University of Groningen Small Heat Shock Proteins Vos, Michel

University of Groningen

Small heat shock proteins Vos, Michel

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2009

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): Vos, M. (2009). Small heat shock proteins: Implications for neurodegeneration & longevity. s.n.

Copyright Other 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 the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

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

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

Download date: 10-10-2021 SMALL HEAT SHOCK PROTEINS Implications for Neurodegeneration & Longevity

Michel J. Vos The research described in this thesis was conducted at the Department of Cell Biology, Section of Radiation & Stress Cell Biology, University Medical Center Groningen, University of Gronin- gen, The Netherlands.

This work was supported by Innovatiegerichte Onderzoeksprogramma (IOP) Genomics Grant IGE03018. The printing of this thesis was financially supported by: Vereniging van Huntington, Den Haag University of Groningen Department of Radiation & Stress Cell Biology (University Medical Center Groningen, University of Groningen) Groningen University Institute for Drug Exploration (GUIDE) Biolegio, Nijmegen Genetic Services Inc, Sudbury, USA

GENETIC SERVICES, INC.

© Copyright 2009 M.J. Vos All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, mechanically, by photocopying, recording or otherwise, without the written permission of the author.

ISBN: 978-90-367-3904-7

Page layout: Michel J. Vos Cover design: Michel J. Vos Cover illustration: Electron microscope image of a Drosophila eye Printed by: Drukkerij van Denderen B.V. Groningen, The Netherlands RIJKSUNIVERSITEIT GRONINGEN

SMALL HEAT SHOCK PROTEINS Implications for Neurodegeneration & Longevity

Proefschrift

ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op woensdag 1 juli 2009 om 13.15 uur

door

Michel Jeroen Vos geboren op 4 september 1979 te Delfzijl Promotores : Prof. dr. H.H. Kampinga Prof. dr. O.C.M. Sibon

Beoordelingscommissie : Prof. dr. R.M. Tanguay Prof. dr. I. Braakman Prof. dr. G. de Haan

Voor Angelique

Table of contents

Chapter 1 Introduction ~ Structural and functional diversities between page 9 members of the human HSPB, HSPH, HSPA and DNAJ chaperone families

Chapter 2 A PCR amplification strategy for unrestricted generation of page 29 chimeric genes

Chapter 3 HSPB7 is a SC35 speckle resident small heat shock pro- page 35 tein

Chapter 4 A comparative screen of the human HSPB family of molec- page 57 ular chaperones reveals HSPB7 as a superior suppressor of aggregation of polyglutamine containing proteins

Chapter 5 Longevity as evoked by different chaperone activities of page 79 small heat shock proteins

Chapter 6 General discussion & perspectives page 103

Appendix Nederlandse samenvatting page 117

Dankwoord page 121

Curriculum vitae page 125

Colour figures page 129

CHAPTER 1

Introduction & aim of the thesis

Structural and functional diversities between members of the human HSPB, HSPH, HSPA and DNAJ chaperone families

Michel J Vos*, Jurre Hageman*, Serena Carra & Harm H Kampinga

Department of Cell Biology, Section of Radiation and Stress Cell Biology, University Medical Center Groningen, Groningen, The Netherlands

* Equal contribution

Published in Biochemistry, 2008 Jul 8;47(27):7001-11 Chapter 1

Abstract

Heat shock proteins (HSP) were originally identified as stress-responsive proteins required to deal with proteotoxic stresses. Besides being stress-protective and possible targets to delay progression of protein folding diseases, mutations in chaperones also have been shown to cause disease (chaperonopathies). The mechanism of action of the ‘classical’, stress-inducible HSPs to serve as molecular chaperones preventing the irreversible aggregation of stress-unfolded or disease related misfolded proteins is beginning to emerge. However, the human genome encodes for several members for each of the various HSP families that are not stress-related but contain conserved domains. Here, we have reviewed the existing literature on the various members of the human HSPB (HSP27), HSPH (HSP110), HSPA (HSP70) and DNAJ (HSP40) families. Apart from structural and functional homologies, several diversities between members and families can be found that not only point to differences in client specificity but that also seem to serve differential client handling and processing. How substrate specificity and client processing is determined is yet far from being understood.

10 Introduction & aim of the thesis

Introduction

Heat shock proteins were originally discovered as proteins that are upregulated upon and protective against proteotoxic stresses, i.e. situations that increase the fraction of proteins that are in a (partially) unfolded state, thereby enhancing their probability to form intracellular protein aggregates that can lead to loss of cell function and eventually to cell death. It is nowadays appreciated that also a variety of normal cellular processes (translation, transport over membranes) constantly challenge the cellular protein homeostasis and require protein quality control systems for assistance. In addition, diseases like Alzheimer’s- and Parkinson’s disease, CAG-repeat diseases and many heart diseases (e.g. atrial fibrillation) or physiological disturbances (e.g. hypoxia) are pathogenic because they disturb protein homeostasis. Finally, folding mutations may arise as a result of somatic mutations (aging) and genomic instability (cancer) requiring increased protein quality control. The heat shock proteins (HSP) are a group of structurally unrelated protein families (HSPA, HSPB, HSPC, HSPD, HSPH, DNAJ) that play a prime role in protein homeostasis by binding to substrates-at-risk, thereby keeping them in a state competent for either refolding or degradation. As such they belong to a much larger superfamily of molecular chaperones. The number of genes coding for the diverse HSP family members largely varies per organism. For HSPA, the number varies from 3 in Escherichia coli, 14 in Arabidopsis thaliana, 12 in Drosophila melanogaster to 13 in Homo sapiens. For small HSP (HSPB), the number of genes is relatively high in plants and the same holds true for DNAJ (Table 1). Although originally identified as heat inducible proteins, many members are in fact not heat shock inducible. However, within each family, individual heat shock inducible proteins are found such as HSPB1, HSPA1, HSPH1 and DNAJB1. Why there are so many members in most families, sometimes with high sequence homology (HSPA, HSPH) but sometimes also with substantial sequence divergence in certain domains (HSPB, DNAJ) is largely unclear. Part of the redundancy may relate to intracom- partmental distribution of the diverse family members and the requirement of their activities in these different compartments. Also, some HSPs show tissue or development specific expression (Table 3). On one hand, this may reflect the ability to specifically regulate expression of the same activity and function. On the other hand, this suggests that chaperones are not merely promiscuous in terms of clients and indicates a high need for specialized chaperones under these conditions. That promiscuity may not be an essential feature of HSP/chaperone activity is further supported by the existence of different HSP families and family members within the same compartment (e.g. the cytosol) (Table 3). In this review, we focus on the structural, sequence and functional divergence within the HSPB, HSPA, HSPH and DNAJ families either in terms of client specificity or client processing. The classical model of the HSP-chaperone functions is primarily based on cell-free experiments with the human HSPB1, HSPA1 (or HSPA8), HSPH1 and DNAJB1 and work on their orthologues in E. coli, yeast, and mouse. In this model, un- or misfolded proteins bind to these HSPs both directly or sequentially (Figure 1). In naïve cells, that were not stressed before, the instant chaperone action towards increases in proteotoxic stresses is obtained by constitutively expressed members such as cellular stores of small heat shock proteins (HSPB). It is thought that HSPB members are stored in oligomeric complexes that dissociate into smaller sized molecules upon stress (1;2). This shift in oligomeric size allows for binding of unfolded proteins, which effectively neutralizes the chance for non-specific interaction of the unfolded substrate with other proteins. HSPB members are ATP-independent chaperones and require other partners for further client processing (see below), which, depending on the substrate and/or chaperone partners can either be refolding or degradation. How this distinction in client processing is regulated is yet unclear. Clearly, transfer to the HSPA/B machine has been suggested to promote folding in vitro as well as in living cells. This ATP-dependent chaperone constantly shuttles between

11 Chapter 1 an ATP and an ADP bound state in which it has different affinity to bind unfolded proteins (3). Substrates enter the HSPA complex in the ATP bound configuration. In this configuration, HSPA has a high substrate on/off rate, meaning low substrate affinity. Upon binding, ATP is hydrolyzed which stabilizes the affinity of HSPA for its substrate, a reaction which is regulated by cofactors like DNAJs and CHIP. Subsequently, nucleotide exchange is stimulated (BAG1, HSPBP1, HSPH) resulting in an ATP bound HSPA complex followed by substrate release. Unfolded proteins may also directly enter the HSPA chaperone machine or through assistance of the HSPA co-chaperones like DNAJ or HSPH. In fact, there are several modulators of the HSPA ATP cycle (HSPH, DNAJ, HIP, CHIP, BAG3, HSPBP1), which not only modulate the cycle, but also may confer client specificity to the HSPA machine and/or affect the fate of its client.

Unfolding Aggregation Substrate Folded substrate HSPB

DNAJ

HSPA

Proteasome

CHIP - ATP

ATP ATP HSP110 + DNAJ + BAG + HSPBP1 + ADP ADP ADP

HIP +

Figure 1. Schematic representation of HSP-mediated client processing. A folded substrate is unfolded upon a proteotoxic stress event. This unfolded substrate either aggregates or binds HSPs like HSPB, DNAJ, or HSPA. Both HSPB and DNAJ are thought to eventually hand over the substrate to the HSPA machine capable of binding and releasing the substrate. Released substrates that still expose hydrophobic patches to the exterior are bound again by HSPs, whereas substrates without such hydrophobic patches are not recognized. Substrates can also be targeted to the proteasome degradation system by HSPB1 and CHIP or by other uncharacterized mechanisms. See Appendix 4 for colour print.

The HSPB family

Small heat shock proteins (sHSPs or HSPB in mammals) are low molecular weight chaperones (Table 3) found in every kingdom. sHSPs are characterized by the presence of a conserved crystallin domain flanked by a variable N- and C-terminus. The N- and C-termini, together with part of the crystallin domain are involved in substrate binding (Figure 2). The majority of structural information of sHSPs comes from studies performed in archaea and plants which show that sHSPs form large symmetrical complexes composed of several dimers. Dimers are formed

12 Introduction & aim of the thesis

Table 1. Occurrence of HSP gene numbers in different species HSPA/H DNAJ HSPB Genome size (bp) 9 H. sapiens 11/3 41 11 3,3 x 10 8 D. melanogaster 12/2 36 11 1,2 x 10 8 A. thaliana 14/4 89 19 1,1 x 10 7 S. cerevisiae 14/2 22 2 1,2 x 10 6 E. coli 3 6 2 4,6 x 10 through strand exchange between a b-sheet extending loop present in the crystallin domain. These interactions are further strengthened by the C-terminal extensions. Together with the N- terminal extensions, this allows for buildup and stabilization of the higher oligomeric structure. The homo and/or heterogeneous oligomeric complexes are believed to be reservoirs, which under stress can dissociate into smaller multimers that are generally assumed to be the active units. Upon dissociation into dimers (Figure 1), hydrophobic residues in the N-terminus, C- terminus and crystallin domain become exposed, allowing interaction with substrate molecules (4) and preventing their irreversible aggregation. Further processing of the bound substrate is carried out by other HSP families, either directing it for refolding or degradation (Figure 1).

Functional diversity

HSPB1 HSPB1 (HSP27) is one of the most well studied members of the family and can exist as high (e.g. hexadecamers) or low molecular weight (e.g. tetramers and dimers) structures. Under non- stressed situations, a high molecular weight form is the most predominant species. During heat stress, its level decreases with a concurrent increase in the amount of two low molecular weight phosphorylated forms. HSPB1 can be phosphorylated at three sites (Figure 2) which regulates its activity. A pseudophosphorylated mutant of HSPB1 (HSPB1-3D) shows a decrease in in vitro chaperone activity, either implying the oligomeric sHSP structure as a key for in vitro chaperone action or pointing to a need for shuttling between dimers and oligomers. The latter is supported by data from cellular studies showing that overexpression of the pseudophosphorylated HSPB1 increases the cellular chaperone capacity (5). In this setting, HSPB1 may also form mixed oligomers with endogenously expressed HSPB members. Interestingly, a non-phosphorylatable mutant was ineffective in cells to increase chaperone activity. Taking the in vitro and cellular data together, shuttling between dimers and (heterogeneous) oligomers seems required for substrate binding and protection against aggregation (Figure 1). However, the non-phosphorylatable mutant is still able to protect against oxidative stress, suggesting that either this stress directly causes (phosphorylation-independent) changes in oligomeric structure or that non-dynamic oligomeric structures can also act cytoprotectively. Once activated by phosphorylation and oligomeric changes, HSPB1 can bind non-native substrates. For substrate release it requires the help of ATP-dependent chaperones (HSPA) or proteases (HSP104) (6) that further process the client. Consistent with this notion, the increase in refolding mediated by transfected HSPB1 was prevented by inhibiting the HSPA chaperone machine (5). When refolding via the HSPA machine is not possible, HSPB1-bound substrates might be ubiquitynated and targeted to the 26S proteasome for degradation. Besides its role in assisting refolding and proteasomal targeting of soluble (denatured) proteins, one of the best characterized functions of HSPB1 is its ability to interact with several cytoskeletal components including actin, intermediate filaments and microtubules (7). In muscle tissue, HSPB1 is found in association with sarcomeres and has been suggested to be

13 Chapter 1

Table 2. Alternative names and gene identifiers of the human HSP gene families Gene Protein Alternative Name Human Name Name GeneID HSPB HSPB1 HSPB1 CMT2F; HMN2B; HSP27; HSP28; Hsp25; HS.76067; DKFZp586P1322 3315 HSPB2 HSPB2 MKBP; HSP27; Hs.78846; LOH11CR1K; MGC133245 3316 HSPB3 HSPB3 HSPL27 8988 HSPB4 HSPB4 crystallin alpha A, CRYAA, CRYA1 1409 HSPB5 HSPB5 crystallin alpha B, CRYAB, CRYA2 1410 HSPB6 HSPB6 HSP20; FLJ32389 126393 HSPB7 HSPB7 cvHSP; FLJ32733; DKFZp779D0968 27129 HSPB8 HSPB8 H11; HMN2; CMT2L; DHMN2; E2IG1; HMN2A; HSP22 26353 HSPB9 HSPB9 FLJ27437 94086 HSPB10 HSPB10 ODF1; ODF; RT7; ODF2; ODFP; SODF 4956 HSPB11 HSPB11 Hsp16.2 51668 HSPH HSPH1 HSPH1 HSP105 10808 HSPH2 HSPH2 HSPH4 APG-2; Hsp110 3308 HSPH3 HSPH3 HSH4L APG-1 22824 HSPH4 HSPH4 HYOU1 10525 HSPA HSPA1A HSPA1A HSP70-1, HSP72, HSPA1 3303 HSPA1B HSPA1B HSP70-2 3304 HSPA1L HSPA1L hum70t, hum70t 3305 HSPA2 HSPA2 Heat-shock 70kD protein-2 3306 HSPA5 HSPA5 BIP, GRP78, MIF2 3309 HSPA6 HSPA6 heat shock 70kD protein 6 (HSP70B') 3310 HSPA7 HSPA7 3311 HSPA8 HSPA8 HSC70, HSC71, HSP71, HSP73 3312 HSPA9 HSPA9 GRP75, HSPA9B, MOT, MOT2, PBP74, mot-2 3313 HSPA12A HSPA12A FLJ13874, KIAA0417 259217 HSPA12B HSPA12B RP23-32L15.1, 2700081N06Rik 116835 HSPA14 HSPA14 HSP70-4, HSP70L1, MGC131990 51182 HSPA15 HSPA15 Stch 6782 DNAJ DNAJA1 DNAJA1 DJ-2; DjA1; HDJ2; HSDJ; HSJ2; HSPF4; hDJ-2 3301 DNAJA2 DNAJA2 DNJ3; mDj3; Dnaj3; HIRIP4 10294 DNAJA3 DNAJA3 Tid-1; Tid1l 9093 DNAJA4 DNAJA4 Dj4; Hsj4 55466 DNAJB1 DNAJB1 HSPF1; Hsp40 3337 DNAJB2 DNAJB2 HSJ1; HSPF3 3300 DNAJB4 DNAJB4 Hsc40 11080 DNAJB5 DNAJB5 Hsc40; Hsp40-3 25822 DNAJB6 DNAJB6 Mrj; mDj4 10049 DNAJB7 DNAJB7 Dj5; mDj5 150353 DNAJB8 DNAJB8 mDj6 165721 DNAJB9 DNAJB9 Mdg1; mDj7; ERdj4 4189 DNAJB11 DNAJB11 Dj9; ABBP-2 51726 DNAJB12 DNAJB12 Dj10; mDj10 54788 DNAJB13 DNAJB13 Tsarg 374407 DNAJB14 DNAJB14 EGNR9427; FLJ14281 79982 cardioprotective. In neuronal cells, overexpression of a mutated form of HSPB1 (S135F), which is associated with distal hereditary peripheral neuropathies (8), resulted in neurofilament network disruption, further confirming the critical role of this protein in cytoskeleton stabilization. Finally, after (heat) stress, HSPB1 shows a reversible nuclear accumulation into SC35 splicing speckles, structures implied in RNA processing (5). Also other HSPB members like HSPB5 and HSPB7 (Chapter 3) have been suggested to localize to these structures. HSPB1 localization to these nuclear structures is not associated with refolding of heat denatured nuclear proteins but overlaps with sites of nuclear protein degradation (5).

14 Introduction & aim of the thesis

HSPB4 and HSPB5 HSPB4 (aA-crystallin) and HSPB5 (aB-crystallin) are both highly expressed in the eye lens, where together they maintain lens transparency. Mutations in both HSPB4 and HSPB5, causing protein instability and aggregation, result in congenital cataract development (9) (Table 3). HSPB5 is also expressed in striated muscle where, together with HSPB1, it can be found in association with sarcomeric structures. In line with this, mutations in HSPB5 have been linked to muscle and cardiac myopathies (9). Like HSPB1 and HSPB4, HSPB5 has chaperone activity in vitro for which a dimer comprising only the crystallin domain was found to be sufficient (10). Also HSPB4 and HSPB5 can be phosphorylated on different serines (Figure 2). Recombinant HSPB5-3D forms oligomers that are smaller than those formed by wildtype HSPB5. Whereas the 3D mutant showed a reduced chaperone activity towards heat-denatured luciferase both in vitro and in cells (2), it showed an enhanced chaperone activity towards heat-induced aggregation of citrate synthase and amyloid fibril formation of α-synuclein (11). These data strongly suggest that HSPB5 shows substrate-specificity. Interestingly, the formation of mixed oligomers between HSPB5 and HSPB5-3D reduced the total chaperone capacity of HSPB5-3D. Thus, subunit exchange and the ratio between both phosphorylated and unphosphorylated subunits may be a key modulator of chaperone activity. An R120G mutation of HSPB5 was found to be associated with desmin related myopathy (9). This mutation results in hyperphosphorylation and aggregation-proneness which prevents nuclear entry, reduces chaperone activity both in vitro and in vivo (12;13). HSPB5, like HSPB1, plays an important role in cytoskeleton stabilization, which is also dependent on its phosphorylation state.

HSPB8 HSPB8 (HSP22/H11/E2IG1) is highly expressed in striated and smooth muscles, brain and keratinocytes. Like HSPB1 and HSPB5, HSPB8 can be phosphorylated in vitro (Figure 2). In contrast to HSPB1 and HSPB5, phosphorylation only marginally affects the tertiary and quaternary structure of HSPB8. Both wildtype and phosphorylated HSPB8 exist as low molecular weight oligomers. Phosphorylation of HSPB8 in vitro severely lowers its chaperone activity towards denatured insulin and rhodanase. In view of HSPB1 and HSPB5, where phosphorylation increases chaperone activity and reduces oligomeric size, phosphorylated HSPB8 is present in slightly larger oligomeric structures than wildtype HSPB8. This could explain the different effect phosphorylation has on HSPB8 chaperone activity (14). In vitro, HSPB8 can form high molecular weight hetero-oligomers by interacting with other members of the HSPB family. In vivo, HSPB8 forms a stable and stoichiometric complex with the co-chaperone BAG3. Interestingly, HSPB8 stability depends on its association with BAG3 (15). In contrast to HSPB1 and HSPB5, which interact and stabilize cytoskeletal elements, no direct association of HSPB8 with actin and microtubules has yet been reported. Rather, a specific role for HSPB8 in autophagy is emerging (15). Recently mutations in HSPB8 (K141E and K141N) have been associated with hereditary peripheral neuropathies (16). The mutations decrease HSPB8 oligomer dissociation, reduce chaperone activity in vitro (17), and also reduce the ability of HSPB8 to clear polyglutamine proteins in cells.

Other members and their diversity Besides HSPB1, HSPB4, HSPB5 and HSPB8, seven more HSPB family members are found in humans. The eleventh member, only recently identified, has been named HSP16.2, but will here be referred to as HSPB11 (Table 2 and 3). HSPB2, also known as myotonic dystrophy protein kinase binding protein (MKBP), interacts with the myotonic dystrophy protein kinase (DMPK), for which mutations have been linked to the development of myotonic dystrophy (18). In muscle cells, HSPB2 forms an oligomeric com-

15 Chapter 1 plex with HSPB3, which expression is induced during muscle differentiation (19). Due to its interaction with DMPK and its ability to enhance its kinase activity, a role for HSPB2 in muscle maintenance has been suggested (18). HSPB6 (HSP20) is highly expressed in smooth muscles and seems to play a role in muscle relaxation (20) and cardioprotection. In vitro, HSPB6 exists mainly as dimers. In the presence of HSPB1 it can oligomerize to form higher molecular weight complexes with a molecular mass of 100-300 kDa. HSPB6 can be phosphorylated at serine 16 and mimicking its phosphorylation (S16D mutant) resulted in decreased chaperone activity in vitro. Recently, HSPB6 was reported to interact in a phosphorylation dependent manner with 14-3-3 proteins that function as regulator of a wide variety of processes (21). HSPB7 (cvHSP) is expressed in heart and skeletal muscle. Analysis of aging muscle shows a large increased expression of both HSPB7 and HSPB5 (22). This could reflect a cellular adaptation to higher proteotoxic stress conditions related to muscle degeneration. HSPB7 upregulation is also found in muscular dystrophy affected diaphragm muscle, again linking high stress levels with HSPB7 induction. Next to the muscle-associated HSPB members, two members (HSPB9 and HSPB10) are exclusively expressed in testis. Through its C-terminus, HSPB9 interacts with DynLT1, which is a light chain component of dynein (23), one of the energy-dependent tubulin motor proteins. However, the physiological significance of this interaction is not yet known. The other testis- specific member, HSPB10 (sperm outer dense fiber protein 1, ODFP/ODF1) seems to fulfill a structural function in the flagellar axoneme cytoskeleton of sperm cells (24), which may indicate additional substrate specificity. But, further structural and functional information is still lacking. The recently reported HSPB11 (HSP16.2), was shown to form oligomeric complexes and to prevent the aggregation of in vitro denaturated aldolase and glyceraldehyde-3-phosphate dehydrogenase in accordance with the chaperone model of HSPB1 and HSPB5. HSPB11 overexpression protected against etoposide-induced cell death which correlated with a decreased release of mitochondrial cytochrome c into the cytosol. Inhibiting HSP90 function completely abrogated the protective effect of HSPB11 (25). This would suggest that at least in the case of HSPB11, interaction with other chaperone machines besides HSPA1A may contribute to functional specificity and cellular functioning.

Concluding remarks The major regulatory mechanism of sHSPs involves phosphorylation and oligomeric redistribution. This can result in a sub-cellular redistribution, influence chaperone activity and even substrate specificity. The outcome of these modifications can, however, be diverse and are only known for a few members. Besides playing a key role in the response to (external) stresses, HSPB members are also crucial for normal cellular functioning, especially in muscle tissue and the eye lens. This is emphasized by several diseases caused by mutations in several HSPB members. Inversely, the functional activity of some HSPB members may be exploited for future therapeutic intervention in combating protein folding diseases like heart diseases (atrial fibrillation, ischemia) and neurodegenerative diseases (Alzheimer, amyotrophic lateral sclerosis, CAG repeat diseases).

The HSP70 chaperoning machine

General introduction Whereas, the HSPB chaperones (HSPB1, HSPB4, HSPB5) may form the primary line of defense under stress, the HSPA chaperone machine, especially HSPA1A/B in conjunction with its co-

16 Introduction & aim of the thesis

A HSPB

HSPB N- Crystallin domain -C Phosphorylation sites 15 78 82 S S S HSPB1 41 42 45 122 SS S S HSPB4 19 45 59 S S S HSPB5 24 57 S S HSPB8

B HSPA/HSPH

Association with: Hsp40 Association with: Hsp40, HspA1A J-domain, Bag-1, Hip Hop, Tpr-2, CHIP HspA1B HspA1L ATPase Substrate C-terminal EEVD HSPA N- Linker -C HspA2 HspA6 HspA8

HSPH N- ATPase Linker Substrate C-terminal -C

C DNAJ HPD

DNAJA N- J-Domain G/F Zn2+ C-terminal -C

HPD

DNAJB N- J-Domain G/F C-terminal -C

HPD

DNAJC N- N-terminal J-Domain C-terminal -C

Figure 2. Linear representation of HSPB, HSPA/HSPH, and DNAJ proteins. chaperone DNAJB1, are the strongest stress-inducible proteins. As ATP-dependent chaperone machine, it can act on substrates bound to HSPB oligomers after being induced (see Figure 1). However, several members, including some stress induced members, are also expressed under non-stress conditions (HSPA8, DNAJA1, HSPA1A/B and DNAJB1). The human genome encodes 13 different HSPA members, 4 HSPH members and 41 different DNAJ members (26;27). HSPA proteins are highly homologous to the 4 members of the HSPH (HSP110) family. In fact, HSPA4 and HSPA4L, are currently annotated as HSPA members in the NCBI gene database, but are more homologues to HSPH1 and therefore here referred to as HSPH2 and HSPH3. In addition, a 4th HSPH member (Grp170) is present in the ER and here referred to as HSPH4. Typically, HSPA proteins consist of a N-terminal ATPase domain of 45 kDa and a C-terminal substrate binding domain of 25 kDa. The ATPase and the C-terminal domain are separated by a small linker domain for HSPA members and a longer linker domain for HSPH members (Figure 2). This linker domain couples the nucleotide hydrolysis to the opening and closing of the substrate binding cavity (28). Why HSPH proteins have an extension of the linker domain is currently unknown. Clearly, like HSPA members, HSPH proteins can bind substrates. However, alone they cannot release substrates, which is typical for HSPA members. Several important sites within the above described domains of the HSPA protein have been mapped with high precision. Different co-factors bind to different domains of the HSPA protein:

17 Chapter 1

DNAJ binds to both the ATPase as well as the extreme C-terminal EEVD motif. Deletion or mutation of the EEVD motif causes a disruption of the interdomain communication resulting in an enhanced intrinsic ATPase activity, leading to a reduced binding of substrates. Also, this deletion resulted in complete abrogation of DNAJ binding to and regulation of HSPA (29). Interestingly, the HSPH proteins show deviations from this motif and hence may not functionally interact with DNAJ proteins. The co-factors BAG1 and Hip bind only to the ATPase domain whereas Hop, Tpr-2, CHIP and DNAJ, bind to the C-terminal domain (Figure 2) (30). HSPA proteins constantly shuttle between an ATP and an ADP bound state (Figure 1). DNAJ members stimulate the hydrolysis of ATP. Hip stabilizes the ADP bound HSPA complex whereas HSPH, BAG-1 and HSPBP1 stimulate the ADP to ATP nucleotide exchange. The E3 ubiquitin ligase co-factor CHIP can inhibit the ATP hydrolyzing capacity of HSPA. Finally, Hop does not act on the HSPA ATPase cycle as such, but links the HSPA chaperone to the HSP90/HSPC chaperone complex. Details on these co-factors and how these co-factors affect the fate of HSPA bound substrates have been described elsewhere (30).

The HSPA family

Paradigm according to HSPA1A/B and HSPA8 Crystallographic evidence shows that both HSPA and DNAJ resemble molecular clamps (31). However, the way by which they bind substrates is significantly different. The substrate binding domain of HSPA proteins consists of a short beta sandwich motif which can be locked by the alpha-helical lid structure. The beta-sandwich contains a hydrophobic core of 4-5 amino acids with two flanking basic residues (32). Only a short linear polypeptide fits within the substrate binding domain of the monomeric HSPA protein and therefore it binds only a short stretch of peptides. Because HSPA recognizes very short hydrophobic peptides, it is thought that it can bind a wide variety of substrates. It must be mentioned, however, that these findings are all based on a small number of HSPA proteins like E. coli DnaK and DNAJ, Bovine HSP70/HSPA1A, and yeast DNAJ members Ydj1 and Sis1. As both the HSPA and the DNAJ families are quite diverse, it is likely that the proposed models do not account for all possible cellular HSPA/DNAJ complexes and that some of the different cytosolic HSPA complexes are adapted for a limited range of cellular substrates.

Functional diversities and co-partners specificity A couple of the HSPA members were recently reviewed elsewhere (33) and are therefore described in less detail here. Features of different HSPA members are summurized in Table 3. HSPA1A and HSPA1B differ by only two amino acids and are believed to be fully interchange- able proteins. Both proteins have been referred to as HSP70i (or HSP72) (Table 2) and are the strongest stress-inducible HSPA members. HSPA8, the cognate HSPA and previously referred to as Hsc70 (or HSP73), is expressed in all cell types. It is considered to be the essential “house-keeping” HSPA member. HSPA1L and HSPA2 are two cytosolic family members with high expression in the testis and HSPA2 has been shown to be required for spermatogenesis (34;35). Their biochemical mode of action is currently unknown. HSPA6 is a poorly studied, stress-inducible protein that is lacking in rodents. It is not expressed under normal conditions and only induced upon severe heat stress, where it is thought to act as a final proteotoxic resistance buffer (36). HSPA7 is considered a pseudogene as its transcription product is terminated after 367 amino acids As a complete conserved HSPA protein can originate by bypassing a frame shift at codon position 340, it might also be a true gene which is highly homologues to HSPA6. HSPA9, the mitochondrial HSPA member (HSP75), and HSPA5, the ER localized HSPA chaperone (BiP), are thought to act in a similar manner in their respective compartments as

18 Introduction & aim of the thesis

HSPA8 in the cytosol. In line with this, co-factors such as ER (DNAJC1, DNAJB11, DNAJB9 and DNAJC10) and mitochondrial specific DNAJs (DNAJA3) as well as ER (HSPH4/Grp170/HY- OU1) and mitochondrial (HMGE) specific nucleotide exchange factors have been found (37;38). A small HSPA-like protein, STCH, which we propose to be referred to as HSPA13 (Table 2), has been found attached to microsomes (39) and may fulfill HSPA8-like actions there. HSPA12A and HSPA12B are two distantly related proteins found in atherosclerotic lesions (40). HSPA14 (HSP70L1) is the smallest HSPA protein and interacts with MPP11, the human ortholog of Zuo1, a cytosolic ribosome-associated chaperone that acts together with Ssz1p and the Ssb proteins in yeast as a chaperone for nascent polypeptide chains during translation (41). Although biochemical details of many HSPA members have yet to be identified, recent systems biology approaches in yeast indicated that two distinct chaperone networks with specialized function exist. One molecular chaperone network may protect the proteome against environ- mental stress (HSP) and the second deals with protein translation (CLIPS) and is associated with ribosomes (42). In line with these findings, one may speculate that HSPA1A/B, HSPA1L and HSPA6/HSPA7 belong to the HSP network whereas HSPA5, HSPA8, HSPA9 and HSPA14 belong to the CLIPS network.

The HSPH family

Paradigm according to HSPH1 In vitro, HSPH1 has been shown to suppress the aggregation of denatured luciferase resulting in enhanced refolding of luciferase. However, rabbit reticulosite lysate was always required as a source of co-factors after the denaturation to stimulate refolding in these assays, indicating that HSPH members are good suppressors of irreversible aggregation but lack the release activity typical of HSPA proteins necessary for the stimulation of protein refolding. Consistently, biochemical evidence from yeast Sse1 and mammalian HSPH2 showed that HSPH proteins are poor ATPases (43). Recently, it was found that HSPH members act as nucleotide exchange factors for both mammalian as well as yeast HSP70 proteins (43). This is surprising as different nucleotide exchange factors (BAG1, HSPBP1) were already identified in the mammalian cytosol (44;45). Interest- ingly, none of the nucleotide exchange factors show significant primary sequence homology to the E. coli nucleotide exchange factor GrpE suggesting that they have evolved independently. However, as HSPH proteins are known to have the capacity to hold (unfolded) proteins in a folding competent state (unlike BAG1 and HSPBP1), they might act as coupling factors between substrate loading and nucleotide exchange for the refolding of specific substrates similar to the coupling of HSPA substrate loading and ATP hydrolysis by DNAJ proteins. Like the other nucleotide exchange factors, HSPH members interact with HSPA members in the ADP configuration and stimulate the dissociation of ADP. The subsequent rebinding of ATP induces the dissociation of HSPH-HSPA complexes (43). By stimulating the nucleotide exchange of the HSPA complex, it was shown that HSPH accelerates the HSPA mediated folding of firefly luciferase (43). As both a heat inducible-substrate holder and heat-inducible nucleotide exchange factor, HSPH may be particularly relevant under (heat) stress conditions during which it may hold substrates (like HSPB) which can be passed on to HSPA for further handling after the stress.

Functional diversities Besides HSPH1 (HSP110), there are 3 other HSPH members in humans: HSPH2 (HSPA4/ APG-2), HSPH3 (HSPA4L/APG-1) and HSPH4 (HYOU1/Grp170). Features of the different HSPH members are summarized in table 3. Currently, it is unknown to what extent the different HSPH members functionally overlap. While one of them, HSPH4, the Grp170 orthologue, is

19 Chapter 1 found in the ER (46) where it is likely to fulfill the role of a nucleotide exchange factor for HSPA5, the other 3 family members are found in the cytosolic/nuclear compartment. HSPH1 and HSPH2 are expressed ubiquitously, whilst HSPH3 is mainly expressed in testis and HSPH3 knockout mice show defects in spermatogenesis (47), suggesting an unique role for this protein within the testis, maybe in conjunction with HSPA1L or HSPA2.

The DNAJ superfamily

All eukaryotic cells contain DNAJ proteins which are known to stimulate the ATPase domain of HSPA chaperones. The common domain that defines this family is the J-domain that stimulates the HSPA ATPase domain. In the human genome, at least 41 different DNAJ encoding genes have been identified (27). The exact protein partners of the different DNAJ family members as well as the exact cellular functions are currently unknown for most members. DNAJ proteins are divided in three sub families (Figure 2); Type A proteins are the closest human orthologues of the E. coli DNAJ and contain, besides an extreme N-terminal J-domain, a glycine/phenylalanine- rich region, a cysteine rich region, and a variable C-terminal domain. Type B proteins contain all the above listed domains with the exception of the cysteine rich region and type C DNAJ proteins contain only the J domain that is not necessarily restricted at the N-terminus but can be positioned at any place within the protein.

Paradigm according to DNAJB1 The J-domain is highly conserved and folded in an a-helical secondary structure. A conserved sequence motif (HPD) in the J domain has been shown to be critical for accelerating the ATPase activity of HSP70. Adjacent to the J domain, a glycine/phenylalanine rich region is believed to function as a flexible spacer that separates the N-terminal J domain from the rest of the molecule. In the centre of the molecule, the cysteine-rich domain contains four cysteine-rich repeats that fold around two zinc atoms. The C-terminal domain folds in a β-plated sheet structure and is involved in dimerization as well as in substrate binding and presentation (48). DNAJA proteins dimerize in a V-like structure (31) and the binding motif of E. coli DNAJ consists of a hydrophobic core of eight residues enriched for arginine, aromatic amino acids and large aliphatic hydrophobic residues positioned in the middle of each monomer. Although each monomer contains only a short binding motif, dimer formation gives rise to a relatively large beta sheet projection. Each of the monomers binds part of the unfolded substrate and hold it in an extended conformation between the middle of the two monomers. HSP70 binds the DNAJ dimer at the tips of the V-like structure and takes over the substrate for binding and release cycles. Al- though DNAJB members also can form dimers, they differ structurally from DNAJA dimers: e.g., whereas DNAJA1 forms compact dimers in which the N- and C-termini face each other, DNA- JB4 forms a dimer in which only the C-termini of the two monomers were in contact (49). These structural differences may very well relate to differences in substrate binding or selection.

Functional diversities Comparative studies on 13 cytosolic J-proteins in yeast (50) revealed that the J domains from a variety of different classes of J proteins could complement the severe growth defects in yeast lacking the DNAJ protein Ydj1. This demonstrates that the stimulation of the ATPase activity of yeast SSA1 is sufficient for many cellular processes. On the contrary, the phenotypes of 4 other DNAJ deletions (cwc23, sis1, jjj1 and jjj3) could only be rescued by expression of the deleted genes indicating that these proteins fulfill highly specialized and unique tasks. For mammals, domain swapping experiments have confirmed that the J-domain can be exchanged between various DNAJ proteins with preservation of the biological function (51). Even more, the J do-

20 Introduction & aim of the thesis main from DNAJB1 could complement the J domain of yeast Ydj1 (52). This all indicates that the J-domain is highly conserved and not likely responsible for any functional diversity. Rather, it suggests that the J-domain is only required to recruit HSPA members to specific microenvi- ronments. Here, the C-termini of the DNAJ-members would provide substrate and/or functional specificity. Clearly,in vitro, type A DNAJ molecules have substrate binding activity (53) and thus may function in substrate delivery to HSPA partners. It is still under debate if type B members also exhibit substrate binding activity, although most studies now support this idea. Whether all DNAJ members can stimulate the nucleotide cycle of HSPA machines and/or have any prefer- ence for specific HSPA members is yet unclear.

The DNAJA family The human genome contains four different members of the DNAJA family and general features are summarized in table 3. Interestingly, it was found that whereas both DNAJA2 and DNAJA4 could stimulate the hydrolysis of ATP on HSPA1A, the DNAJA2-HSPA1A but not the DNAJA4- HSPA1A combination was able to support refolding of denatured luciferase. This indicates that at least DNAJA2 and DNAJA4 may have differential substrate specificity in that DNAJA4 is unable to bind and load denatured luciferase onto the HSPA1 chaperone (54).

The DNAJB family The DNAJB subfamily, DNAJB1 in particular, has been most extensively studied in mammalian cells. It has been found to cooperate with both HSPA1A as well as HSPA8 in luciferase refolding in vitro and in living cells (55). DNAJB4/HLJ1 and DNAJB5 are two DNAJB proteins with unknown function that are close paralogs of DNAJB1. DNAJB4 shows full length homology to DNAJB1 and is also known as Hsc40, a non-heat inducible constitutively expressed member and proposed to act as a housekeeping HSP40/DNAJ protein just as HSPA8/Hsc70 is proposed as the housekeeping equivalent of the stress inducible HSPA1. DNAJB2/HSJ-1 is also relatively well-studied and is expressed as two isoforms. The long isoform is targeted to the cytosolic face of the ER by C-terminal geranylgeranylation while the short isoform is found in the cytosolic and nuclear compartment (56). Both proteins contain an ubiquitin interaction motif which is suggested to sort misfolded clients for HSPA8-dependent proteasomal degradation (57). The DNAJB members DNAJB9/ErdJ4 and DNAJB11/ErdJ3 are ER specific DNAJ members that collaborate with HSPA5 (58;59). Both proteins are induced upon ER stress (58;59). This implies that DNAJB9 and DNAJB11 likely fulfill equivalent roles to that of DNAJB1 in the cytosol. DNAJB6/MRJ-1, DNAJB7 and DNAJB8 are three homologous proteins which share, besides a J domain, high sequence homology in the C-terminus. This C-terminal domain does not show any homology with known domains in the Pfam database (data not shown). DNAJB6 binds keratin 18 and was found to be important to prevent toxic keratin aggregation which interferes with placental development (60). DNAJB6 has also been shown to suppress the toxic aggregation of mutant Huntingtin, supporting the idea that (some) DNAJB members can bind substrates. Whereas this holds for DNAJB7 and DNAJB8 as well and whether this requires collaboration with HSPA machines remains to be elucidated. Finally, there are 3 more diverse DNAJB members, DNAJB12, DNAJB14 and a testis-specific DNAJB13/Tsarg3 that have hardly been studied. DNAJB12 and DNAJB14 show high sequence similarity in the C-terminus and contain the C-terminal DUF1977 (Domain of unknown function 1977) domain. DNAJB13 contains a C-terminal domain homologous to DNAJB1, DNAJB4 and DNAJB5 (Pfam, data not shown). General features of the DNAJB family are summarized in table 3.

21 Chapter 1

The DNAJC family With over 23 members, the DNAJC family represents the largest of the three subfamilies. The family is very diverse in both amino acid composition as well as protein length (Table 3) with the DNAJ domain being the only common feature (Figure 2). Only a dozen of proteins have been studied, including DNAJC1/ErDj1, SEC63/ErDj2, DNAJC10/ErDj5 proteins, all ER specific DNAJ proteins that associate with HSPA5. It is currently unknown to what extent these HSPA5 factors show functional overlap or specificity. Another member, DNAJC19/TIMM14 is part of the mitochondrial TIM23 preprotein translocase that stimulates the ATPase activity of the mitochondrial HSPA protein (HSPA9) hereby supporting mitochondrial import of nuclear encoded proteins. DNAJC21/ZRF1 is the ortholog of the yeast DNAJ protein called Zuo1, a ribosome associated DNAJ protein important for translation in yeast. Furthermore, several other DNAJC members such as DNAJC5/CSP, DNAJC6/auxilin and DNAJC13/RME-8 seem to function in endocytosis and exocytosis. Both DNAJC6 and DNAJC13 have been shown to collaborate with HSPA8 in the process of endocytosis.

Summary & perspectives The human genome not only codes for a wide variety of HSP families, but also for a large number of individual proteins within each of these families. Whilst the diversity is started to be appreciated by a number of investigators, we yet only have faint clues why such diversity exists. Housekeeping members may be primarily involved in co-translational folding of proteins and/or transport of proteins across membranes, whereas some (inducible) members may perform more stress related functions. Substrate specificity to the HSPA machine(s) may be -in part- evoked by specific co-factors (DNAJs and HSPH members). The different structures of the DNAJA and DNAJB C-termini may provide such possibilities for client specificities, but whether mammalian cells also make use of specific combinations between HSPA and DNAJ members remains to be elucidated. Fate determinations on client processing (towards folding or degradation) may not depend on this HSPA machine as such, although the presence of HSPA cofactors such as CHIP and BAG1 provide an easy link to the proteasome. Intriguingly, although client binding to some HSPB members can result in HSPA dependent improved folding of (stress-denatured) clients, several HSPB members are linked to client degradation (proteasomal or autophagy) in both HSPA-dependent and –independent manners. The ability of nearly all HSPB members to associate with cytoskeletal elements increasing their stability suggests that HSPB members act as cytoskeleton-specific chaperones. In addition, association of HSPBs to cytoskeletal elements may allow them to chaperone and transport un- or misfolded cytosolic proteins towards protein storage and/or degradation routes. Yet, much of this is still only speculative. The ability of chaperones to handle unfolded proteins has challenged many researchers to test whether they could be used for preventing progression of protein folding diseases. Although this concept has had some support from in vitro work, so far the results with animal models have had limited success. More insight into the function of other individual HSP family members may help to find better suppressors and/or strategies for these diseases. On the other hand, the finding that several neuropathies are related to mutations in HSPs (chaperonopathies) further supports their importance in neurodegeneration. Investigations on the molecular biochemical mechanisms by which these mutations lead to chaperonopathies will lead to better understanding of these neuropathologies and also increase our understanding of the normal function of the corresponding chaperones. Furthermore, searches for mutations in other family members or for alternative transcripts of individual HSP-genes may lead to identification of causes for non-idiopathic cases of neuropathies or age-related decline in protein quality control and cellular ageing.

22 Introduction & aim of the thesis

Aim of the thesis

In this thesis we explore the functional capacities of both the human (HSPB) and Drosophila small heat shock proteins and their requirement for the HSP70 machine in functioning. Further- more, we determine possible candidates to ameliorate neurodegenerative diseases and boost cellular health to increase lifespan.

23 Chapter 1

y y y y h t h h h h i t t t p a a

w o

s s p s p s r , r r s r t o o e s u e u ur u r r ur yopa t y o o o o u u o a m i e e m m m m m c seas seas e, i i u u u u u

o ar d

t t t r n r n t e d t d l

a d

o o

n n n a t n t t n h for i l a o th t

o o

t a a i a i n a i scu a i

o o t t t e t t o u r l m re m y o i

t e e cus n i l T T o m y y i c cer cer cer cer l

at - - r r

i m a n e e lo a a m n n n

i n n a i i i i i i s i i

i it it

G d n a r a r d d d

soc d d o d d e e e r e s e e e ility 0 e e t r t t t M t M t p , De

b 0 a

- a - a h a a t a t a l t t l l l 8 l ti l e e c dosis her o her o u s 5 p a op l l gu c c g o r gu gu 7 gu gu z a a l e i e t t a rac a e e e e 0 r ar a r c t r r r s r s seas r sce

2 i i i p h a t a h p p p p p u s n i u s sch sarco i

u u

D u C D C C C D U

s y h a x p B o r ate r t t a ,

ubiquitin, e s s c n

i y l l s e t ub a a na mm t t n nts, i S

a m c d e e

e k i l l /

1 ga

3 s uous s uous T n -f i i 3 - 0 i on 6 - t a pon pon e ske ske e F G m m t 3 o e nts nL o o p9 - m o o o T y y ph li A r r r 4 l yt yt

p p

B p D m a cytochro m

co co m C

1 c c Hs

s s s s s

u u u u u us us us us s ules e e e e e e e l e e

l l l l l n l l i l l

c c c c c a ia r asma t n r u u u u u l

l o d e n n n n n

p nuc nuc nuc nuc gra l i

m n n / /

/ / /

/ / / e / /

a n o at l l l l c l l l l l l lic l l l l l w a

h l o o o o o o o o o br c s s s s s s so so s so s liz so p so so so so m n

o r o o o o o o o o m o o o a o o o o o o t t t t t k t t t e c e R R y y y y y n y yt yt yt y y yt yt yt yt p o c c E c

c E c c c c u m cyt c cyt mit c l Subcellul a c s c c c c

n i

l l s ng , a a u t t u

l e e o , l l t

a , s, i

r e , bra e e r in l e e u k k

n, v q s s s s s s i sc

i i ali v u u u u us us tion us us liv e u

n d d b S o o o o o s

o L o c y n n u t t t , m

e e e e , bra it n i i i ito it it ito a a

l l y ,

ow e u u u u n s s tin s , bra t t t c s s l i n i i

i q q q q qu qu a r s sc l sc l qu sc qu t t ti ti t i a i i i i i i i k r a e v s s u u u u r s b b b b n y e e e s e s e O u t B t spleen u kidney, u u ub ub t k e placenta m m m t Tissue distrib u ub e t hear m hear hear ub u

1 3 e . . 2

. 3 4

1 1 2 3 . 3

3 q p 33 s o 4

2 - 3 q p 3 1

1 - 3

n q

3 3

1 3 2 - . . . .2 . . - . . q 3 1 3 3 o 2 3 2 23 1 m o 4 -

. 3 . 2 . .

