R-[N-acetyl]eglin c:poly(oxyethylene) conjugates: Preparation, plasma persistence, and urinary excretion

R-[MAcetylIeglin c:Poly(oxyethylene) Conjugates: Preparation, Plasma Persistence, and Urinary Excretion PETERGODDARD**~, JOHN O’MULLANE*~, LUSIE AMBLER*, ANDREWDAW*, LAURENCE BROOKMAN*, ALFREDLEE*,AND KAREL PETRAK’ Received August 2, 1990, from the ‘Advanced Drug Delivery Research Unit, Research Centre, Ciba-Gei y Pharmaceuticals, Wimblehurst Road, Horsham, West Sussex, RH12 4AB, U.K. Accepted for publication January 30, 1991, Present adiesses: *Norsk Hydro as., Research Centre, N-3901 Porsgrunn, Norway; %mithKline Beecham, Research and Development, Weybridge, Surrey, KT13 OSB, U.K. Abstract 0 In this paper, we describe the preparation, purification, and characterization of conjugates of R[Nacetyl]eglin c (Eglin c) with poly(oxyethy1ene)(POE; Eglin c:POE). The plasma profile and urinary excretion of the conjugates has been determined after iv administration in mice. The modificationof Eglin c with POE does not significantlyimpair the ability of Eglin c to bind elastase as measured by an in vitro assay. In the best example, 79% of theoretical activity was retained by the conjugate. The in vivo results clearly show that the amount of Eglin c:POE in plasma after iv administration is much higher than comparative doses of unconjugated Eglin c. The time course of the plasma concentration of the conjugate matches closely that of the corresponding free polymer. Consequently, we can expect that higher plasma concentration could be achieved, if and when required, by selecting polymers of appropriate size.

Proteins and polypeptides are of increasing importance in pharmaceutical research and clinical therapy. Hormones, serum proteins, and enzymes, in either their natural or modified state, are all well established as therapeutic agents for the treatment of clinically important indications such a s cardiovascular disorders, tumors, autoimmune diseases, and infections.1 A major driving force behind the development of proteins as therapeutic agents has been the successful development of biosynthetic methods for the commercial-scale production of high purity materials. This has arisen largely due to the advances made in genetic engineering during the 1970s and beyond. de-N-Acetylated Eglin c, originally isolated from leeches, is a potent elastase inhibitor which has a potential clinical use in the treatment of emphysema.2 More recently, the gene for this polypeptide has been cloned and expressed in Escherichia coli by Ciba-Geigy A. G. to give recombinant N-acetylated Eglin c (Eglin c), an equally potent elastase inhibitor.3 The structure of Eglin c is given in Figure 1. In the treatment of emphysema, the ultimate target for Eglin c is the e1astase:elastin complex in the interstitium of the lung. Given that the endothelium in general is relatively “leaky” to protein molecules, there is expected to be ready access to the interstitium of a relatively small molecule, such as Eglin c (molecular weight 81331, and also of higher

Ac-Thr-Glu-Phe-Gly-Ser-Glu-Leu-Lys-Ser-PhePro-Glu-Val-Val-Gly-Lys-Thr-Val-Asp-GlnAfa-Arg-GIu-Tyr-Phe-Thr-Leu-His-Tyr-ProGln-Tyr-Asp-Val-Tyr-Phe-Leu-Pro-Glu-GlySer-Pro-Val-Thr-Leu-Asp-Leu-Arg-Tyr-AsnArg-VaI-Arg-Val-Phe-Tyr-Asn-Pro-Gly-Thr-

Asn-Val-Val-Asn-His-Val-Pro-His-Val-Gly Figure 1-Primary structure of R-[Nacetyl]Eglinc. OO22-3549/91/12OO-1 1 71$02.50/0 0 1991, American Pharmaceutical Association

molecular-weight species. (In humans the proteins concentration of lymph leaving the lung is relatively high, averaging -4 g/lOO mL as compared with -2 g/lOO mL in the peripheral tissues.4) Unfortunately, after iv administration, Eglin c has an extremely short circulating half-life (t,,2) in the body of -10 min; the majority of an injected dose in rodents is rapidly removed from the vasculature and excreted intact in the urine. The purpose of our present work was to extend the tl12 of Eglin c to provide a circulating “depot” of the polypeptide within the vasculature available for extravasation across the endothelium into the interstitum of the lung. In studies with synthetic polymeric carriers, we and others have established how, in addition to the effects of chemical structure, the molecular size of a polymer will determine the residence time of the macromolecule in circulation and in the other compartments of the b0dy.u These general observations led us to conclude that both the retention of Eglin c in circulation and its access to the interstitium might well be achieved by administering a higher molecular weight derivative of Eglin c. As a consequence, we chose to attach poly(oxyethy1ene)(POE), with a number-average molecular weight (M,)of >20 000, to the available amino groups of the polypeptide. The activity of the conjugate was assayed in order to determine the effects of conjugation on Eglin c. Finally, the circulating tlIzof conjugates with -70% of the original activity of Eglin c was examined following iv administration.

