INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE HEAT SHOCK PROTEINS (HSP) are a family of endogenous, protective proteins that are upregulated to varying degrees in response to stress (9, 12, 13, 15, 17). Overexpression of HSP72, the inducible HSP70, in cardiac muscle protects the heart from ischemia-reperfusion injury and reduces infarct size (9, 10, 15, 17). Selective inhibition of HSP72 expression by treating isolated, adult feline cardiac myocytes with antisense oligonucleotides increased the susceptibility of these cells to injury by hypoxia and reoxygenation (13), showing the importance of HSP72 for cardiac protection. HSP72 is a protein that moves to the nucleus selectively. Overexpression of HSP72 results in cytoplasmic localization of the protein; only with stress, such as heat shock, does the protein concentrate in the nucleus. Brief cardiac ischemia is followed by nuclear accumulation of HSP72 in the myocardium (20). Given the protective functions of HSP72 in stressed cells and the important role of HSP72 in accelerating cell recovery from injury, further understanding of its function and regulation of its nuclear accumulation are key to clarifying its protective mechanism(s) in the cell.

We hypothesized that nuclear accumulation of HSP72 was an important part of its protective function. Furthermore, we postulated that phosphorylation of HSP72 controls the rate of nuclear accumulation of the protein. Because tyrosine kinases are important in intracellular signaling and changes in total cellular tyrosine phosphorylation have been reported with heat shock, changes in tyrosine phosphorylation might influence nuclear concentration of HSP72 (8). A single tyrosine phosphorylation site (Y⁵²⁴) is predicted for human HSP72. We tested our hypothesis in a series of experiments reported here showing that: 1) phosphorylation of Y⁵²⁴ occurs with heat shock, 2) inhibition of tyrosine kinase prevents nuclear accumulation of HSP72 with stress, and 3) site-directed mutation of Y⁵²⁴ to phenylalanine greatly slows nuclear concentrations of HSP72 with heat shock. Finally, we demonstrate that site-directed mutation of Y⁵²⁴ to aspartic acid, generating a permanent negative charge for a "pseudophosphorylation," does not inhibit nuclear movement; rather, it accelerates nuclear accumulation with stress and improves survival after severe heat shock.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture. Cos-1 cells (ATCC, American Type Culture Collection, Rockville, MD) were grown in DMEM with 10% fetal bovine serum, 50 U/ml and 50 µg/ml of penicillin and streptomycin, respectively. All tissue culture supplies were purchased from GIBCO-BRL (Gaithersburg, MD) except for plasticware (Falcon, Becton-Dickinson, Bedford, MA). For localization experiments, cells were plated on sterile glass coverslips (Bellco Glass, Vineland, NJ) in petri dishes.

Constructs. The BamH I-Hind III region of the human HSP72 cDNA (2.3 kb, ATCC) (4) was subcloned into the Hind III site of pRc/CMV (Invitrogen, Carlsbad, CA) and was used for the insertion of the epitope tag and for the mutations. Restriction enzymes were purchased from New England BioLabs (Beverly, MA).

Mutations. Insertion of an epitope tag and deletion and modification mutations were done using the method of site-directed mutagenesis described by Deng and Nickoloff (2). Briefly, using a commercial kit (Clontech, Palo Alto, CA), mutagenic oligonucleotides were annealed to plasmid DNA. Concomitantly, a second oligonucleotide was used to mutate a unique enzyme-restriction site in the plasmid for the purpose of selecting against paternal unmutated plasmid. Oligonucleotides were synthesized by the Core Molecular Biology Facility at Baylor College of Medicine. A unique epitope tag, flag (DYKDDDDK), was inserted at the carboxyl terminus immediately ahead of the stop codon. A commercial antibody, antiflag, is available for this tag (IBI, New Haven, CT). All mutations were confirmed by sequencing using a PCR-based sequencing kit (fmol, Promega, Madison, WI). When appropriate, a commercially available in vitro transcription and translation kit (Promega) was used to verify that the mutant constructs produced 70-kDa protein products.

