Journal of Chromatography B, 991 (2015) 59–67
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Cationic polyelectrolyte functionalized magnetic particles assisted highly sensitive pathogens detection in combination with polymerase chain reaction and capillary electrophoresis Jia Chen, Yuexin Lin, Yu Wang, Li Jia ∗ MOE Key Laboratory of Laser Life Science & Institute of Laser Life Science, College of Biophotonics, South China Normal University, Guangzhou 510631, China
a r t i c l e
i n f o
Article history: Received 30 July 2014 Accepted 3 April 2015 Available online 9 April 2015 Keywords: Poly(diallyl dimethylammonium chloride) Magnetic particles Polymerase chain reaction Capillary electrophoresis Detection Pathogenic bacteria
a b s t r a c t Pathogenic bacteria cause signiﬁcant morbidity and mortality to humans. There is a pressing need to establish a simple and reliable method to detect them. Herein, we show that magnetic particles (MPs) can be functionalized by poly(diallyl dimethylammonium chloride) (PDDA), and the particles (PDDAMPs) can be utilized as adsorbents for capture of pathogenic bacteria from aqueous solution based on electrostatic interaction. The as-prepared PDDA-MPs were characterized by Fourier-transform infrared spectroscopy, zeta potential, vibrating sample magnetometry, X-ray diffraction spectrometry, scanning electron microscopy, and transmission electron microscopy. The adsorption equilibrium time can be achieved in 3 min. According to the Langmuir adsorption isotherm, the maximum adsorption capacities for E. coli O157:H7 (Gram-negative bacteria) and L. monocytogenes (Gram-positive bacteria) were calculated to be 1.8 × 109 and 3.1 × 109 cfu mg−1 , respectively. The bacteria in spiked mineral water (1000 mL) can be completely captured when applying 50 mg of PDDA-MPs and an adsorption time of 5 min. In addition, PDDA-MPs-based magnetic separation method in combination with polymerase chain reaction and capillary electrophoresis allows for rapid detection of 101 cfu mL−1 bacteria. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Bacteria are widely present in our daily lives. Some of them are harmony with people, while some are harmful to people’s health, which are called pathogenic bacteria. Escherichia coli O157:H7 (E. coli O157:H7), a Gram-negative pathogen, has been reported to cause haemorrhagic colitis, haemolytic uraemic syndrome, and other illnesses . Listeria monocytogenes (L. monocytogenes), a Gram-positive pathogen, is associated with some serious invasive diseases like septicemia, meningitis, and meningoencephalitis . Because of the hazards of pathogens to humans, exposure of humans to ready to eat foods and drinks contaminated by these pathogens should be prevented. The limits for these pathogens have been set to be less than 102 cfu g−1 in foods and drinks in the European Union. The strict criteria have increased the need for rapid and highly sensitive detection methods for bacteria. The conventional plate-counting method is an effective way to get an accuracy result for bacteria detection. However, the method is usually time-consuming and labor-intensive. In order
∗ Corresponding author. Tel.: +86 20 85211543; fax: +86 20 85216052. E-mail addresses: [email protected]
, jiali [email protected]
(L. Jia). http://dx.doi.org/10.1016/j.jchromb.2015.04.002 1570-0232/© 2015 Elsevier B.V. All rights reserved.
to reduce the detection time for bacteria, some other methods have been developed, such as biosensor-based [3,4] and polymerase chain reaction (PCR) based methods [5–7]. However, the detection limits for bacteria of these methods are usually in the range of 103 –104 cfu mL−1 . Although real-time PCR-based method can effectively improve the detection sensitivity of bacteria with a detection limit lower than 102 cfu g−1 , the time-consuming culture enrichment was still a necessary step for high sensitivity, which limits the rapid detection of low concentration of bacteria. Therefore, development of new rapid and sensitive detection methods for bacteria is still challenging. Magnetite (Fe3 O4 ) particles possessing appealing magnetic properties, nontoxicity, easy preparation, and easy surfacefunctionalization have gained attractive attention in a variety of applications especially in biomedicine [9,10] and environmentology [11–13]. The magnetic responsiveness of functional magnetic particles (MPs) enables them to be conveniently separated from aqueous solutions by applying an external magnetic ﬁeld. Some functional MPs have been applied for capture of pollutants in water, such as antibiotics [14,15] and heavy metal ions [16–19]. MPs functionalized with some particular groups, such as amine , N-methylimidazolium , pigeon ovalbumin  and amino acid  have also been utilized for pathogens capture and removal.