Experimental Section ChemicalsiAll chemicals of reagent grade or above were obtained from BDH (Poole, U.K.) or Aldrich (Gillingham,U.K.) and were used without further purification unless stated otherwise. Solvents were dried over 4A molecular sieves. The poly(oxyethy1ene) (POE), with number-average molecular weights (M,)of 20 000 (POE20) and 35 000 (POE35),was obtained from BDH and Fluka (Derbyshire,U.K.), respectively. Prior to use, the POE was dissolved in tetrahydrofuran and precipitated by the addition of diethyl ether. This procedure was repeated twice, and the precipitated material was collected and dried over phosphorus pentoxide in a vacuum desiccator. The POE, activated as its chloroformate derivative (POE-OCOCl), was prepared by the reaction of POE with phosgene in dry acetonitrile following a general method9 previously described for the preparation of MeO-POE chloroformate.1° The POE-OCOC1 was assayed by measuring the W absorbance of its 2-(4-hydroxyphenyl)ethylamine (tyramine)derivative at 275 nm and was found to contain 0.77 x mol chloroformate groupdg of polymer conjugate. The tyramine conjugate was prepared by reaction of POE-OCOCl with tyramine in sodium borate buffer at 25 “C. A radioiodinated sample of POE ([1261]POE20;M,,= 20 OOO), with a specific activity of 62.85 pCi lZ6Umg,was prepared by reacting the POE-tyramine derivative with NalZ6Iin the presence of N-chloro-4toluenesulfonamide sodium salt (Chloramine T).7

Journal of Pharmaceutical Sciences I 1171 Vol. 80, No. 12, December 1991

Preparation of Eglin c:POE Conjugates-The POE-OCOCl was added to Eglin c in at least two portions to ensure an excess of POE in the reaction mixture. Preparation of a n Eglin c:POE20 Conjugate-Typically, Eglin c (0.010 g, 125 x mol) was dissolved in pH 8.0 borate buffer (2 mL). The POE-OCOC1 (M,, = 20 000; 0.010 g, 0.770 x mol OCOCl groups) was added with stirring to the solution a t room temperature. After 30 min, a second addition of activated POE (0.01 g, 0.770 x mol chloroformate groups) was made, and this was repeated every 30 min until a total of 0.100 g of POE (7.70 x mol OCOCl groups, 5.0 x mol POE) was added. The reaction mixture was left to stir overnight (18 h). To recover the product, the mixture was dialyzed against water using an Amicon ultrafiltration cell (YM 10 membrane). The retentate was lyophilized to give a white solid (0.956 g, 87%recovery of total reaction mixture). The product was fractionated using a gel permeation chromatography (GPC) semipreparative system as described below. Preparation of a n Eglin c:POE35 Conjugate-The above preparation was repeated except that 0.200 g of Eglin c (25 x mol) was reacted with POE-OCOCl (M,, = 35 OOO), which was added to the reaction mixture in 0.200-g (25 x mol) portions as above. After mol), the final addition of POE-OCOCl (total of 1.00 g, 28.57 x the mixture was left to stir for 3 days at room temperature. During the reaction, portions of the mixture were removed and analyzed by GPC, though no further reaction appeared to occur following the addition of the final amount of POE-OCOCl. A white solid (1.0926 g, 90% yield of total reaction mixture) was recovered as product after ultrafiltration (YM 10 membrane) and lyophilization. The product was fractionated using semipreparative GPC. Purification of the Eglin c:POE Conjugates-Typically, 0.1-0.35 g of the products from the above experiments were fractionated by GPC using a TSK G3000 PW semipreparative column (mobile phase = 0.1 M LiBr,, flow rate = 5.0 m L h i n ) , employing a refractive index detector. The fractions were collected by a Gilson model 201 fraction collector at suitable time intervals ke., 0.5 or 1.0 m i d . The individual fractions were examined by UV spectroscopy (A = 275 nm). Those fractions which were found to be substantially free of unreacted Eglin c were combined (cf., Table I). To remove all of the eluant, the combined fractions were dialyzed as before. Upon lyophilization, white solids (6-150 mg) were obtained. Eglin c:POE Conjugate Characterization-The Eglin c:POE conjugates prepared and purified as described above were examined using the following techniques: IR analysis; UV analysis; C, H, N microanalysis; amino acid analysis; analytical GPC (using a TSK Table CEglln c Content and Actlvlty of Combined Eglln c:POE Fractions' Conjugate

