan et al., 1997b) or to interact directly with lamins in vitro (Worman et al., 1988; Foisner and Gerace, 1993). Several new compelling findings suggest that at least some of these proteins might also be directly involved in chromatin organization. p58/LBR and LAP2 have been shown to associate with chromosomes and chromatin in vitro (Foisner and Gerace, 1993; Pyrpasopoulou et al., 1996), and might provide essential chromatin docking sites at the nuclear envelope. In a two‐hybrid screen and by co‐immunoprecipitation, LBR was found to interact with human chromodomain proteins homologous to Drosophila HP1, a heterochromatin protein involved in position effect variegation (Ye and Worman, 1996; Ye et al., 1997).
A remarkable feature of the nucleus in higher eukaryotes is its ability to disassemble entirely at mitosis and then rapidly reassemble around daughter chromosomes (for recent reviews see Lourim and Krohne, 1994; Moir et al., 1995; Foisner, 1997; Marshall and Wilson, 1997). NE breakdown, which involves nuclear membrane fragmentation and disassembly of the nuclear lamina, is triggered by mitosis‐specific phosphorylation of the lamins (for review see Nigg, 1992), but there is evidence that the LAPs (Foisner and Gerace, 1993), LBR (Bailer et al., 1991; Courvalin et al., 1992; Nikolakaki et al., 1997) and otefin (Ashery‐Padan et al., 1997a) are also subject to mitosis‐specific phosphorylation and may contribute to mitotic nuclear disassembly. The post‐mitotic assembly of nuclei, including the targeting of nuclear membranes to chromosomes, membrane fusion, nuclear pore complex assembly, lamina assembly and chromatin decondensation, offers an ideal system to study the potential role of lamins and associated proteins at different stages during the establishment of nuclear structure. The specific roles of lamins, LBR and LAPs in targeting nuclear membranes to the chromosomal surface at the end of mitosis has been controversial (reviewed in Lourim and Krohne, 1994; Foisner, 1997). Immunofluorescence microscopy has shown that LAP1, LAP2 (Foisner and Gerace, 1993; Maison et al., 1997) and p58/LBR (Chaudhary and Courvalin, 1993; Ellenberg et al., 1997) accumulate at the chromosomal surface in early stages of NE reformation in late anaphase before the bulk of lamins assemble into the lamina, suggesting that they may be involved in initial targeting of membranes to the chromosomes. Microinjection of recombinant proteins into mammalian cells revealed that fragments of LAP2β which contain the lamin‐binding domain inhibit nuclear volume increase and progression into S phase, but do not inhibit NE assembly (Yang et al., 1997b).
The originally described LAP2 protein, first discovered in rat (Foisner and Gerace, 1993), has recently been shown to be a member of a family of related but distinct nuclear proteins derived from a single gene by alternative splicing (Harris et al., 1994, 1995). Analysis of human LAP2 cDNAs and proteins, and of the human LAP2 gene, revealed three abundant forms: α (75 kDa), β (51 kDa) and γ (39 kDa). This analysis also revealed that an earlier proposed 49 amino acid secreted thymic polypeptide (named thymopoietin) was a proteolytic fragment from the N‐terminus of nuclear LAP2 proteins, probably artificially generated during the isolation, and indicated that earlier claims of a physiological role for this fragment in neuromuscular transmission and T‐cell function do not reflect the true cellular functions of the LAP2 proteins (Harris et al., 1994, 1995). Following the α, β, γ nomenclature of Harris et al. (1994, 1995), LAP2β, the human homologue of the originally described rat LAP2, and LAP2γ both contain a putative transmembrane domain at their C‐termini and are closely related structurally, the only difference being the insertion of a β‐specific domain of 109 amino acids in LAP2β. In contrast, the largest LAP2, LAP2α, contains only a 187 amino acid N‐terminal domain, identical to the N‐termini of LAP2β and LAP2γ, which is followed by a 506 amino acid α‐specific domain lacking a putative transmembrane region (Figure 1). Thus, LAP2 proteins are a family of related nuclear integral membrane proteins as well as non‐membrane associated proteins that may fulfil related functions in nuclear architecture and cell‐cycle‐dependent nuclear dynamics.
An additional nuclear membrane protein, possibly distantly related to LAP2 proteins, has recently been described. Emerin, the protein lost in patients with Emery–Dreyfuss muscular dystrophy (Bione et al., 1994), is an ubiquitous Triton‐resistant nuclear membrane protein (Manilal et al., 1996; Nagano et al., 1996) containing two regions with sequence similarity to LAP2: a 39 amino acid region near the N‐terminus containing 16 identities and a C‐terminal 34 amino acid region containing 14 identities that encompasses the membrane‐spanning domain (Manilal et al., 1996). Interestingly, the 187 residues common to all LAP2s includes the 38 amino acid region similar to the N‐terminus of emerin. One possible explanation for the involvement of a nuclear membrane protein in a genetic muscle‐wasting disease is that it may have a structural role in nuclear organization that affects muscle‐specific functions.
Here, we present for the first time biochemical and structural data on the non‐membrane bound LAP2 protein, LAP2α. We show that the protein is associated with intranuclear structures in interphase and concentrates in areas of the mitotic spindle and in the midbody of mitotic cells. It relocates into the nucleus at early stages of post‐mitotic nuclear assembly, giving rise to inter‐chromosomal aggregates, intranuclear and peripheral structures, intermediarily co‐localizing with lamin A and lamin B, respectively. The redistribution of LAP2α during cell division is correlated with a mitosis specific phosphorylation, and suggests a function of LAP2α in the establishment of post‐mitotic nuclear architecture.
LAP2α is associated with karyoskeletal components
To determine the subcellular distribution of LAP2α we performed immunoblot analyses of nuclear and cytoplasmic fractions of normal rat kidney (NRK) cells using various monoclonal antibodies (mAbs) raised against common domains of the LAP2 proteins or against LAP2α‐ and LAP2β‐specific regions. mAb 10, which was raised against a synthetic peptide corresponding to amino acids 29–50 within the common domain of LAP2 proteins (Figure 1), strongly recognized two proteins of ∼75 and ∼53 kDa exclusively in nuclei‐enriched fractions of NRK cells (Figure 2). In contrast, mAb 15 which was generated against recombinant human LAP2α, and mAb 11, generated against a synthetic peptide corresponding to α‐specific amino acids 233–253, reacted only with the ∼75 kDa protein in NRK nuclei and/or HeLa cell lysates (Figures 1 and 2). The ∼53 kDa band detected by mAb 10 in NRK nuclei co‐migrated with a HeLa protein immunoreactive with mAb 16 (Figure 2), an antibody raised against the β‐specific amino acids 313–330 (Figure 1). Thus, it may be concluded that NRK and HeLa cells contain both LAP2α and β, and that these proteins are predominantly nuclear.