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
Poly(diallyl dimethylammonium) chloride (PDDA) with highly hydrophilic and permanently charged quaternary ammonium groups is a water-soluble cationic polyelectrolyte. It is widely used in water treatment, paper manufacturing, mining industry, and so on . It was also used to remove negatively charged perchlorate , chromate , and arsenic  from aqueous solution based on electrostatic interactions in a polyelectrolyte-enhanced ultraﬁltration process. To the best of our knowledge, PDDA has not been reported for bacteria capture. The favorable attributes of MPs and PDDA inspired us to prepare PDDA functionalized MPs (PDDAMPs). And the feasibility of PDDA-MPs as adsorbents for capture of pathogens can be expected since bacteria are negatively charged above their isoelectric points. In this work, PDDA-MPs were synthesized by three reaction steps. They are characterized by Fourier-transform infrared spectrometry (FT-IR), zeta potential, vibrating sample magnetometry (VSM), X-ray diffraction spectrometry (XRD), scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The positive charges on PDDA-MPs enable them to rapidly capture negatively charged bacteria from aqueous solution based on electrostatic interaction. And the magnetic property of PDDA-MPs enables them to be conveniently separated from aqueous solution. The adsorption isotherms of PDDA-MPs for pathogenic bacteria (E. coli O157:H7 and L. monocytogenes) were studied. Furthermore, PDDA-MPs were used for highly sensitive detection of bacteria in mineral water in combination with PCR and capillary electrophoresis (CE). 2. Experimental 2.1. Chemicals and reagents PDDA, 20% in water, with an average molecular weight between 400,000 and 500,000 Da, was purchased from Sigma–Aldrich, Co. (MO, USA). Ferric chloride hexahydrate (FeCl3 ·6H2 O), ammonium hydroxide (NH3 ·H2 O, [25%, w/w]), tris (hydroxymethyl) amino methane (Tris), boric acid (H3 BO3 ), sodium chloride (NaCl) and sodium hydroxide (NaOH) were purchased from Guangzhou Chemical Regent Factory (Guangzhou, China). Isopropanol was acquired from SK Chemicals (Ulsan, South Korea), Tetraethyl orthosilicate (TEOS) was acquired from Acros Organic (New Jersey, USA). Sodium acetate anhydrous (NaAc) was purchased from Tianjin Guangcheng Chemical Reagent Company (Tianjin, China). Ethylene glycol was purchased from Tianjin Damao Chemical Reagent Factory (Tianjin, China). All these reagents were analytical grade or better. Hydroxypropylmethylcellulose (HPMC), the viscosity in 2% aqueous solution at 25 ◦ C was 100–4000 mPa s, was purchased from Shanghai Sangon Biological Engineering & Technology (Shanghai, China). All aqueous solutions were prepared using water obtained from an Elga water puriﬁcation system (ELGA, London, UK). A GeneRuler 100-bp DNA ladder was purchased from Fermentas (Beijing, China) and consisted of 10 fragments ranging from 100 to 1000 bp with a total concentration of 0.5 mg mL−1 . Taq DNA polymerase (5 U L−1 ), 25 mM MgCl2 , 10 × PCR buffer (100 mM Tris-HCl [pH 8.3], 500 mM KCl), and deoxynucleotide triphosphate (dNTP) mixture (including dATP, dGTP, dCTP, and dTTP, where the concentration of each dNTP was 2.5 mM) were purchased from Takara Biotechnology (Dalian, China). The primers (Table 1) used for PCR were synthesized by Shanghai Sangon Biological Engineering & Technology Services (Shanghai, China). TIANamp bacteria DNA kit was acquired from TianGen Biotech (Beijing, China). 2.2. Preparation of PDDA-MPs Three steps are involved in the preparation of PDDA-MPs, as shown in Fig. 1(a). Firstly, magnetic Fe3 O4 particles were prepared
using the solvothermal reduction method  with minor modiﬁcations. Brieﬂy, 1.35 g FeCl3 ·6H2 O was dissolved in 40 mL ethylene glycol to form a clear solution. Then 3.6 g NaAc was added to the solution and the mixture was stirred vigorously for 30 min. The solution was introduced to a Teﬂon-lined stainless-steel autoclave with 50 mL capacity and heated at 200 ◦ C for 10 h. The resulting black products were cooled to room temperature and rinsed several times with ethanol. Finally, the products were dried at 60 ◦ C under nitrogen atmosphere for 6 h. Secondly, the magnetic Fe3 O4 particles were coated with silica according to the sol–gel method . The as-prepared magnetic Fe3 O4 particles (approximate 0.3 g) were dispersed in 40 mL isopropanol with the aid of ultrasonication. Then, 2 mL deionized water and 0.5 mL NH3 ·H2 O (25%) were added to the suspension. Afterwards, 0.3 mL TEOS was added to the mixture dropwise and the reaction was stirred for 8 h at room temperature. Finally, the particles were rinsed several times with ethanol and dried under nitrogen atmosphere at 60 ◦ C. The magnetic silica particles were soaked in 1 M HCl for 24 h at 4 ◦ C, then dried under nitrogen atmosphere at 60 ◦ C. The silica shell can isolate the inner Fe3 O4 from outer environment, protect the Fe3 O4 from being oxidized and provide –OH as the reaction moieties after activating using HCl solution. Thirdly, the magnetic silica particles were functionalized with PDDA. The PDDA coating solution was prepared by dissolving PDDA at 0.2% in