Scholarly article on topic 'DNA aptamer affinity ligands for highly selective purification of human plasma-related proteins from multiple sources'

DNA aptamer affinity ligands for highly selective purification of human plasma-related proteins from multiple sources Academic research paper on "Chemical sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Chromatography A
OECD Field of science
Keywords
{"Affinity ligands" / Aptamers / "Blood proteins" / Chromatography / Selectivity}

Abstract of research paper on Chemical sciences, author of scientific article — Cynthia Forier, Egisto Boschetti, Mohamed Ouhammouch, Agnès Cibiel, Frédéric Ducongé, et al.

Abstract Nucleic acid aptamers are promising ligands for analytical and preparative-scale affinity chromatography applications. However, a full industrial exploitation requires that aptamer-grafted chromatography media provide a number of high technical standards that remained largely untested. Ideally, they should exhibit relatively high binding capacity associated to a very high degree of specificity. In addition, they must be highly resistant to harsh cleaning/sanitization conditions, as well as to prolonged and repeated exposure to biological environment. Here, we present practical examples of aptamer affinity chromatography for the purification of three human therapeutic proteins from various sources: Factor VII, Factor H and Factor IX. In a single chromatographic step, three DNA aptamer ligands enabled the efficient purification of their target protein, with an unprecedented degree of selectivity (from 0.5% to 98% of purity in one step). Furthermore, these aptamers demonstrated a high stability under harsh sanitization conditions (100h soaking in 1M NaOH). These results pave the way toward a wider adoption of aptamer-based affinity ligands in the industrial-scale purification of not only plasma-derived proteins but also of any other protein in general.

Academic research paper on topic "DNA aptamer affinity ligands for highly selective purification of human plasma-related proteins from multiple sources"

il i'lfljfllm^b i'miiile in press

journal of Chromatography A, xxx (2017) xxx-xxx

j 1 18S1 Contents lists available at ScienceDirect

09 | u Journal of Chromatography A

ELSEVIER journal homepage: www.elsevier.com/locate/chroma

DNA aptamer affinity ligands for highly selective purification of human plasma-related proteins from multiple sources

Cynthia Foriera, Egisto Boschettib, Mohamed Ouhammouchc, Agnès Cibielc, Frédéric Ducongéd, Michel Nogréa, Michel Telliera, Damien Bataille3, Nicolas Bihoreaua, Patrick Santambiena, Sami Chtouroua, Gérald Perret3 *

Nucleic acid aptamers are promising ligands for analytical and preparative-scale affinity chromatography applications. However, a full industrial exploitation requires that aptamer-grafted chromatography media provide a number of high technical standards that remained largely untested. Ideally, they should exhibit relatively high binding capacity associated to a very high degree of specificity. In addition, they must be highly resistant to harsh cleaning/sanitization conditions, as well as to prolonged and repeated exposure to biological environment. Here, we present practical examples of aptamer affinity chromatography for the purification of three human therapeutic proteins from various sources: Factor VII, Factor H and Factor IX. In a single chromatographic step, three DNA aptamer ligands enabled the efficient purification of their target protein, with an unprecedented degree of selectivity (from 0.5% to 98% of purity in one step). Furthermore, these aptamers demonstrated a high stability under harsh sanitization conditions (100 h soaking in 1 M NaOH). These results pave the way toward a wider adoption of aptamer-based affinity ligands in the industrial-scale purification of not only plasma-derived proteins but also of any other protein in general.

© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

a LFB Biotechnologies, Les Ulis, France b Bioconsultant, JAM Conseil, Neuilly, France c 1NSERM U1023, LIME, Orsay, France

d CEA, 12BM, MIRCen, UMR 9199, Université Paris Saclay, Fontenay aux Roses, France

ARTICLE INFO ABSTRACT

Article history:

Received 6 October 2016

Received in revised form 4January 2017

Accepted 6 January 2017

Available online xxx

Keywords: Affinity ligands Aptamers Blood proteins Chromatography Selectivity

1. Introduction

The isolation of a single protein either produced in a cell culture or present in a biological fluid remains a complex challenge. Building a capture system that is simultaneously effective and specific entails a number of interdependent steps. It includes the ligand design or its selection from available collections, the choice of appropriate immobilization chemistry and, ultimately, the use of the resulting affinity chromatography media for the purification of the target protein under fine-tuned conditions of adsorption and elution.

A large diversity of ligands is available today for the purification of a given protein. Antibodies [1,2], nanobodies from camelides [3], affibodies [4], polypeptides from phage display [5], structures derived from natural ankyrin repeats [6] and stefin-based scaffolds [7] are the most known protein-based affinity ligands.

* Corresponding author. E-mail address: perretg@lfb.fr (G. Perret).

Synthetic peptides [8,9], peptoids [10 and other molecules as biomimetic ligands [11,12] could also be used. Alternatively, combinatorial libraries of small and large compounds have been explored to identify molecules that can be used as ligands for affinity chromatography (for a recent review see Ref. [13]). Among them DNA aptamers are proposed as recent promising ligands. They are short single strand oligonucleotides that can fold into a three-dimensional conformation enabling the precise molecular recognition of a given target. These nucleic acid-based ligands are identified from large naive libraries using a well-established technology called SELEX (Systematic Evolution of Ligands by Exponential enrichment) [14,15]. They can be extremely specific toward their target molecule with very high selectivity and low equilibrium dissociation constants (Kd). Moreover, aptamers can be chosen to recognize a high diversity of molecules from very small compounds to large macromolecules such as proteins [16]. They are prepared and screened in vitro. Their selection process could be achieved under various and not necessarily physiological conditions, a critical advantage where immunopurification does not apply or when molecular recognition in a specific buffer is required. Aptamer lig-

http://dx.doi.org/10.1016/j.chroma.2017.01.031

0021-9673/© 2017 The Authors. Published by Elsevier B.V. This is an open access article underthe CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4. 0/).

«■■■m article in press

2 C. Forier et al. /J. Chromatogr. A xxx (2017)xxx-xxx

ands are resistant to proteases, and chemical modifications can be introduced to further enhance their resistance to the nucleases present in biological environment [17,18]. Once characterized, and in contrast to proteinaceous ligands, aptamers can be manufactured by large scale chemical synthesis. So far no toxicity and no immunogenicity events have been reported for aptamers used for therapeutic purposes [19]; as a consequence, immunization against traces of leachable aptamer impurities that could eventually be present in a purified therapeutic protein would be unlikely. Furthermore, the chemical immobilization on the solid phase can be oriented [20] to keep intact the tridimensional structure of aptamers that enable the highest docking efficiency. Once grafted, DNA aptamer-based chromatography media could be regenerated with stringent chemical agents without loss of affinity, which is not the case of protein ligands that are irreversibly denatured [21]. For instance, DNA aptamers withstand high pH, concentrated salts, and high temperatures. Furthermore, the highly sensitive Q-PCR technology could easily be used when assaying for possible released aptamer traces in the purified therapeutic protein product. Such assays are a requirement for regulatory agencies and often constitute a challenge when the affinity ligand is a protein such as an antibody.

In spite of the above-listed advantages, the use of aptamers in affinity chromatography remains tentative. A number of small scale applications have been reported [22-26], but unresolved questions such as long term stability and uncertainties like high binding capacity and high cost probability remain to be addressed before their inclusion in preparative industrial scale processes. Biotinylated aptamers are described as a way to make affinity chromatography using streptavidine-conjugated solid supports [27-32]. Alternatively chemical grafted aptamers have been described involving various chemistries such as N-hydroxysuccinimide [30] and cyanuric chloride [23]. A quite exhaustive list of chromatography separation application examples is given by Walter et al. [33]. Immobilized aptamers have been used for the affinity separation of several proteins such as lysozyme [24], immunoglobulin G [34], thrombin [35], His-Tag fusion proteins [23,25], Thermus aquaticus recombinant DNA polymerase [36] and Concanavalin A [26]. In all these cases, good levels of purity have been achieved, but the modest binding capacities and weak stability when RNA aptamers were used, prevented cost-effective scale-up toward preparative industrial applications.

In this paper we describe the effective advantages of DNA aptamer-mediated affinity chromatography with three purification examples focused on distinct plasma-derived high-value proteins. The experiments were performed keeping in mind possible up scaling for the purification of therapeutic proteins. Particularly, we point out the efficiency of the purification, its level of selectivity and its capability to withstand repeatedly caustic cleaning conditions.

