Introduction

Small genetically engineered immuoglobical constructs are being developed industry wide for a growing range of in vivo applications. Examples include Fab, F(ab’)2, single-chain (sc) Fv, bis-scFV, diabodies, minibodies and single domain antibodies. Their small size potentially tives them access to tissues that are poorly accessible by intact antibodies; rpaid clearance form blood and neotargeted tissues, lower immunogenic response and eye-drop inhalent or oral administration. (Gagnon, “Minibodies and Multimodal chromatography methods” BioProcess Interational Feb 2010).

Domain antibodies are characterized by formation of the antigen binding site by a single antibody domain, which does not require interaction with a futher domain (e.g., in the form of VH/VL interaction) for antigen recognition. Produciton of nanobodies, as one specific example of a domain antibody, in lower eukaryotic hosts such as Pichia pastoris is described in WO94/25591. In contrast to difficulties observed with conventional four chain antibodies or their fragments, including Fab and scFv, domain antibodies can be readily expressed and secreted in a correctly folded, fully funcitonal form from hosts like E. coli or P. pastoris at a sufficient rate and level. Schotte (US13/266503 and US2012/0157664)

Fab fragments:

Fab fragmebts are classically generated by treating immunoglobuilin (Ig) with the protease papain and isolating a 50-kDa disulfide-linked intact clevage product composed of a heavy + light chain fragment that contains only one antigen binding site. Unlke their bivalent Ig precursor, monovalent Fab fragments can bind their targets without cross-linking and thus possess significant potentail as reagents that can block receptors and signaling pathways. However, few Fab fragments are used in clinical applications because Fabs display a relatively short serum half-life, although streategies exist to address this issue such as conjugating Fabs to PEG or engineering Fabs as fusion proteins. The second limigation involves the common observation that Fab prepatations may aggregate as a function of time, concentraiton, temrpature or salt. If a Fab prepration is composed mostly of monovlaent species but contains a small contaminant of prtoein aggregate, the few multivalent complexes can display disproportionate activity becasue they intrinsically possess higher potential for antigen binding avidity and cross-oinking and thus may out-compete the potentail clocking effects of the more numerous monovalent Fabs. (Nelson, J. Biological Chemistry, 287(51) pp. 42936-42950). 

Particular Schemes for Purification of Domain Antibodies

Difficulties with purifying domain antibodies/Importance of conditions:

Gagnon, “Minibodies and Multimodal chromatography methods” BioProcess Interational Feb 2010) discloses that due to their architecture in lacking bidning sties for protein A, the principal enabling too for IgG platform purificaiton is inapplicable for mibodies. With respect to purificaiton of an antiprostate stem-cell antigen )PSCA minibody, the minibody did not bind to protein A and coeluted with BSA to a substantial degree on CEX, AEX, hydorphobic-interaction and hydroxyapatite (eluted by phosphate gradient chromatographic methods).

Affinity Chromatography:

–Protein A:

Beirnaert (US 2010/0297111) teaches nanobodies which includes variable domains present in naturally occurring heavy chain antibodies (referred to as VVHH domains) against TNF alpha using Protein A affinity purification including separation of PEGlyated and non PEGlyated nanobides using cation exchange chromatography.

Basran (WO/2009/074634) discloses that although Protein A based chromatography resins have been extensively used to purify VH dAbs from microbial culture supernatants in a single step, for some molecules the low pH elution conditions can result in the formation of aggregates. There is alo the issue of the limited capactiy of affintiy resins for dAbs. Mixed modal charge induction resin on Capto MMC from GE Healthcare was thus used as the primary capture step follows by anion exchange. Column equilibration was performed using 50 mM sodium phosphate pH 6. After washing the protein is eluted by pH gradient using 50 mM Tris pH 9.0.

Brown (US2010/0172894) discloses purifying single domain antigen binding (SDAB) molecules using Protein A based affintiy chromatography. In one embodiment, the elution and load buffers include sodium phosphate, sodium chloride. In other embodiments, the equilibration and wash buffer include sodium phosphate, sodium chloride and arginine.

Casterman (WO/1994/004678) discloses an isolated immunoglobulin comprising two heavy polypeptide chains sufficient for the formation of a complete antigen binding site which are devoid of light polypeptide chains (camelid antibodies) which are purified by Protein A or G chromatography.

Frenken (WO/1994/025591) discloses the production of antibody fragments derived from heavy chain immunoglobulins of Camelidae by eukaryoties such as yeast and fungi using Protein A chromatography.

