1. Dilute the sample in assay buffer (pH 6.0) to 1, 0.5, and 0.25 mg protein/mL. Note that Subheadings 3.5.2. and 3.5.3. can be performed simultaneously.
2. Add 100-^L aliquots of the antibody dilutions in duplicate to separate 0.9 mL aliquots of the HABA-avidin or streptavidin solution and stir at room temperature for 5-10 min. If the DOL of some of the samples is very high, a precipitate may form during the incubation that should be removed by centrifugation prior to spectrophotometry analysis. The decrease in A500 for each sample should then be recorded. The average value from each pair of samples is used to determine the nmol of biotin in the various aliquots of antibody tested. These values are derived from the biotin standard curve. The ratio of the nmol of biotin and the nmol of antibody assayed represents the DOL.
3. Determine the DOL. Eq. 4 should be used to obtain this value:
[(nmol biotin x MW of antibody or antibody fragment x 10-6/(mg/mL of antibody or fragment x 0.1 mL)] = mol biotin/mol antibody = DOL (Eq. 4)
A brief discussion of the limitations of the HABA assay is found in Note 8.
1. The final concentration of the antibody protein in the biotinylation reaction is an important variable. Preferably, the antibody should be at 3-10 mg protein/mL. Antibodies at lower concentrations can still be labeled, but adjustments in the MR of the reactive label must then be made. Since it is often necessary to biotinylate small quantities of antibodies, the FluoReporter® Mini-Biotin-XX Protein Labeling Kit (Molecular Probes, Inc.) mentioned in 2.2 is available for researchers who want to directly biotinylate 0.1-3 mg antibody. Researchers should determine whether their antibody samples contain bovine serum albumin or gelatin, which are often added by vendors as preservatives. These protein additives are generally present at high concentrations relative to the antibody (especially for monoclonal antibodies), making them inconvenient to remove. Consequently, the biotinylation of an antibody containing a protein stabilizer requires a higher MR of labeling reagent: antibody in the reaction. The resulting biotinylated mixture of antibody and stabilizer is likely to generate high background in imaging and other applications. If researchers want to remove stabilizer proteins from antibodies, they can use various affinity chromatographic methods (e.g., protein A- or G-agarose columns). Descriptions of these affinity-based methods are beyond the present scope. The reader is referred to Chapter 7 (this volume) and to refs. 13-15 for protocols. Submicrogram amounts of mouse, rabbit, and human antibodies also can be indirectly biotinylated or labeled with other tags using Zenon Technology (Molecular Probes, Inc.). This mode of immunolabeling is ideal for antibodies containing stabilizer proteins or those found in ascites fluids.
2. Many reactive biotin derivatives are available and we have not described all of them. The structures of the reactive biotin derivatives discussed in this chapter are shown in Table 1. Reactive biotins based on biocytin (e-N-[D-biotinyl]-L-lysine) are shown in Table 2. Typical intact IgG and IgM molecules contain approx 90 and 350 lysine residues, respectively (4). Many of these lysines are exposed on the protein surface and are available for labeling with amine-reactive biotin derivatives. For labeling amines, we recommend using succinimidyl esters of biotin (biotin-NHS esters) because of their superior amine reactivity and their ease of use. Biotin SE can be used, but we recommend biotin-X-SE or biotin-XX-SE instead. The X and XX represent 1 or 2 aminohexanoic acid moieties that are attached to the carboxyl group of biotin in order to provide a 7- or 14-atom spacer between the biotin and the reactive ester (see Table 1). After biotinylation, these spacers increase the distance between the biotins and the antibody molecules to which they are attached. This makes the biotins more accessible to reporter group-labeled biotin-binding proteins. Biotin sulfosuccinimidyl esters, which have enhanced H2O solubility compared to SE derivatives, can also be used. Desthiobiotin-X-biotin SE or 2-iminobiotin SE are available for those who want to biotinylate antibodies on amines with biotin analogs whose binding to biotin-binding proteins is reversible (5,8). For information on other amine-reactive biotin and desthiobiotin derivatives, the reader is referred to refs. 1-3. For further information on biotinylation kits based on amine-reactive biotins and desthiobiotin, the reader should visit www.probes.com on the Internet. Although typical intact IgG and IgM proteins contain approx 30-40 and approx 180 cysteine residues, respectively, these are linked in inter- and intrachain disulfide bonds that contribute to the formation and maintenance of the classical Y-shaped immunoglobulin structure (4). Consequently, one or more of these disulfides must be reduced to free thiol groups before they will react with thiol-reactive biotin derivatives. This procedure is described in Subheading 3.2.1. It is worth noting that many researchers prefer not to reduce their antibodies because loss of biological activity may occur. However, biotinylation of antibody F(ab')2 and F(ab') fragments with thiol-reactive biotins is commonly performed. The F(ab')2 fragments are typically generated by pepsin digestion of whole immunoglobu-lins using well-established protocols (16). The free thiol-containing
