on the mechanism of transport of inner nuclear membrane proteins

University of Groningen

On the mechanism of transport of Inner Nuclear Membrane ProteinsLaba, Justyna Katarzyna

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Laba, J. K. (2016). On the mechanism of transport of Inner Nuclear Membrane Proteins [Groningen]:University of Groningen

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 03-04-2018

https://www.rug.nl/research/portal/en/publications/on-the-mechanism-of-transport-of-inner-nuclear-membrane-proteins(37fcad33-dea6-4616-876e-88f365a1e094).html

On the mechanism of transport of inner nuclear membrane proteins

Justyna K. Laba

The research described in this thesis was carried out at the University of Groningen, the Netherlands, in the research group Cellular Biochemistry at the European Research Institute for the Biology of Ageing (ERIBA), and the research group Membrane Enzymology of the Groningen Biomolecular Sciences and Biotechnology Institute. The work was financially supported by the Netherlands Proteomics Center (NPC) and FEBS Short Term Fellowship.

ISBN: 978-90-367-8483-2 (printed version) 978-90-367-8482-5 (electronic version)

Cover design by : Edu Madrid PerezLay Out by: Remco Wetzels ||Remcowetzels.nlPrinted by: Proefschriftmaken.nl || Uitgeverij BOXPressPublished by: Uitgeverij BOXPress, ’s-Hertogenbosch

© Justyna K. Laba, 2015

On the mechanism of transport of inner nuclear membrane

proteins

PhD thesis

to obtain the degree of PhD at the University of Groningen on the authority of the

Rector Magnificus Prof. E. Sterken and in accordance with

the decision by the College of Deans.

This thesis will be defended in public on

Friday 22 January 2016 at 14.30 hours

by

Justyna Katarzyna Łaba

born on 15 November 1985 in Wrocław, Poland

Supervisor Prof. B. Poolman Co-supervisor Dr. L.M. Veenhoff Assessment Committee Prof. D.J. Slotboom Prof. E.A.A. Nollen Prof. G. Cingolani

To my first and most patient teacher – my grandma

TABLE OF CONTENTS

Introduction

Long Unfolded Linkers Facilitate Membrane Protein Import Through the Nuclear Pore Complex

Super-resolution optical microscopy of membrane proteins trapped in the Nuclear Pore Complex

Nuclear import of membrane proteins revisited: Active nuclear import of Heh2-derived proteins occurs while membrane embedded, the path through the NPC is unresolved

Nsp1 in the trapping experiments – bait or prey?

Traffic to the inner membrane of the nuclear envelope

Appendix Summary List of publications Acknowledgements

Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

Chapter 6

9

25

37

51

73

93

111117124125

Chapter 1Introduction

Chapter 1

10

The most important characteristic that distinguishes bacterial and eukaryotic cells is the presence of nucleus – a double membrane organelle that encompasses the DNA. Since its first detailed description in 1833 by Robert Brown [1], the nucleus was traditionally defined as an organelle that protects the DNA from mechanical damage and separates the transcription and translation. The advances in knowledge about molecular processes accompanying DNA replication, repair and transcription to RNA have shed new light into this structure. In the editorial from Current Opinion in Cell Biology from June 2014, Michael Rout and Gary Karpen wrote: “Our investment in genomics in humans and other eukaryotes has been rightly huge; however, the return has been slow, as we discovered that sequence assemblies alone reveal little about how genetic information is utilized. Instead genome functions are greatly influenced by processes that control nuclear and chromosome architecture and dynamics – chromatin factors and modifications, epigenetic mechanisms, and pathways for controlling transport of molecules into and out of the nucleus. (… ) The view of the nucleus as an organelle is changing; not a passive repository for genetic information, but an active, dynamic super structure whose processes dictate how that information is organized, assessed and used” [2]. Clearly, new thinking in the field gives the nucleus a much greater, active role in cellular functioning. The nucleus is usually the biggest and round shaped organelle in the cell. The most prominent architectural features of the nucleus include: a nuclear envelope, nuclear lamina, chromatin in the nucleoplasm and a nucleolus (fig.1). The nuclear envelope (NE) is a double membrane system continues with the endoplasmatic reticulum (ER). The inner and outer nuclear membranes (INM and ONM) come together in multiple places forming nuclear pores. The space of the pores is filled with large protein assemblies – the nuclear pore complexes (NPCs), which enable molecule exchange between the nucleus and cytoplasm. The protein composition of ONM is similar to ER, but INM contains unique residents (described in chapter 6). In metazoan cells the INM is overlaid with nuclear lamina, a dense protein meshwork build from lamins and lamin-associated proteins. The DNA content is organized into chromatin and suspended in the viscous nucleoplasm. The transcriptionally active lightly packaged euchromatin is usually found in the inside of the nucleus, while the dense heterochromatin is anchored at the nuclear periphery. The nucleolus, build from protein and RNA and not surrounded by a membrane, is a place of rDNA transcription and processing and ribosome subunit assembly. Gene mutations that affect the nuclear morphology described above can have profound outcomes not only for cells but also for entire organisms. A group of human diseases characterized by disturbed nuclear shape is called nuclear envelopathies. Hutchinson Gilford progeria, for example, is a rare syndrome of accelerated aging that was linked to several mutations in lamin A causing severe deformation of the nuclear envelope. Distortions of other nuclear functions can also have catastrophic

Introduction

11

Chap

ter 1

consequences. Disturbing the balance of nuclear transport is associated with tumorigenesis – for example by overexpression of exportin CRM1 observed in gliomas, osteosarcomas and leukemias, which leads to increased export of tumor suppressors and subsequently causes cell proliferation [1,3-6].

FIGURE 1.

A. A schematic representation of the nuclear architrecture. ER system continues with the NE is depicted. B. Central cross sections of the C2C12 cells nucleus imaged by 3D-SIM (Structured Illumination Microscopy). Lamin B is labeled green and NPCs are red. DNA is stained with DAPI (blue). Scale bars: 5 μm and 1 μm. From [72]. Reprinted with permission from AAAS. C. NE-ER system in yeast Saccharomyces cerevisiae. Left panel shows yeast cell in DIC, right panel depicts fluorescent images of a GFP fused transmembrane reporter located in the NE and ER. Scale bars: 2 μm.

The focus of the work described in this thesis is on the inner nuclear membrane (INM) of yeast Saccharomyces cerevisiae, its components and the mechanism by which membrane proteins enter the nucleus. Once called an ‘uncharted territory’ of the nucleus [7], the INM is an example of the change in thinking about nuclear functions and components. Up until now there are only few confirmed resident proteins of the INM, but the evidence of their involvement in processes important for functioning of entire organisms is growing (described in chapter 6). While identification of membrane proteins that localize to the INM is still challenging, new clues can be gained by defining the features that drive their import to the nucleus. Targeting information in the sequence of yeast protein Heh2 is investigated in this thesis. In this chapter, the known mechanisms of protein transport to the nucleus will be described.

NPClaminschromatin

A B

C

Chapter 1

12

NUCLEAR PORE COMPLEX ARCHITECTURE AND FUNCTIONING

All the transport events to and from the nucleus are regulated by the Nuclear Pore Complexes (NPCs). The architecture of these specialized transport machineries is largely conserved between species. NPCs are cylindrically shaped, with a width of 100-150nm and a height of 50-70nm depending on the organism [8] and a characteristic 8-fold rotational symmetry (depicted in fig.2). They are built by multiple copies of 30 proteins, called nucleoporins, which are organized in 8 spokes. Spokes are interconnected and form co-axial rings: a membrane ring, 2 inner rings, and cytoplasmic and nucleoplasmic exposed outer rings [9]. The structure contains a centrally located tunnel of approximately 40nm diameter - this is the central channel, through which transport of molecules occurs. At the periphery of the nuclear pore, adjacent to the NE smaller tunnels of approximately 10nm diameter were observed – the peripheral channels [10], which may also be involved in transport processes. Cytoplasmic filaments reach out from the structure on the outside of the nucleus, while the nucleoplasmic side of the NPC is covered by the nuclear basket. The core structure of the NPC is formed by the nucleoporins from the inner and outer rings, which together constitute the scaffold of the pore and stabilize the curved pore membrane [9]. Nup170 complex in yeast and Nup155 complex in vertebrates builds the inner rings. The building block of the outer rings is the so-called Y-shaped complex - Nup84 in yeast, Nup107 in vertebrates. Recent experiments have proven that the Y-shaped complex is positioned perpendicularly to the transport direction [11]. The core scaffold is anchored to the NE via a connection between the inner rings and the membrane ring [9]. In yeast, the membrane ring is formed by 3 proteins – Pom34, Pom152 and Ndc1 (in vertebrates – Gp210, Ndc1 and Pom121). In the NE lumen, the C-terminal domains of Pom152 homo-oligomerize and facilitate the formation of the luminal ring [9]. The side of the core scaffold, which faces the central channel, contains attachment sites for linker nucleoporins and FG-nucleoporins (FG-

FIGURE 2. THE STRUCTURE OF THE NPC, REPRESENTED BY A DENSITY CONTOUR. A top view of the entire structure and a slice along the central z-axis are presented. Nuclear envelope is shown in grey. Reprinted by permission from Macmillan Publishers Ltd: Nature [9], copyright 2007.

Introduction

13

Chap

ter 1

Nups). FG-Nups fill the central channel and form the transport barrier of the complex [9]. The natively unfolded domains of FG-Nups, containing multiple copies of phenylalanine and glycine rich repeats (organized in the FG, FxFG or GLFG clusters), provide the docking sites for the transport factors shuttling through the pore [9]. In yeast two linker nucleoporins, Nic96 and Nup82, anchor most of the FG-Nups to the core scaffold [9]. As mentioned earlier, FG-Nups are natively unfolded, and prediction of their configuration within the pore is challenging. However, their anchor points to the core scaffold of the NPC have been mapped with limited resolution [9] and this provides the categorization criterion to distinct groups: cytoplasmic, nucleoplasmic and symmetric FG-Nups. Nsp1, described later in this thesis (chapter 5), is present in two subcomplexes of the NPC, one of which is cytoplasmic and the other symmetric [9]. Each NPC spoke contains 4 Nsp1 copies [9], which makes it the highest abundant protein of the NPC structure. There are two FG-Nups close to the pore membrane – Nup53 and Nup59, attached to the inner ring component Nup170 [9]. Presence of these two FG-Nups in the vicinity of the peripheral channels may imply that peripheral channels are also involved in the transport processes handled by the NPC [9,12]. The key function of the NPC - forming a barrier for some molecules, while supporting the transport of others - is handled by FG-Nups. As mentioned earlier, FG-rich domains provide the binding sites for transport factors, but the precise molecular mechanism of their functioning is still under debate. Multiple models have been suggested so far (as reviewed in [8]). Virtual gating model arose from the energetic considerations of molecules moving within the pore [13]. Their movement is restricted, and therefore their entropy decreases – an ‘entropic price’ has to be paid to allow passage and this price will increase with the size of the molecule [13]. The crowded environment of the central channel of the pore, densely filled with FG-Nups, increases this entropic barrier [13]. The transport factors can overcome the barrier by binding to FG-rich regions – the enthalpy of binding cancels out the entropic barrier, this way lowering the activation energy required for passing the pore [13]. Later on Lim et al. indeed have shown that FG-Nups act together to form a repulsion zone – entropic barrier, which excludes larger molecules, while allowing small molecules to pass [14]. The interaction between the FG-Nups and transport factors overcomes this repulsion effect [14]. They have also shown that FG-Nups are elastic and can be extended from the surface in a brush-like confirmation [14]. This polymer brush collapses to the anchoring sites while binding the transport factors, in this way supporting the facilitated transport [15] – this model of transport is called ‘the reversible collapse’. Another mode of FG-Nup – transport factors interaction is presented by the ‘hydrophobic gel’ model [16,17]. In this model the FG-Nups are stably bound with each other, forming a dense gel, which can be dissolved by the transport factors [16,17]. Reduction of dimensionality model assumes that FG-Nups form a dense coat around the central channel, leaving only a small central tunnel of 10 nm in diameter free [18]. Molecules diffusing through the pore passively could be

Chapter 1

14

accommodated in this small channel, while the facilitated transport of larger molecules would require binding of transport factors to FG-Nups and a random walk through the FG surface [18]. Models that aimed at unifying the described models and observations have been proposed, like the ‘forest model’ [19,20], according to which FG-Nups can be grouped into two categories forming separate transport zones. Low charged ‘shrubs’ have domains, which attract each other and form a collapsed conformation, while ‘trees’ have higher content of charged aminoacids, have a tendency to repulse each other and their confirmation is extended [20]. Given different properties of ‘shrubs’ and ‘trees’ (attraction and repulsion), they could form zones with different gating mechanisms – a zone with the properties of the hydrogel model, and zone with the polymer brush mode of functioning [20].

SOLUBLE PROTEIN IMPORT

Despite all the uncertainties concerning the model of FG-Nup molecular behavior, the sequence of molecular events that lead a soluble molecule from the cytoplasm to the nuclear interior, and vice versa, is relatively well understood. Small molecules, like ions and small proteins diffuse freely through the pore. Transport of larger molecules, like big protein or RNA containing macromolecular complexes, is energy dependent. Many proteins carry tickets for the ride ‘in’ and ‘out’ of the nucleus imprinted in their structure. These ‘tickets’, Nuclear Localization Sequences (NLS) for the way ‘in’ and Nuclear Export Sequences (NES) for the way ‘out’, are recognized by transport factors, which interact with the NPC components and facilitate the transport. The transport factors, called karyopherins, shuttle constantly between the nucleus and cytoplasm. There are 14 karyopherins in yeast S. cerevisiae and 20 in metazoans (see [8] for complete list). The best characterized soluble protein import pathway (depicted in fig.3) depends on the classical NLS (reviewed in detail in [21]). A classical NLS is a short stretch of basic amino acids, e.g. PKKRK, sufficient for nuclear targeting [22]. In some proteins this basic stretch is repeated downstream in the protein sequence – forming a bipartite NLS [23]. In the cytoplasm NLS is recognized by importin α or karyopherin 60 (Kap60) in case of yeast S. cerevisiae. Importin α contains 2 NLS binding sites – the minor and major binding site. Generally, monopartite NLSs bind to the major binding site, while in case of bipartite NLSs the major binding site is occupied by the longest NLS stretch [24-26]. Importin α works only as an adaptor protein - it has to bind importin β (or karyopherin 95; Kap95 in yeast S. cerevisiae) via the N terminal IBB domain (importin β binding domain) of Kap 60 for the transport event to occur [27,28]. The IBB domain has a double role for importin α – next to binding importin β, it also functions as an inhibitor of NLS binding and this is used for the release of the NLS-cargo in the nucleus [29,30].

Introduction

15

Chap

ter 1

The cargo-importin complex crosses the NPC due to the interactions of the importin with the natively unfolded FG-Nups that fill up the central transport channel. The interactions between Phe residues of FG-Nups and the surface residues of import factors are weak and transient, assuring high speed of transport [31-36]. The directionality of import is regulated by the changes in the nucleotide state of the small GTPase Ran. In the nucleus RanGTP binds to the importin β complexed with importin α and cargo, and causes a conformational change in importin β and release of IBB domain, followed by dissociation of the entire importin-cargo complex [37,38]. The cargo is released and importins are rapidly recycled to the cytoplasm. Importin β cycles back complexed with RanGTP, and importin α recycling is facilitated by exportin CAS bound to RanGTP. In the cytoplasm both complexes are dissociated due to GTP hydrolysis mediated by RanGAP (Ran GTPase Activating Protein). Hydrolysis of GTP causes release of Ran from importin β and from CAS, and therefore disassembly of importin export complexes. The importins can now be used for another import cycle and RanGDP is recycled back to the nucleus by nuclear transport factor NTF2. In the nucleus RanGEF (Ran Guanine nucleotide Exchange Factor) catalyzes the change from GDP to GTP.

FIGURE 3. A SCHEMATIC REPRESENTATION OF THE IMPORT AND EXPORT CYCLE THROUGH THE NPC.

A detailed description is provided in the text. Reprinted with permission from Cold Spring Harbor Laboratory Press [8]

Chapter 1

16

The directionality of protein export from the nucleus is also dictated by RanGTP gradient, but in this case RanGTP binding supports complex formation between the karyopherin and the cargo. Recycling of the karyopherin α described above is an example of protein export event. Another example is transport mediated by exportin CRM1. Nuclear Export Signals (NESs) recognized by CRM1 are formed by 4 hydrophobic residues (Φ) separated from each other by small, charged or polar amino acids (x): Φ1-x2-3-Φ2-x2-3-Φ3-x-Φ4 [39]. The spacing in between the hydrophobic residues varies in length, and also different hydrophobic residues are allowed - leucines are most frequent, but isoleucine, valine, methionine and phenylalanine are common as well [40-42]. NESs with 5 hydrophobic residues have been described [43,44]. All these different kinds of NESs adapt their confirmation to the rigid binding pocket of CRM1 [45].The NES is bound by CRM1 in the nucleoplasm and the affinity of this interaction increases after RanGTP binding. In the cytoplasm RanGAP hydrolyzes the GTP, causing complex dissociation. In rapidly growing yeast cells, most of the traffic through the NPC concerns transport of mRNA and traffic related to ribosome synthesis. Each mRNA molecule is assembled into mRNP (RiboNuclear Particle), in a highly regulated, co-transcriptional process (reviewed in [46]). The export of mRNPs does not depend on karyopherins and the RanGTP gradient. Instead during mRNP assembly specialized transport factors, Mex67 and Mtr2 in yeast, associate with the mRNP and interact with FG-Nups in the NPC, in this way facilitating the transport. Docking at the NPC can start before the transcription is completed. The mRNP translocation through the pore is a rapid process, completed in the millisecond time range [47,48]. The export complex is disassembled on the cytoplasmic side of the NPC, due to interaction with Dbp5 (ATP-dependent RNA DEAD-box helicase) in yeast, which releases the transport factors. Ribosomal proteins are imported after their synthesis in the cytosol and together the rRNAs assembled into the ribosomal subunits 40S and 60S, which are assembled in the nucleolus and exported to the cytoplasm for the final ribosomal assembly. The ribosomal subunits are likely the biggest cargo passing the NPC. The molecular details of this export process is complex, and many transport factors are involved, including Crm1 [49-51], Mex67 and Mtr2 [52] but not fully understood. Although NPC is the only gateway to and out of the nucleus for most of the molecules, viruses have found a detour. Nucleocapsids of Herpes Simplex virus pass the NE using a nuclear egress mechanism, where the viral capsid is enveloped by the inner leaflet of the NE and subsequently crosses the perinuclear space in a vesicle. If viruses have hijacked an existing transport system or evolved a new one was not clear until recently large RNPs of Drosophila melanogaster were shown to also exit the nucleus via nuclear egress [53]. An interesting hypothesis appeared that the pathway might be universal and also used for removal of large protein aggregates from the nucleus [54]. Recently, Webster et al. saw similar perinuclear granules in a yeast strain that is impaired

Introduction

17

Chap

ter 1

in NPC assembly (vps4Δpom152Δ), where the granules might be implicated in removal of misassembled NPCs from the NE [55]. It remains to be seen how common nuclear egress is in between species and for what other transport processes it is used.

MEMBRANE PROTEINS IMPORT

Integral membrane protein transport to the nucleus is less well understood than soluble protein transport. The big uncertainty concerns the physical route of transmembrane proteins though the NPC. The way via the central channel would likely require an energetically unfavorable membrane extraction step. If these molecules are transported in the membrane environment, than their cytosolically exposed domains have to travel via the peripheral channels, oval empty spaces within the NPC structure adjacent to the nuclear envelope [10]. The space offered by the peripheral channels is limited (10nm diameter in Xenopus nuclei [10]) which could hinder the formation of importin complexes required for the active pathway. Therefore the question ‘where do the membrane proteins pass the nuclear pore’ implicates another uncertainty – about the energy dependence of the process. Multiple years of research on this topic have yielded several possibilities. Initial search for signaling sequences and domains within protein structure in vertebrate cells [56-60] did not identify a uniform mechanism. Regions implicated in import were found in different locations within proteins and had little homology. An observation of decreased import in case of proteins with larger cytoplasm exposed domains led to formation of diffusion-retention theory [56-58], according to which INM proteins enter the nucleus by passive diffusion and stay inside due to interactions with lamins and DNA. This is especially relevant for organisms with an open mitosis, in which the nuclear membrane is disassembled during cellular division. After completion of chromosome segregation membrane proteins are passively diffusing in the ER and proteins with DNA binding domains bind to chromosomes and enrich around them, likely supporting reformation of the nuclear envelope [61]. If the passive diffusion/retention model explains all transport in S. cerevisiae cells, which undergo closed mitosis, or in postmitotic cells is not clear, although most of the confirmed yeast INM proteins contain typical retention domains (LEM domain for Lap1-emerin-MAN1, MAN domain for Heh/Man1 carboxy-terminal homology domain). There are a few theories that assume energy dependent transport of membrane proteins to the nucleus. Studies on a Lap2β-based reporter in HeLa cells have proven ATP and temperature dependence of import [62]. Here, the ATP dependence was explained by proposing local restructuring of the NPC is required to provide space for passing soluble domains. A later study on the yeast S. cerevisiae proteins Src1/Heh1 and Heh2 identified involvement of components of the soluble import machinery, namely a bipartite NLS, Kap60, Kap95 and RanGTP [63]. Also an active pathway involving a co-translational licensing step was suggested, where a specific ‘INM sorting

Chapter 1

18

motif ’ few aminoacids upstream the membrane segment is recognized at the ribosome [64]. A study in 2011 from Zuleger et al. investigated the transport mechanisms of 15 human integral INM proteins by combining FRAP (Fluorescent Recovery After Photobleaching) with photoactivation techniques and discovered that different INM proteins have different energy requirements [65]. Some of the proteins seemed to follow a diffusion-retention model, while for others import dependency on RanGTP, or ATP was observed. Moreover, the study provided evidence that import of LBR depended on Nup35, yeast homologue of which, Nup53, is located in the peripheral channels [9]. Based on a slight enrichment of F and G in INM proteins, the authors postulated that FG repeats in these INM proteins may interact directly with the NPC components, without the use of importins [65]. Also another NLS-independent mechanism was proposed, involving co-transport of proteins with NLS-containing proteins, so called ‘piggyback’ transport. Yeast Mps3 likely travels to the INM in this way – in complex with histone H2A.Z [66]. With regards to their transport mechanism, the most comprehensively studied INM proteins are probably the yeast proteins Heh1 and Heh2 [63,67] and the SUN proteins from human and C. elegans [68-70]. In case of S. cerevisiae Heh1 and Heh2 the involvement of the NLS, RanGTP, karyopherins and the NPC was solidly proven, and a comprehensive analysis of additional domains involved in transport was executed ([67] and chapters 2 and 4 of this thesis). In case of SUN proteins many signals seem to converge and influence the process. Human SUN2 contains three distinct elements essential for import: a classical NLS, a Golgi retrieval signal and perinuclear SUN domain [68]. The convergence of the signals is crucial – without the Golgi retrieval signal SUN2 is mislocalized to the Golgi, but at least one of the other two domains is necessary for a successful nuclear import. C. elegans SUN protein UNC-84 contains 2 classical NLSs, an INM sorting motif for the co-translational licensing step and a SUN specific nuclear localization signal [70]. All these elements are required for efficient nuclear import [70]. SCOPE OF WORK IN THE THESIS

This thesis investigates the transport mechanism of the yeast INM protein Heh2. In Chapter 2 the sorting signals required for import of Heh2 are identified and a mechanism for it import is proposed. We show that a nuclear localization signal (NLS) plus an intrinsically disordered linker represent a sorting signal that is both required and sufficient for import of a membrane protein. Here, and also in the later chapters, I use a newly established trapping assay based on the Anchor Away technique [71], which takes advantage of the rapamycin dependent high affinity binding between FKBP (12-kDa FK506 binding protein) and FRB (FKBP-rapamycin binding domain of the mammalian target of rapamycin). I use this assay to stabilize NPC resident membrane proteins (Chapter 3 and 4) and to probe where the membrane proteins can reside inside the NPC (Chapter 2 and 4).

Introduction

19

Chap

ter 1

Two questions related to the transport mechanism proposed in Chapter 2 were further addressed in Chapters 3, 4 and 5. The first question is where the membrane proteins traffic through the NPC: the central or peripheral channels. The second question is, if the transmembrane domains indeed remain membrane embedded during import as proposed, or that they may be soluble and chaperoned during import. In chapter 3 we explored the possibilities of super resolution imaging techniques to simultaneously visualize labeled NPCs and trapped labeled membrane proteins and as such shed light on the above questions. We visualized trapped reporters at high resolution, but failed to find conditions for simultaneous visualization with the NPCs, and hence the method did not yield an answer to where and how the proteins pass through the NPC. In Chapter 4 we expanded the trapping studies from chapter 2 towards multiple trap positions inside the NPC, namely Nup53, Nup59 and Nup170, and multiple different membrane proteins with variable domain compositions. The results show that the membrane proteins pass through the NPC while membrane embedded, answering our second question. In Chapter 5 we show off-NPC trapping events related to NPC inheritance and quality control. Chapter 4 and 5 together, show that a definitive interpretation of the trapping events in terms of a translocation path through the NPC is not possible and the route through the NPC remains undefined. Chapter 6 provides a review summarizing our knowledge of the mechanisms for the traffic of membrane proteins in yeast and metazoans. It also discusses the challenges of investigating transmembrane protein import to the nucleus.

Chapter 1

20

REFERENCES

1. Brown R: Observations on the organs and mode of fecundation in Orchideae and Asclepiadeae. 1833.2. Rout MP, Karpen GH: Editorial overview: Cell nucleus: the nucleus: a dynamic organelle.

Current Opinion in Cell Biology 2014, 28:iv–vii.3. Mor A, White MA, Fontoura BM: Nuclear Trafficking in Health and Disease. Current Opinion in

Cell Biology 2014, 28:28–35.4. Shen A, Wang Y, Zhao Y, Zou L, Sun L, Cheng C: Expression of CRM1 in human gliomas and its

significance in p27 expression and clinical prognosis. Neurosurgery 2009, 65:153–160.5. Turner JG, Dawson J, Sullivan DM: Nuclear export of proteins and drug resistance in cancer.

Biochem Pharmacol 2012, 83:1021–1032.6. Yao Y, Dong Y, Lin F, Zhao H, Shen Z, Chen P, Sun Y-J, Tang L-N, Zheng S-E: The expression of

CRM1 is associated with prognosis in human osteosarcoma. Oncol Rep 2008, 21:229–235.7. Burns LT, Wente SR: Trafficking to uncharted territory of the nuclear envelope. Current Opinion

in Cell Biology 2012, 24:341–349.8. Wente SR, Rout MP: The Nuclear Pore Complex and Nuclear Transport. Cold Spring Harbor

Perspectives in Biology 2010, 2:a000562–a000562.9. Alber F, Dokudovskaya S, Veenhoff LM, Zhang W, Kipper J, Devos D, Suprapto A, Karni-Schmidt

O, Williams R, Chait BT, et al.: The molecular architecture of the nuclear pore complex. Nature 2007, 450:695–701.

10. Hinshaw JE, Carragher BO, Milligan RA: Architecture and design of the nuclear pore complex. Cell 1992, 69:1133–1141.

11. Szymborska A, de Marco A, Daigle N, Cordes VC, Briggs JAG, Ellenberg J: Nuclear Pore Scaffold Structure Analyzed by Super-Resolution Microscopy and Particle Averaging. Science 2013, 341:655–658.

12. Hawryluk-Gara LA, Shibuya EK, Wozniak RW: Vertebrate Nup53 interacts with the nuclear lamina and is required for the assembly of a Nup93-containing complex. Molecular Biology of the Cell 2005, 16:2382–2394.

13. Rout M: Virtual gating and nuclear transport: the hole picture. Trends in Cell Biology 2003, 13:622–628.

14. Lim RYH, Huang N-P, Köser J, Deng J, Lau KHA, Schwarz-Herion K, Fahrenkrog B, Aebi U: Flexible phenylalanine-glycine nucleoporins as entropic barriers to nucleocytoplasmic transport. Proc. Natl. Acad. Sci. U.S.A. 2006, 103:9512–9517.

15. Lim RYH, Fahrenkrog B, Köser J, Schwarz-Herion K, Deng J, Aebi U: Nanomechanical basis of selective gating by the nuclear pore complex. Science 2007, 318:640–643.

16. Ribbeck K, Görlich D: The permeability barrier of nuclear pore complexes appears to operate via hydrophobic exclusion. The EMBO Journal 2002, 21:2664–2671.

17. Frey S, Richter RP, Görlich D: FG-rich repeats of nuclear pore proteins form a three-dimensional meshwork with hydrogel-like properties. Science 2006, 314:815–817.

18. Peters R: Translocation Through the Nuclear Pore Complex: Selectivity and Speed by Reduction-of-Dimensionality. Traffic 2005, 6:421–427.

19. Patel SS, Belmont BJ, Sante JM, Rexach MF: Natively Unfolded Nucleoporins Gate Protein Diffusion across the Nuclear Pore Complex. Cell 2007, 129:83–96.

20. Yamada J, Phillips JL, Patel S, Goldfien G, Calestagne-Morelli A, Huang H, Reza R, Acheson J, Krishnan VV, Newsam S, et al.: A bimodal distribution of two distinct categories of intrinsically

Introduction

21

Chap

ter 1

disordered structures with separate functions in FG nucleoporins. Mol. Cell Proteomics 2010, 9:2205–2224.

21. Stewart M: Molecular mechanism of the nuclear protein import cycle. Nat Rev Mol Cell Biol 2007, 8:195–208.

22. Goldfarb DS, Gariépy J, Schoolnik G, Kornberg RD: Synthetic peptides as nuclear localization signals. Nature 1986, 322:641–644.

23. Dingwall C, Robbins J, Dilworth SM, Roberts B, Richardson WD: The nucleoplasmin nuclear location sequence is larger and more complex than that of SV-40 large T antigen. The Journal of Cell Biology 1988, 107:841–849.

24. Conti E, Uy M, Leighton L, Blobel G, Kuriyan J: Crystallographic analysis of the recognition of a nuclear localization signal by the nuclear import factor karyopherin alpha. Cell 1998, 94:193–204.

25. Conti E, Kuriyan J: Crystallographic analysis of the specific yet versatile recognition of distinct nuclear localization signals by karyopherin alpha. Structure/Folding and Design 2000, 8:329–338.

26. Fontes MR, Teh T, Kobe B: Structural basis of recognition of monopartite and bipartite nuclear localization sequences by mammalian importin-alpha. J. Mol. Biol. 2000, 297:1183–1194.

27. Görlich D, Henklein P, Laskey RA, Hartmann E: A 41 amino acid motif in importin-alpha confers binding to importin-beta and hence transit into the nucleus. The EMBO Journal 1996, 15:1810–1817.

28. Weis K, Ryder U, Lamond AI: The conserved amino-terminal domain of hSRP1 alpha is essential for nuclear protein import. The EMBO Journal 1996, 15:1818–1825.

29. Kobe B: Autoinhibition by an internal nuclear localization signal revealed by the crystal structure of mammalian importin alpha. Nat. Struct. Biol. 1999, 6:388–397.

30. Matsuura Y, Stewart M: Structural basis for the assembly of a nuclear export complex. Nature 2004, 432:872–877.

31. Bayliss R, Ribbeck K, Akin D, Kent HM, Feldherr CM, Görlich D, Stewart M: Interaction between NTF2 and xFxFG-containing nucleoporins is required to mediate nuclear import of RanGDP. J. Mol. Biol. 1999, 293:579–593.

