New Non-Saccharide Heparin Mimetics as Inhibitors of Cathepsin G

Abstract

Cathepsin G (CatG), a neutrophil serine protease, plays several roles in host defense, inflammation, and vascular pathology. Selective inhibition of CatG is therefore a compelling therapeutic strategy, yet existing inhibitors such as sulfated saccharides are limited by structural heterogeneity, instability, and bleeding risk. In this study, sulfonated non-saccharide heparin mimetics were explored as a new class of CatG inhibitors with improved stability and drug-like properties. A focused library of derivatives was screened, and several candidates demonstrated potent inhibition, with the most active compound (inhibitor 7) achieving complete CatG inhibition at nanomolar concentrations (IC₅₀ = 86 nM). Inhibitor 7 exhibited remarkable selectivity over clotting factors, digestive enzymes, and related proteases, effectively protected extracellular matrix proteins from CatG-mediated degradation, and neither prolonged plasma clotting times nor induced cytotoxicity at tested levels. Mechanistic studies revealed an allosteric mode of action, whereby inhibitor 7 bound to a distinct anion-binding site on CatG, inducing local conformational changes that disrupted catalysis without affecting substrate recognition. Structural advantages of sulfonated scaffolds, including chemical stability, resistance to enzymatic cleavage, and synthetic accessibility, make them better than sulfated mimetics for long-term development. These findings identify sulfonated heparin mimetics, particularly inhibitor 7, as promising allosteric inhibitors of CatG with potential for safe and selective therapeutic application in vascular and inflammatory disorders.

Introduction

Cathepsin G (CatG) is a member of the neutrophil serine protease (NSP) family, which also includes human neutrophil elastase (HNE), proteinase 3, and neutrophil serine protease 4 (NSP4). These enzymes are important for innate immunity and inflammation. CatG was the first protease to be purified from human neutrophils in the late 1960s. CatG is a highly evolutionary conserved protein in mammals, suggesting that it plays important biological roles [1,2]. CatG is synthesized biochemically as a 255 amino acid precursor which, during granulopoiesis, is subject to proteolytic processing to produce the active enzyme (~29 kDa) [3]. The mature protein possesses the typical catalytic triad of serine proteases (His57, Asp102, and Ser195) in a chymotrypsin-like fold. CatG is characterized by an aromatic preference, albeit with favorable interactions occurring between CatG and Phe or Tyr, that demonstrates specificity for aromatic side-chains; however, it can also express a trypsin-type activity under modified conditions. CatG expression is confined to neutrophils, monocytes, and mast cells. CatG, once synthesized, is maintained within the azurophilic granules until neutrophils are exposed to inflammatory stimulators [4].

Importantly, structural studies demonstrate that CatG possesses a relatively shallow substrate-binding pocket, which accounts for its broader substrate specificity compared with other neutrophil proteases [5]. As a result, CatG is involved in a wide range of (patho-)physiological functions, including extracellular matrix turnover, the induction of matrix metalloproteases, the cleavage and liberation of cytokines, chemokines, and growth factors, as well as the activation of caspases and the generation of angiotensin II. Beyond these roles, CatG contributes to vascular pathologies through prothrombotic activity via platelet activation and endothelial damage, promotion of atheroma formation through LDL modification, potential involvement in aneurysm development, and exacerbation of ischemia–reperfusion injury [6–14]. In order to balance its proteolytic effect, CatG activity is modulated by a number of physiological inhibitors, including α1-antichymotrypsin, α1-proteinase inhibitor, α2-macroglobulin, serpin B1, and secretory leukocyte protease inhibitor [15]. The importance of this regulation has been underscored by recent attention to NSPs in inflammation, as uncontrolled inflammation often results from disruption of the natural protease–antiprotease balance [1,3,16,17].

Consistent with this, CatG is implicated in a wide spectrum of inflammatory diseases, including rheumatoid arthritis [18], psoriasis [19,20], cystic fibrosis [21], emphysema [22], and chronic obstructive pulmonary disease (COPD) [23,24]. Elevated CatG expression is a key driver of protease–antiprotease imbalance in these conditions [25]. Its excessive activity promotes connective tissue and ECM degradation, enhances apoptosis, and stimulates other pro-inflammatory mediators [15]. In COPD, CatG has emerged as both a marker of airway inflammation and a predictor of disease progression [26]. Therefore, the development of small-molecule or biologic CatG inhibitors holds significant therapeutic promise.

Although CatG inhibition shows considerable therapeutic promise across multiple diseases, the number of potent and selective inhibitors developed to date remains limited [15]. Reported small-molecule inhibitors include organophosphorus derivatives, boswellic acid derivatives, 2-substituted saccharines, and thiadiazolidinone dioxides [15,27], as well as N-arylacyl O-sulfonated aminoglycosides [28]. In addition, sulfated saccharides such as heparin and its derivatives have been described [29,30]. Mechanistically, organophosphorus derivatives, boswellic acid derivatives, 2-substituted saccharines, and thiadiazolidinone dioxides function primarily as active-site inhibitors. By contrast, the modified aminoglycosides act as partial mixed inhibitors of CatG, with reported IC₅₀ values ranging from 0.42 to 209 μM. Notably, heparin—an anticoagulant sulfated glycosaminoglycan—has been identified as a highly potent allosteric inhibitor, with an estimated K_i of <25 pM [29,30]. Despite this exceptional potency, the clinical utility of heparin as an anti-CatG therapeutic is limited due to its structural heterogeneity and the severe bleeding risk associated with its anticoagulant activity. Consequently, efforts have turned toward designing small sulfated and sulfonated molecules that mimic heparin’s function. While “mimicking” is often considered a structural concept, studies indicate that these small molecules reproduce heparin-like activity largely through the presence of sulfate or sulfonate groups on their scaffolds, thereby offering a promising strategy for the development of novel therapeutics.

One promising small-molecule inhibitor is sulfated pentagalloyl glucoside (SPGG), recently identified as an allosteric inhibitor of CatG [31]. However, SPGG suffers from several limitations: it exists as a heterogeneous mixture of variably sulfated species [32], and it also inhibits human factor XIa, leading to potent anticoagulant effects that raise bleeding risk [33]. More recently, a homogeneous octasulfated di-quercetin derivative was reported to inhibit CatG with nanomolar potency (~50 nM) [34]. Despite these advances, sulfated mimetics face substantial challenges in drug development, primarily due to the instability of sulfate groups, which are prone to enzymatic (sulfatase-mediated) and chemical hydrolysis. To overcome these barriers, we introduce sulfonated non-saccharide heparin mimetics as a novel class of CatG inhibitors. These molecules were specifically designed with a bioisosteric, negatively charged sulfonate group that mimics sulfate functionality while providing key advantages: (1) enhanced chemical stability, since sulfonates resist cleavage by sulfatases and remain intact under harsh chemical conditions [35–39]; (2) an improved chemical framework suitable for prodrug design, thereby facilitating better bioavailability; and (3) synthetic accessibility, as sulfonated precursors are commercially available and less prone to partial substitution, which often complicates the synthesis of sulfated analogues. The reported derivatives are allosteric, nonpolymeric, homogeneous, and highly water-soluble agents. Unlike sulfated mimetics, they do not contain hydrolytically labile functional groups susceptible to cleavage by DNase (phosphodiester hydrolysis), sulfatase (sulfate ester hydrolysis), or heparinase (heparin degradation). Moreover, their anionic nature reduces the likelihood of crossing the blood–brain barrier or placenta, thereby improving their safety profile. Importantly, the synthesis of these sulfonated derivatives is well established [40–42] and amenable to scale-up. Furthermore, strategies for designing sulfonate-based prodrugs have been described, offering pathways to enhance oral bioavailability [43,44].

