Development of new inhibitors for N-acylethanolamine-hydrolyzing acid amidase as promising tool against bladder cancer

Article abstract of DOI:10.1016/j.bmc.2016.12.042

The endocannabinoid system is a signaling system involved in a wide range of biological effects. Literature strongly suggests the endocannabinoid system role in the pathogenesis of cancer and that its pharmacological activation produces therapeutic benefi

Full text of DOI:10.1016/j.bmc.2016.12.042

Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Development of new inhibitors for N-acylethanolamine-hydrolyzing
acid amidase as promising tool against bladder cancer
Riccardo Vago a,b, Arianna Bettiga a, Andrea Salonia a,b, Pierangela Ciuffreda c, Roberta Ottria c,
⇑
a Urological Research Institute, Division of Experimental Oncology, IRCCS San Raffaele Hospital, Via Olgettina 60, Milan, Italy
b Università Vita-Salute San Raffaele, Via Olgettina 60, Milano, Italy
c Dipartimento di Scienze Biomediche e Cliniche ‘‘Luigi Sacco”, Via G.B. Grassi 74, Università degli Studi di Milano, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The endocannabinoid system is a signaling system involved in a wide range of biological effects.
Literature strongly suggests the endocannabinoid system role in the pathogenesis of cancer and that
its pharmacological activation produces therapeutic benefits. Last research promotes the endocannabi-
noid system modulation by inhibition of endocannabinoids hydrolytic enzymes instead of direct activa-
tion of endocannabinoid receptors to avoid detrimental effects on cognition and motor control. Here we
report the identification of N-acylethanolamine-hydrolyzing acid amidase (NAAA) inhibitors able to
reduce cell proliferation and migration and cause cell death on different bladder cancer cell lines.
These molecules were designed, synthesized and characterized and active compounds were selected
by a fluorescence high-throughput screening method set-up on human recombinant NAAA that also
allows to characterize the mechanism of inhibition. Together our results suggest an important role for
NAAA in cell migration and in inducing tumor cell death promoting this enzyme as pharmacological tar-
get against bladder cancer.
Received 8 July 2016
Revised 14 December 2016
Accepted 23 December 2016
Available online xxxx
Keywords:
Endocannabinoid
Enzyme inhibition
N-acylethanolamine-hydrolyzing acid
amidase
Bladder cancer
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
in cell proliferation demonstrating that targeting ECS can affect
cancer growth.3 ECS has been demonstrated to act by several dif-
Endocannabinoids (EC) and N-acylethanolamides (NAEs) are
lipid mediators of an endogenous signaling system, the endo-
cannabinoid system (ECS), transiently active, involved in a wide
range of biological effects. The ECS includes EC, NAEs, their specific
receptors and biosynthetic and metabolic pathways.1 EC and NAEs
biological effects, indeed, depend on their life span in extracellular
space limited by receptor internalization upon interaction and
enzymatic degradation. Arachidonoylethanolamide (AEA) and 2-
arachidonoylglycerol (2-AG) are the earliest described lipid deriva-
tives from the EC family that also comprise other arachidonic acid
derivatives. AEA also belongs to NAEs, which comprises non-endo-
cannabinoids as palmitoylethanolamide (PEA) and oleoylethanola-
mide (OEA), while 2-AG does not. The literature strongly suggests a
role for the ECS in the pathogenesis of cancer promoting ECS as
promising therapeutic target for cancer treatment.2 Latest studies
propose the involvement of EC and NAEs in maintaining balance
ferent cellular mechanisms including inhibition of cell prolifera-
tion, migration and progression, inhibition of angiogenesis and
promotion of apoptosis and/or cell cycle arrest. Although numer-
ous evidences that pharmacological activation of ECS by cannabi-
noid receptors (CB) can elicit therapeutically beneficial effects,
the associated detrimental effects of CB agonists on cognition
and motor control limit their extensive use as pharmaceuticals.
Accordingly, pharmacological blockade of the EC hydrolytic
enzymes such as fatty acid amide hydrolase (FAAH) or monoacyl-
glycerol lipase (MAGL) has emerged as a potentially attractive
strategy to increase EC signaling and retain advantages of CB acti-
vation avoiding undesirable side effects. At the same time, the inhi-
bition of EC and NAEs hydrolyzing activity will reduce production
of eicosanoids and free fatty acids, important pro-inflammatory
mediators.
In last few years, the availability of selective inhibitors for EC
hydrolytic enzymes has allowed the identification of EC role in some
pathologies, and provided preclinical data supporting potential
applications for this class of compounds.4 Moreover, the overexpres-
sion of FAAH or MAGL has been demonstrated in aggressive
cancers.5 More recently, N-acylethanolamine-hydrolyzing acid ami-
dase (NAAA),6 a new NAEs degrading enzyme, not related to FAAH,
⇑
Corresponding author at: Università degli Studi di Milano-Dipartimento di
Scienze Biomediche e Cliniche ‘‘Luigi Sacco”, Via G.B. Grassi 74, 20157 Milano, Italy.
E-mail addresses: vago.riccardo@hsr.it (R. Vago), bettiga.arianna@hsr.it
(A. Bettiga), andrea.salonia@hsr.it (A. Salonia), pierangela.ciuffreda@unimi.it
(P. Ciuffreda), roberta.ottria@guest.unimi.it (R. Ottria).
http://dx.doi.org/10.1016/j.bmc.2016.12.042
0968-0896/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

2
R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
belonging to the choloylglycine hydrolase family, has been cloned
and characterized. NAAA shows a pH optimum at 4.5–5 and is
almost inactive at alkaline or neutral pH; it has been shown to be
highly expressed in macrophages and localized in lysosomes of var-
ious rat tissues including lung, spleen, intestine and brain. Although
FAAH and NAAA can hydrolyze various NAEs, the substrates pre-
ferred are different; indeed, NAAA hydrolyses preferentially palmi-
toylethanolamide (PEA) and monounsaturated NAEs like
oleoylethanolamide (OEA). In the last years, some selective NAAA
inhibitors have been developed as anti-inflammatory7 and/or anal-
gesic8 compounds promoting this enzyme as a promising pharma-
cological target in this area. Bladder cancer is one of the most
common and expensive malignancies in developed countries due
to the need for continuous monitoring and recurrence treatment.
