A structural mechanism of flavonoids in inhibiting serine proteases

Guangpu Xue ab, Lihu Gong a, Cai Yuan c, Mingming Xu a, Xu Wang ab, Longguang Jiang *d and Mingdong Huang *ad
aState Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, China. E-mail: jianglg@fzu.edu.cn; HMD_lab@fzu.edu.cn
bCollege of Life science, Fujian Normal University, Fuzhou 350117, China
cCollege of Biological Science and Engineering, Fuzhou University, Fuzhou 350116, China
dCollege of Chemistry, Fuzhou University, Fuzhou 350116, China

Received 20th December 2016 , Accepted 1st June 2017

First published on 2nd June 2017


Quercetin is a member of the flavonoids and was previously demonstrated to inhibit trypsin-like serine proteases at micromolar potencies. Different molecular models were proposed to explain such inhibition. However, controversies remain on the molecular details of inhibition. Here, we report the X-ray crystal structure of quercetin in a complex with the urokinase-type plasminogen activator (uPA), an archetypical serine protease. The structure showed that quercetin binds to the specific substrate binding pocket (S1 pocket) of uPA mainly through its two neighboring phenolic hydroxyl groups. Our study thus provides unambiguous evidence to support quercetin binding to serine proteases and defines the molecular basis of the interaction. Our results further establish that natural products with two adjacent phenolic hydroxyl groups (or catechol) are likely to inhibit other trypsin-like serine proteases, a new mechanism formerly under-recognized.


1. Introduction

Flavonoids are a large group of polyphenolic compounds that occur ubiquitously in the plant kingdom as secondary metabolites.1,2 These compounds share a common structure core with two benzene rings (A ring and B ring) joined by a third pyranic ring (C ring) (Fig. 1A).3 Flavonoids have been known for a long time to exert multiple biological effects, including digestion inhibition, anti-anaphylaxis, anti-inflammation, anti-microbial, vascular protection, anti-diabetic and anti-cancer.4–6 They even possess human immune deficiency virus (HIV) killing activity and are potential drugs for acquired immunodeficiency syndrome (AIDS).7 On the other hand, as major functional components of many herbal plants, flavonoids have been used for medical treatment and health care for millennia.8 For the foregoing reasons, flavonoids are referred to as nutraceuticals.9
image file: c6fo01825d-f1.tif
Fig. 1 Flavonoid core structural skeleton (A) and the chemical structures of several flavonoids mentioned in this text (B–D).

Despite the extensive studies on flavonoids, the detailed mechanisms of flavonoids’ functions have not been totally known yet. Potential molecular mechanisms of flavonoids include anti-oxidation, radical scavenging, nitric oxide scavenging, xanthine oxidase inhibition, leukocyte immobilization, prostanoid biosynthesis inhibition, toxin neutralization, cellular signal interference, and gene regulation.10,11 Recently, the inhibition of serine proteases was proposed as a new function mechanism for flavonoids by several independent laboratories.12–14 It was reported that flavonoids can act as inhibitors of several key serine proteases. Quercetin (Fig. 1B), as a member of the flavonoids, was found to inhibit serine proteases (urokinase-type plasminogen activator (uPA), thrombin and trypsin) with IC50 values in the micromolar range (7 μM, 34.5 μM and 10 μM, respectively). Quercetin also inhibits plasmin with a Kd value of 0.62 μM. Moreover, myricetin (Fig. 1C), another member of flavonoids, inhibits proteases trypsin and thrombin with an IC50 of 15 μM and 6 μM, respectively.13,15,16 These findings suggest that flavonoids may be therapeutically useful for serine protease-related diseases, such as tumor and cardiovascular diseases.

