Immunocompatibility of Rad-PC-Rad liposomes in vitro, based on human complement activation and cytokine release
- Matviykiv, Sofiya “Sofiya Matviykiv” (Biomaterials Science Center)
- Buscema, Marzia “Marzia Buscema” (Biomaterials Science Center)
- Gerganova, Gabriela “Gabriela Gerganova” (Biomaterials Science Center)
- Mészáros, Tamás “Tamás Mészáros” (Nanomedicine Research and Education Center; SeroScience)
- Kozma, Gergely (Tibor) “Gergely Tibor Kozma” (Nanomedicine Research and Education Center; SeroScience)
- Mettald, Ute “Ute Mettal” (Department of Chemistry)
- Neuhausd, Frederik “Frederik Neuhaus” (Department of Chemistry)
- Ishikawae, Takashi “Takashi Ishikawa” (PSI)
- Szebeni, János “János Szebeni” (Nanomedicine Research and Education Center; SeroScience; Department of Nanobiotechnology and Regenerative Medicine)
- Zumbuehl, Andreas “Andreas Zumbuehl” (Department of Chemistry)
- Müllera, Bert “Bert Müller” (Biomaterials Science Center)
- Biomaterials Science Center < Department of Biomedical Engineering < University of Basel - Allschwil, Switzerland
- Nanomedicine Research and Education Center < Institute of Pathophysiology < Semmelweis University - Budapest, Hungary
- SeroScience Ltd. - Budapest, Hungary
- Department of Chemistry < University of Fribourg - Fribourg, Switzerland
- Paul Scherrer Institute (PSI) – Villigen, Switzerland
- Department of Nanobiotechnology and Regenerative Medicine < Miskolc University - Miskolc, Hungary
Submitted: March 26, 2018; Accepted: April 19, 2018
Liposomal drug delivery systems can protect pharmaceutical substances and control their release. Systemic administration of liposomes, however, often activate the innate immune system, resulting in hypersensitivity reactions. These pseudo-allergic reactions can be interpreted as activating the complement system. Complement activation destroys and removes foreign substances, either directly through opsonization and the formation of the membrane attack complex (MAC), or by activating leukocytes and initiating inflammatory responses via mediators, such as cytokines.
In this study, we investigated the in vitro immune toxicity of the recently synthesized Rad-PC-Rad liposomes, analyzing the liposome-induced complement activation. In five human sera, Rad-PC-Rad liposomes did not induce activation, but in one serum high sensitivity via alternative pathway was detected. Such a behavior in adverse phenomena is characteristic for patient-to-patient variation and, thus, the number of donors should be in the order of hundreds rather than tens – hence the present study based on six donors has preliminary character. In order to further prove the suitability of mechano-responsive Rad-PC-Rad liposomes for clinical trials, the production of pro-inflammatory cytokines was examined by human white blood cells. The concentrations of the pro-inflammatory cytokines IL-6, IL-12p70, TNF-α, and IL-1β induced by Rad-PC-Rad liposomal formulations incubated with whole blood samples were smaller or comparable to saline (negative control). Because of this favorable in vitro hemo-compatibility, in vivo investigations using these mechano-responsive liposomes should be designed.
- non-spherical liposomes
- immune toxicity
- complement activation
- hypersensitivity reactions
- pro-inflammatory cytokines
Purpose and Rationale
A new artificial diamidophospholipids Rad-PC-Rad were recently synthesized. Rad-PC-Rad liposomes aim to preferentially deliver the vasodilator molecules to the stenosed parts of blood vessels. Liposomes, administered intravenously, are immediately exposed to a complex environment of blood cells and proteins. The adsorption of plasma proteins on the surface of liposomes may not only decrease the therapeutic efficiency and biodistribution, but also may result in immunotoxicity. The toxicities which represent the most common safety issues and reasons for nanomedicines failure include complement-mediated reactions and cytokine-mediated inflammation, which can result in anaphylaxis. Therefore, in this study we investigated complement activation and release of pro-inflammatory cytokines, mediated by Rad-PC-Rad liposomes. The immunotoxicity can be also influenced by the therapeutic payload or addition of surface ligands. Therefore, the comparison between NGT-loaded and drug-free liposomes, as well as PEGylated and non-PEGylated liposomes was evaluated. The physicochemical properties of nanomedicines are crucial to determine their interaction with the immune system. Hence, the Rad-PC-Rad liposomes we characterized in terms of size, zeta potential, and concentration.