water, whose ionic strength was adjusted to 1.5 M by addition of NaCl. The magnetic silica particles (approximate 0.3 g) were suspended in 20 mL 1 M NaOH and shaken for 15 min, then rinsed several times with deionized water. Afterwards, the particles were added to 20 mL 0.2% PDDA solution and shaken for 15 min. Finally, the products were collected by applying an external magnetic ﬁeld and dried under nitrogen atmosphere at 60 ◦ C. 2.3. Characterization The FT-IR spectra of PDDA, SiO2 @Fe3 O4 , and PDDA-MPs were recorded on a Bruker FT-IR spectrometer (Bruker, Germany). The zeta potentials of SiO2 @Fe3 O4 and PDDA-MPs at different pH values were measured on a Zetasizer Nano ZS (Malvern, Worcestershire, UK). The magnetization characterization of Fe3 O4 , SiO2 @Fe3 O4 and PDDA-MPs were performed on a vibrating sample magnetometer (Quantum Design, San Diego, USA) at room temperature and an applied ﬁeld of 10 kOe. The XRD measurements of Fe3 O4 , SiO2 @Fe3 O4 and PDDA-MPs were recorded on a Bruker D8 Advance (Bruker AXS, Germany) using Cu K␣ radiation with scattering angles (2) of 20–80◦ and a counting time of 10 s per increment. The TEM images of PDDA-MPs and the conjugates between PDDAMPs and bacteria were carried out on a JEM-2100HR transmission electron microscope (JEOL, Tokyo, Japan). SEM images of PDDAMPs were carried out on a ZEISS Ultra 55 ﬁeld emission scanning electron microscope (Carl Zeiss, Oberkochen, Germany). 2.4. Preparation of bacteria samples Gram-negative bacteria (E. coli O157:H7) and Gram-positive bacteria (L. monocytogenes) were chosen as the model pathogenic bacteria. E. coli O157:H7 strain was grown at 37 ◦ C for 8 h on a rotary shaker at 200 rpm in Luria-Bertani broth (tryptone 10 g L−1 , yeast extract 5 g L−1 and NaCl 10 g L−1 , pH 7.5). L. monocytogenes (CMCC54007) was grown at 37 ◦ C for 10 h on a rotary shaker at 200 rpm in nutrient broth (3 g L−1 beef extract, 5 g L−1 bacterial peptone and 5 g L−1 NaCl, pH 7.0). Before use, the bacterial cells were all centrifuged at 6000 rpm for 5 min and isolated from the broth. Then the isolated bacterial cells were washed with Tris-borate buffer (50 mM, pH 7.0) three times. Different concentrations of bacteria were adjusted
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
Table 1 Sequences of primers used in PCR reactions. Bacteria
Sequence (5 -3 )
Upstream primer Downstream primer Upstream primer Downstream primer
GGGAAATCTGTCTCAGGTGATGT CGATGATTTGAACTTCATCTTTTGC CAGTTTACCAACCGTCAT GAGCAACCGTTCCATTAC
E. coli O157:H7
using the Tris-borate buffer and their optical densities at 600 nm (OD600) were measured on a micro-spectrophotometer (K5600, Beijing Kaiao Technology Development Company, Beijing, China). The bacteria concentration recorded in colony forming units (cfu) was obtained based on the plate-counting method. For safety consideration, all containers in contact with bacteria were sterilized in a high-pressure steam sterilization pot (Tomy SX500, Tomy Seiko Co., ltd., Tokyo, Japan) at 121 ◦ C for 30 min before and after use. The bacteria samples were also sterilized to kill the bacteria after use. 2.5. Bacteria capture experiments First of all, the concentration of bacteria was adjusted to a desired level (OD600 , 0.4) using Tris-borate buffer (50 mM, pH 7.0). Then 5 mg PDDA-MPs were added to the solution and the solution volume was ﬁxed to 10 mL. The solution was incubated by shaking vigorously for 3 min. Afterwards an external magnet was employed for magnetic separation. The supernatant was then carefully pipetted into a cell to measure its OD600 . The capture efﬁciency (%) was calculated based on Eq. (1). Capture efficiency = 100(C0 − Ce )/C0
where C0 and Ce are the initial and equilibrium concentrations of bacteria expressed as the OD600 value, respectively.
The adsorbed amount of bacteria by PDDA-MPs was calculated using Eq. (2). qe = (Co − Ce )V/m
where qe is the amount of bacteria adsorbed at equilibrium. C0 and Ce are the initial and equilibrium concentration of bacteria, respectively, V (mL) is the initial volume of bacteria solution, and m (mg) is the mass of PDDA-MPs used. 2.6. Genomic DNA extraction and PCR ampliﬁcation The bacterial DNA was isolated using the protocol and reagents from the TIANamp Bacteria DNA Kit and ﬁnally suspended in 100 L of TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.3). The reagents for multiplex PCR included DNA template (5 L), 2 × PCR buffer, 1.5 mM MgCl2 , 0.2 mM each dNTP, 0.4 mM each primer, 0.5 mg mL−1 bovine serum albumin, and 0.2 U L−1 Taq DNA polymerase. The total reaction volume was 25 L. The control solution (blank) contained all the PCR reagents except for the DNA template. The primers used for PCR, and amplicon sizes are listed in Table 1. The target genes chosen in the study were the virulence gene hlyA (106 bp) for L. monocytogenes and the rfbE (168 bp) for E. coli O157:H7 . The PCR ampliﬁcation was performed on a thermocycler (A200, Longgene Scientiﬁc Instrument Company, Hangzhou, China) and programmed with an initial step of denaturation at 94 ◦ C for 3 min.
Fig. 1. Sketches of the process for preparation of PDDA-MPs (a) and bacteria capture and elution (b).
