Preparation and Evaluation of Immunoglobulin Free Sera for Biomaterial-Induced Complement Activation Studies

School of Natural Sciences

Degree project work

Nadia Vickius

Subject: Biomedical Science Level: Advanced level

Preparation and Evaluation of Immunoglobulin Free Sera for

Biomaterial-Induced Complement Activation Studies

Nadia Vickius

Degree Project Work, Biomedicine 30 ECTS Master of Science

Supervisor:

Prof. Kristina Nilsson Ekdahl School of Natural Sciences Linnaeus University

SE-391 82 KALMAR SWEDEN

Examiner:

Prof. Bengt Persson School of Natural Sciences Linnaeus University

SE-391 82 KALMAR SWEDEN

The Degree Project Work is included in the Study programme Biomedical Chemistry 240 ECTS

Abstract

POPULÄRVETENSKAPLIG SAMMANFATTNING

Biomaterial är en viss typ av material som är speciellt ämnade att komma i kontakt med blod eller vävnader inom kroppen. Denna kontakt kan ske i samband med behandling av olika sjukdomstillstånd, då biomaterial i form av slangar, katetrar, dialysmembran, proteser osv. används i försök att mildra symptom eller bota sjukdomen i fråga. Det första som sker när ett biomaterial kommer i kontakt med blod är att det täcks med kroppens egna blodproteiner. Sammansättningen på proteinlagret avgörs av biomaterialets egenskaper och lagret har avgörande betydelse för hur biomaterialet uppfattas av kroppen. Trots att biomaterial är avsedda att hjälpa patienter kan kroppen reagera mot det kroppsfrämmande materialet och ge upphov till inflammationsreaktioner, vilka i värsta fall kan leda till svåra biverkningar. Därför är det ytterst viktigt att det finns kunskap om hur kroppen uppfattar och reagerar på olika biomaterial samt hur biomaterial kan förändras för att ge upphov till så få biverkningar som möjligt.

Komplementsystemet utgör en del av det medfödda immunförsvaret och består av flera komponenter som cirkulerar i blodet. Huvudfunktionen är att skydda kroppen från allt kroppsfrämmande, exempelvis bakterier, virus och material, och förändrade egna celler genom neutralisering samt upphovet av ett inflammatoriskt svar. Ofrivillig aktivering av komplementsystemet kan ske när ett biomaterial kommer i kontakt med blod och detta är ett signifikant problem som måste bemästras genom att producera material som är mer förenliga med kroppen. Genom att förändra komponentsammansättningen hos ett biomaterial kan kroppsförenligheten öka och det är ytterst viktigt att nya biomaterial testas för hur de aktiverar komplementsystemet.

TABLE OF CONTENTS

ABBREVIATIONS

INTRODUCTION

The complement system ____________________________________________________ 10

The classical pathway ___________________________________________________ 11 The lectin pathway______________________________________________________ 12 The alternative pathway _________________________________________________ 12 The terminal pathway ___________________________________________________ 13 Biomaterial-induced complement activation__________________________________ 13 Complement component C1q______________________________________________ 16 The bovine complement system ____________________________________________ 17

Immunoglobulins__________________________________________________________ 17

General description_____________________________________________________ 17 Immunoglobulins and complement activation_________________________________ 19 Protein G _____________________________________________________________ 19

Blood ____________________________________________________________________ 19

AIM

MATERIALS AND METHODS

Detection of bovine Igs __________________________________________________ 21 Biotinylation of anti-human-C3c___________________________________________ 22 Detection of complement activation by measuring C3c _________________________ 22 Biomaterial-induced complement activation study _____________________________ 23 Detection of complement activation by measuring C3a with sandwich-ELISA _______ 23 Detection of complement activation by measuring sC5b-C9 with sandwich-ELISA ___ 24

Human serum ____________________________________________________________ 24

IgG-depletion using a HiTrapTM Protein G column ____________________________ 24 Detection of IgG _______________________________________________________ 25 IgM-depletion using a HiTrapTM IgM column ________________________________ 25 Coupling of anti-human-IgM to CNBr-activated Sepharose™ 4B _________________ 26 IgM-depletion using anti-human-IgM coupled to CNBr-activated Sepharose™ 4B ___ 26 Biomaterial-induced complement activation study _____________________________ 27 Detection of complement activation by measuring C3a with sandwich-ELISA _______ 28 Statistics _____________________________________________________________ 28

RESULTS

Detection of bovine Igs __________________________________________________ 29 Detection of complement activation by measuring activation markers C3c, C3a and C5b-C9 __________________________________________________________________ 29

Human serum ____________________________________________________________ 30

IgM-depletion using a HiTrapTM IgM column and anti-human-IgM coupled to CNBr-activated Sepharose™ 4B ________________________________________________ 31 Biomaterial-induced complement activation study _____________________________ 31

DISCUSSION

ACKNOWLEDGEMENT

ABBREVIATIONS

APW Alternative pathway of complement activation

CDRs Complementarity-determining regions

CPW Classical pathway of complement activation

CRP C-reactive protein

DAB 3,3-diaminobenzidine

DAP N,N-diacryloylpiperazine

DMSO Dimethyl sulfoxide

EDTA Ethylenediaminetetraacetic acid

EGDMA Ethylene glycol dimethacrylate

ELISA Enzyme-linked immunosorbent assay

FBS Fetal bovine serum

Fc region Fragment crystallizable region

FITC Fluorescein isothiocyanate

HEMA 2-hydroxyethyl methacrylate

HIV Human immunodeficiency virus

HRP Horseradish peroxidase IgA Immunoglobulin A IgD Immunoglobulin D IgE Immunoglobulin E IgG Immunoglobulin G IgM Immunoglobulin M Igs Immunoglobulins LPS Lipopolysaccharides

LPW Lectin pathway of complement activation

MAA Methacrylic acid

MAC Membrane attack complex

MASP1-3 MBL-associated serine protease 1-3

MASPs MBL-associated serine proteases

MBL Mannose-binding lectin

MS Multiple sclerosis

NHS Normal human serum

OPD O-phenylendiamine dihydrochloride

RA Rheumatoid arthritis

SDS Sodium dodecyl sulphate

SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis

SIRS Systemic inflammatory reaction syndrome

sTCC Soluble terminal complement complex

TCC Terminal complement complex

TMB 3,3,5,5-tetramethylbenzidine

”The truth is, the science of Nature has been already too long made only a work of the brain and the fancy: It is now high time that it should return to the plainness and soundness of observations on material and obvious things [1]”

