pubs.acs.org/Langmuir © 2010 American Chemical Society
)
)
Single-Step Separation of Platelets from Whole Blood Coupled with Digital Quantification by Interfacial Platelet Cytometry (iPC) L. Basabe-Desmonts,†, S. Ramstrom,‡, G. Meade,‡ S. O’Neill,‡ A. Riaz,† L. P. Lee,*,†,§ A. J. Ricco,*,† and D. Kenny*,‡ Biomedical Diagnostics Institute (BDI), Dublin City University, Dublin, Ireland, ‡BDI Programme, Molecular & Cellular Therapeutics, Royal College of Surgeons in Ireland (RCSI), Dublin, Ireland, and §Biomolecular Nanotechnology Center, Berkeley Sensor & Actuator Center, Department of Bioengineering, University of California, Berkeley, California. These two authors contributed equally to this work. )
†
Received October 20, 2009. Revised Manuscript Received January 8, 2010 We report the efficient single-step separation of individual platelets from unprocessed whole blood, enabling digital quantification of platelet function using interfacial platelet cytometry (iPC) on a chip. iPC is accomplished by the precision micropatterning of platelet-specific protein surfaces on solid substrates. By separating platelets from whole blood using specific binding to protein spots of a defined size, iPC implements a simple incubate-and-rinse approach, without sample preparation, that enables (1) the study of platelets in the physiological situation of interaction with a protein surface, (2) the choice of the number of platelets bound on each protein spot, from one to many, (3) control of the platelet-platelet distance, including the possibility to study noninteracting single platelets, (4) digital quantification (counting) of platelet adhesion to selected protein matrices, enabling statistical characterization of platelet subpopulations from meaningfully large numbers of single platelets, (5) the study of platelet receptor expression and spatial distribution, and (6) a detailed study of the morphology of isolated single platelets at activation levels that can be manipulated. To date, we have demonstrated 1-4 of the above list. Platelets were separated from whole blood using iPC with fibrinogen, von Willebrand factor (VWF), and anti-CD42b antibody printed “spots” ranging from a fraction of one to several platelet diameters (2-24 μm). The number of platelets captured per spot depends strongly on the protein matrix and the surface area of the spot, together with the platelet volume, morphology, and activation state. Blood samples from healthy donors, a May-Hegglin-anomaly patient, and a Glanzmann’s Thrombasthenia patient were analyzed via iPC to confirm the specificity of the interaction between protein matrices and platelets. For example, the results indicate that platelets interact with fibrinogen spots only through the fibrinogen receptor (RIIbβ3) and, relevant to diagnostic applications, platelet adhesion correlates strongly with normal versus abnormal platelet function. A critical function of platelets is to adhere to regions of damage on blood vessel walls; in contrast to conventional flow cytometry, where platelets are suspended in solution, iPC enables physiologically relevant platelet bioassays based on platelet/protein-matrix interactions on surfaces. This technology should be inexpensive to implement in clinical assay format, is readily integrable into fluidic microdevices, and paves the way for high-throughput platelet assays from microliter volumes of whole blood.
1. Introduction Platelets are involved in many vital physiological processes, hence platelet function studies are important. Circulating platelets play a critical role in hemostasis, adhering and aggregating at sites of vascular injury to initiate thrombus formation and staunch bleeding. Platelets also play a central role in the pathogenesis of arterial thrombosis, leading to the clinical events associated with cardiovascular disease.1 Additionally, they are involved in the hematogenous spread of cancer cells during the metastatic cascade2 and in inflammatory diseases.3 Each person has a heterogeneous platelet population, with the platelets differing in size, age, and function; an individual’s platelet distribution may have diagnostic or predictive value. For example, only a small fraction of the platelet population, known as COAT platelets, express phosphatidylserine upon coactivation with collagen and thrombin. The size of this COAT *Address correspondence to any of these authors. E-mail:
[email protected],
[email protected],
[email protected]. (1) Michelson, A. D. Platelets; Academic Press: New York, 2006. (2) Jurasz, P.; Sawicki, G.; Duszyk, M.; Sawicka, J.; et al. Cancer Res. 2001, 61, 376–382. (3) Kulkarni, S.; Woollard, K. J.; Thomas, S.; Oxley, D.; et al. Blood 2007, 110, 1879–1886.
