The intracellular localization and expression of hsp27 (heat-shock protein 27) were investigated by cellular fractionation and immunofluorescence microscopy in Drosophila S3 cells. In unstressed cells, hsp27 is expressed in only 2 % of the cells, whereas following heat shock, during recovery or after induction by ecdysone, the protein is detected in all cells. Under all these conditions, hsp27 appears to be concentrated in the nuclear region as revealed by immunofluorescence. During heat shock, this hsp is localized primarily in the nucleus with an enrichment in the perinucleolar region. However, the cellular fractionation data indicate that the nature of hsp27 interaction with nuclear components greatly differs depending on whether or not cells were subjected to elevated temperatures. After heat shock, hsp27 is resistant to non-ionic detergent extraction. In cells allowed to recover at normal temperature and in those where its synthesis was induced by the molting hormone, ecdysone, this hsp is readily solubilized by detergent. These data suggest that, following heat shock, hsp27 may become physically associated with some nuclear component(s) that are resistant to detergent extraction.

Exposure of cells to supraoptimal temperatures or to many other physiological or chemical stresses is known to induce the rapid expression of a set of genes coding for a restricted number of polypeptides, collectively referred to as the heat-shock proteins (hsps) (reviewed by Schlesinger et al. 1982; Atkinson & Walden, 1985). This response to heat has been observed in most unicellular and multicellular organisms. The basic structure of many hsps is conserved in these organisms, suggesting that these proteins, which can be generally divided into three groups of 80—90, 68–73 and 20–30 (× 103)Mr on the basis of their subunit molecular weights, perform conserved and essential functions in all organisms (Nover, 1984; Tanguay, 1985; Craig, 1985; Lindquist, 1986; Subjeck & Shyy, 1986).

Although many observations suggest that these proteins are involved in the cellular ability to repair reversible damages caused by stress, there is still very little information on the exact function(s) of each individual hsp. The recent efforts directed at the intracellular localization and the chemical characterization of these proteins have provided potential clues concerning the function(s) of some hsps, particularly those of the 70 and 90 (×103)Mr families (Schlesinger, 1986; Pelham, 1986). In the case of the small hsps, the significance of the intracellular distribution of these hsps remains unclear. For example, in Drosophila cells, where the localization of the four small hsps (22, 23, 26 and 27 (×103)) was particularly well investigated, biochemical fractionation studies indicated a nuclear or cytoplasmic localization, depending on the nature and severity of the stress. Their apparent association with the nuclear fraction during heat shock has been suggested to be the result of their interactions with chromatin and nucleolar components (Arrigo et al. 1980), or ribonucleoprotein complexes (Kloetzel & Bautz, 1983), or to reflect their association with nucleoskeletal (Levinger & Varsharvsky, 1981; Sinibaldi & Morris, 1981) or extranuclear cytoskeletal elements that coisolate with the nucleus (Vincent & Tanguay, 1982; Tanguay et al. 1985; Leicht et al. 1986). Alternatively, when cells were stressed by moderate increases in temperature or by other agents, the small hsps remained in a soluble cytoplasmic form (Vincent & Tanguay, 1982; Ireland & Berger, 1982). During Drosophila development, hsp23 is found in the soluble lysate of pupae (Arrigo, 1987), while 40 % of hsp27 is found in a particulate form (Arrigo & Pauli, 1988). Both hsps can be shifted to an insoluble pellet following heat shock of whole animals. These paradoxical results suggest that further investigation is required regarding this group of hsps whose expression is regulated by stress and during normal development. Interestingly, the induction and accumulation of the small hsps has also been correlated with the acquisition of thermotolerance in whole animals (Berger & Woodward, 1983).

We have recently reported the more precise cellular localization of one of the small hsps (hsp23) by immunofluorescence microscopy. This hsp was found to be concentrated in the nucleolus and in granular structures in the cytoplasm of heat-shocked Drosophila cells. During recovery from heat shock, this hsp regained a diffuse cytoplasmic distribution (Tanguay, 1985; Duband et al. 1986). Similar nuclear-cytoplasmic shuttling behaviour was observed in whole animals during development by immunoblotting of various subcellular fractions (Arrigo, 1987). In the present investigation, we have determined the intracellular localization of another small heat-shock protein, hsp27, in cultured Drosophila cells using both indirect immunofluorescence and cellular fractionation techniques. Morphological evidence is presented that suggests that hsp27 is associated with the nucleus of heat-shocked, recovering or ecdysone-treated Drosophila S3 cells.

Cells and cell culture

Drosophila Schneider line 3 (S3) cells were grown at 23°C in Schneider’s revised Drosophila medium (GIBCO) supplemented with 10% heat-inactivated foetal calf serum (FCS).

Labelling of proteins and cell fractionation

In cell fractionation experiments, 5×106 S3 cells were seeded on 35 mm diameter tissue culture dishes and kept overnight at 23°C. The culture medium was then discarded and replaced by incomplete medium (without methionine) and culture dishes were preincubated at 23°C (control) or at 37°C for 30 min before the addition of [35S] methionine (50μCiml-1; 1019 Ci mmol-1 ; ICN, Irvine, CA) for one additional hour. At the end of the labelling period, the medium was removed and cells were washed with ice-cold complete medium. The cells were harvested from the dishes using a rubber policeman, washed once and lysed by vortexing at 4°C in TNM buffer (25 mM-Tris-HC1, pH 7·4, 25mM-NaCl, 5 mM-MgCl2) containing 0-5 % Nonidet P-40, 15 mM-β-mercaptoethanol and 0·1 mM-phenylmethylsulphonyl fluoride (Tanguay & Vincent, 1982). The lysate was centrifuged at 1000g for 5 min. The nuclear pellet was further washed in the hypotonic buffer. The supernatants were pooled and precipitated with 4 vol. of cold acetone overnight at —20°C.

