JIH' IIIIK‘J III | ‘ ff!" “night; ~ ' ' ' ' ' . Iii}:- ILI‘mlI‘mflth bl-lei' ‘ ; . I ’Illiil'ltl?‘”9"l ”let". ll“! I . ' “.J‘-—:4_L..;In;,o lr'fi‘c"3""" '~ 1I? il‘ 'zleli ' " ' ‘z!‘ Irifirle‘V'I‘I‘I 3;. ‘ '2; 4%; z: ‘3‘» : :‘I’r'IN‘ I I” r I .- I. JllfiéfillllIlel'IIIgél Ll l ill I l" ‘4: :‘I‘lal’Il xvi/q -|:' 'l' ' 4.5;; r. “I'm: fiéf‘htié - I; “will; . ..444.-.44.I ...I 4.44 ‘II l4 l} d “n? It“... 144.6‘4‘; ‘.,I .; r. I " 'O‘J‘“"T.nl‘ “[1 I ' 432‘ .I ‘ .. . :f; I I“ .4331}. )4? I“: ‘-‘I :l ”I” Il.l|l1!"';lflf;‘- Ilsa; . I‘ " . I’llfilll" ‘ I'fi'll l‘: l’ ~ III; @314 1- ’I llI Illflll‘l‘l‘l“ 4—224 bl “filial nl ‘.~ ”‘28”? $3114 7;;2’137";;;'I‘4II‘3II *‘lllltlllllm Ill l" ’3'": E ‘."“I".h‘ll'sl’ ' l 2] I. Z L g .41} but; Ii III? 4“ "'11“ it €54 4 -‘~ _. .g‘I'l .r.4 .'.'-lii"">..:l ~ 424: IIIIIII‘I‘I . - I" is 14‘494544454... 4 Z '1'" ) ‘::1'."‘F:’.| ' ‘ ‘ 5; . ‘mllqlllII‘fillII.‘lIlleI-“l‘II: 4"“ . ..; , . l I44 ;‘ | _- : ; E. ‘ 14‘ Illa '9 Alli'hfii "1:" “7 ' ' l‘. '3 '.‘:."l‘. I all.“ ’ . !' 3%! a':.~". ‘h; ‘ ' I! :,.z , -“' \ I‘ . ;l If ' I.'-4 I: 4 I ,, 5 I: III-I - I '53- ' . ;J.“I l'.-' i ' h f l! . ' I'll . ' ghfill‘ll l :l . {Ill-:1 l“ 1‘" il' I ' ' . - l "4‘ IF. ' ' . 1| - '_." . v .' . ~Ifim ~ . . - A... L- - a—n. , .4 . - m .4 . - ...... . ‘ ' ' - ' .. 4 ..4 .- W‘- ". '.4 '. 4 a . .o— . . 4.. .3 .. 4 « —. . u'..« n. ."W l'I‘l "-1 .4 ‘ W‘ r.:-' ‘l I lllIIIIIIllII‘.. ~42 1; . IIIIIIIIIII 3 5 ~.l’.‘.:‘.;" ~ III-I. ,IIII .‘II ’ 3‘ 455-4“: 3 51.. :3" I"? . l‘v‘l‘ ‘ 'II..IIl;;III4;I~:III,~III-III IILIIlIlév‘ ' ' ‘rsQII'r’II‘;:II . 45' ' {5 I. 'glélif’lIi‘ :IIIIIIIIIIIII III???“ KI ‘J glenl .‘1‘ II‘.’ 4 42'1”;le r; [Iiluz‘ ll le WI In“ 11-:5""l!'il“?‘ 3.4;. I ‘434.1 ,."‘IE" ll will?! M, W‘l ‘ll5I‘é.l:l‘-lllfi"“-l‘iu¢l 1“ {‘3‘ ”Lu 3 gull“ . yII4.1'lI“i'lfl?,Ilrzzl Il’lIL'I‘I; g 4 "'JII‘?‘ 9 )l, ”‘J l 91'” I “11”“ z! I;:u "lkhl ”“3 11?. _,I‘ Ifév'L‘ IIIIIJIII'ISIL imp: _ U ql Kl: 7‘3" fizfllx'l'" ‘ .111 fill ‘ fir": I'liH:l- 3.: k§t~1I$£.rrll“ “:1, II - .7244‘15-3‘ 4:, If ‘ '4’” ' 417.". ‘ I" ' 7 N ‘Ir, . I" “if: 13:2 4: I "_ 4': 4 r , . .'| .. c.‘ . 'l Moi; 4 4 '0' Lio‘flllzn‘r} 5i. , 3" 545 If: It 3:4. l;’lll;'I:‘ll. I“ II'.I IIIl L131; 414.; ' :' Irv-iii] Illl! 'lld.“ llIilI vIl‘ ll.| l‘IllMIllnlli -W 4‘. —4.-. 4... II“ I” ‘.'II ll I I? H.l' illIE: I’lqll 'llli ,l’l' I lMI° III I‘lv" III“ l... II; lIIIlHaIlHll U l’Ilmlln‘ 1:4‘ I llll II I.l;‘ llllf‘ll .riESlS' This is to certify that the dissertation entitled Isolation and Characterization of Lectins from Mammalian Fibroblasts presented by Calvin F. Roff has been accepted towards fulfillment of the requirements for Ph.D. degree in Biochemistry 1. Major professor Date 6/ l 4/ 83 MS U is an Affirmative Action/Equal Opportunity Institution 2 T!” E? “W ‘2’? A '-.'."’£ ‘- r. ‘ bun‘nfiatflm ' 9‘ $.2qu a £33:- . Ezlrfifiestg.,.-;tfiur trggtg ! f" ""3 What. 1%; uni 'i‘fiéa‘fi-gay \____._______‘_ _ _- / M 012771 MSU LIBRARIES n. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. m V ,awerI, a 1. [a 4‘9 ' ! - .34 b; ‘1. d v ' ' ' . “A . . “.1 l . » , . 't ‘ 1‘3“ ‘fiz. V‘" . I: f L ‘2’“ 3“} Mi _ t .- a 3'1 “W“ “6. ,,, .-.-l - ~ea ”“1 :vl‘?‘-". "' ISOLATION AND CHARACTERIZATION OF LECTINS FROM MAMMALIAN FIBROBLASTS By Calvin Freeman Roff A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1983 ABSTRACT ISOLATION AND CHARACTERIZATION OF LECTINS FROM MAMMALIAN FIBROBLASTS By Calvin Freeman Roff Three distinct carbohydrate-binding proteins (CBPs) were purified by affinity chromatography on asialofetuin-Sepharose from 3T3 fibro— blasts: CBP35 (Mr = 35,000); CBPl6 (Mr = 16,000); and CBP13.5 (Mr = l3,500). These CBPs were similar in several key properties: (a) they showed agglutination activity when assayed with rabbit erythrocytes; (b) they all appear to specifically recognize galactose- containing glycoconjugates; (c) they have low isoelectric points, pIs 4.5-4.7; (d) their binding activities are rapidly lost in the absence of p-mercaptoethanol; (e) the CBPs do not interact with each other and the fractionated proteins can bind to asialofetuin independent of associated polypeptides; and (f) none of the proteins tend to self- associate to form oligomers of identical subunits. CBP16 and CBPl3.5 may be the murine counterparts of lactose-specific lectins previously identified in electric eel and in several bovine and avian tissues. In contrast, it appears that CBP35 represents a newly identified protein capable of binding to galactose-containing carbohydrates. Chemically defined carbohydrates were covalently coupled to polyacrylamide beads. These saccharide-containing beads were used to demonstrate the carbohydrate—binding capacity of CBP35, CBP16, and CBP13.5, isolated from cultured 3T3 mouse fibroblasts on the basis of their binding to asialofetuin-Sepharose. All three proteins bound to polyacrylamide beads containing the disaccharide DGalB(l-—~4)BDGlcNAc but not to beads containing the monosaccharide BDGal. We have also purified, in a single step, a carbohydrate-binding protein from ex- tracts of human foreskin fibroblasts using an affinity column of polyacrylamide beads derivatized with DGal8(l——+4)BDGlcNAc. Immunologically cross-reactive proteins of the same molecular weight (Mr = 35,000) were found in lung, thymus, spleen and arteries. Fractionation of extracts of mouse lung on affinty columns of asialo- fetuin-Sepharose yielded c3935 (lung). The binding of ‘251-iabe1- ed CBP35 (lung) to rabbit erythrocytes was quantitated in the presence of various carbohydrates. It was found that only saccharides contain- ing galactose were inhibitors of the binding of CBP35 (lung) to erythrocytes; the disaccharide lactose was TOO-fold more potent as an inhibitor than the monosaccharide galactose. TO MY WIFE AND CHILDREN WHO HAVE MADE MANY SACRIFICES AND HAVE GIVEN THEIR SUPPORT THROUGHOUT THESE YEARS ACKNOWLEDGEMENTS To Dr. John L. Wang, I would like to express my sincere apprecia- tion for his advice and encouragement during my entire graduate training. I would also like to thank him for an excellent education. I would like to express my gratitude to Sarah Crittenden and Paul Rosevear for their collaborative efforts that aided in the progress of this work. I would also like to thank Tomas Metcalf, III for his help and advice in many matters. TABLE OF CONTENTS page LIST OF TABLES ......................... vii LIST OF FIGURES ......................... viii ABBREVIATIONS .......................... x INTRODUCTION . . . . . . ............. . . . . . . . 1 CHAPTER I LITERATURE REVIEW Carbohydrates as recognition markers . . . . . ..... 3 Carbohydrate binding proteins .............. 4 Protein transport: endocytosis-asialoglyc0protein receptor 5 Protein transport: intracellular sorting-mannose 6-phosphate receptor ..... . . . . . . ..... . . . . . . . Organization of domains-ligatin . . . . . . . . . . . . . l7 Organization of domains-low molecular weight galactose specific lectins . . . . . . . . . . ......... 18 Cell aggregation: slime molds-discoidin . . . ...... 22 Cell aggregation: spenn egg interactions-bindin . . . . . 25 References 0 O O O O O O O O O O O O O O ..... O O 0 O O 26 CHAPTER II ENDOGENOUS LECTINS FROM CULTURED CELLS I. ISOLATION AND CHARACTERIZATION OF CARBOHYDRATE-BINDING PROTEINS FROM 3T3 FIBROBLASTS O O O O O O O O O O O O O O O O O 32 Summary . . . . . .......... ' ............ 33 IntrMUCtion O O O O O O O O O O O O O O O O O O O O O O O O 34 Experimental Procedures Materials . . . . . . . . . . . . . . . . . . . . . . . . 35 Culture and radiolabeling of 3T3 cells . . . . . . . . . 35 Preparation of asialofetuin-Sepharose . . . . . . . . . . 35 Isolation of asialofetuin binding proteins . . . . . . . 36 Gel electrophoretic characterization of CBPs . . . . . . 37 Assays of agglutination and enzymatic activities . . . . 38 Preparation of antisera and immunoprecipitation . . . . . 39 iv Results Asialofetuin-binding proteins from 3T3 cells . . . . . . Molecular weights and isoelectric points of asialofetuin- binding proteins . . . . . . . . . . . . . . . . . . Effect of saccharide ligands and EDTA on asialofetuin- binding proteins . . . . . . . . . . . . . . . . . Fractionation of the carbohydrate- binding proteins . . . Intrinsic binding properties of the carbohydrate binding proteins . . . . . . . . . . . . . . . . . . . . . . . Immunoprecipitation of the carbohydrate binding proteins Discussion . . . . ...... . . ............. Raferences O O O O O O O O O I O 00000000 O O O O O 0 CHAPTER III ENDOGENOUS LECTINS FROM CULTURED CELLS II. SPECIFIC AFFINITY COLUMNS FOR THE ISOLATION OF CARBOHYDRATE- BINDING PROTEINS . . . . . . . . . . . . . ..... Summary . . ..... . ...... . ........... . IntrOdUCtion 0 O O O O O O O O O O O O O O O O O O O O O O 0 Results Quantitation of coupling reactions . . . . . . . . . . . Binding properties of affinity columns containing CHO-PA. Fractionation of mouse and human fibroblast components using CHO-containing affinity columns . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . ..... . ...... . ........ Materials and Methods Materials . . . . . . . . . . . . . . . . . . . . . . . . Synthesis of carbohydrate ligands containing hexanolamine Preparation of carbohydrate affinity resins . . . . . . . Bio-Gel P-lSO hydrazide . . . . . . . . . . . . . . . . . Bio-Gel P-l50 acyl azide . . . . . . . . . . . . . . . . Quantitation of 2-aminoethanol covalently coupled to Bio- Gel P-l50 . . . . . . . . . . . . . . . . . . . . . . Coupling of GlcNAc-HA to Bio-Gel P-l50 acyl azide . . . Coupling of Gal-HA to Bio-Gel P-l50 acyl azide . . . . Coupling of Gal-GlcNAc-HA to Bio-Gel P-l50 acyl azide . Coupling of 6-aminohexanol (HA) to Bio-Gel P-lSO acyl aZide O O O O O O O I O O O O O O O I O O O O O O O O 0 Isolation of CBPs from mouse and human fibroblasts . . . Res "1 ts C O C O O O O O O O O O I O O O O O O O O O O O O O O 41 46 49 57 70 74 78 81 82 84 86 89 92 101 105 107 107 108 108 108 109 109 110 110 110 111 112 CHAPTER IV ISOLATION AND BINDING PROPERTIES OF A LECTIN FROM MOUSE LUNG O O C O O O O O O O O O O O O O O O O O O O 123 Summary . . . . ............ . . . . . . . . . . . 124 Introduction . . ................. . . . . . l25 Methods and Materials Materials . . . . . . . . . . . . . . . . . . ...... l26 Preparation of tissue samples . . . . . . . . . . . . . . l26 Polyacrylamide gel electrOphoresis and analysis by immunoblotting . . . . . . . . . . . . . . . ..... l27 Isolation of CBP35 from murine lung . . . . . . . . . . . l28 Iodination procedures . . . . . . . . . . . . . ..... l28 Binding of lectin to erythrocytes ....... . . . . . l29 Results Survey of the distribution of CBP35 in tissues of the mouse . . . . . . . . . . . . . . . . . . . . . . . . . 130 Purification of CBP35, CBPl6 and CBPl3.5 from mouse lung 134 Iodination of CBPs with retention of carbohydrate-binding activity . . . . . . . . . . . . . . . . . . . . . . l39 Binding of CBP35 (lung) to erythrocytes and inhibition by specific saccharides . . . . . . . . . . . . . . . . l42 Discussion ........ . . . . . . . . . . ....... l47 References . ............ . ........... l52 CLOSING STATEMENT . . . . . . . . . . . . . . . . . . . . . . . . 154 vi LIST OF TABLES Table Page Chapter I 1. Properties of lectins purified from various sources ..... 7 Chapter II 1. Agglutination and enzymatic activites of CBP35, CBPl6 and CBPl3.5 O O O O O O O I O ........ O O O O O O O O O O 65 Chapter IV 1. Effect of saccharides on the binding of CBP35 (lung) to erythrocytes ................ . . . . . . . . 143 Closing Statement l. Comparison of CBP35, CBPl6 and CBPl3.5 to chicken lactose lectin I, chicken lactose lectin II and hepatic asialoglyco- prOtEIn receptor 0 O O O O O O O 0 O O O O O O O O O O O O O 156 yii LIST OF FIGURES Figure Page Chapter I 1. Schematic of (A) the conversion of glcNAc covered M6P to uncovered M6P and (B) the structure of glucose covered M6P . l2 Chapter II 1. Chromatography of [35$]methionine-labeled extracts from 3T3 fibroblasts on ASF-Sepharose . . . . . . . ....... 43 2. PAGE analysis of 3T3-polypeptides eluted from ASF-Sepharose . 45 3. Rechromatography of ASP-binding proteins on a second ASF- Sepharose column . . . . . . . . . . . . . . . . . . . . . . 48 4. Two dimensional gel electrOphoretic analysis of CBP35, CBPl6 and CBPl3.5 O O O O O O O O O O O O O O O I O O O O O O O O O 5] 5. Effect of saccharides on the elution of CBP35, CBPl6 and CBPl3.5 from ASF-Sepharose ................. 54 6. PAGE of fractions eluted from ASF-Sepharose by various carbohydrates . . . . . . . . . . ......... . . . . . 56 7. Effects of EDTA on the elution of CBP35, CBP16 and CBP13.5 boundtOASF-Sepharose.c.................59 8. Sephadex G-l50 chromatography of CBP35, CBPl6 and CBP13.5 . . 6l 9. PAGE of fractions from Sephadex G-150 chromatography . . . . 63 lo. Binding of CBP35 to ASF-Sepharose . . ........... . 67 ll. Binding of CBPl6 and CBPl3.5 to ASF-Sepharose . . . . . . . . 69 l2. PAGE of polypeptides precipitated by antibodies directed against CBP35 and those precipitated by antibodies directed against chicken lactose lectin I . . . . . . . . . . . . . . 73 viii Figure Page Chapter III l. Schematic representation of the method used to prepare CHO-PA affinity columns ................. . . . . . 88 2. Effect of pH on the coupling of 2-aminoethanol to Bio-Gel P-150 acyl aZTdE o o o ..... o ccccccc o o o o o o 114 3. The rate of coupling 2-aminoethanol to Bio-Gel P-150 acyl aZide O O O O O ..... O O O O O O ....... O O O O O 117 4. The effect of reaction volume on the efficiency of coupling z-aIDTIIOEthaTIO1 t0 BIO-€161 P-ISO acyl aZTdE o o o o o o o o o 119 5. Effect of increasing 2-aminoethanol added on the amount of 2-aminoethanol coupled to Bio-Gel P-l50 . . ...... . . . l22 6. Elution profiles of GlcNAc-HA-PA affinity column . ..... 9l 7. Chromatography of [35$]methionine- labeled CBP35, CBP16 and CBPl3.5 on columns of (a) Gal -GlcNAc-HA- PA and (b) Gal- HA-PA O O O O O O I O I O O O O O O O O O O O O O O O O O 95 8. PAGE of proteins derived fran mouse 3T3 and human fibroblasts after affinity chromatography on CHO-HA-PA columns . . . . . 97 9. Affinity chromatography of [35$]methionine-labeled extracts of 3T3 fibroblasts on Gal-GlcNAc-HA-PA ....... 