MSU LIBRARIES -,—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wiIl be charged if book is returned after the date stamped below. PLEASE NOTE: In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark J . us. 0 >199sz 9090 10. 11. 12. 13. 14. 15. 16. Glossy photographs or pages __ Colored illustrations. paper or print Photographs with dark background _I_/_ Illustrations are poor copy __ Pages with black marks, not original copy __ Print shows through as there is text on both sides of page— Indistinct. broken or small print on several pages L Print exceeds margin requirements __ Tightly bound copy with print lost in spine __ Computer printout pages with indistinct print Page(s) lacking when material received. and not available from school or author. P899(8) seem to be missing in numbering only as text follows. Two pages numbered . Text follows. Curling and wrinkled pages Dissertation contains pages with print at a slant, filmed as received __4__ Other U-M-I THE DESIGN OF HETEROLOGOUS ENZYMES SUITABLE FOR THERAPEUTIC APPLICATION By Kaelyn Boner Hadley A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pharmacology and Toxicology 1987 _)0 I / /(‘ kl.” ‘4 (III . j)- / ABSTRACT The Design of Heterologous Enzymes Suitable for Therapeutic Application by Kaelyn Boner Hadley Scurvy in guinea pigs provides a convenient model of an inborn metabolic disease for the investigation of enzyme therapy protocols. The purpose of these studies was to examine chemical modifications of gulonolactone oxidase, the missing enzyme in ascorbic acid biosynthesis. Administration of the modified forms of this enzyme was tested in guinea pigs to examine such factors as stability of the enzyme, immune response, and the therapeutic response of the animal. One modification of gulonolactone oxidase involves immtmoprecipitation of the enzyme with specific antisera, followed by crosslinking the precipitate with glutaraldehyde. Following this modification, gulonolactone oxidase activity was protected against rapid inactivation at 37°C and also against trypsin digestion. Intraperitoneal injections of this modified enzyme complex to scorbutic guinea pigs, along with substrate supplementation, resulted in a three-fold increase in plasma ascorbic acid concentrations. Furthermore, repeated injections of the complex were tolerated. With this enzyme replacement therapy, the survival time of ascorbic acid-deficient animals could be prolonged to at least 100 days. Although the animals developed circulating antibodies against both the enzyme and antibody component, anaphylaxis was elicited only upon challenge with the antibody component. These results suggest that immobilization of the foreign Kae lyn Boner Hadley enzyme within a large stable complex was critical to reducing the toxicity of these foreign proteins. The complex may be oriented in such a way that the enzyme surface is covered with antibodies, leading to greater sensitization to this surface component. Conjugation of gulonolactone oxidase with polyethylene glycol (PEG) was the second modification investigated. This modified form of the enzyme was suitable for intravenous administration and its infusion elicited ascorbic acid synthesis in a dose-dependent manner. However, the circulating half-life of enzyme activity was not prolonged by this modification. Proteolytic digestion may explain the disappearance of enzyme activity from the circulation, but tissue uptake could contribute initially. In contrast to results obtained with other enzymes, attachment of polyethylene glycol to gulonolactone oxidase did not abolish the enzyme's ability to react with pre-formed antibodies nor did it eliminate its immunogenicity. Apparently, all of the antigenic determinants on the enzyme were not covered by the polymer. Despite this, animals were able to tolerate a second injection of this complex. To my husband, Robert ACKNOWLEDGEMENTS I would like to acknowledge my advisor, Dr. Paul Sato, for his guidance during my graduate career. My thanks go to him for taking the time to train a student and for his continued interest in this project. The members of my guidance committee also deserve acknowledgment for their time, constructive criticism, and encouragement. I would like to express my appreciation to Dr. Tai Akera, Dr. Theodore Brody, Dr. Ronald Patterson, and Dr. John Thornburg. Others who deserve recognition include Dr. Barbara Kintner and Dr. Susan Stein who provided helpful advice and discussions concerning the health of the animals used in these studies. Diane Hummel deserves many thanks for all of her help over the last few years. My gratitude goes to my parents, Garnard and Donnie June Boner, for their interest and confidence in me. They have encouraged me throughout my education and I appreciate their love and support. Most of all, I thank Robert for his friendship, unlimited patience, and encouragement throughout this experience. Especially, I am grateful for his continuing love. iii TABLE OF CONTENTS Page LIST OF FIGURES -------------------------------------------------- vii LIST OF TABLES -------------------------------------------------- LIST OF ABBREVIATIONS ------------------------------------------- 1: INTRODUCTION --------------------------------------------------- 1 I. hherited Metabolic Disease -------------------------------- l A. HIBtOI’Y -------------------------------------------- 1 B. Incidence ------------------------------------------- 3 II. Treatment of Metabolic Disease ---------------------------- 4 A. Gene replacement therapy ---------------------------- 4 B. Other therapeutic approaches ------------------------- 5 C. Enzyme replacement therapy ------------------------- 9 III. Scurvy - A Model Inborn Metabolic Disease ------------------ 11 IV. Treatment of Scurvy -------------------------------------- 14 V. Objectives ---------------------------------------------- 17 A. Heterologous immtm0precipitates --------------------- 18 B. Polyethylene glycol (PEG)-conjugated gulonolactone oxi- dase ----------------------------------------------- 19 MATERIALS AND METHODS ---------------------------------------- 22 I. General ------------------------------------------------- 22 A. Animals ------------------------------------------- 22 B. Surgical procedures ---------------------------------- 22 C. Preparation of enzymes ------------------------. ------ 23 D. Analytical methods ----------------------------------- 24 E. Statistical analysis ---------------------------------- 24 II. Heterologous ImmunOprecipitate Studies -------------------- 24 A. Preparation of antisera ------ 4 ----------------------- 24 B. Purification of immunoglobulin G ---------------------- 26 C. Preparation of immunOprecipitates -------------------- 26 D. Glutaraldehyde reaction ------------------------------ 28 E. Enzyme administration ------------------------------ 28 F. Postmortem evaluation ------------------------------ 29 G. Radiochemical methods ------------------------------ 29 0 1D TABLE OF CONTENTS (Continued) RESULTS Page H. Preparation of samples for liquid scintillation spectro- metry ---------------------------------------------- 30 1. Liquid scintillation counting ---. ------------------------ 31 Polyethylene Glyco l-Conjugated Enzymes -------------------- 3 1 A. Modification of enzymes with polyethylene glycol -------- 31 B. Enzyme administration -------------------------------- 32 --------------------------------------------------------- 33 Heterologous Immunoprecipitate Studies --------------------- 33 A. Modification of gulonolactone oxidase ------------------ 33 l. Crosslinking conditions -------------------------- 36 2. Catalytic characteristics ------------------------ 36 B. Administration of modified gulonolactone oxidase -------- 40 1. Single dose studies ------------------------------ 40 2. Effect of repeated doses of the modified enzyme ---- 44 3. Toxicity of the modified gulonolactone oxidase ----- 44 C. Adaptability of the modification procedure to other enzymes -------------------------------------------- 51 1. Catalytic characteristics ------------------------- 51 2. Toxicity of these modified enzymes --------------- 53 D. Modification of gulonolactone oxidase using purified IgG -- S6 1. Characteristics of the modified enzyme complex ---- 56 2. Administration of the less contaminated complex --- 61 3. Analysis of doses of enzyme activity administered --- 68 4. Supplementation of ascorbic acid during gulonolac- tone oxidase therapy ---------------------------- 69 5. Supplementation of other vitamins during enzyme therapy --------------------------------------- 69 E. Comparison of gulonolactone oxidase therapy regimens --- 73 F. Disposition of crosslinked immunoprecipitates of gulono- lactone oxidase ------------------------------------- 76 1. Recovery of the administered radioactive dose ----- 76 2. Concentrations of radioactivity in tissues ---------- 77 Polyethylene Glyco 1(PEG) -Conjugated Gulonolactone Oxidase -- 77 A. Attachment of PEG to the enzyme --------------------- 83 A 1. Characteristics of the PEG-conjugate ------------- 83 2. Catalytic properties ---------------------------- 86 TABLE OF CONTENTS (Continued) Page B. Stability of PEG-gulonolactone oxidase ----------------- 9O 1. In vitro stability of PEG-GLO activity ------------- 90 2. Circulating half-life of PEG-GLO activity ---------- 9O 3. E vitro stability of PEG-GLO in plasma ---------- 90 C. Administration of PEG-gulonolactone oxidase ------------ 97 1. Single dose studies ------------------------------ 97 2. Dose-response relationship ----------------------- 101 D. Immunogenicity of PEG-gulonolactone oxidase ----------- 101 l. Immtmoreactivity of the PEG-enzyme ------------- 101 2. Immunogenicity of the PEG-enzyme -------------- 101 DISCUSSION ------------------------------------------------------- 109 I. Crosslinked Immunoprecipitated Gulonolactone Oxidase -------- 109 A. Properties afforded by this modification ---------------- 109 B. Modifying gulonolactone oxidase with homologous anti- serum --------------------------------------------- 110 C. Modifying gulonolactone oxidase with heterologous anti- serum or IgG ---------------------------------------- 112 D. Application of the modification procedure to other en- zymes --------------------------------------------- 116 E. Administration of gulonolactone oxidase modified with purified IgG ---------------------------------------- 118 F. Metabolic fate of the administered enzyme complex ------ 123 G. Reasons for detoxification of the modified enzyme com- plex ----------------------------------------------- 125 II. PEG-Conjugated Gulonolactone Oxidase --------------------- 128 A. Characteristics of PEG-gulonolactone oxidase ---------- 129 B. Stability of PEG-gulonolactone oxidase ----------------- 130 C. Administration of PEG-gulonolactone oxidase ---------- 132 D. Immunogenicity of PEG-gulonolactone oxidase ---------- 133 E. Comparison of gulonolactone oxidase with other PEG- enzymes ------------------------------------------- 135 SUMMARY AND CONCLUSIONS ------------------------------------- 137 BIBLIOGRAPHY ................................................... 141 E3323 10 11 12 13 14 LIST OF FIGURES Page Glucuronic acid pathway of glucose metabolism --------------- 12 Protocol for the preparation of crosslinked immunoprecipitates -- 34 Relationship between the extent of crosslinking and enzyme activity with increasing concentrations of glutaraldehyde ------- 38 Synthesis of ascorbic acid in gulonolactone oxidase-treated ani- mals ----------------------------------------------------- 42 Survival and prevention of scurvy in guinea pigs by enzyme administration therapy ------------------------------------- 45 Ouchterlony immunodouble diffusion tests for antibody in serum of gulonolactone oxidasedtreated animals -------------------- 48 Polyacrylamide gel electrophoretic analysis of immunoglobulin G purified from horse serum ---------------------------------- 57 SDS-polyacrylamide gel electrophoresis of gulonolactone oxidase immunoprecipitates ---------------------------------------- 59 Increased stability of modified gulonolactone oxidase to incuba- tion at 37°C and trypsin digestion ---------------------------- 62 Synthesis of ascorbic acid by guinea pigs treated with gulonolac- tone oxidase modified using purified IgG ---------------------- 64 Growth rate of enzyme-treated normal guinea pigs and enzyme- treated scorbutic guinea pigs compared to untreated normal control animals -------------------------------------------- 66 Growth curves of nutrient-supplemented guinea pigs on enzyme replacement therapy --------------------------------------- 71 Comparison of the survival time of ascorbic acid-deficient animals treated with the different gulonolactone oxidase therapy regimens and their respective control groups ------------------ 74 Recovery of the dose of radioactivity from individual tissues ---- 79 LIST OF FIGURES (Continued) 15 16 17 18 19 20 21 22 23 24 Page Concentrations of radioactivity in tissues -------------------- 81 Scheme showing attachment of the polymer polyethylene glycol (PEG) to a protein ----------------------------------------- 84 Electrophoretic migration of PEG-gulonolactone oxidase com- pared to the unmodified enzyme in SDS-polyacrylamide gels ----- 87 Stability of unmodified gulonolactone oxidase and the PEG- enzyme to incubation at 37°C and trypsin digestion ------------- 91 Plasma half-life of PEG-gulonolactone oxidase compared to the unmodified enzyme ---------------------------------------- 93 E vitro half-life of PEG-gulonolactone oxidase in plasma -------- 95 Intravenous administration of PEG-gulonolactone oxidase in- creases plasma ascorbic acid concentrations ------------------ 98 Elevations in plasma ascorbic acid concentrations are dependent on the dose of activity -------------------------------------- 102 PEG-gulono lactone oxidase reacts with antiserum against the unmodified enzyme -----------------e ---------------------- 104 Antibodies against the unmodified