INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The cellular response to stresses, including exposure to environmental (UV radiation, heat shock, heavy metals), pathological (infections, fever, inflammation, malignancy, ischemia) or physiological (growth factors, hormonal stimulation, tissue development) stimuli, is represented at the molecular level by rapid synthesis of molecular chaperones such as the heat shock family of stress proteins (heat shock proteins [hsp's]) (for reviews, see references 4, 10, 23, 28, 29, and 41). This response is remarkably conserved in prokaryotic and eukaryotic cells and is widely believed to play a pivotal role in host defense and survival (30). The induction of hsp's in response to stress serves to protect against the initial insult, augment recovery, and produce a state of resistance to subsequent stress (thermotolerance) (15, 28). This protective role of hsp's is attributed to several functional properties, including an active participation in the folding of proteins by minimizing incorrect interactions within and between molecules, maintenance of proteins in their native folded states, and the repair or promotion of the degradation of misfolded proteins (1, 16). In addition, hsp's can function in cellular protection by modulating the engagement and/or progression of apoptosis induced by a variety of stress stimuli (2). Besides the well-established role of hsp's in cell survival, widespread clinical interest exists in their chaperone function during a range of human pathologies, including neurodegenerative conditions (such as amyloidosis, prion disease, and Alzheimer's disease) and various cardiovascular diseases (including myocardial ischemia, cardiac hypertrophy, stroke, and blood vessel injury) (4, 37, 41).

A critical role in the cellular response to acute stress situations has been assigned to the hsp70 protein family, which is an abundant and highly conserved group of proteins in eukaryotic cells that contains members that are constitutively expressed and inducibly regulated and/or targeted to different intracellular organelles. In the mouse, the hsp70 family contains at least seven proteins, including the heat shock cognate protein (Hsc70), the glucose regulated proteins Grp75 and Grp78, the spermatocyte-specific hsp70.2, and the testis-specific Hsc70t. In addition, the exposure of cells to stress insults activates a survival response via induction of the intronless hsp70.1 and hsp70.3 genes (designated hsp70i). These genes encode identical proteins of 68 kDa and are located approximately 8 kb apart within the major histocompatibility complex class III locus (14, 18, 19). The molecular chaperone function of the hsp70 protein family relies on the ATP-regulated association of hsp70 with hydrophobic segments in the substrate polypeptide (6, 33). All hsp70 proteins contain a conserved 44-kDa NH₂-terminal ATPase domain, a more variable COOH-terminal domain that contains a 15-kDa peptide binding site, and a 10-kDa module with an undefined function. The ATP-bound form of the protein has a low affinity for substrates compared to the ADP-bound form. Thus, these proteins bind a linear peptide intermediate, which is mediated by cycles of ATP binding and hydrolysis, followed by ADP-ATP exchange and release. hsp70 chaperone activity is further regulated by a number of other cochaperones (e.g., hsp40, hip, hop) (7, 17).

The rapid induction of hsp70i (hsp70.1 or hsp70.3) protein expression during acute stress stimuli in the cell represents a unique feature of the physiological function of these molecules. As a result, there is widespread interest in the regulatory networks and mechanisms of hsp70i expression in cells and tissues in vivo. It is well known that hsp70i expression is accomplished by mechanisms of transcriptional activation and translation involving heat shock transcription factors (HSFs). Members of the murine HSF family (HSF1, HSF2, or HSF4) bind to heat shock elements (alternatively oriented pentanucleotide 5'-nGAAn-3' units) in the promoters of hsp genes and regulate their transcription (35, 48). HSF1 is ubiquitously expressed and is the most effective transactivator of stress-induced expression of hsp70i genes. In contrast, HSF2 has been proposed to regulate hsp70i expression during specific stages of development, whereas the function of the more recently described HSF4 is unknown (12, 42). Despite extensive studies of HSF function in the cells of complex organisms, little specific knowledge is available about the contribution of HSFs to the regulation of tissue- and cell-specific expression of hsp's, particularly that of hsp70.1 or hsp70.3, under normal or stress conditions in vivo. However, some evidence indicates that hsp70i genes are also constitutively expressed under physiological conditions in certain tissues, suggesting that such expression is not solely regulated by HSF1. Indeed, it has been shown that hsp70.1 is transcribed at the onset of zygotic genome activation (two-cell stage) and that this spontaneous expression is regulated by transcription factors other than HSFs (e.g., Sp1) (5), although other studies using a heterologous hsp70.1 promoter to regulate transgene expression have also implicated a role for HSF1 in such spontaneous hsp70.1 expression. (8, 24, 25, 44). Furthermore, immunohistochemical studies have provided evidence for constitutive expression of hsp70i in certain tissues of adult mice, such as the epithelial layer of the skin (31). However, an important caveat is that many of these studies have relied on studying hsp70i expression using truncated hsp70 promoter sequences to regulate transgene (-galactosidase or luciferase) expression, conditions that may not faithfully reproduce its normal activity. Together, the available evidence concerning tissue-specific hsp70i expression is insufficient to define a specific role for hsp70 during development or in the adult.