Its treatment remains a challenge despite significant improvements
in preventing disease progression, because of its high rate of recur-
rence and potential progression, regardless of treatment with sur-
gery, chemotherapy, or immunotherapy.9 Alternative and/or
complementary strategies to state-of-art therapies are urgently
needed to limit the recurrence and improve survival. Although the
ECS has been extensively studied in the field of cancer,10 to date
there is only one study regarding bladder cancer and the ECS.11 Gas-
peri and co-workers demonstrated that functional CB-receptors are
expressed in bladder cancer cells and that their stimulation by an
inverse agonist exerts anti-proliferative effects. Given the presence
of ECS components in bladder cancer cells and their involvement
in cancer cell proliferation, our aim was to explore the therapeutic
utility of NAAA inhibitors in bladder cancer. In the present work,
we report the synthesis and the complete characterization of eight
new compounds with the evaluation of their inhibitory activity
toward human NAAA. Moreover, we show that the most effective
NAAA inhibitors are able to significantly inhibit migration and dras-
tically reduce viability and proliferation of bladder cancer cells
resembling different disease stages and proliferation rate.
probe for NAAA enzymatic activity because it is hydrolyzed by
NAAA, producing the fluorescent 7-amino-4-methylcoumarin,
with a binding affinity similar to or slightly greater than that for
PEA.16 A fluorescent-based assay for NAAA activity was selected
to assess inhibitory activity of new synthesized compounds. This
kind of methods are simple, sensitive, time saving, with rapid read
out and cheap, all characteristics suitable for high throughput
screening of small or large library of new compounds. According
to this finding PAMCA was freshly synthesized from palmitic acid
and 7-methylamino coumarine and, after purification and com-
plete characterization, it was used to set-up the fluorescent-based
assay on human recombinant NAAA (hrNAAA, purity >95%). Enzy-
matic reaction conditions were optimized for the screening of new
designed inhibitors using compound 1 and 2 selected as models of
aromatic and linear palmitamine derivatives and dimethylsolfox-
ide (DMSO) as blank. Different quantities of the enzyme and con-
centrations of PAMCA have been assessed and finally 25 ng of
hrNAAA and 10 lM PAMCA have been selected as the amount with
better fluorescence production in reaction time. In a typical analy-
sis, various concentrations of test compound dissolved in DMSO or
DMSO alone were added to the reaction buffer containing hrNAAA.
After incubation with the inhibitor, PAMCA, was added and the
assay ran for 5 h. The kinetic curves obtained in optimized reaction
condition with compound 1 at four concentrations (5, 10, 50,
100 lM final concentrations per well) showed that hydrolysis rate
of PAMCA by NAAA was linear for all 5 h (Fig. 2). To obtain a useful
tool for high throughput screening we decided to stop the assay at
2.5 h. At first, we selected three scalar logarithmic concentrations
to test the activity of our potential inhibitors of NAAA. Compounds
at 1, 10 and 100 lM final concentrations were incubated together
with NAAA and fluorescence intensity was monitored after addi-
tion of the substrate PAMCA. Residual enzymatic activity was cal-
culated from slops of kinetic curves and expressed as percentage.
From this preliminary screening, compounds 3–7 showed a con-
centration-dependent inhibitory activity while only 2-naphthoic
acid derivative 8 did not show any inhibition toward NAAA. Conse-
quently, compounds 1–7 were selected for further investigation.
2. Results and discussion
2.1. Design and synthesis of new NAAA inhibitors
2.3. Characterization of active compounds and structure-activity
relationship consideration
To date, some different classes of compounds have been
reported to selectively inhibit NAAA. Vandevoorde and coworkers
firstly proposed retroamides, esters and retroesters of palmitic acid
as NAAA inhibitors.12 The most potent inhibitor to date described
from the class of palmitic acid-retroamides are N-pentadecylben-
zylamide (a) and N-pentadecylhydroxypropylamide (b) shown in
Fig. 1.13 Starting from data reported in literature on NAAA palmitic
acid-retroamide inhibitors,14 we decided to synthesize a small
library of retroamide derivatives of palmitamine to explore how
structural modification such as the presence of one or more het-
eroatoms (hydrophilic interactions promoters), and their stereo-
chemical position, or double aromatic ring could influence
inhibitory ability. In detail we synthesized benzoic (1), hydrox-
ypropylic (2), salicylic (3), R-mandelic (4) and S-mandelic (5),
pyridinecarboxylic (6), 2,5-diidroxybenzoic (7) and 2-naphthoic
(8) acid derivatives of hexadecylamine (Fig. 1). All compounds
Five points assay (10-25-50-75 and 100 lM final concentra-
tions) was performed with compounds 1–7 to obtain inhibition
curves and IC50 values. From the slop of kinetic curves, residual
enzymatic activity was calculated and IC50 values were obtained
by non-linear regression analysis of Log[concentration]/inhibition
curves. IC50 values found in our experimental conditions are listed
in Fig. 1 along with chemical structures of synthesized compounds.
Compounds 1 and 2 showed IC50 values of 45 1.6 and 34 1.2 lM
respectively. Interestingly compounds 3 and 4 exert three and two
fold major inhibitory activity toward NAAA with IC50 values of
11 1.0 and 20 0.9
of compound is similar to compound
38 1.0 M). Notably, compound 4, the R-mandelic acid deriva-
tive, displayed three fold higher activity than its enantiomer 5
(IC50 78 2.8 M). Finally compound 6 was the less active in
hrNAAA inhibition, displaying an IC50 of 98 1.0 M. Starting from
lM respectively, while inhibitory activity
7
2
(IC50 value of
l
l
were synthesized by
a one-step coupling reaction of the
l
appropriate carboxylic acid with hexadecylamine, according to
the procedure previously described for the synthesis of
N-acylethanolamines15 with yields that range from 45% to 85%.
these results compounds 3, 4 and 7, along with compound 2, the
only linear palmitamine derivative, were selected to examine their
inhibition mechanism. The effects of different concentrations of
PAMCA with hrNAAA in the presence or absence of inhibitors
2.2. Selection of inhibitors by a high-throughput screening fluorimetric
method
(100 lM) were analyzed. Results reported in the Lineweaver-Burk
plot, shown in Fig. 3, demonstrate a competitive inhibition for all
new compounds tested.