enzyme are detected in sera from guinea pigs immtmized with PEG-gulonolactone oxidase ----- 107 Table LIST OF TABLES Page Effect of varymg pH during crosslinking on enzyme activity ---- 37 Comparison of kinetic parameters of gulonolactone oxidase and the modified enzyme ------------------------------------- 41 Reaction of animals to challenging doses of the components of the modified enzyme ------------------------------------- 50 Kinetic parameters of modified enzymes --------------------- 52 Reaction of animals to repetitive injections of enzymes modi- fied with heterologous sera -------------------------------- 54 Reaction of animah to challenging doses of components of modified enzyme complexes ----,- --------------------------- 55 Increasing recovery of the administered dose of 1‘i'C-immtmo- precipitates over a 10-day period --------------------------- 78 Comparison of kinetic parameters of PEG-gulonolactone oxi- dase with those of the immodified enzyme ------------------- 89 Tissue ascorbic acid concentrations increase significantly as a result of PEG-gulonolactone oxidase treatment --------------- 100 AA ANOVA GLO i.v. PEG PEG-GLO SDS SEM XL-IP LIST OF ABBREVIATIONS ascorbic acid analysis of variance gulono lactone oxidase intravenous least significant difference polyethylene glycol PEG-gulonolactone oxidase sodium dodecyl sulfate standard error of the mean crosslinked immunoprecipitated INTRODUCTION The progression of events leading to our present understanding of inherited metabolic diseases has been described in great detail by other authors (Stanbury e_t_ $1., 1983). However, in order to develop an awareness of the problems encountered in the treatment of these diseases, an abbreviated summary of the history of their identification is presented here. This is followed by a discussion of various approaches to the treatment of these diseases and the rationale for each. The focus of this dissertation involves testing enzyme replacement therapy protocols, using scurvy in guinea pigs as a model metabolic disease. The basis for the use of this model, as well as previous protocols which have been tested using it, will also be presented. The specific objectives of this investigation will then be outlined. I. Inherited Metabolic Disease A. History The existence of inborn errors of metabolism was first described by Sir Archibald Garrod at the beginning of the twentieth century (Garrod, 1909). This description was based on his extensive studies of the condition alkaptonuria. Garrod observed that patients with alkaptonuria excreted large quantities of homogentisic acid in the urine. Unaffected individuals did not excrete this compound. He also noted that the condition had a familial distribution. This pattern of occurrence could be explained by the laws of genetics, if the condition were inherited as a recessive trait. From these studies Garrod postulated that certain diseases were the result of the absence or abnormality of a particular enzyme which governed a single metabolic step. In the case of alkaptonuria, he hypothesized that the accumulation of homogentisic acid, and thus its excess excretion, was the result of the inability of the patient to oxidize this component. Fifty years later, this was indeed shown to be the case (LaDu g a_l., 1958). Investigators demonstrated the absence of homogentisic acid oxidase activity in the liver of a patient with alkaptonuria. Garrod also studied patients with cystinuria, pentosuria, and albinism. He suspected that these conditions were the result of a block in some metabolic pathway as well. Albinism was thought to be the result of a block in melanin formation. He proposed that pentosuria and cystinuria resulted from the excretion of excess substrates that accumulated proximal to a blocked metabolic step. In the case of pentosuria, this was later shown to be correct. L-Xylulose is excreted in the urine due to the block of its conversion to xylitol (Hiatt, 1978). Cystinuria was shown to result from a block in amino acid transport (Segal and Thier, 1983). A basis for the explanation of these inborn errors of metabolism was provided when the relationship between genes and enzymes was postulated about 40 years later (Beadle, 1945). Beadle proposed that the information of one gene provided for the synthesis of one enzyme. This concept was developed further by Beadle (1959) and Tatum (1959). Evidence suggested that all biochemical processes are under genetic control. These biochemical processes take place through a series of chemical reactions, with each reaction in the series being controlled by a single enzyme. Mutation of a single gene, therefore, could result in the inability of an organism to carry out a single reaction. Thus, it appeared that inborn metabolic diseases, such as alkaptonuria, could be produced by mutations in genes encoding particular enzymes. At about this time, the first enzyme defect in a human genetic disease was demonstrated by Gibson (1948). Gibson showed that methemoglobinemia was the result of a deficiency in an NADH-dependent enzyme responsible for the reduction of methemoglobin. Demonstrations of other conditions followed. In addition, evidence showing that mutated proteins do indeed show structural differences from the normal, native counterparts was presented (Pauling g 31., 1949). Pauling showed that human mutations actually produce an alteration in the primary structure of the proteins. So it appeared that metabolic errors could result when mutant genes produced abnormal proteins whose functional activities were altered. B. Incidence Inborn errors of metabolism are genetic diseases and are categorized as monogenic disorders. That is, they are the result of a single mutant protein and show Mendelian patterns of inheritance. At this time, approximately 1400 such monogenic disorders have been identified. These diseases occur in approxi- mately 1% of live births and account for significant morbidity in children. The basic biochemical lesions responsible for these conditions involve a variety of proteins. Of the 250 diseases for which the specific defect has been identified, 170 of them involve abnormal enzymes. Although identifying and tmderstanding the specific defect has provided for more accurate detection of carriers and earlier diagnosis of afflicted individuals, the lack of a specific therapy for many of these diseases remains. Therapeutic measures directed toward correcting the metabolic defect exist for only about 40 of these conditions. Since genetic diseases are a major cause of infant mortality in industrialized countries (Bart and Lane, 1985), research directed toward develop- ment of effective therapeutic measures must continue. II. Treatment of Metabolic Disease Thorough treatment of an inborn error of metabolism depends upon an accurate diagnosis and an understanding of the pathophysiology of the disease. Currently, most of these diseases are not fully understood. A complete tmderstanding would mean knowing the exact aberration at each level; at the DNA level, the protein level, and also a knowledge of how the aberrant gene product affects cell function. Although the specific enzymatic defects and accumulating metabolites have been identified for approximately 170 of these metabolic diseases, in some cases it remains unclear how the particular alteration produces the effects of the disease. h addition, the exact disruption at the DNA level has not been deciphered. Despite the fact that some of the information is lacking, attempts are being made to treat these genetic diseases. Investigations are also being carried out to understand their causes more fully. A. Gene replacement therapy With the advances in recombinant DNA techniques and molecular biology, attempts are being made to correct inherited enzyme deficiency diseases by replacement of the defective or missing genetic information. The goal is toward achieving permanent restoration of the particular enzyme activity. Directly reversing the deficiency in this manner would appear to be potentially curative. However, problems with this method still exist. Appropriate genetic information may not be available. Presently, genes encoding for the needed enzymes have not been located and cloned in most cases (Anderson, 1984). Even if the DNA has been isolated, one must be able to correctly insert this information into the appropriate cells and assure that it remains there to be effective. The techniques for gene replacement procedures must be perfected. At this time, techniques for inserting the genes into a particular chromosome, with the assurance that they will be expressed in an orderly and controlled manner, are not available. The expression of the newly inserted genetic information must be appropriately regulated. Cells into which the genetic information has been transferred may have to carry out post-translational modifications to provide a prOperly fmctioning enzyme and produce an appropri- ‘ ate amount of activity. Also, there is not yet a reliable way to introduce the. corrected cells so that they will survive while the defective cells are eliminated. This may be necessary if corrected cells do not have a growth advantage over the endogenous, defective cells. The defective cells might have to be removed so that the growth and expansion of the corrected cell population would be favored. Finally, localization of the enzyme to specific sites may pose a problem. Currently the only human tissue used for gene transfer are bone marrow cells. These cells can be extracted easily and then reinserted after genetic manipulation (Stanbury E g, 1983; Anderson, 1984). In some disorders though, the ftmctioning gene must be located in other organs to be effective. With current techniques this is not possible. Therefore, in view of the problems that must be resolved with genetic restitution and considering the vast number of diseases for which the genetic information has not yet been found, investigation of other therapeutic regimens is necessary. B. Other therapeutic approaches A variety of approaches have been investigated in an effort to treat inborn metabolic diseases (Chang, 1977; Watts, 1982; Stanbury g_t_ 3.1., 1983). Presently, however, a single method is not applicable to the treatment of many such disorders. A brief discussion of several treatment methods is presented here. The rationale for each of these and an example of its use is included. One method of therapy involves the dietary restriction of the sub- strate that accumulates prior to the metabolic block. For some genetic disorders, reduction of the accumulating substrate can prevent the clinical manifestations associated with the disease. An example of this is shown by treatment of the condition phenylketonuria (PKU). Patients with PKU have a deficiency of the enzyme phenylalanine hydroxylase. This enzyme catalyzes the formation of tyrosine from phenylalanine. As a result of this enzyme deficiency, the substrate, phenylalanine, accumulates and an excess of its metabolites appear to contribute to the pathology of the disease. There is evidence that phenylalanine derivatives disrupt myelin formation and this can lead to the mental and growth retardation associated with PKU. Restriction of dietary phenylalanine in these patients, so that normal blood levels of this amino acid are maintained, has been shown to prevent these effects. To achieve optimal efficacy, strict adherence to the dietary regimen is required, beginning within 8 weeks after birth and continuing tmtil at least 10 years of age. After this, the diet may be discontinued gradually with no deterioration in the patient's condition (Bickel and Schmidt, 1982). However, problems may arise when the female patient reaches reproductive maturity. High blood phenylalanine levels in untreated PKU mothers can adversely affect the fetus. Even though the infant may not have PKU, it can suffer mental retardation, congenital malformations, and even early death, as a result of exposure to high maternal phenylalanine levels. Therefore, it is imperative to impose the strict dietary restrictions again. Treatment should begin no later than the third week of pregnancy, and, preferably, before conception. Delivery of a healthy baby is still not guaranteed. An increased frequency of microcephaly and mental retardation has been noted in babies whose mothers comply with the diet. So, problems remain with dietary management, especially regarding maternal PKU and preventing damage to the unborn child. A more reliable means of controlling phenylalanine accumulation is necessary to ensure tight control of phenylalanine levels and prevention of these effects. Replacement of the deficient end-product is another approach to therapy. A metabolic block causes a decrease in the initial product of the particular reaction and, possibly, in subsequent products of the metabolic se- quence. If the absence of these products contributes to the pathologic effects, then their replacement should alleviate the disease state. Orotic aciduria is an example of a case in which such therapy is effective. In this condition, orotic acid can not be converted to uridine due to defects in two sequential enzyme steps. The product, uridine, is necessary for the synthesis of pyrimidines, which are important for nucleic acid synthesis. Administration of uridine overcomes this block and corrects the symptoms of the disease. It also results in a diminished synthesis of orotic acid, so the excretion of this acid in the urine is decreased. In some diseases, the clinical symptoms are the result of accumulation of stored metabolites in the tissues of the body. Removal of these excess stored materials may alleviate the problems associated with the disease. h a case such as this, the goal of therapy is to deplete the accumulated metabolites. A classical examwe of a condition where depletion of a stored substance can be effective is Wilson‘s disease. Copper is stored in excess in afflicted patients and eventually its accumulation leads to severe neurologic problems