Spring term 2020 | LITH-IFM-A-EX−20/3776−SE
Optimizing signal peptides for
expression of recombinant antibodies
in HEK293 cells
Gustav Myhrinder
Science for Life Laboratory
Examiner and internal supervisor: Lars-G¨oran M˚artensson External supervisors: Leif Dahllund & Anders Olsson
2020-06-11 Department of Physics, Chemistry and Biology
Linköping University
URL för elektronisk version
ISBN
ISRN: LITH-IFM-A-EX--20/3776--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
Optimizing signal peptides for expression of recombinant antibodies in HEK293 cells
Författare Author
Gustav Myhrinder
Nyckelord Keyword
Signal peptide, antibody, human immunoglobulin G, mammalian cell, HEK293, protein expression, secretory pathway, transient transfection, Biacore, Biacore concentration analysis, SDS-PAGE, kinetic screening
Sammanfattning Abstract
Monoclonal antibodies are well-established as a therapeutic in the biopharmaceutical market, targeting a variety of diseases and with 79 approved products by the United States Food and Drug Administration in December 2019. Therapeutic monoclonal antibodies are commonly produced as recombinant proteins in mammalian cell lines, due to their capacity of post-translational modifications, most notably glycosylation. Furthermore, an identified bottleneck within the production of recombinant proteins is the translocation of nascent proteins from the cytosol into the lumen of the endoplasmic reticulum. The signal peptide, which is located at the N-terminal of nascent proteins, plays a central role in the process of protein secretion. Several studies have shown that optimization of signal peptides is a crucial step for attempting to achieve increased expression of recombinant antibodies in mammalian systems. The aim of this study was to evaluate the expression of three human recombinant antibodies in Human Embryonic Kidney 293 (HEK293) cells by evaluating 16 different signal peptide combinations, consisting of eight heavy chain (HC) and two light chain (LC) signal peptides. The impact goal was an efficient secretion of recombinant antibodies, and thus lower production cost of recombinant antibodies in HEK293 cells. First, 16 HC and LC signal peptide plasmid constructs were generated for each of the three recombinant antibodies. Thereafter, transient gene expression in HEK293 cells were performed at three independent experiments. Finally, the antibody titers were quantified using Biacore concentration analysis.
The produced antibody titers for the three studied recombinant antibodies were highly dependent on the used signal peptides. Interestingly, the evaluated HC and LC signal peptide combinations resulted in 3 times higher and 2 times higher antibody titers compared to the original signal peptides used by the Drug Discovery and Development platform at Science for Life Laboratory, for two of the studied antibodies respectively. The results presented in this report further demonstrates the necessity to evaluate signal peptides in order to achieve increased expression of recombinant antibodies in mammalian systems.
Monoclonal antibodies are well-established as a therapeutic in the biopharmaceutical market, targeting a variety of diseases and with 79 approved products by the United States Food and Drug Administration in December 2019. Therapeutic monoclonal antibodies are commonly produced as recombinant proteins in mammalian cell lines, due to their capacity of post-translational modifications, most notably glycosylation. Furthermore, an identified bottleneck within the production of recombinant proteins is the translocation of nascent proteins from the cytosol into the lumen of the endoplasmic reticulum. The signal peptide, which is located at the N-terminal of nascent proteins, plays a central role in the process of protein secretion. Several studies have shown that optimization of signal peptides is a crucial step for attempting to achieve increased expression of recombinant antibodies in mammalian systems.
The aim of this study was to evaluate the expression of three human recombinant antibodies in Human Embryonic Kidney 293 (HEK293) cells by evaluating 16 different signal peptide combinations, consisting of eight heavy chain (HC) and two light chain (LC) signal peptides. The impact goal was an efficient secretion of recombinant antibodies, and thus lower production cost of recombinant antibodies in HEK293 cells. First, 16 HC and LC signal peptide plasmid constructs were generated for each of the three recombinant antibodies. Thereafter, transient gene expression in HEK293 cells were performed at three independent experiments. Finally, the antibody titers were quantified using Biacore concentration analysis.
The produced antibody titers for the three studied recombinant antibodies were highly dependent on the used signal peptides. Interestingly, the evaluated HC and LC signal pep-tide combinations resulted in 3 times higher and 2 times higher antibody titers compared to the reference HC and LC signal peptides used by the Drug Discovery and Development platform at Science for Life Laboratory, for two of the studied antibodies respectively. The results presented in this report further demonstrates the necessity to evaluate signal peptides in order to achieve increased expression of recombinant antibodies in mammalian systems.
Ab Antibody
ANOVA Analysis of variance
CHO cells Chinese Hamster Ovary cells CH Constant heavy chain domain CL Constant light chain domain dH2O Sterile water
DNA Deoxyribonucleic acid ER Endoplasmic reticulum E. coli Escherichia coli
Fab region Fragment antigen-binding region Fc region Fragment crystallizable region
FDA United States Food and Drug Administration GFP Green fluorescent protein
HC Heavy chain
HEK293 cells Human Embryonic Kidney 293 cells Ig Immunoglobulin
IgG Immunoglobulin G
IgG1 Immunoglobulin G subclass 1 LC Light chain
mAb Monoclonal antibody mRNA Messenger RNA
PCR Polymerase chain reaction PTM Post-translational modification RNA Ribonucleic acid
RNC Ribosome-nascent chain complex rpm Revolutions per minute
SciLifeLab Science for Life Laboratory
SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis SEM Standard error of the mean
SP Signal peptide
SPHC Heavy chain signal peptides
SPLC Light chain signal peptides
SPase Signal peptidase
SRP Signal recognition particle
SR Signal recognition particle receptor SPR Surface plasmon resonance
Tm Melting temperature
TGE Transient gene expression VH Heavy chain variable domain VL Light chain variable domain
Abstract III
Acronyms and Abbreviations IV
1 Introduction 1 1.1 Background . . . 1 1.2 Aim . . . 1 1.3 Motivation . . . 1 1.4 Work plan . . . 2 1.5 Limitations . . . 2 2 Theory 5 2.1 Scientific background . . . 5
2.1.1 Antibody function and structure . . . 5
2.1.2 Secretory pathway . . . 6
2.1.3 Signal peptides . . . 7
2.1.4 Surface plasmon resonance . . . 8
2.1.5 Biacore biosensors . . . 9
2.2 Methodology . . . 11
2.2.1 In-Fusion cloning . . . 11
2.2.2 Cloning approach . . . 11
2.2.3 Colony PCR screening . . . 13
2.2.4 Restriction digest screening . . . 13
2.2.5 Transient transfection . . . 13
2.2.6 Transfection efficiency . . . 13
2.2.7 Biacore concentration analysis . . . 14
2.2.8 Biacore kinetic screening . . . 14
3 Methods 15 3.1 Preparation of variable domains . . . 15
3.1.1 PCR amplification of variable domains . . . 15
3.1.2 Purification of LC variable domains . . . 15
3.1.3 Purification of HC variable domains . . . 15
3.2 Preparation of signal peptides . . . 16
3.3 Linearization of plasmid backbone . . . 16
3.4 Construction of LC plasmid constructs . . . 17
3.4.1 In-Fusion cloning of LC variable domains . . . 17
3.4.2 Plasmid transformation into competent E. coli . . . 17
3.4.3 Colony PCR screening of LC plasmid constructs . . . 17
3.4.4 Minipreparations of LC plasmid constructs . . . 18
3.4.5 Restriction digest screening of LC plasmid constructs . . . 18
3.4.6 Sequencing of LC plasmid constructs . . . 18
3.5 Linearization of LC plasmid constructs . . . 19
3.6 Construction of HC and LC plasmid constructs . . . 19
3.6.1 In-Fusion cloning of HC variable domains . . . 19
3.6.6 Sequencing of HC and LC plasmid constructs . . . 21
3.7 Sterile filtration of plasmid constructs . . . 21
3.8 Maintenance of HEK293 cells . . . 21
3.9 Transient gene expression of recombinant antibodies . . . 21
3.9.1 Transfection of plasmid constructs into HEK293 cells . . . 21
3.9.2 Measurement of transfection efficiency . . . 22
3.9.3 Harvest of expressed recombinant antibodies . . . 22
3.10 Biacore concentration analysis of recombinant antibodies . . . 23
3.11 Statistics . . . 23
3.12 SDS-PAGE analysis of recombinant antibodies . . . 23
3.13 Kinetic screening of recombinant antibodies . . . 24
4 Results 26 4.1 Cloning of HC and LC plasmid constructs . . . 26
4.2 Evaluation of HC and LC signal peptides for recombinant antibody expres-sion in HEK293 cells . . . 26
4.3 Comparisons between the evaluated HC and LC signal peptides and the reference signal peptides . . . 30
4.4 Control of full-length recombinant antibodies using SDS-PAGE . . . 31
4.5 Kinetic activity of recombinant antibodies . . . 33
4.6 Control of HEK293 cell maintenance culture . . . 34
4.7 Control of transfection efficiency . . . 34
4.8 Sequence alignment of HC signal peptides . . . 35
4.9 Sequence alignment of LC signal peptides . . . 36
4.10 Sequence alignment of the evaluated HC and LC signal peptides and the reference signal peptides . . . 37
4.11 Process analysis . . . 38 5 Discussion 39 6 Conclusions 42 Acknowledgements 43 References 44 Appendix A. Materials 47 Appendix B. Cloning of plasmid constructs 51 B.1 Preparation of cloning fragments . . . 51
B.2 LC plasmid constructs . . . 52
B.3 HC and LC plasmid constructs . . . 53
1
Introduction
1.1
Background
Therapeutic monoclonal antibodies are well-established in the biopharmaceutical market with 79 approved products by the United States Food and Drug Administration (FDA) in December 2019 [1]. Interestingly, there is a variety of diseases that can be targeted with monoclonal antibodies (mAb): acute myeloid leukemia, multiple myeloma, multiple sclerosis, rheumatoid arthritis and severe asthma, just to mention a few [2][3]. In 2018, the global therapeutic monoclonal antibody market was valued at approximately US$ 115.2 billion and the market is still expanding [1]. Thus, the future for therapeutic antibodies shines bright with new monoclonal antibodies that are being commercialized and approved as a therapeutic every year [3].
