Radiolabeled metabolites of proteins play a critical role in radioactivity elimination from the liver

!Vucl.

Pergamon

0969-8051(95)00009-7

Bd Vol. 22, No. 5, op. 5.55 564. IYYS Copyright ,I; 1995Elsevier ScienceLtd Printed in Great Britain. All rights reserved

Med.

096')~8051

>Y5 $9.50 i 0 00

Radiolabeled Metabolites of Proteins Play a Critical Role in Radioactivity Elimination from the Liver YASUSHI ARANO, TAKAHIRO TAKASHI UEZONO, HIROSHI CLAUDIA KAIRIYAMA* Department

of Radiopharmaceutical Chemistry, Yoshida-shimoadachi-cho,

MUKAI, HIROMICHI AKIZAWA, MOTONARI, KOUJI WAKISAKA. and AKIRA YOKOYAMA? Faculty of Pharmaceutical Sciences, Kyoto Sakyo-ku, Kyoto 606-01, Japan

University,

(Acceppted 16 Januur.v 1995) We have recently reported that the behavior of radiolabeled metabolites in the liver appears to be responsible for the hepatic radioactivity levels after administration of protein radiopharmaceuticals. To better understand the role played by radiolabeled metabolites in hepatic radioactivity levels, metabolites, I-(4-isothiocyanatotwo benzyl-EDTA derivatives rendering different radiolabeled benzyl)ethylenediaminetetraacetic acid (SCN-Bz-EDTA) and I-[p-(5-maleimidopentyl)aminobenzyl] ethylenediaminetetraacetic acid (ECMS-Bz-EDTA), were selected as bifunctional chelating agents (BCAs), and “‘In labeling of galactosyl-neoglycoalbumin (NGA) and mannosyl-neoglycoalbumin (NMA) was performed. Biodistribution of radioactivity in mice and subcellular distribution of radioactivity in hepatocytes were then compared. After accumulation in hepatic parenchymal cells, NGA-EMCS-BzEDTA-“‘In rendered a faster elimination rate of radioactivity from the liver than NGA-SCN-Bz-EDTA“‘In. Although each ” ‘In-NMA exhibited a delayed elimination rate of radioactivity from the liver compared to the “‘In-NGA counterpart, NMA-EMCS-Bz-EDTA-“‘In showed faster elimination rate of radioactivity than NMA-SCN-Bz-EDTA-“‘In. Analyses of radioactivity excreted in feces and urine and remaining in the liver indicated that both BCAs rendered mono-amino acid adducts as the major radiolabeled metabolites (cysteine-EMCS-Bz-EDTA-“‘In and lysine-SCN-Bz-EDTA-“‘In), which were generated in both cell types of the liver within 1 h postinjection. Subcellular distribution of radioactivity indicated that the radioactivity was copurified with lysosomes. These results demonstrate that although in ho stability of radiometal chelates is essential, the biological properties of the radiolabeled metabolites generated after lysosomal proteolysis in hepatocytes play a critical role in radioactivity elimination from the livet

Introduction

gation with galactosyl-neoglycoalbumin (NGA) and mannosyl-neoglycoalbumin (NMA) (Arano et al., 1994a, b). Both NGA and NMA are known to be incorporated by hepatic parenchymal and nonparenchymal cells, respectively, via receptor-mediated endocytosis (Ashwell and Harford, 1982; Shen, 1985; Stockert and Morel], 1983). After binding at the surface, both proteins are rapidly internalized via coated pits and then transported intracellularly to the lysosomal compartment. Since lysosomes are the principal sites of intracellular digestions of proteins (Alberts et a/., 1983; Santoro et al., 1993) this biological method using NGA and NMA can he applied to pursue the fate of radiolabels after lysosomal proteolysis in hepatic parenchymal and nonparenchymal cells without being affected by transchelation and redistribution of radiolabels in

In radioimmunoimaging and therapy using monoclonal antibodies (MAbs) labeled with metallic radio-

nuclides, high localization of radioactivity in the liver constitutes a serious problem. Various mechanisms, such as instability of radiolabels in vivo and inherent behaviors of MAbs themselves, have been reported to be responsible for the undesirable radioactivity localization (Adams et al., 1989; Beatty et al., 1989; Brechbiel et al., 1986; Cole et al., 1987; Esteban et al., 1987; Khaw et al., 1991; Schwarz et al., 1991). Recently, we have shown that hepatic behavior of radioactivity derived from each bifunctional chelating agent (BCA) can be evaluated through conju-

*Present address: Instituto de Investigaciones, Fundacion Campomar, Av. Patricia Argenhnas 435, Buenos Aires 1405, Argentina. tAuthor for correspondence.

plasma

and to the liver,

respectively.

Our previous study indicated that when isothiocyanatobenzyl)ethylenediaminetetraacetic 55s

l-(4acid

Yasushi Arano PI (I/

556

~NvN”+JT/--N HOOC

EMCS-Bz-EDTA

NGA-.

