Tissue-Specific Posttranslational Modification of the Small Heat Shock Protein HSP27 inDrosophila

Tissue-Specific Posttranslational Modification of the Small Heat Shock Protein HSP27 inDrosophila

EXPERIMENTAL CELL RESEARCH ARTICLE NO. 223, 1–8 (1996) 0052 Tissue-Specific Posttranslational Modification of the Small Heat Shock Protein HSP27 in...

158KB Sizes 0 Downloads 73 Views

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

223, 1–8 (1996)

0052

Tissue-Specific Posttranslational Modification of the Small Heat Shock Protein HSP27 in Drosophila RAQUEL MARIN,* JACQUES LANDRY,†

AND

ROBERT M. TANGUAY*,1

*Centre de recherche du CHUL and Laboratoire de ge´ne´tique cellulaire et de´veloppementale, RSVS, Pavillon C. E. Marchand, Universite´ Laval, Ste-Foy, Que´bec, Canada G1K 7P4, and †Centre de Recherche en Cance´rologie, Universite´ Laval, L’Hoˆtel-Dieu de Que´bec, 11 Coˆte du Palais, Que´bec, Canada G1R 2J6

INTRODUCTION Drosophila sHSPs (small heat shock proteins) are expressed in the absence of stress in specific regions of the central nervous system and in gonads of young adult flies. In these two organs, the sHSPs show a cellspecific and developmental stage-specific pattern of expression suggesting distinct regulation and function(s) of each individual sHSP (R. Marin et al., Dev. Genet. 14, 69–77, 1993). Since mammalian HSP27 has been reported to be phosphorylated through a complex novel cascade implicating distinct kinases, we examined whether two of the sHSPs (HSP27 and HSP23) exist in different isoforms as a result of posttranslational modification in vivo. HSP27 and HSP23 were analyzed in various tissues in unstressed and heatshocked flies. Four isoforms of HSP27 were found to be constitutively expressed in the nervous system and in testes and two in ovaries. The proportion of these isoforms relative to each other was specific to a given tissue. In the case of HSP23, two isoforms were expressed in the heads and in testes of unstressed flies. In ovaries, a low level of a single isoform of HSP23 was found. Heat shock caused an increase in the amount of preexisting HSP27 and HSP23 and the appearance of additional isoforms in ovaries. Susceptibility to phosphatase treatment indicated that isoforms of HSP27 were phosphoproteins. This was further supported by in vitro experiments in which Drosophila sHSPs were incubated with purified Chinese hamster HSP27 kinase. Only HSP27 was shown to be a substrate of this mammalian HSP27 kinase. The present data suggest that tissue- and HSP-specific posttranslational modification systems may modulate the function of these proteins in different cell types. Furthermore, the signal transduction pathways leading to phosphorylation of the sHSPs are conserved between mammals and Drosophila, and the sHSP kinase cascade may be developmentally regulated. q 1996 Academic Press, Inc.

In contrast to the families of high-molecular-weight HSPs (heat shock proteins, HSP70 and HSP90) whose participation in a number of basic cellular processes has been well documented [1, 2], the function(s) of the small HSPs remains unclear. Evidence for the involvement of human and Drosophila HSP27 in protection against thermal and oxidative stresses has been presented [3–7], and gene transfection studies have shown that overexpression of HSP27 in mammalian cells is a sufficient condition for conferring thermoresistance [7– 16]. Mammalian HSP27 behaves in vitro as an actincapping protein, suggesting that this protein may be involved in the regulation of the dynamics of actin microfilaments [7–9]. Human HSP27, murine HSP25, and bovine a-crystallin have also been proposed to act as molecular chaperones [10]. Under heat shock conditions, these sHSPs bind unfolding proteins, preventing nonspecific aggregation of the substrate protein. Phosphorylation is another interesting property of sHSPs and is observed in response to a number of different stimuli. This phenomenon has been particularly well documented for mammalian HSP27 [4, 10–14, 17], where phosphorylation occurs rapidly following exposure to stress conditions, and in unstressed cells upon stimulation with mitogens and other inducers of differentiation [for a review, see 18]. Phosphorylation of mammalian HSP27 occurs at the same serine residues in vivo in different organisms, and these sites are located within a common sequence motif R-X-X-S, suggesting that the same protein kinase is involved [13]. HSP27 protein kinase appears to be homologous to the MAP kinase-activated protein kinase II (MAPKAP kinase II) [19, 20]. The activity of HSP27 kinase is highly sensitive to treatment with protein phosphatases [20] and can be activated in vitro by mitogen-activated protein MAP kinase (MAPK), indicating that HSP27 kinase may be linked to signal transduction pathways involving MAPK. It has also been suggested that the state of phosphorylation may be an important modulator of the protective function of the small HSPs in cellu-