32. Bayliss R, Littlewood T, Stewart M: Structural basis for the interaction between FxFG nucleoporin repeats and importin-beta in nuclear trafficking. Cell 2000, 102:99–108.

33. Bayliss R, Littlewood T, Strawn LA, Wente SR, Stewart M: GLFG and FxFG nucleoporins bind to overlapping sites on importin-beta. J. Biol. Chem. 2002, 277:50597–50606.

34. Bayliss R, Leung SW, Baker RP, Quimby BB, Corbett AH, Stewart M: Structural basis for the interaction between NTF2 and nucleoporin FxFG repeats. The EMBO Journal 2002, 21:2843–2853.

35. Fribourg S, Braun IC, Izaurralde E, Conti E: Structural basis for the recognition of a nucleoporin FG repeat by the NTF2-like domain of the TAP/p15 mRNA nuclear export factor. Molecular Cell 2001, 8:645–656.

36. Grant RP, Neuhaus D, Stewart M: Structural basis for the interaction between the Tap/NXF1 UBA domain and FG nucleoporins at 1A resolution. J. Mol. Biol. 2003, 326:849–858.

37. Vetter IR, Arndt A, Kutay U, Görlich D, Wittinghofer A: Structural view of the Ran-Importin beta interaction at 2.3 A resolution. Cell 1999, 97:635–646.

38. Lee SJ, Matsuura Y, Liu SM, Stewart M: Structural basis for nuclear import complex dissociation by RanGTP. Nature 2005, 435:693–696.

39. Kutay U, Güttinger S: Leucine-rich nuclear-export signals: born to be weak. Trends in Cell Biology 2005, 15:121–124.

Chapter 1

22

40. Bogerd HP, Fridell RA, Benson RE, Hua J, Cullen BR: Protein sequence requirements for function of the human T-cell leukemia virus type 1 Rex nuclear export signal delineated by a novel in vivo randomization-selection assay. Molecular and Cellular Biology 1996, 16:4207–4214.

41. Zhang MJ, Dayton AI: Tolerance of diverse amino acid substitutions at conserved positions in the nuclear export signal (NES) of HIV-1 Rev. Biochem. Biophys. Res. Commun. 1998, 243:113–116.

42. Kosugi S, Hasebe M, Tomita M, Yanagawa H: Nuclear export signal consensus sequences defined using a localization-based yeast selection system. Traffic 2008, 9:2053–2062.

43. Monecke T, Güttler T, Neumann P, Dickmanns A, Görlich D, Ficner R: Crystal structure of the nuclear export receptor CRM1 in complex with Snurportin1 and RanGTP. Science 2009, 324:1087–1091.

44. Dong X, Biswas A, Süel KE, Jackson LK, Martinez R, Gu H, Chook YM: Structural basis for leucine-rich nuclear export signal recognition by CRM1. Nature 2009, 458:1136–1141.

45. Güttler T, Madl T, Neumann P, Deichsel D, Corsini L, Monecke T, Ficner R, Sattler M, Görlich D: NES consensus for CRM1_gorlich lab. Nat Struct Mol Biol 2010, 17:1367–1376.

46. Rodríguez-Navarro S, Hurt E: Linking gene regulation to mRNA production and export. Current Opinion in Cell Biology 2011, 23:302–309.

47. Grünwald D, Singer RH: in vivo imaging of labeled endogenous bactin mrna during nucleocytoplasmic transport [Internet]. Nature 2010, 467:604–607.

48. Mor A, Suliman S, Ben-Yishay R, Yunger S, Brody Y, Shav-Tal Y: Dynamics of single mRNP nucleocytoplasmic transport and export through the nuclear pore in living cells. Nat Cell Biol 2010, 12:543–552.

49. Fornerod M, Ohno M, Yoshida M, Mattaj IW: CRM1 is an export receptor for leucine-rich nuclear export signals. Cell 1997, 90:1051–1060.

50. Fukuda M, Asano S, Nakamura T, Adachi M, Yoshida M, Yanagida M, Nishida E: CRM1 is responsible for intracellular transport mediated by the nuclear export signal. Nature 1997, 390:308–311.

51. Stade K, Ford CS, Guthrie C, Weis K: Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 1997, 90:1041–1050.

52. Yao W, Roser D, Köhler A, Bradatsch B, Bassler J, Hurt E: Nuclear export of ribosomal 60S subunits by the general mRNA export receptor Mex67-Mtr2. Molecular Cell 2007, 26:51–62.

53. Speese SD, Ashley J, Jokhi V, Nunnari J, Barria R, Li Y, Ataman B, Koon A, Chang Y-T, Li Q, et al.: Nuclear Envelope Budding Enables Large Ribonucleoprotein Particle Export during Synaptic Wnt Signaling. Cell 2012, 149:832–846.

54. Rose A, Schlieker C: Alternative nuclear transport for cellular protein quality control. Trends in Cell Biology 2012, 22:509–514.

55. Webster BM, Colombi P, Jäger J, Lusk CP: Surveillance of Nuclear Pore Complex Assembly by ESCRT-III/Vps4. Cell 2014, 159:388–401.

56. Soullam B, Worman HJ: Signals and structural features involved in integral membrane protein targeting to the inner nuclear membrane. The Journal of Cell Biology 1995, 130:15–27.

57. Smith S, Blobel G: The first membrane spanning region of the lamin B receptor is sufficient for sorting to the inner nuclear membrane. The Journal of Cell Biology 1993, 120:631–637.

58. Wu W, Lin F, Worman HJ: Intracellular trafficking of MAN1, an integral protein of the nuclear envelope inner membrane. Journal of Cell Science 2002, 115:1361–1371.

59. Furukawa K, Panté N, Aebi U, Gerace L: Cloning of a cDNA for lamina-associated polypeptide 2 (LAP2) and identification of regions that specify targeting to the nuclear envelope. The EMBO Journal 1995, 14:1626–1636.

Introduction

23

Chap

ter 1

60. Furukawa K: The Major Nuclear Envelope Targeting Domain of LAP2 Coincides with Its Lamin Binding Region but Is Distinct from Its Chromatin Interaction Domain. Journal of Biological Chemistry 1998, 273:4213–4219.

61. Anderson DJ, Vargas JD, Hsiao JP, Hetzer MW: Recruitment of functionally distinct membrane proteins to chromatin mediates nuclear envelope formation in vivo. The Journal of Cell Biology 2009, 186:183–191.

62. Ohba T: Energy- and temperature-dependent transport of integral proteins to the inner nuclear membrane via the nuclear pore. The Journal of Cell Biology 2004, 167:1051–1062.

63. King MC, Patrick Lusk C, Blobel G: Karyopherin-mediated import of integral inner nuclear membrane proteins. Nature 2006, 442:1003–1007.

64. Saksena S, Shao Y, Braunagel SC, Summers MD, Johnson AE: Cotranslational integration and initial sorting at the endoplasmic reticulum translocon of proteins destined for the inner nuclear membrane. Proc. Natl. Acad. Sci. U.S.A. 2004, 101:12537–12542.

65. Zuleger N, Kelly DA, Richardson AC, Kerr ARW, Goldberg MW, Goryachev AB, Schirmer EC: System analysis shows distinct mechanisms and common principles of nuclear envelope protein dynamics. The Journal of Cell Biology 2011, 193:109–123.

66. Gardner JM, Smoyer CJ, Stensrud ES, Alexander R, Gogol M, Wiegraebe W, Jaspersen SL: Targeting of the SUN protein Mps3 to the inner nuclear membrane by the histone variant H2A.Z. The Journal of Cell Biology 2011, 193:489–507.

67. Meinema AC, Laba JK, Hapsari RA, Otten R, Mulder FAA, Kralt A, van den Bogaart G, Lusk CP, Poolman B, Veenhoff LM: Long unfolded linkers facilitate membrane protein import through the nuclear pore complex. Science 2011, 333:90–93.

68. Turgay Y, Ungricht R, Rothballer A, Kiss A, Csucs G, Horvath P, Kutay U: A classical NLS and the SUN domain contribute to the targeting of SUN2 to the INM. The EMBO Journal 2010, doi:10.1038/emboj.2010.119.

69. Hodzic DM: Sun2 Is a Novel Mammalian Inner Nuclear Membrane Protein. Journal of Biological Chemistry 2004, 279:25805–25812.

70. Tapley EC, Ly N, Starr DA: Multiple mechanisms actively target the SUN protein UNC-84 to the inner nuclear membrane. Molecular Biology of the Cell 2011, 22:1739–1752.

71. Haruki H, Nishikawa J, Laemmli UK: The Anchor-Away Technique: Rapid, Conditional Establishment of Yeast Mutant Phenotypes. Molecular Cell 2008, 31:925–932.

72. Schermelleh L, Carlton PM, Haase S, Shao L, Winoto L, Kner P, Burke B, Cardoso MC, Agard DA, Gustafsson MGL, et al.: Subdiffraction Multicolor Imaging of the Nuclear Periphery with 3D Structured Illumination Microscopy. Science 2008, 320:1332–1336.

Chapter 2Long Unfolded Linkers Facilitate Membrane Protein Import

Through the Nuclear Pore Complex

Anne C. Meinema*, Justyna K. Laba*, Rizqiya A. Hapsari*, Renee Otten, Frans A. A. Mulder, Annemarie Kralt, Geert van den Bogaart, C. Patrick Lusk, Bert Poolman, Liesbeth M. Veenhoff

* These authors contributed equally to this work.Reprinted with permission from AAAS.

The work performed in this chapter is also described in the theses of Anne C. Meinema and Rizqiya A. Hapsari.

The Materials and methods section is available at www.sciencemag.org .

Chapter 2

26

ABSTRACT

Active nuclear import of soluble cargo involves transport factors that shuttle cargo through the nuclear pore complex (NPC) by binding to phenylalanine-glycine (FG)- domains. How nuclear membrane proteins cross through the NPC to reach the inner membrane is presently unclear. Here, we found that, at least a 120 residue long intrinsically disordered linker was required for the import of membrane proteins carrying a nuclear localization signal for the transport factor karyopherin-α. We propose an import mechanism for membrane proteins in which an unfolded linker slices through the NPC scaffold to enable binding between the transport factor and the FG-domains in the centre of the NPC.


27

Chap

ter 2

The nuclear envelope (NE) consists of an inner and outer nuclear membrane (INM, ONM) connected by the pore membrane at sites where the NPCs are embedded. The ONM is continuous with the endoplasmic reticulum (ER). NPCs are composed of a membrane-anchored scaffold that stabilizes a cylindrical central channel, in which nucleoporins (Nups) with disordered phenylalanine–glycine (FG)-rich regions provide the selectivity barrier [1]. In order for a membrane protein to move through the NPC, its transmembrane domains pass through the pore membrane, while its extra-luminal soluble domain(s) pass through the NPC by a mechanism yet to be clarified [2-4]. Some proteins reach the INM by diffusion through the pore membrane and adjacent lateral channels [5-8] and accumulate by binding nuclear structures [9,10]. Other membrane proteins have a nuclear localization signal (NLS), and binding to transport factors karyopherin-α and karyopherin-ß1 is required to pass the NPC and reach the INM [11,12]. We set out to investigate the mechanism and path of nuclear transport of these integral INM proteins. A

h2N

LS-L

-TM

w303 wt

nup170Δ

mtr1-125 C°

mtr1-137 C°

KAP95-AA

KAP95-AA+RAP

B

Heh2NE/ER = 18.5 ± 2.2( = 35)n

h2NLS-L-TMNE/ER = 32.5 ± 3.1( = 45)n

L-TMNE/ER = 2.3 ± 0.2( = 22)n

DC

E RAP

Time (min)30

Heh2h2NLS-L-TM

90601A

ccum

ulat

ion

at N

E (N

E/E

R)

40

30

20

10

1Acc

umul

atio

n at

NE

(NE

/ER

)

40

30

20

10

0

F

h2NLS-L-TML-TM

tcNLS-L-TM

sp h2NLS-L-TM

93

93

TM2MAN1

TM1LD

378

378

-C

-C

663

L-TM

h2NLS-L-TM

Heh2

Reporter (transmembrane)

GFP LEM L

N-

N-

N-

C-h2NLS: 102-KRKR-105 ... 124-PKKKRKKRSSKANK-137

1

138

138

138

INMONM

Nucleus

Nucleus

FIGURE 1. THE NLS-CONTAINING DOMAIN (H2NLS-L) OF HEH2 IS SUFFICIENT FOR ACCUMULATION AT THE INM.

A. Representation of Heh2-based GFP-fusion reporter proteins. B. Confocal fluorescence images of yeast expressing the indicated proteins. Average NE/ER ratios are shown. C. Immuno-electron micrograph of h2NLS-L-TM in the KAP95-AA strain labelled with anti-GFP antibodies and 10-nm diameter gold-conjugated secondary antibody: 64% at the INM (n = 350, fig. S1D). D. h2NLS-L-TM is mis-localized in a Nup170Δ strain (left panel), in a RanGEF mutant strain (mtr1-1) at non-permissive temperature (middle panel), and the KAP95-AA strain, upon addition of rapamycin (RAP) (right panel). E. The accumulation at the NE of h2NLS-L-TM (▲) and Heh2 (■) in the KAP95-AA strain as function of time after anchoring of Kap95 (RAP at t = 0, n ≥ 13). F. The accumulation at the INM of reporter containing a bipartite h2NLS (h2NLS-L-TM), without NLS (L-TM), with single partite NLS (sp h2NLS-L-TM), or with tandem cNLS (tcNLS-L-TM) (n ≥ 32). SEM is indicated, scale bars: (B and D) 5 μm and (C) 250 nm.

Chapter 2

28

We first generated reporters using the Saccharomyces cerevisiae homolog of the human LEM domain-containing integral INM protein, Heh2. Heh2 is composed of a LEM domain, a bipartite NLS (hereafter h2NLS), a linker region (L), two transmembrane segments (TM) flanking a luminal domain (LD), and a domain with homology to the C-terminus of MAN1 (Fig. 1A) [12]. The h2NLS is recognized by Kap60 (Srp1/Karyopherin-α), the yeast homolog of human Importin-α [12]. Similar to Heh2, the reporter protein h2NLS-L-TM, consisting of GFP fused to amino acids 93-378 of Heh2, accumulated specifically at the NE (Fig. 1B). A control lacking the h2NLS, named L-TM, distributed over the NE and cortical ER. While we could not resolve the INM from the ONM, we used the average pixel intensities at the NE and ER (NE-ER ratio) as a measure for INM accumulation. We validated this approach by confirming the localization of h2NLS-L-TM to the INM using immuno-electron microscopy (Fig. 1C). h2NLS-L-TM accumulated 33-fold at the NE (Fig.1B), whereas L-TM accumulated only 2-fold.

GFP NLS L/ /LR1 LR2 TM

Lh2NLS- -TMh2NLS- -TMh2NLS- -TM

LR1LR2

ppm10.5 8.5 6.5

)R

E/E

N( E

N ta noitalumucc

A

A

40

30

20

10

50

Linker length (number of residues)150100 00205

C

MT--

SLN2h

1RL

N-

B

10

h2NLS-L

LR1 = 78LR1 = 178 LR1 = 178

MT--

SLN2h

2RL

LR2 = 75LR2 = 215LR2 = 215

L = 180

-SL

N2hMT-L

L = 90

- RAP + RAP - RAP

L = 180

-C

calbindin-D9K

α-synuclein

FIG. 2. REPORTER PROTEINS CONTAINING SYNTHETIC UNFOLDED LINKERS LOCALIZE AT THE INM.

A. 1D 1H-NMR of the backbone amides for (unlabeled) h2NLS-L. Comparison with the intrinsically-disordered α-synuclein and the folded calbindin-D9k show that h2NLS-L is natively unstructured. B. Localization of the indicated reporters with native linker (L) and the randomized versions LR1 and LR2 in the KAP95-AA strain w/wo rapamycin (RAP). Right panels show localization of shortened linkers. Linker length is in number of amino acids. C. The accumulation at the NE of h2NLS-L-TM (■), h2NLS-LR1-TM (▲) and h2NLS-LR2-TM (●) and truncations thereof, plotted against the length of the linker domain (n ≥ 20). SEM is indicated, scale bars: 5 μm.

GFP NLS L/ /LR1 LR2 TM

Lh2NLS- -TMh2NLS- -TMh2NLS- -TM

LR1LR2

ppm10.5 8.5 6.5)

RE/

EN(

EN ta noitalu

muccA

A

40

30

20

10

50

Linker length (number of residues)150100 00205

C

MT--

SLN2h

1RL

N-

B

10

h2NLS-L

LR1 = 78LR1 = 178 LR1 = 178

MT--

SLN2h

2RL

LR2 = 75LR2 = 215LR2 = 215

L = 180

-SL

N2hMT-L

L = 90

- RAP + RAP - RAP

L = 180

-C

calbindin-D9K

α-synuclein


29

Chap

ter 2

Transport of h2NLS-L-TM was dependent on the Ran-gradient and Nup170, similar to full-length Heh2 (Fig. 1D) [12]. To confirm that the import of our membrane reporter was Kap60/95-mediated, we examined the distribution of h2NLS-L-TM in a Kap95 (Karyopherin-ß/Importin-ß/Rsl1) “anchor away” strain (KAP95-AA) [13]. Upon addition of rapamycin, Kap95-FRB was trapped at Pma1-FKBP in the plasma membrane and no longer available for nuclear transport. Indeed, the accumulation of h2NLS-L-TM at the NE reduced dramatically (+RAP, Fig. 1D). Moreover, INM-localized reporter proteins redistributed to the ONM and ER upon addition of rapamycin, and the nuclear accumulation dropped with a half-time of 14±2.7 minutes (Fig. 1E). In contrast, the fluorescence intensity of Heh2 at the NE remained unaltered for over 90 minutes. Thus, while Heh2 is bound to nuclear factors, h2NLS-L-TM is fully mobile within the NE-ER network.

h2N

LS-

-W

alp2

3LR2

- RAP + RAP

A

-CN-

B

Time (min)30 9060

10

Acc

umul

atio

n at

NE

(NE

/ER

)40

30

20

10

h2NLS-L-TMh2NLS-RL2-WALP23

RAPLR=215 WALP23

N- -C

-CN-LR1=138

Sec61-TM1

h2N

LS-

-S

ec61

(TM

1)LR

1h2

NLS

--

Sec

61LR1

w/o h2NLSLR1=138

h2NLSLR1=78

C

h2NLSLR1=138

w/o h2NLSLR1=138

h2NLSLR1=78

h2NLSLR1=138

LR1=138

Sec61

FIG. 3. SYNTHETIC TRANSMEMBRANE PEPTIDES AND ER-PROTEINS CAN BE TARGETED TO THE INM.

A. Representation of h2NLS-LR2 fused to the WALP23 TM-region and images of its localization w/wo rapamycin (RAP). B. After adding rapamycin, the reporter h2NLS-LR2-WALP23 leaked to the ER with similar kinetics as h2NLS-L-TM. C. NE-localization of h2NLS-LR1(138) reporters containing the first TM of Sec61 (top, left) and full-length Sec61 (bottom, left). Without the h2NLS (middle) and with shorter linker, LR1(78) (right), accumulation is lost. SEM is indicated, scale bars: 5 μm.

Chapter 2

30

The h2NLS is a high affinity NLS compared to the classical NLS. To assess whether this high affinity is actually required for import of h2NLS-L-TM, we replaced the bipartite h2NLS with lower affinity NLSs: either a single-partite version of the h2NLS that lacked the first KRKR basic region (sp h2NLS) or a tandem classical NLS (tcNLS). Both membrane reporters still accumulated at the INM, but the NE/ER ratios were lower (8.1 and 4.0, respectively) than for h2NLS-L-TM (Fig. 1F), indicating a correlation between the affinity of Kap60 for an NLS and the nuclear accumulation of membrane proteins.

1

Nsp1-FRB+RAP

Aw

t (w

303)

Acc

umul

atio

n at

NE

(NE

/ER

)

tcNLS-GFP

3

2

4

1

40

30

10

20

1

Nup159 FGNup42 FG

Nup100 GLFGNup116 GLFGNup49 GLFGNup53 FGNup59 FGNsp1 FG,FxFGNup57 GLFGNup145N GLFG

Nup60 FxFNup1 FxFGNup2 FxFG

SW

Y29

50

SW

Y30

42S

WY

3062

Deleted FG-domains

Nuc

lear

acc

umul

atio

n (N

/C)

C

Nup locali-zation

FG-domain

0 kDa

+MBP

-CN-

-CN-

174

-CN-

92

+2xMBP

+3xMBP

h2N

LS-L

-TM

+MBPNE/ER = 35.5 ±4.8 ( = 29)n

+2xMBPNE/ER = 18.9 ±2.3 ( = 42)n

+3xMBPNE/ER = 11.0 ±2.3 ( = 46)n

BMBP

0 20 40 601

Acc

umul

atio

n at

NE

(NE

/ER

)

60

40

20

80C

ells

sho

win

g re

porte

r at E

R (%

)

Cells with reporter at ER+ RAP+ RAP/Gluc

Accumulation at NE+ RAP

75

25

0

50

Time (min)

100100

D

80

-CN-2xFKBP

133

h2NLS-L-TM

-CN-PrA 2xFKBP

FKBP-h2NLS-L-TM PrA-FKBP-h2NLS-L-TM

0 2 4

Fluo

resc

ence

inte

nsity

(a.u

)

0

3

Dev

iatio

n

0

3

Nsp1-FRB-RAP

0 2 4Position at NE (μm) Position at NE (μm)Deviation = 27.3 ± 0.8%( = 39)n

Deviation = 41.9 ± 1.6%( = 51)n

1

0 0

2 2

cyto

-pl

asm

icsy

mm

etric

nucl

eo-

plas

mic

FIG. 4. MEMBRANE PROTEIN REPORTERS INTERACT WITH CENTRAL CHANNEL FG-NUPS DURING IMPORT.

A. The nuclear accumulation of tcNLS-GFP (soluble) and h2NLS-L-TM, in wild type and mutant strains with FG-domain deletions (n ≥ 21) [17]. B. Localization of reporters containing soluble domains of increasing size. The accumulation at the NE is indicated.C. Localization of a reporter with an N-terminal FKBP-tag in a strain expressing Nsp1-FRB before (left) and after addition of rapamycin (right). Trapping of FKBP-tagged reporter at NPCs is apparent from punctate staining; the deviation in fluorescence at the NE is higher in the presence of rapamycin. D. Rapamycin-dependent trapping of PrA-FKBP-tagged reporter at Nsp1-FRB blocked import as observed from increased ER-localized reporter. Percentage of cells showing fluorescence at the ER (n ≥ 100, bars) and the average NE/ER ratio (n ≥ 13, symbols) upon addition of rapamycin (RAP, filled bars and ) or glucose (inhibition of reporter synthesis) and rapamycin (RAP/Gluc, open bars). SEM is indicated, scale bars: (B) 5 μm and (C) 2 μm.


31

Chap

ter 2

a

b c

d

e

FIGURE 5: MEMBRANE PROTEIN REPORTERS INTERACT WITH CENTRAL CHANNEL FG-NUPS DURING IMPORT.

A. In order to trap the N-terminus of the reporter at the pore side of the nucleus, a 2xFKBP was fused to the N-terminus of the h2NLS-L-TM reporter (1). FRB was fused to the C-terminus of Nsp1, an FG-Nup that is present in multiple copies and anchored to the NPC scaffold by its C-terminal domain. Once rapamycin is added to the medium it will bind to the FKBP at the reporter (2), enabling the FKBP to bind to FRB (3). B. Confocal fluorescence images show localization of FKBP-h2NLS-L-TM in a wild-type (K14708) strain (left) and a strain expressing Nsp1-FRB. Before addition of rapamycin (RAP), the reporter has a uniform NE stain (middle) and 40 minutes after addition of rapamycin (RAP) the reporter binds to Nsp1-FRB and shows a punctuate localization typical for NPC-localized proteins (right). C. Confocal fluorescence images show localization of a reporter containing an N-terminally PrA-FKBP-tag in a strain expressing Nsp1-FRB. Before rapamycin-dependent trapping to Nsp1, the reporter accumulates at the INM (left); after addition of rapamycin (RAP) the reporter binds to Nsp1-FRB and blocks the transport pathway so newly synthesized reporter proteins no longer have access to the INM and stay at the ER (middle). In controls where expression of new reporter was inhibited by addition of glucose while rapamycin was added (RAP/Gluc, right), no reporter was found at the ER, implying that INM-cumulated reporter does not leak out under these conditions. Images are shown

Chapter 2

32

We then addressed how the L-domain contributes to targeting. The amino acid composition of the L-domain and the large Stokes radius (45 Å) of purified recombinant h2NLS-L suggest it is unstructured. In addition, nuclear magnetic resonance (NMR) spectra of (unlabeled) h2NLS-L were typical of disordered proteins. The absence of stable secondary and tertiary structure was gauged from a lack of signal dispersion of the backbone amides for h2NLS-L in 1D 1H-NMR spectra (Fig. 2A, shaded area) and of the side chain methyl signals in [1H-13C]-HSQC spectra (data not shown). To evaluate whether the sequence of the linker region contributed to targeting, we replaced the coding regions of the L-domain in h2NLS-L-TM with two synthetic sequences, LR1 and LR2. These were generated randomly but had the same relative amino acid abundance as L. LR1 and LR2 are also predicted to be unfolded. Both h2NLS-LR1-TM and h2NLS-LR2-TM were efficiently transported to the INM in a Kap-dependent manner (Fig. 2B). Systematic truncations of LR1 and LR2 and the original linker (L) resulted in three sets of reporters with variable linker lengths. The shortest truncations of each linker set did not accumulate at the nucleus (Fig. 2B). In fact, we observed a striking dependence of INM import on linker length (Fig. 2C). Importantly, reporters with a synthetic transmembrane segment and reporters with one, or all ten transmembrane segments of an ER protein, Sec61, were also efficiently imported to the INM (Fig. 3). An “NLS-L-TM”-sorting signal could be recognized in Heh1 and indeed its NLS-linker-domain, even though lacking homology to that of Heh2, promoted INM targeting (data not shown). Next, we determined whether the transport of the reporters across the NPC depends on specific FG-regions of nucleoporins [14-16]. A strain that lack the GLFG repeats of Nups 100, 145 and 57 [17], which are anchored to both the cytoplasmic and nucleoplasmic halves of the NPC scaffold [18], showed 7.5-fold decreased NE accumulation (SW2950, Fig. 4A). Minimal effects were seen with single deletions (data not shown) and in strains lacking the FG-regions from the asymmetric localized Nups (SW3062, SW3042), whereas, Kap60/95-mediated transport of soluble cargo (tcNLS-GFP) was affected in all three strains.

with increased contrast settings to visualize the ER (bottom). No punctuated NE is visible, because the expression of PrA-FKBP-tagged reporter is 10x higher than the FKBP-tagged reporter. D. The nuclear accumulation of the FKBP-reporter was similar in wt (K14708) and Nsp1-FRB. Moreover, the nuclear accumulation of h2NLS-L-TM in Nsp1-FRB was similar to wt (K14708), with and without rapamycin. The NE/ER ratio’s of FKBP-h2NLS-L-TM should be considered a lower limit as they are based on the small fraction of cells (5%) that show fluorescence at the ER. See methods for details. E. The function of Nsp1-FRB was not affected when the FKBP-reporter was bound to it. Confocal images of NSP1-FRB cells showing the expression of FKBP-h2NLS-L-TM fused to GFP (left; green), that of h2NLS-L fused to mCherry (middle; red), and the merger (right). The nuclear accumulation of the soluble h2NLS-L was similar in the absence (top) or presence (bottom) of rapamcyin. This means that Nsp1 is still functional for non-membrane reporters bearing the same h2NLS.


33

Chap

ter 2

Our data point toward passage of the extra-luminal soluble domains of the membrane proteins through the central channel, which is expected to place little constraints on the bulkiness of these domains. Indeed, membrane proteins with up to 174 kDa soluble domains were imported to the INM, although the efficiency went down with increasing size (Fig. 4B). To further support the suggestion that the extra-luminal soluble domains pass through the central channel, we designed experiments to trap the reporters in transit through the NPC. We constructed a strain expressing FRB tagged FG-Nup Nsp1. The C-terminal FRB-tag on Nsp1 is anchored on the pore side of the scaffold of the NPC [16,18-20]. A reporter containing FKBP at its N-terminus was expressed to enable rapamycin-dependent trapping at Nsp1-FRB in the NPC (fig. 5A). Addition of rapamycin yielded a punctate stain typical of NPC-localized proteins; without rapamycin the reporter distributed evenly over the NE (Fig. 4C and fig 5B). Next, we assessed if trapping of the reporter at the NPC affected transport. We used a reporter expressed at higher levels (with an additional N-terminal Protein A tag) and saw a blockage of INM import and steady increase in fluorescence at the ER from newly synthesized proteins, following rapamycin addition (Fig. 4D and fig. 5C). Trapping of the reporter specifically blocked transport of membrane proteins and not soluble proteins (fig. 5 D and E). Thus the h2NLS-containing N-terminus of the reporter passes where Nsp1 is anchored to the NPC scaffold and within the central channel of the NPC. Here we have elucidated the NLS-dependent mechanism of membrane protein transport through the NPC. The Heh2-derived reporter proteins accumulate at the INM, not because they are retained or trapped at the INM, but because Kap60/95-mediated import is faster than export. The signal for targeting to the INM is composed of a natively unfolded linker that spaces the transmembrane segment and a high affinity NLS. It takes little energy to stretch the linker to allow the NLS, with bound karyopherins, to dodge between the NPC scaffold and allowing the karyopherins to bind the FG-Nups. The proposed transport route implies that, at least transiently, openings must exist between the space immediately aligning the pore membrane and the central channel. At present, structures of the NPC do not have the resolution to reveal such conduits, but its plasticity and the overall lattice-like scaffold structure observed in electron microscopy [8,21,22] and computational structures [18] are compatible with our model. The transport mechanism described here, is likely to exist in parallel with a previously proposed route based on diffusion and nuclear retention [2,5-7,9,10].

Chapter 2

34

REFERENCES

1. Wente SR, Rout MP: The Nuclear Pore Complex and Nuclear Transport. Cold Spring Harbor Perspectives in Biology 2010, 2:a000562–a000562.

2. Zuleger N, Korfali N, Schirmer EC: Inner nuclear membrane protein transport is mediated by multiple mechanisms. Biochem. Soc. Trans 2008, 36:1373.

3. Lusk CP, Blobel G, King MC: Highway to the inner nuclear membrane: rules for the road. Nat Rev Mol Cell Biol 2007, 8:414–420.


5. Malik P, Korfali N, Srsen V, Lazou V, Batrakou DG, Zuleger N, Kavanagh DM, Wilkie GS, Goldberg MW, Schirmer EC: Cell-specific and lamin-dependent targeting of novel transmembrane proteins in the nuclear envelope. Cell. Mol. Life Sci. 2010, 67:1353–1369.




9. Ellenberg J, Siggia ED, Moreira JE, Smith CL, Presley JF, Worman HJ, Lippincott-Schwartz J: Nuclear membrane dynamics and reassembly in living cells: targeting of an inner nuclear membrane protein in interphase and mitosis. The Journal of Cell Biology 1997, 138:1193–1206.

