Zosteric acid and salicylic acid bound to a 
low density polyethylene surface 
successfully control bacterial biofilm 
formation
Author: C. Catto, Garth James, F. Villa, S. Villa, and 
F. Cappitelli
This is an Accepted Manuscript of an article published in Pest Management Science on [date of 
publication], available online: http://www.tandfonline.com/10.1002/ps.5043.
Catto, C., Garth James, F. Villa, S. Villa, and F. Cappitelli. "Zosteric acid and salicylic acid bound 
to a low density polyethylene surface successfully control bacterial biofilm formation." Biofouling 
34, no. 4 (April 2018): 440-452. DOI:10.1080/08927014.2018.1462342.
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
Zosteric acid and salicylic acid bound to a low density polyethylene surface 
successfully control bacterial biofilm formation
C. Cattòa,b  , G. Jamesb, F. Villaa  , S. Villac   and F. Cappitellia
aDepartment of food Environmental and nutritional Sciences, università degli Studi di Milano, Milan, italy; bCenter for Biofilm Engineering, 
Montana State university, Bozeman, MT, uSA; cDepartment of Pharmaceutical Sciences, università degli Studi di Milano, Milan, italy
ABSTRACT
The active moieties of the anti-biofilm natural compounds zosteric (ZA) and salicylic (SA) acids have 
been covalently immobilized on a low density polyethylene (LDPE) surface. The grafting procedure 
provided new non-toxic eco-friendly materials (LDPE-CA and LDPE-SA) with anti-biofilm properties 
superior to the conventional biocide-based approaches and with features suitable for applications 
in challenging fields where the use of antimicrobial agents is limited. Microbiological investigation 
proved that LDPE-CA and LDPE-SA: (1) reduced Escherichia coli biofilm biomass by up to 61% with 
a mechanism that did not affect bacterial viability; (2) significantly affected biofilm morphology, 
decreasing biofilm thickness, roughness, substratum coverage, cell and matrix polysaccharide 
bio-volumes by >80% and increasing the surface to bio-volume ratio; (3) made the biofilm more 
susceptible to ampicillin and ethanol. Since no molecules were leached from the surface, they 
remained constantly effective and below the lethal level; therefore, the risk of inducing resistance 
was minimized.
Introduction
The main strategy for preventing biofilm-mediated dam-
age in medical and industrial settings relies on routine 
cleaning and disinfection of microbial contact surfaces 
(Simões et al. 2010). Unfortunately, these approaches are 
not universally effective because sessile microorganisms 
display increased tolerance to conventional antimicrobial 
agents (Hall-Stoodley et al. 2004). In addition, resistance 
towards many antibiotics has emerged in many patho-
genic microbial taxa (Sousa et al. 2014; Ventola 2015).
The modification of surface properties by incorpo-
rating disinfectants, antiseptics, antibiotics and metallic 
nanoparticles into polymeric materials has been proposed 
as a promising strategy to tackle the current challenge in 
controlling biofilm growth (Coenye et al. 2011; Chen et al. 
2013; Lo et al. 2014; Ahire et al. 2016). However, despite 
some of these materials being commercially available and 
already used in several applications, their real efficacy and 
recurrent drawbacks have made their use questionable 
(Chen et al. 2013; Pechook et al. 2015). Most of the bioc-
ide-releasing materials have a short-term efficacy, typically 
no longer than 24 h, which make them less suitable for 
longer applications (von Eiff et al. 2005). Moreover, most 
of these materials exhibit a discontinuous release rate 
with an initially high release followed by an exponential 
decrease, that favors the development of antimicrobial 
resistance (Gharbi et al. 2012). Finally, coating materials 
often undergo chemical damage with a loss of efficacy 
(Ghosh et al. 2012; Pechook et al. 2015; Sobieh et al. 2016).
Recently, the natural compounds zosteric acid (ZA) 
(or p-(sulfoxy)cinnamic acid) and salicylic acid (SA) have 
been proposed as an alternative or integrative antimicro-
bial-free strategy to prevent biofilm development (Villa 
et al. 2010; Cattò et al. 2017). Instead of killing cells, ZA 
and SA influence the multicellular behavior of microor-
ganisms, ie microbial adhesion, cell-to-cell communica-
tion signals and dispersion, massively decreasing biofilm 
development without imposing a selective pressure, thus 
limiting drug-resistance development (Polo et al. 2014). In 
previous studies by these authors, the structural require-
ments of ZA and SA necessary for biofilm inhibition as 
well as functional groups that could be exploited for their 
covalent linkage to an abiotic surface have been identified 
(Cattò et al. 2015, 2017). These works revealed that the cin-
namic acid moiety is responsible for the ZA anti-biofilm 
Biofilm growth in the CDC reactor
E. coli biofilm was grown on non-functionalized (LDPE,
LDPE-OH and LDPE-COOH) and functionalized
(LDPE-CA and LDPE-SA) coupons in the Center for
Disease Control biofilm reactor (CDC reactor, BioSurface 
Technologies, Bozeman, MT, USA) according to Cattò
et al. (2017). Briefly, the bioreactor was inoculated with
400 ml of sterile LB medium with the addition of 1 ml
of diluted pre-washed overnight culture containing 107
cells of E. coli. The culture was grown at 37°C with con-
tinuous stirring for 24 h. When the 24-h adhesion phase
was over, the peristaltic pump was started and sterile 10% 
LB medium was continuously pumped into the reactor
at a rate of 8.3 ml min−1. After a dynamic phase of 48 h,
functionalized and non-functionalized coupons were
removed, gently washed with phosphate buffered saline
(PBS, 0.01 M phosphate buffer, 0.0027 M potassium chlo-
ride pH 7.4) and processed for analysis.
Plate count viability assay
Collected coupons were transferred to 5 ml of PBS and 
sessile cells were removed from the coupon surface by 
vortex mixing for 30 s and sonication for 2 min (Branson 
3510, Branson Ultrasonic Corporation, Dunburry, CT, 
USA) followed by vortex mixing for another 30 s. Serial 
dilutions of the resulting cell suspensions were plated on 
tryptic soy agar (TSA, Fisher Scientific, Hampton, NH, 
USA) and incubated overnight at 37°C. Colony forming 
units (CFUs) were determined by the standard colony 
counting method. The data obtained were reported as 
number of viable bacterial cells normalized to the area 
and means were calculated. The efficacy of the anti-bio-
film material was calculated as percentage reduction of the 
CFUs in respect to the LDPE control. Four coupons for 
each surface were analyzed. The experiment was repeated 
four times for a total of 16 coupons analyzed.
Epifluorescence microscopy analysis
The percentage of live and dead cells in the biofilm bio-
mass grown on both non-functionalized and function-
alized coupons was determined using the Live/Dead 
BacLight viability kit (L7012, Molecular Probes-Life 
Technologies, Eugene, OR, USA). Biofilm was incubated 
with 2 μl of each fluorescent probe per ml of sterile filtered 
PBS at room temperature in the dark for 25 min and then 
rinsed with sterile PBS, according to the manufacturer’s 
instruction. Coupons without biofilm were also stained 
with the dyes in order to exclude any false positive signals. 
