Modification of the loops in the ligand-binding site turns avidin into a steroid-binding protein
- Tiina A Riihimäki1,
- Soili Hiltunen1,
- Martina Rangl2,
- Henri R Nordlund^1,
- Juha AE Määttä1,
- Andreas Ebner2,
- Peter Hinterdorfer2,
- Markku S Kulomaa1,
- Kristiina Takkinen3 and
- Vesa P Hytönen1Email author
© Riihimäki et al; licensee BioMed Central Ltd. 2011
Received: 29 April 2011
Accepted: 9 June 2011
Published: 9 June 2011
Engineered proteins, with non-immunoglobulin scaffolds, have become an important alternative to antibodies in many biotechnical and therapeutic applications. When compared to antibodies, tailored proteins may provide advantageous properties such as a smaller size or a more stable structure.
Avidin is a widely used protein in biomedicine and biotechnology. To tailor the binding properties of avidin, we have designed a sequence-randomized avidin library with mutagenesis focused at the loop area of the binding site. Selection from the generated library led to the isolation of a steroid-binding avidin mutant (sbAvd-1) showing micromolar affinity towards testosterone (Kd ~ 9 μM). Furthermore, a gene library based on the sbAvd-1 gene was created by randomizing the loop area between β-strands 3 and 4. Phage display selection from this library led to the isolation of a steroid-binding protein with significantly decreased biotin binding affinity compared to sbAvd-1. Importantly, differential scanning calorimetry and analytical gel-filtration revealed that the high stability and the tetrameric structure were preserved in these engineered avidins.
The high stability and structural properties of avidin make it an attractive molecule for the engineering of novel receptors. This methodology may allow the use of avidin as a universal scaffold in the development of novel receptors for small molecules.
Keywordsprotein engineering avidin scaffold phage display steroid hormone testosterone
Antibodies are the most widely used biomolecules for therapeutic, diagnostic and research applications, because they can be generated against virtually any molecule using protein engineering techniques (for a review see ). However, antibodies have certain fundamental disadvantages such as the complex architecture of their antigen-binding site, low stability, and a rather large size [2–5]. Moreover, the production of full-size antibodies is relatively expensive . To overcome these limitations and to improve therapeutic antibodies, antibodies have been extensively engineered . For example the size of the antibody molecule has been reduced by producing single-domain antigen-binding derivatives . In addition to extensive antibody engineering, a versatile repertoire of tailored biomolecules from non-immunoglobulin protein scaffolds have been generated [4, 9, 10]. Anticalins, derived from the lipocalin fold, are a good example of engineered proteins . The β-barrel structure of lipocalins is thermostable and robust and serves as an excellent scaffold for engineering novel receptors. They have been modified to bind novel ligands, such as fluorescein and digoxigenin, with affinities comparable with antibodies .
Chicken avidin (Avd), known for its extremely high affinity towards the water-soluble vitamin H, D-biotin, has been widely used in life science research applications . Aside from biotin, Avd also binds dyes and peptides, which share no significant structural similarity with biotin [14, 15]. Avd provides an attractive robust scaffold for the development of novel receptors, and Avd has many advantageous properties such as high chemical and thermal stability, a deep ligand binding site optimized for the binding of small molecules, and an oligomeric nature enabling signal amplification. Moreover, the structure of Avd is well characterized [16, 17] and numerous engineered forms of Avd have been described . Engineered Avd forms, in which the two pairs of the binding sites (dual-chain Avd) [19–21] or all four binding sites (single-chain Avd)  can be independently manipulated, have been developed.
In the present study Avd was modified to bind steroid hormones. Avd mutant sbAvd-1, which was captured using the phage display selection , has a micromolar affinity to testosterone. The steroid-binding protein sbAvd-1 was characterized and further engineered to decrease cross-reactivity towards other molecules, especially towards Avd's natural ligand biotin. The resulting sbAvd-2 mutant with a modified loop between β-strands 3 and 4 was found to prefer steroid hormones over biotin in ligand binding.
Functional display of Avd protein on the M13 phage
A portion of the Avd-pIII fusion was partially proteolytically cleaved, as can be seen from the blot analyzed with the anti-pIII monoclonal antibody (Figure 2B, ~35 kDa band). Because Avd display constructs designed in this study were based on the monovalent display mode (3 + 3), the intact pIII (migrating at ~58 kDa) expressed from the helper phage was also detected from the blot (Figure 2B, uppermost arrow).
Avd-displaying phages were functional because they bound specifically to the biotin-coated surfaces, and phages were efficiently amplified even after several panning rounds. Moreover, phagemid DNA with the Avd display expression unit, which was confirmed by the restriction enzyme digestion analysis, was stable during the panning rounds (data not shown).
To analyze the functionality of phages, the mixture of amplified Avd and Avd mutant N118M phages were screened by panning against 4-hydroxyazobenzene-2-carboxylic acid (HABA). As determined in our previous study, the biotin binding affinity of the Avd mutant N118M was reduced ~1,000,000-fold (Kd = 4.2 × 10-9 M) and HABA-affinity was increased ~1.5-fold (Kd = 5.2 × 10-6 M) compared to wtAvd  (Additional file 1). During the panning procedure a clear enrichment of Avd(N118M) phages over wt Avd phages was detected, which was an indication of the high selectivity of the produced phages. After only three rounds of selection phages displaying the Avd mutant N118M outcompeted the wtAvd phage population (Additional file 2).
