Snake venom glutaminyl cyclases: Purification, cloning, kinetic study, recombinant expression, and comparison with the human enzyme

Article abstract of DOI:10.1016/j.toxicon.2014.04.012

Among various snake venom components, glutaminyl cyclase (vQC) is one of the least understood protein family and none of its members has been purified or characterized. Here we confirmed the presence of vQC activity in a wide spectrum of venom species via enzymatic assay using a synthetic fluorogenic substrate. We have also cloned novel vQC cDNAs from seven species including Crotalus atrox. The results revealed more than 96% sequence similarities among vQCs and ～75% sequence identities between vQCs and human secretory QC (hQC). The vQC glycoprotein of 43 kDa was isolated from C. atrox venom, and its N-terminal sequence was determined. The optimal pH range for vQC reaction was 7.5-8.0, and the enzymes were stable up to 50 °C. Similar to hQC, vQCs were substantially inactivated by 1 mM 1,10-phenanthroline but slightly affected by 20 mM EDTA, suggestive of a similar zinc-catalytic environment for these enzymes. Although their catalytic residues were highly conserved, vQCs were less susceptible to inhibition by synthetic imidazole derivatives which potently inhibited hQC. The 3D-models revealed that vQC and hQC structures display different surface charge distributions around the active sites, which might affect substrate and inhibitor binding affinities. The recombinant vQCs prepared from Escherichia coli displayed weaker substrate binding affinities relative to the native vQCs, possibly due to the lack of glycosylation. The present report offers new structural and functional insights into vQCs and sheds light on the specificity differences between vQCs and hQC.

Full text of DOI:10.1016/j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 1/11
Toxicon xxx (2014) 1–11
51
52
53
54
55
56
57
Contents lists available at ScienceDirect
Toxicon
journal homepage: www.elsevier.com/locate/toxicon
58
59
60
61
1
2
3
4
5
6
7
8
Snake venom glutaminyl cyclases: Purification, cloning,
kinetic study, recombinant expression, and comparison with
the human enzymeq
62
63
64
65
66
67
68
69
70
71
,b,
Ying-Ming Wang a, Kai-Fa Huang a, Inn-Ho Tsai a
a Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan
*
Q8
b Institute of Biochemical Sciences, National Taiwan University, Taipei, Taiwan
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
72
73
74
75
76
a r t i c l e i n f o
a b s t r a c t
Article history:
Among various snake venom components, glutaminyl cyclase (vQC) is one of the least
understood protein family and none of its members has been purified or characterized.
Here we confirmed the presence of vQC activity in a wide spectrum of venom species via
enzymatic assay using a synthetic fluorogenic substrate. We have also cloned novel vQC
cDNAs from seven species including Crotalus atrox. The results revealed more than 96%
sequence similarities among vQCs and w75% sequence identities between vQCs and
human secretory QC (hQC). The vQC glycoprotein of 43 kDa was isolated from C. atrox
venom, and its N-terminal sequence was determined. The optimal pH range for vQC re-
action was 7.5–8.0, and the enzymes were stable up to 50 ꢀC. Similar to hQC, vQCs were
substantially inactivated by 1 mM 1,10-phenanthroline but slightly affected by 20 mM
EDTA, suggestive of a similar zinc-catalytic environment for these enzymes. Although their
catalytic residues were highly conserved, vQCs were less susceptible to inhibition by
synthetic imidazole derivatives which potently inhibited hQC. The 3D-models revealed
that vQC and hQC structures display different surface charge distributions around the
active sites, which might affect substrate and inhibitor binding affinities. The recombinant
vQCs prepared from Escherichia coli displayed weaker substrate binding affinities relative
to the native vQCs, possibly due to the lack of glycosylation. The present report offers new
structural and functional insights into vQCs and sheds light on the specificity differences
between vQCs and hQC.
Received 7 February 2014
Received in revised form 18 April 2014
Accepted 29 April 2014
Available online xxxx
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
Keywords:
Snake venom pyroglutaminyl modification
Cloning and sequencing
3D models
Recombinant enzyme
Inhibition kinetics
Q3
Ó 2014 Published by Elsevier Ltd.
1. Introduction
Abbreviations: Ah, N-
1-(3,4-dimethoxyphenyl)-3-[(3-(1H-imidazol-1-yl) propyl]thiourea; QC,
glutaminyl cyclases; hQC, human secreted glutaminyl cyclase; NA, b-
u-acetylhistamine; Bi, 1-benzylimidazole; PBD150,
b
As one of the enzymes for protein post-translational
modifications, glutaminyl cyclase (QC; glutaminyl-peptide
cyclotransferase, EC 2.3.2.5) catalyzes N-terminal pyroglu-
tamate (pGlu) formation on proteins or peptides and the
release of ammonia or water molecules (Seifert et al.,
2009). This modification appears to be important for
structural stability, resistance to aminopeptidase degrada-
tion, and interaction of the proteins or peptides with their
partners (Stephan et al., 2009). It has been found that the
animal QCs are either Golgi-resident or secreted enzymes
naphthylamine; PNGase F, peptide N-glycosidase F; pGlu, pyroglutamate;
PAP, pyroglutamyl aminopeptidase.
q
Eight nucleotide sequences encoding the vQCs of Crotalus atrox,
Micrurus fulvius, Cerrophidion godmani, Daboia russelii, Trimeresurus gra-
cilis, Protobothrops mucrosquamatus, and Sistrurus catenatus tergeminus
have been deposited in GenBank with the accession numbers JF979132–
JF979139JF979132JF979133JF979134JF979135JF979136JF979137JF979138-
JF979139, respectively.
* Corresponding author. Institute of Biochemical Sciences, National
Taiwan University, Taipei, Taiwan. Tel.: þ886 2 23665521; fax: þ886 2
23635038.
Q2
Q1
E-mail address: bc201@gate.sinica.edu.tw (I.-H. Tsai).
http://dx.doi.org/10.1016/j.toxicon.2014.04.012
0041-0101/Ó 2014 Published by Elsevier Ltd.
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 2/11
2
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
1
2
3
4
5
6
7
8
(Schilling et al., 2007). In humans, aberrant expression of
hQC is related to certain pathological conditions, such as
Alzheimer disease (Jawhar et al., 2011). X-ray crystallog-
raphy and site-directed mutagenesis studies of recombi-
nant human secreted QC (hQC) have provided much
information about the catalytic mechanism of QCs (Huang
et al., 2005a, 2008). Notably, hQC inhibitors have been
synthesized and developed as drugs to treat the relevant
diseases (Buchholz et al., 2009; Schilling et al., 2003a).
Presumably because of the N-terminal pGlu modifica-
tion, a number of snake venom proteins or toxins were
reported to be resistant to Edman-degradation during
sequence analysis. The well-known cases include
Laticauda semifasciata venom was obtained from Sigma
Chemical (MO, USA). Venom and venom glands of Dein-
agkistrodon acutus and Protobothrops mucrosquamatus
were obtained from antivenin manufacture group of Dis-
ease Control Center, Taiwan. A healthy specimen of M. ful-
vius was obtained from Glades Herp Co. (Fort Myers,
Florida, USA). After anesthesia of the snake, the venom
glands were dissected for the preparation of mRNA and
cDNA. Previously, mRNA was extracted from the fresh
venom glands and the cDNAs have been prepared for C.
atrox (Tsai et al., 2001), C. godmani, S. c.tergeminu, D. russelii
(Tsai et al., 2007), P. mucrosquamatus (Tsai et al., 1995), and
Trimeresurus gracilis (Tsai et al., 2012).
