. I: 5.: l . . ,. .1}. 1.. z. . {IKL “Wllllll‘l‘llllmg LIBRARY Mlcmnn State University This is to certify that the dissertation entitled Carbohydrate Binding Protein 35: In Vivo and In Vitro Expression Properties of the Polypeptide presented by ‘ Neera Agrwal has been accepted towards fulfillment of the requirements for _Eh¢ degree in Bio chemis try M 1. due/g fl Major professor / Date August 25, 1992 Mcrl.’.,...1a- .' A v v- .n 1 y - - 042771 CARBOHYDRATE ENDING PROTEIN 35 : EV VIVO and UV VHRO EXPRESSION PROPERTIES OF THE POLYPEPTIDE By Neera Agrwal A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1992 ABSTRACT CARBOHYDRATE BINDING PROTEIN 35: IN VIVO AND IN VITRO EXPRESSION PROPERTIES OF THE POLYPEPTIDE By Neera Agrwal Previous studies had shown that the amount and subcellular localization of the endogenous lectin Carbohydrate Binding Protein 35 (CBP35) in murine 3T3 fibroblasts is regulated by the growth state of the cell. In proliferating cells, the protein levels are higher than in quiescent cells and largely localized in the nucleus. Using the cDNA clone for CBP35 as a probe, we have examined the expression of the CBP35 gene in quiescent and serum-stimulated cells. The main conclusions of these studies include: a) higher levels of accumulated CBP35 mRNA are found in proliferating cells than in quiescent cells; b) the rise in mRNA levels is detected within 30 minutes of serum stimulation, and increases until “'20 hours after serum addition; c) the mRNA is "superinduced" in the presence of cycloheximide; d) the transcriptional rate for the CBP35 gene increases within 3 hours of serum stimulation, and reaches a maximum level at 10 hours following serum addition; and e) the transcription of the CBP35 gene occurs even in the presence of cycloheximide, indicating that this is a primary event in the response to serum growth factors by the cell. Using the cDNA clone for CBP35, the full length recombinant protein and the NHZ- and COOH-terminal domains have been expressed in E. coli cells. Analyses of the expressed proteins have demonstrated that the galactose binding activity of CBP35 is contained entirely within the COOH-terminus. Differential scanning calorimetry with the recombinant polypeptides has shown that the two domains are folded independently. The recombinant CBP35 (rCBP35) has also been used to examine the uptake of the protein by nuclei from 3T3 cells. The results indicate that the protein uses a pathway distinct from that used by the synthetic substrate, human serum albumin bearing the nuclear localization signal of the SV40 large T antigen. Cytosolic factors which are sufficient for the entry of the synthetic substrate into the nucleus cannot support the nuclear import of rCBP35. This suggests that the control of nucleo-cytoplasmic distribution of rCBP35 may be mediated by the presence of a cytoplasmic anchor, or the availability of a factor which serves as a carrier for rCBP35 in nuclear import. To David and To my parents for their love and faith in me ACKNOWLEDGEMENT I thank Dr. John Wang for his infinite patience in guiding me through the time that I have spent in his laboratory. He is an extraordinary teacher, and I am fortunate to have done my graduate work under his care. I thank my committee members for their insights into this research project: Dr. Zachary Burton, Dr. Susan Conrad, Dr. Jerry Dodgson, and Dr. Bill Smith. I would also like to thank Dr. Richard Anderson for his thoughts and views. Much of the best work in science arises as a result of collaborations between investigators. We could not have been successful without the assistance of the following: Dr. Ron Patterson, Dr. Betty Werner and Sue Dagher for their scientific input on the nuclear import assays; Dr. John Wilson for the DSC analysis; Dr. Richard Wozniak and Dr. Gunter Blobel for the use of the microinjection facilities; and Dr. Natasha Raikhel for the use of the fluorescence microscope. I would also like to thank Dr. David Lerner for his support throughout the years, and Dave Schwab for the discussions regarding in vitro protein production. The entire staff in the Biochemistry Department has always done their utmost to facilitate administrative matters. I especially thank Linda Lang for all of her help. For the photographic work, I thank Kurt Stepnitz, who always found a way to get the work done ”as soon as possible". Finally, I wish to thank my colleagues in the Wang and Schindler laboratories for their advice and helpful criticisms: Dr. Elizabeth Cowles, Dr. James Laing, Patty Voss, Dr. John Ho, Dr. Shizhe Jia, Dr. Quan Sun, Sung Yuan Wang, Kim Hamann, John Loh, Anandita, and Mark Kadrofske. I would like to especially acknowledge the help given to me by Liz Cowles, Jamie Laing, and Patty Voss during my first few years in the lab. TABLE OF CONTENTS Page HST OF TABLES .............................................. x LIST OF FIGURES ................. . .......................... xi LIST OF ABBREVIATIONS ..................................... xiv INTRODUCI'ION .............................................. 1 CHAPTER I: LITERATURE REVIEW ............................ 3 INTRODUCI‘ ION TO LECTINS ............................. 3 Classification of Animal Lectins .......................... 4 The C-type lectins .................................... 6 The S-type lectins ................................... 10 The L—14 family of S-type lectins ........................ 11 The L-30 family S-type lectins .......................... 18 Saccharide binding characteristics of the S-type lectins ........ 19 Subcellular localization of the L-14 and L-30 lectins .......... 20 The bifunctional nature of lectins ........................ 25 a) Selectins ................................... 25 b) The S-type L-14 lectins ........................ 27 CARBOHYDRATE BINDING PROTEIN 35 ................... 27 Structure of the CBP35 gene and protein .................. 28 Ligands for CBP35/Mac-2 ............................. 30 Proliferation dependent localization and expression of CBP35 .............................. 30 THE REGULATION OF EXPRESSION OF C-FOS .............. 32 Introduction to Primary Response Genes .................. 32 Kinetics of the expression of c-fos ....................... 34 Superinduction of the c-fos mRNA ...................... 34 vi Lilil The c-fos promoter .................................. 35 Factors controlling the induction and function of fos .......... 37 Regulation of transcription by the PRG ................... 38 NUCLEAR TRANSPORT ................................. 39 General characteristics of nuclear transport ................ 39 Nuclear localization signal ............................. 40 The nuclear pore complex ............................. 41 Cytosolic factors are required for nuclear transport . . . .' ...... 43 NLS binding proteins ................................. 45 Alternative methods of nuclear transport .................. 46 Control of nuclear transport ........................... 48 LITERATURE CITED .................................... 50 CHAPTER II: CARBOHYDRATE BINDING PROTEIN 35: Levels of Transcription and mRNA Accumulation in Quiescent and Proliferating Cells ...................... 63 FOOTNOTES ........................................... 64 ABSTRACT ............................................ 65 INTRODUCTION ........................................ 66 MATERIALS AND METHODS ............................. 68 Cell Culture ..................................... 68 RNA Isolation and Northern Blot Analysis ............... 68 Nuclear Run-off Transcription Assays .................. 71 Plasmids and Preparation of Probes .................... 73 Indirect Immunofluorescence ......................... 74 RESULTS .............................................. 75 Kinetics of the Accumulation of CBP35 mRNA upon Mitogenic Stimulation .................... 75 Analysis of the Transcription Rate of the CBP35 Gene after Stimulation ........................ 78 Effect of Cycloheximide on Transcription Rate and mRNA Accumulation following Mitogen Addition ...... 83 vii Comparison of the Expression of the CBP35 Gene in Normal and Transformed Cells ....................... 86 Comparison of the Expression of the CBP35 Gene in Sparse and Confluent Cultures of 3T3 Cells ............ 96 DISCUSSION ........................................... 97 ACKNOWLEDGEMENTS ................................ 101 REFERENCES ......................................... 102 CHAPTER III: CARBOHYDRATE BINDING PROTEIN 35: Properties of the Recombinant Polypeptide and the Individuality of the Domains .................. 104 FOOTNOTES .......................................... 105 SUMMARY ........................................... 106 INTRODUCT ION ....................................... 107 NMTERIALS AND METHODS ............................ 109 Construction of prCBP355 and Purification of rCBP35 ..... 109 Site-Directed Mutagenesis and Generation of Recombinant Clones ............................ 110 Preparation of the N- and C-domains ................. 111 Antibodies ...................................... 112 Analytical Techniques ............................. 113 Differential Scanning Calorimetry (DSC) ............... 114 RESULTS ............................................. 116 Expression and Purification of Recombinant CBP35 ....... 116 Generation of the NH;— and COOH-terminal Domains of rCBP35 .................. 122 Development and Characterization of Antisera Against rCBP35 ......................... 130 DSC Analysis ................................... 133 DISCUSSION .......................................... 144 ACKNOWLEDGEMENTS ................................ 147 viii REFERENCES ......................................... 148 CHAPTER IV CARBOHYDRATE BINDING PROTEIN 35 : Preliminary Studies on the Transport of the Recombinant Polypeptide into the Nucleus ............. 150 FOOTNOTES .......................................... 15 1 SUMMARY ........................................... 152 INTRODUCTION ............. ‘ .......................... 153 MATERIALS AND METHODS ............................ 155 Cell Cultures .................................... 155 Preparation of Import Substrates ..................... 155 Nuclear Import Assays ............................ 157 Fluorescence Microscopy ........................... 158 RESULTS ............................................. 160 Nuclear Transport of Rh-HSA-NIS Assayed with Permeabilized Cells ........................... 160 Behavior of Rh-rCBP35 in the Permeabilized Cell Assay System ................................ 163 Nuclear Transport in Intact Cells ..................... 166 Nuclear Transport of Rh-rCBP35 in Quiescent and Proliferating Cells ............................. 174 Microinjection of NH;- and COOH- Terminal Domains of rCBP35 .................. , ..... 175 DISCUSSION .......................................... 180 ACKNOWLEDGEMENT ................................. 185 REFERENCES ......................................... 185 CHAPTER V CONCLUDING STATEMENT ............. . ......... 187 REFERENCES ......................................... 188 ix LIST OF TABLES Page CHAPTER I TABLE I. Classes of Animal Carbohydrate Binding Proteins ................................ 5 TABLE II. The C-Type Lectins .............................. 7 TABLE III. L-14 Group of S-Type Lectins ..................... 12 TABLE IV. L-30 Group of S-Type Lectins ..................... 13 TABLE V. S-Type Lectins of Mr 16000-22000 .................. 14 CHAPTER III TABLE 1. Parameters Affecting Yield of rCBP35 .............. 121 CHAPTER IV TABLE I. Nuclear Localization in Serum-Starved and Serum-Stimulated 3T3 Cells ..................... 176 CHAPTERI Figure 1. Figure 2. CHAPTERII Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. LIST OF FIGURES Summary of the structural features of the C-type lectins ................. Summary of the structural features of the S-type lectins ................. Kinetics of the accumulation of mRNA for CBP35 after serum stimulation of growth-arrested 3T3 cells Gene transcription rates after serum stimulation of quiescent 3T3 cells in the absence (A) and presence (B) of cycloheximide (10 ug/ml) . . . . Kinetics of the changes in relative transcription rates of the CBP35 gene following serum stimulation of quiescent 3T3 fibroblasts .................. Effect of cycloheximide on the levels of accumulated mRNA for CBP35 ............. Immunofluorescence analyses comparing CBP35 in 3T3 and 3T3-KiMSV cells . . . Northern blot analysis of CBP35 mRNA in 3T3 and 3T3-KiMSV cells .......................... Autoradiogram of nuclear run-off assays comparing the transcription rate of the CBP35 gene in 3T3 and 3T3-KiMSV cells ....... Page ...... 9 ..... 16 ..... 77 ..... 80 ..... 82 ..... 85 ..... 88 ..... 91 ..... 93 Figure 8. CHAPTERIII Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. CHAPI'ERIV Figure 1. Figure 2. Northern blot analysis to compare half-lives of CBP35 mRNA in 3T3 and 3T3-KiMSV cells ........ 95 Schematic diagram of the construction of the recombinant expression vector prCBP355 and site-directed mutagenesis to obtain the amino terminal and carboxyl terminal domains ............. 118 Purification of rCBP35 from E. coli transformed with the expression vector prCBP355 ..... 120 Purification of the N-domain from E. coli transformed with expression vector containing a fragment of mutagenized cDNA ......... 125 Digestion of rCBP35 with collagenase D and affinity purification of C-domain on Asialofetuin—Affi-gel ..... 129 Characterization of anti-CBP35 antisera using immunoblotting .................... 132 Representative thermograms from DSC analysis of rCBP35 .................... 135 Deconvolution of the thermogram of rCBP35 in pH 10 buffer ............. 137 Effect of heat denaturation of the N-domain on collagenase digestion of rCBP35 ........ 140 Representative thermograms from DSC analysis of N- and C-domains ................ 143 In Vitro nuclear import assay comparing the translocation of Rh-HSA-NLS with Rh-I-ISA ......... 162 In vitro nuclear import assay for Rh-rCBP35 ......... 165 xii Figure 3. Figure 4. Figure 5. Figure 6. In vivo nuclear import assay for Rh-HSA-NLS in the absence and presence of WGA .............. 168 In vivo nuclear import assay comparing the translocation of Rh-rCBP35 with Rh-I-ISA ........... 171 In vivo nuclear import assay for Rh-rCBP35 showing the nuclear and non-nuclear localization of the protein ....................... 173 In vivo nuclear import assay for the NHZ- and COOH-terminal portions of rCBP35 ........ 179 xiii ATP BHK BSA CBP CBP35 cDNA CRD CRE DME DNA DSC EBP EGF ER FBS FGF Fuc Gal GalNAc GlcN Ac HEPES hnRNP HPLC HSA hsc70 hsp70 IE IPTG kD, kDa Lac LBP Man 2-ME M6P LIST OF ABBREVIATIONS adenosine triphosphate baby hamster kidney bovine serum albumin carbohydrate binding protein carbohydrate binding protein 35 complementary deoxyribonucleic acid carbohydrate recognition domain CAMP response element Dulbecco’s modified Eagle Medium deoxyribonucleic acid differential scanning calorimetry IgE binding protein epidermal growth factor endoplasmic reticulum fetal bovine serum fibroblast growth factor fucose galactose N-acetylgalactosamine N-acetylglucosamine fl-Z-hydroxyethylpiperazine-N-Z-ethanesulfonic acid heterogeneous nuclear ribonucleoprotein high performance liquid chromatography human serum albumin heat shock cognate 70 heat shock protein 70 immediate early gene isopropyl-fl-D-thiogalactopyranoside kilodaltons lactose laminin binding protein mannose 2-mercaptoethanol mannose-6-phosphate xiv MEF mGBP mRNA N BP NEM NGF NLS NPC PAGE PBS PDGF PRG rCBP35 rCBPCD rCBPND RNA SDS snRN P SRE SRF TGF TMG TRE WGA mouse embryonic fibroblast mouse galactoside binding protein messenger ribonucleic acid nuclear localization signal binding protein N-ethylmaleimide nerve growth factor nuclear localization signal nuclear pore complex polyacrylamide gel electrophoresis phosphate buffered saline platelet-derived growth factor primary response gene recombinant CBP35 carboxyl terminus of rCBP35 amino terminus of rCBP35 ribonucleic acid sodium dodecyl sulfate small nuclear ribonucleoprotein serum response element serum response factor transforming growth factor trimethylguanosine TPA response element wheat germ agglutinin .m- u, INTRODUCTION Carbohydrate Binding Protein 35 (CBP35; Mr 35,000) was initially purified from extracts of murine Swiss 3T3 fibroblasts on the basis of its affinity for ,3- galactosyl-containing glycoconjugates. It has since been isolated and studied from other species, as well as many tissue types. Our interest in CBP35 stems from three key properties of the protein: a) its structure; b) its subcellular localization; and c) its proliferation dependent expression. First, the amino acid sequence of the polypeptide, deduced from the nucleotide sequence of the cDNA clone, showed that the protein contains two distinct domains. The amino terminal half contains eight contiguous repeats of the 9-amino acid sequence PGAYPGXXX. The carboxyl terminal half contains 13 amino acids that are invariant in CBP35 and other fi-galactoside specific lectins and, therefore, is believed to constitute the carbohydrate recognition domain of the lectin. These features of the primary sequence prompted us to ask questions concerning the physico-chemical properties of the individual domains, and their roles in the biological activity of the protein. Second, studies on the subcellular localization of CBP35 showed that the majority of the lectin is intracellular, although a small percentage is also found at the cell surface, despite the lack of an obvious signal sequence to target it through 2 the secretory pathway. Moreover, the intracellular lectin can be found in the nucleus and the cytoplasm, depending on the proliferative state of the cell. Finally, when quiescent 3T3 cells were stimulated by the addition of serum, the levels of CBP35 rose dramatically, and the protein was translocated into the nucleus. These observations formed the basis for our studies on the transcriptional regulation of CBP35 expression, as well as factors controlling its nuclear localization. Thus, this literature review will discuss the following topics: a) the structure and subcellular localization of carbohydrate binding proteins; b) the proliferation dependent expression of proteins; and c) mechanisms by which proteins translocate into the nucleus. CHAPTER I LITERATURE REVIEW INTRODUCTION TO LECTH‘IS Specific recognition is a key event in every biological process. For example, the specificity of recognition is seminal in cell-cell interaction, a step which is necessary in such diverse functions as immunomodulatory activity, fertilization, development, and infection (1). The discovery that all cells carry carbohydrates on their surface led to the notion that these glycoproteins may play a fundamental role in recognition processes. This idea was furthered by the finding that there exist molecules which specifically recognize these sugar structures. These molecules were termed lectins (2). Lectins are classified as non-enzymatic, non- immune carbohydrate-binding molecules (3). The traditional definition of a lectin has implied that it is at least a bivalent molecule with respect to saccharide binding. The terms carbohydrate binding protein and lectin will be used interchangeably in the discussion to follow. Lectins were first discovered in plants as hemagglutinating substances. They have since been found in almost all organisms, in all tissues and cell types, and both intracellularly and extracellularly. Their exquisite ability to differentiate 4 between different monosaccharides and oligosaccharides, as well as to recognize subtle variations in sugar structures and linkages has made the lectins a potentially central figure in the cellular recognition system. In this chapter, I will concentrate on the animal lectins. The plant lectins have been reviewed elsewhere (4,5). Classification of Animal Lectins The region of the polypeptide which has the ability to bind to carbohydrates is designated as the Carbohydrate Recognition Domain (CRD). Animal lectins can be grouped into several categories based on sequence similarities within the CRD and/or saccharide binding characteristics (Table I). By these criteria, the most well studied lectins fall into the C-type and S-type categories (7). The C-type lectins are those which require the bivalent cation Ca2+ for sugar binding activity. These proteins may be membrane bound or soluble, and may also be glycosylated. The S-type lectins were originally defined to be proteins which require reducing agents for sugar binding. They have only been found as soluble, non-glycosylated entities. The heparin binding lectins that have been isolated have no similarities with the other lectin families. They may or may not exhibit hemagglutinating activity, and may require divalent cations for binding to heparin. Their binding to heparin is inhibitable by competing sugars (8,9). There are two other groups of carbohydrate binding proteins which bear no common structural of functional features with the above mentioned lectins. .