Towards smart biocide-free anti-biofilm strategies: Click-based synthesis of cinnamide analogues as anti-biofilm compounds against marine bacteria

Article abstract of DOI:10.1016/j.bmcl.2017.11.039

A set of triazole-based analogues of N-coumaroyltyramine was designed to discover potential leads that may help in the control of bacterial biofilms. the most potent compounds act as inhibitors of biofilm development with EC50 closed to ampicillin (EC50 =

Full text of DOI:10.1016/j.bmcl.2017.11.039

Accepted Manuscript
Towards smart biocide-free anti-biofilm strategies:Click-based synthesis of cin-
namide analogues as anti-biofilm compounds against marine bacteria.
C. Sall, M. Ayé, O. Bottzeck, A. Praud, Y. Blache
PII:
S0960-894X(17)31138-1
https://doi.org/10.1016/j.bmcl.2017.11.039
BMCL 25445
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
22 September 2017
20 November 2017
23 November 2017
Please cite this article as: Sall, C., Ayé, M., Bottzeck, O., Praud, A., Blache, Y., Towards smart biocide-free anti-
biofilm strategies:Click-based synthesis of cinnamide analogues as anti-biofilm compounds against marine bacteria.,
Bioorganic & Medicinal Chemistry Letters (2017), doi: https://doi.org/10.1016/j.bmcl.2017.11.039
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Towards smart biocide-free anti-biofilm strategies : Click-based synthesis of
cinnamide analogues as anti-biofilm compounds against marine bacteria.
a
b
b
b
b*
C. Sall, M. Ayé, O. Bottzeck, A. Praud, Y. Blache
a
Laboratoire de chimie, UFR des Sciences de la Santé, Université de Thiès BP 967, Thiès, Sénégal.
2
Laboratoire MAPIEM, Seatech Ecole d’ingénieurs, université de Toulon, CS 60584, Toulon Cedex 9
*
blache@univ-tln.fr

Biofilms results in the colonization process of soft or hard, artificial or living substrata by
1
microorganisms which attach to surfaces and bind to one another. Although the control of the
development of planktonic bacteria communities is well known and mastered, bacteria within a biofilm
are much more resistant to antibiotic or biocide treatment (up to 1000-fold increased resistance). To fight
biofilms, the massive use of such compounds has led to major problems in human health by developing
high degrees of resistance in bacterial communities as well as economic, environmental and
2
-5
toxicological issues leading to establishment of strict regulations. In this context, targeting the
formation of bacterial biofilms in a non-toxic way is of great interest in view of a rational use of
antibiotics and/or biocides and the development of original non-toxic biofilm inhibitors should have the
potential to be used in a preventive treatment of a wide diversity of industrial and medical surfaces.
Moreover, the development of such solutions must remain competitive and must enable low-cost
molecules to be placed on the market. For such a challenge, some of the anti-biofilm strategies that are
pursued todays consist in studying structure-activity relationships (SAR) of simple secondary
metabolites from marine organisms such as sponges or soft corals in view of discovering new specific
and non-toxic anti-biofilm leads which should be used as potential adjuvant for antibiotherapies or
6
-8
friendly environmentally biocides. In this field, we are developing an efficient and simple approach
based on the bioisosteric replacement of natural frameworks by a 1,2,3-triazolic ring to allow SAR
studies in the field of antifouling fight, and we have previously designed with success isonaamine A,
9
-11
bromotyramines and hemibastadins analogues (fig. 1). For this purpose, ‘click chemistry” by mean of
copper-catalyzed azide-alkyne cycloaddition was used as a highly efficient approach, offering
1
2-13
substantial advantages, since it is tolerant to multiple functional groups.
In addition a large variety of
terminal alkyne is available as building blocks, while azides may easily be accessed from common
precursors, such as halides and amines, using standard literature procedures.