The major production system for a monoclonal antibody is the expression as a re-combinant protein in a mammalian cell line [4], including Chinese Hamster Ovary (CHO) cells and Human Embryonic Kidney 293 (HEK293) cells [5]. Mammalian expression sys-tems are usually preferred for production of biopharmaceuticals due to their capacity for correct folding, protein secretion and post-translational modifications (PTMs), most notably glycosylation [6][7]. Nevertheless, mammalian cell cultures have a number of in-trinsic disadvantages causing their cultivation to be time-consuming and cost-intensive [8]. Therefore, research and development is crucial for improved production cell lines and cell culture processes [9].
Optimization work is an instrumental part in process development activities and differ-ent factors regarding the expression of recombinant antibodies can be studied to increase the product titer. An identified bottleneck within the secretory pathway of recombinant proteins is the translocation of secretory proteins from the cytosol into the lumen of the endoplasmic reticulum (ER) [10]. Moreover, in the process of protein secretion a central role is played by the signal peptide, which is located at the N-terminal of nascent proteins [11]. Previous studies have shown that optimization of signal peptides is a crucial step for attempting to achieve increased expression of recombinant antibodies in mammalian systems [10][12][13]. Furthermore, a previous study analyzed all naturally occurring hu-man antibody signal peptides and clustered them based on sequence similarity. Based on clustering, eight HC signal peptides and two LC signal peptides were generated [10].
1.2
Aim
The aim of this project was to evaluate the expression of three recombinant antibodies in HEK293 cells by evaluating different combinations of eight HC and two LC signal peptides. The evaluation of signal peptides were performed using three human immunoglobulin G subclass 1 (IgG1) recombinant antibodies with different variable domains, aimed for two different therapeutic areas: infectious disease, oncology and one control. The impact goal was an efficient secretion of recombinant antibodies, and thus an increased expression and lower production cost of recombinant antibodies in HEK293 cells.
1.3
Motivation
evaluation of signal peptides for the expression of recombinant antibodies in HEK293 cells is of great interest for SciLifeLab, especially for their Protein Expression and Char-acterization facility. The three human recombinant antibodies that were evaluated with different signal peptides were provided by the Drug Discovery and Development platform at SciLifeLab.
1.4
Work plan
The project was divided into several objectives:
• Preparation of cloning fragments: signal peptides, the human IgG1-backbone plas-mid and the variable domains for each of the three recombinant antibodies.
• Construction of LC plasmid constructs with different signal peptides. Transforma-tion of plasmid constructs into Escherichia coli (E. coli ) and thereafter plasmid preparation, purification and sequencing for quality control.
• Selection of successful LC plasmid constructs for further cloning with the HC vari-able domains and signal peptides.
• Construction of HC and LC plasmid constructs with different combinations of signal peptides. Transformation of plasmid constructs into E. coli and thereafter plasmid preparation, purification and sequencing for quality control.
• Selection of successful HC and LC plasmid constructs for transient transfection into HEK293 cells.
• Transient gene expression (TGE) of the three recombinant antibodies with 16 dif-ferent HC and LC signal peptide combinations in HEK293 cells.
• Quantification of expressed recombinant antibodies using Biacore concentration analysis.
• Analyzation of expressed recombinant antibody titers with signal peptides under investigation.
The objectives were divided into different activities and milestones. The milestones of the project are shown in Table 1. The different activities with their estimated time, actual consumed time as well as the milestones of the project are illustrated in a Gantt chart in Figure 1.
1.5
Limitations
Some materials, methods and results are not presented in full detail due to confidential-ity commitment to the Drug Discovery and Development platform at Science for Life Laboratory.
Table 1: Milestones for the master’s thesis project ”Optimizing signal peptides for expression of recombinant antibodies in HEK293 cells”.
Milestone Description Calendar Week
M1 Planning report finished 6
M2 LC plasmid constructs are finished 8 M3 HC and LC plasmid constructs are finished 12
M4 Half-time presentation 14
M5 First quantification of expressed antibodies 15 M6 Finished quantification of expressed antibodies 19 M7 Final report draft submitted 21 M8 Project presentation and review 23
Figure 1: Gan tt chart with actual consumed time, planned time and milestones for the master’s thesis pro ject ”Optimizing signal p eptides for expr ession of re combinant antib o di es in HEK293 cel ls” .
2
Theory
2.1
Scientific background
2.1.1 Antibody function and structure
An antibody, also referred to as an immunoglobulin (Ig), is a protein that recognize foreign material (e.g. pathogenic bacteria and viruses) through a unique molecular structure, called an antigen. Antibodies are part of the immune system and can be membrane-bound, functioning as an antigen receptor at the surface of B cells, or secreted in the blood plasma by plasma cells. The antibody recognize and bind to the antigen through its fragment antigen-binding (Fab) region. Thus, the antibody can tag the antigen and stimulate an immune response mediated via its fragment crystallizable (Fc) region, which interacts with receptors at the cell surface. Moreover, an antibody can also directly neutralize an antigen by inhibiting an important structure (e.g. blocking an active site at the pathogen). However, there are a variety of antibodies, specialized in various biological activities and with different structural properties [14].
Immunoglobulin G (IgG) is the most common isotype of antibodies in the blood and extracellular fluid and consists of two identical light chains (LCs) and two identical heavy chains (HCs) [14]. The LC is divided into the variable domain (VL) and the constant domain (CL). The HC is divided into the variable domain (VH), the constant domains (CH1-CH3) and a hinge region between the CH1 and CH2. Furthermore, the different chains are linked together with disulphide bridges. Moreover, the IgG molecule consists of two Fab regions and one Fc region. The Fc region comprise of the paired CH2 and CH3 domains and the Fab region consists of one variable and one constant domain from both the HC and LC [15]. A schematic structure of an IgG antibody can be seen in Figure 2.
CH2 CH3 CH2 CH3 Fc region Fab region
Figure 2: Schematic structure of an immunoglobulin G (IgG) antibody. The light chain (LC) (colored in blue), the heavy chain (HC) (colored in yellow), the hinge region (colored in black) and the disulphide bridges (colored in red). The LC is divided into the variable domain (VL) and the constant domain (CL). The HC is divided into the variable domain (VH) and the constant domains (CH1-CH3). The Fab region is composed of one variable and one constant domain from both the HC and LC. The Fc region consists of the paired CH2 and CH3 domains. Hence, an antibody consists of two Fab regions and one Fc region.
