ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2020

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1888

Characterization of conjugatedprotease inhibitors

ERIKA BILLINGER

ISSN 1651-6214ISBN 978-91-513-0835-7urn:nbn:se:uu:diva-398909

Dissertation presented at Uppsala University to be publicly examined in B42, BMC,Husargatan, Uppsala, Thursday, 13 February 2020 at 09:15 for the degree of Doctor ofPhilosophy. The examination will be conducted in English. Faculty examiner: Professor BorisTurk (Jozef Stefan Institute, Department of Biochemistry, Molecular and Structural Biology).

AbstractBillinger, E. 2020. Characterization of conjugated protease inhibitors. Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty of Science and Technology 1888. 86 pp.Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0835-7.

The overall theme of this thesis is a step by step approach for the design and characterizationof conjugated protease inhibitors. This involves both a new assay method for protease activityand protease inhibition (paper I), a study of the stoichiometry for protease inhibitor interaction(paper II), design of inhibitory peptides (paper IV) and the construction of inhibitor conjugates(paper III & IV).

(I) A model based primarily on erosion in gelatin for protease activity and inhibition studieswas designed. The model was also extended to a separate protective layer covering the layercontaining the target substrate. A good correlation between protease concentration and rateof erosion was observed. Similarly, increased concentration of inhibitors gave a systematicdecrease in the erosion rate. Kinetic analyses of a two-layer model with substrate in the bottomlayer displayed a strict dependence of both inhibitor concentration and thickness of the top“protective” layer.

(II) The binding stoichiometry between pancreatic proteases and a serine protease inhibitorpurified from potato tubers was determined by chromatography-coupled light scatteringmeasurements. This revealed that the inhibitor was able to bind trypsin in a 2:1 complex,whereas the data for a-chymotrypsin clearly showed a limitation to 1:1 complex. The sameexperiment carried out with elastase and the potato inhibitor gave only weak indications ofcomplex formation under the conditions used.

(III) A serine protease inhibitor was extracted from potato tubers and conjugated to soluble,prefractionated dextran or inorganic particles. A certain degree of inhibitory activity wasretained for both the dextran-conjugated and particle-conjugated inhibitor. The apparent Ki

value of the dextran-conjugated inhibitor was found to be in the same range as that for freeinhibitor. The dextran conjugate retained a higher activity than the free inhibitor after 1 monthof storage at room temperature. Conjugation to oxide particles improved the heat stability ofthe inhibitor.

(IV) New synthetic Leupeptin analogues, Ahx-Phe-Leu-Arg-COOH & Ahx-Leu-Leu-Arg-COOH, were synthesized with solid-phase peptide synthesis using Fmoc strategy. Thesetripeptide inhibitors were tight binding inhibitors to the target enzyme trypsin, similar to thenatural occurring leupeptin. The phenylalanine containing synthetic analogue was conjugated toinorganic particles and agarose gel beads. In all cases, the inhibitory activity was well preserved.

Keywords: Serine protease inhibitors, conjugation, immobilization, leupeptin analogues,potato serine protease inhibitor, soluble carriers, inorganic carriers.

Erika Billinger, Department of Chemistry - BMC, Box 576, Uppsala University, SE-75123Uppsala, Sweden.

© Erika Billinger 2020

ISSN 1651-6214ISBN 978-91-513-0835-7urn:nbn:se:uu:diva-398909 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-398909)

Education is not the learning of facts, but the training of the mind to think

- Einstein

LIST OF PAPERS

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Billinger, E., Johansson, G. Kinetic studies of serine protease in-

hibitors in simple and rapid ‘active barrier’ model systems: Dif-fusion through an inhibitor barrier. Analytical Biochemistry, 2018, 546, 43–49.

II Billinger, E., Zuo, S., Lundmark, K., Johansson, G. Light scat-tering determination of the stoichiometry for protease-potato ser-ine protease inhibitor complexes. Analytical Biochemistry, 2019, 582, 113357. DOI: 10.1016/j.ab.2019.113357

III Billinger, E.*, Zuo, S., Johansson G. Characterization of serine protease inhibitor from Solanum tuberosum conjugated to solu-ble dextran & particle carriers. ACS Omega, 2019, 4, 18456-18464. DOI: 10.1021/acsomega.9b02815 *Corresponding author.

IV Billinger, E., Viljanen, J., Bergström Lind, S., Johansson, G. In-hibition properties of free and conjugated leupeptin analogues. 2019. Submitted.

Reprints were made with permission from the respective publishers. Further permissions related to paper III excerpted should be directed to the ACS. Publications not included in this thesis.

V Kjellander, M., † Billinger, E., † Ramachandraiah, H., Boman, M., Bergström Lind, S., Johansson, G. A flow-through nanoporous alumina trypsin bioreactor for mass spectrometry peptide finger-printing. Journal of Proteomics, 2018, 172, 165-172. † These authors contributed equally.

CONTRIBUTION REPORT

I. Participated in formulating the research idea, planned and per-formed all experimental work, wrote the initial draft of the manu-script and shared the writing of the manuscript.

II. Planned the overall experimental work, performed parts of it, par-

ticipated in writing the manuscript.

III. Participated in formulating the research idea, planned the initial scope, performed most experimental procedures, wrote a draft of the manuscript and organized the writing of the manuscript. Served as corresponding author.

IV. Participated in formulating the research idea, planned and per-

formed most experimental work, wrote the initial draft of the man-uscript and shared the writing of the manuscript.

CONTENTS

INTRODUCTION .....................................................................................11ENZYMES ...........................................................................................12

Proteases ..........................................................................................14The catalytic site ...............................................................................16Recognition sites for the substrate .....................................................17Kinetics of serine protease reaction ...................................................17Inhibition ..........................................................................................21

SERINE PROTEASE INHIBITORS .....................................................25Potato serine protease inhibitor .........................................................25Peptide aldehydes .............................................................................26

CONJUGATION ..................................................................................29Amide formation ..............................................................................31Reductive amination .........................................................................32Carriers ............................................................................................34

TOOLS SECTION ................................................................................39Static light scattering (SLS) ..............................................................40Solid-phase peptide synthesis (SPPS) ................................................43Mass spectroscopy (MS) ...................................................................46

MY WORK ...............................................................................................49PAPER I ...............................................................................................50

Kinetic studies of serine protease inhibitors in simple and rapid ‘active barrier’ model systems ..........................................................50One-layer model ...............................................................................50Two-layer model ..............................................................................52Conclusions ......................................................................................53

PAPER II ..............................................................................................54Light scattering determination of the stoichiometry for protease-potato serine protease inhibitor complexes .......................................54The active sites .................................................................................54Trypsin-PSPI interaction ...................................................................56α-chymotrypsin-PSPI interaction ......................................................57Elastase-PSPI interaction ..................................................................58Conclusions ......................................................................................59

PAPER III.............................................................................................60Characterization of serine protease inhibitor from solanum tuberosum conjugated to soluble dextran and particle carriers .........60Prefractionation & oxidation of dextran ............................................60Conjugation of PSPI .........................................................................61Possible conjugation sites .................................................................62Efficiency of the conjugates ..............................................................62The gelatin erosion method ...............................................................63Conclusions ......................................................................................64

PAPER IV ............................................................................................65Inhibition properties of free and conjugated leupeptin analogues ......65Oxidation & reduction of leupeptin ...................................................65SPPS of AHX-R-LEU-ARG-COOH .................................................66Inhibition properties of the two peptides ...........................................67Conjugation of AHX-PHE-LEU-ARG-COOH ..................................67Performance of the conjugates ..........................................................68Inhibition activity in a gelatin layer ...................................................68Conclusions ......................................................................................69

CONCLUDING REMARKS & FUTURE PERSPECTIVES ....................70

POPULÄRVETENSKAPLIG SAMMANFATTNING ..............................72Karakterisering av konjugerade proteasinhibitorer ...........................72

ACKNOWLEDGEMENTS .......................................................................76

REFERENCES ..........................................................................................79

ABBREVIATIONS

PSPI Potato serine protease inhibitor MS Mass spectroscopy BAPA Benzoyl-L-arginine 4-nitroanilide hydrochloride BTEE N-benzoyl-L-tyrosine ethyl ester SLS Static light scattering SEC Size-exclusion chromatography MALS Multi-angle light scattering UV Ultraviolet radiation RI Refractive index

11

INTRODUCTION

Welcome to my story!

Nice to see that you opened this thesis concerning Characterization of conju-gated protease inhibitors. It all started with a master thesis at a company that wanted to add inhibitors into protective formulations and study the effect of these inhibitors. That led me into the world of biochemistry at Uppsala Uni-versity. I have, throughout in my research, studied the interaction between en-zymes and free and conjugated inhibitors and that is what I will present in this thesis. The first part of the introduction will take you through enzymes, in particular serine proteases since these are the enzymes studied here, and in-hibitors to these serine proteases, both larger ones such as proteins and smaller ones such as tripeptides. In the second part I will guide you through conjuga-tion, the conjugation chemistry I have used and the different carriers that I have chosen. The third part is what I call “Tools section” and I will in this part introduce you to some techniques that I have used along the way. Finally, I will present the results of my five years in the lab. The first paper is a new developed application to evaluate my conjugates, the second paper determines the interaction stoichiometry between different serine proteases and a protein inhibitor, namely PSPI, an inhibitor found in the potato you might have eaten yesterday. Paper III and paper IV focuses on the conjugation of the selected inhibitors to different carriers, both polysaccharides and inorganic particles.

So, welcome to my past five years. I hope you will enjoy it as much as I have during these years!

12

What makes an enzyme an enzyme?

ENZYMES Enzymes are natures´ solution to sustain life. Living creatures need to con-stantly perform biochemical reactions which need to happen both efficiently and selectively. In order to achieve this, life uses catalytic proteins, so called enzymes, which are among the most specialized proteins that exist. Without an enzyme present, the reaction rate is rather slow, and is virtually insignifi-cant in comparison with an enzyme catalyzed reaction. For many processes the enzyme catalysis is thus crucial since the rate of the reaction is essential for living systems and actually gives a selection of reactions. We all need the enzymes since without them we can simply not digest our food, contract our muscles or even send a nerve signal at sufficient rates. Enzymes help reactions occur in unfavorable environments or catalyze quite complex reactions. En-zymes can connect reactions which are thermodynamically disadvantageous but crucial for our existing. An enzyme can increase the reaction rate by an order of 1012 to 1016 compared to an uncatalyzed reaction and compared to a chemically catalysis effect the enzymatic one is usually several orders of magnitude greater. Furthermore, it can catalyze a reaction under mild conditions with neutral pH, atmospheric pressure and normal temperatures. Still more important, enzymes have greater reaction specificity and an enzyme catalyst reaction rarely has any side prod-ucts. It also has capacity to control its action since the catalytic activity de-pends on the amount of substrate/product, allosteric control, modifications of the enzyme and the amount of enzyme synthesized within a cell.

13

Figure 1. Comparison of an enzyme catalyzed reaction with an uncatalyzed reaction.

The role of an enzyme is to increase the reaction rate, but it will not change the reaction equilibrium. The free energy profile of a reaction is often plotted in a reaction coordinate diagram where the free energy is plotted on the y-axis and the reaction progress on the x-axis as can be seen in Figure 1. An enzyme will enhance the reaction rate by lowering the activation energy, e.g. by stabi-lizing the transition state. The transition state is the reactive configuration of the molecule where bond breakage, bond formation and charge development have come to a certain point where the direction to substrate or product is equal. The enzyme will accelerate the interconversion of the substrate to prod-uct without being consumed itself.

without enzyme

with enzyme activation

energy with enzyme

activation energy without enzyme

Ener

gy

Reaction coordinate

S

P

overall energy released during reaction

14

PROTEASES

Figure 2. The overall reaction catalyzed by a protease.

Proteases act as sharp scissors and catalyze the hydrolysis of the peptide bond (Figure 2).1 They have been studied for over 100 years and covered by over 350,000 scientific articles, however they are still subject to both academic and pharmaceutical research. There are ∼700 protease genes present in the human genome and in a typical genome 2-4% of the genes are encoding for proteo-lytic enzymes.2 Proteases were first classified into endopeptidases and exo-peptidases but with increasing knowledge of their structure and mechanism of action they are now commonly divided into six classes:

I. Aspartic proteasesII. Glutamic proteases

III. MetalloproteasesIV. Cysteine proteasesV. Serine proteases

VI. Threonine proteases

The classification is thus based on the detailed catalytic mechanism in the ac-tive site of the proteases. In the reaction for classes I-III an activated water molecule is acting as a nucleophile that attacks the peptide bond. On the con-trary, in the classes IV-VI the acting nucleophile is an amino acid residue in the active site of the enzyme. Each class of proteases can be divided further into families and clans based on their structures.3

SERINE PROTEASES Serine proteases are the second most populated protease class in the human genome with >170 members.4 I have studied serine proteases in this thesis, and they use (as the name tells us) a serine residue, actually its hydroxyl group, as nucleophile in the catalysis. Serine proteases are found ubiquitously in na-ture, both in eukaryotes and prokaryotes, and they are among the most well-characterized and well-studied enzymes, displaying a wide range of biological roles.4 Serine proteases fall into two categories, based on their structure: tryp-sin-like or subtilisin-like. Since the enzymes used in this thesis belong to tryp-sin-like serine proteases, I will address these further. Trypsin-like proteases (TLPs) consist of a large family of enzymes with roles including digestion,

15

blood coagulation, fibrinolysis, development, fertilization, apoptosis and im-munity.4 TLPs include digestive enzymes such as trypsin, α-chymotrypsin and elastase.

Trypsin & α-chymotrypsin Trypsin and α-chymotrypsin are secreted by the glandular system in humans into the digestive system and catalyze the hydrolysis of polypeptides to smaller fragments.5 They are, similar to other serine proteases, secreted as zy-mogens in the pancreatic juice as a self-protection mechanism for the pancreas and have essential roles in the human digestive system. The zymogens require activation before receiving full enzymatic activity.6-8 Trypsin is secreted as trypsinogen into duodenum which produces enterokinase that activates trypsin from trypsinogen.6 Trypsin later converts other zymogens, including trypsin-ogen into active proteases. In this thesis the model experiments are mostly based on trypsin, a 23 kDa large protein with 223 amino acids and α-chymo-trypsin, a 25 kDa large protein with 241 amino acids residues in three poly-peptide chains that are linked together by disulfide bonds.9 Trypsin hydrolyzes peptide bonds between the carboxylic acid group of positively charged lysine and arginine and the amino group of the adjacent amino acid residue with a rather high selectivity.10 When it comes to α-chymotrypsin, it prefers an aro-matic side chain (e.g. tryptophan, phenylalanine, tyrosine) on the residue whose carbonyl carbon is part of the peptide bond to be cleaved.

Figure 3. Trypsin-catalyzed hydrolysis of BAPA

The activity of trypsin can be measured using the synthetic substrate BAPA that has a molecular weight of 435 Da. The product released is p-nitroaniline, a chromophore that absorbs light in the blue region at λ410 nm and is easily observed spectrophotometrically (Figure 3).

Figure 4. α-chymotrypsin-catalyzed hydrolysis of BTEE.

16

α-chymotrypsin is measured in the same way but with the synthetic substrate BTEE that has a molecular weight of 313 Da and can be monitored at 256 nm (Figure 4).11

THE CATALYTIC SITE The active site in an enzyme offers a special micro-environment where water can be excluded (or at least be controlled), thus providing a non-polar envi-ronment where electrostatic interactions can become stronger than in water. The architecture in the active site is crucial since the catalytic groups must be placed correctly in space for an enzymatic reaction to efficiently take place. The active site in TLPs has three residues that are crucial for the catalysis: Ser195, His57 and Asp102 (numbers refer to α-chymotrypsin). The active site residues of trypsin include His46 and Ser183.12 The catalytic triad constitutes an extensive hydrogen bonding network with Ser195 placed on one side and His57 and Asp102 placed on the other side (Figure 5). The hydrolysis of a substrate starts by positioning the peptide bond for attack. His57 acts as a gen-eral base in accepting a proton from Ser195, thus activating serine as a nucle-ophile. His57-H+ is stabilized by hydrogen bond to Asp102 and His57-H+ also acts as a general acid donating a proton to the nitrogen of the leaving group.13 The serine nucleophile attacks the carbonyl carbon of the incoming substrate and a tetrahedral acyl-enzyme complex is formed. The negative charge of the carbonyl is stabilized by hydrogen bonds in the oxyanion hole, which is formed by the backbone of NHs of Gly193 and Ser195. The tetrahedral inter-mediate collapses resulting in a leaving group (here His57-H+ donates its pro-ton) leaving the acyl-enzyme intermediate behind. The deacylation part of the reaction essentially repeats the above sequence: deprotonation of water by His57 creating a nucleophile, a second tetrahedral intermediate is formed which collapses and forms the second product (carboxylate anion) expelling Ser195. The active site is regenerated when the second product leaves by dif-fusion.14-18

Figure 5. The generally accepted mechanism for a serine protease.

Asp102

Ser195

His57

Ser195 95 Gly193

17

RECOGNITION SITES FOR THE SUBSTRATE The recognition site in a serine protease for the substrate includes the binding pocket for the side chain of the peptide substrate and the binding site for the backbone of the polypeptide chain. The substrate specificity of serine prote-ases is created by the surrounding residues in the active site. Schechter and Berger19 created the nomenclature for the pocket and for the substrate residues binding to the pocket residues (Figure 6). There are several binding sites to consider. The residues on the substrate N-terminal side are named P1, P2, P3, …etc. The residues on the C-terminal are namned P1´, P2´, P3´, … etc. The residues in the active site that interact with the substrate residues are named S1, S2, S3, … etc and S1´, S2´, S3´, … etc. The P1 residue interact with S1 and so on for all residues.20

Figure 6. Schematic illustration of the recognition site in a serine protease.

