Highly Sensitive Method for Assay of Drug-Induced ... - ACS Publications

Aug 2, 2012 - The principle of this method is based on the assay of DNA ... Methods: Annexin-V/Polyethylene Glycol-Functionalized Quantum Dots as ... ...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Highly Sensitive Method for Assay of Drug-Induced Apoptosis Using Fluorescence Correlation Spectroscopy Lingao Ruan,† Zhancheng Xu,† Tao Lan,† Jinjie Wang,† Heng Liu,† Chaodong Li,‡ Chaoqing Dong,† and Jicun Ren*,† †

College of Chemistry and Chemical Engineering, State Key Laboratory of Metal Matrix Composites, Shanghai Jiaotong University, Shanghai 200240, People's Republic of China ‡ School of Biotechnology, East China University of Science and Technology, Shanghai 200237, People’s Republic of China S Supporting Information *

ABSTRACT: Apoptosis plays a crucial role in many biological processes and pathogenesis of various malignancies and diseases of the immune system. In this paper, we described a novel method for sensitive detection of druginduced apoptosis by using fluorescence correlation spectroscopy (FCS). The principle of this method is based on the assay of DNA fragmentation in the process of the drug-induced apoptosis. FCS is a single molecule method, and it can be used for sensitive and selective assay of DNA fragmentation without separation. We first developed a highly sensitive method for characterization of DNA fragments using a home-built FCS system and SYBR Green I as fluorescent DNA-intercalating dye, and then established a model of druginduced apoptosis using human pancreatic cancer cells and a drug lidamycin. Furthermore, FCS method established was used to directly detect the fragmentation of DNA extracted from apoptotic cells or in the apoptotic cell lysate. In FCS assay, the single-component model and the multiple-components model were used to fit raw FCS data. The characteristic diffusion time of DNA fragments was used as an important parameter to distinguish the apoptotic status of cells. The obtained data documented that the characteristic diffusion time of DNA fragments from apoptotic cells significantly decreased with an increase of lidamycin concentration, which implied that DNA fragmentation occurred in lidamycin-induced apoptosis. The FCS results are well in line with the data obtained from flow cytometer and gel electrophoresis. Compared to current methods, the method described here is sensitive and simple, and more importantly, our detection volume is less than 1 fL, and the sample requirement can easily be reduced to nL level using a droplets array technology. Therefore, our method probably becomes a high throughput detection platform for early detection of cell apoptosis and screening of apoptosis-based anticancer drugs.

A

atrophy, and embryonic and brain development.12 The assay of apoptosis not only greatly improves the diagnosis of certain diseases, but also serves as an early indicator for the effect of therapeutic interventions.13 Currently, various detection methods have been developed for assay of apoptosis, and the principles of these methods are mainly based on the change of morphology,14 PS extroversion level,15 caspases activation,16 and DNA fragmentation17 while cells are undergoing apoptosis. The morphological change of apoptotic cells is generally observed by using electron microscopy.18 Electron microscopy possesses the high spatial resolution, and electron microscopic observation is considered to be the gold standard for the detection of apoptosis.19 However, this assay needs a great deal of time and a high skill for preparation of samples and measurements of cells.20 The externalization of PS to the outer leaflet of the plasma

poptosis (or programmed cell death) refers to the procedure by which cells end their lives following certain programs in a physiological or pathological conditions,1,2 and it plays a crucial role in many biological processes, pathogenesis of various malignancies and diseases of the immune system, and screening of apoptosis-based anticancer drugs. Cells undergoing apoptosis show typical and well-defined morphological changes, including plasma membrane blebbing,3 chromatin condensation with margination of chromatin to the nuclear membrane,4 nuclear fragmentation,5 and formation of apoptotic bodies.6 Apoptosis can be characterized by several biochemical criteria, including different kinetics of phosphatidylserine (PS) exposure on the outer leaflet of the plasma membrane,7 changes in mitochondrial membrane permeability,8 release of intermembrane space mitochondrial proteins,9 caspase-dependent activation,10 and nuclear translocation of a caspase-activated DNase resulting in internucleosomal DNA cleavage.11 On the other hand, apoptosis is responsible for many biological processes such as normal cell turnover, proper development and functioning of the immune system, hormone-dependent © 2012 American Chemical Society