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
The cycling conditions were as follows: denaturation at 94 ◦ C for 30 s, annealing at 56 ◦ C for 30 s, and elongation at 72 ◦ C for 30 s. In total, 40 cycles of the above program were performed. The last round of elongation was for 3 min. The PCR products were analyzed by capillary electrophoresis (CE). 2.7. CE conditions The laboratory-built CE apparatus used in this study consisted of a TriSep-2100 high-voltage power supply and an UV detector (Unimicro Technologies, CA, USA). Data acquisition and analysis were performed with the software HW-2000 workstation (Qianpu Software, Shanghai, China). Separations were carried out in a fused-silica capillary (70 cm [effective length 40 cm] × 75 m i.d.) (Hebei Yongnian Ruipu Chromatogram Equipment Company, Handan, China) at room temperature, which was kept at 25 ◦ C using air conditioning system. The background solution consisted of 1% (w/v) HPMC, 2 mM EDTA, and 20 mM phosphoric acid, whose pH was adjusted to 7.3 using Tris solution. Samples were injected electrokinetically by 10 kV for 10 s at the cathodic end and separated by applying −10 kV. The detection wavelength was set at 260 nm. All solutions used in electrophoresis experiments were sonicated prior to use. Between runs, the capillary was rinsed using background buffer. Before ﬁrst use, a new uncoated capillary was preconditioned by rinsing with 1.0 M HCl (60 min), followed by deionized H2 O (30 min), 1.0 M NaOH (60 min), deionized H2 O (30 min), and ﬁnally with the background solution (60 min) at a ﬂow rate of 20 L min−1 . 2.8. Detection of bacteria in real samples To verify the feasibility of PDDA-MPs in detection of bacteria in real samples, a mineral water sample was used and purchased from a local market. The entire process of the capture and elution of bacteria with PDDA-MPs is shown in Fig. 1(b). Different levels of bacteria were spiked into the mineral water samples whose volume was 1000 mL. Then PDDA-MPs (50 mg) were respectively added to the water samples. The mixed water samples were incubated for 5 min under vigorously shaking in order to capture the bacteria effectively. The particle-bacteria conjugates were magnetically separated with an external magnet and the supernatant was carefully pipetted into a cell to measure the bacteria concentration based on the platecounting method. The adsorption efﬁciencies can be calculated by comparing the bacteria concentration in solution before and after magnetic separation. After capturing the bacteria from 1000 mL spiked mineral water sample, desorption was carried out by washing out PDDA-MPsbacteria conjugates with 10 mL Tris-borate buffer (50 mM, pH 7.0) containing 1.0 M NaCl. The tube was shaken vigorously for 30 min to desorb the bacteria from PDDA-MPs. An external magnet was employed for magnetic separation. And the eluted bacteria were transferred to a 1.5 mL centrifuge tube. Then, the genomic DNA was extracted according to the protocol mentioned above, and the obtained DNA was directly used as the template for PCR ampliﬁcation. Afterwards, the PCR product was analyzed by CE.
Fig. 2. FT-IR spectra of PDDA, SiO2 @Fe3 O4 , and PDDA-MPs.
vibration, and the peaks at ∼1100 cm−1 resulted from the Si-O-Si stretching vibration. It can be clearly seen that the characteristic peaks of PDDA and SiO2 @Fe3 O4 are all observed in the FT-IR spectra of PDDA-MPs, which suggested that PDDA were successfully coated on the surface of SiO2 @Fe3 O4 particles. Fig. 3 shows the zeta potentials of SiO2 @Fe3 O4 and PDDA-MPs at different pH values. The zeta potential of SiO2 @Fe3 O4 was negative and decreased with the increase of the pH values from 2 to 11 due to the dissociation of Si-OH on the surface of SiO2 @Fe3 O4 . The zeta potential of PDDA-MPs was positive and maintained at about +40 mV in the pH range from 2 to 9 due to the existence of PDDA on the surface of PDDA-MPs. Beyond the pH range, the zeta potential of PDDA-MPs was negative, indicating that PDDA-MPs may had a poor stability in strong alkaline solution. The great change in zeta potential of the two particles in the pH range of 2 to 9 directly conﬁrmed that PDDA was coated on SiO2 @Fe3 O4 . Magnetic characterizations of Fe3 O4 , SiO2 @Fe3 O4 , and PDDAMPs at 300 K were measured. Fig. 4 shows the magnetic hysteresis curves of these particles. The saturation magnetization values of Fe3 O4 , SiO2 @Fe3 O4 , and PDDA-MPs were 83.4, 70.7, 68.3 emu g−1 , respectively. The decrease in saturation magnetization value of PDDA-MPs is another powerful proof that PDDA were indeed coated on SiO2 @Fe3 O4 . Ma et al. have reported that a saturation magnetization of 16.3 emu g−1 is sufﬁcient for magnetic separation from solution with a magnet . The experimental results showed that the PDDA-MPs possessed an excellent magnetic response. Fig. 5 shows the XRD diffraction patterns of Fe3 O4 , SiO2 @Fe3 O4 , and PDDA-MPs. Five characteristic diffraction peaks of Fe3 O4 at
3. Results and discussion 3.1. Characterization Fig. 2 shows the FT-IR spectra of SiO2 @Fe3 O4 , PDDA-MPs and PDDA. In the FT-IR spectra of PDDA, the adsorption peaks in the region of 2867–3017 cm−1 originated from the C-H stretching vibrations of PDDA units. In the FT-IR spectra of SiO2 @Fe3 O4 and PDDA-MPs, the peaks at ∼595 cm−1 were related to the Fe-O-Fe
Fig. 3. Zeta potentials of SiO2 @Fe3 O4 and PDDA-MPs at different pH values.