> COOH

scNyJT~~~~ HOOC

> cook’

SCN-Bz-EDTA

NGA-. NMA-SCN-Br-EDTA-In

Fig. 1. Chemical structures of l-[p-(5-maleimidopentyl)aminobenzyl]ethylenediamine~e~raace~ic acid (EMCS-Bz-EDTA), 1-(4-isothiocyanatobenzyl)ethylenediaminetetraacetic acid (SCN-Bz-EDTA), and their neoglycoalbumin conjugates. While SCN-Bz-EDTA covalently attaches to lysine residues, EMCSBz-EDTA binds to cysteine residue of neoglycoalbumin molecules. is used as a BCA for “‘In labelling Bz-EDTA-“‘In (Arano et al., 1994b). Since both is EMCS-Bz-EDTA and SCN-Bz-EDTA possess a of NGA and NMA, lysine-SCN-Bz-EDTA-“‘In generated as a final radiolabeled metabolite in both similar chelating site (Bz-EDTA moiety) for “‘In, hepatic parenchymal and nonparenchymal cells comparative radioactivity distribution of “‘In-lawithin 1 h postinjection (Arano et al., 1994b). While beled neoglycoalbumins derived from each BCA the metabolite is gradually eliminated from parenchy- would give insight into the importance of biological ma1cells via hepatobiliary excretion, it is retained in characteristics of radiolabeled metabolites in eliminonparenchymal cells for longer postinjection nating radioactivity from each liver cell type. periods. Furthermore, the elimination rate of this radiolabeled metabolite is much slower than those of radioiodinated NGA and NMA. Using a similar Materials and Methods approach, metabolic studies of “‘In-DTPA-labeled Preparation of NGA-EMCS-Bz-EDTA and NMANGA and NMA have indicated that lysine-DTPA“‘In is the final radiolabeled metabolite in both cell EMCS-Bz-EDTA types of the liver, and the metabolite is retained in the NGA and NMA were prepared by the reaction lysosomal fraction of murine liver for much longer of human serum albumin (HSA; A-3782, Sigma postinjection time (Arano et al., 1994a; Duncan and Chemical Co., St Louis. MO) with cyanomethylWelch, 1993). These results suggest that radiolabeled 2,3,4,6-tetra-O-acetyl-thio-/3-o-galactopyranoside metabolites derived from each BCA would play a and cyanomethyl-2,3,4,6-tetra-O-acetyl-thio-B-Dsignificant role in eliminating radioactivity from the mannopyranoside, respectively, according to the proliver after injection of radiolabeled proteins. cedures of Stowell and Lee (1980). When determined In the present study, to further estimate the role with the phenol-sulionic acid reaction (Dubois et al., played by radiolabeled metabolites in hepatic radio- 1956), 44 galactose and 31 mannose were attached to activity levels, 1-[p(5-maleimidopentyl)aminobenzyl] each HSA molecule for NGA and NMA, respectethylenediaminetetraacetic acid (EMCS-Bz-EDTA) ively. and SCN-Bz-EDTA were used as BCAs to label NGA-EMCS-Bz-EDTA was prepared by reducing NGA and NMA with “‘In. Biodistribution of radio- NGA with 5 molar excessof dithiothreitol (DTT) to activity in mice and subcellular distribution of radio- expose one thiol group (Sogami et al., 1984),followed activity in murine hepatocytes were compared by the conjugation reaction using maleimide-thiol following injection of “‘In-labeled NGA and NMA chemistry. Freshly prepared 0.5 mL DTT using each BCA. Radiolabeled metabolites derived (0.45 mg/mL) in 0.1 M phosphate buffer (PB, pH 6.8) from EMCS-Bz-EDTA in both hepatic parenchymal containing 0.3 M NaCl and 10mM EDTA was added and nonparenchymal cells were investigated. The to 0.5 mL of NGA (40 mg/mL) in the same buffer. chemical structures of the two BCAs and their pro- After incubation of the mixture for 30 min at room tein conjugates are illustrated in Fig. 1. Deshpande et temperature, excess DTT was removed by the al. (1990) reported high in uivo stability of Bz-EDTADiaflow system (8 MC model, Amicon Grace, Tokyo, “‘In chelate using a monoclonal antibody that recog- Japan) with 20 vol of well degassed0.1 M PB (pH 6.0) nizes Bz-EDTA-In chelate. This result was supported containing 10 mM EDTA before adjusting the proby our previous study using NGA- and NMA-SCNtein concentration to 10 mg/mL. A small portion of (SCN-Bz-EDTA)

Hepatic behavior of radiolabeled

metabolites

Table 1. Radiochemical purities of “‘In-labeled

NGA

and NMA

Radiochemical purity (%)$ Before EDTA chase NGA-EMCS-Bz-EDTA-“‘In NGA-SCN-Bz-EDTA-“‘In NMA-EMCS-Bz-EDTA-“‘In NMA-SCN-Bz-EDTA-“‘In $Radiochemical

LIVER loo f s-C= -t-

EMCS SCN

99.3 94. I 98.5

99.5

99.3

100

chase

by cellulose acetate electrophoresis.

Preparation of NGA-SCN-Bz-EDTA SCN-Bz-EDTA

and NMA-

NGA-SCN-Bz-EDTA and NMA-SCN-Bz-EDTA were prepared according to the procedure as described previously (Arano et al., 1994b). Briefly, to a solution of NGA or NMA (10 mg/mL) in borate-buffered saline (0.05 M, pH 8.5), a 10 molar excess of SCN-Bz-EDTA (Dojindo Labs) in dimethylformamide (7.3 mg/mL) was added. After incubating for 20 h at 37”C, each conjugate was purified by Sephadex G-50 column chromatography (1.8 x 40cm), equilibrated and eluted with 0.1 M acetate buffer (pH 3.0). The respective conjugate fractions were collected and subsequently concentrated to 5 mg/mL by ultrafiltration (8 MC model).

