Uptake pathways of protein coated magnetic nanoparticles in platelets

Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Uptake Pathways of Protein-Coated Magnetic Nanoparticles in Platelets Thi-Huong Nguyen,*,†,‡ Nicola Schuster,‡ Andreas Greinacher,† and Konstanze Aurich*,† †

Institute for Immunology and Transfusion Medicine, University Medicine Greifswald, 17475 Greifswald, Germany ZIK HIKECenter for Innovation Competence, Humoral Immune Reactions in Cardiovascular Diseases, University of Greifswald, 17489 Greifswald, Germany

‡

ACS Appl. Mater. Interfaces Downloaded from pubs.acs.org by UNIV OF KENTUCKY on 08/20/18. For personal use only.

S Supporting Information *

ABSTRACT: Magnetic nanoparticles have recently shown great potential in nonradioactive labeling of platelets. Platelet labeling efficiency is enhanced when particles are conjugated with proteins like human serum albumin (HSA). However, the optimal HSA density coated on particles and the uptake mechanism of single particles in platelets remain unclear. Here, we utilized single-molecule force spectroscopy (SMFS) and other complementary methods to characterize the interaction of particles when interacting with platelets and to determine the optimal HSA amount required to coat particles. An HSA concentration of 0.5−1.0 mg/mL for coating particles is most efficient for platelet labeling. Binding pathways could be elucidated by linking a single HSA particle to SMFS tips via polyethylene glycol (PEG) linkers of different lengths and allowing them to interact with immobilized platelets on the substrate. Depending on the PEG length (i.e., short ∼2 nm, medium ∼30 nm, and long ∼100 nm), particles interact differently with platelets as shown by one, two, or three force distributions, which correspond up to three different binding pathways, respectively. We propose a model that the short PEG linker allows the particle to interact only with the platelet membrane, whereas the medium and long PEG linkers promote the particle to transfer from open canalicular system to another target inside platelets. Our study optimizes magnetic platelet labeling and provides details of particle pathways in platelets. KEYWORDS: platelet labeling, magnetic nanoparticles, force, pathway, human serum albumin

■

INTRODUCTION Besides their role in preventing bleeding, platelets are increasingly recognized as important cells with several biological functions, including bacterial defense.1 In addition, platelet concentrates are an essential treatment for patients with severe thrombocytopenia or platelet function defects. In vivo assessment of the liability and function of platelets is highly desirable. A methodology to track transfused platelets in vivo or to distinguish autologous platelets from proband’s own platelets within the frame of a clinical phase-I trial with newly developed platelet concentrates is to label platelets prior to infusion.2,3 Since decades, radioactive tracers have been used as a gold standard for cell labeling. However, application of radioactive tracers to patients or probands in clinical trials is increasingly considered to be unethical.4 Other platelet labeling methods such as biotinylation show limitations as platelet aggregation and activation,5 while differentiation of transfused from patients’ own platelets based on cell surface proteins like human leukocyte antigens5,6 is problematic in autologous settings with volunteers due to the requirement of allogenic blood. Recently, we developed a method for magnetic labeling of platelets using ferucarbotran particles and found that platelets incorporate these particles by endocytosis without linkers or binding agents.7 Flow cytometry using fluorescein isothiocya© XXXX American Chemical Society

nate-conjugated magnetic ferucarbotran particles showed that platelet labeling efficiency reached a yield of about 83%.8 Platelet labeling efficiency is further enhanced when particles are conjugated with proteins like human serum albumin (HSA).9,10 Platelet visualization by fluorescent and electron microscopy showed an adequate uptake of these proteincoated particles into the open canalicular compartment or in the α-granules. Only a few particles remain attached to the platelet membrane.8 Labeling caused no major impairment of platelet functions or impaired survival in an NOD/SCID mouse model. Furthermore, magnetically labeled platelets can be reisolated from whole blood for further analysis.8 Here, we further characterize the uptake of coated magnetic nanoparticles by platelets. We use mainly single-molecule force spectroscopy (SMFS), a methodology that allows direct measurements of biological processes based on molecular interactions at single-molecule resolution to (i) identify the optimal HSA concentration at which platelets show the highest ability to take up particles and (ii) track binding pathways of a particle during platelet uptake. Received: May 9, 2018 Accepted: August 1, 2018 Published: August 1, 2018 A

DOI: 10.1021/acsami.8b07588 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