. 1 2 2 2 3 . 4 . . . o 2 3 3 1 2 1 1 1 1 8 1 2 2 2 2 1 6 2 1 2 2 q 3 2 2 q 3 q 2 2 2 1 q q q q 3 q 2 cati h r q 4 q 1 p q 3 p p q q 7q 1 1 1 9 p 2 q1 q p3 o 9 1 1 1

6 5 1 6 6 4 5 1 1 2 C l 1 1 1 1 7 8 1

r ) a a

, 0 0 3 4 9 9 4 , , ,0 e (kD , 1 , ,0 , ,0 ,5

,5 ,2 , ,2 ,1 ,6 ,6 ,8 , ,3 z 1 0 1 0 1 4 0 6 0 4 7 7 9 0 1 0 7 8 2 8 6 7 7 7 7 1 9

7 9 7 9 1 1 1 2 2 Molecu l 2 1 1 si 2 1 2

) ) ) y

% %

0 0 0 % % % % % % % % % % % % % % % % % % 0 0 0 3 0 1 8 5 3 9 8 3 6 8 9 4 6 4 0 3 7 1 1 1 6 8 8 8

( 2 ( 6 9 5 3 3 1 2 3 3 3 2 Sequenc e identit ( 1 1

A B

0 5 6 2 1 1 1 2 4 1 3 4 5 9 3 8 2 6 7 1 11 1 A A A A H H H A A H B B B B B B B B B B B P P P P P P P P P P P P P S S S S S S S H HS HS H HS HS HS H HS H H HSP HSP H HSP HSP HSP HSP

HSP H HSP Table 3. Properties of the Human HSP Families 3. Properties of the Human HSP Table

24 Introduction & aim of the thesis

na s s s o r ur u o e m n u

t d

a u

e n

o t for o i

a t n a i r l

t o i y i

d t

va cus atheroscleros i atheroscleros i

i y cer t t at eases

i s s

h m s lo n t i a n i n i i i i

x d a a n

soc with with d o

s e g g r g e opa ility s s r t a a p n y b e

e a

h i a t t s l s e ti e d t m t a a l s i p i op c o gu o c c i z e ec l i t e t o o sfo c seas r sce s s o o

i i r p r s s u n

sch s

card

u p

s , it

n s u tei ors b t o r u s r ate ac p

t f

a s t h

ub be 1

w us u 8 S n o ndrial

C o i / r o - 1 E e s uous s uous s u s uo t h g i i i i B c o tin y

r m o m m m e nts O 3 p o o o o a n li

P 5 r r r r

mit

C G

A kera p p m

us us

e e

l l a

r s r n s R e

d o e

nuc nuc i n n n E

n ndria n n n n n n n n n m n n

/ /

o /

l l l a w w at w o l l l w w w w w w w w w w r h s o ho o o o o o o o o o o o o o o b c so s c o liz n n n s so so n n n n n n n n n n r o oso o o o

k k a o o m o t c k k k k k k k k k k t i c i e n n nk R R n n n n n n n n n n y yt yt o u u mit u E u U u u U u u Cyt Subcellul a l u U u

Cyt M c m E c Cyt c m

n i

r ver n, ver zoa, i y i i etal b o

, L , a r r

, a e a t

y b t l n , ll

leen r er i t

c m m Ske l r a s s s a s s v s s

a ,

mm r tion ce u u testis S p u us us u us us u er latelet a dne u n n li n n n a li li l i

o o B o o p P B e He

t t hea p w

w w

w w m

m s

i i ito ito ito

ito ito it it k

he h

o , o o y o o , t reas, t , , u u ey, d u t d, t n, r s s s e ll v s n s n n n n n n o i o i i r q q scle, n qu scle, qu o i qu qu qu i qu q a n t t u u i i k k i i k i k i i k i a i d d a C a a d o a u u t t s s

r n b b n n n i l n r e e e e u u t hear m Ov Sali u u Pan c m en u ub t ub g l Tissue distrib u k testi b u B

ub ub u ub B ub en ub h b f br fe

2 e . 1 m 1

) q s o 4

m n 12 3 2

2 )

1 1

. .

.1 .2 . .3 . .4 o q 3 3 1 1 3 2 m o M 3

. . 4 6 . 2

. -p 1 .

. 4 - 3 3 1 3 4 3 m ( o 1 1 1 3 1 2 1 2 3 1 3 1 8 6 3 2 2 1 1 1 1 2 1

M q 3 2 2 p q 3 q 2 q 3 1 3 q 3 p p q p q q cati ( D h r

2 q q q 0 0

0 p q 1 q p1 q q q2 p 1 0 9 6 6 5 1 o 1 2 7 1 3 1 1 1 1 1 4 l C 5 2 9 7

2 9 1 3 2 1 1 1 1 1

r )

a a 6 5 6 , , ,9 , 8 3 6 0

2 3 2 3 /

5 6 4 7 7 8 9 5 9 6 7 2 7 7 e (kD , , , , 1 , ,8 , ,7 , ,1 ,1 , , , ,9 ,9 ,5 , , ,5 , ,1 , D z 5 0 5 5 4 6 7 4 3 1 2 9 6 5 1 0 4 0 5 8 2 5 6 4 2 3 3 2 1 N 2 7 3 5 4 3 3 3 7 5 Molecu l si 4 4 4

7 3 5 4 3 4

) ) y

0 0 % % % % % % % % % % % % % % % % % % % % % % 0 0 D 3 9 4 3 4 5 4 5 2 7 5 7 8 7 0 7 0 8 4 4 8 3 1 1 1 2 2 2 1 N 4 2 6 6 2 8 2 identit ( 2 1 2 Sequenc e 1

3 1 ( 2 5 4 7

2 4

B 9 8 7 1 3 6 5 4 2 11 1 1 13 1 2

4 5 2 B B B B B B B B 1 9 7 B B B 8 B B B 1 1 1 A1 A3 A2 A4 J J J J J J J J J J J J A A A A A A A A A A A AJ AJ AJ A AJ A A A A A AJ A AJ A P P P P P P P N N N N N N N S N N S N N N S N N N N N N HS D D HS D HS D D D HS D D D H H D D Protein D D D D H D D D Table 3. Continued Table

25 Chapter 1

References

1. Shashidharamurthy, R., Koteiche, H. A., Dong, J., and McHaourab, H. S. (2005) Mechanism of chaperone function in small heat shock proteins: dissociation of the HSP27 oligomer is required for recognition and binding of destabilized T4 lysozyme, J. Biol. Chem. 280, 5281-5289. 2. Ito, H., Kamei, K., Iwamoto, I., Inaguma, Y. and others (2001) Phosphorylation-induced change of the oligomerization state of alpha B-crystallin, J. Biol. Chem. 276, 5346-5352. 3. Mayer, M. P. and Bukau, B. (2005) Hsp70 chaperones: cellular functions and molecular mechanism, Cell Mol. Life Sci. 62, 670-684. 4. van Montfort, R. L., Basha, E., Friedrich, K. L., Slingsby, C. and others (2001) Crystal structure and assembly of a eukaryotic small heat shock protein, Nat. Struct. Biol. 8, 1025-1030. 5. Bryantsev, A. L., Kurchashova, S. Y., Golyshev, S. A., Polyakov, V. Y. and others (2007) Regulation of stress-induced intracellular sorting and chaperone function of Hsp27 (HspB1) in mammalian cells, Biochem. J. 407, 407-417. 6. Haslbeck, M., Miess, A., Stromer, T., Walter, S. and others (2005) Disassembling protein aggregates in the yeast cytosol. The cooperation of Hsp26 with Ssa1 and Hsp104, J. Biol. Chem. 280, 23861-23868. 7. Landry, J. and Huot, J. (1995) Modulation of actin dynamics during stress and physiological stimulation by a signaling pathway involving p38 MAP kinase and heat-shock protein 27, Biochem. Cell Biol. 73, 703-707. 8. Evgrafov, O. V., Mersiyanova, I., Irobi, J., Van Den, B. L. and others (2004) Mutant small heat-shock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy, Nat. Genet. 36, 602-606. 9. Vicart, P., Caron, A., Guicheney, P., Li, Z. and others (1998) A missense mutation in the alphaB-crystallin chaperone gene causes a desmin-related myopathy, Nat. Genet. 20, 92-95. 10. Feil, J., Malfois, M., Hendle, J., van der Zandt, H. and others (2001) A novel quaternary structure of the dimeric alpha- crystallin domain with chaperone-like activity, J. Biol. Chem. 276, 12024-12029. 11. Ahmad, M. F., Raman, B., Ramakrishna, T., and Rao, C. (2008) Effect of phosphorylation on alpha B-crystallin: differences in stability, subunit exchange and chaperone activity of homo and mixed oligomers of alpha B-crystallin and its phosphorylation-mimicking mutant, J. Mol. Biol. 375, 1040-1051. 12. Chavez Zobel, A. T., Loranger, A., Marceau, N., Theriault, J. R. and others (2003) Distinct chaperone mechanisms can delay the formation of aggresomes by the myopathy-causing R120G alphaB-crystallin mutant, Hum. Mol. Genet. 12, 1609-1620. 13. den, E. J., Gerrits, D., de Jong, W. W., Robbins, J. and others (2005) Nuclear import of {alpha}B-crystallin is phosphorylation-dependent and hampered by hyperphosphorylation of the myopathy-related mutant R120G, J. Biol. Chem. 280, 37139-37148. 14. Shemetov, A. A., Seit-Nebi, A. S., Bukach, O. V., and Gusev, N. B. (2008) Phosphorylation by Cyclic AMP-Dependent Protein Kinase Inhibits Chaperone-Like Activity of Human HSP22 in vitro, Biochemistry (Mosc. ) 73, 200-208. 15. Carra, S., Seguin, S. J., Lambert, H., and Landry, J. (2008) HspB8 chaperone activity toward poly(Q)-containing proteins depends on its association with Bag3, a stimulator of macroautophagy, J. Biol. Chem. 283, 1437-1444. 16. Irobi, J., Van, I. K., Seeman, P., Jordanova, A. and others (2004) Hot-spot residue in small heat-shock protein 22 causes distal motor neuropathy, Nat. Genet. 36, 597-601. 17. Kasakov, A. S., Bukach, O. V., Seit-Nebi, A. S., Marston, S. B. and others (2007) Effect of mutations in the beta5- beta7 loop on the structure and properties of human small heat shock protein HSP22 (HspB8, H11), FEBS J. 274, 5628-5642. 18. Suzuki, A., Sugiyama, Y., Hayashi, Y., Nyu-i N and others (1998) MKBP, a novel member of the small heat shock protein family, binds and activates the myotonic dystrophy protein kinase, J. Cell Biol. 140, 1113-1124. 19. Sugiyama, Y., Suzuki, A., Kishikawa, M., Akutsu, R. and others (2000) Muscle develops a specific form of small heat shock protein complex composed of MKBP/HSPB2 and HSPB3 during myogenic differentiation, J. Biol. Chem. 275, 1095-1104. 20. Beall, A., Bagwell, D., Woodrum, D., Stoming, T. A. and others (1999) The small heat shock-related protein, HSP20, is phosphorylated on serine 16 during cyclic nucleotide-dependent relaxation, J. Biol. Chem. 274, 11344-11351. 21. Chernik, I. S., Seit-Nebi, A. S., Marston, S. B., and Gusev, N. B. (2007) Small heat shock protein Hsp20 (HspB6) as a partner of 14-3-3gamma, Mol. Cell Biochem. 295, 9-17. 22. Doran, P., Gannon, J., O'Connell, K., and Ohlendieck, K. (2007) Aging skeletal muscle shows a drastic increase in the small heat shock proteins alphaB-crystallin/HspB5 and cvHsp/HspB7, Eur. J. Cell Biol. 86, 629-640. 23. de Wit, N. J., Verschuure, P., Kappe, G., King, S. M. and others (2004) Testis-specific human small heat shock protein HSPB9 is a cancer/testis antigen, and potentially interacts with the dynein subunit TCTEL1, Eur. J. Cell Biol. 83, 337- 345. 24. Fontaine, J. M., Rest, J. S., Welsh, M. J., and Benndorf, R. (2003) The sperm outer dense fiber protein is the 10th member of the superfamily of mammalian small stress proteins, Cell Stress. Chaperones. 8, 62-69. 25. Bellyei, S., Szigeti, A., Boronkai, A., Pozsgai, E. and others (2007) Inhibition of cell death by a novel 16.2 kD heat shock protein predominantly via Hsp90 mediated lipid rafts stabilization and Akt activation pathway, Apoptosis. 12, 97-112. 26. Brocchieri, L., Conway de, M. E., and Macario, A. J. (2008) hsp70 genes in the human genome: conservation and differentiation patterns predict a wide array of overlapping and specialized functions, BMC. Evol. Biol. 8, 19. 27. Qiu, X. B., Shao, Y. M., Miao, S., and Wang, L. (2006) The diversity of the DnaJ/Hsp40 family, the crucial partners for

26 Introduction & aim of the thesis

Hsp70 chaperones, Cell Mol. Life Sci. 63, 2560-2570. 28. Swain, J. F., Dinler, G., Sivendran, R., Montgomery, D. L. and others (2007) Hsp70 chaperone ligands control domain association via an allosteric mechanism mediated by the interdomain linker, Mol. Cell 26, 27-39. 29. Freeman, B. C., Myers, M. P., Schumacher, R., and Morimoto, R. I. (1995) Identification of a regulatory motif in Hsp70 that affects ATPase activity, substrate binding and interaction with HDJ-1, EMBO J. 14, 2281-2292. 30. Kampinga, H. H. (2006) Chaperones in preventing protein denaturation in living cells and protecting against cellular stress, Handb. Exp. Pharmacol. 1-42. 31. Stirling, P. C., Bakhoum, S. F., Feigl, A. B., and Leroux, M. R. (2006) Convergent evolution of clamp-like binding sites in diverse chaperones, Nat. Struct. Mol. Biol. 13, 865-870. 32. Rudiger, S., Buchberger, A., and Bukau, B. (1997) Interaction of Hsp70 chaperones with substrates, Nat. Struct. Biol. 4, 342-349. 33. Daugaard, M., Rohde, M., and Jaattela, M. (2007) The heat shock protein 70 family: Highly homologous proteins with overlapping and distinct functions, FEBS Lett. 581, 3702-3710. 34. Fourie, A. M., Peterson, P. A., and Yang, Y. (2001) Characterization and regulation of the major histocompatibility complex-encoded proteins Hsp70-Hom and Hsp70-1/2, Cell Stress. Chaperones. 6, 282-295. 35. Dix, D. J., Allen, J. W., Collins, B. W., Mori, C. and others (1996) Targeted gene disruption of Hsp70-2 results in failed meiosis, germ cell apoptosis, and male infertility, Proc. Natl. Acad. Sci. U. S. A 93, 3264-3268. 36. Noonan, E. J., Place, R. F., Giardina, C., and Hightower, L. E. (2007) Hsp70B' regulation and function, Cell Stress. Chaperones. 12, 393-402. 37. Choglay, A. A., Chapple, J. P., Blatch, G. L., and Cheetham, M. E. (2001) Identification and characterization of a human mitochondrial homologue of the bacterial co-chaperone GrpE, Gene 267, 125-134. 38. Weitzmann, A., Volkmer, J., and Zimmermann, R. (2006) The nucleotide exchange factor activity of Grp170 may explain the non-lethal phenotype of loss of Sil1 function in man and mouse, FEBS Lett. 580, 5237-5240. 39. Otterson, G. A., Flynn, G. C., Kratzke, R. A., Coxon, A. and others (1994) Stch encodes the 'ATPase core' of a micro- somal stress 70 protein, EMBO J. 13, 1216-1225. 40. Han, Z., Truong, Q. A., Park, S., and Breslow, J. L. (2003) Two Hsp70 family members expressed in atherosclerotic lesions, Proc. Natl. Acad. Sci. U. S. A 100, 1256-1261. 41. Wan, T., Zhou, X., Chen, G., An, H. and others (2004) Novel heat shock protein Hsp70L1 activates dendritic cells and acts as a Th1 polarizing adjuvant, Blood 103, 1747-1754. 42. Albanese, V., Yam, A. Y., Baughman, J., Parnot, C. and others (2006) Systems analyses reveal two chaperone networks with distinct functions in eukaryotic cells, Cell 124, 75-88. 43. Raviol, H., Sadlish, H., Rodriguez, F., Mayer, M. P. and others (2006) Chaperone network in the yeast cytosol: Hsp110 is revealed as an Hsp70 nucleotide exchange factor, EMBO J. 25, 2510-2518. 44. Takayama, S., Bimston, D. N., Matsuzawa, S., Freeman, B. C. and others (1997) BAG-1 modulates the chaperone activity of Hsp70/Hsc70, EMBO J. 16, 4887-4896. 45. Shomura, Y., Dragovic, Z., Chang, H. C., Tzvetkov, N. and others (2005) Regulation of Hsp70 function by HspBP1: structural analysis reveals an alternate mechanism for Hsp70 nucleotide exchange, Mol. Cell 17, 367-379. 46. Tsukamoto, Y., Kuwabara, K., Hirota, S., Kawano, K. and others (1998) Expression of the 150-kd oxygen-regulated protein in human breast cancer, Lab Invest 78, 699-706. 47. Held, T., Paprotta, I., Khulan, J., Hemmerlein, B. and others (2006) Hspa4l-deficient mice display increased inci- dence of male infertility and hydronephrosis development, Mol. Cell Biol. 26, 8099-8108. 48. Cheetham, M. E. and Caplan, A. J. (1998) Structure, function and evolution of DnaJ: conservation and adaptation of chaperone function, Cell Stress. Chaperones. 3, 28-36. 49. Borges, J. C., Fischer, H., Craievich, A. F., and Ramos, C. H. (2005) Low resolution structural study of two human HSP40 chaperones in solution. DJA1 from subfamily A and DJB4 from subfamily B have different quaternary structures, J. Biol. Chem. 280, 13671-13681. 50. Sahi, C. and Craig, E. A. (2007) Network of general and specialty J protein chaperones of the yeast cytosol, Proc. Natl. Acad. Sci. U. S. A 104, 7163-7168. 51. Stubdal, H., Zalvide, J., and DeCaprio, J. A. (1996) Simian virus 40 large T antigen alters the phosphorylation state of the RB-related proteins p130 and p107, J. Virol. 70, 2781-2788. 52. Yan, W. and Craig, E. A. (1999) The glycine-phenylalanine-rich region determines the specificity of the yeast Hsp40 Sis1, Mol. Cell Biol. 19, 7751-7758. 53. Rudiger, S., Schneider-Mergener, J., and Bukau, B. (2001) Its substrate specificity characterizes the DnaJ co-chaperone as a scanning factor for the DnaK chaperone, EMBO J. 20, 1042-1050. 54. Hafizur, R. M., Yano, M., Gotoh, T., Mori, M. and others (2004) Modulation of chaperone activities of Hsp70 and Hsp70-2 by a mammalian DnaJ/Hsp40 homolog, DjA4, J. Biochem. 135, 193-200. 55. Michels, A. A., Kanon, B., Konings, A. W., Ohtsuka, K. and others (1997) Hsp70 and Hsp40 chaperone activities in the cytoplasm and the nucleus of mammalian cells, J. Biol. Chem. 272, 33283-33289. 56. Chapple, J. P. and Cheetham, M. E. (2003) The chaperone environment at the cytoplasmic face of the endoplasmic reticulum can modulate rhodopsin processing and inclusion formation, J. Biol. Chem. 278, 19087-19094. 57. Westhoff, B., Chapple, J. P., van der, S. J., Hohfeld, J. and others (2005) HSJ1 is a neuronal shuttling factor for the sorting of chaperone clients to the proteasome, Curr. Biol. 15, 1058-1064. 58. Shen, Y., Meunier, L., and Hendershot, L. M. (2002) Identification and characterization of a novel endoplasmic re-

27 Chapter 1

ticulum (ER) DnaJ homologue, which stimulates ATPase activity of BiP in vitro and is induced by ER stress, J. Biol. Chem. 277, 15947-15956. 59. Shen, Y. and Hendershot, L. M. (2005) ERdj3, a stress-inducible endoplasmic reticulum DnaJ homologue, serves as a cofactor for BiP's interactions with unfolded substrates, Mol. Biol. Cell 16, 40-50. 60. Watson, E. D., Geary-Joo, C., Hughes, M., and Cross, J. C. (2007) The Mrj co-chaperone mediates keratin turnover and prevents the formation of toxic inclusion bodies in trophoblast cells of the placenta, Development 134, 1809- 1817.

28 CHAPTER 2

A PCR amplification strategy for unrestricted generation of chimeric genes

Michel J Vos & Harm H Kampinga

Department of Cell Biology, Section of Radiation and Stress Cell Biology, University Medical Center Groningen, Groningen, The Netherlands

Published in Analytical Biochemistry, 2008 Sep 15:380(2):338-40 Chapter 2

Abstract

For analyzing protein function, protein dynamics or protein-protein interactions, the use of chimeric proteins has become an indispensable tool. The generation of DNA constructs coding for such fused proteins can, however, be a tedious process. Currently used strategies often make use of available endonuclease restriction sites leading to limitations in the choice on the site of fusion between two genes and problems on how to maintain protein secondary structures. We have developed a cloning strategy to get around these disadvantages that is based on a single round of PCR amplification followed by antibiotic resistant gene complementation.

30 A PCR strategy for chimeric gene construction

Introduction & results

Methods for creating either tagged or chimeric proteins usually depend on a combination of PCR amplification of the gene of interest followed by endonuclease mediated cloning into a plasmid. The generation of chimeric genes is often performed using a combination of available endonuclease sites (1) or the introduction of required endonuclease sites (2). Although often efficient, especially the choices at which site in the gene a chimeric gene can be produced is restricted by the presence of endonuclease recognition sites and often requires the introduction of additional bases. Also, several rounds of PCR amplification and cloning are required before a desirably construct is obtained. Furthermore, cloning of genes in frame with for instance, a fluorescent protein encoding gene, always introduces additional aminoacids between both proteins originating from the multiple cloning site of the plasmid used. Here we report on the design of a cloning strategy which, independently of the absence or presence of specific endonuclease recognition sites, allows for fast and reliable generation of chimeric genes in one single round of PCR amplification, without the introduction of additional

Insert A Insert B

Ampicilin R Ampicilin R

1 PCR amplification & DpnI digestion of template

2 T4 kinase & T4 ligase treatment

3 E. coli transformation & selection

Figure 1. Schematic outline of the strategy for generating chimeric genes. After a single round of PCR amplification of Insert A or Insert B together with the N-terminal or C-terminal part of the Ampicillin gene respectively, methylated DNA template was removed by DpnI digestion (1). Secondly, both fragments are combined and treated with T4 kinase and T4 ligase (2). Selection of the correctly ligated product is done in E. coli by exploiting the formation of a correct resistance gene and hence acquired antibiotic resistance (3). aminoacids between the fused proteins. Furthermore, our strategy allows to freely choose where to make a fusion between two genes, thus providing means to avoid disruption of secondary protein structure. In addition, it utilizes antibiotic resistant gene complementation to select for the correct fusion construct in E. coli. This results in a 100% efficiency in obtaining the correct ligated product. Our strategy (Figure 1) requires the availability of two plasmids with the same backbone, one containing gene A and the other gene B or a tag. In the first step sequence A is amplified together with the disired part of the vector backbone including half of the Ampicillin resistance gene. The other part for the fusion plasmid is amplified containing the desired part of sequence B together with the vector backbone including the complementing half of the Ampicillin resist-

31 Chapter 2

A Pto AMP Pto AMP Pto AMP

pcDNA5/FRT/TO-ECFP pcDNA5/FRT/TO-HSPB7 pcDNA5/FRT/TO-ECFP-HSPB7 5851 bps ECFP 5649 bps HSPB7 6339 bps ECFP-HSPB7

HYGRO HYGRO HYGRO

B Primers Sequence (5' to 3') E AmpN ATGCTTTTCTGTGACTGGTG AmpC CTTACGGATGGCATGACAGT CFP/1 CTTGTACAGCTCGTCCATGCC B7/1 ATGAGCCACAGAACCTCTTC CFP-HSPB7

C D

 DNA HSPB7 CFP-HSPB7

Pst1 AmpN-CFP/1 2172bp AmpC-B7/1 4167bp 55kDa 4749bp DAPI

26kDa 1159bp

Figure 2. Fusion of HSPB7 with ECFP and expression in HELA S3 cells.To generate a fusion between ECFP and HSPB7 (panel A), primers were selected as indicated in panel B. The PCR products (panel C) were ligated and transfected in HELA S3 cells. After 48hrs, protein expression was analyzed by Western- blotting (panel D) using monoclonal HSPB7 antibody (Abnova). The fusion protein was visualized using fluorescent microscopy (panel E), showing localization in both the cytosol and the nucleus. The nucleus was visualized using DAPI staining. ance gene. Next, both fragments are digested for two hours with DpnI to allow for methylated template removal from the PCR reaction. This effectively reduces the chance of original vector contamination during E. coli transformation. Then, both fragments are mixed and treated with T4 kinase and T4 ligase for two hours. Selection for the correct ligated product is achieved by the functional restoration of the bacterial resistance gene. This leads to a 100% efficiency in obtaining clones harboring the correctly fused sequences. The same process can be performed with plasmids containing any other bacterial selection gene. As the choice for primer binding sites on both gene A and gene B is unrestricted, a chimeric gene can be generated at any place so that disruption of secondary protein structure can be avoided. As an example, we generated a fusion between the Enhanced Cyan Fluorescent Protein (ECFP) and HSPB7, a small heat shock protein. The plasmids used were pcDNA5/FRT/TO-ECFP and pcDNA5/FRT/TO-HSPB7. Primers used for PCR on pcDNA5/FRT/TO-ECFP were AmpN and CFP/1 lacking the stop codon. For pcDNA5/FRT/TO-HSPB7, primer AmpC was used in combination with B7/1 (Figure 2a,b). The PCR reactions were carried out in a total volume of 50uL containing 10-30 ng plasmid template, 200 mM of each dNTP, 5 mL Pfuturbo buffer, 30 pmol of each primer, 3 mL DMSO and 2,5U of Pfuturbo DNA polymerase (Stratagene). The amplification

32 A PCR strategy for chimeric gene construction reaction was started with a 1 minute incubation at 95 0C followed by 19 cycles of denaturation at 95 0C for 50 seconds, annealing at 55 0C for 50 seconds and amplification at 680 C for 1 minute per kb. The reaction was ended by a final incubation step for 8 minutes at 68 0C. After amplification, the fragments (Figure 2c) were treated with DpnI for two hours and ligated at room temperature for two hours using 5U of T4 DNA ligase in the presence of 10U T4 kinase buffered by T4 DNA ligase buffer (Invitrogen). Several clones were selected for restriction analysis and sequenced, all of which were correct. Transfection of the new fusion plasmid in mammalian cells clearly shows correct expression of the fusion protein by Western blot analysis (Figure 2d). This construct was effectively used to visualize ECFP-HSPB7 in HELA cells using fluorescent microscopy (Figure 2e).

In conclusion, our strategy allows for fast and efficient generation of chimeric genes or tagged genes using an internal selection for appropriately ligated fragments of a functional bacterial resistance gene. Furthermore, it circumvents the need for introducing appropriate endonuclease recognition sequences and allows for the fusion of two sequences taking into account the preservation of secondary protein structures.

Acknowledgements

This work was supported by Innovatiegerichte Onderzoeksprogramma Genomics. Grant IGE03018

References

1. Kageyama, M., Kobayashi, M., Sano, Y., and Masaki, H. (1996) Construction and characterization of pyocin-colicin chimeric proteins, J. Bacteriol. 178, 103-110. 2. Tovar-Palacio, C., Bobadilla, N. A., Cortes, P., Plata, C. and others (2004) Ion and diuretic specificity of chimeric proteins between apical Na(+)-K(+)-2Cl(-) and Na(+)-Cl(-) cotransporters, Am. J. Physiol Renal Physiol 287, F570- F577.

33 Chapter 2

34 CHAPTER 3

HSPB7 is a SC35 speckle resident small heat shock protein

Michel J Vos, Bart Kanon & Harm H Kampinga

Department of Cell Biology, Section of Radiation and Stress Cell Biology, University Medical Center Groningen, Groningen, The Netherlands

In revision for Biochimica et Biophysica Acta – Molecular Cell Research. Chapter 3

Abstract

The HSPB family is one of the more diverse families within the group of HSP families. Some members have chaperone-like activities and/or play a role in cytoskeletal stabilization. Some members also show a dynamic, stress-induced translocation to SC35 splicing speckles. If and how these features are interrelated and if they are shared by all members is yet unknown. Tissue expression data and interaction and co-regulated gene expression data of the human HSPB members was analysed using bioinformatics. Using a gene expression library, sub-cellular distribution of the diverse members was analyzed by confocal microscopy. Chaperone activity was measured using a cellular luciferase refolding assay. Results: Online databases did not accurately predict the sub-cellular distribution of all the HSPB members. A novel and non-predicted finding was that HSPB7 constitutively localized to SC35 splicing speckles, driven by its N-terminus. Unlike HSPB1 and HSPB5, that chaperoned heat unfolded substrates and kept them folding competent, HSPB7 did not support refolding. Our data suggest a non-chaperone-like role of HSPB7 at SC35 speckles. The functional divergence between HSPB members seems larger than previously expected and also includes non- canonical members lacking classical chaperone-like functions.

36 HSPB7 localizes to SC35 speckles

Introduction

The human small heat shock protein (small HSP, HSPB) family of chaperones contains a total of eleven family members (1;2). All HSPB proteins are characterized by a conserved crystallin domain flanked by a variable sized N- and C-terminus resulting in a molecular size of approximately 16-40kDa. Another well-described characteristic is their ability to oligomerize into large, spherical and symmetrical structures (3). The main feature of many small HSPs is their ability to interact with components of the cytoskeleton (4-8) and to protect the cytoskeleton during stress (9). In cells, HSPB1 has been found to affect cell motility and morphology, related to its interaction with cytoskeletal elements (10). Furthermore, HSPB1 seems to fulfill a protective roles in heart diseases like ischemia (11;12) and atrial fibrillation (13) preventing contractile elements against becoming dysfunctional. One other well-defined molecular function of HSPB proteins is their ability to prevent aggregate formation of denatured proteins in vitro (14;15). In vivo, this basic function is part of the cellular pathways accommodating protein folding and degradation (16;17). Active processing of HSPB bound substrates occurs in collaboration with HSPA (HSP70) members which can lead to either the refolding or degradation of the substrate (2;17). The vital role of HSPB members as chaperones and/or protectors of the cytoskeleton is also apparent from several genetically inherited diseases. Mutation in both HSPB1 and HSPB8 have been reported to be causative in the development of Charcot-Marie-Tooth disease and distal hereditary motor neuropathy (18;19). Mutated HSPB4 and HSPB5 can both cause cataract (20;21) while desmin-related myopathy is caused by specific mutation in HSPB5 (22). Another feature of certain HSPB proteins is their involvement in RNA splicing: e.g. HSPB1 upregulation was found to enhance the recovery of splicing after heat stress (23) which may relate to early findings of enhanced recovery of translation arrest after stress when HSPB1 was overexpressed (24). This has been associated with findings that, upon a combination of stress and (stress-induced) phosphorylation, HSPB1 associates with nuclear splicing speckles (or SC35 speckles) (25), which are nuclear domains involved in RNA splicing. Also HSPB5 has been reported to associate with these nuclear speckles in a phosphorylation dependent manner (26). Interestingly, this HSPB member was shown to recruit the F-box protein FBX4 to the speckles, suggesting a role for HSPB5 in facilitating ubiquitination of speckle components (26;27). We also found that heat denatured substrates associate with these nuclear speckles together with HSPB1 and indirect evidence indicated that this was not associated with refolding of denatured substrates (28). Rather, a hypothetical model was proposed that these nuclear splicing speckles would be used during stress for temporal storage of unfolded proteins to target them for degradation upon stress relief. This model was supported by findings that these speckles largely overlapped with sites for protein degradation (29). Whether cytoskeleton-related functions, splicing-related functions and chaperone activity are distinct features of the various HSPB proteins or whether they all are phenotypical manifesta- tions of a single conserved function remains unclear. Also, to what extend the various HSPB members differ in these functional aspects remains to be elucidated. Therefore, we initiated a systematic study on the HSPB1-HSPB10 members of the human HSPB family (HSPB11 was not investigated as it was discovered (30) after we had completed our studies). In this MS, we provide the first part of these data, starting with a general overview, mainly based on bioinformatics, on the tissue specific expression, intracellular localization patterns, interactions partners and co-regulated genes of the HSPB1-10 members. Using a human HSPB plasmid library, we initially focused on the sub-cellular localization patterns of the HSPB family. Interestingly, we show that one of the least investigated HSPB members, HSPB7, is a SC35-speckle associated protein under non-stressful conditions. Association of HSPB7 with nuclear speckles is dependent on the N-terminus of HSPB7. While HSPB1 can assist in the refolding of heat denatured

37 Chapter 3 proteins, HSPB7 is ineffective in doing so. Moreover, targeting HSPB1 to the nuclear speckles using the HSPB7 speckle localization signal abolishes its refolding enhancing capacity, strongly suggesting that association with speckles is not associated with chaperone-assisted refolding.

Experimental procedures

Reagents and antibodies Tetracycline, MOPS and cycloheximide were obtained from Sigma, beetle luciferin from Prome- ga. Antibodies against the V5 tag (Invitrogen) and SC35 (Abcam) were mouse monoclonal. The antibody against the MYC tag (Abcam) was rabbit-polyclonal.

Molecular techniques Standard recombinant DNA techniques were carried out essentially as described by Sambrook et al. (31) Oligonucleotide primers (Biolegio) and plasmids used in this study are listed in Table S1 and Table S2 respectively. Restriction enzymes were used according to the manufacturer’s instructions (Invitrogen, New England Biolabs). Preparative polymerase chain reactions (PCR) were carried out using Vent DNA polymerase (New England Biolabs) according to the manufacturer’s instructions. DNA sequencing reactions were carried out by ServiceXS. HSPB protein sequences were retrieved from the NCBI database and anaysed by ClustalX (32) and visualized using Treeview (33).

HSPB plasmid generation The human HSPB family members were isolated from both mixed cell line cDNA and commer- cially available cDNA clones (Open Biosystems) (Table S1). PCR products were cloned into the pCDNA5/FRT/TO-MV plasmid (Invitrogen) (Figure S1) and sequence verified. Modification of the pcDNA5/FRT/TO plasmid was performed by cloning of oligonucleotides. This vector is designed to allow, in combination with a tetracycline repressor, for tetracycline-regulated expression (Figure S1). The HSPB encoding cDNAs were subsequently sub-cloned in frame with either a HIS tag, V5 tag, EGFP, MYC or mRFP tag (for availability see Figure S1). As such, the resulting library is tailored to a large set of research topics. Deletion of the N-terminus op HSPB7 was performed using the Quickchange kit (Stratagene). Generation of chimeric constructs was performed as described before (34).

Growth conditions and DNA transfections Flp-In T-Rex HEK293 cells expressing the Tet repressor (Invitrogen) and HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco; Invitrogen) supplemented with 10% fetal calf serum (Griener Bio-one) at 37 °C under a humidified atmosphere containing 5% CO2. Mouse HL-1 atrial cardiomyocytes were maintained in Claycomb Medium (Sigma) supplemented with 100 μM norepinephrine (Sigma) dissolved in 0.3 mM l-ascorbic acid (Sigma), 4 mM l-glutamine (Gibco; Invitrogen) and 10% fetal calf serum (Griener Bio-one) at 37 °C under a humidified atmosphere containing 5% CO2. HEK293 cells and mouse HL-1 cells were transfected using Lipofectamine (Invitrogen) according to the manufacturer’s instructions using 1ug of plasmid DNA per 35-mm dish. HeLa cells were transfected using Effectene (Qiagen) according to the manufacturer’s instructions using 0,6 ug of plasmid DNA per 35-mm dish. Gene expression in Flp-In T-Rex HEK293 cells was induced with a final concentration of 1 µg/mL tetracycline.

Microscopy For microscopy, cells were plated 24 hours before transfection. For fixation, the coverslips were washed with cold PBS and fixed for 15 minutes with methanol (-20 0C). Cells were permea-

38 HSPB7 localizes to SC35 speckles bilized in blocking solution (100 mM glycine, 3% BSA, 0,1% triton) for 1 hour followed by 1 hour incubation with the primary antibody (V5 anti-mouse, 1:200, SC-35 anti-mouse, 1:10000, MYC anti-rabbit 1:200). After three washing steps, coverslips were incubated for 1h with CY3- conjugated anti-rabbit secondary antibody (Amersham Biosciences) at 1:200 dilution or with FITC-conjugated anti-mouse secondary antibody (Jackson) at 1:200 dilution. After three washing steps, the coverslips were mounted using Citifluor mounting medium (Citifluor Ltd). Images were obtained using an inverted confocal laser scanning microscope (TCS SP2, DM RXE, Lei- ca, Rijswijk, The Netherlands) with a 63×/1.32 NA oil objective. The mRFPruby fusion protein (35;36) was excited with a 543-nm helium–neon laser line and emission was recorded between 571 and 626 nm using a DD 488/543 dichroic filter.

Luciferase refolding assay Chaperone activity of HSPB members was assessed by using the luciferase refolding assay (37). Briefly, HEK293 cells were co-transfected with nuclear targeted luciferase (Nuc-superluc-EGFP) (38) together with HSPB encoding plasmids (1:9 ratio). Two hours after transfection, expression was induced by addition of tetracycline. After 24 hours, cells were resuspended and divided into 1 mL portions in tissue-grade 10 mL tubes. The following day, cells were given a heat-shock (30 minutes at 43 0C) in the presence of cycloheximide (20 mg/mL) and 4-morpholinepropanesulfonic acid (MOPS; 20mM, pH7.0) in order to inhibit protein synthesis and increase the buffer capacity of the medium respectively. After heat-treatment, cells were allowed to recover before luciferase activity was determined (3 hours at 37 0C). Luciferase activity measurements were performed using a Berthold Lumat 9510 luminometer (Berthold Technologies).

Results

Bioinformatical analysis Compared to other HSP families, small HSPs show a rather large sequence divergence, in particular in both the N- and C-terminal regions. A nearest neighbor analysis using the full sequences (Figure 1a) in order to link functional properties of the diverse member will thus likely be prone to wrong interpretations. And although the two functionally related and interactive members HSPB4 and HSPB5 (39) cluster together in such an analysis (figure 1a), HSPB1 does not, despite the fact that it shares many functional features with HSPB4 and HSPB5, amongst others of chaperone activity (15). Rather, in such an analysis HSPB1 clusters together with the non-functionally related HSPB8. Therefore, we repeated the nearest neighbor analysis using only the conserved crystallin domains (Figure 1b). Now, HSPB1 did group together with its functionally related family members HSPB4 and HSPB5. In addition, HSPB2 and HSPB3, which are reported to form functional hetero-oligomers during muscle differentiation (40), now also appear as a (separate) group within the tree, suggesting that this nearest neighbor analysis based on the crystalline domain only might be more informative for function. If correct, this nearest neighbor analysis would than suggest a possible functional overlap between HSPB8/HSPB10 and HSPB7/HSPB9 (Figure 1b). We have indeed observed a functional overlap between HSPB7 and HSPB9 in handling aggregated substrates (data submitted elsewhere). A functional overlap between HSPB8 and HSPB10 remains to be tested experimentally. Next, we extracted data from online databases to analyze both interaction data (41) and co- expression data (42) to provide information on interaction partners and co-regulated genes. HSPB members show an extensive set of interaction partners including other HSPB members (Figure 2a). When looking at gene-networks, HSPB members show co-regulated expression with numerous genes, again including other HSPB genes (Figure 2a). Clearly these data are yet far from complete as many HSB members have only been poorly studied. Moreover, part of

39 Chapter 3

A Complete protein sequence B Crystallin domain only HSPB4 HSPB6 9914 HSPB5 1977 HSPB8 9618 4896 HSPB6 HSPB10 3162 5134 HSPB2 HSPB7 4348 3775 HSPB1 HSPB9 4227 6155 HSPB8 HSPB1 HSPB7 4959 HSPB4 7488 HSPB3 6743 HSPB5 HSPB10 HSPB2 HSPB9 5363 HSPB3 0.1 0.1 Figure 1. Nearest neighbor analysis of the human HSPB family. Nearest neighbour analysis of the complete protein sequence of the HSPB members (A) or the crystallin domains only (B). Protein alignments were performed in ClustalX using the Neighbor-joining algorithm in combination with Blosum scoring. Boot- strap values are indicated on the branchpoints. (Bootstrap trails: 10000)

Figure 2. Physical interactions A Protein interactions (Biogrid) Co-regulated genes (COXPRESdb) and co-regulated gene expres- HSPB1 HSPB5, HSPB8, HSPB4, CRYBB2, BAG3, TEAD3, SERPINH1 CRYGC, RIF1, TP53, AKT1, sion of the HSPB family mem- MAPKAPK2, CYCS, MED31, ILK, bers with other proteins/genes. MAPKAPK5, DAXX, MAGED1, TGFB1I1, MGC15730, PP (A) Database searches were HSPB2 DMPK, HSPB2, HSPB8, HSPB3, HSPB5, HSPB6, FXYD1 performed to collect information HSPB5, UBL5 HSPB3 ANP32A, MED31, DKFZP564B147, CSRP3, CKM, unc-45 homolog B on HSPB interactions with other DKFZP564O0523, OLFML3, MED8, proteins and co-regulated genes. MRPL38, NELF, RAMP3, BBOX1, PRPF4B, RIF1, SETDB1, ZZEF1, (B) An overview on both physi- UNC119, HSPB2 cal interactions (dotted line) and HSPB4 HSPB1, LALBA, CRYZ, MIP, ALB, SLC22A8, SLC5A2, FAM151A HSPB4, HSPB5 co-regulated genes (straight line) HSPB5 HSPB1, PSMA3, CS, HSPB2, HSPB5, HSPB2, HSPB8, MRAS, ASPA, DIXDC1, within the HSPB family. BMPR2 , HSPB4 S100A1, SASH1, CX3CL1, PPP2R3A, PRIMA1

HSPB6 HSPB2, HSPB7, SOD3, ITGA7, MYOC, PRKG1

HSPB7 HSPB8 HSPB6, CASQ2, ACTC1 HSPB8 HSPB1, HSPB8, HSPB7, HSPB2, HSPB5, PPP1R3C, PLA2G4C the data only come from cell BAG3 free studies and yet have to HSPB9 CNTFR, CLDND2, C1orf65, TEX13A, HSFX1, RP6-166C19.11 be verified in living cells. But, HSPB10 SPAG4, SPAG5, TRIP6, ODF1 FSCN3, ADAD1, SMCP, UGT1A6, HDGFL1, PLSCR2, TSSK4, SYNGR4, HRASLS5, as can be derived from Figure SLC22A16, LRIT1, HSF2BP 2b, there is a likely complex B and extensive interactive net- Interactions work within the HSPB family HSPB8 suggesting that they cooper- Co-regulated genes ate in several processes. HSPB4 In addition to differences in HSPB5 functionality, individual HSPB members show large differ- HSPB1 ences in tissue-specific ex- HSPB2 HSPB3 pression (Figure 3). HSPB4 HSPB7 for instance, is found at high concentrations in the eye lens, HSPB6 while its expression is almost absent in other tissues (43). Compiling data from the UniGene EST ProfileViewer (NCBI) (Figure 3a) for total messenger levels reveals that in the human body as a whole, the most abundant

40 HSPB7 localizes to SC35 speckles

HSPB member is HSPB1, followed by HSPB5 and interestingly also HSPB7 (Figure 3a top row). Although reported to be expressed almost exclusively in heart tissue (44), messenger RNA for HSPB7 can also be found in adipose tissue, connective tissue and the parathyroid gland. HSPB9, that clusters together with HSPB7 in the nearest neighbor tree (Figure 1b) has been referred to as a testis specific HSPB member (45). But, HSPB9 expression is also found in lung and pancreas tissue. When comparing HSPB messenger composition for a selection of organs, HSPB1 and HSPB5 represent the most abundant HSPB messengers in brain, muscle and eye tissue (Figure 3b). Looking at heart tissue, HSPB7 is indeed transcribed at very high levels in comparison to the other HSPB members (Figure 3b). HSPB messenger composition in the testis shows that HSPB7 and HSPB1 are the most abundant in this organ and not HSPB9. The most restricted members in view of tissue specific expression are HSPB4 (eye), HSPB9 (lung, pancreas, testis) and HSPB10 (testis) (Figure 3a).

