Monoclonal Antibody Conjugation via Chemical Modification

December 1, 2003
Muctarr Sesay
Muctarr Sesay

Muctarr Sesay, PhD, is chief scientific officer and vice-president of Bioconjugation at Goodwin Biotechnology.

BioPharm International, BioPharm International-12-01-2003, Volume 16, Issue 12

Bioconjugation chemistry is the joining of biomolecules to other biomolecules, small molecules, and polymers by chemical or biological means.

Bioconjugation chemistry is the joining of biomolecules to other biomolecules, small molecules, and polymers by chemical or biological means. This includes the conjugation of antibodies and their fragments, nucleic acids and their analogs, and liposomal components (or other biologically active molecules) with each other or with any molecular group that adds useful properties. These molecule groups include radionuclides, drugs, toxins, enzymes, metal chelates, fluorophores, haptens, and others.1-3, 7-9

The conjugation of monoclonal antibodies for therapeutic purposes is currently undergoing a flurry of research and development activity.2,11-12 These developments may not only change the way drugs are delivered, but could lead to safer protocols by dramatically reducing the dose to achieve efficacy without causing harmful side effects. Monoclonal antibodies can be utilized for drug or isotope deliveries, but as the biological functions of monoclonal antibodies are often sensitive to subtle variations in structure, attention must be paid to the chemical aspects of preparing, purifying, and characterizing these conjugates. The most commonly employed method for covalently crosslinking monoclonal antibodies to other molecules is the use of special reagents.

This article will primarily focus on the experimental design, protocols, and procedures for the preparation of in vivo therapeutic drug conjugates, such as the conjugation of monoclonal antibodies to other molecules using ligand and metal chelates, toxins, cancer drugs, and other proteins under cGMP or general large-scale manufacturing protocols. It focuses only on the chemical modification of the antibodies. Preparation of antibody fragments such as Fab' and F(ab')2 via enzymatic processes will not be discussed. For more information on the preparation of antibody fragmentation, see Hermanson (pages 478-482).1

A simple illustration of conjugation of a biomolecule is depicted in Figure 1. It essentially involves the activation of a reagent and a biomolecule with Z-A and Z-B crosslinkers, respectively. The reactive moiety Z must react efficiently, yet must not crosslink with the reagent and biomolecule. The linkers A and B must be reactive enough to couple with high efficiency, yet stable enough to be stored in solution or circulated in the body for long periods. It is also important that no nonspecific (noncovalent) binding or sticking of molecules occurs, which would lead to potential reduction in the effectiveness of the drug conjugate and high background interference in any assay.

This article will address the following specific topics: reactive groups of proteins (specifically, monoclonal antibodies) that are available for modification, including their naturally occurring amino acids and reactive groups introduced by chemical modification; reagents that can be used to couple molecules; the reaction environment; the current status of experimental procedures used in laboratory preparation; purification and isolation of the conjugates; storage; and manufacturing.

Reactive Proteins

Monoclonal antibodies (and other proteins) are amino acid polymers containing a number of reactive side chains. In addition to, or as an alternative to, these intrinsic reactive groups, specific reactive moieties can be introduced into the polymer chain by chemical modification. These groups, whether they are naturally a part of the protein or are artificially introduced, may serve as "handles" for attaching a wide variety of drug molecules (center of Figure 2). The intrinsic reactive groups of antibodies are described below.

Figure 1. A generic scheme for bioconjugation reaction

Amines (lysines, α-amino groups). One of the most common reactive groups of proteins is the aliphatic ε-amine of the amino acid lysine. In general, nearly all antibodies contain abundant lysine. Lysine amines are reasonably good nucleophiles above pH 8.0 (pKa = 9.18)4 and therefore react easily and cleanly with a variety of reagents to form stable bonds (Equation 1).

Antibody-NH2 + Z-B →

Antibody-NHB + Z-H (1)

Other reactive amines that are found in proteins are the α-amino groups of the N-terminal amino acids. The α-amino groups are less basic than lysines and are reactive at pH~7.0. Some of them can be selectively modified in the presence of lysines. Since either N-terminal amines or lysines are virtually always present in antibodies, and since they are easily reacted, these aliphatic amines provide the most commonly employed method of antibody modification.

