Biochimie 121 (2016) 179e188

Contents lists available at ScienceDirect

Biochimie
journal homepage: www.elsevier.com/locate/biochi

Research paper

A novel synthetic quinolinone inhibitor presents proteolytic and

hemorrhagic inhibitory activities against snake venom
metalloproteases
Patrícia T. Baraldi a, 1, Angelo J. Magro b, c, d, **, 1, Fa

bio F. Matioli d, Silvana Marcussi e,
d f, g
Ney Lemke , Leonardo A. Calderon , Rodrigo G. Sta
beli f, g, Andreimar M. Soares f, g,
a c, d, *
Arlene G. Correa , Marcos R.M. Fontes
a
Departamento de Química, Universidade Federal de Sa ~o Carlos (UFSCar), Sa
~o Carlos, SP, Brazil
b
Departamento de Bioprocessos e Biotecnologia, Faculdade de Ci^ encias Agra
rias, Universidade Estadual Paulista (UNESP), Botucatu, SP, Brazil
c
Instituto de Biotecnologia, Universidade Estadual Paulista (UNESP), Botucatu, SP, Brazil
d
Departamento de Física e Biofísica, Instituto de Bioci^
encias, Universidade Estadual Paulista (UNESP), Botucatu, SP, Brazil
e
Departamento de Química, Universidade Federal de Lavras (UFLA), Lavras, MG, Brazil
f
Centro de Estudos de Biomol
eculas Aplicadas a Saúde (CEBio), Fundaça~o Oswaldo Cruz (FIOCRUZ), unidade Fiocruz Rondo ^nia, Porto Velho, RO, Brazil
g
Departamento de Medicina, Universidade Federal de Rondo ^nia (UNIR), Porto Velho, RO, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Metalloproteases play a fundamental role in snake venom envenomation inducing hemorrhagic, fibrige-
Received 13 August 2015 n(ogen)olytic and myotoxic effects in their victims. Several snake venoms, such as those from the Bothrops
Accepted 24 November 2015 genus, present important local effects which are not efficiently neutralized by conventional serum therapy.
Available online 14 December 2015
Consequently, these accidents may result in permanent sequelae and disability, creating economic and
social problems, especially in developing countries, leading the attention of the World Health Organization
Keywords:
that considered ophidic envenomations a neglected tropical disease. Aiming to produce an efficient in-
Antihemorrhagic effects
hibitor against bothropic venoms, we synthesized different molecules classified as quinolinones e a group
Snake venom
Antiophidic drugs
of low-toxic chemical compounds widely used as antibacterial and antimycobacterial drugs e and tested
Quinolinones their inhibitory properties against hemorrhage caused by bothropic venoms. The results from this initial
Metalloproteases screening indicated the molecule 2-hydroxymethyl-6-methoxy-1,4-dihydro-4-quinolinone (Q8) was the
Bioinformatics studies most effective antihemorrhagic compound among all of the assayed synthetic quinolinones. Other in vitro
and in vivo experiments showed this novel compound was able to inhibit significantly the hemorrhagic
and/or proteolytic activities of bothropic crude venoms and isolated snake venom metalloproteases
(SVMPs) even at lower concentrations. Docking and molecular dynamic simulations were also performed to
get insights into the structural basis of Q8 inhibitory mechanism against proteolytic and hemorrhagic
SVMPs. These structural studies demonstrated that Q8 may form a stable complex with SVMPs, impairing
the access of substrates to the active sites of these toxins. Therefore, both experimental and structural data
indicate that Q8 compound is an interesting candidate for antiophidic therapy, particularly for the treat-
ment of the hemorrhagic and necrotic effects induced by bothropic venoms.
© 2015 Elsevier B.V. and Socie
te

Française de Biochimie et Biologie Mole
culaire (SFBBM). All rights
reserved.

* Corresponding author. Departamento de Física e Biofísica, Instituto de Bio- 1. Introduction

^ncias de Botucatu, UNESP e Campus de Botucatu, Distrito de Rubia
cie ~o Júnior, S/N,
CEP:18618-970, Botucatu, SP, Brazil.
** Corresponding author. Departamento de Bioprocessos e Biotecnologia, Facul- Snake venoms are complex mixtures of several proteins
dade de Cie ^ncias Agrono
^micas (FCA), UNESP e Campus de Botucatu, Fazenda including phospholipases A2, phospholipases A2-like, metal-
Experimental Lageado, Rua Jose
Barbosa de Barros, 1780, CEP:18.610-307, Botucatu, loproteases, serineproteases, L-amino oxidases, disintegrins, C- type
SP, Brazil.
lectins and others. Their combination are able to unleash a complex
E-mail addresses: [email protected] (A.J. Magro), [email protected]
(M.R.M. Fontes).
series of events responsible for the pathophysiology of snake en-
1
These authors contributed equally to the work. venomation and the consequent emergence of deleterious

http://dx.doi.org/10.1016/j.biochi.2015.11.031
0300-9084/© 2015 Elsevier B.V. and Socie
te

Française de Biochimie et Biologie Mole

culaire (SFBBM). All rights reserved.
180 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188

