Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of an alphaherpesvirus resistant cell line Banfield, Bruce William 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-893184.pdf [ 2.95MB ]
JSON: 831-1.0088303.json
JSON-LD: 831-1.0088303-ld.json
RDF/XML (Pretty): 831-1.0088303-rdf.xml
RDF/JSON: 831-1.0088303-rdf.json
Turtle: 831-1.0088303-turtle.txt
N-Triples: 831-1.0088303-rdf-ntriples.txt
Original Record: 831-1.0088303-source.json
Full Text

Full Text

CHARACTERIZATION OFAN ALPHAHERPESVIRUS RESISTANT CELL LINEtyBRUCE WILLIAM BANFIELDB. Sc., the University of British Columbia, 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Microbiology and Immunology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAFebruary 1994© Bruce William Banfield, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of A,rc,cv ‘‘‘‘Ai _1YThe University of British ColumbiaVancouver, CanadaDate T1 /99VDE-6 (2/88)AbstractA mouse L cell mutant termed gro29 has been isolated (Tufaro et a!. 1987) andshown to be defective in the growth of two enveloped viruses, vesicular stomatitis virus (VSV) and herpes simplex virus type 1 (HSV-1). In this thesis it isdemonstrated that the rate of transport and processing of HSV-1 glycoproteinswas impeded in infected gro29 cells. In addition, HSV-1 virions failed to exit thecell and accumulated in cytoplasmic vacuoles. The phenotype of the uninfectedgro29 cell was examined. It was determined that gro29 cells were resistant to thelectins ricin and modeccin, and showed a reduced ability to bind fluorescentlyconjugated ricin. In addition, gro29 cells failed to synthesize theglycosaminoglycan chondroitin sulfate. The lectin binding properties of gro29cells and the glycosaminoglycan synthesis profile exhibited by these cellssuggested that gro29 might have suffered a defect in the metabolism of Nacetylgalactosamine (GalNAc). The metabolism of GalNAc was examined ingro29 cells and macromolecules directly involved in the synthesis andutilization of this molecule were found to be normal. The phenotype of gro29suggested that HSV-1 requires a host cell component for efficient virusmaturation and egress distinct from those components which facilitate thetrafficking of viral membrane glycoproteins such as VSV G protein. The resultspresented in this thesis suggest that it may be possible to interfere with cellsecretion or GalNAc metabolism such as to leave the cell viable yet impair theability of cells to propagate herpes simplex virus.Table of ContentsAbstract.iiTable of Contents iiiList of Figures viList of Abbreviations viiiAcknowledgment xiDedication xiiIntroduction IThe herpesviruses IThe medical significance of the herpes simplex viruses 2Herpes simplex virus structure 3Herpes simplex virus glycoproteins 7The herpes simplex virus lifecycle 12Study of the herpes simplex virus lifecycle 22Isolation of gro mutants able to survive exposure to HSV-1 23gro29 cells 23Materials and Methods 26Cells and viruses 26Materials 26Harvesting herpes simplex virus 27Determination of virus titre 28Pulse-chase labeling experiments 28Immunoprecipitations 28Endoglycosidase H digestions 29Western blots 29Fractionation of microsomes 30Indirect immunofluorescence 31IIIFluorescence microscopy using lectins .31Electron microscopy 32Analysis of HSV-1 gD in BFA treated cells 32Analysis of glycosaminoglycans 33Enzyme and protein assays 34Purification of Golgi membranes for nucleotide sugar translocationassays 34Nucleotide sugar translocation assays 35Glucose and sulfate uptake assays 36Chapter 1- Analysis of HSV-1 infected gro29 cells 38Results 38Production of infectious virus by gro29 cells 38Pulse-chase analysis of HSV-1 glycoprotein processing 39Analysis of gD transport in HSV-1 infected cells 46Analysis of virus egress 49Analysis of cell-associated virus 50Immunofluorescence and electron microscopic analysis ofinfected cells 58Brefeldin A treatment of HSV-1 infected cells 65Discussion 75Chapter 2- Characteristics of uninfected gro29 cells 83Results 83Lectin sensitivity of gro29 83Fluorescence detection of lectin binding molecules 84Analysis of glycosaminoglycan synthesis in gro29 cells 87Investigation of Ga1NAc metabolism in gro29 cells 93Rescue of chondroitin sulfate synthesis in gro29 cells 98ivMeasurement of UDPGa1-4-epimerase activity in cellextracts 101Measurement of nucleotide sugar translocation into Golgimembranes 106Discussion 116General Discussion 124References 128VList of FiguresFig. 1: Herpesvirus structure 4Fig. 2: Proposed models for the egress of alphaherpesviruses 19Fig. 3. Titre of virus released from HSV-1 infected L and gro29 cells 41Fig. 4. Processing of HSV-1 gD and gB in L and gro29 cells 44Fig. 5. Western blot analysis of HSV-1 gD in membrane fractions of L andgro29 cells 48Fig. 6. Comparison of HSV-1 particles released from L and gro29 cells 52Fig. 7. Nucleocapsid assembly in L and gro29 cells 54Fig. 8. Detection of intracellular virions 57Fig. 9. Immunofluorescence analysis of HSV-1 gD in HSV-1 infected Land gro29 cells 60Fig. 10. High resolution microscopy of HSV-1 infected L and gro29 cells 64Fig. 11. Immunofluorescence localization of cytoplasmic virions 67Fig. 12. Effects of BFA on the processing of HSV-1 gD during HSV-1infection in L cells 70Fig. 13. Effect of BFA on the processing of HSV-1 gD in gro29 cells 73Fig. 14. Fluorescence analysis of lectin-binding molecules in L and gro29cells 86Fig. 15. Analysis of Glycosaminoglycan synthesis in L and gro29 cells 90Fig. 16. Measurement of glucose uptake in L and gro29 cells 92Fig. 17. Rate of sulfate uptake in L and gro29 cells 95Fig. 18. Synthesis of UDPGaINAc 97Fig. 19. Translocation of UDPGa1NAc into Golgi membranes 100Fig. 20. Rescue of chondroitin sulfate synthesis in gro29 cells 103Fig. 21. Measurement of UDPGa1-4-epimerase activity in L and gro29 cellextracts 105viFig. 22. Conversion of UDPGa1NAc to UDPG1cNAc L and gro29 cellextracts 108Fig. 23. Rate of nucleotide sugar translocation into L and gro29 Golgimembranes 111Fig. 24. Kinetics of nucleotide sugar translocation into L and gro29 cellGolgi membranes 113viiList of AbbreviationsBFA brefeldin ABSA bovine serum albuminBrdU 5-bromo-deoxyuridineCMP cytidine monophosphateCMPSA cytidine monophosphate-sialic acidQ3E cytopathic effectCPM counts per minuteCS chondroitin sulfated day(s)DEAE diethylaminoethylDNA deoxyribonucleic acidEDTA ethylenediaminetetraacetic acidendo H endoglycosidase HER endoplasmic reticulumFACS fluorescence activated cell sortingFBS fetal bovine serumFITC fluoresciene isothiocyanateGal galactoseGalA glucuronic acidGa1NAc N-acetylgalactosamineGlc glucosamineG1cNAc N-acetylglucosamineG1cNSO3 N-sulfoglucosamineh hour(s)HA hyaluronic acidHAT hypoxanthine aminopterin thymidineviiiHPI hours post infectionHS heparan sulfateHSV-1 herpes simplex virus type 1HSV-2 herpes simplex virus type 2Kbp kilobase pairsKDa kilodaltonsMan mannosemm minute(s)MOl multiplicity of infectionPAGE polyacrylamide gel electrophoresisPBS phosphate buffered salinePFU plaque forming unit(s)PrV pseudorabies virusRNA ribonucleic acidRPM revolutions per minuteSA sialic acidSDS sodium dodecylsulfateTCA trichloroacetic acidTGN trans Golgi networkTLC thin layer chromatographyTK thymidine kinaseUDP uridine diphosphateUDPGa1 uridine diphosphate-galactoseUDPGa1NAc uridine diphosphate-N-acetylgalactosamineUDPG1c uridine diphosphate-glucosamineUDPG1cNAc uridine diphosphate-N-acetylglucosamineUL unique long region of the herpes simplex virus genomeixUMPK uridine monophosphokinaseU unique short region of the herpes simplex virus genomeVHS virus host shutoffVICV virus-induced cytoplasmic vacuoleVSV vesicular stomatitis virusVZV varicella-zoster virusw/v weight/volumew/w weight/weightxAcknowledgmentsI wish to acknowledge; F. Tufaro, S. L. Gruenheid, H. E. Meadows and K.Schubert for their contributions