Bioconjugate Chem. 2004, 15, 983−987
983
Detection of Apoptosis Using the C2A Domain of Synaptotagmin I Hyo-il Jung,† Mikko I. Kettunen, Bazbek Davletov,‡ and Kevin M. Brindle* Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1GA, U.K. Received April 24, 2004; Revised Manuscript Received July 8, 2004
Binding of annexin V or the C2A domain of synaptotagmin I to phosphatidylserine expressed on the surface of apoptotic cells can, when labeled with appropriate probe molecules, be used to detect the presence of apoptosis using radionuclide, magnetic resonance, and optical imaging techniques. The preparation of a biotinylated C2A-GST fusion protein is described, and its capability, when used in conjunction with fluorescein-labeled streptavidin, of detecting apoptotic cells by flow cytometry is compared directly with the performance of a commercial preparation of fluorescein-labeled annexin V. Biotinylated C2A-GST, when used in conjunction with streptavidin-conjugated superparamagnetic iron oxide nanoparticles or Gd-chelate-avidin conjugates, was shown to be capable of detecting apoptotic cells using T2-weighted or T1-weighted magnetic resonance imaging experiments, respectively.
INTRODUCTION
Apoptosis, or programmed cell death, is a highly regulated process that results in the selective elimination of cells during normal tissue homeostasis and differentiation (1). Apoptosis can also occur in tumors following chemotherapy, where the extent of apoptosis and the speed of onset have been shown to be good prognostic indicators for the outcome of treatment (2-5). For this reason we have been interested in identifying markers of tumor cell apoptosis that could be detected noninvasively in the clinic using magnetic resonance spectroscopy and imaging techniques (6, 7). There are a number of markers that can be used to identify apoptotic cells, including morphological changes, DNA fragmentation, and caspase activation (8). A relatively early event in the process is the flipping of phosphatidylserine from the inner leaflet of the plasma membrane bilayer to the outer surface (9), which can be detected by binding of the protein annexin V to surface phosphatidylserine (10). This has led to the development of a number of annexin V based probes for detecting apoptotic cells, including technetium-labeled (11) and near-infrared fluorochromelabeled-proteins (12), which have allowed the detection of apoptotic cells in vivo using radionuclide and optical imaging techniques, respectively. We have demonstrated previously that the C2A domain of the protein synaptotagmin I, which binds to negatively charged phospholipids in a Ca2+-dependent manner, can be used to detect apoptotic cells in vitro using flow cytometry. We also showed that when labeled with superparamagnetic iron oxide nanoparticles (SPIO)1 the protein domain could be used to detect apoptotic cells, in vitro and in vivo, using T2-weighted magnetic resonance imaging experiments (13). We describe here the synthesis of a biotinylated protein that, when used in conjunction with appropriately labeled avidin or streptavidin molecules, which bind very * To whom correspondence should be addressed. Phone: +44 (0)1223 333674. Fax: +44 (0)1223 766002. E-mail: kmb@ mole.bio.cam.ac.uk. † Present address: Department of Mechanical Engineering, Yonsei University, South Korea. ‡ Neurobiology Division, MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH.
tightly to biotin (Kd ≈ 10-15 M), can be used to detect apoptotic cells by flow cytometry or by T1- or T2-weighted 1 H magnetic resonance imaging. Flow cytometric detection of apoptotic cells using C2A was compared with detection using a commercial preparation of the more widely used annexin V. EXPERIMENTAL PROCEDURES
Protein Purification. E. coli BL21(DE3) cells transformed with a pGEX vector (Amersham Biosciences) expressing a C2A-GST fusion protein (14) were grown at 37 °C in 2xTY medium (2 L) supplemented with 50 µg/mL ampicillin until the A600 was between 0.5 and 0.8. The cells were then induced with IPTG (final concentration of 0.1 mM) and incubated for a further 3 h before harvesting by centrifugation at 6000g for 30 min. The cells were resuspended in 40 mL of BugBuster protein extraction reagent (Novagen) and incubated for 20 min at 25 °C. Cell debris was removed by ultracentrifugation at 20000g for 30 min, the supernatant was applied to glutathione sepharose 4B beads (Amersham Biosciences) preequilibrated with phosphate-buffered saline (PBS), and the protein was purified according to the manufacturer’s instructions. The protein, which ran as a single band in sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), was dialyzed against PBS and concentrated using a Vivaspin filter (Vivascience, Germany). The protein can be produced in very high yield using this system (∼40 mg/L E. coli culture), which is an advantage given the relatively large amounts required for magnetic resonance (MR) imaging experiments. The C2A domain was prepared by thrombin cleavage of the fusion protein and purified using glutathione 1 SPIO, superparamagnetic iron oxide; C2A, C2A domain of synaptotagmin I; GST, glutathione S-transferase; PS, phosphatidylserine; PC, phosphatidylcholine; DTPA, diethylenetriaminepentaacetate; HBS, HEPES-buffered saline; SulfoNHSLC-biotin, sulfosuccinimidyl-6-(biotinamido) hexanoate; PI, propidium iodide; MALDI-TOF MS, matrix assisted laser desorption ionization time-of-flight mass spectrometry; NTA, N,N bis(carboxymethyl)glycine; RU, response unit; SPR, surface plasmon resonance; MRI, magnetic resonance imaging; SDSPAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis.