Biofilm samples were visualized using a Nikon Eclipse 
activity. Moreover, the introduction of an amino group in 
the para position of the cinnamic acid moiety of ZA and 
on the phenyl ring of SA does not affect their anti-biofilm 
activity and could be exploited for their covalent linkage 
to a polymeric support. In the light of that information, 
the cinnamic acid active moiety of ZA and SA were subse-
quently covalently immobilized on a polyethylene surface 
by Dell’Orto et al. (2017) and oriented to externally exhibit 
their active scaffold. Since no molecules were leached from 
the surface, their concentration remained constantly effec-
tive and below the lethal level, reducing the risk of devel-
oping resistant strains. Although a detailed chemical study 
has previously been performed, further development of 
this technology requires a deeper microbiological inves-
tigation. In this work, the new bioengineered materials 
were investigated in-depth to assess their anti-biofilm per-
formance by using a laboratory model system to simulate 
conditions encountered in vivo. Moreover, the synergistic 
effect of these bio-hybrid materials coupled with tradi-
tional antimicrobial agents was investigated.
Materials and methods
Polymeric surface preparation
Low density polyethylene (LDPE) coupons (round in 
shape, d = 1.27 cm) were functionalized with ZA and SA 
derivatives according to Dell’Orto et al. (2017). Briefly, the 
LDPE surface was activated with a low pressure oxygen 
plasma treatment and graft-polymerized with 2-hydrox-
yethyl methacrylate (LDPE-OH). The terminal hydroxyl 
groups of HEMA side chains were converted into the car-
boxylic acid derivatives (LDPE-COOH) by treatment with 
succinic anhydride.
The cinnamic acid active moiety of ZA and SA has been 
modified by the insertion of an amino group in the para 
position of phenyl ring to give p-amino-cinnamic acid and 
p-amino-salicylic acid. Finally, a condensation reaction
between the LDPE-COOH surface and the amino group
was performed (LDPE-CA and LDPE-SA).
Escherichia coli strain and growth conditions
E. coli strain MG1655 was used as a model system for bac-
terial biofilms, being a cosmopolitan bacterium that shares 
a core set of genes with clinically relevant serotypes and
foodborne pathogenic strains, including genes involved
in biofilm formation (Faucher and Charette 2015). The
strain was stored at –80°C in suspensions containing
20% glycerol and 2% peptone, and was routinely grown
in Luria–Bertani broth (LB, Sigma-Aldrich, St Louis, MO, 
USA) at 37°C.
E800 (Tokyo, Japan) epifluorescent microscope with 
excitation at 480 nm and emission at 516 nm for the green 
channel and excitation at 581 nm and emission at 644 nm 
for the red channel. Images were captured with a 60×, 1.0 
Numerical Aperture (NA) water immersion objective and 
analyzed via MetaMorph 7.5 software (Molecular Devices, 
Sunnyvale, CA, USA). The percentage area of stained cells 
was obtained by calculating at least 10 random images 
for each sample. The efficacy of the anti-biofilm material 
was calculated as the percentage reduction in the stained 
cells area in respect to the LDPE control images. Relative 
viability within the biofilm was determined by dividing 
the percentage area of live cells by the percentage area of 
dead cells in each sample. Four coupons of each surface 
were analyzed in each experiment. The experiment was 
repeated four times for a total of 16 coupons analyzed.
Confocal laser scanning microscopy (CLSM) analysis
The 3-D morphology of biofilm growth on non-function-
alized and functionalized surfaces was analyzed by CLSM. 
Biofilm was stained with 200 μg ml−1 the of lectin con-
canavalin A-Texas Red conjugate dye (C825, Molecular 
Probes-Life Technologies) to visualize the polysaccharide 
component of the extracellular polymeric substances 
(EPS) and 1:1,000 SYBR green I fluorescent nucleic acid 
dye (S7563, Molecular Probes-Life Technologies) to dis-
play biofilm cells, in the dark for 30 min. Coupons without 
biofilm were also stained in order to exclude any false 
positive signals. Biofilm samples were visualized using a 
Leica (Wetzlar, Germany) SP5 CLSM with excitation at 
488 nm, and emission < 530 nm (green channel). Images 
were captured with a 63×, 0.9 NA water immersion objec-
tive and projections and 3-D reconstructed images of bio-
film were generated using the Imaris software package 
(Bitplane Scientific Software, Zurich, Switzerland).
Quantitative biofilm structural parameters were cal-
culated, including (1) mean thickness, which identifies 
the mean distance from the substratum in the direction 
normal to the substratum where there is biofilm; (2) 
roughness, a quantity calculated from the thickness dis-
tribution and which describes the heterogeneity of the 
biofilm; (3) substratum coverage, the percentage of sub-
stratum area occupied by the biofilm; (4) surface-to-vol-
ume ratio, which reflects the fraction of the biofilm area 
that is exposed to the nutrients; and (5) bio-volume, of the 
cell polysaccharide matrix, which provides an estimation 
of the biomass in the biofilm (Chang et al. 2015). Biofilm 
morphological parameters were obtained via MetaMorph 
7.5 (Molecular Devices) and COMSTAT software from 
at least five random images for each sample according to 
Heydorn et al. (2000). Four coupons of each surface were 
analyzed. The experiment was repeated four times for a 
total of 16 coupons analyzed.
Antimicrobial susceptibility test
Non-functionalized and functionalized coupons with pre-
grown biofilm were removed from the CDC reactor and 
independently soaked in 15 ml of 100 μg ml−1 ampicillin 
(Amp, BP1760-5, Fisher Scientific) for 24 h and 20% eth-
anol for 20 min. Negative controls were set up by soaking 
coupons with pre-grown biofilm in 15 ml of PBS for 24 h 
or 20 min. The concentrations and the incubation time 
were chosen to mimic antibiotic therapies and daily disin-
fection procedures with ethanol in hospital and industrial 
settings. After the treatment, coupons were left in PBS for 
10 min in order to neutralize the antimicrobial agent. Cells 
in the biofilm were dislodged from the coupon and col-
lected in PBS for colony counting, as previously reported 
in the ‘plate count viability assay section’. Cells dispersed 
in the bulk liquid were washed by centrifugation at 8,000 g 
for 30 min and suspended in PBS for colony counting. 
The efficacy of both ampicillin and ethanol treatments 
was reported as the percentage reduction in respect to the 
LDPE control treated with PBS. For each experiment, four 
coupons of each surface were analyzed. The experiment 
was repeated four times for a total of 16 coupons analyzed.
The antimicrobial activity of 20% ethanol on E. coli 
biofilm pre-grown on non-functionalized and function-
alized coupons was further investigated by using a CLSM 
method that enables the direct and real-time visualization 
of cell inactivation within the biofilm structure (Davison 
et al. 2010). The imaging technique was based on mon-
itoring the loss of fluorescence that corresponds to the 
leakage of a fluorophore out of cells due to the membrane 
permeabilization by the biocides (Davison et al. 2010). 
The experiment was carried out using the FC 270 flow 
cell apparatus (BioSurface Technologies), according to the 
manufacturer’s recommendations. Biofilms pre-grown on 
the coupons were stained with 5 mM of syto-9 green flu-
orescent nucleic acid stain solution (S-34854, Molecular 
Probes-Life Technologies) in PBS at room temperature in 
the dark for 30 min and then rinsed with PBS. A solution 
of 20% ethanol was pumped through the flow cells at a 
rate of 1 ml min−1 for 20 min. The spatiotemporal decrease 
in fluorescence intensity was observed using a Leica SP5 
CLSM with excitation at 488 nm, and emission < 530 nm 
(green channel) and a 63×, 0.9 NA water immersion objec-
tive. Images were analyzed by Imaris (Bitplane Scientific 
Software) according to Davison et al. (2010). The aver-
age green fluorescence intensity was measured at differ-
ent locations within the biofilm: at the interface with the 
bulk fluid (top), the intermediate location (center) and the 
visualizations of the total biofilm biomass on functional-
ized and non-functionalized coupons and stained for live 
and dead cells. Microscope assay revealed that differences 
in dead cell data were not statistically relevant (Table 1 
and Figure 1A). Conversely, biofilms on LDPE-CA and 
LDPE-SA showed a significant decrease in the number of 
live cells, corresponding to 56.7 ± 11% and 70.6 ± 7.3% in 
respect to the LDPE control sample (Table 1, Figure 1B 
and C), confirming the results obtained in the plate count 
viability assay. No significant differences were detected in 
the live cell data between the LDPE and the LDPE-OH 
and LDPE-COOH linker samples (Table 1).