Thiols (cystines, cysteine, methionine). Another common reactive group in antibodies is thiol residue from the sulfur-containing amino acid cystine and its reduction product cysteine (or half cystine). Cysteine contains a free thiol group, which is more nucleophilic than amines and is generally the most reactive functional group in a protein. Thiols, unlike most amines, are generally reactive at neutral pH, and therefore can be coupled to other molecules selectively in the presence of amines (Equation 2). This selectivity makes the thiol group the linker of choice for coupling proteins. Methods that only couple amines (for example, amine reaction with gluteraldehyde) can result in formation of homodimers, oligomers, and other unwanted products (see Hermanson, pages 470-472).1

NH2-Antibody-SH + Z-B →

NH2-Antibody-SZ + BH (2)

Since free sulfhydryl groups are relatively reactive, proteins with these groups often exist in their oxidized form as disulfide groups. Immunoglobulin M is an example of a disulfide-linked pentamer, while the subunits of Immunoglobulin G are bonded by internal disulfide bridges. In such proteins, reduction of the disulfide bonds with a reagent such as dithiothreitol (DTT) is required to generate the reactive free thiol. However, this method also splits the chain linkage in these antibodies and reassembly of the chains to allow proper folding may not be possible. In addition to cystine and cysteine, some proteins also have the amino acid methionine containing sulfur in a thioether linkage. Selective modification of methionine is generally difficult to achieve and is seldom used as a method of attaching drugs and other molecules to antibodies. The literature describes the use of several thiolating crosslinking reagents such as Traut's reagent (2-iminothiolane), succinimidyl (acetylthio)acetate (SATA), and sulfosuccinimidyl 6-[3-(2-pyridyldithio)propionamido]hexanoate (Sulfo-LC-SPDP) to provide efficient ways of introducing multiple sulfhydryl groups via reactive amino groups. 1-3, 8

Carboxylic acids (aspartic acid, glutamic acid). Proteins contain carboxylic acid groups at the C-terminal position and within the side chains of aspartic acid and glutamic acid. The relatively low reactivity of carboxylic acids in water usually makes it difficult to use these groups to selectively modify proteins and other biomolecules. When this is done, the carboxylic acid group is usually converted to a reactive ester by the use of a water-soluble carbodiimide and reacted with a nucleophilic reagent such as an amine, hydrazide, or hydrazine.5 The amine-containing reagent should be weakly basic in order to react selectively with the activated carboxylic acid in the presence of other amines on the protein. Protein crosslinking can occur when the pH is raised above 8.0.

Figure 2. Three side groups provide the major antibody modification pathways. Starting from the unmodified antibody in the center, reactions with the amino group follow the red path, the alcohol group follows the blue path, and the carboxylic acid group follows the green path.

Sugar alcohols. Sodium periodate can be used to oxidize the alcohol part of the sugar within the carbohydrate moiety to an aldehyde (see Hermanson, pages 340-341).1 Each group can be reacted with an amine, hydrazide, or hydrazine as described for carboxylic acids. Since the carbohydrate moiety is predominantly found on the crystallizable fragment (Fc) region of the antibody, conjugation can be achieved through site-directed modification of the carbohydrate away from the antigen-binding site.

Reagents

This section will briefly describe many of the reagents used in protein modification. In order to understand how to use these reagents, it is necessary to know what reactive group(s) on the protein can be modified; which type of chemical reactions these reactive groups will participate in; and the nature of the chemical bonds that will result (Figure 2).

Amine-reactive reagents. These reagents react primarily with lysines and the α-amino groups of proteins. Some amine-reactive reagents are more reactive, and therefore less sensitive, than others. It is necessary to consider reactivity when choosing the best reagent for modification of a specific protein. Table 1 lists commercially available amine-reactive reagents.