to data presented in this thesis, Michael Wiesfor assistance with microscopy and photography, the members of Frank Tufaro’slaboratory past and present for helpful discussions, Dr J. T. Beatty for remindingme from time to time that doing science is really a lot of fun, the members ofmy advisory committee: Drs. Jefferies, McMaster and Spiegelman for soundadvice through the years, Dr McMaster for invaluable assistance in thepreparation of this thesis, and Renée Finnen for her patience and critical readingof this thesis. Most of all I wish to thank Dr Frank Tufaro for his enthusiasmand the opportunity to work under his tutelage.xiThis thesis is dedicated to my parents, William and Gertrude, without whoseconstant support and encouragement this would not have been possible.xii1IntroductionThe herpesvirusesThe Herpesviridae comprise a large family of viruses which share somecommon features (reviewed in Roizman and Batterson 1986). Herpesvirusesare large enveloped viruses with a double stranded, linear, DNA genomeenclosed in an icosahedral nucleocapsid composed of 162 capsomeres. Aproperty shared by all herpesviruses is their ability to remain in a latent state inthe host in which they multiply. In excess of 80 herpesviruses have beencharacterized, infecting organisms as diverse as frogs and man.The members of the virus family, Herpesviridae, have been classifiedinto three subfamilies based on their biological properties; thealphaherpesvirinae, betaherpesvirinae, and gammaherpesvirinae (Roizmanand Batterson 1986). The alphaherpesvirinae (eg. herpes simplex virus,pseudorabies virus, and varicella-zoster virus) are characterized by a variablehost range, short reproductive cycle, rapid spread in cell culture, efficientdestruction of infected cells, and the ability to establish latent infection in cells ofthe ganglia. It has recently been suggested that the alphaherpesvirinae besubclassified into the simplexviruses and the zosterviruses because of thedifferences in the requirements for entry of these viruses into cells (Spear 1993).The betaherpesvirinae (eg. cytomegalovirus), are distinguished by a restrictedhost range, long reproductive cycle, slow spread in culture, enlargement ofinfected cells (cytomegalia), and the ability to be maintained in a latent state insecretory glands, kidneys, lymphoreticular cells, and other tissues.Gammaherpesvirinae (eg. Epstein-Barr virus) are extremely restricted in theirhost range, infecting lymphocytes in the family or order to which the naturalhost belongs. Gammaherpesvirinae infection is often arrested in a prelytic or2lytic stage without production of infectious virus. Latency of these viruses hasbeen found to occur in lymphoid tissues.The medical significance of the herpes simplex virusesThe use of the term herpes (from the Greek herpein, to creep) to describethe typical “cold sores” produced by this agent dates back about 25 centuries(Beswick 1962). However, it was not until the late 19th century when it wasdetermined that there was an infectious agent associated with the lesions. In1920 Grüter published experimental results which showed that the virus couldbe passaged from an experimentally infected rabbit to the cornea of a blind man(Grüter 1920). It was not until 1967 when it was determined that two distinctserotypes of herpes simplex viruses exist, HSV-1 and HSV-2 (Dowdie et al. 1967).It is now known that there are significant molecular differences between the twoviruses; HSV-1 and HSV-2 show roughly 50% homology at the nucleic acidlevel (McGeoch et al. 1991).The herpes simplex viruses (HSV-1 and HSV-2) are the causative agentsof a number of medically important diseases. In addition, infections with HSV1 and HSV-2 are among the commonest of human viral diseases (reviewed inMindel 1989). HSV-1 and HSV-2 each have the ability to cause the all theclinical symptoms observed for these infections although one or the other tendto predominate in a specific condition. These viruses are responsible for highlycontagious oral and genital lesions, which are the most common manifestationsof infection; it has been estimated that half a million new cases are reported eachyear in the USA and, as of 1984, 98 million people were infected in that countryalone. Usually HSV-1 is associated with oral lesions and HSV-2 with genitallesions although there are exceptions in some regions of the world. HSV-13induced ocular infections are the leading cause of corneal blindness in the USAwith 300,000 cases reported each year. HSV-1 is also the most common cause offatal endemic encephalitis and the most prevalent viral infection of the centralnervous system in the USA. Neonatal infection with HSV-2 is also a significantproblem with estimates ranging from 2.6-50 cases per 100,000 live births in theUSA, depending on the region. More recently, the role of herpes simplexinfection in AIDS patients has been examined (reviewed in Quinn 1990). It hasbeen suggested that herpes simplex virus infection may play a role in thetransmission, clinical presentation and pathogenesis of HIV infection. Becauseof the significant medical impact of herpes simplex viruses it is of considerableimportance to determine the function of viral and cellular molecules in thelifecycle of this pathogen, as this information may expose vulnerabilities whichcan be exploited in the treatment of disease.Herpes simplex virus structureThe herpes simplex virion consists of four structural elements: the core,capsid, tegument and the envelope (see Fig.1). It is estimated that between 30-40structural proteins comprise the herpes simplex virion (Heine et al. 1974; Spearand Roizman 1972) which represents approximately half of the coding capacityof the viral genome.The core of the virus contains a linear, double stranded DNA genome ofapproximately l52Kbp which is wound into a tight toroidal structure in thevirion (Furlong et al. 1972). In some electron micrographs of virions it appearsthat fibers connect the electron dense core to the inside of the capsid. The originor function of these fibers is unknown. The entire genome of HSV-1 has beensequenced (McGeoch et al. 1988; McGeoch et al. 1985; McGeoch et al. 1986). HSVDNA consists of two covalently linked components termed long (L) and short4Fig. 1. Herpesvirus structure.A) The core containing 152 Kbp of linear double stranded DNA wound into a torrus. B)The icosahedral nucleocapsid. C) The tegument. D) The viral envelope. E) Viralglycoproteins embedded in the envelope. (schematic of the virion courtesy of Y. Leduc).CDA BE5(S). The L region makes up about 82% of the genome and the S region about18%. Each region consists of mostly unique sequence (UL and Us) flanked byinverted repeats. The repeated sequence flanking UL comprise 12% of thegenome and the repeats flanking Us comprise 8.6% of the genome, thus aconsiderable portion of the viral genome is diploid.The genome encodes at least 77 genes; 58 of which are in the UL region(Barker and Roizman 1992; Liu and Roizman 1991b; McGeoch et al. 1988), 13 inthe Us region (Georgopoulou et al. 1993; McGeoch et al. 1985), and 3 in theinverted repeats which flank the unique regions of the genome (Ackermann etal. 1986a; Chou and Roizman 1986).Interestingly, all but one of the 13 genes which comprise the U5 region arenonessential for viral growth in culture (Longnecker et al. 1987; Longneckerand Roizman 1987; Weber et aL 1987). It has been suggested that the U5 regioncontains genes which although not absolutely required for growth in cultureevolved to allow HSV to survive in its human host, and that HSV may havearose by the insertion of Us sequences into a larger DNA component containinggenes essential for viral replication (Longnecker and Roizman 1987).The capsid is an icosahedron consisting of 162 capsomeres. Thecapsomeres are made up of several