Summary
The transmembrane inner nuclear membrane (INM) protein Samp1 is required for anchoring centrosomes near the nuclei. Using high-resolution fluorescence microscopy we show that Samp1 is distributed in a distinct and characteristic pattern in the nuclear envelope (NE), where it partially colocalizes with the LINC complex protein Sun1. By studying the localization of Samp1 deletion mutants and fusion proteins, we conclude that the cysteine-rich N-terminal half of Samp1 is nucleoplasmically exposed and is responsible for targeting to the INM. It contains four conserved CxxC motifs with the potential to form zinc fingers. The distribution of cysteine-to-alanine substitution mutants, designed to prevent zinc finger formation, showed that NE localization of Samp1 depends on intact CxxC motifs. Overexpression of Samp1 zinc finger mutants produced an abnormal dominant phenotype characterized by disrupted organization of a selective subset NE proteins, including emerin, Sun1, endogenous Samp1 and, in some cases, lamin A/C, but not lamin B, Sun2 or nucleoporins. Silencing of Samp1 expression showed that emerin depends on Samp1 for its correct localization in the NE. Our results demonstrate that Samp1 is functionally associated with the LINC complex protein Sun1 and proteins of the A-type lamina network.
Introduction
The nuclear envelope (NE) separates the nucleoplasm from the cytoplasm and consists of two concentric lipid membranes: the nuclear pores and the nuclear lamina (Hetzer et al., 2005; Stewart et al., 2007). The outer nuclear membrane (ONM) is continuous with the rough endoplasmic reticulum and has attached ribosomes. The inner nuclear membrane (INM) faces the nuclear lamina and chromatin, and has been estimated to contain approximately 80 unique membrane proteins (Schirmer et al., 2003). At numerous points, the INM and ONM fuse to create circular pore openings, which harbor the nuclear pore complexes (NPCs).
Investigations in different labs have shown that proteins of the nuclear lamina and the inner nuclear membrane are not evenly localized in the periphery of mammalian nuclei, but instead distribute in distinct but overlapping networks in the NE (Liu et al., 2007; Lu et al., 2008; Maeshima et al., 2006; Shimi et al., 2008). These networks might serve as scaffolds or fundaments for the assembly of a large variety of functional micro domains.
Recently, trans-cisternal bridges across the two concentric lipid membranes of the NE have been discovered and named linker of nucleoskeleton and cytoskeleton (LINC) complexes (Burke and Roux, 2009; Starr, 2009; Wilson and Berk, 2010; Wilson and Foisner, 2010). LINC complexes form by interactions in the perinuclear space between KASH domains of proteins traversing the ONM (nesprins) and the SUN domains of proteins traversing the INM (Sun proteins). The LINC complex is conserved from yeast to mammals and has been found to have essential roles in many vital cell functions including cell polarization, nuclear migration and positioning (Lei et al., 2009; Starr, 2009; Zhang et al., 2009). The LINC complex has even been suggested to mediate mechanotransduction from the cell surface to the nuclear interior, affecting chromatin organization and gene regulation (Wang et al., 2009). Determination of how exactly Sun domain and nesprin protein isoforms, INM proteins and nuclear lamins are organized to form different LINC complex protein networks remains to be determined (Haque et al., 2010).
In X-EDMD (X-linked Emery–Dreifuss muscular dystrophy) cells, or in cells treated with siRNA against emerin, centrosomes detached from the NE (Salpingidou et al., 2007), and emerin is able to bind the two LINC complex proteins, Sun1 and Sun2 (Haque et al., 2010). Anchoring of centrosomes close to the nucleus was also shown to depend on the novel transmembrane INM protein Samp1 (transmembrane protein 201) (Buch et al., 2009), suggesting that Samp1 is also connected to LINC complexes. As highlighted in a recent review (Wilson and Berk, 2010), anchorage of the microtubule cytoskeleton at the nuclear surface is particularly important for the relative positioning of centrosomes versus nuclei during cell migration (Gomes et al., 2005) or during interkinetic nuclear migration when neuronal cells divide (Burke and Roux, 2009; Starr, 2009) in the hippocampus and cerebellum of the developing mammalian brain. Both processes involve the LINC complex components Sun1, Sun2, nesprin-1 and nesprin-2. In G1, nesprin-1 was shown to connect to the plus-end-directed motor protein kinesin when pushing nuclei away along microtubules from the centrosome in the apical pole of the cell. In G2, nesprin-1 and nesprin-2 connect to the minus-end motor protein dynein to pull nuclei along microtubules back to the centrosome before mitosis (Zhang et al., 2009). Furthermore, a diverse group of diseases called laminopathies (for reviews, see Broers et al., 2006; Gomes et al., 2005; Wilson and Berk, 2010; Worman and Bonne, 2007; Worman et al., 2010) are linked to genes encoding proteins at different locations along the cytoskeleton to nucleoskeleton axis including nesprins, INM proteins and lamins. EDMD can arise from mutations in different genes: EMD (which encodes emerin), LMNA (which encodes lamin A/C), SYNE1 (nesprin-1) or SYNE2 (nesprin-2). These proteins all interconnect, suggesting a common disease mechanism acting on the link between the nucleoskeleton and the cytoskeleton, and illustrating the importance of proper LINC complex composition for healthy human development.
