protamine-based delivery system for enzyme drugs

Journal of Controlled Release 78 (2002) 67–79 www.elsevier.com / locate / jconrel

ATTEMPTS: a heparin / protamine-based delivery system for enzyme drugs J.F. Liang, Y.T. Li, H. Song, Y.J. Park, S.S. Naik, V.C. Yang* Department of Pharmaceutical Sciences, College of Pharmacy, The University of Michigan, 428 Church Street, Ann Arbor, MI 48109 -1065, USA Received 20 February 2001; accepted 29 May 2001

Abstract A prodrug delivery system termed ‘‘Antibody Targeted, Triggered, Electrically Modified Prodrug-Type Strategy (ATTEMPTS)’’ has been developed to permit the antibody-directed administration of inactive enzyme drug including tissue-type plasminogen activator (tPA), and allow a subsequent triggered release of the active tPA at the target site. Cation-modified tPA (mtPA) was attached to a heparin–antifibrin complex via ionic interaction, and the active tPA can subsequently be released by the addition of protamine, a competitive heparin inhibitor. Anti-fibrin IgG was conjugated to heparin via an end-point attachment to form the heparin–antifibrin complex which provides the targeting efficiency of the final heparin / mtPA complex. Cation modification was performed by either chemical conjugation by linking (Arg) 7 Cys to tPA with N-succinimidy-3-(2-pyridyldithio) propionate or by recombinant DNA methods. Results show that the modification process did not significantly alter the specific activity of tPA with regard to plasminogen activation, fibrin-binding ability, and response toward fibrinogen. The complexes of both modified tPA–heparin did not yield any intrinsic catalytic activity owing to the blockage of the active site of tPA by the attached heparin. On the other hand, heparin-induced inhibition of modified tPA activity was reversed by adding protamine, which is similar to that of a prodrug delivery system. These results suggest that heparin / protamine-based enzyme delivery systems may be a useful tool to improve current enzyme therapeutic status, as well as thrombolytic therapy, by both regulating the release of active enzyme and aborting the associated systemic toxic effect. Currently, modification of enzyme drugs has been optimized by recombinant DNA technology assisted by computer simulation. In addition, the original strategy has been revised to obtain enhanced therapeutic efficacy.  2002 Elsevier Science B.V. All rights reserved. Keywords: ATTEMPTS; Prodrug; Tissue-type plasminogen activator; Heparin; Protamine; Thrombolytic therapy; Enzyme therapy

1. Introduction As therapeutic drugs, enzymes possess several attributes such as high specificity towards substrate, *Corresponding author. Tel.: 11-734-764-4273; fax: 11-734763-9772. E-mail address: [email protected] (V.C. Yang).

high solubility for preparing liquid formulations, and optimum activity under physiological conditions. However, despite these advantages, enzyme therapy has been beset by several limitations and only a handful of enzymes such as thrombolytic agents and L-asparaginase have currently been approved for clinical use [1,2]. The primary reasons for the limitation of clinical applications lie in the immuno-

0168-3659 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 01 )00484-9

68

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

genicity and indiscriminate nature of almost all enzymes. Although the immunogenicity of an enzyme can be circumvented to a certain degree, the inability to distinguish between target and normal substrates, however, would yield unwanted toxic effects and thus remains as the major hurdle for its use as a clinical drug [3–5]. While conversion of the target substrate by an enzyme drug would result in positive therapeutic efficacy, conversion of a normal substrate other than the target would lead to systemic side effects. For example, plasminogen activators (PA) have been utilized as potent thrombolytic agents since they dissolve thrombus by converting plasminogen to plasmin with high fibrin selectivity [6,7]. However, despite their superior fibrinolytic activity, administration of PAs during thrombolytic therapy also carries the risk of hemorrhage as its major side effect, which is primarily due to the indiscriminate nature of the enzyme in attacking both the fibrinbound and circulating plasminogen. In addition to thrombolytic agents, anticancer agents also possess the same indiscriminate behavior toward target and normal tissue, which makes it difficult to achieve a desirable therapeutic efficacy. Bestowing upon an enzyme the ability to distinguish between target and normal substrate should thus be the key to success in clinical application. Of all the approaches examined to date, the antibodytargeting approach appears to be most promising, because it consistently and substantially enhances the specificity and selectivity of an enzyme drug. One proposed method of selective drug delivery by using an antibody–enzyme is a two-step approach called antibody-directed enzyme prodrug therapy (ADEPT) [8–11]. In ADEPT, selectivity for the target is achieved by an antibody in an antibody–enzyme conjugate that binds antigen preferentially expressed on the surface of target cells. An administered antibody delivers an enzyme selectively to a tumor deposit. A relatively nontoxic prodrug is then administered systemically. The prodrug is catalyzed by a pre-localized enzyme at the tumor site to produce a potent cytotoxic agent. The generated active agent then diffuses throughout the tumor mass, killing tumor cells. There are advantages of ADEPT when compared to conventional therapy [8,9]: (1) there is increased selectivity for tumor cells owing to the specificity and targeting ability of the antibody; (2) there is an amplification effect since each enzyme