different polypeptides and are in the form ofa hexagonal prism with a hollow tube running the length of the long axis(Wildy et al. 1960). The capsid structure has 20 triangular faces and 12 apices.Each apex consists of a single capsomere surrounded by five others (pentons).The non-apical capsomers are surrounded by six others (hexons). The capsidtherefore has 150 hexons and 12 pentons for a total of 162 capsomeres. Sevenviral polypeptides have been identified as capsid components (Cohen et al. 1980;Gibson and Roizman 1972; Heilman et al. 1979) and are the product of six viralgenes (Davison et al. 1992; Liu and Roizman 1991a; McNabb and Courtney 1992;6Pertuiset et al. 1989; Preston et al. 1983; Preston et al. 1992; Rixon et al. 1990a).These proteins are VP5, VP19C, VP21, VP22a, VP23, VP24 and VP26.The tegument surrounds the capsid and appears as an amorphous fibrousstructure in electron micrographs. The tegument contains five major proteins,VPI /2, VP13, VP14, VPI6 and VP22 (Honess and Roizman 1973; Spear andRoizman 1972). It has been estimated that approximately 15 viral polypeptidesare found in this structure (Heine et al. 1974; Spear and Roizman 1972). Little isknown about the nature of the tegument, however, it has been determined thattwo viral transcription factors, VPI6 (also known as: VMW65 or czTiF) andICP4, are located in this structure (Honess and Roizman 1973; Spear andRoizman 1972; Yao and Courtney 1989). In addition, a viral host shutofffunction, VHS, which destabilizes both viral and cellular mRNAs is thought tobe associated with the tegument (Kwong and Frenkel 1987; Smibert and Smiley1990).Surrounding the tegument is a phospholipid bilayer envelope.Embedded in this envelope are viral glycoproteins. There are at least 11 HSVglycoproteins: gB, gC, gD, gE, gG, gH, gI, gj, gK, gL and gM (Ackermann et al.1986b; Baines and Roizman 1993; Baucke and Spear 1979; Buckmaster et al.1984; Heine et al. 1972; Hutchinson et al. 1992a; Hutchinson et al. 1992b;Longnecker et al. 1987; Para et al. 1983; Spear 1976; Spear 1985), 8 of which havebeen shown to be present in the virion envelope (all but gj, gK and gL)(reviewed in Spear 1993). It is likely that all of the proteins resident in the viralenvelope have not yet been identified because a number of viral open readingframes are predicted to encode polypeptides which have membrane spanningdomains (McGeoch et al. 1991) and as of yet have not been characterized.Furthermore, viral genes encoding proteins which have multiple membranespanning domains (serpentine membrane proteins) may also reside in the7viral envelope (MacLean et al. 1991). Baines and Roizman have recentlydescribed a viral glycoprotein, gM, which belongs to the class of serpentinemembrane proteins and is a resident of the viral envelope (Baines and Roizman1993).Interestingly a number of the genes encoding viral glycoproteins areclustered in the unique short region of the viral genome suggesting that thesegenes may have been inherited together by a precursor of the herpesviruses.These genes encode glycoproteins gE, gG, gI, and gD. The U6 gene, whichencodes gD, is the only essential viral gene found in the U region of the viralgenome.Herpes simplex virus glycoproteinsThe study of the structure and function of the viral glycoproteinsspecified by herpes simplex virus is one of the most concentrated efforts in thefield of herpes simplex virology. These molecules have been shown to beimportant in several stages of the viral life cycle including attachment, entryand egress and also in the host immune response to this pathogen. Theglycoproteins gB, gD, gH and gL are essential for viral propagation in tissueculture (reviewed in Spear 1993). These glycoproteins are thought to have rolesin the stable attachment of the virus with the cell and in the fusion of the viralenvelope with the cell membrane.Glycoprotein gB of HSV-1 is one of the most well characterized of theHSV glycoproteins and is found in the virus envelope and in the plasmamembrane of infected cells as an oligomer (Claesson-Welsh and Spear 1986;Spear 1976). gB is modified by the addition of N- and 0-linked carbohydratesand is essential for the growth of the virus in culture (Cai et al. 1987; Johnson8and Spear 1983; Spear 1976). The gene for this glycoprotein is highly conserved;in fact, 13 homologues from related herpesviruses have been cloned andsequenced and these genes show a great deal of homology. Consistent withthese data is the observation that a gB homolog has been found to functionallysubstitute in a HSV strain with a deleted HSV gB gene (Misra and Blewett 1991).gB has been suggested to play a role in the fusion of the viral envelope with theplasma membrane of the cell (Cai et al. 1988a; Highlander et al. 1988). Thisactivity was inferred from the observation that many viral mutants which formpolykaryons on monolayers of cells (termed syn mutants) mapped to the geneencoding gB (Bzik et al. 1984b; Cai et al. 1988a; Cai et al. 1988b). Furthermore,antibodies against gB inhibit fusion of the virus envelope with the cell, but notthe attachment of the virus to the cell, suggesting that the role of this moleculeis in membrane fusion (Highlander et al. 1988). Recently, regions of the gBmolecule which are important for fusogenic activity have been identified(Baghian et al. 1993; Gage et al. 1993).It has also been shown in vitro that gB isolated from solubilized virusescan bind to heparin (Herold et al. 1991). Heparin is a glycosaminoglycan whichis structurally similar to heparan sulfate, a molecule found on the surface ofcells and implicated in the initial attachment of the virus to the cell (Shieh et al.1992; WuDunn and Spear 1989). In addition, viruses lacking gB in theirenvelopes have a slightly decreased ability to bind to cells (Herold et al. 1991).These data may suggest that gB can act in the initial attachment of the virus tothe cell.Glycoprotein gC is another well characterized viral glycoprotein whichpossesses both N- and 0-linked carbohydrate modifications (Dall’Olio et al. 1985;Johnson and Spear 1983; Olofsson et al. 1981; Spear 1976) and can bind to thethird component of complement, C3b (Friedman et al. 1984b). Although not9essential for viral propagation in culture this molecule appears to have a majorrole in the attachment of virus to cells through heparan sulfateglycosaminoglycans (Herold et al. 1991; Shieh et al. 1992). Addition of heparansulfate to the extracellular medium of cells during HSV infection reduces theability of the virus to infect cells (WuDunn and Spear 1989). Herold and othershave show that mutant viruses lacking gC in their envelopes bind to cells with10-fold less affinity, and virions devoid of gC do not respond as greatly toinhibition by heparin as do wild type virions (Herold et al. 1991).Glycoprotein gD is an essential viral glycoprotein (Ligas and Johnson1988) which has both N- and 0-linked carbohydrate moieties attached (Johnsonand Spear 1983; Spear 1976). It appears that gD binds to a specific receptor on thecell surface and is involved in an intermediate attachment step between theinitial attachment of the virus with heparan sulfate at the cell surface and thefusion of the viral envelope with the plasma membrane of the cell. Theevidence for this is as follows; 1) antibodies against gD inhibit penetration of thevirus into the cell but not attachment (Highlander et al. 1987), 2) mutant virusesdevoid of gD are able to attach to cells but are unable to penetrate into the cell(Ligas and Johnson 1988), 3) viruses lacking gD are unable to inhibit thepenetration of wild type virus (Johnson and Ligas 1988), 4) a soluble truncatedform of gD is able to inhibit penetration of virus into the cell, suggesting thepresence of a cell surface receptor for gD (Johnson et al. 1990), and 5) expressionof gD in the