* Corresponding authors:
Shaokoon Cheng - School of Mechanical Engineering, Macquarie
University, 44 Waterloo Rd., Macquarie Park, Sydney, NSW, 2113,
Australia
Email:
shaokoon.cheng@mq.edu.au
Hui Xin Ong - Woolcock Institute of Medical Research & Faculty
of Medicine and Health, The University of Sydney, Sydney, NSW, 2037,
Australia
Email: ong.hui@sydney.edu.au
KEYWORDS: Biofilm, static culture, biofilm substrate, interfaces,
dual-chamber microreactor, reversible bonding
ABSTRACT
Biofilms are ubiquitous and notoriously difficult to eradicate and
control, complicating human infections, industrial and agricultural
biofouling. Current biofilm studies are commonly performed with the
biofilm cultured on mono-interfaces and generally have neglected to
consider more realistic models or approaches, where diverse interfaces
are involved. In our study, a reusable dual-chamber microreactor with
interchangeable membranes was developed to establish multiple interfaces
for biofilm culture and test. Protocol for culturing Pseudomonas
aeruginosa (PAO1) on the air-liquid interface (ALI) and liquid-liquid
interface (LLI) under static environmental conditions for 48h was
optimized using this novel device. This study shows that LLI model
biofilms are more susceptible to physical disruption compared to ALI
model biofilm. SEM images revealed a unique ’mushroom-shaped’
microcolonies morphological feature, which is more distinct on ALI
biofilms than LLI. Furthermore, the study showed that ALI and LLI
biofilms produced a similar amount of extracellular polymeric substances
(EPS). As differences in biofilm structure and properties may lead to
different outcomes when using the same eradication approaches, the
antimicrobial effect of an antibiotic, Ciprofloxacin (CIP), was chosen
to test the susceptibility of 48h-old P. aeruginosa biofilms
grown on ALI and LLI. Our results show that the minimum eradication
concentration (MBCE) of CIP using our dual-chamber device reached
1600μg/ml, which is significantly higher than the conventional
microtiter plate method (64μg/ml). The results highlight the importance
of having a model that can closely mimic in-vivo conditions to
develop more effective biofilm management strategies.
Introduction
Biofilms are the dominant surviving model of bacteria that exist on
earth [1]. It plays a crucial part in the ecosystem, either
beneficial or detrimental depending on the microbial species and their
growth. For example, crops can benefit from non-pathogenic biofilm
growth-promoting rhizobacteria (PGPR) biofilm [2] but can also
trigger foodborne illnesses in plant diseases caused by pathogenic
microbial biofilms [3]. Also, while biofilms can help degrade
pollutants in liquid and gaseous effluents in wastewater treatment
plants [4], undesired biofilm formation in drinking water, oil
pipelines, and ship hulls can lead to biofouling and biocorrosion, which
undermine operation safety and loss of productivity [4].
Furthermore, biofilms are typically associated with clinical chronic,
nosocomial, and medical device-related infections [5-8].
The formation of biofilm starts with planktonic bacteria adhering to the
surface and encasing the proliferated colonies in self-produced
extracellular plyometric substance and become matured biofilm [1].
Biofilms protect the microorganism from hostile physical and chemical
environments such as altered pH, osmolarity, nutrients scarcity,
mechanical and shear forces, and block bacterial biofilm communities’
access from antibiotics and host’s immune cells [9]. Their
properties are determined by both inherent biological attributes of
bacterial strains and external environmental factors. Among all the
environmental factors, surfaces play a critical role since their
properties directly affect bacteria’s initial attachment, biofilm
maturation, and final detaching. Studies on the impact of surface
roughness, charge, hydrophobicity, tension, wettability, and
microtopography on biofilm growth [10, 11] have persistently modeled
biofilm as solid-attached structures submerged in liquid. Other studies
have explored the development of biofilm aggregates cultured in agarose
have been used to mimic the intrinsic properties of human tissue in
various fields [12, 13], such as wound microenvironment. However,
the variations in terms of interface type have been neglected.
Furthermore, antibiotics have been widely used to treat biofilm
infection diseases with multiple delivery paths, including parenteral,
enteral, transdermal, inhalation, oral, and topical. The interaction
between antibiotics and biofilms also varies depending on interfaces. A
greater understanding of the influence of interfaces on biofilms will
help to develop more efficient and targeted eradication approaches.
For this purpose, we proposed a dual-chamber microreactor with
interchangeable interfaces to study biofilm structure and susceptibility
to drugs. Our model brings the advantages of a microfluidic device,
which include reduced sample volume and reagent consumption and precise
environmental control. Moreover, the curing approach and reversible
bonding technique adopted for device fabrication and assembling, further
helps to reduce experimental time and cost and brings greater
flexibility for sample manipulation. Compared with conventional
antimicrobial susceptibly testing methods, our device can be considered
more physiobiologically relevant in mimicking the microenvironment that
biofilm grows and interacts with antibiotics.
