ORIGINAL ARTICLE Annals of Nuclear Medicine Vol. 1O No 1, 49-55, 1996 Applicability of short-lived radiometallic nuclide for high sensitivity two-site "sandwich" immunoradiometric assay: Human growth hormone assay Kazuko HORIUCHI,* Lin H. LIN,** Yasuhisa FUJIBAYASHI,* Vania C. BORGHI** and Akira YOKOYAMA* *Department of Radiopharmaceutical Chemistry, Faculty of Pharmaceutical Sciences, Kyoto University, Japan **Institute of Energy and Nuclear Research, National Commission for Nuclear Energy, Sao Paulo, Brazil The sensitivity of the IRMA method is limited by the specific activity (SA) of the conventionally employed radioisotopic label and high sensitivity radioimmunoassay should theoretically be attained by the use of short-lived radiometallic nuclides. Our group have achieved radiolabeling of high SA IgG by using the radiometal, gallium-67 (67Ga) with a short half-life (T1/2 = 78 h) and deferoxamine (DF), a bifunctional chelating agent bound through a multispacer (dialdehyde starch, DAS) as the linker (J Nucl Med 32: 825, 1991 ). In the present work, the application of the approach is attempted by employing a two-site IRMA for human growth hormone (hGH); the monoclonal antibody to hGH (MAB2) is bound to DF via DAS and the coupled DF-DAS-MAB2 is radiolabeled with 67Ga. The 67Ga-DF-DAS-MAB2 of high SA (4,884 MBq/mg versus 370-518 MBq/mg calculated for radioiodinated MAB2) was thus used for the two site 'sandwich' 67Ga-IRMA. Excellent correlation with the 125I-IRMA was registered, and higher detection capability obtained by using 67Ga over the 125I in the hGH IRMA offered a good basis for the exploitation of short-lived radio-nuclides in the IRMA system. Key words: high specific activity radiolabeled antibody, short half-life radiometal nuclide IRMA, 67Ga-IRMA, high sensitivity hGH assay, radioactive waste disposal INTRODUCTION IT HAS OFTEN BEEN REPORTED that the high sensitivity of the immunoradiometric assay (IRMA) method is limited by the specific activity of the conventionally employed radioisotopic labels such as 125I.1 In our laboratory, the labeling of high specific activity IgG has been achieved,2,3 by using deferoxamine (DF), a bifunctional chelating agent bound via a multi-site spacer (dialdehyde starch, DAS) to the protein as a means to increase the number of DF molecules per IgG molecule. The IgG-coupled DF was then labeled with the radiometal, gallium-67 (67Ga) with a short half-life (T1/2 = 78 hr). In the present work, the enhancement of the assay sensitivity through the use of high specific activity 67Ga labeled antibody is attempted with a two-site IRMA system for hGH and its performance compared with the radioiodinated IRMA. Attainment of a hGH assay with improved sensitivity may be of clinical interest, i.e. follow-up of growth hormone therapy in elderly people.4 The data gathered by the short-lived (78 hrs) radiometallic nuclide of 67Ga offered the basis for the high specific activity needed to increase the sensitivity of IRMA. MATERIALS AND METHODS The following reagents were employed for this work: gallium-67-citrate (67Ga) (187 MBq/5 ml) (Nihon Medi-Physics, Takarazuka, Japan), monoclonal antibody anti-hGH MAB652 (MAB2) (Chemicon International Inc., California, USA), deferoxamine mesylate (DF) (Ciba-Geigy, Basel, Switzerland and repurified by Ciba-Geigy, Japan), BSA RIA Grade (Sigma Chemical Co., Mo., USA), polymeric dialdehyde starch (DAS) (Sigma Chemi-cal Co., Mo., USA and purified by Nihon Medi-Physics, Takarazuka, Japan, the final product containing 5.25 mmol of aldehyde group per mg of DAS). The following ORIGINAL ARTICLE Annals of Nuclear Medicine Vol. 1O No 1, 49-55, 1996 Applicability of short-lived radiometallic