plasma membrane of cells inhibits the penetration of virus, alsosuggesting the presence of a cell surface receptor for gD (Campadelli-Fiume et al.1990). In addition, studies with gp5O, the gD homologue of PrV, suggest that thisglycoprotein enables the virus to attach to the cell in a manner which is notcompetable by exogenous heparan sulfate (Karger and Mettenleiter 1993). It hasalso been suggested that gD may function in the fusion of the virus envelope10with the cell membrane, as it has been shown that cell lines expressing gD canfuse (Campadelli-Fiume et al. 1988b).gD has been shown to be a predominant target of the host humoral andcellular immune responses (Cohen et al. 1978; Martin et al. 1987; Torseth et al.1987; Zarling et al. 1986). Furthermore, immunization with gD can protectanimals from a lethal challenge of HSV (Chan 1983; Long et al. 1984), andmonoclonal antibodies reactive with gD exhibit potent neutralizing activityagainst the virus (Para et al. 1985). Much is also known about the structure ofgD. In particular, the mapping of epitopes on gD reactive with a panel ofmonoclonal antibodies (Cohen et al. 1986; Eisenberg et al. 1982; Eisenberg et al.1985; Highlander et al. 1987; Showalter et al. 1981) has allowed the study ofdisulfide bond formation and N-glycosylation on the structure and function ofgD (Cohen et al. 1988; Isola et al. 1989; Long et al. 1990; Muggeridge et al. 1988;Muggeridge et al. 1990; Sodora et al. 1989; Sodora et al. 1991a; Sodora et al.1991b; Wilcox et al. 1988).gE has been shown to be modified by N- and 0-linked oligosaccharides,fatty acylation and sulphation (Baucke and Spear 1979; Hope et al. 1982;Johnson and Spear 1983; Para et al. 1983). gE has been identified as being animmunoglobulin G Fc binding protein (Baucke and Spear 1979). gE forms acomplex with another viral glycoprotein, gI, and it is this complex which acts asthe functional Fc receptor (Johnson and Feenstra 1987; Johnson et al. 1988). Therole of this function in viral infection is unclear, but it has been suggested that itmay aid in the viruses evasion of the host immune system. It has also beenproposed that the Fc binding activity might act in the binding of the virus to thecell through cell surface molecules structurally related to immunoglobulins(Johnson and Feenstra 1987) (such as cell surface adhesion molecules, andmolecules of the major histocompatability complex). If this interaction exists it11not essential for the replication of the virus in culture, as deletion of the geneencoding gE has little effect on the replication of the virus (Longnecker andRoizman 1987).Relatively little is known about the function of the other viralglycoproteins. As mentioned above gI forms a complex with gE to form an Fcreceptor. gH is an essential viral glycoprotein thought to be involved in thefusion of the virus membrane with the cell (Fuller et al. 1989). gH has beenshown to form a heterodimer with a recently described glycoprotein, gL, and thisinteraction is essential for the normal folding and cell surface expression of gH(Hutchinson et al. 1992a). In addition, it has been suggested that gH has a role inviral egress (Desai et al. 1988). gL has also been shown to be essential for thegrowth of the virus in culture (Roop et al. 1993) and is likely involved in thefusion of the viral envelope with the membrane of the cell. gK is thought tohave a role in the fusion of viral and cellular membranes based on theobservations that mutations in the gene for gK can cause increased fusion ofinfected cells (DebRoy et al. 1985; Pogue-Guile and Spear 1987), although it hasnot yet been determined if gK is present in the viral envelope. gM, the mostrecently identified glycoprotein, is the only viral glycoprotein described that is amember of the class of serpentine membrane proteins (Baines and Roizman1993).A number of studies have determined that complete glycosylation ofviral glycoproteins is not essential for virion infectivity (Campadelli-Fiume etal. 1982; Serafini-Cessi et al. 1983). These studies indicated that HSV infection ofmutant cell lines defective in glycosyltransferases involved in the biosynthesisof N-linked glycans produced infectious progeny virus. The precise role ofglycosylation on the viral glycoproteins remains unclear.12The herpes simplex virus lifecycleAttachment, binding and penetrationBinding and penetration of HSV into cells is mediated by the interactionof viral envelope glycoproteins with cell surface molecules. The first step in theherpes simplex virus lifecycle is the adsorption of the virus to the cell surface(reviewed in Spear 1993). This process is mediated by the interaction of the viralglycoprotein gC with components of the cells extracellular matrix, in particularheparan sulfate proteoglycans (Herold et al. 1991; WuDunn and Spear 1989). Itshould be noted that this is an oversimplified view of these processes, as otherstudies have implicated the role of molecules independent of the gC/heparansulfate interaction which mediate viral infection in some cell types. Inparticular, Sears and others were able to demonstrate that in MDCK cells, apolarized epithelial cell line, at least two cell surface receptors exist; one on theapical surface of the cell, and one on the basal surface of the cell. The apicalreceptor was dependent on gC in the virus envelope for its function whereas thebasal receptor functioned independently of gC and was as equally efficient as thereceptor on the apical surface (Sears et al. 1991). These data indicated thatalthough the gC/heparan sulfate interaction appeared to be the predominantinteraction in the absorption of the virus to the non-polarized cell lines tested inculture, it was clear that the virus has other means of interaction with the cellindependent of this route. In another study (Gruenheid et al. 1992), mouse Lcell lines selected for their ability to survive herpes simplex virus infection wereisolated and found to be defective in the biosynthesis of heparan sulfateglycosaminoglycans. HSV plaques could still form on monolayers of thesemouse L cells although the numbers were reduced by approximately 10 fold;similar to the reduction in plaque formation observed with virus lacking gC inthe envelope (Herold et al. 1991). Two important conclusions can be drawn13from this study; heparan sulfate is an important molecule for the interaction ofthe virus with the cell under “normalt’circumstances because a mutation in thebiosynthesis of this molecule was identified using the virus as a selective agent,and the presence of this molecule is not essential for infection which suggeststhat other as yet unidentified interactions take place that facilitate infection.The next step in the viral entry pathway is thought to be a “stable”attachment of the virus to the cell surface through gD with a cell surfacereceptor (Johnson et al. 1990; Johnson and Ligas 1988). Unlike the gC/heparansulfate interaction the gD/receptor reaction is essential for productive infectionto occur because mutant virus lacking gD cannot infect cells. The gD/receptorbinding has been termed “stable” in that wild type viruses can enter into anattachment with the cell which is not elutable with heparin whereas viruseslacking gD do not become resistant to the addition of exogenous soluble heparinas readily (Fuller and Lee 1992). The specific roles of gill, a gC homolog, andgp5O, a gD homolog, in the attachment of the closely related alphaherpesvirus,PrV, has been examined in great detail. Karger and Mettenleiter (Karger andMettenleiter 1993) have demonstrated the biphasic attachment of PrV to cellsfrom a heparin sensitive binding to a heparin insensitive binding. The heparinsensitive binding was shown to depend on the presence of gill and the heparininsensitive binding was shown to depend on the presence