N
N
N
N
N
HO
H N
2
R
OH
R
Isonaamine A
R = OCH , bioactive analogue
3
R = OH inactive analogue
N
N
Br
NH2
N
O
N
O
O
Br
Br
Bromotyramine
bioactive analogue
R1
R
X
OH
HO
Br
N
N
H
R2
N
N
N
O
H
N
OH
O
Br
OH
R = OCH , bioactive analogue
R = OH less active analogue
X= Br or H
Hemibastadins
3
9
-11
Figure 1 : click-based design of natural product analogues
Pursuing these investigations in the field of antibiofilm compounds, we are interested in the advanced
SAR studies of tyramine derivatives and we wish to investigate some more simple frameworks which
could be accessible in a two-steps process from simple commercial tyramines and anilines. To assume
such a challenge, we focused on the design of original analogues of cinnamoylphenethylamine family;
this class of natural products are found in over 30 plant families. The most ubiquitous are p-
coumaroyltyramine and feruloyltyramine. Nevertheless, the cinnamic acid and phenethylamine
derivatives are of great interest due to their plethora of associated biological activities such antibacterial
1
4-16
17
and antimicrobial
compounds and more especially as anti-biofilm compounds. In this work we
planified a bioisosteric replacement of the double bond by a 1,2,3-triazole ring (fig. 2). Such bioisosteric
replacements of double bonds have already been investigated with success to design analogues of
1
8-19
combrestatin and resveratrol analogues which exhibited enhanced antitumoral potential.
In this way
the cinnamic acid moiety is modified. Furthermore, considering our previous results which showed that
1
1
methoxy derivatives were more efficient than the hydroxy ones, we decided to restrict the study to
methoxy anilines 1a and 1b as starting materials.

modified
cinnamic acid moiety
cinnamic acid moiety tyramine moiety
tyramine moiety
HO
H CO
3
H
N
N
N
N
H
O
N
OH
Cinnamides
O
R
N-coumaroyltyramine
targeted compounds
R = OCH3
R = OH
Figure 2. Structure of the targeted compounds
The first step of this work was the preparation of intermediary carboxylic acids (3a, 3b) (scheme 1).
These compounds were obtained in excellent yield in two steps. Treatment of anilines (1a, 1b) with
sodium nitrite followed by sodium azide in dimethylformamide afforded the corresponding crude azides
(
2a, 2b) which were used without further purification. Synthesis of the targeted carboxylic acids (3a, 3b)
was then achieved by performing the copper(I)-catalyzed 1,3-dipolar cycloaddition of the organic azides
2
0
with propargylic acid resulting in the formation of 1,2,3-triazoles. In general, these reactions usually
proceed to completion in 6–36 h at room temperature in water with a variety of organic co-solvents, such
1
2,21
as tert-butanol, ethanol, DMF, DMSO, THF, or CH
3
CN.
Ethanol was chosen rather than DMF to
9
allow an easier workup and a better purity of products as described in our previous work. In practice,
propargylic acid was added to a solution of appropriate azide (2a, 2b), CuSO /sodium ascorbate in a
4
water/ethanol mixture (50/50) and the reaction time was optimized at 12 hours at room temperature.
Further access to the different cinnamides analogues 4-11, was allowed by a peptide coupling step using
2
2,23
DCC/HOBt methodology.
table 1).
All amides were obtained in good yields from starting azido compounds
(