In some experiments, cells were lysed in cold hypotonic buffer without NP-40. After 10 min on ice, they were homogenized with a Dounce glass homogenizer. The cell lysate was centrifuged at 1000g for 10 min to yield a nuclear pellet. The supernatant was spun at 10000g for 30 min and the 10000g supernatant at 100 000g for 60 min. The pellets as well as the high-speed supernatant fractions were analysed on gels as described below.

In recovery experiments, the labelling medium containing [35S] methionine was discarded, the dishes were washed once in complete medium and the cells were further incubated at 23 °C for 2 to 8 h in culture medium supplemented with 10% FCS.

In some experiments, ecdysone (Ecdysterone, Calbiochem) was added from an aqueous stock solution (5 mg ml-1) at 1 or 10μM (Ireland & Berger, 1982; Ireland et al. 1982) and cells were incubated at 23 °C for 20 h.

Gel electrophoresis

The acetone precipitates and the nuclear pellets were dissolved in Laemmli sample buffer (250mM-Tris HC1, pH6·8; 2% SDS; 10 % glycerol and 5 % β-mercaptoethanol). Proteins were separated on SDS-12% polyacrylamide gels in the buffer system of Thomas & Kornberg (1975) using an acrylamide: bis ratio of 30:0·8 (Laemmli, 1970). The gels were stained with Coomassie Blue, and autoradiographed with Kodak XAR-5 films.

Protein immunoblotting

Immunoblotting was performed on cellular extracts from cells incubated in the absence of [35S]methionine. Proteins from gels were electrophoretically transferred to a nitrocellulose filter (BioRad) at 200mA overnight (Towbin et al. 1979). Following transfer, proteins and Mr markers were localized by staining with Ponceau Red (Burke et al. 1982). Filters were saturated in PBS (137mM-NaCl, 3mM-KCl, SmM-NazHPCL, 1·5 mM-KH2PO4) containing 5 % Blotto (Johnson et al. 1984) for 1 h at 37°C and further incubated under agitation with the rabbit anti-hsp27 antiserum (Arrigo & Pauli, 1988) for 3 h at 37°C. After washing in PBS containing 0·05% Tween 20, the filters were incubated with 1Z5I-labelled goat anti-rabbit IgG antibodies for 1 h at 37°C in PBS-Blotto and washed in PBS-Tween. Filters were dried and autoradiographed.

Indirect immunofluorescence

S3 cells were seeded on methanol-cleaned microscope slides in complete Schneider’s medium containing 10% FCS and, when indicated, 1 or 10 μM-ecdysone. Control and ecdysone-treated cells were incubated at 23°C. Heat-shocked cells were maintained at 37°C for 1·5 h and allowed to recover at 23°C for 0, 3 or 8h. Cells were fixed in 3-7% formaldehyde in PBS for 30 min. After two rinses in PBS, the slides were incubated successively in PBS containing 0·1 M-glycine, PBS-0·1 % Triton X-100 and finally rinsed twice in PBS. In some experiments, cells were lysed in situ before fixation by washing for 2 min in TNM buffer containing 0·5% NP-40, 15 mM-β-mercaptoethanol and 0·1 mM-phenylmethylsulphonyl fluoride. The slides were then fixed as described above. Fixed cells were blocked in PBS containing 2% bovine serum albumin (BSA) for 30 min and processed for immunofluorescence by incubation for 1 h with the anti-hsp27 antiserum diluted in PSB-BSA. The slides were washed in PBS for 30 min and incubated for 1 h with FITC-conjugated goat anti-rabbit IgG diluted in PBS-BSA. After three 10-min washes in PBS, slides were mounted in glycerol and examined with a Zeiss photomicroscope III. The specificity of the anti-hsp27 has been demonstrated by immunoprecipitation and immunoblotting (Arrigo & Pauli, 1988). In addition, the immunoblot data presented in the next section confirm that this antiserum reacts specifically against hsp27 in Drosophila S3 cells.

Distribution of hsp27 in subcellular fractions during heat shock and recovery

In order to monitor the synthesis and the intracellular distribution of hsp27 in Drosophila cells cultured under normal, heat-shocked and recovering conditions, cells were labelled with [35S] methionine for 1 h, washed with complete culture medium and allowed to recover for 0, 2, 5 or 8h at 23 °C prior to cell fractionation. The crude nuclear pellets and the corresponding detergent-soluble extracts were analysed by SDS-PAGE (Fig. 1A) and radiolabelled proteins were visualized by autoradiography (Fig. IB). As shown previously (Arrigo et al. 1980; Tanguay & Vincent, 1982) the hsps show a specific distribution between the nuclear and cytoplasmic fractions (Fig. IB, lane 2, tracks a and b, respectively). hsp83 is found in the cytoplasmic extract, hsp70 in both nuclear and cytoplasmic fractions and the small hsps (hsp22–27) are concentrated in the nuclear fraction. Furthermore, the data suggest that the solubility of the individual small hsps differs under these conditions of cellular fractionation: hsp27 is found in the nuclear detergent-resistant pellet while hsp26, hsp23 and to a greater extent hsp22 are also found in the detergent-soluble extract.