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
2 = 30.39◦ , 35.71◦ , 43.29◦ , 57.20◦ , and 62.81◦ were observed. In the diffraction patterns of these particles, no obvious difference was observed, indicating that the coated PDDA is amorphous and the crystal phase of Fe3 O4 particles were not affected during coating process. The particle size and morphology information of PDDA-MPs were studied by TEM and SEM images, as shown in Fig. 6a and b. The images indicated that the prepared PDDA-MPs are spherical particles with obvious structure and a mean diameter about 200 nm. Fig. 6c shows the TEM images of PDDA-MPs-bacteria conjugates, which clearly conﬁrm the effective capture of PDDA-MPs for bacteria. 3.2. Bacteria capture experiment
Fig. 4. VSM spectra of Fe3 O4 , SiO2 @Fe3 O4 , and PDDA-MPs.
Fig. 5. XRD spectra of Fe3 O4 , SiO2 @Fe3 O4 , and PDDA-MPs.
3.2.1. Factors affecting the capture of bacteria by PDDA-MPs To evaluate the performance of PDDA-MPs for the capture of bacteria, the factors including contact time, sample pH and volume were studied. All experiments were carried out using 5 mg PDDAMPs as adsorbents and 10 mL Tris-borate buffer (50 mM) spiked with a certain concentration of bacteria as the sample solution. Firstly, the effect of contact time on the adsorbed amount of bacteria were conducted by increasing the shaking time from 1 to 30 min while keeping the bacteria concentration at a certain level (OD600 , 0.4) in Tris-borate buffer (50 mM, pH 7.0, 10 mL). The capture efﬁciency was calculated from the amount of the adsorbed bacteria relative to that of the added bacteria. As shown in Fig. 7, the capture efﬁciencies of E. coli O157:H7 and L. monocytogenes increased with the increase of contact time from 1 to 3 min and reached a maximum at 3 min. Beyond 3 min, the capture efﬁciencies had a drop with the increase of contact time, which is possibly due to the death of bacteria because some chemical groups on PDDAMPs have an antimicrobial activity after a long period of contact with bacteria . Therefore, 3 min was chosen for the following experiments. Secondly, the inﬂuence of sample pH ranging from 3.0 to 9.0 on the capture efﬁciency of bacteria was investigated. As shown in Fig. 8, in the pH range from 6.0 to 7.0, the capture efﬁciencies of PDDA-MPs for the two bacteria were largest. In the pH range of 4.0 to 7.0, little pH effect on the capture efﬁciency was observed. Beyond pH 7.0, the capture efﬁciencies reduced with the increase of pH.
Fig. 6. TEM (a), SEM (b) images of PDDA-MPs, and TEM images of PDDA-MPs-bacteria conjugates (c).
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
Fig. 7. Effect of contact time on bacteria capture. Experimental conditions: concentration of bacteria, OD600 0.4; volume, 10 mL; PDDA-MPs, 5 mg.
The mechanism of PDDA-MPs for the capture of bacteria was discussed. PDDA-MPs are positively charged in the sample solution at pH 4.0 to 7.0, which can be conﬁrmed by the positive zeta potential of PDDA-MPs, as shown in Fig. 3. E. coli O157:H7 is a kind of Gram-negative bacteria, whose isoelectric point is about 4. L. monocytogenes is a kind of Gram-positive bacteria, whose isoelectric point is between 2 and 3. When the pH of solution is above their isoelectric points, their surfaces possess abundant negative charges due to the ionized phosphoryl and carboxylate substituents on outer cell membrane . Thus, in the pH range of 4.0 to 7.0, these negatively charged bacteria were captured by the positively charged PDDA-MPs via electrostatic interactions between them. Little pH effect on the capture efﬁciency in the pH range of 4.0 to 7.0 proved that the electrostatic interaction is the main driving force for the capture of bacteria by PDDA-MPs. Thirdly, in order to test the performance of PDDA-MPs for the capture of bacteria from environmental water samples containing different levels of bacteria, it is necessary to study the effect of sample volume on the capture efﬁciency of bacteria. The experiments were conducted by adding 50 mg PDDA-MPs to the solution containing different amounts of pathogenic bacteria (1.0 × 104 cfu for each bacterium and 1.0 × 106 cfu for each bacterium) and keeping the volume ranging from 10 to 1000 mL. As shown in Fig. 9, the
Fig. 8. Effect of pH on bacteria capture. Experimental conditions: concentration of bacteria, OD600 0.4; volume, 10 mL; contact time, 3 min; PDDA-MPs, 5 mg.
