High Specific Activity Labeling of Insulin with 1311*

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 239, No., November 96 Printed in U.S.A. High Specific Activity Labeling of Insulin with 3* JOSEPH L. Izzo, WILLIAM F. BALE, MARY JANE Izzo, AND ANGELA RONCONE From the Department of Medicine and the Department of Radiation Biology, University of Rochester School of Medicine and Dentistry, Rochester, New York 620 (Received for publication, March 9, 96) The immunoassay of endogenous insulin in plasma or other body fluids by means of radiochromatoelectrophoretic techniques which have been developed by Berson and Yalow requires as a or immunochemical properties investigated. On the basis of these observations it is clear that the average incorporation of total iodine should not exceed atom per molecule of insulin if tracer %I-insulin of high specific activity. However, the meth- the labeled insulin is to be used for physiological studies, whereas ods which Berson, Yalow et al. (, 2), as well as others (3), have employed to prepare highly radioiodinated insulin have usually resulted in low and unpredictable yields of labeled insulin. Furthermore, if large amounts of 3 are used, damage to the insulin the same restrictions are not necessary if the labeled insulin is to be used solely for immunoassay purposes. With these considerations in mind, we present a procedure for incorporating 3 into insulin in the order of 00 to 00 mc of occurs which is roughly proportional to the amount of 3 Is per mg of insulin, with good yield of product in terms of attached to the hormone. It is reported that the insulin which is damaged by the effects of 3 radiation does not bind insulin starting 3 and with little or no damage to the insulin by radiation as determined by chromatoelectrophoresis. By precise antibody (2) and that purification procedures are usually neces- control of the total amount of iodine attached to the insulin sary to remove the damaged components whenever they are present in appreciable quantities. Hence, a procedure for high level radioiodination of insulin which is efficient in the use of (i.e. stable as well as radioactive iodine), highly radioiodinated insulin can be prepared for use in immunoassay procedures or for use in physiological studies. The method is based on, although starting 3 and which does not require repurification of the not identical with, a procedure developed by Helmkamp et al. labeled hormone should be useful. %insulin of very high specific radioactivity should also be (0) for protein labeling, which, in turn, is based on the use of iodine monochloride first suggested by McFarland (). potentially very useful for study of the metabolism and action of insulin. The studies that have been reported thus far on the EXPERIMENTAL PROCEDURE disposal and metabolic fate of injected 3-insulin have been restricted to the use of radioiodinated insulin of very low specific activity, with the result that the total amount of injected insulin has been substantial compared to the amount of endogenous hormone generally present in the circulation (,-8). Although these studies have provided useful information on the metabolic fate of injected insulin, they do not necessarily reflect the disposal of endogenous insulin at physiological levels. Hence, a suitable preparation of highly radioiodinated insulin that would permit the use of truly tracer quantities of injected insulin could be a useful tool for studying the metabolism and action of endogenous insulin at physiological levels. Obviously, one of the prime req- uisites for such a radioiodinated insulin is that it possess full biological activity. In the accompanying paper (9)) we have reported that the total amount of iodine (stable as well as radioactive) attached to insulin differentially affects the biological, electrophoretic, and immunochemical properties of insulin. The incorporation of an average of more than atom of iodine per molecule of insulin of an assumed molecular weight of 6000 resulted in a progressive loss of biological activity which was roughly proportional to the number of iodine atoms incorporated. On the other hand, increasing iodination of insulin had little, if any, effect on the electrophoretic * This investigation was supported by Research Grant A-3556 from the United States Public Health Service, by a grant from Eli Lilly and Company, and by the United States Atomic Energy Commission. Presented in part before the Federation of American Societies for Experimental