Proteomics Study of a Designed Nanoparticle-Protein Corona Made of Animal Model Plasma

(1)

Linköping University | Department of Physics, Chemistry and Biology Master thesis, 30 hp | Master of Science in Chemical Biology Spring term 2020 | LITH-IFM-A-EX—20/3777--SE

Proteomics Study of a

Designed

Nanoparticle-Protein Corona Made of

Animal Model Plasma

Elin Nilsson

Examiner, Karin Enander Tutor, Susana Cristobal

(2)

UPPHOVSRÄTT

Detta dokument hålls tillgängligt på Internet – eller dess framtida ersättare – under 25 år från publiceringsdatum under förutsättning att inga extraordinära omständigheter uppstår. Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner, skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för ickekommersiell forskning och för undervisning. Överföring av upphovsrätten vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av dokumentet kräver upphovsmannens medgivande. För att garantera äktheten, säkerheten och tillgängligheten finns lösningar av teknisk och administrativ art. Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i den omfattning som god sed kräver vid användning av dokumentet på ovan beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan form eller i sådant sammanhang som är kränkande för upphovsmannens litterära eller konstnärliga anseende eller egenart. För ytterligare information om Linköping University Electronic Press se förlagets hemsida http://www.ep.liu.se/.

COPYRIGHT

The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional circumstances. The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/hers own use and to use it unchanged for non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility. According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement. For additional information about the Linköping University Electronic Press and its procedures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/.

(3)

Datum

Date 2020-06-20

Avdelning, institution Division, Department

Department of Physics, Chemistry and Biology Linköping University

URL för elektronisk version

ISBN

ISRN: LITH-IFM-A-EX--20/3777--SE

_________________________________________________________________

Serietitel och serienummer ISSN

Title of series, numbering ______________________________

Språk Language Svenska/Swedish Engelska/English ________________ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport _____________ Titel Title

Proteomics Study of a Designed Nanoparticle-Protein Corona Made of Animal Model Plasma

Författare Author Elin Nilsson

Nyckelord Keyword

Nanoparticle, Protein corona, Nanoparticle-protein corona,SUrface proteomics, Safety, Targeting, and Uptake. Sammanfattning

Abstract

Nanoparticles are currently finding increasing use as drug delivery systems in the treatment of cancer and other disorders. When nanoparticles are introduced into body fluids, they adsorb proteins forming a coating called protein corona. The protein corona is vital since it controls biological responses of nanoparticles through interactions with cells and biological barriers. Due to the dynamic behaviour of protein-protein and protein-nanoparticle interactions, the protein corona evolves during circulation in the body. This results in difficulties to predict the biological behaviour and outcome of nanoparticles. In this work, it is hypothesised that a nanoparticle-protein corona (NP-PC) enriched in specific proteins could serve as a model to determine if the design and formation of a patient-specific nanodrug-protein corona could offer a novel approach to control nanodrug-protein corona evolution. Through usage of a model nanoparticle and model plasmas and by applying shotgun proteomics and SUrface proteomics, Safety, Targeting, and Uptake (SUSTU), NP-PC proteins were identified and quantified. The results indicate that desirable proteins are maintained in the protein corona surface when nanoparticles with a pre-made corona are introduced into model plasma. This implies that a designed NP-PC would be a strategy to control nanodrug-protein corona evolution, offering a route to improve nanodrug targeting and uptake by cells.

(4)

Abstract

Nanoparticles are currently finding increasing use as drug delivery systems in the treatment of cancer and other disorders. When nanoparticles are introduced into body fluids, they adsorb proteins forming a coating called protein corona. The protein corona is vital since it controls biological responses of nanoparticles through interactions with cells and biological barriers. Due to the dynamic behaviour of protein-protein and protein-nanoparticle interactions, the protein corona evolves during circulation in the body. This results in difficulties to predict the biological behaviour and outcome of nanoparticles. In this work, it is hypothesised that a nanoparticle-protein corona (NP-PC) enriched in specific proteins could serve as a model to determine if the design and formation of a patient-specific nanodrug-protein corona could offer a novel approach to control nanodrug-protein corona evolution. Through usage of a model nanoparticle and model plasmas and by applying shotgun proteomics and SUrface proteomics, Safety, Targeting, and Uptake (SUSTU), NP-PC proteins were identified and quantified. The results indicate that desirable proteins are maintained in the protein corona surface when nanoparticles with a pre-made corona are introduced into model plasma. This implies that adesigned NP-PC would be a strategy to control nanodrug-protein corona evolution, offering a route to improve nanodrug targeting and uptake by cells.

(5)

Acknowledgement

I would like to thank several people for their contributions to this project. Thank you, Susana Cristobal, for giving me the opportunity to perform this project and letting me be a part of your team. Thank you, Ana Maria Carrasco Del Amor, for your help and great support throughout this project. Thank you, Veronica Lizano Fallas, for your advice and help in the lab. Lastly, thank you, Jennifer Sundström, for your opposition and feedback.

(6)

Abbreviations

ANOVA Analysis of variance ESI Electrospray ionization

FASP Filter aided sample preparation FDR False discovery rate

LC Liquid chromatography

LC-MS/MS Liquid chromatography coupled tandem mass spectrometry MS Mass spectrometry

m/z Mass-to-charge ratio MS/MS Tandem mass spectrometry NP-PC Nanoparticle-protein corona NSAF Normalized spectral abundance

factor

PSM Peptide spectrum match ROS Reactive oxygen species

RT Room temperature

SAF Spectral abundance factor SUSTU SUrface proteomics, Safety,

Targeting, and Uptake

Chemical Denotations

ACN Acetonitrile BCA Bicinchoninic acid BSA Bovine serum albumin DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid FA Formic acid HCl Hydrochloric acid IAA Iodoacetamide

NaCl Sodium chloride NHS N-Hydroxysuccinimide PBS Phosphate-buffered saline PEG Polyethylene glycol SDS Sodium dodecyl sulfate

SS Disulfide

TiO2 Titanium dioxide

(7)

1. Introduction

Nanoparticles hold a great potential in medicinal applications [1] especially in cancer treatment [2], [3]. Perhaps, nanoparticles are most promising as drug delivery systems to treat cancer [2]-[4], since they enable a high drug load to be stored and protected [5], [6]. In addition, cancer drugs packed in nanoparticles are distributed in a lesser extent to non-tumor cells. Consequently, targeting of tumor cells is improved and damage to non-tumor cells is reduced. [3] Moreover, nanoparticles as drug delivery systems provide longer circulation times [3], which further improves targeting [6].

When encountering body fluids, nanoparticles adsorb proteins and a coating called protein corona is formed [7]. The protein corona is vital since it controls biological responses of nanoparticles [6] through interactions with cells and biological barriers [8]. However, not all corona proteins can interact with the surrounding. Only the surface proteins are able to do this [8]. Ultimately, surface proteins will determine nanoparticle faith [9]. Despite numerous studies, only little is known about the NP-PC surface [9]. Kuruvilla, et al. (2017) developed a method, SUSTU, to identify and quantify NP-PC surface proteins. SUSTU is straightforward and generates only relevant information for understanding biological interactions in contrast to conventional methods. Kuruvilla, et al. (2017) used SUSTU along with titanium dioxide (TiO2) nanoparticles and cell culture medium supplemented with 10 % heat

inactivated fetal bovine serum to prove that less proteins are identified in the surface than in the entire NP-PC. Also Berlin (2019) used SUSTU along with TiO2 nanoparticles and rat plasma and could prove a

lower diversity among surface proteins than hard corona proteins. An additional approach to analyse the NP-PC surface is epitope mapping. This method examines if motifs of surface molecules able to interact with cell receptors or biological barriers are present and available for recognition and if so in what quantity [10]. Epitope mapping is similar to SUSTU as both investigate surface features and employ mass spectrometry (MS). However, their approach to analyse the surface differs.