2. Material and methods

2.1. Chemicals, biologicals and equipments

All biological products (crude samples and pre-purified or purified products from various sources such as human plasma, transgenic milk and cell culture supernatant) were internal preparations from LFB (Les Ulis, France). All chemicals including Coomassie Brilliant Blue protein staining were from Sigma Aldrich (Saint Louis, MO, USA) and were of pure grade. Blue Juice was from Invitrogen, Carlsbad, CA, USA. Gel Red stain was from Biotium Inc., Hayward, CA, USA. N-hydroxysuccinimide (NHS)-activated Sepharose, chromatographic columns XK16, Silver Staining Kit and surface plasmon resonance (SPR) system and products were from GE Healthcare (Uppsala, Sweden).

Chromatographic sorbent Affigel-15, Molecular Imager Chemi-docXRS+ and Image Lab software were from Bio-Rad Laboratories (Hercules, CA, USA). Gel electrophoresis plates NuPAGE® SDS-PAGE sodium dodecyl sulfate polycrylamide gel electrophoresis (SDS-PAGE), lithium dodecyl sulfate (LDS) and Sample Buffer were from Life Technologies (Rockville, MD, USA). Polyethersulfone membrane filters were purchased from Millipore Corp. (Milford, MA, USA).

Blood coagulation Factor VII (Biophen FVII) kit was from ANIARA Corporation (West Chester, OH, USA). Asserachrom Vll:Ag kit was from Stago (Asnieres, France). Diode array detector G1315 and ChemStation software were both from Agilent Technologies (Santa Clara, CA, USA). Minipuls 3 peristaltic pump for chromatography was from Gilson (Middleton, Wl, USA). Plasma protein coagulome-ter was from Siemens Healthcare (Saint Denis, France).

2.2. Information on used aptamers

DNA aptamers were selected using the SELEX technology [14,15] according to the original described method and then their properties are evaluated by SPR (surface plasmon resonance).

All selected aptamers were synthesized by Eurogentec (Seraing, Belgium). The sequences of anti-FVll aptamer Mapt2.2CS [37] is as follows: 5'-CCGCACGCTACGCGCATGAACCCGCGCACACGACTTGAAGTAGC-3'. The sequence of anti-blood coagulation Factor lX aptamer (Anti-FIX) called Nonapta5.1 [38] is as follows: 5'-ACCTTTAACCCACTAGTCTAAGCGGGTCTAGCGTCGTCC-3'. The sequence of anti-Factor H aptamer called MaptH1.1CSO [39] is as follows: 5'-GGTCTCGGGCACGGGTCAGGCGGTTATACGGTGCCC-3'. The sequence of anti-GLA domain aptamer Mapt1 [40,41] is as follows: 5'-GGAGATAGCCACGACCTCGCACATGACTT GAAGT-TAAACGCGAATTACAAACCCAGCCC CCTCCAGGCTGTGCGAAAGC-3'. This sequence has also been shortened to a smaller size (Mapt1.2CS) with a similar specificity, corresponding to the following modified sequence: 5'-CCACGACCTCGCACATGACTTGAAGTTAAACGCGAATTAC-3'.

All aptamers used in SPR experiments comprised a biotin at 5' terminal end. For affinity chromatography applications the aptamers comprised a primary amine at the bottom ofa six-carbon spacer attached at 5' terminal end.

2.3. Selective affinity measurement of aptamers by SPR

Experiments were performed with Biacore T100 at 25 °C. Data were evaluated with Biacore evaluation software.

Biotinylated aptamers Mapt2.2CS, MaptH1.1CSO and Non-apta5.1 were injected in streptavidin chips until the saturation of the active flow cell. Coagulation factor Vll (FVll), blood complement Factor H (FH) and coagulation factor lX (FlX) were diluted in running buffer (50 mM Tris-HCl, 10 mM CaCl2 at pH 7.5 for FVll, 50 mM Tris-HCl, 10 mM MgCl2 at pH 7.5 for FH and 50 mM Tris-HCl, 150 mM NaCl, 2mM CaCl2 1 mM MgCl2 at pH 7.4 for FlX) then injected respectively at a concentration of 200 nM for the first and 1 |M for the others onto both active and control flow cell during 60 s. After a waiting time allowing observation of the dissociation profile, regeneration buffer was injected on both flow cells. Regeneration solution for FVll was 50 mM Tris-HCl, 10 mM ethylenediamine tetraacetic acid disodium salt (EDTA) at pH 7.5, while for FH was 50 mM NaOH and for FlX was 10 mM EDTA at pH 5.

The same procedure was adopted for the binding of biotinylated aptamer anti-GLA domain to FVll and FlX (Mapt1). The running buffer was 50 mM Tris-HCl, 50 mM NaCl, 10 mM CaCl2,5 mM MgCl2 at pH 7.5. Factor Vll and Factor lX were diluted in running buffer

aü^ article in press

C. Forier et al. /J. Chromatogr. A xxx (2017)xxx-xxx 3

and injected on both active and control flow cell during 60 s. Regeneration solution was 50 mM NaOH.

2.4. Aptamer coupling to solid chromatographic support and coupling yield determination

Aptamer coupling onto solid support was performed using a proprietary aptamer grafting technology method [42]. Briefly the required amount of 5' amino-modified aptamer was diluted in sodium acetate pH 4.2. Then the solution was mixed with NHS-activated Sepharose and kept under gentle stirring during at least 1 h. Then a solution of 100 mM Tris-HCl buffer, pH 8.5 was added to the gel sorbent to deactivate all the non-reacted sites and the bead suspension was kept under gentle stirring during at least 1 h. The affinity sorbent was rinsed by alternating washes with 100 mM sodium acetate containing 500 mM NaCl, pH 4.0 and 100 mM Tris-HCl buffer, pH 8.5. All operations were performed at room temperature. Then the gel conjugate was stored at +4 °C in a buffer containing 0.2% sodium azide until its use.

To determine the amount of grafted aptamer the washing supernatant was compared to a set of serial aptamer dilutions (known concentrations). The highest concentration of the set represents the theoretical maximum concentration that could be present in supernatant and by consequence corresponding to 0% grafting yield. All samples and controls were diluted in Blue Juice and loaded into a 3.5% agarose gel plate for electrophoresis containing Gel Red stain. Band intensity proportional to aptamer concentration was analyzed using Molecular Imager Chemidoc XRS+ and Image Lab software; serial dilutions were used to determine a standard curves and intensity comparison was performed against the intensity of the bead supernatant. By this way the quantitative determination of the grafted aptamer was approached.

2.5. Binding capacity determination

Binding capacity was measured by chromatography at room temperature with a LC system consisting of a Minipuls 3 peristaltic pump and a 1200 series diode array detector G1315 set at 280 nm. The flow rate used was 0.5 mL/min. 0.5 mL of aptamer affinity sorbent was packed on a small column and equilibrated with 50 mM Tris-HCl buffer containing 10 mM CaCl2, pH 7.5. Then an amount of purified affinity protein exceeding the capacity was loaded (around two times the amount of grafted aptamers). F-VII was then desorbed with 50 mM Tris-HCl buffer containing 10 mM EDTA. Fractions were collected manually and the amount of eluted protein measured using 280 nm OD against a control.

2.6. General chromatographic separation methods

All chromatography experiments were performed at room temperature with a liquid chromatography (LC) system consisting of a Minipuls 3 peristaltic pump and a 1200 series diode array detector G1315 set at 280 nm. The used flow rate was 0.5 mL/min. All buffers and aqueous solutions used were filtered at 0.22 |im with polyethersulfone membrane filters. Signals from the detector were collected using ChemStation software. Fractions were collected manually. At the end of the separation process the column was always cleaned using 100 mM NaOH solution followed by the storage buffer.

2.6.1. FVII purification

One mL of gel grafted with 5.6 mg of 5' amino-modified Mapt2.2CS was packed on a XK16 column and equilibrated with 50 mM Tris-HCl buffer containing 10 mM CaCl2, pH 7.5 during 1 h. Then the column was sanitized with 1 mL of 100 mM NaOH solution and finally washed with the equilibration buffer above. In one case,

200 |ig of pre-purified plasmatic FVII was previously diluted with the equilibration buffer and injected into the column. In another case, the injected product was 2.8 mL of FVII previously diluted with the equilibration buffer at 0.33 mg/mL and then spiked in 1.4 mL of non-pretreated cow milk (total volume: 4.2 mL, 900 |g of FVII with a final concentration of 0.21 mg/mL). In both cases, non-adsorbed proteins were pushed out by the equilibration buffer and the captured proteins were eluted by 50 mM Tris-HCl buffer containing 10 mM EDTA, pH 7.5. After elution a 1 mL mixture composed of 1 M NaCl and 50% propylene glycol was injected as a regeneration solution before the sanitization stage with 1 mL of 100 mM NaOH. Propylene glycol was used as agent to remove lipid-like products and thus contributing for a better cleaning. Column flow rate all along the experiments was 0.5 mL/min.