pharmacological/biological effects like neurotoxicity, myotoxicity, hemorrhage. Therefore, the structural basis related to the harmful
edema-inducing activity, hemorrhage, coagulation and several activities of the SVMP catalytic domains is a important topic for the
other properties [1
5]. identification of potential compounds for snake envenomation
Recent proteomic analyses indicated that snake venom metal- treatment.
loproteases (SVMPs), which are composed by a group of zinc-
dependent enzymes, are the most representative toxins present 2. Materials and methods
in viperid species [1e9] and also are found in Colubridae [10e12];
Elapidae [13e16] and Atractaspididae [17] families. According to 2.1. Materials and chemicals
the revised classification of SVMPs by Fox and Serrano [18,19], these
molecules are distributed in three different classes, depending on All commercially available reagents were purchased from
their molecular weight and domain structure. i) small SVMPs Aldrich Chemical Co. Reagents® and solvents were purified when
(20e30 kDa) with only a single catalytic domain in the P-I class; ii) necessary according to the usual procedures described in the
medium SVMPs (30e60 kDa) composed by catalytic and literature. B. jararacussu, B. neuwiedi, B. moojeni and B. alternatus
disintegrin-like domains in the P-II class; and iii) large SVMPs venoms were purchased from Bioagents Serpentarium Ltda.
(60e100 kDa) with catalytic, disintegrin-like and cysteine-rich ~o Paulo State, Brazil). The SVMPs (P-III class BjussuMP-I,
(Batatais, Sa
domains in the P-III class (the large SVMPs with an extra lectin- P-I class BjussuMP-II and P-I class neuwiedase) were isolated and
like domain are also included in the P-III class). biochemically characterized as previously described [39e41]. The
Initially, the main role attributed to SVMPs was related to the licenses for scientific purposes are from Instituto Brasileiro do Meio
remarkable hemorrhagic activity presented by many of these Ambiente e dos Recursos Naturais Renova
veis e IBAMA and Instituto
molecules (particularly the large SVMPs). This biological property is Chico Mendes de Conservaça ~o da Biodiversidade e ICMBio. Numbers:
clearly linked to their ability to interact with specific receptors on 11094-2, 11094-1, 10394-1 and 15484-1.
endothelial cells [20,21] and fibroblasts [22] and is also involved in The infrared spectra were measured on a Bomem M102® spec-
the degradation of extracellular matrix components [23]. However, trometer (4000 - 400 cm
1). NMR data assignment was based on
a wide spectrum of other SVMP functions have been described in 1
H and 13C and the spectra were recorded on a Bruker DRX-200®
the last years, comprising very distinct capabilities as fibrin(ogen) and ARX-400®. Mass spectral analyses were carried out with a
olytic activity, inactivation of blood serine proteinase inhibitors, Shimadzu GCMS-QP5000® spectrophotometer. Analytical thin-
prothrombin and blood coagulation factor X activation, platelet layer chromatography was performed on a 0.25 mm film of silica
aggregation inhibition, apoptosis induction, and pro-inflammatory gel containing a fluorescent indicator UV254 supported on
action [24]. All of these activities presented by SVMPs contribute to aluminum sheet (SigmaeAldrich®). Flash column chromatography
the systemic actions observed after envenomation, but the influ- was executed using a silica gel (Kieselgel 60, 230e400 mesh, E.
ence of these toxins on the local effects is especially critical, since Merck®). Gas chromatography was performed in a Shimadzu GC-
they contribute strongly to severe tissue necrosis resulting from 17A® chromatograph equipped with a DB-5 column, employing H2
their hemorrhagic and proteolytic activities [25]. Hence, as serum as a carrier. Elemental analyses and the melting points were per-
therapy is not effective enough to avoid the occurrence of local formed, respectively, with Fisons EA1108 CHNS-O® analyzer and
effects caused by viperid venoms, it is quite interesting to search for FISATOM 430® equipment. Melting points were determined using a
new molecules which could be also used in anti-snake venom Microquimica MQAPF-301® device.
treatments.
In many countries, including Brazil, plant extracts are tradi- 2.2. Compounds and procedures employed for 2-hydroxymethyl-6-
tionally used for the treatment of snakebite envenomations, methoxy-1,4-dihydro-4-quinolinone (Q8) synthesis
although there is scientific validation only in a few cases [26e30].
On the other hand, several isolated plant alkaloids have been 2.2.1. General procedure for the synthesis of enamines
described as in vitro and in vivo anti-snake venom inhibitors In a flask under N2 at 55  C containing aniline (5.0 g, 0.054 mol)
[30e35]. Thus, we synthesized different quinolinone molecules e or p-anisidine (5.0 g, 0.041 mol) in dry methanol (1 mL/mmol an-
low-toxic chemical compounds [36] which are already widely used iline), DMAD (1 eq.) was added and the mixture was stirred over-
as antibacterial and antimycobacterial drugs [37,38] e and tested night. Then methanol was removed under reduced pressure, and
their inhibitory properties against hemorrhage caused by Bothrops CH2Cl2 (30 mL) was added followed by a saturated solution of
jararacussu, Bothrops moojeni and Bothrops alternatus venoms and NH4Cl (3  5 mL). The organic layer was dried over anhydrous
the isolated P-III class SVMP BjussuMP-I from B. jararacussu. The MgSO4, and evaporated to dryness under reduced pressure. The
results obtained from this initial screening pointed out the most residue was purified on a silica gel column, using hexane:ethyl
effective antihemorrhagic molecule was the compound 2- acetate 9.5:0.5 as eluent to afford desired enamine compounds 3 or
hydroxymethyl-6-methoxy-1,4-dihydro-4-quinolinone (identified 4 in 50% yield.
herein as Q8). Subsequently, other in vitro and in vivo experiments
showed the novel Q8 compound was able to significantly inhibit, 2.2.1.1. 2-Phenylamino-2-enediocic acid dimethyl ester (compound
even at lower concentrations, the hemorrhagic and/or proteolytic 3). 6.3 g. IR (nmax, film, cm
1): 3457; 3380; 2953; 1739; 1668; 1282;
activities of four bothropic crude venoms (B. jararacussu, Bothrops 1031. 1H NMR (200 MHz, CDCl3): d 9.67 (sl, 1H); 7.32e7.25 (m, 2H);
neuwiedi, B. moojeni and B. alternatus) and two isolated SVMPs (P-III 7.13e7.05 (m, 1H); 6.9 (m, 2H); 5.39 (s, 1H); 3.74 (s; 3H); 3.70 (s, 3H).
13
class BjussuMP-I from B. jararacussu and P-I class neuwiedase from C NMR (50 MHz, CDCl3) d: 169.8; 164.7; 147.9; 140.2; 129.0; 124.1;
B. neuwiedi). 120.6; 116.6; 93.5; 52.6; 51.1. MS (% rel. intensity) m/z: 235.25
Additionally, docking and molecular dynamics simulations (6.9%); 144.15 (93.3%); 77.10 (100%).
involving BjussuMP-II, a P-I SVMP from B. jararacussu venom [39]
with a very well characterized proteolytic activity, and Q8 com- 2.2.1.2. 2-(4-Methoxy-phenylamino)-2-enediocic acid dimethyl ester
pound were performed to get some insights into the inhibitory (compound 4). 5.5 g. IR (nmax, film, cm
1): 3284; 3210; 2952; 2836;
activity of this novel 4-quinolinone molecule against the action of 1742; 1637; 1033. 1H NMR (200 MHz, CDCl3): d. 9.57 (sl, 1H);
SVMPs. These theoretical studies are justified since proteolysis is a 6.91e679 (m, 4H); 5.30 (s, 1H); 3.78 (s, 3H); 3.73 (s, 3H); 3.67 (s,
common feature of SVMPs related to local tissue damage and 3H). 13C NMR (50 MHz, CDCl3) d: 182.4; 170.0; 164.8; 156.9; 149.0;
 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188 181

133.4; 123.0; 114.3; 91.7; 55.4; 52.6; 21.0. MS (% rel. intensity) m/z: a gradient of n-hexane e methanol as eluent furnishing products 9
265.50 (12.6%); 146.15 (74.6%); 77.15 (96.5%). and 10 in 80% yield.

2.2.2. General procedure for the synthesis of 4-quinolinone 2.2.4.1. 4-ethoxy-6-methoxy-2-quinoline methyl carboxylate (com-