Eglin c:POE20

Sample Eglin c in Active . Specific gl'n in Activity, (Combined Amount' Conjugate % (w/w)~'Conjugate' Yod Fractions) mg Yo (W/W)C

5-8 9-10

5-8 9-1 0 1-4

5-8 He 9-1 0 %loe Eglin c:POE35

5-9 10-12 1-4 5-10

11-20

48.0 64.2 41 .O 6.0 16.8 11.2 11.2 5.0 5.0 56.8 41.0 130.8 32.5 153.4

16.6 9.5 10.6 11.6 3.0 2.6 2.8 3.4 3.2 14.2 13.3 28.8 22.1 23.6

1.7 1.4 1.1 1.1 2.2 1.2 2.2 1.2 2.4 9.7 7.6 14.7 9.5 11.4

30.4 14.7 61.1 34.4 73.3 46.1 78.6 35.3 75.0 68.3 57.1 51.0 43.0 48.3

Obtained from preparative GPC of crude reaction mixtures. The amount of Eglin c was determined by an ELISA assay and is expressed as a percentage of the total weight of the combined fractions. 'The amount of "active" Eglin c, which is defined as the ability of Eglin c to inhibit the action of elastase, was determined using a chromogenic substrate degradation assay; the results are expressed as a percentage of the total weight of the combined fractions. dThe ratio of "active" Eglin c to the amount of Eglin c determined in the combined Eglin c-POE fractions. eAfter overnight dissolution.