1 To whom correspondence and reprint requests should be addressed. Fax: 418-656-7176. E-mail: [email protected].

1

/ m4861$3057

01-24-96 00:14:42

eca

0014-4827/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

AP: Exp Cell

2

MARIN, LANDRY, AND TANGUAY

lar physiology although it may not be essential for the chaperone activity of HSP25 [15]. HSP27 and HSP23 are two members of the family of small heat shock proteins (sHSPs) of Drosophila melanogaster clustered with other sHSP genes within a 12kb stretch of DNA at locus 67B [21, 22]. In response to heat shock, the sHSPs show a coordinate pattern of induction [23]. Some of the sHSPs have also been reported to be transcribed in the absence of stress [24, reviewed in 25]. During development, some sHSPs are transcribed in late larval and late pupal stages [24]. They are also inducible in tissue culture cells by the steroid molting hormone ecdysterone [26–28]. Thus in contrast to their coordinated synthesis following heat shock, the small HSPs of Drosophila show a tissuespecific and developmental-stage-specific pattern of expression in unstressed organisms [24, 29–33]. The specificity in the constitutive patterns of expression of some of these proteins suggest that, in addition to their role during stress, sHSPs may play additional functions in the normal unstressed cell. In cultured cells of D. melanogaster, HSP27 and HSP26 are phosphorylated after treatment with ecdysterone [34]. However, little is known about the posttranslational modifications of the small HSPs under normal physiological conditions. Here we have examined the distinct isoforms of HSP27 and HSP23 present in the brain and in the gonads of both unstressed and heat-shocked flies. We report the presence of distinct isoforms of these two sHSPs in different tissues. In the case of HSP27 these isoforms correspond to phosphorylated forms of the protein. This suggests that each individual sHSP may have distinct mechanisms of posttranslational modifications, and that this modulation may be important for cell-specific functions. MATERIALS AND METHODS Materials. Specific monoclonal antibodies to HSP27 and HSP23 were prepared by immunization of BALB/c mice with sHSP fusion proteins produced in an expression vector, as described previously [33]. The 2C8 (anti-HSP27) and 7B12 (anti-HSP23) monoclonal antibodies were used at dilutions of 1:100 and 1:10, respectively, in immunoblotting assays. Alkaline phosphatase (Escherichia coli) was purchased from Pharmacia. Ampholines (Bio-lyte) were obtained from Bio-Rad. Tissue preparation. Heads, gonads (ovaries or testes), and thoraxes (mostly thoracic muscles) from cold-anesthetized flies of the Oregon-R stock of D. melanogaster raised at 237C were dissected in Ringer’s solution. For heat shock treatments, flies were left for 1 h at 357C in an Eppendorf tube. To prepare 2-D gel samples,2 dissected

2 Abbreviations used: 1-D, one-dimensional; 2-D, two-dimensional; IEF, isoelectrofocusing; PMSF, phenylmethylsulfonyl fluoride; NP40, Nonidet P-40; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; cpm, counts per minute; Hepes, N-2-hydroxyethylpiperazine-N*-2-ethanesulfonic acid; Mops, 3-(N-morpholino) propanesulfonic acid.

