Accepted Manuscript Title: In vitro study of the specific interaction between poly (2-dimethylamino ethylmethacrylate) based polymers with platelets and red blood cells Author: Luca Flebus Franc¸ois Lombart Lucia Martinez-Jothar Chantal Sevrin C´eline Delierneux C´ecile Oury Christian Grandfils PII: DOI: Reference:
S0378-5173(15)30002-8 http://dx.doi.org/doi:10.1016/j.ijpharm.2015.06.036 IJP 14988
To appear in:
International Journal of Pharmaceutics
Received date: Revised date: Accepted date:
11-5-2015 19-6-2015 20-6-2015
Please cite this article as: Flebus, Luca, Lombart, Franc¸ois, Martinez-Jothar, Lucia, Sevrin, Chantal, Delierneux, C´eline, Oury, C´ecile, Grandfils, Christian, In vitro study of the specific interaction between poly (2-dimethylamino ethylmethacrylate) based polymers with platelets and red blood cells.International Journal of Pharmaceutics http://dx.doi.org/10.1016/j.ijpharm.2015.06.036 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
In vitro study of the specific interaction between poly (2-dimethylamino ethylmethacrylate) based polymers with platelets and red blood cells. Luca Flebus (a), François Lombart (a), Lucia Martinez-Jothar (a), Chantal Sevrin (a), Céline Delierneux (b), Cécile Oury (b), Christian Grandfils (a*). a. Interfacultary Research Center of Biomaterials, University of Liège, Institute of Chemistry, Building B6C, Sart-Tilman, Liège (ZIP code 4000), Belgium. b. Laboratory of Thrombosis and Haemostasis, GIGA-Cardiovascular Sciences, University of Liège, Belgium. Luca Flebus: [email protected], Lucia Martínez -Jothar: [email protected], François Lombart: [email protected], Céline Delierneux : [email protected], Cécile Oury :[email protected], Chantal Sevrin: [email protected], Christian Grandfils : [email protected] *CORRESPONDING AUTHOR: Christian Grandfils, Interfacultary Research Center of Biomaterials (CEIB), University of Liège, Institute of Chemistry, Building B6C, Sart-Tilman (Liège), Liège (ZIP code 4000), Belgium. [email protected] Phone number: + 32 (0)4 366 3506. Fax number: + 32 (0) 43663623.
1
Graphical abstract
ABSTRACT Poly(2-dimethylamino)ethyl methacrylate (PDMAEMA) is an attractive polycation frequently proposed as a non-viral vector for gene therapy. As expected for other cationic carriers, intravenous administration of PDMAEMA can result in its ionic complexation with various negatively charged domains found within the blood. To gain more insight into this polycation hemoreactivity, we followed the binding kinetics of a free form (FF) of fluorescein labelled PDMAEMA (below 15 kDa) in normal human blood using flow cytometry. This in vitro study highlighted that platelets display higher affinity for this polycation compared to red blood cells (RBCs), with an adsorption isotherm characteristics of a specific saturable binding site. PDMAEMA (1-20 µg/mL) exerted a concentration dependent proaggregant effect with a
2
biphasic aggregation of washed platelets. Activation of platelets was also noticed in whole blood with the expression of P-selectin and fibrinogen on platelet surface. Although additional studies would be needed in order to elucidate the mechanism of PDMAEMA mediated activation of platelets, our manuscript provides important information on the hemoreactivity of FF PDMAEMA.
KEYWORDS:
Poly
(2-dimethylamino-ethylmethacrylate),
blood
cells,
polycation,
hemocompatibility
3
INTRODUCTION PDMAEMA is one of the most attractive polymer proposed as a non-viral vector for gene therapy (Zhang et al., 2004). Even if polyethylenimine (PEI) has been considered as the gold standard of non-viral vectors for several years, PDMAEMA has gained in popularity due to a high transfection efficiency combined with a lower cytotoxicity (Cerda-Cristerna et al., 2011; Dubruel et al., 2003; Robbens et al., 2010; Schallon et al., 2010). This safety aspect requires major attention for intravenous administration due to the well-known reactivity of the blood and the potential risk of ionic interaction of polycations with various negatively charged domains found on blood components such as RBCs or plasma proteins (Domurado et al., 2001). For instance, it is widely known that polycation–RBC interactions can provoke in vitro cell aggregation (hemagglutination) or hemolysis (Domurado et al., 2001; Fischer et al., 2003; Lv et al., 2006; Moreau et al., 2002; Verbaan et al., 2003). It is also worth to stress that the majority of in vitro and in vivo toxicity studies published until now on PDMAEMA have been focused on the cell reactivity of polyplexes (DNA and polycation complexes) and not on the free form (FF) of this synthetic polymer (Lin et al., 2008; Pirotton et al., 2004; Sharma et al., 2008; Van der Aa et al., 2007; Verbaan et al., 2003). Moreover, from our knowledge, no attention has been paid on the kinetics and selectivity of blood cell interaction with PDMAEMA. In line with this, our research team has recently assessed the in vitro hemocompatibility of FF of PDMAEMA with the aim to draw clear relationships between their macromolecular properties (i.e. Mw, charge density and architecture) and their blood reactivity (CerdaCristerna et al., 2011). Although limited to standard ISO tests, these results have highlighted that PDMAEMAs did not cause hemolysis whatever their concentration (between 10 to 200 µg/mL) or Mw (10 to 40kDa) and irrespectively of the duration of incubation time (up to 120
4
min). These findings differ from the data reported earlier by Dubruel et al. (Dubruel et al., 2003) and Moreau et al. (Moreau et al., 2000) who demonstrated that PDMAEMA could induce a lysis of erythrocytes, depending on the Mw, concentration, and incubation time. This difference has been interpreted as a "buffering" role for plasma proteins, which were excluded from these former studies. Because most blood proteins are anionic under physiological conditions, any circulating cationic macromolecule would most likely interact with plasma proteins. Through their rapid binding to polycations with formation of polyelectrolyte complexes, plasma proteins could significantly alter the affinity and selectivity of polycations for biological sites (Cerda-Cristerna et al., 2011; Domurado et al., 2001; Moreau et al., 2000). Nevertheless, proteins and RBCs are not the only blood elements capable of interacting with polycations. In particular, and thanks to complex and sensitive receptors anchored to their plasma membrane, platelets are well-known to be extremely sensitive and responsive to foreign elements (Gorbet and Sefton, 2004; Simak, 2009). As a consequence, biomaterials, in particular if they are in the nanosize range, can trigger platelet adhesion, activation and thrombocytopenia (Dobrovolskaia et al., 2012; Semberova et al., 2009). With a negative Zeta potential, platelets are also more prone to interact rapidly with polycations (Barry and Gralnick, 1977; Donato et al., 1996; Eika, 1972; Jenkins et al., 1971). The existence of a specific polycation binding site associated to the plasma membrane of human platelets has been postulated to explain their reactivity towards natural basic proteins released by activated leukocytes and platelets themselves (Donato et al., 1996). In addition, other polycations differing in molecular weight, but also in repetitive units have been disclosed to selectively induce agglutination and/or aggregation of platelets. For instance, ristocetin (a cationic antibiotic) is able to induce agglutination of platelets, while poly-L-lysine (PLL), protamine, or polybrene have been shown to rapidly induce platelet aggregation in vitro (Barry et al., 1977; Eika, 1972; Jenkins et al., 1971).
