INOM
EXAMENSARBETE BIOTEKNIK, AVANCERAD NIVÅ, 30 HP
STOCKHOLM SVERIGE 2016,
Purification of bi- and multispecific cancer therapeutic affinity proteins and C-terminally modified anti-
HER3 affibody imaging agents
CHARLES DAHLSSON LEITAO
KTH
Contents
1 Background 2
1.1 HER-Biology . . . 2
1.2 HER-Targeted Cancer Therapeutics and Diagnostics . . . 2
2 Introduction 3 3 Materials and Methods 4 3.1 SDS-PAGE . . . 4
3.2 MALDI-TOF MS . . . 5
3.3 Analysis of Available Bispecific Constructs . . . 5
3.4 Protein Expression . . . 5
3.5 HSA Affinity Purification . . . 5
3.6 Subcloning of genes for the imaging agents and Z3-ABD-Z3-Z2 . . . 6
3.7 IEC purification . . . 6
3.8 IMAC Purification . . . 7
3.9 RP-HPLC . . . 7
3.9.1 Purification . . . 7
3.9.2 Analytical . . . 7
4 Results and Discussion 8
5 Conclusion and Outlook 11
6 Acknowledgements 11
7 Appendix 13
Abstract
Affibody molecules are small protein scaffolds that have been engineered to bind to a variety of targets with di- verse therapeutic and diagnostic applications. In this study, an array of affibody containing therapeutic constructs, targeting HER2 and HER3, and diagnostic anti-HER3 imaging agents have been purified in preparation for subse- quent cancer cell assays and imaging studies in tumour-bearing mice respectively. Herein, the workflow for several purification techniques is delineated.
1 B ACKGROUND 1.1 HER-Biology
The epidermal growth factor receptor (EGFR) family consists of the four structurally related transmembrane receptors HER1 (EGFR/ErbB1), HER2 (ErbB2), HER3 (ErbB3) and HER4 (ErbB4), each with an extracellu- lar domain associated with dimerization, and an in- tracellular tyrosine kinase domain [1]. Activity of all receptors, except HER2, is regulated by the epider- mal growth factor (EGF) family. Upon ligand bind- ing, the receptors adopt a conformation that enables homo- or heterodimerization, activating the intrinsic kinase domain resulting in trans- or autophosphory- lation of tyrosine residues situated on the kinase do- main. The residues initiate signalling cascades of var- ious biochemical pathways collectively culminating in cellular proliferation, survival and migration, including Ras/Raf/MAPK, PI3K/Akt and PLCγ/PKC, by serv- ing as docking and concomitant activation sites for cy- toplasmic signal transducers that mediate downstream transforming effects [2].
HER2 is constitutively prone to heterodimerize with the other HER receptors without ligand-induced con- formational change [3]. In addition to this, it pos- sesses the strongest catalytic kinase activity and is the preferred heterodimerization partner for all other HER family members [4]. As a result of this, genetic aberra- tions and amplification of this receptor strongly poten- tiate downstream signalling and have been associated with various cancers and poor prognosis [5, 6]. Con- sequently, HER2 has been immensely scrutinized as a potential therapeutic and diagnostic target.
HER3 has a functionally impaired tyrosine kinase domain, unlike the other HER family members, but adopts a dimerization enabled conformation upon stim- ulation by its natural ligands, neuregulin I (heregulin) and neuregulin II, forming an active heterodimeric sig- nalling unit with any of the other receptors [1]. HER3 forms together with HER2 the most potent oncogenic unit in the HER family of receptors, pertaining to ac- tivation of the transforming PI3K/Akt signalling path- way mediated by phosphorylated HER3 [5, 7]. More- over, HER3 has been implicated in the resistance to HER2-targeted tyrosine kinase inhibitor (TKI) therapy [8]. Additionally, treatments targeting the extracellular domain of HER2, such as trastuzumab, might be nul- lified through the bypass activation of PI3K/Akt via
EGFR mediated HER3 phosphorylation [9]. In fact, HER3 is pivotal in the activation of the PI3K/Akt path- way and attempts at shutting down this signalling axis by HER3-ligand inhibition could prove effective [10].
Overexpression of HER3 in tumours has also been asso- ciated with a shorter survival time of patients compared to overexpression of HER2 [11].
1.2 HER-Targeted Cancer Therapeutics and Diagnostics
A plethora of HER-family targeted therapeutics for a wide range of malignancies, that are the product of HER receptor aberrations, have been developed in re- cent years [12, 13]. In most cases, the drugs, ranging from small molecules to antibodies, inhibit or attenu- ate the activity of the receptor or the downstream sig- nalling, for instance, by inhibiting the intracellular ty- rosine kinase domain, preventing receptor dimerization or by blocking ligand binding [12]. For example, am- plification of HER-receptors and mutation-driven con- stitutive autophosphorylation are distinct mechanisms for conveying an abundance of pro-survival signals and therefore require different treatment approaches.
Complex diseases, such as cancer, are multifaceted with intricate and overlapping signalling networks. A consequence of this is the resistance to treatment by the compensatory upregulation and activation of al- ternative signalling pathways, or by intrinsic alterna- tions of receptors, which poses a problem for targeted treatments [14]. Combination therapies can enhance, through synergistic effects, the treatment compared to conventional single receptor targeted therapies, and possibly circumvent resistance, by blocking the signal of e.g. both HER2 and HER3 in HER2 amplified breast cancer [7].
Alternatively, these properties can be combined into a single therapeutic agent by utilizing a bi- or mul- tispecific format, with the potential to bind two or more targets simultaneously. This concept has been in- vestigated with bifunctional antibodies, combining in- hibitory functions of HER1/HER3 [15], HER1/HER2 [16], and HER2/HER3 [17], surpassing the inhibitory function of monospecific antibodies, and a myriad of other bispecific antibodies used for both therapy and di- agnostics [18].
A previous study has combined engineered vari- ants of the small non-immunoglobulin based disulfide- independent affibody scaffold into a bispecific format
targeting HER2 and HER3 [19]. The smaller size of this alternative scaffold has advantages over antibodies in terms of superior tissue penetration and less complex production with reduced cost. However, the drawback is the rapid plasma clearance due to its small size, this problem was addressed in the study by introducing an albumin binding domain (ABD) to retain the protein in blood by allowing it to associate with circulating human serum albumin (HSA). The concept of affibodies as ther- apeutic and diagnostic agents is the main focus of this study and will be discussed in greater detail.