10.1021/bc049899q CCC: $27.50 © 2004 American Chemical Society Published on Web 08/12/2004
984 Bioconjugate Chem., Vol. 15, No. 5, 2004
sepharose 4B beads. Briefly, 400 µg of C2A-GST fusion protein were mixed with 10 units of thrombin in PBS and incubated for 20 h at 25 °C. Glutathione sepharose 4B beads were added to the reaction mixture and incubated for a further 10 min before the solution was centrifuged at 300g for 5 min to collect the supernatant, which contained the C2A domain. The bead-bound GST was eluted using a reduced glutathione solution. Surface Plasmon Resonance Analysis. Liposomes were prepared from brain phospholipids (Avanti Polar Lipids Co.). Samples containing 2.6 µmol of a PS/PE/PC mixture (molar ratio 35:50:15) in chloroform were dried under a stream of nitrogen and resuspended in 2.6 mL of buffer (20 mM HEPES, 100 mM NaCl, pH 7.4) by vigorous vortexing. A milky suspension was created, indicating the formation of large multilamellar vesicles (LMV). Small unilamellar vesicles (SUV) were obtained by sonication of the LMV until the solution cleared. To determine the binding affinity of the C2A domain for the liposomes, surface plasmon resonance analysis was performed using a Biacore 2000 system (Biacore, Uppsala, Sweden) and L1 chips. The sensor chips were coated with 40 µL of a suspension of SUV containing 1 mM phospholipid at a flow rate of 5 µL/min. The resulting response was about 3000 RU. To remove any multilamellar material from the lipid surface, 5 µL of 50 mM NaOH was injected, which resulted in a stable baseline (∼2000 RU) corresponding to the lipid bilayer linked to the chip surface. The C2A domain was allowed to interact with the sensor surface for 10 min (association phase) at five different concentrations, increasing from 150 to 550 nM. All the experiments were conducted in a buffer containing 10 mM HEPES, pH 7.4, 150 mM NaCl, and 2 mM CaCl2 at a flow rate of 5 µL/min and at 25 °C and were repeated three times. To regenerate the phospholipid bilayer surface after the dissociation phase, 10 µL of 50 mM NaOH solution was used. The binding data were evaluated using BIAevaluation software, version 3.1 (BIAcore AB, Uppsala, Sweden). Biotinylation of the C2A-GST Fusion Protein. The protein was biotinylated using an N-hydroxysulfosuccinimide (NHS) ester derivative of biotin, which modifies primary amine groups. Since the Ca2+-dependent PS binding sites on C2A contain two Lys residues (Lys200 and Lys236) (20), we sought to protect these by first binding the protein to an SP sepharose column. The protein binds tightly to the negatively charged SP sepharose resin in a Ca2+-dependent manner, via binding sites that likely overlap those for PS. Purified C2A-GST (6 mg/mL, 1 mL) was loaded onto an SP column (HiTrap SP FF, 1 mL volume, Amersham Biosciences) equilibrated with HEPES-buffered saline containing 20 mM HEPES, pH 7.4, 150 mM NaCl, and 0.2 mM CaCl2. After extensive washing, sulfoNHS-LC-biotin (Pierce) (1.8 mg/mL, 1 mL, 40-fold excess of biotin) in the same buffer was injected into the column and the flow-through fraction was collected and reinjected. After four cycles of reinjection, the column was left at 25 °C for 15 min and then washed with 10 bed volumes of HEPES-Ca2+ buffer. The biotinylated protein was eluted using buffer lacking CaCl2 and containing 10 mM EDTA. The collected fractions were pooled and dialyzed against PBS for 24 h to remove EDTA. The biotinylation sites in C2A-GST were investigated by cleaving the fusion protein at the thrombin cleavage site. To estimate the number of moles of biotin conjugated per mole of C2A-GST and to identify the region of the protein modified (C2A versus GST), the biotin-conjugated protein (0.4 mL of a 1 mg/mL solution in PBS) was mixed
Jung et al.