The specificity for the three enzymes studied in this thesis lies in the different side chains and how they bind to a substrate. Trypsin´s S1 site includes a neg-atively charged aspartate residue which will guide trypsin to cleave the peptide bond in a substrate on the carboxyl side of positively charged residues, such as arginine (Arg) or lysine (Lys). α-Chymotrypsin ´s S1 site includes one ser-ine and two glycines. Due to the small side chains of these amino acids the active site in α-chymotrypsin has room for amino acids containing bulky side chains, namely P1 residues, such as phenylalanine (Phe), tyrosine (Tyr) and tryptophan (Trp). Regarding elastase, the cleavage occurs after small residues in the substrate due to a bulky Phe residue in S1 position.20

KINETICS OF SERINE PROTEASE REACTION The most common way of explaining an enzyme-catalyzed reaction is by ki-netic analysis. Many enzyme-substrate interactions follow a simple mecha-

Protease

C-terminus N-terminus

Substrate

P1´ P2´ P3´ P4´ P1 P2 P3 P4 P5

S5 S4 S3 S2 S1 S1´ S2´ S3´ S4´

Targeted bond

18

nism where an initial enzyme-substrate-complex (ES) forms and later decom-poses into a product thus releasing the enzyme to react again as can be seen below in Figure 7.

Figure 7. An enzyme catalyzes the reaction of a substrate to form product.

THE STEADY STATE APPROXIMATION Brown21 argued in 1902 that the enzyme catalyzed reaction could be described as follows:

In 1903 Henri22 put this scheme into a more complex mathematical framework to best describe what happens. Michaelis and Menten23 continued his work in 1913 and hence, the Henri-Michaelis-Menten equation arose. Briggs and Hal-dane24 took this equation even further in 1925 when they recognized that the equation can be described more generally. Henri, Michaelis and Menten deri-vatization was dependent on a rapid formation of the ES-complex but due to that most experimental measurements occurs when the ES-complex is at a constant steady-state concentration, the new equation (Equation 1) was formed (even though it is still referred to as the Henri-Michaelis-Menten equa-tion).

(Eq. 1) If we study the simple enzyme-catalyzed reaction mentioned above, Figure 8a shows how the concentration of substrate [S], free enzyme [E], enzyme-sub-strate complex [ES] and product [P] varies with time. The [ES] complex will after a brief initial period reach a steady state, in which the complex will be consumed approximately as rapidly as it is formed and gives d[ES]/dt ≈ 0. Note that the amounts of E and ES are exaggerated for clarity and that the total enzyme concentration and [ES] declines slowly as [S] is consumed, while [E] rises accordingly.

19

Figure 8. (a) The steady state in enzyme kinetics. (b) Effect of [S] on initial velocity of an enzyme catalyzed reaction.

Some assumptions must be made to simplify the mathematical treatment of the kinetics which leads to Equation 1: as long as the reaction is measured during conditions where the product formation has a linear time dependence (the initial reaction velocity), the formation of the ES-complex is constant which gives d[ES]/dt = 0. It is also important to consider the enzyme concen-tration where the total enzyme concentration includes both the free enzyme molecules and the enzyme molecules in the ES-complex ([E]tot = [E]free + [ES]). When considering the substrate concentration, the assumption that the substrate concentration is far greater than the enzyme concentration ([S]>>[E]) must be made and also that the [S]free ∼ [S] initially added. Meas-urements carried out with constant enzyme concentration and several substrate concentrations allow determination of the parameters described below. Under these conditions, Equation 1 leads to the relation shown in Figure 8b, where the initial velocity is plotted as a function of substrate concentration.

Vmax

Vmax/2

[S] = KM

Vel

ocity

, v

Substrate concentration, [S]

[P][S]

[E] [ES]

[E]t

Con

cent

ratio

ns

Time

Steady state: ES almost constant

StES al

Pre-steady state:

ES forming

(a) (b)

20

KEY PARAMETERS KM: is the substrate concentration where the reaction has attained half of its maximum velocity. At KM, half of the enzyme molecules bind to substrate molecules. KM is thus a measure of at what substrate concentration an effective catalysis occurs. A high KM indicates that the enzyme is optimized for a high substrate concentration whereas an enzyme with low KM is optimized for a low substrate concentration. It can be noted that the KM often correlates with the expected substrate concentration in nature. Vmax: at “high” substrate concentrations where [S] is much greater than KM the reaction approaches Vmax. That is the maximum velocity the reaction can occur at because every enzyme molecule is occupied by a substrate and busy con-verting the substrate into product. kcat: is a measure of the enzymes turnover number, which means that this con-stant tells us the number of substrate molecules converted by one enzyme mol-ecule per time unit. Indirectly, it tells us the time an enzyme needs to turn over one substrate molecule to one product molecule.

21

INHIBITION

“Inhibition might be due to adsorption by the enzyme of the inhibitory substance, thus preventing the adsorption of one or both of the substrates”

Dixon & Thurlow, 1924 The idea that an enzymes function can be inhibited is as old as the first sys-tematic studies of enzymes. Brown21 noticed when he studied the enzyme in-vertase that it was suppressed by its own product. A couple of years later Brown discovered that structures similar to the product of invertase also sup-pressed the enzyme. This study combined with other studies led to the quote above. The inhibition of enzymes can either be reversible or irreversible. A reversible inhibition involves noncovalent binding of the inhibitor. In an irreversible in-hibition there is a covalent bond between the enzyme and the inhibitor itself. The reversible form of inhibition has been studied in this thesis and the various modes of reversible inhibition has traditionally been divided into competitive, noncompetitive and uncompetitive inhibition and they differ by the mecha-nism by which the activity is suppressed. The measurement of the initial rate of several substrate concentration at several fixed inhibitor concentration gen-erally gives a straight line in a double reciprocal plot (1/v vs 1/[S]). The inhi-bition type can be distinguished by the pattern of these lines (Figure 9). The inhibitors studied in this thesis suppress the serine proteases by competitive inhibition and thus compete for the same active binding site on the enzyme.25

Figure 9. Lineweaver-Burk plots for enzyme inhibition.

1/v

1/[S]

Inhibitor

No inhibitor

Slope: change Y-intercept: same

[S]

Slope: same Y-intercept: change

Slope: change Y-intercept: change

Competitive inhibition KM increased

Vmax unaffected

Uncompetitive inhibition KM reduced Vmax reduced

on Noncompetitive inhibition KM unaffected Vmax reduced

22

KEY PARAMETERS FOR INHIBITION The potency of an inhibitor is related to the concentration of an inhibitor needed to yield a certain extent of inhibition. The potency of an inhibitor is generally evaluated as an inhibition constant, Ki, or in pharmaceutical studies IC50, the concentration giving 50% inhibition under certain selected condi-tions. The relationship (Equation 2) between these two constants can be de-scribed as follows:

(Eq. 2)

Ki is a dissociation constant as well as the concentration of the inhibitor where, if [S]<<Km (Equation 4 and 5), the reaction rate is half of the uninhibited re-action rate. In the case of competitive inhibition, the complex of the enzyme and the inhibitor is inactive and represents the inhibitory effect. Thus, a high Ki value points to a weak inhibitory effect since the complex tends to fall apart easily leading to a more normal function of the enzyme. The lower the Ki value, the stronger inhibitory effect due to that the inhibitor binds more tightly and the amount of available enzyme will be small. A strong and effective in-hibitor will have a Ki value that is considerably lower than the KM value for the substrate. In many cases one can perform a direct measurement between the enzyme and the inhibitor that will result in a Kd value that is supposed to correspond to the Ki value obtained in inhibition experiments. The Ki value can be described below according to Equation 3.

(Eq. 3)

23

COMPETITIVE INHIBITION

Figure 10. Illustration of the substrate (pink) and the inhibitor (black) competing

over the same binding site on the enzyme. A molecule that can bind in the enzymes active site and thus prevents the sub-strate from binding is a competitive inhibitor (Figure 10). During the time that the inhibitor is occupying the active site the enzyme is thus unavailable for catalysis. As can be seen in Equation 5, when the substrate and the inhibitor compete over the same binding site the apparent KM will increase. When [S] becomes larger it will decrease the possibility of an inhibitor binding to the active site which will lead to that the maximum velocity is unchanged when [S] is >> KM

app (Eq 4).

(Equation 4)

(Equation 5)

A linear Lineweaver-Burk plot can be expected since the system still follows Michaelis-Menten kinetic at a given [I]. The inhibition factor of a competitive inhibitor is a multiplier of the slope in a reciprocal equation where the lines converge on the y-axis (Figure 9). If several KM

app values at different [I] are obtained the true KM and Ki can be determined according to Figure 11 or by solving Ki from Equation 5.26

Figure 11. KM and Ki can be determined if KMapp is measured repeatedly at different inhibitor concentrations.

[I]

KM

app

Slope = KM/Ki

KM

24

TIGHT BINDING INHIBITION The equations for calculating Ki are dependent on the assumption that the in-hibitor concentration is much higher than the enzyme concentration, meaning that the inhibitor concentration is not significantly changed by addition of the enzyme. Regarding tight binding inhibitors, the Ki value can become so low that it is required to use very low concentration of the inhibitor to be able to detect any differences in the inhibition. The concentrations of the enzyme and the inhibitor are thus in the same range and, in that case, conventional equa-tions can no longer be used. The equations describing tight binding must take into account that [I]0 = [I] + [EI], and the same for the enzyme concentration, meaning that the free inhibitor concentration cannot be approximated by the total inhibitor concentration.27 A scheme describing the simplest kinetic model for tight binding can be seen below:

The reason that the steady state equations is not applicable is due to that (i) the enzyme concentration is in the same magnitude as the inhibitor concentra-tion and (ii) thus at low concentrations of inhibitor. The measurements can be complicated if the binding process does not come to equilibrium within a time frame of the initial rate measurements. The kinetic experiment is best carried out by measuring the initial rate while varying [I]0. The Ki value can be estimated by the graphical method of Dixon28 where the fractional velocity of the enzyme reaction is plotted as a function of inhibitor concentration at a fixed substrate concentration. A more mathematical treat-ment of tight binding inhibitors was presented by Morrison29, which lead to a general equation that describes the fractional velocity of an enzymatic reaction as a function of inhibitor concentration, at fixed concentration of enzyme and substrate. The equation is referred to Morrison equation (Equation 6).

(Eq. 6) The equation of Ki

app varies with the type of inhibitor. For competitive inhibi-tors as previous been described for leupeptin and antipain30 the Ki value can be calculated from Equation 7:

(Eq. 7)

25

SERINE PROTEASE INHIBITORS

I have studied two different reversible types of inhibitors in this thesis, both full size proteins and smaller peptides. The first one is an inhibitor from

Solanum tuberosum, a competitive serine protease inhibitor named Potato Serine Protease Inhibitor (PSPI). The second group includes tight binding

inhibitors such as the peptide aldehyde antipain and leupeptin but also derivatives of leupeptin.

POTATO SERINE PROTEASE INHIBITOR Potato serine protease inhibitors (PSPI) are involved in many physiological processes by controlling protease activity. One important, and maybe the best example, is their role in wound-induced defense responses of plants caused by herbivores or pathogens.31,32 PSPI was first isolated from potato tuber (Solanum tuberosum) in 1963 by Balls & Ryan33 and are expressed in leaves, stems, flowers and tuber sprouts, regulated by environmental and development signals.34-36 As much as 20-50% of all water-soluble proteins in potato tuber are inhibitors to proteolytic en-zymes and these inhibitors ranges from 20 to 24 kDa.37 Studies have shown that these inhibitors can inhibit serine proteases,38-41 aspartic proteases41-43 and cysteine proteases.41,44,45 Based on the primary structure of these inhibitors they can be classified together with Kunitz-type soybean inhibitors of trypsin and since they origin from potato tubers they are denoted as potato Kunitz-type proteinase inhibitors (PKPI´s). PSPI is part of the PKPI family and PSPI´s sufficiently suppress digestive enzymes such as trypsin, α-chymotryp-sin and subtilisin.46,47,35 These inhibitors accumulate in the potatoe leaves and tubers due to mechanical wounding, lesion by insects, phytopathogenic mi-croorganisms39,45 and UV-radiation.38 About 40% of the total amount of water-soluble protein inhibitors from potato consist of PSPI´s making PSPI´s the largest group of inhibitors in the plant. PSPI is a 21 kDa large protein consisting of one large (16.5 kDa) and one small (4.5 kDa) chain. It is expressed as one polypeptide chain but post-trans-lational modifications (six amino acids are deleted) yields the two fragments that are held together by a disulfide bridge and non-covalent interactions. 41 Traditionally, a Kunitz-type inhibitor is a single headed protease inhibitor but PSPI is proposed to be a double-headed Kunitz-type serine protease inhibitor with two independent reactive loops.48 There have been five different sugges-tions of the active sites in this inhibitor and in 2012 Elisabeth Meulenbroek published an article where the crystal structure was determined leading to the identification of two reactive sites that can be found in the protruding loops

26

centered around Phe75 and Lys95 (Figure 12).49 PSPI binds tightly to the ac-tive site of the protease and is mimicking the substrate.

Figure 12. Structure of PSPI with the two suggested active sites marked in blue.

PDB ID 3tc2.

PEPTIDE ALDEHYDES Protease inhibitors play an important part in the market today and used in drugs that treat a variety of diseases. One class of compounds that have re-ceived a lot of attention are peptide aldehydes. They act by mimicking the tetrahedral transition state of a peptide bond hydrolysis, through the formation of a tetrahedral hemiacetal with the active site serine residue.50

LEUPEPTIN Leupeptin is a 426 Da large peptide isolated from Streptomyces species with the structure acetyl-L-leucyl-L-leucyl-argininal (Figure 13) where the alde-hyde group of the L-argininal residue is important for the strong inhibition.51,61 It is a potent inhibitor towards serine proteases (trypsin, plasmin, kallikrein and acrosin) and cysteine proteases.51,52 Leupeptin can inhibit proteases that are involved in a wide range of biological processes which includes coagula-tion,53 blood pressure regulation,54 degradation of proteins,55 cell prolifera-tion,56 virus assembly and tumor generation.57,58 The inhibitor is often used as a biochemical tool when proteolytic enzymes are studied e.g. in their role in biological functions.53-60 Due to its diversity leupeptin can cause responses in a variety of systems depending on the proteases involved.

Phe75 Lys95

27

Figure 13. Leupeptin in its natural state (acetyl-L-leucyl-L-leucyl-argininal).

Leupeptin can exist in three forms when it is dissolved in aqueous solution and these are the hydrate, the aldehyde and the cyclized carbinol amine (Fig-ure 14). The equilibrium between these forms is slow and has been docu-mented with 2D TLC, 1H NMR and slightly broadened 13C NMR signals cor-responding of each of the three equilibrium forms. The rate of conversion be-tween these three forms has not been established.62,63

Figure 14. The equilibrium forms of leupeptin It is previous reported that leupeptin acts as a reversible inhibitor51 and binds to the target enzyme with a such high affinity that the amount of free inhibitors is significantly decreased as the EI-complex forms. Because of this leupeptin falls into the category tight binding inhibitors. Leupeptin acts as a transition state analogue and makes a good inhibitor be-cause it binds to the enzyme more tightly then the substrate. Difference maps showed strong continuous electron density between aldehyde carbon of leupeptin and the hydroxyl group of Ser195, confirming a covalent complex between leupeptin and trypsin.64 The position of the hemiacetal oxygen in the tetrahedral complex was believed to be positioned in the “oxyanion hole” but Kurinov64 showed through a crystal structure between trypsin and leupeptin that the oxygen atom can be positioned to face the “oxyanion hole” and form hydrogen bonds with the main chain amides NH195 and NH193, or it can point to the active site residue His57. In the refined structure (leupeptin-tryp-sin complex) the hemiacetal oxygen positions that it forms a strong hydrogen bond to His57 which stabilizes the whole structure.

28

ANTIPAIN Antipain was found in various species of Actinomycetes by Umezawa in 1972. Antipain is a potent tight-binding inhibitor that resembles leupeptin in its ac-tion.52 However, its plasmin inhibition is weaker and its cathepsin A inhibition is stronger.65 Antipain has an L-argininal residue at the C-terminal structure, similar to leupeptin (Figure 15), and is a strong inhibitor for serine proteases.66

Figure 15. Antipain in its natural form (1-carboxy-2-phenylethyl)-carbamoyl-L-arginyl-L-valyl-arginal

KINETICS OF PEPTIDE ALDEHYDES For both Leupeptin and Antipain, the steady state approximation model and the use of double reciprocal graphics are not valid since these inhibitors are tight-binding inhibitors. In the steady state approximation model both sub-strate and inhibitor are usually in a large excess relative to the enzyme, mean-ing that the free concentration of these can be set equal to the total concentra-tion. On the contrary, for tight binding inhibitors under equilibrium conditions one will obtain depletion of the non-bound form of the inhibitor. Because of this, these inhibitors require another method of analysis. The Ki value for tight binding inhibitors is determined from the fractional velocity of an enzymatic reaction as a function of inhibitor concentrations, at fixed concentrations of enzyme and substrate, as can be seen in the section regarding tight binding inhibition.