Combining anti-HER2 and anti-HER3 therapeutics in a bispecific format would be interesting to evalu- ate potential superior anti-oncogenic effects and ther- apy resistance circumvention, for reasons previously discussed. Furthermore, this format could facilitate the localization to tumours for treatments targeting membrane-bound proteins that are expressed to a lesser extent, e.g. amplified HER2 on tumours can localize a bispecific protein targeting HER2 and HER3, where- upon the anti-HER3 functionality can exert its therapeu- tic function on the less abundant HER3 and vice versa.
These targeted treatments necessitate, preferentially rapid and non-invasive, anti-HER diagnostic agents with high resolution for phenotypic determination and monitoring to allow for patient stratification [20]. This can be achieved by exploiting the small size of the afore- mentioned affibody scaffold, which takes advantage of the fast plasma clearance to improve the resolution and rate of the tumour imaging. By labelling the affibody, targeting a particular receptor, with a detectable ele- ment, tumours expressing that receptor can be visual- ized. This has been done with anti-HER2 affibody vari- ants labelled with technetium-99m [21]. Ultimately, this concept could be developed even further into a therag- nostic format.
2 I NTRODUCTION
Affibodies are small (58 amino acids), stable and versa- tile three-helical bundle proteins with many biotechno- logical and therapeutic applications [22]. The affibody scaffold is based on the so called Z-domain, which is de- rived from the membrane-bound protein A, originally found in Staphylococcus aureus. It is used by the bacte- rial cell to evade the immune system during infection by binding to the heavy-chain Fc region of IgG, thus nullifying its effector function. ABD-Derived Affinity
Protein (ADAPT) is another small three-helical protein, derived from the albumin binding domain of strepto- coccal protein G, with properties similar to affibodies [23]. The intrinsic affinity for albumin increases the re- tention of the scaffold in blood, which is an advantage when considering therapeutic applications as the pro- longed effective concentration reduces the dose required and the number of drug applications, which diminishes discomfort and economical strain for the patients [24].
Reversely, for diagnostic purposes, the rapid clearance of unbound proteins is preferred [25] to achieve high tumour-to-organ ratios with examination of the results shortly after injection, in which case the albumin bind- ing propensity of ADAPT is undesirable. The inclusion of ADAPT, binding to both albumin and an additional target simultaneously, in a bispecific construct abolishes the need for additional domains, such as ABD, to extend the in vivo half-life, while still retaining the properties associated with small molecules.
The first part of this study is concerned with an ar- ray of bivalent/bispecific and multispecific constructs.
The bivalent constructs are comprised of affibody and ADAPT domains targeting HER2 that are intercon- nected by varying repeats (1-3) of a glycine and serine containing linker (G4S). The future prospect is to evalu- ate the effect of the linker-length on cancer cell growth inhibition by a bivalent construct targeting the dimer- ization domain of HER2. Additionally, bivalent and bis- pecfic constructs targeting HER2 and HER3 and multi- specific constructs targeting a range of different cancer- associated cell surface proteins are also included in this study, for a complete list, see table 1. The affibodies have in previous studies been selected by various dis- play techniques from a combinatorial protein library by randomization of 13 solvent-exposed residues on helices one and two, to bind to HER2 [26, 27] and HER3 [28].
ADAPT molecules with affinities for HSA as well as an additional binding affinity for HER2 [29] and HER3 [30]
have previously been generated. A list of the affibody and ADAPT variants used in this study is presented in table 2. Both the anti-HER3 affibody and ADAPT variants compete for binding to HER3 with the nat- ural ligand NRG-β1, suggesting therapeutic potential.
In fact, the anti-HER3 affibody variant has previously demonstrated anti-proliferative effects on breast cancer cell lines [31]. The epitopes recognized by the anti-HER2 affibody ADAPT variants differ. Notably, the ADAPT variant competes with trastuzumab for binding.
The second part of this study is concerned with
Table 1: The proteins involved in the study and their respective theoretical masses, based on the isotopically averaged molecular weights. The isoelectric point (pI), based on pKa of individual amino acids, is included for its relevance in deciding on purification strategy for the imaging agents. The domains of the octamer and hexamer are interconnected by a G4S linker and ZVEGFR2(a) and ZVEGFR2(b) are affibodies against two different epitopes of VEGFR2. The affibody and ADAPT variants used are presented in table 2.
Bivalent/Bispecific Multispecific Imaging agents
Construct Mw (Da) Construct Mw (Da) Construct Mw (Da) C-terminus Mw (Da) pI A2-(G4S)1-Z2 12809 A3-(G4S)1-Z3 12671 Octamer1 52706 -GGEC 6740.5 5.31 A2-(G4S)2-Z2 13124 Z3-(G4S)1-A3 12671 Hexamer2 38746 -GGGC 6668.4 6.62 A2-(G4S)3-Z2 13439 A3-(G4S)1-Z2 12731 Z3-ABD-Z3-Z2 25953 -GGSC 6698.4 6.62
Z2-(G4S)1-A2 12809 Z2-(G4S)1-A3 12731 -GGKC 6739.5 8.12
Z2-(G4S)2-A2 13124 -QAPKC (WT) 6793.6 8.12
Z2-(G4S)3-A2 13439 -QAPK (H6-WT) 7644.5 8.41
1A3-ZEGFR-Z3-ZVEGFR2(a)-ZVEGFR2(b)-A2-ZIGF1R-Z2-Cys
2A3-ZEGFR-ZVEGFR2(a)-ZVEGFR(b)-A2-ZIGF1R-Cys
the production and purification of four C-terminally modified derivatives of the previously generated ZHER3:08698 variant [32] as well as both the wild-type (denoted WT) and His6-tagged WT (denoted H6-WT), listed in table 1. The C-terminal modifications are de- signed to function as peptide chelators to allow for coor- dination of a radionuclide. The C-terminal variants are of the kind -GGXC, where X is either Lysine (K), Glycine (G), Serine (S) or Glutamic acid (E). Each variant will be employed in a different study conducted in mice to in- vestigate their potential use as diagnostic tumour imag- ing agents. This has already been done using anti-HER2 affibodies, with the same array of C-terminal modifica- tions [21]. For this purpose, a minimum of 1 mg of each C-terminally modified imaging agent, and 3 mg of both WT and H6-WT, must be produced and purified with a final purity of at least 95%.
Table 2: The affibody (Z) and ADAPT (A) vari- ants used in this study.
Abbr. Full name Ref.