14700 DOI: 10.1021/la9039682
platelet subpopulation has been suggested to be clinically informative.4-7 Furthermore, a high mean platelet volume has been proposed as a marker for cardiovascular risk.8,9 To identify different platelet subpopulations and characterize their physical and biological properties, single-cell analysis is necessary using statistically meaningful numbers of platelets from an individual. At present, large-scale studies of individual platelets are conducted mainly by flow cytometry,10 which enables the characterization of platelet size and receptor expression using light scattering and fluorescent reporters, respectively; 10 000 single platelets are typically analyzed. Flow cytometry, however, measures platelet activation only in fluid suspension, a nonphysiological situation in the context of key platelet-protein (4) Prodan, C. I.; Ross, E. D.; Vincent, A. S.; Dale, G. L. J. Neurol. 2007, 254, 548–549. (5) Prodan, C. I.; Ross, E. D.; Vincent, A. S.; Dale, G. L. Alzheimer Dis. Assoc. Disord. 2007, 21, 259–261. (6) Prodan, C. I.; Joseph, P. M.; Vincent, A. S.; Dale, G. L. J. Thromb. Haem. 2008, 6, 609–614. (7) Dale, G. L. J. Thromb. Haem. 2005, 3, 2185–2192. (8) Tsiara, S.; Elisaf, M.; Jagroop, I. A.; Mikhailidis, D. P. Clin. Appl. Thromb./ Hemostasis 2003, 9, 177–190. (9) Endler, G.; Klimesch, A.; Sunder-Plassmann, H.; Schillinger, M.; et al. Br. J. Hamaetol. 2002, 117, 399–404. (10) Michelson, A. D. Blood 1996, 87, 4925–4936.
Published on Web 01/28/2010
Langmuir 2010, 26(18), 14700–14706
Basabe-Desmonts et al.
Article
using physical structures to assist the capture process;18 in our work, the protein spots alone are sufficient to capture the platelets. To the best of our knowledge, protein micropatterning to create single-cell arrays has never been used to control platelet-platelet distances and numbers of platelets per spot within such arrays. Moreover, the use of micropatterned surfaces for the specific separation of one cell type from a complex matrix, such as blood, has not been reported previously.
2. Materials and Methods 2.1. Materials. For the fabrication of polydimethylsiloxane Figure 1. Schematic depiction of the incubation of whole blood on an iPC substrate patterned with a platelet-specific protein. Single platelets from whole blood adhere to the protein spots in a single step; the remainder of the blood is then rinsed away and the platelets are labeled. Arrays of single platelets are fabricated on disposable transparent substrates, enabling the direct observation of isolated single platelets and a straightforward and accurate method for quantifying platelet adhesion, size, morphology, and receptor expression.