2.6.2. FH purification

One mL of gel grafted with 5.2 mg of 5' amino-modified MaptH1.1CSO was packed on a XK16 column and equilibrated with 50 mM Tris-HCl buffer containing 50 mM NaCl, 10 mM MgCl2, pH 7.5 during 1 h. Then the column was sanitized with 1 mL of 1 M NaOH solution and finally washed with the equilibration buffer. Five mg of pre-purified plasmatic FH or 45 mL of cell culture supernatant containing recombinant FH were diluted in the equilibration buffer and injected prior to be eluted by 50 mM Tris-HCl buffer containing 100 mM EDTA, pH 8. Column flow rate all along the experiments was 0.5 mL/min.

2.6.3. FIX purification

One mL of gel grafted with 6.3 mg of 5' amino-modified Non-apta5.1 was packed on 1.1 cm diameter column and equilibrated with 50 mM Tris-HCl buffer containing 150 mM NaCl, 2 mM CaCl2, 1 mM MgCl2, pH 7.4 during 1 h. Then the column was sanitized with 2 mL of 50 mM NaOH solution and finally washed with the equilibration buffer. Pre-purified product from plasma, containing 0.5% of plasmatic FIX (2.6 mg), was diluted in 50 mM Tris-HCl buffer containing 150 mM NaCl, 0.5 mM CaCl2, 1 mM MgCl2, pH 7.4. The final volume of 16,3 mL was injected. A wash with the equilibration buffer supplemented with 2 M NaCl was peformed prior to an elution with 200 mM EDTA, pH 8. Column flow rate all along the experiments was 0.5 mL/min.

2.6.4. Separation of active plasma-related proteins with anti-GLA domain aptamer

One mL of gel grafted with 0.5 mg of 5' amino-modified Mapt1 was packed on 1.1 cm diameter column and equilibrated with 50 mM Tris-HCl containing 50 mM NaCl, 10 mM MgCl2,4 mM CaCl2, pH 7.5 during 1 h. Milk from transgenic pig was first clarified by means of citric acid, classically fractionated by MEP-HyperCel chromatography and then dialysed against 50 mM Tris-HCl buffer, 50 mM NaCl, pH 7.5. At this step, the protein extract containing about 2% of human Factor IX with a majority of badly 7-carboxylated forms, was added with MgCl2 and CaCl2 to reach the same composition as the equilibration buffer. Sixteen milliliters of this mixture were injected into the Mapt1 aptamer affinity column. After having washed out all non-adsorbed material, FIX was eluted using 20 mM Tris-HCl, 10 mM EDTA, pH 7.5. Then the column was regenerated using 20 mM Tris-HCl buffer, 1 M NaCl, 5% propy-lene glycol, pH 7.5. Column flow rate all along the experiments was 0.5 mL/min.

2.7. Purity determination of affinity-separated proteins

SDS-PAGE analyses were performed on NuPAGE® SDS-PAGE according to the manufacturer procedures. Samples diluted in LDS Sample buffer, were heated at 95 °C for 3 min. Electrophoreses were performed using precast 4-12% Bis-Tris gradient gels for 55 min

article in press

C. Forieretal./J. Chromatogr. A xxx (2017)xxx-xxx

at 200 V in 2-(N-morpholino)ethanesulfonic acid (MES)-SDS running buffer according to the recommendations of the manufacturer. The gel plates were stained with either Coomassie Blue or Silver Staining Kit.

2.8. DNA aptamer sorbent stability trials towards repeated regeneration and cleaning steps

Resistance of the aptamer ligands to sanitization and to crude biological sample exposure were evaluated using an affinity chromatography model. 1 M NaOH solution was added to anti-FVII aptamer gels and kept during 100 h. To test the gel stability after the caustic exposure the column was washed with the equilibration buffer followed by the injection of FVII corresponding to 80% of gel capacity; the affinity chromatography fractionations were then performed under the same conditions as previously described.

A second resistance test was performed on the same grafted affinity sorbent by running 30 cycles of adsorption buffer alternated by 1 M NaOH.

A third resistance test was performed by exposing the affinity sorbent to transgenic crude clarified milk where FVII was expressed for 100 h. Then the column was assayed for its binding capacity at saturation and also for its ability to separate the target protein after a proper wash with the equilibration buffer. This was performed by injecting a sample containing an amount of FVII corresponding to 80% of gel capacity and developing the affinity purification as previously described.

For each type of purification trial, the FVII purity and the relative biological activity were determined and compared to a control assay.

2.9. Determination ofFV11 and FIX biological activity

Amidolytic activity determination consisted in measuring the FVII and FIX capacity to activate blood coagulation cascade by using a Biophen FVII and FIX kits and a coagulometer according to the instruction of the supplier. The FVII protein and FIX protein assays were performed as an ELISA using Asserachrom VII:Ag and IX:Ag kits according to the supplier's instructions. The FVII and FIX biological activity is here expressed as the ratio between specific amidolytic activity and total FVII or FIX protein; for a full optimized activity this ratio should be of 1.

2.10. Determination ofFH biological activity

This analytical method derives from Sanchez-Corral et al. [43]. It assesses the ability of a purified Factor H to inhibit specific lyses obtained in the presence of sheep red blood cell and a mixture of human plasma pool and serum depleted of Factor H in equivalent proportions. Factor H activity was compared to 100% lyses control which corresponds to the maximum lyses of sheep red blood cells observed with water. Factors H samples (purified or not) were analyzed in dose-effect mode. The samples and controls were incubated for 30 min at 37° C. After incubation, 400 |L of 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-EDTA was added to the sample tubes and 'Blank' control to stop the reaction and 400 |L of water for injection in the '100% lyses' control. The tubes were then vortexed and centrifuged 5 min at 1730g 200 |L of supernatant from each tube was transferred into 96-well microplate and reading was performed at 414 nm.

3. Results and discussion

3.1. Selection and evaluation of aptamer ligands

Caustic agents are frequently used to clean resins at preparative industrial scale between each separation run. Therefore, we decided to use DNA-based aptamer ligands that are known to be stable in such harsh conditions. Furthermore, DNA aptamers are easy to make and cheaper than other class of aptamers such as RNA oligonucleotides. DNA aptamers were selected using the SELEX technology against coagulation Factor VII, complement Factor H, coagulation Factor IX and GLA domain containing proteins. Before coupling on solid chromatographic support the aptamer ligands have been evaluated by SPR (Fig. 1). The main point to consider here is the very shallow decrease of response during the washing step with the equilibration buffer represented by the arrows. This means that during that necessary chromatographic phase to remove foreign proteins that are part of the sample, the risk of losing the target protein is really minimum contributing thus to the maximization of overall separation yield. Among the different affinity aptamers identified, we preferred those with a dissociation constants (Kd) calculated within the nanomolar range. For example, the Kd of the aptamer used for the interaction with FIX was 1.2 nM. Another important selection parameter to consider was the obtention of a very slow dissociation rate. This characteristic is very suitable for chromatographic applications in order to strongly capture the target proteins allowing for efficient washing steps without significant desorbtion of them.

3.2. High yield immobilization of aptamer ligands

All aptamers used in the present study were modified by introducing at the 5' terminal end a 6-carbon spacer comprising a primary amine for proper oriented immobilization reaction on NHS-activated Sepharose. Although this chemical immobilization approach is mostly described for polypeptides it is also quite effective to aptamers when they comprise a primary amine at the bottom of a spacer arm. Repeated coupling experiments on NHS-activated Sepharose were performed. The grafting yields were usually higher than 90%, corresponding to more 6 mg of aptamer per mL of affinity sorbent. With this grafted amount, the binding capacity for the targeted proteins were measured in the range of 6-8 mg per mL of affinity chromatography support according to various experimental determinations. This is a particularly high value when compared to what has been recently reported in for the separation of pure Concanavalin A [26] where the authors indicated a binding capacity lower than 1 mg per mL of sorbent.

3.3. Highly selective purification process for plasma derived proteins

An affinity purification system is generally considered effective when it captures very specifically the target protein in the presence of a large number of protein impurities. This is why the purification examples of plasma derived proteins were developed using crude or semi-crude protein extracts.