molecule is able to cleave a large number of prodrug molecules; (3) the concept has proved to be feasible in the clinic; (4) the concentration of the drug delivered to tumor has been demonstrated to be higher than with direct injection of the drug. Aside from the applicability demonstrated above, however, the ADEPT system has a limitation in selection of drugs. Prodrug formation and conversion to an active drug is mediated by chemical conjugation and enzyme cleavage. Drug candidates are limited to low molecular weight drugs due to the ease of chemical conjugation as well as the presence of one enzyme cleavage site, thus conversion to the active drug is consistent and predictable. Large macromolecules such as enzymes are less suitable for the ADEPT system for several reasons. Macromolecules contain active groups which are more conducive for chemical conjugation, however the presence of multiple active groups can lead to heterogeneous chemical conjugation. Also, the process of chemical conjugation can be harsh to protein drugs, leading to decrease or loss of enzymatic activity. In addition, enzyme drugs are large by nature and may contain cleavage sites that may be cleaved by the enzyme used to activate the prodrug. Thus, it has not been possible to apply ADEPT to the delivery of high molecular weight drugs, including enzyme drugs. Therefore, it is essential to develop a delivery system for targeted enzyme therapy. The authors have developed a novel enzyme delivery system named ‘‘antibody targeted triggered electrically modified prodrug type strategy (ATTEMPTS)’’ (Fig. 1), which permits the administration of an antibody-directed but inactive enzyme drug, followed by triggered release of the active enzyme at the target site [12–15]. Herein, tissue-type plasminogen activator (tPA) was utilized as an enzyme drug. Preliminary studies concerning the feasibility of the proposed enzyme delivery system with heparin-binding peptide were carried out in the delivery of a current clinically used enzyme drug, tissue plasminogen activator (tPA). Similar to a prodrug-type approach, the modified tPA would not possess proteolytic activity during administration, thereby alleviating the bleeding risk produced by systemic activation of circulating plasminogen. Since hindrance of activity is mediated by ionic binding, and the conversion from prodrug to active enzyme is based on competitive electrostatic binding, complete

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

69

(S-2251) was purchased from Pharmacia Hepar (Franklin, OH, USA). Murine anti-fibrin and anti-tPA IgGs were from American Diagnostica (Greenwich, CT, USA). Human plasminogen (plasmin free), plasmin, fibrinogen, N-succinimidyl 3-(2pyridyldithio) propionate (SPDP), and heparin immobilized Sephadex beads were obtained from Sigma (St. Louis, MO, USA). The octapeptide containing seven arginine residues and a cysteine residue at the C-terminal [i.e. (Arg) 7 Cys] was synthesized and purified by the Biomedical Research Core Facilities at the University of Michigan (Ann Arbor, MI, USA). Double-stranded oligonucleotide for expressing poly(Arg) 7 : 59-CGCCGTCGACGGCGCAGAAGGCGCGGCGGCAGCTGCCGCGTCTTCCGCGCCGGC39 which contained two NarI restriction enzyme sites at the two ends and one SalI site in between the sequence, was synthesized and purified by Integrated DNA Technologies (Coralville, IA, USA). Plasmid PtPA-trp12 for expression of tPA in E. coli and bacterial strains HB101 and RB791 were obtained from ATCC (American Type Culture Collection. Rockville, MD, USA). All solvents were of analytical grade.