compounds 5 and 6 pound 9). 17 mg mp: 105e107 C. IR (nmax, KBr) cm
1: 2933; 2856;
1
A flask adapted with a condenser containing diphenyl ether 1730; 1639; 1483; 1236; 1024. H NMR (400 MHz, CDCl3) d: 8.11 (d,
(8 mL) was heated in a sand bath up to the reflux temperature and J ¼ 9.2 Hz, 1H); 7.55 (s, 1H); 7.46 (d, J ¼ 2.8 Hz, 1H); 7.48 (dd, J ¼ 9.2,
then enamine compounds 3 or 4 (1.0 g) were added. The mixture 2.8 Hz, 2H); 4.36 (q, J ¼ 7 Hz, 2H); 4.06 (s, 3H); 3.96 (s, 3H); 1.60 (t,
was stirred for 10 min and transferred to an ice bath. The residue J ¼ 7 Hz, 3H). 13C NMR (100 MHz, CDCl3) d: 166.46; 161.29; 158.91;
was purified by dry flash chromatography using a hexane e 146.60; 144.42; 131.82; 123.42; 123.02; 101.01; 99.61; 64.53; 55.64;
methylene chloride e MeOH gradient to afford the desired 4- 53.08; 14.48.
quinolinone compounds 5 or 6 in 70% yield.
2.2.4.2. 4-ethoxy-6-methoxy-2-quinolilmethanol (compound 10).
2.2.2.1. Methyl-4-oxo-1,4-dihydroquinolinone 2-carboxylate (com- mp: 120e123  C. IR (nmax, KBr) cm
1: 2947; 2866; 1643; 1592;
pound 5). 0.6 g mp: 215e217  C. IR (nmax, film, cm
1): 3436; 2925; 1382; 1031; 825. 1H NMR (400 MHz, CDCl3) d: 7.77 (d, J ¼ 9.14 Hz,
2886; 1733; 1639; 1538; 1033. 1H NMR (200 MHz, CDCl3): 12.03 (s, 1H); 7.42 (d, J ¼ 2.7 Hz, 1H); 7.32 (dd, J ¼ 9.2, 2.9 Hz, 1H); 7.55 (s,
1H); 8.13 (d, J ¼ 8.0 Hz, 1H); 7.95 (d, J ¼ 10 Hz, 1H); 7.66e7.63 (m, 1H); 4.75 (s, 2H); 4.31 (q, J ¼ 7.0 Hz, 2H); 3.88 (s, 3H); 1.55 (t,
1H); 7.35e7.32 (m, 1H); 6.73 (s; 1H); 3.99 (s; 3H). 13C NMR (50 MHz, J ¼ 7 Hz, 3H). 13C NMR (100 MHz, CDCl3) d: 162.11; 161.19; 157.41;
CDCl3) d: 176.4; 161.1; 138.4; 135.8; 130.6; 124.4; 123.0; 122.1; 144.63; 130.72; 122.52; 121.52; 100.73; 99.72; 65.86; 64.91; 56.35;
117.9; 108.8; 51.5. MS (% rel. intensity) m/z: 203 (30.1%); 143 (100%); 15.33.
115.15 (72.9%); 89.15 (66.2%).
2.2.4.3. 4-ethoxy-2-ethoxymethyl-6-methoxyquinoline (compound
2.2.2.2. Methyl-6-methoxy-4-oxo-1,4-dihydro-quinolinone carbox- 11). In a suspension of NaH 60% (3 mg, 0.077 mmol) in dry DMF
ylate (compound 6). 0.53 g mp: 250e253  C. IR (nmax, film, cm
1): (100 mL) and DME (500 mL) was added a solution of compound 8
3440; 2935; 2865; 1729; 1639; 1552; 1024. 1H NMR (200 MHz, (15 mg, 0.064 mmol) in DME (100 mL) at 0  C. After 10 min the
CDCl3): d 12.14 (s, 1H); 7.92 (d, J ¼ 9.1 Hz, 1H); 77.46 (d, J ¼ 2.8 Hz, mixture was treated with LiBr (12 mg, 0.128 mmol) and stirred by
1H); 7.37 (dd, J ¼ 9.0, 2.8 Hz, 1H); 6.67 (s; 1H); 3.96 (s; 3H); 3.85 (s; 15 min. Then ethyl bromide was added (8.3 mg, 0.077 mmol)
3H). 13C NMR (50 MHz, CDCl3) d: 176.3; 162.3; 155.8; 136.2; 134.2; dropwise. The mixture was stirred at room temperature by 3 h;
126.7; 122.8; 120.9; 108.3; 103.2; 54.9; 52.9. MS (% rel. intensity) m/ after ice addition, it was extracted with ethyl acetate (3  10 mL)
z: 207.50 (54.5%); 173.2 (91%); 73.2 (100%). and the combined organic phases where washed with brine
(3  5 mL). The resulting solution was dried with anhydrous so-
2.2.3. General procedure for the synthesis of 4-quinolinone dium sulfate and the solvent was evaporated. The crude material
compounds 7 and 8 was purified using column chromatography in silica gel with hex-
In a flask containing 4-quinolinone compounds 5 or 6 (0.50 g) in ane:ethyl acetate 6:4 as eluent. Compound 11 was obtained in 80%
dry THF (10 mL) under N2 at 0  C, BH3.SMe2 (1 eq) was added drop yield (13.3 mg). mp: 69e72  C. IR (nmax, KBr) cm
1: 2975; 2931;
wise. The mixture was allowed to warm to room temperature 2886; 1704; 1596; 1224; 1093; 831. 1H NMR (400 MHz, CDCl3) d:
overnight. Then dry methanol (3  10 mL) was added, the solvent 7.88 (d, J ¼ 9.1 Hz, 1H); 7.44 (d, J ¼ 2.8 Hz, 1H); 7.32 (dd, J ¼ 9.1,
was evaporated under atmospheric pressure and the crude product 2.8 Hz, 1H); 6.94 (s, 1H); 4.71 (s, 2H); 4.30 (q, J ¼ 7.0 Hz, 2H); 3.93 (s,
was recrystallized to furnish 4-quinolinone compounds 7 or 8 in 3H); 3.64 (q, J ¼ 7 Hz, 2H), 1.57 (t, J ¼ 7 Hz, 3H), 1.29 (t, J ¼ 7 Hz, 3H).
13
70% yield. C NMR (100 MHz, CDCl3) d: 161.27; 158.11; 157.12; 144.39; 129.92;
121.97; 99.98; 98.82; 74.39; 66.35; 64.09; 55.56; 15.28; 14.55.
2.2.3.1. 2-Hydroxymethyl-1,4-dihydro-4-quinolinone (compound 7).
0.3 g mp: 230e232  C. IR (nmax, film, cm
1): 3384; 2946; 2917; 2.3. Biological assays
1619; 1359; 1083. 1H NMR (200 MHz, CDCl3): d. 8.04 (dd, J ¼ 8, 1 Hz,
1H); 7.69e7.56 (m, 2H); 7.31e7.23 (m, 1H); 6.02 (s, 1H); 4.48 (s, Animal procedures were in accordance with the guidelines for
2H).13C NMR (50 MHz, CDCl3) d: 153.5; 140.4; 131.6; 125.3; 125.0; animal care prepared by the Committee on Care and Use of Labo-
122.9; 118.6; 105.6; 60.5. Anal. Calcd for C10H9NO2:C, 68.56%; H, ratory Animal Resources, National Research Council, USA. Animal
5.18%; N, 8.0%. Found: C, 68.60%; H, 5.18%; N, 7.90%. care was in accordance with the guidelines of the Brazilian College
for Animal Experimentation (COBEA) and was approved by Com-
2.2.3.2. 2-Hydroxymethyl-6-methoxy-1,4-dihydro-4-quinolinone mittee for Ethics in Animals Utilization of USP/Campus Ribeira ~o
(compound 8 or Q8). 0.3 g mp: 250e252  C. IR (nmax, film, cm
1): Preto-SP (Process n 07.1.517.53.6).
3448; 2921; 2852; 1637; 1504; 1035. 1H NMR (200 MHz, CDCl3):
d 7.94 (dd, J ¼ 8.4.1 Hz, 1H); 7.52e7.41 (m, 2H); 6.63 (s, 1H); 4.72 (s, 2.3.1. Hemorrhagic activity
2H); 3.89 (s, 3H).13C NMR (50 MHz, CDCl3) d: 171.5; 156.8; 155.5; 50 mg of desiccated crude venoms (B. jararacussu, B. moojeni and
134.5; 124.2; 123.2; 121.0; 103.1; 102.7; 60.0; 55.6. Anal. Calcd. for B. alternatus) and freeze-dried SVMP BjussuMP-I from B. jararacussu
C11H11NO3: C, 64.38%; H, 5.4%; N, 6.83%. Found: C, 64.41%; H, 5.55%; were weighed and dissolved in phosphate-buffered saline (PBS, pH
N, 6.81%. 7.2), while the synthetic compounds, named 5 through 11 were
dissolved in DMSO, including the compound of interest, Q8. For the
2.2.4. General procedure for the synthesis of quinolines 9, 10 and 11 initial screening, Swiss male mice (18e22 g, n ¼ 6) were injected
To a flask containing compound 6 or 8 (0.1 mmol) and K2CO3 intradermically (50 mL) in the back region with different pre-
(20.7 mg, 0.15 mmol) it was added anhydrous DMF (50 mL) and incubated (30 min; 37  C) doses containing different crude
ethyl bromide (9.5 mL, 0.15 mmol). The flask was stirred for 12 h at venoms/SVMPs:5 through 11 compounds at a 1:10 proportion (w/
room temperature. The solution was filtered in silica to remove the w). Control mice received PBS or DMSO. Three hours after injection,
precipitate and the solvent was evaporated, then the resulting the mice were sacrificed and the diameter of the hemorrhagic spot
material was purified by column chromatography in silica gel using on the back skin was measured [39,40].
182 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188

