Highly efficient soluble expression, purification and characterization of recombinant Aβ42 from Escherichia coli

Aggregation of amyloid-β protein (Aβ) is hypothesized to be a seminal neuropathological event in Alzheimer's disease (AD). Recombinant expression and purification of Aβ represents a common basis for investigating the molecular mechanisms of amyloid formation and toxicity. Herein, we report a novel high-yield expression and purification method for Aβ42 based on fusion with maltose binding protein (MBP) followed by the soluble polypeptide linker (NANP)3 and a modified tobacco etch virus (TEV) cleavage site before the Aβ42. We obtained a final yield of ∼18 mg L−1 of recombinant Aβ42 that was confirmed by SDS-PAGE, protein immunoblotting and MALDI-TOF. Finally, thioflavin T fluorescence and atomic force microscopy revealed that the recombinant Aβ42 aggregated into long, branched fibrils. Furthermore, the aggregates of the recombinant peptide had a strong cytotoxic effect on PC12 cells. The method described here can therefore be used to efficiently express the soluble fusion protein MBP-Aβ42 and obtain high-purity Aβ42 peptide, which can be used to understand the molecular mechanism of Aβ42 fibrillization and screen new candidate drugs for AD.


Introduction
Aggregation of amyloid proteins into insoluble brils can cause many neurodegenerative diseases, including Alzheimer's (AD) and Parkinson's disease, but also type-II diabetes. 1 AD is one of the neurodegenerative diseases that presently severely affect tens of millions of elderly patients around the world, causing a great economic burden to families and society. 2 According to the data published in 2016 by the Alzheimer's Association International Conference, about 46.8 million people in the world were living with AD in 2015, and this number is doubling every 20 years. Moreover, the incidence rate of AD is even increasing, so that it is estimated that in 2050 there will be more than 130 million AD patients worldwide. 3 It was estimated that the cost of medical care for the global population of AD patients in 2015 was about $818 billion, about 1.1% of global GDP. 4 Therefore, exploring the etiology of AD and nding anti-AD drugs has become the focus of researchers all over the world. 5 One of the pathological features of AD is the accumulation of extracellular senile plaques in the patients' brains. 6 The major components of neuritis plaques in AD patients' brains are various forms of aggregates of amyloid-b protein (Ab). Ab peptides containing 39-43 amino acids are produced by the sequential hydrolysis of amyloid precursor protein by both band g-secretases. 6 Among them, Ab40 and Ab42 are the two most common components. 7 Since the C-terminus of Ab42 contains two vicinal hydrophobic amino acids, Ile-Ala, Ab42 aggregates rapidly, and the resulting aggregates are the most cytotoxic species. [8][9][10] Therefore, Ab42 has become a focus of research.
Currently, Ab42 used in research is mainly produced by solid-phase chemical synthesis. 11,12 Although this method is efficient, rapid and inexpensive, the variable amounts of intrinsic impurities including residual amino acids, fragments and reagents from the solid-phase synthesis process seriously affect the aggregation properties and the corresponding toxicity of Ab42. 13,14 For example, Adams et al. identied the Ab42D39 species as the major peptide contaminant responsible for limiting both cytotoxicity and brillation of chemically synthesized Ab42. 15 In addition, it is difficult to purify Ab42 from neuronal cells in large amounts due to its high intrinsic hydrophobicity, proneness to aggregation, and extremely low abundance. 16 Consequently, heterologous expression has become the main method to produce Ab42. Some studies have used genetic engineering methods and subsequent expression and purication steps to obtain various types of recombinant amyloid-b peptides including Ab(M1-40) and Ab(M1-42), 17 as well as 15 N-and 13 C-isotope labeled Ab40 (ref. 18) and 15 N isotope-labeled Ab42 (ref. 19 and 20) in E. coli. Moreover, preparative SDS-PAGE was also used to purify the recombinant Ab42 and pGlu-Ab42. 21 In addition to E. coli, P. pastoris was also used to express Ab42. 22 However, the disadvantages of low yield, complex purication steps and insufficient purity of Ab are still present. Moreover, there are some intrinsic shortcomings, such as the addition of a methionine to the peptide and difficult purication. Moreover, prolonged fermentation protocols may result in the accumulation of large amounts of Ab42 brils, which is not conducive to the later purication and aggregation analysis.
Maltose-binding protein (MBP) is a commonly used fusion tag that can improve the yield and soluble expression of target proteins, and has been used in the recombinant expression of several hydrophobic proteins. 23,24 Because of the strong affinity of MBP for amylose, the MBP-fused protein can be specically puried over amylose resin to 70-90% purity. 25 In order to obtain pure Ab42, a protease cutting site is needed between the fusion tag and Ab42. Proteases in common use include tobacco etch virus protease (TEV), 20 thrombin, 26 and enterokinase 27 et al. The TEV protease is widely used in the separation of fusion proteins, in specic protein markers for genomics or proteomics, and segregation analysis. 20,28 However, aer cleavage with TEV, the recombinant Ab peptide would normally carry an additional residue from the cutting site, which may have an inuence on its aggregation properties. 28 In this work, a generally applicable, efficient protocol for the production of recombinant Ab42 without any extra residues with MBP as a fusion protein in E. coli, and an efficient and simple procedure to obtain pure Ab42 was developed. The clone expresses an Ab42 fusion protein containing an MBP tag, the rigid and soluble (NANP) 3 linker, a modied recognition site for TEV protease which changes the ENLYFQG sequence to ENLYFQ, and full-length Ab42. Aer identication via western blot analysis and MALDI-TOF, the target protein was proved to be Ab42. We further characterized the Ab42 by thioavin T (ThT) binding assays, AFM and cytotoxicity experiments, demonstrating its suitability for widespread use in biological research.

