Comprehensive investigation of in vitro hemocompatibility of surface modified polyamidoamine nanocarrier

Abstract

BACKGROUND:

Polyamidoamine (PAMAM) dendrimers have been investigated for decades and currently applied in various areas throughout nanomedicine, including gene therapy, drug delivery, anti-bacteria and imaging. It is therefore necessary to assess cytotoxicity of PAMAM dendrimers systematically. Because blood component is usually the initial step of contact with any therapeutic agent, comprehensive hemocompatibility study is needed.

MATERIAL S AND METHODS:

The triblock dendrimer: polyamidoamine-polyethylene glycol-cyclic RGD (PAMAM-PEG-cRGD), was successfully synthesized. Various in vitro assays to characterize hemocompatibility of both PAMAM (Generation 4.0) and PAMAM-PEG-cRGD were performed, including hemolytic assay, platelet activation examination, platelet counting, assessment of coagulating pathways and evaluation of complement system activation.

RESULTS:

The hemolytic ratio of PAMAM-PEG-cRGD maintained below 5%. Surface engineering of PEG and cRGD to PAMAM attenuated hemolysis and RBC aggregation as compared with unmodified PAMAM. PAMAM (Generation 4.0) reduced platelet counting in a dose-dependent manner, and the platelet number dropped dramatically at a relatively low incubating dose (1 μM). Such surface modifications also alleviated platelet activation and platelet reduction mediated by PAMAM polycationicity. Finally, high concentration (10 μM) of PAMAM interfered the coagulation system, prolonging prothrombin time significantly.

CONCLUSION:

Keywords

Polyamidoamine (PAMAM)dendrimer nanoparticle hemocompatibility surface modification

1 Introduction

Polyamidoamine (PAMAM), a nano-scale dendrimer, has been vigorously investigated in nanomedicine, owing to its versatile and multivalent features, which enable effective and powerful delivering system that attracts interests of scientists for decades. By virtue of its unique properties as a promising nanocarrier, PAMAM has been broadly applied in miscellaneous nanomedicine-related areas, including gene therapy, drug delivery, medical imaging and so forth [1, 2]. Cationic dendritic PAMAM has been proved to promote binding ability to plasmid DNA or small interfering RNA, thus enhancing gene delivery and transfection efficiency to target cells, however, the amino terminals on PAMAM surface drive cytotoxic behavior, of which, the exact mechanism has not yet been strictly established [3, 4]. The direct application of PAMAM as a nanocarrier is therefore seriously constrained, largely because of its cytotoxic uncertainty.

In particular, the blood component biocompatibility is always the research focus, simply based on the fact that blood components most frequently constitute the initial contacts with any therapeutic agents, which are generally administered intravenously [5]. Before a nanoparticle arrives intended target tissues, it regularly confronts with blood components including blood cells, coagulating factors and complements, and polycationicity has electrostatic affinity with these blood components, which contain various negatively charged structures [3, 4]. Previous studies have demonstrated that PAMAM aggregates red blood cells (RBCs), and results in hemolysis [6, 7], moreover, polycationicity triggers complement system activation [8], platelet aggregation [9] and interferes coagulating factors [10]. To avoid these adverse effects, it is imperative that low cytotoxic PAMAM dendrimers with satisfactory biocompatibility should be designed. The most common strategy to improve biocompatibility is surface modification, which allows for increasing solubility, preventing irrelevant cellular uptake and facilitating specific targeting.

Previously, we designed and synthesized a novel triblock dendrimer: polyamidoamine-polyethylene glycol-cyclic RGD (PAMAM-PEG-cRGD), in which PAMAM offered polycationicity for siRNA binding, polyethylene glycol (PEG) attenuated cytotoxicity, and cyclic RGD (cRGD) promoted tumor targeting [11]. The nanocarrier exerted satisfactory transfection efficiency in anaplastic thyroid carcinoma cells, with a ligand-receptor recognition between RGD and integrin on tumor cells [11]. Prior to advanced medical application, in vitro hemocompatibility evaluation of surface modified PAMAM nanocarrier must be characterized, in line with the International Organization for Standardization (ISO) 10993-4 requirements [12]. In this study, attempts have been made to investigate the compatibility of “PAMAM-PEG-cRGD” in blood components, by assessing in vitro hemocompatibility, with special emphasis on hemolysis, platelet interaction, complement system activation, and blood coagulation.

2 Materials and methods

2.1 Materials

PAMAM dendrimer (ethylenediamine core, Generation 4.0) was purchased from Sigma-Aldrich (St. Louis, MO, USA), while PAMAM-PEG-cRGD was synthesized as described in Fig. 1 based on our previous study [11]. Human C3a ELISA kit for quantitative analysis of Human C3a-des-Arg, adenosine diphosphate (ADP) and human collagen were obtained from Qiyun BioTech Inc. (Guangzhou, Guangdong, China). Dade innovin prothrombin time (PT) kit and Dade activated partial thromboplastin time (APTT) kit were purchased from Dade Behring Inc. (Newark, Delaware, USA). Zymosan and kaolin reagent were purchased from Ximei BioTech Inc. (Shanghai, China), while human β-Thromboglobulin (β-TG) enzyme linked immunosorbent assay (ELISA) kit and human platelet factor-4 (PF-4) ELISA kit were from Yanji BioTech Company (Beijing, China). All other chemicals and reagents employed in this study were of analytical grade.