Subcellular localization The small HSPs are known to be widely distributed throughout eukaryotic cellular compartments. Herein, plants take a special place, with the most widely known organelle specific distribution pattern (46). In Drosophila at least one small HSP, the mitochondrial HSP22, shows compartment specific localization besides the cytosol and nucleus (47). As a complete picture of HSPB protein localization is currently unavailable, we combined web-based prediction programs and subcellular localization databases together with confocal microscopy imaging of the V5-tagged HSPB library to generate a complete overview. In silico prediction of subcellular localization was performed using PSORT, Ptarget and Multiloc (48-50) (Figure 4a). There was a surprisingly large disagreement between the predictions made by these different programs especially with regards to a tentative mitochondrial localization signal. Using MITOPRED (51), a web server dedicated specifically for the prediction of mitochondrial localized proteins, HSPB5 and HSPB7 were predicted to be mitochondrial with high confidence, while the other members were not predicted to localize to mitochondria. For HSPB5, a mitochondrial localization was indeed recently experimentally verified (52). However, nor the LOCATE database (53) nor the Hu- man Protein Reference (HPR) database (54) indicated the mitochondrial localization of HSPB5 and large differences in the predicted localization and experimentally verified localizations were seen, especially when using the Ptarget server. To investigate and directly compare the localization of HSPB proteins experimentally, we expressed the V5-tagged HSPB members in human HEK-293 cells (Figure 4b). Endogenous HSPB1, the most extensively investigated HSPB member is known to be mostly cytosolic under non-stress conditions and only enters the nucleus upon stress-induced phosphorylation and changes in oligomeric size after which it co-localizes with SC35 positive nuclear speckles (17;25). Using MCF-7 cells that express high levels of HSPB1, endogenous HSPB1 indeed showed mainly cytoplasmatic staining in unstressed cells and a heat-induced localization to nuclear speckles (Figure 4c). Staining of the ectopically expressed V5-HSPB1 in HEK293 cells, showed a similar pattern of expression during control and heat-shocked conditions (Figure 4d), showing that the V5-tag did not interfere with its physiological localization patterns. Also several experimental data are available for HSPB4 and HSPB5. Like HSPB1, HSPB4 was reported to be localized in the cytoplasm (55;56). A similar pattern is seen for the ectopically expressed V5-tagged HSPB4 (Figure 4b). HSPB5 has been reported to be a cytosolic and nuclear protein where it can specifically localize to SC-35 positive nuclear speckles (57;58) in a phosphorylation dependent manner (26;59). Again, the transiently expressed V5-tagged HSPB5 confirmed the cytosolic staining representative for non-phosphorylated HSPB5. Also a faint signal of nuclear foci could be detected, confirming that like untagged HSPB5, also V5-HSPB5 in part associates with nuclear domains. The findings on these three well-studied and closely related HSPB mem-

41 Chapter 3

Figure 3. HSPB mRNA A levels throughout the HSPB1 HSPB2 HSPB3 HSPB4 HSPB5 HSPB6 HSPB7 HSPB8 HSPB9 HSPB10 human body. (A) The top Relative total body expression row indicates the amount

adipose tissue of total mRNA present in adrenal gland the human body for each ascites HSPB member, relative bladder blood to HSPB1. The vertical bone view shows the distribu- bone marrow brain tion in mRNA expression cervix for each HSPB member connective tissue ear in individual organs rela- embryonic tissue tive to the organ with the esophagus eye highest HSPB mRNA heart abundace. (B) Compari- intestine son of HSP1, HSPB5, kidney larynx HSPB7 and HSPB9 liver mRNA expression in ei- lung lymph ther brain, heart, mus- lymph node cle, eye or testis. Values mammary gland mouth for gene expresssion muscle (transcript per million) nerve were retrieved from Uni- ovary pancreas Gene (NCBI) and calcu- parathyroid lated relative to the most pharynx pituitary gland abundant HSPB mem- placenta ber. These values were prostate salivary gland transformed into gray- skin scale values (0%=white, spleen stomach 100%=black) using Ado- testis be Photoshop. thymus thyroid tonsil trachea umbilical cord uterus vascular

B brain heart muscle eye testis HSPB1 HSPB5 HSPB7 HSPB9 bers (Figure 1b) show that their cellular compartmentalization is highly comparable and also show that V5-tagging does not interfere with their intracellular localization. Also, V5-tagging of HSPB1, HSPB4 or HSPB5 did not affect their function as chaperone in cells to support refolding of heat denatured luciferase (Figure 8a). Apparently for these members tagging does not affect biological activity.The V5-tagged HSPB2 was found to be present in the nucleus and cytoplasm, consistent with the data from LOCATE. In addition, HSPB2 was present in foci throughout both cellular compartments. HSPB2 has been shown previously to localize to mitochondria (60), we were however unable to detect overlap of the cytosolic HSPB2 foci and the mitochondrial marker Cox8-dsRED (data not shown) suggesting either errors in the predictive programs or a inhibitory effect of the V5 tag on mitochondrial import. The nature of the nuclear foci was not elucidated. To date, no experimental data are available on the localization of HSPB3, the protein that is most closely related to HSPB2 (Figure 1b). Unlike HSPB2, however, but consistent with the database

42 HSPB7 localizes to SC35 speckles

A Predicted localizations Experimentally verified PSORT Ptarget Multiloc Mitopred LOCATE & HPR This study HSPB1 cytoplasmic, mitochondrial cytoplasmic non-mitochondrial cytoplasmic; cytoplasmic nuclear plasma membrane HSPB2 cytoplasmic, cytoplasmic cytoplasmic non-mitochondrial nuclear; cytoplasmic cytoplasmic, nuclear nuclear HSPB3 cytoplasmic cytoplasmic cytoplasmic non-mitochondrial _ cytoplasmic, nuclear HSPB4 cytoplasmic, mitochondrial nuclear non-mitochondrial cytoplasmic cytoplasmic nuclear HSPB5 mitichondrial mitochondrial cytoplasmic Mitochondrial (92,3%) nuclear, cytoplasmic, cytoplasmic plasma membrane HSPB6 cytoplasmic, mitochondrial cytoplasmic non-mitochondrial _ cytoplasmic, nuclear nuclear HSPB7 mitochondrial mitochondrial extracellular Mitochondrial (92,3%) cytoskeleton cytoplasmic, nuclear HSPB8 nuclear cytoplasmic nuclear non-mitochondrial cytoplasmic. nuclear cytoplasmic, nuclear HSPB9 mitochondrial cytoplasmic mitochondrial non-mitochondrial cytoplasmic. nuclear cytoplasmic, nuclear HSPB10 extracellular extracellular cytoplasmic non-mitochondrial cytoplasmic cytoplasmic

B HSPB1 HSPB2 HSPB3 HSPB4 HSPB5

HSPB6 HSPB7 HSPB8 HSPB9 HSPB10

C Endogenous HSPB1 D V5-HSPB1 overexpression

HS HS Figure 4. Intracellular distribution of the human HSPB. (A) Several databases were used to extract the predicted intracellular distribution and published intracellular distribution. (B) Intracellular distribution of V5-tagged human HSPB proteins in HEK293 cells. Arrows indicate bright foci in both the nucleus and the cytoplasm (V5-HSPB2), nuclear foci (V5-HSPB7) and aggresome formation (V5-HSPB10). The nucleus was visualized by expression of the human Histone-2B fused to the monomeric red fluorescent protein ruby (36;83). This signal was traced in Adobe photoshop and depicted as a dashed line. (C) Redistribution of endogenous HSPB1 to nuclear speckles after a heat shock in MCF-7 cells. (D) Redistribution of exogenous V5-HSPB1 to nuclear speckles after a heat shock in HEK293 cells. predictions, V5-HSPB3 showed mainly a cytosolic localization with some minor nuclear staining. No clear localization to specific other organelles or sub-nuclear domains was seen. Also for the intracellular localization of HSPB6 experimental data are lacking. The V5-HSPB6 confirmed the predicted cytosolic distribution but could not confirm the mitochondrial localization as predicted by Ptarget.

43 Chapter 3

HSPB8 has been reported to be an exclusively cytosolic protein (61) although others also have suggested that it is localized to the nucleus in some cells (18). The ectopically expressed V5- HSPB8 shows a predominant cytosolic staining, with some intense staining in cytosolic foci. These foci could potentially represent autophagosomes, consistent with the suggested function of HSPB8 in autophagy (62). Its closest homologue HSPB10, known to be a structural protein in sperm tails (63), also mainly localized to the cytosol with minor nuclear staining. The expression was however accompanied by the appearance of many cytoplasmic foci and the formation of an aggresome, suggesting misfolding and/or failed clearance of the ectopically expressed V5-tagged protein (64). The V5-tagged HSPB9 showed a diffuse cytosolic and nuclear distribution (Figure 3b), whereas in vivo HSPB9 shows mainly a nuclear localization in testis and a cytosolic/nuclear distribution in tumor cells (65). Again this localization was not predicted correctly by the in silico prediction programs (Figure 3a). Staining of HSPB7 showed a rather striking distribution. HSPB7 was reported to interact with a-Filamin, an actin-binding protein (44), suggesting a cytoskeletal localization. The V5-HSPB7 showed cytosolic staining (Figure 3b). In addition, however, bright nuclear foci could be detected, that suggested association with nuclear sub-domains in the absence of stress. During stress, several heat shock proteins are known to be re-allocated within the cells and especially move to the nucleus, including to nuclear speckles (25;66-68). Therefore, we analysed heat-induced alterations in intracellular localization for all the individual HSPB members. Interestingly, most members appeared after heat shock in nuclear foci or showed an enhanced nuclear foci signal in comparison to control conditions (Figure S2a). The clearest change in subcellular distribution was found for HSPB1 and HSPB4, which both entered the nucleus after heat shock treatment and appeared at nuclear foci (Figure S2a, b).

HSPB7 specifically associates with Cajal bodies and SC35 speckles Because of the association of HSPB1 with splicing and the stress-induced reallocations of HSPB1 and HSPB5 to SC35 splicing speckles, we decided to identify to which specific nuclear structure(s) HSPB7 localizes. Hereto, we co-expressed different nuclear sub-domain markers tagged with EGFP or EYFP with MYC tagged HSPB7. EGFP tagged SP100 was used to visualize promyelocytic leukaemia (PML) nuclear bodies which play a role in transcriptional regulation, programmed cell death, tumor suppression, and antiviral defense (69;70). These structures were in close proximity to MYC-HSPB7-containing bodies, but did not overlap (Figure 5a). EGFP-tagged Coilin, marking the Cajal bodies, implied in spliceosomal snRNP assembly and various other functions (71) showed partial overlap with the MYC-HSPB7 signal (Figure 5b), indicating that the HSPB7 can localize to Cajal bodies. Splicing speckles were visualized using SC35 and ASF/SF2; two members of the serine/arginine (SR) rich protein family. These proteins form a group of structurally and functionally related splicing factors which are essential for constitutive and alternative splicing regulation in higher eukaryotes (72;73). Expression of EGFP tagged alternative splicing factor/splicing factor 2 (ASF/SF2) showed complete co-localization with MYC-HSPB7 (Figure 5c). Similarly, EGFP tagged SC35 also completely co-localized to MYC-HSPB7 (Figure 5c). Finally, staining of endogenous SC35 confirmed co-localization with MYC-HSPB7 (Figure 5c). So, in contrast to HSPB1 and HSPB5 that enter the nucleus and associate with speckles only after stress, HSPB7 localizes to these splicing factories also in non-stressed human HEK293 cells (Figure 4b). To test whether this localization is not restricted to HEK293 cells only, we also expressed V5-HSPB7 in mouse HL-1 cardiac myocytes and human HeLa cells. Also in these

44 HSPB7 localizes to SC35 speckles

A EYFP-SP100 MYC-HSPB7 Merge C EYFP-ASF/SF2 MYC-HSPB7 Merge PML bodies PML

B EGFP-Coilin MYC-HSPB7 Merge SC-35 MYC-HSPB7 Merge SR proteins Cajal bodies EGFP-tagged Endogenous

Figure 5. HSPB7 associates with SR protein-containing nuclear domains. HSPB7 was analyzed for colocalization with either (A) SP100; a marker for PML bodies, (B) Coilin; a marker for Cajal bodies or (C) SR proteins ASF/SF2 and SC35; two markers of splicing speckles. Co-localization with SC35-EGFP was confirmed by staining for endogenous SC35. See Appendix 4 for colour print.

V5-HSPB7 in HL-1 V5-HSPB7 in HELA Figure 6. HSPB7 association with nuclear domains A B is cell-type independent. V5-HSPB7 was transfected and visualized in either (A) mouse HL-1 or (B) human HeLa cells. (C, D) In both cell lines V5-HSPB1 was predominantly cytoplasmic. The nucleus was visualized by expression of the human Histone-2B fused to the monomeric red fluorescent protein ruby (H2B-mR- FPruby). This signal was traced in Adobe Photoshop and depicted as a dashed line.

two cell types HSPB7 localized to nuclear V5-HSPB1 in HL-1 V5-HSPB1 in HELA speckles (Figure 6a,b), whereas V5-HSPB1 D C was largely cytosolic in both cell types under non-stressed conditions (Figure 6c,d).

The N-terminus of HSPB7 is required and sufficient for localization to SC35 splicing speckles As the conserved crystallin domain is the conserved motif for the HSPB members, it is likely that sub-cellular localization is regulated by targeting sequences present in the N-terminal or C-terminal domain. Indeed, for Drosophila HSP27 and HSP22 a nuclear localization signal and a mitochondrial targeting sequence respectively have been identified in the N-terminal domains (47;74). To analyze if the N-terminus of HSPB7 is required for speckle association, we deleted the first 71 amino acids of HSPB7 (the DN-HSPB7 mutant, Figure 7a). Compared to the full length HSPB7 (Figure 7b1, b1’), the DN-HSPB7 mutant showed less nuclear staining with a complete loss of speckle-association (Figure 7b2, b2’). To test whether the N-terminus of HSPB7 indeed drives speckle localization, we generated HSPB chimeras by switching the N-

45 Chapter 3

74 170 A HSPB7 topology N-terminus crystallin domain

NB7-NB1 N-terminus HSPB7 (1-71) HSPB1 (85-205)

NHSPB7

NB1-NB7 N-terminus HSPB1 (1-84) HSPB7 (72-170)

B HSPB7 NHSPB7 NB1-NB7 NB7-NB1 1 2 3 4

1’ 2’ 3’ 4’

C NB7-NB1 SC-35 / H2B-mRFPruby Merge

D NB7-mRFPruby SC-35-EGFP Merge

Figure 7. The HSPB7 N-terminus is required and sufficient for targeting to SC35 speckles. (A) Sche- matic representation of the topology of HSPB7 and the constructed mutant and chimeric proteins. Localiza- tion pattern and close-up of HSPB7 (B 1/1’), DNHSPB7 (B 2/2’), NB1-DNHSPB7 (B 3/3’) and NB7-DNHSPB1 (B4/4’). (C) Staining of NB7-DNHSPB1 together with endogenous SC35 shows co-localization of both proteins. The nucleus was visualized using H2B-mRFPruby. (D) Expression of a chimeric protein containing the N-terminus of HSPB7 fused to mRFPruby localized to SC35-EGFP stained speckles. See Appendix 4 for colour print. terminal domains between HSPB1 and HSPB7 (Figure 7a). When expressed in cells, HSPB7 containing the N-terminus of HSPB1 (NB1-DNB7) was present in both the cytoplasm and nucleus but completely lacked any detectable association with sub-nuclear domains (Figure 7b3, b3’). HSPB1 containing the N-terminus of HSPB7 (NB7-DNB1) on the other hand showed staining of sub-nuclear structures resembling wildtype HSPB7 (Figure 7b4, b4’). Staining of endogenous SC-35 indeed confirmed association of NB7-DNB1 with speckles (Figure 7c). Thus it seems that speckle association of HSPB7 is indeed dependent on the N-terminal domain. To test whether the N-terminus of HSPB7 can also target proteins without a crystallin domain to SC35 speckles,

46 HSPB7 localizes to SC35 speckles

Nuclear luciferase refolding Speckle association & refolding I A 40 B 30 3h recovery 3h recovery not speckle associated g g 25

i n i n 3h recovery 30 l d l d speckle associated o o f f 20 e e r r

s e 20 s e a a 15 r r e e c i f c i f

u u 10 L L 10 % % 5

0 0 y y A 1 4 5 7 7 1 7 1 b b 1 B B B B u B B B B u r - - A r P P P P P P 1 7 P P P H S H S H S H S H S H S N B N B R F R F H S m m Speckle association & refolding II C 40 Figure 8. HSPB7 and chaperone activity. HSPB 3h recovery proteins were co-expressed with nuclear firefly luci- g

i n ferase in HEK293 cells. 48 hours after transfection, l d o

f 30 cells were heated for 30 minutes at 43 °C reducing e r

luciferase activity to less than 5%. Next, cells were

s e re-incubated for 3 hours at 37 °C to allow for (chap- a r

e 20 erone-assisted) luciferase refolding. Luciferase ac- c i f

u tivities are plotted relative to the activity in unheated L cells (=100%). (A) Refolding of heat-denatured luci- % 10 ferase in cells overexpressing V5-HSPA1A (positive control), V5-HSPB1, V5-HSPB4, V5-HSPB5 and 0 V5-HSPB7. (B) Comparison of chaperone activity in

y 1 1 1 1 refolding heat-denatured luciferase of either SC35 b B B B B u - r 7 P P P speckle localized or non-SC35 speckle localized P N B HSPB7 H S H S H S HSPB proteins. (C) Chaperone activity of HSPB1 R F

m without or with co-expression of resident speckle associated HSPB proteins. Data are the mean ± SD. we fused the N-terminus of HSPB7 to the red fluorescent protein mRFPruby (NB7-mRFPruby). The NB7-mRFPruby chimeric protein indeed co-localized with EGP tagged SC35 (Figure 7d), proving that the N-terminus of HSPB7 contains a novel nuclear speckle localization signal which functions independently of HSPB domain architecture.

Speckle association of HSPB1 and HSPB7 in not linked to chaperone activity In previous studies (17;28) it was found that after heat shock, not only chaperones like HSPB1 are reallocated to SC35 speckles, but also heat-denatured proteins. Although HSPB1 overexpression could enhance refolding of heat denatured nuclear model substrates like firefly luciferase, indirect evidence, including the absence of HSP70 family members in these foci (75), suggested that the association of HSPB1 and denatured luciferase in SC35 speckles was not associated with refolding activity. To further look into the relevance of the presence of both HSPB7, HSPB1 and unfolded luciferase in SC35 speckles for the storage and subsequent processing of unfolded nuclear proteins, we first tested whether the constitutively SC35-associated HSPB7 also could assist in the refolding of heat-denatured nuclear localized luciferase. Overexpression of either V5-tagged HSPB1, HSPB4 or HSPB5 indeed resulted in enhanced refolding (Figure 8a) as described before for the untagged versions in vitro (15;76). Refolding activity of HSPB4 towards nuclear luciferase was due to a translocation of HSPB4 towards the nucleus upon heat shock (Figure S2a, S3). Under these conditions, NLS tagged luciferase was

47 Chapter 3 found exclusively in the nucleus (Figure S3), suggesting that the cytosolic fraction of HSPB4 does not contribute to refolding of heat-denatured nuclear luciferase. Overexpression of the SC35-associated V5-HSPB7 did however not show any refolding promoting activity (Figure 8a). The NB1-DNB7 mutant, harboring the crystallin domain of HSPB7 and devoid of nuclear speckle association also did not show refolding promoting activity (Figure 8b), in line with the complete lack of chaperone activity of the full length HSPB7. The NB7-DNB1 mutant, harboring the crystallin domain of HSPB1, but now associated with SC35 speckles via the N-terminus of HSPB7, lost refolding promoting activity (Figure 8b); suggesting that localization in part results in loss of refolding capacity. Finally, we asked whether speckle-associated HSPB proteins with no folding activity could affect the wildtype HSPB1 folding capacity via e.g. co-oligomerization or induced association with splicing speckles or by competitive substrate binding. Hereto, we co-expressed HSPB7 with HSPB1 and nuclear luciferase. This however did not affect the capacity of HSPB1 to assist in refolding heat denatured luciferase (Figure 8c). Next, we co-expressed the NB7-HSPB1 (NB7- DNB1) mutant, which is targeted to nuclear speckles, together with wildtype HSPB1. Again, HSPB1 refolding capacity was not negatively influenced by co-expression of a speckle-associated (mutant) HSPB protein (Figure 8c). This highly suggests that NB7-DNB1, lacking folding activity, does not interfere with the folding-active wildtype HSPB1.

Discussion

We have presented an overview on several aspects and properties of the human HSPB family using a variety of specialized databases and focused subsequently on sub-cellular localization patterns. Apart from their cytoskeletal localization, some HSP members have been reported to reside in the nucleus during non-stressed states or after a heat-shock. Inside the nuclear compartment, HSPB1 and HSPB5 can furthermore associate with sub-nuclear structures called SC35- or splicing speckles. Using the V5-tagged HSPB plasmid library we confirmed these data for HSPB1 and for the first time show that HSPB7 contains a speckle targeting sequences in its N-terminus that is capable of driving it and other proteins to splicing speckles under non- stressful conditions.

HSPB7 is a resident component of SC35-speckles containing a speckle targeting sequence in its N-terminal domain SC35 speckles (also known as nuclear splicing speckles) are sub-nuclear structures involved in the storage and assembly of the splicing machinery and contain pre-mRNA splicing factors, transcription factors, snRNAs and ribosomal proteins (77). Furthermore, modulation in splicing factor concentrations appears to link SC35 speckles to alternative splicing (78). The SC35 speckle resident splicing factors primarily belong to the serine/arginine-rich (SR) protein family (79;80). In general, little is known on how proteins are able to associate with SC35 speckles. One known SC35 speckle targeting signal consists of the SR motif itself, as was shown for the SC35 protein (73), although the presence of an SR motif not always drives proteins to SC35 speckles, meaning that other elements or the context around SR motifs may also play a role (81). Until now, only one non-SR motif has been reported to function as a targeting signal for SC35 speckles. Cdk9, an elongation factor for RNA polymerase II-directed transcription and DYRK1A, a protein kinase (78;82), both contain a histidine-rich region which was shown to be sufficient to target EGFP to SC35 speckles. HSPB1, HSPB5 and HSPB7 however lack such a SR motif and histidine- rich region, suggesting that other elements in the HSPB sequences are required for speckle association. The dependency on a specific sequence needed for SC35 speckle association

48 HSPB7 localizes to SC35 speckles was reported for the Drosophila HSP27 (74). The authors showed that speckles association in HeLa cells was dependent on two leucine residues (L50, L52) located in the N-terminus. Although leucines are present in the N-termini of HSPB1, HSPB5 and HSPB7, their contribution to speckle association have yet to be established (Figure S4). In this study we showed that the N-terminus of HSPB7 functions as a unique and novel SC35 targeting sequence. Deletion of the N-terminus of HSPB7 or replacement by the N-terminus of HSPB1, abrogated HSPB7 speckle association. We furthermore showed that the N-terminus alone is sufficient to target either a crystallin domain or mRFPruby to SC35 speckles. HSPB7 contains a serine-rich region, which shows homology to the SR motif. Deletion of the serine stretch did however not result in dissociation of HSPB7 from SC35 speckles (data not shown). This suggests that the N-terminus of HSPB7 contains additional elements allowing for speckle association, which have yet to be determined.

Functional implications of HSPB speckle associations HSPB1 is known to associate with SC35 speckles during heat stress. When luciferase is overexpressed under these conditions, it also associates with SC35 speckles (28). In view of this, SC35 speckles might function as storage sites for HSPB-luciferase complexes, ready to be refolded upon stress relief. Could SC35 speckles indeed act as a depot for unfolded proteins, and does the association of HSPB proteins with SC35 speckles reflect a role in chaperoning of protein refolding? We showed that the SC35 speckle resident HSPB7 is not active in refolding of heat-denatured luciferase in contrast to HSPB1 and HSPB5. This suggests that at least for HSPB7, association to SC35 speckles is unrelated to protein refolding. By exchanging the N- terminus of HSPB1 for the N-terminus of HSPB7, we were able to target the HSPB1 crystallin domain to SC35 speckles which resulted in loss of refolding activity. This suggest either that speckles association of HSPB1 is not related to refolding activity or that the HSPB1 N-terminus in required together with the HSPB1 crystallin domain to form a functional unit in assisting protein refolding activity. While co-expression of HSPB5 and luciferase clearly showed that this HSPB member is a chaperone capable of refolding heat-denatured luciferase, the direct association of HSPB5 with SC35 speckles has been suggested to relate to protein ubiquitination and degradation rather than protein refolding (26;27). In conclusion, data on HSPB association with SC35 speckles hint to other functions than specific refolding of denatured proteins in situ. As HSPB1 has been shown to enhance recovery of splicing after a heat shock, a role in splicing of HSPB7 at SC35 speckles should be addressed. A possible function of the SC35 speckle- resident HSPB7 herein could be to fulfill a function in splicing under normal conditions while accumulation of HSPB1 at SC35 speckles during stress indicates a requirement to protect the splicing machinery and accelerate splicing recovery. This idea is under current investigation.

Acknowledgements

This work was supported by Innovatiegerichte Onderzoeksprogramma Genomics. Grant IGE03018.

49 Chapter 3

References

1. Hightower, L. E., Hageman, J., Vos, M. J., Kubota, H. and others (2008) Guidelines for the nomenclature of the human heat shock proteins, Cell Stress. Chaperones. 2. Vos, M. J., Hageman, J., Carra, S., and Kampinga, H. H. (2008) Structural and functional diversities between members of the human HSPB, HSPH, HSPA, and DNAJ chaperone families, Biochemistry 47, 7001-7011. 3. Haslbeck, M., Franzmann, T., Weinfurtner, D., and Buchner, J. (2005) Some like it hot: the structure and function of small heat-shock proteins, Nat. Struct. Mol. Biol. 12, 842-846. 4. Schneider, G. B., Hamano, H., and Cooper, L. F. (1998) In vivo evaluation of hsp27 as an inhibitor of actin polymeri- zation: hsp27 limits actin stress fiber and focal adhesion formation after heat shock, J. Cell Physiol 177, 575-584. 5. Verschuure, P., Croes, Y., van den IJssel, P. R., Quinlan, R. A. and others (2002) Translocation of small heat shock proteins to the actin cytoskeleton upon proteasomal inhibition, J. Mol. Cell Cardiol. 34, 117-128. 6. Mounier, N. and Arrigo, A. P. (2002) Actin cytoskeleton and small heat shock proteins: how do they interact?, Cell Stress. Chaperones. 7, 167-176. 7. Lutsch, G., Vetter, R., Offhauss, U., Wieske, M. and others (1997) Abundance and location of the small heat shock proteins HSP25 and alphaB-crystallin in rat and human heart, Circulation 96, 3466-3476. 8. Djabali, K., Piron, G., de Nechaud, B., and Portier, M. M. (1999) alphaB-crystallin interacts with cytoplasmic intermediate filament bundles during mitosis, Exp. Cell Res. 253, 649-662. 9. Landry, J. and Huot, J. (1995) Modulation of actin dynamics during stress and physiological stimulation by a signaling pathway involving p38 MAP kinase and heat-shock protein 27, Biochem. Cell Biol. 73, 703-707. 10. Piotrowicz, R. S., Hickey, E., and Levin, E. G. (1998) Heat shock protein 27 kDa expression and phosphorylation regulates endothelial cell migration, FASEB J. 12, 1481-1490. 11. Fan, G. C., Ren, X., Qian, J., Yuan, Q. and others (2005) Novel cardioprotective role of a small heat-shock protein, Hsp20, against ischemia/reperfusion injury, Circulation 111, 1792-1799. 12. Yoshida, K., Aki, T., Harada, K., Shama, K. M. and others (1999) Translocation of HSP27 and MKBP in ischemic heart, Cell Struct. Funct. 24, 181-185. 13. Brundel, B. J., Shiroshita-Takeshita, A., Qi, X., Yeh, Y. H. and others (2006) Induction of heat shock response pro- tects the heart against atrial fibrillation, Circ. Res. 99, 1394-1402. 14. Basha, E., Friedrich, K. L., and Vierling, E. (2006) The N-terminal arm of small heat shock proteins is important for both chaperone activity and substrate specificity, J. Biol. Chem. 281, 39943-39952. 15. Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993) Small heat shock proteins are molecular chaperones, J. Biol. Chem. 268, 1517-1520. 16. Carra, S., Seguin, S. J., and Landry, J. (2008) HspB8 and Bag3: a new chaperone complex targeting misfolded proteins to macroautophagy, Autophagy. 4, 237-239. 17. Bryantsev, A. L., Kurchashova, S. Y., Golyshev, S. A., Polyakov, V. Y. and others (2007) Regulation of stress-induced intracellular sorting and chaperone function of Hsp27 (HspB1) in mammalian cells, Biochem. J. 407, 407-417. 18. Irobi, J., Van Impe, K., Seeman, P., Jordanova, A. and others (2004) Hot-spot residue in small heat-shock protein 22 causes distal motor neuropathy, Nat. Genet. 36, 597-601. 19. Evgrafov, O. V., Mersiyanova, I., Irobi, J., Van Den Bosch, L. and others (2004) Mutant small heat-shock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy, Nat. Genet. 36, 602-606. 20. Cobb, B. A. and Petrash, J. M. (2000) Structural and functional changes in the alpha A-crystallin R116C mutant in hereditary cataracts, Biochemistry 39, 15791-15798. 21. Li, H., Li, C., Lu, Q., Su, T. and others (2008) Cataract mutation P20S of alphaB-crystallin impairs chaperone activity of alphaA-crystallin and induces apoptosis of human lens epithelial cells, Biochim. Biophys. Acta 1782, 303-309. 22. Wang, X., Osinska, H., Klevitsky, R., Gerdes, A. M. and others (2001) Expression of R120G-alphaB-crystallin causes aberrant desmin and alphaB-crystallin aggregation and cardiomyopathy in mice, Circ. Res. 89, 84-91. 23. Marin-Vinader, L., Shin, C., Onnekink, C., Manley, J. L. and others (2006) Hsp27 enhances recovery of splicing as well as rephosphorylation of SRp38 after heat shock, Mol. Biol. Cell 17, 886-894. 24. Laszlo, A., Davidson, T., Hu, A., Landry, J. and others (1993) Putative determinants of the cellular response to hyperthermia, Int. J. Radiat. Biol. 63, 569-581. 25. Bryantsev, A. L., Chechenova, M. B., and Shelden, E. A. (2007) Recruitment of phosphorylated small heat shock protein Hsp27 to nuclear speckles without stress, Exp. Cell Res. 313, 195-209. 26. den Engelsman, J., Bennink, E. J., Doerwald, L., Onnekink, C. and others (2004) Mimicking phosphorylation of the small heat-shock protein alphaB-crystallin recruits the F-box protein FBX4 to nuclear SC35 speckles, Eur. J. Biochem. 271, 4195-4203. 27. den Engelsman, J., Keijsers, V., de Jong, W. W., and Boelens, W. C. (2003) The small heat-shock protein alpha B-crystallin promotes FBX4-dependent ubiquitination, J. Biol. Chem. 278, 4699-4704. 28. Bryantsev, A. L., Loktionova, S. A., Ilyinskaya, O. P., Tararak, E. M. and others (2002) Distribution, phosphorylation, and activities of Hsp25 in heat-stressed H9c2 myoblasts: a functional link to cytoprotection, Cell Stress. Chaper- ones. 7, 146-155. 29. Rockel, T. D., Stuhlmann, D., and von, M. A. (2005) Proteasomes degrade proteins in focal subdomains of the human cell nucleus, J. Cell Sci. 118, 5231-5242. 30. Bellyei, S., Szigeti, A., Pozsgai, E., Boronkai, A. and others (2007) Preventing apoptotic cell death by a novel small

50 HSPB7 localizes to SC35 speckles

heat shock protein, Eur. J. Cell Biol. 86, 161-171. 31. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular cloning a laboratory manual Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 32. Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R. and others (2007) Clustal W and Clustal X version 2.0, Bioinformatics. 23, 2947-2948. 33. Page, R. D. (1996) TreeView: an application to display phylogenetic trees on personal computers, Comput. Appl. Biosci. 12, 357-358. 34. Vos, M. J. and Kampinga, H. H. (2008) A PCR amplification strategy for unrestricted generation of chimeric genes, Anal. Biochem. 380, 338-340. 35. Fischer, M., Haase, I., Wiesner, S., and Muller-Taubenberger, A. (2006) Visualizing cytoskeleton dynamics in mammalian cells using a humanized variant of monomeric red fluorescent protein, FEBS Lett. 580, 2495-2502. 36. Muller-Taubenberger, A., Vos, M. J., Bottger, A., Lasi, M. and others (2006) Monomeric red fluorescent protein variants used for imaging studies in different species, Eur. J. Cell Biol. 85, 1119-1129. 37. Michels, A. A., Nguyen, V. T., Konings, A. W., Kampinga, H. H. and others (1995) Thermostability of a nuclear- targeted luciferase expressed in mammalian cells. Destabilizing influence of the intranuclear microenvironment, Eur. J. Biochem. 234, 382-389. 38. Hageman, J., Vos, M. J., van Waarde, M. A., and Kampinga, H. H. (2007) Comparison of intra-organellar chaperone capacity for dealing with stress-induced protein unfolding, J. Biol. Chem. 282, 34334-34345. 39. Horwitz, J. (1992) Alpha-crystallin can function as a molecular chaperone, Proc. Natl. Acad. Sci. U. S. A 89, 10449- 10453. 40. Sugiyama, Y., Suzuki, A., Kishikawa, M., Akutsu, R. and others (2000) Muscle develops a specific form of small heat shock protein complex composed of MKBP/HSPB2 and HSPB3 during myogenic differentiation, J. Biol. Chem. 275, 1095-1104. 41. Breitkreutz, B. J., Stark, C., Reguly, T., Boucher, L. and others (2008) The BioGRID Interaction Database: 2008 update, Nucleic Acids Res. 36, D637-D640. 42. Obayashi, T., Hayashi, S., Shibaoka, M., Saeki, M. and others (2008) COXPRESdb: a database of coexpressed gene networks in mammals, Nucleic Acids Res. 36, D77-D82. 43. Overbeek, P. A., Chepelinsky, A. B., Khillan, J. S., Piatigorsky, J. and others (1985) Lens-specific expression and developmental regulation of the bacterial chloramphenicol acetyltransferase gene driven by the murine alpha A- crystallin promoter in transgenic mice, Proc. Natl. Acad. Sci. U. S. A 82, 7815-7819. 44. Krief, S., Faivre, J. F., Robert, P., Le, D. B. and others (1999) Identification and characterization of cvHsp. A novel human small stress protein selectively expressed in cardiovascular and insulin-sensitive tissues, J. Biol. Chem. 274, 36592-36600. 45. Kappe, G., Verschuure, P., Philipsen, R. L., Staalduinen, A. A. and others (2001) Characterization of two novel human small heat shock proteins: protein kinase-related HspB8 and testis-specific HspB9, Biochim. Biophys. Acta 1520, 1-6. 46. Siddique, M., Gernhard, S., von Koskull-Doring, P., Vierling, E. and others (2008) The plant sHSP superfamily: five new members in Arabidopsis thaliana with unexpected properties, Cell Stress. Chaperones. 13, 183-197. 47. Morrow, G., Inaguma, Y., Kato, K., and Tanguay, R. M. (2000) The small heat shock protein Hsp22 of Drosophila melanogaster is a mitochondrial protein displaying oligomeric organization, J. Biol. Chem. 275, 31204-31210. 48. Horton, P., Park, K. J., Obayashi, T., Fujita, N. and others (2007) WoLF PSORT: protein localization predictor, Nu- cleic Acids Res. 35, W585-W587. 49. Guda, C. (2006) pTARGET: a web server for predicting protein subcellular localization, Nucleic Acids Res. 34, W210-W213. 50. Hoglund, A., Donnes, P., Blum, T., Adolph, H. W. and others (2006) MultiLoc: prediction of protein subcellular localization using N-terminal targeting sequences, sequence motifs and amino acid composition, Bioinformatics. 22, 1158-1165. 51. Guda, C., Guda, P., Fahy, E., and Subramaniam, S. (2004) MITOPRED: a web server for the prediction of mitochondrial proteins, Nucleic Acids Res. 32, W372-W374. 52. Jin, J. K., Whittaker, R., Glassy, M. S., Barlow, S. B. and others (2008) Localization of phosphorylated alphaB- crystallin to heart mitochondria during ischemia-reperfusion, Am. J. Physiol Heart Circ. Physiol 294, H337-H344. 53. Sprenger, J., Lynn, F. J., Karunaratne, S., Hanson, K. and others (2008) LOCATE: a mammalian protein subcellular localization database, Nucleic Acids Res. 36, D230-D233. 54. Mishra, G. R., Suresh, M., Kumaran, K., Kannabiran, N. and others (2006) Human protein reference database-- 2006 update, Nucleic Acids Res. 34, D411-D414. 55. Andley, U. P., Song, Z., Wawrousek, E. F., and Bassnett, S. (1998) The molecular chaperone alphaA-crystallin enhances lens epithelial cell growth and resistance to UVA stress, J. Biol. Chem. 273, 31252-31261. 56. Mackay, D. S., Andley, U. P., and Shiels, A. (2003) Cell death triggered by a novel mutation in the alphaA-crystallin gene underlies autosomal dominant cataract linked to chromosome 21q, Eur. J. Hum. Genet. 11, 784-793. 57. van Rijk, A. E., Stege, G. J., Bennink, E. J., May, A. and others (2003) Nuclear staining for the small heat shock protein alphaB-crystallin colocalizes with splicing factor SC35, Eur. J. Cell Biol. 82, 361-368. 58. Noh, S. J., Jeong, W. J., Rho, J. H., Shin, D. M. and others (2008) Sensitization of RPE cells by alphaB-crystallin siRNA to SAHA-induced stage 1 apoptosis through abolishing the association of alphaB-crystallin with HDAC1 in SC35 speckles, Invest Ophthalmol. Vis. Sci. 49, 4753-4759.

51 Chapter 3

59. den Engelsman, J., Gerrits, D., de Jong, W. W., Robbins, J. and others (2005) Nuclear import of {alpha}B-crystallin is phosphorylation-dependent and hampered by hyperphosphorylation of the myopathy-related mutant R120G, J. Biol. Chem. 280, 37139-37148. 60. Nakagawa, M., Tsujimoto, N., Nakagawa, H., Iwaki, T. and others (2001) Association of HSPB2, a member of the small heat shock protein family, with mitochondria, Exp. Cell Res. 271, 161-168. 61. Carra, S., Sivilotti, M., Chavez Zobel, A. T., Lambert, H. and others (2005) HspB8, a small heat shock protein mutated in human neuromuscular disorders, has in vivo chaperone activity in cultured cells, Hum. Mol. Genet. 14, 1659-1669. 62. Carra, S., Seguin, S. J., Lambert, H., and Landry, J. (2008) HspB8 chaperone activity toward poly(Q)-containing proteins depends on its association with Bag3, a stimulator of macroautophagy, J. Biol. Chem. 283, 1437-1444. 63. Fontaine, J. M., Rest, J. S., Welsh, M. J., and Benndorf, R. (2003) The sperm outer dense fiber protein is the 10th member of the superfamily of mammalian small stress proteins, Cell Stress. Chaperones. 8, 62-69. 64. Johnston, J. A., Ward, C. L., and Kopito, R. R. (1998) Aggresomes: a cellular response to misfolded proteins, J. Cell Biol. 143, 1883-1898. 65. de Wit, N. J., Verschuure, P., Kappe, G., King, S. M. and others (2004) Testis-specific human small heat shock protein HSPB9 is a cancer/testis antigen, and potentially interacts with the dynein subunit TCTEL1, Eur. J. Cell Biol. 83, 337-345. 66. Geum, D., Son, G. H., and Kim, K. (2002) Phosphorylation-dependent cellular localization and thermoprotective role of heat shock protein 25 in hippocampal progenitor cells, J. Biol. Chem. 277, 19913-19921. 67. Hattori, H., Kaneda, T., Lokeshwar, B., Laszlo, A. and others (1993) A stress-inducible 40 kDa protein (hsp40): puri- fication by modified two-dimensional gel electrophoresis and co-localization with hsc70(p73) in heat-shocked HeLa cells, J. Cell Sci. 104 ( Pt 3), 629-638. 68. Tanonaka, K., Toga, W., Takahashi, M., Kawana, K. and others (2003) Hsp70 attenuates hypoxia/reoxygenation- induced activation of poly(ADP-ribose) synthetase in the nucleus of adult rat cardiomyocytes, Mol. Cell Biochem. 248, 149-155. 69. Krejci, J., Harnicarova, A., Kurova, J., Uhlirova, R. and others (2008) Nuclear organization of PML bodies in leukae- mic and multiple myeloma cells, Leuk. Res. 32, 1866-1877. 70. Bernardi, R. and Pandolfi, P. P. (2007) Structure, dynamics and functions of promyelocytic leukaemia nuclear bodies, Nat. Rev. Mol. Cell Biol. 8, 1006-1016. 71. Stanek, D. and Neugebauer, K. M. (2006) The Cajal body: a meeting place for spliceosomal snRNPs in the nuclear maze, Chromosoma 115, 343-354. 72. Misteli, T., Caceres, J. F., and Spector, D. L. (1997) The dynamics of a pre-mRNA splicing factor in living cells, Nature 387, 523-527. 73. Caceres, J. F., Misteli, T., Screaton, G. R., Spector, D. L. and others (1997) Role of the modular domains of SR proteins in subnuclear localization and alternative splicing specificity, J. Cell Biol. 138, 225-238. 74. Michaud, S., Lavoie, S., Guimond, M. O., and Tanguay, R. M. (2008) The nuclear localization of Drosophila Hsp27 is dependent on a monopartite arginine-rich NLS and is uncoupled from its association to nuclear speckles, Biochim. Biophys. Acta 1783, 1200-1210. 75. Nollen, E. A., Salomons, F. A., Brunsting, J. F., Want, J. J. and others (2001) Dynamic changes in the localization of thermally unfolded nuclear proteins associated with chaperone-dependent protection, Proc. Natl. Acad. Sci. U. S. A 98, 12038-12043. 76. Kundu, M., Sen, P. C., and Das, K. P. (2007) Structure, stability, and chaperone function of alphaA-crystallin: role of N-terminal region, Biopolymers 86, 177-192. 77. Matera, A. G. (1999) Nuclear bodies: multifaceted subdomains of the interchromatin space, Trends Cell Biol. 9, 302- 309. 78. Alvarez, M., Estivill, X., and de la, L. S. (2003) DYRK1A accumulates in splicing speckles through a novel targeting signal and induces speckle disassembly, J. Cell Sci. 116, 3099-3107. 79. Isshiki, M., Tsumoto, A., and Shimamoto, K. (2006) The serine/arginine-rich protein family in rice plays important roles in constitutive and alternative splicing of pre-mRNA, Plant Cell 18, 146-158. 80. Cazalla, D., Zhu, J., Manche, L., Huber, E. and others (2002) Nuclear export and retention signals in the RS domain of SR proteins, Mol. Cell Biol. 22, 6871-6882. 81. Shin, C., Kleiman, F. E., and Manley, J. L. (2005) Multiple properties of the splicing repressor SRp38 distinguish it from typical SR proteins, Mol. Cell Biol. 25, 8334-8343. 82. Herrmann, C. H. and Mancini, M. A. (2001) The Cdk9 and cyclin T subunits of TAK/P-TEFb localize to splicing factor- rich nuclear speckle regions, J. Cell Sci. 114, 1491-1503. 83. Rujano, M. A., Bosveld, F., Salomons, F. A., Dijk, F. and others (2006) Polarised asymmetric inheritance of accumulated protein damage in higher eukaryotes, PLoS. Biol. 4, e417.

52 HSPB7 localizes to SC35 speckles

Supplemental tables and figures

Table S1 Primers used in this study

Gene name Alternative name Human Origen Primer Name Primers sequence 5' - 3' Restriction Gene ID site HSPB1 HSP27; Hsp25 3315 cDNA mix B1-F ACAACTCGAGAAGCATGACCGAGCGCCGCGTC XhoI B1-R ACAAACGCGTTTACTTGGCGGCAGTCTCAT MluI HSPB2 MKBP; HSP27 3316 Open Biosystems B2-F ACAACTCGAGAACCATGTCGGGCCGCTCAGTGCCA XhoI B2-R ATAAACGCGTTTTGCTGGGTGCTGGGTCTG MluI HSPB3 HSPL27 8988 cDNA mix B3-F ACAAGCGGCCGCACCATGGCAAAAATCATTTTGAGGC NotI B3-R ACCAACGCGTTCTGAACAGGAACCGATACG MluI HSPB4 crystallin alpha A 1409 cDNA mix B4-F ACAACTCGAGAACCATGGACGTGACCATCCAGCA XhoI B4-R ACAAACGCGTCCGAGGCAATGCCTGCTTAG MluI HSPB5 crystallin alpha B 1410 cDNA mix B5-F ACAACTCGAGAACCATGGACATCGCCATCCACCAC XhoI B5-R ACAAACGCGTAATTTGGGCCTGCCCTTAGC MluI HSPB6 HSP20 126393 Open Biosystems B6-F ACAACTCGAGAACCATGGAGATCCCTGTGCCTGTG XhoI B6-R ACAAACGCGTGAGCCTGAGGAGGCTCCCGG MluI HSPB7 cvHSP 27129 Open Biosystems B7-F ACCGCTCGAGAACCATGAGCCACAGAACCTCTTCC XhoI B7-R CAATACGCGTCAGGGAAAGGGAAGGGAGAG MluI HSPB8 H11; HMN2; HSP22 26353 Open Biosystems B8-F ACAACTCGAGAACCATGGCTGACGGTCAGATG XhoI B8-R ACCCACGCGTCGCTATCCTGTGGAGAAATC MluI HSPB9 FLJ27437 94086 Open Biosystems B9-F ACCCCTCGAGAACCATGCAGAGAGTCGGTAACAC XhoI B9-R ACAGACGCGTTTACCGGGTCAGGTTGGAAG MluI HSPB10 ODF1; ODF 4956 Open Biosystems B10-F ACCGCTCGAGAACCATGGCTGCACTGAGTTGTCT XhoI B10-R ACCCACGCGTTGGCTGGAGTAACTGACTTC MluI  N-HSPB7 This study p7012 deltaN F AACATCAAGACCCTAGGAGACG none p7011 deltaN R GGTTCTCGAGGTTTTGGATC none NB1- NB7 This study p9033 HSPB1 N R CCCGCTGCTGAGTTGCCGGCTGAG none p9034 HSPB7 CRY F GCAGGCAACATCAAGACCCTAGGAGAC none p9027 AMP N ATGCTTTTCTGTGACTGGTG none p9028 AMP C CTTACGGATGGCATGACAGT none NB7- NB1 This study p9036 HSPB7 N R CCCACCGGGGCGGGCTGGGAAG none p9035 HSPB1 CRY F GTCTCGGAGATCCGGCACAC none p9027 AMP N ATGCTTTTCTGTGACTGGTG none p9028 AMP C CTTACGGATGGCATGACAGT none NB7-mRFPruby p9043 mRFP-B7 F ATGGCCAGCTCCGAGGATGTC none p9042 B7-mRFP R CCCACCGGGGCGGGCTGGGAAG none p9027 AMP N ATGCTTTTCTGTGACTGGTG none p9028 AMP C CTTACGGATGGCATGACAGT none

Table S2 Plasmids used in this study Name Description Origen Cloning pCDNA5/FRT/TO Mammalian expression vector; tetracycline regulatable expression Invitrogen pCDNA5/FRT/TO-MV1 Adapted multiple cloning site: XhoI-AgeI-AscI This study pCDNA5/FRT/TO-MV2 Adapted multiple cloning site: NotI-AgeI-AscI This study pCDNA5/FRT/TO-V5-MV1 Contains N-terminal V5 tag sequence This study pCDNA5/FRT/TO-V5-MV2 Contains N-terminal V5 tag sequence This study pCDNA5/FRT/TO-EGFP-MV1 Contains N-terminal EGFP sequence This study pCDNA5/FRT/TO-EGFP-MV2 Contains N-terminal EGFP sequence This study pCDNA5/FRT/TO-MYC-MV1 Contains N-terminal MYC sequence This study pCDNA5/FRT/TO-HIS-MV1 Contains N-terminal HIS sequence This study HSPB pCDNA5/FRT/TO-V5-MV1-HSPB1 V5 tagged human HSPB1 This study pCDNA5/FRT/TO-V5-MV1-HSPB2 V5 tagged human HSPB2 This study pCDNA5/FRT/TO-V5-MV2-HSPB3 V5 tagged human HSPB3 This study pCDNA5/FRT/TO-V5-MV1-HSPB4 V5 tagged human HSPB4 This study pCDNA5/FRT/TO-V5-MV1-HSPB5 V5 tagged human HSPB5 This study pCDNA5/FRT/TO-V5-MV1-HSPB6 V5 tagged human HSPB6 This study pCDNA5/FRT/TO-V5-MV1-HSPB7 V5 tagged human HSPB7 This study pCDNA5/FRT/TO-MYC-MV1-HSPB7 MYC tagged human HSPB7 This study pCDNA5/FRT/TO-V5-MV1-HSPB8 V5 tagged human HSPB8 This study pCDNA5/FRT/TO-V5-MV1-HSPB9 V5 tagged human HSPB9 This study pCDNA5/FRT/TO-V5-MV1-HSPB10 V5 tagged human HSPB10 This study pCDNA5/FRT/TO-V5-NB7 V5 tagged human HSPB7 lacking N-termins (1-73) This study pCDNA5/FRT/TO-V5-NB1-NB7 V5 tagged human HSPB7 with N-terminus of HSPB1 This study pCDNA5/FRT/TO-V5-NB7-NB1 V5 tagged human HSPB1 with N-terminus of HSPB7 This study Imaging pCDNA5/FRT/TO-H2B-mRFPruby Human histone 2B fused to monomeric red fluorescent protein ruby [65] pCDNA5/FRT/TO-NB7-mRFPruby N-terminus of HSPB7 fused to mRFPruby This study pSV2-EYFP-SP100 EYFP tagged SP100; PML body marker Spector Laboratory pCMV-EYFP-SF2/ASF EYFP taggen SF2/ASF; SR marker Spector Laboratory pTRE-GFP-coilin EGFP taggen human Coilin; Cajal body marker Matera Laboratory SC35-eGFP EGFP tagged SC35; SR marker Tazi Laboratory

53 Chapter 3

A Figure S1. The human HSPB plas- MCS MCS Available constructs 1 2 mid library. (A) The human HSPB Ampicillin R Pto HSPB1 V5 His eGFP NotI 1 XhoI 1 family was cloned into an adapted AgeI 12 AgeI 12 HSPB2 V5 His eGFP AscI 20 AscI 20 HSPB3 V5 His eGFP pCDNA5/FRT/TO vector backbone pCDNA5/FRT-MV* ATG ATG ATG ATG HSPB4 V5 His eGFP containing two different multiple Myc His HSPB5 (5121 bp) V5 V5 His eGFP cloning sites (MCS). Subsequently, MCS HSPB6 V5 His eGFP MCS MCS MCS HSPB7 V5 His eGFP mRFP Myc the HSPB members were subcloned HSPB8 V5 His eGFP mRFP Hygromycin R eGFP MCS into pCDNA5/FRT/MV vectors con- ATG HSPB9 V5 His eGFP taining various tags. (B) In combi- HSPB10 V5 His eGFP mRFPruby MCS ATG nation with the Tet repressor, this B tetracycline library can be used for tetracycline V5-HSPB1 regulated expression.