Serine proteases belong to a colossal hydrolase family in which serine residues serve as the nucleophilic amino acid at the active site to specifically cleave peptide bonds of their substrates. They play critical and diverse roles in a number of physiological processes, such as digestion, immune response, blood coagulation and reproduction.17 Serine proteases chymotrypsin, trypsin and elastase are digestive enzymes.18 In addition, serine proteases are also important therapeutic targets in a series of pathological indications, including thrombosis, hypertension and diabetes.19–21 Thus, there is great interest in developing specific and potent inhibitors for serine proteases. Flavonoids can potentially be a new class of candidates of serine protease inhibitors. To date, there have been a number of studies on the interaction of flavonoids with serine proteases. However, the results are not consistent with each other. It was proposed that the B ring of quercetin locates nearby the uPA active site, while the A ring interacts with His57 (chymotrypsinogen numbering) based on molecular docking of quercetin on uPA. However, different results were obtained from the docking study of quercetin on a homologous serine protease, plasmin, where the A and C rings bind to different sites.13 In addition, epigallocatechin gallate (EGCG, a component of green tea) (Fig. 1D) was shown to cover the catalytic triad (His57, Asp102 and Ser195), leading to the inhibition of uPA proteolytic activity. Therefore, a critical study is needed to resolve the detailed molecular interaction between flavonoids and serine proteases.

In this study, we determined the crystal structure of quercetin in a complex with uPA to elucidate the detailed interaction between uPA and quercetin. uPA is an important serine protease involved in fibrinolysis and tumor metastasis,22,23 and is an attractive target for medical treatment. The structural analysis showed that quercetin indeed binds to uPA but in a way different from that proposed by previous studies. Furthermore, the structure revealed a distinct binding mode not observed in conventional uPA inhibitors. These findings provide a new direction to develop therapeutical drugs targeting uPA and other serine proteases using flavonoids as templates.

2. Materials and methods

2.1. Materials

Pichia pastoris strain X-33 was purchased from Invitrogen. Involved chemicals and reagents were purchased from Sinophram Chemical Reagent Co., Ltd and Aladdin Industrial Corporation. Quercetin was purchased from Sigma-Aldrich with a purity no less than 98%. Sepharose fast flow and Superdex 75 HR 10/30 size exclusion columns were from GE Healthcare. Ultra-filtration cells (Model-5124) were from Millipore and Amicon Bioseparations.

2.2. Expression and purification of human recombinant uPA

Expression and purification of recombinant uPA was described previously by Zhao et al.24 Briefly, uPA was secreted from a stably transfected Pichia pastoris strain (X-33) after induction with methanol. A purified product was obtained after cation exchange chromatography (SP Sepharose Fast Flow) followed by gel filtration chromatography using a Superdex 75 HR 10/300 column equilibrated with 20 mM phosphate buffer (pH 6.5) containing 150 mM NaCl. The protein was eluted as a single peak under these conditions with a retention volume of approximately 13.6 ml, corresponding to its molecular weight (about 28 kDa). The recombinant uPA protease domain expressed this way is an active protease with an activity comparable to the full-length two-chain uPA.24 The protein was dialyzed in 20 mM potassium phosphate (pH 6.5) overnight and concentrated to 10 mg ml−1 using stirred ultra-filtration cells.

2.3. Crystallization and data collection of the quercetin:uPA complex

The crystallization trials were carried out using the method of sitting drop vapor diffusion.25 After mixing equal volumes of protein solution and precipitant solution (50 mM sodium citrate at pH 4.6, 1.95 M (NH4)2SO4, 0.03% NaN3, and 5% PEG400) at room temperature, uPA crystals appeared about 3–5 days later. The crystals of uPA were soaked for 5 days in new soaking buffer (40% PEG4000, 100 mM Tris-HCl pH 7.4) containing 1 mM quercetin. A solution of 40% PEG4000, 100 mM Tris-HCl, pH 7.4, 20% (V/V) glycerol was used as a cryoprotectant for the crystals. X-ray diffraction data of the uPA–quercetin complex were collected at Shanghai Synchrotron Radiation Facility Beamline BL17U at a wavelength of 0.979 Å.26 The space group of uPA–quercetin was determined to be R3 by preliminary manipulation using the HKL2000 program package,27 with unit-cell parameters: a = 121.1 Å, b = 121.1 Å, and c = 42.9 Å. The most probable Matthews coefficient was 2.07 Å3 Da−1 and corresponded to one protein molecule per asymmetric unit with a solvent content of 40.52%. The dataset was 96.36% complete to 3.0 Å and with a high mean redundancy of 3.1. The statistics of data collection are listed in Table 1. The interface area between quercetin and uPA molecules was calculated by PISA (http://www.ebi.ac.uk/msd-srv/prot_int/cgi-bin/piserver).28
Table 1 X-ray data collection and model refinement statistics for the quercetin:uPA crystal
Compound uPA–quercetin
Highest-resolution shell is shown in parenthesis.a Rmerge = ∑|Ii<I>|/∑Ii, where Ii is the intensity of the ith observation and <I> is the mean intensity of the reflections.b Percentage of residues in the most favored regions.c Percentage of residues in additional allowed regions.d Percentage of residues in disallowed regions.
Data collection
X-ray source wavelength (Å) 0.979
Resolution limits (Å) 34.6–3.0 (3.1–3.0)
Space group R3
Temperature of experiments (K) 100
Cell constants a = b = 121.116 Å, c = 42.902 Å, α = β = 90°, γ = 120°
Completeness (%) 96.36
Multiplicity 3.1
R merge 0.056 (0.108)
Number of observations 13265
Number of unique reflections 4317
 