Robert Hooke (1665)

INTRODUCTION

Biomaterials

By one definition, a biomaterial is “a material that is intended to come in contact with blood or other tissues within the body” [2]. The increasing need for biomaterials in artificial aids and substitutes in modern medical therapeutics has provided a broad array of implantable and extracorporeal blood contact devices [1]. Commonly applied devices are stents, artificial organs, biosensors, catheters, heart valves [3], hemodialysers, prostheses, vascular grafts, oxygenators and miniature pumps [4], and it can be mentioned that more than 25 millions of patients in USA have some kind of artificial implant. Many more will however come in contact with extracorporeal and temporal biomaterial devices in their medicalization [2]. Materials commonly used as biomaterials in medical implants and extracorporeal devices are metals, ceramics, synthetic polymers [5], composites and glass [2].

As the need for and usage of biomaterials in medicine constantly increase, so do the requirements for increased biocompatibility and hemocompatibility (i.e. blood compatibility) [4]. The definition of biocompatibility are “the acceptance (or rejection) of an artificial material by the surrounding tissues and by the body as a whole” [6] and this implies that the biomaterial must not give rise to incompatibility reactions in the neighbouring surroundings nor systemic reactions which affect the entire host body. A biomaterial is thus required to function together with the local microenvironment supporting cell-biomaterial interactions [3], be non-inflammatic, be non-toxic as well as showing satisfactory tissue compatibility and hemocompatibility [6].

Blood-biomaterial interactions

proteins have shown to influence the immune system differently and it is suggested that the adsorbed protein layer is of major importance for the biocompatibility of biomaterials. Thus controlling the protein adsorption pattern by changing the composition of the biomaterial is one approach to enhance the biocompatibility and obtain a more host functional material [6]. In this degree project work, polymers of different monomers and crosslinkers were utilized to evaluate their complement activating properties (see Table I).

Polymers utilized in the degree project work

Earlier work performed within the research group has led to the characterization of different novel polymers regarding their hemocompatibility. Among other things, complement activation potency, protein adsorption patterns and quantities of the adsorbed proteins have been evaluated. Two polymers (P1 and P2, see Figure 1) required further evaluation regarding their ability to induce antibody-independent complement classical pathway activation, as it has been shown that these polymers adsorb high amounts of complement component C1q when incubated in plasma (unpublished data). C1q was enriched on the surfaces of both polymers, but especially and to a higher extent on the surface of polymer P1. The polymers also showed low (P2) to intermediate (P1) adsorption of IgG.

Both polymers are in the size range of 25-63 µm, but they differ in composition and predicted properties (see Table I). Polymer P1 is predicted to be hydrophilic and negatively charged, while polymer P2 is predicted to be hydrophobic [2].

Figure 1. Fluorescein isothiocyanate (FITC) labeled polymer P1 (A) and polymer P2 (B) visualized by

fluorescence microscope in 20x magnification. (Image source: Kindly provided by Anna E. Engberg and Per Nilsson)

20x 20x

Table I. Composition and earlier results regarding the properties of utilized polymers P1 and P2 [2].

Polymer Monomer* Crosslinker* Functional Complement Adsorbed C1q*** Adsorbed IgG***

groups activation**

P1 MAA DAP -COOH +++ +++ +++

(JR0021)

P2 HEMA EGDMA -OH ++ +++ +

(JR0062)

Methacrylic acid (MAA); N,N-diacryloylpiperazine (DAP); 2-hydroxyethyl methacrylate (HEMA); Ethylene glycol dimethacrylate (EGDMA).

*Crosslinker to monomer ratio 80:20.

** Activation levels indicated as follows: +++ > 150% increase in C3a generation compared to control; ++ 50-150% increase in C3a generation compared to control.

***Adsorbation levels indicated as follows: +++ > 5-fold increased absorbance compared to polystyrene; ++ 2-5-fold increased absorbance compared to polystyrene; + no increased absorbance compared to polystyrene.

The complement system

The immune system is responsible for host protection by identifying and neutralizing material perceived as being foreign or non-self, for example invading pathogenic microorganisms or altered host cells. The immune system can be divided into the innate immune system and the acquired immune system, where the innate immune system plays a crucial role as first-line defence against invading pathogens since prior exposure to the foreign material is not required. The complement system is mainly considered to be a part of the innate immune system [7] but also functions as an effector of aquired immunity [8].

The main function of the complement system is to protect the host from invading pathogens (e.g. bacteria, viruses, fungi and parasites) [11] and to neutralize apoptotic and necrotic host cells. Neutralization of pathogens and altered host cells is achieved by opsonization with subsequent phagocytosis, lysis of foreign cells by the integration of the membrane attack complex (MAC) and generation of an inflammatory response after prosperous activation of one or several of the three complement pathways - the classical pathway, the lectin pathway and the alternative pathway (see Figure 2) [12].

Soluble and membrane-bound complement regulators cooperate to protect the host from damage at multiple levels in the complement cascade. Uncontrolled complement activation beyond normal limits can cause extensive damage to host cells and contribute to several inflammatory diseases and disorders such as rheumatoid arthritis (RA), multiple sclerosis (MS), Alzheimer’s disease, hyperacute graft rejection and systemic inflammatory reaction syndrome (SIRS) [13]. The tissue damaging effects of uncontrolled complement activation arise from the mediated immunological response induced by the anaphylatoxins C3a, C4a and C5a as well as host cell lysis by the membrane attack complex [4].

The classical pathway

C5 convertase (C4bC2aC3b) leads to the enzymatic cleavage of complement component C5 and generation of the very potent anaphylatoxin C5a together with the first component of terminal pathway, C5b. All generated anaphylatoxins (i.e. C3a, C4a and C5a) can induce an inflammatory response and mediate chemotaxis, vasodilatation, cell activation and cell adhesion through binding to specific receptors on neutrophils, monocytes, macrophages, mast cells and smooth muscles cells [15].