Immunofluorescence microscopy of NRK cells using mAb 15 also revealed a predominantly nuclear localization of LAP2α (Figure 3A). However, unlike LAP2β (compare with Foisner and Gerace, 1993; data not shown), LAP2α was distributed throughout the nuclear interior being excluded only from the nucleoli. The nuclear localization of LAP2α was clearly distinct from that of lamin B at the periphery of the NE, which gave rise to a nuclear rim staining (Figure 3A).
To investigate the intranuclear distribution of LAP2α at the ultrastructural level, we performed immunoelectron microscopy of HeLa cells according to the Lowicryl low‐temperature embedding protocol. This technique, which is clearly superior over conventional embedding protocols in terms of antigen preservation, gives a poorer structural preservation. Nevertheless, cells appeared to be intact and the nucleus was clearly discernible (Figure 4a). At higher magnification (Figure 4b) the inner nuclear membrane (IM) could nearly be traced along the entire nucleus, whereas the outer membrane (OM) and nuclear pore complexes (NP) were often disrupted due to a swelling of the luminal space. The LAP2α‐specific label on the Lowicryl sections was scattered over the entire nucleus and was sometimes closely associated with the inner membrane (Figure 4B, arrowheads), whereas the cytoplasm was completely free of label. Together, these data indicated that LAP2α was located in intranuclear structures which were in close association with the NE at the nuclear periphery.
To test LAP2α association with nuclear structures on a biochemical level, we extracted NRK nuclei in urea or non‐ionic detergent and analyzed the distribution of both LAP2 proteins between soluble and insoluble fractions. Unlike LAP2β, LAP2α was solubilized in 7 M urea (Figure 5A), indicating that it is not inserted into the membrane bilayer. In contrast, LAP2α and β behaved similarly upon extraction of nuclei with 1% Triton X‐100 and increasing salt concentrations. At low ionic strength both proteins were completely retained in the low‐speed pellet fraction (P1), and also in more physiological ionic strength buffers (120 mM salt) the majority of the proteins remained insoluble (Figure 5A). Nuclease treatment of nuclei did not significantly change the solubility of the LAP2 proteins in Triton X‐100‐containing buffers. As the majority of the insoluble proteins sedimented at low‐speed centrifugation, it may be concluded that both LAP2α and β were associated with large cellular structures such as the nuclear lamina–nuclear matrix scaffold of the nuclei. At higher salt concentrations (250 mM) the majority of LAP2α and β was soluble (Figure 5A), indicating that their association with the nuclear scaffold was weaker than previously reported for LAP2 and LAP1 (Senior and Gerace, 1988; Foisner and Gerace, 1993). Since these experiments were originally performed using fractions of rat liver NEs, we too tested the biochemical properties of LAP2α in rat liver. An immunoreactive protein of ∼75 kDa, co‐migrating with NRK LAP2α, was detected in ‘salt washed’ NE fractions isolated from rat liver nuclei but was absent from the post‐nuclear microsomal membrane (MM) fraction (Figure 5B). Considering the harsh treatment of nuclei during the preparation of NEs, including extraction with DNase/RNase and with 0.5 M salt, the presence of LAP2α in the NE fraction was unexpected and suggested that the protein was more tightly bound to karyoskeletal structures in rat liver than in NRK nuclei. Accordingly, the majority of rat liver LAP2α was insoluble in 1% Triton X‐100 plus 250 mM salt, while it was still efficiently solubilized in 7 M urea (Figure 5B). To test whether LAP2α has the same subcellular distribution in rat liver tissue and in proliferating NRK cells, we performed immunofluorescence microscopy on cryosections of rat liver. mAb 10, directed against the common domain of LAP2 proteins, revealed staining throughout the nucleus and at the nuclear periphery (LAP; Figure 3B), while an antibody against the LAP2β‐specific domain stained the nuclear periphery exclusively. Therefore, the intranuclear staining detected by mAb 10 probably reflects the localization of the non‐membrane‐bound LAP2 isoform, LAP2α. Unlike in NRK cells (Figure 3A), the LAP2α‐specific antibody mAb 15 did not detect the protein in rat liver nuclei using immunofluorescence microscopy (data not shown). The lack of staining might be caused by inaccessibility of the antigen due to a tight integration of LAP2α in the nuclear structure, and is therefore consistent with the increased resistance of LAP2α to detergent‐salt extraction in rat liver tissue versus NRK cells.
LAP2α localizes to different cytoplasmic and nuclear structures during mitosis
Nuclear structures are profoundly reorganized in the course of the cell cycle, involving the disassembly of the NE and the condensation of chromosomes at metaphase, and the post‐mitosic re‐establishment of nuclear architecture. We followed the cellular localization of LAP2α at various cell‐cycle stages by confocal immunofluorescence microscopy, using LAP2α‐specific mAbs. At prophase, when DNA starts to condense, LAP2α was still located predominantly within the nucleus (Figure 6A). However, a superimposition of the LAP2α (green) and the DNA (red) stain clearly showed that LAP2α was mostly concentrated in the space between chromosomes, and showed only minor overlap (yellow) with DNA‐containing structures. In metaphase and anaphase, when the NE was completely disassembled and DNA fully condensed, LAP2α was found throughout the cell without an apparent association with the chromosomes. Instead, the protein seemed to be concentrated in areas of the mitotic spindle apparatus, particularly at the spindle pole regions (Figure 6B). At the initial stages of nuclear reassembly in early telophase, LAP2α relocated into discrete structures between and around decondensing chromosomes, again mostly filling the space between DNA‐containing structures (see superimposition in Figures 6C and D). At these cell‐cycle stages the LAP2α‐specific staining in the cytoplasm was significantly weaker than that of the chromosome‐associated LAP2α structures. In contrast, both the lamin B‐ and lamin A‐specific staining was still predominantly cytoplasmic (Figures 7A and B, upper panels). Thus, it can be concluded that LAP2α associates with nuclear structures very early in nuclear reassembly, clearly before assembly of the bulk of lamins. Residual cytoplasmic staining of LAP2α, however, was found in the midbody region of dividing cells (Figures 6D–F and 7). LAP2α was detected at discrete spots and structures within the nuclei, and at the nuclear periphery as long as the DNA remained partially condensed (Figure 6D and E), and redistributed more or less uniformly throughout the nucleus after chromosome decondensation was complete (Figure 6F). The discrete LAP2α structures at the nuclear periphery co‐localized with lamin B, which was assembled into the nuclear lamina at late telophase/G1 (Figure 7A, middle panels). Upon complete DNA decondensation, lamin B remained at the nuclear periphery, whereas LAP2α redistributed into intranuclear structures (Figure 7A, lower panels). Therefore, LAP2α may associate intermediarily with the NE and lamin B at early stages of nuclear assembly. Lamin A redistributed to peripheral (data not shown) as well as intranuclear (Figure 7B, lower panels) structures in late telophase and G1. As the intranuclear lamin A structures clearly overlapped with LAP2α during G1, LAP2α might also associate with lamin‐A‐containing structures at specific cell‐cycle stages and/or at specific nuclear sites.