1172 / Journal of Pharmaceutical Sciences Vol. SO, No. 12, December 1991

3000 PW column and 0.01 M aqueous LiBr as eluant, with detection by refractive index and a t 275 nm); and reversed-phase HPLC [lOpm C18 pBondapak column, with an acetonitrile:O.l% (dv) TFA:water gradient as eluant and detection at 277 nml. The presence of Eglin c in the purified conjugates was confirmed by the presence in the IR spectra of a carbonyl stretching frequency at 1650 cm-' (which is absent in the spectrum of POE). In addition, the conjugates absorb light strongly at 275 nm, whereas POE does not absorb light above 240 nm. Amino acid analysis of the conjugates also confirmed the presence of Eglin c (e.g., expectedfound amino acid ratios for the Eglin c:POE20 conjugate were as follows: Asp, 1.03;Thr, 1.09; Ser, 1.18Glu, 0.98; Pro, 0.97; Gly, 1.09; Ala, 1.28; Val, 1.00; Leu, 1.02; Try, 0.97; Phe, 0.98; His, 0.94; Lys, 1.11; Arg, 1.02; Gln and Am were not determined). By microanalysis (C and N determination), the conjugates were found to contain a minimum of two POE molecules attached to one Eglin c molecule (expected C:N ratio = 32) or seven POE chains to two Eglin c molecules (expected C:N ratio = 50). For example, by analysis, C:N ratios of 35 (C, 50.36; N, 1.44), 32 (C, 44.64; N, 1.39), 50 (C, 58.41; N, 1.17), and 49.5 (C, 40.07; N, 0.81) were obtained for the Eglin c:POE35 conjugates. Determination of the Amount of Eglin c in the Conjugates and Their Elastase-Inhibiting Activity-In order to evaluate the effect of grafting polymer chains to Eglin c on its properties, we determined the amount (mass) of the polypeptide in the conjugates using an ELISA technique. The ability of the conjugates to inhibit elastase activity was determined using a chromogenic substrate degradation assay. The details of both assays11 are briefly described below. ELJSA Protocol for Eglin c and Conjugates-A 200-pL aliquot of an aqueous stock solution containing 2.5 pg/L of an anti-Eglin c monoclonal antibody was added to each of 96 wells of an immuloncoated microtiter plate (Dynatech Corporation, U.K.). After incubation at 4 "C overnight, the wells were washed thoroughly with an aqueous phosphate "washing" buffer solution containing 0.05% (v/v) Tween 20 as a blocking agent. A dilution series of Eglin c standards and test samples were prepared in the washing buffer to which 2% (wlv) bovine serum albumin (BSA) had been added. Then, 250-pL aliquots of the solution were added to individual Eppendorf tubes. An Eglin c:peroxidase conjugate stock solution was diluted 1 : l O 000 with the BSA buffer solution and 250 pL was added to each Eppendorftube containing the antigen (as described above). After thorough mixing, 200 pL of the solution from each tube was removed and applied in duplicate to the wells of the microtiter plate. Appropriate blank control solutions were likewise applied to the same plate. Following incubation at 37 "C for 3 h, the plate was washed with aqueous phosphate buffer. Finally, to each well was added 200 pL of substrate solution consisting of 1.0 mL of 2% (w/v) 2,2'-azinobis(3ethylbenzthiazoline sulphonic acid)diammonium salt and 0.4 mL of 0.6% (vh) Hz02 in 18.6 mL of phosphate buffer (pH 4.5). Following incubation at 37 "C for 30 min, the absorbance from each well was measured spectrophotometrically at 410 nm. Elastase-Znhibiting Activity Assay-A series of test solutions and Eglin c standard solutions (0 to 10 pM) were prepared in aqueous buffer containing human serum albumin (pH 7.5). A 2 5 - 4 aliquot of a 10 pM porcine pancreatic elastase solution and 25 pL of each test or standard solution were added to the wells of a microtiter plate. The plate was covered, thoroughly shaken, and incubated for 15-30 min a t 37 "C. Then, 100 pL of MeOSuc-Ala-AlaPro-Val-pnitroaniline (Cambridge Research Biochemicals, Cambridge, U.K.) in 0.2 M "ris/HCl buffer was added to each well. After thorough mixing and incubation for 3 0 4 5 min a t 37 "C, the reaction was halted by the addition of 30 pL of glacial acetic acid to each well. The absorbance from each well was measured spectrophotometrically a t 410 nm. The activity of the conjugates was expressed as a percentage of activity that would be exhibited by the corresponding amount of free Eglin c. All of the conjugates were found to contain between 1.8 and 28.8% (w/w) Eglin c, with specific activities between 15and 80% of the free unconjugated Eglin c, as shown in Table I. Determination of Plasma Levels and Urinary Excretion of Eglin c and Its Conjugates in Mice-Five groups of adult male mice (OLMNIH), numbering three animals per group, were weighed and injected via the tail vein with 50 pL of a 1-mg/mL solution of Eglin c (in sterile, filtered, phosphate-buffered saline). Five other groups were dosed in a similar fashion with either the Eglin c:POE20 or Eglin c:POE35 conjugate (equivalent in each case to 1 mg/mL of conjugate). Two of the five groups of animals were placed in metab-

olism cages for the duration of the experiment (6 and 24 h, respectively) to enable urine to be collected over this period. Additionally, samples of urine were collected, where possible, from the animals sacrificed at earlier time points. At intervals of 10 min and 1,2,6,and 24 h after injection, the animals of a group were sacrificed by decapitation; trunk blood of each animal was collected into 200 pL acid citrated dextrose (ACD) in a preweighed vial. Blood volumes were calculated from the weight of blood collected.The blood samples were maintained on ice until they could be centrifuged at 1500 x g (Beckman TJ-R centrifuge at 4 "C).Plasma derived from centrifugation was stored frozen until the ELISA determinations could be carried out.11 The ELISA method was also used to determine the concentration of Eglin c and its conjugates in the plasma and urine samples collected from the in vivo experiments. Determination of Plasma Level of [12611POE20in Mic-Adult male mice (OLAMIH),three to four animals per group, were weighed and injected via the tail vein with a solution containing 0.22 rJ. of phosphate buffer and 50 pL of ['2611POE20 conjugate solution (originally 7 mg/mL, 400 pCi/mL, diluted 1 5 0 with non-radioiodinated POE).The mice were sacrificed after 1, 2, and 24 h, and their trunk blood collected. The radioactivity was determined using a Packard Auto-Gamma 5650 counter.

