High Mobility Group Box-1 (HMGb1): Current Wisdom and

Subscriber access provided by RMIT University Library

Perspective

High Mobility Group Box-1 (HMGb1) – Current Wisdom, and Advancement as a Potential Drug Target Sonya VanPatten, and Yousef Alabed J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01136 • Publication Date (Web): 21 Dec 2017 Downloaded from http://pubs.acs.org on December 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 50

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Paragon Plus Environment

High Mobility Group Box-1 (HMGb1) -Current Wisdom, and Advancement as a Potential Drug Target

Sonya VanPatten†*, Yousef Al-Abed† †

Center for Molecular Innovation, The Feinstein Institute for Medical Research, 350 Community Drive,

Manhasset, NY 11030

Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 50

Abstract High mobility group box-1(HMGb1) protein, a nuclear non-histone protein that is released or secreted from the cell in response to damage or stress, is a sentinel for the immune system that plays a critical role in cell survival /death pathways. This perspective highlights key features of the endogenous danger-associated molecular pattern (DAMP) protein, HMGb1 in the innate inflammatory response along with various co-factors and receptors which regulate its downstream effects. The evidence demonstrating increased levels of HMGb1 in human inflammatory diseases and conditions is presented, along with a summary of current small molecule or peptide-like antagonists proven to specifically target HMGb1. Additionally, we delineate the measures needed towards validating this protein as a clinically relevant biomarker or bioindicator, and as a relevant drug target.

HMGb1 as an immune system signal (DAMP) Inflammation is a natural and necessary response to invading pathogens and trauma/tissue damage and is manifested by an organism’s immune system. Many aspects of the initiation of inflammation and its resolution have been revealed, yet these multifaceted reactions still hold many obscurities. Pathogen-associated molecular patterns (PAMPs), danger-associated molecular patterns (DAMPs)1, and homeostasis-altering molecular processes (HAMPs) 2 are signals the innate immune system uses to recognize these threats to the health of an organism. Upon detection of an inflammatory signal, whether it be DAMP/PAMP or HAMP related, or a proinflammatory mediator (cytokines/chemokines, or a modified protein (oxidized or glycated) or lipid), the immune system generally responds with a proportional feed-forward amplification and expansion of the response through pattern recognition receptors (PRRs) and other innate immune receptors, cytokine receptors, and/or inflammasome pathways. To turn off the inflammatory process, the sensing and switching of a proinflammatory response into the resolution phase, as well as activation of anti-inflammatory pathways, has begun to be elucidated, though much remains to be explored 3-4. The current understanding of the dangers of excessive acute inflammation and unresolved chronic inflammation is well described in the literature in numerous diseases and