The latest progress in the nanomedicine field has resulted in the development of smart nano-containers for drug delivery applications, including liposomes. Liposomes can improve delivery, targeting, and therapeutic efficacy of the drug, and at the same time, increase the half-life of the drug, lower its effective dose, and reduce toxic side effects [1, 2]. Previously, our research team reported on shear stress sensitive Pad-PC-Pad liposomes for targeted delivery of a vasodilator to constricted arteries [3, 4]. A further in vivo investigation of Pad-PC-Pad phospholipids was limited owing to their phase transition temperature at 37 C. Very recently, we have reported on a thermally more stable phospholipid formulation, such as Rad-PC-Rad . This lipid exhibits a bilayer main phase transition temperature of 44.7C and preserves the responsiveness for mechanical triggers .
The immediate treatment of arterial occlusion generally involves intravenous injection of nitroglycerin (NTG), which acts as a vasodilator. Systemic administration of NTG may cause severe adverse effects including hypotension and diminished blood perfusion to the heart. The targeted delivery of NTG via the incorporation into shear stress sensitive liposomes may reduce these side effects. The direct contact of liposomes with blood carries the risk of immediate activation of the innate immune system . This may result not only in the reduction of the drug’s efficacy, but also in the appearance of hypersensitivity reactions (HSRs) [6, 7, 8]. The main function of the immune system is to protect the organism from invading pathogens. It can, however, also develop an immune response against non-pathogenic objects, such as nanometer-size liposomes. Therein, the recruitment of the complement system is an important step in the recognition and elimination of foreign materials. The complement system is a group of approximately 30 plasma- and membrane-bound proteins . Their protective function leads to the release of active components, which cause opsonization, inflammation, and the generation of the membrane attack complex (MAC) . According to the current literature, the complement activation occurs via the three established routes, termed classical, lectin, and alternative pathways . One can discriminate between these pathways by identifying the presence of unique protein fragments: C4d (classical and lectin pathways) and Bb (alternative pathway) . Activation of either pathway results in the turnover of the C3 protein, which is followed by the production of the anaphylatoxins C3a and C5a, and the formation of the MAC (C5b-9). The release of anaphylatoxins causes leukocyte chemotaxis and the production of pro-inflammatory cytokines, which finally induce inflammation (Figure 1). The excessive production of anaphylatoxins can be harmful and may cause anaphylactic shock or even organ failure at relevant concentrations . Binding of the proteins to the liposomes depends also on their composition, size, geometry, surface charge, and hydrophobicity that can act as immunological adjuvant and trigger strong immune response [6, 13]. The undesirable activation of the complement system can be caused by systemically administered liposomes, such as Doxil® (PEGylated liposomal doxorubicin) and AmBisome® (liposomal amphotericin B), leading to the development of HSRs, termed complement activation-related pseudoallergy (CARPA) [7, 8]. Approximately 2–10% of patients may adversely react to intravenously administered liposomal formulations with mild-to-severe hypersensitivity reactions . CARPA develops at the first exposure and its symptoms involve almost all organ systems . Some of the most important safety concerns for nanoparticle failure are related to the toxicities caused by complement activation-mediated reactions and cytokine-mediated inflammation . Therefore, it is recommended that liposomes intended for intravenous injection are tested in vitro and in vivo for the potential activation of complement system, as a preclinical immune toxicity test . The assessment of the liposomal physicochemical properties and their impact on complement activation is also an important objective in the development of nanometer-size therapeutics.
The production of pro-inflammatory cytokines in vitro is considered a marker of cytokine-associated immunotoxicity in vivo  and screening for these toxicities early in preclinical characterization will help to avoid potentially toxic candidates in nanomedicine development. Recently, Wolf-Grosse et al. reported about cytokine secretion in a complement-dependent manner . They state that cytokine response was generally mostly due to C5a activation, as it is the most potent pro-inflammatory mediator released upon C activation . Therefore, in order to prevent the potential immunotoxicity in vivo, we studied the effect of Rad-PC-Rad liposomes on the production of complement proteins and pro-inflammatory cytokines.
