Chemical Design of Radiolabeled Antibody Fragments for Low Renal Radioactivity Levels¹

LMW³ antibody fragments such as Fab and single-chain Fv fragments provide rapid tumor targeting and uniform distribution in tumor tissues (1). Such characteristics render antibody fragments attractive vehicles for the delivery of radioactivity to tumor tissues for both targeted imaging and radiotherapy. However, high and persistent localization of the radioactivity is observed in the kidneys, which compromises tumor visualization in the kidney region and limits therapeutic potential (2, 3, 4, 5). Indeed, ⁶⁷Ga-labeled Fv fragments indicated a specific:nonspecific tumor ratio of radioactivity of 6 at 90 min postinjection, whereas the renal radioactivity levels reached 205% ID/g of tissue at the same postinjection time (6). Thus, radiolabeled LMW antibody fragments would become much more useful for targeted imaging and targeted therapy if the renal radioactivity levels could be reduced without impairing those in the target tissues.

Previous studies indicated that radiolabeled antibody fragments are most likely reabsorbed by proximal tubules via luminal endocytosis after glomerular filtration (7, 8, 9, 10, 11, 12, 13). The long residence times of the radiometabolites generated after lysosomal proteolysis of radiolabeled antibody fragments in renal cells were also reported to be responsible for the persistent renal radioactivity levels (5, 6, 14, 15, 16).

Two major strategies have been developed to reduce the renal radioactivity levels of radiolabeled antibody fragments. One is the blocking of reabsorption of radiolabeled antibody fragments themselves into renal cells by infusion of basic amino acids such as l-lysine so that the radiolabeled antibody fragments appear directly in the urine without being incorporated and metabolized in lysosomes of renal proximal cells. The other strategy involves interposition of a cleavable linkage between an antibody fragment and a radiolabeled compound of urinary excretion so that the radiolabeled compound is excreted from the lysosomal compartments of renal cells into the urine after lysosomal proteolysis of the radiolabeled antibody fragment in renal cells. Previous studies of renal handling of radiolabeled antibody fragments support the rationale of the two approaches. Indeed, successful reduction of renal radioactivity levels was demonstrated by the former approach using Fab and Fv fragments (17, 18, 19). However, further studies of administration dose and timing of basic amino acids as well as the toxicity of amino acids given in high doses will be required before the former approach can be used routinely (19, 20). Although the latter approach was extensively investigated and shown to reduce the hepatic radioactivity levels of intact antibodies, there have been only a few studies of LMW antibody fragments (21, 22, 23), and further studies may be required with regard to the relationship between chemical structures of radiometabolites and their elimination rates from the lysosomal compartments of renal cells. In addition, a previous study showed that use of ester bonds as the cleavable linkage may impair the target radioactivity levels delivered by antibody fragments due to cleavage of the ester bonds while the radiolabeled antibody fragments are circulating in plasma (23).

Meanwhile, studies of renal handling of LMW peptides indicated that some glomerularly filtered peptides are hydrolyzed to free amino acids during a short contact time with the brush border enzymes present on the lumen of the renal proximal cells (11). Indeed, several peptidases have been identified on the brush border membrane of renal tubules (24). These findings suggested that the renal localization of radioactivity may be reduced if radiolabeled compounds of urinary excretion are liberated from the antibody fragments before they are incorporated into the renal cells by the action of the brush border enzymes. This strategy may be advantageous in reducing the renal radioactivity levels at early postinjection times because radiolabeled compounds are released from antibody fragments before they are incorporated and transported to the lysosomal compartment of the renal cells. This approach may be applicable to antibody fragments that are internalized into target cells if radiolabeled compounds released from the protein molecules at the renal tubules are excreted rapidly into urine but are retained within the target cells for long postinjection intervals.