Fig. 1

Synthesis scheme of PAMAM-PEG-cRGD. Notes: (A) Production of PAMAM-NHAc (acetylation of PAMAM), through the reaction of PAMAM (Generation 4.0) and tritriethylamine for 24 hours, in the presence of methanol and acetic anhydride, at room temperature; (B) reaction of PAMAM-NHAc and PEG for 24 hours with the addition of EDC and DMSO, yielding PAMAM-PEG-COOH; (C) reaction of PAMAM-PEG-COOH and SPDP, yielding PAMAM-PEG-SPDP; (D) Addition of cRGD to PAMAM-PEG-SPDP, gives rise to final product: PAMAM-PEG-cRGD. Abbreviations: PAMAM, polyamidoamine; PEG, polyethylene glycol; cRGD, cyclic arginine-glycine-aspartate peptide; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD; EDTA, ethylenediaminetetraacetic acid; EDC, 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride; SPDP, N-succinimidyl 3-(2-pyridyldithio)-propionate.

2.2 Characterization of nanoparticle

Successful synthesis of PAMAM-PEG-cRGD was confirmed by proton nuclear magnetic resonance (¹H-NMR), for which an Avance 400 MHz NMR Spectrometer (Bruker BioSpin International AG., Aegeristrasse, Zug, Switzerland) was used. Chemical shift was expressed as parts per million (ppm), with D₂O (4.8 ppm) solvent peak as the reference. The morphology and size of PAMAM-PEG-cRGD were also observed through transmission electron microscopy (TEM).

2.3 Preparation of red blood cells (RBCs) and platelet-rich plasma (PRP)

We collected human blood from 10 adult healthy donors, in line with the guidelines of the Helsinki Declaration for human research. The current study was also in accordance with the ethical guidelines of Clinical Hemorheology and Microcirculation [13]. This study was approved by the ethic committee of Sun Yat-sen Memorial Hospital (approval no. SMH-180359). Written informed consent of study was obtained from each donor prior to blood sample collection. Individuals with any history that suggested hematologic intervention, including recent aspirin or antibiotic administration were all excluded for blood collection. In order to control variations between individuals, only blood samples from the same healthy donor were used for each part of the investigation.

Blood was collected in 5.0-mL tubes containing 0.1 mol/L sodium citrate. RBC was extracted by centrifugation at 5000 revolution per minute (RPM) at 4°C for 5 minutes (min), and was then washed thrice with normal saline (NS). RBC suspension in saline solution was finally obtained. While PRP was isolated from fresh citrated whole blood by centrifugation (800RPM, 15 min, 37°C), after which the upper and middle layer were collected. The upper and middle layer were mixed and centrifuged (800RPM, 10 min, 37°C). With the supernatant discarded, the precipitated platelet was then re-suspended in the remaining plasma, and PRP was obtained.

2.4 RBC hemolytic assay

RBC suspension was treated with PAMAM (Generation 4.0) and PAMAM-PEG-cRGD solution at various concentrations (0 nM, 100 nM, 500 nM, 1 μM, 5 μM and 10 μM final) and incubated for 3 hours at room temperature. RBC suspension treated with NS under the same condition was considered as the negative control for reference, while Triton X-100 (10%, v/v) was applied to achieve a full hemolysis (positive control). After incubation, samples were centrifuged (5000RPM, 10 min, 4°C), and the resulting supernatants were recorded for the absorbance at 540 nm (OD₅₄₀). Meanwhile, samples were examined under optical microscope. The following equation was used to assess the degree of hemolysis: $Hemolysis (%) = \frac{{OD}_{sample} - {OD}_{0}}{{OD}_{100} - {OD}_{0}} \times 100 %,$ (1) in which OD_sample, OD₀, OD₁₀₀ were the values of OD₅₄₀ of examined samples, NS, and Triton X-100 (10%, v/v), respectively. Independent experiments were carried out thrice to confirm result reproducibility.

2.5 Platelet function evaluation

Mixed solution of dendrimer (0 nM 10 μM final) and whole blood was made in 1.5 mL Eppendorf tubes (v/v = 1:9). With gentle shaking, each sample was centrifuged (3000RPM, 15 min, 37°C) after a three-hour-incubation. After supernatant isolation, β-TG and PF4 were measured through ELISA according to the manufactures‘ instructions, in order to evaluate the effect of platelet activation. Absorbance at 450 nm was recorded. Absorbance of normal plasma was used as the negative reference, while ADP was applied for positive reference.

500 μL PRP was added to an Eppendorf tube (1.5 mL), with platelet concentration adjusted to 3×10⁶/μL, then solution of nanoparticle (0 nM∼10 μM final) was added for a 3-hour-incubation at room temperature, with NS used as negative control. Platelet counting was done with Coulter Multisizer II. Likewise, independent experiments were performed thrice.