- 5 TABLE I. CLASSES OF ANIMAL CARBOHYDRATE BINDING PROTEINS' Family Type Examples Caz” Cysteines Location receptors mediating . . plasma . . yes disulfide glycoprotein endocyt051s membrane mannose binding proteins yes disulfide serum, liver C-type proteoglycan core lectins 7 yes disulfide extracellular matrix selectins yes cell surface pulmonary surfactant yes disulfide lung, fluid xtracellular L-14 and L-30 [3- e ’ S-type galactoside binding lectins no sulfhydryl cell surface, nucleus plasma 250 kDa M-6—P receptor no disulfide membrane, Mannose-6-P extracellular receptor plasma 46 kDa M-6-P receptor yes disulfide membrane, extracellular Pentraxin serum amyloid protein yes disulfide serum Heparin . 9 9 binding lectins placental lectin, p33, p41 . . ' adapted from reference 6 " Ca“ requirement for sugar binding 6 These include the serum amyloid protein, and the mannose-6-phosphate receptors. I will focus on the C—type and S-type lectins in this review. The C-type Lectins The C-type CRDs are found in a variety of proteins with a diversity of structure and function (Figure 1 and Table 11). These include the following subgroups: a) the selectins, such as gp9OMEL, ELAM-l, and GMP-140 (10,11,12); b) transmembrane receptors, as exemplified by the asialoglycoprotein receptors (13), the Kupffer cell receptor (14), and the hepatic lectins (13); c) the macrophage receptors which are involved in the phagocytosis of pathogens (15); d) the soluble mannose binding proteins, found in liver and serum (16); e) the soluble pulmonary surfactant apoproteins (17); and f) the proteoglycan core proteins (18). Some C-type lectins have also been found in such divergent species as Dictyostelium discoideum (19), the fly (20), barnacle (21), sea urchin (22), and the tunicate P01yandrocarpa misakicnsis (23). The initial studies on the C-type lectins denoted the fact that these proteins required Ca2+ for binding activity. Subsequent cloning and sequence analyses have shown that the CRD for this type of lectin can be recognized by a sequence motif which contains approximately 30 conserved residues over a "' 120 amino acid range (7). The alignment of the CRD regions of 22 distinct proteins shows that there are 14 invariant residues, with an additional 18 residues that are conserved in character. These residues fall into three classes: a) 4 cysteines, involved in disulfide bond formation; b) a "WIGL" sequence which indicates a region packed TABLE II. THE C-TYPE LECTINS' Species/Source Lectin Ligand" Reference flesh fly hemolymph fly lectin Gal 20 sea urchin sea urchin lectin Gal 22 :52; Ziggggarp a tunicate lectin Gal 23 avian liver chicken hepatic lectin GlcNAc 13 sheep, goat, buffalo liver hepatic lectin Gal 27 rat liver RHL-l, RHL-2/3 Gal 28,29 rat liver Kupffer cell receptor Fuc 14 mouse spleen gpl90MEL M6P 10 mouse macrophage macrophage lectin Gal 30 human lung pulmonary surfactant Man 17 human macrophage mannose receptor Man 15 human pancreas pancreatic stone lectin Gal 31 human fibroblast fibroblast proteoglycan Gal 18 COI'C b A representative group of the C-type lectins Gal, galactose; GlcNAc, N-acetylglucosamine; Fuc, fucose; M6P, mannose-6- phosphate; Man, mannose Figure 1. Summary of the structural features of the C-type lectins. The invariant residues in the C-type carbohydrate recognition domain are shown, combined with the effector domains (if any) found in the various members of the C-type lectin family. GAG, glycosaminoglycan; EGF, epidermal growth factor. Adapted from reference 6. C-TYPE CARBOHYDRATE 4— RECOGNITION -> DOMAIN fly / moth lectin sea urchin/ barnacle lectin — av» Polyandrocarpa FAA- tunicate lectin pancreatic stone lectin M— asialoglycoprotein receptor (RHL-1) chicken hepatic lectin lymphocyte FcE receptor 1 RHL-2/3 :3er o“ MEMBRANE f/ \ ANCHOR \\ / 9 ‘~ kupfter cell I \ l receptor T-W-P\ \CE I alveolar 62o 123$ macrophage C lectin selectins —1 EGF COMPLI- MEM- DOMAIN MENT BRANE DOMAIN ANCHOR fibroblast «um-I C—C— EGF-LIKE pmteog'ycan core REPEATS CYS-RICH cartilage 0893‘: proteoglycan co‘ré' human macrophage GAG CHAINS mannose receptor (multiple CRD) conglutinin serum liver mannose binding proteins TR'PLE HELIX pulmona surfactant (SP-A, s p) FlBRONECTIN—C—C— REPEAT C-TY CARBOHYDRATE + RECOGNITION —> DOMAIN fly / moth lectin sea urchin / barnacle lectin Polyandrocarpa tunicate lectin pancreatic stone lectin asialoglycoprotein receptor (RHL—l) chicken hepatic lectin lymphocyte FcE receptor RHL-2/3 I.—_:I—/-—- MEMBRANE ‘\ ANCHOR Q/J \\ kuptfer cell (0» \ receptor T—W—p\ ‘ 0 CE alveolar {Q x’é macrophage («L-GI lectin selectins 4 fibroblast __- EGF—LIKE proteoglycan core REPEATS EGF COMPLI- MEM DOMAIN DMENT BRA NE OMAlN ANCHOR C—C— CYS-RICH DOMAINS cartilage ___ proteoglycan core human macrophage GAG CHAINS mannose receptor FIBRONECTIN—C—C— (multiple CRD) REPEAT conglutinin serum liver mannose binding proteins TRIPLE HELIX pulmonary surfactant (SP- -AS D) 10 into hydrophobic cores; and c) residues forming the ligands for Ca“ (24). Analysis Of the gene structure for several C-type CRDs has shown that they fall into three distinct types (6,25). In the first type, the single CRD is encoded by three separate exons. This can be found in the rat asialoglycoprotein receptor (RHL-l), the Kupffer cell receptor, and the chicken proteoglycan core protein. On the other hand, the (single) CRDs for the mannose binding proteins, the pulmonary surfactant SP-A, and the murine lymphocyte homing receptor are encoded in a single exon. Finally, the cell surface mannose receptor of macrophages and hepatic sinusoidal cells has eight tandemly placed CRDs which comprise the carbohydrate recognition domain of this protein (26). Studies to determine the role of each of the CRDs in this protein have revealed that while CRD4 can function independently of the other CRDs, it does not exhibit the high affinity of binding of the whole molecule. Thus, it is probable that the multiple CRDs are required for tight binding to multivalent ligands. The S-type lectins The S-type differ from the C-type lectins in that they: a) do not depend on cations for carbohydrate binding activity; b) have an affinity exclusively for Gal/Lac containing structures; c) have a specific consensus sequence, differing from that in the C-type lectins, in the CRD domain (7) (figure 2); d) are isolated invariably (thus far) as soluble proteins; and e) are susceptible to oxidation inactivation of sugar binding activity in the absence of reducing agents. The nature of the oxidation may lie in the cysteine or tryptophan residues in these 11 proteins (32). These lectins do not appear to be glycosylated. The S-type lectins (also called the S-lac lectins) can be subdivided into several groups, based on their subunit molecular weight. The two best characterized groups are the L-14 and L-30 lectins. The L-14 group of lectins ranges in size from M, 12,000-14,500 daltons (Table III). The L-30 group of lectins consists of proteins of M, 29,000-35,000 daltons (Table IV). There is another group of S-type lectins which are of M, 16,000-22,000 (Table V). This group of lectins has been isolated from rat intestinal tissue (56,57), mouse lungs (45,58), Xenopus laevis skin tissue (59), and the nematode C. elegans (60). Although these lectins share many amino acid and structural similarities with the L-14 group, they appear to be novel members of the S-lac family, encoded by a separate gene. The 67 kDa component (the large component) of the elastin receptor has also been classified as an S-type lectin based on the following data: a) the 67 kDa protein can be eluted from an asialofetuin affinity column by the addition of ,6- galactoside sugars; b) the asialofetuin-purified protein can be eluted from a column derivatized with elastin peptides by lactose; and c) antibody raised against the rat lung L-14 lectin reacts with the 67 kDa protein (63). However, the S-type CRD consensus sequence has not been found in the 67 kDa lectin. The L-14 family of S-type lectins The most abundant of the S-type lectins are the L-14 proteins. They are isolated as dimers, and have one or more cysteine residues. Evidence obtained 12 TABLE III. L-14 GROUP OF S-TYPE LECTINS‘ Species Lectin Tissue Source M,” References Conger eel congerin skin mucus 15000 33 C. elegans GBP32 whole worm 32000 237 B. arenarum L-15 ovary 15000 34 CLL-II intestine, liver 12000 35,36 Chicken CHL heart 13000 37 C—14 skin 14000 38 CLL-1 muscle, liver 15000 35,39 Marmoset L-15 neonate 15000 40 Rabbit Galaptin bone marrow 13000 41 Bovine BHL fgfrfil’lgung’ Spleen: 12000 42,43 Porcine PHL heart 14700 44 CBP13.5 lung, fibroblast 13500 45 Mouse L-14.5 gifigifgzma 14500 46 mGBP embryonic fibroblast 14735 47 Rat RL-14.5 nmefifrcgflsl‘jgtgeszrrf‘e‘“ 14500 48,49,50 HLBP14 melanoma 14000 235 H14 placenta, hepatoma 14000 51 Human HL—14 lung 14000 52,53 HBL brain 14500 54 L-14-II hepatoma 14650 55 ' A representative group of the L-14 lectins Some of these lectins may be identical " Subunit M, as determined by reducing SDS-PAGE 13 TABLE IV. L-30 GROUP OF S-TYPE LECTINS' Species Lectin Tissue/Cell Source M,” References . 45,58 CBP35 lung, fibroblast 35000 75,76 LBP macrophages 35000 74 Mouse _ L-34 f‘bmsarcoma’ 34000 46,72 melanoma Mac-2 macrophages 32000 77,78 Hamster rarest“ BHK cells 30000 79 ectm . 48,56 RL-29 lung, brain 29000 49,50 Rat . . eBP baSOPh‘Im 31000 80 leukemia cells HL-29 lung, brain 29000 62,81 EBP basophilic 31000 82 Human leukemia cells Mac-2 macrophages 32000 83 CBP35 lung, fibroblast 35000 58,84 ' Some of these lectins may be identical b Subunit M, as determined by reducing SDS-PAGE 14 TABLE V. S-TYPE LECTINS OF M, 16000-22000 Species Lectin Cell/Tissue Source M,” References Electric eel electrolectin electric organ 16000 32 C. elegans L-16 whole worm 16000 60 Xenopus [aevis skin lectin skin 16000 59 Chicken C-16 liver 16000 38 Mouse CBP16 lung, fibroblast 16000 45,58 L-17 intestinal mucosa 17000 56,57 L-19 intestinal mucosa 19000 56,57 R L-21.5 intestinal mucosa 21500 57 at RL-22 lung 22000 48 IgE binding intestine 17500 61 protein CBP16 lung, fibroblast 16000 58 Human HL—22 lung 22000 62 ‘ A representative group of the S-type lectins of M, 16-22 kDa Some of these lectins may be identical " Subunit M, as determined by reducing SDS-PAGE Figure 2. Summary of the structural features of the S-type lectins. A. The multidomain structure of the S-type lectins. The carboxyl terminal domain (white box) contains the carbohydrate recognition domain. The shaded boxes represent various different effector domains, whose function is unknown. B. The features of the domains of the L-14 and L-30 lectins. The 13 invariant amino acid residues that occur in a 39-residue sequence in the CRD are shown. Also shown is the 9-amino acid sequence that is repeated in the amino-terminal domain of the L-30 lectins. The letter n, designating the number of repeats, ranges from 5 in the human Mac-2 sequence to 10 in the rat EBP sequence. A single residue between invariant residues is denoted by hyphen (-). Sequences of two or more residues are denoted by the symbol (~). C Hydropathy plot of murine CBP35, as determined by the deduced amino acid sequence. The distinctive hydropathy plot - pattern of the two domains of an L—30 lectins is illustrated. Positive values indicate hydrophilicity and negative values indicate hydrophobicity. Adapted from references 67 and 88. 16 ; Ins-22 kD WM 1.7.... tH.NPRF~v-N~we-E-R~F~c~ I L-30 I ~(PGAYPe--->,~ I~H~NPRF~V-N~WG-E-R~F~G~ I LLLLLLLLLLLLLLLLL O—NOI HYDROPHILICITY I’V‘WIW ‘ “MINI/IA l 1 l I 1 50 I00 I50 200 250 SEQUENCE POSITION 'll QIN— 17 from analysis Of the cDNA clones indicates that the L-14 lectins are encoded by a multi-gene family. At least three distinct L-14 lectins have been hinted at by the cDNA clones. The L-14-I lectin is representative of the first type of L—14 lectin isolated from vertebrate tissues. These lectins exhibit greater than 85% conservation at the amino acid level (55,64,65,66). The L-14-II lectin, isolated from a human hepatoma library, shares 43% amino acid identity with L-14-I (55). The avian L-14 cDNA appears to be more divergent showing 50% amino acid identity with the L-14-I lectin (65,66). The gene for the L-14-I type of protein contains 4 exons, with the third exon containing the CRD. The upstream region of the gene indicates the presence of a possible heat shock element, a putative steroid binding site, a putative metal regulatory element, and a sequence related to the Y box of the histocompatibility genes. In addition, there are intronic Alu sequences, and a G/T cluster downstream of the polyadenylation signal (53). The upstream region of L-14-II gene differs from that of L-14—I. There are two tandem TATA boxes, an Spl-binding site, and a putative site for regulation by AP-l (55). A novel type of L-14 lectin has been isolated and cloned from the nematode C. eICgans (60,237). The lectin has a M, of 32 kDa, and hence has been named 32 kDa GBP. Analysis of the cDNA clone for the protein shows that it has two tandem repeats of the L-14 unit. Each of the units show approximately 30% amino acid conservation between themselves, as well as the other S-type lectins. The S-type consensus sequence is well conserved, with an exact match of 9 amino acids. However, unlike the other L-14 lectins, there appear to be no 18 cysteines in the 32 kDa GBP. The significance of the repeated domains is unknown, especially since the hemagglutination activity is weak. Histochemical and biochemical studies have shown that the L-14 lectin is ubiquitously distributed in tissues. It is found in skeletal muscle (68,69), nervous tissue (49,50,54), connective tissue (70,71), and tumor tissue (46,72,235). Perhaps, the richest sources of the L-14 lectin are the fetal lung and uterine tissues. This is further corroborated by experiments which were performed to isolate genes which are selectively expressed during embryogenesis. One of the cDNA clones isolated was that for the L-14 lectin (73). The L-30 family S-type lectins The members of the L-30 family (Table IV) are all identical to each other or highly homologous (285%), as determined by cDNA sequence analysis. Despite their similarities in structure, a number of different functions have been ascribed to these proteins. The lectins RL-29, HL-29, BHK lectin, L-34, and CBP35 were isolated as fi-galactoside binding proteins (45,48,62,72,79). 63? was isolated as a protein that bound to Immunoglobulin E (80). Mac-2 was initially identified as a cell surface antigen for thioglycolate-elicited peritoneal macrophages (77). LBP was identified as the major non-integrin laminin binding protein in macrophages, and subsequent sequencing showed its identity to CBP35Mac-2 (and hence, the other members Of the L-30 group) (74). Southern blotting analysis of genomic DNA suggests that there is a single gene which encodes these lectins (Jia, S. and Wang, J.L., unpublished results). Northern 19 blotting analysis has identified a major mRNA transcript of approximately 1.1-1.3 kB (75,80,96). Sequence analysis of the cDNA clones corresponding to these lectins has demonstrated that they are composed of two distinct domains (Figure 2). The carboxyl terminal domain houses the CRD (7). This domain contains the 13 invariant residues found in the CRD regions of all the known S-type lectins. The L-14 lectins consist only of the CRD; however, the L—30 lectins contain a second effector domain. This domain is highly proline and glycine-rich, since it has eight contiguous repeats of the nine amino acid sequence PGAYPGXXX (76). The two domain structure of these lectins is further borne out by hydropathy analysis of the cDNA clones. The carboxyl terminal domain exhibits both hydrophobic and hydrophilic stretches, as characteristic of most globular proteins, whereas the amino terminus does not contain these characteristics. Saccharide binding characteristics of the S-type lectins All the S-type lectins share a higher affinity of binding for Lac than for Gal. The critical determinants within the disaccharide are the hydroxyls at position 4 and 6 of Gal and position 3 of Glc, since substitution at any of these positions greatly reduces binding. The addition of an acetamido group at position 2 of the glucose in the Lac molecule (i.e. to yield N-acetyllactosamine) increases the binding affinity of the lectins for these sugars. Site directed mutagenesis has been carried out on the bovine and human 14 kDa lectin to determine the critical residues for saccharide binding. 