Immunoglobulin-M purification — Challenges and perspectives

Biotechnology Advances 29 (2011) 840–849

Contents lists available at ScienceDirect

Biotechnology Advances j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b i o t e c h a d v

Research review paper

Immunoglobulin-M purification — Challenges and perspectives Satyen Gautam, Kai-Chee Loh ⁎ Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, S117576, Singapore

a r t i c l e

i n f o

Article history: Received 6 October 2010 Received in revised form 28 June 2011 Accepted 29 June 2011 Available online 7 July 2011 Keywords: Immunoglobulin-M Purification Affinity chromatography Biomimetic ligands

a b s t r a c t Extensive research in the past two decades has led to the realization of Immunoglobulin-M (IgM) as a potential therapeutic and diagnostic agent. In order to fully exploit the potential of IgM, large quantities, in a highly pure and active form, must be available at low cost for performing clinical trials, characterization studies and quantitative-structure activity analyses. The complex physico–chemical properties, in particular its large size and labile nature renders downstream purification of IgM difficult. This review discusses the limitations and challenges associated with the current IgM purification strategies and proposes future directions for research. The uniqueness of affinity chromatography, specifically biomimetic affinity chromatography for protein purification is highlighted and its potential for IgM purification is discussed. © 2011 Elsevier Inc. All rights reserved.

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . . . . . . . . Structure of Immunoglobulin-M . . . . . . . . . . . . IgM applications . . . . . . . . . . . . . . . . . . . . 3.1. Disease diagnosis . . . . . . . . . . . . . . . . 3.2. Therapeutic agent . . . . . . . . . . . . . . . . 3.3. Stem cell research . . . . . . . . . . . . . . . . 4. IgM purification . . . . . . . . . . . . . . . . . . . . 4.1. Precipitation . . . . . . . . . . . . . . . . . . 4.2. Chromatographic techniques . . . . . . . . . . . 4.2.1. Non-affinity chromatographic methods . 4.2.2. Affinity-based separation . . . . . . . . 4.2.3. Immunoaffinity chromatography . . . . 5. Commercially available IgM purification columns/matrices 6. Perspectives . . . . . . . . . . . . . . . . . . . . . . 7. Concluding remarks . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

1. Introduction The modern age of immunology began in 1890 when Emil von Behring, “The Founder of Serum Therapy” and Shibasaburo Kitasato described antibody activity against diphtheria and tetanus toxins. Rats, guinea pigs and rabbits were immunized with low doses of the infectious agents causing diphtheria and alternatively, tetanus. ⁎ Corresponding author at: Department of Chemical & Biomolecular Engineering, 4 Engineering Drive 4, S117576, Singapore. Tel.: + 65 6516 2174; fax: + 65 6779 1936. E-mail address: [email protected] (K.-C. Loh). 0734-9750/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.biotechadv.2011.07.001

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

840 841 841 841 841 842 842 842 842 842 844 845 845 845 848 848 848

Administration of the sera produced by these animals into nonimmunized animals infected with the bacteria cured the ill animals. This landmark discovery led to the first successful treatment of a child suffering from diphtheria in 1891 (Kornelia and von Behring, 2001). The use of immunoglobulin (Ig) products dates back to 1952 when plasma-derived, gamma globulin-rich fraction (referred to as ISG) was injected intramuscularly to treat immune deficiency (Schienfeld and Godwin, 2008). At that time, ISG [prepared in accordance with procedures developed by Cohn et al. (1944)] were, however, poorly tolerated when administered intravenously. This was probably due to IgG aggregation and activation of the complement system. ISG had to

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849 Table 1 Characteristics of human immunoglobulins.

Table 2 Physical properties of human IgM.

Isotype

Subclasses

M.W. (kDa)

Occurrence

IgG

IgG1 IgG2 IgG3 IgG4 IgA1 IgA2

150

Monomer

160 – 180 190 950 1150

Monomer Polymer Monomer Monomer Pentamer Hexamer

IgA IgD IgE IgM

841

be injected intramuscularly in low doses. In a single injection, the quantity of ISG administered rarely exceeded 150 mg/kg body weight (Pennington, 1990). A major breakthrough came when Barandun et al. (1962) advocated the need to modify immunoglobulin preparations to prevent aggregation and thus reduce their serious postinfusion adverse effects. It was realized that harsh chemical or enzymatic treatment could actually denature the molecule and reduce or abolish their natural functional ability. Since then advances have been made in purifying and isolating immunoglobulins and today Igs find vast applications as diagnostic and therapeutic agents. The monoclonal antibody (mAb) sector dominates the small molecules market and is the highest growth segment within the entire biopharmaceutical industry. Over the years, the global mAb market for therapeutic applications has witnessed a near five fold growth from $10.3 billion in 2004 to $48 billion in 2010. In terms of total sales revenue, the top ten mAb based therapeutics for the year 2010 were: Infliximab, Bevacizumab, Rituximab, Adalimumab, Trastuzumab, Cetuximab, Ranibizumab, Natalizumab, Omalizumab and Palivizumab. As of 2010, there were at least 6 new mAbs under regulatory review, 30 mAbs in phase III and over 100 in Phase II clinical trials1. Although significant growth is evident in the biopharmaceutical industry, the truth is that unless the industry succeeds in bringing down the cost of mAb production, particularly the downstream purification, mAbs will never reach their full commercial or medical potential. Drucker et al. (2008) estimated the cost of treating a patient with adjuvant and metastatic breast cancer settings using monoclonal antibody Trastuzumab as USD $45,881 and $26,059, respectively. According to Centocor (now Janssen Biotech Incorporation), Infliximab can cost USD $19,000 to $22,000 a year per patient. Simply put, mAbs are extremely expensive to manufacture and thus costly to patients and health care systems. 2. Structure of Immunoglobulin-M Turner (1981) defined immunoglobulins as a family of structurally dynamic glycoproteins responsible for foreign body recognition and isotope elimination. Based on differences in the constant region of the heavy chains, Ig molecules can be classified into five different classes: IgG, IgM, IgA, IgD and IgE. The Ig monomer is a "Y"-shaped molecule that consists of four polypeptide chains; two identical heavy chains and two identical light chains connected by disulfide bonds. Table 1 summarizes the characteristics of human immunoglobulins (Hamilton, 1997). Perkins et al. (1991) investigated the pentameric structure of human and mouse IgM by synchrotron X-ray solution scattering and molecular graphics modeling. Studies performed on IgM and the four fragments of IgM: the IgM-S monomer, the Fc5 central disk, the Fab′2 arm, and the Fab fragment suggested the IgM structure to be essentially planar with tip-to-tip distance between diametrically 1 Maggon Krishan, Guild (KPG). Top ten monoclonal antibodies 2010: global market analysis & blockbuster mabs [Internet]. VersionKnol, 49. Knol Publishing; 2011 http:// knol.google.com/k/krishan-maggon/top-ten-monoclonal-antibodies-2010/ 3fy5eowy8suq3/143.. [Jun 14. Available from]..