After selecting the compound Q8, a hemorrhagic activity inhi- algorithm. All docking simulations between the native BjussuMP-II
bition test was performed, (i) one with pre-incubation (30 min; model and the ligand were executed by the program Gold v.5.0.1
37  C) of the venoms/metalloproteases with different doses of the (CCDC Software Limited, Cambridge, U.K.) [49]. The docking site
compound Q8 (1:5, 1:10, 1:15 and 1:30, w/w); and (ii) a post- was defined within a 20 Å radius around the cofactor Zn2þ localized
envenoming test, where first the mice only received intradermal at the protein active site. The docking simulations were performed
injections containing BjussuMP-I (50 mg/50 mL), and after 15 min, with the options on/fix, on/spin, toggle/fix and toggle/spin to
the same animals were inoculated by the same route with the evaluate the influence of the Glu142/Zn2þ-coordinated catalytic
compound Q8 in the proportions 1:30 and 1:100 (SVMP: Q8, w/w). water molecule on the docking solutions. A minimum of ten rounds
The assessment of the hemorrhagic activity was conducted in the were executed for each simulation, with the generation of twenty
same manner and three hours after injection, mice were killed and docking solutions per round; the other docking parameters were
the diameter of the hemorrhage zone in the skin was measured. defined according to the GOLD v.5.0.1 default settings. The docking
solutions between Q8 and the native BjussuMP-II model were
2.3.2. Proteolytic activity on casein scored using the GoldScore fitness functions [49].
40 mg of crude snake venoms (B. jararacussu and B. neuwiedi) or
SVMPs BjussuMP-I from B. jararacussu and neuwiedase from 2.4.3. Molecular dynamics simulations
B. neuwiedi previously incubated with Q8 at different w/w ratios The BjussuMP-II/Q8 docking solutions which presented the
(1:5, 1:10, 1:15 and 1:30), were incubated with 1.0 mL of 1% (w/v) highest GoldScore values were submitted to molecular dynamics
casein in 0.1 M TriseHCl buffer (pH 8.0) for 30 min at 37  C. The (MD) simulations using the program GROMACS (Groningen Ma-
reaction was blocked by the addition of 1.0 mL of 5% (v/v) tri- chine for Chemical Simulation) v.4.5.4 [50,51]. The GROMOS 96
chloroacetic acid solution and the mixture remained at room 53a6 force field [52] was chosen to perform the MD simulations
temperature for 30 min before centrifugation (2000  g) for and the protonation states of the charged groups were set to pH 7.0.
5 min at 25  C. Proteolytic activity was estimated by monitoring the All the MD simulations were executed in the presence of explicit
absorbance of the clear supernatant at 280 nm [39,40]. water molecules [53] and the minimum allowed distance between
any atom of the models and the box wall was set to 1.0 nm. An
2.3.3. Statistical analysis energy minimization (EM) using a steepest descent algorithm was
Results are presented as the mean value ± SD obtained with the performed to generate the starting configuration of the systems.
indicated number of tested animals. The statistical significance of After this step, 200 ps of MD simulation with position restraints
differences between groups was evaluated using Student's un- applied to the protein (PRMD) was executed in order to relax the
paired t-test. A P value < 0.05 was considered to indicate systems gently. Then, 15 ns of unrestrained MD simulation were
significance. calculated to evaluate the stability of the structures. All MD simu-
lations were carried out in a periodic truncated cubic box under
2.4. Computational procedures constant temperature (298 K) and pressure (1.0 bar), which were
held by coupling to an isotropic pressure and external heat bath
2.4.1. BjussuMP-II modeling [54]. The distances between the catalytic histidines and the Zn2þ
According to the alignment data from the based-threading ion of the native BjussuMP-II model and the BjussuMP-II/Q8 com-
method program HHpred [42], available at the Max-Planck Insti- plexes were kept according to Andreini et al. [55]. Q8 topology and
tute for Developmental Biology server (http://toolkit.tuebingen. coordinates files used for molecular dynamics (MD) simulations
mpg.de/hhpred), a 1.05 Å resolution crystallographic model of the were generated by the Dundee PRODRG2 Server (http://davapc1.
zinc metalloproteinase BaP1 from Bothrops asper snake venom bioch.dundee.ac.uk/prodrg/).
complexed to a peptidomimetic inhibitor [43] (PDB code 2w15)
was selected as a template for the construction of the native P-I 2.4.4. Evaluation of the theoretical structural models
class BjussuMP-II structural model. This template, selected in the Overall stereochemical and energy quality of the native
Protein Data Base (PDB) using the algorithm BlastP (default pa- BjussuMP-II model and the BjussuMP-II/Q8 complex after their
rameters), was chosen based on the alignment score (337.81) and respective MD simulations were checked with the programs
its identity (81%) to the BaP1 sequence. The program Modeller 9v10 RAMPAGE [46] and ProSA-web [47]. Additionally, in order to assess
[44] was then used to generate ten structural models based on the the quality of the native BjussuMP-II model and BjussuMP-II/Q8
selected template, keeping the original position of the cofactor complex after their respective MD simulations, the average rmsd
Zn2þ. Additionally, the 1.93 Å resolution crystallographic model of (root mean square deviation)/time graph of the protein backbone
the native zinc metalloproteinase BaP1 [45] from Bothrops asper atoms were analyzed in terms of the difference between the total
(PDB code 1nd1) was also added to the Modeller 9v10 alignment averages of two equal sets of points (the transient part of the MD
input to define the position of the Glu42/Zn2þ-coordinated catalytic simulations was not considered).
water molecule in the generated models. Variable target function
method (VTFM) with conjugate gradients (CG) [44] and molecular 3. Results and discussion
dynamics (MD) [44] with simulated annealing (SA) [44] were used
in order to refine the models. The Zn2þ cofactor of each model was 3.1. Synthesis of 2-hydroxymethyl-6-methoxy-1,4-dihydro-4-
added based on the coordinates of this ion in the BaP1/peptido- quinolinone (Q8)
mimetic inhibitor crystallographic model. The best BjussuMP-II
model was selected according to the stereochemical and energy The synthetic route to obtain 2-hydroxymethyl-6-methoxy-1,4-
parameters determined respectively by the programs RAMPAGE dihydro-4-quinolinone (Q8) from a commercially available anisi-
[46] and ProSA-web [47]. line in three steps is shown in Fig. 1. The first step involves an
enamine bond formation, which was achieved in 80% yield through
2.4.2. Q8 in silico design and docking simulations condensation of dimethyl acetylenodicarboxylate (DMAD) with the
The program Avogadro v.0.9.4 [48] was used to design Q8 and anisiline compound 2, as described by Edmont [56]. Intramolecular
improve its overall structure by an energy minimization process cyclization to obtain the compound 4 was performed under reflux
based on the MMF94 force field and in a steepest-descent in different conditions and the best results were obtained with 70%
 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188 183

Fig. 1. Synthetic route to obtain 2-Hydroxymethyl-6-methoxy-1,4-dihydro-4-quinolinone (Q8) (1).

yield. Initially, the solvent diphenylether (DFE) was heated to reflux 3.2. Biochemical/pharmacological trials with Q8 demonstrated its
and afterwards enamine compound 3 was added and refluxed high inhibitory efficacy against crude snake venoms and SVMPs
during 10 min. In order to completely stop the reaction and
avoiding decomposition of the desired product, the flask was Fig. 2 shows the initial screening of quinolinones (compounds
immediately transferred to an ice bath. The purification of quino- Q5eQ13) against crude bothropic venoms and BjussuMP-I at 1:10
linone compound 4 was achieved using dry flash with a hexane/ w/w proportions (crude bothropic venoms/BjussuMP-
methylene chloride/methanol solvent gradient. The chemo- I:quinolinone). These assays showed clearly that the molecule Q8
selective reduction of quinolinone compound 4 was carried out is the most effective hemorrhagic inhibitor compared to the other
with BH3.SMe2, furnishing compound 8 or Q8 in good yield (~70%). quinolinones. Taking into account the remarkable inhibitory ac-
Infrared and NMR data assays were used to confirm the identity of tivity of the Q8 molecule, a further in vivo hemorrhagic experiment
the Q8 compound (section 2.2.3). was performed using pre-incubated mixtures containing different
crude bothropic venoms/BjussuMP-I:Q8 w/w proportions (Fig. 3A).
In this experiment, the crude snake venoms and BjussuMP-I which
were not pre-incubated with Q8 induced hemorrhagic spots vary-
ing from 8 to 13 mm of diameter in the back skin of the mice. In
contrast, after pre-incubation with the quinolinone compound, the
hemorrhagic spot in the back skin of the mice was clearly reduced
even considering the 1:5 w/w proportion (crude venom/BjussuMP-
I:Q8). At the minimum inhibitor proportion (1:5 w/w), the hem-
orrhagic spot diameter was, on average, around 6 mm (25%
reduction), 7 mm (42% reduction), 10 mm (23% reduction) and
5 mm (45% reduction) respectively for B. jararacussu, B. moojeni,
B. alternatus crude venoms and BjussuMP-I from B. jararacussu. The
increase in the amount of Q8 in relation to crude venoms or
BjussuMP-I showed a progressive inhibition potency, culminating
in a reduction of approximately 81.25% (~1 mm diameter), 87.5%
(~1 mm diameter), 76.9% (~3 mm diameter) and 77.8% (~2 mm
diameter) in the diameter of induced hemorrhagic spots, respec-
tively, by B. jararacussu, B. moojeni and B. alternatus crude venoms
Fig. 2. Screening all quinolinones versus crude venoms or BjussuMP-I. Effect quino- and BjussuMP-I after pre-incubation with the maximum Q8 pro-
linones (compounds 5e11) on the hemorrhage induced by Bothrops snake venoms
portion (1:30 w/w). Additionally, a post-envenoming experiment
(B. moojeni, B. jararacussu and B. alternatus) and BjussuMP-I, a SVMP from B. jararacussu
venom. confirmed the compound Q8 was also able to inhibit the formation
184 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188

Fig. 3. (A) Effect of 2-Hydroxymethyl-6-methoxy-1,4-dihydro-4-quinolinone (Q8) on the hemorrhage induced by Bothrops snake venoms (B. moojeni, B. jararacussu and
B. alternatus) and BjussuMP-I, a SVMP from B. jararacussu venom. The hemorrhagic spots on the back skin of mice were measured 3 h after intrademic injection of the crude venoms
or SVMP pre-incubated with the Q8 (30 min, 37  C) at different crude venom/BjussuMP-I:Q8 w/w ratios (1:5, 1:10, 1:15 and 1:30). In the control experiments (shown at left of the
graph), mice received PBS or DMSO. Results are expressed as mean ± S.D. (n ¼ 6). (B) Effect of Q8 on the proteolytic activity on casein induced by Bothrops snake venoms
(B. jararacussu and B. neuwiedi) and isolated SVMPs BjussuMP-I and neuwiedase from B. jararacussu and B. neuwiedi venoms, respectively. Different crude venom/SVMP: Q8 w/w
ratios (1:5, 1:10, 1:15 and 1:30) were tested to assay the Q8 inhibitory activity using a 1.0 mL 1% (w/v) casein solution. The measurements were performed after incubation (30 min,
37  C) and the proteolytic activity was estimated monitoring the absorbance of the supernatant at 280 nm at different. The control experiments are shown at left of the graph.
Results are expressed as mean ± S.D. (n ¼ 6).