Fig. 1.

Distribution of hsp27 in control, heat-shocked and recovering Drosophila S3 cells. Cells were incubated either at 23°C (lanes 1) or at 37°C (lanes 2) for 1·5 h or at 37°C for 1·5 h followed by a recovery at 23°C for 2h (lanes 3), Sh (lanes 4) or 8 h (lanes 5). Cells were then fractionated into a crude nuclear pellet (tracks a) and a detergent-soluble extract (tracks b), which were analysed by SDS-PAGE as described in Materials and methods. A. The Coomassie Bhie-stained gel; and B, the corresponding autoradiogram. For labelling, cells were incubated at 23°C or at 37°C in SD medium lacking methionine for 30 min and [35S]methionine was added for the next hour. Cells were harvested immediately (lane 1: 23°C; lane 2: 37°C) or subjected to 2-8 h of chase at the recovering temperature (lanes 3·5). C. The immunoblot. Cells were incubated and processed in the same conditions as B except that proteins were not labelled. The gel was transferred to nitrocellulose paper and probed with anti-hsp27 serum and 1251-labelled goat anti-rabbit IgG antibodies. Proteins from 106 cells were loaded on each track. Mr markers (shown × 10”3) on the left are: phosphorylase a, bovine serum albumin, actin and carbonic anhydrase. Positions of the major stress proteins are indicated on the right.

Fig. 1.

Distribution of hsp27 in control, heat-shocked and recovering Drosophila S3 cells. Cells were incubated either at 23°C (lanes 1) or at 37°C (lanes 2) for 1·5 h or at 37°C for 1·5 h followed by a recovery at 23°C for 2h (lanes 3), Sh (lanes 4) or 8 h (lanes 5). Cells were then fractionated into a crude nuclear pellet (tracks a) and a detergent-soluble extract (tracks b), which were analysed by SDS-PAGE as described in Materials and methods. A. The Coomassie Bhie-stained gel; and B, the corresponding autoradiogram. For labelling, cells were incubated at 23°C or at 37°C in SD medium lacking methionine for 30 min and [35S]methionine was added for the next hour. Cells were harvested immediately (lane 1: 23°C; lane 2: 37°C) or subjected to 2-8 h of chase at the recovering temperature (lanes 3·5). C. The immunoblot. Cells were incubated and processed in the same conditions as B except that proteins were not labelled. The gel was transferred to nitrocellulose paper and probed with anti-hsp27 serum and 1251-labelled goat anti-rabbit IgG antibodies. Proteins from 106 cells were loaded on each track. Mr markers (shown × 10”3) on the left are: phosphorylase a, bovine serum albumin, actin and carbonic anhydrase. Positions of the major stress proteins are indicated on the right.

In order to follow the behaviour of the hsps under recovery conditions, cells labelled for 1 h at 37°C were chased for 2-, 5-or 8-h periods at 23°C (Fig. IB, lanes 3, 4 and 5, respectively). A large part of the hsps associated with the nuclear pellet at the end of heat shock (lane 2a) are found in the detergent-soluble fraction as early as 2h after the shift to recovery conditions (lane 3b). This is particularly evident in the case of hsp27, which is only found in trace amounts in the nuclear pellet of recovering cells (lanes 3–5).

The subcellular distribution of hsp27 was further investigated by immunoblot using an antiserum directed against this hsp. Control, heat-shocked and recovering S3 cells were fractionated into crude nuclear pellets and detergent-soluble cytoplasmic extracts. Proteins were separated by SDS-PAGE, blotted on nitrocellulose and probed with the anti-hsp27 antibody. As shown in Fig. 1C, the antiserum did not detect any low Mr proteins in control cells (lane 1), while in heat-shocked (lane 2) and recovering cells (lanes 3-5) it specifically recognized a protein of 27×103Mr. Since the 27×103Mr protein detected by the antibody comigrates with radiolabelled hsp27 (not shown) and the antigen is only detected in stressed cells (Fig. 1C), it is concluded that this antiserum is specific for hsp27. The immunoblot results with the antiserum confirm that hsp27 is preferentially associated with the nuclear pellet in heat-shocked cells (Fig. 1C, lane 2a). When cells are allowed to recover at 23 °C, hsp27 rapidly shifts to the detergent-soluble cytoplasmic extract. The ratio of soluble (tracks a) to insoluble (tracks b) hsp27 appears to be totally inverted after the first 2h of recovery (lane 3). Subsequently, the amount of nuclear pellet-associated hsp27 decreases progressively as demonstrated by the 5 h (lane 4a) and 8 h (lane 5a) recover) ’ data while that of the soluble form remains abundant and relatively constant (lanes 3 to 5, tracks b).

Immunocytochemical localization of hsp27 in heat-shocked cells

To determine whether the biochemical association of hsp27 with the nuclear pellet of fractionated cells reflects its actual nuclear localization or rather results from its association with large sedimenting structures that coisolate with the nucleus, its cellular localization was investigated by immunofluorescence microscopy in control, heat-shocked and recovering cells.