Z2 ZHER2:02891 [27]
Z3 ZHER3:05417 [28]
A2 ABDErbB2-1 [29]
A3 ABDErbB3-3 [30]
3 M ATERIALS AND M ETHODS 3.1 SDS-PAGE
The amount of protein on SDS-PAGE gels throughout the study was, whenever possible, approximated to be 1-5 µg for optimal visualization. The gels used were either Mini-PROTEAN®TGX gels (BioRad Systems) or NuPAGE® Bis-Tris gels (ThermoFisher Scientific), and the following protocol was always employed. A total volume of 20 µl containing reduction and loading buffer, appropriate amount of protein (1-5 µg) and Milli-Q wa- ter was applied to each well. The samples were heated to near boiling temperature (95°C) for 5 min to increase the rate of reduction. For the Mini-PROTEAN® TGX gels, 1X TGS (Tris, Glycine, SDS) was used as running buffer (BioRad Systems), and for the NuPAGE®Bis-Tris gels, 1X MES SDS (MES, Tris, SDS, EDTA) was used (ThermoFisher Scientific). The electrophoresis program used was 140 V, 400 mA, 45 min for the TGX gels, and 200 V, 400 mA, 30 min for the Bis-Tris gels. Standard al- bumin ladder (BioRad) was always used as a molecular weight reference (97, 66, 45, 30, 20.1, 14.4 kDa).
3.2 MALDI-TOF MS
A 4800 MALDI TOF/TOF™ Mass Analyser (Applied Biosystems) was systematically used to verify the pres- ence of target proteins throughout the purification pro- cess. Additionally, it was used as a preliminary ver- ification of mass for the bispecific constructs and the imaging agents. The matrix used was α-Cyano-4- hydroxycinnamic acid (Bruker Daltonics), which is suit- able for low to mid-range protein masses.
3.3 Analysis of Available Bispecific Con- structs
The concentration of the constructs was determined us- ing a bicinchoninic acid assay (BCA assay) with tripli- cates made for each. SDS-PAGE was used to determine if any cleaved products had been accumulated over the storage period. MALDI-TOF MS was used to verify the mass of the specific constructs. From the BCA assay, it was concluded that the amounts of five constructs (Z3- A3, Z2-A3, Z2-3-A2, Hexamer and Octamer) were insuf- ficient for cell assays, thus more had to be produced.
3.4 Protein Expression
Available BL21(DE3)* expression strain E. coli harbour- ing isolated pET26b(+) plasmids containing the genes for Z2-3-A2, Hexamer and Octamer and TOP10 E. coli harbouring pET26b(+) for Z3-A3 and Z2-A3 were in- oculated into 10 ml TSB+Y medium with 50 µg/ml kanamycin (KM) and incubated at 37°C overnight.
The plasmids produced from TOP10 cells were ex- tracted using QIAprep Spin Miniprep Kit (Qiagen), ac- cording to protocol, and the concentrations were mea- sured to be 96.1 ng/µl for Z3-A3 and 50.6 ng/µl for Z2- A3, using a NanoDrop 1000 spectrophotometer (Ther- moFisher Scientific). BL21(DE3)* cells were transformed with the extracted plasmids using heat shock treatment.
5 µl of plasmid was added to 1 ml BL21(DE3)* cells and incubated on ice for 20 min. The cells were heat-shocked at 42°C for 45 sec and returned to incubation on ice for 5 min. 400 µl TSB was added to each tube and incu- bated at 37°C in a rotamixer for 60 min. 100 µl and 400 µl from each tube were transferred to TBAB plates containing KM and incubated at 37°C overnight. Non- transformed heat-treated cells was included as a nega- tive control. 5 ml TSB+Y+KM cultures were inoculated with single colonies and incubated at 37°C overnight.
5 ml from the five starter cultures were inoculated into 500 ml TSB+Y+KM and incubated at 37°C until an OD600 of ~0.8 was reached, by which time pro- tein expression was induced using isopropyl β-D-1- thiogalactopyranoside (IPTG), from Apollo Scientific, for a final concentration of 1 mM, followed by cultiva- tion overnight at 24°C. The cells were subsequently har- vested by centrifugation at 5000xg and 4°C for 10 min and the pellets frozen at -20°C.
3.5 HSA Affinity Purification
The albumin affinity of ADAPT and ABD enabled affin- ity purification of the construct using HSA as ligand.
The pellets were resuspended in 40 ml TST buffer (pH 8) and the cells were mechanically lysed using french press (Baseclear) three times for each sample. The lysates were centrifuged at 17,000xg and 4°C for 10 min and the supernatant filtered (0.45 µm). The columns were prepared using HSA-coupled agarose matrix (10 ml column volume) and pulsed with 30 ml Milli-Q, 30 ml 0.5 M HAc (pH 2.8) and 60 ml TST. The columns were equilibrated with 75 ml TST followed by the ad- dition of the protein sample (approximately 40 ml). The column was washed with 75 ml TST followed by 50 ml 5 mM NH4Ac (pH 5.5) and the target proteins were eluted by 15 ml HAc into 1 ml fractions. Before storage, the columns were pulsed with 30 ml HAc, 30 ml NH4Ac, 60 ml TST and filled with 20% ethanol. The absorbance (280 nm) of the eluate fractions was measured and the protein containing fractions were pooled.
Z2-3-A2 was lyophilized but exhibited extensive ag- gregation when resuspended in Phosphate Buffered Saline (PBS). Therefore, instead of freeze-drying, the buffer was changed to PBS buffer using disposable PD- 10 columns (GE Healthcare) for the other constructs (Octamer, Hexamer, Z2-A3, Z3-A3). Fractions of 2.5 ml protein solution was applied to the PD-10 column and eluted with 3.5 ml PBS. The constructs were sub- sequently aliquoted and frozen at -20°C. Attempts at solubilizing the aggregated Z2-3-A2 included adding L- Arginine to a concentration of 50 mM and increasing the volume. The residual aggregates were spun down and the clear supernatant was aliquoted and frozen.
3.6 Subcloning of genes for the imaging agents and Z3-ABD-Z3-Z2
Production of the anti-HER3 imaging agents was ini- tially attempted using pUC57 expression vectors. How- ever, low to no protein expression was observed us- ing an E. coli BL21(DE3)* strain for any of the imaging agents. Previously successful production and purifica- tion of bi-specific constructs using a pET26b(+) expres- sion vector rationalized the subcloning of the affibody genes from pUC57 to pET26b(+).