29

CONJUGATION

The chemical technique that couples two molecules together in a covalent bond, in which at least one of them are a biomolecule, is usually referred to

as bioconjugation. The conjugation itself can dramatically change the physiological/functional properties and the practical usefulness of the biomolecule. The possibility of creating these bioconjugates has affected research profoundly and is exten-sively employed in academia, industry and developments in life science. As new techniques are developed, new unique bioconjugates can be formed lead-ing to novel applications being advanced. As of today, there is an enormous amount of different ways that can lead to conjugation. Choosing a strategy of conjugation is largely dependent on the target and the starting point. The prep-aration of bioconjugates is an ongoing field in organic, bioorganic67,68 and pol-ymer chemistry69,70 resulting in a large diversity of methods and approaches. The large size, diverse biological functions, specific conformation and bio-compatibility makes proteins attractive target for conjugation. However, they also come with shortcomings such as poor stability, low solubility and shorter half-lives in vivo. By the usage of a suitable polymer these limitations can be overcome.71,72 An early example of covalent linkage between soluble poly-mers and proteins was reported by Abuchowski73 in 1977 and since then the amount of conjugations has increased enormously. The conjugation of a pro-tein and a synthetic polymer cannot only enhance solubility, stability and bi-ocompatibility but also with respect to the polymer provide biofunctional properties such as enzymatic activity or inhibitory function carried by the pol-ymer. The preparation and the synthesis of the protein-polymer conjugates have ex-tensively been reviewed in the literature over the last 20 years.70,73-77 In the conjugation process of proteins there are a number of possible conjugation sites. The reactivity of proteins depends on the side chains in their spatial structure. The most reactive ones are γ-carboxyl group of glutamic acid, β-carboxyl group of aspartic acid, thiol group of cysteine, ε-amino group of ly-sine, and the hydroxyl groups of serine, where the thiol group is the most nu-cleophilic one followed by the amino and hydroxyl groups. The conjugation of PSPI (paper III) uses the ε-amino group of lysine by reductive amination in a Schiff base formation. Other important reactions of ε-amino group are acyl-ation (formation of amides with active esters) and reaction with squaric acid esters.78

30

In the design of a protein-polymer conjugate it is important to consider the following factors: the protein used, the choice of polymer, the type of conju-gation chemistry used, the stoichiometry, the conjugation site and the resulting architecture of the conjugate.71 The design of biomaterial is important due to the key role it’s playing in influencing bioactivity and construct performance. It is important to consider the impact and suitability of the conjugate before functionalizing the device and starting the production.79 The key factors that need to be considered are site-selectivity for conjugation, efficiency, rate of reaction, ability to provide biomolecule pattern and accessibility of reaction partners.80 It has been previously shown that the surface orientation of the conjugation is crucial in order to maximize bioactivity. The strategy of func-tionalize biomolecules can be divided into two broad categories: the target of a specific site for modification and the target of multiple groups generally. The crucial role of site-selective conjugation determines the biological activ-ity, efficiency, ease of use, and reproducibility.80 There are three key covalent methods by which bioconjugation can be achieved, namely chemical conjugation, enzymatic conjugation and photo conjugation. In this thesis I have focused on chemical conjugation and this is also the most widely studied and applied method for creating bioconjugates. By a chemical conjugation the conjugate can be created smoothly and gener-ally does not require any advanced equipment. Most examples of conjugation involve molecules with amino groups. This type of coupling can be used to conjugate nearly all proteins or peptides due to lysine residue of proteins and the N-terminus of peptides and proteins. Due to the amine group, this is perhaps the simplest reactive handle for conjuga-tion. There is an extensive literature on their use, numerous protocols for their modification and a wealth of commercial reagents and kits for conjugation with minimal difficulties.81-83 The modification of amines proceeds either through acylation or alkylation. These reactions are mostly rapid in their for-mation of a stable amide or secondary amine bond and result in a high yield. The drawback of using amino groups is the low site selectivity in the conju-gations of proteins.80

31

AMIDE FORMATION Amide formation is widely studied and used extensively in synthetic chemis-try. Most methods have been developed for organic solvents but when it comes to aqueous solutions, that are often required in bioconjugation, the op-tions are greatly reduced. There are a number of different approaches to func-tionalize biomaterial, and that includes 1,1´-carbonyldiimidazole (CDI) that forms amine-bridging carbamate linkages rather than amide84,85 and the amide formation can also be performed by the generation of NHS-esters in situ in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and cou-pled directly to the target.82,86 The amide bond between the inorganic particles and inhibitors in paper III and IV is created by the use of CDI activation. In the first step of coupling to in-organic oxides the hydroxyl groups of the particle are modified with 3-trieth-oxysilylpropylamine (APTES) (Figure 16). This reagent reacts readily with OH-groups on the surface and acts as a spacer and increases the flexibility of the peptide that will be attached later on. The reaction takes place in dry ace-tonitrile since the silanes have a tendency to polymerize in aqueous environ-ment.

Figure 16. By using a base to activate the oxygen on TiO2, it can act as a nucleophile and attack the partially positive silicon of APTES forming

a covalent bond between the oxygen and silicon while a deprotonated ethanol group is leaving. (R = (CH2)3NH2).

The amino groups are then activated with 1,1´-carbonyldiimidazole (CDI) (Figure 17). The nucleophilic amino group of the silanized surface will attack the electrophilic carbon of CDI due to its two good leaving groups. The major advantages by using CDI is that after silane modification the activating group is relative resistant to hydrolysis. The drawback of this type of activation com-pared to others is that the coupling time will increase, and the reaction must take place in a dry solvent so the carrier of interest must be able to handle that type of environment.

32

Figure 17. The amine group on APTES performs a nucleophilic attack on the partially positive carbon of CDI forming a covalent bond between the carbon

of the carbonyl group and nitrogen. The leaving group here is a imidazole.

(R´ = ) In the third and final step the peptide of interest is conjugated by allowing its primary amine to react with the activated groups of the surface of the particle (Figure 18). The peptide is coupled through a substitution resulting in a chem-ically very inert urea group. Optimum pH of this reaction is between 9.0-9.5 as the amino groups needs to be in a deprotonated state to function as a nucle-ophile.

Figure 18. In the last step the amino group of the peptide attacks the electrophilic carbon terminating the conjugation by forming the covalent bond meanwhile the

other deprotonated imidazole leaves.

(R´´= ).

REDUCTIVE AMINATION In paper III I have used reductive amination to create the conjugation between PSPI and oxidized dextran (Figure 19). An amine can react with an aldehyde to form an imine moiety by elimination of a water molecule. The equilibrium is fast, but the unconjugated starting materials is strongly favored due to the high concentration of water present. The imine can, however, be irreversible reduced by the use of NaBH4 or NaBH3CN. The use of NaBH3CN is some-times preferred due to its selective reduction of imines even in the presence of unreacted aldehydes.87 Stronger reducing agents, e.g. NaBH4, can also reduce

33

the unreacted aldehydes. Reductive amination can be used for the conjugation between sodium periodate-treated moiety and its corresponding amine moiety. In paper III both NaBH4 and NaBH3CN where evaluated. In comparison to alkylation that can happen multiple times on an amine, imines can only be formed once. When the C=N bond is formed (the imine) it can be reduced which yields a new alkyl group attached. This is a method that is much more controlled in forming new nitrogen-carbon bonds.

Figure 19. In the first step an imine is formed which under acidic

conditions will be protonated giving its conjugated acid (iminium ion). In the presence of a reducing agent the iminium ion will be reduced to give

a new secondary amine.

34

CARRIERS When selecting a support for the biomolecule it is important to consider the following:

- Mechanical properties - Physical form - Resistance to chemical and microbial attacks - Hydrophilicity - Permeability - Price and availability

In general, depending on the type of carrier and the target to be conjugated, pretreatment of the carrier resulting in a chemical modification is often neces-sary. The modification can either act as a spacer to provide spatial availability or it can modify the functionality of the target or do both. The approach in both paper III and IV has been to first activate the carrier,

then conjugate the desired peptide or protein. I wanted to evaluate the inhib-itors on different types of carriers. The carriers I have used have different

mechanical and physiological properties and I have also used different con-jugation chemistries in order to obtain my conjugates.

INORGANIC PARTICLES An inorganic particle that is produced in dimensions of 1-100 nm has proper-ties that non-nanoscale particles do not share. Over the last decade the use of nanotechnology has increased enormously in biomedical science. The appli-cations of nanoparticles are mostly found in diagnosis and treatment of dis-eases where targeted drug delivery has made a large impact in safe pharma-cotherapies for complex diseases.88 I have designed peptide and protein con-jugates with the use of titanium dioxide and zinc oxide as inorganic carriers. Many of the inorganic particle supports, in particular the oxides that I have been using, have a surface that is rich in hydroxyl groups which are used here as the handle for conjugation.88

Titanium dioxide Titanium dioxide (TiO2) is a naturally occurring compound. The source of the titanium used as a carrier in this thesis is a mixture of rutile and anatase, that are naturally occurring minerals. It consists of a nanopowder with a particle size less then 100 nm. TiO2 assumes a tetragonal crystal system (Figure 20).

35

Figure 20. Ball and stick representation of the crystal structure of rutile where the

oxygen atoms can be seen in red and titanium atoms in black. About 80% of the total consumption of titanium dioxide in the world is found in paints, varnishes, paper and plastics but it is also used in cosmetic products as sunscreen, pigment and thickener. Titanium dioxide protects the skin from ultraviolet light due to its ability to strongly scatter and absorb the light. Fur-thermore, titanium dioxide is also studied for its photocatalytic properties.89-91

Zinc oxide Zinc oxide (ZnO) assumes a tetrahedral coordination geometry which leads to a Wurtzite crystal structure (Figure 21). Zinc oxide used here is a white pow-der with a particle size less then 100 nm which is insoluble in aqueous solu-tions.

Figure 21. Ball and stick representation of the Wurtzite crystal structure where the zinc atoms can be seen in green and the oxygen atoms in grey.

ZnO is widely used in the world for a various materials and products,92 for example in plastics, glass, ceramics, pigments, foods, batteries etc. as addi-tives. Another area is ointments and creams since zinc itself has antibacterial action. The fine nanoparticle with an increase surface area enhances this prop-erty.93 Zinc oxide is used in ointments and barrier creams to treat diaper rash, dermatitis, itching, acne. It is also used in sunscreen lotions since ZnO blocks UVA (320-400 nm) and UVB (280-320 nm) rays of ultraviolet light. It is con-sidered as nontoxic, nonallergenic and nonirritating.

36

POLYSACCHARIDES Carbohydrates are an enormously large group of biomolecules with an exten-sive combination of monomer combinations and branching patterns. In the last two decades significant advances have been made in the development of car-bohydrates that are both biocompatible and biodegradable polymers. I have worked with two different kinds of polymers, namely agarose in a bead struc-ture and soluble dextran.

Gel beads The gel beads, namely WorkBeads™, are made of agarose, a linear polysac-charide with repetitive units of agarobiose (Figure 22), generally extracted from red seaweed. The polymer is made up by alternating units of D-galactose and 3,6-anhydro-L-galactopyranose linked together by a α-(1→3) and β-(1→4) glycosidic bonds. A single agarose chain can include up to 800 mole-cules of galactose and the chains form helical fibers that can aggregate into a supercoiled structure and can further be formed into porous beads and resins of varying fineness. These beads are generally highly porous, soft and easily crushed so it is necessary to use mild conditions. The strength of the resin can be improved by increasing the crosslinking, but this may also lead to lower binding capacity in e.g. conjugation of peptides and proteins.94

Figure 22. The repeating unit of agarose.

The resins used in paper IV are preloaded containing bromohydrin groups suitable for coupling of ligands containing a nucleophilic primary amine (Fig-ure 23). The reaction takes place under ambient temperatures and is a simple and reliable coupling procedure that results in a stable covalent linkage. There is no need for any additional reagents during the coupling procedure.

Figure 23. The reaction scheme for coupling the activated resin with a primary amine from paper IV.

37

Soluble polymer Dextran is a natural polysaccharide biopolymer (complexed branched glucan, derived from the condensation of glucose) that is widely used as a carrier in drug delivery systems. Dextran is synthesized from sucrose by a lactic-acid bacterium and consists of main chains from α-(1→6)-linked D-glucose units (Figure 24). The branching and the ratio of linkage all depends on the type of bacterial strain used and occurs in α-(1→2) α-(1→3) and/or α-(1→4). The solubility of dextran depends on the degree of branching. A branching over 43% through α-(1→3)-linkage makes the dextran almost insoluble in water, whereas a linear linkage of 95% makes it water soluble.95 Dextran is available in a wide range of molecular weight ranging from 3 kDa up to 2 000 kDa and is used for many different purposes. For example, it is used as a lubricant in eye drops and as a volume expander in intravenous solutions where it was launched by Pharmacia in Uppsala as a blood plasma replacement (Macro-dex®). It is also used as a size-exclusion chromatography matrix (Sephadex™, another product from Pharmacia), used as coating on nanoparticles to increase the biocompatibility and has also been immobilized in biosensors. The bene-fits of working with dextran is that it is neutral and water soluble, easy to filter, biocompatible, biodegradable and has been shown to be stable for over five years.96

Figure 24. The repeating unit of dextran with branching point indicated. Before conjugation can occur, a chemical derivatization is often required due to that hydroxyl groups are generally unreactive in water due to their similar pKa value as water. A common one is selective oxidation of vicinal diols by periodate.97

Periodate oxidation Periodate oxidation is an efficient way to transform the relatively unreactive hydroxyls of sugar residues into reactive aldehydes. Periodate cleaves car-bon–carbon bonds that possess adjacent hydroxyls, oxidizing the OH groups to form highly reactive aldehydes. For the oxidation to occur it is required that the OH groups are oriented in equatorial-equatorial or axial-equatorial position. The oxidation cannot take place if they are positioned axial-axial since the intermediate cannot form. The oxidation typical occur between C3-

38

C4 or C3-C2 where a second oxidation typically occurs as well (Figure 25). The highly reactive aldehydes formed can easily be attacked by a neighbor-ing hydroxyl group, leading to hemiacetal formation.98 However, several studies restrict these formations to a narrow pH window (4.0-5.2).99 Another consequence of this dextran oxidation (paper III) is the decrease in the aver-age molecular weight.100

Figure 25. Possible sites of periodate oxidation of Dextran’s α-1,6 glucose residue. (A) Periodate attack at C3-C4, (B) C3-C2 and (C) double oxidation.

There are a number of different coupling methods regarding conjugation to dextran that are frequently used. Most of these are however best suited for activation of dextran gel beads. Before I decided on the procedure based on oxidation of the hydroxyl groups of dextran with periodate, a number of dif-ferent methods were experimentally evaluated. By using periodate method, the risk of crosslinking within dextran itself was virtually eliminated. The al-dehyde groups thus formed will, contrary to e.g. CDI-activation, not give rise to direct dextran-dextran crosslinking. The other methods mentioned below were less successful than the periodate method in the conjugation of PSPI in Paper III.

Other methods evaluated - Organic solvent mediated carbonyldiimidazol method: The hydroxyl

groups of dextran were activated by carbonyldiimidazole following attachment of the molecule of interest by using an aminogroup.101

- Aqueous 6-bromohexanoic acid carbodiimide method: Similiar to CDI-activation, but addition of 6-bromohexanoic acid before attach-ment of the molecule of interest by using an aminogroup.102

- Cyanogen bromide method: Dextran was activated by CNBr followed by addition of the target to be coupled. Amine containing drugs or proteins can be conjugated to dextran activated by cyanogens bro-mide.103

39

TOOLS SECTION In order to achieve the results that I present in my papers I have needed to use a number of different instruments and techniques. In paper II you can read about the interaction between serine proteases and the inhibitor from potato (PSPI) and to be able to study this I have used Static Light Scattering. This technique was also used in paper III to be able to confirm my conjugate be-tween dextran and PSPI. In my last paper presented here (paper IV) I have used solid-phase peptide synthesis in order to obtain my two peptides and mass spectroscopy to confirm that my peptides were the peptides I aimed for. Two different mass spectroscopy instruments were used and are presented in this section.

I hope this will provide you with useful information so we can analyze and

understand the results.

40

STATIC LIGHT SCATTERING (SLS)

Atoms or molecules that are exposed to light will absorb light energy and reemit light in different directions with different intensity.

This is called scattering of light. Most macromolecules or macromolecular complexes are soluble and a con-siderably amount of information can be obtained from studying the scattering of radiation from a solution. There are three types of radiation that are mostly used for that type of experiment: visible light, X-rays and neutrons. In this thesis, visible light will be discussed and to break it down, let us start with one molecule: Static light scattering is a technique that measure the intensity of the scattered light to obtain an average molecular weight (MW), size, conjugation or inter-action of a protein(s) or a macromolecule(s) like a polymer in solution.

The intensity from static light scattering depends on:

- the molecular weight of the molecules. - the concentration of the molecules. - the size of the molecules. - the refractive index of the pure solvent. - the refractive index of the suspended molecules.

The light is usually monochromatic, achieved by using a monochromator or laser, and also linearly polarized. The light is directed along the x axis and polarized with its electric vector in the z direction. The wavelength, λ, is in the following assumed to be so long that the molecule is put at its origin (x=0, y=0, z=0), without having to worry about its size (Figure 26).

Interaction Conjugation Size Molecular weight

41

Figure 26. Scattering of linearly polarized radiation by a molecule. Radiation

scattered in this direction is polarized in the plane defined by the z axis and r.104 The energy that the molecule absorbs will produce corresponding oscillation of the electrons within the molecule. The molecule will act as an antenna and disperse the radiation in other directions then the direction of the incoming energy. The scattered radiation occurs from the oscillating dipole moment and when the molecule is much smaller than the wavelength of the incident radia-tion this is called Rayleigh scattering (Rθ). The intensity (iθ) of the scattered radiation is compared to the intensity (I0), taking into account geometrical pa-rameters of the instrument, of the incident radiation. The derived quantity is named Rayleigh ratio (equation 8).

