Semi-automatic instrumentation for nucleic acid extraction and purification to quantify pathogens on surfaces

Won-Nyoung Lee , Hyun Jin Yoo , Kim Huyen Nguyen , Changyoon Baek * and Junhong Min *
School of Integrative Engineering, Chung-Ang University, Heukseok-dong, Dongjak-gu, Seoul, 06974, South Korea. E-mail: cybaek@cau.ac.kr; junmin@cau.ac.kr

Received 16th May 2019 , Accepted 22nd September 2019

First published on 27th September 2019


Public lavatories may cause the spread of infectious pathogens because they are enclosed spaces that both healthy people and patients can use. Thus, surface analysis for microbial contamination in public lavatories is of great importance because it is considered as an indicator of hygiene control. Herein, we developed polymerase chain reaction (PCR)-compatible surface sample preparation tools to increase the detection sensitivity and reproducibility within a short time using a semi-automatic detection system. The bacteria and viruses on different surfaces were collected using half A4-sized wipes. The wipes were treated through four different processes in a cartridge: (1) the pathogens were transferred from the wipes to the aqua phase using simple gentle vortexing; (2) the bacteria and viruses were concentrated by adsorption on the graphene surface; (3) the pathogens on the graphene layer were perfectly lysed using bead-beating tools and (4) the released DNA/RNA was collected in a microtube. The prepared nucleic acid sample was amplified using PCR or loop-mediated isothermal amplification (LAMP). At least one order of magnitude higher sensitivity was achieved using the wipe collecting method compared to that achieved using the normal swab method. This was confirmed using a semi-automatic cartridge for the wipe sampling in a lavatory hygiene test.


Introduction

Biological contamination analysis on surfaces is one of the key technologies widely used in various fields such as food safety, indoor environmental quality, clean rooms, and emergency room management.1–3 Measuring surface bio-contamination has become highly important as a very effective tool in determining the spread of infectious bacteria/viruses.4 This is not only because surfaces can become potential pathogen-transfer mediators, but also because some airborne bacteria or viruses can sink into the surfaces.5–7

Accordingly, several simple sample preparation tools have been studied.8–20 The tape-lift method utilizes a sticky transparent tape to collect pathogens from the surface for the visual count method using a microscope.8 It is very simple to collect pathogens from the surface; however, it cannot be applied to molecular diagnostic tools. The direct-bulk sampling method involves the collection of the surface sample itself, for example, a carpet, wallpaper fragments, etc.9 This collected surface can be washed by a proper aqueous solution, and a relatively large amount of sample can be collected; however, additional serial processes such as the culturing or separation of surface materials are required. The most common is the cotton swab method.10–12 After rubbing certain areas of a surface with a swab, the swab is immersed in a liquid buffer to transfer the infectious bacteria present on the surface. It is easy to use without complex devices and is suitable for measuring narrow or small surfaces, but there is a limit to its use in investigating large surfaces.13 To overcome this limit, wet wipes wider than swabs were introduced.14,15 In the case of biological sampling kits (BiSKit™), the Large Area Sampling Kit (Quick Silver Analytics, Inc., USA) uses a sponge-type surface contact part (aqua form) that can capture various infectious bacteria/viruses from wide surfaces, which are then immersed in a buffer solution (normally culture media) and cultured for a few days.15 Other methods have been investigated for the effective collection/transfer of pathogens from surfaces to solutions in the measurement of biological surface contamination.11,12,16–18 For example, viscose-based nonwoven wipe materials were developed to perform pathogen collection from surfaces and culture the solution directly.19 These wipes and sponge-type materials are very effective to collect pathogens from relatively large surfaces. However, long time-culturing processes with sophisticated equipment such as cell incubators are still required for the quantitative measurement of the pathogenic bacteria or viruses on the surface. Moreover, to date, protocols to culture various unculturable viruses such as noroviruses in the lab have not been developed.21,22

A microfluidic platform was also employed to measure pathogens in various samples (water, swab, air, and surface) through sample treatment and nucleic acid amplification processes in a fluidic chip.20 Viable bacteria could be detected using the propidium monoazide (PMA)-based real-time polymerase chain reaction (PCR).2 It was reported that viable pathogenic E. coli could be detected via mRNA amplification using NASBA (nucleic acid sequence-based amplification).23 Despite these different studies, very simple techniques have not been developed that can be applied to nucleic acid diagnostic methods with high sensitivity by collecting and processing samples on relatively large surfaces.

To develop a simple, highly effective method for analysing biological surface contamination, it is necessary to develop an easy-to-use (automatically available) sample preparation process that (1) will be able to collect a variety of bacteria/viruses from relatively large surfaces, (2) concentrate the target bacteria/viruses in the aqua phase (>20 mL) within a reasonable time, and (3) lyse the bacteria and viruses to extract nucleic acids effectively.