In particular, this work puts forward the hexa-sulfonated inhibitor 7 (IC₅₀ = 86 nM) as a promising allosteric, potent, and nontoxic CatG inhibitor for future development as a potential treatment for inflammatory diseases. The molecule demonstrated enormous selectivity against a wide variety of proteases, including clotting factors, digestive enzymes, plasmin, other cathepsins, caspases, and metalloproteinases. Importantly, inhibitor 7 did not affect the clotting times of human plasma and inhibited CatG cleavage of laminin and fibronectin, indicating a great clinically relevant profile. Overall, inhibitor 7 demonstrates a remarkable potential as an anti-CatG drug that is under testing in relevant animal models to determine its pharmacokinetic and pharmacodynamic aspects.

Results and discussion

Rationale for library design.

The chemical library of inhibitors 1–8 was designed to develop stable, non-saccharide mimetics of heparin capable of engaging the anion-binding allosteric surface of CatG. Prior studies have established that highly anionic sulfated saccharides, including heparin and related GAG mimetics, potently inhibit CatG through electrostatic interactions with basic surface residues [29–31]; however, their clinical utility is limited by structural heterogeneity, enzymatic lability, and anticoagulant liability. We therefore selected sulfonated scaffolds as chemically stable sulfate bioisosteres that preserve high anionic charge density while offering homogeneity, synthetic accessibility, and improved drug-like properties.

To probe structure–activity relationships, the library systematically varies (i) total negative charge (four to eight anionic groups) to evaluate the contribution of electrostatic interactions to potency, and (ii) linker length and flexibility to assess how spatial presentation of charges influences binding to the extended basic surface of CatG. In addition, both globular and linear frameworks were included to examine the impact of three-dimensional shape on allosteric recognition, given that globular architectures may better cluster anionic groups, whereas linear scaffolds may span multiple basic residues. Collectively, this focused design enabled identification of key physicochemical and topological features governing potency, selectivity, and allosteric inhibition of CatG.

Direct inhibition of CatG by heparin mimetic molecules (1 – 8).

A small library of sulfonated molecules (Figure 1) was screened for the inhibition of CatG using the corresponding chromogenic substrate hydrolysis assay under near-physiological conditions, as described earlier [31,34]. Molecules 1 and 2 are globular, while the rest are linear. All molecules are sulfonated, except molecule 8 possesses phosphonate groups. Molecules 1, 3, and 8 carry four negative charges, whereas molecules 4 – 7 carry six negative charges, and molecule 2 possesses eight sulfonate groups. In the chromogenic substrate hydrolysis assay, the initial rate of substrate hydrolysis, as measured by absorbance at 405 nm, is indicative of the catalytic activity of CatG. The fractional decrease in the initial rate of hydrolysis in the presence of an inhibitor is analyzed using the dose–response (Eq. 1) to calculate the potency (IC₅₀, HS) and the efficacy (Y₀, Y_M) parameters. Figure 2 shows the semilog inhibition curves observed for all inhibitors. The range of inhibitory potency was found to be broad (0.086–8.10 μM), while the efficacy for most inhibitors was moderate to high (>66%) (Table 1). Important structure–activity relationship aspects can be inferred from this study. First, inhibitors with globular structures appear to have better inhibition potency than the corresponding linear structures with the same number of sulfonate groups. For example, the globular tetra-sulfonated inhibitor 1 was 10-fold more potent than the linear tetra-sulfonated inhibitor 3. Second, increasing the number of sulfonate groups generally increases the potency. On one hand, increasing the number of sulfonate groups from four to eight in globular structures, as in inhibitors 1 and 2, increases the potency about 9-fold. On the other hand, increasing the number of sulfonate groups from four to six in linear structures, as in inhibitors 3 and 4, increases the potency about 3-fold. Third, sulfonated derivatives with extended linkers were more potent than similar molecules with shorter linkers (inhibitors 5 – 7 were at least 11.5-fold more potent than inhibitor 4). Fourth, the phosphonate group can substitute the sulfonate group while maintaining a moderate inhibition potency, as in inhibitor 8 (IC₅₀ = 3.1 μM). Finally, by comparing the inhibition potency of molecules 5 and 6 with that of inhibitor 7, it is evident that the most optimal position for the trisulfonated naphthalene-4-formamido(benzamido) substituent is the para-position relative to the urea moiety. In fact, although all three inhibitors are hexa-sulfonated derivatives, inhibitor 7 is at least 2.7-fold more potent than inhibitors 5 and 6. Overall, the most potent inhibitor identified in this exercise was the sulfonated molecule 7, which completely inhibited CatG with an IC₅₀ value of 86 nM.

Figure 1.

A) The chemical structure of UFH, which is characterized as an allosteric inhibitor with an estimated K_i of <25 pM, and the chemical structure of SPGG, which is a sulfated small molecule allosteric inhibitor of CatG. B) The small library of anionic non-saccharide heparin mimetics, which are evaluated in this work.

Figure 2.

Direct inhibition of human CatG by molecule 1-8. The inhibition profiles were developed using the corresponding chromogenic substrate hydrolysis assay as described in the experimental part. Solid lines represent sigmoidal dose−response fits (equation 1) of the data to obtain the values of IC₅₀, HS, and ΔY.

Table 1.

Direct Inhibition of Human CatG by Sulfonated Molecules.

Molecule	IC50 (μM)a	HSb	ΔY (%)c
1	0.830 ± 0.140d	1.7 ± 0.4	91.9 ± 6.8
2	0.090 ± 0.010	1.5 ± 0.3	82.5 ± 7.4
3	8.100 ± 0.430	5.2 ± 2.0	66.0 ± 2.9
4	2.640 ± 0.240	3.1 ± 0.9	88.4 ± 4.3
5	0.160 ± 0.020	3.1 ± 1.1	81.5 ± 6.1
6	0.230 ± 0.100	1.1 ± 0.5	94.3 ± 10.8
7	0.086 ± 0.020	0.8 ± 0.1	96.1 ± 6.9
8	3.100 ± 0.190	1.1 ± 0.1	87.5 ± 1.8

^aThe IC₅₀ values,

^bHS,

^cΔY were obtained by non-linear regression of direct CatG inhibition by the sulfonated molecules.

^dError represents ± 1 S.E.

Effect of salt concentration on CatG inhibition by inhibitors 2 and 7.

To investigate the nature of CatG inhibition by these derivatives, we measured the IC₅₀ values of the two most potent inhibitors, 2 and 7, under variable salt concentrations. In this direction, the two derivatives exhibited potent inhibition potencies (0.090 and 0.086 μM, respectively) toward CatG in the presence of 100 mM concentration of NaCl solution. Table 2 shows that increasing the salt concentration from 100 to 200 mM decreased the potency of CatG inhibition by derivative 2 by ~10.5-fold (0.09 μM vs 0.92 μM). Likewise, increasing the salt concentration from 100 to 200 mM decreased the potency of CatG inhibition by derivative 7 by ~9.5-fold (0.086 μM vs 0.816 μM). Previously, we reported that molecule 7 inhibits human CatG with an IC₅₀ value of 0.57 ± 0.05 μM in the presence of 150 mM NaCl (Table 2), indicating a ~6.6-fold reduction in potency and strong salt-dependent inhibition. In the two cases, there was about a 7–13% decrease in the efficacy of the inhibition. Overall, the sulfonated derivatives appear to reversibly inhibit CatG primarily via electrostatic/hydrogen bond interactions, largely directed by the sulfonate groups.

