Conformational changes resulting from ... - BioOne

Cell Stress & Chaperones (2006) 11 (1), 61–70 Q Cell Stress Society International 2006 Article no. csac. 2006.CSC-149R

Conformational changes resulting from pseudophosphorylation of mammalian small heat shock proteins—a two-hybrid study Xiankui Sun, Michael J. Welsh, and Rainer Benndorf Department of Cell and Developmental Biology, University of Michigan Medical School, Ann Arbor, MI 48109, USA

Abstract The human genome codes for 10 so-called mammalian small heat shock or stress proteins (sHsp) with the various tissues expressing characteristic sets of sHsps. Most sHsps interact with each other and form homo- and heterooligomeric complexes. Some of the sHsps are phosphoproteins in vivo, and phosphorylation has been implicated in the regulation of complex size and composition. In this study, we analyze, by the 2-hybrid method, the reporter gene activation pattern of several sHsp pairs that previously have been demonstrated to interact. We show that pseudophosphorylation (mimicry of phosphorylation) of the homologous phosphorylation sites Ser15 and Ser16 in Hsp27 and Hsp20, respectively, modulates characteristics of these sHsps that can be detected by their ability to activate reporter genes in suitable 2hybrid assays. Pseudophosphorylation of the separated N-terminus of Hsp27 alone is not sufficient for the activation of the reporter genes, whereas the separated C-terminus is sufficient. We conclude that pseudophosphorylation of Hsp27 and Hsp20 at their N-termini results in conformational changes that can be detected by their interaction with other sHsps. Pseudophosphorylation of aB-crystallin at Ser19, in contrast, had no detectable consequences.

INTRODUCTION

The human genome codes for 10 so-called mammalian small heat shock or stress proteins (sHsp), with the various tissues expressing characteristic sets of sHsps (Fontaine et al 2003; Kappe´ et al 2003; Verschuure et al 2003). In their C-terminal half, these proteins share a sequence element of approximately 85 amino acid residues called the a-crystallin domain, and toward their N-terminal end, they share a less conserved but nonetheless similar domain (Fig 1). Other structural elements are the variable central domain or region, which separates the a-crystallin and the N-terminal domains (Derham and Harding 1999), and the variable and flexible C-terminal extension or tail (Carver et al 1995). Among these 10 human sHsps, 5 are known to be phosphoproteins, including Hsp27 (HspB1), Hsp20 (HspB6), aB-crystallin (aB-Cry, HspB5), aA-crystallin (HspB4), and Hsp22 (HspB8). Human Hsp27 is phosphorylated at Ser15, Ser78, and Ser82 by Received 2 August 2005; Revised 14 October 2005; Accepted 26 October 2005. Correspondence to: Rainer Benndorf, Tel: 734 615-5670; Fax: 734 7631166; E-mail: [email protected].

MAPKAPK-2 (Landry et al 1992); human aB-Cry at Ser59 by MAPKAPK-2, at Ser45 by p44/42 MAPK, and at Ser19 by an unknown protein kinase (Kato et al 1998); and human Hsp20 at Ser16 by cyclic nucleotide-dependent protein kinases (Beall et al 1999). Phosphorylation of sHsps has been studied in numerous experimental systems (for review, see Gaestel 2002). For example, phosphorylation of Hsp27 has been found to correlate with cytoskeletal functions such as cell migration (Hedges et al 1999) and contraction of fibroblasts, smooth muscle cells, and mesangial cells (Bitar et al 1991; Yamboliev et al 2000; Hirano et al 2002, 2005), as well as with stress protection of neurons (Benn et al 2002). Phosphorylation of aB-Cry at Ser59 is necessary and sufficient for maximal protection of cardiomyocytes from apoptosis (Morrison et al 2003), and phosphorylation of Hsp20 is associated with the relaxation of smooth muscles (Beall et al 1999). In many studies, pseudophosphorylated (ps) sHsps (substitution of 1 or more phosphorylatable serines by aspartate or glutamate residues that mimics phosphorylation) have been used to facilitate the experiments (Rogalla et al 1999; Scha¨fer et al 1999; Benndorf et al 2001; Aquilina et al 2004). 61

62

Sun et al

Fig 1. Multiple alignment and structural features of the human Hsp27, Hsp22, aB-Cry, and Hsp20. This alignment is based on the alignment of all 10 human sHSPs (Fontaine et al 2003) with minor manual editing. Identical and similar (L/V/I, F/Y/W/H, R/K/H, D/E, G/A, S/T) amino acid residues are highlighted in gray if they occur in at least 3 sequences. Identified in vivo phosphorylation sites in Hsp27, aB-Cry, and Hsp20 are highlighted in black. The arrows indicate the separation of Hsp27 into the N- and C-terminal parts as used in this study. Amino acid residues between the arrows are present in both fragments.

sHsps form homo- and hetero-oligomeric complexes, and in cell extracts they are found frequently in variable sizes ranging from monomers/dimers to quadragintamers (Sun et al 2004). Phosphorylation of Hsp27 is usually associated with the disassembly of Hsp27 complexes (Kato et al 2001; Hirano et al 2005). sHsp complexes are dynamic structures with subunits exchanging rapidly (Bova et al 2000). In muscle tissues, Hsp27, aB-Cry, and Hsp20 have been found to form one type of complex, whereas the myotonic dystrophy protein kinase–binding protein (MKBP) and HspB3 form a different type of complex (Sugiyama et al 2000). However, the actual situation might be more complex. For example, Hsp22 binds to most, if not all, sHsps occurring in muscles (Sun et al 2004; Fontaine et al 2005). The processes that control the formation of sHsp complexes are not understood. When studying the interaction properties of Hsp22 by the 2-hybrid (TH) method, we noticed that the reporter gene activation (RGA) by the Hsp22-Hsp27 interaction apparently depended on pseudophosphorylation of Hsp27 if certain TH constructs were used, whereas the corresponding wild-type (wt) construct did not activate the reporter genes (Benndorf et al 2001; Sun et al 2004). From these data, we concluded that pseudophosphorylation of Hsp27 modulates its characteristics, which can be detected in suitable TH assays. In this work, we have studied this pseudophosphorylation-dependent RGA in greater detail using the sHsps Hsp27, Hsp20, aB-Cry, and Hsp22. The collected data are consistent with a conformational change in 2 sHsps (Hsp27, Hsp20) as a result of pseudophosphorylation of their N-terminal phosphorylation sites. The different RGA patterns obtained for the various sHsp-sHsp interactions also tentatively indicated differences in the properties of these sHsps. All sHspsHsp pairs analyzed in this study have been previously demonstrated to be interacting proteins by immunoprecipitation, fluorescence resonance energy transfer, or both in addition to the TH method (Sugiyama et al 2000; Sun et al 2004; Fontaine et al 2005). Cell Stress & Chaperones (2006) 11 (1), 61–70