nuclide for high sensitivity two-site "sandwich" immunoradiometric assay: Human growth hormone assay Kazuko HORIUCHI,* Lin H. LIN,** Yasuhisa FUJIBAYASHI,* Vania C. BORGHI** and Akira YOKOYAMA* *Department of Radiopharmaceutical Chemistry, Faculty of Pharmaceutical Sciences, Kyoto University, Japan **Institute of Energy and Nuclear Research, National Commission for Nuclear Energy, Sao Paulo, Brazil The sensitivity of the IRMA method is limited by the specific activity (SA) of the conventionally employed radioisotopic label and high sensitivity radioimmunoassay should theoretically be attained by the use of short-lived radiometallic nuclides. Our group have achieved radiolabeling of high SA IgG by using the radiometal, gallium-67 (67Ga) with a short half-life (T1/2 = 78 h) and deferoxamine (DF), a bifunctional chelating agent bound through a multispacer (dialdehyde starch, DAS) as the linker (J Nucl Med 32: 825, 1991 ). In the present work, the application of the approach is attempted by employing a two-site IRMA for human growth hormone (hGH); the monoclonal antibody to hGH (MAB2) is bound to DF via DAS and the coupled DF-DAS-MAB2 is radiolabeled with 67Ga. The 67Ga-DF-DAS-MAB2 of high SA (4,884 MBq/mg versus 370-518 MBq/mg calculated for radioiodinated MAB2) was thus used for the two site 'sandwich' 67Ga-IRMA. Excellent correlation with the 125I-IRMA was registered, and higher detection capability obtained by using 67Ga over the 125I in the hGH IRMA offered a good basis for the exploitation of short-lived radio-nuclides in the IRMA system. Key words: high specific activity radiolabeled antibody, short half-life radiometal nuclide IRMA, 67Ga-IRMA, high sensitivity hGH assay, radioactive waste disposal INTRODUCTION IT HAS OFTEN BEEN REPORTED that the high sensitivity of the immunoradiometric assay (IRMA) method is limited by the specific activity of the conventionally employed radioisotopic labels such as 125I.1 In our laboratory, the labeling of high specific activity IgG has been achieved,2,3 by using deferoxamine (DF), a bifunctional chelating agent bound via a multi-site spacer (dialdehyde starch, DAS) to the protein as a means to increase the number of DF molecules per IgG molecule. The IgG-coupled DF was then labeled with the radiometal, gallium-67 (67Ga) with a short half-life (T1/2 = 78 hr). In the present work, the enhancement of the assay sensitivity through the use of high specific activity 67Ga labeled antibody is attempted with a two-site IRMA system for hGH and its performance compared with the radioiodinated IRMA. Attainment of a hGH assay with improved sensitivity may be of clinical interest, i.e. follow-up of growth hormone therapy in elderly people.4 The data gathered by the short-lived (78 hrs) radiometallic nuclide of 67Ga offered the basis for the high specific activity needed to increase the sensitivity of IRMA. MATERIALS AND METHODS The following reagents were employed for this work: gallium-67-citrate (67Ga) (187 MBq/5 ml) (Nihon Medi-Physics, Takarazuka, Japan), monoclonal antibody anti-hGH MAB652 (MAB2) (Chemicon International Inc., California, USA), deferoxamine mesylate (DF) (Ciba-Geigy, Basel, Switzerland and repurified by Ciba-Geigy, Japan), BSA RIA Grade (Sigma Chemical Co., Mo., USA), polymeric dialdehyde starch (DAS) (Sigma Chemi-cal Co., Mo., USA and purified by Nihon Medi-Physics, Takarazuka, Japan, the final product containing 5.25 mmol of aldehyde group per mg of DAS). The following trophoresis and diluted 10 times with buffer D until use (one to two 67Ga half-life). The electrophoretic analysis was carried out with acetate cellulose strips (SE-2, Toyo Co.; Separax, Jookoo Co., Ltd., Tokyo, Japan), in