INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

INTERLEUKIN (IL)-6 is a multifunctional cytokine produced by a wide variety of cells and is known to play important roles in immunological responses, hematopoiesis, host defense, and acute phase reaction (1). IL-6 is a member of the family of cytokines that includes IL-11, leukemia inhibitory factor, ciliary neurotrophic factor, oncostatin M, and cardiotrophin 1. These cytokines share glycoprotein (gp) 130 as a common singling subunit for their receptors and transduce their signals through the Janus kinase (JAK)-signal transduction and activation of transcription (STAT) and Ras-mitogen-activated protein kinase (MAPK) nuclear factor (NF)/IL-6 pathways. A number of acute phase response genes contain binding sites for STAT and/or NF/IL-6 transcription factors [such as CCAAT/enhancer binding protein (C/EBP)-] in their promoters and are induced by IL-6 through the gp130 pathways (11, 15).

The role of IL-6 in the heart is complex. IL-6 may be involved in cardiac hypertrophy, because activation of gp130 signaling has been shown to cause cardiac hypertrophy in transgenic mice; furthermore, treatment of neonatal rat myocardial cells together with IL-6 and its soluble receptor induced myocardial hypertrophy in vitro (12). IL-6 mRNA is reportedly induced in the myocardium of a canine model of ischemia and reperfusion (3, 18). The plasma levels of IL-6 are elevated in patients with either acute myocardial infarction (14) or sepsis, and IL-6 is considered by some to be the cytokine best correlated with the severity of sepsis (20). Because IL-6 reportedly exerts a negative inotropic effect on hamster papillary muscle and human heart pectinate muscle (7, 8), IL-6 may play a role in the depression of myocardial contractility in sepsis or myocardial infarction.

Whereas induction of IL-6 after ischemia-reperfusion has been well documented (3, 10, 18), little is known about its expression during sepsis/endotoxemia. In the present study, we characterized the endotoxin-induced expression of cardiac IL-6 in a mouse model by localizing its expression primarily in nonmyocardial cells and defining IL-1 as a major mediator of the expression. With the use of IL-6-deficient mice, we also demonstrate a possible role for IL-6 as a downregulator of proinflammatory mediators in the heart during endotoxemia.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Four-month-old male C57BL/6 mice (28.5 ± 2.0 g body wt) were purchased from Charles River Laboratories (Wilmington, MA). Two-month-old male IL-6-deficient (/) mice (17) and control (+/+) mice were obtained from the Jackson Laboratory (Bar Harbor, ME) and used when they became 4 mo old. The genetic background of these mice was either B6x129 or C57BL/6. Three-month-old male Sprague-Dawley rats (275-310 g body wt) were purchased from Harlan Sprague Dawley (Indianapolis, IN). Before the experiments were conducted, the animals were kept at least 10 days in a 12-h light cycle and fed a standard chow diet ad libitum. Animals were injected intraperitoneally with bacterial lipopolysaccharide (LPS) derived from Pseudomonas aeruginosa (Sigma Chemical, St. Louis, MO) and dissolved in physiological saline solution. At the indicated time points, animals were killed by cervical dislocation, blood was collected by decapitation, and tissues were quickly dissected. Except for the LPS-dose-response experiment, mice were injected with 50 µg (1.8 mg/kg body wt) of LPS. None of the mice died after injection with this dose of LPS. All procedures have been approved by the University of Texas Medical Branch Institutional Animal Care and Use Committee.