F(ab') fragments are generated by treating the F(ab')2 fragments with reducing agents in the presence of EDTA (16). F(ab) fragments can also be generated by digestion of antibodies with papain (16). Biotin derivatives containing thiol-reactive groups include biotin iodoacetamide, biocytin maleimides, and biotin-HPDP (see Tables 1-3). Biotinylation with biotin-HPDP is reversible because the biotin is attached to the protein by a pyridyl disulfide bond that is cleavable with strong reducing agents like P-mercaptoethanol or sodium boro-hydride (2). Researchers should be aware that exposure to reducing agents may destroy the antibody's biological activity and can dissociate F(ab')2 or F(ab') fragments into smaller components. A variety of biotin derivatives for labeling the carbohydrate moieties found on antibodies are available (see Tables 1-3). Carbohydrates are usually attached to the Fc portion of intact antibodies, although glycosylated F(ab) fragments have been reported (4). Biotinylation of antibodies on their carbohydrates is often preferable because the carbohydrates are at the opposite end of the antibody molecule from the antigen-binding sites. However, some monoclonal antibodies are not glycosylated and researchers should verify the presence of carbohydrate residues before attempting biotinylation by this approach. Before labeling antibody carbohydrates, however, they must be oxidized to aldehydes with periodate ion (see Subheading 3.3.1.). Biotin-XX-hydrazide, desthiobiotin-X-hydrazide, and biocytin hydrazide (enhanced H2O solubility) are typical reactive labeling reagents for aldehydes (see Tables 1-3). Some researchers believe that the biotin hydrazone-antibody bonds formed in these reactions are relatively unstable. Thus, it is advisable to reduce the hydrazone bonds to more stable substituted hydrazide bonds by treating the conjugate with sodium cyanoborohydride (see Subheading 3.3.2.). Photoreactive biotin (photobiotin) derivatives are not commonly used to biotinylate antibodies because they are nonspecific labels that react with a variety of protein functional groups (see Subheading 3.4.2.). However, photobiotins, which typically contain azide moieties on aromatic rings, are highly reactive and they can be used under physiological conditions (2). It has been reported that the degree of biotinylation achieved with photobiotins is lower than other reactive biotins because the reactive nitrenes generated during irradiation have a short lifetime (2,17).
3. Many antibodies are available commercially as solutions in PBS (pH 7.2-7.5). This buffer is easily adjusted to pH 7.5-8.3, the optimum for amine labeling, by adding sodium bicarbonate. Because the reactivity of lysine amines increases at basic pH, we typically adjust the antibody solution to pH 8.0-8.3 by adding to the sample one-tenth volume of a 1 M sodium bicarbonate solution in deionized H2O. This yields a final bicarbonate concentration of 100 mM. For convenience, the 1 M sodium bicarbonate can be prepared in bulk, divided into convenient aliquots, and stored at -20°C for later use. Once defrosted, bicarbonate solutions can be stored at 4°C for up to 2 wk, if desired. Other buffers such as HEPES and EPPS containing tertiary amines are acceptable. However, buffers containing Tris-HCl, glycine, ethanolamine, triethylamine or other amine-containing components should not be used because these substances will react with the biotinylation reagents. Buffers containing <0.1% (w/v) sodium azide or thimerosal as preservatives are acceptable, but biocide concentrations higher than this should be reduced by dialysis or dilution. Antibodies containing up to 50% (v/v) glycerol can be biotinylated as described in this chapter, but the high viscosity of such solutions makes manipulating them very inconvenient. It is advisable to remove high concentrations of glycerol by dialysis before labeling. Researchers should consider always dialyzing their antibody starting materials extensively to remove unwanted buffer components and to equilibrate the protein in their buffer of choice. Lyophilized antibodies can be dissolved directly in 100 mM sodium bicarbonate buffer (pH 8.0-8.3) or another amine-free buffer in this pH range. Immunoglobulin M is unstable at basic pH, so buffers like PBS at pH 7.2-7.4 should be used when biotinylating these antibodies. Increasing the labeling reagent: protein MR can compensate for the lower reactivity of biotin SE derivatives in this pH range.