of gp5O in the viralenvelope. It is likely that the same mechanisms of attachment exist for bothPrV and HSV.HSV is thought to enter the cell primarily by fusion of the viral envelopewith the plasma membrane of the cell. Although the virus has been found toenter the cell by endocytosis, it is not thought that this route leads to aproductive infection insofar as drugs affecting endocytosis and endosomalacidification have no effect on the efficiency of infection (Campadelli-Fiume et14al. 1988a; Koyana and Uchida 1984; Wittels and Spear 1991). It is likely thatHSV which enters by this route is degraded in the lysosome (Campadelli-Fiumeet al. 1988a). Studies using monoclonal antibodies directed against viralglycoproteins have suggested that HSV glycoproteins gB, gD and gH might beinvolved in the fusion of the virus envelope with the cell (reviewed in Spear1993), furthermore the deletion of the genes encoding these proteins from theviral genome results in defective viruses which are unable to penetrate cells(Cal et al. 1988a; Forrester et al. 1992; Ligas and Johnson 1988). In addition,identification of viral molecules involved in this process has come in part fromthe observation that viral mutants which stimulate the fusion of infected cells(syn mutants) occur at relatively high frequency (Brown et al. 1973; Hoggen andRoizman 1959; Person et al. 1976; Read et al. 1980). A number of genes whichencode HSV glycoproteins have been shown to harbor these mutations. Inparticular, gB, gD, gK and gL have been found to have syn mutations, whichsuggested that these molecules may be involved in fusion of the virus envelopewith the cell membrane (reviewed in Spear 1993). Although the mechanism bywhich fusion of these two membranes takes place is unclear, a model has beenproposed based on what is known about the molecules involved.The process of gD interacting with its cellular receptor it is thought tostimulate conformational changes in the viral molecules involved in fusionsuch that a functional “fusion complex” (likely consisting of gD, gB, gH, gL andpossibly gK) can form. It is also possible that other components of the virion areinvolved in the fusion reaction. In particular, a temperature sensitive HSV-1mutant with a defect in penetration has been isolated and shown to have amutation in the UL25 gene, the product of which is thought to be a constituentof the tegument (Addison et al. 1984). It is clear that much work remains to be15done before a comprehensive understanding of the events involved in theabsorption, binding and penetration of herpes simplex viruses exists.Uncoating, gene expression and replicationFollowing the fusion of the viral envelope with the cell membrane thenucleocapsid enters the cytoplasm and travels to a nuclear pore where thegenome is released into the cell’s nucleus. It has been hypothesized that thenucleocapsid travels down cytoskeletal elements in a fashion similar to thatdescribed for adenoviruses (Dales and Chardonnet 1973; Roizman andBatterson 1986). Release of the viral DNA from the nucleocapsid through thenuclear pore and into the nucleus is known to require a viral function because atemperature sensitive mutant of HSV-1 has been isolated which can travel tothe nuclear pore but is unable to release viral DNA into the nucleus at the non-permissive temperature (Batterson et al. 1983).Upon release of the viral genome into the nucleus of the cell a temporalcascade of gene expression ensues (Roizman and Batterson 1986). The virususes the host cell RNA polymerase II in association with cellular and virallyencoded transcription factors to transcribe its genes. There are at least fiveclasses of HSV-1 genes; a, f2, Yi, and Y2• The a genes are the first to betranscribed and reach a peak level of expression from 2-4 hours post-infection(HPI). Some of the a gene products are required for the synthesis of the othergroups of genes. The j3i and the 132 genes reach peak expression levels from 5-7HPI. These temporal classes of genes are distinguished from each other in thatthe 13i genes require the a gene product, a4 (ICP4), for their transcriptionwhereas the 132 genes do not. The Ii and the 132 genes encode proteins involvedin viral genome replication. The Yi and Y2 genes have a requirement forconcurrent viral DNA synthesis for their maximal expression. In the presence16of inhibitors of viral DNA replication the y genes are expressed at a much lowerlevel than that observed during normal infection, whereas f gene expression isunaffected. Conversely, if viral DNA synthesis is blocked the Y2 genes are notexpressed at all. The majority of y genes encode viral structural proteins.As mentioned previously HSV encodes a protein, VHS, that destabilizesboth cellular and viral mRNAs. It has been suggested that a possible role of thismolecule is to help regulate the temporal cascade of gene expression by helpingto degrade messages soon after their synthesis. This action of VHS would helpto modulate the transition between the translation of early genes and late genetranscripts.Viral DNA synthesis is thought to take place at least in part by a rollingcircle mechanism. This means that the parental genome must circularize priorto DNA replication, a process which has been demonstrated in vitro(Wadsworth et al. 1975). The viral genome is replicated using a virally encodedDNA polymerase. The concatomeric replicated viral DNA is cut into genomesand packaged into empty nucleocapsids, in a process requiring at least sevenvirally encoded proteins (Addison et al. 1984; Addison et al. 1990; Al-Kobaisi etal. 1991; Liu and Roizman 1991a; Sherman and Bachenheimer 1987; Shermanand Bachenheimer 1988). These proteins are the products of the UL6, UL25,UL28, UL32 and UL36 genes, and their role in the assembly of nucleocapsids hasbeen inferred from the analysis of temperature sensitive mutations in thesegenes. In addition, the role of the two products of the UL26 gene in theencapsidation of viral DNA has been examined and these proteins have beenfound to be essential for viral replication (Gibson and Roizman 1972; Liu andRoizman 1991a; Preston et al. 1983; Weinheimer et al. 1993).17Egress of herpes simplex virusesAn outline of the models for viral egress are presented in Fig.2. The exactmechanism of herpesvirus egress is unknown. Several views exist on the siteand mechanism of viral envelopment. The early stages of viral egress are wellaccepted (reviewed in Roizman and Bafterson 1986). In the first step,nucleocapsids containing viral DNA associate with a modified region of theinner nuclear membrane and become enveloped resulting in envelopedparticles in the perinuclear space (Darlington and Moss 1968; Morgan et al. 1959;Nii et al. 1969). It is thought that the modifications at the site of envelopmentrepresent the accumulation of viral tegument proteins, and that upon buddingof the nucleocapsid into the perinuclear space the virion acquires a tegument aswell as an envelope derived from the inner nuclear membrane. The innernuclear membrane of infected cells has been shown to contain precursor formsof viral envelope glycoproteins (Compton and Courtney 1984). This is notsurprising because the membrane of the rough endoplasmic reticulum (the siteof synthesis of viral glycoprotein) is continuous with the inner nuclearmembrane. Therefore it is likely that the enveloped virus particles in theperinuclear space contain “immature” precursor forms of the viralglycoproteins. It is at this point where current models for viral egress diverge.In one model, the virions in the perinuclear space enter the endoplasmicreticulum, and are transported through the secretory apparatus en route out ofthe cell. The viral envelope glycoproteins acquired in an immature state fromthe inner nuclear membrane (Compton and Courtney 1984) are processed