■

groups at the end of PEG linkers were activated for 1 h at RT by amine coupling kit 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide/ N-hydroxysuccinimide (Biacore, Uppsala, Sweden). To link a bare particle to the tip, a PEG linker carrying amino group (HS−PEG− NH2) instead of carboxyl group (HS−PEG−COOH) was used. Carboxyl groups on the bare particles were activated prior to incubation with the tip−PEG−NH2 for formation of amide bonds. After rinsing, the cantilevers were incubated with bare or HSA particles (20 μM) for 1 h at RT and then overnight at 4 °C. The next day, free activated groups on the surfaces were blocked by adding 1.0 M ethanolamine (Biacore, Uppsala, Sweden) for 1 h at RT. Afterward, the cantilevers were rinsed with PBS and immediately used for SMFS measurements; otherwise, they were kept maximal 3 day at 4 °C until use. To avoid multiple particles immobilized at the end of the tip, we coated the tips with different concentrations of particle and found that at concentrations ≤20 μM only few multiple rupture events were observed. Thus, a concentration of 20 μM was selected in this study. Furthermore, the tip end (∼20 nm) is five times smaller than the particle diameter (∼100 nm) which can prevent multiple particles being immobilized at the end of the tip. Immobilization of Platelets on the Substrate. Twenty four millimeter round glass coverslips or silicon (Plano GmbH, Wetzlar, Germany) was cleaned as previously described.12 The freshly cleaned substrates were then incubated for 3 h at 37 °C in 50 μg/mL collagen G (acid-soluble calfskin collagen, Biochrom GmbH, Berlin, Germany).13 Subsequently, these material-passivated substrates were rinsed three times with PBS-containing 1.0 mM CaCl2 and 0.5 mM MgCl2. An aliquot platelet suspension of 10 μL from healthy human donors at a density of 3 × 105 platelets/μL, washed as previously described,14 is added dropwise on the freshly passivated substrates and kept at RT for 15 min. For blocking open canalicular system (OCS) channels as a control, the platelets were allowed to fully spread on the glass surface for 2 h, RT. After rinsing, the immobilized platelets are ready to use. Measurement of Force−Distance Curves. The measurements were carried out in PBS containing 1.0 mM CaCl2 and 0.5 mM MgCl2 using JPK NanoWizard 3 (Berlin, Germany). Before each experiment, the cantilever spring constant was independently measured by a thermal tune procedure.15,16 A setpoint of 200 pN was used to control the maximum force of the tip against the surface. To compare the change of binding forces among HSA particles, 1000 force−distance (F−D) curves were recorded at the same tip velocities (1000 nm/s) for each experiment. For each concentration, the measurements were repeated three times using independently prepared cantilevers and freshly immobilized platelets. Data Analysis. The rupture forces at the final rupture points before the cantilevers went back to the rest position were collected from F−D curves exhibiting the behavior of polymer stretching using the JPK data processing software (version 4.4.18+). The mean rupture force values and their corresponding errors were determined by applying Gaussian fits to the data using Origin software (version 8.6). The one-way analysis of variance was used to determine if any significant differences in binding forces among particles coated with different HSA concentrations. To determine thermal off-rates/binding kinetics, the particles coated with 0.5 and 1.0 mg/mL HSA concentration were used for an interaction with platelets at different velocities (ranging from 10 to 4000 nm/s). The mean rupture forces of the second distributions were determined by applying Gaussian fits. The final median and standard deviation (SD) of rupture force at each velocity were averaged from three repetitions. The thermal off-rates (dissociation constant) of the particle−platelet bonds were determined by applying the Bell−Evans model.17−19 The model describes that the rupture force increases proportionally with the natural logarithm of the loading rate during retraction. The relation between rupture force and loading rate (Ḟ ) is described as 1

METHODS

Ethics. The use of human blood obtained from healthy volunteers including the informed consent procedure was approved by the ethics board at the University of Greifswald. Particle Functionalization with HSA. Ferucarbotran iron oxide particles (ferucarbotran particles; Meito Sangyo, Tokyo, Japan), the raw material for Resovist magnetic resonance imaging contrast agent (concentration of 1 M iron, stable suspension in water), were used. Ferucarbotran particles consist of a maghemite core with a carboxydextran shell and a mean hydrodynamic diameter of 100 nm. Particles were functionalized with HSA to improve particle uptake by platelets. In brief, 50 μL of ferucarbotran particles corresponding to 50 μmol iron were incubated with 40 mg sodium periodate at pH 5 for 40 min at 4 °C. After removing excess periodate, 0.1, 0.5, 1.0, or 2.0 mg/mL HSA were added and the mixture was stored at 4 °C for 2 h. Subsequently, 4 mg of borane dimethylamine and 5 mg of ethanolamine hydrochloride were added to stop the reaction and reduce the imine bond overnight.11 After cleaning particles from excess reaction material via magnetic separation with MACS LS columns (Miltenyi Biotec GmbH, Teterow, Germany), particles were suspended in additive solution for platelet storage (SSP+, Macopharma, Tourcoing, France). Characterization of HSA-Ferucarbotran Particles by Dynamic Light Scattering. The successful particle conjugation was proved by size-determination of aggregates that are formed after addition of polyclonal rabbit anti-HSA-antibodies (Sigma-Aldrich, St. Louis, USA, 1 μg/mL final) by dynamic light scattering (DLS, Zetasizer Nano ZS, Malvern Instruments Ltd, Worcestershire, UK). HSA-conjugated particles and anti-HSA-antibodies form aggregates because of antigen−antibody interaction, in which sizes can be detected by DLS over time. Immediately after adding anti-HSA antibodies, DLS measurement was continuously carried out at 25 °C for 30 min in phosphate-buffered saline (PBS, Sigma-Aldrich, Germany) using disposal cuvettes (Sigma-Aldrich, St. Louis, USA). The measured average intensities in sizes were plotted against time. DLS was also used to determine the zeta potential of the different ferucarbotran particles. Particles were diluted 1:100 with water and measured in a folded capillary zeta cell at 25 °C with six repetitions. Platelet Labeling. For platelet labeling, we used platelets isolated from whole blood of healthy volunteers. Whole blood was anticoagulated by adenine-citrate-dextrose A (ACD-A, Baxter, Germany) and centrifuged for 20 min at 650g without brake. Platelets were diluted with SSP+ (Macopharma, Tourcoing, France) to 400 000/μL and bare ferucarbotran or HSA-coated ferucarbotran particles (2 mM iron final concentration) were added and incubated for 1 h at 37 °C. Excess nanoparticles were removed by centrifugation. During washing centrifugation steps platelets were treated with ACD-A, 111 μL/mL platelets, and apyrase (Grade IV; Sigma-Aldrich, St. Louis, USA; 1000 U/mL). Quantitative Cellular Iron Determination. The cellular iron content was determined by means of atomic absorption spectroscopy (AAS). The labeled cell suspension (200 μL) was centrifuged (1000g; 5 min), and the supernatant was discarded. After cell digestion of the cell pellet with 500 μL of trypsin 0.5% in PBS without Ca2+ and Mg2+ and 500 μL of sodium lauryl sulfate 0.2% for 30 min at 37 °C (all Sigma-Aldrich, Germany), the respective sample was heated in 500 μL of nitric acid 65% and 500 μL of H2O2 30% (all Sigma-Aldrich). After cooling, samples were diluted to 5.0 mL with water and iron content per cell was measured in triplicate using an iron calibration graph with a commercial iron standard ranging from 0 to 150 μg/L by a contrAA 700 atomic absorption spectrometer (Analytik Jena, Jena, Germany). Linking a Single Particle with Atomic Force Microscopy Tip. To link HSA particles, both sides’ gold-coated silicon nitride cantilevers with a nominal spring constant of 6 pN/nm (Olympus Biolever, Tokyo, Japan) were coated with thiol−polyethylene glycol− carboxylic acids (HS−PEG−COOH) of different length such as short (3-mercaptopropionic acid, Sigma-Aldrich, Germany), medium (3400 Da, Nanocs, USA), and long PEG (10 000 Da, Sigma-Aldrich, Germany) for 2 h at room temperature (RT). After that, carboxyl B