Results and Discussion Preparation of Eglin c:POE Conjugates-In the late 1970s and early 19808,Abuchowski et al.12913 were able to demonstrate that the residence time of enzymes in the vascular compartment could be increased by attaching MeOPOE to the protein chains. Other water-soluble, usually uncharged or negatively charged, polymers have since been suggested and ~sed.14~16 In addition to increasing the residence time in circulation, it was also demonstrated that the attachment of uncharged synthetic macromolecules to proteins could greatly reduce the immunogenicity of a protein.1&1* Both are desirable properties for potential therapeutic use of proteins, whether they are enzymes or not. Eglin c has in its structure two amino groups arising from the presence of two lysines (cf., Figure 1).Such amino groups can be expected to undergo a facile reaction with activated, water-soluble polymers. For low molecular weight POE (M, 5 5000),there are at least three common methods of polymer activation which involve the use of cyanuric chloride,lS 2,2,2-trifluoroethanesulfonylchloride (tresyl chloride),20and carbonyldiimidazole.21 (A further activated derivative, the N-succinimidyl succinate of POE is also available commercially from Enzon, Inc., according to Shafer and H a r r i s P Eglin c was reacted successfully with POE (M,2 20 000 and 35 000) activated as the carbonyl chloride derivative (POE-OCOC1) after the method of Mutter et al.9 and Artursson et a1.10The reaction between POE-OCOCl and Eglin c was carried out in aqueous conditions at pH 8.0. The experimental conditions reported here are not optimal. In an attempt to react all of the Eglin c with the POE-OCOC1, the activated POE was used in excess in all but one experiment. With time, in aqueous conditions, we observed that the POE chloroformate group is hydrolyzed. Therefore, it seemed prudent to add the activated POE in portions to the reaction mixture. In these experiments, the final amount of added POE-OCOCl was in a fourfold molar excess over the amount of Eglin c. The initial reactivity of the activated POE is critical for complete reaction, and this should be assessed prior to the grafting experiment. The linkage between the two components is likely to be a urethane bond, though it is possible that any sidechain carboxyl or C-terminal carboxyl groups will also react to give a carbonate bond. The former linkage would be preferred in our case, since it would be more resistant to degradation. The experimental conditions employed here are more likely to favor reaction of the activated POE with the amino groups. As yet, we have not investigated the nature of the reaction in

detail. Evidence for the formation of the P0E:Eglin c conjugate is provided by a number of conventional analytical techniques. Confirmation of the presence of Eglin c was provided by UV spectroscopy at 275 nm and also by IR spectroscopy (see Experimental Section). Analysis of the reaction mixtures by aqueous phase GPC during the reaction indicated that Eglin c was being consumed (see discussion below). Eglin c:POE Purification and Characterization-The reaction products were examined initially by aqueous phase GPC. As a control, we chromatographed a mixture of POE and Eglin c (Figure 2). The two compounds do not co-elute (as detected by refractive index); further, POE does not absorb UV light above 240 nm, whereas Eglin c absorbs light strongly a t 275 nm. This property has therefore been utilized to confirm the above result. Under the conditions used here, Eglin c and POE can be easily separated by GPC. In Figure 3, it can be seen that Eglin c is consumed during the reaction with 35 000 POE-OCOC1 to give a high molecular weight material that absorbs light a t 275 nm. A much higher molecular weight shoulder was formed, which is consistent with the preparation of a high molecular weight conjugate fraction. A consequenceof this is that the polydispersity of the product(s) is considerably greater than that of the original starting POE. Similar results were obtained from the reaction between 20 000 POE-OCOC1 and Eglin c. The excess POE used in the reaction to prepare the conjugates is not readily removed and the product therefore must contain POE, as well as the Eglin c:POE conjugate. As stated earlier, the reaction conditions used in the preparation were not optimal, and it is desirable that, in the future, conditions are found which will minimize, if not eliminate completely, the presence of free unreacted POE from the product. Of greater importance was the removal of noncovalently bound Eglin c, and this is described below. To remove the unreacted Eglin c, the reaction mixtures were fractionated using semipreparative aqueous GPC. The results of fractionation are presented in Table I. All of the fractions had measurable extinction a t 275 nm (i.e., in a region where POE is transparent) and formed white solids upon freeze-drying. Suitable combined fractions (Table I) were examined by the analytical GPC column (Figure 4).The early eluting fractions (e.g., 1-10]do not appear to contain (as judged by chromatographic evidence) any free unreacted Eglin c; however, they are UV active, indicating the presence of the polypeptide. The elution times of the various pooled

>

iA /\

€1

0

15

25

Time (min) Flgure 2-Analytical GPC trace of a 1% (w/v) solution of POE (M, = 20 000) and Eglin c. The mixture of the two components is shown clearly

not to co-elute from the column and is thus potentially easily separable by chromatography. Detection is by (A) UV light (275 nm) and (6) refractive index. The POE elutes after -13 min, as shown in (A), and Eglin c elutes after -18 min, as shown in A and 6 . Journal of Pharmaceutical Sciences I 1173 Vol. 80, No. 12, December 7991