Biology, Atlantic City, April, Xolutions-Samples of carrier-free Na?I solutions produced by fission of uranium were obtained by air shipment from Oak Ridge National Laboratory soon after production to mini- mize relative increase in the content of 27 and 29 which are present in the original 3 sample. Activity at the time of shipment ranged from 20 to 50 mc per ml of solution. Borate Bu$ers-Borate buffer of ph 7.65, designated 2X borate buffer (lo), was prepared by adjusting a distilled water solution of 0.32 M NaCl and 0.0 M H3B02 to ph 7.65 by addition of 3.2 M NaOH to a final concentration of approximately 0.08 M NaOH. This borate buffer was diluted with an equal volume of distilled water to produce a second buffer of ph 8.0, designated X borate buffer. Insulin-Bovine zinc insulin crystals, Lot 53566, assayed by Eli Lilly and Company at 25.6 units per mg, were generously supplied by Dr. William R. Kirtley of the Lilly Research Laboratories. Prior to each iodination, a fresh solution of insulin was prepared as follows. Dried insulin cryst,als (2 mg) were dissolved in ml of 0.0 N HCl, and ml of 2X borate buffer, was then added. For iodination of pg of insulin, 0.5 ml of the above solution was diluted with 0.85 mlof x borate buffer. Iodine Monochloride Reagent-As described more fully elsewhere (2), the iodine monochloride reagent was prepared according to the reaction 2 HI + KIOa + 6 HCl 3 3 ICI + 3 KC + 3 Hz0 A convenient stock solution is one that is 0.02 M in ICl, 2.0 M in KCl,.0 M in HCI, and 2.0 M in NaCl and was prepared as

2 High Specific Activity Labeling of Insulin with 3 Vol. 239, No. follows. To a solution of g of KCl, g of KIO, and g of NaCl were added 2 ml of concentrated hydrochloric acid (sp. gr..8) and enough distilled water to bring the final volume of the total mixture to 250 ml. The slight amount of free iodine which formed was removed by repeatedly extracting this solution with carbon tetrachloride. A current of air, saturated with water vapor, was then passed through the ICl solution to volatilize any suspended or dissolved carbon tetrachloride. Such a solution has a molarity within % of that calculated on the basis of the amount of KIO, that was used, and is stable indefinitely at room temperature. Just before use, a dilution of the stock solution of ICI in 2 M NaCl was prepared so that 0.2 ml of this diluted stock solution contained the appropriate number of molecules of ICl for each insulin molecule of molecular weight 6000 that was to be iodinated. Procedure for High Level I3 IO&nation-Depending upon the level of specific activity that was desired, from 50 to 00 mc of Oak Ridge 3 in a volume of solution not exceeding 5 ml were transferred to a chemically clean test tube (A, 2 x 5 cm). To this solution was added.5 ml of 2~ borate buffer (2 ml if the volume exceeded ml). If the ph was not approximately 8 by wide range indicator paper, the solution was adjusted to this ph by the addition of a few drops of N HCl. Hydrogen peroxide, which is produced by /3 radiation from 3, was destroyed by the addition of 0.2 ml of freshly prepared 0.5 M Na2S03. If it is not removed, hydrogen peroxide can prevent iodination of insulin by reducing the iodine of ICl to iodide. The excess sulfite in the contents of tube A was, in turn, oxidized by aerating the unstoppered tube in a shielded rotary Evapo Mix2 with a swirling action at 70 for 2 hours. After cooling to room temperature and measurement of the total radioactivity, the contents of tube A were transferred by jet to tube B, which contained ml of a solution of the insulin to be iodinated.3 In these studies we used to 250 pg of insulin. Immediately, 0.2 ml of the freshly diluted solution of stock ICI was added to.8 ml of 0.85% NaCl solution, and the mixture in turn was added by jet to tube B and thus rapidly mixed with the 3rI-insulin-mI-iodide mixture. Two reactions occur. The Ia which is present as iodide exchanges with the iodine monochloride to form mic. If the 3 is present initially in a total amount of iodine that is small compared with the iodine in the ICl, the exchange will result essentially in a quantitative conver- If the radioiodinated insulin is to be used solely as a labeled tracer for immunological studies, molecules of ICI are used for every insulin molecule of molecular weight However, if the radioiodinated insulin is to be used for studies on the metabolism and action of insulin, used for every insulin not more than molecule of ICl should molecule of molecular weight be 2 Buchler