Reactive esters (formation of an amide bond). Reactive esters, particularly N-hydroxy-succinimide (NHS) esters, are among the most commonly employed reagents for modification of amine groups. They are moderately reactive toward amines, with high selectivity toward aliphatic amines. Their reaction rate with aromatic amines, alcohols, phenols, and histidine is relatively low. The optimum pH for reaction in an aqueous environment is 8.0 to 9.0, and they form very stable aliphatic amine products. The NHS esters are slowly hydrolyzed by water, but are stable if stored well desiccated. Virtually any molecule that contains a carboxylic acid, or that can be chemically modified to contain a carboxylic acid, can be converted into its NHS ester. That explains why these reagents are among the most powerful protein-modification reagents available. A new class of NHS esters are now commercially available with sulfonate groups that have improved water solubility.

Table 1. Some commonly used homo- and heterobifunctional crosslinking and chelating reagents

Isothiocyanates (formation of a thiourea bond). Isothiocyanates behave like NHS esters. They are amine-modification reagents of intermediate reactivity and form thiourea bonds with proteins. They are somewhat more stable in water than NHS esters and react with protein amines in aqueous solution (optimally at pH 9.0 to 9.5). Since this is a higher pH than optimal for NHS esters (which undergo competing hydrolysis at pH 9.0 to 9.5), isothiocyanates may not be as suitable as NHS esters when modifying proteins that are sensitive to alkaline conditions.

Aldehydes (formation of imine, Schiff's base, reduction to secondary amine bond). Aldehyde groups react under mild aqueous conditions with aliphatic and aromatic amines, hydrazines, and hydrazides to form an imine intermediate (Schiff's base). A Schiff's base can be selectively reduced with mild or strong reducing agents (such as sodium borohydride or sodium cyanoborohydride) to derive a stable alkyl amine bond. This method of amine modification can successfully be employed in situations in which the antibody is modified away from the antigen-binding site via the oxidation (typically with sodium periodate) of the alcohols on the carbohydrate moiety of the Fc region.

Miscellaneous amine-reactive reagents (anhydrides). Other reagents that have been used to modify amines are acid anhydrides. For example, diethylenetriaminepentaacetic anhydride (DTPA) is a bifunctional chelating agent that contains two amine-reactive anhydride groups. It can react with N-terminal and ε-amine groups of the proteins to form amide linkages. The anhydride rings open to create multivalent, metal-chelating arms able to bind tightly to metals in a coordination complex. This type of reaction is particularly useful in the preparation of radiolabeled immunoconjugates.6

Thiol-reactive reagents. Thiol-reactive reagents are those that will couple to thiol groups on proteins, forming thioether-coupled products. These reagents react rapidly at slight acidic to neutral pH and therefore can be reacted selectively in the presence of amine groups.

Haloacetyl derivatives (formation of a thioether bond). These reagents (usually iodoacetamides) are common reagents for thiol modification. In antibodies, the reaction takes place at cysteine groups that are either intrinsically present or that result from the reduction of cystine's disulfides at various positions of the antibody. The thioether linkages formed from any reaction of haloacetamides are very stable. However, iodoacetamide modification reagents are unstable in light, especially in solution. They must be protected from light during reaction and in storage. The level of control to achieve reproducibility required for large scale manufacturing conditions may be difficult to achieve.

Tips for Preparing Therapeutic Monoclonal Antibody Conjugates

Maleimides (formation of a thioether bond). The reaction of maleimides with thiol-reactive reagents is essentially the same as with iodoacetamides. Maleimides react rapidly at slight acidic to neutral pH. Above pH 8.0, maleimides can undergo hydrolysis to form nonreactive maleamic acids.

Aldehyde and carboxylic acid-reactive reagents.Amines, hydrazides, and hydrazines (formation of amide, hydrazone, or alkyl amine bonds). Amines, hydrazides, and hydrazines can be coupled to carboxylic acids of proteins after the activation of the carboxyl group by a water-soluble carbodiimide. As mentioned previously, the amine-containing reagent must be weakly basic so that it reacts selectively with the carbodiimide-activated protein in the presence of the more highly basic ε-amines of lysine to form a stable amide bond.