I A

1 0

n

iA

n

J\

1

I

15

25

TIME (MIN)

Figure >Consumption of Eglin c during the course of the reaction with POE-OCOCI (M, = 35 OOO) examined by analytical GPC using refractive

index detection. The POE-OCOCI was added in 200-mg batches every 30 min to a solution of Eglin c (200 mg) until a total of 1 g of the chloroformate had been added. The GPC traces were obtained (A) 1.5 h, (B) 2 h, and (C) 2.5 h after the start of the experiment. The conjugate elutes after -13 min and Eglin c elutes after -18 min. fractions ( 1 4 , 5-8, and 9-10] vary as expected, with the combined fractions 1 4 eluting slightly earlier from the analytical column than the combined fractions 9-10. The higher molecular weight fractions were examined for the presence of Eglin c by biochemical assays (see Experimental Section and Table I) and by HPLC. Reversed-phase HPLC was used to ascertain whether the products of the reaction between Eglin c and POE-OCOCl were covalently bonded conjugates or simply mixtures of the two components. Eglin c elutes as a siqgle component from a 10-pm C,, pBondapak column, using an acetonitri1e:water [with 0.1% (dv) TFAI gradient and detection at 277 nm. However, the POE (derivatized with tyramine end groups for the ease of detection) does not elute from this column under the same chromatographic conditions. This is most likely due to irreversible adsorption of POE onto the bonded phase. Nevertheless, this method is very suitable for separating and detecting free Eglin c. The two products mentioned above were both chromatographed using this method and in both cases no unreacted Eglin c was detected. To overcome possible adsorption of the conjugates on this column, other packing materials were examined which either had less hydrophobic bonded phases or lower carbon loadings. The columns tried were (i) TSK G5000PW Pheny14.6 mm x 7.5 cm, (ii)Nucleosil lop C, 4.6 mm x 25 cm, and (iii) Protesil 300 Diphenyl 4.6 mm x 25 cm. Various gradients of acetoniti1e:water [with 0.1% (v/v) TFAI were used, with detection being made a t 277 nm (Lax for the derivatized POE). Eglin c, but not the derivatized POE, eluted from all of these columns as a single peak. In summary, the analysis of the “purified” Eglin c:POE conjugate by reversed-phase HPLC testified ta the absence of unreacted Eglin c, which is in agreement with the aqueous GPC results. Prior to in vivo studies, the activity of the purified Eglin c:POE conjugates was determined by a chromogenic substrate degradation assay as described by Daw et al.11 The conjugated polypeptide retained between 15 and 80% of the activity of Eglin c (the results are shown in Table I). Since the fractionated conjugates contain excess POE, it was ascertained that the presence of POE does not interfere with the ELISA or functional assay of Eglin c. Given that the polydispersity in molecular weight of the original POE and the polydispersity of the fractionated Eglin c:POE are similar (approximately ~ 1 . 5 )it, is to be expected that there will be some variation in the specific activities of the conjugated Eglin c; the values given in Table I can only represent “average” activity. What perhaps is surprising is that overall a greater amount of covalently bound Eglin c contained in the Eglin c:POE 1174 I Journal of Pharmaceutical Sciences Vol. 80, No. 12, December 1991

I 25

0

Time (min) Flgure 4-Analytical GPC trace of the combined fractions of Eglin c:POE20conjugatesobtained following fractionation of the crude product using preparative GPC. The combined fractions were examined by analytical GPC and detected by (a) refractive index and (b) UV light (275 nm). The combined fractions (A) 1 4 , (B) 5-8, and (C) 9-10 were eluted after 13, 14.6, and 15 min respectively. All are shown to be free of noncovalently bound Eglin c contamination.

