Labeling of hematoporphyrin with 105Rh and binding studies with human gamma globulin

Appl. Rodiaz. hot. Vol. 41, No. I, pp. 69-73, 1990 Inc. J. Radiar. Appl. Instrum. Par1 A Printed in Great Britain. All rights reserved

Copyright

0883-2889/90 $3.00 + 0.00 C 1990 Pergamon Press plc

with lo5Rh Labeling of Hematoporphyrin and Binding Studies with Human Gamma Globulin M. R. A. PILLAI,*

J. M. LOT and

Department of Chemistry, University of Missouri, (Receiced 12 December

D. E. TROUTNER Columbia,

MO

65211,

U.S.A.

1988; in revised form 20 March 1989)

Labeling of hematoporphyrin with loSRh at stoichiometric concentrations IS described. Labeling efficiencies of UD to 93% could be obtained at pH 9.0 in bicarbonate buffer. Solvent extraction of ‘“! Rh-hematopdrphyrin into methyl isobutyl ketone was used to estimate the complex yield. The complex the 6 days of study. showed high stability and no loss of lo5Rh was seen throughout lo5Rh-hematoporphyrin when incubated with hmr+an y globulin was seen to be quantitatively bound to the protein. This procedure may be used for labeling monoclonal antibodies with losRh for therapeutic applications

which could be coupled to proteins for in diagnostic applications (Malcolme-Lawes, 1980). Our results of Rhxysteine complex conjugation with serum albumin were reported recently (Lo et al., 1988). Porphyrins are known to be avid chelators of metals and hence we have tried to use one of them as a ligand for complexing rhodium and coupling to proteins. We have selected the commercially available porphyrin derivative hematoporphyrin IX [8,13bis( 1-hydroxy ethyl)-3,7,12,17 tetramethyl 2 1H, 23H porphine-2,18 dipropionic acid] (Fig. 1). Our aim was to use one of the two propionic acid groups for coupling to an -NH, group in the protein molecule. Doi et al. reported the preparation of ‘09Pd-hematoporphyrin for selective lymphatic ablation studies (Doi et al., 1981). They synthesized an N-methyl hematoporphyrin precursor and labelled it with a DMSO solution of ‘09PdCl,. However, this procedure was not suitable for our studies as precursor preparations will esterify the -COOH, making it unavailable for coupling. The amount of hematoporphyrin used in the labelling is not given in the paper, but we believe that it was at the mg level. Our primary aim was to prepare labelled hematoporphyrin in 10~3-10-4 mmol levels so that it would be suitable for antibody studies. It was also not clear whether complexation was complete with respect to ‘09Pd activity as no quality control of the complex was carried out. Lavallee and Fawwaz (1986) have reported the synthesis and characterization of “I In-hematoporphyrin. They have labelled Photofrin llTM in glacial acetic acid medium with yields up to 95%.

Co(III),

Introduction

vitro

Because of its ideal nuclear properties ‘OSRh is one of the possible nuclides for radiotherapeutic applications (Troutner, 1987). Its fl- emissions of 560 keV, 70% and 250 keV, 30% are suitable for therapy, while the 319 keV, 19% and 306 keV, 6% y rays can be used for imaging to monitor the efficacy of the treatment. The 35.5-h half-life of this isotope is compatible with the in uivo uptake of antibodies by tumor cells. lo5Rh can be produced in a medium flux reactor at high enough specific activities for therapeutic applications (Grazman and Troutner, 1988; Wong et al., 1988). ‘“5Rh can be attached to proteins and antibodies through bifunctional chelating agents. The complexes of rhodium are kinetically inert so that they can be treated as simple molecules or ions. Hence we have been aiming at the preparation of preformed lo5Rh complexes and conjugation of these complexes to antibodies. We have recently reported labelling of proteins with lo5Rh using4-paratoluic acid-diethylenetriamine synthesized for this purpose (John ef al., 1988). We have also been investigating the possibility of using some commercially available ligands as bifunctional chelating agents. We have used the amino acid cysteine because it was reported by Malcolme-Lawes that it forms a complex with

Permanent addresses: *Radiopharmaceuticals Division, Bhabha Atomic Research Centre, Bombay 400 085, India. tInMute of Nuclear Science, National Tsing Hua University, Hsinchu, Taiwan 30043, R.O.C. 69

70

M. R. A. PILLAIet al. CH3CH*OH

H3C

H,C HOOC*CH>*CH,

CHS

OH I CH.CH?