LAP2α is phosphorylated in a mitosis‐specific manner
Since the phosphorylation of lamins, LAP2β and other nuclear components has been suggested to regulate the mitotic reorganization of the proteins (Foisner, 1997), we tested whether LAP2α was also modified in a mitosis‐specific manner. NRK cells presynchronized by a double thymidine block, or asynchronously growing HeLa cells were arrested at metaphase in nocodazole‐containing medium, and mitotic cells were harvested by mechanical shake‐off. FACS analyses (Figure 8B) and microscopy (not shown) revealed a mitotic index of >90% for these mitotic cell fractions. Immunoblot analyses of mitotic and interphase cell extracts (Figure 8A, lanes M and I) or of LAP2α immunoprecipitates from mitotic and interphase cells (data not shown), revealed a slightly reduced mobility of the mitotic LAP2α in sodium dodecyl sulfate (SDS) gels as compared with the interphase protein in both NRK and HeLa cells. To test whether the reduced mobility of mitotic LAP2α might be caused by mitosis‐specific phosphorylation, as has been reported for LAP2β, we analyzed mitotic and interphase LAP2α immunoprecipitates by two‐dimensional gel electrophoresis. Immunoblot analyses of isoelectric focusing gels showed that interphase LAP2α focused at the basic end of the gel at a pI of ∼7.0 (Figure 8C, panel I, small arrow), similar to immunoglobulin heavy chain (star). In contrast, mitotic LAP2α immunoprecipitates focused in at least five distinct spots in a pI range of 5–6 (Figure 8C, panel M, small arrows), whereas the immunoglobulin heavy‐chain as a control (star) was not shifted to acidic pI. Analysis of a mixture of interphase and mitotic samples (Figure 8C, panel I + M) clearly confirmed the differential position of the interphase versus mitotic LAP2 proteins. These results indicated that LAP2α is more highly phosphorylated in mitosis than in interphase, and that at least five different phosphorylation sites might be targeted in a cell‐cycle‐dependent manner. In line with this observation, amino acid sequence analysis revealed two perfect and seven minimal consensus phosphorylation motifs for the mitotic p34cdc2 kinase in LAP2α (Figure 1). The correlation between the phosphorylation state of LAP2α and its mobility on SDS gels was further confirmed by incubating mitotic cell lysates at 30°C in the absence or presence of phosphatase inhibitors. As depicted in Figure 8D, incubation of lysates in the absence of phosphatase inhibitors for 1 h (lane M1) completely removed the molecular weight shift of mitotic LAP2α in SDS gels, while incubation in the presence of okadaic acid and calyculin for 1 h (M1PI) had no apparent effect on the protein's gel mobility.
LAP2α is soluble during mitosis and reassembles in vitro
The different cellular localization of LAP2α in interphase versus mitotic cells was also reflected by a different solubility of the protein in various buffer conditions. When S‐phase cells were lysed in physiological ionic strength buffer (−) or in buffer plus 1% Triton X‐100, the majority of LAP2α was detected in the insoluble nuclear fractions (Figure 9A, Interphase). The protein was only solubilized at significantly higher ionic strength buffers (500 mM salt). In contrast, LAP2α was completely soluble in mitotic cells under all conditions tested (Figure 9A, Metaphase). Thus, LAP2α is apparently not associated with large cellular structures during mitosis, such as chromosomes, membrane vesicles or cytoskeletal networks. LAP2β, in contrast, was recovered in the insoluble fraction upon lysis of mitotic cells in detergent‐free buffers due to its integration in membrane vesicles, while it was mostly soluble in buffers containing 1% Triton X‐100.
It has previously been demonstrated that NE assembly occurs in vitro upon incubation of mitotic cell lysates at 30°C (Burke and Gerace, 1986; Maison et al., 1995). As shown in Figure 8D, incubation of mitotic cell lysates at 30°C for up to 1 h reverted the slightly decreased SDS gel mobility of mitotic LAP2α to that of the interphase protein, suggesting that at least a partial exit from mitosis took place in vitro under these conditions. To test whether LAP2α assembled into larger complexes during the incubation, cell lysates were centrifuged after various incubation times, and insoluble and soluble fractions were tested by immunoblot analyses. We found that the mitotic soluble LAP2α increasingly shifted into the pellet fraction depending on the incubation time (Figure 9B). While after a 0.5 h incubation ∼50% of the mitotic soluble protein was found in the pellet fraction, nearly all of the protein was converted into the insoluble form after 1 h incubation (Figure 9B). Control experiments, performed in the presence of phosphatase inhibitors revealed that the majority of the protein stayed soluble even after a 1 h incubation at 30°C (Figure 9B, 1‐PI). Thus, the assembly of LAP2α into larger structures was directly or indirectly dependent on dephosphorylation. Similarly, lamin B became increasingly insoluble upon incubation for 0.5 and 1 h in a dephosphorylation‐dependent manner (Figure 9B). Both LAP2α and lamin B remained mostly soluble in 1% Triton X‐100 plus 500 mM salt at all stages of in vitro assembly, indicating that unspecific aggregation of the protein due to partial denaturation did not occur in the extract.
