Master’s Thesis, 30 hp | Chemical-Biological Engineering Program: Master of Science in Protein Sciences Spring term 2019 | LITH-IFM-x-EX—19/3726--SE
Understanding the Involvement of
Leukocyte Cell-derived
Chemotaxin 2 (LECT2) in
Amyloidosis
Lisa-Marie Erlandsson
Examinator, Prof. Per Hammarström
Datum
Date
2019-09-20
Avdelning, institution
Division, Department
Department of Physics, Chemistry and Biology
Linköping University
URL för elektronisk version
ISBN
ISRN: LITH-IFM-x-EX--19/3726--SE
_________________________________________________________________
Serietitel och serienummer ISSN
Title of series, numbering ______________________________ Språk Language Svenska/Swedish Engelska/English ________________ Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport _____________ Titel Title
Understanding the Involvement of Leukocyte Cell-derived Chemotaxin 2 (LECT2) in Amyloidosis
Författare Author
Lisa-Marie Erlandsson
Nyckelord Keyword
LECT2 amyloidosis, disulfide bonds, metal-binding, protein expression, refolding, purification, fluorimetry, CD, aggregation Sammanfattning
Abstract
Leukocyte cell-derived chemotaxin 2 (LECT2) is a zinc-binding multi-functional protein comprising three disulfide bonds, that is involved in multiple disorders of worldwide concern. Recently LECT2 was found to be involved in amyloidosis (ALECT2) and is believed to be the third most common form of systemic amyloidosis. The disease progression of ALECT2 is relatively slow, and the aggregation assembly is foremostly associated with the kidneys and the liver, but also other organs in the later onset of the disease. This study involved developing a protocol for
producing His6-TEV-LECT2 including expression in E.coli BL21(DE3), refolding, and purification. The protocol resulted in a sufficient yield for initial measurements for characterization and biophysical analysis with the following methods: mass spectrometry (MS), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), circular dichroism (CD), and fluorimetry. The produced protein was characterized as LECT2 predominantly in its oxidized form. A brief biophysical analysis was made where LECT2 started to unfold already at physiological temperature with a midpoint at 50°C. Additionally, under chemical denaturation LECT2 unfolded with a midpoint of 3 M urea in a cooperative transition without any intermediates. Further on, wavelengths for monitoring the unfolding and the aggregation simultaneously were identified. The unfolding process occurred under 20 sec in 6 M urea and correlates with a double-exponential model. The LECT2 aggregates resemble protofibril-like structures and aggregates species from monomer up to hexamer were found, suggesting simple monomeric addition towards a growing fibril as the aggregation mechanism. The content of aggregates in the sample was notably decreased upon disulfide bond reduction highlighting the importance of further investigating the role of the disulfide bonds in the destabilization and aggregate formation of LECT2.
Understanding the Involvement of Leukocyte
Cell-derived Chemotaxin 2 (LECT2) in Amyloidosis
Author
L-M. Erlandsson
lisamerlandsson@gmail.comDepartment of Physics, Chemistry and Biology Linköping University
Supervisors
Prof. J. Kelly, Scripps Research Institute, La Jolla, USA
Assoc. Prof. E. Powers, Scripps Research Institute, La Jolla, USA
Examiner
Prof. P. Hammarstrom, LiU, Linköping, Sweden
DateAbstract
Leukocyte cell-derived chemotaxin 2 (LECT2) is a zinc-binding multi-functional protein comprising three disulfide bonds, that is involved in multiple disorders of worldwide concern. Recently LECT2 was found to be involved in amyloidosis (ALECT2) and is believed to be the third most common form of systemic amyloidosis. The disease progression of ALECT2 is relatively slow, and the aggregation assembly is foremostly associated with the kidneys and the liver, but also other organs in the later onset of the disease. This study involved developing a protocol for producing His6-TEV-LECT2 including expression in E.coli BL21(DE3), refolding, and purification.
The protocol resulted in a sufficient yield for initial measurements for characterization and biophysical analysis with the following methods: mass spectrometry (MS), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), circular dichroism (CD), and fluorimetry. The produced protein was characterized as LECT2 predominantly in its oxidized form. A brief biophysical analysis was made where LECT2 started to unfold already at physiological temperature with a midpoint at 50°C. Additionally, under chemical denaturation LECT2 unfolded with a midpoint of 3 M urea in a cooperative transition without any intermediates. Further on, wavelengths for monitoring the unfolding and the aggregation simultaneously were identified. The unfolding process occurred under 20 sec in 6 M urea and correlates with a double-exponential model. The LECT2 aggregates resemble protofibril-like structures and aggregates species from monomer up to hexamer were found, suggesting simple monomeric addition towards a growing fibril as the aggregation mechanism. The content of aggregates in the sample was notably decreased upon disulfide bond reduction highlighting the importance of further investigating the role of the disulfide bonds in the destabilization and aggregate formation of LECT2.
Acronymes and Abbreviations
AA Serum Amyloid A AmyloidosisAC Affinity Chromatography AL Light Chain Amyloidosis ALECT2 LECT2 Amyloidosis ALS Amyotrophic lateral sclerosis BME β-mercaptoethanol
CD Circular Dichronism CHO Chinese Hamster Ovary DLC Dynamic Light Scattering DTT Dithiothreitol
FTIR Fourier-Transform Infrared GSH Glutathione
GSSG Glutathione Disulfide IB Inclusion Body
IEX Ion Exchange Chromatography
IMAC Immobilized Metal Affinity Chromatography IPTG Isopropyl β-D-1-thiogalactopyranoside
LC Liquid Chromatography
LECT2 Leukocyte Cell-derived Chemotaxin 2
MALDI Matrix-Assisted Laser Desorption/Ionization MBP Maltose-Binding Protein
MS Mass Spectrometry
NIH National Institues of Health
pKa Acidity (the pH where the acid is protolysed to 50%) SAP Serum Amyloid P
SEC Size-Exclusion Chromatography
SDS-PAGE Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis SOD1 Superoxide dismutase 1
SUMO Small Ubiquitin-Like Modifier TCEP Tris(2-carboxyethyl)phosphine ThT Thioflavin T
Contents
1 Introduction 1
2 Impact of the Study and Ethical Concern 2
3 Theory and Methology 2
3.1 Scientific Background . . . 2
3.1.1 Leokocyte Cell-Derived Chemotaxin 2 . . . 2
3.1.2 Thiols and Disulfide Bond Formation . . . 3
3.2 Methology . . . 5
3.2.1 Protein Expression and Refolding . . . 5
3.2.2 Liquid Chromatography . . . 5
3.2.3 Mass Spectrometry . . . 6
3.2.4 Polyacrylamide Gel Electrophoresis . . . 6
3.2.5 Dialysis . . . 7
3.2.6 Fluorescence Spectroscopy . . . 7
3.2.7 Circular Dichronism Spectroscopy . . . 7
3.2.8 Atomic Force Microscopy . . . 8
4 Methods 8 4.1 Purification and Refolding . . . 8
4.2 Purification of Refolded LECT2 . . . 9
4.3 Mapping of Disulfide bonds . . . 9
4.4 Thermodynamics . . . 10
4.5 Chemical Denaturation and Unfolding Kinetics . . . 10
4.6 LECT2 Aggregation . . . 11
5 Results and Discussion 11 5.1 LECT2 Expression, Refolding, and Purification . . . 11
5.2 Characterization of Refolded LECT2 . . . 13
5.3 Thermodynamics . . . 16
5.4 Chemical Denaturation and Unfolding Kinetics . . . 16
5.5 LECT2 Aggregation . . . 18
6 Conclusion and Future Prospect 18 7 Acknowledgements 22 Appendix A Process i A.1 Project Plan and Analysis . . . ii
Appendix B Construct Design and Production iii B.1 Construct Sequences . . . iii
B.2 Inverse Fusion Cloning and Tandem MBP-SUMO Fusion Expression . . . iii
B.3 Purification of LECT2 Comprising a SUMO Cleavage Site . . . v
B.4 Evaluation of the LECT2 Constructs and Tag Cleavage . . . vii
Appendix C Tables viii C.1 Buffers . . . viii
CONTENTS 1. INTRODUCTION
1.