Received: March 28, 2012 Accepted: August 2, 2012 Published: August 2, 2012 7350

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

membrane is an early marker of apoptotic cells death,21 which can be detected by the binding of Annexin V labeled with certain fluorochromes, such as fluorescein isothiocyanate22 and Alexa Fluor dyes.23 Annexin is a family of calcium-dependent phospholipid-binding proteins that preferentially bind PS. A flow cytometer is usually used for measuring the PS level on cell membrane, and is further used to distinguish between apoptosis and necrosis by combining Annexin V-dye with viability dyes such as propidium iodide (PI). In early stage apoptosis, the plasma membrane excludes PI. But flow cytometry has no temp-spatial resolution, and cannot be used to obtain tempspatial information about the apoptotic cells.24 Caspases are a group of proteases, and act together as a cascade in the selective cleavage of proteins after aspartic acid residues when they are activated by apoptosis.25 Therefore, caspases are an ideal early target for measuring apoptosis, and Western blot analysis is generally used to detect caspase activity.26 Fragmentation of chromosomal DNA is the biological hallmark of apoptosis,27 and can be detected by gel electrophoresis,18 by ELISA,28 and by TUNEL.29 Among them gel electrophoresis is usually used to detect apoptosis due to no requirements of special instrument and reagents. However, gel electrophoresis has low sensitivity,30 and is generally considered to be tedious and time-consuming. Recently, certain new methods and techniques such as single molecule fluorescent spectroscopy, microfluidic device, and electrochemical methods are applied for an assay of apoptosis. Valero et al. designed a microfluidic device to examine caspase activity outside of cells using on-chip flow cytometer.31 Zhu’s group presented a new method for detection of the apoptosis using a microfluidic device and the Annexin V conjugated quantum dots (QDs) as probes,32 and furthermore this group developed an electrochemical cytosensing platform for monitoring the early stages of apoptosis using novel lectinfunctionalized SiO2@QDs nanoassemblies as amplified signal probes.33 Wu et al. combined a simple electrochemiluminescent transducer for evaluation of apoptosis by dual signal amplification of multiwalled carbon nanotubes (MWCNTs) and antiphosphatidyl serine antibody APSA-SiO2@Ru nanoprobes, and the resultant cytosensor could respond as low as 800 cells/mL.34 Fluorescence correlation spectroscopy (FCS) is a single molecule method for studying molecular diffusion,35 chemical kinetics, and molecular interaction in vitro and in vivo.36,37 Martinez et al. used fluorescent probe (aspartic acid)2rhodamine 110 as substrates and Jurkat cells as a model of apoptosis, and in vivo investigated caspase activity by FCS. The apoptotic cells showed the characteristic diffusion times and molecular brightness values of free rhodamine 110,38 which were drastically different from nonapoptotic autofluorescent cells, and more recently this group synthesized a new red fluorescent caspase probe Nile Blue derivative (2SBPO)-Casp for assay of apoptosis.39 In this paper, we described a novel method for sensitive detection of drug-induced apoptosis by FCS. The principle of this method is based on the assay of DNA fragmentation in the process of the drug-induced apoptosis, which is similar to DNA ladder assay. We first developed a highly sensitive FCS method for characterization of DNA fragments using SYBR Green I as fluorescent DNA-intercalating dye, and then established a model of drug-induced apoptosis using human pancreatic cancer cells and a drug lidamycin. In FCS measurements, the single-component model and the multiple-components model were used to fit raw FCS data. The characteristic diffusion time

of DNA fragments was used to distinguish the apoptotic status of cells. FCS method was used to directly detect the fragmentation of DNA extracted from apoptotic cells or in the apoptotic cell lysate. The results obtained from FCS method are well in line with the data obtained from the flow cytometer and gel electrophoresis.

■

EXPERIMENTAL SECTION Materials. Rhodamine Green and SYBR Green I (10 000×) were purchased from Molecular Probes Inc. Human pancreatic cancer cells (PANC-1) were obtained from the Committee on Type Culture Collection of Chinese Academy of Sciences (China). The 50 bp, 200 bp, 500 bp, 2.6 kbp, and 10 kbp dsDNA Markers were purchased from Beijing Comwin Biotech Co. Ltd. (China). The 1 kbp DNA Ladder was purchased from Fermentas International Inc. (Canada), and it contained 250 bp, 500 bp, 750 bp, 1 kbp, 1.5 kbp, 2 kbp, 2.5 kbp, 3 kbp, 3.5 kbp, 4 kbp, 5 kbp, 6 kbp, 8 kbp, and 10 kbp DNA fragments. Calf thymus dsDNA was product of Sigma-Aldrich Chemical Co. (Milwaukee). Annexin V-FITC/PI Apoptosis Detection Kit (50 T) was a product of Nanjing KeyGen Biotech. Co., Ltd. (China). Lidamycin (LDM) was product of Shanghai Laiyi Center for Biopharmaceutical R&D (China). All materials were of analytical grade and used without further purification. Preparation of dsDNA Fragment Markers. The solutions of dsDNA fragments markers were freshly diluted with TE buffer solution (10 mM Tris-HCl with 1 mM EDTA, pH 8.0). The original SYBR Green I solution was diluted 2500 times by TE buffer. A working solution of 1/2500 SYBR Green I was freshly prepared prior to use. The concentrations of DNA extracted from apoptotic cells and in cell lysates were determined by fluorescent spectrometry using calf thymus dsDNA as a standard material and SYBR Green I as fluorescent DNA-intercalating dye. Fluorescence spectra were recorded with an F-380 spectrometer (Tianjin Gangdong SCI & Tech. Development Co., Ltd., China). Cell Culture and Drug-Induced Apoptosis. PANC-1 cells were maintained in vitro in DMEM high glucose medium (Gibco, California) supplemented with 10% (v/v) fetal bovine serum (FBS) (Gibco). Cells were incubated at 37 °C in a humidified incubator with 5% CO2. In apoptosis induction, cells were incubated with 0.5 μg/mL, 5.0 μg/mL, and 50 μg/ mL of LDM, respectively. Cells were induced with LDM for the desired time (20 h). Cells without treatment of LDM were used as a control group. DNA Extraction. Approximately 1 × 106 PANC-1 cells were collected, and the medium was removed. The 300 μL lysis solution (100 mM Tris-HCl containing 40 mM EDTA, 0.2% SDS, and 1 M NaCl, pH 8.0) was added to collected cells, and was incubated at 37 °C for 30 min. The phenol−chloroform method was used to extract DNA from lysed cells. The extracted DNA was precipitated with isopropanol (0.1 volume of 3 M sodium acetate and 0.6 volume of isopropanol, pH 4.8), and was washed once with 75% ethanol. Finally, the extracted DNA was dissolved in 10 μL TE buffer, and was incubated with 2 μL RNase at 37 °C for 3 h. DNA Ladder Assay (Gel Electrophoresis). The 5 μL DNA and 1 μL 6× DNA loading buffer mixed samples were added into wells of 1% agarose gel in TAE containing 0.5 μg/ mL ethidium bromide. Gel electrophoresis was run at 100 V for 60 min. DNA ladders were finally visualized by a UV light source and documented by photography. DNA extracted from apoptotic cells showed a distinct DNA ladder, and DNA from 7351