Given the existence of multiple hsp70 family members (including the stress-induced hsp70.1 and hsp70.3) with close sequence homology, it has been hitherto impossible to determine the precise functional contribution of each gene in cellular protection from stress in vivo and how the individual hsp70s functionally relate to each other. Direct evidence exists that shows that hsp70i has multiple roles in protection against stress-induced cell death via a cell-protective process known as thermotolerance (or cytoprotection), in which initial sublethal exposure of cells to heat (hyperthermic preconditioning) or other stress stimuli can profoundly attenuate all of the heat-induced changes to a subsequent, more-severe, stress challenge that normally results in extensive cell death. Conversely, abrogation of hsp70i expression or neutralization of its function renders cells sensitive to apoptosis while overexpression of hsp70i in most cells provides protection from cell death triggered by a variety of stress stimuli, including hyperthermia, oxidative stress, chemotherapy agents, and radiation (27). These observations clearly demonstrate the cytoprotective properties of hsp70i; however, it is less clear how this is accomplished. Growing evidence from studies using cell-free systems or cell lines engineered to overexpress hsp70 have suggested that this molecule can inhibit apoptosis following a variety of treatments by modulating the stress-induced intrinsic apoptotic pathway, which mediates cell death through the mitochondria by release of cytochrome c from the mitochondrial intermembrane space. It has been reported that hsp70i functions at both the level of cytochrome c release and initiator caspase activation and that the chaperone function of hsp70i is required for these effects (36). Additionally, hsp70i can exhibit an antiapoptotic function via direct association with the caspase recruitment domain of Apaf-1 and inhibition of apoptosome formation (3, 40). Another point in the apoptotic pathway that can be modulated by hsp70i is c-Jun NH₂-terminal kinase (JNK) signaling, which precedes apoptotic cell death following heat shock and ethanol exposure as well as via stimulation with cytokines or UV irradiation (20, 26). It has been proposed that hsp70i acts by preventing JNK activation, although the importance of this intervention is less clear (13). In the case of tumor necrosis factor-induced apoptosis, hsp70i can rescue cells from apoptosis downstream of JNK activation. Clearly, hsp70i affects multiple apoptotic pathways, and cell type-specific differences may account for the various points of hsp70i intervention.

Elucidation of the role of the hsp70i genes in response to environmental and physiological stress requires a direct in vivo study of hsp70.1 and hsp70.3 expression in specific cell and tissue types and examination of their functional contribution to the stress response. To this end, we have generated mice deficient in the hsp70i genes by replacing the entire coding sequence of hsp70.1 or hsp70.3 with an in-frame -galactosidase gene. This allows for the study of hsp70.1 or hsp70.3 function in vivo and enables examination of tissue- and cell-specific regulation of these genes at both the transcriptional and protein levels during development and in the adult mouse under normal or stress conditions. Here we report that inactivation of hsp70.1 or hsp70.3 gene results in a marked reduction in hsp70i protein synthesis in different mouse tissues under both normal and heat stress conditions. This was reflected by deficient maintenance of acquired thermotolerance and increased sensitivity to heat stress-induced apoptosis. Direct in vivo evidence for tissue-specific constitutive expression and inducible expression of hsp70i under stress conditions is also provided, suggesting the existence of separate mechanisms to control hsp70i expression under physiological versus pathological conditions.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of targeting vectors and generation of mice deficient in either the hsp70.1 or hsp70.3 gene. The 5' and 3' ends of the hsp70.1 and hsp70.3 gene fragments used in the construction of the targeting vectors were isolated from a 129/SvJ mouse genomic library (lambda fixII vector; Stratagene, La Jolla, Calif.) by hybridization with a human hsp70.1 cDNA probe (19). Restriction mapping, oligonucleotide hybridization, and sequencing confirmed that four overlapping phage clones contained both murine hsp70.1 and hsp70.3 loci, including several kilobases of 5'- and 3'-flanking sequences. Targeting vector construction was based on a lacZ-neo-tk (pN-Z-tk2) template plasmid vector containing a -galactosidase (lacZ) gene fragment with the bovine growth hormone poly(A) signal [lacZ-poly(A)], a neomycin resistance gene driven by the thymidine kinase (tk) promoter with the simian virus 40 poly(A) signal [tk/neo-poly(A)], and flanking tk gene cassettes. The tk/neo-poly(A) fragment was flanked by Cre recombinase recognition (loxP) sequences to allow removal of the selectable marker gene from the targeted locus by intercrossing the mutant mice with transgenic mice expressing the cre gene. Essentially, nearly the entire coding sequence of hsp70.1 or hsp70.3 (codons 1 to 633) was replaced in frame with the lacZ-neo cassette using the proximal 4.5 kb and distal 4 kb for hsp70.1 and 4.5 and 3.5 kb for hsp70.3, respectively. Targeting vectors were linearized at the unique SalI site for embryonic stem (ES) cell transfection. ES cells (D3; Incyte Genomics, St. Louis, Mo.) were electroporated with the linearized targeting vectors and selected for double resistance to G418 (200 µg/ml) and 2'-fluoro-2'-deoxy-1b-D-arabino-furanosyl-S-iodo-uracil (FIAU) (ganciclovir, 2 µM) following a standard protocol (Incyte Genomics). Doubly resistant clones were screened by Southern blotting. Correct targeting was confirmed by Southern blotting with flanking genomic (external to the targeting vectors) and lacZ-neo probes. Briefly, BamHI-digested genomic DNA was hybridized with an external probe to yield bands of 6 and 9 kb for the hsp70.3 wild-type and targeted loci, respectively. For the hsp70.1 wild-type and targeted loci, hybridization of EcoRI-digested genomic DNA with an external probe yielded bands of 12 and 9 kb, respectively. ES cell clones were microinjected into C57BL/6 blastocysts, and several germ line transmitting chimeric mice were obtained. Genotyping of mice was performed by Southern blotting with external probes or by PCR with primers 1 (annealing to the hsp70.3 and hsp70.1 coding regions, 5' AGATCACCATCACCAACGACAAG), 2 (annealing to the neo gene, 5' CTTGGGTGGAGAGGCTATTC), 3 (binding to the hsp70.3 allele 3' untranslated region, 5' GTGCAATACACAAAGTAACTGAAAGAC), and 4 (annealing to the hsp70.1 gene 3' untranslated region, 5' GACAGTAATCGGTGCCACAAG). For the detection of the wild-type or targeted hsp70.3 allele, primers 1, 2, and 3 were used in combination, and for the wild-type or targeted hsp70.1 allele, primers 1, 2, and 4 were used in combination. The expected PCR products for the wild-type and targeted loci are fragments of 0.6 and 1.1 kb for hsp70.3 and of 0.5 and 1.0 kb for hsp70.1, respectively.