The NP-PC evolves as it is sequentially incubated in different biological fluids [9]. Yet it does not change completely; it retains features from the primary incubation [11] which is why the primary incubation greatly influences the biological behaviour [9]. It is thus important to consider how to administer nanoparticles since it matters for the biological outcome which body fluid they encounter first [12]. Once a nanoparticle is administered it might circulate for hours before reaching cells of remote organs [13]. During this time, the protein corona will encounter multiple body fluids and most likely evolve. Consequently, a range of nanoparticle faiths are possible. [8] This is problematic since varying responses makes it hard to predict the biological behaviour and outcome [9]. However, by controlling the evolution of a nanodrug-protein corona through formation of a patient specific nanodrug-protein corona prior to administration, an extended circulation time could be envisioned. This would pave the way for improved targeting and increased cellular uptake.

This project aimed to:

• Form and characterize an NP-PC enriched in specific proteins able to serve as a model to determine if the design and formation of a patient-specific nanodrug-protein corona could offer a novel solution to control nanodrug-protein corona evolution.

• Examine if desirable proteins would be maintained in the protein corona surface after a pre-made corona is introduced into model plasma.

Through usage of a model nanoparticle and model plasmas and applying shotgun proteomics and SUSTU, NP-PC proteins were identified and quantified and further insight was gained to whether it is possible to design a patient-specific nanodrug-protein corona.

(10)

2

This project used a straightforward model approach to study nanoparticle behaviour in different environments. Two model plasmas were used, depleted and non-depleted rat plasma. These were well-suited since a rat model can provide larger quantities of plasma than for instance a mouse model. In addition, there is an existing kit to enrich rat plasma which was used in this project. Note that depletion was made to generate two separate model plasmas. The idea was not to use depleted plasma to form the ultimate patient-specific nanodrug-protein corona, but to form an NP-PC enriched in specific proteins.

1.1. Objectives of the Project

The main goal of this project was to form and characterize an NP-PC and to examine if it would maintain its surface features when being transferred from one model rat plasma to another. This was further divided into 4 sub-goals, which are stated below.

Sub-goal 1: Incubate model nanoparticles in rat plasma depleted of the seven most abundant proteins to form an NP-PC. This NP-PC would be a model of the patient-specific nanodrug-protein corona. To identify and quantify corona proteins, label-free liquid chromatography coupled tandem mass spectrometry (LC-MS/MS) and SUSTU was performed.

Sub-goal 2: Incubate model nanoparticles in non-depleted rat plasma. This NP-PC would be a model of the situation when a nanodrug is directly administrated to a patient. To identify and quantify corona proteins, label-free LC-MS/MS and SUSTU was performed.

Sub-goal 3: Incubate model nanoparticles in depleted rat plasma before transferring them for further incubation in non-depleted rat plasma. This NP-PC would be a model of the situation when a nanodrug is being administrated to a patient after formation of a patient-specific nanodrug-protein corona. It would moreover mimic the evolution of a nanodrug-protein corona, creating a so-called evolved corona. To identify and quantify evolved corona proteins, label-free LC-MS/MS and SUSTU was performed.

Sub-goal 4: Compare the characterized NP-PCs made of the depleted and non-depleted rat plasma, and the evolved protein corona.

1.2. Process

The project time plan along with process planning and analysis can be found in Appendix A.

1.3. Ethical Issues

1.3.1. Rat Plasma

Rat plasma from Sprague-Dawley rats was received from the University of the Basque Country UPV/EHU where all procedures involving animal handling had been approved by the Ethics Committee for Animal Welfare of the University of the Basque Country UPV/EHU. Sprague-Dawley rats were sacrificed with pentobarbital injected intraperitoneally (120 mg/kg) and blood was subsequently isolated. Sprague-Dawley rats are outbred rats commonly used as animal models in medicinal research [14].

More information regarding sacrificed rats can be found in table B, Appendix B.

1.3.2. Chemical Waste

Handling and disposal of chemical waste was done in accordance with guidelines from Linköping University [15].

(11)

3

2. Theory and Methodology

2.1. Theory

Provided in this section is information regarding nanoparticles.

2.1.1. Nanoparticles

Nanomaterials are key in nanotechnology which has numerous applications including medicine. Size, particle size distribution and surface area must be of certain values for a material to be defined as a nanomaterial. Nanomaterials have a size of 1 to 100 nm, a particle size distribution ≥ 50 % and a surface area to volume ratio > 60 m2_/cm3_{. [7] Nanoparticles belong to a class of nanomaterials with three}

dimensions where all dimensions are less than 100 nm [7], [4]. They are classified according to their chemical composition for instance as organic or inorganic [7]. Inorganic nanoparticles are commonly made of metal oxides and metals [4].

2.1.2. Titanium Dioxide Nanoparticles

TiO2 exists in seven polymorphs out of which three are naturally occurring: anatase (tetragonal

structure), rutile (tetragonal structure) and brookite (orthorhombic structure). Anatase is found at lower temperatures (300-550˚C) and can be transformed to rutile at higher temperatures (600-1100˚C). Mixtures of these are thus common at synthesis which makes it hard to produce identical batches. Reproducibility is consequently a problem in manufacturing. [17]

TiO2 is a photocatalyst and a semiconductor, with a large band gap between the valence band and the

conduction band. Immediately as UV light with energy of the same size as the band gap interacts with the material, photocatalysis will occur. [18] This makes electrons from the valence band move to the conduction band. As electrons move, they generate holes in the valence band able to interact with water to produce hydroxyl radicals. Redundant electrons in the conduction band can further interact with oxygen to generate superoxide radicals, see figure 1 for full photocatalytic mechanism. [19]

Figure 1. Photocatalytic mechanism of TiO2. Holes in the valence band are annotated by hvb+ and electrons in the conduction

(12)

4

TiO2 nanoparticles are chemically stabile, can be produced at a low cost [18] and have a low toxicity

[19]. Due to these qualities they are commonly employed to purify air [20] and water by degrading wastewater and organic pollutants [17], [18]. Moreover, TiO2 nanoparticles are used in several

commercial products such as sunscreen, paint and toothpaste [20]. Further use of TiO2 nanoparticles

is within industry [21].

TiO2 nanoparticles can interact with cells and have shown to be toxic [22], [21]. Toxicity is most likely

caused by reactive oxygen species (ROS) [22]. In order to produce ROS, UV light must trigger photocatalysis. However, UV light cannot reach deep tissues. This implies that it would be difficult for TiO2 nanoparticles to do any internal damage through production of ROS. [21] In addition, as TiO2

nanoparticles encounter body fluids proteins adsorb which reduce production of ROS [23]. A study made by Thevenot, et al. (2008) could confirm low toxic effects from TiO2 nanoparticles on 3T3

fibroblasts.

2.1.3. Nanoparticles and Protein Coronas

Nanoparticles possess unique physicochemical properties owing to their small size [24], [25]. A small size yields a high surface area to volume ratio and consequently a high surface energy [7]. Biomolecules, predominately proteins [4], adsorb to nanoparticle surfaces and form a coating called protein corona to decrease the surface free energy [26] and provide nanoparticle stability [27]. This process occurs when nanoparticles encounter biological fluids [7]. Adsorption is governed by protein composition [9] and concentration [12] along with temperature and pH [6] of biological fluids [2], [7]. Moreover, protein adsorption is specific and proteins are differently prone to adsorb depending on which chemical groups they expose [9].