conjugate does not result in greater Eglin c activity. This may be as a consequence of the reaction conditions used to prepare the conjugates, which possibly caused inactivation of the bound Eglin c. However, it is worth noting that, in two instances upon allowing the Eglin c:POE conjugates to “relax” in solution overnight, a n increase in Eglin c activity was determined. The loss in Eglin c activity upon conjugation may well be attributable to intramolecular crosslinking of the polypeptide. The use of cyanuric chloride-activated POE has caused loss in activity upon conjugation with enzymes, and this loss has been ascribed to the formation of intramolecular crosslinking.21In our experiments, we have used bifunctional activated POE which also has the potential to crosslink Eglin c both inter- and intramolecularly. Observations made by Beauchamp et a1.22 led to the conclusion that the use of low molecular weight bifunctional POE (activated by 1,l’carbonyldiimidazole), when coupled to superoxide dismutase (SOD), did not lead to gross deactivation of the protein. However, this cannot be pertinent to our case, since we used a much higher molecular weight POE coupled to a very different polypeptide (of much lower molecular weight than SOD).For the animal studies reported below, the Eglin c:POE

conjugates with -70% specific activity of the free unconjugated Eglin c were used. Animal Studies-Protein or polypeptide drugs are often by necessity administered parenterally in vivo. Intravenously administered Eglin c is rapidly eliminated from the bloodstream via a process of excretion via the kidneys. In our experiments, we have determined the plasma persistence and urinary excretion of Eglin c and Eglin c:POE conjugates by following the fate of Eglin c in mice (see Experimental Section). The presence in blood and urine of Eglin c conjugated to POE was determined by an ELISA assay. Figure 5 gives the results of plasma retention in terms of percent of injected dose in mice. These results have been derived by taking into account the weights of the animals and the volume of anticoagulant used. Also, we have assumed a hematocrit of 0.44 for mouse blood (a figure that we have previously derived for this strain of mouse) and a blood volume-to-weight ratio of 0.074 mug. This enables direct comparisons to be made between the various samples from different animals. The stability of the Eglin c:POE conjugates in plasma is indicated indirectly by the results shown in Figure 5. Since we determined the fate of Eglin c in blood and not the presence of POE, if free Eglin c had been liberated in any great quantity we would have expected to obtain curves for the conjugate similar to that for Eglin c alone. The initial distribution phase is so rapid for Eglin c that, even at the earliest time point studied, ~ 2 0 % of the injected dose is present in plasma (see Figure 5). These figures do not take into account the volume of distribution of Eglin c. As mentioned previously, it is the interstitium and not the plasma compartment where the inhibiting action of Eglin c is required. The observed disappearance of Eglin c from plasma reflects not only the excretion of the polypeptide via the kidneys, but also its distribution into at least one other body compartment, the interstitium. The general distribution of proteins prevailing in a 65-kg human is given in Figure 6 by way of illustration.23 At any time, there is more protein found in the interstitiurn as

PLASMA 3L 7

0

INTERSTITIAL FLUID

If 1;~

12 L

20-30 g / L

--II

\ ( 4 - 8 ~ultraliltrate/doy

6og/L LYMPH

I(obsorbed by nodal blood ,! circulation)

AFFERENT LYMPH 12L/doy 20g/L

or BL/day

I

30g/L

Figure &Lymphatic turnover of fluid and plasma protein in a 65-kg human (with permission from ref. 23).

compared with that in plasma (the quantities are 210 g in plasma versus 240-360 g in interstitial fluid). The same ratio of distribution between the two volumes can be expected to exist for other macromolecules of similar size (e.g., Eglin c:POE conjugates); this does not take into account the possibility of the conjugate binding to the accessible elastase. It would be prudent to verify this argument by measuring directly the concentration of the conjugates in the interstitium. The excretion of Eglin c and conjugates in urine is shown in Figure 7. It should be noted that mass balance has not been achieved in these experiments. This is due in part to the conditionsunder which the urine was collected, but there may well be other factors which give rise to this result. What ie clear is that very low concentrations of the Eglin c:POE conjugates appear in the urine during the 24-h period of collection. In general, a macromolecule >4 nm, but 70% of the activity that would be expeded from the amount of Eglin c contained in them have been prepared. The conjugates retained their elastase-inhibiting activity throughout the course of this study (5 months). When the conjugates were stored as freezedried samples, we observed that their activity depended on

