Pegylated Zinc Protoporphyrin: A Water-Soluble Heme Oxygenase

John P. Evans , Sylvie Kandel and Paul R. Ortiz de Montellano ..... Jie Zheng , Jessica R. Lou , Xiao-Xi Zhang , Doris M. Benbrook , Marie H. Hanigan ...
0 downloads 0 Views 151KB Size
Download PDF
Bioconjugate Chem. 2002, 13, 1031−1038

1031

Pegylated Zinc Protoporphyrin: A Water-Soluble Heme Oxygenase Inhibitor with Tumor-Targeting Capacity S. K. Sahoo, T. Sawa, J. Fang, S. Tanaka, Y. Miyamoto, T. Akaike, and H. Maeda* Department of Microbiology, Kumamoto University School of Medicine, 2-2-1 Honjo, Kumamoto 860-0811, Japan. Received January 21, 2002; Revised Manuscript Received March 25, 2002

Heme oxygenase (HO) is a key enzyme in heme metabolism; it oxidatively degrades heme to biliverdin, accompanied by formation of free iron and carbon monoxide. Biliverdin is subsequently reduced by cytosolic biliverdin reductase to form bilirubin, a potent antioxidant. We recently found that tumor cells utilize HO to protect themselves from oxidative stress by producing the antioxidant bilirubin. This result suggested an important potential therapeutic strategy: suppression of bilirubin production with the use of HO inhibitors; hence, cancer cells become vulnerable to oxidative stress induced by anticancer drugs or leukocytes of the host. This concept was validated by using the intraarterial administration of an HO inhibitor, zinc protoporphyrin, in nonphysiological solution. In the present study, zinc protoporphyrin (ZnPP) was conjugated with poly(ethylene glycol) (PEG) with molecular weight of 5000, to make ZnPP, a water-soluble compound (PEG-ZnPP), and to improve its tumortargeting efficiency. PEG was conjugated to ZnPP through newly introduced amino groups, where ethylenediamine residues were added at C6 and C7 of protoporphyrin. The divalent zinc cation was chelated into the protoporphyrin ring to obtain PEG-ZnPP. PEG-ZnPP did become highly water-soluble, and it formed multimolecular associations with molecules larger than 70 kDa in aqueous media. PEGZnPP inhibited splenic microsomal HO activity in vitro in a competitive manner in the presence of hemin, with an apparent inhibitory constant of 0.12 µM. Most important, PEG-ZnPP injected intravenously significantly suppressed intratumor HO activity in a murine solid tumor model, which suggests that tumor-targeted inhibition of HO is possible with the use of PEG-ZnPP.

INTRODUCTION

Heme oxygenase (HO)1 is an enzyme of significant clinical importance because of its key role in heme degradation. HO oxidatively degrades the closed tetrapyrrole heme molecule to produce biliverdin; this process is accompanied by the release of free iron (Fe2+) and carbon monoxide (CO) (1, 2). To date, two isoforms of HO have been found: an inducible type HO-1, and a constitutive type HO-2. HO-1, a well-known heme oxygenase, is induced by a wide variety of oxidative stimuli including heme (3), ultraviolet irradiation (4), hydrogen peroxide (4), heavy metals (4, 5), heat shock (3), hypoxia (6), and nitric oxide (NO) (7, 8). Biliverdin, one of the products of heme degradation, is immediately reduced by the cytosolic enzyme biliverdin reductase to form bilirubin (1, 2). Although excessive increases in serum bilirubin concentrations cause hyperbilirubinemia or jaundice (9), recent findings suggest that adequate levels of bilirubin play an important role in protecting cells from oxidative stress because of bilirubin’s potent antioxidant activity. It was reported that bilirubin can scavenge free radicals, especially peroxyl radicals (10); prevent protein oxidation by the potent reactive nitrogen species peroxynitrite (11); and protect * To whom correspondence should be addressed: Professor Hiroshi Maeda, Department of Microbiology, Kumamoto University School of Medicine, 2-2-1 Honjo, Kumamoto 860-0811, Japan. Tel: 81-96-373-5098. Fax: 81-96-362-8362. E-mail: [email protected]. 1 Abbreviations: HO, heme oxygenase; ZnPP, zinc protoporphyrin; PEG-ZnPP, pegylated ZnPP; EPR effect, enhanced permeability and retention effect.

neuronal cells from oxidative stress injury (12). We recently found that HO-1 production was enhanced in an experimental solid tumor model (8) and, more important, that HO-1 played a vital role in tumor growth by providing, at least in part, antioxidant bilirubin to protect tumor cells from host-derived oxidative stress. The evidence for this role was the marked antitumor effect of a specific HO inhibitor, zinc protoporphyrin (ZnPP) (8). These findings suggest a therapeutic potential of this HO inhibitor, which would act as a novel antitumor agent via disruption of the antioxidant defense system of cancer cells. Metalloporphyrins represent a class of compounds in which the central iron of heme is replaced by various other metals such as cobalt, zinc, manganese, chromium, or tin (13). These metalloporphyrins function as competitive inhibitors of the HO reaction because of their inefficient binding to molecular oxygen, thus preventing HO from degrading metalloporphyrins (1, 13). Inhibitory actions of these metalloporphyrins have been studied extensively, especially with respect to therapy for hyperbilirubinemia (14-16). As mentioned above, we also demonstrated antitumor activity of ZnPP in a solid tumor model (8 and our unpublished data, S. Tanaka et al.). However, the very low solubility of these metalloporphyrins in aqueous media limits the use of these compounds in injectable formulations. In addition to the issue of solubility, targeted delivery of the HO inhibitor to tumor tissue is indispensable for antitumor treatment because nonspecific inhibition of HO may cause oxidative stressrelated side effects by reducing the favorable antioxidant capacity of bilirubin in normal organs.