Small and highly abundant proteins diffuse rapidly towards the nanoparticle while large proteins do not. Fast diffusing proteins hence form an initial protein corona. [27] This is a kinetic phenomenon called the Vroman effect which explains how the protein corona is formed [8]. However, large, high affinity proteins will eventually reach the nanoparticle and replace small, low affinity proteins [6], [27]. This explains why exposure time affects protein abundance and distribution [6]. A long exposure yields a protein corona with high affinity proteins [9]. Eventually, the composition of the NP-PC will cease to change and become permanent. An equilibrium is said to have been reached. Association rate will then equal dissociation rate for each corona component. [28]

The protein corona consists of two layers, the hard and soft corona, see figure 2. The hard corona is made up of proteins with high affinity for the nanoparticle surface. [2] These proteins have a large net binding energy of adsorption which makes them less prone to dissociate. This is why the hard corona is stable even in environments without proteins such as water [8], [29]. For soft corona proteins dissociation rates are high. These proteins do not necessarily interact with the nanoparticle surface but rather associate with hard corona proteins. [2] Interactions within the protein corona can thus be indirect (protein-protein) or direct (protein-nanoparticle) [6]. It is hard to isolate the protein corona with both layers and most experiments are therefore conducted only with the hard corona. There is consequently less knowledge regarding the soft corona. [2]

(13)

5

Figure 2.Visualization of the protein corona layers. The hard corona proteins are more tightly bound to the nanoparticle

surface than the soft corona proteins. The solid line represents the hard corona surface and the dashed line represents the soft corona surface. The figure is recreated with modifications from J. Wolfram, et al. (2014).

When the protein corona forms it gradually covers the nanoparticle surface making it impossible for the nanoparticle to interact in a direct way with its surroundings [8]. Eventually, the entire nanoparticle is concealed. The nanoparticle therefore interacts with cells and biological barriers though its protein corona. [8]. However, not all corona proteins can interact with the surrounding. Only the surface proteins are able to do this [8]. These must also be available in terms of orientation [2]. Ultimately, surface proteins will determine nanoparticle faith [9].

Adsorption to nanoparticles may cause conformational changes in proteins, thereby altering their secondary and/or tertiary structure [6]. Some proteins might even denature upon adsorption. Nanoparticles with protein coronas which include denatured proteins will with high probability be removed by cellular mechanisms. [16] In other cases proteins do not fully denature even if some residues which are normally hidden get exposed. These residues might interact with cell receptors and can thereby cause unwanted cellular responses, for instance adverse signalling [16], [30]. In addition, cooperative binding of proteins has been observed upon adsorption to nanoparticles. Deng, et al. (2012) noted this behaviour for fibrinogen. Here cooperative binding means facilitated adsorption of a second protein molecule due to conformational changes following adsorption of a primary protein to a nanoparticle. [27], [30] It is clear that nanoparticles and corona proteins have a mutual influence on each other. Nanoparticles may alter protein structure and proteins change the physicochemical properties of nanoparticles through formation of a protein corona. [30]

(14)

6

2.1.4. Nanoparticles Applied as Drug Delivery Systems in Medicine

Nanoparticles hold a great potential in medicinal applications [1] such as cancer treatment [2], [3]. Perhaps, nanoparticles are most promising as personalized drug delivery systems to treat cancer [4], since they enable a high drug load to be stored and protected [5], [6]. In addition, cancer drugs packed in nanoparticles are distributed in a lesser extent to non-tumor cells. Consequently, targeting of tumor cells is improved and accumulation in desired organs is increased. Also, damage to non-tumor cells and adverse side effects are reduced. [3] Moreover, nanoparticles as drug delivery systems provide longer circulation times [3], which further improves targeting [6].

Many nanodrugs do not achieve a high enough efficacy in clinical trials. Consequently, only a few nanodrugs are commercially used. [34] In 2016, 43 nanodrugs were approved for use, out of which 17 were used to treat cancer [4]. Onivyde® for instance is a PEG-conjugated liposome nanoparticle used since 2015 to encapsule irinotecan and treat pancreatic cancer. Another nanodrug currently used is Marqibo® which is a liposome encapsulating vincristine sulfate to treat acute lymphoblastic leukemia. [34] Besides the nanodrugs approved, many are in clinical trials [32].

An interindividual variability of body fluids among patients due to phenotype and chronical or temporary diseases cause differing NP-PCs and thus varying biological responses to nanoparticles [36]. The protein corona might influence immune response, toxicity, cellular targeting and uptake, accumulation, degradation, clearance, biodistribution and bioavailability [6], [8], [33], [34].

Interaction of nanoparticles with cell receptors can trigger cellular uptake or stimulate intracellular signalling pathways [30]. However, cellular uptake has proven to be more efficient in serum-free solutions partially because the native nanoparticle is more prone to cell membrane adhesion than the protein-coated nanoparticle. The protein corona thus has a negative impact on cellular uptake of nanoparticles [26] although it reduces nanoparticle toxicity [27]. However, it is not a pre-requisite for nanoparticles to enter a cell for drug delivery. Drug release can occur both inside and outside of cells. Mechanisms described to control release include diffusion, solvent effects, chemical reactions and physical stimuli (magnetism, light, heat and ionic strength). [35]

Apolipoproteins, immunoglobulins, serum albumin and fibrinogen are commonly found in many types of hard NP-PCs [2], [8]. Immunoglobulin, fibrinogen and complement factors cause recognition of nanoparticles by phagocytes through opsonization and thereby promote clearance due to uptake by immune cells in organs of the reticuloendothelial system [1], [6], [32]. Opsonization of nanoparticles is problematic as it prevents them from reaching their target and hence impede their interaction with cell receptors [1]. Instead of reaching their target, nanoparticles will end up in the liver or spleen where they might cause damage [31], [32]. In addition, size influences nanoparticle biodistribution. Nanoparticles with a diameter < 5 nm are excreted by the kidneys. In contrast, nanoparticles with a diameter > 100 nm are accumulated in the liver. [32] Unlike opsonins, dysopsonins such as albumin and apolipoproteins increase the chances of nanoparticles reaching their target and interacting with cell receptors [6]. Albumin for instance provides a better biocompatibility [13] and apolipoproteins increase circulation time [33]. Some proteins thus have a positive impact on nanoparticle faith while others do not.

Functionalization with molecules prior to administration could extend circulation time and thus make nanoparticles reach their target [6]. Most attempts made involve creating stealth effects [29] by reducing protein adsorption [6]. Commonly used is a hydrophilic polymer called polyethylene glycol (PEG) [6], [33], see figure 3. Adding PEG increases circulation time [6] but reduces cellular uptake [12]. An approach to improve cellular uptake is to add molecules able to interact with specific cell receptors

(15)

7

[31]. Adsorption of blood proteins might however block the intended interaction. [6] Even though PEGylated nanoparticles have some proteins loosely bound to their surfaces they lack a hard corona [11]. PEGylation hence makes nanoparticles interact with immune cells to a lesser extent [12] thereby decreasing clearance [6]. Nonetheless PEGylation might also act oppositely [1]. In addition to PEG, proteins [1] such as apolipoproteins A4 and C3 can create stealth effects [33].

Figure 3. Functionalization with molecules prior to administration could improvecirculation time. Adding PEG for instance

increases circulation time and biodistribution to desired organs. PEGylation works by reducing protein adsorption. In contrast, naked nanoparticles might get distributed to either liver or spleen. The figure is recreated with modifications from P. Aggarwal, et al. (2009).