The lectin pathway

The lectin pathway is identical to the classical pathway with the exception of the initial activation step. Through the binding of C1q-like mannose-binding lectin (MBL) or ficolins to specific carbohydrates, such as N-acetylglucosamine [9] or lipopolysaccharides (LPS) [12] present on the surface of pathogens, MBL-associated serine proteases called MASPs becomes self-activated in conformity with C1r of the classical pathway. There are three kinds of MASPs associated with MBL and ficolins - MASP1, MASP2 and MASP3. Out of these only MASP2 is known to possess the ability to activate the lectin pathway through cleavage of C4 and C2 [9], even though MASP1 also is able to cleave C2. MASP2 thereby acts in a similar manner as C1s. The role of MASP1 and MASP3 is not yet established, but it is proposed that MASP1 contributes to the activation of MASP2 and has the ability to activate the alternative pathway through direct cleavage of C3 [16].

The alternative pathway

the formation of the alternative C5 convertase (C3bBbC3b). The alternative C5 convertase, in conformity with the classical C5 convertase, cleaves C5 into C5a and C5b fragments [15].

The terminal pathway

Generated C5b from the three complement pathways initiates the assembly of the terminal complement complex (TCC, sC5b-C9) directed against the cell membranes of target cells and pathogens [8]. In absence of a cell membrane, the sC5b-C9 remains in the fluid phase but when attached to a membrane, the complex is instead called the membrane attack complex (MAC) [11]. C5b complexed with complement component C6 can interact with lipid membranes and the additional binding of complement components C7 and C8 to the complex leads to further penetration of the target cell membrane [2]. As multiple complement components C9 bind to the C5b-C8 complex, a cylinder-like pore with a diameter of about 100 Å (1 Ångström = 10-10 m) is formed that causes cell damage through lysis by osmotic disruption [17-19].

Biomaterial-induced complement activation

Figure 2. A schematic overview of the different complement activation pathways. After activation, the classical

Table II. Selected proteins of the complement system [2, 7].

Protein Molecular weight Blood concentration Function

(kDa) (µg/mL)

C1 750 100 The C1 complex initiates the classical

pathway

C1q 410 75-150 Binds to antigen-antibody complexes,

certain pathogens and structures

C1r 85 50 Activates C1s by proteolytic cleavage

C1s 85 50 The active serine protease cleaves C4

and C2

C2 102 20 C2a cleaves C3 and C5 as the active

enzyme of the classical/ lectin pathway C3 and C5 convertases

C3 185 1000-2000 C3b functions as a component of

the alternative pathway C3 and C5 convertases, classical/ lectin pathway C5 convertase as well as an opsonin

C3a is an anaphylatoxin that

stim-ulates inflammation

C4 210 300-600 C4b functions as a component of the

classical/ lectin pathway C3 and C5 convertases

C4a is an anaphylatoxin that

stimulates inflammation

C5 190 80 C5b initiates formation of the

membrane attack complex – MAC

C5a is an anaphylatoxin that

stim-ulates inflammation

C6 110 45 Component of the MAC

C7 100 90 Component of the MAC

C8 155 60 Component of the MAC

C9 79 60 Component of the MAC

Factor B 93 200 Bb is the active enzyme of the

alternative pathway C3 and C5 convertases

Factor D 25 1-2 Cleaves factor B associated with C3b

Properdin 56 per subunit 25 Stabilizes the alternative pathway C3

(up to four subunits) convertase

Complement component C1q

Complement component C1q has a molecular weight of approximately 410 kDa [2] and is made up of 3 kinds of polypeptide chains (6 A-chains, 6 B-chains and 6 C-chains) with N-terminal collagen-like regions and C-N-terminal globular regions [14]. C1q, belonging to the family of collectins [19], is often referred to as a “bouquet of tulips” due to the six globular recognition domains linked together by collagen-like triple helical fibers forming a stalk (see Figure 3) [20]. C1q has several receptors that induce various immune effector functions, such as clearance of apoptotic cells, modulation of cellular cytotoxicity, increased surface expression of adhesion molecules and enhancement of phagocytosis as well as immunoglobulin secretion [14].

Figure 3. The classical pathway initial C1 complex is comprised of three subunits - C1q, C1r and C1s. C1q is

often referred to as a “bouquet of tulips” due to the globular heads and the collagen-like “stalk”. (Image source: Adapted from Favoreel, H.W., et al., Virus complement evasion strategies.)

Classical pathway complement activation is initiated by the binding of C1q to antigen-antibody complexes (IgG and IgM), surfaces of certain pathogens such as human immunodeficiency virus (HIV) [19], apoptotic cells [2], or structures such as lipopolysaccharides (LPS) and C-reactive protein (CRP) [21]. If more than one globular domain of C1q bind to a target, the C1 complex undergoes a conformational change and the associated zymogen C1r becomes self-activated. [19]. Since C1q binds only weakly to single immunoglobulin constant regions (i.e. Fc regions), multiple binding to Fc regions is required for complement activation. Multiple closely spaced Fc regions are found in antigen-antibody complexes enclosed with IgG and IgM and these are efficient activators of complement. Exactly where on the globular domains of C1q that IgG and IgM bind is not yet known, but it is proposed that C1q binds to the Cγ2 domain of IgG and Cμ3 domains of IgM present in the Fc regions [14].

C1q

The bovine complement system

There are apparent differences between adult and fetal bovine complement systems regarding the quantity of complement components. The level of bovine complement components C1, C4, C5, C7 and C9 in adult bovine serum has shown to be high in comparison to complement components C2 and C8 in the same serum, as only low levels of these components have been detected. Interestingly, fetal bovine serum (FBS) only contains approximately 1-3% of C1 and C6 levels and 5-50% of C2, C4, C5, C7, C8 and C9 levels in comparison to adult bovine serum [22]. The level of C3 in FBS has shown to be extremely low, if detectable at all, but functional complement activity can still be present and must be evaluated before the possibility is dismissed [23].

Immunoglobulins

As mentioned earlier, the immune system can be divided into two parts - namely the innate and the aquired immune system. The aquired immune system, in contrast to the non-specific innate immune system, requires prior exposure to a foreign material to be able to combat it. Aquired immunity can be further divided into cell-mediated and humoral immunity, where the humoral immunity is mediated primarily by soluble proteins known as immunoglobulins.