It has been recently demonstrated that the non-fluorescent
N-(4-methyl coumarin)palmitamide (PAMCA) can be a useful
We also examined the reversibility of the inhibition incubating
NAAA with selected compounds for 20 min. After incubation, solu-
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
3
O
N
H
a. N-pentadecylbenzylamide
O
N
H
OH
b. N-pentadecylhydroxypropylamide
= C15H31NH2
NH2
c. N-pentadecylamine
O
O
C15H31
C15H31
N
H
N
H
OH
2. IC50 34 1.2µM
1. IC50 45 1.6µM
O
OH
O
C15H31
C15H31
N
H
N
H
OH
3. IC50 11 1.0µM
4. IC50 20 0.9µM
O
O
C15H31
C15H31
N
H
N
H
N
OH
6
. IC50 98 1.0µM
O
5
. IC50 78 2.8µM
O
OH
OH
C15H31
C15H31
N
H
N
H
7
. IC50 38 1.0µM
8
. IC50 >>100µM
Fig. 1. Chemical structures and IC50 (lM) values of synthesized compounds in direct inhibition of hrNAAA.
tions were diluted twenty fold and new amounts of inhibitors were
added; then reaction was started by PAMCA addition. Results show
that the inhibitory activity of compounds 2 and 4 was determined
mostly by the second concentration added rather than by its initial
concentration, suggesting the reversibility of the inhibition (Fig. 4).
Unfortunately, this experiment has not yielded reliable results for
the samples 3 and 7 because of the intrinsic fluorescence of the
molecules. Taken together these results suggest that the presence
of one aromatic ring maintain the inhibitory activity of the mole-
cule while a double ring results in a loss of activity. Moreover,
the presence of one or two hydroxyl groups on the ring, promoter
of hydrophilic interactions with the enzyme, ameliorates the inhi-
bition power.
2.4. NAAA inhibitors affect bladder cancer cells viability
Since ECS has been demonstrated to play a critical role in
tumors17 and enzymes involved in EC metabolism have been found
more expressed in cancer,18,19 we tested the effect of our NAAA
Fig. 2. Kinetic curves of inhibition of hrNAAA by compound 1 at four concentra-
tions: 5-10-50-100 M (RFU = Relative fluorescence Emission).
l
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

4
R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
aggressive bladder tumors. The compound 4 is, instead, the best
performers on RT4 cells, but is inactive on HT-1376.
2.5. Effects of NAAA inhibitors on the bladder cancer cell migration
Tumor cell migration is an important step for the spread of can-
cer and ECS has been proven to be involved in this process.17 Since
RT112 cells are characterized by rapid migration, we investigated
the role of NAAA inhibitors on cell migration in a classical wound
healing assay. Sub-confluent cells were scratched and incubated
with compounds 3 and 7, the overall best inhibitors of NAAA activ-
ity and cell viability, or with compound 2 as the only linear palmi-
tamine derivative. After 24 h treatment, wound closure was
calculated. As shown in Fig. 5, untreated RT112 cells migrated very
fast and almost completely covered the scratch after 24 h. Com-
pound 2 exerted a very modest effect, while compounds 3 and 7
significantly reduced cell migration (50% and 80% respectively).
These results suggest an important role for NAAA in the cell migra-
tion and its inhibition as a potential approach to reduce cell migra-
tion and thus tumor dissemination. The overall data demonstrate
that a new class of NAAA inhibitors can be effective in inducing
tumor cell death, paving the way for a future development of these
compounds for cancer treatment.
Fig. 3. Lineweaver-Burk plot of the inhibition of hrNAAA by compounds 2, 3, 4 and
7.
inhibitors on human bladder cancer cell lines. To address tumor
heterogeneity, we used different cell lines resembling various
stages of the disease, named RT4, RT112, 5637 and HT-1376. They
differ in size, proliferation rate and motility, making a sample
panel for the analysis. We assayed cell viability by incubating cells
with serial logarithmic dilutions of NAAA inhibitors ranking from
2.6. Effects of inhibitors on endocannabinoids levels in bladder cancer
cells
To assess how the treatment of bladder cancer cells with syn-
thesized inhibitor affected the EC and NAEs levels, the following
EC and NAEs N-palmitoylethanolamide (PEA, C16:0), N-stearoy-
lethanolamide (SEA, C18:0), N-oleoylethanolamide (OEA, C18:1),
anandamide (N-arachidonoylethanolamide, AEA, C20:4), N-lino-
100 to 0.1 lM for 72 h and then a MTT assay was performed
(Table 1). As for the cell-free assay, the inhibitors showed different
activity and compound 3 confirmed to be the most effective in
reducing cell proliferation and causing cell death with an IC50
much lower than the others in all cell lines analyzed. Compound
1 was the less active displaying a definable IC50 only in 5637 cells
experiment. Compound 2 was effective only on RT112 and 5637
cells proliferation displaying a quite high IC50 value. Compounds
4 and 5 corroborated previous data, displaying increased activity
compared to compound 2. Notably, compound 7 resulted highly
active, in a very similar extent to compound 3. HT-1376 cells were
originally derived from a transitional cell carcinoma of grade III,
the highest in the panel; they bear more invasive and metastatic
properties than stage II 563720 and are sensitive only to com-
pounds 3 and 7. Nevertheless, they are little (3) or not active (7)
on grade I RT4, making them suitable for the treatment of more
leoylethanolamide (LEA, C18:2), N-a-linolenoylethanolamide
(LNEA, C18:3) and N-eicosapentaenoylethanolamide (EPEA,
C20:5) have been quantified by HPLC-MS/MS analysis. Cells and
supernatants were extracted and purified as described in the
experimental part and quantification results, expressed in ng/ml,
are reported in Fig. 6. Fig. 6 lacks of EPEA because it was unde-
tectable or its levels were under the limit of quantification of the
method in all samples. At our experimental conditions, 24 h of
treatment with 40 lM inhibitor on RT112 cell line, all assessed
inhibitors affect concentrations of N-acylethanolamides of satu-
rated or monounsaturated acids while polyunsaturated acid
derivatives concentrations remained unvaried. In detail all com-
Fig. 4. Reversibility of the inhibition of hrNAAA by compounds 2, and 4. *p < 0.05.