Once administered, nanoparticles might circulate in multiple extracellular fluids such as blood and lymph before reaching their final target [9]. During the sequential incubation in body fluids the NP-PC evolves [9], [37]. Yet it does not change completely; it keeps features from the primary incubation. The evolved corona is more similar to the protein corona formed in the primary incubation than to protein coronas that would have been formed if any of the subsequent incubations would have been the primary incubation. This is why the primary incubation greatly influences biological behaviour. Bonvin, et al. (2017) proved this to be accurate at low flow. It is thus important to consider how to administer nanoparticles since it matters for biological outcome which body fluid they encounter first [12]. It seems possible to control nanoparticle faith [6] by forming a protein corona before incubating nanoparticles with cells [1]. Although the protein corona evolves, it remains quite stable through multiple incubations which is why it would be meaningful to form a designed protein corona before administration [1]. Designed NP-PCs could for instance be used to foresee biological outcome [9] or to improve targeted drug delivery, possibly by blocking unwanted interactions [12].

Due to an interindividual variability of body fluids among patients and thereby varying responses to nanoparticles [36] there is an interest in developing individualized NP-PCs. Nanoparticles would then be applied in personalized medicine which is a strategy pursuing customization of drugs based on each patient’s genome [36]. Implementing this would yield nanoparticles as personalized drug delivery systems [2], [6].

(16)

8

2.2. Methodology

Provided in this section is information regarding methods utilized in this project.

2.2.1. Mass Spectrometry

MS is an analytical technique with various applications and among these are protein identification and quantification [38]. A mass spectrometer comprises an ion source, a mass analyser, and a detector. Additional systems required for proper functioning are a pumping system used to obtain a low pressure and a system to record the signal. [39]

Soft ionization with electrospray ionization (ESI) is suitable for macro molecules such as proteins. By using ESI most susceptible bonds will be preserved and proteins will be kept relatively intact. Prior to ionization, proteins must be dissolved in a polar and volatile solvent often including formic acid (FA). FA is added to aid protonation when positive ions are being produced. Subsequently, dissolved proteins move through a narrow needle kept at a high positive or negative potential. Charged droplets are formed and ejected out of the needle. Evaporation causes a reduction of droplet size and formation of gas phase ions at atmospheric pressure. This terminates the soft ionization. It is possible to directly couple a protein separation step prior to MS analysis. When MS is coupled with a separation technique its sensitivity and selectivity is increased. In high performance liquid chromatography, a solution including proteins is passed through a column with a stationary phase to which the proteins can bind to in a varying extent. Due to various binding abilities proteins will migrate through the stationary phase with different speeds. A strong interaction causes a long retention time. High performance liquid chromatography has a stationary phase with small and uniform particle size which along with a high flow pressure decreases retention times and increases resolution. [39]

Mass analysers separate ions based on their mass-to-charge ratio (m/z). Many different mass analysers are available today. Among these are the quadrupole and the orbitrap. A quadruple consists of four parallel rods kept at direct-current and radio frequency potentials. Ions with a certain m/z value will travel with a stable trajectory through these rods and give raise to a signal. An orbitrap consists of a central electrode surrounded by outer electrodes. Between these a direct-current potential is applied. Ions oscillate back and forth around the central electrode with various oscillation frequencies which can be measured. [39]

A detector finally registers selected ions. Ions are accelerated to gain energy used to form secondary particles like electrons or photons, to be detected as a signal and sent to a computer. Through faster acceleration ions will gain more energy which will improve sensitivity. [39] Eventually a mass spectrum is produced where relative intensity is plotted against m/z values [38].

Proteins can be identified using either top-down or bottom-up strategies. With bottom-up strategies, also called shotgun proteomics, proteins must be digested to peptides prior to MS. Digestion is regularly made with trypsin as it is stable, selective and active. Trypsin cleaves C-terminally of arginine and lysine (in sequences without an adjacent C-terminal proline). In contrast, with top-down strategies, full-length proteins must be separated prior to MS. [39]

With tandem mass spectrometry (MS/MS) protein structural studies are possible. In MS/MS, selection and isolation of a peptide ion is followed by fragmentation and detection. [39] Fragmentation is usually done with collision-induced dissociation in a collision cell where ions collide with an inert gas causing peptide bonds to break. [46] Fragmented peptide ions can be classified according to where the charge is included, either in the N terminal fragment (a, b or c-ions) or the C terminal fragment (x, y or z-ions). Most commonly observed are y and b ions. [38]

(17)

9

Following acquisition of MS or MS/MS data, proteins can be identified through a correlative database search using a search algorithm [39]. SEQUEST is an algorithm which scans a peptide sequence database to find the best sequence match. First, a search of a database is performed to identify peptides with an identical mass as the precursor ion. Spectra produced by detection of b/y-ions are then compared with m/z values from a theoretical ion fragmentation of all peptides obtained from a theoretical cleavage of all proteins in a database. Subsequently, the most similar peptide is proposed as the most probable. [38]

For label free quantification of proteins spectrum counting can be applied. This approach is used to compare relative protein abundance in a sample. [40] Liu, Sadygov and Yates (2004) demonstrated a linear correlation between spectral counts and relative protein abundance. Small proteins usually have fewer spectral counts than large proteins. This is important to consider. Thus, protein length is accounted for when defining the spectral abundance factor (SAF) [42], [43]. SAF values are calculated as follow:

𝑆𝐴𝐹𝑘 = 𝑆𝑝𝑐𝑘

𝐿𝑘 (1)

Spck = spectral count for protein k (number of MS/MS spectra identifying protein k); Lk = length of

protein k (number of amino acids).

Normalization of SAF values yields the normalized spectral abundance factor (NSAF), which allows for a straightforward comparison of the relative protein abundance in a sample to be made. An NSAF value is calculated for each individual protein as described below:

(𝑁𝑆𝐴𝐹)_𝑘 = 𝑆𝑝𝑐𝑘 𝐿𝑘 ∑ 𝑆𝑝𝑐𝑖 𝐿𝑖 𝑁 𝑖=1 (2)

[44], [45]. A large NSAF value implies a highly abundant protein in a sample [43]. The denominator equals the sum of SAF for all N proteins in the experimental dataset.

2.2.2. SUrface Proteomics, Safety, Targeting, and Uptake

SUSTU is a novel proteomics method used to characterize hard NP-PC surface proteins, developed by Kuruvilla, et al. (2017). It provides information regarding nanoparticle safety, targeting, and uptake. Unlike conventional methods, SUSTU is straightforward and generates only relevant information for understanding biological interactions. SUSTU can be applied to improve prototyping of nanoparticles in terms of speed.

SUSTU relies on capture of peptides representing hard corona surface proteins by biotin labelling. This is accomplished using EZ-Link™ Sulfo-NHS-SS-Biotin, see figure 4. The N-hydroxysuccinimide (NHS) moiety provides an active ester able to react with primary amino groups of lysine or N-terminals at pH 7-9 to form amide bonds [47]. The NHS-ester becomes non-reactive in aqueous solution due to hydrolysation upon contact with water. Hydrolysis is a major competitive reaction, thus the EZ-Link™ Sulfo-NHS-SS-Biotin needs to dissolved just before use. [48] The sulfonate group increases solubility in

(18)

10

water and the disulfide (SS) bond enables release of bound peptides by addition of a reducing agent like DTT [49].

Figure 4. Chemical structure of EZ-link™ Sulfo-NHS-SS-Biotin. A sulfo-NHS group (left) provides reactivity to primary amino groups. An SS bond enables cleavage by adding a reducing agent. The figure is recreated with modifications from ‘Thermo Scientific Avidin-Biotin Technical Handbook’, (2009).

