Human cyclophilin 40 is a heat shock protein that exhibits altered intracellular localization following heat shock

Abstract

The unactivated steroid receptors are chaperoned into a conformation that is optimal for binding hormone by a number of heat shock proteins, including Hsp90, Hsp70, Hsp40, and the immunophilin, FKBP52 (Hsp56). Together with its partner cochaperones, cyclophilin 40 (CyP40) and FKBP51, FKBP52 belongs to a distinct group of structurally related immunophilins that modulate steroid receptor function through their association with Hsp90. Due to the structural similarity between the component immunophilins, FKBP52 and cyclophilin 40, we decided to investigate whether CyP40 is also a heat shock protein. Exposure of MCF-7 breast cancer cells to elevated temperatures (42°C for 3 hours) resulted in a 75-fold increase in CyP40 mRNA levels, but no corresponding increase in CyP40 protein expression, even after 7 hours of heat stress. The use of cycloheximide to inhibit protein synthesis revealed that in comparison to MCF-7 cells cultured at 37°C, those exposed to heat stress (42°C for 3 hours) displayed an elevated rate of degradation of both CyP40 and FKBP52 proteins. Concomitantly, the half-life of the CyP40 protein was reduced from more than 24 hours to just over 8 hours following heat shock. As no alteration in CyP40 protein levels occurred in cells exposed to heat shock, an elevated rate of degradation would imply that CyP40 protein was synthesized at an increased rate, hence the designation of human CyP40 as a heat shock protein. Application of heat stress elicited a marked redistribution of CyP40 protein in MCF-7 cells from a predominantly nucleolar localization, with some nuclear and cytoplasmic staining, to a pattern characterized by a pronounced nuclear accumulation of CyP40, with no distinguishable nucleolar staining. This increase in nuclear CyP40 possibly resulted from a redistribution of cytoplasmic and nucleolar CyP40, as no net increase in CyP40 expression levels occurred in response to stress. Exposure of MCF-7 cells to actinomycin D for 4 hours resulted in the translocation of the nucleolar marker protein, B23, from the nucleolus, with only a small reduction in nucleolar CyP40 levels. Under normal growth conditions, MCF-7 cells exhibited an apparent colocalization of CyP40 and FKBP52 within the nucleolus.

INTRODUCTION

Cells respond to physiological stress by synthesizing a relatively small suite of proteins at elevated rates to facilitate the chaperoning of crucial pathways within the cell. This response is known as the heat shock response, and the family of proteins are referred to as heat shock proteins (Hsps; reviewed in Wu 1995). Some Hsps function as molecular chaperones and facilitate protein folding, intracellular trafficking, complex assembly, and protein degradation within the cell. It is interesting that all known Hsps appear to play a role in the functioning of the cell under normal growth conditions, often chaperoning nonessential proteins in a similar manner during this time (reviewed in Morimoto 1998).

In the absence of hormone the steroid receptors associate with the major heat shock proteins, Hsp90 and Hsp70. These Hsps act coordinately with other molecular chaperones to facilitate the high-affinity binding and activation of the steroid receptors by ligand (reviewed in Pratt and Toft 1997). Also present at certain times within this unactivated complex are the proteins Hsp40/Hdj-1, p48/hip, p60/hop, and the target modulator proteins, CyP40, FKBP51, FKBP52, and PP5. The target modulators are proposed to fine-tune the responses of the Hsp90- associated target proteins, such as steroid receptors and protein kinases, through modulation of Hsp90 activity (Chen et al 1996; Duina et al 1996; Reynolds et al 1999). They bind to Hsp90 in a mutually exclusive manner by virtue of their tetratricopeptide repeat (TPR) domains associating with the TPR-acceptor site (Chen et al 1996; Owens-Grillo et al 1996; Ratajczak and Carrello, 1996; Barent et al 1998).

Four of the major nonsteroid binding components of unactivated steroid receptor complexes identified thus far (Hsp90, Hsp70, FKBP52, and Hsp40/Hdj-1) have been identified as proteins that exhibit an elevated rate of synthesis under conditions of cellular stress. Both Hsp90 (Borkovich et al 1989) and Hsp70 (Tissieres et al 1974; Banerji et al 1986) have long been recognized as heat shock proteins, and their intracellular concentrations rise significantly upon application of cellular stress. FKBP52 (previously known as Hsp56) was determined to be a heat shock protein even though net increases in intracellular levels of FKBP52 were not detected following the stress event, but rather it was demonstrated that the rate of FKBP52 protein synthesis was increased in response to stress (Sanchez 1990). The Hsp40/Hdj-1 chaperone that associates with Hsp70 has also been identified as a heat shock protein in both mammalian and avian cells, responding to both heat and chemical stress by elevated rates of synthesis (Ohtsuka et al 1990).

Recently, the yeast homologues of CyP40 (Cpr6 and wis2⁺) were identified as heat shock proteins (Weisman et al 1996; Warth et al 1997). Both Cpr6 (Saccharomyces cerevisiae) and wis2⁺ (Schizosaccharomyces pombe) are heat-inducible at the mRNA level, whereas Cpr6 also exhibits elevated protein levels in response to heat stress. To date, mammalian CyP40 has not been investigated for its response to heat shock, although other eukaryotic cyclophilins have been identified as heat responsive. In Saccharomyces cerevisiae, both CyP1 and CyP2 are heat-inducible proteins (Sykes et al 1993), as is CyP20 from rat myogenic cells (Andreeva et al 1997). Plants also display heat shock regulation of the cyclophilin genes, pCyPB and bean cyclophilin (Luan et al 1994; Marivet et al 1994).