Page 3 of 50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conditions. HMGb1 was first described as playing an extracellular role in inflammation in 1999 when it was discovered as a late mediator of sepsis 5, although its initial characterization was as an intranuclear non-histone protein which could directly bind DNA. This multi-functional protein is now known for its roles in the nucleus (transcription factor, transcription enhancer, nucleosome sliding, DNA repair, V(D)J recombination, telomere homeostasis )6, cytoplasm (autophagy 7, viral sensing 8), and extracellular milieu (DAMP secreted or released by cell damage) 9. HMGb1’s nuclear function is nearly ubiquitous and the protein in totality, is necessary for life as complete knockout of the gene results in lethality due to hypoglycemia after birth 10. Tissue specific conditional knockouts have been successfully generated however 11. Various inflammatory signals may initiate export of nuclear HMGb1 to the cytoplasm and culminate in extracellular secretion. Nuclear export has been shown to be regulated by oxidation, ADP-ribosylation, phosphorylation, secondary messengers (reactive oxygen species (ROS), Ca2+, nitric oxide (NO)), as well as by both acetylation, and methylation 12-14. The oxidative/reduced state of three critical cysteine residues (yellow highlight -Figure 1) has proven a major regulator of specific cellular death related pathways. The all-thiol form where all three cysteines are reduced, is associated with autophagy and/or HMGb1’s extracellular chemokine-like chemotactic activity (via CXCL12/CXCR4 and receptor for advanced glycation endproducts (RAGE)), while the disulfide-form, in which C23 and C45 form an intra-chain bond, relates to its extracellular proinflammatory cytokine-inducting activity via the extracellular RAGE or tolllike receptor 4 (TLR4). The fully oxidized form has thus far been associated with apoptosis 13. The HMGb1 protein contains two similar yet distinct DNA-binding domains, termed A-box and B-box, each containing three α helices that fold into an L or V shaped structure (Figure 1- 3D model created using Cn3D v.4.3.1 and PDB ID: 2YRQ). It also contains a negatively charged acidic tail which interacts with specific residues within and between the HMG boxes and thus regulates the 3D structure and DNA binding of the protein. These and additional biochemical and structural aspects of HMGb1 have been reviewed extensively elsewhere 12, 14. Initiation and early events in inflammation In the initial insult of trauma or pathogen-exposure, DAMPs, PAMPs, and HAMPs signal PRRs on platelets 15-16, neutrophils, macrophages/monocytes, and many other cell types17. Platelets have also been shown to actively


Page 4 of 50

export HMGb1 to the cell surface after activation 18, which leads to recruitment of other immune cells and amplification of the initial activation signal. Neutrophils are another early responder in inflammation (including sterile), and the generation of neutrophil extracellular traps (NETosis) initiates a series of events which has been reviewed extensively 19. Other extracellular factors such as acute phase proteins (fetuin A, C-reactive protein (CRP), serum amyloid A (SAA)20), pentraxin3 and complement factors 21, endogenous glucocorticoids 22, and cytokines/chemokines serve to coordinate and further regulate the initial inflammatory signal. Pattern recognition receptors including TLRs and RAGE bind HMGb1 and other PAMPs and DAMPs including DNA, dsRNA, LPS, flagellin, peptidoglycan, S100 proteins, heat shock proteins (HSPs), uric acid, ROS, advanced glycation endproduct (AGE)-modified proteins, and cell-adhesion molecules and may also signal and influence the overall inflammatory response. HMGb1 alone has been shown to propagate the inflammatory response by inducing the synthesis of CRP, TNFα, IL-6, and MIP1α and β in human monocytes 23. Moreover, HMGb1 is also capable of modulating the adaptive (acquired) immune response 24-25. Anti-inflammatory pathways and inflammation resolving factors In order to resolve an inflammatory response, damage and pathogens must be cleared and activated cells need to be removed or re-programmed to an inert, resting state. The sequential induction of resolvin-type mediators 26 in the normal immune response helps to achieve this endpoint along with activation of anti-inflammatory signals and pathways 3-4. A few of the main anti-inflammatory pathways which serve to antagonize the pro-inflammatory response include induction of numerous inhibitory miRNA’s, SOCS proteins 27, activation of the hypothalamicpituitary axis (HPA) (i.e. by cytokines) leading to adrenal corticosteroid production and glucocorticoid receptor activation-resulting in inhibition of pro-inflammatory gene expression (reviewed in 28), and stimulation of the cholinergic anti-inflammatory pathway (CAP) 29. Some of these anti-inflammatory responses are known to regulate HMGb1 levels directly (miRNA’s30, SOCS31, CAP32). Many insights into the molecular pathways of the anti-inflammation/ resolution phase have occurred over recent years, including the observation that adaptive immunity is achieved during the post-resolution phase. Non-resolving inflammation or “frustrated resolution”