Immunoprecipitation and Western blotting. Parallel immunoprecipitations were carried out with anti-HSP72 (clone C92F3A-5, Stress-Gen, Victoria, Canada) and with anti-phosphotyrosine (PY20, an IgG2, BD Transduction Laboratories, Franklin Lakes, NJ). Cos-1 cells were heat shocked for up to 1 h at 43.5°C. Cells were collected in radioimmunoprecipitation assay (RIPA) buffer with phosphatase and protease inhibitors [50 mM Tris · HCl, 150 mM NaCl, pH 7.4, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 mM Na₃VO₄, 1 mM NaF, 1% NP-40, 0.25% sodium deoxycholate, 1 µg/ ml each of aprotinin, leupeptin, and pepstatin]. The lysates were precleared by mixing on a rocker panel at 4°C with 20 µl of protein G-agarose (Sigma, St. Louis, MO) for 20 min followed by sedimentation at maximal speed for 5 min in a microcentrifuge. The cell lysates were then immunoprecipitated with either anti-HSP72 (clone C92F3A, mouse monoclonal, StressGen, Vancouver, Canada) or anti-phosphotyrosine overnight. Protein G-agarose was used to precipitate the antibody-antigen complex in a second overnight incubation. After the third wash with RIPA buffer, sample buffer without Coomassie dye was added, and the samples were separated on a 10% SDS-PAGE mini gel. Prestained molecular weight markers (GIBCO-BRL) were used for size identification of the samples. Samples were then transferred to nitrocellulose as previously described (5). The nitrocellulose was blocked with 2% gelatin for the anti-phosphotyrosine antibody and with 5% Blotto for the anti-HSP72 antibody and was developed as previously described using the enhanced chemiluminescent (ECL) system (6) (Amersham, Arlington Heights, IL). The Western blots of the HSP72 immunoprecipitates were developed with anti-phosphotyrosine at a 1:2,500 concentration followed by anti-mouse IgG2-HRP (Southern Biotechnology, Birmingham, AL) at a 1:2,000 concentration. For the anti-phosphotyrosine immunoprecipitants, anti-HSP72 was used at a 1:2,500 concentration, followed by anti-mouse IgG-HRP (Amersham) at a 1:1,000 concentration. Western blots were then developed by exposure to Hyperfilm MP (Amersham). To compare the amount of HSP72 present in each lane of the anti-HSP72 immunoprecipitant, the blot was stripped (2% SDS and 100 mM mercaptoethanol in 62.5 mM Tris · HCl, pH 6.8, for 30 min at 70°C), reexposed to film to confirm the absence of signal, and then redeveloped with anti-HSP72 antibody.

Nuclear-to-cytoplasm ratio. Samples were collected over a 1-h heat-shock time course as described above. Cellular fractionation was done using the approach described by Huang et al. (3). Cells were washed with PBS twice and then scraped into 1 ml of PBS containing protease inhibitors (1 µg/ml each of aprotinin, leupeptin, pepstatin, and 1 mM PMSF). Phosphatase inhibitors were added to all steps as above. The lysate was then centrifuged at 270 g, 4°C, for 10 min. The supernatant was saved as fraction A. The pellet was resuspended in 600 µl of nuclei isolation buffer (60 mM KCl, 15 mM NaCl, 15 mM HEPES, 2 mM EDTA, 0.5 mM EGTA, 0.15 mM spermine, 0.5 mM spermidine, 14 mM mercaptoethanol, 10% sucrose, and 0.1% NP-40) and placed on ice for 5 min. The preparation was centrifuged at 220 g, 4°C, for 5 min. The pellet was resuspended in 300 µl of glycerol storage buffer [50% glycerol, 20 mM Tris · HCl, pH 7.9, 75 mM NaCl, 0.5 mM EDTA, 0.85 mM dithiothreitol (DTT), and 0.1 mM PMSF] and then centrifuged at 13,000 g for 1 min at 4°C. The pellet was resuspended in 500 µl of the PBS with protease inhibitors to which 1% NP-40 had been added. The preparation was then spun for 2 min at 13,000 g at 4°C. The pellet was resuspended in 600 µl of RIPA buffer (50 mM Tris · HCl, 150 mM NaCl, 1 mM EGTA, 1% NP-40, 0.25% sodium deoxycholate, 1 mM PMSF, and 1 µg/ml each of aprotinin, leupeptin, and pepstatin). After the integrity of the nuclei was verified by examination with light microscopy, the nuclei were lysed by sonication. The preparation was then centrifuged at 16,000 g at 4°C, and the supernatant was saved as fraction D. The nuclei (fraction D) were sonicated to disrupt the membranes. Both fraction A and fraction D were immunoprecipitated with anti-HSP72-agarose as described in Immunoprecipitation and Western blotting. Samples were then separated by SDS-PAGE and analyzed by Western blotting as described earlier. Blots were stripped and reprobed with anti-HSP72. Densities were determined for both anti-phosphotyrosine bands and HSP72 bands for both nuclei and cytoplasm. A ratio of phosphotyrosine to HSP72 was determined for each sample, and a nuclei (phosphotyrosine/HSP72)-to-cytoplasm (phosphotyrosine/HSP72) ratio was calculated.