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Here, τ is the correlation time, N is the average number of fluorescent molecules in the detection volume, T and τtr are the fractional population and decay time of the triplet state, respectively, τD is the characteristic diffusion time, and ω0 and z0 are the distances from the center of the laser beam focus in the radial and axial directions, respectively. In the multiple-component model, we suppose that the brightness of various components is identical and the contribution of the triplet state formation can be neglected for the sake of simplicity. When n of components diffuse through the detection volume with the characteristic diffusion time τDi, eq 1 can be modified as the following equation:

viable cells stayed on the top of the gel as a high-molecularweight band. Analysis of Apoptosis by Flow Cytometer. About 1 × 106 PANC-1 cells were incubated with LDM at 37 °C for 20 h to induce apoptosis. Apoptotic cells were first trypsinized, then were washed with PBS, and finally were centrifuged at 2000 rpm for 5 min to get cell pellets. The cell pellets were resuspended in 500 μL Annexin V Binding buffer, and incubated with 5 μL Annexin V-FITC and 5 μL PI staining solution at 25 °C for 15 min in the dark. Flow cytometry (Becton, Dickinson and Company) was used to analyze the apoptotic status of cells. In the measurements, 488 nm excitation wavelength and 530 nm emission wavelength were used for detection of Annexin V-FITC (FL1 channel). A bandpass filter (>600 nm) was used for PI detection (FL3 channel). Electronic compensation of the instrument is required to exclude overlapping of the two emission spectra. Events falling in the FITC (+)/PI (−) region of the lower right quadrant are counted as apoptotic cells. FCS Setup and FCS Assay of Apoptosis. FCS measurements were performed with a home-built FCS system, which was based on an inverted Olympus IX 71 microscope (Japan). In brief, an argon laser with 488 nm laser line (ILT Technology, Shanghai, China) was attenuated to about 40 μW by a circular neutral density filter, and then expanded to underfill the back aperture of the objective lens. The expanded laser beam was focused with a water immersion objective (UplanApo, 60× NA 1.2, Olympus, Japan) to a small volume within the diluted sample. The resulting excitation volume is about 0.5 fL. The excited fluorescence signal was collected by the objective passed through the dichroic mirror (505DRLP, Omega Optical) and then was filtered by a band-pass filter (530DF30, Omega Optical) to block scattering laser light. Finally, the fluorescence was coupled into a 35-μm pinhole at the image plane in front of the single-photon counting module (SPCM-AQR16, Perkin-Elmer EG and G, Canada). The fluorescence fluctuations were correlated with a digital correlator (Flex02-12D/C, Correlator.com). FCS measurements were carried out over a period of 240 s at room temperature (about 25 °C), and were repeated 5 times. The detection volume of FCS system was measured using Rhodamine Green (RG) with the concentration of 1.8 nM (its diffusion coefficient: 2.8 × 10−6 cm2/s in water) as a standard substance. The detection volume obtained was about 0.5 fL.

