Novel S-MSPQC cell sensor for real time monitoring the injury of endothelial cell by LPS and assessing the drug effect on this injury

Biosensors and Bioelectronics 71 (2015) 62–67

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Novel S-MSPQC cell sensor for real time monitoring the injury of endothelial cell by LPS and assessing the drug effect on this injury Feifei Tong, Yan Lian, Fengjiao He n State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, China

art ic l e i nf o

a b s t r a c t

Article history: Received 27 January 2015 Received in revised form 17 March 2015 Accepted 5 April 2015 Available online 8 April 2015

A novel square Au microelectrode multi-channel series piezoelectric quartz crystal (S-MSPQC) cell sensor was constructed by square Au microelectrode in series connected with quartz crystal sensor in MSPQC. The experimental results showed square shape Au microelectrode was more sensitive than normally used interdigital microelectrode. New constructed S-MSPQC was successfully used for real time monitoring the injury of human umbilical vein endothelial cells which induced by lipopolysaccharide (LPS) and assessing the drug effects of two anti-oxidative vitamins, VC and VE and their combination against this injury. The detection results showed that the low concentration of LPS may promote the proliferation of the HUVEC cells at first 12 h, then it caused the cell apoptosis. The pretreatment of VC or VE on the cells for 12 h before adding the LPS could partially reduced the injury of LPS on HUVEC cells. The sensor detection results were all consisted with the results detected by MTT assay and microscopy observation. The synergistic effect of the combination of VC–VE on LPS-induced cell injury was also observed by sensor. Compared with the traditional biological methods, the proposed method is sensitive, label-free, non-invasive, cheap, simple, and automatic. It provides a new automatic tool for cell monitor in the field of cell biology, cytobiology, and drug effects research. & 2015 Elsevier B.V. All rights reserved.

Keywords: MSPQC Cell sensor Real time Frequency shift Lipopolysaccharide

1. Introduction Bacterial endotoxin, also known as lipopolysaccharide (LPS), is one of the outer membrane component in most Gram-negative bacteria. As the main pathogenic factor of the G-bacteria, the endotoxin can result in fever and inflammation, as well as cell and tissue injury in human or mammal. In severe cases, it may cause irreversible septic shock and even death (Parrillo et al., 1990; Raetz and Whitfield, 2002). The vascular endothelial cell is the main target cell of which LPS infected on at the early time. The injury and apoptosis of vascular endothelial cell which induced directly or indirectly by LPS is closely related to septic shock and multiple organ dysfunction syndrome or failure (MODF) (Maier, 2000; Marzocco et al., 2004). Research on the effects of the LPS on human cells and its nosogenesis, as well as developing antagonism medicines, is essential for the in clinical medicine. The traditional methods for investigating the injury of human cells induced by LPS and antagonism medicines to this injury are based on the observation of the morphological changes and the detection of the specific protein and gene expression level in a given time. These methods are effective to obtain the detail n

Corresponding author. Fax: þ86 731 88055818. E-mail address: [email protected] (F. He).

http://dx.doi.org/10.1016/j.bios.2015.04.004 0956-5663/& 2015 Elsevier B.V. All rights reserved.

knowledge of the nosogenesis. But they are all invasive end-point detection. They need fluorescence labeling and tedious manipulations, such as disruption and cleavage (Tkach et al., 2008) of cell and cannot realize the high-throughput screening of the samples. In recent years, the sensor based on the transducing of cell– electrode interface information captured researchers' attention. They are constructed by monitoring the changes of the electric parameters, electrochemical reaction, and acoustic wave etc. which happened on the electrode interface. These information are related to important cellular information. Some cell sensors have been developed to monitor the cytotoxicity and drug effects on cell growth, such as electric cell-substrate impedance sensing (ECIS) approach of detecting the toxic effects of ZD 6474 to breast cancer cells (Pradhan et al., 2014); The electrochemical impedance spectroscopy (EIS) method for monitoring the cell/electrode's surface electrochemical reaction caused by the living cells on the electrode (Bouafsoun et al., 2008). QCM response to the mass coupling and the change of the viscoelasticity and the mechanical properties caused by growth of the cell on electrode (Fatisson et al., 2011; Saitakis and Gizeli, 2012), etc. The series piezoelectric quartz crystal (SPQC) sensor is constructed by electrode in series connected to quartz crystal sensor. It not only reserved the highly-sensitive response of PQC to electric paramenters change, but also get rid of the influence of acoustic attenuation in the liquid. Compared with PQC, its stability