Materials
The expression vector pMAL-c2x was purchased from Novagen Inc. (Darmstadt, Germany). E. coli JM109 was used as the host for plasmid construction and molecular cloning of the candidate gene. E. coli BL21 (DE3) was used as the host for heterologous expression of the candidate gene. Both of them were obtained from Invitrogen Inc. (Carlsbad, USA) and grown at 37 C in Luria-Bertani (LB) medium (10 g L À1 tryptone; 5 g L À1 yeast extract; 10 g L À1 NaCl and 15 g L À1 agar; pH 7.0). Restriction enzymes and T4 DNA ligase were purchased from Takara Inc. (Dalian, China) and used with the provided buffer according to supplier's recommendations. QIAprep Spin MiniPrep Kit and amylose resin were purchased from Qiagen Inc. (Hilden, Germany). Ab42 gene was codon-optimized and synthesized by GENEWIZ Inc. (Suzhou, China). Primers and sequence analyses of the constructions were performed by BGI Inc. (Shenzhen, China). Commercial Ab42 (>95%) was purchased from GL Biochem Ltd. (Shanghai, China) Dulbecco's modied Eagle's medium and fetal bovine serum were purchased from Gibco Invitrogen Inc. (Grand Island, NY, USA). The PC12 cell line was obtained from National Infrastructure of Cell Line Resources of China. Unless noted, all other reagents and chemicals were of the highest purity available from local sources.

Construction of pMAL-Ab42 expression vector
The codon-optimized DNA amino acid sequences of Ab42 were performed to prefer for E. coli. The Ab42 DNA fragment was synthesized in pUC57 vector as pUC57-(NANP) 3 -TEV-Ab42. The target fragment was digested from the plasmid with EcoRI and HindIII and extract with Gel Extraction Kit. Then the fragment was ligated into pMALc2x vector with the same cohesive end for 4 h at 16 C. The ligated product was transformed into E. coli JM109 competent cells. Aer culturing in 37 C for 12 h, positive transformants were picked for colony PCR validation, and the plasmid was extracted for identication by gene sequencing. The identied plasmid was named pMAL-Ab42.