Amines, hydrazides, and hydrazines also can react with aldehyde groups, which can be generated on antibodies by periodate oxidation of the carbohydrate residues on the antibody. In this scenario, a Schiff's base intermediate is formed, which can be reduced to an alkyl amine through the reduction of the intermediate with sodium cyanoborohydride (mild and selective) or sodium borohydride (strong) water-soluble reducing agents.

Bifunctional reagents. Bifunctional crosslinking reagents are specialized reagents that will form a bond between different groups, either on the same molecule or two different molecules. These reagents can be divided into two kinds, homobifunctional reagents (those with the same reactive group at each end of the molecule) and heterobifunctional reagents (those with different reactive groups at each end of the molecule). Recent trends appear to strongly favor the use of heterobifunctional cross-linkers where the bifunctional reagent has two reactive sites, each with selectivity toward different functional groups (for example, an amine reactive and a thiol reactive). Many bifunctional reagents are commercially available with variable chain lengths and water solubility.1, 3, 8

Table 2. Status of immunoconjugate drug products

Miscellaneous reagents. Photoactive reagents capable of being activated by light can produce a reactive intermediate for coupling to various functional groups on biomolecules. Generation of highly reactive intermediates such as nitrene, carbene, and diradical species employed in photoaffinity labeling is of little value since such species are highly nonspecific and often capable of reaction even with alkyl side-chains.9 These reagents have not been widely utilized in the modification and coupling of monoclonal antibodies.

Reaction Environment

It is imperative that the scientist understands the practical aspects of reacting with large, complex, conformationally sensitive, water-soluble antibodies and other biomolecules or small organic molecules. The buffers, the cosolvents, and the reaction conditions all must be considered in order to obtain a scalable process that can be validated and used under aseptic, GMP manufacturing conditions.

Buffers. In general, conjugation should be carried out in a well-buffered system at an optimal pH for the reaction. The ionic strength of the buffer in most cases should be around 10 to 100 mM. Phosphate buffers should be well suited for the modification of thiol groups and α-amino groups, because this occurs selectively at physiological pH (7.0 to 7.5). For lysine amines, which require a more alkaline pH (8.0 to 9.2), carbonate-bicarbonate or borate buffers work well. In some cases, the choice of buffer will be dictated by the compatibility of the protein. For example, since many of the conjugations of antibodies involve the utilization of amine groups, it is necessary to avoid amine-containing buffers like tris(hydroxy-methyl)aminomethane (TRIS).

Cosolvents. Many of the available crosslinking reagents are sulfonated and water-soluble at millimolar to micromolar concentrations. These reagents need no cosolvent for solubilization. If the reagents are not soluble in aqueous medium, water-miscible cosolvent - such as dimethylformamide (DMF), dimethyl-sulfoxide (DMSO), and some alcohols - must be used to dissolve the reagent without causing decomposition. Furthermore, the cosolvent must not cause irreversible denaturation or precipitation of the antibody. DMSO and DMF are the most commonly used cosolvents, usually up to 10% (w/v) concentrations.

Temperature. In general, conjugation reactions involving antibodies (and most other proteins) should be done between room temperature and near-freezing temperature (~25°C to 2°C). Lower temperature tends to slow down the reaction rate and increase the selectivity and control of the reaction. This could result in fewer side reactions and more consistent and reproducible results.

Protein concentration. Typically, a 1 to 10 mg/mL protein concentration is recommended for protein conjugation with an optimum range of 5 to 10 mg/mL. Concentrations of less than 1 mg/mL have been successfully reported, but these require a longer incubation time to achieve adequate conjugation (based on unpublished data).

pH. During the modification of primary amines, only the unprotonated form is reactive and, therefore, it is necessary to maintain a pH at which a significant proportion of amines are unprotonated. When working with lysine, it was found that at an average pKa above 9, the reaction runs at a faster rate. However, some crosslinking reagents may undergo hydrolysis and some antibodies (and other proteins) tend to be unstable at higher pHs. The α-amino groups of the N-terminal amino acids are less basic than lysines and are reactive at pH ~7.0. These amino groups are sometimes preferentially modified when the reaction is run at neutral pH. Thus if the protein and crosslinking reagents are stable, the reaction should be performed close to pH 9.0. For less stable reagents, the reaction should be performed near pH 7.0.