The first affinity chromatography evaluation was performed with coagulation FVII which is a 50 kDa single-chain glycoprotein present in blood plasma at a normal concentration of 0.5 | g/mL. This is a very low-abundance protein representing less than 0.001% of total plasma proteins, involved in the blood coagulation process and is mediated by a Tissue Factor [44]. It is used in clinical applications of acquired hemophilia and also when alloantibodies to coagulation factors are developed. It is classically purified by a combination of several chromatographic separation steps involving ion exchange and immunoaffinity chromatography from either human

'in'mpffw aiiiii e in press

C. Forieret al./J. Chromatogr. A xxx (2017)xxx-xxx

Fig. 1. Affinity measurements of selected aptamers by SPR.

A: FVII was diluted in running buffer at 200 nM concentration and then injected in a biotinylated Mapt2.2CS aptamer immobilized on a streptavidin chip. B: FH was diluted in running buffer at 1 |M concentration and then injected in a biotinylated MaptH1.1CSO aptamer immobilized on a streptavidin chip. C: FIX was diluted in running buffer at 1 |M concentration and then injected in a biotinylated Nonapta5.1 aptamer immobilized on a streptavidin chip. D: FVII and FIX diluted in running buffer at 100 nM concentration and then injected in a biotinylated Mapt1 aptamer immobilized on a streptavidin chip. The arrows indicate the natural dissociation in the presence of the binding buffer. The very shallow slope of the curve from the arrow point demonstrates the strong affinity between the aptamer and the target protein with a very weak decrease of RU (Resonance Unit) signal.

Start Ft El

10 20 30 40 50 60

Time (min)

Fig. 2. Purification of coagulation Factor VII by aptamer affinity chromatography.

A: Global chromatographic profile with the injection of the pre-purified human plasma sample (single arrow) in 50 mM Tris-HCl pH 7.4, containing 150 mM sodium chloride, followed by the elution phase using a solution of 50 mM EDTA. 'Ft' represents the flowthrough or non-adsorbed proteins; 'El' represents the eluted protein fraction; 'Reg' represents the regeneration-released products washed out using sodium chloride and propylene glycol solution. Column dimensions: 16 mm ID x 10 mm long; the flow rate all along of the separation was 0.5 mL/min. B: SDS-PAGE analytical results of fractionated sample. The first lane from the left (Start) represents the initial injected material; the second lane is the flowthrough fraction (Ft). Elution of captured protein (Factor VII) is represented on the right lane (El). The polyacrylamidegel concentration of the plate was 4-12%; the migration was performed under 150 V for 50 min as per the recommendation from the supplier. Proteins separated by electrophoresis were silver stained for an increased sensitivity.

cryoprecipitate-poor plasma [45] or from recombinant technology preparations [46]. Coagulation Factor VII purification trials were performed using a DNA aptamer named Mapt2.2CS and a pre-purified preparation of human origin containing about 10% of protein impurities (Fig. 2, panel A). Since the molecular recognition of this coagulation factor by the selected aptamer was contingent upon the presence of divalent cations, the adsorption condition involved the presence of calcium chloride at 10 mM concentration buffered in 50 mM Tris-HCl, pH 7.4 containing 150 mM sodium

chloride. The adsorbed FVII was eluted with an aqueous buffered solution of 50 mM EDTA. The analysis of the flowthrough and the eluted fraction demonstrated a high degree of purity reached by the present affinity process (see Fig. 2, panel B). Up to now the best affinity chromatography ligands used for the purification of either natural or recombinant Factor VII are specific antibodies. Beyond the above-described merits of aptamers versus antibodies we observed that the binding capacity of the prepared aptamer-based sorbent is significantly higher. Actually, it was estimated as

jmbbiim aiiiii e in press

6 C. Forier et al. /J. Chromatogr. A xxx (2017)xxx-xxx

being 4-6 times higher, with a purification degree of at least 98% determined by semi quantitative method with blue stain SDS-PAGE and a recovery in excess of 80% as determined by peaks integration. As a comparison data from the association of ion exchange and immunoaffinity chromatography showed a quite close purity, but a limited recovery of about 40% [46].

The second example of an aptamer-based purification process deals with the separation of human Complement Factor H (recFH) from either a semi-purified human plasma preparation involving three chromatographic steps (heparin affinity chromatography, cation and anion exchange chromatography) or directly from a crude supernatant of recombinant cells releasing the humanized recFH protein. FH, a member of the regulators of complement activation family, is a large plasma glycoprotein (155 kDa) found in normal blood at a concentration of 200-400 |g/mL [47]. It is a low-abundance protein representing about 0.4% of the global plasma protein amount, but strongly actively involved in the protection action on self cells and surfaces [48]. It is generally purified by combining precipitations and classical chromatography steps. In the present work the efficiency of purification using the anti-Factor H aptamer is shown by the SDS-polyacrylamide gel electrophoresis illustrated in Fig. 3. The panel A on the left represents the separation of native Factor H from a prepurified fraction of human plasma. The eluted fraction from the immobilized aptamer appears as a single band evidencing thus the high level of selectivity. Interestingly a similar behavior is observed even when the initial material was a crude untreated cell culture supernatant comprising expressed recFH (Fig. 3, panel B). The initial loaded material is extremely rich in protein diversity where the target protein is almost undistin-guishable in the middle of numerous other proteins. However, FH eluted after a wash with 1 M sodium chloride was harvested as a single band and many diverse impurities observable from the crude supernatant are removed in a single step. Compared to the native plasma-purified version, recFH (arrow) has a similar degree of purity. This result demonstrates the extremely high selectivity of the aptamer ligand whatever the initial loaded material. To our knowledge, no purification process of FH has been reported to date with such a high degree of selectivity in a single fraction-ation step. Indeed, all the purification methods described in the literature involve several fractionation steps. Actually, when starting from either intermediates of cold ethanol plasma fractionation [49] or from polyethyleneglycol precipitation [50] three chromatographic separation steps are necessary to achieve good purification levels. The first column is based on the use of Lysine-Sepharose, followed by two others more classical steps such as anion exchange chromatography and gel filtration, a quite complex process indeed finalized by a slow separation procedure based on molecular mass discrimination.

The third example deals with the purification of human coagulation Factor IX. Also called Christmas Factor, this protein is a member of the vitamin K-dependent coagulation system. It is a glycosylated polypeptide chain having a molecular mass of 56 kDa. It is constituted of four main domains: (i) the interaction domain of EGF-1, (ii) the activation domain of EGF-2, (iii) the GLA domain interacting with calcium ions and (iv) the serine protease domain. Its normal concentration in blood plasma is 4.4 |g/mL. Its deficiency engenders the so-called hemophilia B which is treated by injections of purified FIX obtained from either blood plasma or from recombinant technology [51]. Purification trials of this protein were performed using an affinity sorbent carrying the aptamer named Nonapta5.1 (see sequence in Material section) from a pre-purified fraction of human plasmatic coagulation Factor IX still comprising a quite large number of impurities (estimated to more than 95%) as confirmed by the SDS-PAGE on Fig. 4. After injection of the biological material and a first washing step with 50 mM Tris-HCl pH 7.4 containing 150 mM NaCl 2mM, CaCl2 and 1 mM MgCl2 followed

by a second wash with 2 M sodium chloride, the captured Factor IX was eluted using 200 mM EDTA solution pH 8. The wash with sodium chloride does not show any significant amount of proteins in SDS-PAGE analysis while the eluate comprises a massive protein band located at the right molecular weight position (arrow). Identification of this band as FIX was unambiguously confirmed by mass spectrometry analysis. Two minor impurities are still present at a low molecular mass position. The biological activity of aptamer-purified FIX, determined as a ratio between the amidolytic activity and the antigen rate [Am]/[Ag], was perfectly preserved with a value of 1,1 (Table 1).

Currently the coagulation Factor IX is extracted from blood plasma and involves the use of anion exchange chromatography [52,53]. However, other chromatographic purification methods have been described such as the use of metal chelate affinity chromatography [54] and hydrophobic interaction chromatography [55]. More sophisticated purification processes have also been described involving several chromatographic steps such as two sequential anion exchange separations followed by heparin affinity chromatography [56]. In spite of these efforts the final pure product was still containing a number of protein impurities evidenced by proteomics technology. Immunoaffinity chromatography procedures providing very pure Factor IX have also been described after having immobilized either mouse monoclonal IgG antibodies [57] or conformation-specific antibodies grafted on chromatographic sorbents [58]. When starting from human plasma it is not absolutely necessary to obtain highly pure products. However, the large request of coagulation Factor IX necessitated the development of recombinant organisms that require extensive purification to remove all impurity traces. This approach inevitably engendered sophisticated purification processes to get rid of all heterolo-gous proteins that can have undesirable side effects upon human injection. Recombinant FIX has been produced either by culturing recombinant cells [59] or secreted in the milk of transgenic mammals [60,61]. From these unconventional sources the purification was performed using either non-immunoaffinity chromatography methods [60,61] or antibody-based chromatography [62,63]. As a general conclusion of the purification processes, significant data are summarized in Table 1.