and liver damage. Removal of excess copper is effected by the agent penicillamine. This agent chelates the copper, increasing its subsequent excretion and alleviating symptoms of the disease. Inhibition of metabolic pathways is another approach that is useful in the treatment of certain disorders. If the substrate which accumulates as a result of the metabolic error is toxic, control of its production may ameliorate the symptoms of the disease. To this end, inhibition of an enzyme critical for the synthesis of the particular accumulating substance may provide therapy. The use of hematin in the treatment of acute hepatic porphyrias is an example of this (Watts, 1982). The enzymatic defect in acute hepatic porphyrias is a deficiency of porphobilinogen (PBG) deaminase activity. Symptoms of the disease occur when S -aminolevulinic acid (ALA) and porphobilinogen (PBG) accumulate as a result of the PEG deaminase deficiency; Certain hormones and drugs can precipitate porphyric attacks. They do this by stimulating the synthesis of ALA and PEG through the induction of ALA synthetase. Hematin 'mhibits ALA synthetase activity and, therefore, reduces levels of ALA and PEG. Inhibiting the metabolic pathway in this manner and preventing the formation of the accumulat- ing substances can lessen the symptoms of the acute porphyric attack. Organ transplantation has also been considered as a therapeutic measure for certain metabolic diseases. Transplanting an organ which can normally synthesize the deficient enzyme has been done. Kidney transplants have been attempted in cases of Fabry's disease and these patients show measurable levels of a-galactosidase activity after the procedure (Groth, 1982 and Stanbury gt a_l., 1983). Other examples include splenic transplantation and renal transplan- tation as treatment for Gaucher's disease (Desnick, 1980; Groth, 1982). However, the results do not warrant the use of organ transplantation on a routine basis. Rejection of the transplanted organ by the patient's body presents a major difficulty in the use of this approach. Demonstration of clear therapeutic benefit is lacking as well. Although an increase in enzyme activity following transplanta- tion is noted in some cases, actual improvement in the clinical symptoms of the disease has not been shown. Transplanting allogenic bone marrow to patients who are afflicted with certain immunodeficiency states provides another example. In some cases efficacy has been demonstrated, but problems in finding compatible donors remain. C. Enzyme replacement therapy Enzyme deficiency diseases should be treatable by administration of the missing enzyme. Studies of methods by which this can be accomplished are the subject of this dissertation research. h addition to its usefulness as treatment for inborn metabolic disorders, enzyme administration may be used as therapy for other human diseases, such as certain cancers. The treatment of acute lymphocytic leukemia with asparaginase is one such example. The rationale for using asparaginase in cancer chemotherapy is that certain tumor cells are dependent upon exogenous sources of the amino acid asparagine, whereas normal cells are not. Asparagine is depleted by the administered asparaginase, depriving the tumor cells of this necessary nutrient and ultimately causing their death (Uren, 1981). Another experimental application for administered enzymes is the detoxification of foreign chemicals i3 gig (Goedde and Altland, 1971). The use of organOphosphate hydrolyzing enzymes to treat organophosphate poisoning is an example of this. Early studies suggest that these enzymes may be effective in the prevention of organophosphate toxicity. While administering enzymes appears to be a relatively simple con- cept, a number of major difficulties are encmmtered. First, adequate quantities of pure enzyme suitable for administration are difficult to obtain. Enzyme purification procedures are time-consuming and frequently result in low yields. Considering that affected individuals may require treatment over an entire lifetime, large quantities of enzyme may be necessary. The stability of injected enzymes may also present difficulties. Administered enzymes are susceptible to denaturation and protease digestion and thus may have brief half-lives. Further- more, if repeated injections are required immunologic complications may occur. If the injected enzymes are recognized as foreign by the body, an immune response can be elicited and an allergic reaction initiated upon subsequent 10 injections of the particular enzyme. Degradation of the injected enzyme activity may also be accelerated as a result of the immune response. Directing the administered enzyme to the appropriate site for its action presents problems in the therapy of certain inborn metabolic diseases. Modification of the particular enzymes may be required in order to target them to specific sites. In summary, there are various problems which need to be overcome to render enzymes suitable for routine therapeutic use. Attempts at using enzyme replacement therapy to treat patients affected by enzyme deficiencies have been reported (Desnick, 1980; Brady, 1983). The outcome has been disappointing, however. These trials often represent the first time that the efficacy of such therapy has been tested. Patients affected by the diseases receive enzyme replacement therapy without evidence that it is safe and will indeed be therapeutically beneficial. The effectiveness of these treatments is ambiguous. While some evidence of increased enzyme activity and subsequent substrate depletion is shown, clinical improvement in the patients' conditions has not been demonstrated. Furthermore, toxicity to the enzyme usually develops. A more reasonable approach would be to test the safety and effectiveness of enzyme therapy in animals before initiating therapy in humans. Although some animal models exist (Migaki, 1982), difficulties are involved in obtaining suitable animals for studies. Many of the animals are heterozygotes and the disease traits are recessive. Careful breeding of animals is often necessary in order to obtain homozygotes that express the disease. The animals' genotype must be determined as well. In the studies presented in this dhsertation, the disease scurvy in guinea pigs is used as a model inborn error of metabolism. The rationale for its use is discussed below. 11 III. Scurvy - A Model Inborn Metabolic Disease Scurvy in guinea pigs provides a convenient model for the study of enzyme therapy protocols. This disease, which results from the lack of ascorbic acid, has been determined to be an inborn error of metabolism (Burns, 1956). Guinea pigs, like man and other primates, have lost the capacity to synthesize this vitamin. Therefore, they require a dietary intake of ascorbic acid. Their inability to synthesize this vitamin is the result of an enzyme deficiency. Gulonolactone oxidase, the missing enzyme, catalyzes the final step in ascorbic acid biosynthesis (Burns, 1956; Sato 3 $1., 1976). Ascorbic acid is synthesized from glucose by the glucuronic acid pathway as shown in Figure 1 (Burns, 197 5). The liver microsomal enzyme gulonolactone oxidase catalyzes the conversion of gulonolac- tone to ascorbic acid. Scurvy-prone species do not express the gene coding for this enzyme (Sato and Udenfriend, 1978a). Administering this enzyme to scorbutic guinea pigs should give them the capacity to synthesize their own ascorbic acid and thus alleviate scurvy. There are a number of advantages to using scurvy as a model in the study of enzyme therapy regimens. Guinea pigs are a common, readily available labora- tory animal. Scorbutic animals are easily obtained simply by maintaining them on an ascorbic acid-deficient diet. Importantly, guinea pigs are homozygous for the enzyme deficiency trait. This eliminates the necessity of determining their genotype and carrying out the selective breeding procedures that are necessary with other animal models. Scurvy is a also a lethal disease, so there is a clear endpoint to evaluate the effectiveness of the enzyme replacement therapy. In addition, gulonolactone oxidase is the final enzyme in the biosynthesis pathway and it does not require other cofactors for activity. For this reason, it is not necessary to target the enzyme to a particular site for optimal activity. 12 Figure 1. Glucuronic acid pathway of glucose metabolism. Ascorbic acid is synthesized from glucose via this pathway. The point of conversion of gulonolac- tone to ascorbic acid is indicated by an asterisk. The microsomal enzyme, gulonolactone oxidase, catalyzes this reaction. Taken from Burns (1975) with permission. 13 socu i i .=¢—l «fl ' 2'.- " o... O ":ILJ -x- 3311 'fiJ llOCll KOCH nocx ‘éflOfl ’éfl.“ 'OILOII pastor-o- S-Reb- km: lactone W said a l 3 . 3 ac ages 9-01! ”$9" so?! use" acca "OT” / ”Io" \"c’l'." Icon ”9°" "To" econ FF" sacs axon ° omen no mucos- rain-route acid mu.- acid r - i o . u '3‘ c-ou hope a l .= a Icon 11;: sob: canon '61:.“ L.s-clueou.“ Fleur-cm sell .5... A. the” Mn, ’Cfl.“ “GI c=o /c=o "c" 352;.32.” "a" o " s on a“ I"”""“iro<'m noel-u c" D-Xylulooe-S-P scion "a.” t-Xylulou 621.001 clean .n-Xylulou Xylltol Glucuronic Acid Pathway of Glucose Metabolism Figure 1 14 In the studies presented here, various therapy regimens are tested using this scurvy model. The efficacy and toxicity of the protocols can be compared and, if necessary, adjustments or variations in a protocol can be instituted. After testing the treatment regimens in this animal model, perhaps they can be applied to the treatment of other inherited metabolic diseases. IV. Treatment of Scurvy A number of different methods of enzyme replacement therapy have already been investigated using the scurvy model. These methods will be discussed here briefly. Implantation of the enzyme contained within a dialysis bag is one approach that was examined (Sato, 1980). Administration of the enzyme in this way was designed to protect it from proteolytic attack and degradation. Furthermore, the animal would not be exposed to the foreign protein and this should prevent the mitiation of an immune response. Gulonolactone oxidase, purified from rat liver, was placed in a dialysis bag which had been surgically implanted within the peritoneal cavity of the animal. The substrate, gulonolactone, was then injected into the peritoneum. The dialysis bag should be permeable to the small molecules of substrate and product, but impermeable to the enzyme macromolecule, antibodies, and proteolytic enzymes. Control animals were treated similarly, except that bovine serum albumin was contained within their dialysis bags. In ziv_o enzymic activity was demonstrated by increases in plasma and tissue concentrations of ascorbic acid following gulonolactone oxidase treatment. In the enzyme-treated animals, a four-fold increase in plasma ascorbic acid concentra- tions occurred, while the vitamin concentrations in the plasma of control animals remained relatively constant. Tissue concentrations of ascorbic acid in the enzyme-treated animals were also significantly higher than those in the control 15 group. Multiple injections prolonged the survival of scorbutic animals to twice the normal time of survival on the deficient diet, demonstrating a therapeutic effect. Scorbutic animals usually do not survive beyond 23-28 days (Barnes gt 51., 1973; Jones g 21., 1973). Gulonolactone oxidase contained within the dialysis bag 8 not antigenic nor does sera from the animals contain detectable precipitat- ing antibody against the enzyme. Therefore, using the scurvy model, it was shown that the enzyme administered in this manner catalyzes the synthesis of ascorbic acid and has potential clinical value. The feasibility of using an intestinal segment as an artificial organ for enzyme replacement was also tested (Shelt e_t_ a_l_., 1982). One problem with the use of the implanted dialysis bag was that the animals mounted a response against the foreign dialysis bag in the long-term studies. This impaired its permeability and effectiveness. It was postulated that by using a membrane from the animal, the problem of a tissue response would be avoided. In addition, this membrane, like the dialysis membrane, should allow passage of substrate and product, while preventing direct exposure of the animal to the foreign enzyme. For these studies, a surgical procedure was developed to prepare a pouch from an intestinal segment of the animal, leaving the blood supply to it intact. A silastic tube attached to this pouch permitted access for injection of the enzyme. Attempts to demonstrate in 2332 synthesis of ascorbic acid by increases in plasma ascorbic acid concentrations following this treatment were unsuccessful. However, one animal did survive for 57 days while being fed the vitamin C-deficient diet and receiving enzyme replacement therapy. So, ascorbic acid synthesis may have occurred. Perhaps the vitamin did not enter the plasma at a rate fast enough to detect an increase in the plasma during the time of the experiment. The vitamin may also have been converted to a different form, for instance dehydroascorbic acid, which would not be detectable in the assay used. Some of the treated 16 animals did not appear to produce antibody against the foreign enzyme. However, results from immunodiffusion tests show that antibodies were raised against the enzyme in other animals. This may have resulted from the diminished integrity of the segments, allowing enzyme to leak into the peritoneal cavity. Despite the presence of antibody, animals did not exhibit an allergic reaction upon intravenous challenge with the enzyme. To summarize, this method does not appear to be feasible for administration of gulonolactone oxidase in amounts adequate to allow detectable in m ascorbic acid synthesis. It does appear to decrease the antigenicity of the foreign protein and it could be useful in the treatment of other metabolic disorders. Additional work has focused on the development and investigation of chemical modifications