Introduction
Amyloidosis is characterized by destabilization and self-assembly of specific proteins into massive insolu-ble non-branching fibrillar deposits. They are stained by Congo red, contains the abundant amyloid-binding protein serum amyloid P (SAP) and are enhanced by seeding [1–3]. About 30 different proteins are found to be amyloidogenic and tissues are affected in a protein-dependent manner [1]. The underlying amyloid disor-ders can either be localized in particular organs or be systemic, i.e. affecting different organs throughout the body [4]. Symptoms vary upon which tissues and or-gans are affected and the most common include spleen, heart, kidneys, liver, lungs, and skin.
Leukocyte cell-derived chemotaxin 2 amyloidosis (ALECT2) is a systemic aggregation disorder that was first discovered in a renal biopsy from a patient suffer-ing from a nephrotic syndrome in 2008 [5]. ALECT2 is estimated to be the most common form of systemic amy-loidosis after light chain amyamy-loidosis (AL) and serum amyloid A amyloidosis (AA) [6], and is especially found in patients with Mexican [6, 7], Egyptian [8], and South Asian descent [9]. The disorder is usually presented with renal insufficiency with or without proteinuria [10]. It is believed to be present in renal amyloid samples up to 10% [10, 11] and in hepatic amyloidosis cases up to 25% [12]. The disease progression of ALECT2 is relatively slow and except for the kidneys and the liver, aggregates are found in other bilateral organs in the later stages of disease.
In order to understand the propensity of proteins to form amyloids, investigations of the unfolding mecha-nism give insight regarding conditions that can favor the stabilization/destabilization of the protein. There is a broad range of biophysical tools and techniques that have been contributing to understanding the structural features, kinetics, and dynamics of proteins, e.g. fluores-cence, circular dichroism (CD), and Fourier-transform infrared (FTIR) spectroscopy, dynamic light scattering (DLS) among others [13]. Except for Congo red stain-ing, thioflavin T (ThT) binds to amyloids, which is commonly used for monitoring protein aggregation. Since ALECT2 was recently discovered, not much is known about its characteristics and involvement in amy-loidosis. In this study, I sought to analyze the overall
stability of the I40V polymorph of LECT2 focusing on the importance of the disulfide bond formation using tryptophan fluorescence and CD spectroscopy under chemical and thermal denaturation.
Moreover, obtaining a soluble fraction when express-ing LECT2 has been shown to be limited, thus multi-ple strategies involving refolding of the insoluble frac-tion have been published including 14 days purifica-tion and oxidative refolding [14], high hydrostatic (200 MPa) pressure [15] of LECT2 expressed in E.coli, or size-exclusion chromatography (SEC) refolding under decreasing urea concentration [16]. In this study, an alternative way of expressing LECT2 was analyzed in order to obtain a soluble protein expression, see Ap-pendix B. The concept was based in expressing LECT2 in fusion with the maltose-binding protein (MBP) pre-viously found to increase protein solubility and the small ubiquitin-like modifier (SUMO) that provides an accurate cleavage site [17, 18].
This is a Master’s thesis work of the chemical-biological engineering program provided at Linköping Univer-sity, Sweden, in collaboration with Scripps Research in La Jolla, USA, addressing production and stability analysis of LECT2. The overall aim is divided in the view from six milestones e.g. involving the produc-tion of a DNA construct followed by establishing a method for expression, refolding and purification re-sulting in a high-yield production protocol for LECT2, see below. Three different constructs were produced and compared in terms of expression yield and solubility in the E.coli strains BL21(DE3) and Rosetta-Gami, and min-imal protein loss under purification see Appendix B. The production work was then followed up by evaluating the folding of LECT2 and the disulfide bond formation within the protein. Furthermore, I investigated the bio-physical stability to give insight to target and stabilize aggregation-prone species of LECT2. Workflow, project plan and process analysis of the study are presented in Appendix A.
1. Primer design and construct production
2. Protein expression and optimization
3. Protein purification and refolding
2. IMPACT OF THE STUDY AND ETHICAL CONCERN CONTENTS
4. Protein characterization and mapping of disulfide bonds
5. Biophysics: thermal/chemical denaturation and unfolding kinetics
6. Aggregation analysis
2.
Impact of the Study and Ethical
Concern
The number of amyloidosis cases increases every year especially as people tend to live longer [19]. Even though the field of amyloidogenic disorders is compre-hensively studied, techniques for diagnosis are limited and there are only a few treatments available. Currently, some amyloidogenic diseases can only be fully diag-nosed post mortem [20], and many patients are misdiag-nosed for secondary symptoms [4]. Further on, most treatments available today are either organ transplan-tation or drugs for treating symptoms rather than the disorder itself. The few treatments available are both expensive [21] and have a fraction of non-responders [22]. Thus, contributing studies to mapping pathways and treatment targets are of high importance.
The insights in terms of the structure and the path-ways LECT2 is involved in is not limited to the field of amyloidosis but contribute to a general understand-ing of the protein. In addition to renal and hepatic amyloidosis, LECT2 seems to be involved in multiple disorders of worldwide concern, e.g. type-II diabetes [23], rheumatoid arthritis [24], hepatocellular carcinoma [25], and intestinal tumourigenesis [26]. This study has been performed in vitro in E.coli and is not causing any ethical concern on an experimental level.
3.
Theory and Methology
3.1.
Scientific Background
This subsection contains information regarding the essen-tials about LECT2, as well as, the basics about thiols and disulfide bond formation.
3.1.1 Leokocyte Cell-Derived Chemotaxin 2
Mature LECT2 consists of 133 amino acids (14.572 kDa and pI 9.4, ExPASy Compute pI/Mw tool), excluding
its 18 amino acid signal peptide, and is encoded by the LECT2 gene located at Chr 5q31.1 in close proximity to genes encoding for cytokines, e.g. IL—4, IL-5 and IL-9 [27]. As many aggregation-prone proteins, the major structural element is β-sheet and has been found deposited with its primary sequence intact, suggesting that the structure rather than degradation makes it amyloidogenic [28]. However, even though degradation may not be crucial for fibril formation, it is believed that a decreased catabolism, as well as, the raised expression of LECT2 is the cause of the increase in local tissue concentrations [4].
The structure of LECT2 is mainly composed of dis-ordered loops held together by three disulfide bonds (Cys25-Cys60; Cys36- Cys41; Cys99-Cys142) [29]. Nine
β-strands form two sheets including a large
six-stranded antiparallel β-sheet and a small three-six-stranded
β-sheet. The former sheet is the central feature of the
structure facing a large groove which end coordinates a bivalent zinc atom by the side chains of His35, Asp39, and His120 and a water molecule.
Furthermore, LECT2 is the first mammalian protein with a confirmed M23 metalloendopeptidase fold (fig-ure 1) [30]. The protease activity of the M23 family is zinc-dependent that gets its glycylglycine peptide bond specificity from conserved Zn2+binding amino acids in HXH and HXnD motifs [31, 32]. LECT2 is comprised by a conserved Zn2+ coordinated conformation but has been found to be proteolysis inactive, which is thought to be because of the replacement of a catalytically active histidine residue at position 86 into a tyrosine [30]. In addition to this, the potential binding pocket is hin-dered by the Zn2+-binding site due to an N-terminal intrachain loop.
Moreover, recombinant LECT2 proteins of both human and mouse have been shown to oligomerize by disul-fide interactions in vitro when expressed in a Chinese hamster ovary (CHO) cell line [33]. However, oligomer-ization was prevented in the presence of zinc ions indi-cating that binding of its cofactor not only has a crucial role in the stability of LECT2 but for its biological func-tions as well.
LECT2 is believed to be exclusively synthesized by the liver followed by being rapidly secreted into the
circu-CONTENTS 3. THEORY AND METHOLOGY
Fig. 1: Schematic representation of LECT2 including a bound zinc ion (PDB: 5b0h) visualizing its M23 fold in car-toon (A), and with disulfide bonds (pink) in ribbon (B). The structure of LECT2 is composed of disordered loops held to-gether by the three disulfide bonds with β-sheet as the ma-jor structural element. Position 40 is the site of the Ile/Val polymorphism which is located in the large sheet of the two present β-sheets in a buried position facing into the interior of the protein in close proximity to the coordinated Zn2+.