/ m4861$3057

01-24-96 00:14:42

eca

organs from 50 heat-shocked or non-heat-shocked flies were placed in tubes containing 150 ml of pH9/NP40 buffer (25 mM Tris–Hcl, pH 9.0, 1 mM EDTA, 1 mM PMSF, and 1% NP40), and homogenized with a micro-tissue grinder (Fisher Scientific). After centrifugation at 13,000 rpm for 30 min at 47C, the soluble phase was recovered and lyophilyzed in a SpeedVac concentrator. Dried pellets were resuspended in 2-D lysis buffer (9.5 M urea, 1% NP-40, 2% ampholines (75% Bio-lyte 5–7, 25% Bio-lyte 3–10), 5% b-mercaptoethanol), and homogenized until complete solubilization was obtained. In the case of 1-D isoelectric focusing (IEF) gels, the dissected organs (heads, testes, ovarioles from early and late stages, embryos, and thoracic muscles) were homogenized and resuspended in the IEF lysis buffer (1% Chaps, 2% ampholines, 5% b-mercaptoethanol, 1 mM PMSF, 1 mM EDTA, saturated with urea). Gel electrophoresis and immunodetection. One-dimensional IEFPAGE (3.3% acrylamide:piperazine diacrylamide (28.3:1.6), 9.0 M urea, 1% Chaps, 3.6% Bio-lyte 5–7, and 0.4% Bio-lyte 3–10) was carried out on horizontal slab gels in a Multiphor II System (Pharmacia LKB), as previously described [20]. The electrophoretic migration was done at 300 V for 4 h. Two-dimensional gels were run according to the method of O’Farrell [35]. Equal amounts of protein were loaded for all samples. For immunoblot assays, proteins were electrophoretically transferred onto nitrocellulose membranes (Gelman). 32P-labeled isoforms of in vitro phosphorylated Drosophila HSP27 were separated on 1-D IEF gels, transferred onto ImmobilonP (Millipore) membranes, and autoradiographed prior to immunoblotting. Membranes were incubated with 2C8 or 7B12 in BLOTTO [36] for 2 h at room temperature, and washed in PBT (PBS / 0.2% Tween 80) as described previously [33]. In some cases, membranes were first treated with the anti-HSP27 antibody (2C8), exposed, and then reblotted with the anti-HSP23 monoclonal antibody (7B12). Membranes from 1-D IEF gels were then incubated with an antimouse IgG horseradish peroxidase-conjugated secondary antibody (ECL, Amersham) diluted 1:2500 in BLOTTO for 1 h at room temperature and processed for detection according to the manufacturer’s protocol. Membranes from 2-D gels were treated with a 125I-labeled goat anti-mouse IgG [28] (2 1 105 cpm/ml in BLOTTO). Alkaline phosphatase treatment. Gonads from 20 males were dissected in Ringer’s solution and homogenized with a micro-tissue grinder in 25 ml of phosphatase buffer (100 mM Tris, pH 8.3, 10 mM MgCl2 , 10 mM ZnCl2). Dephosphorylation was initiated by adding 12 U of the enzyme dissolved in stock solution (phosphatase buffer / 50% glycerol). The reaction mixture was incubated at 377C for 1 or 2 h, and terminated by the addition of 25 ml of 21 IEF lysis buffer. Control reactions were performed with 1 ml of heat-denatured (957C, 5 min) enzyme in the same buffer. Preparation of purified sHSPs for in vitro treatment with HSP27 kinase. Drosophila sHSPs fusion proteins were produced in the pET-3c expression vector [37] expressed in E. coli. For protein extraction, cells were lysed by sonication, resuspended in SDS sample buffer [0.075 Tris–HCl (pH 6.8), 2.3% (w/w) SDS, 5% (v/v) b-mercaptoethanol, 10% (w/v) glycerol and 0.005% (v/v) bromophenol blue], and heated at 957C for 5 min. Proteins were separated on one-dimensional SDS–PAGE [38], and each of the four Drosophila sHSP bands electroeluted against the elution buffer (30 mM Hepes (pH 7.0), 25 mM NaCl, 1 mM MgCl2 , 0.1 mM EGTA) in an electroelutor (ISCO, model 1750). In vitro HSP27 phosphorylation. Recombinant Drosophila HSP27 (4 mg) and recombinant Chinese hamster HSP27 (0.02, 0.05 and 0.1 mg) were phosphorylated with 7.5 munits of highly purified mammalian HSP27 kinase [39] in 10 ml of 10 mM Mops (pH 7.0), 15 mM MgCl2 , 40 mM para-nitrophenylphosphate, 1 mM DTT, 0.1 mM PMFS, and 100 mM [g-32P]ATP (22,000 cpm/mmol). After incubation for 1 h at 307C, a 2-ml aliquot was analyzed immediately using SDS– PAGE. The rest of the reaction mixture was rapidly frozen on dry ice and kept at 0807C for IEF electrophoresis.

AP: Exp Cell

SMALL HSP ISOFORMS IN Drosophila ORGANS

3

FIG. 1. Distinct isoforms of HSP27 and HSP23 found in the brain and in gonads of Drosophila adults. Protein extracts from heads and gonads (testes and ovaries) of non-heat-shocked (237C) flies were separated on 2-D PAGE, and transferred to nitrocellulose membranes. HSP27 a, b, c, and d, and HSP23 a and b isoforms were revealed by blotting with 2C8 (a-HSP27) and reblotting the same membrane with 7B12 (a-HSP23) monoclonal antisera. The most acidic forms (c and d for HSP27 and b for HSP23) are shown on the right.

RESULTS

Drosophila HSP27 and HSP23 Show Distinct Isoforms in the Brain and in Gonads of Nonheatshocked Adult Flies The sHSPs of Drosophila are expressed in the absence of stress in the brain and in gonads [29, 31–33]. To find out whether different isoforms of HSP27 and HSP23 were expressed during development, proteins from dissected brains and gonads of young (0–6 days) adult flies were separated on 2-D gels and the distinct isoforms revealed by immunoblotting. In unstressed flies, four distinct HSP27 species (a–d) and two HSP23 isoforms (a and b) were present in heads and testes (Fig. 1; Fig. 2, H and T). The b and d isoforms of HSP27 and the b isoform of HSP23 were absent in ovaries from non-heat-shocked flies (see also O in Figs. 2 and 3). These results suggest that a complex posttranslational system may be acting to modulate the modifications of HSP27 and HSP23 in a tissue-specific manner. The presence of distinct isoforms of HSP27 and HSP23 under normal physiological conditions was confirmed by an alternative immunoblotting technique using 1-D IEF gels (Fig. 2). Head extracts (H) contained a smaller amount of HSP27 isoforms than gonads (T and O), confirming previous immunohistochemical experiments [31–33] (see also Fig. 3). The four isoforms of HSP27 present in heads were clearly distinguished on overexposed autoradiograms (not shown). The pattern of isoforms in the nervous system is similar to that observed in testes, suggesting similar modifications of these sHSPs in these two tissues. Similar Isoforms Are Expressed during Oogenesis and Embryogenesis To investigate whether different HSP isoforms are expressed at different stages of oogenesis, ovarioles were dissected into early (OI , from germarium to S8) and late (OII , from S8 to mature oocyte) stages [40, 41]. As can be seen in Fig. 2, early stage egg chambers (OI