20 Hirabayashi and Kasai showed that, for the human lectin, neither the cysteine residues nor the conserved tryptophan residue are essential for saccharide binding (85). In contrast, Abbott and Feizi’s results indicate that changing the tryptophan or cysteines either greatly reduces or eliminates the binding of the bovine lectin to Lac. In addition, deletion mutation analyses predict that almost the entire bovine lectin polypeptide chain is necessary for binding activity (86). The varying results obtained by the two groups point to the possible inefficacy of these techniques in analyzing structure-function relationships for the lectins. The L-30 lectins can be distinguished from the L-14 lectins by their preferential binding (approximately 100-fold greater) for the blood group A tetrasaccharide (62). This results from the substitution of the GalNAca 1- at the 3 position of the Gal of the lactose moiety. Within the L—3O family, it seems that the hamster lectin displays a slightly different binding specificity than do RL-29 and HL-29 (87). Whereas substitution at the C6 of the terminal galactose of reactive saccharides eliminates binding of the 29 kDa lectins, it appears to have little effect on the hamster lectin. Whether the differences in binding characteristics of these lectins is due to experimental conditions, species diversity, or protein isoforms is open to question. Sfitbcellular localization of the L-14 and L-301ectins The L-14 and L—30 lectins are found in the nuclear, cytosolic, and extracellular compartments (see references 88 and 240 for a review). The extracellular localization is difficult to explain since the proteins do not contain a 21 signal sequence for secretion. I will discuss the implications of their dual localization below. The predominant portion of the L-14 proteins is found intracellularly, although reports vary regarding the nuclear and cytoplasmic distribution of the lectins. Using antibodies directed against CLL-l and BHL-l to label cryostat sections, staining is noted both in the nucleus and the cytoplasm (39,89). Immunoelectron microscopy has localized the 14 kDa lectin in the nucleus of the epidermal cells of the intermediate layer of chick embryonic skin (90). However, the lectin was not found in the nuclei of the basal cells of chick embryonic skin (90). Immunolocalization of the L-14 lectin in murine myoblasts shows that the lectin is only in the cytoplasm, and not in the nucleus (91). Finally, in neuronal cells, RL-14.5 is found in both the cytoplasm and the nucleus (50). In addition to being found intracellularly, there is considerable evidence for the extracellular localization of the L-14 lectins. It has also been demonstrated that the extracellular localization can arise as a result of a specific stimulus. As an example, the L-14 lectin in Xenopus Iaevis skin tissue is found in the cytoplasm of granular and mucous gland cells. Upon the injection of epinephrine, the lectin is externalized using a novel secretory method (92). Similarly, the 14 kDa lectin in chick embryonic muscle is found intracellularly, but upon maturation of the organism, is exported from polynucleated myotubules (68). In an identical situation, the 14 kDa lectin in mouse cultured myoblasts is found both intracellularly and extracellularly. However, as the cells fuse to form multinucleate myotubules, the lectin is less abundant in the cytoplasm, and is found in vesicles in 22 the extracellular milieu (91). The export of both of these lectins is proposed to occur by a novel secretory mechanism, whereby the protein is packed into vesicles which "bud Off'. The process has been termed ectocytosis (69). An L-14 lectin (mGBP) has been purified from mouse embryonic fibroblast conditioned media and identified as a growth inhibitory substance (47), providing another example of a secreted S-type lectin. Similarly, the L-30 lectins are found on the cell surface and inside the cell. Since the proteins are identical or homologous to each other, presumably information pertaining to one member can also be applied to the Others. CBP35 has been found mostly intracellularly, although a minor fraction is cell surface localized (93). Anti-CBP35 antibody staining of the proliferating cell shows a prominent staining of the nucleus and variable staining Of the cytoplasm (94). Within the nucleus, there is a punctate staining pattern, which can be eliminated by prior treatment with RNase (95). Subcellular fractionation studies with EBP have also indicated that the majority of the protein is found in the cytoplasm and nucleus (96). Immmunocytochemical studies have shown that RL-29 can be detected in both the cytoplasm and the nucleus (49,50). The proteins Mac-2 and LBP were both identified by virtue of their cell surface localization. LBP can be isolated by cell surface iodination of murine macrophages, followed by laminin-Sepharose affinity chromatography (74). The Mac-2 antigen was isolated as a cell surface protein on thioglycollate-elicited macrophages (77). Subsequent experiments with the anti-Mac-Z monoclonal antibody indicate that the protein is also found in the nucleus of the P388D, 23 macrophage cell line (J.L. Wang, unpublished observations). The increased cell surface expression of the L-34 lectin has been proposed to be involved in transformation and metastases of cells (46,72,97). Cells that exhibit the greatest metastatic potential have the highest levels Of L-34 on the surface. The S-type lectins fall into a growing group of proteins which are localized in two distinct milieus, the intracellular and extracellular compartments. These include: a) proteins with known nuclear function, such as SV40 large T antigen (98), adenovirus ElA gene product (99), probasin (100), and the La RNP identified by autoimmune anti-nuclear antibodies (101); b) members Of the growth factor families as exemplified by the heparin-binding growth factors (102), and platelet-derived endothelial cell growth factor (103); and c) other proteins, including interleukin 1a and lfl (104), yeast mating a-factor (105), and CAP-50 (106), which is a member of the annexins. These proteins can be divided into two groups, those that have a signal sequence for extracellular transport, and those that do not. Within the former group are probasin, platelet-derived growth factor (PDGF), and the product of the mouse int-2 gene, which is a member of the fibroblast growth factor (FGF) family. Probasin is a rat prostatic protein which is found in secretions and in the nucleus of prostatic epithelial cells. The dual localization of probasin occurs as a result of alternative AUG-codon usage during translation, with the protein derived from the upstream AUG-codon containing a signal sequence (100). In the case Of the Int-2 oncoprotein, the N-terminally extended protein initiated at a CUG-codon is nuclear, while the downstream AUG-initiated product is found in the secretory 24 pathway (107). Alternative splicing of the transcript for the PDGF A-chain determines whether the protein will be localized to the nucleus, or contain a signal sequence for secretion (108). However, the basic fibroblast growth factor (bFGF) and the interleukins 1a and 16 have not been shown to contain a signal sequence. It has been postulated that these proteins can be externalized via a mechanism of exocytosis that is independent of the ER-Golgi endomembrane secretory pathway (104,109). This conclusion is based on results obtained from studies using drug inhibitors which are specific for the ER-Golgi pathway. Their nuclear localization is mediated through nuclear targetting signals (see section below on nuclear transport). The significance of the cell surface localization of the FGF has been questioned. Acidic FGF mutant molecules, lacking the nuclear targetting signal, failed to induce DNA synthesis and cell proliferation in target cells, even though they could initiate membrane events such as tyrosine-phosphorylation (110). Thus, it has been suggested that acidic FGF may ultimately act as an intracellular, nuclear- translocated polypeptide mitogen, therefore obviating the need for a signal sequence. A possible explanation for the export of proteins independently of the classical secretory pathway may lie in the discovery of the ATP-dependent translocators (111). In yeast S. cerewsiae, the product of the STE6 gene has been shown to be a key factor in the export of the mating a-factor (105). Mutants lacking the STE6 protein fail to export a-factor, while the mutants of the normal secretory pathway (the sec mutants) have no effect. The STE6 protein displays 25 significant homology with the P-glycoprotein, which is involved in multi-drug resistance (reviewed in reference 112), as well as with bacterial permeases (105). The STE6 protein shares about 60% amino acid identity with the mammalian MDR1 gene product. MDRl is homologous to bacterial permeases and contains an ATP-binding domain, suggesting that MDRl functions as an ATP-driven pump. The similarity of STE6 with MDRl implies that it, too, may act as a pump for the a-factor. There may be similar translocation systems for the export of proteins such as the interleukins and the FGFs. The bifunctional nature of lectins It has become increasingly Obvious that the physiological role of many lectins extends further than just the binding of carbohydrates. This has been demonstrated for the C-type lectins (Figure 1), as well as for the S-type lectins (Figure 2). I will discuss two examples below. a) Selectins One of the most exciting discoveries in the field of lectins has been that of the family of cell-adhesion proteins called selectins. Their nomenclature is as follows: a) L-selectin (peripheral lymph node homing receptor), also known as gp190MEL, LAM-1, LECAM-l, LECCAM-l, DREG.56, TQ-1, and Leu-8; b) E- selectin, also known as ELAM-l; and c) P-selectin, also known as PADGEM, GMP-140, and CD62. These proteins were initially isolated as important mediators of adhesion of leukocytes to the blood vascular compartment. The presence of a Ca2+ dependent CRD within these molecules was discerned only 26 after they were cloned and sequenced (10,11,12). Structurally, these proteins are very similar, with an identical arrangement of the C-type lectin domain attached to an EGF-like and a complement binding domain. In addition, there is a signal sequence, a transmembrane domain, and a cytoplasmic tail. It has been proposed that the selectin gene arrangement arose from exon shuffling mechanisms (113). L-selectin is a surface antigen on lymphocytes which facilitates their binding specifically to lymph node endothelium during lymphocyte circulation. It is, hence, commonly called the lymphocyte homing receptor since it allows for selective trafficking of particular lymphocyte populations to specific sites. P-selectin is a glycoprotein found on the surface of platelets and endothelial cells after stimulation by thrombogenic agents, thereby allowing these cells to bind to neutrophils and monocytes at areas of tissue injury. The E-selectin is generated by endothelial cells as a result of inflammatory agents, and promotes adhesion Of neutrophils, monocytes, and a subpopulation of lymphocytes to the endothelial cells. Tentative carbohydrate ligands have been identified for these selectins (see references 114,115 for reviews). The SSEA-l/Lewisx/CD15 antigen has been implicated as the binding determinant for P-selectin. E-selectin seems to prefer a sialylated Lewisx structure. The endothelial ligand for L-selectin appeared to be more elusive, until a recent discovery by Lasky et a]. (116). They have cloned a cDNA for a sulfated 50 kDa glycoprotein, Sgp50, which has a mucin-like domain. The protein backbone on this molecule may serve as a scaffold on which to present the carbohydrates. 27 b) The S-type L-141ectins Although there is a paucity of information on the functional ligand for the L-14 lectins, there have been several reports implicating these lectins in growth regulation. Yamaoka et a]. have reported that the overexpression of a rat 14 kDa galactose—binding-protein (GBP) causes transformation Of murine fibroblasts (117). In fact, they present information suggestingthat the GBP protein is identical to the growth regulatory factor TGFy2. Furthermore, the growth stimulatory activity of TGFyZ/GBP is not inhibited by the addition of ,B-galactoside, implying that the two activities are distinct. The idea that GBP may be a transforming growth factor, or act like one, is strengthened by the Observation by Wells and Malluci that a galactoside binding protein (mGBP) secreted by mouse embryonic fibroblasts (MEF) is a cytostatic growth factor (47). mGBP, when added to MEF in vitro, inhibited their growth in the G0 phase of the cell cycle. The inhibitory activity was not reversed by the addition of lactose. This may be evidence that L- 14 is a multifunctional molecule, much like transforming growth factor which can both stimulate and inhibit cell growth. CARBOHYDRATE BINDING PROTEIN 35 CBP35 was originally isolated from mouse lung tissue as a monomeric protein of M, 35,000 by its affinity for galactose containing glycoconjugates (45). Crittenden et a]. demonstrated that the protein is widely expressed in a variety of species and tissues (58). It was also shown that the protein is more highly expressed in embryonic tissue than in adult tissue. CBP35 has been shown to be 28 identical or homologous to all Of the other members of the L-30 family (Table IV) (88). Immunolocalization data have shown that the majority of CBP35 is found intracellularly, although a small fraction is also found on the cell surface (93). The control of the intracellular localization of CBP35 will be discussed further below. CBP35 is speculated to be a component of the heterogenous nuclear ribonucleoprotein complex (hnRNP). This is based on the following Observations (76,95): a) CBP35 is released from permeabilized nuclei by treatment with RNase A, but not by treatment with DNase I; b) CBP35 is found in the same density fractions as the hnRNP proteins on cesium sulfate gradients (~ 1.30 g/ml); c) CBP35 co-isolates with hnRNP proteins by sucrose gradient centrifugation (40 S); d) fractionation of nucleoplasm on a galactose affinity column yields CBP35 and a set of polypeptides whose molecular weights match those for the hnRNP proteins; and e) sequence analysis of the cDNA clone for CBP35 suggests a homology between CBP35 and some hnRNP proteins. The hnRNP proteins are thought to aid in the processing and transport of mRNA. Further support for the involvement of CBP35 in this process comes from the evidence that antibodies directed against CBP35 and galactose-containing saccharides perturb the in Vitro splicing of pre-mRNA (Patterson, et 31., unpublished data). Structure of the CBP35 gene and protein There is a single gene for CBP35 in the mouse genome. The gene encompasses approximately 9 kb of genomic DNA, and is comprised of five exons 29 and four introns (Jia, S., and Wang, J.L., unpublished observations). The upstream promoter region contains the TATA and CCAAT sequences, as well as the serum response element sequence about 200 nucleotides upstream Of the transcription start site. A polyadenylation signal has been delineated in the 3’ untranslated region. Since the CBP35 protein has been only Observed as a monomeric species, its hemagglutination property is difficult to explain. The stoichiometry of saccharide binding by the CRD was recently determined by equilibrium dialysis, using [”C]-lactose and recombinant L-30 expressed in and purified from E. 0011'. The results indicate one Lac binding site per M, 30000 of protein (Knibbs, R., manuscript in preparation, and ref.118). WOO et 31. suggest that CBP35/Mac-2 forms intermolecular dimers using the single cysteine residue (119). However, since their study used a non-reducing SDS-PAGE system, the question arises whether the dimer that they detect is actually an artifact Of the analysis process (120). The IgE binding protein (eBP) has also been shown to form oligomers using crosslinking reagents (118). Hsu et al. suggest that the N-terminal domain of the EBP may contribute to the multivalency of the molecule, possibly by engaging in protein-protein interactions (118). These results are refuted, however, . by the Observation that CBP35 does not form oligomers under non-denaturing conditions as determined by HPLC gel filtration columns and equilibrium sedimentation centrifugation (Anandita, et al., manuscript in preparation). 30 Ligands for CBP35/Mac-2 Mac-2, the CBP35 homolog, has been shown to bind to two intestinal epithelial glycoproteins, M2BP-1 (M, 98 kDa) and M2BP-2 (M, 70 kDa). These proteins were isolated from a human adenocarcinoma cell line by virtue of their association with Mac-2 (121). The interaction between these proteins and Mac-2 is mediated through the carbohydrate binding portion of Mac-2. The M2BP-1 protein is secreted into the media, leading the authors to speculate that it is an extracellular ligand for the surface antigen Mac-2. However, no glycosylated nuclear ligand has yet been found for CBP35/Mac-2. It remains to be seen if the intracellular pool of CBP35 mediates its function through its sugar binding activity. Proliferation dependent localization and expression of CBP35 Initial studies using indirect immunofluorescence techniques showed that quiescent or serum starved 3T3 cells exhibited very low levels of CBP35. The protein was found only in the cytoplasm of these cells. On the other hand, proliferating or serum fed cells demonstrated a significant increase in the levels of CBP35, and in this case the protein was localized largely in the nucleus (94). Upon serum stimulation, the protein levels increase in the G1 phase of the cell cycle, before the onset of DNA synthesis. These observations have been corroborated by immunoblotting. The results at the protein level prompted us to examine the transcriptional regulation of the CBP35 gene. The conclusions are presented in Chapter II of this thesis. One of the reasons why CBP35 may be stimulated by the addition of serum is the presence of a serum response element 31 (SRE) in the upstream regulatory region of the gene. The SRE has been shown to operate in the serum-mediated activation of the c-fos and fi-actin genes (reviewed in references 122,123). Two dimensional gel electrophoresis analyses of quiescent and proliferating populations of cells revealed that CBP35 exists as two isoelectric variants in the cell. The unmodified polypeptide has an isoelectric point (pl) of 8.7, whereas the singly phosphorylated derivative has 3 pl of 8.2. The pl 8.2 form in found both in the cytoplasm and the nucleus. The pI 8.7 form is restricted to the nucleus. More interestingly, quiescent cells only express the 8.2 variant. But, when cells are proliferating, the levels of the 8.7 variant increase dramatically and it is entirely nuclearly localized (124). A similar situation is encountered with SL66 normal human fibroblasts. Using immunofluorescence, immunoblotting, and two dimensional electrophoresis analyses with these cells, it has been revealed that the expression of CBP35 is dramatically different in young passage (passage 11) as compared to old passage (passage 31-35) cells. When young SL66 cells are serum stimulated, there is a rise in the levels of CBP35. Both isoelectric variants are expressed in these cells and the pI 8.7 form dramatically increases upon serum addition. On the other hand, Old cells did not exhibit any increase in CBP35 levels, and they also did not contain any pl 8.7 form (84). The absence of any CBP35 induction in old cells raises the intriguing possibility that certain induction mechanisms are abrogated in senescence. It has been suggested that a similar fate may occur for the c-fos protein (see below). 32 The significance of the two isoelectric variants remains to be elucidated. However, the compartmentalization of the two forms suggests that the phosphorylation may serve as a partitioning signal. I will discuss this aspect further below in the section on nuclear transport of proteins. THE REGULATION OF EXPRESSION OF C-FOS Introduction to Primary Response Genes It is generally acknowledged that control of vertebrate cell proliferation is exerted largely in the G1 phase of the cell cycle. The complex cellular process of proliferation is initiated by an interaction between extracellular factors and specific cell surface receptors (125). The cytoplasmic activation signals thus engendered cross the nuclear membrane and alter the expression of a set of genes known as immediate early (IE) or primary response genes (PRG). The PRGs are characterized by the following: a) they do not require de novo protein synthesis for induction; and b) their rapid and transient induction occurs in a wide variety of cell types (reviewed in references 126,127). Within the ranks of the PRGs are genes which encode structural proteins, transcription factors, and proto-oncogenes (125). Many cDNA clones for PRGs have also been identified by subtractive hybridization screening, although their products have not yet been characterized (128,129,130). Several of the proto- oncogenes have been well studied, although their exact function is unknown. The product of the c-fos gene has been identified as a transcription factor. 