totheir mature forms as the virion is transported through the successive cisternaeof the Golgi apparatus. Finally, the mature virion is secreted from the cell. Theevidence for this model comes from electron microscopic studies (Morgan et al.1959), and from the biochemical analysis of HSV infection in the presence of the18Fig. 2. Proposed models for the egress of alphaherpesviruses.1) Model proposed by Johnson and Spear. Briefly, virions bud into the perinuclear spaceand obtain a tegument and an envelope containing “immature” viral glycoprotein. Thevirions then move from the perinuclear space to the endoplasmic reticulum (ER) and aretransported through the Golgi apparatus and the trans Golgi network (TGN) beforebeing secreted from the cell. 2) Model proposed by Jones and Grose/ Whealy andothers. Briefly, virions which have budded into the perinuclear space de-envelope at theouter nuclear membrane or the endoplasmic reticulum releasing naked nucleocapsidsinto the cytoplasm. The cytoplasmic nucleocapsids then bud into virus inducedcytoplasmic vacuoles (VICV) (Jones and Grose) or a late Golgi component (TGN)(Whealy et al.) which contain “mature” viral glycoprotein, before being secreted fromthe cell. 3) Point at which the models converge.19usPenucle space**‘ Cytoplasm( - Trans Golgi1Cis Golgi DMedial Golgi/IPlasma Membrane*20carboxylic polyether ionophore monensin (Johnson and Spear 1982). Monesintreatment of cells has been shown to cause the accumulation of vacuolesderived from the Golgi apparatus and block the secretion of soluble proteins andthe transport of membrane proteins from the Golgi to the plasma membrane ofthe cell (Johnson and Schlesinger 1980; Tartakoff and Vassalli 1977; Tartakoffand Vassalli 1978). Monensin is thought to act by disrupting ion gradients in thecell which appear to be critical for the budding of transport vesicles from theGolgi complex (Pressman 1976). Johnson and Spear (Johnson and Spear 1982)were able to show that HSV infected cells treated with monensin did not releaseinfectious virus from the cell, although infectious virus was found inside thecell. In addition, viral glycoproteins were underprocessed and were found inassociation with intracellular enveloped virions, and virus particles inside thecell accumulated in vacuoles likely derived from the Golgi apparatus. Takentogether these results suggested that HSV particles associated with membranesof Golgi apparatus while the viral envelope glycoproteins were in an“immature” form. The processing of glycoproteins to “mature” forms takesplace in the Golgi apparatus. Therefore it is conceivable that the virus istransported through the successive cisternae of the Golgi apparatus and theglycoproteins resident in the viral envelope become fully glycosylated as thisoccurs. These results suggested that HSV may be released from cells in a fashionsimilar to that described for soluble, secreted proteins.In another model for alphaherpesvirus egress, the nucleocapsid whichhas budded into the perinuclear space de-envelopes at the outer nuclearmembrane or the endoplasmic reticulum (effectively the same membrane)releasing the nucleocapsid into the cytoplasm (Whealy et al. 1991). Thesecytoplasmic particles then bud into cytoplasmic membranes derived from theGolgi apparatus (Jones and Grose 1988; Stackpole 1969; Whealy et al. 1991;21Whealy et al. 1990b) and thereby acquire an envelope. These membranescontain fully processed viral glycoproteins. The virus is then secreted from thecell. This model is supported by the study of two alphaherpesviruses, VZV andPrV (Jones and Grose 1988; Whealy et al. 1991). Jones and Grose showed byquantitative electron microscope autoradiography that the predominant site ofVZV envelopment was at virally induced post-Golgi cytoplasmic vacuoleswhich contained viral glycoprotein. Whealy and others were able to show thatin PrV infected cells treated with the drug brefeldin A (BFA) enveloped particlesaccumulated in the perinuclear space and unenveloped particles werepredominant in the cytoplasm. BFA causes the redistribution of membranesfrom the Golgi apparatus to the endoplasmic reticulum (Lippincott-Schwartz etal. 1990; Lippincott-Schwartz et al. 1989). Furthermore, in PrV infected cells nottreated with BFA, nucleocapsids were found in close proximity to membranes ofthe trans-Golgi in structures which were thought to represent envelopmentintermediates.Whether these data supporting the two models of viral egress conflictwith each other or not remains to be determined. It may be that both pathwaysof egress described above are active in alphaherpesvirus infection and that oneor the other predominate for a particular virus. There is also evidence tosuggest that the type of cell line used for the study may have an influence on theobservations of virus-Golgi interactions (Campadelli-Fiume et al. 1993). Inparticular, Campadelli-Fiume and others observed that infection with HSVinduced the redistribution of Golgi components in some cell types but not inothers. Furthermore, the role of virally encoded molecules involved in egresshas been shown to be cell specific (Baines et al. 1991). Baines and others wereable to determine that the product of the HSV-1 UL2O gene was necessary forviral egress beyond the perinuclear space in Vero cells but was not required for22HSV-1 egress in 143 tlc cells. It is clear that further study is required to elucidatethe mechanisms of herpesvirus egress.Study of the herpes simplex virus lifecydeSeveral approaches have been taken to determine the roles of viralproteins and cellular functions in the lifecycle of herpes simplex viruses. Theinvolvement of particular cellular structures in the lifecycle of HSV has beeninferred from microscopic examination of infected cells. The engineering ofviral mutants defective in one gene and the subsequent examination of theability of the mutant virus to propagate in a host cell has furthered theunderstanding of the role of many viral proteins. In addition, the isolation oftemperature sensitive mutants in essential viral genes has proved useful forexamining the role of these gene products. The use of drugs to block specificcellular processes has been very helpful in determining some of the cellularrequirements for productive HSV-1 infection, and learning about themechanisms of viral egress. Infection of well characterized somatic cell mutantsdefective in known enzymatic functions have also aided in the discovery ofsome of the cellular requirements for viral infection. Although the methodsdescribed above have yielded much information about herpes simplex virusinfection, there is another effective way to identify important virus-cellinteractions. The approach is to use the virus as a selective agent to isolatesomatic cell mutants resistant to herpes simplex virus infection. Thisprocedure, outlined below, has been used successfully to isolate several mutantcell lines. This technique has identified virus-cell interactions which arerequired for viral infection but are not essential for the growth of cells inculture. It is thought that this approach will be useful for identifying targets forantiviral drugs which could inhibit viral infection effectively yet not grosslyeffect cell viability.23Isolation of gro mutants able to survive exposure to HSV-1The observation that cells can mutate from a virus sensitive state to avirus resistant state is not a new one (Luria and Deibruck 1943). The rationalefor attempting to isolate and characterize somatic cell mutants able to surviveexposure to HSV came from the great success of earlier studies in which mutantEsherichia coli (E. coli) surviving exposure to phage were isolated (Friedmanet a!. 1984a). These mutant strains of E. coli (some of which were called gromutants) showed defects in many aspects of the ?. life