NANO LETTERS
A Fluorescent Nanosensor for Apoptotic Cells
2006 Vol. 6, No. 3 488-490
Luisa Quinti, Ralph Weissleder, and Ching-Hsuan Tung* Center for Molecular Imaging Research, Massachusetts General Hospital, HarVard Medical School, 149 13th Street, Charlestown, Massachusetts 02129 Received December 14, 2005; Revised Manuscript Received January 25, 2006
ABSTRACT A biocompatible surface-functionalized nanoparticle was designed to sense phosphatidylserine exposed on apoptotic cells. We conjugated synthetic artificial phosphatidylserine binding ligands in a multivalent fashion onto magnetofluorescent nanoparticles. Our results show that (1) the synthetic nanoparticles bind to apoptotic cells, (2) there is excellent correlation with annexin V staining by microscopy, and (3) FACS analysis with nanoparticles allows the measurement of therapeutic apoptosis induction. The described nanomaterials should be useful for a variety of biomedical applications including in vivo imaging of apoptosis.
Designing synthetic nanomaterials that imitate protein function is of great interest in biomedical sensing. In this work, we designed a mimetic to Annexin V (AnxV), a 37-kDa protein that recognizes phosphatidylserine (PS) expressed on apoptotic cell surfaces. Phospholipids are distributed asymmetrically between the inner and outer layers of the plasma membrane. In healthy cells, phosphatidylcholine (PC) and sphingomyelin are exposed on the external leaflet of the lipid bilayer, whereas PS is predominantly located on the inner layer, facing the cytosol. When cells undergo apoptosis, redistribution of phospholipids occurs and PS translocates to the outer layer of the membrane. In the presence of Ca(II) ions, AnxV reversibly binds to the outer plasma membrane of apoptotic cells.1 Because of this, fluorescent annexins are used routinely in fluorescence microscopy and flow cytometry to identify apoptotic cells.2,3 Recently, radiolabeled AnxV has also been tested in patients to image myocardial diseases and cancer therapy4,5 in which apoptosis plays important roles. An alternative way of sensing the phosphate moiety on PS is with Zn(II) di-2-picolylamine (DPA) complexes attached to a fluorochrome.6-8 Reporters with two Zn(II)DPA units were previously found to bind to PS more strongly than those with one Zn(II)-DPA. Furthermore, two Zn(II)DPA units were shown to cooperatively bind the same phosphate moiety.7 Some of these compounds were further applied to detect surface PS in synthetic vesicles and cells.9-11 We hypothesized that multivalent display of such PS binding units would further enhance the sensor affinity and could provide new synthetic nanomaterials with tunable charac* Corresponding author address: Center for Molecular Imaging Research, Massachusetts General Hospital, 149 13th Street, Room 5406, Charlestown, MA02129.Tel: 617-726-5779.Fax: 617-726-5708.E-mail:
[email protected]. 10.1021/nl0524694 CCC: $33.50 Published on Web 02/03/2006
© 2006 American Chemical Society
teristics. Therefore, we attached Zn(II)-DPA units onto a model nanoparticle, which is similar in structure to a clinically used magnetic nanoparticle.12 To facilitate the assembly of multiple phosphate-binding DPA units, we chose a peptide template (Supporting Information). After initial evaluation, an unnatural amino acid, diaminopropionic acid (Dpr), was selected as the building unit of the PS-binding peptide. On each peptide, four DPA units were appended to the side chain of Dpr residues to increase the binding affinity. The N-terminus of the resulting peptide, Gly-Dpr(DPA)-Dpr(DPA)-Gly-Dpr(DPA)-Dpr(DPA)-Gly-CONH2, was conjugated with a fluorescent tag, TAMRA, resulting in T-mDpr4 (MW: [M + H]+ ) 1674.9 calcd.; 1675.0 found) (Figure 1A). To test the preferential binding ability of T-mDpr4 to PS over PC, we mixed aqueous solutions of T-mDpr4 complexed with Zn(II) ions with either PC or PS in chloroform. As shown in Figure 1B, in the case of the CHCl3 control and PC in CHCl3, the pink color representing T-mDpr4 remained in the buffer (upper) layer. With PS in CHCl3, T-mDpr4 moved to the CHCl3 (lower) layer. This was further confirmed by exposing the vials to UV light and observing fluorescence emission in the CHCl3 layer. T-mDpr4 translocated from water into chloroform because the Zn(II)-DPA units coordinated with the negatively charged PS. In contrast, the formation of a complex between T-mDpr4 and PC is disfavored because PC has an overall neutral charge. To quantify the distribution of T-mDpr4, we mixed solutions of TAMRA and T-mDpr4 with increasing concentrations of PS in CHCl3, from 0 to 200 µM (Figure 1C). The fluorescence remaining in the buffer layer was measured. TAMRA did not translocate into the CHCl3 layer; thus, its fluorescence stayed constant during the titration with PS. The fluorescence
Figure 1. (A) Schematic representation of the phosphatidylserine-binding sequences: T-mDpr4 and mDpr4Cys. The latter was used for conjugation to nanoparticles. (B) Distribution of T-mDpr4 in various phospholipid solutions. T-mDpr4 (4 µM) was dissolved in 10 mM Hepes buffer. In each vial, the upper layer is Hepes buffer and the lower layer consists of (from left to right): CHCl3; 25 mM PC in CHCl3; or 25 mM PS in CHCl3. (C) Titration curve of TAMRA and T-mDpr4 with PS. Fluorescence of the aqueous layer was measured at 580 nm. The concentration of TAMRA and T-mDpr4 was 4 µM in 10 mM Hepes. The data for each compound were normalized to its intrinsic fluorescence in the absence of PS.