In the present communication we addressed the possibilities that Rad-PC-Rad liposomes, loaded with NTG solution, would activate the complement system and stimulate the release of pro-inflammatory cytokines, thus raising concern about potential risk for CARPA or cytokine storm. Thus, we have measured in vitro complement activation in human sera and the release of the pro-inflammatory cytokines, in human whole blood and isolated leukocytes, upon incubation with Rad-PC-Rad liposomes. The complement pathway activation products C4d and Bb, and terminal complement complex SC5b-9 were measured using Enzyme Linked Immunosorbent Assay (ELISA), and the release of the pro-inflammatory cytokines IL-1, IL-6, IL-8, IL-12, TNF-α was measured using cytometric bead array test. In addition, we analyzed the liposomal physicochemical properties, in terms of liposomes size, zeta potential, using dynamic light scattering (DLS) and estimated membrane thickness from the micrographs obtained by cryogenic transmission electron microscopy (cryo-TEM).
Figure 1. Schematic representation of complement (C) activation triggering pro-inflammatory cytokines due to anaphylatoxin binding to anaphylatoxin-receptor positive cells (e.g., mast cells, basophils, neutrophils, platelets and pulmonary intravascular macrophages). In the case of Rad-PC-Rad liposomes, C activation proceeds through the alternative pathway. On this pathway C3 directly binds to liposomal phospholipid head-groups. Factor B binds to the newly attached C3b, and again becomes susceptible to cleavage by factor D. Membrane-bound C3bBb is unstable until it is bound by properdin protein (factor P). Stabilized C3-convertase rapidly generates large amounts of C3b that bind more factor B, resulting in dramatic amplification of C3b. Membrane-bound C3b serves as an opsonin and a binding tag for phagocytic cells. Addition of C3b to C3-convertase results in the formation of C5-convertase, which cleaves C5 into C5b, and proceeds to form the MAC. Cleavage of C5 also results in the formation of C5a anaphylatoxin. Together with C3a, the C5a fragment binds to the surface C receptors of mentioned allergy mediating cells. C3a and C5a receptors, after binding small anaphylatoxins, mediate the allergic reaction by stimulating the release of vasoactive mediators (e.g., histamine, thromboxanes, leukotrienes, etc).
Materials and Methods
1,3-Diheptadecanamidopropan-2-yl (2-[trimethylammonio]ethyl) phosphate (Rad-PC-Rad) was synthesized and purified according to the recently reported protocol . Figure 2 contains the structural formula of Rad-PC-Rad phospholipid. Table S1 lists all the materials used for the experiments.
Human sera from six healthy volunteers and whole blood samples from two healthy donors were obtained through an institutionally approved phlebotomy protocol at Semmelweis University (Budapest, Hungary). Human sera were stored at a temperature of −80°C until usage. Whole blood samples were freshly collected into sterile hirudin-treated tubes and immediately employed for experiment. Freshly drawn blood, used for leukocytes isolation, was provided by the Hungarian National Blood Transfusion Service.
Four Rad-PC-Rad/DSPE-PEG2000 phospholipid formulations were prepared, namely R1, R2, R3, and R4, (see Table 1). Lipids were dissolved in chloroform in molar ratios as listed in Table 1. The preparation of the liposomal formulations is described in detail in ref. . The samples were purified through sterile filters and stored at a temperature of 4C until usage.
Lipid composition (molar %)
Characterization of liposomal formulations
Phospholipid concentration. A colorimetric assay (phosphate test 2.0)  served for the determination of the phospholipid content of the liposomal formulations after extrusion and purification. Here, the phosphate moiety in the head group of the phospholipids was a measure of the total phospholipid concentration.
To exclude the variations in lipid concentration, the concentration of each liposomal formulation (R1, R2, R3, and R4) was diluted with 0.9% sodium chloride solution (saline) to a total lipid concentration of 10 mg/mL. In addition, a set of the diluted formulations (R1d, R2d, R3d, and R4d) with 5 mg/mL of phospholipids were prepared for the in vitro immunoassays to examine the impact of the lipid concentration on the complement activation level (see Figures S2 and S3).
Physicochemical characteristics. The liposome average diameter, polydispersity index (PDI) and zeta () potential were obtained by DLS performed at a temperature of 25C using a DelsaMax PRO (Beckman Coulter, USA). The suspensions were diluted 100 times in saline prior to the measurements.
Liposome morphology. The morphology of four Rad-PC-Rad formulations was studied using cryogenic transmission electron microscopy (cryo-TEM) (JEM2200FS, JEOL, Tokyo, Japan). The samples, diluted with saline in the ratio 1:1, were imaged as previously reported [5, 18].