INTESTINE

loor

0 EMCS

q SCN

0

0. 0

6 12 18 24 Time after Injection (h)

8-

99.4 93.9 9x.4

purities were determined

this mixture was sampled, and the number of exposed thiol groups was estimated with 2,2’-dithiodipyridine (Grassetti and Murray, 1967). To the freshly thiolated NGA (2 mL, lOmg/mL), a 5 molar excess of EMCS-Bz-EDTA (Dojindo Labs, Kumamoto, Japan) in dimethylformamide (85 pL, 10 mg/mL) was added, and the reaction mixture was stirred for 2 h. NGA-EMCS-Bz-EDTA was purified by Sephadex G-50 (Pharmacia Biotech Co. Ltd., Tokyo, Japan) column chromatography (1.8 x 40 cm), equilibrated and eluted with 0.1 M acetate buffer (pH 3.0). Each protein fraction was collected and concentrated to 5 mg/mL with the Diaflow system (8 MC model). NMA-EMCS-Bz-EDTA was prepared according to procedures similar to those described above, using NMA in place of NGA.

After EDTA

BLOOD

0 6 12 18 24 Time alter Injection (h) 1.5

URINE*FECES ‘24 h post-injectlon

KIDNEY

EMCS : NGA-EMUS-Bz-EDTA-“‘In SCN : NGA-SCN-Bz-EDTA-“‘in

om * m -a 0 6 12 18 24 Tlme atter lnjectlon (h)

0.0

0 6 12 18 24 Time after Injection (h)

Fig. 2. Comparative biodistribution of radioactivity in mice after i.v. administration of “‘In-labeled NGAs using EMCS-Bz-EDTA and SCN-Bz-EDTA as BCAs. Statistically significant differences in hepatic radioactivity levels were observed between the two “‘In-labeled NGAs (P < 0.001 at 1 h, P < 0.02 at 3 h, P < 0.01 at 6 h and P =z0.001 at 24 h postinjection).

Yasushi Arano PI nl.

IOO-

INTESTINE

loor

cl EMCS R SCN

80

0

8

12

18

24

0

6

12

18

Tlme atter lnjectlon (h)

BLOOD

KIDNEY

8

URINtiFECES

24

Time after lnjectlon (h)

‘24 h post-lnjectbn

6 !i

EMCS : NMA.EYCSBz-EDTA-“‘In SCN : NMA-SCN-Bz-EDTA-“‘In

z4 H $2

0 0

6

12

18

24

Tlme after lnjectlon (h)

0

6

12

18

24

Tlme aller lnjectlon (h)

Fig. 3. Comparative biodistribution of radioactivity in mice after i.v. administration of “‘In-labeled NMAs using EMCS-Bz-EDTA and SCN-Bz-EDTA as BCAs. Statistically significant differences in hepatic radioactivity levels were observed between the two “‘In-labeled NMAs (P < 0.0002 at 1 h, P i 0.02 at 3 h, P < 0.002 at 6 h and P < 0.0001 at 24 h postinjection). Cysteine-EMCS-Bz-EDTA was synthesized by reacting L-cysteine (Nacalai Tesque, Kyoto, Japan) with EMCS-Bz-EDTA at a molar ratio of 3: 1 in 0.1 M acetate buffer (pH 6.0) under a nitrogen atmosphere, and purified by reversed-phase column chromatography (Lober column, Size B, Merck), equilibrated and eluted with a mixture of methanol and water (2:98). This compound showed a molecular weight of 712 (MH+ ) by FAB mass spectrometry (JMS-HX/HXllOA model; JEOL Ltd., Tokyo, Japan). “‘h

labeling of neoglycoalbumins

All the neoglycoalbumins were labeled with “‘In according to the procedure described previously (Arano et al., 1994b). In brief, l”InCl, (2mCi/mL, 20-100 pL) in 0.01 N HCl (Nihon Medi-Physics, Takarazuka, Japan) was added to each protein solution (5 mg/mL, 100 pL) in 0.1 M acetate buffer (pH 3.0). After gentle agitation of the reaction mixture for 1 h at room temperature, 10 mM solution of EDTA in 0.1 M acetate buffer (pH 3.0) was added to reach 100 molar excess of EDTA for each protein molecule. After incubating at 1 h at room temperature, “‘In-labeled proteins were purified by spin

column chromatography (Sephadex G-50, fine), equilibrated and eluted with 0.1 M acetate buffer (pH 3.0) (Meares et al., 1984). Radiochemical purities of the four “‘In-labeled neoglycoalbumins were determined by size-exclusion HPLC (5 Diol-120, 7.5 x 600 mm, Nacalai Tesque), equilibrated and eluted with 0.1 M PB (pH 6.8) at a flow rate of 1 mL/min, and thin layer chromatography (TLC; Merck Art. 5553), developed with 10% ammonium chloride-to-methanol (1: 1). Radiochemical purities were also analyzed by cellulose acetate electrophoresis (CAE) run at an electrostatic field of 0.8 mA/cm for 40 min in Verona1 buffer (I = 0.05, pH 8.6). To further estimate the radiochemical purities of “‘In-labeled neoglycoalbumins, 10 mM solution of EDTA in 0.1 M acetate buffer (pH 3.0) was mixed with “‘In-labeled protein to reach 100 molar excess of EDTA for each protein molecule. After 1 h incubation at room temperature, radiochemical purities of each “‘In-labeled protein fraction were analyzed by CAE. In vivo studies Animal studies were performed in compliance with generally accepted guidelines governing such work.