10. Ostlund C, Ellenberg J, Hallberg E, Lippincott-Schwartz J, Worman HJ: Intracellular trafficking of emerin, the Emery-Dreifuss muscular dystrophy protein. Journal of Cell Science 1999, 112 ( Pt 11):1709–1719.




14. Frey S, Richter RP, Görlich D: FG-rich repeats of nuclear pore proteins form a three-dimensional meshwork with hydrogel-like properties. Science 2006, 314:815–817.

15. Peters R: Translocation through the nuclear pore: Kaps pave the way. BioEssays 2009, 31:466–477.16. Rout MP, Aitchison JD, Suprapto A, Hjertaas K, Zhao Y, Chait BT: The yeast nuclear pore complex:

composition, architecture, and transport mechanism. The Journal of Cell Biology 2000, 148:635–651.

17. Strawn LA, Shen T, Shulga N, Goldfarb DS, Wente SR: Minimal nuclear pore complexes define FG repeat domains essential for transport. Nat Cell Biol 2004, 6:10–206.

18. Alber F, Dokudovskaya S, Veenhoff LM, Zhang W, Kipper J, Devos D, Suprapto A, Karni-Schmidt O, Williams R, Chait BT, et al.: The molecular architecture of the nuclear pore complex. Nature 2007, 450:695–701.

19. Schrader N, Stelter P, Flemming D, Kunze R, Hurt E, Vetter IR: Structural Basis of the Nic96 Subcomplex Organization in the Nuclear Pore Channel. Molecular Cell 2008, 29:46–55.


35

Chap

ter 2

20. Bailer SM, Balduf C, Hurt E: The Nsp1p Carboxy-Terminal Domain Is Organized into Functionally Distinct Coiled-Coil Regions Required for Assembly of Nucleoporin Subcomplexes and Nucleocytoplasmic Transport. Molecular and Cellular Biology 2001, 21:7944–7955.

21. Frenkiel-Krispin D, Maco B, Aebi U, Medalia O: Structural analysis of a metazoan nuclear pore complex reveals a fused concentric ring architecture. J. Mol. Biol. 2010, 395:578–586.

22. Yang Q, Rout MP, Akey CW: Three-dimensional architecture of the isolated yeast nuclear pore complex: functional and evolutionary implications. Molecular Cell 1998, 1:223–234.

Chapter 3Super-resolution optical microscopy of membrane

proteins trapped in the Nuclear Pore Complex

Justyna Laba, Charlotte Kaplan, Anton Steen, Thomas Ponath,

Bert Poolman, Helge Ewers, Liesbeth M. Veenhoff

Chapter 3

38

ABSTRACT

Nuclear Pore Complexes (NPCs), embedded in the double membrane Nuclear Envelope (NE), mediate the transport of molecules between the nucleus and the cytoplasm. Membrane proteins destined for the Inner Nuclear Membrane (INM) need to cross the NPC after their synthesis, but the mechanism of this transport process is still under debate. Yeast S. cerevisiae proteins Heh1p (Scr1p) and Heh2p are imported in an active pathway involving a high affinity NLS, karyopherins and a Ran gradient across the NE. In the previous chapter we have proposed a model of their transport, in which a long unfolded linker domain (L) slices through the pore structure in order to present the NLS to the transport factors in the NPC central channel. In this chapter we aim to show that the transmembrane domains of the previously used reporter proteins are membrane embedded during transit through the NPC. To do this we have combined the rapamycin-dependent complex formation between the FKBP-tagged reporters and FRB-tagged nucleoporins with a novel labeling technique for Stochastic Optical Reconstruction Microscopy (STORM) developed in Ewers lab. Although we were unable to resolve if transmembrane domains of the studied reporters remain in the membrane during transport, the presented work holds promise for future experiments.


39

Chap

ter 3

The Nuclear Envelope (NE) is a double membrane system that protects the nuclear content. Most of the transport events to and from the nucleus are mediated by Nuclear Pore Complexes (NPC). The NPCs of the yeast Saccharomyces cerevisiae are complexes of about 50 MDa built from 30 different proteins in multiple copies [1]. An approximately cylindrical scaffold structure is anchored in the NE at places where the inner and outer membrane of the nuclear envelope are connected. Proteins that are intrinsically disordered and have FG-rich repeats, the FG-Nups, extend from the scaffold into the centre of the pore. Transport of soluble cargo is relatively well understood. Import and export of a variety of macromolecules is facilitated by transport factors that recognize localization signals on the cargo molecules and facilitate transport across the NPC by binding to the FG-repeats. GTP-bound Ran in the nucleus and GDP-bound Ran in the cytoplasm coordinate the direction of transport. In the case of import, a nuclear localization signal (NLS) is recognized by import factors in the cytosol, and RanGTP stimulates the release of the cargo from the transport factors in the nucleus. The mechanism(s) used by membrane proteins are still elusive. Several models describing the process have been proposed (as reviewed in [2,3]). For many proteins passive diffusion and selective retention applies, but the S. cerevisiae proteins Heh1p (Scr1p) and Heh2p are imported by an active transport mechanism [4,5]. Using a set of reporter proteins, we have proposed a model for their transport [5]. A typical reporter is a truncated version of Heh2p, consisting of one transmembrane segment (TM), an unfolded linker domain (L) of approximately 200 amino acids, a high affinity Kap60-NLS (h2NLS) and GFP (G-NLS-L-TM). We showed that reporters with this domain composition are efficiently transported in a Kap60/95, FG-Nup and Ran-dependent fashion[5]. We also provided evidence for passage of the soluble domains through the centre of the NPC [5] by applying the Anchor Away system [6]. This system is based on rapamycin-induced complex formation between neighbouring proteins fused to either FRB or FKBP domains. Using this approach, we could trap the FKBP-tagged reporter to the FRB-tagged pore FG-Nup Nsp1 during transit through the NPC in living cells[5]. We proposed that active import of INM proteins involves stretching of the flexible linker from the nuclear envelope membrane to the central channel of the pore, where import factors can facilitate the transport by interaction with FG-Nups [5]. This novel mode of transport implies the existence of at least transient openings within the NPC structure that would accommodate the linker domain during transit of the reporter. Clearly, more insight in the NPC structure and dynamics is needed. Recent years brought enormous breakthroughs in the field of fluorescence microscopy. Development of super-resolution techniques (reviewed in [7]) that overcome the diffraction limit of optical microscopy is of huge relevance for cell biology. Due to diffraction limit the optical resolution in lateral direction is described by λ/2NA, where λ is the wavelength of light used for imaging and NA is the numerical aperture of the microscope [8]. Most cellular structures are much smaller than the diffraction

Chapter 3

40

limit. For example, in optical microscopy the inner and outer nuclear membranes are projected as a single line. Also the NPC, a pore with approximately 100nm outside diameter and approximately 40nm in the central channel [1], is visualized as a spot. NPCs are densely present in the NE, with roughly 5 NPCs/μm2 in mammalian cells [9] and 12 NPCs/μm2 in fungal cells [10]. So, the spots observed in optical microscopy may not even correspond to individual pores. Super-resolution optical microscopy, and especially techniques based on single fluorophore excitation like PALM [11] and STORM [12], can reduce the spatial resolution to 10-50 nanometer. Both techniques rely on temporal separation of fluorescent events of fluorophores, and subsequent mapping of the centre of the fluorescence peak for each recorded point and image reconstitution. Super resolution techniques were quickly implemented in research on the NPC. Single molecule tracking was successfully applied to study the trajectories of passive and facilitated diffusion through the pore [13]. Using STORM, Loschberger et al. showed first microscopic images of Xenopus leavis NPCs where 8-fold rotational symmetry of the pore was clearly observed [14]. Soon after that, STORM was applied to elucidate the orientation of the human Nup107-160 subcomplex within the human NPCs in living cells [15]. Super-resolution optical images of yeast NPCs are limited to few reports so far [16,17]. The aim of this chapter is visualization of the membrane protein reporters in transit through the NPC and test of our model of their transport. In particular, we want to validate if the membrane proteins are transmembrane embedded while going through the NPC. A direct way to test this is to determine the localisation of both termini of the protein with respect to the NPC. According to our model, the N-terminus -with the NLS- should be in the pore, and the C-terminus should be in the NE lumen. Alternatively, the reporter may not be transmembrane during transport, in which case both termini should be observable in the pore. Indeed, the F-G-NLS-L-TM has only one transmembrane segment located in the C-terminal part, therefore the possibility existed that it is inserted into the membrane environment after import to the nuclear interior. We have now combined rapamycin-induced complex formation between neighbouring proteins fused to either FRB or FKBP domains, with the novel labelling technique based on antiGFP nanobodies and SNAP tagging from the Ewers laboratory [16,17]. Here, we report on the design of strains and reporter constructs that allow dual-color STORM experiments for monitoring both ends of the reporter simultaneously, or for monitoring one end of the reporter protein together with an NPC component. Also different labeling regimes were tested. Using these approaches we visualized for the first time trapping of a membrane protein in the NPC in super-resolution. The two-colour STORM experiments were not yet successful, but hold promise to eventually test the existence of openings within the NPC structure that accommodate the linker domain during transit of the reporter.


41

Chap

ter 3

MATERIALS AND METHODS

Strains and plasmids

All strains and plasmids are listed in Tables 1 and 2 of the Appendix. All the experiments described here were performed in the S. cerevisiae K14708 strain [18] or strains derived from it. For the construction of the strain used in the NPC trapping experiments, the gene encoding nucleoporin Nsp1p was tagged with an FRB cassette from a pFA6a-FRB-KanMX6 plasmid [6]. For the construction of strains used for double-colour imaging, the genes encoding nucleoporins Nic96, Pom34 and Pom152 were tagged with a GFP cassette from a pYM25 plasmid [19]. All the INM reporters were expressed from pACM021-GFP [5]. The SNAP-tagged reporters, F-SN-NLS-L-TM and F-G-NLS-L-TM-SN, were designed based on the pJKL01-2×FKBP12-GFP-Lic-h2NLS-L-TM plasmid [5]. The SNAP-tag is a 20 kDa monomeric protein that binds covalently to substrates containing O6-benzylguanine; the protein is a mutant of the human DNA repair protein O6-alkylguanine-DNA alkyltransferase (hAGT).

Growth conditions

Yeast strains were grown to mid-exponential growth phase at 30°C in synthetic dropout medium without L-histidine. Strains were first grown for 1 day on medium containing 2% glucose (w/v) and then cultured on medium containing 2% raffinose (w/v). On the day of the experiment the expression of the reporters was induced with 0.5% galactose (w/v). For the trapping experiments rapamycin was used at a concentration of 10ug/ml.

Yeast preparation for labelling

The labelling protocol was adapted from [16,17]. 10ml of exponentially growing yeast culture was harvested by centrifugation. 50μl of the condensed yeast suspension was applied on concanavalin A-coated cover glass and allowed to settle for 15min. Subsequently, yeast cells were fixed by incubation in 4% paraformaldehyde, 2% sucrose in PBS (15 minutes, room temperature, gentle rotation). The reaction was quenched by 50mM NH4Cl in PBS (15 minutes, room temperature, gentle rotation). Cells were permeabilized with 0,25% Triton X-100 in PBS supplemented with 1%BSA, 0.004% NaN3 (5min, room temperature, gentle rotation). The permeabilization solution was washed out thoroughly with PBS supplemented with 1%BSA, 0.004% NaN3. Cells were blocked with PBS supplemented with 5%BSA, 0.004% NaN3, 0.01% Triton X-100 (30min, room temperature, gentle rotation). The cells were washed with PBS supplemented with 1%BSA, 0,004% NaN3 before applying nanobodies or SNAP tag solutions.

Chapter 3

42

Labelling with nanobodies

The protocol was adapted from [16,17]. The stock of AF647 (Invitrogen)-labelled nanobody (Chromatek) solution was diluted to 10μM with PBS containing 1% BSA, 0.004% NaN3 and 0.01% Triton X-100. Upon permeabilization of the cells and washing steps, 100 μl of nanobody solution was applied on the cover glass and incubated for 90min at room temperature. Subsequently cells were washed several times with PBS.

SNAP-tagging

The SNAP tagging protocol was adapted from [17]. Upon permeabilization and washing steps, cells were blocked with Image IT FX (Invitrogen). SNAP Surface Alexa647 (NEB) was diluted 1:800 in PBS containing 1% BSA, 0,004% NaN3, 50mM DTT and 0,01% Triton X-100. The reaction was incubated for 30 min at 37°C. Excess of the dye was washed out with PBS.

Double labelling

The protocol was adapted from [17]. In case of double labelling, SNAP tagging was performed prior to application of nanobodies. SNAP tagging was performed as described above. Subsequently, cells were washed with PBS. Cells were blocked again by incubation in PBS supplemented with 5%BSA, 0.004% NaN3 and 0.01% Triton X-100. The Alexa700-labelled nanobody solution was applied as described above. Cells were washed several times with PBS and fixed with 4% paraformaldehyde, 2% sucrose in PBS (10 minutes, room temperature, gentle rotation) and extensively washed again with PBS.

STORM imaging

Except for the tests of SNAP Cell 505 and SNAP Surface 594, all measurements were performed in the lab of Helge Ewers, ETH Zurich, on a custom-build microscope described in detail in [16]. Briefly, the microscope was equipped with: a 473nm laser (100mW, Pusch OptoTech) and a 647nm laser (150mW, Pusch OptoTech); 60x objective NA 1.49 (Olympus); a 700/75 and a 675/50 band pass emission filters (AHF). Images were acquired on EM-CCD camera (Evolve, Photometrics) with MicroManager software [20]. The samples were mounted in a custom-made holder. For image collection the chamber was filled with 150 mM TrisHCl pH 8 buffer containing 10% glucose, 35 mM cysteamine, 0.5 mg/mL glucose oxidase, and 40 mg/mL catalase [17].

Image reconstruction

Image reconstruction was performed in Matlab (The Mathworks, Inc) as described in [16].


43

Chap

ter 3

RESULTSVisualization of the FKBP-tagged reporter trapped in the NPC using super-resolution optical microscopy

Single labelling of the reporter was performed using two methods, namely with anti-GFP nanobodies that bind the GFP-tag, and with fluorescent O6-benzylguanine-containing substrates that bind the SNAP tag from NEB. Our Heh2p-based reporter is a membrane protein, therefore the membrane permeabilization protocol was modified, as described in Materials and Methods, in order to preserve the nuclear envelope structure.

F-G-NLS-L-TMwitout rapAF647 labelled nanobodies

F-SN-NLS-L-TMwithout rapSNAP Surface AF647

F-SN-NLS-L-TM+ rapSNAP Surface AF647

B

F-G-NLS-L-TM+ rapAF647 labelled nanobodies

A

FIG.1. REPORTERS TRAPPED AT THE NSP1FRB LOCALIZE IN RING SHAPED CLUSTERS.

Reconstructed STORM images of yeast cells expressing F-G-NLS-L-TM labeled with nanobodies coupled to AF647 (A) and F-SN-NLS-L-TM labeled with SNAP Surface AF647 (B). Insets show clear clustering of the recorded localizations in case of trapped reporter (+rap, left panels), compared with the free reporter (without rap, right panels). Observed ring shape of clusters is typical for nuclear pore localized proteins. Wide-field images on top show GFP fluorescence (A) or AlexaFluor647 fluorescence (B). Scale bars - 2 μm for wide-field images, 500nm for super resolved images. Reconstruction parameters: A. left panel: 59849 frames, 4431 localizations in the field of view, localization precision 16.1nm; right panel: 42697 frames, 3619 localizations in the filed of view, localization precision 16.8nm. B. left panel: 100000 frames, 2023 localizations in the field of view, localization precision 11.9nm; right panel: 53689 frames, 2487 localizations in the filed of view, localization precision 11.5nm.

Chapter 3

44

Reconstructed images of the reporter trapped on the Nsp1FRB show clear clustering of the fluorescent signal, compared to the images of the non-trapped reporter (fig.1). The best resolution is obtained with SNAP-tag labelling; here a localization precision of 11.9 nm was measured as compared to 16.1 nm for nanobody labelling. Some, but not all, of the observed clusters are ring-shaped; the 8-fold rotational symmetry is not observed. A number of examples of such clusters, which likely represent individual pores, are lined up in figure 2. The morphology of these clusters is in line with what can be expected, but provides no proof of them being NPCs. However, the fact that clustering is rapamycin dependent gives confidence that observed pattern arises from trapping at the nuclear pore.

Double-labelling strategies to localize the membrane protein N-terminus with its C-terminus or with Nups

Next, we experimented with labelling regimes that would allow simultaneous detection of both termini of the membrane protein, or co-detection of the membrane protein with an NPC component. As single labelling with nanobodies AF647 or SNAP Surface AF647 gave good results (figure 1), we attempted double labelling with antiGFP nanobodies AF700 and SNAP Surface AF647. In the experiment presented in figure 3A and B, AF700 nanobodies (green) were used to visualize two GFP tagged nucleoporins, Nic96 and Pom34, in conjunction with SNAP Surface 647 (red) labelled F-SN-NLS-L-TM reporter.

B. F-G-NLS-L-TM (AF647 labelled nanobodies)

A. F-SN-NLS-L-TM (SNAP Surface AF647)

FIG.2. EXAMPLE SELECTION OF THE IMAGES SHOWING HEH2P BASED REPORTERS TRAPPED AT THE NUCLEAR PORE.

Reconstructed STORM images of yeast cells expressing F-SN-NLS-L-TM labeled with SNAP Surface AF647 (A) and F-G-NLS-L-TM labeled with AF647 coupled nanobodies (B). Scale bars - 50nm. White circles indicate the circumference of the putative NPCs.


45

Chap

ter 3

In Figure 3C we labelled the N terminus of the F-G-NLS-L-TM-SN reporter with AF700 nanobodies (green) and the C-terminus with SNAP Surface647 (red). GFP wide-field images are shown on top of the super-resolved images for comparison. Overall the fluorescence signal of AF700 was very low, and, in addition, it was observed throughout the entire cell, even in the case of nucleoporins Nic96 and Pom34 that are known to localize in the nuclear pores only. We do observe co-localisation of both dies, but, because of the low AF700 signal and high background in this channel, we are not sure if the observed co-localization is incidental.

FIG.3. BACKGROUND FLUORESCENT IN THE AF700 CHANNEL HINDERS THE ANALYSIS OF DOUBLE LABELING EXPERIMENTS.

A and B. Reconstructed STORM images of yeast cells expressing F-SN-NLS-L-TM reporter and GFP tagged nucleoporins: Nic96p (A) or Pom34p (B). Reporter was labeled with SNAP Surface AF647 (red), GFP tagged nucleoporins were labeled with nanobodies AF700 (green). Background fluorescence of AF700 channel is observed throughout the entire cell, although Nic96 and Pom34 are localized exclusively to nuclear pores. Partial co-localization of the signals at the nuclear periphery is visible. C. Reconstructed STORM image of yeast cell expressing F-G-NLS-L-TM-SN reporter trapped at Nsp1FRB. Reporter was labeled with SNAP Surface AF647 (red) and nanobodies AF700 (green). Fluorescence of the AF700 channel is observed throughout the entire cell again. Additionally, poor signal of the AF647 channel is probably caused by the use of the blinking buffer preferable for AF700. Wide-field images on top show GFP fluorescence. Scale bars - 2 μm for wide-field images, 500nm for super resolved images. Reconstruction parameters: A. top panel: 46831 frames, 3728 localizations in the field of view, localization precision 13.7nm; bottom panel: 46831 frames, 3836 localizations in the filed of view, localization precision 13.7nm. B. 30057 frames, 8032 localizations in the field of view, localization precision 24.4nm. C. 37448 frames, 3409 localizations in the filed of view, localization precision 18.5nm.

A. Nic96GFP (AF700 labelled nanobodies, green) + F-SN-NLS-L-TM (SNAP Surface AF647, red)

B. Pom34GFP (AF700 labelled nanobodies, green) + F-SN-NLS-L-TM (SNAP Surface AF647, red)

C. F-G-NLS-L-TM-SN (AF700 labelled nanobodies, green; SNAP Surface AF647, red)

Chapter 3

46

In attempts to optimize the dual labelling, different modifications to the labelling protocol were tested, namely, a range of nanobody concentrations, different blocking agents (BSA, ImageIT Invitrogen, horse serum), and exchange of cysteamine to β-merkapthoethanol in the measurement buffer. In addition, we tested differently labelled SNAP substrates, namely SNAP Cell 505 and SNAP Surface 594. Unfortunately, none of these modifications have improved the dual-colour visualization.

DISCUSSION

We asked if membrane proteins are transmembrane embedded while going through the NPC, and used 2-colour high-resolution STORM as a direct way to test this, by determining the localisation of both termini of the protein with respect to the NPC. As double-labelling experiments did not work in the available time frame, we cannot answer this question. However, we have achieved STORM visualization of the membrane protein caught in the NPC. It is the first report of a super-resolved image of cargo trapped in the nuclear pore. When Heh2p-based reporters are trapped on Nsp1FRB clear clustering of the recorded localizations is visible. Some of the clusters form rings. Initial calculation of the rings diameter of 51 potential pores from 7 cells gave a mean value of 46nm +/- 12. For the calculation the pore images were exported from Matlab and the fluorescence intensity was analysed in ImageJ. If rings were not perfectly circular, the smallest diameter was calculated, so in fact the presented average value might be lower than the real diameter of the pore. Yeast NPCs are 40nm high and 100nm in diameter, with the central channel diameter being approximately 40nm [1]. The rapamycin dependent reporter clustering and the shape and size of the clusters is indicative for nuclear pore complex localization. Additionally the very high localization precision of the dataset gives confidence in the observed pattern. Our ring-shaped clusters do not have the 8-fold rotational symmetry, observed before in case of Xenopus leavis nuclear pores [14]. The reasons for the superior images from Xenopus pores may be that they are bigger than yeast (130nm outside diameter [21] compared to 100nm in yeast [1]). More important, the experiments presented in [14] were performed on isolated nuclear envelopes that were spread flatly on the cover glass, so that many pores in the field of view are in one plane. In our experimental setup, where we aimed to visualize the NPC trapped reporter in the whole cell environment, this was much more difficult. Yeast S. cerevisiae cells are roughly spherical, with the position of the spherical nuclei dependent on the cell cycle stage. In order to see NPCs as circular rings, we had to reduce the analysis to apical sections only. We have no control over the orientation of the cells on the cover glass, and yeast cells were rarely positioned in a single plane, and hence the upper surface of the nuclei will not be in the same plane for all the cells. For the measurements the yeast cells were attached to


47

Chap

ter 3

the cover glass via a layer of concanavalin A, which caused a non-planar distribution of yeast cells because of uneven thickness of the concanavalin A layer. Szymborska et. al have overcome the disadvantages of in vivo imaging for the epithelial U2OS cells [15], while using a nanobodies-labelling scheme similar to ours. Although the 8-fold rotational symmetry of the pore is rarely visible also in their case, the reconstructed images presented there are of higher quality than ours, with clearer rings and less background. This difference is attributed to the cell architecture on the one hand and to technical difficulties with labelling yeast specifically on the other. Firstly, the human cell nucleus is much bigger than yeast, with the volume of S. cerevisiae nuclei around 2.9 μm3 [22] compared to 1000-2000 μm3 in case of HeLa cells [23]. Epithelial cells are usually distributed in a single layer on the glass and display an apical-basal polarity, with the nuclei positioned closer to the basolateral membrane. Owing to these characteristics Szymborska et al. were able to visualize hundreds of pores per nucleus while focusing the microscope on the lower nuclear surface. Secondly, our experience with searching for a second well performing fluorescent label showed that non-specific binding of the fluorescent dye to cellular structures is more troublesome in yeast than in mammalian cells. For example: AF700 nanobodies were successfully used for STORM in the Ptk2 cell line, as well as on the S. cerevisiae cell wall [16]; SNAP Cell 505 on the other hand performs well in HeLa and COS7 cells [24]. In our hands non-specific binding of both dyes severely hindered the experiments. In case of AF647 the same problem was less pronounced – we collected good quality data, but also here the non-specific binding of the dye is visible. The presented dataset is the best/only visualization of nuclear pore-localized molecules in yeast S. cerevisiae to date, but, with all the shortcomings in the procedure we were unable to resolve if the membrane proteins remain membrane embedded during transport through the NPC. Future studies using single-particle tracking in metazoan systems may more readily resolve the transport route through the NPC.

ACKNOWLEDGEMENTS

Work described in this chapter was funded by the FEBS Short Term Fellowship to JKL.

Chapter 3

48

REFERENCES


2. Burns LT, Wente SR: Trafficking to uncharted territory of the nuclear envelope. Current Opinion in Cell Biology 2012, 24:341–349.

3. Laba JK, Steen A, Veenhoff LM: Traffic to the inner membrane of the nuclear envelope. Current Opinion in Cell Biology 2014, 28:36–45.




7. Huang B, Bates M, Zhuang X: Super-Resolution Fluorescence Microscopy. Annu. Rev. Biochem. 2009, 78:993–1016.

8. Abbe E: Beiträge zur Theorie des Mikroskops und der mikroskopischen Wahrnehmung. Archiv f. mikrosk. Anatomie 1873, 9:413–418.

9. Kubitscheck U, Wedekind P, Zeidler O, Grote M, Peters R: Single nuclear pores visualized by confocal microscopy and image processing. Biophysj 1996, 70:2067–2077.

10. Steinberg G, Schuster M, Theisen U, Kilaru S, Forge A, Martin-Urdiroz M: Motor-driven motility of fungal nuclear pores organizes chromosomes and fosters nucleocytoplasmic transport. The Journal of Cell Biology 2012, 198:343–355.

11. Betzig E, Patterson GH, Sougrat R, Lindwasser OW, Olenych S, Bonifacino JS, Davidson MW, Lippincott-Schwartz J, Hess HF: Imaging Intracellular Fluorescent Proteins at Nanometer Resolution. Science 2006, 313:1642–1645.

12. Rust MJ, Bates M, Zhuang X: Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nature Methods 2006, 3:793–796.

13. Ma J, Goryaynov A, Sarma A, Yang W: Self-regulated viscous channel in the nuclear pore complex. Proceedings of the National Academy of Sciences 2012, 109:7326–7331.

14. Loschberger A, van de Linde S, Dabauvalle MC, Rieger B, Heilemann M, Krohne G, Sauer M: Super-resolution imaging visualizes the eightfold symmetry of gp210 proteins around the nuclear pore complex and resolves the central channel with nanometer resolution. Journal of Cell Science 2012, 125:570–575.

15. Szymborska A, de Marco A, Daigle N, Cordes VC, Briggs JAG, Ellenberg J: Nuclear Pore Scaffold Structure Analyzed by Super-Resolution Microscopy and Particle Averaging. Science 2013, 341:655–658.

16. Ries J, Kaplan C, Platonova E, Eghlidi H, Ewers H: A simple, versatile method for GFP-based super-resolution microscopy via nanobodies [Internet]. Nature Methods 2012, 9:582–584.

17. Mund M, Kaplan C, Ries J: Localization microscopy in yeast. Methods in Cell Biology; 2014.18. Gruber S, Arumugam P, Katou Y, Kuglitsch D, Helmhart W, Shirahige K, Nasmyth K: Evidence

that Loading of Cohesin Onto Chromosomes Involves Opening of Its SMC Hinge. Cell 2006, 127:523–537.


49

Chap

ter 3

19. Janke C, Magiera MM, Rathfelder N, Taxis C, Reber S, Maekawa H, Moreno-Borchart A, Doenges G, Schwob E, Schiebel E, et al.: A versatile toolbox for PCR-based tagging of yeast genes: new fluorescent proteins, more markers and promoter substitution cassettes. Yeast 2004, 21:947–962.

20. Edelstein A, Amodaj N, Hoover K, Vale R, Stuurman N: Computer control of microscopes using µManager. Curr Protoc Mol Biol 2010, Chapter 14:Unit14.20.


22. Jorgensen P, Edgington NP, Schneider BL, Rupes I, Tyers M, Futcher B: The size of the nucleus increases as yeast cells grow. Molecular Biology of the Cell 2007, 18:3523–3532.

23. Maeshima K, Iino H, Hihara S, Funakoshi T, Watanabe A, Nishimura M, Nakatomi R, Yahata K, Imamoto F, Hashikawa T, et al.: Nuclear pore formation but not nuclear growth is governed by cyclin-dependent kinases (Cdks) during interphase. Nat Struct Mol Biol 2010, 17:1065–1071.

24. Klein T, Löschberger A, Proppert S, Wolter S, van de Linde S, Sauer M: Live cell dSTORM with SNAP-tag fussion proteins. Nature Methods 2011, 8:7–9.

Chapter 4Nuclear import of membrane proteins revisited:

Active nuclear import of Heh2-derived proteins occurs while membrane embedded,

the path through the NPC is unresolved

Justyna K. Laba, Anton Steen, Petra Popken,

Alina Chernova, Bert Poolman, Liesbeth M. Veenhoff

Published in Cells 2015

Chapter 4

52

ABSTRACT

It is poorly understood how membrane proteins destined for the inner nuclear membrane pass the crowded environment of the Nuclear Pore Complex (NPC). Nuclear localization of the Saccharomyces cerevisiae proteins Src1/Heh1 and Heh2 is dependent on an active transport mechanism, requiring a sorting signal that comprises a nuclear localization signal and an intrinsically disordered linker. A transport mechanisms was proposed where the transmembrane domains diffuse through the membrane while the extralumenal domains accompanied by transport factors travel through the NPC. Here, we validate the proposed mechanism. First, we present biochemical and localization data to support that the proteins are membrane-embedded irrespective of the presence of the sorting signal, independent of the specific transmembrane domain (multipass or tail anchored), independent of GET, and also under conditions that the proteins are trapped in the NPC. This confirms the transmembrane domain is in the membrane while transported, and not inserted post import. Second, using the recently established size limit for passive diffusion of membrane proteins in yeast, and an improved assay, we confirm import of polytopic membrane protein with extralumenal soluble domains larger than those that pass passively. This confirms that distinct routes through the NPC are used for active and passive transport. Thirdly, we revisit the proposed route through the NPC and conclude it is unresolved but may be closer to the membrane than we suggested previously. Altogether these observations confirm active, transport factor dependent, nuclear transport of membrane-embedded mono- and polytopic membrane proteins in baker’s yeast via a largely unresolved route.