Recent reports have shown the association of missense mutations in HSP27, HSP22, and aB-Cry with various neuronal and muscular diseases (Vicart et al 1998; Selcen and Engel 2003; Evgrafov et al 2004; Irobi et al 2004; Kijima et al 2005, Tang et al 2005). Disease-associated mutant sHsps show aberrant interaction and complex-forming properties (Bova et al 1999; Perng et al 1999; Irobi et al 2004). The data on conformational changes and differential properties of sHsps as presented here might be of significance for the understanding of the regulation of the composition, dynamics, and the function of sHsp complexes in both healthy and diseased tissues. MATERIALS AND METHODS Multiple alignment

The alignment of the 4 sHSPs—HSP27, HSP22, HSP20, and aB-Cry—and the designation of the various domains as shown in Figure 1 are based on the multiple alignment of all 10 human sHSPs as published previously (Fontaine et al 2003) with minor manual editing. Cloning information

All relevant data on origin of sHsp cDNAs, plasmid constructs and their designation, cloning methods, and PCR primers are given in Table 1. Constructs were grouped according to their use into vector groups (VeGr) as listed in footnote a of Table 1 and as specified in Figures 2–6. Two-hybrid experiments

Small-scale sequential transformation of the yeast strain Y190 was performed as described in the manufacturer’s instructions (Clontech, Mountain View, CA, USA). The TH vectors used were pAS2-1 (pAS) and pACT2 (pAC). The yeast was first transformed with the constructs 2, 3, or 4 and grown on -Leu medium or with the constructs

Conformational changes in small heat shock proteins

63

Table 1 Designation of the constructs, origin of the sHsp cDNA, cloning methods, and vector groups (VeGr)a as used in this study Construct no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Construct designation pAS–Hsp22 pAC–wtHsp27c pAC–3DHsp27c pAC–Hsp22 pAS–wtHsp27e pAS–S15DHsp27 pAS–S78DHsp27 pAS–S82DHsp27 pAS–2DHsp27e,g pAS–3DHsp27e pAS–wtHsp27-N pAS–3DHsp27-N pAS–Hsp27-C pAC–wtHsp20 pAC–S16DHsp20 pAS–wtHsp20 pAS–S16DHsp20 pAC–wtaB-Cry pAC–S19DaB-Cry pAC–3DaB-Cry pAS–wtaB-Cry pAS–S19DaB-Cry pAS–3DaB-Cry

Source of sHsp cDNA/cloning method/reference Sun et al 2004 Subcloning of pBS–wtHsp27d Subcloning of pBS–3DHsp27d Benndorf et al 2001 Subcloning of pBS–wtHsp27d Mutagenesis of construct 5 Mutagenesis of construct 5 Mutagenesis of construct 5 Subcloning of pBS–2DHsp27d,g Subcloning of pBS–3DHsp27d Sun et al 2004 Sun et al 2004 Sun et al 2004 Subcloning of construct 16 Subcloning of construct 17 PCR of a cDNA library Mutagenesis of construct 16 PCR of pET20B–wtaB-Cryh Mutagenesis of construct 18 PCR of pRF14G–3DaB-Cryi PCR of construct 18/TopoTA Mutagenesis of construct 21 PCR of construct 20/TopoTA

Used restriction sites

Primersb

— EcoRI, XhoI EcoRI, XhoI — NcoI/BspHI, XhoI/SalIf — — — NcoI/BspHI, XhoI/SalIf NcoI/BspHI, XhoI/SalIf — — — NcoI, XmaI NcoI, XmaI NcoI, EcoRI — NcoI, XhoI — NcoI, XhoI NcoI, SalI — NcoI, SalI

— — — — — 1, 2 3, 4 5, 6 — — — — — — — 7, 8 9, 10 11, 12 13, 14 11, 12 15, 16 13, 14 15, 16