Immunofluorescence protocol. Perfect Lipid-1 (Invitrogen) was used to transfect Cos-1 cells grown on coverslips in petri dishes. Three variations of HSP72fl were used: WT (wild-type sequence), M17, and M18. After a recovery period of 24 h, the transfected Cos-1 cells were heat shocked by placement in a specially modified incubator with two heat pumps maintained at 43.5°C, followed by incubation at 37°C when a recovery period was included. Cos-1 cells were then fixed with PBS-buffered formaldehyde and made permeable with ethanol, exactly as detailed by Kriegler (7). Cells were then incubated with epitope tag-specific antibody (antiflag from IBI, diluted 1:500 in 3% BSA) for a period of 30 min followed by affinity-purified anti-mouse IgG-FITC (diluted 1:200 in 3% BSA, from The Binding Site, San Diego, CA). Coverslips were mounted with Gel/Mount (Biomeda, Foster City, CA) on glass slides, sealed, and examined. Incubation with anti-mouse IgG-FITC alone resulted in slides with no signal.

Exhaustive photon reassignment. Exhaustive photon reassignment (EPR) (Scanalytics, CSPI, Billerica, MA) was used to analyze the intracellular localization of the HSP72fl constructs. Images were collected with a Nikon fluorescence microscope equipped with a CCD camera. To aid in definitive identification of the nuclei, the cells were counterstained with 0.2 µg/ml propidium iodide (PPI) for some of the experiments. Two different Nikon filter combinations, B2E (dichroic 510, excitation filter 450-490, and barrier filter 520-560) for blue excitation and G-2B (dichroic 580, excitation filter 510-560, barrier filter 610) for green excitation, were used to visualize the FITC-labeled antibody and the PPI, respectively. With this combination of filters and dilute PPI, the signal from the PPI was only visible with green excitation. The images were collected at 0.25-µm increments, background was subtracted, and the images were deconvoluted overnight on a Pentium computer equipped with Cellview software (Scanalytics). The middle image (e.g., section 10 of 20) was used to generate a nuclear-to-cytoplasm ratio (N/C). A representative area of the cytoplasm and the nucleus was sampled, and the N/C calculated. Data were analyzed by combining the results of multiple slides from three experiments. For each group, 10-30 cells were analyzed.

Viability. Ethidium bromide uptake was used to assess viability. Cells were cotransfected with enhanced blue fluorescent protein (EBFP) expressed under the control of a Rous sarcoma virus (RSV) promoter (Clontech). A 10-fold molar ratio of HSP construct to EBFP construct was used. Cells were heat shocked at either 43.5 or 45°C, 24 h after transfection. After a 2-h recovery period, cells were washed twice with PBS, and ethidium bromide (20 ng/ml) in PBS was added to the cells (19). Plates were processed individually, and cells were immediately analyzed under a 4,6-diamidino-2-phenylindole (DAPI) filter for the EBFP and under green light, as described in Exhaustive photon reassignment, to detect the ethidium. Slides were processed without knowledge of the treatment or expressed HSP construct. A minimum of 20 cells were scored per slide. The distinctive blue of the EBFP, distributed throughout the cell, was readily visualized. With heat shock for the WT and M18 constructs at 45°C, very few cells were present. If less than 10 cells could be identified on an entire slide, in contrast to the abundant cells present in the absence of heat shock, the slide was scored as 0. This occurred once for the WT and three times for the M18 constructs.

