Mechanisms of cross-species activity of mammalian interferons

Pharmac. Ther. Vol. 7, pp. 245 252. © Pergamon Press Ltd. Printed in Great Britain.

0163-7258/79/1101-0245/$5.00/0

Specialist Subject Editor: SIDNEYPESTKA

MECHANISMS OF CROSS-SPECIES ACTIVITY OF MAMMALIAN INTERFERONS* WILLIAM A. CARTER Department of Medical Viral Oncology, Roswell Park Memorial Institute and State University of New York at Buffalo, Buffalo, New York 14263, USA

1. INTRODUCTION Interferonf, produced by different cell types, is a glycoprotein that has antiproliferative as well as antiviral and immunomodulatory activities. It is being tested as a therapeutic agent in a variety of human diseases including malignant tumors such as osteogenic sarcoma, breast carcinoma, and lymphomas (Carter et al., 1979a; Carter and Horoszewicz, 1979; Merigan et al., 1978; Strander, 1977; J. Gutterman, unpublished data). These clinical studies are limited by the expense and amount of IF available, which also prevents therapeutic trials in more types of neoplastic disease. Long-term, as well as intensive, IF therapy may also be valuable since its anti-tumor effect is based on different mechanisms than those of conventional chemotherapeutic agents. It is part of the normal cell control mechanisms and can potentiate various host defenses (Tamm and Sehgal, 1979). The limited amount of IF available also prevents therapeutic trials in other chronic disorders, such as connective tissue and neurological diseases, many of which are suspected of being of viral origin. Current procedures for production of IF for use in humans involve human cells because of an apparent species specificity of IF, as originally described by Isaacs and Lindenmann (1957). This severely limits the supplies of IF for clinical trials and only a few laboratories can currently produce human IF in any significant amounts. It would be of considerable scientific, and possibly clinical, importance to have a cheap source of IFs. Animal derived proteins are more readily available, inexpensive, and already widely used in clinical practice. Bioactive preparations from animal sources include insulin, TSH and ACTH. These polypeptide hormones demonstrate a preservation of amino acid sequences crossing species lines with maintenance of activity. For example, comparative studies of insulin, including that derived from a primitive vertebrate, the Atlantic hagfish, indicate preservation of peptide structure in the bioactive region throughout the entire history of vertebrate evolution (De Meyts et al., 1978). IF may have a similar preservation ofa bioactive peptide. Although the concept that IFs are glycoproteins with a restricted, species specific activity is widely accepted, various exceptions have been reported, e.g. the pronounced effect of human leukocyte IF on bovine and porcine cells observed by Gresser et al. (1974). Recently, we found that porcine and bovine carbohydrate-altered IFs obtained from leukocytes have increased efficacy in protecting human cells from viral challenge (Carter et al., 1979b). We also observed that the addition of oligosaccharide chains to the human immune IF polypeptide (derived from lymphocytes) changes many of its physico-chemical properties (Mizrahi et al., 1978). These findings suggest a relationship between function of IF and its glycosylation. Namely, the structural homology and the cross-species activity may be masked by a part or all of the oligosaccharide. This implies that carbohydrate-altered animal IFs should be active in man when there is homology of a bioactive polypeptide. * Supported in part by Department of Health, Education and Welfare Grants CA14801 and CA15502 and the John A. Hartford Foundation, Inc. f The abbreviations used are: IF, interferon; poly(U), polyuridylic acid. 245

of Leukocyte

us Fibroblast-Derived

f Concanavalin A Horseshoe crab Lotus Soybean letting Wheatgerm lectin$

Bovine fibroblast IF Human fibroblast IF

N-acetyl-neuraminic

acid

D-mannose D-galactose L-fucose N-acetyl-galactosamine N-acetyl-glucosamine

80$ 17.5 >50 0 0

>PO >75 >50 >75 >PO 80 >80

Antiviral activity eluted with specific sugar (%)

10%loo0 100-1000 trace* 100 100 100-300

Human (trisomy 21)

-

L

II

..-.

