INTRODUCTION

Interleukin (IL)¹ 1, the processed form of the prointerleukin 1 gene (IL1B), is expressed in activated monocytes and plays a number of roles in inflammation, including the mediation of fever, lymphocyte activation, and the regulation of acute phase genes (1, 2, 3). The expression and biological function of IL-1 in humans is regulated at a number of different levels. These include transcriptional activation of IL1B, IL-1 mRNA stability, posttranslational proteolytic processing of pro-IL-1, and inhibition of IL-1 receptor binding by a naturally occurring IL-1 antagonist (3, 4, 5, 6, 7). At the physiological level, IL-1 expression is subject to feedback inhibition through its stimulation of the hypothalamic-pituitary-adrenal axis with the subsequent release of glucocorticoids and antagonism of IL-1 expression at the transcriptional and posttranscriptional levels (5, 6, 7). The inhibition of expression of IL-1 and a number of other cytokines, including IL-2, granulocyte-macrophage colony stimulating factor (GM-CSF), tumor necrosis factor (TNF), IL-6, and interferon  observed at elevated temperatures in the fever range, suggests the existence of another feedback inhibitory mechanism in which elevated temperatures generated during the onset of fever antagonize expression of pyrogenic cytokines (1, 8, 9, 10). Heat shock factor (HSF) 1, the transcriptional activating protein of the heat shock genes, would seem a likely candidate to mediate such a thermally regulated feedback response (12, 13). We have therefore investigated the potential role of HSF-1 in the repression of IL-1 expression during heat shock. Recent findings linking temperature-dependent binding of HSF to inactivation of developmental gene loci in Drosophila suggest that HSF-1 could function as a mediator of transcriptional repression (14). At elevated temperatures, HSF-1 is rapidly converted from a latent cytoplasmic form to a nuclear form that binds the promoters of heat shock genes and activates transcription (12, 13). The activation of heat shock promoters is effected through the binding of HSF-1 to heat shock elements (HSEs), which consist of inverted repeats of the 5-base pair motif (NGAAN) (15). We have therefore investigated the role of HSF-1 binding to a putative HSE at position 67 in the IL1B promoter in the repression of the IL1B promoter at elevated temperatures (Fig. 1).

Sequence of the IL1B promoter between the HindIII site at 131 and TaqI site at +12 (18, 21) showing the putative HSE at 67.

EXPERIMENTAL PROCEDURES

Total RNA was isolated from controls, and cells were stimulated with lipopolysaccharide (LPS) (10 ng/ml) for 1 h at 37 or 42 °C followed by fractionation on a 1.5% agarose gel containing formaldehyde and transfer to GeneScreen Plus membranes (DuPont NEN). mRNAs were detected by hybridization to an IL-1 cDNA probe (1050-base pair cDNA fragment) (16), a cDNA probe from the coding region of the murine hsp70.1 gene encompassing 1271 base pairs between the XhoI and BglII sites (17, 18), and a 2.0-kilobase fragment of the human -actin cDNA isolated from clone HFBCA46 (ATCC) by EcoRI digestion. Equal loading was demonstrated with a control oligonucleotide probe for the 28 S rRNA (Clontech, Palo Alto, CA).

Human monocytic cells THP-1 (ATCC TIB 202) were grown in RPMI 1640 medium containing 10% fetal bovine serum (containing less than 0.06 endotoxin unit/ml), 0.5% penicillin/streptomycin, and 5 × 10⁵ M 2-mercaptoethanol. Transfection of these cells and CAT assays were carried out as described previously (19) using either the human IL1B promoter (plasmid 3ME-HT) or the human HSP70B promoter (p2500-CAT). p2500-CAT (Stressgen Biotech Corp., BC, Canada) contained a 5 fragment from the human HSP70B gene comprising 2.5 kilobase pairs of non-transcribed sequence and most of the RNA leader (20). HSF-1 expression vector pCMV-HSF-1 contained the human HSF-1 (hHSF-1) coding sequence (from Dr. Carl Wu, NCI) ligated into the pCMV5 expression vector (from Dr. David Russell, University of Texas Southwestern Medical Center). Mouse NIH-3T3 fibroblasts were grown in Dulbecco's modified Eagle's medium containing 10% bovine calf serum and 0.5% penicillin/streptomycin (17). They were transfected by the calcium phosphate method according to the manufacturer's instructions (Promega, Madison, WI). Cells were plated 18 h prior to transfection at 2 × 10⁵/dish, and plasmid DNA precipitates were added dropwise. All treatments contained the same total amount of DNA. Transfectants were incubated overnight in a 2.5% CO₂/air atmosphere. CAT protein concentration was measured by enzyme-linked immunosorbent assay using a standard assay kit (Boehringer Mannheim).