Research Article


Figure 1. Characterization of magnetic nanoparticles coated with different amounts of HSA. (A) Particle aggregates increase with increasing HSA concentrations coated on the particles measured by DLS. (B) Zeta potential of particles conjugated with different amounts of HSA as determined by photon correlation spectroscopy [n = 5, mean ± SD]. (C) Iron content of platelets incubated with particles of increasing HSA concentrations determined by AAS (n = 10; mean ± SD).

Figure 2. Measuring interactions between nanoparticles and platelets by SMFS. (A) Schematic of an HSA particle interacting with a platelet: a single particle is immobilized on the tip via PEG linker of ∼30 nm length, while the platelet is adhered on collagen-coated glass to avoid plateletsurface activation. (B) Typical force distributions of particles coated with different HSA concentrations: for bare particles with 0 mg HSA (bottom), the main force distribution appears at ∼98 pN, but also an additional shoulder at ∼220 pN is seen; the ∼220 pN shoulder further increases with increasing HSA concentrations while the peak of the ∼98 pN shoulder mostly disappears; the forces are highest at 1.0 mg HSA and slightly shift toward lower forces at 2.0 mg HSA (top). (C) On fully spread platelets on bare glass as a control (red), the rupture force significantly decreases as compared with that of resting, nonspread platelets (black) at each HSA concentration. (D) Rupture forces between particles and resting platelets at different loading rates show that particles coated with 0.5 mg/mL HSA (red) have a lower binding affinity (=higher thermal offrate (koff) than particles coated with 1.0 mg/mL HSA (black, n = 3 donors, mean ± SD)).

F(F )̇ =

kBT ijj F ̇ Δx yzz zz lnjj Δx jk koff kBT z{

■

RESULTS

Optimizing HSA Density on Magnetic Particles. To optimize the HSA density on magnetic particles to gain highest yield of platelet labeling, we coated the magnetic particles with different HSA concentrations. Successful HSA coating on particles was confirmed by measuring the sizes of particle aggregates that are formed after the addition of anti-HSAantibodies using DLS. DLS results over time show that sizes of the particle aggregates increase with increasing HSA concentrations (Figure 1A).

(1)

where the involved parameters are the Boltzmann constant (kB), the experimental temperature (T, kBT = 4.1 pN nm), the distance from the bound to the unbound state (Δx = kBT/m), the slope of the fit (m), the thermal off-rate (koff = Ḟ (F = 0)Δx/kBT), the loading rate (Ḟ = υkeff), the pulling speed (υ), and the effective spring constant of the system composed of the springs of both cantilever and involved linkers (keff). For the Bell−Evans model, the platelet−nanoparticle complexes were regarded as one molecule, as these complexes are very stable and do not dissociate during the experiment. C