Coupons without biofilm and stained with the same 
dye did not produce detectable fluorescence and therefore 
no false positive signals were produced.
LDPE-CA and LDPE-SA affect biofilm morphology
Projection analysis as well as 3-D reconstructed CLSM 
images showed a complex biofilm on LDPE, LDPE-OH, 
LDPE-COOH, with an intense red and green signal corre-
sponding to multi-layers of cells organized in macro-colo-
nies inside a well-structured polysaccharide matrix (Figure 
2A and B). Conversely, biofilm growth on LDPE-CA and 
LDPE-SA resulted in a significant decrease in thickness 
with a monolayer of dispersed cells (green signal) and 
very low presence of a polysaccharide matrix (red signal) 
(Figure 2C–F). Coupons without biofilm and stained with 
the same dye did not produce detectable fluorescence; 
therefore false positive signals were not produced.
Biofilms on both the functionalized surfaces displayed 
a thickness below 5 μm with a decrease up to 84.5 ± 1.2% 
(Table 2). Conversely, biofilm on the LDPE control surface 
reached about 25 μm thick. Additionally, the LDPE-CA 
and LDPE-SA biofilm roughness significantly decreased 
(up to 57.0 ± 3.9%), indicating a uniform biofilm layer, 
in contrast with the patchy and more heterogeneous con-
trol biofilm. On LDPE-CA and LPDE-SA, the substratum 
coverage by biofilm was notably lower (up to 84.5 ± 2.5%) 
than in the corresponding non-functionalized counter-
part as well as the total bio-volumes (up to 92.0 ± 6.5%). 
interface with the coupon (bottom). Changes in intensity 
were normalized by the initial intensity at the beginning 
of the 20-min antimicrobial treatment. This normalized 
intensity was used to compare values of relative loss of 
fluorescence during 20-min biocide treatment periods. 
The kinetics of fluorescence loss was quantified by the 
parameter T50, the time elapsed from the initiation of 
treatment until the fluorescence intensity falls to half of 
its initial value at a particular spot. Three coupons of each 
surface were analyzed. The experiment was repeated three 
times for a total of nine coupons analyzed.
Statistical analysis
Two-tailed ANOVA analysis, via a software run in 
MATLAB environment (Version 7.0, The MathWorks 
Inc., Natick, MA, USA), was applied to statistically eval-
uate any significant differences among the samples and 
concentrations. The ANOVA analysis was carried out 
after verifying data independence (Pearson’s chi-square 
test), normal distribution (D’Agostino–Pearson normal-
ity test) and homogeneity of variances (Bartlett’s test). 
Tukey’s honestly significant different test (HSD) was used 
for pairwise comparison to determine the significance of 
the data. Statistically significant results were depicted by 
p-values < 0.05.
Results
LDPE-CA and LDPE-SA reduce biofilm biomass 
without affecting cell viability
Experiments showed that LDPE-CA and LDPE-SA had 
an optimal anti-biofilm performance, reducing viable 
adhered cells by 61.6 ± 10.5% and 60.0 ± 10.6% respec-
tively in comparison to the LDPE control surface (Table 
1). No significant differences were detected in viable 
adhered cells among the LDPE control and the LDPE-OH 
and LDPE-COOH linker samples (Table 1).
Epifluorescence microscopy was additionally used 
to provide image analysis and in situ quantification 
of bacterial cells. Figure 1 shows direct microscopic 
Table 1. Biomass within the biofilm grown on non-functionalized (lDPE, lDPE-oH, lDPE-CooH) and functionalized polyethylene surfac-
es (lDPE-CA, lDPE-SA) by plate count viability assay and epifluorescence analysis.
notes: Data represent the means ± SD of four independent measurements. Different superscript letters indicate significant differences (Tukey’s HSD, p ≤ 0.05) 
between the means of different surfaces.
Surface
Plate count assay Epifluorescence microscope analysis
Viable adhered cells 
(CFU cm−2) Live cells (%) Dead cells (%) Relative viability
lDPE (1.05 ± 0.15) × 107 a 59.82 ± 4.86 a 6.51 ± 0.81 a 8.48 ± 0.34 a
lDPE-oH (1.05 ± 0.10) × 107 a 55.79 ± 4.24 a 7.05 ± 0.42 a 7.84 ± 0.75 a
lDPE-CooH  (1.03 ± 0.17) × 107 a 59.19 ± 2.19 a 7.07 ± 0.44 a 7.37 ± 1.24 a
lDPE-CA  (0.42 ± 0.08) × 107 b 25.91 ± 4.75 b 5.98 ± 1.21 a 4.77 ± 1.22 b
lDPE-SA  (0.40 ± 0.07) × 107 b 17.61 ± 3.34 b 6.81 ± 0.70 a 2.68 ± 0.62 c
1 in biofilm grown on LDPE, meaning a predominance 
of matrix in respect to the cells. A significant reduction in 
matrix/cells ratio was found within the biofilm grown on 
LDPE-CA and LDPE-SA, with a value equal to (LDPE-SA) 
or less than 1 (LDPE-CA), confirming a relevant decrease 
in the amount of matrix. As a consequence, the biofilm 
exposed surface/bio-volume ratio significantly increased 
when biofilm was grown on LDPE-CA (16.8 ± 0.4-fold) 
and LDPE-SA (7.1 ± 0.2-fold). No significant differences 
were detected in all morphological parameters between 
biofilms grown on LDPE, LDPE-OH and LDPE-COOH 
materials.
LDPE-CA and LDPE-SA enhance biofilm 
susceptibility toward antimicrobial agents
To evaluate the synergistic effect of functionalized mate-
rials with traditional antimicrobial agents, biofilm pre-
grown on LDPE control surface and LDPE-CA and 
LDPE-SA was independently submitted to a treatment 
with 100 μg ml−1 of ampicillin and 20% ethanol.
The 24  h treatment with ampicillin did not signifi-
cantly affect the number of viable cells in the biofilm pre-
grown on the LDPE surface in comparison to the biofilm 
grown on the same surface but treated with PBS (Table 
3). Conversely, the combination of both LDPE-CA and 
LDPE-SA with ampicillin reduced biofilm biomass by 
96.1 ± 15% and 95.5 ± 17%, respectively, in comparison 
to the LDPE surface treated with PBS. Indeed, the treat-
ment with ampicillin further decreased the number of 
viable cells by 37.2 ± 6.3% and 32.0 ± 4.9% respectively 
on LDPE-CA and LDPE-SA than the functionalized sur-
faces alone did. In the bulk liquid, no statistical differences 
were detected among the number of live cells from any 
surface after the ampicillin treatment. Interestingly, PBS 
alone weakly increased the number of cells dispersed from 
LDPE-CA and LDPE-SA biofilms.
The 20-min treatment with 20% ethanol did not sig-
nificantly affect the number of viable cells in the bio-
film grown on the LDPE surface (Table 3). In contrast, 
the combination of both LDPE-CA and LDPE-SA with 
ethanol reduced biofilm biomass by 89.2  ±  12% and 
86.5 ± 8.9%, respectively, in comparison to the LDPE con-
trol surface treated with PBS. Indeed, the treatment with 
ethanol decreased the number of viable cells by a further 
20.3 ± 2.5% and 27.8 ± 2.5% on LDPE-CA and LDPE-SA, 
respectively, compared to the functionalized surface alone. 
