Characterization of polymeric nanoparticles for intravenous delivery: Focus on stability

Accepted Manuscript Title: Characterization of polymeric nanoparticles for intravenous delivery: focus on stability Author: Claudia L. Oliveira Francisco Veiga Carla Varela Fernanda Roleira Elisi´ario Tavares Isabel Silveira Antonio J. Ribeiro PII: DOI: Reference:

S0927-7765(16)30765-2 http://dx.doi.org/doi:10.1016/j.colsurfb.2016.10.046 COLSUB 8227

To appear in:

Colloids and Surfaces B: Biointerfaces

Received date: Revised date: Accepted date:

30-7-2016 18-10-2016 24-10-2016

Please cite this article as: Claudia L.Oliveira, Francisco Veiga, Carla Varela, Fernanda Roleira, Elisi´ario Tavares, Isabel Silveira, Antonio J.Ribeiro, Characterization of polymeric nanoparticles for intravenous delivery: focus on stability, Colloids and Surfaces B: Biointerfaces http://dx.doi.org/10.1016/j.colsurfb.2016.10.046 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.

Characterization of polymeric nanoparticles for intravenous delivery: focus on stability Claudia L. Oliveiraa,b,c,d, Francisco Veigad,e, Carla Varelad, Fernanda Roleirad,e, Elisiário Tavaresd,e, Isabel Silveiraa,b,c, Antonio J. Ribeiroa,b,d a

Group Genetics of Cognitive Dysfunction, I3S - Instituto de Investigação e Inovação

em Saúde, Universidade do Porto, Rua Alfredo Allen, 208, 4200-135, Porto, Portugal b

c

IBMC - Instituto de Biologia Molecular e Celular, Universidade do Porto, Portugal

ICBAS, Universidade do Porto, Portugal

d

Faculty of Pharmacy, University of Coimbra, Azinhaga de Santa Comba, 3000-548

Coimbra, Portugal e

CNC - Center for Neurosciences and Cell Biology, University of Coimbra, Portugal

1

Highlights 

Polymeric NP stability within biological media is still poorly investigated.



In vitro stability of NP in different media is independent of NP surface.



Pronounced difference after incubation in mouse and human plasma.



NP aggregation related to medium total protein concentration.

2

Abstract The nano-bio interaction has been of increased focus in the past years but very limited results have been obtained for polymeric nanoparticles (NP). Not only is needed to broaden the results obtained with model NP towards other nano-materials used for clinical application but the colloidal stability of NP as a variable consequence of the formation of the protein corona has been significantly understated. The lack and heterogeneity of assays to study NP stability and represent the biological environment call for the standardization of assays to improve the representativeness and comparability of results. In this paper, uncoated and PAH-coated PLGA NP have been prepared and characterized in regard to their potential for intravenous administration. The comparative study of the stability of NP in three media used to represent the biological environment-bovine serum albumin (BSA) solution, mouse and human plasma - revealed that both formulations were unstable in human plasma as opposed to the results obtained for other media. This unexpected behavior in plasmas of different origins could be correlated with a significant variation of the amount of proteins adsorbed to NP and, ultimately, with an approximately 6-fold difference in total protein concentration between the plasma samples. These results suggest that inter-species variation could impact on the colloidal stability of NP and enhance the need to understand the correlation between biological media and identify protocol-related interferences which, altogether, may evidence a relevant factor compromising in vitroin vivo correlation and the translation of delivery systems aimed at intravenous administration.

3

Keywords Biological stability; Bovine serum albumin; Physicochemical characterization; Poly allylamine hydrochloride; Poly(lactic-co-glycolic) acid; Mouse and human plasma.

4

Graphical Abstract – Lack of correlation of NP stability in the foremost media used to represent the intravenous environment poses a concern for the translation of assays.

5

1. Introduction Nanoparticles (NP) are largely used as delivery systems and their use has become a well-established strategy to overcome several limitations of therapeutics. Elucidation of various extracellular and intracellular obstacles with which NP are confronted has driven the development of enhanced and “smart” NP designed to overcome them [1]. Still, often, lack of correlation between in vitro and in vivo behavior of NP compromises translation into clinical application. This is overall a significant problem on the pharmaceutical development process but it has particularly been gaining attention for intravenous administration due to further unravelling of the nano-bio interaction. This phenomenon refers to the interactions that are established between NP and the components of the biological environment with which they are put into contact. For the intravenous administration route to be feasible not only do NP have to withstand the physiological pH but also the interactions triggered in a highly complex medium, rich in proteins and salts, as the blood. More specifically, upon contact with biological fluids, adsorption of proteins at the surface of NP occurs, leading to the formation of a protein corona that becomes their primary surface for interaction [2]. This can introduce unknown changes to the physicochemical properties of the NP. Even though these may not be necessarily deleterious for the delivery process, negative consequences may arise namely aggregation, loss of active targeting [3] or recognition by the mononuclear phagocytic system (MPS) leading to accelerated clearance [4]. Considering a situation where NP are required to maintain their integrity and targeting capabilities throughout the delivery process, such as for gene delivery purposes, it is important to anticipate the response of NP after administration. 6

Detailed protein corona fingerprinting allows us to better understand the composition of this biomolecular layer and to clarify how formulation properties, such as size and surface chemistry, impact on the formation of the protein corona [5, 6]. However, these studies have mainly been performed using inorganic NP and, as material-specific interactions modulate the dynamics of the protein corona [7], it is fundamental to extend this knowledge to other materials employed for clinical application, such as polymeric NP. Excellent progress has been made in the field, where validated protocols have been published for assessing the hemolytic potential [8], thrombogenic properties [9] and complement activation induced by nanoparticles [10]. However, the colloidal stability of NP in a biological environment is often overlooked, even though NP aggregation can be responsible for lack of efficiency of a formulation or reproducibility of results [11]. Furthermore, great heterogeneity can be observed in the methodologies used for assessing NP stability [12-14]. The inherent difficulty in assessing NP in complex media and the differences on the nature of NP significantly compromise the comparability of results and, consequently, impairs the establishment of relationships. In addition, the composition and ratio of media used in vitro to represent the biological environment are highly heterogeneous in their extent, from biological samples such as plasma or sera of different origins to cell culture media or buffer solutions supplemented with a protein source, such as fetal bovine serum [15, 16]. More frequently, the characterization of the protein corona has been performed after incubation of samples in human plasma. However, mice are usually the first models used to test the efficiency of NP in vivo and the assessment of the protein corona of NP using human plasma could introduce unanticipated differences. This is especially relevant as attention has been paid to the correlation between the composition of the 7