Refinement data
R factor 0.170
R free 0.259
Average B-factor (Å2) of uPA 55.1
Average B-factor (Å2) of quercetin 84.6
r.m.s Deviation of bond lengths (Å) 0.009
r.m.s Deviation of angle (°) 1.41
Ramachandran analysis (%) 93.0b, 6.2c, 0.8d


2.4. Phasing and refinement

The structure of the quercetin:uPA complex was solved by a molecular replacement method using the MolRep program,29 which gave very strong and unambiguous solutions. An uPA was first positioned inside the crystal lattice using the uPA structure (PDB code: 4DVA) as a searching model and all the X-ray data up to 3.0 Å.30 The FoFc electron density map calculated at this stage showed the presence of quercetin at the active site of uPA. The molecular replacement model was subjected to iterative refinement and manual model rebuilding using Refmac31 and Coot,32 alternately, giving a final R factor and R free factor of 0.17 and 0.259, respectively, in a resolution range of 34.6–3.0 Å (Table 1). The structure was validated with PROCHECK.33 The coordinates of the final model were deposited in PDB (PDB code 5XG4). The final results were analyzed and visualized by using software PyMol.34

3. Results and discussion

3.1. Crystal structure of the quercetin:uPA complex

In this study, we aim to understand how quercetin inhibits uPA through determining the structure of quercetin in a complex with uPA by the X-ray crystallography technique. Full-length uPA is a multi-domain protein containing a growth factor domain,35 a kringle domain36 and a serine protease domain (SPD).37 In order to facilitate the protein crystallization, we generated the recombinant protease domain of uPA with activity comparable to full-length uPA.24 The recombinant uPA-SPD was expressed in Pichia pastoris strain (X-33), and purified to homogeneity. The uPA-SPD protein was formed into crystals followed by soaking with quercetin. The crystal structure of this quercetin:uPA complex was determined to a resolution of 3.0 Å and was refined to satisfactory statistics (R factor = 17.0%, R free = 25.9%, Table 1).

We clearly identified that the quercetin molecule binds to the active sites of uPA (Fig. 2A), as shown by the electron densities (Fig. 2B). 71.4% of the area of quercetin is involved in the interaction with uPA. The B ring of quercetin inserts into the specific substrate binding pocket (S1 pocket) (Fig. 2C and D) of uPA, while two hydroxyl groups of the A ring direct to the S2 pocket (Fig. 2A and E). Two neighboring phenolic hydroxyl groups at B3′ and B4′ positions (or catechol) point to the bottom of the S1 pocket, forming hydrogen bonds with the uPA residues Asp189, Ser190 and Gly219 (Fig. 3A). Hydrophobic interactions, from quercetin aromatic rings to uPA residues Trp215 and Gln192 (Fig. 3B), appear to be also important in mediating the quercetin:uPA interaction. At another end of quercetin, three oxygen atoms on the A and C rings make polar interaction with His57 and His99 of the S2 pocket (Fig. 3A). His57 is a member of the catalytic triad (His57, Asp102 and Ser195) which facilitates the catalysis by abstracting a proton from Ser195.38,39 Direct interaction with His57 may be one aspect for quercetin to inhibit the catalytic activity of uPA. The hydroxyl group at the A5 position can also form one hydrogen bond with the side chain of Gln192 (Fig. 3A). All hydrogen bonds and their lengths are listed in Table 2. All the detailed information of quercetin in a complex with uPA has been exhibited in the crystal structure which provides a reasonable explanation for the high affinity.