(Eq. 8) where r is the distance to the detector, v is the scattering volume and θ is the angle with respect to the incident beam. Refractive index of a material or solution is a measure of the light velocity through the material or solution. It is defined as: , here c is the speed of light in vacuum and v is the phase velocity of light in the medium. The differential refractive index detector (dRI) determines the concentration by the change of the refractive index (n-n0) and in a solution it is usually very small. dn/dc is basically the difference between the refractive index of an analyte and the buffer. If we for example would have a mixture of n macromolecular substances, of different molecular weights Mi and concentrations ci, we can measure the total intensity of scattering because that should be the sum of scattering from all components. The relation between molar mass M and the scattering power can be used in the following expression for one molecular component:

(Eq. 9) And by the use of refractive index:

42

(Eq. 10) The ratio can be expressed as:

(Eq. 11) And final rearrange it to:

(Eq. 12)

M is the molecular weight of the analyte, Rθ is the Rayleigh ratio, is the refractive index increment of the analyte and A is an instrumental constant. For modern instruments (Figure 27) a laser is usually used to provide mono-chromatic light. The intensity of light scattered at 90° is usually measured to minimize signal to noise ratio.

Figure 27. Set up of a modern light scattering instrument. The combination of size exclusion-chromatography, light scattering, UV280 and dRI makes it a versatile and reliable method for MW determination.105-111 To determine proteins and other biomacromolecules biophysical properties in solution it is necessary to add an absolute and independent characterization downstream the separation step. In paper II the combination of size exclusion chromatography and static light scattering is used in order to determine the stoichiometry for protease-potato serine protease inhibitor complexes.

To summarize, light scattering is a powerful tool for the determination of

particle weight and dimensions.

polarization

laser scattering volume

detector

43

SOLID-PHASE PEPTIDE SYNTHESIS (SPPS) Solid-phase peptide synthesis is a method which uses a solid support material where the growing peptides are covalently anchored and synthesized by addi-tion of amino acids in repeated cycles. The benefits of using SPPS compared to synthesis in solution are:

- High efficiency and throughput. - Increased simplicity and speed.

The interest in synthesizing peptides is largely due to the peptide’s character-istics such as high bioactivity, high specificity and low toxicity. By using SPPS in the synthesis of peptides, a high yield can be obtained, as each reac-tion step can be driven to completion.112-116

Figure 28. The N-terminal group is first deprotected. The next amino acid is

activated at the C-terminal end by a coupling agent, which facilitates the peptide bond formation. The N-terminus of the growing peptide is then deprotected and the next amino is coupled. The cycle is repeated until the synthesize of the full-length

peptide is complete. The general principle of SPPS is straightforward. The synthesis is carried out on a solid support, resin, normally in the shape of a spherical beads with the active group attached. The first amino acid (the amino acid at the C-terminus in the sequence) is attached to the solid support and new amino acids are then attached step by step. Each cycle includes deprotection, activation and cou-

Deprotection

Deprotection Coupling

Next amino acid (prior to coupling)

N C

R1

N

R2

R2

N

R1

C

R1

C

44

pling steps. The solid-phase approach allows chemicals and solvents to be eas-ily removed by filtration and washing without any significant loss of the grow-ing peptide. As a last step the peptide is cleaved from the resin and further purified and characterized in solution (Figure 28). The synthesis is carried out in a single reaction vessel which thus leads to minor loss of product due to no transfers or exchange of vessels.117,118 The support resins that are used here need to fulfill certain properties. They need to be stable against mechanical mixing, different temperatures and sol-vents, they need to have a certain narrow range of size and also high swelling properties. The resins can also be either pre-loaded, with the first amino acid bound, or initially unloaded. The use of preloaded resins that include the initial amino acid of the sequence is generally preferred.119-122 A new peptide linkage is created between the activated carboxylic group of a new amino acid and the terminal amino group of the already existing growing peptide. During this step it is crucial to protect any other functional groups. The protective groups differ due to which methodology that is used (Fmoc or Boc). The peptide synthesis presented in paper IV uses the base labile Fmoc-methodology.123,124

Figure 29. Illustration of the final step where all protective groups are removed,

and the peptide is cleaved from the solid support giving the final peptide. In the final step, the cleavage of the peptide from the resin, a cleavage cocktail is added to the resins (Figure 29). This step is the most important one in SPPS since the peptide is exposed to a series of competitive reactions. It is important to use the appropriate reagents and the right conditions for the reaction, other-wise undesired reaction such as damaged peptide or irreversible modifications

C

R1

N

R2

R3

Rn

45

can occur. The goal is to split the peptide from the resin while also removing protective groups from the functional groups of the side chains on the amino acids. When using Fmoc chemistry, the peptide-resin is usually treated with 95% TFA while shaking the vessel gently for 1-3 hours. The peptides presented in paper IV were synthesized by the use of SPPS and

purified with preparative HPLC and the molecular weight of the peptides were confirmed with MALDI-TOF mass spectroscopy. Information on struc-

tural properties can be obtained with nuclear magnetic resonance (NMR) and fourier transform infrared spectroscopy (FTIR). Circular dichroism

(CD) can be used to obtain information about the conformation and possible secondary structure, e.g. alpha helix or beta sheets of polypeptides. How-ever, since I only synthesized tripeptides in this context, CD was not ex-

pected to provide any useful information.

46

MASS SPECTROSCOPY (MS) Mass spectroscopy is a technique that can quantify known materials, identify unknown compounds within a sample and elucidate the chemical structure of different molecules. The technique measures mass-to-charge-ratio of ions and the signal intensities of detected ions are often plotted in a mass spectrum as a function of the mass-to-charge- ratio.

Figure 30. Components of a Mass Spectrometer

The first step of the analysis includes production of gas phase ions of the com-pounds. The ions are later separated in the mass spectrometer by their mass-to-charge ratio (m/z) and detected in proportion to their abundance. The typi-cal instrument (Figure 30) used usually consists of four components:

- Ion source: for the production of gas phase ions.- Analyzer: separates the ions by their m/z ratio.- Detector system: detecting the ions in proportion to their abundance. - Data system: processes the signals from the detector.

The detection of the peptides discussed in paper IV was done by the use of two MS-systems, which are discussed further below.

MALDI-TOF MALDI-TOF was used in the detection of my synthesized peptides in paper IV. MALDI is abbreviation for “Matrix Assisted Laser Desorption/Ioniza-tion”. This technique uses a laser energy absorbing matrix as an ion source to create ions from larger molecules with minimal fragmentation. The mass an-alyzer is the time-of-flight (TOF) analyzer.

The principle of MALDI The peptides analyzed in paper IV were embedded in a matrix and deposited on target made of a conducting metal. A brief laser pulse (either UV or IR can be used) irradiates the spot which becomes vibrationally excited and thus rap-idly heated. Down to 100 nm from the outer surface of the matrix will vaporize while carrying the analytes that also become protonated, into the gas phase.

Data System

Ion Detector

Analyzer Source Inlet

Mass Spectrum

Data Output

Gas Phase Ions ns Ion sorting g Ion detection

Vacuum Pumps Sample Introduction

47

The most common ionization is the protonation of the analyte with one single positive charge (Figure 31).125

Figure 31. Ionization of analytes by MALDI

The principle of time-of-flight (TOF) The usage of time of flight leads to that the ions enter the field-free region at the same time, or at least within a short time interval. The ions are accelerated by an electric field and the time it takes for the ions to travel to the detector is measured. If the ions have the same charge, their kinetic energies will be iden-tical, and the time of flight will depend on the mass (Eq. 13). The lighter ions will then arrive earlier than the heavier ones at the detector (Figure 32).126,127

(Eq. 13)

Figure 32. A general illustration of a liner TOF analyzer.

Source Region Field Free Region (Analyzer)

Detector

d V Target Accelaration Grid

+

+ +

Laser beam

Desorption Desolvation To Mass analyzer

Analyte spots Matrix spots

+ +

48

LTQ ORBITRAP VELOS PRO – HYBRID ION TRAP MS This MS-system combines a high-field mass analyzer and ion trap technology that deliver very high resolution, speed, sensitivity and complementary frag-mentation information. It was used when I detected the reduction and the ox-idation of leupeptin in paper IV.

The principle of electrospray ionization The ion source used here was electrospray ionization (ESI). It is a “soft tech-nique” since it gives very little fragmentation, meaning the molecules will re-main intact. The analyte is dispersed by electrospray into a fine aerosol by the use of high voltage. It is common to use volatile organic compounds (e.g. ac-etonitrile) since there will be an extensive solvent evaporation. After the drop-let is dispersed the solvent gradually evaporates forcing the charges in the molecule within each droplet closer together. At a certain point, when they are close enough, repulsion occurs (Coloumb force) causing smaller droplets to occur. This will repeat itself until all solvent has evaporized leaving only charged molecules (Figure 33).128,129

Figure 33. General illustration of ESI in positive mode. A jet of liquid droplets is emitted under high voltage. The solvent will progressively evaporate creating

more and more positive charged ions. The droplet will explosively evaporate when the charge exceeds the Rayleigh limit, leaving a stream of positive ions

to enter the mass analyzer.

The principle of an orbitrap mass analyzer The ions are trapped in an electrostatic field between an inner and outer elec-trode. The inner electrode confines the ions so that they orbite around it at the same time the ions change in the orientation of the rotational axis. This oscil-lation generates an image current in the detector plates and Fourier transfor-mation of these signals generates frequencies and corresponding intensities where the frequencies later can easily be converted into m/z values.130,131

Capillary

Evaporation Chamber -

-

- -

-

-

-

+++++++

++ + p+

+-

-

- -

-

+

++++

--

49

MY WORK

In the following four papers I will take you through the new evaluation method (Paper I), interaction stoichiometry between different serine proteases and PSPI (Paper II), conjugation of PSPI to different carriers (Paper III) and the synthesis and conjugation of my leupeptin analogues (Paper IV). The overall theme of this thesis is based on the idea that one can use inhibitors in different types of formulations, as for example lotions, wash creams or ointments, to treat and prevent dermatitis. The main cause of dermatitis is attacking prote-ases and if they could in some way be inhibited that would ease, and also prevent, the area that is or will be exposed for damage. Inhibitors were conjugated to investigate if their stability can be increased, if the conjugated inhibitors retain their inhibition properties and if the smaller peptides could be conjugated to larger carriers to eliminate the risk of the pep-tides crossing the human skin barrier. You can read about the conjugation of PSPI in paper III and the synthesis and conjugation of new small peptides in paper IV. The carriers in these papers were chosen due to their already known nontoxic and biocompatible properties. To be able to evaluate if these conjugates could be used in different types of formulations an application using gelatin was developed as a simulation for the human skin (Paper I). By integrating free inhibitors in a one-layer model and in a two-layer model I have studied how well the inhibition worked and if the properties of the inhibitor changed due to the gelatin layers. I have also used this evaluation method when I’ve evaluated my conjugates in paper III and IV, where it is called the gelatin erosion method. Furthermore, the interaction stoichiometry between different serine proteases and PSPI have been studied (Paper II) since earlier studies proposed that this inhibitor could be a double-headed inhibitor with the ability to bind two proteases simultane-ously.

50

PAPER I

KINETIC STUDIES OF SERINE PROTEASE INHIBITORS IN SIMPLE AND RAPID ‘ACTIVE BARRIER’ MODEL SYSTEMS

Diffusion through an inhibitor barrier

In order to evaluate my conjugates, I wanted to create some kind of barrier system to see if they could inhibit and function in the same way as the free

inhibitor. I wanted a simple setup that could be expanded to different setups. The idea of using gelatin came up and resulted in this paper published in

Analytical Biochemistry. The models were created to monitor the action of protease inhibitors in a layer between the “skin” and “attacking” proteases but also tested by application of selected proteases in the presence and ab-

sence of inhibitors in the gelatin gel. Gelatin is a convenient model system since it can simulate a barrier similar to the human skin. The epidermis outermost layer of the human skin consists of keratin, a fibrous structural protein which is similar to collagen. Collagen can be found in the second layer of the skin, dermis, which contains fibroblasts that produces elastin and collagen.132 Gelatin is an irreversibly hydrolyzed form of collagen which consist of 25% glycine, 25% proline and the rest is approximately equal amounts of alanine, aspartic acid and arginine. Proteo-lytic enzymes hydrolyze gelatin and destroy the gel structure.

ONE-LAYER MODEL The first model design was studied with and without inhibitors in one layer. Proteases were added and the rate of well formation expansion was studied where gelatin also served as a substrate.

Protease activity screening This simple procedure can be used in the estimation of protease activity in a sample or chromatographic fraction, since the one-layer model shows a corre-lation between increasing well area and protease activity allowing quantitative estimates (Figure 34).

AIM Design a double layer model

for active barrier systems

51

Figure 34. Protease activity screening, where the increased area after 24 hours with increasing enzyme concentration can be seen. At time = 0 hours the area of the well

corresponds to 28.3 mm2 in both cases. The influence of the gelatin properties was also investigated since the gel can be created with different Bloom numbers (gel strength) and is correlated with the average molecular mass. It shows that with increasing Bloom number the diffusion of the enzyme is slower. An increasing Bloom number correlates to the diffusion rate of an enzyme resulting in a slower well area expansion. So, the rate of forming wells will decrease with increasing gel strength. My ex-periment shows that no matter what Bloom number that is used, the general pattern and usefulness of the model is still the same, meaning it is independent of the basic gelatin properties.

Inhibition studies Addition of inhibitor in the gelatin gel resulted in a slower well area expansion compared to the area increase rate of the undisturbed enzyme. The rates de-cline to a plateau value after a certain inhibitor concentration (Paper I, Figures 4 and 5). Regarding the experiment with trypsin as an enzyme the asymptotic value declines to a vi/v0 value at 0.5 around 5 μM for leupeptin and a vi/v0 value at 0.2 around 15 μM for antipain. Chymostatin, on the other hand, showed no detectable inhibition. In the inhibition study, using α-chymotryp-sin, the rates declined to an asymptotic value of 0.7 for 10 μM for leupeptin, 0.2 for 15 μM for antipain and 0.5 for 10 μM for chymostatin. Since gelatin itself is a substrate for both of these enzymes it is important to consider that in the aspect of the plateau value. But as the enzyme molecules encounter the inhibitor molecules by diffusion the inhibition is gradual. In comparison with the free inhibitor, leupeptin declines to a lower vi/v0 value at lower concentra-tions but that could be explained by the competition between the inhibitor and the gelatin and also the diffusion of the enzyme.

61.1 ± 6.9 mm2/μMenzyme 8.2 ± 0.6 mm

2/μMenzyme

52

TWO-LAYER MODEL The two-layer model (Figure 35) represents a more direct simulation of the barrier system which was the aim of this study. In this model the enzyme will encounter the inhibitor in the first layer before entering the bottom layer con-taining the substrate. Besides the studies where the first layer contained inhib-itors, different thicknesses of the first layer were evaluated as well.

Figure 35. Illustration of the two-layer model.

Undisturbed kinetics for the conversion of model substrates in the gel The apparent kinetic parameters Vmax and KM can be seen in Table 1 below. In this experiment the top layer consisted of plain gel solution, set to 3 mm thick, which corresponds to 100 μl gel. Table 1. Apparent kinetic parameters from saturation curves (Paper 1, Fig. 6a and 6b).

Vmax (μμM/min) KM (mM) kcat (min-1) Trypsin 1.8 ± 0.1 5.3 ± 0.7 5.5 α-Chymotrypsin 2.9 ± 0.2 0.5 ± 0.1 8.8

The apparent KM value for trypsin is one order of magnitude higher than its KM of 0.9 mM in free solution. The difference in the buffer conditions in the two different experiment may affect the result and also the diffusional delay in the gel matrix itself. The apparent KM value of α-chymotrypsin is in the same range as that for reactions in free solution, showing that α-chymotrypsin is not as easily affected by the surroundings.

Inhibition of trypsin and α-chymotrypsin Since the inhibitors studied here are tight binding inhibitors the ratio between the initial rate in the presence of the inhibitor (vi) and the initial rate without the inhibitor (v0) are plotted against the inhibitor concentration. The dissocia-tion constant is in the same order of magnitude in the free solution and gel matrix respectively regarding trypsin (Table 2). Thus, the general mechanism of inhibition seems to be unaffected and that it is the diffusion that chiefly influences the Ki

app values. The apparent Ki values for α-chymotrypsin can be seen in Paper 1, Table 6.

Enzyme droplet

Inhibitor layer

Substrate layer

53

Table 2. Apparent Ki values extracted from GraphPad Prism 7 when fitting to Morri-son equation for tight binding inhibitors.

Kiapp in gel matrix

Kiapp in free solution

Antipain 0.35 ± 0.08 μM 0.10 ± 0.01 μM Leupeptin 0.50 ± 0.09 μM 0.09 ± 0.01 μM

The effect of using different thickness of the barrier layer The barrier layer thickness was furthermore varied between 1.5 mm to 4.5 mm and also investigated with and without the presence of inhibitor. The effect of the different thickness shows that the increasing barrier layer thickness corre-lates with an increase in KM. These findings confirm that the gel matrix not only acts as a diffusional delay but also acts as a competing substrate. With the inhibitor presence in the barrier layer the decrease in the apparent Ki ac-companied the thickness of the barrier layer. This could be explained by the higher absolute amount of the inhibitor in the system. An increase in the bar-rier layer will led to a decrease in the observed activity. A parallel experiment with using agarose instead of gelatin was also carried out since, as mentioned, gelatin is a competing substrate for proteases. The KM from the two experiments shows only minor differences which indicates that the gelatin gel used does not affect the overall kinetic activity.