33 The proto-oncogene c-fos is a paradigm for the regulation of proto- oncogenes. In addition, the CBP35 gene and protein share several similarities with the c-fos gene and F08 protein: a) they both satisfy the criteria for a PRG; b) CBP35 contains a DNA promoter element which fits the consensus sequence for the serum response element, also identified in c-fos; c) both are nuclear proteins; and d) quiescent and senescent cells express negligible levels of F08 and CBP35. I shall, therefore, discuss the expression of c-fos with respect to the cell cycle. There are four members of the fos family: c-fos, fosB, fra-l, and Ira-2 (125). The FOS proteins are found associated with the JUN proteins. Proteins within the jun family include c-jun, junB, and junD (125). The PCS and JUN proteins form heterodimers, through leucine zippers, in all possible combinations (131,132). The resulting dimer then forms the AP-l transcription factor (131,132). The F08 and JUN proteins can also dimerize with proteins from the CREB/ATF family using the leucine zipper (133). DNA binding by the AP-l factor occurs by a region adjacent to the leucine zipper which is rich in basic amino acids. Those proteins containing the dual motifs of the basic amino acids and the leucine zipper have been clumped into the "bzip" family of transcription factors (134). Although the FOS/JUN proteins are generally regarded as transcription site AP-l binding proteins, they can also bind to the TRE (TPA response element) and the CRE (CAMP response element) sites (133,135). 34 Kinetics of the expression of c-fos The c-fos gene product is a M, 55,000 nuclear phosphoprotein (p55°"°‘) (136,137). The mRNA transcript has a size of 2.2 kb. Studies with the expression of C-fos show that the levels of the mRNA and protein are transiently high in proliferating cells (136,137). There is increased expression in placental and extraembryonal tissues (138). Experiments with cell cultures have established that c-fos mRNA is detectable within 15 minutes of serum addition to quiescent cells, thus making this the earliest PRG identified so far (136,137). At least part of the increase in the mRNA can be attributed to an increase in the transcription rate (139). The peak of expression of the mRNA at about 60 minutes is followed by a rapid decay (136). The FOS protein has been reported to have a tm" 2 hours (136). The presence of multiple AUUUA sequence motifs in the 3’ noncoding region of the c—fos mRNA probably accounts, in part, for its instability (140). Removal of these sequences has been shown to cause transformation by c-fos. Senescent fibroblasts exhibit very low levels of the c-fos mRNA, even when serum stimulated (141). The results obtained by Seshadri and Campisi indicate that the c-fos gene is subject to transcriptional repression in these cells. Superinduction of the c-fos mRNA The c-fos mRNA exhibits a phenomenon, common to all the PRGs, known as superinduction. Although normally very transient, the level of these transcripts is elevated when cells are incubated with protein synthesis inhibitors. Several reasons have been postulated to explain this occurrence. First, it is possible that 35 the labile transcripts are stabilized. Indeed, the c-fos mRNA half life increases from 9 minutes to several hours (140). This may be accomplished by the loss Of labile RNAses (142), by the shielding of mRNAs which stay trapped on polysomes (128), or because the concurrent translation of fos is necessary for mRNA degradation (142). Second, it is possible that the continued synthesis of the F08 protein is required for transcriptional shut-off of the gene. This implies that the F08 protein has an autorepressor function. It has been proposed that the autorepression may occur through the serum response element (143). Finally, a third explanation is that labile repressors normally keep the c—fos gene inactive in quiescent cells. When protein synthesis ceases, the gene is available for transcription. Although no such repressors have been identified, there are hints that such an effect may be mediated through the serum response element-serum response factor complex (144). In a recent study by Edwards and Mahadevan on the possible mechanism of superinduction of c—fos, the notion of the labile repressor has been questioned. Using several protein synthesis inhibitors, they find that anisomycin and cycloheximide cause the transcriptional induction (as determined by nuclear run-on assays) of c-fos mRNA (145). Thus, in addition to inhibiting protein synthesis, these compounds also act as nuclear signalling agonists. The c-fos promoter Analysis of the upstream promoter region of the c-fos gene reveals multiple regulatory sites. The serum response element (SRE), located at -300 bp, has been 36 hypothesized to be a target of many growth factors (reviewed in reference 123). There is a CAMP response element (CRE) at -60 bp which is regulated by elevated levels of CAMP or Ca2+ (146). A TPA response element (TRE) is located at -295 bp (147). A region at -345 bp is thought to be involved in induction by platelet derived growth factor (PDGF), although no protein binding factor for the site has been seen (148). A site called the SRE-2, at -276 bp, is involved in induction by nerve growth factor (NGF); the proteins binding to this region are different from those binding to the SRE (149). In addition to the positive regulatory elements, the retinoblastoma protein down-regulates c-fos expression by binding to a site at -90 bp (150). There is also evidence that cooperation between the SRE and a fosATF/API sequence downstream from the SRE leads to either repression or activation from the c-Ios promoter, depending on the growth state of the cell (151). Finally, an intragenic sequence at the exonl- intronl boundary is involved in the blocking of transcription elongation (15 2). Interestingly, many of the regulatory sites to which the F08 protein binds are present in the c-fos gene. The SRE in the fos promoter has been studied in great detail. It is a 20-bp region of dyad symmetry which is the site for the binding of the serum response factor (SRF). The 14-bp inner core element of the SRE, CC(A/T)6GG, is sufficient for the binding of the SRF and the induction of serum-stimulated transcription (153). The 211268 gene has 4 SREs, none of which show symmetry outside of the core, but each of the four confer serum inducibility on reporter genes (154). The outer palindromic arms of the SRE may enhance the binding of 37 the SRF. Alternatively, these arms may also serve as binding sites for other regulatory factors, which may act in concert with the SRF. The SRF itself is also a member of the PRG group (155). SRF is a 62-67 kDa protein, depending on its state of phosphorylation, which binds the SRE as a dimer. The phosphorylated SRF has a higher binding affinity for DNA (reviewed in reference 238). The protein is also modified by an O-linked N- acetylglucosamine moiety (156). The p62/temary complex factor (p62/T CF ) interacts with the SRF when it is bound to the SRE, thus enhancing the binding between the SRE and the SRF (156). In yeast, another protein called the SRF accessory protein-1 (SAP-1) is recruited to the SRE-SRF complex (157). Although SAP-l does not seem to be identical to p62/T CF, it does appear to be structurally and functionally related. Factors controlling the induction and function of [as As indicated by analysis of the promoter, mm is subject to both negative and positive regulation. It has been demonstrated that in quiescent cells the transcription of the c-fos gene is induced by the addition of excess copies of the tbs promoter element, thus suggesting that some negative regulatory factor can be competed out. Likewise, in proliferating cells, addition of the same element reduces c-fos transcription, presumably by competing out a positive regulatory factor (158). The activation of c-fos through the SRE can occur by at least two distinct intracellular pathways, one that is protein kinase C (ka) dependent, and another 38 that is ka-independent. There is also a third CAMP-dependent pathway which seems to act independently of the SRE (159). Therefore, the SRE is necessary and sufficient for activation by protein kinase-C, but not for activation by CAMP. A specific inhibitor of FOS/JUN proteins has been identified from nuclear and cytoplasmic extracts (160). IP-l associates with F08 and JUN and prevents the proteins from binding to DNA. It is unclear how IP-l exerts its control over FOS/JUN activity. Phosphorylation of IP-l, possibly by protein kinase A, causes the inhibitor to dissociate from the proteins. This can be seen by an increase in AP-l activity. The presence of a cytoplasmic inhibitor of FOS implies that cells always contain a basal level of the protein. Although this has not been rigorously determined, Bravo et a]. have reported that c-fos is inducible at low levels throughout the cell cycle (161). This suggests that IP-1 may have a specific role in sequestering FOS until the appropriate time. Finally, another mode of regulation of FOS may occur by the formation of disulfide bonds within the basic residues of a FOS-JUN dimer. This prevents it from binding to the TRE sequence (162,163). Regulation of transcription by the PRG Many genes are known to be either activated or repressed by FOS, FOS/JUN, or the FRA proteins (reviewed in ref. 125). It is of great interest to examine the regulatory activity of FOS on other genes as well as on itself. The repressor activity of FOS does not require it to be a heterodimer. However, the repressor activity is dependent upon the presence of the SRE sequence in the target promoter, and the phosphorylation of the serine residues in the carboxyl 39 terminus of FOS (164,165,238). Therefore, the PRGs not only regulate a wide variety of target genes, but also themselves. This indicates the complexity Of the cross talk that occurs at molecular levels in transcriptional regulation. NUCLEAR TRANSPORT General characteristics of nuclear transport Proteins and RNA enter and exit the nucleus in a specific and regulated manner. In this review I will specifically discuss the import of macromolecules into the nucleus. The nuclear envelope, which renders the nuclear compartment distinct from the cytoplasm, is a complex assembly consisting of inner and outer nuclear membranes, nuclear pore complexes, and the nuclear lamina (reviewed in reference 166). The outer membrane of the envelope is contiguous with the endoplasmic reticulum (ER), and the perinuclear space created by the outer and inner membranes is continuous with the ER. The inner and outer membranes are joined at the nuclear pore complex. It has been suggested that nuclear pores can accommodate the passive entry of proteins S M, 20-40 kDa. Larger proteins enter the nucleus by a facilitated process (reviewed in references 167,168,169). Facilitated nuclear transport has the following characteristics: a) it requires the presence of a nuclear localization signal (NLS) which is both necessary and sufficient for transport; b) presumably requires a receptor; C) requires ATP; (1) it is temperature dependent; 40 and e) proteins enter in a folded state. The diffusion of proteins through the pore complex is an open issue. There are data that suggest that nuclear proteins, regardless of size, enter the pore complex in a specific manner. As an example, histone H1 (M, 21 kDa) and histone H2B (M, 13.8 kDa), despite their small size, enter the nucleus using mechanisms which are distinct from simple diffusion (170,171). It has been also been shown that some proteins can enter the nucleus by a "piggyback method" where they are CO-transported with an NLS-containing protein (172). Nuclear import can be divided into two discrete steps: a) a relatively rapid binding of proteins to the nuclear pore; and b) a slower translocation step into the nucleus (173). The translocation step is sensitive to the lack of ATP, and is inhibitable by the inclusion of the N-acetylglucosamine-specific lectin wheat germ agglutinin (WGA) in the transport reaction (174). Nuclear localimtion signal The nuclear localization signals of many nuclear proteins have been compiled and compared in several review papers (168,175). There appears to be no consensus sequence. However, there are certain characteristics which are Obviously specific enough to ensure that only nuclear proteins enter the nucleus. These are as follows: a) they are short sequences, usually no more than 8-10 amino acids; b) the NLS may be contained within one sequence or may be divided into a bipartite signal; C) they contain a high proportion of basic amino acids, usually lysine and arginine; (1) they can occur at any site in the polypeptide; e) 41 NLSs are retained following transport; and f) a protein may contain more than one NLS. The well-characterized NLS of the SV40 large T antigen, PKKKRKV, has been regarded as the classic unipartite signal. The lys128 residue in the SV40 NLS is the critical residue in determining whether the NLS is sufficient for transport (176,177). On the other hand, nucleoplasmin, a major nuclear protein of Xenopus oocytes and embryos, requires a bipartite signal for import (178,179). The bipartite signal is located at the terminus of a 16 amino acid stretch in the carboxyl end of the protein. The recent elucidation of an NLS for an agrobacterium protein shows that it is homologous to that of nucleoplasmin (180). This indicates that the NLS structure has been conserved in evolution. The nuclear pore complex The nuclear pore complex (NPC) is an aqueous channel which acts as a molecular sieve that spans the nuclear envelope. The functional diameter of the pore is 7-10 nm, thus allowing for the diffusion of ions and small molecules (166). However, particles of sizes up to 280 A have been shown to traverse through the NPC. Studies by Feldherr and Akin have suggested that the permeability of the NPC may be linked to the physiological state of the cell (181,182). Using nucleoplasmin-coated colloidal gold particles, they have demonstrated that proliferating cells can transport particles of 230-250 A, whereas growth arrested cells can only transport particles of 160-200 A It has been estimated that the nuclear envelope of a eucaryotic cell contains approximately 2000-4000 pores (183) 42 The architecture of the pore complex consists of three prominent substructures: nuclear and cytoplasmic rings or annuli, central spokes, and a central plug (reviewed in reference 166). These structures are complexed with proteins to form the pore complex. The mass Of the nuclear pore complex has been estimated at 120 mDa. Two major classes of pore proteins have been identified. These are: a) the integral membrane protein gp210; and b) the nucleoporins. The integral membrane protein gp210 contains N-linked high mannose sugar modification. Primary structure and antibody epitope mapping studies suggest that gp210 spans the nuclear membrane once. A short region in the carboxyl terminus protrudes towards the pore, with the rest of the protein being found in the perinuclear space (184). The microinjection of anti-gp210 antibody into the ER, which is continuous with the perinuclear space, drastically reduced nuclear import (185). This has raised the question whether the NPC can be regulated via the ER. Nucleoporins are proteins containing the O-linked N-acetylglucosamine (GlcNAc) sugar which are found on both the cytoplasmic and nuclear faces Of the NPC (reviewed in reference 186). Three nucleoporins, p62 (187), the 110 kDa yeast NSPl (188), and the 130 kDa yeast NUPl (189) have been studied in some detail. It is uncertain whether the yeast nucleoporins are modified with O-GlcNAc (189). A monoclonal antibody developed against mammalian p62 cross-reacts with the yeast nucleoporins (189,190). Since there is no primary sequence homology between p62 and NUPl, it is assumed that the anti-p62 antibodies recognize some secondary structural feature. 43 Perturbation of transport by anti-nucleoporin antibodies and the GlcNAc- specific lectin WGA has implicated nucleoporins in the transport process. The addition of monoclonal antibodies against nucleoporins blocked nuclear import and RNA export (191). WGA inhibits the facilitated import of proteins into the nucleus, although diffusion through the pores is unaffected (192). Finlay and Forbes have demonstrated that WGA can deplete the nuclear transport capabilities of an extract, but that reconstitution with the WGA-bound fraction restores nuclear transport (193). The analysis of the WGA-bound fraction from rat cytosol yields three proteins of M, 62 kDa, 58 kDa, and 57 kDa, which seem to be a part of a ~600 kDa complex (194). Cytosolic factors are required for nuclear transport The nuclear pore complex does not contain all the components necessary for transport, as evidenced by the inability of isolated nuclei to support nuclear import with fidelity (195,196). Using an in Vitro system involving digitonin- permeabilized cells, Adam et a]. have shown that transport requires soluble cytosolic factors (196). These factors are found in both nucleate and anucleate cell extracts, and cannot be pelleted by centrifugation at 100000 x g. No RNA component is involved in the cytosolic factor, since the extract can be treated with micrococcal nuclease with no adverse effects. Treatment of the cytosol with the sulfhydryl alkylating agent N-ethylmaleimide (NEM) leads to an inhibition of transport. Also, cytosolic extracts from one species can support in vitro nuclear transport with another cell system, indicating that the factors are not species- 44 specific. Newmeyer and Forbes have isolated two factors from Xenopus oocyte extract which no longer support nuclear transport after NEM treatment (197). The factors have been named NIF-l and NIP-2. The factor NIF-l is required for the binding of the nuclear protein to the pore, and thus may act as a cytoplasmic carrier protein that binds to the NLS. Using the nuclear pore glycoproteins as an affinity matrix, Sterne-Marr et a]. have depleted cytosol Of transport factors such that nuclear transport was reduced ~80% (198). The factors did not exhibit binding to WGA, and also, were not sensitive to NEM inactivation. These factors may also act as docking proteins with the nuclear pore complex. In agreement with the two step model of nuclear transport, Moore and Blobel have fractionated oocyte transport extract on a DE-52 anion exchange column and isolated two fractions, A and B, both of which are required for efficient transport (199). Fraction A seems to be involved in the NLS recognition; addition of fraction A alone leads to accumulation of the protein around the nuclear periphery. Fraction B is necessary for translocation into the pore, and cannot by itself support nuclear transport. Fraction A is NEM-sensitive, whereas fraction B is not affected by NEM. Thus, the data from the various laboratories indicate that there are several different cytosolic components which are involved in the transport of proteins into the nucleus. Finally, hsp70 (heat shock protein 70) and its cognate hsc70 have been shown to be required for the transport Of proteins into the nucleus (200). The hsp70 is not NEM sensitive. 