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
5
Table 1
Cell viability of bladder cancer cell lines treated with NAAA inhibitors. RT4, RT112, 5637 and HT-1376 cells were incubated with compounds 1–7 and after 72 h MTT assay was
performed. IC50 is indicated standard deviations. NM = Not Measurable.
RT4
RT112
5637
HT-1376
1 (l
2 (l
3 (l
4 (l
5 (l
6 (l
7 (l
M)
M)
M)
M)
M)
M)
M)
N M
N M
63.33 20.82
3.50 2.18
>100
N M
45.00 22.32
40.00 10.95
11.25 5.97
25.00 9.13
18.75 8.54
51.25 16.52
7.25 2.22
N M
N M
10.25 3.69
N M
N M
58.33 23.63
5.75 1.63
28.33 2.89
25.00 0.00
76.25 7.50
7.00 0.96
l
M
41.67 5.77
N M
N M
9.50 1.00
Fig. 5. Effects of NAAA inhibitors on the bladder cancer cell migration. The migration capacity of RT112 bladder cancer cells were tested in the presence of NAAA inhibitors 2,
3 and 7 in a wound healing assay. Panels (A) shows representative pictures of RT112 cells at time 0 and after 24 h treatment. The migration was determined by the rate of cells
filling the scratched area and the result represented as mean standard deviation (B). Ctr.: negative control without inhibitors. **p value < 0.01; n.s.: not significant.
RT112 by all assessed compounds. Together these results demon-
strate the selective inhibition of NAAA in intact bladder cancer cells
since they increase the levels of PEA, SEA and OEA without affect-
ing concentrations of the other EC and NAEs AEA, LEA and LNEA.
2.7. Statistics
All experiments on hrNAAA and on cell lines have been per-
formed at least three times in triplicate and results are expressed
as mean of obtained results standard deviation. Significance of
experiments was assessed by Student’s T-test.
3. Conclusions
Here we report the design and synthesis of a small library of ret-
roamide derivatives of palmitamine with the aim to identify new
NAAA inhibitors. Moreover, we describe the set-up of a fluorescent
base method to assess enzymatic activity, fast, sensitive and cheap,
that allow the high-throughput screening of newly synthesized
compounds on human recombinant NAAA and the characterization
of the inhibition mechanism of active molecules. In addition, we
demonstrate that three of new compounds are active on hrNAAA
at low concentration and are able to reduce cell proliferation and
Fig. 6. Quantifications of EC and NAEs by HPLC-MS/MS analysis. EC and NAEs levels
in RT112 and 5637 bladder cancer cells. *p < 0.05, **p < 0.01.
pound caused a statistically significant increment of PEA in RT112
cell line. OEA was incremented by the three molecules assessed,
but only variation linked to compound 2 and 3 treatment were sta-
tistically significant. Finally, SEA concentration was incremented in
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

6
R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
cause cell death on different bladder cancer cell lines. Obtained
results from cell migration experiments demonstrate that com-
pounds 3 and 7 significantly reduce cell migration while inhibitor
2 shows a very low effect suggesting an important role for NAAA in
the cell migration. The overall data demonstrate that a new class of
NAAA inhibitors can induce tumor cell death and reduce cell
migration promoting NAAA as new target in bladder cancer man-
agement and paving the way for a future development of these
compounds for cancer treatment. Results from in vitro experi-
ments on hrNAAA allow considerations on structure activity rela-
tionship and lay the bases to optimize additional molecules as
NAAA inhibitors. Finally, the increase of PEA, SEA and OEA without
modifications in the concentrations of the other EC and NAEs AEA,
LEA and LNEA is an additional evidence of the selective inhibition
of NAAA in intact bladder cancer cells. Moreover, these results
demonstrate that effects on survival and migration of bladder can-
cer cells are due to NAAA inhibition. Comparing our data with the
to date knowledge on the involvement of ECS in cancer manage-
ment we can hypothesize that PEA, OEA and SEA increase could
affect bladder cancer by several mechanisms as apoptosis promo-
tion, intensification of the apoptotic response induced by AEA, or
angiogenesis modulation. Moreover OEA has been recognized as
acid ceramidase (AC) inhibitor21 concurring to sensitize prostate
bladder cells to EC–induced apoptosis. Further experiments are
required and will be devoted to confirm the hypothesized mecha-
nism of action of NAAA inhibitors.
was continued until TLC confirmed the completion of the reaction
(1–6 h). The reaction mixture was diluted with CH2Cl2 and washed
with 5% HCl, saturated NaHCO3 and brine. The organic layers were
collected, dried over anhydrous Na2SO4 and filtered. The solvent
was evaporated under reduced pressure and crude purified by flash
chromatography to give white solids and reaction yields range
from 45% to 85%.