Page 5 of 50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


occurs when inflammatory signals are not cleared or switched off and the resolution phase does not progress, hindering the development of adaptive immunity 33. Endogenous modulation of HMGb1 activity and signaling HMGb1, which normally resides in the nucleus, can be translocated to the cytoplasm in response to various stimuli which regulate its post-translational modifications including, acetylation 34, methylation 35, phosphorylation 36, and redox modification (reviewed in 37). The upstream signals which result in these posttranslational modifications (PTMs) may include infection or trauma (sterile injury) resulting in release of PAMPs/DAMPs, oxidative or osmotic stress, ATP excess or depletion (starvation), and/or ischemia/reperfusion (reviewed in 38). Additional levels of regulation may include pathways of degradation and cleavage 39, interaction with binding partners 40, or other receptors (CD24/Siglec41, syndecans42, Mac-143, TLR244). Recently two other proteins have been demonstrated to interact with extracellular HMGb1 and may serve as a form of regulation in specific conditions. Thrombomodulin, a regulator of coagulation, can bind HMGb1 and thereby reduce its inflammatory activity by sequestering it from binding other receptors (reviewed in 45). Additionally haptoglobin, an acute-phase protein which binds free hemoglobin to prevent oxidative damage, can also scavenge extracellular HMGb1thereby reducing inflammatory signals 46. The concept of intracellular molecular signaling hubs which integrate survival and death signals, leading to generation of the appropriate cellular response has been proposed 47. These signals, combined with the subsequent amounts of HMGb1 passively released (necrosis) or secreted have been shown to regulate whether necroptosis, pyroptosis, apoptosis, NETosis, or autophagy pathways are initiated in surrounding areas. One particular signal integration hub involves regulation of intracellular calcium levels 48. The cell’s ability to buffer calcium levels, via modulation of ion channel opening, and/or synthesis/release of Ca2+ binding proteins serves a critical role in signal propagation, cellular functions and cell fate (survival vs. death49). Calcium signaling is crucially involved in a myriad of cellular processes including survival/repair, proliferation, differentiation, death pathways, normal metabolism, gene transcription, signaling pathways (protein phosphorylation), stimulus-induced protein secretion, neurotransmission and motility. HMGb1 can itself be regulated by calcium signaling 36, 50-51 and


Page 6 of 50

also induce calcium-related signaling through different receptors and downstream pathways (TLR’s 52-55, RAGE 56 , CXCL12/CXCR4 57, transient receptor potential (TRP) channels 58). The precise involvement of calcium signaling pathways (interrogated via chelating agents, or specific calcium channel blockers) in different cell types and environments, and in response to different HMGb1 levels, still requires some elucidation, but appears to be highly context dependent.

HMGb1 and extracellular HMGb1 receptors as drug targets The relevance of HMGb1 as a significant drug target and potential biomarker has been demonstrated by measuring its increase extracellularly in plasma/serum and tissue in human diseases and pathological conditions. These include neurodegenerative disease, arthritis, autoimmune diseases, lung injury and disease, cancer, cardiac and vessel disease, diabetes (type 1&2), trauma/ischemia-reperfusion injury, infection, kidney and liver-related disease, pancreatitis, and pregnancy complications (Table 1). HMGb1-directed therapies, focusing on those that directly bind and neutralize extracellular HMGb1 (including antibodies), have been tested in preclinical models of the above-listed, as well as other diseases and conditions. The myriad of potential benefits of HMGb1-specific drugs and therapies is well supported in the scientific literature 59-63. We provide a summary of directly-binding, peptide-based and small-molecule, HMGb1-targeted therapeutics in Table 2. Some of these are approved drugs which were found to have off-target effects by directly binding HMGb1, while others have only been tested preclinically. The structures of these directly-binding HMGb1 inhibitors are illustrated in Figure 2. Going forward, HMGb1-directed therapies with minimal off-targets should be used to reinforce the value of HMGb1 as a specific and effective drug target. An additional point, HMGb1 family member, HMGb2, is highly similar in structure/sequence and its expression and contribution to inflammation and disease is beginning to be reported 64-67. Equivalent PTM’s and receptor binding for HMGb2 can be assumed from sequence similarity (http://www.uniprot.org/uniprot/p26583). The specificity of HMGb1directed small molecules and peptides for HMGb2 is scant 68, and could possibly be exploited for selective targeting as more is learned.


