Radioiodination of Antibodies

I. INTRODUCTION
Antibodies are usually radiolabelled in order to characterise the interaction of an antibody with its specific epitope. This characterisation may consist of the localisation of the interaction, e.g., on a Western blot or an immunoscintigraphic image, or alternatively the quantification of the interaction, e.g., in radioimmunoassay. The radionuclide employed and the technique used to incorporate it into the antibody depend on the application envisaged and the means of localisation and/or quantification. Both the in vitro and the in vivo uses of labelled antibodies span more than 50 years and, during this time, methods for labelling antibodies with at least 20 diffferent radionuclides have been developed. A detailed description of all the techniques used to incorporate all of these radioisotopes is not possible within the space limitations of this book and therefore this article is restricted to the techniques likely to be of interest to the majority of its readership, i.e., radioiodination of antibodies for in vitro applications. A more comprehensive review of methods for labelling antibodies with a wider range of radionuclides can be found in Mather (2000).

II. MATERIALS AND INSTRUMENTATION
The antibody to be labelled must be available in a purified form (any contaminating proteins will also be radiolabelled). The antibody concentration should be 100µg-5mg/ml in either 0.1M phosphate buffer, pH 7.4, or borate, phosphate, or HEPES buffer, pH 8, as indicated in the procedures.

Radioisotopes are available from Amersham Biosciences: Radioiodine Na¹²⁵I (100mCi/ml 3.7GBq/ml) (IMS30 Amersham), N-Succimidyl-3-(4-hydroxy-3- [¹²⁵I]iodophenyl)propionate, Bolton and Hunter reagent (IM5861). Unless otherwise indicated, all reagents are from Sigma Aldrich Chemical Company, Poole, UK: chloramine-T (40,286-9), sodium metabisulphite ($9000), disodium hydrogen phosphate (21,988- 6), sodium dihydrogen phosphate (S8282), Iodogen (Pierce Chemical Company, 28,600), dichloromethane (43,922-3), methanol (44,347-6), Sephadex-G50 fine grade (G-50-80), prepacked PD-10 column (Amersham Biosciences 17-0851-01), bovine serum albumin (BSA) (A7906), silica gel-coated plastic TLC sheets (Polygram 805013, Marchery-Hagel, Duren, Germany), silica gelimpregnated glass fibre (ITLC) sheets (61,885, Pall Corp., Ann Arbor, MI), and Whatman 3 MM Chr paper (Z27,085-7).

Many alternatives exist for the equipment used for these procedures. The following are the ones used in this laboratory: Gamma counter (LKB Ultragamma, Wallac, Finland), vortex mixer (MS2 minishaker, Staufen, Germany), microcentrifuge (MSE Microcentaur, Crawley, UK), rotamixer (Denley Spiramix, Thermo Life Sciences, Basingstoke, England), and Savant Speed-Vac (Westwood, MA).

All radiolabelling procedures must be performed with attention to good radiation safety practice. In particular, radioiodinations should be performed in a well-ventilated fume hood and all containers should be locally shielded by keeping them in small lead pots. A lead-glass L-shield should be used to reduce wholebody radiation doses.

III. PROCEDURES
A. Radioiodination with Iodine- 125 Using Chloramine-T as Oxidant
This procedure is adapted from that originally published by Hunter and Greenwood (1962).

Solutions

1. 0.1M phosphate buffer, pH 7.4: Add 19ml of 0.1M sodium dihydrogen phosphate solution to 81 ml of 0.1 M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.
2. Chloramine-T: 0.5mg/ml in 0.1M phosphate buffer, pH 7.4. Weigh 5 mg of chloramine-T in a universal container and add 10ml of 0.1M posphate buffer, pH 7.4. Swirl gently to dissolve. Keep cool until required. Prepare fresh.
3. Sodium metabisulphite: 0.5g/ml weigh 5mg of sodium metabisulphite in a universal container and add 10ml of 0.1M phosphate buffer, pH 7.4. Swirl gently to dissolve. Keep cool until required. Prepare fresh.
4. 0.5M phosphate buffer, pH 7.4: Add 19ml of 0.5M sodium dihydrogen phosphate solution to 81 ml of 0.15 disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.