F. Tong et al. / Biosensors and Bioelectronics 71 (2015) 62–67

has improved significantly(Shen et al., 1993a, 1993b). Some multichannel SPQC sensors were developed for microorganism detection (He and Zhong, 2010; Mi et al., 2012; Ren et al., 2008; Shi et al., 2014). The electrodes were usually pair electrode and interdigital electrode. In order to monitor the living mammalian cells’physiological and pathological changes affected by the endogenous or exogenous stimulating factor, a new square shape Au microelectrode was fabricated and used to construct the S-MSPQC cell sensor. The experimental results showed square Au microelectrode was more sensitive than normally used interdigital microelectrode (IDE). New S-MSPQC cell sensor was successfully used to monitor the LPS-induced injury on endothelial cell and assess the drug effect of vitamin C, E and their combination on the LPS-induced injury. It will be a sensitive, non-invasive, label-free, automatic, and living cell monitoring and analyzing tool in the field of cell biology and drug effect.

2. Experiments 2.1. Materials and reagents Lipopolysaccharide (LPS; from Eschericha coli, L2880, Sigma) was dissolved by ultrapure water, and then filtered by 0.22 μm of filter membrane to sterilization. Vitamin C (Sigma) was prepared by sterile water, and vitamin E (Sigma) by absolute ethyl alcohol. Human umbilical vein endothelial cells (HUVEC) were taken from the College of Biology, Hunan University, China. Low-glucose DMEM medium, fetal bovine serum (FBS), trypsin, penicillinstreptomycin, and EDTA were purchased from Gibco, USA. MTT reagent and dimethyl sulfoxide (DMSO) were from Sigma, USA. PBS buffer solution was composed of 136.7 mM NaCl þ2.7 mM KCl þ8.1 mM Na2HPO4 þ1.5 mM KH2PO4 (pH¼7.4). All other chemicals were of reagent grade. Ultrapure water (RN18 MΩ/cm) was used throughout the experiment. 2.2. Apparatus Biological phase contrast microscope (CKX41, Olympus, Japan), microtiter plate reader (Bio-Tek, USA), biological safety cabinet (Heal Force, China), CO2 cell incubator (Xiangyi, China), and Multichannel Series piezoelectric quartz crystal sensor (self-designed product for the study). 2.3. Cell Cultures The frozen HUVEC cells were recovered in the low-glucose DMEM medium supplemented with 10% (v/v) FBS, 100 U/mL penicillin, and 0.1 g/mL streptomycin and cultured in a humidified incubator with a 5% CO2, 37 °C, and 95% air atmosphere. The culture medium was changed with a fresh one every three days. At 70%–80% confluence, the cells were harvested by trypsinization (0.25% trypsin/0.02% EDTA). After a 5 min centrifugation at 1000 rpm, the cells were diluted in a fresh medium and seeded at the rate of 1:3.

63

the cells were seeded, the sensor software can automatically record ΔF at different culturing time and output by computer for real time online. Monitoring the injury process of HUVEC cell induced by LPS: the LPS solution was added to the culture-detection well after HUVEC cell were cultured for 24 h. Then HUVEC cell growth process were monitoring by S-MSPQC cell sensor. Monitoring the effects of the antioxidant vitamins on the HUVEC cell injury: vitamin C, or vitamin E, or their combination was added to the culture-detection well after HUVEC cells were cultured for 24 h. After incubated for another 12 h, the LPS solution (25 μg/mL)was added to the culture-detection well. HUVEC cell growth process were monitored by S-MSPQC cell sensor.