To test whether the shift of LAP2α from the soluble to insoluble fractions during incubation of mitotic lysates was dependent on the presence of chromosomes, we removed chromosomes by centrifugation prior to starting the in vitro assembly reaction. As depicted in Figure 9C, the protein distributed equally between soluble and insoluble fractions after a 1 h incubation of the chromosome‐free mitotic cell fractions at 30°C. Therefore it can be concluded that LAP2α may also partially assemble into larger protein complexes in the absence of chromosomes.
To study the kinetics of in vitro assembly in more detail, we performed differential centrifugation of whole‐cell lysates after various incubation times and tested for the presence of histones, lamins and nuclear membrane proteins in subcellular fractions. Mitotic NRK cells were homogenized in the absence of phosphatase inhibitors, incubated at 30°C for up to 60 min and centrifuged at 2000 g to sediment large complexes which, according to the presence of histones in Coomassie Blue‐stained samples, contained the majority of chromatin (Figure 10A, P1). The supernatant fraction was then centrifuged at 100 000 g to sediment smaller structures, such as membrane vesicles (P2). Immunoblot analyses of fractions obtained shortly after starting nuclear assembly by increasing the temperature to 30°C revealed that one half of the total cellular LAP2α was found in the low‐speed pellet, the other half in the soluble fraction (Figure 10A, 5 min). In contrast, lamins were still predominantly soluble, except for lamin B which was also found in the membrane‐containing pellet fractions P2 (Figure 10A, P2), probably due to its reported membrane association during mitosis (Gerace and Burke, 1988). These observations confirmed our immunofluorescence data showing that LAP2α assembled around chromosomes prior to the assembly of lamins (see above). As expected, the integral membrane protein LAP1 was detected in the membrane‐containing high‐speed pellet (P2), but was absent from the soluble fraction (Figure 10A). In addition, large amounts of LAP1 were also found in the low‐speed pellet fractions (P1), indicating that nuclear membranes were already attached to chromosomes at this early stage of nuclear reassembly.
Next, we followed the subcellular distribution of LAP2α in the course of in vitro nuclear reassembly, and plotted the relative amounts of LAP2α in supernatant and pellet fractions against the incubation time (Figure 10B). A rapid decrease of the soluble protein was detected within the first 30 min, accompanied by an increase of LAP2α in the low‐speed chromosome pellet (Figure 10B). Interestingly, we also detected an increase of the relative amount of LAP2α in the high‐speed, membrane‐containing pellet fractions within the first 30 min, which declined again upon further incubation (Figure 10B, P2). This result suggested that at least some of LAP2α may associate with membranes before these membranes attach to chromosomes. Immunoblot analyses confirmed that the majority of LAP1 and lamins were also found in the chromosome fraction after 30 min incubation. In addition, these proteins were still present in the high‐speed membrane pellet, indicating that they have not been completely incorporated into the NE and the karyoskeleton at this stage of nuclear assembly. By contrast, almost all of LAP2α, as well as of LAP1 and lamins were found in the low‐speed pellet in interphase cells.
LAP2α interacts with chromosomes in vitro
To test whether LAP2α can interact with chromosomes, we expressed the protein in Escherichia coli and performed in vitro co‐sedimentation assays using metaphase chromosomes isolated from mitotic Chinese hamster ovary (CHO) cells. The soluble bacterial cell lysate contained a high concentration of recombinant protein of the expected molecular weight (Figure 11A), which was also detected by mAb 10 in immunoblot analysis (LAP2; Figure 11B). After incubation of chromosomes with bacterial cell lysates, chromosomes were sedimented through a sucrose cushion, and the pellet fractions were analyzed by immunoblotting using mAb 10. More than half of LAP2α present in bacterial cell lysates was co‐sedimented with chromosomes, whereas only a minor fraction of LAP2α sedimented in the absence of chromosomes (Figure 11B). As chromosomes did not contain detectable amounts of immunoreactive endogenous LAP2α and β (Figure 11B), it can be concluded that sedimented recombinant LAP2α associated with components of the chromosomal fraction. Recombinant LAP2β polypeptide, which contains the entire nucleoplasmic domain but lacks the transmembrane domain and the luminal C‐terminus (Figure 1), was also efficiently co‐sedimented with chromosomes. This confirmed previous results showing an in vitro interaction of purified LAP2β with metaphase chromosomes (Foisner and Gerace, 1993). Immunofluorescence microscopy of chromosomes after incubation with bacterial cell lysates containing LAP2α or LAP2β revealed a uniform staining of the chromosomes with mAb 10. As this antibody did not stain chromosomes alone (data not shown), it can be concluded that both LAP2α and β associated with the chromosomal surface and did not non‐specifically aggregate with non‐chromosomal structures. Considering that both proteins are identical in the first 187 N‐terminal amino acids (Figure 1), one might speculate that this region is responsible for the proteins' interaction with chromosomes. To test this hypothesis, we performed co‐sedimentation and immunofluorescence assays with a recombinant protein covering LAP2 amino acids 1–187 (Figure 11). However, an interaction of the N‐terminal fragment with chromosomes was not found in either assay.
Here we demonstrate the cell‐cycle‐dependent subcellular distribution and biochemical properties of a nuclear protein (LAP2α) that has recently been identified as an alternatively spliced isoform of the inner nuclear membrane protein LAP2 (which we now call LAP2β) (Harris et al., 1994, 1995; Furukawa et al., 1995). Unlike LAP2β and several other LAP2 isoforms, LAP2α lacks a putative transmembrane domain and was found throughout the nuclear interior, suggesting that the LAP2 proteins may be more generally involved in the establishment and maintenance of nuclear structure and not restricted to functions at the nuclear envelope. In addition to LAP2α, LAP2β and the β‐related LAP2γ, which together are the most abundant LAP2 proteins detected in a human T‐cell line (Harris et al., 1994), the possibility of additional β‐related proteins derived by alternative splicing, LAP2δ, LAP2ϵ, and LAP2ζ, has recently been revealed by analysis of the mouse LAP2 gene and by re‐analysis of the human LAP2 gene (Berger et al., 1996; C.Harris, unpublished observation). Although LAP2β has been shown to associate directly with lamins (Foisner and Gerace, 1993), and LAP2α is shown here to co‐localize with lamins at some cell‐cycle stages, a direct interaction with the nuclear lamina has not been demonstrated for LAP2α, or for LAP2γ, δ, ϵ or ζ. One or more of the proteins may not associate directly with the lamina, and revision of the protein name may become appropriate as the molecular functions of each of the proteins become better understood.