In this study, for the first time, we developed a full sample preparation method to achieve PCR on available nucleic acids from pathogens gathered from relatively wide surfaces using multifunctional (pathogen concentration and lysis) microbeads coated with graphene oxide. As shown in Scheme 1, (1) the pathogens (bacteria/viruses) were collected from the surfaces by wiping the surfaces with non-woven fabric materials. (2) The pathogens were moved back to a buffer solution (buffer 1, B1) by shaking the non-woven fabric with a magnet. (3) The buffer solution (buffer 1 + buffer 2, (B1 + B2)) including the pathogens was passed through a mini-chamber containing graphene oxide-coated microbeads, where the pathogens were non-specifically adsorbed on the graphene oxide surfaces via electrostatic forces/van der Waals forces24,25 and became concentrated. (4) After the addition of an elution buffer solution (buffer 3, B3), DNA was extracted from the pathogens on the surfaces through the direct beating of the graphene oxide-coated microbeads. (5) The extracted DNA was eluted into a microtube and (6) amplified using loop-mediated isothermal amplification (LAMP) or real-time PCR. The efficiencies of the pathogen collection were investigated in terms of pathogen transfer (surfaces wiped to solutions eluted), material of wipes and wetting conditions of the surfaces and wipes. The bacteria and virus adsorption on the graphene oxide surfaces was also optimized in terms of buffer pH and flow rate. For the ease of use, the processes suggested in this scheme were integrated into a single semi-automatic cartridge consisting of a wipe container, three different buffer chambers, one microbead chamber, and two microchannels. The performance of the simple cartridge suggested in this work was tested using various surface samples and compared with that of a commercial swab kit.


image file: c9an00896a-s1.tif
Scheme 1 Schematic diagram of the entire process for detecting infectious pathogens.

Experimental

Materials

Acetate buffer (pH 5, 1 M) and Tris-HCl (pH 7 and 9, 10–100 mM) were used as the main solutions for the transfer buffer (B1), adsorption buffer (B2), and elution buffer (B3). The experiment used three different wipes (21.0 × 14.8 cm2) made of (1) rayon-polyester (PE) (Cleanjoy, South Korea), (2) polyethylene terephthalate (PET)-PE (Sunwooland, South Korea), and (3) PP-PET (Green Gate Network, South Korea).

Graphene oxide synthesis

Graphene oxide (GO) was synthesized via a modified Hummer's method.26 Briefly, 1 g of graphite powder (Sigma-Aldrich, MO, USA) was added to 23 mL of H2SO4 and vigorously stirred for 12 h. Then, 3 g of KMnO4 (Sigma-Aldrich, MO, USA) was added to the suspension and kept at 35 °C for 30 min. The temperature was increased up to 70 °C and kept for 45 min. Next, 140 mL of distilled water and 10 mL of 30% H2O2 were added and stirred for 1 h. The synthesized GO was rinsed with 5% HCl and distilled water.

Bacteria and virus preparation

Representative bacteria and viruses were chosen as target pathogens in the experiment. Staphylococcus aureus (S. aureus, ATCC 6538P, Gram-positive bacteria) was provided from the BioNano Health-Guard Research Center and Escherichia coli (E. coli, ATCC 15489, Gram-negative bacteria) was purchased from the American Type Culture Collection (Washington DC, USA). Both bacteria were cultured in 3% tryptic soy broth media (BD Bioscience, NJ, USA) at 37 °C and 150 rpm in a shaking incubator (Biofree, South Korea). Red Fluorescence Protein (RFP)-containing bacteria, donated by Prof. Dokyun Na (same department in Chung-Ang University), were incubated for 12 h at 37 °C and 150 rpm in Chloramphenicol (0.1%) containing 3% Difco™ LB Broth and Miller media (BD Bioscience, NJ, USA). Human adenovirus type 5 and HEK 293 cells were obtained from Prof. Dai-wu Seol (College of Pharmacy, Chung Ang University, Korea) and cultured in Dulbecco's MEM containing 10% fetal bovine serum (Gibco, Grand Island, NY, USA). The virus was infected with HEK 293 cells and cultured for 2 days. The cultured virus was obtained by freeze–thaw cycling.

Amplification-based quantitation of bacteria and virus

The concentration of bacteria (E. coli and S. aureus) and adenovirus was measured by real-time PCR (LightCycler® 480 Instrument II, Roche, Basel, Switzerland) with a master mixture (TB Green™ Premix Ex Taq™ (Takara Bio Inc., Japan)) and primer sets (for E. coli, F: 5′-ACT TCG ACA AAT ATG CTG-3′ and R: 5′-CGG GAT GAT GTT CTG GAA A-3′; for S. aureus, F: 5′-GTT GCA TCG GAA ACA TTG TGT-3′, R: 5′-ATG ACC AGC TTC FFT ACT ACT AAA GAT-3′ and for adenovirus, F: 5′-GGT GTC GCG CTT GCC TAC TA-3′, R: 5′-CGA TCG CGT TGT TCA TAA G-3′).27,28 The QIAamp DNA mini kit (Qiagen, Hilden, Germany) for S. aureus and E. coli and QIAamp MinElute virus spin kit for adenovirus were used as normalization basic tools in the lysis test. The quantification of the bacteria and viruses was achieved based on the calibration curve (Fig. S1).