Recombinant HSF-1 (rHSF-1) was extracted from Escherichia coli expressing hHSF-1 from the pET22b vector (Novagen, Madison, WI) and purified to homogeneity using sequential (NH₄)₂SO₄ precipitation and chromatography on heparin-agarose and Mono-Q columns.² To analyze the binding of rHSF-1 to IL1B promoter fragments, probes were labeled by end filling with Klenow fragment using [³²P]dCTP at 6,000 Ci/mmol (DuPont NEN) to an activity of 100,000 cpm/ng and incubated (0.2 ng each) with rHSF-1 (0.7 µg), poly(DI·DC) (1.0 µg), dithiothreitol (2.5 mM), phenylmethylsulfonyl fluoride (1 mM), NaCl (40 mM), Na₂HPO₄ (10 mM), pH 7.4, in a total of 15 µl for 10 min at 20 °C. HSF-1·DNA complexes were then resolved by electrophoresis on 4% polyacrylamide gels in 0.5 × TBE (45 mM Tris, pH 8.0, 45 mM boric acid, 1 mM EDTA). Competition assays were carried out using a 25-fold molar excess of either a consensus HSE from the mouse hsp70.1 gene (17) or an isoform of this sequence bearing the mutation (mHSE) described in Fig. 1. To study the binding of nuclear HSF to IL1B promoter fragments, probes were incubated with 2.0 µg of nuclear extract as for rHSF-1 except that the reaction was buffered with 10 mM Tris, pH 7.4. Nuclear extracts were prepared as described previously (19). Electrophoretic mobility supershift assays were carried out by adding a 1:100 dilution of either anti-HSF-1 antibody 68-3 or preimmune serum from the same animals. Antibody 68-3 was raised against the carboxyl-terminal 14 amino acids of hHSF-1.³

RESULTS AND DISCUSSION

Exposure of THP-1 human monocytes to elevated temperatures resulted in divergent changes in the expression of IL-1 and HSP 70 mRNAs (Fig. 2A). The IL-1 mRNA, although abundantly expressed at 37 °C in the presence of LPS, was inhibited at 42 °C while the 2.6-kilobase pair HSP 70 mRNA was strongly induced at 42 °C (Fig. 2A). Equal loading was demonstrated by probing for the 28 S rRNA (Fig. 2A). Since prolonged heat shock is known to inhibit the expression of many proteins, we also examined the effects of heating at 42 °C on the expression of a constitutively expressed RNA polymerase II gene, -actin. Cells were exposed to LPS and heat shocked at the onset of IL-1 expression (1 h), at the peak of expression (2 h) and as IL-1 expression was declining during continuous exposure to LPS (6 h). In each case, although a 1-h heat shock at 42 °C inhibited IL-1 expression, the expression of -actin was neither increased by LPS nor inhibited by heat shock (Fig. 2A). Inhibition of IL-1 expression at 42 °C is therefore not the consequence of a general inhibition of gene expression.