To test the hypothesis of a connection between Samp1 and LINC complexes, we have investigated the role of conserved zinc finger motifs in Samp1 for the organization of Samp1 and other NE proteins. Here we present data showing that Samp1 zinc finger mutations specifically affect the distribution of emerin, Sun1 and lamin A/C, tying Samp1 function to a LINC complex protein network.
Results
Samp1 distributes in distinct micro-domains of the inner nuclear membrane
We have previously shown that Samp1 is an integral membrane protein located in the inner nuclear membrane of many cultured cells (Buch et al., 2009). Careful inspection by confocal fluorescence microscopy of equatorial, as well as oblique, optical sections of HeLa cell nuclei stained using antibodies specific for Samp1 revealed that Samp1 was not uniformly distributed in the INM. The immunofluorescence decorated the nuclear periphery in a diffuse dotty pattern. This prompted us to investigate the distribution of Samp1 relative to other NE markers using methods with higher resolving power. For this, we performed deconvolution on image stacks along the z-axis acquired by CLSM. Double-labeled nuclei of fixed and permeabilized HeLa cells revealed that the immunofluorescence staining of NE protein markers distributed in characteristic patterns showing variable degrees of overlap (Fig. 1), indicative of the existence of partially overlapping networks of NE proteins, consistent with previous studies (Liu et al., 2007; Lu et al., 2008; Maeshima et al., 2006; Shimi et al., 2008). We determined the degree of colocalization statistically by Pearson analysis using dual-channel image data from an optical plane corresponding to the lower nuclear surface.
To get a quantitative measure of the scope between colocalizing objects and separate structures with a low degree of overlap, we analyzed images of nuclei double stained with well-characterized markers of the NE. To obtain a quantitative measure of a high degree of colocalization, we used nuclei of cells coimmunostained with the nuclear-pore-specific antibody mAb414, and antibodies specific for the pore membrane protein POM121. Both antibodies gave rise to finely punctuated patterns over the nuclear surface, characteristic of nuclear pores (Fig. 1a–c). The patterns displayed a high degree of colocalization and a Pearson coefficient of 0.68 (Fig. 1d). As a measure for staining of separate but partially overlapping structures (Shimi et al., 2008), we used images of nuclei coimmunostained with antibodies specific for lamin A/C and lamin B, respectively. In this case, two less finely dotted, but different patterns over the nuclear surface were produced (Fig. 1e–g), resulting in a Pearson coefficient of 0.08 (Fig. 1h). In the nuclear periphery and surface wrinkles, the different patterns merged locally as a result of the intrinsic lower resolution along the z-axis.
Samp1 was distributed in a distinct pattern, showing partial colocalization with Sun1 (Pearson coefficient=0.24, Fig. 1u–x), but only limited overlap with nuclear pores (Fig. 1i–l), lamin A/C (Fig. 1m–p), lamin B (Fig. 1q–t), emerin or Sun 2 (supplementary material Fig. S1). The colocalization of Samp1 and Sun1 suggests that a fraction of Samp1 interacts with Sun1 containing LINC complexes.
Membrane topology and functional domains of Samp1
To identify functional domains of Samp1, we first designed a series of deletion mutants fused to YFP and studied their intracellular distribution after expression in HeLa cells. Similarly to the endogenous protein (Buch et al., 2009), YFP–Samp1 (Fig. 2A) and Samp1–YFP (not shown) distributed in the nuclear rim. A truncated variant YFP–Samp1(1–236), lacking the C-terminus and the last three of its five hydrophobic domains, also localized to the nuclear rim (Fig. 2A). Surprisingly, the deletion mutants YFP–Samp1(1–210) and YFP–Samp1(1–180), which contained only the first hydrophobic domain close to the N-terminus (amino acids 15–35), displayed diffuse nucleoplasmic distribution, indicating that the fusion protein was no longer inserted in the lipid membrane. To rule out that this was due to inefficient membrane insertion caused by the upstream positioning of YFP, we designed a corresponding fusion protein with GFP fused to the C-terminus. Because Samp1(1–180)–GFP also localized to the nucleoplasm (Fig. 2Ai,j), we conclude that the N-terminally located hydrophobic stretch of Samp1 does not form a transmembrane domain. This is in agreement with the membrane topology of Inner MAS protein 1 (Ima1; SPCC737.03c), the homologue of Sampl in fission yeast (King et al., 2008) and raises the question of the function of the N-terminal hydrophobic sequence, which is conserved from yeast to man. Apparently, the N-terminal domain (amino acids 1–210 of Samp1) needs to be linked to a membrane-spanning segment either in its N-terminus or its C-terminus to properly distribute in the INM. This is supported by the facts that the chimera with residues 35–210 of Samp1 replacing the cytoplasmically exposed C-terminal tail of the bitopic plasma membrane protein CD8 (Buch et al., 2009) as well as YFP–Samp1 (1–236), ending with a stretch comprising at least 19 hydrophobic amino acids, both distributed in the nuclear rim (Fig. 2A). Thus, both the entire N-terminal domain (comprising roughly half of the amino acid residues of Samp1) and the short C-terminal tail (Buch et al., 2009) are exposed in the nucleoplasm (Fig. 2B).