of the hemorrhagic spots induced by BjussuMP-I. The results of this theoretical model of a P-I class SVMP, which is constituted by a
assay indicated a significant reduction of the hemorrhagic activity single catalytic domain (present and conserved in all SVMPs), and
when Q8 was administered at 1:30 and 1:100 w/w (toxin/Q8) Q8 were performed in order to provide noteworthy insights into
proportions after 15 min of the BjussuMP-I injection (47% and 66%, the structural features of this and other similar complexes.
respectively). Initial theoretical model of the native BjussuMP-II SVMP was
Proteolytic activity of B. jararacussu and B. neuwiedi crude built [42,44] using two crystallographic models as templates: a zinc
venoms and the SVMPs BjussuMP-I and neuwiedase on casein was metalloproteinase BaP1 from B. asper snake venom complexed to a
also inhibited significantly by Q8. In Fig. 3B, it is possible to observe peptidomimetic inhibitor [43] and a native metalloprotease BaP1
that the proteolytic activity of the control experiments (performed from B. asper [45]. This protein presents an identity degree of 81% in
without pre-incubation with Q8) presented an efficiency superior relation to BjussuMP-II and the score of the alignment between
to 60% for all tested crude venoms or SVMPs. In contrast, after pre- these two sequences was 417.0 [42]. The initial theoretical model of
incubation with the inhibitor, the efficiency of the proteolytic ac- the native BjussuMP-II was then improved through a 20 ns-MD
tivity dropped to values near 20% and 10%, respectively, for crude simulation [50,51] and it presented, after 5000 ps, a maximum
venoms and SVMPs. rmsd backbone atom amplitude of approximately 1.0 Å and a rmsd
Therefore, the data above show that Q8 can inhibit efficiently backbone atom average difference around 0.1 Å between 5001-
the hemorrhagic activity of some crude snake venoms and isolated 12500 ps and 12501e20000 ps (the first transient 5000 ps was not
the SVMP BjussuMP-I from B. jararacussu in mice. Additionally, the considered). After the MD simulation, the final model of the native
same conclusion can be drawn from experiments involving pro- BjussuMP-II showed a good stereochemical quality: 96.1% of its
teolytic activity on casein of tested crude venoms and SVMPs. These residues were in the core and additionally allowed regions of the
findings indicate the potential of Q8 as an anti-snake venom drug, Ramachandran plot [46] and the potential energy of the entire
particularly for the treatment of the hemorrhagic and necrotic ef- structure presented a negative balance (Z-score ¼ - 6.52) [47].
fects induced by the action of snake venoms and SVMPs found in Based on the data showed above, it is possible to suppose that
these complex biological mixtures. the theoretical native BjussuMP-II model obtained after the 20 ns-
MD simulation is feasible, not showing an apparent structural
instability. The model obtained at the end of the 20 ns-MD simu-
3.3. Structural insights into BjussuMP-II inhibition by Q8 lation also kept the structural similarity in relation to other
experimentally-determined SVMP catalytic domains [57e64], pre-
3.3.1. BjussuMP-II structural model senting three disulfide bridges (Cys116-Cys196, Cys156-Cys180,
According to the biochemical trials, Q8 was particularly efficient and Cys158-Cys163) and two ellipsoidal subdomains: the major
in the inhibition of hemorrhagic and proteolytic activities induced one containing the first 152 residues (four a-helices e A, B, C, and D,
by some crude snake venoms and snake venom metalloproteases and six stranded b-sheets e sI, sII, sIII, sIV, sV, and sVI) and the
(SVMPs). These results suggest that the toxin and Q8 interactions minor one enclosing the remaining 98 residues (one a-helix and
could be circumscribed to the active site and/or surroundings of several loops). Additionally, the imidazole rings from catalytic
SVMPs, since the catalytic domains of these toxins are related to histidines presented a favorable position to coordinate the catalytic
these harmful pharmacological and biological properties. Hence, zinc after the 20 ns-MD simulation, reinforcing the probable
docking and molecular dynamics simulations involving a
 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188 185

adequate structural conformation of the final native BjussuMP-II approximately 0.8 Å and a rmsd backbone atom average difference
model. Likewise, the side chain of the residue Glu142, which is around 0.1 Å between 2001e24000 ps and 24001e50000 ps (the
also essential for the nucleophilic attack on the scissile peptide first transient 2000 ps was not considered). Consequently, as in the
bond of metalloproteinase substrates [57], also remained in the first 20 ns-MD simulation, it is possible to assume that the model of
vicinity of the catalytic histidines. the protein/inhibitor complex does not present an instability ten-
dency during the 50 ns-MD simulation. In addition, after this 50 ns-
3.3.2. BjussuMP-II/Q8 complex model and its comparison with MD simulation, the general stereochemistry quality (97.1% of the
other complexed-SVMP structures residues were in the core and additionally allowed regions of the
Following the 20 ns-MD simulation, a molecular docking Ramachandran plot) (46) and potential energy (47) (Z-score ¼ -
simulation was carried out with the final native BjussuMP-II model 6.27) were essentially kept in comparison to the initial model ob-
and Q8. For this purpose, the docking site of the ligand was defined tained after the 20 ns-MD simulation.
as a sphere of approximately 4000 Å3 around the BjussuMP-II Indeed, the data from the latter MD simulation revealed more
catalytic ion Zn2þ. The definition of the active site and its sur- interesting features which could shed some light on the structural
roundings as the docking site was based on several crystallographic basis related to the SVMP inhibitory activity of Q8. An analysis of
experiments which indicate that these protein spots are effectively the interactions between ligand and protein during the 50 ns-MD
the regions of interaction between different SVMPs and specific simulation indicated that the formation of hydrogen bonds seems
inhibitors or ligands. In 1994, Zhang et al. [59], solved crystallo- not to play an essential role for the stabilization of the BjussuMP-II/
graphic complexes between atrolysin C (form d), a SVMP from Q8 complex. The most prevalent hydrogen bond (formed between
Crotalus atrox venom, and the natural PyroGlu-Asn-Trp tripeptide the Q8 hydroxymethyl group and the Gln108 Nε2 atom) occurred
and synthetic SCH47890 ligands, demonstrating that these mole- during only approximately 5.5% of the MD simulation (~2.75 ns),
cules bind to the atrolysin C active site region. Similarly, Cirilli et al. whereas none of the remaining exceeded 1.5% (~0.75 ns). On the
[65], solved the adamalysin II from Crotalus adamanteus crystallo- other hand, an important interaction involving the oxygen atom
graphic structure bound to an analogue of POL647, a synthetic tri- from the Q8 carbonyl group and the cofactor Zn2þ is observed in the
peptide derivative from the natural reprolysin inhibitors found in BjussuMP-II/Q8 complex, since an average distance of 1.9 Å was
crotalid and viperid snake venoms, and showed that the ligand kept between these atoms during all the 50 ns-MD simulation
interacts with catalytic site residues and the cofactor Zn2þ in an (Fig. 4).
asymmetric bidentate, partly filling the primary specificity subsite In fact, according to the program FindGeo [67], Q8 takes part in a
S1'. Furthermore, the presence of the inhibitor in this complex regular square pyramidal coordination of the cofactor Zn2þ, which
displaces the catalytically essential water molecule, which is co- is completed by the Nε2 atoms from active site histidines and the
ordinated by the cofactor Zn2þ and a glutamate residue in the Glu142/Zn2þ-coordinated catalytic water molecule and presents a
native protein. The non-covalent interaction of other peptidomi- rmsd value of only 0.312 Å in relation to the ideal geometry of this
metic inhibitors (POL647 and POL656) with residues of the ada- type of coordination. Interestingly, the Zn2þ regular square coor-
malysin II active site region was also described by Gomis-Rüth et al. dination found in the BjussuMP-II/Q8 complex is unique in com-
[58]. Analyzing the potential contributions of the adamalysin II/ parison to the total of 27 native or ligand-complexed SVMP crystal
POL647 and adamalysin II/POL656 complexes for drug design of structures currently available in PDB (Protein Data Bank). This
TACE (TNFa-converting enzymes) inhibitors, the authors indicated finding, associated with the fact that this ligand apparently binds to
that the insertion of these inhibitors in the adamalysin II active site BjussuMP-II without displacing the Glu142/Zn2þ-coordinated
provoked a slight opening of the substrate-binding cleft, as also
described by Zhang et al. [61] in the complex tripeptide KNL/SVMP
F-II from Agkistrodon acutus.
Our molecular docking results showed that the best protein/
ligand solutions (GoldScore Fitness y 55) present similar orienta-
tions of Q8, which are deeply inserted into the S1'-pocket of the
native BjussuMP-II model, occupying a region nearby the helix D, b-
strand sIV, and bulged segment (a region anterior to the b-strand
sIV edge which protrudes into the active site cleft) (for a complete
description of the SVMP structural elements see the review pre-
pared by Sto € cker et al. [66]). Interestingly, the analysis of these
molecular docking solutions seem to indicate that the fitting of the
inhibitor does not necessarily implicate in a displacement of the
Glu142/Zn2þ-coordinated catalytic water molecule, thus indicating
that the inhibitory activity of Q8 is probably not linked to the
impairment of the dyad Glu142/reactive solvent molecule. There-
fore, according to the molecular docking simulation, the ligand
could just impair the correct contact between the components of
the catalytic site and substrate molecules, not allowing the occur-
rence of any hydrolytic reaction due to a considerable steric
hindrance.
In order to verify the feasibility of these results, the BjussuMP-II/
Q8 molecular docking solution with the highest score (GoldScore
Fitness y 55.12) was submitted to an extended MD simulation Fig. 4. BjussuMP-II/Q8 interactions. An average distance of 1.9 Å was kept between the
(50 ns) to assess the stability of this complex and identify the oxygen atom from Q8 carbonyl group and the cofactor Zn2þ during the 50 ns-MD
simulation. The distances (Å) from the catalytic histidines (His144, His148, and His154)
prevalence of possible protein/inhibitor interactions. The analysis and Glu142/Zn2þ-coordinated catalytic water molecule (HOH) to the cofactor Zn2þ are
of the BjussuMP-II/Q8 structure after the 50 ns-MD simulation also shown to attest the appropriate conformation of the active site after the 50 ns-MD
showed a maximum rmsd backbone atom amplitude of simulation. This illustration was generated with program PyMOL v.1.3.
186 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188