CH,

CH,.CH,.COOH

Fig. 1. Hematoporphyrin IX. [8,13-bis( 1-hydroxyethyl)3.7,12,17-tetramethyl-21 H, 23H-porphine-2.1%dipropionic acid]. Wong (1984) reported labelling of a hematoporphyrin derivative with “‘In at pH 7.4 with heating at 120 C for 30 min, and obtained labelling yields around 95%. However, in all these studies the hematoporphyrin was in large excess over the radioisotope. Moreover, palladium and indium are better chelating metals than rhodium. In this paper we report the labelling of hematoporconcentrations. phyrin with lo5Rh at stoichiometric We have used solvent extraction of labelled hematoporphyrin with methylisobutylketone (MIBK) as a simple, but accurate, quality control procedure for estimation of the complexation yield. Purification of labelled hematoporphyrin by silica gel column chromatography was tried with limited success. with human ; Binding of lo5Rhhhematoporphyrin globulin was studied for possible application in radioimmunotherapy.

Materials

and Methods

Materials Hematoporphyrin IX, RhCI,.3H20, and l-(3dimethylaminopropyl)-3-ethyl-carbodiimide hydrochloride (ECDI) were purchased from Aldrich Chemical Company. Spectrapor molecular porous membrane tubing (mol. wt cut off 12,00&14,000 Da) was purchased from Spectrum Medical Industries, Inc. Human IgG (I-4506) and Sephadex G-75 were purchased from Sigma Chemical Company. Silica gel 100, particle size 0.2-0.5 mm, 35-70 mesh, was from E. Merck, Darmstadt. All other solvents and materials were of reagent grade. ‘“‘Rh was prepared at the Missouri University Research Reactor and supplied as no-carrier-added in HCl. Saline solution used in these studies was prepared by dissolving 9 g of NaCl in 1 L of double-distilled water. E.xperimental One mL of Preparation of ‘“Rh-hematoporphyrin. RhClt (2.5 x 10-j mmol), 0.3 mL ‘OSRh (-37 MBq), and 1 mL of 0.5 M bicarbonate buffer pH 9.0 were mixed together in a IO-mL round-bottom flask fitted with a condenser and refluxed for I min in a boiling water bath. To this 0.6 mL of an alcoholic solution of hematoporphyrin (3 x 10m3 mmol) was added and

refluxing continued for about 2 h. At the end of refluxing the solution was taken to dryness by removing the condenser and continuing heating until the solution evaporated. The contents were redissolved in 3 mL saline. The final hematoporphyrin concentration was 10~3mmol/mL and the Rh:hematoporphyrin ratio was a little less than I. Characterization of the complex. We observed that the rhodium complex of hematoporphyrin is extracted in methyl isobutyl ketone, MIBK. and therefore used a solvent extraction technique to estimate the percentage complexation. An amount of 20 PL of the complex was diluted to 1 mL in saline and mixed with I mL of MIBK in a vortex mixer for about 1 min. The tubes were then centrifuged for I min and the activity in the organic and aqueous layer measured and the percentage extraction calculated. A second extraction of the aqueous layer and back extraction of the organic layer were carried out by mixing with equal volumes of MIBK and saline respectively. Purl$cation of labelled porphyrin. The labelled hematoporphyrin was purified from uncomplexed rhodium by silica gel column chromatography. Silica gel 100, particle size 0.2-0.5 mm, 35-70 mesh was swollen in 0.9% saline and loaded on a small syringe column to a volume of 6 mL. ““Rhhhematoporphyrin (0.1 mL) was applied on the top of the column. Elution was carried out with 10 mL of saline followed by 20 mL saline:acetone (50:50 v/v). The saline fraction contained unreacted rhodium and at times a small quantity of “‘Rh-hematoporphyrin which was not retained in the column. Saline:acetone lo5Rh-hematoporphyrin were fractions containing heated gently to remove acetone. Binding studies with protein. Specific coupling: 0.1 mL (IO- ’ mmol) of lo5Rh-hematoporphyrin was mixed with 2 mL of IgG (lOmJmmol) solution in saline. The pH of the solution was adjusted to 5.0 using HCl and 40 mg of ECDI powder added. The were mixed together and incubated contents overnight at room temperature. Non-specific binding was measured by running experiments as above, but without the coupling agent ECDI. A similar non-specific binding experiment was also run at pH 8.8 in 0.1 M borate buffer. Estimation ofprotein bound acticity. The amount of ‘“‘Rh bound to protein was estimated by gel filtration over a Sephadex G-75 column (30 x 1.4 cm). Normal saline was used as eluent. Fractions of 2-mL were collected and the activity measured with a NaI(T1) scintillation counter. Dialysis. The stability of the IgG bound labelled hematoporphyrin was estimated by equilibrium dialysis. A quantity of 0.5mL of the reaction mixture from the protein binding studies was diluted to 3 mL and loaded in a 4-cm molecular membrane bag and dialyzed for 48 h against 5 changes of distilled water, 2 L at each time. Original solution (0.5 mL) was diluted to 3.0 mL and kept as a control At the end

