Engineering of an elastic scaffolding polyprotein based on an SH3-binding intrinsically disordered titin PEVK module

Abstract

Titin is a large elastic protein found in muscle that maintains the elasticity and structural integrity of the sarcomere. The PEVK region of titin is intrinsically disordered, highly elastic and serves as a hub to bind signaling proteins. Systematic investigation of the structure and affinity profile of the PEVK region will provide important information about the functions of titin. Since PEVK is highly heterogeneous due to extensive differential splicing from more than one hundred exons, we engineered and expressed polyproteins that consist of a defined number of identical single exon modules. These customized polyproteins reduce heterogeneity, amplify interactions of less dominant modules, and most importantly, provide tags for atomic force microscopy and allow more readily interpretable data from single-molecule techniques.

Expression and purification of recombinant polyprotein with repeat regions presented many technical challenges: recombination events in tandem repeats of identical DNA sequences exacerbated by high GC content, toxicity of polymer plasmid and expressed protein to the bacteria; early truncation of proteins expressed with different numbers of modules; and extreme sensitivity to proteolysis. We have investigated a number of in vitro and in vivo bacterial and yeast expression systems, as well as baculoviral systems as potential solutions to these problems. We successfully expressed and purified in gram quantities a polyprotein derived from human titin exon 172 using Pichia pastoris yeast. This study provides valuable insights into the technical challenges regarding the engineering and purification of a tandem repeat sequence of an intrinsically disordered biopolymer.

Introduction

Intrinsically disordered and repetitive regions are abundant in the genomes of many organisms [1–3]. These regions have important functions in the regulation of cell signal pathways [4–7]. They have been implicated in many illnesses such as cancer, diabetes and neurodegenerative diseases [8–13]. Therefore, it becomes increasingly important to study the functions of these repetitive and intrinsically disordered regions.

Titin is a large elastic muscle protein (3–4 MDa) in skeletal and cardiac muscles. It consists mainly of numerous immunoglobulin-like domains, fibronectic type III domains, and the PEVK region [14]. Titin plays a critical role in maintaining the stability of sarcomeres and assembling nascent sarcomeres during the development of muscle cells [15, 16]. Recent studies have revealed the role of the titin kinase and the PEVK region in the signaling pathways that regulate muscle function [17, 18]. The PEVK region of titin is expressed by differential splicing from a large number of exons, e.g. exons 109–224 in humans, and consists of tandem repeats of numerous 28-amino acid motifs enriched in proline, glutamic acid, valine, and lysine. The PEVK region of titin has the unique function of regulating sarcomere tension upon stretching [19, 20].

We have studied the physical and biochemical characteristics of synthetic modules and cloned fragments of the PEVK region of titin extensively [21, 22]. Our studies indicated that the PEVK region of titin is highly charged with gel-like, intrinsically disordered structures [23, 24]. The PEVK region of titin behaves like a spring where part of its elasticity may be due to charge-charge interactions between side chains of the molecules, since the persistence length of the region increases with an elevation in ionic strength [21, 25, 26]. We also discovered that the left handed PPII helices in many PEVK modules of titin bind to Src homolog domains (SH3) of nebulin and other signaling proteins [27, 28]. Such PPII sequences thus may mediate interactions between nebulin and titin [29–31].

To further understand the possible role of PPII in the PEVK region in the signaling functions and mechanical properties of titin, we engineered a recombinant polyprotein with 15 identical repeat modules to facilitate purification and immobilization. This single module consisted of 28 amino acids (PEPPKEVVPEKKAPVAPPKKPEVPPVKV) derived from the human titin exon 172. This engineered protein, designated as [PPE172₁₅] contains 15 tandem PEVK repeats and is tagged at both ends with his₆ tag, S-protein tag, and biotin protein ligase (BirA) [32] on the N-terminus and a fluorescein arsenical hairpin binder (FlAsH) tag [33, 34] on the C-terminus of the polymer. The defined number of identical modules in [PPE172₁₅] reduces heterogeneity of its behavior and allows for more readily interpretable binding interaction measurements. The multiple affinity tags on [PPE172₁₅] can facilitate the investigation of force-related protein interactions and provide more readily interpretable data from single-molecule techniques such as atomic force microscopy (AFM).

The cloning and expression of a tandem repeat sequence in large quantities was technically challenging [35–40]. The tandem repeats in the DNA sequence coding for [PPE172₁₅] caused many problems in engineering the plasmids. The repetitive sequences with high GC content were difficult to digest by restriction enzymes and amplify by PCR reactions. The newly cloned plasmids were not stable; random recombination events and deletions of nucleotides readily occurred at all the stages of cloning. The polymer plasmids and expressed protein were also toxic to the bacteria. [PPE172₁₅] was expressed in truncated forms or with different numbers of modules. Fig. 1 illustrates the major steps and difficulties encountered in the process of cloning and expression of [PPE172₁₅]. We compared different methods of protein expression and overcame many technical challenges during the process of engineering the tandem repeat sequence. We further studied the binding and biochemical characteristics of the newly engineered polyprotein. Therefore, this study offers practical details and guidance on the purification and expression of tandem repeats and intrinsically disordered recombinant proteins.

Figure 1.

The process and challenges of engineering the [PPE172₁₅] protein

Materials and Methods

Materials

The Pichia expression system, BL21(DE3) pLysS and Stbl2 competent cells, and Gateway cloning technology were obtained from Invitrogen. His-tag monoclonal antibody, pET-30a vector, and S-protein agarose were purchased from Novagen. Rapid translation system (RTS 500) E. coli HY kit, and complete mini EDTA-free protease inhibitor cocktail tablet were purchased from Roche Molecular Biochemicals. Exo mung bean deletion kit and quikchange XL site-directed mutagenesis kit were purchased from Stratagene. HRP-rabbit anti-mouse IgG was obtained from Zymed Laboratories Inc. Wetware FlAsH protein labeling kit was purchased from Aurora Biosciences Corporation, San Diego. The nickel sepharose 6 fast flow column resin was purchased from GE Amersham Biosciences. The Affi-gel 10 was purchased from Bio-Rad.

Methods

In vivo expression and purification of the [PPE172₁₅]