Hepatic behavior of radiolabeled metabolites 5ooa 7

4ooo-

3ooo -

2000 -

1000 -

0

0

10 Retention

20 Time

30

40

(min)

Fig. 4. Size-exclusion HPLC radioactivity profiles of liver supernatant at 1 h postinjection of NGA-EMCS-Bz-EDTA“‘In (0) and NMA-EMCS-Bz-EDTA-“‘In (0). Both supernatants showed a single peak at 23 min. The parental NGA- and NMA-EMCS-Bz-EDTA-“‘In (dotted line) were eluted at 14min. The protein concentrations of each “‘In-labeled NGA and NMA were adjusted to 90pg/mL with 0.1 M phosphate-buffered saline (PBS, pH 6.0). Biodistribution studies were performed by the i.v. administration of each “‘In-labeled protein to 6week-old male ddY mice weighing 27-30g (Imai et al.. 1986). Groups of five mice each were administered 9 pg (0.3-0.5 ,nCi) of the respective “‘In-labeled protein before sacrifice at 10, 30 min and I, 3, 6, and 24 h postinjection by decapitation. Tissues of interest were removed, weighed and the radioactivity counts were determined (ARC 2000, Aloka, Tokyo, Japan). To assess the cellular localization of radioactivity in murine liver cells, collagenase digestion of hepatocytes was performed as described previously (Arano et crl., 199413).Briefly, parenchymal and nonparenchyma1 cells were fractionated by collagenase digestion in situ from the portal vein with Type I collagenase (C-0130, Sigma) dissolved in an EDTA-free solution at 15 min postinjection of each “‘In-labeled neoglycoalbumin (Seglen, 1973). Radiolabeled species in the liver at 1 h postinjection of NGA-EMCS-Bz-EDTA-‘“In and at 1 and 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In were analyzed according to the procedure described previously (Arano et al., 1994b). Briefly, each protein (5.-6pCi/9 ilg protein) was administered iv. to 6week-old ddY mice, and the liver was perfused in situ with cold 0.1 M Tris-citrate buffer (pH 6.5) containing 0.15 M NaCl, 0.02% sodium azide, 1 TIU/mL aprotinin, 2 mM benzamide-HCI, 2 mM iodoac&amide, 1 mM phenylmethylsulfonyl fluoride and 5 mM diisopropyl fluorophosphate before hepatic samples of 1 g each were isolated. Each tissue sample was placed in a test tube and subjected to three cycles of freezing (dry ice-acetone) and thawing. After

559

addition of the same buffer containing an additional 3.5mM of fi-octyl-glucoside, the hepatic sample was homogenized with a Polytron homogenizer (PT 10-35. Kinematica GmgH Littau, Switzerland) set at full speed with three consecutive 30-s bursts prior to centrifugation at 48,000g for 20 min at 4°C (Himac CS-120 Centrifuge, Hitachi Co. Ltd., Tokyo, Japan). The supernatants were separated from the pellets, and the radioactivity was counted. Similarly, feces at 24 h postinjection of NGA-EMCS-Bz-EDTA-]“In was homogenized in the presence of 0.1 M PBS (pH 6.0) before centrifugation at 10,OOOg for 20 min at 4°C. The liver supernatants were analyzed by size-exclusion HPLC after filtration through a polycarbonate membrane with a pore diameter of 0.22 pm (Myrex, Millipore Ltd, Tokyo, Japan). The liver, feces supernatants and urine samples at 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In were analyzed by TLC and CAE without any pretreatment. Each sample was also analyzed by reversedphase HPLC (RP-HPLC) after filtering through a 10 kDa cut-off ultrafiltration membrane (Millipore Ltd.). The RP-HPLC column (Cosmosil 5C,*-AR, 4.6 x 250 mm, Nacalai Tesque) was eluted with a mixture of methanol and 20% aqueous ammonium acetate (1: 3) at a flow rate of 1 mL/min. Analyses were also performed by co-chromatography with cysteine-EMCS-Bz-EDTA-“‘In. Subcellular distributions of radioactivity in the murine liver were investigated by perfusing the liver in situ with cold 0.25 M sucrose buffered with 10 mM phosphate (pH 7.4) at 1 h postinjection of each “‘Inlabeled NGA and at I, 3, and 24 h postinjection of each “‘In-labeled NMA. The liver was treated according to the procedure of Yamada et al. (1984) with some modification. In brief, the liver was suspended in 4~01 of 0.25 M sucrose and homogenized with a Dounce homogenizer by hand (20 strokes), followed by two final strokes in an ice-cooled Potter-Elvehjem homogenizer at 800 rpm. The 20% homogenate was centrifuged twice for 5 min each time at 340g at 4 C. The isolated supernatant (0.45 mL) was layered onto iso-osmotic (0.25 M sucrose) 37.5% Percoll (9 mL; Pharmacia Biotech Co. Ltd., Tokyo) at a density of 1.08 g/mL. After centrifugation at 20,OOOg (RP 30 Rotor; Hitachi Co. Ltd., Tokyo) for 90min at 4 C, the gradients (1.02-1.14 g/mL) were collected in 14 fractions before analysis of the B-galactosidase activity (Wallner and Walker, 1975), density and radioactivity counts of each fraction. The liver supernatant at 24 h postinjection of NMA-EMCS-Bz-EDTA-‘“In was also incubated at 37’C for 5 min in the presence of 1 mM CaClz to swell mitochondria before the Percoll density gradient centrifugation. This treatment leads to a marked shift of the mitochondria to a lighter density (1.08 g/mL) (Yamada et al., 1984). Statistical anai,vses Biological data were analyzed statistically Student’s unpaired t-test.

using

Yasushi Arano et a/.