Complementary to their identification as heat shock proteins, many of the members of the unactivated steroid receptor complex have been shown to exhibit chaperone properties. Both Hsp90 (Schumacher et al 1994; Young et al 1997) and Hsp70 (Schumacher et al 1994) display an adenosine triphosphate (ATP)-dependent chaperone function, with Hsp90 also exhibiting an ATP-independent chaperone activity (Wiech et al 1992). Other members of the unactivated steroid receptor complex have also been shown to possess chaperone activity. CyP40 and p23 (Freeman et al 1996), FKBP52 (Bose et al 1996), p60/hop (Johnson et al 1998), p48/hip (Hohfeld et al 1995), and Hsp40/Hdj-1 (Lu and Cyr 1998) proteins or their homologues are all active molecular chaperones. It is believed that this chaperoning property of the proteins within the unactivated steroid receptor complex contributes to the folding of steroid receptors to a high-affinity steroid binding state (Chen et al 1996; Duina et al 1996; Reynolds et al 1999).

The predominant mediator of the heat shock response within the cell is the family of transcription factors known as heat shock factors (HSFs). Four members of this family have been identified, with the major regulator of the response in eukaryotes being HSF1 (Rabindran et al 1991). The HSFs act by binding to their heat shock regulatory elements (HSEs) in the promoter regions of their hsp target genes, resulting in an up-regulation of transcription. Under normal cellular conditions, the heat shock response is held in check by a negative feedback loop that involves a number of different molecular chaperones (Abravaya et al 1992; Ali et al 1998; Duina et al 1998; Bharadwaj et al 1999). The final composition of the complex that inhibits basal HSF activity is remarkably similar to that observed for the unactivated steroid receptors (Bharadwaj et al 1999). Cellular conditions that sequester these chaperone proteins away from HSF (eg, cellular stress or overexpression of target proteins for the chaperones; Anathan et al 1986; Xiao and DeFranco 1997) result in the activation of the heat shock pathway. Hsp90 (Ali et al 1998; Duina et al 1998; Zou et al 1998; Bharadwaj et al 1999), Hsp70 (Abravaya et al 1992; Mosser et al 1993; Baler et al 1996), and CyP40 (Duina et al 1998) have all been implicated in this negative feedback loop.

Here we investigate the possibility that the CyP40 component of unactivated steroid receptor complexes is, like other members of the complex, a heat shock protein. We describe an elevation in the levels of CyP40 mRNA in response to heat stress, with no alteration in the amount of CyP40 protein. We demonstrate that the rate of CyP40 protein degradation increases under heat-stress conditions, implying an elevated rate of protein synthesis. The consequences of heat shock and the effect of the RNA synthesis inhibitor, actinomycin D, on the intracellular localization of CyP40 were investigated in the MCF-7 breast cancer cell line. Our results show that application of heat stress alters the intracellular localization of CyP40 protein from a predominantly nucleolar distribution to a more complete nuclear pattern of expression.

MATERIALS AND METHODS

Reagents, cell lines, and culture

The FKBP52 cDNA (Peattie et al 1992) was donated by Dr Debra Peattie, Vertex Pharmaceuticals, Massachusetts; the plasmid encoding rat 18S rRNA (Katz et al 1983) was obtained from Dr George Yeoh, University of Western Australia; and the plasmid, pH2.3 (Wu and Morimoto, 1985), containing Hsp70 cDNA was obtained from Dr Richard Morimoto, Northwestern University, Illinois, USA. The Hi52c antibody against FKBP52 (Nair et al 1997) was obtained from Dr David Smith, University of Nebraska Medical Center, Omaha, NE, USA. The B23 antibody (Valdez et al 1994) was a gift from Dr Ben Valdez, Baylor College of Medicine, Houston, TX, USA; and the hsp/ hsc70 antibody, BB70, was obtained from Dr David Toft, Mayo Clinic, Rochester, MN, USA.

The MCF-7 breast cancer cell line was obtained from Dr H. Dawkins, Queen Elizabeth II Medical Centre, Perth, Western Australia. BT-20 and COS-1 cells were purchased from American Type Culture Collection, Rockville, MD, USA. Mouse fibroblastic NIH 3T3 cells were obtained from Dr W. Langdon, Queen Elizabeth II Medical Centre, Perth, Western Australia. MCF-7, COS-1, and NIH 3T3 cells were routinely cultured in Dulbecco modified Eagle medium (DMEM) containing 5% fetal bovine serum (Trace Biosciences, Sydney, Australia). The BT-20 cells were grown in Earles modified Eagle medium (EMEM) supplemented with 10% fetal calf serum and 1% nonessential amino acids (Trace Biosciences). Cell lines were maintained at 37°C in a humidified incubator at 5% CO₂. For all experiments, cells were in the exponential phase of growth. Cultured cells were routinely shown to be free of Mycoplasma contamination using a polymerase chain reaction–based assay.

cDNA probes for Northern analysis

The CyP40 cDNA probe was excised from pGEM3Z/ CyP40 using XbaI digestion. The FKBP52 cDNA probe was excised from the plasmid pGEM3Z/FKBP52 using BamHI digestion. The hsp70 cDNA probe was excised from the plasmid pH2.3 using BamHI/HindIII digestion. The rat 18S rRNA insert was released with EcoRI/BamHI digestion. All inserts were gel extracted (QIAEX II, Qiagen Gmbh, Germany) and 25 ng of the appropriate cDNA fragment was labeled with α-[³²P]-deoxy cytidine 5′-triphosphate (dCTP) using the random hexamer primer method (Promega Corporation, Madison, WI) as previously described (Ratajczak et al 1996).