protein corona obtained between biological samples from human and mouse origin [17]. Standardizing reliable and comparable assays is fundamental to enhance the comparability of results as well as to minimize lab-to-lab variation. Even though polymeric NP have been widely developed as delivery systems, little is known about the nano-bio interactions using formulations of this nature. Poly(D,L-lactic-co-glycolic acid) (PLGA) is negatively-charged, hydrophobic, widely used co-polymer due to its biodegradability and biocompatibility which has been extensively applied to the intravenous administration of NP [18]. Poly allylamine hydrochloride (PAH) is a cationic, hydrophilic polymer which can interact with oppositely charged polymers, such as PLGA, by electrostatic interactions [19]. This allows to vary surface hydrophobicity and to modulate the charge of NP, which are relevant properties for the interaction with proteins and cells [20]. Herein, we investigate the relationship between in situ protein corona formation and its impact on the stability of uncoated and PAH-coated PLGA NP in three media BSA solution and plasma of mouse and human origin, using size, zeta potential and quantification of total protein adsorption at the surface of NP as screening tests. BSA solution at physiological concentration was used as a model of the biological environment and also to provide information on how the competition between different proteins affect the stability of NP in comparison with their behavior with a single protein, which should only depend on the affinity of the protein towards the NP. Finally, we address the suitability of media used to represent the biological environment in regard to the correlation of results obtained in media of different complexity and origin.

8

2. Material and methods 2.1. Materials Poly(lactic-co-glycolic) acid 50:50 (Resomer RG 502 H, Evonik Industries, Germany); Extra-pure acetone (Scharlau, Barcelona, Spain); Poloxamer 188 (Kolliphor® P188, BASF SE, Germany);

Bovine serum albumin (A1933),

Poly

(allylamine hydrochloride) average Mw ~17,500, ninhydrin reagent 2% solution, Fluorescein isothiocyanate (FTIC), buffer

N-HydroxySuccinimide (NHS) and phosphate

saline pH 7.4 (P-3813) were purchased from Sigma-Aldrich, Germany;

Cellulose nitrate filters 0.8 μm and Vivaspin® 20 centrifugal filters with 100 kDa MWCO (Sartorius Stedim biotech GmbH, Göttingen, Germany); Lactic acid (VWR, USA); Pierce™ Coomassie Plus (Bradford) Assay Kit (Thermo Scientific, USA); All water was of MilliQ grade (Millipore, Germany). 2.2. Preparation of PLGA nanoparticles PLGA nanoparticles were prepared through a nanoprecipitation method. Briefly, 4 mL of an extra-pure acetone solution of PLGA (15 mg/mL) were added drop-wise into 12 mL of a 0.5% (w/v) aqueous solution of poloxamer 188, under magnetic stirring at 500 rpm. The organic solvent was removed by rotoevaporation using a BUCHI rotavapor R-210 (BUCHI, Switzerland), for 35 minutes under a pressure of 70 mBar and a water bath heated to 40 ºC. The recovered suspension was centrifuged for 30 minutes at 10000 rpm in a Minispin® centrifuge (Eppendorf, Germany) and the pellet washed twice with water. The pellet was re-dispersed in 9.6 mL of a 0.5% (w/v) poloxamer 188 solution using an ultrasound probe (Sonics Vibra-Cell, USA) at 100% amplitude for 5 minutes, under an ice bath. The suspension of nanoparticles was filtered

9

through a 0.8 μm filter. The concentration of nanoparticles was determined through gravimetric methods. Fluorescently-labeled nanoparticles were prepared by introducing FITC-labeled PLGA in a ratio of 5:95 to the PLGA solution and following the previous protocol. FITC-labeled PLGA was synthetized by a carbodiimide/NHS crosslinking chemistry [21]. 2.3. Coating of PLGA nanoparticles Surface coating was introduced by the addition of PAH to modulate the surface properties of PLGA nanoparticles. In all settings, 10 mL of PLGA nanoparticles 5 mg/mL were used and the coating process was performed under sonication using an ultrasound probe (Sonics Vibra-Cell, USA) at 50% amplitude, in an ice bath. The coating agent was added drop-wise to the suspension of PLGA nanoparticles at 1 rpm (approximately 0.1 mL/min flow rate) using a Watson Marlow 205 S peristaltic pump (Watson Marlow, UK). PAH coating was achieved by the addition of 2 mL of a 0.025% (w/v) aqueous solution at pH 5.5 to PLGA nanoparticles. Samples were purified from unreacted polymer by centrifugation using centrifugal filters with molecular weight cut off of 100 kDa. The samples were centrifuged at 1000 g for 30 minutes and washed twice with water. Final concentrated samples were diluted to 10 mL with water to maintain the original concentration. 2.4. Quantification of the coating adsorbed to PLGA nanoparticles The amount of PAH adsorbed to PLGA nanoparticles was quantified using the ninhydrin reagent according to a method previously described [22, 23]. Briefly, 0.5 mL of the ninhydrin reagent was added to 0.5 mL of sample in a test tube and heated in a boiling water bath for 30 minutes. Upon cooling, 15 mL of ethanol 50% (v/v) were 10

added and the samples were vortexed for 15 seconds. The absorbance of each sample was measured at 570 nm using a UV spectrophotometer UV-1800 (Shimadzu, Japan) and the concentration of polymer determined from a calibration curve of aqueous solutions of PAH at pH 5.5 prepared simultaneously. A 3-fold dilution of uncoated PLGA nanoparticles were used as negative control, to maintain a similar concentration to the unknown sample. 2.5. Characterization of nanoparticles Nanoparticles were characterized according to their size and polydispersity index through dynamic light scattering and their zeta potential assessed by means of laser doppler micro-electrophoresis