Steps

1. Into a 1- to 2-ml polypropylene tube pipette 10-100 µg of the antibody and 50µl of 0.5M phosphate buffer, pH 7.4. Add the desired amount of iodine- 125, typically 100µCi-1mCi, and mix by gently drawing the solution up and down in the pipette tip.
2. Add 20µl of chloramine-T solution and mix again.
3. Cap the tube and leave for 5 rain.
4. Add 40µl of sodium metabisulphite solution and mix.
5. If desired, check labelling efficiency by ITLC (see Section III,E). This will typically range from 50 to 80%.
6. Separate labelled antibody from free iodine (see Section III,D).
7. If desired, determine the immunoreactive fraction of the radiolabelled antibody (see Section III,F).

B. Radioiodination with Iodine-125 Using lodogen as Oxidant
This procedure is adapted from that originally published by Fraker and Speck (1978).

Solutions

1. Iodogen tubes: Dissolve 1 mg of Iodogen in 10ml of dichloromethane in a glass or polypropylene container. Pipette 500µl into as many 2-ml glass or polypropylene test tubes as required. Evaporate the solvent in a Speed-Vac, with a stream of nitrogen, or by leaving in a laminar flow hood for 2-4h with the lights turned off. Cap the tubes and store in a closed container at -20°C for up to a year until required.
2. 0.1M phosphate buffer, pH 7.4: Add 19ml of 0.1M sodium dihydrogen phosphate solution to 81ml of 0.1M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.
3. 0.5M phosphate buffer, pH 7.4: Add 19ml of 0.5M sodium dihydrogen phosphate solution to 81 ml of 0.15 M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.

Steps

1. Into an Iodogen tube pipette 10-100µg of the antibody and 50µl of 0.5M phosphate buffer, pH 7.4. Add the desired amount of iodine-125, typically 100 µCi-1 mCi, and mix by gently drawing the solution up and down in the pipette tip. Wait for 10min mixing gently every 2-3 min.
2. Transfer the reaction mixture to a fresh test tube, wash the Iodogen tube with 0.5ml of 0.1M phosphate buffer, pH 7.4, and add to the mixture.
3. If desired, check labelling efficiency by ITLC (see Section III, E). This will typically range from 50 to 80%.
4. Separate labelled antibody from free iodine (see Section III, D).
5. If desired, determine the immunoreactive fraction of the radiolabelled antibody (see Section III,F).

C. Iodination of Antibody with "Bolton and Hunter" Reagent
This procedure is adapted from that originally published by Bolton and Hunter (1973).

Solutions

1. Antibody solution: 10-100 µg at a concentration of 2-5 mg/ml in 0.1M HEPES, phosphate, or borate buffer, pH 8.0.
2. Glycine solution: 0.2M in 0.1M phosphate buffer, pH 8.0.
3. N-Succimidyl-3-(4-hydroxy-3- [¹²⁵I]iodophenyl) propionate. Bolton and Hunter reagent [IM5861, Amersham Pharmacia Biotech, or equivalent].
4. 0.1M phosphate buffer, pH 7.4: Add 19ml of 0.1M sodium dihydrogen phosphate solution to 81ml of 0.1M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.
5. 0.1M phosphate buffer, pH 7.4 containing 0.05% polysorbate 20 (PBS/Tween): Add 50 µl of Tween 20 to 100ml of 0.1M phosphate buffer, pH 7.4. Mix gently but thoroughly. Store at room temperature for 4 weeks.

Steps

1. Into a 1.5-ml microcentrifuge tube (e.g., Eppendorf) pipette the required radioactivity of Bolton and Hunter reagent. Evaporate the solvent, ideally with a Speed-vac or, alternatively, under a gentle stream of nitrogen.
2. Add the required amount of antibody to the vial. Mix briefly and incubate for 30min at room temperature.
3. Add 0.5 ml of 0.2M glycine solution. Mix and incubate for a further 10min.
4. If desired, check labelling efficiency by ITLC (see Section III,E and use ITLC paper and 20% trichloracetic acid as mobile phase). This will typically range from 30 to 50%.
5. Separate labelled antibody from free iodine (see Section III,D).
6. If desired, determine the immunoreactive fraction of the radiolabelled antibody (see Section III,F).