5
According to a selective recognition process, Donato et al. (Donato et al., 1996), Leventhal et al. (Leventhal and Bertics, 1993) and Sushkevich et al. (Sushkevich et al., 1977) have highlighted that polycation mediated platelet aggregation could trigger specific endogenous cell activation such as protein phosphorylation and intracellular signals following an increase of intracellular calcium ions. In the particular case of PDMAEMA, those biological events have not been studied yet. Until now, we only noticed a reduction in platelet count assigned to platelet aggregation (CerdaCristerna et al., 2011; Yancheva et al., 2007). These last authors have noticed that this drop in platelet count occurred independently of the PDMAEMA molecular weight, (from 10 kDa to 40 kDa), which is in contrast to the induced RBC aggregation (Cerda-Cristerna et al., 2011). In the perspective to adopt PDMAEMA for new cardiovascular applications (Flebus et al., 2015) we have assessed the kinetics, selectivity and the mechanisms of interaction of a low molecular weight PDMAEMA (below 15 kDa) with human RBCs and platelets. Low Mw polycation was selected in order to facilitate the future clearance by the kidneys (Flebus et al., 2015). To serve this aim, we used a fluorescent form of this polycation previously described by our research team (Flebus et al., 2015). Flow cytometry and aggregometry tests were carried out in order to evaluate the platelet responses elicited by PDMAEMA. Our observations were compared to protamine sulfate (PS), a natural polycation currently used in clinical settings.
2. MATERIALS AND METHODS 2.1. Products 2-(dimethylamino) ethyl methacrylate (DMAEMA), fluorescein acrylate and grade I apyrase from potatoes were purchased from Sigma Aldrich. PS (10 mg/mL) was acquired from LEO pharma. Collagen HORM was purchased from Takeda Austria GmbH. Bovine serum albumin
6
was purchased from US Biologicals. Anti-human CD61 (PerCP) and anti-human CD62P (or P-selectin) were from BD Biosciences, while polyclonal rabbit anti-human fibrinogen FITC was provided by DAKO. Tirofiban (Aggrastat) was purchased from Correvio. 2.2. Synthesis and characterization of PDMAEMA-based polymers PDMAEMA (Mn: 15 kDa) and PDMAEMA-g-fluorescein (Mn: 14 kDa) were synthetized by atom-transfer radical polymerization (ATRP) (Flebus et al., 2015). After polymerization, the polymers were purified in three successive steps consisting of chromatography on alumina support, precipitation in heptane and dialysis (cut-off: 10 kDa) against water. Once purified, the polymers were dried by lyophilization. All these procedures have been detailed in one of our previous paper (Flebus et al., 2015). Mw, polydispersity and homogeneity of fluorescent labelling were determined by size exclusion chromatography (SEC) performed in THF (TEA 2%) using fluorescent and RI detectors in series. Purity and fluorescent grafting were verified by 1H.NMR realized in CDCl3. The proton spectra were acquired with a NMR spectrometer Bruker (250 MHz) at 298°K (CREMAN, NMR Center of the University of Liege). NMR spectra were processed with MestReNova software (V. 5.304536), Mestrelab Research (Spain). Aqueous stock polymer solutions (1 mg/mL) were prepared in phosphate buffer saline (PBS) and stored at −20°C until their future use. 2.3. Blood collection and washed platelet preparation Blood samples were collected by venipuncture from healthy donors who have signed an informed consent. This study received the approval of the local Ethics Committee (CHU, Liège, 2004-Mc28F-Ip-ad-041201). Blood samples for flow cytometry experiments were collected in 3.2% citrate, stored at RT and were used within less than 1 hour to prevent any loss in platelet reactivity.