10.1021/bc020010k CCC: $22.00 © 2002 American Chemical Society Published on Web 07/16/2002

1032 Bioconjugate Chem., Vol. 13, No. 5, 2002

Sahoo et al.

Figure 1. Synthesis of pegylated zinc protoporphyrin (PEG-ZnPP). Reagents and conditions: (a) ethylchloroformate, triethylamine, tetrahydrofuran, 0 °C, 2 h; (b) ethylenediamine, room temperature, 24 h; (c) succinimidyl PEG (Mr 5000), chloroform, room temperature, 24 h; (d) zinc acetate, room temperature, 1 h (See text for details).

Tumor-specific accumulation of biocompatible macromolecules and lipids, after intravenous and intraarterial administration, has been observed in many types of solid tumors of mice, rats, rabbits, and human patients. We called this phenomenon the “enhanced permeability and retention (EPR) effect” of macromolecules and lipids in solid tumor (17). The EPR effect is mainly mediated by four major factors: (i) hypervasculature of tumor tissues due to active angiogenesis, (ii) defective anatomical structure of tumor blood vessels (thus leaky), (iii) enhanced production of vascular mediators such as nitric oxide and bradykinin, and (iv) incomplete lymphatic drainage system (17-20). We hypothesized that if we could increase the molecular size as well as the watersolubility of an HO inhibitor such as ZnPP by conjugating it with water-soluble polymers, we will be able to achieve, based on the EPR effect, more effective and selective delivery of the HO inhibitor to tumor sites. The present study describes the synthesis and characterization of a water-soluble derivative of ZnPP obtained by conjugation with PEG (PEG-ZnPP). The inhibitory constant (Ki) of PEG-ZnPP against HO was determined in vitro and compared with that of native ZnPP. The biological significance of PEG-ZnPP is further discussed in view of antitumor effect in vivo. EXPERIMENTAL PROCEDURES

Materials. Protoporphyrin IX was purchased from Sigma Chemical Co. (St. Louis, MO). ZnPP was from Wako Pure Chemical Industries Ltd., Osaka, Japan. The succinimidyl derivative of PEG (MEC-50HS), with an average molecular weight of 5000, was kindly provided by NOF Co., Tokyo, Japan. PEGs used in this experiment have a molecular weight dispersity (Mw/Mn) of 1.025. Other reagents, of reagent grade, were from Wako Pure Chemical Industries and were used without further purification. General Methods. Absorption spectra were determined by using a spectrophotometer (U-2000, Hitachi Ltd., Tokyo, Japan). Fluorescence spectra were recorded by a fluorescence spectrophotometer (F-4500, Hitachi).

Mass spectra were obtained via a liquid chromatography spectrometer (LCMS QP8000R, Shimadzu Corp., Kyoto, Japan), with the electron spray ionization probe operating in the positive mode. Methanol was used as a solvent. Infrared spectra were recorded with a JIR-6500w spectrometer (JEOL, Tokyo, Japan). Thin-layer chromatography (TLC) was performed on 60 A silica with F254 aluminum-backed plates (Merck); UV-visualization was at 254 nm. Animals and Tumor Implantation. Male ddY mice, 6 weeks of age and weighing 30-35 g (from SLC, Inc., Sizuoka, Japan), were used in this study. S-180 tumor cells (2 × 106 cells) were implanted subcutaneously in the dorsal skin of the ddY mice. The inhibitory activity of PEG-ZnPP in vivo was examined on days 10-14 after tumor inoculation, when the tumors were approximately 10 mm in diameter but contained no necrotic region. All experiments were carried out according to the guidelines of the Laboratory Protocol of Animal Handling, Kumamoto University School of Medicine. Synthesis of PEG-ZnPP. The overall scheme of synthesis of PEG-ZnPP is shown in Figure 1. PEG-ZnPP synthesis proceeded by three major steps: (i) the introduction of amino groups into the protoporphyrin ring by reacting ethylenediamine with the intrinsic carboxyl groups of the ring; (ii) PEG conjugation to the amino groups; and (iii) chelation of Zn2+ into the PEG-porphyrin ring. Detailed procedures for each step are described in the following text. Synthesis of the Aminated Derivative of Protoporphyrin IX (Compound 1; cf. Compound 3). Triethylamine (2.45 mL, 17.6 mmol) was added to a suspension of protoporphyrin IX (100 mg, 178 mmol) (1) in dry tetrahydrofuran (20 mL) to remove the hydrochloric acid formed. The suspension was then cooled to 0 °C; 1.7 mL (17.9 mmol) of ethylchloroformate was added in dropwise fashion, with stirring, and the reaction was continued at 0 °C for 2 h. The resultant suspension was then filtered to remove the triethylamine hydrochloride salt. By TLC (chloroform/methanol, 9:1 v/v), Rf was 0.9. To the solution obtained (2) was added an excess of