Table 2.

Salt-Dependent Inhibition of CatG by Sulfonated Inhibitors 7 and 2.^a

Inhibitor	[NaCl] (mM)	IC50 (μM)	HS	ΔY (%)
2	100	0.090 ± 0.010b	1.5 ± 0.3	82.5 ± 7.4
	200	0.920 ± 0.150	1.5 ± 0.3	75.2 ± 3.4
7	100	0.086 ± 0.020	0.8 ± 0.1	96.1 ± 6.9
	150 c	0.57 ± 0.05	1.8 ± 0.3	89.7 ± 2.9
	200	0.816 ± 0.058	0.9 ± 0.2	83.2 ± 6.6

^aThe values of IC₅₀, HS, and ΔY were obtained following the nonlinear regression analysis of direct inhibition of human CatG in an appropriate buffer of pH 7.4 at 37 °C. Inhibition was monitored by the spectrophotometric measurement of residual enzyme activity.

^bErrors represent ±1 SE.

^cValue from reference [50].

Inhibition mechanism: Michaelis–Menten kinetics of chromogenic hydrolysis by CatG in the presence of inhibitors 2 and 7.

To understand the mechanistic basis of CatG inhibition by sulfonated derivatives, the Michaelis–Menten kinetics of S-7388 hydrolysis by CatG were performed in the presence of inhibitor 2 at pH 7.4 and 37 °C. Figure 3A shows the initial rate profiles in the presence of inhibitor 2 (0–800 nM). Each profile displays a characteristic rectangular hyperbolic trend, which could be fitted using Eq. 2 to give the apparent K_M and V_MAX (Table 3). The K_M for S-1384 did not significantly change (1.44 ± 0.17–1.32 ± 0.27 mM) in the presence or absence of inhibitor 2. However, the V_MAX decreased from 33.5 ± 1.8 mAU/min in the absence of inhibitor 2 to 10.6 ± 0.95 mAU/min at 800 nM of inhibitor 2. Likewise, Figure 3B shows the initial rate profiles in the presence of inhibitor 7 (0–400 nM). Each profile shows a rectangular hyperbolic line, which could be fitted using Eq. 2 to give the apparent K_M and V_MAX (Table 3). The K_M for S-1384 did not significantly change (1.44 ± 0.17–1.58 ± 0.20 mM) in the presence or absence of inhibitor 7. However, the V_MAX decreased from 33.5 ± 1.8 mAU/min in the absence of inhibitor 7 to 12.9 ± 0.9 mAU/min at 400 nM of inhibitor 7. Thus, the inhibitors appear to bring about structural changes in the active site of CatG, which do not affect the formation of the Michaelis complex but lead to a significant disruption in CatG catalytic activity. This indicates that inhibitors 2 and 7 are allosteric (non-competitive) inhibitors.

Figure 3.

Michaelis–Menten kinetics of S-7388 hydrolysis by human CatG in the presence of inhibitors 2 and 7. The initial rate of hydrolysis at various substrate concentrations was measured in pH 7.4 buffer as described in the experimental part using human CatG. Various concentrations of each inhibitor were used. Solid lines represent nonlinear regressional fits to the data using the standard Michaelis–Menten equation to calculate the V_MAX and K_M.

Table 3.

Michaelis–Menten Kinetics of CatG Hydrolysis of Chromogenic Substrate (S-7388) in the Presence of Inhibitors 2 and 7.^a

Inhibitor	KM (mM)	VMAX (mAU/min)
[2] (nM)
0	1.44 ± 0.17	33.5 ± 1.8
40	1.20 ± 0.3	26.1 ± 2.5
100	1.29 ± 0.19	17.6 ± 1.1
400	1.40 ± 0.22	11.4 ± 0.8
800	1.32 ± 0.27	10.6 ± 0.95
[7] (nM)
0	1.44 ± 0.17b	33.5 ± 1.8
25	1.03 ± 0.2	24.8 ± 1.9
50	0.9 ± 0.2	17.7 ± 1.9
100	1.23 ± 0.1	14.4 ± 0.5
200	1.52 ± 0.2	13.6 ± 0.7
400	1.58 ± 0.2	12.9 ± 0.9

^aK_M and V_MAX values were measured as described in the experimental section. mAU indicates milli-absorbance units.

^bError represents ± 1 S.E.

CatG inhibition by 2 and 7 is physiologically relevant. Effect of 2 and 7 on CatG-mediated proteolysis of laminin and fibronectin.

Although molecule 7 inhibits CatG hydrolysis of chromogenic substrate, extracellular matrix (ECM) components serve as more relevant substrates of CatG. In fact, based on their in vitro properties, we reasoned that inhibitor 7 could protect ECM components from proteolysis mediated by proteinases activated during the inflammatory process, including CatG. Accordingly, we studied the in vitro effect of inhibitor 7 on the CatG-mediated proteolysis of fibronectin and laminin, the major non-collagenous components of ECM and basement membranes [45]. SDS-PAGE analyses showed that fibronectin was significantly cleaved by CatG, as evidenced by the disappearance of ≥200-kDa fibronectin bands (Figure 4A, lane 5). In the presence of inhibitor 7 (100–1000 μM), fibronectin is protected from CatG-mediated cleavage (Figure 4A, lanes 2–4). Similar results were shown by inhibitor 2, the second most potent inhibitor (Figure 4A, lanes 7–9). Likewise, SDS-PAGE analyses showed that laminin was significantly cleaved by CatG, as evidenced by the disappearance of the ≥200-kDa laminin bands (Figure 4B, lane 10). In the presence of inhibitor 7 (10–1000 μM), however, laminin is concentration-dependently protected from cleavage by CatG (Figure 4A, lanes 7–9). Similar results were shown by inhibitor 2, the second most potent inhibitor (Figure 4B, lanes 2–4). Together, SDS-PAGE analyses indicate that the inhibition of CatG by molecules 7 and 2 is physiologically relevant.

Figure 4.

A) Effect of inhibitors 2 and 7 on CatG-mediated hydrolysis of fibronectin as evaluated by SDS-PAGE under reducing conditions. The inhibitor concentrations used were 0, 10, 100, and 1000 μM. B) Effect of inhibitors 2 and 7 on CatG-mediated hydrolysis of laminin as evaluated by SDS-PAGE under reducing conditions. The inhibitor concentrations used were 0, 10, 100, and 1000 μM.

However, it is important to be aware of the apparent discrepancy in the activity of the inhibitors as demonstrated by the results from the two differing experiments, i.e., the chromogenic substrate hydrolysis assay and the SDS-PAGE. A potential reason for such a discrepancy can be attributed to the size of the substrates used in the two experiments. When evaluating the activity of the inhibitors using the chromogenic substrate hydrolysis assay, a small tripeptide chromogenic substrate (S-7388; molecular weight is 624.64 g/mol) is used. It is expected here that the activity of the inhibitor will be better. However, larger protein substrates such as fibronectin (molecular weight can be 440–530 kDa) and laminin (molecular weight can be 400–900 kDa) are used in the SDS-PAGE analysis. Under these conditions, a higher concentration of the inhibitor is needed to bring about a substantial inhibition. Importantly, the SDS-PAGE results are essentially indicative. While the results quantitatively demonstrate that inhibitor concentrations as low as 100 μM block the hydrolysis of laminin and fibrinogen by CatG, the results also indicate that the inhibitors work well even with more physiologically relevant substrates.