3. Results and discussions 3.1. The construction and detection principles of the S-MSPQC cell sensor The schematic diagram of S-MSPQC cell sensor system was shown in scheme 1. This system consisted of four parts. I: cell culturing-detection system which contains eight detection channels; II: oscillating circuits; III: data processing; IV: PC interface and operating system. For part I, The 3.0 mm  3.0 mm area of square Au microelectrode was assembled at the bottom of the cell culturing-detection well. The single finger of electrode was produced with the width of 25 μm, the gap of 25 μm and 15 loops total. Its fabrication process was shown in supplementary material. Each detection channel was provided with independent oscillating circuit and the fundamental frequency of PQC which connected in series was 9 MHz. Researchers could obtain the response curves of eight cell samples simultaneously. The equivalent circuit of cell culture-detection system of S-MSPQC cell sensor and its simplified equivalent circuit model were shown in Scheme 1a. In Scheme 1a, Rs and Cs was the resistance and capacitance of the medium solution; Ri and Ci was the interface resistance and interface capacitance between the cell and the electrode; Re and Ce was the resistance and capacitance of the cell enchylema respectively. C0 was the static capacitance; Lq, Cq, and Rq were the motional inductance, capacitance, and resistance of quartz crystal. As Ri, Ci, Re and Ce were concerned with cell, they can be represented by the cell-caused resistance Rcell and capacitance Ccell, as shown in Scheme 1b. For further simply, Rs, Cs, Rcell, and Ccell can be equivalent to Rt and Ct, as shown in Scheme 1c. For equivalent circuit shown in Scheme 1c, according to the phase shift conditions of oscillation theory, the resonant frequency of the S-MSPQC cell sensor could be deduced as follows (Shen et al., 1993b; Tong et al., 2014):

⎤ ⎡ πF0 Cq (2πF0 Rt2 Ct − ARt ) ⎥ F = F0 ⎢1 + ⎢ 1 − 2πF0 C0 Rt A + 4π 2F02 Rt Ct (C0 + Ct ) ⎥⎦ ⎣

From Eq. 1, resonant frequency F was a function of Rt and Ct: R ·R F = f (Rt , Ct ). As Rt = R cell+ Rs , Ct = Ccell + Cs , so: cell

2.4. Monitoring of HUVEC cell growth by S-MSPQC cell sensor Monitoring HUVEC cells growth: The culture-detection plate of the S-MSPQC cell sensor was sterilized by rinsed with 75% ethanol and ultraviolet irradiated for 15 min, then preheated in the incubator for 30 min. About 500 μL/ well of HUVEC cell suspensions were added in the culture-detection wells and incubated. F0 is corresponding to the resonant frequency of the sensor before cell seeded, Fi is corresponding to the frequency after cell seeded at i time and the responsed frequency shifts is ΔF¼ Fi  F0. After

(1)

s

∂F ∂F × dRt + × dCt dF = ∂Rt ∂Ct ⎛ R cell × Rs ⎞ ∂F ∂F = × d⎜ × d (Ccell + Cs ) ⎟+ ∂Rt ⎝ R cell + Rs ⎠ ∂Ct

(2)

During the cell growth process, it was found that the resistance and capacitance of the medium solution had little change by measurement using the 4192 A Impedance Analyzer at 9 MHz. So Rs and Cs could be regarded as constant. For the S-MSPQC cell sensor:

64

F. Tong et al. / Biosensors and Bioelectronics 71 (2015) 62–67

Scheme 1. The diagram of the S-MSPQC cell sensor and the corresponding equivalent circuit model. (I) Cell culture-detection system; (II) oscillating circuit; (III) data processing system; (IV) PC interface. (a) The equivalent circuit of the detection system of thel sensor; (b, c) the simplified equivalent circuit model.