c-z

.I

I-___~

-

.I__..,_

_

L1-u-

_.,

-..Y_lY

* Trace cross-over activity of bovine fibroblast IF may be due to the subcomponent which is in the breakthrough fraction on Con A-agarose chromatography (Miller, Chadha and Carter, unpublished data). t Proteolytic cleavage (Chadha et al., 1978) also modulates the cross-species activity of human leukocyte IF since the full length molecule (26,000 mol wt) is more active on some heterologous cell types, while the nicked molecule (21,OCOMW) has equal activity on homologous and heterologous cells (Chadha and Carter, ‘Human leukocyte interferon preparations: effect of protease activity’, submitted, 1979) (also, M. N. Thang, personal communication, 1979); the 5000 mol wt fragment plays a role in the enhanced cross-species behavior of human leukocyte IF. $ Requires both a MM plus ethylene glycol. g Carbohydrate moieties may be masked by peripheral sugars. The table is based on these references: Carter et al., 1979b ; Janowski et al., 1975 ; Davey et al., 1976 ; Miller, Chadha and Carter (Interferon Scientific Memoranda, 1978); Carter et al., unpublished data. The 5000 mol wt fragment apparently consists of both carbohydrate and amino acid moieties. In contrast, the human leukocyte form minus this fragment (21,000 mol wt by gel filtration or 15,00&17,500 by SDS-PAGE) has no appreciable glycosidic residues as determined by periodic acid treatment (Rubinstein et al., 1979).

Concanavalin A Castor bean lectin Lotus Soybean lectin Wheatgerm lectin

Sugar specificity

Antiviral activity in breakthrough (%)

on immobilized l&ins

100 N.D. trace N.D. N.D. 100

200-300 100 100 100 trace 200-300

100 100 trace 200-1OOct trace 200-300 Chromatography

Horse (dermal)

Lectin-agarose

Interferons

% Interferon activity (homologous cell activity defined as 100%)

and Pectin Binding Properties

Bovine (embryonic kidney)

Activity

Porcine (kidney)

Antiviral

Porcine leukocyte IF Bovine leukocyte IF Human leukocyte IF

Preparation

Porcine leukocyte IF Bovine leukocyte IF Bovine fibroblast IF Human leukocyte IF Human fibroblast IF Equine leukocyte IF

Preparation

TABLE 1. Cross-species

Carbohydrate-altered interferons

247

2. ANIMAL LEUKOCYTE INTERFERON AND BIOACTIVITY IN H U M A N CELLS The antiviral activity of porcine, bovine and equine leukocyte IF is several fold higher when measured on human cells than on their homologous cells (Carter et al., 1979b) (Table 1). By contrast, IF derived from animal fibroblast cultures does not display cross-species bioactivity; these IFs bind readily to lectin columns such as concanavalin A (Davey et al., 1976). Similarly, human fibroblast IF is extensively glycosylated and D-mannose, L-fucose and sialic acid residues, etc. have been identified (Jankowski et al., 1975). It often is at least 1000-fold less active on various heterologous cells than human leukocyte IF. The lack of retention of leukocyte IF preparations (human and non-human) on various iectin columns seems to suggest a different oligosaccharide composition (Carter et al., 1979b; Jankowski et al., 1975) (Table 1). A specificity for binding of other serum glycoproteins to membrane receptors has been previously reported; for example, asialo derivatives of orosomucoid, ceruioplasmin and transferrin are bound to a hepatic membrane receptor while the sialic acid-terminated forms do not bind (Hudgin et al., 1974). Also, across species lines, membrane receptors for glycoproteins are known to differ in terms ofwhich terminal sugar residue is recognized and specifically bound (Kawasaki and Ashwell, 1977). Even though sialic acid is present on human fibroblast IF and missing on leukocyte IF (Morser et al., 1978), the presence of peripheral sugars on fibroblast IF does not fully account for the species specificity; human fibroblast IF, synthesized in the presence of 2-deoxyglucose, is still species specific (W. A. Carter, unpublished data). 2-deoxyglucose, an inhibitor of glycosylation which appears to block synthesis of peripheral sugars (Schmidt et al., 1976) may leave part of the oligosaccharide core intact. Therefore, the core sugars (e.g. mannose and Nacetyiglucosamine) of fibroblast IFs may be the more important residues in masking their cross-species activity.