/ m4861$3057

01-24-96 00:14:42

eca

are enriched in HSP23 while HSP27 is mainly expressed at later stages (OII). The isoforms of HSP27 (a and c) detected during early oogenesis are the same as those found during later stages. Only one isoform of HSP23 is observed at early or late stages of oogenesis. Early (0 to 5 h) embryos (Fig. 2, E1) have an isoform pattern of HSP27 identical to that of late oocytes, consistent with the maternal origin of this HSP [42]. In these embryos HSP23 increases, and this may be related to the expression of this HSP in midline precursor cells observed during early embryogenesis [25, 32]. The presence of both HSP27 and HSP23 decreased in late (5 to 12 h embryos (Fig. 2, E2).

FIG. 2. Comparison of HSP27 and HSP23 forms obtained in different organs by 1-D IEF. Proteins from heads (H), testes (T), and ovaries (O) of non-heat-shocked (237C) flies, and embryos of 0–5 h (E1) and 5–12 h (E2) were solubilized in IEF buffer, and fractionated by 1-D IEF on horizontal slabs. Ovarioles were separated in early (from germarium to S8) (OI) and late (from S8 to S14) (OII) stages. After being transferred to nitrocellulose, being blotted with 2C8 antibody (A), and being reblotted with 7B12 antibody (B), HSP27 a, b, c, and d, and HSP23 a and b forms were visualized as individual bands. As in Fig. 1, the most acidic isoforms are shown on the right.

AP: Exp Cell

4

MARIN, LANDRY, AND TANGUAY

isoforms observed in each tissue are summarized in Table 1. HSP27 Isoforms Are Phosphoproteins

FIG. 3. Changes in HSP isoforms in response to heat shock. Protein extracts from heads (H), testes (T), ovaries (O), and thoracic muscles (M) of non-heat-shocked (237C) or heat-shocked (357C, 1 h) flies were separated on 1-D IEF gels and blotted with anti-HSP27 (a-HSP27) or anti-HSP23 (a-HSP23) antibodies.

Effects of Stress on Expression of HSP27 and HSP23 Isoforms As posttranslational changes of the sHSPs have been suggested to be important for function in stressed cells [reviewed in 18], we next analyzed the modifications of these sHSPs in response to heat shock. Total protein extracts of four tissues, heads (H), testes (T), ovaries (O), and thoracic muscles (M) from unstressed (237C) or heat-shocked flies (357C, 1 h) were separated on 1D IEF gels and immunoblotted (Fig. 3). The relative amount of the basic isoform a in each tissue was generally higher in heat-shocked than in non-heat-shocked tissues. Heat shock also induced an increase in the level of the HSP27 b isoform in testes (T), and of the c isoform in heads (H). In contrast with the reports in mammals demonstrating a dramatic increase of HSP27 phosphorylation within minutes of exposure to heat or chemical stresses [4, 12] or to the addition of serum in HeLa cells [14], no further modification of HSP27 was induced upon heat shock in Drosophila. No striking changes in the HSP23 isoform’s pattern were seen in the heads and testes of stressed flies. In thoracic muscles a single isoform of HSP23 was observed after heat shock, confirming earlier data [31–33].

The main posttranslational modification of HSP27 reported in other organisms is phosphorylation [10, 20, 43, 44]. We therefore determined whether any of the physiologically expressed isoforms of either Drosophila HSP27 or HSP23 were phosphoproteins. The d and c HSP27 isoforms gradually shifted to more basic positions on IEF gels with increasing time of incubation with phosphatase (Fig. 5, lanes 1 and 2), suggesting that they correspond to multiphosphorylated forms. The b form did not show any reduction in intensity upon long incubation, and could not be shifted to the a form. This may result from a reduced accessibility or susceptibility of a phosphate residue. HSP23 was not affected by phosphatase treatment. HSP23b may be a form where the phosphate is not accessible to phosphatase or a form containing a different type of modification. Alternatively, HSP23b may be a product of a distinct HSP23-related gene [45]. Drosophila HSP27 Is Phosphorylated in Vitro by a Mammalian HSP27 Kinase In previous works [19, 20, 39, 46] a novel mitogeninduced, second messenger-independent serine kinase has been implicated in the phosphorylation of mammalian HSP27. Therefore, we tested whether mammalian HSP27 kinase could modify Drosophila sHSPs. Recombinant Drosophila HSP27, HSP26, HSP23, and HSP22 were incubated in vitro with a highly purified HSP27 kinase in the presence of [g032P]ATP. Figure 6A shows the Coomassie blue-stained SDS–PAGE with the four recombinant purified sHSPs (lanes 4–7). Three different concentrations of the Chinese hamster HSP27 were run concomitantly as controls of phosphorylation reactions (lanes 1–3). Recombinant HSP23 (lane 5) did not