LAP2α—a karyoskeletal element?
A currently evolving view is that the nucleus contains a highly structured internal skeletal lattice that can organize chromosomes and numerous other nuclear components into physical and functional subdomains. Morphologically, this karyoskeletal fraction contains the nuclear lamina, attached nuclear pore complexes and in many instances a meshwork of filaments apparently derived from the nuclear interior, poorly defined at the biochemical level. The karyoskeleton is broadly defined as the components that remain after nuclei are extracted with non‐ionic detergents, nucleases and high salt. Applying these operational criteria, it is not entirely clear whether LAP2α can be considered as a genuine component of the karyoskeleton in NRK nuclei. Although the majority of LAP2α in NRK nuclei was found to remain insoluble in buffers containing 1% Triton X‐100, physiological salt concentrations, and nucleases, it was almost completely soluble under high salt conditions (>250 mM). On the other hand, several observations are consistent with the LAP2α function as a structural element of the karyoskeleton. First, upon extraction of NRK nuclei with Triton‐containing buffers, LAP2α behaved exactly like the membrane protein LAP2β, which has previously been identified as a component of the nuclear lamina (Foisner and Gerace, 1993). Secondly, LAP2α was present in NE fractions obtained from rat liver nuclei by extraction with nucleases and high salt (500 mM). Thirdly, unlike in NRK cells, LAP2α‐specific antibodies failed to detect the protein in rat liver nuclei by immunofluorescence microscopy, probably due to limitations of antigen accessibility. Thus, it is possible that the LAP2 proteins might be more tightly incorporated in the nucleoskeleton of differentiated non‐proliferating cells in tissues as compared with the nucleoskeleton of highly proliferating cells in culture.
On the structural level, studies on the nucleoskeleton have always been difficult as it is normally concealed by a much larger mass of chromatin, and although several studies offer compelling evidence for the nucleoskeleton (Goldberg and Allen, 1992; Hozak et al., 1995; Cordes et al., 1997; Nickerson et al., 1997), it is still uncertain whether the scaffold is one continuous element, several discrete elements or an in vitro artefact.
At the light‐ and electron‐microscopic level, LAP2α was found scattered throughout the nuclear interior with no obvious concentration in defined intranuclear structures being absent only from nucleoli. However, the relatively weak staining of interphase nuclei compared with mitotic cells, as well as the low density of LAP2α‐specific label in ultrathin Lowicryl sections in immunoelectron microscopy, indicated that only a fraction of the total cellular protein is detectable by the microscopic techniques in interphase cells, while the rest of the protein may be tightly integrated in higher chromatin structures, inaccessible for the antibodies. The masking of nuclear proteins during interphase, making them undetectable by conventional microscopic techniques, and the unmasking of the respective antigens to immunofluorescence staining upon entry of cells into mitosis has also been reported for other potential nuclear matrix proteins (Nickerson et al., 1992).
Disassembly of LAP2α structures during mitosis
According to our biochemical fractionation experiments LAP2α became completely soluble during mitosis and was not apparently associated with large cellular structures, such as chromosomes or membrane vesicles. However, it remains unclear whether LAP2α associates with other proteins during mitosis. The membrane‐bound protein LAP2β has also been shown to dissociate from chromosomes during mitosis, and the lack of an in vitro interaction of mitotically phosphorylated LAP2β with lamin B (Foisner and Gerace, 1993) suggested that LAP2β did not interact with lamins during mitosis. This is in contrast to other lamina‐associated proteins of the inner membrane, such as LBR and LAP1, which have been suggested to exist in multimeric complexes with lamins and possibly other proteins during mitosis (Meier and Georgatos, 1994; Maison et al., 1997).
Although at the biochemical level we did not detect any interaction of LAP2α with cytoskeletal proteins in mitotic cells, immunofluorescence microscopy showed a concentration of LAP2α in the areas of the mitotic spindle apparatus, particularly at spindle poles. The biological significance of this is not clear, but similar distributions have been reported for other nuclear proteins such as LAP1 (Maison et al., 1997), NuMA (for review see Compton and Cleveland, 1994), protein 4.1 (Krauss et al., 1997) and a structural nuclear matrix antigen (Nickerson et al., 1992). The recently reported association of LAP1 with the mitotic spindle apparatus (Maison et al., 1997) has been suggested to play a role in membrane partitioning during cell division, separating a subset of nuclear membrane vesicles from other nuclear and endoplasmic reticulum‐derived vesicles. However, these data are controversial as more recent studies have demonstrated the diffusion and distribution of nuclear membrane proteins throughout the cytoplasmic membrane system at the onset of mitosis (Ellenberg et al., 1997; Yang et al., 1997a).
NuMA has been found to play an integral role in the establishment and maintenance of nuclear structure by antibody microinjection (Kallajoki et al., 1993) and NuMA transfection (Compton and Cleveland, 1993) experiments, and it has recently been shown to play a crucial role in the assembly of the mitotic spindle itself (Gaglio et al., 1995; Merdes et al., 1996). In addition, as NuMA associates with microtubules during metaphase and progressively accumulates at the polar ends of the mitotic spindle, when cells proceed to anaphase, it has been suggested that the association of NuMA and possibly other nuclear proteins with the mitotic spindle apparatus may provide a mechanism to distribute the proteins equally between daughter cells. This mechanism of post‐mitotic nuclear accumulation is distinct from the passive diffusion and active nuclear transport of most other nuclear proteins, including lamins, and might be particularly important for proteins that are essential for early stages of nuclear reassembly and that associate with chromosomes prior to nuclear membrane formation. Therefore, considering the potential role of LAP2α in early stages of nuclear reassembly (see below), its association with the mitotic spindle may ensure an efficient interaction of LAP2α with chromosomes at the end of mitosis. It remains unclear whether LAP2α can directly interact with microtubules. Based on our observations that a significant fraction of the protein associated with the mitotic spindle and that some of the protein ended up in the midbody, it might be speculated that LAP2α may transiently gain microtubule binding activity, probably similar to the previously reported chromosome passenger proteins (Earnshaw and Bernat, 1990).