E. coli expression system

The polyprotein vector containing the 15 monomers of titin exon 172 [PPE172₁₅] with FlAsH and BirA tags in pENTR11 entry vector was purified and digested with NotI and NcoI restriction endonucleases, and ligated into the pET-30a(+) protein expression vector containing the kanamycin-resistance marker. This expression vector, designated as pET30a-[PPE172₁₅], was transformed into E. coli BL21(DE3) pLysS cells for expression. Single colonies were inoculated into four 50 ml tubes of LB medium with 25 μg/ml chloramphenicol, 40 μg/ml kanamycin and 1% glucose, and then incubated at 30 °C overnight. The 50 ml cultures were transferred into 8 L of the same medium and incubated at 30 °C with a shaker at 200 rpm until the OD reached A₅₅₀ = 0.6–0.8. Isopropyl-β-D-1-thiogalactopyranoside (IPTG) at 1 mM was used to induce protein expression for 4 hours. The bacteria were collected by centrifugation at 6500 × g for 10 min and lysed in a French press cell (3 × 1500 psi) in lysis buffer (20 mM sodium phosphate, 0.3 M NaCl, 5 mM imidazole, 10 μg/ml leupeptin, and 1 complete EDTA-free protease inhibitor cocktail tablet per 50 ml buffer pH 8.0). The cell lysate was centrifuged at 45,000 × g for 15 min at 4 °C and the supernatant was loaded to a nickel-chelating sepharose column (Amersham Pharmacia Biotech) equilibrated with 20 mM sodium phosphate, 0.3 M NaCl pH 8.0. The column was washed with 10 volumes of column buffer and the [PPE172₁₅] was eluted with 0.25 M imidazole, 20 mM sodium phosphate, 0.3 M NaCl pH 8.0. The eluent was then applied to an S-protein agarose (Novagen) column in 20 mM tris-HCl, 150 mM NaCl pH 7.5 and the S-protein column was washed with 25 volumes of column buffer and eluted with 7 M guanidine chloride in the same column buffer. Due to the multiple cysteine amino acids on the FlAsH tag, the eluted [PPE172₁₅] was reduced with 40 times molar excess with tris (2-carboxyethyl) phosphine hydrochloride (TCEP) under nitrogen gas at 4 °C for 60 min. The reduced [PPE 172₁₅] was loaded onto an Affi-gel 501 organomercurial agarose column in 20 mM tris-HCl, 1 mM EDTA, 0.5 M NaCl pH 8.0. The column was washed with 25 column-volume buffer and eluted with 100 mM 1,4-dithiothreitol in the same buffer with pH 8.0. The eluted [PPE172₁₅] was quickly frozen in liquid nitrogen and stored at −80 °C. The total yield of purified protein was about 2 mg/liter of culture medium and was measured spectrophotometrically (extinction coefficient at 280 nm is 0.21 for 1 mg/ml).

Baculovirus expression system

The baculovirus expression was conducted by Kemp Biotechnologies Inc. (Frederick, MD). Briefly, the [PPE172₁₅] insert was excised from plasmid at the EcoRI and NdeI sites and ligated into the pVL1393 transfer vector with EcoRI and NotI sites to construct pVL1393-[PPE172₁₅] transfer vector. The pVL1393-[PPE172₁₅] plasmids were amplified in E. coli and further purified. The pVL1393-[PPE172₁₅] vectors were co-transfected with linearized DNA of Autographa californica nuclear polyhedrosis virus (AcNPV) in the Sf9 insect cells (Spodoptera frugiperda). The homologous recombinant virus was then amplified by infecting fresh Sf9 insect cells at 0.3% (v/v) and 3% (v/v) multiplicity of infection (MOI) for 0, 24, 48 and 72 hrs. The cell lysates were analyzed with SDS-PAGE gels to check the [PPE172₁₅] expression.

Cell-free expression and purification of polyprotein [PPE172₁₅]

Roche rapid translation system RTS-500 was used to express pET30a-[PPE172₁₅] plasmid. The feeding and reaction solutions were provided in the expression kit (Roche Applied Science) and prepared according to the product manual. Briefly, the reaction solution included E. coli lysate, amino acid mix, and reaction buffer with 15 μg pET30a-[PPE172₁₅] plasmid DNA. The feeding solution had feeding mix, amino acid mix, and methionine. One complete mini EDTA free protease inhibitor cocktail tablet (Roche Applied Science) was added to 11 ml feeding solution. The reaction and feeding solutions were loaded into reaction and feeding compartments, respectively. The reaction was carried out in the RTS Proteomaster instrument at 30 °C with 150 rpm for 17–24 hrs. The reaction solution was taken out of the compartment and centrifuged at 10,000 × g for 5 min at 4 °C. The supernatant was mixed with 0.5 ml nickel-nitrilotriacetic acid (Ni-NTA) agarose (Qiagen) in 50 mM sodium phosphate, 0.3 M NaCl pH 8.0. The sample mixture was rotated at 4 °C for 60 min and packed into an econo-column (Bio-Rad). The resin was washed with 20 columns volume of 50 mM sodium phosphate, 5 mM imidazole, 5 mM β-mercaptoethanol, 1.5 M NaCl pH 8.0, and 50 mM sodium phosphate, 0.3 M NaCl, 20 mM imidazole pH 8.0, respectively. The polyprotein was eluted with 0.25 M imidazole, 50 mM sodium phosphate, and 0.3 M NaCl pH 8.0. Pooled fractions were treated at eight times molar excess of protein concentration with TCEP under nitrogen gas at 4 °C for 60 min. The pool was applied to an organomercurial agarose Affi-gel 501column (1.0 × 5 cm; mercury content is 2.51μmol/ml) equilibrated with 20 mM tris-HCl, 0.5 M NaCl, 1 mM EDTA pH 8.0. The column was washed with 20 volumes of column buffer, and the polyprotein was eluted with 100 mM DTT in the same buffer pH 8.0. The fractions were quickly frozen in liquid nitrogen and stored in −80 °C. The total yield of RTS-500 per reaction was about 1 mg.

Yeast expression system and large scale expression

Pichia pastoris strains X-33 and GS115 were used as hosts for expression. The pPICZαA and pPICZ, which contains the zeocin resistance genes and alcohol oxidase (AOX1) promoters with and without α-factor secretion signal, were used as the expression vectors (Invitrogen). The pPICZαA and pPICZ vectors were digested with restriction endonuclease EcoRI and pET30a-[PPE172₁₅] was digested with NdeI. The 3′-overhang DNA created by enzyme digestion was blunt-ended with 1 U of Klenow in NEBuffer2. The blunt ends of the vectors were ligated and transformed into E. coli Top 10 for propagation. The linearized pPICZαA-[PPE172₁₅] (10 μg) and pPICZ were electroporated into Pichia pastoris X-33 and GS115 with a 0.2 cm cuvette containing 80 μl pre-washed yeast cells at a voltage of 2.0 kV using Bio-rad Micropulser pre-programmed “pic” setting. 1 ml of 1 M sorbitol was immediately added to cuvettes and incubated at 30 °C for 60 min. The mixture was spread onto YPDS plates 1% (w/v) yeast extract, 2% (w/v) peptone, 2% (w/v) dextrose, 1 M sorbitol, and 1.5% (w/v) bacteriological agar with 100 μg/ml zeocin and incubated at 30 °C for 3 days. The methanol utilization plus (Mut⁺) phenotype of the Pichia pastoris X-33 and GS115 were determined using minimal dextrose histidine (MDH) and minimal methanol histidine (MMH) plates. The Mut⁺ and Mut^s transformants were selected by picking single colonies from YPDS plates containing the zeo^R and patched onto MDH and MMH plates since Mut^s transformants do not produce alcohol oxidase and cannot metabolize methanol as a carbon source. After 2 days incubation at 30 °C, Pichia yeast cells grown normally on both YPDS and MMH plates were selected and used for the expression host cells.

The Pichia yeast cells X-33 with pPICZαA-[PPE172₁₅] were scaled up and expressed with the fermentation process in the 10 L bench-top fermentor vessel (New Brunswick BioFlo 3000). The components of the fermentation basal salts medium with PTM₁ trace salts were as described [41]. A single colony was picked from YPDS plates (100 μg/ml zeocin) and inoculated into 325 ml MGY medium with shaking at 250 rpm and at 30 °C for 24 hrs. The log phase medium was injected into 7 L basal salts medium with PTM₁. The OD 550_nm and the dissolved oxygen were monitored. 50% glycerol with 12 ml/L PTM₁was added at 1 ml/min for 72 hours until the cell mass reached 252 g/L wet weight. The protein expression was induced by adding methanol at 0.8 ml/min for 24 hrs. The pH was adjusted by adding 5 M potassium hydroxide or ammonium hydroxide during the fermentation process.