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
bradykinin-potentiating
peptides
(Ferreira,
1965;
Murayama et al., 1997), metalloproteases (Huang et al.,
2002), endogenous metalloprotease inhibitors (Wagstaff
et al., 2008), crotoxin A (Chen et al., 2004; Faure et al.,
2011), crotalphine (Konno et al., 2008), and dendrotoxin
(Harvey and Robertson, 2004). Although several cDNA se-
quences encoding snake venom QCs (vQCs) could be found
in databanks (Pawlak and Kini, 2006), none of the vQCs
have been isolated or assayed. Whether vQCs are useful
models for the drug development and applications related
to hQC also remains to be evaluated and explored.
Advances in the method for QC activity assays (Schilling
et al., 2002) and the discovery of new hQC inhibitors
(Buchholz et al., 2009) have rendered better tools to
explore vQCs. In the present study, we have assayed the QC
activities in more than 18 venom species with a fluorogenic
substrate. One of the vQC has been isolated and its N-ter-
minal sequence was determined for the first time.
Furthermore, we successfully cloned and sequenced the
vQC cDNAs from seven representative species and
compared the sequences with those of human and other
animal QCs (Huang et al., 2008). We also examined the
thermal stability, optimal pH, and the effects of metal ions,
metal chelators, and synthetic QC inhibitors on the repre-
sentative vQCs. Because the vQC sequences indicated three
N-glycosylation sites, we investigated the effect of glycan-
removal on the vQC activities. Furthermore, recombinant
vQCs were prepared using Escherichia coli and their kinetic
properties were studied. Finally, the structure homology
models of vQCs were generated and compared. The results
provide new insights into the occurrence, enzymology,
mechanism and phylogenetics of this venom protein
family.
2.2. Biochemicals and other reagents
Gln-b-naphthylamide (Gln-bNA) and recombinant
pyroglutamyl aminopeptidase (PAP) from Bacillus amyloli-
quefaciens were purchased from Bachem (Bubendorf,
Switzerland) and Qiagen (Hilden, Germany), respectively.
Peptide N-glycosidase F (PNGase F) was obtained from New
England Biolabs. (Ipswich, UK). Restriction enzymes and the
pGEM-Teasy vector were acquired from Promega Corp. (WI,
USA). Buffers and other chemicals were purchased from
Merck Co. (Darmstadt, Germany) or Sigma–Aldrich Co. (MO,
USA). The synthetic inhibitors,1-(3,4-dimethoxyphenyl) -3-
[(3-(1H-imidazol-1-yl) propyl]thiourea (PBD150), N-u-
acetylhistamine (Ah),1-benzylimidazole (Bi), and hQC were
gifts from our colleague, Prof. A. H.-J. Wang (Huang et al.,
2011).
2.3. Protein determination and vQC assay
Crude venoms were dissolved in reagent grade water
and quantified using a BCA protein assay kit (Pierce
Chemical Co., IL, USA) with bovine serum albumin as a
standard. The vQC activities were assayed using the fluo-
95
96
97
98
rogenic substrate Gln-bNA in the presence of an accessory
99
enzyme pyroglutamyl aminopeptidase (PAP) (Huang et al.,
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
2011). A 96-well microplate was incubated at 30 ꢀC, and
each well contained 0.025 units PAP in 170
HCl (pH8.0), which was sufficient to hydrolyze pGlu-
instantaneously. A small aliquot (10 l) of the enzyme
sample was added to each well and mixed, and the reaction
was initiated by adding Gln- NA to a final concentration of
m
l 50 mM Tris–
b
NA
m
b
0.50 mM (i.e., [S] >> Km). Using a SpectraMax M2 fluo-
rescent spectrometer reader (Molecular Devices Corp. CA,
USA) with excitation and emission wavelengths at 320 and
405 nm, respectively, the fluorescence increase was moni-
tored for 10–20 min. For kinetic analyses, the substrate
concentrations varied over a 20-fold range and the reaction
rates were determined in triplicate. The initial rates were
calibrated by a standard curve of the fluorescence of 2.5–
2. Materials and methods
2.1. Snake venoms and venom glands
The pooled venom samples of Bothrops atrox, Crotalus
atrox, Crotalus horridus atricaudatus, Crotalus durissus ter-
rificus, Daboia russelii, Echis pyramidium, Micrurus fulvius
and Naja haje were purchased from Kentucky Reptile Zoo
(KY, USA). Cerrophidion godmani, Causus rhombeatus,
Cerastes vipera and Dendroaspis angusticeps venoms were
obtained from Miami Serpentarium Laboratories (FL, USA).
Protobothrops elegans venom was purchased from La Toxan
(Valence, France). Pseudechis australis venom was acquired
from Venom Supplies Ltd. (Tanunda, South Australia).
50 mM bNA. The specific activity was determined based on
the initial rate and the protein concentration.
2.4. The pH profile and the thermal stability
The venom solution was mixed with various buffers
(final concentration 50 mM), including: acetate (pH 3.5, 4.0,
4.5 and 5.0), MES (pH 5.5, 6.0 and 6.5), HEPES (pH 7.0 and
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 3/11
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
3
1
2
3
4
5
6
7
8
7.5), and Tris–HCl (pH 7.5, 8.0, 8.5, 9.0 and 9.5). After adding
PAP and Gln- NA, the initial rates of product generation at
different pH were measured by a fluorescent spectrometer
reader. Additionally, to study the pH profile of PAP cata-
condition in 50 ml 40 mM sodium phosphate (pH 7.2) at
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
b
37 ꢀC for 3 h. The reduced vQC mass after the treatment was
analyzed by SDS-PAGE. The Km of the deglycosylated vQC
was determined by Lineweaver–Burk double reciprocal
plot.
lyzed reaction, hQC was used to transform Gln-
pGlu- NA in dilute Tris–HCl buffer, then the product was
added to different buffers containing PAP.
bNA to
b
2.8. Cloning and sequencing of vQCs
To evaluate the vQC stability, the venom solution was
incubated in a water bath at various temperatures (20–
65 ꢀC) for 30 min with gentle shaking and then cooled
immediately. The vQC activity in the sample was then
9
PCR primers were synthesized based on conserved re-
gions of the cDNA encoding two homologous vQCs from
database (Pawlak and Kini, 2006), namely, 50-CTSCAACGGC
ATGGCCRGA-30 (sense) and 50-TMAGGGTAACAGTTATCC C-
30 (antisense). Using the cDNA prepared from venom gland
mRNA templates, PCR was conducted using SuperTaq DNA
polymerase (HT Biotech, UK) (Mullis and Faloona, 1987;
Maniatis et al., 1989). The PCR consisted of 30 cycles of
denaturation at 92 ꢀC for 1.0 min, annealing at 60 ꢀC for
1 min, and then extension at 72 ꢀC for 1.0 min. The sizes of
the PCR products were confirmed by gel electrophoresis on
a 1% agarose gel. After being treated with polynucleotide
kinase, the PCR product was inserted into the pGEM-T easy
vector (Promega Biotech) and transformed into the E. coli
strain JM109. The white transformants were selected, and
the DNA was extracted, digested with restriction enzymes,
and analyzed by agarose gel electrophoresis. The cDNA was
sequenced using Sequencing System model 373A (PE-
Applied Biosystems, USA).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
assayed using 0.50 mM Gln-
bNA in Tris–HCl (pH 8.0) and
30 ꢀC, as described above.