Histology. Embryos recovered at embryonic day 17 (E17) or tissues harvested from adult mice were embedded in OCT compound, snap-frozen in a dry ice 2-methyl-butane bath, sectioned, air dried, and fixed in 0.2% glutaraldehyde in phosphate-buffered saline (pH 7.3) with 2 mM MgCl₂ for 10 min. Sections from each tissue specimen were stained with either hematoxylin or 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) for -galactosidase activity (Molecular Probes, Eugene, Oreg.). All sections were counterstained with eosin and subjected to gross and microscopic pathological analysis.

Whole-body hyperthermic challenge. Wild-type (+/+), heterozygous (+/), or homozygous (/) hsp70.3 or hsp70.1 adult mice were semi-immersed in a circulating water bath at 42°C for 45 min and left to recover for 6 to 8 h before euthanatization. Tissues were prepared for histological examination as described above. Untreated mice were used as controls.

Thermal response and kinetics of thermotolerance induction in CFU-GM obtained from hsp70.3- or hsp70.1-deficient mice. Bone marrow cells from +/+, +/, or / hsp70.1 or hsp70.3 mice were tested for their ability to develop thermotolerance (34). Bone marrow cells were heated at 43°C for 20 min and challenged with a more-severe heat (44°C for 40 min) at the times indicated during the recovery period at 37°C. Cells were subsequently plated and incubated at 37°C in 5% CO₂ for 8 days. Colonies of granulocytes, macrophages, or a mixture of granulocytes and macrophages were counted microscopically. The colony forming efficiency of untreated CFU granulocytes/macrophages (CFU-GM) was approximately 2/10³ nucleated cells. Cells were plated at various concentrations (1 × 10⁵ to 16 × 10⁵ cells per dish) depending on the given treatment. The percentage survival was calculated by the following formula: percent survival = [(number of colonies after severe heat challenge/number of cells plated)/(number of colonies after primary heat/number of cells plated)] × 100.

Western blot analysis for hsp70 protein expression. Whole-cell extracts (35 µg of total protein) were subjected to sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis and transferred to a nitrocellulose filter (Bio-Rad, Hercules, Calif.). Immunodetection using the enhanced chemiluminescence (ECL) method (ECL kit; Amersham, Piscataway, N.J.) was performed according to the manufacturer's instructions. The membrane was probed first with a monoclonal antibody specific for the heat-inducible hsp70 (C92; Amersham), antibody reacting to both Hsc70 and heat-inducible hsp70 (3A3; Affinity BioReagents Inc., Golden, Colo.), or antibody specific to -galactosidase (Promega, Madison, Wis.), and then it was probed with an appropriate horseradish peroxidase-conjugated second antibody. After ECL detection, the membrane was stripped and reprobed with a rabbit polyclonal antibody specific to actin (Sigma, St. Louis, Mo.) and horseradish peroxidase-conjugated anti-rabbit serum and detected with ECL again. The blot was probed for actin with specific rabbit polyclonal antibody (Sigma) to ensure that equivalent amounts of protein were present in each lane.

Apoptosis and survival of MEFs deficient in expression of hsp70.1 or hsp70.3 following thermal challenge. Mouse embryonic fibroblasts (MEFs) were prepared from day 14 embryos and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. To induce thermotolerance, cells were preconditioned with a relatively mild heat shock (43°C for 20 min) and allowed to recover at 37°C for 6 or 24 h before challenge with a lethal heat shock (45°C for 30 min). Analyses were performed following a further recovery at 37°C for 24 h. Cell viability and the level of apoptosis after heat challenge were measured by staining with annexin V-fluorescein isothiocyanate (FITC) and propidium iodide (Apoptosis Detection kit; R & D Systems, Minneapolis, Minn.) according to the manufacturer's instructions and analyzed using a FACSCalibur cytometer (Becton Dickinson).