Pegylated Zinc Protoporphyrin

Figure 2. IR spectra of 1 (a) and 3 (b). Clear appearance of amide I (1641 cm-1) and amide II (1552 cm-1) was observed in 3 suggesting the conjugation of diaminoethylene to propionic acid residues of 1 through amide bond.

ethylenediamine (1.2 mL, 17.9 mmol) against mixed anhydride to introduce amino groups into the porphyrin ring (Figure 1). The reaction proceeded at room temperature for 24 h. The solvent was removed by evaporation in vacuo, and the solid material obtained (3) was washed several times with chilled distilled water. Finally, 3 was suspended in a small aliquot of distilled water and was then lyophilized, yielding 60 mg of product. By TLC (chloroform/methanol 9:1 v/v), Rf ) 0.3. IR (KBr): 1641 cm-1 (amide I) and 1552 cm-1 (amide II) were distinct (Figure 2). The mass analysis (electron spray ionization, polarity positive) result, calculated for M+H, was 647. Conjugation of PEG with Bis(ethylenediamino)protoporphyrin. Bis(ethylenediamino)protoporphyrin (5 mg, 7.7 µmol, 3) was dissolved in 20 mL of chloroform. To this solution, 860 mg (172 µmol) of succinimidyl PEG was added in stepwise fashion (five times, about 170 mg each time) at 30-min intervals, and the reaction proceeded at room temperature for 24 h with stirring. Pegylated protoporphyrin (PEG-PP) (4) thus obtained was subjected to dialysis (cutoff size of 50 000) against distilled water for 2 days with several changes of water to remove unreacted PEG. The resultant solution was then lyophilized to obtain a powder of 4. The yield was 120 mg. The content of porphyrin moiety was determined spectroscopically, and the yield of 4 based on porphyrin moiety was 4 µmol (52%). Chelation of Zinc into the Porphyrin Ring of PEG-PP. PEG-PP (100 mg, 0.16 mM porphyrin equivalent) was dissolved in 20 mL of chloroform, to which was added 20 mg (27 mM) of zinc acetate. The solution was stirred at room temperature for 1 h to complete chelation of Zn2+ into the porphyrin ring. After the reaction, chloroform was removed by evaporation, to yield the crude PEG-ZnPP (5). PEG-ZnPP was then purified by overnight dialysis against 1 L of distilled water, with a membrane filter having a cutoff size of 8000; the water was changed three times. The yield of PEG-ZnPP after lyophilization was 77 mg. Size Exclusion Chromatography. The apparent molecular sizes of PEG-PP (4) and PEG-ZnPP were studied by means of size exclusion chromatography with the use of both polar (buffer) and nonpolar (chloroform)

Bioconjugate Chem., Vol. 13, No. 5, 2002 1033

mobile phases. As molecular size standards for PEG, conjugates of nitrotyrosine and PEG were prepared in a manner similar to the procedure used for PEG-PP. Succinimidyl PEG, with a molecular size of 5 or 12 kDa, was incubated overnight with nitrotyrosine in 100 mM phosphate buffer (pH 7) at room temperature. During this incubation, the amino group of nitrotyrosine reacted with succinimidyl PEG, which was confirmed from the loss of free amino groups determined by fluorescamine assay (see below). The solution was then dialyzed to remove unreacted nitrotyrosine. PEG-nitrotyrosine thus obtained was easily detected by absorption at 390 nm during chromatographic separation. Size exclusion chromatography was performed by using a Sephadex LH-60 column (13 mm inner diameter, 400 mm long), with chloroform as the mobile phase, at a flow rate of 60 mL/h. Absorption of each fraction (2 mL) was measured at 415 nm for porphyrin derivatives and at 390 nm for PEG-nitrotyrosine. Also, a high-performance liquid chromatography (HPLC) system (LC-10DA and SPD-10A, Shimadzu) equipped with a YMC Diol-60 column was used for size exclusion chromatography of PEG-ZnPP in buffer. The mobile phase consisted of 20 mM sodium phosphate buffer (pH 7) containing 0.2 M sodium chloride. In this experiment, bovine serum albumin (Mr 67 000) was used as the molecular size standard. Quantification of Free Amino Groups. To determine whether PEG reacted with the amino groups of 3, loss of the primary amino group after the reaction was quantified by means of an amino group-reactive fluorescence agent, fluorescamine, according to the literature (21). In brief, 2 µM (protoporphyrin equivalent) PEG-PP was dissolved in distilled water and was then reacted with fluorescamine. The intensity of fluorescence at 475 nm (under excitation at 390 nm) was determined. The concentration of free amino groups in PEG-PP was estimated by using glycine as a standard. Inhibitory Activity of PEG-ZnPP against HO in Vitro. A rat splenic microsomal fraction was prepared for measurement of splenic HO activity according to the literature (15). A liver cytosolic fraction was used as the source of bilirubin reductase (15). The HO reaction mixture was composed of the splenic microsomal fraction (1.0 mg of protein), the liver cytosolic fraction (3.0 mg of protein), and 333 µM nicotinamide adenine dinucleotide phosphate (NADPH), without or with inhibitors at a given concentration, in 1.0 mL of 90 mM potassium phosphate buffer (pH 7.4). The reaction was initiated by adding hemin (33 µM) to the mixture. In some experiments, the HO fraction and the inhibitors were incubated for 30 min before starting the reaction to permit interaction of the inhibitor to bind to HO before substrate ligand (hemin) binds to HO. The mixture was incubated for 15 min at 37 °C, and then the reaction was terminated by addition of 33 µL of 0.01 M HCl. The bilirubin formed in the reaction was extracted with 1.0 mL of chloroform, and the bilirubin concentration was determined spectroscopically by the difference in absorbance at 465 and 530 nm with a molar extinction coefficient of 40 mM-1 cm-1 as described previously (15). The Ki of PEG-ZnPP was determined by using double reciprocal Lineweaver-Burk plots and the slope of the plotted lines with added inhibitor (15). Inhibitory Activity of PEG-ZnPP against HO in an in Vivo Solid Tumor Model. To examine whether PEG-ZnPP can be targeted to tumor tissue and can inhibit intratumor HO activity, PEG-ZnPP was administered intravenously to ddY mice bearing S-180 solid