The [PPE172₁₅] was secreted into growth medium and separated from the yeast cells by centrifugation at 6500 × g for 10 min and then filtered through a 0.2 μm filter. The [PPE172₁₅] was precipitated with 50% ammonium sulfate at 4 °C for 30 min and centrifuged at 18,000 × g for 10 min. The pellet was suspended into 10 mM potassium phosphate pH 7.8 and dialyzed into the same buffer at 4 °C overnight. The [PPE172₁₅] was loaded onto a hydroxylapatite column equilibrated in 10 mM potassium phosphate pH 7.8. The column was washed with 10 volumes of the same buffer and the [PPE172₁₅] was eluted with 1 M potassium phosphate pH 7.8. The yield was over 1 gram/L. For the N¹⁵ labeled [PPE172₁₅], FM22 media [41] with 10 g/l (¹⁵NH₄)₂SO₇ (Cambridge Isotope Laboratories, Inc) was used and processed as described above.

The [PPE172₁₅] was digested with PNGase F (BioLabs) to check the glycosylation of the yeast expressed proteins according to the manufacturer’s instructions. Briefly, different amounts of the [PPE172₁₅] (10, 18, and 36 μg) were digested with 10 units PNGase F for 1 hour at 37 °C in G7 reaction buffer, respectively. The digestion mixtures were visualized with 4–12% Bis-Tris NuPAGE gel.

Affinity purification of the [PPE172₁₅] with organomercurial agarose gel

The Affi-gel 501 organomercurial agarose was prepared according to manufacturer’s protocol (Bio-Rad). Briefly, the Affi-gel 10 gel resin was washed with 3 volumes of anhydrous isopropyl alcohol and then reacted with 1.15% of p-aminophenylmercuric acetate in dimethylformamide for 4 hours at the ambient temperature. The unreacted succinimide groups were blocked with 0.8% (v/v) of ethanolamine HCl, pH 8.0 for 1 hour at the same temperature. The agarose gel was washed with 2 volumes of dimethylformamide and 6 volumes of anhydrous isopropyl alcohol. The organomercurial agarose gel resins were experimentally determined to have 8 mg/ml of bovine hemoglobin binding capacity with 2.51 μmoles Hg/ml.

Generating short segments of the [PPE172₁₅] DNA sequence for the plasmid sequencing

The [PPE172₁₅] plasmid was sequenced from both ends of the polymer insert. The pET30a-[PPE172₁₅] vector was first digested with Kpn1 and NcoI restriction endonuclease, then using the Exonuclease III to digest the 3† hydroxyl terminal of double strand DNA for 1 to 5 min. The reaction was stopped by deactivating the Exonuclease III at 68 °C for 15 min. The remaining undigested single strand DNA extensions were digested with Mung bean nuclease at 30 °C for 30 min. The deleted vectors at different time points were purified by 0.2% SDS, 25 mM tris-HCl, 0.8 M lithium chloride pH 8.8, and 2.5 parts of 1:1 phenol-chloroform. The mixture was spun at 16,000 × g for 5 min with an Eppendorf 5415C centrifuge, a final concentration of 0.2 M sodium acetate and 60% cold ethanol added to the supernatant, and the mixture placed on dry ice for 10 min. The mixture was then spun at 16,000 × g for 20 min. 80% ethanol was added to the pellet, spun again, and the pellet air dried. The purified DNA was suspended into 10 mM tris-HCl, 1 mM EDTA pH 8.0 and run on a 1% low melting point agarose gel. The different lengths of deleted pET30a-[PPE172₁₅] inserts were gel purified and ligated with T₄ ligase. The recirculated pET30a-[PPE172₁₅] vectors were transformed into E. coli Stbl2 competent cells and grown overnight on Luria-Bertani (LB) medium agar plates with kanamycin. The single colonies were picked and plasmids were purified by the miniprep procedure (Qiagen). The purified DNA was sequenced by GENEWIZ (South Plainfield, NJ).

Western blot and FlAsH labeling for the tag detection

Polyacrylamide gel electrophoresis was performed in 4–12% NuPAGE Bis-Tris gel with MES SDS running buffer (Invitrogen). The [PPE172₁₅] was transferred electrophoretically onto a PVDF membrane (Invitrogen) using a TE22 Mighty Small Transphor tank transfer unit (Amersham Biosciences) at 50 V constant voltage for 3 hours at 4 °C. NuPAGE transfer buffer with 10% methanol was used.

Western blotting was conducted by using monoclonal antibody against rabbit skeletal muscle titin (RT11) and his-tag monoclonal antibodies, and was detected by horseradish peroxidase-conjugated secondary antibodies (rabbit anti-mouse IgG, IgM and IgA (H + L)). Titin antibody RT11 reacted positively with titin exon 172 PEVK module in the [PPE172₁₅]. S-protein HRP conjugate was used for detection of S-tag on the expressed proteins. Affinity purified rabbit polyclonal antibody (IgG) against human nebulin SH3 C-terminus peptide (VQRTGRTGMLPANYVEC) [42] was used to detect nSH3.

The tetracysteine Tag (CCXXCC) was detected with the FlAsH reagent. The [PPE172₁₅] was treated with Laemmli sample buffer (30 mM tris-HCl, 1% SDS, 12.5% glycerol pH 6.8 plus 50 mM tris (2-carboxyethyl) phosphine hydrochloride (TCEP) at 99 °C for 3 min and cooled down before adding FlAsH dye (final concentration 1.5 μM). The reaction was overnight at 4 °C in the dark. The [PPE172₁₅] samples were separated in 4–12% Bis-Tris gel with MES SDS running buffer (Invitrogen), and scanned with Molecular Dynamics Storm 860 (excited at 450 nm with emission at 535 nm).

Circular dichroism spectroscopy

The circular dichroism (CD) spectra were recorded on a JASCO-J-600 spectropolarimeter (Easton, MD), and the detailed procedure has been described [28]. Briefly, yeast expressed [PPE172₁₅] at a concentration of 0.289 mg/ml (5.35 μM) in 20 mM potassium phosphate pH 7.0, was measured at 4, 25, 50, 75 °C, respectively in a cuvette with a 0.1 cm path length. Thermal titration of the [PPE172₁₅] was monitored by following the change in CD spectra at 201 nm between 4 °C to 75 °C.

Gel filtration chromatography

The elution behavior of E. coli expressed [PPE172₁₅] in several types of gel filtration chromatography was investigated. Gel filtration resins of Sephadex G-25, G-50, G75, Hitrap desalting column (GE Healthcare), and Bio-gel P4 polyacrylamide gel (Bio-Rad) were equilibrated in 20 mM tris-HCl, 0.5 M NaCl pH 7.5. The [PPE172₁₅] was dialyzed into the same buffer plus 2 mM TCEP and 1 mM EDTA. The [PPE172₁₅] was centrifuged at 10,000 × g for 20 min at 4 °C prior to loading.

The K_av of the [PPE172₁₅] was calculated according to Kav = (Ve−Vo)/(Vt−Vo) where Vo is the void volume, Vt is the total volume of the column, and Ve is the elution volume of [PPE172₁₅]. The high salt in the column buffer was used to minimize nonspecific ionic interactions between the [PPE172₁₅] and the column resins.