Research Article

ACS Applied Materials & Interfaces Next, we tested whether the uptake of particles into platelets depends on the HSA density on particle surfaces. Platelets were incubated with HSA particles, and the mean cellular iron concentration was determined by AAS. The iron content per labeled platelet increases with increasing HSA concentration coated on particles (Figure 1C). The mean value shows that the iron content per platelet tends to reach a maximum at 0.5− 1.0 mg. However, the iron content per platelet did not show obvious differences when 0.1−1.0 mg HSA was applied. The variation of iron concentration per platelet is significantly large. At a concentration of 2.0 mg HSA, the uptake efficiency decreases. Consistently, the absolute zeta potential decreases with increasing HSA density on the particles suggesting reduced electrostatic repulsion between particles (Figure 1B). Probably, at 2.0 mg/mL HSA, particles tend to aggregate and are not more available for cellular uptake. The wide variability in iron uptake among platelets as quantified by AAS made it difficult to identify the optimal HSA concentration. To better determine this critical concentration, we next investigated single particle−platelet interaction by SMFS-based atomic force microscopy (AFM). SMFS with a better resolution provides a direct insight into the interaction between a single particle and an individual platelet. This methodology in principle describes the force−displacement curves obtained by oscillating the scanner in the z-direction, while the scanner movement in x- and y-directions is disabled. The cantilever interacts with the sample when the z-scanner moves up, and it ruptures from the sample when the z-scanner retracts. By interacting with the sample, the deflection signal from the cantilever and the movement of the piezoelectric scanner are recorded.20 Here, the particle immobilized on the cantilever will interact with platelet immobilized on the substrate and the interaction force is measured when the tip separates from the sample. To determine the interaction force between particles and platelets, a single HSA-coated ferucarbotran particle was covalently immobilized on an AFM-cantilever via a PEG linker (∼30 nm) carrying at each end functional thiol and carboxyl groups (HS−PEG−COOH).21 Thiol groups bind covalently to the gold-coated cantilever while carboxyl groups allow the formation of amide bonds with amino-groups from HSA (Figure 2A). Bare particles without HSA were linked to the tips via a PEG linker (∼30 nm) carrying amino groups (HS− PEG−NH2). Carboxyl groups on the particles’ shell were activated prior to incubation with tip−PEG−NH2 for formation of amide bonds. The particle on the tip was then brought into contact with a platelet immobilized on collagencoated glass substrates and the interaction forces were measured when the tip ruptured from the platelet (Figure 2A). Typical rupture profiles of the interactions are shown in Figure 2B. Two main force peaks at ∼98 pN and a shoulder at ∼220 pN were obtained by using bare particles (Figure 2B, bottom). At increasing HSA concentrations, the force profile of the ∼220 pN shoulder increased, whereas the force at ∼98 pN decreased. However, increasing the HSA concentration to 2.0 mg/mL led to a lower interaction force (∼170 pN), which is consistent with the reduced uptake of nanoparticles by platelets as measured by AAS. As control experiments for the influence of own PEG on final measured forces, the medium and long PEG linkers without HSA particles were coated on the tips and approached to the platelets for interaction. These experiments showed only some random unspecific interactions which are considered as background (Figure 3D, blue). We

Figure 3. Interactions between nanoparticles coated with 1.0 mg HSA and platelets using PEGs of different lengths. Schematics of (A) long (L ≈ 100 nm), (B) medium (L ≈ 30 nm), and (C) short (L ≈ 2 nm) PEG linkers bound to HSA-coated particles interacting with platelets. (D) On resting platelets (gray), one rupture force distribution at ∼80 pN was recorded with the short (bottom) and one higher rupture force distribution at ∼220 pN with the medium PEG linker, whereas three force distribution regimes (∼80, ∼220, and ∼360 pN) appeared when a long PEG linker (second top) was used. On fixed platelets (D, top), only one rupture force distribution was obtained even though long PEG linker was used; only a few random interaction forces for the corresponding PEGs without immobilized particles (blue).

also observed that the platelets kept their round shapes on PEGs of different length coated on glass substrate (Figure S2), indicating that PEGs did not (or weakly) interact with the platelets. It has been shown that the nanoparticles were localized mainly in the OCS of platelets.22 On the basis of this information, we minimized the OCS access by allowing the platelets to spread completely on the glass surface to prevent particle uptake. We, then, measured the interaction force between particles and these fully spread platelets as a control. Comparing with those of nonspread platelets (Figure 2C, black), the interaction forces measured on the fully spread platelets are significantly reduced at each HSA concentration tested (Figure 2C, red). Next, we determined kinetic properties of HSA particle− platelet interactions by measuring the rupture forces between particles and platelets at different loading rates (Figure 2D). The average rupture forces at different loading rates obtained from the second force distribution were plotted against loading rates. The thermal off-rate (koff) of the interaction was obtained by fitting the data applying the Bell−Evans model.23,24 The results demonstrate that particles coated with 1.0 mg/mL HSA show a lower thermal off-rate (koff between 10 and 102 s−1, black) than that of particles coated with 0.5 mg/mL HSA (∼103 s−1, red), which indicates a higher binding affinity for the 1.0 mg/mL HAS-coated particles. Tracking the Binding Pathway of Particles during Platelet Labeling. To track the uptake pathways of particles when they approach platelets, either a long (L ≈ 100 nm; Figure 3A), medium (L ≈ 30 nm; Figure 3B) or short (L ≈ 2 nm; Figure 3C) PEG linker was utilized to link particles to the tips. When the tip contacts the platelets, the PEG linker will control the orientation of the particle coated at the end of the tip. The short PEG linked the particle closely to the tip, D