ΔF = K1 ×

Rs 2 (R cell + Rs

∂F ∂Ccell

× ΔR cell + K2 × ΔCcell

(3)

⎧ ⎫ 2 Rt2 A − 4πF0 Ccell Rt ⎪ ⎪ A − 4π 2F02 Ccell ⎬ = πF02 Cq ⎨ ⎪ ⎡⎣1 − 2πF0 C0 Rt A + 4π 2F02 Rt2 Ccell (C0 + Ccell ) ⎤⎦2 ⎪ ⎭ ⎩ ⎧ ⎫ 2 2 2 2 ⎪ ⎪ 1 − 4π F0 Ccell Rt + 4πF0 Ccell Rt A ⎬ = 2π 2F03 Cq ⎨ 2 ⎪ ⎡1 − 2πF0 C0 Rt A + 4π 2F 2 R 2 Ccell (C0 + Ccell ) ⎤ ⎪ 0 t ⎦ ⎭ ⎩⎣

where, K1 =

K2 =

)2

∂F ∂Rt

ΔF = 9.87 ×

Rs 2 (R cell + Rs )2

R cell + 1.66 × 1 × Δ09 × ΔCcell

(4)

Eq. 4 showed that the change of resonant frequency (i.e. frequency shift) of S-MSPQC cell sensor was affected by the change of cell-caused resistance and capacitance on the electrode as well as the resistance of the medium solution. The attachments and growth of cell on the surface of square Au microelectrode will change these electric parameters of the microelectrode and results in sensitive resonant frequency respose of MSPQC.

The parameters of PQC have been known as F0 = 9 × 106 (Hz);

C0 = 1 × 10−11 (F); Cq = 1 × 10−13 (F); A¼ tan( 71.3)¼ 2.95。The Rt and

Ccell

were

measured

at

culturing

32 h:

Rt = 160(Ω),

Ccell = 2.7 × 10−10 F therefore, K1 ¼ 9.87; K2 ¼ 1.66  109. Hence, ΔF expression was obtained as follows:

Fig. 1. Growth frequency shift response curve of 1  104 cell/mL HUVEC cells.

3.2. Monitoring the growth of HUVEC cells for real time by S-MSPQC cell sensor 3.2.1. Growth frequency shift response curve of HUVEC cells The growth frequency shift response curves of 1  104 cell/mL HUVEC cells monitored by the S-MSPQC cell sensor was showed in Fig. 1. The frequency shift increased rapidly during the initial 12 h after the cell seeding. This was because the cells were gradually sinking in the suspension, adhesion to the microelectrode and resulting in the change of ΔRcell and ΔCcell . Thus the frequency shift increased. In the next 22–84 h, the frequency shift continued to increase, which caused by the proliferation of the HUVEC cells, and the number of the cells increased on the sensor electrode. The maximum of frequency shift was about 670 Hz. After that, the frequency shift was tending to stabilize as the cell growth reached a plateau. For comparation, the DMEM medium without cells was also monitored as the control. 3.2.2. Effect of cell seeding density on growth response curve The growth response curve of HUVEC cell seeding density of 0.1, 1, 5, 10  104 cell/mL detected by S-MSPQC cell sensor were

F. Tong et al. / Biosensors and Bioelectronics 71 (2015) 62–67

Fig. 2. Frequency shift curves of HUVEC cells with different densities monitoring by S-MSPQC cell sensor. (insert: the line curve of the frequency shift-cell density when culturing for 12 h).

shown in Fig. 2. It was found that sinking and adherence process were completed for all seeding density in 12 h. The greater density of seeding cell, the higher of frequency shift obtained, The linear relationship was observed between the frequency shift and the cell seeding density at the time point of 12 h (Fig. 2 insert). Each data was the average of three times detection. The detection limit of the constructed sensor was 1.36  103 cell/mL, which was calculated on the standard of the triple signal-to-noise analysis. 3.2.3. Effect of electrode shape on the detection The sensor’s responsive sensitive can be influenced by culturedetection well constant which determined by the electrode parameters. Two shape of gold microelectrode had used for comparison. The frequency shift response in different concentration of KCl detected by IDE-MSPQC and S-MSPQC were shown in Fig. 3(a). It can be seen that the frequency response of S-MSPQC to conductance change of solution was sensitive than that of IDE-MSPQC. Fig. 3(b) showed the comparison of the maximum frequency shift caused by the HUVEC cells growing on two shapes of electrode. It implied that the response signal of the cell growth on the square Au microelectrode was stronger than that on the IDE electrode. 3.3. Monitoring the injury of LPS-induced on cell by S-MSPQC cell sensor The prevention and the cure for LPS-induced cell injury is one of the active medical research fields in recent 20 years. When the bacterial LPS enters the blood circulation through the wound or