ATP binding studies. ATP binding studies were performed as described by Milarski and Morimoto (11). In vitro transcription and translation were performed by using the Promega kit and adding [³⁵S]methionine to label the final product (Amersham). ATP-agarose beads (Sigma) were added to the translate after it had been diluted 40-fold in buffer B (20 mM Tris · HCl, 20 mM NaCl, 0.1 M EDTA, 2 mM DTT, pH 8.0). Protease inhibitors were added to all steps (1 mM PMSF and 1 µg/ml each of aprotinin, antipain, leupeptin, and pepstatin). After incubation for 2 h at 4°C on a rocker panel, the beads were precipitated in a microcentrifuge and washed three times with ice-cold buffer. The HSP72 was released from the beads by competing with 20 mM ATP. Immunoprecipitation was carried out overnight with antiflag-agarose beads as detailed earlier. After precipitating and washing the samples three times, we resuspended them in sample buffer and separated them on a 10% SDS-PAGE mini gel. Both the precipitate and the supernatant after treatment with ATP-agarose were analyzed. The gel was treated with Amplify (Amersham), dried, and exposed to Hyperfilm (Amersham).

Statistical analysis. SigmaStat (Jandel, San Rafael, CA) was used to perform statistical analysis where appropriate. Depending on the data, either an ANOVA or an ANOVA on ranks (Kruskal-Wallis) test was run, followed by a Student-Newman-Keuls, Dunn's, or Dunnett's test, as appropriate. P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Evidence that HSP72 is phosphorylated at Y⁵²⁴ after heat shock. In a series of experiments, control and heat-shocked Cos-1 cells were immunoprecipitated with anti-HSP72 antibody, separated on an SDS-PAGE mini gel, and transferred to nitrocellulose. Cells were collected for analysis over a 2-h interval because it was suspected that phosphorylation of Y⁵²⁴, the only tyrosine phosphorylation site predicted in HSP72, might be transient. Development of these blots with anti-phosphotyrosine demonstrated that HSP72 had a phosphotyrosine but only after the 30-min heat shock (Fig. 1). We immunoprecipitated with the antibody used to probe the blots (anti-phosphotyrosine) and found that this brought down HSP72, as shown in Fig. 1. To compare the amounts of HSP72 precipitated with anti-HSP72 (Fig. 1B) and anti-phosphotyrosine and show that the observed change in phosphorylation was not due to differences in the amount of protein precipitated, the blots were stripped and redeveloped with anti-HSP72 antibody. Figure 1C suggests that, if anything, more HSP72 was precipitated from the control sample than from the heat-shocked samples. Thus the differences observed in phosphotyrosine were not related to differences in the amount of HSP72 in these samples.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Immunoprecipitation of heat-shock protein 72 (HSP72) from Cos-1 cells following heat-shock time course. A: cell lysates were immunoprecipitated with anti-phosphotyrosine, and Western blots were developed with anti-HSP72. B: HSP72 was immunoprecipitated with anti-HSP72 antibodies, and Western blots were developed with anti-phosphotyrosine antibodies. C: Western blot in B was stripped and developed with anti-HSP72 antibodies to show relative amounts in each lane. Lane 1, no heat shock; lane 2, 30-min heat shock; lane 3, 60-min heat shock; lane 4, 60-min heat shock and 60-min recovery at 37°C. Arrow indicates band at 70 kDa. Thus tyrosine phosphorylation occurred in HSP72 after heat shock lasting 30 min or longer.

Nuclear accumulation of phosphorylated protein. To determine whether phosphorylation facilitates nuclear accumulation of HSP72, cells were fractionated into nuclear and cytoplasmic fractions after: 1) no treatment, 2) heat shock for 30 min, or 3) heat shock for 60 min. All heat shock was done at 43.5°C. The fractions were immunoprecipitated with anti-HSP72 antibody bound to agarose beads. Western blots of the immunoprecipitates (Fig. 2A) were developed with anti-phosphotyrosine, stripped, and reprobed with anti-HSP72. The results were recorded, the blots were stripped, and then the blots were redeveloped with anti-HSP72. Immunoprecipitation is qualitative rather than quantitative; therefore the ratio of tyrosine-phosphorylation-HSP72 was determined for each fraction (as described in METHODS) to correct for differences in immunoprecipitation. The N/C ratio was then determined. The amount of tyrosine phosphorylation of nuclear HSP72 increased sharply at 30 min. In contrast, almost no phosphorylation of cytoplasmic HSP72 was seen at baseline.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   HSP72 in nuclear and cytoplasmic extracts after heat shock. A: anti-HSP72 immunoprecipitates with Western blots developed with anti-phosphotyrosine antibody (P) at 30 and 60 min showing that this antibody reacts with HSP72 strongly after heat shock, and that this band moves from cytoplasm to nucleus with time. Several other bands were seen from other phosphotyrosines coprecipitating with HSP72. Blots were stripped and developed with anti-HSP72 antibody to verify location of HSP72 and to determine relative amount of HSP72 present (lower Western blot, H). Variation in amount of protein represents variation in immunoprecipitation efficiency. C, cytoplasm, N, nucleus. Groups are control (no heat shock), and 30 or 60 min of heat shock. Phosphorylated HSP72 appears in nucleus by 30 min of heat shock. B: summary of changes in nuclear-to-cytoplasm ratio (N/C) of phosphotyrosine-HSP72 over heat-shock time course. Results are means of four separate experiments. Clearly as early as 30 min an increase in nuclear phosphotyrosine-HSP72 is seen, and this declines with time. *P < 0.05 vs. time 0.