Northern blot analysis. Each excised heart was quickly cross cut into four pieces, blotted onto paper towels to remove excess blood, transferred into a cryovial, rapidly frozen in liquid nitrogen, and stored at 80°C. The frozen tissues were individually transferred into guanidine-phenol solution and processed with a tissue homogenizer (Polytron PT3000, Brinkmann Instruments, Westbury, NY), and the total RNA was isolated by the method of Chomczynski and Sacchi (5). The RNA samples (20 µg) were electrophoretically fractionated through 1.0% agarose gels containing 3% formaldehyde buffered in 20 mM MOPS and 1 mM EDTA at pH 7.4. The integrity of the RNA and the equality of loading were verified by the intensities of rRNA in ethidium bromide-stained gels. The RNAs were transferred overnight from the gels to Zeta-Probe nylon membranes (Bio-Rad Laboratories, Hercules, CA) and fixed by ultraviolet cross-linking. Radiolabeled probes were prepared from mouse cDNA by the random priming technique with [-³²P]dCTP using Megaprime DNA labeling systems RPN 1606 (Amersham Pharmacia, Piscataway, NJ). DNA plasmids containing the mouse IL-1 and IL-6 cDNA (pMuIL-1 and pcD-mIL-6, respectively) were obtained from Dr. D. Pennica (see Ref. 9) and Dr. F. Lee (see Ref. 4), respectively. The plasmid K3-1.1 containing the mouse intercellular adhesion molecule (ICAM)-1 cDNA was purchased from American Type Culture Collection (Rockville, MD). For the tumor necrosis factor (TNF)- probe, a 326-bp cDNA fragment was amplified by RT-PCR using 1 µg of total heart RNA from a mouse that was injected with LPS 6 h before being killed. The primer sequences and reaction condition of the RT-PCR were described previously (23). The PCR product was agarose gel purified and used directly for radiolabeling. The methods for hybridization and washing were described by Church and Gilbert (6). Because all of the cDNA probes were derived from mice, the hybridization and washing were performed under stringent conditions (at 65°C) for detecting mouse mRNA or under less stringent conditions (at 60°C) for detecting rat mRNA. The washed filters were exposed to Kodak XAR-5 film in the presence of an intensifying screen at 80°C. As a control, the filters were reprobed with radiolabeled DNA specific for the 18S ribosomal subunit. The relative amounts of RNA were determined by densitometric analysis of Northern blot autoradiograms using a scanning, transmitting densitometer. The levels of mRNA for cytokines were normalized to the 18S RNA signals.

Isolation of cytoplasmic proteins from mouse hearts and cytokine measurements. A previously described method for isolating nuclear proteins from mouse brains (23) was modified to isolate cytoplasmic proteins from mouse hearts. Each heart was quickly minced into at least 50 small pieces with a razor blade in ice-cold Tris-buffered saline [25 mM Tris · HCl (pH 7.4), 5 mM KCl, and 137 mM NaCl]. The tissue pieces were rinsed with the buffer to remove blood and then processed in a Dounce homogenizer (Kontes Glass, Vineland, NJ) with 2 ml of ice-cold buffer A including proteinase inhibitors and phosphatase inhibitors (23). Homogenates were left on ice for 15 min, a volume of 10% Nonidet P-40 was then added (final concentration 0.6%), and the samples were centrifuged at 3,000 g for 10 min at 4°C. The supernatants (cytoplasmic fraction) were aliquoted and stored at 80°C. The protein concentrations were determined using protein assay reagent (Bio-Rad Laboratories). The cytokine concentrations in the samples were determined by ELISA using ELISA kits specific for mouse cytokines that were purchased from Endogen (Woburn, MA).

In vitro cell culture. Primary cultures of cardiac cells were prepared essentially by the method of Simpson (24). The left ventricles from 1- to 2-day-old Sprague-Dawley rat pups were minced and subjected to serial trypsin-collagenase digestion (0.125%-0.025%, respectively) to release single cells. After the cells underwent the final digestion, the cells were washed and plated for 30 min in medium 199 with 10% fetal bovine serum. Nonattached cells (the myocyte-enriched population) were counted and replated in 100-mm Corning dishes at 4 × 10⁶ cells/plate in the same medium with cytosine arabinofuranoside (5 µg/ml), insulin (10 µg/ml), transferrin (10 µg/ml), and penicillin (50 U/ml). Cells attached to the dish within the 30 min of the initial plating were taken as the nonmyocyte-enriched population. Two days later, the cultures were used for experiments. With this protocol, we usually obtained the myocyte-enriched cultures with ~80% cells beating. The nonmyocyte-enriched cultures contained <1% beating cells.

Immunohistochemistry. Mouse hearts were cut horizontally into halves, and cryosections were prepared. The sections were incubated overnight at 4°C with polyclonal rabbit anti-rat IL-6 antibody SC-1267 (Santa-Cruz Biotechnology, Santa Cruz, CA). After they were serially washed with phosphate-buffered saline, the slides were incubated with biotinylated secondary antibody for 45 min at room temperature. The sections were developed with ABC reagent (ABC kit, Vector Laboratories, Burlingame, CA) for 45 min. After they were incubated in Fast Red (Sigma Chemical, F-4648) containing 1 mM levamisole, the slides were counterstained with Mayer-hematoxylin and mounted for light microscopy (22).