Research Article


phospholipid bilayer embedded with cholesterol, glycolipids, and glycoproteins and acts as a physical barrier.30 This barrier impedes ferucarbotran particles with a highly negative zeta potential as bare ferucarbotran particles to attach to the membrane due to repulsion forces. Thus, the ideal ferucarbotran uptake, on the one hand, requires a zeta potential >|15| mV to avoid particle aggregation and on the other hand to be low enough to prevent particle repulsion from the negatively charged cell membrane. Here, HSA−ferucarbotran particles meet these criteria when coated at an HSA concentration of 1.0 mg/mL. In addition, it has been previously reported that a protein corona changes the zeta potential of particles and hereby influences the particle−cell interaction.31,32 By coating HSA onto ferucarbotran particles, we produced a protein corona and induced an uptake behavior: particle uptake increased with a maximum at coating concentration from 0.5 to 1.0 mg/mL HSA. Furthermore, the surface charge has a significant impact on the degree and efficacy of nanoparticle uptake. Because of the negative potential of cell membranes, cationic nanoparticles are taken up by cells more efficiently than negatively charged nanoparticles as shown by increasing interaction forces with platelets. Higher concentration of HSA coating lowers the charge of anionic nanoparticles, leading to a more effective particle-attachment on the cell membrane shown by the increased binding force at higher HSA concentrations. Besides charge, the size of particles is also an important criterion for platelet labeling. The highest HSA coating concentration (2 mg/mL) led to a near neutral net charge of particles. This could result in particle aggregate formation because of missing electrostatic repulsion. Then, particles are too large for endocytosis and particle uptake is reduced. Even though the investigated particles were not susceptible to agglomeration in biological fluids, they may have a particle inherent tendency to sediment over time.28 We found that particles coated with 1.0 mg/mL HSA have a higher binding affinity (koff ≈ 10 s−1) to platelets than the ones coated with 0.5 mg/mL HSA (koff ≈ 103 s−1). However, for HeLa cells and iron oxide nanoparticles Wilhelm et al. reported a koff = 4.4 × 10−5 s−1 at 37 °C; labeling human umbilical vein endothelial cells with silicon-quantum dot nanoparticles resulted in a koff = 1.1 × 10−1 s−1.33,34 Cellular uptake rates of nanoparticles are dependent on structure and morphology, hydrophilicity of nanoparticle surface, or surface charge,35 resulting in a wide range of binding constants. Thus, varying the HSA amount on ferucarbotran particles led to ∼100-fold difference in koff among studies. To date, bond dynamics of magnetic particles taken up by platelets remain poorly understood. Here, we are able to track binding pathways of the particles interacting with platelets. As the nanoparticles are either adsorbed on the cellular membranes or internalized into membrane-bound endocytic compartments,27 we linked the particle via a long PEG linker (∼100 nm) to allow it to interact freely with platelets. In contrast to short PEG that displays only one force distribution, the long PEG linker shows three distributions, indicating that short PEG allows the particle to interact only with the membrane, whereas long PEG facilitates it to enter platelet membrane. With the length of ∼100 nm, the PEG allows the particle to move inside the platelet membrane with a distance that even exceeds 100 nm due to PEG stretch.36 It has been reported before that nanoparticles first seem to enter the platelet OCS and then move to platelet granules.7,37

allowing only a small part of the particle to be in contact with the platelet, whereas the long PEG linker allows the particle freely to move on the platelet membrane, and therefore, enhances the efficiency of interaction with platelet. The short PEG linker should preferentially allow interaction of the particle with the platelet membrane, whereas medium and long PEG linkers should permit the particle to be taken up through platelet membranes. Indeed, for 1.0 mg HSA-coated particles, SMFS results show only a single force profile of ∼80 pN for the short PEG linker, whereas the medium PEG linker exhibits an intermediate rupture force at ∼ 220 pN and long PEG linkers display three rupture force distributions including ∼80 pN, the ∼220 pN regimes seen with the other two linkers, but in addition a regime of ∼360 pN (Figure 3D). The results indicate that the short PEG linker allows particles to interact only with the platelet membrane with low interaction forces. This is fully confirmed by the experiment with fixed platelets. Fixed platelets cannot take up particles and only allow binding to the outer membrane. Particles linked by the medium PEG linker bound with higher rupture forces to the platelet, which indicates that they must have been partly internalized. Particles linked by long PEG showed three binding regimes, the first regime indicates the interaction of the particle with platelet membranes, the second regime demonstrates the binding forces obtained also with medium length PEG, and the third shows even higher binding forces. These three distributions in the rupture force profile are not due to multiple interactions because the force magnitude of the second or the third distribution is not “n-fold” of the first distribution. They rather indicate uptake of the HSA particles into different platelet compartments.

■

DISCUSSION This study provides the optimal HSA concentration that is required for magnetic particle coating to achieve the highest platelet labeling rate. Furthermore, a detailed mechanism of particle−platelet interaction could be elucidated. We did not observe any significant aggregation and activation after platelet labeling observed by scanning electron microscopy (SEM) (Figure S1) which is consistent with previous observations.25 The enhancement of particle uptake with increasing HSA amount coated on particles could be explained by changed water solubility on particle surfaces that avoids aggregation and reduces nonspecific adsorptions.26 By changing the coating of magnetic nanoparticles from a low-molecular weight ligand (dextran) to polymers (HSA), the colloidal stability of the dispersion is improved and adsorption and internalization of particles toward living platelets is profoundly affected.27 In contrast to adherent cell lines where HSA conjugation decreases cellular uptake of particles because of a reduced sedimentation ability of HSA particles,28 suspension cells such as platelets benefit from protein conjugation regarding particle uptake. In this case, not only particle concentration reduction due to sedimentation but also altered particle surface properties plays a role. HSA conjugation led to a reduced absolute amount of the zeta potential. A zeta potential below an absolute value of 15 mV is considered to be insufficient for electrostatic stabilization of ferucarbotran particles.29 This applies to HSA−ferucarbotran particles with 2.0 mg/mL HSA and may lead to the formation of particle aggregates. On the other hand, despite that the high zeta potential bare ferucarbotran particles show a moderate uptake into platelets. The cell membrane consists of a negatively charged E