65

the intestinal, it may induce the vascular endothelial cells to secrete multiple inflammatory mediators, including tumor necrosis factor (TNF), interleukin-1 (IL-1), adhesion molecule, reactive oxygen species (ROS), and nitric oxide (NO) etc. These mediators induce the injury and apoptosis of the endothelial cells and result to the dysfunction of the blood vessel, even to the systemic inflammatory response syndrome (SIRS) (Griffin, 1998; Henry et al., 2009). In this paper, the injury process of HUVEC cell induced by LPS was monitored for real time by the S-MSPQC cell sensor. The HUVEC cells (1  104 cell/mL) were cultured for 24 h in the culture-detection wells. Then 1 μg/mL LPS solutions was added to the culture medium. Continued to cultured and recorded the response curve of the frequency shift by S-MSPQC and showed in Fig. 4(a). For comparation, the growth curve of 1  104 cell/mL HUVEC cells with no LPS in culture medium was also monitored as control. As the LPS adding, there was a small stimulus on HUVEC cell monolayer, and then the frequency shift recovered. After 10 h, the response curve which contained LPS decreased significantly because LPS reduced the cell proliferation, and resulted in the injured and dead cells increasing. The effect of LPS concentration (0.1, 0.5, 1, 5, 25 μg/mL) on the growth of HUVEC cells were detected and shown in Fig. 4(b). The frequency shift were obviously declined with the increasing concentration for 1, 5, 25 μg/mL of LPS regardless the disturbing signal caused by the dropping LPS. The interesting phenomena was found that when the cells were treated with LPS concentrations of 0.1 and 0.5 μg/mL, the frequency shifts were higher than the control group at first and then lower than the control group after 33 h. This means LPS might promote the proliferation of the HUVEC cells in the low concentration at first time; and then finally showed the effect of the cell growth inhibition and apoptosis. MTT assay and optical microscopy graph were run for the comparation and the results were shown in Fig. 4(c)–(d) (the method of MTT assay was shown in supplementary material). The results were consistent with that detected by S-MSPQC cell sensor. Under the low-degree bacterial infection, the host cells transferred the mechanism of resistance actively and reduced the injury of LPS-induced, while under the serious infection, the high concentration LPS caused most of endothelial cells injury and led to strong systemic inflammatory response syndrome(Astiz et al., 1995; Russell et al., 2000). The S-MSPQC cell sensor offered a effective method for studying the nosogenesis of the LPS to vascular endothelial cells.

Fig. 3. Effect of electrode shape on the detection. (a) The frequency shift response in different concentration of KCl detected by IDE-MSPQC and S-MSPQC. (b) The comparison of the maximum frequency shift caused by HUVEC cells growing on two shapes of electrode. (The IDE microelectrode had 12 pairs of finger with the length of 3000 μm and the gap of 25 μm.)

66

F. Tong et al. / Biosensors and Bioelectronics 71 (2015) 62–67

Fig. 4. (a) Monitoring the injury of HUVEC cells induced by LPS for real time with S-MSPQC cell sensor; (b) effect of LPS concentration on the growth of HUVEC cells; (c) the result of MTT assay; (d) optical microscopy graph of the HUVEC cells treated by the LPS on the square Au microelectrode (if LPS was added, t ¼0).