R1
N
NH2
R1
N3
N
N
R1
OH
(
i)
(ii)
CO H
R2
R2
O
2
R2
1
4
3a (71%)
3b (68%)
1
1
a : R = OCH , R = H
2a
2b
3
2
1
b : R = H, R = OCH
3
NH2
(
(
(
i) : a) NaNO , HCl/H O, 0°C, 2h, b) NaN , 70°C.
2 2 3
(
iii)
ii) : CuSO pentahydrate, Na ascorbate, EtOH/H O, 12h, RT
4
2
iii) : DCC/HOBt/DMF, 5 min, then DIPEA/appropriate tyramine, 48h
R4
R3
R2
N
R1
N
N
H
N
R4
O
R3
4-11
Scheme 1 Synthesis of targeted cinnamide analogues from methoxyanilines 1a, 1b
:
Table 1 : selected 1,4-disubstituted 1,2,3-triazoles
1
2
3
4
R
R
H
H
H
H
R
R
Cpd (Yield)
4 (50%)
5 (43%)
6 (78%)
7 (79%)
8 (64%)
9 (50%)
10 (66%)
11 (39%)
OCH
OCH
OCH
OCH
H
3
3
3
3
OH
H
H
OCH
3
3
3
3
OCH
OCH
OH
3
H
3
OCH
H
OCH
OCH
OCH
OCH
3
3
3
3
H
H
OCH
H
H
OCH
3
3
H
OCH
OCH
In order to assess anti-biofilm activity of these compounds against representative Gram-negative
bacterial biofilms, three strains were chosen for their capacity to form biofilms: Pseudoalteromonas
2
4
lipolytica (TC8), Pseudoalteromonas ulvae (TC14) and a Paracoccus sp. strain (4M6), by using our
®
25,26,27
previous method adapted from Leroy et al. using the specific fluorophore Syto 61.
screen are outlined in Table 2 as effective concentrations to inhibit 50% of the bacterial adhesion
expressed as EC50). First, we can note that response of the three strains are different and that TC14 is
Results of this
(
more sensitive to this class of compounds than 4M6 and TC8 strains. The second point concerns the
natural N-coumaroyltyramine which does not exhibit significant efficiency. Then, the results highlighted

some preliminary SAR information. Compounds possessing a 4-hydroxy group on the tyramine moiety
4 and 8) are not active (EC50> 100 M) confirming our hypothesis and previous results which showed
(
that hydroxy groups are not beneficial for the activity, while all methoxy derivatives are active. We can
also remark that the position of the methoxy group on the tyramine moiety affects the activity and that
the 4-methoxy derivatives 6 and 10 gave rise to the best anti-biofilm, in the same order of magnitude of
ampicillin against TC14 (11.4 M and 11.3 M respectively vs 9.3 for ampicillin). In the same way,
addition of a supplementary methoxy group (compounds 7 and 11) results in reduction of effectiveness,
especially in the case of compound 7 (73.7 M). Otherwise, no precisions concerning the SAR on the
modified cinnamic moiety could be highlighted at this stage since the position of the methoxy group on
the 3-position or the 4-position of the aromatic ring does not affect the activity (4 vs 8, 5 vs 9, 6 vs 10),
while we can observe a modulation of activity in the case of 7 and 11 (3,4-dimethoxy on the tyramine
moiety).
Table 2 : biological screening against bacteria biofilms
*
*
*
compounds
4M6 (EC , M) TC8 (EC , M) TC14 (EC , M)
5
0
50
50
4
5
6
7
8
9
>300
275.5 ± 25
204.3 ± 26.0
226.4 ± 16
215.5 ± 25
153.5 ± 9
219.5 ± 12
213.6 ± 23
259.6 ± 66
162.5 ± 5
>300
80.6 ± 22.0
11.4 ± 1.0
73.7 ± 17.0
115.3 ± 25.0
52.3 ± 22.0
11.3 ± 7.0
54.6 ± 11.0
75.5 ± 20
>300
>300
>300
142.3±0
>300
1
1
0
1
>300
N-coumaroyltyramine >300
Ampicillin
144.1 ± 0.6
>300
17.9 ± 1.0
9.3 ± 1.0
*
4M6 : Paracoccus sp. TC8 : Pseudoalteromonas lipolytica, TC14 : Pseudoalteromonas ulvae
The further step was finally to investigate if these compounds exhibited a specific anti-biofilm activity
or if this observation was simply related to a general toxic effect on the bacteria. For this purpose, both
2
8
growth inhibition and viability assay were performed. We have already shown that the antibiofilm
activity of ampicillin is directly connected to an antibacterial and general toxic effect on these different
10,11
strains (TC8, TC14, 4M6),
but in contrast, the results presented in figure 3 and 4 showed that when
compared to untreated samples, the compounds 6, 9, 10, 11 exhibit no effect on the bacterial growth of
TC14 strain (fig. 3).