560

TLC

CAE

RP-HPLC

(A)

1000 r

5000 r

,r" 2 ij

600

800

"0 600

5 l-a 400 cc

-6 Retention

time (min)

-4

-2

Distance

0

2

4

from origin

6 (cm)

Rf

(8) 8000 r 6ooo I

? 6000

-6 Retention

time (min)

Fig. 5. Chromatographic

-4

-2

Distance

0

2

from origin

4

6 (cm)

0

0.2

0.4

0.6

0.8

1.0

Rf

analyses of liver (A) and feces (B) supernatants at 1 and 24 h postinjection

of

NGA-EMCS-Bz-EDTA-“‘In,

respectively. Each sample was analyzed by RP-HPLC, CAE, and TLC in the absence (0) and presence (0) of cysteine-EMCS-Bz-EDTA-“‘In. Under these analytical conditions, cystine-EMCS-Bz-EDTA-“‘In was detected at retention time of 6-7 min on RP-HPLC, a migration distance of l-2cm

anode from the origin on CEA, and an R, value of ca 0.6 on TLC.

Results Radiochemical purities of all the “‘In-labeled proteins exceeded 93% before and after EDTA chase (Table 1). Collagenase digestion of liver cells at 15 min postinjection of each “‘-In-labeled-NGA and “‘In-labeled-NMA indicated that more than 83% of the injected radioactivity were found in the respective cell types. Biodistributions of radioactivity after i.v. injection of “‘In-NGA using either EMCS-Bz-EDTA or SCN Bz-EDTA as a BCA are illustrated in Fig. 2. While both “‘In-labeled NGAs accumulated in the liver almost quantitatively upon injection, NGAEMCS-Bz-EDTA-“‘In demonstrated faster elimination rate of radioactivity from the liver than NGA-SCN-Bz-EDTA-“‘In. At 24 h postinjection, 2.8 and 6.4% of the injected radioactivity remained in the liver following injection of NGA-EMCS-BzEDTA-“‘In and NGA-SCN-Bz-EDTA-“‘In, respectively. During the same postinjection interval, over 80% of the injected radioactivity was excreted in

the feces with both “‘In-labeled NGAs (Fig. 2). Biodistributions of radioactivity after i.v. injection of the two “‘In-labeled NMAs are shown in Fig. 3. Both “‘In-labeled NMAs manifested slower elimination rates of radioactivity from the liver than the corresponding “‘In-labeled NGAs. When the two “‘In-labeled NMAs were compared, NMA-EMCSBz-EDTA-“‘In showed much faster elimination rate of radioactivity from the liver than NMA-SCNBz-EDTA-“‘In. This was reflected in the higher amount of radioactivity excreted in the urine following injection of NMA-EMCS-Bz-EDTA-‘“In (Fig. 3). Extraction of radioactivity from the liver homogenates at I h postinjection of NGA-EMCS-BzEDTA-“‘In and at 1, 3, and 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In was performed with efficiencies of over 9 1%. Size-exclusion HPLC of liver supernatants at 1 h postinjection of NGA-EMCS-BzEDTA-“‘In and NMA-EMCS-Bz-EDTA-“‘In registered a single radioactivity peak at a retention time of 23 min (Fig. 4). Under similar conditions, the parent

Hepatic behavior of radiolabeled metabolites

(4

RP-HPLC

CAE 2000

8000 r

0

IO

Retention

20

2

TLC

r

30

-6

time (min)

(B) 5000 x .= ,2mo

561

2000

-4 -2

Distance

0

2

4

from origin

6

r

-0

0.2

0.4

(cm)

0.6

0.8

1.0

Rt

1000

4000

x

800

3000

.z 'g

600

2000

.yrj Fi 400

'2

2 1000

200

0

0 0

102030

Retention

0

IO

Retention

0 -6 -4 -2

time (min)

20

30

time (min)

Distance

-6

0

4

from origin

-4 -2

Distance

2

0

2

4

from origin

6

0

0.2

0.4

0.6

0.8

1.0

0

0.2

0.4

0.6

0.8

1.0

(cm)

6

(cm)

Uf

Fig. 6. Chromatographic analyses of liver supernatants at 1 (A) and 3 h (B) and urine samples (C) at 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In. Each supernatant was analyzed by RP-HPLC, CAE, and TLC in the absence (0) and presence (0) of cysteine-EMCS-Bz-EDTA-“‘In. proteins (NGA- and NMA-EMCS-Bz-EDTA-“‘In) were eluted at the retention time of 14min. The RP-HPLC, CAE and TLC radioactivity profiles of liver supernatants at 1 h postinjection of NGAEMCS-Bz-EDTA-“‘In are illustrated in Fig. 5(A). Under similar analytical conditions, cysteine-EMCSBz-EDTA-“‘In had a retention time of 6-7min on RP-HPLC, a migration distance of l-2cm anode from the origin on CEA, and an R, value of ca 0.6 on TLC. The liver supernatant showed a single radioactivity peak in each analysis, and each radioactivity peak correlated well with cysteine-EMCS-Bz-EDTA“‘In, as confirmed by co-chromatographic analyses. Radioactivity profiles of feces supernatants at 24 h postinjection of NGA-EMCS-Bz-EDTA-“‘In are shown in Fig. 5(B). The feces supernatant also exhibited a major radioactivity peak, which had similar

retention time, migration distance, and R, value to those of cysteine-EMCS-Bz-EDTA-“‘In. Chromatographic analyses of liver supernatants at 1 and 3 h and urine samples at 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In are represented in Fig. 6(A), (B) and (C), respectively. All the samples demonstrated a major radioactivity peak at the same retention time, migration distance and R, value to those of cysteine-EMCS-Bz-EDTA-“‘In, which was also confirmed by cochromatographic analyses. The Percoll density gradient centrifugation profiles of radioactivity in the liver at 1 h postinjection of NGA-EMCS-Bz-EDTA-“‘In and 1, 3 and 24 h postinjection of NMA-EMCS-Bz-EDTA-“‘In are shown in Fig. 7. Each liver homogenate demonstrated a single radioactivity peak at a density of ca 1.10 g/mL, which correlated well with each