45 Hsp70 has been found to be necessary for the translocation of proteins into mitochondria and the endoplasmic reticulum (201,202). In addition, transport into the ER and mitochondria requires signal sequences and receptors. It has also been documented that NEM inactivates transport into the Golgi (203). Together, these requirements suggest that there are certain basic principles for the translocation of proteins within the cell (fora commentary, see reference 236). NLS binding proteins The existence of the NLS binding proteins (NBPs) has been inferred from the demonstration that nuclear transport is a saturable event (204,205). The presence of several of these adaptor molecules or NBPs has been substantiated by their purification using the SV40 T antigen N LS as an affinity matrix (204,206,207). The NBPs may either stay fixed to the nuclear pore, or alternatively, they may act as shuttling molecules between the cytoplasm and nucleus (167,169,208). A mammalian 55 kDa protein has been found to stimulate in vitro transport while demonstrating a sensitivity to N-ethylmaleimide (204,207). It is unclear whether the 55 kDa protein is related to the cytosolic factor NIF-l. A 70 kDa protein, isolated from yeast, has been localized to the nuclear envelope, and it has been suggested to interact with the nucleoporin NSPl (169,209,210). Proteins which are immunologically cross-reactive with the yeast 70 kDa protein are also found in Drosophila, Hela cells, and Z. mays. Stochaj and Silver report that these eucaryotic NBPs are phosphorylated, and this modification is necessary 46 for binding by the NLS (211). Lee and Mélese have identified a 67 kDa protein from yeast, encoded by the NSRI gene, which not only binds to the SV40 NLS, but also is localized to the nucleus and the nucleolus (212,213). The NSRl protein also contains two RNA recognition motifs, the significance of which is unclear. Meier and Blobel have reported the identification of a 140 kDa protein, Nopp140, which binds to NLSs and is also found in the nucleolus (214). The binding of Nopp140 to NLSs requires that the protein be phosphorylated (239). The dual localization of the p67 and p140 proteins has led to the debate of whether non-nucleolar proteins use the nucleolar/ribosomal protein transport pathway. This could be a demonstration of transport efficiency, since the transport of ribosomal proteins constitutes the majority of all trafficking into the nucleus/nucleolus. Following a general transport of all nuclear proteins into the nucleolus, there could be a re-routing of the proteins into the appropriate nuclear compartment. Alternative methods of nuclear transport Studies with transport of the spliceosomal U small nuclear ribonucleoproteins (snRNPs) have revealed the existence of other methods of nuclear import. The transport of these molecules into the nucleus cannot be competed by the SV40 T antigen NLS, thus indicating that they use a distinct pathway (215,216). The assembly pathway of the U snRNPs involves the export of the newly synthesized RNAs into the cytoplasm where they are complexed with proteins, either of the Sm Class or the U-specific proteins. The initial export of 47 the snRNA from the nucleus into the cytoplasm may be due, in part, to the presence of a 7-methylguanosine (m7GpppG) cap at the 5’ end (217). After the snRNA is complexed with proteins in the cytoplasm, the m7GpppG structure is hypermethylated to form the 2,2,7-trimethylguanosine (m3GpppG) cap. All U snRNA species, except U6 snRNA, carry the trimethylguanosine (TMG) cap; the U6 snRNA has a y—monomethylphosphate at its 5’ end. The Sm proteins and the TMG cap have been shown to make up a bipartite NLS for some of the U snRNPs (218,219). Using microinjection of the U snRNAs in oocyte nuclei, Michaud and Goldfarb have hypothesized that these snRNPs are imported into nuclei by three different pathways (220): a) the karyophilic pathway employed by SV40 large T antigen; b) the TMG dependent pathway; and C) a pathway distinct from the other two. The import of U1, U2, U4, and U5 snRNPs can be inhibited by the injection of excess TMG cap, but not by excess SV40 T antigen NLS (220). On the other hand the U6 snRNP import is inhibited by the SV40 T antigen NLS, but not by the TMG cap (220). The U6 snRNP does not contain a consensus Sm- protein binding site, but is bound by U6-specific proteins (221). The U3 snRNP also does not contain Sm proteins, but does contain a TMG cap. Its import is inhibited neither by free TMG cap, nor by excess SV40 T antigen NLS (220). Despite the various pathways used in the nuclear import of these U snRNPs, they all cross into the nucleus through the nuclear pore complex. Evidence for this notion resides in the observation that anti-nucleoporin antibody and WGA both affect the import of the snRNPs (220). The WGA inhibition of 48 transport varies, with the U6 snRNP being highly sensitive (216,220). On the other hand, it has been reported that the U1 and U5 snRNPs are not inhibited by the same concentrations of WGA as used to block U6 entry (216,220). Control of nuclear transport Two general mechanisms, not exclusive of each other, for regulating the nuclear-cytoplasmic distribution of proteins are phosphorylation-dcphosphorylation and anchor proteins (reviewed in references 222,238). Posttranslational mechanisms, such as phosphorylation, may change a protein’s conformation or mask/unmask the NLS, thus rendering the protein translocation competent. Alternatively, anchor proteins may act as scaffolds which secure the protein in the cytoplasm until the appropriate signal causes its release. The yeast transcription factor SW15 is cytoplasmic in the S, G2, and M phases of the cell cycle. The phosphorylation of three serine residues near the NLS by the CDC28 kinase is instrumental in preventing nuclear entry. Dephosphorylation of these residues at the end of M phase, when the CDC28 kinase is inactive, causes SW15 to enter the nucleus (223). Studies with the in vitro nuclear uptake of SV40 T antigen-I6- galactoside fusion proteins have shown that phosphorylation at residues away from the NLS by casein kinase II accelerates transport (224). Phosphorylation within the SV40 NLS by Cdc2 kinase reduces the nuclear accumulation of the fusion proteins (225). Finally, Xenopus oocytes contain large amounts of the protein c- MYC in the cytoplasm. The fertilization process triggers the phosphorylation of c- MYC which then leads to its nuclear localization (169,226). 49 Studies with the transcription factor NF-ICB have shown that in quiescent cells the protein is associated with its anchor IKB in the cytoplasm (reviewed in reference 227). The stimulation of cells leads to the phosphorylation of MB through a protein kinase C dependent pathway, thus resulting in the dissociation of the anchor and the entry of NF-IcB into the nucleus (228). A similar situation is seen with the gene product of the dorsal gene which controls the asymmetric expression of genes in Drosophlla. The dorsal product is secured in the cytoplasm of the embryonic cells by the cactus gene product (229,230,231). The subsequent entry of the dorsal protein into the nucleus is influenced by many other gene products (reviewed in reference 232). The C-FOS protein has been reported to be cytoplasmic in serum-starved cells but nuclear in serum-stimulated cells. The regulation of c-FOS nuclear transport has been suggested to depend on two factors: a labile control protein, and mediation by the CAMP dependent protein kinase (233). Although the labile inhibitor of transport has not been identified by Roux et al., it could be the IP-1 inhibitor protein of FOS/JUN (160). Positive regulation of nuclear transport has been reported for the pancreas- specific transcription factor PTFI. There are two forms of PTFl, a and [3, which are distinguished by their subcellular localization. The only difference between the two forms is the association of a protein, p75, with the a form. PTFla is found in both the nucleus and the cytoplasm, while PTFlfl is found exclusively in the cytoplasm. Based on these localization data, it has been hypothesized that the stable association of p75 with PTFla allows the protein to enter the nucleus (234). 10. 11. 12. 13. 14. 15. 16. 17. 50 LITERATURE CITED Sharon N & Lis H (1989) Science 246: 227-234. Liener IE, Sharon N & Goldstein IJ, eds (1986) The Lectins: Properties, Functions and Applications in Biology and Medicine. Academic Press, Orlando. Goldstein IJ, Hughes RC, Monsigny A, Osawa T & Sharon N (1980) Nature 285: 66-68. Etzler ME (1985) Ann Rev Plant Physiol 36: 209-234. Sharon N & Lis H ( 1989) in Lectins. Chapman and Hall, London. DriCkamer K ( 1989) in Carbohydrate Recognition in Cellular Function. Bock G & Harnett S, eds. Wiley & Sons, Chichester, UK. DriCkamer K (1988) J Biol Chem 263: 9557-9560. Kohnke-Godt B & Gabius H-J (1989) Biochemistry 28: 6531-6538. Kojima K, Ogawa HK, Seno N & Matsumoto I (1992) J Chromatog 597: 323-330. Lasky LA, Singer MS, Yednock TA, Dowbenko D, Fennie C, Rodriguez H, Nguyen T, Stachel S & Rosen SD (1989) Cell 56: 1045-1055. Bevilacqua MP, Stengelin S, Gimbrone Jr. MA & Seed B (1989) Science 243: 1160-1172. Johnston GI, Cook RG & McEver RP (1989) Cell 56: 1033-1044. Ashwell G & Harford J (1982) Ann Rev Biochem 51: 531-554. Hoyle GW & Hill RL (1988) J Biol Chem 263: 7487-7492. Lennartz MR, Wileman TE & Stahl PD (1987) Biochem J 245: 705-711. Taylor ME & Summerfield JA (1986) Clin Sci 70: 539-546.. Rust K, Grosso L, Zhang V, Chang D, Persson A, Longmore W, Cai G-Z & Crouch E (1991) Arch Biochem Biophys 290: 116-126. 19. 20. 21. 22. 23. 24. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 51 Halberg DF, Proulx G, Doege K, Yamada Y & DriCkamer K (1988) J Biol Chem 263: 9486-9490. Cooper DN, Lee S-C & Barondes SH (1983) J Biol Chem 258: 4698-4701. Komano H, Mizuno D & Natori S (1987) J Biol Chem 255: 2919-2924. Muramoto K & Kamuja H (1986) Biochim Biophys Acta 874: 285-295. Giga Y, Ikai A & Takahasi K (1987) J Biol Chem 262: 6197-6203. Suzuki T, Takagi T, Furukohri T, Kawamura K & Nakauchi M (1990) J Biol Chem 265: 1274-1281. Weis WI, Kahn R, Fourme R, DriCkamer K & Hendrickson WA ( 1991) Science 254: 1608-1615. Bezouska K, Crichlow GV, Rose JM, Taylor ME & DriCkamer K (1991) J Biol Chem 266: 11604-11609. Taylor ME, Bezouska K & DriCkamer K (1992) J Biol Chem 267: 1719- 1726. Ali N & Salahuddin A (1989) Biochim Biophys Acta 992: 30-34. DriCkamer K, Mamon JF, Binns G & Leung JO (1984) J Biol Chem 259: 770-778. Halberg DF, Wager RE, Farrell DC, Hildreth IV J, Quesenberry MS, Loeb JA, Holland EC & DriCkamer K (1987) J Biol Chem 262: 9828-9838. Oda S, Sato M, Toshima S & Osawa T (1988) J Biochem 104: 600-605. Patthy L (1988) Biochem J 253: 309-311. Levi G & Teichberg VI (1981) J Biol Chem 256: 5735-5740. Muramoto K & Kamiya H (1992) Biochim Biophys Acta 1116: 129-132. Fink de Cabutti NE, Caron M, Joubert R, Elola MT, Bladier D & Herkovitz J (1987) FEBS Iett 223: 330-334. Beyer EC & Barondes SH ( 1982) J Cell Biol 92: 23-27. Beyer EC & Barondes SH (1982) J Cell Biol 92: 28—33. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 52 Barak-Briles E, Gregory W, Fletcher P & Kornfeld S (1979) J Cell Biol 81: 528-537. Sakakura Y, Hirabayashi J, Oda Y, Ohyama Y & Kasai K (1990) J Biol Chem 265: 21573-21579. Beyer EC & Barondes SH (1980) J Supramol Struct 13: 219-227. Oshawa F, Hirano F & Natori S (1990) J Biochem 105: 431-434. Fitzgerald JE, Catt JW & Harrison FL (1988) Eur J Biochem 140: 137-141. Abbott WM, Mellor A, Edwards Y & Feizi T (1989) Biochem J 259: 283- 290. Carding SR, Thorpe SJ, Thorpe R & Feizi T (1985) Biochem Biophys Res Comm 127: 680-686. Merkle RA, Zhou Q, Schultz TK, Harper WB & Cummings RD (1989) Arch Biochem Biophys 274: 404-416. Roff CF & Wang JL (1983) J Biol Chem 258: 10657-10663. Raz A, Carmi P & Pazerini G (1988) Cancer Res 48: 645-649. Wells V & Malucci L (1991) Cell 64: 91-97. Leffler H & Barondes SH (1986) J Biol Chem 261: 10119-10126. Regan LJ, Dodd J, Barondes SH & Jessell TM (1986) Proc Natl Acad Sci USA 83: 2248-2252. Hynes MA, Gitt M, Barondes SH, Jessell TM & Buck LB (1990) J Neurosci 10: 1004-1013. Hirabayashi J, Kawasaki H, Suzuki K & Kasai K (1987) J Biochem 101: 775-787. Gitt MA & Barondes SH (1986) Proc Natl Acad Sci USA 83: 7603-7607. Gitt MA & Barondes SH (1991) Biochemistry 30: 82-89. Bladier D, Le Caer JP, Joubert R, Caron M & Rossier J (1991) Neurochem Int 18: 275-281. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 53 Gitt MA, Massa SM, Leffler H & Barondes SH (1992) J Biol Chem 267: 10601—10606. - Leffler H, Masiarz FR & Barondes SH (1989) Biochemistry 28: 9222-9229. Ruggiero-Lopez D, Louisot P & Martin A (1992) Biochem Biophys Res Comm 185: 617-623. Crittenden SL, Roff CF & Wang JL (1984) Mol Cell Biol 4: 1252-1259. Marschal P, Herrmann J, Leffler H, Barondes SH & Cooper DNW (1992) J Biol Chem 267: 12942-12949. Hirabayashi J, Satoh M, Ohyama Y & Kasai K (1992) J Biochem 111: 553- 555. Brassart D, Kolodziejczyk E, Granato D, Woltz A, Pavillard M, Perotti F, Frigeri LG, Liu F-T, Borel Y & Neeser J-R (1992) Eur J Biochem 203: 393-399. Sparrow CP, Leffler H & Barondes SH (1987) J Biol Chem 262: 7383-7390. Hinek A, Wrenn DS, Mecham RP & Barondes SH (1988) Science 239: 1539-1541. Abbott WM & Feizi T (1989) Biochem J 259. 291-294. Greer Wilson TJ, Firth MN, Powell JT & Harrison FL (1989) Biochem J 261: 847-852. Harrison L (1991) In Lectin Reviews Vol I. Kilpatrick DC, Van Driessche E & Beg-Hansen TC, Eds. Sigma Chemical Co, St Louis. Powell JT & Harrison FL (1991) Am J Physiol 261: L236-L239. Barondes SH & Haywood-Reid PL ( 1981) J Cell Biol 91: 568-572. Cooper DN & Barondes SH (1990) J Cell Biol 110: 1681-1691. Catt JW & Harrison FL (1985) J Cell Sci 73: 347-359. Wasano K, Hirakawa Y & Yamamoto T (1990) Cell Tissue Res 259: 43-49. Raz A, Meromsky L, Zvibel I & Lotan R (1987) Int J Cancer 39: 353-360. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 54 Poirier F, Timmons PM, Chan C-TJ, Guenet J-L & Rigby PWJ (1992) Develop 115: 143-155. Woo H-J, Shaw LM, Messier JM & Mercurio AM (1990) J Biol Chem 265: 7097-7099. Jia S, Mee RP, Morford G, Agrwal N, Voss PG, Moutsatsos IK & Wang JL (1987) Gene 60: 197-204. Jia S & Wang JL (1988) J Biol Chem 263: 6009-6011. Ho MK & Springer TA (1982) J Immunol 128: 1221-1228. Cherayil BJ, Weiner SJ & Pillai S (1989) J Exp Med 170: 1959-1972. Foddy L, Stamatoglou SC & Hughes RC (1990) J Cell Sci 97: 139-148. Laing JG, Robertson MW, Gritzmacher CA, Wang JL & Liu F-T (1989) J Biol Chem 264: 1907-1910. Oda Y, Leffler H, Sakakura Y, Kasai K & Barondes SH (1990) Gene 99: 279-283. Robertson MW, Albrandt K, Keller D & Liu F-T ( 1990) Biochemistry 28: 8093-8100. Cherayil BJ, Chaitovitz S, Wong C & Pillai S (1990) Proc Natl Acad Sci 87: 7324-7328. Hamann KK, Cowles EA, Wang JL & Anderson RL (1991) Exp Cell Res 196: 82-91. Hirabayashi J & Kasai K (1991) J Biol Chem 266: 23648-23653. Abbott WM & Feizi T (1991) J Biol Chem 266: 5552-5557. Sato S & Hughes RC (1992) J Biol Chem 267: 6983-6990. Wang JL, Laing JG & Anderson RL (1991) Glycobiology 1: 243-252. Childs RA & Feizi T (1980) Cell Biol Int Reports 4: 775. Akimoto Y, Kawakami H, Oda Y, Obinata A, Endo H, Kasai K & Hirano H (1992) Exp Cell Res 199: 297-304. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 55 Harrison FL & Greer Wilson TJ (1992) J Cell Sci 101: 635-646. Bols NC, Roberson MM, Haywood-Reid PL, Cerra RF & Barondes SH (1986) J Cell Biol 102: 492-499. Moutsatsos IK, Davis JM & Wang JL (1986) J Cell Biol 102: 477-483. Moutsatsos IK, Wade M, Schindler M & Wang JL ( 1987) Proc Natl Acad Sci USA 84: 6452-6456. Laing JG & Wang JL (1988) Biochemistry 27: 5329-5334. Gritzmacher CA, Robertson MW & Liu F-T (1988) J Immunol 141: 2801- 2806. Raz A, Zhu D, Hogan V, Shah N, Raz T, Karkash R, Pazerini G & Carmi P (1990) Int J Cancer 46: 871-877. Butel JS & Jarvis DL (1986) Biochim Biophys Acta 865: 171-195. Bellgrau D, Walker TA & Cook JL (1988) J Virol 62: 1513-1519. Spence AM, Sheppard PC, Davie JR, Matuo Y, Nishi N, McKeehan WL, Dodd JG & Matusik RJ (1989) Proc Natl Acad Sci USA 86: 7843-7847. Bachmann M, Chang S-H, Slor H, Kukulies J & Miiller WEG (1990) Exp Cell Res 191: 171-180. Sano H, Forough R, Maier JAM, Case JP, Kackson A, Engleka K, Maciag T & Wilder RL ( 1990) J Cell Biol 110: 1417-1426. Ishikawa F, Miyazono K, Hellman U, Drexler H, Wernstedt C, Hagiwara K, Usuki K, Takaku F, Risau W & Helden C-H (1989) Nature 338: 557- 562. Rubartelli A, Cozzolino F, Talio M & Sitia R (1990) EMBO J9: 1503- 1510. Kuchler K, Sterne RE & Thorner J (1989) EMBO J 8: 3973-3984. Mizutani A, Usuda N, Tokumitsu H, Minami H, Yasui K, Kobayashi R & Hidaka H (1992) J Biol Chem 267: 13498-13504. Acland P, Dixon M, Peters G & Dixon C (1990) Nature 343: 662-665. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 56 Maher DW, Lee BA & Donoghue DJ (1989) Mol Cell Biol 9: 2251-2253. Mignatti P, Morimoto T & Rifldn DB (1992) J Cell Physio] 151: 81-93. Imamura T, Engleka K, Zhan X, Tokita Y, Forough R, Roeder D, Jackson A, Maier JAM, Hla T & Maciag T (1990) Science 249: 1567-1570. Featherstone C ( 1990) Trends Biochem Sci 15: 169-170. Juranka PF, Zastawny RL & Ling V ( 1989) FASEB J 3: 2583-2592. Dowbenko DJ, Diep A, Taylor BA, Lusis AJ & Lasky LA ( 1991) Genomics 9: 270-277. Feizi T (1991) Trends Biochem Sci 16: 84-86. Springer TA & Lasky LA (1991) Nature 349: 196-197. Lasky LA, Singer MS, Dowbenko D, Imai Y, Henzel WJ, Grimley C, Fennie C, Gillett N, Watson SR & Rosen SD (1992) Cell 69: 927-938. Yamaoka K, Ohno S, Kawasaki H & Suzuki K (1991) Biochem Biophys Res Comm 179: 272-279. Hsu DK, Zuberu RI & Liu F-T (1992) J Biol Chem 267: 14167-14174. Woo H-J, Lotz MM, Jung JU & Mercurio AM (1991) J Biol Chem 266: 18419-18422. Kumar AM & Davidson VL (1992) Biotechniques 12: 198-202. Rosenberg I, Cherayil BJ, Isselbacher KJ & Pillai S (1991) J Biol Chem 266: 18731-18736. Rivera VM & Greenberg ME (1990) New Biol 2: 751-758. Treisman R (1990) Cancer Biol 1: 47-58. Cowles EA, Agrwal N, Anderson RL & Wang JL (1990) J Biol Chem 265: 17706-17712. McMahon SB & Monroe JG (1992) FASEB J 6: 2707—2715. Denhardt DT, Edwards DR & Parfett CLJ (1986) Biochim Biophys Acta 865: 83-125. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 57 Herschman HR (1991) Ann Rev Biochem 60: 281-319. Cochran BH, Reffel AC & Stiles CD (1983) Cell 33: 939-947. Lau LF & Nathans D (1987) Proc Nat] Acad Sci USA 84: 1182-1186. Mohn KL, Laz TM, Hsu JC, Melby AE, Bravo R & Taub R (1991) Mo] Cell Biol 11: 381-390. Nakabeppu Y, Ryder K,& Nathans D (1988) Cell 55: 907-915. Halazonetis TD, Georgopoulos K, Greenberg ME & Ieder P (1988) Cell 55: 917-924. Hai T & Curran T (1991) Proc Nat] Acad Sci USA 88: 3720-3724. Vinson CR, Sigler PB & McKnight SL (1989) Science 246: 911-922. Rauscher III FJ, Voulalas PJ, Franza Jr BR & Curran T (1989) Genes Dev. 2: 1687-1699. Kruijer W, Cooper JA, Hunter T & Verma IM (1984) Nature 312: 711-716. Miiller R, Bravo R, Burckhardt J & Curran T (1984) Nature 312: 716-720. Miiller R, Verma IM & Adamson ED (1983) EMBO J 2: 679-684. Greenberg ME & Ziff EB (1984) Nature 311: 433-438. Wilson T & Treisman R (1988) Nature 336: 396-399. Seshadri T & Campisi J (1990) Science 247: 205-209. Pontecorvi A, Tata JR, Phyillaier M & Robbins J ( 1988) EMBO J 7: 1489- 1495. Sassone-Corsi P, Sisson JC & Verma IM (1988) Nature 334: 314-319. Subramaniam M, Schmidt LJ, Crutchfield CE & Getz MJ (1989) Nature 340: 64-66. Edwards DR & Mahadevan LC (1992) EMBO J 11: 2415-2424. Sheng M, Dougan ST, McFadden G & Greenberg ME (1988) Mo] Cell Biol 8: 2787-2796. 147. 148. 149. 150. 151. 15 2. 153. 154. 155. 156. 15 7. 15 8. 159. 160. 161. 162. 163. 164. 165. 