Capture of a steroid-binding avidin
In the panning experiments Avd L1,2 loop library phages were introduced onto a testosterone-coated surface. The phage genomes carrying the mutated cDNA were found to be stable during the selection. A clear enrichment of sequences in the randomized loop area was observed, indicating the success of the selection conditions. Interestingly, we detected N12 as being a highly conserved amino acid residue among the enriched pool of proteins. An Avd variant, named sbAvd-1, that carried the sequence N12, R13, M14, N15, H16 was selected for further analysis.
The specificity of sbAvd-1 can be tuned by additional mutations in the loop between β-strands 3 and 4
To further lower the biotin-binding affinity and to decrease the cross-reactivity of steroid-binding Avd, a library (sbAvd-1 L3,4) was generated in which the loop area between β-strands 3 and 4 was randomized (Figure 3). The loop between β-strands 3 and 4 is highly important for biotin binding of Avd because this loop 'locks' biotin into the binding site. In this process, three amino acid residues in the loop form direct interactions with biotin . In the library four amino acids (T35, A36, V37 and, T38) were randomized using the NNY codon. The use of this codon covers 14 of the 20 naturally occurring amino acids while eliminating all of the stop codons. The library consisted of approximately 1.4 × 106 individual members, when calculated from the sequencing results and transformation efficiency, which exceeds the theoretical size of the library calculated based on the possible combinations of amino acid residues (3.8 × 104).
Binders from the sbAvd-1 L3,4 library were selected by phage display panning against a testosterone surface. In every panning round washes were adjusted according to the number of output colonies. The quality of the phage genomes carrying the mutated cDNA was evaluated by DNA sequencing at various stages during selection. A combination of acid and testosterone was used for elution. The selected sbAvd-1 phage clones were evaluated by microplate analysis using BSA-testosterone as a target ligand and utilizing M13-antibody to determine the amount of bound phages. The sbAvd-1 variant that showed the highest binding activity in the microplate assay (data not shown) had the sequence A35, T36, V37, N38. This mutant, named sbAvd-2 was selected for comparative analysis with sbAvd-1.
Production and purification of Avd mutants
Proteins were produced in soluble form in the E. coli strain BL21-AI using N-terminal OmpA bacterial secretion signal from Bordetella avium . Proteins were purified by Ni-NTA affinity chromatography that yielded ~2 mg/L pure protein. According to gel-filtration analysis, in solution, both sbAvd-1 (51 kDa) and sbAvd-2 (53 kDa) showed tetrameric state (Additional file 3) similar to that of wtAvd (Avd expressed in E. coli 53 kDa, chicken Avd 60 kDa ). The slight decrease in molecular weight of Avd expressed in bacteria can be explained by the lack of glycosylation.
Determination of ligand-binding specificity of Avd forms by microplate assay
The binding to the surface-immobilized ligands was inhibited by pre-incubating the proteins with 10 μM D-biotin (Figure 4B). Free biotin significantly inhibited the binding of sbAvd-1 to steroids, indicating notable affinity towards biotin. However, in the case of sbAvd-2 the binding of testosterone and progesterone was not affected by biotin, showing a clear decrease in biotin-binding affinity.
Biosensor analyses of steroid-binding Avds
Kinetic parameters and the determined thermostability of sbAvds
ΔH ×104 (cal/mol)
4.2 × 105
5.6 × 10-4
1.4 × 10-9
1.0 × 103
9.5 × 10-3
9.0 × 10-6
1.0 × 103
6.8 × 10-4
6.6 × 10-7
8.5 × 10-3
1.1 × 10-5
The biotin binding of the steroid-binding proteins was analyzed by the biotin-coated sensor chip. The comparison of sbAvd-1 and sbAvd-2 revealed that the biotin-binding affinity decreased almost 500-fold due to the modification of the loop between β-strands 3 and 4 (Table 1). This result appears to be consistent with the previous study in which mutation of T35A alone decreased the biotin-binding affinity of wtAvd approximately 200-fold .
Interaction analysis of steroid-binding Avds by MRFS
To prove the specificity of the measured interactions, control experiments were performed. For these experiments, free testosterone was injected into the measuring solution to preoccupy the binding sites of the steroid-binding protein immobilized on the surface. The retraction curve identical to the approaching curve was detected (insets of Figure 6B and 6C), which indicated the total inhibition of the binding by free testosterone. Additionally, when the binding sites were preoccupied with free testosterone, the number of binding events dropped down to 67 from 183 detected interactions in the case of sbAvd-1. In case of sbAvd-2 the number of interactions dropped from of 215 to 21 after addition of free testosterone.
Differential scanning calorimetry reveals the high stability of the engineered proteins
The high thermal stability of wtAvd (temperature transition midpoint (Tm) = 85.5°C) was preserved in the steroid-binding mutants; these proteins showed similar Tm values compared to wtAvd (sbAvd-1 Tm value of 80.6°C and sbAvd-2 Tm value of 82.5°C) (Table 1, Additional file 4). Ligand binding often stabilizes a protein and raises the Tm value. As expected, the addition of biotin stabilized wtAvd effectively, increasing the Tm value almost 40°C. This Tm indicates a very high binding affinity.