2.5. Effects of modulators and inhibiters
EDTA (ethylenediaminetetraacetic
acid),
1,10-
phenanthroline and metal ions were added to the assay
buffer separately, and their effects on the vQC activity were
compared with the control. The reaction was initiated by
adding an aliquot of dissolved venom (0.1 mg) to the 170
reaction solutions in microplate wells containing 0.05–
5 mM Gln- NA. The initial rates were measured, and the Km
ml
b
values were determined by Lineweaver–Burk double-
reciprocal plots. To determine the inhibition by three
imidazole derivatives (PBD150, Ah, and Bi), the venom or
vQC was first incubated with the inhibitor at 30 ꢀC for
20 min, and then the substrate solution was added to
initiate the reaction. The IC50 values were obtained by
fitting the initial reaction rate versus the inhibitor con-
centrations. Inhibition constant (Ki) was calculated ac-
cording to the equation IC50 ¼ Ki (1 þ S/Km) (Segel, 1993),
where S is the substrate concentration and Km is the
Michaelis–Menten constant.
2.9. Sequence alignment, phylogenetic analysis, and
homology modeling
The deduced amino acid sequences of ten vQCs were
aligned with those of human QC and Xenopus QC using the
program ALIGN X (VECTOR NTI package). Cladistic analysis
of the sequence data was conducted using PHYLIP
(Felsenstein, 1992) to construct a phylogenetic tree by the
neighbor-joining method. Supports for the nodes were
assessed by a bootstrap computer program in 1000 repli-
cates (Felsenstein, 1985).
The 3D-homology models of five representative vQCs
were generated using the Discovery Studio program
(Accelrys Software Inc.) and the reported crystallographic
structure of hQC (PDB code 2AFM) (Huang et al., 2005a). The
initial models were optimized by energy minimization and
verified by 3D-profile (Lüthy et al., 1992) and PROCHECK
(Laskowski et al., 1993) programs. The structure figures
were prepared using PyMOL (http://www.pymol.org).
95
96
97
98
2.6. Partial purification and N-terminal sequencing of C.
atrox vQC
99
Because of the low vQC content, we used 1.5 g of the C.
atrox venom powder to purify the vQC. The venom was
dissolved in 50 mM CH3COONH4 (pH 6.5) and clarified by
centrifugation. The supernatant was applied to a Super-
dexÔ 200 gel filtration column (HiLoadÔ 16/60, Amersham
Biosciences Co., Sweden) and eluted with the same buffer
on an FPLC system. The QC activity in each fraction was
assayed. The fractions containing QC activity were sub-
jected to ion exchange chromatography on Mono S column
(HR5/5, GE Healthcare Bio-Science AB, Sweden), and
repeated using different salt gradients. The purity of the
isolated vQC was evaluated by SDS-PAGE on a NuPAGE 4–
12% Bis-Tris Gel (Novex life technologies, USA). The putative
vQC band was blotted on a PVDF membrane and stained
with Amido Black. The protein band was cut out and sub-
jected to amino acid sequence analysis using a protein
sequencer (ProciseÒ Sequencing System, model 494;
Applied Biosystems, USA).
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
2.10. Expression and characterization of recombinant vQCs
A BamHI site and the encoding sequences for (His)6-tag
and Factor Xa cleavage site were introduced into the 50-end
of the open reading frame of vQC cDNA and an XhoI site was
introduced into the 30-end. The resulting product was
ligated into the BamHI-XhoI sites of the pET32a expression
vector (Novagen, Madison, WI). The constructed expression
vectors were transformed into E. coli BL21 (DE3) (Novagen)
(Huang et al., 2005b), which were grown in LB broth con-
2.7. Effect of PNGase F deglycosylation on vQC kinetics
taining 200
of 0.8 was reached. Then, 1 mM isopropyl-
galactopyranoside was added to the culture. After 24 h of
m
g/mL ampicillin at 37 ꢀC until a culture OD600
-thio-
Deglycosylation of isolated C. atrox vQC by 1.0 unit of
b-D
Q4
PNGase
F was performed under a non-denaturing
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 4/11
4
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
incubation at 20 ꢀC, the cells were harvested by centrifu-
gation at 6000 rpm (2500 g). The cell pellet was suspended
in buffer A (150 mM NaCl/20 mM imidazole/50 mM Tris–
HCl, pH 8.0) with 10 mg/ml lysozyme at 37 ꢀC for 30 min
and lysed by freeze and thaw three times. After centrifu-
gation at 31,000 g for 1 h, the supernatant was loaded onto
a HisTrapÔ FF (GE Healthcare Bio-Science AB, Sweden)
column equilibrated in buffer A. The column was washed
with buffer A and then eluted with a linear gradient of
buffer B (i.e., buffer A containing 500 mM imidazole).
The fusion proteins of vQC were pooled and digested
with Factor Xa at 25 ꢀC for 1 day. Subsequently, the re-
combinant vQCs were eluted from a HistrapÔ HP column
(GE Healthcare) with a linear gradient of buffer B, and then
concentrated and desalted by an Amicon Ultra centrifugal
3. Results and discussions
1
2
3
4
5
6
7
8
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
3.1. Occurrence and the activity levels of vQCs
As shown in Fig.1A, the QC activity is proportional to the
amount of the venom added to the reaction in the assay
using 0.50 mM Gln-
capable of detecting QC activity in less than 20
venoms. The control reaction omitting the auxiliary
bNA in Tris–HCl (pH 8.0). This assay was
m
g of crude
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
filter device (Millipore Corp., MA). Using the Gln-bNA assay
system, kinetic studies of the fusion proteins and the pro-
cessed recombinant vQCs were performed. The kinetic
parameters of the isolated C. atrox vQC at 37 ꢀC were also
analyzed by Graph Pad Prism-5 software.
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Fig. 2. (A) The pH profiles for the reactions of vQCs, using Gln-
substrates. (B) The pH profiles for the reactions of PAP, using pGlu-
bNA as the
bNA as
the substrates. The buffers used are as follows, acetate (pH 3.5, 4.0, 4.5 and
5.0), MES (pH 5.5, 6.0 and 6.5), HEPES (pH 7.0 and 7.5), and Tris (pH 7.5, 8.0,
8.5, 9.0 and 9.5). (C) Thermal stability of vQCs. The venom of C. atrox or P.
mucrosquamatus was incubated for 30 min in 50 mM Tris–HCl (pH 8.0) at
various temperatures. The remaining enzyme activities were assayed using
Fig. 1. Activity assay of vQCs in various snake venoms. (A) Initial rates of QC
reaction using Gln-bNA as substrate versus the amounts of venom added. (B)
The vQC specific activities in 18 venoms: the Elapidae (black), Viperinae
(while), and Crotalinae (gray). The specific activity of human QC, after a 104-
fold reduction, is shown with a striped bar.
Gln-
bNA and PAP. The results are averages based on at least two indepen-
dent experiments.
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 5/11
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
5
1
2
3
4
5
6
7
8
enzyme PAP, showed no detectable fluorescence increase
(sensing the product release), thus other venom compo-
nents did not influence the assay system.