No statistical differences were detected among the number 
of live cells from any surface after the ethanol treatment 
in the bulk liquid.
Direct time-lapse CLSM analysis of the 20% ethanol 
effect on a biofilm pre-grown on both LDPE control and 
LDPE-CA and LDPE-SA functionalized surfaces was 
Indeed, LDPE-CA and LDPE-SA reduced cellular (by 
91.2 ± 14% and 90.1 ± 14% respectively) and polysaccha-
ride matrix bio-volumes (by 99.9 ± 11% and 94.0 ± 17% 
respectively), in comparison to the LDPE control surface. 
The matrix/cells bio-volumes ratio displayed a value over 
Figure 1.  Representative epifluorescence microscope images 
of E. coli biofilm stained with live/Dead Baclight viability kit 
and grown on non-functionalized lDPE, lDPE-oH, lDPE-CooH 
surfaces (A) and lDPE-CA (B) and lDPE-SA (C) functionalized 
surfaces (60×, 1.0 nA water immersion objective). green 
fluorescence corresponds to E. coli live cells (excitation wavelength 
(λex): 480  nm and emission wavelength (λem): 516  nm) and red 
fluorescence corresponds to E. coli dead cells (λex: 581 nm and λem: 
644 nm). Scale bar = 20 μm.
both LDPE-CA and LDPE-SA significantly affected the 
integrity of the biofilm biomass, leading to a rapid and 
complete loss in fluorescent intensity (Figure 3A). The 
highest T50 occurred at the bottom of the biofilm, at the 
contact with the coupon surface, and the lowest at the bio-
film surface, at the interface with the bulk liquid. Indeed, 
on LDPE-CA, T50 values were found 10.8 ± 1.5-fold and 
1.2 ± 0.03-fold higher, respectively, at the bottom and in 
additionally performed. The penetration of ethanol was 
inferred from the loss of green fluorescence in the biofilm 
during the treatment. The time-lapse images showed that 
the ethanol treatment did not affect the green fluores-
cence of the biofilm grown on the LDPE control within 
the 20 min of the experiment (Figure 3A). Moreover, no 
spatiotemporal changes were detected (Figure 3B and C). 
Conversely, the ethanol treatment on biofilm grown on 
Figure 2. Representative projection analysis (column on the left) and 3-D-reconstructed ClSM images (column on the right) of E. coli 
biofilm grown on non-functionalized lDPE, lDPE-oH, lDPE-CooH surfaces (A, B), lDPE-CA (C, D) and lDPE-SA (E, f) functionalized 
surfaces (λex at 488 nm, and λem < 530 nm, 60×, 0.9 nA water immersion objective). live cells were stained green with SYBR green i, 
whereas the polysaccharide matrix was stained red with Texas Red-labeled concanavalin A. Scale bar = 20 μm.
(Raj et al. 2004), due to its excellent chemical resistance, low 
wetting properties in aqueous media, high impact strength, 
light weight and high flexibility (Sastri 2010). Notably, to 
date, the functionalization was performed only when pol-
ymeric material is a film, with a thickness of around a few 
nanometres. In this work 1.6 mm thickness LDPE coupons 
were used, opening the potential to extend the technology 
to other materials of a similar thickness.
In a first step, ZA and SA have been modified to be 
suitable for covalent grafting to the LDPE surface.
In a previous study the authors demonstrated that the 
cinnamic scaffold is the specific structural feature respon-
sible for the anti-biofilm performance of ZA (Cattò et al. 
2015). Moreover, these studies of the relationship between 
structure and anti-biofilm activity revealed that the anti-bi-
ofilm activity of ZA and its derivatives depended upon the 
presence of a carboxylate anion and, consequently, on its 
hydrogen-donating ability. In addition, the conjugated 
the center of the biofilm in respect to the top (Figure 3B 
and C). Also, in the biofilm on LDPE-SA T50 progres-
sively decreased from the bottom to the center, with val-
ues respectively 40.0 ± 1.2-fold and 4.5 ± 0.3-fold higher 
in comparison to the top (Figure 3B and C). T50 at the 
bottom was 3.8 ± 0.06-fold higher in the biofilm grown 
on LDPE-CA in comparison to that grown on LDPE-SA, 
while no differences in the T50 were detected at the center 
and the top of the biofilm (Figure 3C).
Discussion
In this work, the functionalization of LDPE surface with 
biocide-free anti-biofilm compounds, ie, ZA and SA, was 
carried out to obtain new non-toxic materials able to hin-
der biofilm formation. LDPE has been chosen as the most 
common widespread polymer in many applications, eg bio-
medical devices (Siddiqa et al. 2015) and food packaging 
Table 2. Morphological parameters of biofilms grown on non-functionalized (lDPE, lDPE-oH, lDPE-CooH) and functionalized polyeth-
ylene surfaces (lDPE-CA, lDPE-SA).
notes: in brackets, the percentage reduction/increase in comparison to the lDPE control sample. Data represent the means ± SD of four independent measure-
ments. Different superscript letters indicate significant differences (Tukey’s HSD, p ≤ 0.05) between the means of different surfaces.
LDPE LDPE-OH LDPE-COOH LDPE-CA LDPE-SA
Thickness (μm) 25.36 ± 7.06a 20.4 ± 4.43a 23.91 ± 7.13a 3.82 ± 1.11b 5.00 ± 1.00 b
(–84.5 ± 1.2) (–80.2 ± 1.7)
Roughness 0.37 ± 0.02a 0.36 ± 0.02a 0.35 ± 0.05a 0.18 ± 0.03b 0.16 ± 0.01b
(–51.8 ± 10.5) (–57.0 ± 3.9)
Substratum coverage (%) 68.40 ± 5.02a 64.44 ± 2.49a 62.19 ± 1.42a 14.37 ± 3.83b 10.62 ± 1.72b
(–78.9 ± 5.6) (–84.5 ± 2.5)
Surface/bio-volume (μm2 μm−3) × 10−1 0.18 ± 0.05a 0.16 ± 0.02a 0.16 ± 0.01a 3.08 ± 0.10b 1.65 ± 0.69c
(+16.8 ± 0.4-fold) (+7.1 ± 0.2-fold)
Total bio-volume (μm3 μm−2) 91.03 ± 12.25a 92.51 ± 11.66a 90.62 ± 9.20a 3.85 ± 0.09b 9.18 ± 0.21b
(–92.0 ± 6.5) (–91.3 ± 2.6)
Cell bio-volume (μm3 μm−2) 34.46 ± 7.48a 35.59 ± 3.64a 32.48 ± 4.44a 3.85 ± 0.09b 4.48 ± 0.35b
(–91.2 ± 14.1) (–90.1 ± 13.7)
Polysaccharide matrix bio-volume 
(μm3 μm−2)
59.88 ± 10.97a 56.91 ± 8.02a 58.15 ± 4.75a 0.00 ± 0.01b 3.99 ± 1.05b
(–99.9 ± 11.3) (–94.0 ± 16.7)
Matrix/cells bio-volume ratio 1.44 ± 0.23a 1.60 ± 0.06a 1.63 ± 0.13a 0.00 ± 0.00b 1.00 ± 0.13c
(–88.4 ± 3.0) (–99.9 ± 0.1)
Table 3. Viable adhered cells (Cfu cm−2) on non-functionalized (lDPE) and functionalized coupons (lDPE-CA, lDPE-SA) and those re-
leased into the surrounding bulk liquid after treatment for 24 h with 100 μg ml−1 ampicillin or PBS and treatment for 20 min with 20% 
ethanol or PBS.
notes: Data represent the means ± SD of four independent measurements. Different superscript letters indicate significant differences (Tukey’s HSD, p ≤ 0.05) 
between the means of different surfaces and treatment.