2.1.2 Secretory pathway
Recombinant antibodies are secreted from the mammalian host cell via the co-translational translocation pathway. First, each chain of the recombinant antibody is translated on the ribosome with a signal peptide at the N-terminal. The signal peptide is recognized by the signal recognition particle (SRP) while the protein still is being synthesized on the ribosome. The SRP transports the ribosome-nascent chain complex (RNC), consisting of the ribosome and the growing protein, to the endoplasmic reticulum (ER) membrane. Thereafter, at the ER membrane, SRP binds to the SRP receptor (SR) and delivers the RNC to the ER membrane-bound translocon which transports the synthesized protein into the ER lumen. Then, the signal peptide is cleaved off by a recruited signal pepti-dase (SPase) and the newly synthesized protein is released into the ER lumen [16]. The co-translational translocation pathway is illustrated in Figure 3.
mRNA Ribosome Signal peptide SRP Translated protein RNC SRP receptor Translocon Cytosol ER lumen Cytosol ER lumen 1. 2. 3. 4. Cytosol ER lumen SPase SRP receptor Translocon
Figure 3: The co-translational translocation pathway. 1. The protein is translated on the ribosome with a signal peptide at the N-terminal, which is detected by the signal recognition particle (SRP). 2. SRP binds to the signal peptide and transports the ribosome-nascent chain complex (RNC), containing the ribosome and the growing protein, to the endoplasmic reticu-lum (ER) membrane. 3. SRP binds to the SRP receptor and the RNC is transferred to the membrane-bound translocon, which transports the translated protein into the ER lumen. 4. The signal peptide is cleaved off by a recruited signal peptidase (SPase) and the synthesized protein is released from the ribosome into the ER lumen.
In the rough ER, the newly synthesized proteins are incorporated into transport vesi-cles. The transport vesicles fuse together with the cis-Golgi vesicle and release the syn-thesized proteins into the Golgi lumen. Then, the Golgi vesicle moves from the cis-position (nearest the ER) to the trans-cis-position (farthest from the ER), in a process called cisternal migration. When the Golgi vesicle reaches the trans-position the secretory
pro-teins are moved into the trans-Golgi reticulum. The secretory propro-teins are incorporated into transport vesicles which move towards the plasma membrane and release the secretory proteins by exocytosis [17].
2.1.3 Signal peptides
Signal peptides are located at the N-terminal of nascent proteins and are on average 15−30 amino acids in length for eukaryotes. Comparisons of signal peptide sequences indicated that they typically comprise of three distinct domains: a positively charged amino-terminal region (n-region), a central hydrophobic region (h-region) and a po-lar carboxy-terminal region (c-region). Apart from this three-domain structure with a ”positive-hydrophobic-polar” design, the signal peptide sequence is highly variable [11]. In order to address the different residue positions in the signal peptide, the last position is referred to as −1, the second last position as −2 and so on. The first residue position in the mature protein is referred to as +1. A schematic structure of a signal peptide can be seen in Figure 4.
The positively charged n-region is responsible for interactions with the phosphate backbone of the SRP and the phosphate group of lipid bilayers, which is crucial for an efficient translocation [18]. The n-region varies strongly with the overall length of the signal peptide, but usually consists of five residues [19]. Furthermore, the hydrophobic h-region determines the conformation of the signal peptide and is essential in protein pro-cessing and translocation [18]. Interestingly, previous studies suggested that hydrophobic h-regions stabilize the interactions between the signal peptide and SRP [20][21]. The h-region has some variability in length, though the most important hydrophobic residues consist of position −6 to −13 in eukaryotes [11]. Further, the c-region is independent of the total length of the signal peptide and usually comprise of residue −1 to −5 in eukaryotes [11]. The cleavage site for the SPase is located in the c-region, between residue −1 and +1. Interestingly, critical points in the c-region are located at position −1 and −3, well-known as the ”(−3, −1)-design” of signal peptides [22]. Furthermore, a pre-vious study presented conserved residues at position −1 and −3 in the signal peptide sequence through a rigorous analysis of 1877 eukaryotic signal peptide sequences. The results showed that small residues are conserved at position −1; alanine (A) and glycine (G). Position −3 favoured small aliphatic residues; alanine (A) and valine (V). Moreover, serine (S), threonine (T) and cysteine (C) are also noticeable at position −1 and −3 [23]. The conserved ”(−3, −1)-design” fits within a pocket in the catalytic domain of the SPase and thereby defines the cleavage site between position −1 and +1 [18].
NH2 –
n-region h-region c-region mature protein
cleavage site
Figure 4: Schematic structure of a signal peptide with the three distinct domains: a positively charged n-region, a hydrophobic h-region and a polar c-region. The cleavage site between the signal peptide and the mature protein is also visualized.
efficiency of the co-translational translocation pathway, which have been identified as a crucial step in the secretory pathway. An optimal secretion path for a recombinant protein would increase the product yield in a mammalian host system [12]. Nonetheless, there is a variety of signal peptides with various sequences that show completely different impact on the protein secretion [10][12][13]. However, a previous study analyzed all naturally occurring human antibody signal peptides (172 HC signal peptides and 62 LC signal peptides) and clustered them based on their sequence similarity. This resulted in eight HC signal peptides (H1-H8) and two LC signal peptides (L1-L2) [10]. These eight HC and two LC signal peptides were used in this project for the evaluation of different combinations of HC and LC signal peptides. The amino acid sequence for the eight HC signal peptides (H1-H8) and two LC signal peptides (L1-L2) are shown in Table 2.
Table 2: The amino acid sequence for the eight heavy chain (HC) signal peptides (H1-H8) and the two light chain (LC) signal peptides (L1-L2) that were used in this study.
HC signal peptide Amino acid sequence H1 MELGLSWIFLLAILKGVQC H2 MELGLRWVFLVAILEGVQC H3 MKHLWFFLLLVAAPRWVLS H4 MDWTWRILFLVAAATGAHS H5 MDWTWRFLFVVAAATGVQS H6 MEFGLSWLFLVAILKGVQC H7 MEFGLSWVFLVALFRGVQC H8 MDLLHKNMKHLWFFLLLVAAPRWVLS LC signal peptide Amino acid sequence
L1 MDMRVPAQLLGLLLLWLSGARC L2 MKYLLPTAAAGLLLLAAQPAMA
2.1.4 Surface plasmon resonance
Surface plasmon resonance (SPR) is a phenomenon that occurs in thin conducting films, often a gold layer, at the interface of materials with different refractive indices. First, incident light passes through a prism causing an internal reflection at the metal-prism interface. As a result, an electromagnetic evanescent wave occurs in the metal that prop-agates along the metal-ambient interface. At a certain energy and angle of the incident light, surface plasmons in the metal layer are excited which can be seen as a drop in the intensity of the reflected light [24][25], which is illustrated in Figure 5.
SPR angle Angle of reflection Reflected light
intensity
Figure 5: The surface plasmon resonance (SPR) detection principle. A specific angle of the incident light (referred to as the SPR angle) excite surface plasmons in the metal layer which result in a drop in the intensity of the reflected light [25].
2.1.5 Biacore biosensors
Biacore biosensor systems utilizes SPR to monitor molecular interactions in real-time and can be used to study specificity, kinetics, affinity and concentration analysis. Biacore systems measures the change in the refractive index, which is due to molecular interactions at the sensor surface. In Biacore systems, the conducting film is a gold layer and the materials with different refractive indices are a glass slide and the sample solution that flows through the flow cell. A schematic illustration of a Biacore system can be seen in Figure 6. The gold film is coated with a carboxymethylated dextran matrix to which ligand molecules are immobilized. In direct binding assays, analytes in the sample solution binds directly to the immobilized ligands, which alters the refractive index. Thus, the response is directly proportional to the analyte concentration at the sensor surface. The SPR signal is recorded in a sensorgram which plots response against time, and hence showing the progress of the interaction between the analyte and the ligand [25], which is illustrated in Figure 7.