Unstressed cells were not stained with anti-hsp27 antiserum, with the exception of approximately 2% of the cells, which exhibited a bright nuclear fluorescence (Fig. 2A-B). Immediately after heat shock (37°C, 1-5 h), nearly all cells become positive with a predominant staining of the nuclei (Fig. 2C-D). In these cells, the distribution of hsp27 within the nucleus is not uniform and patches of fluorescence can be seen (Fig. 3A-B). When cells were allowed to recover at 23°C for 3h, the pattern of nuclear fluorescence appears remarkably similar to that observed for heat-shocked cells (Fig. 3C-D). Moreover, even after 8 h of recovery, the nucleus remains predominantly stained and nucleoli do not appear to be stained (Fig. 3E-F).

Fig. 2.

Localization of hsp27 in unstressed and heat-shocked Drosophila S3 cells by indirect immunofluorescence. Cells grown on microscope slides were fixed as described in Materials and methods prior to (A.B) or after incubation at 37°C for 1·5 h (C,D). Celis were stained with the antibody against hsp27 and visualized with FITC-conjugated anti-rabbit IgG antibodies. A,C. Fluorescent micrographs; B,D, corresponding phase-contrast images. Bar, 50μm.

Fig. 2.

Localization of hsp27 in unstressed and heat-shocked Drosophila S3 cells by indirect immunofluorescence. Cells grown on microscope slides were fixed as described in Materials and methods prior to (A.B) or after incubation at 37°C for 1·5 h (C,D). Celis were stained with the antibody against hsp27 and visualized with FITC-conjugated anti-rabbit IgG antibodies. A,C. Fluorescent micrographs; B,D, corresponding phase-contrast images. Bar, 50μm.

Fig. 3.

Localization of hsp27 after heat shock and during recovery. Cells grown on microscope slides were subjected to heat-shock conditions (37°C, 1·5 h) (A,B) and allowed to recover at 23°C for 3 h (C,D) or 8h (E,F). Cells were fixed and stained as described in Materials and methods. A,C,E. Fluorescent micrographs; and B,D and F, corresponding phase-contrast images. Bar, 10 μm.

Fig. 3.

Localization of hsp27 after heat shock and during recovery. Cells grown on microscope slides were subjected to heat-shock conditions (37°C, 1·5 h) (A,B) and allowed to recover at 23°C for 3 h (C,D) or 8h (E,F). Cells were fixed and stained as described in Materials and methods. A,C,E. Fluorescent micrographs; and B,D and F, corresponding phase-contrast images. Bar, 10 μm.

Localization of hsp27 in ecdysone-treated cells

The possibility cannot be excluded that the nuclear localization of hsp27 in heat-shocked and in recovering cells may simply result from an artefactual deposition and immobilization of the protein in the nucleus following exposure to extreme temperatures. In order to test this hypothesis, the synthesis of the small hsps was stimulated using the steroid moulting hormone, ecdysone (Ireland & Berger, 1982). S3 cells were cultured at 23 °C in complete medium supplemented with 1 or 10 μM-ecdysone for 20 h, and the induction and localization of hsp27 were monitored by subcellular fractionation and immunofluorescence. The immunoblot of Fig. 4 shows that ecdysone strongly stimulates the synthesis and accumulation of hsp27 at the two concentrations tested (lanes 2 and 3; 1 and 10 μM-ecdysone, respectively) in comparison with untreated cells (lane 1), which in this experiment exhibit a low basal level of hsp27. Moreover, the amount of hsp27 detected in ecdysone-treated cells appears even higher than that measured in heat-shocked cells (37°C, 1·5 h; lane 4).

Fig. 4.

Distribution of hsp27 in S3 cells cultured in the presence of ecdysone. Cells were grown in the complete medium either without or supplemented with 1 or 10μ.M-ecdysone for 20h. Control (lane 1), ecdysone-treated (lane 2: 1μM; lane 3: 10^M) and, as positive control, heat-shocked cells (lane 4: 37°C, 1·5 h) were harvested at the end of the incubation period and fractionated into detergent-insoluble (tracks a) or soluble (tracks b) fractions. hsp27 was separated by SDS-PAGE and detected by immunoblotting as for Fig. 1.

Fig. 4.

Distribution of hsp27 in S3 cells cultured in the presence of ecdysone. Cells were grown in the complete medium either without or supplemented with 1 or 10μ.M-ecdysone for 20h. Control (lane 1), ecdysone-treated (lane 2: 1μM; lane 3: 10^M) and, as positive control, heat-shocked cells (lane 4: 37°C, 1·5 h) were harvested at the end of the incubation period and fractionated into detergent-insoluble (tracks a) or soluble (tracks b) fractions. hsp27 was separated by SDS-PAGE and detected by immunoblotting as for Fig. 1.

The distribution of hsp27 between the nuclear pellet (lanes 1–4, tracks a) and the detergent-soluble cytoplasmic extract (lanes 1–4, tracks b) clearly differs according to the mode of induction. hsp27 is found exclusively in the detergent-soluble fraction of untreated (lane lb) and ecdysone-stimulated cells (lanes 2b and 3b) while, as shown above, it is concentrated in the nuclear fraction of heat-shocked cells (lane 4a). However, the localization of hsp27 by immunofluorescence microscopy illustrates that the antiserum decorates predominantly the nuclear region of ecdysone-treated cells (Fig. 5), as previously observed in heat-shocked and recovering cells (compare Fig. 5 with Fig. 3). This nuclear localization of hsp27 in ecdysone-treated cells was further confirmed by cell lysis not in those incubated at 37°C. This suggests that the strength of association of this hsp with some nuclear component(s) could be dependent on the nature of the stress applied to the cells. To verify this possibility the localization of hsp27 was further examined by immunofluorescence on heat- or ecdysone-induced cells per-meabilized with NP-40 in situ prior to fixation. In heat-shocked cells permeabilized before fixation, hsp27 shows in the absence of detergent and immunoblotting of the various subcellular fractions with the anti-hsp27 antibody. As shown in Fig. 6, hsp27 is mainly enriched in the 1000g nuclear pellet (lane 1) and in the high-speed supernatant (lane 4).