Ampicillin (24 h)
Coupon Bulk liquid
PBS Amp PBS Amp
lDPE (1.13 ± 0.18) × 107 a (9.28 ± 3.51) × 106 a (3.39 ± 0.35) × 107 a (3.57 ± 0.92) × 106 c
lDPE-CA (4.05 ± 1.07) × 106 b (4.40 ± 0.70) × 105 c (4.13 ± 0.46) × 107 b (4.69 ± 0.28) × 106 c
lDPE-SA (4.72 ± 0.52) × 106 b (5.12 ± 0.91) × 105 c (4.16 ± 0.29) × 107 b (3.84 ± 0.38) × 106 c
Ethanol (20 min)
Coupon Bulk liquid
PBS EtoH PBS EtoH
lDPE (1.07 ± 0.13) × 107 a (1.00 ± 0.04) × 107 a (3.39 ± 0.34) × 107 a (1.87 ± 0.08) × 106 b
lDPE-CA (3.32 ± 0.32) × 106 b (1.16 ± 0.16) × 106 c (4.13 ± 0.60) × 107 a (5.39 ± 0.50) × 106 b
lDPE-SA (4.41 ± 0.78) × 106 b (1.44 ± 0.15) × 106 c (4.16 ± 0.51) × 107 a (9.53 ± 0.17) × 106 b
Figure 3. Time-lapse ClSM analysis of the 20% ethanol action against E. coli biofilm pre-grown on non-functionalized and functionalized 
polyethylene surfaces within the 20 min antimicrobial susceptibility test. (A) Time-lapse ClSM representative images of biofilm pre-
grown on lDPE control surface (first line), lDPE-CA (second line) and lDPE-SA (third line) during the treatment with 20% ethanol. live 
cells were stained green with Syto 9 (λex at 488 nm, and λem < 530 nm, 60×, 1.0 nA water immersion objective). Scale bar = 20 μm. 
(B) Relative fluorescence intensity against time at the interface with the bulk fluid (gray symbols), at the intermediate location (open
symbols) and at the interface with the coupon (black symbols). fluorescent intensity in a specific biofilm region was normalized by the
initial intensity in that same region. (C) The T50 parameter (time elapsed from the initiation of treatment until the fluorescence intensity 
fell to half of its initial value at a particular spot) calculated at the interface with the bulk fluid (top), at the intermediate location (center) 
and at the interface with the coupon (bottom). nD, not determined as fluorescence intensity did not reach 50% of the initial intensity
within the experimental time.
A laboratory standard procedure using a CDC biofilm 
reactor was employed to grow E. coli biofilms, simulating 
the conditions to which surfaces of a wide range of appli-
cations are subjected during their use (Goeres et al. 2005; 
Gilmore et al. 2010; Williams et al. 2011). Experiments 
revealed that biofilms were significantly affected when 
grown on both LDPE-CA and LDPE-SA in comparison 
to the LDPE surface. Indeed, according to the anti-biofilm 
ranges proposed by Cattò et al. (2015), LDPE-CA and 
LDPE-SA displayed optimal anti-biofilm performance, 
decreasing viable biomass over the 60% in comparison to 
the control material. Moreover, in situ stained biofilm for 
live and dead bacterial cells confirmed that the reduction 
in the biofilm biomass was achieved on both functional-
ized surfaces by a mechanism that did not affect bacterial 
viability, an important factor in the challenge to limit the 
risk of developing resistant microbial strains. The results 
are in line with those previously achieved with ZA and SA 
free in solution. Villa et al. (2010) found that 500 mg l−1 
of ZA reduced E. coli cell adhesion up to 70% whereas 
Cattò et al. (2017) proved that 25 mg l−1 of SA inhibited 
E. coli biofilm by 68%. In both cases the anti-biofilm per-
formances were displayed with a mechanism more sub-
tle than simple killing activity, modulating the activity of
some ZA and SA targeted proteins and providing con-
ditions to which the best microbial strategy is to escape
from the adverse conditions, instead of persisting in the
biofilm lifestyle.
Additional experiments performed by CLSM showed 
the massive impacts of the functionalized material on 
biofilm morphology. Biofilm grown on LDPE-CA and 
LDPE-SA showed a significant decrease in thickness and 
roughness with a uniform monolayer of dispersed cells 
and a significant lower bio-volume of polysaccharide 
matrix. Conversely, LDPE control images showed a com-
plex heterogeneous biofilm, with an intense fluorescence 
signal corresponding to dense multi-layers of cells uni-
formly distributed over the substratum and organized in 
macro-colonies inside a well-structured robust polysac-
charide matrix. Cattò et al. (2017) showed that SA free in 
solution significantly decreased the amount of polysac-
charide in the matrix of the E. coli biofilm, while Vila and 
Soto (2012) hypothesized that SA plays a role in matrix 
production, decreasing the expression of the membrane 
protein, OmpA, involved in the transport of polymeric 
substances required for the formation of the EPS outside 
the cells.
The exposed biofilm surface to bio-volume ratio, which 
reflects the fraction of biofilm area that is exposed to the 
nutrient flow, was higher in the biofilm grown on the 
functionalized surfaces in contrast to the biofilm grown 
on the non-functionalized counterpart. A high surface to 
aromatic system was integral to the anti-biofilm activities 
of ZA since the presence of the double bond stabilized the 
carboxylate anion. In contrast, the sulfate group is not 
integral to the anti-biofilm activity of ZA and its presence 
in the ZA structure is part of an ecological phytochemical 
strategy to make the active cinnamic moiety more solu-
ble and thus more mobile in the water-based sap of the 
vascular transport system (Baek et al. 2010). In line with 
the previous research, in this work ZA has been modified 
by replacing the sulfate group in the para position of the 
phenyl ring with an amino group suitable for a condensa-
tion reaction with the LDPE-COOH surface. Cattò et al. 
(2015) identified p-amino-cinnamic acid as an excellent 
candidate to be covalently linked to a polymeric support. 
Moreover, p-amino-cinnamic acid displayed anti-biofilm 
activity, suggesting that the addition of an amino group 
in the para position of phenyl ring of the cinnamic acid 
moiety does not affect the anti-biofilm performance of the 
molecule. Similarly, Cattò et al. (2017) demonstrated that 
the addition of the amino group in the para position of 
the SA phenyl ring does not affect its anti-biofilm perfor-
mance. Moreover, docking calculations suggested that the 
presence of the amino group only slightly influences the 
binding mode to the tryptophanase TnaA targeted protein 
predicted for SA, with a mechanism of action that involves 
the same protein residues expected for SA. Thus, an amino 
group in the para position of phenyl ring has been added 
to the SA chemical structure, allowing its covalent binding 
with the LDPE-COOH surface.
In the past, ZA and SA were incorporated in different 
polymeric materials able to gradually release the anti-bi-
ofilm molecules into the surrounding area (Geiger et al. 
2004; Barrios et al. 2005; Rosenberg et al. 2008; Nowatzki 
et al. 2012). However, in most cases these systems showed 
several problems such as the non-uniform distribution 
of the anti-biofilm compounds inside the material and 
the formation of aggregates due to the incomplete mis-
cibility of the molecules in the polymers. In addition, a 
constant release rate of the active compound has neither 
been achieved nor monitored, limiting the effectiveness of 
the technology to only a short period of time and making 
this system often unfeasible for a wide range of practical 
applications (Barrios et al. 2005; Nowatzki et al. 2012). 