2.6 Measurements of coagulation

Fresh whole blood and nanoparticle solutions at different concentrations were mixed and incubated as mentioned above. Samples were centrifuged (3000RPM, 15 min, 37°C) to get the supernatants. PT and APTT were measured based on the manufactures‘ instructions, using a Behring Coagulation Timer Analyzer (Dade Behring, Deerfield, IL, USA). NS solution was used as negative control and Kaolin reagent as positive control. All these experiments were repeated three times.

2.7 Complement activation test

Fresh whole blood and dendrimer solutions in different concentrations were mixed and incubated as mentioned above. 1 mM ethylenediaminetetraacetic acid (EDTA) was added to abolish complement activation. Samples were then centrifuged (2000RPM, 5 min, 37°C). Complement activation was tested via human C3a ELISA kit in line with its instruction. Absorbance at 450 nm was recorded, and was converted to concentration displayed as ng/mL. Zymosan was used as positive reference. Three independent experiments were carried out to confirm the sustainability of these results.

2.8 Statistics

Continuous variables are expressed as mean ±standard deviation (SD). Before the comparison of between-group differences, test of normality and Hartley test for homogeneity of variance were applied. One-way analysis of variance (one-way ANOVA) and Turkey–Kramer’s test were used to analyze variables between groups. P-values lower than 0.05 were deemed as statistical significance. Statistical analysis was performed by SPSS version.10.0 for Windows (SPSS Inc., Chicago, IL, USA).

3 Results

3.1 Characterization of nanoparticle

Successful preparation of PAMAM-PEG-cRGD was confirmed by ¹H-NMR. As shown in Fig. 2A, the signals at 2.12, 2.24 and 2.38 ppm reflected the presence of PAMAM. A single peak (δ= 3.62 ppm) was identified according to PEG, while the aromatic ring of cRGD contributed to the broad proton peaks at 6.61 ppm and 6.92 ppm. Figure 2B shows the ¹H-NMR reference of PAMAM (Generation 4.0). The morphology of PAMAM-PEG-cRGD was observed by TEM. As shown in Fig. 2C, the size of the nanoparticle ranged from 98 to 132 nm, with well-defined spherical shape.

Fig. 2

Characterization of PAMAM-PEG-cRGD. Notes: Panel A, ¹H-NMR analysis of PAMAM-PEG-cRGD; Panel B, ¹H-NMR reference of PAMAM (Generation 4.0); Panel C, characterization of nanoparticle morphology. Abbreviations: PAMAM, polyamidoamine; PEG, polyethylene glycol; cRGD, cyclic RGD; ¹H-NMR, proton nuclear magnetic resonance; ppm, parts per million.

3.2 Hemolysis

The RBC hemolysis assay is broadly adopted to investigate hemocompatibility of a nanoparticle. To examine possible RBC hemolysis, RBCs were treated with dendrimers in various concentrations. The absorbance at 540 nm (OD₅40) mirrors the concentration of free hemoglobin, which reveals the disruption of RBC membrane and hemolysis. As illustrated in Fig. 3, when erythrocytes were incubated with PAMAM (Generation 4.0) of 1 μM for three hours, the hemolysis rate was 3.53 ±1.57%, indicating slight degree of hemolysis. When RBCs were incubated with PAMAM (Generation 4.0) at 5 μM and 10 μM for 3 hours, the rate of hemolysis increased to 9.88 ±4.82% and 13.34 ±4.91%, respectively, revealing the presence of hemolysis. Clearly, PAMAM (Generation 4.0) induced hemolysis in a dose-dependent way. When erythrocytes were administered with PAMAM-PEG-cRGD, the hemolysis rate was also on the increase with the rising concentration, and 3-hour-incubations of PAMAM-PEG-cRGD at 5 μM and 10 μM leaded to hemolysis rates of 4.21 ±3.41% and 4.34 ±2.72%, respectively. These results suggested that hemolytic percentage of PAMAM (Generation 4.0) was higher than that of PAMAM-PEG-cRGD at the same dose. The hemolysis rate of PAMAM-PEG-cRGD was lower than 5% in concentrations below 10 μM, revealing that the RBC compatibility of PAMAM-PEG-cRGD was acceptable according to the ISO 10993-4 criteria [12].

Fig. 3

Dendrimer-induced hemolysis assay. Notes: Normal saline was used as negative control, while 4% Triton X-100 was considered as 100% release of hemoglobin. All data were illustrated as “mean ±standard deviation ”. ^*p < 0.05,statistically different from PAMAM at the same concentration, ^**p < 0.01 statistically different from PAMAM at the same concentration. Independent experiments were performed three times. Abbreviations: PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD.

Next, we evaluated the impact of dendrimer on RBC morphology through optical microscopy. Representative images of RBC morphology observed with the treatment of nanoparticles are presented in Fig. 4. When erythrocytes were incubated in the presence of PAMAM (Generation 4.0) at a dose of 5 μM or higher, spherical and echinocytic-shape with irregular contour were observed, meanwhile, RBC aggregation was apparent (Fig. 3A and B). When RBCs were exposed to PAMAM-PEG-cRGD at a concentration above 5 μM (Fig. 3C and D), the majority of erythrocytes maintained disk shaped regularly, although limited numbers of RBCs showed mild degree of irregular transformation in contact with PAMAM-PEG-cRGD at 10 μM (Fig. 3D).