■

n

G D(τ ) =

i=1

n

G D(τ ) =

(

τ τD

1+

2

( ) ω0 Z0

×

τ τD

1 1+

2

( ) ω0 Z0

×

τ τDi

(2)

τDi τDi + τ

1 1+

2

( ) ω0 Z0

×

τ τDi

(3)

Here, αi is the distribution of the characteristic diffusion time associated with τDi. The definitions of χ2 and S in MEMFCS were described in detail by Maiti et al. The αi is associated with the concentration and the brightness of the component i, and its physical meaning is different from that of Ni in eq 2.40 Herein, the multiple-component analysis is performed by using MEMFCS software provided by Maiti, and the distribution of the characteristic diffusion times is used to distinguish the apoptotic status of cells. FCS Assay of DNA Fragments. FCS is a single molecule detection method using statistical analysis of the fluctuations of the fluorescence emitted from a small, optically well-defined open volume element due to Brownian motion of fluorescent molecules, and therefore, the measurement time, fluorescent molecule concentration, and apparatus parameters (such as laser intensity and detection volume) significantly affect the reproducibility of FCS results. In assay of apoptosis, we used the characteristic diffusion time of dsDNA fragments to distinguish the apoptotic status of cells; therefore, the reproducibility of the characteristic diffusion time will be greatly important. We investigated the concentration effects of

)

1

×

∑ αi i=1

RESULTS AND DISCUSSION

⎛ 1 T e−τ / τtr ⎞ 1 ⎟ × ⎜1 + 1−T ⎠ 1+ N ⎝

τDi Ni(τDi + τ )

Here, Ni and τDi denote the number of fluorescent molecules in the detection volume and the characteristic diffusion time of component i, respectively. In eq 2, the data analysis, in fact, is a multicomponent distribution problem. Maiti’s group for the first time introduced a maximum entropy method (MEM) based fitting procedure to analyze FCS raw data, which was called MEMFCS.40 So far, this method has been successfully to investigate the aggregation of amyloid β peptide in the solution,41 the size distribution and polydispersity of microemulsion droplets in solution,42 and the diffusion behaviors of herceptin-conjugated gold nanorods in different cell organelles.43 In MEMFCS, the data analysis is based on minimizing quantitative parameter χ2 as well as maximizing entropy S to obtain an optimal fit when the characteristic diffusion times of the different species are involved in a given focal volume. In fact, the fluorescent brightness values of different species (DNA fragments with different length) are nonidentical in this study; G(τ) in eq 2 should be rewritten to obtain a continuous distribution of characteristic diffusion times as the following equation:

FCS Data Processing. In FCS assay, the characteristic diffusion times of DNA fragments were used to distinguish the apoptotic status of cells. Two models, the single-component model and the multiple-component model, were used to fit raw FCS data. In the single-component model, FCS data were analyzed with the standard eq 1 and nonlinearly fitted with the Microcal Origin 6.1 software package based on the Levenberg− Marquardt algorithm: G (τ ) =

∑

(1) 7352

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Figure 1. (a) Relationships between the characteristic diffusion time and the concentration of 500 bp dsDNA and the concentration of SYBR Green I. In the study on the effects of the dsDNA concentration, the concentrations of SYBR Green I was 1/20 000, and the concentrations of 500 bp dsDNA were 20, 50, 100, 250, and 500 ng/mL, respectively. In the study on the effects of the SYBR Green I concentration, the concentration of 500 bp dsDNA was 200 ng/mL, and the concentrations of SYBR Green I were 1/100 000, 1/40 000, 1/20 000, 1/10 000 and 1/5000, respectively. The measurement time was 240 s. (b) The relationship between the characteristic diffusion time and the measurement time. The measurement times in this experiment were 30 s, 60 s, 120 s, 180 s, and 240 s, respectively. The error bars represent the standard deviation of 5 times measurements.

Figure 2. (a) Typical normalized autocorrelation curves and their fitting curves of DNA fragments using single-component model. The inset of part a shows the relationship between the characteristic diffusion time and the lengths of DNA fragments. (b) The fitting residuals. (c) Typical normalized autocorrelation curves and their fitting curves of DNA fragments using multiple-component model. The inset of part c shows the distributions of the characteristic diffusion times of DNA fragments. (d) The fitting residuals. The lengths of DNA fragments used in this study were 50 bp, 200 bp, 500 bp, 2.6 kbp, and 10 kbp dsDNA Markers, respectively. Rhodamine Green was used as a standard substance. The measurement time was 240 s. The error bars represent the standard deviation of 5 times measurements.

DNA and SYBR Green I on the characteristic diffusion times. The characteristic diffusion times are obtained by fitting the correlation curves using eq 1, and the results are shown in Figure 1a. In about 20 to 500 ng/mL of DNA and 1/100 000 to 1/5000 SYBR Green I, the characteristic diffusion times of dsDNA show no obvious change, and the relative standard deviations (RSDs) of the characteristic diffusion time are about 4% to 16%. In subsequent experiments, 100 ng/mL dsDNA

solution and 1/40 000 SYBR Green I were used, and RSDs of the characteristic diffusion time were about 4%. The effect of the measurement time is shown in Figure 1b. The RSDs of the characteristic diffusion time considerably decreased with an increase of the measurement time. In order to obtain good reproducibility, the measurement time of 240 s was used in this study. 7353

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Figure 3. Strategy for FCS assay of apoptosis. PANC-1 expresses human pancreatic cancer cell, and LDM represents the induced drug lidamycin. The principle of FCS assay is based on the significant difference between the characteristic diffusion times of DNA fragments in apoptotic cells and normal cells.