Heat Shock Induces Changes in HSP27 and HSP23 Isoforms in Ovaries In contrast to the situation in heads and testes, heat shock induced a strong stimulation of HSP23 synthesis, and the appearance of isoforms HSP27b and also HSP23b in ovaries (Fig. 3, O). The results obtained on 1-D IEF and 2-D gels (Fig. 4) show that some modifications of HSP27 and HSP23 are induced in ovaries specifically under stress conditions. The different sHSP

/ m4861$3057

01-24-96 00:14:42

eca

FIG. 4. Additional isoforms of both HSP27 and HSP23 are induced in ovaries specifically under stress conditions. Total protein extracts from non-heat-shocked (237C) or heat-shocked (357C, 1 h) ovaries were resolved on 2-D gel electrophoresis, and the different isoforms obtained were blotted with monoclonal anti-HSP27 (aHSP27) antibody, and reblotted with monoclonal anti-HSP23 (aHSP23) antibody. Acidic forms are on the right.

AP: Exp Cell

5

SMALL HSP ISOFORMS IN Drosophila ORGANS

TABLE 1 Summary of the Different Isoforms of HSP27 and HSP23 HSP27

HSP23

Tissue

237C

357C

237C

357C

Brain Testis Ovary Embryos Thoracic muscle

a, b, c, da a, b, c, d a—c— a—c— —b

a, b, c, d a, b, c, d a, b, c— ND —b

a, b a, b a— a— —

a, b a, b a, b ND a—

Note. ND, not determined. a a–d describe the isoforms oriented from more basic to more acidic positions on the gels (see also Figs. 1–3). b No sHSP forms observed.

migrate at the expected 23-kDa level. This polypeptide has been isolated from a plasmid construction (pET3c-HSP23) resulting in a recombinant protein of more than 23 kDa. Autoradiographic analysis of the 32P-labeled proteins (Fig. 6B) showed that, among the four Drosophila sHSPs, only HSP27 incorporated 32P (lane 7), and no radioactive signals were detected for HSP26, HSP23, or HSP22. The Different Isoforms of Drosophila HSP27 Are Phosphoproteins The different isoforms of purified Drosophila HSP27, phosphorylated in vitro by mammalian HSP27 kinase, were resolved on 1-D IEF gels, and compared to the isoforms formed in vivo. 32P-labeled HSP27 isoforms were detected by autoradiography (Fig. 7A), and unlabeled isoforms by immunoblotting (Fig. 7B). Three

HSP27 isoforms (b, c, and, d) incorporated 32P (lane 2). No radioactivity signal was detected in HSP27a. Five bands of the purified Drosophila HSP27 were recognized by the anti-HSP27 antibody (Fig. 7B, lane 2). The most acidic ones correspond to the 32P-labeled b, c, and d forms, as shown by their comigration with HSP27 from testes (lane 1). The 32P-labeled isoform d was detected on overexposed autoradiograms. The native recombinant HSP27a was resolved as two different forms on IEF gels. The same two bands were seen when the purified recombinant HSP27 was incubated in the absence of the kinase (lane 3), indicating that this basic band is not related to kinase activity. These results confirm that HSP27c and HSP27d are posttranslationally phosphorylated forms of the same gene product, HSP27a. They further suggest that HSP27b is also a phosphoprotein that, for unknown reasons, is not accessible during phosphatase treatment. DISCUSSION

Drosophila HSP27 and HSP23 are two members of a family of proteins that includes the small HSPs of

FIG. 5. Alkaline phosphatase treatment of HSP27 and HSP23. Total soluble proteins from testes of unstressed (237C) flies were incubated in the presence of alkaline phosphatase (12 units) at 377C for 0 (lane 0), 1 (lane 1), or 2 (lane 2) h. Lane 3 corresponds to testis proteins incubated at 377C for 1 h in the same phosphatase buffer in the presence of heat-denatured enzyme. Equal amounts of protein were loaded in the different samples. The reaction mixtures were analyzed by 1-D IEF and immunoblotting with antisera specific for Drosophila HSP27 (2C8) and HSP23 (7B12). The most acidic forms are on the right.