To verify this hypothesis, meta-iodohippuric acid was selected as the radiolabeled compound because of its rapid urinary excretion from the renal tubular lumen and high stability against in vivo deiodination (25, 26). Because the glycyl-lysine sequence is a substrate for one of the brush border enzymes, carboxypeptidase M (27), present on the lumen of renal tubules, [¹³¹I]HML was conjugated to the Fab fragment of a mAb against osteogenic sarcoma after treatment of the antibody fragment with 2-IT. When the peptide bond between the glycine and the lysine in HML is cleaved, meta-iodohippuric acid is released from the conjugate. The tissue distribution and subcellular distribution of the radioactivity after administration of [¹³¹I]HML-Fab were compared to those of simultaneously administered ¹²⁵I-Fab prepared by direct radioiodination of Fab fragment. In addition, meta-[¹²⁵I]iodohippuric acid was also conjugated to Fab fragments via an ester bond ([¹²⁵I]MIH-Fab) or an amide bond ([¹²⁵I]MPH-Fab) by procedures similar to those used for [¹³¹I]HML-Fab. Previous studies indicated that, although MIH-conjugated polypeptides generated meta-iodohippuric acid as the sole radiometabolite after lysosomal proteolysis in hepatocytes, the radiometabolites derived from MPH-conjugated polypeptide showed long residence times in hepatocytes (23, 25, 26, 28). The chemical structures of the radioiodinated Fab fragments used in this study are illustrated in Fig. 1. The validity of the new chemical design of antibody fragments in reducing the renal radioactivity levels was estimated.

Na[¹³¹I]I and Na[¹²⁵I]I were obtained from Daiichi Radioisotope Laboratories (Tokyo, Japan) and Daiichi Kagaku (Tokyo, Japan), respectively, and were diluted with 0.1 m PB (pH 7.4) to 37 MBq (1 mCi)/2–10 μl. The stannyl precursors of the radioiodination reagents, SHML, SMPH, and SMIH, were synthesized as reported previously (25, 26). Size-exclusion HPLC and RP-HPLC were performed using Cosmosil Diol-300 (7.5 × 600 mm, Nacalai Tesque, Kyoto, Japan) and Cosmosil 5C₁₈-AR (4.6 × 250 mm, Nacalai Tesque) columns, respectively. Size-exclusion HPLC and RP-HPLC columns were equilibrated and eluted with 0.1 m PB (pH 6.8) and with a mixture of 0.1% aqueous phosphoric acid and acetonitrile (7:3), respectively, at a flow rate of 1 ml/min. Other reagents were of reagent grade and were used as received.

KT005-cloned human osteogenic sarcoma was maintained by serial s.c. transplantation in athymic mice. Pieces of tumor tissue (∼0.5 g) at ∼2.5 weeks postimplantation were used for the in vivo study. The mAb against osteogenic sarcoma (OST7, IgG1), generated by the standard hybridoma technique, was purified by ammonium sulfate precipitation with subsequent protein A affinity chromatography (Pharmacia Biotech Co. Ltd., Tokyo, Japan), as reported previously (29). The Fab fragment was prepared by the standard procedure using papain (30).

SHML, SMPH, and SMIH were radioiodinated in the presence of NCS as described previously (26). Briefly, SHML was dissolved in methanol containing 1% AcOH (0.64 mg/ml), and 8.4 μl (0.021 μmol) of this solution were mixed with 2.3 μl of NCS in methanol (0.5 mg/ml) in a sealed vial, followed by addition of Na[¹³¹I]I (1 μl). After incubation at room temperature for 25 min, the reaction was quenched with aqueous sodium bisulfite (1.1 μl, 0.72 mg/ml). The radiochemical yield of [¹³¹I]HML was determined by silica gel TLC (Merck Art 5553) developed with a mixture of chloroform, methanol, and AcOH (40:5:2). The solvent was removed under a stream of N₂ prior to subsequent conjugation reaction with the Fab fragment.

Conjugation of [¹³¹I]HML with the Fab fragment of OST7 was performed by treating the Fab fragment with 2-IT, followed by the conjugation reaction between the thiol groups of Fab and the maleimide groups of [¹³¹I]HML (26). Briefly, a solution of Fab (200 μl, 2 mg/ml) in well-degassed 0.16 m BB (pH 8.0) containing 2 mm EDTA was allowed to react with 7.2 μl of 2-IT solution (1 mg/ml) prepared in the same buffer. After gentle agitation of the reaction mixture for 60 min at room temperature, excess 2-IT was removed by a centrifuged column procedure (31) using Sephadex G-50 (Pharmacia Biotech Co. Ltd.) equilibrated and eluted with 0.1 m PB (pH 6.0) containing 2 mm EDTA. Aliquots of this mixture were sampled for estimation of the number of thiol groups with 2,2′-dipyridyl disulfide (32). The filtrate (100 μl) was then added to a reaction vial containing crude [¹³¹I]HML. After gentle agitation of the reaction mixture for 1.5 h at room temperature, 14.8 μl of iodoacetamide (10 mg/ml) in 0.1 m PB (pH 6.0) were added. The reaction mixture was further incubated for 30 min to alkylate the unreacted thiol groups. [¹³¹I]HML-Fab was subsequently purified by the centrifuged column procedure, equilibrated, and eluted with 0.1 m acetate buffer (pH 6.0).