Here, these issues have all been addressed, as ZA and 
SA are uniformly distributed and oriented to externally 
exhibit their active scaffold. This allows the functionalized 
material to directly interact with bacterial cells even at the 
early stages of biofilm formation (Dell’Orto et al. 2017). 
In this work, the efficacy of the anti-biofilm materials was 
evaluated in depth with the aim of transferring the tech-
nology into consumer products suitable for a widespread 
distribution.
their effect even at a concentration below those normally 
used in traditional applications.
In addition, an integrative solution by coupling a func-
tionalized material with ethanol could have wider applica-
tions than a strategy with ampicillin. Ethanol is the most 
popular broad range bactericidal agent (McDonnell and 
Russell 1999). It is used for example in the disinfection 
of skin, medical equipment, and cooking equipment 
because it is volatile, leaves minimal residue, and is harm-
less, even if intraoral intake occurs. Despite its extensive 
and long-standing use as an antiseptic, there is no evi-
dence of acquired microbial resistance (Balestrino et al. 
2009). Importantly, ethanol is universally available with-
out restriction and with an affordable price. Conversely, 
antibiotic resistance is a global issue and sooner or later, 
after several antibiotic treatments, ampicillin resistance 
naturally occurs (Ventola 2015). Moreover, antibiotic 
treatment is limited to only few sectors of the economy, 
ie the medical sector, and antibiotic use often requires 
a prescription. On the basis of these considerations, the 
spatial and temporal patterns of ethanol penetration into 
the biofilm pre-grown on both non-functionalized and 
functionalized surfaces were additionally investigated by 
direct time-lapse CLSM analysis. This method is suitable 
for mimicking antimicrobial treatments that cause mem-
brane permeabilization, eg ethanol (Davison et al. 2010; 
Corbin et al. 2011). The biofilm on LDPE surface retained 
its initial fluorescence over the 20-min experimental test, 
while ethanol caused a loss of fluorescence in the biofilms 
pre-grown on both LDPE-CA and LDPE-SA. A distinct 
spatiotemporal pattern of fluorescence loss was observed 
for both the surfaces investigated. In both cases the anti-
microbial action of ethanol occurred very rapidly at the 
biofilm surface and slower at the bottom of biofilm, in 
contact with the coupon surface. However, at the bot-
tom of the biofilm the decrease of fluorescence intensity 
was 3.8 ± 0.06-fold more rapid in the biofilm grown on 
LDPE-SA in comparison to that grown on LDPE-CA, 
suggesting a stronger action of ethanol against biofilm 
on LDPE-SA rather than on LDPE-CA. The direct time-
lapse CLSM analysis did not give evidence of whether the 
treatment caused the killing or the removal of the bio-
mass from the biofilm or both. The loss of a green color 
proved that cell membranes were compromised, not killed 
(Corbin et al. 2011). In addition, the plate count assay of 
cells in the bulk liquid after ethanol treatment performed 
in the previous experiments suggested the presence of a 
number of live cells.
The low toxicity of both ZA and SA are promising for 
the spread of this technology to different sectors of science 
and technology, resulting in new, safe and eco-friendly 
types of products suitable for applications not only in the 
bio-volume ratio indicates the presence of small cell clus-
ters attached to the substratum, while a low value indicates 
the occurrence of larger micro-colonies within the biofilm 
(Mangwani et al. 2016). Indeed, surface to bio-volume 
ratio is an indicative parameter of adaptation of the bio-
film to the environment and it is reported to increase in 
stress conditions (Heydorn et al. 2000; Bester et al. 2011; 
Mangwani et al. 2016).
Experiments showed no significant differences in 
adhered cells, cell viability and the quantitative structural 
parameters of the biofilm between the LDPE control sur-
face and the LDPE-OH and LDPE-COOH linker surfaces. 
These results suggest that the functionalization procedure 
did not introduce modification in the LDPE structure that 
affected cell adhesion and viability as well as biofilm mor-
phology. Consequently, the anti-biofilm performance of 
LDPE-CA and LDPE-SA was totally attributable to ZA 
and SA covalently immobilized on the surface. Indeed, 
the data suggest that the ZA and SA functionalized on a 
LDPE surface are able to exert their anti-biofilm activity 
even immobilized on a surface, providing data comparable 
with those previously reported with the same molecules 
free in solution.
The combination of anti-biofilm surfaces with conven-
tional treatments could be a step forward to maximizing 
the anti-biofilm performance of the polymeric materials. 
Surfaces that retard adhesion and consequently biofilm 
formation greatly enhance the efficacy of cleaning treat-
ments and disinfection procedures. Once biofilm integrity 
is affected, bacteria are more susceptible to conventional 
antimicrobial agents than those in the intact biofilms, 
allowing reduction in the antimicrobial doses and pro-
viding more potent control against the development of 
drug-resistant strains (Cui et al. 2015; Cheesman et al. 
2017).
In this study, the ability of LDPE-CA and LDPE-SA 
to enhance biofilm susceptibility to conventional antimi-
crobial agents, ie ampicillin and ethanol, has been shown. 
Indeed, both ampicillin and ethanol significantly further 
decreased the number of viable adhered cells on both 
functionalized materials with respect to the non-func-
tionalized controls. CLSM analysis revealed that biofilms 
grown on LDPE-CA and LDPE-SA were deeply affected 
with an extremely low amount of polysaccharide matrix. 
The biofilm matrix is a barrier with the potential to reduce 
the penetration of antibiotics or biocides within the bio-
film, either by physically slowing their diffusion or chemi-
cally reacting with them (Rabin et al. 2015). Although the 
involvement of other mechanisms is not excluded, biofilm 
grown on functionalized materials might be more sus-
ceptible to antimicrobial agents, as these compounds can 
penetrate more easily within the biofilm matrix, exerting 
The work described in this study could be extended to 
other polymeric materials as well as natural molecules. 
Notably, the use of plasma technology in the grafting 
procedure makes each surface, including those that do 
not possess the required chemical features, suitable for 
covalent binding, without changing the material bulk.
There is much scope for further studies, following on 
from this pioneering work. A multitude of natural com-
pounds have shown promising anti-biofilm properties 
suitable for the development of improved effective eco-
friendly anti-biofilm materials (Villa et al. 2013). However, 
the functional groups required by these molecules to exert 
their anti-biofilm activity and, as a consequence, the bind-
ing site needed for their surface immobilization are poorly 
known, or even completely unknown.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by the Fondazione Cariplo [grant 
number 2011−0277].
ORCID
C. Cattò   http://orcid.org/0000-0002-3709-1802
F. Villa   http://orcid.org/0000-0003-2930-4684
S. Villa   http://orcid.org/0000-0002-0636-7589
F. Cappitelli   http://orcid.org/0000-0003-1237-1813
References
Ahire JJ, Hattingh M, Neveling DP, Dicks LM. 2016. Copper-
containing anti-biofilm nanofiber scaffolds as a wound 
dressing material. PLoS One. 11:e0152755. doi: 10.1371/
journal.pone.0152755.
Baek D, Pathange P, Chung J-S, Jiang J, Gao L, Oikawa A, 
Hirai MY, Saito K, Pare PW, Shi H. 2010. A stress-inducible 
sulphotransferase sulphonates salicylic acid and confers 
pathogen resistance in Arabidopsis. Plant Cell Environ. 
33:1383–1392.
Balestrino D, Souweine B, Charbonnel N, Lautrette A, 
Aumeran C, Traoré O, Forestier C. 2009. Eradication of 
microorganisms embedded in biofilm by an ethanol-based 
catheter lock solution. Nephrol Dial Transplant. 24:3204–
3209.doi: 10.1093/ndt/gfp187.