Fig. 9. Effect of sample volume on bacteria capture. Experimental conditions: concentration of bacteria, 1 × 104 cfu for each bacterium and 1 × 106 cfu for each bacterium; contact time, 5 min; PDDA-MPs, 50 mg.
capture efﬁciencies reached 100% at different sample volumes and did not change with the increase of sample volume. The results indicated that PDDA-MPs had strong interactions with pathogenic bacteria. The solution volume 1000 mL was used in the following capture experiments from real water samples. 3.2.2. Adsorption isotherms Adsorption isotherms experiments were carried out, as shown in Fig. 10. The Langmuir (Eq. (3)) and Freundlich isotherm equations (Eq. (4)) were applied to analyze the adsorption process of PDDAMPs for bacteria. qe = Ce KL qm /(1 + Ce KL ) 1/n
qe = KF Ce
where qe is the amount of bacteria adsorbed at equilibrium, qm is the maximum bacteria adsorption amount, Ce is the equilibrium concentration of bacteria, KL , KF are the adsorption constants of Langmuir and Freundlich models, respectively, and n is the Freundlich linearity index. The linearized forms of Langmuir and Freundlich equations can be represented as Eqs. (5) and (6), respectively. Ce /qe = Ce /qm + 1/(KL qm )
ln qe = ln KF + ln Ce /n
The constants KF and KL can be obtained from the plots of Ce /qe against Ce , lg qe versus lg Ce , respectively. All the parameters are presented in Table 2. It can be found that the Langmuir model ﬁts the experiment data better than Freundlich model by comparing the determination coefﬁcients. As shown in the insert plot in Fig. 10, the adsorption equilibrium data were very well represented by the Langmuir isotherm, indicating the monolayer coverage of bacteria on PDDA-MPs. And the maximum adsorption capacities of PDDAMPs for L. monocytogenes and E. coli O157:H7 were calculated to be OD600 1.76 per mg and OD600 1.59 per mg by the Langmuir model, respectively. In order to get the adsorption capacity expressed as the bacteria cell numbers per mg PDDA-MPs, the correlation between OD600 and cell numbers expressed as cfu was determined based on the plate-counting method, as shown in Fig. 11. Based on the correlation between OD600 and cell numbers, the maximum adsorption capacities of PDDA-MPs for L. monocytogenes and E. coli O157:H7 were calculated to be 3.1 × 109 and 1.8 × 109 cfu mg−1 , respectively,
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
Table 2 Langmuir and Freundlich isotherm parameters for pathogenic bacteria adsorption on PDDA-MPs. Bacteria
L. monocytogenes E. coli O157:H7
qm (mg−1 )
KL (mL mg−1 )
KF (mg−1 )
Fig. 10. Langmuir adsorption isotherms of L. monocytogenes and E. coli O157:H7 by PDDA-MPs. Experimental conditions: volume, 10 mL; contact time, 3 min; PDDAMPs, 5 mg.
which are higher than those by N-methylimidazolium modiﬁed MPs  prepared by us. This is because PDDA-MPs (+41.7 mV, pH 7.0) took more positive charges than N-methylimidazolium modiﬁed MPs (+34.1 mV pH 7.0), which can be conﬁrmed by their zeta potentials. 3.3. Detection of bacteria in real water samples In order to verify the feasibility of PDDA-MPs as adsorbents in the detection of bacteria in real samples, PDDA-MPs were applied to capture bacteria from mineral water. No bacteria were detected in the blank mineral water samples based on the plate-counting method. Then 1000 mL spiked mineral water samples containing different levels of bacteria (1 × 104 cfu for each bacterium and 1 × 106 cfu for each bacterium) were used as the test samples. PDDA-MPs (50 mg) were respectively added to the samples and shaken for 5 min, then an external magnet was employed for magnetic separation. Each supernatant was carefully pipetted and the amount of the bacteria in it was measured based on the platecounting method. The bacteria in the spiked mineral water can be completely captured. The results demonstrated that PDDA-MPs can
Fig. 11. Correlation between OD600 and bacteria cell numbers.
capture different levels of bacteria from large volumes of mineral water samples rapidly, efﬁciently and conveniently. Since the captured bacteria by PDDA-MPs was enriched from large volume of aqueous solution, PDDA-MPs would help to enhance the detection sensitivity of bacteria by combination with PCR and CE and the time-consuming culture enrichment step can be avoided prior to PCR. The spiked mineral water sample (1000 mL) containing 1 × 104 cfu for each bacterium was used as the test sample. The captured bacteria were ﬁrstly eluted from the MPs in order to extract genomic DNA. The desorption conditions including different eluents, eluent concentration, and desorption time were investigated. When 10 mL Tris-borate buffer (50 mM, pH 7.0) containing 1.0 M NaCl was used as the elution buffer, the desorption efﬁciencies for these bacteria were better than other eluents. When the particle-bacteria conjugates were mixed with the eluent and shaken vigorously for 30 min, the desorption efﬁciencies for E. coli O157:H7 and L. monocytogenes were 85% and 62%, respectively. Further prolonging the desorption time did not improve the desorption efﬁciencies. Next, the genomic DNA of the desorbed bacteria was extracted and the obtained DNA was directly used as the template for PCR ampliﬁcation. Afterwards, the PCR product was analyzed by CE.
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67
Acknowledgments We are grateful to the ﬁnancial support of the National Natural Science Foundation of China (21175048).
Fig. 12. Analysis of PCR products by CE. (a) Blank (without DNA templates); (b) PCR products; (c) Spiked PCR products; (d) GeneRuler 100-bp DNA ladder.