3. GLYCOSYLATION ALTERS THE MOLECULAR SHAPE OF INTERFERONS All IF molecules (human or non-human) which bind to concanavalin A display strongly hydrophobic properties. This degree of hydrophobicity is unexpected and is consistent with an apparently increased density of aromatic amino acids (e.g. tyrosine) on the surface of the molecule (Sulkowski et al., 1976). It is usually believed that glycosylation increases molecular surface hydrophilic properties although this depends to a great extent on the nature of the sugar residues; the presence of ionically charged sialic acid is more hydrophilic than that of mannose. If binding to concanavalin A indicates increased glycosylation of most IFs not derived from leukocytes, then the nature of sugar residues may be in part responsible for the decreased hydrophilic nature. Glycosylation also seems to alter the tertiary structure of fibroblast IFs so as to displace the hydrophobic amino acid residues from the interior to the exterior of the molecule. Both factors could operate to make the non-leukocyte derived IFs hydrophobic. In effect, having a decrease ofhydrophilic sugars promotes a change in tertiary structure so that more hydrophobic amino acids are closer to the surface of the molecule. These changes are visualized schematically in Fig. 1. Tunicamycin, an antibiotic which specifically inhibits the glycosylation of newly synthesized polypeptides (Takatsuki et al., 1977)~ has been used to demonstrate that glycosylation alters the physical shape of some immunoglobulins, e.g. IgA and IgE (Hickman et al., 1977). We used it to explore the possible relationship between glycosylation of IF and increased molecular surface hydrophobicity. Human immune IF (produced by exposure of lymphocytes to various mitogens) exists in several molecular forms (O'Malley and Carter, 1978) which are apparently due to variations in the degree of glycosylation. These forms possess a surface hydrophobic center (Mizrahi et al., 1978) named the polynucleotide binding site which can be detected by the binding of immune IF to various immobilized polynucleotides, e.g. poly(U). However, when immune IF is produced in the presence of

248

WILLIAM A. CARTER

Portiolly Glycosyloted Interferon

or

(unmodified)

A

Fully Glycosyloted Interferon

I modified 1

I A

I

CHOn

FIG. I. Schematic diagram suggesting the effects of extensive glycosylation on shape of bioactive region of IF. The number of growth points for oligosaccharidechains are estimatedfrom the apparent

mass of carbohydratein human fibroblast IF (Havell et al., 1977). A, asparagine residues; CHO, carbohydratemoietieswith n indicatingrelativenumber;--O, hydrophobicamino acid,as tyrosine or phenylalanine.The bioactivesiteis shownas a stippledarea containinga hydrophobicaminoacid and, presumably,a nearby asparagineresidue. Leukocyte-derivedIFs represent a class of partially glycosylatedIFs, while the fibroblast-derivedIFs are usually fully glycosylated.Minor subcomportents of leukocyte IF may be fully glycosylated (Chadha et al., 1978) and, similarly, a subcomponentof fibroblastIF may be partiallyglycosylated(Daveyet al., 1976; Miller,Chadha and Carter, unpublisheddata). tunicamycin, only one form is synthesized (Mizrahi et al., 1978); this form appears similar to the leukocyte IF; namely, it has no surface hydrophobic groups, polynucleotide binding does not occur and it no longer binds to lectins. If the IF formed in the presence of tunicamycin is similar to human leukocyte IF, this carbohydrate-depleted form should also display protection of non-human cells similar to leukocyte-derived IF as reported by Gresser et ai. (1974) and Carter et al. (1979b). Although glycosylation is of no apparent consequence for secretion of IF (Mizrahi et al., 1978) or some other glycoproteins (Tucker and Pestka, 1977 ; Hickman and Kornfeld, 1978), its impairment does result in changes in various physico-chemical properties of IF, as suggested schematically in Fig. 1. The increased hydrophobicity which occurs after glycosylation of IF may be due in part to surface exposure of hydrophobic polypeptides located within the bioactive site required for both recognition of IF by the cell receptor and for its action. The bioactive site therefore would become distorted (note distortion of hypothetical active site region in Fig. 1). Such hydrophobic peptide residues are often part of the bioactive site of hormones (e.g. insulin, Pullen et al., 1976); they seem to provide a major driving force for the interaction of a protein with its cell surface receptor. Distortion of these residues c.ould alter the binding properties of IF and its induction of synthesis of effector proteins, all of which probably depend on tertiary structure. An alteration in interaction with surface receptor would explain why the species-specificity is not observed at the subcellular level. For example, in cell-free systems, Ball and White (1978) have observed that the IF-induced inhibitors are functionally interchangeable. To prove that the specificity must lie in the cell surface receptor, I propose that the effector proteins induced by carbohydrate-deficient heterologous IFs be compared in a cell-free system; if a mixing experiment shows identity of the IF-induced inhibitors, this would be strong evidence in favor of using heterologous IFs to promote the exact biological effects obtained with homologous IFs.