SPR angle Reflected light Incident light Glass slide Gold layer Sensor chip Flow cell
Figure 6: Schematic illustration of a Biacore system. The conducting film is the gold layer and the media with different refractive indices are the glass slide of the sensor chip along with the sample solution that flows through the flow cell. The SPR angle refers to the angle of the incident light that result in a drop in the intensity of the reflected light [25].
Response signal (RU)
Time
Buffer Sample Buffer Regeneration Buffer Baseline Association
Dissociation
Analyte Ligand
Figure 7: Schematic illustration of a sensorgram with the different phases and the detection of binding events in a direct binding assay. The bars below the sensorgram curve indicate the solution that flows over the sensor surface. First, buffer flows over the sensor surface with immobilized ligands, which results in a baseline response signal. Thereafter, as the analyte begins to bind to the immobilized ligand, the refractive index on the sensor surface change and the response signal increases. After sample injection, buffer flows over the sensor surface and the analyte starts to dissociate from the ligand, resulting in a decrease of the response signal. Then, regeneration solution removes bound analyte from the ligand at the sensor surface and the signal response returns back to baseline [25].
2.2
Methodology
2.2.1 In-Fusion cloning
A key step in the expression of a recombinant protein is the construction of a vector containing the target protein. Traditionally, restriction enzymes and DNA ligase are used to manipulate the plasmid vector, which is quite time consuming and relies on unique sites for the restriction enzymes. In contrast, In-Fusion cloning uses a recombinant exonuclease that exposes a complementary sequence, containing 15 bases, at the vector and insert. The 15 bases homologous overlap fuses the vector and insert together. The design of primers to create these overlaps, at both ends, for the insert are therefore key for a successful In-Fusion cloning [26]. In-Fusion cloning can join multiple DNA fragments into a vector during a single reaction, as long as they share 15 bases of overlap at each end [27]. An illustration of an In-Fusion cloning can be seen in Figure 8.
1. Linearized plasmid Insert 2. Linearized plasmid Insert 3. Plasmid Insert
Figure 8: In-Fusion cloning with two fragments: insert and plasmid. Before the In-Fusion cloning the plasmid is linearized with restriction enzymes. 1. Thereafter, the In-Fusion cloning starts by mixing the two fragments and a recombinant exonuclease. 2. The exonuclease exposes the 15 bases overlaps at each fragment. 3. The homologous overlaps fuses the two fragments together, resulting in a plasmid which contains the desired insert.
2.2.2 Cloning approach
The plasmid constructs in this project were based on an existing vector provided by the facility that contained the constant domains (CL and CH) for a human IgG1 antibody along with VL and VH stuffer fragments. Also, all of the plasmid constructs in this project had the same promoter sequence and regulatory elements. Before cloning, the plasmid was linearized with restriction enzymes that cleaved the VL stuffer fragment. Thereafter, each VL was fused to the two LC signal peptides along with the CL in the plasmid by In-Fusion cloning. Three LC In-Fusion reactions were carried out, one for each antibody. The LC In-Fusion cloning resulted in two different constructs for each antibody: one with L1 as a signal peptide and one with L2. An illustration of the LC In-Fusion cloning can be seen in Figure 9. Thereafter, the fused LC plasmid constructs were transformed into competent E. coli cells for amplification. The colonies were screened for a successful cloning with colony PCR. After plasmid purification from successful colonies, restriction digest screening was used to determine which colonies that contained which LC signal peptide, L1 or L2. Then, the purified plasmid constructs were sequenced for quality control.
1. Linearized plasmid SPLC 2. Linearized plasmid VL 3. LC plasmid VL VL 2.1 Linearized LC plasmid VL SPLC SPLC SPHC SPLC VH Linearized LC plasmid VL SPHC SPLC VH LC + HC plasmid VL SPHC SPLC VH 2.2 2.3
Figure 9: In-Fusion cloning with LC variable domain (VL), LC signal peptides (SPLC) and
plasmid. First, the plasmid is linearized with restriction enzymes. 1. The In-Fusion cloning starts by mixing the three fragments and a recombinant exonuclease. 2. The exonuclease exposes the 15 bases overlap at the VL, SPLC and plasmid. 3. The homologous overlaps fuses the three
fragments together, resulting in a LC plasmid construct.
After the LC In-Fusion cloning six different LC plasmid constructs were generated in total: one construct with L1 as a signal peptide and one construct with L2, for each of the three antibodies. These LC plasmid constructs were linearized with restriction enzymes that cleaved the VH stuffer fragment. Six HC In-Fusion reactions were carried out, one for each LC plasmid construct. In the HC In-Fusion cloning the VH was fused to the eight HC signal peptides along with the CH in the human IgG1-backbone plasmid. The HC In-Fusion cloning resulted in plasmid constructs that contained both the LC and the HC with their respective signal peptide, which is illustrated in Figure 10. These plasmid constructs were transformed into competent E. coli cells for amplification. Thereafter, the plasmids were purified and screened for a successful cloning with restriction digest screening. Then, the plasmid constructs with a successful cloning were sequenced for quality control and most importantly to determine which colonies that contained which combination of the HC and LC signal peptides.
Finally, after a successful two-step In-Fusion cloning, 16 different plasmid constructs were generated for each antibody: all combinations of the two LC signal peptides and the eight HC signal peptides. Hence, 48 different plasmid constructs in total.
1. Linearized plasmid SPLC 2. Linearized plasmid VL 3. LC plasmid VL VL 1. Linearized LC plasmid SPLC SPLC SPHC SPLC VH SPHC VH HC and LC plasmid 2. 3. VL Linearized LC plasmid SPLC VL SPLC VL VH SPHC
Figure 10: In-Fusion cloning with HC variable domain (VH), HC signal peptides (SPHC) and
LC plasmid construct. First, the LC plasmid is linearized with restriction enzymes. 1. The In-Fusion cloning starts by mixing the three fragments and a recombinant exonuclease. 2. The exonuclease exposes the 15 bases overlap at the VH, SPHC and LC plasmid. 3. The homologous
2.2.3 Colony PCR screening
Colony PCR is a rapid screening method to determine which colonies, following trans-formation, that carries the desired genetic insert. First, PCR primers are designed that generate a product of known size only if the desired insert is present. Thereafter, the bacteria are thermally lysed and their DNA are used as template in the PCR. Following the amplification, the PCR products are analysed using gel electrophoresis. Colonies with gel bands of the correct size indicates a successful transformation. Hence, colony PCR can be used to distinguish between a successful and an unsuccessful transformation [28].
2.2.4 Restriction digest screening
Restriction digest can be used as a screening method for purified plasmid constructs to determine if the cloning was successful, i.e. if the plasmid construct contains the desired insert. The plasmid constructs are cleaved with restriction enzymes that generate different digestion patterns depending on if the desired insert is present or not. Thereafter, the digested plasmid constructs are analysed using gel electrophoresis. Restriction mapping and in silico digestions are performed to determine which restriction enzymes that can be used to identify the presence of the desired insert [29].
2.2.5 Transient transfection
Transfection is a method that introduces foreign DNA into eukaryotic cells and can either be used to yield transient or stable expression of the gene of interest. For stable transfec-tion, the foreign DNA is integrated into the host genome, stably expressed and passed on to daughter cells. In contrast, transiently transfected cells express the introduced DNA for a limited amount of time and the foreign DNA is not integrated into the host genome. Thus, transiently transfected DNA is not passed on to daughter cells [30]. Neverthe-less, transient transfection is a rapid method for the expression of recombinant proteins in mammalian cells without having to establish a stable cell line [31]. However, foreign DNA can be introduced into eukaryotic cells with a variety of methods and transfection reagents: calcium phosphate, cationic polymers, cationic lipids and electroporation, just to mention a few [32]. An excess of cationic transfection reagent interacts with the nega-tively charged DNA and forms posinega-tively charged complexes, which is endocytosed by the transfected cell [33]. For plasmid DNA, efficient transfection reagents delivers the gene into the nucleus of the host cell, where it is expressed [34]. In this project, FectoPRO® was used, which is an optimized transfection reagent for transient gene expression (TGE) of recombinant proteins and mAb in suspension CHO cells and HEK293 cells [35].