Multifunctional microbead preparation

With the piranha solution (the ratio of 98% H2SO4[thin space (1/6-em)]:[thin space (1/6-em)]35% H2O2 is 3[thin space (1/6-em)]:[thin space (1/6-em)]1), 70–100 μm glass beads (Daihan Scientific, Korea) were cleaned for 30 min and rinsed with deionized water. The beads were dried at 70 °C for 3 h. Three different materials were coated on the clean glass beads. The TEOS-coated glass beads were immersed in ethanol with 5 mM tetraethyl orthosilicate (TEOS, 99%) (Sigma-Aldrich, MO, USA) and 5 mM ammonium hydroxide solution (28.0–30.0%) (Sigma-Aldrich, MO, USA) for 1 h at 25 °C. The beads were washed three times with ethanol and dried in air at 80 °C for 1 h.27 The glass beads were immersed in 5 mM solution of (3-aminopropyl)triethoxysilane (APTES, 98%) (Sigma-Aldrich, MO, USA) containing 95% ethanol for 1 h at room temperature to coat the APTES on the glass beads. The beads were washed three times in ethanol and dried at 120 °C for 2 h.29 To coat GO on the glass beads, the APTES-coated glass beads were immersed in GO (1 mg mL−1) dispersed in ethanol for 2 h. The GO-coated beads were then stabilized under mild bead-beating in water. Finally, they were washed and dried in air at 70 °C.30

Microscopic investigation of wipes and RFP-bacteria

The fabric structure of the three different wipes was qualitatively examined with a microscope (Eclipse 80i, Nikon, Japan). The collection of RFP-bacteria on the wipes was visually tested using florescence microscopy (CELENA™ S Digital Imaging System, Logos Biosystems, South Korea).

The quantification of the adsorption of bacteria/virus on the microbeads

First, 0.4 g of microbeads (70–100 μm diameter, TEOS-, APTES-, or GO-coated glass microbeads) was filled into a microchamber (9 mm height and 12 mm diameter, Fig. S2). Then, 20 mL of buffer solution (B1 + B2) containing S. aureus (104 CFU mL−1) and adenovirus (104 PFU mL−1) was passed through the microbeads at 1–20 mL min−1. Adsorption % was calculated based on 100× the amount of bacteria or viruses adsorbed on the surface of the microbeads/those in the input solution. To count the amount of bacteria/viruses adsorbed on the surface of the microbeads and that in the input solution, the bead beating method (50 Hz and 3 min) was employed.31 The extracted DNA was directly amplified by real-time PCR and converted to the concentration (Fig. S1).

Typical sample preparation conditions for artificial biologically contaminated surfaces

The biologically contaminated wide surfaces were treated in the following manner. Solutions (200 μL) containing S. aureus, E. coli, (105, 104, and 103 CFU mL−1) and/or adenovirus in various concentrations (105, 104, and 103 PFU mL−1) were uniformly spread on polished surfaces (40 × 40 cm2) and allowed to dry for 40 min at room temperature. These prepared artificial surface samples were completely wiped using various kinds of wipes. The used wipes were placed in a 50 mL syringe or a chamber containing 20 mL of transfer buffer (B1) and a neodymium magnet and vigorously mixed using a magnetic stir plate (25 Hz, 1 min). The wipes in the B1 buffer were squeezed, and the recovered solution was mixed with 1 mL adsorption buffer (B2). The mixed buffer solution was passed (1–20 mL min−1) through the concentration/lysis chamber containing the multifunctional glass microbeads (0.4 g) and a single neodymium magnet (3 mm diameter, 6 mm thickness), on which the bacteria/viruses in the solution were adsorbed. In the chamber, a 200 μL lysis/elution buffer (B3) was loaded and vigorously mixed (25 Hz, 1 min). Finally, a B3 buffer containing the nucleic acids extracted from the bacteria/virus was obtained.

LAMP assay

LAMP assays were performed at 65 °C for 30 min in a 25 μL total mixture including 12.5 μL WarmStart® Colorimetric LAMP 2X Master Mix (New England Biolabs, MA, USA), 2.5 μL LAMP primer solution and 1 μL sample. All primers were selected from published works,32 while amplification was confirmed by the naked eyes (color changed from red to yellow).