Northern blots of heat shock–treated cells

Forty eight hours prior to treatment, cells were passaged into triplicate 90-mm dishes and grown to 60% confluence. Plates were either left at 37°C or transferred to a 42°C incubator for a further 3 hours before harvesting for total RNA using 2 mL of RNAzol reagent (Tel-Test Incorporated, Friendswood, TX, USA) according to the manufacturer's instructions. Northern analysis was conducted essentially as previously described (Ratajczak et al 1996). Briefly, 30 μg total RNA was electrophoresed through a 1% denaturing agarose gel and transferred to Zeta-Probe GT membrane (Bio-Rad, Hercules, CA, USA) by capillary transfer. The membrane was hybridized overnight at 42°C with the appropriate α-[³²P]-dCTP-labeled cDNA probe. The membranes were washed in 0.1× standard saline citrate (SSC)/0.1% sodium dodecyl sulphate (SDS) at 65°C for 45 minutes and exposed to X-ray film (XOMAT, Kodak). Membranes were then stripped and reprobed with the next labeled cDNA fragment. This procedure was repeated until all 4 cDNAs had been used as probes, in the order of least abundant to most abundant (CyP40, FKBP52, Hsp70, and 18S rRNA). The autoradiographs were quantitated using Sigma Gel software (Jandel Scientific, San Rafael, CA, USA).

Effect of quercetin on CyP40 basal and heat-induced mRNA expression

To determine the effects of quercetin on the expression of both basal and heat-induced CyP40 mRNA expression, MCF-7 cells were plated out into 90-mm-diameter plates and grown to 60% confluence. Cells were pretreated in triplicate with either 100 μM quercetin or an ethanol vehicle for 6 hours and then either left at 37°C or transferred to 42°C for 3 hours. RNA was isolated from the dishes using 2 mL RNAzol as described above and then quantitated by spectrophotometry. Thirty micrograms of total RNA was electrophoresed, transferred to a Zeta-Probe GT membrane, and probed for CyP40 mRNA levels as already described. Membranes were subsequently stripped and reprobed for 18S rRNA levels.

Preparation and purification of antibodies to recombinant CyP40 protein

Recombinant, purified GST-bCyP40 was prepared (Ratajczak et al 1995) and the GST portion was cleaved off with thrombin and removed by glutathione-agarose chromatography to produce pure CyP40. Purified CyP40 protein (100 μg in 250 μL phosphate buffered saline [PBS]) was used as an antigen in combination with Montanide ISA 50V adjuvant (1:1 vol/vol) and injected intradermally at numerous sites on the back of 2 male New Zealand white rabbits. The animals were subsequently boosted at 6- week intervals with a further 100 μg of CyP40. Testing of the antisera (RC1 1:1000 dilution) by Western analysis against MCF-7 protein extract showed significant titre for CyP40 protein, detecting a single band at 40 kDa.

Following accumulation of sufficient antibody, an affinity column consisting of purified GST-bCyP40 immobilized onto cyanogen bromide–activated Sepharose (Harlow and Lane 1988; Goding 1996) was used to affinity- purify CyP40 polyclonal antibody RC1 for use in immunohistochemistry. The purified antibody, RC1P, was shown to bind specifically to CyP40 protein in confocal microscopy experiments at a dilution of 1:25.

Effect of heat shock on CyP40 and FKBP52 protein expression

MCF-7 cells propagated under normal conditions were plated out and allowed to grow to 60% confluence before either being left at 37°C or transferred to 42°C for 3, 5, or 7 hours. Plates were harvested in 300 μL of 1× sample buffer (62.5 mM Tris pH 6.8, 1 mM ethylenediamine-tetraacetic acid, 10% glycerol, 1% SDS). Protein extracts were quantitated using the Bradford protein assay (Bradford, 1976).

For Western analysis, either 25 μg (for FKBP52 probing) or 100 μg (for CyP40) of total protein was routinely loaded onto a Protean II (Bio-Rad) 10% SDS-polyacrylamide gel electrophoresis slab. The gels were electroblotted onto Hybond C+ Super membranes (Amersham Pharmacia Biotech, Uppsala, Sweden). Primary antibodies were incubated with the membranes at 4°C overnight. To detect CyP40 protein, RC1 rabbit polyclonal antiserum (1:1000) was used, whereas for FKBP52 protein, the mouse monoclonal antibody was Hi52c (1:5000). Following incubation with the primary antibody, the membranes were washed and incubated with appropriate secondary antibody conjugated to horseradish peroxidase (1:10 000) for 60 minutes. Membranes were washed and then incubated in Renaissance Western Blot Chemiluminescence Plus substrate solution (NEN Life Sciences, Boston, MA, USA) according to the manufacturer's instructions, before exposing to X-ray film. Autoradiographs were quantitated using the Sigma Gel densitometry package (Jandel Scientific).