2.2.6 Transfection efficiency
The transfection efficiency is a crucial parameter to determine in gene expression ex-periments and describes the proportion of cells that have been successfully transfected. Commonly, transfection efficiency is often monitored through fluorescent reporter genes, including the expression of green fluorescent protein (GFP) [36][37]. In this project, GFP was used as an internal control for transfection efficiency and was transfected in a sepa-rate well for each cell culture plate. The transfection efficiency was determined using a Countess™ II FL Automated Cell Counter.
2.2.7 Biacore concentration analysis
Biacore systems can be utilized for many purposes, including concentration analysis. In direct binding concentration assays, responses from samples with known analyte concen-trations are monitored and used to calculate a standard curve. Thereafter, the standard curve can be utilized to determine the concentration of samples with unknown concentra-tions [25]. In this project, Biacore T200 and Series S Sensor Chip Protein A were used for the concentration assay. The sensor surface is precoated with recombinant Protein A, which predominantly binds human antibodies of the subclasses IgG1, IgG2 and IgG4. The recombinant Protein A only binds to the HC within the Fc region, thus ensuring an orientation-specific binding of the antibody [38]. Hence, in the concentration assay the immobilized ligand was the recombinant Protein A and the analytes were the expressed recombinant antibodies. Moreover, the standard curve was calculated from samples with known concentrations of Herceptin® (Trastuzumab), which is a recombinant humanized IgG1 mAb [39].
2.2.8 Biacore kinetic screening
Determination of real-time interaction kinetics with high resolution can be studied using Biacore systems. Further, antibody screening can be performed to identify antibody constructs with suitable kinetic and affinity properties against an antigen. The affinity for the binding interaction, such as the equilibrium dissociation constant (KD), can be
derived from the kinetic rate constants for the interaction (ka for association rate constant
and kd for dissociation rate constant) [40]. In this project, a Human Fab Capture Kit was
used for immobilization at the sensor surface on a Sensor Chip CM5. The sensor surface was coated with capturing molecules, monoclonal antibodies that bind specifically to the LC of human Fab regions [41]. Thereafter, the expressed recombinant antibodies, referred to as the ligand, was flowed across the sensor surface and captured through interactions with the immobilized capturing molecules. Then, the antigen, referred to as the analyte, was injected and the binding interaction was monitored [40].
3
Methods
The materials and instruments used in this project are found in Appendix A, except for general laboratory equipment. Computer software used in the project are cited in the text.
3.1
Preparation of variable domains
3.1.1 PCR amplification of variable domains
The two variable domains (VH and VL) for each of the three recombinant antibodies (referred to as Antibody A, Antibody B and Antibody C) were amplified respectively with CloneAmp™ HiFi PCR Premix (2X). 1µL of each variable domain (10 ng/µL) were added to PCR tubes along with 1µL forward primer (10 µM), 1 µL reverse primer (10 µM), 25 µL CloneAmp™ HiFi PCR Premix (2X) and 22µL dH2O, resulting in a total volume of 50µL.
The melting temperatures (Tm) for the forward and reverse primer were calculated to be
65°C and 72°C respectively, using Thermo Fisher Scientific TmCalculator [42]. Therefore,
a two-step PCR was feasible since the primers annealing temperature were within 3°C of the extension temperature at 68°C [43]. Thus, the amplifications were performed with a two-step PCR with the following program: a denaturation step at 98°C for 10 seconds, followed by a combined annealing and extension step at 68°C for 30 seconds. The thermal cycle was repeated 30 times and thereafter the temperature decreased to 4°C.
3.1.2 Purification of LC variable domains
After the amplification, the LC variable domains were purified using GeneJET PCR Purification Kit. 50 µL dH2O, 100 µL Binding buffer and 100 µL 2-propanol were added
to each PCR mixture, resulting in a total volume of 300µL. Then, the PCR mixtures were transferred to GeneJET purification columns and washed according to the manufacturer’s protocol. The VL were eluted in 30 µL Elution buffer. The concentration of the VL were determined using an Implen NanoPhotometer® NP80.
Then, the VL were analysed with gel electrophoresis for quality control. 1 µL of each VL was mixed with 1 µL FlashGel™ Loading Dye (5X) and 3 µL dH2O and applied to
the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons. Thereafter, the VL were further purified with gel electrophoresis. 5 µL FlashGel™ Loading Dye (5X) was added to 20 µL of each VL and applied to the FlashGel™ Recovery Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder. The identified gel bands for the VL were extracted from the FlashGel™ Recovery Cassette (1.2% agarose). The concentration of the extracted VL were determined using an Implen NanoPhotometer® NP80. To control that the gel electrophoresis purification had worked 4 µL of each VL was mixed with 1 µL FlashGel™ Loading Dye (5X) and applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder.
3.1.3 Purification of HC variable domains
After the amplification, the HC variable domains (VH) were purified using GeneJET PCR Purification Kit. 50 µL dH O, 100 µL Binding buffer and 100 µL 2-propanol were added
transferred to GeneJET purification columns and washed according to the manufacturer’s protocol. The VH were eluted in 30 µL Elution buffer. The concentration of the VH were determined using an Implen NanoPhotometer® NP80.
Then, the VH were analysed with gel electrophoresis for quality control. 1µL of each VH was mixed with 1 µL FlashGel™ Loading Dye (5X) and 3 µL dH2O and applied to
the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons. Thereafter, the VH were further purified with gel electrophoresis. 7.5 µL FlashGel™ Loading Dye (5X) was added to 30 µL of each VH and applied to the FlashGel™ Recovery Cassette (1.2% agarose). The identified gel bands for the VH were extracted from the FlashGel™ Recovery Cassette (1.2% agarose). The concentration of the extracted VH were determined using an Implen NanoPhotometer® NP80. Then, the VH were concentrated using GeneJET PCR Purification Kit. 100 µL Binding buffer and 100 µL 2-propanol were added to each VH. The VH were washed and eluted in 30 µL Elution buffer according to the manufacturer’s protocol. The concentra-tion of the VH were determined using an Implen NanoPhotometer® NP80. For control, 1 µL of each VH was mixed with 1 µL FlashGel™ Loading Dye (5X), 3µL dH
2O and applied
to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder.
3.2
Preparation of signal peptides
A vector containing all of the signal peptides (L1-L2 and H1-H8) with their respective 15 bases In-Fusion overlap was ordered from GeneArt™ (Thermo Fisher Scientific). The vector was designed with cleavage sites for the restriction enzyme BveI FastDigest™ be-tween every signal peptide construct. Therefore, the vector was cleaved to generate a mixture containing all of the signal peptides. 5 µL vector (3 µg) was added to a PCR tube along with 6µL FastDigest™ buffer (10X), 3µL BveI FastDigest™ restriction enzyme, 3 µL oligonucleotide (20X) and 43 µL dH2O, resulting in a total volume of 60 µL. The
PCR tube was incubated at 37°C for three hours, and then the enzymes were inactivated by heating for five minutes at 80°C.
Thereafter, the signal peptides were purified using gel electrophoresis. 15µL FlashGel™ Loading Dye (5X) was added to 60µL solution containing the signal peptides and applied to the FlashGel™ Recovery Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder. The identified gel bands for the signal peptides were extracted from the FlashGel™ Recovery Cassette (1.2% agarose). Then, the concentration of the signal peptides were determined using an Implen NanoPhotometer® NP80. To control that the gel electrophoresis purification had worked 4 µL solution containing the signal peptides was mixed with 1 µL FlashGel™ Loading Dye (5X) and applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder.
3.3
Linearization of plasmid backbone
A human IgG1-backbone plasmid with a VL and VH stuffer fragment was provided by the facility. Before the LC In-Fusion cloning the plasmid was linearized with restriction enzymes that cleaved the VL stuffer fragment. Restriction mapping were performed to determine which restriction enzymes that could be used, using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). 6 µL plasmid (3 µg) was added to a PCR tube along with 6 µL
FastDigest™ buffer (10X), 3 µL of each restriction enzyme (Pfl23II FastDigest™ and SgsI FastDigest™) and 42 µL dH2O, resulting in a total volume of 60 µL. The PCR tube was
incubated at 37°C for three hours and thereafter the enzymes were inactivated by heating for five minutes at 80°C.