Investigation of conserved cysteines of the zinc finger motifs in the N-terminal domain of Samp1
Samp1 contains eight conserved cysteines organized as four CxxC motifs, which have potential to form two zinc fingers (Fig. 3A). For simplicity, we term these hypothetical zinc finger domains ZF1 (residues 48–70) and ZF2 (residues 121–167), respectively. To investigate the role of the cysteine motifs in Samp1 we designed cysteine-to-alanine substitution mutations of YFP-tagged Samp1 aimed at disrupting the potential formation of Zinc fingers. When expressed in HeLa cells, YFP–Samp1(C121A), YFP–Samp1(C48A) and YFP–Samp1(C48A,C121A) were all mislocalized and distributed in cytoplasmic membranes (Fig. 3B). The results show that mutations designed to disrupt ZF1, ZF2 or both, destroy the ability of Samp1 to locate in the NE.
We also noted that the nuclear morphology was affected in cells expressing any of the cysteine to alanine substitution mutants and decided to investigate this effect in further detail using a series of specific antibody NE markers (Fig. 4). Instead of a smooth spherical appearance, the nuclear rim displayed a distorted irregular shape reminiscent of nuclei of cells expressing some genes with laminopathic mutations (Dechat et al., 2008). Immunostaining showed that, whereas overexpression of wild-type YFP–Samp1 had no effect on nuclear morphology or NE localization of emerin (Fig. 4A), emerin and Sun1 were completely mislocalized in cells expressing YFP–Samp1(C121A) (Fig. 4A), YFP–Samp1(C48A) (supplementary material Fig. S2) or YFP–Samp1(C48A,C121A) (supplementary material Fig. S3). Instead of a characteristic distribution in the nuclear rim, both emerin and Sun1 appeared in a number of aggregates in the cytoplasm (Fig. 4A). Interestingly, lamin B displayed a continuous staining in the periphery of nuclei despite their irregular shape (Fig. 4A and supplementary material Figs S2, S3). Also NPCs still located in the NE, although the staining intensity was sometimes affected. This was also true for Sun2 in cells expressing YFP–Samp1(C121A). However, in cells expressing the C48A mutant (supplementary material Fig. S2) or the C48A,C121A double mutant (supplementary material Fig. S3) Sun2 often displayed less-intense staining and partial mislocalization. In cells expressing YFP–Samp1(C121A), the lamin A/C localization to the NE was also disrupted (Fig. 4B), but this effect was suppressed in cells expressing the double mutant YFP–Samp1(C48A,C121A) (Fig. 4B). Lamin-A/C was also normally distributed in cells expressing YFP–Samp1(C48A) (supplementary material Fig. S2). Although mAB414 staining decorated the nuclear rims of cells expressing all three mutants, signs of NPC clustering were evident.
Samp1 distributes in a distinct pattern in the NE. HeLa cells were double labeled with antibodies specific for Samp1 and marker proteins of the NE. Confocal z-stacks of optical sections from two different channels were subjected to deconvolution. Two deconvolved optical sections of the lower surface of nuclei from two different channels are shown separately or overlaid with a pixel correlation graph. anti-Pom121 (a) and mAb414 (c) antibodies give rise to a similar punctated pattern (b), with a high degree of correspondence and a Pearson coefficient of 0.68 (d). By contrast, lamin A/C (e) and lamin B (g) localize with different patterns (f) with a limited degree of correspondence and a Pearson coefficient of 0.08 (h). The distribution pattern of Samp1 (k,m,q) shows little or no overlap with that of Pom121 (i), lamin A/C (o) or lamin B (s), with Pearson coefficients of 0.03 (l), 0.08 (p) and 0.00 (t), respectively. By contrast, Samp1 partially colocalized with Sun1 (u–w) displaying a Pearson coefficient of 0.24 (x).