Page 7 of 50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The first directly binding HMGb1 antagonist to be described in the literature was the glycoconjugated triterpene glycyrrhizin, a compound which occurs naturally in the licorice plant. It was found using NMR chemical shift data to bind to both the A and B boxes with Kd’s of 103+/-25µM and 87+/- 35µM respectively. This binding was also found to be independent and non-cooperative and involved residues F17, Q20, R23, E25, K43 and C44 on the A box and residues R109, I112, D123, and A125 on the B box. Glycyrrhizin inhibited the HMGb1-induced chemotactic and mitogenic activity in the 3T3 fibroblast cell line with an IC-50 of approximately 50µM. Additional studies showed only a mild effect of the compound on DNA/HMGb1 binding. The binding region on HMGb1 is described as occurring at the crux of the L-fold which has an exposed hydrophobic surface amenable to van der Waals interactions with glycyrrhizin. The structurally related compound carbenoxolone (glucuronic acid substituted with succinyl moiety) was found to bind A box in a manner similar to glycyrrhizin but data for B box was less clear and showed evidence of aggregation at higher concentrations69. Heparin, an endogenous sulfated glycosoaminoglycan anti-coagulant (3Kda to 30Kda) which has been adapted for clinical use (approx MW 1.5Kda), was shown to inhibit HMGb1-induced TNF and IL-6 secretion using an in vitro assay with RAW264.7 macrophages and HUVEC cells in 2011. A direct heparin-HMGb1 interaction was demonstrated with surface plasmon resonance (SPR) experiments, circular dichromism (CD) spectroscopy, and binding studies monitoring tryptophan and tyrosine florescence. SPR gave an estimated KD of 4.5nM. HMGb1 amino acid residues 6-12, which contain a consensus sequence similar to other heparin-binding proteins, was responsible for heparin binding and heparin successfully competed with RAGE binding to HMGb170. The heparin binding ability of HMGb1 (previously termed p30/amphoterin) was first realized back in 198771 before any of its extracellular proinflammatory roles were realized, when it was fractionated out with a heparinsepharose column in a screening for neurite outgrowth promoting factors. In 2013, the autoimmune disease and cancer drug methotrexate (Mtx), a biguanidine-folic acid-like derivative known for inhibition of dihydrofolate reductase , deoxycytidine kinase, and nuclear factor kappa-light-chainenhancer of activated B cells (NFκB) was shown to also interact with and inhibit HMGb1. Binding between HMGb1 and Mtx was observed using SPR analysis and electromobility shift assays (EMSA), while docking


Page 8 of 50

studies provided further evidence for a direct interaction. Two distinct non-cooperative binding sites in the Al (187aa) and Bj(88-181aa) box regions were measured with Kd’s of 0.5+/-0.3µM and 0.24+/-0.01µM respectively. Functionally, Mtx inhibited HMGb1-induced TNFα secretion and mitogenic activity and this was likely through interaction with the RAGE receptor as Mtx did not inhibit LPS-induced TNFα release (TLR4 dependent). Furthermore, Mtx was not found to inhibit HMGb1 binding to DNA72 . The inflachromene series of HMGb1 antagonists was first reported in 2014. These pentacyclic triazolidine-dionebased class of compounds were found in a cell-based phenotypic screening for neuroinflammatory inhibitors in an LPS-induced microglial cell stimulation assay. Inflachromene (ICM-1d) suppressed LPS-induced nitrite release as well as cytokine production in microglial cell lines, primary microglia, and macrophage cell lines. To identify the cellular targets, a photocrosslinking analogue of ICM was created using structure activity relationship (SAR) studies and used to pull out HMGb2 (among others). Additional studies using an ICM-linked column proved HMGb2 could be successfully competed out with non-bound ICM, demonstrating its specificity. Based on structure-sequence homology, HMGb1 was speculated to also bind ICM. This was shown functionally by using siRNA knockdown of either or both HMGb1&2 to prove that ICM lost inhibitory ability on LPS-induced nitrite release after HMGb1/2 knockdown. It was also discovered that ICM directly inhibited post-translational modifications on HMGb1&2 responsible for cytosolic and extracellular translocation. Compared to glycyrrhizin, ICM had 600x more potency on microglia and docking studies predicted binding at the DNA binding domain in Box A adjacent to a nuclear localization sequence (NLS) where PTM’s occur