Nuclear import of membrane proteins revisited

53

Chap

ter 4

The nuclear envelope (NE), although physically connected to the endoplasmic reticulum (ER), has features and functions that are distinct from the ER membrane system (as reviewed in [1-3]. It consists of two membranes: the inner and outer nuclear membrane (INM and ONM, respectively). The NE is perforated by Nuclear Pore Complexes (NPCs); the number of pores in Saccharomyces cerevisiae is estimated at 65-180, depending on the cell cycle stage [4]. In the pores the INM and ONM are connected via the highly curved pore membrane. The ONM is more similar in composition to the ER than to its nuclear interior-facing counterpart. No diffusion constraints in places where the ER meets the ONM have been discovered so far, but specific proteins such as the KASH-domain (for Klarsicht, ANC-1, Syne Homology) proteins enrich at the ONM through interaction with NE resident proteins (as reviewed in [5]. To date there are few proteins identified as stable components of the INM and their targeting mechanism is still under debate. The process of how any membrane protein reaches its final destination is more complicated than for soluble proteins as an additional step in the biogenesis is required, namely insertion into the membrane. For years, the Sec61 system in the ER was the only characterized membrane insertion/translocation machinery (reviewed in [6]), but more recently other systems have been described [7]. Most notably a specialized insertion system for tail-anchored proteins, called GET (Guided Entry of Tail anchored proteins), was characterized (reviewed in [8]). The function of others such as Ssh1, the non-essential Sec61 homologue in yeast [7], is still elusive. It is thus currently difficult to be certain how a specific membrane protein is inserted to the membrane and also how much redundancy exists between different systems. Most significant in this context is the question if sorting to the nucleus can precede insertion into the lipid bilayer. While the transport of membrane proteins to the nuclear interior is still under debate, the rules for soluble traffic across the NPC are well understood (reviewed in [9-11]). Soluble proteins can cross the pore passively, with the diffusion rate dependent on protein size and surface properties. Proteins rich in phenylalanine and glycine residues, the FG-Nups, fill the central channel of the NPC and establish the permeability barrier. Active transport results in fast accumulation, also of large macromolecular complexes. It involves interaction of the cargo with transport factors and requires the small GTPase Ran and factors that regulate its nucleotide-bound state to orchestrate entry and release from the NPC FG-Nups. One of the biggest questions in membrane protein nuclear traffic concerns its energy dependence. A model, in which INM proteins passively diffuse through the pore and are retained in the nucleus due to interactions with DNA and lamins, is mostly supported by the data from higher eukaryotes [12-14]. Here, the NPC acts as a selective gate only to prevent passage of membrane proteins or protein complexes with large extralumenal domains [12,14]. On the other hand, there is evidence from the yeast S. cerevisiae that, in addition to selective retention, an energy dependent route

Chapter 4

54

exists for specific proteins [15-17]. With respect to membrane protein trafficking in yeast, NPCs can therefore be viewed as passive barriers for diffusion and as active pumps, facilitating enrichment of certain proteins in the nuclear interior. The NPCs architecture is conserved between species, but the biology of the nuclear envelope is distinct as yeast undergoes a closed mitosis omitting the possibility of recruitment of INM proteins into the reassembling NE after open mitosis. Altogether, we deem it likely that both processes, diffusion/retention and active import, are involved in importing membrane proteins to the nucleus in all species. Indeed, previous work reports on multiple overlapping signals and mechanisms in traffic of INM proteins [18-20]. In yeast, we showed that the nuclear localization signal (NLS) of Heh2, first described in [16], together with an intrinsically disordered (ID) linker (L) was required and sufficient for INM accumulation of a transmembrane protein [15]. The accumulation was dependent on the transport factor Kap95 (importin-beta), a functional Ran-gradient [15,16], and specific FG-Nups [15]. Our interpretation of the data was that the NLS and ID linker act in facilitating active transport through the NPC of a transmembrane protein. To exclude that membrane insertion happens only post nuclear import and to dissect if the sorting signal (NLS-L) may act earlier in the biogenesis, we here designed experiments to follow the fate of INM targeted reporter proteins from biogenesis to their localization in the nucleus. Altogether, we show the transport route through the NPC may be close to the membrane, different from what we suggested previously, and we confirm active - transport factor and NLS-L dependent - nuclear transport of membrane-embedded mono- and polytopic membrane proteins. MATERIALS AND METHODS


All strains and plasmids are listed in Tables 1 and 2 of the Appendix. Besides the GET deletion mutants, all the experiments described were performed in the S. cerevisiae K14708 strain (w303, matα tor1-1 fpr1::NAT) [21] or strains derived from it. For the construction of strains used in the NPC trapping experiments, the genes encoding nucleoporins Nup170, Nup53 and Nup59 were tagged with an FRB (FKBP12-rapamycin binding) cassette from a pFA6a-FRB-KanMX6 plasmid [22]. All the INM reporters were expressed from pACM021-GFP plasmid [15]. All the FKBP (FK506 binding protein) tagged reporters were designed based on the pJKL01-2×FKBP12-GFP-Lic-h2NLS-L-TM plasmid [15]. The plasmids encoding the Sec61-based reporter proteins with a variable number of maltose-binding proteins (MBP) in the extralumenal domain were based on the plasmids described in [17]. In these constructs, the region encoding the transmembrane helix of Heh2 was replaced by the transmembrane domain of Sec61, using homologous recombination.


55

Chap

ter 4

Growth conditions

Yeast strains were grown to mid exponential growth phase at 30°C in synthetic dropout medium without L-histidine. Strains were first grown for 1 day on medium containing 2% glucose (w/v) and then cultured on medium containing 2% raffinose (w/v). On the day of the experiment the expression of the reporters was induced with 0,5% galactose (w/v). For the trapping experiments rapamycin was used at a concentration of 10ug/ml.

Fluorescence microscopy

The images from fig. 1, 3 and 5 were collected with an LSM 710 confocal microscope (CarlZeiss MicroImaging, Jena, Germany) using an objective C-Apochromat 40×/1.2NA, a solid-state laser (488 nm) for excitation and ZEN2010B software. The images from fig. 2 and 4 were collected with a Delta Vision Microscope (Applied Precision), using InsightSSITM Solid State Illumination at 488 nm and an Olympus UPLS Apo 100x oil objective with 1.4 NA. Detection was done with a coolSNAP HQ2 camera. Image stacks (20 stacks of 0.2 um) were deconvolved using standard settings.

Analysis of the standard deviation of fluorescence intensity (SD fraction quantification)

Analysis of the fluorescent images was performed using ZEN2010B software and Fiji [23]. Line-scans of the fluorescence intensity over NE were collected for at least 30 cells for each strain in each condition. For each cell two values were calculated: average fluorescence intensity in the NE and the standard deviation of it. Standard deviation values for each cell were divided by the average fluorescence intensity values, and the mean value over all cells (later in the chapter referred to as SD fraction) was calculated from the resulting ratios. Differences in the data were considered to be significant with a p value less than 0.05 using a student’s t- test.

Analysis of NE/ER ratios

Analysis was done as reported previously [17].

Protein extraction and effects of salts and detergents

Whole cell extracts were obtained from 10ml cultures of exponentially growing yeast cells by lysis in NaOH and β-mercaptoethanol and TCA- precipitation. For preparation of crude membrane fractions, 200ml of exponentially growing yeast cells were harvested by centrifugation (5500xg 10 min). The pellet was washed with cold demi water and centrifuged again. Subsequently, the yeast pellet was resuspended in 10ml of 20mM TrisHCl pH 8.0, 0.01M EDTA with 35ul of β-mercapthoethanol and incubated for 15 min on ice with occasional swirling. The suspension was centrifuged (rotor type: Thermo scientific SL16R 75003629, 10000 rpm, 5 min), washed with buffer containing 1.1M sorbitol, 50mM TrisHCl pH 7,5, 10mM MgCl2, 3mM DTT and

Chapter 4

56

centrifuged again. The pellet was resuspended in the sorbitol buffer again to achieve 1.5x109 cells per 1ml. Zymolyase was added to the final concentration of 0,02%. Cells were incubated at 300C for 1 hour. The spheroplasts were pelleted (4000xg 10min), carefully washed with sorbitol buffer containing 1mM PMSF, 2ug/ml of pepstatin and protein inhibitor cocktail and centrifuged again. From this step the membrane isolation and solubilization of proteins with 1% Triton X100 or 1M NaCl were performed as described in [24]. In short, the spheroplasts were resuspended in 3ml of 20mM TrisHCl pH 8,0 150mM NaCl and lysed in a 5ml Potter tube. The cell debris was removed by short centrifugation (2min, 2000rpm). The supernatant was centrifuged again (20min, 16200xg) and the pellet containing membrane fraction resuspended in 500ul 20mM Tris HCl containing either 1% Triton X100 or 1M NaCl or 150mM NaCl. After 30 minutes incubation on ice the samples were centrifuged (100.000xg, 1 hour). The supernatants (solubilized fraction) and pellets (non-solubilized fraction) were mixed with SDS-PAA loading buffer, incubated at 70°C for 10 minutes and loaded on SDS PAGE gels.

Western Blotting

Samples were separated by SDS-PAA gel electrophoresis and transferred to a PVDF membrane (wet transfer, 16V, 17 hours). For detection of the reporters anti-FKBP antibody (Abcam, Cambridge, UK) was used at a dilution of 1:1000 (v/v) and a secondary anti-rabbit-alkaline phosphatase conjugate at 1:20,000 (v/v) (Sigma-Aldrich, St. Louis, USA). For the detection of Nup170FRB anti-FRB antibody was used at 1:1000 (v/v) (Enzo Life Sciences, Farmingdale, New York, USA) and again the anti-rabbit-alkaline phosphatase conjugate. For the detection of yeast Nsp1 anti-Nsp1 antibody was used at 1:1000 (v/v) (Abnova, Taipei, Taiwan) with anti-mouse-alkaline phosphatase conjugate at 1:20,000 (Sigma-Aldrich, St.Louis, USA). For the detection of Pma1 anti-Pma1 antibody was used at 1:5000 (v/v) (Abcam) with the anti-mouse-alkaline phosphatase conjugate. For the detection of Dpm1 anti-Dpm1 antibody was used at 1:250 (v/v) (Abcam) with the anti-mouse-alkaline phosphatase conjugate. Pre-stained molecular weight marker (ThermoScientific) was used.

RESULTS

Heh-2 derived reporter proteins for studying import

In our previous work [15] we have defined the minimal features of the yeast S. cerevisiae INM protein Heh2 that govern its accumulation in the INM. In this process Heh2 was truncated, removing the extralumenal LEM and MAN domains as well as its second transmembrane helix (TM) and lumenal domain, thereby changing it in a monotopic membrane protein. The LEM and MAN domains could cause retention in the nucleus. The resulting reporter is composed of GFP and Heh293-378 encoding the NLS, the ID linker and the first transmembrane helix, and is named G-NLS-L-TM


57

Chap

ter 4

(Fig. 1A). We have previously shown that this reporter still accumulates in the INM, despite the removal of over a half of the molecular weight of native Heh2; the presence of the NLS and linker regions is required and sufficient for INM accumulation. We also showed that the accumulation is reversible upon inhibition of active transport by depletion of cytosolic Kap95 [15], consistent with an active import mechanism and counter arguing a mechanism depending on nuclear retention. In the here presented experiments, we use the above-mentioned reporter and one with an N-terminal FKBP tag, F-G-NLS-L-TM. Addition of the FKBP tag has a dual role - it is used in trapping experiments described later, but it also reduces the expression level of the entire protein and the NE deformation caused by high expression of INM proteins. Analysis of the microscopy data of the reporters (Fig. 1B) reinstated the necessity of both the NLS and the linker region for targeting to the NE as removal of the NLS (F-G-ΔNLS-L-TM) or the linker domain (F-G-NLS-ΔL-TM) abolishes NE accumulation: reporters are evenly distributed between the NE and ER. Higher expression levels of G-NLS-L-TM, or the reporters with C-terminal FKBP-tag (G-NLS-L-TM-F) or N-terminal Protein A tag (PrA-F-G-NLS-L-TM) do not interfere with accumulation at the NE, but cells display deformation of the NE. Comparison of the expression levels of all the reporters is presented in Fig. 1C. They range from very low in case of F-G-NLS-L-TM-SN (SN for SNAP tag), to very high in case of G-NLS-L-TM-F and PrA-F-G-NLS-L-TM.

Heh-2 derived reporters are integral membrane proteins of the NE-ER

Next we performed biochemical fractionation studies to confirm the above reporters are membrane embedded. We measured the steady state membrane integration of the reporters by salt and detergent extraction (as published in [24]). In this method a crude membrane fraction is isolated from exponentially growing yeast and subsequently incubated with buffers containing a high concentration of salt or detergent. After ultracentrifugation the soluble and pellet fraction are analyzed. Proteins that solubilize with salt are only peripherally associated with the membrane, such as the control Nsp1. Transmembrane proteins require detergents to be extracted from the membrane, such as the ER protein Dpm1 with one transmembrane helix and the plasma membrane protein Pma1 with 10 transmembrane helices. As shown on Western blots in Fig. 1D, the integral membrane proteins solubilize only (Pma1) or predominantly (Dpm1) in 1% Triton, while Nsp1 is also found in the soluble fractions when membranes were incubated with the control buffer containing 150 mM NaCl or the buffer with 1M NaCl. The F-G-NLS-L-TM reporter behaved similarly to the controls, Dpm1 and Pma1, and solubilized in buffer with 1% Triton only. We next checked if deletion of the NLS (F-G-ΔNLS-L-TM), deletion of the linker region (F-G-NLS-ΔL-TM) or expanding the size of the soluble domains (from the cytoplasmic side PrA-F-G-NLS-L-TM and from

Chapter 4

58

B) F-G-NLS-L-TMG-NLS-L-TM

F-G-ΔNLS-L-TM F-G-NLS-ΔL-TM

F-G-NLS-L-TM-SN G-NLS-L-TM-F

PrA-F-G-NLS-L-TM

A)

control 1M NaCL 1%TritoncytoplasmP S S SP P

Nsp1

Pma1

Dpm1

F-G-NLS-L-TM

F-G-NLS-ΔL-TM

G-NLS-L-TM-F

F-G-NLS-L-TM-SN

PrA-F-G-NLS-L-TM

F-G-ΔNLS-L-TM

control 1M NaCL 1%TritoncytoplasmP S S SP P S

control 1M NaCL 1%TritoncytoplasmP S S SP P

130

100

70

F-G

-NLS

-L-T

M

F-G

-ΔN

LS-L

-TM

F-G

-NLS

-ΔL-

TM

F-G

-NLS

-L-T

M-S

N

G-N

LS-L

-TM

-F

PrA

-F-G

-NLS

-L-T

M

F-S

N-N

LS-L

-TM

C)

D)

FIG.1. HEH2-DERIVED REPORTER PROTEINS LOCALIZING TO THE NE-ER NETWORK ARE MEMBRANE EMBEDDED.

A. Cartoons showing domain composition of native Heh2 and derived reporters. F: FKBP; G: GFP; SN:SNAP; PrA: ProteinA. B. Confocal fluorescence microscopy images showing K14708 cells expressing the indicated Heh2-derived reporters. Scale bars: 2 μm. C. Western blot (anti-FKBP) of whole cells extracts of cells expressing the indicated reporters; equal protein amounts were loaded. D. Western Blots showing the results from salt and detergent extraction of crude yeast membranes fractions. Crude membranes are incubated in control buffer (20mMTris with 150mM NaCl), or in buffer with 1M NaCl or 1%Triton and ultracentrifuged as described in Materials and Methods. The pellet (P) and supernatant (S) fractions were loaded onto the gel. The cytoplasm fraction represents the proteins present in the supernatant of the lysate after the first centrifugation step.

the lumenal side F-G-NLS-L-TM-SN or G-NLS-L-TM-F) changes the sollubilization pattern of these reporters. All of them solubilized predominantly or exclusively when treated with 1% Triton (Fig. 1D). On the blots with G-NLS-L-TM-F and PrA-F-G-NLS-L-TM, which are the higher expressed proteins, a fraction of the protein appears in the soluble fraction after incubation with the control buffer or with 1M NaCl, similar to Dpm1. This suggests that a fraction of the proteins is not well inserted. Altogether, the biochemical fractionation studies on whole cell lysates and the in vivo localization studies confirm that the majority of Heh2-derived reporter proteins are membrane embedded, and that the linker region and NLS are not critical for membrane insertion.


59

Chap

ter 4

Evidence against membrane insertion post nuclear-import

The biochemical assay presented above does not exclude if temporarily, for example during the transport via the NPC channel, a fraction of the proteins is not embedded in the membrane. This could be particularly relevant for proteins that are inserted in the membrane post-translationally. Many of our reporters have C-terminal transmembrane helices, and could thus classify as tail-anchored proteins, which are inserted into the membrane environment posttranslationally, via the GET pathway [8]. The possibility of insertion to the inner nuclear membrane post nuclear import has thus far not been tested. We first aimed to resolve the uncertainty if our reporters depend on the GET system, and tested their localization in a series of GET deletion mutants. The microscopy images and the membrane extractions with salt and detergent presented in Fig. 2 clearly show that the G-NLS-L-TM reporter and derivatives were NE localized and membrane embedded, also when the GET system was nonfunctional. The expression level of G-NLS-L-TM is similar to that of G-NLS-L-TM-F and in both we see a low level of salt soluble membrane protein. As we see no difference between the wild type and the GET mutant we conclude that the insertion of these reporters is not dependent on GET, which would point to either Sec61-dependent insertion or another still unidentified system.

WT G-NLS-L-TM GET3Δ G-NLS-L-TMA)

B) control 1M NaCL 1%TritoncytoplasmP S S SP P

WT G-NLS-L-TM

GET3Δ G-NLS-L-TM

FIG.2. MEMBRANE INSERTION OF THE G-NLS-L-TM REPORTER IS GET-INDEPENDENT.

A. Fluorescence microscopy of wild type and Get3Δ mutant yeasts expressing G-NLS-L-TM. Scale bars: 2 μm B. Western Blots (anti-GFP) showing the results form salt and detergent extraction assay on WT and Get3Δ mutant expressing G-NLS-L-TM. Crude membrane preparations were treated as described in Material and Methods and in the legend to figure 1. P – pellet, S- supernatant.

Chapter 4

60

Although chaperoned soluble transport is less likely for reporters co-translationally inserted by Sec61, instead of post-translationally via the GET system, we aimed to further experimentally test this possibility for two reasons. Firstly, since the h2NLS-linker motif is a potent signal for sorting to the nuclear envelope outcompeting other classical NLSs [15,17], a competition between the insertion and nuclear transport

FIG.3. HEH2-DERIVED REPORTERS ARE MEMBRANE-EMBEDDED WHILE TRANSITING THE NPC.

A. Confocal fluorescence microscopy images of cells expressing Nup170FRB and F-G-NLS-L-TM reporter after incubation with rapamycin (+rap 10μg/ml) and in control conditions (+DMSO). The fluorescence patterns change from continues to punctate upon rapamycin treatment. Scale bars: 2 μm. B. Line-scans of the fluorescence intensity in the NE of the representative cells expressing F-G-NLS-L-TM after incubation with rapamycin (black line, +rap) or in control conditions (grey line, no rap). Top panel: control K14708 strain. Bottom panel: Nup170FRB strain. C. Comparison of the average SD fraction for each reporter in the Nup170FRB strain (n=37 cells for both conditions) and in the wild type (K14708, n=7 cells for no rap, n=55 cells for + rap); standard deviation is indicated. The SD fraction is calculated from the standard deviation of the fluorescence intensity along the NE in a cell divided by the average fluorescence intensity at the NE in that cell D. Western Blots showing the results from salt and detergent extraction of crude yeast membranes fractions of cells expressing Nup170FRB and F-G-NLS-L-TM. Membrane extractions were performed in trapped (+rap) and in non-trapped (+DMSO) conditions. F-G-NLS-L-TM solubilizes with the buffer with 1% Triton both in trapped and non-trapped conditions. Nup170FRB is salt-soluble before trapping and becomes salt-resistant upon trapping as the bands of solubilized fractions treated with control buffer and 1M NaCl disappear.


61

Chap

ter 4

machineries might take place. Secondly, one report indicates the presence of the Sec61 machinery on the inner side of the nuclear envelope [25] in yeast. We thus developed a system to monitor if membrane proteins were membrane embedded when trapped in the NPC. Our trapping experiments are based on the Anchor Away system [22], which utilizes rapamycin-dependent interaction between FRB and FKBP molecules. Previously Nsp1 had been used as the anchor [15]. However as recent publications show a non-NPC pool of Nsp1 [26,27] in our new set up, FRB was fused to one of the inner ring scaffold nucleoporins, Nup170. We show rapamycin-dependent trapping of the reporter with N-terminal FKBP tag, F-G-NLS-L-TM, at Nup170 (Fig. 3A). After cells were exposed to rapamycin, the fluorescence signal in the NE was observed in an exclusive punctate pattern, similar to what is seen with fluorescently labeled nucleoporins. The change in fluorescence pattern is also apparent when measuring the fluorescence intensity in the NE (Fig. 3B), and calculating the SD fraction (as described in the Experimental procedures). The average SD fraction over multiple cells is indicated in Fig. 3C. The average SD fraction increases significantly in conditions with rapamycin in the strain expressing Nup170FRB but not in the background strain (K14708) without an FRB anchor. Therefore, we conclude that a significant fraction (or possibly all) of the expressed reporter molecules were trapped at Nup170. Using biochemical fractionation methods we now address if the reporter molecules are membrane embedded when trapped. If the reporter protein passes the NPC as a soluble protein, one expects a change in the membrane extraction characteristics (Fig. 3D). When trapped at Nup170, the reporter is detergent-soluble and thus a true transmembrane protein after entry in the pore. Consistently, the tight interaction with the reporter causes a change in Nup170 solubility: in non-trapped conditions Nup170 is partially extracted with salt as expected for a non-transmembrane protein, while salt-extraction is clearly prevented after rapamycin-induced binding to the reporter. Altogether the experiments in figures 1,2,3 show the transmembrane nature of our INM reporter proteins, also when trapped in the NPC. We find no support for a mechanism where membrane insertion occurs post nuclear import.

Active transport breaks size restrictions for passive leak through the NPC

Having confirmed that the transmembrane domain is in the membrane during sorting we revisit our claim of the extralumenal domains trafficking through the central channel of the NPC. One line of evidence that we presented was that large extralumenal domains were imported; larger than what was estimated to pass the lateral channels passively. We here revisit this evidence. Firstly, the assays are now performed with reporter proteins with multi-pass transmembrane domain to validate that our previous studies were not biased from the use of monotopic membrane proteins. Secondly, we mapped the size limitations for passive entry of membrane proteins more precisely and

Chapter 4

62

established that membrane proteins with an extralumenal domain of 90 kDa still diffuse to the INM on a time-scale of hours [28]. This now allows us to better assess if active transport indeed breaks size restrictions for passive diffusion. Thirdly, we improved the assay to better account for the impact of protein synthesis in our assay.

FIG.4. NE ACCUMULATION OF MEMBRANE PROTEIN REPORTERS WITH EXTRALUMENAL DOMAINS OF INCREASING SIZE.

A. Fluorescence microscopy of cells expressing G-NLS-L-RibU, MG-NLS-L-RibU, MGM2-NLS-L-RibU. M:MBP. B. Average accumulation of reporter proteins at the NE over the ER after different regimes of expression, and import inhibition. The reporter proteins were expressed for 1 hr (striped bars); subsequently expression was inhibited by glucose for 1 hr (grey bars) and finally import was blocked by rapamycin for 1 hr (white bars). Alternatively, transcription and import were inhibited simultaneously (black bars). Average of 20 cells; SEM are indicated. C. Viability of cells expressing membrane proteins with different transmembrane domains and differently-sized extralumenal domains.


63

Chap

ter 4

The new reporters consisted predominantly of the transmembrane domain of Sec61 (10 transmembrane segments). However, previous publications have suggested that Sec61 may exist in the INM [25] and thus Sec61 may have specific protein interaction partners in the INM. To exclude that the Sec61 transmembrane domain contributed to the sorting, we additionally repeated the INM targeting with reporters with a TM domain unrelated to yeast. We used the polytopic TM domain from RibU from a Gram-positive bacterium Lactococcus lactis, thereby excluding the chance of specific retention of the reporter in the INM via interaction with the other INM proteins. The fusion indeed sorts effectively to the INM (Fig. 4A, G-NLS-L-RibU) confirming our previous claims that import is not dependent on the nature of the TM domain. Next, a set of reporter membrane proteins was constructed with extralumenal domains of increasing size. Their extralumenal domains were composed of GFP, one, two or three copies of MBP together with the NLS-L motif, resulting in extralumenal domain size of 95 kDa (MG-NLS-L-S and MG-NLS-L-RibU); 136 kDa (MGM-NLS-L-S) and 176 kDa (MGMM-NLS-L-S and MGMM-NLS-L-RibU). From previous work, where we correlated NE/ER ratios to immunoEM observations, we know that high NE/ER ratios correlate with accumulation at the INM ([15] and Popken, accepted in JBC). In the Kap95AA strain background, we find that a protein lacking complete sorting signals, such as G-ΔNLS-L-TM, gives an NE/ER ratio similar to the NE/ER ratio found for mCherry-HDEL (Figure 4B, NE/ER ratio of 1.9+/-0.1). When expressed in yeast, the MG-NLS-L-S and MGM-NLS-L-S reporters enter the nucleus and accumulate at the INM (NE/ER ratios of 19.8 and 5.7 respectively), while the MGMM-NLS-L-S reporter doesn’t accumulate and has a NE/ER ratio of 2.1, similar to mCherry-HDEL (Fig. 4B, striped bars). To estimate how synthesis of new reporters and import of existing ones affect the NE/ER ratios, we inhibited the expression of the reporters from the GAL promoter after one hour of induction, by adding glucose to the growth medium (Fig. 4B, grey bars). The cultures were imaged again after one hour: the total fluorescence levels measured in the NE and the ER increase, which indicates that some reporter protein has still been synthesized and/or matured. Importantly, the accumulation levels (NE/ER) increase for the reporters MG-NLS-L-S and MGM-NLS-L-S, consistent with continued import at a reduced synthesis rates after transcription repression (Fig. 4B, compare striped and grey bars). The NE/ER ratios of the MGMM-NLS-L-S and mCherry-HDEL reporters do not change in this timeframe. Next, we determined the passive diffusion or leak of the reporters from the nucleus to the ER. The reporter proteins were expressed in the KAP95AA strain, which expresses Pma1-FKBP and Kap95-FRB, and in which the Kap95/Kap60-dependent active import can be conditionally blocked by the addition of rapamycin. The experiments were performed as follows: first protein expression was induced (1 hour galactose), then expression was repressed (1 hour glucose). As we showed in Fig.

Chapter 4

64

4B, grey bars, during this hour in glucose, translation of existing mRNAs, maturation of fluorophores and import continues. Finally, rapamycin was added which results in a fast block of nuclear import [17,22]. The distribution of the reporter proteins over the NE-ER network was determined after one hour (Fig. 4B, white bars). This experimental scheme is better than the one used previously [15,17], where the protein synthesis was not inhibited before addition of rapamycin, since the NE/ER ratio measured is less influenced by reporter synthesis (Fig. 4B, black bars). Only the reporter protein with the smallest extralumenal domain, MG-NLS-L-S, leaks out of the nucleus when import is blocked: the NE/ER ratio drops from 37.9 to 12.0, 1 hour after addition of rapamycin (Fig. 4B, compare grey and white bars). The accumulation levels of the MGM-NLS-L-S reporter do not decrease when the import is blocked, indicating that this reporter protein cannot passively diffuse out of the nucleus, consistent with the size limits in [28]. When active import and transcription were inhibited simultaneously, i.e. by the addition of rapamycin and glucose after 1 hr of galactose-induced expression, the NE/ER ratios decrease for both MG-NLS-L-S (from 19.8 to 3.7) and MGM-NLS-L-S (from 5.7 to 2.8) (Fig. 4B, compare striped and black bars). This decrease in the NE/ER is the result of the reporter being synthesized, albeit at a reduced rate, while Kap-dependent INM import is blocked. This is another confirmation that INM accumulation of the MG-NLS-L-S and MGM-NLS-L-S reporters is the result of Kap-dependent import. Additional support that the type of transmembrane domain, monotopic of polytopic, is not critical for the targeting comes from viability tests. As shown in Fig.2, INM accumulation of the G-NLS-L-TM reporter results in deformed nuclei. In figure 4 we show that the deformation correlates with the levels of INM accumulation, and with the viability of the cells: bigger extralumenal domains result in lower accumulation (Fig. 4B) and increased viability (Fig. 4C). This is observed irrespective of the type of transmembrane domain. In conclusion, our findings in figure 4 confirm that the “NLS-L” motif enables the import of extralumenal domains bigger than by passive diffusion. This is best illustrated by the reporter MGM-NLS-L-S; this molecule is small enough to be imported but too large to passively efflux from the NPC.

Transport route through the NPC

Having confirmed that membrane proteins with NLS and L can have larger extralumenal domains compared to those without sorting signal, suggests that they pass through a different area of the NPC, or that the NPC adopts a different conformation during transport. We previously proposed a route through the central channel as we see FG-Nup dependence of INM accumulation and showed that the extralumenal domains could be trapped at Nsp1 [15]. However, recent publications show a non-NPC pool of Nsp1 [26,27] that may have affected our measurements. Here, besides the scaffold


65

Chap

ter 4

nucleoporin Nup170 (Figure 3), we also tested the FG-Nups Nup53 and Nup59 astrapping sites for our reporters. All three nucleoporins are located relatively close to the pore membrane [29-33]. Upon incubation with rapamycin, we saw the characteristic punctate localization pattern of the F-G-NLS-L-TM reporter in all tested Nup-FRB strains (Fig. 5A). The two reporters that fail to accumulate in the nuclear envelope (F-G-ΔNLS-L-TM and F-G-NLS-ΔL-TM), can also be trapped at Nup170, Nup53 and Nup59. In these cases rapamycin does not only cause clustering of the fluorescence signal in distinct points of the nuclear envelope, but also an almost complete shift of the ER-localized molecules to the NE (Fig. 5B and 5C). This data reinforces the mobility of our reporters between ER and the pore membrane, and also confirms that the tested reporters are not restricted

B)

+DM

SO+r

ap 1

0ug/

ml 4

0min

Nup170FRB Nup53FRB Nup59FRBF-G-ΔNLS-L-TM

C) F-G-NLS-ΔL-TM

+DM

SO+r

ap 1

0ug/

ml 4

0min

Nup170FRB Nup53FRB Nup59FRB

A) F-G-NLS-L-TMNup53FRB

+DM

SO+r

ap 1

0ug/

ml 4

0min

Nup59FRB

FIG.5. THE REPORTERS WITHOUT THE NLS OR THE LINKER DOMAIN CAN ENTER THE NPC.

A. Confocal fluorescence microscopy images showing cells expressing F-G-NLS-L-TM in Nup53FRB and Nup59FRB strains background after incubation with rapamycin (+rap 10μg/ml) and in control conditions (+DMSO). The fluorescence patterns change from continues to punctate upon rapamycin treatment. B., C. As A. but with cells expressing F-G-ΔNLS-L-TM (B) or F-G-NLS-ΔL-TM (C) in Nup170FRB, Nup53FRB and Nup59FRB strain background. The fluorescence signal disappears from the ER upon rapamycin treatment. Scale bars: 2 μm.

Chapter 4

66

from entering the NPC in the absence of a NLS or a linker region. Unfortunately, as trapping occurs also in the absence of the sorting signals we cannot interpret the trapping event as one that uniquely reflects active import.

FIG.6. ANALYSIS OF THE FLUORESCENCE IMAGES FROM FIGURE 4.

A. Comparison of the average SD fraction (as in Fig. 3C) for F-G-NLS-L-TM, F-G-ΔNLS-L-TM, F-G-NLS-ΔL-TM in all the trap strains (Nup53FRB, Nup59FRB, Nup170FRB) and the background strain (K14708). Grey columns – control conditions (no rap); columns with black diagonal stripes – cells incubated with rapamycin (+rap). Number of cells analyzed is between 29 and 55. B. Percentage of cells expressing F-G-ΔNLS-L-TM or F-G-NLS-ΔL-TM that display fluorescence in the ER in the trap strains (Nup53FRB, Nup59FRB, Nup170) and the background strain (K14708). The number of cells analyzed is between31 and 158.