4.1.1. N-Hexadecylbenzamide (1)
The compound 1 was prepared from benzoic acid (50 mg,
0.41 mmol) as white solid 122 mg (0.35 mmol, 85%) Rf 0.42 (Hex-
ane/Acetone 85:15). Mp 83-85 °C. 1H NMR: d 0.90 (3H, t, J = 6.9,
160-CH3), 1.22–1.40 (24 H, m, 40-150-CH2), 1.63 (2H, tt, J = 7.0, 7.0,
30-CH2), 3.47 (2H, t, J = 7.0, 20-CH2), 6.18 (1H, bs, NH), 7.45 (2H,
dd, J = 7.0, 7.0, m-Ph H), 7.51 (1H, t, J = 7.0, p-Ph H), 7.78 (2H, d,
13
J = 7.0, o-Ph H). C NMR: d 14.1 (160-CH3), 22.7(150-CH3), 27.0,
29.3, 29.4, 29.6, 29.7, 29.8, 31.9 (14 ꢀ CH2), 40.1 (20-CH2), 126.9,
128.5, 131.1, 134.9 (aromatic-C), 167.5 (C@O). Physical and spec-
troscopic data were in accordance with those reported.22
4.1.2. N-(3-Hydroxypropionyl)-hexadecananamide (2)
The compound 2 was prepared from 3-hydroxypropanoic acid
(48 mg, 0.53 mmol, d 1.08 g/mL, 34 lL) as white solid 69 mg
(0.22 mmol, 42%). Rf 0.38 (Hexane/Acetone 80:20). Mp 85–87 °C.
1H NMR (CDCl3): d 0.91 (3H, t, J = 6.9, 160-CH3), 1.25–1.36 (24 H,
m, 40-150 CH2), 1.55 (2H, tt, J = 7.0, 7.0, 30-CH2), 2.54 (2H, t, J = 5.4,
1-CH2), 3.31 (2H, t, J = 7.0, 20-CH2), 3.93 (2H, t, J = 5.4, 2-CH2),
13
6.30 (1H, bs, NH). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3),
4. Materials and methods
26.9, 29.3, 29.4, 29.5, 29.6, 29.7, 29.8, 31.9 (14 ꢀ CH2), 37.7 (2-
CH2), 39.6 (20-CH2), 59.0 (1-CH2), 173.3 (C@O). Anal. Calcd. for
4.1. Chemistry
C19H39NO2 (313,30): C, 72.79; H, 12.54; N, 4.47. Found: C, 72.77;
H, 12.53; N, 4.48.
Column chromatography was performed on Silica Gel 60 (70–
230 mesh) using the specified eluents. The progress of the reac-
tions was monitored by analytical thin-layer chromatography
(TLC) on pre-coated glass plates (silica gel 60 F254-plate-Merck,
Darmstadt, Germany) and the products were visualized by UV
light. Purity of all compounds (>99%) was verified by thin layer
chromatography and NMR measurements. Elemental analyses
were obtained for all intermediates and are within 0.4% of theo-
retical values. The chemicals and solvents were obtained from
Sigma–Aldrich and used without further purification. Melting
points were determined with a Stuart Scientific SMP3 melting
point apparatus and are uncorrected. Optical rotations were mea-
sured on a Perkin-Elmer 241 polarimeter (sodium D line at
25 °C). 1H NMR spectra were recorded in CDCl3 (isotopic enrich-
ment 99.95%) or pyridine-d5 (isotopic enrichment 99.98%) solu-
4.1.3. N-Hexadecyl-2-hydroxybenzamide (3)
The compound 3 was prepared from 2-hydroxybenzoic acid
(57 mg, 0.41 mmol) as white solid 109 mg (0.30 mmol, 73%). Rf
0.48 (Hexane/Acetone 80:20). Mp 104–106 °C. 1H NMR (CDCl3): d
0.90 (3H, t, J = 6.5, 160-CH3), 1.25–1.34 (24 H, m, 40-150 CH2), 1.65
(2H, tt, J = 7.0, 7.0, 30-CH2), 3.47 (2H, t, J = 7.0, 20-CH2), 6.87 (1H,
ddd, J = 1.4, 7.6, 7.7, 5-H), 7.01 (1H, dd, J = 1.4, 7.7, 3-H), 7.36 (1H,
dd, J = 1.4, 7.6, 6-H), 7.41 (1H, ddd, J = 1.4, 7.6, 7.7, 4-H), 8.26 (1H,
13
bs, NH). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 27.0, 29.3,
29.4, 29.6, 29.7, 29.8, 31.9 (14 ꢀ CH2), 39.7 (20-CH2), 114.4, 118.5,
118.7, 125.1, 134.1 (aromatic-C), 161.2 (2-C), 169.9 (C@O). Physical
and spectroscopic data were in accordance with those reported.23
4.1.4. N-Hexadecyl-(R)-2-Hydroxy-2-phenylaceticamide (4)
The compound 4 was prepared from R-mandelic acid (63 mg,
0.41 mmol) as white solid 106 mg (0.28 mmol, 68%). Rf 0.56 (Hex-
tions at 300 K using
a
Bruker AVANCE 500 instrument
(500.13 MHz for 1H) using 5 mm inverse detection broadband
probes and deuterium lock. Chemical shifts (d) are given as parts
per million relative to the residual solvent peak (7.26 ppm for
1H) and coupling constants (J) are in Hertz. The experimental error
in the measured 1H-1H coupling constants is 0.5 Hz. The splitting
pattern abbreviations are as follows: s, singlet; d, doublet; t, tri-
plet; q, quartet; m, multiplet, and bs, broad peak. For two-dimen-
sional experiments, Bruker microprograms using gradient
selection (gs) were applied.
ane/Acetone 80:20). Mp 90–92 °C. [a]