58 Fisch TM, Prywes R & Roeder RG (1989) Mo] Cell Biol 9: 1327-1331. Hayes TE, Kitchen AM & Cochran BH (1987) Proc Nat] Acad Sci USA 84: 1272-1276. Visvader J, Sassone-Corsi P & Verma IM (1988) Proc Nat] Acad Sci USA 85: 9474-9478. Robbins PD, Horowitz JM & Mulligan RC (1990) Nature 346: 668-671. Morgan IM & Birnie GD (1992) Cell Prolif 25: 205-215. Mechti N, Piechaczyk M, Blanchard J-M, Jeanteur P & Lebleu B (1991) M0] Cell Biol 11: 2832-2841. Rivera VM, Sheng M & Greenberg ME (1990) Genes Dev 4: 255-268. Christy BA & Nathans D ( 1989) Mo] Cell Biol 9: 4889-4895. Norman C, Runswick M, Pollock R & Treisman R (1988) Cell 55: 989- 1003. Schroter H, Mueller CGF, Meese K & Nordheim A (1990) EMBO J 9: 1123-1130. Dalton S & Treisman R (1992) Cell 68: 597-612. Sassone-Corsi P & Verma IM (1987) Nature 326: 507-510. Gilman MZ (1988) Genes Dev 2: 394-402. Auwerx J & Sassone-Corsi P (1991) Cell 64: 983-993. Bravo R, Burckhardt J, Curran T & Muller R (1986) EMBO J 5: 695-700. Bannister AJ, Cook A & Kouzarides T (1991) Oncogene 6: 1243-1250. Abate CL, Patel L, Rauscher III FJ & Curran T (1990) Science 249: 1157- 1161. Gius D, Cao X, Rauscher III FJ, Cohen DR, Curran T & Sukhatme VP (1990) M0] Cell Biol 10: 4243-4255. Konig H, Ponta H, Rahmsdorf U, Buscher M, Schonthal A, Rahmsdorf HJ & Herrlich P (1989) EMBO J 8: 2559-2566. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 59 Gerace L & Burke B (1988) Ann Rev Cell Bi014: 335-374. Silver P (1991) Cell 64: 489-497. Garcia-Bustos J, Heitman J & Hall MN (1991) Biochim Biophys Acta 1071: 83-101. Nigg EA, Baeuerle PA & Liihrmann R (1991) Cell 66: 15-22. Moreland RB, Langevin GL, Singer RH, Garcea RL & Hereford LM (1987) Mo] Cell Biol 7: 4048-4057. Breeuwer M & Goldfarb DS (1990) Cell 60: 999-1008. Zhao L-J & Padmanabhan R (1988) Cell 55: 1005-1015. Richardson WD, Mills AD, Dilworth SM, Laskey RA & Dingwall C (1988) Cell 52: 655-664. Newmeyer DD & Forbes DJ (1988) Cell 52: 641-653. Dingwall C & Laskey RA (1991) Trends Biochem Sci 16: 478-481. Kalderon D, Richardson WD, Markham AF & Smith AE (1984) Nature 311: 33-38. Kalderon D, Roberts BL, Richardson WD & Smith AE (1984) Cell 39: 499-509. Dingwall C, Robbins J, Dilworth SM, Roberts B & Richardson WD (1988) J Cell Biol 107: 841-849. Robbins J, Dilworth SM, Laskey RA & Dingwall C (1991) Cell 64: 615-623. Citovsky V, Zupan J, Warnick D & Zambryski P (1992) Science 256: 1802- 1805. Feldherr CM & Akin D (1990) J Cell Biol 111: 43-53. Feldherr CM & Akin D (1991) J Cell Biol 115: 933-939. Maul GG (1977) Int Rev cytol Supp] 6: 75-186. Greber UF, Senior A & Gerace L (1990) EMBO J 9: 1495-1502. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 60 Greber UF & Gerace L (1992) J Cell Biol 116: 15-30. Hart GW, Haltiwanger RS, Holt GD & Kelly WG (1989) Ann Rev Biochem 58: 841-874. Starr CM, D’Onofrio M, Park MK & Hanover JA (1990) J Cell Biol 110: 1861-1871. Nehrbass U, Kern H, Mutvei A, Horstmann H, Marshallsay B & Hurt EC (1990) Cell 61: 979-989. Davis LI & Fink GR (1990) Cell 61: 965-978. Aris JP & Blobel G ( 1989) J Cell Biol 108: 2059-2067. Featherstone C, Darby MK & Gerace L (1988) J Cell Bio] 107: 1289-1297. Finlay DR, Newmeyer DD, Price TM & Forbes DJ (1987) J Cell Biol 104: 189-200. Finlay DR & Forbes DJ (1990) Cell 60: 17-29. Finlay DR, Meier E, Bradley P, Horecka J & Forbes DJ (1991) J Cell Biol 114: 169-183. Newmeyer DD, Finlay DR & Forbes DJ (1986) J Cell Biol 103: 2091-2102. Adam SA, Marr RS & Gerace L (1990) J Cell Biol 111: 807-816. Newmeyer DD & Forbes DJ (1990) J Cell Biol 110: 547-557. Sterne-Marr R, Blevitt JM & Gerace L (1992) J Cell Biol 116: 271-280. Moore MS & Blobel G (1992) Cell 69: 939-950. Shi Y & Thomas JO (1992) Mo] Cell Bio] 12: 2186-2192. Chirico WJ, Waters MG & Blobel G (1988) Nature 332: 805-810. Deshaies RJ, Koch BD, Werner-Washburne M, Craig EA & Schekman R (1988) Nature 332: 800-805. Block MR, Glick BS, Wilcox CA, Wieland FT & Rothman JE (1988) Proc Nat] Acad Sci USA 85: 7852-7856. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 61 Adam SA, Lobl TJ, Mitchell MA & Gerace L (1989) Nature 337: 276-279. Garcia-Bustos JF, Wagner P & Hall MN (1991) J Biol Chem 266: 22303- 22306. Yamasaki L, Kanda P & Lanford RE (1989) Mo] Cell Biol 9: 3028-3036. Adam SA & Gerace L (1991) Cell 66: 837-847. Borer RA, Lehner CF, Eppenberger HM & Nigg EA (1989) Cell 56: 379- 390. Silver PA, Sadler I & Osborne MA (1989) J Cell Biol 109: 983-989. Stochaj U, Osborne M, Kurihara T & Silver PA (1991) J Cell Biol 113: 1243-1254. Stochaj U & Silver PA (1992) J Cell Biol 117: 473-482. Lee W-C & Mélese (1989) Proc Nat] Acad Sci USA 86: 8808-8812. Lee W-C, Xue Z & Mélese T (1991) J Cell Biol 113: 1-12. Meier UT & Blobel G (1990) J Cell Biol 111: 2235-2245. Michaud N & Goldfarb DS (1991) J Cell Biol 112: 215-223. Fischer U, Darzynkiewicz E, Tahara SM, Dathan NA, Liihrmann R & Mattaj IW (1991) J Cell Biol 113: 705-714. Hamm J & Mattaj IW (1990) Cell 63: 109-118. Hamm J, Darzynkiewicz E, Tahara SM & Mattaj IW (1990) Cell 62: 569- 577. Fischer U & Liihrmann R (1990) Science 249: 786-790. Michaud N & Goldfarb D (1992) J Cell Biol 116: 851-861. Hamm J & Mattaj IW (1989) EMBO J 8: 4179-4187. Hunt T ( 1989) Cell 59: 949-951. M011 T, Tebb G, Surana U, Robitsch H & Nasmyth K (1991) Cell 66: 743- 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 62 Rihs H-P, Jans DA, Fan H & Peters R (1991) EMBO J 10: 633-639. Jans DA, Ackermann MJ, Bischoff JR, Beach DH & Reiners P (1991) J Cell Biol 115: 1203-1212. Gusse M, Ghysdael J, Evan G, Soussi T & Méchali M (1989) Mol Cell Biol 9: 5395-5403. Baeuerle PA (1991) Biochim Biophys Acta 1072: 63-80. Ghosh S & Baltimore D (1990) Nature 344: 678-682. Rushlow CA, Han K, Manley JL & Levine M (1989) Cell 59: 1165-1177. Steward R (1989) Cell 59: 1179-1188. Roth S, Stein D & Niisslein-Volhard C (1989) Cell 59: 1189-1202. Govind S & Steward R (1991) Trends Genet 7: 119—125. Roux P, Blanchard J-M, Fernandez A, Lamb N, Jeanteur P & Piechaczyk M (1990) Cell 63: 341-351. Sommer L, Hagenbiichle O, Wellnauer PK & Strubin M (1991) Cell 67: 987-994. Castronovo V, Luylen F, van den Brfile F & Sobel ME (1992) Arch Biochem Biophys 297: 132-138. Goldfarb DS (1992) Cell 70: 185-188. Hirabayashi J, Satoh M & Kasai K (1992) J Biol Chem 267: 15485-15490. Hunter T & Karin M (1992) Cell 70: 375-387. Meier UT & Blobel G (1992) Cell 70: 127-138. Harrison FL (1991) J Cell Sci 100: 9-14. CHAPTER II CARBOHYDRATE BINDING PROTEIN 35: Levels of Transcription and mRNA Accumulation in Quiescent and Proliferating Cells Neera Agrwal, John L Wang, and Patricia G. Voss Department of Biochemistry Michigan State University East Lansing MI 48824 63 FOOTNOTES * This work was supported by Grants GM-38740 and GM-27203 from the National Institutes of Health. 1 N. Agrwal was supported by a Patricia Roberts Harris Fellowship. 2 The abbreviations used are: CBP, carbohydrate-binding protein; hnRNP, heterogeneous nuclear ribonucleoprotein complex; SDS, sodium dodecyl sulfate; BSA, bovine serum albumin; HEPES, N-2-hydroxyethylpiperazine-N’- 2-ethanesulfonic acid. 3 This work was previously published in Journal of Biological Chemistry (1989) 264: 17236-17242 and is reproduced with permission from the publisher. Copyright 1989 American Society for Biochemistry and Molecular Biology. 65 ABSTRACT In previous studies, we observed proliferation-dependent expression and nuclear localization of the lectin, designated carbohydrate-binding protein 35 (CBP35), in 3T3 fibroblasts at the polypeptide level by Western blot and immunofluorescence analysis. In the present study, we have compared the expression of the CBP35 gene in quiescent and proliferative 3T3 cells at the levels of (a) accumulated mRNA by Northern blotting and (b) nuclear transcription by run-off assays. When serum-starved, quiescent cultures of 3T3 cells were stimulated by the addition of serum, there was an increase in the nuclear transcription of the CBP35 gene and in the accumulation of its mRNA early (1-3 h) in the activation process, well before the first wave of DNA synthesis. These increases were not dependent on de novo protein synthesis inasmuch as they occurred even in the presence of cycloheximide. There was also an elevated transcription rate and mRNA level in transformed cells when compared to their normal counterparts. Finally, the expression of CBP35 was compared between sparse, proliferating cultures of 3T3 cells and density-inhibited confluent monolayers of the same cells. Although the rate of transcription of the CBP35 gene was approximately the same in the two cultures, there was a higher level of CBP35 mRNA in the dense cells. Thus, it appears that post-transcriptional mechanisms may be involved in the accumulation of mRNA. 66 INTRODUCTION Carbohydrate-binding protein (CBP) 35 (M, 35,000) was initially purified from extracts of mouse 3T3 fibroblasts on’the basis of its binding to galactose- containing glycoconjugates (1). More recent studies have suggested that CBP35 is a component of the heterogeneous nuclearribonucleoprotein complex (hnRNP). This conclusion was based on the following observations (2): (a) CBP35 was released from permeabilized nuclei by treatment with ribonuclease A but not by similar treatment with deoxyribonuclease I; (b) when nucleoplasm was fractionated on a cesium sulfate gradient, CBP35 was found in fractions with the same densities as those reported for hnRNP (z1.30 g/ml); (c) CBP35 co-isolates with hnRNP by sucrose gradient centrifugation (40 S); and (d) fractionation of nucleoplasm on a column derivatized with N6-amino-caproylgalactosamine yielded a bound fraction containing CBP35, as well as a set of polypeptides whose molecular weights matched those reported for the proteins in hnRNP. By using antibodies specifically directed against CBP35 to screen a Agt 11 expression library derived from the mRNA of 3T3 cells, we have identified and characterized a cDNA clone for CBP35 (3). The fusion protein expressed by this clone, upon V8 protease digestion, yielded a M, 30,000 polypeptide that exhibited carbohydrate binding activity and immunoblotting pattern with anti-CBP35 identical to those of CBP35 itself. Moreover, sequence analysis of this cDNA clone revealed two distinct domains within the polypeptide. The carboxyl-terminal domain showed sequence homology with other fl-galactoside-binding lectins, 67 whereas the amino-terminal portion showed homology with certain hnRNP proteins (4). Previous studies had shown that, when quiescent 3T3 fibroblasts were stimulated by the addition of serum, there was increase in the overall level of the CBP35, as well as a dramatic rise in the amount of the polypeptide in the nucleus (5). This increase in the level of CBP35 occurred well before the onset of the first S phase after serum stimulation of 3T3 cells. We had also demonstrated that, in a comparison of the levels of CBP35 in normal fibroblasts and their virally transformed counterparts, there was always more CBP35 in the transformed cells (5, 6). The availability of the cDNA probe provided the opportunity to analyze the expression of the CBP35 gene under these conditions. In the present article, we document the levels of nuclear transcription and mRNA accumulation in quiescent and proliferating cells. 68 MATERIALS AND METHODS Cell Culture. Swiss 3T3 fibroblasts and 3T3 cells transformed by Kirsten murine sarcoma virus cells (3T3-KiMSV cells) were obtained from the American Type Culture Collection (CCL92 and CCL163.3, respectively). They were cultured in Dulbecco’s modified Eagle’s Medium (GIBCO) supplemented with 10% calf serum (GIBCO) at 37°C in 10% C02. Proliferating cultures of 3T3 cells were seeded at a density of 1 X 104 cells/cm2 in culture medium containing 10% serum. These cultures were synchronized by incubation in Dulbecco’s medium contain 0.2% serum for 48 h and were stimulated to proceed through the cell cycle by the addition of 10% serum. At various times after stimulation, cells were harvested for the isolation of polyadenylated (poly(A)+) RNA. Alternatively, nuclei were isolated for run-off transcription assays. Cells cultured at a density of 5 X 10‘1 cells/cm2 formed a quiescent monolayer. In some experiments, cycloheximide was added to a final concentration of 10 ug/ml (36 uM) at the same time as the addition of serum to quiescent cells. RNA Isolation and Northern Blot Mia. Poly(A)+ RNA isolation and Northern blot analysis were carried out as described by Stuart et al. (7). The fibroblasts were washed with calcium and magnesium-free phosphate-buffered saline (PBS;0.14 M NaCl, 2.7 mM KCl, 1.5 mM KH2P04, 4.3 mM NazHPO4, pH 7.4) and lysed with 2 ml of lysis buffer (0.5 M NaCl, 10 mM Tris-HCl, 1 mM EDTA, 1% sodium dodecyl sulfate (SDS), 100 pg of proteinase K/ml, pH 7.5) per 69 flask (150 cmz). The lysed cells were removed from the flask, cellular DNA was sheared by passage through a 22-gauge needle, and the lysate was incubated with additional proteinase K (100 ug/ml) for 1 h at 37°C. The lysate was incubated with 30 mg of oligodeoxythymidylate-cellulose (Sigma) for 1 h at room temperature to absorb the poly(A)+ RNA. This material was packed into an Isolab Quick-Sep microcolumn (VT: 2 ml) and washed with 4 ml of wash buffer no. 1 (0.5 M LiCl, 10 mM Tris-HCl, 1 mM EDTA, 0.2% SDS, pH 7.5) followed by 6 ml of wash buffer no. 2 (0.1 M NaCl, 10 mM Tris—HCl, 1 mM EDTA, 0.2% SDS, pH 7.5). The poly(A)+ RNA was eluted with 10 mM Tris-HCl, 1 mM EDTA, pH 7.5. After the addition of the carrier yeast tRNA (10 pg) and sodium acetate (0.3 M final concentration), the eluted material was ethanol-precipitated. Precipitated RNA was suspended in gel sample buffer (50% formamide, 1 X running buffer, 2.2 M formaldehyde), heated to 65°C, and electrophoresed on a 1.2% agarose gel containing 20 mM morpholinepropanesulfonic acid, 2.2 M formaldehyde. Running buffer was 20 mM morpholinepropanesulfonic acid, 1 mM EDTA, 5 mM sodium acetate, 2.2 M formaldehyde, pH 7.0. After electrophoresis, gels were washed in water and then in 20 X SSC (1 X SSC = 0.15 M NaCl, 15 mM trisodium citrate), and the separated RNA was transferred to nitrocellulose filters. The filters were incubated in prehybridization solution (50% formamide, 5 X SSC, 5 X Denhardt’s solution, 25 mM sodium phosphate, 0.5 rug/ml denatured sheared salmon sperm DNA) for 2-4 h in plastic bags at 42°C. The composition of the 5 X Denhardt’s solution was 0.1% bovine serum albumin, 0.1% 70 polyvinylpyrollidone, and 0.1% Ficoll in H20. Filters were then incubated overnight at 42°C in hybridization solution (50% formamide, 3 X SSC, 1 X Denhardt’s solution, 10 mM sodium phosphate, 0.2 mg/ml denatured sheared salmon sperm DNA) containing 10‘ cpm/ml of DNA probe that was labeled with [a-32P]dCT P using random oligodeoxynucleotide primer labeling. Following hybridization with the radiolabeled probes, filters were washed for 15 min in 2 X SSC, 0.1% SDS at room temperature and washed twice for 45 min in 2 X SSC, 0.1% SDS at 65°C. The washed filters were exposed to Kodak X-Omat AR film with an intensifying screen where indicated. Relative intensities of the bands were determined by scanning densitometric analysis (Gelman ACD-18 automatic computing densitometer). In experiments to determine the half-life of mRNA, actinomycin D was added to cell cultures at a final concentration of 8 ug/ml for the indicated times prior to RNA isolation as described above. In experiments comparing poly(A)+ RNA levels in 3T3 and 3T3-KiMSV cells, the method used was essentially that of Maniatis et al. (8). Cells were scraped into ice-cold PBS and pelleted by centrifugation, followed by incubation in cold lysis buffer (10 mM Tris-HCl, pH 7.5, 10 mM NaCl, 3 mM MgC12, 0.5% Nonidet P-40). The nuclei were removed for nuclear run-off experiments while the cytoplasmic layer was incubated with 2X PK buffer (0.2 M Tris-HCI, pH 7.5, 25 mM EDTA, 0.3 M NaCl, 2% SDS, 400 ug/ml proteinase K), after which the RNA was ethanol-precipitated. The RNA was quantitated, and equal amounts were paSsaged over oligo(dT)-cellulose columns as described above to isolate poly(A)+ RNA. 71 Nuclear Run-off Transcription Assafi Run-off transcriptions were performed essentially as described by Linial et al. (9) and Stewart et al. (10). The cells were washed in cold hypotonic buffer (20 mM Tris-HCl, 5 mM MgC12, 6 mM CaClz, 0.5 mM dithiothreitol, pH 8.0) and scraped off the dishes in this buffer with a rubber policeman. The scraped cells were collected by centrifugation and lysed in 0.6 M sucrose with 0.2% Nonidet P-40 and 0.5 mM dithiothreitol. The nuclei were suspended in freezing buffer (40% glycerol, 50 mM Tris-HCl, 5 mM MgC12, 0.1 M EDTA, pH 8.3), frozen, and stored in liquid nitrogen until they were used. In experiments comparing 3T3 and 3T3-KiMSV cells, nuclei were isolated as described under "RNA Isolation and Northern Blot Analysis." A typical transcription assay (total volume of 200 pl) consisted of approximately 5-10 X 10‘ nuclei in a volume of 130 pl; 40 pl of 5 X run-off buffer (12 mM magnesium acetate, 25 mM Tris-HCl, 12.5 mM MgCl2, 750 mM KCl, 0.5 mM EDTA, pH 8.0), 20 pl of 10 X nucleotides (4.0 mM CT P, 4.0 mM GTP, 10 mM ATP, 5 mM S-adenosylmethionine), and 100 pCi of either [a-32P]UTP or [a- 35S]UTP (800 Ci/mmol or 1500 Ci/mmol, respectively). The reaction was carried out at 25°C for 30 min, after which human placental ribonuclease inhibitor (RNA- guard, Pharmacia LKB Biotechnology Inc.) and ribonuclease-free deoxyribonuclease I (Boehringer Mannheim) were added. The suspension was then mixed with an equal volume of 2 X SETY (2% SDS, 10 mM EDTA, 20 mM Tris-HCI, pH 7.5, 200 pg/ml yeast tRNA) with 200 pg/ml proteinase K, incubated at 37°C for 20-45 min, extracted with an equal volume of phenolzchloroform (phenolz-chloroform:isoamyl alcohol (25:24:1) (v/v) saturated with 10 mM Tris- 72 HCl, 1 mM EDTA, pH 7.5), and precipitated overnight at -20°C with ammonium acetate (final concentration 2.3 M) and 2 volumes of ethanol. The RNA was pelleted by centrifugation for 30 min at 12,000 X g. Incorporated radioactive label in the RNA was verified by precipitating a small amount of the RNA onto glass fiber filters using trichloroacetic acid and counting the filters in a scintillation counter. NaOH was added to the resuspended RNA pellet (to a final concentration of 0.2 N) for 10 min on ice, and then HEPES was added to a final concentration of 0.24 M to neutralize the base. The RNA was then ethanol- precipitated. Equal amounts of radioactivity were added to each hybridization solution. Complementary CBP35 RNA transcripts and DNA plasmids were bound to nitrocellulose fibers by using a slot blot manifold (Bethesda Research Laboratories). For the RNA probes, approximately 1-2 pg of RNA probe/slot was suspended in H20. Formaldehyde (5 pl) was added, and the RNA was heated to 65°C for 10 min. Then, 130 pl of 20 X SSC was added, and the samples were immediately applied to the individual slots. The plasmid DNAs were linearized by restriction enzyme digestion, extracted with an equal volume of phenolzchloroform and precipitated with ethanol at -20°C. Approximately 1-2 pg of DNA/slot was suspended in 25 pl of H20 and mixed with 5 pl of freshly made 2 M NH4OH. The sample was heated for 3 min at 90°C, quenched in ice, and 20 pl of cold 5 M NaCl was added prior to sample application to the nitrocellulose filters. The filters were washed with 2 X SSC and baked under vacuum at 80°C. The filters were prehybridized in 50% formamide, 5 X SSC, 50 mM sodium phosphate, 1 X 73 Denhardt’s solution, 250 pg/ml denatured sheared salmon sperm DNA for 2-24 h at 42°C. Then, hybridization solution (1 part of 50% dextran sulfate added to 4 parts of prehybridization solution) containing the denatured radioactive RNA transcripts was added, and hybridization was carried out at 42°C for 72 h. The filters were washed as directed for Northern blot analysis. The filters were allowed to air dry and were exposed to Kodak X-Omat AR film with an intensifying screen where appropriate. Plasmids and Propagation of Probes. The plasmid pWJ31 containing the cDNA insert for CBP35 was constructed by excising the 883-base pair EcoRI fragment out of the clone identified in the Agt11 library and subcloning into the pUC-13 vector (3). The probe for murine flz-microglobulin was an 8-kilobase pair XhoI fragment subcloned into the ch7 vector (11). Complementary RNA transcripts for CPB35 were made using insert fragments subcloned into the pSp65 vector (Promega Biotec.). The in vitro transcription reactions were performed as described by Promega Biotec, and the amount of RNA transcribed was quantitated by comparing the intensity of ethidium bromide straining of the RNA transcripts against known amounts of control RNA after electrophoresis on agarose-formaldehyde gels. DNA probes were linearized by restriction enzyme digestion and extracted with phenolzchloroform. The insert cDNA for CBP35 was purified by gel electrophoresis. DNA labeled with [a-32P]dCI‘P was prepared using a random oligodeoxynucleotide primer labeling kit (Pharmacia LKB Biotechnology Inc.). 