Structure published by Zheng et al. [30].
lation [34]. As its name implies LECT2 has chemotactic activity, and beyond that numerous other immunomod-ulatory pathways where LECT2 e.g. has been shown to interact with the C-type lectin receptor [35], and to regulate the natural killer T cells in the liver in a negative manner [36]. Furthermore, LECT2 also has been found to be involved in multiple physiological processes such as hepatocyte activity regulation [37, 38], neuronal development [39], bone growth [24], and glu-cose metabolism [23].
Aggregates are usually predominantly found in the in-terstitial cortex of the kidneys, and as a rare globular deposit in the periportal parenchyma or the portal triad of the liver [4]. Thus, in contrast to other amyloido-genic disorders, ALECT2 show several characteristic
morphological patters rather than one distinct protein deposition pattern in a specific compartment [4]. In ad-dition to this, there is no sign of medullary involvement in the kidneys typically found in other systemic amy-loidogenic disorders. LECT2 deposits are both highly congophilic and strongly birefringent [9], indicating a highly ordered fibrillar cross-β-sheet aggregates in contrast to diffuse amorphous aggregates.
There is no evidence for underlying mutations of LECT2, but there are two polymorphic forms, due to codon exchange from ATC to GTC, resulting in a valine instead of an isoleucine residue at position 40 of mature LECT2 [28]. Position 40 is located in the large β-sheet in a buried position facing into the interior of the pro-tein, but also in close proximity to the zinc-binding site. Virtually all patients diagnosed with ALECT2 that have been genotyped have been shown to be homozygous for the I40V polymorph, i.e. the G/G allele, of LECT2. Thus, the I40V polymorph is thought to destabilize and impart an amyloidogenic propensity of the protein. The localized inflammation in the kidneys in patients suffer-ing from ALECT2 reminds of chronic tubulointerstitial nephritis and has been shown to be connected with secluded synthesis and elevation of the amyloidogenic G/G allele, resulting in I40V LECT2 [40]. Interestingly, the disease-associated I40V allele is more frequent than the I40 allele, 0.6 compared to 0.4, respectively [5]. Moreover, recent studies have shown that patients di-agnosed with ALECT2 in the liver have shown an in-tense and uniform hepatocellular expression of LECT2, suggesting a correlation between overexpression and ALECT2 amyloidosis [12], but the plasma levels of LECT2 have been found normal [7]. Further on, in-creased serum levels of LECT2 has been shown to cor-relate with e.g. hepatocellular carcinoma [41], acute liver damage [37], and diabetes [42], but LECT2’s exact physiological role remains unknown.
3.1.2 Thiols and Disulfide Bond Formation
Disulfide bonds are post-translational modifications formed between cysteine residues as an essential com-ponent of the three-dimensional structure of a wide set of proteins. Meanwhile, these sulfur-sulfur linkages can be altering and necessary for structure, stability, and function of proteins, they can be disruptive and
3. THEORY AND METHOLOGY CONTENTS
Fig. 2: Schematic representation of the thiol-disulfide interconversion in a protein mediated by glutathione disulfide (GSSG) throughout a nucleophilic attack on one of the sulfur atoms. This results in a linear arrangement of the two sulfur atoms within the protein and the sulfur atom in glutathione in the transition state. As a similar nucleophilic attack occurs, this time on the second sulfur within the protein, the glutathione leaves. The resulting products are disulfide bond and two reduced glutathiones (GSHs).
deleterious for others. The thiol-disulfide exchange and homeostasis are carried out by a redox reaction (figure 2), where the former is the reduced state and the latter the oxidized state, respectively. The redox reaction is mediated by the coenzyme glutathione (GSH), which in its oxidized and dimerized form is termed glutathione disulfide (GSSG) [43].1
The interconversion takes place throughout a nucle-ophilic attack on one of two sulfur atoms by a thiolate resulting in a disulfide. The interconversion is depen-dent on the linear arrangement of the three sulfur atoms in the transition state. When it comes to proteins, one sulfur of the disulfide bond is usually more accessible to the reductant than the other. Hence, generating a single mixed intermediate comprised of both the thiol and the thiolate. Proteins containing sulfur atoms with similar accessibility will then be more reactive based on their deprotonated state of the associated thiol, to form thiolate, thus their nucleophilic capacity is dependent on acidity (pKa) [43].
Thiols of biological molecules, especially proteins, come with an extensive span of pKa values, starting from about three and goes as high as up to eleven, due to H-bonds, electrostatic effects from neighboring dipoles and charges, and solvation. Thus, highly impacting the deprotonation equilibrium, where a low pKa increases 1Since this subsection exclusively contains general knowledge
about thiols and disulfide bonds, it is entirely based on the National Institutes of Health (NIH) review article by Winther and Thorpe (2014) [43].
the population of thiolates to disulfides. Although most proteins have their on/off switch at normal cellular conditions, some enzymes have remarkably low pKa in order to suppress oxidative side reactions that could affect catalysis. Beyond reactivity, pKa can affect nucle-ophilicity by tuning the equilibrium by improving the leaving properties of glutathione.
Most of the proteins comprising disulfide bonds are found extracellularly. However, intracellularly GSH is in excess due to the flavin-dependent enzyme glu-tathione reductase maintaining a high reduced-to-oxidized ratio. The catalytic cycle is based on NADPH acting as an electron donor and reducing FAD to FADH2. Furthermore, disulfide bonds can also be re-duced directly by thioredoxin, another flavin-dependent enzyme, throughout the same process.
An effective way of reducing protein samples in vitro in a controlled manner is by adding reducing agents, e.g. β-mercaptoethanol (BME), dithiothreitol (DTT), or tris(2-carboxyethyl)phosphine (TCEP). BME acts with a comparable mechanism to that of GSH, meanwhile, DTT forms an intramolecular disulfide in its oxidized state since it contains two thiols. TCEP, which is a phosphine, on the other hand, forms a phosphonium ion sulfur adduct when attacking one of the two sulfurs. This followed by hydrolysis resulting in an irreversible phosphine oxide contrasting thiol-based reductants that normally are needed in large excess and usually must be removed or quenched.
CONTENTS 3. THEORY AND METHOLOGY
3.2.
Methology
Based on the six milestones comprising the aim of this study, this subsection gives a scientific background of the methods and techniques used in this thesis work.
3.2.1 Protein Expression and Refolding
To obtain a protein of interest from a bacterial expres-sion system, a DNA construct is synthesized comprising the gene of interest, a promoter, a terminator, etc., and is then transformed into the cell, normally the E.coli strain BL21(DE3). Among others, this particular strain can be cultivated to high concentrations in inexpensive media. Furthermore, it is lacking multiple proteases and is employed with the T7 expression system, making it an advantageous tool within synthetic biology [44]. Normally, the T7 expression system is used together with a lac promoter, which enables a controlled way of overexpressing the protein of interest with a chemi-cal inducer, e.g. Isopropyl β-D-1-thiogalactopyranoside (IPTG).
However, expression of eukaryotic proteins containing post-translational modifications, e.g. disulfide bonds, phosphorylations, and glycosylations, in prokaryotes can be challenging [45]. To overcome this, E.coli strains such as Rosetta-gami has been engineered. In addition to having the same features as BL21, this strain has a low codon bias and a mutation in the thioredoxin re-ductase and glutathione rere-ductase, respectively. Hence, creating an oxidizing environment of the cytoplasm, which stabilizes disulfide bond formation in the protein.
Rather than a soluble protein expression, heterologous expression of genes that are foreign to E. coli often causes the desired protein to end up in the insoluble fraction, i.e. inclusion bodies [16]. Thus, the protein of interest must be denatured and renatured in a multiple-step sample preparation in order to become active. Re-cently, an integrated size-exclusion chromatography refolding system under a denaturation gradient was de-veloped to both increase the yield of the refolding but also to inhibit aggregation [46]. Further on, adding argi-nine during the refolding process has been shown to stabilize folding intermediates by enhancing sulfhydryl reshuffling, thereby, affording a higher yield of natively folded protein [47]. Refolding under a denaturing
gra-dient of LECT2 has been performed with arginine and a one-to-one ratio of GSH:GSSG, to enhance disulfide bond formation, added to the solution by Zhang et al. in 2011 [16]. The refolding and purification of LECT2 from inclusion bodies in this study are based on the method of Zhang et al. In parallel, a method result-ing in a soluble fraction of LECT2 was produced and investigated, see Appendix B.