Reverse primers, annealing to the 30 end, were de- signed for each protein gene and ordered (Integrated DNA Technologies). The same forward primer was used for all except for H6-WT and -GGEC. The primers comprised specific overlap sequences, NdeI and XhoI restriction sites on the forward and reverse primer re- spectively and an overhang with a length of six nu- cleotides, see table 5 in appendix.
The gene inserts for the imaging agents and Z3-ABD- Z3-Z2 on pUC57 were amplified by touchdown PCR us- ing the designed primers and Phusion® High-Fidelity DNA Polymerase (ThermoFisher Scientific). Additional PCR products could be observed from the amplification of the Z3-ABD-Z3-Z2 insert. The correct Z3-ABD-Z3-Z2 amplicon with the size of 700 bp was isolated by excis- ing a preparative 1% agarose gel followed by purifica- tion using QIAquick Gel Extraction Kit (Qiagen)
The amplified inserts were purified using QIAquick PCR Purification Kit (Qiagen), according to protocol, and digested overnight at 37°C using XhoI and NdeI re- striction enzymes (New England Biolabs), both lacking star activity. The concentrations were measured with NanoDrop. Digestion protocol for 1 µg of DNA was ob- tained from NEBCloner [33]. Following digestion, the inserts were again purified to remove the overhang se- quences.
The pET26b(+) vector backbone was generated by digestion using XhoI and NdeI. The digested backbone was separated from circular plasmid and insert by elec- trophoresis on a DNA recovery 1% agarose gel at 140 V for 30 min. The digested vector band was excised from the gel and the backbone was recovered using QIAquick Gel Extraction Kit, according to protocol, and the con- centration measured.
Ligation was performed using 50 ng of vector back- bone, 15x molar excess of insert (30 ng for the imaging agents and 87.5 ng for Z3-ABD-Z3-Z2), T4 DNA ligase, 5 µl of T4 DNA ligase buffer (New England Biolabs) in a
total reaction volume of 50 µl. The reaction was allowed to proceed for 2 hours. The vectors containing each in- sert were sent for sequencing using a promoter specific forward primer (T7 promoter) and the insert specific re- verse primer.
KCM-competent TOP10 cells were transformed, ac- cording to the aforementioned protocol, using 8 µl of lig- ated plasmid for each construct and 2 µl of KCM solu- tion. Five colonies for each imaging agent and Z3-ABD- Z3-Z2 were picked, except for ZHER3-WT, for which only four colonies were available. Each colony was dipped in 20 µl of Milli-Q and then transferred to a new TBAB KM plate.
Colony PCR was performed on the in-solution colonies using DyNAzyme II DNA polymerase (Ther- moFisher Scientific) and the respective primers. The TOP10 colonies verified to harbour plasmids with in- serts of correct size were cultured in 10 ml TSB+Y+KM.
Subsequently, the plasmids were extracted using QI- Aprep Spin Miniprep Kit (Qiagen), according to pro- tocol. Additionally, freeze-stocks containing 200 µl of each culture mixed with 800 µl glycerol were made and stored at -80°C.
The extracted plasmids were used to transform BL21(DE3)* cells and the obtained colonies were cul- tured and induced to produce protein, as previously de- scribed. As with the TOP10 cells, a freeze-stock was made for the BL21(DE3)* cells. Following overnight production, the cells were harvested and the expression analysed with SDS-PAGE on intact cells. All imaging agents and Z3-ABD-Z3-Z2 were overexpressed, except for -GGKC. For this reason, the five -GGKC colonies obtained from the transformation underwent the same preparatory process for protein production as previ- ously described, to determine if any exhibited suffi- cient expression of the target protein. All -GGKC colonies harboured the correct insert as established by colony PCR. Furthermore, the plasmid from each -GGKC colony was sent for sequencing. The presence of insoluble -GGKC contained in inclusion bodies was investigated, by separating and comparing the soluble and insoluble proteins on SDS-PAGE. The -GGKC gene on pET26b(+) was ultimately ordered (BioBasic).
3.7 IEC purification
The first step purification of the non-tagged imaging agents was performed using ion exchange chromatogra- phy (IEC) on ÄKTAexplorer (GE Healthcare). The type
of IEC was decided based on the pI, listed in table 1.
ZHER3-WT and -GGKC have a pI of 8.12, thus cation ex- change was chosen using 20 mM MES (pH 5.5) as wash buffer, and 20 mM MES + 1 M NaCl (pH 5.5) as elution buffer. For the imaging agents with a pI of 5.31 and 6.62, anion exchange was chosen using 20 mM Tris (pH 8.6) as wash buffer and 20 mM, 1 M NaCl (pH 8.6) as elution buffer. All buffers were filtered (0.45 µm) and degassed before use.
A general procedure before IEC purification was em- ployed for all imaging agents. The frozen pellet was re- suspended in 10 ml of the appropriate wash buffer. The cells were lysed by sonication. The lysates were cen- trifuged at 10,000xg and 4°C for 20 min. The supernatant was heat-treated at 90°C for 10 min, followed by instant incubation on ice for 20 min. An additional 10 ml of wash buffer was added before spinning down precipi- tated proteins at 10,000xg and 4°C for 20 min. Lastly, the supernatant was filtered (0.45 µm).
The columns used for the cation and anion exchange chromatography were Resource S (sulphonate group as ligand) and Resource Q (quaternary ammonium as lig- and) respectively (GE Healthcare). The flow-rate was 4 ml/min and the linear elution buffer gradient went from 0-50%, however, these parameters were in some cases optimized for maximized peak separation based on pilot runs for each protein. Eluate fractions and flow-through were analysed with MS and SDS-PAGE to determine the peaks corresponding to the target pro- tein. Large amounts of protein were observed in the flow-through from the purification of WT, -GGGC, and -GGSC, possibly due to an overload of the column, hence a second round of purification was undertaken to salvage as much protein as possible.
The correct fractions were pooled and the buffer changed to NH4Ac (pH 7) using PD-10 columns fol- lowed by lyophilization. 500 µl from each imaging agent was taken and freeze-dried separately with the purpose of checking for aggregation after resuspension and to be used in pilot RP-HPLC runs.
3.8 IMAC Purification
The lysate preparation was identical to the other imag- ing agents, described above. The IMAC column was packed with ~3 ml TALON Metal Affinity Resin (Clon- tech) and pulsed with 30 ml Milli-Q followed by equi- libration with 40 ml wash buffer (50 mM Na2HPO4, 500 mM NaCl, 15 mM imidazole, pH 8). The lysate sam-
ple (~20 ml) was applied to the column and impurities were removed with the wash buffer. A buffer containing 50 mM Na2HPO4, 500 mM NaCl, 500 mM imidazole, pH 8 was used to elute the proteins. The protein containing fractions were pooled and the buffer changed to NH4Ac (pH 7) with subsequent lyophilization.