image file: c6fo01825d-f2.tif
Fig. 2 Crystal structure shows the binding sites of quercetin to uPA. A. Quercetin binds to the active site of uPA. The surface of uPA is colored according to the electrostatic potential. Quercetin is shown as sticks (carbon atoms in green and oxygen atoms in red). B. The 2FoFcσ-weighted electron density map wrapping the quercetin molecule at a contour level of 0.7σ is shown as a magenta mesh. C. The side view of the electrostatic potential surface shows the profile of the S1 pocket. A key residue locating at the bottom of S1 pocket Asp189 is also given as sticks. D. The top view of the electrostatic potential surface shows the outlet shape and interior charge distribution of the S1 pocket. E. The electrostatic surface rendering the silhouette of the S2 pocket. His57 and His99 of the S2 pocket which interact with quercetin are indicated in sticks. Quercetin in C, D and E figures is hided for visualization.

image file: c6fo01825d-f3.tif
Fig. 3 Detailed interactions between quercetin and uPA. A. Hydrogen bonding networks (dashed lines) between uPA and quercetin in two orientations. uPA residues and quercetin are shown in gray and green sticks, respectively. Oxygen atoms in red and nitrogen atoms in blue. Specific substrate binding pockets are shown in arcs. B. Hydrophobic interactions between quercetin and uPA are shown by purple arrows.
Table 2 Hydrogen bonds and lengths between quercetin and uPA
Quercetin uPA Distance (Å)
The distances between donor and recipient atoms are limited to no more than 3.6 Å as consensus.
O17 Asp189OD2 3.2
O17 Gly219O 3.4
O18 Ser190OG 2.9
O19 Ser195OG 3.4
O19 Ser214O 2.0
O20 His57ND1 3.3
O21 Gln192NE2 3.4
O23 His57NE2 2.8
O23 His99NE2 3.4
O23 Ser214O 3.6


3.2. Flavonoids with neighboring phenolic hydroxyl groups (catechol) can serve as inhibitors of trypsin-like serine proteases

An important finding in this work is that the catechol (benzenediol) moiety can function as the P1 group in a protease inhibitor to bind to the S1 pocket of trypsin-like serine proteases (TLPs). The P1 group is an amino acid residue in a polypeptide substrate of serine protease closest to the scissile bond towards the NH2-terminus, the corresponding binding site on enzyme is the S1 pocket.40 To inhibit TLPs, alkaline moieties with a high isoelectric point, e.g., amidine, guanidine or arginine, are typically used as the P1 group to recognize the Asp189 at the bottom of the major substrate binding pocket (S1 pocket) (Fig. 4A). These alkaline groups typically form charged hydrogen bonds to the Asp189.41 Thus formed hydrogen bonds locate inside the S1 pocket, an environment shielded away from the solvent and having a low dielectric constant. Such charged hydrogen bonds can contribute high binding energy (up to 4.5 kcal mol−1) compared to regular hydrogen bonds (2–3 kcal mol−1).42,43 Thus, the P1–S1 interaction is typically most important in determining the potency of protease inhibitors.17 Such protease inhibitors with a highly alkaline P1 group are not good drug candidates, because the P1 group will be protonated at physiological pH, hindering their penetration through the lipid bilayer of the cell membrane and leading to poor oral bioavailability of the inhibitors.44,45 In this regard, catechol can be a better replacement owing to its electroneutrality at physiological pH.
image file: c6fo01825d-f4.tif
Fig. 4 Catechol serves as a new type of P1 group binding to the bottom residue (Asp189) of the S1 pocket (gray outline) of TLPs in a way resembling to the typical P1 group. Inhibitors and protease residues are shown as sticks with carbon atoms in gray in typical inhibitors and protease residues, while green in catechol inhibitors, all nitrogen atoms in blue and oxygen atoms in red. Hydrogen bonds are denoted by red dashed lines.