67

Chap

ter 4

Despite the limitations of the assay, we noted differences in trapping efficiency dependent on the presence of an NLS and dependent on the trap position. The increase in SD fraction seemed smaller in the absence of the NLS (Figure 6A) possibly reflecting that trapping is less efficient. This NLS-dependent trapping efficiency is more clearly observed when looking at the percentage of cells that show fluorescence at the ER (Fig. 6B). For F-G-NLS-ΔL-TM we see that the percentage of cells that show fluorescence at the ER drops from around 70% to below 20% after trapping at Nup170, Nup53 or Nup59. Trapping of the F-G-ΔNLS-L-TM reporter is less efficient at Nup59 and Nup53 as here still around 55% of the cells show fluorescence at the ER, while this value is around 20% when trapped at Nup170. The difference in trapping efficiency may reflect a different residence time of the reporters in the NPC and suggests increased residence close to Nup53 and Nup59 for reporters with the NLS as compared to those without. While the trap assay is not suitable to dissect the native path through the NPC, we can conclude that the extralumenal domains can reach positions in the NPC close to the membrane. DISCUSSION

In this work we have used a Heh2-based set of reporters to study membrane protein biogenesis and transport to the INM. Our main goal was to distinguish the different signals encoded in the protein sequence that could potentially alter the final localization of the reporters without being involved in the nuclear transport process itself. The use of protein truncations is a common practice in the field of INM protein traffic [12-15,18,34-39]. We were curious if some of the effects of the truncations that we have applied to Heh2 in the past [15] might have resulted in artificial import to the nucleus. We investigated the membrane insertion status of the Heh2-based reporters. We found that the insertion of the reporters did not depend on the GET system, or at least was not solely dependent on it, as deletion of GET components did not interfere with the insertion. The Sec61 system is likely responsible for membrane insertion of our reporters although we were not able to test it directly and other systems could not be ruled out. None of the changes that we applied to the G-NLS-L-TM (whether it was a truncation or addition of another domain) changed its transmembrane nature. The insertion state of our reporters depended on the transmembrane domain only. Moreover, our data pointed towards insertion in the ER preceding the transport to the NE, since we show that while (trapped) in the NPC, the reporter fractionates like an integral membrane protein. Consistently, the NPC-anchored Nup170, normally fractionating as a peripheral protein, became salt-insoluble upon anchoring to the reporter protein. Altogether these data show that our reporters do not pass the NPC in a chaperoned-soluble state but as transmembrane proteins. We thus conclude that our

Chapter 4

68

previously published conclusion, that the NLS and ID linker are required and sufficient for active nuclear transport of a transmembrane embedded protein, stands. With a new series of reporters that have the Sec61 transmembrane domain instead of a single transmembrane helix, and an improved experimental setup, we confirm that the NLS and linker domain enable the transport of proteins to the INM that would be too large for passive diffusion. The improved experimental setup takes into account that synthesis of new reporters influences the NE/ER ratio measured and reports more modest tolerance for extralumenal size than we previously reported [15]. Indeed, when simultaneously blocking nuclear transport by addition of rapamycin (which depletes the cytosol of Kap95) and transcription with glucose (which inhibits transcription of galactose-inducible reporters) one sees a decrease in the NE/ER ratio. This decrease was previously explained by efflux from the INM, but here we show that in addition residual synthesis plays a role. In the current assay we inhibited the expression for one hour using glucose before addition of rapamycin. We show that with the NLS-L motif proteins having extralumenal domains bigger than 136 kDa, but smaller than 176 kDa, are actively imported. This is significantly larger than what is found for passive diffusion without the NLS-L motif, which is limited to proteins with an extralumenal domain size of approximately 90 kDa [28]. We conclude, the “NLS-L” motif enables the import of extralumenal domains bigger than by passive diffusion, which is best illustrated by the reporter MGM-NLS-L-S with an extralumenal domain of 136 kDa. This proteins is small enough to be imported but too large to passively efflux from the NPC. Lastly, using yet another transmembrane domain (RibU) we re-emphasize that sorting is not dependent on the type of transmembrane domain. The route that these actively transported proteins take through the NPC is unresolved. In this paper we show that an extralumenal FKBP domain on the reporter could be trapped in the NPC at Nup170, Nup53 or Nup59, all positions close to the pore membrane, even in the absence of an NLS or a linker region. This shows that, at least for these small reporters, the NLS or linker regions are not required for entry in the NPC. The trapping assay merely reports if the FRB and FKBP tags are ever in proximity and may thus stabilize rare positions. The assay however detects subtle differences in the localization of the FKBP tag in these reporters, since the efficiency to trap at Nup53 or Nup59 is reporter dependent. Trapping at Nup53 and Nup59 is less efficient with the reporter lacking an NLS (F-G-ΔNLS-L-TM) compared to those with NLS-L (F-G-NLS-L-TM) or lacking the linker (G-F-NLS- ΔL-TM). One explanation could be that with the NLS, the FKBP tag spends more time close to Nup53 and Nup59, which would argue that the active transport could occur close to the membrane. Previously we have shown that we could trap the same reporter as used here (F-G-NLS-L-TM) at Nsp1-FRB and suggested a route through the center of the pore. However, last year new data appeared showing that a pool of Nsp1 functions outside of the nuclear pore [26,27]. This pool of “soluble” Nsp1 may have affected our experiments.


69

Chap

ter 4

Altogether, the path through the NPC remains unresolved but at present a route closer to the pore membrane seems more likely. Overall, the current study supports that in yeast Heh2-derived reporters with NLS-L motif are transported through the NPC while embedded in the membrane, and this process is energy-dependent. The spatial route through the NPC may be more close to the membrane than previously reported.

Chapter 4

70

REFERENCES

1. D’Angelo MA, Hetzer MW: The role of the nuclear envelope in cellular organization. Cell. Mol.

Life Sci. 2006, 63:316–332.2. Akhtar A, Gasser SM: The nuclear envelope and transcriptional control. Nat Rev Genet 2007,

8:507–517.3. Shimi T, Butin-Israeli V, Goldman RD: The functions of the nuclear envelope in mediating the

molecular crosstalk between the nucleus and the cytoplasm. Current Opinion in Cell Biology 2012, 24:71–78.

4. Winey M, Yarar D, Giddings THJ, Mastronarde DN: Nuclear Pore Complex number and distribution throughout the Saccharomyces cerevisiae cell cycle by three-dimensional reconstruction from electron micrographs of nuclear envelopes. Molecular Biology of the Cell 1997, 8:2119–2132.

5. Tapley EC, Starr DA: Connecting the nucleus to the cytoskeleton by SUN–KASH bridges across the nuclear envelope. Current Opinion in Cell Biology 2013, 25:57–62.

6. Shao S, Hegde RS: Membrane protein insertion at the endoplasmic reticulum. Annu. Rev. Cell Dev. Biol. 2011, 27:25–56.

7. Aviram N, Schuldiner M: Embracing the void—how much do we really know about targeting and translocation to the endoplasmic reticulum? Current Opinion in Cell Biology 2014, 29:8–17.

8. Denic V: A portrait of the GET pathway as a surprisingly complicated young man. Trends in Biochemical Sciences 2012, 37:411–417.

9. Stewart M: Molecular mechanism of the nuclear protein import cycle. Nat Rev Mol Cell Biol 2007, 8:195–208.

10. Wente SR, Rout MP: The Nuclear Pore Complex and Nuclear Transport. Cold Spring Harbor Perspectives in Biology 2010, 2:a000562–a000562.

11. Grünwald D, Singer RH: Multiscale dynamics in nucleocytoplasmic transport. Current Opinion in Cell Biology 2012, 24:100–106.






17. Meinema AC, Poolman B, Veenhoff LM: Quantitative Analysis of Membrane Protein Transport Across the Nuclear Pore Complex. Traffic 2013, 14:487–501.



71

Chap

ter 4



21. Gruber S, Arumugam P, Katou Y, Kuglitsch D, Helmhart W, Shirahige K, Nasmyth K: Evidence that Loading of Cohesin Onto Chromosomes Involves Opening of Its SMC Hinge. Cell 2006, 127:523–537.


23. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al.: Fiji: an open-source platform for biological-image analysis. Nature Methods 2012, 9:676–682.

24. Grund SE, Fischer T, Cabal GG, Antunez O, Perez-Ortin JE, Hurt E: The inner nuclear membrane protein Src1 associates with subtelomeric genes and alters their regulated gene expression. The Journal of Cell Biology 2008, 182:897–910.

25. Deng M, Hochstrasser M: Spatially regulated ubiquitin ligation by an ER/nuclear membrane ligase. Nature 2006, 443:827–831.

26. Colombi P, Webster BM, Frohlich F, Lusk CP: The transmission of nuclear pore complexes to daughter cells requires a cytoplasmic pool of Nsp1. The Journal of Cell Biology 2013, 203:215–232.

27. Makio T, Lapetina DL, Wozniak RW: Inheritance of yeast nuclear pore complexes requires the Nsp1p subcomplex. The Journal of Cell Biology 2013, 203:187–196.

28. Popken P, Ghavami A, Onck PR, Poolman B, Veenhoff LM: Size-dependent leak of soluble and membrane proteins through the yeast nuclear pore complex. Molecular Biology of the Cell 2015, 26:1386–1394.

29. Eisenhardt N, Redolfi J, Antonin W: Interaction of Nup53 with Ndc1 and Nup155 is required for nuclear pore complex assembly. Journal of Cell Science 2014, 127:908–921.

30. Onischenko E, Stanton LH, Madrid AS, Kieselbach T, Weis K: Role of the Ndc1 interaction network in yeast nuclear pore complex assembly and maintenance. The Journal of Cell Biology 2009, 185:475–491.

31. Amlacher S, Sarges P, Flemming D, van Noort V, Kunze R, Devos DP, Arumugam M, Bork P, Hurt E: Insight into Structure and Assembly of the Nuclear Pore Complex by Utilizing the Genome of a Eukaryotic Thermophile. Cell 2011, 146:277–289.

32. Marelli M, Aitchison JD, Wozniak RW: Specific binding of the karyopherin Kap121p to a subunit of the nuclear pore complex containing Nup53p, Nup59p, and Nup170p. The Journal of Cell Biology 1998, 143:1813–1830.


34. Soullam B, Worman HJ: The amino-terminal domain of the lamin B receptor is a nuclear envelope targeting signal. The Journal of Cell Biology 1993, 120:1093–1100.

35. Hasan S, Güttinger S, Mühlhäusser P, Anderegg F, Bürgler S, Kutay U: Nuclear envelope localization of human UNC84A does not require nuclear lamins. FEBS Letters 2006, 580:1263–1268.


Chapter 4

72

37. Jaspersen SL: The Sad1-UNC-84 homology domain in Mps3 interacts with Mps2 to connect the spindle pole body with the nuclear envelope. The Journal of Cell Biology 2006, 174:665–675.

38. Ostlund C, Ellenberg J, Hallberg E, Lippincott-Schwartz J, Worman HJ: Intracellular trafficking of emerin, the Emery-Dreifuss muscular dystrophy protein. Journal of Cell Science 1999, 112 ( Pt 11):1709–1719.

39. Brachner A: LEM2 is a novel MAN1-related inner nuclear membrane protein associated with A-type lamins. Journal of Cell Science 2005, 118:5797–5810.

Chapter 5Nsp1 in the trapping experiments – bait or prey?

Chapter 5

74

ABSTRACT

In the previous chapters we have studied the nuclear import of membrane reporters destined for the Inner Nuclear Membrane (INM) using a rapamycin-dependent complex formation between FRB and FKBP domains. In our set up an FKBP-tagged reporter protein has been trapped at the FRB-tagged nucleoporin located in the central channel of the NPC, namely Nsp1. New data, which were not available at the time when we have designed the trapping experiments, suggest that a cytoplasmic pool of Nsp1 might exist. In this chapter we investigate at which pool of Nsp1 the FKBP reporter proteins actually trap and conclude that trapping takes place at both cytoplasmic and nuclear pools, but likely with different affinity for different reporters. We also consider the influence of these findings on our previous conclusions about the transport mechanism of membrane proteins. The trapping assay most likely does not solely report on transport phenomena in the pore.


75

Chap

ter 5

Nuclear Pore Complexes (NPCs) of yeast S. cerevisiae are big multiprotein assemblies that span the Nuclear Envelope (NE). They form the only gateway between the nucleus and the cytoplasm. The NPC is composed of only 30 proteins, nucleoporins, which are present in multiple copies. The NPC has a 8-fold rotational symmetry. The 8 symmetry units, called spokes, are interconnected and form co-axial rings: a membrane ring, 2 inner rings in the middle, and 2 outer rings (exposed to the cyroplasm or nucleoplasm)[1]. The inner and outer rings form the scaffold of the NPC [1]. Linker nucleoporins are found in between the inner and outer rings[1]. Another important type of nucleoporins – FG-Nups – fills up the transport channel. FG-Nups are natively unfolded proteins that are required for the transport function of the NPC. Yeast NPCs are 40nm high and 100nm in diameter, with the central channel diameter being approximately 40nm [1]. In the previous chapters an assay was described, where FKB-tagged inner nuclear membrane (INM) reporter proteins are trapped by FRB-tagged nucleoporins, namely inner ring Nup170 and 3 FG-nups: Nup53, Nup59 and Nsp1. Three of these nups are located in close vicinity to the pore membrane, while Nsp1 is positioned closer to the central channel of the pore. We reported trapping of h2NLS-L bearing reporter on Nsp1 and interpreted it as a potential active pathway for transport of membrane proteins [2]. The linker domain is natively unfolded [2] and when completely extended can potentially enable exposure of the NLS roughly 20nm away from the pore membrane. The same reporter and truncated versions of it, lacking the NLS or the linker domains, trap at the membrane-exposed Nup170, Nup53 and Nup59 (previous chapter). These findings possibly weaken our conclusions that the reporter traps at the Nsp1 during active import only. Moreover, a recently published study [3] suggests the existence of a cytoplasmic pool of Nps1, which was not known at the time we designed the trapping assay. The above-mentioned cytosolic pool of Nsp1 is involved in the distribution of NPCs between the mother cell and the bud during cell division. Upon knockdown of Nsp1 expression, the NPCs will not cross the budneck, although nuclear envelope division progresses [4]. Foci containing newly synthesized members of Nsp1 subcomplex (Nsp1, Nic96, Nup49, Nup57) were observed in the buds [3]. Inhibiting the formation of these foci by trapping the newly synthesized Nsp1, Nic96, Nup49 or Nup57 molecules on the plasma membrane causes a failure in NPC transmission to the bud [3]. Moreover, nucleopodia movements via the budneck towards bud localized Nsp1 foci were observed preceding the NPCs transmission [3]. As the Nsp1 foci are likely associated with the ER [3], their presence in the ER-NE network may have influenced our previous Nsp1 experiments, where we monitored trapping of INM reporter proteins at Nsp1. This study aims to answer at which pool of Nsp1 the FKBP-tagged INM reporter proteins actually trap: at the NPC or at

Chapter 5

76

the cytoplasmic foci. We conclude that it can trap at both, and evaluate our original conclusions drawn about the path that membrane proteins follow through the NPC [2].

MATERIALS AND METHODS


All strains and plasmids are listed in Table 1 and 2 of the Appendix. All the experiments described here were performed in the S.cerevisiae K14708 strain [5] or strains derived from it. For the construction of strain used in the NPC trapping experiment, the gene encoding nucleoporin Nsp1 was tagged with an FRB cassette from the pFA6a-FRB-KanMX6 plasmid [6]. For the construction of the strain for double color imaging, the gene encoding nucleoporin Pom34p was tagged with a GFP cassette from plasmid pYM25 [7]. All the INM reporters were expressed from pACM021-GFP plasmid [2]. All the FKBP tagged reporters were designed based on the pJKL01-2×FKBP12-GFP-Lic-h2NLS-L-TM plasmid [2].

Growth conditions

Yeast strains were grown to mid-exponential growth phase at 30°C in synthetic dropout medium without L-histidine. Strains were first grown for 1 day on medium containing 2% glucose (w/v) and then cultured on medium containing 2% raffinose (w/v). On the day of the experiment the expression of the reporters was induced with 0,5% galactose (w/v). For the trapping experiments rapamycin was used at a final concentration of 10ug/ml.

Fluorescence microscopy

The live cell images from fig. 1,3, 4 and 6 were collected with an LSM 710 confocal microscope (CarlZeiss MicroImaging, Jena, Germany) using an objective C-Apochromat 40×/1.2NA, a solid state laser (488 nm) for excitation and ZEN2010B software. The SNAP-tagged and fixed cells from fig. 2 were imaged with a Delta Vision Microscope (Applied Precision), using InsightSSITM Solid State Illumination at 488 nm and an Olympus UPLS Apo 100x oil objective with 1.4 NA. Detection was done with a coolSNAP HQ2 camera. Image stacks (30 stacks, of 0.2 μm) were deconvolved using standard settings.

SNAP-tagging

The SNAP-tagging protocol was adapted from the nanobodies labeling protocol [8,9]. 10ml of exponentially growing Nsp1FRB Pom34GFP yeasts expressing the F-SN-NLS-L-TM reporter were harvested by centrifugation. 50ul of the concentrated yeast suspension was applied on a concanavalin A-covered Mattek glass bottom microscopy


77

Chap

ter 5

dish (50mm, glass no. 1.5, Mattek Corporation, Ashland, MA, USA). Yeast cells were fixed by incubation in 4% paraformaldehyde, 2% sucrose in PBS (15 minutes, room temperature, gentle rotation). The reaction was quenched by 50mM NH4Cl in PBS (15 minutes, room temperature, gentle rotation). Cells were permeabilized with 0,25% Triton X-100 in PBS supplemented with 1%BSA, 0,004% NaN3 for 5 min at room temperature with gentle rotation. The permeabilization solution was washed out thoroughly and the cells were blocked with PBS supplemented with 5%BSA, 0,004% NaN3 and 0.01% Tritonx100 (30 min) and subsequently with Image IT FX (Invitrogen). SNAP Surface Alexa647 (NEB) was diluted 1:800 in PBS containing 1% BSA, 0,004% NaN3, 50mM DTT and 0,01% Triton x100. The reaction was incubated for 30 min at 37°C. Excess of the dye was washed out with PBS.

Analysis of the standard deviation of fluorescent intensity (SD fraction quantification)

Analysis of the fluorescent images was performed using ZEN2010B software and Fiji [10]. Line scans of the fluorescence intensity over NE were collected for at least 30 cells for each strain in each condition. For each cell two values were calculated: average fluorescence intensity in the NE and the standard deviation of it. The standard deviation values for each cell were divided by the average fluorescence intensity values, and the mean value of n cells (later in the chapter referred to as SD fraction) was calculated from the resulting ratios.

N/C ratios calculations

The average fluorescence intensity in the nucleus was divided by the average fluorescence intensity in the cytoplasm for each cell and the mean N/C value was calculated.

Membrane protein extraction and salt and detergent solubilization

Whole cell extracts were created from 10ml cultures of exponentially growing yeast cells by lysis in NaOH and β-mercaptoethanol and TCA- precipitation. For preparation of crude membrane fractions, 200ml of exponentially growing yeast cells were harvested by centrifugation (5500 g 10 min). The pellet was washed with cold demi water and centrifuged again. Subsequently, the yeast pellet was resuspended in 10ml of 20mM TrisHCl pH 8.0, 0.01M EDTA with 35ul of β-mercapthoethanol and incubated for 15 min on ice with occasional swirling. The suspension was centrifuged (10000rpm, 5 min), washed with buffer containing 1.1M sorbitol, 50mM TrisHCl pH 7,5, 10mM MgCl2, 3mM DTT and centrifuged again. The pellet was resuspended in the sorbitol buffer again to achieve 1.5x109 cells per 1ml. Zymolyase was added to the final concentration of 0,02%. Cells were incubated at 300C for 1 hour. The spheroplasts were pelleted (4000g 10min), carefully washed with sorbitol buffer containing 1mM PMSF, 2ug/ml of pepstatin and protein inhibitors cocktail and centrifuged again. From this step the membrane isolation and solubilization of proteins with 1% Triton X-100 or 1M NaCl

Chapter 5

78

were performed as described in [11]. In short, the spheroplasts were resuspended in 3ml of 20mM TrisHCl pH 8,0 150mM NaCl and lyzed in a Potter tube. The cell debris was removed by short centrifugation (2min, 2000rpm). The supernatant was centrifuged again (20min, 13000rpm) and the pellet containing membrane fraction resuspended in 500ul 20mM Tris HCl containing either 1% Triton X-100 or 1M NaCl or 150mM NaCl. After 30 minutes incubation on ice the samples were centrifuged (100,000xg, 1 hour). The supernatants (solubilized fraction) and pellets (non-solubilized fraction) were mixed with SDS-PAGE loading buffer, incubated at 70°C for 10 minutes and loaded on SDS PAA gels.

Western Blots

Samples were separated by SDS-PAGE electrophoresis and transferred to a PVDF membrane (wet transfer, 16V, 17 hours). For detection of the reporters anti-FKBP antibody (Abcam, Cambridge, UK) was used at a dilution of 1:1000 (v/v) and a secondary anti-rabbit-alkaline phosphatase conjugate at 1:20,000 (v/v) (Sigma-Aldrich, St. Louis, USA). For the detection of Nsp1FRB anti-Nsp1 antibody was used at 1:1000 (v/v) (Abnova, Taipei, Taiwan) with anti-mouse-alkaline phosphatase conjugate at 1:20,000 (Sigma-Aldrich, St.Louis, USA). For the detection of Pma1 anti-Pma1 antibody was used at 1:5000 (v/v) (Abcam) with the anti-mouse-alkaline phosphatase conjugate again. For the detection of Dpm1 anti-Dpm1 antibody was used at 1:250 (v/v) (Abcam) with the anti-mouse-alkaline phosphatase conjugate again. Pre-stained molecular weight marker (ThermoScientific) was used.

RESULTS

Previously we have used Heh2p based reporter proteins to study the transport through the NPC ([2] and previous chapters). A minimal reporter that was imported to the nucleus in an energy dependent manner consisted of one transmembrane segment (TM), a linker domain (L) and a Nuclear Localization Signal (NLS) and a GFP for detection. For the nuclear pore trapping experiments this reporter was N terminally tagged with an FKBP molecule. We have shown previously that both actively transported F-G-NLS-L-TM as well as the passively diffusing variant of it, the F-G-ΔNLS-L-TM reporter, can be trapped at FRB-tagged versions of Nup53, Nup59 and Nup170 upon rapamycin treatment (previous chapter). The F-G-ΔNLS-L-TM reporter, however, displayed a trapping efficiency that depended on the position within the NPC, as more cells showed trapping with Nup170 than with Nup53 or Nup59. If trapping of F-G-NLS-L-TM on Nsp1FRB happens exclusively during active import, as we have previously suggested [2], then the reporter without NLS, F-G-ΔNLS-L-TM, should not bind the FRB at this position. In fig.1 we present data on the trapping of the F-G-NLS-L-TM reporter on Nsp1FRB along with that of the F-G-ΔNLS-L-TM reporter. After the rapamycin treatment localization of both reporters changes from continues


79

Chap

ter 5

no rap

+ rap

+ rapno rap

+ rap

no rap

A) B)

C)

D)

no rap

+ rap

no rap

+ rap

Position at the NE [μm]



Fluo

resc

ence

inte

nsity

[AU

]Fl

uore

scen

ce in

tens

ity [A

U]

Fluo

resc

ence

inte

nsity

[AU

]

Nsp1FRB Nsp1FRBF-G-NLS-L-TM F-G-ΔNLS-L-TM

F-G-NLS-L-TM F-G-ΔNLS-L-TM

+DM

SO

+rap

100

μg/

ml

40m

in

F-G-ΔNLS-L-TM

Nsp1FRB Nsp1FRB

Nsp1FRB

K14708 K14708

K14708

SD

frac

tion

SD

frac

tion

strain: K14708reporter: F-G-NLS-L-TM

strain: Nsp1FRBreporter: F-G-NLS-L-TM

strain: Nsp1FRBreporter: F-G-ΔNLS-L-TM

% c

ells

with

ER

signal over the NE or NE-ER network (in case of F-G-ΔNLS-L-TM reporter), into punctuate pattern at the NE (fig.1A). The analysis of the standard deviation of the fluorescence intensity over the NE shows much higher variation in case of rapamycin-treated samples (line scans fig.1b). The average standard deviation of the fluorescence

FIG.1. BOTH FKBP-GFP-NLS-L-TM AND DNLS REPORTERS TRAP AT NSP1FRB.

A. Confocal fluorescence microscopy images showing Nsp1FRB cells expressing F-G-NLS-L-TM and F-G-ΔNLS-L-TM reporters after incubation with rapamycin (+rap 100μg/ml) and in control conditions (+DMSO). Scale bars: 2 μm. B. Typical line-scans of the fluorescence intensity in the NE of the representative cells from each strain, after incubation with rapamycin (black line, +rap) or in control conditions (grey line, no rap). Top: reference strain K14708 strain expressing F-G-NLS-L-TM reporter. Middle: Nsp1FRB strain expressing F-G-NLS-L-TM reporter. Bottom: Nsp1FRB strain expressing F-G-ΔNLS-L-TM reporter. C. Comparison of the average standard deviation of the fluorescence intensity divided by average fluorescence intensity (SD fraction) for each reporter in the Nsp1 trap strain (Nsp1FRB) and in the reference strain (K14708). Left: F-G-NLS-L-TM (n=47 for K14708 no rap; n=55 for K14708 +rap; n=31 for Nsp1FRB no rap; n=44 for Nsp1FRB +rap). Right: F-G-ΔNLS-L-TM reporter (n=34 for K14708 no rap; n=33 for K14708 +rap; n=22 for Nsp1FRB no rap; n=36 for Nsp1FRB +rap). Grey columns – control conditions (no rap); columns with black diagonal stripes – cells incubated with rapamycin (+rap). D. Percentage of cells expressing the F-G-ΔNLS-L-TM that display fluorescence signal in the ER in the Nsp1 trap strain (Nsp1FRB, n=38 for no rap, n=126 for +rap) and in the reference strain (K14708, n=38 for no rap, n=78 for +rap). Grey columns – control conditions (no rap); columns with black diagonal stripes – cells incubated with rapamycin (+rap).

Chapter 5

80

intensity equals 38% of the average fluorescence intensity in case of trapped F-G-NLS-L-TM reporter and 46% in case of trapped F-G-ΔNLS-L-TM reporter (SD fractions fig.1c). For the reporters in the wild type K14708 strain or in the Nsp1FRB strain without rapamycin treatment this number equals roughly 25-30% (SD fractions fig.1c). Moreover, the F-G-ΔNLS-L-TM reporter disappears from the ER after rapamycin treatment. In control cells treated with DMSO, 84% of the cells have a fluorescence signal in the ER, while this number drops to 44% after rapamycin is added into the culture (graph fig.1d), which is much higher than the 20% observed with Nup170, but slightly lower than the 55% observed with Nup53 and Nup59 (previous chapter). This data indicates that the F-G-ΔNLS-L-TM reporter, as well as the F-G-NLS-L-TM reporter, can be trapped at Nsp1FRB. We have also performed a colocalization experiment with Pom34, a nucleoporin that is not associated with Nsp1 complex, to confirm that the punctate pattern can be really interpreted as trapping at the NPCs (fig.2). Clearly the presence of an NLS in the Heh2p based reporters has little influence on trapping on Nsp1FRB, similarly to what was previously observed for the other trapping positions (Nup53FRB, Nup59FRB, Nup170FRB, previous chapter). Nsp1FRB trapping is therefore not a distinguishing factor between active and passive transport processes.

Assuming that the non-NPC pool of Nsp1 has no influence on our findings, the interaction of both membrane reporters with the centrally-located Nsp1 can be explained by extending the flexible linker domain through the scaffold of the NPC to the center of the nuclear pore complex [2]. This would likely require temporal restructuring of the dense assembly of the NPC. Such a energy dependent event was suggested before as a mechanism for active import of proteins to the INM [2,12]. This could explain F-G-NLS-L-TM trapping at Nsp1, but not the reporter without the NLS, because it is unclear what would trigger the exposure of N-terminal domains of the reporter to the center of the pore without Kap-FG-Nup interactions. The possibility of

FIG.2. F-SN-NLS-L-TM COLOCALIZES WITH POM34 WHEN TRAPPED AT NSP1.

Fluorescence microscopy images showing F-SN-NLS-L-TM reporter expressed in the Nsp1FRB Pom34GFP strain. Scale bars: 2 μm. Green – Pom34GFP; red – SNAP tagged reporter, labeled with Alexa Fluor 647.

+rap

100

ug/m

l 4

0min

Nsp1FRB Pom34GFP F-SN-NLS-L-TMGFP Alexa647 merger


81

Chap

ter 5

trapping one or both of these reporters at the cytoplasmic fraction of Nsp1, however, could blur the readout of these experiments. By removing the long flexible linker from the reporter we could distinguish between two Nsp1 pools. Without the long flexible linker, the F-G-NLS-ΔL-TM reporter should not be able to reach out to the center of NPC. If in the previous experiments we trapped the reporter on the “pore” Nsp1 only - the new reporter should not trap on this position. On the other hand, the trapping of the F-G-NLS-ΔL-TM reporter should not be affected, if our assay is influenced by the “cytoplasmic” Nsp1 pool.

As shown in fig.3 the F-G-NLS-ΔL-TM reporter also changes localization after exposure to rapamycin. Similar to what was observed for the F-G-ΔNLS-L-TM reporter, the fluorescence signal disappears from the ER and a punctate pattern around the NE occurs (fig.3a). The number of cells with fluorescence signal in the ER drops from 77% to 18% after rapamycin treatment (graph fig. 3d). The standard deviation of fluorescence intensity differs by 54% from the average fluorescence intensity, compared

0

1

2

0 1 2 3 4 5 6

0

20

40

60

80

100

A) B)

C) D)

Nsp1FRBF-G-NLS-ΔL-TM

+DM

SO

+rap

100

μg/

ml

40m

in

strain: Nsp1 FRBreporter: F-G-NLS-ΔL-TM

F-G-NLS-ΔL-TM F-G-NLS-ΔL-TM

+rap

no rap

+rap +rap

no rap no rap

SD

frac

tion

0

1

0.2

0.4

0.6

0.8

Fluo

resc

ence

inte

nsity

[AU

]


% o

f cel

ls w

ith E

R

K14708 Nsp1FRB K14708 Nsp1FRB

FIG.3. THE F-G-NLS-ΔL-TM REPORTER TRAPS AT NSP1FRB.

A. Confocal fluorescence microscopy images showing Nsp1FRB cells expressing the F-G-NLS-ΔL-TM reporter after incubation with rapamycin (+rap 100μg/ml) and in control conditions (+DMSO). Scale bars: 2 μm. B. Line-scan of the fluorescence intensity in the NE of the representative cells expressing the F-G-NLS-ΔL-TM reporter after incubation with rapamycin (black line, +rap) or in control conditions (grey line, no rap). C. Comparison of the average standard deviation of the fluorescence intensity divided by average fluorescence intensity (SD fraction) for the F-G-NLS-ΔL-TM reporter in the Nsp1 trap strain (Nsp1FRB, n=29 for no rap, n=35 for +rap) and in the reference strain (K14708, n=29 for no rap, n=30 for +rap). Grey columns – control conditions (no rap); columns with black diagonal stripes – cells incubated with rapamycin (+rap). D. Percentage of cells expressing the F-G-NLS-ΔL-TM reporter that display fluorescence signal in the ER in the Nsp1 trap strain (Nsp1FRB, n=35 for no rap, n=34 for +rap) and in the reference strain (K14708, n=51 for no rap, n=78 for +rap). Grey columns – control conditions (no rap); columns with black diagonal stripes – cells incubated with rapamycin (+rap).