7
For washed platelet preparation, blood samples were collected in 6:1 v/v of Acid-CitrateDextrose (ACD, 7.7 mM citric acid, 93 mM sodium citrate, 1.4 mM dextrose) containing 1 U/mL apyrase. Washed platelet suspensions were prepared by whole blood centrifugation at 100 g for 15 min at RT. The platelet-rich plasma (PRP) was carefully recovered and gently transferred in a new 15 mL Falcon tube. ACD solution + apyrase (1 U/mL) was slowly added to the PRP (2:1 v/v) and the sample was centrifuged at 800 g for 10 min. The supernatant was discarded and the liquid droplets adhering to the wall of the tube were wiped out using paper towels. Washed platelets were resuspended in Tyrode’s buffer (137 mM NaCl, 12 mM NaHCO3, 2 mM KCl, 0.34 mM Na2HPO4, 1 mM MgCl2, 5.5 mM glucose and 5 mM Hepes) containing 0.35 % bovine serum albumin and 1 U/mL of apyrase. Washed platelets were kept at RT for about 30 min before starting the tests and were used in the following 2 hours. 2.4. Study of PDMAEMA interaction with blood cells by flow cytometric analysis (FACS) 50 µL of the polymer solution were added to 450 µL of either whole blood (collected in 3.2% citrate) or blood diluted 10x, 100x or 500x in PBS. In order to ensure a rapid and reproducible homogenization of the polycation solution in blood we adopted the following detailed procedure. Within polystyrene tubes of 5 mL, nine volumes of whole blood were rapidly (less than 1 second) added to one volume of a polycation solution. Immediately after, the mixture was homogenized by 3 up-and-down pipetting. The samples were incubated in the dark at RT during the different incubation times required before analysis. Negative controls were made in a similar way, but PBS was used instead of polycation solution. Flow cytometer analysis was carried out with a FACSCanto II (Becton Dickinson) recording the fluorescein signal associated with our polymer using the FITC channel. These analyses were performed applying a flow-rate ranging between 10 and 120 µL/min. A total of 10,000 events were recorded. The cell histograms were analyzed using the BD FACSDiva 6.1.2
8
software. The data presented in the result section on the figure 2 are representative of at least three independent experiments. 2.5. Aggregometry tests in washed platelets Platelet aggregation induced by PDMAEMA or PS was evaluated by optical aggregometry in a Chrono-log 700 (Kordia, NL). In brief, cuvettes (Chrono-log, ref. P/N 312) containing 270 μL of washed platelets (250, 000 platelets/µL) and a siliconized stir bar were incubated at 37°C for 5 min. Afterwards, one of these cuvettes was introduced in each of the two channels. 1 min after, 30 μL of PDMAEMA or PS were added to the test cuvette in order to obtain final concentrations of 1, 5, 10 and 20 μg/mL. We have selected a range of protamine sulfate concentration slightly below the typical concentration met in clinic (i.e. 30 µg/mL) in view to fit to the aggregation window measurable with our equipment. Changes in light transmission were recorded for 8 min. PBS and collagen (collagen horm 10 μg/mL) were used as negative and positive controls, respectively. Aggregation curves were digitalized using a graph digitizing software, (DigitizeIt, Germany). The data presented in table 1 are representative of at least three independent experiments. 2.6. Platelet aggregation induced by polycations in the presence of heparin As described in section 2.5, cuvettes containing 270 μL of washed platelets (250,000 platelets/µL) were incubated at 37°C for 5 min. 30 µL of a solution of Grade I-A heparin sodium salt from porcine intestinal mucosa (Sigma Aldrich, ref. H3393, ≥180 USP units/mg) was added to the test sample. The same volume of PBS was added to the control. Samples were pre-incubated for 5 min (37°C, 1200 rpm), and 30 µL of PDMAEMA or PS (10 µg/mL) was then added to the test and control samples respectively. The ratio between polymer and heparin was 1:1. Platelet response was recorded for additional 14 min.
9
2.7. Analysis of platelet activation by expression of CD61, CD62P and fibrinogen through flow cytometry The procedure was adapted from the “Platelet activation, staining and analysis” protocol published by BD Biosciences (Support protocols Platelet Activation, Staining, and Analysis). In brief, citrated-blood was incubated for 15 min at RT with PS or PDMAEMA (final concentrations of 1, 5, 10 or 20 μg/mL). Negative (PBS) and positive controls (collagen: 10µg/mL) were included in each series. Afterwards, samples were fixed in cold paraformaldehyde (Paraformaldehyde (PFA) 1% prepared in PBS) for 12 hours. PFA was eliminated by two washing steps in PBS. Cells were then incubated for 15 min in the dark with a mixture of a PercP-conjugated anti-human CD61 antibody, a PE-anti-human CD62P antibody and a FITC-anti-human fibrinogen antibody. Samples were fixed by addition of cold PFA 1%. Fluorescent signals were acquired on 10,000 platelet events on a FACSCanto II, using a blue filter (515-545 nm for FITC, 564 – 606 nm for PE and 670 longpass filter for PerCP). BD FACS Diva software was used to determine the percentage of CD61-positive cells expressing CD62P and/or binding fibrinogen. The data presented in the result section are representative of at least three independent experiments. 2.8. Statistical test Statistical differences were calculated with the Student t-test considering a p-value of 0.05 (Microsoft Office 2010, Excel; Microsoft, Redmond, WA, USA). Figure 2, 6 and table 1 have been analyzed by the statistical test.
3. Results 3.1. Kinetics and adsorption isotherm of PDMAEMA to blood cells We used a fluorescent form of PDMAEMA recently synthesized in our research unit (Flebus et al., 2015) and flow cytometry to assess the rate but also the selectivity of PDMAEMA
10
interaction with the various blood cells. To first optimize the conditions of this investigation, a kinetics study has been carried out considering a fixed polycation concentration of 10 µg/mL within a blood cell suspension of ~ 9000 cells/µL. This blood cell concentration was selected to discriminate RBCs from platelets while avoiding any clogging of the flow cytometer capillary. Flow cytometry allowed us to follow the kinetics of interaction of the fluorescent polycation with erythrocytes and platelets in an easy and fast way. This approach is however limited by the impossibility to convert fluorescence intensity data in absolute amount of polymer adsorbed/uptake per cell. It is also worth to remember that fluorescence intensity could be significantly affected by the microenvironment encountered by the fluorophore. Keeping in mind these limitations, we compared the fluorescent intensity signals integrated for platelets and RBCs beyond a threshold of 300 FITC units per cell type (RBCs and platelets). Thus, to follow the kinetics of the interactions, we calculated the percentage of labeled cells from the total population corresponding to RBCs or platelets.