[¹²⁵I]MIH-Fab and [¹²⁵I]MPH-Fab were also prepared by experimental procedures similar to those used for [¹³¹I]HML-Fab, except that SMIH or SMPH was used in place of SHML, as reported previously (25, 26).

Direct radioiodination of Fab was performed by the chloramine T method (23). To 100-μl aliquots of Fab (2.5 mg/ml) in 0.3 m PB (pH 7.3) was added 1 μl of Na[¹²⁵I]I. Chloramine T (0.1 mg/ml, 12.5 μl), freshly prepared in the same buffer, was then added. After incubation of the mixture for 5 min at room temperature, ¹²⁵I-Fab was purified by the centrifuged column procedure, as described above.

Radiochemical purities of the radioiodinated Fab fragments were determined by size-exclusion HPLC and silica gel TLC developed with a mixture of methanol and water (4:1).

Each radiolabeled Fab fragment was diluted to 0.5 mg/ml with 0.1 m PB (pH 6.0), and 20 μl of each solution were added to 230 μl of freshly prepared human serum. After incubation for 1, 3, 6, 24, or 48 h at 37°C, samples were taken from the solutions, and the percentages of radioactivity released from each antibody were determined by paper chromatography (No. 50; Advantec Toyo, Tokyo, Japan) and silica gel TLC. Each value was calculated by dividing the antibody-bound radioactivity at various intervals by the radiochemical purity of freshly prepared respective antibody. Paper chromatography was developed with a mixture of methanol and water (4:1). Under these conditions, protein-bound radioactivity, free iodine, m-iodobenzoic acid, and m-iodohippuric acid had R_fs of 0, 0.6–0.65, 0.9–0.95, and 0.9–0.95, respectively. TLC plates were developed with a mixture of chloroform, methanol, and water (15:8:1). Under these conditions, protein-bound radioactivity, free iodine, meta-iodohippuric acid, and meta-iodobenzoic acid had R_fs of 0, 0.1, 0.25, and 0.4, respectively.

In vivo studies were performed after simultaneous administration of a mixed solution of [¹³¹I]HML-Fab and [¹²⁵I]MIH-Fab, [¹²⁵I]MPH-Fab, or ¹²⁵I-Fab. Three ml of 0.1 m PB (pH 6.0) containing 600 μg of unmodified Fab were mixed with 10 μl each of [¹³¹I]HML-Fab (12 μCi) and [¹²⁵I]MIH-Fab (13 μCi), [¹²⁵I]MPH-Fab (13 μCi), or ¹²⁵I-Fab (13 μCi) prior to administration. The biodistribution of radioactivity after i.v. administration of each mixture to 6-week-old ddY mice (33) was monitored at 10 and 30 min and 1, 3, 6, and 24 h postinjection. Groups of five mice, each receiving 20 μg of Fab fragments, were used for the experiments. Organs of interest were removed and weighed, and the radioactivity was determined with a well counter (ARC 2000; Aloka, Tokyo, Japan). A window from 29 to 97 keV was used for counting ¹²⁵I, whereas one from 280 to 440 keV was used for ¹³¹I. Correlation factors to eliminate cross-over of ¹³¹I activity into ¹²⁵I were determined by counting ¹³¹I standard in each window. The crossover of ¹²⁵I into the ¹³¹I channel was negligible. To determine the amounts and routes of radioactivity excretion from the body, mice were housed in metabolic cages (Metabolica MM type; Sugiyama-Gen Iriki Co. Ltd., Tokyo, Japan) for 24 h after administration of the respective preparation. Urine and feces were collected for 6 and 24 h postinjection, and the radioactivities were determined.