2.2. Preparation of cation-modified tPA

Fig. 1. Schematic diagram of the ATTEMPS approach.

reversal to an active enzyme could be anticipated. Herein, we discuss the details concerning the development of the ATTEMPTS method, and its optimization, including enzyme modification using computer simulation.

2. Materials and methods

2.1. Materials Recombinant tPA (Alteplase) was purchased from Genentech (South San Francisco, CA, USA). The chromogenic substrate D-Val-leu-lys-p-nitroanilide

Modification of tPA with (Arg) 7 Cys was performed by both chemical and biological conjugation methods. In the chemical conjugation method, (Arg) 7 Cys and tPA were linked with the heterobifunctional cross-linking reagent SPDP using a two-step procedure described previously (Fig. 2) [14]. Modified tPA (mtPA) was separated from unmodified tPA by passing through a heparinSephadex column (530.5 cm). For the biological conjugation method, those DNA fragments encoding poly(Arg) 7 peptides, tPA, try promoter, gene conferring antibiotic resistance and the double-stranded oligonucleotide encoding the poly (Arg) 7 sequence, were linked with T4 ligase, followed by transformation into HB101 bacterial strain (Fig. 3). The recombinant plasmid ( ptPA-polyArg) was expressed in RB791 bacteria strain as previously described [15]. After harvest, the sample was then loaded onto a column and eluted with a buffer containing a linear gradient of imidazole. The EmtPA fractions were

70

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

Fig. 2. Schematic diagram of the synthesis of modified tPA using tPA and (Arg) 7 Cys.

collected after chromatography and then concentrated using an Amicon (Beverly, MA, USA) concentrator.

2.3. Preparation of heparin /antifibrin IgG conjugate The hydroxyl groups of heparin and the sugar moieties of anti-fibrin IgG were oxidized by sodium nitrite and sodium peroxyiodide, respectively, to produce aldehyde groups. A solution (0.5 ml) of the peroxyiodide-treated anti-fibrin IgG was mixed with 0.2 ml of 100 mM adipic acid dihydrazide, the mixture was tumbled and kept at room temperature for 2 h. Excess adipic acid dihydrazide was removed by gel filtration. A solution (0.5 ml) of the above anti-fibrin IgG was mixed with 0.1 ml of the NaNO 2 -treated heparin (5.0 mg / ml), and the mixture was tumbled overnight at 48C. Excess heparin was removed by passing the reaction mixture through a Sephadex G-100 (2530.8 cm) column equilibrated with 0.15 M NaCl. The heparin / anti-

fibrin IgG conjugates were purified by passage through a protamine-Sepharose (2530.8 cm) column and eluted with 2.0 M NaCl. Heparin / anti-fibrin IgG conjugates were analyzed by SDS–7.5% polyacrylamide gels under nonreduced conditions, and the anti-fibrin activity of the heparin / anti-fibrin IgG conjugate was measured by enzyme-linked immunosorbent assay (ELISA).

2.4. Spectrophotometric assay of plasminogen activation Plasminogen activation of tPA was determined by monitoring plasmin production using its specific chromogenic substrate S-2251. The initial rate of S-2251 hydrolysis in TBS (50 mM Tris–HCl containing 50 mM NaCl and 0.01% Tween 80, pH 7.2) at 258C was determined by measuring the absorbance at 405 nm at different time intervals using a microplate reader (BioRad, Hercules, CA, USA) and expressed as an absorbance per time square (A 405 / min 2 ).

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

71

Fig. 3. Construction of the recombinant PtPA-(Arg)7 plasmid for expression of mtPA.

2.5. Prodrug features of the mtPA–heparin complex The prodrug features of the mtPA–heparin complex were examined by measuring the plasminogen concentration in plasma and the plasminogen activation capacity. First, plasminogen and a 2 -antiplasmin levels in plasma were determined according to the method described elsewhere [16]. In brief, 150 ml of

a 500 U / ml thrombin solution was mixed with 150 ml diluted plasma and the activated partial thromboplastin time (APTT) of the sample was measured using a Fibrometer (BBL FibroSystem, Cockeysville, MD, USA). The fibrinogen level in plasma was estimated from the measured APTT value, which is inversely proportional to the fibrinogen concentration. To examine plasminogen activation, a solution (0.1 ml) of mtPA in TBS was mixed with a suspen-

72

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

sion (0.02 ml) containing heparin-Sephadex beads and incubated at 258C for 5 min. The heparinSephadex beads were then rinsed with 0.1 ml plasma for 5 min, followed by subsequent washing with plasma (twice) and TBS (twice). Plasminogen activation of mtPA bound to the heparin beads was examined by determining the rate of hydrolysis of chromogenic substrate S-2251 [17]. The percentage inhibition of mtPA activity by heparin binding was estimated from the ratio of the mtPA activity ‘adsorbed to’ and ‘desorbed from’ the heparin beads.