20-5.0 (c 1, CDCl3). 1H NMR
D
(CDCl3): d 0.90 (3H, t, J = 6.7, 160-CH3), 1.22–1.34 (24 H, m, 40-150
CH2), 1.48 (2H, tt, J = 7.0, 7.0, 30-CH2), 2.65 (1H, bs, OH), 3.27 (2H,
t, J = 7.0, 20-CH2), 5.04 (1H, s, 1-H), 6.07 (1H, bs, NH), 7.36–7.44
13
(5H, m, aromatic). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 26.7,
29.2, 29.3, 29.4, 29.5, 29.6, 29.7, 29.8, 29.9, 31.9 (14 ꢀ CH2), 39.7
(20-CH2), 74.1 (1-C), 126.9, 128.7, 128.9, 139.6 (aromatic-C), 172.0
(C@O). Anal. Calcd for C24H41NO2 (375,31): C, 76,75; H, 11.00; N,
3.73. Found: C, 76.74; H, 10.99; N, 3.75.
All compounds were synthesized by a one-step coupling reac-
tion. Briefly, in a general procedure the appropriate acid (1 eq),
and (1-cyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylamino-
morpholino-carbenium hexafluoro phosphate (COMU, 1 eq) were
dissolved in anhydrous CH2Cl2/CH3CN (3/1) at room temperature
under a nitrogen atmosphere. After 10 min diisopropylethylamine
(2 eq) was added and the resulting orange-red solution was stirred
for 30 min. Hexadecylamine (1 eq) in CH3CN was then injected into
the reaction mixture and vigorous stirring at room temperature
4.1.5. N-Hexadecyl-(S)-2-hydroxy-2-phenylaceticamide (5)
The compound
5
was prepared from S-mandelic (63 mg,
0.41 mmol) as white solid 114 mg (0.30 mmol, 73%). Rf 0.56 (Hex-
ane/Acetone 80:20). Mp 89–92 °C. [a]
D
20 + 5.1 (c 1, CDCl3). 1H NMR
(CDCl3): d 0.90 (3H, t, J = 6.7, 160-CH3), 1.22–1.34 (24 H, m, 40-150
CH2), 1.48 (2H, tt, J = 7.0, 7.0, 30-CH2), 2.66 (1H, bs, OH), 3.27 (2H,
t, J = 7.0, 20-CH2), 5.04 (1H, s, 2-H), 6.07 (1H, bs, NH), 7.36–7.44
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
7
13
(5H, m, aromatic). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 26.7,
4.75), 3 mM DTT and 0.1% Triton X-100), the correct amount of
hrNAAA (5-15-25-50 ng) and 20 l of a solution of PAMCA in
DMSO (to obtain final concentration of PAMCA of 1-5-10,100 M)
were added. The reaction was run for 5 h in a Perkin Elmer Victor
3 fluorimeter with fluorescence readings taken every 3 min at a
wavelength of 460 nm applying an excitation wavelength of
29.2, 29.3, 29.4, 29.5, 29.6, 29.7, 31.9 (14 ꢀ CH2), 39.7 (20-CH2),
74.1 (1-C), 126.9, 128.7, 128.9, 139.6 (aromatic-C), 172.0 (C@O).
Anal. Calcd. for C24H41NO2 (375,31): C, 76.75; H, 11.00; N, 3.73.
Found: C, 76.77; H, 10.98; N, 3.71.
l
l
4.1.6. N-Hexadecyl-4-pyridincarboxyamide (6)
360 nm. As described before 25 ng of hrNAAA and PAMCA 10 lM
The compound 6 was prepared from 4-pyridinecarboxylic acid
(51 mg, 0.41 mmol) as white solid 124 mg (0.36 mmol, 88%). Rf
0.64 (Hexane/Acetone 75:25). Mp 93–94 °C. 1H NMR (CDCl3): d
0.90 (3H, t, J = 6.7, 160-CH3), 1.24–1.38 (24 H, m, 40-150 CH2), 1.66
(2H, tt, J = 7.0, 7.0, 30-CH2), 3.49 (2H, t, J = 7.0, 20-CH2), 6.79 (1H,
bs, NH), 7.85 (2H, d, J = 5.3 Hz, 3 and 5-H), 8.80 (2H, d, J = 5.3 Hz,
were selected as the amount with better fluorescence production
in a good reaction time. Compound 1 was used to optimize screen-
ing experiments using conditions described above. 10
solution of 1 (to reach 5-10-50-100 M final concentrations per
well) were added to hrNAAA in 180 l of reaction buffer and after
30 min of incubation at room temperature 10 l of PAMCA in
DMSO (to reach 10 M final concentration) were added. The first
ll of DMSO
l
l
l
13
2 and 6-H). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 27.0, 29.3,
l
29.4, 29.5, 29.6, 29.7, 29.8, 31.9 (14 ꢀ CH2), 40.5 (20-CH2), 122.0,
144.0, 148.4, (aromatic-C), 164.6 (C@O). Physical and spectroscopic
data were in accordance with those reported.24
screening and the selection of newly synthesized compounds were
performed in the same conditions at three concentration levels 1-
10-100 lM in a run of 2.5 h. Results were plotted as time vs mea-
sured fluorescence and residual enzymatic activity was calculated
as percentage comparing the slope of kinetic curves of PAMCA
hydrolysis in the presence or absence of inhibitors.
4.1.7. N-Hexadecyl-2,5-dihydroxybenzamide (7)
The compound 7 was prepared from 2,5-dihydroxybenzoic acid
(64 mg, 0.42 mmol) as white solid 75 mg (0.20 mmol, 48%). Rf 0.70
(Hexane/Acetone 70:30). Mp 101–103 °C. 1H NMR (CDCl3): d 0.89
(3H, t, J = 6.7, 160-CH3), 1.23–1.38 (24 H, m, 40-150 CH2), 1.62 (2H,
tt, J = 7.0, 7.0, 30-CH2), 3.42 (2H, t, J = 7.0, 20-CH2), 6.85 (1H, bd,
J = 8.0, 4-H), 6.91 (1H, bs, 6-H), 6.93 (1H, bd, J = 8.0, 3-H), 8.24
4.3. Characterization of active compounds and structure activity
relationship consideration
Five points assay was performed with all compounds, excluding
the inactive compound 8, to obtain inhibition curves and IC50 val-
ues. As described before hrNAAA and compounds at 10-25-50-75
13
(1H, bs, NH). C NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 27.0,
29.3, 29.4, 29.6, 29.7, 29.8, 31.9 (14 ꢀ CH2), 39.7 (20-CH2), 111.3,
118.9, 122.1 (aromatic-C), 148.3 (2-C), 154.4 (4-C), 169.5 (C@O).