Reaction time. In general, one to two hours should suffice for conjugation reactions to go to completion. Longer reaction times are acceptable, since the degree of labeling is generally limited by the molar ratio of the reagent to antibody, not the reaction time. Many published procedures indicate overnight reaction times and incubations performed at 2°C to 8°C.1, 3, 8 The more reactive the reagent is, the shorter the reaction time.

Molar ratio of the reactants. Antibodies are relatively sensitive to substitution, since there are usually reactive amino acid side chains (amines, thiols) in or near the antigen binding sites. For this reason, a low to moderate degree of labeling is preferable in order to preserve binding specificity. Over-labeling can result in decreased solubility which may reduce the overall activity of the conjugates. In some cases it causes precipitation of the antibody conjugate.

Laboratory Methods

The methods below are widely employed for conjugating amine-reactive and thiol-reactive and other reagents to antibodies. The wide variety of experimental conditions required for activating and coupling antibodies with bifunctional and other reagents make it difficult to generate a simple general procedure. The reader is advised to consult the literature for specific procedures. We recommend three books for initial consultation: Hermanson,

1

Aslam,

3

and Wong.

8

Although the procedures described below are for preparing antibody conjugates, they can also be utilized to prepare many other proteins and biological molecules.

Amine-reactive reagents. Use the following general procedure for the first trial. It is adaptable to amine-reactive molecules such as tags and probes, biotin, homobifunctional and heterobifunctional crosslinkers, haptens, carbodiimides, bifunctional chelating agents (for radioimmunoconjugates), and toxins. The procedure may be modified after the degree of substitution has been determined in analysis of the final product.

For the basic lysine amines, prepare the antibody at 1 to 10 mg/mL in 50 to 100 mM sodium carbonate-bicarbonate (or borate) buffer pH 8.2 to 9.2 at room temperature. For less-basic α-amino groups, which react selectively at physiological pH, prepare the antibody in 10 to 50 mM sodium phosphate buffer, with 150 mM sodium chloride, pH 7.2 to 7.5. Note that amine-containing buffers such as TRIS are not recommended, since these buffers will compete with the antibody in reacting with the modifying reagent.

Add sufficient antibody-modifying reagent from a freshly prepared stock solution, which contains about 5 to 10 mol of isothiocyanate or succinimide ester for each mole of protein. Note that for carboxylic acid-containing reagents such as bifunctional chelating agents DTPA and DOTA, 5 to 10 mols of the chelating agent per mol of antibody should be used in the initial conjugation. These reactions only work through the initial activation of the carboxylic acid–reactive groups of the chelator with a water-soluble carbodiimide reagent such as EDC. Follow up by stabilizing the active acylisourea intermediate with NHS ester or immidazole. React the resulting product (activated and stable chelating reagent) with the amino group of the antibody to generate the modified antibody-chelator conjugate. Such conjugates are useful as cancer therapeutic agents through the preparation of radiolabeled antibodies. In general, the preparation of antibody-chelating agent via the reaction of the amino group and carboxylic group is not very efficient.

Gosh et al. reported that no more than 0.7 mols of the carboxylic acid containing modification reagent per mol of the proteins were used up.8 For more details on the experimental procedures for preparing antibody reactive-amines and bifunctional chelating agent with reactive carboxylic acids conjugates, see Hermanson.1

Most commercially available protein modification reagents have some solubility in water. For insoluble reagents, a stock solution should be prepared immediately before use in a water-miscible solvent such as DMF, DMSO, or dioxane (prepared dry). Prepare a stock solution of the antibody-modifying reagent at about 5 to 10% (w/v) of the volume of conjugation buffer in order to minimize denauturation or precipitation of the antibody. Add the solution of modifying reagent in small amounts to the antibody solution. Gently mix each time to ensure homogeneous reaction mixture over a period of a few minutes at room temperature. Following complete addition of the modifying reagent, ensure that the mixture is homogeneous and incubate at room temperature for one to two hours without further mixing. As a matter of convenience, the reaction mixture can be incubated overnight at 2°C to 8°C. Following incubation, quench the reaction mixture using an amine-containing reagent such as excess TRIS buffer. Experience shows that quenching the reaction mixture allows for consistency and reproducibility when preparing the conjugates.