2.6. Fibrin clot lysis assay Fibrin clot lysis assay was performed using fibrincontaining agarose plates to validate the catalytic activity of mtPA. Samples were applied directly onto the sample wells (3 mm diameter) prepared on solidified fibrin–agarose gel and incubated overnight at 308C. Fibrin lysis was quantified by comparing the size of the lysed zone around the sample wells with standards prepared by addition of known concentrations of tPA.

3. Results and discussion

3.1. The heparin /protamine-based enzyme delivery system As shown in Fig. 1, this system is comprised of a large complex made of two components: (1) a targeting component consisting of an antibody chemically linked to an anionic heparin molecule; and (2) a drug component consisting of an enzyme modified by a cationic species. Because the cations used to mediate the binding of tPA were relatively small (e.g. a small positively charged peptide), the modified enzyme was able to retain a significant level of its catalytic activity. The modified enzyme, however, would be deprived of such activity once it binds to the antibody / heparin counterpart, presumably due to the blocking of the active site of the enzyme by these appended macromolecules. The two components are linked via a tight but reversible electrostatic interaction. Therefore, similar to a prodrug-type approach like the ADEPT system, the modified tPA would be without its proteolytic activi-

ty in circulation following its administration, thereby alleviating systemic side effects. After reaching the target thrombus via the attached antibody, the active modified tPA could then be released locally at the clot site by adding the triggering agent protamine, in contrast to the addition of an enzyme specific for a particular drug as in the ADEPT system. Protamine is the clinical heparin antagonist that is known to bind heparin stronger than most cationic species. The released active tPA would then be concentrated at the site of action, enabling it to yield greater catalytic activity while sparing the other systemic components. The prodrug and triggered release features of the ATTEMPTS delivery system were successfully demonstrated in previous reports. Cationic species with a specific heparin-binding strength were chosen so that the tPA / heparin–antibody complex could remain intact while in the circulation, whereas efficient release of modified tPA at the clot site could be achieved with the use of protamine. Conversion from prodrug to active tPA is based on competitive binding toward heparin, thus the binding between heparin and tPA and the subsequent triggered release of the active tPA depend on the insertion of an appropriate cationic moiety into tPA, which serves as the key to the success of this delivery system. In this regard, modification of tPA was accomplished by both chemical and biological conjugation, which will be dealt with in the following sections.

3.2. Modification of tPA by peptide (Arg)7 Cys It should be emphasized that two binding conditions must be met for the proposed heparin / protamine-based delivery system to function. First, in order for the modified tPA to remain attached to the heparin–antibody complex prior to reaching the targeted site, binding of the cation-modified tPA to heparin must be stronger than that of ATIII, which possesses the highest heparin-binding affinity throughout the circulation [18]. On the other hand, heparin binding to cation-modified tPA must be weaker than heparin binding to protamine, so that protamine can effectively release the modified tPA at the target site. Therefore, modification of the enzyme by a cationic species with an appropriate cationic density should be the key to successful protaminetriggered release. Based on previous results, at least

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

six arginine residues would be required to neutralize heparin and fully dissociate ATIII from its binding to heparin. Therefore, we proposed the use of a poly(Arg) 7 Cys peptide as the competitive heparin binder. During the early stages of ATTEMPTS research, chemical conjugation between cations and tPA was adopted. Since tPA contains only one free SH group [19], the (Arg) 7 Cys peptide was chemically linked to tPA and further improved by using the cross-linking agent N-succinimidyl 3-(2-pyridyldithio) propionate (SPDP). A cysteine residue was incorporated at the C-terminal of the peptide so that the peptide would be specifically linked to the tPA via the SH functional group (Fig. 2). Modification of tPA with the (Arg) 7 Cys peptide did not demonstrate significant alteration in its fibrin-binding ability, as plasminogen activation mediated by mtPA was similarly enhanced in the presence of fibrinogen. Compared with tPA, mtPA showed a much stronger heparin affinity, as the heparin / mtPA complex was quite stable in human plasma. The prodrug behavior of the mtPA– heparin complex was confirmed by the absence of depletion of the plasma plasminogen, a2-antiplasmin, and fibrinogen levels by mtPA, indicating massive activation of plasminogen to plasmin was