The effect of biotin-binding on the protein stability of steroid-binding proteins was much smaller than in the case of wtAvd, thus showing decreased biotin affinity. In fact, sbAvd-2 showed a negligible increase in Tm in the presence of biotin (ΔTm = 0.5°C). The presence of testosterone only slightly increased the Tm values of sbAvds (ΔTm = 0.6-0.9°C). Actually, a similarly small increase was detected in the ΔTm value of wtAvd when testosterone was added. However, because the testosterone-binding of wtAvd was not detected in other analyses (SPR and microplate analysis) the mechanism of stabilization remains unclear.
In the present study, Avd proteins were displayed on the M13 phage in a monovalent form fused with the C-terminal region (aa residues 198-406) of the coat protein pIII. The crucial requirement for the functional display on phage is successful expression of the protein in E. coli. As has been earlier reported Avd protein can be efficiently expressed in a soluble form in E. coli . The close relative of Avd, streptavidin, has been displayed on bacterial phages as a pVIII fusion [30, 31]. However, to our best knowledge, chicken Avd has not been previously displayed on phages for screening purposes.
To evaluate the different modes for Avd display, we generated constructs expressing pIII fused to Avd and constructs that expressed free Avd subunits in addition to the pIII fusions. Based on the expression analysis, binding properties and selection experiments, the construct that expressed both fusion and free Avd showed an enhanced assembly of functional, tetrameric Avd on the phage compared to the Avd-pIII fusion alone (Additional file 2). However, the construct that expressed pIII-fused Avd was chosen for use in the selection system for the generation of gene libraries because the use of only one Avd gene in the protein display would ensure that the selected proteins could be assembled to homotetramers.
The potential of the developed platform for screening novel Avd-based receptors was investigated. Five residues in the loop between β-strands 1 and 2 of Avd were randomized to generate a population of diverse Avd genes. The gene population was screened for novel binding properties using the phage display method . We observed the enrichment of Avd phages binding to testosterone-coupled BSA. The Avd variant sbAvd-1, which carried the sequence N12, R13, M14, N15, H16, was further analyzed and found to bind free testosterone with an affinity similar to BSA-conjugated testosterone. Moreover, we detected that the biotin-binding affinity of sbAvd-1 was not completely diminished; it was still high enough to inhibit sbAvd-1 from binding to testosterone-BSA (Table 1, Figures 4 and 5). This finding confirmed that testosterone occupies the same binding site in Avd as does biotin. Based on the cross-creativity measurements, sbAvd-1 also binds other steroid hormones, such as progesterone and DHEAS (Figures 4 and 5).
SbAvd-1 was further engineered to decrease biotin binding and to improve steroid specificity. From the phage display selections, the sbAvd-1-derived protein, named sbAvd-2, with a mutated loop between β-strands 3 and 4 was captured. This mutant showed similar or slightly decreased binding affinity towards immobilized testosterone, whereas the biotin-binding affinity was clearly decreased (Table 1, Figures 4 and 5). We also noticed that sbAvd-2 bound more tightly to free testosterone than did sbAvd-1 (Figure 5). Significantly, sbAvd-2 preferred steroids as a ligand over biotin; this finding is important when considering the applications for biofluids, in which biotin is often present in relatively high concentrations.
This study and our preliminary results from experiments with a number of different target ligands (Hiltunen S, Riihimäki TA et al., unpublished data) suggest that Avd-based receptors for various different small molecules can be tailored. These Avd-based receptors may be valuable tools for diagnostic use in the future.
The current study provides a promising platform for the selection of tailored ligand-binders evolved from the Avd scaffold in the monovalent pIII protein display. The Avd scaffold has characteristics that are beneficial in protein engineering, such as high thermal and chemical stability, simple folding and an optimal structure for small ligands. Furthermore, Avd can be modified rather freely without major change in the fold . Novel Avds that can simultaneously bind multiple ligands could become next-generation molecular tools for clinical and diagnostic applications [22, 32]. Importantly, novel Avd-based receptors could be used in applications that require harsh conditions.
Construction of phagemid vectors
All of the basic recombinant DNA methods were performed essentially as previously described . Appropriate restriction sites were added to the cDNA of the Avd core sequence  by PCR using the primers Avd_NheI_5' and Avd_NotI_3' (Additional file 5). PCR products were first subcloned into the pCR®2.1-TOPO plasmid by TOPO TA-cloning (Invitrogen) and the plasmids were transformed into E. coli TOP10 cells. Plasmids were isolated from colonies that contained the inserts based on the blue-white screening. Avd fragments were cut out from the pCR®2.1-TOPO-plasmid using the NheI and NotI restriction enzymes and ligated into the phagemid vector (pBluescript SK+ derived phagemid, Research Center of Finland, Biotechnology, Espoo, Finland). The cDNA of the Avd was cloned into the phagemid vector as an N-terminal fusion to the C-terminal domain (amino acids 198-406) of the minor phage coat protein III. In the Avd/Avd-pIII constructs, the coding sequence of the free Avd was subcloned using the primers Avd_NheI_5' and Avd_AscI_stop_3' (Additional file 5). The Avd-pIII expression cassette was generated by subcloning a fragment amplified from the primers Avd_SfiI_5' and Avd_NotI_3' (Additional file 5). The nucleotide sequences of the Avd constructs were verified by sequencing on an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems) according to the protocols recommended by the manufacturer (ABI PRISM BigDye Terminator Cycle Sequencing Kit v.1.1, Applied Biosystems).