/ m4861$3057

01-24-96 00:14:42

eca

FIG. 6. In vitro phosphorylation of Drosophila HSP27. Different quantities of purified recombinant Chinese hamster HSP27 and Drosophila sHSPs were incubated at 307C with [g032P]ATP and highly purified Chinese hamster kinase. After 60 min. SDS sample buffer was added, and the reaction mixtures were analyzed by 1-D SDS– PAGE and autoradiography. The different sHSPs were stained with Coomassie blue (A) and 32P-labeled proteins were visualized by autoradiography (B). Lanes 1–3: 0.02, 0.05, and 0.1 mg, respectively, of Chinese hamster HSP27; lanes 4–7: 4 mg of HSP22 (lane 4), HSP23 (lane 5), HSP26 (lane 6), and HSP27 (lane 7).

AP: Exp Cell

6

MARIN, LANDRY, AND TANGUAY

FIG. 7. The different posttranslational forms of Drosophila HSP27 are phosphoproteins. Purified Drosophila HSP27 was phosphorylated in vitro by Chinese hamster HSP27 kinase in the presence of [g032P]ATP (A, lane 2), or nonradioactive ATP (B). The phosphorylation reactions were stopped by adding IEF buffer, and the distinct isoforms were separated on 1-D IEF gels. Drosophila HSP27 was also incubated in the same kinase buffer but in the absence of the enzyme (lane 3). As a control, Chinese hamster HSP27 was also treated with mammalian kinase (lane 5). The purified kinase was also loaded on gels to detect possible nonspecific 32P-labeled bands unrelated to the phosphorylation activity (lane 4). Nonstressed testis proteins were concomitantly run on the IEF gels (lane 1). (A) The different 32P-labeled isoforms visualized by autoradiography. (B) The different HSP27 a, b, c, and d isoforms revealed after blotting with monoclonal anti-HSP27 antibody.

essentially all organisms as well as the aA- and aBcrystallins [47]. Seven sHSPs, four of which are classical sHSPs, are clustered at the same locus of the Drosophila genome [21, 22]. They have been traditionally considered to be regulated by similar mechanisms and to have related functions in the cell. However, it is becoming evident that each sHSP has its own pattern of expression and posttranslational regulation [24, 29–34]. Drosophila HSP27 is constitutively present in four stable forms (the unmodified HSP27a and three phosphoproteins HSP27b, c, and, d) in cells of the central nervous system and testes. Interestingly, only native HSP27a and HSP27c forms are detectable in unstressed ovaries. The b form of HSP27, which is the major acidic isoform constitutively present in testes, is absent in non-heat-shocked ovaries, but can be induced in this organ by heat shock. In addition, in Drosophila brain and testes, heat shock does not lead to increased phosphorylation of the sHSPs as reported in cultured mammalian cells [4, 13]. In fact we observed a reduction in the proportion of most acidic forms in heatshocked brains and testes. These data suggest that the mechanisms regulating phosphorylation of the different sites in HSP27 are complex. As for human HSP27, which is constitutively expressed at low levels in most tissues and rapidly phosphorylated at two sites in stressed cells [11], specific posttranslational mecha-

/ m4861$3057

01-24-96 00:14:42

eca

nisms could be operating in ovaries under stress conditions. Another important aspect of the posttranslational modification of Drosophila HSP27 is the possible relation between the intracellular localization of the distinct isoforms and their level of phosphorylation. Previous data on mammalian sHSPs reported the translocation of phosphorylated forms into or on the nucleus after heat shock [44]. In serum-stimulated HeLa cells, the phosphorylated forms of HSP27 were mainly recovered in a soluble fraction while the unphosphorylated form tended to associate with the particulate fraction [14]. Using whole-mount immunohistochemical experiments on Drosophila, HSP27 has been localized in the nuclei of nurse cells in the first stages of oogenesis, and it is seen to diffuse to the cytoplasm at later stages (R. Marin and R. M. Tanguay, in preparation). One could hypothesize that the distinct isoforms of HSP27 may be implicated in the division and/or differentiation of cells during gametogenesis and during embryogenesis. Drosophila HSP23 is constitutively present in two forms (native HSP23a and a more acidic form HSP23b) in heads and testes. A single unmodified a form of HSP23 is present in unstressed ovaries and, as in the case of HSP27b, the HSP23b isoform is induced when ovaries are submitted to a heat shock. This protein could also be subject to posttranslational regulation. Alternatively, the a and b isoforms recognized by the anti-HSP23 antibody could correspond to two genes, the HSP23 gene and an additional HSP23-related gene. A Drosophila sHSP-homologous gene [1(2)efl] has been reported recently [45]. Interestingly, Drosophila HSP27 can be phosphorylated in an in vitro assay with a highly purified HSP27 kinase from Chinese hamster. Two potential phosphorylation sites similar to those found in mammalian HSP27 (R-X-X-S) [12, 13] are found at serine 58 and serine 75 of Drosophila HSP27. Their location with respect to the a-crystallin domain could correspond to Ser 78 and Ser 82 of human HSP27. A third potential site (Ser 173) of Drosophila HSP27 is found at the Cterminal end. However, none of these sites shows a perfect match with the consensus of the mammalian HSP27 kinase, recently described by Stokoe et al. [46]. Furthermore, in vitro, the Chinese hamster HSP27 kinase has a much lower (at least 50 times) affinity for Drosophila HSP27 than for its mammalian counterpart. Identical Drosophila HSP27 isoforms were generated in vivo by transfection of the Drosophila HSP27 gene in Chinese hamster cells (Y. Wu and R. M. Tanguay, unpublished results). Thus, the same isoforms of Drosophila HSP27 which are phosphorylated in vitro by mammalian HSP27 kinase are also generated in vivo after transfection in Chinese hamster cells. The mammalian HSP27 kinase did not phosphorylate Drosophila HSP26, a protein which has been