Limited digestion by Enterokinase

The yeast expressed [PPE172₁₅] was digested with recombinant Enterokinase (BioLabs) in 10 mM potassium phosphate, 150 mM KCl, 10 mM dithiothreitol (DTT) pH 7.0 with 0.004% (w/w) and 0.04% (w/w) protein to Enterokinase ratio. The reaction was carried out at room temperature at 1, 3, 6, and 10 hours, and was terminated with 25 mM phenylmethanesulfonylfluoride (PMSF). The digested [PPE172₁₅] was treated with Laemmli sample buffer with DTT and boiled in a water bath for 1 min. The protein samples were loaded onto NuPAGE 4–12% Bis-Tris gel and run in the NuPAGE MES SDS running buffer (Invitrogen) at 100V constant voltage for 1 hour.

Enzyme linked immunosorbent assay

The yeast expressed [PPE172₁₅] was adsorbed to microtiter plates (50 μl at 10 μg/ml in 20 mM sodium phosphate, 1 mM EDTA, 150 mM NaCl pH 7.0) overnight at 4 °C. The [PPE172₁₅] coated plates were washed three times with TBS-T and blocked with 0.2% bovine serum albumin in TBS-T for 1 hour at 4 °C and 25 °C, respectively. 50 μl nebulin SH3 (nSH3) from 0 to 80 μM in 20 mM potassium phosphate, 50 mM NaCl pH 7 was then added to the [PPE172₁₅] coated plates. The plates were then incubated for 1 hour at either 4 °C or 25 °C. After washing the plates three times with TBS-T, rabbit polyclonal anti-peptide antibodies to nebulin SH3 and mouse monoclonal anti-titin antibody (RT11) were incubated for 1 hour at 4 °C or 25 °C in 0.2% BSA in TBS-T. Plates were washed three times with TBS-T and then incubated with peroxidase-conjugated rabbit anti-mouse antibody or goat anti-rabbit antibody (Zymed, S. San Francisco, CA) in the same blocking buffer for 1 hour at the same temperature. The binding was detected by adding ABTS-H₂O₂ substrate in 100 mM citrate buffer pH 4.2 and absorbance measured at 405 nm (Molecular Dynamics Biolumin 960).

Surface plasmon resonance spectroscopy

The binding of nebulin SH3 (nSH3) to yeast expressed [PPE172₁₅] was studied by surface plasmon resonance with a Biacore 1000 (GE Healthcare) instrument. The [PPE172₁₅] in 10 mM potassium phosphate buffer pH 7.0 was immobilized onto CM5 chips (GE Healthcare). The surfaces were activated with 1:1 molar ratio of 0.4 M 1-ethyl-3-(3-dimethylpropyl)-carbodiimide to 0.1 M N-hydroxysuccinimide at flow rate 10 μl/min for 10 min. The [PPE172₁₅] was then immobilized onto the surface until there were no changes in the RU reading. The excess reactive sites on the surface were deactivated by injecting 1 M ethanolamine-HCl, pH 8.0 at10 μl/min for 7 min. Then nSH3 in running buffer (10 mM potassium phosphate, 150 mM KCl pH 7.0) was injected at 5 μl/min onto the surface of the chip at room temperature. The concentration was increased stepwise from 0.0025 to 3 mM; while the increase in resonance unit (RU) was monitored. The binding constant was calculated from the titration curve by subtracting the baseline RU from the RU at the different concentrations of nSH3.

The binding constants of the ELISA and SPR were calculated according to a 3 parameter single rectangular hyperbola regression using Sigmaplot 9.0. The equation of y = y₀ + (ax / b + x) was used, where b is the K_d.

Results

Engineering the polyprotein [PPE 172₁₅] vector

We engineered a recombinant polyprotein construct, the [PPE172₁₅], which consists of a defined number and order of identical modules. An 84-base pair DNA oligonucleotide from titin PEVK region at exon172 (5′-ccagagcctcccaaagaagtagttcctgaaaagaaagcgccagtggctcctcctaaaaagcctgaagtcccacctgttaaagtg) was synthesized, and used as a single monomer unit for [PPE172₁₅] vector construction. Synthetic DNA oligonucleotides were filled in with polymerase and cloned into pbluescript vector. The monomer sequence with EarI site was annealed specifically and seamlessly into the polymer sequence [43–45], which was further ligated into pbluescript via EarI sites. The ligation product was cloned into DH5α and colonies selected. The longest clone selected had a total of 15 monomers.

The FlAsH binding tetracysteine sequence (CCXXCC) [33] was cloned into pENTR11 (Invitrogen) entry vector between the end of 15 monomers and C-terminus stop codon. The biotin holoenzyme synthetase (BirA) sequence (AGLNDIFEAQKIEWHEG) [32] was cloned into pENTR11 entry vector at the N-terminus of the beginning of the 15 monomers. The EarI sites were added to pENTR11 via PCR using Pfu polymerase and methyl-dCTP. The polymer sequence and pENTR11 entry vector were ligated and transformed into E. coli Stbl2 cells. The constructed peptide sequence was as follows:

MHHHHHHSSGLVPRGSGMKETAAAKFERQHMDSPDLGTDDDDKAMAGLNDIFEAQKIEWHEG[PEPPKEVVPEKKAPVAPPKKPEVPPVKV]₁₅ PSWEAAAREACCRECCARA.

The newly engineered [PPE172₁₅] vector had fifteen 28-amino acid repeats, multiple tags, his₆ tag, S-tag, and BirA [32] biotinylation sequence, on the N-terminus, and a FlAsH tag [33, 34] on the C-terminus of the polymer as shown in the Fig. 2. The biotinylation BirA tag of the [PPE172₁₅] makes it possible to study protein interactions for a variety of commercially available reagents. The FlAsH tag is a short tetracysteine sequence (CCXXCC), and is capable of binding to a chemically modified fluorescent dye that fluoresces when bound to the peptide sequence. It is stable and versatile for both affinity purifications and detection studies. Therefore, these affinity tags on [PPE172₁₅] not only facilitate the purification of the protein from both ends to ensure a full-length protein is isolated, but also makes it possible to attach [PPE172₁₅] to tips and surfaces of AFM for single molecule techniques studies. More readily interpretable data can then be generated from these techniques.

Figure 2.

Schematic diagram of the engineered polyprotein [PPE172₁₅]

The [PPE172₁₅] recombinant protein expression

The challenges in the expression of a repetitive DNA sequence include toxicity of expressed proteins to expression host cells, proteolysis, random recombination events, and a very low yield of expressed protein. We expressed and purified both in vitro and in vivo bacteria systems, and viral expression systems as potential solutions to these problems. We successfully expressed [PPE172₁₅] in large quantities in a yeast expression system. We will address each expression system respectively; and combine the purification of the expressed [PPE172₁₅] for each host system in one section.

Prokaryotic hosts expression

Elastomeric proteins such as elastin, abductin, resilin, spider silk, and the streptococcal B1 immunoglobulin-binding domain of protein G (GB1) have been cloned and expressed in bacterial hosts [46–53]. Since the elastomeric protein contains repetitive sequences and intrinsically disordered regions, the problems of premature terminations and limited yields of the expressed proteins were encountered in these studies as well [54]. In order to find more suitable hosts for the repetitive sequence expression, we cloned the [PPE172₁₅] into pET-30a(+) expression vector and tested different bacterial hosts for [PPE172₁₅] expression. The pET-30a-[PPE172₁₅] plasmid was transformed into BL21-CodonPlus (DE3)-RP, BL21-SI, BLR(DE3) pLysS, RosettaBlue(DE3) pLysS, and BL21(DE3) pLysS. The [PPE172₁₅] was expressed in LB medium in these hosts. Although these BL21 derivatives contain extra tRNA genes such as argU and proL [55] to enhance expressions of toxic proteins with repetitive sequences, we could not detect expression of the [PPE172₁₅] in these hosts with western blots. In addition, the growth rates of bacterial cells with the [PPE172₁₅] plasmid were significantly slower than normal rates of cell growth.