Figure 3. Effect on TC14 growth of compounds
, 9-11 at concentrations of 100 M
Figure 4. Effect on TC14 viability of compounds
6, 9-11 at concentrations of 100 M.
6
0
.5
.45
.4
.35
.3
.25
.2
.15
.1
Control
140
0
0
0
0
Ampicillin
1
20
00
0
6
9
1
80
1
0
1
0
1
60
4
0
0
0
0
2
0
Compounds
0
H
3 H
5 H
7 H
9 H
1 1 H
TIME (HOURS)
®
For viability, the same methodology used for antiadhesion assay with Syto 61 was applied using
resazurin test (fig. 4). Results showed that, when compared to ampicillin, none of the tested compounds
(
6, 10-11) exhibited any effect on the viability of the bacteria clarifying that these compounds were not
lethal at 100 M to the bacteria and suggested that the anti-biofilm activities of such compounds were
not connected to antibacterial effect.
In conclusion, we have designed original analogues of cinnamides (N-coumaroyltyramine) based on the
bioisosteric replacement of the cinnamic double dond. Interest of the work resides in the simple two-
steps synthetic approach (copper(I)-catalyzed 1,3-dipolar cycloaddition followed by a peptide coupling
step). This class of compounds exhibits specific anti-biofilm activity against a gram negative strain
(
Pseudoalteromonas ulvae,TC14). Futhermore, the low toxicity of the more potent derivatives when
compared to ampicillin, should allow their rational use as co-antibiotic and/or co-biocides in view of
eradication of biofilms in various applications such as biomedical materials, antifouling coatings, water
pipelines network. Further studies are now needed to define more precisely the structure-activity
relationships and their effect as potential adjuvents.
Acknowledgments
The Paracoccus sp. strain 4M6 was provided by the LBCM (Université de Bretagne Sud).
References and notes
1
2
3
-
-
-
Hall-Stoodley, L.; Costerton, J. W.; Stoodley, P. Nature Reviews Microbiology 2004, 2, 95.
Saint, S.; Chenoweth, CE. Infect. Dis. Clin. North Am. 2003, 17, 411.
Omae, I. Chem. Rev. 2003, 103, 3431.