Fig. 5.

Localization of hsp27 in ecdysone-treated cells. Cells were grown on microscope slides in the complete medium supplemented with 10 μM-ecdysone for 20 h. Cells were then fixed and stained for the detection of hsp27. A. Fluorescent micrograph; and B, corresponding phase-contrast image. Bar, 10 μm.

Fig. 5.

Localization of hsp27 in ecdysone-treated cells. Cells were grown on microscope slides in the complete medium supplemented with 10 μM-ecdysone for 20 h. Cells were then fixed and stained for the detection of hsp27. A. Fluorescent micrograph; and B, corresponding phase-contrast image. Bar, 10 μm.

Fig. 6.

Subcellular distribution of hsp27 in ecdysone-treated cells. Cells grown in the presence of ecdysone were lysed in the absence of detergent and fractionated by differential centrifugation. Proteins from the 1000(lane 1), 10000g (lane 2) and 100000g (lane 3) pellets and from the 100 000g supernatant (lane 4) were separated by electrophoresis, transferred to nitrocellulose and immunoblotted with the anti-hsp27 antibody as for Fig. 1.

Fig. 6.

Subcellular distribution of hsp27 in ecdysone-treated cells. Cells grown in the presence of ecdysone were lysed in the absence of detergent and fractionated by differential centrifugation. Proteins from the 1000(lane 1), 10000g (lane 2) and 100000g (lane 3) pellets and from the 100 000g supernatant (lane 4) were separated by electrophoresis, transferred to nitrocellulose and immunoblotted with the anti-hsp27 antibody as for Fig. 1.

In situ nuclear association of hsp27 during heat shock

The subcellular fractionation data reported above show that hsp27 can be rapidly extracted by non-ionic detergents in cells where this hsp is induced by ecdysone but a prominent nuclear localization (Fig. 7A-B) similar to that observed in unpermeabilized cells (compare with Figs 2 and 3). In the majority of cells, the nucleoli do not appear to be stained by the antibody but rather appear to be surrounded by a hsp27-containing ring. Deposition of hsp27 is also observed in phase-dense structures and at the border of the nucleus. In contrast, no nuclear or cytoplasmic fluorescence can be observed after detergent pre-permeabilization in ecdysone-induced cells (Fig. 7C-D). A similar absence of nuclear staining is observed in heat-shocked cells allowed to recover for more than 1 h (not shown). From these experiments, it is tentatively concluded that hsp27 is tightly bound (in a detergent-resistant form) to some nuclear component(s) during severe heat-shock conditions only. In the course of cellular recovery or in ecdysone-stimulated cells, the anti-hsp27 also decorates the nucleus but hsp27 is in a form that is readily solubilized by non-ionic detergents.

Fig. 7.

Localization of hsp27 in heat-shocked or ecdysone-treated cells permeabilized prior to fixation. Cells grown on microscope slides were incubated at 37°C for 1 ·5 h (A,B) or cultured with 10μM-ecdysone for 20 h (C,D) and permeabilized with 0·5% NP-40 before fixation and antibody staining. A,C. Fluorescent micrographs with the same exposure time; B,D, corresponding phase-contrast image. Bar, 10 μm.

Fig. 7.

Localization of hsp27 in heat-shocked or ecdysone-treated cells permeabilized prior to fixation. Cells grown on microscope slides were incubated at 37°C for 1 ·5 h (A,B) or cultured with 10μM-ecdysone for 20 h (C,D) and permeabilized with 0·5% NP-40 before fixation and antibody staining. A,C. Fluorescent micrographs with the same exposure time; B,D, corresponding phase-contrast image. Bar, 10 μm.

In the present study, the localization of hsp27 in Drosophila S3 cells was investigated by both indirect immunofluorescence microscopy and cellular fractionation techniques. Immunocytological evidence is presented that suggests that hsp27 is a formal nuclear protein. However, as shown in cell fractionation studies, binding of this protein to certain nuclear component(s) can differ greatly depending on whether cells are exposed or not to high temperatures. We have observed that hsp27 remains bound to insoluble nuclear structures in cells heat-shocked at 37°C, while it is readily solubilized by detergents in stressed cells that are allowed to recover or under conditions of hormonal induction by ecdysone.