Briefly explained SUSTU works as follows. EZ-Link™ Sulfo-NHS-SS-Biotin is added to label exposed NP-PC surface proteins. Washing and centrifugation removes excess reagent. Trypsin is added to digest proteins into peptides. A long incubation with trypsin added in a 1:20 enzyme to protein ratio is followed by a short incubation with half the amount of trypsin added to complete digestion. A NeutrAvidin agarose resin enables capturing of labelled peptides as it binds to the biotin moiety, hence separating biotinylated peptides from non-biotinylated peptides. Bound peptides are subsequently eluted with guanidine-hydrochloric acid (guanidine-HCl) and biotin is cleaved off with dithiothreitol (DTT) to enable label-free quantitative proteomics analysis with LC-MS/MS. The workflow for SUSTU is shown in figure 5.

(19)

11

2.2.3. Filter Aided Sample Preparation

Filter aided sample preparation (FASP) is a general method used to digest proteins and prepare samples for MS. A centrifugal ultrafiltration device is used to digest proteins with enzymes and isolate peptides. Devices with molecular weight cut offs of 30 000 are optimal to use. [50]

FASP works well with detergents such as sodium dodecyl sulfate (SDS). It is however necessary to exchange SDS since it is easily ionized and can hide MS signals of peptides. SDS is consequently exchanged by urea in a standard filter devices used for centrifugal ultrafiltration. [51] Urea works by causing dissociation and reducing the size of SDS micelles [50]. Moreover, urea is a chaotropic chemical [51] able to cause unfolding of proteins through direct or indirect denaturing routs [52].

SDS and low molecular weight molecules are removed by ultrafiltration and centrifugation, followed by digestion of proteins with proteases, for instance trypsin. Ultrafiltration and centrifugation subsequently yields a run through with peptides. [50]

2.2.4. Bicinchoninic Acid Protein Assay

Pierce™ BCA protein assay is colorimetric and uses formation of a colour complex to determine protein concentration in a solution.

Peptide bonds, cysteine, cystine, tryptophan and tyrosine of proteins reduces copper Cu2+ _{to Cu}1+ _in

alkaline pH (the biuret reaction). Copper in its reduced form can bind two bicinchoninic acid (BCA) molecules to generate a complex with an optical absorbance at 562 nm that is linearly proportional to an increasing protein concentration. A solution containing bovine serum albumin (BSA) is used as standard. BSA is diluted to yield several known concentrations and give a linear standard curve from which the protein concentration is determined. Both standards and samples are diluted and subsequently mixed with BCA working reagent before being incubated. It is crucial that absorbance is measured immediately after incubation since colour development will proceed. However, due to a slow colour development following incubation protein concentration in multiple solutions can be determined. [53]

2.2.5. Depletion of Rat Plasma

Depletion of rat plasma can for instance be performed with Seppro® Rat Spin Columns. These columns contain selective avian antibodies able to interact with antigen. Here, antigens are the seven most abundant plasma proteins. Depletion improves MS resolution since high-abundant proteins are removed and will thus not mask detection peaks of low-abundant ones, enabling more proteins to be identified. Rat plasma is added to a spin rat column, washed, and eluted. The eluate might need to be concentrated depending on the intended application. This specific removal depletes 60–70 % of the total protein mass from rat plasma. [54]

(20)

12

3. Materials and Methods

3.1. Materials

The following materials were used in this project.

Chemicals: SDS, HCl, Tris, DTT, ammonium bicarbonate (ABC), UA, iodoacetamide (IAA), acetonitrile (ACN), FA, ethylenediaminetetraacetic acid (EDTA), 10 x phosphate-buffered saline (PBS), sodium chloride (NaCl), guanidine-HCl, and milli Q water were available, hence did not need to be purchased. Pierce™ protease and phosphatase inhibitor mini tablets were purchased from Thermo Scientific™ (# A32959) and 30 nm TiO2 nanoparticles were obtained from Nanograde LIc (Zurich, Switzerland, # 4025).

DTT, PBS, EZ-link™ sulfo-NHS-SS-biotin, and NeutrAvidin from Pierce cell surface protein isolation kit were purchased from Thermo Scientific (# 89881). BCA reagent A, BCA reagent B and BSA from Pierce™ BCA protein assay kit were purchased from Thermo Scientific (# 23227). 10 × Dilution Buffer, 10 × Stripping Buffer and 10 × Neutralization Buffer from Seppro® rat spin columns were obtained from Sigma-Aldrich (# SEP130).

Proteins: Recombinant trypsin expressed in Pichia pastoris, proteomics grade, was purchased from Roche (# 03708985001).

Biological material: Plasma isolated from normal feed Sprague-Dawley rat blood was provided by the University of the Basque Country UPV/EHU. Blood was taken from the inferior cava vein and supplemented with 1.5 mg/ml EDTA. Plasma was obtained after centrifugation and supplemented with protease and phosphatase inhibitors from Pierce.

Fibrinogen was removed from rat plasma through incubation at 550 rpm for 1 h at 37°C followed by centrifugation at 10 000 x g for 10 minutes at 4°C to avoid clotting of fibrinogen proteins. The supernatant depleted of fibrinogen was collected. This is the non-depleted plasma.

Filters: Amicon® ultra 3KDa device was purchased from Merck Millipore (# UFC500308), Seppro® rat spin columns were purchased from Sigma-Aldrich (# SEP130), Microcon® 30K ultrafilter device was purchased from Merck Millipore (# MRCF0R030), C18 Top Tips were purchased from Glygen Columbia, Maryland, USA (# TT2C18.96) and spin columns with bottom caps and collection tubes from Pierce cell surface protein isolation kit were purchased from Thermo Scientific (# 89881).

3.2. Methods

3.2.1. Depletion of Plasma

Non-depleted plasma was depleted of the seven most abundant proteins: albumin, immunoglobulin G, α1-antitrypsin, immunoglobulin M, transferrin, haptoglobin and fibrinogen to yield a plasma

enriched in low abundant proteins called depleted plasma. Depletion was made with Seppro® rat spin columns.

Non-depleted plasma was diluted in 1 x dilution buffer from Seppro® rat spin columns to a final volume of 500 µl. Diluted plasma was added to a 0.45 µm spin filter and centrifuged at 10 000 x g for 1 minute at room temperature (RT) to remove viscous substances which arise from lipids. A rat spin column without its associated end cap attached was put in a collection tube and centrifuged at 2 000 rpm for 30 seconds at RT to remove storage solution. To remove non-covalently bound antibodies from column beads two additional centrifugations were made with 1 x dilution buffer. Diluted, non-depleted plasma (500 µl) was added to the column with the end cap attached. Complete mixture of column beads and non-depleted plasma was obtained by shaking the column. The column was incubated rotating

(21)

13

clockwise for 15 minutes at RT to allow association of immobilized antibodies and high abundant plasma proteins. To collect the depleted plasma enriched in low abundant proteins the column was placed in a new collection tube without the end cap attached and centrifuged at 2 000 rpm for 30 seconds at RT. Elution of bound proteins and regeneration of the column was then made according to the Seppro® rat spin column protocol from Sigma-Aldrich.

3.2.2. Concentration of Depleted Plasma

Amicon® ultra 3KDa device was put in a filter tube. Depleted plasma was added to the device. The device was centrifuged at 14 000 x g for 15 minutes at RT to collect molecules above the molecular weight cut off. To recover the concentrated plasma the device was put upside down in a collection tube and centrifuged at 1 000 x g for 2 minutes at RT.