Selectivity Studies: Inhibition of other relevant enzymes, including clotting factors, plasmin, digestive enzymes, neutrophil elastase, proteinase 3, other cathepsins, caspases, BACE1, and matrix metalloproteinases by inhibitor 7.

To determine the selectivity of this class of CatG inhibitors over other relevant enzymes, we have chosen the most potent inhibitor identified in this study, which was inhibitor 7. A series of chromogenic substrate hydrolysis assays (for clotting factors, plasmin, digestive enzymes, neutrophil elastase, proteinase 3), fluorogenic substrate hydrolysis assays (other cathepsins, caspases, BACE1, and matrix metalloproteinases), and a fluorescence-based bi-substrate trans-glutamination assay (Factor XIIIa) were used, as reported earlier [31–34, 46–48]. In these assays, the inhibition potential was determined by spectrophotometric measurement of the residual protease activity in the presence of varying concentrations of inhibitor 7. The IC₅₀ values (or estimates based on the highest concentrations tested) of the inhibitor 7 against the corresponding enzymes are reported in Table 4. The inhibitor demonstrates remarkable selectivity toward human CatG over a wide range of enzymes. Interestingly, the molecule exhibits sub-micromolar inhibition potency toward HNE (IC₅₀=0.22 μM), suggesting a potential additional benefit in treating inflammatory conditions. Overall, the potency and the selectivity indices of inhibitor 7 make it a strong candidate for testing in appropriate in vivo models of inflammatory diseases.

Table 4.

Inhibition of Serine Proteases by Inhibitor 7.^a

Enzyme	IC50 (μM)	HS	ΔY (%)	Selectivity index
Cathepsin G	0.086 ± 0.02b	0.8 ± 0.1	96.0 ± 7.0	1
Thrombin	>100	ND c	ND	>1163
Factor Xa	49.0 ± 15.0	1.8 ± 1.0	83.0 ± 15.0	570
Factor XIa	4.6 ± 0.4	2.3 ± 0.4	102.0 ± 4.0	53.5
Factor IXa	>350	ND	ND	>4070
Factor XIIIa	84.0 ± 28.0	1.5 ± 0.8	91.0 ± 15.0	977
Neutrophil elastase	0.22 ± 0.00	2.6 ± 0.1	105.0 ± 1.0	2.6
Proteinase 3	34.0 ± 13.0	0.8 ± 0.2	49.0 ± 8.0	395
Plasmin	>200	ND	ND	>2326
Trypsin	>200	ND	ND	>2326
Chymotrypsin	>200	ND	ND	>2326
Cathepsin B	~3.5	~0.8	~87	41
Cathepsin F	~0.48	~2.6	~98	5.6
Cathepsin L	~33.9	~0.7	~71	394
Cathepsin S	~10.3	~1.9	~62	120
Cathepsin V	>100	ND	ND	>1163
Caspase 3	>100	ND	ND	>1163
Caspase 6	~2.9	~1.4	~97	33.7
Caspase 7	~7.7	~1.2	~79	89.5
Caspase 8	~9.7	~1.8	~101	113
BACE 1	~14.6	ND	~69	170

^aThe values of IC₅₀, HS, and ΔY were obtained following the nonlinear regression analysis of direct inhibition of human enzymes in an appropriate buffer of pH 7.4 at 25–37 °C.

^bErrors represent ±1 SE.

^cNot determined.

Effects of inhibitor 7 on human plasma coagulation and cell proliferation.

To further evaluate the selectivity as well as the antiproliferative properties of inhibitor 7, we studied its effects on the clotting times of normal human plasma as well as its effects on the proliferation of three cell lines. On the one hand, the inhibitor’s effects on activated partial thromboplastin time (APTT) and prothrombin time (PT) were measured, as described earlier under in vitro conditions [49–52], using variable concentrations (0–550 μM). Results indicated that the inhibitor’s concentrations estimated to double APTT or PT are >450 μM (>5200-fold of the IC₅₀), suggesting the lack of a significant effect on heparin-binding coagulation proteins such as thrombin, FXa, FIXa, FXIa, antithrombin, and heparin cofactor II. On the other hand, the antiproliferative properties of the inhibitor were evaluated in three cell lines of breast (MCF-7), intestine (CaCo-2), and kidney (HEK-293), as described earlier [53,54]. Similar to testing in human plasma, results indicated that 10 μM of the inhibitor (>116-fold of the IC₅₀) does not significantly affect the proliferation of the above cell lines. Taken together, inhibitor 7 appears to exhibit significant selectivity indices over other heparin-binding proteins in human plasma, and it is not associated with cellular toxicity at the highest concentration tested of 10 μM. Overall, inhibitor 7 can be considered for further development as a potent, selective, and nontoxic inhibitor of CatG.

Molecular modelling studies. Docking and molecular dynamics simulations involving inhibitors 2 and 7.

To elucidate the binding modes of inhibitors 2 and 7 on CatG, we first conducted molecular docking studies as described in the experimental section. The amino acid sequence ¹⁰⁹SRRVRRNRN¹¹⁷, which includes the consensus motif XBBXBX (¹¹²VRRNRN¹¹⁷), was selected to define the putative allosteric anion-binding site, consistent with our previous findings [31]. Both inhibitors were predicted to bind at this allosteric site, fitting well within the defined sequence region. Docking revealed distinct binding behaviors for the two inhibitors. Inhibitor 2 was located within the putative site, forming interactions with multiple arginine residues (R110, R111, R113, R114, and R116) (Figures 5A and 5B). By contrast, inhibitor 7 interacted with R110, R113, and R116, but its more linear structure allowed additional interactions with R23 and R86 (Figures 6A and 6B). In both cases, the inhibitors established a combination of hydrogen-bonding, electrostatic interactions, and salt bridges with the key arginine residues and surrounding amino acids.

Figure 5.

Comparison of the binding of inhibitor 2 to CatG at time points 0 and 100 ns MD simulations. A) Cartoon representation of the binding mode of 2 with CatG at time of 0 ns. B) Schematic 2D representation of the CatG-inhibitor 2 interaction diagram at time of 0 ns. C) Cartoon representation of the binding mode of 2 with CatG at time of 100 ns. D) Schematic 2D representation of the CatG-inhibitor 2 interaction diagram at time of 100 ns. The catalytic triad H57, D102, and S195 are shown in spheres with C atoms in cornflower blue, and the allosteric anion-binding site populated with Arg residues and other ligand-interacting amino acids are shown in spheres with C atoms in orchid color.

Figure 6.

Comparison of the binding of inhibitor 7 to CatG at time points 0 and 100 ns MD simulations. A) Cartoon representation of the binding mode of 7 with CatG at time of 0 ns. B) Schematic 2D representation of the CatG-inhibitor 7 interaction diagram at time of 0 ns. C) Cartoon representation of the binding mode of 7 with CatG at time of 100 ns. D) Schematic 2D representation of the CatG-inhibitor 7 interaction diagram at time of 100 ns. The catalytic triad H57, D102, and S195 are shown in spheres with C atoms in cornflower blue, and the allosteric anion-binding site populated with Arg residues and other ligand-interacting amino acids are shown in spheres with C atoms in orchid color.