1034 Bioconjugate Chem., Vol. 13, No. 5, 2002

Sahoo et al.

Figure 4. UV spectra of PEG-protoporphyrin conjugate (PEGPP, 3.3 µM), PEG-ZnPP (4.0 µM), and native ZnPP (3.3 µM) in dimethyl sulfoxide. Inset shows the amplified spectra from 475 to 650 nm.

Figure 3. Size exclusion chromatography with chloroform. (A) bis(ethylenediamino)protoporphyrin before and after PEG conjugation. Symbols: open circle, aminated protoporphyrin; closed circle, pegylated protoporphyrin. (B) molecular weight standards of PEG-nitrotyrosine conjugates. Symbols: open circle, pegylated (5 kDa) nitrotyrosine; closed circle, pegylated (12 kDa) nitrotyrosine. Detection wavelengths were 415 and 390 nm for porphyrin derivatives and nitrotyrosine derivatives, respectively.

tumor. In this study, 300 or 600 µM PEG-ZnPP dissolved in distilled water (0.1 mL) was injected via the tail vein (0.5 or 1 mg of ZnPP equiv/kg). At 24 h after injection, tumor tissue was collected, and intratumor HO activity was measured as described previously. Control animals received distilled water without PEG-ZnPP. RESULTS

Figure 5. HPLC profiles of PEG in various conjugates (A) and the molecular size marker bovine serum albumin (BSA) (B) with the use of size exclusion column (Diol-60) eluted by sodium phosphate buffer (pH 7.0). Detection wavelengths of compounds were 415 nm (PEG-ZnPP), 390 nm (pegylated nitrotyrosine, PEG-12 kDa-NT and PEG-5 kDa-NT), and 280 nm (BSA), respectively.

Synthesis of PEG-ZnPP. In the present study, amino groups were first introduced to protoporphyrin IX (1) as reactive nucleophils (22), because a variety of polymer conjugates can be targets for nucleophilic amino groups. PEG with a succinimidyl ester group at one end was utilized to introduce water-soluble moieties to the porphyrin structure (Figure 1). LC-MS analysis indicated that 3 shows a single peak at a molecular weight of 647, which is identical to that of calculated mass (data not shown). Formation of an amide bond was demonstrated by infrared spectroscopy; 3 shows characteristic absorption at 1641 cm-1 (amide I) and 1552 cm-1 (amide II) (Figure 2). Size exclusion chromatography with chloroform (Figure 3) clearly indicated that, under the present conditions, the reaction between 3 and succinimidyl PEG proceeded completely as that no 3 remained (Figure 3A). Comparison of the chromatographic profile of 4 with those of the molecular size markers nitrotyrosine conjugated with PEG, either 5 or 12 kDa (Figure 3B), indicated that 4 has an apparent molecular size similar to PEG 12 kDanitrotyrosine, and larger than PEG 5 kDa-nitrotyrosine. Thus, PEG conjugation might occur at both two amino groups of compound 3. This notion was supported by the fact that 4 has no reactive free amino groups as determined by a fluorescamine-assisted amino group assay (data not shown).

Compound 4 shows intense absorption at 406 nm corresponding to the Soret band of protoporphyrin moiety (Figure 4). In addition, there are four bands, 505, 540, 575, 627 nm in the visible region, which correspond to the bands numbered I (627 nm), II (575 nm), III (540 nm), and IV (505 nm) for the porphyrin ring as reported previously (23). After the reaction with zinc acetate, the Soret band of 5 was found to red shift (422 nm) compared to that of 4. A marked difference was also observed in the visible region, i.e., 5 shows two major bands at 548 and 584 nm, respectively (Figure 4). These observations are well consistent with the finding that divalent metal insertion into porphyrin ring induces a red shift of the Soret band and the formation of two major bands, alpha (584 nm) and beta (548 nm) (23). Furthermore, the spectrum of native ZnPP showed good agreement with that of 5. Taken together, it was suggested that zinc was chelated into the porphyrin ring of 4, and hence, the formation of 5 (PEG-ZnPP) as depicted in Figure 1. Molecular Association of PEG-ZnPP in Aqueous Media. The PEG-ZnPP obtained here was highly watersoluble. The saturated concentration of PEG-ZnPP in water was determined to be 3.15 mM; in contrast, native ZnPP shows virtually no dissolution in water. The apparent molecular size of PEG-ZnPP in an aqueous system was studied by using size exclusion