cofactor Zn2þ hold the main interaction responsible for the stabi-
lization of the complex BjussuMP-II/Q8. Additionally, the low mo-
lecular weight of this quinolinone probably is potentially useful for
application against local effects caused by ophidic venoms since
this characteristic could permit rapid tissue diffusion of the inhib-
itor. Therefore, the experimental and theoretical data indicate that
Q8 is an interesting drug-candidate compound for antiophidic
therapy, particularly for the treatment of hemorrhagic and necrotic
effects induced by the action of bothropic and other viperid snake
venoms.

Conflict of interests
Fig. 5. Fit of Q8 (depicted in sticks) into the active site cavity of BjussuMP-II (depicted
in a surface representation). The low molecular weight and shape of the synthetic None.
molecule seem to be appropriate to provide an optimum access to the catalytic site,
allowing the impairing of the SVMP activity possibly by a steric hindrance mechanism.
This illustration was generated with program PyMOL v.1.3. Acknowledgments

The authors express their gratitude to Coordenaça ~o de Aper-

catalytic water as occurs in other available ligand-complexed SVMP feiçoamento de Pessoal de Nível Superior (CAPES), Conselho
crystal structure molecules, highlights the potential singular Q8 Nacional de Desenvolvimento Científico e Tecnolo
gico (CNPq),
characteristics in relation to other SVMP inhibitors. A possible Fundaç~ ao de Amparo a  Pesquisa do Estado de Sa ~o Paulo (FAPESP),
explanation for these features could be the low molecular weight Fundaç~ ao de Amparo  a Pesquisa do Estado de Minas Gerais
(173 Da) of Q8 in comparison to other SVMP inhibitors. This (FAPEMIG), Instituto Nacional de Cie ^ncia e Tecnologia em Toxinas
compact shape of this synthetic quinolinone probably provides (INCTTox) and Fundaça ~o de Amparo  a Pesquisa do Estado de
better access and an appropriate fit into the BjussuMP-II active site Rondo ^nia (FAPERO) for the financial support, and to Conselho de
(Fig. 5), thus favoring the interaction between the carbonyl group Gest~ao do Patrimo^ nio Gene
tico (CGEN/MMA) for the authorization
from the inhibitor and the protein cofactor Zn2þ. Therefore, it is number 010627/2011-1. We also thank Prof. Jose
R. Giglio (FMRP-
very plausible that Q8 is able to impair the SVMP activity by a steric USP, Brazil) (in memorian), Ta
ssia R. Costa and Renata S. Fernandes
hindrance mechanism related to its remarkable compact di- (FCFRP-USP) for their academic advising. The authors thank the
mensions. Also, this low molecular weight probably is particularly Program for Technological Development in Tools for Health-PDTIS-
useful for its application against local effects caused by ophidic FIOCRUZ for use of its facilities. Amy Grabner provided the English
venoms since it allows a rapid tissue diffusion. Indeed, our in vivo editing of the manuscript.
experiments support this last supposition given that the addition of
Q8 helped to reduce significantly the myotoxic effect caused by References
B. jararacussu PLA2s in mice.
Thereby, the structural analysis of the BjussuMP-II/Q8 model [1] M. Risch, D.D. Georgieva, M.M. Von Bergen, N. Jehmlich, N. Genov, R.K. Arni,
underscored the potential use of this synthetic quinolinone as an C. Betzel, Snake venomics of the siamese russell's viper (Daboia russelli sia-
mensis) e relation to pharmacological activities, J. Proteomics 72 (2009)
adjuvant in snakebite serum therapy, mainly in order to relieve the 256e269.
hemorrhagic and proteolytic actions of the Bothrops metal- [2] D. Georgieva, M. Risch, A. Kardas, F. Buck, M. Von Bergen, C. Betzel,
loproteases, as previously attested by the biological and functional Comparative analysis of the venom proteomes of Vipera ammodytes ammo-
dytes and Vipera ammodytes meridionalis, J. Proteome Res. 7 (2008) 866e886.
assays. Moreover, future pharmacological and structural studies [3] L. Sanz, N. Ayvazyan, J.J. Calvete, Snake venomics of the Armenian mountain
involving Q8 could also reveal more important clues for the vipers Macrovipera lebetina obtusa and Vipera raddei, J. Proteomics 71 (2008)
development of structure-based drugs against defective homo- 198e209.
[4] J.J. Calvete, C. Marcinkiewicz, L. Sanz, Snake venomics of Bitis gabonica
logue proteases involved in haemostatic system diseases which gabonica. Protein family composition, subunit organization of venom toxins,
affect humans and other mammals. and characterization of dimeric disintegrins bitisgabonin-1 and bitisgabonin-
2, J. Proteome Res. 6 (2007) 326e336.
[5] J.J. Calvete, E. Fasoli, L. Sanz, E. Boschetti, P.G. Righetti, Exploring the venom
proteome of the western diamondback rattlesnake, Crotalus atrox, via snake
4. Conclusions venomics and combinatorial peptide ligand library approaches, J. Proteome
Res. 8 (2009) 3055e3067.
The experimental results indicate that Q8 is able to inhibit [6] A. Alape-Giron, L. Sanz, J. Escolano, M. Flores-Diaz, M. Madrigal, M. Sasa,
J.J. Calvete, Snake venomics of the lancehead pitviper Bothrops asper:
significantly hemorrhagic and/or proteolytic activities presented by geographic, individual, and ontogenetic variations, J. Proteome Res. 7 (2008)
crude snake venoms from B. jararacussu, B. moojeni and B. alternatus 3556e3571.
and the isolated metalloproteases BjussuMP-I from B. jararacussu [7] A. Alape-Giron, M. Flores-Diaz, L. Sanz, M. Madrigal, J. Escolano, M. Sasa,
J.J. Calvete, Studies on the venom proteome of Bothrops asper: perspectives
and neuwiedase from B. neuwiedi in mice. According to structural and applications, Toxicon 54 (2009) 938e948.
studies involving docking and molecular dynamics simulations, a [8] J.J. Calvete, L. Sanz, P. Cid, P. De La Torre, M. Flores-Diaz, M.C. Dos Santos,
possible explanation for the inhibitory properties of this quinoli- A. Borges, A. Bremo, Y. Angulo, B. Lomonte, A. Alape-Giron, J.M. Gutierrez,
Snake venomics of the central American rattlesnake Crotalus simus and the
none molecule could be its low molecular weight structure, which South American Crotalus durissus complex points to neurotoxicity as an
allows an appropriate fit into the BjussuMP-II active site. The for- adaptive paedomorphic trend along crotalus dispersal in South America,
mation of a stable complex between Q8 and BjussuMP-II or other J. Proteome Res. 9 (2010) 528e544.
[9] M. Ohler, D. Georgieva, J. Seifert, M. Von Bergen, R.K. Arni, N. Genov, C. Betzel,
SVMPs may, consequently, impair the access of substrates to the
The venomics of Bothrops alternatus is a pool of acidic proteins with pre-
active site of these toxins, explaining thus the hemorrhagic and dominant hemorrhagic and coagulopathic activities, J. Proteome Res. 9 (2010)
proteolytic inhibition verified in the experimental trials. Indeed, a 2422e2437.
detailed analysis of the contacts formed during the MD simulation [10] A.T. Ching, M.M. Rocha, A.F. Paes Leme, D.C. Pimenta, D.F.M. De Fatima,
S.M. Serrano, P.L. Ho, I.L. Junqueira-De-Azevedo, Some aspects of the venom
between the residues from the BjussuMP-II active site and Q8 proteome of the Colubridae snake Philodryas olfersii revealed from a Duver-
shows the carbonyl group from the inhibitor and the protein noy's (venom) gland transcriptome, FEBS Lett. 580 (2006) 4417e4422.
 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188 187