To assess the stability and conformational dynamics of these complexes, we performed 100 ns all-atom molecular dynamics (MD) simulations in explicit water. The backbone root mean square deviation (RMSD) values for ligand-free CatG and the CatG–inhibitor complexes remained below 2.3 Å across the trajectories, confirming overall structural stability (Figure 7A). Most fluctuations were attributed to the flexible loop regions. The final CatG–inhibitor complexes after 100 ns are shown in Figures 5C/D and 6C/D. Inhibitor 2 largely remained within the proposed binding site, maintaining interactions with R110, R111, R113, and R114, as well as with R86. In contrast, inhibitor 7 shifted slightly during the simulation, forming additional contacts with R243 and R87 while preserving interactions with R111, R113, and R116. We next evaluated the fluctuations of specific residues relative to the reference structure. The root mean square fluctuations (RMSF) of C_α atoms were obtained (Figure 7B). The fluctuations were generally similar in the inhibitor-enzyme complex and the free enzyme. No significant fluctuations in the Cα atoms of catalytic triads (H57, D102, S195) have been observed. However, the RMSF values of M146 and R147 in the CatG–inhibitor 7 complex were about 1 Å higher than those in the free enzyme and the CatG–inhibitor 2 complex, making a distant effect upon the binding of inhibitor 7 to CatG. Likewise, the RMSF values of R114, N115, and R116 that interact with inhibitor 2 were higher than those in the free enzyme and CatG–inhibitor 7 complex. This analysis shows that though both inhibitors bind to the putative allosteric site, they have some distinct effects on the protein. We also examined whether binding at the putative allosteric site influenced the side chain orientations of the residues of the catalytic triad of CatG. Superposition of the initial and final structures (Figure 8) revealed no major alterations in secondary structure. In ligand-free CatG and the CatG–inhibitor 7 complex, the catalytic triad residues retained all hydrogen-bond interactions, with distances between the H57 imidazole nitrogen, the D102 carboxyl group, and the S195 hydroxyl oxygen remaining <3.5 Å for >90% of the simulation. By contrast, in the CatG–inhibitor 2 complex, H57 flipped its orientation around 60 ns, disrupting the H-bond with S195 and increasing the H57–S195 distance to >9 Å, a shift that persisted throughout the remainder of the simulation.

Figure 7.

A) The backbone’s root mean square deviation (RMSD) plots of ligand-bound and unbound complexes along the 100 ns MD simulations. B) The root mean square fluctuation (RMSF) of the C_α atoms of CatG in ligand-bound and unbound complexes during the 100 ns MD simulations.

Figure 8.

Superposition of the active site - catalytic triad residues at times 0 (blue color ribbon and sticks) and 100 ns (pink color ribbon and sticks) MD simulations. A) Ligand-free CatG system. B) CatG-inhibitor 2 complex system. C) CatG-inhibitor 7 complex system. Distances between the H57 imidazole nitrogens and the side chain carboxylic carbon of D102 and the hydroxyl oxygen of S195 are shown in black dashed lines, and the values for the initial (at time 0) and the final structures are shown in blue and pink, respectively. Hydrogen bonds are shown in yellow dashed lines.

Together, these docking and MD studies suggest that inhibitors 2 and 7 bind to a putative allosteric site through interactions with key arginine residues, but induce distinct downstream effects in the protein and in the catalytic triad. For instance, while inhibitor 7 preserved the integrity of the active site, inhibitor 2 caused a conformational disruption in H57–S195 interactions. Although these findings require experimental validation through crystallography or mutagenesis, the computational insights highlight a plausible allosteric site on CatG that can guide rational optimization of inhibitors 2 and 7 for enhanced potency and selectivity.

Conclusion

Given CatG’s well-documented role in vascular pathologies and inflammatory disorders, selective CatG inhibitors hold a significant promise as therapeutic agents for vascular and inflammatory disorders. In this study, we evaluated a series of non-saccharide heparin mimetics and identified inhibitor 7 as the most potent and selective CatG inhibitor. With an IC₅₀ value of 86 nM, complete inhibition of CatG activity, and extensive selectivity over related proteases and coagulation factors, inhibitor 7 demonstrated a highly favorable pharmacological profile. It effectively protected physiologically relevant extracellular matrix proteins, such as fibronectin and laminin, from CatG-mediated proteolysis, while exhibiting no anticoagulant activity or cytotoxicity at pharmacologically relevant concentrations. Mechanistic studies revealed that both inhibitors 2 and 7 act as allosteric, non-competitive inhibitors, reducing the catalytic activity of CatG without blocking substrate binding. Molecular docking and dynamics simulations further demonstrated that both molecules target a putative allosteric anion-binding site, but with distinct binding modes. Importantly, inhibitor 7 induced a local change but maintained the structural integrity of the catalytic triad, while inhibitor 2 induced conformational disruption of the H57–S195 hydrogen bond. These findings suggest that subtle differences in binding geometry can translate into distinct downstream effects on CatG activity, providing valuable insights for rational inhibitor design.

The allosteric inhibition strategy offers several advantages over active-site inhibition. For example, the shallow and promiscuous substrate-binding pocket of CatG makes it inherently difficult to achieve selectivity through active-site binding alone. In contrast, allosteric inhibitors exploit distinct binding surfaces, enabling both higher selectivity and reduced cross-reactivity with other serine proteases. Furthermore, because allosteric inhibitors modulate enzyme activity without directly competing with substrates, they maintain sustained inhibition even under conditions of elevated proteolytic activity. Equally important are the structural advantages of the sulfonated heparin mimetics. Unlike sulfated saccharides, these molecules are non-polymeric, chemically homogeneous, and highly water-soluble, minimizing the heterogeneity and instability issues that complicate drug development. The sulfonate groups provide greater chemical stability than sulfate esters, resisting enzymatic hydrolysis by sulfatases, and heparinases. In fact, their anionic nature reduces the likelihood of crossing the blood–brain barrier or placenta, which may translate into an improved safety profile. Their modular design also facilitates synthetic feasibility, scalability, and the potential for prodrug strategies to enhance oral bioavailability.

Taken together, our biochemical, structural, and computational studies establish sulfonated heparin mimetics, particularly inhibitor 7, as a promising new class of potent, selective, and non-toxic allosteric CatG inhibitors. These molecules overcome key limitations associated with sulfated saccharide-based inhibitors while leveraging the therapeutic and structural advantages of allosteric modulation. They therefore represent attractive leads for further preclinical evaluation. Future studies, including crystallography and mutagenesis, as well as in vivo pharmacokinetic and pharmacodynamic assessments, will be critical to validate the proposed allosteric binding site and to advance inhibitor 7 toward therapeutic development, especially for vascular and inflammatory disease indications.

Materials and Methods

Materials.