Figure 4. Autocorrelation curves and their fitting curves of DNA extracted from apoptotic cells induced with different concentrations of LDM by using single-component model (a) and multiple-component model (c). Their fitting residuals are shown in b and d. The inset of part a shows the relation between the characteristic diffusion times of DNA fragments and LDM concentrations. The inset of c shows the distributions of the characteristic diffusion times of DNA fragments under different concentrations of LDM. The 200 bp dsDNA marker and the 1 kbp DNA Ladder were used as standard substances, and the concentrations of LDM used in this study were 0.5 μg/mL, 5.0 μg/mL, and 50 μg/mL, respectively. The measurement time was 240 s. The error bars represent the standard deviation of 5 time measurements.

0.995. The fitting residuals are shown in Figure 2b, which reflected that the goodness of fit was satisfactory. As shown in the inset of Figure 2a, the characteristic diffusion times of DNA fragments dramatically increased with an increase of the DNA fragment length. Figure 2c shows the normalized autocorrelation curves and their fitting using the multiple-component model. As seen in Figure 2c, these autocorrelation curves are

We investigated the relationship between the lengths of DNA fragments and the characteristic diffusion times. Figure 2a shows the typical normalized autocorrelation curves and their fitting curves of DNA fragments using the single-component model. As seen in Figure 2a, these autocorrelation curves are well fitted with the theoretical mode of autocorrelation function as described in eq 1 with correlation coefficients (R2) of 0.972− 7354

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Figure 5. Autocorrelation curves and their fitting curves of DNA fragments in cell lysates by using single-component model (a) and multiplecomponent model (c), respectively. Their fitting residuals are shown in b and d. The inset of part a shows the relation between the characteristic diffusion times of DNA fragments and LDM concentrations. The inset of c shows the distributions of the characteristic diffusion times of DNA fragments in cell lysate under different concentrations of LDM. Other conditions are same as described in Figure 4.

also well fitted with the theoretical mode of autocorrelation function as described in eq 3. The inset of Figure 2c reflects the distribution of the characteristic diffusion times of DNA fragments, and the peak positions approximately express the average characteristic diffusion times of DNA fragments. The characteristic diffusion time of DNA fragments increased with an increase of the DNA fragment length, which is well in line with the results obtained by using the single-component model. The characteristic diffusion times of Rhodamine Green and 50 bp, 200 bp, 500 bp, 2.6 kbp, and 10 kbp dsDNA Markers fitted by single-component model were 0.036, 0.74, 1.10, 1.85, 3.13, and 5.24 ms, respectively, and the results fitted by multiplecomponent model were 0.033, 0.68, 1.03, 2.07, 3.14, and 6.29 ms, respectively. Figure 2d shows that the fitting residuals are about ±0.15. The results above document that FCS is an efficient and reliable method for characterization of DNA fragments. FCS Assay of Drug-Induced Apoptosis. In this study, LDM as a drug model was used to induce the apoptosis of human pancreatic cancer cells. LDM, also called C-1027, is a member of the enediyne antibiotic family, which is produced by Streptomyces globisporus C-1027.44 LDM is an acid protein containing 110 amino acid residues and a chromophore of enediyne structure,45 and shows significant anticancer activity due to its ability to damage DNA through radical-mediated hydrogen abstraction.46 Meanwhile, LDM shows marked cytotoxicity against cancers in vitro and in vivo as well.47,48 The drug-induced apoptosis was described in the Experimental Section, and a strategy for FCS assay of apoptosis is shown in Figure 3. First, we want to apply FCS to an assay of DNA fragments extracted from apoptotic cells. The total cellular DNA of normal and apoptotic cells were extracted according to the procedure described in the Experimental Section, and the DNA extracted from normal cells was used as a contrast

experiment (control). Figure 4a shows the autocorrelation curves and their fitting curves of DNA extracted from control cells and DNA extracted from apoptotic cells induced with different concentration of LDM, respectively. Although these autocorrelation curves can be fitted with the single-component model with correlation coefficients (R2) of 0.914−0.986, the fitting residuals as shown in Figure 4b are relatively large, especially in the DNA samples extracted from the control cells and from the apoptotic cells induced with 0.5 μg/mL of LDM. As shown in Figure 4a, the characteristic diffusion times of DNA extracted from apoptotic cells were dramatically different from that of the DNA extracted from normal cells, and they significantly decreased with an increase in the concentrations of induced drug (as shown in the inset of Figure 4a). The results implied that the DNA of apoptotic cells was gradually fragmented with increasing of induced-drug concentration. In order to improve the goodness of fit, the multiple-component model was used to fit the autocorrelation curves of DNA fragments, and the fitting results are shown in Figure 4c,d. These autocorrelation curves are well fitted with the theoretical mode of autocorrelation function as described in eq 3, and the fitting residuals as shown in Figure 4d are about ±0.1. The inset of Figure 4c reflects the distribution of the characteristic diffusion times of DNA fragments significantly decreased with an increase in the concentrations of induced drug, which are well in agreement with the results obtained by using the singlecomponent model. The characteristic diffusion times of 200 bp dsDNA marker, 1 kbp DNA ladder, 50 μg/mL, 5.0 μg/mL, 0.5 μg/mL LDM groups and control cell group obtained by singlecomponent model were 1.07, 3.24, 2.16, 5.14, 7.93, and 45.1 ms, respectively, and the results obtained by multiplecomponent model were 1.13, 3.27, 2.26, 4.89, 10.9, and 50.0 ms, respectively. The result above indicated that the goodness of fit in the multiple-component model was much better than 7355