No. of residues Sedimentation coefficient (S20,w) Carbohydrate (% by weight) Biological survival (plasma half-life) Diffusion coefficient (cm2/sec) Molar extinction coefficient (polyclonal)

7926 17.7 10 5–10 days 2.6 × 10−7 1.18

opposite Fab arms to be approximately 36 nm. IgM dominantly exists as a pentamer, consisting of five identical subunits of about 180 kDa and a J or joining chain, with a molecular weight of about 15 kDa (Hamilton, 1997). Table 2 provides an indication of the physical properties of the human IgM molecule (Gagnon et al., 2008; Perkins et al., 1991). Interested readers can refer to Perkins et al. (1991) and Volkov et al. (2003) for detailed descriptions of the IgM structure. 3. IgM applications 3.1. Disease diagnosis Detection of specific IgM antibodies in the sera of patients has been widely used as an early and sensitive diagnostic tool for many infectious diseases. Commonly used serology techniques like ELISA, immunofluorescence and precipitation have undergone significant modifications over time. Accurate analysis of antigen-specific IgM antibodies can now be performed allowing reliable diagnosis of diseases including legionnaires' disease (Zimmerman et al., 1982), Mycoplasma pneumoniae infection (Samra and Gadba, 1993), viral diseases like measles, rubella, mumps and M. parainfluenzae (Bringuier et al., 1978), dengue (Sathish et al., 2003), enteroviral meningitis (Climaker et al., 1992), rickettsial disease (Vene, 1989) and cytomegalovirus disease in AIDS patients (Boibieux et al., 1992). 3.2. Therapeutic agent Transition of mAbs from research reagents to mainstream commercial therapeutics has been a slow process. It took the first mAb 22 years from discovery to approval as a cancer therapeutic. Ongoing efforts to fight life threatening diseases have led to the realization of monoclonal IgM as a potential therapeutic agent. Harte et al. (1983) performed studies on murine systems and showed that small amounts of specific monoclonal IgM could specifically enhance the response to a blood-stage murine malaria vaccine. These studies open up avenues for the probable use of (specific) monoclonal IgM as a potent adjuvant in human malarial vaccination. In a study by Irie and Morton (1986), the researchers demonstrated the anti-tumor effects of human monoclonal IgM antibody, L72, in patients with cutaneous melanoma. Regression was seen in tumors of six out of the eight patients. Natural IgMs found in the serum of healthy humans have been highly effective in arresting the growth of established human neuroblastomas (NB) engrafted into rats (David et al., 1996). The antiNB activity of injected cytotoxic IgM was reported to persist for several weeks after infusion. Such a finding is important particularly in the United States where neuroblastoma accounts for approximately 7.8% of childhood cancers, which is approximately 9.5 cases per million children (Lacayo and Marina, 2007). Ongoing efforts saw PAM-1, a fully human IgM antibody, being identified as an ideal tool for the diagnosis and treatment of the precancerous and cancerous epithelial lesions (Brandlein et al., 2003). In July 2010, Patrys Limited (Australia) acquired the rights to PAT-PA1 (formerly PAM-1) from Debiovision Incorporation. PAT-PA1 is currently under preclinical trials 2. Work by Irie et al. (2004) led to the successful development of a 2

Available from: http://www.patrys.com/.

842

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

100% human monoclonal IgM protein (L612 HuMAb), infusion of which had produced significant antitumor activity in patients with metastatic melanoma. [L612 consists of hexameric IgM (about 20%), pentameric IgM (about 74%) and other minor IgM molecules]. The authors documented that out of the nine patients involved in the study, two patients, after mAb infusion followed by surgical therapy, were without any signs of disease even after 5 years of treatment. mAb 216, a naturally occurring human IgM, can efficiently bind and kill acute lymphoblastic leukemia (ALL) B-progenitor lymphoblasts (Bieber et al., 2007). mAb 216 has been shown to exhibit an enhanced degree of cytotoxicity when introduced in combination with vincristine, a chemotherapeutic agent. mAb 216, alone and in combination with vincentrine is currently under Phase I clinical trials. Studies performed by Vollmers et al. (1998) established SC-1, an IgM of human origin, as a promising neoadjuvant therapeutic agent for gastric cancer. In October 2009, Patrys Limited acquired the rights to PAT-SC1 (formerly SC1) from Debiovision Incorporation. In November 2010, the company received confirmation of Orphan Medicinal Product Designation for PAT-SC1 from the US Food and Drug Administration. In addition to PAT-PA1 and PAT-SC1, Patrys Limited has been involved in developing PAT-SM6 (SAM-6), an IgM of human origin, that binds to multiple types of cancer including melanoma, breast, colon and pancreatic. In addition to being effective against cancer, PAT-SM6 has also been shown to reduce bad cholesterol in three animal models, presenting potential cardiovascular applications3. While most of the mAb based therapeutics are IgGs, the world's first IgM-based therapeutic product is now available from Biotest, sold under the commercial name Pentaglobin. Pentaglobin is an IgMenriched intravenous immunoglobulin (IVIGMA) preparation. Owing to its IgM content, Pentaglobin has been found to be unique in the elimination of infectious pathogens and neutralization of their endoand exo- toxins, far superior to normal IgG preparations (IVIG) (Alejandria et al., 2002). Although all IgM-based therapeutics (except for Pentaglobin) are either in laboratory or clinical trials, the above studies clearly approve the role of monoclonal IgMs as futuristic therapeutic agents. 3.3. Stem cell research Stem cells have the ability for both self-renewal and to give rise to differentiated progeny. In the past few years, significant efforts have been put by the scientific community to exploit stem cells for curing and treating diseases as it is hoped that stem cells can repair, improve, and/or replace damaged organs (Brignier and Gewirtz, 2010). Many different types of stem cells are now known to exist but in small populations. Identification and isolation of these stem cells can be achieved by developing monoclonal antibodies that can be directed against their cell surface receptors (stem cell markers). Literature suggests the use of several monoclonal IgM antibodies for the identification/characterization of specific stem cells by directing specific IgMs towards CD15, human STRO-1, EMA-1, SSEA-1, SSEA-3 and TRA-1-60, TRA-1-81 stem cell markers (Jung et al., 2007; Kaneko et al., 2009; Read et al., 2009; Takehara et al., 2008). 4. IgM purification The advent of hybridoma technology by Kohler and Milstein in 1975 led to the first murine antibody of defined specificity. Since then, dramatic improvements have been made in upstream and downstream processing. In the commercial development of biotherapeutics, downstream processing accounts for 50% to 80% of the total manufacturing costs and is a key issue that manufacturers focus on (Roque et al., 2007). Several downstream processing steps including 3

Available from: http://www.patrys.com/.