A, reciprocal effects of heat shock on the expression of IL-1 and HSP 70 mRNAs in THP-1 cells. Total RNA was isolated from control cells, and cells were stimulated with LPS for 1 h at 37 or 42 °C and analyzed by Northern analysis as described under ``Experimental Procedures.'' Blots were probed sequentially for IL-1, HSP 70 mRNAs, and 28 S RNA as a control for equal loading. An additional experiment was carried out to determine whether the effects of heat shock were due to a general inhibition of gene expression (far right panel). THP-1 cells were exposed to LPS for 1, 2, and 6 h without further treatment or with heat shock at 42 °C for the final 1 h of incubation (0-1, 1-2, and 5-6 h, respectively). RNA was extracted and processed as described above and probed sequentially for IL-1 and -actin mRNAs. B, inhibitory effects of heat shock on LPS induction of the IL1B promoter. THP-1 monocytes were transfected with 10 µg of plasmid 3ME-HT and either left unstimulated (CONTROL) or induced with LPS (10 ng/ml) 18 h later (LPS). In a third treatment group, heat shock (2 h at 42.5 °C) was given immediately prior to LPS treatment (LPS + HS). CAT assays were then carried out after 24 h of incubation at 37 °C as described (19). Results are expressed as average percentage of the 3ME-HT activity in cells maximally stimulated with LPS. Error bars indicate standard deviations for a minimum of three repetitions. C, effect of heat and co-transfection of HSF-1 expression vector pCMV-HSF-1 on the activity of the human HSP70B promoter in THP-1 cells. Cells were transfected with the HSP70B promoter-CAT construct p2500-CAT (10 µg) (HSP-70) either with 4 µg of pCMV-HSF-1 (HSF-1) or without expression plasmid. Two further treatment groups were heat shocked (HEAT) 18 h after transfection (42.5 °C/2 h) in the absence or presence of HSF-1 expression plasmid co-transfection. All cultures were then incubated at 37 °C for 24 h to allow expression of CAT. Results are expressed as an average percentage of the CAT activity observed in heat-shocked cells transfected with p2500-CAT. Error bars indicate standard deviation for a minimum of three replicate assays. D, repression of the IL1B promoter in THP-1 cells transfected with pCMV-HSF-1. Cells were transfected with 3ME-HT (10 µg) without expression plasmid or with pCMV-HSF-1 at a range of doses (2, 4, 6, 8, and 10 µg) 18 h prior to the experiment. Cells were then incubated an additional 1 h at 37 or 40 °C prior to LPS stimulation followed by a 24-h incubation at 37 °C for expression of CAT. Results are expressed as in B. 3MEHT, 3ME-HT construct of the IL1B gene promoter and the SV40 enhancer. E, inhibition of IL1B promoter activity by pCMV-HSF-1 co-transfection in NIH-3T3 cells. The IL1B promoter was activated by co-transfection with Spi-1 expression vector pRCMVSpi-1 as described under ``Experimental Procedures.'' Cells were transfected with 3ME-HT (10 µg) either in the absence of HSF-1 expression or with pCMV-HSF-1 at a range of concentrations (1.25, 2.5, 5, and 10 µg) and with a constant amount of pRCMVSpi-1 (0.25 µg). A cytomegalovirus-driven expression vector containing the HSF-1 cDNA in the reverse orientation had no effect on the IL1B promoter activity (not shown). Results are expressed as a percentage of the maximal Spi-1-inducible IL1B promoter alone. Results from one out of four representative experiments are indicated. F, activation of the human HSP70B promoter in NIH-3T3 cells by heat and co-transfection of HSF-1 expression vector pCMV-HSF-1. Cells were transfected with p2500-CAT (10 µg) and, after an 18 h incubation, either untreated (Control) or heat shocked (HS) at 42.5 °C for 1 or 3 h. An additional group of cultures (HSF-1) was co-transfected with p2500-CAT and pCMV-HSF-1 (4 µg). All cultures were incubated for an additional 24 h at 37 °C to allow expression of CAT. Results are expressed as an average percentage of the CAT activity observed after 3 h at 42.5 °C in cells transfected with p2500-CAT. Results from one out of four representative experiments are indicated.

In order to examine the role of the putative HSE at 67 in IL-1 repression at elevated temperatures, we next investigated the effects of HSF-1 activation on IL1B promoter activity. The tissue-specific IL1B promoter consists of positions 131 to +12 and functions as such when ligated to either the IL1B upstream enhancer or the SV40 enhancer (Fig. 1). THP-1 cells were transfected with the IL1B promoter CAT reporter construct 3ME-HT which contains the IL1B gene promoter (Fig. 1) and the SV40 enhancer (19). The IL1B promoter had a low basal activity in controls, which was strongly stimulated by LPS, and this LPS-inducible activity was reduced approximately 50% by heat shock (Fig. 2B). These heat shock conditions (2 h at 42.5 °C) led to transcriptional activation of HSF-1 as demonstrated by marked stimulation of the heat-inducible human HSP70B promoter (Fig. 2C). The reciprocal effects of heat shock on the HSP70B and IL1B promoters thus resemble the effects observed at the level of mRNA synthesis, with concomitant activation of HSP 70 and inhibition of IL-1 mRNA expression (Fig. 2, A-C).