3.2.2 Liquid Chromatography
Liquid chromatography (LC) techniques offer protein purification from a crude sample according to differ-ences in specific properties with elution being detected using ultraviolet (UV), electrical conductivity, or flu-orescence based on the properties of the protein [48]. Affinity chromatography (AC) is based on recognition of a tag of recombinant proteins, e.g. histidine (His) tag or maltose-binding protein (MBP), where one pu-rification step usually is sufficient. Proteins can also be purified based on their natural characteristics such as size and charge with size-exclusion (SEC) and ion-exchange (IEX) chromatography, respectively. SEC can be tailored to fit separation of proteins of interest and is suited for purification of proteins sensitive to the concentration of cofactors or metal ions, changes in pH, or disruptive environments[49].
Meanwhile, proteins migrate through the resin with different pace because they take different paths relating to their size in SEC, proteins bind to the resin in AC and IEX [48]. Hence, the latter two need to be eluted with a compound that outcompetes their binding. The elution in AC is dependent on the tag [48] and the elution in IEX is obtained either by changing the pH or by increasing the ionic strength [50]. In the case of the latter, the increase in ionic strength will cause the proteins with the highest net charge at the selected pH to be eluted last and a high concentration of salt e.g. NaCl is needed.
Different constructs of LECT2 expressed with different tags are tested in this study including a His6-TEV, a
His6-SUMO, and a His6-SUMO-MBP purification tag,
respectively. Since LECT2 is one of the few alkaline proteins of the cell, i.e. has a positive net charge at physiological pH, it can be purified with a negatively charged S-sepharose resin. Hence, IEX is a powerful
3. THEORY AND METHOLOGY CONTENTS
separation tool for purifying LECT2.
3.2.3 Mass Spectrometry
The principle behind mass spectrometry (MS) is the sep-aration of ionized compounds, even large compounds like proteins, followed by qualitative and quantitative detection by mass-to-charge ratio (m/z) and abundance, respectively. The analyte, i.e. the sample under analysis, can either be ionized by electric fields, temperature, or by impacting energetic ions, electrons, or photons. The mass spectrometer consists of an ion source, a mass analyzer, and a detector, where each compartment can be based on slightly different techniques depending on the device in question [51].
Matrix-assisted laser desorption/ionization (MALDI) and Electrospray ionization (ESI) are two of the more common ionization techniques. Both MALDI and ESI are soft ionization techniques meaning that the ana-lyte is ionized without fragmentation. In MALDI, the analyte is mixed with a light-absorbing matrix and so-lidified, then the sample is desorbed and ionized upon radiation by a laser operating with wavelengths from UV to IR. The ionization step is usually followed by acceleration of the ions created under high vacuum con-ditions towards the detector where the time of flight (TOF) is translated to m/z. In contrast, ESI transfer ions from solution to the gas phase as the sample is sprayed with high voltage into a condensed phase. It all starts at atmospheric pressure and continues constantly into the high vacuum of the mass analyzer where its m/z is detected. ESI may produce multiple-charged ions, thereby, extends the m/z scale by the number of charges allowing for the widespread biological and biomedical applications [51].
The coupling of LC and MS is a powerful tool with a wide variety of interfaces and applications. In this way, individual fractions from the chromatography can be analyzed separately as they are sequentially eluted and then ionized for mass detection. The integrated LC-MS approach lowers the complexity of the protein sample markedly as each peak from the chromatogram generates its own MS spectrum (figure 3) [51]. LC-MS will, in this study, be used for identifying trypsinized fragment of LECT2 in order to map its three disulfide bonds of the protein. That is, to determine if the sample
Fig. 3: Schematic representation of the mass spectrometry (MS) spectra generated from proteins separated with chro-matography. As the proteins are eluted they are instantly ionized and accelerated into the mass analyzer. This setup generates a simplistic view of the sample since each mass spectrum can be traced back to the chromatogram with the retention time. Thus, the mass spectrum reveals the con-tent of the responding peak of the chromatogram. Except for mass-to-charge ratio, and intensity, eventual overlapping species can be detected.
protein is oxidized and natively folded. 3.2.4 Polyacrylamide Gel Electrophoresis
In order to analyze molecular weights and the com-plexity of samples, DNA or proteins can be separated alongside a mixture of reference weight marker with an electric current over a gel in agarose and polyacry-lamide, respectively. The separation of proteins can be based on the native occurrence, size, isoelectric point among other characteristics. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) con-tains an initial step where the sample is mixed with the negatively charged surfactant SDS, which denatures secondary and non-disulfide linked tertiary structures and interacts with hydrophobic patches, thereby, gen-erating an overall negative charge. Thus SDS-PAGE offers protein separation based on size exclusively, com-pared to for example a native PAGE where amino acids and conformational differences generate a variation in
CONTENTS 3. THEORY AND METHOLOGY
charge and shape that affects the migration pattern of proteins [52].
There are multiple different stains to reveal the migra-tion patterns [52]. In addimigra-tion to this, specific proteins can be identified and quantified with immunoblots, e.g. western blot, as the bands of the gel are transferred to a membrane that is stained with a tailor-made antibody against the protein of interest [53].
3.2.5 Dialysis
Dialysis is a technique offering removal of unwanted compounds, e.g. detergents, salts and small molecules from a protein sample [54]. The dialysis membrane is semi-permeable allowing only compounds up to a spe-cific size to pass through making it suitable for buffer exchange and purification, but also refolding of pro-teins.
3.2.6 Fluorescence Spectroscopy
Fluorescence is a phenomenon related to the absorp-tion of a photon leading to excitaabsorp-tion of an electron to a higher energy state followed by an energy transfer back to the ground state [55]. The radiated emission, as the electron falls back into its initial state, can be measured and the structural characteristics can be es-timated from the potential of the molecule to absorb light. The fluorescent group in the molecule is referred to as a fluorophore and is usually either comprised of fairly rigid aromatic rings or ring systems. Tryptophan is a common fluorophore in proteins that absorbs light at 280 nm and is commonly used to determine the struc-ture and the unfolding mechanisms of proteins. LECT2 has an embedded N-terminal tryptophan accompanied by one phenylalanine and seven tyrosines throughout the structure. However, since tryptophan is superior to phenylalanine and tyrosine in absorbing light it will be the residue that contributes to the majority of the signal. By tracking changes in intensity and spectral shifts under unfolding experiments LECT2’s general stability and possible unfolding intermediates can be estimated.
3.2.7 Circular Dichronism Spectroscopy
The principle of Circular dichroism (CD) is based on the fact that asymmetrical structures absorb right (εR) and
Fig. 4: Characteristic Circular dichroism (CD) spectra of the typical secondary structures of proteins, i.e. α-helix, β-sheet, and random coil. The α helical structure shows a positive peak at 193 nm followed by two negative peaks at 208 nm and 222 nm, respectively. Meanwhile, β-sheet formation shows a positive peak around 195 nm and a negative peak at 216 nm. In contrast to α-helices and β-sheets, random coil shows a negative peak around 195 nm which then heads towards zero. The picture is taken from the book Essentials of Medical Biochemistry written by N.V. Bhagavan (2011) [56]
left (εL) circularly polarized light differently [56], and is commonly used to determine the structural features of proteins. The difference in absorbance between right and left polarized light (εL- εR), i.e. the ellipticity (Θ), measured in degcm2dmol−1 plotted against the
wave-length generates a CD spectrum. Scanning the far UV region (250-190 nm), where the peptide chain absorbs light, reveals information about the secondary structure (figure 4), meanwhile, the near UV (>250 nm), where aromatic chromophores have their bands, can reveal information about the tertiary structure [57]. The spec-trum shows an average of contained structures if the protein in question comprises more than one. Combin-ing CD measurements with thermal and/or chemical denaturation can be used in order to reveal information about the structure, and the potential folding interme-diates.