buffer E and run for 30 min at 0.8 mA/cm. At the end of the electrophoresis, the strip was dried, cut into 0.5 cm segments and the radioactivity of each one determined in a gamma counter (Cobra-TM 5003, Packard Instrument Co., CT, USA). The labeling efficiency (%) was estimated as the ratio of the radioactivity associated with DF-DAS-MAB2 to the total radioactivity of the strip. D. Estimation of 67Ga-DF-DAS-MAB2 specific activity The specific activity (SA) of the radiolabeled conjugate was calculated by the equation below: SA (MBq/mg) *Radioactivity/**[MAB2] *Radioactivity (MBq/ml) measured in a Curiemeter (IGC-3, Aloka, Japan). **[MAB2] concentration (mg/ml) contained in DF-DAS-MAB2, estimated by the protein absorbance at 280 nm. III. Two-site 'Sandwich' IRMA procedure The standard pituitary hGH (0.50-50.00 ng/ml) was that provided with the commercial Allegro IRMA kit. The lower hGH concentration range of 0.03, 0.06, O. 12, 0.25, 0.50 ng/ml was prepared by diluting the 50 ng/ml standard of the kit with buffer D. l ) 125I-IRMA: This is the commercially available two-site 'sandwich' IRMA, provided with: a) Reagent A: the avidin coated beads (Bead-Av) and b) Reagent B: the 125I-antibody solution; that is, the radioiodinated MAB2, plus the biotin coupled MAB 1 (Bio-MAB 1 ). The assay is per-formed according to the manufacturer's instructions, with-out modification. 2) 67Ga-IRMA: corresponds to the two-site 'sandwich' IRMA with the 67Ga-DF-DAS-MAB2 prepared as described in the previous section (II) and modified reagents of the 125I-IRMA system as follows: a) 125I-MAB2 removal: 100 ml of the solution provided with the kit containing the biotin-MABl and the 125I-MAB2 (reagent B) was mixed with 100 ml of buffer C, and then one avidin coated bead (reagent A) was added, followed by station-ary incubation for 4 hours at room temperature. The supernatant was then aspirated and the bead was washed twice with 2 ml of buffer D each time. The MAB 1 coated beads linked through the Avidin-Biotin system (Bead Avidin-Biotin-MAB I ) were kept in buffer D at 4'C until their use (No more than 1-2 67Ga half-life). b)Two-site 'sandwich' 67Ga-IRMA procedure: 100 ml of 67Ga-DF-DAS-MAB2 (approximately 400,000 cpm diluted in buffer D) was mixed with 100ml of the sample or standard sample followed by the addition of one previously prepared MAB 1 coated bead as described in (a). After 4 hours of stationary incubation at room temperature, the supernatant was aspirated and the bead was washed twice with 2 ml of washing solution provided with the kit, and the 67Ga radioactivity bound to the bead was measured. The standard curve, the assay data and the hGH sample concentration were compared with those obtained by the 125I-IRMA, simultaneousl performed. RESULTS A. Coupling of DF-DAS-MAB2 The commercially supplied MAB anti-hGH (MAB652, named MAB2) was purified and after its coupling with the bifunctional chelating agent DF, via the water soluble spacer DAS, the non reduced SDS-PAGE analysis of DF-DAS-MAB2 indicated the presence of a single fraction at the position of the major fraction of uncoupled MAB2, closed to the molecular weight standard of 200,000 (Fig. l ). The analysis of DF molecules bound per MAB2 calcu-lated by the NTA method (6) indicated the DF-DAS-MAB2 conjugate as holding a conjugation level of DF/ DAS-MAB2 of 11.81 to 13.45 and that of DF-DAS/ MAB2 of 1.19 to 1.37. Approximately one molecule of DAS is therefore holding 9.30 to 11.00 DF molecules per MAB2. The conjugate obtained was then diluted in order to have 100 mg/ml of MAB2 before the radiolabeling. No change in stability was detected over a period of three months. B. Radiolabeling of DF-DAS-MAB2 