Pseudomonas
aeruginosa is a ubiquitous pathogen that could colonize multiple
environments and establish a biofilm within 24 h [14]. It can cause
a variety of infections, such as chronic lung infection [15] and
urinary tract infections (UTIs) [16], and its accentuated antibiotic
resistance during biofilm growth poses a significant threat to the
medical community. The biofilm developed in the respiratory system can
be modeled as biofilm formed on the air-liquid interface (ALI), while
biofilm developed in the urinary tract can be best characterized by
liquid-liquid interface (LLI) model biofilm. In this study, we used this
novel device to optimize a protocol for culturingPseudomonas
aeruginosa (PAO1) on ALI and LLI, respectively, under static
environmental conditions for 48h. Biofilm susceptibility was also
evaluated after exposure to ciprofloxacin from the substrate side to
mimic systemic circulation exposure, which permeates through a physical
barrier to reach the biofilm site.
Materials and Methods
Materials
Polydimethylsiloxane monomer and curing agent (Sylgard 184, Dow Corning)
were purchased from Revolution Industrial (Australia). Formlab® Clear
resin V4 was obtained from Core Electronics (Australia). Isopropyl
alcohol (IPA) (≥99%) and ethanol (80% v/v) were obtained from
Chem-supply Pty Ltd (Australia). Ambersil® polymer remover was purchase
from RS components Pty Ltd (Australia). Hydrophilic porous polyester
(PETE) membrane (pore size 0.2μm, pore density 3 x 108
(pores/cm2), open area 9.4%, thickness 10μm, bubble
point 20psi, water flow rate 10mL/min/cm2, airflow
rate 3L/min/cm2) was purchased from Sterlitech (USA).
Translucent silicon rubber (TRANSIL®) was obtained from Barnes Products
Pty Ltd (Australia). Tygon tubing (1.59mm OD x 0.51mm ID), 23G stainless
steel couplers (0.63mm OD x 0.33mm
ID), and razor-sharp stainless-steel biopsy punches (0.5mm and 1.25mm
OD) were purchased from Darwin Microfluidics (France). Stainless steel
screws and nuts were obtained from Small Parts & Bearings (Australia).
Blunt needles (22G) and sterile disposable syringes were obtained from
Livingstone (Australia). Pseudomonas aeruginosa (PAO1 ATCC 15692)
was purchased from the American Type Culture Collection (ATCC,
Rockville, USA). Cation-adjusted Mueller-Hinton Broth (CAMHB) was
purchased from BD Biosciences (Australia). Phosphate-buffered saline
(PBS), LB Broth with agar (Lennox), Alcian blue (1% w/v in acetic acid,
3% v/v pH 2.5), AeraSeal® Film, and formaldehyde solution (≥36.0% in
water) were purchased from Sigma Aldrich (Australia). FilmTracer™
LIVE/DEAD® Biofilm viability kit was obtained from Thermo Fisher
Scientific (Australia). Ciprofloxacin hydrochloride was supplied by MP
(Biomedical Australasia Pty Limited, Australia). Milli-Q water was
obtained from Biopak® Polisher system (Merck KGaA, Germany).
Device design and fabrication
The microreactor (Fig.1) consists of a dual-chamber design (basal and
apical) physically separated by a porous hydrophilic PETE membrane, used
as the substrate for biofilm growth. Each device comprised three
identical dual-chamber flow cells allowing for the testing of multiple
experimental setups in parallel. The dimensions of the chamber were: 1mm
(width) x 4mm (length) x 0.5mm (height). The flow chambers (Fig. 1-B)
were fabricated using PDMS by 3D printing microfluidic fabrication
technique [17]. The molds for PDMS casting were printed using a
clear resin on a Form2 3D printer (Formlabs), washed for 15 min with IPA
(Formlabs, Form wash), and cured under UV light for 90 mins at 65°C
(Formlabs, Formcure). Before pouring the PDMS, the mold was spray-coated
with polymer remover to protect the mold surface and to assist with the
PDMS removal after the curing process. PDMS monomer and curing agent
were mixed (10:1 w/w), cast into the mold, degassed, and cured at 65°C
for 12h. After cooling, the PDMS chamber was peeled off from the mold,
and the inlet and outlet holes were punched using a 0.5mm biopsy
puncher. Two pieces of thin silicon gaskets (0.3mm) were placed between
the membrane and the PDMS layers to prevent leakage and protect the
membrane from the mechanical bounding force. The silicone gaskets (Fig.
1-C) were fabricated by pouring
silicone on the plastic sheet and
pressing the 3D printed template into the uncured silicone mix and
against the plastic sheet. The silicone gaskets were cured at room
temperate for 25min and then peeled from the plastic sheet.
Device assembly
Reversible mechanical bonding of the multiple layers (Fig. 1-A) was
obtained using two pieces of acrylic Plexiglas covers (76mm × 25 mm x
3mm) with holes aligned with the position of the inlet, outlet and
bonding holes in the PDMS layers. The mechanical bonding was achieved by
twelve screws and nuts positioned around the perimeter of the flow
chamber to ensure the alignment of the layers and the interface area
between the two chambers. The flow into the channels was supplied
through microbore tubing connected to stainless-steel couplers (23G),
inserted into the inlet and outlet ports. Before the bacteria
inoculation, 1 mL sterilized water was injected into two chambers to
ensure no leaking occurs.