clarification, anion exchange chromatography (AEC) followed by hydrophobic interaction chromatography (HIC) [or cation exchange chromatography (CEC)], virus filtration and finally ultrafiltration or diafiltration are often involved (Kelley, 2007). The elimination of even one of these steps can significantly reduce operating costs. Cutting downstream processing costs would result in reduced treatment costs allowing greater accessibility and acceptance of the therapy and finally greater revenues. This review will focus on the various IgM purification strategies employed over the past few decades and attempts will be made to assess their strengths and shortcomings in purifying IgM. 4.1. Precipitation When used at appropriate concentrations, polyethylene glycol (PEG), which is non-ionic and water-soluble, can cause proteins to precipitate from solution. Neoh et al. (1986) employed this technique for purifying mouse monoclonal IgMs from ascitic fluid. The methodology was successful in providing high yields (~80%) but the low IgM purity (~90%) reciprocates the need to employ additional purification steps. Gonzalez et al. (1988) took the approach of euglobulin precipitation for purifying murine IgG3 and IgM monoclonal antibodies from ascitic fluid. IgM recovery was less reproducible, varying from 40% to 90%. HPLC and radial immuno-diffusion analyses showed the presence of IgG contaminants in the purified IgM preparations. Along the same lines, Tatum (1993) illustrated the use of a three stage precipitation approach for purifying IgM from human plasma obtained by therapeutic plasmapheresis. The first stage initiated the removal of lipoproteins and fibrinogen by subjecting the plasma to a saturated ammonium sulfate (SAS) solution (final concentration 30%), followed by immunoglobulin recovery by precipitation with SAS (final concentration 50%). Subsequent PEG precipitation of the recovered immunoglobulin fraction led to the selective removal of IgM from IgG. The method was effective in handling large volumes of the starting material with yields ranging from 65% to 80%. An IgM purity of 95% with contaminations from small quantities of IgG, IgA and albumin was documented. Precipitation techniques, though simplistic in approach, are slow, non-reproducible, non-specific in nature and prone to aggregational losses. Any partnership with chromatographic methods to attain higher purities often augments the total cost and time of purification. 4.2. Chromatographic techniques 4.2.1. Non-affinity chromatographic methods Table 3 summarizes research encompassing non-affinity type chromatographic methods for IgM purification. Most of the literature on IgM separation suggests extensive use of chromatographic columns packed with porous beaded resins. These conventional chromatographic supports with pore diameters ranging from 300– 500 Å rely on intraparticle diffusion for mass transport. Diffusion is slow and increasingly so for a large molecule like IgM. This is the reason why both capacity and resolution decline with increasing flow rate, necessitating the use of low superficial velocities. However, low velocities result in prolonged exposure of IgM molecules to the matrix material rendering them susceptible to denaturation. McCarthy et al. (1996) addressed this problem through the use of perfusion ionexchange chromatography (IEC). Large “throughpores” (pore diameter 6000–8000 Å) allowed IgM molecules to “perfuse” rapidly through the interior as well as around the particles. In addition, short “diffusive pores” (mean diameter 500–1500 Å) (Roche Applied Science, 1996) allowed rapid diffusion to and from the internal binding sites. The whole purification process could be completed in 10 min with no loss in resolution and binding capacities. Recently, Brne et al. (2007) and Gagnon et al. (2008) advocated the use of monolithic columns for applications in chromatography. Monoliths are chromatographic matrix cast as a single piece into a

Table 3 IgM purification by non-affinity chromatographic techniques. Purification approach

Column/matrix

% Purity (final)

% yield of biologically active IgM

Reference

Gel filtration + AECa AEC

Ultrogel AcA 34 DEAE Sepharose CL6B Mono Q HR 5/5

N98

70

Jehanli and Hough (1981)

~ 0.01 g IgG/l of purified IgM fraction + traces of IgA + proteins other than Igs Highly pure (only IgM was detected in the purified fraction by SDS-PAGE) ~ 95

~ 50 (based on the data of 13 sera samples)

Sampson et al. (1984)

80–90

Stanker et al. (1985)

30–60

Johnson et al. (1986/1987)

NA*

Loss of biological activity but no quantitative data

Belew et al. (1987)

NA N95

NA 10

Harshman (1989) Knutson et al. (1991)

Pure but no quantitative data available

Bouvet and Pires (1991)

Pure but no quantitative data available 60-70

Activity retained but quantitative data not available 1.1 mg IgM/250 microlitre of ascitic fluid NA

Josic et al. (1991) Knudsen et al. (1992)

N95

34

Roggenbuck et al. (1994)

N95

Activity retained but quantitative data not available 40–65

McCarthy et al. (1996) Tornoe et al. (1997)

~ 99

~ 56

Gagnon et al. (2008)

~ 90%

Low but no quantitative data available

Mahassni et al. (2009)

Human serum

b

Ascitic fluid

HAC

Ascitic fluid

Dye ligand affinity chromatography + gel filtration TACc

Culture supernatant

Ascitic fluid Ascitic fluid

Ascitic fluid

HAC Ammonium sulfate precipitation + Gel filtration + AEC Modified gel filtration

Ascitic fluid Cell culture supernatant

HAC TAC

Cell culture supernatant

HICd + gel filtration CECe + AEC HAC+ HIC+ CEC+ AEC CEC+ AEC+ HAC CEC+ AEC+ CEC+ gel filtration

Ascitic fluid Cell culture supernatant

Cell culture supernatant

Ascitic fluid

HTP grade hydroxylapatite

Affigel blue ACA-22 T-Gel (mercaptoethaol coupled to divinylsulfone activated-Sepharose 6B) Bio-Gel HT hydroxylapatite Ultrogel AcA 22 DEAE-Sepharose CL 6B S200 Sephacryl HA-HPLC 2-hydroxypyridine coupled to divinyl-sulfone-activated agarose Phenyl–Superose HR 10/10 Superose 12 POROS HS/M PEEK POROS HQ/M PEEK HAP, Bio Gel Phenyl-Sepharose HiLoad Sulfonyl-Sepharose Fast Flow Q-Sepharose Fast Flow CIM SO3 CIM QA Ceramic hydroxylapatite, TypeII CM-Biogel DEAE-Biogel CM-Biogel Sephadex G-100

Substantially pure but no quantitative data available

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

Source Plasma

*NA: not available. a Anion Exchange Chromatography. b Hydroxylapatite Chromatography. c Thiophilic Adsorption Chromatography. d Hydrophobic Interaction Chromatography. e Cation Exchange Chromatography.