interactions: in vivo, platelet interactions with vessel walls induce activation. Thus, flow cytometry provides no direct measurements of platelets interacting with surface-confined proteins, only limited information (from light scattering) about morphological changes, and no information about the distribution of surface receptors and other proteins on activated platelets. Moreover, flow cytometry requires expensive equipment and trained personnel. An ideal tool for platelet function studies would provide a detailed characterization of platelet subpopulations without the need for expensive equipment while enabling the statistical measurement of multiple parameters, including platelet adhesion, platelet activation, morphological changes, and platelet receptor levels and distributions. The correlation of the quantification of these parameters with physiological events promises to provide a powerful screening, research, and diagnostic tool. To facilitate their characterization, platelets must be separated from other blood components without activating or otherwise perturbing them. Sample preparation can cause artifactual platelet activation, hence an approach that does not require sample preparation for the separation of platelets from other blood components is highly desirable. We introduce here a new method of studying platelet function and subpopulations that addresses these requirements using onchip interfacial platelet cytometry (iPC). Platelet-specific proteins are patterned on a glass surface to capture specifically single or multiple platelets per protein “spot”. Platelets are thus separated in a single step from other blood components, enabling direct optical observation, quantification (i.e., cytometry), and statistical analysis of platelet subpopulations (Figure 1). Protein patterning provides complete control of the size, shape, and spacing between protein-coated areas, largely determining the number, spacing, and interactions of platelets captured on each spot. Arrays to capture selectively single or multiple cells have been created by others for various applications,11-17 in some cases (11) Di Carlo, D.; Lee, L. P. Anal. Chem. 2006, 78, 7918–7925. (12) Di Carlo, D.; Wu, L. Y.; Lee, L. P. Lab Chip 2006, 6, 1445–1449. (14) Thery, M.; Jimenez-Dalmaroni, A.; Racine, V.; Bornens, M.; et al. Nature 2007, 447, 493–U496. (15) Cerf, A.; Cau, J. C.; Vieu, C.; Dague, E. Langmuir 2009, 25, 5731–5736. (16) Falconnet, D.; Csucs, G.; Grandin, H. M.; Textor, M. Biomaterials 2006, 27, 3044–3063. (17) Chen, C. S.; Mrksich, M.; Huang, S.; Whitesides, G. M.; et al. Science 1997, 276, 1425–1428.
Langmuir 2010, 26(18), 14700–14706
(PDMS) stamps for microcontact printing, Dow Corning Sylgard 184 was purchased from Farnell (Farnell Ireland, Dublin, Ireland). MICROPOSIT S1818 positive photoresist (Chestech Ltd., Warwickshire, U.K.) was used in the fabrication of masters for PDMS curing. Coverglass slips (20 20 mm2) were purchased from Agar Scientific Ltd. (Essex, England). PBS tablets, paraformaldehyde, sodium citrate, and all of the chemicals for the buffer were purchased from Sigma Chemical Company (St. Louis, MO). For each day, a 37% stock solution of paraformaldehyde (PFA) was made in distilled water with 1.4% 2 N NaOH. A working solution of 3.7% PFA in PBS was then prepared from this. For the preparation of samples for microscopy inspection, samples were mounted on glass slides using Dako Cytomation mounting medium (Dako A/S, Glostrup, Denmark). 2.1.1. Buffers. HEPES buffer consisted of 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 5.6 mM glucose, 1 g/L bovine serum albumin, and 20 mM HEPES (N-[2-hydroxyethyl]piperazineN0 -[2-ethanesulphonic acid]) at pH 7.40. Platelet buffer A consisted of 130 mM NaCl, 6 mM dextrose, 9 mM NaHCO3, 10 mM sodium citrate, 10 mM Tris, 3 mM KCl, 0.81 mM KH2PO4, and 0.9 mM MgCl2. This buffer was made from stock solutions, and the pH was adjusted to 7.35 on the day of use. 2.1.2. Proteins and Antibodies. Plasminogen-depleted human fibrinogen (g95% clottable protein, homogeneous by SDS-PAGE) was bought from Calbiochem (Merck, KGaA, Darmstadt, Germany). Mouse monoclonal antibodies against CD41 (clone P2) and CD42b (clone SZ2) were from Immunotech (Marseilles, France). VWF was a kind gift from Robert Montgomery (Blood Research Institute, Milwaukee, WI). The mouse monoclonal antifibrinogen antibody (clone GMA-035) was from Upstate (Millipore, Lake Placid, NY). The rabbit polyclonal antihuman antibody to VWF antibody was purchased from Dako (Glostrup, Denmark). Goat serum was from Vector Laboratories (Burlingame, CA). The Alexa Fluor 488 goat antimouse and goat antirabbit antibodies were from Molecular Probes (Invitrogen, Carlsbad, CA). 2.1.3. Reagents for Flow Cytometry Experiments. Phycoerythrin (PE)-conjugated mouse antibody against P-selectin (CD62P) was from BD Pharmingen (Franklin Lakes, NJ). Chemicals used for the HEPES buffer were from Sigma Chemical Company (St. Louis, MO). 2.2. Methods. 2.2.1. PDMS Stamp Fabrication. Patterned PDMS stamps were fabricated by pouring a 10:1 (v/v) mixture of Sylgard 184 elastomer and curing agent over a patterned silicon master. Fabrication of the patterned silicon master was done as follows: Microposit S1818 positive photoresist was spun at 5500 rpm for 30 s on a silicon wafer. It was then cured for 1 min in vacuum on a hot plate at 115 °C. UV light irradiated the photoresist layer for 20 s through a photomask (Photronics, Mid Glamorgan, South Wales, U.K.). Resultant features were developed by dipping the master in developer MF319 (Chestech Ltd., Warwickshire, U.K.) for 40 s and finally rinsing with water and drying under nitrogen. Subsequently, masters were exposed to a vapor of (tridecafluoro-1,1,2,2-tetrahydrooctyl)-1-trichlorosilane (Sigma-Aldrich Inc., Ireland) under (18) Rettig, J. R.; Folch, A. Anal. Chem. 2005, 77, 5628–5634.