Fig. 4

Morphology of RBCs treated with dendrimers at different doses. Notes: Panel A, RBCs exposed to PAMAM (Generation 4.0) at 5 μM; Panel B, RBCs exposed to PAMAM (Generation 4.0) at 10 μM; Panel C, RBCs exposed to PAMAM-PEG-cRGD at 5 μM; Panel D, RBCs exposed to PAMAM-PEG-cRGD at 10 μM. Magnification: ×100. Abbreviations: PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD; RBC, red blood cell.

3.3 Platelet activation

Effects of platelet activation were assessed by plasma β-TG and PF4 measurements. Absorbance of normal plasma was set as 100%. As shown in Fig. 5A and B, both PAMAM (Generation 4.0) and PAMAM-PEG-cRGD exhibited platelet activation properties, at concentrations higher than 5 μM, and β-TG reached above 150% of control level. The β-TG releasing percentages of PAMAM and PAMAM-PEG-cRGD at 10 μM were 227.67 ±27.74%, and 150.67 ±11.85%, respectively, and between-group difference was significant (p < 0.05). On the other hand, PAMAM induced PF4 releasing in a dose-dependent manner, and the PF4 releasing percentage of PAMAM at 10 μM was 240.00 ±45.71%, while the releasing percentages of PAMAM-PEG-cRGD were significantly lower than those of PAMAM from 1 μM to 10 μM, with the presence of significant between-group differences. The highest PF4 releasing percentage of PAMAM-PEG-cRGD at 10 μM was 164.33 ±24.66%.

Fig. 5

Platelet compatibility assays. Notes: Panel A, β-TG releasing percentage of fresh whole blood treated with dendrimers at various doses; Panel B, PF4 releasing percentage of fresh whole blood treated with dendrimers at various doses; Panel C, Platelet counting percentage in platelet rich plasma treated with dendrimers at various doses. Bars with *a indicate statistically significant differences at the same doses compared with saline control (p < 0.05). Bar with *b indicate statistically significant differences at the same doses compared with PAMAM (Generation 4.0) or PAMAM-PEG-cRGD (p < 0.05). Independent experiments were performed three times. Abbreviations: β-TG, β-Thromboglobulin; PF4, human platelet factor-4; PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD; ADP, adenosine diphosphate; NP, nanoparticle.

As seen in Fig. 5C, PAMAM (Generation 4.0) resulted in platelet reduction in a dose-dependent fashion, and the platelet counting dropped to nearly 20% of control level at a relatively low PAMAM incubating dose (1 μM). PAMAM-PEG-cRGD also caused platelet reduction, but the reduction percentages were significantly lower compared with unmodified PAMAM at 5 μM and 10 μM (p < 0.01). PAMAM at 10 μM induced 44% reduction of platelet number, while the reduction percentage was 28.3% in PAMAM-PEG-cRGD at the same dose.

3.4 Coagulation

Two pathways are routinely involved in human coagulation process: the intrinsic pathway and the extrinsic pathway, for which evaluations broadly acknowledged are APTT and PT, respectively. Using NS and kaolin as references, the durations of clotting formation were recorded when fresh blood were treated with dendrimers at different concentrations. As shown in Table 1, PAMAM-PEG-cRGD and PAMAM (Generation 4.0) did not exhibit interferences to PT for most study-settings, with one exception: PAMAM at 10 μM, under which PT prolonged significantly compared with the control level (p = 0.023). This difference, however, was absent for APTT, on which neither PAMAM (Generation 4.0) nor PAMAM-PEG-cRGD exhibited evident influences (Table 2).

Table 1
Assessment of extrinsic pathway interference of dendrimers. Prothrombin time (in sections) of blood sample treated with dendrimer at different concentration was recorded

Concentrations (μM)\Treatment NS control PAMAM (Generation 4.0) PAMAM-PEG-cRGD Kaolin control

0.1 μM 11.13 ±1.70 11.13 ±1.70 9.70 ±0.53 21.30 ±0.63

0.5 μM 11.27 ±1.16 11.27 ±2.85 10.03 ±0.21 20.93 ±0.21

1 μM 10.83 ±1.42 10.83 ±2.57 13.50 ±3.03 21.20 ±0.30

5 μM 9.83 ±0.46 9.93 ±1.10 10.80 ±2.29 20.70 ±0.52

10 μM 10.27 ±1.00 16.70 ±2.88^* 13.03 ±3.89 20.40 ±0.44

Concentrations (μM)\Treatment	NS control	PAMAM (Generation 4.0)	PAMAM-PEG-cRGD	Kaolin control
0.1 μM	11.13 ±1.70	11.13 ±1.70	9.70 ±0.53	21.30 ±0.63
0.5 μM	11.27 ±1.16	11.27 ±2.85	10.03 ±0.21	20.93 ±0.21
1 μM	10.83 ±1.42	10.83 ±2.57	13.50 ±3.03	21.20 ±0.30
5 μM	9.83 ±0.46	9.93 ±1.10	10.80 ±2.29	20.70 ±0.52
10 μM	10.27 ±1.00	16.70 ±2.88^*	13.03 ±3.89	20.40 ±0.44

Notes: data with * indicate statistically significant differences at the same doses compared with saline control (p < 0.05). Independent experiments were performed three times. Abbreviations: PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD.