Purify the antibody conjugate from unreacted modifying reagent using the purification device most appropriate to the development and manufacturing scale as described later in this article.

Thiol-reactive reagents. Antibody conjugates can be prepared from thiol-reactive reagents such as maleimides, iodoacetates, and homo-bifunctional and heterobifunctional crosslinkers. In general, antibodies exist in their stable oxidized (disulfide) form. As described earlier, the disulfide form is usually converted into the more reactive and relatively unstable free thiol. This is achieved via the reduction of the disulfide bonds with reagents such as DTT, 2-MEA, BME, and TCEP.

If a full-length and intact antibody is needed, several sulfur containing reagents can be attached onto the reactive amino group. Usually a reduction of SATA and SPDP via hydroxyl amine, a reduction of Sulfo-LC-SPDP via DTT, or the hydrolysis of Traut's reagent generates modified antibodies containing multiple, reactive sulfhydryl groups. Be aware of the highly reactive nature of some sulfhydryl groups that can be oxidized to disulfides. It is best to do all reactions in an oxygen-free environment (for example under an argon or nitrogen blanket) and in the presence of ethylenediamine tetraacetic acid (EDTA) to minimize metal-catalyzed oxidation of sulfhydryl groups into the more stable and less-reactive disulfide groups. As with the reactive esters and isothiocyanates, use only freshly prepared reagents. Protection from light (particularly for iodoactamide) is important. Due to the reactive nature of thiol groups, use only filtration to purify the thiolated antibody. Avoid using dialysis since it is time consuming and may result in the conversion of sulfhydryl to disulfide.

Step 1. Prepare the antibody at 1 to 10 mg/mL in a suitable conjugation buffer (10 to 100 mM phosphate or TRIS, 150 mM sodium chloride, 1 to 5 mM EDTA at pH 6.5 to 7.5) at room temperature. At this pH range, the thiol groups on the antibody are sufficiently nucleophilic so that they react almost exclusively with the reagent in the presence of the more abundant protein amines, which are protonated and relatively unreactive. At pH 6.5 to 7.0, the reaction of the maleimide group with sulfhydryls proceeds at a rate 1,000 times greater than its reaction with amines.11 It is advisable to deoxygenate all buffers to prevent reformation of disulfide.

Step 2. Add sufficient protein modification reagent to the antibody prep from a stock solution to contain 5 to 15 mol of the reagent for each mole of protein. Upon completion of the reaction, quench the reaction mixture with a thiol-containing reagent such as DTT, BME, or glutathione to consume excess modification reagent.

Iodoacetamides. Reactions with iodo-acetamides should be carried out in the dark, since light can cause reagent decomposition. Slowly add the stock reagent solution in small amounts over a period of about one to three minutes with gentle mixing. Once the reaction mixture is homogeneous, let it incubate without further mixing for one to two hours at room temperature.

Maleimides. Reaction conditions for maleimides are essentially the same as with iodoacetamides. The selectivity of maleimides toward thiol groups is greater than iodo-acetamides, thereby allowing somewhat more latitude in the buffer pH. Also, at pH above 8.0, the maleimide group can undergo a competing hydrolysis reaction to form maleamic acid, which is unreactive to thiol groups. As with iodoacetamide, slowly add the stock reagent solution in small amounts with gentle mixing to the antibody solution over approximately one to three minutes and let it incubate without further mixing for one to two hours at room temperature.

Conjugate Purification and Storage

Usually, molar excess of the antibody modifying reagents is used during the preparation of antibody conjugates. It is necessary to remove excess non-covalently bound modifying reagents from the antibody conjugates. There are several techniques and commercially available devices one can utilize in the purification of antibody conjugates.