of T-mDpr4, however, decreased with additional PS, indicating its translocation to the CHCl3 layer. The titration reached equilibrium at 20 equiv (50 µM) of PS to T-mDpr4, with only ca. 10% of the original fluorescence remaining in the aqueous layer. After demonstrating that T-mDpr4 recognized PS-rich solutions, its ability of sensing apoptotic cells was assessed. However, no significant binding of free T-mDpr4 to apoptotic cells could be detected (data not shown); thus, we attached multiple PS binding ligands to a nanoparticle to obtain multivalency (Figure 2A). A dextran-coated iron-oxide nanoparticle was chosen because its properties are well studied and it has been applied in various in vitro and in vivo applications.13,14 On average, 41 amino groups were available per nanoparticle for chemical conjugation (Supporting Information). For the purpose of conjugation, a cysteine residue was added to the C terminus of the mDpr4 peptide (Figure 1A). The mDpr4Cys-conjugated nanoparticle was synthesized in three steps: (i) TAMRA labeling of the nanoparticle; (ii) activation of the amino groups on the nanoparticle surface with SPDP; and (iii) conjugation of mDpr4Cys to the activated nanoparticle via the Cys-sulfhydryl moiety (details in the Supporting Information). The final multivalent particle, T-P-mDpr4Cys, contained, on average, 1 TAMRA molecule and 15 mDpr4Cys peptides. The ability of T-P-mDpr4Cys to distinguish apoptotic and viable cells was assessed by both FACS and microscopy analysis on Jurkat T cells. During flow cytometry studies, Jurkat cells were incubated with T-P-mDpr4Cys and then costained with FITC-labeled AnxV (AnxV-FITC) and 7-aminoactinomycin D (7-AAD). Figure 2B shows the dual staining with T-P-mDpr4Cys and AnxV-FITC of Jurkat cells, before and after induction of apoptosis with staurosporine. Necrotic cells (7-AAD positive cells) were excluded from Nano Lett., Vol. 6, No. 3, 2006
the FACS analysis. Upon treatment with staurosporine, the number of cells costained by AnxV-FITC and T-P-mDpr4Cys increased from 4 to 67% and the TAMRA mean fluorescence intensity went from 7 to 33 AU (Figure 2B, inset). About 80% of the staurosporine-treated cells were AnxV-FITC positive, and around 84% of these AnxV positive cells were also T-P-mDpr4Cys positive. These results support that T-P-mDpr4Cys has a binding property similar to AnxV because both of them label apoptotic cells. Cellular distribution of T-P-mDpr4Cys was investigated using fluorescence microscopy. Jurkat T cells were costained with Hoechst 33342, AnxV-FITC, and T-P-mDpr4Cys. After induction of apoptosis, AnxV-FITC positive cells were also found to be T-P-mDpr4Cys positive (Figure 2C). Confocal microscopy analysis of the apoptotic cells at 100X magnification revealed that both T-P-mDpr4Cys and AnxV-FITC had colocalized on the cell surface, as expected from binding to PS. In contrast, healthy cells showed lack of staining by T-P-mDpr4Cys and AnxV-FITC (Supporting Information). To further confirm that T-P-mDpr4Cys recognizes PS exposed on the apoptotic-cell surface, we performed a competition experiment with AnxV. Following staurosporine treatment, Jurkat T cells were first incubated with T-P-mDpr4Cys and then with AnxV. Upon the addition of increasing amounts of AnxV, the mean fluorescence intensity of T-PmDpr4Cys decreased progressively as quantified by flow cytometry (Supporting Information). The IC50 value of 2.52 nM for AnxV displacing T-P-mDpr4Cys, calculated from these experiments (Figure 3), was in agreement with data reported previously.14 In summary, we have developed a peptide-based PS binding ligand for detection of apoptotic cells. Although T-mDpr4 was shown to distinguish phospholipid solutions of PS over PC, its binding to apoptotic cells was not 489
Figure 3. Competition assay showing the displacement of T-PmDpr4Cys by AnxV obtained by FACS analysis.