Careful analysis of the counting events (figure 1A) highlighted that labeled platelet population increased strongly with the interaction time. Thus, while the fluorescent platelet population was equal to 0 % of observed events at the beginning (incubation time 0), platelets increased steadily to finally reached 16.9 % after 3 hours of incubation. Therefore, this result supports the existence of a specific interaction of platelets with the fluorescent polycation. By comparison, the population of fluorescent RBCs remains very low for all incubation times and the percentage of labeled RBC population only reached 0.3% after 3 hours (Figure 1B). In addition, we can derive that the surface exposed by a PDMAEMA molecule in a solidliquid interface is approximatively equal to 20nm2 (based on Mw±10kDa and hydrodynamic
11
diameter: 10 nm). From this approximation, we calculated a total surface developed by PDMAEMA molecules per µm2 of blood cell surface equal to 0.002 µm2. Accordingly to this rough estimation, we increased polycation concentration to try to achieve an excess of free PDMAEMA to its potential binding sites (i.e. number of PDMAEMA molecules per cell). As detailed in the Materials and Methods section 2.4, this study has been conducted by exposing 500-fold diluted blood within a PDMAEMA solution 10 µg/mL. As expected from the exploratory kinetics study, the isotherm adsorption curve given on figure 2 highlighted a sharp difference in affinity of PDMAEMA between platelets and RBCs. Indeed, in the range of cell dose investigated, i.e. from 125.103 to 65.106 PDMAEMA molecules/blood cell, a classical isotherm adsorption is observed for human platelets while a linear relationship is observed for the RBCs curve. Whereas the RBC interactivity profile is indicative of a nonspecific binding site, the asymptotic curve displayed by the adsorption of PDMAEMA to platelets is clearly indicative of the presence of a specific binding site of this polycation on platelet surface. The dissociation constant (Kd) calculated from this isotherm curve corresponds to a dose of 450 x 103 PDMAEMA molecules/blood cell. From this isotherm curve, a total amount of 1.25 x 106 PDMAEMA molecules/blood cell to achieve saturation can be extracted. RBCs also interacted with PDMAEMA but with markedly less affinity than platelets. For instance, at a dose of 1.25 x 106 molecules per cell only ~ 7% of the RBCs had a fluorescence intensity exceeding the selected threshold, while by comparison 92% of platelets were positive.
3.2 Aggregometry analysis of washed platelets induced by polycations The selectivity displayed by platelets to interact with PDMAEMA prompted us to further investigate the biological outcome(s) of these cells in terms of aggregation and/or activation. For this purpose, aggregometry tests were carried out in washed platelets and using collagen 12
(10 µg/mL) as a well-known positive control. These observations were compared with PS, a natural polycation, daily used in clinics for neutralization of heparin during cardiovascular surgery. Collagen is a naturally strong platelet agonist giving rise to a single-phase and rapid aggregation (Gratacap et al., 2009; Leo et al., 2002; Yoshida, et al., 2008) At 10 µg/mL, this agonist induced an immediate response, reaching 90% aggregation within 4 min of incubation and 100% after 8 min. The aggregation triggered by PDMAEMA evolved significantly when raising its concentration. Indeed whereas 1 µg/mL polycation concentration failed to induce any significant response, 5 µg/mL PDMAEMA elicited platelet aggregation that reached 25 to 35% after 8 min for all test subjects. When platelets were exposed to 10 or 20 µg/mL PDMAEMA, the aggregation profile was converted into two phases. The first one reached a plateau within 5 to 6 min, with a maximal platelet aggregation around 30%. Afterwards, platelet aggregation resembled to a sigmoid curve achieving about 75% aggregation at the plateau. It is noteworthy that, while response to 10 and 20 µg/mL was similar in all test subjects, platelet stimulation with 5 µg/mL PDMAEMA led to a more variable response between blood donors. While for some blood samples a slight monophasic aggregation reaching 30% was observed, others were more reactive, giving rise to the biphasic response within a time-scale of 15 min. interestingly enough, a similar concentration/aggregation profile has been observed for PS (Figure 3b). For the sake to quantify these data and facilitate the comparison between the potency of PDMAEMA and protamine, Table 1 summarizes the percentages of maximal aggregation recorded after 8 min of incubation of platelets with the two polycations. At the highest concentration evaluated (20 µg/mL) PDMAEMA and PS showed similar effects on washed platelets and led to an aggregation response ranging between 77.1 and 80.5 %.
13
Further information on the mechanism leading to platelet aggregation after stimulation with PDMAEMA was obtained by performing tests in samples preincubated during 10 min with Tirofiban (figure 3), a well-known inhibitor of GPIIb/IIIa receptor (Hashemzadeh et al., 2008). In the presence of Tirofiban (1.25 μg/mL), platelets stimulated with PDMAEMA or PS (20 μg/mL) exhibited a slight aggregation reaching approximately respectively 20 and 30 % after 10 min. This aggregation is similar to the first aggregation phase observed in the presence of PDMAEMA or PS alone (figure 4) or of low polycation concentrations. The difference noticed in platelet aggregation until 5 min is within the windows typically met between blood sample donors and is therefore not significant. These results support the assumption that the first aggregation phase induced by polycation corresponds to platelet agglutination while the second phase depends on GPIIb/IIIa engagement and fibrinogen binding.
3.3 Aggregometry analysis of washed platelets induced by polycations in the presence of heparin We compared the action of PDMAEMA on platelets in the absence or in the presence of nonfractionated heparin, a polysaccharide known to have the highest negative electrical charge density of any other biological macromolecules (Björk and Lindahl, 1982). As a clinical reference, we also assessed the neutralization of heparin with PS (Schroeder et al. 2011). Based on the results acquired in the previous section, this study was carried out by measuring the proaggregant effect of PDMAEMA on washed platelets. Thus, washed platelets were preincubated with heparin for 5 min before adding the polycation. As highlighted on figure 5, while both PDMAEMA and PS (10 μg/mL) induced a
14
strong biphasic platelet aggregation in the absence of heparin (black trace), no perceivable aggregation was noticed during the whole analysis in the presence of heparin (10 μg/mL, grey trace). The similarity of the profile between PDMAEMA and PS is also interesting to note. Indeed, the both polycations demonstrated practically the same profile whether in term of kinetics than in term of aggregation (no perceivable aggregation). The small shift noticed for signals of PDMAEMA and Heparin and not for PS is artifactual and can be assigned from the difficulty to add the reagents just at the same within the two channels of the equipment. Similar observations were also made at a final concentration of 20 μg/mL of both the polycations and heparin (data not shown).