Pegylated Zinc Protoporphyrin

Bioconjugate Chem., Vol. 13, No. 5, 2002 1035

Figure 6. Fluorescence spectra of PEG-ZnPP in chloroform and in 200 mM phosphate buffer (pH 7.4). Excitation was at 370 nm.

chromatography. Interestingly, PEG-ZnPP was eluted much faster than PEG 12 kDa-nitrotyrosine (Figure 5). This result is different from the chromatographic data obtained by using chloroform as a mobile phase (Figure 3). These observations suggest that PEG-ZnPP behaves differently in solution as the tertiary and quaternary structure of macromolecules may differ depending on the type of solvents; PEG-ZnPP might form multimolecular association in aqueous media, resulting in increased apparent molecular size; whereas, it behaves as free monomolecular dispersion in chloroform. It has been known that molecular size is an important factor for preferred tumor accumulation due to the EPR effect (17-20). Compounds smaller than 40 kDa will rapidly be cleared from circulation by urinary excretion mechanism, and hence, show no tumor accumulation. Elution time of PEG-ZnPP was comparable to that of bovine serum albumin (67 kDa), suggesting that PEGZnPP behaves as large as BSA in aqueous system. This increased molecular size of PEG-ZnPP would facilitate its tumor accumulation based on the EPR effect. Fluorescence Spectroscopy of PEG-ZnPP. Porphyrins have intense fluorescence when they are well dispersed (monomeric form) in solution, whereas upon aggregation the fluorescence is strongly quenched (24). This different fluorescent property has been utilized to study the monomer-aggregate transition of porphyrins in solutions. As shown in Figure 6, we observed intense fluorescence of PEG-ZnPP dissolved in chloroform solution. The fluorescence of PEG-ZnPP was, however, strongly quenched when aqueous buffer is used as a solvent. These observations suggest that the porphyrin moiety of PEGZnPP may aggregate, but not disperse homogeneously, in buffer solution. Considering the size exclusion chromatographic data together, PEG-ZnPP might form multimolecular association facilitated in aqueous solution similar to the micelle system. Inhibitory Activity of PEG-ZnPP against HO in Vitro. The inhibitory activity of PEG-ZnPP against HO was examined by using a rat splenic microsomal fraction as described earlier (8, 14-16). The addition of PEGZnPP (1.0 µM) to a splenic HO incubation mixture drastically decreased the formation of bilirubin in a manner that depended on preincubation time (Figure 7A). This requirement for such a preincubation period was not observed for native ZnPP (Figure 7A). This result suggests a number of possibilities: (i) binding of PEG-

Figure 7. Inhibitory activity of PEG-ZnPP on splenic microsomal heme oxygenase (HO). (A) Effect of preincubation period. Symbols: open circle, native ZnPP (1 µM); closed circle, PEGZnPP (1 µM). (B) Effect of various compounds on HO activity. The concentration of each compound was 1.0 µM. Preincubation period was 30 min. Data were averages of two measurements.

ZnPP is a very weak and slow process, and it needs time to access or fit to the catalytic site; (ii) dissociation of monomeric PEG-ZnPP from associated molecules is required for inhibitory action, and/or (iii) hydrolytic cleavage between the PEG chain and the porphyrin ring results in the ZnPP moiety being more accessible to the catalytic site of HO, or a combination of these. Further investigation is needed to clarify this point. Neither PEG-

1036 Bioconjugate Chem., Vol. 13, No. 5, 2002

Sahoo et al. Table 1. Heme Oxygenase (HO) Activity in Tumor and Liver Tissues with or without PEG-ZnPP Administrationa

tissue sample liver tumor

dose of PEG-ZnPP (mg of ZnPP equiv/kg)

HO activity (nmol of bilirubin /mg of protein/h)

0 0 0.5 1.0

1.32 ( 0.26 4.17 ( 1.07 2.23 ( 0.54* 2.46 ( 0.33*

a n ) 6. *P < 0.05. ddY mice bearing S-180 solid tumor received 0.1 mL of distilled water (control) or an aqueous solution of PEGZnPP via the tail vein. After 24 h, tumor tissue was collected, and HO activity in the tissue was measured as described in Experimental Procedures.

DISCUSSION

Figure 8. Lineweaver-Burk plot analysis of the effects of PEG-ZnPP and native ZnPP on the oxidation of Fe-heme in vitro by microsomal splenic HO. The concentration of both native ZnPP and PEG-ZnPP was 1.0 µM. Symbols: triangle (control without inhibitor), closed circle (native ZnPP), open circle (PEGZnPP). Data were averages of two measurements.