using

a

ZetaSizer Nano

ZS

(Malvern,

Worcestershire, UK. Each measurement was performed in triplicate using a polystyrene cuvette and a disposable folded capillary cell for size and zeta potential determination, respectively. Size measurements were performed at a backscattering angle (173°) and reported as the intensity-weighted average. Zeta potential was inferred using the Smoluchowski equation. The characteristics of both nanoparticles were assessed after dilution in water to a concentration range appropriate for analysis and also upon pH adjustment to 7.4 with NaOH 0.05 M. All measurements were performed in automatic mode at 25°C. The equipment was routinely calibrated using size and zeta potential standards. 2.6. Transmission electron microscopy The morphology of nanoparticles was assessed by transmission electron microscopy under a Tecnai G2 Spirit BioTWIN electron microscope (FEI, Oregon, USA) operated at 100kV. A 50-fold dilution of sample (5 μL) was deposited on a formvar/carbon coated-copper grid (300 mesh) for 30 seconds and excess solution 11

blotted off with filter paper. Negative staining was achieved with 2 μL of phosphotungstic acid 1% (w/v), pH 7.4 filtered with a 0.22 μm syringe filter, which was allowed to react with the sample for 5 minutes, after which excess was removed with filter paper. 2.7. In vitro stability of nanoparticles in a biological environment The stability of coated and uncoated nanoparticles was assessed in vitro in conditions based on those previously proposed by other groups to mimic the biological environment. 2.7.1. Stability in a BSA solution To assess the stability of samples in the presence of proteins and physiological pH, 100 µL of a BSA solution 0.4 mg/mL prepared in PBS 10 mM pH 7.4 [24, 25] was incubated in a shaking water bath (SS40-D Grant, UK) at 37 ºC and the temperature was equilibrated for approximately 10 minutes. Afterwards, 900 µL of 10-fold diluted sample were added to the previous solution and incubated for 1 hour at 37ºC under agitation of 150 strokes/min. At the end-point, an aliquot of 1 mL of sample was centrifuged for 10 minutes at 13400 rpm using a Minispin centrifuge and washed twice with water. Between each washing step, the sample was subjected to an ultrasound bath for one minute to remove aggregates formed upon centrifugation. The sample was resuspended in water, to a final volume of 1 mL. Particle size was assessed as aforementioned. Incubation of NP samples in PBS 10 mM pH 7.4 (1:9) (v/v) was used as negative control; BSA solution 1% (w/v) pH 5.1 was used as positive control because it was previously known to induce NP aggregation.

12

2.7.2. Incubation in mouse and human plasma Samples were incubated in mouse plasma and human plasma in accordance with the study of Dobrovolskaia [12]. Mouse whole blood was collected by cardiac puncture using heparin-coated needles while animals were under anesthesia. Plasma was obtained using heparin as anticoagulant by centrifugation at 2000g for 15 minutes at 4°C, stored at -20°C and pooled before use. Prior to the assay, plasma was thawed at room temperature and a short spin was performed to remove large protein precipitates. To evaluate the stability of nanoparticles, 1 mL of plasma was incubated at 37ºC in a shaking water bath (SS40-D Grant, UK) for approximately 10 minutes to equilibrate the medium temperature. Subsequently, 1 mL of sample of 10-fold diluted sample was added to plasma and incubated for 1 hour at 37ºC under agitation of 150 strokes/minute. Afterwards, an aliquot of 1 mL of sample was centrifuged for 10 minutes at 13400 rpm using a Minispin centrifuge and washed twice with water. Between each washing step, the sample was subjected to an ultrasound bath for 1 minute to remove aggregates formed upon centrifugation. The sample was re-suspended in water, to final volume of 1 mL. Particle size was assessed as aforementioned. Incubation of samples in PBS 10 mM pH 7.4 ratio 1:9 (V/V) was used as negative control. Incubation of NP with BSA solution 1% (w/v) pH 5.1 was used as positive control. 2.8. Quantification of protein adsorption at the surface of nanoparticles To determine the extent of the interaction between nanoparticles and proteins, the amount of protein adsorbed was quantified using the Bradford assay [26]. More specifically, after incubation of samples in a BSA 0.4 mg/mL solution in PBS 10 mM pH 7.4, mouse plasma, human plasma and PBS (as negative control) as well as subsequent washing steps as detailed in 2.7., particles were recovered and reacted with

13

the Coomassie Plus Assay reagent according to the specifications provided by the manufacturer for a working range within 1-25 μg/mL of protein. 2.9. Statistical analysis Assays were performed in triplicate, except for quantification of protein adsorption to NP which was performed in duplicate, using three technical replicates. Results are expressed as mean and standard deviation. For the study of zeta potential measurements of NPs after incubation and total protein adsorption to NPs a one-way ANOVA and Tukey post-test were used. GraphPad Prism 5 software was used for statistical analysis.

3. Results 3.1 Nanoparticle characterization PAH-coated and uncoated PLGA NP were characterized in regard to size and surface charge by dynamic light scattering and laser doppler micro-electrophoresis. As evidenced in Table 1, uncoated PLGA NP have an average size of approximately 150 nm and a polydispersity index inferior to 0,1. Furthermore, these NP have a negative surface charge, imparted by PLGA and inferior to -30mV. PAH-coated PLGA nanoparticles have a relatively higher diameter than uncoated PLGA NP, with an average size of 200 nm, and low polydispersity index, although higher than for uncoated NP. Of notice, by the addition of PAH, the resulting NP evidenced an inversion of their surface charge which shifts from -60 mV to 40 mV. It is possible to observe that after this adjustment both NP show a slightly lower average size and polydispersity index.