Effect of heat shock on CyP40 and FKBP52 protein turnover

The protein synthesis inhibitor, cycloheximide, was used to determine the rate of CyP40 degradation under both normal and heat shock conditions. MCF-7 cells were propagated under normal growth conditions, passaged into 90-mm plates, and grown to 60% confluence. Duplicate plates were exposed to either 42°C for 3 hours or left at 37°C. All plates were then incubated at 37°C with 2 μg/mL cycloheximide for indicated times up to a further 21 hours.

At the end of the time course, the duplicate plates were harvested in 300 μL of 1× sample buffer. The recovered extracts were quantitated for protein and aliquots corresponding to 100 μg (for CyP40 analysis) or 25 μg (FKBP52 analysis) of total protein were electrophoresed and subjected to immunoblotting as described above.

Immunohistochemistry

Cells were passaged from stock flasks and plated onto 25 × 25 mm glass coverslips in 35-mm-diameter wells and grown to ∼70% confluence. Coverslips were prepared for immunolocalization by washing in TBS-T (50 mM Tris pH 7.6, 150 mM NaCl, 0.5% Tween 20) for 5 minutes prior to fixing and permeabilization with 2% paraformaldehyde in PBS + 0.1% Triton X-100 for 10 minutes at room temperature. Cells were washed again in TBS-T for 5 minutes before being blocked for nonspecific binding of antibodies using TBS-T containing 10% (vol/vol) normal goat serum and 1% (wt/vol) bovine serum albumin at room temperature for 60 minutes.

The appropriate primary antibodies were incubated with cells overnight in blocking solution at 4°C in a humidified container. The following day, cells were washed and incubated with secondary antibody (either goat anti- rabbit Alexa 488 [Molecular Probes, Eugene, OR, USA] for rabbit primary antibodies or goat anti-mouse Alexa 546 [Molecular Probes] for the mouse primary antibodies) for 60 minutes. After washing 5 times for 5 minutes with TBS-T, coverslips were mounted onto slides in mounting medium (40 mM Tris pH 8.2, 15% [wt/vol] polyvinyl alcohol (PVA), 25% [vol/vol] glycerol, 0.75 mg/mL chlorobutanol, and 0.1% phenol red), and sealed.

Cells were imaged on a Bio-Rad MRC-1000/1024 UV confocal laser scanning microscope using a Nikon 60× NA 1.4 oil immersion lens. The Alexa 488 fluorophore was excited using the 488 nm line from a 250 mW argon ion–tuned laser (Coherent Enterprises, Palo Alto, CA, USA) with the emitted fluorescence being collected through a 522/35 nm bandpass filter. The Alexa 546 fluorophore was excited using the 543 nm line from a Green HeNe laser with the emitted fluorescence being collected through a 585/35 nm bandpass filter. Images were collected sequentially in 1-μm vertical increments. Computer-aided colocalization was performed in Confocal Assistant (Bio-Rad Microscience, Hemel Hempstead, UK) using corresponding 1-μm z-sections through the cells. Controls for cross-reactivity were performed for the secondary antibody specificity, as well as imaging using the alternate laser to determine extent of bleed-through of signals, and shown to be negative in each case.

The specificity of the affinity-purified antibody, RC1P, in detecting CyP40 was demonstrated by incubating purified CyP40 protein (10 μg) with RC1P antibody (1:25) in 150 μL of blocking solution overnight at 4°C. The primary antibody solution was microfuged for 30 minutes at 4°C prior to use in immunohistochemistry with MCF- 7 cells propagated under standard conditions.

To confirm that the punctate pattern of CyP40 expression observed in cell nuclei was derived from the nucleolus, anti-B23 antibody (1:100) was used in colocalization experiments with RC1P in MCF-7 cells. The cells were imaged on the Bio-Rad laser confocal microscope using both the 488-nm and 543-nm lasers by collecting sequential z-sections through the MCF-7 cells as already described.

The effect of heat shock on CyP40 localization was assessed using MCF-7 cells either left at 37°C for 7 hours or placed at 42°C for either 3, 5, or 7 hours. Cells were then washed, fixed, permeabilized, and immunohistochemistry performed for localization of CyP40 as previously described. To act as a positive control for heat shock, cells that had been cultured at either 37°C or 42°C for 7 hours were subjected to immunohistochemistry using the mouse monoclonal antibody, BB70 (1:100), against Hsp70.

For investigations of the effect of actinomycin D on CyP40 localization, MCF-7 cells were treated for 1 or 4 hours with either 5 μg/mL actinomycin D or vehicle (ethanol) and then processed for immunohistochemistry as previously described using either RC1P (1:25) or anti-B23 (1:100) antibodies. Cells were sequentially imaged on the confocal microscope as already described.

To determine the intracellular localization of FKBP52 in MCF-7 cells, the breast cancer cell line was processed for immunohistochemistry with both RC1P antibody (1:25) against CyP40 and with Hi52c antibody (1:200) against FKBP52 as already described.