PP nor zinc alone showed any inhibitory activity against the splenic HO system, whereas PEG-ZnPP and ZnPP showed similar inhibitory activity (Figure 7B). This finding suggests that PEG-ZnPP inhibits HO activity in a manner similar to that of native ZnPP. The apparent Michaelis-Menten constant (Km) for heme in the splenic HO system was determined to be 12.2 µM (Figure 8), which is consistent with the previously reported value (15). Native ZnPP inhibited splenic HO activity in a competitive manner (Figure 8). The Ki of native ZnPP was determined to be 0.11 µM, which is also within the range of previously reported values (15). We found that PEG-ZnPP inhibited splenic HO activity in a competitive manner, similar to native ZnPP (Figure 8): the Ki of PEG-ZnPP was determined to be 0.12 µM, which is comparable to that of native ZnPP. Inhibition of Intratumor HO Activity by PEGZnPP. As mentioned above, we recently found a high level of HO expression in the rat AH136B solid tumor model in vivo (8). This high level of HO expression in the solid tumor could be mediated by, at least in part, a high level of NO produced by monocyte-derived macrophages that infiltrated the tumor (8, 25). In the murine S-180 solid tumor model, the model used in this study, we also previously demonstrated extensive production of NO in tumor tissue (26). We thus expected upregulation of HO-1 in the S-180 tumor in this study. Here, at 15 days after tumor inoculation, S-180 tumor tissue excised from mice that had received no treatment was determined to contain HO activity of 4.17 ( 1.07 nmol of bilirubin (mg protein)-1 h-1 (Table 1). This activity is comparable to that found in AH136B tumors reported previously (8) and is 3-fold higher than that in normal liver tissue (Table 1). Most important, PEG-ZnPP (30 or 60 nmol/mouse) administered intravenously significantly reduced HO activity in tumor tissue to almost 50% of the control level, i.e., tumor without treatment (Table 1). No acute lethal effect of PEG-ZnPP was observed at this dosage (n ) 6, data not shown). These findings clearly demonstrate the potential of PEG-ZnPP to inhibit intratumor HO activity.

The present study provides evidence that polymerconjugated ZnPP possesses potent inhibitory activity against HO. PEG-ZnPP became highly water-soluble and showed potent inhibition of the splenic HO activity in vitro, inhibition that was equivalent to that of the wellknown HO inhibitor ZnPP as its native form (Figure 8). Most important, PEG-ZnPP injected intravenously markedly suppressed intratumor HO activity (Table 1). Chemical modifications of porphyrin structure have been found to affect the inhibitory potential of metalloporphyrin analogues. For example, reduction of vinyl groups at the C2 and C4 positions, to form mesoporphyrin, slightly increased the inhibitory potential of zinc porphyrin (16). Zinc deuterioporphyrin (ZnDP), which has no vinyl groups on C2 and C4, showed activity similar to that of ZnPP (14). Introduction of glycol moieties at positions C2 and C4 of ZnDP significantly increased the inhibitory potential, compared with the most potent HO inhibitor tin protoporphyrin, whereas introduction of sulfonic acid at these positions completely nullified the activity (14). In our study here with a zinc-modified porphyrin, although mechanisms of action of PEG-ZnPP remain unclear, the results obtained indicate a similarity to native ZnPP, in that PEG-ZnPP inhibited HO in a competitive manner to the substrate (hemin) concentration (Figure 8), with the zinc metal center playing an essential role (Figure 7B). Crystal structure analysis of HO indicates that the heme substrate binds to the catalytic site of HO, called a heme pocket, by an R-methene bridge as a head, opposite a σ-methene bridge with both propionic acid residues exposed to the outer surface (27). This analysis may help understanding of the effect of PEG conjugation into the porphyrin ring. In this study, PEG chains were conjugated to propionic acid residues through an ethylenediamine linkage, which may not have a great effect on the binding affinity of the porphyrin portion of PEGZnPP to the heme pocket. The outer surface of the heme pocket is thought to play an important role in electron transfer from NADPH-cytochrome P450 reductase to the catalytic domain (27). Thus, binding of PEG-ZnPP to the heme pocket may disturb the interaction between HO and NADPH-cytochrome P450 reductase through steric hindrance caused by the PEG chains. Further investigation is needed to clarify these mechanistic points. Several recent studies have demonstrated a high level of HO expression in tumor tissues: a human prostate malignant tumor (28), a human renal adenocarcinoma (29), and the AH136B rat hepatocyte solid tumor model (8). The remarkable inhibition of tumor growth after administration of ZnPP to AH136B-bearing rats supports a vital role for HO in tumor growth (8). Very recently,