Results and discussion

Wipe sample preparation cartridge design

To maximize user convenience, a semi-automatic sample preparation cartridge was designed, as shown in Fig. 1. The cartridge was designed to reduce the number of channels and chambers. It contained three different parts: a wipe container, a buffer solution reservoir (top part), and a channel control containing two microchannels and a microbead chamber (bottom part). The wipe container (32 mm diameter, 80 mm high) containing a magnet bar (25 mm diameter, 5 mm thickness) was in the form of a syringe (Fig. S2). The wipes were stored in the wipe container with a syringe piston-like lid (P1) after collecting the sample from the surfaces. This was installed in the C2 chamber in the top part, which has four reservoirs (C1 chamber for B1 buffer, C2 chamber for wipe container, C3 chamber for B2 buffer and C4 chamber for B3 buffer) and a single channel (TH5–TH6). First, the B1 buffer was moved from the C1 chamber to the C2 chamber via the BH7–BH8 channel in the bottom part by pulling P1, resulting in the wipes becoming fully wet and the bacteria and virus becoming desorbed from the wipes by magnetic stirring. The wipes in the C2 chamber were squeezed by (P1) pushing and the B1 buffer was moved to the C3 chamber via the BH7–BH8 channel (bottom part was rotated) to mix the B1 buffer with the B2 buffer in the C3 chamber. The bacteria and virus were present in the mixture of B1 + B2 buffers in the C3 chamber. This mixture (B1 + B2 buffers) was moved from the C3 chamber to the C2 chamber via the C5-BH10 channel by pulling P1. During the movement of the buffer, the bacteria and virus were adsorbed on the surface of the microbeads in the C5 chamber (bottom part was rotated). By half-pushing another piston-like lid (P2), the B3 buffer in the C4 chamber was moved to the C5 chamber, in which the bacteria and virus were present on the surface of the microbeads (bottom part was rotated). DNA was directly extracted through the rotational beating of the beads on which the bacteria and virus were absorbed. By fully pushing P2, the B3 buffer in the C5 chamber moved to the microtube via the C5-BH10 and TH5–TH6 channels, resulting in the preparation of PCR-ready DNA. All the moving part protocols are shown in table in Fig. 1.
image file: c9an00896a-f1.tif
Fig. 1 Design of the cartridge for the PCR-compatible surface sampling tool and a drawing of the sample preparation process.

Step 1: Bacteria/virus collection from wide surfaces using wipes

The first step towards measuring biological surface contamination is collecting bacteria or viruses from the surface. Three different wipes (Rayon/PE, PET/PE, and PP/PET) were prepared under various conditions (wet/dried wipes and wet/dried surfaces). The three kinds of wipes did not have any significant physical differences when observed under a bright-field microscope, as shown in Fig. S3. Artificial biologically contaminated surfaces (with RFP E. coli (107 CFU cm−2)) were prepared to rapidly confirm the collection of the bacteria by a simple wipe. Fig. 2a shows the fluorescence images of the artificial sample surface and wipes used in this experiment before and after collecting RFP E. coli from the surface using wipes under wet and dry conditions. RFP E. coli was well-spread and presented on the sample surface and collected when either the surface or the wipes were in the wet condition (400 μL spread) irrespective of the type of wipe. When both the surface and wipes were dry, a lot of RFP E. coli were still left on the surface after collection by the wipes. This fluorescence investigation showed that having wet surfaces or wipes is more important than the wipe materials. This implied that RFP E. coli preferred water molecules to solid surfaces and moved to the wet parts in the surface or wipes by hydrophilic interactions and not by electrostatic interactions. In general, the surfaces were not wet; thus, we decided to use wet wipes when collecting pathogens from the surfaces.
image file: c9an00896a-f2.tif
Fig. 2 Collection of bacteria/virus per composition of wipes and optimization of the desorption condition. (a) Fluorescence microscopy images of RFP bacteria (E. coli) before and after sampling on the surface (left) and wipes (right) according to various types and conditions (dry or wet) of wipes. (b) Desorption amounts of bacteria (E. coli) from the wipes with various pH buffer and ethanol concentrations. (c) Overall transfer amounts of bacteria (S. aureus)/virus (adenovirus) from the different types of wipes at pH 9 buffer containing 10% ethanol.

Step 2: Bacteria/virus transfer from wipes to aqua solution

To determine the condition for the transfer buffer to release the bacteria/virus collected from the surface to the aqua buffer solution, ethanol was added to various pH buffer solutions. This is because ethanol is normally used to remove non-specifically bound impurities in the washing step of DNA purification kits. To calculate the desorption efficiency from the wipes to the buffer solution, 200 μL of the solution containing E. coli (105 CFU mL−1) was directly loaded in the wipes and dried for 30 min. The desorption efficiency was defined by 100× the amount in the B1 buffer solution/the amount directly loaded onto the wipes. As shown in Fig. 2b, the desorbed amount of E. coli from the wipes (PET/PE) to the buffer solution is enhanced as the pH of the buffer solution increases. It is believed that wipes made of organic synthetic polymers are negatively charged depending on the pH of the solution.33,34 The effect of ethanol on E. coli desorption from the wipes also increased. This increase was likely due to the fact that ethanol could help the buffer solution to penetrate into porous wipes more easily by lowering the surface tension of the solution. Therefore, pH 9 with 10% ethanol was chosen as the optimal B1 buffer. The determined B1 buffer was confirmed with adenovirus and S. aureus in terms of the overall transfer efficiency (from the surface to the buffer solution via wipes; 100× pathogen amount in the solution/the amount loaded on the surface). As shown in Fig. 2c, the wipes 2 (PET/PE) exhibit the best transfer efficiency compared with the other wipes when 200 μL of 105 CFU mL−1 of S. aureus and 105 PFU mL−1 of adenovirus is spread on an area of 40 × 40 cm2. It is considered that the PET and PE materials have more negative charges on their surfaces at pH 9 than Rayon and PP.33 As a result, wet wipes 2 (PET/PE) were selected as the wipe materials.