99 Chapter IV l. Survey of the tissue distribution of CBP35 in adult male mouse 0 O O O O O O O O O O O O O O I O O O O O O O O O O O 132 2. Representative column profile of mouse lung extracts on ASF-Sepharose . . . . ................... . l36 3. PAGE and immunoblotting of CBP35 isolated from mouse lung . . 138 4. Chromatography of 1251-1abe1ed capss on ASF-Sepharose . . . . 141 5. Concentration dependence ofzghe effect of galactose and lactose on the binding of I-CBP35 to erythrocytes . . . . 145 ix CBP CHO LE M6P CLL I CLL II DMEM ASF PMSF SDS PAGE HA PA PBS Hepes EDTA Tris ABBREVIATIONS carbohydrate binding protein carbohydrate lysosomal enzyme mannose 6-phosphate chicken lactose lectin I chicken lactose lectin II Dulbecco-Modified Eagle's Medium asialofetuin phenyl methyl sulfonylfluoride sodium dodecyl sulfate polyacrylamide gel electrophoresis hexanolamine polyacrylamide phosphate buffered saline N-2-hydroxyethylpiperazine-N-Z-ethanesulfonic acid (ethylenedinitrilo)-tetraacetic acid Tris (hydroxymethyl)aminoethane INTRODUCTION This thesis describes the purification and characterization of three carbohydrate binding proteins (CBPs) from Swiss 3T3 fibroblasts. Since relatively little is known about these lectins, I have chosen to review the literature on CBPs, isolated from other sources, which have been studied in more detail. As will become apparent, certain of these CBPs may be related to those isolated from 3T3 cells and, therefore, may serve as paradign systems for the initial studies concerning the cellular functions of the fibroblast CBPs. However, it will also become apparent that in very few instances are the functions of the CBPs well understood. We have chosen to search for CBPs in Swiss 3T3 fibroblasts for numerous reasons. The foremost is that this cell line has been exten- sively characterized in terms of its genetic stability (1), growth control (2), nutrient requirements (3), differentiation (4) and transformation (5) as well as having had many molecular components identified. This allows for the study of the functions of the CBPs in a relatively defined system. Although a variety of fibroblast cellular events are affected by reagents which alter the cellular state of glycosylation, very little is known about the nature of the molecules which are affected. One class of examples of the effects of reagent that affect glycosylation include the regulation of cell proliferation by uridine 2 diphospho-Z-N-acetyl-glucosamine (83) and by tunicamycin (84) in the 3T3 system. One possible explanation for these effects is that inter- actions between glycoconjugates and CBPs are prevented. Therefore we have proceeded to purify and characterize CBPs from 3T3 fibroblasts. This will allow us to investigate the importance of glycoconjugate-CBP interactions in fibroblasts. Chapter I LITERATURE REVIEW (I) Carbohydrates as Recognition Markers. Until recently the function of the CH0 moieties of glycoconjugates has not been well understood. Eylar (6) and Melchers (7) proposed that glycosylation is a means by which a protein is targeted for export fron the cell. Inconsistent with this idea, however, are the findings that many glycoproteins are found within the cell. Moreover, Blobel (8) has shown that the instructional information for export resides in the polypeptide and is independent of glycosylation. Winterburn and Phelps (9) proposed that the CH0 detennines the fate of the protein after it is secreted. Envisaged in their scheme is a direct effect of CHOs on the interaction of the glycoprotein with receptors on the plasmalemma. Indeed, the data supporting this notion are accumulating. Not only do the CH0 moieties affect interactions of glyc0proteins with receptors on the plasmalemma but they also affect interactions of this class of proteins with receptors residing on intracellular membranes. Many mechanisms by which the CHO can affect these interactions with receptors are possible. However, only those interactions which result directly from the binding of the CHO moiety will be discussed in this thesis. The heterogeneity of sugar sequence, linkage and confonnation makes possible for a large repetoire of encoded "signals". However there appear to be limitations on the structures which are actually found in .3 4 nature. This is most probably due to the mechanisms by which the oligosaccharides are constructed (g.g. formation of high mannose oligosaccharides on dolichol phosphate). This limitation may also result directly from the relative abundance of the various oligosaccharides (i.g. there may be many structures which are minor components). It should be noted, however, that the variety of known CHO structures is rapidly increasing and that the potential for a tremendous amount of "signal" information in the CH0 structure exists. Thus alteration of some structural feature of the oligosaccharide such as addition or removal of a monosaccharide (sialic acid), change in linkage (l--+4 vs. l--*3), anomeric change (a to a) or chemical modification (phosphorylation) can lead to dramatic effects on the interaction of a glyc0protein with its receptor. (II) Carbohydrate-Binding,Proteins If the CH0 moiety serves as a recognition marker, then the cell must possess a complementary set of molecules which have the inherent ability to specifically recognize and bind distinct structural features of the oligosaccharide chains. These complementary molecules are referred to as carbohydrate binding proteins (CBPs). If a major func- tion of the CHOs on glycoproteins is to serve as “signals" it would be expected that there be many CBPs, each of which recognizes some special feature of the oligosaccharide. Actually very few CBPs relative to the variety of CHO structures have been isolated. Undoubtedly there exists as yet unidentified CBPs. It is also possible that it is not necessary for a cell to possess a large variety of CBPs. Instead the CBP-glyco- protein interactions can be regulated by developmental controls (either 5 by time dependent expression of the CBP and/or complementary glyco- protein ligand), compartmentalization as well as by many other mechanisms. Many functions have been attributed to CBPs. Here they are arbitrarily divided into three classes: (a) transport glyc0proteins, (b) proteins that organize cellular domains, and (c) proteins that mediate cell recognition and aggregation. In discussing each class of functionally defined CBPs the difficulty of determining the function of a CBP should be considered. This is due, in part, to the difficulty of identifying the_1g.yiyg ligand. It is also evident that the function ascribed to a certain CBP is depenent on the assay employed. In the following discussion these pitfalls will be addressed when they are relevant. (III) Protein Transport: Endocytosis - Asialoglycgprotein Receptor Pinocytosis of desialylated glycoproteins is mediated by CBPs. The first major breakthrough in this area was the finding that removal of terminal sialic acids on glyc0proteins results in their rapid clearance from the circulatory system (l0). The liver is the major organ responsible for this clearance (ll). Both perfused liver and cultured. hepatocytes show this ability to endocytose asialoglycoproteins. In mammals, a CBP has been shown to bind and to internalize galac- tose bearing compounds. Direct evidence for the involvement of these CBPs in this clearance are: (i) antibodies directed at the CBP inhibit this clearance (12), (ii) insertion of the CBP into fibroblasts which do not contain it, endows them the ability to endocytose galactose bearing ligands at relatively high rates (l3), and (iii) when these galactose tenninating ligands are bound to hepatocytes and crosslinked with a bifunctional reagent it is the CBP to which they are crosslinked (l4). These as well as other less direct evidence have firmly established the roles of CBPs in the removal of glycoproteins from the circulatory system. The CBP responsible for glyc0protein clearing has been purified fran a number of mammalian livers. The intact proteins are constructed from polypeptides with apparent molecular weights of 48,000 and 40,000 for rabbit (15), 52,000 for rat (Table I) (16) and 4l,OOO for human (l7). Although isolated from different sources the CBPs share many conmon characteristics. They are all gal actose binding proteins. In aqueous solution they form large aggregates which, in the case of the rabbit CBP, can be dispersed into 260,000 dalton species upon the ad- dition of the detergent, Triton X-lOO (l8). It is presumed that this 260,000 dalton species is comprised of two large and four small sub- units. Based on target analysis from radiation inactivation of isolat- ed rat liver plasma membranes, a molecular weight of l05,000 dalton (two 52,000 dalton subunits) has been assigned to the rat CBP (l9). Thus it appears that the active CBPs are oligomers. All these CBPs are glycoproteins and have an absolute requirement for divalent cations for expression of binding activity (20). Avian species possess a similar system with one major exception, the CH0 specificity differs (21). Following binding and endocytosis the asialoglycoprotein is degraded in the lysosomes (22). Initially the ligands are bound to CBPs which appear to be diffusely distributed on the plasmalemma. The ligand-CBP complexes then translocate laterally through the membrane and aggregate in coated pits (23). The coated pits subsequently bud .cwmuocaooapm .au mesomomap .>4 .5333? 0.252825 .E 332.6 .u $52.5... 3.....qu .za mH—s—um .< move—05.225 am "CFHUQP OmOHUQP CUJUPSU «:30 "wow: mcowumv>mcan< ammucaom maorcm> soc» nmwm_c:a mappomg to mo_acmaoca "H opnmh 02:25 22:6 moamam .8 £0 anmzto_u.owam . uz:mm»z_.< 55-4 :58 I ham“. 529x-“ mao>muz-m 88.8.8 55... 20.59599 55... -w 38:: -m mswm: , 2.5% So: Em quom «mm maze. 55.x... 296.5 2.36.5 2429: 5. 358 Av . .v .%w A/ .Lu .u L7 afiv n%u may xx \A/ %%7 .QMV 4a .ny a», o. x.. a? 8 into the cytoplasm and are directed towards the lysosomes. Once internalized the ligand-CBP complexes become associated with smooth membranes termed endocytotic vesicles. The ligands are ultimately delivered to the lysosome where they are degraded to amino acids and monosaccharides. During this delivery the CBP becomes separated from the ligand (24) and is spared from degradation. A number of elegant experiments have shown that the CBPs originally found at the cell surface can participate in many cycles of endocytosis (25). In the absence of protein synthesis hepatocytes can endocytose a quantity of ligand in great excess of their number of CBPs (26). Furthermore they continue to endocytose these ligands at or near a rate equal to the initial rate even while degradation of the ligand is occuring (27). These results have been interpreted to mean that recep- tors recycle. However, the CBPs may be segregated into two or more functionally distinct pools, one pool which recycles and others which do not. It has been proposed that the majority of CBPs found on intracelluar membranes do not cycle nor do they participate in endocytosis (28). The mechanism(s) by which these CBPs are segregated and by which they recycle remain to be determined. Subcellular distribution and binding studies further support the idea that these CBPs are recycled. Of the total mammalian liver, galactose-specific CBP only 5% is found on the plasma membrane. The remaining binding activity resides on internal membranes (43% on microsomes, 32% on golgi and 20% on lysosomes) (29). The CBP is oriented towards the lumen in all the intracellular organelles except the lysosome where it has its binding site oriented outward. This exterior orientation on the lysosome, which would require trans-bilayer 9 displacement, may explain in part, the mechanism by which the CBP is spared degradation. More recently it has been demonstrated that the CBP and the ligand are separated into different membranous compartments prior to ligand delivery to the lysosome (30). Thus the ligand proceeds to the lysosome, whereas the CBP is free to recycle. Although the studies demonstrating the involvement of CBPs in the clearance of asialoglyc0proteins from the circulatory system have been extensive, there remains the possibility that these CBPs are also involved in the intracellular transport of endogenous glyc0proteins. Of major concern in this regard is the assay which is the mainstay of this particular field, 122- the use of an exogenous ligand. It appears that many plasmalemma receptors deliver their ligands to the lysosome. This list includes those receptors Specific for polypeptide hormones, low density lipoprotein, mannose 6-phosphate bearing proteins and o2- macroglobulin (31). It is possible that those receptors found on the cell surface have one major "artery" into the cell and this "artery" goes through/to the lysosome. Therefore ligands which cannot survive the lysosomal milieu are degraded. An endogenous ligand such as a newly synthesized glycoprotein may never see the lysosome. Instead it could be delivered to an organelle prior to seeing the lysosome or may have an intracellular pathway which excludes the lysosome. If the CBP is involved in transport/sorting mechanisms it would be expected to be ubiquitous. Reticuloendothelial cells possess a CBP which mediates the endocytosis of glycoconjugates bearing terminal man- nose or N-acetylglucosamine residues (32). A man/glcNAc specific CBP has been isolated from rabbit liver (33), although no evidence iden- tifying this CBP as the one involved in endocytosis has been presented. l0 Recently a CBP which, based on subunit molecular weight and immunologi- cal crossreactivity, appears to be identical to that isolated fron liver, has been isolated from hepatocytes (34). Whereas reticuloen- dothelial cells can endocytose man/glcNAc terminating glycoconjugates, hepatocytes can not. Furthermore, the hepatocyte CBP appears to be an intracellular protein and is not an integral membrane protein. If the CBP isolated from whole liver turns out to be responsible for endocyto- sis then cell specific roles must be postulated. These would have to