The radiolabeled species excreted in the urine for 6 h postinjection of [¹³¹I]HML-Fab were immediately analyzed by silica gel TLC developed with a mixture of chloroform, methanol, and water (15:8:1) before and after filtration through a M_r 10,000 cutoff ultrafiltration membrane (Microcon 10; Amicon Grace, Tokyo, Japan). The urine samples were also analyzed by size-exclusion HPLC and RP-HPLC after filtration through a polycarbonate membrane with a pore diameter of 0.45 μm (Cosmonice Filter, Nacalai Tesque) and a M_r 10,000 cutoff ultrafiltration membrane, respectively. On RP-HPLC analysis, free iodine, meta-iodohippuric acid, and meta-iodobenzoic acid had retention times of 3, 9, and 31 min, respectively.

Athymic mice bearing osteogenic sarcoma were also treated i.v. with 100 μl of a mixed solution of [¹³¹I]HML-Fab and ¹²⁵I-Fab, prepared as described above. The animals were sacrificed at 3 h postinjection (six mice). Organs of interest were removed and weighed, and the radioactivity in each tissue was determined (ARC-2000).

The subcellular distribution of the radioactivity in the kidney was determined 10 and 30 min after administration of 20 μg each of [¹³¹I]HML-Fab, [¹²⁵I]MIH-Fab, or ¹²⁵I-Fab, according to the procedures described previously, with slight modifications (34). At the indicated times, murine kidneys were isolated, minced, and homogenized in 4 volumes of ice-cold 0.25 m sucrose containing 1 mm EDTA with a Dounce homogenizer by hand (20 strokes). This was followed by two final strokes in an ice-cold Potter-Elvehjem homogenizer with a Teflon pestle rotated at 800 rpm. The resulting 20% homogenate was centrifuged twice for 5 min at 340 × g at 4°C. The isolated supernatant was layered on top of iso-osmotic (0.25 m sucrose) Percoll (Pharmacia Biotech Co. Ltd.) at a density of 1.08 g/ml. After centrifugation at 20,000 × g (RP 30 rotor: Hitachi Co. Ltd., Tokyo, Japan) for 90 min at 4°C, the gradient (∼1.02–1.14 g/ml) was collected in 14 fractions before analyses of β-galactosidase (35), alkaline phosphodiesterase I activities (36), density, and the radioactivity of each fraction.

Data are expressed as means ± SD where appropriate. Results were statistically analyzed using the unpaired Mann-Whitney test. Differences were considered statistically significant when P was <0.05.

Radioiodination of SHML was performed in the presence of NCS as an oxidant. [¹³¹I]HML was obtained with a radiochemical yield of 67.2%, as determined by TLC. Similar procedures were used for the preparation of [¹²⁵I]MIH and [¹²⁵I]MPH with radiochemical yields of 68.9 and 71.2%, respectively. All of the radioiodinated compounds were treated with sodium bisulfite to quench any residual NCS, preventing the iodine species from reacting with tyrosine residues of the antibody molecule. Neutralization of NCS also prevented exposure of the protein to the oxidant.

Conjugation of [¹³¹I]HML with Fab fragments was performed by reacting the thiolated Fab fragments with the maleimide groups of [¹³¹I]HML, as shown in Fig. 2. A preliminary study showed that only the thiol groups of Fab reacted with the maleimide groups and that the alkylation of lysine residues with maleimide was negligible or nonexistent under these conditions (37). Treatment of Fab fragments with 2-IT introduced ∼2.4 mol of thiol groups per molecule of Fab, as determined using 2,2′-dipyridyl disulfide. The radiochemical yield of [¹³¹I]HML-Fab was 35.1%, as determined by TLC. After purification by the centrifuged column procedure, [¹³¹I]HML-Fab was obtained with a radiochemical purity and specific activity of 95.8% and 575 μCi/mg, respectively, as determined by size-exclusion HPLC. [¹²⁵I]MPH-Fab and [¹²⁵I]MIH-Fab were obtained with radiochemical yields of 36.7 and 36.4%, respectively. Radiochemical purities and specific activities were 99.7% and 625 μCi/mg, respectively, for [¹²⁵I]MPH-Fab, whereas those for [¹²⁵I]MIH-Fab were 99.5% and 650 μCi/mg. ¹²⁵I-Fab was prepared by the chloramine T method with radiochemical purity and specific activity of 94.9% and 640 μCi/mg, respectively.