68

. In 2016, this group expanded

their SAR studies and identified two additional ICM analogues (2j and 2l) with improved solubility and microsomal stability, tighter functional binding and lower energies (modeling). In the cecal ligation and puncture (CLP)-sepsis model the new ICM’s (2j, 2l) showed 74 and 56% survival rates compared with 45% for the original ICM (1d) compared with control values (vehicle) of 10% 73. Salicylic acid (SA) (the endogenously deacetylated product of aspirin (acetylsalicylic acid)), and both synthetic (3-aminoethyl (AE)-SA) and natural (amorfrutin-B1) analogues of SA, were discovered to bind HMGb1 in 2014. This added to the already known anti-inflammatory actions of aspirin through COX1/2 inhibition. The interaction


Page 9 of 50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


was found using an SA-crosslinked column affinity pulldown/competition analysis and was also measured using NMR and SPR. Although Kd’s for SA were estimated in the low mM range, the SA analogues- 3-AESA, and amorfrutin B1- had Kd’s in the nM range. These authors found that SA and the select analogues had the same separate binding sites in both the A- and B-boxes of HMGb1. Additionally, SA and analogues could inhibit the chemoattractant activity of fully-reduced HMGb1 (via CXCL12/CXCR4), as well as the cytokine-inducing effect of disulfide HMGb1 (via MD2/TLR4). This biological inhibition was deemed relevant based on estimated plasma SA concentrations after standard dose aspirin administration, while the analogues (3-AESA, amorfrutinB1) were observed to be orders of magnitude more potent. 74 The popular green tea component, epigallocatechin-3- gallate (EGCG) had previously been shown to inhibit the LPS-induced release of HMGb1 from macrophage/monocytes, to reduce plasma HMGb1 levels in a model of sepsis75, and to stimulate autophagy after induction of aggregation of the protein76, yet the mechanisms remained to be fully elucidated. The reducing agent DTT could disrupt this interaction, providing evidence that cysteine residue sulfide bonds were involved in the binding to EGCG. A computational modeling study in 2016, elaborated on the EGCG interaction using molecular dynamic simulations and potential mean force calculations to project details of the HMGb1-induced aggregation. Modeling with EGCG predicted induction of a large conformational change in the protein which caused burying of certain hydrophobic residues (including Cys106) involved with the catechin ring of EGCG and surface exposure of more charged hydrophilic residues to solution. The molecular polarity of the HMGb1 with EGCG was estimated to be twice that of protein alone; thus favorable electrostatic interactions were implicated in the observed aggregation77. As EGCG has other known targets and is already broadly commercially available as a dietary supplement, its development and use as a clinically relevant HMGb1 inhibitor is unlikely. Another clinically approved biguanidine-containing drug –metformin- used already in metabolic disease (prediabetes) and type 2 diabetes, was discovered to bind HMGb1 in 2017. The exact cellular mechanisms by which it lowers hepatic blood glucose production and restores insulin sensitivity are not fully elucidated, but effects on the mitochondrial complex 1/gapdh/AMPK pathway, and NfκB have been reported. HMGb1 was discovered as