[11] G. Ompraba, A. Chapeaurouge, R. Doley, K.R. Devi, P. Padmanaban, O. Tabarrini, V. Cechetti, 6-Aminoquinolones: photostability, cellular distri-
C. Venkatraman, D. Velmurugan, Q. Lin, R.M. Kini, Identification of a novel bution and phototoxicity, Toxicol. vitro 18 (2004) 581e592.
family of snake venom proteins veficolins from Cerberus rynchops using a [37] J.P. Michael, Quinoline, quinazoline and acridone alkaloids, Nat. Prod. Rep. 21
venom gland transcriptomics and proteomics approach, J. Proteome Res. 9 (2004) 650e668.
(2010) 1882e1893. [38] G. Anquetin, J. Greiner, P. Vierling, Quinolone-based drugs against Toxoplasma
[12] C.L. Weldon, S.P. Mackessy, Biological and proteomic analysis of venom from gondii and Plasmodium spp, Curr. Drug Targets Infect. Disord. 5 (2005)
the Puerto Rican Racer (Alsophis Portoricensis: Dipsadidae), Toxicon 55 (2010) 227e245.
558e569. [39] S. Marcussi, C.P. Bernardes, N.A. Santos-Filho, M.V. Mazzi, C.Z. Oliveira,
[13] D. Georgieva, J. Seifert, M. Ohler, M. Von Bergen, P. Spencer, R.K. Arni, L.F.M. Izidoro, A.L. Fuly, A.J. Magro, A.S.K. Braz, M.R.M. Fontes, J.R. Giglio,
N. Genov, C. Betzel, Pseudechis australis venomics: adaptation for a defense A.M. Soares, Molecular and functional characterization of a new non-
against microbial pathogens and recruitment of body transferrin, J. Proteome hemorrhagic metalloprotease from Bothrops jararacussu snake venom with
Res. 10 (2011) 2440e2464. antiplatelet activity, Peptides 28 (2007) 2328e2339.
[14] K. Kulkeaw, W. Chaicumpa, Y. Sakolvaree, P. Tongtawe, P. Tapchaisri, Prote- [40] M.V. Mazzi, S. Marcussi, G.B. Carlos, R.G. Stabeli, J.J. Franco, F.K. Ticli,
ome and immunome of the venom of the thai cobra, Naja kaouthia, Toxicon 49 A.C.O. Cintra, S.C. França, A.M. Soares, S.V. Sampaio, A new hemorrhagic
(2007) 1026e1041. (fibrinogenolytic) metalloprotease from Bothrops jararacussu snake venom:
[15] D. Petras, L. Sanz, A. Segura, M. Herrera, M. Villalta, D. Solano, M. Vargas, isolation and biochemical characterization, Toxicon 44 (2004) 215e223.
G. Leon, D.A. Warrell, R.D. Theakston, R.A. Harrison, N. Durfa, A. Nasidi, [41] V.M. Rodrigues, A.M. Soares, R. Guerra-Sa
, V. Rodrigues, M.R.M. Fontes,
J.M. Gutierrez, J.J. Calvete, Snake venomics of African spitting cobras: toxin J.R. Giglio, Structural and functional characterization of neuwiedase, a non-
composition and assessment of congeneric cross-reactivity of the pan-African hemorrhagic fibrin(ogen)olytic metalloprotease from Bothrops neuwiedi snake
EchiTAb-Plus-ICP antivenom by antivenomics and neutralization approaches, venom, Arch. Biochem. Biophys. 381 (2000) 213e224.
J. Proteome Res. 10 (2011) 1266e1280. [42] J. So€ ding, A. Biegert, A.N. Lupas, The HHpred interactive server for protein
[16] L. St Pierre, G.W. Birrell, S.T. Earl, T.P. Wallis, J.J. Gorman, J. De Jersey, homology detection and structure prediction, Nucl. Acids Res. 33 (2005)
P.P. Masci, M.F. Lavin, Diversity of toxic components from the venom of the 244e248.
evolutionarily distinct black whip snake, Demansia vestigiata, J. Proteome Res. [43] T.J. Lingott, C. Schleberger, J.M. Gutie
rrez, I. Merfort, High-resolution crystal
6 (2007) 3093e3107. structure of the snake venom metalloproteinase BaP1 complexed with a
[17] L. Quinton, J.P. Le Caer, G. Phan, C. Ligny-Lemaire, J. Bourdais-Jomaron, peptidomimetic: insight into Inhibitor binding, Biochemistry 48 (2009)
F. Ducancel, J. Chamot-Rooke, Characterization of toxins within crude venoms 6166e6174.
by combined use of fourier transform mass spectrometry and cloning, Anal. [44] M.A. Marti-Renom, A. Stuart, A. Fiser, R. Sanchez, F. Melo, A. Sali, Comparative
Chem. 77 (2005) 6630e6639. protein structure modeling of genes and genomes, Annu. Rev. Biophys. Bio-
[18] J.W. Fox, S.M. Serrano, Insights into and speculations about snake venom mol. Struct. 29 (2000) 291e325.
metalloproteinase (SVMP). Synthesis, folding and disulfide bond formation [45] L. Watanabe, J.D. Shannon, R.H. Valente, A. Rucavado, A. Alape-Giron,
and their contribution to venom complexity, FEBS J. 275 (2008) 3016e3030. A.S. Kamiguti, R.D. Theakston, J.W. Fox, J.M. Gutie
rrez, R.K. Arni, Amino acid
[19] J.W. Fox, S.M. Serrano, Timeline of key events in snake venom metal- sequence and crystal structure of BaP1, a metalloproteinase from Bothrops
loproteinase research, J. Proteomics 72 (2009) 200e209. asper snake venom that exerts multiple tissue-damaging activities, Protein
[20] D.H. Souza, M.R. Iemma, L.L. Ferreira, J.P. Faria, M.L. Oliva, R.B. Zingali, Sci. 12 (2003) 2273e2281.
S. Niewiarowski, H.S. Selistre-de-Araujo, The disintegrin-like domain of the [46] S.C. Lovell, I.W. Davis, W.B. Arendall III, P.I.W. de Bakker, J.M. Word,
snake venom metalloprotease alternagin inhibits alpha2beta1 integrin- M.G. Prisant, J.S. Richardson, D.C. Richardson, Structure validation by Calpha
mediated cell adhesion, Arch. Biochem. Biophys. 384 (2000) 341e350. geometry: phi, psi and Cbeta deviation, Proteins Struct. Funct. Genet. 50
[21] M.R. Cominetti, C.H. Terruggi, O.H. Ramos, J.W. Fox, A. Mariano-Oliveira, (2003) 437e450.
M.S. De Freitas, C.C. Figueiredo, V. Morandi, H.S. Selistre-de-Araujo, Alter- [47] M. Wiederstein, M.J. Sippl, ProSA-web: interactive web service for the
nagin-C, a disintegrin-like protein, induces vascular endothelial cell growth recognition of errors in three-dimensional structures of proteins, Nucleic
factor (VEGF) expression and endothelial cell proliferation in vitro, J. Biol. Acids Res. 35 (2007) W407eW410.
Chem. 279 (2004) 18247e18255. [48] M.D. Hanwell, D.E. Curtis, D.C. Lonie, T. Vandermeersch, E. Zurek,
[22] P. Zigrino, A.S. Kamiguti, J. Eble, C. Drescher, R. Nischt, J.W. Fox, C. Mauch, The G.R. Hutchison, Avogadro: an advanced semantic chemical editor, visualiza-
reprolysin jararhagin, a snake venom metalloproteinase, functions as a tion, and analysis platform, J. Cheminform. 4 (2012) 17.
fibrillar collagen agonist involved in fibroblast cell adhesion and signaling, [49] G. Jones, P. Willett, R.C. Glen, Molecular recognition of receptor sites using a
J. Biol. Chem. 227 (2002) 40528e40535. genetic algorithm with a description of desolvation, J. Mol. Biol. 245 (1995)
[23] J.M. Gutie
rrez, A. Rucavado, T. Escalante, C. Díaz, Hemorrhage induced by 43e53.
snake venom metalloproteinases: biochemical and biophysical mechanisms [50] H.J.C. Berendsen, D. van der Spoel, R. Van Drunen, GROMACS: a message-
involved in microvessel damage, Toxicon 45 (2005) 997e1011. passing parallel molecular dynamics implementation, Comp. Phys. Commun.
[24] F.S. Markland Jr., S. Swenson, Snake venom metalloproteinases, Toxicon 62 91 (1995) 43e56.
(2013) 3e18. [51] B. Hess, C. Kutzner, D. van der Spoel, E. Lindahl, GROMACS 4: algorithms for
[25] J.M. Gutie
rrez, A. Rucavado, Snake venom metalloproteinases: their role in the highly efficient, load-balanced, and scalable molecular simulation, J. Chem.
pathogenesis of local tissue damage, Biochemie 82 (2000) 841e850. Theory Comput. 4 (2008) 435e447.
[26] J.D. Phillipson, L.A. Anderson, Ethnopharmacology and western medicine, [52] C. Oostenbrink, T.A. Soares, N.F.A. van der Vegt, W.F. van Gunsteren, Valida-
J. Ethnopharmacol. 25 (1989) 61e72. tion of the 53A6 GROMOS force field, Eur. Biophys. J. 34 (2005) 273e284.
[27] W. Martz, Plants with a reputation against snakebite, Toxicon 30 (1992) [53] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, J. Hermans, Interaction
1131e1142. models for water in relation to protein hydration, in: B. Pullman (Ed.),
[28] W.B. Mors, M.C. Nascimento, B.M.R. Pereira, N.A. Pereira, Plant natural prod- Intermolecular Forces, D. Reidel Publishing Co., Dordrecht, 1981, pp. 331e342.
ucts active against snake bite e the molecular approach, Phytochem 55 (2000) [54] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. DiNola, J.R. Haak,
627e642. Molecular dynamics with coupling to an external bath, J. Chem. Phys. 81
[29] A.M. Soares, A.H. Janua
rio, M.V. Lourenço, A.M.S. Pereira, P.S. Pereira, (1984) 7.
Neutralizing effects of snake venoms exhibited by Brazilian plants, Drugs [55] C. Andreini, L. Banci, I. Bertini, S. Elmi, A. Rosato, Comparative analysis of the
Future 29 (2004) 1105e1117. ADAM and ADAMTS families, J. Protein Res. 4 (2005) 881e888.
[30] A.M. Soares, F.K. Ticli, S. Marcussi, M.V. Lourenço, A.H. Janu
ario, S.V. Sampaio, [56] D. Edmont, R. Rocher, C. Plisson, J. Chenault, Synthesis and evaluation of
B. Lomonte, P.S. Pereira, Medicinal plants with inhibitory properties against quinoline carboxyguanidines as antidiabetic agents, Bioorg. Med. Chem. Lett.
snake venoms, Curr. Med. Chem. 12 (2005) 2625e2641. 10 (2000) 1831e1834.
[31] B.S. Vishwanath, T.V. Gowda, Interaction of aristolochic acid with Vipera [57] F.X. Gomis-Rüth, L.F. Kress, W. Bode, First structure of a snake venom met-
russelli phospholipase A2: its effect on enzymatic and pharmacological ac- alloproteinase: a prototype for matrix metalloproteinases/collagenases, EMBO
tivities, Toxicon 25 (1987) 929e937. J. 12 (1993) 4151e4157.
[32] S. Marcussi, C.D. SantAna, C.Z. Oliveira, A.Q. Rueda, D.L. Menaldo, R. Beleboni, [58] F.X. Gomis-Rüth, E.F. Meyer, L.F. Kress, V. Politi, Structures of adamalysin II
R.G. Sta
beli, J.R. Giglio, M.R.M. Fontes, A.M. Soares, Snake venom phospholi- with peptidic inhibitors. Implications for the design of tumor necrosis factor a
pase A2 inhibitors: medicinal chemistry and therapeutic potential, Curr. Top. convertase inhibitors, Protein Sci. 7 (1998) 283e292.
Med. Chem. 7 (2007) 743e756. [59] D. Zhang, I. Botos, F.X. Gomis-Rüth, R. Doll, C. Blood, F.G. Njoroge, J.W. Fox,
[33] M.F. Batina, A.C. Cintra, E.L. Veronese, M.A. Lavrador, J.R. Giglio, P.S. Pereira, W. Bode, E.F. Meyer, Structural interaction of natural and synthetic inhibitors
D.A. Dias, S.C. França, S.V. Sampaio, Inhibition of the lethal and myotoxic ac- with the venom metalloproteinase, atrolysin C (form d), Proc. Natl. Acad. Sci.
tivities of Crotalus durissus terrificus venom by Tabernaemontana catharinensis: U. S. A. 91 (1994) 8447e8451.
identification of one of the active components, Planta Med. 66 (2000) [60] T. Kumasaka, M. Yamamoto, H. Moriyama, N. Tanaka, M. Sato, Y. Katsube,
424e428. Y. Yamakawa, T. Omori-Satoh, S. Iwanaga, T. Ueki, Crystal structure of H2-
[34] Q.B. Li, R. Pan, G.F. Wang, S.X. Tang, Anisodamine as an effective drug to treat proteinase from the venom of Trimeresurus flavoviridis, J. Biochem. 119
snakebites, J. Nat. Toxins 8 (1999) 327e330. (1996) 49e57.
[35] K.S. Girish, K. Kemparaju, Inhibition of Naja naja venom hyaluronidase by [61] W. Gong, X. Zhu, S. Liu, M. Teng, L. Niu, Crystal structures of acutolysin A, a
plant-derived bioactive components and polysaccharides, Biochem. Mosc. 70 three-disulfide hemorrhagic zinc metalloproteinase from the snake venom of
(2005) 948e952. Agkistrodon acutus, J. Mol. Biol. 283 (1998) 657e668.
[36] G. Viola, L. Facciolo, S. Dall’Acqua, F. Di Lisa, M. Canton, D. Vedaldi, A. Fravolini, [62] X. Zhu, M. Teng, L. Niu, Structure of acutolysin-C, a haemorrhagic toxin from
188 P.T. Baraldi et al. / Biochimie 121 (2016) 179e188

