Sickle cell disease (SCD) is a chronic, devastating family of closely related disease conditions affecting nearly 100,000 people in the USA and an estimated 20 million people worldwide [1]. The most severe form of SCD is homozygous sickle cell anemia (SCA), the most common SCD condition. Each year, approximately 300,000 infants are born with SCA, a number that could reach 500,000 by 2050 [2]. Interestingly, childhood mortality from SCD in high-income countries is similar to the mortality rates in the general population, and even though the median survival of adults with SCD is approaching more than 60 years [2,3], this notable achievement is still 16–17 years shorter than the average life expectancy at birth for all people in the USA [4]. Note that for the sake of simplicity in this manuscript we will use SCD when discussing the vasculopathy and chronic organ disease that develops in Townes sickle cell mice, the humanized strain that most closely replicates human SCA [5]. In current literature, SCD and SCA appear to be used interchangeably.

Sundd et al. have proposed that the pathobiology of SCD is mediated by a vicious cycle composed of four interconnected processes: 1) hemoglobin (Hb) polymerization; 2) altered red blood cell rheology and adhesion-mediated vaso-occlusion; 3) hemolysis-mediated endothelial dysfunction; and 4) concerted activation of sterile inflammation [6]. Vasculopathy is often considered to precede chronic organ injury [7]. Therapeutic targeting of the mechanisms mediating these processes underpins current SCD drug development.

Hemolysis and the resulting release of cell-free Hb (cf-Hb) in SCD have long been considered one of the primary mechanisms for impairing vascular function by preventing vasodilation, enhancing vascular adhesion, and increasing vasculopathy. Sickled RBCs are more rigid and oxidized than biconcave RBCs, making them more prone to being trapped in microvessels [7]. These blockages not only impede microvascular blood flow but also increase the risk of intravascular hemolysis [7]. Mechanistically, cf-Hb undergoes dioxygenation reactions resulting in superoxide anion (O2•–) generation. Superoxide anion scavenges nitric oxide (·NO) at the rate of diffusion to form peroxynitrite (ONOO−) [8], a potent oxidant that can induce severe cellular injury and damage [9]. These biochemical and radical reactions are considered foundational to the mechanisms by which cf-Hb increases oxidative stress and how SCD increases ·NO consumption, and resistance to ·NO donors and reduces ·NO bioavailability [10,11].

Despite many reports describing how cf-Hb initiates and propagates SCD pathophysiology, some have argued that hemolysis does not account for all the observed complications in SCD. In his critical review, Hebbel offered the opinion that although cf-Hb “probably does consume NO in SCD …. it is very unlikely to cause endothelial dysfunction” [12]. To test this possibility, we designed hE-HB-B10, a small peptide that binds and reduces cf-Hb in the circulation via liver uptake [13]. Our vasodilation studies, which focused on endothelial function in isolated facialis arteries showed that reducing cf-Hb in the circulation in SCD mice to levels approaching controls had little to no effect on endothelial-dependent vasodilation but did switch the mechanism of vasodilation from non-endothelial nitric oxide synthase (eNOS) -dependent to eNOS-dependent in the facialis arteries isolated from SCD mice [14]. In addition, reducing cf-Hb in the circulation of SCD mice reduced liver injury based on significant reductions in plasma alanine transaminase (ALT) levels [14]. These studies suggest that cf-Hb has little to no effect on endothelial-dependent vasodilation, a minimal inhibitory effect on eNOS-dependent vasodilation and a modest effect on liver injury in SCD mice. More importantly, if effectively reducing cf-Hb did not significantly improve endothelial-dependent vasodilation, then other pathways and mechanistic inhibitors of vascular function must exist in SCD mice besides cf-Hb.

Previously, we investigated the role of myeloperoxidase (MPO) in the impairment of vascular function and chronic organ injury in SCD. Using KYC, our studies showed that MPO generates toxic oxidants (HOCl, NO2) that damaged the liver and lungs of SCD mice and that KYC was a novel tripeptide inhibitor of MPO-dependent toxic oxidants that improves vasodilation and reduces biomarkers of leukocyte diapedesis in SS mice [15]. As neutrophil-derived MPO activity increases cell death, and dead and dying cells passively release high mobility group box-1 (HMGB1) [16], we consider MPO something more than a simple biomarker of neutrophil recruitment. HMGB1 is a potent damage-associated molecular pattern (DAMP) molecule that binds and induces Toll-like receptor 4 (TLR4)-dependent inflammation [17], increases vascular permeability [18] and recruitment of neutrophils and mononuclear cells [18], all the while impairing vasodilation [[19], [20], [21]] and increasing apoptotic cell death [22]. Recently, Vogel et al. [23] reported that HMGB1 played a causal role in inflammation and thrombosis in SCD. Their studies showed that HMGB1 activates platelet NLRP3 inflammasomes, which increases platelet Il-1Beta expression to increase inflammation in SCD. MPO and HMGB1 are both components in sterile inflammation, where MPO initiates oxidative vascular and organ injury and HMGB1 propagates the cycle by recruiting innate immune cells to sites of injury [24].

Here, we continue our investigations into how SCD induces vasculopathy and chronic lung disease by examining HMGB1's role in SCD pathogenesis. Our working hypothesis is that SCD increases vascular and pulmonary inflammation by an HMGB1-dependent mechanism that impairs vascular function and induces chronic lung injury. Although vasculopathy in SCD has been characterized as a state of ·NO resistance, accelerated ·NO consumption and decreased ·NO bioavailability mediated by primarily hemolysis and cf-Hb [25], S-nitrosoglutathione (GSNO) reductase (GSNOR) has been reported to degrade GSNO to protect against nitrosative stress as well as modulate intracellular GSNO levels to regulate protein S-nitrosothiols (SNOs) and ·NO signaling [26]. Protein SNOs are essential for vasodilation, regulating ·NO balance, reducing pulmonary inflammation and airway hyperresponsiveness [27]. Given previous reports linking HMGB1 to impaired vasodilation [[19], [20], [21]], a similar vascular response should be anticipated in SCD. Our findings, however, reveal a new aspect of HMGB1 signaling where HMGB1 increases GSNOR that in turn decreases protein SNOs and ·NO bioavailability in SCD.