Studies have shown that addition of glucose in the LB medium can decrease levels of basal expression of a target gene. It can also help to maintain plasmid stability, since it reduces the cAMP stimulation of T7 RNA polymerase [56, 57]. Therefore, we modified the expression environments by adding 1% glucose. The expression of the [PPE172₁₅] in the LB culture with 1% glucose was significantly increased upon 1 mM IPTG induction as shown in the Fig. 3. In addition, adding 1% glucose in LB agar plates before inoculation also increased the level of the [PPE172₁₅] expression in comparison with the LB agar plates without glucose.

Figure 3. Effect of glucose on expression level of [PPE172₁₅] in E. coli.

Addition of 1% glucose in LB medium increased the level of E. coli expressed [PPE172₁₅]. The [PPE172₁₅] was separated on a 4–12% Bis-Tris gel with MES SDS running buffer stained with Coomassie blue. Lane 1: Myofibril standard. Lanes 2: SeeBlue prestained protein standard. Lanes 3 and 5: E. coli cell lysates before 1 mM IPTG induction. Lanes 4 and 6: E. coli cell lysates after 1 mM IPTG induction. Lanes 3 and 4: LB medium with 1% glucose. Lanes 5 and 6: LB medium without 1% glucose. *The [PPE172₁₅].

In addition, we found that the toxicity of the [PPE172₁₅] to bacterial hosts was further reduced by decreasing the temperature during expression. When the temperature of the expression medium was reduced from 37 °C to 30 °C during cell growth, the bacteria grew better and expressed more [PPE172₁₅]. At each stage of expression, we controlled the growth of the culture cells within the logarithmic phase. We found that mid-log phase induction of expression increased the quantity of the expressed [PPE172₁₅].

We successfully expressed the full-length [PPE172₁₅] in E. coli BL21(DE3) pLysS hosts (Fig. 4). We further purified the [PPE172₁₅] with affinity NiNTA and organomercurial agarose columns sequentially. The total yield of purified [PPE172₁₅] was approximately 2 mg/liter of culture medium.

Figure 4. Expression and purification of the [PPE172₁₅] from E. coli hosts.

Analysis of [PPE172₁₅] on a 4–12% NuPAGE Bis-tris gel. (A) Gel stained with Coomassie blue. (B) The C-terminal tetracysteine sequence (CCXXCC) of [PPE172₁₅] was labeled with FlAsH reagent. Lane 1: Invitrogen SeeBlue plus2 prestained standard. Lane 2: Nebulin SH3. Lane 3: NA4 (nebulin recombinant fragments) [102]. Lane 4: [PPE172₁₅]. *The bright band of lane 1 in B is myoglobin-red in Invitrogen SeeBlue plus 2 standards, which can be detected at 535 nm by the instrument.

Baculoviral expression

Insect cell expression systems have many advantages over other expression systems. They can produce large quantities of recombinant proteins in a short period of time. Since expressed recombinant proteins in insect cells can be posttranslationally modified, they can have increased biological activity and antigenicity [58]. Insect cells also have chaperonins that are required for the proper folding of some proteins and are absent in prokaryotic cells. Recombinant proteins with repetitive sequences such as human type III collagens were expressed successfully in an insect system [59, 60]. Collagens resemble the characteristics of the [PPE172₁₅] sequence, since they consist of many repeating triplet sequences enriched in proline amino acids. Therefore, the [PPE172₁₅] was further investigated in the viral expression system. The [PPE172₁₅] DNA sequence was cloned into a pVl1393 expression vector. The pVL1393-[PPE172₁₅] vectors were then recombined with the Autographa californica nuclear polyhedrosis virus (AcNPV) genome. Spodoptera frugiperda Sf9 insect cells were then infected, and the [PPE172₁₅] was expressed and monitored by the gel electrophoresis (Fig. 5A).

Figure 5. Expression and purification of the [PPE172₁₅] in different hosts.

A. Western blots of the baculovirus expressed [PPE172₁₅] using anti-his antibodies. Lane 1: His Std: Qiagen 6× His protein ladder. Lane 2 & 3: The [PPE172₁₅] expressed in BL21(DE3)pLysS cells. Lanes 4, 5, 6, 7: The baculovirus expressed [PPE172₁₅] at multiplicity of infection (MOI) of 0.3% v/v at 0, 24, 48, 72 hours, respectively. Lanes 8, 9, 10, 11: The MOI of 3% expressed at 0, 24, 48, 72 hours, respectively.

B. The prokaryotic cell-free translation system expressed [PPE172₁₅] was separated on a 4–20% Tris-glycine gel with silver staining. Lane 1: Invitrogen SeeBlue plus2 pre-stained standard. Lane 2: The [PPE172₁₅] purified from organomercurial agarose columns. Lane 3: [PPE172₁₅] purified from NiNTA columns. Lane 4: Western blots for [PPE172₁₅] with anti-titin antibodies. Lane 5: Western blots for [PPE172₁₅] using anti-His antibodies.

C. Analysis of the yeast expressed [PPE172₁₅] with 4–12% NuPAGE Bis-Tris gel stained with Coomassie blue stain. Lane 1: TP1 (recombinant titin PEVK fragment) [22]. Lane 2: Invitrogen SeeBlue plus2 prestained marker. Lanes 3 & 4: [PPE172₁₅] at different amounts of loading.

D. Screening of [PPE172₁₅] expression in the Pichia pastoris yeast system. The pPICZα-[PPE172₁₅] and pPICZ-[PPE172₁₅] expression vectors were electroporated into yeast X-33 and GS115 cells. The transformants were screened for the expression of the [PPE172₁₅]. The yeast cell pellets and supernatants were analyzed with 4–12% NuPAGE Bis-Tris gel stained with Coomassie blue stain. Lanes 1 & 2: The transformants of X-33 with pPICZα-[PPE172₁₅] vectors. Lane 3: Invitrogen SeeBlue plus2 prestained standard. Lane 4: TP1. Lane 5: E. coli expressed [PPE172₁₅]. Lane 6 & 7: The transformants of GS115 cells with pPICZα-[PPE172₁₅]. Lane 8 & 9: The transformants of GS115 cells with pPICZ-[PPE172₁₅] vectors.

E. Digestion of the yeast expressed [PPE172₁₅] with PNGase F to check for glycosylation of the proteins. Lane 1: Invitrogen SeeBlue plus2 prestained marker. Lanes 2, 4, 6: The digested [PPE172₁₅] at 10, 18, 36 μg of PNGase. Lanes 3, 5, 7: Nondigested [PPE172₁₅]. Lanes 8 & 9: Digested and nondigested rabbit IgG at 20 μg, respectively. Lane 10: CandyCane glycoprotein molecular weight standard (Molecular Probes).