In order to provide direct evidence for HSF-1 involvement, we next examined the effect of heat-independent HSF-1 expression on IL1B promoter activity. This was accomplished by high level expression of HSF-1 from the cytomegalovirus promoter in THP-1 cells in the absence of heat (Fig. 2D). Promoter response to LPS was inhibited in cells co-transfected with the HSF-1 expression vector in a dose-dependent manner and declined to less than 50% of control values at higher concentrations of plasmid (Fig. 2D). Activity was further decreased to a minimum of 20-25% when HSF-1 expression was combined with heat shock at 40 °C for 60 min (Fig. 2D). By contrast, HSF-1 expression did not result in significant stimulation of the HSP70B promoter in THP-1 cells unless HSF-1 expression was accompanied by heat shock (Fig. 2C). Transfection of HSF-1 alone resulted in 2-fold activation of the HSP 70B promoter and 30-fold stimulation when combined with heat shock (Fig. 2C). Although the significance of these findings is not entirely clear, a number of published studies indicate that the activation of HSF-1 is at least a two-part process and that an intermediate form exists that binds HSE but does not activate heat shock transcription (17, 25, 26). HSF-1 produced in THP-1 cells from the expression plasmid was able to repress the IL1B promoter while apparently in the intermediate form that is not competent for transcriptional activation. Interestingly, repression by HSF-1 was amplified by a heat shock condition (1 h at 40 °C) that alone causes minimal HSF-1 activation (Fig. 2D). These minimal conditions were used to keep cell stress to a minimum while amplifying HSF-1 activity. Our unpublished studies indicated that HSF-1 could be activated in HSF-1 transfectants using very mild conditions that do not activate the heat shock response in untransfected cells.⁴ These data are consistent with earlier studies indicating that HSF-1 activation from its latent form depends both on temperature and HSF-1 concentration (reviewed in Refs. 12 and 13). We also examined IL1B repression in a distinct cell line (NIH-3T3), which differs from THP-1 cells in that HSF-1 expression does activate the HSP70B promoter. Because non-myeloid cells such as NIH-3T3 fibroblasts are deficient in the expression of the monocyte-specific transcription factor PU.1/Spi-1 required for IL1B promoter activity, we used Spi-1 transfection to activate the promoter in these cells (19). Transfection with HSF-1 expression vector caused inhibition of Spi-1 activated IL1B promoter activity in NIH-3T3 cells with a similar concentration dependence as that observed for LPS induced activity in THP-1 cells (Fig. 2E). In contrast with THP-1 cells, HSF-1 expression did activate the HSP70B promoter in NIH-3T3 cells (Fig. 2F). HSF-1 transfection led to promoter activation that was approximately 50% of that observed in cells heat shocked for 3 h at 42.5 °C, conditions that lead to maximal HSP70B promoter stimulation (Fig. 2F). Thus although there are differences between THP-1 and NIH-3T3 cells in terms of their negative regulation of HSF-1 function as a transcriptional activator of HSP70B (Fig. 2, C and F), transcriptional repression by HSF-1 was similar between the cell types (Fig. 2, D and E). These findings suggest that the events that take place during transcriptional repression and activation by HSF-1 may not be identical. Conversion of HSF-1 to a transcriptionally active form may require overcoming additional regulatory barriers compared with the control mechanisms involved in repression by HSF-1.

In order to localize the HSF-1 binding site within the IL1B promoter, the HT fragment encompassing the entire promoter was divided into two fragments, HD (131 to 59) and DT (59 to +12) (Fig. 1). Recombinant HSF-1 (rHSF-1) bound specifically to the HD fragment that contains the putative HSE at 67 with an avidity similar to that of a consensus HSE oligonucleotide from the murine hsp70.1 promoter (HSE) but did not bind to the DT fragment (Fig. 3A). HSF-1 also bound to a smaller oligonucleotide (100) that spans the putative HSE (positions 100 to 54). Insertion of a triple mutation (Fig. 1) into the HSE site of this oligonucleotide (100 mHSE) resulted in the loss of rHSF-1 binding (Fig. 3A). Similar mutations have been shown previously to disrupt HSE function (26, 27). The specificity of rHSF-1 binding was also supported by competition with unlabeled HSE. A variant HSE oligonucleotide (mHSE) incorporating the mutations described above did not compete with HD or 100 for rHSF-1 binding (Fig. 3A). Nuclear HSF extracted from heat-shocked THP-1 cells also bound specifically to the HD and 100 probes with an avidity similar to that of the consensus HSE used above (Fig. 3B). The complex between the 100 oligonucleotide and nuclear HSF was supershifted by anti-HSF-1 antibody but not by the corresponding preimmune serum, implicating HSF-1 as the species bound to the IL1B promoter (Fig. 3B). Association of HSF-1 with the IL1B promoter (HT fragment) had little influence on the binding of other factors extracted from THP-1 cells (not shown). The HT fragment formed numerous complexes with nuclear factors including NF-IL6 and Spi-1 none of which were displaced by the binding of HSF-1. These results are representative of a series of experiments consistently showing that binding of HSF-1, whether extracted from nuclei or added in purified form to nuclear extracts, did not displace other species from the IL1B promoter (not shown).