14

Table 1 Characterization of PLGA and PAH-coated PLGA NP after dilution in water to a concentration range appropriate for analysis and also upon pH adjustment to 7.4 with NaOH 0.01 M. Size and zeta potential measurements have been performed in triplicate using three technical replicates. SAMPLE

MEAN DIAMETER (nm)

PDI

ZETA POTENTIAL (mV)

pH

PLGA NP

145.5 ± 2.4

0.07 ± 0.02

-59.5 ± 1.0

7.4

PLGA NP

153.1± 2.4

0.07 ± 0.01

-44.4 ± 0.7

6.3

PAH-PLGA NP

201.7 ± 0.3

0.08 ± 0.02

39.4 ± 0.8

7.4

PAH-PLGA NP

226.8 ± 3.6

0.10 ± 0.01

51.8 ± 0.9

5.7

Morphology of NP was characterized in regard to their shape and structural properties by TEM (Fig. 1). Uncoated PLGA NP appear to be uniformly spherical, however the sizes measured in TEM are relatively inferior to those obtained by DLS, varying between 70 and 100 nm (A). Nevertheless, this may be due to difficulty in defining the outline of the NP due to the negative contrast process. These NP appear to be homogenous in size, which is in accordance to the polydispersity values obtained by DLS. Despite stained using the same protocol, coated NP are subjected to a positive staining by phosphotungstic acid, as evidenced by the majority of NP being completely blackened, as opposed to the results obtained for uncoated NP. TEM analysis revealed that, in majority, PAH-coated NP had lost their sphericity to become more roughlyshaped. Furthermore, the analysis of several NPs suggests the presence of two distinguishable structures of different densities. Sizing measurements accounting both structures are congruent with the average sizes obtained with DLS before and after coating supporting the hypothesis that the outer layer correspond to the PAH coating 15

(B). PAH-coated PLGA NP are more heterogeneous in size and shape, in comparison to uncoated PLGA NP.

129.05 nm 233.83 nm 104.01 nm

145.59 nm 227.21 nm

A

200 nm

B

500 nm

Fig.1. TEM observation of PLGA NP (A) and PAH-PLGA NP (B). In order to accurately compare both formulations, the concentration of NP was determined by gravimetric methods, after freeze-drying. Uncoated PLGA NP were determined to be at a concentration of approximately 5 mg/mL, whereas PAH-coated PLGA NP had a 10-fold lower yield, with a concentration of 0.5 mg/mL. This difference in concentration may be caused by the interference of the surfactant poloxamer 188 in the suspension of PLGA NP. Therefore, the concentration of NP was further assessed by the fluorescence intensity of FITC groups present in PLGA used for the preparation of both coated and uncoated NP. Using a calibration curve of FITC with a linear range of 1.56 – 1.22x10-2 µg/mL of FITC and a correlation coefficient of 0.9993, it was determined that uncoated PLGA NP had approximately a 3-fold higher concentration in FITC, in comparison to coated NP. Thus, this difference in concentration was considered in further assays in order to maintain a similar NP concentration for both samples. The coating process of PLGA NP with PAH was further assessed by quantification of the amine groups available at the surface NP, imparted by PAH 16

adsorption, using the ninhydrin assay. In the linear range of 0.750 - 0.031 mg/mL, for an aqueous solution of PAH at pH 5.5, the correlation coefficient was 0.9994. Using a 3-fold dilution of PLGA NPs as negative control, the determined concentration of PAH adsorbed to NP was 0.044 mg/mL. Thus, approximately 88% of PAH is detected at the surface of nanoparticles. 3.2. Stability of NP for intravenous administration After incubation with either BSA solution prepared in PBS 10mM at a final protein concentration of 0.4 mg/mL, mouse plasma and human plasma, uncoated and PAH-coated PLGA NP were recovered and characterized according to their mean diameter, polydispersity index and zeta potential (Table 2).

Table 2 Characterization of PAH-coated and uncoated PLGA NP after incubation in different media used to simulate the intravenous environment. ND – Not determined due to aggregation of NP. Negative control – PBS 10 mM pH 7.4 (ratio 1:9 to NP); Positive control – BSA 1% solution pH 5.1 (ratio 1:9 to NP). Results are presented as mean and standard deviation of three measurements using three technical replicates. Sample

Incubation medium

Mean diameter (nm)

PdI

Zeta Potential (mV)

PLGA NP

Mouse plasma

219.3 ± 37.2

0.24 ± 0.05

-21.4 ± 5.5

Human plasma

1247.0 ± 281.0 *

0.40 ± 0.11

ND

BSA solution

166.0 ± 4.5

0.10 ± 0.03

-24.0 ± 2.4

Negative control

157.3 ± 4.2

0.07 ± 0.03

-34.4 ± 4.6

17

PAH NP

Positive control

688.2 ± 114.0 *

0.34 ± 0.12

ND

Mouse plasma

280.4 ± 30.9

0.23 ± 0.11

-25.5 ± 6.0

Human plasma

1429.0 ± 815.0 *

0.65 ± 0.25

ND

BSA solution

290.0 ± 36.8

0.17 ± 0.07

6.8 ± 7.6

Negative control

244.0 ± 5.1

0.12 ± 0.04

38.0 ± 1.3

Positive control

924.5 ± 133.0 *

0.47 ± 0.23

ND

PLGA NP maintain a negative surface charge after incubation in all media. However, there is a significant (P<0,001) difference in the surface charge of NP incubated in the negative control in comparison with BSA solution and mouse plasma. PAH-coated PLGA NP behaved differently in all media when incubated in the absence of proteins, experiencing a significant shift towards neutral zeta potential values when incubated in a BSA solution and further inversion of surface charge when in contact with mouse plasma (P<0.0001). Contact with mouse plasma and BSA solution did not largely affect the mean diameter of NPs, as their profile is almost superimposable to the negative control (Fig.2). Nevertheless, upon incubation in human plasma uncoated PLGA NP experienced a significant shift towards sizes greater than 1 µm, which vary greatly from the results obtained in the absence of protein (negative control). The results obtained for uncoated PLGA NP after incubation in mouse plasma and a BSA solution appear to have a high similarity among each other. Surprisingly, the size distribution of this sample in human plasma does not correspond at all with the profile obtained after incubation in mouse plasma. 18

Fig. 2. Stability of uncoated PLGA NP after incubation in mouse plasma, BSA solution and human plasma in comparison with the negative control. Data is represented as mean of three measurements using three technical replicates. As evidenced in Fig. 3, the incubation of NP in a negative control is similar to that of NP before incubation, the aggregation process is not expected to be an artifact of the methodology. Concomitantly, a positive control was used in order to guarantee that possible aggregation would be detected.