a Vector groups used in this study: VeGr I (constructs 1–3); VeGr II (constructs 4–13); VeGr III (constructs 1, 14, 15); VeGr IV (constructs 4, 16, 17); VeGr V (constructs 1, 18–20); VeGr VI (constructs 4, 21–23); VeGr VII (constructs 5, 14, 15); VeGr VIII (constructs 2, 16, 17); VeGr IX (constructs 5, 18–20); VeGr X (constructs 2, 21–23); VeGr XI (constructs 10, 18–20); VeGr XII (constructs 3, 21–23). b (1) 59-TCTCGCTCCTGCGGGGCCCCGACTGGGACCCCTTCCGCGACT; (2) 5′-AGTCGCGGAAGGGGTCCCAGTCGGGGCCCCGCAG GAGCGAGA; (3) 5′-CAGCCGCGCGCTCGACCGGCAACTCAGCAGC; (4) 5′-GCTGCTGAGTTGCCGGTCGAGCGCGCGGCTG; (5) 5′CTCAGCCGGCAACTCGACAGCGGGGTCTCGGAG; (6) 5′-CTCCGAGACCCCGCTGTCGAGTTGCCGGCTGAG; (7) 5′-CAAATCCATGGA GATCCCTGTGCCTGTGCAG; (8) 5′-CAGAATTCAGATCGGCTTTAATAGAGGGAGC; (9) 5′-GTCTTGGCTGCGCCGCGCCGA CGCCCCGTTGCCCGGACTTTC; (10) 5′-GAAAGTCCGGGCAACGGGGCGTCGGCGCGGCGCAGCCAAGAC; (11) 5′–CTAGCCACCA TGGACATCGCCATCCACCACCCC; (12) 5′-CAAAGCTCGAGCTATTTCTTGGGGGCTGCGGTGAC; (13) 5′-GCCCCTTCTTTCC TTTCCACGACCCCAGCCGCCTCTTTGACC; (14) 5′-GGTCAAAGAGGCGGCTGGGGTCGTGGAAAGGAAAGAAGGGGC; (15) 5′-CCA TGGACATCGCCATCCACCA; (16) 5′-GTCGACGCTATTTCTTGGGGGCTGCGGTG. c After subcloning, the 59- and 39-flanking sequences of these constructs in pAC were: . . .GGTGGTCATATGGCCATGGA GGCCCCGGGGATCCGAATTCTGCAGATCTCGGTACCCTTTTCTGAGCAGACGTCCAGAGCAGAGTCAGCCAGCATG-HSP27TAAAGCCTTAGCCTGGATGCCCACCCCTGCTGCCGCCACTGGCTGTGCCTCCCCCGCCACCTGTGTGTTCTTTTGATACATT TATCTTCTGTTTTTCTCAAATAAAGTTCAAAGCAACCACCTGTAAAAAAAAAACCCCCCCCCCCCCCCTGCAGCCCAAGCTTATCGATA CCGTCGAACCTCGAGAGATC . . . (start and stop codons of Hsp27 in bold; used restriction sites underlined). d Obtained from L.A. Weber, Reno, NV, USA. e After subcloning, the 59- and 39-flanking sequences of these constructs in pAS were: . . . TATGGCCATGACC-Hsp27-TAAAGCCTTA GCCCGGATGCCCACCCCTGCTGCCGCCACTGGCTGTGCCTCCCCCGCCACCTGTGTGTTCTTTTGATACATTTATCTTCTGTTTTTCT CAAATAAAGTTCAAAGCAACCACCTGTCAAAAAAAAAACCCCCCCCCCCCCCCCTGCAGCCCAAGCTTATCGATACCGTCGAACCTCGA CCTGCA . . . (start and stop codons of Hsp27 in bold; used restriction sites underlined). f Restriction sites of the vector and insert are compatible. g 2D refers to substitution of the serines Ser78 and Ser82 in human Hsp27 to aspartates. h Obtained from J. Horwitz, Los Angeles, CA, USA. i Obtained from H. Ito, Kasugai, Aichi, Japan.

1, 5, or 10, and grown on -Trp medium as described previously (Sun et al 2004). In the second step, the yeast was transformed with the complementary vectors within the same VeGr. After the second transformation, selection was on -Trp,-Leu,-His medium. The strain Y190 has 2 reporter genes (his1, growth of colonies; gal1, blue colonies), and the interaction assays were considered positive only if both reporter genes were activated. The controls in all TH experiments were designed according to the following general scheme: pAS-sHspX alone, pAC-sHspY alone, pAS-sHspX 1 pAC, and pAC-sHspY 1 pAS. In none of the controls was the gal reporter gene activated, although the his reporter gene sometimes was slightly activated,

resulting in a few colonies. Some of the TH controls (C1– C12) are shown in Figures 2, 5, and 6. For each interaction, false positive results have been excluded previously (Sugiyama et al 2000; Sun et al 2004; Fontaine et al 2005). RESULTS Multiple alignment of sHSPs and positions of phosphorylation sites

The alignment of human Hsp27, Hsp22, aB-Cry, and Hsp20 shows the domain structure of these sHsps and the positions of the known in vivo phosphorylation sites Cell Stress & Chaperones (2006) 11 (1), 61–70

64

Sun et al

Fig 2. Reporter gene activation by different forms of Hsp27 with Hsp22. (A) Assays with full-length Hsp22 and Hsp27 constructs of VeGr I. Both forms of Hsp27 (wtHsp27, 3DHsp27) activated the reporter genes to a similar extent (pseudophosphorylation had no effect). (B) Assays with a full-length Hsp22 construct and various full-length and truncated Hsp27 constructs of VeGr II. In full-length Hsp27, pseudophosphorylation of Ser15 was required for activation of the reporter genes, whereas pseudophosphorylation of Ser78 and Ser82 had no or only minor effects. Pseudophosphorylation of the separated N-terminal part of Hsp27 was not sufficient for activation of the reporter genes, whereas the separated C-terminal part of Hsp27, together with Hsp22, was sufficient for activation of the reporter genes. TH controls: C1, pAS-Hsp22 alone; C2, pAS-Hsp22 1 pAC; C3, pAC-Hsp22 alone; C4, pAC-Hsp22 1 pAS.

Fig 3. Reporter gene activation by different forms of Hsp20 with Hsp22. (A) Assays with Hsp20 and Hsp22 constructs of VeGr III. Both forms of Hsp20 (wtHsp20, S16DHsp20) activated the reporter genes to a similar extent (pseudophosphorylation had no effect). (B) Assays with Hsp20 and Hsp22 constructs of VeGr IV. Pseudophosphorylation of Hsp20 at Ser16 was required for activation of the reporter genes. TH controls: C1–C4 as in Figure 2.

(Fig 1). Hsp27 has 1 phosphorylation site (Ser15, position 21) near the N-terminus and 2 sites (Ser78, position 84; Ser82, position 88) in the central region. Hsp20 also has 1 site (Ser16, position 21) near the N-terminus that is apparently homologous to Ser15 of Hsp27. In this alignment, the N-terminal site of aB-Cry (Ser19, position 27) is located slightly farther away from the N-terminus, possibly indicating that this site is nonhomologous to Ser15 and Ser16 of Hsp27 and Hsp20, respectively. In addition, aB-Cry has 2 more phosphorylation sites, one in the Nterminal domain (Ser45, position 57) and another in the central region (Ser59, position 73). In the experiments of this study, different wild-type and pseudophosphorylated forms of these sHsps have Cell Stress & Chaperones (2006) 11 (1), 61–70

been used (the positions in which the phosphorylatable serines have been substituted by aspartate are designated). The Hsp27 forms used include wtHsp27, S15DHsp27, S78D Hsp27, S82DHsp27, 2DHsp27 (Ser78 and Ser82 replaced by Asp), 3DHsp27 (Ser15, Ser78, and Ser82 replaced by Asp), wtHsp27-N (wild-type N-terminal part of Hsp27), 3D Hsp27-N (3-D form of the N-terminal part of Hsp27), and Hsp27-C (C-terminal part of Hsp27). The aB-Cry forms used include wtaB-Cry, S19DaB-Cry, and 3DaB-Cry (Ser19, Ser45, and Ser59 replaced by Asp), and the Hsp20 forms used include wtHsp20 and S16DHsp20. Conformational change in Hsp27 as revealed by its interaction with Hsp22