Labeling Table

of hematoporphyrin

I.

with lo5Rh

Results of complexation

71

studies

Porphyrin Poruhvrin (nmol)

RhCI, @mCJi,

2h

% Yield

4.0

250

250 25 2.5 NCA

I IO 100

I5 22 I8 14

6.0

250

250 2s 2.5 NCA

I IO 100

35 31 31 36

9.0

250

250 25 2.5 NCA

1 IO 100

56 68 54 38

6.0

2500

2500 250 25 NCA

I IO 100

73 62 63 87

9.0

2500

2500 250 25 NCA

I IO 100

93 79 63 70

PH

Reaction volume was 3.1 mL. Buffers were sodium acetate for pH 4 and 6, and sodium bicarbonate for pH 9. Radioactivity for each run was -5 MBq. Yields are reuorted as % of activity transferred by one extraction from saline solution into MIBK.

of dialysis the activity remaining in the dialysis bag was compared against the total activity loaded as determined from the control. Care was taken to have as similar a geometry as possible during counting for the control and dialyzed solution. The dialysed solution was counted along with the dialysis bag.

Results Complexation A large number of studies were carried out to determine the optimum conditions for complexation. These included complexation reactions at different buffers, pH, rhodium to hematoporphyrin ratios, and concentration levels. Table I summarizes the results. The complexation percentage reported are the first extraction yields in MIBK. The reaction yields were generally high at pH 9.0, but were dependent on the porphyrin and metal concentration. As expected yields were high at higher concentrations of porphyrin (2.5 x 1O-3 mmol). However, contrary to our expectations, the reaction yield decreased with increasing porphyrin to metal ratios and best yields were at a 1: 1 hematoporphyrin:Rh ratio. At 2.5-3.0 mmol of hematoporphyrin and 2.5 mmol of rhodium reaction yields up to 93% were obtained.

Table 2. Stability

of “‘Rh-hematoporphyrin

Time (h)

% Complex

0 26 48 96 144

76.5 76.2 75.1 78.8 80.4

The % complex MlBKjsaline

is the first extraction system.

yield in the

The yield in acetate buffer at pH 4.0 was significantly lower. At this pH solubility of hematoporphyrin was poor and it tended to come out of solution. This together with the difficulty in deprotonation of the tetrapyrrole ring may be a reason for the low yield. Characterization of the complex was mainly carried out by solvent extraction using MIBK. It was found that MIBK extracts hematoporphyrin and ‘OSRh-hematoporphyrin leaving behind inorganic rhodium in all forms in the aqueous layer. Upon back extraction of ‘OSRh-hematoporphyrin to saline 94.4 _+ 1.6% (n = 7) of the activity was seen in the organic phase indicating a partition of 20: 1. The second extraction in most cases was about 5-7%, and hence for yield calculations we have taken the first extraction data only. When the hematoporphyrin concentration used was high some solid was seen in the organic phase. This solid was suspended in saline and extracted with MIBK. The extraction yield was greater than 95% indicating that it was pure

30 * .= z

1

Saline:Acetone, 48 %

1:l v/v

R

‘;, 20-

a E al lo2 d 0

o

-0

20

Elution dokme (mL) Fig. 2. Purification of ‘05Rh-hematoporphyrin on a silica gel column. The first peak corresponds to uncomplexed rhodium and the second to ‘05Rh-hematoporphyrin.