General description

At their amino terminal end, both heavy chains and light chains comprise highly variable regions named fragment antigen-binding (Fab) regions. Within these variable regions, hypervariable complementarity-determining regions (CDRs) reside which compose the antigen-binding site and recognize specific antigen epitopes (see Figure 4). The constant regions of the carboxyl terminal ends of heavy and light chains do not contribute to antigen binding and recognition, but are responsible for immunobiological effector functions as these fragment crystallizable (Fc) regions interact with immunoglobulin receptors called Fc receptors found on many types of immune cells [24].

Figure 4.Schematic figure of the antibody antigen-binding site recognizing the unique region, the epitope, of a specific antigen. (Image source: Free image from Wikipedia, http://en.wikipedia.org/wiki/File:Antibody.svg)

Immunoglobulins are found in two forms, either membrane-bound on the surfaces of B cells or soluble secreted by plasma cell. Soluble immunoglobulins reside in the fluid fraction of the blood, the blood plasma, and in case of blood coagulation, the immunoglobulins are present in the residual blood serum [7]. About 80% of the total immunoglobulin content in blood serum is made up of IgG (see Table III) [24].

Table III. Properties of human IgG and IgM classes and subclasses [7, 18, 24].

Class/subclass Heavy chain MW Forms Serum concentration Classical pathway

(kDa) (mg/mL) complement activation*

IgG1 γ1 146 Monomer 5-10 + IgG2 γ2 146 Monomer 1.8-3.5 +/-IgG3 γ3 170 Monomer 0.6-1.2 ++ IgG4 γ4 146 Monomer 0.3-0.6 - IgM µ 900 Pentamer 0.5-2.0 ++

Immunoglobulins and complement activation

As a part of the humoral immunity, soluble immunoglobulins circulate in the blood and functions as effectors since binding of an antigen triggers several mechanisms that causes elimination of the antigen [7]. Different immunoglobulin classes and subclasses have different functions in the immune system due to the differences in heavy chain constant regions. Most IgG subclasses (not IgG4) and IgM can activate the complement system, while all other immunoglobulin classes lack this ability [10]. Activation of the classical complement pathway occurs when C1q binds to antibody-antigen complexes enclosed with these IgG and IgM [19]. IgG is not as efficient in activating the complement system as IgM, since the classical pathway activation requires binding of at least two Fc regions simultaneously and IgM consists of multiple Fc regions due to the pentameric structure [24].

Protein G

Protein G is an immunoglobulin-binding protein derived from the cell walls of certain

Streptococcus with the capacity of binding the Fc region of IgG with high affinity (Ka~108 M -1

) [24]. Protein G binds all human IgG subclasses and in affinity chromatography, protein G is widely used for purification of IgG by attachment of the recombinant proteins to inert supports like sepharose or agarose [25].

Blood

The blood cells (i.e. erythrocytes, leukocytes and thrombocytes) comprise 40-45% of total human blood volume of 3 to 6 liters and the remaining liquid fraction is called blood plasma. Blood plasma consists of 90% water and the remaining percentage is made up of soluble small molecules and macromolecules, including proteins, glucose, fatty acids and ions. The concentration of proteins in human plasma is about 60-80 g/L [2], out of which approximately 90% is comprised of the ten most abundant proteins [26]. Important plasma proteins among others are albumin, transport proteins, proteins of the complement and coagulation systems and immunoglobulins.

studies of the complement system in vitro, serum is preferable to plasma since it is a simplified system, however it should be noted that plasma provides more relevant physiological in vivo indications [27] due to interactions between the complement system, the coagulation system and the different blood cells [4].

AIM

MATERIALS AND METHODS

Figure 5. Flow chart summarizing approaches and methods evaluated for antibody-independent and

biomaterial-induced classical pathway activation studies.

Fetal bovine serum

Detection of bovine Igs

detected using 100 µl of a polyclonal rabbit anti-bovine-immunoglobulins-HRP (Dako, Glostrup, Denmark) diluted 1:500. After incubation for 60 minutes at RT (shaking) and washing, staining followed by 100 µl of o-phenylendiamine dihydrochloride (OPD, 1 mg/mL, Sigma, St. Louis, MO, USA) in citrate buffer (0.35 mM containing 70 mM Na-phosphate, pH 5.0) with H2O2 (1 µl/mL buffer). The reaction was stopped with 100 µl 1 M sulfuric acid (H2SO4) and the absorbance was measured with an ELISA reader (SpectraCount, Packard, Canberra Company, Australia) at 490 nm.

Biotinylation of anti-human-C3c

To remove preservatives (Na-azide, NaN3) from anti-human-C3c (Dako, Glostrup, Denmark) before biotinylation, a protein desalting spin column (Thermo Scientific, Rockford, IL, USA) was used according to manufacturer’s instructions. The column was initially equilibrated three times with coupling buffer containing 0.1 M NaHCO3 (pH 8.3) and 0.5 M NaCl. After washing, 100 µl of anti-human-C3c was added to the column and the column was centrifuged at 1500g for 2 minutes. To the antibody, 900 µl of coupling buffer and 100 µl of biotin amidohexanoic acid N-hydroxysuccinimide ester (Sigma, St. Louis, MO, USA) dissolved in dimethyl sulfoxide (8 mg/mL, DMSO) was added and the biotinylation reaction was allowed to proceed for 30 minutes in RT. The mixture was dialyzed at 4ºC over night against 250 mL of PBS to remove excess biotin reagent. NaN3 was added to a final concentration of 0.05%.