Hsp22 and Hsp27 are interacting proteins, as has been shown previously by different methods (Sun et al 2004). With the TH method, RGA was observed with the constructs of VeGr I (Hsp22, construct 1; wtHsp27, construct 2). Similarly, the reporter genes were activated if the corresponding 3DHsp27 sequence (construct 3) was used (Fig 2A). This suggests that under these conditions, the RGA is independent of pseudophosphorylation of Hsp27. However, with the TH constructs of VeGr II, the results were different. Although Hsp22 (construct 4) and 3D Hsp27 (construct 10) did activate both reporter genes, Hsp22 (construct 4) and wtHsp27 (construct 5) together did not (Fig 2B). This was the original observation when we identified Hsp22 as the binding partner of Hsp27 (Benndorf et al 2001), which was later confirmed (Sun et al 2004). We reasoned that pseudophosphorylation at 3


Fig 4. Reporter gene activation by different forms of aB-Cry with Hsp22. (A) Assays with aB-Cry and Hsp22 constructs of VeGr V. All forms of aB-Cry (wtaB-Cry, S19DaB-Cry, 3DaB-Cry) activated the reporter genes to a similar extent (pseudophosphorylation had no effect). (B) Assays with aB-Cry and Hsp22 constructs of VeGr VI. None of the forms of aB-Cry (wtaB-Cry, S19DaB-Cry, 3DaB-Cry) activated the reporter genes (pseudophosphorylation had no effect). TH controls: C1–C4 as in Figure 2.

65

construct 7) and Ser82 (S82DHsp27, construct 8) or both together (2DHsp27, construct 9) had no or only a minor effect. To analyze this in greater detail, we separated the Nand C-terminal domains of Hsp27 in the central region (Fig 1) and tested both parts separately. As shown previously (Sun et al 2004) and in Figure 2B, the C-terminal part of Hsp27 (Hsp27-C, construct 13) activates, together with Hsp22, both reporter genes, whereas the N-terminal part is insufficient, be it in its wild-type form (wtHsp27N, construct 11) or in its pseudophosphorylated form (3DHsp27-N, construct 12). These data demonstrate that pseudophosphorylation per se (ie, without the context of the full-length protein) is not sufficient for RGA. Instead, the C-terminus is also required for this function. Collectively, pseudophosphorylation of Ser15 is sufficient to suppress the restraint imposed by the vectors of VeGr II. The data are consistent with a conformational change induced by pseudophosphorylation of Hsp27 at Ser15, which would enable the C-terminus to interact with Hsp22. This conformational change can be visualized by TH assays provided that a suitable vector system is used. Additionally, a suitable interacting partner (such as Hsp22) is required (cf the experiments below). Conformational change in Hsp20 as revealed by its interaction with Hsp22

Fig 5. Reporter gene activation by different forms of Hsp20 with wt Hsp27. (A) Assays with Hsp20 and wtHsp27 constructs of VeGr VII. (B) Assays with Hsp20 and wtHsp27 constructs of VeGr VIII. In both vector groups, all forms of Hsp20 (wtHsp20, S16DHsp20) activated the reporter genes to a similar extent (pseudophosphorylation had no effect). TH controls: C5, pAS-wtHsp27 alone; C6, pAS-wtHsp27 1 pAC; C7, pAC-wtHsp27 alone; C8, pAC-wtHsp27 1 pAS.

sites of Hsp27 possibly induced changes at the molecular level, which would permit RGA under these conditions. To determine the contribution of the individual phosphorylation sites to this effect, single and double psHsp27 constructs were tested by the same assays. As shown in Figure 2B, pseudophosphorylation of Ser15 in full-length Hsp27 (S15DHsp27, construct 6) is necessary and sufficient for full RGA. Pseudophosphorylation of Ser78 (S78DHsp27,

Because Hsp20 has a phosphorylation site (Ser16) homologous to Ser15 in Hsp27 (Fig 1), we tested the effect of pseudophosphorylation of Hsp20 in a similar way, again using Hsp22 as interacting partner. Both proteins previously have been identified as interacting partners (Fontaine et al 2005). With the use of the VeGr III vectors, both wild-type (wtHsp20, construct 14) and pseudophosphorylated (S16DHsp20, construct 17) Hsp20 activated the reporter genes when probed with Hsp22 (construct 1; Fig 3A). With the VeGr IV vectors, however, psHsp20 (S16DHsp20, construct 17) but not wtHsp20 (construct 16) activated the reporter genes when probed with Hsp22 (construct 2; Fig 3B). Therefore, Hsp20 has an RGA pattern similar to Hsp27 when probed with Hsp22. Thus, pseudophosphorylation of Ser16 is sufficient to suppress the restraint imposed by the vectors of VeGr IV. Analogous to Hsp27, these data suggest that pseudophosphorylation near the N-terminus of Hsp20 (Ser16) also results in conformational changes, as is detected by its interaction with Hsp22. aB-Cry–Hsp22 interaction

aB-Cry has 3 phosphorylation sites, including Ser19 near the N-terminus (Fig 1). To test the effect of pseudophosphorylation of aB-Cry on the RGA in its interaction with Cell Stress & Chaperones (2006) 11 (1), 61–70