In situ hybridization. Heart tissues were fixed in 4% paraformaldehyde for 12 h and embedded in paraffin. The embedded tissues were cut into 5-µm thick sections and dried on glass slides at 37°C overnight, followed by dewaxing and rehydration. Slides were incubated with proteinase K (2.5 µg/ml, Sigma) in 100 mM Tris · HCl (pH 7.6) and 10 mM EDTA for 30 min at 37°C and then in 0.25% acetic anhydride in 0.1 M triethanlamine (pH 8.0) for 10 min, followed by dehydration and air-drying. The mouse IL-6 cDNA probe was labeled using the MAXIscript kit (Ambion, Austin, TX) and [³⁵S]UTP to generate sense and antisense riboprobes. Hybridization with the sense strand served as a control. Hybridization was performed at 55°C for 12 h as described (22). The slides were dehydrated and first exposed to Kodak XAR-5 film for 3 days. For microautoradiography, the slides were then coated with NTB2 emulsion (Eastman Kodak, New Haven, CT), exposed in the dark for 2 wk, and counterstained with hematoxylin and eosin. All sections were examined by bright- and dark-field microscopy.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Endotoxin-induced IL-6 mRNA in hearts and other tissues. IL-6 mRNA was not detectable in the hearts of untreated and saline-injected mice by Northern blot analyses. Three hours after the mice were injected with LPS, IL-6 mRNA was strongly induced in the mouse heart. The induced levels increased in a dose-dependent fashion up to a dose of 100 µg of LPS (3.5 mg/kg body wt) and did not increase further with >100 µg of LPS (Fig. 1A). For the rest of the mouse studies, a nonlethal dose of 50 µg (1.8 mg/kg body wt) was chosen, and noninjected mice were used as controls.

View larger version (78K): [in this window] [in a new window]   Fig. 1.   Northern blot analyses demonstrating that lipopolysaccharide (LPS) induces interleukin (IL)-6 mRNA in the heart and other tissues. A: induction of cardiac IL-6 mRNA by various doses of LPS. Mice were killed 3 h after injection with various doses of LPS (from left to right lanes: noninjected control, saline-injected, and LPS-injected at doses ranging from 13 to 400 µg per mouse, i.e., 0.4 to 14.0 mg/kg body wt). Each lane consists of total RNA equally mixed from 2 mouse hearts. B: LPS-induced IL-6 mRNA in the heart and spleen. C: LPS-induced IL-6 mRNA in the whole brain, heart, lung, kidney, liver, skeletal muscle from thigh, and whole blood. Mice were killed 1 or 3 h after injection with 50 µg of LPS (1.8 mg/kg body wt). Control mice (0 h) did not receive injection. Each RNA blot was hybridized first with IL-6 probe (top) and then with 18S probe (bottom). Each lane in B and C represents an individual mouse tissue. The same experiments were repeated 2 more times with similar results.

IL-1 and TNF- as mediators for induction of cardiac IL-6 during endotoxic shock. We compared the induction of IL-6 mRNA with those of other inflammatory mediators, IL-1, TNF-, and ICAM-1. With the use of Northern blot hybridization, we analyzed the RNA from the hearts of mice killed at 0.5, 1, 1.5, 3, and 6 h after LPS injection and from control noninjected mice. The results shown in Fig. 2 demonstrate that IL-6 mRNA became detectable at 1 h and peaked at 3 h after LPS injection. Strong induction of IL-1 mRNA occurred well before that of IL-6, appearing as early as 0.5 h and peaking at 1 h. TNF- mRNA was also detectable at 0.5 h and peaked at 1 h. Unlike these cytokines, ICAM-1 mRNA was detectable at low levels without LPS injection (0 h), and the induction occurred at 1 h and peaked at 1.5 h (Fig. 2, A and B). We also measured IL-6 protein levels in the heart over the same time course. Until 1.5 h after the mice were injected with LPS, IL-6 protein levels were very low or below the limits of detection. A rapid increase and decline of IL-6 protein levels was seen at 3 and 6 h, respectively, demonstrating that the pattern of IL-6 protein levels follows IL-6 mRNA levels (Fig. 2C)

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Time course of tumor necrosis factor (TNF)-, IL-1, IL-6, and intercellular adhesion molecule (ICAM)-1 gene expression in mouse hearts in response to LPS administration. Mice were killed 0.5, 1, 1.5, 3, and 6 h after LPS injection. Noninjected control mice are designated as 0 h. Induction of each mRNA was analyzed by Northern blot analysis, and the IL-6 protein levels were measured by ELISA. A: representative set of autoradiograms of the Northern blot analyses. Each lane consists from total RNA equally mixed from 2 mouse hearts. An RNA blot was sequentially hybridized with probes for TNF-, IL-1, IL-6, ICAM-1, and 18S. A similar experiment was repeated at least twice for each gene with almost identical results. B: densitometric analyses of the autoradiograms of A. The mRNA levels were standardized with 18S RNA and shown relative to the peak mRNA level of each cytokine. C: IL-6 protein levels in total heart cytoplasmic fractions measured by ELISA. Each time point represents an average and SD of 2 to 6 mice.

