The giant isoforms of nesprins 1 and 2 are emerging as important players in cellular organization, particularly in the positioning of nuclei, and possibly other organelles, within the cytoplasm. The experimental evidence suggests that nesprins also occur at the inner nuclear membrane, where they interact with the nuclear lamina. In this paper, we consider whether this is consistent with current ideas about nesprin anchorage and about mechanisms for nuclear import of membrane proteins.
Nesprins are a family of spectrin repeat proteins [1,2], and missense mutations in these proteins have been described in Emery–Dreifuss muscular dystrophy , a disease that is also caused by mutations in the nuclear envelope proteins, emerin and lamin A/C . Alternative initiation and splicing of two genes, SYNE1 and SYNE2, generate multiple isoforms of different sizes but mostly with a common C-terminal region [5–8]. The largest nesprin isoforms have a spectrin-repeat rod domain separating a C-terminal transmembrane KASH (Klarsicht-ANC-Syne-homology) domain from N-terminal CH (calponin homology) domains that bind actin. These ‘nesprin-giant’ molecules (also known as enaptin  and NUANCE ) attach to the ONM (outer nuclear membrane) through their KASH domain, bound by SUN proteins in the lumen between the INM (inner nuclear membrane) and ONM, and are thus able to form a link to the actin cytoskeleton through their CH domains [9,10]. This LINC (linker of nucleoskeleton and cytoskeleton) complex has been reviewed recently by Starr .
The function of short nesprin forms that lack an actin-binding domain is unclear, although most of them do contain the sequences required for direct interaction with emerin and lamin A/C in the nucleoplasm [1,2,12,13] and a role in linking the INM to the nuclear lamina has been proposed . Some of these have been called Syne (synaptic nuclear envelope)  or Myne (myocyte nuclear envelope)  proteins. One current hypothesis is that the smallest isoforms occupy the INM, whereas the giant forms link the ONM with the actin cytoskeleton [14,15]. This view is partly based on the observation that only membrane proteins smaller than approx. 60 kDa, such as emerin (29 kDa) and nesprin-2-α (49–64 kDa), can enter the nucleus by diffusion through lateral channels of the nuclear pores, without the need for an NLS (nuclear localization signal) sequence . The lateral diffusion model for localization of INM proteins would suggest that only the short nesprin-2-α forms can reach the INM easily by this pathway, if a strict size limitation of approx. 60 kDa applies . On this model, all larger nesprin-2 isoforms and all nesprin-1 isoforms would be at the ONM, leaving only nesprin-2-α at the INM. However, the only published experimental evidence does not support this simple model. One group has shown nesprin-2-giant at both INM and ONM in cultured keratinocytes, using both a digitonin permeabilization method  and immunogold EM (electron microscopy) . The EM studies are so important that it is worth spending a little time considering the technical problems. Transfection of cells with full-length giant nesprins has not yet been achieved, so the studies were done with antibodies against endogenous nesprin with potentially greater non-specific background problems. However, the same result was obtained with two different N-terminal nesprin-2 antibodies  and this makes non-specific staining at the INM far less likely. A control experiment with an almost exclusively INM protein, such as LAP2 or LBR (lamin B receptor), would strengthen the evidence further. Four possible configurations of nesprins are illustrated in Figure 1 (B and C being the lateral diffusion options). Quantitative comparison of EM data with N-terminal and C-terminal antibodies  supported a model similar to Figure 1(D), and nesprin-2-giant was actually more abundant at the INM than at the ONM.
Possible orientations of nesprin isoforms at the nuclear membrane
Is it possible to reconcile giant nesprins at the INM with current models for INM protein transport? Malik et al.  have reviewed in detail all conceivable mechanisms for transporting proteins that are embedded in the ER (endoplasmic reticulum) into the nucleus. For dividing cells, there would be no size barrier during re-assembly of the nuclear membrane after mitosis. The possibility of detaching proteins from the ER for transport through the nuclear pore without size limit and re-insertion into the INM was also considered. King et al.  showed that some INM proteins use an importin-mediated mechanism to squeeze past the pore while presumably being still inserted in the membrane. This would be one explanation of the experimental evidence for energy-dependent transport of some INM proteins , since simple lateral diffusion should be energy independent. Absence of a known, or classic, NLS does not necessarily rule out importin-mediated transport, since different sequences may be used for recognition by different importin complexes . Experimentally, some transfected nesprin fragments without a transmembrane domain do cross into the nucleoplasm . However, although importin complexes can certainly carry very large proteins, such as giant nesprin, through the nuclear pore, there may still be a size limit for importin-mediated transport of membrane proteins alongside the nuclear pore.