with a thrombin solution (10 µL of a 1 U/µL solution in PBS) and incubated at 25 °C for 20 h. The resulting mixture of two protein fragments and the intact fusion protein were analyzed by SDS-PAGE and MALDI-TOF mass spectrometry. The binding activity of biotinylated C2A-GST for PS was assessed using a liposome binding assay described previously (21). Briefly, the biotinylated protein was mixed with the liposomes and then centrifuged. Flow Cytometry. Murine lymphoma cells (EL4) were grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 U/mL penicillin, and 0.1 mg/mL streptomycin. The cells were treated with 15 µM etoposide for 16 h to induce apoptosis, collected by centrifugation, and then washed in ice-cold PBS. The washed cells were suspended in C2Abinding buffer (10 mM HEPES, 140 mM NaCl, 2.5 mM CaCl2, pH 7.4), and the cell density was adjusted to 1 × 106 cells/mL. To 100 µL of cell suspension, 1 µL of biotinC2A-GST protein (30 µM protein), 1 µL of propidium iodide (PI) (100 µg/mL), and 5 µL of streptavidin-FITC (Sigma) (1 mg/mL) were added, and the mixture was incubated for 5 min at 25 °C. The resulting mixture was then subjected to flow cytometry (FACScan, Becton Dickinson, U.K.). For each sample 20 000 cells were counted. Apoptosis was also detected using a commercially available preparation of annexin V-FITC (Molecular Probes), which was used according to the manufacturer’s instructions. Preparation of Contrast Agents and Magnetic Resonance Imaging. Avidin-GdDTPA conjugates were prepared as described in (15). Briefly, purified avidin (Sigma) was incubated with 20-fold excess DTPA anhydride (Sigma) for 24 h at 4 °C. After extensive dialysis against 100 mM sodium citrate buffer, pH 6.5, a 40-fold excess of Gd(NTA)2 was added to the avidin-DTPA solution. After incubation for 24 h at 4 °C, the resulting mixture was dialyzed extensively against PBS and stored at -20 °C before use. The T1 relaxivity of this modified protein at 9.4 T was about 60 mM-1 s-1. For T2-weighted 1 H MR imaging, commercially available strepavidinconjugated SPIO nanoparticles were used (Miltenyi Biotec). Etoposide-treated cells were incubated with different concentrations (1.5, 4.5, 15 µM) of the biotinylated C2AGST protein at 25 °C for 5 min. The protein concentration was determined by amino acid analysis. After being washed, cells were incubated with avidin-GdDTPA conjugates (80 µM protein) or streptavidin-SPIO nanoparticles (30 ng/mL of Fe) at 25 °C for 5 min. After washing, 107 cells were mixed with 50 µL of 4% bovine gelatin and then immediately loaded into a 5 mm NMR tube. T1- (repetition time, TR ) 0.2 s; spin-echo time, TE ) 15 ms) and T2- (TR ) 4 s, TE ) 40 ms) weighted 1 H MR images of the samples were acquired at 9.4 T using a two-dimensional spin-echo imaging sequence, a 25 mm diameter proton probe, and a Varian UnityPlus spectrometer. RESULTS
Measurement of the Affinity of C2A for Phosphatidylserine-Containing Liposomes Using Surface Plasmon Resonance. Small unilamellar vesicles were immobilized on an L1 sensor chip, which allows retention of the lipid bilayer structure (16). A representative sensorgram is shown in Figure 1, showing the association of the C2A domain with phosphatidylserine on the surface of the sensor chip and its subsequent dissociation. The C2A domain bound to the surface of the
Detection of Apoptosis with Synaptotagmin
Bioconjugate Chem., Vol. 15, No. 5, 2004 985
Figure 1. SPR sensorgram (relative response vs time) for the C2A domain. The L1 sensor chip was coated with PS-containing liposomes, and the C2A domain was injected at a series of increasing concentrations (150, 250, 350, 450, and 550 nM). The start of the association and dissociation phases are indicated by the symbols a and d, respectively.
Figure 2. Analysis of the biotinylation of C2A-GST by SDSPAGE: (1) size markers; (2) intact C2A-GST; (3) biotinylated C2A-GST; (4) thrombin-cleaved C2A-GST; (5) thrombin-cleaved biotinylated C2A-GST; (6) size markers. The relative molecular mass of GST is 26 kDa, and that of the C2A domain is 16 kDa. Both the GST and C2A bands from the biotinylated protein are retarded when compared to the unmodified protein, indicating that both domains of the fusion protein are biotinylated.
sensor chip in the presence of 2 mM CaCl2 but had no affinity when free Ca2+ was depleted by the addition of 10 mM EDTA (data not shown). The calculated dissociation constant is in the nanomolar range ((17.1 ( 0.5) × 10-9 M) and comparable to a value of 115 × 10-9 M for the binding of C2A-GST to a lipid monolayer (17) determined by surface plasmon resonance and to a value of 40 × 10-9 M for the binding of C2A to phospholipid vesicles (25% PS/75% PC) determined using intrinsic fluorescence measurements (18). However, the affinity of C2A for phosphatidylserine-containing membranes is lower than that of annexin V, for which a dissociation constant of