DOI: 10.1021/la9039682
14701
Article vacuum for 1 h to facilitate the release of the PDMS mold after curing. The mixture was cured for 1 h in an oven at 60 °C and then carefully peeled away from the master and left in the oven for another 18 h at 60 °C to ensure complete curing. Prior to inking of the stamps, they were oxidized by exposure to UV/ozone for 10 min. This process makes the stamp surface hydrophilic, which ensures homogeneous spreading of the ink (i.e., the protein solution). All of the stamps were freshly prepared no more than 2 days prior to use. 2.2.2. Preparation of Protein-Patterned Surfaces. Coverglass slips (20 20 0.150 mm3) were used as glass substrates for protein patterning. Prior to use, they were rinsed with ethanol and dried. After ozone treatment, the PDMS stamps were immediately inked with 50 μL of a 200 μg/mL protein solution in phosphate-buffered saline (PBS) for at least 15 min. Excess ink solution was removed from the PDMS surface with a pipet, and the stamp was blown dry with nitrogen. The stamp was placed in contact with a clean glass substrate for 5 min. After each printing, the stamp was inked for another 5 min and used again to print new samples. Each stamp was used to print up to seven samples and then discarded. After protein micropatterning, the glass slides were immersed in BSA solution (10 mg/mL) in PBS, blocking any uncovered regions of the glass surface, for at least 1 h at room temperature. The slides were rinsed with PBS prior to use for blood incubation and testing. Typically, patterned substrates were used within 2 days after preparation. Patterns appeared to be stable even after weeks of storage for samples at 4 °C in BSA solution in the dark. 2.2.3. Immunostaining of Protein-Patterned Surfaces. Fibrinogen and VWF surfaces were incubated for 1 h at room temperature with a primary antibody solution (1 μg/mL in PBS). (See details above.) After being rinsed with PBS, fibrinogen and VWF samples were incubated for 20 min at room temperature in the dark with a solution of goat antimouse or goat antirabbit Alexa Fluor 488 antibody (4 μg/mL in PBS), respectively. AntiCD42b-patterned surfaces were incubated for 20 min at room temperature in the dark with a solution of goat antimouse Alexa Fluor 488 antibody (4 μg/mL in PBS). Finally, all three types of immunostained surfaces were rinsed in PBS and mounted on a glass slide for microscopy using Dako Cytomation mounting medium. 2.2.4. Blood Collection and Sample Incubation. Venous blood was collected from healthy adult volunteers who had not ingested any drugs known to interfere with platelet function for at least 2 weeks prior to phlebotomy. Blood was collected by venipuncture through a 19-gauge Butterfly needle into a plastic syringe containing 3.2% sodium citrate (1:10 of the volume of blood, final concentration 10.9 mM). The first 3 mL of blood was discarded. One milliliter of blood was added to a protein-patterned glass substrate in a 35-mm-diameter Petri dish on a “seesaw” rocking platform (Stuart SSL4, Stone, Staffordshire, U.K.) and incubated at 35 oscillations per minute for 30 min at room temperature. After the incubation with blood, the surfaces were rinsed with buffer A and placed in 3.7% PFA for at least 15 min. For surfaces not incubated with blood, the fixation step was omitted. For immunostaining, after being rinsed with PBS buffer, surfaces were incubated for 1 h at room temperature with a primary antibody solution (1 μg/mL in PBS containing 1.5% goat serum). Subsequently, after being rinsed with PBS, samples were incubated for 20 min at room temperature in the dark with a solution of goat antimouse Alexa Fluor 488 antibody (4 μg/mL in PBS containing 1.5% goat serum). The surfaces were rinsed in PBS and mounted on a glass slide for microscopy using Dako Cytomation mounting medium (Dako A/S, Glostrup, Denmark). 2.2.5. Flow Cytometry Measurements. To assess whether platelets were activated by the incubation procedure itself, P-selectin expression was measured by flow cytometry on nonadherent platelets from aliquots taken from whole blood incubated with the surfaces. Aliquots of blood were collected at the 14702 DOI: 10.1021/la9039682