In the baculoviral expression, there was one major [PPE172₁₅] band detected at 3% multiplicity of infection (MOI) at 24 hours from western blots. The number of [PPE172₁₅] related bands increased with the expression time, which was revealed by western blots with specific monoclonal antibody (RT11) to titin PEVK region. At 48 and 72 hours, numerous forms of [PPE172₁₅] were expressed. In the baculovirus system, [PPE172₁₅] showed an array of distinguishable bands (Fig. 5A). The mass difference between each band of the [PPE172₁₅] varied from 3000 to 7000 Da. Many clearly defined lower molecular weight [PPE172₁₅] bands were detected on the western blots. Most interestingly, we detected multiple bands of polyprotein higher than the molecular weight of [PPE172₁₅]. The amount of expressed [PPE172₁₅] with higher molecular weights increased after 24 hours of expression in the viral system. The creation of new and larger forms of [PPE172₁₅] in the viral expression system may be caused by recombination events during the process of protein expression or DNA transcription. The recombination events of the polyprotein seemed to follow the modular sizes of the [PPE172₁₅]. The molecular weight of each band of the polyprotein was changed roughly in multiples of the single module of [PPE172₁₅].

Repetitive nucleotide sequences in the eukaryotic genome are a very common feature [61, 62]. During DNA replication processes, the single-stranded DNA regions with repetitive sequences can spontaneously insert or delete some tandem repeats. Therefore, it causes mismatches of the tandem repeats, and results in recombinant of DNA sequences [63]. These mismatches and recombination of DNA structures cause a newly synthesized DNA with a different length and various numbers of repeats [64, 65]. The aberrant forms of [PPE172₁₅] in viral expression may be the result of recombination events during expression.

Cell-free expression system

Directly coupled transcription and translation systems provide an easy and fast way to produce many recombinant proteins in a cell-free system. Therefore, the pET-30a-[PPE172₁₅] plasmid was expressed in vitro by an E. coli coupled transcription and translation system according to the product manual (Roche rapid translation system). Briefly, 15 μg pET30a-[PPE172₁₅] plasmid DNA was added to the reaction solution, one complete mini EDTA free protease inhibitor cocktail tablet was added to the feeding solution. Both solutions were loaded to the designated compartment respectively. The reaction was carried out in the RTS Proteomaster instrument at 30 °C with 150 rpm for 17–24 hrs. The expressed [PPE172₁₅] was then purified with NiNTA and organomercurial agarose columns. As shown in Fig. 5B, there were significant problems with premature termination and/or proteolysis of the [PPE172₁₅]. [PPE172₁₅] was expressed in a ladder pattern. Increasing the concentration of protease inhibitors in the reaction mixture and varying the reaction time did not significantly prevent the formation of truncated [PPE172₁₅] during expression.

The truncation of [PPE172₁₅] in the cell-free protein expression could be caused by translational error [66], since repetitive DNA sequences can cause ribosome frameshifts during the translation process [67]. The frequency of frameshifts increases significantly if there is a limited amount of transfer ribonucleic acids. Since 29.3% of the amino acid composition in the [PPE172₁₅] sequence was proline, an insufficient supply of proline aminoacyl-transferase may have increased the frequency of ribosome frameshifts. Since the proline codon is used less frequently by bacterial hosts, the difficulty in [PPE172₁₅] expression may have been exacerbated [68, 69]. In addition, truncated forms of [PPE172₁₅] can be caused by proteolysis during the expression process, due to the inherent sensitivity of intrinsically disordered proteins to proteolysis.

Pichia pastoris yeast expression

The yeast system, as a eukaryotic host, has an advantage over bacterial systems in terms of protein expression and folding. Since there were 15 identical modules in [PPE172₁₅], the expression host cells were placed under extreme stress by the demand of the repetitive sequence during transcription and translation. This exacerbated the potential for translational errors [70]. We compared the demands of codon usage in the [PPE172₁₅] sequence to the usage in E. coli and yeasts. The percentage of the codon usage for yeast cells was relatively close to the requirement of the four major amino acids (PEVK) in the sequence of [PPE172₁₅]. Therefore, the yeast system was a more suitable host for the expression of [PPE172₁₅] [55, 71]. In addition, Pichia pastoris transformed with pPICZα expression vectors can transport expressed proteins with the secretion signal into the expression medium via secretory pathways. This mechanism could further reduce the stress imposed by the tandem repeat sequences on the expression hosts. Although recombinant proteins with repetitive sequences have been expressed in the yeast Pichia [72, 73], the yields of expressed protein vary for the secretory pathway in the yeast system [74].

Therefore, we investigated the expression of [PPE172₁₅] in the yeast Pichia system. We cloned the [PPE172₁₅] into pPICZα and pPICZ expression vectors, and electroporated into Pichia x-33 and GS115 cells. Numerous transformants for each expression vector and cell types were screened and selected for the expression of the [PPE172₁₅]. As indicated in Fig. 5D, the transformants in x-33 cells with pPICZa vector were the most successful clones in expression of the [PPE172₁₅]. We expressed the [PPE172₁₅] in the Pichia X-33 cells with pPICZαA-[PPE172₁₅] in YPD medium at 30 °C. The expressed [PPE172₁₅] was further sequentially purified by NiNTA and organomercurial agarose columns. Fig. 5C indicated the successful purification of the [PPE172₁₅] in the Pichia system. The yeast expression system produced large quantities of the expressed polyprotein via secretion pathway. It overcame the difficulties associated with the expression of repetitive regions in recombinant polyprotein. Production of [PPE172₁₅] was increased significantly. The yield was over 1 gram/L. We expanded the expression of [PPE172₁₅] in a bench top fermentor with FM22 medium, which made large quantities of [PPE172₁₅] available for biophysical studies. We further expressed [PPE172₁₅] with the stable isotype nitrogen-15 in minimal media for NMR studies. In addition, we digested the different concentrations of the [PPE172₁₅] with endoglycosidases, PNGase F, which removes all attached N-linked carbohydrate such as mannose from the proteins. The results, Fig. 5E indicated that there was no significant glycosylation of the expressed [PPE172₁₅].

Purification of the expressed [PPE172₁₅]

The [PPE172₁₅] was engineered to have 15 identical modules with multiple tags at both ends of the polymer. With these engineered tags, studies of [PPE172₁₅] with single molecule technique such as AFM and Laser Tweezers can generate more interpretable data to understand the mechanical function of the PEVK region of the titin.

In order to attach both C- and N-termini of [PPE172₁₅] to surfaces and AFM tips for nanomechanical measurements, tags at both ends must be present. The full-length nature of [PPE172₁₅] expressed by each host system was confirmed and purified by affinity purification. Due to the high frequency of recombination events during cloning and expression, we purified both E. coli, cell-free, and yeast expressed [PPE172₁₅] first with a Ni-NTA column to catch the N-terminus His affinity tags, and then with an organomercurial agarose column to catch the C-terminus tetracysteine sequence (CCXXCC) FlAsH tags. The purified proteins were thus highly enriched in full-length [PPE172₁₅]. The quality of the purified [PPE172₁₅] was routinely checked by labeling with FlAsH dye prior to SDS gel analysis to confirm the presence of the tetracystein sequence at the C-terminus. In addition, [PPE172₁₅] was further checked by western blotting with anti-His tag and RT11 antibodies. The RT11 antibody is monoclonal and targeted directly to the PEVK module in the full-length [PPE172₁₅] [22]. Results from the purification and labeling confirmed the full-length expression of the [PPE172₁₅] in Fig. 4 for E. coli expression and in Fig. 5B for cell-free express and Fig. 5C for yeast expression.