Encapsulation efficiency. The encapsulation efficiency of Rad-PC-Rad liposomes for passive loading with nitroglycerin was determined indirectly by measuring the peak of a glucose-trifluoroacetic acid adduct by electrospray ionization mass spectrometry (ESI-MS) on a Bruker esquire HCT ion trap mass spectrometer (Bruker Corporation, Billerica, MA, USA) .
Liposomal release. Two Rad-PC-Rad liposomal formulations, i.e. with and without DSPE-PEG, were loaded with 5(6)-carboxyfluorescein (CF) buffer and prepared as described in ref. . Seven aliquots with a volume of 2 mL were separated into 5 mL glass vials and kept for selected periods of time (0, 5, 10, 20, 40 min) at a temperature of 37C. The CF release was quantified using a fluorospectrometer (SpectraMax 2, Bucher Biotec AG, Switzerland) with the wavelengths of 485 nm for excitation and 538 nm for emission. Sample fluorescence at a temperature of 20C served as a negative control (F0). As a positive control for the maximum dye release (F100), liposomal samples were heated to a temperature of 65C, above the lipids transition temperature of 44.7C. The release fraction at the selected time point x was calculated according to:
where is the fluorescence at time x.
Activation of human sera with liposomes. Human sera from six healthy donors were thawed and kept at a temperature of 4°C during the experiment. Due to the limited amount of serum available, the sera #5 and #6 were prepared as pools from distinctive donors in the ratios 1.5:1 and 7.8:1, respectively. The liposomal suspensions in the two concentrations were added to the sera of each donor in the ratio of 1:3. Saline and nitroglycerin were used as negative controls. FDA-approved liposomal drugs, with recorded cardio-toxicity effects and activation of the complement system in sensitive patients, Doxil® (2 mg/mL doxorubicin, 12.77 mg/mL phospholipids, used as provided) and AmBisome® (17.975 mg/mL amphotericin B, 4.02 mg/mL phospholipids, reconstituted with injection water) were employed as well . Zymosan (1.2 mg/mL), known as activator of the complement system, was used as positive control. Each activation mixture was incubated at a temperature of 37°C. The concentration of the terminal complement complex SC5b-9 was investigated over time. The incubation was terminated after 5, 10, 20, and 40 minutes by adding 10 mM EDTA.
The ELISA assays were carried out following the manufacturer’s protocol. The optical density was measured with a 96-well plate reader (FLUOstar Omega, BMG Labtech, Germany) at a wavelength of 450 nm for SC5b-9, Bb, C3a and C5a as well as at a wavelength of 405 nm for C4d.
Isolation of leukocytes from buffy coat. A volume of 400 mL of buffy coat (BC), a pool of white blood cells (WBCs) concentrates of four healthy volunteers, were obtained from Hungarian National Blood Transfusion Service within 24 hours of blood withdrawal. Altogether three BC pools were used, each consisting of four donors. Leukocytes were further concentrated two times by mixing with DPBS (w/o CaCl2, MgCl2) in 1:1 ratio and centrifuged for a period of 10 min at a velocity of 750 G and a temperature of 4°C. To lyse the remaining erythrocytes, distilled water at a temperature of 4°C was added to the BC (4:1 ratio) for 20 seconds. Lysis was stopped by adding one volume hyperosmotic salt solution (containing 1.8% of NaCl). After washing with ice-cold DPBS (w/o CaCl2, MgCl2) for platelets elimination, WBCs were re-suspended in R5 medium.
Qualitative and quantitative analysis of isolated leukocytes. The concentration of WBCs, in three independent blood packages was determined. Viable cells were detected using FITC Annexin V apoptosis detection kit, see Table S4. The staining procedure was performed according to the manufacturer’s instructions. Leukocytes were further diluted or concentrated to reach necessary concentration of 108 cells/mL. The cell viability was also checked after cell isolation and treatments by test materials and control agents. The viability of cells before and after treatments were always higher than 98%, except for the positive control, Table S5.
Activation of BC leukocytes with liposomes. Freshly isolated leukocytes from three independent blood packages were separately incubated with four Rad-PC-Rad liposomal formulations in the ratio 7:1. Samples, with a concentration of 4 mg/mL, were incubated for four hours at a temperature of 37°C on a shaker plate. The incubation was stopped by EDTA (final concentration 10 mM). Cell culture supernatants were further used for mixing with cytokine capture beads. The assay was performed according to the suggested protocol the manufacturer provided with the kit.