View larger version (76K): [in this window] [in a new window]   Fig. 3.   Northern blot analysis demonstrating induction of IL-6 mRNA by IL-1 and TNF- in myocytes and nonmyocytes isolated from neonatal rat heart. Cells were treated with either IL-1 or TNF- (10 ng/ml for 2 h) in vitro. The left 2 lanes are RNA from in vivo left ventricles removed from 3-mo-old rats killed 3 h after injection with either LPS (2.5 mg/kg body wt) or saline as a control. The RNA blot was hybridized first with IL-6 probe (top) and then with 18S probe (bottom). The results shown here represent 1 of 2 identical experiments with very similar results.

Localization of cardiac IL-6 expression during endotoxic shock. The in vitro cell study demonstrated that IL-6 mRNA is induced in the nonmyocyte cultures to levels approximately threefold higher than in myocyte cultures (Fig. 3). This result suggests that cardiac IL-6 mRNA is expressed mainly in nonmyocyte cells of the heart during endotoxic shock. To verify this observation in vivo, we performed in situ hybridization and immunohistochemical analyses to localize cardiac IL-6 mRNA and protein, respectively. Autoradiograms of the in situ hybridization analysis showed that the induced IL-6 mRNA signals were evenly distributed in the whole heart (Fig. 4C). The in situ time course of IL-6 mRNA induction after LPS injection correlated well with the results of Northern blot analyses (Figs. 2 and 4, A-C). Through microautoradiography, the strongest signals of IL-6 mRNA were localized only in nonmyocytes, whereas weaker signals were seen in myocytes (Fig. 4G). Immunohistochemical analyses detected IL-6 protein mainly in and around nonmyocytes after LPS injection (Fig. 5). Therefore, we conclude that the induction of cardiac IL-6 during endotoxemia occurs mostly in nonmyocytes.

View larger version (141K): [in this window] [in a new window]   Fig. 4.   In situ hybridization localizing the induced IL-6 mRNA in mouse hearts in response to LPS administration. Control mice did not receive injection (A and E). After LPS injection, mice were killed at 1 h (B and F) or 3 h (C, D, G, and H). Heart sections derived from these mice were hybridized with radiolabeled IL-6 antisense mRNA (A-C and E-G) or sense mRNA (D and H). A-D: autoradiograms of whole hearts. E-H: microautoradiograms of the same hearts shown with ×400 higher magnification.

View larger version (151K): [in this window] [in a new window]   Fig. 5.   Immunohistochemical localization of IL-6 protein in the mouse heart in response to LPS administration. A: control mouse heart without LPS administration. B: heart from a mouse 6 h after LPS injection.

Augmented induction of the IL-1, TNF-, and ICAM-1 genes in hearts of IL-6-deficient (/) mice. To elucidate the role of IL-6 in regulating gene expression in the heart during endotoxemia, we compared the induction of various genes in the hearts of IL-6-deficient (/) mutant mice and non-IL-6-deficient (+/+) mice. We killed the mice 6 h after they were injected with LPS and performed Northern blot analyses for mRNA from six stress-inducible genes: IL-6, IL-1, TNF-, ICAM-1, C/EBP-, and C/EBP-. As shown in Fig. 6, A and B, the LPS-induced IL-6 mRNA was detected only in non-IL-6-deficient (+/+) mice but not in IL-6-deficient (/) mice, thereby confirming the IL-6-deficient genotype of these mice. Although the mRNAs for ICAM-1, C/EBP-, and C/EBP- were constitutively detected without LPS injection, there were no significant differences in the two mouse groups. The induced mRNA levels for IL-1, TNF-, and ICAM-1 were significantly higher in the hearts of IL-6-deficient (/) mice than non-IL-6-deficient (+/+) mice (2.2-, 2.4-, and 2.5-fold, respectively). The mRNA level for C/EBP- and C/EBP- were also induced in the hearts of both IL-6-deficient (/) and control (+/+) mice after LPS injection, but there were no significant differences between the two groups. We further compared protein levels of IL-1 and TNF- by ELISA in the hearts of IL-6-deficient (/) and control (+/+) mice (both with C57BL/6 genetic background) 6 h after LPS injection. The levels for IL-1 and TNF- were higher in the hearts of IL-6-deficient (/) mice than those in non-IL-6-deficient (+/+) mice (1.8- and 2.8-fold, respectively; Fig. 6C). The augmented induction of genes for IL-1, TNF-, and ICAM-1 in the IL-6-deficient mice suggests that IL-6 has a role in downregulating the expression of these three genes in the heart during endotoxemia.