Page 10 of 50

an additional metformin target while searching for binding proteins using a biotinylated-metformin, avidin-linked column pull-down assay. This interaction was shown to occur through the acidic C-terminal tail region as HMGb1 constructs lacking the tail region (HMGb1-∆AT) could still stimulate downstream HMGb1-induced readouts including p38 phosphorylation and TNFα production, however administration of metformin with HMGb1-∆AT no longer had any inhibitory effects on those readouts. These authors also established that metformin did not influence the redox state of HMGb1 and demonstrated that binding was non-covalent in nature using mass spectrometry. 78 Two of the most prevalent and well-studied HMGb1 extracellular receptors are TLR-4 and RAGE. These receptors likely work in conjunction, and also in accordance with other factors (MD-2 79, CD-14 80 LPS81), to propagate downstream inflammatory signals and effects 82. HMGb1 has also been shown to interact with CXCL12 and mediate chemotactic responses through the CXCR4 receptor. Thus far, numerous studies have confirmed the importance of HMGb1 receptor antagonists in diverse disease models. A more general review of smallmolecule TLR4 and RAGE antagonists is described in references 83-84. Furthermore, antibody-based antagonists against TLRs, and also RAGE, show efficacy in models including neurodegeneration (AD85), liver injury, kidney injury, ischemia/reperfusion injury (trauma), multiple sclerosis (EAE), acute lung injury, xenograft/transplant rejection, systemic inflammation/sepsis86, asthma, myocardial infarction, and atherosclerosis to highlight a few . From a small-molecule/peptide perspective, a summary of several more recently described TLR4-specific antagonists is listed in Table 2 and structures are presented in Figure 3. Past reviews describe other TLR4specific antagonists from previous years 87-89. RAGE antagonists are less abundant than those directed at the TLRs, yet also offer potential in the anti-inflammation field and are summarized in Table 2 with structures in Figure 3. Other isoforms (splice variants) of RAGE i.e. soluble RAGE (sRAGE) and RAGE∆ICD 90, have been shown to function as a competitive inhibitors of RAGE-ligand binding to signaling-competent RAGE by acting as decoys for HMGb1 binding. Thus, regulation of RAGE splicing can provide another level of control by the organism to influence extracellular HMGb1 levels. As both TLR and RAGE receptors bind other ligands, their


Page 11 of 50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


inhibition may have much broader effects, the advantage or disadvantage of which, in each specific disease, remains to be determined. The ability of HMGb1 to bind CXCL12 and activate CXCR4 chemotactic and migratory responses serves as a potential target for specific antagonist development91. Future research is needed to design antagonists which specifically block HMGb1-CXCL12 binding; however general CXCR4 antagonists (Table 2, Figure 3) are a growing domain in cancer, autoimmune, HIV and neuroinflammation research. 92 Achievements and remaining challenges to realizing HMGb1 as a biomarker/bioindicator and drug target The search for biological targets which are upregulated in disease conditions and can be used as biomarkers/indicators of disease activity, or exploited for design of small molecule or protein-based antagonists, is best exemplified by the development of the tumor necrosis factor (TNF) α antagonists. While generally effective for some patients in many inflammatory diseases, long-term use may lead to an increased risk of infections and/or malignancy 93-94. Moreover, in certain inflammatory conditions, antagonism of TNFα -an early phase pro-inflammatory factor- proves ineffective as it is believed later stage mediators (including HMGb1) may play a more critical role in the perpetuation of inflammation 14. HMGb1 seems primed to become the next viable candidate for an FDA-approved inflammatory biomarker/bioindicator and drug target. Current research in animal models strongly support its critical role in inflammatory diseases and conditions and there are several known drugs and preclinical compounds which bind and antagonize HMGb1 directly which can be used to interrogate its relevance (Table 2.). In addition, correlations of disease and disease activity with HMGb1 levels in humans has been advancing (summarized in Table 1), although further clinical studies with larger patient populations are needed. As a first step in this direction, a reliable and standardized assay for both tissue and serum/plasma is necessary. Many studies looking at HMGb1 levels used antibody-based systems (ELISA or immunohistochemistry) employing antibodies with various or unknown specificities. Some antibodies may not give true indications of HMGb1 levels if they target regions of interaction with binding proteins such as sRAGE, etc 95. Although several ELISA-