Carbohydrate-altered interferons

249

An evolutionary constraint on the structure of cell membrane receptor proteins of many polypeptide hormones has been proposed (Blundell, 1976). This has been used to explain why the 'best fit' of a protein for its receptor does not always occur even in the same species. Examples of enhanced bioactivity of proteins in heterologous cells already exist in clinical medicine; for example, salmon calcitonin has high activity in man when compared with human calcitonin. Hormones (e.g. glucagon and insulin) may even evoke a biological response in prokaryotic cells. Therefore, it is not such an unusual finding that human leukocyte IF is more active on bovine than human cells and porcine leukocyte IF more active on human than porcine cells. The activity of porcine leukocyte IF is greatly enhanced in human cells trisomic for chromosome 21 (Down's syndrome) (Carter et al., 1979b), the chromosome which encodes for the IF membrane receptor, a point that further illustrates the critical function of the surface receptor in the expression of bioactivity. Glycosylation may or may not affect the biological activity of IF directly, but does seem to alter the 'fit' in the interaction between IF and receptor on heterologous cells. On the other hand, more extensive glycosylation of fibroblast IF may, in fact, enhance activity in homologous cells since it has been demonstrated that purified fibroblast IF has high specific activity (Berthold et al., 1978) near to the value which is theoretically attainable (Joklik, 1977). When synthesized in the presence of inhibitors of glycosylation, Havell et al. (1977) demonstrated clearly that the moi wt of human fibroblast IF decreases by about 5000, indicating that about 25 per cent of its mol wt is due to oligosaccharide units. Earlier studies with insulin have shown that even small modifications of the bioactive site, either by adding various groups or by inducing configurational changes, can dramatically alter the protein affinity for receptor (De Meyts et al., 1978; Pullen et al., 1976). Although not yet known, it should be possible to distinguish whether fully glycosylated fibroblast IFs fail to bind to receptor on heterologous cells or occupy the receptor but fail to activate the cell.

4. BYPASSING THE 'SPECIES-BARRIER' BY CHEMICAL MODIFICATION OF IF During synthesis of the leukocyte IF glycopeptide, certain alterations in its glycosylation must occur which preserve cross-species activity of the polypeptide. Biosynthesis of glycoproteins involves the en bloc transfer of a large oligosaccharide chain (containing Nacetylglucosamine, mannose and glucose) to the asparagine residue of a polypeptide (Robbins et al., 1977). Typically, the chain is then processed by removal of several residues, followed in the Golgi by sequential enzymatic addition of the peripheral sugars (Nacetylglucosamine, galactose and sialic acid) to form the final oligosaccharide. With leukocyte IF, an enzymatic defect in processing of several residues would result in a glycoprotein with an altered oligosaccharide core; a difference in its peripheral sugars has already been reported (Jankowski et al., 1975; Morser et al., 1978). Enzymatic defects in the processing step as well as in addition of peripheral sugars have been recently suggested in clones of Chinese hamster ovary cells (Li and Kornfeld, 1978); such cells have an altered binding to lectins similar to that demonstrated by leukocyte IFs derived from either human or non-human sources. Part of the carbohydrate chain is probably added during the synthesis of the polypeptide (Kornfeld and Kornfeld, 1976; Sefton, 1977) and this may considerably magnify its overall effect on the shape of the molecule since the glycoprotein (e.g. IF) has not yet assumed a more rigid structure as when its synthesis is complete. Still, this might be unimportant unless changes in glycosylation also altered the active site. The majority of buried hydrophobic groups which are externalized probably do not lie in proximity of the bioactive region (see Fig. 1 and hydrophobic residues removed from active site region). However, the rearrangement of a single hydrophobic group within the bioactive site (Fig. 1, stippled area) should alter the activity; tyrosine or phenylalanine are the likely residues.