As for the regulation of LAP2α relocalization, it is very likely that, similar to lamins, LAPs and LBR (reviewed in Foisner, 1997), mitosis‐specific phosphorylation is one of the key mechanisms. Although the effect of phosphorylation on LAP2α dynamics has not yet been shown directly, a number of observations argue for a phosphorylation‐dependent mechanism. There is a perfect temporal correlation of phosphorylation and solubilization of the protein in both our subcellular fractionation and our in vitro assembly studies. Furthermore, the dephosphorylation and the assembly of LAP2α into insoluble structures in our in vitro nuclear assembly studies were inhibited by phosphatase inhibitors. Two‐dimensional gel electrophoresis revealed at least five mitosis‐specific spots, suggesting that a minimum of five potential phosphorylation sites are targeted during mitosis. It remains to be analyzed whether these differentially phosphorylated LAP2α proteins exist simultaneously in the mitotic cell or whether specific sites are phosphorylated at different stages of mitosis. Alternatively, a partial dephosphorylation might occur during cell lysis and sample preparation. Nevertheless, the presence of various consensus sites for p34cdc2‐dependent phosphorylation in the primary sequence of the polypeptide (Figure 1) is consistent with a mitosis‐specific phosphorylation of the protein.
Potential role of LAP2α in nuclear assembly
The assembly of LAP2α structures at distinct sites around chromosomes during early stages of post‐mitotic nuclear reassembly, when the majority of lamins was not yet assembled, suggests a role for LAP2α in the establishment of nuclear structure. Its inter‐chromosomal localization would be consistent with a function of LAP2α in tethering together the telophase bundle of chromosomes. It is unclear whether LAP2α associates with chromosomes prior to the formation of a closed nucleus, or whether LAP2α, which contains a putative nuclear localization sequence (Figure 1), enters the nucleus by active transport through newly formed pore complexes. The early association of LAP2α with chromosomes and its nuclear accumulation before lamina assembly, as well as the concentration of LAP2α at the spindle poles in close vicinity to the separated sets of chromosomes, argue for a direct access of LAP2α to chromosomal structures prior to nuclear membrane formation at least at this cell cycle stage. The demonstrated in vitro interaction of LAP2α with isolated chromosomes is also consistent with this model. Nuclear localization signal (NLS)‐mediated nuclear transport of LAP2α might be relevant for newly synthesized protein during interphase, although the functional significance of the NLS in LAP2α has not yet been shown directly.
The nature of the molecular interactions between LAP2α and chromosomes remains to be solved. Neither the chromosomal components interacting with LAP2α, nor the molecular domains of LAP2α involved in the interaction have yet been identified. Our observation that both recombinant LAP2α and the nucleoplasmic domain of LAP2β interacted with chromosomes in vitro would suggest that the chromosome interaction site of the protein is located within their identical N‐termini, consistent with a recently reported chromatin binding site in the N‐terminal 85 amino acids of LAP2β (Yang et al., 1997b; Furukawa et al., 1998). However, the N‐terminal LAP2 fragment (amino acids 1–187) alone failed to interact with chromosomes in our assays. There are several explanations for this apparent inconsistency. The recombinant bacterially expressed protein may lack certain post‐translational modifications or may be misfolded, and thus unable to interact with chromosomes. Alternatively, LAP2α and LAP2β may contain additional chromosome‐binding sites in their C‐terminal regions. In line with this hypothesis, the interaction of LAP2β with chromosomes has recently been suggested to involve multiple domains of the protein, located in the β‐specific region (Furukawa et al., 1997).
Interestingly, we found a co‐localization of LAP2α with lamin B at discrete structures at the nuclear periphery and with lamin A in intranuclear structures in late anaphase and G1 phase. This might indicate a direct or indirect association of at least some of the cellular LAP2α with lamins at specific cell‐cycle stages. The close association of a subset of LAP2α‐specific label with the NE in immunoelectron microscopy and the association of LAP2α with lamin‐containing cell fractions in the in vitro nuclear assembly studies and in subcellular fractionation experiments are consistent with a LAP2α‐lamin interaction. Furthermore, we observed an in vitro interaction of recombinant LAP2α with purified lamins (our unpublished results).
Overall, our data indicate that LAP2α is a nuclear protein that is involved in the structural organization of the nucleus and in the post‐mitotic nuclear assembly. Functional studies involving the inducible expression of mutated LAP2 proteins in mammalian cells are in progress to unravel the specific function(s) of LAP2α in the various steps of nuclear assembly, involving membrane targeting, chromosome decondensation, lamin assembly and nuclear growth.
Materials and methods
Cell culture and synchronization
NRK, CHO and HeLa cells were routinely maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS), 50 μg/ml of penicillin and streptomycin (all from Life Technologies, Paisley, UK) at 37°C in a humidified atmosphere containing 5% CO2. For synchronization of cell growth, NRK and CHO cell cultures were arrested in G1/S phase by an overnight incubation in complete DMEM or in Joklik's modified minimum essential medium (S‐MEM, Life Technologies) with 10% FCS, a non‐essential amino acid mix and penicillin, streptomycin, plus 2 mM thymidine. In some cases a second overnight incubation in thymidine‐containing medium was applied following a 10–12 h release from the block in complete medium without thymidine. To select S‐phase cells, cultures were released from the thymidine block for 4 h. For the collection of mitotic cells, cultures were released from the thymidine block for 4 h and cultivated for an additional 4–6 h in growth medium containing 0.2 μg/ml nocodazole (Calbiochem‐Behring Corp., La Jolla, CA). Weakly attached mitotic cells were harvested by mechanical shake off. HeLa cells were incubated for 16–24 h in medium containing 0.2 μg/ml nocodazole without preincubation in thymidine‐containing medium. For FACS analyses, ∼5×105 cells in 1 ml phosphate‐buffered saline (PBS) were fixed in 4 ml of 85% ethanol at −20°C, and cell pellets were briefly treated with 0.05% aqueous pepsin, stained with 2 μg/ml 4,6‐diamidino‐2‐phenylindole and 15 μg/ml sulforhodamine (Sigma‐Aldrich Chemie GmbH, Deisenhofen, Germany), in 100 mM Tris–HCl pH 8.0, 2 mM MgCl2, 0.1% Triton X‐100, and analyzed with a Partec PAS II flow cytometer (Partec, Münster, Germany).
