摘要
See article by Y. Chen and R.W. Currie[1](pages 66–75) in this issue. Investigators are beginning to unravel the intricate relationship between NFκB, heat shock proteins (HSPs), and heat shock factor (HSF)-1. New information is added to this complex picture with the novel finding that HSF-1 modulates AP-1 activation, adding yet another layer of complexity [1]. HSF-1, a transcription factor for the heat shock proteins, is a key component of the heat shock response. The activation of HSF-1 drives the increase in HSPs in response to injury or stress. NFκB, a critical regulatory transcription factor, has both pro-inflammatory and anti-apoptotic effects. Activation of the heat shock response yields activation of HSF-1 and increased transcription of the heat shock proteins. The heat shock proteins have an array of protective properties, including refolding denatured proteins, protecting folded proteins and targeting irreversibly denatured proteins for removal. However, it is increasingly evident that there are other mechanisms by which the heat shock response reduces inflammation and injury. HSF-1 knockout is associated with a chronic increase in tumor necrosis factor (TNF)-α levels and increased susceptibility to endotoxin [2]. Consistent with this, the promoter for TNF-α has been found to contain an HSF-1 binding site that represses transcription, and thus loss of this repressor results in sustained expression of TNF-α [3]. However, there is a reverse action, as well: TNF-α transiently inhibits activation of HSF-1 via TNF-R1 and activation of phosphatases [4]. Phosphorylation is key to both activation and inhibition of HSF-1, and dephosphorylation, in this case mediated by TNF-α, can lead to inactivation. In this issue of Cardiovascular Research, Chen and Currie [1] demonstrate that reduction of HSF-1 with iRNA in a vascular smooth muscle cell line amplified the inflammatory response to angiotensin II [1]. Activation of both NFκB and AP-1 was significantly higher with reduction in HSF-1. HSF-1 has also been reported to block transcription of IL-1β by binding its transcription factor and to block expression of cfos and cfms[5,6,2], which may explain the increased activation of AP-1 with loss of HSF-1. In a mouse model of toxin inhalation, the HSF−/− mice had greater activation of NFκB, particularly p50, and this was mediated, at least in part, by increased degradation of IKB, suggesting at least one mechanism for increased activity of NFκB in the absence of HSF-1 [7]. Thus, independent of activation of HSP expression, HSF-1 inhibits transcription of a number of pro-inflammatory genes and modulates the activation of both NFκB and AP-1. These complex interactions are simplified in the diagram shown in Fig. 1. Diagram summarizes key interactions amongst HSF-1, heat shock proteins (HSPs), NFκB, AP-1 and TNF-α. HSF-1 inhibits both NFκB and the expression of TNF-α. TNF-α can both activate HSF-1 and inhibit HSF-1. NFκB can activate HSF-1 and increase expression of HSP72. HSF-1 can inhibit AP-1 activation. Arrow indicates activation. Dashed line stands for a weaker effect. “X” at the end of a line indicates inhibition. HSPs themselves have direct anti-inflammatory and protective effects. Key aspects include protection of protein synthesis, membrane integrity, and mitochondrial function after injury as well as inhibition of apoptosis [8,9]. HSP70 and HSP27 have both been found to decrease NFκB activation [10]. Knockout of HSF-1 is associated with loss of both basal expression and stress-mediated induction of HSP72 and HSP27 in fibroblasts. In the heart, knockout of HSF-1 has no effect on basal levels of HSP72 and HSP27, but results in loss of the stress-induced increase in these proteins [2]. Paradoxically, in cardiac myocytes, TNF-α actually induces expression of HSP72 without evidence of injury [11]. TNF-α may mediate the increase in HSP72 by increasing the binding of HSF-1 [12]. We have found that TNF-α activates HSF-1 in cardiac myocytes (Chang and Knowlton, unpublished). In both endothelial cells and adult cardiac myocytes, activation of NFκB was essential for induction of HSP72 expression by estradiol [13,14]. Inhibition of NFκB activation with DNA binding decoys blocked the protective effects of estradiol treatment. Similarly, NFκB suppresses TNF-α-induced apoptosis. Knockout of IKK-β, key for activation of NFκB, blocked systemic inflammation in a model of intestinal injury, but was associated with severe apoptosis of the mucosa [15]. Thus, TNF-α and NFκB exhibit protective, anti-inflammatory, anti-apoptotic properties as well as inflammatory and pro-apoptotic properties. Clearly, there is a complex interaction among HSF-1, heat shock proteins, NFκB, and TNF-α. This interaction may produce fine tuning of the inflammatory response. Much remains to be understood about the interactions among these proteins [2]. The work of Chen and Currie adds new information to our understanding of this interrelationship and shows, for the first time, that AP-1 activation is also modulated by HSF-1 [1].