Expression, purication of MBP-Ab42 fusion protein and digestion by TEV protease
The recombinant plasmid pMAL-Ab42 was transformed into E. coli BL21 (DE3). Single colony was picked and cultured in 5 mL LB medium at 37 C overnight, and inoculated to 200 mL fresh LB medium with 1% inoculation until the OD600 to 0.6. Aer adding a nal concentration of 0.5 mM isopropyl-b-D-thiogalactoside, the recombinant BL21-Ab42 culture was incubated at 16 C for 16-18 h. The cell pellets were harvested (8000 rpm, 10 min, 4 C) and suspended in 20 mL of column buffer (20 mM Tris-HCl, pH 7.4, 200 mM NaCl, 1 mM EDTA, 1 mM DTT). A nal concentration of 30 mg mL À1 lysozyme and 1% (v/v) phenylmethanesulfonyl uoride were added to the solution. Aer incubating in ice bath for 30 min and submitting to sonication (15 cycles of 10 s), the mixture was centrifuged at 12 000 rpm at 4 C for 40 min. The supernatant and precipitate were separated and analyzed by SDS-PAGE. The fusion protein in the supernatant was puried by affinity chromatography with amylose resin and quantitated by Nanodrop 2000 (Thermo Fisher Scientic, Wilmington, USA). 1 mg of the above puried fusion protein was mixed with 4 mL of 250Â TEV buffer, 2 mL or 4 mL TEV enzyme (5 U mL À1 ) and make up to 1 mL volume with buffer (50 mM NaH 2 PO 4 , 150 mM NaCl). Followed by incubating at 23 C for 3 h, 6 h or 12 h, the products of TEV digesting were analyzed by 12% SDS-PAGE.

SDS-PAGE analysis
20 mL of protein sample was mixed with 5 mL Laemmli Sample buffer (Bio-Rad) containing 10% b-mercaptoethanol, and denatured at 100 C for 10 min. Then the protein was separated by 12% polyacrylamide gels. Electrophoresis was run for 30 min at 60 V and then 120 V for nearly 1 h in tris-glycine SDS buffer. Gels were stained with Coomassie brilliant blue R250 and destained with water.

Purication of recombinant Ab42 using size-exclusion chromatography
According to the optimal conditions for digestion, aliquot amount of the fusion protein was taken for digesting. The TEV cleavage product was concentrated to 500 mL and ltered through a 0.22 mm syringe lter. A GE gel ltration column, Superdex 200, was used to separate Ab42 with a column volume of 30 mL and the mobile phase was 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM DTT. The effluent phase was collected according to the peak time. The puried Ab42 was detected by Tricine-SDS-PAGE gel kit (Solarbio, Beijing, China). At last, the product was lyophilized and stored at À80 C.

Western blot and MALDI TOF mass spectrometry
The recombinant Ab42 was separated by Tricine-SDS-PAGE, and the gel was transferred onto a polyvinylidene uoride membrane at 100 V for 30-60 min using a Bio-Rad mini Trans-Blot electrophoretic transfer for western blot analysis. 29 The obtained membrane was sealed in 5% degreased milk powder solution and incubated for 1 h. Then the blotted membrane was rinsed with TBST buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.05% (v/v) Tween 20) for four times. Aer probing with the rabbit anti-Ab42 antibody 6E10 (1 : 1000 dilution, Abcam) in 4 C overnight, the membrane was washed with TBST buffer for four times and then incubated in the goat anti-rabbit IgG secondary antibody (1 : 1000 dilution, Beijing Zhongshan Golden Bridge Biotechnology Co. Ltd., China) at room temperature for another 2-3 h. Following four times washes with TBST buffer, the membrane was visualized with an infrared laser imaging system (LI-COR, USA). The target band of Ab42 from Tricine-SDS-PAGE gel was cut and sent to the Tianjin International Biomedical Research Institute for MALDI TOF mass analysis.

Thioavin T uorescence assay
Ab42 was dissolved in 20 mM NaOH and centrifuged at 16 000 g for 20 min to remove the preformed aggregates. Then, the supernatant of Ab solution was immediately diluted to a nal 25 mM use 10 mM PBS buffer (pH 7.4, 100 mM NaCl). Together with 25 mM ThT, the freshly prepared Ab monomers were incubated at 37 C for 72 h. The ThT uorescent assay was performed by a uorescence plate reader (Innite 200PRO Laboratories, TECAN, Austria). The wavelengths of the excitation and emission were 440 nm and 485 nm, respectively. Three measurements were performed and all data represent the mean AE standard deviation.