include the possibility that a given CBP can play different roles, such as involvement in endocytosis when it is expressed on the cell surface and some other function when it is located intracellularly. (IV) Protein Transport:Intracellular Sorting - Mannose 6-Phosphate Receptor Glycoprotein synthesis occurs on the rough endOplasmic reticulum where the polypeptides are deposited in the lumen (35). Subsequently the proteins must be segregated and delivered to their respective subcelluar sites. In the case of fibroblast lysosomal enzymes (LES) it is a CBP which is responsible for this segregation and delivery. In the classical studies of Neufeld and coworkers (36), to show that a CBP was involved in delivery of LEs, advantage was taken of the wide range of human mutant fibroblasts which are available. These mutants have single mutations which results in the absence of a particular lysosomal enzymatic activity. Early studies demonstrated the ability of these mutant cells to acquire corrective factors (leg. enzyme replacement) which were supplied to the media of cultured 11 fibroblasts. This uptake of the lysosomal enzymes was demonstrated to be a receptor mediated transport process (37). The CHO moieties of the LEs were shown to be essential components for this endocytosis (38). Specifically, mannose 6-phosphate (M6P) residues are bound by a CBP. Their uptake is inhibitable by yeast phosphomannans (39), mannose 6-phosphate and fructose l-phosphate (37). Furthennore analysis of the LE oligosaccharides revealed high mannose types which are phosphorylated on certain 6-hydroxyls. There appears to be at least five different sites of phosphorylation and the degree of phosphorylation of each oligosaccharide chain is variable (40). The acid hydrolases are phosphorylated by the transfer of N-acetylglucos- amine l-phosphate from UDP-N-acetylglucosamine to the 6-hydroxyls on the high mannose oligosaccharide. The blocking a-N-acetylglucosamine is subsequently removed exposing the mannose 6-phosphate residues (Fig. l). Both of these enzymatic activities are localized in the cis-golgi and endoplasmic reticulum (4l). This is the expected location of these enzymatic activities if the NSF moieties are involved in targeting these enzymes. The most convincing evidence which suggests that the M6P binding protein serves to sort newly synthesized endogenous LEs comes fran studies of fibroblasts from patients with mucolipidosis II disease. These cells do not possess the UDP-N-acetylglucosamine: glycoprotein N-acetylglucosamine l-phosphotransferase activity (42). This results in LEs which lack phosphates on their oligosaccharides. The conse- quence of this is that the enzymes are secreted and the cells exhibit a deficiency of LEs in their lysosomes. Thus, at least in fibroblasts, the failure to phosphorylate the oligosaccharides results in the 12 O O u “ o-e-o NAC 0‘ O O-R O gGIcNAc Figure l: Schematic representation of: (A) the conversion of GlcNAc-covered M6P to uncovered M6P; and (B) the structure of Glc-covered M6P. l3 diversion of newly synthesied LEs from their normal intracellular pathway. The M6P receptor has been purified from bovine liver (43), human fibroblasts and Swanm rat chondrosarcoma cells (44) (Table I). The receptor is a glycoprotein wfith subunit molecular weight of 215,000. The receptor binding does not require divalent cations for CH0 binding and this binding is reversed below pH 6. It is an integral membrane protein since it requires detergent for solubilization and can be adsorbed into liposomes. The distribution as well as the degree of occupancy by LEs of the receptor in rat liver has been determined. Of the total phosphoman- nosyl enzyme receptors, 90% were found in end0plasmic reticulum, Golgi apparatus, and lysosomes (78%, 7%, 5% respectively) and l0% was found in the plasma membranes (45). Other organelles had negligable binding activity. The receptors appear to be on the luminal surfaces of all the organelles except the plasma membrane where they are oriented outward from the cell. Furthermore, of the total receptors in the endoplasmic reticulum and Golgi more than 80% and 50%, respectively, of the binding sites are occupied. Only l0% of the binding sites in the lysosomes and plasma membranes were so occupied. Thus the receptors are distributed and occupied in a gradient which would suggest their involvement in the delivery of newly synthesized enzymes from the endOplasmic reticulum to the lysosome. An affinity column covalently derivatized with the M6P receptor has been used to determine the structural requirements for binding to the CHOs (46). Oligosaccharide bearing covered M6P residues (the N-acetylglucosamine has not been removed, Fig. l), do not bind, those l4 with one exposed M6P are slightly retarded and oligosaccharides containing two or more exposed M6P residues require the addition of M6P to the buffer to elute them from the column. Thus the receptor binds to eXposed M6P moieties and increasing the number of M6P residues results in tighter binding. These results were found to correlate with the ability of human fibroblasts to endocytose these oligosaccharides (47). Although these studies clearly indicate that oligosaccharides with two or more exposed M6P moieties is optimal for binding and endocyto- sis, the situation is more complex when considering the binding and uptake of an intact enzyme. Most acid hydrolases are multimeric which results in multiple oligosaccharide moieties. In addition certain of the polypeptides have multiple sites of glycosylation. The effect on uptake of multiple recognition sites on an intact protein can clearly cause synergistic effects such that what may be weak interactions be- tween the receptor and oligosaccharides can be strong interactions be- tween the acid hydrolase and components of the endocytotic mechanism. The involvement of the M6P receptor in the intracellular transport of LEs may be a specialized function in fibroblasts and other cell types. As noted above patients afflicted with mucolipidosis II disease fail to phOSphorylate their LEs and thus cultured fibroblasts from these patients have very deficient levels of intracellular LEs. However it has been found that certain tissues (liver, kidney, brain and spleen) from these patients acquire near normal amounts of lysoso- mal activities except for s-galactosidase (48). It appears that these tissues contain an alternate mechanism for the targeting of LEs to the lysosome. It is possible that other CHO receptors may be responsible 15 for their uptake from the circulatory system or that an as yet unknown mechanism is responsible for these tissues ability to target the enzymes to the lysosomes. It may also be possible that the targeting of these enzymes via the M6P receptor to the lysosome in fibroblasts is a consequence of culture conditions. This is unlikely when the clinical aspects of mucolipidosis II disease are considered. In these patients the connective tissues are most severly affected (49). This correlates well with the high content of fibroblasts in these tissues. The role the fibroblast M6P receptor plays in the delivery of LEs is firmly established. However there is evidence that certain of the ligands are unique, thus providing a very specialized function of the M6P receptor-LE system. Uteroferrin is a protein which is: (i) iron containing, (ii) secreted in large amounts by the porcine uterus during pregnancy, (iii) is absorbed by the develOping fetus, (iv) is endocytosed by a mannose receptor mediated event in reticuloendothelial cells, (v) has acid phosphatase activity, (vi) bears glcNAc covered M6P moeities and (vii) is located in intracellular vacuoles resembling lysosomes (50). Therefore uteroferrin appears to be an overproduced LE which has both acid phosphatase and iron transport activity, resulting in a LE with specialized functions. It has also been shown that M6P, fructose l-phosphate and fructose 6-phosphate inhibit 1a.!1trg human natural cell-mediated cytoxicity (51). Although no molecular evidence has been presented it is suggest- ed that this effect may be due to either the inhibition of binding of the effector cell to the target cell or inhibition of binding of a cytolytic effector molecule to the target cell. If it is a cytolytic factor there must be specific mechanisms such that the natural killer 16 cells are not exposed to the factors activity which would result in suicide. When mouse fibroblasts are either virally transformed or treated with growth factors the major excreted protein (MEP) is a 35,000 dalton glyc0protein (52). MEP contains M6P moeities and binds to immobilized M6P receptor in an M6P inhibitable manner (53). However the phospho- glycoprotein is unique in that it is not endocytosed by cells. The paradox that it binds to the receptor but is not internalized remains to be clarified. The possibility that MEP is a hydrolase, localized to the plasma membrane by the M6P receptor, and is involved in the maintenance of the transfonned phenotype is an intriguing one. There is recent evidence which supports the general idea that the M6P receptor may serve to localize certain hydrolases at the cell surface (54). When human fibroblasts are cultured in the presence of M6P, there is a decrease in the amount of glycosaminoglycans released into the media as compared to fibroblasts grown in its absence. This is accompanied by an increase in cell surface glycosaminoglycans in cultures grown in the presence of M6P (55). Furthermore there are distinct molecular differences between the glycosaminoglycans isolated from M6P treated cultures and control cultures. Based on specificity, dose response and cell specificity it was concluded that this effect was due to the dissociation of LEs from cell surface M6P receptors. It was proposed that M6P receptors serve to anchor LEs proximate to specific substrates, such as glycosaminoglycans, so that they can cleave them. l7 (V) Organization of Domains-Ligatin Just as a cell is divided into regions of metabolic activity so are these organelles further subdivided into specific domains. These domains consist of the segregation of macromolecules to a specific region of a membrane. An involvement of CBPs in the formation of these domains has been suggested by a number of laboratories. Ligatin (Table I) is a protein which forms regular arrays of 4.5 nm filaments on suckling rat intestinal mucosa (56) and on the cell surf- ace of chick neural retina (57), mouse macrOphages, sea urchin spenn and hunan fibroblasts (58). The protein, with associated lipids, is released fron the cell surface by treatment with 30 mM calcium or by alkaline pH. When the calcium is removed these filaments dissociate into monomers of Mr = l0,000. Readdition of calcium causes the mono- mers to polymerize back to filaments of 3 nm diameter. This difference in diameter is thought to result from the absence of associated poly- peptides in the reconstituted filaments. This suggests that ligatin serves as a baseplate at the cell surface for the attachment of other proteins. One protein which is associated with neonatal rat ileum ligatin is N-acetyl-s-D-glucosaminidase. This protein is bound to ligatin via glucose l-phosphate residues. Whereas the M6P receptor does not bind to covered M6P residues, ligatin preferentially binds to M6P residues covered by glucose in a phosphodiester linkage (Fig. 1). Recently an enzymatic activity which transfers glucose l-phosphate to high mannose oligosaccharides has been identified in enbryonic chick neural retina (59). 18 Additional evidence is accumulating which supports the idea that ligatin serves to localize certain glchproteins to the plasmalemma. When ligatin is extracted from cerebrum membranes, acetylcholinesterase is cosolubilized. This acetylcholinesterase binds to ligatin affinity columns and can be eluted with either M6P or glucose l-phosphate (60). Marchase's laboratory has found that washing neural retina cells with glucose l-phosphate releases a glycoprotein of 48,000 daltons (Marchase, personal communication). Although this glyc0protein has not been identified it may be the protein cognin (50,000 daltons) which has been shown to promote the aggregation of neural retina cells (61). Consistent with this possible localization of cognin by ligatin is the finding that addition of ligatin to dispersed neural retina cells inhibits their aggregation (62). This may possibly be due to the binding of cognin molecules, thus removing then from the surfaces of cells. It has also been reported that neural retina cells release glycoprotein complexes, termed adherons, into the culture medium (63). Adherons have been shown to increase cell-cell aggregation and cell- substrate adhesion. Recently Marchase's group has found that these adherons contain glucose l-phosphate bearing components (Marchase, personal communication). When all the preliminary data are combined, a role for ligatin in the localization of cell-cell adhesion factors in neural retina cells is suggested. (VI) Organization of domains - Low Molecular Weight Galactose Specific Lectins. There is a group of lectins whose functions have been very elusive. Teichberg gt 21. first identified galactose specific agglutination 19 activity in a number of tissues from rat, chicken, eel and mouse (64). They subsequently purified a galactose specific lectin from eel and others have isolated similar lectins from bovine (65) and chicken (66) tissues. These lectins have many common pr0perties (Table l): (I) have subunit molecular weights of l0,000- l6,000 daltons, (ii) are sensitive to air oxidation, (iii) do not require divalent cations for binding, (iv) are extractable from disrupted cells in the absence of detergent, (v) have low isoelectric points of 3-5, ard (vi) most appear to fonm dimers in aqueous solution. Although these lectins have been studied in a number of systems their functions are not known. Since the most recent studies implicate their involvement in the organization of domains they have been included in this section. However a number of other functions have been proposed. Chicken-lactose-lectin I (CLL I) is a 32,000 dalton protein com- prised of two identical l6,000 dalton monomers (Table I). It is found in embryonic muscle and is developmentally regulated. The activity increases in early develOpmental stages and has maximal activity at a time which coincides with the fusion of myoblasts to form myotubes. The activity subsides at later times and in adult muscle is quite low (67). This temporal correlation of activity with morphological events spurred investigations into the involvement of this lectin in the fusion process. Inclusion of either thiodigalactose (68) or CLL I (69) in cultures of myoblasts has been reported to inhibit myoblast fusion. These results suggest that CLL I is involved in the fusion process. However these results are subject to debate since a series of reports (70) which claim that neither thiodigalactose nor CLL I has an effect on fusion, have appeared. 