—-Protein A – CEX –AEX

Stals (US 2012/0141460) discloses purifying a bivalent nanobody using Protein A affintity chromatography and eluting the nanbody using 100 mM Glycine pH 2.5. After neutralization the nanbody was further purified using CEX and an additional AEX step. 

—-VH3 Domain Antibodies

The term “VH3 domain” refers to the framework subgroup 3 of the human heavy chain variable region of an immunoglobulin. The heavy chain variable domains of antibodies are classified into distinct subfamilies (VH1 to VH6) on the basis of DNA sequence and protein homologies. (Slough, US Patent Application No: 15/566231, published as US 2018/010007)

Boein (US 2015/0239991) discloses  the purification of a hetero-dimeric antibody such as scFv using Protein A chromatography such that one of the H chains is modified to reduct or eliminate binding to the Protein A and the unmodified H chain retain binding to Protein A. 

Hagai (WO 02/059264) discloses the pruificaiton of an scFv blonging to the VH3 family by Protein A where the bound scFvs were recovered form the column by acid elution (0.1 M glycine, pH 3.0)

Haywood (US 15/566,231, published as US 2018/0100007) discloses purification of a human VH3 domain containing antibody such as an Fab, Fab’, F(ab’)2, Fv and scFv as well as such types of antibodies that include more than one VH3 domain such as those in the format of diabodies, tetrabodies, minibodies, and domain antibodies. using Protein A chromatogrpahy where the VH3 domain containing antibody is recovered in monomeric form. 

–Protein G (see affinity chromatography and “Protein G”)

Ion Exchange:

Cho (US 2006/0153860) teaches purification of antigen dbinding polypeptides which can include the light chain or variable region of a heavy chain or at least one CDR of a light chain or heavy chain or Fab or diabody or llama antibody using anion or cation chromatography, HIC hydroxylapatite chromatography, etc.

–Anion Exchange

–Cation Exchange

Beirnaert (US 2010/0297111) teaches nanobides which includes variable domains present in naturally occurring heavy chain antibodies (referred to as VVHH domains) against TNF alpha using using cation exchange chromatography.

Brown (US 12/608964, published as US 2010/0172894 and US 10,118,962) disclsoes purifying single domain antigen binding molecules by adjusting the cell culture medium containg the antigen-binding fusion polypeptide to a conductivity between about 12 and 9 mS/cm and a pH equal to or less than about 4.5 and contacting the medium with a CEX and selectively eluting the nanobody from the support by washing teh support to remove contaminant(s) and eluting the antigeng binding fusion polypeptide wiht an elution buffer. Further chromatography such as hydroxyapatite can then be performed. In one embodiment the antigen-binding fusion polyeptide is a singel cchain polypeptide that includes at least one immunoglobulin variable domain from an antibody naturally devoid of light chains. .

Jonniaux (US2011/0183861) discloses purification of nanobides via CEX with a wash buffer of 10 mM citric acid, pH 4.0 and elution buffer of 10 mM citric acid/1M NaCL, pH4 followed by size exclusion chromatography. The nanobdy was produced intra cellularly or from an inclusion body in a bacterial cell or produced extracellularly and isolated form the medi7um in which the host cell is cultivated.

Silence (US2006/0115470) discloses purification of VHH on a cation exchange column which had been equilibrated in 25 mM citric acid pH 4.0 and then eluted with 1M NaCL.

Mixed-Mode:

Gagnon, “Minibodies and Multimodal chromatography methods” BioProcess Interational Feb 2010) discloses that Capto MMC media provides adequate capture and significant dimer removal for  purifying a stem-cell antigen (PSCA) minibody. 

Structural Variants of Domain Antibodies:

Lack of Disulfide Bonds:

Schotte (US13/266503 and US2012/0157664) discloses that despite the high yield and functionality of domain antibodies produced in non-coli hosts, in particular yeast, there is a quantitatively significant fraction of product that represents a structural variant. In particular, a fraction of the product lacks at least one disulfide bond. It is consistently reported that conventional antibodies or fragments lacking at least one disulfide bond are characterized by a loss of function. Schotte provides a method of counteracting this. Schotte teaches a method of producing a domain antibody in yeast by applying conditions such as addition of oxidizing agents, preferably oxidizing metal ions such as Cu2+, Fe2+, Fe3+ and Zn2+ that promote the formation of disulfide bridges in the domain antibodies and/or removing domain antibodies lacking at least one disulfide bride. The method results in the production of domain antibodies wherein the quality of the domain antibodies is improved with a reduced level of free thiol or its absence.