Measurement of caspase activity and cytochrome c release. Apoptotic caspase activity was assayed by immunoblotting total cell lysates (50 µg of protein) with either antibody specific to caspase 9 (AAP; StressGen, Victoria, Canada) or antibody (H-250; Santa Cruz Biotechnology, Santa Cruz, Calif.) to one of the caspase substrates, poly(ADP-ribose) polymerase (PARP). For the detection of cytochrome c release from the mitochondria, we followed the previously described procedures (50). The S-100 fraction was prepared as follows. Following heat treatment of MEFs, cell pellets were washed once in ice-cold phosphate-buffered saline and resuspended with 5 volumes of buffer (20 mM HEPES-KOH [pH 7.5], 10 mM KCl, 1.5 mM MgCl₂, 1 mM sodium EDTA, 1 mM sodium EGTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride) containing 250 mM sucrose. Cells were homogenized and centrifuged at 10,000 × g for 15 min at 4°C. The supernatant of the 10,000 × g spin was further centrifuged at 100,000 × g for 1 h at 4°C, and equal amounts of protein (30 µg) present in the supernatant (S-100 fraction) were examined for cytochrome c release by immunoblotting using antibody specific to mouse cytochrome c (Pharmingen, San Diego, Calif.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Generation of mice deficient either in hsp70.1 or hsp70.3 by targeted gene replacement with -galactosidase. To study the function and transcriptional regulation of hsp70.1 and hsp70.3 in vivo, we devised a gene deletion strategy involving replacement of nearly the entire hsp70.1 or hsp70.3 coding sequence with a lacZ-neo cassette (Fig. 1A and D). Five D3 embryonic cell clones resistant to G418 and FIAU, three clones heterozygous for the inactivated hsp70.3 gene and two clones heterozygous for the hsp70.1 gene, transmitted the disrupted allele to the germ line. Correct targeting was demonstrated by Southern blotting with flanking 3' genomic, external to the targeting vectors, and neo-lacZ probes (Fig. 1B and E), and routine genotyping of mice was performed by PCR (Fig. 1C and F). All experiments using homozygous, heterozygous, or wild-type controls were performed on F₂ C57BL/6-129/SvJ mixed background littermates from F₁ heterozygous crosses. To avoid potential interference of the tk-neo marker gene on the expression of LacZ under the hsp70.3 or hsp70.1 promoter, mice were bred with C57BL/6 mice expressing the cre gene (provided by Pandelakis Koni, Institute of Molecular Medicine and Genetics, Medical College of Georgia, Augusta) to remove the neomycin selection marker. Experiments presented in this report have been performed with mice containing the selection marker in the targeted locus. However, analyses with mice in which the neomycin gene has been removed from the targeted allele confirm these results (data not shown). The hsp70.1^/ or hsp70.3^/ mice were viable, born at the expected Mendelian distribution, and fertile, and no distinguishable morphological or microscopic abnormalities were detected in the different tissues analyzed.

View larger version (32K): [in this window] [in a new window]   FIG. 1.   Construction of the targeting vectors for hsp70.1 and hsp70.3 genes and generation of mice deficient for the hsp70.1 or hsp70.3 gene. Restriction map of the hsp70.3 (A) or hsp70.1 (D) gene, showing the wild-type allele (top), the targeting vector (middle), and the predicted targeted allele following homologous recombination (bottom). The position of the neo-lacZ and TK cassettes and probes for Southern blotting analyses are indicated. The vectors were designed so that the promoter of each hsp70.1 or hsp70.3 gene drives the -galactosidase expression. The probe to detect correct gene targeting by Southern blotting was a 3' fragment of 0.4 kb (BamHI-HindIII) for the hsp70.3 gene locus and a 1.2-kb (BamHI-BamHI) fragment for the hsp70.1 gene locus. The restriction enzymes are designated as follows: R, EcoRI; B, BamHI; A, ApaI; N, NotI; H, HindIII; E, EcoCRI. Arrows indicate gene orientation for the hsp70.1 or hsp70.3 allele. (B and D) Southern blotting analyses of tail DNA derived from wild-type (+/+), heterozygous (+/), or homozygous (/) hsp70.3 mice (B) or hsp70.1 mutant mice (D). The replacement of the hsp70.3 gene by the neo-lacZ cassette yields a 9-kb fragment in addition to the 6-kb wild-type fragment, and replacement of the hsp70.1 gene yields 12- and 9-kb fragments. (C and F) PCR-based genotyping assay amplifies fragments of 0.6 and 1.1 kb for the wild-type and targeted hsp70.3 loci (C) and fragments of 0.5 and 1.0 kb for the hsp70.1 loci (F), respectively.

Tissue-specific expression of hsp70.3 and hsp70.1 during embryonic development and in tissues of adult mice. The remarkable inducibility of hsp70i under acute stress conditions (i.e., thermal stress) led to the current consensus that tightly controlled regulatory mechanisms have evolved to ensure that this protein becomes active at the right time and place but is otherwise silent. The tissue- and cell-specific hsp70i expression profile, a critical aspect for understanding hsp70i function in vivo, has not been well defined, primarily because of the lack of adequate experimental approaches to monitor hsp70i expression in vivo.