sensing, the superparamagnetic nature of the T-P-mDpr4Cys iron oxide core could be utilized for magnetic resonance imaging in vivo.12 We furthermore envision that the presented synthetic strategy will be applicable to a wide variety of other nanoparticles to improve sensing. Acknowledgment. We thank the CMIR Chemistry Core for providing the iron oxide nanoparticle and Dr. Benedict Law for helpful discussions. This research was supported in part by NIH P50-CA86355, RO1 CA99385, and R21 CA114149. Supporting Information Available: Experimental details of synthesis and assays. This material is available free of charge via the Internet at http://pubs.acs.org. References
Figure 2. (A) Schematic representation of T-P-mDpr4Cys particles and their recognition of phospholipids rich in PS. (B) Dual wavelength FACS analysis with AnxV-FITC (FL1) and T-P-mDpr4Cys (FL2) of Jurkat T cells untreated (left) and treated with Staurosporine 3 µM for 4.5 h (right). The percentage of cells in each quadrant is shown. FL2 fluorescence histograms are reported in the insets. All of the cells shown are 7-AAD negative. (C) Microscopic images of staurosporine-treated Jurkat T cells. Top: 20× images of Jurkat T cells stained with Hoechst (blue), AnxVFITC (green), and T-P-mDpr4Cys (red); Bottom: confocal images (100× magnification) of a Jurkat T cell stained with AnxV-FITC (green) and T-P-mDpr4Cys (red), plus the merged image. Bar ) 10 µm.
detectable (data not shown), presumably because of weak affinity in the monomeric state. In contrast, when multiple T-mDpr4 ligands were conjugated to a nanoparticle template, the resulting nanoparticle, T-P-mDpr4Cys, bound preferentially to apoptotic cells. Specifically, our data support that T-P-mDpr4Cys mimics the recognition properties of AnxV toward apoptotic cells. T-P-mDpr4Cys should be useful for in vivo sensing because of the optimized pharmacokinetic properties of the core nanoparticle. In addition to optical
490
(1) Kuypers, F. A.; Lewis, R. A.; Hua, M.; Schott, M. A.; Discher, D.; Ernst, J. D.; Lubin, B. H. Blood 1996, 87, 1179-87. (2) Vermes, I.; Haanen, C.; Steffens-Nakken, H.; Reutelingsperger, C. J. Immunol. Methods 1995, 184, 39-51. (3) Koopman, G.; Reutelingsperger, C. P.; Kuijten, G. A.; Keehnen, R. M.; Pals, S. T.; van Oers, M. H. Blood 1994, 84, 1415-20. (4) Flotats, A.; Carrio, I. Eur. J. Nucl. Med. Mol. Imaging 2003, 30, 615-30. (5) Belhocine, T.; Steinmetz, N.; Li, C.; Green, A.; Blankenberg, F. G. Technol. Cancer Res. Treat. 2004, 3, 23-32. (6) Ojida, A.; Mito-Oka, Y.; Inoue, M. A.; Hamachi, I. J. Am. Chem. Soc. 2002, 124, 6256-8. (7) Ojida, A.; Mito-oka, Y.; Sada, K.; Hamachi, I. J. Am. Chem. Soc. 2004, 126, 2454-63. (8) Hanshaw, R. G.; Smith, B. D. Bioorg. Med. Chem. 2005, 13, 503542. (9) Koulov, A. V.; Stucker, K. A.; Lakshmi, C.; Robinson, J. P.; Smith, B. D. Cell Death Differ. 2003, 10, 1357-9. (10) Hanshaw, R. G.; O’Neil, E. J.; Foley, M.; Carpenter, R. T.; Smith, B. D. J. Mater. Chem. 2005, 15, 2707-13. (11) Hanshaw, R. G.; Lakshmi, C.; Lambert, T. N.; Johnson, J. R.; Smith, B. D. Chembiochem 2005, 6, 2214-20. (12) Harisinghani, M. G.; Barentsz, J.; Hahn, P. F.; Deserno, W. M.; Tabatabaei, S.; van de Kaa, C. H.; de la Rosette, J.; Weissleder, R. N. Engl. J. Med. 2003, 348, 2491-9. (13) Lewin, M.; Carlesso, N.; Tung, C. H.; Tang, X. W.; Cory, D.; Scadden, D. T.; Weissleder, R. Nat. Biotechnol. 2000, 18, 410-4. (14) Schellenberger, E. A.; Sosnovik, D.; Weissleder, R.; Josephson, L. Bioconjugate Chem. 2004, 15, 1062-7.
NL0524694
Nano Lett., Vol. 6, No. 3, 2006