RESULTS

Effect of heat and chemical stress on CyP40 and FKBP52 mRNA expression

MCF-7 and BT-20 breast cancer cells were subjected to either 100 μM sodium arsenite (As) or 42°C for 3 hours. Isolation of total RNA followed by Northern blotting revealed that, in comparison to vehicle-treated controls (37°C, 0.9% saline-treated cells), increased levels of CyP40 and FKBP52 mRNA were detected following stress treatment (Fig 1). Sodium arsenite caused only small increases in mRNA expression for both immunophilins. Greater increases were induced in response to heat stress of the cells, with CyP40 mRNA increasing by 75-fold in MCF-7 cells and by ∼2-fold in BT-20 cells. In comparison, FKBP52 mRNA levels increased by 4- and 3.5-fold in MCF-7 and BT-20 cells, respectively. The positive control for both heat shock and chemical stress, Hsp70, showed a 3-fold increase in mRNA levels in response to As treatment and a greater than 50-fold increase in response to heat stress in MCF-7 cells. Hsp70 gene transcription was more responsive to As treatment in BT-20 cells, increasing by almost 10-fold, whereas heat stress–induced expression of Hsp70 mRNA was elevated by more than 50-fold.

Fig 1. Effect of heat and chemical stress on CyP40 and FKBP52 mRNA expression. (A) Composite figure of Northern blots demonstrating the effect of sodium arsenite (As-100 μM for 3 hours) and heat shock (42°C for 3 hours) on levels of mRNA for CyP40, FKBP52, and Hsp70 in both MCF-7 and BT-20 breast cancer cells. Northern analysis was conducted on 30 μg total RNA and membranes were sequentially probed with α-[³²P]-labeled cDNA probes for CyP40, FKBP52, and hsp70. Evenness of loading was demonstrated by probing for 18S rRNA. (B) Graphical representation of standardized expression levels relative to 18 S rRNA for CyP40 and FKBP52 mRNA in both MCF-7 and BT-20 cells. Error bars represent standard error

It is interesting that the basal CyP40 mRNA expression in MCF-7 and BT-20 differed considerably, with the mRNA level of CyP40 in normal cultured BT-20 cells being almost 60 times that in MCF-7 cells cultured under normal growth conditions. Following heat shock, the maximal levels of CyP40 mRNA were much more equitable between the cell lines (Fig 1).

Effect of quercetin on CyP40 basal and heat-induced mRNA expression

The bioflavonoid, quercetin, is a known inhibitor of the heat shock response (Hosokawa et al 1990a; Elia and Santoro 1994; Lee et al 1994; Hansen et al 1997). Treatment of MCF-7 cells with 100 μM quercetin resulted in basal CyP40 mRNA being reduced to below detectable levels (Fig 2). Pretreatment with quercetin for 6 hours prior to heat shock also completely abrogated the heat shock–induced elevation of CyP40 mRNA, again decreasing CyP40 mRNA to below detectable levels. In separate experiments with breast cancer T47-D cells, exposure to 100 μM quercetin over 24 hours down-regulated the basal expression of both CyP40 and FKBP52 mRNA to less than 20% of normal levels (not shown).

Fig 2. Effect of quercetin on basal and heat-induced CyP40 mRNA expression. MCF-7 cells were either cultured at 37°C or 42°C for 3 hours following pretreatment for 6 hours with either 100 μM quercetin or vehicle (ethanol). RNA was isolated and 30 μg electrophoresed through a 1% denaturing agarose gel prior to Northern blot transfer. Membranes were probed with an α-[³²P]-labeled cDNA probe for CyP40 and subsequently stripped and reprobed for 18S rRNA

Effect of heat shock on CyP40 and FKBP52 protein levels

MCF-7 cells were either subjected to 42°C for 3, 5, or 7 hours or left to grow at 37°C for the same amount of time. To investigate whether heat shock had any effect on the intracellular levels of the CyP40 and FKBP52 proteins, the relative levels of these proteins were determined by Western blotting (Fig 3). Under the same conditions that demonstrated a 75-fold increase in the mRNA levels for CyP40 within 3 hours (Fig 1A), no net increase in CyP40 protein was observed on Western analysis of protein extracts probed with antibody RC1 (Fig 3—compare triplicates for hours 0 and 3). Even allowing for a lag time in the synthesis of protein from the mRNA, no net increase in protein was observed after a further 2 hours or 4 hours at 42°C. In contrast, incubation of cells for 3 hours at 42°C resulted in elevation of the levels of Hsp70 in comparison to those at 37°C. Longer exposure to heat (5 and 7 hours) caused further increases in Hsp70 protein (Fig 3). In a subsequent experiment to determine whether the lag time for protein synthesis was insufficient to detect an increase in CyP40 protein expression, recovery times at 37°C of up to 37 hours following heat shock treatment at 42°C for 3 hours were investigated, with no detectable change in CyP40 protein levels (results not shown). Western analysis of total protein extracts, with the Hi52c mouse monoclonal antibody for FKBP52, revealed that the FKBP52 protein level also did not increase significantly following heat shock for 3, 5, or 7 hours (Fig 3).

Fig 3. Effect of heat shock on CyP40 and FKBP52 protein levels. MCF-7 breast cancer cells were subjected to heat shock (42°C) in triplicate wells for either 0, 3, 5, or 7 hours before lysing cells in sample buffer. Western analysis was performed on either 100 μg (for CyP40) or 25 μg (for FKBP52 and hsp70) of total protein, probing for expression levels of CyP40 (RC1 antibody, 1:1000) and FKBP52 (Hi52c antibody, 1:2000) protein. Hsp70 protein was used as a positive control for heat shock (BB70 antibody 1:10000)

Effect of heat shock on CyP40 and FKBP52 protein turnover

We next investigated whether the lack of increase in CyP40 protein in response to heat shock was due to an increased rate of CyP40 degradation. Total protein synthesis was blocked by cycloheximide, and the rate of protein degradation was assessed over a 24-hour time period in MCF-7 cells cultured under normal or heat shock conditions.