66

Sun et al

Fig 6. Reporter gene activation by different forms of aB-Cry and Hsp27. (A) Assays with aB-Cry and Hsp27 constructs of VeGr IX. (B) Assays with aB-Cry and Hsp27 constructs of VeGr X. In both vector groups, all forms of aB-Cry (wtaB-Cry, S19DaB-Cry, 3DaB-Cry) activated the reporter genes to a similar extent with wtHsp27 as interacting partner (pseudophosphorylation had no effect). (C) Assays with aB-Cry and Hsp27 constructs of VeGr XI. All forms of aB-Cry (wtaB-Cry, S19DaB-Cry, 3DaB-Cry) activated the reporter genes to a similar extent with 3D Hsp27 as interacting partner (pseudophosphorylation had no effect). (D) Assays with aB-Cry and Hsp27 constructs of VeGr XII. None of the forms of aB-Cry (wtaB-Cry, S19DaB-Cry, 3DaB-Cry) activated the reporter genes (pseudophosphorylation had no effect). Comparison of panels A and B with panels C and D reveals that pseudophosphorylation of Hsp27 (3DHsp27) had an effect in the interaction with all forms of aB-Cry. TH controls: C5–C8 as in Figure 5; C9, pAS-3DHsp27 alone; C10, pAS-3DHsp27 1 pAC; C11, pAC-3DHsp27 alone; C12, pAC3D Hsp27 1 pAS.

Hsp22, we have included S19DaB-Cry and 3DaB-Cry in addition to wtaB-Cry in these assays. Both proteins have been identified previously as interacting partners (Fontaine et al 2005). With the VeGr V vectors, all forms of aB-Cry (wtaB-Cry, construct 18; S19DaB-Cry, construct 19; 3D aB-Cry, construct 20) activated the reporter genes in combination with Hsp22 (construct 1; Fig 4A). When the complementary experiment was done with the vectors of VeGr VI, none of the aB-Cry forms (wtaB-Cry, construct 21; S19DaB-Cry, construct 22; 3DaB-Cry, construct 23) activated the reporter genes in combination with Hsp22 (construct 4; Fig 4B). Thus, aB-Cry did not show pseudophosphorylation-sensitive RGA with HSP22 as the interacting partner. These data suggest that aB-Cry, in its interaction with HSP22, has a different property when compared with both Hsp27 and Hsp20. Pseudophosphorylation of Ser19 is not sufficient to suppress the restraint imposed by the vectors of VeGr VI. The reason may be that Ser19 near the N-terminus of aB-Cry is nonhomologous with the sites Ser15 and Ser16 in Hsp27 and Hsp20, respectively (Fig 1). Hsp20—wtHsp27 interaction

In the experimental approaches described here, the effect of pseudophosphorylation of Hsp27, Hsp20, and aB-Cry has been probed by their interaction with Hsp22. To learn about the specific contribution of Hsp22 in these assays, Hsp22 subsequently has been replaced by wtHsp27 in otherwise similar settings. Hsp27 was chosen because it is Cell Stress & Chaperones (2006) 11 (1), 61–70

most closely related to Hsp22 among all sHsps (Sun et al 2004) and because the Hsp20-Hsp27 interaction has been established previously (Sugiyama et al 2000). With the use of the VeGr VII constructs (wtHsp27, construct 5; wtHsp20, construct 14; S16DHsp20, construct 15; Fig 5A) or VeGr VIII constructs (wtHsp27, construct 2; wt Hsp20, construct 16; S16DHsp20, construct 17; Fig 5B), all forms of Hsp20 activated, together with Hsp27, the reporter genes. Thus, the pseudophosphorylation-dependent RGA that is characteristic of Hsp20 in the Hsp20Hsp22 interaction is not observed in the Hsp20-Hsp27 interaction with the use of the VeGr VII and VIII vectors. aB-Cry–Hsp27 interaction

Following the same rationale, the effect of pseudophosphorylation of aB-Cry on its interaction with Hsp27 has been examined. Previously, Hsp27 and aB-Cry have been shown to be interacting proteins (Sugiyama et al 2000). With the VeGr IX constructs (wtHsp27, construct 5; wt aB-Cry, construct 18; S19DaB-Cry, construct 19; 3DaB-Cry, construct 20; Fig 6A) and VeGr X constructs (wtHsp27, construct 2; wtaB-Cry, construct 21; S19DaB-Cry, construct 22; 3DaB-Cry, construct 23; Fig 6B), all forms of aB-Cry, together with wtHsp27, activated both reporter genes. No pseudophosphorylation-dependent effects on the RGA were observed. Finally, to evaluate the effect of pseudophosphorylation of Hsp27 in the Hsp27–aB-Cry interaction, we used the 3D Hsp27 constructs in otherwise identical conditions. With the use of the VeGr XI constructs (3DHsp27, con-


struct 10; wtaB-Cry, construct 18; S19DaB-Cry, construct 19; 3D aB-Cry, construct 20), all forms of aB-Cry activated, together with 3DHsp27, the reporter genes, although this interaction was somewhat impaired, judging from the number and uneven size of the colonies (Fig 6C). On the other hand, with the use of VeGr XII constructs (3DHsp27, construct 3; wtaB-Cry, construct 21; S19DaB-Cry, construct 22; 3DaB-Cry, construct 23), none of the aB-Cry forms activated, together with 3DHsp27, both reporter genes (Fig 6D). Thus, no pseudophosphorylation-dependent effects of aB-Cry have been observed, regardless of whether wt Hsp27 of 3DHsp27 has been used as the interacting partner. These results demonstrate that wtHsp27 and 3DHsp27 are clearly different in their RGA patterns if aB-Cry is used as the interacting partner, again suggesting a conformational change in Hsp27 as a result of pseudophosphorylation. Thus, the pseudophosphorylation-sensitive RGA patterns of Hsp27 have been demonstrated in the interaction with both aB-Cry (Fig 6) and Hsp22 (Fig 2). However, both aB-Cry and HSP22 have a different property: aB-Cry activated the reporter genes of both vector groups with wtHsp27 as partner, whereas Hsp22 activated the reporter genes of both vector groups with 3DHsp27 as partner. DISCUSSION

In TH experiments, the nature of the N-terminal gal4 moiety (including the linker sequence resulting from the cloning procedure) of the fusion proteins is known to influence the outcome, sometimes leading to false negative or false positive results. The negative data on some sHsp interactions as collected in this study fall into the category of false negative results because the tested proteins are known to interact. In genetic terms, the chosen vectors imposed a restraint, resulting in nonpermissive conditions. However, the ability to suppress this restraint (creating permissive conditions) just by substituting 1 amino acid residue (Ser15 in Hsp27, Ser16 in Hsp20) turns this system into a useful analytical tool. In other words, we took advantage of the sensitivity of the RGA to detect effects resulting from pseudophosphorylation of sHsps. Because under both permissive and nonpermissive conditions the N-terminal gal4 moieties were identical, the pseudophosphorylation-sensitive RGA can be attributed to this mutation only, thus rendering the effect of the gal4 moieties of the fusion proteins irrelevant. The major conclusion drawn from the results we present is that the properties of 2 sHsps—Hsp27 and Hsp20—change because of pseudophosphorylation near the N-terminus. This has been determined in the interaction of Hsp27 with both Hsp22 (Fig 2) and aB-Cry (Fig 6) and of Hsp20 in the interaction with Hsp22 (Fig 3).