I*nstruments, New York, New York- 3 Tube B is fitted with a two-hole rubber stopper. U-shaped capillary glass tubing with an internal diameter of mm is used to connect tubes A and B. One end of the tubing, which is slightly constricted at its tip, is inserted for a distance of several centimeters into tube B through one of the openings in the rubber stopper while the other end of the tubing is placed in unstoppered tube A so that its end just clears the bottom of the tube. By means of glass tubing (6 mm internal diameter) inserted through the other opening of the rubber stopper, tube B is connected to a glass stopcock which, in turn, is connected to a suction flask. By opening and closing the stopcock, solutions in tube A can be rapidly transferred (jetted) and mixed with the contents of tube B. These operations are conducted behind a lead shield by remote handling appliances. sion of the 3 to the iodine monochloride form with great rapidity. The 3C then reacts with the tyrosine residues of insulin to attach 3 by stable covalent bonds. This reaction is also complete in a short time, probably in less than minute. After 3 minutes, ml of 6.25% human serum albumin in 0.85% NaCl solution was added to serve as a competitive sink or site of reaction for free radicals and peroxides, and thus acted as a protection against radiation damage to the insulin. The reaction mixture was then transferred to a cellophane dialysis sac* containing an additional 8 ml of 6.25% human serum albumin. The reaction mixture was dialyzed overnight at 3 against two successive 2-liter portions of 0.850/, NaCl, each portion being adjusted to ph 3.0 with 6 ml of glacial acetic acid. At the end of the dialysis, the radioactivity in the sac containing 3 was assayed. The contents of the sac were then transferred to a test tube containing 25 mg of 6.25% human serum albumin and centrifuged for 5 minutes at 0,000 r.p.m. to remove a very small precipitate of what is probably denatured albumin (with no appreciable loss of 3). The supernatant solution was then removed, assayed for radioactivity, and divided into several portions; the frozen material was stored immediately at -8 until used. Such solutions may be kept in the frozen state for several weeks without damage to the insulin by radiation. Radioactivity measurements were carried out by placing samples to be measured at a distance of approximately 55 cm in a lateral direction from a well shielded well-type sodium iodide scintillation counter, usually with.25 or.5 inches of lead filtration to reduce the y-ray sensitivity of the measuring apparatus to a usable level. Calculations of?i content of samples were made in terms of the initial la activity corrected for radioactive decay. A 3 standard measured under the same conditions was used to check counter stability. Care was taken to place all samples in containers of approximately the same diameter to make self-absorption corrections unnecessary. Preliminary studies showed that after dialysis of 3-labeled insulin, 3 activity was quantitatively carried down in a trichloroacetic acid precipitate. Small corrections were made for losses of 3-insulin which occurred while the material was being transferred from one container to another. This was of the order of 3% for loss on glassware and an additional 2 to 3% loss upon centrifugation after dialysis. This correction was applied to the calculation of the specific activity, which is expressed as millicuries of r3ri per mg of insulin. The I3 attached to insulin is calculated as 3rI millicuries used, corrected for decay, multiplied by the ratio of the counts obtained in the final preparation of 3-insulin after centrifugation to the counts of 3 after aeration, corrected for decay. Detection of Radiation Damage to 3-InsuZin-Damage to each radioiodinated insulin preparation was determined by means of radiochromatoelectrophoresis according to a modification of the technique of Berson and Yalow (, 2). 3-Insulin (0. ml) was diluted to 5 ml with Verona buffer containing 0.25% human serum albumin. Of this diluted mixture, containing 8.9 pg of %I-insulin, 50 ~ were added to each of two test tubes containing either 0.5 ml of 0.25% human serum albumin in Verona buffer of ph 8.6 and ionic strength 0. (Solution A) or 0.5 ml of human plasma (Solution B). Samples of Solution A (0 ~) were then applied to each of two sets of duplicate strips of Whatman No. 3MM paper along a line cm from the cathode end of the cell. * Available from the Visking Corporation, New York.