3.4 Flow cytometric analysis of platelet activation by polycations in whole blood In order to gain future insight into the mechanisms of platelet stimulation with PDMAEMA and protamine, we analyzed fibrinogen binding and P-selectin exposure on platelet surface, two conventional markers of platelet activation. The concentration–response curves corresponding to the fibrinogen and P-selectin response clearly highlighted statistical difference in potency between the two polycations tested (figure 6) (p
S0378-5173(15)30002-8 http://dx.doi.org/doi:10.1016/j.ijpharm.2015.06.036 IJP 14988
To appear in:
International Journal of Pharmaceutics
Received date: Revised date: Accepted date:
11-5-2015 19-6-2015 20-6-2015
Please cite this article as: Flebus, Luca, Lombart, Franc¸ois, Martinez-Jothar, Lucia, Sevrin, Chantal, Delierneux, C´eline, Oury, C´ecile, Grandfils, Christian, In vitro study of the specific interaction between poly (2-dimethylamino ethylmethacrylate) based polymers with platelets and red blood cells.International Journal of Pharmaceutics http://dx.doi.org/10.1016/j.ijpharm.2015.06.036 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
In vitro study of the specific interaction between poly (2-dimethylamino ethylmethacrylate) based polymers with platelets and red blood cells. Luca Flebus (a), François Lombart (a), Lucia Martinez-Jothar (a), Chantal Sevrin (a), Céline Delierneux (b), Cécile Oury (b), Christian Grandfils (a*). a. Interfacultary Research Center of Biomaterials, University of Liège, Institute of Chemistry, Building B6C, Sart-Tilman, Liège (ZIP code 4000), Belgium. b. Laboratory of Thrombosis and Haemostasis, GIGA-Cardiovascular Sciences, University of Liège, Belgium. Luca Flebus: [email protected], Lucia Martínez -Jothar: [email protected], François Lombart: [email protected], Céline Delierneux : [email protected], Cécile Oury :[email protected], Chantal Sevrin: [email protected], Christian Grandfils : [email protected] *CORRESPONDING AUTHOR: Christian Grandfils, Interfacultary Research Center of Biomaterials (CEIB), University of Liège, Institute of Chemistry, Building B6C, Sart-Tilman (Liège), Liège (ZIP code 4000), Belgium. [email protected] Phone number: + 32 (0)4 366 3506. Fax number: + 32 (0) 43663623.
1
Graphical abstract
ABSTRACT Poly(2-dimethylamino)ethyl methacrylate (PDMAEMA) is an attractive polycation frequently proposed as a non-viral vector for gene therapy. As expected for other cationic carriers, intravenous administration of PDMAEMA can result in its ionic complexation with various negatively charged domains found within the blood. To gain more insight into this polycation hemoreactivity, we followed the binding kinetics of a free form (FF) of fluorescein labelled PDMAEMA (below 15 kDa) in normal human blood using flow cytometry. This in vitro study highlighted that platelets display higher affinity for this polycation compared to red blood cells (RBCs), with an adsorption isotherm characteristics of a specific saturable binding site. PDMAEMA (1-20 µg/mL) exerted a concentration dependent proaggregant effect with a
2
biphasic aggregation of washed platelets. Activation of platelets was also noticed in whole blood with the expression of P-selectin and fibrinogen on platelet surface. Although additional studies would be needed in order to elucidate the mechanism of PDMAEMA mediated activation of platelets, our manuscript provides important information on the hemoreactivity of FF PDMAEMA.
KEYWORDS:
Poly
(2-dimethylamino-ethylmethacrylate),
blood
cells,
polycation,
hemocompatibility
3
INTRODUCTION PDMAEMA is one of the most attractive polymer proposed as a non-viral vector for gene therapy (Zhang et al., 2004). Even if polyethylenimine (PEI) has been considered as the gold standard of non-viral vectors for several years, PDMAEMA has gained in popularity due to a high transfection efficiency combined with a lower cytotoxicity (Cerda-Cristerna et al., 2011; Dubruel et al., 2003; Robbens et al., 2010; Schallon et al., 2010). This safety aspect requires major attention for intravenous administration due to the well-known reactivity of the blood and the potential risk of ionic interaction of polycations with various negatively charged domains found on blood components such as RBCs or plasma proteins (Domurado et al., 2001). For instance, it is widely known that polycation–RBC interactions can provoke in vitro cell aggregation (hemagglutination) or hemolysis (Domurado et al., 2001; Fischer et al., 2003; Lv et al., 2006; Moreau et al., 2002; Verbaan et al., 2003). It is also worth to stress that the majority of in vitro and in vivo toxicity studies published until now on PDMAEMA have been focused on the cell reactivity of polyplexes (DNA and polycation complexes) and not on the free form (FF) of this synthetic polymer (Lin et al., 2008; Pirotton et al., 2004; Sharma et al., 2008; Van der Aa et al., 2007; Verbaan et al., 2003). Moreover, from our knowledge, no attention has been paid on the kinetics and selectivity of blood cell interaction with PDMAEMA. In line with this, our research team has recently assessed the in vitro hemocompatibility of FF of PDMAEMA with the aim to draw clear relationships between their macromolecular properties (i.e. Mw, charge density and architecture) and their blood reactivity (CerdaCristerna et al., 2011). Although limited to standard ISO tests, these results have highlighted that PDMAEMAs did not cause hemolysis whatever their concentration (between 10 to 200 µg/mL) or Mw (10 to 40kDa) and irrespectively of the duration of incubation time (up to 120
4
min). These findings differ from the data reported earlier by Dubruel et al. (Dubruel et al., 2003) and Moreau et al. (Moreau et al., 2000) who demonstrated that PDMAEMA could induce a lysis of erythrocytes, depending on the Mw, concentration, and incubation time. This difference has been interpreted as a "buffering" role for plasma proteins, which were excluded from these former studies. Because most blood proteins are anionic under physiological conditions, any circulating cationic macromolecule would most likely interact with plasma proteins. Through their rapid binding to polycations with formation of polyelectrolyte complexes, plasma proteins could significantly alter the affinity and selectivity of polycations for biological sites (Cerda-Cristerna et al., 2011; Domurado et al., 2001; Moreau et al., 2000). Nevertheless, proteins and RBCs are not the only blood elements capable of interacting with polycations. In particular, and thanks to complex and sensitive receptors anchored to their plasma membrane, platelets are well-known to be extremely sensitive and responsive to foreign elements (Gorbet and Sefton, 2004; Simak, 2009). As a consequence, biomaterials, in particular if they are in the nanosize range, can trigger platelet adhesion, activation and thrombocytopenia (Dobrovolskaia et al., 2012; Semberova et al., 2009). With a negative Zeta potential, platelets are also more prone to interact rapidly with polycations (Barry and Gralnick, 1977; Donato et al., 1996; Eika, 1972; Jenkins et al., 1971). The existence of a specific polycation binding site associated to the plasma membrane of human platelets has been postulated to explain their reactivity towards natural basic proteins released by activated leukocytes and platelets themselves (Donato et al., 1996). In addition, other polycations differing in molecular weight, but also in repetitive units have been disclosed to selectively induce agglutination and/or aggregation of platelets. For instance, ristocetin (a cationic antibiotic) is able to induce agglutination of platelets, while poly-L-lysine (PLL), protamine, or polybrene have been shown to rapidly induce platelet aggregation in vitro (Barry et al., 1977; Eika, 1972; Jenkins et al., 1971).