Dialysis is simple and inexpensive. However, it is the most inefficient and time-consuming method of purifying antibody conjugates. Dialysis may be more appropriate during early conjugation research and small-scale development, but it may not be amenable to large-scale manufacturing. In general, not all molecules dialyze efficiently. The rate of dialysis depends on the relative affinity of the modifying reagent for the protein versus the dialysis solution. Molecules that are sparingly soluble in water or strongly adsorbed to the protein surface will take a long time to dialyze. Dialysis works best when the labeling reagent and its unreacted byproduct are hydrophilic. When purifying conjugates by dialysis, a dialysis buffer volume of at least 100 times the volume of conjugate solution should be used, and the dialysis buffer should be changed at least three to five times, allowing at least 4 hours for dialysis between buffer exchanges.

Stirred cell filtration is a widely used procedure for purification of antibody conjugates. This specialized filtration system (also known as microfiltration, ultrafiltration, and diafiltration) consists of a flat membrane held in place over the outlet of a pressurized reservoir. The device is pressurized to force fluid through the membrane while retaining the macromolecules. A second reservoir contains the final storage buffer for the conjugate product. The membrane material is available in varying pore sizes.

Stirred cell devices have been developed to purify products with volumes from a few milliliters to approximately 2,000 mL and are commonly used in research and development laboratories and early manufacturing for clinical study of drug products. This device enables both purification and concentration of the conjugate product.

Tangential flow filtration (TFF) is also a pressure-driven filtration process for purification of antibody conjugates. TFF uses a large membrane surface (commercially available in varying pore sizes). Unlike stirred cell filtration, TFF systems are designed to handle from a few milliliters to hundreds of liters of sample without the concerns associated with pressurized systems. In TFF, the fluid is pumped tangentially along the surface of the membrane. Pressure is applied through the membrane to the filtrate side, allowing macromolecules that are too large to pass through the membrane to be retained on the upstream side. Once it is optimized for flow, flux and transmembrane pressures, a TFF system can be predictably scaled from bench to industrial manufacturing (phase 1 and 2 clinical studies and beyond). Furthermore, the TFF system can use a large surface area while maintaining a minimum holdup volume. This allows for a shorter processing time than can be achieved with stirred cells. TFF can concentrate the desired product usually with over 90% recovery. Filtration and concentration can be performed in a completely closed system.

Gel filtration chromatography is a simple and reliable chromatographic technique for separating molecules by size. The labeled conjugate is separated from the excess noncovalently-bound labeling reagent and other small molecules while allowing the larger antibody conjugate to pass through the void space in the porous gel matrix. A gel filtration technique is generally faster and more effective in removing most hydrophilic and hydrophobic labeling reagents than other filtration systems.

In order to achieve high-resolution separation, ~5 to 10% of the column's volume should be loaded with the sample. Sometimes up to 30% sample volume can be loaded. There are many sizes of prepacked gel filtration columns available. For development and small-scale purification (<1 mL to ~15 mL), prepacked columns such as PD-10 and HiPrep desalting columns are available. For larger-scale purification (>15 mL), columns such as the XK series or equivalent can be utilized. Generally, you will both pack and qualify such columns with the appropriate gel prior to use. For even higher volume (>100 mL), the sample volume should be reduced (for example, via TFF) to ~10% column volume. You may have to run multiple columns or connect two or more columns in series. It should be noted that the utilization of gel filtration chromatography produces a relatively low concentration of protein and the product may need more steps to achieve a higher concentration.

Miscellaneous. With the exception of dialysis, all of the separation techniques described above work well if the antibody is significantly larger (>3-fold) than the modifying or coupling reagent. For reagents (mostly protein and other biological molecules) that are similar in size or larger than the antibody, one must resort to other purification techniques such as affinity chromatography, ion-exchange chromatography, and hydrophobic interaction chromatography (see Aslam and Dent, pages 588-665).3

Storage. The researcher must determine the most appropriate buffers to be utilized for both short- and long-term storage of their therapeutic antibody conjugate. From our experience and published procedures, buffers close to physiological pH, such as phosphate buffered saline (PBS), are generally recommended.2

Immunoconjugate Drugs Today

While only a few immunoconjugate drug products are currently available commercially, many monoclonal antibodies that are conjugated to other biomolecules and pharmaceutical drugs are currently undergoing clinical trials. Table 2 reports the latest progress.