Fig. 3. Comparative size distribution profiles of uncoated PLGA NP before incubation

19

and positive and negative controls. Data is represented as mean of three measurements using three technical replicates. After incubation in mouse plasma PAH-coated NPs maintained a similar size distribution to the negative control and the same is observed for incubation in a BSA solution (Fig. 4). Finally, when subjected to contact with human plasma, PAH-coated PLGA NPs also exhibited a significant shift towards sizes greater than 1 µm, resembling the behavior of uncoated NPs in this medium (A). Appropriate controls were used during this assay, as previously mentioned (B).

A

B

20

Fig. 4. Stability of PAH-PLGA NP after incubation in mouse plasma, BSA solution and human plasma (A) and assay controls (B). Data is represented as mean of three measurements using three technical replicates.

3.3. Total protein adsorption at the surface of NP Given the unexpected variance between the stability profiles of both sets of NPs in human and mouse plasmas, we hypothesized if this different behavior could be related to the amount of proteins that interacted with the NPs. In view of this, after incubation and subsequent washing steps for the recovery of NP, the concentration of proteins still present in the sample, which would correspond to the hard corona, was determined using the Bradford method. In the range of 2.5 – 25.0 µg/mL of protein a correlation coefficient of 0.999 was obtained and the protein concentration in each sample was assessed, as represented in Fig. 5.

*

A *** *** *

21

***

B

Fig. 5. Total protein adsorption to NPs after incubation in different media; PLGA NPs (A); PAH-coated PLGA NPs (B). Data is represented as mean and standard error of mean of two measurements, each performed in triplicate. Diverse protein concentrations are observed upon incubation of PLGA NP with different media (Fig. 5A). Firstly, incubation in a BSA solution does not lead to a substantial variation in protein concentration. However, after contact with mouse plasma, uncoated PLGA NP experience significantly higher (P<0.05) protein adsorption to their surface in comparison with the results found for a BSA solution, with mean total protein concentration of 4.1 ± 2.6 µg/mL. More importantly, the protein concentration found after incubation in human plasma differs significantly (P<0.001) from all the other media tested, being approximately 2.5-fold higher than in mouse plasma, with average total concentration of 10.4 ± 3.0 µg/mL. It is possible to observe a different pattern of protein adsorption, especially in regard to incubation in a BSA solution (Fig. 5B). Indeed, NP incubated in the latter demonstrate a high concentration in protein, with mean total protein concentration of 8.3 ± 0.4 µg/mL. This result contrasts to the 22

behavior found for uncoated NP and also for the levels of protein adsorption found after contact with mouse plasma, which are lower for the latter with average total protein concentration of 4.1 ± 1 µg/mL. Nevertheless, the concentration of proteins adsorbed to PAH-coated PLGA NPs are also significantly higher after incubation in human plasma (12.8 ± 3.4 µg/mL) in comparison with mouse plasma (P<0.001) and a BSA solution (P<0.01). Total protein concentration of both media before incubation was quantified and human plasma had a total protein concentration of approximately 90.8 g/L whereas mouse plasma evidenced a substantially inferior total protein concentration of approximately 15.5 g/L.

4. Discussion Polymeric NP have been successfully prepared and modified in order to produce uncoated and PAH-coated PLGA NP, whose properties were assessed in regard to their potential for intravenous administration. This was achieved either by assessing the main physicochemical properties of NP after preparation as well as by analyzing how the latter were affected by their interactions within a biological environment. The size results obtained by DLS were supported by TEM analysis and together evidenced NP have a similar size range to each other and are both relatively homogenous. Zeta potential results indicate that uncoated PLGA NP have a negative surface charge whereas PAH-coated NP are positively charged. The inversion of zeta potential of PLGA NP after the addition of PAH is indicative of the efficiency of the coating process as the latter is a cationic polymer which interacted with the negatively charged NP, becoming electrostatically adsorbed. Zeta potential results obtained for both uncoated and PAH-coated PLGA NP are suggestive of good colloidal stability of 23

NP in the medium in which they are dispersed. The adjustment of pH from the one obtained after NP preparation and dispersion in water to 7.4, representative of physiological pH, introduced mild changes to the results observed previously. Mainly, pH adjustment translated into a decrease in size and polydispersity for both formulations whereas opposite results between uncoated and PAH-coated PLGA NP were obtained for zeta potential measurements. There was an increase in the surface charge of uncoated PLGA NP while PAH-coated PLGA NP experienced a decrease in zeta potential. Even though for these samples there were no drastic changes promoted by pH adjustment, these results provide a better understanding of their characteristics at physiological pH and are indicative of good colloidal stability. This is especially important for polymeric NP as polyelectrolytes are protonated differently depending on the pH of the medium in which they are dispersed, meaning that this assessment could be predictive of lack of stability promoted by pH. TEM analysis indicated that PLGA NP have a perfectly spherical shape while PAH-PLGA NP are more heterogeneous and supports the observation that the coating process was successful. Furthermore, ninhydrin assay showed that the PAH coating process had an efficiency of approximately 88%. Altogether, these results suggest that PLGA NP and surface-modified PLGA NP were efficiently obtained. The protocol used to assess the stability of NP has been adapted to allow the study of uncoated and PAH-coated PLGA NP. Of notice, since the coating process was promoted by electrostatic interactions between PLGA and PAH and is therefore dependent on the strength of the interaction between both polymers, the washing steps could not be performed using a buffer solution, such as PBS 10 mM. This was resultant from the ionic strength of the buffer which destabilized the coating leading to NP collapse (data not shown). The ionic strength of PBS is strong enough to destroy the 24