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Figure 6. Assay results of flow cytometer (a) and DNA ladder assay (b). In assay of flow cytometer, the cytogram of four quadrants was used to distinguish the normal, primary apoptotic, late apoptotic, and necrotic cells by the criteria of FITC(−)/PI(−) (lower left), FITC(+)/PI(−) (lower right), FITC(+)/PI(+) (upper right), and FITC(−)/PI(+) (upper left), respectively. The Arabic numbers in the right corners indicate the percentages of primary apoptotic (lower side) and late apoptotic (upper side) states. The electrophoretic gel images of total cellular DNA extracted from LDM-treated PANC-1 cells and normal cultural PANC-1 cells. PANC-1 cells were treated with 0.5 μg/mL, 5.0 μg/mL, and 50 μg/mL LDM, respectively. The left lane contained a 1 kbp DNA Ladder as control DNA fragments.

induced cell lysates are clearly different from that of control cells. Although cell lysates are very complicated, FCS was successfully used for direct assay of DNA fragments in apoptotic cell lysates. This result demonstrates that FCS has a good selectivity. To further validate the reproducibility of FCS method, we have repeated experimental tests. The repeated test results are shown in Table S1 in Supporting Information. The RSDs of the characteristic diffusion times are about 6.8% to 14.4%, and the RSDs of lysates group are slightly higher than the extracted DNA group. These results document that our method is reliable for assay of drug-induced apoptosis. Contrast Assay of Apoptosis by Flow Cytometer and Gel Electrophoresis. In order to further confirm the reliability of our method, contrast assays were carried out by flow cytometry and gel electrophoresis. In the analysis of the flow cytometer, PS externalization and PI uptake in intact PANC-1 cells following LDM treatment were used to distinguish the apoptotic status of cells. The flow cytometer is used for measuring the PS level on cell membrane and distinguishing apoptosis and necrosis by combining Annexin V-dye with viability dyes such as propidium iodide (PI). Viable cells with intact membranes exclude PI, whereas the membranes of dead and damaged cells are permeable to PI. In analysis of flow cytometry, viable cells will show negative in both FITC Annexin V and PI detection channels, cells in early apoptosis show positive in FITC Annexin V channel and negative in PI channel, and cells in late apoptosis display positive in both FITC Annexin V and PI channels. The results obtained are shown in Figure 6a. Figure 6a shows a dot plot of four quadrants scaled with logarithm as fluorescence level of FITC labeled Annexin V (FL-1) and PI (FL-3), respectively. Control group cells showed the most integrity of plasma membrane without PS externalization. In contrast, 0.5 μg/mL LDM-

that in the single-component model. Moreover, these data illustrated that FCS was a highly sensitive tool for assay of druginduced apoptosis, and even in the presence of 0.5 μg/mL LDM, the characteristic diffusion time of DNA extracted from LDM treated cells was clearly different from that of control cells. It should be pointed out that we assume that the brightness of various components is identical in data processing, and in this case the distributions of the characteristic diffusion times of DNA fragments in Figures 4 and 5 probably were a departure from “the real distributions” of the characteristic diffusion times in a certain extent due to different brightnesses of various components. According to our experimental results, this assumption in the data processing did not affect our conclusions. The results above documented that FCS was successfully used for an assay of DNA fragments extracted from apoptotic cells. Next, we want to explore the possibility for direct FCS assay of DNA fragments in apoptotic cell lysates without DNA extraction. The procedure of drug-induced apoptosis was same as for the experiments above. The cell lysates were obtained by lysis of collected cells, which was described in the Experimental Section. Other experimental conditions were the same as the assay of the extracted total cellular DNA. Figure 5a,c displays the autocorrelation curves and their fitting curves of DNA fragments in cell lysates by using the single-component and multiple-component models, respectively. These autocorrelation curves are fitted with two models as well, and the fitting residuals are shown in Figure 5b,d. In this case, the assay results were similar to that of DNA extracted from apoptotic cells. As shown in the inset of Figure 5a, the characteristic diffusion times of DNA fragments in apoptotic cell lysates significantly decreased with an increase in the concentrations of induced drug. Even if in the presence of 0.5 μg/mL LDM, the characteristic diffusion times of DNA fragments in drug7356