The purified full-length bacterial and yeast expressed [PPE172₁₅] was further confirmed by LC-MALDI-TOF mass spectrometry with the molecular mass of 54,180 Da. The mass value in the mass spectrometry was slightly higher than the calculated [PPE172₁₅] molecular weight according to the amino acids sequence. The difference could be caused by attachments of residual metal ions in the buffer or matrix to the highly charged [PPE172₁₅] [75–77]. The extinction coefficient was calculated (http://expasy.org/proteomics/protein_characterisation_and_function) at 280 nm as 0.21 for 1 mg/ml, due entirely to the end tags and the absence of aromatic amino acids in the PEVK module.

For conformational studies, the [PPE172₁₅] was subject to gel filtration chromatography. We observed that the E. coli and yeast expressed [PPE172₁₅] appeared to interact significantly and variably with the gel filtration matrix under the experimental conditions. As shown in Table 1, the K_av varied with the gel filtration media even under a strongly reducing condition to keep the tetracysteines from cross linking [PPE172₁₅] into aggregates. In the Sephadex gels, the K_av was about 2.5 times larger than the K_av of bovine serum albumins (BSA, globular protein at 68 kDa). Fortunately, its K_av on a desalting column (HiTrap, GE Health Sciences) was as small as 0.077. [PPE172₁₅] was eluted near the void volume of the column as expected in the desalting column.

Table 1.

The partition coefficients (Kav) of E. coli expressed [PPE172₁₅] were determined by gel filtration chromatography with different types of column resins

Gel Filtration Column Resins	The [PPE17215]	TCEPa
HiTrap Desalting	0.077	0.615
Sephadex G-25	1.3	2.1
Sephadex G-50	1.34	2.36
Sephadex G-75	1.22	2.5
Bio-Gel P-4	2.09	2.63

^aTris (2-carboxyethyl) phosphine hydrochloride (TCEP) is a reducing agent with molecular weight of 250.2. We used the small TCEP as a control for column volumes in the gel filtration columns. The average equilibrium K_av is 0.54 in SP-sepharose for BSA and ovalbumin [101]. [PPE172₁₅] has a much higher K_av than globular proteins.

In general, the affinity of [PPE172₁₅] toward the gel filtration media still allowed for effective purification, but rendered an uncertain estimation of hydrodynamic property (Stokes radius). Since [PPE172₁₅] is intrinsically disordered with unpredictable shape and flexibility, it was not useful to estimate its molecular weight or the oligomeric state on gel filtration columns. Specifically, there were no suitable calibration markers, with similar shapes and flexibility, that correlate stokes radius to the molecular weight of intrinsically disordered proteins. The interactions of the [PPE172₁₅] with gel filtration media further compounded the problem. Similar problems were also evident on gel electrophoresis. Since the [PPE172₁₅] migrated much slower in the gels, and thus a higher molecular weight than its calculated molecular weight (54,000 Da based on protein sequence) was observed in gel electrophoresis system. This abnormal mobility might be caused by unusual SDS binding, the abnormal conformation of SDS/protein complexes at any given SDS-protein ratio and/or potential interaction of the complex with the gel material.

Plasmid sequencing

We further confirmed the sequence of the [PPE172₁₅] at the DNA level. Since the total length of the [PPE172₁₅] DNA sequence was approximately 1500 base pairs (bp), and standard DNA sequence reactions give a reliable reading for only 400–500 base pairs, it was difficult to confirm the accuracy of the [PPE172₁₅] sequence. Therefore, we generated a set of nested deletion inserts from the pET-30a-[PPE172₁₅] plasmid. The [PPE172₁₅] plasmid was sequentially digested with Kpn1, NcoI, and Exonuclease III sequentially. The remaining undigested single strand DNA extensions were further digested with Mung bean nuclease. Three DNA segments with 200 bp, 850 bp, and 1200 bp were created from the [PPE172₁₅] plasmids. These new DNA segments were gel-purified and ligated with T₄ ligase. The recirculated [PPE172₁₅] plasmids were transformed into Stbl2 cells. The plasmids were purified and sequenced. The results of the sequences for the set of deletion inserts were aligned with each other using the website http://www.ncbi.nlm.nih.gov/blast. A 1.5 kb [PPE172₁₅] sequence was confirmed with correct reading frames and stop codons for expression.

Disordered in the [PPE172₁₅] conformation

The intrinsically disordered structures of [PPE172₁₅] were similar to that of the PEVK modules of titin, as measured by near UV circular dichroism spectroscopy (CD) and its thermal responses. As shown in Fig. 6, the strong negative ellipticity minimum at 201 nm and a shoulder near 220 nm in the CD spectra indicated a polyproline II (PPII) left-handed helix structure of the [PPE172₁₅]. The CD spectra of the [PPE172₁₅] measured over the range of 2 °C to 75 °C and showed a well-defined isodichroic point between 210 nm and 212 nm, which was completely reversible when it was cooled. Thermal melting curves were monitored by continuous measurements of the CD spectral value at 201 nm from 4 °C to 75 °C. The melting curve consisted of two linear lines that intercepted at 58 °C, with no sigmoidal melting characteristic of globular proteins (data not shown). The CD studies of the [PPE172₁₅] were indicative of an ensemble of malleable conformations of the PEVK modules, including PPII helices that interconvert to other secondary structures gradually with the rising temperature [28, 31].

Figure 6. Circular dichroism of the intrinsically disordered [PPE172₁₅].

The circular dichroism (CD) spectra of the yeast expressed [PPE172₁₅]. The CD spectra were obtained at 4 °C, 25 °C, 50 °C, and 75 °C. Each spectrum was scanned eight times at 20 nm/min with a time constant of 1 s. The [PPE172₁₅] was in 20 mM potassium phosphate pH 7.0.

Limited digestion of [PPE172₁₅]

Intrinsically disordered proteins are generally prone to proteolytic digestion due to their rather open structure that exposes proteolytic sites. To investigate the polymeric structure, we subjected yeast expressed [PPE172₁₅] to limited digestion conditions with Enterokinase that cuts mainly between glutamic acid (E) and lysine (K) residues. The [PPE172₁₅] was digested with 0.004% (w/w) and 0.04% (w/w) protein to Enterokinase ratio. The time course of digestion was followed by SDS gels. As shown in Fig. 7, the digested [PPE172₁₅] at different enzyme concentrations displayed a ladder of bands with increasing mobility on the gel electrophoresis. The spacing between these bands was estimated to be 3 kDa, corresponding to the size of one PEVK module. This was confirmed by LC-MALDI-TOF/TOF mass spectrometry. The ladder pattern of 15 bands of progressively digested PEVK module confirmed the 15-mer structure of [PPE172₁₅]. The pattern of the digested [PPE172₁₅] was reminiscent of those observed during the expression of the [PPE172₁₅] in the baculovirus system.

Figure 7. Proteolysis of tandem repeats of the [PPE172₁₅].

Yeast expressed [PPE172₁₅] was digested with Enterokinase in 10 mM potassium phosphate, 150 mM KCl, 10 mM DTT pH 7.0 with 0.004% (w/w) and 0.04% (w/w) protein to Enterokinase ratio. The digested [PPE172₁₅] was loaded onto NuPAGE 4–12% Bis-Tris gel and run in the NuPAGE MES SDS running buffer (Invitrogen). Lane 1: SeeBlue Plus2 molecular weight marker (Invitrogen). Lane 6: Undigested [PPE172₁₅]. Lanes 2 & 7: 1 hour digestion. Lanes 3 & 8: 3 hours digestion. Lanes 4 & 9: 6 hours digestion. Lanes 5 & 10: 10 hours digestion.