3.9 RP-HPLC 3.9.1 Purification
ZHER3-WT (+FT), -GGEC, and -GGGC (+FT) were puri- fied with reversed-phase high-performance liquid chro- matography (RP-HPLC) on Agilent 1200 HPLC Systems (Agilent Technologies) using 0.1% trifluoroacetic acid (TFA) in water for washing and a 20-40% linear gradi- ent of 0.1% TFA in acetonitrile (ACN) for eluting. The flow-rate was 3 ml/min for 30 min, with some opti- mization based on pilot chromatograms. The column used was Zorbax 300SB-C18 Semi-prep, 9.4x250 mm, 5 µm particle size (Agilent Technologies). The lyophilized proteins were resuspended in 20% ACN. Two distinc- tive peaks were obtained that, according to MS analy- sis, corresponded to the same protein (Fig. 9a). This was assumed to be the result of dimerization via the C-terminal cysteine. The peaks putatively correspond- ing to monomers and dimers were pooled separately.
With this knowledge, -GGSC was reduced with 20 mM Dithiothreitol (DTT), from Saveen & Werner, before pu- rification to obviate dimer formation. The monomer and dimer eluate fractions were freeze-dried and resus- pended in NH4Ac (pH 7.2). H6-WT was purified with- out any prior reduction since it lacks a C-terminal cys- teine.
3.9.2 Analytical
To evaluate the purity of each imaging agent, analytical RP-HPLC was performed with a Zorbax 300SB-C18 An- alytical column, 4.6x150 mm, 3.5 µm particle size (Ag- ilent Technologies) using a 20-40% linear ACN + 0.1%
TFA buffer gradient and a flow-rate of 0.9 ml/min for 30 min. 10 µg from both the monomer and dimer frac- tions were pooled and reduced with 20 mM DTT for 30 min at 37°C. Following reduction, an equal amount of 40% ACN + 0.1% TFA was added for a final concen- tration of 20%. The sample was filtered (0.45 µm) be- fore injection. This procedure noticeably decreased the presumed dimer peak, but not completely, strongly sug-
gesting the presence of dimers (Fig. 9b). In an attempt to entirely reduce the proteins, thus producing a soli- tary peak, a stronger reducing agent was used. Reduc- tion using 5 mM Tris(2-carboxyethyl)phosphine (TCEP), from Sigma-Aldrich, for 20 min was performed in an otherwise identical procedure prior to analytical RP- HPLC. Complete reduction using TCEP was achieved for all imaging agents (Fig. 9c) with the exception of -GGEC, which still exhibited a minor presence of dimers.
The monomer and dimer fractions of each con- cerned imaging agent were pooled, reduced with 20 mM TCEP for 30 min, and the buffer changed to NH4Ac (pH 5.5) to prevent redimerization and to remove the reducing agent. The absorbance (A280) of all imaging agents was measured and the proteins were distributed in 100 µg aliquots, freeze-dried and stored in -20°C.
4 R ESULTS AND D ISCUSSION
The concentration and mass measurements of the pre- viously produced constructs as well as the amounts and mass identity of the constructs produced and purified in this study are summarized in table 3. The structural in- tegrity of the previously produced constructs was anal- ysed with SDS-PAGE (Fig. 2a). SDS-PAGE was also used to evaluate the purity level of the newly produced constructs (Fig. 1 and 2b), as no HPLC-based purity analysis was performed.
Figure 1: Purification of Z3-ABD-Z3-Z2. (1) Standard albumin ladder (2) cell lysate (3) flow-through (4) TST wash (5) NH4Ac wash (6) pooled eluate.
Concentration analysis of the available constructs, using both spectrophotometric measurement and BCA, revealed a tenfold lower concentration than what had previously been reported (table 3), requiring additional
protein to be produced and purified for subsequent cell assays. SDS-PAGE did not reveal any cleaved products (Fig. 2a), however, the approximate masses of the pro- teins on the gel were not in concordance with the the- oretical masses (table 1), with bands slightly above the 14.4 kDa mark. This was not corroborated by the MS re- sults (table 3), which indicated an actual mass closer to the theoretical, with some deviations. These deviations could possibly be due to the relatively poor precision of the instrument, with significant differences in measured mass observed between runs (data not shown). How- ever, this seems to be more prominent for larger pro- teins, as the newly produced imaging agents exhibited differences within the range isotopic distribution whilst the newly produced Z3-ABD-Z3-Z2 exhibited a consid- erable mass difference of -37.3 Da, and also significantly lower peak signal (spectrum not shown), see tables 4 and 3. This was especially noticeable for the octamer, which exhibited close to no signal and a significant mass difference of -256 and -182 Da for the previously and newly construct respectively. Potential improvement in this regard could be to use another more suitable matrix to facilitate desorption of these large proteins. A plausi- ble explanation for some of these mass deviations could be the oxidation of methionine to methonine sulphox- ide and cysteine to cysteine sulfenic acid, which read- ily occurs at the presence of atmospheric oxygen when stored at 5°C [34]. This would make sense for the Z2- containing constructs, since only Z2 contains a methion- ine residue, which would explain the mass deviations for A2-1-Z2 and possibly Z2-3-A2. A3Z3 and Z3A3, with negligible deviations, do not contain methionine, which supports this theory. A more precise mass deter- mination would be achieved with ESI-MS, which was desired, but unfortunately the instrument was unavail- able for use.
Sequencing established that the cloned vectors har- boured the correct gene sequences for the imaging agents and Z3-ABD-Z3-Z2, except for some discrep- ancies observed in the sequences for WT and H6-WT.
However, these could be dismissed after inspection of the sequencing chromatograms (Fig. 6 in appendix) and MS verification, which corroborated the correct se- quences. The sequencing of the plasmids corresponding to all five available -GGKC colonies were inconclusive.
Table 3: Summary of the analysis on the bi- and multispecific constructs. (1) Available amounts (mg) based on absorbances ob- tained from BCA. (2) Available amounts (mg) based on absorbances obtained from single spectrophotometric measurements. (3) Amounts (mg) produced and purified in this study. (4) Mass differences (Da) between theoretical and MALDI-TOF MS measure- ments, see theoretical masses in table 1. Mass differences of the newly produced constructs are in parenthesis.