The DF-DAS-MAB2 conjugate was easily radiolabeled with the purified iron-free 67Ga-citrate by a simple mixing reaction. The purified 67Ga solution (range 329.30 to 21 ,645.00 MBq/ml) was added to the DF-DAS-MAB2 conjugate preparation (100 mg/ml) reaching a labeling efficiency between 90 and 95% whenever the radioactivity of the purified gallium solution was lower than 3,700 MBq/ml (Table 1). In this range the electrophoretic analysis detected a single peak near the site of sampling at the center of the strip (Fig. 2), and neither a 67Ga-DF peak nor a 67Ga free peak was observed. The colorimetric Ponceau analysis demonstrated the peak as being that of protein. As the radioactivity of the 67Ga solution increased, the electrophoretic analysis detected another peak in the anodic site, corresponding to unbound free 67Ga; in fact, at the 67Ga concentration of 21 ,645 MBq/ml the labeling yield decreased to 23% (with 72% of free 67Ga fraction), requiring the purification of the radiolabeled conjugate. C. Stability of 67Ga-DF-DAS-MAB2 The stability of 67Ga-DF-DAS-MAB2, followed through one half-life of the radionuclide (78 h), showed some sign of change if kept at a very high specific activity level (range 3,700-4810 MBq/mg), but the simple mediation of a 10 fold dilution kept the radiolabeled conjugate stable, as seen in Figure 3, so that any following assay requiring a 67Ga-DF-DAS-MAB2 conjugate with high specific activity was carried out immediately after its radiolabeling or stored in the diluted condition at 4C for two half-lives at the most. D. Two-site 'Sandwich' IRMA The hGH two-site 'sandwich' IRMA was performed with 100 ul of the 67Ga-DF-DAS-MAB2 at the highest specific activity obtained (4,884 MBq/mg) providing approximately 400,000 cpm per tube, as described in Material and Methods. In the 67Ga-IRMA procedure, the percent bound/total counts (% B/T) versus the hGH concentration provided good linearity in both ranges, high (0.50-50.00 ng/ml) (Fig. 4A) and low (0.03-0.50 ng/ml) (Fig. 4B); the linearity range of the 125I-IRMA kit (Fig. 4A) was that described in the Allegro kit brochure, up to 0.50 ng hGH/ml. The linearity of the 67Ga-IRMA could reach 0.03 ng/ml (Fig. 4B). The highly significant correlation (r = 0.9977; p < 0.001) found between the hGH concentration of the sample measured by the 125I-IRMA and the newly formulated 67Ga-IRMA procedure is shown in Figure 5. DISCUSSION In recent years, the trend toward increasing the sensitivity of the isotopically based immunoassay method has led to the adoption of the noncompetitive approach. Among the methods, the two-site immunoradiometric assay (IRMA) has offered several advantages in comparison with the traditional IRMA or the competitive radioimmunoassay. On the other hand, the conventionally employed radioisotopic label, such as 125-I (T l/2 = 60.2 days), involves a limitation due to its slow desintegration rate.1 The use of radionuclides, with faster desintegration rate, such as 67Ga (Tl /2 = 3.24 days) allows the labeling of IgG with higher specific activity (SA), particularly if a multi-site spacer like the DAS is inserted as the linker between the MAB and the metal chelating (DF) molecule 3 as a means to increase the number of DF per MAB molecule. The antibody labeled with radiometallic nuclides at high specific activity appeared to be of interest in the IRMA procedure. Thus, to validate the presented conceptual approach, a commercial assay system containing MABs Of facile acquisition was judged as primordial. In fact, in the selected two-site IRMA kit (Allegro), the antihGH mouse monoclonal antibody (MAB2) was commercially available. And, the MAB2 could be easily coupled with