Research Article


Figure 4. Proposed model of binding pathways of the magnetic nanoparticle to platelets. Nanoparticles bound to a short linker typically bind to the platelet membrane. However, when they attach by chance close to the opening of the OCS, they may partly enter the OCS (A). In both cases, weak interaction forces will be measured (B, black). Medium-length PEG linkers allow the particle to enter the OCS, whereas long PEG linkers (C) allow the particles to be taken up by not only the OCS (D) but also the cytoplasm or even the α granules (E), which lead to appearance of the additional force regimes (B, cyan and red). TEM images (adopted from Aurich et al.8) of nonlabeled (F) and magnetically labeled platelets (G) show magnetic nanoparticles are mainly located in the OCS, some in granules (blue arrows).

Here, we provide further evidence of a multistep binding process. First, particles bind to the outer platelet membrane. This corresponds to the weakest force of ∼36 pN we measured by SMFS. As approximately the same force was measured for the interaction of nanoparticles with spread and even paraformaldehyde fixed platelets, the first binding step is rather a passive than an active binding process (PFA, Figure 3D, top) of the interaction between particles and the outer membrane. The second binding force we measured is obtained when the particle has entered the OCS as it has been shown previously (Figure 4F).7 Previously, our transmission electron microscopy (TEM) images showed the majority of particles in the OCS while some particles are located in α-granules.8 It is therefore very likely that the ∼360 pN force we measured only with the longest PEG linker results from the rupture of particles out of the α-granules. The peak at ∼80 pN in the case of the long PEG linker is more pronounced than that of the medium or short PEG. Probably, via the long PEG, the particle adhered to the membrane and the PEG was stretched before the particle is detached from the membrane. The peak of ∼80 pN was induced by interaction of the particle with the platelet membrane that supposes to appear for every system regardless of medium or long PEG. However, the force in the long PEGsystem is slightly lower than that in the medium PEG-system. This can be explained by the effect of long PEG linker. A series pulling of particle from membrane together with stretching of the long PEG may be a reason that leads to a lower final force. In addition, the particle may be automatically dissociated from platelet membrane during stretching the long PEG that also can lead to lower measured force. The required force and time to stretch the medium PEG (∼30 nm) are not significant, and therefore, the effect of PEG on the measured force was not clearly observed as compared to the long PEG system (∼100 nm). According to these results, we suggest a model for platelet uptake of magnetic particles (Figure 4). Because the access to

OCS channels is not exposed regularly on platelet membranes, the yield that the particles find and contact the OCS access to enter the platelet membrane is not 100%, regardless of the PEG length. When short PEG is used, the particle can either interact with the platelet membrane (Figure 4A, arrows) or partially enter the OCS if it reaches exactly the accesses (Figure 4A). For both cases, weak interaction force will be induced (Figure 4B, black). On the contradictory, medium and long PEG (Figure 4C) allows particles to interact fully with all targets that lead to a significant increase of interaction forces (Figure 4B, cyan and red). While binding of particles to the outer membrane of platelets and to the OCS is clearly supported by electron microscopy (Figure 4F,G), their uptake into α-granules is less clear. When the particle is taken up by α-granules, it will show a series of several rupture forces. A first rupture force when it is pulled out of the α-granule and a second rupture force, when it is pulled out of the OCS or through the outer cell membrane. Indeed, using the long PEG linker ∼35% retraction curves in our experiments shows two steps of ruptures (Figure 4B, red): one at ∼200 pN and another one at a high force of ∼360 pN.

■

CONCLUSION Magnetic nanoparticle conjugation with 0.5−1.0 mg/mL HSA entailed the optimal particle uptake in platelets. Binding affinities between particles and platelets are dependent on the HSA coating amount. Particles coated with 1.0 mg/mL HSA resulted in a ∼100-fold higher binding affinity to platelets than particles coated with 0.5 mg/mL HSA. By linking magnetic particles with PEG of different lengths on a cantilever tip, we could identify three different binding targets of platelets by SMFS: (i) adsorption on the platelet membrane, (ii) entry into the OCS, and (iii) likely transfer to platelet granules. Our results reveal platelet−particle interaction mechanisms on a single particle level. These data are of high relevance for the assessment of platelets used not only for transfusion of patients F