Following cycloheximide treatment, the level of CyP40 protein in MCF-7 cells at 37°C over the 24-hour time course did not decrease significantly (Fig 4A). This indicated that in MCF-7 cells cultured at 37°C, the half-life of CyP40 protein is longer than 24 hours. In cells that had been previously heat shocked for 3 hours, inhibition of protein synthesis resulted in a sharp reduction of CyP40 protein levels to less than 20% of the initial level of CyP40 protein by 24 hours from the onset of heat shock. The exponential decay curve for CyP40 protein levels following heat shock (Fig 4B) was used to calculate a half-life of 8.4 hours under these conditions.

Fig 4. Effect of heat shock on CyP40 and FKBP52 protein turnover. MCF-7 cells were cultured in duplicate wells at either 37°C or 42°C for 3 hours and then protein synthesis was blocked with 2 μg/ mL cycloheximide. All cells were then cultured at 37°C for between 0 and 21 hours prior to harvesting. (A) Western blot of protein extracts immunoprobed for either CyP40 (RC1 antibody) or FKBP52 (Hi52c antibody). (B) Scatter graph of CyP40 and FKBP52 expression for normal (o) and heat shock–treated (▪) cells demonstrating rate of protein degradation

Cycloheximide treatment of MCF-7 cells grown under otherwise normal culture conditions did not show an apparent reduction of the FKBP52 protein levels over the course of 24 hours (Fig 4A), indicating that this immunophilin also has a half-life of longer than 24 hours. Inhibition of protein synthesis following heat shock at 42°C for 3 hours, however, did result in a marginal reduction in expression levels for FKBP52 protein to 70% of the original amount of FKBP52. The decay curve for FKBP52 protein after heat shock exhibited an initial increase in the degradation rate for 10 hours, following which the profile paralleled that observed for normal MCF-7 cells.

Heat shock redistributes CyP40 from a nucleolar to a nuclear location in MCF-7 cells

The affinity-purified antibody, RC1P, was tested in immunolocalization studies in MCF-7 cells for its ability to accurately demonstrate the intracellular localization of CyP40 (Fig 5B). The observed localization pattern was predominantly nuclear, with intense punctate expression noted within discrete foci in the nucleus. A low level of diffuse staining was also detected within the cytoplasm of the MCF-7 cells. A similar staining pattern was observed with T47-D breast cancer cells, whereas the mouse fibroblastic cell line, NIH 3T3, did not display punctate immunostaining within the nucleus, but a higher degree of cytoplasmic staining than the breast cancer cell lines (results not shown). The punctate foci were confirmed to be of nucleolar origin in colocalization experiments using a monoclonal antibody for the nucleolar marker protein, B23 (Fig 6). Immunohistochemistry without the use of a primary antibody (Fig 5A) demonstrated that there was no nonspecific contribution to the observed signal by the secondary antibody. Preadsorption of the RC1P antibody with 10 μg of purified CyP40 protein overnight at 4°C was able to completely reduce the fluorescence signal to background levels (Fig 5C).

Fig 5. Influence of heat shock on immunolocalization of CyP40 in MCF-7 cells. MCF-7 cells were cultured at either 37°C or 42°C for 3, 5, or 7 hours prior to immunohistochemistry using RC1P antibody against CyP40. (A) No primary antibody. (B) Normal MCF-7 cells probed with RC1P (1:25). (C) RC1P (1:25) preincubated with 10 μg of purified CyP40 protein prior to immunohistochemistry. (D) MCF-7 cells cultured at 37°C, (E) after 3 hours exposure to heat stress at 42°C, (F) after 5 hours at 42°C, and (G) after 7 hours at 42°C. (H) MCF-7 cells cultured at 37°C probed with BB70 antibody (1:100) for hsp70. (I) MCF-7 cells probed with BB70 after heat stress for 7 hours at 42°C. Cells were imaged using a Bio-Rad MRC-1000/1024 UV confocal scanning laser microscope. Scale bar represents 10 μm and arrows indicate nucleolar localization

Fig 6. Effect of actinomycin D on CyP40 and B23 immunolocalization in MCF-7 cells. MCF-7 breast cancer cells were cultured under normal growth conditions before being treated for 4 hours with ethanol vehicle (A and B) or 5 μg/mL actinomycin D (C and D) and then subjected to immunohistochemistry using RC1P antibody (1:25) for CyP40 (A and C) or anti-B23 antibody (1:100) (B and D). Cells were imaged using a Bio-Rad MRC-1000/1024 UV confocal scanning laser microscope. Arrows indicate nucleolar localization

For investigations of CyP40 localization in response to heat shock, cells were either cultured at 37°C (Fig 5D) or transferred to 42°C for 3 hours (Fig 5E), 5 hours (Fig 5F), or 7 hours (Fig 5G). Cells cultured at 37°C (Fig 5D) confirmed the pattern of expression for CyP40 previously observed (Fig 5B), but after heat shock for 3 hours, the signal for CyP40 was initially translocated out of the nucleoli, and exhibited a more diffuse nuclear staining rather than the punctate nuclear pattern of normal cultured cells. With increased exposure time to heat stress (first for 5 hours and then for 7 hours at 42°C), CyP40 immunostaining appeared to intensify within the nucleus, but with a more complete granular appearance, rather than being predominantly associated with nucleoli. No identifiable nucleolar staining was apparent following 7 hours of heat shock.

