Heat shock proteins in inflammation and asthma: Dr Jekyll or Mr Hyde?

The so-called heat shock or stress protein families (HSP) are induced in all cells and organisms upon threatening alterations of the cellular environment. The inflammatory process can be considered as a paradigm for cellular environmental stress, though only some specific types of inflammation have been associated with HSP overexpression, including arthritis and severe asthma. In this editorial, we will solely consider the 70 kDa HSP family, and in particular its cytosolic, inducible, 72 kDa member (hereafter named hsp70), and neglect all other HSP. For these latter, the interested reader should refer to more exhaustive reviews in the field [ 1, 2]. The major function usually ascribed to hsp70 is a protective one, insofar as it is essential for the ability of cells to adapt to and survive environmental stresses. In particular, HSP protect cells and tissues from the deleterious effects of numerous mediators of inflammation, including the reactive oxygen species or tumour necrosis factor (TNF)α [reviewed in 3]. In vivo, induction of a heat shock response also protects against pulmonary inflammation [ 4, 5]. It should, however, be mentioned that the anti-inflammatory effects of the stress response might relate more to the inhibition of activation of the transcription factor NFκB — NFκB being central to the inflammatory process — than to the HSP themselves [ 6]. HSP can play multiple other roles in various steps of the inflammatory response. Firstly, the gene coding hsp70 maps within the major histocompatibility complex, and hsp70 gene polymorphism, a likely contributor to disease susceptibility or stress resistance [ 7], is being extensively studied for associations with immune/inflammatory diseases, including asthma [ 7]. Secondly, several HSP can act as antigen, either self or non-self, and thus initiate or potentiate an immune response. Thirdly, hsp70 belongs to the molecular chaperones, i.e. it binds polypeptides, which has been associated with a possible role in antigen processing and presentation. In that case, hsp70 might enhance the immune/allergic response, which might be either protective (in the case of protective immune responses) or deleterious (in the case of allergy or autoimmunity). Finally, hsp70 has also been reported to prevent cell death by apoptosis [ 8, 9]. Although at first sight this antiapoptotic effect might appear protective, it actually could promote persistence rather than resolution of inflammation, by preventing the physiological removal of inflammatory cells by apoptosis, an essential component of the healing process. Thus, two essential questions on the role of hsp70 in inflammation remain as yet unsolved: is hsp70 overexpressed in inflammation, e.g. asthma, as a marker of increased stress, or does it represent an autoprotective mechanism?, and does hsp70 play a protective, an amplificatory, or a dual role in inflammation? The work by Bertorelli et al. in this issue [ 10] contributes several interesting points to these issues. These authors report that the expression of both hsp70 and HLA-DR was increased in pulmonary antigen presenting cells of subjects with asthma as compared to controls, that their expression significantly correlated with each other's, and that inhaled steroids, along with improving the clinical symptoms of asthma, decreased both hsp70 and HLA-DR [ 10]. Logically, if hsp70 expression decreases as the disease improves, it becomes difficult to ascribe it a protective role, while the possibility to use it as an adequate marker for disease severity increases. That hsp70 could be a specific marker for severity of asthma is further supported by the fact that hsp70 overexpression is not found in mild asthma [ 11], and that in severe asthma, it correlates with the number of eosinophils recovered by bronchoalveolar lavage in the same patients [ 12]. Furthermore, similar findings were made in another disease associated with inflammation and necrosis, though a genetic muscle disease, i.e. Duchenne's muscular dystrophy (DMD). Indeed, we found overexpression of hsp70 in the muscle of these patients [ 13]. In the animal model for the disease (mdx mice), novel therapeutic approaches leading to a decrease in muscle necrosis were also associated with a decrease in hsp70 expression [ 14]. If one is to accept, against most published evidence, that overexpression of hsp70, in contrast to other diseases, does not subserve protective functions in inflammatory diseases, does this then mean that hsp70 plays amplificatory roles in asthma and other inflammatory diseases? One way to do so would be by amplifying the initiation of the inflammatory process, that is, the immune/allergic response, and the other, to prevent its resolution. Far from us the idea to shed any doubt on the results of the many groups that have used hsp70 overexpressing transgenic mice to prove its beneficial effects in myocardial infarction for example [ 15, 16] — our concerns todate restrict to inflammation. In asthma for example, according to Bertorelli's work, one could assume that hsp70 overexpression, together with an increased expression of HLA-DR, contributes to antigen/allergen processing and/or presentation, and thus, to the longlasting activation of an immune response [ 10]. Recent work from our laboratory suggests that hsp70 plays indeed a subsidiary role in antigen processing, merely in stressed cells (Jacquier-Sarlin & Polla, submitted). Such a role in antigen processing is compatible with the intracellular localization of hsp70. But in order to contribute to antigen presentation, hsp70 should somehow get expressed at the cell membrane of antigen-presenting cells. Although members of the hsp70 family are indeed found on the plasma membrane of certain cell types, the mechanism(s) underlying transport and anchorage to the plasma membrane has not been identified yet; productive interactions between members of the hsp70 family and other proteins are emerging as an unifying concept for hsp70 membrane anchorage [ 17]. The study by Bertorelli et al. did not investigate hsp70 subcellular localization, but further studies will undoubtedly follow to solve this issue and the many others that remain unanswered. If apoptosis of inflammatory cells is indeed a prerequisite for the scarless healing of an inflammatory process, then hsp70 could prevent resolution of severe asthma by inhibiting apoptosis. Interestingly, hsp70 expression is specific for asthma, and not found in chronic bronchitis, while the figure is opposite for apoptosis (decreased in asthma, present in chronic bronchitis). The use of corticosteroids could be particularly informative in providing answers to the question as to whether agents that lead to clinical improvement, and decrease hsp70 overexpression in the airways of asthmatic patients, also increase inflammatory cell apoptosis. Indeed, we have been missing drugs that modulate hsp70 expression to better address their role in human disease. Besides corticosteroids, non-toxic chaperone (co)-inducers such as Bioclomol® (Biorex, Hungary; [ 18]) should tell us more about the precise roles of HSP in disease, whether protective or deleterious. The possibility to test these or similar drugs — though keeping in mind the potential side effects — might provide further insight into the role of HSP in inflammation in general and asthma in particular. In conclusion, although HSP generally exert protective/antiinflammatory functions, they might also have opposing effects, i.e. contribute to the amplification and the chronicity of the inflammatory process. We can anticipate the work by Bertorelli et al. to promote many other studies in the field, in order to solve the dilemma: Dr Jekyll or Mr Hyde? As usual, the answer is unlikely to be simple, and disease specificity will have to be considered.