Physical and spectroscopic data were in accordance with those
reported.25
and 100 lM final concentrations were incubated for 30 min and,
after PAMCA addition, reaction was monitored for 2.5 h. The
enzyme activity was evaluate by fluorescence produced, and inhi-
bition expressed as percentage of the residual enzyme activity. IC50
values were calculated by non-linear regression analysis of Log[in-
hibitor concentration]/enzyme activity curves using GraphPad
Prism 5 applying standard slop curve. To explore the manner of
inhibition (competitive or non-competitive), reaction of six differ-
4.1.8. N-Hexadecylnaphthalene-2-carboxamide (8)
The compound 8 was prepared from 2-naphthoic acid (72 mg,
0.42) as white solid 139 mg (0.35 mmol, 83%). Rf 0.58 (Hexane/
Acetone 70:30). Mp 107–109 °C. 1H NMR (CDCl3): d 0.90 (3H, t,
J = 6.7, 160-CH3), 1.27–1.46 (24 H, m, 40-150 CH2), 1.69 (2H, tt,
J = 7.0, 7.0, 30-CH2), 3.54 (2H, t, J = 7.0, 20-CH2), 6. 29 (1H, bs, NH),
7.54–7.60 (2H, m, 6- and 7-H), 7.85 (1H, dd, J = 1.8, 8.5 Hz, 3-H);
7.89–7.95 (3H, m, 4-, 5-and 8-H), 8.30 (1H, d, J = 1.8 Hz, 1-H). 13C
NMR: d 14.1 (160-CH3), 22.7 (150-CH3), 27.1, 29.4, 29.5, 29.6, 29.7,
31.9 (14 ꢀ CH2), 40.3 (20-CH2), 123.6, 126.7, 127.2, 127.5, 127.7,
128.4, 128.9, 132.1, 132.7, 134.7 (aromatic-C), 167.5 (C@O). Anal.
Calcd. for C27H41NO (395,32): C, 81.97; H, 10.45; N, 3.54. Found:
C, 81.98; H, 10.46; N, 3.52.
ent concentrations of PAMCA (5-10-25-50-100-200
centrations per well) with hrNAAA in the presence or absence of
more active inhibitors 3, 4, 7 and the linear 2 (100 M) were per-
lM final con-
l
formed as described before. Analyzing data with GraphPad Prism 5,
Michaelis-Menten curves gave Vmax and KM values and Linewea-
ver-Burk plot was built reporting 1/V vs 1/[PAMCA]. Reversibility
experiments were performed, as described in literature,21 incubat-
ing hrNAAA with selected compound at two different concentra-
tions (50 and 100
After incubation solutions were diluted twenty fold and new
amount of inhibitors (2.5–50 M final concentration to 50 M ini-
tial concentration and 5–100 final concentration to 100 M initial
lM final concentration ꢀ well) for 20 min.
4.2. Selection of inhibitors by an high-throughput screening
fluorimetric method
l
l
l
concentration) were added, then reaction was started by PAMCA
addition. The fluorescence produced was evaluated and compar-
ison of the enzyme activity displayed by the two couple of exper-
iments allowed to define reversibility of the inhibition.
The fluorogenic probe was synthesized starting from palmitic
acid and 7-methyl-4-amino coumarine, by a coupling reaction
and in presence of COMU and DIPEA in dry condition as described
for compounds 1–7. Crude was purified by flash chromatography
(Hexane/Acetone 80:20) to give product as white solid (95 mg,
0.23 mmol) and reaction yield of 56%. NMR analyses were in agree-
ment with that described in literature16 confirming structure and
purity grade of the product. The florescence assay procedure used
to optimize the method to evaluate NAAA activity was very similar
to that used by West and coworkers.16 We decided to performed
in vitro experiments only in presence of hrNAAA, instead of a total
protein extract from cells over-expressing NAAA, because PAMCA
is not a substrate specific for this enzyme and the presence of FAAH
or other hydrolases could influence florescence production. To
optimize the experiment condition different amounts of hrNAAA
(R&D System, Cod 4494-AH, purity >95%) and concentration of
PAMCA were tested. In a 96 well black plate to each well, contain-
4.4. Cell culture and cell viability assay
RT4 (grade I), RT112 (II), 5637 (II) and HT-1376 (III) human
bladder cancer cells were maintained in RPMI 1640 medium sup-
plemented with 10% (v/v) fetal bovine serum, 50 U/ml penicillin
and 50
lg/ml streptomycin at 37 °C in a 5% CO2 humidified
atmosphere.
To test the effect of NAAA inhibitors on cell viability, cells were
plated in 96-well plates at a cell density of 5 ꢀ 103 cells/well and
treated with serial logarithmic dilutions of NAAA inhibitors rank-
ing from 100 to 0.1 lM for 72 h. Control cultures received the
respective volume of DMSO. At the end of the exposure period,
MTT assay was performed according to the manufacturer’s recom-
mendations (Sigma-Aldrich). Purple formazan product was re-sus-
ing 180 ll of reaction buffer (sodium acetate buffer 100 mM (pH
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042

8
R. Vago et al. / Bioorganic & Medicinal Chemistry xxx (2017) xxx–xxx
pended in DMSO and measured in a spectrophotometric micro-
plate reader at 570 nm wavelength. Cell viability was evaluated
as the concentration inhibiting the 50% incorporation in untreated
References
1. Fonseca BM, Costa MA, Almada M, Correia-da-Silva G, Teixeira NA. Prostag Oth
Lipid
prostaglandins.2013.02.002.
M.
2013;102–103:13–30.
http://dx.doi.org/10.1016/j.
control cells and is expressed as IC50
.
2. Velasco G, Hernández-Tiedra S, Dávila D, Lorente M. Prog Neuro Psychoph.
2016;64:259–266. http://dx.doi.org/10.1016/j.pnpbp.2015.05.010.
3. Pisanti S, Picardi P, D’Alessandro A, Laezza C, Bifulco M. Trends Pharmacol Sci.
2013;34:273–282. http://dx.doi.org/10.1007/s12020-013-0131-8.
4. Feledziak M, Lambert DM, Marchand-Brynaert J, Muccioli GG. Recent Pat CNS
Drug Discov. 2012;7:49–70. http://dx.doi.org/10.2174/157488912798842223.
5. Thors L, Bergh A, Persson E, et al. PLoSOne. 2010;5:e12275. http://dx.doi.org/
10.1371/journal.pone.0012275.