Effect of tyrosine kinase inhibition on nuclear accumulation with heat shock. Cos-1 cells were heat shocked for 2 h at 42°C, recovered for 2 h, and then examined by immunofluorescence microscopy to determine the effect of geldanamycin, a tyrosine kinase inhibitor, on nuclear localization of HSP72 with heat shock. As shown in Fig. 3, treatment with geldanamycin at two different concentrations prevented nuclear accumulation of HSP72 with heat shock.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Cos-1 cells were treated with geldanamycin during heat shock. As shown, summarizing N/C from immunocytochemistry, geldanamycin treatment blocked nuclear accumulation of HSP72 with heat shock. G2 and G10 represent 2 and 10 µg/ml doses of geldanamycin (3.56 µM and 17.8 µM), similar to doses used in literature. C, no heat shock, no geldanamycin. C-HS, heat shock, but no geldanamycin. Results of one of multiple experiments are shown. Each bar represents 15-35 analyzed cell images. *P < 0.05 vs. all others.

Mutation of Y⁵²⁴ and its effect on nuclear accumulation. Two tagged mutant constructs of HSP72 were made: M17 and M18. M17 has an aspartic acid substituted for Y⁵²⁴, and M18 has a phenylalanine at Y⁵²⁴. In pilot experiments comparing transfected WT HSP72 and WT HSP72fl (with an epitope tag), there was no difference in nuclear concentration of the proteins following heat shock. Nuclear concentration of the mutant proteins was compared with WT by immunofluorescence microscopy using the antiflag antibody followed by the FITC-conjugated second antibody. A more severe heat-shock protocol was used for these experiments because the cells overexpress HSP72 and are therefore resistant to heat shock. We used EPR to study nuclear accumulation of HSP72. As illustrated in Fig. 4, it was clear that M17 concentrated in the nucleus as early as within 90 min of heat shock, although the WT did not significantly accumulate within the nucleus until 1 h into the recovery period (hour 3). After heat shock, the N/C ratio for M18 increased, but the changes were not significant. Similar results were observed with standard immunofluorescence microscopy (data not shown).

View larger version (41K): [in this window] [in a new window]   Fig. 4.   A: nuclear localization as determined by exhaustive photon reassignment (EPR) over a 4-h time course. Heat shock was carried out for up to 2 h as shown with up to 2 h additional recovery. N/C for M17 increases significantly by 90 min of heat shock; however, N/C for WT does not increase until 1 h of recovery after 2 h of heat shock (hour 3). M18 N/C increases but does not reach statistical significance compared with baseline. *P < 0.05 vs. time 0. B: center section of representative cells illustrating differences in cellular localization of transfected protein when analyzed by EPR. Arrows indicate nuclei. Top: 90 min of heat shock. Bottom: 2 h of heat shock followed by 2 h of recovery.