1.14

1.14

* (A)

r (W

.z

‘3 0

1.12

1.121

m g

f b

E* 5 .fl ZL g .-5 p"

1.08 1.06

1.10, 1.08' 1.06'

'; .-iz

1.04'

az

1.02 0

Fraction Number

2 a 0

1.1

E 2

1.0

5 s -2 .z

1.0 1.0

:: e 2 5

10

I 10

1 15

Fraction number $ > 'L :

F & I+"

I 5

1.141.12 E

l.lO-

3

1.08-

* 5 5 0

-

1.061.041.02

-

l.OO-

15

Fraction Number

Fraction Number

Fig. 7. Percoll density gradient centrifugation profiles of liver supernatant at I h postinjection of NGA-EMCS-Bz-EDTA-“‘In (A) and 1 (B), 3 (C) and 24 h (D) postinjection of NMA-EMCS-Bz-EDTA“‘In. In each analysis, a single radioactivity peak (0) that coincided with fi-galactosidase activity (A) was detected at a density (0) of ca I .Og/mL.

p-galactosidase activity profile. The Percoll density gradient centrifugation of liver supernatant pretreated with 1 mM of CaCI, indicated that the radioactivity profile was still observed at similar density (ca I .lO g/mL) and was correlated well with the b-galactosidase activity profile (data not shown).

Discussion In biodistribution studies of “‘In labeled NGA and NMA, significantly faster elimination rates of radioactivity from the liver were observed when EMCSBz-EDTA was used as a BCA (Figs 2 and 3). All the “‘In-labeled NGAs and NMAs had radiochemical purities of higher than 93% (Table l), and both NGA and NMA indicated specific accumulation in the corresponding cell types of the liver after injection. Therefore, the different elimination rates of radioactivity observed for NGA-EMCS-Bz-EDTA-“‘In and NGA-SCN-Bz-EDTA-“‘In and for NMA-EMCSBz-EDTA-“‘In and NMA-SCN-Bz-EDTA-“‘In would be attributed to the nature of each BCA. Analyses of radiolabeled species following injection of NGA- and NMA-EMCS-Bz-EDTA-“‘In strongly

suggested that the final radiolabeled metabolite was generated within 1 h postinjection in both cell types of the liver, and that the metabolite was most likely to be cysteine-EMCS-Bz-EDTA-“‘In, as shown in Figs 4-6. Analyses of feces and urine after 24 h injection of NGA- and NMA-EMCS-Bz-EDTA-“‘In indicated that the final radiolabeled metabolite was excreted without further metabolism (Figs 4 and 5). In our previous studies of “‘In-labeled NGA and NMA using SCN-Bz-EDTA as a BCA, lysine-SCNBz-EDTA-“‘In was generated as the final radiolabeled metabolite in both ccl1 types of the liver within 1 h postinjection (Arano et al., 1994b). Thus, it is clear that both EMCS-Bz-EDTA and SCN-BzEDTA render “‘In chelates of high in vim stability, and the different radioactivity levels caused by the two BCAs are not due to differences in stability of the “‘In chelates. Furthermore, our recent study using radioiodinated NGA prepared by the N-succinimidyl iodobenzoate (ATE) method indicated that only 7% of the injected radioactivity remained in the murine liver at 1 h postinjection (Arano et al., 1994c). ATE is covalently bound to lysine residues of NGA at random, and the final radiolabeled metabolite derived from ATE-conjugated protein is reported to be