+/

+/

View larger version (243K): [in this window] [in a new window]   FIG. 2.   Tissue-specific constitutive expression of hsp70.3 or hsp70.3 during late embryonic development. hsp70.1+/ and hsp70.3+/ embryos were recovered at E17, frozen, sectioned, and fixed and stained with X-Gal. Sections were counterstained with eosin (pink). Arrows indicate blue -galactosidase expression. Tissues presented are from the skin, nasal cavity, tongue, esophagus, stomach, forestomach, and kidney. Note that hsp70i expression was observed only in the epithelium of the forestomach (arrow), whereas no constitutive expression was noted in the stomach epithelium. Magnification for the tongue, stomach, and forestomach, ×20. Magnification for the skin, ×200. Magnification for all other tissues, ×50.

View larger version (223K): [in this window] [in a new window]   FIG. 3.   Tissue-specific constitutive hsp70.1 and hsp70.3 expression in adult mice. Tissues from wild-type controls (C57BL/6-129/SvJ)F2, hsp70.1+/, or hsp70.3+/ adult mice (8 to 12 weeks old) were harvested, frozen, sectioned, and fixed and stained with X-Gal. Sections were counterstained with eosin (pink). Arrows indicate blue -galactosidase activity. hsp70i expression is shown for the skin, tongue, esophagus, forestomach, hippocampus, cornea, kidney, glomeruli, and bladder. Magnification for the hippocampus, ×30. Magnification for the esophagus and kidney, ×70. Magnification for the skin, tongue, and forestomach, ×145. Magnification for all other tissues, ×290.

/

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Tissue-specific constitutive hsp70i protein expression is diminished in hsp70.1- or hsp70.3-deficient adult mice compared to wild-type controls. Levels of hsp70i protein were assessed by immunoblotting protein extracts from the brain, skin, esophagus, forestomach, and kidney tissues harvested from wild-type (+/+), heterozygous (+/), or homozygous (/) hsp70.1 or hsp70.3 mice. As a control for equal protein loading, the blot was probed for actin.

hsp70.1 or hsp70.3 gene activity is rapidly up-regulated following thermal stress in tissues lacking constitutive expression. We next examined the effects of heat shock on expression of hsp70i in tissues of adult mice. hsp70.1^+/ or hsp70.3^+/ mice were exposed to a mild whole-body thermal stress (42°C for 45 min), and hsp70i expression was assessed 6 h later. In agreement with earlier observations of increased hsp70i protein levels following heat shock, hsp70.1 or hsp70.3 expression increased dramatically in multiple tissues, especially those lacking basal hsp70i expression. As shown in Fig. 5, remarkably high levels of hsp70i expression were observed in the liver, small intestine, pancreas (although, interestingly, no expression was found on the islets of Langerhans), and adrenal cortex. In the spleen, hsp70i expression was found in a few accessory cells in the marginal zone and in vascular endothelial cells. No detectable expression in lymphoid cells was observed (data not shown). Expression of hsp70i in the heart was largely limited to cardiac muscles and the cardiovascular system (coronary arterial vessels). In the testis, marked expression was found in the interstitial (Leydig) cells, although spermatocytes lacked hsp70 expression.

View larger version (227K): [in this window] [in a new window]   FIG. 5.   Tissue-specific hsp70.1 and hsp70.3 induction following whole-body hyperthermic treatment of adult mice. hsp70.1+/ and hsp70.3+/ adult mice (8 to 12 weeks old) were semi-immersed in a circulating water bath at 42°C for 45 min and left to recover for 6 h before euthanatization. Untreated hsp70.1+/ mice were used as controls (left panels). Tissues were prepared and stained for -galactosidase activity as described in the legend to Fig. 3. Arrows indicate blue -galactosidase staining. No -galactosidase activity was detected in any of the tissue sections in wild-type mice (data not shown). hsp70.1 or hsp70.3 expression is shown for the liver, small intestine, pancreas, adrenal gland, spleen, heart, and testis. Magnification for the adrenal gland and heart, ×70. Magnification for all other tissues, ×145.

+/

+/

/

/

View larger version (30K): [in this window] [in a new window]   FIG. 6.   Reduced induction of hsp70i protein expression following whole-body hyperthermic treatment in hsp70.1- and hsp70.3-deficient adult mice. Wild-type (+/+), heterozygous (+/), or homozygous (/) hsp70.1- or hsp70.3-deficient mice were anesthetized, subjected to heat shock (42°C for 45 min), and allowed to recover for 6 h before euthanasia (indicated 42°C). Controls were mice anesthetized but not exposed to heat shock (indicated C). Protein extracts from different tissues were assessed by immunoblotting with antibody specific to hsp70i (C92). The data show heat-induced hsp70i protein levels in a selection of mouse tissues (heart, liver, pancreas, and intestine) that lack basal (constitutive) hsp70i expression.