The vQC activities seem to be correlated with the level of
putative substrates in the venoms. Among the 18 selected
several days. We tested the thermal stability of vQCs be-
tween 25 and 65 ꢀC after incubation for 30 min. As shown in
Fig. 2C, the QC activities of C. atrox and P. mucrosquamatus
venoms had little changes at temperatures up to 45 ꢀC, and a
similar transition temperature occurred at approximately
58–62 ꢀC for both venoms. The vQC structures presumably
contain a disulfide bridge near the zinc-catalytic center,
which might contribute toward their thermal stabilities,
similar to the case of hQC (Ruiz-Carrillo et al., 2011).
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
venoms (100
mg each) analyzed, vQC activities of the
viperid venoms were significantly higher than those of the
elapid venoms (Fig. 1B). It is known that metalloproteases
(Watanabe et al., 2003) and bradykinin potentiating pep-
tides (Wermelinger et al., 2005) are common and abundant
in viperid venoms, these proteins or peptides and some
Kunitz-inhibitors of true-viper venoms contain pGlu N-
termini (Tsai et al., 2011). The absence of pGlu at the N-
termini of phospholipases A2 and three-finger toxins
(major elapid venom components) is consistent with the
lower vQC activities of elapid venoms. We also speculate
that vQCs may have lower Km values toward the venom
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
3.4. Effects of metal ions, chelating agents, and synthetic
inhibitors
As shown in Table 1, 5–20 mM EDTA or 20 mM CaCl2 had
little effect on vQC activity, but 1.0 mM 1,10-phenanthroline
caused
a
w60% inactivation, and 5.0 mM 1,10-
phenanthroline fully inactivated the enzyme (data not
shown). These results are similar to those observed for hQC
(Schilling et al., 2003a) and indicate that vQCs and hQCs
have similar zinc-binding environments in the active site.
Notably, 0.1 mM Cuþ2 or Znþ2 could effectively inhibit the
vQC activities by 60–80%, while 5.0 mM Mnþ2 or Niþ2 were
less effective (Table 1). MgCl2 and NaCl inhibited the ac-
tivity of elapid vQCs but enhanced the activity of crotalid
vQCs. Elapid vQCs were also more sensitive to Feþ2 and
Coþ2 and EDTA when compared to crotalid vQCs. These
results possibly are related with structural difference be-
tween the vQCs from the two snake subfamilies. For control
experiments, we found that Cuþ2 (0.1 mM), Znþ2 (0.1 mM),
Feþ2 (1–2 mM), and 1,10-phenanthroline (0.5–1.0 mM) had
little effect on the activity of PAP (Cummins and O’Connor,
1996), indicating that these metal ions and 1,10-
phenanthroline directly affect vQC activity.
The imidazole derivatives were shown to be strong
competitive inhibitors of hQC with no effects on the
auxiliary enzyme PAP (Buchholz et al., 2009; Schilling et al.,
2003a). As shown in Table 2, PBD150 inhibited viperid vQCs
with Ki values in the range of 210–3300 nM, which are
much higher than the Ki value (95 nM) for hQC. The Ki
values for the inhibitors 1-benzylimidazole and N-u-ace-
tylhistamine toward vQCs were 16–59 fold higher than the
Ki values for hQC, and all the inhibitors had weaker effects
proteins (estimated concentration of ꢁ10–100
m
M
in
venom gland) than those revealed by using Gln-
b
NA. The
contents of vQCs in snake venoms are probably regulated
based on the composition of the venoms, as described by a
recent paper reporting the neonate-to-adult changes in the
vQC activities of a pitviper (Gao et al., 2013).
3.2. The pH profile of vQC-catalyzed reaction
Both the C. atrox and P. mucrosquamatus venoms showed
bell-shaped pH profiles of vQC reactions with the optimal
pH values in the range of 7.5–8.0 (Fig. 2A). Similar pH pro-
files were observed for the vQCs of C. godmani, C. d. terrificus,
D. russelii, and M. fulvius (data not shown). Notably, the ac-
tivity pH profiles of vQC are similar to that of hQC (Schilling
et al., 2002) presumably reflecting the pKa value for the a-
95
96
97
98
amino group of the substrates. A previous study indicated
that PAP activity is stable in the pH range of 5.5–8.5 with an
optimum at approximately 7.5–8.0 (Tsuru et al., 1978), and
we also confirmed the pH profile of PAP (Fig. 2B).
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
3.3. Thermal stability of vQCs
The vQC activities were stable in neutral buffers at 4–
25 ꢀC even when the crude venoms were incubated for
Table 1
Effects of metal ions and metal chelators on the activities of four representative vQCs.
Metal or reagent
Concentration mM
Relative vQC activity (%)a
C. atrox
P. mucrosquam.
P. australis
M. fulvius
None (Control)
0
100
5
5
1
2
2
5
5
0.1
0.1
20
0.5
1
100 ꢂ 4.4
117 ꢂ 11
109 ꢂ 2.6
63.6 ꢂ 3.2
80.5
100 ꢂ 8.5
128 ꢂ 7
103 ꢂ 6
54.4 ꢂ 1
81.2
100 ꢂ 11
75.2 ꢂ 11
81.2 ꢂ 11
65.2 ꢂ 9.4
95
100 ꢂ 11
84.8 ꢂ 12
81.3 ꢂ 1.1
60 ꢂ 6.7
80.1
Naþ
Mg2þ
Mn2þ
Fe2þ
Fe2þ
59.3 ꢂ 8
61.8 ꢂ 2.2
71.6 ꢂ 2.1
63.7 ꢂ 0.8
51 ꢂ 6
42 ꢂ 4
65.7 ꢂ 11
44
Co2þ
49 ꢂ 8
46 ꢂ 2
66 ꢂ 2
Co2þ
29.1 ꢂ 2.9
48 ꢂ 3.2
20.7
30.9 ꢂ 5
45 ꢂ 4.9
33.7
36.7 ꢂ 4.3
65.5 ꢂ 0.5
42
Ni2þ
Cu2þ
Zn2þ
36.3 ꢂ 12
90.9 ꢂ 7.8
57
62.6 ꢂ 3.6
91.3 ꢂ 8.8
64
32.6 ꢂ 7.6
55.3 ꢂ 2.4
61
38.5 ꢂ 14
61 ꢂ 2
62
EDTA
1,10-phenanthroline
1,10-phenanthroline
38.4
39.4
43
53
a
Each venom was assayed using Gln-
b
NA in 50 mM Tris–HCl (pH 8.0) at 30 ꢀC. The results are the means ꢂ SD (n ¼ 3), relative to the control (as 100%).
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 6/11
6
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
Table 2
1
2
3
4
5
6
7
8
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
Km and inhibition constants of the vQCs.
Venom or hQC
Km
(
m
M)a
Ki (m
M)a
PBD150
Bi
Ah
C. atrox
81 ꢂ 12
115 ꢂ 5
56 ꢂ 10
71 ꢂ 7
1.13 ꢂ 0.06
1.64 ꢂ 0.14
3.3 ꢂ 0.4
35.2 ꢂ 11.2
35.8 ꢂ 9.6
27.9 ꢂ 0.7
31 ꢂ 10
39.5 ꢂ 1.5
88 ꢂ 36
C. d. terrificus
C. godmani
P. mucrosquamatus
P. elegans
D. russelii
E. carinatus
M. fulvius
97 ꢂ 32
0.50 ꢂ 0.01
0.28 ꢂ 0.06
2.87 ꢂ 0.05
0.21 ꢂ 0.02
>225
53.7 ꢂ 8.7
53.5 ꢂ 19.5
27.4 ꢂ 12.4
41.1 ꢂ 2.3
>225
49 ꢂ 3
16.1 ꢂ 1.7
29 ꢂ 1.5
18.6 ꢂ 1.9
>225
9
39 ꢂ 3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
42 ꢂ 9
232 ꢂ 43
50 ꢂ 4
P. australis
>225
0.095 ꢂ 0.003
>225
0.61 ꢂ 0.09
>225
1.7 ꢂ 0.5
hQCb
76 ꢂ 4
a
The results are means ꢂ SD from two or three experiments.
b
Data are taken from the previous report (Huang et al., 2011).
on elapid vQC. Thus, the synthetic inhibitors are more
specific for hQC than for vQCs.