Table 2

Assessment of intrinsic pathway interference of dendrimers

Concentrations (μM)\Treatment	NS control	PAMAM (Generation 4.0)	PAMAM-PEG-cRGD	Kaolin control
0.1 μM	34.23 ±0.93	33.33 ±1.53	34.83 ±1.76	15.90 ±3.56
0.5 μM	32.97 ±1.90	35.50 ±2.29	35.33 ±1.16	13.53 ±3.14
1 μM	34.33 ±3.06	37.13 ±5.24	34.33 ±5.77	13.13 ±1.55
5 μM	34.90 ±0.96	35.50 ±6.07	33.50 ±5.68	14.37 ±4.19
10 μM	35.20 ±2.38	32.43 ±6.05	36.67 ±6.03	13.60 ±2.97

Activated partial thromboplastin time (in sections) of blood sample treated with dendrimer at different concentration was recorded. Independent experiments were performed three times. Abbreviations: PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD.

3.5 Complement activation

C3a, a key product of anaphylaxis, is cleaved from C3 protein by C3 convertase, and is a biomarker reflecting complement activation. We used zymosan, which was a polysaccharide of Saccharomyces cerevisiae and a well-known activator of complement system, for positive reference throughout the complement activation assays. Table 3 displays the degree of complement activation (data presented in % of NS control) related to dendrimers. Investigating of C3a level from samples treated with dendrimers at various concentrations reached the conclusion that significant complement activation was absent within the dose rage of experimental settings.

Table 3
Assessment of complement activation effects of dendrimers

Concentrations (μM)\Treatment NS control PAMAM (Generation 4.0) PAMAM-PEG-cRGD Kaolin control

0.1 μM 100.33 ±0.58 99.00 ±6.56 102.33 ±9.02 707.00 ±81.30

0.5 μM 97.67 ±1.53 100.00 ±18.52 100.67 ±17.67 683.67 ±139.11

1 μM 99.33 ±3.06 98.67 ±25.58 118.00 ±15.40 767.67 ±140.05

5 μM 98.33 ±1.53 91.67 ±30.01 93.67 ±14.19 728.00 ±52.26

10 μM 101.00 ±1.00 76.67 ±14.05 104.00 ±33.18 655.67 ±87.01

Concentrations (μM)\Treatment	NS control	PAMAM (Generation 4.0)	PAMAM-PEG-cRGD	Kaolin control
0.1 μM	100.33 ±0.58	99.00 ±6.56	102.33 ±9.02	707.00 ±81.30
0.5 μM	97.67 ±1.53	100.00 ±18.52	100.67 ±17.67	683.67 ±139.11
1 μM	99.33 ±3.06	98.67 ±25.58	118.00 ±15.40	767.67 ±140.05
5 μM	98.33 ±1.53	91.67 ±30.01	93.67 ±14.19	728.00 ±52.26
10 μM	101.00 ±1.00	76.67 ±14.05	104.00 ±33.18	655.67 ±87.01

C3a level in saline control was normalized as 100%, and data were displayed as percentage of control level. Independent experiments were performed three times. Abbreviations: PAMAM, polyamidoamine; PAMAM-PEG-cRGD, polyamidoamine-polyethylene glycol-cyclic RGD.

4 Discussion

Safety and biocompatibility are the permanent issue of significance for nanoparticle design and development. Evaluations with respect to hemocompatibility must be strictly addressed before the application of a new nanomaterial, as long as it potentially interact with blood. The interaction between PAMAM dendrimer and blood components is on a intricate fashion. Better understanding of nanoparticle hemocompatibility also clearly adds valuable evidence to nanomedicine and benefits research translation into clinical application. Complying with the ISO 10993-4, we performed hemolysis assay, platelet compatibility assay and complement activation assays to examine the blood compatibility of a new dendritic nano-scale gene vector: PAMAM-PEG-cRGD, in which we previously conjugated PEG and cRGD to PAMAM (Generation 4.0) spherical molecule [11]. Our findings show that surface modification of PEG and cRGD to dendrimer significantly improves RBC biocompatibility as compared with unmodified PAMAM, and similarly, these modifications alleviate both platelet activation and deleterious effects induced by cationic PAMAM. Moreover, high concentration of PAMAM interferes coagulation system, and prolongs hemostasis.