3 November 96 J. L. Izxo, W. F. Bale, M. J. Ixxo, and A. Roncone 375 TABLE I Results of high level radioiodination of insulin with different ratios of starting 3I-Insulin and iodine monochloride (ICI) Preparation I Moles of ICl la recoveries 3 recovered Estimated average per mole of Insulin used 3 usedt attached to attached to insulin Specific activity insulin* insulin at completion of of insulin.adioiodinatior No. of atoms of 3 per molecule of experiment insulin*t - Pg mc % mclmg % Al A A Bl B2 B3 B B5 B Cl c2 c3 c c5 C * Molecular weight = t The millicuries of 3 were always corrected for decay up to the day on which the experiment was completed. $ Estimations were calculated on the basis that 00 mc of 3 contains 0.8 pg of lzii. One set of strips previously had been saturated with native insulin by immersing the strips for hour in a 0.3% solution of unlabeled insulin, and then removing the strips and rinsing them three times with Verona buffer to remove the excess insulin. The other set of strips was previously soaked in Verona buffer prior to electrophoresis. A similar procedure was carried out for Solution B. The cell containing the four sets of duplicate strips was sealed except for the opening along the top.5 A constant voltage (300 volts) was applied, and the electrophoresis was conducted for hour at room temperature in Verona buffer (ph 8.6 and ionic strength 0.). The slit at the top of the cell was then sealed, and the electrophoresis was continued for 6 hours at 00 volts. The paper strips were dried at 20 for 30 minutes. One strip from each of the four sets of duplicate strips was stained with Spinco B-l dye (bromphenol blue and hydrated zinc sulfate), while the other was not stained but was cut into 0.5-cm widths, and the radioactivity of each width was measured either in a gas flow counter (Nuclear-Chicago) or in a liquid scintillation system (Nuclear-Chicago), model 703. RESULTS AND DISCUSSION Percentage of m Attached to Insulin-Table I summarizes the results of three series of radioiodinations of insulin with different ratios of 3, insulin, and ICl. Note that the amount of Is incorporated was determined not only by the initial ratio of 3 to insulin, but also by the amount of ICl used. The highest radioiodinating efficiencies, i.e. 60 to 70% of starting 3 firmly attached to protein, were achieved with 50-mc lots of 3, 250 pg of insulin, and atom equivalents of ICl. Specific radioactivities of 00 to 25 mc of 3 per mg of insulin were obtained 5 If the slit at the top of the cell cover is left open, hydrodynamic flow occurs. The resulting evaporation provides the flow necessary to move globulins away from the origin. under these conditions (Series A). Higher specific activities (237 to 09 me of 3 per mg of insulin) were achieved with only a slight loss in iodinating efficiencies (0 to 50%) by the use of loomc lots of 3, pg of insulin, and atom equivalents of ICl (Series B). In these preparations an average of 7 mc of 3 attached to insulin was recovered. On the other hand, iodinations of vg of insulin with loo-mc lots of 3 and only atom equivalent of ICl resulted in specific activities in the range of 88 to 2 mc of I3 I per mg of insulin with efficiencies of 0 to 23 % in terms of starting 3 (Series C). An average of mc of 3 attached to insulin was recovered in Series C. In terms of insulin used, the yields of radioiodinated insulin in Series A and B are higher than those that have been reported in other published procedures (2, 3). Studies are currently in progress to improve efficiencies of iodinations of the Series C type of preparation. An additional and very important factor that limits the amount of radioactive iodine that can be attached to insulin or other protein is the presence of appreciable amounts of 27 and l29 in the so-called carrier-free 3 preparations. As was described more fully elsewhere (2), the production of 3 with a half-life of 8.05 days either by the fission of uranium or by the bombardment of tellurium with thermal neutrons also results in the production of stable 27 and a long lived radioactive isotope, lz9 (half-life,.6 x 07 years). Stable 27 may also be present as an impurity in the irradiated material or in reagents that are used in the extraction and purification of 3. Hence, as presently manufactured, so-called carrier-free 3 is never in a strict sense free of stable or carrier iodine. Furthermore, 3 decays on storage, while the other two isotopes do not, with the result that the ratio of stable to radioactive iodine increases progressively and can reach very substantial amounts. With the ICl chemical exchange method of preparing high level radioiodinated insulin, the presence of increased quantities of I27 and 2gI interferes with the quantitative conversion of 3 to 3ICl and results in a reduc-