A2-1-Z2 A2-2-Z2 A2-3-Z2 Z2-1-A2 Z2-2-A2 Z2-3-A2 A3-Z2 Z2-A3 Z3-A3 A3-Z3 Octamer Hexamer 3-A-3-2
(1) 0.47 0.55 0.57 0.61 0.38 0.04 2.58 - 0.12 0.53 0.32 0.11 -
(2) 0.55 0.25 0.28 0.49 0.38 0.08 2.17 - 0.07 0.34 0.12 0.3 -
(3) - - - - - 10.23 - 2.63 9.74 - 5.93 19.91 23.84
(4) 17.5 23.2 7.2 -5.5 -3.6 24.8 (-12) 0.76 9.9 (6.8) -0.04 (-2.7) 1.5 -256 (-182) -91.5 (-13.3) -37.3
(a) (b)
Figure 2: (a) SDS-PAGE on the previously produced and biotinylated (451.54 Da mass difference from those listed in table 1) bi- and multispecific constructs, (1) standard Albumin ladder (97, 66, 45, 30, 20.1, 14.4 kDa), (2) A2-1-Z2, (3) A2-2-Z2, (4) A2-3-Z2, (5) Z2-1-A2, (6) Z2-2-A2, (7) Z2-3-A2, (8) A3-Z2, (9) A3-Z3, (10) Z3-A3, (11) Hexamer, (12) Octamer. (b) New production of non- biotinylated (451.54 Da mass difference) bi- and multi-specific constructs (1) Standard albumin ladder (2) Octamer (3) Hexamer (4) Z2-3-A2, (5) Z2-A3 (6) Z3-A3.
Figure 3: SDS-PAGE page on intact cells to analyse protein expression. The overexpressed bands correspond to affibody dimers. (1) Standard albumin ladder, (2) WT, (3) H6-WT, (4) -GGGC, (5) -GGEC, (6) -GGKC, (7) -GGSC, (8) Z3-ABD-Z3-Z2.
The genes for the imaging agents and Z3-ABD-Z3- Z2 were successfully cloned into pET26b(+) and exhib- ited overexpression of the correct proteins following cul- tivation and induction using IPTG, except for -GGKC (Fig. 3). This was perplexing since all kanamycin se-
lected colonies obtained from transformation of the lig- ated vector insert harboured circular plasmids with the correct insert size, based on results from colony PCR.
None of these five colonies overexpressed -GGKC (Fig.
4), nor did they contain any -GGKC trapped in inclu- sion bodies (data not shown). The gene in pET26b(+) was instead acquired from BioBasic, which successfully expressed when induced in BL21(DE3)* cells.
A summary of the protein amounts obtained after IEC and RP-HPLC purification as well as the purity level of each imaging agent and MALDI-TOF MS deter- mined masses are presented in table 4. The amount of the -GGGC obtained after HPLC was comparably low (0.153 mg), entailed by the low amounts obtained fol- lowing AIEX purification which could be a consequence of the ambiguous chromatogram (Fig. 7c in appendix) and low expression, indicated somewhat by a relatively faint band on SDS-PAGE (Fig. 3). A second batch of -GGGC was produced and instead purified with CIEX
Table 4: Summary of the purification results for the imaging agents concerning the amounts obtained after IEC and RP-HPLC purification, the A280purity level revealed by analytical HPLC, and the difference in mass from MALDI-TOF MS measurements compared to the the theoretical values presented in table 1. The values in parenthesis are the amounts obtained from the purifi- cation of flow-through from the first batch. The results from the second batch is presented on the second row for each concerned conjugate.
Conjugate Purified using
Amount (mg)
Amount after HPLC (mg)
A280
Purity (%) ∆Mw (Da)
CIEX 4.88 (5.43) 1.1 95 -3.2
WT CIEX 5.45 3.2 100 -1.5
AIEX 0.99 (2.35) 0.15 - -
-GGGC
CIEX 1.4 1.1 97.2 -1
AIEX 9.8 0.85 95 -2.5
-GGEC
CIEX 1.39 1.04 95.5 -2.6
-GGSC AIEX 1.36 (4.44) 1.37 100 -1.6
-GGKC CIEX 2.82 1.94 100 -1
WT-His6 IMAC 9.9 5.45 98.1 -4.7
following reduction using TCEP, resulting in a solitary peak with a 280 nm purity level of higher than 95% (Fig.
8c in appendix), revealed by subsequent HPLC purifica- tion. Although the expression was still low (1.4 mg after CIEX), a satisfactory final amount of 1.1 mg was avail- able following HPLC purification. A cation exchanger was also used for a second batch of -GGEC (pI 5.31) with similar results as -GGGC (Fig. 8b in appendix). No- tably, a low pH buffer was required for CIEX purifica- tion of -GGEC (MES pH 4.6), but owing to the stability of affibodies at low pH, this approach could prove to be more effective. The amount of WT obtained from the first batch was insufficient as well (1.1 mg) with the re- quirement of 3 mg. A second production batch with the same purification strategy as the first, with the exception of TCEP reduction, resulted in 3.2 mg (table 4). The total amount of protein acquired after both batches exceeded the requirement of 1 mg for the C-terminally modified imaging agents, and 3 mg for WT and H6-WT, with at least 95% purity determined by analytical HPLC.
The salient differences between the two types of ion exchangers, apparent from the AIEX chromatograms in figure 7 and the CIEX chromatograms in figure 8, accen- tuates the advantage of utilizing a cation exchanger for purification, at least when concerned with the affibody molecules studied here, and presumably also other vari- ants. The peak corresponding to the overexpressed pro- tein is easily discerned and protein identity determina- tion by SDS-PAGE or MS analysis can in these cases
often be omitted, additionally allowing for immediate pooling of primary eluate fractions and potential puri- fied flow-through.
Figure 4: SDS-PAGE on intact cells showing the lysate from five cultivations corresponding to five different colonies ob- tained from transformation after subcloning of the -GGKC in- sert. No expression of the target protein can be seen.
The protein retention could vary between CIEX and AIEX for each conjugate, one possibly eluting more in the flow-through or during the column wash than the other, however, any protein collected in the flow- through can easily be reapplied and salvaged in an ad- ditional purification run. Moreover, the amount of pro- tein remaining in the column after gradient elution was low for CIEX compared to AIEX, which can be seen in figures 7 and 8.
A significant portion of the protein amount from IEC purification was obtained from purifying the flow- through in the case of AIEX (table 4). The high retention
of undesired proteins and nucleic acids, not prominent for CIEX, presumably saturates the binding capacity of the column, resulting in low retention of the target pro- tein.