4. Gel filtration media such as Biogel P-30 (fine or medium) or Biogel P-6 (fine or medium; Bio-Rad Laboratories; Hercules, CA); Sephadex G-10, G-15, or G-25 (Amersham Biosciences; Piscataway, NJ); or equivalent matrices made by many other manufacturers are typically used. For optimum separation of the biotinylated antibody from unattached label, researchers should choose a gel matrix with an exclusion limit smaller than the molecular weight of the protein sample. Thus, the protein will elute in the void volume of the column while smaller unwanted components will be strongly retarded by the matrix. Neither biotin nor antibody molecules are colored, so the progress of the gel filtration column must be monitored by absor-bance at 280 nm. Small quantities of biotinylated antibodies can be separated from excess reactants by centrifugation through a variety of spin columns packed with one of the gel filtration matrices previously mentioned.
5. It is often necessary to concentrate the antibody starting material and/ or the biotinylated antibody after the unreacted biotinylation reagent has been removed. A variety of centrifugable concentrators containing molecular weight-selective, low protein-binding permeable membranes are available and are preferred because they save time and minimize product losses. Researchers should use devices containing membranes with MWCO <50,000 to concentrate intact antibodies. Devices with lower MWCO membranes (e.g., 10,000) should be used for F(ab)2 or F(ab') fragments, in particular. Other methods for concentrating protein solutions such as dialysis versus solid polyethylene glycol can be used, but these are time consuming and inconvenient and are not recommended.
6. Most reactive biotin derivatives are typically dissolved in anhydrous DMSO or DMF. Organic solvents are required to dissolve biotin SE derivatives because they are poorly soluble or insoluble in aqueous solutions. The H2O-soluble biotin sulfosuccinimidyl esters or biocy-tin derivatives are preferred if researchers do not wish to expose their antibodies to organic solvents. To reduce the amount of solvent added to the antibody, working solutions of reactive biotins can be prepared at concentrations >10 mg/mL. However, the volume of label added to the antibody must be decreased accordingly to avoid changing the MR. Once the label is dissolved, it should be used as soon as possible to minimize hydrolysis of the reactive esters. Any remaining label solution should be discarded. If researchers must reuse a reactive labeling reagent solution prepared in DMSO or DMF, it can be stored at -20°C for short periods. However, the reactivity of the stored reagent will decrease over time owing to hydrolysis. Upon reuse, the MR should be increased to compensate for decreased reactivity. How much to increase the MR must be determined empirically. If H2O-soluble biotin SE labels are used, the labeling reagent solution should be prepared immediately before use and then discarded. Stock solutions of biotin hydrazides can be stored at -20°C and reused several times.
7. Most researchers have only a small amount of antibody available. The FluoReporter® Mini-Biotin-XX Protein Labeling Kit from Molecular Probes is designed to label as little as 100 ^g of antibody and is recommended for direct biotinylation of such small quantities. Sometimes researchers have antibodies whose protein concentration is <1 mg/mL. If concentrating the protein is not an option, the labeling reagent: antibody MR should be increased to 30-40. If the antibody has a protein concentration >10 mg/mL, the MR ratio should be lowered accordingly. The antibody can also be diluted to <10 mg/mL.
8. Although it is a useful tool, the HABA assay is not ideal for determining the DOL for several reasons. First, it is not particularly sensitive, so researchers should consider more rigorous methods for determining the DOL if higher levels of sensitivity are desired (911). Second, the HABA assay is subject to considerable variability from experiment to experiment. Thus, it is essential to run the biotin standard curve each time the assay is used and to test multiple aliquots of the sample to obtain an average DOL value. However, the HABA assay is easy and inexpensive to perform, it requires no sophisticated equipment other than a spectrophotometer, and it is sensitive enough for most situations researchers are likely to encounter using the protocols described herein.
Was this article helpful?