4 376 High Specific Activity Labeling of Insulin with I3 Vol. 239, No. tion in the efficiency of the coupling of 3 to protein. Difficulty of this kind has been reduced to a minimum in the present studies by the use of freshly processed 3. Percentage of Total Iodine Attached to Insulin-Although the percentage of 3 coupled to insulin is influenced not only by the ratio of starting %I to insulin, but also by the ratio of ICl to insulin and by the ratios of 2 and 29 to KI, the percentage of total iodine attached to insulin is largely a function of the ratio of ICl to insulin. On the basis of data we have obtained from iodination of pg of insulin with and atom equivalents of ICl and trace quantities of 3, it is estimated that the Series B preparations contained an average of approximately 3.5 atoms of total iodine per molecule of insulin. On the other hand, the total iodine content of the Series C preparations did not exceed an average of approximately 0.65 atom of total iodine per molecule of insulin (mol. wt. 6000). Approximately 3 to y0 of the total iodine incorporated into insulin is 3 in the Series B preparations or, in other words, about out of 8 insulin molecules is labeled with 3. In the Series C preparations, approximately out of 28 insulin molecules contains 3. Since the total amount of iodine incorporated into insulin, as can be seen from the foregoing discussion, is a function of the ratio of ICl to insulin used and cannot exceed the iodine content of the added ICl, it is apparent that with the iodine monochloride method the presence of unknown amounts of nonradioactive iodine isotopes in a given sample of 3 would influence merely the percentage of 3 coupled to protein, the total amount of iodine incorporated into insulin remaining unaltered. This constitutes a built in safety device to prevent overiodination of insulin. On the other hand, with iodination methods which depend upon the conversion of iodide to free iodine (2), the total as well as radioactive iodine incorporated into insulin would be influenced by the presence of unknown amounts of nona dioactive iodide in the 3 sample. Hence, at high 3 levels, overiodination of insulin can easily occur with oxidation methods unless the total iodine content of a given 3 sample is accurately known. This factor assumes especial importance in the preparation of 3-insulins of high specific activity which are intended for physiological use, because the total average iodine incorporation of such preparations should not exceed atom per molecule of insulin (9). Damage to 3-Insulin Produced by Radiation-None of the 3-insulin preparations listed in Table I exhibited damage by radiation in excess of 5 y0 as determined by chromatoelectrophoresis in the presence of added human serum albumin or normal plasma. In several preparations no damage at all could be detected, even though the specific activity was of the order of 300 or more mc of 3 per mg of insulin. This is illustrated by the chromatoelectrophoretic behavior of Preparation B 5 with added normal plasma which had an initial specific activity of 306 mc of 3 per mg of insulin (Fig. ). According to Berson and Yalow (, 2), undamaged free insulin is adsorbed at site of application to the paper strip and remains at the origin when an electrical voltage is applied, whereas the damaged components are said to migrate away from the origin with the serum proteins. On this basis the preparation was essentially undamaged, since virtually all of the radioactivity was detected at the origin and no perceptible increase in radioactivity could be detected in the region of the serum proteins. Furthermore, on paper that was presaturated with stable insulin, the radioactivity of the same preparation moved as one peak just behind the serum albumin fraction. The movement of the radioactivity of these 3-insulin preparations as one peak (Fig. 2) suggests that the preparations were probably homogeneous. This, however, does not preclude the presence of altered insulin molecules having an electrophoretic mobility similar to those of the unaltered molecule. Nevertheless, since the migration of the radioactivity of these 3-insulin 6000 k 5000 PREPARATION B-5 NON-SATURATED PAPER -UNDAMAGED FREE INSULIN LABELED WITH i PREPARATION B-5 SATURATED PAPER 2 s $ ? I IO II I2 I3 5 Cms Cms ORtGlN ALGIN FIG. (Zeft). Distribution of radioactivity of 3-insulin Preparation B 5 (specific activity, 306 mc of Ia per mg of insulin) by the naner chromatoelectronhoretic techniaue of Yalow and Berson (2) on-untreated paper. According to their criteria, undamaged free 2 ORkIN &&IN insulin remains at the origin, while damaged insulin migrates with the protein components of admixed nonimmune plasma. FIG. 2 (right). Distribution of radioactivity of 3~insulin Preparation B 5 (specific activity, 306 mc of 3 per mg of insulin) on paper presaturated with nonlabeled insulin.