of enzymes that make them suitable for therapeutic application. Efforts are directed toward improving the stability of the injected enzyme activity in this way. Decreasing the im munogenicity and allergenicity of the foreign enzyme is a goal also. One modification that has been studied involves the immunoprecipitation of the enzyme with specific antisera followed by crosslinking the precipitate with the bifunctional reagent glutaraldehyde. Crosslinking is intended to reinforce the interactions between antigen (enzyme) and antibody. Using crosslinked immunOprecipitated enzymes, several of the basic pro- blems encountered in enzyme administration can be addressed. Immunoprecipita- tion facilitates the isolation of large quantities of enzyme that may be needed for therapy. Furthermore, enzyme activity is significantly stabilized to heat denatur- ation and trypsin digestion by a number of different modifications (Holcenberg, 1982; Poznansky, 1983), including immunoprecipitation (Snyder e_t a_l_., 1974) and crosslinking (Snyder e_t. a_l., 1974; Poznansky, 1979; Klibanov, 1979; Wold, 17 1973). The characteristics of crosslinked immunoprecipitated gulonolactone oxidase were examined as well as its potential for use in therapy. Work using this modification for gulonolactone oxidase demonstrates that crosslinked immunoprecipitates are effective in reversing scurvy in guinea pigs (Sato and Walton, 1983). Their weight gain is restored following a single dose, as compared to the continuing decline of body weight observed in control animals. A three-fold increase in plasma ascorbic acid concentrations also occurs following the enzyme therapy. Furthermore, repeated injections of this modified enzyme are tolerated and are therapeutically beneficial in prolonging the survival of guinea pigs fed without ascorbic acid (Sato e_t 11., 1986). These gulonolactone oxidase-treated animals survived for at least 100 days compared to the usual time of survival of guinea pigs on the ascorbic acid-deficient diet which is usually 23- 28 days (Jones gt 3.1., 1973; Barnes e_t a_l., 1973). These studies show that the enzyme has been detoxified by this modification, yet substantial enzyme activity remains. Specific guinea pig antisera was used to immtmoprecipitate gulonolac- tone oxidase in these experiments. It is possible that use of homologous sera in the procedure is critical for the animals to tolerate repeated administration of the modified enzyme complex. Further investigation into the basis for the "detoxification' of gulonolactone oxidase by this modification are part of this dissertation research. V. Objectives The research presented in this dissertation is directed toward the investiga- tion of chemical modifications of gulonolactone oxidase that render it suitable for therapeutic use. Work has continued with the crosslinked immunoprecipitated gulonolactone oxidase with efforts focused on determining the possible reasons for the "detoxification" of this modified enzyme complex. Efforts were directed 18 toward the development of a modified enzyme that would be suitable for intravenous administration as well. To this end, gulonolactone oxidase was conjugated to the polymer polyethylene glycol. A. Heterologous immunOprecipitates After completion of the studies using specific guinea pig antiserum for the modification of gulonolactone oxidase, it was postulated that the multiple injections were tolerated because the homologous antibody masked the antigenic sites of the foreign enzyme. The guinea pigs, therefore, do not recognize it as allergenic. Alternatively, the modified enzyme may be tolerated not because it is bound to homologous antisera, but because of some other characteristic of the immunocomplex, such as immobilization within a large stable complex. These studies were designed to investigate whether detoxification of the enzyme was dependent on the use of homologous antiserum. Using heterologous antiserum to modify the enzyme in these experiments should distinguish between these possibilities. Furthermore, when considering application to human therapy, the use of heterologous antiserum would be advantageous. Otherwise, large amounts of human serum would be necessary to prepare the enzyme-antibody complex. Using heterologous antiserum would be far more practical since large amounts of antibody specific for the enzyme can be raised in animals and then used for preparations to treat humans. For these studies, specific antiserum against gulonolactone oxidase was raised in rabbits and horses and used for immunoprecipitation of the enzyme from solubilized chicken kidney microsomes. Chicken kidneys possess gulonolac- tone oxidase which has a high specific activity (Chaudhuri and Chatterjee, 1969; Chatterjee gt _a_l., 1975). The modification procedure was optimized with these components in order to maintain as much activity as possible and the character- istics of the resulting complex were examined. Experiments were conducted to 19 examine the ability of this modified enzyme complex to catalyze the E gi_v_g synthesis of ascorbic acid and to be therapeutically beneficial to scorbutic animals. The effects of repeated administration of the complex were also tested since the ability of animals to tolerate repeated injections of the foreign proteins is an important consideration in enzyme replacement therapy. To be therapeuti- cally beneficial, repeated administration of the foreign proteins may be required and, as a result, there may be immunologic complications. It is important that this be avoided so that allergic reactions do not interfere with subsequent treatments and efficacy. Studies of the toxicity and immune response to the modified enzyme, as well as its metabolic fate, are presented. If a procedure is to have practical applicability, it should be suitable for modifying other enzymes that have therapeutic value. This is also tested in the experiments discussed here. Three other enzymes having potential applica- tions in therapy were modified by immunoprecipitation and crosslinking. Included were the enzymes asparaginase, histidase, and serum cholinesterase. Asparagi- nase, which catalyzes the hydrolysis of asparagine, has possible uses in cancer chemotherapy. Histidase, the missing enzyme in the disease histidinemia, catalyzes the deamination of histidine, while administration of serum cholinester- ase has potential use in the treatment of organophosphate poisoning. The characteristics of these modified enzyme complexes were examined in a further effort to test the adaptability of this procedure. In addition, the toxicity of these complexes was studied. B. Polyethylene glycol (PEG)-conjugated gulonolactone oxidase Studies were also conducted to modify gulonolactone oxidase to render it suitable for intravenous administration. Attempts to administer crosslinked immunoprecipitated gulonolactone oxidase via this route suggested that a more rapid and efficient synthesis of ascorbic acid would occur (Sato and Lindemann, 20 1986). Since infusion of that complex was found to be highly toxic, examination of another modification procedure was necessary. Development of a modified form of gulonolactone oxidase that could be infused intravenously would permit comparison between the two routes of administration. Because of a more efficient synthesis by the enzyme admini- stered intravenously, the same therapeutic benefit may be attained using smaller doses of the enzyme. The amomt of ascorbic acid synthesis might be more predictable as well, because enzyme and substrate could interact almost imme- diately, whereas with the previous protocol, enzyme access to the substrate may have been delayed by slow absorption. In addition, potential problems with modified enzyme accumulating in the peritoneum could be avoided. Directing the enzyme to a site other than the peritoneum would be an Option as well with administration into the vasculature. Polyethylene glycol-enzyme adducts have characteristics suitable for enzyme administration and can be infused intravenously (Abuchowski e_t a_l., 1977; Savoca gt a_l., 1979; Abuchowski and Davis, 1981). A number of other enzymes have been modified by attachment of PEG and their characteristics examined. The PEG-enzymes are less susceptible to proteolytic attack and covalent attach- ment of this polymer to certain proteins prolongs their circulating half-life. By comparison, the immodified enzymes are inactivated rapidly by proteolytic digestion and have very short half-lives in the circulation. Importantly, the immimogenicity of these foreign enzymes is also diminished by the reaction with PEG. One hypothesis for the ability of PEG to confer these properties is that the polymer forms a protective layer around the enzyme. In this way, attack by proteolytic enzymes is prevented and recognition of the foreign enzyme by the immune system is avoided. 21 All of the properties discussed above are desirable when considering administration of an enzyme in therapy. The ability of PEG-conjugation to confer these characteristics may be variable however, since the catalytic properties of the bound enzymes may be affected to various extents by the modification. The degree to which PEG is able to decrease immunogenicity and enhance stability may differ with different enzymes. Rather than yielding a product of uniform preperties regardless of the protein modified, it may depend upon the character- istics of the particular enzyme. The effects of modifying gulonolactone oxidase with the PEG polymer were therefore examined. The product of the conjugation was characterized with respect to its catalytic properties and stability jg _v_i_t_rg and _i_n_ 3122. The results of administering this modified enzyme to vitamin C-deficient guinea pigs, as well as the immunogenicity of this PEG-enzyme, were also studied. MATERIALS AND METHODS I. General A. Animals English short hair guinea pigs (strain, Mdh:(SR[A])) were obtained from the Michigan State Health Laboratories (Lansing,MI). Animals used for enzyme administration experiments were 3-7 day old males weighing 130-170 g. They were depleted of ascorbic acid by feeding them either short-term or long-term ascorbic acid-deficient diets (ICN Nutritional Biochemicals, Cleveland,OH). Guinea pigs used for immtmization were mature males (400-500 g). Male ICR mice weighing 18-22 g were obtained from Harlan Sprague-Dawley (Indianapolis, IN). New Zealand White male rabbits weighing 2.5-3.0 kg were purchased from Bailey Rabbitry (Alto, M1) and a horse was utilized from the College of Veterinary Medicine (Michigan State University, East Lansing, MI) for the purpose of raising antisera. B. Surgical procedures For some enzyme administration studies, guinea pigs were catheter- ized via the carotid artery in order to obtain plasma samples and via the jugular vein for the infusion of enzyme, substrate, or both. These surgical procedures were carried out using ketamine-acepromazine anesthesia (Shugard gt a_l., 1975). Lidocaine (0.1 ml) was injected subcutaneously at the incision site to produce local anesthesia. The animals were allowed to recover for at least one day prior to treatment. 23 C. Preparation of enzymes Gulonolactone oxidase (EC 1.1.3.8) was purified from chicken kidneys obtained from Pel Freez Biologicals (Rogers, AR). The purification procedure is similar to the one previously described by Sato and Grahn (1981) with some minor modifications. All procedures were carried out at 0-4°C and are described below. Chicken kidneys (100 g) were thawed in 1.15% potassium chloride and homogenized in 4 volumes of 0.25 M sucrose. This homogenate was centrifuged at 103 g for 10 min, the resulting supernatant saved at 0°C and the pellets washed with 1.15% potassium chloride, 10 mM EDTA (pH 7.5) and centrifuged again (103 g for 10 min). This supernatant was then combined with the one from the previous spin and centrifuged for 45 min at 100,000 g. The resulting pellets was washed by homogenizing in 1.15% potassium chloride, 10 mM EDTA (pH 7.5) and centrifuged again for 45 min at 100,000 g. This pellet was suspended at a protein concentration of 10 mg/ml in 20 mM Tris-acetate buffer (pH 8.0) with 1 mM EDTA. Trypsin (0.1 mg/ml) was added to this suspension and it was stirred overnight, at 4°C, under nitrogen. The trypsin-digested pellet suspension was centrifuged at 100,000 g for 60 min. The resulting pellet was suspended at a protein concentration of 10 mg/ml in 20 mM Tris-acetate buffer (pH 8.0) with 1 mM EDTA that contained 3% Tween 20 for solubilization and stirred for 30 min at 4°C. Following centrifuga- tion at 100,000 g for 75 min, the pellet was discarded and the supernatant portion was frozen at -20°C overnight. Ammonium sulfate fractionation was carried out by adding finely ground solid ammonium sulfate to the solubilized preparation, slowly, with constant stirring. The first addition of ammonium sulfate was 130 mg/ml, after which the preparation was centrifuged at 30,000 g for 15 min, and the pellicle removed. A second addition of ammonium sulfate (145 mg/ml) was carried out, 24 followed by centrifugation at 30,000 g for 30 min and the resulting pellet was dissolved in 7 ml of Tris-acetate buffer (pH 8.0), which contained 10 mM potassium chloride, 1 mM EDTA, and 0.4% of the detergent Brij 35. This preparation was dialyzed extensively overnight against 4 liters of the same Tris- acetate buffer in which it was dissolved. The dialysis buffer was changed once. This preparation was passed through a DEAE Sephadex A-50 column (2.0x15 cm) equilibrated with the same Tris acetate buffer. The tmadsorbed fractions which possessed high specific activity were combined and frozen at -85°C under nitrogen. This resulted in the purified enzyme preparation that was used for the PEG modification procedure. When a more highly purified prepara- tion was necessary, the fractions from the DEAE chromatography step were dialyzed overnight against 10 mM potassium phosphate buffer (pH 7.0) containing 1 mM EDTA and 0.4% Brij 35. Adsorption chromatography was carried out using a hydroxylapatite column (2.2x4 cm) that was equilibrated with the same 10 mM potassium phosphate buffer. Elution with 100 mM potassium phosphate buffer (pH 7.0) containing 1 mM EDTA and 0.4% Brij 35 was performed. Fractions from the first peak and the elution peak were assayed for enzyme activity. The fractions with high specific activity were concentrated to greater than 0.5 mg/ml protein using an Amicon ultrafiltration apparatus equipped with a YM 5 membrane. Enzyme was stored frozen at -85°C under nitrogen. Human serum cholinesterase (EC 3.1.1.8) was purified from frozen plasma by affinity chromatography as described by Lockridge and LaDu (1978) and was a generous gift from them. Partially purified histidase (EC 4.3.1.3) from E. fluorescens, a- galactosidase (EC 3.2.1.22) from E. 59g, and chromatographically purified aspara- ginase Type 11 (grade VIII) (EC 3.5.1.1) from E. ggl_i were purchased (Sigma Chemical Co., St. Louis, MO). 