73

significantly inhibited by the heparin-bound mtPA in plasma (Fig. 4). In addition, it is noteworthy that as much as 95% of the amidolytic activity of mtPA was blocked once it bound to the heparin beads, confirming the prodrug feature of our proposed approach. An in vitro fibrin clot lysis assay, which mimicked the real clinical situation, was performed to demonstrate the feasibility of this heparin / protamine-based tPA delivery system. In contrast to free mtPA, the same amount of heparin-bound mtPA showed significantly less clot lysis. When protamine was added to the heparin-bound mtPA, however, the area of clot lysis increased, indicating that the heparin-induced blockage of the mtPA activity was reversed (Fig. 5). The inhibitory action of heparin on mtPA activity was also achieved for the heparin–IgG-59D8 conjugate, which was confirmed by chromogenic assay. The results showed that the heparin / anti-fibrin IgG conjugate was also capable of blocking the activity of mtPA and the inhibited activity was reversed by the addition of protamine (Fig. 6). It should be emphasized that heparin can be used to block the activity of mtPA (i.e., to achieve the prodrug feature of our approach), and this inhibition can be reversed by the addition of protamine. Therefore, these preliminary in vitro studies strongly suggest both the

Fig. 4. Percentage changes of plasminogen, plasmin and fibrinogen levels in plasma after 12 h incubation with mtPA-bound heparin beads (mtPA content 1.3 mg / ml). The same amount of free mtPA was used as a control.

74

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

Fig. 5. In vitro clot lysis studies. Wells 1–4 contain 0.025, 0.05, 0.1 and 0.2 mg of tPA, respectively; well 5 contains 0.2 mg of free mtPA; well 6 contains 0.2 mg mtPA bound to heparin beads; well 7 contains 0.2 mg mtPA released from heparin beads by the addition of 50 mg protamine; well 8 contains buffer only.

Fig. 6. Inhibition of mtPA activity by the heparin / anti-fibrin IgG conjugate and reversal of inhibition by protamine. The amidolytic activity of mtPA was measured using chromogenic substrate S-2251. Experimental conditions: mtPA, 0.5 mg / ml; plasminogen, 0.24 mM; S-2251, 1.2 mM; heparin / anti-fibrin IgG, 2 mg / ml; protamine, 20 mg / ml.

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

feasibility and utility of the approach in delivering enzyme drugs, particularly related to the prodrug and triggered release features. With these promising and remarkable preliminary results achieved by modification of tPA, the authors explored the potential of producing mtPA directly by biological conjugation using recombinant DNA technology (Fig. 3). Biological conjugation has several advantages over chemical conjugation: (1) the conjugation process can be easily controlled to obtain a homogeneous product with the desired properties in high yield; (2) the recombinant technique will alleviate the risk of random incorporation of the cationic peptide onto tPA as seen by the chemical conjugation method. This recombinant DNA technology produced a high yield of mtPA, which was confirmed by the presence of a protein band with a slightly higher molecular weight (by about 2000 Dalton) than that of native tPA in high yield. In addition, expressed mtPA (EmtPA) yielded an almost identical initial rate of plasminogen conversion as tPA, indicating that EmtPA retained the intrinsic catalytic activity of tPA. This catalytic activity of EmtPA was significantly inhibited by adding heparin (ca. 54% of initial activity). Heparin-induced inhibition was markedly

75

reversed by the addition of an equivalent amount of protamine (i.e. relative to heparin), indicating its prodrug feature (Fig. 7). This recombinant DNA technology was successful in producing cation-modified tPA directly and it opened up the possibility of direct production of the desired protein in high yield. Therefore, the authors attempted to optimize the recombinant technology method by producing mtPA that possessed the proper cationic sequence for ideal heparin binding at the appropriate insertion site.