Detection of complement activation by measuring C3c

As FBS has shown to comprise low levels of complement components [22], the complement activity was evaluated by detection of the C3 fragment C3c [29]. Several attempts to detect C3c and achieve complement activation in FBS were done by different experimental methods and set-ups. Dilution of samples was performed with PBS and biotinylated anti-human-C3c was diluted with washing buffer (PBS containing 0.05% Tween 20). Parameters varied in the different experimental set-ups were dilution of biotinylated anti-human-C3c, dilution of fetal, newborn and adult bovine serum and addition of bovine IgG to physiological concentration (~13.5 mg/mL) for enhancement of complement activation.

buffer. C3c depositions in the wells were detected with 100 µl of biotinylated anti-human-C3c, incubation 60 minutes shaking at RT, followed by washing. The wells were incubated for 15 minutes with 100 µl HRP-conjugated streptavidin (Amersham, Little Chalfort, UK) diluted 1:500 with working buffer containing 1% bovine serum albumin (BSA, Sigma, St. Louis, MO, USA) and 10 mM EDTA (ethylenediaminetetraacetic acid). After washing and staining with 100 µl 3,3,5,5-tetramethylbenzidine (TMB, 6 mg/mL, Serva, Heidelberg, Germany) in Na-acetate buffer (0.11 M, pH 5.5) containing H2O2 (1µl/mL buffer), the absorbance was measured with an ELISA reader at 450 nm.

Biomaterial-induced complement activation study

Polymers P1 and P2 were incubated with undiluted FBS or normal human serum (NHS, 2.5 mg/mL) in heparinized 2 mL Eppendorf tubes for 60 minutes at 37ºC with continuous rotation (20 rpm). As controls, serum samples incubated in the same manner but with no added polymers were used. After incubation, the polymer particles were centrifuged down and the supernatants was transferred to Eppendorf tubes containing EDTA (complexes metal ions necessary for the complement system and thereby inhibits further activation) to a final concentration of 10 mM and put on ice. To the unwashed polymer particles, 400 µl 2% SDS in PBS were added to elute adsorbed proteins and the tubes were incubated for 20 minutes at 37ºC with rotation. After centrifugation the eluates were transferred to Eppendorf tubes containing EDTA (to a final concentration of 10 mM). All samples were stored at –80ºC until sandwich-ELISA measurements of C3a and sTCC.

Detection of complement activation by measuring C3a with sandwich-ELISA

RT shaking and subsequently washed three times. The wells were incubated for 15 minutes with 50 µl streptavidin-HRP diluted 1:500 with working buffer. After washing and staining with TMB (50 µl) as described earlier, the absorbance was measured with an ELISA reader at 450 nm. Evaluation of obtained results was performed using the software DeltaSoft (BioMetallics Inc., Princeton, NJ, USA).

Detection of complement activation by measuring sC5b-C9 with sandwich-ELISA

Detection of soluble C5b-C9 (TCC) is an important tool to measure C5 activation and thus complement activation [11]. A polystyrene microtiter ELISA plate was coated at 4ºC over night with monoclonal anti-human-C9 aE11 (Diatec Monoclonals AS, Oslo, Norway) as the capturing antibody. After blocking of remaining protein-binding sites with 300 µl working buffer, 100 µl of diluted serum and eluates, zymosan-activated serum (served as standard in two-fold dilution, AU/mL) and sC5b-C9 control diluted 1:25 in working buffer were added to the plate and incubated for 60 minutes at RT shaking. Subsequent incubation for 60 minutes at RT shaking with 100 µl polyclonal rabbit anti-human-C5 (Dako, Glostrup, Denmark) followed by anti-rabbit-immunoglobulin-HRP (Dako, Glostrup, Denmark), both diluted 1:500 with working buffer, resulted in detection of sC5b-C9 by TMB staining (100 µl) as described earlier and subsequent absorbance measurement at 450 nm. Evaluation of obtained results was performed using the software DeltaSoft (BioMetallics Inc., Princeton, NJ, USA).

Human serum

IgG-depletion using a HiTrapTM Protein G column

Since no established protocol for IgG-depletion of NHS without interference of complement activity existed, a new method had to be developed. Initially the IgG-depletion was performed in absence of NaCl (sodium chloride) but since this also lead to the depletion of C1q, the protocol below was established. C1q has affinity for Sepharose and the presence of NaCl is crucial to avoid C1q binding to Sepharose.

was applied onto the column (3 mL) and the first mL of throughput was discarded. The following 1.5 mL of NHS was the desired IgG-depleted fraction and was kept on ice until dialyzation against 2 L PBS over night and storage at –80ºC. After washing with 5 mL of binding buffer, bound IgG was eluted using 10 column volumes of elution buffer (0.1 M glycine-HCl, pH 2.7). Finally the column was washed with 10 mL of binding buffer and 5 mL of 20% ethanol for storage in at 4ºC. The IgG-depletion was evaluated by SDS-PAGE (sodium dodecyl sulphate polyacrylamide gel electrophoresis) with separating gel 12% under reduced conditions followed by silver staining according to manufacture’s instructions (Bio-Rad, Hercules, CA, USA). Precision Plus ProteinTM Standards (Bio-Rad, Hercules, CA, USA) served as molecular weight marker.

Detection of IgG

Initially a polystyrene microtiter ELISA plate was coated over night at 4ºC with 150 µl anti-human-IgG (Dako, Glostrup, Denmark) diluted 1:200 in PBS. After washing three times with washing buffer, blocking of remaining protein-binding sites using 300 µl working buffer. Two-fold diluted IgG-depleted human serum (1:1-1:8,192) and NHS (1:100-1:819,200) in working buffer was added to the plate of 100 µl and incubated for 1 h in RT. Working buffer served as blank. Bound IgG was detected by incubation for 1 h in RT with 100 µl of a mixture of biotinylated and non-biotinylated anti-human-IgG diluted 1:800 in working buffer. HRP-conjugated streptavidin diluted 1:500 followed washing and was incubated for 15 minutes at RT shaking. After washing, staining followed with 100 µl of OPD as described earlier and the absorbance was measured with an ELISA reader at 490 nm.