D. Separation of Radiolabelled Antibody from Free Iodide
Solutions

1. Sephadex gel: Weigh out 1 g of Sephadex G-50 powder and add 15ml of deionised water. Mix well and either leave overnight or heat in a boiling water bath for 1 h to allow the gel to swell. Keep at 4°C for 4 weeks.
2. 0.1M phosphate buffer, pH 7.4: Add 19ml of 0.1M sodium dihydrogen phosphate solution to 81 ml of 0.1M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.
3. Bovine serum albumin solution: 1% BSA in phosphate- buffered saline, pH 7.4 (1% BSA/PBS). Weigh out 1 g of BSA and dissolve by gentle continuous mixing in 100ml of 0.1M phosphate buffer, pH 7.4. Keep at 4°C for 4 weeks.

Steps

1. Use either a prepacked PD-10 gel-filtration column or, if not available, prepare one as follows: Remove the barrel from a 10-ml disposable syringe and cap the luer tip. Plug the end of the syringe with a small circle of filter paper or lint dressing. Clamp the syringe vertically in a retort stand. Swirl the swollen Sephadex gel and pour as much as possible into the syringe. Allow the gel to settle for a few minutes. Then remove the luer cap and allow the liquid supernatant to run through into a waste container. Gently layer 10ml of deionised water on top of the gel and allow to run through to waste. Replace the luer cap and use the prepared column as soon as possible.
2. Clamp either a prepacked or the home-made column vertically in a retort stand and remove the luer cap.
3. Wash the column with 30ml of cold 1% BSA/ PBS.
4. Apply the labelled antibody reaction mixture to the surface of the column and allow it to run into the gel. Gently pipette 1 ml of cold 1% BSA/PBS onto the gel and collect the eluate in a test tube.
5. Repeatedly elute the column with ten 1-ml aliquots of 1% BSA/PBS and collect each 1 ml of eluate in a fresh, numbered test tube.
6. Pipette 10-µl samples from each of the eluate fractions into counting tubes and count them in a gamma counter in order to identify tubes containing the labelled antibody fractions (typically tubes 3-5). Use or store the contents as required.

E. Determination of Radiochemical Purity by TLC
Solution
85% methanol solution: Pour 85 ml of methanol into a measuring cylinder and make up to 100ml with deionised water. Store in a tightly closed container at room temperature for up to 4 weeks.

Steps

1. Cut a piece of chromatographic support material: Whatman 3-mm chromatography paper, silica gelcoated plastic TLC sheets, or silica gel-impregnated glass fibre (ITLC) approximately 1 x 10cm in size. Make a faint pencil mark 1.5 cm from one end.
2. Pour enough 85% methanol into a 10- to 15-cm-tall glass beaker or similar container until it is 0.5cm deep. Cover the beaker with a petri dish lid, aluminium foil, or similar.
3. Place a 1-btl spot of the sample to be analysed onto the centre of the pencil mark on the chromatographic strip and allow it to dry.
4. Using forceps, gently place the strip upright in the beaker with the pencil mark at the lower end just above the solvent level. Cover the beaker and allow the solvent to run up the support material.
5. When the solvent is about 5 mm from the top of the strip, remove it from the beaker using forceps and lay it on a clean piece of tissue to dry.
6. Cut the strip into upper and lower halves, place each half into counting tubes, and count the tubes in a gamma counter.
7. Calculate the labelling efficiency as follows:
   

   % Labelling efficiency =	Radioactive counts on lower half of strip	x 100%
   	Counts on lower half counts on upper half

F. Measurement of the Immunoreactive Fraction of Radiolabelled Antibody
This assay has been adapted from a method published by Lindmo et al. (1984). Modifications include a number of simplifications that make the assay easier and quicker but which could correctly be criticised if used out of context. This modified assay is, therefore, only recommended as a "quality control check" rather than as a way of determining the real immunoreactive fraction of the antibody for which the original published method is recommended.