3.5. Partial purification of a vQC and determination of the N-
terminal sequence
All purification steps were carried out at 4 ꢀC to minimize
possible vQC degradation conferred by venom proteases. On
the basis of the elution volume from a Superdex 200 column
(Fig. 3A) and a calibration curve using four molecular weight
markers (trypsinogen, 25,000; ovalbumin, 45,000; serum
albumin, 67,000; and alcohol dehydrogenase, 150,000), the
molecular weight of C. atrox vQC was estimated to be
46,000 ꢂ 3000, suggesting that it is present as a monomeric
form. After two further ion-exchange chromatographies
(Fig. 3B and C), the vQC protein was only partially purified
with a yield of <0.05%. By SDS-PAGE and silver staining, the
molecular weight of C. atrox vQC was estimated to be 43,000
(Fig. 3C, Inset). The first four N-terminal residues (Arg-Glu-
Asp-Arg) of the vQC were determined by Edman degrada-
tion method, and could be matched to its cDNA deduced
sequence. This is the first time that the N-terminal sequence
of an animal QC has been identified experimentally.
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
3.6. Molecular cloning and full sequences of seven novel vQCs
We successfully cloned and sequenced the vQC cDNAs
for one elapid (i.e., M. fulvius), one true viper (i.e., D. rus-
selii), and five pitvipers in the present study. However,
utilizing the same primers and strategy, we could not
amplify the vQC cDNAs from the venom glands of Echis
carinatus, Bungarus fasciatus, Bungarus candidus, and Wal-
terinnesia aegyptia. All the deduced vQC sequences contain
six conserved Trp residues, Cys110 and Cys136 (which
possibly form a disulfide bond), and three potential N-
glycosylation sites. The deduced amino acid sequences of
these vQCs exhibited ꢁ96% sequence identities, and the
hQC sequence is 75% identical to these vQC sequences, as
aligned in Fig. 4. Interestingly, the nerve growth factors
found in pitviper venoms also exhibit ꢁ96% sequence
identities among one another and are 76% similar to the
human nerve growth factor (Kashima et al., 2002).
Fig. 3. Purification of vQC from C. atrox venom. (A) The venom was frac-
tionated on a Superdex G200 column. Fractions that showed QC activity are
marked with a bar. (B) The vQC fractions were pooled and fractionated on a
Mono-S ion-exchange column. (C) Further purification of C. atrox vQC by a
second Mono S column. The salt gradient was made by mixing buffer A
(50 mM MES, pH6.1) and buffer B (50 mM MES with 1.0 M NaCl). Both insets
show the SDS-PAGE results of isolated vQC fractions. The vQC band in the
inset in (C) was excised and subjected to N-terminal sequence analysis.
For most of the snake species, a single vQC was cloned,
but two vQCs were cloned from a single C. atrox specimen
and the enzymes were designated as Ca-1 and Ca-2. The
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 7/11
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
7
1
2
3
4
5
6
7
8
62
63
64
65
66
67
68
69
70
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
Fig. 4. Amino acid sequence alignment of vQCs with those of human and frog QCs. The numbering starts from the N-termini of the mature vQCs. The species
abbreviations and the QC accession numbers are: Sct (S. c. tergeminus), JF979139; Ca (C. atrox), JF9791323; Tgc (T. gracilis), JF979137; Ab (A. blomhoffi), AB059426;
Pm (P. mucrosquamatus), JF979138; Bj (B. jararaca), AB014769; Cg-1 (C. godmani), JF979134; Dr (D. russelii), JF979136; Mfu (M. fulvius), JF979133; Bd (Boiga
dendrophila), DQ404534; Xenopus (Xenopus laevis), AAH80397; and Human, NP_036545. Secretion signals are in italics and the Cys residues are shaded in gray.
Potential N-glycosylation sites are underlined. The three arrows mark putative zinc-coordinating residues in vQCs according to the structure of hQC (Huang et al.,
2011).
99
sequence of Ca-1 is identical to that of S. c. tergiminus vQC.
Ca-1 and Ca-2 differ by only two amino acid residues at
positions 163 and 264 (Fig. 4). Signal 4.0 algorithm (http://
www.cbs.dtu.dk/services/SignalP) was used to identify the
signal sequence for vQCs. Two most likely signal cleavage
sites appear to be at the peptide bond after Gly(-1) or Ala(-
11), and a less probable site is located after Pro(-6). The
resolved C. atrox vQC N-terminal sequence confirms the
signal cleavage site with the removal of the 33-residue
signal. The predicted mass of mature C. atrox vQC (Ca-1)
is 38,634 Da. Judging from the mass of the purified vQC
(43 kDa, Fig. 3C), the glycans contribute 4–5 kDa to the
molecular mass and present a significant coverage on the
vQC surface. Notably, the C. godmani vQC (designated as Cg-
1) sequence contains special deletion between residues ꢃ3
and 11 (Fig. 4). Cg-1 might have a distinct N-terminal
sequence and the vQC activity could be detected in the
venom (Fig. 1B) although we did not isolate the enzyme.
closely related to that of the species tree and linked all the
Serpentes vQC together. Notably, Cys23 and Tyr209 resi-
dues in all of the crotalid vQCs are substituted with Ser23
and Asn209 in the elapid vQCs (Fig. 4). The tree also con-
firms the high structural similarities among the elapid and
colubrid vQCs and their distinction from the crotalid and
viperid vQCs.
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
3.8. Three-dimensional structure models and proposed
mechanism
The 3D homology models of five representative vQCs
were generated using the reported hQC structure as a
template. The catalytic center of the vQC models is
composed of several loops (Fig. 6A), like that in hQC (Huang
et al., 2005a). Notably, to compare these vQC and hQC
structures, some of the amino acid substitutions are in the
proximity of the catalytic center, especially the His/Tyr
substitution at position 274 and Arg/Gln substitution at
position 249 in vQCs (versus His249 and Gly274 in hQC)
(Fig. 6B). The amino acid substitutions around the active-
site pockets are more basic in vQCs compared to those in
hQC (Fig. 6A). These substitutions at loops surrounding the
3.7. Phylogenetic tree
A QC phylogenetic tree was constructed based on the
protein sequences (Fig. 5). The tree topology appears to be
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 8/11
8
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
1
2
3
4
5
6
7
8
nucleophilic attack on the
gen of the -amino group, (ii) transfer of a proton from the
-amino to the nitrogen of the side chain amide group
g-carboxyl carbon by the nitro-
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
a
a
facilitated by the catalytic Glu residue, and (iii) expulsion of
the amino leaving group prompted by the catalytic Glu
residue or other acidic residue. It is reasonable to assume
that the five conserved acidic residues (i.e., Glu174, Glu175,
Asp131, Asp221, and Asp278) in vQCs are involved in the
zinc-coordination or a possible proton transfer during
catalysis (Huang et al., 2008). In addition, Ser132 and
His292 in vQCs help to stabilize the important hydrogen-
bond network in the active site, while Asp279 and His292
in vQCs, which are flanked by proline-brackets (Kini and
Evans, 1996) are exposed to interact with the substrate.