Amplification of phage particles
All of the basic phage display methods were performed essentially as previously described . Phagemid vectors with the Avd insert were transformed into chemically competent E. coli XL1-Blue cells (Stratagene, La Jolla, CA) with the heat shock method. Phage stocks of the different Avd display constructs were made from individual colonies picked from the transformation plates into super broth (SB) medium supplemented with the appropriate antibiotics and glucose. The bacterial cultures were infected with the helper phage (1012 pfu/ml) VCS-M13 (Stratagene, LaJolla, CA) for amplification of Avd phages. Phages were PEG precipitated and analyzed by SDS-PAGE and western blotting following immunostaining with the polyclonal rabbit α-avd IgG (University of Oulu, 1:5000) and the monoclonal mouse anti-pIII IgG (Biosite, Sweden, 1:2000) antibodies.
Functionality test of Avd-displaying phages
The functionality of the Avd phages was tested by panning the phages on surfaces coated with BSA conjugated to HABA (HABA-BSA) . As a positive control for selection, the Avd mutant N118M was also displayed on the phages. Panning was performed essentially as previously described . For elution, vigorous shaking in 100 mM hydrochloric acid containing 10 μM D-biotin (Biochemica, Fluka, 14400) was performed. In total, three selection rounds were performed. After each round the integrity of the Avd expression units was analyzed by restriction enzyme digestions of the phagemid DNA samples. Importantly, both single and double Avd and Avd(N118M) construct phages were competed against each other. Equivalent amounts of phages displaying Avd or Avd(N118M) mutant were mixed and biopanned. The ratio between constructs was determined with DNA sequencing.
Construction of Avd L1,2 library
To construct an Avd DNA library, amino acids N12, D13, L14, G15, and S16 in the loop between beta strands 1 and 2 (L1,2) were randomized. The libraries were constructed essentially as previously described . A nucleic acid fragment of 105 base pairs was amplified with the primers Avd_NheI_5' and Loop 1-2 _R1_3' (Additional file 5) using wt Avd cDNA as template. Parallel to this fragment, a nucleic acid fragment with 357 base pairs was PCR-amplified with the primers Loop 1-2_R2_5' and Avd_NotI_3' (Additional file 5) using wt Avd as a template. The PCR strategy is presented in Additional file 6. The desired amplification products were separated by agarose gel electrophoresis and isolated from the gel using the Nucleo Spin Extract II kit (Macherey-Nagel) according to the manufacturer's instructions. PCR products were combined in the second amplification step in the presence of the PCR primers Avd_NheI_5' and Avd_NotI_3' (Additional file 5). This amplification resulted in a DNA fragment of 462 base pairs. The fragment was isolated from the gel and cut with the restriction enzymes NheI and NotI (Fermentas) followed by purification using the Nucleo Spin Extract II kit (Macherey-Nagel). Fragments were ligated into the phagemid vector and the resulting ligation product was transformed into electrocompetent cells of the E. coli strain XL1-Blue (Stratagene) by electroporation. Transfected bacteria were infected with the VCS-M13 helper phage (Stratagene) and phages were harvested from the culture and purified with PEG precipitation. The amount of phage particles was determined by titration.
Selecting steroid-binders from the Avd L1,2 library
The Avd L1,2 library was panned against the steroid hormone testosterone. As a negative control, wtAvd-displaying phages were also panned against testosterone. NUNC immunosorp plates were coated with a testosterone-BSA conjugate (Sigma, T-3392, 1 μg). Phages were preincubated in the BSA-coated wells to prevent non-specific binding. Three to four selection cycles were conducted, and the stringency of washing conditions was increased every panning round to decrease nonspecific binding. In the first panning round wells were washed three times with phosphate buffered saline containing 0.05% Tween (PBS-Tween) and five times with phosphate buffered saline (PBS). Phages were eluted by vigorous shaking for 10 minutes in 100 mM hydrochloric acid containing 10 μM D-biotin (Biochemica, Fluka, 14400). Biotin was used for elution because it was probable that after randomization of the loop L1,2 the resulting Avd would still have a rather high affinity towards biotin. In the final panning round 10 μM testosterone (Steraloids Inc., USA) was used in addition to 100 mM hydrochloric acid for elution. The eluted phage solutions were neutralized with 2 M Tris. Eluted phages were used to infect E. coli XL1-Blue cells and aliquots of the infected bacteria were plated to quantify the amount of eluted phages. Phages were amplified and purified as described earlier. Precipitated phages were then used for the next round of selection. After every panning round, results were verified by sequencing (20 sequences), and the number of phage particles was determined by phage titration.