3.4. Aptamer ligands to distinguish native proteins versus variants

Discriminating protein variants that are truncated or incompletely post-translationally modified from their native integral forms is essential to select the highest biological activity for proteins intended for in vivo therapeutic treatments. This feature appears reachable with the use of aptamer ligands targeting the GLA epitope. This capability is exemplified below with two coagulation factors.

As many proteins, coagulation Factor VII and Factor IX are post-translationally modified to acquire specific binding properties or its full biological activity. In this case and also for a number of other vitamin K-dependent plasma proteins involved in the coagulation cascade, a 7-carboxylation process takes place at glutamic acid residue levels. In the so-called GLA domain the carboxyla-tion of the 7-carbons is catalyzed by the 7-glutamyl carboxylase [64,65]. This post-translational modification is required to confer to GLA domain binding properties to calcium and phospholipids and to acquire the necessary protease activity for the coagulation effect [66]. While this protein modification is present during its maturation in human cells, recombinant proteins contain a certain amount of incompletely modified proteins with the absence or partial presence of GLA domain (des-GLA forms). This engenders protein isoforms with largely reduced biological activity that must be eliminated during the purification process. With the use

■■IBBI^BaRTiCLe in press

C. Forieret al./J. Chromatogr. A xxx (2017)xxx-xxx

Fig. 3. Analytical determinations by SDS-polyacrylamide gel electrophoresis of the aptamer-purified complement Factor H from either a pre-purified plasma fraction (A) or from a crude cell culture supernatant where FH was expressed (B). In both cases after the injection of the material to be purified (Start), the columns were washed with a 1 M sodium chloride solution followed by the elution (El) by 50 mM Tris-HCl buffer containing 100 mM EDTA, pH 8. 'Ft' represents the flowthrough or the non-retained fraction. The arrows indicate the positioning of FH. The polyacrylamide gel concentration of the plate was 4-12%; the migration was performed under 150 V for 50min as perthe recommendation from the supplier. Separated protein bands were revealed by silver staining for an increased sensitivity.

Fig. 4. Purification of coagulation Factor IX by aptamer affinity chromatography.

A: Chromatographic profile obtained by the injection of an intermediated human plasma fraction containing FIX. The equilibration buffer was 50 mM Tris-HCl pH 7.4 containing 150 mM sodium chloride, 2 mM calcium chloride and 1 mM magnesium chloride. A wash to push out non retained proteins was performed using 0.5 M sodium chloride in the equilibration buffer. The elution was operated by 200 mM EDTA pH 8. Ft: flowthrough; W: wash proteins; El: eluted protein fraction. B: SDS polyacrylamide gel electrophoresis of the fractionated FIX containing sample. The first lane (Mr) from the left represents the molecular weight markers; the second lane (Start) represents the initial injected material; the third lane shows the flowthrough fraction (Ft). Elution (El) of purified Factor IX, (see arrow) is represented by the lane on the right. The polyacrylamide gel concentration of the plate was 4-12%; the migration was performed under 150 V for 50min as per the recommendation from the supplier. Protein separated bands were stained with Coomassie blue.

of aptamer ligands we discovered that the molecular recognition of protein epitopes is so specific that it is possible to discriminate between active proteins and des-GLA forms. This possibility has been experienced using first a model sample and then applied to a concrete case. As a model, food cow milk was spiked at 0.21 mg/mL of pure human plasma coagulation Factor VII containing a significant part of des-GLA forms. This sample was directly submitted to affinity separation on anti FVII-GLA Mapt2.2CS aptamer ligand

according to technical details described in Method section and on Fig. 5.

From the SDS-polyacrylamide gel electrophoresis (Fig. 5) it is apparent that the initial sample (Pre-FVII) comprises both the active FVII and the des-GLA forms. During the affinity chromatography process, the des-GLA forms, very similar to the active protein are fully eliminated, demonstrating the extreme discrimination capability of the aptamer ligand. The lower molecular mass of des-GLA variant result from two distinct phenomena: (i) the absence of

аддняяя^и article in press

8 C. Forier et al. /J. Chromatogr. A xxx (2017)xxx-xxx

Table 1

Summary of the affinity chromatography separation of plasma proteins. Purity determined by semi quantitative method with blue stain SDS-PAGE. The biochemical activities were expressed differently: as a ratio between the amidolytic activity and the antigen rate [Am]/[Ag] for coagulation factors (FVII and FIX) and as the quantity of proteins required to inhibit complement mediated lysis for FH.

Target purification Fraction Biological activities Purity (%)

FVII Starting material 0.7 [Am]/[Ag] 51

Eluate 0.9 [Am]/[Ag] 100

FH Starting material 6 ^g (qty of protein to protect RBC) 74

Eluate 4 |xg (qty ofprotein to protect RBC) 99

FIX Starting material 1,1 [Am]/[Ag] 0.5

Eluate 1,1 [Am]/[Ag] 98

Fig. 5. SDS-PAGE analysis of coagulation Factor Vll using the anti FVll-GLA aptamer ligand Mapt2.2CS. Mr: molecular weight markers; PreFVll: prepurified human plasma Factor VII; Ft: flowthrough from aptamer affinity column; El: eluate from aptamer affinity column. The two forms of FVII are present in the initial sample while the des-GLA variant is not any longer present after chromatographic separation.

Adsorption buffer: 50 mM Tris-HCl buffer containing 50 mM NaCl, 10 mM CaCl2, pH 7.5; elution buffer: 50 mM Tris-HCl buffer containing 10 mM, EDTA, pH 7.5. The poly-acrylamide gel concentration of the plate was 4-12%; the migration was performed under 150 V for 50 min as per the recommendation from the supplier. Separated protein bands were revealed by silver staining.

glutamic acid carboxylation and (ii) the partial cleavage of the GLA-domain [67] ln the lane El which represents the eluted proteins, FVll is present as a major component while des-GLA FVll is fully absent. Other proteins found in the eluate (El) are identified as being FVll fragments containing the GLA-domain.

The real case here described is a transgenic swine milk containing a recombinant human Factor lX. ln this biological sample, FlX has a variable degree of 7-carboxylation. To separate the highly 7-carboxylated forms, GLA-specific aptamer was used as chro-matographic ligand. The initial transgenic milk material was first clarified by precipitation in the presence of citrate buffer and the supernatant fractionated on MEP-HyperCel sorbent. Then the fraction comprising the FlX variants (about 2% of the protein content) was submitted to aptamer affinity chromatography. The chromatographic profile of this purification step is shown on Fig. 6 (panel A). The majority of material did not bind the sorbent (flowthrough Ft); a small peak composed of active FlX was eluted and a third peak of protein impurities was collected during column regeneration. The analytical data from SDS-PAGE (Fig. 6 panel B) show a dominant band at the elution lane out of a very large number of other proteins as shown on the lane corresponding to the initial material (Start). The pure band of FlX appears similar to the plasma reference FlX on the right of the figure (Control). These data were confirmed

over three runs. The degree of 7 - carboxylation within the initial material was determined by mass spectrometry between 1-7 out of 12, with a high majority of low 7-carboxylated species (data not shown). This low 7-carboxylation degree corresponds to a biological activity determined as a ratio between the amidolytic activity and the antigen rate [Am]/[Ag] that was around 0.2 before affinity separation. After separation, the biological activity was increased up to 0.7 due to the capture and concentration of the rare highest 7-carboxylated species.

3.5. Applicability of an aptamer to the separation of either native or recombinant proteins

Experiments were performed to compare the separation of native plasma-related proteins from blood or of recombinant forms from either cell culture or from transgenic animal milk.

Under similar conditions it has been demonstrated that with a same aptamer ligand it was possible to maintain the selectivity against many different protein impurities independently on the species. For instance in Fig. 3 (panel B) FH was purified as a single step from a recombinant cell culture supernatant compared to a similar separation from a human plasma fraction (panel A). In spite of a much larger number of protein impurities, the recombinant

üb^b article in press

C. Forieret al./J. Chromatogr. A xxx (2017)xxx-xxx

Fig. 6. Purification of transgenic human Factor IX from swine milk. The initial material is first clarified by a citrate treatment and then the protein mixture prefractionated on MEP-HyperCeI. The fraction containing correctly carboxylated forms is then separated using an aptamer specific for GLA domain.