Catechol has its own problem of bioavailability46–49 because it is prone to be oxidized.50 Nevertheless, catechol and catechol-containing flavonoids can be a starting point to design or screen inhibitors of TLPs. Phenolic compounds were traditionally considered not good drug candidates due to, e.g., their propensity to oxidation and their fast elimination through glucuronidation by detoxification enzymes.50 However, successful examples have shown the feasibility of modifying such phenolic compounds into applicable drugs. For example, Raltegravir, a hydroxynaphthyridine compound, has been successfully modified by chemical techniques to possess a much improved pharmacokinetic profile as a potent inhibitor of HIV integrase.51 We also identified Embelin as a new type inhibitor of PAI-1 and showed its efficacy in attenuating blood clot formation in an animal model with tail vein administration of the inhibitor.52

The catechol moiety is commonly found in natural flavonoids.53 The current study provides solid evidence to support the potential direct interaction of catechol-containing flavonoids with serine proteases and points out new functions of such flavonoids. We aligned the sequences of some common trypsin-like serine proteases (Fig. 5), the results show that many uPA residues involved in the interaction with quercetin are highly conserved in other trypsin-like serine proteases, such as Asp189, Gly219, Ser214 and Trp215, let alone the catalytic residues His57 and Ser195. Besides, Ser190 and Gln192 are also relatively conserved among some trypsin-like serine proteases. The S1 pockets of most trypsin-like serine proteases have a similar size, strengthening the possibility of consistent catechol binding mode to serine proteases. Therefore, the current work suggests that the catechols can most likely inhibit all these TLPs by the catechol moiety (Fig. 4B, we drew two parallel hydrogen bonds in this model which is possible considering the flexibility of the protein structure). For example, catechins, major tea components, may inhibit TLPs inside the human digestive tract, e.g., trypsin.54 This provides a potential molecular link to the commonly observed inhibitory effect of tea on digestion. Indeed, as mentioned above, quercetin was observed to inhibit many TLPs like uPA, plasmin, thrombin and trypsin. In order to achieve high specificity and potency to a given protease, catechols need to be derivatized to increase their contact surface area to the target protease, especially at the so-called exosites (e.g., S2 pocket) unique to the target protease.


image file: c6fo01825d-f5.tif
Fig. 5 Structure-based sequence alignments (chymotrypsinogen numbering) of uPA and other representative trypsin-like serine proteases. Some uPA residues involved in the interaction with quercetin that are highly conserved in trypsin-like serine proteases. Red triangles represent highly conserved residues: His57, Asp189, Ser195, Ser214, Trp215 and Gly219. Green triangles represent relatively conserved residues. The variable regions (37-, 60-loop and 97-loop) are shown in blue arrows, which can be used for protease inhibitors’ recognition to enhance targeting specificity. Purple pentagrams denote residues of the serine protease catalytic triad.

3.3. Physiological roles of endogenous catechol-containing compounds

Catechol is the most critical moiety of quercetin in inhibiting serine proteases. A list of endogenous substances contain the catechol moiety, including catecholamines (dopamine, epinephrine and norepinephrine) that are biochemically significant hormones/neurotransmitters.55 On the basis of our current study, it is likely that this class of compounds bind to endogenous serine proteases present in the blood circulation, including uPA, tissue-type plasminogen activator (tPA), plasmin, and activated blood coagulation factors. Such interaction can lead to a longer plasma half-life of the catecholamines by storing them in serine proteases for slow release and thus affects the action profile of the catecholamines. Are there sufficient amounts of endogenous serine proteases in blood circulation to buffer catecholamines? It appears that some serine proteases have high transient local concentrations attributable to the presence of a large amount of their precursors, e.g., ∼2 μM for plasminogen in human blood.56 Therefore, endogenous serine proteases and catechol-containing compounds may constitute an antagonistic system buffering their physiological concentrations, and such equilibrium can likely be intervened by exogenous flavonoids.