3.2.3. BCA Protein Assay

Protein concentration in plasma was determined with Pierce™ BCA Protein Assay Kit. BCA working reagent was prepared according to instructions given in the protocol provided from the kit. Milli Q water, 2 mg/ml BSA, plasma and BCA working reagent was added in two replicates to a 96-well plate according to the scheme bellow, see table 1.

Table 1. Amount of milli Q water, 2 mg/ml BSA, plasma and BCA working reagent in µl, added to 9 separate wells in a 96-well plate. To make sure protein concentration was within range for the standard curve, X was varied between 2-10 µl depending on plasma sample.

Well Milli Q water (µl) 2 mg/ml BSA (µl) Plasma (µl) BCA WR (µl) Blank 25 0 200 2 24 1 200 3 23 2 200 4 20 5 200 5 17.5 7.5 200 6 15 10 200 7 13 12 200 8 10 15 200 9 25 - X X 200

The plate was incubated for 30 minutes at 37°C in an incubator shaker (New Brunswick Scientific, Innova 42). Before measuring absorbance in a microplate reader (BioTek, Synergy 2) the plate was cooled to RT. The absorbance was measured at 562 nm. The corrected average absorbance of BSA standards was calculated by subtracting the average absorbance of the blank replicates from the average absorbance of the BSA standard replicates. A standard curve was made to determine protein concentration in plasma by plotting the corrected average absorbance of BSA standards against the mass of BSA standards in µg.

3.2.4. Formation of NP-PC

Below follows the protocol describing how to form NP-PCs of the depleted and non-depleted plasma. Three biological replicates of each model plasma were used to form six different NP-PCs.

To create a 1 mg/ml TiO2 nanoparticle stock solution, 1 mM EDTA and 1 protease and phosphatase

inhibitor mini tablet from Pierce per 10 ml of solution, supplemented with 1 x PBS was added to 30 nm TiO2 nanoparticles. This was sonicated in an ultrasonic cleaner (Branson, 1510) for 30 minutes to

(22)

14

solubilize nanoparticles and for total mixture. To keep nanoparticles from sedimenting the solution was manually stirred every 10 minutes.

Plasma (non-depleted or depleted) was diluted in 1 x PBS to a concentration of 0.6 mg/ml and separated into three fractions, each in a 15 ml falcon tube: one fraction to measure protein concentration in the hard corona, one for characterization of hard corona proteins and one for characterization of surface hard corona proteins. The 1 mg/ml TiO2 nanoparticle stock solution was

added in 100 µl to each fraction of plasma yielding a final nanoparticle concentration of 0.1 mg/ml and plasma protein concentration of 0.5 mg/ml. All fractions were incubated (Eppendorf, Thermomixer comfort) in the dark at 550 rpm for 1 h at 37˚C to form NP-PCs. Following incubation all fractions were centrifuged at 2 700 x g for 5 minutes at 4˚C to pellet nanoparticles with adsorbed plasma proteins. The supernatants were discarded and 0.5 ml 1 x PBS, 1 mM EDTA and Pierce™ protease and phosphatase inhibitor mini tablet solution was added to wash the pellets. Samples were transferred to low binding tubes and centrifuged at 10 000 x g for 5 minutes at 4˚C to remove unbound proteins including soft corona proteins. The supernatants were discarded, and the pellets were washed with 0.25 ml 1 x PBS, 1 mM EDTA and protease and phosphatase inhibitor mini tablet solution and centrifuged at 10 000 x g for 5 minutes at 4˚C. This was repeated once. The supernatants were discarded. The pellets were frozen at -80˚C and stored for quantification of hard corona proteins, FASP digestion and SUSTU.

3.2.5. Formation of Evolved NP-PC

Below follows the protocol describing how to form the evolved NP-PC. Three biological replicates of each model plasma were used to form three evolved NP-PCs. One biological replicate of the depleted plasma and one replicate of the non-depleted plasma was required to form one evolved NP-PC. Formation of NP-PC was performed with the depleted plasma as described in section 4.2.4. However, the pellets were not frozen. Instead, the pellets were re-suspended in 100 µl of 1 x PBS, 1 mM EDTA and protease and phosphatase inhibitor mini tablet solution.

Non-depleted plasma was diluted in 1 x PBS to a concentration of 0.6 mg/ml and separated into three fractions, each in a 15 ml falcon tube. The re-suspended pellet was added in 100 µl to each fraction of plasma yielding a final nanoparticle concentration of 0.1 mg/ml and plasma concentration 0.5 mg/ml. All fractions were incubated at 550 rpm for 1 h at 37°C to form the evolved NP-PCs. Following incubation all fractions were centrifuged at 2 700 x g for 5 minutes at 4°C to pellet nanoparticles with adsorbed plasma proteins. The supernatants were discarded and 0.5 ml of 1 x PBS, 1 mM EDTA and protease and phosphatase inhibitor mini tablet solution was added to wash the pellets. Samples were transferred to low binding tubes and centrifuged at 10 000 x g for 5 minutes at 4°C to remove unbound proteins. The supernatants were discarded, and the pellets were washed with 0.25 ml of 1 x PBS, 1 mM EDTA and protease and phosphatase inhibitor mini tablet solution and centrifuged for at 10 000 x g for 5 minutes at 4°C. This was repeated once. The supernatants were discarded. The pellets were frozen at -80°C and stored for quantification of hard corona proteins, FASP digestion and SUSTU.

3.2.6. Quantification of Hard NP-PC

To solubilize bound proteins, 20 µl 2 % SDS was added to the NP-PC pellet. The sample was centrifuged at 14 000 x g for 10 minutes at 20°C to obtain a supernatant with unbound corona proteins. A BCA protein assay was performed with the supernatant to determine the protein concentration of the hard corona.

(23)

15

3.2.7. FASP Digestion

A Microcon® 30K ultrafilter device was conditioned with 200 µl of milli Q water and centrifuged at 14 000 x g for 15 minutes at 20°C. This was repeated once. The flow through was discarded. To further condition the device, 200 µl of 8 M UA in 0.1 M Tris-HCl pH 8.5 referred to as UA buffer was added. The device was centrifuged at 14 000 x g for 15 minutes at 20°C. This was repeated once. The flow through was discarded.

Thawed NP-PC pellet (non-depleted, depleted or evolved) was resuspended in 10 µl 2 % SDS, 100 mM Tris-HCl at pH 7.6 and 100 mM DTT to yield a 1:10 pellet to solution ratio to release bound corona proteins. This was incubated for 5 minutes at 95°C to denature proteins, then chilled to RT. The sample was centrifuged at 16 000 x g for 10 minutes at 20°C to obtain a supernatant with unbound corona proteins. A mix of 30 µl supernatant and 200 µl UA buffer was added to the device and centrifuged at 14 000 x g for 15 minutes at 20°C. To exchange SDS and denature proteins, 200 µl UA buffer was added to the device and centrifuged at 14 000 x g for 15 minutes at 20°C. The flow through was discarded and 100 µl of 0.05 M IAA in UA buffer was added to the device. This was incubated (Eppendorf, Thermomixer comfort) in the dark at 600 rpm for 1 minute at 20°C to prevent re-formation of broken disulphide bonds. Further incubation was made for 20 minutes at 20°C without mixing. The device was centrifuged at 14 000 x g for 10 minutes at 20°C. UA buffer of 100 µl was added and the device was centrifuged at 14 000 x g for 15 minutes at 20˚C. This was repeated twice. To condition the device, 100 µl 50 mM ABC pH 8.0 was added. The device was centrifuged at 14 000 x g for 10 minutes at 20°C. This was repeated twice. ABC buffer and trypsin were added. The device was incubated (Eppendorf, Thermomixer comfort) at 600 rpm for 1 minute at 20°C, followed by incubating in a wet chamber in an incubator shaker (New Brunswick Scientific, Innova 42) for 16 hours at 37°C to digest proteins. Trypsin was added in a 1:100 trypsin to protein ratio.