the venom of Agkistrodon acutus, providing further evidence for the mecha- metalloprotease subgroups, Toxicon 52 (2008) 807e816.
nism of the pH-dependent proteolytic reaction of zinc metalloproteinases, [65] M. Cirilli, C. Gallina, E. Gavuzzo, C. Giordano, F.X. Gomis-Rüth, B. Gorini,
Acta Cryst. D55 (1999) 1834e1841. L.F. Kress, F. Mazza, M.P. Paradisi, G. Pochetti, V. Politi, 2 Å x-Ray structure of
[63] K.F. Huang, S.H. Chiou, T.P. Ko, A.H. Wang, Determinants of the ihibition of a adamalysin II complexed with a peptide phosphonate inhibitor adopting a
Taiwan habu venom metalloproteinase by its endogenous inhibitors revealed retro-binding mode, FEBS Lett. 418 (1997) 319e322.
by X-ray crystallography and synthetic inhibitor analogues, Eur. J. Biochem. [66] W. Sto€cker, F. Grams, U. Baumann, P. Reinemer, F.X. Gomis-Rüth, D.B. McKay,
269 (2002) 3047e3056. W. Bode, The metzincins e topological and sequential relations between the
[64] J.R.C. Muniz, A.L.B. Ambro
sio, H.S. Selistre-de-Araújo, M.R. Cominetti, astacins, adamalysins, serralysins, and matrixins (collagenases) define a su-
A.M. Moura-da-Silva, G. Oliva, R.C. Garratt, D.H.F. Souza, The three- perfamily of zinc-peptidases, Protein Sci. 4 (1995) 823e840.
dimensional structure of Bothropasin, the main hemorrhagic factor from [67] C. Andreini, G. Cavallaro, S. Lorenzini, FindGeo: a tool for determining metal
Bothrops jararaca venom: insights for a new classification of snake venom coordination geometry, Bioinformatics 28 (2012) 1658e1660.