5
According to a selective recognition process, Donato et al. (Donato et al., 1996), Leventhal et al. (Leventhal and Bertics, 1993) and Sushkevich et al. (Sushkevich et al., 1977) have highlighted that polycation mediated platelet aggregation could trigger specific endogenous cell activation such as protein phosphorylation and intracellular signals following an increase of intracellular calcium ions. In the particular case of PDMAEMA, those biological events have not been studied yet. Until now, we only noticed a reduction in platelet count assigned to platelet aggregation (CerdaCristerna et al., 2011; Yancheva et al., 2007). These last authors have noticed that this drop in platelet count occurred independently of the PDMAEMA molecular weight, (from 10 kDa to 40 kDa), which is in contrast to the induced RBC aggregation (Cerda-Cristerna et al., 2011). In the perspective to adopt PDMAEMA for new cardiovascular applications (Flebus et al., 2015) we have assessed the kinetics, selectivity and the mechanisms of interaction of a low molecular weight PDMAEMA (below 15 kDa) with human RBCs and platelets. Low Mw polycation was selected in order to facilitate the future clearance by the kidneys (Flebus et al., 2015). To serve this aim, we used a fluorescent form of this polycation previously described by our research team (Flebus et al., 2015). Flow cytometry and aggregometry tests were carried out in order to evaluate the platelet responses elicited by PDMAEMA. Our observations were compared to protamine sulfate (PS), a natural polycation currently used in clinical settings.
2. MATERIALS AND METHODS 2.1. Products 2-(dimethylamino) ethyl methacrylate (DMAEMA), fluorescein acrylate and grade I apyrase from potatoes were purchased from Sigma Aldrich. PS (10 mg/mL) was acquired from LEO pharma. Collagen HORM was purchased from Takeda Austria GmbH. Bovine serum albumin
6
was purchased from US Biologicals. Anti-human CD61 (PerCP) and anti-human CD62P (or P-selectin) were from BD Biosciences, while polyclonal rabbit anti-human fibrinogen FITC was provided by DAKO. Tirofiban (Aggrastat) was purchased from Correvio. 2.2. Synthesis and characterization of PDMAEMA-based polymers PDMAEMA (Mn: 15 kDa) and PDMAEMA-g-fluorescein (Mn: 14 kDa) were synthetized by atom-transfer radical polymerization (ATRP) (Flebus et al., 2015). After polymerization, the polymers were purified in three successive steps consisting of chromatography on alumina support, precipitation in heptane and dialysis (cut-off: 10 kDa) against water. Once purified, the polymers were dried by lyophilization. All these procedures have been detailed in one of our previous paper (Flebus et al., 2015). Mw, polydispersity and homogeneity of fluorescent labelling were determined by size exclusion chromatography (SEC) performed in THF (TEA 2%) using fluorescent and RI detectors in series. Purity and fluorescent grafting were verified by 1H.NMR realized in CDCl3. The proton spectra were acquired with a NMR spectrometer Bruker (250 MHz) at 298°K (CREMAN, NMR Center of the University of Liege). NMR spectra were processed with MestReNova software (V. 5.304536), Mestrelab Research (Spain). Aqueous stock polymer solutions (1 mg/mL) were prepared in phosphate buffer saline (PBS) and stored at −20°C until their future use. 2.3. Blood collection and washed platelet preparation Blood samples were collected by venipuncture from healthy donors who have signed an informed consent. This study received the approval of the local Ethics Committee (CHU, Liège, 2004-Mc28F-Ip-ad-041201). Blood samples for flow cytometry experiments were collected in 3.2% citrate, stored at RT and were used within less than 1 hour to prevent any loss in platelet reactivity.