Step 3: Bacteria and virus concentration on multifunctional microbeads by adsorption

The adsorption of bacteria/virus on a solid surface required them to be concentrated from 20 mL B1 buffer because it is hard to handle the chemicals and energy-intensive 20 mL volume treatments (for lysis and purification). The adsorption efficiencies (%, adsorbed bacteria or virus/input bacteria or virus) at various pH values of the buffer solution and different kinds of surfaces are shown in Table 1. Three different materials (GO, TEOS, and APTES) were selected as the functional surfaces to concentrate the bacteria/virus through physico-chemical adsorption.
Table 1 Bacteria/virus adsorption efficiency (%) on various bead surfaces
Bead surface Staphylococcus aureus Adenovirus
pH 5 pH 7 pH 9 pH 5 pH 7 pH 9
GO 92.84 ± 7.78% 89.01 ± 7.92% 84.66 ± 7.66% 91.57 ± 3.09% 83.44 ± 11.27% 14.50 ± 10.85%
TEOS 82.23 ± 4.91% 80.25 ± 6.89% 48.51 ± 13.65% 89.70 ± 5.88% 52.51 ± 7.65% 9.68 ± 3.37%
APTES 0.14 ± 0.04% 0.13 ± 0.00% 0.10 ± 0.02% 0.06 ± 0.05% 0.06 ± 0.00% 0.03 ± 0.00%


S. aureus was highly adsorbed on the GO surface in a wide range of pH values, whereas the adsorption efficiency of adenovirus on GO was dramatically reduced at pH 9. For the TEOS surface, the effect of pH on the adsorption efficiency of both the bacteria and the virus increased. It is assumed that the interaction between the surface and the bacteria/virus that induces adsorption depends on the type of the surface and the type of the bacteria/virus. For the GO surface, functional residues such as epoxy or hydroxyl groups on the basal plane and carboxyl groups at the edges could help the bacteria and virus to be more readily adsorbed on the surface compared to only hydroxyl group-containing TEOS.30,35 Interestingly, S. aureus and adenovirus did not seem to be adsorbed on the APTES surface. It was shown elsewhere that bacteria are negatively charged in a broad pH range and bind well to positively charged surfaces such as APTES.35 We assume the reason for S. aureus appearing not to be adsorbed on the surface of APTES is that the negatively charged DNA extracted from S. aureus adsorbed on the APTES surface strongly binds to the positively charged APTES, resulting in DNA not being eluted into the buffer solution.

This was confirmed by the fluorescence images of S. aureus adsorbed on various surfaces. As shown in Fig. 3, S. aureus appears to be highly adsorbed on the surface of APTES, GO and TEOS. Obviously, the brightness of the fluorescence from S. aureus (live-dead staining) on the GO surface was higher than that on others. As a result, GO was chosen as the functional material to concentrate the bacteria and virus through adsorption. As shown in Table 2, the adsorption % of both bacteria and virus on the GO surface at pH 5 and 7 decreases as the flow rate increases. However, over 70% of S. aureus in a 20 mL buffer (pH 5) solution was adsorbed at 20 mL min−1, whereas only ∼40% adenovirus in pH 5 solution was adsorbed on the GO beads at 20 mL min−1. To achieve the maximum performance, a lower flow rate and pH (92.84 ± 7.78% of S. aureus and 91.57 ± 3.09% of adenovirus at 1 mL min−1 at pH 5) are required. Because S. aureus (104 CFU mL−1) and adenovirus (104 PFU mL−1) were present in 20 mL of the B1 buffer solution (pH 9, 10 mM), 1 mL of the B2 buffer (pH 5, 1 M acetate) was added to the 20 mL of the B1 buffer, resulting in a 50 mM acetate buffer solution at pH 5.


image file: c9an00896a-f3.tif
Fig. 3 Fluorescence microscopy images of the bacteria (S. aureus) adsorbed on the various bead surfaces: (a) GO beads, (b) TEOS beads, and (c) APTES beads. The inset images are the bead surfaces after passing through a solution without bacteria (the scale bar is 100 μm).
Table 2 Bacteria/virus adsorption amounts (%) on the GO bead surface at various flow rates
  Staphylococcus aureus Adenovirus
1 mL min−1 5 mL min−1 10 mL min−1 20 mL min−1 1 mL min−1 5 mL min−1 10 mL min−1 20 mL min−1
pH 5 92.84 ± 7.78% 90.62 ± 7.84% 99.02 ± 1.22% 73.34 ± 13.76% 91.57 ± 3.09% 66.40 ± 10.96% 56.00 ± 4.48% 43.92 ± 0.81%
pH 7 89.01 ± 7.92% 89.41 ± 4.66% 74.58 ± 7.71% 64.97 ± 8.93% 83.44 ± 11.27% 70.01 ± 10.20% 38.68 ± 2.74% 39.06 ± 10.64%