Inactivation of the hsp70.1 or hsp70.3 gene results in deficient maintenance of acquired thermotolerance in bone marrow cells. The rapid induction of hsp70i in response to stress is thought to be fundamental to the cellular protection process. Thus, an increased level of hsp70i protein following sublethal heat stress, or hyperthermic preconditioning, profoundly attenuates all of the heat-induced cellular changes to a subsequent severe heat challenge (thermotolerance).

/

View larger version (48K): [in this window] [in a new window]   FIG. 7.   Thermotolerance induction in bone marrow cells. (A) Bone marrow progenitors from hsp70.3- or hsp70.1-deficient mice exhibit reduced capacity in maintenance of thermotolerance. Bone marrow cells from +/+ (), +/ (), or / () hsp70.1 or hsp70.3 mice were tested for their ability to develop thermotolerance. Cells preconditioned to heat (43°C for 20 min) were rechallenged with a more-severe heat (44°C for 40 min) at the times indicated during the recovery period at 37°C. The cells were subsequently plated for colony formation of granulocytes, macrophages, or a mixture of both granulocytes and macrophages, which were counted microscopically after the incubation of the cells at 37°C for 8 days. The data represent percent survival for each time point as described in Materials and Methods. Error bars indicate the standard errors of the means for three replicates. (B and C) The kinetics of heat-induced hsp70i protein expression in bone marrow cells is significantly reduced in hsp70.3- or hsp70.1-deficient mice. Bone marrow cells harvested from hsp70.1 or hsp70.3 +/+, +/, or / mice were left untreated or were heated at 43°C for 20 min. Cells were then incubated at 37°C for 1, 2, 4, 6, or 24 h before collection and analysis by immunoblotting with 35 µg of total cellular protein. The antibodies used are indicated on the left. -Gal is a mouse monoclonal antibody specific to -galactosidase protein. C92 recognizes the heat-inducible hsp70. 3A3 recognizes both the constitutive and the heat-inducible hsp70. A rabbit polyclonal antibody specific to actin was used as an indicator of loading. Protein extracted from non-heat-treated cells was used as a control (indicated C). Note that in hsp70.1/ bone marrow cells, an hsp70 protein band is still obtained. This represents hsp70.3 protein; the converse is true in hsp70.3/ mice.

Inactivation of the hsp70.1 or hsp70.3 gene results in increased sensitivity to thermal stress-induced apoptotic cell death in MEFs. It has been suggested that the mode of hsp70i action in acquired thermotolerance is the prevention of apoptotic death. hsp70i may function in this respect at both the level of cytochrome c release and initiator caspase activation via direct association with the caspase recruitment domain of Apaf-1 and inhibition of apoptosome formation. However, the physiological importance of this pathway is undefined.

View larger version (45K): [in this window] [in a new window]   FIG. 8.   Cumulative expression of both hsp70 genes after heat shock is essential for maximal levels of thermotolerance development and maintenance. (A) Inactivation of hsp70.1 or hsp70.3 genes results in deficient maintenance of acquired thermotolerance in MEFs. To induce thermotolerance, MEFs were preconditioned with a relatively mild heat shock (43°C for 20 min) and allowed to recover at 37°C for 6 or 24 h before challenge with a lethal heat shock (45°C for 30 min). Analyses were performed, following a further recovery at 37°C for 24 h, by staining with annexin V-FITC and propidium iodide and by a fluorescence-activated cell sorter. Percent cell survival of heat-treated cell populations was calculated based on the staining profile (annexin and propidium iodide double negative). Filled bars, control MEFs; hashed bars, hsp70.3/ MEFs; open bars, hsp70.1/ MEFs. (B) Exposure of hsp70.1- or hsp70.3-deficient MEFs to sublethal heat shock resulted in dramatically reduced induction of hsp70i expression. MEFs derived from +/+, +/, or / hsp70.1 or hsp70.3 mice were heat treated (43°C for 20 min), and protein was extracted following recovery at 37°C for the indicated time. hsp70i protein expression was detected by immunoblotting with antibody specific to hsp70i (C92). Protein extracted from non-heat-treated cells was used as a control (indicated C). hsp70.3 or hsp70.1 transcriptional activity was detected by -galactosidase protein analysis, and actin was used as an indicator of loading as described in Materials and Methods. (C) Inactivation of hsp70.1 and hsp70.3 genes increases MEF susceptibility to cell death through activation of the intrinsic apoptotic pathway. MEFs derived from +/+ or / hsp70.1 or hsp70.3 mice were preconditioned with a relatively mild heat shock (43°C for 20 min) and allowed to recover at 37°C for 6 or 24 h before challenge with a lethal heat shock (45°C for 30 min). Analyses were carried out following a further recovery at 37°C for 24 h, except for cytochrome c, which was analyzed following the 6-h recovery period. Procaspase 9, PARP, mitochondrial cytochrome c release, and actin, as a control for loading, from heat-treated cells were assayed by immunoblotting. Protein extracted from non-heat-treated cells was used as a control (indicated C). The lane designated 45°C represents protein extracted from cells heated 45°C for 30 min with no preconditioning treatment.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Although the functional and physical features of hsp70i as a molecular chaperone have been well studied in vitro, their precise physiological roles and the regulatory networks that determine cell- and tissue-specific expression of this chaperone molecule in vivo are not well understood. The current paradigm is that tightly controlled regulatory mechanisms have evolved to ensure that hsp70i becomes active at the right time and place under acute stress conditions (i.e., thermal stress) but is otherwise silent. However, evidence of spontaneous hsp70i protein expression during embryo development and postnatal growth has challenged this view (25, 31, 43). Another important issue concerns the functional requirement for both hsp70.1 and hsp70.3 genes, which encode identical proteins, not only in defense of the host against proteotoxic damage during adverse (patho)physiological conditions but also under normal conditions during development and in the adult. In the study described here, gene targeting was used to generate mice harboring a null mutation of the genes encoding hsp70i with an insertion of a -galactosidase gene into the hsp70.3 or hsp70.1 locus. Analyses conducted with these animal models provide direct evidence that hsp70i is not only up-regulated in response to stressful stimuli but also that transcriptional programs exist for tissue-restricted constitutive expression of this protein under normal conditions. The tissue-specific expression patterns observed for the hsp70.3 and hsp70.1 genes are virtually indistinguishable, suggesting a similar mode of regulation for both genes at the transcriptional level. While hsp70.3- or hsp70.1-deficient mice are viable and fertile and exhibit no obvious phenotypic abnormalities, inactivation of the hsp70.1 or hsp70.3 gene results in a marked reduction in hsp70i protein synthesis in different mouse tissues under both normal and heat stress conditions. This is reflected by increased sensitivity to thermal stress-induced apoptosis. It is interesting that disruption of one hsp70 allele seems to reduce overall hsp70 protein levels, despite the presence of a nondisrupted allele. This phenomenon is particularly notable following heat stress (Fig. 6, liver). While we currently have no explanation for this observation, it is possible that one hsp70 gene may influence the accumulation of other hsp70's at the protein level, perhaps by acting to chaperone correct protein folding. Clearly, this is worthy of further investigation, and these studies are under way in our laboratory. Overall, these findings suggest a redundancy of the individual genes in embryonic development but a requirement for both genes for optimal responses to stresses such as heat shock.