25 D. Analytical methods Protein concentrations were determined by the method of Lowry gt gl. (1951) using bovine serum albumin as the standard. Published methods were used to assay asparaginase (Ho 3 a_l., 1970), histidase (Tabor and Mehler, 1955), serum cholinesterase (Kalow and Lindsay, 1955), and a-galactosidase (Petek gt g, 1969). Gulonolactone oxidase activity was determined by the method of Sato and Udenfriend (1978b). The method of Zannoni g g._l. (1974) was used to measure ascorbic acid. E. Statistical analysis Kinetic data and tissue ascorbic acid concentrations are expressed as means i S.E.M. In experiments where comparisons are made between two groups, the Student's t-test was used to compare the two means. Comparison of ascorbic acid synthesis between gulonolactone oxidase-treated and control animals was analyzed using a mixed design ANOVA and individual comparisons were made with either the least significance difference (lsd) test or Dunnett's test for between group comparisons (Steel and Torrie, 1980). Significance was established at p < 0.05. II. Heterologous Immtmoprecipitate Studies A. Preparation of antisera Gulonolactone oxidase, asparaginase, histidase and serum cholinester- ase were electrophoresed in 7% polyacrylamide gels (Dewald gt a_l., 1974). Gulonolactone oxidase gels contained 1.5% Tween 20. The location of this enzyme in the gels was determined by staining for activity (Nishikimi g g_l., 1976). Serum cholinesterase and asparaginase activities migrated in the gels with the only major 280 nm light-absorbing hand. So, these enzymes were located by scanning the gels in a spectrOphotometer at this wavelength. Histidase was 26 located by slicing the gels and assaying the individual segments for activity. Gel segments containing enzyme activity were excised and homogenized in equal parts of saline and Freund's complete adjuvant (GIBCO Laboratories, Grand Island, NY). Guinea pigs received injections of the enzyme (40 1.1g) intramuscularly and rabbits were injected with 40 u g in the footpads. Three weeks later, similar booster injections were given. Guinea pigs were bled by cardiac puncture under anesthesia (Shugard g g_l., 197 5) and rabbits were bled via the central ear artery one week after the booster injection. A horse was immtmized and boosted with gulonolactone oxidase, according to a similar schedule, with gel segments containing 0.5 mg of the enzyme. It was bled from the jugular vein one week after the booster injection. One week later, the horse was boosted again and bled the following week. All sera were stored frozen at -20°C. B. Purification of immunoglobulin G IgG was purified from the horse antiserum by chromatography on QAE Sephadex (Joustra and Lundgren, 1969). Serum (50 ml) was thawed and centri- fuged for 20 min. The clear supernatant was diluted with 100 m1 of ethylene diamine-acetic acid buffer (pH 7.0). This solution was loaded onto a QAE Sephadex A-50 column (4.5x20 cm) equilibrated with the same buffer. The break- through fraction was immediately concentrated to approximately 10% of its original volume, using an Amicon ultrafitration cell equipped with an XM 50 membrane, and the fraction lyophilized. Using this procedure, the average yield of IgG from 50 ml of sera was 1.61:0.12 g dry weight. C. Preparation of immimOprecipitates For the studies" using whole antiserum to prepare immtmoprecipitates, chicken kidney microsomes (5 mg protein/ml) were solubilized using 0.7% sodium deoxycholate as described (Nishikimi and Udenfriend, 1976) and concentrated to 10 mg protein/ml. Immunoprecipitates were made by mixing antiserum (13 ml) 27 with solubilized microsomes (17 ml). The mixtures were allowed to stand overnight at 4°C. They were then centrifuged for 20 min at 30,000 g at 4°C, suspended to their original volume in 0.1 M potassium phosphate buffer (pH 7.4) containing 2 mM EDTA, and centrifuged again. When immunoprecipitates were prepared with purified horse IgG, chicken kidney microsomes (10 mg protein/ml) were solubilized using 0.7% sodium deoxycholate. Purified horse IgG was dissolved in 0.1 M potassium phosphate buffer (pH 8.0) to a concentration of 20 mg/ml and mixed with the solubilized microsomes. For each ml of microsomes, 2.5 mg of IgG was used. The mixtures were allowed to react overnight at 4°C, centrifuged for 40 min at 100,000 g at 4°C, suspended to 60 ml in 0.1 M potassium phosphate buffer (pH 7 .4) containing 2 mM EDTA, and centrifuged again. Inactive immunoprecipitates for the treatment of certain control groups were made using microsomes which had been obtained from kidney homogenates prepared in 1.15% potassium chloride with 10 mM EDTA at pH 4.5. Immtmoprecipitates of E. ggl_i asparaginase were obtained by mixing 1.5 mg of asparaginase in 1 ml of 0.2 M Tris buffer (pH 8.6) with 13 ml of anti- asparaginase serum. Histidase immunoprecipitates were prepared using 30 mg 3. fluorescens cells. These cells were suspended in 3 m1 of 0.1 M Tris buffer (pH 9.0), sonicated for 2.5 min at 2°C, and centrifuged for 60 min at 100,000 g at 4°C. The supernatant was then mixed with 10 ml of anti-histidase serum. Serum cholinesterase was precipitated from 27 ml of human serum with 3 ml of antiserum. Reaction of enzymes with the respective antisera was carried out overnight at 4°C. The immunoprecipitates were isolated by centrifugation at 40,000 g for 30 min, washed by resuspension in their appropriate buffer, and recentrifuged. 28 D. Glutaraldehyde reaction The procedure used was adapted from Habeeb and Hiramoto (1968). Immtmoprecipitates made from whole antisera were suspended to 30 ml in 0.13 M potassium phosphate buffer (pH 6.5). Glutaraldehyde (25mM; Sigma Chemical Co. St. Louis, MO) was added slowly, with mixing, to a final concentration of 8.3 mM at 0°C. This was centrifuged for 25 min at 15,000 g at 4°C. Precipitates were then resuspended in 0.1 M potassium phosphate buffer (pH 7.4) containing 2 mM EDTA and centrifuged again. When pure IgG was used to obtain the immunopreci- pitate, conditions were the same except that crosslinking was carried out at pH 7.0 to obtain a preparation that was considered to be crosslinked as extensively as possible, while maximally preserving enzyme activity. The crosslinked precipitate was centrifuged for 25 min at 50,000 g at 4°C. Catalytically inactive crosslinked immunoprecipitated gulonolactone oxidase was prepared by using 166 mM gluta- raldehyde and carrying out the reaction for 1 hr at 25°C. E. Enzyme administration Glutaraldehyde-reacted immunoprecipitates of gulonolactone oxidase were suspended in 6% dextran in 0.9% sodium chloride (Abbott Laboratories, North Chicago, 11..) at a concentration of 5.3 mg dry weight of precipitate per ml, when prepared from whole antisera. A total volume of 5 m1 of the suspension was injected intraperitoneally into the guinea pigs. This amount of the modified gulonolactone oxidase is able to produce 0.07 mg ascorbic acid/ min at 37°C in Q gi_tr_g assays. Crosslinked immtm0precipitates, made using purified horse IgG, were also suspended in 6% dextran in 0.9% sodium chloride (Abbott Laboratories, North Chicago, IL) and injected intraperitoneally. The preparation for a single dose had a dry weight of about 20 mg per ml of suspension. 29 Crosslinked immunoprecipitates of serum cholinesterase, asparaginase, and histidase were suspended at concentrations of approximately 5.2, 2.2, and 0.1 mg dry weight/ml, respectively. Mice were given 1 ml and guinea pigs were given 5 ml of these suspensions by intraperitoneal injection. F. Postmortem examination Selected guinea pigs were sent to the Animal Health Diagnostic Laboratory (Michigan State University, East Lansing, MI) in order to determine their cause of death or to look for possible pathological changes caused by the enzyme replacement therapy. First, the occurrence of any gross lesions was noted. Histopathological and bacteriological examination of sections of various tissues was conducted. The tissues examhied included the following: cerebrum, cerebellum, midbrain, duodenum, liver, spleen, kidney, lung, tongue, ileum, adrenal glands, skeletal muscle, bladder, bone marrow, lymph nodes, and stomach. Vitamin A and E levels were determined in the liver. The levels of selenium and other minerals were determined in the liver and kidney. G. Radiochemical methods 14C-Labelled enzyme complex was prepared by reductive alkylation of the immunoprecipitates following the glutaraldehyde crosslinking reaction. Im- mimoprecipitates were prepared and crosslinked with glutaraldehyde as described previously for the treatment of animals. Radiolabelling was carried out as described by Rice and Means (1971) using 14C-formaldehyde (10 mCi/mmol, New England Nuclear, Boston, MA). 14C-Formaldehyde was added to the suspended crosslinked immimOprecipitates and allowed to react for 10 min. Following this reaction, sodium borohydride was added to reduce the formaldehyde. The labelled precipitates were then washed extensively by suspension in 0.1 M potassium phosphate buffer (pH 7.0) and recentrifuged to remove the unreacted l4C--labe1. 30 The washed, labelled precipitates were suspended and administered in the same way as the unlabelled immunoprecipitates. H. Preparation of samples for liquid scintillation spectrometry Guinea pigs that were treated with the 14C-labelled crosslinked immimoprecipitated gulonolactone oxidase were housed in urine and feces collec- tion cages. Plasma samples, urine, and feces from each animal were obtained each day. After the specified period of time, animals were sacrificed and tissues were obtained for counting. Tissue samples of 200 mg were dissolved in 1.0 ml of TS—l tissue solubilizer (Research Products International Corp., Mount Prospect, IL). Glacial acetic acid (0.2 ml) was added and samples allowed to stand at room temperature for 45 min to minimize chemiluminescence. Following this procedure, 10 ml of Safety Solve counting cocktail (Research Products hternational Corp.) was added to each vial and samples were counted as described below. Plasma samples (0.5 ml each) were counted in the form of an emulsion (Kobayashi and Maudsley, 1974). Triton X-100 (Research Products International Corp.) (5 ml) was added to each sample , along with 5 ml of the Safety Solve counting cocktail. After thorough mixing, samples were heated for 20-30 min at 80°C, cooled for 2-4 hr at 4°C and then allowed to equilibrate to room temperature prior to counting. Urine samples (0.5 ml) were counted in 10 ml of comiting cocktail. Feces samples (20 mg) were prepared for counting by adapting the method of Mahin and Lofberg (1966). Water (0.180 ml), 0.2 ml of perchloric acid, and 0.4 m1 of 30% hydrogen peroxide were added to the feces samples. They were heated at 80°C for 1-2 hr, and then allowed to cool to room temperature. Counting cocktail (10 ml) was added and after 45 min, the samples were counted. 31 1. Liquid scintillation counting Carbon 14-labelled samples were counted using a Beckman LS7000 scintillation counter. Samples were counted in a full C-14 window (95% of all C- 14 emissions). Accuracy of comting was i 5%. Disintegrations per minute were calculated using a quench curve. The quench curve was generated by comting suples that contained varying amounts of tissue and a known amoimt of C14. The H-number technique (Long, 1977) was then used to determine the degree of quench. III. Polyethylene Glyco l-Conjugated Enzymes A. Modification of enzymes with polyethylene glycol Gulonolactone oxidase was purified from chicken kidneys as described above. For this modification, adsorption chromatography through hydroxyapatite was not performed, however. After DEAE Sephadex chromatography, the fractions containing gulonolactone oxidase activity were concentrated, imder nitrogen, using an Amicon ultrafiltration imit equipped with a PM 10 membrane. Prior to the modification procedure, the buffer of this enzyme preparation was changed to 0.1 M potassium phosphate (pH 8.0) using a spun column containing G- 25 Sephadex. The primary amine groups of this enzyme preparation were measured using the fluorescamine method (Udenfriend g g” 1972; Bohlen gtgl” 1973). Polyethylene glycol (PEG) 5000, activated with succinimidyl succinate (Enzon, lnc., Piscataway, NJ), was reacted with the enzyme. An amount of this polymer representing a 5-fold molar excess of the primary amine groups m the enzyme preparation was used to modify the enzyme. The method of Abuchowski g a_l. (1977a) was adapted for this procedure. The activated PEG was reacted with the enzyme for 90 min in 0.1 M potassium phosphate, pH 8.0. Unreacted PEG and derivatives were removed by washing with 0.9% sodium chloride using an 32 ultrafiltration cell equipped with a PM 30 membrane and overnight dialysis against 0.9% sodium chloride. a-Galactosidase from E. ggg (Sigma Chemical Co., St. Louis, MO) was modified by the same procedure. B. Enzyme administration PEG-reacted gulonolactone oxidase was infused at a rate of 0.2 ml/min into the jugular vein. Doses of modified gulonolactone oxidase that were able to produce 