74 Typical reactions yielded probe with a specific activity of 1 X 109 cpm/pg. Usually 1 X 10‘5 cpm/ml of hybridization solution was used. Indirect Immunofluorescence. 3T3 and 3T3-KiMSV cells were seeded on coverslips at a subconfluent density (1 X 10" cells/cmz) in Dulbecco’s modified Eagle’s medium containing 10% calf serum. 3T3 cells were synchronized by serum starvation and then stimulated to enter the cell cycle by serum addition following the procedure described (5). The coverslips were washed in PBS, fixed in 3.7% formaldehyde for 15 min at 4°C,washed in PBS and finally permeabilized in PBS containing 0.2% Triton X-100 for 30 min at 4°C (12). The cells were washed with 20 mM Tris-HCI, 0.5 M NaCl, 2.5% bovine serum albumin, pH 7.5. Each coverslip was then inoculated for 1 h in 100 pl of a 1:10 dilution of rabbit antiserum raised against CBP35 (1) in 20 mM Tris-HCI, 0.5 M NaCl, 2.5% bovine serum albumin pH 7.5. The cells were washed in the same Tris buffer containing bovine serum albumin and then incubated in 100 pl of a 1:30 dilution of rhodamine-conjugated goat anti-rabbit immunoglobulin (ICN Biomedical) in the same buffer. The coverslips were then rinsed and mounted in 70% glycerol, PBS containing the anti-bleaching agent n-propylgallate (5%; Sigma). The slides were viewed with a Leitz epiphase fluorescence microscope using a 25 X objective lens. 75 RESULTS Kinetics of the Accumulation of CBP35 mRNA upon Mitogenic Stimulation. In previous studies, we observed that when serum-starved quiescent 3T3 fibroblasts were activated by the addition of serum, there was an increase in the overall level of CBP35 as detected by Western blotting of total cell extracts (5). This increase in the amount of CBP35 polypeptide accounts, at least in part, for an increase in the percentage of cells that are fluorescently labeled by rabbit antibodies directed against CBP35. The increase in the amount of CBP35 at the protein level, as detected by immunoblotting and by immunofluorescence, occurred as early as 5-8 h after the addition of serum. Serum-starved 3T3 cells were stimulated by the addition of serum; at various times thereafter, poly(A)+ RNA was isolated and subjected to Northern analysis using the cDNA probe for CBP35. A single mRNA species, ~1.3 kilobases, was revealed by this probe (3). This mRNA was virtually absent in quiescent cells but could be detected within 30 min after the serum activation (Fig. la). The level reached a peak value ~1-2 h after stimulation and then decreased. A more extended study revealed that the early rise (<3 h) was followed by a slight decline before the mRNA level for CBP35 increased some 6-fold at 21 h (Fig. 1b). The generation time and the length of S phase for the 3T3 cells used in the present study have been estimated to be about 24 and 9 h, respectively (5,13). Therefore, the early time points (Fig. 1a) cover the GO/early G1 transition, whereas the 9-h time point (Fig. 1b) is in the middle of the G1 period. DNA synthesis, as 76 .296 :8 25 mo confine 05 9:350 £05— 95% of the incorporation of 1“C-labeled amino acids into trichloroacetic acid-precipitable material, in agreement with the result of previous studies on the effect of the drug on protein synthesis in 3T3 cells (15,16). In the absence of de novo protein synthesis, the transcription of the CBP35 gene was nevertheless elevated, increasing at least 5- fold over a course of 10 h (Fig. 28). Over the same time course, the transcription rate of the gene for flz-microglobulin remained constant. Thus, it appears that the gene for CBP35 can be induced directly as a result of mitogen addition, without 84 5028088 U new v. E 805 5.5» 3.525 E uoEuotom 803 Q 95 m E 8552098 2: J .mE :5: 55 58058 D can T .oEEEucofimo Mo 8885 2: E 8555.8 85:8 E £82 8 £96— 18 .m 8.82 “: “.5 no engage—9m. me 80am— .v 053m o a. 4/218\\ IIIII m 350... v 328?: _N O_ m m 0 O¢N 09 ON. ow on n. O 5 , I s o ‘I I I 4/ o \ I I I I r. < L 8_E_xmco_o>o L 84 5028088 U 0:: T E 805 :23 3:85 E 005585 803 Q 0:0 m E 8505:0520 05 MA .wE :52 :80 58058 U 0:: 2 025205208 20 8:805 05 E 082558 805:8 E 282 8 £82 «.725 memo 20 230:: U ”0250202208 20 8:805 05 E 008558 80528 E 282 <25: wmn—mo 282 858 0:5 250 .M £05. m2§-mr_.m 0:0 5m 58 825.05 05 mum a 2.50 5.5588 “3505 8:038:55; .n 25.: 88 Q 89 with antibodies directed against CBP35, the 3T3-KjMSV cells showed much higher intensity of fluorescence than their normal counterparts (Fig. 5). At least part of the reason for this drastic difference in staining intensity must be ascribed to a difference in the level of the CBP35 polypeptide in the two cell types (6). In both 3T3 and 3T3-KiMSV cells, the staining pattern reflected the nuclear localization of the lectin. We have corroborated these differences between 3T3 and 3T3-KiMSV cells in the expression of CBP35 at the level of accumulated mRNA, as well as at the level of nuclear transcription rates. By using nonsynchronized cell populations, poly(A)+ RNA was prepared from an equal number of cells derived from sparse and confluent cultures and subjected to Northern blot analysis. In both cases, much higher amounts (10-15-fold) of CBP35 mRNA were detected in 3T3-KiMSV cells than in normal 3T3 cells (Fig. 6). The RNA species in 3T3-KiMSV cells is of the same size (~ 1.3 kilobases) as that in 3T3 cells. To determine whether the differences seen in the levels of CBP35 mRNA between normal and virally transformed cells are due to transcription rate differences, nuclear run-off transcription assays were performed. Under both sparse and confluent culture conditions, 3T3-KiMSV cells had a higher transcription rate for the CBP35 gene than 3T3 cells (Fig. 7). This conclusion takes into consideration the fact that equal amounts of radioactivity from the different run-off experiments were hybridized in the slot blots shown in Fig. 7. Experiments were also carried out to compare the half-lives of the CBP35 mRNA in normal and transformed 3T3 cells. Cells were cultured in the presence 90 Figure 6. Northern blot analysis of CBP35 mRNA in 3T3 and 3T3-KMV cells. Lane a, 3T3 cells seeded at a density of 1 X 10‘ cells/cmz; lane b, 3T3-KiMSV cells seeded at 1 X 104 cells/cmz; lane c, 3T3 cells seeded at 5 X 10" cells/cmz; lane (1, 3T3-KiMSV cells seeded at 5 X 10“ cells/cmz. Equal amounts of total RNA (120 pg) from each culture were used to isolate poly(A)+ RNA, which was then electrophoresed and hybridized with a 32P-labeled cDNA probe for CBP35. The autoradiograms were exposed for 72 h without an intensifying screen. 91 abcd 92 0:00:00 5.5500005 :0 553 0.500 0 :00 000058 003 5050500550 05H. 42% 00:00:50 wmmmv 5550500 0:05 :20 0: 0050555 0:03 53800500: 00 055050 .0000 0:0 $55000 500—5500 0: 0080.300 0:03 0.50000 50-5: 50.5 0:580:05 m2§-m.bm .0 0:0: m0500:00 A: x m 00 000000 0:00 mph .0 0:0: m0:50:00 A: X H :0 000000 0:00 >m2§éhm .5 0:0: m0502—00 A: X 0 00 50:00 50555 :0 :0 000000 0:00 mhm .0 0:3 0:8 >m$5fimhn 000 pm 5 one» name 050 .50 30: 0005.60.55 05..— 0559500 0.50000 .095: .5205: we 50305089020 .0 053m 93 000001.-‘l. 0 0 0 0 94 00:00:00 <20»: +A:0000 :0: 0.: 0:0 A:0000 0.: 0: 00000003 3000—0500:: 0.: :00: w2g-mrrm :0 000—0 E :000 0:00 5.00:: 00: 0>000 00006085.. 0:00 00:85 000. 00,: 00:00:25 0:00 00: :0 0.00020 050800200 w::::000 m:.::::0::0m >0 0058:0000 0:03 0033—02 .588 m:.:©._0:0::._ :0 300:3 0 mm :0: 0000000 003 0:00 >mEE-mHm 00: :0“: E0:m0_00:0::0 00: 0:0 .:00:00 $000500: :0 0:3 0 S :0: 0000900 003 0:00 mhm :0: E0:m0.:00:0::0 00:. A5 008: 00:00:05 0:: :0: GEE: 8 Q 50:80:80 5 0:00 00: :0 :0000005 :20 0000—00: 003 m§ -000 08 0.00 0: :0000 :0: 0: :00 Am:>:v:-m:.m :0 0:0—0 8 :000 0:00 8:08 00: 0>000 30:0:0088: 0:00 00:08: 000: 00,: 00:00:08: 0:00 00: :0 0:00:80 050800800 w:_::000 w:_8:0::00: :0 00:85:00 0:03 003—:0:]: .:00:00 m:.§::0:0::: :0 80003 0 ma :0: 000098 003 0:00 >m2§-m:.m 00: :0: 80:m0:00:0::0 00: 0:0 8088 83:808.. :0 0:3 0 S :0: 000096 003 0:00 2% :0: 80:w0:00:0::0 00,: A5 008:: 00:00:08 00: :0: GER: my D 800080800 8 0:00 00: :0 :0:00:08 :0::0 00:060.: 003 «id 450900: 00,: 40:00 >m§ 0,00 05 0.00 0: <88 00:00 :0 030.00,: 20088 9 0:008 :2: E0582 0 8.0:: ,(({\{“‘-”tiit----:‘ ‘( t “ 95 vfnhn ohm 96 of actinomycin D for various times; RNA from these treated cells was isolated and subjected to Northern blotting with the cDNA probe for CBP35. The autoradiogram showed that the 3T3-KiMSV cells displayed an enhanced stability of the CBP35 mRNA (Fig. 8). The levels of the mRNA for CBP35 were quantitated from the autoradiogram by densitometric scanning. CBP35 mRNA in 3T3 cells had a half-life of 120 min, compared to that of 200 min in 3T3-KiMSV cells. It appears, therefore, that increased stability of the mRNA allows for its accumulation in the transformed cells. Comm‘ n of the fission of the CBP35 Gene in Sparse and Confluent Cultures of 3T3 Cells. In previous studies, we observed that sparse cultures of 3T3 cells showed intense immunofluorescence staining for CBP35, predominantly within the nuclei, whereas in confluent monolayers of the same cells, the staining intensity decreased with a dramatic loss of CBP35 from the nuclei (5,12). In the present study, it was quite to our surprise, therefore, to find that the transcription rate, as well as level of accumulated mRNA, for CBP35 remained appreciable in the confluent cultures. As quantitated by densitometric analysis, the transcription rate for the CBP35 gene in sparse cultures of 3T3 cells was approximately the same as that for confluent cultures (Fig. 7). However, the levels of CBP35 mRNA, as detected by Northern blot analysis, were higher for confluent cultures of 3T3 cells than for sparse cells (Fig. 6). 97 DISCUSSION We have compared the rate of nuclear transcription and the level of accumulated mRNA for CBP35 in cultures of 3T3 cells representing quiescent and proliferative states: (a) serum-starved versus serum-stimulated cells, (b) confluent monolayers versus sparse cultures, and (c) normal fibroblasts versus cells transformed by KiMSV. The basis for these comparisons is the observation that at the polypeptide level, as detected by Western blotting and/or immunofluorescence, the expression of CBP35 was correlated with the proliferating state of the cell (5,6). Our present studies showed that, in general, high levels of CBP35 protein in proliferating 3T3 cells reflect higher levels of mRNA as well as elevated nuclear transcription rates of the gene. In one instance, however, there was an exception: although confluent monolayers of 3T3 cells exhibited much lower levels of CBP35 protein than sparse proliferating cells, their mRNA levels for CBP35 were higher when compared on the basis of equal amount of RNA. When serum-starved 3T3 cells were stimulated by the addition of serum, there was an early rise in the transcription rate of the CBP35 gene. Within 0.5 h, there was a detectable increase in the level of CBP35 mRNA, which exhibited a peak at about 3 h. This was followed by a transient decrease and then a second rise, leading up to a ~6-fold elevation over approximately the next 10 h. Such a bimodal kinetic pattern for the level of mRNA as a function of time after mitogen activation has also been seen when the mRNA corresponding to the c-myc gene 98 was monitored after stimulation of quiescent 3T3 cells by fetal calf serum (19). The effect of serum growth factors on the regulation of CBP35 and c-myc genes is similar in another respect: when the stimulation of the cells was carried out in the presence of cycloheximide, superinduction of mRNA accumulation for these genes was observed. This phenomenon may be rationalized in terms of the inhibitory effect of cycloheximide on the production of shortlived nucleases that normally could affect the stability of mRNA Alternatively, the results may be interpreted to indicate that the increased expression of the CBP35 gene is a direct result of signals transduced by the binding of growth factors to their plasma membrane receptors, without the requirement of prior synthesis of other gene products. The observation that a similar pattern of transcription rate variations occurred in the presence and absence of cycloheximide lends support to the latter hypothesis. This pattern of changes includes an early increase, a slight decline at 6 h, and an approximately S-fold elevation of CBP35 mRNA under both sets of conditions. Thus, CBP35 appears to be regulated in manner comparable to other mitogen activated genes, including the oncogenes c-fos and c-myc (17,18). Recent evidence indicates that inhibition of the expression of these oncogene products (e.g. by antisense RNA) results in a blockage of the progression of a stimulated cell into DNA synthesis and cell division (20,21). This suggests that oncogene products are required for normal cell cycle progression during cell activation. It would be of interest, therefore, to test if inhibition of the expression of CBP35 protein might also result in growth inhibition. 99 Like the oncogenes, the amounts of CBP35 are elevated in "deregulated" cell populations, lie. transformed cells. The data suggest that the expression of the CBP35 gene is directly affected by transformation at the transcriptional level. In addition, our results also indicate that there is greater stability of the CBP35 mRNA in 3T3-KiMSV cells than in their normal counterparts. Both of these effects contribute to a marked elevation in the accumulated mRNA for CBP35 and the protein product in transformed cells. In the transformed murine fibrosarcoma cell line, UV-2237-IP3, there is a selectively elevated expression of one lectin, L-34 (which corresponds to CBP35), but not of another, L-14 (which corresponds to CBP13.5) (22). It was suggested that increased amounts of L-34 may reflect increased malignancy of the cells (23). By using differential screening and cDNA cloning techniques, a set of some 10 mRNAs that appears in 3T3 cells soon after growth stimulation by serum has been identified (17,18). These have been termed growth-related immediate early genes. The kinetics of the induction of expression of the CBP35 gene are similar to these immediate early genes (3CH61, 3CH77, 3CH268, etc.) in that (a) they are all elevated within 10-30 min after serum addition, and (b) they are superinduced in the presence of cycloheximide. The mRNAs corresponding to most of these immediate early genes reached peak levels between 40 and 120 min after serum addition and rapidly decayed thereafter (17,18), whereas that for CBP35 showed bimodal kinetics, with a long term increase over a period of approximately 20 h. Recent nucleotide sequence analyses have revealed that one of the immediate early genes (3CH268) contains three tandem "zinc finger" sequences typical of a 100 class of eukaryotic transcription factors (24) and another (3CH77) encodes a member of the superfamily of ligand-binding transcription factors that includes the steroid and thyroid hormone receptors (25). We have recently provided evidence that CBP35 is a component of hnRNP on the basis of immunochemical localization of the lectin in hnRNP fractions (2) and on the basis of amino acid homology of the CBP35 polypeptide to other hnRNP proteins (4). Our previous observations of the proliferation-dependent expression of the CBP35 polypeptide (5,6) and our present documentation of the correlation of increased nuclear transcription and mRNA accumulation of the CBP35 gene with data indicating that several other hnRNP polypeptides are also regulated upon quiescent cell to proliferative cell transition (26,27). Although the function of CBP35 is not known, its identification as a component of hnRNP suggests that it may play a role in the processing, packaging, or transport of mRNA from the nucleus into the cytoplasm. Glycosylated components (28) and CBP35 (12) are found in both the nucleus and the cytoplasm, suggesting the possibility that they may shuttle between the two subcellular compartments in a transport role. Indeed, preliminary experiments using a cell-free assay for RNA transport from isolated nuclei indicate that CBP35 is cotransported with mRNA in the form of a ribonucleoprotein complex (J. G. Laing, E. A. Werner, and R. J. Patterson, unpublished observations). Finally, it is noteworthy to discuss the one exception to the correlation between CBP35 protein levels and CBP35 mRNA levels and gene transcription rates. Although dense cultures of 3T3 cells exhibit approximately the same transcription rate for the CBP35 gene as sparse cultures, there are higher levels of 101 detectable mRNA in the dense monolayers. Thus, posttranscriptional mechanisms, such as mRNA stability, must be involved in the accumulation of CBP35 RNA in 3T3 cells. Such mechanisms have been reported in the regulation of histone mRNA (29) and c-myc RNA (30). More surprisingly, despite having higher amounts of mRNA for CBP35, dense 3T3 cells exhibit less of the protein than sparse 3T3 cells (5). Therefore, additional mechanisms for this difference at the translation level (e.g. ribosome binding) or posttranslational level (e.g. protein stability) may be involved. Clearly, there is need for caution in interpreting correlative data from different sets of experiments; simultaneously Northern and Western blot determinations on CBP35 mRNA and protein, respectively, from the same cell population are required. ACIQIOWLEDGEMEN'IS We thank Mrs. Linda Lang for her help in the preparation of the manuscript. 10. 11. 12. 13. 14. 15. 16. 17. 18. 102 REFERENCES Roff CF & Wang JL (1983) J Biol Chem 258: 10657-10663. Laing JG & Wang JL (1988) Biochemistry 27: 5329-5334. Jia S, Mee RP, Morford G, Agrwal N, Voss PG, Moutsatsos IK & Wang JL (1987) Gene (Amst) 60: 197-204. Jia s & Wang JL (1988) J Biol Chem 263: 6009-6011. Moutsatsos IK, Wade M, Schindler M & Wang JL ( 1987) Proc Natl Acad Sci USA 84: 6452-645 6. Crittenden SL, Roff CF & Wang JL (1984) Mol Cell Biol 4: 1252-1259. Stuart P, Ito M, Stewart C & Conrad SE (1985) Mol Cell Biol 5: 1490- 1497. Maniatis T, Fritsch EF & Sambrook J (1982) Molecular Cloning: A Laboratory Manual pp. 191-193, Cold Spring Harbor laboratory, Cold Spring Harbor NY. Linial M, Gunderson N & Groudine M (1985) Science 230: 1126-1132. Stewart CJ, Ito M & Conrad SE (1987) Mol Cell Biol 7: 1156-1163. Parnes JR & Seidman JG (1982) Cell 29: 661-669. Moutsatsos IK, Davis JM & Wang JL (1986) J Cell Biol 102: 477-483. Steck PA, Voss PG & Wang JL (1979) J Cell Biol 83: 562-575. Kelly K, Cochran BH, Stiles CD & Leder P (1983) Cell 35: 603-610. Brooks RF (1977) Cell 12: 311-317. Campisi J, Medrano E, Morreo G & Pardee AB (1982) Proc Natl Acad Sci USA 79: 436-440. Lau LF & Nathans D (1985) EMBO J. 4: 3145-3151. Lau LF & N athans D (1987) Proc Natl Acad Sci USA 84: 1182-1186. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 103 Miiller R, Bravo R, Burckhardt J & Curran T (1984) Nature 312: 716-720. Nishikura K, & Murray JM (1987) Mol Cell Biol 7: 639-649. Freytag SO (1988) Mol Cell Biol 8: 1614-1624. Raz A, Carmi P & Pazerini G (1988) Cancer Res 48: 645-659. Raz A, Meromsky L, Zvibel I & Lotan R (1987) Int J Cancer 39: 353-360. Christy BA, Lau LF & Nathans D (1988) Proc Natl Acad Sci USA 85: 7857-7861. Hazel TG, Nathans D & Lau LF (1988) Proc Natl Acad Sci USA 85: 8444- 8448. Celis JE, Bravo R, Arenstorf HP & LeStourgeon WM (1986) FEBS Lett 194: 101-109. Leser GP & Martin TE (1987) J Cell Biol 105: 2083-2094. Hart GW, Holt GD & Haltiwanger RS (1988) Trends Biochem Sci 13: 380- 384. Marzluff WF & Pandey PB (1988) Trends Biochem Sci 13: 49-51. Piechaczyk M, Yang J-Q, Blanchard J-M, Jeanteur P & Marcu KB (1985) Cell 42: 589-597. CHAPTER III CARBOHYDRATE BINDING PROTEIN 35: Properties of the Recombinant Polypeptide and the Individuality of the Domains' Neera Agrwal, Quan Sun, Sung-Yuan Wang, and John L. Wang Department of Biochemistry Michigan State University East Lansing, MI 48824 104 105 FOOTNOTES This work was supported by grants GM-3874O and GM-27203 from the National Institutes of Health. The abbreviations used are: rCBP35, recombinant carbohydrate binding protein 35; N- and C-domains, NHz-terminal and COOH-terminal portions, respectively, of the CBP35 polypeptide; IPTG, isopropyl-fl-D- thiogalactopyranoside; Z-ME, 2-mercaptoethanol; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; DSC, differential scanning calorimetry. 106 SUMMARY The cDNA clone for Carbohydrate Binding Protein 35 (CBP35) was engineered into the bacterial expression vector pIN III ompAZ, which directs the secretion of the expressed protein into the periplasmic space. Recombinant CBP35 was purified from this system, at a level of "' 50 mg per liter of bacterial culture. Digestion of recombinant CBP35 with collagenase D, followed by purification using saccharide-specific affinity chromatography yielded a Mr "' 16,000 polypeptide, corresponding to the COOH-terminal domain (residues 118- 264) of the CBP35 polypeptide. This indicates that the COOH-terminal half of CBP35 contains the carbohydrate recognition domain, consistent with its sequence homology to other S-type lectins. The NHz-terminal domain (residues 1-137) was derived by site-directed mutagenesis of the cDNA, in which stop codons are inserted in place of Gly 138 and Gly 139, and expression of the mutant cDNA in the same pIN III ompA2 system. The purified NHz-terminal domain failed to bind to saccharide-specific affinity resins. Differential scanning calorimetry of rCBP35 and its individual domains yielded transition temperatures of "' 39°C and ”56°C for the NHZ- and COOH-terminal domains, respectively. Lactose binding by the COOH-terminal domain shifted the transition temperature to 65°C, whereas sucrose failed to yield the same effect. These results suggest that the individual domains of the CBP35 polypeptide are folded independently. 107 INTRODUCTION Carbohydrate Binding Protein 35 (CBP35, Mr 35,000) is a galactose specific lectin which was initially purified from murine fibroblasts (1). CBP35 or its homologs have since been found in a variety of tissues and species (2). Moreover, the available data indicate that the same protein has been studied under various different names (for a review, see (3)): (a) L-34, a lectin from murine tumor cells (4,5); (b) Mac-2, a surface antigen of mouse macrophages (6,7); (c) IgE-binding protein from rat basophilic leukemia cells (8,9); and (d) RL-29 (10) and HL-29 (11), lectins from rat and human lungs, respectively. Using indirect immunofluorescence and cell fractionation studies, we have localized CBP35 predominantly to the cytoplasm and nucleus of mouse 3T3 fibroblasts (12). CBP35 gene expression, both at the transcriptional level and protein level, is elevated upon mitogenic stimulation of quiescent cultures of 3T3 cells (13,14). Sequence and hydropathy analysis of the cDNA clone for CBP35 showed that it has two distinct structural domains (3,15). The carboxyl terminal half contains significant homology to other fi-galactoside binding lectins, and thus may contain the carbohydrate recognition domain. The amino terminal half has eight contiguous nine amino acid repeats with the sequence Pro-Gly-Ala-Tyr-Pro-Gly-X- X-X, which consequently makes this domain highly proline- and glycine- rich. Thus, CBP35 has been classified as an S-type lectin with a bifunctional motif (16,17). 108 In this paper, we have taken advantage of an improved prokaryotic expression vector to produce milligram amounts of recombinant CBP35 (rCBP35) and its corresponding NHz-terminal and COOH-terrninal halves (designated as N- domain and C-domain, respectively). Comparison of the physico-chemical properties of the N- and C-domains with the parent rCBP35 polypeptide suggest independent folding of the individual domains. 109 MATERIALS AND METHODS Construction of prCBPBSs and Purification of rCBP35 The CBP35 cDNA was excised out of the plasmid pWJ31 using EcoRI restriction. The cDNA was then ligated into the prokaryotic expression vector pIN III ompA2 (18) using standard subcloning techniques (19). The resulting recombinant clone, containing the plasmid for the production of rCBP35 in a soluble form (prCBP355) (Fig. 1A), was then transformed into the E. 0011' JA221 strain. prCBP35s was freshly transformed into JA221 cells prior to each purifica- tion. An overnight culture of prCBP35s was grown in LB broth containing 100 pg/ml ampicillin at 37°C. One liter of Terrific Broth (2.31 g KHZPO.” 12.54 g KZHPO4, 12 g bacto-tryptone, 24 g bacto-yeast extract, and 4 ml glycerol in 1 liter (ref. 19)) containing 100 ug/ml ampicillin was then inoculated with the overnight culture, and then allowed to grow at room temperature for 2 hours, after which isopropyl-fi-D-thiogalactopyranoside (IPTG; Research Organics) was added to a final concentration of 50 pM, and the culture allowed to grow at room temperature for 16-24 hours. After harvest, the cell pellet was washed with ice cold phosphate-buffered saline (0.13 M NaCl, 5 mM sodium phosphate, pH 7.5), and then resuspended in 25 ml ice cold 1 M Tris-HCl (pH 7.0) containing 10 mM 2-mercaptoethanol (2-ME) (Pierce Chemical), 1 mM phenylmethanesulfonyl fluoride (BMB), 0.5 ug/ml leupeptin (BMB), 30 KIU/ml aprotinin (BMB), and .01 mg/ml soybean trypsin inhibitor (Sigma). The resuspended pellet was kept on 110 ice for 30 minutes, after which it was centrifuged at 90,000 x g for 30 minutes at 4°C. The supernatant, representing the periplasmic fraction, was then fractionated by ammonium sulfate (65% of saturation) at 4°C. The precipitated protein was recovered by centrifugation at 12,000 x g, and resuspended in loading buffer (75 mM Tris-HCI pH 7.5, 75 mM NaCl, 2 mM EDTA, 1 mM 2-ME, 1 mM phenylmethanesulfonyl fluoride). The resuspended protein was then combined with 10 ml of asialofetuin—Affi-gel 15 (BioRad), and the slurry dialyzed overnight at 4°C (2 changes of 2 liters each) against loading buffer. The slurry was packed into a column, washed with 20 column volumes of loading buffer (25 ml/hour), and eluted with 0.4 M lactose in loading buffer. The eluted protein was concentrated by ultrafiltration, and the buffer exchanged to 10 mM Tris-HCl pH 8.5, 1 mM EDTA, 1 mM 2-ME, using a PM10 filter (Amicon) in an Amicon stirred cell. Where noted, rCBP35 was purified entirely without the addition of 2-ME. Site-Directed Muggenesis and Generation of Recombinant Clones The CBP35 cDNA was subcloned into the phagemid pUC119 (20), and the recombinant vector, designated pWJ 1131, was transformed into the E. coli strain CJZ36 (21). After uracil containing single stranded DNA was obtained, site- directed mutagenesis was carried out using the BioRad Muta-Gene Phagemid in vitro Mutagenesis kit (BioRad Laboratories, Richmond CA). All oligonucleotides were synthesized in the Michigan State University Macromolecular Structure Facility. 111 In order to obtain the amino terminal half of CBP35, glycines 138 and 139 (the numbering system follows that of murine L-34 (4) and Mac-2 (6), since these cDNAs identified the translation initiation methionine (residue 1)) were converted to stop codons by using the oligonucleotide 5 ' -GATCAGCATGCGAAGCTT GA CT CATCAAGGCAACGGCAGGTC-3’ (Fig. 1B). The above oligonucleotide also inserted a HindIII restriction site downstream of the stop codons. The fragment covering the 5 ’-end of the cDNA through the HindIII site was subcloned into pIN III ompAZ. The construct was transformed into JA221 cells. HindIII digestion of the mutagenized cDNA also allowed us to subclone the fragment corresponding to the carboxyl terminal half of the polypeptide (Fig. 1B). The carboxyl terminal half was excised out of pWJ1131 with HindIII-BamHI restriction, and then ligated into the vector pIN III ompAl. The recombinant expression plasmid was then transformed into E. coli JA221. The periplasmic fractions of E. coli transformed with the mutant recombi- nant clones were harvested in the same manner as that for rCBP35. Premiun of the N- and C-domains For the purification of N-domain, the ammonium sulfate (65% of saturation) precipitate of the periplasmic fraction was dialyzed against 10 mM sodium phosphate buffer (pH 7.2). The dialyzed material was mixed with 20 ml of hydroxylapatite (Biogel HT, BioRad) and rocked at 4°C for 2 hours. The suspension was then centrifuged (1,500 x g; 10 minutes). The supernatant represented the unbound fraction. The beads were washed four times with 10 112 mM sodium phosphate (25 ml each), and the supernatant of each wash was combined with the unbound fraction. Material bound to hydroxylapatite was eluted by incubating the beads with 0.4 M sodium phosphate (pH 7.2). The unbound and bound fractions were concentrated by pressure dialysis in Amicon filters and analyzed by SDS-PAGE. The unbound fraction (containing the N- domain) was dialyzed against 10 mM Tris (pH 8.0) and then fractionated over a column (1.8 x 14 cm) of DEAE-cellulose (DE-52, Whatman). A linear gradient (0 - 0.2 M KCl in 0.01 M Tris, pH 8.0) in a total volume of 150 ml was used to develop the column and purified N-domain was found in fractions eluted by 0.05 - 0.1 M KC1. For sequence analysis, the N-domain containing fractions were pooled and subjected to high pressure liquid chromatography over a Poly LC (Columbus, MD) hydroxyethyl A column (4.6 x 250 mm) in 50 mM formic acid. The flow rate was 250 ill/minute and the absorbance of the effluent was monitored at 214 nm. For the preparation of C-domain, rCBP35 was dialyzed against 75 mM Tris-HCl pH 7.0, 75 mM NaCl, 10 mM CaClz overnight at 4°C, followed by digestion with an equal weight of collagenase D (BMB) at 37°C for 1.5 hours. The digestion mixture was then dialyzed overnight against loading buffer, and then fractionated over an asialofetuin—Affi-gel 15 column as described above for rCBP35. mm CBP35 was purified from mouse lung.(2) and was used to immunize a female New Zealand White rabbit via the popliteal lymph node (1). The 113 specificity of this antiserum, which serves as a reference in the present study, has been characterized in terms of immunoprecipitation of CBP35 from a nondenatured protein mixture and in terms of immunoblotting the polypeptide after SDS-PAGE (12). This antiserum is hereafter referred to as #24. Two new antisera, designated #32 and #33, have been generated using rCBP35 as the immunogen. Flemish Giant rabbits were immunized subcutaneously with 50 pg of rCBP35 in complete Freund’s adjuvant; they were boosted two weeks later with 50 pg of rCBP35 in incomplete Freund’s adjuvant. Antisera were collected 7-10 days after each subsequent booster injection. Rat monoclonal antibody directed against the Mac-2 antigen (22) was isolated from the hybridoma line M3/38.1.2.8.HL.2 (American Type Culture Collection T18166). The hybridoma line was cultured in serum-free medium (RPMI 1640 containing Nutridoma SP (BMB)). After centrifugation to pellet the cells, supernatants from the cultures were pooled, subjected to ammonium sulfate precipitation (45% of saturation), dialyzed against phosphate-buffered saline exhaustively, and stored in aliquots (25 ug/ml). FAME-32m Protein concentrations were determined by the Bradford assay (23). Amino acid sequence analysis was carried out in the Michigan State University Macromolecular Structure Facility. Protein samples were electrophoresed on 12.5% or 15% SDS-PAGE as described by Iaemmli (24). The gels were stained for protein by silver (25, 26) or Coomassie blue, or transferred onto Immobilon-P 114 membrane (Millipore), using semidry blotting and the buffer system of Bjerrum and Schafer-Nielsen (27). The immunoblots were blocked for several hours in 2% gelatin (BioRad) Tris-buffered saline (20 mM Tris-HCl, pH 7.5, 0.5 M NaCl), and subsequently incubated in polyclonal antiserum raised against CBP35 or monoclonal anti-Mac-Z (in Tris-buffered saline containing 1% gelatin) for 2 hours at room temperature. The blots were washed in Tris-buffered saline containing 0.05% Tween-20 extensively, prior to the addition of secondary antibody conjugated to either horseradish peroxidase (BioRad) or alkaline phosphatase (BMB). The blots were colorimetrically developed using either 3,3 '-diaminobenzidine/I-1202 for horseradish peroxidase or nitroblue tetrazolium/S-bromo-4-chloro-3-indolyl phosphate for alkaline phosphatase. Differential my Calorimeg (ESQ) Samples were dialyzed at 4°C against either 10 mM Tris (pH 7), 1 mM EDTA, 10 mM 2-ME (pH 7 buffer) or the same buffer adjusted to pH 10 (pH 10 buffer). Both samples and buffers were deaerated with stirring for 10 minutes under vacuum. Aliquots were removed for the determination of protein concentration. The DSC experiments were carried out in a Microcal MC-2 scanning calorimeter (Microcal Inc., Amherst, MA), with a Model 150 B (Keithley Instruments, Cleveland, OH) microvolt ammeter added to increase sensitivity. The calorimeter was interfaced with an IBM XT. Once the temperatures of the 115 sample and reference buffer were equilibrated, a computer-controlled scan was initiated, at a rate of 90°/hour. The data were analyzed using the "F1" option of the DA—2 program, included in the software provided by Microcal. Deconvolution analysis was performed using the "F1" option of the deconvolution subroutine. To test the effect of prior heat denaturation of the N-domain on collagenase digestion of rCBP35, parallel samples (20 pg rCBP35 in 400 pl of 75 mM Tris, 75 mM NaCl, 10 mM CaClz, pH 10) were incubated at 30°C and 40°C for 15 minutes. After reequilibration at 30°C, 5 pg of collagenase D was added in 20 pl of the same Tris buffer. Aliquots were taken from the two digestion mixtures every 3 minutes, quenched with SDS-PAGE buffer and analyzed by gels. 1 16 RESULTS gmssion and Purification of Recombinant CBP35 The cDNA clone for CBP35 was inserted into the pIN III ompA2 secretion vector in E. coli for the production of rCBP35 (Fig. 1). This vector uses the signal sequence of ompA, an E. coli outer membrane protein, to direct expressed proteins into the periplasmic space (18). Using the lpp-lac fusion promoter, the protein can be induced by the addition of IPTG (Fig. 1). The soluble periplasmic fraction of E. c011 JA221 cells transformed with prCBP355 was collected 16—24 hours after induction, subjected to ammonium sulfate precipitation and then affinity chromatography on a column of asialofetuin—Affi-gel. The material bound to the column and eluted upon lactose addition yielded a single band upon SDS-PAGE, as revealed by both silver-staining (Fig. 2A, lane 4) and immunoblotting (Fig. 2B, lane 4). The position of migration of this polypeptide corresponded to that observed with mouse CBP35 (Mr "' 35,000). This material is designated as rCBP35. The initiator methionine of the protein product is provided by the ATG of the signal sequence. Due to the location of the cleavage site of the signal sequence, three extra amino acids are added on to the amino-terminus of the mature protein. Thus, the amino terminal end of rCBP35 has the following sequence: Ala-Glu-Phe-Arg-Asp-Ser-, with the fourth residue (Arg) being the first amino acid of the cDNA clone for CBP35 (15). The yield of the isolated protein is highly dependent on the type of culture broth used, the concentration of IPTG, 117 Figure 1. Schematic diagram of the construction of the recombinant expression vector prCBP35s and site-directed mutagenesis to obtain the amino terminal and carboxyl terminal domains A) The cDNA clone for CBP35 was cloned into the EcoRI site of the pIN III ompA2 vector. The cDNA is downstream of the IPTG inducible promoter and the ompA signal sequence. B) Using oligonucleotide- directed mutagenesis, glycines 138-139 were changed to translation stop codons, and a HindIII restriction site was introduced at amino acid positions 141-142. The mutant cDNA was digested with HindIII and the fragment corresponding to the NHz-terminal portion of the CBP35 polypeptide was subcloned and inserted into the EcoRI-HindIII sites of pIN III ompA2 vector as above to express the NHZ- , terminal domain. 118 EcoRI ompA sunm pepfide Hmcn Pst I prCBP35s (8.4kb) HMcn B I37 I40 I45 'Id 1 4 Pro Gly Gly Vol Met Pro Arg Met Leu "" V” ccr GGA GGA GTC ATG ccc ccc ATG crc mutant { CCT IGA IGA GTC A_A_(; CH 060 ATG CTG Pm §'_0r2 m2 V‘“ 5” fl Arg Met Leu 4‘1 J «v .4 -d'vvv--w-~m..---w\/VJ - 119 and. v.3 new £9 New 69 fimv .mQx Hm £9 mAN .mflx v.3 ”03 when .3 @8865 8:85am Emmy? 330208 05. .m2-~ Ho 8535 65 E noun—mi mmmmvh Am ”8082 B “6820 3:088 Av 6828 05 be vogue? 33me Am E628 _ow-.£o 2080:0808 0:0 $8 08 : 805028 5? 02806 as, as as 08000 200.810.8883? 8 £8800 :0 8080.5: ban.“ 05 a 88:8 a? 8:09 :o 80305 .v 950:: IIII'I‘r((( 129 130 Finally, the isolated C-domain was not sensitive to treatment with collagenase, also similar to the behavior of this domain in the intact polypeptide. Development and Characterization of Antisera M“ t rCBP35 Purified rCBP35 was used as an immunogen to generate polyclonal antiserum in two rabbits, designated as #32 and #33. Their antisera were compared against: (a) an antiserum derived from a rabbit (designated #24) immunized with CBP35 purified from mouse lung (2); and (b) the immunoglobulin fraction of a rat monoclonal antibody directed against mouse Mac-2 (22), whose cDNA sequence had indicated it to be identical to CBP35 (6). All four antibody preparations recognized CBP35 in Western blots of the immunogen, rCBP35 (Fig. SA-D, lane 3), as well as the extracts of mouse 3T3 cells (data not shown). Preimmune serum from each of the rabbits, when used at the same dilutions, failed to yield any immunoreactive bands on control Western blots. The four antibody preparations also immunoblotted CBP35 in extracts of human HeLa cells (data not shown). In this case, the molecular weight of the immunoreactive band was slightly lower than that of mouse CBP35, consistent with the length of the polypeptides predicted from the nucleotide sequences of the mouse and human cDNAs (5,7,9,11). Immunoblotting of the N- and C—domains was carried out with the four preparations of anti-CBP35. Antisera #24 and #33 immunoblotted the N-domain (Fig. 5A and 5C, lane 1), indicating that the principal epitopes recognized by these antibodies were in this portion of the polypeptide. Neither antisera reacted with 131 .803 .3 08030:: 80 0:0 N 8:me 8 0008 0: 003802 80 00:00:03 3903 830038 03:. 80:08.8 200:_-m-8030-v-0803-Q8:=0~088 0:308: 08083.: 08 33 0032.8 0080:0830 08—030-me H0380 80m :0 808039030 08830-03 0330380 80w .3 02005: 556 303 858 0250800 2F .980: no 0-82-08 a: 8 55:20 800: 8* 8:880 8%. G @826 82”: 08. 88380 fies E ”05026 on”: E0 528:3 830: A< ”803 mmmmu 380m0 0080308 0:9. 083030888 0:0 A0080Eo0 002V m000 556 203 8:8 0800860 29 .960: 000 0-82-80 :8 5 x8026 8000 8.0 8:808 fins 6 55026 08:: ~90 82380 88: Am A8026 and 00* 828050 05000 A< "0003 mmmmU :80w0 02000000 00,—. 0830300088 080 A0080050 002V m0