Basabe-Desmonts et al. start of the incubation and after 30 min. Whole blood (10 μL) was added to plastic tubes containing saturated concentrations of PEP-selectin and HEPES buffer, giving a final volume of 100 μL. After incubation at room temperature for exactly 10 min, the reaction was stopped by the addition of 1 mL of HEPES buffer. The samples were then immediately analyzed by flow cytometry. Flow cytometry was performed on a fluorescence-activated cell system (FACSCalibur with CellQuest Pro software from Becton Dickinson, Franklin Lakes, NJ). The platelet population was identified by its forward- and side-scattering properties, and the mean fluorescence index (MFI) for FL2 (PE-anti-P-selectin) was recorded. 2.2.6. Microscopy Images. Fluorescence microscopy images were obtained using an inverted microscope (Olympus IX81) equipped with a CCD camera (Hamamatsu C4742-80-12AG) and a xenon lamp as the light source. Images were collected with a 40 or 100 1.3 NA oil-immersion objective (excitation filter BP492/18; emission light was collected through a filter cube, U-MF2, Olympus). 2.2.7. SEM Images. To obtain SEM images, fibrinogen- and anti-CD42b antibody-patterned surfaces were incubated with blood, rinsed, and fixed as before. After gold sputtering, the samples were imaged by a Zeiss EVO LS15 SEM instrument.
2.2.8. Multiple- and Single-Platelet Adhesion Quantification. To quantify either the number of platelets per protein spot or the fractional occupancy of an array of single-platelet spots, fluorescence microscopy images were acquired using a 40 oilimmersion objective (see details above) of adhering stained platelets. The field of view of each such image is 200 μm 150 μm. The stained platelet membrane is significantly brighter than the cytosol (the internal cytoplasmic matrix), allowing the easy distinction of two adjacent platelets from one another in the case of larger spots that can bind multiple platelets. For multiplatelet binding, the number of platelets per spot is simply counted manually. Mean values of the platelet count per spot are obtained in such cases by averaging the number of platelets on at least 32 different spots. For single-platelet binding assays in which the protein and the spot size are chosen such that no more than one platelet binds per printed protein spot, the number of occupied spots was counted out of a total of 170 or 117 of the 2 and 6 μm spot arrays, respectively. To facilitate the quantification process, the patterns included markers in the form of crosses. Four of these markers delimited an area approximately equal to that of the 40 objective field of view. These markers also helped to identify unique areas for imaging (Figure 2). Six different images were acquired for each sample: the number of occupied spots was counted and used to calculate the percentage of single-platelet-occupied spots in each image. The protein-patterned surface occupancy in the sample was calculated as the mean value of the occupancy in the six different images.
3. Results and Discussion 3.1. Single-Step Specific Separation of Platelets from Whole Blood and Control of Platelet Spreading and PlateletPlatelet Distances. Three surface-patterned proteins were studied to create functional iPC substrates: fibrinogen, VWF, and anti-CD42b antibody. VWF and fibrinogen are blood glycoproteins that, at low shear rates (