AP: Exp Cell

SMALL HSP ISOFORMS IN Drosophila ORGANS

shown to be phosphorylated in vivo in Drosophila heatshocked Kc cells [34], and which resolves into at least six species with different isoelectric points on 2-D gels [33]. Drosophila HSP23 does not have any R-X-X-S motif, and it has not been observed to be phosphorylated in vivo using 32P after heat shock [34], nor after incubation with hamster HSP27 kinase, as described here. The present data suggest that the signal transduction pathways leading to phosphorylation of Drosophila HSP27 may be homologous to those controlling the phosphorylation of mammalian HSP27. Moreover, there could be other Drosophila protein-specific mechanisms controlling the behavior of each individual sHSP. In Drosophila, the HSP27 kinase cascade may be developmentally regulated, and other protein kinases may be implicated, such as for example those stimulated by D-raf, a Drosophila homologue of Raf-1. D-raf plays key roles in multiple signal transduction pathways, such as that required for the determination of cell fates at embryonic termini and for development of the compound eye [48, 49], where HSP27 is expressed. Future studies will help to elucidate the mechanisms regulating the tissue-specific expression, and the complex protein modification–function relationships of these sHSPs during development and under stress conditions. It will also be interesting to investigate whether the sHSP kinase cascade is also developmentally regulated in higher vertebrates. We are grateful to Drs. Y. Wu, E. Khandjian, and Louis Nicole, and to Dominique Mayrand for assistance. This work was supported by the Medical Research Council of Canada (Grant MT-11086 to R.M.T. and Grant MT-7088 to J.L.). Raquel Marin is a recipient of a scholarship from the FCAR of Que´bec.

1. Nover, L. (1990) Heat Shock Response. CRC Press, Boca Raton, FL. 2. Morimoto, R. I., Tissie`res, A., and Georgopoulos, C. (1994) The Biology of Heat Shock Proteins and Molecular Chaperones, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 3. Berger, E. M., and Woodward, M. P. (1983) Exp. Cell. Res. 147, 437–442. 4. Landry, J., Chre´tien, P., Lambert, H., Hickey E., and Weber, L. A. (1989) J. Cell. Biol. 109, 7–15. 5. Rollet, E., Lavoie, J. N., Landry, J., and Tanguay, R. M. (1992) Biochem Biophys. Res. Commun. 185, 116–120. 6. Mehlen, P., Briolay, J., Smith, L., Diaz-Latoud, C., Fabre, N., Pauli D., and Arrigo, A. P. (1993) Eur. J. Biochem. 215, 277– 284. 7. Lavoie, J. N., Gingras-Breton, G., Tanguay, R. M., and Landry, J. (1993) J. Biol. Chem. 268, 3420–3429. 8. Lavoie, J. N., Hickey, E., Weber, L. A., and Landry, J. (1993) J. Biol. Chem. 268, 24210–24214. 9. Miron, T., Vancompernolle, K., Vanderkerckhove, J., Wilchek, M., and Geiger, B. (1991) J. Cell Biol. 114, 255–261. 10. Jakob, U., Gaestel, M., Engel, K., and Buchner, J. (1993) J. Biol. Chem. 268, 1517–1520.