4
5
-
-
Haras D. Mater. Techn. 2005. 93. 27.
Qian, P. Y.; Lau, S. C. K.; Dahms, H. U.; Dobretsov, S.; Harder, T. Marine biotech. 2007, 9, 399.
Melander R.J. ; Liu H-B. ; Stephens D.M. ; Bewley C.A. ; Melander C. Bioorg. Med.Chem. Lett.,
6
-
2
017, 26, 5863-5866.
7
-
Richards, J.J; Ballard, T.E.; Melander, C. Org. Biomol. Chem., 2008, 1356.
9
-
Linares, D.; Bottzeck, O.; Pereira, O.; Praud-Tabariès, A.; Blache, Y. Bioorg. Med. Chem. Lett.
2
011, 21, 6751.
1
1
1
1
0- Andjouh, S.; Blache, Y. Biofouling, 2016, 32, 871-881.
1- Andjouh, S.; Blache, Y. Bioorg. Med. Chem. Lett. 2015, 25, 5762-5766
2- Huisgen, R. 1,3-Dipolar Cycloaddition Chemistry; Wiley: New York, 1984. pp 1–176.
3- Shearman, J.W.; M. Myers M.R.; Beale T.M.; Brenton J.D.; b, Steven V. Ley S.V. Tetrahedron
Lett. 2010, 51, 4812.
1
1
4- Georgiev, L. ; Chochkova , M. ; Totseva, I. ; Seizova, K. ; Marinova, E. ; Ivanova, G. ; Ninova,
M. ; Najdenski, H. ; Milkova, T. Med. Chem. Res . 2013, 22, 4173–4182
5- Nishioka, T. ; Watanabe, J. ;Kawabata, J. ; NIKI, R. ; Biosci. Biotech. Biochem. 1997, 61, 1138-
1
141.
1
1
6- Todd, J. ; Zimmerman, R.C. ; Crews, P. ; Alberte R. S. Phytochemistry, 1993,34, 401-404
7- De Vita, D. ; Simonetti, G. ; Pandolfi, F. ; Costi, R. ; Di Santo R. ; Diodata D’Auria, F. ; Scipione,
L. Bioorg. Med. Chem. Lett. 26 (2016) 5931–5935.
1
1
2
8- Odlo, K. ; Hentzen, J. ; Chabert, J. r. m F. d. ; Ducki, S. ; Gani, O. A. B. S. M. ; Sylte, I.; Skrede,
M. ; Flørenes, V. A. ; Hansen, T. V. Bioorg. Med. Chem. 2008, 16, 4829 –4838.
9- Pagliai, F. ; Pirali, T. ; Del Grosso, E. ; Di Brisco, R. ; Tron, G.C. ; Sorba, G. ; Genazzani, A.A. J.
Med. Chem. 2006, 49, 467-470.
0- Typical method for preparation of compounds 3a, 3b: to a stirred solution of 1g of
methoxyaniline (1a, 1b) in H
NaNO . The whole was stirred at 0°C for 2h, and then NaN
resulting solution was refluxed for 3h. After extraction with dichloromethane, 100 mg of the
resulting crude azide was dissolved in a solution of H O/EtOH (1.5 mL/1.5 mL) containing CuSO
pentahydrate (0.3 eq.), propiolic acid (1.5 eq.) and sodium ascorbate (0.4 eq.). The resulting
mixture was stirred 12 hours at RT. A saturated solution of Na CO was added and the resulting
solution extracted 3 times with ethyl acetate. The organic layers was then dried over Na SO and
2
O/HCl (50/50, 15mL / 15 mL) was added 0.8g (11.55 mmol.) of
2
3
0.78g (11.55 mmol.) were added. The
2
4
2
3
2
4
evaporated to give the crude triazoles which were purified by flash chromatography on silica gel
eluted with hexane/ethyl acetate (80/20).