Steps 4 & 5: Bacteria and virus lysis through glass bead-beating and nucleic acid elution

The efficiency of DNA extraction was first confirmed prior to the investigation of the adsorption of the bacteria and virus on microbeads. This is because the efficiencies of adsorption and lysis are calculated by quantitative DNA amplification (real-time PCR). The normalized lysis efficiency was based on the Qiagen commercial kit, as shown in Fig. 4. The vibrational bead-beating method seemed to be more effective for the lysed S. aureus (Gram-positive bacteria) than the guanidine-based Qiagen commercial kit, whereas similar results were achieved using both the vibrational bead-beating method (50 Hz, 3 min) and Qiagen kit for adenovirus. This result agreed well with previous results with reference to various bead-beating tools.31 The stirring bead-beating method (25 Hz, 1 min) using a magnetic bar in the cartridge was also compared with the normal vibrational method and yielded a 90% performance over vibrational bead-beating. This implies that stirring bead-beating provides weaker bead movements than vibrational tools even though baffles were included to increase the bead movement in the chamber. However, the stirring bead-beating tools were still selected because the vibrational bead-beating tools may induce whole cartridge weakening by vibrations.
image file: c9an00896a-f4.tif
Fig. 4 Efficiency of bacteria/virus lysis bead-beating using an electromagnet.

Step 6: Direct connection to the nucleic acid amplification step

The ready-to-use target sample prepared using the semi-automatic sample preparation cartridge was finally collected in a commercial microtube. This target sample was directly applied to the amplification step such as PCR, RT-PCR, or LAMP. Fig. 5 shows the real sample preparation cartridge including the microtube. A total of 105 CFU of S. aureus was spread on a 40 × 40 cm2 surface and collected using wet wipes. After the sample preparation procedures using the cartridge, shown in Fig. 5a, the ready-to-use target sample containing DNA/RNA was collected in the microtube (Fig. 5b). The LAMP pre-mixture could be placed in the microtube in advance, as shown in Fig. 5c. After a simple LAMP reaction at 65 °C for 30 min, the presence of S. aureus was simply confirmed by the naked eye, as shown in Fig. 5d.
image file: c9an00896a-f5.tif
Fig. 5 PCR-compatible surface sampling module using wipes: (a) photograph of 3D-printed cartridge for sample preparation, (b) image of microtube for PCR, (c) mixture of isothermal amplification (colorimetric LAMP mixture) for detecting S. aureus, and (d) result of LAMP of S. aureus after the sample preparation process.

The total recovery rate from the surface (40 × 40 cm2) to the elution buffer (200 μL) was investigated using a typical artificial surface sample (a 200 μL solution containing 104 CFU mL−1S. typhimurium, 104 CFU mL−1S. aureus, and 10 4 PFU mL−1 adenovirus was spread on the surface and dried for 40 min). As shown in Fig. 6, the averaged total recovery rate is 67.4%.


image file: c9an00896a-f6.tif
Fig. 6 Total recovery rate from the surface to the elution buffer using semi-automatic sample preparation tools (pathogen collection-transfer-DNA extraction).

A comparison of this wipe sample collection/its preparation cartridge system and regular commercial swab sampling/preparation kit was performed (Fig. 7) via real-time PCR methods. The detection performance of the sample collection/preparation method using wipes and cartridge was much better than that of the swab method (with a low concentration of S. aureus on the surface, 28 Cp (equivalent to Ct) when wipes were used and 36 Cp when a swab was used). This implies that these sample preparation tools using wipes are more sensitive than swabs.


image file: c9an00896a-f7.tif
Fig. 7 PCR-compatible surface sampling tool using wipes; Cp value of S. aureus compared to commercial kit (swab protocol of Qiagen). The sample area was 40 cm × 40 cm.