The polycationic milieu of PAMAM renders superiorities as a drug or gene vector, allowing effective affinity and multivalency to multiple-agent attachment. The spherical structure also enables “proton sponge effect”, which is modulated by pH and electrostatic interaction [14]. It is well acknowledged that PAMAM toxicity property is rooted in polycationicity, for which surface modification for the sake of cationic charges shielding is required to improve biocompatibility, including PEGylation [15 –17], cyclodextrin dendrimers [18, 19], glycoconjugates [20], acetylation [21], thiolation [22], hydrophobicity [23] and so forth. Similar to Neffe’s finding [24], in which hyperbranched oligo- and polyglycerols shields rough membranes to mediate hemocompatibility [24], PEGylation also shields polycationicity and attenuates hemocompatibility. In addition to these shielding approaches, surface chemistry for the introduction of targeting ligands improves specificity and indirectly mitigates cytotoxic property, since dendrimer concentration required for effective cellular transfection could be reduced with the increase of targeting delivery [25], for which available targeting ligands include cRGD [11 , 27], hyaluronic acid [28], folic acid [29], epidermal growth factor receptor [30], and so on. In parallel with our previous results [11], the introduction of PEG and cRGD also improves PAMAM biocompatibility in blood components.

Erythrocyte, the oxygen carrier in human circulation system, is more vulnerable to foreign macromolecules as the bilayer, nucleus-absent architecture insuffices protective responses to deleterious environment. The PAMAM hemoreactivity is similarly well-explained by the theory regarding destabilization of the negatively-charged RBC membrane integrity, yielding membrane disruption and hemolysis [6]. Consistent with Kim‘s report [31], the deleterious effect on RBC is recognized to be concentration, generation, and charge dependent. In the present study, PAMAM-PEG-cRGD did not manifest evident hemolytic property under 10 μM, although the absorbance increased in a dose-dependent manner. We speculate that hemolysis might be observed if RBC is exposed to a much higher concentration, but effective transfection could be achieved in a concentration lower than 1 μM as we reported previously [11].

Compared with erythrocytes, the toxicological mechanism of platelets is even more complicated. Platelets are key to hemostasis, which is composed of multiple steps, including adhesion, activation, aggregation, and coagulation cascade involvement. Our results revealed that both PAMAM-PEG-cRGD and PAMAM (Generation 4.0) exerted platelet activation activity, as confirmed by the elevation of β-TG and PF4. In agreement with other published studies [22, 32], the PEG modification also attenuates PAMAM-mediated activation effect on platelets. We observed PAMAM-induced platelet reduction occurred at a relatively low dose, and this finding clearly supports the notion that platelet is sensitive and fragile to foreign particles reacting with cell surface. Some glycoproteins, like integrin receptor GPIIb/IIIa contribute to platelet activation and aggregation [33], while it is noteworthy that, some of these glycoproteins are also over-expressed in proliferating cancer cells, allowing for targets of dendrimer surface engineering. It has been proved that there is an affinity between cRGD and integrin receptor [11 , 27], which theoretically enhances interactions between cRGD-modified nanoparticle and the platelet as well. We observed that PAMAM-PEG-cRGD at 10 μM caused an approximately 30% reduction in platelet counting, yet based on our findings we could not conclude actual mechanism of platelet biocompatibility at this stage. Either cationic toxicity or interference of cRGD binding is possible explanation for platelet reduction, and fortunately, the same as hemolytic behavior, toxicity of PAMAM-PEG-cRGD to platelet is negligible at the dose intended for transfection. Antonova et al. also demonstrated that modification of RGD peptides enhanced platelet aggregation [34]. It is imperative that awareness of the affinity with regard to RGD derivatives and platelet should be highlighted when surface chemistry of RGD derivatives is undertaken. Platelet aggregation assays are currently on the way and we are now thoroughly exploring underlying mechanism of PAMAM-induced platelet toxicity.

Biocompatibility of coagulation and complement system should not be ignored when addressing toxicity profile of a nanoparticle. Either coagulation cascade or complement system activation involves an army of proteinous participants, it is thereby anticipated that coagulation or complement system might be affected if PAMAM polycationicity absorbs negatively-charged key proteins in coagulation or complement system. One study by Shcharbin D. et al enunciated that unmodified PAMAM was non-toxic under Generation 5.0 in vivo, however, disseminated intravascular coagulation was triggered in Generation 7.0 [35]. We discovered that neither PAMAM (Generation 4.0) nor PAMAM-PEG-cRGD affected coagulation and complement system, except for PAMAM at 10 μM, under which we observed PT prolonged significantly. This prolonged effect was probably because the polycationicity consumed negative-charged coagulating factors, leading to coagulating factor deficiency. Our finding, however, has some variations as compared with a study by Liu. et al, in which PEGylated PAMAM presented prolonged PT and APTT [22]. Such variations are probably associated with the difference in experimental setting of incubation. Our coagulation and complement activation assays were conducted in whole fresh blood, whereas they examined these results from human plasma [22]. The complex toxicological machinery includes multiple factors, like hemolysis, coagulation factor binding, and platelet toxicity, which are all possibly affect coagulation, hence, it is more appropriate to perform coagulation assays in whole blood to measure these factors entirely.