[PPE172₁₅] binds multiple nebulin SH3 with modest affinity

Our previous studies indicated that many titin PEVK modules contain a single SH3 binding motif [17]. SH3 is a 60-amino acid recognition domain, and it serves important regulatory functions by binding to proline rich motifs of signaling proteins [78]. Since [PPE172₁₅] was expected to contain 15 tandem SH3 binding sites, we investigated this interaction by both ELISA and surface plasmon resonance (SPR) methods. In ELISA, the surface adsorbed yeast expressed [PPE172₁₅] bound to the soluble nSH3 with an isotherm. The binding curves were fitted with a dissociation constant (K_d) of 7 μM at 4 °C (Fig. 8A) and 23 μM at 25 °C (Fig. 8B). The K_d was consistent with the affinity constants of the binding of nSH3 to a single PEVK module in the solution as studied by the fluorescence enhancement of nSH3 [31]. The increase in affinity at the lower temperature may result from an increase in PPII content that formed at the nSH3 binding site. Anti-titin monoclonal antibody, RT11, showed a strong binding to [PPE172₁₅] with a K_d of 0.47 μM at 4 °C and 0.3 μM at 25 °C. Interestingly, titin RT11 inhibited the binding of nSH3 to [PPE172₁₅] in the ELISA. The K_d decreased to 9.69 μM at 4 °C and 471 μM at 25 °C in the presence of 50 μg/ml of RT11. These data suggested that the RT11 epitope is at the nSH3 binding motif of the PEVK module and would be a useful probe of the interplay of PEVK and SH3 signaling pathways.

Figure 8. Interactions of the [PPE172₁₅] with nebulin SH3.

A& B: ELISA study. Yeast expressed [PPE172₁₅] was adsorbed to a microtiter plate (10 μg/ml in 20 mM sodium phosphate, 1 mM EDTA, 150 mM NaCl pH 7.0) and incubated with nebulin SH3 (20 mM potassium phosphate, 50 mM NaCl pH 7.0) at 4 °C and 25 °C. Binding was detected with rabbit polyclonal anti-peptide antibodies to nebulin SH3. The dissociated binding constants were 7 μM at 4 °C and 23 μM at 25 °C.

C: Surface plasmon resonance (SPR) study. Yeast expressed [PPE172₁₅] in 10 mM potassium phosphate buffer pH 7.0 was immobilized onto CM5 chips with 0.4 M of 1-ethly-3-(3-dimethylpropyl)-carbodiimide and 0.1 M N-hydroxysuccinimide. 1 M ethanolamine-HCl was used to deactivate excess active groups on the surface. Nebulin SH3 (nSH3) in running buffer (10 mM potassium phosphate buffer, 150 mM KCl pH 7.0) was injected from 0.0025 to 3 mM into the surface of the chip at room temperature.

The interactions between the immobilized yeast expressed [PPE172₁₅] and soluble nSH3 were studied by SPR. The [PPE172₁₅] was directly immobilized onto sensor chips and the binding of nSH3 to the [PPE172₁₅] chip was studied by continuously injecting nSH3 into the [PPE172₁₅] immobilized chip surfaces at different concentrations without regeneration. The K_d of the [PPE172₁₅] was calculated as 340 μM at room temperature by assuming 15 independent sites (Fig. 8C). The discrepancies in the binding constants between the SPR (340 μM) and the ELISA (23 μM) may be caused by the differences in the data treatment and/or the negatively charged surface of the CM5 chip that was lower in affinity. Furthermore, the binding of nSH3 in a dynamic flow system may be affected by on/off kinetics of the gel layer on the chip surfaces; it could further affect the binding capability of nSH3.

Discussion

Successful cloning and expression of repetitive and intrinsically disordered sequences of recombinant proteins has become increasingly important for two reasons. First, recent advances in tissue engineering and regenerative medicine have created a pressing need for effective biopolymers to generate supporting scaffolds. The future of regenerative medicine requires these scaffolds for cells to adhere and grow into functional units. The increasing demand for a suitable biopolymer will promote interest in engineering tandem repeat recombinant polyproteins [79–86]. Second, the availability of genomic sequences in various organisms makes many new open reading frames interesting for further study of their functions [87, 88]. Many of these sequences are with repetitive and intrinsically disordered regions [89–91]. Therefore, successful cloning and expression of these regions are the first steps for examining their biological roles [38, 92]. The knowledge and ability to express and purify a repetitive and intrinsically disordered recombinant protein such as [PPE172₁₅] will be useful for future research.

Due to the repetitive sequences in the [PPE172₁₅] DNA, random recombination events and deletions of nucleotides readily occurred during the cloning process. We tested several strains of E. coli hosts and found that max efficiency Stbl2 with recA1 was the best host for the unstable DNA plasmid. The newly constructed DNA plasmids containing the repetitive sequences were very unstable. The DNA plasmids in the transformed E. coli cells had to be instantly stored at −20 °C. At each stage of the plasmid propagation, it was critical to avoid reaching the saturation stage in LB growth medium and to store the plasmid in E. coli collected during the logarithmic phase of the growth curve. These careful storage steps could reduce the instability of DNA plasmid throughout the process. Additionally, secondary structures of DNA and RNA interfered with each step of DNA manipulation due to the repetitive sequences. For example, it was difficult to cut the [PPE172₁₅] DNA near the EarI restriction enzyme sites. We used Eam1101I enzyme as an isoschizomer to EarI, since they have the same recognition sequence. We prioritized the order of enzyme digestions by linearizing the [PPE172₁₅] plasmids with other enzymes before the EarI digestion. Fresh EarI restriction enzymes were also added to the reaction mixture multiple times during the digestions process. These steps helped to overcome the difficulties in engineering the [PPE172₁₅] DNA polymer.

In this study, use of the Pichia yeast expression host via the secretion pathway was most suitable for protein expression of the tandem repeat sequence [70, 93–96]. Codon usage bias in Pichia and E. coli could contribute to the success of the expression of the [PPE172₁₅] in the yeast system. The relative synonymous codon usage (RSCU) value is a useful tool to assess the adaptation of expression hosts to the demand of the expression sequence. For instance, the RSCU value for three proline codons used in [PPE172₁₅] was 3.975 for yeast, but only 0.711 for bacteria [71, 97, 98]. This indicates that yeast is a better host system for protein expression in this study. The effective number of codon (ENC) is a good indicator for the usage of the codon for the viral expression system. For example, proline has an average ENC of 25, which is considered as a strong codon bias [99]. Therefore, it may not be a good host for expression of [PPE172₁₅].

Nutrient composition in the medium is an important factor in the process of recombinant protein expression. Studies have shown that the formulation of growth medium can alter how bacterial cells utilize nutrients via different fermentation processes during the growth [100]. In this study, we changed the expression environment for E. coli expression hosts by adding glucose to attenuate the bacterial expression process, and by lowering medium temperature to increase the [PPE172₁₅] expression. These changes can decelerate the translation assembling process; decrease the speed of biosynthesis by the expression hosts, and therefore increased expression of the recombinant polyprotein.

This study represents the successful expression of a recombinant polyprotein with tandem repeats and intrinsically disordered regions. The technical challenges that we overcame in this work can aid future research by providing information about pitfalls during the design and engineering of tandem repeat DNA sequences.

Highlights.

• Successfully engineered titin PEVK polyproteins for nanomechanical measurements.

• Investigated systematically expression hosts for proteins with tandem repeats.

• Expressed and purified gram quantities of secreted polyproteins in a yeast system.

• Facilitated studies of the role of force in titin’s function.