View larger version (33K): [in this window] [in a new window]   Fig. 6.   Differential expression of various inflammatory genes in the hearts of IL-6-deficient (/) vs. non-IL-6-deficient (+/+) mice in response to LPS administration. Both / and +/+ mice were killed 6 h after injection with LPS, and the heart tissues were subjected to either RNA analyses (A and B) or protein analyses (C). A: representative autoradiograms of the Northern blots. The RNA blots were sequentially hybridized with probes for IL-6, IL-1, TNF-, ICAM-1, CCAAT/enhancer binding protein (C/EBP)-, C/EBP-, and 18S. Lanes 1 and 2, / mice without LPS injection; lanes 3 and 4, +/+ mice without LPS injection; lanes 5 and 6, / mice injected with LPS; and lanes 7 and 8, / mice injected with LPS. B: summary of densitometric analyses. Each bar represents an average of 4 mice. The average value of +/+ mice was set at 1 for each mRNA analysis. Error bars represent SD. *P < 0.05 and **P < 0.01 with Student's t-test, / vs. +/+ mice. These experiments were performed using 2 types of / mice with different genetic backgrounds (B6x129 and C57BL/6), and the results from both were very similar as to elevated mRNA induction of IL-1, TNF-, and ICAM-1 but not C/EBP- and C/EBP- in / mice. C: summary of ELISA analyses on IL-1 and TNF- protein levels. Each bar represents an average of 2 mice. Each error bar represents the range of values. The average value of +/+ mice was set at 1 for each protein analysis.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Previously, Troutt and Lee (26) detected IL-6 mRNA in the heart, kidney, and spleen but not in the brain, liver, thymus, lung, and bone marrow of mice 3 h after they were intravenously injected with 5 µg of Salmonella typhosa-derived LPS. In the present study, we detected IL-6 mRNA not only in the heart, kidney, and spleen but also in the brain, lung, liver, and skeletal muscle of the mice 3 h after they were intraperitoneally injected with 50 µg of LPS. In both studies, the induced levels of IL-6 mRNA were highest in the heart, suggesting that IL-6 may have a significant biological role in the heart during endotoxemia. The additional detection of IL-6 mRNA in the brain, lung, and liver, which were not found previously, may be due to differences in doses and types of LPS or route of LPS administration.

Through in situ hybridization and immunohistochemistry, we demonstrated that cardiac IL-6 induction occurs mainly in nonmyocytes during endotoxemia. These results are distinct from IL-6 induction after myocardial infarction, in which IL-6 is induced in both the myocardium and in infiltrating mononuclear cells (10). Induction of IL-6 by IL-1 or TNF- in cultured myocytes in vitro has been demonstrated previously (10, 29). However, new findings in the present study demonstrated that nonmyocytes respond to IL-1 with an induction of IL-6 mRNA and, more importantly, that this response is approximately threefold stronger compared with that of myocytes. Thus these results strongly support our interpretation of these in vivo observations that the nonmyocyte is the predominant cell type that expresses IL-6 in the heart during endotoxemia. However, myocytes also express IL-6 at low levels during endotoxemia. We detected this as weak IL-6 mRNA signals in myocytes by in situ hybridization as well as through a moderate induction of IL-6 mRNA that occurred in myocytes in vitro after stimulation with either IL-1 or TNF-. It was intriguing that TNF- at 10 ng/ml concentration moderately induced IL-6 mRNA in myocytes but not in nonmyocytes. These results suggest that cytokine-mediated stress responses exhibit differential levels of sensitivity to LPS in myocytes versus nonmyocytes. This may be due to differences in the number of cytokine receptors on these cells. The cardiac nonmyocytes include fibroblasts, macrophages, vascular endothelial cells, and smooth muscle cells. All of these cells are known to express IL-6 in certain conditions (1, 16). Further study will be required to identify which of these cell types express IL-6 in the heart during endotoxic stress.