Field test

We investigated the biological contamination of a lavatory to verify the performance of this cartridge in preparing ready-to-use targets from wipes. Three public lavatories were chosen randomly, and their toilet seats, toilet lids, and lavatory walls were selected as the sample-collecting points, as shown in Fig. S4. The results in terms of total bacteria, S. typhimurium, E. coli, S. aureus, B. cereus, and adenovirus were compared to the results using a normal commercial swab kit. As shown in Table 3, the sample-collecting tools using wipes are much more effective and provide more sensitive results to detect pathogens on the surfaces than the commercial swab kit. For example, in the investigation of the toilet lid in lavatory 2, the wipe sample-collecting method could provide detailed results about individual bacterial and viral loads, whereas only the total bacterial load could be estimated from the PCR result when the swab protocol was used. From the detailed result from the wipes, we could determine that lavatory 2 was totally contaminated by S. typhimurium and S. aureus. Therefore, at least one order of magnitude higher sensitivity was achieved through the wipe-collecting method, which was enabled by the semi-automatic cartridge, compared to that of the normal swab method.
Table 3 Performance test on various large surfaces of lavatories
Location (area) Toilet seat (846.34 cm2) Toilet lid (1120.98 cm2) Lavatory wall (3600.00 cm2)
Lavatory number 1 2 3 1 2 3 1 2 3
Method Swab Wipes Swab Wipes Swab Wipes Swab Wipes Swab Wipes Swab Wipes Swab Wipes Swab Wipes Swab Wipes
Total bacteria 8.1 × 103 7.1 × 104 2.3 × 103 6.6 × 104 1.5 × 103 3.5 × 104 4.4 × 103 8.8 × 105 4.1 × 103 7.7 × 105 2.6 × 104 2.9 × 105 1.8 × 102 2.0 × 104 2.1 × 102 7.4 × 105 8.4 × 102 1.1 × 104
Salmonella typhimurium N.D. N.A. N.D. N.D. N.D. N.A. N.D. N.D. N.D. 1.1 × 101 1.8 × 101 5.7 × 101 N.D. N.D. N.D. 1.0 × 101 N.D. N.D.
Escherichia coli 6.9 × 100 4.8 × 103 N.A. 9.2 × 102 N.D. 3.1 × 103 N.A. 1.2 × 103 N.A. 5.6 × 102 1.3 × 101 1.2 × 104 N.A. 9.5 × 103 6.2 × 100 5.8 × 102 2.6 × 101 2.7 × 103
Staphylococcus aureus N.D. 5.7 × 101 N.D. 5.3 × 100 N.D. 4.2 × 100 N.D. 2.1 × 102 N.D. 5.1 × 100 N.D. 1.3 × 102 N.A. 3.1 × 103 N.D. 5.5 × 100 N.A. 4.1 × 100
Bacillus cereus 5.7 × 100 3.9 × 102 N.D. N.D. N.D. N.D. N.D. N.D. N.D. 4.0 × 103 N.D. 4.2 × 100 N.D. N.D. N.D. 2.3 × 103 N.A. 8.5 × 102
Adenovirus N.D. 1.6 × 102 N.A. 5.5 × 101 N.A. 1.1 × 102 N.A. 5.1 × 101 N.A. 1.1 × 102 N.A. 9.7 × 102 N.A. 9.7 × 101 N.A. 8.1 × 101 N.D. 1.0 × 102


Conclusions

A simple cartridge to perform the entire sample preparation process using wipes was developed to increase the sensitivity of detecting surface contamination. The bacteria and viruses present on wide surfaces were effectively collected in 20 mL buffer solutions and their nucleic acids were concentrated to 200 μL with high recovery rates. The wipe sampling method achieved two orders of magnitude higher sensitivity compared to swab kits. Thus, this method will be very useful for the intensive investigation of super-bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) and in important spaces such as emergency rooms because it can effectively measure very low concentrations of Gram-positive bacteria on large surfaces. Moreover, it can be applied to in situ tools to measure bacteria and viruses since these tools are perfectly compatible with the isothermal amplification methods such as LAMP and NASBA if an automatic operating system is developed for the cartridge proposed in this work.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the BioNano Health-Guard Research Center funded by the Ministry of Science and ICT (MSIT) of Korea as a Global Frontier Project (Grant number H-GUARD_2018M3A6B2057261).