In order to investigate a functional role for the HSE at 67 in IL1B promoter repression by HSF-1, the inactivating HSE mutation described above was incorporated into the 3ME-HT construct (Fig. 1). This IL1B construct containing the mHSE, although fully inducible by LPS, was completely resistant to repression by HSF-1 (Fig. 4, columns 4-6). In contrast, the activity of the wild-type promoter (WT) was repressed 50% by HSF-1 co-transfection as described earlier (Fig. 4, columns 1-3). Both constructs were equally inducible by LPS and differed only in the resistance of mHSE to repression by HSF-1, indicating that repression of the IL1B promoter by HSF-1 is mediated through the HSE at 67. Since HSF-1 only binds to HSE sequences with high affinity in its trimeric form, the data indicate a scheme involving activation of HSF-1 to its trimeric form during exposure to elevated temperatures, binding to the HSE at 67 in the IL1B promoter, and repression of the promoter by the resulting HSF-1·HSE complex (12, 13, 15). In addition, mutation of the adjacent NF-IL6 site (91) in the IL1B promoter (mNFIL6) substantially reduced the LPS responsiveness (Fig. 4, columns 7-9). This residual, but otherwise significant, LPS-induced activity in the mNFIL6 construct was not repressed by co-transfection with HSF-1 expression vector (Fig. 4, column 9). The importance of the NF-IL6 site at 91 in maximal IL1B promoter activity has been reported (19), and the degree of repression of the IL1B promoter by HSF-1 (Fig. 2D) is similar to the loss of activity that occurs on introducing an inactivating mutation into the NF-IL6 site (Fig. 4). The resistance of this residual, NF-IL6-independent IL1B promoter activity to repression by HSF-1 may indicate an interaction between HSF-1 and NF-IL6 in repression.

We conclude that HSF-1 binds to the IL1B promoter and represses its activity in a manner dependent upon the presence of an intact HSE at position 67. The exact mechanism for HSF-1 acting as a repressor rather than an activator of the IL1B promoter is, however, not clear although our data suggest that transcriptional repression by HSF-1 may involve mechanisms distinct from those involved in activation (Fig. 2, C-F). A model for repression based on competition for binding sites on the IL1B promoter between HSF-1 and essential transcription factors seems unlikely as neither NF-IL6 nor Spi-1 was displaced from the IL1B promoter by HSF-1 binding. However, the requirement for a functional NF-IL6 site at 91 in order to demonstrate HSF-1 repression (Fig. 4) suggests that repression is mediated through cooperative interactions between HSF-1 and other factors on the IL1B promoter. Although the significance of these findings in terms of the physiological function of IL-1 has not been addressed here, the results may have implications in terms of IL-1 regulation during whole body stress and infection. It should be pointed out that some of the experiments in the present report involve exposing cells to temperatures, such as 42.5 °C, that are above the fever range (Fig. 2, B and C). However, fever involves the exposure of cells to temperatures in the 39-40 °C temperature range for extended periods (up to 80 h), and these conditions are also known to inhibit cytokine expression (8, 9, 10, 11). The biological effects of exposure to heat shock are a function of both time and temperature, and thus 2 h at 42.5 °C, although non-physiological, may reflect the effects of a longer exposure at temperatures in the fever range (28). IL1B repression by HSF-1 could thus play a role in limiting potentially damaging responses such as toxic shock, fever, and inflammation that are mediated in part by the rapid and abundant expression of IL-1 in blood monocytes and tissue macrophages (1, 8, 9, 10, 11). Our findings therefore suggest a novel role for HSF-1 in gene repression during the heat shock response and in the feedback regulation of cytokines in the acute phase and febrile responses.