20 Recent reports have suggested a role for CLL I in the organization of glycoconjugates during the formation of T-tubules (71). In myo- blasts the lectin is predominatly intracellular. The lectin is concen- trated in longitudinal lines and perpendicular spokes radiating from these lines as myoblasts fuse and begin synthesis of contractile proteins. This pattern is similar to that observed with T-tubules. Later in development as the sarc0plasmic reticulum fuse hnth the plasma membrane, the lectin becomes localized extracellularly. Thus, it appears that CLL I may serve to organize glycoconjugates which are involved in the formation of T-tubules during differentiation. A similar involvement of lectins in the organization of glycoconjugates has been preposed for chicken-lactose-lectin II (CLL II), a monomeric protein of Mr = 14,000 (66). CLL II (Table I) is very abundant in intestine and is located in secretory vesicles of mucin-secreting goblet cells (72). It is also located on the lumenal surfaces of the epithelial cells which line the intestine. It is presumed that the lectin found on the mucosal surface is bound to either multiple mucins or crossiinks mucin to glycoproteins on the intestinal epithelial surface. Furthermore, it was demonstrated that CLL II is secreted in conjunction with mucin. The findings are similar to those concerning CLL I and suggest that lectins in animal tissue may play an important role in the organization of glycoconjugates. The identification of the functions of CLL I and CLL II are com- plicated by their tissue distribution. Extracts of many embryonic and adult chicken tissues contain substantial amounts of both CLL I and CLL II (73). Both lectins show striking changes in concentration at dif- ferent stages of development. Whereas CLL I concentration is greatly 21 decreased in adult muscle as compared to embryonic muscle it is greatly increased in adult liver when compared to the embryonic liver. CLL II is very abundant in embryonic kidney but is present at less than one tenth that level in the adult organ. Conversely CLL 11 increases 30 fold in the intestine in the adult as compared to the 15 day embryo. These results suggest multiple functions for these CBPs. Furthermore the functions may be Specific to the cell types in which they are found. The involvement of a galactose specific lectin (Mr = 13,000) has been implicated in erythroid develOpment (74). In adult mammalian bone marrow developing erythroblasts are clustered closely together in "erythroblastic islands" around a central macrOphage nurse cell. It is proposed that the lectin is involved in the bridging of these erythro- blasts to fonn these clusters. Consistent with this notion is the finding that the lectin is found on the cell surface of the erythro- blasts but not on cells which have undergone further differentiation. It has been found that as these cells mature their susceptability to agglutination by this lectin decreases, indicating a decrease in the amount of cell surface galactose terminating glycoconjugates. Indeed the levels of galactose tenminating glycoconjugates has been shown to change as a function of development (75). Not only is there a quantitative change but there also appears to be an organizational change. That is, there are patches or domains of these galactose bearing compounds on the cell surface. This raises the possibility that the lectin may be involved in the fonnation of these microdomains rather than bridging molecules between two adjacent cells. 22 The major drawback in these studies is the lack of information concerning their natural ligand(s). As previously noted, this is a difficulty not easily overcome. Since many glycoconjugates contain galactose moeities, it is impossible to identify the in situ ligand solely on the basis of their binding to the lectins. Thus an important breakthrough in this field will be the establishment of procedures to identify the_i_ situ ligand. (VII) Cell Aggregation: S1ime Molds -Discoidin Certain homotypic and heterotypic cell aggregation processes have been implied to be mediated by lectins. This type of activity was originally identified in the symbiotic interactions of nitrogen fixing bacteria with legumes (for review see Ref. 76). It is generally accepted that these Species Specific lectins are responsible for the initial recognition and binding events between the bacteriun and the root hairs. Thus the lectins confer the Specificity seen in these processes. These studies have prompted investigations into the possibility that lectins are involved in the cell-cell aggregation/ recognition processes observed in other organisms. The most convincing results concern the aggregation of the slime molds. When supplied with adequate nutrients the slime molds exist as independent amoebae. Upon starvation the cells display a mutual cohesiveness and aggregate into multicelluar pseudoplasmodia (Slugs). Further morphogenetic events result in the fbrmation of mature fruiting bodies. Lectins have been shown to mediate the initial aggregation process. 23 This event is species Specific in that if two or more Species of slime molds in the amoeboid stage of their life cycle, are mixed prior to starvation, they seek out members of their own species to fbrm Slugs of a homogenous cell type upon removal of nutrients. Therefore if lectins are the molecules responsible for this Species specificity the different slime molds should have distinct lectins. To date identical lectins have not been identified to be present in more than one species of slime mold. Moreover different species appear to possess unique lectins. It has been shown that in Dictyostelium discoideum an endogenous lectin, discoidin I (Table I), is absolutely required for these cells to fonn aggregates. The lectin is a tetrameric protein consisting of subunits of Mr = 25,000 (76). The lectin binds to terminal and pos- sibly internal galactose residues. When the slime mold is in the vega- tative state (single cell) the lectin is virtually absent. In the aggregating amoebae it may comprise more than 1% of the total cellular protein (78) and this increase is under transcriptional control. The initial increase in lectin activity precedes cell cohesiveness by approximately two hours and has its most rapid increase at a time coincidental with the rapid increase in the aggregation of the amoebae" (79). The lectin has been demonstrated to reside at the cell surface by a number of procedures: (1) it can be labeled with 125I by cell surface labeling procedures, (ii) is immunoreactive with fluorescent antibodies when intact cells are used, and (iii) aggregating cells agglutinate erythrocytes in a galactose inhibitable fashion; however nonaggregating cells do not agglutinate erythrocytes. Although the 24 lectin is a cell surface protein it is not an integral protein since it can be extracted fron disrupted cells in the absence of detergents. Of the total cellular lectin in aggregating cells only about 2% (leg. 1 x 105 molecules per cell) is detectable on the cell surface; the remainder is intracellular. It has been shown that the intracellu- lar lectin can be elicited to the cell surface. When monovalent Fab fragments against the lectin was used to quantitate surface lectin less than 1 x 105 molecules per cell were detected. However if divalent antibodies were used then at least 1 x 106 molecules per cell were seen. Polyvalent glchproteins and conconavalin A can also elicit the appearance of additional cell surface lectin. Although the relevance of this elicitation to the in 1119 Situation is still somewhat obscure, there does appear to be a possible correlation between the induced movement of the lectin to the cell surface and the progressive externalization of the lectin during differentiation. Towards later stages of aggregation, the lectin is no longer detectable within the cells, but is confined to the aggregate's periphery. It finally becomes associated with the amorphous material associated with the surface of the slug as well as with material which is shed (80). The most convincing evidence suggesting an involvement of lectins in this aggregation process comes from work with mutants. Mutants which have discoidins containing defective binding sites fail to aggre- gate and differentiate past this step (81). The mutant lectin is syn- thesized upon starvation and is localized to the correct subcellular Site. However it fails to agglutinate erythrocytes. Revertants of this mutant acquire aggregation competence. It is concluded from these 25 studies that the lectin plays an essential part in this differentiation process. Whether the lectin serves to bridge molecules between adjacent cells or aggregates glycoconjugates into a domain on a single cell is unknown. It is most likely that there is a sequence of events which lead to interactions which provide for the avidity required to maintain these tight cell-cell contacts. A number of glycoproteins have been suggested to be involved in this formation of cell-cell contacts, but the ligand-receptor relationship of these glyc0proteins and lectin remain to be determined. (VIII) Cell Aggregation: Sperm Egg Interaction-Bindin The process of fertilization requires that spenn bind to the egg's surface. It has been proposed that a lectin mediates this initial binding event. A lectin, bindin, (Mr = 30,500) has been isolated from the acrosomal processes of Sperm from the sea urchin, Strongylocentrotus purpuratus (82). Bindin agglutinates eggs in a species Specific manner and this agglutination is inhibitable by mild periodate oxidation of the egg as well as by glycopeptides derived fran proteolyzed vitelline layers. AS the Spenn binds to the egg the lectin is localized to the point of attachment on both the acrosomal process and the vitelline layer adjacent to this site of attachment. Although there is no direct evidence as to the function of this lectin it appears that it is involved in the fertilization process. b 00 N O O m o \OSDNOS 10. 11. 12. 13. 14. 15. REFERENCES Todaro, G.J. and Green, H., J. (1963) J. Ce11 Biol. 11, 299-313. Stoker, M.G.P. and Rubin, H. (1967) Nature (Lond.) 215, 171-172. Barnes, 0. and Sato, 6. (1981) Cell 22, 649-655. Green, H. and Kehinde, 0. (1974) Cell 1, 113-116. Todaro, G.J., Green, H. and Goldberg, 8.0. (1964) Proc. Natl. Acad. Sci. USA 51, 66-73. Eylar, E.H. (1965) J. Theoret. Biol. 19, 89-113. Melchers, F. (1973) Biochem. 12, 1471-1476. Blobel, G. (1980) Proc. Natl. Acad. Sci. USA 17, 1496-1500. Winterburn, P.J. and Phelps, C.F. (1972) Nature 236, 147-151. Morell, A.G., Irvine, R.A., Sternlieb, I.H., and Ashwell, G. (1968) J. Biol. Chem. 23;, 155-159. Morell, A.G. and Scheinberg, I.H. (1972) Biochem. Biophys. Res. Comm. 48, 808-815. Stockert, R.J., Gartner, U., Morell, A.G. and Wolkoff, A.W. (1980) J. Biol. Chem. 255, 3830-3831. Doyle, 0., Hou, E. and Warren, R. (1979) J. Biol. Chem. 254, 6853-6856. Baenziger, J.U. and Fiete, D. (1982) J. Biol. Chem. 251, 4421-4425. Hudgin, R.L., Pricer, W.E., Jr. and Ashwell, G. (1974) J. Biol. Chem. 242, 5536-5543. 26 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 27 Stockert, R.J., Morell, A.G. and Scheinberg, I.H. (1977) Science 151, 667-668. Baenziger, J.U. and Maynard, Y. (1980) J. Biol. Chem. 555, 4607-4613. Kawasaki, T. and Ashwell, G. (1976) J. Biol. Chem. 551, 1296-1302. Steer, C.J., Kemper, E.S. and Ashwell, G. (1981) J. Biol. Chem. 2_5§, 5851 -5856. Strickland, D.K., Andersen, T.T., Hill, R.L. and Castellino, F.J. (1981) Biochemistry 29, 5294-5297. Kawasaki, T. and Ashwell, A. (1977) J. Biol. Chem. 555, 6536-6543. LaBadie, J.H., Chapman, K.P., Aronson, N.N., Jr. (1975) Biochem. J. 15;, 271-279. Kolb-Bachofen, V. (1981) Biochim. Biophys. Acta 555, 293-299. Bridges, K., Harford, J., Klausner, R. and Ashwell, G. (1982) Proc. Natl. Acad. Sci. USA 15, 350-354. Stockert, R.J., Howard, D.J., Morell, A.G. and Scheinberg, I.H. (1980) J. Biol. Chem. 255, 9028-9029. Steer, C.J. and Ashwell, G. (1980) J. Biol. Chem. 255, 3008-3013. Warren, R. and Doyle, D. (1981) J. Biol. Chem. 555, 1346-1355. Doyle, D., Baumann, H., Hou, E. and Warren, R. (1981) In International Cell Biolggy, New York, Rockefeller Univ. Press. Pricer, W.E., Jr. and Ashwell, G. (1976) J. Biol. Chem. 551, 7539-7544. Geuze, H.J., Slot, J.W., Strous, G.J.A.M., Lodish, H.F. and Schwartz, A.L. (1983) Cell 5;, 277-287. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 28 Goldstein, J.L., Anderson, R.G.W. and Brown, M.S. (1979) Nature 215, 679-685. Schlesinger, P.H., Rodman, J.S., Doebber, T.W., Stahl, P.D., Lee, Y.C., Stowell, C.P. and Kuhlenschmidt, T.B. (1980) Biochem. J. 152, 597-606. Kawasaki, T., Etoh, R. and Yamashina, I. (1978) Biochem. Biophys. Res. Commun. 51, 1018-1024. Maynard, Y. and Baenziger, J.U. (1982) J. Biol. Chem. 251, 3788-3794. Blobel, G. (1980) Proc. Natl. Acad. Sci. USA 11, 1496-1500. Cantz, M., Kresse, H., Barton, R.W. and Neufeld, E.F. (1972) Meth. Enzym. 25, 884-897. Sando, G.N. and Neufeld, E.F. (1977) Cell, 619-627. Kaplan, A., Achord, D.T. and Sly, W.S. (1977) Proc. Natl. Acad. Sci. USA 11, 2026-2030. Kaplan, A., Fischer, 0., Achord, D.T. and Sly, W.S. (1977) J. Clin. Invest. 