0.01 to 0.07 mg ascorbic acid per min at 37°C in Q gittg assays were tested. The PEG-conjugated u-galactosidase was also given via the intravenous route. RESULTS 1. Heterologous Immtmoprecipitate Studies Previous studies showed that guinea pigs could be maintained without dietary ascorbic acid by administration of crosslinked immunoprecipitated (XL-1P) gulonolactone oxidase (Sato _e_t gl., 1986). Specific guinea pig antisera were used to immtmoprecipitate the enzyme in those experiments. It was concluded that the multiple injections were tolerated because homologous antibody masked the antigenic determinants of the foreign enzyme. Thus, the guinea pigs did not recognize the complex as being allergenic. An alternative explanation is that the modified enzyme was tolerated because it was immobilized within a large, stable complex. The studies described here were designed to test whether detoxification of the enzyme was dependent on the use of homologous antiserum or some other characteristic of the crosslinked enzyme immunocomplex. A. Modification of gulonolactone oxidase Gulonolactone oxidase was precipitated from chicken kidney micro- somes using specific antisera raised in rabbits. This was followed by crosslinking the precipitate with glutaraldehyde. Reaction with this bifunctional reagent is intended to reinforce the interactions between the antigen (enzyme) and antibody. The scheme for this modification procedure is shown in Figure 2. The amounts of antisera and solubilized microsomes necessary to obtain maximum precipitation of enzyme activity were determined. Then, appropriate conditions in which to ‘11 34 Figure 2. Protocol for the preparation of crosslinked immunoprecipitates. The source of gulonolactone oxidase is solubilized chicken kidney microsomes. The enzyme is isolated by reaction of this crude tissue with antibody (IgG or serum) directed against the enzyme. The resulting immunoprecipitate is collected and washed by resuspension in potassium phosphate buffer followed by ultracentrifu- gation. It is then reacted with glutaraldehyde, a bifunctional reagent that reacts mainly with primary amines of proteins. Formation of these covalent crosslinkages is intended to reinforce the existing bonds between the enzyme and antibody. The crosslinked complex is suspended in dextran and injected intraperitoneally into the animal. 35 069° oY-xo Q° 0 °>-O '< 0 00%. .OAQ lmmunoprscipitsto o Hiccriicuicnic’fiz Glutaraldehyde . ’ Crossllnkad lmmunoprecipltato Figure 2 36 crosslink the enzyme-antibody complex with glutaraldehyde were studied. Condi- tions that provided extensive crosslinking with minimal inactivation of the enzyme were determined. 1. Crosslinking conditions The effect of hydrogen ion concentration on the crosslinking reaction was examined. Glutaraldehyde ((16.6 mM) was reacted with the im- munoprecipitates at pH values ranging from 6.0 to 7.5 and enzyme activity tested. The results from one such experiment are shown in Table l. The greatest enzyme activity was obtahied when the crosslinking reaction was carried out at pH 6.5 or 7.0. Subsequently, this reaction was performed at pH 6.5. Next, the concentration of glutaraldehyde was examined. Con- centrations of this reagent ranging from 3.3 to 83 mM were reacted with the immunOprecipitates and enzyme activity was assayed. Also, the extent of the glutaraldehyde reaction was assessed by reacting the crosslinked immimoprecipi- tates with fluorescamine (Udenfriend gt a_l., 1972; Bohlen gt a_l.,1973). Glutaraldehyde reacts predominantly with the primary amines of proteins (Habeeb and Hiramoto, 1968). Fluorescamine also reacts with primary amines to yield a fluorescent product. Reaction of the complex with fluorescamine, therefore, should reflect the extent of the glutaraldehyde crosslinking reaction. The results from a representative experiment are shown in Figure 3. Both the enzyme activity and fluorescence decreased with increasing concentrations of glutaralde- hyde. For subsequent crosslinking reactions, 8.3 mM glutaraldehyde was selected. At this concentration, 70% of the enzyme activity was retained, while there was a 70% decrease in fluorescamine-reactive groups. 2. Catalytic characteristics The kinetic parameters of the enzyme were compared in order to examine how they had been affected at different steps of the modification. 37 TABLE 1 The effect of pH on the Glutaraldehyde Crosslinking Reaction pH Activity (nmol/min/mg) 6.0 0.0 80 6.5 0.125 7.0 0.120 7.5 0.090 Unreacted 0.330 38 Figure 3. Relationship between the extent of crosslinking and enzyme activity with increasing concentrations of glutaraldehyde. Immunoprecipitated gulonolac- tone oxidase was crosslinked using glutaraldehyde concentrations of 3.3 to 83 mM. Results from a representative experiment are shown here. Enzyme activity was determined following crosslinking and activity is expressed as nmol ascorbic acid formed/min/mg microsomal protein (H). Unreacted primary amines were determined using fluorescamine. The results are expressed as relative fluorescent units (O--O). 39 0 0 5 _ mmgfizm .urcOmmmOmZn. €2.4m 5 O 5 7. 5 2 3 2 .1. _ . - 655.5355 >:>:.o< .75 50 ‘GLUTARALDEHYDE (mM) 25 Figure 3 40 Activity of the modified and immodified preparations was tested at varying substrate concentrations. The apparent Km and Vmax of the unmodified and modified enzymes were calculated from Lineweaver-Burk double reciprocal plots. Recovery and apparent kinetic constants were compared among the unmodified, immimoprecipitated, and crosslinked immunoprecipitated enzymes (Table 2). Recovery of activity was based on the amount of enzyme mixed with the antiserum, with the assumption that an antiserum concentration at which no detectable activity remained in the supernatants after centrifugation represented precipitation of all of the enzyme. To calculate recovery, activity of the modified enzyme was compared to unmodified enzyme at the same substrate concentration used in the usual assay. Upon immunoprecipitation, about 80% of enzyme activity was lost. However, no further decrease in enzyme activity occurred with crosslinking. Comparing the crosslinked immunocomplex to the microsomal preparations, Vmax decreased, but Km values were not significantly different. B. Administration of modified gulonolactone oxidase 1. Single dose studies Following development of this XL-IP gulonolactone oxidase, experiments were performed to determine whether a single dose was able to elicit ascorbic acid synthesis in guinea pigs. Vitamin C-deficient animals were injected intraperitoneally with the XL-IP enzyme complex suspended in dextran. The substrate, gulonolactone, was infused 'mtravenously. Control animals were treated with histidase or asparaginase that had been modified similarly. Plasma was analyzed for ascorbic acid (Figure 4). During a 5-h period, plasma ascorbic acid concentrations of the gulonolactone oxidase-treated animals increased more than three-fold. At the end of the sampling period, plasma concentrations of ascorbic acid in these animals were 0.33 mg/100 ml, which approached concentra 41 TABLE 2 Comparison of Kinetic Parameters of L-Gulonolactone Oxidase and the Modified Enzyme Unmodified Immuno- X-Linked Enzyme precipitate Complex Recovery 100 19.9134 l9.0t2.2 Km (mM) 0.06:0.01 0.08:0.02 0.07:0.01 Vmax (nmol/min/mg) 4.48:1.02 0.89:0.22 1.04:0.10 The apparent Km and Vmax of the modified and unmodified enzymes were calculated from Lineweaver-Burk double reciprocal plots. Recovery is expressed as a percentage of the unmodified enzyme activity at the substrate concentra- tions used in the usual assay. Km is expressed as a mM concentration. Vmax is expressed as nmol ascorbic acid formed/min/mg microsomal protein. 42 Figure 4. Synthesis of ascorbic acid in gulonolactone oxidase-treated animals. A group of four 14-day ascorbic acid-deficient guinea pigs was injected with the XL- IP chicken kidney gulonolactone oxidase H). A control group of four l4-day ascorbic acid-deficient animals was given either histidase or asparaginase prepared similarly (O--O). Plasma ascorbic acid concentrations were measured. Gulonolactone (100 mg/2 ml) was administered intravenously immediately prior to injection of the enzyme and 800 mg/5 ml was infused over a 3-hr period afterward. At zero time, the enzyme suspension was injected intraperitoneally. Blood samples were taken every 30 min. Three samples were drawn before injection of the enzyme and sampling was continued for a 5-hr period thereafter. Statistics were done by mixed design analysis of variance and the vertical bar denotes t SEM for between group comparisons. The 120-min and all subsequent time points of the gulonolactone oxidase-treated group are significantly higher than the control (p < 0.05). The standard error for within group comparisons is 10.02 for each group. PLASMA ASCORBIC ACID (mg/100 ml) 0.4 0.3 0.2 0.1 43 Gulonolactone Oxidase TIME (hours) Figure 4 44 tions found in guinea pigs fed a saturating amount of vitamin C, when analyzed using the same method (Zannoni e_t a_l., 1974). On the other hand, plasma ascorbic acid concentrations in the control animals remained relatively constant. 2. Effect of repeated doses of the modified enzyme Since the XL-IP gulonolactone oxidase possessed '2 g_i_v_o_ activity, studies were planned to test whether repetitive treatments could be tolerated and therapeutically beneficial to the animals. The ability to tolerate repeated injections of foreign protein is a vital consideration in the treatment of enzyme deficiency diseases. Afflicted individuals may require therapy for their entire lifetime. To examine these aspects, guinea pigs were maintained on an ascorbic acid-deficient diet and given weekly injections of the XL-IP gulonolactone oxidase along with substrate supplementation. The growth curves of three animals treated by this protocol are shown in Figure 5 (A-C). These animals survived for at least 50 days on the ascorbic acid-deficient diet with enzyme replacement therapy. Ten animals received the control treatment that consisted of either histidase or asparaginase modified according to the same method. The average weight gain of these control animals is shown for comparison hi part D of Figure 5. These animals survived an average of 23.612.62 days on the deficient diet. The weekly gulonolactone oxidase injections prolonged the survi- val time of the ascorbic acid-deficient animals. Guinea pigs ordinarily survive between 23 and 28 days on this vitamin C-deficient diet (Jones g a_l., 1973; Barnes gt a_l., 1973). The fact that the gulonolactone oxidase treatment prolonged the survival of scorbutic animals demonstrates E gilg synthesis of vitamin C, as well as a therapeutic benefit to these animals. 3. Toxicity of modified gulonolactone oxidase Fourteen animals were started on the therapy with this XL-IP Subnolactone oxidase; however, survival was prolonged significantly in only 7 of 45 Figure 5. Survival and prevention of scurvy in guinea pigs by enzyme administration therapy. Guinea pigs were placed on an ascorbic acid-deficient diet on Day 0 and weighed daily. On the days indicated by arrows, modified gulonolactone oxidase was administered intraperitoneally and gulonolactone (86 mg/0.6 ml) was injected at 20 min intervals subcutaneously over 2 hr. Parts A-C of this figure are the growth curves for 3 animals that survived for 50 days on this therapy. In D, the open circles represent the average body weight of 10 control guinea pigs. The standard errors of the mean body weights ranged from :6 to _+_8 grams. These animals received either histidase or asparaginase prepared by the same procedure and also supplementation of gulonolactone. C 300- 46 A 300- n-10 ' so 1 125 I l 10 .5 1 l m .5 o. 012 19¢ . Fan nu .U. _ 4 F s O O 0 O 0 0 0 0 n4 4| 00 n4 50 25 B V as so. In” o 1. . .. m 1» than. Ioowo o in... . as . 0 0 4: a. 59m; 50m 200- ' soo- 200~,. DAYS fiwn5 n-l 47 these animals, with only 3 surviving for the entire 50-day period. Since the therapy was beneficial for only some animals, the reasons for this limited success were investigated. A possible explanation for the survival of only certain animals is that animals that died might have had an allergic reaction, whereas those that survived did not. To determine whether the surviving animals had been sensitized immunologically to the foreign proteins, their serum was tested for antibodies against the components of the modified enzyme. A serum sample was obtained from each animal after the 50-day treatment and Ouchterlony immunodouble diffusion tests were carried out. Serum was tested for antibodies against the enzyme source (chicken kidney microsomes) and rabbit antiserum directed against another enzyme, _E. gqlt asparaginase. The results are shown in Figure 6. The precipitin line between the center well and the wells that contained microsomes in part C could not be seen unless stained with Coomassie Blue. Even then, the line was too faint to be seen in photographs of the plate, so the unstained plate is shown here. These guinea pigs formed antibodies against both microsomes and antiserum. Thus, the surviving animals did mount an immune response against these foreign proteins. These animals were also tested to find out whether they had been allergically sensitized to the components of the XL-IP gulonolactone oxidase complex. Following the 50-day enzyme treatment, the animals were challenged first with an intravascular dose of a solubilized microsomal extract from chicken kidneys (1 mg protein). One day later, each animal received 0.1 m1 of rabbit anti- asparaginase serum via the same route. Responses were recorded after each challenge and the data are presented in Table 3. No response was noted when the shocking doses of enzyme were administered. On the other hand, when serum from rabbits was injected, all three of the guinea pigs experienced anaphylactic 48 Figure 6. Ouchterlony immunodouble diffusion tests for antibody in serum of gulonolactone oxidase-treated animals. The wells contained the following: 1,4- saline; 2,3-solubilized chicken kidney microsomes; and 5,6-rabbit antiserum agaimt asparaginase. Center wells contained serum from the treated guinea pigs. The plates are labelled A-C to correspond to the animals represented in A-C of the growth curves (Figure 5). The precipitin line between the serum and the microsome well of plate C is not visible in the photograph. 