As a positive control for heat shock, MCF-7 cells cultured either at 37°C or 42°C for 7 hours were probed with the antibody, BB70 against Hsp70. At 37°C, Hsp70 exhibited a low level of immunostaining throughout the cell (Fig 5H). In contrast, heat stress for 7 hours (Fig 5I) caused a substantial increase in Hsp70 levels in the nuclei of the MCF-7 cells, and particularly in the nucleoli, which had previously exhibited no staining for Hsp70 in normal cultured cells.

Effect of actinomycin D on CyP40 cellular localization

Inhibition of RNA synthesis with actinomycin D has previously been demonstrated to cause a translocation of the nucleolar protein, B23, into the nucleus (Yung 1985). MCF-7 cells were either treated with vehicle (ethanol) or 5 μg/mL actinomycin D for 4 hours prior to immunohistochemistry. Vehicle treatment did not affect the localization of either CyP40 (Fig 6A) or B23 (Fig 6B). In contrast, treatment with actinomycin D for 4 hours caused translocation of B23 from the nucleolus, resulting in diffuse nuclear staining (Fig 6D). Actinomycin D treatment resulted in a considerably lower signal intensity of CyP40 nucleolar staining (Fig 6C).

Colocalization of CyP40 and FKBP52

The previously reported presence of both CyP40 and FKBP52 within estrogen receptor complexes (Ratajczak 1993) prompted us to determine whether both were colocalized within the cell. For this purpose, we performed immunolocalization studies with antibodies specific to CyP40 and FKBP52 in the estrogen receptor–positive MCF-7 cell line. Initial localization experiments with the Hi52c monoclonal antibody to FKBP52 showed the protein to be localized throughout the cytoplasm and more abundantly within the nucleus of MCF-7 cells (Fig 7B). In colocalization experiments, computer-aided projection of CyP40 and FKBP52 signals from corresponding sequentially imaged z-sections for the same cells (Fig 7C) revealed a weak but distinct nucleolar colocalization of FKBP52 with CyP40. To demonstrate specificity of the signals, cross-reactivity of the anti-mouse secondary antibody with the rabbit primary antibody was determined and found to be below detectable levels (not shown). Correspondingly, the anti-rabbit secondary antibody was unable to detect the mouse primary antibody. Imaging of the fluorophores using the excitation with the opposite laser revealed no bleed-through of signals from one fluorophore to the other (results not shown).

Fig 7. Colocalization of CyP40 and FKBP52 in MCF-7 cells. Cells were grown under normal culture conditions before being subjected to immunohistochemistry, using indicated antibodies, and imaged using a Bio-Rad MRC-1000/1024 UV confocal scanning laser microscope. (A) Localization of CyP40 using RC1P antibody (1:25). (B) Localization of FKBP52 using Hi52c antibody (1:200). (C) Colocalization of CyP40 and FKBP52 designated by areas in yellow. Arrows indicate nucleolar localization, scale bar represents 10 μm

DISCUSSION

The accepted definition of a heat shock protein is one that exhibits an elevated rate of synthesis under conditions of cellular stress. It has been previously shown that FKBP52, the major FK506 binding protein in unactivated steroid receptor complexes, follows this pattern of response. Indeed, one of the early identities of FKBP52 was that of Hsp56 (Sanchez 1990). Although an increase in intracellular levels of Hsp56 protein in response to heat stress could not be demonstrated in those initial studies, pulse- labeling experiments confirmed that the rate of synthesis of the protein was elevated.

Here we report for the first time that human CyP40 is a heat shock protein, displaying elevated levels of mRNA, as well as an increase in the rate of protein turnover, following heat shock. These results are consistent with the earlier demonstration that the yeast CyP40 homologues, Cpr6 and wis2+, are also heat-responsive (Weisman et al 1996; Warth et al 1997).

The basal expression of CyP40 mRNA in BT-20 cells was significantly higher than that for MCF-7 cells (Fig 1), although the final level to which each increased following heat shock was similar. Consequently, the overall magnitude of increase for CyP40 mRNA was considerably larger in the MCF-7 cells (75-fold increase in comparison to 1.7-fold). The basal promoter for CyP40 in the BT-20 cells may be operating at near to maximal activity, and therefore, is only able to respond to heat shock in a limited capacity. Amplification of the CyP40 gene in BT-20 cells (Ward et al 1999), may also contribute to the elevated basal levels of CyP40 mRNA expression in this cell line.

Our results are consistent with HSF being involved in regulation of both basal and stress-induced transcription of the CyP40 gene in MCF-7 cells. This is demonstrated by the ability of quercetin to inhibit the expression of CyP40 mRNA (Fig 2). Quercetin has been implicated in the inhibition of HSF activity, affecting both the heat shock response and basal expression of heat shock proteins (Hosokawa et al 1990a, 1990b, 1992; Koishi et al 1992; Nagai et al 1995; Hansen et al 1997). The identification of a putative heat shock element in the first intron of the human CyP40 gene (Yokoi et al 1996b) further implicates HSF in the up-regulation of CyP40 mRNA in response to stress.