59, 1088-1093. Varki, A. and Kornfeld, S. (1980) J. Biol. Chem. 255, 10847-10858. Pohlmann, R., Waheed, A., Hasilik, A. and von Figura, K. (1982) J. Biol. Chem. 251, 5323-5325. Reitman, M.L., Varki, A., and Kornfeld, S. (1981) J. Clin. Invest. 51, 1574-1579. Sahagian, G.G., Distler, J. and Jourdian, G.W. (1981) Proc. Natl. Acad. Sci. USA 15, 4289-4293. Steiner, A.W. and Rome, L.H. (1982) Arch. Biochem. Biophys. 211, 681-687. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 29 Fischer, H.D., Gonzalez-Noriega, A., Sly, W.S. and Morre, D.J. (1980) J. Biol. Chem. 255, 9608-9615. Fischer, H.D., Creek, K.E. and Sly, W.S. (1982) J. Biol. Chem. 251, 9938-9943. Creek, K. and Sly, W.S. (1982) J. Biol. Chem. 251, 9931-9937. Waheed, A.,.Pohlmann, R., Hasilik, A., von Figura, K., van Elsen, A. and Leroy, J.G. (1982) Biochem. BiOphys. Res. Commun. 155, 1052-1058. McKusick, V.A., Kelly, T.E. and Neufeld, E.F. (1978) The Metat lic Basis of Inherited Disease. New York: MCGraw-Hill, 1282-1307. Baumbach, G.A., Saunders, P.T.K., Bazer, F.W. and Roberts, R.M. (1983) Submitted for publication. Forbes, J.T., Bretthauer, R.K. and Oeltmann, T.N. (1981) Proc. Natl. Acad. Sci. USA 15, 5797-5801. Gottesman, M.M. (1978) Proc. Natl. Acad. Sci. USA 15, 2767-2771. Sahagian, G.G. and Gottesman, M.M. (1982) J. Biol. Chem. 251, 1145-1150. Roff, C.F., Wozniak, R.W., Blenis, J. and Wang, J.L. (1983) Exp. Cell Res., 155, 333-344. Roff, C.F. and Wozniak, R.W. (1982) Extracellular Matrix. Ed. Hawkes, S. and Wang, J.L., New York: Academic Press, 329-333. Jakoi, E.R., Zampighi, G. and Robertson, J.D. (1976) J. Cell Biol. 15, 97-111. Jakoi, E.R. and Marchase, R.B. J(l979) J. Cell Biol. 55, 642-650. Jakoi, E.R. and Corley, R.B. (1979) J. Cell Biol. 55, 65a. Koro, L.A. and Marchase, R.B. (1982) Cell 51, 739-748. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 30 Gaston, S.H., Marchase, R.B. and Jakoi, E.R. (1982) J. of Cell Biochem. 15, 447-459. McClay, D.R. and Moscona, A.A. (1974) Exp. Cell Res. 51, 438-441. Marchase, R.B., Harges, P. and Jakoi (1981) Rev. Biol. 55, 250-255. Schubert, 0., LaCorbiere, M., Klier, G. and Birdwell, C. (1983) J. Cell Biol. 25, 990-998. Teichberg, V.I., Silman, 1., Beitsch, 0.0. and Resheff, G. (1975) Proc. Natl. Acad. Sci. USA 1g, 1383-1387. Dewaard, A., Hickman, S. and Kornfeld, S. (1976) J. Biol. Chem. .551, 7581-7587. Beyer, E.C., Zweig, S.E. and Barondes, S.H. (1980) J. Biol. Chem. 255, 4236-4239. Nowak, T.P., Kobiler, 0., Roel, L.E. and Barondes, S.H. (1977) J. Biol. Chem. 252, 6026-6030. Gartner, T.K. and Podleski, T.R. (1975) Biochem. Biophys. Res. Comun. _§_7_, 972-978. MacBride, R.B. and Przybylski, R.J. (1980) J. Cell Biol. 55, 617-625. Den, H. and Chin, J.H. (l981) J. Biol. Chem. 255, 8069-8073. Barondes, S.H. and Haywood-Reid, R.L. (1981) J. Cell Biol. 51, 568-572. Beyer, E.C. and Barondes, S.H. (1982) J. Cell Biol. 55, 28-33. Beyer, E.C. and Barondes, S.H. (1982) J. Cell Biol. 5;, 23-27. Harrison, F.L. and Chesterton, C.J. (1980) Nature 555, 502-504. Skutelsky, E. and Bayer, E.A. (1983) J. Cell Biol. 25, 184-190. 76. 77. 78. 79. 80. 81. 82. 83. 84. 31 Dazzo, F.B. (1980) The Cell Surface: Mediator of Developmental Processes, Ed. S. Substelny, N.K. Nessels, New York: Academic Press, 277-304. Rosen, 5.0., Kafka, J.A., Simpson, D.L. and Barondes, S.H. (1973) Proc. Natl. Acad Sci. USA 15, 2554-2557. Simpson, D.L., Rosen, 5.0. and Barondes, SH. (1974) Biochem. 15, 3487-3493. Frazier, w.A., Rosen, S.0., Reitherman, R.M. and Barondes, S.H. (1975) J. Biol. Chem. 555, 7714-7721. Barondes, S.H., C00per, 0.N. and Haywood-Reid, P.L. (1983) J. Cell Biol. 55, 291-296. Ray, J., Schinnick, T. and Lerner, R.A. (1979) Nature 515, 215-221. Vacquier, v.0. and Moy, G.N. (1977) Proc. Natl. Acad. Sci. USA 15, 2456-2460. Natraj, C.V. and Datta, P. (1978) Proc. Natl. Acad. Sci. USA Z5, 6115-6119. Foecking, N.K., Otto, A.M. and Jimenez de Asua, L. (1983) Proceedings of Thirteenth Congress of Chemotherapy, in press. Chapter II ENDOGENOUS LECTINS FROM CULTURED CELLS 1. Isolation and characterization of carbohydrate - binding proteins from 3T3 fibroblasts* Calvin F. Roff and John L. Wang Department of Biochemistry Michigan State University East Lansing, MI 48824 Running Title: Carbohydrate-Binding Proteins from 3T3 Cells 32 SUMMARY Extracts of cultured 3T3 fibroblasts, obtained by homogenization and Triton X-100 solubilization, were fractionated on Sepharose columns covalently derivatized with asialofetuin. Three distinct carboyhdrate- binding proteins (CBPs) were purified from the material bound to the affinity column: CBP35 (Mr=35,030}, CBP16 (Mr=16,000); and CBP13.5 (Mr=13,500). These CBPs were similar in several key properties: (a) they showed agglutination activity when assayed with rabbit erythrocytes; (b) they all appear to specifically recognize galactose- containing glycoconjugates; (c) they have low isoelectric points, pIs 4.5-4.7; (d) their binding activities are rapidly lost in the absence of s-mercaptoethanol; (e) the CBPs do not interact with each other and the fractionated proteins can bind to asialofetuin independent of associated polypeptides; and (f) none of the proteins tend to self- associate to form oligomers of identical subunits. Comparisons of these and other properties of the CBPs suggest that CBP16 and CBPl3.5 may be the murine counterparts of lactose-specific lectins previously identified in electric eel and in several bovine and avian tissues. In contrast, it appears that CBP35 represents a newly identified protein capable of binding to galactose-containing carbohydrates. 3.3 34 The purification of carbohydrate-binding proteins (CBPs) , including lectins and enzymes such as glycosyl transferases and glycosidases, has made much use of the powerful technique of affinity chromatography. Because of the similarity of saccharide structures found on various serum glycoproteins such as fetuin and those found on the cell surface (1-3), CBPs can potentially recognize similar monosac- charide units, oligosaccharide structures, or the entire carbohydrate complex on glycoproteins and on cell surface heterosaccharides. This suggests that affinity columns containing Sepharose covalently coupled to a glycoprotein such as fetuin might be used for the isolation of CBPs. Indeed, this approach has been successfully applied in the purification of the hemagglutinin receptor of influenza virus (4), carbohydrate-specific antibodies (5), as well as lectins from both plant (6) and animal (7,8) sources. We have undertaken a study of CBPs from an established tissue culture cell line, Swiss 3T3 fibroblasts. This cell line has well- defined growth and morphological characteristics (9, 10). Because it is thought to be derived originally from mouse embryo fibroblasts, 3T3 cells presumably represent cells of a rather ubiquitous distribution. In the present communication, we report the purification and character- ization of three distinct CBPs, all of which were isolated on the basis of their binding to asialofetuin (ASF) - Sepharose and subsequent elution with the disaccharide, lactose. The CBPs exhibited agglutina- tion activity for rabbit erythrocytes and therefore, are probably fibroblast lectins. EXPERIMENTAL PROCEDURES Materials - Swiss 3T3 cells were obtained from American Type Culture Collection (CCL92). Dulbecco modified Eagle‘s medium (DMEM) was from K.C. Biologicals, calf serum from M.A. Bioproducts and fetuin from Gibco. All carbohydrates, cyanogen bromide and phenyl methyl sulfonylfluoride (PMSF) were products of Sigma, Aquacide III of Calbiochem, and Sepharose 4B and Sephadex G-150 of Pharmacia. [35$]Methionine (1012 Ci/mmol) was bought from New England Nuclear. Ampholines were purchased from LKB. Culture and Radiolabeling of 3T3 Cells - Maintenance of Swiss 3T3 cells has been described elsewhere (11). Swiss 3T3 cells were grown to confluent monolayers (4-5 x 104 cells/cmz). The medium was replaced with fresh growth median for 24h. This medium was removed and the cells were cultured in serum-free DMEM (10 m1/150 cm2 growth area) containing 3 ug/ml unlabeled methionine (one-tenth of the concentration normally found in DMEM) and 20 uCi/ml [35$]methionine (12). After 24 h, the medium was removed and the cells were washed prior to the extraction and isolation of the proteins. Preparation of Asialofetuin - Sepharose - Fetuin was desialylated as described by De Haard.ggngl. (13) and coupled to Sepharose 48 by the method of Cuatrecasas (14). Fetuin (500 mg) dissolved in 25 m1 H20 (pH 2.0) was heated at 80° for 1h. The solution was then cooled to 35 36 25°, neutralized with NaOH, and dialyzed against 0.2 M NaHCO3, pH 7.9. The ASF was coupled to 150 ml of CNBr (20 g) activated Sepharose 4B in a combined volume of 300 ml of 0.2 M NaHC03, pH 7.9. After 24 h at 4°, 150 ml of 2 M ethanolamine, pH 8.0, was added for an addition- al 24 h. The resin was washed extensively with 1 M NaCl and then washed with Buffer A (see below). Routinely, greater than 80% of the ASF was coupled as determined by the difference in absorbance at 280 nm of the ASF solution before and after the coupling reaction. Isolation of Asialofetuin Binding Proteins - The purification of ASF-binding proteins used the following buffers: Buffer A - 50 mM CaClz, 1 mM NaN3, 75 mM Tris(hydroxymethyl)aminomethane, pH 7.2; Buffer B - Buffer A supplemented with 1% Triton X-100 and 1 mM PMSF; and Buffer C - Buffer A containing 2 mM e-mercaptoethanol. Extracts of 3T3 cells were prepared from [35$]methionine- labeled monolayers (in 150 cm2 flasks) by decanting the labeling median and washing with 15 ml of Buffer A containing 1 m PMSF. The cells in each flask were scraped with a rubber policeman into 2 ml Buffer B. The pooled cellular material was homogenized in a 2 m1 Potter homogenizer (102-152 um clearance) at five strokes/2 ml. Insoluble material was pelleted by centrifugation at 3,000 x g for 15 min. and then the supernatant was cooled to 4°. All previous Operations were performed at 25° and subsequent steps at 4°. The supernatant was applied to an ASF-Sepharose column (1.4 x 15 cm) and the column was washed extensively with Buffer B containing 2 mM 8 - mercaptoethanol. To remove detergent, the column was washed with 2-4 column volumes of Buffer C. Protein bound on the ASF-Sepharose column was eluted with a 0-0.15 M lactose gradient (100 ml total 37 volume). Aliquots from the column effluent were assayed for radioactivity due to [35$]methionine-labeled proteins using scintillation counting (12). The material eluted by lactose was pooled and dialyzed against Buffer C in tubing impermeable to molecules of molecular weight greater than 3,500. The dialysis tubing also contained ASF-Sepharose. The dialyzed contents were poured into a column and the resin was allowed to settle before washing with Buffer C. The bound proteins were eluted with a lactose gradient as described above. The material eluted with lactose from the second affinity column was concentrated by reverse dialysis against Aquacide III to a final volume of l-2 ml. This concentrate was chromatographed on a column (1 x 150 cm) of Sephadex G-150. Fractions from the Sephadex G-150 column were pooled and tested for their ability to rebind to another ASF-Sepharose column. Gel Electrophoretic Characterization of CBPs - Polyacrylamide gel electrophoresis in sodium dodecyl sulfate (SDS) was performed according to the procedure of Laemmli (15) on a 1 mm thick, 9 cm long, 5-16% gradient slab gel (.21-.67% bisacrylamide) with a 1 cm long 4% stacking gel. Samples were prepared by dialysis against water followed by lyophilization. They were dissolved in 1% SDS, 4% B-mercaptoethanol and boiled for 1 minute. After electrOphoresis the gels were fixed for 30 min in 10% trichloroacetic acid and stained with Coomassie Brilliant Blue. After destaining the gel was subjected to fluorographic treatment as described by Bonner and Laskey (16), using Kodak X-Omat AR (XAR-S) film. 38 Two-dimensional gel electrOphoretic analysis was performed according to the method of 0'Farrell (17). Samples were first subjected to isoelectric focusing in 1 mm X 10 cm tube gels containing pH 3-10 ampholines. The second dimension was electrOphoresed on 5-16% polyacrylamide slabs as described above. Assays of Agglutination and Enzymatic Activities Fresh rabbit erythrocytes were isolated following the method of Lis and Sharon (18) and trypsin-treated, glutaraldehyde-fixed rabbit erythrocytes were prepared by the method of Nowak et al. (19). The cells were used as a 4% stock su5pension in 0.9% NaCl containing 0.3% bovine serum albumin (pH 7.4). Hemagglutination assays were carried out in microtiter V-plates; each well contained 25 ul of erythrocyte suspension and 25 ul of the test sample. To study the effects of saccharides on hemagglutination, 10 ul of a stock solution of saccharide in 0.9% NaCl was added; control wells received 10 ul of 0.9% NaCl. In addition, the effect of the various saccharides on the erythrocytes were tested in the absence of any agglutinin sample. All agglutination assays were scored after 1 h at room temperature. B-Galactosidase activity was determined by the method of Bishop and Desnick (20). To 150 pl of 1.5 mM 4-methylumbelliferyl-B-D-galacto- pyranoside (Pierce Chemical) in 0.03 M citrate, 0.05 M phosphate, pH 4.6, was added 50 ul of the test sample. After incubation for l h at 37°, the reaction was terminated by the addition of 2.4 ml of 0.1 M ethylenediamine. Fluorescence was monitored on a Perkin-Elmer 650-40 fluorimeter using excitation and emission wavelenghts of 360 nm and 440 nm, respectively. s-Galactosidase activity was also determined at 39 neutral pH using 1.5 mM 4-methylumbelliferyl-s-D-galact0pyanoside in Buffer B. Sialyl transferase activity was determined by the method described by Bosmann (21) using ASF as a potential acceptor. To a solution of 10 mM MgC12, 10 mM MnClz, 400 ug ASF and 8.6 x 104 dpm of cytidine 5'-monophospho-N-acetyl-[4,5,6,7,8,9-]4c]-neuraminic acid (Amer- sham, 247 mCi/mmol) in 100 pl of Buffer B was added 50 ul of test sample. After 4 h at 37°C, an aliquot was removed and 5 volumes of cold 1% phosphotungstic acid in 0.5 N HCl was added. The mixture was centrifuged and the precipitate was washed twice with 1% phosphotung- stic acid in 0.5 N