Human CatG was purchased from Enzo Life Sciences (Farmingdale, NY). Human plasma clotting enzymes and digestive enzymes were purchased from Haematologic Technologies (Essex Junction, VT). Laminin, fibronectin, and the chromogenic substrate for CatG (S-7388) were purchased from Sigma Aldrich (St. Louis, MO). The chromogenic substrates for thrombin, FIIa, FVIIa, FXa, FIXa, and FXIIa were purchased from Biomedical Diagnostics (Windsor, NS, Canada). FXIa chromogenic substrate (S-2366) was purchased from Diapharma (West Chester, OH). Pre-cast SDS-PAGE gels were from BioRad (Hercules, CA). Coomassie brilliant blue for gel electrophoresis was from Fisher Scientific (Waltham, MA). Stock solutions of CatG were prepared in 20 mM Tris-HCl buffer, pH 7.4, containing 0.02% Tween80, 0.1% PEG8000, 2.5 mM CaCl₂, and 100 mM NaCl. N,N-dimethylcasein, dansyl-cadaverine, and dithiothreitol were also from Sigma Aldrich. Stock solutions of HNE were prepared by reconstituting 1 mg with 100 μL of 1:1 glycerol:200 mM sodium acetate buffer, pH 5, which was then diluted using HEPES buffer, pH 7.4, containing 125 mM HEPES, 0.125% Triton-X 100, and 100 mM NaCl. Stock solutions of thrombin, FXIa, cathepsin G, proteinase 3, plasmin, trypsin, and chymotrypsin were all prepared in 50 mM TrisHCl buffer, pH 7.4, containing 150 mM NaCl, 0.1% PEG8000, and 0.02% Tween80. A stock solution of FIXa was prepared in 20 mM Tris-HCl buffer, pH 7.4, containing 100 mM NaCl, 2.5 mM CaCl₂, 0.1% PEG 8000, 0.02% Tween 80, and 33% v/v ethylene glycol. Stock solution of FXa was prepared in 20 mM TrisHCl buffer, pH 7.4, containing 100 mM NaCl, 2.5 mM CaCl₂, 0.1% PEG8000, and 0.02% Tween80. Stock solution of FXIIIa was prepared in 50 mM Tris-HCl, 1 mM CaCl₂, 100 mM NaCl, 0.1% PEG8000, 0.02% Tween80, and 2 mg/mL N,N–dimethylcasein. Materials for assays involving other cathepsins, caspases, and matrix metalloproteinases were obtained from BPS Bioscience (San Diego, CA). Inhibitors 1-8 were chemically characterized and synthesized as reported earlier with purity exceeding 95% [55]. All experiments were repeated at least twice.

For the clotting assays, pooled normal human plasma for coagulation assays was purchased from George King Bio-Medica (Overland Park, KS). Activated partial thromboplastin time reagent containing ellagic acid, thromboplastin-D (PT reagent), and 25 mM solution of CaCl₂ was purchased from Thermo Fisher Scientific (Waltham, MA). All experiments in this paper were repeated at least twice.

Inhibition of cathepsin G (CatG) by anionic non-saccharide heparin mimetics.

Direct inhibition of CatG was determined at pH 7.4 and 37°C by a chromogenic substrate hydrolysis assay, as described previously [31,34]. To each well of a 96-well microplate containing 88 μl of 20 mM tris buffer containing 100 mM NaCl (or 200 mM), 2.5 mM CaCl₂, 0.1% PEG 8000, and 0.05% tween 80 was added 3 μl of CatG (final concentration of 30 nM), and 5 μl of H₂O or potential inhibitor (final concentration of 0–100 μM). Following a 5-minute incubation, 3 μl of CatG substrate was added (final concentration 750 μM), and the residual CatG activity was obtained from the initial rate of increase of absorbance at 405 nm. The relative residual activity of CatG at each of the inhibitor concentrations was obtained from the ratio of CatG activity in the presence and absence of the inhibitor. Logistic Equation (1) was used to plot the dose-dependence curve to obtain the IC₅₀ (potency) and efficacy of CatG inhibition. Here, Y is the ratio of residual CatG activity in the presence of the inhibitor to that in its absence, Y_O and Y_M are the minimum and maximum values of fractional residual CatG activity, respectively, IC₅₀ is the concentration of the inhibitor that results in 50% inhibition of CatG activity, and HS is the Hill slope.

$$Y=Y_0+\frac{Y_M-Y_0}{1+{10}^{\left(Log{\left[I\right]}_0-Log{IC}_{50}\right)\left(HS\right)}}$$ Eq. 1

Michaelis–Menten enzyme kinetics.

The initial rate of the hydrolysis of the chromogenic substrate by CatG was monitored from the linear increase in absorbance corresponding to less than 10% consumption of the substrate at 37°C in a pH 7.4 20 mM Tris buffer containing 100 mM NaCl, 2.5 mM CaCl₂, 0.1% PEG 8000, and 0.05% Tween 80. The initial rate was measured at various substrate concentrations (0–2500 mM) at a fixed enzyme concentration (30 nM), and fixed inhibitor concentrations (0 – 800 nM). The data were fitted by the Michaelis–Menten Eq. (2) to determine the K_M (substrate affinity) and V_MAX (maximum reaction velocity).

$$V=\frac{V_{MAX}\left[S\right]}{K_M+\left[S\right]}$$ Eq. 2

Inhibition of CatG cleavage of laminin and fibronectin by anionic non-saccharide heparin mimetics.

Inhibition of CatG cleavage of laminin and fibronectin by inhibitors 2 and 7 was studied using SDS-PAGE, as previously reported [31]. Briefly, CatG (0.8 or 0.5 μM) was incubated with different concentrations of the inhibitors (final concentrations: 0 μM, 10 μM, 100 μM, and 1000 μM), and laminin (20 μg) or fibronectin (36 nM). Following incubation for 60 min, the samples were quenched using SDS-PAGE loading buffer containing DTT and subjected to electrophoresis on 10% SDS-PAGE pre-cast gels. The gels were visualized by silver staining.

Inhibition of clotting factors, digestive enzymes, neutrophil elastase, proteinase 3, and plasmin by molecule 7.

The inhibition potential of molecule 7 against human thrombin, factor Xa, factor IXa, and factor XIa was evaluated using chromogenic substrate hydrolysis assays. The inhibition potential of molecule 7 was also evaluated against human FXIIIa using a bisubstrate fluorescence-based transglutaminase assay, as reported in the literature under physiological conditions. The inhibitory potentials of molecule 7 were also evaluated against neutrophil elastase, proteinase 3, plasmin, and digestive proteases (trypsin and chymotrypsin) using the corresponding chromogenic substrate hydrolysis assays as reported earlier [31–34, 47–52]. These assays were performed using substrates and conditions appropriate for the enzyme under investigation. At least six serially diluted concentrations of inhibitor 7 were exploited, and the fractional residual enzyme activity was measured at each concentration. The inhibition profile was determined over a range of inhibitor concentrations to determine the IC₅₀, HS, and ΔY. The concentrations of enzymes and substrates in microplate cells were identical to those reported in our previous studies [31–34, 47–52].

In the chromogenic hydrolysis assay, each well of the 96-well microplate generally contained 85 μL (factor IXa, factor XIa, and plasmin) or 185 μL (thrombin, factor Xa, neutrophil elastase, proteinase 3, trypsin, and chymotrypsin) of Tris-HCl pH 7.4 buffer to which 5 μL of HNE inhibitor (or solvent reference) and 5 μL of enzyme were sequentially added. After 5–10 min incubation, 5 μL of the chromogenic substrate was rapidly added and the residual enzyme activity was measured from the initial rate of increase in absorbance at 405 nm. The buffer of factor IXa also contained 33% ethylene glycol. In the factor XIIIa assay, 1 μL of inhibitor 7 was diluted with 87 μL of Tris-HCl pH 7.4 buffer and 5 μL DTT (20 mM) at 37 °C followed by the addition of 2 μL of FXIIIa (0.3 μM) and incubation for 10 min. The activity of FXIIIa was evaluated by the addition of 5 μL of dansylcadaverine (2 mM) and measuring the initial rate of increase in fluorescence emission (λ_Ex. =360 nm and λ_Em.=490 nm). In the above assays, the dose dependence of the fractional residual enzyme activity at each concentration of inhibitor 7 was analyzed using Eq. 1 to obtain the apparent concentration of HNE inhibitor required to reduce enzyme activity to 50% of its initial value.

Inhibition of other cathepsins (B, F, L, V, and S) by molecule 7.