Generation and screening of the SbAvd-1 L3,4 library
An SbAvd-1-derived DNA library was constructed and ligated into the phagemid as described above. In the library amino acids T35, A36, V37, and T38 in the loop between beta strands 3 and 4 (L3,4) were randomized; thus, in the construction of the library the primers Avd_NheI_5' and 3_4R_1_3' and primers 3_4R_2_5' and Avd_NotI_3' (Additional file 5) were used in the PCR. SbAvd-1 cDNA was used as a template. Ligated phagemid was transformed into electrocompetent cells of E. coli strain XL1-Blue (Stratagene) by electroporation. Transfected bacteria were infected with the VCS-M13 helper phage (Stratagene) and phages were harvested from the culture and purified with PEG precipitation. The amount of phage particles was determined by titration.
The sbAvd-1 L3,4 library was biopanned against the steroid hormone testosterone similarly as described above, with some exceptions. NUNC immunosorp plates were coated with a testosterone-BSA conjugate (Sigma, T-3392, 200 ng). Phages were preincubated in BSA-coated wells before the panning procedure. Four selection cycles were performed. Phages were eluted by vigorous shaking for 10 minutes in 100 mM hydrochloric acid containing 35 μM testosterone (Steraloid Inc., USA). After each panning round, results were verified by sequencing (20 sequences), and the number of phage particles was determined by phage titration. Additionally, every panning round was screened for binders by anti-M13 as previously described . For the ELISA, (NUNC) immunosorp plates were coated with a testosterone-BSA conjugate (Sigma, T-3392, 1 μg), wells were blocked with 5% milk solution, and detection was performed with Anti-M13/HRP (GE Healthcare) and read with a microplate reader (Bio-Rad 680 XR).
Production and purification of Avd mutants
For biochemical analyses, the proteins were produced in E. coli strain BL21-AI (Invitrogen) using expression vector pET101/D (Invitrogen) . After sonication (Sonics & Materials Vibra Cell™) and DNaseI (New England Bio Labs) treatment of E. coli cells, the purification of the proteins was conducted using Ni-NTA affinity chromatography according to the instructions of the manufacturer (QIAGEN).
The oligomeric state of the proteins was assayed with fast protein liquid chromatography (FPLC) gel-filtration using an ÄKTApurifier™ HPLC equipped with a Superdex 200 10/300 GL column (Tricorn, Amersham Biosciences, GE Healthcare). The column was calibrated using the gel-filtration mixture (thyroglobulin, γ-globulin, ovalbumin, myoglobin, and vitamin B12; Bio-Rad Laboratories) as a molecular-mass standard. Sodium phosphate buffer (20 mM, pH 7.4) with 1M NaCl and, 20 mM imidazole was used as the liquid phase. Protein samples of 90-193 μg in a volume of 500 μl were used in the analysis.
DSC measurements of wtAvd and steroid-binding Avds
Proteins (0.225 mg/ml) were analyzed in sodium phosphate buffer (20 mM, pH 7.4), containing 20 mM imidazole and 1 M NaCl. D-biotin (Biochemica, Fluka, 14400) and testosterone (Steraloids Inc., USA) were diluted with the measurement buffer to a final concentration of 50 μM. All solutions were degassed prior to measurements to avoid air bubbles. An automated capillary VP-DSC instrument (GE Healtcare, MicroCal, Northampton, USA) was used to measure the stability of the proteins with or without ligands. During the measurement, protein samples were heated from 20°C to 130°C at a scanning rate of 120°C/h. Feedback mode was set to low and the filler period was 8 s. Temperature transition midpoints (Tm) were recorded from the highest peaks and the calorimetric heat changes (ΔH) were calculated using the MicroCal Origin 7 software (GE Healtcare, MicroCal, Northampton, USA).
Determination of ligand-binding specificity of Avd forms by microplate assay
MaxiSorp F96 microplate wells (NUNC) were coated with 500 ng of conjugated ligand (HSA-conjugated progesterone, hydrocortisone, testosterone, and cholic acid (Technical Research Center of Finland, Espoo, Finland); or BSA-conjugated testosterone (A6958-000, Steraloids Inc., USA), and biotin (Jenni Leppiniemi, University of Tampere, Finland)), and with 500 ng of the carrier proteins HSA and BSA (A7906, Sigma) in 100 μl of PBS for 2 hours 37°C. Carrier proteins were used as negative controls. Plates were washed three times with PBS-Tween, blocked with 0.5% BSA-PBS for 30 min, and then washed again. Proteins (0.9 μg/ml) in sodium phosphate buffer (20 mM, pH 7.4) with 1M NaCl and 20 mM imidazole, were added to the wells and incubated for 1 h. A portion of wtAvd (Belovo, Bastogne, Belgium), sbAvd-1 and sbAvd-2 was preincubated with 10 μM D-biotin (Biochemica, Fluka, 14400). Bound Avd was detected by rabbit α-avd IgG (University of Oulu) and with alkaline phosphatase conjugated goat anti-rabbit IgG (A3937, Sigma). After adding the phosphatase substrate solution (1 mg/ml pNPP (S0942, Sigma) in 1 M diethanolamine pH 9.8, with 0.5 mM MgCl2), the plates were read after 15 minutes at A405 with a microplate reader (Bio-Rad 680 XR).