CONCLUSIONS In this paper I’ve showed that the basic well formation approach is an exceed-ingly simple screening method for protease activity. The addition of inhibitor in the screening lowered the area well expansion. The simple gelatin layer serves as a general substrate for monitoring protease activity and protease in-hibition. The results from the two-layer model clearly demonstrate that the action of the proteases was strongly inhibited by inhibitors in the top layer. The double layer design serves as a model for active barrier systems and is suitable for plate reader application. An apparent increase of inhibition was observed not only as an effect of increased inhibitor concentration, but also from an increase of the barrier layer thickness.

In this project I shown that the approach allows easy evaluation of systems and designs intended to counteract or eliminate enzyme action where an in-

hibitor barrier separates the enzyme from its target.

54

PAPER II

LIGHT SCATTERING DETERMINATION OF THE STOICHIOMETRY FOR PROTEASE-POTATO SERINE

PROTEASE INHIBITOR COMPLEXES

In this project I’ve purified and characterized an inhibitor from Solanum tu-berosum, commonly known as household potato. It was suggested that this

inhibitor is a double-headed Kunitz type inhibitor that could simultaneously bind two proteases. I wanted, by the use of static light scattering, determine the interactions between PSPI and trypsin, α-chymotrypsin and elastase. In

order to do so, the samples were introduced into the measurement system via a size-exclusion column that allowed separation of the complex and the

“free” molecules due to strong interaction and thereby a slow dissociation of the complex.

THE ACTIVE SITES It has been suggested that PSPI’s are double-headed Kunitz type inhibitors and a number of reactive sites have been proposed. Since the structure of PSPI has not been completely solved, the reactive site(s) have not been confirmed. The suggested active sites are thus predicted by sequence alignment (Figure 36) of other similar Kunitz-type serine protease inhibitors with determined and veri-fied reactive loops. From these alignments three different suggestions arose. The first suggestion is Asp45-Asn50 where the corresponding reactive site in winged bean α-chymotrypsin inhibitor is Asn38-Leu43.133 The P1 residue (Figure 6) in this loop is suggested to be Pro47. The second suggestion is Ser71-Phe80 due to that this is the most common reactive-site loop in mono-meric Kunitz-type serine protease inhibitor.134 Phe75 is proposed being the P1 residue in this loop and points out into the solvent which could facilitate it as a recognition site for attacking proteases. The third and final reactive site could be located at Lys95. This loop has been identified as a reactive site in other double-headed protease inhibitors135,136 and similar to suggestion I, pro-trudes from the structure and points out into the solvent.

AIM Confirm the interaction of PSPI and three different

proteases by the use of static light scattering

55

10 20 30 40 I 50 LPSDATPVLD VTGKELDPRL SYRIISTFWG ALGGDVYLGK SPNSDAPCAN

60 70 II 80 90 III 100 GVFRYNSDVG PSGTPVRFIG SSSHFGQGIF EDELLNIQFA ISTSKMCVSY

110 120 130 140 150 TIWKVGDYDA SLGTMLLETG GTIGQADSSW FKIVKSSQFY NLLYCPVTTS

160 170 180 SDDQFCLKVG VHQNGKRRLA LVKDNPLDVS FKQVQ

Figure 36. Sequence of PSPI where suggested active sites are marked in blue

(I: Asn45-Asp50, II:Ser71-Phe80, III: around Lys95) To get a better insight into the protease-inhibitor binding stoichiometry, static light scattering was used to determine the functional molar mass in solution, thereby revealing the nature of possible complexes. The samples were intro-duced into the light scattering measurement cell via a chromatography column that served to separate complexes from noncomplexed components. I have studied the interaction between PSPI and three different proteases, namely trypsin, α-chymotrypsin and elastase where the injected sample were studied in 4 different molar ratios (1:1, 2:1, 3:1 and 4:1 protease:PSPI for trypsin and α-chymotrypsin and 1:1 and 2:1 for elastase:PSPI). All scattering data shown were averages from experiments carried out in triplicate. Theoretical mass av-erages were calculated using equations 14-16 below:

(Eq. 14)

(Eq. 15)

(Eq. 16) Each component was individually measured and compared to the theoretically calculated molar mass (Table 3). Table 3. Measured and calculated masses for possible defined components. These values were used for the evaluation of the measurements.

Components Mapp (kDa) Trypsin 23.7

α-Chymotrypsin 25.6 PSPI 19.1

Tentative complexes Mcalculated (kDa) Trypsin-PSPI 42.8 Trypsin2-PSPI 66.5

α-Chymotrypsin-PSPI 44.7 α-Chymotrypsin2-PSPI 70.3

56

TRYPSIN-PSPI INTERACTION

Figure 37. Molar mass distribution for 10 µM PSPI and 40 µM trypsin

The molar mass distribution (Figure 37) revealed that only two components were found for samples containing trypsin-PSPI in 4:1 ratio. Narrow band analysis of the same samples revealed a significant amount of a component with a molar mass far exceeding the theoretically expected mass for a 1:1 complex but corresponding to that expected for a 2:1 complex in the first peak (Table 4). Table 4. Narrow band molar mass values (kDa) for 10 μM PSPI and 40 μM trypsin in the first peak. Best fit theoretical values are indicated in bold face.

Measured value Peak 1:4 PSPI:Trypsin 64.2

Theoretical Value Try2PSPI 66.5 TryPSPI 42.8

The evaluation by integrating the full sample molar mass average and com-paring to theoretical calculations indicates that PSPI interacts with two trypsin molecules simultaneously (table 5). Table 5. Evaluation of Trypsin-PSPI 4:1 ratio by integration of the full sample. Best fit theoretical values are indicated in bold face.

Full sample molar mass average values Data source Mn (kDa) Mw (kDa) Mz (kDa) Experiment 38.0 51.3 60.1

Theoretical value Try2PSPI + 2 Try 38.0 48.7 57.8 TryPSPI + 3 Try 28.5 30.9 33.6

57

α-CHYMOTRYPSIN-PSPI INTERACTION In the interaction between α-chymotrypsin and PSPI no experiments display any significant presence of components with a molar mass corresponding to a 2:1 complex. Even though the ratio between α-chymotrypsin and PSPI was increased from 1:1 to 4:1 the data merely displayed an increasing “saturation” of the 1:1 complex accompanied by a correspondingly larger second peak in the chromatogram proportional to the increasing excess of α-chymotrypsin. It is similarly obvious from the distribution analysis that there is a larger com-ponent with a mass in the 45 kDa range regardless of the α-chymotrypsin-PSPI ratio (Figure 38, Table 6).

Figure 38a-d. Molar mass distribution for (a) 1:1 (b) 1:2 (c) 1:3 and (d) 1: 4PSPI:α-chymotrypsin.

Table 6. Narrow band molar mass values (kDa) for different ratio of concentration between PSPI and α-chymotrypsin in the first peak. Best fit theoretical values are indicated in bold face.

Measured valuesPSPI: α-Chymotrypsin Peak

1:1 43.0 1:2 43.0 1:3 44.5 1:4 44.1

Theoretical Value α-ChyPSPI 44.7 α-Chy2PSPI 58.3

a. b.

c. d.

58

The full sample molar mass average values (Table 7) confirm that there is no component corresponding to the 2:1 complex observed for α-chymotrypsin and PSPI with ratio 4:1. Table 7. Evaluation of α-chymotrypsin-PSPI 4:1 ratio. Best fit theoretical values are indicated in bold face.

Full sample molar mass average values Data source Mn

(kDa) Mw

(kDa) Mz

(kDa) Experiment 30.0 32.1 34.3

Theoretical value α-Chy2PSPI + 2 α-Chy 40.5 51.5 60.9 α-ChyPSPI + 3 α-Chy 30.4 32.6 35.2

Stoichiometry estimates The association and dissociation between the complexes are relatively slow and this results in a separation between the complex and free components, which we can call “frozen equilibrium”. This allows for a fair estimation of the stoichiometry by comparing the experimental peak integral ratio to the theoretical calculated one (table 8). The RI signal is a rather constant mass/sig-nal factor for different proteins which gives an appropriate accuracy for the integral ratios. The data in table 8 is good enough and corresponds to 1:1 com-plex. Table 8. RI peak area ratios for α-Chymotrypsin-PSPI combinations.

Molar ratios 2:1 3:1 4:1

Observed ratio 0.60 0.9-1.25 1.39-1.66 Theoretical value

Theoretical ratio 2nd/1st for 1:1 complex 0.57 1.14 1.71 Theoretical ratio 2nd/1st for 2:1 complex low 0.36 0.73

ELASTASE-PSPI INTERACTION The analysis of the interaction of elastase and PSPI in a 2:1 ratio revealed a single macromolecular peak with a faint leading shoulder and the correspond-ing distribution analysis shows a weak contribution of a component in the 40 kDa range corresponding to a 1:1 complex. This indicates that it is a weak interaction with a Ki in the 100 μM range. Elastase is less specific with pref-erence for small amino acids residues that may explain the weaker binding observed.

59

CONCLUSIONS Here I have shown that PSPI can interact with two trypsin molecules simulta-neously but only with one α-chymotrypsin. The reason for this could lie in the active site of the enzymes and the two suggested binding loops of PSPI (Figure 12). Trypsin is able to bind to both a phenylalanine residue and a lysine residue and the complex is strong enough for the light scattering experiment. The re-sults show that α-chymotrypsin binds to one of these sites of PSPI (1:1 ratio) and it would most likely be Phe75 since that is an aromatic residue. The pos-itive charge of lysine could be the reason behind the 1:1 binding between α-chymotrypsin and PSPI by preventing the binding to the second loop. If the active site of PSPI is placed at Pro47 it would in theory be able to bind to α-chymotrypsin, but the results here strongly indicate that it does not. It can also be due to that Pro47 is sterically inaccessible. The interaction between PSPI and elastase showed a weak indication of a component in the 40 kDa range which corresponded to around 10% of the total integral. In this project I was able to show that PSPI is able to bind two trypsin mole-cules simultaneously in a 2:1 complex. However, the data also shows a limi-tation to the interaction with α-chymotrypsin to a 1:1 complex and, possibly,

a weak 1:1 complex with elastase.

60

PAPER III

CHARACTERIZATION OF SERINE PROTEASE INHIBITOR FROM SOLANUM TUBEROSUM CONJUGATED TO SOLUBLE

DEXTRAN AND PARTICLE CARRIERS

The PSPI-protease interaction was studied in Paper II. Here, I present in-stead PSPI conjugated to prefractionated oxidized dextran and inorganic

particles, i.e. titanium dioxide and zinc oxide, and characterized PSPI in its conjugated state. The choice of the carriers was decided upon their physical

properties and biocompatibility.

PREFRACTIONATION & OXIDATION OF DEXTRAN Before it was possible to conjugate PSPI to water soluble dextran it was de-cided to prefractionate dextran due to easier data interpretation. Prefractiona-tion was achieved using gel filtration giving fractions with a narrower molec-ular weight interval. Dextran was then activated through periodate oxidation. The oxidation of dextran led to degradation of the soluble carrier which was confirmed by static light scattering (Table 9) and the mass average of the total sample was calculated by Equation 17. A distribution of the molar mass could also be seen, however, on average the dextran polymer was degraded by a factor 4. Table 9. Light scattering analysis of dextran and its derivatives (kDa).

Mw

front Mw

peak Mw

back Mw

whole Dextran, 1% 116 61 31 68 Oxidized dextran, 1% 32 11 - 17

(Equation 17)

AIM Establish a conjugation protocol for the protease inhibitor

from Solanum tuberosum onto carriers and characterize the

stability and kinetics parameters of the conjugated inhibitor.

61

CONJUGATION OF PSPI

PSPI-dextran conjugate Five different conjugation conditions were tested using different mass ratio of PSPI and the oxidized dextran, time of incubation, amount of reducing agent, and if HCl was used to neutralize the alternatives in order to obtain the conju-gate (Figure 39). In the first alternative a mass ratio of 0.59:1 (PSPI:oxidized dextran) was used and led to partial conjugation. The second alternative, where a mass ratio of 0.18:1 (PSPI:oxidized dextran) was used, led to a com-plete conjugation of PSPI and oxidized dextran. A complete conjugation could also be seen for alternatives 3, 4 and 5 were the mass ratios were lower than in alternative 2. Light scattering data and SDS-PAGE confirmed a complete conjugation for alternative 2-5 (Paper III) which also indicates that dextran should have more available site for conjugation. The chromatogram for alter-native 1 showed a partial conjugation most likely due to excess of the inhibi-tor. The conjugation could also be confirmed visually since the solution be-came turbid during the reaction.

Figure 39. Model that demonstrates the conjugation of PSPI and oxidized dextran

through a reduced Schiff base coupling.

PSPI-inorganic particle conjugates Next, I also functionalized inorganic nano particles through the method de-scribed in amide formation. The amount of PSPI conjugated to the inorganic particles can be seen in Table 10. Table 10. Amount PSPI conjugated to inorganic particles

Particle Conjugated PSPI TiO2 2.86 nmol/mg ZnO 1.70 nmol/mg

62

POSSIBLE CONJUGATION SITES There are a number of positions where the conjugation can occur since the reaction proceeds between a lysine of PSPI and the oxidized dextran, so called multi-point binding. PSPI has ten lysines available for conjugation where most of them are exposed (Figure 40). Lys95, in particular, is proposed to be crucial for trypsin inhibition and PSPI conjugated by this residue will thus not be ac-tive by this site.

Figure 40. Model of PSPI where the ten possible conjugation sites

are marked in pink.

EFFICIENCY OF THE CONJUGATES

PSPI-dextran conjugatesThe apparent Ki values obtained after conjugation are shown in Table 11. They have increased by a factor 2-3 but are still in the same range as the free inhib-itor. This increase can be explained by cross-linking between dextran-PSPI-dextran and sterical inaccessibility for the enzyme. This apparent Ki value is obviously affected by which lysine that is part of the conjugation. As men-tioned earlier, Lys95 is most likely part of the inhibition and if this is a site for conjugation the inhibition will be affected since this site is no longer available. Yet, it appears from the apparent Ki values that most of the conjugation has not affect the inhibition significantly, indicating that Lys95 is a site where minor conjugation has occurred. To be able to evaluate the conjugates I have chosen to use the parameter efficiency. It is defined as the fraction of inhibitor molecules that are functionally active. The loss in efficiency can be ascribed as denaturation and/or sterical hindrance. The efficiency is calculated from Equation 18.

(Equation 18)

63

The inhibitor remains 39-58% active after conjugation and 12-20% active af-ter one month of storage (Paper III, Table 3). In comparison, 11% of with the free inhibitor that was stored under the same conditions remained after one month. The conjugated inhibitor could thus be roughly twice as stable as the free inhibitor. Table 11. Apparent Ki values of the dextran-PSPI conjugates with corresponding ef-ficiency.

Alternative Ki (µM) Efficiency Free PSPI 0.10 ± 0.03 1.00*

1 0.28 ± 0.03 0.37 2 0.23 ± 0.03 0.44 3 0.19 ± 0.03 0.53 4 0.17 ± 0.04 0.60 5 0.25 ± 0.03 0.41

*reference value

PSPI-particle conjugates The apparent Ki values of the conjugates can be seen in Table 12. The effi-ciency of TiO2-PSPI is calculated to be 0.06 and 0.02 for ZnO-PSPI. The low efficiency could be explained by a likely degree of denaturation of the conju-gated inhibitor. Sterical inaccessibility can of course not be ruled out due to the rough surface of the carriers. The efficiency is also here affected by which lysine that is part of the conjugation. During the last step in the conjugation to the particles the conjugate is exposed to 65 °C for 5 hours to remove buffer. The same heat treatment increases the apparent Ki of the free PSPI by a factor of ∼170. This clearly indicates that the conjugates are more stable to heat treat-ment. Table 12. Apparent Ki-values of the different conjugates.

State of PSPI Ki (µM) Efficiency Free PSPI 0.10 ± 0.03 1.00* TiO2-PSPI 1.81 ± 0.27 0.06 ZnO-PSPI 4.83 ± 0.99 0.02 Free PSPI after heat treatment 17.3 ± 1.13 0.006 *reference value

THE GELATIN EROSION METHOD In order to evaluate the function of the conjugates the gelatin erosion method (Paper I) was chosen as an additional evaluation. This is due to the fact that this method simulates a process at a surface where a protective cover layer is

64

needed. Both the free PSPI and the conjugated PSPI showed inhibition to-wards trypsin and α-chymotrypsin digestion actions on gelatin. The ratio (v0/vi) of the area increase rate in the presence of the inhibitor (vi) and the area increase rate of the undisturbed enzyme (v0) was plotted as a function of free and conjugated PSPI which corresponds to a linear function of the inhibitor concentration (Table 13). There is no large difference between the free PSPI and the conjugated one. This indicates that the conjugation itself will not affect the mode of action and that the conjugated inhibitor can still function similarly as the free inhibitor. The mechanism itself is also preserved since the v0/vi of the conjugate remains as a linear function of the inhibitor concentration (Paper III, Figures 7 and 8). Table 13. Linear regression values extracted from the v0/vi-graphs. Values extracted from GraphPad Prism 8.