U

"12

6

48

24

Time (hr) Figure 8-Comparison between the plasma retention of (A)Eglin ,-Dnt=m a d /-&-i r 1 2 5 1 1 ~ n ~fnllnwinn c~n

i\r aArninictrntinn tn mien

1176 I Journal of Pharmaceutical Sciences Vol. 80, No. 72, December 1997

the length of time allowed for the dissolution of the sample (i.e., much higher activity was observed after overnight dissolution as compared with 2 h). Using mice, we have shown that the amount of Eglin c conjugate found in plasma after iv administration is much higher at any time after administration as compared with unconjugated Eglin c. Typically, -17% of the injected dose of the conjugated Eglin c was found in plasma after 24 h using the conjugates reported here. The time course of the conjugate plasma concentration matches closely that of the corresponding free polymer. Consequently, we can expect that higher plasma concentration could be achieved, if and when required, by simply selecting the polymer of appropriate size. The synthetic polymer selected for this work (POE) has been used by others (e.g., Enzon, Inc.) for a similar purpose. We do not think that this is the only polymer that could be used for such modifications. Considerations will need to be given to both commercial (e.g., patent) and application (e.g.,biological, chemical) aspects in selecting alternatives (e.g., proteins such as albumin or polysaccharides such as dextran).

References and Notes 1. Blohm, D.; Bollschweiler, C.; Hillen, H. Angew. Chem. Znt. Ed. Engl. 1988,207-225. 2. Snider, G. L.; Stone, P. J.; Lucey, E. C.; Breuer, R.; Calore, J. D.; Seshadri, T.; Catanese, A.; Maschler, R.; Schnebli, H-P. Am. Rev. Respir. Dis. 1985, 132, 1155-1161. 3. Rink. H.: Liersch. M.: Sieber. P.: Mever. F. Nucl. Acids Res. 1984. 12,6369L6388. 4. Guyton, A. C. Textbook of Medical Physiology; W. B. Saunders: Philadelphia, 1986; p 371. 5. Petrak. K.: Goddard. P. Adv. Drug Deliv. Rev. 1989.3, 191-214. 6. Rypacek, F.; Drobnik, J.; Chmelir, V.; Kalal, J. Pfiugers Arch. 1982,392,211-217. 7. Seymour, L. W.; Duncan, R.; Strohalm, J.; Kopecek, J.J.Bwmed. Mat. Res. 1987,21,1341-1358. 8. Goddard, P.; Williamson, I.; Brown, J.; Hutchinson, L. E.; Nicholls, J.;Petrak, K. J.Bioact. Compat. Polym. 1991,6,4-24. 9. Mutter, M.; Altmann, K-H.; Gehrhardt, H. React. Polym. 1987,6, 99-107. 10. Artursson, P.; Brown, L.; Dix, J.;Goddard, P.;Petrak K. J.Polym. Sci., Part A: Polym. Chem. 1990,28,2651-2663. 11. Daw, A. P. W.; Gpddard, P.; Petrak, K.; O'Mullane, J.; Alkan, S. S., in preparation. 12. Abuchowski, A.; McCoy, J. R.; Palczuk, N. C.; van Es, T.; Davis, F. F. J. Biol. Chem. 1977,252, 35823586. 13. Abuchowski, A.; Davis, F. F. in Enzymes as Drugs; Holcenburg, J.: Roberts. J.: Eds.: Wilev. New York. 1981; PP 367383. 14. Wileman, T.; Bennett, M:;'Lilleyman,.J. J. Phirm. Pharmacol. 1983,35.762-765. 15. Melt&, 'R. G.; Wiblin, C. N.; Foster, R. L.; Sherwood, R. F. Biochem. Pharmucol. 1987,36, 105-112. 16. Abuchowski, A.; van Es, T.; Palczuk, N. C.; Davis, F. F. J. Biol. Chem. 1977,252,3578-3581. 17. Davis. F. F.: Abuchowski. A.: van Es. T.: Palczuk. K.: Savoca., R.:. Chen,'H-L.;'Pyatak, P. Polym. Prep;. 1979,20, 357260. 18. Suzuki, T.; Ikeda, K.; Tomono, T. J. Biomuter. Sci., Polym. Ed. 1989,1, 71-84. 19. Shafer, S. G.; Harris, J. M. J. Polym. Sci., Polym. Chem. Ed., 1986,24,375-378. 20. Nilsson, K.; Mosbach, K. Methods Enzymol. 1984,104, 56. 21. Yoshinga, K.; Harris, J. M. J. Bioact. Compat. Polym. 1989, 4 , 17-24. 22. Beaucham , C 0 ; Gonias, S. L.; Manapace, D. P.; Pizzo, S. V. Anal. BiocRem: 1983,131,2533. 23. Renkin, E. M. A m . J.Physiol. 1986,250, H70&H710. I

,

,

I

Acknowledgments We thank Dr. Hans-Peter Schnebli (Ciba-Geiy,Basle)for supplying the Eglin c, and Professor Sefik Alkan (Ci a Geigy, Basle) for discussions on the development of the assays.