A HSPB1 HSPB2 HSPB3 HSPB4 HSPB5 heat shock heat shock + recovery

HSPB6 HSPB7 HSPB8 HSPB9 HSPB10 heat shock heat shock + recovery

B HSPB1 HSPB2 HSPB3 HSPB4 HSPB5 HSPB6 HSPB7 HSPB8 HSPB9 HSPB10 cytosol xxxx xxxx x x rol nucleus x x x x xxx on t

C nuclear foci x x x

t cytosol xxxx x x x x x x a ck

o nucleus xxxx x x x x x x He s h nuclear foci x x x x x x x x x

Figure S2. Intracellular distribution of the human HSPB family after heat shock. (A) Intracellular distribution of V5-tagged human HSPB proteins in HEK293 cells either directly after a 30 minute heat shock treatment (43oC) or combined with a recovery period of 45 minutes at 37oC. (B) Summary of HSPB localization at control conditions or after heat shock treatment. Grey boxes indicate nuclear foci which have been found positive for SC35 staining. The DAPI signal for nuclear staining was traced in Adobe photoshop and depicted as a dashed line.

54 HSPB7 localizes to SC35 speckles

V5-HSPB4 Luciferase Figure S3. Localization of V5-HSPB4 and nuclear luciferase during heat shock. HEK293 cells overexpressing V5-HSPB4 and EGFP tagged nuclear luciferase were heat shocked (30 minutes, 43oC) and allowed to recover for 45 minutes at 37oC after which cells were fixed and analysed by confocal microscopy. See Appendix 4 for colour print.

DAPI Merge

DmHSP27 1 --msiipllhlareldhdyrtdw---ghlleddfgfgvhahdlf------hp------HSPB1 1 -mterrvpfsllrgpswdpfrdwyphsrlfdqafglprlpeewsqwlggsswpg------HSPB5 1 --mdiaihhpwir----rpffpfhspsrlfdqffgehllesdlfp-tstslspf------HSPB7 1 mshrtsstfraersfhsss----sssssstsssasralpaqd--ppmekalsmfsddfgs

DmHSP27 42 -rrlllpntlglgrrryspyershghhnqmsrrasggpnall 17,1% Leu HSPB1 54 yvrplppaai------espavaapaysralsrqlssgvs--- 10,5% Leu HSPB5 48 ylrp------psflraps-wfdtgls--- 12,1% Leu HSPB7 55 fmrp------hseplafpa------rpggag----- 4,1% Leu Figure S4. Protein alignment of N-terminal regions of SC35 speckle associated small heat shock proteins. The N-termini of Drosophila HSP27 and human HSPB1, HSPB5 and HSPB7 were aligned using the BLOSUM 62 scoring matrix. Leucines are indicated in bold and the percentage of leucines present in the N-terminal sequences are indicated. The protein sequence required for Drosophila HSP27 to associate with SC35 splicing speckles is indicated by a box.

55 Chapter 3

56 CHAPTER 4

A comparative screen of the human HSPB family of molecular chaperones reveals HSPB7 as a superior suppressor of aggregation of polyglutamine containing proteins

Michel J Vos*, Marianne P Zijlstra* & Harm H Kampinga

Department of Cell Biology, Section of Radiation and Stress Cell Biology, University Medical Center Groningen, Groningen, The Netherlands

* Equal contribution

To be submitted Chapter 4

Abstract

A small number of heat shock proteins have previously been shown to act protectively on aggregation of several proteins containing an extended polyglutamine tract, which are linked to several neurodegenerative diseases. A specific subfamily of heat shock proteins is formed by the HSPB family of molecular chaperones, of which a few members are known to act as anti- aggregation proteins in vitro. Furthermore, mutations in a subset of HSPB members are linked to neuropathies, underlining the importance of this subfamily in protein folding, aggregation and neurological diseases. Using overexpression, we screened the HSPB family for novel inhibitors of polyglutamine aggregation and provide a comparative overview on the human HSPB family in view of preventing polyglutamine aggregation. This screen revealed several HSPB members as novel suppressors of polyglutamine aggregation, with HSPB7 as the most active one. Interestingly, this HSPB protein was found to function independently of the HSP70 machine and heat-shock respons. Furthermore, it did not target polyglutamine proteins for proteasomal degradation, but was found to depend in part on autophagy for its activity. Moreover, it prevented polyglutamine induced toxicity in cells and in an in vivo model for SCA3, showing the potential of HSPB7 as a potential therapeutic target.

58 HSPB7 is a novel suppressor of polyglutamine aggregation

Introduction

Several inherited neurodegenerative diseases exist which are based on a CAG triplet expansion in the affected genes. This results in elongation of the glutamine (polyQ) tract present in the proteins where, in general, an extension beyond 35-40 glutamines causes disease whereby the age at onset is related to the length of the expansion (1). These diseases include amongst others polyQ extensions in the androgen receptor, the TATA-box binding protein, huntingtin and several ataxia-related genes (SCA genes). The corresponding diseases are characterized by protein misfolding and aggregate / inclusion formation of the affected proteins with a concurrent increase in proteotoxic stress. From (cellular) studies it has become clear that certain heat shock proteins, as molecular chaperones, form a potent natural defense against proteotoxic stress induced by these protein misfolding diseases. Heat shock proteins are divided into several main families which are both structurally and functionally highly divergent (2). Within the HSPH (Hsp110) , HSPA (Hsp70) and DNAJ (Hsp40) families, several members have been found that can reduce aggregation and toxicity associated with polyglutamine (polyQ) containing proteins and other misfolded proteins (3-10). Also within the mammalian small heat shock protein (HSPB) family, some members, in particularly HSPB1 and HSPB8, have been reported to suppress toxicity or aggregation of neuropathy-causing mutated proteins (11;12). In cells, HSPB1 suppresses mutant huntingtin- induced reactive oxygen species formation and cell death which is dependent on the phosphorylation status of HSPB1 (13). In a mouse model of Huntington disease (HD), however, HSPB1 overexpression did not improve the HD phenotype and furthermore it failed to form large HSPB1-substrate complexes which are implicated in chaperone activity (14). HSPB8 acts via another mechanism and modulates toxicity of HD via the induction of autophagy and a translational shut-down (15;16). Whether or not HSPB8 can also rescue HD pathology in vivo is yet unknown. Furthermore, the potency of the other HSPB members in reducing proteotoxic stress remains unknown. The human HSPB family consists of eleven members and the amino acid sequence variation between members of this family is larger than in any of the other HSP families. Here, we present a systematic comparison of ten members of the human HSPB family for their ability to modulate either HD or SCA3 aggregation or assist in the refolding of a heat-denatured substrate. The effectiveness to prevent aggregation largely differed between members and, in part, was dependent on the type of the misfolded substrate. However, within the HSPB family, HSPB7 was overall the most potent suppressors of HD and SCA3 aggregation. In addition, HSPB7 was able to reduce polyQ related toxicity in cells and protected against polyQ-related eye-degeneration in a Drosophila SCA3 model. In contrast to e.g. HSPB1, HSPB7 is not a classical chaperone since it is unable to assist in refolding of heat denatured proteins. In addition, it does not depend on the HSPA chaperone machine for its anti-aggregation activity. Although its precise mode of action remains to be elucidated, HSPB7 is in part dependent on autophagy to decrease polyQ aggregation.

Experimental procedures

Reagents and antibodies Chemicals were obtained from Sigma, beetle luciferin from Promega. Antibodies against the V5 tag (Invitrogen), EGFP (JL-8) (BD bioscience), GAPDH (6C5) (RDI research diagnostics), HA tag (Covance), HSPB7 (Abcam) and HSPA1A (SPA-800), HSPA6 (SPA-754) from Stressgen were mouse monoclonal. The antibodies against DNAJB1 (SPA-400) (Stressgen) and HSPB5 (Abcam) were rabbit-polyclonal. Reagents for Western blotting were obtained from Amersham.

59 Chapter 4

Secondary antibodies for immunocytochemistry used were CY5 anti-mouse and FITC anti- mouse (Jackson) and CY3 anti-mouse (Amersham).

Molecular techniques Standard recombinant DNA techniques were carried out essentially as described by Sambrook et al. Oligonucleotide primers (Biolegio). Restriction enzymes and DNA polymerases were used according to the manufacturer’s instructions (Invitrogen, New England Biolabs). DNA sequencing reactions were carried out by ServiceXS.

Plasmid generation The truncated SCA3 fragments were PCR amplified from skin fibroblasts from heterozygous SCA3 partient with 64 and 82 CAG-repeat present in the Ataxin3 gene and cloned into pCMV- EGFP and pCDNA5.1-FRT/TO-EGFP respectively, in frame with EGFP. Plasmid containing either HDQ43 or HDQ74 were kindly provided by Dr. David Rubinsztein (Cambridge, UK). Gen- eration of the human HSPB plasmid library was described before as well as the HSPB7 deletion mutants (Chapter 3). For retroviral transduction, the pQCXIN vector (Clontech) was used for cloning. The Ub-R-EGFP construct was kindly provided by Dr. Dantuma (Karolinska Institutet, Sweden) The construct encoding His-BAG3 and HA-BAG1 have been described before (21;44). The construct encoding the dominant negative HSF-1 was kindly provided by Prof. dr. N.H. Lub- sen (Nijmegen, the Netherlands).

Cell lines, cell culture, transfections and transductions NG108, mouse embryonic fibroblasts (MEF) and human embryonic kidney 293 cells expressing the tetracycline repressor (Flp-In T-Rex HEK293, Invitrogen) were grown in Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10% fetal calf serum (Griener Bio-one) at 37

°C under a humidified atmosphere containing 5% CO2. The generation of HEK293-HDQ119, with integration of a tetracycline regulatable expression vector coding for EGFP-HDQ119, has been described before (Hageman et al. submitted elsewere). Gene expression in Flp-In T-Rex HEK293 cells was induced with a final concentration of 1 µg/mL tetracycline. HEK293 cells and MEF cells were transfected using Lipofectamine (Invitrogen) according to the manufacturer’s instructions using 1μg of plasmid DNA per 35-mm dish. NG108 cells were differentiated into neuronal-like cells by addition of 2 mL differentiation medium (DMEM containing 1 mM sodium pyruvate, 50 μM 3-Isobutyl-1-methylxanthine, 10 μM 5’-(N-Ethylcarboxamido)adenosine, 0,5% FCS) to 35000 cells in a 35mm dish. Seven days later, cells were transfected using Effectene (Qiagen) according to the manufacturer’s instructions using 0,6 μg of plasmid DNA per 35-mm dish. For transduction, retroviral particles were generated by co-transfection of HEK293-T cells with pCL-ampho and pQCXIN endoding either HSPB5 or HSPB7 using Fugene (Roche) according to the manufacturers instructions. The supernatant was collected after 48 hours and filtered using a Millex-HV filter (Millipore). Cells were transduced by adding 1,5 mL of filtrated medium containing 4 µg/mL polybrene. Medium was replaced after 24 hours and transduced cells were selected using puromycin selection.

Drosophila Fly stocks were maintained at 22°C according to standard protocols. GAL4 driver stocks were obtained from the Bloomington Stock Centre (Indiana University, USA). The GMR-UAS- SCA3Q78 fly used for the eye-degeneration screen was generously provided by N. Bonini (Uni- versity of Pennsylvania, USA) and maintained at 25 °C. HSPB7 transgenic lines were generated

60 HSPB7 is a novel suppressor of polyglutamine aggregation by Genetic Services Inc. (Sudbury, USA) by injection of the pUAS transformation vector into the W1118 genetic background.

Westernblot analysis Samples were prepared in Laemmli sample buffer. Equal amount of protein were separated on 10% or 12,5% SDS-PAGE gels followed by transfer onto nitrocellulose membranes. Primary antibodies were used at the following dillutions: V5 (1:5000), EGFP (1:5000), GAPDH (1:10000), HSPA1A (1:5000), HSPA6 (1:5000), DNAJB1 (1:5000). After a one hour incubation with the primary antibody in PBS-Tween20, membranes were incubated with HRP-conjugated secondary antibodies at a 1:5000 dillution. Detection was performed using enhanced chemiluminescence.

Filtertrap assay The filtertrap assay was performed basically as described by Carra et al. (12). Briefly, cells grown in a 35mm dish were lysed in 200 μL FTA buffer (10 mM Tris-Cl pH 8.0, 150 mM NaCl, 50 mM dithiothreitol) containing 2% SDS. 100 μL, 20 μL and 4 μL sample was applied onto a prewashed (FTA, 0,1% SDS) 0,2 μM cellulose acetate filter contained in a Bio-Dot microfiltration apparatus (Biorad). Gentle suction was applied to filtrate the samples followed by one washing step using FTA, 0,1% SDS. Trapped material was probed with anti-EGFP (GL-8) or anti-HA at a 1:5000 and 1:2000 dillution respectively followed by incubation with HRP-conjugated secondary antibodies at a 1:5000 dillution. Detection was performed using enhanced chemiluminescence.

Microscopy For microscopy, cells were plated 24 hours before transfection. For fixation, the coverslips were washed with cold PBS and fixed for 15 minutes with methanol (-20o C). Cells were permeabilized in blocking solution (100 mM glycine, 3% BSA, 0,1% triton) for 1 hour followed by 1 hour incubation with the primary antibody (V5 anti-mouse, 1:200, SC-35 anti-mouse, 1:10000, MYC anti- rabbit 1:200). After three washing steps, coverslips were incubated for 1h with CY3-conjugated anti-rabbit secondary antibody (Amersham Biosciences) at 1:200 dilution or with FITC-conjugated anti-mouse secondary antibody (Jackson) at 1:200 dilution. After three washing steps, the coverslips were mounted using Citifluor mounting medium (Citifluor Ltd). Images were obtained using an inverted confocal laser scanning microscope (TCS SP2, DM RXE, Leica) with a 63×/1.32 NA oil objective.

Luciferase refolding assay Chaperone activity of HSPB members was assessed by using the luciferase refolding assay (41). Briefly, HEK293 cells were co-transfected with luciferase (NucSuperluc-EGFP) (59)together with HSPB encoding plasmids (1:9 ratio). Two hours after transfection, expression was induced by addition of tetracycline. After 24 hours, cells were resuspended and divided into 1 mL portions in tissue-grade 10 mL tubes. The following day, cells were given a heat-shock (30 minutes at 43o C) in the presence of cycloheximide (20 μg/mL) and 4-morpholinepropanesulfonic acid (MOPS; 20 mM, pH7.0) in order to inhibit protein synthesis and increase the buffer capacity of the medium respectively. After heat treatment, cells were allowed to recover before luciferase activity was determined (3 hours at 37 o C). Luciferase activity measurements were performed using a Berthold Lumat 9510 luminometer (Berthold Technologies).

Sucrose gradient centrifugation Cells from a 35mm dish were scaped in 200 μL cold lysis buffer (150 mM NaCl, 50 mM NaH2- PO4, 10 mM imidazole, 0,5% NP-40, 1,5 mM MgCl2 and 3% glycerol, pH8). Cells were lysed by five passages through a 25 gauge needle, followed by centrifugation at 300G for 15 minutes.

61 Chapter 4

The cell lysate was loaded on top of a 10-80% sucrose gradient (10 mM Tris-HCl, pH8, 5 mM EDTA, 50 mM NaCl) and centrifuged for 18 hours at 100000g using a Sorvall Discovery 90SE ultra centrifuge and the SW55 rotor. Fractions of 400 μL were precipitated by adding an equal volume of 25% (w/v) trichloroacatic acid. After a 30 minute incubation on ice, proteins were pelleted at 14000 rpm for 15 minutes. The pellet was washed twice with 80% acetone (-200C) and allowed to dry. The pellet was dissolved in 20 μL 0,1M NaOH containing 1% SDS. An equal amount of Laemmli buffer was added and samples were boiled for five minuten follwed by West- ern blot analysis.

Clonogenic assay A clonogenic assay was used to determine cellular survival when overexpressing EGFP- HDQ119 without or with HSPB members. Exponentially growing cells were plated in triplicate in 10 cm dishes at a density of 200 cells per dish. EGFP-HDQ119 expression was activated by addition of tetracycline. Cells were allowed to grow and form colonies for two weeks, after which they were fixed and stained (0,1% Coomassie Brilliant Blue, 50% methanol, 10% glacial acetic acid followed by washing (10% methanol, 7,5% glacial acetic acid). The effect of EGFP-HDQ119 expression and co-expression of HSPB5 and HSPB7 was determined by comparing colony reduction in the presence of EGFP-HDQ119 expression.

Proteasome inhibition, proteasome activity measurement and autophagy analysis Proteasomal activity was blocked by addition of 10 μM MG132 to the culture medium for 12 hours. Proteasomal activity was monitored by co-transfection of UB-R-EGFP followed by West- ernblot- and FACS analysis. The role of autophagy in reducing polyQ aggregation was analysed using mouse embryonic fibroblasts lacking the ATG5 gene which was kindly provided by Dr. Mzushima, Okazaki, Japan.

Results

HSPB7 is the most potent HSPB member in preventing polyglutamine aggregation Certain members of the HSPB family have been reported to suppress protein aggregation in vitro using several model proteins (17-21). Specific substrates include amongst others, amyloid beta (19), desmin filaments (20) and mutated huntingtin (21). Apart from these specific substrates, HSPB proteins can bind to many other non-native proteins in general (22-24). In cell free systems, several HSPB members were shown to be rather generic in preventing aggregation of a number of unrelated proteins and support their refolding via interacting with the HSPA1A (HSP70) machine (25). In cells, HSPB proteins are found to interact with cytoskeletal elements but may also here act rather promiscuously in preventing protein aggregation and support protein refolding (26-28). To compare the effectiveness of the HSPB family in dealing with disease-related misfolded proteins, we co-expressed them together with either mutant Huntingtin exon 1 (Htt) or a fragment of Ataxin-3 (SCA3), both containing an expanded glutamine tract. Using the filtertrap assay (29) to detect protein aggregates, we found that aggregation of a moderately expanded HD protein (HA-HttQ43) was strongly reduced by co-expression of four HSPB members (HSPB6, HSPB7, HSPB8, HSPB9), whereas HSPB1 and HSPB4 were only marginally effective (Figure 1a). HSPB2, HSPB3, HSPB5 and HSPB10 showed no activity at all. Using a HD fragment containing a longer polyQ tracts (EGFP-HttQ74), co-expression of HSPB7 and HSPB9 resulted in a considerable reduction of aggregated material, while co-expression of either HSPB1, HSPB4

62 HSPB7 is a novel suppressor of polyglutamine aggregation

A V5-tagged HttQ43 screen B V5-tagged HttQ74 screen

mRFP V5-HSPB5 mRFP V5-HSPB5

V5-HSPB1 V5-HSPB6 V5-HSPB1 V5-HSPB6

V5-HSPB2 V5-HSPB7 V5-HSPB2 V5-HSPB7

V5-HSPB3 V5-HSPB8 V5-HSPB3 V5-HSPB8

V5-HSPB4 V5-HSPB9 V5-HSPB4 V5-HSPB9 V5-HSPB10 HA-HttQ43 EGFP-HttQ74 V5-HSPB10 C V5-tagged HttQ119 screen D Non-tagged HttQ73 screen

mRFP V5-HSPB5 mRFP HspB5

V5-HSPB1 V5-HSPB6 HspB1 HspB6

V5-HSPB2 V5-HSPB7 HspB2 HspB7

V5-HSPB3 V5-HSPB8 HspB3 HspB8

V5-HSPB4 V5-HSPB9 HspB4 HspB9

V5-HSPB10 HspB10 EYFP-Q119 EGFP-HttQ74 E F G mRFP V5-HSPB1 V5-HSPB2 V5-HSPB3 V5-HSPB4 V5-HSPB5 V5-HSPB6 V5-HSPB7 V5-HSPB8 V5-HSPB9 V5-HSPB10 mRFP V5-HSPB1 V5-HSPB2 V5-HSPB3 V5-HSPB4 V5-HSPB5 V5-HSPB6 V5-HSPB7 V5-HSPB8 V5-HSPB9 V5-HSPB10 mRFP V5-HSPB1 V5-HSPB2 V5-HSPB3 V5-HSPB4 V5-HSPB5 V5-HSPB6 V5-HSPB7 V5-HSPB8 V5-HSPB9 V5-HSPB10 V5-tagged HttQ43 screen V5-tagged HttQ74 screen Non-tagged HttQ74 screen -HA -EGFP -EGFP -GAPDH -V5 -V5

-GAPDH -GAPDH Figure 1. The human HSPB family contains potent suppressors of polyQ expanded Huntingtin aggregation. HEK293 cells were co-transfected with V5-tagged HSPB members and polyglutamine encoding plasmids at a 9:1 plasmid ratio. Cells were lysed after 24 or 48 hours and analysed by the filtertrap-binding assay using five-fold dillutions. Several HSPB members were able to suppress polyglutamine aggregation of a protein with (A) 43 glutamines; HA-HttQ43 (B) 74 glutamines; EGFP-HttQ74 and (C) 119 glutamines; EGFP-HttQ119. (D) To compare the effect of the V5-tag, also non-tagged HSPB members were co-expressed with the EGFP-HttQ74 encoding plasmid, (E-G) Expression levels of the polyglutamine proteins and HSPB members were analyzed by Western blot analysis.

Figure 2. The human HSPB family contains A potent suppressors of polyQ expanded SCA3 aggregation. (A) HEK293 cells were mRFP V5-HSPB5 co-transfected with V5-tagged HSPB members V5-HSPB1 V5-HSPB6 and truncated EGFP tagged SCA3 containing a V5-HSPB2 V5-HSPB7 stretch of 82 glutamines at a 9:1 plasmid ratio.

V5-HSPB3 V5-HSPB8 After 24 hours cells were lysed and analyzed by V5-HSPB9 the filtertrap-binding assay using five-fold dillu- V5-HSPB4 tions. (B) Expression levels of EGFP-SCA3-Q82 V5-HSPB10 SCA3-Q82 24h and HSPB members were analyzed by Western blot analysis. B mRFP V5-HSPB1 V5-HSPB2 V5-HSPB3 V5-HSPB4 V5-HSPB5 V5-HSPB6 V5-HSPB7 V5-HSPB8 V5-HSPB9 V5-HSPB10 V5-tagged SCA3-Q82 screen

-V5

-V5 long -EGFP

63 Chapter 4

A - HSPB7 + HSPB7 D - HSPB7 + HSPB7 EGFP-HttQ74 NLS-EGFP-HttQ74

B E

V5-HSPB7 V5-HSPB7 Empty vector 1:9 1:0,6 Empty vector 1:9 1:0,6

Soluble Soluble

-GAPDH -GAPDH

V5 V5 EGFP-HttQ74 NLS-EGFP-HttQ74

C F Empty vector Empty vector

1:9 1:9

1:4,5 1:4,5

1:2,3 1:2,3 V5-HSPB7 1:1,2 V5-HSPB7 1:1,2

1:0,6 1:0,6

EGFP-HttQ74 NLS-EGFP-HttQ74 Figure 3. HSPB7 can prevent polyglutamine aggregation in the cytosol and nucleus. (A) Confocal microscopy of HEK293 cells overexpressing cytosolic EGFP-HttQ74 (green signal) without or with co-expression of V5-HSPB7 for 48 hours. The nucleus was stained with DAPI (grey signal) (B) Expression levels of V5-HSPB7 were modified by varying the amount of transfected plasmids while keeping the amount of EGFP-HttQ74 encoding plasmid constant. (C) Filtertrap-binding analysis shows a correlation between HSPB7 concentration and its effect on reducing polyglutamine aggregation. (D) Confocal microscopy of HEK293 cells overexpressing nuclear localized NLS-EGFP-HttQ74 (green signal) without or with co-expression of V5-HSPB7 for 48 hours. (E) Expression levels of V5-HSPB7 were modified by varying the amount of transfected plasmids while keeping NLS-EGFP-HttQ74 encoding plasmid constant. (F) Also for nuclear aggregation of polyglutamine proteins, a correlation exists between levels of HSPB7 and its activity in reducing aggregation of polyglutamine proteins. See Appendix 4 for colour print. or HSPB6 was less effective (Figure 1b). At a glutamine length of 119, only HSPB7 was still effective in reducing the amount of aggregated material, while HSPB9 showed only a marginal effect (Figure 1c). Repeating the screen with the non-tagged HSPB library and EGFP-HttQ74 led to similar results, i.e. that HSPB9 and especially HSPB7 are the most potent suppressors of polyQ aggregation (Figure 1d), indicating that the V5 tag had not affected the function of most HSPB members. Under all conditions of overexpressed HSPB members, the expression levels of the soluble fraction of the various Huntingtin proteins was not largely affected (Figure 1 e-g) meaning that different substrate concentrations could not explain differences in effectiveness of the various HSPB members. The difference in aggregation suppressive function of the various HSPB members was also not due to differences in expression levels or higher stability of the HSPB proteins themselves; in fact, HSPB7 and HSPB9 expression levels were amongst the lowest of the whole HSPB family (Figure 1e,f). This means that their effectiveness may even be somewhat underestimated here. We next repeated the screen using another polyglutamine expanded protein, Ataxin-3 (EGFP- SCA3Q82) (Figure 2a). Also for this substrate, HSPB7 and HSPB9 were the most effective

64 HSPB7 is a novel suppressor of polyglutamine aggregation suppressors; HSPB1, HSPB4 and HSPB8 also showed some, but minor activity (Figure 2a). In accordance with data from the HD screen (Figure 1e,f,g), HSPB effectiveness against aggregation was not accompanied by a detectable reduction in soluble SCA3Q82 levels (Figure 2b).

HSPB7 activity is concentration dependent and active in the nucleus Transient overexpression can often lead to non-physiologically high levels of expression per cell. Therefore, we titrated the levels of HSPB7 by varying the ratio between HSPB7 and polyQ expressing plasmids and analyzed the effects on suppression of polyQ aggregation. Exclu- sively cytosolic inclusions were found when EGFP-HttQ74 was overexpressed (Figure 3a). This was markedly reduced when HSPB7 was co-expressed (Figure 3a). An increase in the HSPB7:EGFP-HttQ74 ratio (Figure 3b), was paralleled by a concurrent reduction in the extend of polyQ aggregation (Figure 3c). HSPB7 overexpression did not result in a reduction of soluble polyQ levels (Figure 3b), suggesting that rapid substrate degradation or translational inhibition are not the major pathways by which HSPB7 prevents aggregate formation. Next, we tested if HSPB7 could also prevent nuclear polyQ aggregation, a main feature of polyQ diseases related to cytotoxicity (4). Expression of a nuclear localization signal (NLS) containing EGFP-HttQ74 construct shows the formation of solely nuclear aggregates (Figure 3d). Clearly, co-expression of HSPB7 also prevented aggregate formation in the nucleus (Figure 3d) in a concentration dependent manner (Figure 3e,f). Again, no large effects on soluble polyQ levels were observed (Figure 3e).

HSPB7 does not target polyQ proteins for proteasomal degradation Classical chaperone activity towards misfolded and non-foldable proteins is generally found to intertwine with cellular degradation pathways. Heat shock proteins such as HSPA1A, DNAJB1 and DNAJB2 were shown to promote degradation of misfolded proteins through the ubiquitin- proteasome system (30;31) while e.g. the Drosophila HSP40 member Mrj (ortholog of mammalian DNAJB6/DNAJB8) and human HSPB8 were shown to induce autophagy (15;32). Although HSPB7 did not visibly reduce soluble HD or SCA3 levels (Figure 1f, 2b, 3b,e), part of the misfolded or aggregated substrate could specifically be targeted for degradation by either of these pathways. To test whether HSPB7 could promote proteasomal degradation we co-transfected cells with HSPB7 together with HA tagged HttQ43 and treated cells with the proteasome inhibitor MG-132. Under such conditions, HSPB7 was still active in preventing aggregate formation (Figure 4a), whereas the protective action of DNAJB1, known to prevent polyQ aggregation by stimulating proteasomal degradation (30), was clearly reduced (Figure 4a). In addition, we used the proteasome activity reporter Ub-R-EGFP (33); when co-expressed together with DNAJB1 the Ub-R-EGFP levels were strongly reduced, confirming the role of DNAJB1 in proteasomal degradation (Figure 4b). Co-expression of HSPB7, however, did not significantly affect the expression level of this proteasome activity reporter (Figure 4b). Co-expression of polyglutamine proteins (HA-HttQ43), that is sometimes suggested to negatively affect proteasomal activity (34), did not have an effect on proteasomal activity in our assays irrespective of chaperone co- expression (Figure 4b). Analysis of Ub-R-EGFP levels by fluorescent-activated cell sorting gave comparable results (Figure 4c). DNAJB1 enhanced proteasomal degradation while HSPB7 did not. These data suggest that HSPB7 does not reduce polyQ aggregation by enhancing proteasomal degradation.

HSPB7 is partly dependent on autophagic clearance of aggregated polyQ proteins Although HSPB7 was also effective in the nucleus of cells, which would not support that it acts via autophagy, we decided to directly test the possibility that HSPB7 could prevent polyQ ag-

65 Chapter 4 gregation by stimulating autophagosomal clearance of misfolded proteins at least in the cytosol. Hereto, we co-expressed EGFP-HttQ74 with either HSPB7 or BAG3 in mouse embryonic fibroblasts with or without a disruption of ATG5, a key regulator of autophagy (35). BAG3 was used as a positive control as it was shown that, in complex with HSPB8, it induces eIF2a phosphorylation and hereby reduced polyQ aggregation via protein synthesis inhibition and clearance of aggregates via autophagy (15;21). In cells positive for ATG5, polyQ aggregates are reduced when HSPB7 is co-expressed (Figure 5a) with no visible changes is soluble EGFP-HttQ74 levels (Figure 5b) unlike overexpression of BAG3 A that also reduced polyQ aggregation in these 24h mRFP ATG5 positive cells (Figure 5a), but with a con- comittant large reduction in soluble EGFP-Ht- V5-DNAJB1 tQ74 levels (Figure 5b). Cells lacking the ATG5 gene show more polyQ aggregation in control V5-HSPB7 conditions (Figure 5c), consistent with the role 48h of constitutive autophagy in clearance of polyQ mRFP aggregates (36;37). As demonstrated before, V5-DNAJB1 BAG3 overexpression in ATG5 -/- cells was less effective in reducing polyQ aggregation V5-HSPB7 than in ATG5 +/+ cells, although some activity 48h remained, which can be attributed to its effect mRFP MG-132 on protein synthesis leading to a large reduction V5-DNAJB1 of soluble EGFP-HttQ74 levels also in ATG5 -/- cells (Figure 5d) (16). In ATG5 -/- cells, HSPB7 V5-HSPB7 activity was largely reduced independent of an effect on soluble EGFP-HttQ74 levels (Figure B 5c, d). This suggests that, at least in the cytosolic compartment, HSPB7 co-operates with autophagic mechanisms to reduce aggregated mRFP mRFP V5-DNAJB1 V5-DNAJB1 V5-HSPB7 V5-HSPB7 HA-HttQ43 + + + polyQ material but in a manner that is distinct Ubr-EGFP + + + + + + from the BAG3/HSPB8 complex. -EGFP HSPB7 is not a classical chaperone -V5 The canonical mechanism by which small heat shock proteins are believed to function is that -GAPDH they form oligomeric complexes that, via dy- C Processing of Ubr-EGFP Figure 4. HSPB7 is not involved in proteasomal 600 degradation. (A) HEK293 cells were co-transfected with V5-tagged DNAJB1 or HSPB7 together with HA- 500

n a l HttQ43 at a 9:1 plasmid ratio. After 24 hour and 48 i g

s hours, samples were generated and analysed by the 400 filtertrap binding assay. In addition, cells were treated for 12 hours with the proteasome inhibitor MG132. E G F P 300

e (B) Co-transfection of V5-tagged DNAJB1 or HSPB7 i v 200 with the proteasomal activity reporter Ubr-EGFP, e l a t shows a reduction in Ubr-EGFP levels for DNAJB1 R 100 co-expression, while HSPB7 does not influence Ubr- EGFP clearance. Co-expression of HA-HttQ43 did 0 not influence Ubr-EGFP clearance. (C) Fluorescent 7 2

3 activated cell sorting measurements on Ubr-EGFP G 1

SP B fluorescence levels show that DNAJB1, but not NAJB1 Control M H

D HSPB7, reduces the Ubr-EGFP fluorescence.

66 HSPB7 is a novel suppressor of polyglutamine aggregation

A ATG5 +/+ EGFP-HttQ74 B

Control Control V5-HSPB1 V5-HSPB7 His-BAG3 V5-HSPB1 -EGFP V5-HSPB7 -GAPDH His-BAG3 -V5 0 20 40 60 80 100 Signal intensity C ATG5 -/- EGFP-HttQ74 D

Control

V5-HSPB1 Control V5-HSPB1 V5-HSPB7 His-BAG3 -EGFP V5-HSPB7 -GAPDH His-BAG3 -V5 0 20 40 60 80 100 Signal intensity Figure 5. HSPB7 efficacy is reduced in autophagy-deficient ATG5 -/- cells. (A) Wildtype MEF cells (ATG5 +/+) were co-transfected with V5-tagged HSPB1, HSPB7 or His-tagged BAG3 together with EGFP- HttQ74 (9:1 ratio). Cell lysates were analysed by filtertrap binding 48 hours after transfection. (B) Western- blot analysis shows that BAG3, but not HSPB7, reduces the soluble pool of EGFP-HttQ74. (C) V5-tagged HSPB1, HSPB7 or His-tagged BAG3 were co-transfected with EGFP-HttQ74 (9:1 ratio) in MEF cells lacking the ATG5 gene (ATG5 -/-). Both BAG3 activity and HSPB7 activity was reduced. (D) Westernblot analysis shows that BAG3, but not HSPB7, also reduced the soluble pool of EGFP-HttQ74 in ATG5 -/- cells. namic de- and re-oligomerization steps, bind unfolded substrates that are subsequently trans- ferred to HSPA protein machines for further processing (38;39). First, we tested whether HSPB7 forms oligomers in cells, like the classical small HSPs, such as HSPB1 and HSPB5 do. Hereto we expressed V5-HSPB5 or V5-HSPB7 in HEK293 cells and separated the cell lysates on sucrose gradients. While HSPB5 was found throughout the gradients, indicating formation of oligomeric species of various sizes, HSPB7 was mainly localized at low density indicative of its existence as a mono- or dimeric species (Figure 6a). Native gel analysis confirmed these findings: V5-HSPB1 clearly migrated consistent with oligomeric species like V5-HSPB5, while V5-HSPB7 migrated mostly as a single molecular species (Figure 6b). So, in living cells HSPB7 does not seem to form the canonical large HSPB oligomers. To test whether, despite its non-classical characteristic, HSPB7 still could act as a classical chaperone, we used heat-denatured firefly luciferase as a model substrate (40-42). Cells expressing luciferase with or without co-expressed HSPs were heated for 30 minutes at 45 °C to denature luciferase and (chaperone-assisted) refolding was measured by allowing an 1 hour recovery period at 37 °C. Overexpression of the classical V5 tagged chaperones HSPB1, HSPB4 and HSPB5 increased luciferase refolding compared to cells co-expressing mRFPruby (a non- chaperone control) (Figure 6c), consistent with previous data (39;43). Overexpression of the other HSPB members, including HSPB7, showed no comparable activities on luciferase refolding (Figure 6c). Overexpression of non-tagged HSPB1 and HSPB7 gave similar results compared to the V5-tagged versions (Figure 6d), demonstrating that V5-tagging did not negatively influence refolding activity. This suggests that HSPB7 does not act as a classical chaperone at least not when firefly luciferase is used as a substrate.

67 Chapter 4

A B HSPB1 HSPB7 D Luciferase refolding Figure 6. HSPB7 is not a canoni- 50 cal small heat shock protein. (A) HSPB5 HSPB7 non-tagged HEK293 cells were transfected with 40 either V5-tagged HSPB5 or HSPB7. Cell lysates were separated on a oligomers 30 sucrose-gradient. Whereas HSPB5 oligomers can be detected almost Luciferase

throughout the sucrose gradients, in- 20 % dicative a hetero-oligomeric complex formation, HSPB7 is found only the 10

Sucrose gradient less dense fractions (mainly mono- and dimers). (B) Native page gel 0 y 1 1 7 7 analysis of either V5-tagged HSPB1 A b 1 B B B B u A P P P P r or HSPB7 expressed in HEK293 cells. P P H S H S HSPB1 forms large oligomeric struc- R F H S tures while HSPB7 migrates as a sin- m V5- H S V5- H S gle molecular species. (C) Cells were C Luciferase refolding activity of the human HSPB family 50 co-transfected with firefly luciferase V5-tagged and the indicated chaperones (1:9 ra- classical HSPBs 40 tio). Cells were heat-shocked at 43 °C for 30 minutes and allowed to recover

30 for 1 hour at 37 °C. The percentage of reactivated luciferase, relative to

Luciferase the un-heated controls is depicted. 20 % Error bars indicate the SD of three independent experiments. (D). To rule 10 out that the V5-tag negatively affected HSPB1 or HSPB7 activity in relation to 0

y luciferase refolding, activities of both 1 2 3 4 5 6 7 8 9 0 A b 1 1 B B B B B B B B B u r B non-tagged and tagged HSPB1 and A P P P P P P P P P P P P HSPB7 were compared. H S H S H S H S H S H S H S H S H S R F H S H S m

HSPB7 is not dependent on a functional HSPA machine or activity of HSF-1 To explore whether the action of HSPB7 on poly-Q aggregation was dependent on collaboration with HSPA members, we initially aimed to down regulate HSPA8 and HSPA1, the endogenously expressed HSPA members in HEK293 cells (Hageman and Kampinga, unpublished observations). However, this turned out to be rather toxic to these cells. We therefore used an indirect approach to address this question and co-transfected the HEK293 cells with HSPB7 and EGFP- HDQ74 with or without BAG1. BAG1 is a nucleotide exchange factor that accelerates the AT- Pase cycle of HSPA and we previously showed that BAG1 overexpression leads to a negative effect on HSPA1A-mediated refolding (44) and hereby also interferes with the effect of HSPB1 on protein refolding (39). Co-expression of HSPA1A and BAG1 with either cytosolic or nuclear luciferase resulted in a decrease in refolding activity (Figure 7a), showing that HSPA mediated refolding is affected in both cellular compartments. However, BAG1 overexpression did not significantly affect HSPB7 activity on polyQ aggregation (Figure 7b), showing that HSPB7s major route of anti-aggregation does not require HSPA activity. To exclude that HSPB7 overexpression might induce a general stress response that upregulates other chaperones that in turn ensure clearance of poly-Q aggregates, we first measured levels of the stress inducible HSPA1A and DNAJB1. HSPB7 overexpression did not lead to a detectable induction of these classical heat- shock proteins (Figure 7c). In addition, we co-expressed EGFP-HttQ74 with either HSPB5 or HSPB7 in cells carrying a tetracycline inducible dominant negative HSF-1 (dnHSF-1) transgenic

68 HSPB7 is a novel suppressor of polyglutamine aggregation

HSPA1A luciferase refolding & BAG1 A 60 C 3h recovery 3h recovery +BAG1 g 50 i n mRFP mRFP V5-HSPB5 V5-HSPB5 V5-HSPB7 V5-HSPB7 l d o f 40 HA-Q43 - + - + - + e r -HSPA1A s e

a 30 r

e -DNAJB1 c i f

u 20 -GAPDH L

% 10 D 0 HS - + + - + + Cytosol Nucleus Rec - 8h 12h - 8h 12h HEK293 DNHSF-1 B EGFP-HttQ74 mRFPruby -HSPA6

V5-HSPB5 -GAPDH

V5-HSPB7

0 20 40 60 80 100 Signal intensity HEK293 DNHSF-1 EGFP-HttQ74 + BAG1 E mRFPruby mRFPruby

V5-HSPB5 V5-HSPB5

V5-HSPB7 V5-HSPB7 0 20 40 60 80 100 Signal intensity Figure 7. HSPB7 does not require a functional HSPA machine, induces no general stress response and is not dependent on HSF-1 responsive genes to reduce polyglutamine aggregation. (A) HSPA1A was cotransfected with BAG1, cytoplasmic luciferase or nuclear luciferase (4,5:4,5:1 ratio). Cells were heat shocked at 43oC for 30 minutes and allowed to recover after which luciferase activity was determined. In both compartments, co-expression of BAG1 inhibits the HSPA machine. (B) Co-expression of BAG1 with HSPB7 and EGFP-HttQ74 shows that BAG1 mediated inhibition of the HSP70 machine only marginally reduced the effect of HSPB7 on preventing polyQ aggregation. (C) HEK293 cells were transfected with either mRFPruby, V5-tagged HSPB5 or HSPB7 with (+) or without (-) HA-HttQ43 encoding plasmids. Levels of the stress-inducible HSPA1A and DNAJB1 were not altered under these conditions, indicating that none of them induced a general stress response. (D) The dominant negative HSF-1 cell line lacks HSF-1 inducible gene expression as can be seen by a failure of these cells to induce the expression of stricktly heat- inducible HSPA6 within 12 hours after a 30 minutes heat shock at 43 °C. (E) Using either HEK293 cells or the dominant negative HSF-1 celline, V5-tagged HSPB5 or HSPB7 were co-transfected with EGFP-HttQ74 (9:1 ratio). In both cell lines, HSPB7 was found equally effective in suppressing HttQ74 aggregation. construct which blocks HSF-1 dependent HSP expression. Expression of the dnHSF-1 construct effectively blocked the heat shock induced-expression of the HSF-1 inducible HSPA6 protein (Figure 7d). In these cells, that can no longer induce a stress response, HSPB7 retained its full activity to suppress poly-Q aggregation (Figure 7e). Thus, HSPB7 does not induce the general heat shock response and does not require induction of other heat shock proteins through HSF-1 in order to be functional.