Enzyme State of PSPI Slope from v0/vi

Trypsin Free 2.77 ± 0.35 Conjugated 2.54 ± 0.17

α-Chymotrypsin Free 0.76 ± 0.07

Conjugated 0.51 ± 0.05

CONCLUSIONS Different types of carriers will lead to improvement in the applicability in both medical/care purposes and for the use in biotechnology processes. The SDS-PAGE analysis indicates that the inhibitory effect really depends on the con-jugate since no traces from the free inhibitor could be seen on the gel. The increase of the apparent Ki value is most likely due to the conjugation itself and the possible crosslinking and sterical inaccessibility that follows as a con-sequence. Due to PSPI’s 10 lysines the conjugation will be randomized, and it can also lead to crosslinking between the polymer molecules due to multi-point binding. If conjugation has occurred at Lys95 it should lower the inhib-itory effect, possibly explaining the increase in Ki

app. Conjugation of the in-hibitor leads to an improved storage stability compared to the free inhibitor in solution. The conjugation between PSPI and the inorganic particles leads to a conjugate that is more resistant to heat treatment. Overall, the increase in Ki

app is most likely due to the limited availability of individual inhibitor molecules rather than a gradual change of the intrinsic properties of the molecules re-garding both kinetics and binding equilibrium.

In this project I was able to establish an immobilization protocol for PSPI

and showed that the conjugate retained its inhibitory effect and also in-creased the stability of the inhibitor. Regarding the particle conjugate I have

shown that the conjugate is more stable to heat treatment in comparison with the free inhibitor.

65

PAPER IV

INHIBITION PROPERTIES OF FREE AND CONJUGATED LEUPEPTIN ANALOGUES

In this project I wanted to modify the natural inhibitor Leupeptin so it could be conjugated and detected in an easier way. This took me into the field of

solid-phase peptide synthesis (SPPS) and led me to the synthesis of two tripeptides which differed in the R2-group and later to the conjugation of one

of the peptides to both gel beads and inorganic particles. Leupeptin has been part of my research since the beginning. In nature Leupep-tin is synthesized by three different enzymes. The first step in the synthesis is the acylation of leucine by leucine acid acylase. Leupeptin acid is synthesized by a multifunctional enzyme (leupeptin acid synthetase) where acetyl-leucine, leucine and arginine are coupled together using ATP as energy source. In the third and final step leupeptin acid is reduced to leupeptin by leupeptin acid reductase in the presence of ATP and NADPH. The reduction of a carboxylate group by NADPH to aldehyde is however thermodynamically unfavored and is thus supported by ATP hydrolysis. The chemical synthesis of leupeptin is thus more complicated than standard peptide synthesis due to its L-argininal moiety.137-139

OXIDATION & REDUCTION OF LEUPEPTIN In the first part of this project the importance of the aldehyde group of leupep-tins C-terminal was tested. The aldehyde structure was both reduced (Paper IV, Figure 3) and oxidized (Paper IV, Figure 2), confirmed with MS (Paper IV, Figure 7) and the products were studied kinetically. The apparent Ki value of the reduced form was measured to 270 ± 100 μM and for the oxidized form it was measured to 2.69 ± 0.13 μM. Since neither the reduced nor the oxidized state can form a hemiacetal (Figure 41) this could be the reason behind the higher apparent Ki compared to the natural state of Leupeptin which has a Ki value in the nanomolar range (87.5 ± 7.8 nM). Due to the possible interaction of the negative oxygen of the oxidized form in the oxyanion hole in the active

AIM

Design & characterize protease inhibition tools

based on easily synthesized inhibitors that are

conjugated to suitable carriers

66

site of trypsin, this could be the reason for the lower Ki value in comparison with the reduced form. Since the oxidized form of Leupeptin gave an apparent Ki value in the low micromolar range it served as a starting point when this project moved on to the second part.

Figure 41. Reaction forming the hemiacetal of leupeptin and the hydroxyl group of serine residue at the active site of trypsin.

SPPS OF AHX-R-LEU-ARG-COOH In the second part I wanted to synthesize my own peptides using a carboxylic group at the C-terminal instead, since that the apparent Ki value of the oxidized form of Leupeptin was in the micromolar range, but also due to that a synthetic pathway of a serine protease inhibitor, as leupeptin, is challenging due to the argininal residue at the C-terminus of their sequence resulting in a low yield.140 Our synthesis started with an arginine preloaded to a Wang resin. In the next step the amino acid leucine was added, following that, phenylalanine was added for synthesis of Ahx-Phe-Leu-Arg-COOH or a second leucine for syn-thesis of Ahx-Leu-Leu-Arg-COOH. Thus, I decided to do two different but very similar peptides. In the last step of the peptide synthesis 6-aminocaproic acid was added as an extension that could serve as a site for conjugation due to the primary amine. The difference between the two new peptides are found in the R2, where the phenylalanine made it easy to detect the peptide through-out the experiments using Phe-absorbance. The mass spectra confirming the new peptides can be seen in Paper IV Figure 8 and 9 and the two synthesized peptides can be seen below in Figure 42.

Figure 42. Structures of the two tripeptides.

Ahx-Leu-Leu-Arg-COOH

513.7 Da

Ahx-Phe-Leu-Arg-COOH

547.7 Da

67

INHIBITION PROPERTIES OF THE TWO PEPTIDES

The apparent Ki value of Ahx-Phe-Leu-Arg-COOH was 3.42 ± 0.24 μM and for Ahx-Leu-Leu-Arg-COOH it was 9.48 ± 0.96 μM. Compared to the oxi-dized form of Leupeptin which had a Ki value of 2.69 ± 0.13 μM they lie in the same range. The minor difference that can be seen between peptides 1 and 2 would most possible be found in the difference in the R2-group. Replace-ment of leucine by a phenylalanine lowers the Ki value by a factor of 3. This could be due to that phenylalanine is more hydrophobic and therefore interacts more strongly in the active site of the enzyme. Even though these peptides are synthesized with a carboxylic acid instead of the original aldehyde group both peptides still show significant inhibition properties.

CONJUGATION OF AHX-PHE-LEU-ARG-COOH Ahx-Phe-Leu-Arg-COOH was conjugated to both agarose gel beads and to inorganic particles, namely titanium dioxide and zinc oxide. The N-terminal acetyl group found in original leupeptin was in these peptides replaced with 6-aminohexanoyl group which facilitates the conjugation while retaining the inhibitory effect by leaving the important arginine well accessible (Figure 43). Another benefit by adding this ‘arm’ is that it gives rise to site-selective con-jugation, meaning that the conjugation is controlled. The peptide was conju-gated to the agarose gel bead by a bromo activated matrix resulting in a sec-ondary amine coupling. The inorganic particles where first silanized, followed by activation and resulting in a substituted urea-type coupling (Figure 43). The amount conjugated per carrier can be found in Paper IV Table 2.

Figure 43. Conjugate of TiO2 particles and Ahx-Phe-Leu-Arg-COOH.

R =

68

PERFORMANCE OF THE CONJUGATES The three carriers used in the conjugates all have an uneven surface resulting in an increase in apparent Ki value due to sterical inaccessibility. The distance between conjugated peptides also gives rise to an increase in Ki

app because the space for the enzyme molecules is less accessible. The risk of crosslinking between these conjugates are non-existing since the only possible site for con-jugation is the added arm at the N-terminal. An increase in the Ki

app regarding the agarose gel beads can be explained by that the peptides can be hidden inside the beads. That will obviously influence the number of peptides acces-sible for the enzyme. The efficiency of the ZnO-peptide conjugate reveals that the conjugation itself does not affect the inhibition properties. A lower efficiency can be seen for the agarose gel bead-peptide conjugate and that could, as mentioned above, be explained by the sterical accessibility for the enzyme. Around 30% of the con-jugated peptides are available for inhibition of the enzyme regarding the TiO2-conjugate and around 10% for the agarose gel bead-conjugate. A storage of six months showed a Ki

app value in the same range for the inor-ganic particle conjugates. The bond between the carrier and the peptide is thus stable and the inhibitory affect is retained. The storage of the agarose gel beads led to an increase in Ki

app by approximated 50%. This could be due to either loss of peptides or change in the structure of the gel beads.

INHIBITION ACTIVITY IN A GELATIN LAYER

Figure 44. Relative area increase rate for free Ahx-Phe-Leu-Arg-COOH, conjugated Ahx-Phe-Leu-Arg-COOH and free particles.

69

This experiment is used to evaluate the function of the conjugates in a simu-lated surface with a protective cover layer. Both the free and conjugated in-hibitor, respectively, were incorporated in the gel. The degree of inhibition is similar when comparing the plateau values (Paper IV, Table 4) and indicates that the plateau value is only affected by the peptide itself, meaning that the conjugation does not prevent the inhibition (Figure 44). The difference can be seen in the start of the inhibition where the conjugation of ZnO and the peptide gives a steeper inhibition. This could be explained by the inhibition/adsorption properties zinc oxide itself possesses and combined with the peptide this sup-presses the increase in the relative area rate. This is also confirmed by the experiment with only ZnO itself since the plateau value decreases to 86%. Another confirmation of this is that there is no significant difference between the free peptide and the TiO2-conjugate. Consequential, the ZnO-peptide con-jugate makes a powerful conjugate regarding its inhibition properties.

CONCLUSIONS The design and synthesis of two new protease inhibitors with retained inhibi-tion properties was successfully achieved here. The peptides were designed with a new functional group instead of the argininal group of the natural state leupeptin and an extra handle was added in the opposite end to facilitate the conjugation of these peptides. An apparent Ki value were measured in the mi-cromolar range and after the conjugation of Ahx-Phe-Leu-Arg-COOH the in-hibition was retained with a Ki

app in the micromolar range for all conjugates. The conjugation of the peptide is performed not to increase the stability but to create a larger size object by conjugating the peptide to a larger carrier.

In this project I was able to design two new protease inhibitors with retained inhibition properties and that a conjugation of these inhibitor

to inorganic particle and agarose gel beads gave a retained Ki

app in the micromolar range.

70

CONCLUDING REMARKS & FUTURE PERSPECTIVES

The overall theme of this thesis is a step by step approach to design and char-acterization conjugated protease inhibitors. I have designed conjugates with inhibitors attached to larger carriers in order to create a better and safer way of handling the inhibitors, both in technical and medical approaches. In this thesis you have read about a new assay method for protease activity and pro-tease inhibition (paper I), a study of the stoichiometry for protease inhibitor interaction (paper II), design of inhibitory peptides (paper IV) and the con-struction of inhibitor conjugates (paper III & IV). In Paper I have showed that the basic well formation approach is an exceed-ingly simple screening method for protease activity. The simple gelatin layer serves as a general substrate for monitoring protease activity and protease in-hibition. The double layer design serves as a model for active barrier systems and the results clearly demonstrate that the action of the proteases was strongly inhibited by inhibitors in the top layer. This approach can be used to evaluate systems and designs intended to weaken or eliminate enzyme action where a barrier containing, i. e. inhibitors, separates the enzyme from its substrate. It is also an easy evaluation method of conjugated inhibitors towards attacking proteases. The experimental setup in Paper II allowed a quantitively measurement of the interaction between proteases and the inhibitor from Solanum tuberosum (PSPI). The results revealed that PSPI acts as a double-headed inhibitor to-wards trypsin and simultaneously binds two molecules, but the inhibitor limits itself to only one α-chymotrypsin and shows only a weak binding to elastase in a 1:1 ratio. I continued my work with PSPI in paper III and conjugated it to both soluble carrier and inorganic particles. A degree of inhibition was retained for all con-jugates and for the dextran conjugate the stability of the inhibitor increased after conjugation. In the case of the inorganic particle conjugates, the inhibitor became more resistant to heat treatment compared to the free inhibitor. Since PSPI has ten lysine residues that could act as a conjugation site and thereby eliminates the possibility of all PSPI conjugated available for inhibition a site-selective conjugation would be interesting. By expressing PSPI recombinant

71

with a single site available for conjugation would thereby control the conju-gation and opening for more conjugated molecules with an available active site. In my last paper IV presented in this thesis I have designed and synthesized two new protease inhibitors with retained inhibition properties and also char-acterized them in their conjugated state. For the future it would be interesting to end the synthesis enzymatically by reducing the carboxylate group to an aldehyde and kinetically characterize the peptide inhibitor by SPR or stopped flow. It could also be possible to synthesize/add the aldehyde in solution as a last step. The conjugation of the peptide was performed not to increase the stability but to create a larger size object in order to minimize the risk of the peptide for example crossing the human skin barrier. The conjugation can also facilitate the removal of the peptide in a controlled way (magnetic particles) and minimize the risk of the peptide crossing over for example the skin barrier in a formulation. So, to get back to the overall theme of this thesis which actually is, in a step-by-step approach, to design and characterization conjugated protease inhibi-tors and to the last question: why is this important? These papers were founded in the idea that formulations could contain inhibitors to prevent damage of attacking proteases or relief the area of the human skin that already has been damaged. Dermatitis is mainly caused by attacking pancreatic proteases that can lead to severe rashes and blisters. To be able to overcome limitation pro-teins and peptides have due to the lower stability and solubility for proteins and the small size of the peptides studied here the idea of conjugation arose. The dextran conjugate showed an increase in stability which is valuable in for example a formulation containing enzyme inhibitors. The inorganic particle conjugates have multifunctional action since they also protect the human skin from sun burn and have antibacterial properties. This thesis has contributed to knowledge of what a conjugation can do to the properties of inhibitors, the design of new inhibitor analogues and also increased the possible applications of inhibitors through conjugations. To continue and study the function of con-jugates in formulations, but also in biotechnical applications would be inter-esting. The development of a ‘multi-target’ meaning a carrier that includes several different inhibitors conjugated which will broaden the conjugate and inhibit several enzymes at once, would also be an interesting idea for upcom-ing research.

Thank you for being with me to the end of this thesis, I hope you enjoyed it as much as I have during these years!

72

POPULÄRVETENSKAPLIG SAMMANFATTNING

KARAKTERISERING AV KONJUGERADE PROTEASINHIBITORER

Enzymer är en häftig skapelse av naturen. De är specialiserade proteiner som har fantastiska katalytiska egenskaper. Utan ett enzym närvarande sker oftast reaktionen långsammare, oftast till den grad att vi själva inte skulle överleva utan enzymernas hjälp. Vi människor behöver enzymer då vi till exempel bry-ter ner mat, kontraherar muskler eller skickar nervsignaler. Enzymernas roll är att öka reaktionshastigheten utan att förändra jämvikten i reaktionen, då enzymerna sänker aktiveringsenergin genom att stabilisera övergången från substrat till produkt. Dessutom sker detta utan att enzymerna själva förbrukas. Ett enzym kan alltså genomföra samma reaktion om och om igen. Det finns ett enormt stort antal olika enzymer och i mitt arbete har fokus legat på en grupp enzymer som kallas för proteaser. De bryter ner proteiner (sub-strat) till kortare fragment (produkt). Det finns sex olika klasser av proteaser där jag har studerat enzymer från en klass som kallas serinproteaser. Namnet serinproteaser kommer från att dessa enzymer använder aminosyran serin i sin aktiva ficka som en nukleofil i reaktionen. Serinproteaser är involverade i många processer som sker i kroppen. Bland annat deltar de i koagulering av blod, celldöd (apoptos), befruktning, bildning av sår och matsmältningen. I matsmältningen finns det framförallt två enzymer som har studerats i stor ut-sträckning genom åren som kallas för trypsin och α-chymotrypsin. Trypsin och α-chymotrypsin utsöndras av bukspottkörteln och deltar i nedbrytningen av proteiner i tunntarmen. De lagras innan användningen som zymogener (in-aktiv form av enzymet) i bukspottkörteln som en självskyddsmekanism då dessa proteaser inte ska angripa varandra och bukspottkörteln själv. Zymoge-nerna kräver aktivering innan de får full enzymatisk aktivitet vilket sker med hjälp av att trypsin aktivera sig själv och andra enzymer som deltar i matspjälkningen. Experimenten i den här avhandlingen baseras på dessa två enzymer. För att motverka effekten av ett enzym kan man använda något som kallas för en inhibitor. En inhibitor hämmar alltså arbetet ett enzym genomför genom att

73

binda till enzymet. Antingen kan inhibitorn binda på samma ställe som sub-stratet gör eller så kan den interagera med någon annan del av enzymet som leder till att enzymet förlorar funktionen. En inhibitor kan vara allt från större proteiner till mindre molekyler. Jag har i mitt arbete studerat inhibitorer från vanlig potatis köpt på den lokala matbutiken, peptider från bakterier samt själv designat mindre peptider.