[¹³¹I]HML-Fab, [¹²⁵I]MPH-Fab, and ¹²⁵I-Fab released only 2–3% of the initial antibody-bound radioactivity, even after incubation in freshly prepared human serum at 37°C for 48 h. However, [¹²⁵I]MIH-Fab released radioactivity as meta-iodohippuric acid in a time-dependent manner, and 81.6, 44.4, and 22.0% of the initial antibody-bound radioactivity were detected in the protein fractions after 3, 24, and 48 h of incubation, respectively.

The biodistributions of the radioactivity after simultaneous administration of [¹³¹I]HML-Fab and [¹²⁵I]MPH-Fab to normal mice are summarized in Table 1. No significant differences were observed in the blood radioactivity levels between the two radioiodinated Fab fragments. However, although [¹²⁵I]MPH-Fab showed high radioactivity levels in the kidney and reached its peak at 1 h, [¹³¹I]HML-Fab showed significantly lower renal radioactivity levels at all postinjection time points examined. The biodistributions of the radioactivity after concomitant injection of [¹³¹I]HML-Fab and ¹²⁵I-Fab are shown in Table 2. Although [¹³¹I]HML-Fab showed slightly higher radioactivity levels in the blood, [¹³¹I]HML-Fab demonstrated significantly lower radioactivity levels in the kidney than ¹²⁵I-Fab from 10 min postinjection onward. ¹²⁵I-Fab showed the highest radioactivity levels in the kidney at 30 min postinjection. Table 3 shows the localizations of the radioactivity after concomitant administration of [¹³¹I]HML-Fab and [¹²⁵I]MIH-Fab. Although [¹³¹I]HML-Fab showed slightly slower elimination rates of the radioactivity from the blood, [¹³¹I]HML-Fab demonstrated significantly lower levels of radioactivity in the kidney at early postinjection time points as compared with [¹²⁵I]MIH-Fab. The kidney:blood ratios of the radioactivity in the three experiments are summarized in Fig. 3. [¹²⁵I]MPH-Fab showed the highest kidney:blood ratios of radioactivity, reaching a peaked ratio of 16.9 at 6 h postinjection. Although [¹²⁵I]MIH-Fab showed lower kidney:blood ratios than those of ¹²⁵I-Fab throughout the experimental period, both [¹²⁵I]MIH-Fab and ¹²⁵I-Fab reached their peak ratios of 3.8 and 7.3, respectively, at 1 h postinjection. On the other hand, [¹³¹I]HML-Fab demonstrated almost constant radioactivity ratios of ∼1 from 10 min to 3 h postinjection, and the ratios were significantly lower than those observed with [¹²⁵I]MIH-Fab.

Fig. 4 shows the radiochromatograms of urine samples obtained for 6 h postinjection of [¹³¹I]HML-Fab when analyzed by size-exclusion HPLC (Fig. 4,A) and TLC (Fig. 4,B). On size-exclusion HPLC, the majority of the radioactivity (85.5%) was observed in the LMW fractions with ∼12% of the radioactivity being eluted in fractions similar to those of intact [¹³¹I]HML-Fab. On TLC analysis, 14.6% of the radioactivity was detected at the origin and 84.4% of the radioactivity was observed with an R_f of 0.25. Fig. 4, C and D, show RP-HPLC and TLC radiochromatograms of urine samples after filtration through a M_r 10,000 cutoff membrane. On RP-HPLC, 93.0% of the radioactivity of the urine samples was observed with a retention time of 9 min, identical to that of meta-iodohippuric acid. On TLC analysis, 97.9% of the radioactivity in the urine samples was detected with an R_f of 0.25, identical to that of meta-iodohippuric acid. On analysis of urine samples obtained 6 h postinjection of [¹²⁵I]MIH-Fab, >95% of the radioactivity of the LMW fractions was detected at a position identical to that of meta-iodohippuric acid (data not shown).