250

WILLIAM A. CARTER

Stepwise stripping of sugar residues from fibroblast IFs with specific glycosidases (Koide et al., 1977; Kornfeld and Kornfeld, 1976) could alter configuration and also modulate

binding with receptors on heteroiogous cells; activation of these cells might then occur. Specifically, the use of either mannosidases or endoglycosidases which attack the core region of the IF oligosaccharide may resurrect activity on heterologous cells. In addition, stripping of sugar moieties from fibroblast IFs could increase their recognition by antibody raised against leukocyte IFs, thus allowing a better index of their polypeptide homologies. It is well recognized in immunochemistry that sugar residues are strong antigenic determinants and thereby dampen the potential antigenic contribution of the polypeptide backbone.

5. DISCUSSION In conclusion, efficacious forms of animal IF could be produced from leukocytes obtained at the slaughter house. One obvious major advantage is the availability of sufficient leukocytes to provide for many long-term clinical studies of the antiproliferative activity of IF. Estimates based on the present annual slaughter of pigs, in the U.S. alone, indicate that even this one source would greatly augment the currently available supplies of IF from various human cell sources. This would also permit more extensive pharmacokinetic studies of IF using in vivo primate models which allow easier measurements of clearance, tissue concentrations, etc. Clinical use of animal leukocyte IFs should be safe, especially due to the lack of occult human viruses (intact or as genetic fragments). It has been suggested that current therapy with human leukocyte IF may carry some hazard by possible transfer of slow growing viruses (Wadell, 1977). Viral infections in man trigger the production of different antigenic forms of IF (Mogensen and Cantell, 1977; O'Malley and Carter, 1978). The immunoheterogeneity may arise from variation in the glycosylation patterns of human IF and it is likely that both the polypeptide and its carbohydrate component contribute to the immunologic specificity. These human IF forms are not recognized by the immune system as foreign, i.e. there exists cross tolerance within the human species for all the human IF forms. Small differences in the primary structure, which differentiate IF from one species to another, could also escape the immune system, thereby extending human immunological tolerance to carbohydrate-depleted IFs from non-human sources. IFs are already known to be poorly antigenic. The identification of these small differences in primary structure, which presumably differentiate IF from one species to another, will require as a first step their purification to homogeneity, such as has been recently achieved conclusively in Pestka's laboratory with human leukocyte IF (Rubinstein et al., 1978, 1979). These authors emphasize the importance of further study to establish the relationship of this particular form of pure human leukocyte IF to the primary product of the IF gene. In contrast, other workers in the IF field generally may have overlooked some of the lessons taught earlier in protein chemistry, wherein variations in post translational modification were originally misinterpreted as differences in primary gene products. As one example, the early estimates of large numbers of individual histone proteins (as in the hundreds) vanished with delineation of biochemical steps in post translation control; the rapid isolation of individual histone messenger RNAs (Childs et al., 1979) now allows a clear understanding of each primary gene product. Application of purification techniques, such as those described for human leukocyte IF (Rubinstein et al., 1978 ; 1979), to non-humanleukocyte IFs, should allow various assessments of homology by tryptic peptide mapping, etc. in the near future. Such studies would add significantly to our scientific understanding as well as provide more insight into the possible value of heterologous IFs in clinical medicine. Further studies will also be required concerning possible mechanisms which may interrelate further the 'species barrier', glycosylation, and the specific biological activity of mammalian IFs on their homologous cells. For example, Pestka et al. (1975) have demonstrated that mRNA from induced human fibroblasts can direct the synthesis, in a cellfree extract from mouse cells, of an IF with the antigenic qualities and species-specificity of

Carbohydrate-altered interferons

251

human fibroblast IF. Later work by Tucker and Pestka (1977) showed that the Ehrlich ascites cell-free system, as it had been used for IF synthesis, did not glycosylate fully MOPC46B immunoglobulin light chain protein synthesized de novo. However, as pointed out by these authors, because of the minute amounts of IF activity detected, it is possible that a degree of glycosylation occurred in such extracts and that this hypothetical product, glycosylated IF, accounted for the antiviral activity. In any case, biochemical characterization of the active cell-free product will serve to answer these questions as well as to contribute to our growing knowledge on the role of glycosylation and the 'species-barrier'. Acknowledgements--I am indebted to Drs. M. M. Weiser, A. O. Vladutiu, J. S. Horoszewicz and S. Pestka for their

thoughtful comments.