Barrios CA, Xu Q, Cutright T, Newby BM. 2005. Incorporating 
zosteric acid into silicone coatings to achieve its slow 
release while reducing fresh water bacterial attachment. 
Colloids Surf B Biointerfaces. 41:83–93. doi: 10.1016/j.
colsurfb.2004.09.009.
Bester E, Kroukamp O, Hausner M, Edwards EA, Wolfaardt 
GM. 2011. Biofilm form and function: carbon availability 
affects biofilm architecture, metabolic activity and 
planktonic cell yield. J Appl Microbiol. 110:387–398. doi: 
10.1111/jam.2011.110.issue-2.
more traditionally studied sector of the economy, ie the 
medical area, but in other challenging fields, especially 
when the use of antimicrobial agents is limited for safety 
reasons, eg in the food industry. Indeed, a food-packaging 
surface with immobilized anti-biofilm molecules would 
offer the advantage of extending the shelf life of food and 
preventing the growth of food-borne bacteria/pathogens 
at early stages, without direct addition of preservative 
agents, and such that only a low level of preservative 
comes in contact with the food. Additionally, industrial 
installation and pipelines, such as drinking water net-
works, could benefit from this research, providing safe 
material for water distribution systems that, at the same 
time, maintains high purity standard levels of water and 
limits the need for the less effective and expensive disin-
fection procedures.
Moreover, in addition to E. coli, the technology might 
be potentially used with a broad-spectrum activity against 
mixed infections. Most of the proteins targeted by ZA and 
SA are highly conserved with often a 100% of identity in a 
number of microorganisms, including pathogens that are 
a concern in the food processing industry and the health-
care sector and that are responsible for extensive damage 
to crops in the agricultural field (eg Pseudomonas spp., 
Klebsiella spp., Salmonella spp., Shigella spp., Candida spp., 
and Fusarium spp.) (Cattò et al. 2015, 2017). Additionally, 
ZA free in solution has been shown to be beneficial against 
Bacillus cereus, P. putida, Aspergillus niger, Penicillium 
citrinum, Candida albicans, Colletotricum lindemuthia-
num and Magnaporthe grisea (Stanley et al. 2002; Villa 
et al. 2010, 2011; Polo et al. 2014) while SA has shown 
activity toward P. aeruginosa, Klebsiella pneumoniae, 
Staphylococcus epidermidis, P. syringae, Pectobacterium 
carotovorum and C. albicans (Farber and Wolff 1993; 
Muller et al. 1998; Dong et al. 2012; Lagonenko et al. 
2013; Cattò et al. 2017). Thus, it is likely that LDPE-CA 
and LDPE-SA could exert their anti-biofilm activity also 
against other microorganisms.
In conclusion, the active moiety of the natural com-
pounds ZA and SA have been successfully immobilized on 
a LDPE surface, providing a new class of polymeric mate-
rials with extended shelf-life and anti-biofilm properties 
superior to the conventional biocide-based approaches in 
terms of addressing the problem of detrimental biofilms. 
Indeed, the development of materials that can resist or 
prevent bacterial adhesion in a non-leaching modality 
and with biocidal free features constitutes a promising and 
emerging approach to deal with material-associated bio-
films and an important achievement toward the dramatic 
increase of antimicrobial resistance among pathogens. 
Moreover, the covalent grafting guarantees the material has 
a long life, as the molecules become permanently attached 
and integrated into the scaffold structure of polymers.
Genet Toxicol Environ Mutagen. 749:60–69. doi: 10.1016/j.
mrgentox.2012.08.007.
Gilmore BF, Hamill TM, Jones DS, Gorman SP. 2010. 
Validation of the CDC biofilm reactor as a dynamic model 
for assessment of encrustation formation on urological 
device materials. J Biomed Mater Res B Appl Biomater. 
93:128–140.
Goeres DM, Loetterle LR, Hamilton MA, Murga R, Kirby DW, 
Donlan RM. 2005. Statistical assessment of a laboratory 
method for growing biofilms. Microbiology. 151:757–762. 
doi: 10.1099/mic.0.27709-0.
Heydorn A, Nielsen AT, Hentzer M, Sternberg C, Givskov M, 
Ersbøll BK, Molin S. 2000. Quantification of biofilm structures 
by the novel computer program comstat. Microbiology. 
146:2395–2407. doi: 10.1099/00221287-146-10-2395.
Hall-Stoodley L, Costerton JW, Stoodley P. 2004. Bacterial 
biofilms: from the natural environment to infectious 
diseases. Nat Rev Microbiol. 2:95–108. doi: 10.1038/
nrmicro821.
Lagonenko L, Lagonenko A, Evtushenkov A. 2013. Impact 
of salicylic acid on biofilm formation by plant pathogenic 
bacteria. J Biol Earth Sci. 3:B176–B181.
Lo J, Lange D, Chew BH. 2014. Ureteral stents and Foley 
catheters-associated urinary tract infections: the role of 
coatings and materials in infection prevention. Antibiotics. 
3:87–97. doi: 10.3390/antibiotics3010087.
Mangwani N, Shukla SK, Kumari S, Das S, Rao TS. 2016. 
Effect of biofilm parameters and extracellular polymeric 
substance composition on polycyclic aromatic hydrocarbon 
degradation. Rsc Adv. 6:S7540–S7551.
McDonnell G, Russell AD. 1999. Antiseptics and disinfectants: 
activity, action, and resistance. Clin Microbiol Rev. 12:147–
179.
Muller E, Al-Attar J, Wolff AG, Farber BF. 1998. Mechanism of 
salicylate-mediated inhibition of biofilm in Staphylococcus 
epidermidis. J Infect Dis. 177:501–503. doi: 10.1086/
jid.1998.177.issue-2.
Nowatzki PJ, Koepsel RR, Stoodley P, Min K, Harper A, Murata 
H, Donfack J, Hortelano ER, Ehrlich GD, Russell AJ. 2012. 
Salicylic acid-releasing polyurethane acrylate polymers as 
anti-biofilm urological catheter coatings. Acta Biomater. 
8:1869–1880. doi: 10.1016/j.actbio.2012.01.032.
Pechook S, Sudakov K, Polishchuk I, Ostrov I, Zakin V, Pokroy 
B, Shemesh M. 2015. Bioinspired passive anti-biofouling 
surfaces preventing biofilm formation. J Mat Chem B. 
3:1371–1378. doi: 10.1039/C4TB01522C.
Polo A, Foladori P, Ponti B, Bettinetti R, Gambino M, Villa F, 
Cappitelli F. 2014. Evaluation of zosteric acid for mitigating 
biofilm formation of Pseudomonas putida isolated from a 
membrane bioreactor system. Int J Mol Sci. 15:9497–9518. 
doi: 10.3390/ijms15069497.
Rabin N, Zheng Y, Opoku-Temeng C, Du Y, Bonsu E, Sintim 
HO. 2015. Biofilm formation mechanisms and targets for 
developing antibiofilm agents. Future Med Chem. 7:493–
512. doi: 10.4155/fmc.15.6.
Raj B, Udaya Sankar K, Siddaramaiah. 2004. Low density 
polyethylene/starch blend films for food packaging 
applications. Adv Polym Technol. 23:32–45. doi: 10.1002/
(ISSN)1098-2329.
Rosenberg LE, Carbone AL, Romling U, Uhrich KE, Chikindas 
ML. 2008. Salicylic acid-based poly(anhydride esters) for
Cattò C, Dell’Orto S, Villa F, Villa S, Gelain A, Vitali A, Marzano 
V, Baroni S, Forlani F, Cappitelli F. 2015. Unravelling the 
structural and molecular basis responsible for the anti-
biofilm activity of zosteric acid. PLoS One. 10:e0131519. 
doi: 10.1371/journal.pone.0131519.