Our study is distinctive, and provides valuable information to nano-researchers in that, to the best of our knowledge, this is the first study to fully address hemocompatibility in dendrimer with PEG and RGD derivatives modified. Our study, however, has several limitations which merit further considerations. To begin with, all our findings in this study are in vitro currently. In vivo PAMAM cytotoxicity is deemed to be a more challenging issue, which correlates not only with all in vitro aspects mentioned above, but also dynamically with nanoparticle biodistribution, pharmacokinetics, endocrine regulation, and so on. We are now validating our results in mice. In addition, available experimental methods for coagulation and complement assays limit investigators to explore underlying toxicological mechanism more sufficiently. Intricate as these cascade machineries which numerous proteins are included, it is clearly not sufficient to figure out full-scale toxicological pattern with only C3a, PT and APTT examined. Certainly, further investigations are warranted to refresh our findings.

5 Conclusion

Surface modification of PEG and cRGD to PAMAM (Generation 4.0) improves hemocompatibility. Introduction of PEG and cRGD significantly mitigates hemolytic and RBC aggregation effects as compared with unmodified PAMAM. Similarly, these modifications alleviate platelet activation and platelet reduction mediated by PAMAM polycationicity. Finally, high concentration of unmodified PAMAM induces hemolysis (10 μM), leads to platelet reduction (≥5 μM), interferes coagulation system (10 μM), and prolongs hemostasis.

Disclosure

All authors report no conflicts of interest in this work.

References

Araujo

, Santos

SDS

, Igne Ferreira

, Giarolla

. New advances in general biomedical applications of PAMAM dendrimers. Molecules (Basel, Switzerland). 2018;23(11).

Abedi-Gaballu

, Dehghan

, Ghaffari

, Yekta

, Abbaspour-Ravasjani

, Baradaran

, et al. PAMAM dendrimers as efficient drug and gene delivery nanosystems for cancer therapy. Applied Materials Today. 2018;12:177–90.

Otto

, de Villiers

. Poly(amidoamine) dendrimers as a pharmaceutical excipient. Are we there yet? Journal of Pharmaceutical Sciences. 2018;107(1):75–83.

Fox

, Richardson

, Briscoe

. PAMAM dendrimer - cell membrane interactions. Advances in Colloid and Interface Science. 2018;257:1–18.

Cui

, Liang

, Wang

, Zhu

, Kang

, Sun

, et al. Enhanced biocompatibility of PAMAM dendrimers benefiting from tuning their surface charges. Materials Science & Engineering C, Materials for Biological Applications. 2018;93:332–40.

Wang

, Xiong

, Zhu

, Xu

, Yang

. Protective effect of PEGylation against poly(amidoamine) dendrimer-induced hemolysis of human red blood cells. Journal of Biomedical Materials Research Part B, Applied Biomaterials. 2010;93(1):59–64.

Han

, Chen

, Wang

, Chen

, Wang

. Blood compatibility of polyamidoamine dendrimers and erythrocyte protection. Journal of Biomedical Nanotechnology. 2010;6(1):82–92.

Lin

, Hua

, Zhang

, Li

, Xue

, Yin

, et al. Effect of poly(amidoamine) dendrimers on the structure and activity of immune molecules. Biochimica et Biophysica Acta. 2015;1850(2):419–25.

Watala

, Karolczak

, Kassassir

, Siewiera

, Maczynska

, Pieniazek

, et al. How do the full-generation poly(amido)amine (PAMAM) dendrimers activate blood platelets? Platelet membrane zeta potential and other membrane-associated phenomena. International Journal of Pharmaceutics. 2016;500(1-2):379–89.

10.

Jones

, Campbell

, Brooks

, Assemi

, Tadjiki

, Thiagarajan

, et al. Cationic PAMAM dendrimers aggressively initiate blood clot formation. ACS Nano. 2012;6(11):9900–10.

11.

, Hu

, Yin

, Zhang

, Huang

, Wang

, et al. A novel dendritic nanocarrier of polyamidoamine-polyethylene glycol-cyclic RGD for “smart” small interfering RNA delivery and in vitro antitumor effects by human ether-a-go-go-related gene silencing in anaplastic thyroid carcinoma cells. International Journal of Nanomedicine. 2013;8:1293–306.

12.

Seyfert

, Biehl

, Schenk

. In vitro hemocompatibility testing of biomaterials according to the ISO 10993-4. Biomolecular Engineering. 2002;19(2-6):91–6.

13.

Ethical guidelines for publication in Clinical Hemorheology and Microcirculation: Update 2016. Clinical Hemorheology and Microcirculation. 2016;63(1):1–2.

14.

Ouyang

, Zhang

, Parekh

, Smith

. The effect of pH on PAMAM dendrimer-siRNA complexation: Endosomal considerations as determined by molecular dynamics simulation. Biophysical Chemistry. 2011;158(2-3):126–33.

15.

Barraza

, Jimenez

, Alderete

. Association of methotrexate with native and PEGylated PAMAM-G4 dendrimers: Effect of the PEGylation degree on the drug-loading capacity and preferential binding sites. The Journal of Physical Chemistry B. 2017;121(1):4–12.

16.

Luong

, Kesharwani

, Deshmukh

, Mohd Amin

MCI

, Gupta

, Greish

, et al. PEGylated PAMAM dendrimers: Enhancing efficacy and mitigating toxicity for effective anticancer drug and gene delivery. Acta Biomaterialia. 2016;43:14–29.

17.