Because IL-6 is induced together with IL-1 and TNF- and is a strong activator of hepatic acute phase response (2), it has been classified as a proinflammatory cytokine. However, several studies have suggested that IL-6 may play a role in the downregulation of other proinflammatory genes. For example, IL-6 has been reported to suppress LPS-induced TNF- in serum or IL-1 mRNA levels in the liver and spleen in mice (27). More recently, it was reported that induction of TNF- and macrophage inflammatory protein 2 during acute lung inflammation is higher in IL-6-deficient (/) mice than in non-IL-6-deficient (+/+) mice, suggesting that IL-6 downregulates these inflammatory genes (28). In the present study, we used a similar approach by using IL-6-deficient (/) mice to seek a role for IL-6 in suppressing inflammatory genes in the heart. We showed that LPS-induced mRNA levels for three inflammatory genes (IL-1, TNF-, and ICAM-1) were significantly higher in the hearts of IL-6-deficient (/) mice than those of non-IL-6-deficient (+/+) mice, suggesting that IL-6 functions to negatively regulate the expression of these three genes. Both IL-1 and TNF- cause myocardial depression (19) and also induce ICAM-1 gene expression (25). ICAM-1 promotes neutrophil-myocyte adhesion that may cause cardiac cell injury and/or necrosis due to the cytotoxic activity of the neutrophils (30). Therefore, IL-6 may play a cardioprotective role during endotoxemia by downregulating ICAM-1 expression. Furthermore, IL-6 is likely to self-regulate its expression by suppressing its own inducers, IL-1 and TNF-.

It has been reported that ICAM-1 is induced by IL-1, TNF- (25), or IL-6 (30) in cultured myocytes. It was also reported that peak IL-6 mRNA induction precedes that of ICAM-1 mRNA in a canine model of myocardial ischemia and reperfusion, supporting the idea that IL-6 is important in the induction of ICAM-1 in the area of ischemia (18). However, our time course study showed that LPS-mediated induction of ICAM-1 mRNA occurs later than that of IL-1 and TNF- mRNA and slightly earlier than that of IL-6 mRNA (Fig. 2). These results suggest that the major inducers of cardiac ICAM-1 during endotoxemia are IL-1 and/or TNF- rather than IL-6. Furthermore, our studies using IL-6-deficient (/) mice demonstrated that the cardiac ICAM-1 mRNA is induced more strongly in IL-6-deficient (/) mice than in non-IL-6-deficient (+/+) mice, suggesting that IL-6 is not an essential inducer of ICAM-1 but rather a suppressor during endotoxemia (Fig. 6). Taken together, our data suggest that the cardiac inflammation cascade during endotoxemia is not the same as that in ischemia-reperfusion.

The molecular mechanisms for downregulation of the three inflammatory genes (IL-1, TNF-, and ICAM-1) by IL-6 remain unclear. IL-6 can activate the transcription factors STAT-3 or C/EBP- (or NF/IL-6) via the gp130/JAK-STAT or Ras-MAPK-NF/IL-6 pathway, respectively. Because a NF/IL-6 (or C/EBP-) binding site is present in the ICAM-1 gene regulatory region (13), C/EBP- may be activated by IL-6 and, in turn, downregulate the ICAM-1 gene. Alternatively, IL-6 may indirectly downregulate ICAM-1 by suppressing IL-1 and TNF-, which are the direct inducers of ICAM-1 (25).

Although the C/EBP- gene is reportedly induced by IL-6 (21), there was no decrease in the induced C/EBP- mRNA levels in the hearts of IL-6-deficient (/) mice after LPS administration (Fig. 6). This result suggests that IL-6 deficiency is ameliorated by other cytokines for induction of the C/EBP- gene. Whereas a number of hepatic acute phase response genes are known to be induced by endotoxin challenge and by IL-6 (1, 2, 11), little is known about gene expression in response to endotoxin or IL-6 in the heart. Further studies screening expression of multiple genes are necessary to further understand the effects of endotoxin or IL-6 in the heart.

In conclusion, IL-6 is strongly induced in the heart during endotoxic stress, and nonmyocardial cells are the primary source of the cardiac IL-6. The IL-6 expression is induced mainly by IL-1, which is also induced in the heart earlier during endotoxic shock. The endotoxin-induced IL-6 is likely to downregulate expression of IL-1, TNF-, and ICAM-1 in the heart, suggesting a self-regulatory mechanism and a possible anti-inflammatory function for IL-6.