”Tänk dig en lång kedja. På denna kedja har jag kopplat fast mindre och större molekyler för att sedan studera hur egenskaperna hos dessa molekyler har förändrats vid kopplandet, särskilt då deras

förmåga att inhibera enzymer.” Mitt arbete under mina år som doktorand kan jag förklara ungefär såhär. Jag har till stor del i mitt arbete konjugerat inhibitorer till olika typer av bärare till ett så kallat konjugat. Den kemiska tekniken som kopplar samman två mole-kyler i en kovalent bindning, där åtminstone en av dem är en biomolekyl, be-nämns vanligen som biokonjugering. Konjugeringen i sig kan dramatiskt för-ändra biomolekylens fysiologiska egenskaper. Möjligheten att skapa dessa bi-okonjugat har påverkat och är omfattande för forskningen både inom akade-min och industrin. När nya tekniker utvecklas kan nya unika konjugat skapas vilket leder till att nya applikationer förs fram. Idag finns det en enorm mängd olika sätt att tillverka ett konjugat på så att välja en strategi beror till stor del på målet och utgångspunkten. Då många forskningsområden använder sig av konjugering resulterar det i en stor mångfald i metoder och tillvägagångssätt. Även om proteiner har många fördelar såsom deras storlek, biologiska funkt-ioner, specifika konformationer och biokompabilitet har de även brister såsom sämre stabilitet, lägre löslighet samt kortare livslängd. En konjugering kan leda till att de fysiologiska egenskaperna drastiskt förändras. Då peptider kan bestå av endast några få aminosyror, medan proteiner typiskt innehåller flera hundra aminosyror, kan de på grund av deras mindre storlek penetrera exem-

74

pelvis hudens barriär och behöver då bäras på ett större konjugat. Genom an-vändandet av en passande bärare kan dessa brister överkommas. När man vill designa ett konjugat är det viktigt att tänka på proteinet/peptiden som ska kon-jugeras, vilken bärare som ska användas, vilken kemi är bästa tillvägagångs-sättet för att skapa en permanent bindning, vart ska proteinet/peptiden konju-geras på bäraren samt det resulterande konjugatet. Det finns många olika typer av bärare som är biokompatibla. Jag har använt mig av både fasta (titandioxid, zinkoxid, agaroskulor) och lösliga bärare (dextran). Alla dessa bärare används redan i olika typer av formuleringar som du säkert har använt genom åren. Titandioxid används som ett UV-filter i solkräm, zinkoxid finns i zinkpasta som används för att behandla eksem, sol-bränna och nässelutslag där zinken har en uttorkande effekt. Dextran är en polymer som används i stor utsträckning som bärare i många vetenskapliga publikationer samt som ersättningsmedel för blodplasma. Dextran är alltså en biokompatibel, inert och icke toxisk polymer som kan administreras i kroppen utan biverkningar.

I den här avhandlingen presenterar jag flera olika konjugat samt hur dessa konjugat fungerar i jämförelse med den fria molekylen.

I artikel III har jag konjugerat PSPI (inhibitorn från potatis) till lösligt dextran och oorganiska partiklar. Konjugatet leder till att inhiberingseffekten halveras, mest troligtvis på grund av steriska hinder mellan enzymet och konjugatet, vilket ger oss att åtminstone hälften av alla PSPI-molekyler är tillgängliga för inhibering. Förändringen i inhiberingen kan också bero på att en PSPI-mole-kyl kan tvärbinda till flera dextranpolymerer. Konjugeringen leder till att sta-biliteten av PSPI ökade samt att den blev mer tålig för högre temperaturer. I artikel IV kan ni läsa om hur en modifiering av existerande peptidinhibitorer leder till fortsatt god inhibering av proteaser. Jag fortsatte projektet med att syntetisera peptider med bibehållen inhiberande effekt samt ett ”handtag”, en tilläggsdel på molekylen, som skulle möjliggöra en konjugering. Konjuge-ringen gjordes till oorganiska partiklar samt gelkulor från agaros som gav ett stabilt konjugat även efter 6 månaders förvaring.

Jag har utöver detta även studerat interaktionen mellan PSPI och serinprotea-ser med hjälp av ljusspridningsexperiment (Static Light Scattering, SLS) i ar-tikel II. Statisk ljusspridning är en teknik som mäter intensiteten hos det spridda ljuset för att erhålla genomsnittlig molekylvikt (MW), storlek, bekräfta konjugering eller interaktion mellan ett protein(er) eller en eller flera makro-molekyler som en polymer i lösning. Tidigare datasimuleringar har visat att

75

den här inhibitorn kan binda flera proteaser samtidigt men aldrig bevisats ex-perimentellt. Här har jag visat det intressanta resultatet att inhibitorn kan inte-ragera med två trypsinmolekyler samtidigt men begränsar sig till endast en molekyl α-chymotrypsin. För att kunna studera om dessa konjugat kan fungera i en barriär mellan de attackerande enzymerna och substratet har jag skapat en modell i gelatin som ni kan läsa om i artikel I. Modellen har utvecklats så att den kan användas som ett eller två lager. I en-lager modellen var inhibitorerna en del av lagret och gelatinet i sig fungerade som ett substrat. En bra korrelation mellan enzym-koncentrationen och storleken på gropar som bildades i ett vanligt gelatinskikt observerades. På liknande sätt gav ökad koncentration av hämmare en syste-matisk minskning av gropstorleken. I två-lager modellen var inhibitorerna en del av det översta lagret och substrat en del av det nedre lagret. Kinetisk analys av tvåskiktsmodellen visade ett strikt beroende av både inhibitorkoncentration och tjocklek hos det övre "skyddande" lagret. En uppenbar men svagare inhi-beringseffekt observerades också utan hämmare på grund av diffusion och erosionsfördröjning av enzymtransport till det substratinnehållande skiktet. Tanken bakom den här avhandlingen grundar sig i en framtida möjlighet att använda dessa konjugat i olika typer av formuleringar som tex salvor och krä-mer. Genom en konjugering kan stabiliteten av inhibitorerna öka samt de kan också tålas att utsättas på olika miljöer jämfört med den fria inhibitorn och dennes optimala förhållanden. Dessa formuleringar kan appliceras på exem-pelvis huden och förebygga eller lindra olika typer av eksem eller sår som orsakas av enzymer. I den här avhandlingen har jag gått mellan instrumentellt avancerade tekniker till något så enkelt som att mäta gropar i gelatin men med samma vetenskapliga relevans.

76

ACKNOWLEDGEMENTS

#bmcdärdrömmarblirtillverklighet

First of, my supervisor Gunnar! Thank you so much for the years together with you. It has been a true pleasure. You have challenged me in many ways and to be able to experience the knowledge you possess close up is a privilege. I’ve realized now in the end of my PhD that you have known when I needed an extra push before I knew myself. I’ve learned so much from you and I will miss having our daily meetings and discussions. You have always had time for me when I popped my head in your office, which I very much appreciate. Thank you for letting me be a team with you and thank you for the nice friend-ship that we have formed over the years. I would also like to thank my co-supervisor Helena Danielson. You have given me valuable advice along the way, and with critical eyes on my research pushing me forward. A big thanks to all my collaborators: Shabira Abbas – thank you for giving me the opportunity to do my master thesis with you which led me to a PhD. This period of life has been extra fun as the collaboration with you has broad-ened my views. Johan Viljanen – for all your help and your vast knowledge in SPPS. Sara Bergström Lind – for guiding me in the world of mass spec-trometry. Shusheng Zou – for helping me with the PSPI-project while I was changing diapers and running after kids. To all you fantastic students that have helped me along the way! Martin G, Simon H, Cecilia Å, Kajsa E, Kris-toffer L (!), Enrique L, Anton R and Alejandro - your work in the lab are so valuable to me! Till alla ni gamla och nuvarande biokemister, några organkemister och admin. Ni har alla varit en stor del till att det har varit roligt att vara på labbet! åsa – till dig kan jag skippa grammatiken. helt . är innerligt glad att just du fanns på institutionen när jag kom ! att vi fann varandra bland alla pysselbutiker i Japan, när vi insåg vid templet att vi båda älskar smör lika mycket (vilket jag nu lärt linus, hej tandmärken!), tack för din granskning av den här boken, att du lärde mig vilket tempo på labbandet som gäller, att jag har fått vakna upp till 157 nya sms flera gånger om (tack till dig också jojo), tack för resor, häng och allt därtill! vi får hoppas att vi någon gång får möjlighet att skriva ”ses om fem

79

REFERENCES

1. Beynon, R.J., Bond, J.S. Proteolytic enzymes: A practical approach. Oxford Uni-versity Press, London. 2001.

2. Di Cera, E. Serine proteases. IUBMB Life. 2009, 61, 510-515. 3. López-Otín, C., Matrisian, L.M. Emergin roles of proteases in tumor suppression.

Nat. Rev. Cancer. 2007, 7, 800-808. 4. López-Otin, C., Bond, J.S. Proteases: Multifunctional enzymes in life and dis-

ease. J. Biolog. Chem. 2008, 283, 30433-30437. 5. Götze, H., Rothmann, S. S. Enteropancreatic circulation of digestive enzymes as

a conservative mechanism. Nature, 1975, 257, 607-609. 6. Kunitz, M., Northrop, J. H. Autocatalytic activation of trypsinogen in the pres-

ence of concentrated ammonium or magnesium sulfate. Science. 1934, 80, 190. 7. Kunitz, M., Northrop, J.H. Isolation of a crystalline protein from pancreas and its

conversion into a new crystalline proteolytic enzyme by trypsin. Science. 1933, 78, 558-559.

8. Baló, J., Banga, I. 1950. The elastolytic activity of pancreatic extracts. Biochem. J. 1950, 46, 384-387.

9. Kenninston, J.A., et al. Inhibition of Plasma Kallikrein by a Highly Specific Ac-tive Site Blocking Antibody. J. Biol. Chem. 2014, 22, 289, 23596-608.

10. Brown, W.E., Wold, F. Alkyl isocyanates as active-site-specific reagents for ser-ine proteases. Identification of the active-site serine as the site of reaction. Bio-chemistry. 1973, 12, 5, 835-840.

11. Kaufman, S., Neurath, H., Schwert, G.W. The specific peptidase and esterase ac-tivity of chymotrypsin. J. Biol. Chem. 1949, 177, 793-814.

12. Walsh, K.A. Trypsinogens and trypsins of various species. Meth. Enzymol. 1970, 19, 41-63.

13. Voet, D. Voet, J.G. 2004. Biochemistry, Wiley New York, 3rd Edition. 14. Kraut, J. Serine proteases: structure and mechanism of catalysis. Annu. Rev. Bi-

ochem. 1977, 46, 331-358. 15. Hess, G. P., McConn, J., Ku, E., McConkey, G. Studies of the activity of chymo-

trypsin. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 1970, 12, 257, 89-104. 16. Bender, M. L.; Killhefer, J. V. Chymotrypsins. Crit. Rev. Biochem. 1973, 1, 149-

199. 17. Steitz, T. A., Shulman, R. G. Crystallographic and NMR studies of the serine

proteases. Annu. Rev. Biophys. Bioeng. 1982, 11, 419-444. 18. Polgar, L. Mechanism of protease action, CRC Press: Boca Raton, 1989. 19. Schechter, I., Berger, A. On the size of the active site in proteases. I. Papain.

Biochem Biophys Res Commun. 1967, 27, 157–162. 20. Blay, V., Pei, D. Serine proteases: how did chemists tease out their catalytic

mechanism. ChemTexts. 2019, 5, 19. 21. Brown, A.J. Enzyme action. Chem. Soc. 1902, 91, 373-386. 22. Henri, V. Lois Générales de láction des diastases. J. Am. Chem. Soc. 1903, 25, 7,

780-782.

80

23. Michaelis, L., Menten, M.L. Die kinetik der invertinwirkung. Biochem. Z. 1913, 49, 333–369.

24. Briggs, G.E., Haldane, J.B.S. A note on the kinetics of enzyme action. Biochem. J. 1925, 19, 2, 338-339.

25. Lodish, H., Berk, A., Zipursky, S., Matsudaira, P. Baltimore, D., Darnell, J. 2000. Functional design of proteins. Molecular Cell Biology. 4th Edition.

26. Lineweaver, H., Burk, D. The determination of enzyme dissociation constants. J. Am. Chem. Soc. 1934, 56, 3, 658-666.

27. Williams, J. W., and Morrison, J. F. The kinetics of reversible tight-binding inhi-bition. Methods Enzymol. 1979, 63, 437-467.

28. Dixon, M. The graphical determination of Km and Ki. Biochem. J. 1972, 129, 197. 29. Morrison, J. F. Kinetics of the reversible inhibition of enzyme-catalyzed reactions

by tight-binding inhibitors. Biochim. Biophys. Acta. 1969, 185, 269-286. 30. Aoyagi, T., Miyata, S., Nambo, M., Kojima, F., Matsuzak, M., Istuzuka, M.,

Takeuchl, T., Umezawa, H. Biological activities of leupeptins. J. Antibiot. 1969, 22, 558-568.

31. Hartl. M., Giri, A.P., Kaur, H., Baldwin, I.T. The multiple functions of plant ser-ine protease inhibitors: defense against herbivores and beyond. Plant Signal Be-hav. 2011, 6, 1009-1011.

32. Jorgensen, M., Bauw, G., Welinder, K.G. Molecular properties and activities of tuber proteins from starch potato cv. Kuras. J. Agr. Food Chem. 2006, 54, 9389-9397.

33. Balls, A.K., Ryan, C.A. Concerning a crystalline chymotryptic inhibitor from po-tatoes, and its binding capacity for enzyme. J. Biol. Chem. 1963, 238, 2976-2982.

34. Johnson, R., Ryan, C.A. Wound-inducible potato inhibitor-II genes – enhance-ment of expression of sucrose. Plant Mol. Biol. 1990, 14, 527-536.

35. Turra, D., Bellin, D., Lorito, M., Gebhardt, C. Genotype-dependent expression of specific members of potato protease inhibitor gene families in different tissues and in response to wounding and nematode infection. J. Plant Physiol., 2009, 166, 762-774.

36. Valueva, T.A., Revina, T.A., Gvozdeva, E.L., Gerasimova, N.G., Ozeretskovskaya, O.L. Role of protese inhibitors in potato protection. Russ. J. Bioorg. Chem., 2003, 29, 454-458.

37. Pouvreau, L., Gruppen, H., Piersma, S. R., van den Brock, L. A. M., van Kon-ingsveld, G. M., Voragen, A. G. J. Relative abundance and inhibitory distribution of protease inhibitors in potato juice from c.v. Elkana. J. Agric. Food Chem., 2001, 49, 2864-2874.

38. Walsh, T. A., and Twitchell, W. P. Two Kunitz-Type proteinase inhibitors from potato tubers. Plant Physiol. 1991, 97, 15-18.

39. Suh, S.G., Stiekema, W. J., and Hannapel, D. J. Proteinase-inhibitor activity and wound-inducible gene expression of the 22 kDa potato-tuber proteins. Planta. 1991, 184, 423–430.

40. Valueva, T. A., Revina, T. A., and Mosolov, V. V. Potato tuber protein protinase inhibitors belonging to the Kunitz soybean inhibitor family. Biochem. (Moscow), 1997, 62, 1367–1374.

41. Pouvreau, L., Gruppen, H., van Koningsveld, G. A., van den Brock, L. A. M., and Voragen, A. G. J. The most abundant protease inhibitor in Potato tuber is a serine protease inhibitor from the Kunitz family. J. Agric. Food Chem. 2003, 51, 17, 5001-5005.

42. Mares, M., Meloun, B., Pavlik, M., Kostka, V., Baudys, M. Primary structure of cathepsin D inhibitor from potatoes and its structure relationship to soybean tryp-sin inhibitor family. FEBS Lett. 1989, 251, 94–98.

81

43. Strukelj, B., Pungercar, J., Mesko, P., Barlic Magana, D., Gubensek, F., Kregar, J., Turk, V. Characterization of aspartic proteinase inhibitors from potato at the gene, cDNA and protein levels. Biol. Chem. Hoppe Seyler. 1992, 373, 477-482.

44. Krizaj, M., Drobnic-Kosorok, J., Brzin, J., Jerata, R., Turk, V. The primary struc-ture of inhibitor of cysteine proteinases from potato. V. FEBS Lett. 1993, 333, 15-20.

45. Valueva, T. A., Revina, T. A., Kladnitskaya, G. V., Mosolov, V. V. Kunitz-type proteinase inhibitors from intact and Phytophthora-infected potato tubers. FEBS Lett. 1998, 426, 131-134.

46. Hartl, M., Giri, A.P., Kaur, H., Baldwin, I.T. Serine protease inhibitors specifi-cally defend Solanum nigrum against generalist herbivores but do not influence plant growth and development. Plant Cell, 2010, 22, 4158-4175.

47. Mosolov, V.V., Valueva, T.A. Proteinase inhibitors and their function in plants: a review. Appl. Biochem. Microbiol. 2005, 41, 227-246.

48. Valueva, T. A., Revina, T. A., Mosolov, V. V., Mentele, R. Primary structure of potato Kunitz-type serine protease inhibitor. Biol. Chem. Hoppe Seyler. 2000, 381, 1215-1221.

49. Meulenbroek, E.M., Thomassen, E.A.J., Pouvreau, L., Abrahams, J.P., Gruppen, H., Pannu, N.S. Structure of a post-translationally processed heterodimeric dou-ble-headed Kunitz-type serine protease inhibitor from potato. Acta. Cryst. 2012, D68, 794-799.

50. Ortiz, C., Tellier, C., Williams, H., Stolowich, N.J., Scott, A.I. Diastereotopic covalent binding of the natural inhibitor leupeptin to trypsin: Detection of two interconverting hemiacetals by solution and solid-state NMR spectroscopy. Bio-chemistry. 1991, 30, 10026-10034.

51. Aoyagi, T., Miyata, S., Nanbo, M., Kojima, F., Matsuzaki, M., Ishizuka, M., Takeuchi, T. & Umezawa, H. Biological activities of leupeptins. J. Antibiot. 1969, 22, 558-568.

52. Umezawa, S., Tatsuta, K., Fujimoto, K., Tsuchiya, T., Umezawa H., & Na-ganawa, H. Structure of Antipain, a new Sakagughi-positive product of strepto-myces. J. Antibiot. 1972, 25, 267-270.

53. Chi, C. W., Liu, H., Liu, C., Chibber, B., Baksky, A. K., Castellino, F. The Inhi-bition of enzymatic activity of blood coagulation and fibrinolytic serine proteases by new leupeptin-like inhibitor from Streptomyces griseus. J. Antibiot. 1989, 42, 1506-12.