Notes and references

  1. K. J. Lee, W. S. Lee, A. Hwang, J. Moon, T. Kang, K. Park and J. Jeong, Analyst, 2017, 143, 332–338 RSC.
  2. A. Martinon, U. P. Cronin, J. Quealy, A. Stapleton and M. G. Wilkinson, Food Control, 2012, 24, 86–94 CrossRef CAS.
  3. J. M. Boyce, J. Hosp. Infect., 2007, 65, 50–54 CrossRef PubMed.
  4. A. S. Downey, S. M. Da Silva, N. D. Olson, J. J. Filliben and J. B. Morrow, Appl. Environ. Microbiol., 2012, 78, 5872–5881 CrossRef CAS PubMed.
  5. W. Wells, Am. J. Epidemiol., 1934, 20, 611–618 CrossRef.
  6. D. K. Milton, M. P. Fabian, B. J. Cowling, M. L. Grantham and J. J. McDevitt, PLoS Pathog., 2013, 9, e1003205 CrossRef CAS PubMed.
  7. W. G. Lindsley, F. M. Blachere, R. E. Thewlis, A. Vishnu, K. A. Davis, G. Cao, J. E. Palmer, K. E. Clark, M. A. Fisher, R. Khakoo and D. H. Beezhold, PLoS One, 2010, 5, e15100 CrossRef CAS PubMed.
  8. N. Yamaguchi, A. Ishidoshiro, Y. Yoshida, T. Saika, S. Senda and M. Nasu, J. Microbiol. Methods, 2003, 53, 405–410 CrossRef PubMed.
  9. C. Gurtner, J. Heyrman, G. Piñar, W. Lubitz, J. Swings and S. Rölleke, Int. Biodeterior. Biodegrad., 2000, 46, 229–239 CrossRef CAS.
  10. A. Dolan, M. Bartlett, B. McEntee, E. Creamer and H. Humphreys, J. Hosp. Infect., 2011, 79, 227–230 CrossRef CAS PubMed.
  11. N. Panpradist, B. J. Toley, X. Zhang, S. Byrnes, J. R. Buser, J. A. Englund and B. R. Lutz, PLoS One, 2014, 9, e105786 CrossRef PubMed.
  12. G. W. Park, D. Lee, A. Treffiletti, M. Hrsak, J. Shugart and J. Vinje, Appl. Environ. Microbiol., 2015, 81, 5987–5992 CrossRef PubMed.
  13. T. Verdier, M. Coutand, A. Bertron and C. Roques, Build. Sci., 2014, 80, 136–149 Search PubMed.
  14. J. M. Edmonds, Appl. Microbiol. Biotechnol., 2009, 84, 811–816 CrossRef CAS PubMed.
  15. M. P. Buttner, P. Cruz, L. D. Stetzenbach, A. K. Klima-Comba, V. L. Stevens and P. A. Emanuel, Appl. Environ. Microbiol., 2004, 70, 7040–7045 CrossRef CAS PubMed.
  16. S. Kane, S. R. Shah, S. E. Létant, G. A. Murphy, T. M. Alfaro, J. R. Avila, E. Salazar, M. Mullins and T. Nichols, J. Bioterrorism Biodef., 2013, S3, 016 Search PubMed.
  17. J. K. Lutz, J. Crawford, A. E. Hoet, J. R. Wilkins 3rd and J. Lee, J. Appl. Microbiol., 2013, 115, 171–178 CrossRef CAS PubMed.
  18. L. Ramm, H. Siani, R. Wesgate and J. Y. Maillard, Am. J. Infect. Control, 2015, 43, 724–728 CrossRef CAS PubMed.
  19. A. Dural-Erem, P. Wessman, U. Husmark and V. Nierstrasz, Text. Res. J., 2017, 89, 216–222 CrossRef.
  20. S. Julich, H. Hotzel, C. Gartner, D. Trouchet, M. Fawzy El Metwaly Ahmed, N. Kemper and H. Tomaso, Biologicals, 2016, 44, 574–580 CrossRef CAS PubMed.
  21. S.-J. Gao and P. S. Moore, Emerging Infect. Dis., 1996, 2, 159 CrossRef CAS PubMed.
  22. F. Lou, P. Huang, H. Neetoo, J. B. Gurtler, B. A. Niemira, H. Chen, X. Jiang and J. Li, Appl. Environ. Microbiol., 2012, 78, 5320–5327 CrossRef CAS PubMed.
  23. S.-J. Park, J. Min, Q.-Z. Hu and C.-H. Jang, Colloid Polym. Sci., 2014, 292, 1163–1169 CrossRef CAS.
  24. C. P. Gerba, in Advances in applied microbiology, Elsevier, 1984, vol. 30, pp. 133–168 Search PubMed.
  25. L. Shen, Z. Jin, D. Wang, Y. Wang and Y. Lu, Chemosphere, 2018, 190, 201–210 CrossRef CAS PubMed.
  26. J. Chen, B. Yao, C. Li and G. Shi, Carbon, 2013, 64, 225–229 CrossRef CAS.
  27. C. Baek, H. Y. Kim, D. Na and J. Min, Sens. Actuators, B, 2015, 208, 238–244 CrossRef CAS.
  28. K.-Y. Hwang, S. H. Kwon, S.-O. Jung, H.-K. Lim, W.-J. Jung, C.-S. Park, J.-H. Kim, K.-Y. Suh and N. Huh, Lab Chip, 2011, 11, 3649–3655 RSC.
  29. F. N. Mutua, P. Lin, J. K. Koech and Y. Wang, Mater. Sci. Appl., 2012, 03, 856–860 CAS.
  30. C. Baek, S. H. Chung, J. Kim, S. Yoon and J. Min, J. Nanosci. Nanotechnol., 2016, 16, 11887–11891 CrossRef CAS.
  31. C. Baek and J. Min, BioChip J., 2015, 9, 332–338 CrossRef CAS.
  32. Z. Xu, L. Li, J. Chu, B. M. Peters, M. L. Harris, B. Li, L. Shi and M. E. Shirtliff, Food Res. Int., 2012, 47, 166–173 CrossRef CAS PubMed.
  33. A. M. Grancaric, A. Tarbuk and T. Pusic, Color. Technol., 2005, 121, 221–227 CAS.
  34. F. S. Lameiras, A. L. d. Souza, V. A. R. d. Melo, E. H. M. Nunes and I. D. Braga, Mater. Res., 2008, 11, 217–219 CrossRef CAS.
  35. J. Y. Won, J. Min and J. H. Park, Biosens. Bioelectron., 2010, 26, 1763–1767 CrossRef CAS PubMed.

Footnotes

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9an00896a
Both authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2019