The device was centrifuged at 14 000 x g for 10 minutes at 20°C to elute peptides. To remove unwanted resin particles, increase the solubility of proteins, and improve elution yield in the elution column 50 µl of 0.5M NaCl was added and the device was centrifuged at 14 000 x g for 10 minutes at 20°C. Peptide eluate was acidified by 10 % FA. pH was measured with a pH strip and adjusted to a value between 2-3.

The acidified peptides were desalted and dried, then stored at -80°C. See description of execution under section 4.2.9. Desalting.

3.2.8. SUSTU: Biotinylation and Digestion and Purification of Biotinylated Proteins

EZ-link™ Sulfo-NHS-SS-Biotin provided from Pierce cell surface protein isolation kit was dissolved in milli Q water just before use to create a 10 mM biotin stock solution. PBS at pH 7.2 from Pierce cell surface protein isolation kit and 10 mM biotin stock solution were mixed to create a 1 mM biotin solution. To resuspend the NP-PC pellet (non-depleted, depleted or evolved) 50 µl of 1 mM biotin solution was added. The sample was incubated (Eppendorf, Thermomixer comfort) in the dark at 1 400 rpm for 30 minutes at 23°C to allow for biotin to bind to accesseble surface corona proteins. For evolved corona samples 0.25 ml of 50 mM ABC was added while 0.5 ml 50 mM ABC was added for other samples. The sample was centrifuged at 10 000 x g for 5 minutes at 4°C to remove excess of biotin. The supernatant was discarded. This was repeated twice. The NP-PC pellet was resuspended in 50 µl of 50 mM ABC. Trypsin was added in an approximate 1:1 trypsin to protein ratio (25 µl) and the sample was incubated in an incubator shaker (New Brunswick Scientific, Innova 42) in a clockwise rotator over night (16 h) at 37°C to digest NP-PC proteins.

(24)

16

Trypsin in half amount (12.5 µl) was added to further digest NP-PC proteins. This was incubated for 3 h at 40°C to complete digestion. The sample was centrifuged at 16 000 x g for 15 minutes at 18°C. The supernatant with biotinylated peptides was recovered. To resuspend the nanoparticle pellet 50 µl 50 mM ABC buffer was added and the sample was centrifuged at 16 000 x g for 15 minutes at 18°C. The supernatant was recovered and pooled with the first one. To again resuspend the nanoparticle pellet 50 µl 50mM ABC was added and the sample was centrifuged at 16 000 x g for 10 minutes at 18°C. The supernatant was recovered and pooled with the others.

The peptide concentration of pooled supernatants was measured on a spectrophotometer (Thermo Scientific, NanoDrop 2000 UV-vis) at 280 nm to determine whether peptides were present or not. ABC was used as a blank. The peptide concentration was determined and used to derive the peptide mass. The peptide mass was then compared to a reference value [37].

NeutrAvidin agarose provided from Pierce cell surface protein isolation kit was packed into a spin column placed in a collection tube, followed by centrifugation at 500 x g for 1 minute at RT to remove storage solution. PBS at pH 7.2 was added and the spin column was centrifuged at 500 x g for 1 minute at RT to wash it. The flow through was discarded. This was repeated twice. The spin column was capped and placed in a new collection tube. Biotinylated peptides were added and let to settle in the NeutrAvidin agarose for 30 minutes at RT while rotating clockwise. The cap was removed and PBS at pH 7.2 was added. The spin column was centrifuged at 500 x g for 1 minute at RT and the flow through was discarded. This was repeated four times. The spin column was transferred to a new collection tube for elution. To elute peptides, 50 µl of 8 M guanidine-HCl pH 1.5 and 50 mM DTT was added and the spin column was centrifuged at 500 x g for 1 minute at RT. This was repeated three times.

The acidified peptides were desalted and dried, then stored at -80°C. See description of execution under section 4.2.9. Desalting.

3.2.9. Desalting

C18 Top Tips from Glygen were used to remove salts and buffers to decease the chemical noise in the mass spectrum. Tips were put in spin adapters placed in Eppendorf tubes and conditioned with 60 % ACN and 0.1 % FA referred to as releasing solution and 0.1 % FA referred to as binding solution. Releasing solution was added and tips were centrifuged at 3 000 rpm for 3 minutes. This was repeated twice. Binding solution was added, and tips were centrifuged at 3 000 rpm for 3 minutes. This was repeated twice. The flow through was discarded when needed. Acidic (pH 2-3) peptide sample of 50 µl was added and tips were centrifuged at 3 000 rpm for 3 minutes. This was repeated until all sample had been added to the tips. Binding solution was added, and the tips were centrifuged at 3 000 rpm for 3 minutes to wash peptides. This was repeated twice. Tips and spin adapters were moved to new Eppendorf tubes for elution. Releasing solution was added and tips were centrifuged at 3 000 rpm for 3 minutes to elute hydrophobic peptides bound to the reverse phase C18 matrix. This was repeated three times.

The eluate was incubated in a speedvacuum concentrator (Thermo Savant, SPD 1010) for 1 hour at 45°C to dry and concentrate peptides. Dried peptides were storded at -20°C.

3.2.10. Preparation for MS Measurements

To reconstitute dried peptides 0.1 % FA was added. This was sonicated in an ultrasonic cleaner (Branson, model 1510) for 15 minutes to solubilize peptides. Peptide concentration was measured on a spectrophotometer (Thermo Scientific, NanoDrop 2000 UV-Vis) at 280 nm. The concentration was adjusted to 0.06 ± 0.02 mg/ml by adding 0.1 % FA. Peptide concentration was measured again to

(25)

17

confirm that the concentration was within the desired range. Peptide sample of 10 µl was added to an MS vial. This was sealed with a cap and sent for MS measurements.

3.2.11. MS Measurement

Peptides were analysed in a QExactiv quadrupole-orbitrap mass spectrometer (Thermo Scientific) coupled to a liquid chromatography (LC) system, see table 2.

Table 2. Unbiased label-free LC-MS/MS was performed following SUSTU and FASP digestion of NP-PC proteins. This was made with all samples. Shotgun proteomics was performed with proteins from the depleted and non-depleted plasma. Proteins

were digested with trypsin prior to shotgun proteomics.

Strategy Model plasma Biological replicate

SUSTU, FASP, shotgun proteomics Depleted 2

SUSTU, FASP, shotgun proteomics Non-depleted 1.1

SUSTU, FASP Depleted, non-depleted (evolved) 2, 1.1

Samples were separated in an EASY nLC II system (Thermo Scientific). Peptides were injected into a pre-column (NanoSeparations, the Netherlands; NS-MP-10 BioSphere C18, 5 µm particle size, 120 Å pore size, 100 mm inner diameter x 20 mm length) and subsequently separated on an analytical column (NanoSeparations, the Netherlands; NS-AC-10 BioSphere C18, 5 mm particle size, 120 Å pore size, 75 µm x 100 mm) before being ionised by ESI. A linear gradient of 6 to 40 % buffer B (0.1 % FA in ACN) against buffer A (0.1 % FA in water) was carried out with a constant flow rate of 300 nL/min for 100 min. Full scan MS spectra were recorded in positive mode ESI with an ion spray voltage power frequency of 2.4 kilovolt, a radio frequency lens voltage of 69 V and a capillary temperature of 235°C, at a resolution of 30 000 to scan all precursor ions. The top 15 most intense ions were selected for MS/MS under an isolation width of 1.2 m/z units. Collision induced dissociation with a collision energy of 27 % was performed to fragmentize ions. MS/MS spectra were recorded.