Atomic force microscopy
At given time points (0 and 3 d in this study), 20 mL of Ab solution was deposited onto a freshly cleaved mica sheet. Aer incubation for 5 min at room temperature, the mica sheet was rinsed three times with deionized water to remove salts and loosely bound Ab species. Excess water was removed with a gentle stream of nitrogen. Previous studies have shown that the morphology of Ab42 brils in the liquid 30 is similar to the one observed in the air. 31 Moreover, most studies performed AFM experiments in air to observe the morphology of brils formed by Ab. 32,33 Therefore, tapping mode AFM imaging was performed in air using a Multimode 8 AFM (Bruker, USA) with a Scanasyst-Air silicon probe tip (115 mm cantilever, spring constant: 0.4 N m À1 , tip radius: 2 nm, resonance frequency: $70 kHz). Scanning parameters were as follows: peak force setpoint, 0.3-1.0 V; scan rate, 0.8-2.0 Hz. During the AFM imaging, we set the scan size to 2 mm Â 2 mm, and the resolution was set to 1024 lines. Further nano-mechanical measurements were conducted in PF-QNM mode and the tip was initially calibrated on calibration sample (sapphire surface) to obtain the deection sensitivity. The cantilever spring constant and tip radius were further calibrated following the operation manual. All AFM images were analyzed using Nanoscope analysis 1.8 soware (Bruker, USA).

Cytotoxicity assay
In vitro cytotoxicity assay was performed using PC12 cell line. PC12 cells were cultured in Dulbecco's modied Eagle's media supplemented with 10% fetal bovine serum. The cells were plated at a density of $5000 cells per well in 96-well plates with 90 mL of fresh medium. Aer incubated for 24 h, the aged Ab42 aggregates (10 mL each well) were added into the plates, and the cells were incubated for another 48 h. Then, 10 mL of MTT store solution (5.0 mg mL À1 ) was added into each well, and the plates were incubated for another 4 h. Aer removing the culture medium, precipitated cells were lysed using DMSO, and the absorbance at 570 nm was measured using a Microplate Reader (Innite 200PRO Laboratories, TECAN, Austria). The wells containing medium only were subtracted as the background from each reading. The cell viability data were normalized as a percentage of the control group.

Results and discussion
Cloning and plasmid construction for MBP-Ab42 fusion protein expression A schematic of the plasmid pMAL-Ab42 that was used for the expression of the MBP-Ab42 fusion protein is shown in Fig. 1A. In order to obtain the authentic Ab42 peptide, the Ab42 gene was inserted downstream of the MBP-encoding DNA in the same open reading frame, by ligating it between the EcoRI and HindIII restriction sites in the multiple cloning site of the plasmid pMALc2x. Between MBP and Ab42, there was a rigid and soluble linker (NANP) 3 and a modied recognition sequence for TEV (from ENLYFQYG to ENLYFQY; the arrow indicates the cleavage site) which can release the authentic Ab42 without leaving any excess residues. Firstly, the codon usage of the (NANP) 3 -TEV-Ab42 fragment was optimized for E. coli, and the DNA sequence (ESI Table S1 †) was obtained by synthesis, and was delivered cloned into the plasmid pUC57-(NANP)3-TEV-Ab42. The fragment encoding (NANP) 3 -TEV-Ab42 was double-digested using EcoRI/HindIII and sub-cloned into the pMALc2x expression vector. The correct assembly of the recombinant expression vector was conrmed by PCR using the primer pair EcoRI-Ab42-F/HindIII-Ab42-R (ESI Table 1 †). As shown in Fig. 1B, PCR product of the expected size (approximately 200 bp) was generated from four positive transformants,  Paper while the JM109 wild-type control showed no clear band at the corresponding location. Furthermore, the orientation and sequence of the gene fragment was veried by DNA sequencing (data not shown). These results proved that the designed expression vector pMAL-Ab42 was successfully constructed as intended.