HSPB7 sequence features in relation to anti-aggregation activity HSPB7 was originally discovered as a cardiovascular HSPB member (cvHSP) but although it is indeed highly expressed in cardiac tissue (45), database analyses revealed that HSPB7 is more ubiquitously expressed than originally thought. In fact, it was found to be one of the

69 Chapter 4

Figure 8. HSPB7 activity in pre- A serine-rich region B M S H R T S S T F R A E R S F H S S S S S S S S S T S S S A S R A L P A Q D P P M M S H R T S S A F R A E R S F R S S S S S S S S - S S S S A S R A L P A Q D P P M venting polyglutamine aggrega- M S A S N S S A Y R S E R S Y H Q T S S S S S S ------S S T S A N P Y M M N L K S S S A F R S E H T I K Q T S S S S S S ------S S S Q P R E P F M mRFP tion is evolutionary conserved

E K A L S M F S D D F G S F M R P H S E P L A F P A R P G G A G N I K T L G D A Y E K A L S M F S D D F G S F M L P H S E P L A F P A R P G G Q G N I K T L G D A Y and independent of two conserved E K S R G L F A D D F G S F M C P - K D A L G F P N R T G T V G N I K T L G D T Y Hs D K T Q G M F G D D F G K F R R P H G D P L G Y P G - - - A A D N V K S V G T T F motifs. (A) Sequence alignment of E F A V D V R D F S P E D I I V T T S N N H I E V R A E K L A A D G T V M N T F A Mm E F T V D M R D F S P E D I I V T T F N N H I E V R A E K L A A D G T V M N T F A Homo sapiens (Hs), Mus musculus Q F T V D V Q D F S P E D V I V T T S N N Q I E V H A E K L A S D G T V M N T F T Q F A V D V R D F S P E D I I V T T S N N Q I E V H A E K L G A D G T I M N T F T Dr (Mm), Danio rerio (Dr) and Xeno- H K C Q L P E D V D P T S V T S A L R E D G S L T I R A R R H - - P H T E H V Q Q H K C Q L P E D V D P T S V T S A L R E D G S L T I R A R R H - - P H T E H V Q Q H K C R L P E D V D P T S V K S S L G A D G T L T I K A Q R N - - T A K L E H A Q pus leavis (Xl) HSPB7. (B) Human H K C Q L P E D V D P T T V T S A L G E D G N L T V K A Q R N P A K H S E H L Q Q Xl HEK293 cells were co-transfected T F R T E I K I Hs T F R T E I K I Mm HttQ74 48h T F R T E I K I Dr with HSPB7 and EGFP-HttQ74 (9:1 T F R T E I K I Xl  -V5

C-box  -EGFP ratio) from the indicated species. Re-  -GAPDH Serine stretch 74 169 duction in polyglutamine aggregation C V5-HSPB7 N-terminus crystallin domain was analysed by the filtertrap binding assay. (C) Schematical overview of V5-S17-29-HSPB7 17-29 the HSPB7 topology and the dele- V5-N-HSPB7 1-73 tion mutants. (D) HEK293 cells were co-transfected with EGFP-HDQ74, V5-HSPB7-Cbox 162-170 wildtype HSPB7 or the deletion mutants (1:9 ratio). Deletion of the con- D served serine stretch or the C-box Control did not abolish effects of HSPB7 on V5-HSPB7 polyglutamine aggregation. Deletion

V5-S17-29-HSPB7 of the complete N-terminus did abro- gate anti-polyglutamine aggregation V5-N-HSPB7 activity. V5-HSPB7-Cbox

HttQ74 48h

0 20 40 60 80 100  -V5

Signal reduction  -eGFP  -GAPDH highest expressed HSPB members in human tissues at the messenger level which suggests a general function in cells (Chapter 3). Furthermore, HSPB7 homologs can be found throughout vertebrate species (46) (Figure 8a). To test whether HSPB7 from different species are functional homologs of the human HSPB7, we co-expressed EGFP-HttQ74 with either human, mouse, fish or frog HSPB7 and compared their efficacy on preventing polyQ aggregation in human HEK293 cells. All HSPB7 species variants were able to suppress polyQ aggregation (Figure 8b), strongly suggesting that they are true orthologs of human HSPB7 and that they have a general function that is maintained when expressed in cells derived from a different species. The main sequence-related feature that makes HSPB7 unique within the HSPB family is the presence of a conserved N-terminal serine-rich region of approximately eighteen residues. In addition, HSPB7 contains a conserved C-terminal region (C-box) of nine residues. Both these sequence features are also conserved in mouse, fish and frog orthologs (Figure 8a). In addition, we recently found that the N-terminus contains targeting motifs required for HSPB7 to be located to nuclear SC35 splicing speckles (Chapter 3). To investigate the role of these features on preventing polyQ aggregation, we co-expressed EGFP-HttQ74 with several human HSPB7 deletion mutants (Figure 8c). Deletion of the serine-rich region, previously shown to be redundant for SC35 speckle association, did not affect the activity of HSPB7 against polyQ aggregation (Figure 8d). Deletion of the N-terminal domain, which leads to displacement of HSPB7 from SC35 speckles, strongly reduced the activity of HSPB7 against polyQ aggregation (Figure 8d). Deletion of the nine C-terminal residues of HSPB7 did not affect the activity of HSPB7 against polyQ aggregation (Figure 8d). Thus, both the HSPB7-specific serine-rich region and the C-box are redundant features not required for preventing polyQ aggregation. A combination of these two deletions still resulted in an active HSPB7 mutant protein (data not shown). On the other hand, the role of the N-terminus seems indispensable in preventing polyQ aggregation. This

70 HSPB7 is a novel suppressor of polyglutamine aggregation could merely reflect the deletion of a substrate interaction domain, as implicated for other small HSP proteins (40;47;48), or that SC35 speckle association of the HSPB7 N-terminus has a role in preventing polyQ aggregation.

HSPB7 prevents polyQ aggregation in neuronal-like cells Although the precise mechanism remains to be elucidated, the above data show that HSPB7 is a novel and very potent, non canonical modulator of polyQ aggregation. To test whether it

A Tubulin SCA3-Q64 DAPI Overlay

B myc-HSPB7 SCA3-Q64 DAPI Overlay

Huntingtin Q74 SCA3 Q64 C 10 D 14

12 8 i o n s i o n s 10 l u s l u s 6

n c n c 8 I I

l e a r 4 l e a r

u c u c 4 N N 2

% % 2

0 0 l l 7 7 5 5 7 7 5 5 o o r r t t n n o o SP B SP B SP B SP B SP B SP B SP B SP B C C H H H H H H H H - - - - 5 5 5 5 V V V V

Figure 9. HSPB7 co-localizes with polyQ aggregates and reduces aggregate formation in neuronal cells. (A) NG108 cells were differentiated (see materials and methods) into neuronal cells during a seven day period. Cells were transfected after differentiation with EGFP tagged SCA3-Q64. After 48 hours cells were fixed with methanol and analysed by confocal microscopy. (B) After differentiation, cells wereco- transfected with EGFP-SCA3-Q64 and V5-HSPB7 encoding plasmids (1:9 ratio). In the low fraction of cells expressing HSPB7 that did show aggregates, V5-HSPB7 colocalized with these intranuclear SCA3 aggregates. The number of cells positive for either EGFP-HDQ74 (C) or SCA3-Q64 (D) inclusions in transfected differentiated NG108 cells overexpressing either HSPB7, V5-HSPB7, HSPB5 or V5-HSPB5 were counted using immunofluorescence. Clearly, HSPB7 strongly reduced inclusion formation. See Appendix 4 for colour print.

71 Chapter 4 is also effective in reducing polyQ aggregation and toxicity in neurons, we first co-transfected differentiated NG-108 neuroblastoma x glioma cells with HSPB7 and EGFP-HttQ74 or EGFP- SCA3Q63. Interestingly, overexpression of polyQ proteins (without NLS) in these cells led to an exclusive nuclear accumulation of polyQ aggregates, leaving the cytoplasm almost void of polyQ inclusions (Figure 9a), which is entirely different from what was found in HEK293 cells, unless we used NLS-mediated targeting. HSPB7 showed co-localization with the nuclear polyQ aggregates (Figure 9b), Since transfection efficiencies were very low, we could not use biochemical endpoints to test whether HSPB7 reduced the process of aggregation also in these NG108 cells. Therefore, we calculated the fraction of green, EGFP-HttQ74 or EGFP-SCA3Q63 expressing cells with visible inclusions (Figure 9c,d). Clearly, also in differentiated NG-108 cells

A C Cytotoxicity polyQ polyQ HSPB5 polyQ HSPB7 100 Without EGFP-Q119 expression -EGFP With EGFP-Q119 co-expression -HSPB5/7 80 P = 0.002 -GAPDH l -TET +TET 60 v i a r u S B

% 40

20 Control

0 HEK293 Q119 HSPB5 HSPB7 D HSPB7 transgene SCA3trQ78 SCA3trQ78 / HSPB7 1 2 3 polyQ

WT Actin-GAL4

Positive phototaxis

E t

h 60 l i g polyQ + HSPB5 50 o t

P = <0.001 d 40 e t c t

a 30 r t t

a 20 s

l i e 10 f

% 0

polyQ + HSPB7 Control SCA3trQ78 HSPB7 UAS-SCA4trQ78 x UAS-HsHSPB7 UAS-HsHSPB7 Figure 10. HSPB7 reduces polyQ induced toxicity in cells and in vivo. PolyQ induced cellular toxicity was analysed using a HEK293 cell line containing a tetracycline inducible EGFP-HttQ119 construct (53) that stably expressed either non-tagged HSPB5 or HSPB7. (A) Addition of tetracycline to induce the expression of EGFP-HttQ119 (panel A, upper lane) (B) Long term toxicity of polyQ expression was evaluated using the colony formation assay showing reduced colony formation (within two weeks) upon poly-Q expression. (C) Quantitative analysis of colony formation reveals that expression of EGFP-HttQ119 alone or together with HSPB5 results in a major reduction of colonies. Co-expression of HSPB7 results in a significant protection against polyQ toxicity. (D) The Drosophila SCA3 model (32) was used to evaluate the effect of HSPB7 expression on eye-degeneration. HSPB7 transgenic flies were generated and expression of HSPB7 alone caused no phenotypic changes and these flies had normal eye morphology (D1). Expression of HSPB7 was tested by crossing the Actin-GAL4 driver fly with the HSPB7 transgenic fly (D1, bottom right insert). Expres- sion of SCA3trQ78 resulted in a visible degeneration (D2). A mild reduction upon co-expression of HSPB7 on SCA3trQ78 induced eye degeneration was observed, but this effect was not statistically significant (data not shown & D3). (E) By analysing positive phototaxis, a significant reduction in blindness could be observed when HSPB7 was co-expressed with SCA3trQ78. The level of statistical significance was determined using Chi-square analysis.

72 HSPB7 is a novel suppressor of polyglutamine aggregation

HSPB7 (both V5- and non-tagged) effectively reduced polyQ aggregation, while overexpression of HSPB5 had only a minor effect.

HSPB7 reduces polyQ related toxicity We next analyzed if HSPB7, in addition to reducing polyQ aggregation, could also reduce polyQ-induced cytotoxicity. Hereto, we used a HEK293 cell line containing a tetracycline inducible EGFP-HttQ119 construct. These cells were next used to generate stable lines constitutively expressing either non-tagged HSPB5 or non-tagged HSPB7 (Figure 10a). Expression of EGFP- HttQ119 was subsequently induced by addition of 1 µg/ml tetracycline (Figure 10a) and long term toxicity was evaluated using colony formation (i.e the ability of individual cells expressing polyQ proteins to form colonies containing at least 50 cells within 2 weeks). Expression of EGFP-HttQ119 severely affected colony-forming ability, indicating polyQ-induced cytotoxicity (Figure 10b,c). While co-expression of HSPB5 did not significantly protect against cytotoxicity, co-expression of HSPB7 led to a significant reduction in polyQ mediated cytotoxicity (Figure 10b,c). HSPB7 expression alone did not affect cell growth or colony formation (data not shown). Thus, HSPB7 is not only able to both reduce polyQ aggregate formation in cells but also reduces polyQ induced cytotoxicity. To test if HSPB7 can also prevent polyQ-induced neuronal toxicity in vivo, we targeted HSPB7 overexpression in the compound eyes of the Drosophila SCA3 model (32). This fly model is characterized by continuous expression of a truncated polyQ-expanded SCA3-gene (SCA3- trQ78) in the compound eyes, resulting in neurodegeneration and blindness. Overexpression of polyQ expanded SCA3 resulted in visible eye degeneration (Figure 10d) and loss of positive phototaxis, a measure for blindness (Figure 10e). HSPB7 transgenic flies showed normal eye-morphology and phototaxis (Figure 10d,e). Although expression of HSPB7 in the SCA3 transgenic flies did not result in statistically significant amelioration of the SCA-3 induced eye- degeneration (data not shown), a significant improvement in positive phototaxis was seen (Fig- ure 10d,e). So, HSPB7 is not only functional in preventing polyQ aggregation in cells in vitro, but is also able to reduce polyQ-induced toxicity in vivo

Discussion

In this work, we have compared the various members for their ability to prevent polyQ aggregation and identified, HSPB7 as the most active member. Its mode of action was demonstrated to be of a non-canonical nature: although HSPB7 was in part dependent on autophagy, it functions independently of proteasomal degradation. Furthermore, HSPB7 did not form high-molecular weight complexes like HSPB1 and HSPB5, which are implicated as being important for their chaperone activity. In agreement with this, HSPB7 was unable to chaperone heat-denatured luciferase, a well known substrate for both HSPB1 and HSPB5. Although the precise mechanisms remains to be elucidated, HSPB7-like activities are evolutionary conserved and also prevent aggregate formation in neuronal-like cells and even reduce cytotoxicity and neuronal toxicity in cells and in vivo respectively, making it an excellent candidate for future therapeutic strategies.

HSPB members show diversity in cellular activity By screening the HSPB family for activity against amyloid fibril formation of polyQ proteins, a clear difference in their respective capacities to handle this substrate was found. Whereas the classical, well-studied HSPB1 and HSPB5 assist in refolding of heat denatured luciferase, and show substantial activity against aggregation of diverse other substrates (27;49;50), they were merely ineffective in preventing polyQ aggregation (12) (this study). Strikingly, HSPB9 and especially HSPB7 were most effective in reducing polyQ aggregation, but did not sup-

73 Chapter 4 port refolding of heat unfolded luciferase. This set of data suggests at least two points. Firstly, diverse HSPB members can act on different substrates consistent with a rather promiscuous substrate recognition ability. Secondly, the effectiveness of the various members to handle (the same) substrates differs. This could mean that binding of a certain substrate to one HSPB is more efficient for its further processing (refolding, proteasomal degradation, autophagosomal routing etc.) than binding to another HSPB member. The latter may occur via additional interaction partners of the diverse HSPB members and or via differential substrate binding affinity e.g. impaired hand-over to HSPA members. Whereas relatively lower substrate affinity (e.g. HSPB1) could favor transfer to HSPA members and subsequent substrate folding (luciferase), tentative stronger substrate affinities (or other non-HSPA dependent mechanisms: e.g. HSPB7) may avoid the danger of transfer to HSPA members for non-refoldable substrates like polyQ that are poorly handled by the proteasome. Whereas such a model requires further biochemical evidence, it is supported by several other observations. Those HSPB members that function in stimulating luciferase refolding depend on a functional HSPA machine (39;this study) and be- have like classical dynamic oligomers as has been described in cell free experiments (23;51). Such members are usually poor suppressors of polyQ aggregation (12;this study). Inversely, the effect of HSPB7 on polyQ aggregation does not require a functional HSPA machine (this study) and a mere independence of the HSPA machine in preventing polyQ aggregation was recently also found for strong suppressors in the family of DNAJ proteins (9;52) that also do not support protein refolding reactions (53). Clearly HSPB7 is not the only non-canonical member of the HSPB family, Also HSPB8 does not form the classical oligomeric complexes in cells but rather forms a stoichiometric complex with BAG3 (2:1). This complex induces autophagy (15) and inhibits protein synthesis in a HSP70- independent but eIF2α-dependent manner (16). Hereby it also can lead to reduced polyQ aggregates, although this mechanism seems less effective in handling long polyQ expansions (this study). Although HSPB7 in part also depends on active autophagy, its mechanism of action seems distinct from the BAG3/HSPB8 complex as its activity does not seems to affect soluble protein levels, meaning that unlike the BAG3/HSPB8 it does not act on protein synthesis.

Oligomerization of HSPB members and anti-aggregation One striking distinction between the non-canonical polyQ chaperoning HSPB proteins (HSPB6, HSPB7, HSPB8) and the classical HSPB proteins (HSPB1, HSPB5), seems to be the finding that they do not exist as large oligomers within living cells. While active chaperone activity of HSPB1 towards unfolded proteins is dependent on oligomeric dynamics(51;54), HSPB7 was found to exist as either mono- or dimeric sizes. Also HSPB6 and HSPB8, active in reducing HA-HttQ43 aggregation, are reported to mainly exist as dimers or small oligomers respectively (55;56). Why would HSPB7, that can reduce amyloid fibril formation of polyQ containing proteins, not also support refolding of heat-denatured luciferase, which forms amorphous aggregates. Success- ful chaperoning of this specific substrate rather seems to require shuttling between dimers and large oligomers (39). As we were unable to detect clear degradation-related effect on reducing polyQ aggregation by HSPB7 and in addition a lack of requirement of HSPA and other stress- induced chaperones, we propose that HSPB7 might function by transiently shielding the polyQ stretch and effectively preventing polyQ-oligomer formation and halting the process of inclusion formation. These chaperone-polyQ oligomers next may be (slowly) cleared by autophagy, which could explain why HSPB7 is less efficient in ATG5 -/- cell defective in macroautophagy In this way, seed formation for further aggregation is reduced and other cellular degradation pathways are allowed a larger time-window to clear the aggregation-prone substrate without the requirement of boosting either proteasomal or autophagic clearance. A similar mode of action has been suggested for HSPB5 in preventing nucleation of amyloid fibril forming substrate in vitro (57;58),

74 HSPB7 is a novel suppressor of polyglutamine aggregation a process leading to aggregate formation. Yet, HSPB5 seems unable to act in reducing polyQ aggregation, again suggesting specificity in substrate chaperoning. An alternative hypothesis could be that HSPB7 somehow has indirect effects on poly-Q aggregation via an effect on RNA metabolism. In chapter 3 we showed that HSPB7 is associated with SC35 speckles that play a role in RNA splicing and processing. By acting as a possible RNA chaperone, HSPB7 may effectively decrease aggregation of triplet expanded RNA species hereby maintaining cellular homeostasis and the ability of reducing the burden of unfolded peptides via constitutively active protein quality control pathways. HSPB7 is able to reduce polyQ aggregation and toxicity in several models used here, which are all associates with an overwhelming induction of polyQ expression in a very short time. As negative effects of expanded polyQ protein expression in patients accumulates over many years, HSPB7 activity in vitro might translate in such situations in a highly effective strategy to handle polyQ fibril formation. Taken together, HSPB proteins in general show a diverse and often specific chaperone activity towards proteins tending to form either amorphous or amyloidal aggregates, with sometimes diverse outcomes on cellular survival. The comparative study presented here forms a beginning in our understanding on substrate-specificity of HSPB proteins and differences in their mode of action. Further comparative studies using different substrate proteins are embraced to clarify the role of substrate in relation to HSPB chaperone activity.

Acknowledgements

This work was supported by Innovatiegerichte Onderzoeksprogramma Genomics. Grant IGE03018.

References

1. Brinkman, R. R., Mezei, M. M., Theilmann, J., Almqvist, E. and others (1997) The likelihood of being affected with Huntington disease by a particular age, for a specific CAG size, Am. J. Hum. Genet. 60, 1202-1210. 2. Vos, M. J., Hageman, J., Carra, S., and Kampinga, H. H. (2008) Structural and functional diversities between members of the human HSPB, HSPH, HSPA, and DNAJ chaperone families, Biochemistry 47, 7001-7011. 3. Muchowski, P. J., Schaffar, G., Sittler, A., Wanker, E. E. and others (2000) Hsp70 and hsp40 chaperones can inhibit self-assembly of polyglutamine proteins into amyloid-like fibrils, Proc. Natl. Acad. Sci. U. S. A 97, 7841-7846. 4. Rujano, M. A. and Kampinga, H. H. (2008) The HSP70 chaperone machine as guardian of the proteome: Implica- tions for protein folding diseases, in Heat Shock Proteins in Biology and Medicine (Radons, J. and Multhoff, G., Eds.) Research Signpost. 5. Warrick, J. M., Chan, H. Y., Gray-Board, Chai, Y. and others (1999) Suppression of polyglutamine-mediated neurodegeneration in Drosophila by the molecular chaperone HSP70, Nat. Genet. 23, 425-428. 6. Wyttenbach, A., Carmichael, J., Swartz, J., Furlong, R. A. and others (2000) Effects of heat shock, heat shock protein 40 (HDJ-2), and proteasome inhibition on protein aggregation in cellular models of Huntington’s disease, Proc. Natl. Acad. Sci. U. S. A 97, 2898-2903. 7. Ishihara, K., Yamagishi, N., Saito, Y., Adachi, H. and others (2003) Hsp105alpha suppresses the aggregation of truncated androgen receptor with expanded CAG repeats and cell toxicity, J. Biol. Chem. 278, 25143-25150. 8. Yamashita, H., Kawamata, J., Okawa, K., Kanki, R. and others (2007) Heat-shock protein 105 interacts with and suppresses aggregation of mutant Cu/Zn superoxide dismutase: clues to a possible strategy for treating ALS, J. Neurochem. 102, 1497-1505. 9. Fayazi, Z., Ghosh, S., Marion, S., Bao, X. and others (2006) A Drosophila ortholog of the human MRJ modulates polyglutamine toxicity and aggregation, Neurobiol. Dis. 24, 226-244. 10. Rujano, M. A., Kampinga, H. H., and Salomons, F. A. (2007) Modulation of polyglutamine inclusion formation by the Hsp70 chaperone machine, Exp. Cell Res. 313, 3568-3578. 11. Patel, Y. J., Payne, S., de, B. J., and Latchman, D. S. (2005) Hsp27 and Hsp70 administered in combination have a potent protective effect against FALS-associated SOD1-mutant-induced cell death in mammalian neuronal cells, Brain Res. Mol. Brain Res. 134, 256-274. 12. Carra, S., Sivilotti, M., Chavez Zobel, A. T., Lambert, H. and others (2005) HspB8, a small heat shock protein mutated in human neuromuscular disorders, has in vivo chaperone activity in cultured cells, Hum. Mol. Genet. 14, 1659-1669.

75 Chapter 4

13. Wyttenbach, A., Sauvageot, O., Carmichael, J., az-Latoud, C. and others (2002) Heat shock protein 27 prevents cellular polyglutamine toxicity and suppresses the increase of reactive oxygen species caused by huntingtin, Hum. Mol. Genet. 11, 1137-1151. 14. Zourlidou, A., Gidalevitz, T., Kristiansen, M., Landles, C. and others (2007) Hsp27 overexpression in the R6/2 mouse model of Huntington’s disease: chronic neurodegeneration does not induce Hsp27 activation, Hum. Mol. Genet. 16, 1078-1090. 15. Carra, S., Seguin, S. J., and Landry, J. (2008) HspB8 and Bag3: a new chaperone complex targeting misfolded proteins to macroautophagy, Autophagy. 4, 237-239. 16. Carra, S., Brunsting, J. F., Lambert, H., Landry, J. and others (2008) HSPB8 participates in protein quality control by a non chaperone-like mechanism that requires eIF2alpha phosphorylation, J. Biol. Chem. 17. Rogalla, T., Ehrnsperger, M., Preville, X., Kotlyarov, A. and others (1999) Regulation of Hsp27 oligomerization, chaperone function, and protective activity against oxidative stress/tumor necrosis factor alpha by phosphorylation, J. Biol. Chem. 274, 18947-18956. 18. Horwitz, J. (1992) Alpha-crystallin can function as a molecular chaperone, Proc. Natl. Acad. Sci. U. S. A 89, 10449- 10453. 19. Wilhelmus, M. M., Boelens, W. C., Otte-Holler, I., Kamps, B. and others (2006) Small heat shock protein HspB8: its distribution in Alzheimer’s disease brains and its inhibition of amyloid-beta protein aggregation and cerebrovascular amyloid-beta toxicity, Acta Neuropathol. (Berl) 111, 139-149. 20. Wang, X., Klevitsky, R., Huang, W., Glasford, J. and others (2003) AlphaB-crystallin modulates protein aggregation of abnormal desmin, Circ. Res. 93, 998-1005. 21. Carra, S., Seguin, S. J., Lambert, H., and Landry, J. (2008) HspB8 chaperone activity toward poly(Q)-containing proteins depends on its association with Bag3, a stimulator of macroautophagy, J. Biol. Chem. 283, 1437-1444. 22. Lee, G. J., Roseman, A. M., Saibil, H. R., and Vierling, E. (1997) A small heat shock protein stably binds heat- denatured model substrates and can maintain a substrate in a folding-competent state, EMBO J. 16, 659-671. 23. Leroux, M. R., Melki, R., Gordon, B., Batelier, G. and others (1997) Structure-function studies on small heat shock protein oligomeric assembly and interaction with unfolded polypeptides, J. Biol. Chem. 272, 24646-24656. 24. Morrow, G., Heikkila, J. J., and Tanguay, R. M. (2006) Differences in the chaperone-like activities of the four main small heat shock proteins of Drosophila melanogaster, Cell Stress. Chaperones. 11, 51-60. 25. Mogk, A., Schlieker, C., Friedrich, K. L., Schonfeld, H. J. and others (2003) Refolding of substrates bound to small Hsps relies on a disaggregation reaction mediated most efficiently by ClpB/DnaK, J. Biol. Chem. 278, 31033- 31042. 26. Schneider, G. B., Hamano, H., and Cooper, L. F. (1998) In vivo evaluation of hsp27 as an inhibitor of actin polymeri- zation: hsp27 limits actin stress fiber and focal adhesion formation after heat shock, J. Cell Physiol 177, 575-584. 27. Pivovarova, A. V., Chebotareva, N. A., Chernik, I. S., Gusev, N. B. and others (2007) Small heat shock protein Hsp27 prevents heat-induced aggregation of F-actin by forming soluble complexes with denatured actin, FEBS J. 274, 5937-5948. 28. Djabali, K., Piron, G., de Nechaud, B., and Portier, M. M. (1999) alphaB-crystallin interacts with cytoplasmic intermediate filament bundles during mitosis, Exp. Cell Res. 253, 649-662. 29. Scherzinger, E., Lurz, R., Turmaine, M., Mangiarini, L. and others (1997) Huntingtin-encoded polyglutamine expansions form amyloid-like protein aggregates in vitro and in vivo, Cell 90, 549-558. 30. Bailey, C. K., Andriola, I. F., Kampinga, H. H., and Merry, D. E. (2002) Molecular chaperones enhance the degradation of expanded polyglutamine repeat androgen receptor in a cellular model of spinal and bulbar muscular atrophy, Hum. Mol. Genet. 11, 515-523. 31. Howarth, J. L., Kelly, S., Keasey, M. P., Glover, C. P. and others (2007) Hsp40 molecules that target to the ubiquitin- proteasome system decrease inclusion formation in models of polyglutamine disease, Mol. Ther. 15, 1100-1105. 32. Bilen, J. and Bonini, N. M. (2007) Genome-wide screen for modifiers of ataxin-3 neurodegeneration in Drosophila, PLoS. Genet. 3, 1950-1964. 33. Dantuma, N. P., Lindsten, K., Glas, R., Jellne, M. and others (2000) Short-lived green fluorescent proteins for quan- tifying ubiquitin/proteasome-dependent proteolysis in living cells, Nat. Biotechnol. 18, 538-543. 34. Venkatraman, P., Wetzel, R., Tanaka, M., Nukina, N. and others (2004) Eukaryotic proteasomes cannot digest polyglutamine sequences and release them during degradation of polyglutamine-containing proteins, Mol. Cell 14, 95-104. 35. Hara, T., Nakamura, K., Matsui, M., Yamamoto, A. and others (2006) Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice, Nature 441, 885-889. 36. Khan, L. A., Yamanaka, T., and Nukina, N. (2008) Genetic impairment of autophagy intensifies expanded polyglutamine toxicity in Caenorhabditis elegans, Biochem. Biophys. Res. Commun. 368, 729-735. 37. Jia, K., Hart, A. C., and Levine, B. (2007) Autophagy genes protect against disease caused by polyglutamine expansion proteins in Caenorhabditis elegans, Autophagy. 3, 21-25. 38. Cashikar, A. G., Duennwald, M., and Lindquist, S. L. (2005) A Chaperone Pathway in Protein Disaggregation: HSP26 alters the nature of protein aggregates to facilitate reactivation by HSP104, J. Biol. Chem. 280, 23869- 23875. 39. Bryantsev, A. L., Kurchashova, S. Y., Golyshev, S. A., Polyakov, V. Y. and others (2007) Regulation of stress-induced intracellular sorting and chaperone function of Hsp27 (HspB1) in mammalian cells, Biochem. J. 407, 407-417. 40. Basha, E., Lee, G. J., Demeler, B., and Vierling, E. (2004) Chaperone activity of cytosolic small heat shock proteins

76 HSPB7 is a novel suppressor of polyglutamine aggregation

from wheat, Eur. J. Biochem. 271, 1426-1436. 41. Michels, A. A., Nguyen, V. T., Konings, A. W., Kampinga, H. H. and others (1995) Thermostability of a nuclear- targeted luciferase expressed in mammalian cells. Destabilizing influence of the intranuclear microenvironment, Eur. J. Biochem. 234, 382-389. 42. Singer, M. A. and Lindquist, S. (1998) Multiple effects of trehalose on protein folding in vitro and in vivo, Mol. Cell 1, 639-648. 43. Michels, A. A., Kanon, B., Konings, A. W., Ohtsuka, K. and others (1997) Hsp70 and Hsp40 chaperone activities in the cytoplasm and the nucleus of mammalian cells, J. Biol. Chem. 272, 33283-33289. 44. Nollen, E. A., Brunsting, J. F., Song, J., Kampinga, H. H. and others (2000) Bag1 functions in vivo as a negative regulator of Hsp70 chaperone activity, Mol. Cell Biol. 20, 1083-1088. 45. Krief, S., Faivre, J. F., Robert, P., Le, D. B. and others (1999) Identification and characterization of cvHsp. A novel human small stress protein selectively expressed in cardiovascular and insulin-sensitive tissues, J. Biol. Chem. 274, 36592-36600. 46. Franck, E., Madsen, O., van, R. T., Ricard, G. and others (2004) Evolutionary diversity of vertebrate small heat shock proteins, J. Mol. Evol. 59, 792-805. 47. Basha, E., Friedrich, K. L., and Vierling, E. (2006) The N-terminal arm of small heat shock proteins is important for both chaperone activity and substrate specificity, J. Biol. Chem. 281, 39943-39952. 48. Stromer, T., Fischer, E., Richter, K., Haslbeck, M. and others (2004) Analysis of the regulation of the molecular chaperone Hsp26 by temperature-induced dissociation: the N-terminal domail is important for oligomer assembly and the binding of unfolding proteins, J. Biol. Chem. 279, 11222-11228. 49. Ecroyd, H., Meehan, S., Horwitz, J., Aquilina, J. A. and others (2007) Mimicking phosphorylation of alphaB-crystallin affects its chaperone activity, Biochem. J. 401, 129-141. 50. Ghosh, J. G., Houck, S. A., and Clark, J. I. (2007) Interactive domains in the molecular chaperone human alphaB crystallin modulate microtubule assembly and disassembly, PLoS. ONE. 2, e498. 51. Ehrnsperger, M., Lilie, H., Gaestel, M., and Buchner, J. (1999) The dynamics of Hsp25 quaternary structure. Struc- ture and function of different oligomeric species, J. Biol. Chem. 274, 14867-14874. 52. Chuang, J. Z., Zhou, H., Zhu, M., Li, S. H. and others (2002) Characterization of a brain-enriched chaperone, MRJ, that inhibits Huntingtin aggregation and toxicity independently, J. Biol. Chem. 277, 19831-19838. 53. Hageman, J. (2008) The human HSP70/HSP40 chaperone family: a study on its capacity to combat proteotoxic stress, Thesis. 54. Shashidharamurthy, R., Koteiche, H. A., Dong, J., and McHaourab, H. S. (2005) Mechanism of chaperone function in small heat shock proteins: dissociation of the HSP27 oligomer is required for recognition and binding of destabilized T4 lysozyme, J. Biol. Chem. 280, 5281-5289. 55. Bukach, O. V., Seit-Nebi, A. S., Marston, S. B., and Gusev, N. B. (2004) Some properties of human small heat shock protein Hsp20 (HspB6), Eur. J. Biochem. 271, 291-302. 56. Kasakov, A. S., Bukach, O. V., Seit-Nebi, A. S., Marston, S. B. and others (2007) Effect of mutations in the beta5- beta7 loop on the structure and properties of human small heat shock protein HSP22 (HspB8, H11), FEBS J. 274, 5628-5642. 57. Hatters, D. M., Lindner, R. A., Carver, J. A., and Howlett, G. J. (2001) The molecular chaperone, alpha-crystallin, inhibits amyloid formation by apolipoprotein C-II, J. Biol. Chem. 276, 33755-33761. 58. Rekas, A., Jankova, L., Thorn, D. C., Cappai, R. and others (2007) Monitoring the prevention of amyloid fibril formation by alpha-crystallin. Temperature dependence and the nature of the aggregating species, FEBS J. 274, 6290- 6304. 59. Hageman, J., Vos, M. J., van Waarde, M. A., and Kampinga, H. H. (2007) Comparison of intra-organellar chaperone capacity for dealing with stress-induced protein unfolding, J. Biol. Chem. 282, 34334-34345.

77 Chapter 4

78 CHAPTER 5

Longevity as evoked by different chaperone activities of small heat shock proteins

Michel J Vos, Bart Kanon, Floris Bosveld, Karin Klauke, Serena Carra, Ody C M Sibon & Harm H Kampinga

Department of Cell Biology, Section of Radiation and Stress Cell Biology, University Medical Center Groningen, Groningen, The Netherlands

In revision for Aging Cell Chapter 5

Abstract

During aging, oxidized, misfolded and aggregated proteins accumulate in cells, whilst the capacity to deal with protein damage declines severely. To cope with the toxicity of damaged proteins, cells rely on protein quality control. In particular, heat shock proteins (HSP) have been proven to be successful in restoring cellular protein homeostasis and in boosting longevity of organisms. Cells contain many HSPs with overlapping and non-overlapping activities, but it remains unclear which functions are required to support healthy aging. In the current study, we compared the functions of the members of the Drosophila small HSP family for their ability to assist in refolding stress denatured substrates and/or prevent aggregation of disease associated, misfolded proteins. Whereas some small HSP members chaperone both substrates, we identified one member (CG14209) that exclusively assisted in HSP70-dependent refolding of stress denatured proteins. On the other hand, one member (HSP67BC) did not support protein refolding but only prevented toxic aggregation of misfolded proteins in an HSP70 independent manner. Intrigu- ingly, both single activities supported longevity, supporting that protein homeostasis in general rather than specific chaperone activities can decelerate the negative effects of aging.

80 Small heat shock proteins & longevity

Introduction

Imbalances in protein homeostasis has been suggested to be a crucial factor in the development of heritable age-related neurodegenerative diseases and in normal aging (1-6). Achieving and maintaining the correct three-dimensional protein structure is a continuous struggle within cells. Firstly, folding of proteins towards an active biological state is challenged by the crowded environment within the cell which may lead to off-pathway reactions resulting in protein aggregation (7-9). Protein misfolding can further originate from direct protein damage (e.g oxidation, thermal denaturation) but can also originate from age-related mutations in the DNA, molecular misreading (10), splicing errors (11) or errors in translation (12;13). While cells are challenged by an accumulation of oxidized, misfolded and aggregation-prone proteins, at the same time the capacity of a cell to deal with accumulated protein damage declines with aging (14-19). As molecular chaperones, heat shock proteins (HSPs) play a central role in protein homeostasis: they safeguard protein conformation and folding, the assembly and disassembly of protein complexes and protein translocation. By their ability to bind non-native polypeptides, they maintain their clients in a state competent for subsequent folding or, when folding is not successful, for degradation by the ubiquitin-proteasome system (20) or through autophagy (21). Hereby chaperones can prevent toxic protein aggregation and as such they have been implicated as protectors against age-related protein folding diseases (22) and as supporters of healthy aging (2;23-26). Indeed, global activation of all stress-inducible HSPs, either by overexpression of the heat shock factor-1 (HSF-1) (2;25) or via caloric restriction and the accompanying insulin signaling (24) was shown to delay the onset of protein folding diseases and to induce longevity. Interestingly, overexpression of individual members of especially the small heat shock protein family was shown to support longevity both in Caenorhabditis elegans (26) and in Drosophila melanogaster (D. melanogaster) (23;27-29). All these small HSPs tested so far (HSP22, HSP23, HSP26, HSP27) share the capacity to facilitate refolding of stress denatured substrates in vitro (30), again suggesting that supporting protein homeostasis is essential for longevity. Remarkably, however, overexpression of a member of the major HSP70 chaperone family, D. melanogaster HSP70, did not result in longevity and transgenic flies even exhibited lower lifespan and higher mortality rates (31). The HSP70 family comprises a highly homologues group of (HSF-1) inducible proteins and play a rather promiscuous role in protein folding (32;33). Members of the small HSPs are much more heterogeneous (34-37) and were found to have partially overlapping but also non-overlapping functions in protein quality control (37-39). To further elucidate the role of small HSPs in aging, we cloned 10 members of the Drosophila small HSP family and compared their ability to assist in refolding of stress denatured substrates and/or in preventing aggregation of disease associated, misfolded proteins in living cells. Whereas some small HSP members chaperone both substrates, single members were identified that either exclusively assisted in HSP70-dependent refolding of stress denatured proteins or were only able to prevent toxic aggregation of misfolded proteins in an HSP70 independent manner. Therefore our combined cellular and in vivo studies unambiguously demonstrate that longevity can be achieved through separate chaperone activities.

Experimental procedures

Organisms and growth conditions For cloning purposes and plasmid isolation, Escherichia coli DH5α was used and grown at 37 °C in LB medium (1% bacto-tryptone, 0.5% yeast extract, 0.5% NaCl), supplemented with the appropriate antibiotics when required.

81 Chapter 5

Drosophila Schneider’s S2 cells, were cultured in Schneider’s Drosophila medium (GIBCO) supplemented with 10% heat-inactivated fetal bovine serum (Greiner), 100 units/ml penicillin and 100 g/ml streptomycin in T25 flasks at 25 0C. For exponential cell growth, cell density was kept between 3x105 and 3x106 cells/ml. Fly stocks (Table S1) were maintained at 22 0C according to standard protocols. GAL4 driver stocks were obtained from the Bloomington Stock Centre (Indiana University, USA) (Table S1). The GMR-UAS-SCA3Q78 fly used for the eye-degeneration screen was generously provided by N. Bonini (University of Pennsylvania, USA) and maintained at 25 0C. Transgenic lines were generated by Genetic Services Inc. by injection of the pUAS vector into the W1118 genetic background.

Molecular techniques, bioinformatics, and plasmid generation Oligonucleotide primers (Biolegio) and plasmids used in this study are listed in Tables S2-4. Standard recombinant DNA techniques were carried out essentially as described by Sambrook et al. (81) Restriction enzymes were used according to the manufacturers instructions (Invitro- gen, New England Biolabs). Vent DNA polymerase (New England Biolabs) was used for preparative polymerase chain reactions (PCR). DNA sequencing reactions were carried out by ServiceXS. Blast algorithms (82) were used to screen databases at the National Center for Biotechnology Information (NCBI). Drosophila small heat shock proteins sequences were retrieved from the NCBI database using the D. melanogaster HSP27 sequence as input for a protein BLAST search. Results were analysed by ClustalX (83) and visualized by Treeview (84). A total of eleven small HSPs was found to be present in the D. melanogaster genome most of which are located at position 67B on the third chromosome (Figure S1a). Nearest neighbour analysis shows three main groups within the D. melanogaster small HSP family (Figure S1b). Group A represents the most studied classical small HSPs: HSP22, HSP23, HSP26, and HSP27 (23;29;30;85-89). Group B consists of L(2) EFL, CG4461, CG7409 and CG14207, while group C contains HSP67BC and CG13133. The Drosophila HSP plasmid library (Figure 1c) was generated using either cDNA originating from heat-shocked flies or from the Gold cDNA library (Bloomington). Primers used for isolation and amplification of individual small HSPs using are listed in Table 2. All PCR products were cloned into the pAc5.1-V5 plasmid and sequence verified. The pAc5.1-V5 was generated by annealing two oligonucleotides forming a KpnI overhang at the 5’ end (Table S2). This fragment was ligated into KpnI-EcoRV digested pAc5.1. The L4440 RNAi feeding vector (Fire laboratory), containing two opposing T7 sequences, was modified for T/A-cloning as follows. To remove unneeded nucleotides between both T7 sequences, L4440 was digested with BglII and XhoI. Annealed oligonucleotides (Table S2), designed to provide BglII and XhoI overhangs and two internal XcmI sites, were ligated into digested L4440 leading to L4440-T/A. Upon digestion with XcmI, this plasmid contains two T-overhangs, allowing annealing of Taq DNA polymerase amplified DNA fragments (A-overhangs). To generate the dsRNA-template library, a specific part of the HSP70 and HSC70 genes was amplified using primers listed in Table S5 using Taq polymerase. Subsequently, the A-tailed PCR products were ligated into the XcmI digested L4440-T/A plasmid.

DNA transfection Schneider’s S2 cells were either transfected using the CaCl2 method or Effectene (Qiagen). Transfection using the CaCl2 was performed as follows. 2.5 μg of plasmid DNA were mixed with 10 μl 2.5 M CaCl2, the volume was adjusted to 100 μl with 0.1× TE (pH 7.6) and this was added μ drop-wise to 100 l 2× HEPES buffer (280 mM NaCl, 1.5 mM Na2HPO4•2H2O, 50 mM HEPES, pH 7.05) while vortexing. Precipitates were allowed to form for 30 min. and then the solution was

82 Small heat shock proteins & longevity added to 1×106 cells in a 35-mm dish. Transfection using Effectene was performed according to the manufacturers instructions using 0,6 μg of plasmid DNA in combination with 5μL Effectene. HEK293 cells were transfected using Lipofectamine (Invitrogen) according to the manufacturers instructions using 1ug of plasmid DNA per 35-mm dish.

RNA Interference RNA interference was performed as described previously (90). In short, DNA fragments of variable length coding for specific parts of the target genes (Table S5) were amplified using Taq DNA polymerase. These fragments were cloned into XcmI digested L4440-T/A vector. Using T7-specific primers, the T7 flanked target sequence was re-amplified. Subsequently, dsRNA was generated using the MEGASCRIPT T7 transcription kit (Ambion). S2 cells were diluted to a final concentration of 1x106 cells/mL in Drosophila Serum-Free Medium (GIBCO, Paisley, UK). dsRNA was added to the serum-free medium and incubated for 1h at 25 0C, followed by addition of 2 mL complete Schneider’s medium. Part of an intron sequence of the human MAZ gene was used as mock dsRNA. One day after dsRNA transfection, cells were transfected with plasmid DNA. Two days later cells were subjected to either the luciferase refolding assay or the filtertrap assay.

Quantitative PCR Exponentially growing Schneider’s S2 cells were resuspended in complete Schneider’s Dro- sophila medium to a final concentration of 1x106 cells per mL. 5 mL of cell suspension was heat-shocked for the indicated temperature and timepoints using a precision waterbath. Total RNA was isolated using the Invisorb Spin Cell RNA mini kit (Westburg). First-strand cDNA was generated using M-MLV reverse transcriptase (Invitrogen) using oligo(dT)18 primers (Biolegio). Relative changes in transcript level were determined using the Icycler (Bio-Rad) in combination with SYBR green supermix (Bio-Rad). Calculations were performed using the comparative CT method according to User Bulletin 2 (Applied Biosystems) and (91). The PCR efficiencies for all primer pairs (Table S3) were between 85% and 100%. Fold induction was adjusted using RpL32 transcript levels as a standard.

Luciferase refolding assay The luciferase refolding assay was performed for HEK293 cells as previously described (92). The assay was adapted for Schneider’s S2 cells as follows. S2 cells were transfected with the pAc5.1-Luc plasmid, coding for firefly luciferase together with different heat shock protein- coding plasmids in a 1:9 ratio. After two days, cells were resuspended in complete Schneider’s Drosophila medium containing cycloheximide (2 mg / 100 mL). The cell suspension was divided into 400 mL portions in 1,5 ml centrifuge tubes. Tubes were placed into custom-made acrylic glass racks (50 positions) which allowed continuous water-flow around the tubes when placed in a waterbath. After heat-treatment, the tubes were cooled down rapidly by placing them in a 25 0C waterbath followed by incubation in a 25 0C incubator. Cell lyses and luciferase measurements were performed as described (92). The experiments were performed and measured in triplicate.

Biochemical techniques SDS–PAGE and Western blotting were performed by established procedures. Primary antibodies used were monoclonal anti-EGFP (Clontech) and monoclonal anti-V5 (Invitrogen) according to the manufacturers instruction. The filtertrap binding assay was performed as described previously (39).

83 Chapter 5

Lifespan analysis For lifespan analysis, 100-250 male flies were selected per line and maintained at non-crowding conditions at 22 0C. Dead flies were counted three times per week followed by transfer to fresh media.

Results

Heat-inducibility of the Drosophila small HSPs Certain members of all major HSP families are known to be induced upon proteotoxic stresses, including heat-shock and exposure to heavy metals (40). This enhanced expression reflects the need of additional chaperones when protein homeostasis is compromised and has to be restored to ensure cell survival. To analyse which of the D. melanogaster small HSPs are heat- inducible, we heat-shocked Drosophila S2 cells at 40 0C for 15 minutes and analysed the small HSP mRNA levels by QPCR at various time points after heating. As controls, we measured the mRNA levels of HSP70AA, a highly heat-inducible gene, and HSC70-4 and HSC70-5, two

A HSP70 / HSC70 B Classical sHSPs 10000 10000 Control HS 1000 HS 1h recovery 1000 HS 2h recovery n n

i o 100 i o 100 c t c t u u d d i n i n

l d l d 10

o 10 o F F

1 1

0.1 0.1 HSP70AA HSC70-4 HSC70-5 HSP22 HSP23 HSP26 HSP27 C Novel sHSPS D Pre-heat shock Post-heat shock 10000

1000 HSP23 HSP26

n HSP27 i o 100 c t CG14207 u d i n

l d 10 o F

0.1 HSP22 9 1 3 7 A C 0 6 3 0 B B 4 4 1 2 HSP67BC EFL 7 7 ) 7 4 3 4 6 6 2 1 1 ( P P CG4461 L C G C G C G C G H S H S CG7409

Figure 1. D. melanogaster small HSP family and heat inducibility. Transcript levels of HSP70/HS70 (A), the classical D. melanogaster small HSP genes (B) and the novel small HSP members (C) were analysed directly after heat shock and one and two hours after heat shock using QPCR. Relative mRNA abundance before and after heat shock are depicted in panel D. See Appendix 4 for colour print.

84 Small heat shock proteins & longevity constitutively expressed genes not responsive to heat-shock (41;42) (Figure 1a). The four classical small HSPs (HSP22, HSP23, HSP26, HSP27) were all found to be highly induced after a heat-shock (Figure 1b), consistent with previous findings (43;44). Apart from CG14207, all other small HSPs were also found to be heat-inducible (Figure 1c). The most strongly induced members were HSP67BC, L(2)EFL and CG4461 while HSP67BA, CG7409 and CG13133 were moderately induced (Figure 1c). Thus, CG14207 is the only Drosophila small HSP that is not responsive to heat shock. QPCR data were subsequently used to estimate the relative contribution of each member to the total small HSP mRNA levels before and after heat shock. In S2 cells, the most abundantly constitutively expressed small HSPs are HSP23 (50% of the total pool), HSP26 (12%), HSP27 (12%) and CG14207 (25%) (Figure 1d). After heat shock HSP27 mRNA levels show the largest increase, and together with HSP23 and HSP26 constitute the predominant small HSP messengers (Figure 1d). From the less abundantly expressed group of small HSPs (HSP22, HSP67BC, CG4461, CG7409), HSP67BC and CG4461 become the most abundant messengers after heat shock treatment (Figure 1d).