3.3. Optimization of ATTEMPTS: modification of tPA assisted by computer simulation and recombinant technology The two essential elements required for the prepared modified tPA are: (1) it must possess a defined heparin-binding strength; and (2) the binding of heparin must be able to provide blockage of the catalytic domain or, in other words, inhibition of the activity of the modified tPA. To achieve this goal, our plan in preparing the desirable modified tPA was to attempt to modify the target surface loops of PA by mutation with the proposed cationic peptide sequence. While insertion of the peptides into folded

Fig. 7. Inhibition of the activity of biologically expressed modified tPA (EmtPA) by heparin and reversal of inhibition by protamine. Samples contained 0.5 mg / ml EmtPA, 0.24 mM plasminogen, and 1.2 mM S-2251.

76

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

areas of a PA protein would render them inaccessible to heparin binding, conjugation of the peptides at, or too close to, the active site of the PA could impair the catalytic function of the protease. Thus, the ideal situation is the insertion of the peptide onto surface regions that can cause steric hindrance of PA’s active site, which may lead to reversible inactivation of enzyme activity. Therefore, the biological conjugation method using mutation would alleviate the above limitations. The recombinant approach would permit homogeneous, site-specific incorporation of the peptides at desirable surface locations. Moreover, the biological method is well suited for mass production of the final products. Selection of the incorporation site is extremely important, as previous results suggest that an inappropriate location of the peptide insertion site can abort heparin binding or diminish the induced inhibitory effect upon heparin binding [20]. In general, the site should be spatially close to the protease active site such that bound heparin interrupts substrate interaction, and yet the modification must not significantly detract from the protease activity in the absence of heparin. In other words, the modification should be performed on a surface-exposed segment with maximum retention of the tertiary structure of the protease. In this regard, we applied computer simulation to identify the target sites which are surface-exposed regions that evolutionary and experimental evidence indicate would be tolerant to mutation or insertion. The overall strategy is therefore as follows: (1) identify proteins that share a high degree of sequence similarity to the catalytic domain of the selected PA species; (2) perform a multiple sequence alignment of these sequences to identify variable regions; and (3) select variable regions which are on surface-exposed loops in close proximity to the active sites. Using computer simulation and multiple sequence alignment of tPA with other similar proteins, the 37-loop and the 186-loop in tPA were identified as the two surface-exposed domains which should be tolerant to mutation (Fig. 8). Of the two loops, the 37-loop was selected as the site for incorporation of the peptide because it is an arginine-rich region and has also been implicated in the binding of PAI-1, a circulating tPA inhibitor. Creation of a heparin-binding segment in this region by changing three to four amino acids [21,22] through a very simple mutagenesis approach is

currently underway, and this may provide an additional advantage by disrupting PAI-1 binding, thereby resulting in the prolongation of the tPA half-life in the circulation. The crystal structure of tPA reveals ˚ from the that the 37-loop is approximately 20 A active site catalytic triad and is located at the edge of the active site cleft. Considering that a heparin dodecasaccharide with an average molecular weight ˚ in length, substrate binding to tPA of |5 kDa is 50 A would likely be prevented by a heparin molecule bound to the 37-loop [21–24]. Further analysis of the target incorporation site in tPA indicates that it contains a sequence of about nine to 10 amino acids that can be replaced by mutation without disrupting the overall fold of the protein. In this regard, a nonapeptide consisting of the sequence RCRRCPRRR was selected as the peptide to be incorporated into the target site by mutation. This sequence is similar to that of heparinbinding proteins (BXBBXXBBB) demonstrated by the previous report and to that in the 37-loop of tPA, which require minimal mutation steps to create [25]. On the basis of the above findings, four mutants were created, and characterized, in our laboratory (Fig. 9).