Hepatic behavior of radiolabeled

the lysine adduct of iodobenzoic acid (Garg et al., 1994). Thus, it is most likely that the metabolism of neoglycoalbumins in the liver is very fast, and that the radiolabeled metabolites are metabolized to monoamino acid adducts before excretion from hepatocytes. Since a large difference in hepatic radioactivity levels was observed between the two “‘In-NGAs and between the two “‘In-NMAs up to 24 h postinjection (Figs 2 and 3) the in uiuo behaviors of each radiolabeled metabolite (cysteine-EMCS-Bz-EDTA-“‘In and lysine-SCN-Bz-EDTA-“‘ln) are apparently responsible for the different radioactivity levels in the liver. Percoll density gradient centrifugation after injection of NGA- and NMA-EMCS-Bz-EDTA-“‘In demonstrated that the radioactivity in the liver was copurified with lysosomes (Fig 7) which suggests that neither intracellular transport of the radiolabeled metabolite, cystine-EMCS-Bz-EDTA-“‘In, to other organelles nor interaction of the metabolite with biomolecules present in the cytoplasm would take place. Since similar results were observed following administration of NGA- and NMA-SCN-Bz-EDTA“‘In (Arano et al., 1994b), the different elimination rates of radioactivity from the liver for the two BCAs would be due to the different elimination rates of each radiolabeled metabolite from the lysosomal fraction. Such differences would be attributed to the different biological characteristics of each radiolabeled metabolite, arising from different physicochemical properties such as the molecular weight (822 Da for cysteine-EMCS-Bz-EDTA-“‘In and 696 Da for lysine-SCN-Bz-EDTA-“‘In) and lipophilicity, which were reflected in the RP-HPLC analyses (cysteineEMCS-Bz-EDTA-“‘In and Iysine-SCN-Bz-EDTA“‘In have similar retention times when eluted with a mixture of methanol/20% aqueous ammonium acetate of I : 3 and 1: IO, respectively) (Arano et al., 1994b). The hepatobiliary excretion of radiolabeled metabolites following injection of NGA-EMCS-BzEDTA-“‘In and NGA-SCN-Bz-EDTA-“‘In (Fig. 2) could be attributed to the bile secretion characteristics of hepatic parenchymal cells and hepatobiliary excretion characteristics of the two radiolabeled metabolites (Firnau, 1976). Recently, Kinuya et ul. (1994) reported that use of EMCS-Bz-EDTA as a BCA for “‘In labeling of an antibody renders lower levels of radioactivity in the murine liver when compared with “‘In-labeled antibody using SCN-Bz-EDTA as a BCA. Since the catabolism of radiolabeled antibodies in the liver has been reported to be rapid (Himmelsbach and Wahl, 1989; Motta-Hennessy et al., 1990; Paik et al., 1992) this result would be well explained by the different elimination rates of each radiolabeled metabolite derived from the two BCAs from the liver, as discussed above. This result also reinforces the usefulness of the biological method using NGA and NMA to evaluate BCAs for protein radiopharmaceuticals labeled with metallic radionuclides.

metabolites

563

In conclusion, the present study indicated that differences in the physicochemical properties of radiolabeled metabolites derived from BCAs would be responsible for the different elimination rates of radioactivity from the liver after injection of protein radiopharmaceuticals utilizing metallic radionuclides. This study also demonstrated that although i% tjirlo stability of the radiometal chelate is essential. the biological characteristics of radiolabeled metabolites generated after lysosomal proteolysis should also be taken into consideration for the design of new BCAs for protein radiopharmaceuticals to reduce the hepatic radioactivity levels. authors thank Dr N. Akimoto (Faculty of Pharmaceutical Sciences,Kyoto University) for taking FAB mass spectra. The authors also thank Nihon Medi-Phvsics Co. Ltd. Takarazuka. Janan and Doiindo Labs., K;mamoto, Japan for their kind gjfts of “‘b&Z< and EMCS-Bz-EDTA, respectively. This study was supported in part by a Grant-in-Aid for Developing Scientific Research (No. 05151036)from the Ministry of Education, Scienceand Culture of Japan.

Acknowledgements-The

References Adams G. P., DeNardo S. J., Deshpande S. V., DeNardo G. L., Meares C. F., McCall M. J. and Epstein A. L. (1989) Effect of mass of “‘In-benzyl-EDTA monoclonal antibody on hepatic uptake Cancer Res. 49, 1707-l 711.

and processing

in mice.

Alberts B., Bray D., Lewis J., Raff M., Roberts K and Watson J. D. (1983) Molecular Biology of The Cell, pp. 307-310. Garland Publishing Inc., New York. Arano Y., Mukai T.. Uezono T., Motonari H., Wakisaka K. and Yokoyama A. (1994a) Biological comparison of DTPA and SCN-benzyl-EDTA as chelating agents for indium labeling of antibodies. J. LubeNed Compd Radiopharm. 35, 381-383. Arano Y., Mukai T., Uezono T., Wakisaka K., Motonari H., Akizawa H., Taoka Y. and Yokoyama A. (1994b) A biological method to evaluate chelating agents to label antibodies with metallic radionuclides. /. Nurl. Med. 35, 890-898. Arano Y., Wakisaka K., Ohmomo Y., Mukai T.. Uezono T., Motonari H., Sakahara H., Konishi J.. Tanaka C. and Yokoyama A. (1994c) Maleimidoethyl 3-(tri-n-butylstannyl)hippurate: a useful radioiodination reagent for protein radiopharmaceuticals to enhance target selective radioactivity localization. J. Med. Chem. 37, 2609. 2618. Ashwell G. and Harford J. (1982) Carbohydrate-specific receptors of the liver. Ann. Rec. Biochem. 51, 531L554. Beattv B. G.. Beattv J. D.. Williams L. E.. Paxton R. J.. Sh&ely J. E. and’O’Connor-Tressel M. (1989) Effect of specific antibody pretreatment on liver uptake of “‘In-labeled anticarcinoembryonic antigen monoclonal antibody in nude mice bearing human colon cancer xenografts. Cancer Res. 49, 1587-1594. Brechbiel M. W., Gansow 0. A., Atcher R. W., Schlom J.. Esteban J., Simpson D. E and Colcher D. (1986) Synthesis of I-(p-isothiocyanatobenzyl) derivatives of DTPA and EDTA. Antibody labeling and tumor-imaging study. Inorg. Chem. 25, 2772--2781. Cole W. C., DeNardo S. J., Meares C. F., McCall M. J.. DeNardo G. L., Epstein A. L., O’Brien H. A. and Moi M. K. (1987) Comparative serum stability of radiochelates for antibody radiopharmaceuticals. .I. Nucl. Med. 28, 83-90.