Panel A: chromatographic profile of aptamer affinity chromatography; protein composing the initial material (sample) were mostly not captured since the presence of FIX was only about 2%. Elution was performed by 50 mM Tris-HCl, containing 10 mM CaCI2,4mM MgCI2 and 10 mM EDTA. Then the column was regenerated (Reg) for following cycles by using 20 mM Tris-HCI, 1 M NaCI, 5% propylene glycol, pH 7.5. Panel B: SDS-PAGE analysis. Mr: molecular weight markers; Start: initial biological material (prefractionated transgenic milk); EI: desorbed protein from aptamer; Control: reference Factor IX from human plasma purified using the same aptamer ligand. The polyacrylamide gel concentration of the pIate was 4-12%; the migration was performed under 150 V for 50 min as per the recommendation from the suppIier. Separated protein bands were reveaIed by siIver staining

form of FH is separated as a single band. Moreover, when the biological activity of the purified recFH was determined, the amount needed to fully protect red blood cells from lysis was 4 |g compared to 6 |g for the currently available plasmatic FH reference. This value is identical to the aptamopurified plasmatic FH In other words, the aptamer affinity chromatography separation process not only preserved the biological activity but, additionally, allowed to obtain a product with better purity by the selection of most physiologically close forms and hence with a significantly higher specific activity.

As described above, human recombinant coagulation Factor IX was expressed in swine and then purified using aptamer affinity chromatography with purity in excess of 95% (see Fig. 6). As a confirmation, in another example recombinant human coagulation Factor X has been purified with the anti-GLA domain Mapt1.2CS from a cell culture supernatant with an impressive result. The initial purity of the target protein was estimated as 0.7% among a large number of other cell supernatant proteins and after separation through an aptamer affinity column the purity obtained by SDS-PAGE became close to 87% (data not shown).

In conclusion the same DNA aptamer can be used both for the purification of native and recombinant proteins preserving the capability to remove impurities that are not the same in each condition.

3.6. Aptamer-based affinity chromatography compared to immunoaffinity purification

Compared to immunoaffinity chromatography, a currently accepted specific protein purification method, DNA aptamer lig-ands exhibit affinities for target proteins similar or higher than those of antibodies, with Kd often in the low nanomolar to pico-molar range. They further present a number of advantages over antibodies for preparative industrial separation applications as described [68-70].

Antibodies as specific ligands for affinity chromatography are used under their different forms such as full structure or as trun-

cated protein from various animals including camelides that are single chain antibodies [3] and can be used as quite small fragments [71]. The major interest of immunoaffinity chromatography is the specificity for the antigen (the protein to separate) that was unchallenged up to the advent of aptamers. The latter being considered relatively equivalent or superior to antibodies [23] have other interesting properties conferring decisive advantages such as the biological stability (proteases have no deleterious action on DNA aptamer ligands), the stability under caustic conditions and the easiness of a full oriented grafting on solid supports. These features are not possible with antibody ligands except oriented immobilization at a price of quite complex reaction processes [72]. The selection of the appropriate antibody for a given purification is also very laborious [13] while sorting out the DNA aptamer ligands is generally performed in vitro with a well-established cycling/amplification process [13]. Another major advantage over the antibodies-based ligands is the possibility during the SELEX process to parameter and to choose the most suitable elution condition. This is not always the case with antibody-based ligands since the elution is frequently a compromise between elution efficiency and deleterious effects. Once identified these ligands are to be produced at a scale to withstand the needs of chemical coupling on solid support. This process is performed by chemical synthesis for DNA aptamers but requires a cell culture to produce antibodies that have to be further purified prior immobilization. Clearly the cost involved to use immunoaffin-ity chromatography is significantly higher compared to aptamer affinity purification an advantage that is essential at preparative industrial scale. Due to their small size (5-20 kDa compared to 150-160 kDa for native antibodies), aptamers can be chemically grafted at high ligand densities, thus increasing the binding capacity which is another decisive economical advantage. In terms of specificity these two types of affinity sorbents have similar performance as attested in Fig. 7 where immunopurified Factor IX are compared to the same protein purified using the described Nonapta5.1 aptamer ligand. More intensive comparative studies are however necessary to demonstrate the selectivity benefits of

s^ article in press

C. Forieretal./J. Chromatogr. A xxx (2017)xxx-xxx

Fig. 7. Comparative purification of FIX by between aptamer affinity chromatography and immunoaffinity chromatography.

From lanes 1 to lane 6: molecularweight markers; current commercially available Factor IX concentrate (e.g. Betafact); aptamer-purified FIX from Betafact; recombinant pure FIX(Benefix: purified by a four-step chromatography purification process); immunopurified FIX(Mononine): molecularweight markers. The polyacrylamide gel concentration of the plate was 4-12%; the migration was performed under 150 V for 50 min as per the recommendation from the supplier. Separated protein bands were revealed by silver staining for an increased sensitivity

aptamer ligands over antibody affinity ligands. It is to be noticed that the specificity of an aptamer highly depend on a very accurate selection and criteria of selection that must consider the sample complexity, the adsorption conditions and the elution conditions.

3.7. Stability of aptamer affinity solid-phase media

The efficiency of aptamer-based affinity chromatography to purify proteins was strengthened by its stability over time throughout stringent cleaning steps between cycles. At the end of each purification cycle the sorbents was submitted to a sanitization step involving 0.1 M or 1 M sodium hydroxide prior re-equilibration in the initial buffer. As described above in previous section it has been demonstrated that the aptamer affinity sorbent withstand a caustic step a the end of the separation. However in order to demonstrate the long term stability, stringent caustic treatment has been performed. The affinity sorbent was incubated first 100 h in the presence of 1 M sodium hydroxide, followed by 100 additional hours in clarified milk and then submitted to thirty sequential separation cycles including cleaning cycles with 1 M sodium hydroxide. It is at the issue of this harsh treatment that we noticed an only slight decrease of the binding capacity estimated at 18% corresponding to about 0.14% per cycle (Table 2). In addition the biochemical activity of eluates did not seem modified. Taken together, these results reinforce the capability of such aptamer-based sorbents for preparative industrial applications.

The adsorption capacity of the described solid-phase affinity media was about 6 mg/mL which is significant for high value

Table 2

FVII binding capacity (given for 80% breakthrough) of grafted Mapt2.2CS submitted to various stringent prolonged treatments. The biochemical activity was also measured for FVII eluates.

Treatment condition Binding capacity at 80% Activity ratio of

breakthrough FVII eluates

(mg/mL of gel)

Starting conditions 7.6 1.0

After 100 hours in 1 M NaOH 6.9 0.9

After 100 hours in clarified milk 7.1 0.9

Then 30 cycles in 1 M NaOH 6.2 1.0

and rare proteins. Nevertheless when considering current low-medium-value proteins as for example antibodies, such a binding capacity would be qualified as critical. However, if one takes as a reference the separation of immunoglobulins G using a RNA aptamer which was reported as being as high as 30 mg/mL [34] it is believed that there are still room for improvement. Increased aptamer ligand density, enhanced aptamer ligand accessibility, stabilization of the active conformation using modified nucleotides or relevant mutations, more adapted nature and length of the spacer as well the fine-tuned strategy of chemical immobilization may play in favor of enhanced capture efficiency while preserving the purity of the final product. Efforts in this direction are presently ongoing in our laboratories.

шмивии ar mile in press

C. Forier et al. /J. Chromatogr. A xxx (2017)xxx-xxx 11

4. Conclusion

While the affinity of selected aptamers for proteins is not any longer to be demonstrated, the present report shows their selectivity interest as affinity chromatography ligands intended for preparative industrial applications.

The reported real life examples are significant of the potential of this material for the isolation of a pure target protein while ignoring most of protein impurities at a point that single step purification appears possible at preparative scale. The same aptamer can advantageously be used for the separation of the same protein whatever its native origin or recombinant from cultured cells or transgeni-cally produced.

The potential ligands retrievable from aptamer libraries being very large, only few of them have been tested. It is therefore probable that better sequences could be identified with reduced size without scarifying the selectivity for a given protein. Such a finding can further concur to improve final results in terms of binding capacity, purity and long term stability over multiple reutilization cycles.