Chapter 5

82

to 42% without rapamycin treatment (linescan and SD fraction graph fig.3b and 3c). Moreover, characteristic fluorescent foci were observed in cytoplasm of the cells (fig. 3a). These observations imply that the F-G-NLS-ΔL-TM reporter traps (also) at the cytoplasmic pool of Nsp1. A)

B)

C)

D)

single cell mother biased foci bud biased foci unbiased foci

F-G

-NLS

-L-T

MF-

G-Δ

NLS

-L-T

MF-

G-N

LS-Δ

L-TM

020406080

100

0

204060

80100

020406080

100

no rap

no rap

+rap

+rap

% o

f cel

ls w

ith fo

ci%

of d

ivid

ing

cells

with

foci

F-G-NLS-L-TM F-G-ΔNLS-L-TM F-G-NLS-ΔL-TM



foci distribution

non dividing cells

unbiased

mother biased

bud biased

FIG.4. REPORTERS TRAP ALSO AT THE CYTOPLASMIC FOCI POOL OF NSP1.

A. Confocal fluorescence microscopy images showing Nsp1FRB cells expressing F-G-NLS-L-TM, F-G-ΔNLS-L-TM and F-G-NLS-ΔL-TM reporters after incubation with rapamycin. Arrows point to the cytoplasmic foci. Note that the bitmap settings for F-G-NLS-L-TM and F-G-ΔNLS-L-TM reporters had to be adjusted in order to show low intensity cytoplasmic foci. Scale bars: 2 μm. B. Percentage of cells with cytoplasmic foci from all cells with fluorescence signal in the NE after incubation with rapamycin (+rap, columns with black diagonal stripes; n=202 for F-G-NLS-L-TM; n=200 for F-G-ΔNLS-L-TM; n=264 for F-G-NLS-ΔL-TM) and in the control conditions (no rap, grey columns; n=177 for F-G-NLS-L-TM; n=163 for F-G-ΔNLS-L-TM; n=212 for F-G-NLS-ΔL-TM). C. Distribution of foci in between all cells with foci. Grey – bud biased foci; white – mother biased foci; squares– unbiased foci; black – foci in non dividing cells D. Percentage of dividing cells with cytoplasmic foci from all dividing cells after incubation with rapamycin (+rap, columns with black diagonal stripes; n=62 for F-G-NLS-L-TM; n=80 for F-G-ΔNLS-L-TM; n=103 for F-G-NLS-ΔL-TM) and in the control conditions (no rap, grey columns; n=49 for F-G-NLS-L-TM; n=63 for F-G-ΔNLS-L-TM; n=85 for F-G-NLS-ΔL-TM).


83

Chap

ter 5

According to Colombi et al. the ‘cytoplasmic pool’ of Nsp1 forms foci located mostly in the buds of dividing cells [3]. These foci are likely associated with the ER. We decided to analyze all our data for the presence of foci of F-G-NLS-L-TM, F-G-NLS-ΔL-TM and F-G-ΔNLS-L-TM. We see their appearance in case of all three reporters. Their presence in the cells is rapamycin dependent (graph, fig.4b), which shows that their appearance is not an artifact caused by problems of focusing the sample in themicroscope. Similar to Colombi et al. [3], we have grouped cells with foci into 4 categories: non-dividing cells with foci, and 3 categories of dividing cells with foci, namely bud-biased, mother-biased and unbiased. Examples of all the categories are presented in fig. 4a. In case of the F-G-NLS-ΔL-TM reporter, the percentage of cells with foci is higher and the foci are bigger and brighter, and therefore easier to detect. In case of the F-G-NLS-L-TM reporter, 29% of all cells contain foci; for F-G-ΔNLS-L-TM and F-G-NLS-ΔL-TM the numbers are 35% and 43%, respectively (graph, fig.4b). If we reduce the analysis to dividing cells only, then the fraction of cells containing foci is greater: 52% for F-G-NLS-L-TM, 50% for F-G-ΔNLS-L-TM reporter and 68% for F-G-NLS-ΔL-TM (graph, fig.4d). The distribution of foci between mother and bud is similar for all the reporters. Typically, from all cells containing foci around 20-30% are in non-dividing cells, 40% are bud-biased, 20-30% mother-biased and 10-15% unbiased (graph fig.4c). The average number of foci per cell is 1.5 for all the tested reporters. Clearly, in our experimental set up we have two Nsp1 trap positions – one within the pore, and one located outside of it, presumably in the cytosolic foci. The F-G-NLS-ΔL-TM reporter traps at the cytoplasmic Nsp1 more efficiently, i.e. when compared to reporters containing the flexible linker region. The majority of the reporter molecules are located within the NE, though, as can be seen in the microscopy pictures. Even in case of the F-G-NLS-ΔL-TM reporter only 50% of all cells with the NE staining contains foci at all. We thus conclude that trapping at the pore fraction of Nsp1 might be a dominating phenomenon, at least for the reporters with the long flexible linker. So far, we have seen that regardless the presence of an NLS or a linker most of the molecules trap at the NE. We already mentioned the difficulty of explaining trapping of the F-G-ΔNLS-L-TM at Nsp1, even more difficult is explaining the trapping of the F-G-NLS-ΔL-TM as the residual linker is not long enough to reach out through the scaffold of the NPC to the center. One explanation that we sought to investigate is that the reporters are prevented from inserting into the membrane, or even extracted from it due to interaction with Nsp1. We have already tested the membrane insertion state of this reporter in the wild type strain (previous chapter) and concluded that theprotein behaves as an integral membrane protein in the salt and detergent solubilization assay. We now check the insertion state of Nsp1-trapped F-G-NLS-ΔL-TM reporter, in comparison with Nsp1-trapped F-G-NLS-L-TM. As shown in the Western Blots of fig.5 both reporters solubilize in the Triton X-100 buffer, as control integral membrane proteins do (Pma1, Dpm1), despite the presence of rapamycin. Both of reporters are

Chapter 5

84

therefore membrane-embedded while trapped on Nsp1. As mentioned before, the F-G-NLS-ΔL-TM reporter has a higher affinity for the foci localized Nsp1 than the F-G-NLS-L-TM reporter. We do not see any increase in the salt soluble fraction in case of this reporter. We thus conclude that the reporter without linker is likely also embedded in the membrane, even when trapped at Nsp1. This fact reinforces the observation that the Nsp1 cytoplasmic foci are located in the vicinity of the ER system [3]. Surprisingly, we have also noted a change in Nsp1 solubility in these experiments. In non-trappedconditions Nsp1 solubilizes most efficiently in the buffer with 1M NaCl, which is typical for proteins that are peripherally associated with the membranes. However, upon rapamycin treatment Nsp1 solubilization by 1M NaCl is less efficient (Western blots fig.5, compare lanes P and S for 1M NaCl). Clearly, interaction of Nsp1 with our reporters keeps the nucleoporin more tightly in the membrane environment.

FIG.5. BOTH F-G-NLS-L-TM AND F-G-NLS-ΔL-TM REPORTERS SPAN THE MEMBRANE WHEN TRAPPED AT THE NSP1FRB. Western Blots showing the results from salt and detergent solubilization of crude yeast membranes fractions. Membrane extractions were performed in trapped (+rap) and in non-trapped (+DMSO) conditions. Crude membranes are incubated in buffer with 1M NaCl, 1%Triton or control buffer used for the cell lysis (in this case 20mMTris with 150mM NaCl). The degree of reporter solubility is assessed by comparison of the fraction that is remained in the pellet (P) upon incubation with the extraction buffer with the fraction that is extracted to the supernatant (S). In these conditions polytopic integral membrane proteins Pma1 and Dpm1 display the most intense bands in the solubilized fraction when treated with 1% Triton X-100 (left panel, two bottom rows). In case of Nsp1FRB, a protein peripherally associated with the membranes, the most intense bands are observed in the solubilized fraction when treated with 1M NaCl, but there is also a clear solubilized fraction in the control conditions with 150mM NaCl (left and right panels, non trapped +DMSO). In case of both reporters the most intense band of the solubilized fractions appears when cells are treated with 1% Triton X-100.


85

Chap

ter 5

Another question is if the function of Nsp1, both incorporated in the NPC and cytoplasmic, is affected by the trapping of membrane-embedded reporters. We assessed the pore function of Nsp1 by evaluating transport of soluble cargo via the NPC. To do that we expressed two proteins in the Nsp1FRB strain – mCherry with classical NLS signal and a PrA-F-G-NLS-L-TM reporter (fig. 6 and published before in [2]). As seen in fig.6 the N/C ratio of the mCherry-cNLS is the same (around 3) for both trapped and non-trapped conditions. Next, we investigated the function of the fraction of Nsp1 that is localized to cytoplasmic foci. This pool of Nsp1 is involved in the distribution NPCs between the mother cell and the bud during cell division [3,4]. If the Nsp1 pool located in the cytoplasmic foci is inhibited, the NPCs will not be distributed between the cells. We determined the fraction of dividing cells, in which buds have no fluorescence signal in the NE under trapped conditions. In case of F-G-ΔNLS-L-TM and F-G-NLS-ΔL-TM 55% and 49% of the cells have this phenotype,

Fig.6. Soluble transport across the NPC is not affected due to PrA-F-G-NLS-L-TM trapping at the Nsp1FRB. A) Fluorescence microscopy images showing PrA-F-G-NLS-L-TM and cNLS-mCherry reporter co-expressed in Nsp1FRB strain. Scale bars: 2 μm. Green – PrA-F-G-NLS-L-TM; red – cNLS mCherry. B) The ratio of the mCherry fluorescence intensity in the nucleus to the cytoplasm in the trapped (+ rap, column with black diagonal lines, n=18) and non-trapped conditions (no rap, gray column, n=14).

cNLS mCherry

Chapter 5

86

while for the F-G-NLS-L-TM this amounts to 26% (fig.7a). Colombi et al. also observed that inhibition of Nsp1 in the cytoplasmic foci causes abnormal location of the mother nuclei before cell division. In 25% of the cells the mother nuclei positions distal to the budneck, instead of locating in the vicinity to it [3]. We therefore estimated the fraction of cells with mispositioned mother nuclei in our assay and compared the results with cells not treated with rapamycin. We find cells with mispositioned nuclei during division even without addition of rapamycin. The percentage of mother cells displaying this phenotype dropped after rapamycin incubation from roughly 20% to 10% in case of F-G-NLS-L-TM and F-G-NLS-ΔL-TM (fig.7b). In contrast, we noted an increase of this type of cells upon rapamycin treatment when using F-G-ΔNLS-L-TM (from 10 to 20%, also fig.7b). These data suggest that expression of the reporters alone induces a mispositioning phenotype, making the interpretation of the additive effect of Nsp1 trapping to the different reporters difficult.

Fig.7. No signs of cytoplasmic foci Nsp1 inhibition were observed upon binding between the reporters and Nsp1 A) Percentage of buds with no fluorescence signal in the NE after incubation with rapamycin (+rap, columns with black diagonal stripes; n=62 for F-G-NLS-L-TM; n=80 for F-G-ΔNLS-L-TM; n=103 for F-G-NLS-ΔL-TM) and in the control conditions (no rap, grey columns; n=49 for F-G-NLS-L-TM; n=63 for F-G-ΔNLS-L-TM; n=85 for F-G-NLS-ΔL-TM). B) Percentage of mother cells with the nuclei mis-positioned from the budneck during cell division after incubation with rapamycin (+rap, columns with black diagonal stripes; n=62 for F-G-NLS-L-TM; n=80 for F-G-ΔNLS-L-TM; n=103 for F-G-NLS-ΔL-TM) and in the control conditions (no rap, grey columns; n=49 for F-G-NLS-L-TM; n=63 for F-G-ΔNLS-L-TM; n=85 for F-G-NLS-ΔL-TM).


87

Chap

ter 5

DISCUSSION

In order to study the import characteristics of membrane proteins into the inner nuclear membrane we have developed an assay to trap them during their passage through the NPC. We have reported trapping on all of the tested positions, namely FRB-tagged versions of Nsp1, Nup53, Nup59 and Nup170. However, discovery of a pool of Nsp1 that is not incorporated into the pore made us reevaluate part of the data. We have noticed foci characteristic of a cytoplasmic pool of Nsp1, similar to what was reported before by Colombi et al [3]. Clearly, next to trapping our reporters in the NPC, we also attract them to this cytoplasmic pool of Nsp1. In addition we find trapping at Nsp1 of the F-G-NLS-ΔL-TM and F-G-ΔNLS-L-TM reporters. Thus, trapping occurs even if the reporters lack the flexible linkers or the nuclear localization signal. Below, I will discuss how our previous interpretation of the Nsp1 trap experiments may have been affected by presence of a cytoplasmic pool of Nsp1. In addition I explore what new lessons we learn from the data. The Nsp1 in our experimental setup is C-terminally tagged with FRB, and the functionality of it could be questioned. Indeed, formation of cytoplasmic foci is a phenotype that has previously been observed in a temperature sensitive mutant of Nsp1 [13]. However, we believe our Nsp1 variant is completely functional. Firstly, an Nsp1 deletion is not viable, which is not the case for our Nsp1 trap strain. Secondly, our analysis shows no decrease in nuclear transport rates for soluble and membrane cargo (data not shown). Finally, experiments from Colombi et al., in which the cytoplasmic foci Nsp1 fraction was reported for the first time, were also performed on strains where Nsp1 was C terminally tagged with either GFP or FRB-GFP. In addition to testing C terminal fusions of Nsp1, the authors confirmed the appearance of non tagged Nsp1 in cytoplasmic foci in a wild type strain using immunofluorescence microscopy and immunoEM [3]. The question, how many of the reporter molecules trap at Nsp1 in the cytoplasmic foci instead of Nsp1 in the pore, is difficult to answer. The majority of the fluorescence signal that we observe is localized to the NE. Even in case of F-G-NLS-ΔL-TM reporter, which has a greater affinity to the cytoplasmic Nsp1 pool comparing with other reporters, foci were observed in 50% of the cells only. The percentage of cells with visible foci after rapamycin treatment is slightly lower in case of F-G-NLS-L-TM than with F-G-ΔNLS-L-TM or F-G-NLS-ΔL-TM. Also, the foci are bigger and easier to detect for F-G-NLS-ΔL-TM than for the other two reporters. The h2NLS-linker is a very strong signal for nuclear localization, so it is possible that this reporter is mostly trapped in the pore due to rapid interaction with importins and immediate delivery to the nucleus. The F-G-ΔNLS-L-TM and F-G-NLS-ΔL-TM reporter are not actively (or at least less efficiently) transported to the nucleus, so they are likely exposed to cytoplasmic Nsp1 for a longer period time, especially given the fact that cytoplasmic

Chapter 5

88

Nsp1 foci are likely associated with the ER. The interpretation of these data in terms of transport mechanism for membrane proteins will largely depend on the actual ratio of Nsp1 localizing to the cytoplasmic foci to Nsp1 in the pore. The larger the non-pore fraction is, the less confidence one has in NPC trapping. The current assumption in the field is that the pore and foci pools of Nsp1 are physically separated from each other, one being embedded deeply in the structure of the pore and the other co-localizing with the ER system. In this light, our finding of the F-G-NLS-ΔL-TM reporter trapping at the pore fraction of Nsp1 is confusing. This reporter is not long enough to reach out to the center of the pore, where Nsp1 is located. Our FRB tag is located at the C terminus of Nsp1, so at the domain that anchors the protein to the structure of the pore. We estimate that the distance between the FRB molecule and the pore membrane should be around 15-20nm. The linker domain of the F-G-NLS-L-TM reporter has a Stokes radius of 4.5nm [2]. Given the unfolded nature of this domain, we estimate it’s maximum length at ~20 nm, which is based on the peptide bond length. The GFP and FKBP molecules are fused to the reporter on the N terminal side from the NLS. The Stokes radius of YFP-FKBP was experimentally estimated at 3.2nm [14], and the Stokes radius of our 2xFKBP-GFP is likely slightly larger. All together the distance between the F-G-NLS-L-TM transmembrane segment and the N terminal FKBP molecule equals around 27nm, if the linker domain is stretched. In case of the F-G-NLS-ΔL-TM reporter, in which the NLS is separated from the transmembrane segment by 7 amino acids, this distance is ~7nm, which is not enough to reach out to the anchor point of Nsp1FRB. Most probably competition exists between two pools of Nsp1 for binding of the reporter and we do see that the F-G-NLS-ΔL-TM reporter has a higher affinity for the foci fraction of Nsp1 compared to other reporters. The preference for the cytoplasmic pool is however not strong enough to completely eliminate binding at the NPC. Trapping of this reporter at the NPC could be explained if the reporter is extracted from the membrane prior to traversing the NPC and binds the pore Nsp1FRB as a soluble protein. However, we show that the F-G-NLS-ΔL-TM reporter is embedded in the membrane when trapped at Nsp1. There are three possibilities explaining trapping of this reporter at the NE. Firstly, the NPC structure might be much more dynamic than previously assumed. In this case Nsp1 may not always be centrally located in the pore and could just temporally meet the membrane-embedded reporter and form a complex with it. Secondly, the phenotype, which we interpret as NPC trapping, might in fact be trapping at the NPC assembly intermediates. Yeasts undergo a closed mitosis, which implies that new NPCs have to be inserted to the intact NE, in a process similar to interphase de novo NPC assembly in vertebrates. Nuclear pore formation is a complicated molecular process in which membrane remodeling processes and assembly of the protein scaffold have to be coordinated together in order to maintain the nucleus-cytoplasm barrier (reviewed


89

Chap

ter 5

in [15]). One aspect of this process is extremely significant in the analysis of our trapping data, namely – when and where are the Nup subcomplexes assembled? It is known that nucleoporins are inserted into the NE from both the cytoplasmic and nucleoplasmic side during pore assembly in a cell free Xenopus system [16]. In living yeast cells accumulation of nucleoporins in cytoplasmic foci and in the nuclear sites adjacent to the INM was observed when pore assembly was inhibited by Nup170 and Nup157 depletion [17]. At these nuclear sites no pore membrane was observed, i.e., the INM and ONM were separated. The presence of Nups in these structures was detected by the mAb414 antibody, which binds several FG Nups, also Nsp1 [17]. These data suggest that nucleoporin subcomplexes can be assembled in the vicinity of the membrane prior to pore formation. It is interesting to speculate, that the cytoplasmic Nsp1 foci observed by [3] are also first assembled in the NE and later on send to the bud. If that possibility is true, then our F-G-NLS-ΔL-TM reporter is likely trapped at Nsp1 assembly intermediates for NPC or the cytoplasmic foci pool rather than mature NPCs. Thirdly, it is possible that FG-Nups are prone to aggregation [18], especially in molecular crowding conditions [19]. Recent simulations revealed that adjacent Nsp1 molecules form frequently cross-linked brush like bundles [20]. A scenario, in which our F-G-NLS-ΔL-TM reporter binds the Nsp1 molecules located in the cytoplasmic foci and subsequently pulls them towards the NPC, where two pools of Nsp1 aggregate, cannot be excluded. To distinguish which of the scenarios is actually the case, one needs to know more about where the Nsp1 foci in the cytoplasm come from. In our experiments, where crude yeast membranes were solubilized with salt or detergent, we have seen a change in Nsp1 behavior after rapamycin treatment. Nsp1, as a protein peripherally associated with the membranes, normally solubilizes in the buffer with 1M NaCl, but upon trapping of the reporters it becomes less easily extractable from the membranes. The biochemical properties of this molecule clearly change due to complexing our reporters, so it is unlikely that the function is not affected at all. Although, the effect of complex formation with the reporters is probably minor, as the cells continue to divide, we decided to test the effect of trapping of the reporter on nuclear transport, NPC inheritance and nuclear positioning. The transport rates of soluble cargo through the NPC were not changed after the cells were incubated with rapamycin. We have also compared the phenotype of cells where the formation of cytoplasmic foci was inhibited [3] with our cells after exposure to rapamycin. Upon inhibiting the formation of cytoplasmic foci of Nsp1 the transfer of NPCs from the mother cell to the bud is blocked. In our experiments we have seen that roughly a quarter of the budding cells expressing F-G-NLS-L-TM and half of the budding cells expressing F-G-ΔNLS-L-TM or F-G-NLS-ΔL-TM reporters exhibit no NE staining in the bud. It is difficult, however, to provide a suitable control for this experiment, as without rapamycin our GFP labeled reporters will not associate with the pores, but will distribute evenly over the NE. As a different way to assess the effect of trapping the

Chapter 5

90

reporters at the Nsp1 foci in the cytosol we estimated the amount of mother cells with the nuclei positioned to the distal end of the cell instead of the budneck during division. Colombi et al. observed this phenotype for roughly 25% of the cells where cytoplasmic foci formation was inhibited [3]. As we haven’t seen any clear rapamycin dependent trend (the amount of mother cells with mispositioned nuclei dropped for 2 reporters and increased for another reporter after rapamycin treatment), we cannot provide an unequivocal answer if Nsp1 in the foci is functional. In summary, our experiments designed to study the molecular pathway of membrane proteins through the NPC turned out to report on another molecular phenomenon as well. In the Nsp1FRB strain background, next to trapping at the NPC, the reporters are also caught in the cytoplasmic foci. This must mean that the cytoplasmic foci are close to the ER. We see that different reporters have a different affinity for the foci-localized pool of Nsp1, which likely reflects the competition between cytoplasmic foci and pore Nsp1 pools for binding. Without proper estimation of binding sites in the cytoplasmic foci and the pore we cannot say if the trapping assay reports solely on transport through the NPC.


91

Chap

ter 5

REFERENCES





5. Gruber S, Arumugam P, Katou Y, Kuglitsch D, Helmhart W, Shirahige K, Nasmyth K: Evidence that Loading of Cohesin Onto Chromosomes Involves Opening of Its SMC Hinge. Cell 2006, 127:523–537.


7. Janke C, Magiera MM, Rathfelder N, Taxis C, Reber S, Maekawa H, Moreno-Borchart A, Doenges G, Schwob E, Schiebel E, et al.: A versatile toolbox for PCR-based tagging of yeast genes: new fluorescent proteins, more markers and promoter substitution cassettes. Yeast 2004, 21:947–962.

8. Ries J, Kaplan C, Platonova E, Eghlidi H, Ewers H: A simple, versatile method for GFP-based super-resolution microscopy via nanobodies [Internet]. Nature Methods 2012, 9:582–584.

9. Mund M, Kaplan C, Ries J: Chapter 14 - Localization microscopy in yeast. In Quantitative Imaging in Cell Biology. Edited by Waters JC, Wittman T. Academic Press; 2014:253–271.

10. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al.: Fiji: an open-source platform for biological-image analysis. Nature Methods 2012, 9:676–682.

11. Grund SE, Fischer T, Cabal GG, Antunez O, Perez-Ortin JE, Hurt E: The inner nuclear membrane protein Src1 associates with subtelomeric genes and alters their regulated gene expression. The Journal of Cell Biology 2008, 182:897–910.


13. Nehrbass U, Kern H, Mutvei A, Horstmann H, Marshallsay B, Hurt EC: NSP1: A yeast nuclear envelope protein localized at the nuclear pores exerts its essential function by its carboxy-terminal domain. Cell 1990, 61:979–989.

14. Lin Y-C, Niewiadomski P, Lin B, Nakamura H, Phua SC, Jiao J, Levchenko A, Inoue T, Rohatgi R, Inoue T: Chemically inducible diffusion trap at cilia reveals molecular sieve–like barrier. Nat Chem Biol 2013, 9:437–443.

15. Rothballer A, Kutay U: Poring over pores: nuclear pore complex insertion into the nuclear envelope. Trends in Biochemical Sciences 2013, 38:292–301.

16. D’Angelo MA: Nuclear Pores Form de Novo from Both Sides of the Nuclear Envelope. Science 2006, 312:440–443.

Chapter 5

92

17. Makio T, Stanton LH, Lin CC, Goldfarb DS, Weis K, Wozniak RW: The nucleoporins Nup170p and Nup157p are essential for nuclear pore complex assembly. The Journal of Cell Biology 2009, 185:459–473.

18. Halfmann R, Wright JR, Alberti S, Lindquist S, Rexach M: Prion formation by a yeast GLFG nucleoporin. prion 2012, 6:391–399.

19. Milles S, Bui KH, Koehler C, Eltsov M, Beck M, Lemke EA: Facilitated aggregation of FG nucleoporins under molecular crowding conditions. EMBO Rep. 2013, 14:178–183.

20. Gamini R, Han W, Stone JE, Schulten K: Assembly of Nsp1 Nucleoporins Provides Insight into Nuclear Pore Complex Gating. PLoS Comput Biol 2014, 10:e1003488.

Chapter 6 Traffic to the inner membrane of the nuclear envelope

Justyna K. Laba, Anton Steen, Liesbeth M. Veenhoff

Published in Current Opinion in Cell Biology 2014.

Chapter 6

94

ABSTRACT

Past research has yielded valuable insight into the mechanisms that regulate the nuclear transport of soluble molecules like transcription factors and mRNA. Much less is known about the mechanisms responsible for the transportation of membrane proteins to the inner membrane of the nuclear envelope. The key question is: does the facilitated transport of integral inner membrane proteins exist in the same way as it does for soluble proteins and, if so, what is it used for? Herein, we provide an overview of the current knowledge on traffic to the inner nuclear membrane, and make a case that: a) known sorting signals and molecular mechanisms in membrane protein biogenesis, membrane protein traffic and nuclear transport are also relevant with respect to INM traffic; and b) the interplay of the effects of these signals and molecular mechanisms is what determines the rates of traffic to the INM.


95

Chap

ter 6

The nuclear envelope (NE) is a specialized area of the endoplasmic reticulum (ER). It is composed of two membranes, the inner and the outer nuclear membrane (INM and ONM), which come together in places where Nuclear Pore Complexes (NPCs) are embedded. In many eukaryotes, a proteinaceous surface, namely the nuclear lamina, underlies the INM. The perinuclear space in between the two membranes is continuous with the ER lumen. The ER, ONM and INM are also continuous, but have distinct functions and sets of transmembrane proteins. Assigning proteins as true INM residents is problematical for multiple reasons, ranging from technical difficulties in microscopically resolving their localization in the INM or ONM, to biological reasons such as their cell type specificity [1]. Bioinformatic predictions are difficult to make, as only a few domains specific to INM proteins have been identified, such as the LEM (for Lap1-emerin-MAN1) and SUN (for Sad-Unc-84 homology) domains for which structures are available [2,3]. A decade ago, the first proteomic studies aimed at identifying putative INM proteins were performed [4,5] but to date, only a relatively small number of these proteins have been both well-characterized and proven to be enriched in the inner membrane compared to the outer membrane and ER. The importance of the correct trafficking and function of INM proteins is clear from numerous examples of the roles played in the development of nuclear envelopathies and cancer. Accordingly, the lamina-associated polypeptide 2, Lap2β, is over-expressed in digestive tract cancers [6]. Mutations in the lamin B receptor, LBR, cause both Greenberg dysplasia, a major disease leading to aberrant embryonic development [7], or Pelger-Huet anomaly [8]. Laminopathies are often linked to mutations in lamin A, but recent studies show that the mistargeting of INM proteins could be causative of the disease phenotypes [9,10]. For example, Hutchinson-Gilford Progeria Syndrome (HGPS), a serious accelerated ageing disease, is caused by a dominant de novo mutation in lamin A that results in the accumulation of progerin, which is a farnesylated lamin A variant. In HGPS cells, the levels of SUN1 in the INM are increased [9,11], and knocking-down SUN1 alleviates cellular senescence [9]. Similarly, nuclear deformation and cell survival are rescued by SUN1 knock-down in mice cells lacking lamin A or carrying progerin-like mutations [9]. Over the years, multiple mechanisms of INM protein targeting have been proposed, involving a variety of potential sorting signals. Earlier work suggested that the interplay between multiple signals is required for the efficient targeting of INM proteins [12,13]. Clearly, there will be multiple signals encoded on a specific membrane protein to guide its biogenesis and targeting. These signals may encode information for: insertion into the lipid bilayer, cytosolic subcellular sorting to the different membrane compartments, and nuclear transport. For each of these categories (‘molecular toolboxes’), short descriptions of the molecular signals and mechanisms are presented in Figure 1. Table I contains an overview of some of the better studied integral membrane proteins that are enriched in the INM in Saccharomyces cerevisiae and humans. We have

Chapter 6

96

Fig.1. The sorting of integral inner membrane proteins is an add-up of known principles of membrane protein biogenesis, cytosolic subcellular traffic and nuclear transport.

Toolbox I: Protein insertion Membrane protein integration into the lipid bilayer is a facilitated process. There are two well-characterized insertion systems that are conserved from yeast to man: the Sec61 system and the GET pathway (reviewed in [14,58]). The Sec61 system translocates soluble proteins and membrane proteins with single (monotopic) and multiple transmembrane spanning segments (polytopic). The current data supports that polytopic membrane proteins are inserted cotranslationally by the Sec61 system. The GET system evolved for the specialized post-translational insertion of tail anchored proteins, which are proteins with a single transmembrane spanning segment at their C-terminus and a short lumenal tail. Small monotopic membrane proteins may also be inserted posttranslationally. Prediction of the topology of membrane proteins is based on hydrophobicity profiles and the characteristic features of the regions flanking the transmembrane segments. E.g., if the region N-terminal of the transmembrane domain is long or positively charged (positive-inside rule), then it is likely extralumenal. The translocation of the N-terminus to the lumen is likely when positive charges are lacking, when the preceding region is not well folded and the hydrophobic sequence is long. Besides the Sec and GET pathways, other insertion machineries also exist, for example, in yeast, the Sec61 homolog, Ssh1, and the Sec63 complex. The question of post- or cotranslational insertion could be relevant to traffic of INM proteins, but has been little studied. In particular post-translational insertion via the GET pathway could occur at the INM post nuclear import, in which case the transport occurs as a soluble chaperoned protein.

also sorted into three molecular toolboxes the plethora of targeting information that has been experimentally validated.


97

Chap

ter 6

Toolbox I: Protein insertion Membrane protein integration into the lipid bilayer is a facilitated process. There are two well-characterized insertion systems that are conserved from yeast to man: the Sec61 system and the GET pathway (reviewed in [14,58]). The Sec61 system translocates soluble proteins and membrane proteins with single (monotopic) and multiple transmembrane spanning segments (polytopic). The current data supports that polytopic membrane proteins are inserted cotranslationally by the Sec61 system. The GET system evolved for the specialized post-translational insertion of tail anchored proteins, which are proteins with a single transmembrane spanning segment at their C-terminus and a short lumenal tail. Small monotopic membrane proteins may also be inserted posttranslationally. Prediction of the topology of membrane proteins is based on hydrophobicity profiles and the characteristic features of the regions flanking the transmembrane segments. E.g., if the region N-terminal of the transmembrane domain is long or positively charged (positive-inside rule), then it is likely extralumenal. The translocation of the N-terminus to the lumen is likely when positive charges are lacking, when the preceding region is not well folded and the hydrophobic sequence is long. Besides the Sec and GET pathways, other insertion machineries also exist, for example, in yeast, the Sec61 homolog, Ssh1, and the Sec63 complex. The question of post- or cotranslational insertion could be relevant to traffic of INM proteins, but has been little studied. In particular post-translational insertion via the GET pathway could occur at the INM post nuclear import, in which case the transport occurs as a soluble chaperoned protein.