We have previously shown that a chimera with residues 35–210 of Samp1, replacing the cytoplasmically exposed C-terminal tail of the type 1 bitopic plasma membrane protein CD8, was correctly targeted to the NE (Buch et al., 2009). Therefore, we generated a substitution mutant designed to disrupt ZF2 of this chimera, CD8–Samp1(35–210,C121A) to enable simultaneous analysis of the distribution of the chimera and endogenous Samp1, as well as to determine necessary domains for the effect exerted by the Samp1 zinc finger mutants. As expected, this construct was mislocalized to cytoplasmic membranes and gave rise to a similar abnormal phenotype as the corresponding full-length variant YFP–Samp1(C121A). In cells transfected with CD8–Samp1(35–210,C121A), lamin B was still distributed in the NE (not shown) despite the irregular shape of the nuclear periphery. By contrast, emerin, Sun1 and, to some extent, lamin A/C were mislocalized in cytoplasmic aggregates (Fig. 5A). In addition, endogenous Samp1 (Fig. 5B) appeared in cytoplasmic aggregates, demonstrating the dominant-negative nature of the C121A substitution on NE protein organization. Nuclei of cells expressing CD8–Samp1(35–210) had a normal appearance (not shown) (Buch et al., 2009). The results show that the phenotypic effect caused by the C121A substitution is independent of both the hydrophobic C-terminal half and the first 35 amino acids of Samp1. Analysis by immunoelectronmicroscopy revealed that organization of chromatin was distorted in the nuclear periphery manifested as long stretches of the INM free of heterochromatin in cells expressing CD8–Samp1(35–210,C121A) (Fig. 5C), but not in cells expressing CD8–Samp1(35–210) (Fig. 5D).
The first hydrophobic segment of Samp1 is not a membrane-spanning domain. (A) Fluorescence micrographs and corresponding phase-contrast images of HeLa cells expressing YFP–Samp1 (full length) (a,b), YFP–Samp1 (1–236) (c,d), YFP–Samp1 (1–210) (e,f), YFP–Samp1 (1–180) (g,h) and Samp1 (1–180)–GFP (i,j). Scale bars: 10 μm. Schematic outline of the fusion proteins with hydrophobic segments (white and brown boxes) are shown on the left. (B) Different domains of Samp1: a working hypothesis. Section of the inner nuclear lipid membrane, the nuclear lamina network and the Samp1 polypeptide (green). The N-terminal half is exposed in the nucleoplasm, contains a non-membrane spanning hydrophobic domain (blue) and eight conserved cysteines (yellow) that form two hypothetic zinc fingers (ZF1 and ZF2). The C-terminal portion contains four transmembrane domains and a short tail exposed in the nucleoplasm.
The data presented in Figs 4 and 5 show that expression of substitution mutants designed to disrupt ZF1 and/or ZF2 of Samp1 cause abnormal NE organization and impaired distribution of a specific subset of NE proteins, including emerin and the LINC complex protein Sun1. In the case of the C121A substitution, endogenous Samp1 and lamin A/C were also mislocalized. This demonstrates a strong and selective functional connection between Samp1 and A-type lamina and LINC complex protein networks.
The role of conserved cysteine motifs for targeting of Samp1 to the inner nuclear membrane. (A) Sequence of the most conserved amino acids (35–210) of Samp1. Conserved CxxC motifs are showed in bold. The asterisks indicate cysteines substituted with alanines. (B) Immunofluorescence micrographs and corresponding phase-contrast images of HeLa cells expressing cysteine-to-alanine substitution mutants of YFP–Samp1. YFP–Samp1(C48A) (a,b), YFP–Samp1(C121A) (c,d) and YFP–Samp1(C48A,C121A) (e,f). All three Samp1 mutants localize in the cytoplasm and give rise to small nuclei with irregular shapes. Scale bars: 10 μm.
Disruption of the nuclear envelope organization by mutating conserved cysteine motifs of Samp1. (A) HeLa cells expressing YFP–Samp1(C121A) (a–t) or YFP–Samp1 (u–y) were immunostained with anti-GFP antibodies (b,f,j,n,r,v) and antibodies specific for the proteins indicated on the left. Corresponding images of DNA staining (c,g,k,o,s,x) and phase-contrast images (d,h,l,p,t,y) are also shown. In cells expressing mutant YFP–Samp1(C121A), emerin (a) and Sun1 (e) were aggregated in the cytoplasm. Despite the distorted nuclear morphology, Sun2 (i), nuclear pore proteins (m) and Lamin-B (q) localise in the nuclear envelope, although the intensity was sometimes affected. Cells expressing YFP–Samp1 display a normal nuclear morphology and normal distribution of emerin in the nuclear rim (u–y). (B) HeLa cells expressing YFP–Samp1(C121A) and YFP–Samp1(C48A,C121A) mutants stained with antibodies against GFP and lamin A/C. Corresponding images of DNA staining (c,g) and phase-contrast images (d,h) are shown. In cells expressing YFP–Samp1(C121A), lamin A/C (a) is located in the cytoplasm as aggregates, but in cells expressing the double mutant YFP–Samp1(C48A,C121A), the effect on lamin A/C (e) is suppressed although nuclear morphology is abnormal. Scale bars: 10 μm.