3.4. Future exploration While optimizing the ATTEMPTS system, it was anticipated that it would be more successful if the triggering agent stimulates drug activity as well as triggering active drug release from the complex. Research in our laboratory has shown that heparin can stimulate tPA activity, while protamine cannot [15]. A necessity when using heparin as a tPA binder is that the binding strength between modified tPA and heparin should lie in a range that is intermediate between that of heparin to ATIII and heparin to protamine. With this stringent requirement, the appropriate heparin-binding strength was difficult to regulate precisely. To overcome the hurdles of this approach, it may be possible to reverse the ATTEMPT strategy by using protamine to block the active site of the enzyme, followed by triggered release with heparin. Implementing this new strategy has two merits: (1) the binding strength of protamine to mtPA is not restricted to a particular range, as is the case when

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

77

Fig. 8. Tertiary structure of human tPA catalytic domain. The chain length of heparin tetrasaccharide is estimated based on literature data for heparin dodecasaccharide [22].

heparin is used as the binder; and (2) heparin can stimulate tPA activity in contrast to protamine. An additional advantage is that, by using heparin as a triggering agent, our strategy merges with routine thrombolytic therapy where heparin is often administered adjunctly with the plasminogen activator. The revised method closely resembles our original design, except for some minor adjustments. Briefly, instead of a cationic moiety attached to the plasminogen activator, an anionic peptide is conjugated by chemical or biological methods. To achieve a negative surface charge, we shall use a peptide containing either poly-aspartic acid or poly-glutamic acid residues, or a peptide that consists of a combination of the two amino acids. Upon delivery of the prodrug complex, triggered release may occur with heparin administration. Preliminary experiments concerning this new strategy are now in progress.

4. Conclusion ATTEMPTS, a heparin / protamine-based delivery system, has been developed for the delivery of enzyme drugs such as tPA without toxic side effects. It offers similar behavior to that of a prodrug delivery system, which allows the administration of inactive drug and release of the active drug by adding a release-triggering agent. In addition, optimization of the production of modified tPA based on recombinant technology assisted by computer simulation was accomplished. In conclusion, the ATTEMPTS approach can be applied to the design of other novel delivery systems for clinically important catalytic drugs (e.g. enzymes, protease, etc.), toxins, or even small or macromolecular drugs which possess strong side effects. Furthermore, future exploration of this system by revising the original AT-

78

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79

Fig. 9. Four mutants of tPA having different cationic peptide sequences (shown in underline) incorporated into the target 37-loop of tPA.

TEMPTS concept is in progress for enhancing therapeutic efficacy.

Acknowledgements Financial support from a NIH grant (HL 55461) and a Korea Science and Engineering Foundation (KOSEF) Postdoctoral Fellowship (to which Dr.

Yoon-Jeong Park was a recipient in 1999) are acknowledged.

References [1] N. Nagai, I. Vanlinthout, D. Collen, Comparative effects of tissue plasminogen activator, streptokinase, and staphylokinase on cerebral ischemic infarction and pulmonary clot lysis in hamster models, Circulation 100 (25) (1999) 2541–2546.

J.F. Liang et al. / Journal of Controlled Release 78 (2002) 67 – 79 [2] B.L. Asselin, The three asparaginases. Comparative pharmacology and optimal use in childhood leukemia, Adv. Exp. Med. Biol. 457 (1999) 621–629. [3] A. Lauwers, S. Scharpe (Eds.), Pharmaceutical Enzymes, Marcel Dekker, New York, 1997, p. 203. [4] R.B. Pedley, S.K. Sharma, G.M. Boxer, R. Boden, S.M. Stibbling, L. Davis, C.J. Springer, R.H. Begent, Enhancement of antibody-directed enzyme prodrug therapy in colorectal xenografts by an antivascular agent, Cancer Res. 59 (16) (1999) 3998–4003. [5] T. Sawa, J. Wu, T. Akaike, H. Maeda, Tumor-targeting chemotherapy by a xanthine oxidase–polymer conjugate that generates oxygen-free radicals in tumor sites, Cancer Res. 60 (3) (2000) 666–671. [6] H.R. Lijnen, D. Collen, Strategies for the improvement of thrombolytic agents, Thromb. Haemost. 66 (1991) 88–110. [7] P.W. Majerus, G.J. Bronze Jr., J.P. Miletich, D.M. Tollefsen, Anticoagulant, thrombolytics, and antiplatelet drugs, in: Goodman and Gilmans’s the Pharmacological Basis of Therapeutics, 9th Edition, McGraw-Hill, New York, 1995, pp. 1341–1359. [8] I. Niculescu-Duvaz, C.J. Springer, Antibody-directed enzyme prodrug therapy (ADEPT): a review, Adv. Drug Deliv. Rev. 26 (1997) 151–172. [9] K.D. Bagshawe, S.K. Sharma, C.J. Springer, G.T. Rogers, Antibody-directed enzyme prodrug therapy (ADEPT): a review of some theoretical, experimental and clinical aspects, Anal. Oncol. 5 (1994) 879–891. [10] C.J. Springer, V. Bavetsias, A.L. Jackman, R. Kimbell, F.T. Boyle, D. Marshall, R.B. Pedley, G.M.F. Bisset, Prodrug of thymidylate synthase inhibitors: potential for antibody directed enzyme prodrug therapy (ADEPT), Anti-Cancer Drug Des. (1996) 625–636. [11] S.K. Sharma, J.A. Boden, C.J. Springer, P.J. Burke, K.D. Bagshawe, Antibody-directed prodrug therapy (ADEPT). A three phase study in ovarian tumour xenografts, Cell Biophys. 24 / 25 (1994) 219–228. [12] Y. Byun, V.C. Yang, Delivery system for targeted thrombolysis without the risk of hemorrhage, ASAIO J. 44 (1998) M638–M641. [13] H. Wang, H. Song, V.C. Yang, A recombinant pro-drug type approach for triggered delivery of streptokinase, J. Controlled Release 59 (1999) 119–122.