Issues related to ligand traces released during separation runs and progressive degradation upon caustic treatments have not been observed during the fractionation of described proteins. However, the separation process was performed over a limited number of cycles where traces of possible degradations may be difficult to demonstrate. Our team is currently checking for aptamer affinity property deviations on long term about binding capacity, specificity, ligand release and non-specific binding.

Acknowledgements

We gratefully acknowledge Bruno Picot for FH activity determinations, Frédéric Dhainaut and members of his team, Laurence Boulet and Martine Martin for FVII and FIX activity determinations; Alain Lejars, Béatrice Souilliart and Camille Bechetoille for their continuous technical purification support; Catherine Groseil for mass spectrometry investigations of protein impurities from electrophoresis analysis. We are also sincerely grateful to Laurent Siret for his initial strong support.

References

[ 1] S. Sheng, F. Kong, Separation of antigens and antibodies by immunoaffinity chromatography, Pharm. BioI. 50 (2012) 1038-1044.

[ 2] J. Fitzgerald, P. Leonard, E. Darcy, R. O'Kennedy, Immunoaffinity chromatography, Methods MoI. BioI. 681 (2011) 35-59.

[ 3] D. Reinhart, R. Weik, R. Kunert, Recombinant IgA production: singIe step affinity purification using cameIid Iigands and product characterization, J. ImmunoI. Methods. 378 (2012) 95-101.

[ 4] H. WaIIberg, H.P.K. LofdahI, K. TschapaIda, M. UhIem, V. ToImachev, P.K. Nygren, S. StahI, Affinity recovery of eight HER2-binding affibody variants using an anti-idiotypic affibody moIecuIe as capture Iigand, Prot. Exp. Purif. 76(2011) 127-135.

[ 5] J. Pander, M.M. Szewczyk, A.K. Grover, Phage dispIay: concept, innovations, appIications and future, BiotechnoI. Adv. 28 (2010) 849-858.

[ 6] P. MiIovnik, D. Ferrari, C.A. Sarkar, A. PIuckthun, SeIection and

characterization of DARPins specific for the neurotensin receptor 1, Protein Eng. Des. SeI. 22 (2009) 357-366.

[ 7] L.K.J. StadIer, T. Hoffmann, D.C. TomIinson, Q. Song, T. Lee, M. Busby, Y. Nyathi, E. Gendra, C. Tiede, K FIanagan, S.J. CockeII, A. Wipat, C. Harwood, S.D. Wagner, M.A. KnowIes, J.J. Davis, N. Keegan, P.K. Ferrigno, Structure-function studies of an engineered scaffoId protein derived from Stefin A. II: DeveIopment and appIications of the SQT variant, Protein Eng. Des. SeI. 24 (2011)751-763.

8] H. Yang, P.V. GurgeI, R.G. CarboneII, Purification ofhuman immunogIobuIin G via Fc-specific smaII peptide Iigand affinity chromatography, J. Chromatogr. A 1216(2009)910-918.

[ 9] L.N. Lund, P.E. Gustavsson, R. MichaeI,J. Lindgren, L. Norskov-Lauritsen, M. Lund, G. Houen, A. Staby, P.M. St HiIaire, NoveI peptide Iigand with high binding capacity for antibody purification, J. Chromatogr. A. 1225 (2012) 158-167.

10] T. Kodadek, К Bachhawat-Sikder, Optimized protocols forthe isolation of specific protein-binding peptides or peptoids from combinatorial libraries displayed on beads, Mol. Biosyst. 2 (2006) 25-35.

11] A.C. Roque, G. Gupta, C.R. Lowe, Design synthesis, and screening of biomimetic ligands for affinity chromatography, Methods Mol. Biol. 310 (2005) 43-62.

12] G. El Khoury, Y. Wang, D. Wang, S.I.Jacob, C.R. Lowe, Design synthesis, and assessment of a de novo affinity adsorbent forthe purification of recombinant human erythropoietin, Biotechnol. Bioeng. 10 (2013) 3063-3069.

13] G. Perret, P. Santambien, E. Boschetti, The quest for affinity chromatography ligands: are the molecular libraries the right source? J. Sep. Sci. 38 (2015) 2559-2572.

14] C. Tuerk, L. Gold, Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase, Science 249 (1990) 505-510.

15] A.D. Ellington, J.W. Szostak, SELEX-a (r)evolutionary method to generate high-affinity nucleic acid ligands. Nature 346 (1990) 818-822.

16] A. Cibiel, D. Miotto Dupont, F. Ducongé, Methods to identify aptamers against cell surface biomarkers, Pharmaceuticals 4 (2011) 1216-1235.

17] S. Klussman, A. Nolte, R. Bald, V.A. Erdmann, J.P. Furste, Mirror-image RNA that binds D-adenosine, Nat. Biotechnol. 14 (1996) 1112-1116.

18] G. Mayer, The chemical biology of aptamers, Angew. Chem. Int. Ed. Engl. 48

(2009) 2672-2689.

19] X. Pei, J. Zhang, J. Liu, Clinical applications of nucleic acid aptamers in cancer, Mol. Clin. Oncol. 2 (2014) 341-348.

20] S. Centi, S. Tombelli, M. Minunni, M. Mascini, Aptamer-based detection of plasma proteins by an electrochemical assay coupled to magnetic beads, Anal. Chem. 79 (2007) 1466-1473.

21] M. Liss, B. Petersen, H. Wolf, E. Prohaska, An aptamer-based quartz crystal protein biosensor, Anal. Chem. 74 (2002) 4488-4495.

22] Q. Zhao, X.F. Li, Y. Shao, X.C. Le, Aptamer-based affinity chromatographic assays for thrombin, Anal. Chem. 80 (2008) 7586-7593.

23] Ö. Kökpinar, J.G. Walter, Y. Shoham, F. Stahl, T. Scheper, Aptamer-based downstream processing of his-tagged proteins utilizing magnetic beads, Biotechnol. Bioeng. 108 (2011) 2371-2379.

24] B. Han, C. Zhao, J.Y.H. Wang, High performance aptamer affinity chromatography for single-step selective extraction and screening of basic protein lysozyme,J. Chromatogr. 903 (2012) 112-117.

25] F. Bartnicki, E. Kowalska, K. Pels, W. Strzalka, Imidazole-free purification of His3-tagged recombinant proteins using ssDNA aptamer-based affinity chromatography, J. Chromatogr. A 1418 (2015) 130-139.

26] R. Ahirwar, P. Nahar, Development of an aptamer-affinity chromatography for efficient single step purification of Concanavalin A from Canavalia ensiformis, J. Chromatogr. B 997 (2015) 105-109.

27] T.S. Romig, C. Bell, D.W. Drolet, Aptamer affinity chromatography: combinatorial chemistry applied to protein purification, J. Chromatogr. B 731 (1999) 275-284.

28] Q. Deng, I. German, D. Buchanan, R.T. Kennedy, Retention and separation of adenosine and analogues by affinity chromatography with an aptamer stationary phase, Anal. Chem. 73 (2001) 5415-5421.

29] M.B. Murphy, S.T. Fuller, P.M. Richardson, S.A. Doyle, An improved method for the in vitro evolution of aptamers and applications in protein detection and purification, Nucl. Acids Res. 31 (2003) e110.

30] J. Ruta C. Grosset, C. Ravelet, J. Fize, A. Villet, A. Ravel, E. Peyrin, Chiral resolution ofhistidine using an anti-D-histidine L-RNA aptamer microbore column, J. Chromatogr. B. 845 (2007) 186-190.

31] Q. Zhao, X.F. Li, X.C. Le, Aptamer-modified monolithic capillary chromatography for protein separation and detection, Anal. Chem. 80 (2008) 3915-3920.

32] M. Michaud, E. Jourdan, A. Villet, A. Ravel, C. Grosset, E. Peyrin, A DNA aptamer as a new target-specific chiral selector for HPLC, J. Am. Chem. Soc. 125(2003)8672-8679.

33] J.A. Walter, F. Stahl, T. Scheper, Aptamers as affinity ligands for downstream processing, Eng. Life Sci. 12 (2012) 1-11.

34] S. Miyakawa, Y. Nomura, T. Sakamoto, Y. Yamaguchi, K. Kato, S. Yamazaki, Y. Nakamura, Structural and molecular basis for hyperspecificity of RNA aptamer to human immunoglobulin G, RNA 14 (2008) 1-10.

35] A.C. Connor, L.B. McGown, Aptamer stationary phase for protein capture in affinity capillary chromatography, J. Chromatogr. A 1111 (2006) 115-119.