charge balance of PAH coated polymeric NP [27]. Thus, all washing steps were performed using water. Appropriate negative and positive controls suggest the methodology did not interfere on the assessment of NP stability and is adequate for the study of polymeric NP. The incubation process of NP in the media used to mimic the intravenous environment had a significant impact, especially on PAH-coated PLGA NP which experience the most drastic changes to their surface charge. The interactions between proteins and NP are modulated by a variety of factors [28] among which electrostatic interactions may be the reason there is a higher interaction between BSA and PAHcoated PLGA NP, whose previous positive surface charge is driven to neutral zeta potential values. The formation of PAH/BSA complexes at pH 7.4 has been described as a spontaneous process [29]. In contrast, uncoated PLGA NP, which are negatively charged, experience a significant shift towards increased zeta potential values which suggest that moderate adsorption of BSA has occurred, as a relatively similar surface charge is maintained. Since no different surface charge between PLGA NP incubated in BSA solution and mouse plasma was observed it can be admitted that positively charged proteins with a concentration in the NP corona of at least 1 µg/mL do not contribute extensively to the formation of the protein corona of these NP which could suggest that other forces dominate over charge interactions in this case. Plasma proteins in general are negatively charged, especially albumin, except for some immunoglobulin molecules that are positively charged. The adsorption of BSA onto nanoparticle surface has been reported for PLGA [30] and PLGA NPs have exhibited the adsorption of three major plasma negatively charged proteins: fibrinogen > albumin > globulin (γ-globulin) [31]. As for PAH-coated PLGA NP, there was an even further reduction in their zeta potential which experiences an extensive inversion. This result suggests that an 25

increased amount of negatively charged proteins are adsorbed to their surface, in comparison with the BSA solution. Uncoated PLGA NP are hydrophobic, as opposed to PAH-coated PLGA NP whose coating imparts them increased hydrophilicity, and this property may also modulate the interaction with different proteins [32]. It is also noteworthy that PLGA nanoparticles have been prepared in the presence of a surfactant which may itself modulate the interaction between NP and proteins as it is expected to become, to some extent, absorbed to the surface of NP. Even though changes to the surface charge of NP provide an insight into the interaction with NP, particle size gives an understanding of their colloidal stability. Uncoated PLGA NP good colloidal stability after incubation in a BSA solution and mouse plasma are congruent with the zeta potential values. Moreover, it is important to denote the similarity between the stability profiles obtained after incubation in mouse plasma and BSA solution, for both formulations, suggests that, for uncoated and PAHcoated PLGA NP incubation in a BSA solution could be predictive of their behavior in mouse plasma, indicating the single protein solution as an early screening approach to the stability of NP in mouse plasma using more easily available and ethically-desirable assays. However, the opposite is observed after incubation in human plasma, where the observed shift towards sizes greater than 1 µm indicate that aggregation has occurred. Not only was this result unexpected but also it indicates that there is no correlation between the stability of NP in mouse and human plasma. Furthermore, this aggregation occurred independently of the surface composition of NP. This was found to be a particularly important observation as, to the best of our knowledge, there has been no comparative study between the stability of NP in plasmas of different origins, human and mouse. Even further, it could be indicative of a contributing factor to the lack of correlation between in vitro models and, in more advanced stages, lack of translation of 26

delivery systems between animal models and humans. It is important to denote that this observation has only been addressed for these polymeric NP and could be materialdependent thus leading to diverse results for NP of different composition. However, it instigates further consideration into the stability of NP during their development process and into the correlation between studies used to characterize NP for intravenous administration. Indeed, as previously mentioned, it has previously been demonstrated that after incubation in mouse and human plasma lipidic NP acquire protein coronas of different composition [17, 33]. Congruently, we observe that the difference between media could impact beyond corona composition and translate into loss of colloidal stability. Given these observations, we investigated if these differences could be correlated with the amount of proteins adsorbed to NP. A negligible concentration of proteins is adsorbed to uncoated PLGA NP when incubated in a BSA solution, whilst PAH-coated PLGA NP evidence a significant increase in the concentration of proteins adsorbed to them, which supports the hypothesis that an increased interaction with BSA is observed for surface-modified PLGA NP. In addition, there is a significant difference between both latter results and the ones obtained after incubation in mouse plasma. PLGA NP reveals a significantly increased amount of proteins adsorbed to them, in comparison with the BSA solution, whereas PAH-coated PLGA NP show a decreased protein adsorption. Uncoated PLGA NP may interact less with the single protein solution in comparison to a higher complex medium, while PAH-coated PLGA NP interact in excess with BSA. Lastly, a significantly higher concentration of proteins was found after incubation of uncoated and PAH-coated PLGA NP in human plasma. Since a significantly increased protein concentration was found for NP incubated in human plasma we hypothesized if this result could be promoted by an 27

observable difference between this medium and mouse plasma. Thus, the total protein concentrations originally present in the mouse and human plasma samples used for this study were determined and results evidenced an approximately 6-fold difference between the total protein concentration in both media. The concentration found for human plasma is in agreement with the values found in the literature [34] and supports the adequacy of the methodology. In contrast, the results obtained for mouse plasma are inferior to those previously reported [35, 36]. As mice plasma samples used in this study were the result of a pool of plasma from different strains, this source of variability is not expected to have impacted on the results. Alternatively, the presence of protein precipitates that are found once plasma is thawed and removed could have led to a significant loss of protein. This practice is an inevitability of the process and is common to other protocols found for the study of the protein corona [37]. It could however be a variability source which should be accounted for. This enhances the need for optimization and standardization of protocols for the use of biological media in the study of the nano-bio interaction, as other authors have also reported other protocolrelated issues that influence the results obtained for the in vitro study of the biological behavior of NP [38]. Taking into consideration the previously described difference in adsorption of proteins in plasma of different origins [17] it cannot be excluded that an inter-species variability in plasma composition may motivate a different response of NP as already described for drugs. In general, there is a good correlation between the protein binding observed for drug molecules in human plasma and that for rat, dog and mouse plasma, although compounds tend to be slightly more bound to human plasma proteins compared to those from the pre-clinical species [39]. Thus, the results herein presented instigate the need for further understanding of the in vitro models used to study the 28

behavior of delivery systems for intravenous administration in regard to their comparability and in vitro-in vivo correlation and support the need for a case-by-case approach to the study of NP.