67

These changes, as detected by the TH method, are most likely conformational changes for 2 reasons: (1) pseudophosphorylation of the separated N-terminus of Hsp27 at all 3 phosphorylation sites per se does not result in RGA, and (2) the Hsp27—Hsp22 interaction involves the C-terminus of Hsp27 (Fig 2; Sun et al 2004). These data are consistent with the concept that pseudophosphorylation of Hsp27 at Ser15 causes a conformational change resulting in the exposure of the C-terminus so that the interacting partner (Hsp22) can bind. Although not analyzed in detail, this pseudophosphorylation-dependent conformational change of Hsp27 was also detected with aB-Cry as partner (Fig 6). Additionally, Hsp20 which has a homologous phosphorylation site at Ser16 is subject to a similar change resulting from pseudophosphorylation. aB-Cry, of which the near N-terminal phosphorylation site is apparently not homologous to the Ser15 and Ser16 sites of Hsp27 and Hsp20, respectively, does not show evidence of a comparable conformational change from pseudophosphorylation (as determined with Hsp22 and Hsp27 as interacting partners and with the use of the indicated vectors; cf Figs 4 and 6). Thus, aB-Cry might have a different property compared with Hsp22 and Hsp27. In hamster Hsp27, phosphorylation of Ser15, in addition to Ser90 (Ser90 corresponds to Ser78 and Ser82 in the human sequence), is necessary for thermoprotection (Theriault et al 2004). Whereas the role of Ser90 phosphorylation in the dissociation of sHsp complexes into dimers is relatively well understood at the molecular level, the role of Ser15 phosphorylation is less clear. Structural data suggest that the N-terminal region of sHsps (including the region corresponding to Ser15 in the human sequence) is normally hidden within the oligomeric structure and binds to the a-crystallin domain (Leroux et al 1997; Kim et al 1998) but becomes exposed on complex dissociation into dimers (van Montfort et al 2001). Phosphorylation of hamster Hsp27 at Ser90 would result in the dissociation of the complexes into dimers, which would expose the N-terminus to subsequent additional phosphorylation at Ser15. The dissociated dimers might well adopt a conformation that does not require the interaction of the N-terminus with the a-crystallin domain, and the Ser15 region then would be free to interact with other proteins (Theriault et al 2004). However, other data suggest a different sequence of events. In mouse mesangial cells, the monophosphorylated isoform of Hsp27 was almost exclusively phosphorylated at Ser15, but not at the central phosphorylation site Ser86 (Hirano et al 2005). This monophosphorylated Hsp27 isoform, however, occurred in the large Hsp27 complexes and did not cause complex dissociation. Stimulation of the cells with cadmium resulted in additional phosphorylation at the central phosphorylation site (emergence of the bis-phosphorCell Stress & Chaperones (2006) 11 (1), 61–70

68

Sun et al

ylated isoform) and in complex disassembly (Hirano et al 2005). These data suggest a sequential phosphorylation of Hsp27 (N-terminal site prior to central site). Whatever the sequence of events is, phosphorylation near the N-terminus is very likely to induce a conformational change that is related to the structural organization of the sHsp complexes. The RGA patterns presented here also support, although tentatively, the notion that the various sHsps have partially different properties resulting in different qualities of the various sHsp-sHsp interactions. Preliminary conclusions drawn from these data are: (1) Hsp20 (wt/ ps) and Hsp27 (wt/ps) have a common property (pseudophosphorylation-sensitive RGA) with Hsp22 as interacting partner (Figs 2 and 3); (2) Hsp22 and wtHsp27 have a common property (pseudophosphorylation-insensitive RGA) with aB-Cry (wt/ps) as interacting partner (Figs 4 and 6A, B); (3) Hsp22 and wtHsp27 have a different property (pseudophosphorylation-sensitive vs -insensitive RGA) with Hsp20 (wt/ps) as interacting partner (Figs 3 and 5); (4) aB-Cry (wt/ps) and Hsp20 (wt/ps) have a common property (pseudophosphorylation-insensitive RGA) with wtHsp27 as interacting partner (Figs 5 and 6A, B); (5) aB-Cry (wt/ps) and Hsp20 (wt/ps) have a different property (pseudophosphorylation-insensitive vs -sensitive RGA) with Hsp22 as interacting partner (Figs 3 and 4); (6) Hsp22 and aB-Cry have a different property with Hsp27 (wt/ps) as interacting partner—although both interactions show a pseudophosphorylation-sensitive RGA, they are different: Hsp22 shows RGA in both VeGr with 3D HSP27 (Fig 1), whereas wtaB-Cry shows RGA in both VeGr with wtHSP27 (Fig 6); and (7) aB-Cry (wt/ps) and Hsp27 (wt/ps) have a different property (pseudophosphorylation-insensitive vs -sensitive RGA) with Hsp22 as interacting partner (Figs 2 and 4). A complicating factor for such comparisons of different sHSPs in different TH interactions is that, for cloning reasons, the linker sequences in the gal4-sHSP fusion proteins are partially different (cf cloning information in Table 1). This could influence the outcome, and further studies are required before final conclusions on functional similarities and differences of the various sHsps can be drawn. In the various tissues, the biological role of the 10 mammalian sHsps is only poorly understood, including their degree of functional differences and redundancy. Similarities in their sequences (Fig 1) and in biochemical properties such as chaperoning (Horwitz 1992; Rogalla et al 1999; Bukach et al 2004; Chowdary et al 2004; Kim et al 2004; Carra et al 2005) suggest a similar biological role. However, striking sequence differences, notably in the Cterminal tail region and central region, imply functional differences, which could explain why mutations in various sHsps are associated with similar, although not identical, disease phenotypes (Benndorf and Welsh 2004). This Cell Stress & Chaperones (2006) 11 (1), 61–70