Fig. 12 shows the electropherograms of the standard GeneRuler 100-bp DNA ladder, the blank, the PCR products and the spiked sample. In the electropherogram of the PCR products (Fig. 12b), two peaks appeared between the peaks of 100 and 200 bp fragments, which are the target amplicon 106 bp, the ampliﬁcation product of speciﬁc virulence gene hlyA of L. monocytogenes, and the target amplicon 168 bp, the ampliﬁcation product of speciﬁc virulence gene rfbE of E. coli O157:H7, respectively . The results indicated that the bacteria with a concentration as low as 101 cfu mL−1 can be detected in less than 4.5 h using the PDDA-MPs-based magnetic separation method in combination with PCR and CE. The reusability of PDDA-MPs for adsorbing bacteria was also tested. After adsorption, desorption was carried out by washing out PDDA-MPs-bacteria conjugates with Tris-borate buffer (50 mM, pH 7.0) containing 1.0 M NaCl, and by rinsing PDDA-MPs with water. Then, PDDA-MPs were dried and reused. It was found that the adsorption capacity of PDDA-MPs for bacteria declined more than 60% when the particles were used once again. The results indicated that PDDA-MPs were not suitable for the recycling in the capture of bacteria. 4. Conclusions We describe here the preparation of structured PDDA-MPs in three reaction steps and they are used as the adsorbents to capture bacteria from real water samples rapidly and effectively. The asprepared particles combine the favorable attributes of PDDA and MPs. The high magnetic responsiveness of the particles enables them to be conveniently separated from aqueous solution by applying an external magnetic ﬁeld. The positive charges taken by PDDA on the particles enable them to rapidly and effectively capture negatively charged bacteria from aqueous solution based on electrostatic interaction, as conﬁrmed by the TEM images of MPsbacteria conjugates. Little pH effect on the capture efﬁciency in the pH range of 4.0 to 7.0 proved that electrostatic interaction is the main driving force for the capture of bacteria by PDDA-MPs. The as-prepared MPs were used as adsorbents to completely capture the bacteria in spiked mineral water samples in 5 min. In comparison with other functionalized MPs which aimed at the capture of bacteria [20–23], PDDA-MPs showed the superiority in adsorption time and efﬁciency. In addition, PDDA-MPs-based magnetic separation method can effectively reduce the detection time for low concentration of bacteria by combination with PCR and CE. The bacteria with a concentration of 101 cfu mL−1 can be detected in less than 4.5 h using the developed PDDA-MPs-PCR-CE method.
 D.D. Hancock, T.E. Besser, M.L. Kinsel, P.I. Tarr, D.H. Rice, M.G. Paros, The prevalence of Escherichia coliO157:H7 in dairy and beef cattle in Washington State, Epidemiol. Infect. 113 (1994) 199–207.  J.C. Piffaretti, H. Kressebuch, M. Aeschbacher, J. Bille, E. Bannerman, J.M. Musser, R.K. Selande, J. Rocourt, Genetic characterization of clones of the bacterium Listeria monocytogenes causing epidemic disease, Proc. Natl. Acad. Sci. USA 86 (1989) 3818–3822.  B. Serra, M. Gamella, A.J. Reviejo, J.M. Pingarrón, Lectin-modiﬁed piezoelectric biosensors for bacteria recognition and quantiﬁcation, Anal. Bioanal. Chem. 391 (2008) 1853–1860.  P. Banerjee, A.K. Bhunia, Cell-based biosensor for rapid screening of pathogens and toxins, Biosens. Bioelectron. 26 (2010) 99–106.  L. Settanni, A. Corsetti, The use of multiplex PCR to detect and differentiate foodand beverage-associated microorganisms: a review, J. Microbiol. Methods 69 (2007) 1–22.  S. Bhattacharya, S. Salamat, D. Morisette, P. Banada, D. Akin, Y.S. Liu, A.K. Bhunia, M. Ladisch, R. Bashir, PCR-based detection in a micro-fabricated platform, Lab Chip 8 (2008) 1130–1136.  B. Suo, Y. He, G. Paoli, A. Gehring, S.I. Tu, X. Shi, Development of an oligonucleotide-based microarray to detect multiple foodborne pathogens, Mol. Cell. Probes 24 (2010) 77–86.  G. Amagliani, E. Omiccioli, G. Brandi, I.J. Bruce, M. Magnani, A multiplex magnetic capture hybridisation and multiplex real-time PCR protocol for pathogen detection in seafood, Food Microbiol. 27 (2010) 580–585.  A.K. Gupta, M. Gupta, Synthesis and surface engineering of iron oxide nanoparticles for biomedical applications, Biomaterials 26 (2005) 3995–4021.  L. Liu, S. Shi, H. Zhao, J. Yu, X. Jiang, X. Chen, Selective ﬁshing and analysis of xanthine oxidase binders from two fabaceae species by coupling enzyme functionalized core–shell magnetic nanoparticles with HPLC–MS, J. Chromatogr. B 945-946 (2014) 163–170.  