4. Conclusions

The crystal structure of quercetin:uPA solved in this work provided direct evidence for serine protease inhibition by flavonoids. We described for the first time the detailed interaction between quercetin and uPA and defined the molecular basis for quercetin as a uPA inhibitor. Structural analysis also showed that the phenolic ring of the flavonoids docks into the primary specificity pocket (S1 pocket) of serine proteases. This finding resolved the longstanding controversy on the interactions between flavonoids and serine proteases. In the crystal structure of quercetin:uPA, the non-alkaline catechol moiety of quercetin forms an electrostatic shielding hydrogen bond with the Asp189 of the S1 pocket, which indicates that flavonoids with neighboring phenolic hydroxyl groups (or catechol) can act as TLP inhibitors. We also pointed out that endogenous catechol-containing compounds might weakly combine to endogenous serine proteases for slow release, thus affecting their action profile. Overall, we revealed the molecular basis of serine protease inhibition by flavonoids, a new flavonoid function mechanism formerly under-recognized.

Author contributions

G. X. and L. G. have performed most of the experiments and prepared the draft; L. G. and Y. C. have participated in the experimental guidance; M. X. was involved in structural analysis; X. W. has assisted in preparing and operating experiments; M. H. has conceived and led the project and finalized the manuscript.

Conflict of interest

The authors have declared no conflicts of interest.

Abbreviations

uPAUrokinase-type plasminogen activator
SPDSerine protease domain
S1 pocketSpecific substrate binding pocket
TLPsTrypsin-like serine proteases.

Acknowledgements

We thank Shanghai Synchrotron Radiation Facility beamline BL17U for X-ray data collection. The work was supported by grants from NSFC (31370737, 31400637, 2017T010, 31570745, 31670739, and U1405229).