4. METHODS CONTENTS
3.2.8 Atomic Force Microscopy
Atomic Force Microscopy (AFM) offers high-resolution, three-dimensional imaging of structures at the single-particle level down to the nanometer scale giving im-portant insights to biological sciences. The technique is based on a fine metal tip scanning over a surface, which deflects the cantilever that the tip is mounted on, thus, the laser beam that is focused on the cantilever. The laser is directed towards a detector that reconstitutes an image of the topography of the sample from the deflections.
AFM gives information beyond two-dimensional imag-ing techniques, except for topography e.g. mechanical properties, and can be performed at conditions close to physiological environments. Additionally, it can be operated both in buffer solutions and in air. One of the major advantageous with AFM is that any sample can be analyzed as it is not based on conductivity and is widely used for imaging of proteins, cells, viruses, and membranes.
4.
Methods
Three constructs of LECT2 were initially produced and evaluated, first, in terms of solubility in the E.coli strain BL21(DE3) and Rosetta-gami, respectively. Second, the con-structs were expressed in large cultures of 6 L for 4-6 h and were then purified, and refolded if necessary, on an immobi-lized metal affinity chromatography (IMAC) column. This followed by either ion-exchange chromatography (IEX) or maltose-affinity purification where the constructs were eval-uated in terms of yield and production time. All cultures were induced at OD600 0.5 with 1.0 mM
isopropyl-β-D-thiogalactoside (IPTG). The method section is explained from the construct resulting in the highest yield, i.e. His6
-TEV-LECT2. Construct design, evaluation on the constructs, and purification protocols for low yielding constructs are ex-plained in detail in Appendix B.
4.1.
Purification and Refolding
I40V LECT2 was expressed with a His6-tag and a TEV
cleavage site from a construct designed in a commer-cial vector (pLJSRSF7, Addgene) containing the coding sequence for the residues MGSS- prior to the His-tag. Additionally, the vector contained the T7 and lacI
ex-pression systems, and kanamycin resistance. Further on, LECT2 was isolated by a three-step purification pro-tocol from inclusion bodies (IBs) including on-column refolding, IMAC, and IEX purification. The sample preparation was initiated with resolving the cell pel-let in 60 mL IMAC A.1 buffer (20 mM NAHPO4, and
200 mM NaCl, pH 8.0) followed by stirring on ice with one pill protease inhibitor cocktail (A32965, Thermo Scientific) for 30 min at 4° C. The sample was then son-icated 4-5 times at the amplitude 80% of 3 sec pulses for 2 min, and centrifuged at 4,000g for 30 min at 4°C. The insoluble fraction was resuspended in 30 mL 6 M guanidinium hydrochloride (GuHCl, pH 8.0) with an-other pill protease inhibitor cocktail and was stirred for 30 min, followed by sonication and centrifugation as previously.
The supernatant was filtered (0.45 µm, MilliPore) and was then loaded onto a 5 mL injection loop attached to the IMAC column (17524801, GE), containing a Ni Sepharose resin, with a flow rate of 2.0 mL/min under denaturing conditions from govern IMAC buffer A.2 (20 mM NaHPO4, 200 mM NaCl, 6 M GuHCl, 4:1 mM
ratio of GSH:GSSG, and 400 mM L-arginine, pH 8.0). The proteins bound onto the column were then refolded on the FPLC ÄKTA pure 25 purification system as the buffer was exchanged with a gradient to IMAC buffer A.3 (20 mM NaHPO4, 200 mM NaCl, and 1:4 mM ratio
of GSH:GSSG, pH 8.0) over 13 h with 0.5 mL/min flowrate.
Before elution, the column was removed and the sys-tem was washed with buffer A.1 (20 mM NaHPO4, and
200 mM NaCl, pH 8.0) at a 10 mL/min flowrate. Be-fore remounting the column the flowrate was lowered back to 2.0 mL/min. The elution occurred in two steps: first, a star elution with 5% IMAC buffer B.1 (800 mM Imidazole added to buffer A.1) was made to remove unspecifically bound proteins followed by a gradient to 100% over 20 min. Eluted fractions were analyzed with SDS-PAGE where 1.5 µg protein sample was loaded alongside a Precision Plus Protein™ Dual Color Stan-dard marker (1610374, BioRad). The bands were stained with either coomassie or silver stain and were imaged with a ChemiDoc (Biorad). All following SDS-PAGE protocols were executed throughout the same protocol unless otherwise stated.
CONTENTS 4. METHODS
Fig. 5: The workflow of the precipitate-denature-fragmentation concept that is followed with mapping of the disulfide bonds in LECT2 with LC-MS. One sample was reduced with 10 mM TCEP and alkylated with 25 mM iodoacetamide (black probe) and was run as a control samlpe.
4.2.
Purification of Refolded LECT2
The eluted fractions were dialyzed against 2 L IEX buffer A (10 mM NaHPO4) overnight using a 7 kDa membrane. LECT2 was then purified a second time with ion-exchange chromatography in order to remove truncated and aggregated species. The protein sample was added to an S-sepharose anion column (17-1153-01, GE) the ÄKTA pure 25 purification system with a flow rate of 2 mL/min of buffer A and eluted with buffer B (10 mM NAHPO4, and 800 mM NaCl, pH 7.6)
un-der a ten-minute gradient. Eluted fraction samples of 850 ng per well were analyzed with SDS-PAGE and the protein mass was further characterized by MS, i.e. MALDI and ESI. Additionally, a control sample was reduced with 10 mM TCEP and the redox state of the protein was analyzed with SDS-PAGE. The initial pro-tein concentration was determined with UV using a Nanodrop 2000/c Spectrophotometer (Thermo Fisher) and was then verified with a BCA Protein assay kit (23225, Thermo Fisher). All protein samples were stored at -80°C, and were thawed on ice for approximately 30 min before analysis.
4.3.
Mapping of Disulfide bonds
The characterization of the disulfide bond formation re-veals important information about the fold of a protein.
Establishing a method to first verify that the produced protein is natively folded is meaningful when monitor-ing the natural behavior of LECT2. Secondly, durmonitor-ing the analysis of aggregation assembly eventual disulfide bond shuffling give a great insight into the cause of the aggregation mechanism and fibril formation.
This protocol (figure 5) was based on a precipitate-denature-fragmentation concept followed with disulfide bond mapping with LC-MS of Lu et al. [58]. This in combination with a method of Agarwal et al. which showed that performing the protein digestion under slightly acidic conditions was crucial to avoid an artifact of disulfide scrambling [59].
Initially, two samples of 40 µg LECT2 in 250 µL (9 µM) was precipitated with 100 µL CHCl3and 600 µL MeOH,
followed by vortexing the samples and centrifugation at 15 000 RPM for 2 min. Then the protein pellets were washed three times by resolving the pellets by roughly mixing it with 600 µL MeOH followed by vortexing and centrifugation as previously. After the last wash, the supernatants were removed and the pellet was let to air dry for an hour. Once the protein precipitate was resolved in a buffer DM (100 mM tris, and 8 M urea buffer, pH 6.5) one sample was reduced with 25 mM TCEP and alkylated with 25 mM iodoacetamide
4. METHODS CONTENTS
under incubation at 37°C for 30 min for each treatment. Then the buffer of both the protein sample and the reduced and alkylated control sample were diluted to a concentration of 2 M urea by adding 60 µl ddH2O.
This in order not to affect the proteolytic activity of the 20 µg trypsin that was added to the LECT2 samples overnight.
The fragmentation was analyzed by SDS-PAGE on a silver-stained 4-20% tricine gel (EC6625B, Thermo Fisher Scientific) by running 3.5 µg protein of each sam-ple alongside a low molecular ladder (26628, Thermo Fisher Scientific) at 180 V for approximately 20 min. Once verified that the fragmentation pattern contained the majority of the expected molecular weights from theoretically calculated fragment masses (table 4 in Ap-pendix C) they were accurately searched for with LC-MS.