Since these residues are conserved in vQCs, the catalytic
mechanism of vQC must be similar to that of hQC.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
3.9. Preparation and characterization of recombinant vQCs
Because of the limited amount of vQCs present in crude
venoms, we have used E. coli cells to produce recombinant
vQCs of C. atrox, S. c. tergeminus, P. mucrosquamatus, M.
fulvius, and D. russelii for further characterization. Similar to
the recombinant hQC previously prepared (Huang et al.,
2005b), the N-terminus of each of the recombinant vQC
was designed to start from residue 5 (Fig. 4). Like hQC, the
thioredoxin-fusion proteins of recombinant vQCs could be
expressed from E. coli and folded successfully, in spite of
their lack of N-glycosylation. The yields of the five
recombinants were in the range of 1.5–4.0 mg/L culture
medium, with relatively higher yields for C. atrox and S. c.
tergeminus vQC fusion proteins. The fusion proteins were
Fig. 5. Molecular phylogenetic analysis. In addition to the sequences shown
in Fig. 5, the following QC sequences are included in the data set: Mus
musculus, XP_128770; Rattus norvegicus, XP_233812; Bos taurus, NP_803472;
Canis lupus familiaris, XP_532934; Gallus, XP_419527; Boiga irregularis,
DQ404533; Danio rerio, XP_689638 (assigned as the outgroup). Numbers on
the branches indicate the bootstrap support values for the nodes.
enzyme active site possibly affect the local surface charge
distribution and thus the substrate binding. The mecha-
nism proposed for hQC consists of the following main steps
(Huang et al., 2005a; Jawhar et al., 2011): (i) intramolecular
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Fig. 6. 3D homology models of five vQCs. (A) Comparison of the surface-charge potentials in vQC and hQC structures. Abbreviations used are: S. c. t., S. c. ter-
geminus; C. a., C. atrox (Ca-2); A. b., A. blomhoffi; D. r., D. russelii; B. d., B. dendrophila. The blue color depicts the positive charge and red is for negative charge. Note
that the vQC structures show different charge distributions from the structure of hQC, particularly those around the active site pocket. (B) Superimposition of the
five vQC structures shown in (A). The His/Tyr substitutions at position 274 between the vQC structures are shown with stick model and labeled, and also the
Thr163 and Lys264 in C. atrox vQC. In each structure, the highly acidic catalytic site is indicated with an arrow while the yellow ball depicts the catalytic zinc ion.
(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 9/11
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
9
1
2
3
4
5
6
7
8
purified by a Ni-NTA column. Their homogeneities were
examined by SDS-PAGE (data not shown).
crude venom), e.g. the Km value of D. russelii recombinant
vQC for Gln- NA was 2.6-fold greater than that of the native
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
b
The recombinant vQCs were subjected to kinetic ana-
lyses, and their concentrations were determined by absor-
bance at 280 nm assuming an extinction coefficient of 1.0
for 1.0 mg/ml of the protein solution. We also noticed that
the recombinant vQCs in the buffer were partially inacti-
vated after several freeze–thaw cycles. We thus stored the
recombinant enzymes at <ꢃ20 ꢀC in the presence of 50%
glycerol to avoid the denaturation and inactivation.
Remarkably, the activities of the recombinant vQCs were
almost the same before and after the amino-terminal thi-
oredoxin tag was removed by Factor Xa cleavage, when
assayed at the condition of [S] >> Km. Thus, the thioredoxin
tag had no adverse effects on the kcat. The Km values of the
recombinant C. atrox vQC before and after removal of the
vQC (Tables 2 and 3). We speculate that the glycans in
native vQCs may stabilize the active enzyme conformation
and facilitate the substrate binding. However, the deter-
mined Km of M. fulvius vQC was exceptionally high and
higher than the Km of its recombinant enzyme (Tables 2 and
3). This could be due to the presence of competitive in-
hibitors or other interfering components in the crude
venom.
Three potential Asn-glycosylation sites are present in
vQCs, while hQC contains five such sites (at positions 20,
99, 227, 259 and 269) and only Asn20 and Asn259 are
conserved in hQC and vQCs (Fig. 4). According to the re-
ported 3D structure of hQC, most of the glycosylation sites
are far away from the substrate binding site (Asp279 and
His292) (Ruiz-Carrillo et al., 2011). To study the role of
glycans for vQC catalysis, the isolated C. atrox vQC was
subjected to PNGase F treatment under a non-denaturing
condition. After deglycosylation, the Km was increased
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
tag were determined as 169 ꢂ 28 and 152 ꢂ 32
mM,
respectively, confirming that the N-terminal thioredoxin
tag has little effect on the substrate binding. Therefore, the
fusion proteins were used in subsequent kinetic studies. As
shown in Table 3, kcat values of the vQC fusion proteins are
from 76
mM to 131 mM, and the Vmax was increased from 6.9
rather similar to that of hQC but the specificity indices (kcat
/
to 9.0 M/min. The 1.7-fold increase in the Km value agrees
m
Km) for recombinant vQCs are 1.5–4.8-fold lower than that
of hQC (Huang et al., 2011).
with the higher Km values of the recombinant vQC relative
to the native C. atrox vQC.
3.10. Inhibition kinetics of recombinant vQCs
4. Conclusions
According to the sequence alignments and the 3D-ho-
mology models, Arg249/Gln249 and His274/Tyr274 in the
vQCs (versus His249 and Gly274 in hQC) are close to the
catalytic center (Fig. 6B). Thus, different substitutions at
both positions might result in slightly different shapes of
the QC active site pockets. The relatively higher kcat and
kcat/Km values of the recombinant vQCs from P. mucros-
quamatus and D. russelii suggest possible roles of Tyr274 in
P. mucrosquamatus vQC and Gln249 in D. russelii vQC, in the
catalytic processes. Moreover, the Ki values of Bi and Ah for
the recombinant C. atrox vQC are higher than their Ki values
for the recombinant S. c. tergeminus vQC, even though both
vQCs differ in only two residues at positions 163 and 264
(Fig. 4). Possibly, different substitutions at these positions
may affect the active site conformation.
For the first time, we have cloned the vQC cDNAs from
an Elapidae (M. fulvius) and a Viperinae (D. russelii) species,
and identified the signal peptides and N-termini of vQCs.
This study confirms the presence of QC activities in various
snake venoms, and vQCs can be considered as one of the
snake venom families besides the two dozen families
which have been compiled (Fry et al., 2008). Although hQC
has been associated with septic arthritis and other in-
flammatory symptom (Hellvard et al., 2013), the low con-
tent and only a single form of vQC present in each venom
suggest its house-keeping role for posttranslational modi-
fication of the venom proteins. This is in contrast to some
other venom toxin families which have undergone accel-
erated evolution to generate variants with different func-
tions (Fry et al., 2008).