References

  1. N. C. Cook and S. Samman, J. Nutr. Biochem., 1996, 7, 66–76 CrossRef CAS.
  2. E. Middleton, Adv. Exp. Med. Biol., 1998, 439, 175–182 CrossRef CAS PubMed.
  3. M. D'Archivio, C. Filesi, R. Di Benedetto, R. Gargiulo, C. Giovannini and R. Masella, Ann. Ist. Super. Sanita, 2007, 43, 348–361 Search PubMed.
  4. E. Middleton, C. Kandaswami and T. C. Theoharides, Pharmacol. Rev., 2000, 52, 673 CAS.
  5. P. Knekt, J. Kumpulainen, R. Järvinen, H. Rissanen, M. Heliövaara, A. Reunanen, T. Hakulinen and A. Aromaa, Am. J. Clin. Nutr., 2002, 76, 560–568 CAS.
  6. H. Cao and X. Chen, Anti-cancer Agents Med. Chem., 2012, 12, 929–939 CrossRef CAS PubMed.
  7. S. Pasetto, V. Pardi and R. M. Murata, PLoS One, 2014, 9, e115323 Search PubMed.
  8. B. H. Havsteen, Pharmacol. Ther., 2002, 96, 67–202 CrossRef CAS PubMed.
  9. A. Tapas, D. Sakarkar and R. Kakde, Trop. J. Pharm. Res., 2008, 1, 132–143 CAS.
  10. R. J. Nijveldt, E. van Nood, D. E. van Hoorn, P. G. Boelens, K. van Norren and P. A. van Leeuwen, Am. J. Clin. Nutr., 2001, 74, 418–425 CAS.
  11. A. Ahmad, M. Kaleem, Z. Ahmed and H. Shafiq, Food Res. Int., 2015, 77(Part 2), 221–235 CrossRef CAS.
  12. L. R. F. de Sousa, H. Wu, L. Nebo, J. B. Fernandes, M. F. D. G. F. da Silva, W. Kiefer, M. Kanitz, J. Bodem, W. E. Diederich, T. Schirmeister and P. C. Vieira, Bioorg. Med. Chem., 2015, 23, 466–470 CrossRef CAS PubMed.
  13. M. Cuccioloni, M. Mozzicafreddo, L. Bonfili, V. Cecarini, A. M. Eleuteri and M. Angeletti, Chem. Biol. Drug Des., 2009, 74, 1–15 CAS.
  14. V. Vranova, K. Rejsek and P. Formanek, Appl. Soil Ecol., 2013, 70, 23–32 CrossRef.
  15. L. Liu, H. Ma, N. Yang, Y. Tang, J. Guo, W. Tao and J. A. Duan, Thromb. Res., 2010, 126, e365–e378 CrossRef CAS PubMed.
  16. M. Mozzicafreddo, M. Cuccioloni, L. Bonfili, A. M. Eleuteri, E. Fioretti and M. Angeletti, Biochim. Biophys. Acta, 2008, 1784, 995–1001 CrossRef CAS PubMed.
  17. L. Hedstrom, Chem. Rev., 2002, 102, 4501–4524 CrossRef CAS PubMed.
  18. J. Eberhardt, J. Comp. Physiol., B, 1992, 162, 159–167 CrossRef CAS.
  19. N. H. van Tilburg, F. R. Rosendaal and R. M. Bertina, Blood, 2000, 95, 2855 CAS.
  20. E. P. Navarese, M. Kolodziejczak, V. Schulze, P. A. Gurbel, U. Tantry, Y. Lin, M. Brockmeyer, D. E. Kandzari, J. M. Kubica, R. B. D'Agostino, J. Kubica, M. Volpe, S. Agewall, D. J. Kereiakes and M. Kelm, Ann. Intern. Med., 2015, 163, 40–51 CrossRef PubMed.
  21. S. Ohtomo, M. Nangaku, Y. Izuhara, N. Yamada, T. Dan, T. Mori, S. Ito, C. van Ypersele de Strihou and T. Miyata, Kidney Int., 2008, 74, 768–774 CrossRef CAS PubMed.
  22. K. Tryggvason, M. Höyhtyä and T. Salo, Biochim. Biophys. Acta, Rev. Cancer, 1987, 907, 191–217 CrossRef CAS.
  23. P. A. Andreasen, L. Kjøller, L. Christensen and M. J. Duffy, Int. J. Cancer, 1997, 72, 1–22 CrossRef CAS PubMed.
  24. G. Zhao, C. Yuan, T. Wind, Z. Huang, P. A. Andreasen and M. Huang, J. Struct. Biol., 2007, 160, 1–10 CrossRef CAS PubMed.
  25. M. Benvenuti and S. Mangani, Nat. Protocols, 2007, 2, 1633–1651 CAS.
  26. Q. S. Wang, F. Yu, S. Huang, B. Sun, K. H. Zhang, K. Liu, Z. J. Wang, C. Y. Xu, S. S. Wang, L. F. Yang, Q. Y. Pan, L. Li, H. Zhou, Y. Cui, Q. Xu, T. Earnest and J. H. He, Nucl. Sci. Tech., 2015, 26, 12–17 Search PubMed.
  27. Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276, 307–326 CAS.
  28. E. Krissinel and K. Henrick, J. Mol. Biol., 2007, 372, 774–797 CrossRef CAS PubMed.
  29. A. Vagin and A. Teplyakov, J. Appl. Crystallogr., 1997, 30, 1022–1025 CrossRef CAS.
  30. L. Jiang, K. A. Botkjaer, L. M. Andersen, C. Yuan, P. A. Andreasen and M. Huang, Biochem. J., 2013, 449, 161–166 CrossRef CAS PubMed.