The plasmid was purified using GeneJET PCR Purification Kit. 40µL dH2O and 100
µL Binding buffer were added to the plasmid solution, resulting in a total volume of 200 µL. Then, the solution was transferred to a GeneJET purification column and the plasmid was purified and eluted in 30 µL Elution buffer according to the manufacturer’s protocol. The concentration of the plasmid was determined using an Implen NanoPhotometer® NP80. For control, 1 µL FlashGel™ Loading Dye (5X) and 3 µL dH2O were added to
1 µL plasmid and applied to the FlashGel™ DNA Cassette (1.2% agarose) along with a control (non-cleaved) plasmid and 5 µL FastRuler Middle Range DNA Ladder for size comparisons.
3.4
Construction of LC plasmid constructs
3.4.1 In-Fusion cloning of LC variable domains
2 µL of each VL (≈ 14 ng) were added to a PCR tube along with 4 µL signal peptides (≈ 13 ng), 1 µL linearized plasmid (≈ 65 ng), 2 µL In-Fusion® HD Enzyme Premix (5X) and 1 µL dH2O, resulting in a total volume of 10µL. For control, 1 µL linearized plasmid
(≈ 65 ng), 2 µL In-Fusion® HD Enzyme Premix (5X) and 7 µL dH2O were added to a
PCR tube. The four PCR tubes (three different VL and one control) were incubated at 50°C for 15 minutes according to the manufacturer’s protocol.
3.4.2 Plasmid transformation into competent E. coli
3 µL In-Fusion product and 50 µL Stellar™ Competent Cells were added to a 1.5 mL microcentrifuge tube and incubated on ice for 30 minutes. Thereafter, the cells were heat shocked at 42°C for 45 seconds and placed back on ice for 2 minutes. 450 µL Invitrogen™ S.O.C. Medium was added to the transformed cells followed by shake-incubation at 37°C and 240 rpm for 1 hour. Then, the cells were spread onto agar plates with 100 µg/mL Carbenicillin and incubated overnight at 37°C.
3.4.3 Colony PCR screening of LC plasmid constructs
Initially, colony PCR screening was not performed and plasmid DNA were directly isolated with minipreparation. However, due to a high transformation background for three of the LC plasmid constructs, colony PCR was used to select colonies for these constructs.
Colonies for colony PCR were picked from agar plates using pipette tips and put in wells of a PCR plate. 1 µL (1 ng) plasmid with a successful insert was added to a well for positive control. 1 µL (1 ng) plasmid with no insert was added to a well for negative control along with a well with no added colony or plasmid. Thereafter, the pipette tips were moved to a 24 well cell culture plate and the colonies were inoculated in 2.5 mL LB medium containing 100 µg/mL Carbenicillin. The samples were shake-incubated overnight at 37°C and 240 rpm.
resulting in a total volume of 25 µL. The colony PCR was performed with the following program: an initial step at 95°C for 7 minutes, followed by a thermal cycle of 95°C for one minute and 68°C for one minute, which was repeated 25 times. After the thermal cycle, the temperature was kept at 68°C for 7 minutes and thereafter the temperature decreased to 4°C. 3 µL of each PCR reaction was directly applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons. Colonies with identified gel bands of the correct size, indicating a successful transformation, were used for minipreparation of plasmid DNA.
3.4.4 Minipreparations of LC plasmid constructs
Colonies that were not screened with colony PCR were picked from agar plates using sterile culture loops and inoculated in 5 mL LB medium containing 100 µg/mL Carbenicillin in 14 mL Falcon tubes. The samples were shake-incubated overnight at 37°C and 240 rpm. Thereafter, 1.8 mL overnight cell culture, either screened with colony PCR or not, was transferred to a 2 mL microcentrifuge tube and centrifuged at 17,000 × g for 2 minutes at room temperature. The supernatants were thrown to waste. The cell pellets were resuspended and lysed followed by plasmid DNA purification using GeneJET Plasmid Miniprep Kit. The plasmid constructs were eluted in 50 µL Elution buffer according to the manufacturer’s protocol. Then, the concentration of the plasmid constructs were determined using an Implen NanoPhotometer® NP80.
3.4.5 Restriction digest screening of LC plasmid constructs
In silico digestions with the restriction enzymes NcoI and XbaI were performed on LC plasmid constructs, for both L1 and L2, and one control plasmid with no inserts using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). 1 µL plasmid construct was added to a PCR tube along with 0.5µL FastDigest™ buffer (10X), 0.25µL of each restriction enzyme (NcoI FastDigest™ and XbaI FastDigest™) and 3µL dH2O, resulting in a total volume of
5µL. The PCR tubes were incubated at 37°C for 20 minutes, and then the enzymes were inactivated by heating for five minutes at 80°C. 1.25 µL FlashGel™ Loading Dye (5X) was added to each PCR tube and thereafter 5 µL was applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons. The in silico digestions were used as references for digestion patterns in the gels. Plasmid constructs that generated gel bands with digestion patterns that indicated a successful cloning were sequenced for further quality control.
3.4.6 Sequencing of LC plasmid constructs
LC plasmid constructs were sequenced at KIGene, Karolinska Institutet. The obtained DNA sequence electropherogram for each LC plasmid construct was analysed using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). LC plasmid constructs that showed a correct DNA sequence when mapped to the reference plasmids were accepted as candidates for further cloning. Two candidates for each antibody, one with L1 as a signal peptide and one with L2, were chosen for cloning with the VH. Hence, six LC plasmid constructs in total.
3.5
Linearization of LC plasmid constructs
After sequencing, the six chosen LC plasmid constructs were linearized with restriction enzymes that cleaved the VH stuffer fragment. Restriction mapping were performed to determine which restriction enzymes that could be used, using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). 2 µg LC plasmid construct was added to a PCR tube along with 6 µL FastDigest™ buffer (10X), 1.5 µL of each of the four restriction enzymes (BamHI FastDigest™, BspTI FastDigest™, Eco91I FastDigest™ and HindIII FastDigest™) and dH2O,
resulting in a total volume of 60 µL. All of the LC plasmid constructs had different concentrations from the minipreparations, and thus the volume of dH2O differed from
each reaction so that the total volume was 60µL. The PCR tubes were incubated at 37°C for three hours and thereafter the enzymes were inactivated by heating for five minutes at 80°C.
Thereafter, the LC plasmid constructs were purified using GeneJET PCR Purification Kit. 40 µL dH2O and 100 µL Binding buffer were added to the LC plasmid constructs,
resulting in a total volume of 200µL. The solutions were transferred to a GeneJET purifi-cation columns and the LC plasmid constructs were purified and eluted in 30µL Elution buffer according to the manufacturer’s protocol. The concentration of the LC plasmid constructs were determined using an Implen NanoPhotometer® NP80. For control, 1 µL FlashGel™ Loading Dye (5X) and 3 µL dH
2O were added to 1 µL LC plasmid
con-struct and applied to the FlashGel™ DNA Cassette (1.2% agarose) along with two control plasmids (one non-cleaved plasmid and one cleaved, but non-purified plasmid) and 5 µL FastRuler Middle Range DNA Ladder for size comparisons.
3.6
Construction of HC and LC plasmid constructs
3.6.1 In-Fusion cloning of HC variable domains
2 µL of each VH (≈ 16 ng) were added to a PCR tube along with 4 µL signal peptides (≈ 13 ng), 2µL linearized LC plasmid construct (≈ 66 ng), 4 µL In-Fusion® HD Enzyme Premix (5X) and 8 µL dH2O, resulting in a total volume of 20 µL. For control, 1 µL
linearized LC plasmid construct (≈ 66 ng), 2 µL In-Fusion® HD Enzyme Premix (5X) and 7 µL dH2O were added to a PCR tube. The PCR tubes were incubated at 50°C for
15 minutes according to the manufacturer’s protocol.
3.6.2 Plasmid transformation into competent E. coli
5 µL In-Fusion product and 50 µL Stellar™ Competent Cells were added to a 1.5 mL microcentrifuge tube and incubated on ice for 30 minutes. Thereafter, the cells were heat shocked at 42°C for 45 seconds and placed back on ice for 2 minutes. 450 µL Invitrogen™ S.O.C. Medium was added to the transformed cells followed by shake-incubation at 37°C and 240 rpm for 1 hour. Then, the cells were spread onto agar plates with 100 µg/mL Carbenicillin and incubated overnight at 37°C.