Solutions

1. 0.1M phosphate buffer, pH 7.4: Add 19ml of 0.1M sodium dihydrogen phosphate solution to 81 ml of 0.1M disodium hydrogen phosphate solution. Check the pH and adjust with monobasic or dibasic solutions as required. Store at room temperature for 4 weeks.
2. Bovine serum albumin solution: 1% BSA in phosphate-buffered saline, pH 7.4 (1% BSA/PBS). Weigh out 1 g of BSA and dissolve by gentle continuous mixing in 100ml of 0.1M phosphate buffer, pH 7.4. Keep at 4°C for 4 weeks.
3. Cells expressing the appropriate antigen. Harvest at least 15 × 10⁶ cells from a sufficient number of tissue culture flasks, using trypsin/EDTA or manual scraping to detach adherent cell lines. Wash the cells and resuspend them in cold 1% BSA/PBS. If required, break up small clumps of cells by repeatedly drawing them up and down in a pipette or by syringing them through a 23-gauge needle until a single-cell suspension is obtained. Measure the cell concentration with a haemocytometer or automated cell counter and dilute to a final concentration of 4 × 10⁶ml in cold 1% BSA/PBS.
4. Radiolabelled antibody at a concentration of 50ng/ml in cold 1% BSA/PBS. Freshly prepare at least 4ml.
5. Unlabelled antibody: 400 µg at a concentration of at least 0.5 mg/ml.

Steps

1. Place duplicate rows of 7 × 1.5-ml microcentrifuge tubes in a test tube rack. With a marker pen, label the tubes 1-7 and 8-14. Pipette 0.5 ml of cold 1% BSA/PBS into tubes 2-5 and 9-12.
2. Pipette 0.5 ml of the cell suspension into tubes 1, 2, 6, 8, 9, and 13.
3. Pipette 200 µg of unlabelled antibody into tubes 6 and 13. Mix well.
4. Briefly vortex tube 2 and transfer 0.5 ml of the contents to tube 3. Vortex tube 3 and transfer 0.5 ml to tube 4. Vortex tube 4 and transfer 0.5ml to tube 5. Vortex tube 5 and discard 0.5 ml.
5. Repeat step 4 with tubes 9-12.
6. Pipette 250µl of radiolabelled antibody into all tubes. Mix well.
7. Incubate the tubes for 2h at a constant temperature (preferably 4°C but alternatively at room temperature or 37°C). Mix the tubes either constantly using a mechanical shaker or roller or about every 15 min by hand during the incubation.
8. After the incubation, leave the tubes on ice for 10 min to cool. Centrifuge tubes 1-6 and 8-13 at high speed (e.g., >1000rpm) for 2min. Carefully remove and discard the supernatant, taking care not to disturb the cell pellet. Pipette 0.5 ml of cold 1% BSA in PBS into each tube and quickly vortex each tube to resuspend the cell pellet. Recentrifuge and again discard the supernatant.
9. Count all tubes in a gamma counter on the appropriate isotope setting.
10. Analyse data as follows: Divide the counts in tubes 1-5 and 8-12 (bound counts) by an average of the counts in tubes 7 and 14 (total counts) in order to calculate the fraction of counts bound to the cells for each cell concentration. Average the values for duplicate tubes (1 and 8, 2 and 9, 3 and 10, 4 and 11, and 5 and 12).
11. Calculate the cell concentration in each tube (approximately 40, 20, 10, 5, and 2.5 × 10⁵ ml).
12. Plot as y values the reciprocal of fraction bound (i.e., total/bound) against the reciprocal of cell concentrations as x values. A straight line plot should be obtained. Determine the intercept on the y axis and calculate the reciprocal. This is the immunoreactive fraction.
13. Calculate the contribution made by nonspecific binding (i.e., binding in the presence of a large excess of unlabelled antibody) by dividing the average of the counts in tubes 6 and 13 by the average of the counts in tubes 7 and 14. This value is generally very low and the nonspecific binding contribution can therefore reasonably be excluded from the calculation of the immunoreactive fraction. If the nonspecific binding is greater that 5% of the total binding, then the causes should be identified and eliminated.