3.4. Assessing the drug effects by S-MSPQC cell sensor In the process of the bacteria growth, cell autolysis phase and cell disruption at the infection site, the LPS released from the bacteria’s outer membrane to ambient tissue. The excessive employment of antibiotics in period of mid or later infection could cause a great number of bacteria killed, and the releasing of the LPS would aggravate the patient’s condition (Holzheimer, 2001). It reported that the oxidative stress response participates in injury caused by the LPS-induced. The LPS stimulated the cell to overexpress iNOS protease, and produced numerous NO compound which reacted with free radical, such as superoxide anion. The later produce stronger toxicity peroxynitrite anion (ONOO  ) which cause the injury of the endothelial cells and at the same time increase the permeability of the vessel endothelium(Chandra et al., 2006; Vo et al., 2005). So it is of great clinical significance how to inhibit the oxidative stress response induced by the bacterial LPS and maintain the integrity of the blood vessel endothelium by the drug effectively. The ascorbic acid (vitamin C, VC) and alpha-tocopherol (vitamin E, VE) are two important antioxidants which can remove free radicals in the human antioxidant defense system and inhibit the oxidative stress response product effectively(Carlson et al., 2006; Dwenger et al., 1994). LPS and VC or VE drug with different concentrations or VC–VE combination were used to treat the HUVEC cells. The protection effect of VC or VE or VC–VE combination were detected by the S-MSPQC cell sensor. The HUVEC cells (1  104 cell/mL) were cultured in the culturedetection well for 24 h in normal medium, then cultured for another 12 h in medium containing the VC or VE (10, 50, and 100 μmol/L). After that, the LPS with the final concentration of 25 μg/mL) was added and recorded response curve with the

constructed sensor. The results were shown in Fig. 5(a)–(b). Results indicated that the pretreatment of VC or VE medium separately on the cell for 12 h before the adding the LPS showed certain protection effect on the LPS-induced oxygen cell injury on the endothelial cells and could partial reverse LPS-induced injury. Meanwhile, it was found that the effect of VC against the LPS-induced injury was better than VE. 50 μmol/L VC showed better protection effect than 100 μmol/L VC. Because excessive VC caused decomposition of the lipid peroxide in the cell and led to the cell reducing stress state of pro-oxidation, thus toxic to the cell (Pleiner et al., 2002; Wilson, 2009). It is reported that VE could be regenerated by VC reacted with free radical VE. VE was recycled and the efficiency of removing the superoxide anion free radical improved(Selin et al., 2013). The combination use of VC–VE was of synergistic effect to the antioxidant injury activity. The effect of VC–VE combination on LPSinduced injury detected by the S-MSPQC cell sensor was shown in Fig. 5(c). It can be seen that the combination of VC–VE showed the protective effect on LPS-induced injury to the cell at low concentration (10 μmol/L). Therefore, for the treatment of inflammation related to sepsis and injury of tissue, it may be an active suggestion to consider the low concentration VC–VE combined employment.

4. Conclusion The physiological and pathological changes of cell which caused by the endogenous or exogenous stimulating factor in its growth state is the bases of various illnesses. These changes are dynamic and depend on the dosage or treating time of stimulating factor. It is expected to develop an automatic biosensor for

F. Tong et al. / Biosensors and Bioelectronics 71 (2015) 62–67

67

their combination in vitro. The detected results by constructed sensor were consistent with those by MTT assay and microscopic observation. Proposed S-MSPQC cell sensor provides high-sensitivity and effective detection tool in the fields of cell biology and drug effect.

Acknowledgment This research work was supported by the National Natural Science Foundation of China (No. 21275042) and National High Technology Research and Development Program of China (863 Program) (no. 2013AA020203)

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.04.004.

References

Fig. 5. Frequency shift curve of the VC or VE separately, or their combination treatment to the HUVEC cell injury of LPS-induced. (a) pretreatment by VC; (b) pretreatment by VE; (c) pretreatment by VC–VE combination).

monitoring the cell growth state in vitro sensitively and continuously. In this paper, the S-MSPQC cell sensor was constructed and applied for real time monitoring of the LPS-induced injury on endothelial cell and assessing the drug effect of vitamin C, E, and