HCl, resuspended in 0.5 ml H20 and neutralized with l N NaOH. The radioactivity was then determined by scintillation counting. Alternatively, the reaction mixture was chromatographed on columns (100 x 1.5 cm) of Sephadex G-25 and the effluent fractions were monitored for radioactivity. The test sample has also been assayed for transferase activity in the presence of 2 mM unlabeled CMP-sialic acid. Preparation of antisera and ImmunOprecipitation Antisera directed against CBP 35 were raised in New Zealand White female rabbits. CBP35 isolated from 17 flasks (150 cm of confluent 3T3 fibroblasts) was mixed in 200 pl complete Freunds adjuvant and injected near the papliteal lymph node (VIOO ul/node) of an etherized rabbit. After 10 days, the rabbit was injected at the same site with CBP35 (isolated from 10 flasks of confluent 3T3s) which was suspended in incomplete Freunds adjuvant. The rabbit was bled 10 days after the second immunization. Subsequent bleedings were also made 10 days after boosting the rabbit. Antiserun against chicken lactose lectin I (CLL 40 I) was prepared by injecting CLL I (Mr = 16,000) isolated from female adult chicken liver into a rabbit (22). It was a gift of Dr. Steven Ullrich (Michigan Molecular Institute, Midland, MI). Material used for the immunoprecipitation was 35s-methionine labeled 3T3 cell extract partially purified by affinity chromatography over one ASFSepharose column. The material was concentrated by reverse dialysis against Aquacide III to a volume of approximately 2 ml. The concentrated material was Split into 4 aliquots, 450 ul per tube, placed on ice for 1 hour, after which it was centrifuged for 15 minutes at 12,000 x g. The supernatants were transferred to new tubes and 5 ul of antisera was added. After incubation at 37°C for 1 hour, they were incubated at 4°C for 8-12 hours. Then, 150 ml of goat anti-rabbit IgG serum (Gibco) was added to each sample and they were placed at 4° for an additional 14 h. The precipitates were pelleted by centrifugation at 10,000 x g for 15 min. The supernatant was discarded and the pellet was washed twice in .05 M TriSHCl, 1.2 M KCl, 1% (v/v) Triton X-100, pH 7.4, followed by two washings in .05 M Tris-HCl, .1 M NaCl, pH 7.4, and finally with water. The pellet was dissolved in 100 pl of buffer for polyacrylamide gel electrOphoresis, and boiled for 2 minutes prior to electrophoresis on 10% polyacrylamide gels. RESULTS Asialofetuin-Binding,Proteins from 3T3 Cells - 3T3 fibroblasts were cultured in the presence of [35$]methionine to label the cellular proteins. After washing, confluent monolayers of these labeled cells were extracted with Triton X-100 and fractionated by affinity chromato- graphy on a column of ASF-Sepharose (Fig. l). The majority of the radioactive material was not bound by the column (Component A, Fig. 1). After extensive washing in buffer containing Triton X-100, the column was further developed with detergent free buffer (position of arrow 1, Fig. 1). Finally, the column was eluted with a linear gradient of lactose (position of arrow 2, Fig. 1), which resulted in the appearance of a peak of radioactivity (Component C, Fig. l). The radioactivity in Component C (Fig. l) accounted for < .01% of the total radioactivity applied to the affinity column. Polyacrylamide gel electrOphoretic analysis in SDS was carried out on Component C (Fig. 1), as well as on 35S-labeled material from pooled fractions immediately before (Component B, Fig. l) and immedi- ately after (Component 0, Fig. 1) the radioactive peak. Component 8 (Fig. 1) yielded a heterogeneous mixture of polypeptides on $05 gel analysis (lane b, Fig. 2). In contrast, Component C (Fig. 1) yielded three predominant bands, corresponding to molecular weights of 35,000, 16,000, and 13,500 (lane c, Fig. 2). Several other bands were notice- able; they corresponded to molecular weights of 30,000, 20,000, 11,000 41 42 Figure 1. Affinity chromatography of [35s]methionine-labeled, Triton x-100 solubilized, extracts of 3T3 fibroblasts on a column (1.4 x 15 cm) of asialofetuin-Sepharose. The column was equilibrated with Buffer B and was eluted as described in Materials and Methods. The arrows mark the fractions at which the buffers were changed: 1, buffer C; and 2, 0-0.15 M lactose gradient (100 ml total volume). The gradi- ent is marked (---) so as to indicate the concentration of lactose at the top of the resin bed. Fractions (2.5 ml) were collected and ali- quots (0.2 ml) were assayed for radioactivity. 43 .28 @883. . :50 F rochon 100 50 Figure l 44 Figure 2. Polyacrylamide gel electrOphoresis in sodium dodecyl sulfate of fractions derived from affinity chromatography of [355]- methionine-labeled extracts of 3T3 cells on asialofetuin (ASF)-Sepha- rose columns. Effluent fractions from the first ASF-Sepharose column: lane a, component A of Fig. 1; lane b, component B of Fig. 1; lane c, component C of Fig. 1; and lane d, component 0 of Fig. 1. Effluents fron rechramatography of component C (Fig. 1) on a second ASF-Sepharose column: lane e, component A of Fig. 3; and lane f, component B of Fig. 3. Lane 9 consists of material corresponding to Component 8 of Fig. 3 and shows the Mr 34,000 polypeptide which occurs in some preparations. Radioactivity in the samples: lane a, 40,000 cpm; lanes b-f, 10,000 cpm; and lane 9, 25,000 cpm. The fluorogram was exposed at -70° for 72 hours. 45 220' 68> ‘ 40’ 12.5> 46 and (10,000. Finally, Component 0 (Fig. l), which consisted of material eluted as a shoulder of the main radioactive peak, yielded a gel pattern that contained at least four prominent bands (Mrs of 100,000, 36,000, 16,000 and 11,000), as well as many minor contaminants. Component C (Fig. 1) can be further purified by another cycle of affinity chromatography. The pooled material was dialyzed to remove the lactose. Two important requirements were noted concerning the dialysis: (a) the presence of s-mercaptoethanol (2 mM) preserved the ASF-binding capacity of the proteins; in its absence, the binding activity of the preparation was completely lost; (0) the inclusion of the affinity matrix (ASF-Sepharose) in the dialysis bag prevented the nonspecific absorption of 35s-labeled proteins to the tubing and therefore, increased the recovery of material after dialysis. After dialysis, the contents of the bag was packed directly into a column and washed (Fig. 3). Approximately 20% of the dialyzed material did not bind to ASF-Sepharose (Component A, Fig. 3). The bound material (Component B, Fig. 3) was eluted with lactose at a position similar to that found in the first affinity column (20 mM lactose). Molecular Heights and Isoelectric Points of Asialofetuin-Binding Proteins- Polyacrylamide gel electrophoretic analysis in $05 of Component B (Fig. 3) yielded three polypeptide bands, with molecular weights of 35,000, 16,000, and 13,500 (lane f, Fig. 2). The material corresponding to these three bands will be referred to hereafter as CBP35, CBP16, and CBPl3.5. Densitometric tracing of the fluorogram (lane f, Fig. 2) showed that these three bands accounted for > 99.9% of the total radioactivity and that the relative proportions of the 47 Figure 3. Rechromatography of asialofetuin-binding proteins on a second ASF-Sepharose column. Material eluted with lactose fron an ASF-Sepharose column (component C, Fig. 1) was applied to a second ASF-Sepharose column (1.8 x 5 cm), washed with Buffer C, and eluted with a lactose gradient (0-0.15 M, ---). Fractions (2.1 ml) were collected and aliquots (0.2 ml) were assayed fbr radioactivity. 48 .6Cl ‘40 Fraction 20 Figure 3 49 individual bands were: 2.9 (CBP35): 1.1 (CBP16): 1(CBP13.5). Similar results were obtained both in the presence and absence of B-mercaptoe- thanol during the electrOphoresis. In some preparations, the material corresponding to Component B (Fig. 3) yielded a fourth band on polyacrylamide gels (lane 9, Fig. 2). It accounted for no more than 2-3% of the total radioactivity on the gel. The molecular weight of this fourth band was estimated to be 34,000. Although the origin of this band is not known at present, this polypeptide also has the capacity to bind to ASF. In order to determine the isoelectric properties of the polypeptides corresponding to Component B (Fig. 3), a sample containing only CBP35, CBP16, and CBP13.5 was subjected to two dimensional gel electrophoretic analysis (Fig. 4). The results showed that CBP16 and CBP13.5 had similar pI's of approximately 4.5. CBP35, which corre- Sponded to a single band on one dimensional electrOphoresis, yielded two Spots that had the same molecular weight (Mr = 35,000) but different isoelectric points (pI's 4.5 and 4.7) (Fig. 4). These results indicate that the ASF-Sepharose column could be used to isolate a minimum of three, and possibly four, polypeptides from 3T3 cells. Effect of Saccharide Ligands and EDTA on Asialofetuin-Binding_ Proteins - In order to probe the sugar-binding specificity of the ASF-binding proteins, Triton X-100 extracts of 3T3 cells were chromatographed on columns of ASF-Sepharose. Various saccharides were tested for their capacity to elute the bound radioactive polypeptides. When the column was develOped sequentially with mannose (position of arrow 2) sucrose (position of arrow 3) and lactose (position of arrow 4), a prominent radioactive peak was observed only upon the addition of 50 Figure 4. Two dimensional gel electrOphoretic analysis of CBP35, CBP16, and CBP13.5. Material eluted with lactose (component B, Fig. 3; lane f, Fig. 2) was subjected to isoelectric focusing in the first dimension and sodium dodecyl sulfate polyacrylamide electrophoresis in the second dimension. Approximately 8,000 cpm were electrOphoresed and the fluorogram was exposed for 16 days. 44 ii 4.8 v 51 PH 5.6 6.0 8,4 6.8 v 468 MW «40 412.5 52 lactose (Fig. 5a). Polacrylamide gel analysis in $05 of Components A, B, and C (Fig. 5a) showed that CBP35, CBP16 and CBP13.5 all were eluted with lactose (lanes a-c, Fig. 6). When the ASF-Sepharose column was develOped sequentially with N-acetyl-Dglucosamine (position of arrow 2), galactose (position of arrow 3), and lactose (position of arrow 4), no radioactive material was eluted with the first monosaccharide (Fig. 5b). A peak of radioactivity was eluted, however, upon the addition of galactose (Component B, Fig. 5b). Polyacrylamide gel analysis in $08 of this material yielded two predominant polypeptides, corresponding to CBP35 and CBP13.5 (lane e, Fig. 6). When lactose was used to develOpe the column after galactose, some radioactivity was eluted in a rather ill-defined peak (Component C, Fig. 5b). This material yielded CBP16 upon SDS gel electrOphoresis (lane f, Fig. 6). These results indicate that the ASF-binding polypeptides are carbohydrate-binding proteins (CBPS). The question arose whether any of the CBPS had an intrinsic requirement for Ca2+ ion in order to bind the saccharide. To test this, the CBPS bound on ASF-Sepharose were eluted with Ca2+ free buffer containing 10 M EDTA (position of arrow 1, Fig. 7). No distinct peak of radioactivity was observed (component B, Fig. 7). Moreover, SDS gel analysis of Component B (Fig. 7) revealed trace amounts of CBP35, CBP16, and CBP13.5. There was no apparent enrichment of any one of the polypeptides relative to the other two. In contrast, the addition of lactose to the ASF-Sepharose column after EDTA (position of arrow 2, Fig. 7) yielded a substantial peak of radioactivity (Component C, Fig. 7). All three of the ASF-binding 53 Figure 5. Effect of various saccharides on the elution of CBP35, CBP16, and CBP13.5 from ASF-Sepharose. An extract of [35$]methionine labeled cells was divided into two aliquots, each of which was applied to an ASF-Sepharose column (1.2 x 7 cm). After washing, the column was eluted sequentially with mono- and disaccharides of the D-configura- tion. Fractions (2 ml) were collected and aliquots (0.5 ml) were assayed for 35S-radioactivity. The arrows indicate changes in car- bohydrate (0.15 M) included in the develOping buffer. Panel a: l, Buffer C; 2, mannose; 3, sucrose; and 4, lactose. Panel b: l, Buffer C; 2, N-acetylglucosamine; 3, galactose; and 4, lactose. 54 2 i 1 I l 1 0 4O 80 I20 I60 Frociion Figure 5 55 Figure 6. Polyacrylamide gel electrOphoresis in sodium dodecyl sulfate of fractions derived from affinity chromatography of extracts of 3T3 cells on asialofetuin-Sepharose columns eluted with various saccharides (Fig. 5). Fractions fron panel a, Fig. 5: lane a, component A; lane b, component B; and lane c, component C. Fractions from panel b, Fig. 5: lane d, component A; lane e, component B, and lane f, component C. Approximately 4,000 cpms were applied to each lane and the fluorograms were exposed for 30 days (lanes a-c) or 45 days (lanes d-f). 56 57 polypeptides (CBP35, CBP16, and CBP13.5) were found in Component C (Fig. 7). It appears, therefore, that none of the CBPS require Ca2+ ions for saccharide-binding. This conclusion is further cor- roborated by experiments in which the same three CBPs were isolated from the ASF-Sepharose column when the extraction and affinity chroma- tography were carried out in calcium free buffer (Buffer A was replaced by 75 mM NaCl, 2 mM ethyleneglycol-bis-(8-aminoethyl ether)N,N'-tetra- acetic acid, 2 mM NaN3 and 75 mM Tris(hydroxymethyl)aminomethane, pH 7.0). Fractionation of the Carbohydrate-Binding Proteins - Gel filtration of the CBPS purified by two cycles of affinity chromatography on Sephadex G-150 further fractionated the CBPS into two new components (Fig. 8). Component A (Fig. 8) chromatographed to a region corresponding to molecular weights of 30,000-35,000. Upon gel electrOphoresis in 505, it yielded only CBP35 (lanes b and d, Fig. 9). (In the sample used for Fig. 9, the gel also shows a minor band corresponding to a molecular weight of about 34,000; this represents the fourth band of the CBPS that occurs in some preparations.) Gel electrophoresis in $05 of Component B (Fig. 8) showed that it consisted of only CBP16 and CBP13.5 (lane c and e, Fig. 9). No CBP35 was observed in these fractions. Similar results were obtained both in the presence as well as in the absence of lactose (Fig. 8 and 9). These results indicate that CBP35 does not interact with either CBP16 or CBP13.5. In addition, the position of migration of CBP35 (Component A, Fig. 8) suggests that the molecular weight of the protein in the absence of denaturants is 35,000 and therefore the polypeptide does not self-associate intodimers or oligomers. Similarly, the 58 Figure 7. Effect of EDTA on the elution of CBP35, CBP16 and CBP13.5 bound to asialofetuin-Sepharose. An extract of [355] methionine-labeled 3T3 cells was chromatographed on an ASF-Sepharose column (1.3 x 14 cm) in Buffer C. After nonbound material was collect- pd, the column was eluted with three column volumes of calciun free buffer containing EDTA (75 mM Tris, pH 7.2, 10 mM EDTA, 60 mM NaCl, and lmM NaN3), followed by Buffer C containing 0.15 M lactose. The arrows indicate changes in developing buffer: 1, EDTA and 2, lactose. Fractions (1.5 ml) were collected and aliquots (0.4 ml) assayed fbr radioactivity. 