Cattò C, Grazioso G, Dell’Orto S, Gelain A, Villa S, Marzano V, 
Vitali A, Villa F, Cappitelli F, Forlani F. 2017. The response of 
Escherichia coli biofilm to salicylic acid. Biofouling. 33:235–
251. doi: 10.1080/08927014.2017.1286649.
Chang YW, Fragkopoulos AA, Marquez SM, Kim HD, 
Angelini TE, Fernandez-Nieves A. 2015. Biofilm formation 
in geometries with different surface curvature and oxygen 
availability. New J Phys. 17:033017. doi: 10.1088/1367-
2630/17/3/033017.
Cheesman MJ, Ilanko A, Blonk B, Cock IE. 2017. Developing 
new antimicrobial therapies: are synergistic combinations of 
plant extracts/compounds with conventional antibiotics the 
solution? Pharmacogn Rev. 11:57–72.
Chen M, Yu Q, Sun H. 2013. Novel strategies for the prevention 
and treatment of biofilm related infections. Int J Mol Sci. 
14:18488–18501. doi: 10.3390/ijms140918488.
Coenye T, De Prijck K, Nailis H, Nelis HJ. 2011. Prevention of 
Candida albicans biofilm formation. Open Mycol J. 5:9–20. 
doi: 10.2174/1874437001105010009.
Corbin A, Pitts B, Parker A, Stewart PS. 2011. Antimicrobial 
penetration and efficacy in an in vitro oral biofilm model. 
Antimicrob Agents Chemother. 55:3338–3344. doi: 10.1128/
AAC.00206-11.
Cui JH, Ren B, Tong YJ, Dai HQ, Zhang LX. 2015. Synergistic 
combinations of antifungals and anti-virulence agents to 
fight against Candida albicans. Virulence. 6:362–371. doi: 
10.1080/21505594.2015.1039885.
Davison WM, Pitts B, Stewart PS. 2010. Spatial and temporal 
patterns of biocide action against Staphylococcus epidermidis 
biofilms. Antimicrob Agents Chemother. 54:2920–2927. 
doi: 10.1128/AAC.01734-09.
Dell’Orto S, Cattò C, Villa F, Forlani F, Vassallo E, Morra 
M, Cappitelli F, Villa S, Gelain A. 2017. Low density 
polyethylene functionalized with antibiofilm compounds 
inhibits Escherichia coli cell adhesion. J Biomed Mater Res 
A. 105:3251–3261. doi: 10.1002/jbm.a.v105.12.
Dong Y, Huang C, Park J, Wang G. 2012. Growth inhibitory 
levels of salicylic acid decrease Pseudomonas aeruginosa fliC 
flagellin gene expression. JEMI. 16:73–78.
Farber BF, Wolff AG. 1993. Salicylic acid prevents the adherence 
of bacteria and yeast to silastic catheters. J Biomed Mater 
Res. 27:599–602. doi: 10.1002/(ISSN)1097-4636.
Faucher SP, Charette SJ. 2015. Bacterial pathogens in the non-
clinical environment. Front Microbiol. 6:331.
Geiger T, Delavy P, Hany R, Schleuniger J, Zinn M. 
2004. Encapsulated zosteric acid embedded in poly 
3-hydroxyalkanoate coatings - protection against biofouling. 
Polym Bull. 52:65–72.
Gharbi A, Humblot V, Turpin F, Pradier CM, Imbert C, Berjeaud 
JM. 2012. Elaboration of antibiofilm surfaces functionalized 
with antifungal-cyclodextrin inclusion complexes. FEMS 
Immunol Med Microbiol. 65:257–269. doi: 10.1111/j.1574-
695X.2012.00932.x.
Ghosh M, Manivannan J, Sinha S, Chakraborty A, Mallick 
SK, Bandyopadhyay M, Mukherjee A. 2012. In vitro and 
in vivo genotoxicity of silver nanoparticles. Mutat Res 
Vila J, Soto SM. 2012. Salicylate increases the expression 
of marA and reduces in vitro biofilm formation in 
uropathogenic Escherichia coli by decreasing type 1 fimbriae 
expression. Virulence. 3:280–285. doi: 10.4161/viru.19205.
Villa F, Albanese D, Giussani B, Stewart PS, Daffonchio 
D, Cappitelli F. 2010. Hindering biofilm formation 
with zosteric acid. Biofouling. 26:739–752. doi: 
10.1080/08927014.2010.511197.
Villa F, Pitts B, Stewart PS, Giussani B, Roncoroni S, Albanese 
D, Giordano C, Tunesi M, Cappitelli F. 2011. Efficacy 
of zosteric acid sodium salt on the yeast biofilm model 
Candida albicans. Microb Ecol. 62:584–598. doi: 10.1007/
s00248-011-9876-x.
Villa F, Villa S, Gelain A, Cappitelli F. 2013. Sub-lethal 
activity of small molecules from natural sources and their 
synthetic derivatives against biofilm forming nosocomial 
pathogens. Curr Top Med Chem. 13:3184–3204. doi: 
10.2174/15680266113136660225.
von Eiff C, Jansen B, Kohnen W, Becker K. 2005. Infections 
associated with medical devices: pathogenesis, management 
and prophylaxis. Drugs. 65:179–214. doi: 10.2165/00003495-
200565020-00003.
Williams DL, Woodbury KL, Haymond BS, Parker AE, 
Bloebaum RD. 2011. A modified CDC biofilm reactor to 
produce mature biofilms on the surface of peek membranes 
for an in vivo animal model application. Curr Microbiol. 
62:1657–1663. doi: 10.1007/s00284-011-9908-2.
control of biofilm formation in Salmonella enterica serovar 
Typhimurium. Lett Appl Microbiol. 46:593–599. doi: 
10.1111/j.1472-765X.2008.02356.x.
Sastri VR. 2010. Commodity thermoplastics: polyvinyl 
chloride, polyolefin, and polystyrene. In: Sastri VR, editor. 
Plastics in medical devices: properties, requirements, and 
applications. Amsterdam (NL): Elsevier Inc; p. 73–119.
Siddiqa AJ, Chaudhury K, Adhikari B. 2015. Hydrophilic low 
density polyethylene (LDPE) films for cell adhesion and 
proliferation. RRJOM. 1:43–54.
Simões M, Simões LC, Vieira MJ. 2010. A review of current 
and emergent biofilm control strategies. LWT - Food Sci 
Technol. 43:573–583. doi: 10.1016/j.lwt.2009.12.008.
Sobieh SS, Kheiralla ZMH, Rushdy AA, Yakob NAN. 2016. 
In vitro and in vivo genotoxicity and molecular response of 
silver nanoparticles on different biological model systems. 
Caryologia. 69:147–161. doi: 10.1080/00087114.2016.1139416.
Sousa ACA, Pastorinho MR, Takahashi S, Tanabe S. 2014. History 
on organotin compounds, from snails to humans. Environ 
Chem Lett. 12:117–137. doi: 10.1007/s10311-013-0449-8.
Stanley MS, Callow ME, Perry R, Alberte RS, Smith R, Callow 
JA. 2002. Inhibition of fungal spore adhesion by zosteric 
acid as the basis for a novel, nontoxic crop protection 
technology. Phytopathology. 92:378–383. doi: 10.1094/
PHYTO.2002.92.4.378.
Ventola CL. 2015. The antibiotic resistance crisis: part 1: causes 
and threats. Pharmacol Ther. 40:277–283.