36] H.A. Oktem, G. Bayramoglu, V.C. Ozalp, M.Y. Arica, Single-step purification of recombinant Thermus aquaticus DNA polymerase using DNA-aptamer immobilized novel affinity magnetic beads, Biotechnol. Prog. 23 (2007) 146-154.

37] G. Perret, F. Ducongé, Patent Application WO/2010/094899 (2009).

38] G. Perret, M. Ouhammouch, Patent Application WO/2015/044923 (2014).

39] G. Perret, A. Cibiel, Patent Application WO/2013/080134 (2012).

40] G. Perret, N., Bihoreau, L. Siret, Patent Application WO/2011/012831 (2010).

41] G. Perret, S. Chtourou, N. Bihoreau, Patent Application WO/2011/012830

(2010).

42] E. Boschetti, G. Perret, Patent Application WO/2012/090183 (2011).

43] P. Sánchez-Corral, C. González-Rubio, S. Rodríguez de Córdoba, M. López-Trascasa, Functional analysis in serum from atypical Hemolytic Uremic Syndrome patients reveals impaired protection of host cells associated with mutations in factor H, Mol. Immunol. 41 (2004) 81-84.

44] C.L. Brucato, C.A. Birr, P. Bruguera, J.A. Ruiz, D. Sánchez-Martínez, Expression of recombinant rabbit tissue factor in Pichia pastoris, and its application in a prothrombin time reagent, Prot. Exp. Purif. 26 (2002) 386-393.

hss—^bm article in press

C. Forieretal./J. Chromatogr. A xxx (2017)xxx-xxx

[45] B. Juriander, L. Thim, N.K. Klausen, E. Persson, M. Kjalke, P. Rexen, T.B. Jorgensen, P.B. 0stergaard, E. Erhardtsen, S.E. Bj0rn, Recombinant activated factor VII (rFVIla): characterization, manufacturing, and clinical development, Semin. Thromb. Hemost. 27 (2001) 373-384.

[46] K. Tomokiyo, H. Yano, M. Imamura, Y. Nakano, T. Nakagaki, Y. Ogata, T. Terano, S. Miyamoto, A. Funatsu, Large-scale production and properties of human plasma-derived activated Factor Vll concentrate, Vox Sang. 84 (2003) 54-64.

[47] M.K. Pangburn, Host recognition and target differentiation by factor H, a regulator of the alternative pathway of complement, Immunopharmacology 49(2000) 149-157.

[48] S. Hakobyan, C.L. Harris, A. Tortajada, E. Goicochea de Jorge, A. García-Layana, P. Fernández-Robredo, S. Rodríguez de Córdoba, B.P. Morgan, Measurement of factor H variants in plasma using variant-specific monoclonal antibodies: application to assessing risk of age-related macular degeneration, Invest. Ophthalmol. Vis. Sci. 49 (2008) 1983-1990.

[49] H. Brandstätter, P. Schulz, I. Polunic, C. Kannicht, G. Kohla, J. Römisch, Purification and biochemical characterization of functional complement factor H from human plasma fractions, Vox Sang. 103 (2012) 201-212.

[50] F.M. Wang, F. Yu, M.H. Zhao, A method of purifying intact complement factor H from human plasma, Prot. Exp. Purif. 91 (2013) 105-111.

[51] L.D. Taran, Factor IX of the blood coagulation system: a review, Biochem. Mosc. 62(1997)685-693.

[52] D. Stampe, B. Wieland, A. Köhle, Isolation of factor IX concentrates for clinical use by ion-exchange chromatography and ammonium sulphate precipitation, J. Chromatogr. 363 (1986) 101-103.

[53] L. Hoffer, H. Schwinn, D. Josic, Production of highly purified clotting factor IX by a combination of different chromatographic methods, J. Chromatogr. A 884 (1999) 119-128.

[54] P.A. Feldman, P.l. Bradbury, J.D. Williams, G.E. Sims, J.W. McPhee, M.A. Pinnell, L. Harris, G.l. Crombiel, D.R. Evans, Large-scale preparation and biochemical characterization ofa new high purity factor lX concentrate prepared by metal chelate affinity chromatography, Blood Coagul. Fibrinolysis. 5 (1994) 939-948.

[55] H. Husi, M.D. Walkinshaw, Separation of human vitamin K-dependent coagulation proteins using hydrophobic interaction chromatography, J. Chromatogr. B. 736 (1999) 77-88.

[56] J. Clifton, F. Huang, D. Gaso-Sokac, K. Brilliant, D. Hixson, D. Josic, Use of proteomics for validation of the isolation process of clotting factor IX from human plasma, J. Prot. 73 (2010) 678-688.

[57] H.C. Kim, C.W. McMillan, G.C. White, G.E. Bergman, M.W. Horton, P. Saidi, Purified factor IX using monoclonal immunoaffinity technique: clinical trials in hemophilia B and comparison to prothrombin complex concentrates, Blood 79(1992) 568-575.

[58] H.A. Liebman, S.A. Limentani, B.C. Furie, B. Furie, lmmunoaffinity purification of factor lX (Christmas factor) by using conformation-specific antibodies directed against the factor IX-metal complex, Proc. Natl. Acad. Sci. U. S. A. 82 (1985)3879-3883.

[59] S. Harrison, S. Adamson, D. Bonam, S. Brodeur, T. Charlebois, B. Clancy, R. Costigan, D. Drapeau, M. Hamilton, K. Hanley, B. Kelley, A. Knight, M. Leonard, M. McCarthy, P. Oakes, K. Sterl, M. Switzer, R. Walsh, W. Foster, The manufacturing process for recombinant factor IX, Semin. Hematol. 35 (1998) 4-10.

[60] M. Lindsay, G.C. Gil, A. Cadiz, W.H. Velander, C. Zhang, K.E. Van Cott, Purification of recombinant DNA-derived factor IX produced in transgenic pig milk and fractionation of active and inactive subpopulations, J. Chromatogr. A. 1026(2004) 149-157.

[61] Y.L. Sun, Y.S. Chang, Y.S. Lin, C.H. Yen, Pilot production of recombinant human clotting factor IX from transgenic sow milk, J. Chromatogr. B 898 (2012) 78-89.

[62] P. Meulien, A. Balland, P. Lepage, F. Mischler, K. Dott, C. Hauss, M. Grandgeorge, J.P. Lecocq, Increased biological activity of a recombinant factor IX variant carrying alanine at position +1, Protein Eng. 3 (1990) 629-633.

[63] S. Jallat, F. Perraud, W. Dalemans, A. Balland, A. Dieterle, T. Faure, P. Meulien, A. Pavirani, Characterization of recombinant human factor IX expressed in transgenic mice and in derived trans-immortalized hepatic cell lines, EMBO J. 9(1990)3295-3301.

[64] J. Suttie, Vitamin K-dependent carboxylase, Annu. Rev. Biochem. 54 (1985) 459-477.

[65] S.M. Wu, W.F. Cheung, D. Frazier, D. Stafford, Cloning and expression of the cDNA for human gamma-glutamyl carboxylase, Science 254 (1991) 1634-1636.

[66] K. Mann, M. Nesheim, W. Church, P. Haley, S. Krishnaswamy, Surface-dependent reactions ofthe vitamin K-dependent enzyme complexes, Blood 76(1990)1-16.

[67] L. Thim, S. Bjoern, M. Christensen, E.M. Nicolaisen, T. Lund-Hansen, A.H. Pedersen, U. Hedner, Amino acid sequence and posttranslational modifications of human factor. Vila from plasma and transfected baby hamster kidney cells, Biochemistry 27 (1988) 7785-7793.

[68] V.J.B. Ruigrok, M. Levisson, M.H.M. Eppink, H. Smidt, J. Van Der Oost, Alternative affinity tools: more attractive than antibodies? Biopchem. J. 436 (2011)1-13.

[69] A.B. iliuk, L. Hu, W.A. Tao, Aptamer in bioanalytical applications, Anal. Chem. 83(2011)4440-4452.

[70] K.-M. Song, S. Lee, C. Ban, Aptamers and their biological applications, Sensors 12(2012)612-631.

[71] C. Hamers-Casterman, T. Atarhouch, S. Muyldermans, G. Robinson, C. Hamers, E.B. Songa, N. Bendahman, R. Hamers, Naturally occurring antibodies devoid of light chains, Nature 363 (1993) 446-448.

[72] G. Fleminger, E. Hadas, T. Wolf, B. Solomon, Oriented immobilization of periodate-oxidized monoclonal antibodies on amino and hydrazide derivatives of Eupergit C, Appl. Biochem. Biotechnol. 23 (1990) 123-137.