cycle including attachmentof the phage to cells, transcription of viral genes, phage DNA replication andassembly of new phage particles.To isolate mutant cells able to survive exposure to HSV-1 mouse LMtkcells were mutagenized with ethyl methanesulfonate prior to exposure to HSV1. Colonies of cells surviving exposure to HSV-1 were cloned and furtheranalyzed for their ability to propagate HSV-1. These mutant cells were termedgro mutants (Tufaro et al. 1987). Mouse L cells were chosen as a parental cellline because they readily propagate HSV-l and grow well in culture.The majority of the gro mutants isolated shared the same phenotype. Onesuch gro mutant, gro2C, has been characterized extensively (Gruenheid et al.1992). These cells were defective in the synthesis of the glycosaminoglycanheparan sulfate, a molecule which has been implicated in the initial interactionof the virus with the cell (Shieh et al. 1992; WuDunn and Spear 1989). Anothermutant, gro29, had a considerably different HSV-1 resistance phenotype; theanalysis of this cell line is the subject of this thesis.gro29 cellsIn a previous study, a mouse L cell mutant termed gro29 was isolated(Tufaro et al. 1987) and shown to be defective in the growth of two envelopedviruses, vesicular stomatitis virus (VSV), which buds from the plasma24membrane, and herpes simplex virus type 1 (HSV-1), which buds from theinner lamellae of the nuclear membrane (Darlington and Moss 1968; Morgan etaL 1959; Nii et al. 1969). Initial characterization of this cell line showed thatVSV-infected gro29 cells were defective in the transport and processing ofnewly-made G protein, the envelope glycoprotein of VSV. Despite this defect,the release of infectious VSV from gro29 cells was diminished only three-foldwhen compared with the normal parental L cells, suggesting that the secretorydefect in this cell line is not critical for the maturation and egress of VSV fromthe plasma membrane. The effect of this lesion on the release of infectiousHSV-1 is very different, however. Although gro29 cells are infected efficientlyand the replication cycle proceeds to the late stages of viral gene expression, thespread of HSV-1 from cell to cell does not occur and under-processedglycoproteins accumulate inside the gro29 cells at late times of infection (Tufaroet al. 1987).To further the understanding of the virus-host interactions governingthese processes, the nature of the block to virus production in the gro29 cell linewas investigated. In this thesis it is demonstrated that the rate of transport andprocessing of HSV-1 glycoproteins from their site of synthesis in the ER to thecell surface was impeded in infected gro29 cells. This defect in protein transportreduced but did not eliminate the appearance of viral glycoproteins in theplasma membrane of the infected cells. By contrast, newly-assembled virionsfailed to exit the host cell. In infected gro29 cells, the intracellular virionscontained predominantly immature forms of the envelope glycoproteins andaccumulated in cytoplasmic vacuoles resembling those that accumulate in cellstreated with the carboxylic ionophore monensin, which also blocks HSV-1egress (Johnson and Spear 1982). If the observations of HSV-1 infection in gro29cells represent intermediates of a “normal” infection then the data presented in25this thesis suggest that the egress of HSV-1 in L cells occurred in a fashionsimilar to that described by Johnson and Spear (Johnson and Spear 1982).Furthermore, the phenotype of the gro29 mutant cell line was unique withregard to HSV-1 infection and the experiments suggested that HSV-1 required ahost cell component for efficient virus maturation and egress that was distinctfrom those components which facilitated the trafficking of viral membraneglycoproteins such as VSV G protein.To further examine the nature of the mutation(s) in gro29 cells thephenotype of the uninfected cell was examined. It was determined that gro29cells were resistant to the lectins ricin and modeccin, a unique lectin resistantphenotype described for only one other cell line (Laurie and Robbins 1991).These lectins bind to galactose (Gal) and N-acetylgalactosamine (Ga1NAc)containing glycoconjugates. Furthermore, gro29 cells showed a reduced abilityto bind fluorescently conjugated ricin, but not fluorescently conjugated RCA120;which binds Gal. In addition, gro29 cells failed to synthesize theglycosaminogycan chondroitin sulfate normally but were able to synthesizeheparan sulfate glycosaminoglycans. A major difference between chondroitinsulfate and heparan sulfate is that chondroitin sulfate contains Ga1NAc whereasheparan sulfate does not. Based on the lectin binding properties of gro29 cellsand the glycosaminoglycan synthesis profile exhibited by these cells it wassuggested that gro29 might have suffered a defect in the metabolism of Ga1NAc.26Materials and MethodsCells and virusesThe clone 1D line of LMtk mouse fibroblasts was the parental cell line ofthe mutant gro29 cells. The mutant gro29 cell line was obtained from F. Tufaroand was selected for the ability to survive HSV-1 infection (Tufaro et al. 1987).Vero cells were obtained from S. McKnight. All cells were grown at 37°C inDulbecco’s modified Eagles medium (DMEM) supplemented with 10% fetalbovine serum (FBS) in a 5% CO2 atmosphere. The virus used was the HSV-1KOS strain obtained from D. Coen. Stocks of HSV-1 used for infections weregrown in Vero cells.MaterialsAsialofetuin, 5-Bromo-deoxyuridine (BrdU), Bovine serum albumin(BSA, fraction five), UDPG1c dehydrogenase, UDPGa1NAc, UDPG1cNAc,UDPGa1, UDPG1c, pronase, chondroitinase ABC lyase, heparitinase,hyaluronidase, shark cartilage chondroitin sulfate, PET cellulose and lectins(except for modeccin) were obtained from Sigma (St. Louis, MO). Modeccin wasobtained from Inland Scientific (Austin, Texas). Scintilation cocktails (ReadySafe and Ready Prot.), all ultracentrifuge rotors and tubes, and HPLC columnswere from Beckman (Palo Alto, California). Anti-BrdU monoclonal antibody,cytochrome C (from horse heart), NADPH, CMP-sialic acid, and endoglycosidaseH were from Boehringer Mannheim (Laval, Quebec). DEAE Sephacel was fromPharmacia LKB (Piscataway, New Jersey). Zwittergent was from Calbiochem (LaJolla, California). All radioactive reagents were purchased from Dupont-NEN(Mississauga, Ontario) with the exception of[35S1-sulfate and D-[6-3H]-glucosamine which were purchased from ICN (St. Laurent, Quebec). Brefeldin A27(BFA) was from BioCan (Mississauga, Ontario). All tissue culture products andall conjugated secondary antibodies were purchased from Gibco/BRL(Burlington, Ontario). Antibodies to HSV-1 glycoproteins were kind gifts fromM. Zweig and R. Philpotts. Polyclonal antisera against p58 was a kind gift from J.Saraste.Harvesting herpes simplex virusMedium was removed from infected cell monolayers (lOml) andsubjected to low-speed centrifugation (600 Xg, 10 mm, 4°C) to pellet cell debris.The supernatant was centrifuged for 2 h at 20,000 Xg to pellet the virions, andthe resulting pellet was resuspended in 1 ml of 10 mM Tris, pH 7.8, 50 mM NaC1on ice. This material was sedimented through a 10 ml 5-40% Dextran T10 gradient formed in 50 mM NaCl, 10 mM Tris pH 7.8 for 1 h at 22,000 RPM in aBeckman SW41 rotor. Gradients were fractionated from the bottom of the tubeinto 0.5 ml fractions. For determination of radioactivity in insoluble material,10% of each fraction to be analyzed was added to 50 jig BSA (lmg/ml) followedby I ml of 10% cold trichloroacetic acid (TCA). Insoluble material was collectedonto filters after 1 h on ice and radioactivity was determined by liquidscintillation spectroscopy. For determination of virus titres, fractions werediluted serially with medium and used to inoculate monolayers of Vero cellsgrowing in 96-well dishes. Titres were calculated