REFERENCES BALL, L. A. and WHITE,C. N. (1978) Oligonucleotide inhibitor of protein synthesis made in extracts of interferontreated chick embryo cells: comparison with the mouse low molecular weight inhibitor. Proc. natn Acad. Sci. U.S.A. 75: 1167- ll71. BERTHOLD, W., TAN, C. and TAN, Y. H. (1978) Purification and in vitro labeling of interferon from a human fibroblastoid cell line. J. biol. Chem. 252: 5206-5212. BLUNOELL,T. (1976) Hormone and antihormone action at the target cell. Nature, Lond. 260: 288. CARTER,W. A. and HOROSZEWlCZ,J. S. (1979) Production, purification and clinical application of human fibroblast interferon. Pharmac. Ther. A, (in press). CARTER, W. A., LEONG, S. S. and HOROSZEWlCZ,J. S. (1979a) Human fibroblast interferon in the control of neoplasia, ln: Antiviral Mechanisms and the Control ofNeoplasia. Chandra, P. (ed.)Plenum Press, New York (in press). CARTER,W. A., DAVIS,L. R., JR., CHADHA,K. C. and JOHNSON,F. H., JR. (1979b) Porcine leukocyte interferon and antiviral activity in human cells. Molec. Pharm. (in press). CHADHA, K. C., SCLAIR,i . , SULKOWSKI,E. and CARTER,W. A. (1978) Molecular size heterogeneity of human leukocyte interferon. Biochemistry 17: 196-200. CHILDS,G., LEVY,S. and KEDES,L. H. (1979) Rapid purification of biologically active individual histone messenger RNAs by hybridization to cloned DNA linked to cellulose. Biochemistry 18: 208-213. DAVEY,M. W., SULKOWSKI,E. and CARTER,W. A. (1976) Binding of human fibroblast interferon to concanavalin A-agarose. Involvement of carbohydrate recognition and hydrophobic interaction. Biochemistry 15:704-713. DE MEYTS,P., VANOBBERGHEN,E., ROTH,J., WOLLMER,A. and BRANDENBURG,D. (1978) Mapping of the residues responsible for the negative cooperativity of the receptor-binding region of insulin. Nature, Lond. 273: 504-509. GRESSER, I., BANDU,M. T., BROUTY-BOYf/,D. and TOVEY, M. (1974) Pronounced antiviral activity of human interferon on bovine and porcine cells. Nature, Lond. 251: 543-545. HA'CELL,E. A., YAMAZAKI,S. and VIL~EK,J. (1977) Altered molecular species of human interferon produced in the presence of inhibitors of glycosylation. J. biol. Chem. 252: 4425-4427. HICKMAN,S., KULCZYCKI,A., JR., LYNCH,R. G. and KORNFELD,S. (1977) Studies of the mechanism oftunicamycin inhibition of IgA and lgE secretion by plasma cells. J. biol. Chem. 252: 4402-4408. HICKMAN, S. and KORNFELD, S. (1978) Effect of tunicamycin on IgM, IgA, and IgG secretion by mouse plasmacytoma cells. J. lmmunol. 121: 990-996. HUDGIN, R. L., PRICER,W. E., JR., ASnWELL,G., STOCKERT,R, J. and MORELL,A. G. (1974) The isolation and properties of a rabbit liver binding protein specific for asialoglycoproteins. J. biol. Chem. 249: 5536-5543. ISAACS,A. and LINDENMANN,J. (1957) Virus interference. I. The interferon. Proc. Roy. Soc. (Ser. B) 147: 258-267. JANKOWSKI,W. J., DAVEY,M. W., O'MALLEY,J. A., SULKOWSKI,E. and CARTER,W. A. (1975) Molecular structure of human fibroblast and leukocyte interferons: probe by lectin and hydrophobic chromatography. J. Virol. 16: 1124-1130. JOKLIK,W. K. (1977) 3rd Conference on Antiviral Substances. Ann. N. Y. Acad. Sci. 284: 718. KAWASAKI,T. and ASHWELL,G. (1977) Isolation and characterization of an avian hepatic binding protein specific for N-acetylglucosamine-terminated glycoproteins. J. biol. Chem. 252: 6536-6543. KOIDE, N., NOSE, M. and MURAMATSU,T. (1977) Recognition of IgG by Fc receptor and complement: effects of glycosidase digestion. Biochem. biophys. Res. Commun. 75: 838-844. KORNFELD, R. and KORNFELD,S. (1976) Comparative aspects of glycoprotein structure. Ann. Rev, Biochem. 45: 217-237. LI, E. and KORNTELD,S. (1978) Structure of the altered oligosaccharide present in glycoproteins from a clone of Chinese hamster ovary cells deficient in N-acetylglucosaminyltransferase activity. J. biol. Chem. 253: 6426-6431. MERIGAN,T. C., SIKORA,K., BREEDEN,J. H., LEVY,R. and ROSENBERG,S. A. (1978) Preliminary observations on the effect of human leukocyte interferon in non-Hodgkin's lymphoma. N. Engl. J. Ned. 299:1449-1453. MIZRAHI, A., O'MALLEY, J. A., CARTER, W. A., TAKATSUKI,A., TAMURA, G. and SULKOWSKI,E. (1978) Glycosylation of interferons: effects of tunicamycin on human immune interferon. J. biol. Chem. 253: 7612-7615. MOGENSEN,K. C. and CANTELL,K. (1977) Production and preparation of human leukocyte interferon. Pharmac. Ther. l: 369-381. MORSER,J., KABAYO,J. P. and HUTCHINSON,D. W. 0978) Differences in sialic acid content of human interferons, d. gen. Virol. 41: 175-178. JPT Vol. 7, No. 2 - - D