Toolbox III: Nuclear import. The NPCs are anchored where both the INM and ONM come together to form the highly curved pore membrane. Their overall architecture and function is broadly conserved from yeast to humans. There are also distinct differences between yeast and metazoan NPCs: the yeast NPCs are smaller in size and molecular weight and each of them has several unique components [59,60]. NPCs are composed of a scaffold of folded proteins that anchor the 8-fold rotational symmetric structure to the nuclear envelope membrane. A set of intrinsically disordered proteins, the FG-Nups, are anchored to the scaffold of the NPC, and are critical for the selectivity of the pore. For soluble proteins, the mechanisms of nuclear transport are well described [61]. Molecules may diffuse through the NPC passively (efflux/influx) and equilibrate between the cytosol and nucleus. Transport factors and the gradient of RanGTP-RanGDP across the nuclear envelope are required to specifically 'pump' proteins against a concentration gradient and to transport very large macromolecular complexes across the NPC. In these facilitated import and export reactions, soluble transport factors shuttle Nuclear Localization Signal (NLS)- or Nuclear Export Signal (NES)-containing proteins across the NPCs. The FG-Nups encode multiple phenylalanine and glycine (FG)-repeats that act as binding sites for the soluble transport factors. Direction to the transport reaction is given by the gradient of RanGTP-RanGDP across the nuclear envelope: in an import reaction, the transport factor dissociates from cargo in the presence of RanGTP, thereby releasing the cargo in the nucleus. In addition, retention mechanisms usually play a role in defining nuclear and cytosolic concentrations of soluble proteins. Retention mechanisms also play a major role in defining INM localization of membrane proteins. In addition, specifically in yeast, there is good evidence for the facilitated import of Heh1 and Heh2 resulting in accumulation in the INM. Alike for soluble proteins, traffic of these INM proteins depends on FG-Nups, Kaps and the gradient of RanGTP-RanGDP. The sorting signal is composed of an NLS and a long intrinsically disordered linker.

Toolbox II: Cytosolic subcellular traffic In general, membrane proteins may traffic through different subcellular compartments before they reach their destination, e.g. to be modified post-translationally. In addition, membrane proteomes are generally dynamic, e.g. there is a rapid exchange between the plasma membrane, pools of vesicles and the ER network. For the trafficking of membrane proteins to the different cellular membranes, signal sequences exist such as those for Golgi retrieval, ER retention, and peroxisome and mitochondrial targeting. The localization of a protein encoding multiple signals depends on the kinetics of the different trafficking routes. It may be a mistake to think of INM proteins as stable components of the INM; they may well also have a dynamic localization within the cell that is regulated by the interplay of the above sorting signals. Cases where cytosolic subcellular traffic are relevant for traffic of INM proteins are given in Table I.

EN ROUTE TO THE INMEn route to the INM: membrane insertion

The first step of the targeting process is the synthesis and insertion of the nascent polypeptide to the membrane environment (Figure 1, I) [14]. The two conserved insertion machineries, the Sec61 and GET (Guided-Entry of TA proteins) systems, are situated in the ER, including the ONM. An INM localized pool of Sec61 might exist [15] and the GET transmembrane components are small and may also passively reach the INM through the lateral channels of the NPC. Thus, in principle a post-

Chapter 6

98

translational mechanism where a chaperoned INM protein is first trafficked to the nucleus, after which it is membrane inserted, could be possible for membrane proteins that are posttranslationally inserted such as very small monotopic membrane proteins and tail anchored proteins. However, this has not been tested directly. Monotopic membrane proteins could be targeted at either of the two insertion machineries. From the proteins listed in Table I, only emerin and LAP2β are potentially inserted via the GET pathway. The small splice variants of Heh1 (helix-extension-helix-1)/Src1, known as Src1-small or Mps3 (monopolar spindle), SUN1 and SUN2 have larger lumenal domains and their insertion is probably facilitated by the Sec61 system. The polytopic membrane proteins, Heh1/Src1, Heh2 (helix-extension-helix-1), LEM2 (Lap1-emerin-MAN1-2), MAN1, LBR and nurim, are probably inserted by Sec61 co-translationally. Many INM proteins have relatively large N-terminal extralumenal domains, which often contain regions that have been proved to be relevant for trafficking (Figure 2). The early recognition of INM proteins, including as early as during translation, was first proposed for viral peptides and later for native INM proteins [16-18]. Here, a shorter isoform of importin-α was shown to both bind a nascent polypeptide chain predisposed for the INM at a stretch of positive charges located 5-8 residues from the transmembrane segment, and direct it to the translocon [16-18]. Whether this is a significant sorting event in Heh2 is unclear, as the absence of this sequence does not affect localization [19]. Accordingly, instead of a sorting sequence, this could be regarded as a manifestation of the positive inside rule guiding membrane insertion. We take that the early recognition of Heh2, being destined for the INM, is more likely to occur through the early binding of the yeast homolog of importin-α, Kap60, to the exceptionally high affinity NLS of Heh2 [20].

En route to the INM: cytosolic subcellular sorting

The localization of a protein encoding multiple signals depends on the kinetics of the different trafficking routes. Anything that disrupts this balance can cause a change in localization. For example, mitochondria have a separate system for tail anchored protein insertion that could potentially compete with the GET system inserting them into the ER [21]. Knowledge of the cytosolic subcellular sorting (Figure 1, II) of integral INM proteins is thus far limited, but SUN2 is a clear example of how elements of subcellular sorting between the ER and Golgi are important. This monotopic INM protein possesses an Arg-rich Golgi retrieval signal that is necessary for its INM localization [12]. Similar Arg-rich sequences are found in LBR, Lap2ß, emerin and LEM2, but their involvement in targeting has not yet been characterized. Changes in the concentrations of interaction partners can also disturb proper sorting. For instance, SUN2 was found in endosomes when Rab5, a small GTPase responsible for endosomal membrane fusion and complexing SUN2, was over-expressed


99

Chap

ter 6

[22]. Another example is the mislocalization of SUN1 to the Golgi that has been observed in mice lacking functional wild-type lamin A [9]. Lamin A is probably needed for the retention of SUN1, preventing it from travelling to the Golgi. Indeed, when the N-terminal lamin A-interacting domain of SUN1 is deleted, the SUN1 relocates from the NE to the Golgi [9]. A balance between subcellular localizations is also required for LBR, which has two separate functions: inside the nucleus it is responsible for regulating the structure of the NE, as illustrated by its role in maintaining the lobulated structure of granulocyte nuclei [8], but it also acts as a sterol reductase for which it has to be ER localized [7]. How the dual localization is controlled is presently unclear. Intriguing connections with plasma membrane localization also exist. Emerin, for instance, targets the plasma membrane in the heart tissue of some animals [23]. Also interesting is how the epidermal growth factor (EGF) receptor (EGFR) travels from the plasma membrane to the nucleus upon EGF binding. The receptor is endocytosed and travels through the Golgi and to the ER via COPI regulated retrograde vesicle trafficking [24,25]. The next steps include translocation to the nucleus and extraction from the membrane, although the order in which this happens is unclear. Future experiments need to validate and refine proposed models, and questions, such as how membrane extraction is triggered, need answers. En route to the INM: nuclear import Current models of the transport of INM proteins disagree significantly on the nature of energy dependence: is there or is there not an active energy dependent import that drives the accumulation of membrane proteins in the INM? When looking at soluble proteins like transcription factors, we mostly see that retention mechanisms, as well as the kinetics of import, export, influx and efflux, define their localization (Fig. 1, III). These kinetics can be adapted by modification or the shielding of import and export signals. Many membrane proteins in the INM are retained due to interactions with nuclear components, most notably lamins and chromatin and SUN-KASH interactions in the lumen, but there is now good evidence that this ‘selective retention’ is not the sole basis for their nuclear presence. An initial report on the energy (and temperature) dependence of INM protein import [26] suggested that ATP is used for NPC restructuring, e.g. by creating transient channels through which the proteins could travel. Later reports show that several INM proteins make direct or indirect use of the classical nuclear transport elements, including NLSs, Kaps and FG-nups. S. cerevisiae Heh1 and Heh2 and human SUN2 have confirmed NLS sequences [12,27], while others have predicted sequences [28]. Moreover, Heh1 and Heh2 localization is dependent on the transport factors Kap60 and Kap95 (yeast importin-ß), the RanGTP/RanGDP gradient, and a subset of FG-Nups [19,27]. In S. cerevisiae, a combination of an NLS and an intrinsically disordered (ID) linker (L) is required and is sufficient for INM targeting. This ‘NLS-L’

Chapter 6

100

Fig.2. Targeting signals in integral membrane proteins of the inner membrane of the nuclear envelope. The topology of some of the better studied integral membrane proteins of the INM. The red bar indicates part of the sequence that was shown experimentally to be important for the INM localization of the protein (references in Table I). LEM (for Lap1-emerin-MAN1), MAN (Heh/Man1 carboxy-terminal homology domain, CTHD) and SUN (for Sad-Unc-84 homology) domains are indicated. Hs Homo sapiens, Sc, Saccharomyces cerevisae, Rn Rattus norvegicus, Gg Gallus gallus.

motif targets an Heh2 transmembrane domain, a polytopic Sec61 transmembrane domain and a synthetic transmembrane domain composed of leucine alanine repeats to the INM. We propose that the ID linker facilitates binding to the transport factors and interactions with the FG-Nups [19,29]. Alternatively, or additionally, the combination of the strong NLS and the ID linker act earlier in the membrane protein biogenesis or traffic.


101

Chap

ter 6

Tabl

e I:

Sor

ting

sign

als i

n in

tegr

al IN

M p

rote

ins

Protein

Toolbox I elements:

Mem

brane protein insertion

*1 p

redi

ctio

n

Toolbox II elements:

Cytosolic protein sorting

Toolbox III elements: Nuclear im

port

machinery

Unclassified

Refs

Y E A S T

Mps

3 co

tran

slat

iona

l Sec

61 sy

stem

*1

-‐ I

ndir

ect d

epen

denc

e on

Kap

123,

Kap

95

and Ra

nGTP

-‐Ran

GDP

grad

ient

; pi

ggyb

ack

mec

hani

sm v

ia b

indi

ng to

hi

ston

e H

2Z.A

. -‐ N

ucle

ar re

tent

ion

[3

1,62

]

Heh

1/Sr

c1

-‐ Src

1 sm

all:

cotr

ansl

atio

nal S

ec61

*1

-‐ Ful

l len

gth

Heh

1: co

tran

slat

iona

l Se

c61*

1

-‐ N

LS, R

anGT

P-‐Ra

nGDP

gra

dien

t, Ka

p60,

Ka

p95,

Nup

170,

Nup

2 -‐ N

ucle

ar re

tent

ion

[1

9,27

]

Heh

2 -‐ C

otra

nsla

tiona

l Sec

61*1

; -‐ W

e in

terp

ret “

INM

sort

ing

mot

if” is

to

polo

gy in

dica

tor

-‐ N

LS, R

anGT

P-‐Ra

nGDP

gra

dien

t, Ka

p60,

Ka

p95,

Nup

170,

Nup

2, G

LFG

dom

ains

of

Nup

100,

Nup

57, N

up14

5 -‐ N

ucle

ar re

tent

ion

[1

7,19

,27]

H U M A N

SUN

1 co

tran

slat

iona

l Sec

61*1

-‐ Nuc

lear

rete

ntio

n Lo

caliz

atio

n de

pend

s on

farn

esyl

ated

pre

lam

in A

[1

1,32

,63]

SUN

2 co

tran

slat

iona

l Sec

61*1

Go

lgi r

etri

eval

sign

al

-‐ NLS

, im

port

in-‐α

, im

port

in-‐ß

, Ran

GTP-‐

RanG

DP g

radi

ent,

-‐ Nuc

lear

rete

ntio

n

SUN

2 m

obili

ty re

quir

es

ATP

[1

2,22

,64,

65]

Emer

in

Tail

anch

ored

pro

tein

, pos

sibl

y po

sttr

ansl

atio

nal i

nser

tion

by G

ET

path

way

*1

Subp

opul

atio

n in

pla

sma

mem

bran

e in

hea

rt ti

ssue

fr

om h

uman

, rat

and

mou

se

(sor

ting

sign

als u

nkno

wn)

-‐ Nuc

lear

rete

ntio

n Em

erin

mob

ility

re

quir

es A

TP

[23,

65]

LAP2

β Ta

il an

chor

ed p

rote

in, p

ossi

bly

post

tran

slat

iona

l ins

ertio

n by

GET

pa

thw

ay*1

-‐ N

ucle

ar re

tent

ion

[26,

66-‐6

8]

LEM

2 Co

tran

slat

iona

l Sec

61*1

-‐ Nuc

lear

rete

ntio

n

[6

9]

MAN

1 Co

tran

slat

iona

l Sec

61*1

-‐ Nuc

lear

rete

ntio

n

[70]

LB

R -‐ C

otra

nsla

tiona

l Sec

61*1

-‐ N

term

inal

dom

ain

prob

ably

co-‐

defin

es to

polo

gy;

-‐ “ IN

M so

rtin

g m

otif”

Dis

tinct

func

tions

at E

R an

d N

E (s

ortin

g si

gnal

s un

know

n)

-‐ Ran

GTP

depe

nden

t int

erac

tion

with

Im

port

inβ

(not

impo

rtin

-‐α d

epen

dent

) -‐ N

ucle

ar re

tent

ion

Mob

ility

of L

BR is

de

pend

ent o

n Ra

nGTP

an

d N

up35

[7,1

6,65

,71-‐

74]

Nur

im

-‐ Cot

rans

latio

nal S

ec61

*1

-‐ “I

NM

sort

ing

mot

if”

-‐ N

ucle

ar re

tent

ion

(but

not

to D

NA

and

not t

o la

min

s)

[7

5,76

]

Chapter 6

102

Consistent with a facilitated transport mechanism, large extralumenal domains are tolerated [20]. However, more importantly, using these mobile proteins it was shown that, upon blocking import, the protein leaks out from the INM to the ER. This demonstrates that INM accumulation is the result of fast import and slower efflux, and reflects energy driven accumulation. Facilitated NLS mediated import of proteins with large extralumenal domains has been reproduced with multipass transmembrane reporters based on Sec61 and RibU (as described in Chapter 4 of this thesis), which should resolve the discussion of whether the transmembrane segments are embedded in the membrane during transport. Moreover, Chapter 4 describes fractionation studies of Heh2-based reporters with systematically designed deletions of Heh2 domains. These experiments have proved that the Heh2-based reporters with 1 transmembrane segment are still embedded in the membrane. In the same chapter we have shown that Heh2-based 1 transmembrane reporter trapped at nucleoporin Nup170, fractionates with the cell membranes, which implicates that it is inserted in the membrane environment while in the pore. Having reinforced the aspect of the facilitated transport of these yeast INM proteins, we emphasize that retention also plays a role. Full length Heh1 and Heh2 have LEM domains, and their diffusion in the membrane is much slower than that of truncated versions without the LEM domain. This is consistent with them binding to nuclear structures [19]. Overall, as for soluble proteins, the localization of these INM proteins is defined by the kinetics of import, leakage and nuclear retention.

Nuclear import: NLS independent mechanisms

For INM proteins without predicted NLS sequences, other mechanisms for facilitated transport have been proposed, e.g. via FG repeats encoded on the INM proteins [30], or via a piggyback mechanism in which membrane proteins bind to a soluble NLS-containing protein. The latter mode of transport was proposed for Mps3, which binds histone H2Z.A [31]. Some of the INM proteins that are thought to localize due to retention may in fact make use of the piggyback import. Lamins come to mind as potential piggyback candidates. The current thinking is that lamins contribute to sorting by retaining INM proteins upon arrival at the INM, but a role in piggyback transport of INM proteins cannot be excluded until we measure where they first associate. For example, prelamin A may have such a role in targeting of SUN1 to the INM. In differentiating human myoblasts, farnesylated prelamin A accumulates in and recruits SUN1 to the NE [32]. Additionally, a type of lamin A, possibly the unprocessed or mature forms, prevents SUN1 from travelling to the Golgi [9]. Farnesylated prelamin A also interplays with SUN2 targeting in differentiating myoblasts. Here, the enrichment of SUN2 at the nuclear poles depends on farnesylated prelamin A [32]. Moreover, in patients with Mandibuloacral dysplasia with type A lipodystrophy (MADA), which is a rare disease caused by the accumulation of unprocessed prelamin A, SUN2 distribution


103

Chap

ter 6

in the NE is disorganized, which is rescued by drugs that reduce prelamin A farnesylation [10]. In conclusion, particularly in yeast, there is good evidence for the facilitated import of membrane proteins that results in accumulation in the INM. In human cells, there is no definitive answer as to whether facilitated transport alone can result in accumulation in the INM. Nevertheless, it is clear that retention mechanisms play a major role in both yeast and mammalian systems.

Putative NPC independent traffic

As discussed [33], NPC independent routes across the nuclear envelope, such as those used by viruses [34], may also be available to traffic native membrane proteins. For the replication of the Herpes Simplex virus, large nucleocapsids are formed in the nucleus, which have to pass the NE before their maturation in the cytosol. An NPC independent export model, namely nuclear egress, is currently accepted as an explanation for this phenomenon (reviewed in [35]). The nuclear localized capsids are enveloped by the INM and cross the perinuclear space as vesicles, which fuse with the ONM and release the capsid to the cytoplasm. The same mechanism is reported in Drosophila melanogaster for the export of ribonucleoprotein particles (RNPs), which are too large to pass the NPC [36]. Perinuclear granules have been observed in other cell types and species, so the nuclear egress might in fact be a conserved export mechanism. Nuclear egress has been hypothesized to be involved in the removal of nuclear protein aggregates [37]. Future studies will have to demonstrate if membrane proteins could exit the nucleus via any such egress pathway.

CHALLENGES WHEN STUDYING INM IMPORT Kinetics matter

Definitive proof of the existence of the facilitated transport of membrane proteins requires verification that import across the NPC is faster than efflux, as well as a demonstration that import is transport factor and Ran dependent. This requires methods that allow the direct assessment of transport kinetics through the NPC, distinct from the kinetics of diffusion in the INM and ONM. Single molecule tracking experiments would be uniquely suitable, but are thus far unexploited. Alternatively, it is possible to measure rates of bulk efflux or bulk import. Bulk efflux is measured in experiments that start with an accumulation in the INM and then follow the kinetics of equilibration after blocking facilitated import. The steady state accumulation levels together with the efflux kinetics reveal the kinetics of import. These measurements can only be obtained when the proteins of interest are freely diffusing and are not retained in either compartment. The absence of protein turnover over the measured time period is also critical. However, for all known INM proteins, the binding of nuclear localized proteins is an important retention mechanism which makes them

Chapter 6

104

unsuitable for bulk efflux measurements. As a consequence, truncated versions that lack retention signals, or even synthetic constructs encoding only the minimally required sorting signals [19,20], must be used for these studies.

Where do membrane proteins travel through the NPC?

Based on electron tomographs of metazoan NPCs [38-41], the most logical pathway of the extralumenal domains of INM proteins is along the pore membrane through the lateral channels (Figure 1). These channels are flanked on the cytoplasmic and nuclear sides by the proteins from the outer ring Y-shaped Nup84 or Nup107 subcomplexes in yeast and humans, respectively [42]. An approximately 10nm space is available between the membrane and this part of the NPC scaffold. More centrally in the NPC, the lateral channels are flanked by its integral membrane proteins, and here the passage seems to be more restricted. Unfortunately, high resolution tomographs are not available of S. cerevisiae. Membrane proteins have been proposed as passing through (a) the lateral channels, which seems likely judging from the tomographic images of NPCs, or (b) the central channel, which is likely considering the involvement of FG-Nups and Kaps. Three main uncertainties about the structure of the NPC are relevant here. Firstly, yeast and human pores may differ critically, and whereas (a) is largely supported by work regarding metazoans, (b) is mostly from work with respect to baker’s yeast. Secondly, whether the disordered FG-nups occupy the space in the lateral channels, and whether FG-nups facilitate karyopherin mediated traffic through the lateral channels, is unknown. Thirdly, NPCs are flexible structures in which the position of the 8-fold rotational symmetric units is variable [39]. At the impressive but still limited resolution available, it is uncertain whether small or temporary openings exist between the centre of the NPC and the lateral channels. Accordingly, to resolve the route(s) through the NPC, (even) better knowledge of its dynamic structure is required. The major goal of experiments described in this thesis was to find the answer to this question – where does yeast INM protein Heh2 pass the pore during import? In Chapter 2 we have presented data supporting the central channel as a passage point for the soluble domains of Heh2. We have shown that Heh2-based reporters with 1 transmembrane segment can get trapped at centrally located nucleoporin Nsp1. We then went ahead and further tested and validated this model using different approaches. A first point of discussion was if the transmembrane segment of the Heh2-based reporters were permanently membrane embedded during passage. Indeed, the 1 transmembrane reporter used in this experiments is a tail-anchored protein, and a posttranslational membrane insertion mechanism could result in the reporter passing the pore in a soluble state. In Chapter 3 we aimed at visualizing both ends of the Nsp1-trapped reporter using the high-resolution microscopy technique STORM to resolve the location of the transmembrane segment in the pore. We were able to visualize at


105

Chap

ter 6

high resolution the membrane proteins while trapped in the NPCs but failed to visualize both ends simultaneously due to technical challenges and limited time. We thus used a classical membrane extraction assay to study if Heh2-based reporters are membrane embedded in trapped and non-trapped conditions (Chapter 4) and proved that Heh2-based reporters are membrane embedded when trapped at the pore. In the meantime a new fraction of Nsp1 was discovered outside of the nuclear pore [43,44], which made us question the validity of the experiments where this trapping strain was used. In Chapter 5 we have investigated if Heh2-based reporters trap at the nuclear pore localized or cytoplasmic localized fraction of Nsp1 (or at both of them), but to unequivocally answer this question more information about the newly discovered cytoplasmic Nsp1 fraction is needed. Worth noting is that we resorted to unquestioned trapping position in the NPC for our studies regarding the transmembrane nature (nucleoporin Nup170, Nup59, Nup53, Chapter 4) so the conclusions about the membrane insertion status of pore trapped reporters are solid. Nup170 however is located closer to the pore membrane than Nsp1 and the Heh2-based reporters do trap at this location. In the same chapter other peripherally located nucleoporins were successfully used as trapping positions, namely Nup53 and Nup59. The physical route through the pore that Heh2 uses during import still awaits future discovery, however this new data point towards a route that might be closer to the pore membrane than we have initially anticipated. The experimental setup that we have used, based on trapping of highly expressed reporters on the FRB fused nuclear proteins may not be the first choice for investigating in vivo localization of proteins during transport. The trapping strains described in this thesis may still be used in combination with genomically tagged Heh2. The absolute protein abundance of Heh2 equals 1250 molecules per cell [45], which is less than the abundance of nucleoporins used in the trapping assay in this thesis (Nup170: 1560 molecules per cell; Nup53: 2060; Nup59: 2740 [45]). Also immunoprecipitation studies of Heh2 may bring clues relating to the NPC passage point. We expect that further development of high-resolution microscopy techniques, and especially single molecule tracking, may allow for visualization of native proteins during transport in future.

WHY INM TARGETING WOULD BE NEEDED

Recent studies have uncovered new exciting functions of integral membrane proteins residing in the INM, and while for some of these activities passive diffusion and selective retention is sufficient, for others a tighter control of protein localization could be expected. Passive diffusion may be enough for LAP2beta and MAN1, which have been shown to (redundantly) mediate the assembly of the NE [46]. High enrichment in the INM may possibly be required for proteins that play a role in NPC assembly into an intact NE. NPC assembly in the intact NE in yeast depends on Heh1

Chapter 6

106

and Heh2 [47], while Sun1 and an INM-localized pool of Pom121 play a role in NPC assembly in humans [48-52]. INM proteins can also directly contribute to the INM acting as a “transcription factor resting place” by sequestering transcription factors that illegitimately entered the nucleus and as such prevent transcription of target genes [53]. Functions related to chromatin anchoring to the nuclear periphery might depend on facilitated transport, as they require a higher level of regulation (recent reviews [54,55]). An analysis of cells with inverted chromatin architecture has provided interesting insights into this topic [56]. The heterochromatin of rod photoreceptor cells of nocturnal mammals is not located on the nuclear periphery, but is shifted to the nuclear interior. This phenotype occurs gradually during differentiation and is caused by the lack of the anchoring proteins LBR and lamin A/C in mature cells, whereas LBR is still present in the cells at early stages of differentiation. This sequential expression of the above-mentioned proteins during differentiation has also been observed in different mouse tissues, and has a potential effect on the expression of cell type specific genes. The deletion of LBR or lamin A in differentiating myotubes has the opposite effect: a lack of LBR increases the expression of muscle specific genes, while the loss of lamin A reduces it. There is no LBR or lamin A regulated effect on the expression level of the same genes in mature muscle cells. These results suggest that INM proteins act as heterochromatin tethers to regulate differentiation. Indeed, several INM proteins are able to reposition specific chromosomes and are restricted to certain tissues [57]. These observations strongly support the idea that INM proteins localize in the nucleus specifically to shape chromatin and regulate transcription, and do not enter the nucleus by chance and stay there due to an interaction with DNA. So, in addition to the regulation of the expression or turnover of INM proteins, facilitated import may also play an important role in tuning the proteome of the inner membrane.

CONCLUDING REMARKS

We asked the question as to whether the facilitated transport of integral inner membrane proteins exists in the same way as it does for soluble proteins, and, if so, what is it used for. We conclude that there is ample evidence that the facilitated import of integral membrane proteins exists in S. cerevisiae. Some may argue that facilitated import in yeast is a consequence of its closed mitosis and lack of lamins. However, the biological evidence of INM proteins directing chromosome localization and transcriptional regulation, as well as the presence of NLS sequences, suggests that facilitated transport is also present in humans. A better understanding of the transport of integral membrane proteins to the INM should go hand in hand with research aimed at uncovering new roles of INM proteins in chromatin organization and signal transduction in development, ageing and differentiation.


107

Chap

ter 6

REFERENCES:

1. las Heras de JI, Meinke P, Batrakou DG, Srsen V, Zuleger N, Kerr AR, Schirmer EC: Tissue specificity in the nuclear envelope supports its functional complexity. nucleus 2013, 4.

2. Laguri C, Gilquin B, Wolff N, Romi-Lebrun R, Courchay K, Callebaut I, Worman HJ, Zinn-Justin S: Structural characterization of the LEM motif common to three human inner nuclear membrane proteins. Structure/Folding and Design 2001, 9:503–511.

3. Sosa BA, Rothballer A, Kutay U, Schwartz TU: LINC complexes form by binding of three KASH peptides to domain interfaces of trimeric SUN proteins. Cell 2012, 149:1035–1047.

4. Dreger M, Bengtsson L, Schöneberg T, Otto H, Hucho F: Nuclear envelope proteomics: novel integral membrane proteins of the inner nuclear membrane. Proc. Natl. Acad. Sci. U.S.A. 2001, 98:11943–11948.

5. Schirmer EC: Nuclear Membrane Proteins with Potential Disease Links Found by Subtractive Proteomics. Science 2003, 301:1380–1382.

6. Kim H-J, Hwang S-H, Han M-E, Baek S, Sim H-E, Yoon S, Baek S-Y, Kim B-S, Kim J-H, Kim S-Y, et al.: LAP2 is widely overexpressed in diverse digestive tract cancers and regulates motility of cancer cells. PLoS ONE 2012, 7:e39482.

7. Clayton P, Fischer B, Mann A, Mansour S, Rossier E, Veen M, Lang C, Baasanjav S, Kieslich M, Brossuleit K, et al.: Mutations causing Greenberg dysplasia but not Pelger anomaly uncouple enzymatic from structural functions of a nuclear membrane protein. nucleus 2010, 1:354–366.

8. Hoffmann K, Dreger CK, Olins AL, Olins DE, Shultz LD, Lucke B, Karl H, Kaps R, Müller D, Vaya A, et al.: Mutations in the gene encoding the lamin B receptor produce an altered nuclear morphology in granulocytes (Pelger-Huët anomaly). Nat. Genet. 2002, 31:410–414.

9. Chen C-Y, Chi Y-H, Mutalif RA, Starost MF, Myers TG, Anderson SA, Stewart CL, Jeang K-T: Accumulation of the inner nuclear envelope protein Sun1 is pathogenic in progeric and dystrophic laminopathies. Cell 2012, 149:565–577.

10. Camozzi D, D’Apice MR, Schena E, Cenni V, Columbaro M, Capanni C, Maraldi NM, Squarzoni S, Ortolani M, Novelli G, et al.: Altered chromatin organization and SUN2 localization in mandibuloacral dysplasia are rescued by drug treatment. Histochem. Cell Biol. 2012, 138:643–651.

11. Haque F, Mazzeo D, Patel JT, Smallwood DT, Ellis JA, Shanahan CM, Shackleton S: Mammalian SUN Protein Interaction Networks at the Inner Nuclear Membrane and Their Role in Laminopathy Disease Processes. Journal of Biological Chemistry 2010, 285:3487–3498.



14. Shao S, Hegde RS: Membrane protein insertion at the endoplasmic reticulum. Annu. Rev. Cell Dev. Biol. 2011, 27:25–56.

15. Deng M, Hochstrasser M: Spatially regulated ubiquitin ligation by an ER/nuclear membrane ligase. Nature 2006, 443:827–831.

16. Braunagel SC, Williamson ST, Ding Q, Wu X, Summers MD: Early sorting of inner nuclear membrane proteins is conserved. Proc. Natl. Acad. Sci. U.S.A. 2007, 104:9307–9312.

Chapter 6

108

17. Liu D, Wu X, Summers MD, Lee A, Ryan KJ, Braunagel SC: Truncated Isoforms of Kap60 Facilitate Trafficking of Heh2 to the Nuclear Envelope. Traffic 2010, 11:1506–1518.

18. Saksena S, Shao Y, Braunagel SC, Summers MD, Johnson AE: Cotranslational integration and initial sorting at the endoplasmic reticulum translocon of proteins destined for the inner nuclear membrane. Proc. Natl. Acad. Sci. U.S.A. 2004, 101:12537–12542.


20. Meinema AC, Poolman B, Veenhoff LM: Quantitative Analysis of Membrane Protein Transport Across the Nuclear Pore Complex. Traffic 2013, 14:487–501.

21. Wattenberg B, Lithgow T: Targeting of C-terminal (tail)-anchored proteins: understanding how cytoplasmic activities are anchored to intracellular membranes. Traffic 2001, 2:66–71.

22. Liang Y, Chiu PH, Yip KY, Chan SY: Subcellular localization of SUN2 is regulated by lamin A and Rab5. PLoS ONE 2011, 6:e20507.

23. Berk JM, Tifft KE, Wilson KL: The nuclear envelope LEM-domain protein emerin. nucleus 2013, 4:298–314.

24. Du Y, Shen J, Hsu JL, Han Z, Hsu M-C, Yang C-C, Kuo H-P, Wang YN, Yamaguchi H, Miller SA, et al.: Syntaxin 6-mediated Golgi translocation plays an important role in nuclear functions of EGFR through microtubule-dependent trafficking. Oncogene 2013, doi:10.1038/onc.2013.1.