Cell fractionation and in vitro assembly
Cells were incubated in ∼5 vol. ice‐cold hypotonic buffer (10 mM HEPES pH 7.4, 5 mM MgCl2, 10 mM NaCl, 1 mM DTT) containing the protease inhibitors PMSF (1 mM), benzamidine (2 mM), and aprotinin, leupeptin, and pepstatin (2 μg/ml each), and homogenized in a glass–glass homogenizer by 100 strokes with a tight pestle. After addition of 8% sucrose, the soluble cytoplasmic and the insoluble nuclear fractions were separated by centrifugation at 2000 g for 20 min at 4°C. The nuclei‐containing pellets were extracted in the same buffer containing either 1% Triton X‐100 and 50–250 mM NaCl or 7 M urea, and centrifuged at 2000 or 100 000 g to yield a low‐speed and a high‐speed pellet, and a high‐speed supernatant fraction.
For analyzing solubility of proteins in interphase versus mitotic cells, synchronized cell cultures were homogenized in ice‐cold homogenization buffer (50 mM HEPES pH 7.4, 4 mM MgCl2, 10 mM EGTA, 100 mM NaCl, 0.1 mM DTT) containing protease inhibitors, 20 μM cytochalasin B (Sigma‐Aldrich Chemie GmbH), phosphatase inhibitors (0.1 μM calyculin, 8 μM microcystin, 0.1 μM okadaic acid, 1 mM Na‐pyrophosphate, 0.5 mM β‐glycerophosphate) and kinase inhibitors (0.1 μM olomoucine, 0.5 μg/ml staurosporine, 0.5 mM H7) (Sigma‐Aldrich Chemie GmbH, Life Technologies, and Calbiochem‐Behring Corp.). Samples were centrifuged at 15 000 g for 10 min without prior treatment, or after the addition of 1% Triton X‐100 and/or 500 mM NaCl, and/or 500 μg/ml DNase and 200 μg/ml RNase (Boehringer Mannheim, Germany).
For in vitro nuclear assembly, mitotic NRK or HeLa cells were homogenized in homogenization buffer without phosphatase and kinase inhibitors, and cell lysates were incubated for up to 2 h at 30°C. Phosphatase and kinase inhibitors were then added and samples were centrifuged at 2000 or 100 000 g for 20 min with or without prior addition of 1% Triton X‐100 and/or 500 mM salt. In some cases, in vitro assembly was performed in KHM buffer as described previously (Burke and Gerace, 1986). To test the assembly of LAP2α in the absence of chromosomes mitotic cell lysates were centrifuged for 10 min at 15 000 g and supernatants were used in the assembly reaction.
NEs and microsomal membranes were isolated from rat liver, and ‘salt washed’ and extracted with urea, Triton X‐100 and various salt concentrations as described previously (Foisner and Gerace, 1993).
Isolation of metaphase chromosomes
Nocodazole‐arrested CHO cells obtained from ten roller bottles were incubated in complete medium containing 20 μM cytochalasin B and 0.2 μg/ml nocodazole for 30 min at 37°C, preswollen in 5 mM HEPES pH 7.4, 5 mM NaCl, 5 mM MgCl2, 0.5 mM EDTA, 1 mM DTT and protease inhibitors (chromosome buffer) for 10 min on ice, and homogenized in the same buffer containing 0.1% Triton X‐100. The cell lysate was layered on top of a discontinuous sucrose gradient (3 ml of 20, 30, 40, 50 and 60% sucrose in chromosome buffer), centrifuged for 10 min at 1500 r.p.m. (Megafuge 1.0R, Heraeus Sepatech GmbH., Osterode, Germany) at 4°C, and chromosomes enriched in a diffuse band in the middle of the gradient were collected.
Expression and isolation of recombinant LAP proteins
The construction of plasmids pET17b‐LAP2α (formerly called pET 17b‐hTPα) is described elsewhere (Harris et al., 1994); pET23a‐LAP2β 1–408 and pET23a‐LAP2 1–187 were constructed by PCR amplification of sequences encoding LAP2β amino acids 1–408 and LAP2αβγ amino acids 1–187 (Harris et al., 1994), and insertion into pET 23a (Novagen Inc., Madison, WI) via NdeI and XhoI restriction sites. Recombinant proteins were expressed in E.coli BL21 (DE3) using the inducible T7 RNA polymerase‐dependent pET vector system (Novagen). Protein expression was induced with isopropyl‐β‐d‐thiogalactopyranoside (IPTG) for 2–4 h, and bacteria were harvested by centrifugation at 4000 r.p.m. for 5 min (Heraeus Megafuge 1.0R). Bacteria were frozen in 1/10 of the original culture volume of Tris buffer (20 mM Tris–HCl pH 8.0, 500 mM NaCl, 2 mM EGTA, 1 mM DTT, protease inhibitors), thawed, and lysed by addition of 0.1 mg/ml lysozyme, 0.1% Triton X‐100, 10 mM MgCl2, 50 μg/ml DNase, and 20 μg/ml RNase and incubation for 30 min at 30°C. Following the addition of 7 M urea and homogenization in a glass–glass homogenizer, cell lysates were spun at 4000 r.p.m. (Heraeus Megafuge) for 10 min and subsequently at 35 000 r.p.m. (TLA 45 rotor, Beckmann Instruments Inc., Palo Alto, CA) for 30 min and supernatants were stored at −20°C.
Chromosome binding assays
Soluble bacterial lysate fractions were dialyzed against binding buffer (20 mM HEPES pH 7.4, 150 mM NaCl, 5 mM MgCl2, 2 mM EGTA, 0.2% Triton X‐100, 1 mM DTT and protease inhibitors), incubated for 30 min at 30°C and centrifuged in an Eppendorf microfuge for 30 min at 13 000 g. The supernatants were analyzed by SDS–PAGE and diluted with binding buffer to obtain similar concentrations of recombinant proteins. An aliquot (300 μl) of the soluble fraction was mixed with 150 μl of chromosome buffer or chromosomes, incubated for 30 min at 30°C and centrifuged through 100 μl of a 35% sucrose cushion in binding buffer for 10 min at 4000 r.p.m. (Heraeus Megafuge 1.0R). Pellet fractions were analyzed by SDS–PAGE and immunoblotting.