2
2
1- Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41,
596.
2- Typical method for preparation of compounds 4-11 : To a stirred solution of the acid 3a or 3b
250 mg, 1.28 mmol) in anhydrous DMF, were added dicyclohexylcarbodiimide (1,2eq) and
2
2
3
(
Hydroxybenzotriazole (1.2eq). The mixture was strirred at room temperature under nitrogen for 5
min, then appropriate tyramine (1eq) and diisopropylethylamine (100 uL) were added. The
reaction was stirred at room temperature for 48 hours. A saturated NaCl solution was added and
the organic layer extracted three times with CH
2
Cl
2
. The solvent was removed in vacuo and the
Cl /MeOH as eluent to yield
residue purified by flash chromatography (SiO
2
) using CH
2
2
compounds 4-11.
2
2
2
2
2
3- Shearman, J.W.; M. Myers M.R.; Beale T.M.; Brenton J.D.; b, Steven V. Ley S.V. Tetrahedron
Lett. 2010, 51, 4812.
4- Brian-Jaisson, F.; Ortalo-Magne, A.; Guentas-Dombrowsky, L.; Armaougon, F.; Blache, Y.;
Molmeret, M. Microbial Ecology 2014, 68, 94
5- Camps, M.; Briand, J. F.; Guentas-Dombrowsky, L.; Culioli, G.; Bazire, A.; Blache, Y. Mar. Poll.
Bull. 2011, 62, 1032.
6- Leroy, C. ; Delbarre-Ladrat, C. ; Ghillebaert, F. ; Rochet, M. J. ; Compère, C. ; Combes, D. Lett.
Appl. Microbiol. 2007, 44, 372.
2
6
7- Bacterial adhesion assays (adapted from Leroy et al. ). Pseudoalteromonas lipolytica (TC8) was
grown on VNSS (Vaatanen Nine Salt solution) at 20°C and sampled at the stationary phase. After
centrifugation, cells were suspended in sterile artificial sea water (ASW) until an optical density of
0
9
.2-0.4 at 600 nm was achieved. 200 µL of ASW was inoculated on the border-row wells of the
6-well microtiter plates (sterile black polystyrene NUNC), and 100 µL of the bacterial suspension
on other wells using an eight-channel pipette. 100 µL of diluted standard biocide (Seanine) and
purified molecules were added in the latter wells. All the concentrations were tested in triplicates.
1
2
00 µL of ASW was added in six wells to constitute the bacterial adhesion control. After 15 h at
0°C, the non-adhered bacteria were eliminated by three successive washes (36 g/L of sterile NaCl
solution). The adhered bacteria were stained 10 min at room temperature by adding 200 µL of 40
µmol/L Syto® 61 (Invitrogen, USA). The excess stain was removed by two washes (36 g/L NaCl
solution). The Syto® 61 was then solubilized in 200 µL of 36 g/L NaCl solution and fluorescence
was measured (kexc = 615 nm, kem = 670 nm) using an Infinit 200 micro-plate fluorescence
reader (TECAN, Lyon, France). The dose-response curves fitting and the determination of the
EC50 for each molecule were achieved using GraphPad Prism software.
2
8- Antibacterial assays. Bacterial strains were grown on VNSS at 20°C under shaking conditions
(
120 rpm) and collected during the exponential phase. After centrifugation, cells were suspended

in sterile VNSS (OD600 nm = 0.1). 180 L at different concentrations for each tested compounds
standard biocides, natural or natural-derived products) were added in four wells of the microtiter
(
plates (sterile transparent PS; Nunc, Fisher Scientific). All the concentrations were tested in
triplicate and the fourth well was filled for control. The maximum percentage of solvent used for
the dilution of biocides was also tested in triplicate as additional control. For the growth inhibition
control, 180 L of VNSS was added in six wells. Then 20 L of the bacterial suspension was
inoculated on all the wells except the border-row wells and all the wells were filled out to 200 L
with VNSS. Turbidity (OD600 nm) was measured every hour during 6 h. Then, resazurin (50 μM)
was added in all the wells, and fluorescence was measured after 2h to quantify the percent of
bacterial viability. The same methodology used with SYTO 61 was applied to calculate a percent
of viability after resazurin staining.
2
9- Spectra data of the representative compound 10 : N-(4-Methoxyphenethyl)-1-(-4-
1
methoxyphenyl)-1H-1,2,3-triazole-4-carboxamide H NMR (400 MHz, CDCl
3
) δ : 2.89 (t, 2H, J =
7
7
3
1
.1 Hz), 3.71 (m, 2H), 3.79 (s, 3H), 3.87 (s, 3H), 6.86 (d, 2H, J = 8.6 Hz), 7.04 (d, 2H, J = 9.0Hz),
1
3
.17 (d, 2H, J = 8.6 Hz), 7.63 (d, 2H, J = 9.0 Hz), 8.39 (s, 1H), C NMR (100 MHz, CD OD) δ :
3
4.9, 40.6, 55.2, 55.6, 114.1 (2C), 114.9 (2C), 122.3 (2C), 123.4, 129.7 (2C), 130.6, 143.9, 157.5,
58.3, 159.9, 160.2.

Graphical abstract