Emerin depends on Samp1 for correct distribution in the NE
To investigate the relationship between Samp1, emerin, Sun1 and lamin A/C we silenced the expression of each of these proteins using siRNA and studied the distribution of the other proteins by immunofluorescence. Nuclear morphology was normal after 4 days of treatment with siRNA specific for Samp1, although the levels of Samp1 were significantly decreased. Interestingly, the levels of emerin were reduced in the nuclear rim and instead appeared in aggregates (Fig. 6). At the same time, distribution of lamin A/C and Sun1 appeared normal (not shown). These results demonstrate that emerin is dependent on Samp1 for proper distribution in the NE. However, silencing of emerin expression had no apparent effect on the localization of Samp1 (Fig. 6), Sun1 or lamin A/C (not shown), suggesting that Samp1 is not dependent on emerin for its location in the NE. Furthermore, distribution of Samp1, lamin A/C and emerin was not dependent on Sun1 (not shown). Consistent with previous studies on embryonic fibroblasts from lamin A/C double-knockout mice (Sullivan et al., 1999), siRNA-mediated silencing of lamin A/C expression resulted in a distorted nuclear morphology and a selective mislocalization of emerin, whereas Samp1 (Fig. 6j–l) and Sun1 (not shown) were still able to distribute in the NE.
The zinc finger domain of Samp1 accounts for a dominant-negative effect on endogenous Samp1 and perturbation of nuclear integrity. HeLa cells were transfected with cDNA encoding CD8–Samp1(35–210,C121A) and subjected to immunolabeling. (A) Cells immunostained for emerin (a), lamin A/C (d) and Sun1 (g) with corresponding DNA staining (b,e,h) and phase-contrast images (c,f,i). Transfected cells are indicated by arrows. (B) Double immunofluorescence labeling anti-CD8 antibodies (a,e, green) and anti-Samp1 antibodies (c,g, red), with corresponding overlays (b,f) and phase-contrast images (d,h). Scale bars: 10 μm. Note the appearance of endogenous Samp1 in cytoplasmic aggregates (c,g). (C,D) Immuno-electronmicrographs showing large areas lacking peripheral heterochromatin in the right part of the nucleus of a cell expressing CD8–Samp1(35–210,C121A) (C) not seen in cells expressing CD8–Samp1(35–210) (D). Insets show the boxed regions at higher magnification.
Samp1 interacts with emerin in a zinc-dependent manner
Next, we investigated protein–protein interactions of Samp1 by coimmunoprecipitation experiments. Detergent extraction of HeLa cells were performed under optimized NaCl concentrations for YFP–Samp1 and emerin, respectively. Immunoprecipitation, which was performed at 150 mM NaCl, using GFP-specific antibodies, showed that emerin coprecipitated with YFP–Samp1 in the presence of zinc ions, an effect that was impeded by the bivalent metal ion chelator EDTA (Fig. 7). This demonstrates that Samp1 is able to interact with emerin in a zinc-dependent manner and suggests that Samp1 might indeed have one or two functional zinc fingers.
Discussion
Combining confocal microscopy and deconvolution, we found that Samp1 distributes in a distinct pattern, displaying partial colocalization with Sun1, but only limited overlap with lamin A/C, lamin B, emerin, Sun2 or nuclear pores. This is consistent with previously published studies of other NE proteins, which show that they distribute in characteristic, but interconnected, networks or microdomains (Liu et al., 2007; Lu et al., 2008; Maeshima et al., 2006; Shimi et al., 2008). Sun1 has previously been reported to partially colocalize with nuclear pores and partially with Sun2. Sun2, in turn, also localized to areas distinct from either nuclear pores or Sun1 (Liu et al., 2007; Lu et al., 2008). The characteristic distribution patterns obtained in fixed cells are not likely to represent non-flexible bindings or ‘allowed’ interactions in the living cell. More likely, the distribution patterns can be regarded as temporal steady state snapshots of certain combinations of more or less dynamic interactions (see e.g. Ostlund et al., 2009) at any given moment. In this way, colocalization indicates that two proteins have the capability to interact, whereas lack of colocalization does not mean that they cannot or do not interact.
An interesting question is what determines the patterns of distribution of NE proteins. This is likely to be specified in part by the LINC complexes connecting the cytoskeleton with the nucleoskeleton. The Samp1 homologue Ima1 is part of a LINC complex in fission yeast, which was shown to oscillate in live yeast cells (King et al., 2008) that lack nuclear lamins. It remains to be determined whether Samp1 and LINC complex proteins are also involved in the elastic movements of NE proteins that take place in the nuclear periphery of mammalian cells (Daigle et al., 2001) and to what extent these movements are specified by the cytoskeleton, the nucleoskeleton or both.
The effect of post-transcriptional silencing of Samp1 on emerin. HeLa cells were treated with siRNA directed against Samp1, emerin or lamin A/C, as specified on the left and stained with antibodies specific for emerin (a,d,g) or Lamin A/C (j) and Samp1 (b,e,h,k). The corresponding phase-contrast images (c,f,i,l) are shown. Note the abnormal distribution of emerin after knockdown of Samp1, whereas Samp1 localizes normally in the NE after knockdown of emerin or lamin A/C. Scale bars: 10 μm.