Biosensor analyses of steroid-binding Avds
A BIAcore X optical biosensor (Biacore, Uppsala, Sweden) was used for the analysis of binding kinetics. Testosterone-BSA was coupled to the carboxymethylated dextran layer of a sensor chip using standard amine coupling chemistry (1000 RU, 40 μl/min flow rate). Samples of sbAvd-1, and sbAvd-2 were diluted in 50 mM sodium phosphate containing 1 M NaCl, and the same buffer was used in the measurements. The binding of the sbAvd-1 and sbAvd-2 samples on testosterone-BSA coated chips was measured and the kinetic constants were determined from the measurements performed with different protein concentrations using the BiaEvaluation software according to the manufacturer's instructions. WtAvd was used as a negative control.
The binding of steroid-binding proteins to the testosterone-BSA surface was competed with free steroid hormone molecules (testosterone, DHEAS, androstenedione, estradiol, and DHT (Steraloids Inc., USA)) and free biotin (Biochemica, Fluka, 14400) to evaluate the specificity of binding. Additionally, steroid-binding was more closely detected by measuring the binding in the presence of varying concentrations (0.75 μM-50 μM) of inhibiting testosterone.
For the determination of biotin-binding kinetics, a sensor chip was prepared as follows: diaminoethylene was first attached to the surface using a mixture containing 0.2 M 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and 0.05 M N-hydroxysuccinimide (NHS) in water. Second, to introduce amino groups to the surface, 1 M ethylenediamine (Fluka 03550) in water was applied. Finally, 5 mM biotin N-succinimidyl ester (Biochemica, Fluka, 14405) in 50% DMSO was injected on the surface (40 μl/min flow rate, ~130 RU; please note that the determination of the bound mass is not very accurate in case of small molecules because the immobilization can change the physicochemical properties of the surface).
Molecular Recognition Force Spectroscopy experiments
Molecular recognition force spectroscopy (MRFS) was used to study the interaction of the produced proteins with testosterone and biotin. Steroid-binding proteins were covalently bound to modified mica sheets via lysines as previously described . Testosterone was coupled to the AFM tip via the heterobifunctional Fmoc-PEG-NHS crosslinker as previously described , resulting in a covalent attachment of testosterone via a flexible spacer (Figure 6A).
All MRFS experiments were performed on a Pico SPM I (Agilent Technologies, Santa Clara, CA). All modified cantilevers used had nominal spring constants between 10-30 pN/nm (Veeco Instruments, Santa Barbara, CA). The effective spring constants of the cantilevers were determined by the thermal noise method [39, 40]. Force-distance cycles were completed using a z-range of 200 and 300 nm. Sweep durations were adjusted between 0.25 and 4 s. During one data set of 1000 force-distance curves, the lateral tip position was changed (a few hundred nm) about every 100 curves to ensure that the binding events were statistically reasonable. The specificity of the binding was proved by adding free testosterone (200 nM) into the measuring solution and incubating for approximately 1 h to block the ligand-binding sites of the proteins.
atomic force microscopy
bovine serum albumin: BTN: D-biotin
1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride
human serum albumin
molecular recognition force spectroscopy
phosphate buffered saline
sodium-dodecyl sulfate polyacrylamide gel electrophoresis
wild type avidin.
We thank the laboratory staff of VTT (Technical Research Center of Finland) Biotechnology, Espoo, Finland, for excellent technical assistance. We thank MSc. Antti Tullila for kindly providing us HSA-conjugates. We also thank the staff at the Molecular Biology group, Department of Biological and Environmental Science, University of Jyväskylä for their assistance. We thank MSc. Sampo Kukkurainen for help in molecular modeling. We would also like to thank MSc. Kaisa Helttunen, Professor Kari Rissanen and Dr. Juhani Huuskonen for their help in the preparation of the HABA-BSA conjugate. We thank Hong Chang for participation in the experimental work. The study was financially supported by the Academy of Finland (115976, 121236); the National Technology Agency of Finland (BioFace 40055/08); European Micro and Nano Technology support program (FFG 421695); Pirkanmaa Hospital District and Tampere Graduate Program in Biomedicine and Biotechnology.