843

844

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

column in the form of disks, rods or tubes. The highly interconnected flow-through pores ensure that mass transfer is governed by convection rather than by diffusion. Flow-through pores with diameters ranging from 2–5 μm can accommodate even the largest of biomolecules ensuring high binding capacities. An open pore structure in the monoliths thus facilitates separations at high flow rates and low pressure drops without sacrificing capacity and separation performance (BIA Separations, 2008). Use of monolithic IEC for IgM separation has been demonstrated by Brne et al. (2007). Three different column chemistries: quaternary amine (QA), diethylaminoethanol (DEAE) and ethylenediamine (EDA) were investigated for their efficacy to separate IgM, IgG and human serum albumin. Chromatographic techniques like IEC, HIC, hydroxylapatite chromatography (HAC), thiophilic adsorption chromatography (TAC) and dye ligand chromatography exploit the differences in charge, hydrophobicity and interactional ability with hydroxylapatite, thiophilic adsorbent and dye, respectively for protein purification. These techniques are nonspecific in their binding and thus vulnerable to cross-contaminations. As evident from Table 3, a single stage purification approach has been inadequate in providing highly pure IgMs. Multi-step procedures have been adopted, but these increase the total cost and time for downstream processing. HAC, a mixed mode ion exchange technique, has been quite popular among researchers (Gagnon et al., 2008; Harshman, 1989; Josic et al., 1991; Stanker et al., 1985; Tornoe et al., 1997). Use of mild elution conditions during HAC preserves the activity of labile molecules like IgM. However, use of low ionic strength solutions for sample preparation and column equilibration can cause IgM precipitation and aggregation. Josic et al. (1991) showed almost complete precipitation of IgMs when applied to hydroxylapatite column at low ionic concentrations (30 mM sodium phosphate, pH 6.8). To overcome these problems, addition of 0.1 M sodium chloride to the buffers was suggested. Hydroxylapatite suffers from high risks of microbial contamination leading to non-reproducible separations and short column life (Josic et al., 1991). Column regeneration problems have been reported and use of fresh hydroxylapatite for every batch has been recommended (Tornoe et al., 1997). Rigorous cleaning and appropriate storage of the column is critical to minimize loss of separation performance, reproducibility, capacity and short column lifetimes (Josic and Lim, 2001). Immobilized pseudo-ligand Cibacron Blue F3GA has been investigated by Johnson et al. (1986/1987) for isolation of IgM from murine ascitic fluid. Though considerable recoveries were achievable, the purified fraction was found to be contaminated with haptoglobin, transferrin and albumin. The non-specific nature of the dye-ligand thus necessitated the use of an additional stage of gel filtration. Acetate buffer having a pH 5 was used for the dialysis of ascitic fluid, for column wash and protein elution. Extensive exposure to such low pH values can have a deteriorating impact on the IgM activity. Furthermore, extremely long purification time of more than 3 days was required to process 1.8 mL of ascitic fluid. Immobilized pseudo-ligand ‘T Gel’, a thiophilic adsorbent, has not been very promising either. The method suffers from low recoveries associated with activity losses of IgM on adsorption to the chromatographic adsorbent (Belew et al., 1987). Modified sulfone-aromatic ligands as thiophilic adsorbents were introduced by Knudsen et al. (1992). Low purities (~70%) and low adsorption capacities of 1–1.5 mg/mL limit their suitability for large scale IgM isolation. Work by Sampson et al. (1984) deserves some mention. The researchers evaluated an anion exchange chromatographic system for the separation of IgM from human serum. Human serum diluted in 0.16 M phosphate buffer, pH ~ 6.5 was loaded on a Mono-Q anion exchange column. IgM molecules got bound to the column, while IgA and IgG passed out as flow through. The fact that IgA did not bind to the column and passed as flow through is rather intriguing. IgA is probably the most acidic among serum immunoglobulins with isoelectric points (pI) ranging between 4 and 7.1. The pI range of

IgM extends between 4 and 9.1 (Prin et al., 1995). IgG has been reported to be a basic isotype with the pI ranging from 7.2 to 8.6 for the four subclasses (Hamilton, 1998). These data suggest that serum loading at pH 6.5 should have actually led to the binding of IgA and IgM to the anion exchange column, with IgG passing out as flow through. Owing to the large size of IgMs in comparison to other impurities, size exclusion chromatography (SEC) has been used effectively to purify IgMs. However, low loading capacity ranging from 2 to 5% of the total column volume limits its use for large scale application. Increasing operational velocities or sample loading can have detrimental impact on peak resolution. It is a well known fact that SEC exploits the difference in the hydrodynamic volume of proteins for separation efficiency. No interactions occur between the proteins and the column media. However, an unusual gel filtration technique in which IgMs were actually made to interact with the media was demonstrated by Bouvet and Pires (1991). The separation was based on introducing ascitic fluid to an SEC column equilibrated with a low ionic strength buffer (0.005 M phosphate buffer) followed by a rinse with a high ionic strength buffer (0.05 M phosphate, 2 M NaCl). The protocol prolonged the stay of IgMs in the column, preventing its elution before albumin, α2 macroglobulin and other serum proteins. The method was successful in providing pure IgM with yields ranging from 50 to 80% for the sixteen euglobulin monoclonal IgMs tested. 4.2.2. Affinity-based separation Affinity chromatography, a technique based on molecular recognition, involves specific, reversible and non-covalent interactions for protein purification. The specific nature of binding facilitates increased operational yield and elimination of undesirable contaminants allowing several fold purification in a single step. In addition, the method allows for the separation of active biomolecules from denatured or functionally different forms (Roque et al., 2007). One of the earliest approaches for affinity purification of IgM was based on the adsorption of IgM to protamine-Sepharose matrix (Wichman and Borg, 1977). The researchers proposed that the binding was likely electrostatic in nature involving the binding of several Fc parts of IgM to suitable portions in protamine. IgA, ceruloplasmin, α2-macroglobulin among some other serum proteins, were present as contaminants in the protamineSepharose eluted fraction. Two additional stages of gel filtration chromatography had to be employed to attain 98% pure IgM with a 30% yield. The biggest drawback of this technique is probably protamine's incompetency to specifically bind IgM molecules. Human secretory component (SC), a 75 kDa protein isolated from human milk whey, has been evaluated as an affinity adsorbent for IgM purification (Jones et al., 1987). Cell culture supernatants of murine and rat hydridomas were purified by the affinity matrix. SC specifically binds polymeric immunoglobulin from different species but does not bind to monomeric or aggregated monomeric immunoglobulins, making the process species independent and free from any contamination with IgGs. However, the ability of the SC to bind polymeric immunoglobulins limits its application to purify IgMs from sera as the final product could be contaminated with IgA. Secondly, generation of pure SC was a cumbersome process in itself, involving i) affinity chromatography using human pentameric IgM-Sepharose matrix ii) gel filtration chromatography and iii) Immuno-affinity chromatography using Sepharose anti-human IgA adsorbent. Shibuya et al. (1988) advocated the use of a mannose-specific snowdrop (Galanthus nivalis) bulb lectin-GNA as an affinity ligand for IgM purification. GNA coupled to Sepharose 4B was employed for purification of IgM from two sources: murine hybridoma serum and human serum. The ligand showed high specificity to bind murine IgM with no affinity for IgG molecules. The elution of bound IgMs was achieved using methyl α-D-mannoside which is costly and impractical for large scale application. In addition, the ligand failed to bind human IgMs when human serum was subjected to affinity