report is a first approach to identifying subtle differences between 4 of these sHsps with the use of a genetic method. The structure and dynamics of the sHsp complexes have been the subjects of numerous studies, although almost all of them focused on homo-oligomeric sHsp complexes or on mixed aA- and aB-crystallin complexes. The in vivo situation in most tissues is far more complex, with each tissue probably expressing a unique set of several of the sHsps (Verschuure et al 2003). Additional variability results because only a subset of sHsps is stress inducible. This implies the possibility that unique types of sHsp complexes occur in each tissue in a given physiological situation. The picture becomes even more complicated if phosphorylation of at least some sHsps is taken into account. The data presented here tentatively suggest that each single sHsp has unique properties that should contribute in a specific way to the properties of the complexes. Recently, mutations in Hsp22, Hsp27, and aB-Cry have been identified that are associated with a number of human neuronal and muscular diseases (Vicart et al 1998; Selcen and Engel 2003; Evgrafov et al 2004; Irobi et al 2004; Kijima et al 2005, Tang et al 2005). Interestingly, some of the mutations (R120GaB-Cry, K141NHsp22, 141EHsp22) alter the ability of the affected proteins to interact with other sHsps, resulting in the formation of abnormal sHsp complexes and aggresomes (Bova et al 1999; Kumar et al 1999; Chavez-Zobel et al 2003; Irobi et al 2004). It is very likely that abnormal complex properties of these mutant sHsps contribute to the manifestation of the diseases. Taken together, the results presented here suggest that pseudophosphorylation of some sHsps near their N-termini causes conformational changes that can be detected by the activation of reporter genes in TH experiments. The data presented here also indicate that the interactions between different sHsps might have different qualities. ACKNOWLEDGMENTS

This work was supported by National Institutes of Health grant P01ES11188 (to M.J.W. [PI] and R.B.). We thank L.A. Weber (Reno, NV, USA) for providing us with various Hsp27 cDNA constructs (wtHsp27, 2DHsp27, 3DHsp27), J. Horwitz (Los Angeles, CA, USA) for providing us with wt aB-Cry cDNA, and H. Ito (Kasugai, Aichi, Japan) for providing us with 3DaB-Cry cDNA. Jeff Ballou’s (Ann Arbor, MI, USA) help with the cloning of some vectors is gratefully acknowledged. REFERENCES Aquilina JA, Benesch JL, Ding LL, Yaron O, Horwitz J, Robinson CV. 2004. Phosphorylation of aB-crystallin alters chaperone func-


tion through loss of dimeric substructure. J Biol Chem 279: 28675–28680. Beall A, Bagwell D, Woodrum D, Stoming TA, Kato K, Suzuki A, Rasmussen H, Brophy CM. 1999. The small heat shock-related protein, Hsp20, is phosphorylated on serine 16 during cyclic nucleotide-dependent relaxation. J Biol Chem 274: 11344–11351. Benn SC, Perrelet D, Kato AC, Scholz J, Decosterd I, Mannion RJ, Bakowska JC, Woolf CJ. 2002. Hsp27 upregulation and phosphorylation is required for injured sensory and motor neuron survival. Neuron 36: 45–56. Benndorf R, Sun X, Gilmont RR, et al. 2001. Hsp22, a new member of the small heat shock protein superfamily, interacts with mimic of phosphorylated Hsp27 (3DHsp27). J Biol Chem 276: 26753– 26761. Benndorf R, Welsh MJ. 2004. Shocking degeneration. Nat Genet 36: 547–548. Bitar KN, Kaminski MS, Hailat N, Cease KB, Strahler JR. 1991. Hsp27 is a mediator of sustained smooth muscle contraction in response to bombesin. Biochem Biophys Res Commun 181: 1192– 1200. Bova MP, McHaourab HS, Han Y, Fung BK. 2000. Subunit exchange of small heat shock proteins. Analysis of oligomer formation of alphaA-crystallin and Hsp27 by fluorescence resonance energy transfer and site-directed truncations. J Biol Chem 275: 1035– 1042. Bova MP, Yaron O, Huang Q, Ding L, Haley DA, Stewart PL, Horwitz J. 1999. Mutation R120G in aB-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. Bukach OV, Seit-Nebi AS, Marston SB, Gusev NB. 2004. Some properties of human small heat shock protein Hsp20 (HspB6). Eur J Biochem 271: 291–302. Carra S, Sivilotti M, Chavez Zobel AT, Lambert H, Landry J. 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. Carver JA, Esposito G, Schwedersky G, Gaestel M. 1995. 1H NMR spectroscopy reveals that mouse Hsp25 has a flexible C-terminal extension of 18 amino acids. FEBS Lett 369: 305–310. Chavez-Zobel AT, Loranger A, Marceau N, Theriault JR, Lambert H, Landry J. 2003. Distinct chaperone mechanisms can delay the formation of aggresomes by the myopathy-causing R120G alphaB-crystallin mutant. Hum Mol Genet 12: 1609–1620. Chowdary TK, Raman B, Ramakrishna T, Rao CM. 2004. Mammalian Hsp22 is a heat-inducible small heat-shock protein with chaperone-like activity. Biochem J 381: 379–387. Derham BK, Harding, JJ. 1999. a-Crystallin as a molecular chaperone. Prog Retinal Eye Res 18: 463–509. Evgrafov OV, Mersiyanova I, Irobi J, et al. 2004. Mutant small heatshock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy. Nat Genet 36: 602–606. Fontaine JM, Rest JS, Welsh MJ, 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. Fontaine JM, Sun X, Benndorf R, Welsh MJ. 2005. Interactions of HSP22 (HSPB8) with HSP20, aB-crystallin, and HSPB3. Biochem Biophys Res Commun 337: 1006–1011. Gaestel M. 2002. sHsp-phosphorylation: enzymes, signaling pathways and functional implications. Prog Mol Subcell Biol 28: 151– 169. Hedges JC, Dechert M, Yamboliev IA, Martin J, Hickey E, Weber LA,