There is another issue, apart from their size, that makes people reluctant to accept giant nesprins at the INM and that is a question of symmetry and orientation. On current models of the LINC complex, giant nesprins at the ONM are anchored to SUN proteins at the INM by an interaction in the lumen between SUN and the KASH domain of nesprin (Figure 1A). Indeed, it has been suggested that this interaction in the lumen determines the size of the gap between INM and ONM . In order to anchor nesprins at the INM in a similar way, SUN proteins would have to bind KASH domains in two different orientations (as illustrated in ) or SUN proteins would have to occur at both INM and ONM, and there is no evidence for either possibility. Some published models show nesprins at the INM anchored by an unspecified, SUN-independent mechanism (as in ). The anchorage question applies, of course, to all nesprin isoforms at the INM, even the shortest ones. To deal with one final possibility, there is evidence for the existence of nesprin-2 transcripts lacking the KASH and transmembrane domains  and such isoforms (of any size) could enter easily through the nuclear pore and be trapped at the INM by emerin or lamin A/C interactions alone. It is not known, however, whether such isoforms could be sufficiently abundant to account for all larger nesprins at the INM, since there are currently no antibodies that will distinguish between isoforms containing or lacking the KASH and transmembrane domains.
Knockouts and knockdowns have not been as helpful as one might have hoped in illuminating nesprin localization. Knockdown of SUN proteins in HeLa cells caused mislocalization of nesprin-2-giant to the ER, leaving lamin A/C in the nuclear lamina unchanged, as expected from the SUN-anchorage model [9,21]. However, absence of lamin A/C in Lmna−/− MEFs (mouse embryonic fibroblasts; a heterogeneous population) also caused nesprins to mislocalize to the ER, even though SUN proteins remained at the INM in most cells [9,10,21]. This shows that both lamin A/C and SUN proteins are necessary, but not sufficient, to retain nesprin-2-giant at the nuclear rim. The possibility of more subtle, indirect effects of lamin A/C on the ability of SUN to anchor nesprins has not been ruled out. Specific knockdown of nesprin-2-giant in keratinocytes by targeting the N-terminus also reduced the levels of the shorter C-terminal nesprin-2 isoforms, leading to the idea that they might interact with nesprin-2-giant via their spectrin repeats . In this model, short forms would co-localize with nesprin-2-giant either at the ONM only or at both ONM and INM. The problem with this as a general model is that nesprin isoforms do not show the same stoichiometry in all cells. Thus giant nesprins predominate in skin fibroblast cultures, whereas short forms are dominant in adult skeletal muscle nuclei  and nesprin-1 is absent from the nuclear membrane in keratinocytes . This shows the importance of avoiding generalization from experimental data obtained only with skin fibroblasts, keratinocytes, HeLa cells or MEFs. A recent study , for example, showed that Sun1 protein localization at the INM was normal in Lmna−/− MEFs, but Sun2 relocated to the ER in a small proportion of the cells (possibly a reflection of MEF heterogeneity).
In conclusion, many of the tools required to study nesprin localization and function are currently lacking, including full-length transfectable cDNAs and isoform-specific antibodies, and it is already clear that generic cell lines, such as HeLa and MEFs, cannot provide a full understanding of nesprin function.
Nuclear Envelope Disease and Chromatin Organization 2009: Independent Meeting held at College of St Hild and St Bede, University of Durham, Durham, U.K., 22–23 April 2009. Organized and Edited by Chris Hutchison (Durham, U.K.).
We thank Eric Schirmer (University of Edinburgh) for helpful comments that we have incorporated into this review.
We thank the Robert Jones and Agnes Hunt Institute of Orthopaedics and the Muscular Dystrophy Association (U.S.A.) for financial support.
Present address: Department of Pathology, Hope Hospital, Salford M6 8HD, U.K.