It is of great interest to understand the mechanisms and factors that determine stress-induced versus persistent expression of hsp70i under normal conditions, which is restricted to certain tissues throughout embryonic development and in the adult. The remarkable induction of hsp70 in Drosophila melanogaster can be achieved by a wide range of regulatory strategies, including the presence of multiple hsp70 genes in the genome (21), the maintenance of an open chromatin configuration on these genes even at normal temperatures (47) with RNA polymerase arrested at the transcription start site (39), and activation and nuclear transport of preexisting transcription factors (HSFs) within 1 min of temperature shift (46). These elaborate regulatory mechanisms have been proposed not only to ensure rapid induction of hsp70 upon heat shock but also to maintain a basal level of hsp70 expression at normal temperature and in the absence of detectable stress. It is likely that similar mechanisms are operative for the rapid stress-inducible or constitutive hsp70i expression in mammals. However, direct evidence for this is lacking. It should be noted that the pattern of constitutive and induced hsp70i expression in the mouse, as detected by staining for -galactosidase activity in this study, faithfully reveals the expression pattern of the endogenous hsp70i genes. However the data obtained by this method differ significantly from existing data in the literature (8, 24, 25, 44). This is due to the fact that many of the previous studies conducted in this context have relied on studying hsp70i expression using hsp70 promoter sequences to regulate transgene (lacZ or luciferase) expression, conditions that may not strictly reproduce its normal distribution. Tissue-restricted or poor gene expression is not an infrequent effect in transgenic mice and may result from the site of chromosomal integration of the transgene and/or an inability to include in the expression vector all sequences required to establish its native epigenetic organization. The fact that the hsp70i promoter (an approximately 0.5 to 0.7 kb sequence upstream from the start of transcription) mostly restricts expression of reporter genes to certain tissues and to stress-inducible conditions leads us to predict that additional, as yet unidentified, transcriptional regulatory elements are required for constitutive expression of hsp70i and that they are located in distant regions flanking the hsp70.1 or hsp70.3 gene.