3.6.3 Colony PCR screening of HC and LC plasmid constructs
Initially, colony PCR screening was performed for one of the In-Fusion products to rapidly determine the success rate of the transformation. Thereafter, because of a successful
transformation, colony PCR screening was not performed and the HC and LC plasmid constructs were directly purified with minipreparations.
Colonies for the initial screening were picked from agar plates using pipette tips and put in wells of a PCR plate. 1 ng plasmid with a successful insert was added to a well for positive control. 1 ng plasmid with no insert was added to a well for negative control along with a well with no added colony or plasmid. Thereafter, the pipette tips were moved to a 24 well cell culture plate and the colonies were inoculated in 2.5 mL LB medium containing 100 µg/mL Carbenicillin. The samples were shake-incubated overnight at 37°C and 240 rpm.
12.5µL DreamTaq™ Green PCR Master Mix (2X) was added to each PCR well along with 0.5 µL (10 µM) forward primer, 0.5 µL (10 µM) reverse primer and 11.5 µL dH2O,
resulting in a total volume of 25 µL. The colony PCR was performed with the following program: an initial step at 95°C for 7 minutes, followed by a thermal cycle of 95°C for one minute and 68°C for one minute, which was repeated 25 times. After the thermal cycle, the temperature was kept at 68°C for 7 minutes and thereafter the temperature decreased to 4°C. 3 µL of each PCR reaction was directly applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons.
3.6.4 Minipreparations of HC and LC plasmid constructs
Colonies were picked from agar plates using pipette tips and inoculated in 3 mL LB medium containing 100 µg/mL Carbenicillin in a 24 well cell culture plate. The cultures were shake-incubated overnight at 37°C and 240 rpm. Thereafter, 100 µL of each cell culture was transferred into a PCR plate and stored at 4°C. The 24 well cell culture plates were centrifuged at 2,500 × g for 10 minutes at 4°C. The supernatants were thrown to waste. The cell pellets were resuspended and lysed, followed by plasmid DNA purification using QIAGEN® Plasmid Plus 96 Miniprep Kit and QIAvac 96 Vacuum Manifold. The plasmid constructs were eluted in 80 µL Elution buffer according to the manufacturer’s protocol.
3.6.5 Restriction digest screening of HC and LC plasmid constructs
In silico digestions were performed on HC and LC plasmid constructs and control plas-mids with no inserts, either with the restriction enzymes NcoI and XbaI or ApaI and HindIII, using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). 200 ng plasmid construct was added to a PCR tube along with 0.5 µL FastDigest™ buffer (10X), 0.25 µL of each restriction enzyme (either NcoI FastDigest™ and XbaI FastDigest™ or ApaI FastDigest™ and HindIII FastDigest™) and dH2O, resulting in a total volume of 5 µL. The PCR tubes
were incubated at 37°C for 20 minutes, and then the enzymes were inactivated by heating for five minutes at 80°C. 1.25 µL FlashGel™ Loading Dye (5X) was added to each PCR tube and thereafter 5 µL was applied to the FlashGel™ DNA Cassette (1.2% agarose) along with 5 µL FastRuler Middle Range DNA Ladder for size comparisons. The in silico digestions were used as references for digestion patterns in the gels. Plasmid constructs that generated gel bands with digestion patterns that indicated a HC and LC plasmid construct were sequenced for further quality control.
3.6.6 Sequencing of HC and LC plasmid constructs
HC and LC plasmid constructs were sequenced at KIGene, Karolinska Institutet. The obtained DNA sequence electropherograms were analysed and the HC signal peptide was determined for each HC and LC plasmid construct using Geneious Prime® v. 9.1.8 (Biomatters Ltd.). HC and LC plasmid constructs that showed a correct DNA sequence when mapped to the reference plasmids were accepted as candidates for transfection. 16 plasmid constructs for each antibody, with all the HC and LC signal peptide combinations, were chosen for transient transfection into HEK293 cells.
3.7
Sterile filtration of plasmid constructs
Before transfection, the plasmid constructs were sterile filtered using Corning® Costar® Spin-X® centrifuge tube filter with 0.22 µm pore size cellulose acetate membrane. The plasmid constructs from the minipreparations were transferred to the centrifuge tube filters and centrifuged at 17,000 × g for 5 minutes at room temperature. The concentration of the sterile plasmid constructs were determined using an Implen NanoPhotometer® NP80.
3.8
Maintenance of HEK293 cells
A maintenance culture of Expi293F™ cells were cultivated and split twice a week, either a 3-day split or a 4-day split. Total cell count and percent viability for the maintenance culture was determined before each split using a Countess™II FL Automated Cell Counter. 50 µL cell culture was mixed with 50 µL Trypan Blue Stain (0.4%). 10 µL sample mixture was loaded to each chamber in the sample slide and analysed according to the manufacturer’s protocol. At each split, Expi293F™cells were transferred to fresh Expi293™ Expression Medium, pre-warmed to 37°C, in an Erlenmeyer cell culture flask. The seeding density was either 0.75 × 106 cells/mL or 0.6 × 106 cells/mL, for a 3-day split or a 4-day
split respectively. The maintenance culture was shake-incubated at 37°C, 125 rpm, 50 mm shaking diameter, 8% CO2 and 70% relative humidity.
3.9
Transient gene expression of recombinant antibodies
3.9.1 Transfection of plasmid constructs into HEK293 cells
On the day prior to transfection (Day −1), total cell count and percent viability was determined using a Countess™ II FL Automated Cell Counter, according to the man-ufacturer’s protocol. Then, Expi293F™ cells were diluted in fresh Expi293™ Expression Medium, pre-warmed to 37°C, to a final density of 2.35 × 106 cells/mL. Thereafter, the
cells were shake-incubated overnight at 37°C, 125 rpm, 50 mm shaking diameter, 8% CO2
and 70% relative humidity. On the next day (Day 0), total cell count and percent viabil-ity was determined using a Countess™ II FL Automated Cell Counter, according to the manufacturer’s protocol. The Expi293F™ cells were diluted in fresh Expi293™ Expression Medium, pre-warmed to 37°C, to a final density of 2.5 × 106 cells/mL. The total transient
gene expression (TGE) volume for each plasmid construct was 4 mL. 3.6 mL Expi293F™ cells (2.5 × 106 cells/mL) (0.9 mL × total TGE volume) were added to a round bottom 24 well cell culture plate. 3.2µg plasmid DNA (0.8 µg × total TGE volume) was added to µL FectoPRO® ® µL : 1 µg) was
mixed with 400µL Opti-MEM™ I Reduced Serum Medium (0.1 mL × total TGE volume) and transferred to the plasmid DNA in the 96 deep well plate. The transfection mixture containing plasmid DNA, FectoPRO® and Opti-MEM™ I Reduced Serum Medium was homogenized and incubated at room temperature for 20 minutes. After incubation, the transfection mixture was added to the Expi293F™ cells. Thereafter, the transfected cells were incubated at 37°C, 8% CO2, 80% relative humidity and placed on a MixMate®
shaker at 400 rpm.
In each round bottom 24 well cell culture plate, one well was transfected with pmaxGFP™ vector, which was used as an internal control for transfection efficiency. Moreover, a hu-man IgG1-backbone plasmid with the reference HC and LC signal peptides used by the facility for each recombinant antibody were transfected in one well in each cell culture plate. The antibody titer for the reference HC and LC signal peptides for each recom-binant antibody was well-documented at the facility, and therefore used as a control in the TGE. The amino acid sequence for the reference HC and LC signal peptides used by the facility are shown in Table 3. The TGE experiment described above was repeated three times for each HC and LC signal peptide construct. Hence, three independent transfections for the 16 different combinations of the HC and LC signal peptides for each antibody were performed. Thus, 144 transfections were carried out in total, excluding controls (3 repeats × 3 antibodies × 16 HC and LC plasmid constructs). Therefore, the TGE experiments were carried out in seven independent batches, one batch consisted of a 24 well round bottom cell culture plate.
Table 3: The reference heavy chain (HC) and light chain (LC) signal peptides used by the facility and their respective amino acid sequence.