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

chromatography. Nevens et al. (1992) described a method for affinity purification of monoclonal IgM utilizing animal-derived mannan binding protein (MBP). MBP, a 650 kDa protein, was isolated from rabbit serum. Purification of IgM was temperature and calcium dependent. Binding was performed at 4 °C in a buffer that contained calcium chloride. Elution was achieved at room temperature in a buffer that contained ethylenediaminetetraacetic acid (EDTA) in the absence of calcium ions. A binding capacity of nearly 1 mg IgM/mL of support has been reported. The affinity matrix was shown to possess the ability to isolate IgMs from different species, including mouse, human and bovine, but with varying efficiencies. IgMs isolated from ascitic fluid were purer than those obtained from serum purification. The method, though highly effective in preserving the biological activity of IgMs, suffered from very low binding capacity. While Protein A of Staphylococcus aureus, Protein G of group C and G Streptococci and Protein A/G (a recombinant fusion protein combining IgG binding domains of both Protein A and Protein G) do not bind IgM well, Protein L, a cell wall molecule of certain strains of the anaerobic bacterial species Peptostreptococcus magnus, exhibits affinity for IgM. In addition, Protein L binds IgG and IgA well. The equilibrium constants for the binding reactions between Protein L and human IgM, human IgA and human IgG are almost same (~ 10 10 M −1) suggesting a similar binding to the three immunoglobulin classes (Akerstrom and Bjorck, 1989). Ability to bind to a wide range of immunoglobulin isotypes limits the application of Protein L as an affinity ligand for IgM purification. It may, however, be used to isolate IgMs from crude samples devoid of IgG and IgA. In addition, strong forces of attraction between IgM and Protein L (Ka = 10 10 M −1) could result in elution difficulties. Use of affinity chromatography for purifying an antigen-specific IgM has also been seen in the works of Hibma and Griffin (1990). The technique was highly specific in its approach and could not be used to purify all IgMs. Complement protein C1q for IgM purification was investigated by Nethery et al. (1990). The purification strategy was based on an 18-fold higher affinity of C1q to bind IgM over monomeric IgG. The affinity adsorbent strongly bound IgM at 5 °C while pure IgM was eluted isocratically, after 2 h of column incubation at room temperature, with the column buffer. Isolation of IgM was based on a temperature-dependent interaction with no involvement of harsh elution conditions. This served to preserve the immuno-activity of the IgM molecule. However, the method suffered from low ligand immobilization efficiency of 35%, low binding capacity of 0.4 mg of IgM/mL of gel and product contamination with IgG. Most of the affinity ligands described thus far showed specific binding to IgM and allowed easy isolation of pure and active IgMs. However, use of these natural antibody-binding ligands suffered from high costs associated with their isolation and purification, low immobilization efficiency, losses in ligand activity on immobilization and poor resistance to sanitizing agents. In the last few years, biomimetic ligands have gained recognition as a viable alternative to natural binding partners. Biomimetic ligands are synthetic molecules that mimic the recognition activity of the native biological macromolecules. These ligands offer considerable benefits over their natural counterparts including low cost and ease of synthesis, improved resistance to chemical and biochemical degradation, the possibility to employ mild elution conditions and lower toxicity and immunogenicity (Roque et al., 2007). Palombo et al. (1998) used a synthetic ligand PAM (Protein A mimetic, TG19318) for affinity purification of IgM from serum, ascitic fluid and cell culture supernatant. TG19318 is a low molecular weight peptide consisting of four identical tripeptide chains linked to a central polylysine core. The ligand was selected through the synthesis and screening of a multimeric peptide library. Palombo and coworkers reported a purity of 85–95% when isolating monoclonal IgMs. The advantages offered by TG19318 were worthwhile although its use as an affinity ligand for IgM purification was probably limited by the fact

845

that the molecule bound not only to IgM but also to IgG, IgA, IgE and IgY. The effectiveness of affinity purification relies on the ability of an affinity adsorbent to recognize specifically the molecule of interest. In the case of TG19318 for purification of IgM from sera, ascitic fluid and culture supernatants, the purified fraction could be contaminated with extraneous antibodies present in the feeds. Another synthetic ligand, 8/7, has been shown to bind IgM. Ligand 8/7 is a Protein L mimetic ligand and contains aromatic and aliphatic moieties with polar substitutents, comprising of 4-aminobenzamide and 4-amino butyric acid. Ligand 8/7 exhibits a binding behavior similar to Protein L in binding to IgM, IgA and IgG (Roque et al., 2005). While ligand 8/7 may be suitable as a general Ig-bjnding ligand, low binding capacity of 0.55 mg IgM/gm of resin limits its application for large scale purification of IgM alone. 4.2.3. Immunoaffinity chromatography As early as 1973, one of the articles on IgM purification described the application of antisera (containing anti-γ, anti-α, anti-α2macroglobulin and anti-complement) as immunoadsorbent to obtain pure IgM from Cohn fraction III of normal human plasma (Hoven et al., 1973). Immunoaffinity approaches based on anti-IgMs as ligands have been described by Cripps et al. (1983) and Huschka et al. (1982). Use of antisera as immunoadsorbent posed practical limitations on the use of this technique for large scale purification. Generation of large amounts of purified immuno-adsorbent required large quantities of highly pure antibodies, antithesis to the very task at hand to purify IgMs. In addition, high affinities associated with immunoaffinity chromatography could result in incomplete column regeneration and thus reduced capacity. Use of harsh elution conditions could denature the biomolecules, adversely affecting their biological activity. Cripps et al. (1983) and Huschka et al. (1982) used 3 M sodium thiocyante (NaSCN) for eluting the bound IgMs from the immuno-adsorbent. This was contrary to the results reported by Lim (1987) who showed complete loss of IgM activity on exposure to 3 M KSCN. 5. Commercially available IgM purification columns/matrices Over the last few years, the commercial market has seen an increase in the number of players offering solutions to IgM purification. Some of these product details are tabulated in Table 4. 6. Perspectives mAbs produced from murine hybridomas using conventional fusion technology face major obstacles in application to immunotherapy. Repeated dosing of humans with murine monoclonals can lead to a human anti-murine antibody response (HAMA). (Smith et al., 2004). Chimeric antibodies had emerged as the first viable alternative to murine antibodies for therapeutic application. The modular nature of antibodies enabled the linkage of a murine antibody variable region gene segment to a human constant region gene segment to form ‘half human/half mouse’ antibodies referred to as chimeric antibodies. A further reduction in immunogenicity associated with chimeric antibodies was achieved by the process of humanization i.e. grafting of the complementarity-determining regions of a murine antibody onto a human variable-region framework without loss of specificity. These engineered antibodies permit multiple dosing and increased half-life of each dose. On an industrial scale, expression of these antibodies has been performed in three principal mammalian cell lines: Chinese hamster ovary (CHO) cells; and the murine myeloma lines SP2/0 and NS0, in a protein-free/ serum-free chemically defined medium (Isaacs, 2009). The next significant advancement in the field of therapeutic mAbs has been the emergence of “fully human” or human monoclonal antibodies generated through the use of either phage display or transgenic mouse platforms. As human mAbs are better tolerated and are less

846

Table 4 Commercially available IgM purification columns/matrices. Supplier

Trade name

Availability

Matrix

Average particle Binding size (μm) capacity

Affiland

Human IgM Purification Kit

Prepacked 5 mL column

Ligands coupled to Sepharose Fast Flow. No further details available

–

Mouse IgM Purification Kit

10 mg rabbit IgM/mL

Rat IgM Purification Kit

12 mg rat IgM/mL 2 mL, 5 mL, 10 mL, 25 mL, 100 mL gel

TG19318 immobilized on 4% beaded agarose

45–165

GE

HiTrap IgM Purification HP Column

Prepacked 1 mL column

2-mercaptopyridine coupled to 6% beaded agarose

Millipore

Prosep-Thiosorb-M

10 mL, 50 mL, 100 mL, 500 mL, 1 L, 5 L, 10 L gel

BAC

Capture Select Human IgM

5 mL, 10 mL, 50 mL gel Also available in a high throughput format or as prepacked columns

– Ligand of non-biological origin bound to porous glass particles. No further details available 14 kDa Llama antibody fragment 90 coupled to Sepharose 4 Fast Flow

Pierce

ImmunoPure IgM Purification Prepacked 5 mL column Kit Ultralink Immobilized Mannan 5 mL settled gel Binding Protein (MBP)