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

Institute of Fundamental Research, India) for kindly providing the MEMFCS software.

treated cells showed the certain increased PS externalization, which indicated that apoptosis was induced by LDM. PS externalization significantly increased with an increase of LDM concentration. When LDM concentration was 50 μg/mL, the induced-cells showed the loss of plasma membrane integrity, which was a typical characteristic of late apoptosis stage or necrosis. The results of flow cytometry were in good agreement with the results of FCS assay above. The gel electrophoresis of DNA fragments is performed using a 1% agarose gel according to the procedure described in ref 49, and the results are shown in Figure 6b. The total cellular DNA from control and apoptotic cultures was extracted by using the same procedure in FCS analysis. A DNA ladder was used as a control in gel electrophoresis. As shown in Figure 6b, the sample from cells treated with 50 μg/mL LDM presents the shorter DNA band, and length range of DNA fragments is about 200−1000 bp, which explained that most cells were apoptosis in the presence of the high doses of LDM. A large band of DNA fragments is located at the top of the gel when LDM concentration is 0.5 μg/mL, and this result cannot be clearly distinguished from control group. The data above illustrates that DNA ladder assay is not sensitive as FCS and flow cytometry.

■

(1) Kerr, J. F.; Wyllie, A. H.; Currie, A. R. Br. J. Cancer 1972, 26, 239−257. (2) Martinez, M. M.; Reif, R. D.; Pappas, D. Anal. Methods 2010, 2, 996−1004. (3) Fackler, O. T.; Grosse, R. J. Cell Biol. 2008, 181, 879−884. (4) Van Cruchten, S.; Van Den Broeck, W. Anat., Histol., Embryol. 2002, 31, 214−223. (5) Zhang, J.; Xu, M. Trends Cell Biol. 2002, 12, 84−89. (6) Ihara, T.; Yamamoto, T.; Sugamata, M.; Okumura, H.; Ueno, Y. Virchows Arch. 1998, 433, 443−447. (7) Zhu, C. L.; Yang, Q.; Liu, L. B.; Wang, S. Chem. Commun. 2011, 47, 5524−5526. (8) Macho, A.; Decaudin, D.; Castedo, M.; Hirsch, T.; Susin, S. A.; Zamzami, N.; Kroemer, G. Cytometry 1996, 25, 333−340. (9) Schuler, M.; Bossy-Wetzel, E.; Goldstein, J. C.; Fitzgerald, P.; Green, D. R. J. Biol. Chem. 2000, 275, 7337. (10) Li, G. X.; Xiao, H.; Liu, L.; Meng, F. B.; Huang, J. Y. Anal. Chem. 2008, 80, 5272−5275. (11) Ito, Y.; Shibata, M. A.; Kusakabe, K.; Otsuki, Y. J. Histochem. Cytochem. 2006, 54, 683−692. (12) Vega Thurber, R.; Epel, D. Dev. Biol. 2007, 303, 336−346. (13) Brown, J. M.; Attardi, L. D. Nat. Rev. Cancer 2005, 5, 231−237. (14) Häcker, G. Cell Tissue Res. 2000, 301, 5−17. (15) van Engeland, M.; Nieland, L. J. W.; Ramaekers, F. C. S.; Schutte, B.; Reutelingsperger, C. P. M. Cytometry 1998, 31, 1−9. (16) Hug, H.; Los, M.; Hirt, W.; Debatin, K. M. Biochemistry 1999, 38, 13906−13911. (17) Zhang, J. H.; Xu, M. Trends Cell Biol. 2002, 12, 84−89. (18) Yasuhara, S.; Zhu, Y.; Matsui, T.; Tipirneni, N.; Yasuhara, Y.; Kaneki, M.; Rosenzweig, A.; Martyn, J. J. Histochem. Cytochem. 2003, 51, 873. (19) Wyllie, A.; Kerr, J.; Currie, A. Int. Rev. Cytol. 1980, 68, 251−306. (20) Pappas, D. Pracical Cell Analysis; Wiley Online Library: New York, 2010. (21) Kim, Y. E.; Chen, J.; Langen, R.; Chan, J. R. Nat. Protoc. 2010, 5, 1396−1405. (22) Gatti, R.; Belletti, S.; Orlandini, G.; Bussolati, O.; Dall’Asta, V.; Gazzola, G. C. J. Histochem. Cytochem. 1998, 46, 895−900. (23) Lu, W.; Yu, P.; Li, J. Biotechnol. Lett. 2011, 1−8. (24) Martinez, M. M.; Reif, R. D.; Pappas, D. Anal. Methods 2010, 2, 996−1004. (25) Pop, C.; Salvesen, G. S.; Scott, F. L. Methods Enzymol. 2008, 446, 351−367. (26) Tsang, W. P.; Kwok, T. T. Apoptosis 2008, 13, 1215−1222. (27) Nagata, S. Exp. Cell Res. 2000, 256, 12−18. (28) Salgame, P.; Varadhachary, A. S.; Primiano, L. L.; Fincke, J. E.; Muller, S.; Monestier, M. Nucleic Acids Res. 1997, 25, 680−681. (29) Ray, S. K.; Schaecher, K. E.; Shields, D. C.; Hogan, E. L.; Banik, N. L. Brain Res. Protoc. 2000, 5, 305−311. (30) Barbouti, A.; Doulias, P. T.; Nousis, L.; Tenopoulou, M.; Galaris, D. Free Radical Biol. Med. 2002, 33, 691−702. (31) Valero, A.; Merino, F.; Wolbers, F.; Luttge, R.; Vermes, I.; Andersson, H.; van den Berg, A. Lab Chip 2005, 5, 49−55. (32) Zhao, L.; Cheng, P.; Li, J.; Zhang, Y.; Gu, M.; Liu, J.; Zhang, J.; Zhu, J. J. Anal. Chem. 2009, 81, 7075−7080. (33) Zhang, J. J.; Zheng, T. T.; Cheng, F. F.; Zhang, J. R.; Zhu, J. J. Anal. Chem. 2011, 83, 7902−7909. (34) Wu, Y.; Zhou, H.; Wei, W.; Hua, X.; Wang, L.; Zhou, Z.; Liu, S. Anal. Chem. 2012, 84, 1894−1899. (35) Shimizu, M.; Sasaki, S.; Tsuruoka, M. Biomacromolecules 2005, 6, 2703−2707. (36) Kim, S. A.; Heinze, K. G.; Schwille, P. Nat. Methods 2007, 4, 963−973. (37) Zhou, X.; Tang, Y.; Xing, D. Anal. Chem. 2011, 83, 2906−2912.