Our data show that the entire N-terminal half of Samp1 is exposed in the nucleoplasm and contains a non-membrane-spanning hydrophobic domain. The N-terminal domain contains four conserved CxxC motifs with potential to form two zinc fingers. Analysis of cells expressing Samp1 variants carrying cysteine to alanine substitution mutations designed to specifically disrupt these putative zinc fingers showed that CxxC motifs of both ZF1 and ZF2 were necessary for the sorting of Samp1 to the NE and also had an impact on the organization of a specific subset of NE proteins including emerin, Sun1 and in some cases lamin A/C. In addition, Samp1 is able to interact with emerin in a zinc-dependent manner (Fig. 7). Taken together, this raises the question whether Samp1 has functional zinc fingers. This possibility is supported by a 39% sequence similarity between a 23 amino acid stretch (residues 48–71) containing the first two CxxC motifs of human Samp1 and two prokaryotic transposases (Figueroa et al., 2010).
The compelling impact of the Samp1 zinc finger mutants on nuclear organization (Figs 4 and 5) is reminiscent of the effects reported in cells where Sun1 expression had been silenced (Liu et al., 2007) and strongly suggests that Samp1 interacts with components of the A-type lamina network including the LINC complex protein Sun1. We propose that a series of sequential steps resulting in a decreased stabilization of the A-type lamina network is the most likely explanation for the observed effects exerted by the mislocated Samp1 zinc finger mutants. Disruption of the NE distribution of endogenous Samp1 implies that also emerin would become mislocalized (see Fig. 6). Although neither Sun1 nor lamin A/C are dependent on emerin alone for their location in the nuclear periphery (Haque et al., 2010), the combined depletion of both Samp1 and emerin from the INM, might destabilize the anchorage of Sun1 and lamin A/C in the NE.
Coimmunoprecipitation of Samp1. Extracted proteins from HeLa cells (lanes 9 and 10) or HeLa cells expressing YFP–Samp1 (lanes 1–8) were incubated with Protein-G–Sepharose in the presence and absence of anti-GFP antibodies, as indicated. The experiments were carried out either in the presence of 1 mM Zn2+ (lanes 5–10) or in the presence of 10 mM EDTA (lanes 1–4). Bound (P, lanes 2, 4, 6, 8 and 10) and unbound (S, lanes 1, 3, 5, 7 and 9) fractions were analyzed by western blotting, using antibodies specific for emerin. 15.9% of total emerin coprecipitated in the presence of anti-GFP antibodies compared with 3.8% in the absence of anti-GFP antibodies.
Emerin has previously been shown to be dynamically associated with the NE (Ostlund et al., 1999), which might render it easier to extract compared with Sun domain or lamina proteins. Coimmunoprecipitation (Fig. 7) showed that Samp1 is able to interact with emerin. This is further supported by the dependence on Samp1 for the proper location of emerin in the NE (see Fig. 6). The partial coimmunoprecipitation (Fig. 7) and low degree of colocalization (supplementary material Fig. S1, Pearson coefficient=0.07) indicate that Samp1 and emerin distribute in separate, but partially overlapping structures. Although the NE location of emerin depends on Samp1, it cannot be mediated by permanent binding to Samp1 to a large extent. Whether Samp1 has a chaperone or licensing role in connecting emerin to A-type lamins and/or LINC complexes and/or chromatin binding proteins such as barrier to autointegration factor (BAF; LAP2 binding protein 1) for its NE location remains to be established.
The major determinant for Sun1 localization at the NE has not yet been identified (Haque et al., 2010). Samp1 partially colocalizes with Sun1 (Fig. 1u–x) and is required for anchorage of the microtubule-organizing center close to the nucleus (Buch et al., 2009) and for INM localization of emerin (Fig. 6). Furthermore, zinc finger mutants of Samp1 had a selective deleterious effect on the distribution of Sun1, emerin and lamin A/C (Fig. 4). Together this evidence argues in favor of a functional association of Samp1 with the LINC complex and the A-type lamina networks. This idea is supported by the previously described interaction between the Samp1 homologue Ima1 and the Sun homologue Sad1 in fission yeast (King et al., 2008). Perhaps in metazoans, which have a nuclear lamina, NE proteins are attached by a more complex network of multivalent interactions.