Pegylated Zinc Protoporphyrin

we found that bilirubin could function as an antiapoptotic factor against oxidative stress-induced tumor cell apoptosis which is induced by caspase activation, and the caspase is induced by oxidative stress (S. Tanaka et al., unpublished data). In addition, expression of HO is reportedly involved in the resistance of tumor cells to oxygen radical-generating agents such as Adriamycin (30). Therefore, antitumor activity of PEG-ZnPP may exist as a function of decreasing the antioxidant defense capacity of tumor cells, which results in sensitizing tumor cells against host-derived oxidative stress as well as oxygen radicals generated by administration of antitumor agents. Size exclusion chromatography analyses suggested that PEG-ZnPP spontaneously formed multimolecular associations in aqueous media, resulting in an increase in apparent molecular size, to that of bovine serum albumin (67 kDa) (Figure 5). Fluorescence spectroscopic analyses demonstrated a clear decrease in fluorescence intensity of PEG-ZnPP in buffer compared with that in chloroform (Figure 6). This phenomenon is explained by the different status of molecular associations of porphyrins in different media: porphyrin is well dispersed (monomeric dispersion) in chloroform and shows a high fluorescence intensity, whereas porphyrin has diminished fluorescence intensity in buffer because of molecular association, probably through π-π stacking and hydrophobic interaction (24). These findings, together with size exclusion chromatography data, indicate that PEG-ZnPP forms a molecular association in neutral buffer in this way: a hydrophobic inner core consisting of porphyrin moieties covered with hydrophilic PEG chains. Self-assembly association has been observed in a block copolymer system consisting of hydrophilic and hydrophobic polymeric chains, termed a “polymer micelle” (31). Core segregation in an aqueous milieu is the direct driving force for micellization and proceeds through hydrophobic interaction. Yokoyama and Kataoka and colleagues have extensively studied the physicochemical and biological importance of these polymer micellar systems and the application of these systems as anticancer drug carriers (31-33). These researchers found that polymer micelles with hydrophilic surfaces (i.e., polymers such as PEG) show an extremely prolonged circulation time in vivo because of the low excretion efficiency (renal clearance). Furthermore, these polymer conjugates and micelles preferentially accumulate in solid tumor after intravenous injection because of the unique vessel biology, the EPR effect, that operates in tumors (see recent reviews, refs 17-20). Interestingly, both monoclonal antibody and normal immunoglobulin G accumulated similarly to solid tumor, indicating the EPR effect is the prime importance for tumor targeting of macromolecular drugs (34). In this context, the EPR effect is also relevant for PEG-ZnPP, with enhanced delivery of PEG-ZnPP to the tumor. Further studies of PEG-ZnPP targeting to tumors as well as its tissue distribution are now under way. A preliminary result shows almost complete suppression of S-180 tumor growth with PEG-ZnPP alone, by three times of iv injection (30 nmol/mouse for each time) (unpublished data). In conclusion, we demonstrate here, for the first time, that a polymer conjugate of ZnPP effectively inhibits HO activity both in vitro and in vivo. The strategy of the synthesis described here (Figure 1) may allow us to produce functional water-soluble polymers as well as a series of metalloporphyrin analogues by alternating metal center (e.g., tin, manganese). The rational design

Bioconjugate Chem., Vol. 13, No. 5, 2002 1037

of polymer chains may provide better organ-specific targeting capacity of HO inhibitors. ACKNOWLEDGMENT

The authors thank Ms. Judith Gandy for editing and Ms. Chie Honda for typing the manuscript. This work was supported in part by a Grant-in-Aid for Research on Cancer from Ministry of Education, Culture, Sports and Science of Japan. S.K.S. would like to thank the Japan Society for the Promotion of Science (JSPS) for financial support through a postdoctoral fellowship. LITERATURE CITED (1) Schacter, B. A. (1988) Heme catabolism by heme oxygenase: Physiology, regulation, and mechanism of action. Semin. Hematol. 25, 349-369. (2) Maines, M. D. (1988) Heme oxygenase: function, multiplicity, regulatory mechanisms, and clinical applications. FASEB J. 2, 2557-2568. (3) Shibahara, S. (1988) Regulation of heme oxygenase gene expression. Semin. Hematol. 25, 370-376. (4) Keyse, S. M., and Tyrrell, R. M. (1989) Heme oxygenase is the major 32-kDa stress protein induced in human skin fibroblasts by UVA radiation, hydrogen peroxide, and sodium arsenite. Proc. Natl. Acad. Sci. U.S.A. 86, 99-103. (5) Mitani, K., Fujita, H., Fukuda, Y., Kappas, A., and Sassa, S. (1993) The role of inorganic metals and metalloporphyrins in the induction of haem oxygenase and heat-shock protein 70 in human hepatoma cells. Biochem. J. 290, 819-825. (6) Motterlini, R., Foresti, R., Bassi, R., Calabrese, V., Clark, J. E., and Green, C. J. (2000) Endothelial heme oxygenase-1 induction by hypoxia. Modulation by inducible nitric oxide synthase and S-nitrosothiols. J. Biol. Chem. 275, 1361313620. (7) Foresti, R., Clark, J. E., Green, C. J., and Motterlini, R. (1997) Thiol compounds interact with nitric oxide in regulating heme oxygenase-1 induction in endothelial cells. Involvement of superoxide and peroxynitrite anions. J. Biol. Chem. 272, 18411-18417. (8) Doi, K., Akaike, T., Fujii, S., Tanaka, S., Ikebe, N., Beppu, T., Shibahara, S., Ogawa, M., and Maeda, H. (1999) Induction of haem oxygenase-1 by nitric oxide and ischaemia in experimental solid tumours and implications for tumour growth. Br. J. Cancer 80, 1945-1954. (9) Valaes, T., Petmezaki, S., Henschke, C., Drummond, G. S., and Kappas, A. (1994) Control of jaundice in preterm newborns by an inhibitor of bilirubin production: Studies with tinmesoporphyrin. Pediatrics 93, 1-11. (10) Farrera, J. A., Jauma, A., Ribo, J. M., Peire, M. A., Parellada, P. P., Roques-Choua, S., Bienvenue, E., and Seta, P. (1994) The antioxidant role of bile pigments evaluated by chemical tests. Bioorg. Med. Chem. 2, 181-185. (11) Minetti, M., Mallozzi, C., Di Stasi, A. M., and Pietraforte, D. (1998) Bilirubin is an effective antioxidant of peroxynitritemediated protein oxidation in human blood plasma. Arch. Biochem. Biophys. 352, 165-174. (12) Dore, S., Takahashi, M., Ferris, C. D., Zakhary, R., Hester, L. D., Guastella, D., and Snyder, S. H. (1999) Bilirubin, formed by activation of heme oxygenase-2, protects neurons against oxidative stress injury. Proc. Natl. Acad. Sci. U.S.A. 96, 2445-2450. (13) Drummond, G. S. (1987) Control of heme metabolism by synthetic metalloporphyrins. Ann. N. Y. Acad. Sci. 514, 8795. (14) Chernick, R. J., Martasek, P., Levere, R. D., Margreiter, R., and Abraham, N. G. (1989) Sensitivity of human tissue heme oxygenase to a new synthetic metalloporphyrin. Hepatology 10, 365-369. (15) Drummond, G. S., and Kappas, A. (1981) Prevention of neonatal hyperbilirubinemia by tin protoporphyrin IX, a potent competitive inhibitor of heme oxygenase. Proc. Natl. Acad. Sci. U.S.A. 78, 6466-6470.