For detection of bound protein on individual chromosomes, 30 μl of the chromosome–protein mixture was placed on adhesion slides (Bio‐Rad, Hercules, CA), and attached complexes were analyzed by immunofluorescence.
Immunofluorescence and immunoelectron microscopy
Cells grown on plastic Petri dishes or chromosomes attached to adhesion slides were fixed with 3.7% formaldehyde in PBS for 20 min at room temperature with or without prior permeabilization in PBS plus 0.1% Triton X‐100 for 5 min. After incubation in 50 mM NH4Cl in PBS and permeabilization of the fixed cells in PBS/0.1% Triton X‐100 for 5 min each, samples were incubated in PBS plus 0.2% gelatine for 30 min. Primary and secondary antibodies were applied in PBS plus gelatine for 1 h each at room temperature. Primary antibodies used were hybridoma supernatants (undiluted), purified LAP2 antibodies (5–10 μg/ml) (Harris et al., 1994), affinity‐purified guinea pig anti‐lamin B antibodies (10 μg/ml), and guinea pig anti lamin A/C antibody (1:100) (Ottaviano and Gerace, 1985; Foisner and Gerace, 1993), secondary antibodies, goat anti‐mouse IgG conjugated to Bodipy FL (Molecular Probes, Leiden NL), and goat anti‐guinea pig IgG conjugated to Texas Red (Accurate Chemicals & Scientific Corp., Westbury, NY). After several washes in PBS, DNA was stained with 1 μg/ml Hoechst dye 33258 (Calbiochem‐Behring Corp.) for 10 min, or RNA was digested with 10 μg/ml DNase free RNase for 30 min at room temperature, and DNA was stained in 0.1 μg/ml propidium iodide (Sigma‐Aldrich Chemie GmbH) in PBS for 10 min. Samples were mounted in Mowiol and viewed in a Zeiss Axiophot microscope and a MRC‐600 confocal microscope (Bio‐Rad).
For immunolocalization of LAP2 isoforms in rat liver tissue, frozen tissue sections (5–10 μm) cut on a cryotome were fixed and stained as described above, except that samples were treated with 10 μg/ml DNase for 30 min after fixation, 0.2% gelatine, 1% bovine serum albumin (BSA), and goat serum (1:50 in PBS) were used for blocking, and secondary antibodies were preabsorbed on rat liver acetone powder.
For immunoelectron microscopy, HeLa cells were fixed for 1 h in 4% paraformaldehyde, suspended in 10% gelatine at 37°C and chilled on ice. Small cubes of the samples were dehydrated in ethanol at progressively lower temperatures, embedded in Lowicryl K4M (Agar Scientific Ltd, Stansted, UK), and polymerized at −35°C by UV light (Carlemalm et al., 1982). Ultrathin sections were cut on a Reichert Ultracut S ultra‐microtome (Reichert Division Leica AG, Vienna, Austria), and processed for immunolabeling using 2% BSA plus 0.2% gelatine in PBS as blocking reagent, anti‐LAP2α mAb 15 at 10 μg/ml, and goat anti‐mouse IgG coupled to 10 nm gold particles (British BioCell International, Cardiff, UK; diluted 1:20). Sections were stained with uranyl acetate and lead citrate, and viewed in a JEOL 1210 transmission electron microscope at 80 kV.
Mitotic or interphase HeLa cells were lysed in homogenization buffer containing 1% Triton X‐100, 500 μg/ml DNase, 200 μg/ml RNase, and protease‐, kinase‐ and phosphatase inhibitors. Following a 10 min incubation at room temperature, 1% SDS was added to the samples. After 10‐fold dilution with 50 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% Triton X‐100 and 5 mM EDTA the samples were centrifuged for 10 min in an Eppendorf centrifuge, and incubated with 100 μl 10% (w/v) protein A–Sepharose (Pharmacia Biotech, Uppsala, Sweden). Following removal of the beads by centrifugation in an Eppendorf microfuge, supernatants (1 ml) were incubated with 5–10 μg mAb 15 by end‐over‐end rotation for 15–20 h at 4°C. After addition of 100 μl protein G–Sepharose beads (Sigma‐Aldrich Chemie GmbH) samples were incubated for an additional 2 h, and beads were pelleted through 30% sucrose at 1000 r.p.m. (Heraeus Megafuge 1.0R) for 2 min at 4°C, washed with buffer and prepared for two‐dimensional gel electrophoresis.
Gel electrophoresis and immunoblotting
One‐dimensional SDS–PAGE was performed according to Laemmli (1970). Two‐dimensional gel electrophoresis was performed as previously described (Gotzmann et al., 1997). For immunoblotting, proteins separated on one‐ or two‐dimensional gels were electrophoretically transferred onto nitrocellulose (0.2 μm Schleicher and Schuell, Inc., Dassel Germany) in 48 mM Tris–HCl pH 9.4, 39 mM glycine using the Mini Transblot system (Bio‐Rad). For the immunological detection of proteins the Protoblot Immunoscreening System (Promega, Madison, WI) or the SuperSignal detection system (Pierce, Rockford, IL) were used. Primary antibodies used were anti‐LAP2 antibodies (hybridoma supernatants, undiluted), anti‐LAP1 (Senior and Gerace, 1988; RL13 ascites fluid, diluted 1:1000), antiserum to lamin B (Foisner et al., 1991; diluted 1:1000) and antiserum to lamins A/C (Glass et al., 1993; diluted 1:1000). Quantification of stained protein bands was done using the NIH image software.
We wish to thank Christine Stadler for purification of the antibodies and for technical assistance with some of the experiments; Thomas Sauer, University of Vienna for performing FACS analyses; Karen Chan for construction of vectors for expression of recombinant LAP2 proteins; Sylvia Vlcek for providing some of the immunoblots shown in Figure 9; and Allison Witte, Georgetta Denhardt and Laura Abriola for preparation of monoclonal antibodies. Furthermore, we thank Larry Gerace, at The Scripps Research Institute, La Jolla, CA, USA, for his generous gifts of lamin and LAP1 antibodies; and Peter Traub, of the Max‐Planck‐Institute, Heidelberg, Germany, for providing lamin B antiserum. This study was supported by grants from the Austrian Science Research Fund (FWF) and from the Jubiläumsfonds der Österreichischen Nationalbank to R.F.
- Copyright © 1998 European Molecular Biology Organization