Materials and Methods
Quikchange mutagenesis
Deletion mutants and cysteine-to-alanine mutants of cDNA encoding YFP–Samp1 were created by site-directed mutagenesis using the Quik-change kit (Stratagene) with Pfu Turbo polymerase (Stratagene). For Samp1 deletion mutants, the following primers were used: Samp1(1–236), forward 5′-GCCAGCGGACACTTCGCCTAAGGCACCACTGTGCCCCTG-3′ and reverse, 5′-CAGGGGCACAGTGGTGCCTTAGGCGAAGTGTCCGCTGGC-3′; Samp1(1–210), forward, 5′-TCCCCGGTCCAGGTCATCTAACTCCGTGCCCTCGCCTTC-3′ and reverse, 5′-GAAGGCGAGGGCACGGAGTTAGATGACCTGGACCGGGGA-3′; Samp1(1–180), forward, 5′-CGCCAGCTGCGCGCCCTGTAACTCAGCCAGCAGTTCAAG-3′ and reverse, 5′-CTTGAACTGGTGGCTGAGTTACAGGGCGCGCAGGCTGCG-3′. For cysteine to alanine mutants of Samp1 we used the following primers: Samp1(C48A) forward, 5′-ACGATGGTCAACGCCTGGTTCTGCAAC-3′ and reverse, 5′-GTTGCAGAACCAGGCGTTGACCATCGT-3′; Samp1(C121A) forward, 5′-CAAGTCCTGCTGGCCAAGAGGTGCAAC-3′ and reverse, 5′-GTTGCACCTCTTGGCCAGCAGGACTTG-3′.
Cloning
To express Samp1(1–180)–GFP fusion protein, we fused the genes encoding amino acids 1–180 of Samp1 to GFP. The cDNA encoding full-length Samp1 (IMAGE clone ID: 3870153) was used as a template for amplification. The forward primer, 5′-CTCTAAGCTTATGGAGGGAGTGAGCGCG-3′ contained a HindIII restriction site and reverse primer 5′-ATATGGATCCCGGTTCTGGTGCTTGATG-3′ contained a BamHI restriction site. After amplification and digestion with HindIII and BamHI (Fermentas), the fragments were ligated into pEGFP-N1 (BD Biosciences), vector.
Antibodies
Primary antibodies used for immunofluorescence were rabbit polyclonal Samp1 antibodies (1:6000) (Buch et al., 2009), except in Fig. 1u–x, where APEX-labeled Alexa Fluor 488 (Invitrogen) anti-Samp1 antibodies were used; mouse monoclonal antibodies against emerin (1:500) from Santa Cruz Biotechnology (sc-25284); rabbit polyclonal antibodies against GFP (1:3000) from Invitrogen (A11122); mouse monoclonal anti-GFP antibodies (1:3000) from Roche (11814460001); mouse monoclonal mAb414 (1:5000) from BioSite (MMS120P); mouse monoclonal anti-Lamin-A/C antibodies (1:250) from abcam (ab40567); rabbit polyclonal anti-Sun1 antibodies (1:400) (antiUNA84A) from Sigma (HPA008346); rabbit polyclonal anti-Sun2 antibodies (1:500) (antiUNC84B) from Sigma (HPA001209); and goat polyclonal anti-Lamin-B antibodies (1:100) from Santa Cruz Biotechnology (sc-6217). As secondary antibodies for immunofluorescence, we used Alexa Fluor 488, Alexa Fluor 568 and Alexa Fluor 647 donkey anti-mouse, anti-goat and anti-rabbit antibodies, diluted 1:5000 (Invitrogen). Draq5 (Biostatus Limited) dilution 1:1000 in PBS was used to stain DNA. As secondary antibodies for western blotting, horseradish-peroxidase-coupled donkey anti-mouse IgG (NA931V, GE Healthcare) was used at 1:3000 dilution.
Cell culture and transfection
HeLa (ATCC no.CCL-2) cells were grown and maintained in 1× Dulbecco's modified Eagle's (DMEM) GlutaMAX™ medium (GIBCO), supplemented with 10% fetal bovine serum (FBS, v/v) and 1% penicillin-streptomycin (v/v) at 37°C in a humidified atmosphere containing 5% CO2. For transfection of cells with plasmids encoding the different Samp1 fusion proteins, FuGENEHD transfection reagent (Roche) was used and cells were analyzed after incubation for 24 hours. For siRNA-mediated post-transcriptional silencing of Samp1, the following sense sequences were used: Samp1 #1, 5′-GGAAGUGUUGACAGUGUGAtt-3′ and Samp1 #2, 5′-GCGGCUGUGGAGUACUACAtt-3′ (Buch et al., 2009) (Ambion). For post-transcriptional silencing of emerin the following siRNA sense sequence was used: 5′-GGUGCAUGAUGACGAUCUUtt-3′, (Thermo Fisher Scientific) (Salpingidou et al., 2007). For post-transcriptional silencing of lamin A/C the following siRNA sense sequence was used: 5′-CUGGACUUCCAGAAGAACAtt-3′ (MWG-eurofins). As a control, we used Stealth™ RNAi negative universal control (Invitrogen). 18 nM siRNA in HiPerfect (Qiagen) transfection reagent was used to transfect HeLa cells, which were analyzed after incubation for 96 hours.
Immunofluorescence
For immunofluorescence, cells were grown on glass coverslips, carefully washed once in PBS, fixed on ice in 3.7% paraformaldehyde in PBS for 20 minutes and permeabilized with 0.5% Triton X-100 in PBS for 5 minutes. Cells were then blocked in PBS containing 0.1% Tween-20 (PBS-T) and 2% BSA. The samples were then incubated for 1 hour with primary antibodies in blocking buffer, washed with blocking buffer three times, and incubated with secondary antibodies in blocking buffer for 1 hour. After additional washes in PBS-T, the coverslips were stained with DRAQ5 for 30 minutes and after a final wash, mounted in Fluoromount-G (SouthernBiotech).