The inhibitory activity of molecule 7 against cathepsins B, F, L, S, and V was evaluated using a fluorescence-based enzymatic assay. Inhibitor 7 was dissolved in water to a concentration of 10 mM, while the reference inhibitors E-64 and Cystatin C were prepared in DMSO at the same concentration. Serial dilutions of all molecules were made in a 10% DMSO/water mixture using the cathepsin assay buffer. For each reaction, 5 μL of the compound dilution was added to a 50 μL total reaction volume, resulting in a consistent final concentration of 1% DMSO across all assay conditions. Test molecules were pre-incubated in duplicate at room temperature for 30 minutes in a mixture containing cathepsin assay buffer, the corresponding cathepsin enzyme, and the molecule being tested. The amounts of enzyme used per reaction were as follows: 0.4 ng for cathepsin B, 75 ng for cathepsin F, 0.4 ng for cathepsin L, 20 ng for cathepsin S, and 2 ng for cathepsin V. The concentration ranges evaluated were 0.005–100 μM for inhibitor 7, 0.0003–10 μM for E-64, and 0.00003–1 μM for Cystatin C. Following the pre-incubation step, enzymatic reactions were initiated by adding fluorogenic peptide substrates. For cathepsins B, L, S, and V, 5 mM Z-Leu-Arg-AMC was used, while for cathepsin F, 10 mM Z-Phe-Arg-AMC was employed. The reactions were allowed to proceed for 60 minutes at room temperature, after which fluorescence intensity was measured using a Tecan Infinite M1000 microplate reader at an excitation wavelength of 360 nm and an emission wavelength of 460 nm.

All assays were conducted in duplicate, and the resulting fluorescence data were analyzed using GraphPad Prism software. Fluorescence intensity in the absence of inhibitor (Ft) was defined as 100% enzymatic activity, while fluorescence in the absence of enzyme (Fb) was defined as 0% activity. The percent activity at each compound concentration was calculated using the formula: % activity = (F − Fb)/(Ft − Fb), where F represents the fluorescence intensity in the presence of the test compound. The data were further analyzed by plotting percent activity against the logarithm of compound concentration, and fitted using a four-parameter sigmoidal dose-response model: Y = B + (T − B)/[1 + 10^((LogEC₅₀ − X) × Hill slope)], where Y is percent activity, B is the minimum response, T is the maximum response, X is the logarithm of the compound concentration, and the Hill slope reflects the steepness of the curve. The IC₅₀ value, defined as the concentration of inhibitor that results in 50% inhibition of enzyme activity, was determined from the curve. The reference inhibitor E-64 showed IC₅₀ values of 0.018 μM for cathepsin B, 0.061 μM for cathepsin L, 0.13 μM for cathepsin S, and 0.12 μM for cathepsin V. Cystatin C exhibited an IC₅₀ value of 0.23 μM against cathepsin F. These values served as benchmarks for evaluating the potency of inhibitor 7 against the same enzymes.

Inhibition of caspases (3,6, 7, and 8) and BACE1 by molecule 7.

Caspase activity assays were performed in duplicate to evaluate the inhibitory effects of the test molecules. Each enzymatic reaction was carried out in a 50 μL mixture containing 10 mM HEPES buffer (pH 7.4), 10 mM EDTA, 0.05% CHAPS, 5 mM DTT, 1 μM of the appropriate fluorogenic substrate, the corresponding caspase enzyme, and the test compound. Reactions were incubated at room temperature for 30 minutes. Fluorescence intensity was then measured using a Tecan Infinite M1000 microplate reader with an excitation wavelength of 380 nm and an emission wavelength of 505 nm. The specific enzyme–substrate combinations used were caspase-3 with 1 μM caspase-3 substrate (1 ng/reaction), caspase-6 with 1 μM Ac-VEID-AFC (15 ng/reaction), caspase-7 with 1 μM caspase-3 substrate (3 ng/reaction), and caspase-8 with 1 μM Ac-LEHD-AFC (5 ng/reaction). For BACE1 inhibition studies, the BACE1 FRET Assay Kit (BPS Bioscience) was used according to the manufacturer’s protocol. Each reaction contained 200 ng of BACE1 enzyme and 7.5 μM of the supplied fluorogenic BACE1 substrate. Fluorescence was measured using the same Tecan microplate reader, with excitation and emission wavelengths set to 360 nm and 460 nm, respectively.

All protease assays, including caspase and BACE1 reactions, were performed in duplicate at each tested concentration. Fluorescence data were analyzed using GraphPad Prism software. The fluorescence intensity observed in the absence of inhibitor (Ft) was defined as 100% enzymatic activity, while the fluorescence in the absence of enzyme (Fb) was defined as 0% activity. Percent enzymatic activity in the presence of each compound was calculated using the equation: % activity = (F − Fb)/(Ft − Fb), where F represents the fluorescence in the presence of the compound.

Dose-response curves were generated by plotting percent activity against the logarithm of compound concentration. Data were fitted using a four-parameter sigmoidal model according to the equation: Y = B + (T − B) / [1 + 10^((LogEC₅₀ − X) × Hill slope)], where Y is percent activity, B is minimum activity, T is maximum activity, X is the log of compound concentration, and the Hill slope reflects the steepness of the curve. The IC₅₀ value was defined as the concentration of the molecule that reduced enzymatic activity by 50%. The molecules tested included inhibitor 7 (10 mM water stock; test range: 0.003–100 μM), Ac-DNLD-CHO (100 μM stock in DMSO; 0.00003–1 μM range), Ac-IETD-CHO (5 mM stock in DMSO; 0.0002–50 μM range), and verubecestat (1 mM stock in DMSO; 0.0001–3 μM range). The IC₅₀ values for the reference inhibitors were as follows: Ac-DNLD-CHO exhibited IC₅₀ values of 0.016 μM for caspase-3 and 0.094 μM for caspase-7; Ac-IETD-CHO showed IC₅₀ values of 0.14 μM for caspase-6 and 0.035 μM for caspase-8; and verubecestat inhibited BACE1 with an IC₅₀ value of 0.026 μM.

Inhibition of matrix metallopeptidases (MMPs 1–3, 7–10, and 12–14) by molecule 7.

Matrix metalloproteinase (MMP) inhibition assays were conducted using a panel of MMP enzymes, each activated under optimized conditions before testing. For example, MMP1 and MMP9 were diluted in MMP Buffer 1 and activated at 37°C for 2 hours. MMP2 was diluted in MMP Buffer 2 and similarly activated for 2 hours at 37°C. MMP3 (K45E mutant) was prepared at a concentration of 100 ng/mL in a buffer composed of 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 10 mM CaCl₂, 0.05% Brij-35, and 1 mM APMA, and activated at 37°C for 1 hour. Enzymatic reactions were conducted in duplicate in 50 μL volumes at 37°C for 30 minutes. Each reaction contained the respective assay buffer, MMP enzyme, fluorogenic MMP substrate, and a test molecule. Fluorescence was measured using a Tecan Infinite M1000 microplate reader. For MMP1, MMP2, and MMP9, readings were taken at 328 nm excitation and 393 nm emission. For MMP3, excitation/emission wavelengths were 320/405 nm, and for TACE, 485/530 nm.

