Sensores Electromiográficos

High mobility group box 1 (HMGB1, previously named HMG1 or amphoterin) is a 25kDa nuclear DNA-binding protein that is highly conserved across species and widely distributed in all mammalian tissues. Structurally, it is composed of three domains: two homologous DNA binding motifs named A box and B box, and a negatively charged C-terminus [362]. HMGB1 may exist in its native form or after a variety of post-translational modifications. Nuclear HMGB1 may be acetylated in response to inflammatory cell activation resulting in HMGB1 passing into the cytosol, prior to packaging into lysozymes and being secreted [363]. HMGB1 can exist in oxidised or reduced states (via an intra-molecular disulphide bond) which may affect extracellular activity [364]. Indeed, all HMG proteins are subjected to extensive post-translational modifications [365] and the extent to which this affects extra- cellular activity remains to be fully elucidated.

In the nucleus, HMGB1 binds DNA in a non-specific manner and acts primarily by manipulating nucleoprotein complexes to regulate gene transcription and to fluidize chromatin by loosening the wrapped DNA-core histone complex [365]. HMGB1 genetically deleted mice die shortly after birth from hypoglycaemia, supporting the functional

82 significance of the role of HMGB1 in the transcriptional enhancement of the glucocorticoid receptor [366].

In a large cohort of healthy humans the mean level of circulating HMGB1 was 1.69±0.04 ng/ml in males and 1.62±0.04 ng/ml in females; sRAGE levels being independently inversely related, and CRP and WBC being (independently) positively related [367]. Anti-HMGB1 antibodies have been found in the plasma of patients with auto-immune disease and in more than one third of patients with septic shock [368, 369]; it is uncertain how this affects the accuracy of immunoassays.

The role of extracellular HMGB1 is contentious; a large body of work suggesting it fulfils a cytokine-like role, eliciting pro-inflammatory responses. By contrast, there are data suggesting that HMGB1 may have only weak or absent inflammatory activity per se and is more likely to function as a molecular chaperone, augmenting the inflammatory responses to other mediators.

3.2.2.1. The case for HMGB1 being a pro-inflammatory mediator

3.2.2.1.1. Interventional models

Initially, murine experiments revealed that HMGB1 levels were elevated during sepsis and administration of HMGB1 was lethal, and that antibodies to HMGB1 were protective against lethal endotoxaemia. Furthermore, they demonstrated that endotoxin caused release of HMGB1 from macrophages and levels of HMGB1 were elevated in patients with sepsis [370]. These elevations occurred late and persisted far beyond those of the classical cytokines; thus HMGB1 was considered to have a larger potential ‘therapeutic-window’ for intervention and was described as a late-mediator of sepsis. An inhibitor of HMGB1 release, ethyl pyruvate (EP) was shown to have efficacy in reducing mortality in murine endotoxaemia and polymicrobial sepsis, even when administered 24h after the onset of injury[371]. Later, other inhibitors (anti-HMGB1 antibodies or DNA-binding A box) were also shown to be similarly effective in murine models of sepsis even when given late [372]. Indeed, a large number of HMGB1-inhibiting reagents have been evaluated in murine models of sepsis, (reviewed by Wang et al [373]). In addition to models of sepsis, inhibition of HMGB1 was found to be efficacious in experimental acute lung injury, pancreatitis, haemorrhagic shock and polytrauma [374-378]. HMGB1-inhibition is particularly attractive

83 as a potential treatment in humans as the therapeutic window is wide and the experimental interventions were effective when started after the onset of systemic inflammation; a situation akin to the use of antibiotics in patients presenting with sepsis.

3.2.2.1.2. In vitro data

Inflammatory stimuli cause monocytes and macrophages to secrete HMGB1 in a non- classical, leaderless manner [379], an effect that can be attenuated by immunomodulatory drugs such as corticosteroids [380]. In response to extra-cellular HMGB1, monocytes show altered chemotaxis (spreading), transendothelial migration [381] and release of IL-6, IL-1β and TNF-α [382, 383]. HMGB1 also mediates neutrophil adhesion and transmigration across endothelium as well as inducing pro-inflammatory cytokine expression [374, 384, 385]. The actions of HMGB1 have been shown to be mediated through interaction with RAGE, TLR2, TLR4, and TLR9 [386-389].

3.2.2.1.3. Patients

Levels of HMGB1 have been found to be elevated, and in some cases related to clinical outcome, in the plasma of patients with acute inflammatory conditions including pneumonia, sepsis and septic shock, haemorrhagic shock, disseminated intravascular coagulation, falciparum malaria, burns, trauma and severe acute pancreatitis[390-401]. Soon after surgery, monocytes exhibit increased HMGB1 expression. Later this is diminished, concurrent with elevation in plasma HMGB1 levels, thereby implicating monocytes as a source of plasma HMGB1 [383].

3.2.2.2. The case against HMGB1 being a pro-inflammatory mediator

The form of HMGB1 used in early experiments was recombinant and sourced from E. Coli. In 2007, a pivotal advance suggested that when eukaryotic HMGB1 was appropriately purified, it lacked the observed cytokine-like activities [402, 403]. In fact, lesser inflammatory effects of native, compared to recombinant, HMGB1 had previously been reported [404]. Affinity chromatography revealed that bacterially-derived material, including lipids, were tightly bound to HMGB1 and contributed to the observed activity [403]. Thus, in addition to the effect of many possible post-translational modifications to HMGB1, there remains the important effect of smaller molecules bound to HMGB1 including LPS, IL-1β, IFN-γ, TNF-α and fragments of DNA [388, 405, 406]. These findings have altered the view of the scientific

84 community of HMGB1 which is now generally considered to act as a chaperone protein, in molecular collaboration with other immuno-stimulatory factors [407, 408].

A compelling reason to doubt that inhibition of HMGB1 will offer therapeutic benefit in patients is that in a large cohort of patients with severe community acquired pneumonia, HMGB1 levels were persistently elevated at hospital discharge [409, 410]. It remains possible that anti-HMGB1 therapies will retain efficacy through diminishing the adjuvant role of HMGB1 in potentiating inflammatory responses.