Reference signal peptide Amino acid sequence
HC signal peptide MSVSFLIFLPVLGLPWGVLS LC signal peptide MEAPAQLLFLLLLWLPDTTG
3.9.2 Measurement of transfection efficiency
After 48 hours post-transfection (Day 2), the transfection efficiency was determined. The percentage of Expi293F™ cells that were transfected with GFP was used as an inter-nal control for transfection efficiency. 1 mL cell culture that had been transfected with pmaxGFP™ vector was collected. 50 µL cell culture was mixed with 50 µL Trypan Blue Stain (0.4%). 10 µL sample mixture was loaded to each chamber in the sample slide and analysed using a Countess™ II FL Automated Cell Counter, according to the manufac-turer’s protocol.
3.9.3 Harvest of expressed recombinant antibodies
After 120 hours post-transfection (Day 5), the expressed recombinant antibodies were harvested. The round bottom 24 well cell culture plates were centrifuged at 2,500 × g for 5 minutes at 4°C. Thereafter, the culture supernatants were collected and stored at −20°C until further analysis.
3.10
Biacore concentration analysis of recombinant antibodies
First, HBS-EP+ buffer (pH 7.4) was prepared, containing 10 mM HEPES, 150 mM sodium chloride, 3mM EDTA and 0.05% TWEEN® 20, and used as running buffer. 10 mM Glycine (pH 1.5) was prepared and used as regeneration buffer. Then, standard curve samples were prepared from Herceptin® stock solution (1 mg/mL), diluted in Expi293F™ Expression Medium to a final concentration according to Table 4. Thereafter, the cul-ture supernatants from the TGE of recombinant antibodies were diluted in Expi293F™ Expression Medium. First, dilution factors for each recombinant antibody were based on expected antibody titers, according to previous data from the facility. After quan-tification, the dilution factors were evaluated and selected for each HC and LC signal peptide construct in order for the antibody titers to lie within the linear range of the standard curve. After an appropriate dilution, the culture supernatants were filtered using MultiScreenHTS HV Filter 96 well plate, with 0.45 µm pore size membrane. The
culture supernatans were transferred to the 96 well filter plate and centrifuged at 700 × g for 2 minutes at room temperature. Series S Sensor Chip Protein A was docked to the Biacore T200, the samples were loaded and the system was primed with running buffer. Thereafter, the Biacore analysis was performed with the following settings: two startup cycles with running buffer. Sample injection with 135 seconds sample contact time, 10 µL/minute sample flow rate. Regeneration with 30 seconds regeneration buffer contact time, 30 µL/minute regeneration buffer flow rate. The standard curve was measured before the first sample and after every 24th sample.
Table 4: The ten different concentrations of Herceptin® that were used as standard samples for the standard curve in the Biacore concentration analysis.
Standard curve
Analyte 1 2 3 4 5 6 7 8 9 10
Herceptin® (µg/mL) 0.5 1 3 5 10 15 20 30 40 50
3.11
Statistics
The quantified titer for each recombinant antibody was analyzed using a two-way analysis of variance (ANOVA) with a full factorial model, with the LC signal peptide and HC signal peptide as independent variables. Moreover, post hoc Tukey’s honestly significant difference (HSD) test for multiple comparisons were also performed. Statistical data analysis was performed with IBM SPSS Statistics 26 (IBM Corporation). A P value ≤ .05 was used as level of significance in all statistical analysis.
3.12
SDS-PAGE analysis of recombinant antibodies
First, evaluation of antibody titers for each of the three recombinant antibodies with dif-ferent signal peptides were performed. Thereafter, the reference signal peptide constructs and the HC and LC signal peptide constructs that resulted in the significantly highest antibody titers for each antibody were chosen for further control using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The evaluated HC and LC signal
H6/L1 for Antibody C were selected along with the reference signal peptide constructs for each antibody.
Due to a low antibody titer, the three Antibody C constructs were concentrated. First, 1 mL sample solution from the transient gene expression of Antibody C constructs were sterile filtered using Corning® Costar® Spin-X® centrifuge tube filter with 0.22µm pore size cellulose acetate membrane. The sample solutions were transferred to the centrifuge tube filters and centrifuged at 17,000 × g for 10 minutes at 4°C. Thereafter, the sample solutions from Antibody C constructs were transferred to Amicon® Ultra-4 Centrifugal filters (10 kDa cutoff) and centrifuged at 2,500 × g for 15 minutes at 4°C. The sample solutions from the transient gene expression for the three Antibody C constructs were concentrated from 1 mL to 250 µL.
Thereafter, the different antibody constructs were prepared for separation under re-ducing conditions. 15.6 µL sample solution was added to a PCR tube along with 6 µL NuPAGE™ LDS Sample Buffer (4X) and 2.4µL NuPAGE™ Reducing Agent (10X), result-ing in a total volume of 24 µL. The sample solutions were heated at 95°C for 5 minutes. Then, the NuPAGE™ 4−12% gradient Bis-Tris Gel was mounted in the XCell SureLock® Mini-Cell. Running buffer was prepared with 50 mL NuPAGE™ MES SDS Running Buffer (20X) and 950 mL dH2O. The upper and lower buffer chambers were filled with the
pre-pared 1X running buffer. Moreover, 500 µL NuPAGE™ Antioxidant was applied in the upper buffer chamber. Thereafter, 20 µL of each sample solution were loaded to the NuPAGE™ 4−12% gradient Bis-Tris Gel along with 10µL Novex™ Sharp Pre-stained Pro-tein Standard for size comparisons. Then, 200 V was applied for 35 minutes. The gel was removed from the XCell SureLock®Mini-Cell and submersed in InstantBlue™ Protein Gel Stain for 1 hour. The gel was imaged using LI-COR Odyssey® Fc Imaging System.
The different antibody constructs were also separated under non-reducing conditions. 15.6µL sample solution was added to a PCR tube along with 6 µL NuPAGE™ LDS Sample Buffer (4X) and 2.4 dH2O, resulting in a total volume of 24 µL. The NuPAGE™ 4−12%
gradient Bis-Tris Gel was mounted in the XCell SureLock® Mini-Cell. The prepared 1X running buffer containing NuPAGE™ MES SDS Running Buffer was used. Thereafter, 20 µL of each sample solution were loaded to the NuPAGE™ 4−12% gradient Bis-Tris Gel
along with 10µL Novex™ Sharp Pre-stained Protein Standard for size comparisons. Then, 200 V was applied for 35 minutes. The gel was removed from the XCell SureLock® Mini-Cell and submersed in InstantBlue™ Protein Gel Stain for 1 hour. The gel was imaged using LI-COR Odyssey® Fc Imaging System.
3.13
Kinetic screening of recombinant antibodies
The HC and LC signal peptide constructs that resulted in the significantly highest anti-body titer together with the reference signal peptide constructs for each antianti-body were selected for a kinetic screen. The signal peptide constructs H6/L1, H7/L1 for Antibody A, H6/L1 for Antibody B and H2/L1, H6/L1 for Antibody C were selected along with the reference signal peptide constructs for each antibody. The three recombinant anti-bodies (Antibody A, Antibody B and Antibody C) were screened against their respective antigen, referred to as Antigen A, Antigen B and Antigen C.
First, a Human Fab Capture Kit was used for immobilization at the sensor surface on a Sensor Chip CM5 according to the manufacturer’s protocol. Second, a test capture
experiment was performed where each antibody construct flowed across the sensor surface and was test captured at the Fab capture surface. Then, the antibody constructs were diluted in HBS-EP+ buffer to yield capture levels around 400 RU. Samples containing 50 nM antigen were prepared for each antibody. HBS-EP+ buffer was used as running buffer and 10 mM Glycine (pH 1.5) was used as regeneration buffer. Thereafter, the kinetic screen was performed with the following settings: three startup cycles with running buffer, 60 seconds contact time, 30 µL/minute flow rate. Antibody constructs were captured, 60 seconds contact time, 10 µL/minute flow rate. Antigen was injected, 120 seconds contact time, 300 seconds dissociation time, 30 µL/minute flow rate. Two regeneration steps, 60 seconds contact time, 30 µL/minute flow rate and ending with 300 seconds stabilization. Data from the reference flow cell was subtracted from the flow cells where the antigen binding activity was monitored.