49 _- . .I Figure 6 50 TABLE 3 Reaction of Guinea Pigs to Challenging Doses of Components of Modified Gulonolactone Oxidase Animal Shocking Agent Response A Enzyme No reaction Rabbit Antiserum Reaction -> recovery B Enzyme No reaction Rabbit Antiserum Reaction -> death C Enzyme No reaction Rabbit Antiserum Reaction -> death One week after the last dose of XL-IP gulonolactone oxidase, the animals were challenged first with the solubilized microsomal extract (enzyme source) and, one day later, with rabbit anti-asparaginase serum. Challenging doses were administered intravenously. 51 reactions. The animal whose sera is tested in Fig. 5A recovered, while the other two animals had terminal reactions. These tests also show that the surviving guinea pigs have been sensitized by this XL-IP enzyme preparation. The survival of the animals cannot be attributed to absence of an immune response against the foreign proteins. They tolerated multiple intraperi- toneal injections of the complex, but not the antiserum component when it was injected separately into the vasculature. While serum antibodies were present against the microsomal extract, they did not experience allergic reactions when challenged with it. C. Adaptability of the modification procedure to other enzymes If this procedure is to have practical applicability, it must be adaptable to other enzymes and not uniquely suited to gulonolactone oxidase. Therefore, studies were conducted with other enzymes that have potential therapeutic application. Adaptability of this procedure was demonstrated by using asparaginase, histidase, and serum cholinesterase. These enzymes were modified with either guinea pig or rabbit antisera and crosslinked with glutaralde- hyde. Kinetic parameters were determined for the resulting XL-IP enzyme complexes and compared to the native enzymes. Then, the toxicity of these XL- IP enzymes was tested in mice and guinea pigs. 1. Catalytic characteristics Apparent kinetic constants were affected to varying degrees by the modification (Table 4). The Vmax values varied from almost complete inhibition of activity to values that approached those of the immodified enzyme. All of the enzymes tested had a lower Vmax after they were reacted with antibody. The glutaraldehyde reaction had little further effect on this parameter. Km values varied widely as well. Serum cholinesterase modified using guinea pig antisera showed only small decreases in affinity, while asparaginase modified with 52 .833 no "8 ~09 “5030.293 5 56> was 833 a 03.530.5vo 830a 60333255 dares—0 “305.39 3 08530 doggone: end £32m mo: :09 6080.298 8 03230.85 .fiakguouagmono 820$ cosmonofifi 03.533 no U083“ “Duncan 308: as “030.270 5 “88> 43395200200 28 s as n030~nu0 a BM dreams 3:3 05 3 won: 3030320280 03333 05 «a 5:500 08.320 “5033082: 05 mo 0meun0ou0m s as pomegmuo 0.3 a0mu0>oo0m .303 Eugene?“ 03:36 xurmiu0>003025 Bonn "5058.233. 0:03 353.50 03051 «20.8.54 e...“ 3. 2: e; :3 2: rescue: 8832: etc 85 85 mm... mm... 2.5 58> See 355 :85 m8... :8... 255 a: 823.3220 3. 8 2: 2. S .2: r380: 83% 2: 2: 3m 2:. v.8 we” 58> S 5 ma 3” 28 e.~ 8: to 3: 2: $3 3 2: E283: namesake 0.233302“ 0u33mu0un : oumfihuoan 33330.5 8: -9888: -2588: 32:88 9 .. a: -9583 Base 9 332x 3254.3" 083:”.— nuonmug finned 8522 m5 eases was: 32:32 Susanna Baa—cos— mo 3308053 0305 w m4m<9 53 rabbit antisera had a much lower affinity. Modified histidase was not sufficiently active to allow determination of its kinetic constants. Z. Toxicity of these modified enzymes Mice and guinea pigs were given repeated intraperitoneal injec- tions of these modified enzymes to test the toxicity of such XL-IP complexes. The animals received either three or four injections of enzyme complex at two- week intervals. Mice received multiple injections of the enzymes modified by using either rabbit or guinea pig antisera, and guinea pigs were treated with rabbit antisera-enzyme preparations. All of the modified enzymes in this experiment were tolerated (Table 5). A total of 69 animals were treated without any observable allergic or other reaction to any of the treatments. These results suggest that such XL-lP enzyme complexes are safe for repeated administration. Furthermore, this modification procedure can be extended to enzymes other than gulonolactone oxidase. Animals in this study were also tested for an allergic response to the components of the XL-IP complexes used in their treatments. Some of the animals were challenged with unmodified enzyme. Others were challenged with nonimmune serum from the same species that was used to obtain antiserum for modifying the enzyme that they received. Doses were approximately 1.5 mg protein/animal. The results are summarized in Table 6. The ratios are the number of animals that exhibited an allergic response immediately following the challenge over the total number of animals tested. Human sera was used as the enzyme source for serum cholinesterase. Most of the animals that did not “exhibit any reaction to an initial challenge with enzyme were subsequently shocked with the apprOpriate nonimmune serum. This accounts for the higher total number of animals in this table. Characteristic signs of systemic anaphylactic shock were observed in the animals that reacted following the challenge. 54 TABLE 5 Reaction of Animals to Repetitive Injection of Enzymes Modified with Heterologous Antisera ASPARAGINASE Species Serum Source Responding Mice Guinea Pig 0/7 Mice Rabbit 0/ 6 Guinea Pig Rabbit 0/ 3 HISTID ASE Species Serum Source Responding Mice Guinea Pig 0/13 Mice Rabbit 0/ 6 Guinea Pig Rabbit 0/3 SERUM CHOLINESTERASE Species Serum Source Responding Mice Guinea Pig 0/12 Mice Rabbit 0/13 Guinea Pig Rabbit 0/6 Animals were injected 3 or 4 times with the modified enzyme at 2- week intervals. Mice received enzyme modified with guinea pig or rabbit antisera. Guinea pigs were given enzyme modified with rabbit antisera. The ratios are the number of animals that experienced symptoms characteristic of anaphylactic shock over the total number of animals challenged. 55 Table 6 Reaction of Animals to Challenging Doses of Components of the Modified Enzyme Complexes Species Shocking Agent Responding Guinea Pigs Heterologous sera 7/10 Histidase or Asparaginase 1/5 . Human sera 5/5 Mice Heterologous sera ” 29/38 Histidase or Asparaginase 0/6 Human sera ll/l3 Two weeks after the last dose of modified enzyme animals were challenged with either the enzyme or the nonimmune serum from the species from which the antisera was obtamed to modify the enzyme. Human serum was used as the source for serum cholinesterase. Ratios are the number of animals showing symptoms characteristic of an anaphylactic reaction over the total number of animals challenged. 56 Like the gulonolactone oxidase-treated guinea pigs these animals also have been allergically sensitized to some component of serum, but not to the administered enzyme. When challenged with an intracardiac (guinea pigs) or intravenous (mice) injection of enzyme, only 1 of 11 animals (a guinea pig) showed a reaction. It is not clear whether this reaction was a true anaphylactic reaction. It may have been related to the cardiac puncture. In contrast, 52 of 66 animals shocked with either the antisera or the human serum clearly experienced symptoms characteristic of an anaphylactic reaction. D. Modification of gulonolactone oxidase using purified IgG Since these studies showed that a component of antisera was capable of initiating an allergic reaction, IgG was purified from antisera in an effort to eliminate potentially toxic components. It was postulated that using purified IgG to modify the enzyme might result in a more consistent product and improve the success of the XL-IP gulonolactone oxidase therapy. For these experiments, antiserum against chicken kidney gulonolactone oxidase was raised in a horse. The IgG was purified using QAE Sephadex chromatography, according to the method of J oustra and Lundgren (1969). Polyacrylamide gel electrophoresis showed that the product obtained was essentially homogeneous (Figure 7). 1. Characteristics of the XL—IP enzyme complex Gulonolactone oxidase was precipitated from chicken kidney microsomes with the purified IgG. Optimum glutaraldehyde concentrations and crosslinking conditions were established. Electrophoresis of the resulting im mu- noprecipitate revealed fewer components in this preparation compared to those prepared using whole antiserum (Figure 8). The stability of this XL-IP gulonolactone oxidase preparation to heat denaturation and trypsin digestion was examined and compared to that of the unmodified enzyme in E vitro studies. Purified gulonolactone oxidase and the 57 350— 0a «A»? 80am 8 0030.:me £008.83? mama—80¢ _0w 8:3 non—0830835002» 0 8 8: 80 «0 008800 88 053 08008000 52> 008.30 0808 30m 05. 60.89800 88 30m mama-«0200802 no 0000003905030 0.33 a 38$ 2.: 68208.2 6:083 8:2 5:33:28 0»: 2:3 case: a e5 .2. :25: 82.: 5.355 0:0 0330a. no 00508 05 .3 082:0 0030808345 «amend 8300320 030: 80¢ @3389 .03 .833 098A 80am “Karma 0 833030888“ mo 80kg 03009305020 ~0w 0383.300on 4.. 0.83% 58 vumvamum owH 0000: pnsamxm uwH oqouauq< mayo: KIA 59 Figure 8. SDS-Polyacrylamide gel electrOphoresis of gulonolactone oxidase immu- noprecipitates. Samples were dissociated in the presence of mercaptoethanol and electr0phoresed in 7.5% polyacrylamide gels in the presence of 0.1% SDS, according to the method of Nishikimi e_t a_l. (1977). Scan A is purified chicken kidney gulonolactone oxidase. An immunOprecipitate prepared from whole antiserum is shown in scan B and scan C is an immunoprecipitate made using purified IgG. Scan D is the purified IgG component alone. The gels were stained with Coomassie Blue and scanned at 600 nm in a spectrophotometer with a gel scanning attachment. Migration is from right to left. 60 L ( Purified GLO Immunoprecipitate from Whole_Serum Immunoprecipitate from Pure IgG Purified IgG 61 XL-IP enzyme were incubated alone and in the presence of the proteolytic enzyme trypsin at 37°C. Enzyme activity was assayed at selected times over a 48-h:- period. The results from a representative experiment are shown in Figure 9. The activity of the immodified enzyme decayed rapidly at 37°C (t 1/2 = 12 min) and in the presence of trypsin (t 1/2 = 9 min). In contrast, there was virtually no decay in the activity of the modified gulonolactone oxidase over a 3-hr period. The half-er of activity of this XL-IP enzyme was on the order of 24 hours. Even in the presence of trypsin, its half-life was approximately 3 hr. Thus, compared to the immodified enzyme preparation, gulonolactone oxidase activity in the XL- IP enzyme complex was greatly stabilized and protected from proteolytic degradation to some extent 2. Administration of the less contaminated complex This XL-IP gulonolactone oxidase was administered to guinea pigs to test its E 3132 activity. The scorbutic guinea pigs also received subcutaneous injections of the substrate, gulonolactone. Elevations of plasma ascorbic acid in guinea pigs after a single intraperitoneal injection demonstrated that this modified enzyme had ascorbic acid synthetic capability (Figure 10). Over the 5-hr period, there was a two-fold increase in plasma concentrations of the vitamin. In order to show that this complex could be given safely over an extended period, normal guinea pigs were treated with this XL—IP gulonolactone oxidase. A group of four animals received repeated injections of the modified enzyme for a period of 100 days, while being maintained on a normal vitamin C- containing diet. All four animals survived up to 14 weekly injections and they gained weight at a rate comparable to untreated normal guinea pigs (Figure 11). 62 .8000 m0— 0 00 53300 00308 003 .00 0000000 0 00 000000800 8 000 08000 00:00 0 00>0 0083 00800» 30 0000000 003 508000 0850M .0888 000 008003 0000 0300000 «0 0—080 H 80:00 53300 30 5v :8 000 000300000 0800 000 00000 ABLE =8\m8 m.8 809503 033 000 A; 0080 00000008 080 003 =8\50>300 D 8.3 08.300 0088000000 .0000380000000880 00F .Uopm 00 00800808 500 00308 0 8 5:9 =8xw8 m.3 800003 .080000 038000000 05 no 00000009 003 8 00.0 A1 0080 00000008 008 =8\hfi>300 D NA: 000003 00000800016 0033a .0000 03000 000 000830800 0030000000000 0 800m 03—0000 00h. 8030080 80%» 000 Conn 30 00300008 03 0000000 0003080003m Huang" «0 533030 00000008 .0 000th 63 o 0308 A0505 08:. O P Rag/«93v awfizua lemon 00 64 Figure 10. Synthesis of ascorbic acid by guinea pigs treated with gulonolactone oxidase modified using purified horse IgG. Two ll-day ascorbic acid-deficient guinea pigs received intraperitoneal injections of XL-IP gulonolactone oxidase (H). ‘Ihe substrate, gulonolactone (100 mg/l ml), was injected subcutaneously prior to the enzyme injection. Subcutaneous injections (0.6 ml) of gulonolactone (800 mg/5 ml) were given at 20 min intervals for a 3-hr period afterward. Two plasma samples were tested to establish pretreatment ascorbic acid levels. Plasma was tested at 30 min intervals following enzyme injection and ascorbic acid was measured by the method of Zannoni g _a_l. (1974). It has been demonstrated previously that control animals given modified proteins other than gulonolactone oxidase and substrate maintain relatively constant plasma ascorbic acid concentrations (—-) (Hadley g a_l., 1987). 65 =5: 0'2 r O O E 2‘ V Gulonolactone Q Oxidase 0 < 2 0-1 - m c: O U U) < — < \ 6% Control E O I I J 0 3 6 TIME (hours) Figure l 0 66 .303 am 00 800. 8020. 883 00 03000 000 000005 0008008000 00000000 0000000000006 0080000 0000 0008000 000003000000 0300000 0 00 00830.3 000.