59 e—N L L B J C I I J l 1 1 20 4O 60 80 100 120 Frociion Figure 7 60 Figure 8. Sephadex G-150 chromatography of [35$]methionine- labeled CBP35, CBP16, and CBP13.5. Purified material fran the second ASF-Sepharose colunn was concentrated and chromatographed on a Sephadex G-150 colunn (0.9 x 120 cm) in Buffer C either in the presence (H) or absence (o--o) of 0.1 M lactose. Fractions (1.1 ml) were collected and aliquots (0.1 ml) were assayed for radioactivity. Molecular weight markers were: A_l_d_, aldolase (158,000); _B_S_A, bovine serun albunin (68,000); M, chymotrypsinogen A (25,000); and 5_Lt_C, cytochrane C (12,500). cpm x 10 61 Ald BSA Chym Cyi C l l 1 4 A B 1-—l l—l J 60 Fraction Figure 8 62 Figure 9. Polyacylamide gel electrOphoresis in sodium dodecyl sulfate of highly purified CBP35, CBP16, and CBP13.5 after gel filtration on Sephadex G-150 (Fig. 8). Lane a is an aliquot of the sample which was applied to the column. Gel filtration in the absence of lactose: lane b, component A and lane c, component B, and in the presence of lactose: lane d, component A and lane e, component B. Aproximately 25,000 cpm were applied and the fluorograms were exposed for 48 hours. 63 220» 68> 40> 64 chromatographic position of Component B (Fig. 8) is consistent with polypeptides of molecular weight 13,000-16,000. Moreover, we have achieved a similar fractionation using columns of Sephadex G-50. Under this condition, the chromatographic position of CBP16 and CBP13.5 was well resolved from that of chymotrypsinogen (Mr = 25,000). Therefore, it appears that CBP16 and CBP13.5 do not bind to each other in non-denaturing solvents. Intrinsic Binding5Properties of the Carbohydrate-Binding Proteins - Component A (Fig. 8) was applied to an ASF-Sepharose col unn. More than 95% of the radioactivity applied was bound to the affinity column and could be eluted with lactose (Fig. 10). Polacrylamide gel electro- phoresis in $05 of the recovered material showed that it was highly This suggests that CBP35 can bind to purified CBP35 (inset Fig. 10). Moreover, isolated carbohydrates independently of CBP16 and CBP13.5. CBP35 can agglutinate rabbit erythrocytes as well as rabbit erythro- cytes previously treated with trypsin followed by glutaraldehyde fixa- tion (Table I). This agglutination was inhibited by lactose (0.05 M). Component B (Fig. 8), which consisted of CBP16 and CBP13.5, also exhibited lactose-inhibitable agglutination activity (Table 1). Component B (Fig. 8) can be bound to ASF-Sepharose columns and eluted with specific saccharides. Elution with galactose yielded a fraction (Component A, Fig. 11) containing only CBP13.5 (see inset, Fig. 11). sUbsequent elution with lactose yielded a fraction (Component C, Fig. 1 1 ) containing CBP16, although this protein can also be detected at the 1 atter part (Component B, Fig. 11) of the galactose elution (0.15 M Qa‘l aetose). These results corroborate the previous demonstration that 9a} aCtose can elute CBP13.5 and lactose can e1 ute the bulk of the CBP16 65 TABLE I Agglutination and Enzymatic Activities of CBP35, CBP16 and CBP13.5 Assay Sample Agglutination B-Galactosidase Sialyl transferase Component C + - nt Fig. 1 Component B + - - Fig. 3 CBP35 + nt nt (Component A, Fig. 8) CBP16 and CBP13.5 + nt nt (Component B, F ig. 8) __¥ '*' activity detected - no activity detected "It not tested for activity 66 Figure 10. Binding of [35$]methionine labeled CBP35 to an ASF-Sepharose column. CBP35, fractionated on a Sephadex G-150 column (component A, Fig. 8) was bound on an ASF-Sepharose column (1.5 x 8 cm). The arrow marks the addition of lactose (0.15 M) in the develop- ing buffer. Fractions (0.9 ml) were collected and aliquots (0.2 ml) were assayed for radioactivity. The inset Shows the fluorogran of Component A from this column after polyacrylamide gel electrOphoresis in sodiun dodecyl sulfate. cpm x IO"2 67 l l 20 4O Fraction 68 Figure 11. Binding of [35$]methionine labeled CBP16 and CBP13.5 to an ASF-Sepharose column. Component B (Fig. 8) was affinity chromatographed on an ASF-Sepharose column (1.5 x 8 cm). The arrows indicate changes in carbohydrate (0.15 M) included in the developing buffer: 1, galactose; 2, lactose. Fractions (0.9 ml) were collected and aliquots (0.2 ml) were assayed for radioactivity. The inset Shows the fluorograms of pooled fractions corresponding to Components A-C from this column after polyacrylamide gel electrOphoresis in sodiun dodecyl sulfate. 69 ZIIV lb Fraction 70 from ASF affinity columns (Fig. 5b). Together, they suggest that both CBP16 and CBP13.5 have intrinsic carbohydrate-binding capacities, independent of other associated polypeptides. We have also tested the various fractions containing the CBPS for enzymatic activities such as glycosidases and transferases. Using 4-methylumbelliferyl- a-D-galactOpyranoside as a substrate, we found no e-galactosidase activity associated with the CBPS (Component C, Fig. l and Component 8, Fig. 3) at pH 4.6 and at pH 7.2. Moreover, all of the B-galactosidase activity observed in the original cell extract could be accounted for in the material not bound by the first ASF-Sepharose column (Component A, Fig. 1). Similarly, we found no transferase activity associated with the CBP fractions (Component B, Fig. 3) as assayed with ‘4C-labeled CMP-sialic acid and ASF. This conclusion is based on experiments which showed that the radioactivity precipitat- ed along with ASF by phOSphotungstic acid was identical when the 'transferase assay was carried out in the presence and absence of the (ZBPs. Moreover, analysis of the assay reaction mixture by chromato- sgraphy on Sepahdex G-25 showed that no radioactivity migrated in the \roid volume fractions, at a position corresponding to ASF. .ijmunoprecipitation of the Carbohydrate-Binding Proteins - The iavailability of two antisera, one directed against CBP35 (anti-CBP35) iand the other directed against CLL I (anti-CLL I), provided the I1ecessary reagents to study the structural relationships between CBP35, (2BP16, and CBP13.5, as well as to test for relatedness to lectins i<:haracterized in the chicken intestine and muscle systems. Component C (Fig. 1), which contained a mixture of CBP35, CBP16 and CBP13.5, was asubjected to immunoprecipitation by anti-CBP35 and anti-CLL I. 71 .Analysis of the immunOprecipitates by gel electrOphoresis showed that anti-CBP35 reacted only with CBP35 but not with CBP16 and CBP13.5 (Fig. 12, lane b). In contrast, anti-CLL I precipitated only CBP 16 and no other CBP from the mixture (Fig. 12, lane c). These results suggest that CBP35 is not structurally related and therefore, is most probably not a higher molecular weight precursor of CBP16. In addition, the immune reactivity of CBP16 with anti-CLL I indicates that this CBP is probably a murine fibroblast counterpart of the lactose-specific lectins described in a number of other systems. 72 Figure 12. Polyacrylamide gel electrOphoresis in sodium dodecyl sulfate of CBPS identified by specific antisera. (a) Partially purified CBPS (Component C, Fig. 1) subjected to immunOprecipitation; (b) Material immunOprecipitated by antibodies directed against CBP35; and (c) Material immunOprecipitated by antibodies directed against chicken lactose lectin 1. Approximately 1000 cpms were applied to each lane and the fluorograms were exposed far 30 days. 73 DISCUSSION The results obtained in the present study indicate that we have purified, from the 3T3 fibroblast system, three distinct CBPS: (a) CBP35 (M, = 35,000); (b) CBP16 (M, = 16,000); and (c) CBP13.5 (M, = 13,500). These three proteins are similar in several key properties. First, they all appear to recognize specifically galactose-containing gglycoconjugates. In the present paper, we have isolated them on the t>asis of their binding to columns containing the asialoglyc0protein, ASF and we have demonstrated differences between them on the basis of their elution from the affinity column using lactose or galactose. We fLave also obtained evidence that these proteins will bind polyacryl- amide beads derivatized with disaccharide ligands of defined structure (DGa18(l 4)BDGlcNAc) but not to beads containing only the monosaccha- ride aDGal (23). CBP 13.5 exhibited agglutination activity when assayed with rabbit These Moreover, fractions containing CBP 35 or CBP l6 and Erythrocytes. This agglutination can be inhibited by lactose. t>‘iruding and agglutinating results demonstrate unequivocally that the ‘5 Sol ated proteins are carbohydrate-binding proteins. Second, these CBPS have low isoelectric points (pIs 4.5-4.7), 1 r'd‘icating that they are most probably acidic proteins. In our two affilensional gel elec- traphoretic analysis, CBP35 was actually resolved ““30 two different spots with p15 of 4.5 and 4.7. The relationship of 74 75 these two polypeptides, of the same molecular weight but of different isoelectric points, has not been determined. Third, the binding activity of all three of the CBPS was rapidly lost in the absence of B-mercaptoethanol. This suggests that function- al groups, most likely free sulfhydryl or tryptOphan residues (23), may be sensitive to air oxidation. Fourth, CBP35, CBP16, and CBP13.5 do not appear to require Ca2+ for activity. Finally, gel filtration studies in non-denaturing solvents indicate that the CBPS do not interact with each other, both in the presence and absence of lactose. The chromatographic data also suggest that none of the three polypeptides appear to self-associate to form oligomers of identical subunits. Because these column separation experiments were carried out using minute anounts of radiolabeled proteins, however, the concentrations of CBPS used in our column experiments may be below the threshold required for aggregation. In any case, it should be empha- sized that each of the fra tionated proteins can bind to ASF, indepen- dent of associated polypeptides. Comparisons of the polypeptide molecular weights, the isoelectric points, agglutination activity, and the carbohydrate-binding specific- ity of CBP16 and CBP13.5 with the lactose-specific lectins isolated fran electric eel organ (24), calf heart and lung (13), and chicken intestine (25) and muscle (26, 19) suggest that they may be related Proteins. This conclusion is supported by the observation that antibodies raised against CLL I (M, = 16,000) imuneprecipitated only CBP16 out of a mixture containing all three CBPs. In addition, these PPOteins all share the similar properties of being highly sensitive to 31" oxidation and of being calcium-independent. CBP16 and CBP13.5 76 differ from these lactose-specific lectins in one major respect. Whereas all but one of the previously identified lectins self-associated to form dimers and oligomers. CBP16 and CBP13.5 remain in monomeric form. If these proteins do indeed turn out to be analogous counterparts of each other in different species (27, 28), then the question concerning their postulated tissue-specific function(s) must be raised. To the best of our knowledge, a protein analogous to CBP35 has not been previously identified and isolated in other species or from other cell types. It does not appear that CBP35 is a higher molecular weight precursor to CBP16 and/or CBP13.S. Antibodies directed against CBP35 did not show cross reactivity with either CBP16 or CBP13.5. Conversely, antibodies that recognized CBP16 failed to react with CBP35. Therefore, CBP35 most probably represents a new carbohydrate-binding protein, co-isolated with CBP16 and CBP13.5, which do have analogous counterparts. At present, it does not appear that CBP35 is the fibroblast coun- terpart of the galactose-specific receptor on hepatocytes, originally identified by Ashwell and co-workers (29-33). The differences between the two proteins include: (a) molecular weight of the polypeptide chain; (b) aggregation properties of the polypeptides; (c) effect of EDTA on binding of carbohydrates; and (d) sensitivity to air oxidation. He have recently purified a carbohydrate-binding protein from mouse 1“"9 tissue using the same procedures described for the isolation of CBP35. The molecular weight of this protein was 35,000, as determined by SDS polyacrylamide gel electrOphoresis and Coomassie blue staining. 77 When the mouse lung protein on the polyacrylamide gel was transferred canto nitrocellulose paper and then immunoblotted with anti-CBP35, a single radioactive band (M, = 35,000) was observed after autoradio- graphy. These results suggest that we can isolate CBP35 in large amounts (microgram levels) from mouse lung. This in turn will allow us to carry out structural studies on the polypeptide, to study its cellular localization, and to search for its endogenous ligand in the cell. ll. ‘2. REFERENCES Spiro, R.G. (1973) Adv. Protein Chem. 55, 349-467. Kornfeld, R., Keller, J., Baenziger, J. and Kornfeld, S. (1971) J. Biol. Chem. 515, 3259-3268. Kornfeld, S. and Kornfeld, R. (1971) in "Glyc0proteins in Blood Cells and Plasma" (Jamieson, G.A. and Greenwalt, T.J., eds.) pp. 50-67. J.B. Lippincolt, Philadelphia. Scheid, A. and Choppin, P.H. (1974) Virology 55, 125-133. Sela, B.-A., Hang, J.L. and Edelman, G.M. (1975) Proc. Natl. Acad. Sci. USA 55, 1127-1131. Sela, B.