Astiz, M.E., Rackow, E.C., Still, J.G., Howell, S.T., Cato, A., Von Eschen, K.B., Ulrich, J.T., Rudbach, J.A., McMahon, G., Vargas, R., 1995. Crit. Care Med. 23, 9–17. Bouafsoun, A., Othmane, A., Jaffrezic-Renault, N., Kerkeni, A., Thoumire, O., Prigent, A.F., Ponsonnet, L., 2008. Mater. Sci. Eng. C 28, 653–661. Carlson, D., Maass, D.L., White, D.J., Tan, J., Horton, J.W., 2006. Am. J. Physiol.: Heart C 291, H2779–H2789. Chandra, A., Enkhbaatar, P., Nakano, Y., Traber, L.D., Traber, D.L., 2006. Clinics 61, 71–76. Dwenger, A., Pape, H.C., Bantel, C., Schweitzer, G., Krumm, K., Grotz, M., Lueken, B., Funck, M., Regel, G., 1994. Eur. J. Clin. Investig. 24, 229–235. Fatisson, J., Azari, F., Tufenkji, N., 2011. Biosens. Bioelectron. 26, 3207–3212. Griffin, G.E., 1998. J. Antimicrob. Chemother. 41, 25–29. He, F.J., Zhong, M., 2010. Talanta 80, 1210–1215. Henry, C.J., Huang, Y., Wynne, A.M., Godbout, J.P., 2009. Brain Behav. Immun. 23, 309–317. Holzheimer, R.G.J., 2001. J. Chemother. 13, 159–172. Maier, R.V., 2000. Surg. Infect. 1, 197–204 (discussion 204-195). Marzocco, S., Di Paola, R., Ribecco, M.T., Sorrentino, R., Domenico, B., Genesio, M., Pinto, A., Autore, G., Cuzzocrea, S., 2004. Free Radic. Res. 38, 1143–1153. Mi, X.W., He, F.J., Xiang, M.Y., Lian, Y., Yi, S.L., 2012. Anal. Chem. 84, 939–946. Parrillo, J.E., Parker, M.M., Natanson, C., Suffredini, A.F., Danner, R.L., Cunnion, R.E., Ognibene, F.P., 1990. Ann. Intern. Med. 113, 227–242. Pleiner, J., Mittermayer, F., Schaller, G., MacAllister, R.J., Wolzt, M., 2002. Circulation 106, 1460–1464. Pradhan, R., Mandal, M., Mitra, A., Das, S., 2014. IEEE Sens. J. 14, 1476–1481. Raetz, C.R.H., Whitfield, C., 2002. Annu. Rev. Biochem. 71, 635–700. Ren, J.L., He, F.J., Yi, S.L., Cui, X.Y., 2008. Biosens. Bioelectron. 24, 403–409. Russell, J.A., Singer, J., Bernard, G.R., Wheeler, A., Fulkerson, W., Hudson, L., Schein, R., Summer, W., Wright, P., Walley, K.R., 2000. Crit. Care Med. 28, 3405–3411. Saitakis, M., Gizeli, E., 2012. Cell. Mol. Life Sci. 69, 357–371. Selin, J.Z., Rautiainen, S., Lindblad, B.E., Morgenstern, R., Wolk, A., 2013. Am J Epidemiol. 177, 548–555. Shen, D.Z., Zhu, W.H., Nie, L.H., Yao, S.Z., 1993a. Anal. Chim. Acta 276, 87–97. Shen, D.-Z., Li, Z.-Y., Nie, L.-H., Yao, S.-Z., 1993b. Anal. Chim. Acta 280, 209–216. Shi, X.H., He, F.J., Lian, Y., Yan, D.Y., Zhang, X.Q., 2014. Sens. Actuators B: Chem. 198, 431–437. Tkach, A.V., Ivanova, L.A., Statsenko, I.V., 2008. Med. Tr. Prom. Ekol., 28–35. Tong, F., Lian, Y., Zhou, H., Shi, X., He, F., 2014. Anal. Chem. 86, 10415–10421. Vo, P.A., Lad, B., Tomlinson, J.A.P., Francis, S., Ahluwalia, A., 2005. J. Biol. Chem. 280, 7236–7243. Wilson, J.X., 2009. Biofactors 35, 5–13.