F. Qiao, K.H. Row, M. Wang, Water-compatible magnetic imprinted microspheres for rapid separation and determination of triazine herbicides in environmental water, J. Chromatogr. B 957 (2014) 84–89.  Y.B. Luo, Z.G. Shi, Q. Gao, Y.Q. Feng, Magnetic retrieval of graphene: Extraction of sulfonamide antibiotics from environmental water samples, J. Chromatogr. A 1218 (2011) 1353–1358.  Y. Zhou, Q. Xia, M. Ding, H. Kageruka, H. Jiang, Y. Jiang, H. Jing, X. Xiao, H. Zhong, Magnetic nanoparticles of nitrogen enriched carbon (mnNEC) for analysis of pesticides and metabolites in zebraﬁsh by gas chromatography–mass spectrometry, J. Chromatogr. B 915–916 (2013) 46–51.  Y. Lin, S. Xu, L. Jia, Fast and highly efﬁcient tetracyclines removal from environmental waters by graphene oxide functionalized magnetic particles, Chem. Eng. J. 225 (2013) 679–685.  H. Niu, D. Zhang, Z. Meng, Y. Cai, Fast deﬂuorination and removal of norﬂoxacin by alginate/[email protected]
O4 core/shell structured nanoparticles, J. Hazard. Mater. 227–228 (2012) 195–203.  J.F. Liu, Z.S. Zhao, G.B. Jiang, Coating Fe3 O4 magnetic nanoparticles with humic acid for high efﬁcient removal of heavy metals in water, Environ. Sci. Technol. 42 (2008) 6949–6954.  X. Luo, C. Wang, S. Luo, R. Dong, X. Tu, G. Zeng, Adsorption of As (III) and As (V) from water using magnetite Fe3 O4 -reduced graphite oxide–MnO2 nanocomposites, Chem. Eng. J. 187 (2012) 45–52.  Y. Ren, H.A. Abbood, F. He, H. Peng, K. Huang, Magnetic EDTA-modiﬁed chitosan/SiO2 /Fe3 O4 adsorbent: preparation, characterization, and application in heavy metal adsorption, Chem. Eng. J. 226 (2013) 300–311.  Z. Wang, D. Wu, G. Wu, N. Yang, A. Wu, Modifying Fe3 O4 microspheres with rhodamine hydrazide for selective detection and removal of Hg2+ ion in water, J. Hazard. Mater. 244–245 (2013) 621–627.  Y.F. Huang, Y.F. Wang, X.P. Yan, Amine-functionalized magnetic nanoparticles for rapid capture and removal of bacterial pathogens, Environ. Sci. Technol. 44 (2010) 7908–7913.  Y. Wang, M. Deng, L. Jia, N-methylimidazolium functionalized magnetic particles as adsorbents for rapid and efﬁcient capture of bacteria, Microchim. Acta 181 (2014) 1275–1283.  J.C. Liu, P.J. Tsai, Y.C. Lee, Y.C. Chen, Afﬁnity capture of uropathogenic Escherichia coli using pigeon ovalbumin-bound Fe3 O4 @Al2 O3 magnetic nanoparticles, Anal. Chem. 80 (2008) 5425–5432.  Y. Jin, F. Liu, C. Shan, M. Tong, Y. Hou, Efﬁcient bacterial capture with amino acid modiﬁed magnetic nanoparticles, Water Res. 50 (2014) 124–134.  C. Wandrey, J. Hernández-Barajas, D. Hunkeler, Diallyldimethylammonium chloride and its polymers, in: I. Capek, J. Hernández-Barajas, D. Hunkeler, J.L. Reddinger, J.R. Reynolds, C. Wandrey (Eds.), Radical Polymerisation Polyelectrolytes, Springer, Berlin Heidelberg, 1999, pp. 123–183.  J.D. Roach, D. Tush, Equilibrium dialysis and ultraﬁltration investigations of perchlorate removal from aqueous solution using poly(diallyldimethylammonium) chloride, Water Res. 42 (2008) 1204–1210.
J. Chen et al. / J. Chromatogr. B 991 (2015) 59–67  S. Sriratana, J.F. Scamehorn, S. Chavadej, C. Saiwan, K.J. Haller, S.D. Christian, E.E. Tucker, Use of poly electrolyte-enhanced ultra ﬁltration to remove chromate from water, Sep. Sci. Technol. 31 (1996) 2493–2504.  J. Sánchez, B.L. Rivas, Arsenic extraction from aqueous solution: electrochemical oxidation combined with ultraﬁltration membranes and water-soluble polymers, Chem. Eng. J. 165 (2010) 625–632.  H. Deng, X. Li, Q. Peng, X. Wang, J. Chen, Y. Li, Monodisperse magnetic single-crystal ferrite microspheres, Angew. Chem. Int. Ed. 44 (2005) 2782–2785.  J. Wang, J. Shi, D. Tian, H. Deng, Y. Li, P. Song, C. Chen, Fabrication and enhanced magnetoresistance of SiO2 -coated Fe3 O4 nanosphere compact, Appl. Phys. Lett. 90 (2007) 213106.
 C. Zhang, H. Wang, D. Xing, Multichannel oscillatory-ﬂow multiplex PCR microﬂuidics for high-throughput and fast detection of foodborne bacterial pathogens, Biomed. Microdevices 13 (2011) 885–897.  Z. Ma, Y. Guan, H. Liu, Synthesis and characterization of micron-sized monodisperse superparamagnetic polymer particles with amino groups, J. Polym. Sci. Pol. Chem. 43 (2005) 3433–3439.  J. Thome, A. Holländer, W. Jaeger, I. Trick, C. Oehr, Ultrathin antibacterial polyammonium coatings on polymer surfaces, Surf. Coat. Technol. 174–175 (2003) 584–587.  W.W. Wilson, M.M. Wade, S.C. Holman, F.R. Champlin, Status of methods for assessing bacterial cell surface charge properties based on zeta potential measurements, J. Microbiol. Methods 43 (2011) 153–164.