IgM-depletion using a HiTrapTM IgM column

Ammonium sulphate ((NH4)2SO4) was, according to manufacturer’s instructions, supposed to be used in the IgM-depletion using a HiTrapTM IgM column. After several trial and errors to achieve IgM-depletion as well as maintaining functional complement activation, the protocol below for IgM-depletion of NHS with best results was established. No regard of C1q-depletion was taken (compensating C1q-addition afterwards if necessary was possible), since NaCl in early experimental set-ups was shown not to possess the ability to promote IgM-depletion as (NH4)2SO4 did and thus interfered with the depletion.

following buffers - binding buffer (20 mM Na-phosphate containing 10 mM EDTA and 0.8 M (NH4)2SO4, pH 7.5), elution buffer (20mM Na-phosphate containing 10 mM EDTA, pH 7.5) and regeneration buffer (20 mM Na- phosphate with 30% isopropanol and 10 mM EDTA, pH 7.5). Following equilibration with 5 mL binding buffer, 3 mL of NHS containing 0.8 M (NH4)2SO4 and 10 mM EDTA was applied onto the column. The first mL was discarded and the following 2 mL were collected and put on ice until storage at –80ºC. After column washing with 5 mL binding buffer, the bound IgM was eluted by applying 10 column volumes of elution buffer. Regeneration was obtained with 7 mL of regeneration buffer and the column was re-equilibrated with 5 mL of binding buffer. The column was stored in 20% ethanol at 4ºC. The IgM-depletion was evaluated by SDS-PAGE with subsequent WB (western blot) analysis using a biotinylated anti-human-IgM (Dako, Glostrup, Denmark) diluted 1:400 in working buffer and DAB (3,3-diaminobenzidine) staining. Precision Plus ProteinTM Standards served as molecular weight marker.

Coupling of anti-human-IgM to CNBr-activated Sepharose™ 4B

Preservatives were removed from 500 µl anti-human-IgM (5.6 mg/mL) by dialyzation against coupling buffer. Of the CNBr-activated Sepharose™ 4B (Pharmacia Biotech, Uppsala, Sweden), 2.0 g were transferred to a 50 mL Falcon tube and activated by 1 mM HCl two times with subsequent centrifugation at 1000 rpm for 5 minutes. The dialyzed antibody was added to 12 mL coupling buffer and UV280 nm was measured with NanoDrop for later control of successful coupling. After washing of the Sepharose with coupling buffer and centrifugation at 1000 rpm for 5 minutes, the antibody-coupling buffer mixture was added to the Sepharose and the ligand was allowed to bind for 30 minutes at RT. Centrifugation at 1000 rpm for 10 minutes was followed by UV280 nm which showed successful coupling and the Falcon tube was filled with blocking buffer (0.2 M glycine, pH 8.0) for 2 h at RT. The anti-human-IgM-coupled Sepharose was washed with washing buffer (0.1 M Na-acetate containing 0.5 M NaCl, pH 4.0) and coupling buffer alternately (three times with each buffer) with centrifugation at 1000 rpm for 10 minutes in between and finally three times with PBS. The anti-human-IgM-coupled Sepharose was stored in PBS containing 0.05% NaN3.

IgM-depletion using anti-human-IgM coupled to CNBr-activated Sepharose™ 4B

IgM-depletion. This was also the case in early experimental set-ups of this method and therefore no regard of simultaneous C1q-depletion in the protocol was taken as NaCl was excluded.

Washing of 1 mL anti-human-IgM coupled Sepharose with PBS was followed by the application of 2 mL EDTA-treated NHS and subsequent incubation (shaking) for 30 minutes at RT. After centrifugation at 1000 rpm for 5 minutes, the serum was retained on ice until storage at –80ºC. The Sepharose was regenerated with 8 mL of 0.1 M Tris buffer containing 0.5 M NaCl (pH 8.5) and 0.1 M Na-acetate buffer containing 0.5 M NaCl (pH 4.0) alternately three times with centrifugation at 1000 rpm for 5 minutes in between. The anti-human-IgM-coupled Sepharose was finally washed with PBS three times and stored in 0.05% NaN3 in PBS. The IgM-depletion was evaluated by SDS-PAGE with subsequent WB analysis using a biotinylated anti-human-IgM diluted 1:400 in working buffer and DAB staining. Precision Plus ProteinTM Standards served as molecular weight marker.

Biomaterial-induced complement activation study

Different serum variations were prepared – depleted serum containing C1q, depleted serum containing IgG and depleted serum containing both C1q and IgG. IgG-depleted C1q (1.18 µg/µl) and IgG (165 µg/µl, Beriglobin, CLS Behring, King of Prussia, PA) were added to the sera in physiological concentrations of 75 µg/mL and 13.5 µg/µl respectively. Finally all sera were diluted 1:8 in VBS++ (veronal buffered saline containing 0.15 mM Ca2+ and 0.5 mM Mg2+, pH 7.4) to eliminate contribution of the alternative pathway to complement activation. NHS, also diluted 1:8, served as comparison control and IgG-depleted serum containing C1q and IgG served as control sample (also incubated with continuous rotation but without added polymers). Start values of IgG-depleted serum (start value 1) as well as IgG-depleted serum containing both C1q and IgG (start value 2) were also obtained by EDTA-treatment of the sera.

Detection of complement activation by measuring C3a with sandwich-ELISA

The same protocol as earlier described was used for measuring C3a with sandwich-ELISA. For dilution of supernatants and eluates from the biomaterial-induced complement activation study, see Appendix.

Statistics

RESULTS

Fetal bovine serum

Detection of bovine Igs

HRP-conjugated polyclonal anti-bovine-immunoglobulins followed by OPD-staining was used to detected bovine immunoglobulins in FBS (newborn and adult bovine serum served as positive control samples). The result (see Figure 6) showed no detectable immunoglobulins in the screened FBS and this indicates that the serum consequently was deficient of immunoglobulins. 0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 1 :2 1 :4 1 :8 1 :1 6 1 :3 2 1 :6 4 1 :1 2 8 1 :2 5 6 1 :5 1 2 1 :1 ,0 2 4 1 :2 ,0 4 8 1 :4 ,0 9 6 1 :8 ,1 9 2 1 :1 6 ,3 8 4 Serum dilution A b s o rb a n c e 4 9 0 n m

Fetal bovine serum Newborn bovine serum Adult bovine serum

Figure 6. Detection of bovine immunoglobulins in fetal, newborn and adult bovine serum at various dilutions

using polyclonal anti-bovine-immunoglobulins-HRP followed by OPD-staining. The result showed that the screened FBS contained no immunoglobulins in contrast to newborn and adult bovine serum.