Page 12 of 50

based kits are available commercially for preclinical studies, a mass spectrometry-based assay would theoretically provide a more consistent measure of biological levels and also be able to distinguish PTM’s when desired 96-97. The effective development of a standardized, validated clinical assay for HMGb1 is a critical first step on the path towards effective HMGb1 antagonists and therapeutics. We highlight the necessary objectives towards advancement of HMGb1 as a clinical biomarker/indicator and drug target in Table 3. At the time of this writing, the NIH website, https://clinicaltrials.gov/, listed 37 studies with the query “hmgb1”. Of these, 10 studies are related to measurement of plasma HMGb1 (either as a primary or secondary outcome measure, no intervention), 26 involve some form of intervention (with plasma HMGb1 levels as either a primary or secondary outcome measurement), and 1 study was withdrawn before it was initiated. The interventions include drugs (14 studies), dietary supplements (2 studies), biologic or cell-based therapies (2 studies), exercise (1 study with results posted), rehabilitation (1 study), psycho-cognitive therapies (1 study), medical devices (4 studies), and 2 studies are related to amphotericin B and mistakenly use the term amphoterin B (an early term for HMGb1) and thus are not relevant, leaving the final tally at 35 HMGb1-related studies. Additional stratification reveals that of these 35 studies, 1 was withdrawn, 1 was terminated, 5 are not yet recruiting, 13 are recruiting, 8 are completed (only 1 posted results), and 7 have an unknown status. It is clear this field is still in its infancy. Furthermore, as the interventions are mainly targeted at other inflammation or cancer-related proteins and pathways, the secondary effects observed on HMGb1 levels do not prove direct correlations with disease activity. Future studies with directly-binding HMGb1 inhibitors (with minimal off-targets) will be needed for validation of this specific target in each disease condition. In conclusion, HMGb1 is poised to become a relevant drug target in inflammatory-related diseases and conditions including, but not limited to, cancer, sepsis, atherosclerosis, diabetes, chronic pain, and autoimmune, neurodegenerative, and heart disease. Refinement of quantitation techniques for HMGb1 from biological samples resulting in a standardized measurement assay, and careful design, execution and reporting of future clinical trials will ensure the successful advancement of therapeutics which target the extracellular escape and accumulation of this multifunctional protein. Design and testing of antagonists with greater HMGb1 specificity is essential and


Page 13 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


additional development of antagonists which target regions of HMGb1’s interaction with binding partners is also an area for development. Figure 1. Model and schematic view of HMGb1 structure with box regions and cysteine residues highlighted

Table 1. –Human HMGb1 associated diseases and conditions (not comprehensive) disease /condition

receptors/sign aling

serum /tissue levels (disease vs. control ng/mL)

pathways implicated


references


Alzheimer’s

Page 14 of 50

*98

*sRAGE interference in measurement?

99

brain - ↑1.5 fold vs. control

100

TLR2&4 and

spinal cord -↑mRNA and protein staining in slow-

101

RAGE

progressing ALS samples vs. controls

Tau and Amyloid-beta colocalization with hmgb1 (separately) ALS

102

↑ protein expression (2-3 fold) vs. control


Page 15 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


arthritis osteo (OA)

synovium-mRNA and protein ↑approx. 2-fold vs.

103

control 104

synovial fluid

105

cartilage – 100x ↑mRNA by qPCR vs. control

106

juvenile idiopathic

serum –(3.4 vs 0.8), p=0.0014

(JIA) 107

rheumatoid (RA) synovial fluid- (54) in RA vs.(12)in OA, p

High Mobility Group Box-1 (HMGb1): Current Wisdom and

Recommend Documents