Research Article


(7) Aurich, K.; Spoerl, M.-C.; Fürll, B.; Sietmann, R.; Greinacher, A.; Hosten, N.; Weitschies, W. Development of a Method for Magnetic Labeling of Platelets. Nanomed. Nanotechnol. Biol. Med. 2012, 8, 537− 544. (8) Aurich, K.; Wesche, J.; Palankar, R.; Schlüter, R.; Bakchoul, T.; Greinacher, A. Magnetic Nanoparticle Labeling of Human Platelets from Platelet Concentrates for Recovery and Survival Studies. ACS Appl. Mater. Interfaces 2017, 9, 34666−34673. (9) Albanese, A.; Tang, P. S.; Chan, W. C. W. The Effect of Nanoparticle Size, Shape, and Surface Chemistry on Biological Systems. Annu. Rev. Biomed. Eng. 2012, 14, 1−16. (10) Matuszak, J.; Dörfler, P.; Zaloga, J.; Unterweger, H.; Lyer, S.; Dietel, B.; Alexiou, C.; Cicha, I. Shell Matters: Magnetic Targeting of Spions and in Vitro Effects on Endothelial and Monocytic Cell Function. Clin. Hemorheol. Microcirc. 2015, 61, 259−277. (11) Romanus, E.; Groß, C.; Glöckl, G.; Weber, P.; Weitschies, W. Determination of Biological Binding Reactions by Field-Induced Birefringence Measurements. J. Magn. Magn. Mater. 2002, 252, 384− 386. (12) Krauel, K.; Medvedev, N.; Palankar, R.; Greinacher, A.; Delcea, M. Micropatterned Array to Assess the Interaction of Single Platelets with Platelet Factor 4-Heparin-Igg Complexes. Thromb. Haemostasis 2014, 111, 862−872. (13) Nguyen, T.-H.; Palankar, R.; Bui, V.-C.; Medvedev, N.; Greinacher, A.; Delcea, M. Rupture Forces among Human Blood Platelets at Different Degrees of Activation. Sci. Rep. 2016, 6, 25402. (14) Krauel, K.; Fürll, B.; Warkentin, T. E.; Weitschies, W.; Kohlmann, T.; Sheppard, J. I.; Greinacher, A. Heparin-Induced Thrombocytopenia - Therapeutic Concentrations of Danaparoid, Unlike Fondaparinux and Direct Thrombin Inhibitors, Inhibit Formation of Platelet Factor 4-Heparin Complexes. J. Thromb. Haemostasis 2008, 6, 2160−2167. (15) Hutter, J. L.; Bechhoefer, J. Calibration of atomic-force microscope tips. Rev. Sci. Instrum. 1993, 64, 1868−1873. (16) Butt, H.-J.; Jaschke, M. Calculation of thermal noise in atomic force microscopy. Nanotechnology 1995, 6, 1−7. (17) Merkel, R.; Nassoy, P.; Leung, A.; Ritchie, K.; Evans, E. Energy Landscapes of Receptor-Ligand Bonds Explored with Dynamic Force Spectroscopy. Nature 1999, 397, 50−53. (18) Evans, E. Probing the Relation Between Force-Lifetime-and Chemistry in Single Molecular Bonds. Annu. Rev. Biophys. Biomol. Struct. 2001, 30, 105−128. (19) Evans, E.; Leung, A.; Hammer, D.; Simon, S. Chemically Distinct Transition States Govern Rapid Dissociation of Single LSelectin Bonds under Force. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 3784−3789. (20) Nguyen, T.-H. Single-Molecule Force Spectroscopy Applied to Heparin-Induced Thrombocytopenia. J. Mol. Recognit. 2016, 30, e2585. (21) Strunz, T.; Oroszlan, K.; Schafer, R.; Guntherodt, H.-J. Dynamic Force Spectroscopy of Single DNA Molecules. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 11277−11282. (22) Escolar, G.; White, J. G. The Platelet Open Canalicular System - a Final Common Pathway. Blood Cell 1991, 17, 467−485. (23) Evans, E.; Ritchie, K. Strength of a Weak Bond Connecting Flexible Polymer Chains. Biophys. J. 1999, 76, 2439−2447. (24) Evans, E. A.; Calderwood, D. A. Forces and Bond Dynamics in Cell Adhesion. Science 2007, 316, 1148−1153. (25) Santos-Martinez, M. J.; Tomaszewski, K. A.; Medina, C.; Bazou, D.; Gilmer, J. F.; Radomski, M. W. Pharmacological Characterization of Nanoparticle-Induced Platelet Microaggregation Using Quartz Crystal Microbalance with Dissipation: Comparison with Light Aggregometry. Int. J. Nanomed. 2015, 10, 5107−5119. (26) Zhang, S.; Gao, H.; Bao, G. Physical Principles of Nanoparticle Cellular Endocytosis. ACS Nano 2015, 9, 8655−8671. (27) Safi, M.; Courtois, J.; Seigneuret, M.; Conjeaud, H.; Berret, J.-F. The effects of aggregation and protein corona on the cellular internalization of iron oxide nanoparticles. Biomaterials 2011, 32, 9353−9363.

but also for basic studies like the interaction of platelets with pathogens, especially virus.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.8b07588. SEM images of human platelets after uptaking ferucarbotran particles and platelet morphology on PEG-coated glass (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (T.-H.N.). *E-mail: [email protected]. Phone: +49-3834-8619518 (K.A.). ORCID

Konstanze Aurich: 0000-0003-3365-0168 Author Contributions

T.-H.N. developed the study concept, designed and performed the SMFS, SEM, and partially DLS experiments, analyzed and interpreted the data, and wrote the manuscript. N.S. partially performed DLS experiments. A.G. discussed the data and wrote the manuscript. K.A. conjugated HSA particles, designed and performed the DLS and AAS experiments, interpreted the data, and wrote the manuscript. All authors read and agreed to the final version of the manuscript. Funding

This work was supported by the Starting up Project FOMM/ FOCM-2016-04 University Medicine Greifswald. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS We thank Ulrike Strobel, Ricarda Raschke, Jessica Fuhrmann, and Robert Koch for the technical support. REFERENCES