Human serum

IgG-depletion using a HiTrapTM Protein G column

The IgG-depletion of NHS was performed by the employment of a HiTrapTM Protein G column with affinity for all human subclasses of IgG. After SDS-PAGE with subsequent silver staining, the resulting gel showed a prosperous IgG-depletion, as IgG heavy chain and light chain bands (at approximately 50 kDa and 25 kDa, respectively) were less intense as compared to NHS (see Figure 7). The stained gel further indicated that the depleted serum comprised approximately the same protein profile as NHS and no noticeable unspecific protein depletion could be seen.

Figure 7. Silver stained gel visualizing normal (A-C) and IgG-depleted human serum (D-F) applied to the gel in

three different protein concentrations (from left to right for each serum; 10 µg/mL, 5 µg/mL and 2,5 µg/mL). Prosperous IgG-depletion could be seen as the IgG-depleted serum showed less intense immunoglobulin heavy and light chain bands compared to NHS.

Detection of IgG

IgM-depletion using a HiTrapTM IgM column and anti-human-IgM coupled to CNBr-activated Sepharose™ 4B

The IgM content of the assumed IgM-depleted sera was evaluated by WB analysis followed by DAB staining. The result clearly showed (see Figure 8) that none of the applied methods successfully depleted NHS of IgM, although the HiTrapTM IgM column showed some IgM-depletion as IgM was detected in the eluate. Additional purifications by anti-human-IgM coupled to CNBr-activated Sepharose did not result in extended IgM-depletion.

Figure 8. WB analysis followed by DAB staining visualizing HiTrapTM IgM column eluate (A), HiTrapTM IgM column throughput serum (B), fractions of serum after one, two and three times of IgM-depletion using anti-human-IgM coupled to CNBr-activated Sepharose™ 4B (C-E) and IgM-containing IgG-depleted serum (F) for comparison. The resulting blot clearly showed unsuccessful IgM-depletion by both applied methods, as IgM still was present in the serum and the IgM content had not decreased significantly compared to the non-IgM-depleted serum (F).

Biomaterial-induced complement activation study

P1 0 200 400 600 800 1000 1200 1400 NHS +C1q +IgG +C1q +IgG Pre-C1q Start value 1 Start value 2 Control sample [C 3 a ] (n g /m L ) Eluate Supernatant P2 0 200 400 600 800 1000 1200 NHS +C1q +IgG +C1q +IgG Pre-C1q Start value 1 Start value 2 Control sample [C 3 a ] (n g /m L ) Eluate Supernatant C3a in the supernatants than in the eluates. Control sample indicated elevated activation in comparison to start value.

Figure 9. Generation of complement activation marker C3a in various sera subjected to polymer P1 and P2 for

DISCUSSION

To develop and engineer biocompatible materials and devices, it is necessary to have full knowledge of the different fundamental mechanisms underlying the biological responses to implanted biomaterials and extracorporeal devices. Complement activation after biomaterial-blood contact is a significant problem to overcome in the enhancement of biomaterial hemocompatibility and it is crucial to test novel biomaterials for their potency to activate the complement system. The focus in this degree project work was on preparing and evaluating immunoglobulin-free sera for further antibody-independent and biomaterial-induced complement activation studies.

Immunoglobulin deficient FBS was evaluated regarding complement activity but no complement activation could be detected in the screened serum despite various measurements of complement activation markers. This was not surprising as earlier publications had indicated low levels of complement components in FBS, but the possibility of functional complement activity could not be dismissed until the serum was evaluated thoroughly.

precipitation using ammonium sulphate could be seen, conceivably due to lack of coagulation proteins such as fibrinogen (physiological plasma concentration of 1.5-4.5 g/L [31]). The possibility of precipitation of complement proteins was considered but as the method did not show prosperous results, further evaluation was not performed regarding complement functionality. According to manufacturer of CNBr-activated Sepharose 4B, NaCl could be used to elute bound ligands. This offered no ground for high expectations and as predicted IgM-depletion was prevented in presence of NaCl. Although NaCl was excluded and several depletion processes was performed sequentially, the method still was not successful in depleting NHS of IgM. One possible reason for this might be that too low concentration of capturing antibody was coupled to the Sepharose. The coupling procedure was however effectively achieved and could not significantly contribute to the method failure. Another issue using this method was unavoidable dilution of the serum, as the serum was diluted for every time the depletion procedure was repeated. Because of intended dilution to eliminate the effect of the alternative pathway on complement activation, indeterminate dilution was undesirable and this is something that necessitates further optimization and evaluation to avoid using this method.

Since none of the applied methods successfully depleted NHS of IgM, the antibody-independent and biomaterial-induced classical pathway activation could not be thoroughly studied. Some potential results were however obtained using the IgG-depleted serum and will be discussed below. However, it is kittle and difficult to draw any concrete conclusions as the study was performed to few times to provide accurate and reliable results.

must be noted that IgM still was present in the serum and could induce activation as well. IgG appeared to enhance complement activations as predicted. The complement activation in NHS appears to be approximately 100-fold higher than in the serum containing both C1q and IgG when the two sera in fact ought to have similar contents and activity. Possible explanations for the decreased activation could be distressed proteins after several and various depletion methods or limited functional activity in added C1q. Enhancement of complement activation will further on be evaluated with the addition of known complement activators. When comparing control sample and start value 2 (containing the same serum), the control sample showed elevated activation probably due to heat and rotation during incubation.

Some small differences between the polymers were however revealed, as complement activation in the prepared sera varied slightly when studied in detail. In general, hydrophilic and carboxylate-containing polymers such as P1 adsorb less protein [6]. Interestingly, P1 adsorbed more C3a on the surface regardless of sera in opposite to P2, which had a higher concentration of C3a in supernatants than in eluates. One possible reason for the larger amount of adsorbed C3a to P1 could be attraction between the nitrogen-containing and thus negatively charged crosslinker DAP and the positively charged C3a. P1 also appeared to induce complement activation to higher extent than P2 in some sera and this is in conformity with earlier finding, as P1 has shown to be a potent activator of complement system probably due to the large amounts of adsorbed C1q and IgG (unpublished data).

CONCLUSIONS

- FBS is deficient of immunoglobulins.

- No detectable complement activity is present in FBS. - A close to complete IgG-depletion of NHS is achievable. - IgG-depleted NHS has low but functional complement activity.