- Hoogenboom HR: Selecting and screening recombinant antibody libraries. Nat Biotechnol. 2005, 23: 1105-1116. 10.1038/nbt1126.View ArticleGoogle Scholar
- Hoess RH: Protein design and phage display. Chem Rev. 2001, 101 (10): 3205-3218. 10.1021/cr000056b.View ArticleGoogle Scholar
- Sarikaya M, Tamerler C, Jen AK, Schulten K, Baneyx F: Molecular biomimetics: nanotechnology through biology. Nat Mater. 2003, 2: 577-585. 10.1038/nmat964.View ArticleGoogle Scholar
- Binz HK, Amstutz P, Pluckthun A: Engineering novel binding proteins from nonimmunoglobulin domains. Nat Biotechnol. 2005, 23: 1257-1268. 10.1038/nbt1127.View ArticleGoogle Scholar
- Skerra A: Alternative non-antibody scaffolds for molecular recognition. Curr Opin Biotechnol. 2007, 18: 295-304. 10.1016/j.copbio.2007.04.010.View ArticleGoogle Scholar
- Werner RG: Economic aspects of commercial manufacture of biopharmaceuticals. J Biotechnol. 2004, 113: 171-182. 10.1016/j.jbiotec.2004.04.036.View ArticleGoogle Scholar
- Hudson PJ, Souriau C: Engineered antibodies. Nat Med. 2003, 9: 129-134. 10.1038/nm0103-129.View ArticleGoogle Scholar
- Saerens D, Ghassabeh GH, Muyldermans S: Single-domain antibodies as building blocks for novel therapeutics. Curr Opin Pharmacol. 2008, 8: 600-608. 10.1016/j.coph.2008.07.006.View ArticleGoogle Scholar
- Nygren PA, Skerra A: Binding proteins from alternative scaffolds. J Immunol Methods. 2004, 290: 3-28. 10.1016/j.jim.2004.04.006.View ArticleGoogle Scholar
- Gebauer M, Skerra A: Engineered protein scaffolds as next-generation antibody therapeutics. Curr Opin Chem Biol. 2009, 13: 245-255. 10.1016/j.cbpa.2009.04.627.View ArticleGoogle Scholar
- Beste G, Schmidt FS, Stibora T, Skerra A: Small antibody-like proteins with prescribed ligand specificities derived from the lipocalin fold. Proc Natl Acad Sci USA. 1999, 96: 1898-1903. 10.1073/pnas.96.5.1898.View ArticleGoogle Scholar
- Schlehuber S, Skerra A: Tuning ligand affinity, specificity, and folding stability of an engineered lipocalin variant -- a so-called 'anticalin' -- using a molecular random approach. Biophys Chem. 2002, 96: 213-228. 10.1016/S0301-4622(02)00026-1.View ArticleGoogle Scholar
- Laitinen OH, Nordlund HR, Hytönen VP, Kulomaa MS: Brave new (strept)avidins in biotechnology. Trends Biotechnol. 2007, 25: 269-277. 10.1016/j.tibtech.2007.04.001.View ArticleGoogle Scholar
- Green NM: Avidin. Adv Protein Chem. 1975, 295: 85-133.View ArticleGoogle Scholar
- Repo S, Paldanius TA, Hytönen VP, Nyholm TK, Halling KK, Huuskonen J, Pentikäinen OT, Rissanen K, Slotte JP, Airenne TT, Salminen TA, Kulomaa MS, Johnson MS: Binding Properties of HABA-Type Azo Derivatives to Avidin and Avidin-Related Protein 4. Chem Biol. 2006, 13: 1029-1039. 10.1016/j.chembiol.2006.08.006.View ArticleGoogle Scholar
- Livnah O, Bayer EA, Wilchek M, Sussman JL: Three-dimensional structures of avidin and the avidin-biotin complex. Proc Natl Acad Sci USA. 1993, 90297: 5076-5080.View ArticleGoogle Scholar
- Rosano C, Arosio P, Bolognesi M: The X-ray three-dimensional structure of avidin. Biomol Eng. 1999, 16: 5-12. 10.1016/S1050-3862(99)00047-9.View ArticleGoogle Scholar
- Laitinen OH, Hytönen VP, Nordlund HR, Kulomaa MS: Genetically engineered avidins and streptavidins. Cell Mol Life Sci. 2006, 63: 2992-3017. 10.1007/s00018-006-6288-z.View ArticleGoogle Scholar
- Nordlund HR, Laitinen OH, Hytönen VP, Uotila ST, Porkka E, Kulomaa MS: Construction of a dual chain pseudotetrameric chicken avidin by combining two circularly permuted avidins. J Biol Chem. 2004, 279: 36715-36719. 10.1074/jbc.M403496200.View ArticleGoogle Scholar
- Hytönen VP, Nordlund HR, Hörhä J, Nyholm TK, Hyre DE, Kulomaa T, Porkka EJ, Marttila AT, Stayton PS, Laitinen OH, Kulomaa MS: Dual-affinity avidin molecules. Proteins. 2005, 61: 597-607. 10.1002/prot.20604.View ArticleGoogle Scholar
- Riihimäki TA, Kukkurainen S, Varjonen S, Hörhä J, Nyholm TKM, Kulomaa MS, Hytönen VP: Construction of chimeric dual-chain avidin by tandem fusion of the related avidins. PLoS One. 2011, 6: e20535-10.1371/journal.pone.0020535.View ArticleGoogle Scholar
- Nordlund HR, Hytönen VP, Hörhä J, Maatta JA, White DJ, Halling K, Porkka E, Slotte JP, Laitinen OH, Kulomaa MS: Tetravalent single chain avidin: From subunits to protein domains via circularly permuted avidins. Biochem J. 2005, 392: 485-491. 10.1042/BJ20051038.View ArticleGoogle Scholar