M. R. A.

72

lo5Rh-hematoporphyrin. Hence for all solvent extraction studies we removed 0.8 mL of the aqueous solution and the remaining 0.2 mL of the aqueous solution was counted with the organic fraction. A correction was applied for this residual aqueous layer activity in the organic phase during calculation. Stability

PuriJication qf the comples The elution pattern of silica gel column chromatography for a reaction mixture containing 61.5% ‘05Rh-hematoporphyrin is given in Fig. 2. The first peak corresponds to uncomplexed rhodium and fallthrough of a small amount of lo5Rh-hematoporphyrin eluted in saline. The second peak eluted in saline:acetone (50% v/v) was found to be pure lo5Rh-hematoporphyrin. After removal of acetone the extractability of this fraction was -90%. The overall recovery of lo5Rh-hematoporphyrin from the column was 45% as against 62% of the complex present in the original solution. About 5-6% of the loaded activity was held in the column and some “‘Rh-hematoporphyrin activity was lost in the first few fractions. Binding studies with IgG Binding studies of the lo5Rh-hematoporphyrin with human y globulin with or without the coupling agent ECDI showed no difference. In both cases the ro5Rh-hematoporphyrin activity was seen in the protein fraction during column chromatography. Studies at pH 8.8 also indicated that ‘05Rhhematoporphyrin was quantitatively bound to protein. Elution patterns of lo5Rh-hematoporphyrin bound to IgG and coupled with ECDI are shown in Fig. 3. The bulk of the activity corresponding to lo5Rh-hematoporphyrin was eluted with protein (15-20 mL). Uncomplexed lo5Rh impurities were seen W2,

40 30

& 2 p" .=

Table 3. Results of equilibrium dlalysls studies % Activity retained inude the membrane ‘“5Rh~hematoporphyrin ‘OSRh-hematoporphyrin + IgG ‘OS Rh-hematoporphyrin + I@ + ECDI

66.5 x5.7 XY

qf ‘*‘Rh-hematoporphyrin

The stability of ‘“‘Rhhhematoporphyrin was studied by solvent extraction of the complex at different time intervals after preparation. The results are given in Table 2. The complex showed exceedingly high stability and no loss of lo5Rh was seen throughout the 6 days studied.

.? zi a

PILLAI et al.

20 10 50--

rl

O-00

Elution Volume (mL)

Rh-hematoporphyrin conjugated Fig 3. Purification of ‘OS to IgG on a Sephadex G-75 column (30 x 1.4cm). Eluent was normal saline. 0, ECDI present; n , no ECDI present.

from 3450mL. In a separate experiment in which lo5Rhhhematoporphyrin alone was passed through this gel permeation column, the entire activity was adsorbed on the top layer of the column and could not be eluted even with 200mL of the eluent. In the case of the reaction mixture with ECDI 83% of the total activity was seen with IgG. and without ECDI 80% of the total activity was seen with IgG. Both these figures are very similar to the total amount of ‘05Rh-hematoporphyrin present in the solution. From these experiments it was evident !hat ‘05Rh-hematoporphyrin was quantitatively adsorbed by the protein, even without a coupling agent. Dialysis Results of the equilibrium dialysis experiment are given in Table 3. In the case of “‘Rhhhematoporphyrin 35% of the activity was removed from the molecular membrane. Under the same conditions > 85% of the activity remained inside the dialysis bag for protein bound lo’ Rhhhematoporphyrin. No significant difference was seen with the specifically coupled (ECDI) and non-specifically bound protein ‘05Rh-hematoporphyrin. Note that the dialysis was incomplete as only 35% of the ““Rhhhematoporphyrin could be removed from the dialysis bag.

Discussion One of the problems during the initial studies was the unavailability of a suitable method for the estimation of complex yields. Silica gel thin layer chromatography could not be used for reactions carried out at basic pH, because both labelled hematoporphyrin and uncomplexed rhodium were seen at the point of spotting. We tried several solvent systems for extracting the complex ana observed that hematoporphyrin is extracted into MIBK. The extractability did not change with the addition of rhodium inside the porphyrin ring. As the partition coefficient was -20 we took the first extraction yield to be the complex yield. The same solvent extraction technique could have been used as a purification technique except for the high b.p. (117°C) of MIBK. The silica gel column chromatography proved to be a good purification technique, except that IO-15% of the ‘05Rh-hematoporphyrin activity was lost. We first tried to elute ‘05Rhhhematoporphyrin with acetone and found that a significant amount of the activity was retained in the column. Upon modification of the solvent system to saline:acetone (1: I, v/v) the activity retained in the column was reduced to

