**1. Introduction**

Matrix metalloproteinase-8 (MMP-8), also known as neutrophil collagenase or collagenase-2, is a Zn+<sup>2</sup> -dependent endopeptidase of the MMP family. It is primarily expressed and secreted by neutrophils as a zymogen (pro-MMP-8), and is activated by the reactive oxygen species released from activated neutrophils, thus rendering MMP-8 as having a key role in both acute and chronic inflammation [1]. Tissue inhibitors of metalloproteinases (TIMPs) regulate the activity of MMP-8 by forming a 1:1 stoichiometric inhibitory complex [2]. However, imbalances in MMP-8 and TIMP levels can lead to increased collagen degradation and pathological remodeling of the extracellular matrix [1]. Although it was initially believed that MMP-8′ s only function was to degrade collagen, later studies have identified both non-collagenous and non-structural substrates [1], thereby implicating MMP-8 in various pathological processes (see Table 1 for the full listing). In recent years, a wealth of preclinical and clinical data have emerged that correlate high MMP-8 levels with the potential for diagnosing diseases [3]. Likewise, MMP-8 inhibitors have also been suggested as potential therapeutics for treating some of these pathologies [3].


**Table 1.** Matrix metalloproteinase-8 (MMP-8) in various diseases.

Most early attempts to develop MMP inhibitors focused on developing peptide-based agents mimicking endogenous substrate sequences aided by a hydroxamate group, which ligates the critical Zn+<sup>2</sup> co-factor [3,16]. Unfortunately, the high affinity of hydroxamates to Zn+<sup>2</sup> containing proteins, which include several MMPs, as well as a disintegrin and metalloproteinase (ADAM) and ADAM with thrombospondin motifs (ADAMTS) family members, introduced a significant non-selectivity of inhibition, causing severe side-effects in the clinical trials [3,16]. Later, researchers focused on the development of peptides with a lower affinity to Zn+<sup>2</sup> by including phosphonates, thiolates, and carboxylates in the structure (see Table S1) [3,16–18]. Although such peptides demonstrated significantly reduced side-effects, they failed to show sufficient efficacy in the clinical trials [3]. This resulted in a rather negative perception of MMPs as drug targets in the community [3].

After nearly a decade of no progress in clinical trials with MMP inhibitors, a better understanding of the MMP biology has led to the insight that the initial clinical trials were perhaps held prematurely [3,16]. Learning from both the fundamental advances made so far and the limitations of the broad-spectrum non-selective MMP inhibitors, researchers have focused on the development of selective small molecule MMP inhibitors. This has led to the development of selective MMP-2, -9, and -13 inhibitors. Several of these molecules are now being tested in clinical trials [3,16].

In contrast to these successes, MMP-8 appears to be a rather challenging target. Although small molecule inhibitors are likely to prevent and treat multiple diseases (Table 1), few reports are available on the development of MMP-8 inhibitors. In fact, only two publications have reported MMP-8 inhibitors in the last decade [19,20]. Thus, the vast chemical space offered by small molecules appears to be almost untouched for MMP-8 inhibitor development.

Natural glycosaminoglycans (GAGs) modulate various functions of MMPs, including cell localization, conformation, and stability [21,22]. As one would expect, based on the extensive literature behind GAG–protein interactions [23,24], GAGs are expected to affect MMP activity, depending on the length of their polymeric chains and the structure of the local sulfated microdomains [21,25]. In fact, distinct GAGs have been found to inhibit MMP-9, which is closely related to MMP-8 [26]. Furthermore, the potency of natural GAGs against MMPs is quite promising. However, the heterogeneity of these GAGs is problematic for drug development. Additionally, polymeric natural GAGs generally display unwanted side-effects; they are impossible to chemically synthesize, and their bio-analytical characterization is a major challenge. These challenges reduce the enthusiasm to pursue natural GAGs as drugs [27].

In contrast, sulfated small molecules that mimic GAGs offer much better prospects. These molecules, referred to as non-saccharide GAG mimetics (NSGMs), have now been shown to bind and functionally modulate multiple GAG-binding proteins [28–34]. For example, distinct NSGMs have been identified as ligands of fibroblast growth factor receptor-1 [28], antithrombin [33,34], neutrophil elastase [29], coagulation factor Xia [30,35], plasmin [31], and viral glycoprotein D [32], each of which are known to bind to heparin. In fact, several NSGMs have been found to display one-to-one correspondence with specific GAG sequences [28,36]. In terms of drug-like properties, NSGMs are fully synthetic, which bodes well for analytical ease and scalability of preparation [27,37]. Their much smaller molecular size (MW 500–2000) in comparison with natural GAGs (MW 15,000–100,000) enables a rational and/or computational design for identifying advanced analogs [28]. Additionally, NSGMs are highly water-soluble owing to the presence of one or more sulfate groups, which enables easier formulation, e.g., direct instillation into the lung. This property could have important value for lung disorders, such as chronic obstructive pulmonary disease (COPD) and idiopathic pulmonary fibrosis (IPF), where MMP-8 inhibition has shown to be beneficial.

Over the past decade, we have developed a library of NSGMs (Figure 1) with 12 different chemical scaffolds (apigenin, benzofuran, catechin, glucoside, gossypetin, inositol, luteolin, morin, phloretin, quercetin, quinazolinone, and resveratrol). The library contains mimetics with different chain lengths (monomer, dimer, and trimer), as well as varying sulfation levels (mono-sulfate through dodeca-sulfate), which affords an excellent potential to mimic the diversity of the chain length and sulfation patterns of GAGs.

The rationale behind the development of the NSGM library was that while the sulfate groups will offer GAG-like protein recognition features, the aromatic scaffold will impart hydrophobic characteristics. These dual ionic and non-ionic forces of recognition are likely to enhance selectivity for the target protein of interest. Using this concept, distinct NSGMs have been developed as promising anticoagulants [35], antifibrinolytics [31], anticancer [39], anti-inflammatory [29], and antiviral agents [32]. More importantly, our work has shown that initial "hits" can be systematically developed through an appropriate hit-to-lead optimization strategy into potent and selective candidates, with favorable pharmacokinetic properties and minimum adverse side-effects [30,36,40,42].

In this work, we screen a library of NSGMs using computational and in vitro techniques to identify promising inhibitors of MMP-8. Our work shows, for the first time, that MMP-8 can be inhibited by targeting electropositive GAG binding sites, and offers at least one promising small molecule inhibitor of MMP-8.

**Figure 1.** Structures of the 58 non-saccharide glycosaminoglycans (GAGs) mimetics (NSGMs) studied against MMP-8. The synthesis and characterization of these molecules has been previously reported [30,31,38–41].