79

[14] J.F. Liang, Y.T. Li, M. Connel, V.C. Yang, Synthesis and characterization of positive charged tPA as a prodrug using heparin / protamine drug delivery system, J. Pharm. Sci. 2 (1) (2000), article 7, http: / / www.pharmsci.org. [15] H. Song, J.F. Liang, V.C. Yang, A recombinant tissue plasminogen activator (tPA)-poly (Arg) 7 chimera for targeted thrombolysis without the bleeding risk, ASAIO J. (in press). [16] H. Bernstein, V.C. Yang, R. Langer, An investigation of heparinase immobilization, Appl. Biochem. Biotechnol. 16 (1987) 129–143. [17] A. Lanevschi, J.W. Kramer, S.A. Greens, K.M. Meyers, Evaluation of chromogenic substrate assays for fibrinolytic analytes in dogs, Am. J. Vet. Res. 57 (1996) 1124–1130. [18] Y. Byun, V.K. Singh, V.C. Yang, Low molecular weight protamine: a potential non-toxic heparin antagonist, Thromb. Res. 94 (1999) 53–61. [19] T.J. Harris, T. Patel, F.A. Marston, S. Little, J.S. Emtage, G. Opdenakker, G. Volckaert, W. Rombauts, A. Billiau, P. De Somer, Cloning of cDNA coding for human tissue-type plasminogen activator and its expression in Escherichia coli, Mol. Biol. Med. 3 (1986) 279–292. [20] H. Song, J.F. Liang, V.C. Yang, A recombinant t-PA for targeted thrombolysis without the risk of bleeding, ASAIO J. (in press). [21] E.L. Madison, E.J. Goldsmith, R.D. Gerard, M.J.H. Gething, J.F. Sambrook, R.S. Bassel-Duby, Amino acid residues that affect interaction of tissue-type plasminogen activator with plasminogen activator with plasminogen activator inhibitor 1, Proc. Natl. Acad. Sci. USA 87 (1990) 3530–3533. [22] P.W. Majerus, J.P. Miletich, D.M. Tollefsen, in: G. Gillman (Ed.), The Pharmacological Basis of Therapaeutics, 9th Edition, McGraw-Hill, New York, 1995, pp. 1341–1359. [23] E.D.T. Atkins, I. Niednszinski, Crystalline structure of heparin, in: R.A. Bradshaw (Ed.), Heparin: Structure, Function and Clinical Implications, Plenum Press, New York, 1975, pp. 19–36. [24] D. Lamba, R. Bauer, S. Fischer, R. Rudolph, U. Kohnert, W. Bode, A crystal structure of the catalytic domain of recombinant two-chain human tissue plasminogen activator, J. Mol. Biol. 258 (1996) 117–135. [25] A.D. Cardin, H.R. Weintraub, Molecular modeling of protein–glycosminoglycan interactions, Arteriosclerosis 9 (1989) 21–32.