5 November 96 J. L. Izzo, W. F. Bale, M. J. Izzo, and A. Roncone 377 preparat,ions of high specific activity was similar to the migration patterns of radioactivity obtained by Berson et al. () and in our laboratory (9) with the use of insulin from the same manufacturer and lot and labeled with only trace quantities of r3, it is reasonable to assume that our radioiodinating procedure did not produce gross changes in the insulin molecule that would affect its electrophoretic mobility. We believe that the early addition of human serum albumin to the radioiodinated insulin is largely responsible for preventing the degree of radiation damage that has been encountered by others (2, 3) in high level iodination of insulin with 3rI. This early addit ion of protective protein is possible because of the rapidit.y with which 3C reacts with the tyrosine residues of the insulin molecule to attach 3 (0). Since the 3C that has not. coupled with protein is promptly converted to iodide, serum albumin can be added within a few minutes of iodination of insulin with no appreciable labeling of albumin detectable within the limits of accuracy and sensitivity of our methods. As shown in Fig. 2, no discrete peak of radioactivity is detectable in the albumin region of the 6-hour chromatoelectrophoretogram of Preparation B 5, suggesting that perceptible labeling of albumin with 3 did not occur under the conditions employed. This possibility was further investigated as follows. Two *3- insulin preparations were made with 0.25 mg of insulin,.5 mc of 3, and atom equivalent of ICl. To one of the preparations 32 atom equivalents of Na$03 were added just prior to the addition of the protective albumin. This was done to reduce any remaining ionic I3 to iodide. The percentage of iodination of both preparations was similar, 66. and 65.$& respectively. Low level rather than high level radioiodination was carried out to eliminate or minimize the variable presence of small amounts of damaged components less than (5%) which have been observed to migrate with the serum proteins in some high level preparations and which, if present, might interfere with evaluation of the results. Chromatoelectrophoresis of both preparations with hydrodynamic flow was carried out for 2 hours instead of 6 hours, with 0.25 y0 human serum albumin instead of normal plasma in order to achieve maximal sensitivity. Both preparations gave identical results. No radioactivity was perceptible in the albumin region (Fig. 3). High Level Radioiodination of Other Proteins--Preliminary studies indicate that with slight modifications in the method presented, high level radioiodination of glucagon can also be achieved. In fact, the relative simplicity of the method, coupled with its flexibility in permitting variable but controlled total as well as radioactive iodination of a protein of known composition, should make the procedure well suited for the labeling of other protein hormones of known chemical structure. Greenwood, Hunter, and Glover (3) have recently reported a method for preparing small quantities of highly radioactive growth hormone based on 3 labeling with r3 which has been oxidized from the initial iodide form with chloramine-t. They maintain that the degree of chemical substitution is minimized (0.5 to.0 atom of iodine per molecule of protein) by the use of carrier-free (m) iodide. For reaspns that we have outlined earlier, the 27 and I29 content of these starting preparations must be far higher than is consistent with the assumption of Greenwood et al. that 90% of the total iodine content of their starting preparation is 3. Therefore, the number of iodine atoms attached to each growth hormone molecule must be greater than their calculations indicate. Bale et al. (2) have shown by cal- -UNDAMAGED FREE INSULIN LABELED WITH 3 0 I Cms. Ii c I OF&IN GiiGlN FIG. 3. Chromatoelectrophoresis of 3-insulin of low specific activity