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Our kinetics and inhibition results suggest that both
vQCs and hQC react in a similar fashion. However, a few
amino acid substitutions in these vQCs may result in the
observed different susceptibilities toward the inhibitors.
hQC showed a broad substrate specificity and a preference
for aromatic amino-acid residues at the second position of
3.11. Effects of deglycosylation on the kinetics of vQC
Notably, the Km values for the other recombinants vQCs
are in the range of 76–169 mM (Table 3) which are signifi-
cantly higher than those of the corresponding vQCs (in the
Table 3
Kinetic properties and inhibitor susceptibilities of five recombinant vQCs and hQC. The assays were performed at 30 ꢀC, pH 8.0 using Gln-
bNA as the
substrate. The results are the means ꢂ SD from three experiments.
Recombinant vQC
Km
(
mM)
kcat (sꢃ1
)
kcat/Km (m Mꢃ1 sꢃ1
)
Ki (mM)
Bi
Ah
C. atrox
169 ꢂ 28
131 ꢂ 28
76 ꢂ 15
103 ꢂ 25
108 ꢂ 11
76 ꢂ 4
10.0
9.0
14.2
18.6
7.1
60
69
186
181
66
285 ꢂ 8
102 ꢂ 12
65 ꢂ 13
162 ꢂ 12
118 ꢂ 9
86 ꢂ 7
S. c. tergeminus
P. mucrosquamatus
D. russelii
69 ꢂ 4.3
47 ꢂ 2.6
72 ꢂ 14
65 ꢂ 3.4
113 ꢂ 5.4
1.7 ꢂ 0.5
M. fulvius
hQCa
21.7 ꢂ 0.5
0.61 ꢂ 0.09
a
Data are taken from the previous report (Huang et al., 2011).
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 10/11
10
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
Huang, K.F., Liu, Y.L., Cheng, W.J., Ko, T.P., Wang, A.H.J., 2005a. Crystal
structures of human glutaminyl cyclase, an enzyme responsible for
protein N-terminal pyroglutamate formation. Proc. Nat. Acad. Sci. (U.
S. A.) 102, 13117–13122.
Huang, K.F., Liu, Y.L., Wang, A.H.J., 2005b. Cloning, expression, charac-
terization, and crystallization of a glutaminyl cyclase from human
bone marrow: a single zinc metalloenzyme. Protein Expr. Purif. 43,
65–72.
Huang, K.F., Wang, U.R., Chang, E.C., Chou, T.L., Wang, A.H.J., 2008. A
conserved hydrogen bond network in the catalytic center of animal
glutaminyl cyclases is critical for catalysis. Biochem. J. 411, 181–190.
Huang, K.F., Liaw, S.S., Huang, W.L., Chiam, C.Y., Lom, Y.C., Chen, Y.L.,
Wang, A.H.J., 2011. Structures of human Golgi resident glutaminyl
cyclase and its complexes with inhibitors reveal a large loop move-
ment upon inhibitor binding. J. Biol. Chem. 286, 12439–12449.
1
2
3
4
5
6
7
8
its substrates (Schilling et al., 2003b; Seifert et al., 2009),
but many natural substrates of vQC contain a basic or a
hydrophilic residue at the second amino acid position
(Watanabe et al., 2003; Tsuru et al., 1978). We also found
that the recombinant vQCs have higher Km values than the
native vQCs which could be attributed to the lack of
glycosylation on the recombinants. Our results thus pro-
vide basic biochemical data for vQCs and new insights into
the similarities and differences between the structures and
functions of vQC and hQC.
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
Ethical statement
Jawhar, S., Wirths, O., Bayer, T.A., 2011. Pyroglutamate Amyloid-b (Ab): a
hatchet man in Alzheimer disease. J. Biol. Chem. 286, 38825–38832.
Kashima, S., Soares, A.M., Roberto, P.G., Pereira, J.O., Astolfi-Filho, S.,
Cintra, A.O., Fontes, M.R., Giglio, J.R., de Castro França, S., 2002. cDNA
sequence and molecular modeling of a nerve growth factor from
Bothrops jararacussu venomous gland. Biochimie 84, 675–680.
Kini, R.M., Evans, H.J., 1996. Prediction of potential protein-protein
interaction sites from amino acid sequence, identification of a fibrin
polymerization site. FEBS Lett. 385, 81–86.
Konno, K., Picolo, G., Gutierrez, V.P., Brigatte, P., Zambelli, V.O.,
Camargo, A.C., Cury, Y., 2008. Crotalphine, a novel potent analgesic
peptide from the venom of the South American rattlesnake Crotalus
durissus terrificus. Peptides 29, 1293–1304.
Laskowski, R.A., Moss, D.S., Thornton, J.M., 1993. Main-chain bond lengths
and bond angles in protein structures. J. Mol. Biol. 231, 1049–1067.
Lüthy, R., Bowie, J.U., Eisenberg, D., 1992. Assessment of protein models
with three-dimensional profiles. Nature 356, 83–85.
Maniatis, T., Fritsch, E.F., Sambrook, J., 1989. Molecular Cloning: a Labo-
ratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor.
Mullis, K.B., Faloona, F.A., 1987. Specific synthesis of DNA in vitro via a po-
lymerase catalyzed chain reaction. Methods Enzymol. 155, 335–350.
Murayama, N., Hayashi, M., Ohi, H., Ferreira, L., Hermann, V., Saito, H.,
Fujita, Y., Higuchi, S., Fernandes, B., Yamane, T., de Camargo, A., 1997.
Cloning and sequence analysis of a Bothrops jararaca cDNA encoding a
precursor of seven bradykinin-potentiating peptides and a C-type
natriuretic peptide. Proc. Nat. Acad. Sci. (U. S. A.) 94, 1189–1193.
Pawlak, J., Kini, M.R., 2006. Snake venom glutaminyl cyclase. Toxicon 48,
278–286.
Ruiz-Carrillo, D., Koch, B., Parthier, C., Wermann, M., Dambe, T.,
Buchholz, M., Ludwig, H.H., Heiser, U., Rahfeld, J.U., Stubbs, M.T.,
Schilling, S., Demuth, H.U., 2011. Structures of glycosylated mamma-
lian glutaminyl cyclases reveal conformational variability near the
active center. Biochemistry 50, 6280–6288.
Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack, C.,
Demuth, H.U., 2002. Continuous spectrometric assays for glutaminyl
cyclase activity. Anal. Biochem. 303, 49–56.
Schilling, S., Niestroj, A.J., Rahfeld, J.U., Hoffmann, T., Wermann, M.,
Zunkel, K., Wasternack, C., Demuth, H.U., 2003a. Identification of
human glutaminyl cyclase as a metalloenzyme. Potent inhibition by
imidazole derivatives and heterocyclic chelators. J. Biol. Chem. 278,
49773–49779.
The experiments carried out in the present report did
not involve live animals.
Acknowledgments
This study was supported by grants from Academia
Sinica and National Science Council Taiwan (grant No NSC
99-2311-B-001-016-MY3) of Taiwan, ROC. We thank Prof.
Andrew H.-J. Wang for sharing the reagents to perform QC
kinetic studies, and Prof. Anthony T. Tu for the gift of C.
atrox venom.
Q5
Q6
Q7
Conflict of interest statement
The authors declare no conflict of interest.
References
Buchholz, M., Hamann, A., Aust, S., Brandt, W., Böhme, L., Hoffmann, T.,
Schilling, S., Demuth, H.U., Heiser, U., 2009. Inhibitors for human
glutaminyl cyclase by structure based design and bio-isosteric
replacement. J. Med. Chem. 52, 7069–7080.