6. Ueda N, Tsuboi K, Uyama T. Prog Lipid Res. 2010;49:299–315. http://dx.doi.org/
10.1016/j.plipres.2010.02.003.
7. Li Y, Yang L, Chen L, et al. PLoS One. 2012;7:e43023. http://dx.doi.org/10.1371/
journal.pone.0043023.
4.5. Wound healing assay
Human bladder cancer cells were plated in 6-well culture
dishes at a cell density of 5 ꢀ 105 cells/well to create a sub-conflu-
ent monolayer. A ꢁ1 mm wide scratch was made across the cell
layer using a sterile pipette tip. Debris were removed and the
growth medium was replaced with a medium containing 40 nM
NAAA inhibitors (DMSO only for controls). Phase-contrast images
were acquired at time 0 and after 24 h incubation. Wound closure
areas were measured with ImageJ software by subtracting the total
amount of grayscale pixel counted in the cell-free area remaining
after 24 h from the initial wound area. Data are reported as
means standard deviations. Cell motility was assessed as per-
centage of cell covered area with respect to controls.
8. Taylor
BK.
Pain.
2013;154:326–327.
http://dx.doi.org/10.1016/
j.pain.2012.12.012.
9. Chang SS, Dinney CP, Donat SM, et al. Urol Oncol. 2012;30:944–947. http://dx.
doi.org/10.1016/j.urolonc.2012.07.006.
10. Pisanti S, Malfitano AM, Grimaldi C, et al. Best Pract Res Clin Endocrinol MeTable.
2009;23:117–131. http://dx.doi.org/10.1016/j.beem.2009.02.001.
11. Gasperi V, Evangelista D, Oddi S, et al. Life Sci. 2015;138:41–51. http://dx.doi.
org/10.1016/j.lfs.2014.09.031.
12. Vandevoorde S, Tsuobi K, Ueda N, Jonsson K-O, Fowler CJ, Lambert DM. J Med
Chem. 2003;46:4373–4376. http://dx.doi.org/10.1021/jm0340795.
13. Bandiera T, Ponzano S, Piomelli D. Pharmacol Res. 2014;86:11–17. http://dx.doi.
org/10.1016/j.phrs.2014.04.011.
4.6. Effects of inhibitors on endocannabinoid levels in bladder cancer
cells
14. (a) Saturnino C, Petrosino S, Ligresti A, et al. Bioorg Med Chem Lett.
2010;20:1210–1213. http://dx.doi.org/10.1016/j.bmcl.2009.11.134;
(b) Yamano Y, Tsuboi K, Hozaki Y, et al. Bioorg Med Chem. 2012;20:3658–3665.
http://dx.doi.org/10.1016/j.bmc.2012.03.065.
15. Ottria R, Casati S, Ciuffreda P. Chem Phys Lip. 2012;165:705–711. http://dx.doi.
org/10.1016/j.chemphyslip.2012.06.010.
16. West JM, Zvonok N, Whitten KM, Wood JAT, Makriyannis A. J Proteome Res.
2012;11:972–981. http://dx.doi.org/10.1021/pr200735a.
17. Hermanson DJ, Marnett LJ. Cancer Metastasis Rev. 2011;30:599–612. http://dx.
doi.org/10.1007/s10555-011-9318-8.
18. Nomura DK, Lombardi DP, Chang JW, et al. Chem Biol. 2011;18:846–856. http://
dx.doi.org/10.1016/j.chembiol.2011.05.009.
19. Endsley MP, Thill R, Choudhry I, et al. Int J Cancer. 2008;123:1318–1326. http://
dx.doi.org/10.1002/ijc.23674.
20. Pinto-Leite R, Carreira I, Melo J, et al. Tumour Biol. 2014;35:4599–4617. http://
dx.doi.org/10.1007/s13277-013-1604-3.
21. Mimeault M, Pommery N, Wattez N, Bailly C, Hénichart JP. Prostate.
2003;56:1–12.
After migration experiments cells were used to quantify: PEA,
OEA, SEA, AEA, LEA, and LNEA. Quantification were performed on
an ABSciex 5600 TripleTOF mass spectrometer (AB Sciex, Foster-
City, CA, USA) coupled with a LC system composed by an Agilent
1290 Infinity autosampler and an Agilent 1200 Infinity pump (Agi-
lent Technologies, Waldbronn, Germany) as described before.26
Briefly, 1 ml of cells and supernatants taken together was spiked
with a solution of d4-SEA and d4-LNEA, used as internal standard
for quantification, and 800 ll of cold acetone were added for pro-
tein precipitation. After centrifugation, supernatants were
extracted and purified by SPE extraction on OASIS HBL cartridges.
Eluates were collected and maintained at ꢂ80 °C till the analyses.
Quantifications were performed using Multiquant 1.2.1 software
by AB Sciex. Data are reported as means standard deviations.
22. Zaffalon P-L, Stalder E, Fedotenko IA, Favarger F, Zumbuehl A. Tetrahedron Lett.
2011;52:4215–4217. http://dx.doi.org/10.1016/j.tetlet. 2011.06.031.
23. Kaushik P, Shakil NA, Kumar J, Singh BB. Synthetic Comm. 2015;45:579–587.
http://dx.doi.org/10.1080/00397911.2014.974611.
Acknowledgments
24. Ichikawa T, Kato T, Ohno H. J Am Chem Soc. 2012;134:11354–11357. http://dx.
doi.org/10.1021/ja304124w.
25. Tsuboi K, Hilligsmann C, Vandevoorde S, Lambert DM, Ueda N. Biochem J.
2004;379:99–106. http://dx.doi.org/10.1042/bj20031695.
We thank the University of Milan and the Urological Research
Institute (URI) for financial support.
26. Ottria R, Ravelli A, Gigli F, Ciuffreda P. J Chrom B. 2014;958:83–89. http://dx.doi.
org/10.1016/j.jchromb.2014.03.019.
A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2016.12.042.
Please cite this article in press as: Vago R., et al. Bioorg. Med. Chem. (2017), http://dx.doi.org/10.1016/j.bmc.2016.12.042