-A., Hang, J.L. and Edelman, G.M. (1975) J. Biol. Chem. 555, 7535-7538. Barak-Briles, E., Gregory, H., Fletcher, P. and Kornfeld, S. (1979) J. Cell Biol. 55, 528-537. Beyer, E.C., Zweig, S.E. and Barondes, S.H. (1980) J. Biol. Chem. 555, 4236-4239. Todaro, G.J. and Green, H. (1963) J. Cell Biol. 15, 299-313. Todaro, G.J., Green, H., and Goldberg, 8.0. (1964) Proc. Natl. Acad. Sci. U.S.A. 51, 66-73. Steck, P.A., Voss, P.G., and Hang, J.L. (1979) J. Cell Biol. 55, 562-675. Steck, P.A., Blenis, J., Voss, P.G., and Hang, J.L. (1982) J. Cell Biol. 55, 523-530. 78 13. 14. 15. 15. 17. 18. 19. 20. 21. 22. 23. 24, 225. £26. 273 28L 25L 3(L, 79 0e Haard, A., Hickman, S. and Kornfeld, S. (1976) J. Biol. Chem. ggl, 7581-7587. Cuatrecasas, P. (1970) J. Biol. Chem. 555, 3059-3065. Laemmli, U.K. (1970) Nature (Lond.) 555, 680-685. Bonner, H.M., and Laskey, R.A. (1974) Eur. J. Biochem. 55, 83-88. 0'Farrell, P.H. (1975) J. Biol. Chem. 555, 4007-4021. Lis, H. and Sharon, N. (1972) Methods Enzymol. 55, 360-368. Nowak, T.P., Kobiler, 0., Roel, L.E. and Barondes, S.H. (1977) J. Biol. Chem. 555, 6026-6030. Bishop, D.F. and Desnick, R.J. (1981) J. Biol. Chem. 555, 1307-1316. Bosmann, B.H. (1972) Biochem. BiOphys. Res. Comm. 55, 523-529. Beyer, E.C. and Barondes, S.H. (1982) J. Cell Biol. 55, 23-27. Raff, C.F., Rosevear, P.R., Hang, J.L. and Barker, R. (1983) Biochem. J., 51, 625-629. Levi, G. and Teichberg, v.1. (1981) J. Biol. Chem. 555, 5735-5740. Beyer, E.C., Zweig, S.E. and Barondes, S.H. (1980) J. Biol. Chem. 555, 4236-4239. Den, H. and Malinzak, 0.A. (1977) J. Biol. Chem. 555, 5444-5448. Barondes, S.H. (1981) Ann. Rev. Biochem. 55, 207-231. Childs, R.A. and Feizi, T. (1979) Biochem. J. 155, 755-758. Ashwell, G. and Harford, J. (1982) Ann. Rev. Biochem. 51, 531-554. Stockert, R.J., Gartner, U., Morell, A.G., and Holkoff, A.H. (1980) J. Biol. Chem. 555, 3830-3831. 31. 32. 33. 80 Kawasaki, T., and Ashwell, G. (1976) J. Biol. Chem. 551, 1296-1302. Tanabe, T., Pricer, H.E., Jr., and Ashwell, G. (1979) J. Biol Chem. 555, 1038-1043. Hudgin, R.L., Pricer, H.E., Jr., Ashwell, G., Stockert, R.J., and Morell, A.G. (1974) J. Biol. Chem. 555, 5536-5543. Chapter III ENDOGENOUS LECTINS FROM CULTURED CELLS 11. Specific Affinity Columns for the Isolation of Carbohydrate-Binding Proteins* Calvin F. Roff, Paul R. Rosevear, John L. Hang and Robert Barker Department of Biochemistry Michigan State University East Lansing, MI 48824 Running Title: Affinity Columns for Carbohydrate-Binding Proteins 81 SUMMARY Treatment Of polyacrylamide beads with hydrazine hydrate followed by nitrous acid converts the anide groups on the resin into acyl azides. These reactive acyl azides can be used to couple ligands such as glycosides Of hexanolamine. Unreacted acyl azides are converted back into amide groups by treatment with ammonia. The conditions for this sequence Of chemical reactions were investigated and Optimized, with the following two particular objectives in mind: (a) efficient and stable coupling Of ligands that are available in limited amounts ( 100 umoles); and (b) minimizing the number of charged groups remain- ing on the polyacrylamide beads in the resulting product. Using these Optimized conditions, we have synthesized polyacrylamide beads containing defined carbohydrate ligands. These saccharide-containing beads were used to demonstrate the carbohydrate-binding capacity of three proteins (M,s 35,000, 16,000, and 13,500), isolated from cultured 3T3 mouse fibroblasts on the basis Of their binding to asialofetuin-Sepharose. All three proteins bound to polyacrylamide beads containing the disaccharide DGalB(l-+4)BDGlcNAc but not to beads containing the monosaccharide BDGal. He have also purified, in a single step, a carbohydrate-binding protein fran extracts Of hunan “foreskin fibroblasts using an affinity colmmn of polyacrylamide beads <1erivatized with DGa18(1-*4)BDGlcNAc. This protein (M, = 35,000) may represent the human counterpart Of the mouse protein Of similar 82 83 molecular weight and binding prOperties characterized in the 3T3 fibroblast system. 84 Affinity columns containing Sepharose covalently coupled to a glyCOprotein such as fetuin have been used for the isolation of carbohydrate binding proteins (CBPs). This approach has been successfully applied in the purification Of hemaglutinin receptor Of influenza virus (1), carbohydrate-Specific antibodies (2), as well as lectins from plant (3) and animal (4, 5) sources. Recently, we have isolated from 3T3 fibroblasts a fraction which binds tO asialofetuin- Sepharose and is eluted with lactose. This fraction yielded three polypeptide chains on analysis by polyacrylamide gel electrophoresis. In the course Of these studies, it became apparent that an affinity support containing only carbohydrates (CHOS) Of chemically defined structure would facilitate the analysis Of the carbohydrate-binding (:apacity and specificity of these asialofetuin-binding proteins. Although the use of agarose supports derivatized with CHOS has been (developed and used extensively in the purification of lectins (6) and enzymes (7), our Observations as well as those reported by others (8-10) indicate that affinity supports derived from cyanogen bromide (activation Of polysaccharide resins are highly charged, resulting in (appreciable non-specific binding, and are relatively unstable. Furthermore the use Of polysaccharide resins and glyCOproteins (complicate interpretations Of CHO binding since the saccharides are heterogenous. He have, therefore, Optimized the procedure originally «developed by Inman and Dintzis (ll, 12) for the derivatization of polyacrylamide (PA) beads and have used CHO ligands Of chemically defined structure. In the present communication we report the coupling of defined CHO structures to inert PA beads. Our procedure incorporated three 85 important features: (a) the use of disaccharides synthesized by glycosyltransferases such that the CHO ligands are of chemically defined structure, (b) the coupling Of limiting anounts of CHO in sufficient ligand density, and (c) synthesis of an affinity support which does not have a high capacity for ion exchange. Finally, we have used these PA supports derivatized with 8-0-ga1actose (Gal-HA-PA) and D-galactose-8(l———+4) BD-N-acetyglucosamine ( Gal-GlcNAc-HA-PA) to demonstrate the CHO binding specificities Of three CBPs isolated from 3T3 cells. MATERIALS AND METHODS (see supplemental material, p. 107) RESULTS ,Quantitation Of Coupling Reactions - The coupling of sufficient quantities of ligand to PA beads when starting with a limited anount of the ligand was the main achievement of our chemical studies. Although the basic scheme (Fig. 1) is that of Inman and Dintzis (ll, 12), there are features of the reactions that were particularly important for our isolation Of CBPS and that needed to be modified. For example, it was Of great importance to us that the derivatized PA beads dO not carry a large number Of positive charges because proteins bearing negatively charged residues may bind to then through ionic interactions rather than specific CHO interactions. Conversely, it was also important that the nunber of negative charges are sufficiently low so as not to interfere with specific carbohydrate recognition and binding. The following variables were monitored using 1,2-[14CJ-2- aminoethanol to quantitate the amount of coupled ligand: (a) length of incubation in hydrazine hydrate; the effect Of (b) pH, (c) time, (d) volume on the coupling Of ligand (ii, iii, Fig. l), (e) concentration Of the ligand; and (f) conditions for the regeneration Of amides from the acyl azides (iii iv, Fig. 1). The results of our chemical studies indicate that the Optimal conditions for the preparation Of affinity PA beads using this method are (see supplemental material for details, p. 107): '86 87 Figure 1: Schematic representation Of the method used to prepare CHO-PA affinity columns. 88 -g’NHZ ’g‘NH‘NHz -g-N3 N HN O j-m, H2" H2 > -g-NH-NH2 02 > -8-N3 O “E‘NHZ . “g’NHz ‘8'"“2 (i) (ii) g a HzN-(cnzia-R " -NHZ ”C'N3 " H u 9 8 -c-NH2 - -NH2 (iv) (111) NAc .. .-(..,,-ow .,-ra,,..w NAc dl-(CH214-Owo el-(CH214-OH Figure l 89 (a) formation Of the acyl hydrazide in 6.1 M hydrazine hydrate for 3 h at 50°; (b) addition of the ligand in H20 at pH 11.0; (c) coupling reaction time Of 4-5 h; (d) a reaction volume Of 1.5-2.5 times that Of the resin; (e) use Of 5 M NH40H-NH4C1, pH 10.0, to regenerate anides from the acyl azides. Binding Properties of Affinity Columns Containing CHO-PA - The binding properties of affinity columns containing CHO ligands coupled tO PA beads were studied using various proteins with different CHO-binding as well as charge characteristics. Some representative results Obtained with the GlcNAc-HA-PA column are presented in Figure 6. The ligand density of this column was approximately 8 mal GlcNAc/m1 Of resin. Hhen a solution Of wheat germ agglutinin (HGA), which has a high affinity for GlcNAc (20), is chromatographed on this column, the protein becomes bound to it (Fig. 6B). The bound material can be eluted fran the col unn when soluble GlcNAc is introduced in the developing solvent. This binding of HGA is not due tO prOperties of the PA backbone or the HA Spacer arm since HGA is not retained by HA derivatized PA beads (HA-PA column; Fig. 6A). Hhen soybean agglutinin (SBA) is applied to this colunn, an elution profile consistent with the chromatographic behavior of this protein on underivatized Bio-Gel P-150 was Observed (Fig. 6C). This would be expected because SBA has been shown to be specific for Gal residues but not for GlcNAc (21). The charge properties Of the GlcNAc-HA-PA resin were investigated using the proteins pepsin and histones, which have low and high lSOelectric points, reSpectively (22, 23). Hhen a solution of pepsin 90 Figure 6: Elution profiles of GlcNAc-HA-PA affinity column. The GlcNAc-HA-PA (2.0 x 8.0 cm) and HA-PA (2.0 x 5.5 cm) columns were equilibrated with PBS. The proteins were applied in PBS, washed with PBS, 0.25 M GlcNAc in PBS (starting at arrow a) and 0.4 M NaCl in phos- phate buffer (starting at arrow b). Fractions (3.4 ml) were collected and monitored for absorbance at 280 nm. The protein loaded, column used and % recovery were as follows: Panel A, 1.2 units of HGA on HA-PA column, 96% recovery; Panel B, 6.2 units of HGA on GlcNAc-HA-PA column, 87% recovery; Panel C, 23.2 units Of SBA on GlcNAc-HA-PA colunn, 79% recovery; Panel 0, 8.1 units Of pepsin on GlcNAc-HA-PA, 99% recovery; Panel E, 5.1 units Of histone II-A on GlcNAc-HA-PA column, 68% recovery. All units expressed are absorbances at 280 nm. 91 A B c C .f D 2 C c 3 A c C 3 . t L A. A 3 . .v A : . v C . t A. . u" . . bel" bII' .vbl' .vbll' A .. A. z c . 3 C : A v 3 C 1 t . C 2 A .. . v 3 . b . _ n p n B 4. 6 8 6 8 O O l. O. l. 0. cm W AEcOmNV wozo< 09¢ 4mo-0_m o\mz_2... . . ref: too: uEJ . on: .3353. - o L“. l1.||Tl.l-4..I..LIIL,.&I.I.?, 9, ~ v ,0 cu... .. ¢ 3.... ) Panel A 133 5.3 32.50 . r coo-am , . s 378. can 7...... V 0.3:! M gun—=03 .. . o 7 . Panel B 134 Purification Of CBP35,,CBP16 and CBP13.5 from Mouse Lung Because the lung showed a relatively large amount of material reactive with anti-CBP35 (Fig. 1, panel A, lane 2), we tested whether this tissue could serve as a source for large scale preparation of CBP35. Homogenates Of lung were fractionated by two cycles Of affinity chromatography on asialofetuin-Sepharose. Upon elution with lactose, a peak Of protein was eluted from the columns (Fig. 2). A polypeptide (M, = 35,000) was present in this peak as determined by gel electrOphoresis and Coomassie Blue staining (Fig. 3, lane a). This polypeptide was the only stained protein Observed. TO ascertain that this protein was CBP35, its reactivity with anti- bodies raised against CBP35 derived from 3T3 cells (anti-CBP35 (3T3)) was tested. The protein was transferred onto nitrocellulose paper and stained with Amido black; this revealed a polypeptide with M, = 35,000 (Fig. 3, lane b). In parallel, another strip of nitrocellulose paper was blotted with anti-CBP35 (3T3) followed by 125I-labeled goat anti-rabbit IgG. Autoradiography revealed a reactive polypeptide of M, = 35,000 (Fig. 3, lane c). This polypeptide migrated to a position identical with that of the Amido black stained polypeptide. The CBP35 isolated and identified from lung tissue will be hereafter designated CBP35 (lung). Although CBP16 and CBP13.5 were not detected by staining with Coomassie Blue, they were present in the material purified by two cycles Of affinity chromatography (Fig. 2). Hhen this material was labeled with 1251 (see below), subjected to gel electro- phoresis and autoradiography, three major bands are seen. These corresponded to CBP35, CBP16 and CBP13.5 (data not shown). 135 Figure 2. Representative column profile showing the affinity chro- matography of mouse lung extracts on an asialofetuin-Sepharose colunn (1.2 x 15 cm). Lung extracts from 100 mice were subjected to affinity chromatography on asialofetuin-Sepharose as described in Materials and Methods. The lactose eluted material from the first cycle Of affinity chromatography was dialyzed in the presence Of asialofetuin-Sepharose and poured into a column. Fractions (4 ml) were collected and 0.4 m1 aliquots were assayed for protein by the method Of Bradford (11). The arrow marks the beginning Of the lactose gradient (0-150 mM lactose, 100 ml total volune). The horizontal bar indicates fractions which were pooled for gel electrOphoresis analysis, for gel filtration on Sephadex G-150, and for iodination. 136 s. mmmeo; 15.. mu. w. 0 u _ _ — // m I, /. l. /. [I I, /. Ill nu L / AH / IIV / an 2 _ b _ . _ . O 2 8 4 in nu nu nv AEc mmmv wozozo: .opoopwo>o ouou do women mg» co nocooEoo mew: mzwuoop ovcpooomiomouoopom m=o_co> on» mo cowuomocmao mo mmucum one.» 156 .mcpuooo o_$pooomiwmouoopoo.$o mowucoooco use to comwcooeou "a epoch l l H ._._o-:.z< mu» oz mu» oz 1.: 3 w>=bc.o:u no 3 8-2. 3 a... 5.3 Ho zmzozoz 525 3“qume muzozoz zuzozoz mmzozoz .520240. on. go. gun zen. go. :3 cs. :zamam )0 O s/ 9,7 00 O 90 .2 .Av .w v o. xv v % xx true/«a any 9 one MIC R (11111311111111!(IIIIIIIIIIIIIIIIIIIES