Expression and purication of the MBP-Ab42 fusion protein
A schematic of the expression and purication process of the Ab42 peptide is shown in Fig. 2. The plasmid pMAL-Ab42 was rst transferred into the expression strain BL21. Aer cultivation on a plate at 37 C overnight, a single colony of the recombinant bacteria was used to inoculate LB medium, and induced with isopropyl-b-D-thiogalactoside. The cells were collected and submitted to sonication. The supernatant and sediment were analyzed by SDS-PAGE, which showed that the MBP-Ab42 fusion protein was present in the supernatant, demonstrating its successful soluble expression (ESI Fig. 1 †). The supernatant of the cell lysate was separated using affinity chromatography with amylose resin. Aer washing twice with column buffer, the MBP-Ab42 fusion protein was eluted with column buffer containing 10 mM maltose. The ow-through, wash, and elution fractions were analyzed by SDS-PAGE as shown in Fig. 3A. It was clear that the amounts of the MBP-Ab42 fusion protein in the ow-through and wash fractions were very low. By contrast, the elution fraction contained mainly MBP-Ab42 fusion protein, with minor impurities comprising low-molecular-weight proteins. As can be seen in Fig. 3A, the molecular weight of the fusion protein was approximately 49.5 kDa, which was consistent with its predicted molecular weight. The fusion protein yield was estimated to be up to $240 mg per liter of culture.
TEV protease cleavage of the MBP-Ab42 fusion protein and purication of recombinant Ab42.
In order to obtain authentic Ab42 without any additional amino acid residues, the cleavage site of the TEV protease (ENLYFQYG) was changed to the 6-amino-acid sequence (ENLYFQY). It is known that 90% cleavage efficiency of TEV is retained when the glycine is substituted by aspartate, 34 which is serendipitously also the rst amino acid of Ab42. Therefore, aer the MBP-Ab42 fusion protein is digested with TEV protease, native-like Ab42 without any additional residues can be obtained. In order to achieve the best digestion effect, the digestion conditions were optimized by modulating the addition of TEV protease and the digestion time. The cleavage efficiency was estimated by SDS-PAGE and the results are shown in Fig. 3B. The molecular weight of the MBP-Ab42 was about 49.5 kDa, which can be cleaved into the two fragments MBP and Ab42 with 45 and 4.5 kDa, respectively. Aer cleavage for 3 h, two lanes were observed around 49.5 kDa and 45 kDa, corresponding to the MBP-Ab42 fusion protein and MBP, respectively (Fig. 3A). Since low-molecular-weight proteins are difficult to visualize by the ordinary SDS-PAGE used in the current study, the Ab42 band cannot be seen in Fig. 3B. However, the effect of TEV cleavage could nevertheless be analyzed qualitatively and quantitatively due to the gradually decreasing 49.5 kDa band of the MBP-Ab42 fusion protein and increasing 45 kDa band of the MBP tag. We found that TEV cleavage was signicantly enhanced by adding more protease. The results showed that the optimal conditions for enzymatic digestion encompassed a reaction temperature of 23 C, a ratio of the MBP-Ab42 fusion protein to TEV protease of 1 mg to 4 mL of an 5 U mL À1 TEV solution, and a reaction time of 12 h.
The digested product was further puried by size-exclusion chromatography (SEC) to obtain high-purity Ab42 (Fig. S2 †). As shown in Fig. 3C, the puried product was veried by Tricine-SDS-PAGE. A clear single band around 4.5 kDa was visible, indicating a molecular weight that is consistent with the theoretical molecular weight of Ab42. We obtained approximately $18 mg of puried Ab42 from 1 L of culture, which is a very high yield compared to the other strategies adapted for the affordable production of Ab peptide. 16,19,20,35 Characterization of the puried recombinant Ab42 The identity of the puried recombinant Ab42 peptide was further conrmed by immunoblotting. A chemically synthesized Ab42 peptide and cell lysate of E. coli BL21 containing pMAL-c2x vector were used as positive and negative control, respectively. The samples were separated via electrophoresis and transferred onto a PVDF membrane. Aer incubating overnight at 4 C with the monoclonal rabbit antibody 6E10, which specically binds to Ab42, 36 the membrane was incubated with a goat anti-rabbit IgG secondary antibody for 2-3 h at room temperature, and the membrane scanned by an infrared laser imager. As shown in Fig. 3D, the recombinant Ab42 obtained in this study was identical to chemically synthesized Ab42, with both samples showing a single specic band at the same position, while there was no band in the control group comprising BL21 cell lysate. This result proved that the puried product is authentic Ab42 peptide.
The target band from the gel was cut out and analyzed by MALDI-TOF/TOF mass spectrometry. As shown in Fig. 4, the target product ion spectra derived from the fragmentation of the charged ion at m/z 1325.69 correspond to the peptide KLVFFAEDVGSNKG (residues 16-29 of Ab42). Analysis of the secondary ion chromatograms revealed m/z 504 corresponding to the peptide KLVF (residues 16-19 of Ab42) and m/z 966.58 corresponding to the peptide FAEDVGSNKG (residues 19-29 of Ab42). The mass spectrum therefore corroborated that the target product was indeed authentic Ab42.