CG14207 and CG7409 are the most active members in assisting refolding of heat denatured luciferase It has been shown for several small HSPs that they can maintain substrates in a folding competent form both in vitro and in vivo (45-47). In vitro, the addition of the HSP70/HSP40 refolding machinery is required for the refolding reaction (48). This has been reproduced in living cells by using luciferase as a substrate (49;50). Here we tailored the cellular luciferase refolding assay for Drosophila S2 cells and characterized which Drosophila small HSPs could enhance luciferase refolding. The mitochondrial HSP22 was excluded from our analyses since our cellular assays were only tailored for the cytosolic and nuclear compartments. Consistent with in vitro data (30), overexpression of the classical small HSPs (HSP23, HSP26 and HSP27) increased luciferase refolding (Figure 2). Although less efficient, overexpression of L(2)EFL also led to improved luciferase refolding whereas HSP67BA, HSP67BC, CG4461 and CG13133 had no effect. Interestingly, overexpression of CG7409 and the non-heat shock inducible CG14207, resulted in the highest level of refolding 1 hour after heat-shock (Figure 2). To analyse whether D. melanogaster small HSPs, like bacterial, plant and mammalian small HSPs (47;49;51) also require HSP70 machines for refolding, we first tested which of theD. melanogaster HSP70s could promote refolding of heat-denatured luciferase. Hereto, we cloned a selection of the D. melanogaster HSP70 family (Supplemental Table 2) and analysed their effect on luciferase refolding. Overexpression of both D. melanogaster HSC70-2 and HSC70-4, but not HSC70-1, enhanced luciferase refolding in S2 cells (Figure 3a). This suggests that HSC70-1 Figure 2. The effect of overexpression Luciferase refolding in S2 cells of the D. melanogaster small HSP family Control 1 hour recovery in S2 cells on refolding heat-inactivated Hsp23 firefly luciferase. D. melanogaster small Hsp26 HSPs or eGFP (as control) and firefly lu- Hsp27 ciferase were co-expressed in Schneider’s Hsp67Ba S2 cells. 24 hours after transfection, S2 Hsp67Bc cells were heated for at 30 minutes at 38 °C L(2)EFL reducing luciferase activity to less than 5%. CG4461 CG7409 Next, cells were re-incubated for 1 hours at CG13133 25 °C to allow for (chaperone-assisted) lu- CG14207 ciferase refolding. Luciferase activities are plotted relative to the activity in unheated 0 10 20 30 40 cells (=100%). Data are the mean ± SD of 3 % Luciferase refolding independent experiments.

85 Chapter 5

A HSP70 / HSC70 refolding B Endogenous refolding & HSP70 RNAi

Control Control Mock HSP70AA HSP70AA HSC70-1 HSC70-1 HSC70-2 HSC70-2 dsRNA treatment HSC70-4 HSC70-4

70AA+70-1 0 10 20 30 40 % Luciferase refolding 0 10 20 30 40 % Luciferase refolding C HSP27 refolding & HSP70 RNAi CG14207 refolding & HSP70 RNAi Control D

Mock Control

HSP70AA Mock

HSC70-1 HSP70AA

HSC70-2 HSC70-1 dsRNA treatment HSC70-2 HSC70-4 dsRNA treatment HSC70-4 70AA+70-1

0 5 10 15 20 25 30 0 5 10 15 20 25 % Luciferase refolding % Luciferase refolding Figure 3. Dependency of HSP27 or CG14207 assisted luciferase refolding on HSP70. Firefly luciferase was co-expressed in Schneider’s S2 cells together with eGFP (control) or D. melanogaster HSP70s (A), dsRNA targeting D. melanogaster HSP70s (B), HSP27 + dsRNA targeting D. melanogaster HSP70s (C), or CG14207 + dsRNA targeting D. melanogaster HSP70s (D). 24 hours after transfection, S2 cells were heated for at 30 minutes at 38 °C reducing luciferase activity to less than 5%. Next, cells were re-incubated for 1 hours at 25 °C to allow for (chaperone-assisted) luciferase refolding. Luciferase activities are plotted relative to the activity in unheated cells (=100%). Data are the mean ± SD of 3 independent experiments. either lacks the ability to assist in refolding heat denatured luciferase or that specific co-factors that are required for its activity are rate-limiting under the conditions of HSC70-1 overexpression. Subsequently, we down-regulated the individual D. melanogaster HSP70s using dsRNA molecules (Supplemental Table 5). Transfection of mock dsRNA did not affect luciferase refolding while luciferase refolding was partially inhibited by the down regulation of D. melanogaster HSP70, HSC70-1 or HSC70-2. (Figure 3b). This confirmed the role of HSP70 members in substrate refolding in S2 cells and showed that endogenous HSC70-1 does play a role in refolding of heat-denatured luciferase under normal conditions. Next, we combined overexpression of D. melanogaster HSP27 and CG14207 with down regulation of D. melanogaster HSP70 members (Figure 3c,d). Refolding of luciferase in the presence of both D. melanogaster HSP27 and CG14207 was considerably reduced by down regulating either D. melanogaster HSP70AA, HSC70-1, HSC70-2, or HSC70-4 (Figure 3c,d). Thus our results demonstrate that the refolding capacity of D. melanogaster HSP27 and CG14207 requires an intact HSP70 machine.

HSP67BC is the most potent suppressor of polyglutamine aggregation Several protein folding diseases are characterized by the formation of toxic aggregates like Alzheimer’s disease, Huntington’s disease and amyotrophic lateral sclerosis. Some small HSPs have been reported to suppress the aggregation of such misfolded proteins (52;53) and we have recently found that this is not always related to their capacity to refold heat-denatured luciferase (Vos et al., in preparation). To test which of the Drosophila small HSPs could suppress aggregation of misfolded proteins, we co-expressed an EGFP tagged Huntingtin exon-1 containing

86 Small heat shock proteins & longevity

Figure 4. The effect of overexpression of the D. melanogaster Small HSP in A S2 cells on preventing aggregation of empty vector V5-HSP67BC poly-Q proteins. (A) EGFP-Htt-Q119, har-

V5-MRJ V5-L(2)EFL bouring exon 1 fragment of the huntingtin protein containing 119 glutamine repeats V5-HSP23 V5-CG4461 was co-expressed in Schneider’s S2 cells V5-HSP26 V5-CG7409 together with an empty vector (control) or

V5-HSP27 V5-CG13133 the various D. melanogaster small HSPs. MRJ, a well known potent inhibitor of V5-HSP67BA V5-CG14207 polyQ aggregation (54;93) was included HttQ119-eYFP as a positive control. 48 hours after transfection, S2 cells were lysed and analysed B using the filter trap assay. Serial five-fold

Protein Size (kDa) dilutions were loaded on cellulose acetate HSP23 20,6 membranes and probed with anti-GFP

V5-MRJ V5-HSP23 V5-HSP26 V5-HSP27 V5-HSP67Ba V5-HSP67Bc V5-L(2)EFL V5-CG4461 V5-CG7409 V5-CG13133 V5-CG14207 HSP26 22,9 empty vector antibody. (B) Western blot analysis of cell  HSP27 23,6 -eGFP HSP67BA 46,9 extracts prepared 48 hours after transfec- HSP67BC 22,2 tion showing soluble HDQ119-EYFP us- L(2)EFL 21,3 -V5 CG4461 23,7 ing anti-GFP antibodies and expression CG7409 17,9 of chaperones using anti-V5 antibodies. CG13133 24,5 -Tubulin CG14207 21,7 Note that, besides being very active in preventing aggregation, HSP67BC also C - HSP67BC + HSP67BC substantially lowered the levels of soluble Control HttQ119-EGFP; (C) EGFP-Htt-Q119 was

Mock co-expressed in Schneider’s S2 cells with

dsRNA (+) or without (-) HSP67BC together with HSP70AA dsRNA targeting D. melanogaster HSP70s. HSC70-4 poly-Q aggregates were detected using the

g loaded 50 10 100 20 filter trap assay using two dilutions. Sam- ples co-expressing HSP67BC were loaded at higher concentrations to better visualize trapped aggregated. 119 glutamines (EGFP-HDQ119) with the various D. melanogaster small HSPs in S2 cells. The HSP40 member MRJ, previously identified in a screen for suppressors of polyQ toxicity in flies (54) was used as a positive control and indeed also completely inhibited aggregate formation in S2 cells as demonstrated using filter trap binding (Figure 4a). While most small HSPs show only small effects on aggregate formation, HSP67BC was very active in preventing polyQ aggregation (Figure 4a). Interestingly, this small HSP did not support luciferase refolding in Drosophila S2 cells (Figure 2a). Inversely, the best supporter of protein refolding after heat denaturation (CG14207) did not significantly suppress polyQ aggregation. Except for CG4461, all small HSPs were expressed to equal levels (Figure 4b). We next tested whether HSP67BC requires a functional HSP70 machine to prevent polyQ aggregation. RNAi against HSP70AA and HSC70-4, that resulted in the inhibition of the refolding promoting activity of CG14207 (Figure 3d), did not lead to an increase in polyQ aggregation (Figure 4c), indicating that the protective effect of HSP67BC against polyQ aggregation does not depend on HSP70 activity. Our findings clearly demonstrate that HSP70-independent prevention of protein aggregation and HSP70-dependent stimulation of protein refolding are two different capacities that can be separated.

In vivo effects of small HSPs on polyQ toxicity and longevity

87 Chapter 5

HSP67BC CG7409 CG14207 CG4461 A C L1 L2 L3 L1 L2 L1 L2 L3 L4 L1 L2 L3 L4 Short exposure

Long exposure

B HSP67BC C CG7409 ** 180 180 L1 RNAi#1 L1 RNAi 160 160

140 140 s s e ** e * y 120 y 120 e e

d d e 100 e 100 c t c t e e f f f 80 f A 80 A

% 60 L1 x SCA3 RNAi#1 x SCA3 % 60 L1 x SCA3 RNAi x SCA3

40 40

20 20

0 0 i 1 SCA3-Q78 L1 L2 L3 2 SCA3-Q78 L1 L2 L3 i # i # HSP overexpression HSP overexpression R N A R N A R N A D CG14207 E CG4461 180 180 L2 RNAi#1 L1 RNAi#1 160 160

140 140 s s e e y y 120 120 e e

d d e e 100 100 c t c t e e f f f f 80 80 A A

% % 60 L2 x SCA3 RNAi#1 x SCA3 60 L1 x SCA3 RNAi#1 x SCA3

40 40

20 20

0 0 1 2 1 SCA3-Q78 L1 L2 L3 L4 2 i # i #

SCA3-Q78 L1 L2 L3 L4 i # i # HSP overexpression HSP overexpression R N A R N A R N A R N A Figure 5. The effect of overexpression of the D. melanogaster small HSP in vivo on reducing polyQ toxicity. (A) Expression levels of the V5 tagged small HSPs were examined in the individual small HSP transgenic lines with actin-GAL4 driven transgene expression. Male flies were prepared in Laemmli buffer (94) and analysed by Western blotting followed by staining with a V5 antibody. (B-E) Ataxin-3 induced eye- degeneration in flies expressing Ataxin-3 alone (SCA3-Q74) or together with (B) HSP67BC overexpression (L1, L2, L3) or HSP67BC downregulation (RNAi#1, RNAi#2) or with (C) CG7409 overexpression (L1, L2, L3) or CG7409 downregulation (RNAi) or with (D) CG14207 overexpression (L1, L2, L3, L4) or CG14207 downregulation (RNAi#1, RNAi#2) or with (E) CG4461 overexpression (L1, L2, L3) or CG4461 downregulation (RNAi#1, RNAi#2). In each panel, representative eye phenotypes are depicted for the indicated fly lines (samples for all fly lines are depicted in figure S3 and S4). Eye degeneration was quantified by microscopic classification into rough, patched, speckles or collapsed phenotypes (see also figure S2). The more severe phenotypes (patched, speckles, collapsed) were expressed relatively to the rough phenotype and set to 100% for the SCA3-Q74 expressing flies. Levels of significance (**P<0.001; *p<0.05) were determined using the Chi-square test. The numbers of flies analysed for each fly line are: SCA3-Q78 n=290, HSP67BC-L1 n=74, HSP67BC-L2 n=29, HSP67BC-L3 n=60, CG7409-L1 n=53, CG7409-L2 n=49, CG7409-L3 n=59, CG14207-L1 n=96, CG14207-L2 n=36, CG14207-L3 n=66, CG14207-L4 n=74, CG4461-L1 n=57, CG4461- L2 n=47, CG4461-L3 n=37, CG4461-L4 n=78, RNAi HSP67BC 1 n=14, RNAi HSP67BC 2 n=33, RNAi CG7409 n=145, RNAi CG14207 1 n=48, RNAi CG14207 2 n=218, RNAi CG4461 1 n=43, RNAi CG4461 2 n=40. To determine whether the different capacities of small HSPS to either prevent protein aggregation or to assist in HSP70-dependent protein refolding have an impact on polyQ induced eye- degeneration and longevity in vivo, we next selected CG14207 (strong refolder), HSP67BC (strong anti-aggregation), CG7409 (strong refolder and minor anti-aggregation) and CG4461 (no effect). Transgenic lines were generated (Table S1) and expression of the proteins of interest was confirmed by Western blot analysis of extracts of pUAS containing flies crossed with flies containing the actin-GAL4 driver (Figure 5a). Adverse effects of small HSP overexpression were not observed. Multiple transgenic lines per small HSP were maintained and analysed, thereby consolidating obtained results. In addition, we used RNAi lines (Table S1) targeting the four small HSPs..

88 Small heat shock proteins & longevity

As a model for polyQ diseases, we employed the Ataxin-3 fly-model (55). This fly-model ex- presses the Ataxin-3 gene with 78 CAG repeats under the control of the UAS/gmr-GAL4 expression system (56), resulting in eye-specific expression. Cryo-electron microscopy showed degeneration of the individual hexagonal ommatidia from flies expressing the SCA3-Q78, which was not observed in wild type flies (Figure S2a). This degeneration was also readily visible by light microscopy and used to score eye degeneration (Figure S2b). Transgenic lines overexpressing HSP67BC showed a significant reduction in eye degeneration (Figure 5b, S3b). Inversely, the two RNAi lines for HSP67BC drastically aggravated the eye degeneration (Figure 5b, S4a). This is entirely in line with the observation that HSP67BC can suppress the aggregation of polyglutamine containing proteins in cells. Interestingly, also transgenic flies overexpressing CG7409 showed a small but significant reduction in eye degeneration (Figure 5c, S3d). This small HSP had only a marginal effect on polyQ aggregation in cells (Figure 4a). Down regulation did, however, not worsen eye-degeneration (Figure 5c, S4c). Transgenic lines with either up- or down-regulated CG14207, which is not able to prevent polyQ aggregation in cells (Figure 4a), also did not show a changes in the level of eye-degeneration in flies, except for one transgenic line (UAS-CG14207-L1) (Figure 5d, S3e, S4d). Overexpression or down regulation of CG4461 had no effect in vivo (Figure 5e, S3c, S4b), which is consistent with its lack of any chaperone-like activity in cells (Figure 2,4a). In conclusion, the in vivo data suggest that only those small HSP members that show activity to inhibit poly-Q aggregation in cells (HSP67BC, CG7409) are also effective in preventing SCA3-Q78 mediated eye degeneration in vivo. Previous studies have shown that overexpression of HSP23, HSP26 and HSP27 can extend lifespan in flies (27-29). In this study, we show that these proteins have intermediate activities both in assisting refolding of stress denatured substrates and in preventing aggregation of disease associated, misfolded proteins. Moreover, we have demonstrated that these two functional properties can be separated and can be found in individual small HSPs. To more specifically ask which one of these chaperone activities is the most enviable in supporting longevity, we determined lifespan of flies overexpressing CG14207 (strong refolder), HSP67BC (strong anti-aggregation), CG7409 (strong refolder and intermediate anti-aggregation) or CG4461 (no effect) using elav-GAL4 or ey-GAL4. Elav-GAL4 drives expression in the central and peripheral nervous systems (57), while ey-GAL4 was used to drive ubiquitous expression throughout development and adulthood (FlyAtlas http://www.flyatlas.org). All control lines (W1118, elav-GAL4, ey-GAL4) showed comparable aging kinetics (Figure 6a), meaning that the drivers had not integrated into areas affecting normal lifespan. From the UAS transgenic lines, only UAS-CG7409-L2 and UAS-HSP67BC-L2 showed a small trend towards a shorter lifespan (Figure 6b). Expression of HSP67BC by elav-GAL4 resulted in a 5,8% mean lifespan increase (Figure 6c, 7), while expression by ey-GAL4 resulted in a 16,3% mean lifespan increase (Figure 6c, 7). Elav-GAL4 driven expression of CG7409, did not change the mean lifespan of the flies (Figure 6d, 7). However, when expressed under the control of ey-GAL4 a 16,3% increase in the mean lifespan was observed (Figure 6d, 7). The largest effect on mean lifespan was achieved by overexpression of CG14207. This led to a 17,8% and 20,4% mean lifespan increase by elav-GAL4 or ey-GAL4 driven expression respectively (Figure 6e, 7). Flies expressing CG4461 did not show any benefi- cial effects on mean lifespan (Figure 6f, 7). Thus, overexpression of small HSPs with chaperone activities either on heat unfolded proteins and/or on preventing polyQ aggregation were both able to extend the mean lifespan of flies.

Discussion

For the first time variousDrosophila small HSPs were functionally compared for heat-inducibility, their ability to assist protein refolding and their ability to prevent polyQ aggregation. Next, we have linked these activities to in vivo effects on poly-Q mediated neurodegeneration and longev-

89 Chapter 5

A Wildtype and driver lines B pUAS transformed lines

100 100

75 75

l l UAS-HSP67BC-L1 UAS-HSP67BC-L2 v i a v i a r r

u 50 u 50 UAS-CG14207-L3 S S

% % UAS-CG14207-L4 25 25 UAS-CG7409-L1 Wildtype UAS-CG7409-L2 elav UAS-CG4461-L1 0 ey 0 UAS-CG4461-L2

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) C HSP67BC overexpression (line 1)

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25

Control Control 0 UAS-HSP67BC-L1 x elav 0 UAS-HSP67BC-L1 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) HSP67BC overexpression (line 2)

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25

Control Control 0 UAS-HSP67BC-L2 x elav 0 UAS-HSP67BC-L2 x ey

0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (days) Time (days)

Figure 6. The effect of overexpression of the D. melanogaster small HSP in vivo on lifespan. Cohorts of 100-250 male flies were analysed for their lifespan. Every two days, dead flies were counted, and the percentage of flies alive at each time point is plotted. (A) Wiltype (W1118) and both driver lines (elav-GAL4, ey-GAL4). As they differ only marginally in their lifespan characteristics, the data from these lines was combined to plot the control lifespan curve in the panels C-F. (B) All UAS-sHSP carrying transgenic lines without GAL4 induced transgene expression. Except for line UAS-HSP67BC-2 and line UAS-CG7409-L2, these lines show comparable lifespan characteristics. (C) Effect of overexpression of HSP67BC using the elav-GAL4 (left –hand panels) or ey-GAL4 (right-hand panels) driver in 2 fly-lines. The individual lines are depicted separately as they were analysed in two individual lifespan experiments.

90 Small heat shock proteins & longevity

D CG7409 overexpression

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25 Control Control UAS-CG7409-L1 x elav UAS-CG7409-L1 x ey 0 UAS-CG7409-L2 x elav 0 UAS-CG7409-L2 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) E CG14207 overexpression

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25 Control Control UAS-CG14207-L3 x elav UAS-CG14207-L3 x ey 0 UAS-CG14207-L4 x elav 0 UAS-CG14207-L4 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) F CG4461 overexpression

100 elav 100 ey

80 75 l l 60 v i a v i a r r

u 50 u S S 40 % %

25 Control 20 Control UAS-CG4461-L1 x elav UAS-CG4461-L1 x ey 0 UAS-CG4461-L2 x elav 0 UAS-CG4461-L2 x ey

0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (days) Time (days)

Figure 6. Continued (D) Effect of overexpression of CG7409 using the elav-GAL4 (left–hand panel) or ey-GAL4 (right-hand panel) (E) Effect of overexpression of CG14207 using the elav-GAL4 (left–hand panel) or ey-GAL4 (right- hand panel) (F) Effect of overexpression of CG4461 using the elav-GAL4 (left–hand panel) or ey-GAL4 (right-hand panel). See Appendix 4 for colour print.

91 Chapter 5

50% of population survival 25 elav driven ey driven 20 e a s i n c r 15 e s p a n l i f

a g e 10 c e n t e r

P 5

0 HSP67BC CG7409 CG14207 CG4461 Cellular refolding activity 0 +++ +++ 0 Cellular anti-polyQ activity +++ +/- 0 0 Figure 7. The effect of overexpression of the D. melanogaster small HSP in vivo on lifespan: analysis at the 50% population survival level. The percentage of lifespan increase of the individual transgenic lines at 50% of population survival was determined relative to that in control flies for the transgenic lines with either elav-driven expression (white bars) or ey-driven HSP expression (grey bars). Values were calculated form the survival curves in Figure 6 and are indicated as means ± SD. The table indicates properties on protein refolding and polyQ aggregation of the sHSPs. ity. Intriguingly, we identified two small HSPs that were not previously studied, which exclusively supported either refolding (CG14207) or prevented aggregation (HSP67BC). CG14207 was the only small HSPs that was not heat-inducible and in addition depended on HSP70 for its refolding activity. HSP67BC, on the other hand, was clearly heat-inducible, showed strongest but HSP70 independent activity in preventing polyQ aggregation, while its overexpression did not increase the cellular ability to refold heat-denatured luciferase. Although diverse in action, both small HSPs were shown to increase lifespan in various degrees.

Classical HSP70 activity & anti-aggregation Chaperone-like actions of small HSPs have generally been suggested to depend on HSP70 activity. In protein refolding assays in vitro (58-60), aggregation prevention by the ATP-independent small HSPs has been shown to occur independent of the ATP-dependent HSP70 machinery, but for efficient refolding, substrate transfer to the HSP70 machine is required (47;51). Also, in living cells, refolding assistance by HSPB1 was found to depend on a functional HSP70 machinery (49) and in this study, a similar scenario was found to be true for the best refolding stimulating D. melanogaster small HSP; CG14207. Strikingly, CG14207 was in effective in preventing poly-Q aggregation. Inversely, the best protector against poly-Q aggregation, HSP67BC acted independent of the HSP70 machine, and was ineffective in the luciferase refolding assay. We recently found a similar pattern for members of the human family of small HSP proteins as well, where the efficacy to assist in HSP70-dependent refolding was almost inversely related to the ability to prevent poly-Q aggregation (Vos et al. MS in preparation). One can envision that transfer of non-foldable substrates to HSP70 would result in a fatuous ATP-driven substrate

92 Small heat shock proteins & longevity binding and release which may slightly delay aggregation but not prevent it. Indeed, in cells overexpression of HSP70s only marginally affects polyQ aggregation (61;62) whereas in vivo HSP70 mediated rescue of polyQ toxicity only occurs in the presence of semi-soluble nuclear polyQ aggregates (63-65). So, the different small HSPs with different affinities to substrates and HSP70s may have evolved to serve adequate processing of a broad spectrum of clients. The low HSP70 and high substrate affinity may serve to prevent aggregation of unfoldable substrates and provide a longer time window for ubiquitination and normal proteasomal turnover and/or autophagic processing.

In vivo anti-aggregation Using the in vivo Ataxin-3 fly model, we could show that the cellular data nicely predict the effects of the small HSP actions in terms of preventing poly-Q aggregation. The transgenic lines for CG7409 and especially HSP67BC showed reduced eye-degeneration. Inversely, the lines transgenic for CG14207 and CG4461, being ineffective in cells, also did not show reduced Ataxin-3 induced eye-degeneration, except for one line (line UAS-CG14207-L1). Strikingly, CG14207 has recently been reported as a modifier of Ataxin-3 neurodegeneration also using eye-degeneration as endpoint (55). However, it was shown that CG14207 did not directly interfere with aggregate formation (55) consistent with our cellular data. As we only find a similar effect in one out of the four CG14207 transgenic lines tested, the combined data suggest that CG14207 may not only prevent degeneration by a non-direct, compensatory mechanism but also depend on other factors (see also below).

Inducing longevity Several screens have been performed as an attempt to elucidate the molecular events un- derlying aging (66;67). This has revealed changed expression of several genes involved in many cellular pathways, including members of heat shock protein families. Instead of looking at physiologically-induced aging effects on gene expression, we tried to determine which molecular functions of small HSPs may contribute to enhancing lifespan. By screening the D. melanogaster small HSP family, we were able to show for the first time that longevity can be the result of two different chaperone-like activities; HSP70 dependent assistance of (re)folding reactions (CG14207) and HSP70 independent prevention of polyQ aggregation (HSP67BC). This is in line with the theory that overall protein homeostasis is important for healthy aging (68). Whereas previous data have demonstrated a role of other members of the D. melanogaster small HSP family protein refolding (30), we now show that these proteins have partial activities on the two substrates investigated here. Furthermore, other studies have already pointed out that a multitude of molecular chaperones of which the expression is regulated via HSF-1, a transcriptional regulator of stress-inducible gene expression, is vital in both the protection against protein folding diseases and aging (25;69;70). We show that overexpressing CG14207, a non HSF-1 regulated gene, can also provide longevity. This provides a means to boost longevity and avoid the adverse effects that HSF-1 activation and down-stream events may have in the development of cancer (71-73). It is rather surprising that especially small HSPs have been reported to induce longevity, while overexpression of HSP70 members, including D. melanogaster HSP70, one of the most strongly induced HSF targets (41) has no effect on lifespan (31). Interestingly, we show that overexpression of this member did not result in increased refolding activity (Figure 3a) and also did not affect polyQ aggregation (data not show). Yet, this HSP70 member does contribute to endogenous and small HSP-mediated refolding (Figure 3b,c,d) and deletion of D. melanogaster HSP70 genomic copies does translate in higher levels of neurotoxicity in flies (74). This suggests that a lack of co-factors, when HSP70 is overexpressed, may explain why it does not

93 Chapter 5 contribute to longevity in Drosophila. Alternatively, HSP70 overexpression may have deleterious side-effects; in flies HSP70 overexpression has been associated with reduced growth and development (75) and developmental abnormalities (76) that are not seen upon overexpression of small HSPs. In fact, HSP70 overexpression may allow stabilization of a greater variety of metastable proteins (2;77) than is the case at physiological levels. This could mask hidden mutations, allowing malformation effects. Physiological activity in buffering of genetic variation and masking hidden mutations has also been shown to be the case for another ATP dependent chaperone HSP90 (78;79). In view of this, employing small HSP activity to deal with adverse affects of protein misfolding seems rather a safer choice. Improving protein homeostasis has also been suggested to lead to resistance to age related protein folding diseases such as polyQ related diseases (68). So, why does a general increase in protein homeostasis mediated by CG14207 that does lead to longevity increases, not also delays Ataxin-3 mediated eye degeneration? Our data suggest that dealing with high polyQ overexpression, produced within a short time both in cells and in the Ataxin-3 flymodel, is only effective by chaperones capable of modifying aggregate formation directly. Dealing with polyQ induced misfolding of meta-stable proteins (80) by CG14207 seems not effective enough to resist polyQ toxicity. However, at normal physiological circumstances either dealing with polyQ aggregation itself or dealing with the side-effects, might turn out to be equally effective in resist- ing neurodegeneration.

Acknowledgements

This work was supported by Innovatiegerichte Onderzoeksprogramma Genomics. Grant IGE03018.

References

1. Arslan, M. A., Csermely, P., and Soti, C. (2006) Protein homeostasis and molecular chaperones in aging, Biogeron- tology. 7, 383-389. 2. Morimoto, R. I. (2008) Proteotoxic stress and inducible chaperone networks in neurodegenerative disease and aging, Genes Dev. 22, 1427-1438. 3. Vacher, C., Garcia-Oroz, L., and Rubinsztein, D. C. (2005) Overexpression of yeast hsp104 reduces polyglutamine aggregation and prolongs survival of a transgenic mouse model of Huntington’s disease, Hum. Mol. Genet. 14, 3425-3433. 4. Zhang, X., Smith, D. L., Meriin, A. B., Engemann, S. and others (2005) A potent small molecule inhibits polyglutamine aggregation in Huntington’s disease neurons and suppresses neurodegeneration in vivo, Proc. Natl. Acad. Sci. U. S. A 102, 892-897. 5. Haass, C. and Selkoe, D. J. (2007) Soluble protein oligomers in neurodegeneration: lessons from the Alzheimer’s amyloid beta-peptide, Nat. Rev. Mol. Cell Biol. 8, 101-112. 6. Dobson, C. M. (2003) Protein folding and misfolding, Nature 426, 884-890. 7. Ellis, R. J. and Minton, A. P. (2006) Protein aggregation in crowded environments, Biol. Chem. 387, 485-497. 8. Homouz, D., Perham, M., Samiotakis, A., Cheung, M. S. and others (2008) Crowded, cell-like environment induces shape changes in aspherical protein, Proc. Natl. Acad. Sci. U. S. A 105, 11754-11759. 9. Engel, R., Westphal, A. H., Huberts, D. H., Nabuurs, S. M. and others (2008) Macromolecular crowding compacts unfolded apoflavodoxin and causes severe aggregation of the off-pathway intermediate during apoflavodoxin folding, J. Biol. Chem. 10. van Leeuwen, F. W., Fischer, D. F., Kamel, D., Sluijs, J. A. and others (2000) Molecular misreading: a new type of transcript mutation expressed during aging, Neurobiol. Aging 21, 879-891. 11. Pettigrew, C. A. and Brown, M. A. (2008) Pre-mRNA splicing aberrations and cancer, Front Biosci. 13, 1090-1105. 12. Kramer, E. B. and Farabaugh, P. J. (2007) The frequency of translational misreading errors in E. coli is largely determined by tRNA competition, RNA. 13, 87-96. 13. Parker, J. (1989) Errors and alternatives in reading the universal genetic code, Microbiol. Rev. 53, 273-298. 14. Bulteau, A. L., Szweda, L. I., and Friguet, B. (2002) Age-dependent declines in proteasome activity in the heart, Arch. Biochem. Biophys. 397, 298-304. 15. Ferrington, D. A., Husom, A. D., and Thompson, L. V. (2005) Altered proteasome structure, function, and oxidation in aged muscle, FASEB J. 19, 644-646.

94 Small heat shock proteins & longevity

16. Vernace, V. A., Arnaud, L., Schmidt-Glenewinkel, T., and Figueiredo-Pereira, M. E. (2007) Aging perturbs 26S proteasome assembly in Drosophila melanogaster, FASEB J. 21, 2672-2682. 17. Naidoo, N., Ferber, M., Master, M., Zhu, Y. and others (2008) Aging impairs the unfolded protein response to sleep deprivation and leads to proapoptotic signaling, J. Neurosci. 28, 6539-6548. 18. Liu, A. Y., Lin, Z., Choi, H. S., Sorhage, F. and others (1989) Attenuated induction of heat shock gene expression in aging diploid fibroblasts, J. Biol. Chem. 264, 12037-12045. 19. Liu, A. Y., Choi, H. S., Lee, Y. K., and Chen, K. Y. (1991) Molecular events involved in transcriptional activation of heat shock genes become progressively refractory to heat stimulation during aging of human diploid fibroblasts, J. Cell Physiol 149, 560-566. 20. Urushitani, M., Kurisu, J., Tateno, M., Hatakeyama, S. and others (2004) CHIP promotes proteasomal degradation of familial ALS-linked mutant SOD1 by ubiquitinating Hsp/Hsc70, J. Neurochem. 90, 231-244. 21. Carra, S., Seguin, S. J., and Landry, J. (2008) HspB8 and Bag3: a new chaperone complex targeting misfolded proteins to macroautophagy, Autophagy. 4, 237-239. 22. Rujano, M. A. and Kampinga, H. H. (2008) The HSP70 chaperone machine as guardian of the proteome: Implica- tions for protein folding diseases, in Heat Shock Proteins in Biology and Medicine (Radons, J. and Multhoff, G., Eds.) Research Signpost. 23. Morrow, G. and Tanguay, R. M. (2003) Heat shock proteins and aging in Drosophila melanogaster, Semin. Cell Dev. Biol. 14, 291-299. 24. Hsu, A. L., Murphy, C. T., and Kenyon, C. (2003) Regulation of aging and age-related disease by DAF-16 and heat- shock factor, Science 300, 1142-1145. 25. Morley, J. F. and Morimoto, R. I. (2004) Regulation of longevity in Caenorhabditis elegans by heat shock factor and molecular chaperones, Mol. Biol. Cell 15, 657-664. 26. Walker, G. A. and Lithgow, G. J. (2003) Lifespan extension in C. elegans by a molecular chaperone dependent upon insulin-like signals, Aging Cell 2, 131-139. 27. Aigaki, T., Seong, K. H., and Matsuo, T. (2002) Longevity determination genes in Drosophila melanogaster, Mech. Ageing Dev. 123, 1531-1541. 28. Wang, H. D., Kazemi-Esfarjani, P., and Benzer, S. (2004) Multiple-stress analysis for isolation of Drosophila longevity genes, Proc. Natl. Acad. Sci. U. S. A 101, 12610-12615. 29. Morrow, G., Samson, M., Michaud, S., and Tanguay, R. M. (2004) Overexpression of the small mitochondrial Hsp22 extends Drosophila life span and increases resistance to oxidative stress, FASEB J. 18, 598-599. 30. Morrow, G., Heikkila, J. J., and Tanguay, R. M. (2006) Differences in the chaperone-like activities of the four main small heat shock proteins of Drosophila melanogaster, Cell Stress. Chaperones. 11, 51-60. 31. Minois, N., Khazaeli, A. A., and Curtsinger, J. W. (2001) Locomotor activity as a function of age and life span in Drosophila melanogaster overexpressing hsp70, Exp. Gerontol. 36, 1137-1153. 32. Deuerling, E. and Bukau, B. (2004) Chaperone-assisted folding of newly synthesized proteins in the cytosol, Crit Rev. Biochem. Mol. Biol. 39, 261-277. 33. Bukau, B., Deuerling, E., Pfund, C., and Craig, E. A. (2000) Getting newly synthesized proteins into shape, Cell 101, 119-122. 34. de Wit, N. J., Verschuure, P., Kappe, G., King, S. M. and others (2004) Testis-specific human small heat shock protein HSPB9 is a cancer/testis antigen, and potentially interacts with the dynein subunit TCTEL1, Eur. J. Cell Biol. 83, 337-345. 35. Parcellier, A., Schmitt, E., Gurbuxani, S., Seigneurin-Berny, D. and others (2003) HSP27 is a ubiquitin-binding protein involved in I-kappaBalpha proteasomal degradation, Mol. Cell Biol. 23, 5790-5802. 36. Marin-Vinader, L., Shin, C., Onnekink, C., Manley, J. L. and others (2006) Hsp27 enhances recovery of splicing as well as rephosphorylation of SRp38 after heat shock, Mol. Biol. Cell 17, 886-894. 37. Vos, M. J., Hageman, J., Carra, S., and Kampinga, H. H. (2008) Structural and functional diversities between members of the human HSPB, HSPH, HSPA, and DNAJ chaperone families, Biochemistry 47, 7001-7011. 38. Lelj-Garolla, B. and Mauk, A. G. (2006) Self-association and chaperone activity of Hsp27 are thermally activated, J. Biol. Chem. 281, 8169-8174. 39. Carra, S., Sivilotti, M., Chavez Zobel, A. T., Lambert, H. and others (2005) HspB8, a small heat shock protein mutated in human neuromuscular disorders, has in vivo chaperone activity in cultured cells, Hum. Mol. Genet. 14, 1659-1669. 40. Mosser, D. D., Theodorakis, N. G., and Morimoto, R. I. (1988) Coordinate changes in heat shock element-binding activity and HSP70 gene transcription rates in human cells, Mol. Cell Biol. 8, 4736-4744. 41. Ashburner, M. and Bonner, J. J. (1979) The induction of gene activity in drosophilia by heat shock, Cell 17, 241- 254. 42. Ish-Horowicz, D., Pinchin, S. M., Schedl, P., Artavanis-Tsakonas, S. and others (1979) Genetic and molecular analysis of the 87A7 and 87C1 heat-inducible loci of D. melanogaster, Cell 18, 1351-1358. 43. Marin, R., Demers, M., and Tanguay, R. M. (1996) Cell-specific heat-shock induction of Hsp23 in the eye of Dro- sophila melanogaster, Cell Stress. Chaperones. 1, 40-46. 44. Michaud, S., Marin, R., and Tanguay, R. M. (1997) Regulation of heat shock gene induction and expression during Drosophila development, Cell Mol. Life Sci. 53, 104-113. 45. Ecroyd, H., Meehan, S., Horwitz, J., Aquilina, J. A. and others (2007) Mimicking phosphorylation of alphaB-crystallin affects its chaperone activity, Biochem. J. 401, 129-141.

95 Chapter 5

46. Cashikar, A. G., Duennwald, M., and Lindquist, S. L. (2005) A Chaperone Pathway in Protein Disaggregation: HSP26 alters the nature of protein aggregates to facilitate reactivation by HSP104, J. Biol. Chem. 280, 23869- 23875. 47. Mogk, A., Schlieker, C., Friedrich, K. L., Schonfeld, H. J. and others (2003) Refolding of substrates bound to small Hsps relies on a disaggregation reaction mediated most efficiently by ClpB/DnaK, J. Biol. Chem. 278, 31033- 31042. 48. Lee, G. J., Roseman, A. M., Saibil, H. R., and Vierling, E. (1997) A small heat shock protein stably binds heat- denatured model substrates and can maintain a substrate in a folding-competent state, EMBO J. 16, 659-671. 49. Bryantsev, A. L., Kurchashova, S. Y., Golyshev, S. A., Polyakov, V. Y. and others (2007) Regulation of stress-induced intracellular sorting and chaperone function of Hsp27 (HspB1) in mammalian cells, Biochem. J. 407, 407-417. 50. Nollen, E. A., Kabakov, A. E., Brunsting, J. F., Kanon, B. and others (2001) Modulation of in vivo HSP70 chaperone activity by Hip and Bag-1, J. Biol. Chem. 276, 4677-4682. 51. Lee, G. J. and Vierling, E. (2000) A small heat shock protein cooperates with heat shock protein 70 systems to reactivate a heat-denatured protein, Plant Physiol 122, 189-198. 52. Carra, S., Seguin, S. J., Lambert, H., and Landry, J. (2008) HspB8 chaperone activity toward poly(Q)-containing proteins depends on its association with Bag3, a stimulator of macroautophagy, J. Biol. Chem. 283, 1437-1444. 53. Wilhelmus, M. M., Boelens, W. C., Otte-Holler, I., Kamps, B. and others (2006) Small heat shock proteins inhibit amyloid-beta protein aggregation and cerebrovascular amyloid-beta protein toxicity, Brain Res. 1089, 67-78. 54. Fayazi, Z., Ghosh, S., Marion, S., Bao, X. and others (2006) A Drosophila ortholog of the human MRJ modulates polyglutamine toxicity and aggregation, Neurobiol. Dis. 24, 226-244. 55. Bilen, J. and Bonini, N. M. (2007) Genome-wide screen for modifiers of ataxin-3 neurodegeneration in Drosophila, PLoS. Genet. 3, 1950-1964. 56. Brand, A. H. and Perrimon, N. (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes, Development 118, 401-415. 57. Robinow, S. and White, K. (1991) Characterization and spatial distribution of the ELAV protein during Drosophila melanogaster development, J. Neurobiol. 22, 443-461. 58. Horwitz, J. (1992) Alpha-crystallin can function as a molecular chaperone, Proc. Natl. Acad. Sci. U. S. A 89, 10449- 10453. 59. Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993) Small heat shock proteins are molecular chaperones, J. Biol. Chem. 268, 1517-1520. 60. Kim, R., Kim, K. K., Yokota, H., and Kim, S. H. (1998) Small heat shock protein of Methanococcus jannaschii, a hyperthermophile, Proc. Natl. Acad. Sci. U. S. A 95, 9129-9133. 61. Chai, Y., Koppenhafer, S. L., Bonini, N. M., and Paulson, H. L. (1999) Analysis of the role of heat shock protein (Hsp) molecular chaperones in polyglutamine disease, J. Neurosci. 19, 10338-10347. 62. Rujano, M. A., Kampinga, H. H., and Salomons, F. A. (2007) Modulation of polyglutamine inclusion formation by the Hsp70 chaperone machine, Exp. Cell Res. 313, 3568-3578. 63. Cummings, C. J., Sun, Y., Opal, P., Antalffy, B. and others (2001) Over-expression of inducible HSP70 chaperone suppresses neuropathology and improves motor function in SCA1 mice, Hum. Mol. Genet. 10, 1511-1518. 64. Warrick, J. M., Chan, H. Y., Gray-Board, Chai, Y. and others (1999) Suppression of polyglutamine-mediated neurodegeneration in Drosophila by the molecular chaperone HSP70, Nat. Genet. 23, 425-428. 65. Chan, H. Y., Warrick, J. M., Gray-Board, Paulson, H. L. and others (2000) Mechanisms of chaperone suppression of polyglutamine disease: selectivity, synergy and modulation of protein solubility in Drosophila, Hum. Mol. Genet. 9, 2811-2820. 66. Zou, S., Meadows, S., Sharp, L., Jan, L. Y. and others (2000) Genome-wide study of aging and oxidative stress response in Drosophila melanogaster, Proc. Natl. Acad. Sci. U. S. A 97, 13726-13731. 67. Pletcher, S. D., Macdonald, S. J., Marguerie, R., Certa, U. and others (2002) Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster, Curr. Biol. 12, 712-723. 68. Balch, W. E., Morimoto, R. I., Dillin, A., and Kelly, J. W. (2008) Adapting proteostasis for disease intervention, Sci- ence 319, 916-919. 69. Cohen, E., Bieschke, J., Perciavalle, R. M., Kelly, J. W. and others (2006) Opposing activities protect against age- onset proteotoxicity, Science 313, 1604-1610. 70. Walker, G. A., White, T. M., McColl, G., Jenkins, N. L. and others (2001) Heat shock protein accumulation is upregulated in a long-lived mutant of Caenorhabditis elegans, J. Gerontol. A Biol. Sci. Med. Sci. 56, B281-B287. 71. Page, T. J., Sikder, D., Yang, L., Pluta, L. and others (2006) Genome-wide analysis of human HSF1 signaling reveals a transcriptional program linked to cellular adaptation and survival, Mol. Biosyst. 2, 627-639. 72. Wang, J. H., Yao, M. Z., Gu, J. F., Sun, L. Y. and others (2002) Blocking HSF1 by dominant-negative mutant to sensitize tumor cells to hyperthermia, Biochem. Biophys. Res. Commun. 290, 1454-1461. 73. Dai, C., Whitesell, L., Rogers, A. B., and Lindquist, S. (2007) Heat shock factor 1 is a powerful multifaceted modifier of carcinogenesis, Cell 130, 1005-1018. 74. Gong, W. J. and Golic, K. G. (2006) Loss of Hsp70 in Drosophila is pleiotropic, with effects on thermotolerance, recovery from heat shock and neurodegeneration, Genetics 172, 275-286. 75. Krebs, R. A. and Feder, M. E. (1997) Deleterious consequences of Hsp70 overexpression in Drosophila melanogaster larvae, Cell Stress. Chaperones. 2, 60-71. 76. Williams, K. D., Helin, A. B., Posluszny, J., Roberts, S. P. and others (2003) Effect of heat shock, pretreatment and

96 Small heat shock proteins & longevity

hsp70 copy number on wing development in Drosophila melanogaster, Mol. Ecol. 12, 1165-1177. 77. Prahlad, V., Cornelius, T., and Morimoto, R. I. (2008) Regulation of the cellular heat shock response in Caenorhab- ditis elegans by thermosensory neurons, Science 320, 811-814. 78. Queitsch, C., Sangster, T. A., and Lindquist, S. (2002) Hsp90 as a capacitor of phenotypic variation, Nature 417, 618-624. 79. Rutherford, S. L. and Lindquist, S. (1998) Hsp90 as a capacitor for morphological evolution, Nature 396, 336-342. 80. Gidalevitz, T., Ben Zvi, A., Ho, K. H., Brignull, H. R. and others (2006) Progressive disruption of cellular protein folding in models of polyglutamine diseases, Science 311, 1471-1474. 81. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular cloning a laboratory manual Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 82. McGinnis, S. and Madden, T. L. (2004) BLAST: at the core of a powerful and diverse set of sequence analysis tools, Nucleic Acids Res. 32, W20-W25. 83. Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R. and others (2007) Clustal W and Clustal X version 2.0, Bioinformatics. 23, 2947-2948. 84. Page, R. D. (1996) TreeView: an application to display phylogenetic trees on personal computers, Comput. Appl. Biosci. 12, 357-358. 85. King, V. and Tower, J. (1999) Aging-specific expression of Drosophila hsp22, Dev. Biol. 207, 107-118. 86. Kurapati, R., Passananti, H. B., Rose, M. R., and Tower, J. (2000) Increased hsp22 RNA levels in Drosophila lines genetically selected for increased longevity, J. Gerontol. A Biol. Sci. Med. Sci. 55, B552-B559. 87. Morrow, G., Inaguma, Y., Kato, K., and Tanguay, R. M. (2000) The small heat shock protein Hsp22 of Drosophila melanogaster is a mitochondrial protein displaying oligomeric organization, J. Biol. Chem. 275, 31204-31210. 88. Michaud, S., Lavoie, S., Guimond, M. O., and Tanguay, R. M. (2008) The nuclear localization of Drosophila Hsp27 is dependent on a monopartite arginine-rich NLS and is uncoupled from its association to nuclear speckles, Biochim. Biophys. Acta 1783, 1200-1210. 89. Michaud, S., Morrow, G., Marchand, J., and Tanguay, R. M. (2002) Drosophila small heat shock proteins: cell and organelle-specific chaperones?, Prog. Mol. Subcell. Biol. 28, 79-101. 90. Clemens, J. C., Worby, C. A., Simonson-Leff, N., Muda, M. and others (2000) Use of double-stranded RNA interference in Drosophila cell lines to dissect signal transduction pathways, Proc. Natl. Acad. Sci. U. S. A 97, 6499-6503. 91. Livak, K. J. and Schmittgen, T. D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method, Methods 25, 402-408. 92. Michels, A. A., Nguyen, V. T., Konings, A. W., Kampinga, H. H. and others (1995) Thermostability of a nuclear- targeted luciferase expressed in mammalian cells. Destabilizing influence of the intranuclear microenvironment, Eur. J. Biochem. 234, 382-389. 93. Chuang, J. Z., Zhou, H., Zhu, M., Li, S. H. and others (2002) Characterization of a brain-enriched chaperone, MRJ, that inhibits Huntingtin aggregation and toxicity independently, J. Biol. Chem. 277, 19831-19838. 94. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227, 680-685.