69

Gerthoffer WT. 1999. A role for p38 (MAPK)/Hsp27 pathway in smooth muscle cell migration. J Biol Chem 274: 24211–24219. Hirano S, Rees RS, Gilmont RR. 2002. MAP kinase pathways involving hsp27 regulate fibroblast-mediated wound contraction. J Surg Res 102: 77–84. Hirano S, Sun X, Deguzman CA, et al. 2005. p38 MAPK/Hsp25 signaling mediates cadmium-induced contraction of mesangial cells and renal glomeruli. Am J Physiol Renal Physiol 288: F1133– F1143. Horwitz J. 1992. Alpha-crystallin can function as a molecular chaperone. Proc Natl Acad Sci U S A 89: 10449–10453. Irobi J, Impe KV, Seeman P, et al. 2004. Hot-spot residue in small heat-shock protein 22 causes distal motor neuropathy. Nat Genet 36: 597–601. Kappe´ G, Franck E, Verschuure P, Boelens WC, Leunissen JA, de Jong WW. 2003. The human genome encodes 10 alpha-crystallin–related small heat shock proteins: HspB1-10. Cell Stress Chaperones 8: 53–61. Kato K, Ito H, Iwamoto I, Lida K, Inaguma Y. 2001. Protein kinase inhibitors can suppress stress-induced dissociation of Hsp27. Cell Stress Chaperones 6: 16–20. Kato K, Ito H, Kamei K, Inaguma Y, Iwamoto I, Saga S. 1998. Phosphorylation of alphaB-crystallin in mitotic cells and identification of enzymatic activities responsible for phosphorylation. J Biol Chem 273: 28346–28354. Kijima K, Numakura C, Goto T, Takahashi T, Otagiri T, Umetsu K, Hayasaka K. 2005. Small heat shock protein 27 mutation in a Japanese patient with distal hereditary motor neuropathy. J Hum Genet (on-line publication ahead of print). Kim KK, Kim R, Kim SH. 1998. Crystal structure of a small heatshock protein. Nature 394: 595–599. Kim MV, Seit-Nebi AS, Marston SB, Gusev NB. 2004. Some properties of human small heat shock protein Hsp22 (H11 or HspB8). Biochem Biophys Res Commun 315: 796–801. Kumar LV, Ramakrishna T, Rao CM. 1999. Structural and functional consequences of the mutation of a conserved arginine residue in aA- and aB-crystallins. J Biol Chem 274: 24137–24141. Landry J, Lambert H, Zhou M, Lavoie JN, Hickey E, Weber LA, Anderson CW. 1992. Human hsp27 is phosphorylated by heat shock and mitogen-activated kinases that recognize the same amino acid motif as S6 kinase II. J Biol Chem 267: 794–803. Leroux MR, Melki R, Gordon B, Batelier G, Candido EP. 1997. Structure-function studies on small heat shock protein oligomeric assembly and interaction with unfolded polypeptides. J Biol Chem 272: 24646–24656. Morrison LE, Hoover HE, Thuerauf DJ, Glembotski CC. 2003. Mimicking phosphorylation of alphaB-crystallin on serine-59 is necessary and sufficient to provide maximal protection of cardiac myocytes from apoptosis. Circ Res 92: 203–211. Perng MD, Muchowski PJ, van Den IJssel P, Wu GJ, Hutcheson AM, Clark JI, Quinlan RA. 1999. The cardiomyopathy and lens cataract mutation in alphaB-crystallin alters its protein structure, chaperone activity, and interaction with intermediate filaments in vitro. J Biol Chem 274: 33235–33243. Rogalla T, Ehrnsperger M, Preville X, et al. 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. Scha¨fer C, Clapp P, Welsh MJ, Benndorf R, Williams JA. 1999. HSP27 expression regulates CCK-induced changes of the actin cytoskeleton in CHO-CCK-A cells. Am J Physiol 277: C1032–C1043. Selcen D, Engel AG. 2003. Myofibrillar myopathy caused by novel dominant negative aB-crystallin mutations. Ann Neurol 54: 804– 810.

Cell Stress & Chaperones (2006) 11 (1), 61–70

70

Sun et al

Sugiyama Y, Suzuki A, Kishikawa M, et al. 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. Sun X, Fontaine JM, Rest JS, Shelden EA, Welsh MJ, Benndorf R. 2003. Interaction of human Hsp22 (HspB8) with other small heat shock proteins. J Biol Chem 279: 2394–2402. Tang BS, Zhao GH, Luo W, et al. 2005. Small heat-shock protein 22 mutated in autosomal dominant Charcot-Marie-Tooth disease type 2L. Hum Genet 116: 222–224. Theriault JR, Lambert H, Chavez-Zobel AT, Charest G, Lavigne P, Landry J. 2004. Essential role of the NH2-terminal WD/EPF motif in the phosphorylation-activated protective function of mammalian Hsp27. J Biol Chem 279: 23463–23471.

Cell Stress & Chaperones (2006) 11 (1), 61–70

van Montfort RL, Basha E, Friedrich KL, Slingsby C, Vierling E. 2001. Crystal structure and assembly of a eukaryotic small heat shock protein. Nat Struct Biol 8: 1025–1030. Verschuure P, Tatard C, Boelens WC, Grongnet JF, David JC. 2003. Expression of small heat shock proteins HspB2, HspB8, Hsp20 and cvHsp in different tissues of the perinatal developing pig. Eur J Cell Biol 82: 523–530. Vicart P, Caron A, Guicheney P, et al. 1998. A missense mutation in the alphaB-crystallin chaperone gene causes a desmin-related myopathy. Nat Genet 20: 92–95. Yamboliev IA, Hedges JC, Mutnick JL-M, Adam LP, Gerthoffer WT. 2000. Evidence for modulation of smooth muscle force by the p38 MAP kinase/Hsp27 pathway. Am J Physiol Heart Circ Physiol 278: H1899–H1907.