(1) Gaertner, F.; Ahmad, Z.; Rosenberger, G.; Fan, S.; Nicolai, L.; Busch, B.; Yavuz, G.; Luckner, M.; Ishikawa-Ankerhold, H.; Hennel, R.; Benechet, A.; Lorenz, M.; Chandraratne, S.; Schubert, I.; Helmer, S.; Striednig, B.; Stark, K.; Janko, M.; Böttcher, R. T.; Verschoor, A.; Leon, C.; Gachet, C.; Gudermann, T.; Mederos y Schnitzler, M.; Pincus, Z.; Iannacone, M.; Haas, R.; Wanner, G.; Lauber, K.; Sixt, M.; Massberg, S. Migrating Platelets Are Mechano-Scavengers That Collect and Bundle Bacteria. Cell 2017, 171, 1368−1382. (2) van der Meer, P. F.; Tomson, B.; Brand, A. In Vivo Tracking of Transfused Platelets for Recovery and Survival Studies: An Appraisal of Labeling Methods. Transfus. Apher. Sci. 2010, 42, 53−61. (3) Vassallo, R. R.; Murphy, S. A Critical Comparison of Platelet Preparation Methods. Curr. Opin. Hematol. 2006, 13, 323−330. (4) Decristoforo, C.; Penuelas, I.; Patt, M.; Todde, S. European Regulations for the Introduction of Novel Radiopharmaceuticals in the Clinical Setting. Q. J. Nucl. Med. Mol. Imaging 2017, 61, 135−144. (5) Stohlawetz, P.; Horvath, M.; Pernerstorfer, T.; Nguyen, H.; Vondrovec, B.; Robisch, A.; Eichler, H. G.; Spitzauer, S.; Jilma, B. Effects of Nitric Oxide on Platelet Activation During Plateletpheresis and in Vivo Tracking of Biotinylated Platelets in Humans. Transfusion 1999, 39, 506−514. (6) Hughes, D. L.; Evans, G.; Metcalfe, P.; Goodall, A. H.; Williamson, L. M. Tracking and Characterisation of Transfused Platelets by Two Colour, Whole Blood Flow Cytometry. Br. J. Haematol. 2005, 130, 791−794. G


Research Article

ACS Applied Materials & Interfaces (28) Poller, J.; Zaloga, J.; Schreiber, E.; Unterweger, H.; Janko, C.; Radon, P.; Eberbeck, D.; Trahms, L.; Alexiou, C.; Friedrich, R. Selection of Potential Iron Oxide Nanoparticles for Breast Cancer Treatment Based on in Vitro Cytotoxicity and Cellular Uptake. Int. J. Nanomed. 2017, 12, 3207−3220. (29) Perillo, E.; Hervé-Aubert, K.; Allard-Vannier, E.; Falanga, A.; Galdiero, S.; Chourpa, I. Synthesis and in Vitro Evaluation of Fluorescent and Magnetic Nanoparticles Functionalized with a Cell Penetrating Peptide for Cancer Theranosis. J. Colloid Interface Sci. 2017, 499, 209−217. (30) Kievit, F. M.; Zhang, M. Surface Engineering of Iron Oxide Nanoparticles for Targeted Cancer Therapy. Acc. Chem. Res. 2011, 44, 853−862. (31) Vogt, C.; Pernemalm, M.; Kohonen, P.; Laurent, S.; Hultenby, K.; Vahter, M.; Lehtiö, J.; Toprak, M. S.; Fadeel, B. Proteomics Analysis Reveals Distinct Corona Composition on Magnetic Nanoparticles with Different Surface Coatings: Implications for Interactions with Primary Human Macrophages. PLoS One 2015, 10, e0129008. (32) Gräfe, C.; Weidner, A.; Lühe, M. v. d.; Bergemann, C.; Schacher, F. H.; Clement, J. H.; Dutz, S. Intentional Formation of a Protein Corona on Nanoparticles: Serum Concentration Affects Protein Corona Mass, Surface Charge, and Nanoparticle-Cell Interaction. Int. J. Biochem. Cell Biol. 2016, 75, 196−202. (33) Wilhelm, C.; Gazeau, F.; Roger, J.; Pons, J. N.; Bacri, J.-C. Interaction of Anionic Superparamagnetic Nanoparticles with Cells: Kinetic Analyses of Membrane Adsorption and Subsequent Internalization. Langmuir 2002, 18, 8148−8155. (34) Ohta, S.; Inasawa, S.; Yamaguchi, Y. Real Time Observation and Kinetic Modeling of the Cellular Uptake and Removal of Silicon Quantum Dots. Biomaterials 2012, 33, 4639−4645. (35) Mailänder, V.; Landfester, K. Interaction of Nanoparticles with Cells. Biomacromolecules 2009, 10, 2379−2400. (36) Kreuzer, H. J.; Payne, S. H.; Livadaru, L. Stretching a Macromolecule in an Atomic Force Microscope: Statistical Mechanical Analysis. Biophys. J. 2001, 80, 2505−2514. (37) Escolar, G.; Leistikow, E.; White, J. G. The Fate of the Open Canalicular System in Surface and Suspension-Activated Platelets. Blood 1989, 74, 1983−1988.

H


Uptake pathways of protein coated magnetic nanoparticles in platelets

Recommend Documents