The co-fractionation of the small hsps (22–27 × 103Mr) with the nucleus of heat-shocked cells has been reported in various cell lines (summarized by Tanguay, 1985; Subjeck & Shyy, 1986). The precise nuclear localization of these proteins during heat shock remains unclear, however. High concentrations of small hsps were detected in chromatin and nucleoli preparations (Arrigo et al. 1980). These hsps were also reported as constituents of ribonucleoprotein particles (Kloetzel & Bautz, 1983). This group of proteins has also been found to be highly resistant to salt extraction (Tanguay & Vincent, 1982) and to nuclease treatment, suggesting a structural function for small hsps either as nucleoskeletal elements (Levinger & Varshavsky, 1981; Sinibaldi & Morris, 1981) or as extranuclear cytoskeletal elements that could have collapsed on the nuclear surface (Tanguay et al. 1985) as the result of an extensive cytoskeletal rearrangement induced by heat shock (Falkner et al. 1981; Vincent & Tanguay, 1982). In the present work, hsp27 has been immunolocalized in the perinucleolar region and in phase-dense nuclear structures of heat-shocked cells permeabilized prior to fixation. This suggests an association of this protein with an unidentified nuclear structural element. Interaction with DNA appears unlikely, as the small hsps are found in the nuclear matrix-intermediate filament fraction but not in the chromatin fraction of Drosophila cells fractionated according to the method of Fey et al. (1984) (Rollet et al. unpublished data).

In contrast with previous biochemical fractionation studies, which suggested a nuclear-cytoplasmic shuttling behaviour of the small hsps during heat shock and recovery, the present study offers an alternative view for hsp27. The immunofiuorescent data indicate that the anti-hsp27 decorates the nucleus not only in heat-shocked cells but also after recovery or after induction by hormonal stimulation. However, the strength of its association with nuclear components as determined by resistance to detergent extraction, varies under these different physiological conditions. This behaviour is somewhat reminiscent of that of hsp70 in COS cells. In this system, although immunofluorescence indicates that a substantial portion of hsp70 is localized in the nucleus of unstressed cells, only 2% of this hsp is recovered in the NP-40-insoluble nuclear pellet of unstressed cells compared to 41 % in heat-shocked cells (Lewis & Pelham, 1985). An ATP-dependent release of hsp70 from the nucleus has been reported in mammalian (Lewis & Pelham, 1985), avian (Collier & Schlesinger, 1986) and Drosophila cells (Beaulieu & Tanguay, 1988). However, as we have shown previously, the small hsps are not released from nuclei by ATP extraction, suggesting that their association with nuclear components differs from that of hsp70 (Beaulieu & Tanguay, 1988). Non-ionic detergents such as NP-40 are generally believed to bind to the hydrophobic regions of proteins, thereby displacing the lipids associated with these regions. They generally do not break up proteinprotein interactions. Sequence analysis of the hsp27 gene indicates the presence of a hydrophobic region at the amino terminus, suggesting a potential capacity to interact with membranes (Southgate et al. 1983). Alternatively, hsp27 could behave like many nucleoplasmic proteins, leaking out from nuclei when their membranes are damaged by the detergent. In heat-shocked cells, it may remain bound to some components of the nucleus.

In mammalian cells, hsp28 also fractionates with the nuclear pellet immediately after heat shock and is subsequently found in the cell supernatant after recovery (Arrigo & Welch, 1987). Interestingly, a considerable portion of the nuclear-associated mammalian hsp28 is of the phosphorylated form (28×103Mr b isoform). In cultured Drosophila cells, heat shock and hormonal treatment (albeit at a lower level) have been reported to induce phosphorylation of both hsp26 and hsp27 (Rollet & Best-Belpomme, 1986). Such post-translational modifications could promote and/or strengthen interactions with nuclear components. Alternatively, the increased binding as reflected by resistance to non-ionic detergents could result from heat-induced formation of insoluble nuclear complexes as reported in human cells (Littlewood et al. 1987). While the nature of the binding of the small hsps with cellular components has not been determined, several lines of evidence suggest that this phenomenon occurs in a temperature-dependent way. First, small hsps are found mainly in the detergentsoluble extract of Drosophila cells that have been heat-shocked at milder temperatures of 33 °C or 35 °C instead of 37°C (Vincent & Tanguay, 1982). Second, when induced by other stresses such as an ethanol treatment, the small hsps are not found in a nucleus-associated form (Rollet et al. unpublished data). Third, ecdysone, an insect steroid hormone, has been reported to induce synthesis of the smalls hsps, which are found exclusively in a detergent-soluble form (Ireland & Berger, 1982). This latter case is particularly interesting, since the induction of the small hsps by physiological concentrations of ecdysone can be clearly distinguished from a stress response. Thus Riddihough & Pelham (1987) have recently shown that the activation of hsp27 by ecdysone involves promoter sequences different from those required for heat shock. The present nuclear immunolocalization of hsp27 in both heat-shocked and ecdysone-stimulated cells suggests a nuclear function under both conditions. Finally, the rapid change from an insoluble to a soluble form upon recovery from heat shock appears to be a consequence of modification in the binding capacity, specific or not, of hsp27 rather than a shift from one cellular compartment to another.

The results presented here also show that hsp27 has a different cellular localization from the other small hsps of Drosophila, although all four hsps share extensive homology (Ingolia & Craig, 1982; Southgate et al. 1983). Thus in our fractionation study of heat-shocked cells, hsp27 is almost exclusively found in the nuclear pellet while the other small hsps are also found in the detergentsoluble extract (see Fig. IB). During development in Drosophila, 40% of hsp27 is found associated with a particulate fraction while, under the same conditions, hsp23 is essentially soluble (Arrigo & Pauli, 1988). Using an antibody against hsp23, we have previously shown that this hsp is localized in the nucleolus and in dense cytoplasmic aggregates during heat shock and rapidly returns to the cytoplasm during recovery (Duband et al. 1986). hsp27 does not stain the nucleolus and appears to remain associated with the nucleus even during recovery. Leicht et al. (1986) suggested that the small hsps of Drosophila cells were associated with the cytoskeleton, on the basis of their co-localization with intermediate filaments using a monoclonal antibody recognizing hsp23 and hsp26. In chick embryo fibroblasts, hsp24 is also localized in cytoplasmic aggregates, which can form perinuclear aggregates after heat shock (Collier & Schlesinger, 1986). In maize, hsp25 cofractionated with organelles containing Golgi and endoplasmic reticulum markers while 50% of hsp 18 copurified with nuclei (Cooper & Ho, 1987). Thus, while the precise nature of the nuclear association of Drosophila hsp27 remains to be determined, both fractionation and immunolocalization data show that this hsp has a different localization and behaviour from hsp23. This suggests that the different members of the small hsp gene family, although quite similar at the gene level, could have different functions as reflected by their different affinity for various cellular compartments during heat shock or under more physiological conditions.