252

WILLIAMA. CARTER

O'MALLEY, J. A. and CARTER, W. A. (1978) Human interferons: characterization of the major molecular components. J. Reticuloendothel. Soc. 23: 299-305. PESTKA,S., Mclm,ms, J., HAVELL,E. A. and VIL~EK,J. (1975) Cell-free synthesis of human interferon. Proc. ham Acad. Sci. U.S.A. 72: 3898-3901. PULLEN, R. A., LINDSAY, n. G., WOOD, S. P., TICKLE, I. J., BLUNDELL,T. L., WOLLMER, A., KRAIL, G., BRANDENBURG,D., ZAHN, H., GLIEMANN,J. and GAMMELTOFT,S. (1976) Receptor-binding region of insulin. Nature, Lond. 259: 369-373. ROBBINS,P, W., HUBBARD,S. C., TURCO, S. J. and WIRTH, D. F. (1977) Proposal for a common oligosaceharide intermediate in the synthesis of membrane glycoproteins. Cell 12: 893-900. RUBINSTEIN,M., RUBINSTEIN,S., FAMILLETTI,P. C., GROSS,M. S., MILLER,R. S., WALDMAN,A. A. and PESTKA,S. (1978) Human leukocyte interferon purified to homogeneity. Science 202: 1289-1290. RUBINSTEIN,M., RUmNSTEIN,S., FAMILLETTI,P. C., MILLER,R. S., WALDMAN,A. A. and PESTKA,S. (1979) Human leukocyte interferon: Production, purification to homogeneity, and initial characterization. Proc. natn Acad. Sci. U.S.A. 76: 640-644. SCHMIDT,M. F. G., SCVlWARZ,R. T. and SCHOLTISSEK,C. (1976)Interference of nucleoside diphosphate derivatives of 2-deoxy-D-glucose with glycosylation of virus-specific glycoproteins in vivo. Eur. J. Biochem. 70: 55-62. SEF'rON,B. M. (1977) Immediate glycosylation of sindbis virus membrane. Cell 10: 659-668. STRANDER,H. (1977) Interferons: anti-neoplastic drugs? Blut 35: 277-288. SULKOWSKI,E., OAVEY, m. W. and CARTER,W. A. (1976) Interaction of human interferons with immobilized hydrophobic amino acids and dipeptides. J. biol. Chem. 251: 5381-5385. TAKATSUKI,A., KAWAMURA,K., OKINA, M., KODAMA,Y., ITO, T. and TAMURA,G. (1977) The structure of tunicamycin. Agr. biol. Chem. 41: 2307-2309. TAMM, I. and SEHGAL,P. B. (1979) Intefferons. Am. J. Med. 66: 3-5. TUCKER,P. and PESTKA,S. (1977) De novo synthesis and glycosylation of the MOPC-46B mouse immunoglobulin light chain in cell-free extracts. J. biol. Chem. 252: 4474-4486. WADELL,G. (1977) Hazards of human leukocyte interferon. N. Enyl. J. Med. 296:1295.