Transcriptional activation of hsp70i is mostly dependent on the activation of the presynthesized HSFs which thereafter acquire the ability to bind the tetrameric heat shock elements present in the proximal promoter sequences of the hsp70.1 and hsp70.3 genes. In addition to playing a key role in the response to noxious stimuli, HSFs are widely believed to play a role in spontaneous hsp70i expression during embryogenesis and postnatal growth. However, only indirect and circumstantial evidence exists for this, and such data remains open to the criticisms detailed above. As far as the regulation of the tissue-specific hsp70i response to stress is concerned, the role of HSF1 as the critical factor for hsp70i protein synthesis following thermal stress is well characterized (35, 49). In this regard, analyses using the genetic approach of breeding hsp70.1^+/ -galactosidase mice on the HSF1-deficient genetic background reveal that heat-induced hsp70i expression in mouse tissues is entirely controlled by HSF1, but no significant changes in constitutive hsp70i expression are observed, which suggests that HSF1 is not required for organ-specific expression of hsp70i (unpublished data). While these results suggest that HSF1 is indispensable for the inducible expression of hsp70i upon thermal stress and obviously cannot be functionally compensated for by other transcription factors, for example, the structurally related HSF2, the specific function of these factors in constitutive hsp70i expression remains unresolved. As HSF2 is found to be expressed ubiquitously both in tissues with inducible or constitutive hsp70i expression (11, 38), it is less likely that this factor is the key player in constitutive hsp70i expression. In addition to the double overlapping heat shock element motifs, the highly conserved promoter (approximately 300 bp) of both hsp70.1 and hsp70.3 contains an array of regulatory elements, including at least two Sp1 boxes. Hence, it has been proposed that spontaneous expression of hsp70 in the embryo is regulated via Sp1 transcriptional activity (5). However, it is difficult to explain a tissue-specific regulated hsp70i expression solely on the basis of such a ubiquitous transcription factor as Sp1. Thus, relatively little is currently understood about the regulation of tissue-specific constitutive expression of hsp70i in the absence of a measurable stress response and its basal homeostatic function in particular tissues. Clearly, the complexity of the molecular interaction between transcription factors (mostly unknown) and trans or cis regulatory elements on the hsp70i locus make it difficult to analyze their contribution to the specific pattern of hsp70i expression. However, cooperation between transcription factors (combinatory regulation) as a mechanism for conferring tissue-specificity may help to explain the hsp70i expression patterns in our study. Thus, a combination of ubiquitous transcription factors (for example, many cell types produce Sp1, AP1, or AP2, and putative binding sites for such factors are present in the hsp70i promoter regions), perhaps in combination with cell type-specific factor(s), may regulate constitutive hsp70i expression in a tissue-specific manner. Finally, continuous expression of hsp70 can inhibit cell growth, and forced expression of hsp70 in Drosophila cells can cause a reduction in growth without affecting viability (9), but in other experimental settings has been shown to be involved in differentiation and cell proliferation (22, 45). It is apparent that these opposing functional features of hsp70 depend on the level of hsp70i expression in particular cells or tissues; for example, forced expression of hsp70 in cancerous cells can promote oncogenic transformation. Whether tissue-specific constitutive hsp70i expression will also have growth-inhibiting or -promoting effects cannot be easily assessed, but it is likely that such effects will depend on several other factors, such as cell cycle regulators. Further investigation of such interactions using the models described in this report will likely prove of value in elucidating these opposing facets of hsp70i expression. Clearly, the characteristic tissue-specific developmentally regulated pattern of hsp70i expression in this study is suggestive of the involvement of this molecule in the differentiation and/or proliferation of cells, perhaps acting to ensure protein fidelity in the development and maintenance of tissue-specific cellular components.

An increased demand for the chaperoning function of hsp70 occurs at many stages in the life of a cell and determines the outcome of a cell faced with death. It is well known that stressful conditions, including heat shock, may lead to cell death by three distinct modes: reproductive (clonogenic) cell death, apoptosis, and necrosis. While little is known about the mechanisms of stress-induced necrosis or reproductive death, in apoptosis the initial damage does not directly kill the cell but initiates specific signaling pathways that lead to cellular suicide. The rapid induction of hsp70i in the response to stress is thought to be fundamental to the cellular protection process. In this regard, there is compelling evidence that resistance to stress by cells primed with a mild heat shock that induces hsp's (especially hsp70i) may be due to down-regulation of the signaling events that initiate apoptosis. Increased sensitivity to thermal stress in MEFs deficient in either the hsp70.3 or hsp70.1 gene reported in this study can be explained at least partially via the function of hsp70i in modulation of the stress-induced intrinsic apoptotic pathway. However, heat shock can also induce a caspase-independent type of cell death, which is distinct from classical apoptosis and has been proposed to involve JNK protein kinase activity (13, 32, 36). Evidence for heat-induced JNK activity that remains at relatively higher levels for a prolonged period in MEFs deficient in the hsp70.1 or hsp70.3 gene versus the wild type supports this hypothesis (data not shown). Participation of hsp70i and heat shock-activated kinases in the caspase-independent cell death of MEFs is likely. Definitive assessment of the physiological importance of the above pathways in cellular protection from apoptosis in response to harmful stress stimuli would provide new insights into the role of hsp70i in different pathological states in vivo. Together, the hsp70i gene products may affect multiple apoptotic pathways, and cell type-specific differences may account for the various points of hsp70i intervention in different tissues. The availability of hsp70.1- and hsp70.3-deficient mice provides an opportunity to explore not only the extent to which the antiapoptotic function of hsp70i is required in the protection of various tissues and cell types from stress damage but also will enable us to investigate the molecular basis for the antiapoptotic function of hsp70i. In addition, identification and characterization of regulatory elements determining tissue-specific constitutive and inducible expression of hsp70.1 and hsp70.3 under environmental stress conditions should provide valuable insights into the role of hsp's as factors in the development and maintenance of tissue-specific functions in a host. Finally, the use of these animal models will allow the examination of whether hsp70i activity can be compensated for by other members of the hsp family or whether hsp70i has a unique function in cellular protection from environmental stress. Thus, these mutant mice provide a valuable experimental model to achieve a better understanding of the fundamental cellular processes in which hsp70i molecular chaperones engage in the response to environmental stresses as well as to determine their role in clinically relevant pathologies in humans.