The present results point out that hsp27 is associated with the nucleus in both heat-shocked and ecdysonestimulated cells. In unstressed cells, it is also found associated with the nucleus but in only 2 % of cells. Thus, as suggested previously for hsp23 (Duband et al. 1986), the low and variable level of expression of hsp27 as assessed by immunoblotting or labelling appears to reflect expression in a few cells rather than a basal level of expression in the entire cellular population.

We thank Dr E. Rollet and Ms F. Lettre for technical assistance, Drs M. Vincent and P. Rogers for reviewing the manuscript and Mrs M. A. Godbout for typing. This work was supported by the Medical Research Council of Canada (PG35). J.F.B. was supported by a postdoctoral fellowship from the ‘Fonds de la Recherche en Santé du Québec’.

Arrigo
,
A. P.
(
1987
).
Cellular localization of hsp 23 during Drosophila development and following subsequent heat shock
.
Devi Biol
.
122
,
39
48
.
Arrigo
,
A. P.
,
Fakan
,
S.
&
Tissières
,
A.
(
1980
).
Localization of the heat shock-induced proteins in Drosophila melanogaster tissue culture cells
.
Devi Biol
.
78
,
86
103
.
Arrigo
,
A. P.
&
Pauli
,
D.
(
1968
).
Characterization of hsp 27 and three immunologically related polypeptides during Drosophila development
.
Expl Cell Res
.
175
,
169
183
.
Arrigo
,
A. P.
&
Welch
,
W. J.
(
1987
).
Characterization and purification of the small 28,000-dalton mammalian heat shock protein
.
J. biol. Chem
.
262
,
15 359
15 369
.
Atkinson
,
B. G.
&
Walden
,
D. B.
, eds (
1985
).
Changes in Eukaryotic Gene Expression in Response to Environmental Stress
.
OrLondo
:
Academic Press
.
Beaulieu
,
J. F.
&
Tanguay
,
R. M.
(
1988
).
Members of the Drosophila HSP 70 family share ATP-binding properties
.
Eur. J. Biochem
.
172
,
341
347
.
Berger
,
E. M.
&
Woodward
,
M. P.
(
1983
).
Small heat shock proteins in Drosophila may confer thermal tolerance
.
Expl Cell Res
.
147
,
437
442
.
Burke
,
B.
,
Griffiths
,
G.
,
Reggio
,
H.
,
Louvard
,
D.
&
Warren
,
G.
(
1982
).
A monoclonal antibody against a 135-K Golgi membrane protein
.
EMBO J
.
1
,
1621
1628
.
Cooper
,
P.
&
Ho
,
T. D.
(
1987
).
Intracellular localization of heat shock proteins in maize
.
Pl. Physiol
.
84
,
1197
1203
.
Craig
,
E.
(
1985
).
The heat shock response
.
CRC Crit. Rev. Biochem
.
18
,
239
280
.
Collier
,
N. C.
&
Schlesinger
,
M. J.
(
1986
).
The dynamic state of heat shock proteins in chicken embryo fibroblasts
.
J. Cell Biol
.
103
,
1495
1507
.
Duband
,
J. L.
,
Lettre
,
F.
,
Arrigo
,
A. P.
&
Tanguay
,
R. M.
(
1986
).
Expression and localization of hsp 23 in unstressed and heat shocked Drosophila cultured cells
.
Can. J. Genet. Cvtol
.
28
,
1088
1092
.
Falkner
,
F. G.
,
Saumweber
,
H.
&
Biessmann
,
H.
(
1981
).
Two Drosophila melanogaster proteins related to intermediate filament proteins of vertebrate cells
.
J. Cell Biol
.
91
,
175
183
.
Fey
,
E. G.
,
Wan
,
K. M.
à
Penman
,
S.
(
1984
).
Epithelial cytoskeletal framework and nuclear matrix-intermediate filament scaffold: three-dimensional organization and protein composition
.
J. Cell Biol
.
98
,
1973
1984
.
Ingolia
,
T. D.
&
Craig
,
E. A.
(
1982
).
Four small Drosophila heat shock proteins are related to each other and to mammalian a- crystallin
.
Proc. natn. Acad. Sci. U.SA
.
79
,
2360
2364
.
IreLond
,
R. C.
&
Berger
,
E. M.
(
1982
).
Synthesis of low molecular weight heat shock peptides stimulated by ecdysterone in a cultured Drosophila cell line
.
Proc. natn. Acad. Sci. U.SA
.
79
,
855
859
.
IreLond
,
R. C.
,
Berger
,
E.
,
Sirotkin
,
K.
,
Yund
,
M. A.
,
Osterbur
,
D.
&
Fristrom
,
J.
(
1982
).
Ecdysterone induces the transcription of four heat-shock genes in Drosophila S3 cells and imaginai discs
.
Devi Biol
.
93
,
498
507
.
Johnson