Activation of human whole blood with liposomes. Freshly collected whole human blood from two donors was separately incubated with four Rad-PC-Rad liposomal formulations at a concentration of 4 mg/mL, following the same procedure as described in the section above. The distinctive step was the incubation time. Here samples were incubated for a period of six hours. Whole blood samples had no R5 medium, instead they contained their own plasma.
Qualitative and quantitative analysis of leukocytes originated from whole blood samples. An aliquot of human blood from two donors was stained as described in the section above. Cell viability was determined before and after treatments by test materials and control agents. The percentage of apoptotic and necrotic cells was less than 3%, except for the positive control, Table S5.
Cytometric bead array test. The human inflammatory cytokines kit was used to quantitatively measure interleukin-1 (IL-), interleukin-6 (IL-6), interleukin-8 (IL-8), interleukin-12p70 (IL-12p70), tumor necrosis factor α (TNF-α), and interleukin-10 (IL-10) protein levels in the studied samples. The assay was carried out following the manufacturer’s instructions. The beads fluorescence was recorded by flow cytometry using a FACScan instrument (BD Biosciences, USA), and the data were analyzed using the Kaluza Analysis 1.5 software (Beckman Coulter, USA).
Statistical analysis was carried out using GraphPad Prism 6 (GraphPad Software Inc., San Diego, CA, USA). Data from the ELISA samples (Figure 4 and Figure 7), except zymosan, were compared with saline as negative control after 40 minutes of incubation. Significance of differences between the groups was determined by non-parametric Kruskal-Wallis test, followed by Dunn's multiple comparisons test. P-values lower than 0.05 were considered as statistically significant.
Characterization of Rad-PC-Rad liposomal formulations
The lipid concentration of Rad-PC-Rad ranged from 10 to 20 mg/mL, and the mean diameter of the liposomes in the suspensions was around 100 nm and varied from 95 to 140 nm (see Table 2). Measurements after 20 days showed that the PEGylated liposomes did not change their size, whereas the non-PEGylated ones displayed an increase from 140 to 270 nm (R1) and from 115 to 200 nm (R3). Table 2 lists the measured zeta potential values of the Rad-PC-Rad formulations. Pure Rad-PC-Rad samples revealed positive potential values, between +1.3 (R1) to +4.7 mV (R3), while PEGylated samples turned to negative potentials, from −2.0 (R4) to −4.5 mV (R2).
Lipid content (mg/mL)
Mean diameter (nm)
Membrane thickness (nm)
potential (mV) 2
The size and morphology of Rad-PC-Rad liposomes, evaluated using cryo-TEM imaging, is represented in Figure 2. These micrographs show intact spherical, lenticular, and faceted unilamellar liposomes below their main phase transition temperature. The percentage of faceted liposomes within samples was 46% (R1), 72% (R2), 42% (R3), and 52% (R4).
The addition of DSPE-PEG (see Figure 2D and Figure 2F), led to the co-existence of flat circular disks and unilamellar liposomes. Depending on the disk orientation, they appear either as small rods with high contrast (red-colored arrows), or, when seen from the top, as circular structures with low uniform contrast (Figure 2D, right).
The liposome membrane thickness was estimated from the cryo-TEM projections of the appropriately oriented membranes (Figure 2). We have measured the individual thicknesses of 100 membranes and found the mean values of R1 to be (3.27 ± 0.14) nm and of R3 to be (3.27 ± 0.19) nm, which indicates the interdigitation of the Rad-PC-Rad leaflets. Interdigitation may be one of the driving forces in the formation of faceted liposomes . Samples loaded with DSPE-PEG, i.e. R2 and R4, tended to result in higher mean values. The values correspond to (3.60 ± 0.21) nm and (3.50 ± 0.24) nm, respectively. The mean diameters of the liposomes, derived from cryo-TEM images, were 10–15 % smaller than those obtained from DLS data.
The encapsulation efficiency of NTG-loaded samples was estimated from the ESI-MS measurements. The calculation is based on the 100% ESI-MS signal of pure NTG and liposome size. The employed NTG solution contained glucose as an excipient, therefore, the NTG encapsulation was determined indirectly. The integral of the glucose-trifluoroacetic acid adduct was evaluated after NTG incorporation. The ratio between these values determines the percentage of NTG encapsulation efficiency (see Table S2). The values correspond to 38% (R3) and 12% (R4).