MBP immobilized on 4% beaded agarose MBP immobilized on ultra link support

34

– 50–80

10 mg IgM/mL

Affinity

5 mg human IgM/mL

Thiophilic

2 mg IgM/mL

Thiophilic

2.5 mg IgM/ mL

Affinity

0.3 mg mouse Affinity IgM/mL 0.75 mg mouse IgM/ mL

% Cross reactivity Purity

Isolation of IgM from human cancer serum Isolation of IgM from mouse cell culture supernatant and/or ascitic fluid Isolation of IgM from rabbit cell culture supernatant and/or ascitic fluid Isolation of IgM from rat cell culture supernatant and/or ascitic fluid - Purification of polyclonal IgM from sera of different species including human, rabbit, mouse, rat, cow and goat - Purification of monoclonal IgM from ascitic fluid (%Yield N 85) - Purification of mouse monoclonal IgM from cell culture supernatant (%yield N 90) Purification of mouse monoclonal IgM from hybridoma cell culture. Purification of human IgM is also possible Purification of monoclonal mouse IgM (% Yield N 90)

95

–

90

90

95

–

Binds all Ig isotypes

– N 85

–

Binds all Ig isotypes

95

Binds all Ig isotypes

Purification of IgMs from plasma, serum and cell culture supernatants. Binds human, rat and mouse IgM only.

–

Most effective for purifying mouse IgM from ascitic fluid

~ 90

Does not bind human and mouse IgG and IgA –

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

20 mg human Affinity IgM/mL 10 mg mouse IgM/mL

Rabbit IgM Purification Kit

Tecnogen KAPTIV-M

Mode of Application purification

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

847

Fig. 1. Schematic representation of the structure of pIgR (adapted from Kaetzel, 2001).

likely to cause immunological side effects, the focus of the industry seems to be gradually shifting away from chimeric and humanized mAbs to fully human antibodies (Lonberg, 2008). While continuous progress is being made in upstream processing, downstream processing has had its own share of developments. The emergence of biomimetic affinity ligands and monolithic chromatographic media in recent years are of remarkable industrial importance as they have opened up new avenues for large scale purification of therapeutic proteins. Monolithic stationary phases have revolutionized protein chromatography because they combine speed, capacity, and resolution in a unique manner. BIA Separation offers monolithic columns based on a highly cross-linked porous polymer with amine/epoxy surface functionalities making ligand immobilization permissible. Two common approaches for the design of biomimetic ligands include de novo/rational approach pioneered by Lowe and co-workers and combinatorial ligand synthesis (Roque and Lowe, 2007). An approach is described below for the design of a biomimetic peptidic ligand for IgM purification. The interaction between a naturally occurring protein and IgM was used as the template for the design of the peptidic ligand. In this approach, it is endeavored that a ‘universal’ affinity ligand, one that can be used to purify all IgM molecules can be designed. In this case, the naturally occurring protein will need to be an Fc receptor to IgM and not a Fab receptor. Polymeric immunoglobulin receptor (pIgR) binds specifically to dimers and larger polymers of IgA (collectively called pIgA) and pentameric IgM (Brandtzaeg, 1985). Fig. 1 shows a schematic

representation of the structure of the polymeric immunoglobulin receptor (Kaetzel, 2001). pIgR is a 100–120 kDa type 1 transmembrane protein with five Ig-like extracellular domains and a long cytoplasmic tail containing signals for intracellular sorting and endocytosis. Once formed, the pIgR-pIg complex undergoes endocytosis and is transported to the apical face of the cell. At this site, the extracellular ligand binding segment of pIgR gets proteolytically cleaved. The released portion of the receptor, referred to as secretory component (SC), remains associated with the polymeric Ig to form secretory IgA or IgM. (Musil and Baenziger, 1987). Interaction studies between free SC and J chain-containing IgA and IgM suggest that the binding constant (Ka) falls in the range of 10 7–10 8 M −1, with SC exhibiting a 5–30 times higher affinity for IgM than for IgA (Brandtzaeg, 1977). This suggests that the non-covalent forces involved in the complex formation are much stronger for IgM than for IgA. Although human pIgR interacts with pIgA and pentameric IgM with high affinity, rabbit pIgR interacts primarily with pIgA with negligible binding capacity for pentameric IgM (Underdown et al., 1992). This difference in the interaction of IgM with rabbit and human pIgR has been exploited by Roe et al. (1999) to characterize the sites of interaction of pentameric IgM with human pIgR. A series of chimeras were constructed to assess the relative role of the various regions. It was found that replacing domain 1 (D1) of rabbit origin with human D1 established pentameric IgM-binding properties in rabbit pIgR. On the other hand, human chimeric receptor containing rabbit D1

Fig. 2. Comparison of the amino acid sequences of the regions A, B and C of human (H) and rabbit (R) pIgR-D1 (adapted from Roe et al., 1999).

848

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849

exhibited reduced binding affinity for pentameric IgM. These results established the critical role of D1 in initiating non-covalent forces of interaction between IgM and human pIgR. pIgR-D1 consists of three complimentary determining regions (CDR)-like loops. To evaluate the relative role of the regions containing the different CDR-like loops, Roe et al. (1999) divided domain 1 into three regions A, B, and C that contained the CDR-like loop 1, 2 and 3, respectively (Fig. 2). A series of chimeras were constructed to assess the relative role of the various regions. The human regions were replaced alone, or in combination, with the same region(s) from rabbit D1, resulting in six different mutants (Fig. 2). Binding studies between pentameric IgM/pIgA and the various chimeric constructs suggested that the region containing the CDR2-like loop is the most important in the pentameric IgM-binding process and replacing this region in the human receptor with the rabbit counterpart will significantly reduce the pentameric IgM-binding capacity. Through the above-mentioned observations, it is therefore possible to design a peptidic ligand that incorporates the CDR2-like loop (SSEGY) of pIgR-D1 for affinity purification of IgM. The authors of this review have attempted the investigation of a 14mer peptide (pep14), CITLISSEGYVSSK, for its interaction to monoclonal human IgM, human IgG1, human IgE and human IgA1, polyclonal IgG from human serum and BSA. Surface plasmon resonance assays have established that pep14 interacted specifically to IgM with negligible affinity for IgG, IgE, IgA1 and BSA (unpublished results). Further studies are being conducted to assess the suitability of pep14 as an affinity ligand for IgM purification. 7. Concluding remarks Low manufacturing cost will be the key to the successful application of IgM as a diagnostic and therapeutic agent. Monoliths immobilized with biomimetic ligands seem to carry potential for IgM production at improved economics and in a highly pure and active form thereby fulfilling the increasingly demanding quality criteria set forth by the regulatory and approval agencies. Acknowledgment Authors gratefully acknowledge the National University of Singapore for the financial support [R-279-000-223-112]. References Akerstrom B, Bjorck L, Protein L. An immunoglobulin light chain-binding bacterial protein characterization of binding and physicochemical properties. J Biol Chem 1989;264:19740–6. Alejandria MM, Lansang MAD, Dans LF, Mantaring III JBlas. Intravenous immunoglobulin for treating sepsis, severe sepsis and septic shock. Cochrane Database Syst Rev 2002(1):CD001090. [Art. No.:]. Barandun S, Kistler P, Jeunet F. Intravenous administration of human gammaglobulin. Vox Sang 1962;7:157–74. Belew M, Juntti N, Larsson A, Porath J. A one-step purification method for monoclonal antibodies based on salt promoted adsorption chromatography on a ‘thiophilic’ adsorbent. J Immunol Methods 1987;102:173–82. BIA Separations. CIM: convective interaction media for biochromatography; 2008. Bieber MM, Twist CJ, Bhat NM, Teng NNH. Effects of human monoclonal antibody 216 on B-progenitor acute lymphoblastic leukemia in vitro. Pediatr Blood Cancer 2007;48:380–6. Boibieux A, Tardy JC, Simplot A, Peyramond D, Aymard M. Amplified ELISA anti-CMV IgM antibodies for early diagnosis of cytomegalovirus disease in AIDS patients. Int Conf AIDS 1992;8:57. Bouvet JP, Pires R. One-step purification of murine monoclonal antibodies of the IgM class. J Immunol Methods 1991;145:263–6. Brandlein S, Beyer I, Eck M. Cysteine-rich fibroblast growth factor receptor 1, a new marker for precancerous epithelial lesions defined by the human monoclonal antibody PAM-1. Cancer Res 2003;63:2052–61. Brandtzaeg P. Human secretory component— VI, immunoglobulin-binding properties. Immunochemistry 1977;14:179–88. Brandtzaeg P. Role of J chain and secretory component in receptor-mediated glandular and hepatic transport of immunoglobulins in man. Scand J Immunol 1985;22: 111–46.