97 Chapter 5

Supplemental tables and figures Primer sequence (5'-3')

GCGCGATATCACCATGCCTGCTATTGGAATCGATC GCAGCTCGAGTTAGACTCTTTGGCCTTAGTCG CGACGATATCACCATGAAGCATTGGCCCTTCGA ATCAGCGGCCGCCTAATCAACCTCCTCAATGG ACCAGATATCACCATGGGTAAAATTCCGGCCATCGGCATCG ACCAGCGGCCGCGGTCGGTGGATCGTACTATC ACAAGCGGCCGCAAGATGTCTAAAGCTCCTGCTGTT ACAAACGCGTATGGTTTAGTCGACCTCCTC GATCCCATGGATAACATGGCCAACAAAACCATGG ATCCGTAGAATCGAGACCGAGGAGAGGGTTAGGGATAGGCT TCGACCATGGTTTTGTTGGCCATGTTATCCATGG AGTACTACCATGGGTAAGCCTATCCCTAACCCTCTCCTCGGT TACCCATGGTAGTACTGTAC CTCGATTCTACGGAT ACCCCTCGAGCACCATGGCAAATATTCCATTGTTGT ACCGCTCGAGGACCATGTCGCTATCTACTCTGCTTTC ACCAGCTCGAGACCATGTCAATTATACCACTGCTGCAC ACAACTCGAGAACCATGTCCGTAGTGCCACTGAT ACCCACGCGTTCACTTGGCCAGTTCCTCGGTTT ACCCCTCGAGCACCATGCCAGATATTCCCTTTGT ACCGACGCGTGACACATGGTCCACTATTTCC ACCCACGCGTCTGTGATTAGTTCTTGATGC AACCACGCGTTCATTTGTCCATCGCCGTGCCAGCAG ACCCACGCGTAAATAATGCAGGGCATCTCT ACCCACGCGTGGCTCCTTTACTTGTCCTTG ACCCACGCGTTTCTCGTCTTCTTCATCTCCTAGTC ACCCACGCGTACTAACACTGAAGGCGGTACA ACCACTCGAGGACCATGTCGCTGATACCGTTCATAC ACCTACGCGTTCACTTGGCTTCTGGCTCTG ACGCCTCGAGAGAAATGTCTCTGGTGCCTACC ACAACTCGAGAACCATGGCTCTTGTCCCGGCTAC AAACCTCGAGAACCATGCCCATCCACTGGGACTG AACCCTCGAGAACCATGGCCGAGGCTAACAAGAGG ACCCACGCGTACTGGATAGCTGGACAACTGGA ACCCGATATCACCAACATGGTTGACTACTAT ACAAACGCGTGCTATTGAAGGGAGCCCATC Primer nam e HSP23 F HSP23 R HSP26 F HSP26 R HSP27 F HSP27 R L(2)EFL F L(2)EFL R HSP67Ba F HSP67Ba R HSP67Bc F HSP67Bc R CG4461 F CG4461 R CG7409 F CG7409 R CG13133 F CG13133 R CG14207 F CG14207 F MRJ F MRJ R DmHSP70A a DmHSP70A a HSC70-1 HSC70-1 HSC70-2 HSC70-2 HSC70-4 HSC70-4 XcmI-F stu f XcmI-R stu f V5 stuffer F V5 stuffer R Gene famil y sHSP HSP40 HSP70 Modifications L4440-XcmI pAc5.1-V5 Table S2 Primers used for molecular cloning y y y y y y y y y y y y y y y s e d d d d d d d d d d d d d d d e c c u u u u u u u u u u u u u u u n i t t t t t t t t t t t t t t t v e s s s s s s s s s s s s s s s r

r VDRC VDRC VDRC VDRC VDRC VDRC VDRC s s s s s s s s s s s s s s s e e i i i i i i i i i i i i i i i f S h h h h h h h h h h h h h h h N. Bonini

e T T T T T T T T T T T T T T T c i R t Bloomington Bloomington Bloomington e n e G / 4414 high ubiquitous expression 1 1 1 1 1 + 1 1 b b b b 1 1 1 1 b b b b b b b S S S S S y y y

S S , , , , S S S S ,

C C C , , 1 1 1 1 , , , , , , 1 , r r r r r 1 1 1 1 1 1 s r r e e e e r O O O r r r e ) ) ) c e e e e e e S S S S i S

t R R R

S S , , , , S S S S ,

L L L , , 3 3 3 3 , , i , , 3 2 2 2 r 3 3 3 3 3 3 ( ( ( M M M M M e n n n t M M M M T T T T M M I I T I / / / / / / / / c T T T T T T / / 1 2 6 7 / / / / 4 5 7 0 a 1 4 r 7 1 1 2 X X B B 0 0 0 1 - - - - a Z C C A F F L L L L - - - 7 7 7 7 ------h 9 9 9 0 0 0 0 0 0 0 1 1 1 1 c

0 0 0 2 2 2 2 9 9 9 6 6 6 6 1 d 4 4 4 4 4 4 4 n 1 1 1 4 4 4 4 b i 7 n 7 7 1 1 1 1 4 4 4 4 4 4 4 S a

r a G , G G G G G G G G G G G G G t

s 1 e C C C C C C C C C C C C C C

r ------p ; P{ey1x-GAL4.Exel}2 / 8228 moderate ubiquitous expression e e 5 5 5 5 5 5 5 5 5 5 5 5 5 5 y p t S V V V V V V V V V V V V V V

y ; P{Act5C-GAL4}25FO1/CyO, y ------, o t * 1118 3 S S S S S S S S S S d S S S S n l ; P{GAL4-elav.L}2/CyO / 8765 high expression in CNS i w w e * A A A A A A A A A A A A A A M 1 1 U y y U U U U U U UAS-HSP67BC-RNAi / VDRC Transformant ID 26416 UAS-HSP67BC-RNAi / VDRC Transformant ID 26417 UAS-CG4461-RNAi / VDRC Transformant ID 40528 UAS-CG4461-RNAi / VDRC Transformant ID 40529 UAS-CG7409-RNAi / VDRC Transformant ID 40637 UAS-CG14207-RNAi / VDRC Transformant ID 31800 UAS-CG14207-RNAi / VDRC Transformant ID 31802 w W G T U U U GMR-GAL4 UAS-SCA3trQ78/ln(2LR), Cy U U U U e 18 1 118-TM3 m 1 1 a CG7409-L1 CG7409-L2 CG7409-L3 CG14207-L1 CG14207-L2 CG14207-L3 CG14207-L4 HSP67BC RNAi 1 HSP67BC RNAi 2 CG4461 RNAi 1 CG4461 RNAi 2 CG7409 RNA i CG14207 RNAi 1 CG14207 RNAi 2 act -GAL4 ey -GAL4 W N W SCA3trQ7 8 HSP67BC-L1 elav -GAL4 HSP67BC-L2 HSP67BC-L3 CG4461-L1 CG4461-L2 CG4461-L3 CG4461-L4 Table S1 Drosophila lines used in this study

98 Small heat shock proteins & longevity y y y y e d d d d c u u u u n t t t t e s s s s

r s s s s e i i i i f h h h h e Invitrogen This study This study This study This study This study This study This study This study This study This study T This study This study This study This study This study This study This study T T T R ) a g s e e n m i o r m a P ) t ( y

u r l 3 o g L t

g a 9 G r a 1 t p o

- 1 b

5 n m i a h l V t o

i r l t f e w

a o r

i r d n i 1 F P

e (

t m n n r r n o o o l e t x e t c - c c

e /

s e

N n

v e i e /

r t

s r g o g a o n u n r i t l i t e c d F f n i

e e c u n v e

s f u e H l

n i c

e i o r d d t A i t e e s G N i a

g g r R d g g

e m e t r a a d t t c c o

e f n t a P P s Actin promoter HSP23 / GeneID: 39077 Actin promoter / N-terminal V5-tag r a p CG13133 / GeneID: 34342 CG14207 / GeneID: 32955 HSP26 / GeneID: 39075 HSP27 / GeneID: 39078 HSP67BA / GeneID: 39076 HSP67BC / GeneID: 39071 L(2)EFL / GeneID: 37744 CG4461 / GeneID: 39074 MRJ / GeneID: 36797 CG7409 / GeneID: 38870 HSP70AA / GeneID: 48581 HSC70-1 / GeneID: 39542 HSC70-2 / GeneID: 41609 HSC70-4 / GeneID: 41840 HSPB1 / GeneID: 3315 n F F a h a a h n G G d r E C Dm E Dm E A Dm Dm Hs Dm Dm Dm Dm Dm Dm Dm T Dm Dm Dm Dm Dm Dm P F G E - 9 e 1 1 P m F a Q c-EGFP t 5 t n G u A

/ V H L E d - - - - T i - T 1 1 1 . . . 0 m S 5 5 5 4 s A c c c 4 a l A A A 4 U P pAc5.1 p pAc5.1-V5-HSP23 p p L pAc5.1-V5 pAc5.1-V5-CG14207 pAc5.1-V5-HSPB1 pAc5.1-V5-HSP26 pAc5.1-V5-HSP27 pAc5.1-V5-HSP67BA pAc5.1-V5-HSP67BC pAc5.1-V5-L(2)EFL pAc5.1-V5-CG4461 pAc5.1-V5-CG7409 p pAc5.1-V5-MRJ pAc5.1-V5-CG13133 pAc5.1-V5-HSP70AA pAc5.1-V5-HSC70-1 pAc5.1-V5-HSC70-2 pAc5.1-V5-HSC70-4 Cloning sHSP General Table S4 Plasmids used in this study HSP40 HSP70 ) ' 3 - ' 5 (

e c n e u q e s

r e m i r P CTGCTGTTGGTATTGATTTGG CATAGGATGGAGTGGTACGA CACTATTCCCACCAAGAAGTC TTATTGTCGTTAGCCATCTCAC CACTTCCACTTTAAATCCGACTG CCAGGGAAATCACAATCGTC GATTGGGATTTGAACGACTGG GTGCACCACTTTCAGATCTC TTAGCTCACCCTTGATCCAG ACCTCAACCTATCCATCAACC ACATGCTGCCCACCGGATTC CTGTCAAGAGTATAAATAGCCACC GTTGTCTGAGGTTTCAATCGT CAATGGCAACGATAAGTAGAG TGTGTATGAGTGAAGTGTCCT CGCATCATTCAAATTCAGCA ATGGCTCCTTTACTTGTCCT CTAGACAGGGTTGTGAATAAAGAG AAACCGAAGTCATCCTCCAG GAAGTTCGAGGTCATTCTGG CTTCTCCTCATGCTTTCCCT TTTCCATCTCTGTCTCTTCCT GGTTCCACCATTAGTTTCCA CCTCCTTATGACCGTTTCCA CTGGATATGTCCGAGATTACCT TCTCATAGGGCTTGAACAGG CTTGTAGGATCGCGTAAACTC TTTGTGTCCGACTTCTCCTC GTCTCAATTCCCAATGAAAGTG GCGATCTCGCCGCAGTAAAC e m a n

r e m i r qPCR HSP22 R qPCR HSP23 F qPCR HSP23 R qPCR HSP26 F qPCR HSP26 R qPCR HSP27 F qPCR HSP27 R qPCR L(2)EFL F qPCR L(2)EFL R qPCR HSP67Ba F qPCR HSP67Bc F qPCR HSP67Bc R qPCR CG4461 F qPCR CG7409 F qPCR CG13133 F qPCR CG14207 F qPCR HSC70-4 F qPCR HSC70-4 R qPCR HSC70-5 F qPCR HSC70-5 R qPCR HSP22 F qPCR HSP67Ba R qPCR CG4461 R qPCR CG7409 R qPCR CG13133 R qPCR CG14207 R qPCR HSP70Aa F qPCR HSP70Aa R qPCR RpL32 F qPCR RpL32 R P y l i m a f

e n e Table S3 Primers used for qPCR G sHSP HSP70 General

99 Chapter 5

Table S5 Primers used for the generation of dsRNA and specificities of the dsRNA sequences On-target HSP off-target Primer name Primer sequence (5'-3') Template Region siRNA's Gene Mismatch or fly 0 1 2

HsMAZ-F GTGGCGTCTAGATTCCTACAAG NC_000016 REGION: 2310-3028 0 none HsMAZ-F AGGTATGCTGCCGTGGTGAA 29725356..29730005 RNAi Hsp70x F AAGCATCGCCAGCGAATAAC CG31366-RA 2088-2408 303* HSP68 3 22 14 RNAi Hsp70x R TCCAGAGTAGCCTCCAAATC RNAi Hsc70-1 F TGTTCGCCGAGAACGGAAAG CG8937-PC 726-1109 366 HSP68 0 9 23 RNAi Hsc70-1 R TTGCCTTCACCTCGAAGATG RNAi Hsc70-2 F CCTTCGACGTCTCCGTACTG CG7756-PA 807-1193 369 HSP70 1 3 9 RNAi Hsc70-2 R GAATGGCCTTCTTGTCCATC HSC70-4 0 7 27 RNAi Hsc70-4 F TGAACGTGCTGCGTATCATC CG4264-PA 655-1059 387 HSC70-3 5 18 29 RNAi Hsc70-4 R AACGGGCACGAGTAATCGAG HSP70 4 26 37 Transgenic lines UAS-HSP67BC-RNAi Obtained from VDRC 322 HSP27 0 1 6 UAS-CG4461-RNAi Obtained from VDRC 284 none UAS-CG7409-RNAi Obtained from VDRC 306 none UAS-CG14207-RNAi Obtained from VDRC 340 none

* This pool of siRNA’s targets both the HSP70A and HSP70B members. On-target and off-target siRNA’s were determined using dsCheck located at http://dscheck.rnai.jp

Figure S1. The Drosophila family of small heat A The Drosophila small HSP family CG number Gene name Chromosome Locus GeneID shock proteins. (A) The Drosophila small HSP CG4460 HSP22 3L 67B 3772576 family consisting of eleven individual members. CG4463 HSP23 3L 67B 39077 CG4183 HSP26 3L 67B 39075 (B) Bootstrap analysis (bootstrap value: 10000) CG4466 HSP27 3L 67B 39078 shows three main groups within the small HSP CG4167 HSP67BA 3L 67B 39076 CG4190 HSP67BC 3L 67B 39071 family. (I) HSP22, HSP23, HSP26 and HSP27; CG4533 L(2)EFL 2R 59F 37744 II) CG7409, CG4461, L(2)EFL and CG14207; III) CG4461 - 3L 67B 39074 CG7409 - 3L 66A 38870 HSP67BC and CG13133. (C) Molecular cloning CG13133 - 2L 31A 34342 of the Drosophila small HSP plasmid library was CG14207 - X 18D 32955 performed using the pAc5.1 vector and deriva- Hsp23 tives thereof containing either no tag, a V5 tag or B 9973 A EGFP tag. 8080 Hsp26 Hsp27 3737 Hsp22 2786 CG7409 9478 B 6127 CG4461 L(2)EFL 4470 CG14207 HSP67Bc 6018 C CG13133 HSP67Ba 0.1

C Multiple cloning Cloned genes Actin promoter site KpnI 1 HSP23 EcoRI 25 HSP26 EcoRV 35 HSP27 NotI 51 HSP67Ba XhoI 58 HSP67Bc pAc5.1 MluI 121 L(2)EFL 5379 bps PolyA PmeI 154 CG4461 CG7409 CG13133 Amp CG14207

V5 MCS ATG

eGFP MCS ATG

100 Small heat shock proteins & longevity

A Wildtype Atx3 overexpression B Figure S2. Scoring system for eye-degeneration. The efficacy of HSP67BC, CG4461, CG7409 and CG14207 on amel- iorating polyQ toxicity was determined

UAS-SCAtrQ78/+ gmr-GAL4/+ using the Ataxin-3 fly model (ATX3trQ78) 300x (55). (A) Cryo-electron microscopy clari- fies the nature of the SCA3 eye-degener-

50M 50M ation. Wildtype eyes show a symmetrical alignment of smooth hexagonal shaped ommatidia, whereas overexpression of ROUGH PATCHED SCA3-trQ78 leads to loss of the hexagonal structure and smooth surface. (B)

1500x The UAS-SCA3trQ78, without SCA3 ex-

gmr-GAL4 UAS-SC3AtrQ78/+ pression, shows normal eye morphology which is also the case for the gmr-GAL4 10M 10M SPECKLED COLLAPSED activator line. Gmr-GAL4 driven expression of SCA3trQ78 in the eyes results in a visible disturbed eye morphology and degeneration that can be categorized into several morphological stages: rough and the more severe degenerative phenotypes referred to as patched, speckles, collapsed. Pictures indicate representative examples of either phenotype. The effect of small HSP expression on eye degeneration was determined by variation in the amount of patched, speckles and collapsed eyes. The amount of patched, speckles and collapsed eyes were expressed relatively to the amount of rough eyes and set to 100% for the SCA3-Q74 expressing flies. Flies with either sHSP overexpression or dsRNA mediated sHSP downregulation were scaled accordingly.

A UASxGAL4 UASxGAL4 UASxGAL4 UASxGAL4

UAS-SCAtrQ78/+ gmr-GAL4/+ ROUGH PATCHED SPECKLED COLLAPSED B HSP67BC C CG4461 L1 L2 L3 L1 L2 L3 L4

L1 x SCA3 L2 x SCA3 L3 x SCA3 L1 x SCA3 L2 x SCA3 L3 x SCA3 L4 x SCA3

CG7409 CG14207 FigureD S3. Overview of eye phenotypes uponE overexpression of small HSPs. (A) The SCA3-trQ78 expressing flies L1show various L2degrees of eye-degeneration.L3 RepresentativeL1 L2 examples ofL3 eye morphologyL4 are shown for co-expression of (B) HSP67BC, (C) CG4461, (D) CG7409 or (E) CG14207. In each panel the top row indicated morphology without SC3-trQ78 expression (L(x)) while the bottom row indicates morphology with co-expression of small HSPs and SCA3-trQ78 (L(x) x SCA3).

L1 x SCA3 L2 x SCA3 L3 x SCA3 L1 x SCA3 L2 x SCA3 L3 x SCA3 L4 x SCA3 101 A UASxGAL4 UASxGAL4 UASxGAL4 UASxGAL4

UAS-SCAtrQ78/+ gmr-GAL4/+ ROUGH PATCHED SPECKLED COLLAPSED B HSP67BC C CG4461 L1 L2 L3 L1 L2 L3 L4

L1 x SCA3 L2 x SCA3 L3 x SCA3 L1 x SCA3 L2 x SCA3 L3 x SCA3 L4 x SCA3

Chapter 5

D CG7409 E CG14207 L1 L2 L3 L1 L2 L3 L4

L1 x SCA3 L2 x SCA3 L3 x SCA3 L1 x SCA3 L2 x SCA3 L3 x SCA3 L4 x SCA3

Figure S3. Continued.

A RNAi HSP67BC RNAi CG4461 C CG7409 D RNAi CG14207 RNAi#1 RNAi#2 B RNAi#1 RNAi#2 RNAi#1 RNAi#1 RNAi#2

SCA3 x SCA3 x SCA3 x SCA3 x SCA3 x SCA3 x SCA3 x RNAi#1 RNAi#2 RNAi#1 RNAi#2 RNAi#1 RNAi#1 RNAi#2

Figure S4. Overview of eye phenotypes upon down regulation of small HSPs. Representative examples of eye morphology are shown for dsRNA mediated downregulation of (A) HSP67BC, (B) CG4461, (C) CG7409 or (D) CG14207. In each panel the top row indicated morphology without SC3-trQ78 expression (RNAi#(x)) while the bottom row indicates morphology with downregulation of small HSPs and SCA3-trQ78 overexpression (RNAi#(x) x SCA3).

102 CHAPTER 6

General discussion & perspectives

Michel J Vos Chapter 6

Screen & basic concept

Cells constantly have to deal with acute stresses caused for instance by changes in temperature, chemical or physical stress as well as chronic stresses such as the continuous expression of mutated proteins caused by heritable diseases. Life has evolved a protective mechanism to sense and deal with (stress-induced) protein unfolding called the heat shock response. This response consists of the induced expression of several proteins (heat shock proteins or molecular chaperones) whose role it is to assist in protein folding, stabilization and degradation (1;2). The simultaneous transcription of these various genes is brought about by the heat shock transcription factor 1 (HSF-1) (3). In general, higher eukaryotes are found to express multiple HSF genes which regulate gene expression of non-heat shock proteins as well. Regulation and alternative splicing of these HSFs introduces an even more elaborate signaling pathway, linking them to development, tissue-specific expression and reproduction (4-7). Small heat shock proteins (small HSP) make up one node in the protein folding network. The ability of at least some members to de-oligomerize in response to stress allows them to instantly provide the cell with binding surfaces for non-native protein conformations without the need for (HSF-1 regulated) enhanced expression. Besides their normal level of expression, some of these members can in addition be elevated through HSF-1. In the preceding chapters several additional aspects of both human and Drosophila small heat shock protein families have been addressed. The described research offers a glimpse on the functional specialization and activities of small HSPs in these two species. Although separated by evolution, similar distributions in molecular activities can be observed. Both species contain members specialized in assisting refolding of (heat) denatured proteins (here referred to as “refolders”) while other members seem specialized in reducing aggregation of misfolded proteins, especially polyQ containing proteins (here referred to as “anti-Q”) (Figure 1a,b). Whereas some refolders can also exhibit (some) anti-Q activity (human HSPB4, Drosophila melanosgater (Dm) CG7409), the best anti-Q small HSP proteins did not show effects on refolding of heat-denatured substrates like luciferase (Fig- ure 1a,b). Clearly, anti-Q small HSP members (human HSPB7, HSPB8, HSPB9; DmHSP67Bc) have evolved to specifically deal with non-refoldable substrates like polyQ proteins and maybe other amyloid forming substrates. When combining this data from both species, it is clear that small HSPs can be divided in three main groups based on their activities (Figure 1c, Chapters 4 & 5). Group one consists of highly active refolders (HSPB1, HSPB5, DmHSP27, DmCG14207. DmL2EFL) with no activity against polyQ aggregation, group two consists of highly active anti-Q proteins (HSPB7, HSPB9, HSPB8, DmHSP67Bc,) where some members are linked to protein degradation. Group two members show no activity on luciferase refolding and group three contains small HSPs with combined activities (HSPB4, HSPB6, DmHSP23, DmHSP26, DmCG7409). As this functional distribution is observed within two different species, this likely reflects involvement of small HSPs in conserved parts of the protein homeostasis network with specialization of individual members on proteins folding, anti-aggregation and protein degradation. Whereas small HSPs appear to differ in their type of chaperone-like action, most of them have been reported to associate with cytoskeletal elements (Chapter 1, table 3). Intriguingly, several neuropathies and myopathies are linked to mutations in HSPB1, HSPB5 and HSPB8, affecting cell types and tissues that are highly dependent on cytoskeletal transport and functioning. In fact, neuronal and muscular tissues generally show a large diversity of HSPB expression, with relative high levels of HSPB5, HSPB6, HSPB7 and HSPB8 compared to other tissues like liver and lung (Figure 2). Apparently, neuronal and muscular tissues require a more elaborate small HSP subset, with devastating effects when this balance is disturbed by mutations in HSPB1, HSPB5 or HSPB8. For the other two HSPB members, HSPB6 and HSPB7, no link has yet been established between possible mutation and disease. To obtain data on HSPB6 and HSPB7, and

104 General discussion & perspectives

A Homo sapiens B Drosophila melanogaster C Anti poly-Q Refolding Hs Dm Anti-Q Anti-Q Refolding Refolding HSPB1 HSPB1 HSP23 HSPB5 HSPB2 HSP26 HSP27 CG14207 HSPB3 HSP27 Refolders HSPB4 HSP67BA L2EFL HSPB5 HSP67BC HSPB6 L2EFL Hs Dm HSPB7 CG4461 HSPB7 HSPB8 CG7409 HSPB9 HSP67BC HSPB9 CG13133 Anti-Q HSPB10 CG14207 HSPB8

Hs Dm HSPB4 HSP23 HSP26 CG7409 Combined HSPB6 Figure 1. Activities of Homo sapiens and Drosophila melanogaster small HSPs. Activity on either refolding of heat-denatured luciferase or on preventing polyQ aggregation was classified as high (100% grayscale) or low (25% grayscale) for either (A) Homo sapiens small HSPs or (B) Drosophila small HSPs. (C) Both sets of small HSPs were subsequently grouped into either refolders, anti-Q or small HSPs with combined activities. other less studied HSPB members, we analyzed their subcellular localization and consulted databases containing data on tissue expression and interactions between individual members. This revealed that especially HSPB7 has a remarkable localization pattern and is one of the most active HSPB members in preventing polyQ aggregation (Chapter 4)

Different faces of HSPB7

HSPB7 and splicing speckles We identified HSPB7 as a cytoplasmic and nuclear protein with constitutive localization to SC35 splicing speckles. Within the nucleus, SC35 speckles are part of the highly organized nuclear space and together with promyelocytic leukaemia (PML) nuclear bodies (8) and Cajal bodies (CB) form the small nuclear domains (9). Speckles are rich in splicing-related proteins, linking them to RNA processing and splicing (10). In addition to HSPB7, also HSPB1 and HSPB5 have been reported to associate with SC35 speckles under either stressful conditions or in a constitutive manner respectively (11-14). As splicing is extremely sensitive to heat shock (15;16), the involvement of heat shock proteins in safeguarding this cellular process seems obvious. Indeed, yeast HSP104 and HSP70 have been reported to enhance recovery of splicing after a heat shock (17). While HSPB1 has been directly linked to restoration of splicing activity after stress, a role of HSPB5 in splicing was not found (18). In addition to the faster recovery of splicing after a heat shock, HSPB1, yeast HSP104 and yeast HSP70 are reported to be efficient chaperones in preventing protein aggregation and confer refolding of non-native substrates (19-21). This suggests that these proteins support post-heat shock splicing activity via protection of other speckle-associated pro-

105 Chapter 6

HSPB8 Figure 2. Relative expression levels A HSPB7 B HSPB8 HSPB1 of the human HSPB family in different tissues. Expressed sequence tag HSPB1 data were collected from the NCBI da- HSPB7 tabase and HSPB mRNA levels were Muscle HSPB6 Brain HSPB2 calculated relative to HSPB1 levels for HSPB6 either (A) muscular tissue, (B) brain tis-

HSPB2 sue, (C) liver tissue and (D) lung tissue.

HSPB5 HSPB5

HSPB9 HSPB7 HSPB8 HSPB8 C HSPB5 D HSPB7 HSPB6 HSPB2 HSPB5

Liver Lung HSPB3

HSPB2 HSPB1

HSPB1 teins against heat-induced unfolding. Also HSPB1 and HSPB5 are highly active refolders in the cytoplasm and nucleus (Chapter 3, Chapter 4, Figure 1a), which might as well be extendable to proteins in SC35 speckles. HSPB7, however, is unable to assist in the refolding of heat- denatured luciferase in both cellular compartments (Chapter 3, Chapter 4, Figure 1a) which suggest that this HSPB protein functions in a different way than HSPB1 and HSPB5. These different molecular activities could suggest an intricate regulatory mechanism whereby HSPB1 is constitutively associated with protein folding in the cytoplasm and only during proteotoxic stress enters the nucleus to both restrict protein unfolding and support splicing activity. HSPB5, on the other hand, does not only seem to serve a constitutive role in protein folding in the cytoplasm but in addition may be involved in protein folding in the nucleus (Chapter 3 &4) and possibly at SC35 splicing speckles. HSPB7, not active in the process of protein folding in the cytoplasm or nucleus, seems to fulfill a constitutive role at splicing speckles unrelated to protein folding. Albeit not experimentally verified yet, HSPB7 could have a novel function in the splicing process as is discussed below The N-terminus of HSPB7 is required and sufficient for localization to SC35 speckles. Nota- bly, the N-terminus contains a serine-rich stretch which could resemble an arginine/serine-rich (RS)-like domain. The RS domain is present in various SC35 speckle resident proteins and can play a role in protein-protein interaction as well as in RNA binding (22;23). The RS-domain is required for interaction with other splicing factors that also contain an RS domain. This leads to the formation of high molecular weight complexes in a phosphorylation-dependent manner (24). It is tempting to suggest that RS-domain containing proteins not only allow association of SC35 speckles with RNA but by binding also serve a chaperone-like function in preventing single stranded RNA from forming stable folds. Furthermore, the apparent similarity between SR-proteins and HSPB members concerning phosphorylation dependent oligomer formation is striking and as such, a fraction of HSPB7 might form mixed oligomers with other SR-proteins. Our data suggest, however, that HSPB7 does not form large oligomeric complexes like HSPB1 and HSPB5 nor forms mixed oligomers of HSPB7 with other HSPB proteins in living cells. This implies that at speckles, HSPB7 functionality is not determined by the association with either HSPB1 (during stress) or HSPB5.

106 General discussion & perspectives

RNA chaperones So, what could be the functional relevance of the association of HSPB members with SC35 speckles. As stated above, speckles are linked to RNA processing and splicing (10) and in addition speckles have been reported to localize to genomic regions with high transcriptional activity (25;26) which was shown to be independent on the presence of introns (27). Thus, splicing speckles seem also involved in transcription where splicing activity is not required. Here, SC35 splicing speckles and their associated proteins, could facilitate additional biological activity concerning pre-mRNA synthesis, stabilization or transport. In view of this, HSPB1 might associate with speckles during heat shock to induce binding specificity of the splicing machinery for (heat shock induced) destabilization of RNAs. Sinsimer and colleagues showed that HSPB1 can bind destabilizing AU-rich elements present in RNA (28). Whether HSPB7 can bind RNA molecules as well has yet to be determined. However, not all pre-mRNAs associate with SC35 speckles (29) implying either structural requirements present in RNA to associate with speckles or specificity of SC35 speckles defined by its associated proteins. Lin and colleagues recently found that SC35 might act as a single-strand RNA binding protein to facilitate transcriptional elongation (30). In addition, binding of speckle components to RNA might provide stabilization and prevent RNA misfolding which could potentially slow down the splicing- and translation process. In this way, the cell ensures proper trafficking from transcription to translation. This so called RNA chaperone activity has been described for several other proteins (31;32). They are able to promote RNA folding by resolving the complex tertiary structure of RNAs or prevent their formation (33). Such RNA chaperones include the stress-granules localized fragile X mental retardation protein FMRP (34;35) and hnRNP A1 (36;37), the DNA associated YB-1 (38;39) and DDX15 which localizes to nuclear speckles (40). Furthermore, RNA misfolding and the formation of RNA aggregates could play a role in the development of disease. For instance, myotonic dystrophy (MD) has been shown to be caused by a CTG triplet expansion in the 3’ untranslated region (3’ UTR) of a gene encoding a serine- threonine protein kinase (DMPK) (41). Proposed mechanisms leading to MD pathogenesis include changes in mRNA metabolism and transport of DMPK mRNA and mRNA aggregation into foci (42-44). RNA foci are also observed for Huntington disease-like 2 (HDL2), which shares the pathological hallmarks of Huntington disease (HD) (45;46). In this case, a CTG/CAG extended repeat, is found in a variably spliced exon of the junctophilin-3 gene, In both cases, other mRNA species might co-aggregate in RNA-containing foci leading to a disruption of cellular homeostasis and a concurrent sensitivity for internal and external stressors. It is thus likely that SC35 speckles not only constitute the splicing machinery but in addition also contain components for RNA chaperoning needed to loosen tertiary RNA structures. The specific and constitutive localization of HSPB7 to SC35 speckles, as we demonstratet in Chapter 3, might reflect an additional biological function of this small heat shock protein involved in RNA chaperoning.

HSPB7 and anti-Q activity Certain HSPB proteins have already been described before to be effective in preventing protein aggregation (47-49) probably by providing an interaction surface for hydrophobic protein stretches (48). Mutations in both HSPB5 and HSPB8 are causative for myopathy and neuropathy respectively (50-53). Furthermore, HSPB members have been reported to be upregulated in protein folding-related diseases (54;55). Thus it seems that the activity of these HSPB proteins is of key importance to guarantee a healthy cellular state and can be used to deal with high levels of proteotoxic stress. To identify the most potent protectors against misfolded proteins, we compared the human HSPB members for their (differential) capacities in anti-aggregation and re- naturation using either amyloidal aggregated polyglutamine (polyQ) proteins or heat-denatured luciferase (Chapter 4). Using several polyglutamine lengths, which are correlated to aggregation rates (56), we found, HSPB6, HSPB8, HSPB9 and especially HSPB7 as the most active

107 Chapter 6

Figure 3. Nuclear targeting of HSPB7 reduces cytosolic A FITC DAPI Overlay polyQ aggregation. (A) HEK293 cells were transfected with WT, cytosolic (NES) or nuclear (NLS) HSPB7 followed by immunofluoresecence. (B) Co-expression of either wildtype HSPB7 or nuclear localized (NLS) HSPB7 with cytosolic or

Wildtype nuclear EGFP-HttQ74 reduced polyQ aggregation in both cellular compartments. See Appendix 4 for colour print.

in preventing aggregation of polyQ proteins. At least

NES HSPB6, HSPB7 and HSPB8 differ from the classical HSPB members HSPB1, HSPB4 and HSPB5 in that they do not form large oligomeric complexes in living cells (Chapter 1, Chapter 4). Oligomer dynamics have been suggested to be crucial for HSPB members to

NLS promote chaperone activity. This makes it puzzling why HSPB7, the most active HSPB member towards long polyQ stretches, apparently does not oligomer- B ize in living cells. Whereas the precise mode of ac- mRFPruby tion remains to be elucidated, HSPB7 might prevent HSPB7 polyQ aggregation in both the cytoplasm and nucleus by shielding the polyQ stretch as has been described NES-HSPB7 for yeast HSP70 and HSP40 resulting in detergent NLS-HSPB7 soluble inclusions (57). We were however unable to Cytosolic Q74 Nuclear Q74 co-immunoprecipitate polyQ expanded Huntingtin and HSPB7 (data not shown) which could suggest an effect of HSPB7 on preventing polyQ aggregation independent from direct substrate interaction. Support for an indirect action of HSPB7 was further obtained by targeting HSPB7 to either the cytoplasm or nucleus using a nuclear export signal (NES) or nuclear localization signal (NLS) respectively (Figure 3a). Interestingly, targeting of HSPB7 to the nucleus resulted in anti-Q activity in both cellular compartments (Figure 3b). This would suggest that a biological activity of HSPB7 within the cells nucleus is sufficient to prevent polyQ aggregation in the cytoplasm. As discussed in the previous paragraph, triplet extensions in mRNA can potentially form mRNA foci and deregulate normal mRNA processing. As HSPB7 is localized to SC35 speckles, and thus has a potential role in mRNA biology, it could protect triplet expanded mRNA species from entangling with other RNAs and thereby maintaining RNA and protein homeostasis. In this way, key cellular systems would be maintained which are able to deal with polyQ expanded proteins and accompanying proteotoxic stresses. As HSPB7 was effective in cells lacking a normal heat shock response (Chapter 4) these key systems seem not to involve HSF-1 dependent chaperones. Also, as HSPB7 did not lead to detectable changes in polyQ protein expression, its putative role in RNA homeostasis also did not seem due to reduced RNA processing of the CAG-repeat containing messengers. One option is that its action at the RNA level may somehow impact autophagy, as the effects of HSPB7 were significantly reduced in autophagy deficient cells. Clearly, much more biochemical evidence is needed to further unravel whether any of these speculations are valid.

HSPB7, one function or moonlighting? Another question that arises from all our observation is whether the described activities of HSPB7 are functionally linked, or whether simply reflect different functions of HSPB7 depending on subcellular localization or/and available substrates. Multiple functions executed by one protein is a common phenomenon in biology known as gene sharing or moonlighting (58-61) which effectively extends the functional repertoire of the proteome. Protein function can change

108 General discussion & perspectives as a result of differences in sub-cellular localization. These include amongst others, oligomeric state or concentration of a substrate or co-factor and expression within a developmental context (58), For instance, localization-induced functional changes were reported for the Salmonella typhimurium PutA protein (62;63). Association with the plasma membrane enables proline dehydrogenase and pyrroline-5-carboxylate dehydrogenase activity while PutA functions as a transcriptional repressor lacking enzymatic activity when present in the cytoplasm. It can thus not be ruled out that HSPB7 performs anti-aggregation functions in both the cytoplasm and the nucleus while a fraction of HSPB7 associates with SC35 speckles to perform other functions related to, for instance, RNA processing. This is currently under investigation.

Small HSPs & protein homeostasis

As stated before, small HSPs can be divided into refolders and anti-Q proteins which both feed into cellular protein homeostasis. Whereas refolders generally show the formation of oligomers and high molecular weight complexes containing bound substrate, anti-Q small HSPs seem to exist mainly as dimers. Based on this difference, two models can be extrapolated. As dimers, the anti-Q HSPB proteins indirectly affect polyQ aggregation by either RNA processing or translational repression. They thus act in a classical chaperone independent manner. Classical oligomeric HSPB proteins on the other hand act in a HSPA-dependent way in substrate transfer and release to refold denatured substrate. This latter function would allows efficient refolding activity but low anti-Q activity. In the other model, dimeric anti-Q HSPB proteins do associate with substrates (Figure 4), but in a high affinity manner. This makes HSPA transfer unlikely. HSPA refolding cycles thus do not occur, decreasing classical chaperone activity of these members. The oligomeric HSPB members on the other hand have a lower affinity to their substrates which allows transfer to the HSPA refolding system. In either case, chaperone activity towards either polyQ stretches, unfolded proteins or aggregation-prone RNA will result in protein homeostasis.

Different chaperone pathways Whereas both refolders and anti-Q proteins are able to bind to the cytoskeleton (Figure 4) , processing of non-native proteins appears to follow different pathways: refolders collaborate with HSPA members to refold substrate or target them for degradation through the proteasome  while anti-Q small HSPs reduce polyQ aggregation followed by processing to the cellular degradation pathways  either via the proteasome (HSPB9, Vos, Zijlstra et al. unpublished results) or via autophagy (HSPB8 and maybe HSPB7;64). Failure of direct processing of either substrate can result in transport along the cytoskeleton towards the microtubule organizing center  to form aggresomes that later may be further processed by autophagic degregation. Whereas we can separate HSPB members functionally in this manner, this may not be the entire story and part of the processing of unfolded substrates may in fact depend on the substrates themselves (rather than the chaperone per se). HSPB5 for instance, is able to reduce amyloid fibril formation of certain substrates in vitro but was unable to reduce polyQ aggregation in vivo in our screen. Furthermore, via SC35 speckle association  some HSPB members (HSPB1, HSPB5, HSPB7) may also affect protein homeostasis through additional biologic functionalities like RNA processing. The possibility of mixed small HSP oligomer formation could further modify substrate specificity and small HSP activity. Nevertheless, both patways described above (refolding & anti-Q) can result in protein homeostasis and provide longevity on an organismal level.

109 Chapter 6

small HSP

1 Substrate

Refolders Anti-Q Amyloid substrate

SC35 speckle

Autophagic vesicle and lysosome

RNA c a Proteasone

b 2 b

3 5

d 4

Figure 4. Model of small HSP activities based on the human HSPB family. Small HSPs can be divided into either refolders or anti-Q proteins with a main difference in oligomeric structure. Both types of small HSPs can however associate with the cytoskeleton  for either stabilization, cytoskeletal dynamics or involvement in cytoskeletal transport. Deoligomerization (a) of refolders allows some small HSP members (HSPB1, HSPB5) to enter the nucleus and (b) associate, like anti-Q HSPB7, with SC35 splicing speckles . During protein unfolding (c), refolders bind the non-native substrate and associate into large substrate- HSPB complexes . Further processing can either result in proteasomal degradation or transfer to HSPA members (d) for substrate refolding. Anti-Q HSPB members likely also associate with the non-foldable substrate and subsequently target the substrate for degradation by using proteasomal or autophagic clearance . Failure of substrate refolding or clearance results in cytoskeletal transport of the non-native substrate to the microtule organizing center for subsequent degradation . Both single activities (refolding, anti-Q) or a combination of both can support protein homeostasis and provide longevity in Drosophila. See Appendix 4

110 General discussion & perspectives

From protein homeostasis to longevity The protective effect of small heat shock proteins in cells on protein aggregation and protein unfolding can be extended towards organisms as well. Here, support of the protein degradation and folding pathways can result in protective effects during aging, leading to an increase in lifespan. Aging has long been attributed to damage caused by free radicals resulting in oxidative damage to the proteome, genome and other cellular components (65). This theory was further extended by including production of reactive oxygen species (ROS) by mitochondria as a major source of oxidative damage (66). Besides this more general insult, genome instability has also been proposed to contribute to the aging process (67). Genome instability is, in part, also the result of oxidative damage to DNA but is in addition burdened by mutations in proteins required for DNA replication, DNA repair and telomere shortening and replication stress (68). Although reducing oxidative stress by overexpression of Cu/Zn-superoxide dismutase resulted in an extended lifespan (69), little evidence has been provided that boosting DNA damage repair pathways can enhance longevity, except for one report in which overexpression of the ATM gene, that is nu- tated in Ataxia Telangiectasia, had a minor effect on life span (70). However, in the experiments it was not demonstrated that this was due to improved DNA repair. Furthermore, aging is paralleled by an increase in damaged and misfolded proteins leading to cellular dysfunction. Although cells are capable to suppress proteotoxic stress via asymmetrical division of protein damage during mitosis (71) and via cellular protein quality-control systems, including small HSPs, these pathways fail at saturating levels of protein damage (Figure 5, left panel). Extending the lifespan of an organism can be achieved by a simple intervention: reducing food- intake which is referred to as caloric restriction (CR) (72-74). CR reduces metabolism and hereby ROS production is reduced, partially explaining the lifespan effects. CR leads to activation of insulin-like signaling pathway in which the chromatin silencing Sirtuins (SIRT) (75;76) play a central role. (75;77). Intriguingly, whereas SIRT1 deacetylates and inhibits key transcription factors like nuclear factor-κB (NF-κB) (78), p53 (79) and FOXO (forkhead) transcription factors (80),CR induced longevity has also been associated with sustained activitation of HSF-1 (81;82), the transcription factor regulating the stress-inducible heat shock response and is linked to overexpression of HSPs (83). In parallel, such manipulations reduced neurodegeneration as caused Figure 5. Aging, protein homeostasis Onset of protein folding disaeses and irriversible cellular damage and small HSPs. During aging, protein Refoldable substrates damage increases with a concurrent Non-foldable substrates decline of the efficiency of safeguarding protein homeostasis. In time, the amount of non-native proteins (proteotoxic stress) increases which can be divided into either refoldable substrates or non- foldabel substrates. Hereby, the healthy state of the cell slowly shifts to a condition

Healthy state HSP67BC favoring protein folding diseases,cellular damage and aging. Whereas reducing oxidative stress affects several cellular processes, leading to lower levels of proteotoxic stress, small HSPs overexpres- Cumulative proteome damage sion leads to a specific decline of non- Preventing oxidative stress native proteins directly affecting protein homeostasis. Increasing either refolding capacity (CG14207) or anti-Q capacity CG14207 (HSP76BC) directly reduces proteotoxic stress levels thereby delaying the onset Age Interventions of protein folding diseases and aging.

111 Chapter 6 by protein misfolding diseases (84-87). These data suggest that the heat-shock response is an intrinsic system that cells can upregulate to provide the organism with a more robust state of protein homeostasis. Yet, with age these systems apparently fail and protein folding related problems arise. So, boosting the entire system or its essential components, restores protein homeostasis with no or little side effects on cellular functioning and promote healthy ageing. In this thesis, we specifically focused on the role small HSPs may play in aging and asked what type of function (refolding or anti-Q) might be more relevant in boosting longevity. Hereto, we used Drosophila melanogaster and first determined whether, like in humans, members of the Drosophila small HSP family differ in molecular activity. Indeed, we were able to select a set of small HSPs which differ in their activities either being ”refolders” or “antiQ” proteins (Figure 1b). Interestingly, both boosting of refolding activity (CG14207) or anti-Q activity (HSP67BC) provided longevity (Figure 5, right panel). Non-native proteins thus seem to be a major determinant of aging. Improvement of protein homeostasis does not seem to depend on a single cellular process, rather it can be safeguarded by both refolding and anti-aggregation activity on the pool of non-native proteins, effectively decreasing the total proteotoxic burden. This is especially important in view of HSP70 dependent refolding activity of small heat shock proteins such as DmCG14207. HSP70 induction has been suggested to deregulate signalling pathways (88) and correlates with the development of malignancies (89). Boosting protein homeostasis through HSP70 activity might thus result in adverse side effects. Boosting anti-aggregation activity in a non-HSP70 dependent manner (DmHSP67BC), on the other hand, would result in longevity without the risk of HSP70 associated side effects. Thus using HSPB6, HSPB7, HSPB9 or DmH- SP67BC (Figure 1c) to boost protein homeostasis seems a more safer choice.

Perspectives

The overall picture emerging from this thesis is the link between small heat shock proteins and key cellular functions related to protein homeostasis. Apart from the role of HSPB1, HSPB4, HSPB5 and Drosophila HSP23, HSP26, HSP27, CG7409 and CG14207 in refolding of denatured proteins, certain members are tailored for additional functions. While some members seemingly combine refolding and anti-aggregation, HSPB7, HSPB8, HSPB9 and Drosophila HSP67BC are specialized in clearance of protein aggregates depending on the length of the glutamine stretch. The human anti-Q small HSPs generally show different mechanisms to prevent polyQ aggregation. One possible theory for the functionality of HSPB7 could be related to processing or chaperoning of RNA species in the presence of aggregation-prone triplet-expanded mRNA. Both DMPK and HD-2, which are caused by triplet expansion of the corresponding genes like HD, are also associated with nuclear RNA aggregates or foci. Both of these disease-related mRNA foci could serve as a tool to assess the role of HSPB7 in chaperoning RNA measured as a reduction in foci formation. One functional implication of RNA foci formation is a negative effect on protein synthesis of the affected genes. Here, the luciferase reporter system could serve as an excellent tool to acquire data on protein synthesis levels in the absence or presence of HSPB7 expression. Hereto, a luciferase construct would be needed containing a triplet expansion in the 3’ UTR, resembling expanded DMPK. After overexpression of HSPB7, reduction in foci formation and an increase of RNA chaperone activity could be directly linked to a functional improvement of cellular homeostasis measured as an increase in luciferase activity. One other hypothesis on HSPB7 activity is a potential role in RNA splicing modification, leading to, for instance, exon skipping or active inclusion of otherwise skipped exons. This mechanism could result in changes in protein function, possibly generating yet to be identified anti-Q pro-

112 General discussion & perspectives teins. A first step to confirm this hypothesis would be to test if HSPB7 is able to interact with RNA and if such attracts splicing factors. These questions could have a major impact on aging research as well. Whereas science has mainly focused on oxidative stress, DNA damage and proteotoxic stress in relation to aging, the role of a declining RNA modification system should also be taken into account. With the here described data on protein folding, anti-Q activity and aging, a possible role of RNA chaperoning in aging could make a worthwhile addition. In addition to HSPB7, several known RNA chaperones could be tested to determine their role in the network of cellular homeostasis and their contribution to aging.