Labeling

of the was to couple the “‘5Rh-hematoporphyrin with IgG and the final gel filtration removed nonconjugated ‘OSRh-hematoporphyrin and uncomplexed rhodium from IgG. We attempted specific coupling of “‘Rhhematoporphyrin by activating the -COOH group by isobutylchloroformate or diphenylphosphorylazide (Krejcarek and Tucker, 1977; Arano et al., 1986). In these methods it is essential to bring the lo5Rhhematoporphyrin into organic solution. After complete removal of water labelled hematoporphyrin was not soluble in any of the solvents used in these coupling studies. Therefore, we attempted the ECDI method. These experiments, however, provide no evidence that specific coupling took place in our experiments, as there was no difference in the dialysis experiment for the solutions with or without ECDI. The protein binding mechanism of hematoporphyrin may be similar to the carrying of heme by hemoglobulin. In that case iron is coordinated to four nitrogens of the tetrapyrrole ring, and one of the two remaining valencies is satisfied by the nitrogen of histidine from the protein, and the last valency by oxygen (Lehninger, 1976). In the present case, we believe that rhodium may be coordinated to the four nitrogens of the porphyrin ring and the remaining two valencies may be satisfied by coordinating groups donated by the protein. Further stability may also come from hydrogen bondings and Van der Waals forces possibly due to the availability of -COOH and -OH groups in hematoporphyrin. The clinical utility of the nonspecifically bound ‘OSRh-hematoporphyrin to antibody is uncertain as we have no evidence to suggest in viuo stability. Biodistribution studies with monoclonal antibodies labelled by this method once carried out will help in ascertaining the application of this method in radioimmunotherapy. There exists a good possibility that the lo5Rh-hematoporphyrin will be released from the monoclonal antibody in the presence of other body proteins. If, however, this does not occur this method may be one of the best to label

about

complex

5%

was

of the

of hematoporphyrin

not

total

essential

load.

as our

Purification aim

with lo5Rh

73

monoclonal antibodies with radioisotopes, because it requires no chemical modification of the monoclonal antibodies. Earlier success in the detection of tumor with hematoporphyrin derivatives suggest the possibility of using lo5Rh-hematoporphyrin also as a radiopharmaceutical for tumor therapy (McBride, 1979; Cortes and Kinsey, 1982; Benson et al., 1982). The excellent stability of ‘05Rh-hematoporphyrin will be a major advantage in such studies. Acknowledgements-This work was supported by the U.S. Department of Energy, Grant No. DE-FG02-86ER60400 and the University of Missouri Research Council. M.R.A.P. acknowledges the kind encouragement of Mr. C. N. Desai, Head, Radiopharmaceuticals Division, B.A.R.C. J. M. Lo wishes to thank the National Science Council of ROC for financial support for his study in the University of Missouri.

References Arano Y., Yokoyama A., Magata Y., Saji H., Horiuchi K. and Torizuka K. (1986) Nucl. Med. Biol. 12, 425 (1986). Benson R. C., Farron G. M., Kinsey J. H. et al. (1982) Mayo Clin. Proc. 57, 548. Cortes D. A. and Kinsey J. H. (1982) Mayo Clin. Proc. 57, 543. Doi J. D., Lavallee D. K., Srivastava S. C., Prach T. and Richards P. (1981) Inr. J. Appl. Radiat. Isot. 32, 887. Grazman B. and Troutner D. E. (1988) Appl. Radiar. Isot. 39, 257. John C. S., Pillai M. R. A., Lo J. M. and Troutner D. E. (1988) Abstract. /. Nucl. Med. 29, 814. Krejcarek G. E. and Tucker K. L. (1977) Biochem. Biophys. Res. Commun. 77, 58 1. Lavallee D. K. and Fawwaz R. (1986) Nucl. Med. Biol. 13, 639. Lehninger L. A. (1976) Biochemistry, 2nd edn, p. 490. Worth, New York. Lo J. M., Pillai M. R. A., John C. S. and Troutner D. E. (1988) Fourth Asia and Oceania Congress of Nuclear Medicine, Taipei, Taiwan. Malcolme-Lawes D. J. (1980) Int. J. Appl. Radial. Isot. 31, 190. McBride G. (1979) J. Am. Med. Ass. 242, 403. Troutner D. E. (1987) Nucl. Med. Biol. 14, 171. Wong D. W. (1984) Int. J. Appl. Radial. Isot. 35, 691. Wong Y. K., Ketring A. R.; Lo J. M. and Troutner D. E. (1988) Secenth Inr. Svmp. on Radiopharmaceutical Chemistry, The Netherlands.