- Smith GP: Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science. 1985, 228: 1315-1317. 10.1126/science.4001944.View ArticleGoogle Scholar
- Hoogenboom HR, Griffiths AD, Johnson KS, Chiswell DJ, Hudson P, Winter G: Multi-subunit proteins on the surface of filamentous phage: methodologies for displaying antibody (Fab) heavy and light chains. Nucleic Acids Res. 1991, 19: 4133-4137. 10.1093/nar/19.15.4133.View ArticleGoogle Scholar
- Määttä JA, Airenne TT, Nordlund HR, Jänis J, Paldanius TA, Vainiotalo P, Johnson MS, Kulomaa MS, Hytönen VP: Rational modification of ligand-binding preference of avidin by circular permutation and mutagenesis. Chembiochem. 2008, 9: 1124-1135. 10.1002/cbic.200700671.View ArticleGoogle Scholar
- Hytönen VP, Laitinen OH, Airenne TT, Kidron H, Meltola NJ, Porkka E, Hörhä J, Paldanius T, Määttä JA, Nordlund HR, Johnson MS, Salminen TA, Airenne KJ, Ylä-Herttuala S, Kulomaa MS: Efficient production of active chicken avidin using a bacterial signal peptide in Escherichia coli. Biochem J. 2004, 384: 385-390. 10.1042/BJ20041114.View ArticleGoogle Scholar
- Hinterdorfer P, Dufrene YF: Detection and localization of single molecular recognition events using atomic force microscopy. Nat Methods. 2006, 3: 347-355. 10.1038/nmeth871.View ArticleGoogle Scholar
- Wildling L, Hinterdorfer P, Kusche-Vihrog K, Treffner Y, Oberleithner H: Aldosterone receptor sites on plasma membrane of human vascular endothelium detected by a mechanical nanosensor. Pflugers Arch. 2009, 458: 223-230. 10.1007/s00424-008-0615-1.View ArticleGoogle Scholar
- Baumgartner W, Hinterdorfer P, Schindler H: Data analysis of interaction forces measured with the atomic force microscope. Ultramicroscopy. 2000, 82: 85-95. 10.1016/S0304-3991(99)00154-0.View ArticleGoogle Scholar
- Avrantinis SK, Stafford RL, Tian X, Weiss GA: Dissecting the streptavidin-biotin interaction by phage-displayed shotgun scanning. Chembiochem. 2002, 3: 1229-1234. 10.1002/1439-7633(20021202)3:12<1229::AID-CBIC1229>3.0.CO;2-X.View ArticleGoogle Scholar
- Sidhu SS, Weiss GA, Wells JA: High copy display of large proteins on phage for functional selections. J Mol Biol. 2000, 296: 487-495. 10.1006/jmbi.1999.3465.View ArticleGoogle Scholar
- Leppiniemi J, Määttä JA, Hammaren H, Soikkeli M, Laitaoja M, Jänis J, Kulomaa MS, Hytönen VP: Bifunctional avidin with covalently modifiable ligand binding site. PLoS One. 2011, 6: e16576-10.1371/journal.pone.0016576.View ArticleGoogle Scholar
- Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual. 1990, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NYGoogle Scholar
- Gope ML, Keinänen RA, Kristo PA, Conneely OM, Beattie WG, Zarucki-Schulz T, O'Malley BW, Kulomaa MS: Molecular cloning of the chicken avidin cDNA. Nucleic Acids Res. 1987, 15: 3595-3606. 10.1093/nar/15.8.3595.View ArticleGoogle Scholar
- Barbas CF, Burton DR, Scott JK, Silverman GJ: Phage Display, A Laboratory manual. 2001, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NYGoogle Scholar
- Hofstetter H, Morpurgo M, Hofstetter O, Bayer EA, Wilchek M: A labeling, detection, and purification system based on 4-hydroxyazobenzene-2-carboxylic acid: an extension of the avidin-biotin system. Anal Biochem. 2000, 284: 354-366. 10.1006/abio.2000.4617.View ArticleGoogle Scholar
- Kingsbury GA, Junghans RP: Screening of phage display immunoglobulin libraries by anti-M13 ELISA and whole phage PCR. Nucleic Acids Res. 1995, 23: 2563-2564. 10.1093/nar/23.13.2563.View ArticleGoogle Scholar
- Kamruzzahan AS, Kienberger F, Stroh CM, Berg J, Huss R, Ebner A, Zhu R, Rankl C, Gruber HJ, Hinterdorfer P: Imaging morphological details and pathological differences of red blood cells using tapping-mode AFM. Biol Chem. 2004, 385: 955-960. 10.1515/BC.2004.124.View ArticleGoogle Scholar
- Hutter JL, Bechhoefer J: Calibration of atomic-force microscope tips. Review of Scientific Instruments. 1993, 64: 1868-1873. 10.1063/1.1143970.View ArticleGoogle Scholar
- Butt HJ, Jaschke M: Calculation of thermal noise in atomic force microscopy. Nanotechnology. 1995, 6: 1-7. 10.1088/0957-4484/6/1/001.View ArticleGoogle Scholar
- Livnah O, Bayer EA, Wilchek M, Sussman JL: Three-dimensional structures of avidin and the avidin-biotin complex. Proc Natl Acad Sci USA. 1993, 90: 5076-5080. 10.1073/pnas.90.11.5076.View ArticleGoogle Scholar
- Humphrey W, Dalke A, Schulten K: VMD: visual molecular dynamics. J Mol Graph. 1996, 14: 33-8. 10.1016/0263-7855(96)00018-5. 27-8View ArticleGoogle Scholar