Brignier AC, Gewirtz AM. Embryonic and adult stem cell therapy. J Allergy Clin Immunol 2010;125:S336–44. Bringuier JP, Andre J, Sohier R. The use of IgM antibody responses in the diagnosis of primary infections to measles, rubella, mumps, and M. Parainfluenzae viruses. Med Microbiol Immunol 1978;64:299–305. Brne P, Podgornik A, Bencina K, Gabor B, Strancar A, Peterka M. Fast and efficient separation of immunoglobulin M from immunoglobulin G using short monolithic columns. J Chromatogr A 2007;1144:120–5. Climaker M, Samuelson A, Magnius L, Ehrnst A, Olcen P, Forsgren M. Early diagnosis of enteroviral meningitis by detection of specific IgM antibodies with a solid-phase reverse immunosorbent test (SPRIST) and μ-capture EIA. J Med Virol 1992;36: 193–201. Cohn EJ, Oncley JL, Strong LE, Hughes WL, Armstrong SH. Chemical, clinical and immunological studies on the products of human plasma fractionation. I. The characterization of protein fractions of human plasma. J Clin Invest 1944;23:417–32. Cripps AW, Neoh SH, Smart IJ. Isolation of human IgA and IgM from normal serum using polyethylene glycol precipitation and affinity chromatography. J Immunol Methods 1983;57:197–204. David K, Ollert MW, Juhl H, Vollmert C, Erttmann R, Vogel CW, et al. Growth arrest of solid human neuroblastoma xenografts in nude rats by natural IgM from healthy humans. Nat Med 1996;2:686–9. Drucker A, Skedgel C, Virik K, Rayson D, Sellon M, Younis T. The cost burden of trastuzumab and bevacizumab therapy for solid tumours in Canada. Curr Oncol 2008;15:136–42. Gagnon P, Hensel F, Richieri R. Purification of IgM monoclonal antibodies. Biopharm International; 2008. p. 26–36. [Supplement Mar]. Gonzalez MG, Bettinger S, Ott S, Olivier P, Kadouche J, Pouletty P. Purification of murine IgG3 and IgM monoclonal antibodies by euglobulin precipitation. J Immunol Methods 1988;111:17–23. Hamilton RG. Human immunoglobulins. In: Leffell MS, Donnenberg AD, Rose NR, editors. Handbook of human immunology. CRC Press; 1997. p. 63–74. Hamilton RG. The human IgG subclasses. Calbiochem; 1998. Harshman JS. Purification of murine IgM monoclonal antibodies by hydroxylapatite chromatography. J Tissue Cult Methods 1989;12:115–7. Harte PG, Cooke A, Playfair JHL. Specific monoclonal IgM is a potent adjuvant in murine malaria vaccination. Nature 1983;302:256–8. Hibma MH, Griffin JFT. The purification and characterisation of cervine IgM and IgG. Vet Immunol Immunopathol 1990;26:343–52. Hoven AVD, Conradie JD, Bubb M, Visser L. The isolation of immunogenically pure IgM from Cohn Fraction III of pooled normal human plasma. Immunochemistry 1973;10:107–14. Huschka U, Maier J, Rauterberg EW, Doerr HW. Rapid separation of Immunoglobulin M by immunoaffinity chromatography for detection. of specific antibodies to rubella and Treponema pallidum. Eur J Clin Microbiol 1982;1:118–21. Irie RF, Morton DL. Regression of cutaneous metastatic melanoma by intralesional injection with human monoclonal antibody to ganglioside GD2. Proc Natl Acad Sci 1986;83:8694–8. Irie RF, Ollila DW, O'DAY S, Morton DL. Phase I pilot clinical trial of human IgM monoclonal antibody to ganglioside GM3 in patients with metastatic melanoma. Cancer Immunol Immunother 2004;53:110–7. Isaacs JD. Antibody engineering to develop new antirheumatic therapies. Arthritis Res Ther 2009;11:225–36. Jehanli A, Hough D. A rapid procedure for the isolation of human IgM myeloma proteins. J Immunol Methods 1981;44:199–204. Johnson E, Miribel L, Arnaud P, Tsang KY. Purification of IgM monoclonal antibody from murine ascitic fluid by a two-step column chromatography procedure. Immunol Lett 1986/1987;14:159–65. Jones CL, Georgiou GM, Fowler KJ, Wajngarten PI, Roberton DM. Purification of polymeric immunoglobulin from cell culture supernatants by affinity chromatography using secretory component. J Immunol Methods 1987;104:237–43. Josic D, Lim YP. Analytical and preparative methods for purification of antibodies. Food Technol Biotechnol 2001;39:215–26. Josic D, Loster K, Kuhl R, Noll F, Reusch J. Purification of monoclonal antibodies by hydroxylapatite HPLC and size exclusion HPLC. Biol Chem Hoppe Seyler 1991;372: 149–56. Jung JG, Lee YM, Park TS, Park SH, Lim JM, Han JY. Identification, culture, and characterization of germline stem cell-like cells in chicken testes. Biol Reprod 2007;76:173–82. Kaetzel CS. Polymeric Ig receptor: defender of the fort or Trojan horse? Curr Biol 2001;11:R35–8. Kaneko R, Akita H, Shimauchi H, Sasano Y. Immunohistochemical localization of the STRO-1 antigen in developing rat teeth by light microscopy and electron microscopy. J Electron Microsc 2009;58:363–73. Kelley B. Very large scale monoclonal antibody purification: the case for conventional unit operations. Biotechnol Prog 2007;23:995–1008. Knudsen KL, Hansen MB, Henriksen LR, Andersen BK, Lihme A. Sulfone-aromatic ligands for thiophilic adsorption chromatography: purification of human and mouse lmmunoglobulins. Anal Biochem 1992;201:170–7. Knutson VP, Buck RA, Moreno RM. Purification of a murine monoclonal antibody of the IgM class. J Immunol Methods 1991;136:151–7. Kornelia G, von Behring Emil. The founder of serum therapy; 2001. Lacayo NJ, Marina N. Neuroblastoma; 2007. Lim PL. Isolation of specific IgM monoclonal antibodies by affinity chromatography using alkaline buffers. Mol Immunol 1987;24:11–5. Lonberg N. Fully human antibodies from transgenic mouse and phage display platforms. Curr Opin Immunol 2008;20:450–9.