i64

Yasushi Arano e/ rrl

Deshpande S. V., Subramanian R., McCall M. J., DeNardo S. J., DeNardo G. L. and Meares C. F. (1990) Metabolism of indium chelate attached to monoclonal antibody: minimal transchelation of indium from benzyl-EDTA chelate in niro. /. Nucl. Med. 31, 218-224. Dubois M., Gilles K. A., Hamilton J. K., Rebers P. A. and Smith F. (1956) Calorimetric method for determination of sugars and related substances. Anal. Chem. 28, 350-356. Duncan J. R. and Welch M. J. (1993) Intracellular metabolism of indium- I I I -DTPA-labeled receptor targeted proteins. J. Nucl. Med. 34, 1728-1738. Esteban J. M., Schlom J., Gansow 0. A., Atcher R. W., Brechbiel M. W., Simpson D. E. and Colcher D. (1987) New method for the chelation of indium-1 I I to monoclonal antibodies: biodistribution and imaging of athymic mice bearing human colon carcinoma xenografts. J. Nucl. Med. 28, 86ll870. Firnau G. (1976) Why do Tc-99m chelates work for cholescintigraphy? Eur. J. Nucl. Med. 1, 137-139. Garg P. K., Garg S., Zhao X. G., Welsh P. C. and Zalutsky M. R. (1994) Catabolism of label from MEL-F(ab’), fragment radiohalogenated using N-succinimidyl 3halobenzoate. J. Lahelled Compd Radiopharm. 35, 310-312. Grassetti D. R. and Murray J. F. M Jr. (1967) Determination of sulfhydryl groups with 2,2’- or 4,4’dithiodipyridine. Archs Biochem. Biophys. 119, 41-49. Himmelsbach M. and Wahl R. L. (1989) Studies on the metabolic fate of “‘In-labeled antibodies. Nucl. Med. Biol. 16, 839-845. Imai S., Morimoto J., Tsubura T., Esaki K., Michalides R., Holmes R. S., Deimling 0. and Hilgers J. (1986) Genetic marker patterns and endogenous mammary tumor virus genes in inbred mouse strains in Japan. Exp. Anim. 35, 263-273. Khaw B. A., Kilbanov A., O’Donnell S. M., Saito T., Nossiff N., Slinkin M. A., Newell J. B., Strauss H. W. and Torchilin V. P. (1991) Gamma imaging with negatively charge-modified monoclonal antibody: modification with synthetic polymers. J. Nucl. Med. 32, 1742.-1751. Kinuya S., Yokoyama K., Shuke N., Sun B.-F., Miyauchi T., Tonami N. and Hisada K. (1994) Mechanism of decreased hepatic activity of In-l I I monoclonal antibody labeled with chelate with hydrocarbon spacer: difference from diester spacer. J. Nucl. Med. 35, 29P (abstract).

Meares C. F., McCall M. J., Reardan D. T., Goodwin D. A., Diamanti C. I. and McTigue M. (1984) Conjugation of antibodies with bifunctional chelating agents: isothiocyanate and bromoacetamide reagents, methods of analysis, and subsequent addition to metal ions. .4&. Bioc~hcrn. 142, 68 78. Motta-Hennessy C., Sharkey R. M. and Goldenberg D. M. (1990) Metabolism of indiun- I I I -labeled murine monoclonal antibody in tumor and normal tissues of the athymic mice. j, Nucl. Med. 31, 1510-1519. Paik C. H., Sood V. K., Le N., Cioloca L., Carrasquillo J. A., Reynolds J. C., Neumann R. D. and Reba R. C. (1992) Radiolabeled products in rat liver and serum after administration of antibody-amide-DTPA-indiumI I I. Nucl. Med. Biol. 19. 517-522. Santoro L.. Rebout A:, Journet A. M. and Colomb M. G. (1993) Major involvement of cathepsin B in the intracellular proteolic processing of endogenous IgGs in U937 cells. Mol. Zmmunol. 30, 103331039. Schwarz S. W., Mathias C. J., Sun J. Y., Dilley W. G., Wells S. A., Martell A. E. and Welch M. J. (1991) Evaluation of two new bifunctional chelates for radiolabeling a parathyroid-specific monoclonal antibody with In-l 1 I. Nucl. Med. Biol. 18. 477 48 I. Seglen P. 0. (1973) Preparation of rat liver cells. Ezp. Cell Res. 16, 25~ 30. Shen T. Y. (1985) Drug delivery via cell-surface receptors. In Directed Drug Delicery (Edited by Borchardt R. T., Repta A. J. and Stella V. J.). pp. 231-245. Humana Press, Clifton. Sogami M., Nagoka S., Era S., Honda M. and Noguchi K. (1984) Resolution of human mercapt- and nonmercaptalbumin by high-performance liquid chromatography. Int. J. Peptide Protein Res. 24, 966103. Stockert R. J. and Morel1 A. G. (1983) Hepatic binding protein: the galactose specific receptor of mammalian hepatocytes. Hepatologv 3, 750-757. Stowell C. P. and Lee Y. C. (1980) Preparation of some new neoglycoproteins by amidination of bovine serum albumin using 2-imino-2-methoxyethyl I-thioglycosides. Biochemistry 19, 4899-4904. Wallner S. J. and Walker J. E. (1975) Glycosidases in cell wall-degrading extracts of ripening tomato fruits. Plant Physiol. 55, 94-98. Yamada H., Hayashi H. and Natori Y. (1984) A simple procedure for the isolation of highly purified lysosomes from normal rat liver. J. Biochem. 95, I I55 -1160.