Diaz

, Guzman

, Jimenez

, Alderete

. Partially PEGylated PAMAM dendrimers as solubility enhancers of Silybin. Pharmaceutical Development and Technology. 2018;23(7):689–96.

18.

Arima

, Motoyama

, Higashi

. Potential therapeutic application of dendrimer/cyclodextrin conjugates with targeting ligands as advanced carriers for gene and oligonucleotide drugs. Therapeutic Delivery. 2017;8(4):215–32.

19.

Saraswathy

, Knight

, Pilla

, Ashton

, Gong

. Multifunctional drug nanocarriers formed by cRGD-conjugated betaCD-PAMAM-PEG for targeted cancer therapy. Colloids and Surfaces B, Biointerfaces. 2015;126:590–7.

20.

Cai

, Xun

, Tang

, Tian

, Yang

, Ding

, et al. Glycoconjugate probes containing a core-fucosylated N-glycan trisaccharide for fucose lectin identification and purification. Carbohydrate Research. 2017;449:143–52.

21.

Yan

, Gu

, Lv

, Shi

, Jing

. Improved intestinal absorption of water-soluble drugs by acetylation of G2 PAMAM dendrimer nanocomplexes in rat. Drug Delivery and Translational Research. 2017;7(3):408–15.

22.

Liu

, Pang

, Toh

, Chiu

. Dual-functionalized poly(amidoamine) dendrimers with poly(ethylene glycol) conjugation and thiolation improved blood compatibility. The Journal of Pharmacy and Pharmacology. 2015;67(11):1492–502.

23.

Wang

, Dong

, Sun

. Hydrophobic modification of carboxyl-terminated polyamidoamine dendrimer surface creates a potent inhibitor of amyloid-beta fibrillation. Langmuir: The ACS Journal of Surfaces and Colloids. 2018;34(47):14419–27.

24.

Neffe

, Ruesten-Lange

, Braune

, Lutzow

, Roch

, Richau

, et al. Supplementary information for: Multivalent grafting of hyperbranched oligo- and polyglycerols shielding rough membranes to mediate hemocompatibility. Jornal of Maral Hmry B. 2014;2(23):Accepted Manuscript.

25.

Sharma

, Gothwal

, Kesharwani

, Alsaab

, Iyer

, Gupta

. Dendrimer nanoarchitectures for cancer diagnosis and anticancer drug delivery. Drug Discovery Today. 2017;22(2):314–26.

26.

, Yu

, Zhang

, Mu

, Chu

, Ni

, et al. Increased active tumor targeting by an alphavbeta3-targeting and cell-penetrating bifunctional peptide-mediated dendrimer-based conjugate. Pharmaceutical Research. 2017;34(1):121–35.

27.

, Song

, Huang

, Chen

, Deng

. Novel, nano-sized, liposome-encapsulated polyamidoamine dendrimer derivatives facilitate tumour targeting by overcoming the polyethylene glycol dilemma and integrin saturation obstacle. Journal of Drug Targeting. 2017;25(8):734–46.

28.

, Sha

, Cheng

, Yao

, Li

, Qi

. Construction of hyaluronic tetrasaccharide clusters modified polyamidoamine siRNA delivery system. Nanomaterials (Basel, Switzerland). 2018;8(6).

29.

Chanphai

, Tajmir-Riahi

. Characterization of folic acid-PAMAM conjugates: Drug loading efficacy and dendrimer morphology. Journal of Biomolecular Structure & Dynamics. 2018;36(7):1918–24.

30.

Yoon

, Kasala

, Li

, Hong

, Lee

, Jung

, et al. Antitumor effect and safety profile of systemically delivered oncolytic adenovirus complexed with EGFR-targeted PAMAM-based dendrimer in orthotopic lung tumor model. Journal of Controlled Release: Official Journal of the Controlled Release Society. 2016;231:2–16.

31.

Kim

, Nafiujjaman

, Nurunnabi

, Lim

, Lee

, Park

. Effects of polymer-coated boron nitrides with increased hemorheological compatibility on human erythrocytes and blood coagulation. Clinical Hemorheology and Microcirculation. 2018;70(3):241–56.

32.

Watala

, Karolczak

, Kassassir

, Talar

, Przygodzki

, Maczynska

, et al. How do the full-generation poly(amido)amine (PAMAM) dendrimers activate blood platelets? Activation of circulating platelets and formation of “fibrinogen aggregates” in the presence of polycations. International Journal of Pharmaceutics. 2016;503(1-2):247–61.

33.

Coller

. Blockade of platelet GPIIb/IIIa receptors as an antithrombotic strategy. Circulation. 1995;92(9):2373–80.

34.

Antonova

, Silnikov

, Sevostyanova

, Yuzhalin

, Koroleva

, Velikanova

, et al. Biocompatibility of small-diameter vascular grafts in different modes of RGD modification. Polymers. 2019;11(1).

35.

Shcharbin

, Janaszewska

, Klajnert-Maculewicz

, Ziemba

, Dzmitruk

, Halets

, et al. How to study dendrimers and dendriplexes III. Biodistribution, pharmacokinetics and toxicity in vivo. Journal of Controlled Release: Official Journal of the Controlled Release Society. 2014;181:40–52.