25. Wang Y-N, Wang H, Yamaguchi H, Lee H-J, Lee H-H, Hung M-C: COPI-mediated retrograde trafficking from the Golgi to the ER regulates EGFR nuclear transport. Biochem. Biophys. Res. Commun. 2010, 399:498–504.



28. Lusk CP, Blobel G, King MC: Highway to the inner nuclear membrane: rules for the road. Nat Rev Mol Cell Biol 2007, 8:414–420.

29. Meinema AC, Poolman B, Veenhoff LM: The transport of integral membrane proteins across the nuclear pore complex. nucleus 2012, 3:322–329.

30. Kerr AR, Schirmer EC: FG repeats facilitate integral protein trafficking to the inner nuclear membrane. Commun Integr Biol 2011, 4:557–559.


32. Mattioli E, Columbaro M, Capanni C, Maraldi NM, Cenni V, Scotlandi K, Marino MT, Merlini L, Squarzoni S, Lattanzi G: prelamin a mediated recruitment of sun1 to the ne directs nuclear positioning in human muscle. Cell Death and Differentiation 2011, 18:1305–1315.

33. Burns LT, Wente SR: Trafficking to uncharted territory of the nuclear envelope. Current Opinion in Cell Biology 2012, 24:341–349.

34. Kobiler O, Drayman N, Butin-Israeli V, Oppenheim A: Virus strategies for passing the nuclear envelope barrier. nucleus 2012, 3:526–539.

35. Mettenleiter TC, Müller F, Granzow H, Klupp BG: The way out: what we know and do not know about herpesvirus nuclear egress. Cell. Microbiol. 2013, 15:170–178.

36. Speese SD, Ashley J, Jokhi V, Nunnari J, Barria R, Li Y, Ataman B, Koon A, Chang Y-T, Li Q, et al.: Nuclear Envelope Budding Enables Large Ribonucleoprotein Particle Export during Synaptic Wnt Signaling. Cell 2012, 149:832–846.


109

Chap

ter 6

37. Rose A, Schlieker C: Alternative nuclear transport for cellular protein quality control. Trends in Cell Biology 2012, 22:509–514.

38. Elad N, Maimon T, Frenkiel-Krispin D, Lim RY, Medalia O: Structural analysis of the nuclear pore complex by integrated approaches. Current Opinion in Structural Biology 2009, 19:226–232.

39. Frenkiel-Krispin D, Maco B, Aebi U, Medalia O: Structural analysis of a metazoan nuclear pore complex reveals a fused concentric ring architecture. J. Mol. Biol. 2010, 395:578–586.

40. Maimon T, Medalia O: Perspective on the metazoan nuclear pore complex. nucleus 2010, 1:383–386.

41. Beck M, Lucić V, Förster F, Baumeister W, Medalia O: Snapshots of nuclear pore complexes in action captured by cryo-electron tomography. Nature 2007, 449:611–615.




45. Ghaemmaghami S, Huh W-K, Bower K, Howson RW, Belle A, Dephoure N, O’Shea EK, Weissman JS: Global analysis of protein expression in yeast. Nature 2003, 425:737–741.

46. Anderson DJ, Vargas JD, Hsiao JP, Hetzer MW: Recruitment of functionally distinct membrane proteins to chromatin mediates nuclear envelope formation in vivo. The Journal of Cell Biology 2009, 186:183–191.

47. Yewdell WT, Colombi P, Makhnevych T, Lusk CP: Lumenal interactions in nuclear pore complex assembly and stability. Molecular Biology of the Cell 2011, 22:1375–1388.

48. Yavuz S, Santarella-Mellwig R, Koch B, Jaedicke A, Mattaj IW, Antonin W: NLS-mediated NPC functions of the nucleoporin Pom121. FEBS Letters 2010, 584:3292–3298.

49. Doucet CM, Talamas JA, Hetzer MW: Cell cycle-dependent differences in nuclear pore complex assembly in metazoa. Cell 2010, 141:1030–1041.

50. Talamas JA, Hetzer MW: POM121 and Sun1 play a role in early steps of interphase NPC assembly. The Journal of Cell Biology 2011, 194:27–37.

51. Mitchell JM, Mansfeld J, Capitanio J, Kutay U, Wozniak RW: Pom121 links two essential subcomplexes of the nuclear pore complex core to the membrane. The Journal of Cell Biology 2010, 191:505–521.

52. Shaulov L, Gruber R, Cohen I, Harel A: A dominant-negative form of POM121 binds chromatin and disrupts the two separate modes of nuclear pore assembly. Journal of Cell Science 2011, 124:3822–3834.

53. Heessen S, Fornerod M: The inner nuclear envelope as a transcription factor resting place. EMBO Rep. 2007, 8:914–919.

54. Towbin BD, Meister P, Gasser SM: The nuclear envelope--a scaffold for silencing? Curr. Opin. Genet. Dev. 2009, 19:180–186.

55. Mekhail K, Moazed D: The nuclear envelope in genome organization, expression and stability. Nat Rev Mol Cell Biol 2010, 11:317–328.

56. Solovei I, Wang AS, Thanisch K, Schmidt CS, Krebs S, Zwerger M, Cohen TV, Devys D, Foisner R, Peichl L, et al.: LBR and Lamin A/C Sequentially Tether Peripheral Heterochromatin and Inversely Regulate Differentiation. Cell 2013, 152:584–598.

57. Zuleger N, Boyle S, Kelly DA, las Heras de JI, Lazou V, Korfali N, Batrakou DG, Randles KN, Morris

Chapter 6

110

GE, Harrison DJ, et al.: zuleger_schirmer_specific NETs can promote the location of chromosomes to and from the nuclear periphery. Genome Biology 2013, 14:R14.

58. Denic V: A portrait of the GET pathway as a surprisingly complicated young man. Trends in Biochemical Sciences 2012, 37:411–417.

59. Rout MP, Aitchison JD, Suprapto A, Hjertaas K, Zhao Y, Chait BT: The yeast nuclear pore complex: composition, architecture, and transport mechanism. The Journal of Cell Biology 2000,148:635–651.

60. Cronshaw JM, Krutchinsky AN, Zhang W, Chait BT, Matunis MJ: Proteomic analysis of the mammalian nuclear pore complex. The Journal of Cell Biology 2002, 158:915–927.

61. Aitchison JD, Rout MP: The Yeast Nuclear Pore Complex and Transport Through It. Genetics 2012, 190:855–883.

62. Friederichs JM, Gardner JM, Smoyer CJ, Whetstine CR, Gogol M, Slaughter BD, Jaspersen SL: Genetic analysis of Mps3 SUN domain mutants in Saccharomyces cerevisiae reveals an interaction with the SUN-like protein Slp1. G3 (Bethesda) 2012, 2:1703–1718.

63. Hasan S, Güttinger S, Mühlhäusser P, Anderegg F, Bürgler S, Kutay U: Nuclear envelope localization of human UNC84A does not require nuclear lamins. FEBS Letters 2006, 580:1263–1268.

64. Hodzic DM: Sun2 Is a Novel Mammalian Inner Nuclear Membrane Protein. Journal of Biological Chemistry 2004, 279:25805–25812.


66. Furukawa K: The Major Nuclear Envelope Targeting Domain of LAP2 Coincides with Its Lamin Binding Region but Is Distinct from Its Chromatin Interaction Domain. Journal of Biological Chemistry 1998, 273:4213–4219.

67. Burke B, Stewart CL: The Laminopathies: The Functional Architecture of the Nucleus and Its Contribution to Disease *. Annu. Rev. Genom. Human Genet. 2006, 7:369–405.

68. Furukawa K, Panté N, Aebi U, Gerace L: Cloning of a cDNA for lamina-associated polypeptide 2 (LAP2) and identification of regions that specify targeting to the nuclear envelope. The EMBO Journal 1995, 14:1626–1636.

69. Brachner A: LEM2 is a novel MAN1-related inner nuclear membrane protein associated with A-type lamins. Journal of Cell Science 2005, 118:5797–5810.



72. Soullam B, Worman HJ: The amino-terminal domain of the lamin B receptor is a nuclear envelope targeting signal. The Journal of Cell Biology 1993, 120:1093–1100.


74. Ma Y, Cai S, Lv Q, Jiang Q, Zhang Q, Sodmergen, Zhai Z, Zhang C: Lamin B receptor plays a role in stimulating nuclear envelope production and targeting membrane vesicles to chromatin during nuclear envelope assembly through direct interaction with importin beta. Journal of Cell Science 2007, 120:520–530.

75. Rolls MM, Stein PA, Taylor SS, Ha E, McKeon F, Rapoport TA: A visual screen of a GFP-fusion library identifies a new type of nuclear envelope membrane protein. The Journal of Cell Biology 1999, 146:29–44.

76. Hofemeister H: Analysis of the Localization and Topology of Nurim, a Polytopic Protein Tightly Associated with the Inner Nuclear Membrane. Journal of Biological Chemistry 2004, 280:2512–2521.

Appendix

Summary

118

NEDERLANDSE SAMENVATTING

In dit proefschrift hebben we het transport van membraaneiwitten tussen het cytoplasma en de kern in gistcellen bestudeerd. Door de jaren heen is er heel veel onderzoek gedaan om het importmechanisme van wateroplosbare eiwitten naar de kern te ontrafelen. Die route en de regels daarvan zijn nu goed bekend, maar er zijn veel twijfels gerezen over het mechanisme van import van de membraaneiwitten. In ons onderzoek hebben we voor de gist Saccharomyces cerevisiae gekozen als modelorganisme, omdat het een eenvoudige Eukaryoot is en de genetisch manipulatie eenvoudig is. De enige transportroute naar de kern gaat via het kernporiecomplex (Nuclear Pore Complex, NPC). De NPC is een grote gecompliceerde structuur van ongeveer 500 eiwitten, nucleoporines genoemd, die samen een ring vormen. De NPC zit in de kernenvelop (Nuclear Envelope, NE) op de plaats waar de binnen- en buitenmembraan samensmelten. De binnenzijde van de ring is de plaats waar al het transport gebeurt, maar die ruimte – het zogenoemde centrale kanaal - is niet leeg. Deze is gevuld met ongestructureerde eiwitten, de FG-Nups, die samen een soort slagboom vormen. Kleine moleculen, zoals water en ionen, drijven passief tussen de FG-Nups door. Ook kleinere oplosbare eiwitten kunnen passief door de NPC diffunderen, maar voor de grotere eiwitten zijn er gespecialiseerde importeiwitten nodig om de energiebarrière te overwinnen – de karyopherines. Voor de kern bestemde eiwitten hebben een speciale “kern-kaart” gecodeerd in hun sequentie – de Nuclear Localization Signal (NLS). De karyopherine bindt het voor de celkern bestemde eiwit in het cytoplasma via de NLS en faciliteert daarna het transport via de NPC door de FG-Nups korte tijd te binden. De gradiënt van RanGTP tussen het nucleoplasma en cytoplasma zorgt ervoor dat het transport in de juiste richting gaat. In de kern wordt het complex tussen het eiwit en de karyopherin verbroken door RanGTP en de karyopherine wordt terug getransporteerd naar het cytoplasma. De manier waarop membraaneiwitten getransporteerd worden is minder bekend. De energieafhankelijkheid van de nucleaire import van membraaneiwitten staat nog steeds ter discussie. Het eerste model van dit transport, het zogenaamde diffusie-retentie model, is op een passieve stroom van eiwitten door de NPC gebaseerd. Vervolgens binden de eiwitten kernstructuren, bijvoorbeeld chromatine, waardoor ze in de kern accumuleren. De energieafhankelijkheid van de nucleaire import van verschillende nucleaire eiwitten is later waargenomen, ons team heeft daaraan ook bijgedragen. In hoofdstuk 2 presenteren we een model van het energie-afhankelijke transport van het Heh2p eiwit van S. Cerevisiae; Heh2p heeft een lange ongestructureerde linker die zich uitstrekt van de membraan tot het centrale kanaal. Op die manier wordt de NLS gepresenteerd en wordt de karyopherin-FG-Nup interactie mogelijk gemaakt. We laten zien dat dit proces afhankelijk is van de lengte van de linker (maar niet van de aminozuursequentie), van de karyopherine Kap95 en van sommige

119

FG-Nups (Nup100, 145 en 57). We tonen aan dat het linkerdomain ongestructureerd is en tenslotte immobiliseren we de Heh2p reporter aan een FG-Nup van het centrale kanaal: Nsp1p. Echter, enkele jaren later bleek Nsp1p niet de beste keuze voor dit trapping experiment. Onze aarzeling erover begon toen Colombi et al. en Makio et al. een fractie van Nsp1p in het cytoplasma ontdekten. In hoofdstuk 5 hebben we de invloed van deze “cytoplasmic pool” van Nsp1p op onze resultaten bestudeerd. We concluderen dat het zonder een precieze kwantitatieve schatting van het aantal Nsp1p moleculen in het cytoplasma onmogelijk is om te bepalen of onze Heh2 reporter het Nsp1p in het cytoplasma of in de porie bindt. Desondanks vermoeden we dat de binding met de Nsp1p fractie in de kernenvelop een dominante rol speelt. Tegelijkertijd hebben we ernaar gestreefd om de moleculaire route van de Heh2p reporters via de NPC te ontrafelen door de trapping experimenten uit te voeren met andere nucleoporines - twee FG-Nups, Nup53p and Nup59p, en een structurele nucleoporine, Nup170p. Deze resultaten worden in hoofdstuk 4 gepresenteerd. De bovengenoemde nucleoporines liggen veel dichterbij het poriemembraan dan Nsp1p. Deze keer hebben we de trapping experimenten uitgevoerd met een grotere collectie van Heh2p reporters – een deel daarvan stroomt passief naar de kern en een deel wordt via een energie-afhankelijk proces geïmporteerd. We zien dat beide categorieën reporters aan Nup53p, Nup59p en Nup170p worden gebonden. De vraag of het actief en passief transport via verschillende routes door de NPC gaat blijft voorlopig onopgelost. De ontdekking dat nucleaire import van Heh2p energie-afhankelijk is, is door ons verder bevestigd. In hoofdstuk 6 van dit proefschrift hebben we beschreven dat de route van een membraaneiwit, vanaf het moment van synthese in het cytoplasma tot het moment dat het de eindbestemming in de cel bereikt, iets gecompliceerder is dan die van water-oplosbare eiwitten. De meeste membraaneiwitten hebben nog andere signaalsequenties naast NLS, dus een competitie tussen al de signalen kan de volgorde van moleculaire stappen beïnvloeden. Met die hypothese zijn we een onderzoek begonnen naar de vraag: kunnen membraaneiwitten met slechts één transmembraansegment en een zeer sterke NLS geïmporteerd worden naar de kern voordat de membraaninsertie plaatsvindt? In hoofdstuk 4 presenteren we een aantal experimenten, die bevestigen dat onze één-transmembraan reporters eerst in het membraan worden geinserteerd. We stellen dat het transport van membraaneiwitten altijd in de membraanomgeving gebeurt. Dat brengt ons terug tot een vraag: waar gaan de membraaneiwitten door de NPC langs? Zoals hiervoor uitgelegd gaat al het NPC transport door het centrale kanaal, omdat hier de binding tussen FG-Nups en karyopherines het transport faciliteert. De membraaneiwitten worden eerst in de membraan geinserteerd. Dus de route via het centrale kanaal lijkt onwaarschijnlijk – de membraanextractiestap zou te veel energie kosten. Cryoelectron tomography imaging van NPCs uit verschillende organismen heeft goede 3D-projecties van de NPC-structuur opgeleverd. In deze modellen zijn kleine

120

tunnels dichtbij de membraanzijde van de porie zichtbaar– de zogenoemde perifere kanalen. Hier zouden de oplosbare domeinen van de membraaneiwitten kunnen passen, maar de ruimte is zeer beperkt. Dat zou de grootte van de oplosbare domeinen waarschijnlijk beperken. Desondanks, de experimenten van hoofdstukken 2 en 4 laten zien dat eiwitten met oplosbare domeinen tot 150kDa de kern wel kunnen betreden en daarin accumuleren. We stellen dat die eiwitten met een actief, energie-afhankelijk mechanisme naar de kern worden geïmporteerd. Op deze manier overwinnen ze de potentiële diffusielimiet. In hoofdstuk 6 hebben we toegelicht dat verschillende transportmodellen gelijktijdig kunnen bestaan in de import van elk eiwit. Volledig Heh2p bevat bijvoorbeeld de retentiedomeinen die nucleaire structuren kunnen binden volgens de regels van het diffusie-retentie model. In onze set-up hebben we gekozen om die domeinen te verwijderen, want we wilden de kinetiek van NLS-afhankelijk transport meten. Het volledige Heh2p wordt waarschijnlijk geïmporteerd naar de kern op een efficiënte, energie-afhankelijke manier, maar blijft vervolgens in de kern dankzij de interacties met nucleaire binding partners. In hoofdstukken 2 en 4 hebben we aangetoond dat reporters zonder de retentiedomeinen mobiel zijn en uit de kern kunnen lekken als de actieve transportmachinerie uitgeschakeld is. We stellen dat een soortgelijke ingreep tussen het actieve transport en de retentiemodellen mogelijk ook relevant is voor andere eiwitten. In hoofdstuk 6 erkennen we dat aktief transport van membraaneiwitten waarschijnlijk vooral belangrijk is voor cellen met een gesloten mitose, en daardoor een altijd intacte kernenvelope, zoals in gist, maar ook voor post-mitotische cellen van Metazoën. In cellen van Metazoën die mitose ondergaan kunnen de membraaneiwitten in de binnenmembraan accumuleren terwijl de kernenvelop opnieuw wordt opgebouwd na de mitose.

121

ENGLISH SUMMARY

In this thesis we have investigated the mechanism of nuclear import of membrane proteins. The rules governing the transport of soluble molecules between the cytoplasm and the nucleus are well known. The transport is mediated by the Nuclear Pore Complexes (NPCs), which are complicated ring shaped protein assemblies embedded in the Nuclear Envelope (NE) where the inner and outer membranes meet. The protein building blocks of the NPC are called nucleoporins or nups. The inside of the NPC is a cylindrical tunnel filled with natively unstructured proteins, the FG-Nups. These “spaghetti like” structures play a key role in molecule traffic to and from the nucleus – they create the permeability barrier. Small soluble molecules, like water, ions and small proteins are permitted to freely pass the barrier, while larger molecules (RNA, proteins) require an energy dependent transport mechanism. Proteins destined for the nucleus have a “nuclear ticket” imprinted in their sequence – a nuclear localization signal (NLS). The NLS is recognized in the cytoplasm by the transport proteins – karyopherins. Upon binding of their cargo protein, karyopherins transiently bind the FG-Nups, thereby overcoming the permeability barrier of the pore. Directionality of the transport is determined by the RanGTP gradient across the NE; in the nucleoplasm RanGTP causes disassembly of the cargo-karyopherin complex and the karyopherin can recycle back to the cytoplasm. The transport of membrane proteins is less well understood. Particularly, the energy dependence of nuclear import of membrane proteins still divides the field. The first widely accepted transport model of inner nuclear membrane proteins, the diffusion-retention model, is based on passive flux of proteins through the NPC and binding to nuclear structures – for example chromatin. Later on, multiple lines of evidence pointed towards energy dependence of the transport of (at least) some transmembrane proteins. My research has also contributed to the elucidation of this energy-dependent mode of import of membrane proteins. In chapter 2 we propose a model of active transport of the yeast S. cerevisiae protein Heh2p, in which a long natively-unfolded linker spans the structure of the pore from the membrane to the central channel, exposes the NLS and facilitates karyopherin-FG-Nup interaction. We have shown that this process depends on the linker length (but not the amino acid sequence), karyopherin Kap95p and certain FG-Nups (Nup 100, Nup145 and Nup57). We have also shown the unfolded nature of the linker, and finally we have trapped the Heh2p-based reporter at the centrally located FG-Nup, Nsp1p. In the course of the project, Nsp1p has been proven not to be the best choice to probe the path of transport through the NPC. The discovery of a cytoplasmic pool of Nsp1p by Colombi et al. and Makio et al. made us doubt the readout of the trapping experiment. In chapter 5 we investigate the influence of the cytoplasmic pool of Nsp1p and came to the conclusion that without proper estimation of Nsp1p quantities in

122

the cytoplasm (which is currently unavailable), it is impossible to discriminate between the binding of our reporter to Nsp1p in the pore versus the soluble fraction in the cytoplasm. However, we deem it likely that binding to the nuclear envelope-localized Nsp1p pool (i.e. NPCs or NPC assembly intermediates) is dominant. Simultaneously, we aimed to map the route of the Heh2p based reporters through the pore by performing the trapping experiments with different set of nucleoporins– two FG-Nups, Nup53p and Nup59p, and one structural nucleoporin, Nup170p. These data are presented in chapter 4. All of the mentioned nucleoporins are located much closer to the pore membrane than Nsp1p. We have seen the trapping of both actively and passively transported reporters at these NPC positions, so it remains unresolved if the passive and active import of membrane proteins take different routes via the pore. The observation that import of Heh2p to the nucleus is energy dependent is not influenced by the shortcomings of trapping experiments. As described in chapter 6 the route of membrane proteins from the point of synthesis at the ER to its final destination in the cell is more complicated than soluble proteins. To facilitate the different steps in the biogenesis and sorting of the membrane proteins, specific signal sequences are encoded on the proteins. Besides a targeting sequence like an NLS, proteins also have other signals in their structure and the kinetics of binding with their interaction partners may impact the order of events. In case of nucleocytoplasmic transport, we aimed to answer the question if nuclear import could precede membrane insertion of proteins containing 1 transmembrane segment and having a strong NLS. In chapter 4 we present experiments that point against this hypothesis, which supports our working model that transport of membrane proteins is via the membrane environment. That brings us back to the question: where do the membrane proteins pass the NPC? As explained before all the traffic through the pore is directed via the central channel, where FG-Nups and karyopherins interact. Since membrane insertion happens prior to crossing of the pore, the route via the central channel seems unlikely – the required membrane extraction step is energetically unfavorable. Small tunnels on the side of the pore have been observed by immunoelectron microscopy projections of the NPC, and these form the peripheral channels. However, the space offered by them is limited and would impose a limit to the size of cytoplasmically exposed domains of membrane proteins. Contrary to that notion, experiments presented in chapter 2 and 4 of this thesis show that extralumenal domains up to 140 kDa are tolerated and such proteins are accumulated in the nucleus. Our claim is that they are transported in an energy-dependent manner, this way overcoming potential diffusion limits. In chapter 6 we explain that multiple transport mechanism may in fact apply for the import of inner nuclear membrane proteins. For example, full length Heh2p contains retention domains, which were removed from our reporters in order to measure the kinetics of import and leak. Native Heh2p most probably enters the

123

nucleus in an energy-dependent process, but stays there due to interaction with nuclear binding partners. In chapters 2 and 4 we show that the reporters without the retention domains are highly mobile and able to leak out from the pore if the active transport is abolished. We claim that similar interference of active transport and retention models may exist for other proteins. In chapter 6 we acknowledge that active transport maybe more crucial in cells with an intact nuclear envelope, such as in yeast S. cerevisiae and in postmitotic metazoan cells, than for mitotic metazoan cells where membrane proteins may enrich at the inner membrane while the nuclear envelope reforms after mitosis.

124

LIST OF PUBLICATIONS

•Laba JK, Steen A, Popken P, Chernova A, Poolman B, Veenhoff LM (2015) Cells 4(4):653-73

•Laba JK, Steen A, Veenhoff LM (2014) Curr Opin Cell Biol 28C :36-45

•Meinema AC*, Laba JK*, Hapsari RA*, Otten R, Mulder FAA, Kralt A, van den Bogaart G, Lusk CP, Poolman B, Veenhoff LM (2011) Science 333, 90-93 * equal contribution

•Kolaczkowska A, Manente M, Kolaczkowski M, Laba J, Ghisian M, Wawrzycka D (2012) FEMS Yeast Res 12(3), 279-292.

125

ACKNOWLEDGEMENTS

It is 15 November 2015 and I am sitting at the Warsaw airport, waiting for a connection flight to Amsterdam. I have just turned 30 today and my thesis will go to the printing office this week. Somehow I cannot think of a moment that would be more suitable to write this acknowledgement - travelling between the Netherlands and Poland was such a big part of my life over the past 6 years. Those were exciting years filled with hard work, collaborations, some successful projects and some failures. My PhD would not be the same without people around me – my supervisors, collaborators and colleagues. This is the time to thank you all for your contribution. Firstly, my supervisor Liesbeth Veenhoff. I still remember this day back in 2008 when I received an email from you saying “Can you imagine a PhD in Groningen?”. It was several months after I came back from my Erasmus exchange, in which I conducted a rather successful student project under your (and Anne’s and Geert’s) supervision. There was not a single doubt in my mind about it – I could not wait to come back and work with you again on yeast NPC transport. I still had a year of studies to go, but you have kindly agreed to wait. Thank you for giving me this chance and for the many others later. I always admired your enthusiasm for science and your creative ideas. Thank you for the understanding and support throughout all these years. My promoter, Bert Poolman. Thank you for letting me join the group, both back in 2007 as a student and in 2009 as a PhD student. Thank you for making the time in your busy schedule to listen about how my projects are going and for the useful input. I would also like to thank the reading committee: Ellen Nollen, Dirk Jan Slotboom and Gino Cingolani for making the time to read my thesis and for the helpful comments. Jacek and Frans – I could not imagine better paranymphs than the two of you. Jacek, as cheesy as it may sound, if I ever had a mentor in my life, it was probably you. You were my first friend in the lab, always willing to listen and advise. Life is not the same without the daily dose of your super mean comments! Thanks for all the parties and all the times you were there for me. You really should become a professor. Frenemy – our special relationship, mostly based on picking on each other made the PhD time much more fun. Thanks for making the time to be there for me before and during the defense – I know that last months were already overfilled. I hope you have enjoyed it a bit as well. Anne Meinema and Astri Hapsari were my closest colleagues and collaborators in the beginning of my PhD. Anne, thank you for showing me the world of science during my student time and for introducing me to the Netherlands. We had a lot of fun both in and outside the lab and I will always remember that. Astri – you were personally very close to me. We joined the lab at roughly the same time, you also started as a student with Anne and Liesbeth and we were both foreign. Thank you for all the times I

126

could borrow the MgCl2 (and other stuff…). I always admired your perseverance, your calmness in stressful situations and the fact that you were cheerful no matter what! I am so happy that you are also about to defend your thesis shortly. As the time passed and projects developed other people have joined the team and the dynamics has changed. Anton, as I stated many times before, you were my superhero in the last 2 years of my contract. I am not sure how my thesis would have looked like if you have not joined the group. Thank you for the fruitful discussions and the morning coffees. Petra – you were a great colleague and your ideas and analytical mind had huge influence on chapter 4. I really enjoyed the spontaneous brainstorm sessions between you, Astri and me; those were very creative times. Chapter 4 would not be complete without experiments of Alina Chernova – thank you very much for your hard work and motivation! My student, Hesley de Vries. You were a great student, very motivated and professional. I really enjoyed the time with you in the lab. I wish all this hard work was published somewhere, but tough luck; this project might have been a bit too crazy. All the best for the future! I would like to thank the collaborators from ETH Zurich – especially Helge Ewers and Charlotte Kaplan for hosting me for 6 weeks in summer of 2012 and introducing me to the world of blinking molecules. Also other lab members from Zurich – Thomas thank you for setting up Matlab on my computer. Mayra and Monika you were awesome labmates! Andrew Robinson and Antoine Van Oijen – thank you for your willingness to put time and effort into continuation of the super resolution project. Andrew, sorry for all the times you waited for me at the microscope. Good luck for both of you in Australia! Fabrizia Fusetti and Alicja Filipowicz – thank you for the collaboration on the proteomics project. It is sad that it didn’t work out. I have learned a lot from both of you. Alicja, your perseverance and professionalism will always be an inspiration for me. All the members of the Veenhoff lab, whom I didn’t mention by name yet: Annemarie, Georges and all the students who joined the group when I was there. Thank you all for creating a good working atmosphere and for the useful discussions and help throughout our lab struggles. Georges, thanks for your laid back attitude. Throughout the years the Poolman lab was changing a lot – but a creative and friendly atmosphere always stayed. I would probably have to write down 50 names here in order to thank all of you properly. Let me just say to all the lab members, guests and students – whether we did experiments together, or you have borrowed me buffers, nucleotides or media, or if we just chatted and laughed over coffee - thank you for that! I am very grateful to you and very happy to have met you!

127

Members of the Chang lab, especially Micheal, Clemence and Jannie, thank you for sharing the lab and work discussions. Micheal, thanks for bringing a different perspective to the lab. Clemence, my friend, thank you for listening each time when I had to talk. I do hope your projects will turn out great and you will become a successful scientist in the country of your choosing. Jannie – you are the best, always patient, always smiling and very professional. In my final years our group has moved to ERIBA. I would like to thank ERIBA team for adopting me, although I was not really working on ageing. Special thanks to ERIBA PhD committee for organization of all the social events. Kuba, Magda, Quentin, Stijn – thanks for making our shared lab space a nice place to work. Edu, thank you very much for designing my cover. Brainstorming with you was a great process and I love the outcome. Lucas, I would have never imagined I would dare to write the Dutch summary on my own. Thanks for planting this idea in my head and for all the corrections. I am very proud that many of my colleagues became friends for life. Tejas, I do hope to visit you soon in Zurich. Thanks for being even more confused than I was and a victim of Jacek’s jokes even more often than I was. Josy – thanks for all the girls nights and small chats during work. You were the best officemate ever! Ronnie – I do believe that the great atmosphere in the lab had something to do with your social skills and perseverance about the beer club. Marysia – a separate note for you, as my roommate, labmate and friend. Thank you for organizing so many social events, in the lab and for the Polish community. Thanks for all these buffers I have borrowed and never returned. Maciek – thanks for always being there for me when I found myself in some sort of technical drama again. If you end up away from your roots, it is so important to find people who make you feel like home - I would like to thank our Polish community for creating this home feeling for me. Marysia, Maciek, Hania, Jacek, Wojtek, Weronika, Alicja, Witek, Mariusz, Bartek, Agata, Piotrek, Gosia, Sławek, Kasia, Maciek, Ewa, Łukasz – dzięki za obiady i imprezy! Girls from the gym – Maria Jose, Maria Jesus, Lorenza, Marta, Oktavia, thank you for keeping my motivation and endorphin level high. Salsa Juan Carlos, and personally Feline and Juan Carlos – thank you for the friendly atmosphere and for all the great lessons and parties. Thanks to all the great people that I met there, especially Rob and Valeriya. Jolanda, Henk, Matthijs and Ira – thank you for making me feel like I am part of the family. To my family – thank you for your support and patience over all these years. Mum, Dad, this thesis would have never happened without you. Thank you for giving me the freedom to choose my own way. My sister, Magda - I have missed a lot during last years and now you are likely moving even further. I was always very happy that

128

despite the distance we could continue our conversation as if we saw each other just a day before. I believe that even Australia can’t change that! Moim Babciom dziękuję za cierpliwość, ciepłe słowa za każdym razem gdy Was odwiedzałam i za to, że nigdy nie miałyście mi za złe, że nie przyjeżdżam częściej. Last, but not least, Michiel. I would have never finished this thesis without your love and support. Words cannot describe how grateful I am to you. Whether it comes to running, dancing or my PhD - you motivate me to search for my boundaries and push through them. I became a better person with you and I hope you feel as much support from me as you give me. I love you and I am looking forward to coming years together!

on the mechanism of transport of inner nuclear membrane proteins

Documents