in 0.25% human serum albumin and Verona buffer, ph 8.6, ionic strength 0., for 2f hours. Note the absence of perceptible radioactivity in the albumin region, indicating absence of any appreciable radioiodoalbumin in the preparation. culations that are based on the physics of 3 production by various methods, and also by chemical total iodine determinations of highly iodinated 3 preparations, that all presently available 3 preparations must contain much more total nonradioactive iodine than assumed by Greenwood et al. (3). In a short communication, Banerjee and Gibson () have recently reported the production of radioiodinated insulin with a specific activity of curie per mg with the method of Greenwood et al. (3). However, considerable radiation damage was observed, and repurification procedures were necessary to render the preparations sufficiently pure for use as tracers in immunoassay procedures of insulin. SUMMARY. A procedure is described for preparing r3-labeled insulin of high specific activity with efficient use of 3 and with less than 5% radiation damage to insulin as measured by chromatoelectrophoresis. 2. %I produced by fission of uranium is used soon after production to keep at a minimum the relative concentrations of 2rI and 29. After destruction of the hydrogen peroxide which is invariably present in 3 samples, solutions of I3 as iodide and of nonradiaoctive iodine monochloride (ICI) are successively added to solutions of insulin in borate buffer at ph 8.0. Chemical exchange produces 3C, which, in turn, reacts with insulin to label it. Radiation damage is maintained at a minimum by prompt dilution with a solution of human serum albumin. 3 that is not bound to insulin is removed by dialysis. 3. Iodination of pg of insulin with loo-mc lots of 3 and molecules of ICI per molecule of insulin resulted in specific activities of 237 to 09 mc of 3 per mg of insulin. An average of 7 mc of 3rI was recovered firmly attached to insulin. The total iodine (radioactive plus nonradioactive) incorporated into insulin was estimated at 3.5 atoms of iodine per molecule of insulin with an assumed molecular weight of Iodination of Kg of insulin with loo-mc lots of I3 and only molecule of ICl per molecule of insulin resulted in specific activities of 88 to 2 mc of 3 per mg of insulin. An average of mg of 3 was recovered attached to insulin. The total

6 378 High SpeciJic Activity Labeling of Insulin with 3 Vol. 239, No. iodine coupled to insulin did not exceed an average of 0.65 atom of iodine per molecule of insulin. 5. For 3-insulin preparations that are to be used as tracers in physiological studies, the average incorporation of total iodine per molecule of insulin (mol. wt. 6000) should not exceed atom. Such restriction is not required if the labeled insulin is to be used for immunoassay purposes only. REFERENCES. BERSON, S. A., YALOW, R. S., BAUMAN, A., ROTHSCHILD, M., AND NEWERLY, K., J. Clin. Invest., 36, 70 (956). 2. YALOW, R. S., AND BERSON, S. A., J. Clin. Invest,, 39, 57 (960). 3. SAMOLS, E., AND WILLIAMS, H. S., Nature, 90, 2 (96).. ELGEE, N. J., WILLIAMS, R. H., AND LEE, N. D., J. Clin. Invest., 33, 252 (95) LEE, N. D., Endocrinology, 65, 37 (959). WICK, A. N., AND DRURY, D. R., Proc. Sot. Exptl. Biol. Med., 97, 5 (958). STEIN, O., AND GROSS, J., Endocrinology, 66, 707 (959). PROUT, T. E., AND EVANS, I. E., Ann. N. Y. Acad. Sci., 7, 530 (959)..~~0, J. L., RONCONE, A., Izzo, M. J., AND BALE, W. F., J. Biol. Chem., 239, 379 (96). HELMHAMP, R. W., GOODLAND, R. L., BALE, W. F., SPAR, I. L., AND MUTSCHLER, L. E., Cancer Research, 20, 95 (960). MCFARLAND, A. S., Nature, 83, 53 (958). BALE, W. F., HELMKAMP, R. W., DAVIS, T. P., Izzo, M. J., GOODLAND. R. L., AND SPAR, I. L., U. S. Atomic Energy Comm., U&assi&d Repts., 60 (96i). GREENWOOD. F. C.. HUNTER. W. M.. AND GLOVER, J. S., Biochem. j., 89, li (963). BANERJEE, R. N., AND GIBSON, K., J. Endocrinol., 26, 5 (962).