7
For washed platelet preparation, blood samples were collected in 6:1 v/v of Acid-CitrateDextrose (ACD, 7.7 mM citric acid, 93 mM sodium citrate, 1.4 mM dextrose) containing 1 U/mL apyrase. Washed platelet suspensions were prepared by whole blood centrifugation at 100 g for 15 min at RT. The platelet-rich plasma (PRP) was carefully recovered and gently transferred in a new 15 mL Falcon tube. ACD solution + apyrase (1 U/mL) was slowly added to the PRP (2:1 v/v) and the sample was centrifuged at 800 g for 10 min. The supernatant was discarded and the liquid droplets adhering to the wall of the tube were wiped out using paper towels. Washed platelets were resuspended in Tyrode’s buffer (137 mM NaCl, 12 mM NaHCO3, 2 mM KCl, 0.34 mM Na2HPO4, 1 mM MgCl2, 5.5 mM glucose and 5 mM Hepes) containing 0.35 % bovine serum albumin and 1 U/mL of apyrase. Washed platelets were kept at RT for about 30 min before starting the tests and were used in the following 2 hours. 2.4. Study of PDMAEMA interaction with blood cells by flow cytometric analysis (FACS) 50 µL of the polymer solution were added to 450 µL of either whole blood (collected in 3.2% citrate) or blood diluted 10x, 100x or 500x in PBS. In order to ensure a rapid and reproducible homogenization of the polycation solution in blood we adopted the following detailed procedure. Within polystyrene tubes of 5 mL, nine volumes of whole blood were rapidly (less than 1 second) added to one volume of a polycation solution. Immediately after, the mixture was homogenized by 3 up-and-down pipetting. The samples were incubated in the dark at RT during the different incubation times required before analysis. Negative controls were made in a similar way, but PBS was used instead of polycation solution. Flow cytometer analysis was carried out with a FACSCanto II (Becton Dickinson) recording the fluorescein signal associated with our polymer using the FITC channel. These analyses were performed applying a flow-rate ranging between 10 and 120 µL/min. A total of 10,000 events were recorded. The cell histograms were analyzed using the BD FACSDiva 6.1.2
8
software. The data presented in the result section on the figure 2 are representative of at least three independent experiments. 2.5. Aggregometry tests in washed platelets Platelet aggregation induced by PDMAEMA or PS was evaluated by optical aggregometry in a Chrono-log 700 (Kordia, NL). In brief, cuvettes (Chrono-log, ref. P/N 312) containing 270 μL of washed platelets (250, 000 platelets/µL) and a siliconized stir bar were incubated at 37°C for 5 min. Afterwards, one of these cuvettes was introduced in each of the two channels. 1 min after, 30 μL of PDMAEMA or PS were added to the test cuvette in order to obtain final concentrations of 1, 5, 10 and 20 μg/mL. We have selected a range of protamine sulfate concentration slightly below the typical concentration met in clinic (i.e. 30 µg/mL) in view to fit to the aggregation window measurable with our equipment. Changes in light transmission were recorded for 8 min. PBS and collagen (collagen horm 10 μg/mL) were used as negative and positive controls, respectively. Aggregation curves were digitalized using a graph digitizing software, (DigitizeIt, Germany). The data presented in table 1 are representative of at least three independent experiments. 2.6. Platelet aggregation induced by polycations in the presence of heparin As described in section 2.5, cuvettes containing 270 μL of washed platelets (250,000 platelets/µL) were incubated at 37°C for 5 min. 30 µL of a solution of Grade I-A heparin sodium salt from porcine intestinal mucosa (Sigma Aldrich, ref. H3393, ≥180 USP units/mg) was added to the test sample. The same volume of PBS was added to the control. Samples were pre-incubated for 5 min (37°C, 1200 rpm), and 30 µL of PDMAEMA or PS (10 µg/mL) was then added to the test and control samples respectively. The ratio between polymer and heparin was 1:1. Platelet response was recorded for additional 14 min.
9
2.7. Analysis of platelet activation by expression of CD61, CD62P and fibrinogen through flow cytometry The procedure was adapted from the “Platelet activation, staining and analysis” protocol published by BD Biosciences (Support protocols Platelet Activation, Staining, and Analysis). In brief, citrated-blood was incubated for 15 min at RT with PS or PDMAEMA (final concentrations of 1, 5, 10 or 20 μg/mL). Negative (PBS) and positive controls (collagen: 10µg/mL) were included in each series. Afterwards, samples were fixed in cold paraformaldehyde (Paraformaldehyde (PFA) 1% prepared in PBS) for 12 hours. PFA was eliminated by two washing steps in PBS. Cells were then incubated for 15 min in the dark with a mixture of a PercP-conjugated anti-human CD61 antibody, a PE-anti-human CD62P antibody and a FITC-anti-human fibrinogen antibody. Samples were fixed by addition of cold PFA 1%. Fluorescent signals were acquired on 10,000 platelet events on a FACSCanto II, using a blue filter (515-545 nm for FITC, 564 – 606 nm for PE and 670 longpass filter for PerCP). BD FACS Diva software was used to determine the percentage of CD61-positive cells expressing CD62P and/or binding fibrinogen. The data presented in the result section are representative of at least three independent experiments. 2.8. Statistical test Statistical differences were calculated with the Student t-test considering a p-value of 0.05 (Microsoft Office 2010, Excel; Microsoft, Redmond, WA, USA). Figure 2, 6 and table 1 have been analyzed by the statistical test.
3. Results 3.1. Kinetics and adsorption isotherm of PDMAEMA to blood cells We used a fluorescent form of PDMAEMA recently synthesized in our research unit (Flebus et al., 2015) and flow cytometry to assess the rate but also the selectivity of PDMAEMA
10
interaction with the various blood cells. To first optimize the conditions of this investigation, a kinetics study has been carried out considering a fixed polycation concentration of 10 µg/mL within a blood cell suspension of ~ 9000 cells/µL. This blood cell concentration was selected to discriminate RBCs from platelets while avoiding any clogging of the flow cytometer capillary. Flow cytometry allowed us to follow the kinetics of interaction of the fluorescent polycation with erythrocytes and platelets in an easy and fast way. This approach is however limited by the impossibility to convert fluorescence intensity data in absolute amount of polymer adsorbed/uptake per cell. It is also worth to remember that fluorescence intensity could be significantly affected by the microenvironment encountered by the fluorophore. Keeping in mind these limitations, we compared the fluorescent intensity signals integrated for platelets and RBCs beyond a threshold of 300 FITC units per cell type (RBCs and platelets). Thus, to follow the kinetics of the interactions, we calculated the percentage of labeled cells from the total population corresponding to RBCs or platelets.