Plausibly, the presence of negatively charged nu- cleotides associates with the more clustered chro- matogram observed from anion chromatography, as op- posed to the more unambiguous cation exchange chro- matogram. Additionally, there are generally more acidic than alkaline proteins produced by E. coli strains [36], hence fewer proteins are subject to retention by a cation exchanger. It might be good practise to assume a sub- stantial loss of target protein in the flow-through and thus always reapply for one or two additional purifica- tion runs.
The rationale behind heat-treatment lies in the propensity of affibodies to effectively and rapidly re- fold following denaturation. In fact, its three helical bundle structure endows it with one of the fastest fold- ing kinetics currently known [35]. The instant and dra- matic temperature change will cause bulkier proteins to aggregate and precipitate. Meanwhile, the affibody manages to re-fold and remain soluble. It is highly ef- fective, removing a majority of the proteins present in the bacterial lysate (Fig. 5). This level of purity might be sufficient to omit ion exchange purification, and poten- tially achieve higher than 95% purity solely using HPLC purification, thus increasing the protein yield.
Figure 5: Effectiveness of heat-treatment. A majority of the lysate proteins (lane 2) are affected and can be removed (lane 3). A small portion of proteins, includ- ing the affibody (strong band), remain soluble. The affibody shown here is H6- WT
The imaging agents produced and purified in this study were meant to coordinate a radionuclide using the chelating properties of the C-terminal residues, how- ever, initial reports based on the first production batch, suggest an intrinsic inability of ZHER3 to properly do so. This is surprising considering the success of a pre- vious imaging study using chelating anti-HER2 affibod- ies with the same C-terminal modifications. Instead, the imaging agents will be coupled to an external chelator
via biotin-maleimide functionalization. The C-terminal sequence is known to affect the labelling efficacy and function of the chelator and by extension the biodistri- bution. Even though the concept of a peptide chela- tor was unsuccessful in the case of the anti-HER3 affi- body variant, which is useful information in and of it- self, the great potential for imaging agents based on af- fibody molecules warrants great excitement for future anti-HER3 diagnostics regardless of the labelling ap- proach.
5 C ONCLUSION AND O UTLOOK
The purification workflow for the imaging agents was optimized throughout the study, with the most notewor- thy adjustments being the use of a cation over an anion exchanger, whenever possible, and the use of a strong reducing agent (TCEP) prior to every purification step.
Additionally, the required amount and purity were at- tained for all of the imaging agents.
Unfortunately, due to time limitations, the investi- gation of the bi- and multispecific constructs could not be finalized and was limited to concentration measure- ments, mass identity determination and stock replenish- ment by production and purification of those in scarce amounts. No kinetics or secondary structure analysis was performed on the newly produced constructs. Ac- quisition of new such data on the old as well as the new constructs might be warranted due to the extent of the storage period, before commencing cell assays.
Although the potential use of the C-terminal ends of anti-HER3 affibody variants as peptide chelators was abrogated, exciting results await with regard to the newly conceived approach for anti-HER3 affibody imaging agents.
6 A CKNOWLEDGEMENTS
I would like to sincerely thank my two amazing super- visors, Tarek Bass and Stefan Ståhl, who have been im- mensely helpful and considerate throughout the entire project. I would also like to thank my previous supervi- sor, John Löfblom, who presented me with this fantastic opportunity. Lastly, I would like to voice my exuberant appreciation for everyone at the department and my fel- low thesis workers who collectively created a wonderful atmosphere.
R EFERENCES
[1] Hynes, N. E. & Lane, H. a. ERBB receptors and can- cer: the complexity of targeted inhibitors. Nat. Rev.
Cancer 5, 341–54 (2005).
[2] Scaltriti, M. & Baselga, J. The epidermal growth fac- tor receptor pathway: a model for targeted therapy.
Clin. Cancer Res. 12, 5268–72 (2006).
[3] Yarden, Y. & Sliwkowski, M. X. Untangling the ErbB signalling network. Nat. Rev. Mol. Cell Biol. 2, 127–137 (2001).
[4] Graus-Porta, D., Beerli, R. R., Daly, J. M. & Hynes, N.
E. ErbB-2, the preferred heterodimerization partner of all ErbB receptors, is a mediator of lateral signal- ing. EMBO J. 16, 1647–1655 (1997).
[5] Moasser, M. M. The oncogene HER2: its signaling and transforming functions and its role in human cancer pathogenesis. Oncogene 26, 6469–87 (2007).
[6] Yarden, Y. Biology of HER2 and its importance in breast cancer. Oncology 61 Suppl 2, 1–13 (2001).
[7] Lee-Hoeflich, S. T. et al. A central role for HER3 in HER2-amplified breast cancer: Implications for tar- geted therapy. Cancer Res. 68, 5878–5887 (2008).
[8] Sergina, N. V et al. Escape from HER-family tyro- sine kinase inhibitor therapy by the kinase-inactive HER3. Nature 445, 437–441 (2007).
[9] Nahta, R., Yu, D., Hung, M.-C., Hortobagyi, G.
N. & Esteva, F. J. Mechanisms of Disease: under- standing resistance to HER2-targeted therapy in hu- man breast cancer. Nat. Clin. Pract. Oncol. 3, 269–280 (2006).
[10] Schoeberl, B. et al. Therapeutically targeting ErbB3:
a key node in ligand-induced activation of the ErbB receptor-PI3K axis. Sci. Signal. 2, ra31 (2009).
[11] Ocana, A. et al. HER3 overexpression and survival in solid tumors: A meta-analysis. Journal of the Na- tional Cancer Institute 105, 266–273 (2013).
[12] Mass, R. D. The HER receptor family: A rich target for therapeutic development. Int. J. Radiation Oncol- ogy Biol. Phys. 58, 932–940 (2004).
[13] Roskoski, R. The ErbB/HER family of protein- tyrosine kinases and cancer. Pharmacological Research 79, 34–74 (2014).
[14] Rexer, B. N. & Arteaga, C. L. Intrinsic and acquired resistance to HER2-targeted therapies in HER2 gene- amplified breast cancer: mechanisms and clinical implications. Crit. Rev. Oncog. 17, 1–16 (2012).
[15] Schaefer, G. et al. A Two-in-One Antibody against HER3 and EGFR Has Superior Inhibitory Activ- ity Compared with Monospecific Antibodies. Cancer Cell 20, 472–486 (2011).