— 0m000>0 05 .008000800 00.0 .3300: no 0008000009 305 0090035 0030000 008000 05 3 00003000 0008000 0000000 008000 0000000000 000 03000 00300m 0 8 00: 0:00 080. .0000 008000 0 00 00800808 w80n 0033 600080 00000000003 00qu 0303 0000 2: 000 0000000 0008000 0 no 000803 000: 0w000>0 003 «00000000 0000000 0000 000. 0008000 0000000 008000 000000000 00 00000800 090 008% 030000000 0000007080000 000 080 00806 008000 0000000680000 «0 0000 033000 .00 000mg 67 O O 8 o o *0 V N (5)1HDISM A008 100 50 75 DAYS Figure 11 25 68 Pathologic examination of these enzyme-treated animals at the end of this 100- day period confirmed that the enzyme was not toxic. Thus, this preparation was safe for long-term administration. A product that contained less contaminants had been prepared, its synthetic capability demonstrated and safety established, so the possibility of treating scurvy was tested. Attempts to show a long-term therapeutic benefit were unsuccessful, however. The success rate of this therapy was not improved compared to that of the protocol using enzyme modified with whole antiserum. None of four guinea pigs treated with this preparation survived for 50 days on the therapy, while being maintained on the ascorbic acid-deficient diet. From these studies, it must be concluded that the poor rate of successful therapy was not the result of a few preparations containing a toxic component. 3. Analysis of doses of enzyme activity administered The possibility that inadequate ascorbic acid synthesis contri- buted to the early death of gulonolactone oxidase-treated ascorbic acid-deficient guinea pigs was also considered. Surviving animals may have been better able to tolerate marginal vitamin C deficiency. Based on the plasma and tissue ascorbic acid concentrations of the vitamin, the amount of vitamin C synthesis catalyzed in the enzyme-treated animals was believed to be adequate however. Compari- sons were also made between the doses given to survivors of a lOO-day enzyme therapy protocol and those given to the nonsurvivors (Sato e_t a_l., 1986). There was no correlation noted~ between successfully treated animals and those that received the highest amounts of activity. Furthermore, the average doses were not significantly different. The average dose given to surviving animals was 0.69:0.06 U/treatment, while animals that did not survive received 0.66:0.03 U/treatment. It was concluded, therefore, that the doses of enzyme activity adminktered were adequate. 69 4. Supplementation of ascorbic acid during gulonolactone oxidase therapy To further examine the possibility that insufficient doses of enzyme activity and, subsequently, inadequate vitamin C synthesis contributed to the death of the gulonolactone oxidase-treated animals, an additional study was performed. A group of guinea pigs was maintained on the ascorbic acid-deficient diet and given injections of modified gulonolactone oxidase every 4 or 5 days. In addition to the enzyme therapy, these guinea pigs received 6 mg of vitamin C mixed with the dose of substrate and injected subcutaneously. This amount of vitamin C approximated both the minimum requirements of this vitamin for animals of this size (Chatterjee, 1967), as well as the estimated amount of the vitamin synthesized by a single enzyme injection. If the reason for the low rate of therapeutic success involved inadequate ascorbic acid synthesis in the enzyme- treated animals, then vitamin C supplementation should remedy the situation. After several weeks on this therapy, the condition of these animals suggested that simply providing adequate amounts of ascorbic acid was not equivalent to treating animals fed the normal vitamin C-containing diet. 5. Supplementation of other vitamins during enzyme therapy Another difference between these two groups was their diets. Animals fed the ascorbic acid-deficient diet might not have been obtaining adequate nutrition if the ascorbic acid-deficient diet was deficient in nutrients other than vitamin C. Other evidence supported this hypothesis. For example, even the deficient animals that survived the previous lOO-day protocol (Sato _e_t_ a_l., 1986) did not grow at the same rates as untreated and enzyme-treated normal guinea pigs (Figure 11). In addition, postmortem examination of animals that did not survive the gulonolactone oxidase treatment revealed that these animals were deficient in vitamins A and E. Perhaps supplementation of other nutrients is 70 necessary for the treated animals that are maintained on the vitamin C-deficient diet. Studies were designed to examine whether the success of the XL-IP gulonolactone oxidase therapy could be improved by supplementing the animals maintained on the ascorbic acid-deficient diet with additional nutrients. Improvement of the nutritional status of such animals might be a critical factor for increasing the rate of success of the enzyme replacement therapy protocol. Five guinea pigs were started on the ascorbic acid-deficient diet and treated weekly with active XL-IP gulonolactone oxidase. In addition to the enzyme therapy, these animals were given injections of vitamins A, B, D and E and selenium once every week. Appmpriate doses of each of these vitamins were given to meet the weekly requirements (Fox _e; a_l., 1984). Control animals were given preparations of XL-IP gulonolactone oxidase which had been inactivated. These inactivated preparations contained 0.3% of the activity given to the other animals. Weekly vitamin supplements were also given to the control animals. The growth curves for these animals are presented in Figure 12. Four of the animals started on the therapy with active XL—IP gulonolactone oxidase survived for at least 100 days (parts a-d) and the fifth animal (part e) survived for 68 days. Importantly, all of these animals appeared healthy and showed growth rates approaching those of animals fed a normal diet. In contrast, the five control animals survived an average of only 33.610.78 days (part f) . This is similar to the usual time of survival on an ascorbic acid-deficient diet (23-23 days) (Barnes 93 a_l., 1973; Jones g a_l., 1973), showing that the vitamin supplementation (A, B, D, E and selenium) itself does not protect from scurvy. The fact that the control animals survived slightly longer than usual could be the result of the small amount of gulonolactone oxidase activity that they received. It appeared that supplementation with other vitamins was necessary to ensure I? 71 Figure 12.. Growth curves of nutrient-supplemented guinea pigs on enzyme replacement therapy. Five guinea pigs were placed on an ascorbic acid-deficient diet and treated with XL-IP gulonolactone oxidase on the days indicated by arrows. Animals shown in part A-E were given active preparations. Treatment was terminated after 100 days. The animal in part e died on day 68. Five control animals received inactive XL—IP gulonolactone oxidase and an average of their body weights is shown in part F of the figure. The standard errors of the mean body weights of the control animals ranged from :8 to :20 grams. Body weights were determined daily. 400. 200 600 . 400.. 72 {-_«J-J 1 3L1}... soo_ 200.; . i 25 25 200 . 50 25 50 75 l l}, 600.- 600 . n=1 200 T", DAYS Figure 12, 200.° 25 50 75 ‘00 [i=5 50 ., "T A, 73 that animals fed the ascorbic acid-deficient receive adequate nutrition. The success rate of the gulonolactone oxidase therapy was greatly improved. The cause of death of the guinea pig that did not survive the entire 100 days of the protocol (part e of Figure 12) was most likely a bacterial infection (Enterobacter cloacae) in the lungs and liver. Enterobacter cloacae is a common inhabitant of the guinea pig intesthial tract. Internal examination of this animal showed that the abdominal cavity contained blood, peritonitis and adhe- sions between the liver and gastrointestinal tract. These symptoms were not necessarily an effect of the modified enzyme, since one of the other animals in the study was examined and did not exhibit these signs. The presence of this infection was consistent with a pimcturing of the gastrointestinal tract during intraperitoneal injection of the enzyme. Partial analysis of the vitamin and mineral content of the ascorbic acid-deficient diet was performed also. There were some disparities between the contents of this diet and those of the normal guinea pig chow. For the most part, however, nutrient levels were similar. The diet was considered to be adequate. Nutrient analyses provided by the manufacturers confirmed this belief. However, after 14 days of vitamin C deficiency, guinea pigs become anorexic and enter a chronic fasting state (Peterkofsky gt a_l., 1986). As a result, additional nutritional deficiencies may be imposed. The frequent intraperitoneal enzyme injections may have complicated this situation by causing irritation and lowering the appetite further. It may be that supplementation of additional vitamins to the ascorbic acid-deficient animals prevented more generalized malnutrition. E. Comparison of gulonolactone oxidase therapy regimens A comparison of the XL-IP gulonolactone oxidase therapy protocols described is presented in Figure 13. The survival time of all of the animals in 74 .0 «8a 8 035 3 v0.8a8oo 0.8 n— 88" 5 30828 was 4 «8m 5 39388 0380900.— h0fi ou U08A8ou 0.8 m «8m 8 £08“: .39: v 3 88080 8 up 608085 3 005 #383583 8 A033 083 898m 68828 no 950» A000 you 850:» am 083 893.80 808 0E. $808805 08.380 A000 £33 £088.... :0 on n0>wm 33 683080803» 68.33% 0E. 503080802550 50m 98 .m .D .m .5 £83m.» 53003 05 was 08—88 083039815 akin 0380 603000." D «8A 5 0.0880 0a. .Q anon S 30828 05 now £93.30 no @0200 80830 002E. 8.33035 A0m was m .Q .m )3 £803» £3003 ~80 083xo 0830308286 Elfin 05 no 838.8908 63.23003 5:: @0303 0.0828 $803.30.. 0 «8m .m «8A 8 30828 on v0u0uo~£8v0 0003 083.3 083032315 anti—x no 80300?“ 5003 .m «.8.— 5 $0880 .3“ £93.80 00 @0200 v8 00053890 80 086303 and“ no 88300?“ 3x003 83m 0003 < «8m 3 88:2 .umnoum 3.5800 0330908 #05 28 808300 59805 08380 08300—0888 80835 05 8:.» @0803 £0880 82337300 03.308 no 082 332.5 05 mo 80382—800 .2 0.8th 75 « mxmu mxmo * mxmu mama 3.3.8 598.8 €938». 33.8 mz_h 4<>_>m=m z a a. . .'.4" s a c. , 88 -o‘. .._ .. _ y“ a ._ . . Figure 17 89 TABLE 8 Comparison of Kinetic Parameters of Gulonolactone Oxidase and PEG-Gulonolactone Oxidase Unmodified GLO PEG-GLO Recovery 100 73.9:3.9 Km (mM) 0.089:0.026 0.065:0.011 The apparent Km of the modified and unmodified enzyme was determined from Lineweaver-Burk double reciprocal plots. Recovery is expressed as a percentage of the unmodified enzyme activity at the substrate concentrations used in the usual assay. Km is expressed as a mM concentration. 90 B. Stability of PEG-gulonolactone oxidase 1. Ln zi_tr_9__ stability of PEG-GLO activity The stability of PEG-GLO at 37°C and to trypsin digestion was compared to that of the immodified enzyme in Q 3332 assays. The results are shown in Figure 18. Enzyme activity was greatly stabilized at 37°C by the modification. PEG-GLO retained 50% of its activity for more than 24 hr, whereas the unmodified enzyme lost 50% of its activity within 10 min. Little, if any, protection from trypsin digestion was afforded by conjugation with PEG. The activity of both preparations decayed to less than 50% within 6 min. 2. Circulating half-life of PEG-GLO activity Experiments were continued to examine whether the circulating half-life of enzyme activity was extended by the modification. Conjugation of gulonolactone oxidase with PEG does not prolong the plasma half-life of activity in guinea pigs. The plasma half-er of PEG-GLO activity in !i_v£ was not significantly different (54 min) from that of the unmodified enzyme (50 min) (Figure 19). 3. Q vitro stability of PEG-GLO activity in plasma To determine whether this loss of enzyme activity from the circulation was the result of the PEG-enzyme leaving the vasculature or being inactivated within the circulation, an '3 _v_i_t£g experiment was conducted. PEG- GLO was incubated with fresh guinea pig plasma at 37°C. Enzyme activity was assayed at various times over a 4—hr period and compared to the results obtained from the Q 1122 half-life experiments (Figure 20). For the most part, the decline of enzyme activity in plasma in 11132 paralleled that rapid decline in activity observed in the circulation. These results provide evidence that inactivation within the circulation may account for the rapid loss of PEG-GLO activity, as opposed to it leaving the vasculature. Since proteolytic degradation in the plasma 91 Figure 18. (A) Stability of immodified gulonolactone oxidase to incubation at 37°C and trypsin digestion. Unmodified gulonolactone oxidase (0.2 U activity/m1) was incubated at 37°C alone ( O ) and in the presence of trypsin (0.5 mg/ml) ( O ). Enzyme activity was assayed at various times over a 90-minute period and is expressed as a percent of the initial activity. (B) Increased stability produced by modification with PEG. PEG-gulonolactone oxidase (0.02 U activity/ml) was incubated at 37°C alone ( O ) and in the presence of trypsin (0.5 mg/ml) ( O ) over a 24-hr period. Enzyme activity is expressed as a percent of the initial activity. One unit (U) of activity equals 1 umole of ascorbic acid formed per minute. 92 100- % INITIAL ENZYME ACTIVITY 50" O. o. o O o O o ' . O O o o o . I 9 0 _(3 30 60 90 TIME (minutes) 100 >- ‘. ° i- . ° . E i B E; o < I.” 15 . ' i] 507:; . 2 DJ co 4 as < °' E o .2. X i l L [I r 4 1 1 2 3’6 24 ' TlME(hours) Figure 13 93 Figure 19. Plasma half-life of PEG-gulonolactone oxidase compared to the unmodified enzyme. Doses of unmodified gulonolactone oxidase ranging from 0.16 to 0.36 U of activity were infused intravenously into 4 guinea pigs (H). Another group of 4 animals received i.v. infusions of 0.08 to 0.51 U of PEG—gulonolactone oxidase (00). One unit (U) of activity equals 1 umole of ascorbic acid formed per minute. Blood was sampled at the time points indicated over a period of four hours and enzyme activity was determined in plasma. Activity is plotted as the percentage of the activity in the 10 min sample. 94 I0 3 m ... U . .m w... 2 I E .99... M .. . .. n I .0 ......... a ..... i O 0 0 5 1 >:>_.8< 22>sz ._