All enzyme activity assays were performed in duplicate across a range of molecule concentrations. Fluorescence data were analyzed using GraphPad Prism software. In each dataset, the fluorescence in the absence of inhibitor (C_e) was defined as 100% enzymatic activity, while fluorescence in the absence of enzyme (C₀) was defined as 0% activity. Percent enzymatic activity in the presence of the molecule was calculated as: % Activity = (C − C₀) / (C_e − C₀), where C is the fluorescence measured in the presence of the compound. Background fluorescence was subtracted from all wells. Dose-response curves were generated by plotting percent activity versus the logarithm of the molecule concentration, and fit using a four-parameter sigmoidal equation: Y = B + (T − B) / [1 + 10^((LogEC₅₀ − X) × Hill slope)], where Y is the percent activity, B is the minimum activity, T is the maximum activity, X is the logarithm of the molecule concentration, and Hill slope is the slope coefficient. IC₅₀ values were defined as the compound concentration causing 50% inhibition of enzyme activity.

The MMPs and reagents used included MMP1, MMP2, MMP3 (K45E), and MMP9 (Q279R) from BPS Bioscience; MMP7 from R&D Systems; and MMP8, MMP10, MMP12, MMP13, and MMP14 from BPS Bioscience. Fluorogenic substrates included general MMP substrates (from BPS) and the MMP3-specific substrate (from R&D Systems). NNGH was obtained from Cayman Chemical, APMA from Sigma, and other buffer reagents (Tris-HCl, NaCl, CaCl₂, Brij-35) from Sigma and BioPioneer. Test molecules included inhibitor 7 (10 mM water stock, tested from 0.005–100 μM), and reference inhibitors NNGH, Batimastat, and MMP408 (10 mM stocks in DMSO, tested from 0.0005–10 μM). Enzyme and substrate concentrations used for each assay were as follows: MMP1 (32 ng/reaction, 1 μM substrate), MMP2 (100 ng, 2 μM), MMP3 (100 ng, 10 μM), MMP7 (20 ng, 1 μM), MMP8 (4 ng, 1 μM), MMP9 (4 ng, 1 μM), MMP10 (64 ng, 1 μM), MMP12 (20 ng, 1 μM), MMP13 (32 ng, 1 μM), and MMP14 (64 ng, 1 μM). The IC₅₀ values for the reference inhibitors are summarized as follows: 1) NNGH inhibited MMP1 (0.12 μM), MMP2 (0.11 μM), MMP3 (0.029 μM), MMP8 (0.009 μM), MMP9 (0.0041 μM), MMP10 (0.024 μM), and MMP14 (0.007 μM); 2) Batimastat showed potent inhibition of MMP7 (0.016 μM) and MMP13 (0.23 μM); and 3) MMP408 selectively inhibited MMP12 with an IC₅₀ of 0.012 μM. These values served as benchmarks for evaluating the potency of inhibitor 7 and other test inhibitors across the MMP enzyme panel.

Human plasma coagulation studies.

Clotting time was measured in a standard one-stage recalcification assay with a BBL Fibrosystem fibrometer (Becton−Dickinson, Sparles, MD), as reported previously [31–34, 47–52]. In the PT assay, thromboplastin-D was reconstituted according to the manufacturer’s directions and warmed to 37 °C. A 10 μL amount of inhibitor 7, to give the desired concentration, was brought up to 100 μL with citrated human plasma, incubated for 30 sec at 37 °C followed by the addition of 200 μL of prewarmed thromboplastin-D, and the time to clot was recorded. In the APTT assay, 10 μL of the same inhibitor was mixed with 90 μL of citrated human plasma and 100 μL of prewarmed APTT reagent (0.2% ellagic acid). After incubation for 4 min at 37 °C, clotting was initiated by adding 100 μL of prewarmed 0.025 M CaCl₂, and the time to clot was noted. The data were fit to a quadratic trend line, which was used to determine the concentration of the inhibitor necessary to double the clotting time. Clotting time in the absence of inhibitor 7 was determined in a similar fashion using 10 μL of deionized water and was found to be ~12 sec for PT and ~31 sec for APTT.

Effect of inhibitor 7 on cellular proliferation.

MCF-7 human breast carcinoma cells, HEK-293 human embryonic kidney cells, and CaCo-2 heterogeneous human epithelial colorectal adenocarcinoma cells were maintained as previously described [47, 49–52]. The test was done using a modified procedure in which Alamar Blue (resazurin) fluorescent dye was used [47, 49–52]. Briefly, MCF-7, HEK293, and CaCo-2 cells were seeded in a 96-well plate at an optimized concentration to allow them to be in log growth phase. Each cell line was individually optimized. The number of cells was determined by adding Alamar Blue dye, incubating for 2 hrs at 37 °C, and then reading the plate in a fluorescent reader with λ_EX. of 560 nm and λ_EM. of 590 nm on the Synergy H1 multiplate reader (BioTek, Winooski, VT). The lids were always kept on the plates during this experiment, as the cells need to be kept sterile for the remainder of the experiment. The number of cells was determined in the absence and the presence of inhibitor 7 (10 μM), which was incubated with the cells for 3 days. Distilled water control was included in each assay to ensure that the effect observed was due to the molecule and was not because of apoptotic/necrotic effects of the vehicle. The experiments were repeated four times to determine the corresponding standard deviations.

Molecular modeling studies.

Structure-based molecular docking studies followed by molecular dynamics simulations were conducted to study the binding modes of the two most active sulfonated non-saccharide heparin mimetics (inhibitors 2 and 7) to CatG. Docking studies were carried out using the Molecular Operating Environment (MOE) 2022 suite [56]. The protein structure was retrieved from the crystal structure of human CatG in complex with a peptidyl phosphonate inhibitor (PDB ID: 1CGH) [57]. Then, the protein structure was prepared using the MOE protein preparation module by removing water molecules, adding hydrogen atoms at a physiological pH using Protonate3D, and assigning atomic charges to all protein atoms, followed by energy minimization. The 3D structures of the inhibitor molecules were generated and subsequently energy minimized using the builder module of MOE. As reported in our previous work [31], the amino acid sequence ¹⁰⁹SRRVRRNRN¹¹⁷ consisting of the consensus amino acid sequence XBBXBX motif (¹¹²VRRNRN¹¹⁷) was selected to define the purported allosteric binding site, where X and B are noncharged and positively charged amino acids, respectively. The default parameters were used to do docking experiments. The best protein ligand binding complex for each ligand, based on the docking score, was then subjected to molecular dynamics (MD) simulations. All-atom MD simulations with explicit water (TIP3P model) in a cubic box with periodic boundary conditions were carried out for 100 ns using the Desmond software [58] for ligand-free CatG and ligand-bound CatG obtained from the docking studies. The simulation box boundaries were set 10 Å away from the protein/protein-ligand complexes, and the system was neutralized by adding counter ions. Additionally, a salt concentration of 0.15 M NaCl was maintained to mimic the physiological condition. The default Desmond protocol was used for minimization and relaxation using the OPLS4 forcefield. MD simulations were run for 100 ns in the NPT ensemble with a 300 K Nose-Hoover thermostat and 1 atm pressure, and the trajectories were saved at every 100 ps interval.

Abbreviations

APTT

Activated partial thromboplastin time

CatG

Cathepsin G

COPD

Chronic obstructive pulmonary disease

FIIa

Factor IIa

FXa

Factor Xa

FXIa

Factor XIa

FIXa

Factor IXa

FXIIIa

Factor XIIIa

MD

Molecular dynamics

MMP

Matrix metalloproteinase

MOE

Molecular operating environment

NSP

Nucleophilic serine protease

HNE

Human neutrophil elastase

PT

Prothrombin time

RMSD

Root mean square deviation

SPGG

Sulfated pentagalloylglucoside