3.2.12. Data Analysis

Protein identification was made with Proteome Discover software (Thermo Fischer Scientific, version 1.4) and protein quantification was made with Scaffold software (Proteome Software, version 4.11). Proteins were identified from mass spectra by a comparative search made in Proteome Discover. A workflow was defined, see figure 6. MS/MS spectra files and the Rattus norvegicus database from UniProt (29 951 entries) were uploaded. Parameter settings were selected. A Δ Cn value, calculated as follows:

∆ 𝐶𝑛(𝑟𝑎𝑛𝑘 𝑖) =𝑠𝑐𝑜𝑟𝑒 𝑃𝑆𝑀𝑟𝑎𝑛𝑘 1− 𝑠𝑐𝑜𝑟𝑒 𝑃𝑆𝑀𝑟𝑎𝑛𝑘 𝑖 𝑠𝑐𝑜𝑟𝑒 𝑃𝑆𝑀𝑟𝑎𝑛𝑘 1

(3)

was selected to distinguish top-scoring peptide spectrum matches (PSM). A maximum Δ Cn value was selected to 0.05. Peptides with a Δ Cn less than or equal to this were used for the percolator. A maximum of 2 tryptic cleavages were allowed. The precursor and fragment mass tolerance were 10 ppm and 0.02 Da, respectively. Peptides with a false discovery rate (FDR) less than 0.01 were used. The minimum peptide length considered was 6. Parameters for dynamic modifications for biotinylated

(26)

18

samples were selected as follows: acetyl / +42.011 Da (any N-terminal); oxidation / +15.995 Da (methionine). No static modifications were selected for biotinylated samples. The Rattus norvegicus sequence database including common contaminants was compared to mass spectral data using Sequest HT. Common contaminants like human keratin and trypsin were included to minimize false identifications. Search results were sorted and filtered.

Figure 6. Workflow for Proteome Discover. First, spectrum files were selected. Then, a correlative search was made against a database. Data was filtered in the percolator where FDR values were calculated. The event detector created extracted ion chromatograms and removed peaks below a threshold signal-to-noise value from the spectrum. Precursor ions area detector was used for peak area calculation quantification.

A second identification was performed in Scaffold with X! Tandem to assign peptide probabilities and validate identified peptides. A relative label-free quantification strategy was simultaneously applied to compare the protein abundance in the samples. The Rattus norvegicus database from UniProt and Magellan storage files were loaded to Scaffold for spectrum counting using NSAF to quantify identified proteins. Only proteins satisfying filter and threshold settings for the experiment were displayed and considered. No proteins with a calculated probability below 99 % were considered identified. A minimum number of 2 unique peptides per protein was selected. Peptide FDR was selected to 1.0 % to assure certainty of a peptide identification. A quantification was subsequently made.

(27)

19

4. Results

4.1. Process Results

First, the hard corona and the surface of the hard corona made of depleted plasma were characterized. To ensure that all methods were functional a primary characterization was made with NP-PC proteins from depleted plasma sample 6. The results deviated from expected ones regarding the number of proteins in the surface of the hard corona, see table 3. Consequently, it was concluded that SUSTU was not working properly. After some investigation, it was found that the EZ-link™ Sulfo-NHS-SS-Biotin used to biotinylate depleted plasma sample 6 is moisture-sensitive. It becomes non-reactive due to hydrolysation as it encounters water and must therefore be dissolved just before use. This was not in accordance with how the solution had been handled.

Table 3. Number of proteins identified in the hard corona and the surface of the hard corona after data analysis in Proteome Discover. No results were obtained with non-depleted plasma 5.1.

Sample Number of proteins in

hard corona (FASP)

Number of proteins in surface of corona (SUSTU)

Depleted plasma 6 275 30

Non-depleted plasma 5.1 -

-Non-depleted plasma 1.1 305 59

To examine the functionality of the biotinylation a trial was set up. The idea was to measure protein concentration in eluate from a spin column where plasma biotinylated with a freshly made EZ-link™ Sulfo-NHS-SS-Biotin solution had been added and compare this to a reference (eluate from a spin column where non-biotinylated plasma had been passed through). However, the trial was time consuming and lacked useful results since protein concentration could not be measured due to interference, and so it was decided to cancel the trial and proceed not to cause any major deviations from the time plan.

A second attempt to characterize the NP-PCs was done with NP-PCs made of non-depleted plasma sample 5.1 and freshly made solutions, including the EZ-link™ Sulfo-NHS-SS-Biotin solution. Yet again deviating results were obtained. This time no identification was made due to noisy chromatograms, hence no results are displayed for the non-depleted plasma sample 5.1 in table 3. SUSTU has however been successful in other projects [55], therefore it was decided to make a third attempt using experimental parameters proven to work in a previous project. No adaptations of volumes or concentrations were made to actual protein concentrations. This time the results were as expected, see non-depleted plasma sample 1.1 in table 3.

Based on these attempts the protocol was improved. Changes from the initial protocol were as follows: • Trypsin was added in an approximate 1:1 enzyme to protein ratio, not in a 1:20 enzyme to

protein ratio.

• Biotinylated peptides were let to settle in the NeutrAvidin agarose while in movement. • The amount of NeutrAvidin agarose added to the spin column was not based on protein

concentration but had a fixed value.

• ABC volumes used to wash pellets and recover biotinylated peptides for the evolved NP-PC were reduced from 0.5 to 0.25 ml based on a smaller amount of proteins compared to the original protocol used by Kuruvilla, et al. (2017).

(28)

20

The next step was to develop a protocol for the evolved corona. In a previous project with objectives to simplify nanoparticle uptake route, nanoparticles were incubated with cell media and bovine foetal serum and transferred to a solution containing cellular lysate from in vitro cell cultures. An evolved corona has thus been made before, although never with the model plasma used here. Development of the protocol for the evolved corona was identical to the protocol from a previous but unpublished project.

4.2. Main Results

A primary search to identify NP-PC surface proteins performed in Proteome Discover yielded three lists of identified proteins, see table 4. Contaminants (keratin and trypsin) were removed from each list.

Table 4. Number of surface proteins identified in Proteome Discover. Also, number of identified surface proteins after removal of contaminants. An isoelectric point over 4.69 was observed for all proteins in each surface.

There were 128 proteins identified in the surface of the NP-PC made of depleted plasma and 55 proteins identified in the surface of the NP-PC made of non-depleted plasma. A total of 137 proteins were identified, of which 46 were present in both NP-PCs as can be seen in figure 7. Thus, 36 % of the surface proteins in the depleted NP-PC were present in the non-depleted NP-PC surface and 84 % of the surface proteins in the non-depleted NP-PC were present in the depleted NP-PC surface.

Figure 7. Venn diagram displaying the relationship between surface proteins from two samples. A total of 137 proteins were identified in both surfaces. There were 128 proteins identified in the surface of the NP-PC made of depleted plasma and 55 proteins identified in the surface of the NP-PC made of non-depleted plasma.

NP-PC Number of

identified proteins

Number of identified proteins after removal of contaminants

Range for isoelectric point

Depleted 2 132 128 4.69–9.88

Non-depleted 1.1 59 55 4.96–9.52

Proteomics Study of a Designed Nanoparticle-Protein Corona Made of Animal Model Plasma