[16] Ding, L. et al. Small sized EGFR1 and HER2 spe- cific bifunctional antibody for targeted cancer ther- apy. Theranostics 5, 378–398 (2015).
[17] McDonagh, C. F. et al. Antitumor Activity of a Novel Bispecific Antibody That Targets the ErbB2/ErbB3 Oncogenic Unit and Inhibits Heregulin-Induced Activation of ErbB3. Mol. Cancer Ther. 11, 582–593 (2012).
[18] Byrne, H., Conroy, P. J., Whisstock, J. C. &
O’Kennedy, R. J. A tale of two specificities: Bispe- cific antibodies for therapeutic and diagnostic appli- cations. Trends in Biotechnology 31, 621–632 (2013).
[19] Malm, M. et al. Engineering of a bispecific affibody molecule towards HER2 and HER3 by addition of an albumin-binding domain allows for affinity pu- rification and in vivo half-life extension. Biotechnol.
J. 9, 1215–1222 (2014).
[20] 1. Weber, J., Haberkorn, U. & Mier, W. Cancer strat- ification by molecular imaging. International Journal of Molecular Sciences 16, 4918–4946 (2015).
[21] Wållberg, H. et al. Molecular design and op- timization of 99mTc-labeled recombinant affibody molecules improves their biodistribution and imag- ing properties. J. Nucl. Med. 52, 461–469 (2011).
[22] Löfblom, J. et al. Affibody molecules: Engineered proteins for therapeutic, diagnostic and biotechno- logical applications. FEBS Letters 584, 2670–2680 (2010).
[23] Garousi, J. et al. ADAPT, a novel scaffold protein- based probe for radionuclide imaging of molecular targets that are expressed in disseminated cancers.
Cancer Res. 75, 4364–4371 (2015).
[24] Kontermann, R. E. Strategies for extended serum half-life of protein therapeutics. Current Opinion in Biotechnology 22, 868–876 (2011).
[25] Tolmachev, V. et al. Affibody molecules: potential for in vivo imaging of molecular targets for cancer therapy. Expert Opin. Biol. Ther. 7, 555–568 (2007).
[26] Orlova, A. et al. Tumor imaging using a picomo- lar affinity HER2 binding Affibody molecule. Cancer Res. 66, 4339–4348 (2006).
[27] Feldwisch, J. et al. Design of an Optimized Scaf- fold for Affibody Molecules. J. Mol. Biol. 398, 232–247 (2010).
[28] Kronqvist, N. et al. Combining phage and staphy- lococcal surface display for generation of ErbB3- specific Affibody molecules. Protein Eng. Des. Sel. 24, 385–396 (2011).
[29] Nilvebrant, J. et al. Engineering of bispecific affinity proteins with high affinity for ERBB2 and adaptable binding to albumin. PLoS One 9, (2014).
[30] Nilvebrant, J., Åstrand, M., Löfblom, J. & Hober, S. Development and characterization of small bispe-
cific albumin-binding domains with high affinity for ErbB3. Cell. Mol. Life Sci. 70, 3973–3985 (2013).
[31] Göstring, L. et al. Cellular effects of HER3-specific Affibody molecules. PLoS One 7, (2012).
[32] Malm, M. et al. Inhibiting HER3-Mediated Tumor Cell Growth with Affibody Molecules Engineered to Low Picomolar Affinity by Position-Directed Error- Prone PCR-Like Diversification. PLoS One 8, (2013).
[33] http://nebcloner.neb.com/#!/protocol/re/
double/XhoI,NdeI[Accessed June 28, 2016]) [34] Kroon, D. J., Baldwin-Ferro, A. & Lalan, P. Identifi-
cation of Sites of Degradation in a Therapeutic Mon- oclonal Antibody by Peptide Mapping. Pharm. Res.
An Off. J. Am. Assoc. Pharm. Sci. 9, 1386–1393 (1992).
[35] Arora, P., Oas, T. G. & Myers, J. K. Fast and faster: a designed variant of the B-domain of protein A folds in 3 microsec. Protein Sci. 13, 847–853 (2004).
[36] Kiraga, J. et al. The relationships between the iso- electric point and: length of proteins, taxonomy and ecology of organisms. BMC Genomics 8, 163 (2007).
7 A PPENDIX
Table 5: Primer sequences for the imaging agents and Z3-ABD-Z3-A2. Colour notation: Overhang sequence (purple), NdeI restriction site (red), XhoI restriction site (green).
GGE Forward 5’-CGTTGTCATATGGCCGAAGC-3’
GGE Reverse 5’-ACAACGCTCGAGTTAACACTCG-3’
WT-His6 Forward 5’-CGTTGTCATATGCACCATCATCA-3’
WT-His6 Reverse 5’-ACAACGCTCGAGCTCGAGTCATTTGGG-3’
General forward 5’-CGTTGTCATATGGCTGAAGCCAAGTACGCAAAG-3’
GGG Reverse 5’-ACAACGCTCGAGTTAGCATCCTCCCCCCTGG-3’
GGK Reverse 5’-ACAACGCTCGAGTTAGCATTTCCCACCCTGACTATC-3’
GGS Reverse 5’-ACAACGCTCGAGTTAGCAACTTCCCCCCTGGG-3’
WT Reverse 5’-ACAACGCTCGAGTCAACATTTAGGCGCTTGGGAATC-3’
3-A-3-2 Reverse 5’-ACAACGCTCGAGTCAACACTTCGGTGCTTGGCTG-3’
(a) (b)
Figure 6: Sequencing results analysed with Geneious (Biomatters). Deviations from expected sequences. (a) ZHer3-WT with an erroneous representation of Serine (TCC) instead of the correct Tyrosine (TAC) and (b) WT-His6 with an apparent insertion AAG(A)TT. However, this frame- shifting insertion would introduce a stop-codon a few bases downstream. Both proteins exhibit the correct mass when analysed with MS.
(a) -GGSC (AIEX) (b) -GGEC (AIEX)
(c) -GGGC (AIEX) (d) -WT (CIEX)
Figure 7: Representative ion-exchange chromatograms for the first batch: ZHer3-WT, -GGGC, -GGEC, and -GGSC.
(a) WT (CIEX) (b) -GGEC (CIEX)
(c) -GGGC (CIEX) (d) -GGGC (CIEX)
Figure 8: Representative CIEX chromatograms from the second batch of WT, -GGEC and -GGGC as well as from the purification of -GGKC.
(a) (b)
(c)
Figure 9: Analytical HPLC chromatograms showing the effect of (a) no reduction (b) DTT reduc- tion (c) TCEP reduction.