5. Conclusion Uncoated PLGA and PAH-coated PLGA NP of similar size but different surface composition and charge have been prepared and characterized according to their potential for intravenous administration. It was found that adjustment of pH to determine the size and surface charge of NP could be an important tool to predict the instability of NP at physiological pH at an early stage. The stability of both formulations was assessed in three media where it was possible to observe that NP were stable in a BSA solution and mouse plasma but not in human plasma. These results were independent of surface composition. The difference in behavior of NP in mouse and human plasma was unanticipated and could be associated with a significant difference in total protein concentration found between both plasma samples. The protein levels of mouse plasma were inferior to those expected which suggests the formation of protein aggregates after plasma thawing could be introducing a significant variability for this particular sample. Thus, we suggest the parameter total protein concentration should be used as medium control to improve the comparability between results. In addition, the results here presented enhance the need to put further consideration into the composition of samples of different species origin in a bid to promote the standardization of assays and to use appropriate biological media to anticipate, with increased representativeness and reproducibility, the stability of NP.

29

6. Acknowledgements This work was financially supported by national funds through FCT - Fundação para a Ciência e a Tecnologia/MEC - Ministério da Educação e Ciência and co-funded by FEDER funds within the partnership agreement PT2020 related with the research unit number 4293.

7. Disclosure The authors declare no conflict of interest.

30

References: [1]

M. Karimi, A. Ghasemi, P. Sahandi Zangabad, R. Rahighi, S.M. Moosavi Basri, H. Mirshekari, M. Amiri, Z. Shafaei Pishabad, A. Aslani, M. Bozorgomid, D. Ghosh, A. Beyzavi, A. Vaseghi, A.R. Aref, L. Haghani, S. Bahrami, M.R. Hamblin, Smart micro/nanoparticles in stimulus-responsive drug/gene delivery systems, Chem Soc Rev, 45 (2016) 1457-1501.

[2]

S. Tenzer, D. Docter, J. Kuharev, A. Musyanovych, V. Fetz, R. Hecht, F. Schlenk, D. Fischer, K. Kiouptsi, C. Reinhardt, K. Landfester, H. Schild, M. Maskos, S.K. Knauer, R.H. Stauber, Rapid formation of plasma protein corona critically affects nanoparticle pathophysiology, Nat Nano, 8 (2013) 772-781.

[3]

A. Salvati, A.S. Pitek, M.P. Monopoli, K. Prapainop, F.B. Bombelli, D.R. Hristov, P.M. Kelly, C. Aberg, E. Mahon, K.A. Dawson, Transferrinfunctionalized nanoparticles lose their targeting capabilities when a biomolecule corona adsorbs on the surface, Nat Nanotechnol, 8 (2013) 137-143.

[4]

D.E. Owens Iii, N.A. Peppas, Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles, Int J Pharm, 307 (2006) 93-102.

[5]

M. Lundqvist, J. Stigler, G. Elia, I. Lynch, T. Cedervall, K.A. Dawson, Nanoparticle size and surface properties determine the protein corona with possible implications for biological impacts, PNAS, 105 (2008) 14265-14270.

[6]

U. Sakulkhu, M. Mahmoudi, L. Maurizi, J. Salaklang, H. Hofmann, Protein corona composition of superparamagnetic iron oxide nanoparticles with various physico-chemical properties and coatings, Sci Rep, 4 (2014) 5020.

[7]

D.J. Sobczynski, P. Charoenphol, M.J. Heslinga, P.J. Onyskiw, K. Namdee, A.J. Thompson, O. Eniola-Adefeso, Plasma protein corona modulates the vascular 31

wall interaction of drug carriers in a material and donor specific manner, PLoS One, 9 (2014) e107408. [8]

B.W. Neun, M.A. Dobrovolskaia, Method for analysis of nanoparticle hemolytic properties in vitro, Methods Mol Biol, 697 (2011) 215-224.

[9]

B.W. Neun, M.A. Dobrovolskaia, Method for in vitro analysis of nanoparticle thrombogenic properties, Methods Mol Biol, 697 (2011) 225-235.

[10] B.W. Neun, M.A. Dobrovolskaia, Qualitative analysis of total complement activation by nanoparticles, Methods Mol Biol, 697 (2011) 237-245. [11] T.L. Moore, L. Rodriguez-Lorenzo, V. Hirsch, S. Balog, D. Urban, C. Jud, B. Rothen-Rutishauser, M. Lattuada, A. Petri-Fink, Nanoparticle colloidal stability in cell culture media and impact on cellular interactions, Chem Soc Rev, 44 (2015) 6287-6305. [12] M.A. Dobrovolskaia, A.K. Patri, J. Zheng, J.D. Clogston, N. Ayub, P. Aggarwal, B.W. Neun, J.B. Hall, S.E. McNeil, Interaction of colloidal gold nanoparticles with human blood: effects on particle size and analysis of plasma protein binding profiles, Nanomedicine, 5 (2009) 106-117. [13] D.D. Jurašin, M. Ćurlin, I. Capjak, T. Crnković, M. Lovrić, M. Babič, D. Horák, I. Vinković Vrček, S. Gajović, Surface coating affects behavior of metallic nanoparticles in a biological environment, Beilstein J Nanotechnol, 7 (2016) 246262. [14] L. Novo, L.Y. Rizzo, S.K. Golombek, G.R. Dakwar, B. Lou, K. Remaut, E. Mastrobattista, C.F. van Nostrum, W. Jahnen-Dechent, F. Kiessling, K. Braeckmans, T. Lammers, W.E. Hennink, Decationized polyplexes as stable and safe carrier systems for improved biodistribution in systemic gene therapy, J Control Release, 195 (2014) 162-175. 32

[15]

L. Landgraf, C. Christner, W. Storck, I. Schick, I. Krumbein, H. Dahring, K. Haedicke, K. Heinz-Herrmann, U. Teichgraber, J.R. Reichenbach, W. Tremel, S. Tenzer, I. Hilger, A plasma protein corona enhances the biocompatibility of Au@Fe3O4 Janus particles, Biomaterials, 68 (2015) 77-88.