IV. COMMENTS
Many techniques have been developed in the last 50 years for labelling proteins with radioiodine but, for various reasons, most of these are now only of academic interest, at least so far as the routine labelling of antibodies is concerned. For the interested reader, a detailed review has been put together by Dewanjee (1992). The methods practised most widely are those in which the radioiodine is oxidised to a reactive intermediate positively charged species such as the hydrated iodinium ion H₂OI⁺, which then reacts via electrophilic substitution for the activated protons on the phenolic ring of the tyrosine side chains. These iodinium species or related cations can be produced by reacting the radioiodide with a variety of oxidising agents, but two in particular have become the most popular: chloramine-T (N-chloro-p-toluenesulphonamide) and Iodogen (diphenylglycoluril). Methods for labelling antibodies with radioiodine using these two methods were described in Sections III,A and III,B. Both techniques have their own inherent advantages and drawbacks. The Iodogen method (Fraker and Speck, 1978) is extremely simple and largely invariable. The concentration of the oxidant is determined by its very poor solubility in aqueous solvents and therefore the only variables one can change in order to influence the reaction are incubation time and temperature. Even so, this method normally produces very acceptable labelling efficiencies of the order of 80-95%. In contrast, details of the chloramine-T method (Hunter and Greenwood, 1962) vary widely from laboratory to laboratory. Different researchers have their own favoured oxidant concentrations, incubation times, and choice of quenching reagents. This method has the (somewhat theoretical) advantage that it can be tailored for different proteins, but has the disadvantage that the reagents have to be freshly prepared prior to use and overenthusiastic attempts to improve labelling efficiencies with high concentrations of the oxidant can lead to antibody damage.

Only in rare circumstances will either labelling procedure result in labelling efficiencies approaching 100%. The consequence of this is that the reaction mixture will contain a significant proportion of unreacted "free iodide." It is normally desirable to remove this free iodine in order to provide a preparation with sufficiently high radiochemical purity. The most widely used method for the purification of labelled antibody preparations is size-exclusion chromatography on a short Sephadex column as described in Section III,D, but several alternatives, such as ionexchange chromatography or the use of spin columns, exist, some of which are potentially more convenient. Before and after purification of the labelled antibody it is useful to measure the purity of the labelled antibody preparation, initially to check the efficiency of the labelling procedure and then later to determine the purity of your reagent. A very simple means of measuring this purity, based on ascending thin-layer chromatography, is described in Section III,E.

The main reason for labelling an antibody is to obtain a radioactive molecule that binds to a specific recognition site. It is therefore essential that the antibody retains the ability to bind to its epitope throughout the labelling procedure. The best tests to ensure that this is the case are radioligand binding assays, which measure directly the binding of the radiolabelled molecules rather than the whole population of antibody molecules in solution, most of which will not be labelled. These assays fall into two categories: those used to determine the binding affinity of the antibody and those intended to measure the proportion of labelled molecules that retain some ability to bind specifically to their epitope. For the purposes of a relatively simple check to see if the antibody remains functional after labelling, the latter type of assay is the more appropriate and a protocol describing such a test can be found in Section III,F.

Two main types of mechanism can compromise the immunoreactivity of a labelled antibody. The first is due to the effect of the steric hindrance of the large iodine atom when it is substituted into a critical tyrosine residue close to the binding site of the antibody. Although the radioiodine can potentially react with any of several tyrosine amino acids scattered throughout the antibody molecule, factors such as local charge distribution and accessibility mean that one or more residues will be labelled preferentially. If one of these sites happens to be in one of the critical CDRs, then a significant degree of antibody binding will be lost. If this happens, the number of possibilities for solving the problem are limited. It is possible that a change in pH during the labelling procedure may alter the local charge distribution and favour an alternative tyrosine residue, albeit at the risk of a lower labelling efficiency. If this does not work, then the only alternative is to use an entirely different chemistry for radioiodination. The most well-established alternative is the Bolton and Hunter method described in Section III,C, which results in antibody labelling at the site of lysine residues.