Imaging
Imaging was performed using a Leica TCS-SP1 laser-scanning confocal microscope (Leica, Heidelberg, Germany) with a 63× 1.32 NA oil-immersion objective using a 488 nm 20 mW Ar laser line, a solid-state 561 nm, 10 mW laser and a 633 nm, 10 mW He-Ne laser line. Emission spectra were collected at 500–550 nm (Alexa Fluor 488 and YFP), 570–650 nm (Alexa Fluor 568) and 645–700 nm (Alexa Fluor 647, draq5). The laser lines were scanned sequentially.
Optical sectioning of cells was performed and subsequent deconvolution using an iterative Wiener filter with experimentally measured PSFs was applied to enhance resolution using ImageJ with the parallel iterative deconvolution plug-in. The iteration was allowed to run until quality criteria of mean delta <0.01% was reached. Colocalization was evaluated by Pearson correlation statistical analysis with Manders' coefficients using the JaCoP plug-in of imageJ (http://rsb.info.nih.gov/ij).
Coimmunoprecipitation
Untransfected HeLa cells or HeLa cells transfected with YFP–Samp1 were used for extraction of endogenous emerin or YFP–Samp1, respectively. Crude nuclear preparations were prepared as described previously (Buch et al., 2009). Emerin was extracted with buffer containing with 1% Triton X-100 and 150 mM NaCl, whereas YFP–Samp1 was extracted with buffer containing 1% Triton X-100 and 250 mM NaCl, before centrifugation at 100,000 g for 20 minutes. The supernatant containing extracted YFP–Samp1 was diluted with buffer to reduce NaCl concentrations to 150 mM. YFP–Samp1 and emerin extracts were mixed and added to Protein-G–Sepharose beads (GE Healthcare). Immunoprecipitation was performed in the presence or absence of anti-GFP antibody. Parallel experiments were performed either in the presence of 1 mM zinc acetate or 10 mM EDTA. Samples were incubated overnight at 4°C before separation of the bound (P, pellet) and unbound (S, supernatant) fractions by centrifugation. Equivalent amounts of the supernatant and pellet fractions were separated by SDS-polyacrylamide (10%) gel electrophoresis and transferred to PVDF membranes. The membranes were blocked in 5% milk in PBS-T, incubated with primary antibodies for 1 hour, washed three times for 10 minutes with blocking buffer and then incubated with secondary antibodies for 1 hour. After additional washing in PBS-T, membranes were subjected to ECL detection (GE Healthcare) using Amersham Hyperfilm™ ECL (GE Healthcare).
Acknowledgments
We thank Marie Beckman for critically reading the manuscript. The core facility for electron microscopy at Karolinska Institute is acknowledged for help with electron microscopy. This study has been supported by grants to E.H. from the Swedish Research Council (2006-5977, 2006-3468 and 2010-4481) and the Swedish Cancer Society (08 0306).
Footnotes
-
↵* These authors contributed equally to this work
-
Supplementary material available online at http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.078923/-/DC1
- Accepted February 22, 2011.
- © 2011.
References
Article navigation
Related articles
Cited by...
More in this TOC section
Similar articles
Other journals from The Company of Biologists
Introducing FocalPlane’s new Community Manager, Esperanza Agullo-Pascual
We are pleased to welcome Esperanza to the Journal of Cell Science team. The new Community Manager for FocalPlane, Esperanza is joining us from the Microscopy Core at Mount Sinai School of Medicine. Find out more about Esperanza in her introductory post over on FocalPlane.
New funding scheme supports sustainable events
As part of our Sustainable Conferencing Initiative, we are pleased to announce funding for organisers that seek to reduce the environmental footprint of their event. The next deadline to apply for a Scientific Meeting grant is 26 March 2021.
Read & Publish participation continues to grow
"Alongside pre-printing for early documentation of work, such mechanisms are particularly helpful for early-career researchers like me.”
Dr Chris MacDonald (University of York) shares his experience of publishing Open Access as part of our growing Read & Publish initiative. We now have over 150 institutions in 15 countries and four library consortia taking part – find out more and view our full list of participating institutions.
Cell scientist to watch: Romain Levayer
In an interview, Romain Levayer talks about starting his own lab, his love for preprints and his experience of balancing parenting with his research goals.
Live lactating mammary tissue
In a stunning video, Stewart et al. demonstrate warping of the alveolar unit due to basal cell-generated force as part of their recent work investigating roles for mechanically activated ion channels in lactation and involution.
Visit our YouTube channel to watch more videos from JCS, our sister journals and the Company.
JCS and COVID-19
For more information on measures Journal of Cell Science is taking to support the community during the COVID-19 pandemic, please see here.
If you have any questions or concerns, please do not hestiate to contact the Editorial Office.