The biodistributions of radioactivity after simultaneous administration of [¹³¹I]HML-Fab and ¹²⁵I-Fab in nude mice bearing osteogenic sarcoma are summarized in Fig. 5. [¹³¹I]HML-Fab showed significantly lower radioactivity levels in the kidney than of ¹²⁵I-Fab, whereas no significant differences were observed in the radioactivity levels in the tumor between the two radioiodinated Fab fragments. As a result, [¹³¹I]HML-Fab increased the tumor:kidney ratios of the radioactivity by a factor of ∼4.

Percoll density gradient centrifugation profiles of the radioactivity at 10 and 30 min postinjection of [¹³¹I]HML-Fab, [¹²⁵I]MIH-Fab, and ¹²⁵I-Fab are illustrated in Fig. 6. Both [¹²⁵I]MIH-Fab and ¹²⁵I-Fab showed two radioactivity peaks at densities of 1.04–1.05 and 1.10 g/ml at 10 min postinjection. At 30 min postinjection, the major radioactivity peak was detected at a density of 1.10 g/ml. On the other hand, [¹³¹I]HML-Fab showed a major radioactivity peak only at the density of 1.04–1.05 g/ml at both 10 and 30 min postinjection. The radioactivity peak at a density of 1.04–1.05 was copurified with alkaline phosphodiesterase I activity, whereas that at a density of ∼1.10 was copurified with β-galactosidase activity.

This study was performed to evaluate the hypothesis that the renal radioactivity levels of radiolabeled antibody fragments could be reduced if radiolabeled compounds of urinary excretion were released from the antibody fragments before they were incorporated into renal cells by the action of brush border enzymes present on the lumen of renal tubules. The results with [¹³¹I]HML-Fab were compared with those of ¹²⁵I-Fab, [¹²⁵I]MIH-Fab, and [¹²⁵I]MPH-Fab. [¹³¹I]HML-Fab, [¹²⁵I]MPH-Fab, and [¹²⁵I]MIH-Fab were prepared with similar conjugation chemistry. Previous studies using galactosyl-neoglycoalbumin and intact IgG showed that [¹²⁵I]MIH-conjugated polypeptides released meta-iodohippuric acid as the sole radiometabolite after lysosomal proteolysis in hepatic parenchymal cells at a rate similar to that of [¹³¹I]HML-conjugated polypeptides (23, 25, 26, 28). On the other hand, [¹²⁵I]MPH-conjugated polypeptides generated radiometabolites with long residence times in hepatic parenchymal cells. High resistance against in vivo deiodination of HML-, MIH-, and MPH-conjugated polypeptides and their radiometabolites was also confirmed.

Both [¹³¹I]HML-Fab and [¹²⁵I]MPH-Fab exhibited stable attachment of the radiolabels to the respective protein in serum. This was reflected in the similar radioactivity levels of the two radiolabeled Fab fragments in the blood from 10 min to 24 h postinjection (Table 1). Thus, similar radioactive portions of the radiolabeled Fab fragments would be filtered through the glomerulus and transported to the proximal tubules of the kidney after administration. However, although [¹²⁵I]MPH-Fab reached its peak renal radioactivity of ∼65% ID/g at 1 h postinjection, [¹³¹I]HML-Fab demonstrated markedly lower renal radioactivity levels in a time-dependent manner. [¹³¹I]HML-Fab also demonstrated significantly lower renal radioactivity levels when compared with simultaneously administered ¹²⁵I-Fab that reached its maximum renal radioactivity levels of ∼65% ID/g at 30 min postinjection (Table 2). Because >96% of the radioactivity excreted from the body was recovered in the urine (Tables 1,2,3) and meta-[¹³¹I]iodohippuric acid represented >80% of the radioactivity in the urine after administration of [¹³¹I]HML-Fab (Fig. 4), the significantly lower renal radioactivity levels achieved by [¹³¹I]HML-Fab were attributed to the rapid and selective release of meta-[¹³¹I]iodohippuric acid from the Fab fragment in the kidney.