1038 Bioconjugate Chem., Vol. 13, No. 5, 2002 (16) Vreman, H. J., Ekstrand, B. C., and Stevenson, D. K. (1993) Selection of metalloporphyrin heme oxygenase inhibitors based on potency and photoreactivity. Pediatric Res. 33, 195200. (17) Matsumura, Y., and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res. 46, 6387-6392. (18) Maeda, H. (2001) SMANCS and polymer-conjugated macromolecular drugs: advantages in cancer chemotherapy. Adv. Drug Deliv. Rev. 46, 169-185. (19) Maeda, H. (2001) The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumorselective macromolecular drug targeting. Adv. Enzyme Regul. 41, 189-207. (20) Maeda, H., Sawa, T., and Konno, T. (2001) Mechanism of tumor-targeted delivery of macromolecular drugs, including the EPR effect in solid tumor and clinical overview of the prototype polymeric drug SMANCS. J. Controlled Release 74, 47-61. (21) Stocks, S. J., Jones, A. J., Ramey, C. W., and Brooks, D. E. (1986) A fluorometric assay of the degree of modification of protein primary amines with poly(ethylene glycol). Anal. Biochem. 154, 232-244. (22) Kimura, M., Matsumura, Y., Miyauchi, Y., and Maeda, H. (1988) A new tactic for the treatment of jaundice: An injectable polymer-conjugated bilirubin oxidase. Proc. Soc. Exp. Biol. Med. 188, 364-369. (23) Falk, J. E. (1964) Spectra. Porphyrins and Metalloporphyrins, Their General, Physical and Coordination Chemistry, and Laboratory Methods, pp 72-84, Elsevier Publishing Company, Amsterdam. (24) Nishide, H., Mihayashi, K., and Tsuchida, E. (1977) Dissociation of aggregated ferroheme complexes and protoporphyrin IX by water-soluble polymers. Biochim. Biophys. Acta 498, 208-214. (25) Doi, K., Akaike, T., Horie, H., Noguchi, Y., Fujii, S., Beppu, T., Ogawa, M., and Maeda, H. (1996) Excessive production of nitric oxide in rat solid tumor and its implication in rapid tumor growth. Cancer 77, 1598-1604.

Sahoo et al. (26) Wu, J., Akaike, T., and Maeda, H. (1998) Modulation of enhanced vascular permeability in tumors by a bradykinin antagonist, a cyclooxygenase inhibitor, and a nitric oxide scavenger. Cancer Res. 58, 159-165. (27) Schuller, D. J., Wilks, A., Ortiz de Montellano, P. R., and Poulos, T. L. (1999) Crystal structure of human heme oxygenase-1. Nat. Struct. Biol. 6, 860-867. (28) Maines, M. D., and Abrahamsson, P. A. (1996) Expression of heme oxygenase-1 (HSP32) in human prostate: normal, hyperplastic, and tumor tissue distribution. Urology 47, 727733. (29) Goodman, A. I., Choudhury, M., da Silva, J. L., Schwartzman, M. L., and Abraham, N. G. (1997) Overexpression of the heme oxygenase gene in renal cell carcinoma. Proc. Soc. Exp. Biol. Med. 214, 54-61. (30) Lee, T. C., and Ho, I. C. (1994) Expression of heme oxygenase in arsenic-resistant human lung adenocarcinoma cells. Cancer Res. 54, 1660-1664. (31) Kwon, G. S., and Kataoka, K. (1995) Block copolymer micelles as long-circulating drug vehicles. Adv. Drug Delivery Rev. 16, 295-309. (32) Kataoka, K., Matsumoto, T., Yokoyama, M., Okano, T., Sakurai, Y., Fukushima, S., Okamoto, K., and Kwon, G. S. (2000) Doxorubicin-loaded poly(ethylene glycol)-poly(β-benzylL-aspartate) copolymer micelles: their pharmaceutical characteristics and biological significance. J. Controlled Release 64, 143-153. (33) Yokoyama, M., Fukushima, S., Uehara, R., Okamoto, K., Kataoka, K., Sakurai, Y., and Okano, T. (1998) Characterization of physical entrapment and chemical conjugation of adriamycin in polymeric micelles and their design for in vivo delivery to a solid tumor. J. Controlled Release 50, 79-92. (34) Marecos E., Weissleder R., and Bogdanov A., Jr. (1998) Antibody-mediated versus nontargeted delivery in a human small cell lung carcinoma model. Bioconjugate Chem. 9, 184191.

BC020010K