- None of the applied methods for IgM-depletion successfully depleted NHS of IgM.

To be able to thoroughly study the antibody-independent and biomaterial-induced classical pathway activation, further evaluation regarding the immunoglobulin free serum is necessary. Two further aspects to evaluate would be the possibility to accomplish a complete IgM-depletion by optimization of existing methods as well as the possibility to enhance the complement activation in the immunoglobulin free serum with known complement activators.

ACKNOWLEDGEMENT

First and foremost, I would gratefully like to acknowledge Prof. Kristina Nilsson Ekdahl for the opportunity to perform this degree project work and for the chance to enter the thrilling world of complement activation. Further, Anna E. Engberg (Ph.D.) is acknowledged for her never-ending enthusiasm and goofiness, as well as support regarding everything from experimental set-ups to encouraging peptalks. Two members of the research group possessing enormous amounts of knowledge, Per H. Nilsson and Kerstin Sandholm (both MSc.), are also acknowledged for their helpful guidance and inspiration during this degree project work. My former academy classmate Gustaf Olsson (MSc.) is acknowledged for his motivating chitchats, computational aid and for always being close to hand (nearby laboratory facilities is great).

REFERENCES

1. Tang, L. and J.W. Eaton, Natural responses to unnatural materials: A molecular

mechanism for foreign body reactions. Mol Med, 1999. 5(6): p. 351-8.

2. Engberg, A.E., Biomaterial and Hemocompatibility, in School of Natural Science. 2010, Linnaeus University: Kalmar.

3. Onuki, Y., et al., A review of the biocompatibility of implantable devices: current

challenges to overcome foreign body response. J Diabetes Sci Technol, 2008. 2(6): p.

1003-15.

4. Nilsson, B., et al., The role of complement in biomaterial-induced inflammation. Mol Immunol, 2007. 44(1-3): p. 82-94.

5. Williams, D.F., On the nature of biomaterials. Biomaterials, 2009. 30(30): p. 5897-909.

6. Wang, Y.X., et al., Effects of the chemical structure and the surface properties of

polymeric biomaterials on their biocompatibility. Pharm Res, 2004. 21(8): p. 1362-73.

7. Abbas, A.K., Lichtman, A.H., Cellular and Molecular Immunology. 5 ed. 2003: Saunders.

8. Markiewski, M.M., et al., Complement and coagulation: strangers or partners in

crime? Trends Immunol, 2007. 28(4): p. 184-92.

9. Sim, R.B. and S.A. Tsiftsoglou, Proteases of the complement system. Biochem Soc Trans, 2004. 32(Pt 1): p. 21-7.

10. Mölne, J., and Wold, A., Inflammation. 1 ed. Vol. 1. 2007: Liber AB.

11. Janatova, J., Activation and control of complement, inflammation, and infection

associated with the use of biomedical polymers. ASAIO J, 2000. 46(6): p. S53-62.

12. Mastellos, D., et al., Complement: structure, functions, evolution, and viral molecular

mimicry. Immunol Res, 2003. 27(2-3): p. 367-86.

13. Kirschfink, M., Targeting complement in therapy. Immunol Rev, 2001. 180: p. 177-89.

14. Kishore, U. and K.B. Reid, C1q: structure, function, and receptors. Immunopharmacology, 2000. 49(1-2): p. 159-70.

15. Gorbet, M.B. and M.V. Sefton, Biomaterial-associated thrombosis: roles of

coagulation factors, complement, platelets and leukocytes. Biomaterials, 2004. 25(26):

16. Takahashi, M., et al., Mannose-binding lectin (MBL)-associated serine protease

(MASP)-1 contributes to activation of the lectin complement pathway. J Immunol,

2008. 180(9): p. 6132-8.

17. Gros, P., F.J. Milder, and B.J. Janssen, Complement driven by conformational

changes. Nat Rev Immunol, 2008. 8(1): p. 48-58.

18. Wilson, K., Walker, J., Principles and Techniques of Biochemistry and Molecular

Biology. 6 ed. 2005: Cambridge University Press.

19. Favoreel, H.W., et al., Virus complement evasion strategies. J Gen Virol, 2003. 84(Pt 1): p. 1-15.

20. Gaboriaud, C., et al., Structure and activation of the C1 complex of complement:

unraveling the puzzle. Trends Immunol, 2004. 25(7): p. 368-73.

21. Skoglund, C., et al., C1q induces a rapid up-regulation of P-selectin and modulates

collagen- and collagen-related peptide-triggered activation in human platelets.

Immunobiology.

22. Triglia, R.P. and W.D. Linscott, Titers of nine complement components, conglutinin

and C3b-inactivator in adult and fetal bovine sera. Mol Immunol, 1980. 17(6): p.

741-8.

23. Linscott, W.D. and R.P. Triglia, The bovine complement system. Adv Exp Med Biol, 1981. 137: p. 413-30.

24. Kindt, T.J., Goldsby, R.A., Osborne, B.A., Kuby, J., Kuby Immunology. 6 ed. 2007: W.H. Freeman and Company.

25. GE Healthcare, Instructions for usage of HiTrap Protein G HP. 2006.

26. Pernemalm, M., R. Lewensohn, and J. Lehtio, Affinity prefractionation for MS-based

plasma proteomics. Proteomics, 2009. 9(6): p. 1420-7.

27. Lachmann, P.J., Preparing serum for functional complement assays. J Immunol Methods. 352(1-2): p. 195-7.

28. Gould, E.A., et al., Detection of antibody to bovine syncytial virus and respiratory

syncytial virus in bovine fetal serum. J Clin Microbiol, 1978. 8(2): p. 233-7.

29. Janssen, B.J., et al., Structures of complement component C3 provide insights into the

function and evolution of immunity. Nature, 2005. 437(7058): p. 505-11.

30. Gadjeva, M.G., et al., Interaction of human C1q with IgG and IgM: revisited. Biochemistry, 2008. 47(49): p. 13093-102.

31. Boyd, J.H., et al., Fibrinogen decreases cardiomyocyte contractility through an

APPENDIX

Table IV. Supernatant and eluate dilutions for biomaterial-induced complement activation study.