54. Stewart, J. M. Bradykinin Antagonists: Development and applications. Biopoly-mers. 1995, 37, 143-55.

55. Tyc, F., Vrbova, G. Stabilization of neuromuscular junctions by leupeptin in-creases motor unit size in partially denevated rat muscles. Dev. Brain Res. 1995, 88, 186-93.

56. March, K. L., Wilensky, R. L., Roeske, R. W., Hathaway, D. R. Effects of thiol protease inhibitors on cell cycle and proliferation of vascular smooth muscle. Cells in Culture. Circ. Res. 1993, 72, 413-23.

57. Togushi, K. I., Majae, J., Kataoko, T., Nagal, K. Thiol-protease inhibitors selec-tively inhibit presentation of antigen incorporated through B cell antigen recep-tors. Biochem. Biophys. Res. Commun. 1995, 210 (2), 324-8.

58. Krusslich, H. G., Ingraham, R. H., Skoog, M. T., Wimmer, E., Palla, P. V., Carter, C. A. Activity of purified biosynthetic proteinase of human immunodeficiency virus on natural substrates and synthetic peptides. Natl. Acad. Sci. U.S.A. 1989, 86, 807-11.

82

59. Kawada, M., Umezawa, K. Suppression of in vitro invasion of human fibrosar-coma cells by leupeptin analog inhibiting urokinase-plasmin system. Biochem. Biophys. Res. Commun. 1994, 209 (1), 25-30.

60. Funahashi, T., Shimamura, M., Kocha, T., Fukuda, T., Aoyagi, T. Proportionality of protease activity in malignant cells to their metastic potentials. Biol. Pharm. Bull. 1994, 17, 1118-1120

61. Kawamura, K., Kondo, S., Maeda, K., Umezawa, H. Biological activities of Leupeptin. Chem. Pharm. Bull. 1969, 17, 1902-1909

62. McConnell, R. M., Barnes, G. E., Hoyng, C. F., Gunn, J. M. New leupeptin ana-logues. Synthesis and inhibition data. J. Med. Chem. 1990, 33, 86-93.

63. McConnell, R. M., York, J. L., Frizzell, D., Ezell, C. Inhibition studies of serine and thiol proteases by new leupeptin analogues. J. Med. Chem. 1993, 36, 1084-89.

64. Kurinov, I.V., Harrison, R.W. Two crystal structures of the leupeptin-trypsin complex. Protein Science. 1996, 5, 752-758.

65. Vaccari, M., Argnani, A., Horn, W., Silingardi, P., Giungi, M., Mascolo, M.G., Bartoli, S., Grilli, S., Colacci, A. Effects of the protease inhibitor antipain on cell malignant transformation. Anticancer Res. 1999, 19, 589-596.

66. Gross, G.G., Zenk, M.H. Reduktion aromatischer säuren zu aldehyden und alko-holen im zellfreien system. Eur. J. Biochem. 1969, 8, 413-419.

67. Pozsgay, V., Kubler-Kielb, J. Conjugation methods toward synthetic vaccines. In R. Roy (Ed.), Carbohydrate based vaccines (pp. 36-70). Washington, DC, USA: American Chemical Society.

68. Sletten, E.M., Bertozzi, C.R. Bioorthogonal chemistry: Fishing for selectivity in a sea of functionality. Angew. Chem. 2009, 48, 6974-6998.

69. Lutz, J.F., Börner, H.G. Modern trends in polymer bioconjugates design. Pro-gress in Polymer Science. 2008, 33, 1-39.

70. Canalle, L. A., Lowik, D. W. P. M., van Hest, J. C. M. Polypeptide-polymer bio-conjugates. Chem. Soc. Rev. 2010, 39, 329−353.

71. Briand, V. A., Kumar, C. V., Kasi, R. M. Protein-polymer conjugates. In Ency-clopedia of Polymer Science and Technology, John Wiley & Sons, Inc., 2002.

72. Zhao, W., Liu, F., Chen, Y., Bai, J., Gao, W. Synthesis of well-defined pro-tein−polymer conjugates for biomedicine. Polymer. 2015, 66, Supplement C, A1−A10.

73. Abuchowski, A., van Es, T., Palczuk, N. C., Davis, F. F. Alteration of immuno-logical properties of bovine serum albumin by covalent attachment of polyeth-ylene glycol. J. Biol. Chem. 1977, 252, 3578−3581.

74. Gauthier, M. A., Klok, H.-A. Peptide/protein−polymer conjugates: synthetic strategies and design concepts. Chem. Commun. (Cambridge, U. K.), 2008, 23, 2591−2611.

75. Heredia, K. L., Maynard, H. D. Synthesis of protein-polymer conjugates. Org. Biomol. Chem. 2007, 5 (1), 45−53.

76. Jung, B., Theato, P. Chemical Strategies for the synthesis of protein−polymer conjugates. Bio-synthetic Polymer Conjugates, Schlaad, H., Ed., Springer: Ber-lin, Heidelberg, 2013, 37−70.

77. Jung, S., Kwon, I. Expansion of bioorthogonal chemistries towards site-specific polymer-protein conjugation. Polym. Chem. 2016, 7, 4584−4598.

78. Tietze, L.F., Schröter, C., Gabius, S., Brinck, U., Goerlach-Graw, A., Gabius, H.J. Conjugation of p-aminophenyl glycosides with squaric acid diester to a car-rier protein and the use of the neoglycoprotein in the histochemical detection of lectins. Bioconj. Chem. 1991, 2, 148-153.

83

79. Fisher, S. A., Baker, A. E. G., Shoichet, M. S. Designing peptide and protein modified hydrogels: Selecting the optimal conjugation strategy. J. Am. Chem. Soc. 2017, 139, 7416−7427.

80. Spicer, C. D., Davis, B. G. Selective chemical protein modification. Nat. Com-mun. 2014, 5, 4740.

81. Hersel, U., Dahmen, C., Kessler, H. RGD Modified polymers: Biomaterials for stimulated cell adhesion and beyond. Biomaterials. 2003, 24, 4385−4415.

82. Hermanson, G. T. Bioconjugate Techniques, 2nd ed., Academic Press, Inc.: New York, 2008.

83. Baslé, E., Joubert, N., Pucheault, M. Protein chemical modification on endoge-nous amino acids. Chem. Biol. 2010, 17, 213−227

84. Bellamkonda, R. V., Ranieri, J. P., Aebischer, P. Laminin oligopeptide derivat-ized agarose gels allow 3-dimensional neurite extension in-vitro. J. Neurosci. Res. 1995, 41, 501−509.

85. Borkenhagen, M., Clémence, J. F., Sigrist, H., Aebischer, P. Three-dimensional extracellular matrix engineering in the nervous system. J. Biomed. Mater. Res. 1998, 40, 392−400.

86. Koniev, O., Wagner, A. Developments and recent advancements in the field of endogenous amino acid selective bond forming reactions for bioconjugation. Chem. Soc. Rev. 2015, 44, 5495−5551.

87. Lane, C. F. Sodium cyanoborohydride - a highly selective reducing agent for or-ganic functional groups. Synthesis. 1975, 135–146.

88. Khan, H.A., Sakharkar, M.K., Nayak, A., Kishore, U., Khan, A. Nanoparticles for biomedical applications: An overview. Nanobiomaterials. 2018, 14, 357-384.

89. Winkler, Jochen. 2003. Titanium Dioxide. Hannover, Germany: Vincentz Net-work. p.5. ISBN 978-3-87870-148-4.

90. Rajh, T., Dimitrijevic, N.M., Bissonnette, M., Koritarov, T., Konda, V. Titanium dioxide in the service of the biomedical revolution. Chem. Rev. 2014, 114, 10177-10216.

91. Fujishima, A., Honda, K. Electrochemical photolysis of water at a semiconductor electrode. Nature. 1972, 238, 37-38.

92. Hernandezbattez, A., Gonzalez, R., Viesca, J., Fernandez, J., Diazfernandez, J., MacHado, A., Chou, R., Riba, J. CuO, ZrO2 and ZnO nanoparticles as antiwear additive in oil lubricants. Wear. 2008, 265, 422-428.

93. Padmavathy, N., Vijayaraghavan, R. Enhanced bioactivity of ZnO nanoparticles – an antimicrobial study. Sci. Technol. Adv. Mater. 2008, 9, 3, 035004.

94. Freifelder, D. 1982. Physical Biochemistry: Applications to Biochemistry and Molecular Biology (2nd ed.). WH Freeman. p. 240, 258. ISBN 978-0716714446.

95. Suflet, D.M., Chitanu, G.C., Desbrieres, J. Phosphorylated polysaccharides. 2. Synthesis and properties of phosphorylated dextran. Carbohydr. Polym. 2010, 82, 4, 1271-1277.

96. Tacias-Pascacio, V.G. et al. Dextran aldehyde in biocatalysis: More than a mere immobilization system. Catalysts. 2019, 9, 622, 1-43.

97. Farkas, P., Bystricky, S. Chemical conjugation of biomacromolecules: A mini-review. Chemical Papers. 2010, 64, 683-695.

98. Ishak, M.F., Painter, T.J. Kinetic evidence for hemiacetal formation during the oxidation of dextran in aqueous periodate. Carbohydr. Res. 1978, 64, 189-197.

99. Drobchenko, S.N., Isaeva Ivanova, L.S., Kleiner, A.R., Lomakin, A.V., Kolker, A.R., Noskin, V.A. An investigation of the structure of periodate-oxidized dex-tran. Carbohydr. Res. 1993, 241, 189-199.

84

100.Maia, J., Ferreira, L.S., Carvalho, R., Ramos, M.A., Gil, M.H. Synthesis and char-acterization of new injectable and degradable dextran-based hydrogel. Polymer. 2005, 46, 9604-9614.

101.Lukyanov, A.N., Sawant, R. M., Hartner W. C., Torchilin, V.P. PEGylated dex-tran as long-circulating pharmaceutical carrier. J. Biomater. Sci. Polym. Ed. 2004, 15, 5, 621-630.

102.Hermanson, G.T. 2013. Bioconjugate techniques, 3rd Ed. USA: Elsevier Inc. 103.Axen R., Porath J., Ernback S. Chemical coupling of peptides and proteins to

polysaccharides by means of cyanogens halides. Nature. 1967, 214, 1302-1304. 104.Van Holde, K.E., Johnson, W.C., Ho, P.S. 2006. Principles of Physical Biochem-

istry. USA: Pearson Education Inc. p. 352. 105.Folta-Stogniew, E., Williams, K. R. Determination of molecular masses of pro-

teins in solution: Implementation of an HPLC size exclusion chromatography and laser light scattering service in a core laboratory. J. Biomol. Technol. 1999, 10, 51–63.

106.Kendrick, B. S., Kerwin, B. A., Chang, B. S., Philo, J. S. Online size-exclusion high-performance liquid chromatography light scattering and differential refrac-tometry methods to determine degree of polymer conjugation to proteins and pro-tein-protein or protein-ligand association states. Anal. Biochem., 2001, 299, 136–146.

107.Folta-Stogniew, E. Oligomeric states of proteins determined by size-excluion chromatography coupled with light scattering, absorbance, and refractive index detectors. Methods Mol. Biol. 2006, 328, 97–112.

108.Wyatt, P. J. Light scattering and the absolute characterization of macromolcules. Anal. Chim. Acta272, 1993.

109.Takagi, T. Application of low-angle laser light scattering detection in the field of biochemistry: review of recent progress. J. Chromatogr. A. 1990, 506, 51–63.

110.Wen, J., Arakawa, T. & Philo, J. S. Size-exclusion chromatography with on-line light scattering, absorbance, and refractive index detectors for studying proteins and their interactions. Anal. Biochem. 1996, 240, 155–166.

111.Mogridge, J. Using light scattering to determine the stoichiometry of protein complexes. Methods Mol. Biol. 2015, 1278, 233–238.

112.Fosgerau, K., Hoffmann, T. Peptide therapeutics: Current status and future direc-tions. Drug Discovery Today. 2015, 20, 122-128.

113.Craik, D.J., Fairlie, D.P., Liras, S., Price, D. The future of peptide-based drugs. Chemical Biology & Drug Design. 2013, 81, 136-147.

114.Okarvi, S. Peptide-based radiopharmaceuticals and cytotoxic conjugates: Poten-tial tools against cancer. Cancer Treatment Reviews. 2008, 34, 13-26.

115.Mercer, D.K., O’neil, D.A. Peptides as the next generation of anti-infectives. Fu-ture Medicinal Chemistry. 2013, 5, 315-337.

116.Gori, A., Longhi, R., Peri, C., Colombo G. Peptides for immunological purposes: Design, strategies and applications. Amino Acids. 2013, 45, 257-268.

117.Mäde, V., Els-Heindl, S., Beck-Sickinger, AG. Automated solid-phase peptide synthesis to obtain therapeutic peptides. Beilstein J. Org. Chem. 2014, 10, 1197-1212.

118.Atherton, E., Sheppard, R. Solid Phase Peptide Synthesis: A Practical Approach. Oxford: IRL Press, 1989.

119.Labadie, J.W. Polymeric supports for solid phase synthesis. Current Opinion in Chemical Biology. 1998, 2, 346-352.

120.Merrifield, B. The role of the support in solid phase peptide synthesis. British Polymer Journal. 1984, 16, 173-178.

85

121.Bonkowski, B., Wieczorek, J., Patel, M., Craig, C., Gravelin, A., Boncher, T. Basic concepts of using solid phase synthesis to build small organic molecules using 2-chlorotrityl chloride resin. Chem. Inform. 2015, 1, 1-4.

122.Shelton, P.T., Jensen, K.J. Linkers, resins, and general procedures for solid phase peptide synthesis. In: Peptide Synthesis and Applications. New York, Springer, 2013. 23-41.

123.Behrendt, R., White, P., Offer, J. Advances in Fmoc solid-phase peptide synthe-sis. J. Peptide Sci. 2016, 22, 4-27.

124.Moss, J.A. Guide for resin and linker selection in solid-phase peptide synthesis. Curr. Protoc. Protein Sci. 2005, 40, 1871-1879.

125.Duncan, M.W., Roder, H., Hunsucker, S.W. Quantitative matrix-assisted laser desorption/ionization mass spectrometry. Brief. Funct. Genomic. Proteomics. 2008, 7, 355-370.

126.Gross, J.H. Mass spectrometry: a textbook. Springer Science & Business Media, 2006.

127.Boesl, U. Time - of - flight mass spectrometry: Introduction to the basics. Mass Spectrom. Rev. 2017, 36, 86-109.

128.Ho, C.S., Chan, M.H.M., Cheung, R.C.K., Law, L.K., Lit, L.C.W. Suen, M.W.M. Tai, H.L. Electrospray ionization mass spectrometry: Principles and clinical ap-plications. Clin. Biochem. Rev. 2003, 24, 3–12.

129.Pitt, J. Principles and applications of liquid chromatography-Mass spectrometry in clinical biochemistry. Clin. Biochem. Rev. 2009, 30, 19–34.

130.Makarov, A. Electrostatic axially harmonic orbital trapping: a high-performance technique of mass analysis. Anal. Chem. 2000, 72, 1156-1162.

131.Makarov, A., Denisov, E., Kholomeev, A., Balschun, W., Lange, O., Strupat, K., and Horning, S. Performance evaluation of a hybrid linear ion trap/orbitrap mass spectrometer. Anal. Chem. 2006, 78, 2113-2120.

132.Borumand, M., Sibilla, S. Daily consumption of the collagen supplement Pure Gold Collagen® reduces visible signs of aging. Clin. Interv. Aging. 2014, 9, 1747-1758.

133.Dattagupta, J.K., Podder, A., Chakrabarti, C., Sen, U., Mukhopadhyay, D. Dutta, S.K., Singh, M. Refined crystal structure (2.3 angstrom) of a double-headed winged bean alpha-chymotrypsin inhibitor and location of its second reactive site. Proteins: Struct., Funct., Genet. 1999, 35, 321-331.

134.Gahloth, D., Sharma, K. Identification and partial characterization of trypsin in-hibitory activity in seed of some fruit plants. J. Plant Biochem. Biotechn. 2010, 19, 2, 235-237.

135.Bao, R., Zhou, C-Z., Jiang, C., Lin, S-X., Chi, C-W., Chen, Y. The ternary struc-ture of the double-headed arrowhead protease inhibitor API-A complexed with two trypsins reveals a novel reactive site conformation. J. Biol. Chem. 2009, 284, 26676-26684.

136.Micheelsen, P.O., Vévodová, J., De Maria, L., Ostergaard, P.R., Friis, E.P., Wil-son, K. Skjøt, M. Structural and mutational analyses of the interaction between the barley alpha-amylase/subtilisin inhibitor and the subtilisin savinase reveal a novel mode of inhibition. J. Mol. Biol. 2008, 380, 681–690.

137.Suzukake, K., Hayashi, H., and Hori, M.: Biosynthesis of leupeptin. III. Isolation and properties of an enzyme synthesizing acetyl-L-leucine. J. Antibiot. 1980, 33, 857-862.

138.Suzukake, K., Fukiyama, T., Hayashi, H., Hori, M., and Umezawa, H.: Biosyn-thesis of leupeptin. II. Purification and properties of leupeptin acid synthetase. J. Antibiot. 1979, 32, 523-530.

86

139.Suzukake, K., Hori, M., Tamemasa, O., Umezawa, H.: Purification and properties of an enzyme reducing leupeptin acid to leupeptin. Biochim. Biophys. Acta. 1981, 661, 175-181.

140. Moulin, A., Martinez, J., Fehrentz, J.A. Synthesis of peptide aldehydes. J. Pept. Sci. 2007, 13, 1-15.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1888

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-398909

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2020