FDA has approved Bexxar and Zevalin for patients with CD20-positive follicular, non-Hodgkin's lymphoma (NHL). The monoclonal antibodies in these drugs recognize and attach to a particular part of a B-cell, the CD20 antigen. This allows Bexxar and Zevalin to specifically target B-cells, destroying the malignant NHL B-cells and also some normal B-cells. Therefore, Zevalin and Bexxar are considered dual-action therapies because they pair the tumor-targeting ability of the monoclonal antibodies with the cancer-killing radioisotope.

It should be noted that tethering a drug molecule to an antibody could dramatically change the biological properties of the drug. Also, because they are large proteins, antibodies do not penetrate cells except under special circumstances (such as with antibody fragments). This can greatly reduce the toxic side effects of a drug - an important goal in many cases. Radiation safety is an important factor when handling and administering Bexxar and Zevalin. The volatile nature of many radiohalogen species (for example, I131) makes handling of these materials somewhat different than other radioactive materials. This is particularly true for handling radioiodine, as it is sequestered by the thyroid gland. Therefore, use a fume hood or vented box or (if large amounts of activity are used) a "hot cell" when handling these radiolabeled materials.

Not enough information is available to adequately assess other immuno-drug conjugates currently undergoing phase 1 and 2 clinical trials.

Through conjugation chemistries, traditional pharmaceutical drugs can be linked to monoclonal antibodies in order to deliver targeted doses, prevent breakdown, and increase bioavailability in circulation. The techniques for coupling drugs to monoclonal antibodies have become more sophisticated with the design of novel linkers and protecting groups that afford unique properties such as stability in the serum. Prior to starting conjugation of biomolecules, it is important to consider the methods used in maintaining the functionality of the antibody, biological properties of the drugs and other biological molecules, and the scalability of the manufacturing process. To date, the majority of drugs used in immunoconjugates has been limited to clinical treatment of cancer and various blood-borne diseases.

References

1. Hermanson GT.

Bioconjugate Techniques.

New York: Academic Press; 1996.

2. Pincus SE, Lukacher N, Mohamed N, Sesay M, Zabinski R, Ebelle R, et al. Evaluation of antigen base heteropolymer (AHP) for treatment of systemic lupus erythmatosus in a nonhuman primate model. Clinical Immunology 2002; 105(2):141-154.

3. Aslam M, Dent AH. Bioconjugation: protein coupling techniques for the biomedical sciences. Houndsmills, England: Macmillan Publishers; 1999.

4. Fasman GD. Practical handbook of biochemistry and molecular biology. Boca Raton, FL: CRC Press; 1989

5. Ghosh SS, Kao PM, Kwoh DY. Synthesis of 5'-oligonucleotide hydrazide derivatives and their use in preparation of enzyme-nucleic acid hybridization probes. Analytical Biochemistry 1989; 178(1):43-51.

6. Griffiths GL, Goldenberg DM, Jones AL, Hansen HJ. Radiolabeling of monoclonal antibodies and fragments with technetium and rhenium. Bioconjugate Chemistry 1992; 3(2):91-99.

7. Means GE, Feeney RE. Chemical modification of proteins: history and applications. Bioconjugate Chemistry 1990; 1(1):2-12.

8. Wong SS. Chemistry of protein conjugation and crosslinking. Boca Raton, FL: CRC Press; 1991.

9. Smith RA, Dewdney JM, Fears R, Poste G. Chemical derivatization of therapeutic proteins. Trends in Biotechnology 1993; 11(9):397-403.

10. Smyth, DG, et al. Reaction of N-ethylmaleimide with peptides and amino acids. Biochem Journal 1964; 91:589.

11. Morrow KJ. Genetic Engineering News 2003; 23(14):42.

12. Pietersz GA. The linkage of cytotoxic drugs to monoclonal antibodies for the treatment of cancer. Bioconjugate Chemistry 1990; 1(2):89-95.