S. Gautam, K.-C. Loh / Biotechnology Advances 29 (2011) 840–849 Mahassni SH, Klapper DG, Hiskey RG. Purification of a murine IgM monoclonal antibody. Hybridoma 2009;28:189–97. McCarthy E, Vella G, Mhatre R, Lim YP. Rapid purification and monitoring of immunoglobulin M from ascites by perfusion ion-exchange chromatography. J Chromatogr A 1996;743:163–70. Musil LS, Baenziger JU. Cleavage of membrane secretory component to soluble secretory component occurs on the cell surface of rat hepatocyte monolayers. J Cell Biol 1987;104:1725–33. Neoh SH, Gordon C, Potter A, Zola H. The purification of mouse monoclonal antibodies from ascitic fluid. J Immunol Methods 1986;91:231–5. Nethery A, Raison RL, Smith SBE. Single-step purification of immunoglobulin M on ClqSepharose. J Immunol Methods 1990;126:57–60. Nevens JR, Mallia AK, Wendt MW, Smith PK. Affinity chromatographic purification of immunoglobulin M antibodies utilizing immobilized mannan binding protein. J Chromatogr 1992;597:247–56. Palombo G, Verdoliva A, Fassina G. Affinity purification of immunoglobulin M using a novel synthetic ligand. J Chromatogr B 1998;715:137–45. Pennington JE. Newer uses of intravenous immunoglobulins as anti-infective agents. Antimicrob Agents Chemother 1990;34:1463–6. Perkins SJ, Nealis AS, Sutton BJ, Feinstein A. Solution structure of human and mouse Immunoglobulin M by synchrotron X-ray scattering and molecular graphics modelling: a possible mechanism for complement activation. J Mol Biol 1991;221: 1345–66. Prin C, Bene MC, Gobert B, Montagne P, Faure GC. Isoelectric restriction of human immunoglobulin isotypes. Biochim Biophys Acta 1995;1243:287–90. Read TA, Fogarty MP, Markant SL, McLendon RE, Wei Z, Ellison DW, et al. Identification of CD15 as a marker for tumor-propagating cells in a mouse model of medulloblastoma. Cancer Cell 2009;3:135–47. Roche Applied Science. POROS perfusion chromatography media. Biochemica 1996(No. 1). Roe M, Norderhaug IN, Brandtzaeg P, Johansen Finn-Eirik. Fine specificity of ligandbinding Domain 1 in the polymeric Ig receptor: importance of the CDR2-containing region for IgM interaction. J Immunol 1999;162:6046–52. Roggenbuck D, Marx U, Kiessig ST, Schoenherr G, Jahn S, Porstmann T. Purification and immunochemical characterization of a natural human polyreactive monoclonal IgM antibody. J Immunol Methods 1994;167:207–18. Roque ACA, Lowe CR. Affinity chromatography: history, perspectives, limitations and prospects. In: Zachariou M, Bailon P, editors. Affinity chromatography: methods and protocols. Springer; 2007. p. 1–24. Roque ACA, Taipa MA, Lowe CR. An artificial protein L for the purification of immunoglobulins and Fab fragments by affinity chromatography. J Chromatogr A 2005;1064:157–67. Roque ACA, Silva CSO, Taipa MA. Affinity-based methodologies and ligands for antibody purification: advances and perspectives. J Chromatogr A 2007;1160:44–55.

849

Sampson IA, Hodgen AN, Arthur IH. The separation of Immunoglobulin M from human serum by fast protein liquid chromatography. J Immunol Methods 1984;69:9–15. Samra Z, Gadba R. Diagnosis of Mycoplasma pneumoniae infection by specific IgM antibodies using a new capture-enzyme-immunoassay. Eur J Epidemiol 1993;9:97–9. Sathish N, Vijayakumar TS, Abraham P, Sridharan G. Dengue fever: its laboratory diagnosis, with special emphasis on IgM detection. Dengue Bull 2003;27:116–25. Schienfeld NS, Godwin JE. Intravenous immunoglobulin; 2008. Shibuya N, Berry JE, Goldstein IJ. One-step purification of murine IgM and human α2macroglobulin by affinity chromatography on immobilized snowdrop bulb lectin. Arch Biochem Biophys 1988;267:676–80. Smith KA, Nelson PN, Warren P, Astley SJ, Murray PG, Greenman J. Demistified recombinant anitbodies. J Clin Pathol 2004;57:912–7. Stanker LH, Vanderlaan M, Juarez-Salinas H. One-step purification of mouse monoclonal antibodies from ascites fluid by hydroxylapatite chromatography. J Immunol Methods 1985;76:157–69. Takehara T, Teramura T, Onodera Y, Kakegawa R, Fukunaga N, Takenoshita M, et al. Rhoassociated kinase inhibitor Y-27632 promotes survival of cynomolgus monkey embryonic stem cells. Mol Hum Reprod 2008;14:627–34. Tatum AH. Large scale recovery of biologically active IgM (95% pure) from human plasma obtained by therapeutic plasmapheresis. J Immunol Methods 1993;158: 1–4. Tornoe I, Titlestad IL, Kejling K, Erb K, Ditzel HJ, Jensenius JC. Pilot scale purification of human monoclonal IgM (COU-1) for clinical trials. J Immunol Methods 1997;205: 11–7. Turner MW. Structure and function of immunoglobulins. In: Glynn LE, Steward MW, editors. Structure and function of antibodies. London: Wiley; 1981. Underdown BJ, Switzer I, Jackson GD. Rat secretory component binds poorly to rodent IgM. J Immunol 1992;149:487–9. Vene SL. Diagnosis of rickettsial infections in Swedish patients: detection of specific IgM. Eur J Epidemiol 1989;5:436–7. Volkov VV, Lapuk VA, Kayushina RL, Shtykova EV, Yu E, Malfois VM, et al. Lowresolution structure of immunoglobulins IgG, IgM and rheumatoid factor IgM-RF from solution X-ray scattering data. J Appl Cryst 2003;36:503–8. Vollmers HP, Zimmermann U, Krenn V, Timmermann W, Illert B, Hensel F, et al. Adjuvant therapy for gastric adenocarcinoma with the apoptosis-inducing human monoclonal antibody SC-1: first clinical and histopathological results. Oncol Rep 1998;5:549–52. Wichman A, Borg H. Purification of human Immunoglobulin M by affinity chromatography on protamine-sepharose. Biochim Biophys Acta 1977;490:363–9. Zimmerman SE, French ML, Allen SD, Wilson E, Kohler RB. Immunoglobulin M antibody titers in the diagnosis of Legionnaires' disease. J Clin Microbiol 1982;16:1007–11.