Biophysical characterization of puried recombinant Ab42
In order to characterize the aggregation properties of the puri-ed recombinant Ab42, we analyzed the amyloid properties by studying its aggregation kinetics, AFM and cytotoxicity. The ThT uorescence assay is considered a highly sensitive tool for detecting the formation of amyloid aggregates of various amyloidogenic proteins. [37][38][39] Therefore, this assay was used to characterize the aggregation kinetics of recombinant Ab42 as shown in Fig. 5A. When 25 mM recombinant Ab42 was incubated at 37 C for 72 h, the ThT uorescence prole showed an almost negligible lag phase, a fast growth phase within 20 h, and a steady equilibrium phase aer 24 h. Fig. 5B shows the AFM images of Ab42 aer 0 and 3 days of incubation, which conrmed that the recombinant Ab42 formed typical long, branched brils with lengths of 200-700 nm. As shown in Fig. 5C, in the height image derived from the cross-section of the mature ber in Fig. 5B (indicated by the white bar), the height of the brils was about 5-15 nm. The corresponding 3D images of the two artifacts are also shown in Fig. S3A, † which present a visual view of the brils. Further nano-mechanical properties of the brils were analyzed in PF-QNM mode and the adhesion image is shown in Fig. S3B. † The dashed white line marks the position at which the adhesion image was analyzed and the prole of the adhesion force along the bril is shown in Fig. S3C. † According to the results, the average value of the adhesion force along the bril was 2.35 AE 0.66 nN.
This observation was consistent with the ThT uorescence results and our previous studies. 40 Therefore, both ThT and AFM results indicated that the recombinant Ab42 showed biochemical behavior consistent with the aggregation characteristics of previously reported recombinant and synthetic Ab42. 17,[40][41][42] Cytotoxicity of recombinant Ab42 towards PC12 cells To conrm the cytotoxicity of the recombinant Ab42, the cell viability of cultures of the PC12 neuronal cell line exposed to it was assessed using the standardized MTT assay. We set the survival of cells mock-treated with PBS buffer only as a basis of 100% to normalize the other data for comparison. As shown in Fig. 6, treatment of the PC12 cells with 3 mM recombinant Ab42 for 48 h reduced their viability by about 47.5%. These results suggest that the aggregates of the recombinant Ab42 exhibit strong cytotoxicity, which was expected based on the results of the ThT assay and AFM.

Conclusions
In this study, we developed a novel method for the efficient expression and purication of recombinant Ab42 without any additional residues. Herein, an MBP fusion tag was used to improve the solubility and yield of Ab42. The rigid and soluble linker (NANP) 3 and the modied TEV protease recognition site ENLYFQ were added between the MBP and Ab42. Aer optimization of the TEV cleavage conditions, the MBP tag was removed, yielding 18 mg L À1 of authentic Ab42, which was identied by western blot analysis and MALDI-TOF mass spectrometry. ThT uorescence and AFM assays demonstrated the great aggregation activity of the recombinant peptide, and the MTT assay demonstrated that the aggregates of the recombinant Ab42 showed neurotoxicity in vitro. Furthermore, the method developed in this study provides us with a biochemical tool to obtain Ab variants with various lengths or sequences that will certainly facilitate further structural and functional studies in the future. Therefore, the availability of puried Ab42 could enable new structural, biochemical, and biological insights, as well as cheaper screening of anti-AD drug candidates. Moreover, the recombinant protein could also be used as a reagent for the  preparation of monoclonal antibodies against Ab42, which can also be used in anti-Ab therapies for AD. Finally, this study also provides a potential strategy for the expression and purication of other amyloid proteins in E. coli.

Conflicts of Interest
There are no conicts of interest to declare.