The digested fragments were directly analyzed on a Q Exactive mass spectrometer (Thermo Fisher Scientific). The digests were loaded onto a three-phase 100µm mi-crocapillary column containing Aqua C18 resin. The loaded column was then attached to an ID analytical column with a pulled tip packed with Aqua C18 resin. Samples were separated at a flow rate of 200 nL/min on an Easy nLCII (Thermo Fisher Scientific). Peptides were eluted from the tip of the column and nanosprayed di-rectly into the mass spectrometer by application of 2.5 kV voltage at the back of the column.
The Q Exactive was operated in a data-dependent mode. Full MS1 scans were collected in the Orbitrap at 70K resolution with a mass range of 400 to 1800 m/z and an AGC target of 1e6. A top 10 method was utilized with HCD fragmentation at 25 NCE, resolution of 17.5 k, AGC target of 1e5 and an underfill ratio of 0.1%. Maximum fill times were set to 60 ms and 120 ms for MS and MS/MS scans respectively. Quadrupole isolation at 2 m/z was used, singly charged and unassigned charge states were excluded, and dynamic exclusion was used with an exclusion duration of 15 sec.
4.4.
Thermodynamics
Initially, a CD spectrometric screen of the native protein sample of OD2800.1 (6 µM, extinction coefficient 16 290
M−1cm−1) was made in the far UV range (250-185 nm)
at 25°C with a Jasco J-1500 CD Spectrometer in a 1 mm cuvette to characterize the protein. This followed by analyzing the unfolding under a thermal denaturing gradient by increasing the temperature in 5°C intervals followed by 10 min incubation from 25°C up to 80°C. All CD spectra were collected under a 20-sec time frame.
Fig. 6: Visualization of the embedded N terminal trypto-phan residue of LECT2 (PDB: 5b0h), which is well suited for monitoring the red shift of the cheminal denaturation of the protein with increasing urea concentration. Structure published by Zheng et al. [30].
4.5.
Chemical Denaturation and Unfolding
Kinetics
LECT2 comprises a tryptophan residue at position three that is well embedded within the structure, making it well suited for monitoring unfolding by tryptophan fluorescence (figure 6). A protein sample of OD2800.1
(6 µM, extinction coefficient 16 290 M−1cm−1) was de-natured under a 0-8 M urea gradient under both native and reducing conditions, generated with 10 mM TCEP. The denaturation was measured by an FP-8500 Jasco spectrofluorometer (excitation 280 nm; emission 310-420 nm, bandwidth ex. 5 nm; bandwidth em. 10 nm). In addition to this, a primary screen of the unfolding kinetics was made under a chemical denaturation with 6 M urea of the native protein sample of concentration
CONTENTS 5. RESULTS AND DISCUSSION
OD280 1 (60 µM, extinction coefficient 16 290 M−1cm −1) under an interval of 140 sec.
4.6.
LECT2 Aggregation
LECT2 aggregates were taken from reduced protein species that had aggregated under the dialysis between the two purification steps. Both an unreduced sample and a sample reduced with 10 mM TCEP were analyzed with SDS-PAGE and MALDI. Finally, the LECT2 aggre-gates were analyzed with an AFM screen using a Multi-mode scanning probe microscope with a Nanoscope Illa controller. A sample of 20 µL aggregates was dispensed onto a freshly cleaved mica, washed three times with ddH2O, and then air-dried for 4 h. The probe was then
scanned over a 1 µm surface at three representative areas of the sample at a rate of 1.489 Hz.
5.
Results and Discussion
5.1.
LECT2 Expression, Refolding, and
Pu-rification
LECT2 was found to have the highest expression as inclusion bodies in BL21 after 4 h induction at 37°C, when analyzed with a solubility assay in Rosetta-Gami and BL21, respectively. The evaluation of induction temperature is excluded, however, the incubation tem-perature 37°C showed slightly higher expression than 25°C and was therefore used for all protein expression procedures. Further on, the protocol was applied on 6 L cultures that were harvested by centrifugation at 4°C. The obtained pellet was resolved in 60 mL IMAC buffer A.1 and was stored at -80°C.
Before purification and refolding, the sample was pro-cessed through a multi-step preparation protocol. After thawing the cell sample it was stirred on ice with one pill of protease inhibitor cocktail mix for half an hour in a 4°C cold room. Then the sample was sonicated five times for two min each time, which was sufficient to increase the overall protein concentration around eight to ten times.
After the insoluble fraction was isolated after a second round of centrifugation it was dissolved in 6 M GuHCl followed with sonication and centrifugation as previ-ously. After filtering the sample it was loaded onto
the IMAC column and was refolded overnight as the conditions were changed from a reducing to an oxidiz-ing environment, and the denaturatoxidiz-ing conditions were lowered and gradually removed (figure 7A). The buffer change was monitored by a chromatogram to ensure that no protein was eluted during the refolding phase. The refolding protocol was based on the method of Zhang et al. [16], with some adjustments. The GSH:GSSG ratio was modified from 1:1/1:1 to 1:4/4:1 during the buffer-change between the loading/elution buffer. Changing the conditions from a reducing to an oxidative environment was to ease the refolding pro-cess, however, it might have been even more efficient to make a 1:0/0:1 buffer change to have the protein in an entirely oxidative environment before changing to the native buffer before elution. In addition to this, we could decrease the amount of GSH and GSSG since both reagents are relatively expensive. Moreover, adding a stronger reducing agent, e.g. DTT, to the IMAC A.2 buffer would possibly make the refolding more efficient, i.e. increase the yield of oxidized protein.
Except for removing the detergent and introducing an excess of GSSG, the arginine was gradually removed during the buffer-change. Although arginine is added to stabilize the folding, there is a risk that it could block the binding of LECT2 to the IMAC column, but this needs to be further investigated. The elution of the pro-teins resulted in a double-peak, which fractions were kept separated and were analyzed with SDS-PAGE and western blot (figure 7B). Both peaks consisted mainly of a broad band corresponding to the weight of LECT2. However, two differences were noticed, first, the peak (P1) showed two defined lower molecular bands proba-bly corresponding to C terminal truncations. Secondly, the second peak (P2) contained aggregates, which are reflected in the increase in UV as P2 was eluted. Addi-tionally, all molecular bands were stained with a mono-clonal anti-His antibody confirming that all the bands shown in the SDS-PAGE also contained a His-tag. Further on, after refolding and purification there was an extensive loss of material especially due to aggre-gation under the dialysis between the IMAC and IEX chromatography. The aggregation is most likely a re-sult of partially reduced protein species as a rere-sult of unsuccessful refolding. Except for modifying the buffer,
5. RESULTS AND DISCUSSION CONTENTS
Fig. 7: Representative chromatogram for the refolding of LECT2 (A) bound onto an IMAC column showing no protein elution (blue) and a gradual change in conductance corresponding to the buffer change (brown). The elution from the IMAC column resulted in two peaks (B), and the corresponding SDS-PAGE and western blot revealed a high abundant band corresponding to the mass of LECT2 in both peaks. What distinguishes the character of the two peaks is that the first peak (P1) showed two defined lower molecular bands most likely corresponding to C-terminal truncations and that the second peak (P2) contains aggregates. After changing to a buffer suitable for IEX, LECT2 was highly efficiently isolated on an S-Sepharose column (C). The chromatogram showed is the one of P2, which also is representative for P1. Eluted fractions from P1 and P2 were almost identical.
changing to a eukaryotic expression system may be cru-cial to increase the yield of oxidized protein since the ER seems to be important for correct folding of LECT2.
Furthermore, as LECT2 is one of relatively few basic proteins, i.e. carries a positive pI, a highly efficient pu-rification of the band corresponding to LECT2 from the lower molecular bands could be made on a negatively charged S-Sepharose IEX column (figure 7C). The two IMAC samples were purified separately and generated two comparable chromatograms with the only differ-ence that the first IEX peak of sample P1 (P1.1) was slightly increased in comparison to the first peak of the P2 sample (P2.1). The second peak of both P1 and P2 (P1.2 and P2.2) were almost identical. Additionally, no
sign of aggregation was seen in either of the IEX peaks.