DF; the SDS-PAGE (Fig.1 ) demonstrated the presence of a DF-DAS-MAB2 molecule with slightly slower migration rate than the free MAB2, probably due to the presence of approximately one PAS molecule (MW = 7,000) plus 11-13 DF molecules (MW = 656 x 11 or x 13 = 7,216 or 8,528) contributing to the change in the migration. Since as high a DF/MAB2 ratio as 11.80 was estimated in 100 uL of DF-DAS-MAB2 containing 10 mg MAB2 (MAB2 = 100 ug/ml), about 780 pmol of DF may be present, so it offered a MAB2 conjugate with capacity to coordinate a great amount of the radiometal per MAB2 molecule. As the coordinating metal, gallium-67 in a concentration range from 329 MBq/ml to 21,645 MBq/ml was used (theoretical calculation: 222 to 14,600 pmol/ml of 67-Ga) (1 MBq = 4.5 x l0e-7 g; 1 mCi = 1.67 x l0e-9 g).7 Then, at a constant concentration of DF-DAS-MAB2 (1OO ug/ml), the calculation indicated that in spite of sufficient DF coordination sites for the radiometal (DF approx 780 pmol/1OO uL), the addition of a solution containing 46.30 pmol of 67Ga ( 10 uL of 6,867 MBq/ml of 67Ga) to the DF-DAS-MAB2 conjugate, induced a saturation (Table 1 ). A calculation can estimate the coordination of about 25 pmol of 67Ga to the DF-DAS-MAB2; an amount 30 times lower than the calculated moles of DF. This achieved DF/67-Ga coordination ratio represented, nevertheless, an improvement of various order of magnitude over the 67-Ga-radiolabeling of fibrinogen previously reported by Takahashi,6 which requires a 3,000 times excess of DF over 67Ga. Labeling reactions dealing with radionuclides at high specific activity in the micro or nano concentration order is calling the attention of several disciplines.8 The SA achieved for the anti-hGH antibody, although lower than that reached for the mouse CA-130-22,3 was as high as 4,884 MBq/mg, almost 10 times higher than the theoretical one calculated from the radioiodinated antibody (370-518 MBq/mg).9 And when compared with other 67Ga labeled preparations, the SA of the 67Ga-DF-DAS-MAB2 was much higher than the values reported by Wuro of 37-111 MBq/mg at 1.8 mole DF/mole of antibody and by Motta-Hennessyll of 1.85-7.40 MBq/mg at 0.9 mole DF/mole antibody. On the other hand, in spite of the various modifications introduced in the MAB2, the preservation of DF-DAS-MAB2 immunoreactivity could be assessed from the paralleled standard curves of the 67Ga-IRMA to the 125I-IRMA, in the higher range (Fig. 4A); this was strongly supported by the good correlation between the concen-tration of hGH measured by both methods, as shown in Figure 5 . According to the theory of the IRMA kinetic, the limiting factor of the two-site 'sandwich' IRMA sensitivity is the SA of the labeled MAB rather than its affinity constant.1 The theoretical calculation, based on 1251 and 67Ga half-life alone, indicates a 20 times higher SA for 67Ga (81.4 TBq/mmol vs. 1,480 TBq/mmol respectively). In fact, in the low range, as indicated in Figure 4B, the 67Ga-IRMA with the synthesized 67Ga-DF-DAS-MAB2 at high SA (> 3,700 MBq), enabled the measurement of a lower hGH concentration than that of 125I-IRMA. Never-theless, the higher CV values for the 67Ga-IRMA induced a similar MDD range for both methods (Table 2), an indication of further experimental intervention, required in the separation procedure. In conclusion, the coupling and the radiolabeling of monoclonal antibody for hGH assay with the radiometallic nuclide 67Ga could be readily accomplished and, as predicted the radiolabeling of 67Ga-DF-DAS-MAB2 at a very high SA was achieved, enabling the measurement of a smaller amount of hGH than the 125I-IRMA counterpart. Above all, the data gathered indicated the feasibility