Chen, Y.H., Wang, Y.M., Hseu, M.J., Tsai, I.H., 2004. Molecular evolution
and structure-function relationships of crotoxin-like and asparagine-
6-containing phospholipases A2 in pitviper venoms. Biochem. J. 381,
25–34.
95
96
97
98
99
Cummins, P.M., O’Connor, B., 1996. Bovine brain pyroglutamyl amino-
peptidase (type-1): purification and characterisation of
a
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
neuropeptide-inactivating peptidase. Int. J. Biochem. Cell Biol. 28,
883–893.
Faure, G., Xu, H., Saul, F.A., 2011. Crystal structure of crotoxin reveals key
residues involved in the stability and toxicity of this potent hetero-
dimeric b-neurotoxin. J. Mol. Biol. 412, 176–191.
Felsenstein, J., 1992. PHYLIP: the PHYLogeny Inference Package, Version
3.573. Computer Program Distributed by the University of Washing-
ton, Department of Genetics, Seattle.
Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using
the bootstrap. Evolution 39, 783–791.
Ferreira, S.H., 1965. A bradykinin-potentiating factor (BPF) present in the
venom of Bothrops jararaca. Br. J. Pharmacol. 24, 163–169.
Fry, B.G., Scheib, H., van der Weerd, L., Young, B., McNaughtan, J., Ryan
Ramjan, S.F., Vidal, N., Poelmann, R.E., Norman, J.A., 2008. Evolution of
an arsenal: structural and functional diversification of the venom
system in the advanced snakes. Mol. Cell Proteomics 7, 215–246.
Gao, J.K., Qu, Y.F., Zhang, X.Q., He, Y., Ji, X., 2013. Neonate-to-adult tran-
sition of snake venomics in the short-tailed pitviper, Gloydius brevi-
caudus. J. Proteomics 84, 148–157.
Harvey, A.L., Robertson, B., 2004. Dendrotoxins: structure-activity re-
lationships and effects on potassium ion channels. Curr. Med. Chem.
23, 3065–3072.
Hellvard, A., Maresz, K., Schilling, S., Graubner, S., Heiser, U., Jonsson, R.,
Cynis, H., Demuth, H.U., Potempa, J., Mydel, P., 2013. Glutaminyl cy-
clases as novel targets for the treatment of septic arthritis. J. Infect.
Dis. 207, 768–777.
Huang, K.F., Chiou, S.H., Ko, T.P., Wang, A.H.J., 2002. Determinants of the
inhibition of a Taiwan habu venom metalloproteinase by its endog-
enous inhibitors revealed by X-ray. Eur. J. Biochem. 269, 3047–3056.
Schilling, S., Manhart, S., Hoffmann, T., Ludwig, H.H., Wasternack, C.,
Demuth, H.U., 2003b. Substrate specificity of glutaminyl cyclases from
plants and animals. Biol. Chem. 384, 1583–1592.
Schilling, S., Lindner, C., Koch, B., Wermann, M., Rahfeld, J.U., von
Bohlen, A., Rudolph, T., Reuter, G., Demuth, H.U., 2007. Isolation and
characterization of glutaminyl cyclases from Drosophila: evidence for
enzyme forms with different subcellular localization. Biochemistry
46, 10921–10930.
Segel, I.H., 1993. Enzyme Kinetics: Behavior and Analysis of Rapid Equi-
librium and Steady-state Enzyme System. John Wiley & Sons, New
York, pp. 100–118.
Seifert, F., Schulz, K., Koch, B., Manhart, S., Demuth, H.U., Schilling, S.,
2009. Glutaminyl cyclases display significant catalytic proficiency for
glutamyl substrates. Biochemistry 48, 11831–11833.
Stephan, A., Wermann, M., von Bohlen, A., Koch, B., Cynis, H.,
Demuth, H.U., Schilling, S., 2009. Mammalian glutaminyl cyclases and
their isoenzymes have identical enzymatic characteristics. FEBS J. 276,
6522–6536.
Tsai, I.H., Lu, P.J., Wang, Y.M., Ho, C.L., Liaw, L.L., 1995. Molecular cloning
and characterization of
a neurotoxic phospholipase A2 from the
venom of Taiwan habu (Trimeresurus mucrosquamatus). Biochem. J.
311, 895–900.
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012

TOXCON4820_proof ■ 12 May 2014 ■ 11/11
Y.-M. Wang et al. / Toxicon xxx (2014) 1–11
11
Tsai, I.H., Chen, Y.H., Wang, Y.M., Tu, M.C., Tu, A.T., 2001. Purification,
sequencing and phylogenetic analyses of novel Lys-49 phospholi-
pases A2 from venoms of rattlesnakes and other Pitvipers. Arch.
Biochem. Biophys. 394, 236–244.
Tsai, I.H., Tsai, H.Y., Wang, Y.M., Tun-Pe, Warrell, D.A., 2007. Venom
phospholipases of Russell’s vipers from Myanmar and eastern India:
cloning, characterization and phylogeographic analysis. Biochim.
Biophys. Acta 1774, 1020–1028.
Tsai, I.H., Wang, Y.M., Cheng, A.C., Starkov, V., Osipov, A., Nikitin, I.,
Makarova, Y., Ziganshin, R., Utkin, Y., 2011. cDNA cloning, structural,
and functional analyses of venom phospholipases A2 and a Kunitz-
type protease inhibitor from steppe viper Vipera ursinii renardi. Tox-
icon 57, 332–341.
Tsai, I.H., Wang, Y.M., Tsai, T.S., Tu, M.C., Chang, H.C., Chen, H.S., 2012.
Cloning and characterization of Trimeresurus gracilis venom phos-
pholipases A2: comparison with Ovophis okinavensis venom and the
systematic implications. Toxicon 59, 151–157.
Tsuru, D., Fujiwara, K., Kado, K., 1978. Purification and characterization of
L-pyrrolidonecarboxylate peptidase from Bacillus amyloliquefacien. J.
Biochem. 84, 467–476.
Wagstaff, S.C., Favreau, P., Cheneval, O., Laing, G.D., Wilkinson, M.C.,
Miller, R.L., Stöcklin, R., Harrison, R.A., 2008. Molecular characteriza-
tion of endogenous snake venom metalloproteinase inhibitors. Bio-
chem. Biophys. Res. Commun. 365, 650–656.
Watanabe, L., Shannon, J.D., Valente, R.H., Rucavado, A., Alape-Girón, A.,
Kamiguti, A.S., Theakston, R.D.G., Fox, J.W., Gutiérrez, J.M., Arni, R.K.,
2003. Amino acid sequence and crystal structure of BaP1, a metal-
loproteinase from Bothrops asper snake venom that exerts multiple
tissue-damaging activities. Protein Sci. 12, 2273–2281.
Wermelinger, L.S., Dutra, D.L.S., Oliveira-Carvalho, A.L., Soares, M.R.,
Bloch Jr., C., Zingali, R.B., 2005. Fast analysis of low molecular mass
compounds present in snake venom: identification of ten new
pyroglutamate-containing peptides. Rapid Commun. Mass Spectrom.
19, 1703–1708.
1
2
3
4
5
6
7
8
14
15
16
17
18
19
20
21
22
23
24
25
26
9
10
11
12
13
Please cite this article in press as: Wang, Y.-M., et al., Snake venom glutaminyl cyclases: Purification, cloning, kinetic study,
recombinant expression, and comparison with the human enzyme, Toxicon (2014), http://dx.doi.org/10.1016/
j.toxicon.2014.04.012