As the S-Sepharose offers such convenient and accurate purification of LECT2, using a refolding and purifica-tion method only based on IEX purificapurifica-tion could be a strategy to overcome the loss of material after IMAC. Then the LECT2 protein could first of all be expressed without any tag as inclusion bodies, this followed by being unfolded with GuHCl, GSH and a low concentra-tion of DTT. The refolding procedure could be carried out through systematically adding the protein sample to a beaker of 2 L native buffer containing GSSG and arginine. After incubating at 4°C overnight, the protein sample would then be concentrated and isolated with IEX, eventually followed by gel filtration.
CONTENTS 5. RESULTS AND DISCUSSION
5.2.
Characterization of Refolded LECT2
In addition to characterizing the molecular weight, the redox state of the protein was investigated initially with MS and SDS-PAGE, and later on fluorimetry, explained in Chemi-cal Denaturation and Unfolding Kinetics. Moreover, a CD screen was made in the far UV region and the protein sample was compared to those of parallel studies, see Thermodynam-ics.An accurate characterization of proteins prior to anal-ysis is necessary to ensure that a pure sample of the natively folded protein of interest is obtained. In this study, LECT2 was characterized by overlapping meth-ods. Initially, the molecular weight was determined with MALDI and ESI (figure 8A-B). The molecular weight characterized with MALDI matched the 16 415.6 Da of the calculated mass of SS-His6-TEV-LECT2 if
all the cysteines are oxidized, instead of 16 421.6 Da if they all are reduced. This applied to both sample P1.2 and P2.2 which had a major peak at 16 414 Da and 16 419 Da, respectively, indicating that the major part of the protein is in its oxidized form. The two initial serine residues are present due to the N-terminal MGSS-extension coded prior to the LECT2 sequence in the construct.
Interestingly, when analyzed with ESI the mass was 54-58 Da higher compared to the mass characterized by MALDI. The additional weight could correspond to a bound nickel atom which has the mass of 58.6 Da, which matches the molecular weight of Ni2+-bound and oxidized LECT2 accurately with only 0.3 Da off target. This suggests that LECT2 has metal-binding characteristics of bivalent cations in general rather than only binding Zn2+ions. However, the molecular weight of Zn2+ is 65.7 Da and falls within the 10% off the targeted mass, i.e. it could be either or both a Ni2+or a Zn2+ion bound to the protein. Why the bound metal ion only is shown when analyzed with ESI is probably since it offers a more gentle ionization method than MALDI.
Meanwhile, the absence of a zinc atom bound has shown to initiate aggregation, adding zinc to LECT2 ag-gregates has been shown to decrease the aggregation in previous studies [33]. Whether it has to be a zinc atom bound to obtain the native fold of LECT2 or if any
bi-valent cation can stabilize sufficiently is not yet known. As the IMAC buffer was based on phosphate, no zinc was added as the two are known to aggregate when contained in the same buffer. A zinc-based tris-buffer may be more suitable during sample preparation, re-folding and purification, but has not been investigated at this stage of the project.
A redox state analysis with SDS-PAGE was carried out with a comparison between the native protein sam-ple and the same samsam-ple reduced with 10 mM TCEP (figure 9A). The reduction caused a shift of the migra-tion pattern of LECT2 suggesting that the major part of the protein is oxidized. Furthermore, the sample was treated with 25 mM iodoacetamide to alkylate free cysteine thiols and then analyzed with MALDI-TOF (figure 8C). The alkylated mass spectrum revealed two smaller peaks corresponding to up to two modifications. The peak suggesting two modifications of the produced I40V LECT2, which indicates that one disulfide bond is harder to obtain than the other two. However, iodoac-etamide has a tendency to modify histidines and pri-mary amines, i.e. N terminus and lysine residues [60]. This could be a reason for the alkylation shown in the spectrum, especially in the case of one modification. Whether the modifications truly are due to unformed disulfide bonds or an effect of unspecific binding needs to be further inspected.
Mapping of the disulfide bonds is an important inves-tigation in order to prove that the produced LECT2 has a native fold. This procedure was initiated with this project and a protocol was developed includ-ing a precipitate-denature-fragmentation concept fol-lowed by LC-MS. The fragments were first analyzed with SDS-PAGE where the molecular weights of the trypsin-cleaved fragments were matched towards possi-ble disulfide-bound fragment masses that were theoreti-cally calculated (table 4 in Appendix C). The majority of the molecular weights of the fragments were between 4.6 kDa and 10 kDa, but there were also indications of some lower molecular weight bands that could have been giving more information if the bands would have been developed longer (figure 9B).
The Cys99-Cys142 disulfide bond was characterized from the peak eluted at retention time 67.38-68.26 sec in four peaks, z= +2, +3, +4, +5, +6, in the MS spectrum
5. RESULTS AND DISCUSSION CONTENTS
Fig. 8: Mass spectrum of the protein sample P1.2 (A) and P2.2 (B) generated with both MALDI and ESI, respectively. The MALDI spectrum shows a major peak at 16 414 Da and 16 419 Da, respectively, which corresponds to the theoretical mass of SS-His6-TEV-LECT2 (oxidized cysteines 16 415.6 Da/reduced cysteines 16 421.6 Da, ExPASy Compute pI/Mw
tool). The ESI spectra of LECT2 show an increased mass of approximately 58 Da, matching a Ni2+ ion. When alkylating the native protein sample with 25 mM iodoacetamide, two smaller peaks appear in the MALDI spectrum corresponding to up to two modifications compared to the native sample either due to free thiols or unspecific binding (C).
CONTENTS 5. RESULTS AND DISCUSSION
Fig. 9: Reduction of the native LECT2 protein sample with 10 mM TCEP causes a shift in migration pattern on a Coomassie-stained SDS-PAGE gel (A). To verify the fold of the protein a protocol for mapping the disulfide bonds were developed and the trypsinized fragments were first analyzed with SDS-PAGE (B) and second with LC-MS (C) where the molecular weights were matched towards possible disulfide-bound fragment masses that were theoretically calculated (table 4). The third (Cys99-Cys142) disulfide bond was characterized from the peak eluted at retention time 67.38-68.26 sec in four peaks, z= +2, +3, +4, +5, +6, in the MS spectrum corresponding to the mass of 3096.43 Da.
5. RESULTS AND DISCUSSION CONTENTS
corresponding to the mass of 3096.43 Da 9C). In order to map the remaining two disulfide bonds, the analysis needs to be repeated. The fragmentation could possibly be made more sufficient by modifying the concentra-tion of denaturant, pH, and by using multiple different proteases, e.g. LysC that has a high specificity towards lysine residues resulting in larger peptides followed by a more accurate cleavage with trypsin.
5.3.
Thermodynamics
The CD spectrometric analysis was initiated with a screen in the UV range (figure 10A), indeed the obtained CD spectra were highly comparable to those previously published [14, 15]. Additionally, the study of Zheng et. al showed that adding Zn2+to the protein sample did not affect the CD spectrum. Thus, the absence of zinc in the buffer does not seem to have a drastic effect on the structure of LECT2.
The positive peak around 230 nm is almost identical to the spectra published before. However, the negative peak observed at 202 nm is slightly higher than seen previously, which could be an effect of a higher extent of β-sheet formation. Also, between the negative and the positive peak the appearance of the spectra was recorded somewhat different as it does not include as many overlapping peaks, but rather shows a convex be-havior. Beyond differences in the content of β-structure, the overall fold of the protein and aggregation artifacts could affect the character of the CD spectrum.
The far UV screen does not resemble a traditional struc-ture, and considering the structure revealed by X-ray crystallography [30], the spectra would be expected to have the nature of proteins with extensive β-structure. The collected data show some similarities to random coil, which was later on verified with fluorimetry not to be the case see Chemical Denaturation and Kinetics. Fur-ther on, the lack of helical signal was expected due to the absence of helical content in the crystal structure of LECT2. Thus, the majority of the signal is most likely due to aromatic residues. Overall, LECT2 shows dis-ordered and highly dynamic characteristics. It may be that the structure resembles a molten globule-like chain bound with disulfides, which could explain the drastic change of the protein upon reduction.