of the theoretical approach to the increase in radiolabeled biomolecule specific activity, by the use of a short half-life radio-metallic nuclide and its consequent effect on the assay sensitivity. Another challenge involved in the use of short-lived radionuclides is the radioactivity waste disposal problem faced in today's clinical assay practice. The prompt availability of a generator produced 68Ga is most desirable; due to its even shorter half-life of 68.3 min, approximately only 5,890 molecules of 68Ga against the 4.02 x 10^5 molecules of 67Ga is theoretically required to detect 1 dps (i.e. 7.5 x 10^6 labeled molecules for 1251, T1/2 60.2 d). Reactions with radiometals at microconcen-trations are a challenging step toward the predicted goals and there are still many unknown physical and chemical factors to be determined, but the present results will serve as a basis on which to build a future understanding to facilitate better exploitation of short-half life radiometallic nuclides. ACKNOWLEDGMENT The authors wish to thank Nihon Medi-Physics for financial support and the supply of 67Ga citrate and other reagents. Thanks are also extended to Ms. Takako Kosaka from the Department of Nuclear Medicine, Kyoto University Hospital for guidance, assistance and suggestions during the development of this work. REFERENCES l . Ekins R, Chu F, Biggart E. Fluorescence spectroscopy and its application to a new generation of high sensitivity, multi-microspot, multianalyte immuno-assay. Clin Chim Acta 194: 91-114, 1990. 2. Furukawa T, Fujibayashi Y, Fukunaga M, Yokoyama A. Ga-labeling of immunoglobulin G with high specific radio-activity. Chem Pharm Bull 38: 2285-2286, 1990. 3. Furukawa T. Fujibayashi Y, Fukunaga M, Saga T, Endo K, Yokoyama A. An approach for immunoradiometric assay with metallic radionuclides: Gallium-67 Deferoxamine-dialdehyde starch-IgG. J Nucl Med 32: 825-829, 1991 . 4. Vance ML. Growth Hormone Therapy in Adults. Endocrinol Metab 3: 46-48, 1992. 5. Emery T, Hoffer PB. Siderophore mediated mechanism of gallium uptake demonstrated in the microorganism Ustilago Sphaerogena. J Nucl Med 21 : 935-939, 1980. 6. Takahashi K, Ueda N, Hazue M, Ohmomo Y, Yokoyama A, Suzuki T, et al. Preparation and biodistribution of 67Ga-labeled fibrinogen conjugated with a water-soluble polymer containing deferoxamine, a potential thrombus imaging agent. In Radiopharmaceutical and Labeled Compounds. Vienna, IAEA Proceeding, pp. 471-482, 1985. 7. Wolf W. Radiochemistry. In Radiopharmacy, Tubis M, Wolf W (ed.), New York, John Wiley & Sons, pp. 203-223, 1976. 8. Finn R, Cheung N-K V. Divgi C, St. Germain J, Graham M, Pentlow K, et al. Technical challenges associated with the radiolabeling of monoclonal antibodies utilizing short lived positron emitting radionuclides. Nucl Med Biol Int J Radiat Appl Instrum Part B 18: 9-13, 1991. 9. Eckelman WC, Paik CH, Reba RC. Radiolabeling of anti-bodies. Cancer Res 40: 3036-3042, 1980. lO. Wu RS. Novel bifunctional linkers for antjbody chelation. In Cancer imaging with radiolabeled antibodies, Goldenberg DM (ed.), Boston, Kluwer Academic Publishers, Boston, pp. 212-232, 1990. 11. Motta-Hennessy C, Eccles SA. Dean C, Coghlan G. Preparation of 67Ga-labelled human IgG and its Fab fragments using desferoxamine as chelating agent. Eur J Nucl Med 11 : 240-245, 1985. 12. a) Feldkamp CS, Smith SW. Practical guide to radioimmunoassay method evaluation. In Immunoassay: A Practical Guide. Chan DW, Perlstein MT (ed.), Academic Press, pp. 70-72, 1987. b) Hisazumi K, Kosaka T, Nishikawa S, Usami M, Morita O, Hamatsu N, et al. Practical Evaluation Method of Radioimmunoassay Kit. Nihon Houshasen Gljutsu Gakkai Zasshi 42: 565-573, 1986.