Careful analysis of the counting events (figure 1A) highlighted that labeled platelet population increased strongly with the interaction time. Thus, while the fluorescent platelet population was equal to 0 % of observed events at the beginning (incubation time 0), platelets increased steadily to finally reached 16.9 % after 3 hours of incubation. Therefore, this result supports the existence of a specific interaction of platelets with the fluorescent polycation. By comparison, the population of fluorescent RBCs remains very low for all incubation times and the percentage of labeled RBC population only reached 0.3% after 3 hours (Figure 1B). In addition, we can derive that the surface exposed by a PDMAEMA molecule in a solidliquid interface is approximatively equal to 20nm2 (based on Mw±10kDa and hydrodynamic
11
diameter: 10 nm). From this approximation, we calculated a total surface developed by PDMAEMA molecules per µm2 of blood cell surface equal to 0.002 µm2. Accordingly to this rough estimation, we increased polycation concentration to try to achieve an excess of free PDMAEMA to its potential binding sites (i.e. number of PDMAEMA molecules per cell). As detailed in the Materials and Methods section 2.4, this study has been conducted by exposing 500-fold diluted blood within a PDMAEMA solution 10 µg/mL. As expected from the exploratory kinetics study, the isotherm adsorption curve given on figure 2 highlighted a sharp difference in affinity of PDMAEMA between platelets and RBCs. Indeed, in the range of cell dose investigated, i.e. from 125.103 to 65.106 PDMAEMA molecules/blood cell, a classical isotherm adsorption is observed for human platelets while a linear relationship is observed for the RBCs curve. Whereas the RBC interactivity profile is indicative of a nonspecific binding site, the asymptotic curve displayed by the adsorption of PDMAEMA to platelets is clearly indicative of the presence of a specific binding site of this polycation on platelet surface. The dissociation constant (Kd) calculated from this isotherm curve corresponds to a dose of 450 x 103 PDMAEMA molecules/blood cell. From this isotherm curve, a total amount of 1.25 x 106 PDMAEMA molecules/blood cell to achieve saturation can be extracted. RBCs also interacted with PDMAEMA but with markedly less affinity than platelets. For instance, at a dose of 1.25 x 106 molecules per cell only ~ 7% of the RBCs had a fluorescence intensity exceeding the selected threshold, while by comparison 92% of platelets were positive.
3.2 Aggregometry analysis of washed platelets induced by polycations The selectivity displayed by platelets to interact with PDMAEMA prompted us to further investigate the biological outcome(s) of these cells in terms of aggregation and/or activation. For this purpose, aggregometry tests were carried out in washed platelets and using collagen 12
(10 µg/mL) as a well-known positive control. These observations were compared with PS, a natural polycation, daily used in clinics for neutralization of heparin during cardiovascular surgery. Collagen is a naturally strong platelet agonist giving rise to a single-phase and rapid aggregation (Gratacap et al., 2009; Leo et al., 2002; Yoshida, et al., 2008) At 10 µg/mL, this agonist induced an immediate response, reaching 90% aggregation within 4 min of incubation and 100% after 8 min. The aggregation triggered by PDMAEMA evolved significantly when raising its concentration. Indeed whereas 1 µg/mL polycation concentration failed to induce any significant response, 5 µg/mL PDMAEMA elicited platelet aggregation that reached 25 to 35% after 8 min for all test subjects. When platelets were exposed to 10 or 20 µg/mL PDMAEMA, the aggregation profile was converted into two phases. The first one reached a plateau within 5 to 6 min, with a maximal platelet aggregation around 30%. Afterwards, platelet aggregation resembled to a sigmoid curve achieving about 75% aggregation at the plateau. It is noteworthy that, while response to 10 and 20 µg/mL was similar in all test subjects, platelet stimulation with 5 µg/mL PDMAEMA led to a more variable response between blood donors. While for some blood samples a slight monophasic aggregation reaching 30% was observed, others were more reactive, giving rise to the biphasic response within a time-scale of 15 min. interestingly enough, a similar concentration/aggregation profile has been observed for PS (Figure 3b). For the sake to quantify these data and facilitate the comparison between the potency of PDMAEMA and protamine, Table 1 summarizes the percentages of maximal aggregation recorded after 8 min of incubation of platelets with the two polycations. At the highest concentration evaluated (20 µg/mL) PDMAEMA and PS showed similar effects on washed platelets and led to an aggregation response ranging between 77.1 and 80.5 %.
13
Further information on the mechanism leading to platelet aggregation after stimulation with PDMAEMA was obtained by performing tests in samples preincubated during 10 min with Tirofiban (figure 3), a well-known inhibitor of GPIIb/IIIa receptor (Hashemzadeh et al., 2008). In the presence of Tirofiban (1.25 μg/mL), platelets stimulated with PDMAEMA or PS (20 μg/mL) exhibited a slight aggregation reaching approximately respectively 20 and 30 % after 10 min. This aggregation is similar to the first aggregation phase observed in the presence of PDMAEMA or PS alone (figure 4) or of low polycation concentrations. The difference noticed in platelet aggregation until 5 min is within the windows typically met between blood sample donors and is therefore not significant. These results support the assumption that the first aggregation phase induced by polycation corresponds to platelet agglutination while the second phase depends on GPIIb/IIIa engagement and fibrinogen binding.
3.3 Aggregometry analysis of washed platelets induced by polycations in the presence of heparin We compared the action of PDMAEMA on platelets in the absence or in the presence of nonfractionated heparin, a polysaccharide known to have the highest negative electrical charge density of any other biological macromolecules (Björk and Lindahl, 1982). As a clinical reference, we also assessed the neutralization of heparin with PS (Schroeder et al. 2011). Based on the results acquired in the previous section, this study was carried out by measuring the proaggregant effect of PDMAEMA on washed platelets. Thus, washed platelets were preincubated with heparin for 5 min before adding the polycation. As highlighted on figure 5, while both PDMAEMA and PS (10 μg/mL) induced a
14
strong biphasic platelet aggregation in the absence of heparin (black trace), no perceivable aggregation was noticed during the whole analysis in the presence of heparin (10 μg/mL, grey trace). The similarity of the profile between PDMAEMA and PS is also interesting to note. Indeed, the both polycations demonstrated practically the same profile whether in term of kinetics than in term of aggregation (no perceivable aggregation). The small shift noticed for signals of PDMAEMA and Heparin and not for PS is artifactual and can be assigned from the difficulty to add the reagents just at the same within the two channels of the equipment. Similar observations were also made at a final concentration of 20 μg/mL of both the polycations and heparin (data not shown).
3.4 Flow cytometric analysis of platelet activation by polycations in whole blood In order to gain future insight into the mechanisms of platelet stimulation with PDMAEMA and protamine, we analyzed fibrinogen binding and P-selectin exposure on platelet surface, two conventional markers of platelet activation. The concentration–response curves corresponding to the fibrinogen and P-selectin response clearly highlighted statistical difference in potency between the two polycations tested (figure 6) (p