01-24-96 00:14:42

11. Arrigo, A. P., and Welch, W. J. (1987) J. Biol. Chem. 262, 15359–15369. 12. Gaestel, M., Schroder, W., Benndorf, R., Lippmann, C., Buchner, K., Hucho, F., Erdmann, V. A., and Bielka, H. (1991) J. Biol. Chem. 266, 14721–14724. 13. Landry, J., Lambert, H., Zhou, M., Lavoie, J. N., Hickey, E., Weber, L. A., and Anderson, C. W. (1992) J. Biol. Chem. 267, 794–803. 14. Mehlen, P., and Arrigo, A. P. (1994) Eur. J. Biochem. 221, 327– 334. 15. Knauf, U., Jakob, U., Engel, K., Buchner, J., and Gaestel, M. (1994) EMBO J. 13, 54–60. 16. Landry, J., Chre´tien, P., Lambert, H., Hickey, E., and Weber, L. A. (1989) J. Cell Biol. 109, 7–15. 17. Lavoie, J. N., Lambert, H., Hickey, E., Weber, L. A., and Landry, J. (1995) Mol. Cell Biol. 1, 505–516. 18. Arrigo, A. P., and Landry, J. (1994) in The Biology of Heat Shock Proteins and Molecular Chaperones (Morimoto, R. I., Tissie`res, A., and Georgopoulos, C., Eds.), pp. 335–373, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 19. Stokoe, D., Engel, K., Campbell, D. G., Cohen, P., and Gaestel, M. (1992) FEBS Lett. 313, 307–313. 20. Zhou, M., Lambert, H., and Landry, J. (1993) J. Biol. Chem. 268, 35–43. 21. Corces, V., Holmgren, R., Freund, R., Morimoto, R., and Meselson, M. (1980) Proc. Natl. Acad. Sci. USA 77, 5390–5393. 22. Craig, E. A., and McCarthy, B. J. (1980) Nucleic Acids Res. 8, 4441–4457. 23. Tanguay, R. M., and Vincent, M. (1982) Can. J. Biochem. 60, 306–315. 24. Mason, P., Hall, L., and Gausz, J. (1984) Mol. Gen. Genet. 194, 73–78. 25. Arrigo, A. P., and Tanguay, R. M. (1991) in Heat Shock and Development (Hightower, L., and Nover, L., Eds.), pp. 106–119, Springer-Verlag, Berlin. 26. Sirotkin, K., and Davidson, N. (1982) Dev. Biol. 89, 196–210. 27. Ireland R. C., Berger, E., Sirotkin, K., Yund, M. A., Osterbur, D., and Fristrom, J. (1982) Dev. Biol. 93 498–507.

REFERENCES

/ m4861$3057

7

eca

28. Beaulieu, J. F., Arrigo, A. P., and Tanguay, R. M. (1989) J. Cell. Sci. 92, 29–36. 29. Glaser, R. L., Wolfner, M. F., and Lis, J. T. (1986) EMBO J. 4, 747–754. 30. Glaser, R. L., and Lis, J. T. (1990) Mol. Cell. Biol. 10, 131–137. 31. Pauli, D., Tonka, Ch-H., Tissie`res, A., and Arrigo, A. P. (1990) J. Cell. Biol. 111, 817–827. 32. Haass, C., Klein, U., and Kloetzel, P. M. (1990) J. Cell. Sci. 96, 413–418. 33. Marin, R., Valet J. P., and Tanguay, R. M. (1993) Dev. Genet 14, 69–77. 34. Rollet, E., and Best-Belpomme, M. (1986) Biochem. Biophys. Res. Commun. 141, 426–433. 35. O’Farrel, P. H. (1975) J. Biol. Chem. 250, 4007–4021. 36. Johnson, D. A., Gautsch, J. W., Sportsman, J. R., and Elder, J. H. (1984) Gene Anal. Technol. 1, 3–8. 37. Studier, F. W., Rosenberg, A. H., and Dunn, J. J. (1990) Methods Enzymol. 185, 60–89. 38. Thomas, J. O., Ko¨rnberg, R. D. (1975) Proc. Natl. Acad. Sci. USA 72, 2626–2630. 39. Huot, J., Lambert, H., Lavoie, J. N., Guimond, A., Houle, F., and Landry, J. (1995) Eur. J. Biochem 227, 416–427.

AP: Exp Cell

8

MARIN, LANDRY, AND TANGUAY

40. King, R. C., Rubinson, A. C., and Smith, R. F. (1956) Growth 20, 121–157. 41. Cummings, M. R., and King, R. C. (1969) J. Morphol. 128, 427– 442. 42. Zimmerman, J. L., Petri, W., and Meselson, M. (1983) Cell 32, 1161–1170. 43. Arrigo, A. P. (1990) Mol. Cell. Biol. 10, 1276–1280. 44. Arrigo, A. P., Suhan, J. P., and Welch, W. J. (1988) Mol. Cell. Biol. 8, 5059–5071.

45. Kurzik-Dumke, U., and Lohmann, E. (1995) Gene 154, 171– 175. 46. Stokoe, D., Caudwell, B., Cohen, P. T. W., and Cohen, P. (1993) Biochem. J. 296, 843–849. 47. Hickey, E., Brandon, S. E., Potter, R., Stein, G., Stein, J., and Weber, L. A. (1986) Nucleic Acids Res. 14, 4127–4145. 48. Perrimon, N., Engstrom, L, and Mahowald, A. P. (1985) Dev. Biol. 110, 480–491. 49. Nishida, Y., Hata, M., Ayaki, T., Ryo, H., Yamagata, M., Shimizu, K., and Nishizuka, Y. (1988) EMBO J. 7, 775–781.

Received July 3, 1995 Revised version received October 30, 1995

/ m4861$3057

01-24-96 00:14:42

eca

AP: Exp Cell