[¹³¹I]HML-Fab also significantly reduced the renal radioactivity levels even from 10 min postinjection onward, compared with simultaneously administered [¹²⁵I]MIH-Fab (Table 3). This was clearly demonstrated when the kidney:blood ratios of the radioactivity were compared (Fig. 3). Because both HML- and MIH-conjugated polypeptides generated meta-iodohippuric acid as radiometabolites at a similar rate after lysosomal proteolysis in hepatocytes (26, 28), similar kidney:blood ratios of the radioactivity would be observed with the two radioiodinated Fab fragments if [¹³¹I]HML-Fab liberated meta-[¹³¹I]iodohippuric acid after lysosomal proteolysis in the renal cells. Indeed, no significant differences were observed in the liver:blood ratios of the radioactivity after simultaneous administration of [¹³¹I]HML- and [¹²⁵I]MIH-conjugated intact IgG into mice, although [¹²⁵I]MIH-IgG showed lower radioactivity levels in both the liver and the blood (26, 28). Thus, the significant differences in the kidney:blood ratios of the radioactivity between [¹³¹I]HML-Fab and [¹²⁵I]MIH-Fab from early postinjection times suggested that [¹³¹I]HML-Fab would liberate meta-[¹³¹I]iodohippuric acid before the radiolabeled Fab fragment was incorporated and transported to the lysosomal compartment in the renal cells.

To further investigate the low renal radioactivity levels of [¹³¹I]HML-Fab from early postinjection times, the subcellular localization of the radioactivity was investigated. Although the radioactivity was observed in both membrane and lysosomal fractions at 10 min postinjection of [¹²⁵I]MIH-Fab and ¹²⁵I-Fab, the majority of the radioactivity had migrated to the lysosomal fraction after 30 min (Fig. 6). This suggested that most of the administered Fab fragments would be transported to and metabolized in the lysosomal compartment after 30 min injection. This was supported by an increase in the renal radioactivity levels from 10 to 30 min after injection of ¹²⁵I-Fab and by the rapid decrease in the renal radioactivity levels of both ¹²⁵I-Fab and [¹²⁵I]MIH-Fab from 30 min to 1 h postinjection. Similar maximum renal radioactivity levels of ¹²⁵I-Fab at 30 min and [¹²⁵I]MPH-Fab at 1 h postinjection also suggested that lysosomal proteolysis of radiolabeled Fab fragments would proceed 30 min after injection. On the other hand, most of the renal radioactivity after administration of [¹³¹I]HML-Fab was observed in the membrane fraction at both 10 and 30 min postinjection. In addition, [¹³¹I]HML-Fab reached its maximum radioactivity levels in the kidney at 10 min postinjection and showed similar kidney:blood radioactivity ratios of ∼1 from 10 min to 3 h. These findings implied that [¹³¹I]HML-Fab released meta-[¹³¹I]iodohippuric acid at the membrane fraction of the renal cells before transported to the lysosomal compartment, presumably by the action of brush border enzymes present on the lumen of the renal tubules.

The ability of HML to reduce the renal radioactivity levels without impairing the target radioactivity levels delivered by the antibody fragment was demonstrated in the biodistribution study in nude mice (Fig. 5). Although both [¹³¹I]HML-Fab and ¹²⁵I-Fab showed similar radioactivity levels in the tumors, [¹³¹I]HML-Fab amplified the tumor:kidney ratio of the radioactivity by a factor of ∼4, compared with simultaneously administered ¹²⁵I-Fab. High stability of the chemical linkage in HML in tumors was also indicated by the unchanged radioactivity levels in the tumor tissue between 24 and 48 h following administration of [¹³¹I]HML-labeled intact antibody (OST7; Ref. 28). Such chemical and biological characteristics render HML useful as a radioiodination reagent for targeted imaging and targeted therapy using LMW antibody fragments as vehicles although suitable chemistries to improve the radiochemical yields of HML-labeled antibody fragments are required. Because the stannyl precursor of HML, SHML, can be radiofluorinated, HML may also be useful to prepare ¹⁸F-labeled LMW antibody fragments and peptides for PET imaging. Although further studies are required, the findings in this study strongly suggested that renal brush border enzymes may constitute key targets for the radiochemical design of LMW antibody fragments and peptides with low renal radioactivity levels using a variety of clinically important radionuclides.

We are grateful to Dr. Isamu Yomoda of Diichi Radioisotope Laboratories (Tokyo, Japan) for providing Na[¹³¹I]I.