The other type of mechanism responsible for loss of immunoreactivity is oxidation. In addition to its desired role in oxidising the radioiodide to a reactive species, the oxidant may potentially oxidise critical residues, particularly methionine, in the antibody molecule. A way to find out which of the two possible mechanisms may in fact be the cause of a loss in immunoreactivity is to perform the labelling procedure without the addition of the radioiodine and to perform an ELISA assay. If an oxidative mechanism is responsible, then immunoreactivity will still be lost, as all the antibody molecules will be affected, not only those substituted with radioiodine. If this is found to be the case then either a Bolton and Hunter approach can be pursued or an alternative electrophilic substitution method that does not subject the antibody to such strong oxidising conditions can be employed. Two approaches may work. The first is to use a milder oxidising agent, such as the lactoperoxidase system (Morrison and Bayse, 1970). The alternative is to use a modification of the Iodogen system in which the radioiodine is first oxidised in the iodogen tube but is then transferred from the oxidising environment to another tube containing the antibody (van der Laken et al., 1997). It is likely that both of these procedures will result in a lower labelling efficiency but either may solve the problem of oxidative damage to the antibody.

The shelf life of radioiodinated antibodies is limited by radiolysis, which causes a gradual loss in both purity and immunoreactivity. The rate of deterioration can be reduced by the addition of carrier proteins or antioxidants that scavenge the radiolytic free radicals (Chakrabarti et al., 1996). A concentration of 0.1-1% albumin or 0.5% ascorbic acid is commonly used and antibodies may be stored in these solutions at either 4 or -20°C for at least a month without a significant loss of quality. If stored below 0° then the preparation should be divided into aliquots to save repeated freezing and thawing, which tends to favour aggregation of the antibody. If stored above 0° provided it does not interfere with the ultimate application, sodium azide can be added to a final concentration of 0.05% to limit microbial growth.

V. PITFALLS
The most likely cause of failure of any of the labelling methods described here is the presence of impurities in the antibody solution. The best solution is to repurify the antibody by either dialysis or gel filtration into freshly prepared buffers.

The most common problem experienced with the immunoreactive fraction assay described in Section III,F is that a curve, rather than a straight line, is obtained when data are plotted. This is nearly always caused by inaccuracies in diluting and losses in washing the cells. With practice and care, the problem usually goes away.

References

Bolton, A. E., and Hunter, W. M. (1973). The labelling of proteins to high specific activities by conjugation to a 125-I-containing acylating agent. Biochem. J. 133, 529-538.

Chakrabarti, M. C., Le, N., Paik, C. H., De Graft, W. G., and Carrasquillo, J. A. (1996). Prevention of radiolysis of monoclonal antibody during labeling. J. Nuclear Med. 37(8), 1384-1388.

Dewanjee, M. K. (1992). Radioiodination: Theory, Practice and Biomedical Applications. Kluwer Academic, Dordrecht.

Fraker, P. J., and Speck, J. C. (1978). Protein and cell membrane iodinations with a sparingly soluble chloramide 1,3,4,6-tetrachloro 3a.6a diphenylglycoluril. Biochem. Biophys. Res. Commun. 80, 849.

Hunter, W. M., and Greenwood, E C. (1962). Preparation of iodine- 131 labelled human growth hormone of high specific activity. Nature 194, 495-496.

Lindmo, T., Boven, E., and Cuttita, E (1984). Determination of the immunoreactive fraction of radiolabelled monoclonal antibody by linear extrapolation to binding at infinite antigen excess. J. Immunol Methods 27, 77-89.

Mather, S. (2000). Radiolabelling of monoclonal antibodies. In "Monoclonal Antibodies, a Practical Approach" (P. Shepherd and C. Dean, eds.), pp. 207-236. Oxford Univ. Press, Oxford.

Morrison, M., and Bayse, G. S. (1970). Catalysis of iodination by lactoperoxidase. Biochemistry 9, 2995-3000.

van der Laken, C. J., Boerman, O. C., Oyen, W. J., van de Ven, M. T., Chizzonite, R., Corstens, F. H., and van der Meer, J. (1997). Preferential localization of systemically administered radiolabeled interleukinl alpha in experimental inflammation in mice by binding to the type II receptor. J. Clin. Invest. 100(12), 2970-2976.