When being subject to elevated temperature, LECT2 was shown to be highly unstable and started to unfold already at physiological temperature (figure 10B). More-over, an isodichronic point is observed in temperature-dependent CD spectra around 197 nm suggesting a two-phase unfolding process, i.e. the absence of an unfolding intermediate, and a redshift of the CD min-imum is observed as LECT2 unfolds. As the protein sample is further heated to 80°C the spectra take the form of a random coil’s as a result of complete denatura-tion of the protein and the isosbestic point is no longer observed (figure 10C). When analyzing the spectra at two wavelengths showing a major change in ellipticity over the temperature interval, two entierly different spectra are observed. At 230 nm a sigmoidal curve with a midpoint at approximately 50°C is seen which is a traditional character of a spectrum monitoring protein unfolding. In contrast, at 205 nm aggregation is noticed as the denaturation curve starts to elevate around 65°C. Thus, unfolding and aggregation of LECT2 can be ana-lyzed simultaneously. The same experiment was to be repeated under reducing conditions, i.e. 10 mM TCEp or 1 mM DTT, which interfered with the signal and no conclusions could be drawn from the generated spectra.
5.4.
Chemical Denaturation and Unfolding
Kinetics
The tertiary structure denaturation of LECT2 was ana-lyzed with a fluorescence spectrometer (excitation 280 nm) within the UV range 310-420 nm. Nine different samples comprising a gradient between 0-8 M urea were prepared and incubated for ten minutes. The unfolding could be monitored throughout the urea gradient with a redshift as the concentration was increased (figure 11A), and the redshift is clearly visualized when comparing the 0 M and 8 M urea sample alone (figure 11B). The intensity of the signal decreased as urea was added, but did not show a clear trend as it increased and decreased changeably throughout the increased urea gradient. In addition to this, the 1 M urea sample showed a slight blueshift and decreased drastically in intensity, and is most probably an outlier, thus, was excluded in further analyses.
The LECT2 unfolding was further investigated in terms of the ratio between the two wavelengths that showed
CONTENTS 5. RESULTS AND DISCUSSION
Fig. 10: CD spectrometric measurements on LECT2 in the far UV region reveals a highly dynamic structure. The spectrum does not resemble a traditional appearance of proteins with a high β-sheet content (A), and the structure is highly vulnerable to increase in temperature (B). Moreover, an isosbestic point around 197 nm suggests a two-state unfolding process. As the sample is heated to 80°C LECT2 is completely denatured and the isosbestic point is no longer observed (C). When analyzing the thermal denaturation at the wavelengths 230 nm and 205 nm, the process of unfolding and aggregation can be monitored simultaneously (D). All spectra represent raw, unsmoothed data.
6. CONCLUSION AND FUTURE PROSPECT CONTENTS
the greatest intensity change, i.e. 325 nm and 375 nm, to make the spectrum independent of protein concen-tration changes over the urea gradient (figure 11C). A cooperative unfolding was observed which verifies that LECT2 is not a random coil. In addition to this, LECT2 showed a midpoint of denaturation at 3 M urea. Moreover, a sample set with 10 mM TCEP added fol-lowed with 10 min incubation was analyzed separately to determine the effect of reduction on stability. The reduction of the disulfide bonds had dramatic effects on LECT2, which was almost completely unfolded even without any urea added. Hence, the produced LECT2 in this study is most likely oxidized and correctly folded. Furthermore, a kinetic analysis was made of the un-folding of LECT2 in 6 M urea at the anticorrelated wavelengths 325 nm and 375 nm over a time frame of 140 sec (figure 11D). What can be concluded from this preliminary investigation of the unfolding kinetics is that LECT2 unfolds completely already within 20 sec. The bleaching of the sample can be observed as the baseline drift of the curve but does not occur until af-ter approximately 70 sec, thereby, does not affect the visualization of the unfolding.
The unfolding of LECT2 can be described as a two-phase process, which could be explained by the pres-ence of two native states (figure 12). The two-phase process is proposed as the unfolding kinetics data resembles a double-exponential model of unfolding rather than a single-exponential model. The double-exponential model uses five different parameters m1-m6 where m2 and m4 are the amplitude of the com-ponents of the biexponential fit, the sum of which is equal to the total amplitude of the unfolding exponen-tials. Thus, m2explains the fraction of fast unfolding
species, m2
m2+m4, meanwhile, m4 explains the slow un-folding species. m1 is an offset indicating where the intensity starts, which ideally is equal to the value of the folded protein in the presence of the denaturant. Finally, m3 and m5 are the rate constants of the model, and m6 is a linear term accounting for bleaching. Two unfolding intermediates could be explained by whether the LECT2 has a metal-bound or not. As LECT2 previously has been shown to be stabilized by metal-binding [33], unfolding the metal bound LECT2 would demand higher energy, i.e. be slower, compared
to the metal-free LECT2. When comparing the ratio between the possible bounded and the metal-free it is divided 40/60 12B), suggesting that the slight majority of the protein does not have a metal-bound. However, what the reason is behind the presence of two intermediates needs to be further investigated.
5.5.
LECT2 Aggregation
LECT2 aggregation was briefly investigated from the aggregated fraction from the dialysis between the two protein purifications. The sample was initially analyzed with SDS-PAGE with and without a reducing agent added (figure 13A). The gel revealed the presence of both monomers, dimers, trimers, and higher oligomers, which intensity decreased by approximately half for every higher state of assembly. As 10 mM TCEP was added the trimer disappeared and the sign of the dimer was weak, meanwhile, the band corresponding to the higher oligomer was about the same. The same behav-ior of aggregation has been seen with LECT2 in the absence of a bivalent metal ion, where the aggregates were removed in a similar matter upon the introduction of zinc [33].
The aggregates species were further characterized when analyzed with MALDI-TOF suggesting simple monomeric polymerization. This due to a trend of peaks corresponding to the addition of one monomer up to a mass corresponding to the one of a LECT2 hexamer. Furthermore, for every higher order of as-sembled species, the intensity is halved probably as a cause of the time required for assembling monomer by monomer. The reduced sample also contained the same aggregation trend but was notably decreased.
Finally, an AFM screen of the LECT2 aggregates was made and generated images were different from the control sample containing solely 10 mM NaHPO4
(fig-ure 14). The pattern revealed reminds of protofibril-like oligomers probably corresponding to those identified by SDS-PAGE and MALDI-TOF. In order to draw further conclusions a more thorough investigation is required.
6.
Conclusion and Future Prospect
I40V LECT2 was produced as His6-TEV-LECT2 in E.
CONTENTS 6. CONCLUSION AND FUTURE PROSPECT
Fig. 11: Chemical denaturation causes a redshift in tryptophan fluorescence as LECT2 unfolds (A), especially seen when comparing the 0 M and the 8 M urea sample, respectively. LECT2 shows a cooperative unfolding with a midpoint at 3 M urea (B). The intensity difference of the wavelength ratio F325/F375 of reduced and unreduced LECT2 change drastically as the reduction of the disulfide bonds denatured LECT2 alone (C). A screen of the unfolding kinetics under 6 M urea denaturation revealed that LECT2 unfolds within 20 sec (D).
6. CONCLUSION AND FUTURE PROSPECT CONTENTS
Fig. 12: The unfolding kinetics of LECT2 can be explained by a two exponential model, thus correlates with a two-phase process when monitored at 325 nm (A). The model is explained by six different variables, where m2 and m4 explain the biexponential fit. m1 is an offset and m6 is a linear term accounting for bleaching, meanwhile, m3 and m5 are the rate constants of the model. The sum of m2 and m4 corresponds to the total amplitude of the unfolding exponentials. When calculating the fraction of the fast (m2) and the slow (m4) unfolding species they are divided 60/40 (B). The unfolding rate is a reflection of the energetic barrier that needs to be overcome in order for unfolding to occur, which can be visualized in a free energy diagram.
CONTENTS 6. CONCLUSION AND FUTURE PROSPECT
Fig. 13: Reduced and unreduced aggregate sample analyzed with SDS-PAGE (A) and MALDI-TOF (B), respectively. The aggregation trend suggest a simple monomer polymerization (left) but needs to be further analyzed. The concentration of higher assebled LECT2 species decrease upon reduction (right) indicating an involvement of the disulfide bonds.
Fig. 14: A primary AFM screen of three representative areas of an aggregated LECT2 sample indicates protofibril-like structures probably corresponding to the oligomer species identified with MALDI-TOF (figure 13)