■

CONCLUSION In this paper, we presented a novel method for sensitive detection of drug-induced apoptosis based on the FCS analysis of DNA fragmentation in apoptotic cells. In FCS measurement, the characteristic diffusion time of DNA fragments was used as a key parameter to distinguish the apoptotic status of cells. The single-component model and the multiple-components model were used to fit raw FCS data. The results indicated that the multiple-component model was better than the singlecomponent model in FCS assay of apoptosis. The established FCS method was successfully used for assay of drug-induced apoptosis using extracted cellular DNA and cell lysates as samples. The results of FCS method were consistent with those of gel electrophoresis and flow cytometer. Our data showed that the sensitivity of FCS method was similar to flow cytometer, but was better than gel electrophoresis. More importantly, FCS is a single molecule method, and the sample requirement is very small. We believe that FCS may become a robust and simple method for assay of cell apoptosis and screening of induced apoptosis drugs.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: +86-21-54746001. Fax: +86-21-54741297. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by NSFC (21135004 and 21075081), National Basic Research Program of China (2009CB930400), and Nano-Science Foundation of Shanghai (1052 nm04000). We also greatly thank Professor Maiti (Tata 7357

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358

Analytical Chemistry

Article

(38) Martinez, M. M.; Reif, R. D.; Pappas, D. Anal. Bioanal. Chem. 2010, 396, 1177−1185. (39) Dong, M.; Martinez, M. M.; Mayer, M. F.; Pappas, D. Analyst 2012, 137, 2997−3003. (40) Sengupta, P.; Garai, K.; Balaji, J.; Periasamy, N.; Maiti, S. Biophys. J. 2003, 84, 1977−1984. (41) Sengupta, P.; Garai, K.; Sahoo, B.; Shi, Y.; Callaway, D. J.; Maiti, S. Biochemistry 2003, 42, 10506−10513. (42) Pal, N.; Dev Verma, S.; Singh, M. K.; Sen, S. Anal. Chem. 2011, 83, 7736−7744. (43) Chen, J.; Irudayaraj, J. ACS Nano 2009, 3, 4071−4079. (44) Hu, J. L.; Xue, Y. C.; Xie, M. Y.; Zhang, R.; Otani, T.; Minami, Y.; Yamada, Y.; Marunaka, T. J. Antibiot. 1988, 41, 1575−1579. (45) Tanaka, T.; Fukuda-Ishisaka, S.; Hirama, M.; Otani, T. J. Mol. Biol. 2001, 309, 267−283. (46) Shao, R. G.; Zhen, Y. S. Anticancer Agents Med. Chem. 2008, 8, 123−131. (47) Huang, Y. H.; Shang, B. Y.; Zhen, Y. S. World J. Gastroenterol. 2005, 11, 3980−3984. (48) Chen, J.; Ouyang, Z. G.; Zhang, S. H.; Zhen, Y. S. Oncol. Rep. 2007, 17, 1445−1451. (49) Trauth, B. C.; Klas, C.; Peters, A.; Matzku, S.; Moller, P.; Falk, W.; Debatin, K. M.; Krammer, P. H. Science 1989, 245, 301−305.

7358

dx.doi.org/10.1021/ac301654g | Anal. Chem. 2012, 84, 7350−7358