In the studies performed to determine the effect of heat shock on CyP40 protein, no increase was observed following exposure to heat shock for up to 7 hours (Fig 3). This lack of protein accumulation in response to heat shock is similar to that previously reported for FKBP52 (Hsp56) in human IM-9 cells (Sanchez 1990). This would indicate that either the increase in mRNA is a futile cycle or that the heat-induced increase in the degradation of the immunophilins in conjunction with an increase in synthesis serves some cellular function. These possibilities led us to investigate the effect of blocking of protein synthesis following heat shock in order to compare the rates of degradation of CyP40 and FKBP52 under such conditions with those for the proteins in non–heat shocked MCF-7 cells.

The rate of degradation for both CyP40 and FKBP52 proteins was elevated immediately following heat shock (Fig 4). This increased rate of degradation coupled with no net change in protein levels at the same time points implies that there must be a corresponding increase in the rate of synthesis of both CyP40 and FKBP52 proteins. The mechanism and physiological basis for the increase in CyP40 (and FKBP52) turnover remains to be elucidated.

The half-lives of both CyP40 and FKBP52 proteins are longer than 24 hours under normal growth conditions (Fig 4) and both immunophilins would therefore be considered part of the long-lived fraction of cellular proteins. It is interesting that in cultured rat hepatoma cells, elevation of the culture temperature to 43°C was found to accelerate the degradation of the long-lived fraction of cellular proteins via the ubiquitin-lysosomal and ubiquitin-proteosomal pathways. The initial burst of protein degradation was followed by a slower phase of proteolysis, and this profile of degradation was unaffected by inhibition of Hsp synthesis (Parag et al 1987; Westwood and Steinhardt 1989).

During the culture of MCF-7 breast cancer cells under heat shock conditions, the distribution of CyP40 within the cell altered from a predominantly punctate nucleolar localization to a more complete granular appearance within the nucleus. This redistribution of protein is in direct contrast to the observed migration of Hsp70 to the nuclei following heat stress (Welch and Feramisco 1984). Although the confocal images suggest that there is an elevated level of CyP40 present in the nucleus following 7 hours of heat shock (Compare Fig 5, G and D), it would have to result from a redistribution of CyP40 protein, as opposed to an actual increase, because Western blot analysis from extended heat shock of MCF-7 cells (Fig 3) suggests that CyP40 protein levels do not increase in response to heat stimulus under the same conditions. The nuclear CyP40 would potentially have to be drawn from the nucleolar and cytoplasmic protein pools in response to the heat shock. Perhaps under the stressful conditions, CyP40 is involved in transporting or chaperoning proteins to the nucleus, where essential pathways are being kept active. The observed translocation of CyP40 protein in response to heat shock is in contrast to the results obtained with mouse Lcl3 cells, in which CyP40 was found to display an unaltered intracellular localization in response to heat shock (Lebeau et al 1999). In that particular study, the application of heat stress was for a considerably shorter duration than that used in our experiments and may have been insufficient to trigger the translocation.

The intracellular localization of CyP40 has been previously established in two quite different cell lines. A predominantly nucleolar localization was reported in rat pulmonary endothelial cells (Owens-Grillo et al 1996). In contrast, an almost exclusively cytoplasmic localization was observed in porcine kidney LLC-PK₁ cells (Yokoi et al 1996a), with no apparent nucleolar localization. In our hands, CyP40 was consistently present within the nucleus of MCF-7 cells, with lower levels being detected in the cytoplasm (Fig 5B). The breast cancer cells displayed a punctate pattern of expression within the nucleus, identified as nucleolar staining using the nucleolar marker protein, B23 (Fig 6). It has been proposed that the discrete patches of cytoplasmic staining detected in rat pulmonary endothelial cells (similar to that observed for CyP40 in MCF-7 cells) might be associated with mitochondria (Owens-Grillo et al 1996), however, this has not yet been confirmed.

Actinomycin D treatment of MCF-7 cells at 5 μg/mL (∼4 μM) concentrations caused a translocation of B23 from the nucleolus to the nucleus, but only partially affected CyP40 nucleolar localization (Fig 6). In contrast, mouse Lcl3 cells displayed a complete translocation of CyP40 from the nucleoli in response to 1 μM actinomycin D (Lebeau et al 1999). The results suggest differences in cell line sensitivity to actinomycin D, and that higher concentrations of the inhibitor or longer treatment times may be required for a complete loss of nucleolar localization for CyP40 in MCF-7 cells. In our hands, treatment of MCF- 7 cells with 5 μg/mL actinomycin D for extended times resulted in cell death from toxicity.

FKBP52 exhibited strong nuclear staining with considerable cytoplasmic staining in MCF-7 cells (Fig 7). In addition, the cells exhibited a weak but significant nucleolar colocalization of FKBP52 protein with CyP40. This observation has not been previously reported. It has been speculated that the TPR-containing proteins (particularly FKBP52) within the steroid receptor heterocomplexes are involved in the transport of steroid receptors from the cytoplasm to the nucleus (Czar et al 1995; Pratt et al 1999). If this is the case, it is possible that CyP40 plays a similar role in steroid receptor cytoplasmic-nuclear transport.