References

1. Morimoto, R. I. (2008) Proteotoxic stress and inducible chaperone networks in neurodegenerative disease and aging, Genes Dev. 22, 1427-1438. 2. Walter, S. and Buchner, J. (2002) Molecular chaperones--cellular machines for protein folding, Angew. Chem. Int. Ed Engl. 41, 1098-1113. 3. Westerheide, S. D. and Morimoto, R. I. (2005) Heat shock response modulators as therapeutic tools for diseases of protein conformation, J. Biol. Chem. 280, 33097-33100. 4. Thompson, S., Croft, N. J., Sotiriou, A., Piggins, H. D. and others (2008) Neurospora crassa heat shock factor-1 is an essential gene; a second heat shock factor-like gene, hsf2, is required for asexual spore formation, Eukaryot. Cell. 5. Roccisana, J. L., Kawanabe, N., Kajiya, H., Koide, M. and others (2004) Functional role for heat shock factors in the transcriptional regulation of human RANK ligand gene expression in stromal/osteoblast cells, J. Biol. Chem. 279, 10500-10507. 6. Fujimoto, M., Izu, H., Seki, K., Fukuda, K. and others (2004) HSF4 is required for normal cell growth and differentiation during mouse lens development, EMBO J. 23, 4297-4306. 7. Morange, M. (2006) HSFs in development, Handb. Exp. Pharmacol. 153-169. 8. Dellaire, G. and Bazett-Jones, D. P. (2004) PML nuclear bodies: dynamic sensors of DNA damage and cellular stress, Bioessays 26, 963-977. 9. Handwerger, K. E. and Gall, J. G. (2006) Subnuclear organelles: new insights into form and function, Trends Cell Biol. 16, 19-26. 10. Lamond, A. I. and Spector, D. L. (2003) Nuclear speckles: a model for nuclear organelles, Nat. Rev. Mol. Cell Biol. 4, 605-612. 11. Adhikari, A. S., Sridhar, R. K., Rangaraj, N., Parnaik, V. K. and others (2004) Heat stress-induced localization of small heat shock proteins in mouse myoblasts: intranuclear lamin A/C speckles as target for alphaB-crystallin and Hsp25, Exp. Cell Res. 299, 393-403. 12. Bryantsev, A. L., Chechenova, M. B., and Shelden, E. A. (2007) Recruitment of phosphorylated small heat shock protein Hsp27 to nuclear speckles without stress, Exp. Cell Res. 313, 195-209. 13. den Engelsman, J., Bennink, E. J., Doerwald, L., Onnekink, C. and others (2004) Mimicking phosphorylation of the small heat-shock protein alphaB-crystallin recruits the F-box protein FBX4 to nuclear SC35 speckles, Eur. J. Biochem. 271, 4195-4203. 14. Michaud, S., Lavoie, S., Guimond, M. O., and Tanguay, R. M. (2008) The nuclear localization of Drosophila Hsp27 is dependent on a monopartite arginine-rich NLS and is uncoupled from its association to nuclear speckles, Biochim. Biophys. Acta 1783, 1200-1210. 15. Yost, H. J. and Lindquist, S. (1986) RNA splicing is interrupted by heat shock and is rescued by heat shock protein synthesis, Cell 45, 185-193. 16. Bond, U. (1988) Heat shock but not other stress inducers leads to the disruption of a sub-set of snRNPs and inhibition of in vitro splicing in HeLa cells, EMBO J. 7, 3509-3518. 17. Vogel, J. L., Parsell, D. A., and Lindquist, S. (1995) Heat-shock proteins Hsp104 and Hsp70 reactivate mRNA splicing after heat inactivation, Curr. Biol. 5, 306-317. 18. Marin-Vinader, L., Shin, C., Onnekink, C., Manley, J. L. and others (2006) Hsp27 enhances recovery of splicing as well as rephosphorylation of SRp38 after heat shock, Mol. Biol. Cell 17, 886-894. 19. Bryantsev, A. L., Kurchashova, S. Y., Golyshev, S. A., Polyakov, V. Y. and others (2007) Regulation of stress-induced intracellular sorting and chaperone function of Hsp27 (HspB1) in mammalian cells, Biochem. J. 407, 407-417. 20. Lu, Z. and Cyr, D. M. (1998) Protein folding activity of Hsp70 is modified differentially by the hsp40 co-chaperones Sis1 and Ydj1, J. Biol. Chem. 273, 27824-27830. 21. Haslbeck, M., Miess, A., Stromer, T., Walter, S. and others (2005) Disassembling Protein Aggregates in the Yeast Cytosol: THE COOPERATION OF HSP26 WITH SSA1 AND HSP104, J. Biol. Chem. 280, 23861-23868. 22. Ma, X. and He, F. (2003) Advances in the study of SR protein family, Genomics Proteomics. Bioinformatics. 1, 2-8. 23. Hertel, K. J. and Graveley, B. R. (2005) RS domains contact the pre-mRNA throughout spliceosome assembly,

113 Chapter 6

Trends Biochem. Sci. 30, 115-118. 24. Nikolakaki, E., Drosou, V., Sanidas, I., Peidis, P. and others (2008) RNA association or phosphorylation of the RS domain prevents aggregation of RS domain-containing proteins, Biochim. Biophys. Acta 1780, 214-225. 25. Xing, Y., Johnson, C. V., Moen, P. T., Jr., McNeil, J. A. and others (1995) Nonrandom gene organization: structural arrangements of specific pre-mRNA transcription and splicing with SC-35 domains, J. Cell Biol. 131, 1635-1647. 26. Misteli, T., Caceres, J. F., Clement, J. Q., Krainer, A. R. and others (1998) Serine phosphorylation of SR proteins is required for their recruitment to sites of transcription in vivo, J. Cell Biol. 143, 297-307. 27. Jolly, C., Vourc’h, C., Robert-Nicoud, M., and Morimoto, R. I. (1999) Intron-independent association of splicing factors with active genes, J. Cell Biol. 145, 1133-1143. 28. Sinsimer, K. S., Gratacos, F. M., Knapinska, A. M., Lu, J. and others (2008) Chaperone Hsp27, a novel subunit of AUF1 protein complexes, functions in AU-rich element-mediated mRNA decay, Mol. Cell Biol. 29. Smith, K. P., Moen, P. T., Wydner, K. L., Coleman, J. R. and others (1999) Processing of endogenous pre-mRNAs in association with SC-35 domains is gene specific, J. Cell Biol. 144, 617-629. 30. Lin, S., Coutinho-Mansfield, G., Wang, D., Pandit, S. and others (2008) The splicing factor SC35 has an active role in transcriptional elongation, Nat. Struct. Mol. Biol. 15, 819-826. 31. Rajkowitsch, L., Chen, D., Stampfl, S., Semrad, K. and others (2007) RNA chaperones, RNA annealers and RNA helicases, RNA. Biol. 4, 118-130. 32. Russell, R. (2008) RNA misfolding and the action of chaperones, Front Biosci. 13, 1-20. 33. Ivanyi-Nagy, R., Davidovic, L., Khandjian, E. W., and Darlix, J. L. (2005) Disordered RNA chaperone proteins: from functions to disease, Cell Mol. Life Sci. 62, 1409-1417. 34. Linder, B., Plottner, O., Kroiss, M., Hartmann, E. and others (2008) Tdrd3 is a novel stress granule-associated protein interacting with the Fragile-X syndrome protein FMRP, Hum. Mol. Genet. 35. Antar, L. N., Dictenberg, J. B., Plociniak, M., Afroz, R. and others (2005) Localization of FMRP-associated mRNA granules and requirement of microtubules for activity-dependent trafficking in hippocampal neurons, Genes Brain Behav. 4, 350-359. 36. Guil, S., Long, J. C., and Caceres, J. F. (2006) hnRNP A1 relocalization to the stress granules reflects a role in the stress response, Mol. Cell Biol. 26, 5744-5758. 37. Fridell, R. A., Truant, R., Thorne, L., Benson, R. E. and others (1997) Nuclear import of hnRNP A1 is mediated by a novel cellular cofactor related to karyopherin-beta, J. Cell Sci. 110 ( Pt 11), 1325-1331. 38. Vaiman, A. V., Stromskaya, T. P., Rybalkina, E. Y., Sorokin, A. V. and others (2006) Intracellular localization and content of YB-1 protein in multidrug resistant tumor cells, Biochemistry (Mosc. ) 71, 146-154. 39. Spitkovsky, D. D., Royer-Pokora, B., Delius, H., Kisseljov, F. and others (1992) Tissue restricted expression and chromosomal localization of the YB-1 gene encoding a 42 kD nuclear CCAAT binding protein, Nucleic Acids Res. 20, 797-803. 40. Fouraux, M. A., Kolkman, M. J., Van der, H. A., De Jong, A. S. and others (2002) The human La (SS-B) autoantigen interacts with DDX15/hPrp43, a putative DEAH-box RNA helicase, RNA. 8, 1428-1443. 41. Brook, J. D., McCurrach, M. E., Harley, H. G., Buckler, A. J. and others (1992) Molecular basis of myotonic dystrophy: expansion of a trinucleotide (CTG) repeat at the 3’ end of a transcript encoding a protein kinase family member, Cell 69, 385. 42. Taneja, K. L., McCurrach, M., Schalling, M., Housman, D. and others (1995) Foci of trinucleotide repeat transcripts in nuclei of myotonic dystrophy cells and tissues, J. Cell Biol. 128, 995-1002. 43. Sabouri, L. A., Mahadevan, M. S., Narang, M., Lee, D. S. and others (1993) Effect of the myotonic dystrophy (DM) mutation on mRNA levels of the DM gene, Nat. Genet. 4, 233-238. 44. Fu, Y. H., Friedman, D. L., Richards, S., Pearlman, J. A. and others (1993) Decreased expression of myotonin- protein kinase messenger RNA and protein in adult form of myotonic dystrophy, Science 260, 235-238. 45. Rudnicki, D. D., Pletnikova, O., Vonsattel, J. P., Ross, C. A. and others (2008) A comparison of huntington disease and huntington disease-like 2 neuropathology, J. Neuropathol. Exp. Neurol. 67, 366-374. 46. Rudnicki, D. D., Holmes, S. E., Lin, M. W., Thornton, C. A. and others (2007) Huntington’s disease--like 2 is associated with CUG repeat-containing RNA foci, Ann. Neurol. 61, 272-282. 47. Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993) Small heat shock proteins are molecular chaperones, J. Biol. Chem. 268, 1517-1520. 48. Lee, G. J., Roseman, A. M., Saibil, H. R., and Vierling, E. (1997) A small heat shock protein stably binds heat- denatured model substrates and can maintain a substrate in a folding-competent state, EMBO J. 16, 659-671. 49. Sun, Y. and MacRae, T. H. (2005) Small heat shock proteins: molecular structure and chaperone function, Cell Mol. Life Sci. 62, 2460-2476. 50. Selcen, D. and Engel, A. G. (2003) Myofibrillar myopathy caused by novel dominant negative alpha B-crystallin mutations, Ann. Neurol. 54, 804-810. 51. Bova, M. P., Yaron, O., Huang, Q., Ding, L. and others (1999) Mutation R120G in alphaB-crystallin, which is linked to a desmin-related myopathy, results in an irregular structure and defective chaperone-like function, Proc. Natl. Acad. Sci. U. S. A 96, 6137-6142. 52. Irobi, J., Van Impe, K., Seeman, P., Jordanova, A. and others (2004) Hot-spot residue in small heat-shock protein 22 causes distal motor neuropathy, Nat. Genet. 36, 597-601. 53. Kasakov, A. S., Bukach, O. V., Seit-Nebi, A. S., Marston, S. B. and others (2007) Effect of mutations in the beta5- beta7 loop on the structure and properties of human small heat shock protein HSP22 (HspB8, H11), FEBS J. 274,

114 General discussion & perspectives

5628-5642. 54. Santos, S. D., Magalhaes, J., and Saraiva, M. J. (2008) Activation of the heat shock response in familial amyloidotic polyneuropathy, J. Neuropathol. Exp. Neurol. 67, 449-455. 55. Wilhelmus, M. M., Otte-Holler, I., Wesseling, P., de Waal, R. M. and others (2006) Specific association of small heat shock proteins with the pathological hallmarks of Alzheimer’s disease brains, Neuropathol. Appl. Neurobiol. 32, 119- 130. 56. Krobitsch, S. and Lindquist, S. (2000) Aggregation of huntingtin in yeast varies with the length of the polyglutamine expansion and the expression of chaperone proteins, Proc. Natl. Acad. Sci. U. S. A 97, 1589-1594. 57. Muchowski, P. J., Schaffar, G., Sittler, A., Wanker, E. E. and others (2000) Hsp70 and hsp40 chaperones can inhibit self-assembly of polyglutamine proteins into amyloid-like fibrils, Proc. Natl. Acad. Sci. U. S. A 97, 7841-7846. 58. Jeffery, C. J. (1999) Moonlighting proteins, Trends Biochem. Sci. 24, 8-11. 59. Jeffery, C. J. (2003) Moonlighting proteins: old proteins learning new tricks, Trends Genet. 19, 415-417. 60. Jeffery, C. J. (2004) Molecular mechanisms for multitasking: recent crystal structures of moonlighting proteins, Curr. Opin. Struct. Biol. 14, 663-668. 61. Gancedo, C. and Flores, C. L. (2008) Moonlighting proteins in yeasts, Microbiol. Mol. Biol. Rev. 72, 197-210, table. 62. Ostrovsky de, S. P. and Maloy, S. (1993) PutA protein, a membrane-associated flavin dehydrogenase, acts as a redox-dependent transcriptional regulator, Proc. Natl. Acad. Sci. U. S. A 90, 4295-4298. 63. Muro-Pastor, A. M., Ostrovsky, P., and Maloy, S. (1997) Regulation of gene expression by repressor localization: biochemical evidence that membrane and DNA binding by the PutA protein are mutually exclusive, J. Bacteriol. 179, 2788-2791. 64. Carra, S., Seguin, S. J., and Landry, J. (2008) HspB8 and Bag3: a new chaperone complex targeting misfolded proteins to macroautophagy, Autophagy. 4, 237-239. 65. Harman, D. (1956) Aging: a theory based on free radical and radiation chemistry, J. Gerontol. 11, 298-300. 66. Harman, D. (1972) The biologic clock: the mitochondria?, J. Am. Geriatr. Soc. 20, 145-147. 67. McMurray, M. A. and Gottschling, D. E. (2003) An age-induced switch to a hyper-recombinational state, Science 301, 1908-1911. 68. Hyrien, O. (2000) Mechanisms and consequences of replication fork arrest, Biochimie 82, 5-17. 69. Sun, J. and Tower, J. (1999) FLP recombinase-mediated induction of Cu/Zn-superoxide dismutase transgene expression can extend the life span of adult Drosophila melanogaster flies, Mol. Cell Biol. 19, 216-228. 70. Symphorien, S. and Woodruff, R. C. (2003) Effect of DNA repair on aging of transgenic Drosophila melanogaster: I. mei-41 locus, J. Gerontol. A Biol. Sci. Med. Sci. 58, B782-B787. 71. Rujano, M. A., Bosveld, F., Salomons, F. A., Dijk, F. and others (2006) Polarised asymmetric inheritance of accumulated protein damage in higher eukaryotes, PLoS. Biol. 4, e417. 72. Guarente, L. and Picard, F. (2005) Calorie restriction--the SIR2 connection, Cell 120, 473-482. 73. Heilbronn, L. K. and Ravussin, E. (2003) Calorie restriction and aging: review of the literature and implications for studies in humans, Am. J. Clin. Nutr. 78, 361-369. 74. Koubova, J. and Guarente, L. (2003) How does calorie restriction work?, Genes Dev. 17, 313-321. 75. Tissenbaum, H. A. and Guarente, L. (2001) Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis elegans, Nature 410, 227-230. 76. Kaeberlein, M., McVey, M., and Guarente, L. (1999) The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two different mechanisms, Genes Dev. 13, 2570-2580. 77. Imai, S., Armstrong, C. M., Kaeberlein, M., and Guarente, L. (2000) Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase, Nature 403, 795-800. 78. Yeung, F., Hoberg, J. E., Ramsey, C. S., Keller, M. D. and others (2004) Modulation of NF-kappaB-dependent transcription and cell survival by the SIRT1 deacetylase, EMBO J. 23, 2369-2380. 79. Vaziri, H., Dessain, S. K., Ng, E. E., Imai, S. I. and others (2001) hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase, Cell 107, 149-159. 80. Motta, M. C., Divecha, N., Lemieux, M., Kamel, C. and others (2004) Mammalian SIRT1 represses forkhead transcription factors, Cell 116, 551-563. 81. Heydari, A. R., You, S., Takahashi, R., Gutsmann, A. and others (1996) Effect of caloric restriction on the expression of heat shock protein 70 and the activation of heat shock transcription factor 1, Dev. Genet. 18, 114-124. 82. Frier, B. and Locke, M. (2005) Preservation of heat stress induced myocardial hsp 72 in aged animals following caloric restriction, Exp. Gerontol. 40, 615-617. 83. McElwee, J. J., Schuster, E., Blanc, E., Thomas, J. H. and others (2004) Shared transcriptional signature in Caenorhabditis elegans Dauer larvae and long-lived daf-2 mutants implicates detoxification system in longevity assurance, J. Biol. Chem. 279, 44533-44543. 84. Steinkraus, K. A., Smith, E. D., Davis, C., Carr, D. and others (2008) Dietary restriction suppresses proteotoxicity and enhances longevity by an hsf-1-dependent mechanism in Caenorhabditis elegans, Aging Cell 7, 394-404. 85. Cohen, E., Bieschke, J., Perciavalle, R. M., Kelly, J. W. and others (2006) Opposing activities protect against age- onset proteotoxicity, Science 313, 1604-1610. 86. Hsu, A. L., Murphy, C. T., and Kenyon, C. (2003) Regulation of aging and age-related disease by DAF-16 and heat- shock factor, Science 300, 1142-1145. 87. Morley, J. F. and Morimoto, R. I. (2004) Regulation of longevity in Caenorhabditis elegans by heat shock factor and

115 Chapter 6

molecular chaperones, Mol. Biol. Cell 15, 657-664. 88. Williams, K. D., Helin, A. B., Posluszny, J., Roberts, S. P. and others (2003) Effect of heat shock, pretreatment and hsp70 copy number on wing development in Drosophila melanogaster, Mol. Ecol. 12, 1165-1177. 89. Powers, M. V., Clarke, P. A., and Workman, P. (2009) Death by chaperone: HSP90, HSP70 or both?, Cell Cycle 8.

116 APPENDI X1

Nederlandse samenvatting Appendix

Inleiding

Alle levende organismen zijn onderhevig aan biologische veroudering, wat gekenmerkt wordt door het minder goed functioneren van lichaam en geest. Dit is voornamelijk zichtbaar in dieren, maar ook kleinere organismen zoals gisten en bacteriën vertonen verouderingsverschijnselen. De basis van biologische veroudering is een toename van schade in de bouwstenen die samen het organisme vormen; de cellen. Dit omvat onder meer schade aan het cellulaire DNA (desoxyribonucleïnezuur, het erfelijke materiaal van de cel) en schade aan cellulaire eiwitten die celstructuur, transport, communicatie en regulatie voor hun rekening nemen. Hierdoor zijn cellen minder efficiënt in het uitvoeren van hun functies en kunnen ze afsterven. Naast de “normale” verouderingsverschijnselen zijn er specifieke aandoeningen die al vanaf de geboorte aanwezig zijn maar die zich pas op latere leeftijd openbaren. Het betreft hier aandoeningen die veroorzaakt worden door veranderingen of mutaties in het cellulaire DNA en die leiden tot problemen met de vouwing van het eiwit dat wordt gecodeerd door het gemuteerde gen Deze zogenaamde eiwitmisvouwingsziektes leiden veelal tot hersenziektes die tot uiting kunnen komen door het minder goed functioneren van cellen op latere leeftijd. Deze ziektes worden ondermeer gekenmerkt door het vroeger tot uiting komen in vergelijking met niet-erfelijke vormen van ouderdomszi ekten, zoals de ziekte van Alzheimer of de ziekte van Parkinson.

Eiwitvouwing

Zoals hiervoor al is beschreven, zijn eiwitten betrokken bij diverse cellulaire processen. Echter, een goed functionerend eiwit ontstaat niet spontaan. Eerst moet er een kopie worden gemaakt van het op het DNA gelegen gen, dat codeert voor het desbetreffende eiwit. Dit kopie wordt gemaakt in RNA (ribonucleïnezuur), wat vervolgens wordt afgelezen en vertaald in een lange keten van aminozuren; de bouwstenen van eiwitten. Vervolgens wordt deze lange keten gevouwen tot een drie dimensionaal eiwit, wat zijn functie in de cel kan vervullen. Het correct vouwen van een aminozuurketen zou veelal spontaan kunnen verlopen, maar cellen zitten zo vol met andere eiwitten en moleculen dat dit vouwingsproces begeleid moet worden door “bodyguards”, een groep eiwitten die moleculaire chaperonnes worden genoemd. Door interacties met deze chaperonnes wordt klontering van eiwitten met elkaar voorkomen, zodat uiteindelijk een actieve vorm van het eiwit verkregen wordt. Chaperonnes zijn echter niet alleen betrokken bij het vouwen van nieuwe eiwitten. Ook bestaande eiwitten kunnen lokaal ontvouwen, wat tot functieverlies kan leiden en ongewilde interacties met andere eiwitten te weeg kan brengen. Ook hier bieden chaperones hulp door ofwel deze beschadigde eiwitten te helpen hervouwen of door te helpen deze af te breken. Hierdoor wordt het gevaar van eiwitklontering voorkomen en blijft de cel gezond. Veroudering gaat ondermeer gepaard met mutaties in het cellulaire DNA, waardoor er veranderingen in de aminozuursequentie kunnen ontstaan. Hierdoor ontstaan er eiwitten die problemen kunnen ondervinden om tot een actieve structuur te vouwen. Veroudering gaat dus gepaard met een toename aan moeilijk te vouwen eiwitten en daarmee een verhoogde vraag naar chaperonne activiteit. Het is dit stadium waarin ook andere “verborgen” mutaties tot uiting kunnen komen in het functioneren van de cel. Het betreft hier eiwitten die reeds het gehele leven afhankelijk zijn geweest van extra chaperonne activiteit, maar door de extra vraag naar chaperonne activiteit door steeds meer beschadigde eiwitten, nu niet meer correct gevouwen kunnen worden. Daarnaast is, zoals al gezegd, het chaperonne systeem ook betrokken bij afbraak van niet goed functionerende eiwitten. Ook dit proces komt langzamerhand onder druk te staan wanneer een verouderend organisme meer aanspraak moet doen op het chaperonne systeem om het correct functioneren

118 Nederlandse samenvatting van zijn cellen te waarborgen. Tegelijkertijd kunnen er fouten ophopen in de DNA dat codeert voor de chaperonnes, waardoor er minder of minder goed functionerende chaperonnes worden gemaakt. De combinatie van dit alles leidt tot een verstoring in de balans van eiwit vouwing (eiwithomeostase) waardoor eiwitklontering in cellen ontstaat en cellen hun functie gaan verliezen en vervolgens afsterven.

HSPB chaperonnes

Een van de chaperonne families zijn de “small heat shock proteins” of HSPB chaperonnes. Deze eiwitten worden gekenmerkt door hun kleine massa in vergelijking met de andere chaperonne families. Verder is bekend dat een aantal HSPB eiwitten zeer effectief is in het voorkomen van ongewenste eiwit-eiwit interacties en het afschermen van ontvouwen gedeeltes van een aminozuurketen. In samenwerking met de HSPA chaperonnes kan vervolgens een ontvouwen eiwit worden hervouwen of worden afgebroken (Hoofdstuk 1). Het beschreven onderzoek in dit proefschrift richt zich op deze HSPB eiwitten en hun rol in eiwit vouwing, het voorkomen van eiwit aggregatie en positieve effecten op veroudering.

HSPB7 en splicing speckles Het DNA van de mens codeert voor elf van deze HSPB eiwitten (HSPB1-HSPB11), waarvan er maar een aantal goed zijn bestudeerd. Door te kijken waar deze HSPB eiwitten zich in een cel bevinden krijgen we een idee over eventuele functionele eigenschappen in relatie tot cel- compartimenten (Hoofdstuk 3). Ook verschil in lokalisatie tussen normale condities en condities van eiwitontvouwing kan informatie geven over functies. Een opmerkelijke HSPB eiwit, qua lokalisatie, is HSPB7. Deze chaperonne bleek onder niet-stress condities geassocieerd te zijn met een onderdeel van de celkern, waar tot nu toe alleen HSPB eiwitten voornamelijk werden aangetroffen tijdens omstandigheden van massale eiwitontvouwing. Deze zogenaamde SC35 splicing speckles fungeren als een verzamelplaats voor eiwitten betrokken in het moduleren van RNA, alvorens dit wordt vertaald in aminozuren. Het moduleren betreft het verwijderen van niet relevante gedeeltes van het RNA wat wordt aangeduid als splicing, of splitsen. Hoewel de rol van HSPB7 in SC35 splicing speckles en het splicing proces nog niet kon worden ontrafeld, ontdekten we wel welk gedeelte van HSPB7 nodig is voor lokalisatie naar deze SC35 structuren. Deze signaal sequentie, te vergelijken met een postcode, bevindt zich in het NH2-terminale gedeelte van het eiwit (wanneeer een eiwit wordt vertaald vanuit RNA, gebeurd dit van de NH2 uiteinde naar de COOH-uiteinde) (Hoofdstuk 3). en blijkt in geen enkel ander HSPB eiwit voor te komen, ook niet in HSPB eiwitten die bekend zijn voor hun SC35 splicing speckle lokalisatie tijdens eiwit ontvouwing. De signaal sequentie in HSPB7 blijkt uniek en voldoende te zin om ook andere eiwitten naar SC35 splicing speckles te transporteren. Deze sequentie lijkt verder niet essentieel voor de chaperonne-achtige functies van HSPB7 (Hoofdstuk 4) en lijkt daarom meer te dienen als middel om de HSPB7 functie te brengen naar plekken (SC35 speckles) waar deze functie gewenst is.

HSPB7 en anti-aggregatie Naast het analyseren van de lokalisatie van HSPB eiwitten, hebben we ook onderzoek gedaan naar hun individuele capaciteit in het chaperonneren van andere eiwitten. Hiervoor hebben we twee modeleiwitten (substraten) bekeken. Het eerste modeleiwit is luciferase, een biolumi- nescent (lichtgevend) eiwit afkomstig van de vuurvlieg. Dit eiwit ontvouwt snel wanneer cellen worden blootgesteld aan temperaturen boven 40ºC. Verder is bekend dat met hulp van sommige chaperonnes (vooral van de HSPA groep) luciferase kon hervouwen wanneer de temper- atuur weer wordt verlaagd naar 37ºC. Verassend genoeg bleek maar een aantal van de HSPB

119 Appendix eiwitten de hervouwing van luciferase te kunnen helpen. De meest actieve “hervouwers” waren HSPB1, HSPB4 en HSPB5, terwijl bijvoorbeeld HSPB7 en HSPB9 dit in het geheel niet konden (Hoofdstuk 4). Het tweede type substraat betreft eiwitten die betrokken zijn bij een van de vormen van erfelijke ouderdoms gerelateerde eiwitmisvouwingsziekten, de zogenaamde polyQ eiwitten. De term polyQ slaat op een herhaling van het aminozuur genaamd glutamine (afgekort tot Q) in een eiwit. Deze polyQ sequentie komt in veel eiwitten voor maar het is niet precies bekend wat de functie van deze sequentie is. In eiwitmisvouwingsziekten is deze polyQ sequentie sterk ver- lengd waardoor het eiwit heel snel klontert. Hierdoor gaan cellen eerst disfunctioneren, gevolgd door afsterving. De ziekte van Huntington (HD) en Machado-Joseph disease (MJD of SCA3) zijn bekende voorbeelden waar de polyQ expansie zich bevindt in respectievelijk het huntingtin of ataxin-3 eiwit. Wanneer we nu de HSPB eiwitten testten voor hun capaciteit in het reduceren van polyQ aggregatie, bleken HSPB9 en vooral HSPB7 zeer effectief te zijn (Hoofdstuk 4). HSPB1, HSPB4 en HSPB5 als hervouwers waren nauwelijks actief.. Het blijkt dus dat er binnen de HSPB groep van eiwitten specialisatie is opgetreden, waarbij sommige HSPB eiwitten begeleiders zijn van de (her)vouwing processen, terwijl andere HSPB eiwitten er voor zorgen dat niet-vouwbare eiwitten niet klonteren en worden afgebroken. Hoe dit laatste plaatsvindt is vooralsnog niet helder. Verder onderzoek naar de functie van de beste voorkomer van polyQ aggregatie, HSPB7, toonde aan dat dit niet direct gerelateerd lijkt te zijn aan de lokalisatie in de ee- rder genoemde SC35 splicing speckles. Wel bleek dat, waar HSPB-hervouwers samenwerken met leden van de HSPA familie van chaperones, de activiteit van HSPB7 HSPA-onafhankelijk was. Verder bleek dat de HSPB7 werking sterk afnam wanneer in cellen autofagie, een proces voor opruiming van beschadigde celonderdelen, werd uitgeschakeld terwijl het proces voor afbraak van individuele eiwitten (via het zogenaamde proteasome systeem) geen rol hierin lijkt te spelen.

HSPB eiwitten en veroudering

Door gebruik te maken van het feit dat kleine heat shock eiwitten verschillende chaperonne activiteiten laten zien (hervouwing en/of anti-aggregatie), kon worden onderzocht welk type eiwit-ondersteunende activiteit de grootste hulp kan leveren aan verouderende cellen. Het eiwithomeostase model veronderstelt dat beide types van activiteiten veroudering zou moeten beïnvloeden. Om dit model te testen maakten we gebruik van de fruitvlieg, omdat dit organisme relatief kort leeft en omdat volwassen fruitvliegen voornamelijk bestaan uit niet-delende cellen net als hersencellen, waarin de eiwitmisvouwingsziekten zich met name manifesteren. Allereerst werd onderzocht of, net als de humane HSPB eiwitten, ook de fruitvlieg familie van kleine heat shock eiwitten onderverdeeld kunnen worden in eiwitvouwers en anti-aggregatie eiwitten; deze diversiteit in activiteiten bleek inderdaad ook te bestaan in fruitvliegen (Hoofdstuk 5). Op basis hiervan hebben we vervolgens fruitvliegen genetisch aangepast zodat ze een van de kleine heat shock eiwitten met een van deze activiteiten verhoogd in hun weefsels aanwezig hebben. Beide types activiteiten bleken inderdaad in staat om de fruitvliegen langer te laten leven, hetgeen de belangrijke rol van eiwithomeostase in gezond ouder worden verder ondersteund.

120 APPENDI X2

Dankwoord Appendix

Dankwoord

Iets langer dan gedacht, maar dan is het toch eindelijk voltooid. Na bijna 5 jaar zou het schrijven van het dankwoord toch vergezeld moeten gaan met een soort van voldaan gevoel, in plaats van het “laatste hoofdstuk” gevoel. Wellicht dat deze omslag binnen twee pagina’s zal plaats- vinden, ik heb er in ieder geval het volste vertrouwen in.

De afgelopen jaren bij de “Straling & Stress” zijn een zeer waardevolle en leerzame ervaring geweest. Hoewel ik me in het begin toch vaak heb afgevraagd of een AIO project niet iets duidelijker en strakker omschreven had moeten zijn qua onderzoekslijnen, kom je er in de loop der jaren achter dat dit je juist meer ruimte geeft om te doen wat je zelf graag wilt (lees: met in achtneming van de projectomschrijving, uiteraard). Ik wil je hierbij dan ook bedanken Harrie, dat je me deze vrijheid gegeven hebt en daarnaast altijd energievol en met veel motivatie als begeleider hebt opgetreden. Deze vrijheid maakt een promotietraject veel leerzamer en zeker ook spannender.

Ody, ook jou wil ik bedanken voor je rol als co-promotor en je bijdragen bij de werkbesprekingen. Ik vond het altijd erg prettig wanneer je de werkbesprekingen bijwoonde, mede omdat je andere kennis bijdroeg en op een frisse, “niet-chaperone” manier naar de data kon kijken.

Beste Bart, je kunt wel zeggen dat we tegelijk zijn begonnen aan het Drosophila avontuur. Het leuke is dan ook dat individuele momenten van onbegrip en gebrek aan kennis op dit gebeid vaak een algemeen gevoel van “waar zijn we eigenlijk mee bezig” met zich meebracht, ten minste vanuit mijn oogpunt. Maar inmiddels ben jij de persoon binnen de HSP groep die het vliegengebeuren weet te coördineren. Daarnaast wil ik je bedanken voor de leuke en gezellige sfeer die je op de 5e verdieping wist (en weet) in te brengen. Met de leuke verhalen, uitspraken, oprechtheid en de vaak iets (veel) te zoete (anders smaakt het fris) glühwein en borstplaat, zorg je voor een vertrouwd en gezellig geheel.

Marianne, ik kijk met zeer veel plezier terug naar de afgelopen jaren waarin we steeds meer zijn gaan samenwerken. Je vrolijkheid, spontaniteit en gezelligheid, maar ook de serieuzere ge- sprekken, zal ik nooit vergeten. Het HSPB7 stuk zou nooit zo volledig zijn geworden zoals het nu is zonder het groot aantal experimenten dat jij hebt gedaan. Ik ben blij dat we ook nu nog vaak even contact hebben, zowel voor het bepreken van resultaten als voor het kletsen.

Ooit, bijna 5 jaar geleden, begon ik het AIO avontuur tegelijk met jou Jurre. Het heeft absoluut voordelen als je samen op een groot project begint met deels overlappende onderzoeksvragen. Als “cloning-freaks” wisten we in een paar jaar tijd de -80 vriezer aardig op te vullen met een enorme verzameling plasmiden. Hoewel niet alle constructen achteraf gezien even bruikbaar bleken, gaf het toch een heerlijk gevoel om weer het zoveelste construct af te maken. Misschien een tip om verstokte kloneerders verplicht op te laten nemen in een afkickkliniek? Bedankt voor de gezellige tijd!

Maria, what to say. I was warned beforehand that you were not that easy to approach and then I’m not even taking your almost legendary “latino temperamento” into account. Well, that all turned out to be just a myth. And after four years, I’m sure even you enjoyed Dutch boys in pink

122 Dankwoord outfits. It was very pleasant to be working with you in the HSP group. And I will never forget our fitness (gossiping) hours that got us into perfect physical and mental shape .

Karin, van mijn studenten ben jij toch wel degene die de grootste wetenschappelijke en zeker ook sociale bijdrage heeft geleverd. Nog nooit heb ik zoveel koffiepauzes gehad (Michel..... koffie doen???) als toen jij bij ons was. Bedankt voor de gezellige tijd en succes met je eigen promotieonderzoek.

Willie, Bianca, Annet, Jeanette, Martha, Marieken, Rita, Anne-Jan, Maarten (MAN!), Anil (re- spawn), Petra, Hette, Erwin, Floris, Isabelle (hugs), Maria van Waarde, Xia, Ria, Rob, Jielin, Sascha, Vaishali, Lei, Kasia, Serena, Erwin Wiegman, Marianne van der Zwaag, Hidde, Ab- raham, Reinier, Kevin en de stamcel collega’s, mede door jullie is het een onvergetelijke tijd geweest. Bedankt!

Lieve Angelique, ten slotte wil ik jou bedanken voor je steun en gezelligheid. Jij kan als geen ander een promotietraject relativeren, waardoor we vaak (in plaats van experimenten) leuke uitjes, vakanties en andere leuke dingen hebben ondernomen. Je bent een geweldige vriendin!

Prachtig dat promoveren!

Michel

123 Appendix

124 APPENDI X3

Curriculum vitae Appendix

Curriculum vitae

Personal details Surname: Vos Given name: Michel Jeroen Place of birth: Delfzijl, the Netherlands Date of birth: September 4th, 1979

Education 2004-2009 PhD University of Groningen / University Medical Center Groningen, The Netherlands 1999-2004 MSc Biology, cum laude University of Groningen / University Medical Center Groningen, The Netherlands

Traineeships 2003-2004 University of Groningen, The Netherlands Department of Eukaryotic Microbiology Project title: Analysis of a novel mitochondrial peroxin Supervision: Prof. Dr. Marten Veenhuis & Prof. Dr. Ida van der Klei 2003-2004 University of Groningen, The Netherlands Department of Radiation & Stress Cell Biology Project title: Post-transcriptional gene silencing and in vivo functions of heat shock proteins Supervision: Prof. Dr. Harm H Kampinga 2002-2003 University of Groningen, The Netherlands Department of Eukaryotic Microbiology Project title: Activation of Hansenula polymorpha alcohol oxidase in Saccha romyces cerevisiae Supervision: Prof. Dr. Marten Veenhuis & Prof. Dr. Ida van der Klei

Professional experience 2008-current University of Groningen / University Medical Center Groningen, The Netherlands Department of Membrane Cell Biology Postdoc Project title: Fibronectin & myelin repair in neurodegeneration. Supervision: Prof. Dr. Dick Hoekstra & Dr. Wia Baron 2004-2008 University of Groningen / University Medical Center Groningen, The Netherlands Department of Radiation & Stress Cell Biology PhD student Project title: Helping health ~ Boosting the cellular defense system Supervision: Prof. Dr. Harm H Kampinga & Prof. Dr. Ody CM Sibon

126 Curriculum vitae

Conferences & posters 2008 Molecular Chaperones and Stress Responses, Cold Spring Harbor, New York, USA 2007 Protein Folding & Chaperones in Biology, Tomar, Portugal 2006 Drosophila as a Model for Human Diseases, Barcelona, Spain

Publications • M.J.Vos*, J.Hageman*, S.Carra, and H.H.Kampinga, Structural and functional diversities between members of the human HSPB, HSPH, HSPA, and DNAJ chaperone families, Bio- chemistry 47 (2008) 7001. • M.J.Vos and H.H.Kampinga, A PCR amplification strategy for unrestricted generation of chimeric genes, Anal. Biochem. 380 (2008) 338. • L.E.Hightower, J.Hageman, M.J.Vos, H.Kubota, R.M.Tanguay, E.A.Bruford, M.E.Cheetham, B.Chen, and H.H.Kampinga, Guidelines for the nomenclature of the human heat shock proteins, Cell Stress. Chaperones. (2008). • J.Hageman, M.J.Vos, M.A.van Waarde, and H.H.Kampinga, Comparison of intra-organellar chaperone capacity for dealing with stress-induced protein unfolding, J. Biol. Chem. 282 (2007) 34334. • X.Yi, W.Lemstra, M.J.Vos, Y.Shang, H.H.Kampinga, T.T.Su, and O.C.Sibon, A long-term flow cytometry assay to analyze the role of specific genes of Drosophila melanogaster S2 cells in surviving genotoxic stress, Cytometry A (2008). • A.Muller-Taubenberger, M.J.Vos, A.Bottger, M.Lasi, F.P.Lai, M.Fischer, and K.Rottner, Monomeric red fluorescent protein variants used for imaging studies in different species, Eur. J. Cell Biol. 85 (2006) 1119. * equal contribution

127 Appendix

128 APPENDI X4

Colour figures Appendix

Chapter 1

Unfolding Aggregation Substrate Folded substrate HSPB

DNAJ

HSPA

Proteasome

CHIP - ATP

ATP ATP HSP110 + DNAJ + BAG + HSPBP1 + ADP ADP ADP

HIP + Figure 1. Schematic representation of HSP-mediated client processing. A folded substrate is unfolded upon a proteotoxic stress event. This unfolded substrate either aggregates or binds HSPs like HSPB, DNAJ, or HSPA. Both HSPB and DNAJ are thought to eventually hand over the substrate to the HSPA machine capable of binding and releasing the substrate. Released substrates that still expose hydrophobic patches to the exterior are bound again by HSPs, whereas substrates without such hydrophobic patches are not recognized. Substrates can also be targeted to the proteasome degradation system by HSPB1 and CHIP or other uncharacterized mechanisms.

130 Colour figures

Chapter 3

A EYFP-SP100 MYC-HSPB7 Merge C EYFP-ASF/SF2 MYC-HSPB7 Merge PML bodies PML

B EGFP-Coilin MYC-HSPB7 74Merge SC-35 170 MYC-HSPB7 Merge A HSPB7 topology N-terminus crystallin domain

NB7-NB1 N-terminus HSPB7 (1-71) HSPB1 (85-205) SR proteins NHSPB7 Cajal bodies EGFP-tagged

NB1-NB7 N-terminus HSPB1 (1-84) HSPB7 (72-170)

B HSPB7 NHSPB7 NB1-NB7 NB7-NB1 1 2 3 4 Endogenous

Figure 5. HSPB7 associates with SR protein-containing nuclear domains. HSPB7 was analyzed for colocalization with either (A) SP100; a marker for PML bodies, (B) Coilin; a marker for Cajal bodies or (C) SR proteins ASF/SF2 and SC-35; two markers of splicing speckles. Co-localization with SC35-EGFP was confirmed1’ by staining for2’ endogenous SC35.3’ 4’

C NB7-NB1 SC-35 / H2B-mRFPruby Merge

D NB7-mRFPruby SC-35-EGFP Merge

Figure 7. The HSPB7 N-terminus is required and sufficient for targeting to SC35 speckles. (C) Stain- ing of NB7-DNHSPB1 together with endogenous SC-35 shows co-localization of both proteins. The nucleus was visualized using H2B-mRFPruby. (D) Expression of a chimeric protein containing the N-terminus of HSPB7 fused to mRFPruby localized to SC35-EGFP stained speckles.

131 Appendix

DAPI Merge

132 Colour figures

Chapter 4

A - HSPB7 + HSPB7 D - HSPB7 + HSPB7 EGFP-HttQ74 NLS-EGFP-HttQ74

B E

V5-HSPB7 V5-HSPB7 Empty vector 1:9 1:0,6 Empty vector 1:9 1:0,6

Soluble Soluble

-GAPDH -GAPDH

V5 V5 EGFP-HttQ74 NLS-EGFP-HttQ74

C F Empty vector Empty vector

1:9 1:9

1:4,5 1:4,5

1:2,3 1:2,3 V5-HSPB7 1:1,2 V5-HSPB7 1:1,2

1:0,6 1:0,6

133 Appendix

A Tubulin SCA3-Q64 DAPI Overlay

B myc-HSPB7 SCA3-Q64 DAPI Overlay

Huntingtin Q74 SCA3 Q64 C 10 D 14

12 8 i o n s i o n s 10 l u s l u s 6

n c n c 8 I I

l e a r 4 l e a r

u c u c 4 N N 2

% % 2

0 0 l l 7 7 7 7 5 5 5 5 o o r r t t n n o o SP B SP B SP B SP B SP B SP B SP B SP B C C H H H H H H H H - - - - 5 5 5 5 V V V V

134 Colour figures

Chapter 5

A HSP70 / HSC70 B Classical sHSPs 10000 10000 Control HS 1000 HS 1h recovery 1000 HS 2h recovery n n

i o 100 i o 100 c t c t u u d d i n i n

l d l d 10

o 10 o F F

1 1

0.1 0.1 HSP70AA HSC70-4 HSC70-5 HSP22 HSP23 HSP26 HSP27 C Novel sHSPS D Pre-heat shock Post-heat shock 10000

1000 HSP23 HSP26

n HSP27 i o 100 c t CG14207 u d i n

l d 10 o F

0.1 HSP22 1 9 3 7 A C 6 0 3 0 B B 4 4 1 2 HSP67BC EFL 7 7 ) 4 7 3 4 6 6 2 1 1 ( P P CG4461 L C G C G C G C G H S H S CG7409

135 Appendix

A Wildtype and driver lines B pUAS transformed lines

100 100

75 75

l l UAS-HSP67BC-L1 UAS-HSP67BC-L2 v i a v i a r r

u 50 u 50 UAS-CG14207-L3 S S

% % UAS-CG14207-L4 25 25 UAS-CG7409-L1 Wildtype UAS-CG7409-L2 elav UAS-CG4461-L1 0 ey 0 UAS-CG4461-L2

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) C HSP67BC overexpression (line 1)

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25

Control Control 0 UAS-HSP67BC-L1 x elav 0 UAS-HSP67BC-L1 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) HSP67BC overexpression (line 2)

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25

Control Control 0 UAS-HSP67BC-L2 x elav 0 UAS-HSP67BC-L2 x ey

0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (days) Time (days)

Figure 6. The effect of overexpression of the D. melanogaster Small HSP in vivo on lifespan. Cohorts of 100-250 male flies were analysed for their lifespan. Every two days, dead flies were counted, and the percentage of flies alive at each time point is plotted. (A) Wiltype (W1118) and both driver lines (elav-GAL4, ey-GAL4). As they differ only marginally in their lifespan characteristics, the data from these lines was combined to plot the control lifespan curve in the panels C-F. (B) All UAS-sHSP carrying transgenic lines without GAL4 induced transgene expression. Except for line UAS-HSP67BC-2 and line UAS-CG7409-L2, these lines show comparable lifespan characteristics.. (C) Effect of overexpression of HSP67BC using the elav-GAL4 (left –hand panels) or ey-GAL4 (right-hand panels) driver in 2 fly-lines. The individual lines are depicted separately as they were analysed in two individual lifespan experiments.

136 Colour figures

D CG7409 overexpression

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25 Control Control UAS-CG7409-L1 x elav UAS-CG7409-L1 x ey 0 UAS-CG7409-L2 x elav 0 UAS-CG7409-L2 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) E CG14207 overexpression

100 elav 100 ey

75 75 l l v i a v i a r r

u 50 u 50 S S

% %

25 25 Control Control UAS-CG14207-L3 x elav UAS-CG14207-L3 x ey 0 UAS-CG14207-L4 x elav 0 UAS-CG14207-L4 x ey

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time (days) Time (days) F CG4461 overexpression

100 elav 100 ey

80 75 l l 60 v i a v i a r r

u 50 u S S 40 % %

25 Control 20 Control UAS-CG4461-L1 x elav UAS-CG4461-L1 x ey 0 UAS-CG4461-L2 x elav 0 UAS-CG4461-L2 x ey

0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (days) Time (days)

137 Appendix

Chapter 6

Wildtype nuclear EGFP-HttQ74 reduced polyQ aggregation in both cellular compartments. NES NLS

B mRFPruby

HSPB7

NES-HSPB7

NLS-HSPB7 Cytosolic Q74 Nuclear Q74

138 Colour figures

small HSP

1 Substrate

Refolders Anti-Q Amyloid substrate

SC35 speckle

Autophagic vesicle and lysosome

RNA c a Proteasone

b 2 b

3 5

d 4

Figure 4. Model of small HSP activities based on the human HSPB family. Small HSPs can be divided into either refolders or anti-Q proteins with a main difference in oligomeric structure. Both types of small HSPs can however associate with the cytoskeleton for either stabilization, cytoskeletal dynamics or involvement in cytoskeletal transport. Deoligomerization (a) of refolders allows some small HSP members (HSPB1, HSPB5) to enter the nucleus and (b) associate, like anti-Q HSPB7, with SC35 splicing speckles. During protein unfolding (c), refolders bind the non-native substrate and associate into large substrate- HSPB complexes. Further processing can either result in proteasomal degradation or transfer to HSPA members (d) for substrate refolding. Anti-Q HSPB members likely also associate with the non-foldable substrate and subsequently target the substrate for degradation by using proteasomal or autophagic clearance . Failure of substrate refolding or clearance results in cytoskeletal transport of the non-native substrate to the microtule organizing center for subsequent degradation. Both single activities (refolding, anti-Q) or a combination of both can support protein homeostasis and provide longevity in Drosophila.

139 Appendix

140