,
D. A.
,
Gautsch
,
J. W.
,
Sportsman
,
J. R.
&
Elder
,
J. H.
(
1984
).
Improved technique utilizing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose
.
Gene Anal. Techn
1
,
3
8
.
Kloetzel
,
P. M.
&
Bautz
,
E. K. F.
(
1983
).
Heat-shock proteins are associated with hnRNA in Drosophila melanogaster tissue culture cells
.
EMBOJ
.
2
,
705
710
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
.
Nature, Lond
.
227
,
680
685
.
Leicht
,
B. G.
,
Biessmann
,
H.
,
Palter
,
K. B.
&
Bonner
,
J. J.
(
1986
).
Small heat shock proteins of Drosophila associate with the cytoskeleton
.
Proc. natn. Acad. Sci. U.SA
.
83
,
90
94
.
Levinger
,
L.
&
Varskavsky
,
A.
(
1981
).
Heat-shock proteins of Drosophila are associated with nuclease-resistant, high-salt-resistant nuclear structures
.
J. Cell Biol
.
90
,
793
796
.
Lewis
,
M. J.
&
Pelham
,
H. R. B.
(
1985
).
Involvement of ATP in the nuclear and nucleolar functions of the 70 KD heat shock protein
.
EMBOJ
.
4
,
3137
3143
.
Lindquist
,
S.
(
1986
).
The heat shock response. A
.
Rev. Biochem
.
55
,
1151
1191
.
Littlewood
,
T. D.
,
Hancock
,
D. C.
&
Evan
,
G. I.
(
1987
).
Characterization of a heat shock-induced complex in the nuclei of cells
.
J. Cell Sci
.
88
,
65
72
.
Nover
,
L.
(
1984
).
Heat Shock Response of Eucaryotic Cells
.
Leipzig
:
Veb Georg thieme
.
Pelham
,
H. R. B.
(
1986
).
Speculations on the functions of the major heat shock and glucose-regulated proteins
.
Cell
46
,
959
961
.
Riddihough
,
G.
&
Pelham
,
H. R. B.
(
1987
).
Activation of the Drosophila hsp27 promoter by heat shock and ecdysone involves independent and remote regulatory sequences
.
EMBO J
.
5
,
1653
1658
.
Rollet
,
E.
&
Best-Belpomme
,
M.
(
1986
).
Hsp 26 and 27 are phosphorylated in response to heat shock and ecdysterone in Drosophila melanogaster cells
.
Biochem. biophvs. Res. Commun
.
141
,
426
433
.
Schlesinger
,
M. J.
(
1986
).
Heat shock proteins: the search for functions
.
J. Cell Biol
.
103
,
321
325
.
Schlesinger
,
M. J.
,
Ashburner
,
M.
&
Tissieres
,
A.
(
1982
).
Heat Shock from Bacteria to Man
.
Cold Spring Harbor, NY
:
Cold Spring Harbor Laboratory Press
.
Sinibaldi
,
R.
&
Morris
,
P. W.
(
1981
).
Putative function of Drosophila melanogaster heat shock proteins in the nucleoskeleton
.
J. biol. Chem
.
256
,
10735
10738
.
Southgate
,
R.
,
Aymé
,
A.
&
Voellmy
,
R.
(
1983
).
Nucleotide sequence analysis of the Drosophila small heat shock gene cluster at locus 67B
.
J. molec. Biol
.
165
,
35
57
.
Subjeck
,
J. R.
&
Shyy
,
T. T.
(
1986
).
Stress protein systems of mammalian cells
.
Am. J. Physiol
.
250
,
G1
G17
.
Tanguay
,
R. M.
(
1985
).
Intracellular localization and possible functions of heat shock proteins
.
In Changes in Eukaryotic Gene Expression in Response to Environmental Stress
(ed.
B. G.
Atkinson
&
D. B.
Walden
), pp.
91
113
.
OrLondo
:
Academic Press
.
Tanguay
,
R. M.
,
Duband
,
J. L.
,
Lettre
,
F.
,
Valet
,
J. P.
,
Arrigo
,
A. P.
&
Nicole
,
L.
(
1985
).
Biochemical and immunocytochemical localization of heat-shock proteins in Drosophila cultured cells
.
Ann. N.Y. Acad. Sci
.
455
,
712
714
.
Tanguay
,
R. M.
&
Vincent
,
M.
(
1982
).
Intracellular translocation of cellular and heat shock induced proteins upon heat shock in Drosophila Kc cells
.
Can J. Biochem
.
59
,
67
73
.
Thomas
,
J. O.
&
Kornberg
,
R. D.
(
1975
).
An octamer of histone in chromatin and free in solution
.
Proc. natn. Acad. Sci. U.SA
.
72
,
2626
2630
.
Towbin
,
H.
,
Staehelin
,
T.
&
Gordon
,
J.
(
1979
).
Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications
.
Proc. natn. Acad. Sci. U.SA
.
76
,
4350
4354
.
Vincent
,
M.
&
Tanguay
,
R. M.
(
1982
).
Different intracellular distributions of heat-shock and arsenite-induced proteins in Drosophila Kc cells
.
J. molec. Biol
.
162
,
365
379
.