[16] C. Oliveira, I. Silveira, F. Veiga, A.J. Ribeiro, Recent advances in characterization of nonviral vectors for delivery of nucleic acids: impact on their biological performance, Expert Opin Drug Deliv, 12 (2015) 27-39. [17] D. Pozzi, G. Caracciolo, A.L. Capriotti, C. Cavaliere, G. La Barbera, T.J. Anchordoquy, A. Laganà, Surface chemistry and serum type both determine the nanoparticle–protein corona, J Proteomics, 119 (2015) 209-217.

[18] W. Bao, R. Liu, Y. Wang, F. Wang, G. Xia, H. Zhang, X. Li, H. Yin, B. Chen, PLGA-PLL-PEG-Tf-based targeted nanoparticles drug delivery system enhance antitumor efficacy via intrinsic apoptosis pathway, Int J Nanomedicine, 10 (2015) 557-566. [19] R. Luo, B. Neu, S.S. Venkatraman, Surface functionalization of nanoparticles to control cell interactions and drug release, Small, 8 (2012) 2585-2594. [20] K. Sempf, T. Arrey, S. Gelperina, T. Schorge, B. Meyer, M. Karas, J. Kreuter, Adsorption of plasma proteins on uncoated PLGA nanoparticles, Eur J Pharm Biopharm, 85 (2013) 53-60. [21] S. Basu, R. Harfouche, S. Soni, G. Chimote, R.A. Mashelkar, S. Sengupta, Nanoparticle-mediated targeting of MAPK signaling predisposes tumor to chemotherapy, PNAS, 106 (2009) 7957-7961.

33

[22] A.M. Hollman, D. Bhattacharyya, Pore Assembled Multilayers of Charged Polypeptides in Microporous Membranes for Ion Separation, Langmuir, 20 (2004) 5418-5424. [23] M.M. Leane, R. Nankervis, A. Smith, L. Illum, Use of the ninhydrin assay to measure the release of chitosan from oral solid dosage forms, Int J Pharm, 271 (2004) 241-249. [24] S. Fertsch-Gapp, M. Semmler-Behnke, A. Wenk, W.G. Kreyling, Binding of polystyrene and carbon black nanoparticles to blood serum proteins, Inhal Toxicol, 23 (2011) 468-475. [25] W.G. Kreyling, S. Fertsch-Gapp, M. Schäffler, B.D. Johnston, N. Haberl, C. Pfeiffer, J. Diendorf, C. Schleh, S. Hirn, M. Semmler-Behnke, M. Epple, W.J. Parak, In vitro and in vivo interactions of selected nanoparticles with rodent serum proteins and their consequences in biokinetics, Beilstein J Nanotechnol, 5 (2014) 1699-1711. [26] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal Biochem, 72 (1976) 248-254. [27] L.-l. Xie, W.-j. Tong, J.-q. Xu, C.-y. Gao, Multilayers and poly(allylamine hydrochloride)-graft-poly(ethylene glycol) modified bovine serum albumin nanoparticles: Improved stability and pH-responsive drug delivery, Chin J Polym Sci, 30 (2012) 719-726. [28] I. Lynch, K.A. Dawson, Protein-nanoparticle interactions, Nano Today, 3 (2008) 40-47.

34

[29] V. Ball, M. Winterhalter, P. Schwinte, P. Lavalle, J.C. Voegel, P. Schaaf, Complexation Mechanism of Bovine Serum Albumin and Poly(allylamine hydrochloride), J Phys Chem B, 106 (2002) 2357-2364. [30] C. Fornaguera, G. Caldero, M. Mitjans, M.P. Vinardell, C. Solans, C. Vauthier, Interactions of PLGA nanoparticles with blood components: protein adsorption, coagulation, activation of the complement system and hemolysis studies, Nanoscale, 7 (2015) 6045-6058. [31] Y.M. Thasneem, S. Sajeesh, C.P. Sharma, Effect of thiol functionalization on the hemo-compatibility of PLGA nanoparticles, Journal of biomedical materials research. Part A, 99 (2011) 607-617. [32] A. Gessner, R. Waicz, A. Lieske, B.R. Paulke, K. Mäder, R.H. Müller, Nanoparticles with decreasing surface hydrophobicities: influence on plasma protein adsorption, Int J Pharm, 196 (2000) 245-249. [33] G. Caracciolo, D. Pozzi, A. L. Capriotti, C. Cavaliere, S. Piovesana, G. La Barbera, A. Amici, A. Laganà, The liposome–protein corona in mice and humans and its implications for in vivo delivery, J Mater Chem B, 2 (2014) 7419-7428 [34] J.N. Adkins, S.M. Varnum, K.J. Auberry, R.J. Moore, N.H. Angell, R.D. Smith, D.L. Springer, J.G. Pounds, Toward a human blood serum proteome: analysis by multidimensional separation coupled with mass spectrometry, Mol Cell Proteomics, 1 (2002) 947-955. [35] S.E. Bernstein, Physiological Characteristics, in: E.L. Green (Ed.) Biology of the Laboratory Mouse, Dover Publications, Inc., New York, 1966. [36] U. Windberger, A. Bartholovitsch, R. Plasenzotti, K.J. Korak, G. Heinze, Whole blood viscosity, plasma viscosity and erythrocyte aggregation in nine mammalian

35

species: reference values and comparison of data, Exp Physiol, 88 (2003) 431440. [37] M.P. Monopoli, A.S. Pitek, I. Lynch, K.A. Dawson, Formation and characterization of the nanoparticle-protein corona, Methods Mol Biol, 1025 (2013) 137-155. [38] S. Palchetti, V. Colapicchioni, L. Digiacomo, G. Caracciolo, D. Pozzi, A.L. Capriotti, G. La Barbera, A. Lagana, The protein corona of circulating PEGylated liposomes, Biochim Biophys Acta, 1858 (2016) 189-196. [39] N. Colclough, L. Ruston, J.M. Wood, P.A. MacFaul, Species differences in drug plasma protein binding, Med Chem Comm, 5 (2014) 963-967.

36