Viability studies. To evaluate the effect of the two mutants on cell viability, we tested these cells' uptake of ethidium bromide after heat shock. Treatment with 2 h of heat shock at 43.5°C followed by 2 h of recovery at 37°C reduced the viability of M18 transfected cells to 54.4% (Fig. 5). M17 and WT were unaffected. Severe heat shock at 45°C followed by 2 h of recovery at 37°C killed 53% of WT and 61% of M18 transfected Cos-1 cells, but 78.9% of the M17 transfected cells remained viable after the same treatment (Fig. 5C).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Viability was assessed 2 h after heat shock at 43.5°C. Cos-1 cells were transiently cotransfected with enhanced blue fluorescent protein (EBFP) under control of RSV promoter and with a 10-fold excess of HSP constructs under control of cytomegalovirus (CMV) promoter. Ethidium bromide uptake into nucleus was used as an index of cell death (live dead assay). Blue cells (i.e., transfected cells) were scored as either live (no ethidium) or dead (ethidium uptake in nucleus). A: representative cells. A live, blue cell (left) and a dead cell (right) shown blue plus red from ethidium bromide uptake are shown. B: effect of heat shock at 43.5°C. C: effect of heat shock at 45°C. At 43.5°C only M18 shows decreased viability, but at 45°C both WT and M18 have decreased viability.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results reported here demonstrate that HSP72 is phosphorylated at Y⁵²⁴ after heat stress and that accumulation in the nucleus occurs after this phosphorylation. Inhibition of tyrosine kinases prevents the accumulation of HSP72 in the nucleus after heat shock. Although we show a change in the distribution of HSP72, we cannot exclude the possibility that the protein shuttles in and out of the nucleus at baseline and that the observed changes reflect a change in concentration rather than in transport through the nuclear pore.

Conservative mutation of the single predicted tyrosine phosphorylation site Y⁵²⁴ to phenylalanine greatly retards concentration of HSP72 in the nucleus. In contrast, conversion of Y⁵²⁴ to aspartic acid to create a pseudophosphorylation by introducing a negative charge at this site results in early accumulation of the protein in the nucleus after heat stress. These conclusions were strengthened by using EPR microscopy to identify precisely the locations of HSP72 in these cells. Baseline localization of the Y⁵²⁴ mutants in the cytoplasm of Cos-1 cells was no different from WT cells. Thus these experiments suggest that phosphorylation of Y⁵²⁴ helps to regulate nuclear accumulation of HSP72 after heat stress. It is unlikely to be the only factor controlling nuclear concentration because in the unstressed cell M17 remains in the cytoplasm. Viability studies demonstrated that M18-expressing cells had greatly increased mortality after heat shock, indicating that movement of HSP72 to the nucleus after heat shock is important for cell survival. In pilot experiments to determine the effects of heat shock on cell viability, no difference was observed among groups with a 1-h recovery following 2 h of heat shock; therefore, a 2-h recovery time period was used. The need to extend recovery until after significant nuclear accumulation of the protein has occurred (that is, after 2 h of heat shock and 1 h of recovery) provides additional evidence that nuclear accumulation of HSP72 influences cell survival. Furthermore, M17 cells were able to survive severe heat shock and WT cells were not, suggesting that for this one measurement, at least, the M17 mutation improves survival.

There is a disparity in the immunoprecipitations: when we immunoprecipitate total cell lysates with anti-phosphotyrosine antibodies and develop with anti-HSP72 antibodies, we only see signal at 30 min of heat shock; but when we do the reverse and immunoprecipitate with HSP72 and develop with anti-phosphotyrosine antibodies, the anti-phosphotyrosine antibody binds at the three heat-shock time points. We think this difference occurs because HSP72 binds tightly to other proteins with heat shock. Under the native conditions of immunoprecipitation, the phosphotyrosine site is concealed when HSP72 binds to another protein. Under the conditions of a denaturing gel, the phosphotyrosine site is no longer concealed and the phosphotyrosine antibody can bind to it. Therefore, under normal conditions the phosphotyrosine antibody only recognizes HSP72 at 30 min of heat shock, and under denatured conditions the antibody recognizes HSP72 at all three heat-shock time points. These same results were seen with multiple experiments.

To our knowledge, this report represents the first observation that mammalian HSP72 is phosphorylated. However, others have observed that the Escherichia coli homologue Dnak is phosphorylated (14, 18), and heat shock cognate 70, the constitutive form of HSP70, is phosphorylated (16). Change in phosphorylation status has been observed to influence nuclear localization of other proteins. HSP72 belongs to an interesting subset of proteins that only concentrate in the nucleus selectively; understanding their regulation and control will increase our understanding of nuclear transport of proteins and its role in the cellular response to stress. Future experiments will need to identify the additional regulatory factors controlling nuclear concentration of HSP72 with stress.