  31. G. N. Murshudov, A. A. Vagin and E. J. Dodson, Acta Crystallogr., Sect. D: Biol. Crystallogr., 1997, 53, 240–255 CrossRef CAS PubMed.
  32. P. Emsley and K. Cowtan, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2004, 60, 2126–2132 CrossRef PubMed.
  33. R. A. Laskowski, M. W. Macarthur, D. S. Moss and J. M. Thornton, J. Appl. Crystallogr., 1993, 26, 283–291 CrossRef CAS.
  34. W. L. DeLano, The PyMOL Molecular Graphics System, Version 1.8, Schrödinger, LLC, Palo Alto, CA, USA, 2002, http://pymol.org/citing.
  35. S. A. Rabbani, A. P. Mazar, S. M. Bernier, M. Haq, I. Bolivar, J. Henkin and D. Goltzman, J. Biol. Chem., 1992, 267, 14151–14156 CAS.
  36. X. Li, A. M. Bokman, M. Llinás, R. A. G. Smith and C. M. Dobson, J. Mol. Biol., 1994, 235, 1548–1559 CrossRef CAS PubMed.
  37. G. Spraggon, C. Phillips, U. K. Nowak, C. P. Ponting, D. Saunders, C. M. Dobson, D. I. Stuart and E. Y. Jones, Structure, 1995, 3, 681–691 CrossRef CAS PubMed.
  38. J. Kraut, Annu. Rev. Biochem., 1977, 46, 331–358 CrossRef CAS PubMed.
  39. D. M. Blow, J. J. Birktoft and B. S. Hartley, Nature, 1969, 221, 337–340 CrossRef CAS PubMed.
  40. N. D. Rawlings, F. R. Morton, C. Y. Kok, J. Kong and A. J. Barrett, Nucleic Acids Res., 2008, 36, D320–D325 CrossRef CAS PubMed.
  41. B. A. Katz, K. Elrod, C. Luong, M. J. Rice, R. L. Mackman, P. A. Sprengeler, J. Spencer, J. Hataye, J. Janc, J. Link, J. Litvak, R. Rai, K. Rice, S. Sideris, E. Verner and W. Young, J. Mol. Biol., 2001, 307, 1451–1486 CrossRef CAS PubMed.
  42. A. R. Fersht, J. P. Shi, J. Knilljones, D. M. Lowe, A. J. Wilkinson, D. M. Blow, P. Brick, P. Carter, M. M. Y. Waye and G. Winter, Nature, 1985, 314, 235–238 CrossRef CAS PubMed.
  43. M. Huang, R. Syed, E. A. Stura, M. J. Stone, R. S. Stefanko, W. Ruf, T. S. Edgington and I. A. Wilson, J. Mol. Biol., 1998, 275, 873–894 CrossRef CAS PubMed.
  44. K. Menear, Curr. Med. Chem., 1998, 5, 457–468 CAS.
  45. G. Y. Liang and J. P. Bowen, Curr. Top. Med. Chem., 2016, 16, 1506–1529 CrossRef CAS PubMed.
  46. C. Felgines, O. Texier, C. Besson, D. Fraisse, J. L. Lamaison and C. Remesy, J. Nutr., 2002, 132, 1249–1253 CAS.
  47. T. K. McGhie and M. C. Walton, Mol. Nutr. Food Res., 2007, 51, 702–713 CAS.
  48. K. Nemeth and M. K. Piskula, Crit. Rev. Food Sci. Nutr., 2007, 47, 397–409 CrossRef CAS PubMed.
  49. I. L. Nielsen and G. Williamson, Nutr. Cancer, 2007, 57, 1–10 CrossRef CAS PubMed.
  50. W. B. Jakoby and D. M. Ziegler, J. Biol. Chem., 1990, 265, 20715–20718 CAS.
  51. V. Summa, A. Petrocchi, F. Bonelli, B. Crescenzi, M. Donghi, M. Ferrara, F. Fiore, C. Gardelli, O. Gonzalez Paz, D. J. Hazuda, P. Jones, O. Kinzel, R. Laufer, E. Monteagudo, E. Muraglia, E. Nizi, F. Orvieto, P. Pace, G. Pescatore, R. Scarpelli, K. Stillmock, M. V. Witmer and M. Rowley, J. Med. Chem., 2008, 51, 5843–5855 CrossRef CAS PubMed.
  52. Z. H. Lin, J. K. Jensen, Z. B. Hong, X. L. Shi, L. H. Hu, P. A. Andreasen and M. D. Huang, Chem. Biol., 2013, 20, 253–261 CrossRef CAS PubMed.
  53. J. Frank, A. Budek, T. Lundh, R. S. Parker, J. E. Swanson, C. F. Lourenco, B. Gago, J. Laranjinha, B. Vessby and A. Kamal-Eldin, J. Lipid Res., 2006, 47, 2718–2725 CrossRef CAS PubMed.
  54. R. J. Green, A. S. Murphy, B. Schulz, B. A. Watkins and M. G. Ferruzzi, Mol. Nutr. Food Res., 2007, 51, 1152–1162 CAS.
  55. M. Esler, G. Jennings, G. Lambert, I. Meredith, M. Horne and G. Eisenhofer, Physiol. Rev., 1990, 70, 963–985 CAS.
  56. E. F. Plow, T. Herren, A. Redlitz, L. A. Miles and J. L. Hoover-Plow, FASEB J., 1995, 9, 939–945 CAS.

This journal is © The Royal Society of Chemistry 2017