when generalized cytopathiceffect (CPE) was noticed in control infected samples. For electrophoreticanalyses, samples of fractions to be analyzed were centrifuged at 436,000 g for 20mm in a Beckman TLA 100.2 rotor. Pelleted material was solubilized in sodiumdodecylsulfate (SDS) sample buffer and subjected to SDS polyacrylamide gelelectrophoresis (SDS PAGE) as described (Laemmli, 1970). Followingelectrophoresis, gels were fixed, dried and autoradiography was performed. In28some cases, protein in gels was transferred to nitrocellulose membranes forwestern blot analysis prior to autoradiography.Determination of virus titreMedium was completely removed from infected cell monolayers atvarious times of sampling and replaced with fresh medium. Each sample ofmedium to be titred was centrifuged (600 Xg, 10 mi 4°C) to remove cell debris.Serial ten-fold dilutions of each sample were made and used to inoculateconfluent monolayers of Vero cells. After I h, the innoculum was removed andthe monolayers were overlayed with fresh DMEM containing 4% FBS and 0.8%agar or Methocell. Duplicate wells for each sample were analyzed after threeand five days.Pulse-chase labeling experimentsMonolayers of L cells and gro29 cells growing in 60mm dishes wereinfected with HSV-1 (multiplicity of infection (MOl) =10). After I h the viruswas removed and replaced with complete medium. At 5 h post infection (HPI)cells were washed three times with methionine-free medium and labeled for 30mm with 100 .tCi/ml[35S]-methionine (1117 Ci/mmol) in methionine-freemedium containing 5% dialyzed FBS. At the completion of labeling, cells wereeither harvested immediately or washed three times and incubated inDMEM/10% FBS (which contains excess methionine) for various times. Cellswere harvested by washing the monolayer with cold phosphate buffered saline(PBS) and incubated for 15 mm with 1 ml cold lysis buffer (10 mM Tris-HC1 pH7.4, 150 mM NaCl, 1% Nonidet P-40 (NP-40), 1% Na-deoxycholate).Immunoprecipitations1/10 volume of 10% NP-40, 10% Na-deoxycholate, 1% SDS was added toaliquots of cell lysates to be precipitated. The appropriate volume of anti-HSV-129gD or anti-HSV-1 gB monoclonal antibody was added to each aliquot and thesample incubated overnight on ice. A 10% suspension of Staphylococcus aureuscells was then added to each sample equal to 20 volumes of primary antibody.This mixture was rocked at 4°C for 2 h before the cells were pelleted and washedonce each with 0.5 ml of the following solutions; wash buffer #1 (20 mM TrisHC1 pH 7.5, 150 mM NaC1, 1% NP-40), wash buffer #2 (20 mM Tris-HC1 pH 8.8,150 mM NaC1, 1% NP-40, 0.2% SDS), and wash buffer #3 (20 mM Tris-HCL pH6.8, 150 mM NaC1, 1% NP-40, 0.2% SDS). The final pellet was resuspended inSDS sample buffer, heated at 100°C for 5 mm and subjected to electrophoresisthrough a 10% SDS polyacrylamide gel. Gels were fixed, dried and exposed toKodak XAR-5 film for autoradiography.Endoglycosidase H digestionsThe Staphylococcus aureus pellets were resuspended in 40 p.1 2X endo Hbuffer (1% SDS, 5% 13-mercaptoethanol, 2 mM NaN3, 100 mM Na-citrate, pH5.5), heated at 100°C for 5 mm, centrifuged to pellet the cells and thesupernatants transferred to a fresh tube. Each supernatant was divided into twotubes containing 20 p.1 distilled water (dH2O). The material was then digested ormock-digested for 18 h at 37°C by the addition of 1 mU of endoglycosidase H(endo H) or 1 p.1 dH2O. Samples were subjected to polyacrylamide gelelectrophoresis and autoradiography as described above.Western blotsCellular extracts were subjected to SDS polyacrylamide gel electrophoresisand electroblotted to nitrocellulose membranes as described by Towbin andothers (Towbin et al. 1979). Proteins were stained using monoclonal antibodiesand an alkaline phosphatase-based detection kit from Gibco/BRL, Ontario,Canada.30Fractionation of microsomesThis procedure was adapted from Saraste and others (Saraste et al. 1986).Briefly, monolayers of infected or uninfected cells were harvested by incubationwith trypsin and the trypsin was quenched by the addition of cold DMEMcontaining 10% FBS. All further steps were carried out in the cold. Cells werepelleted by centrifugation for 10 mm, washed with DMEM containing 10% FBS,followed by a wash with homogenization buffer (0.25 M sucrose, 10 mM Tris, pH7.4, 10 mM KC1, 1.5 mM MgC12). Cells were then pelleted, resuspended in 4pellet volumes of homogenization buffer and homogenized in a Douncehomogenizer. Homogenization was monitored by phase contrast microscopywith the aim of maximizing cell breakage and minimizing nuclear disruption.Nuclei and cell debris were removed from the homogenate by centrifugation(600 Xg, 10 mm, 4CC). The pellet was then washed with homogenization bufferand the supernatants combined. The supernatant was centrifuged at 10,000 Xgfor 20 mm to remove mitochondria. The post mitochondrial supernatant wasthen decanted and layered on top of 5 ml of 0.33 M sucrose which had beenlayered over Imi of 2 M sucrose. This was centrifuged at 25 K rpm (107, 000 Xg)in a SW4I rotor for 1 h. Approximately 600 p.1 of the turbid band at the 2 M /0.33M interface was removed with a syringe. This material was suitable for furtherpurification of Golgi membranes. This total microsome sample (600 p.1) wasmade 50% in sucrose with the addition of 2.4 ml of 2 M sucrose and was addedto a SW41 tube. The following sucrose solutions (w/w) were layered overtop: 1ml of 45%; 1.5 ml each of 40%, 35%, 30%, 25%, and 2 ml of 20%. The stepgradients were then centrifuged at 170,000 Xg for 19 h at 4°C. Followingcentrifugation, 0.5 ml fractions were collected from the bottom of the tube. 150p.1 of each fraction was diluted with 0.5 ml dH2O and centrifuged at 100 K rpm(436,000 Xg) for 10 mm in a TLA 100.2 rotor. The supernatants were discarded31and the pellets resuspended in 30 jil of SDS sample buffer, boiled for 5 mm, andsubjected to polyacrylamide gel electrophoresis. Polypeptides were transferred tonitrocellulose by western blotting and specific polypeptides were detected usingmonoclonal antibodies as described above. The density of the fractions rangedfrom 1.08 to 1.24. The location of the intermediate compartment between theER and the Golgi was determined by western blotting using an anti-p58monoclonal antibody. p58 served as a useful marker for the location of thiscompartment. This fractionation procedure yielded a protein peak in fraction 5and a peak of the ER marker NADPH cytochrome C reductase in fraction 4.Indirect immunofluorescenceCells were grown on glass coverslips for 3 days and infected with HSV-1(MOI=5). At 13 HPI, cells were rinsed with PBS, fixed with 2% formaldehyde inPBS. Cells were permeabilized in PBS! 1% BSA (PBS-BSA) and 0.2 % Triton X100 for 3 mm. Permeabilization was stopped by rinsing the cells in PBS-BSA.The cells were then incubated with a 1 / 100 dilution of anti-gD monoclonalantibody in PBS-BSA. Monolayers were washed extensively and incubated with1/100 dilution of rhodamine-conjugated goat anti-mouse IgG in PBS-BSA for 30mm. Coverslips were then washed and mounted in 50% glycerol, 100 mM Tris,pH. 7.8. Images were photographed using a Zeiss microscope with epifluorescence optics. Confocal images were captured using a Bio-Rad MRC-500 confocalfluorescence microscope.Fluorescence microscopy using lectinsCells were grown on glass coverslips in 35 mm plastic dishes for 3 daysprior to fixation. For fixation, cells were rinsed with PBS, and incubated in 2%formaldehyde in PBS for 10 mm. Fixation was stopped by rinsing cells in PBS.Cells were permeabilized in PBS-BSA and 0.2 % Triton X-100 for 3 mm.32Permeabilization was stopped by rinsing the cells in PBS-BSA. Cells were thenincubated for 30 mm in 20 p.g/ml FITC-conjugated ricin, wheat germ agglutinin(WGA) and RCA120 in separate incubations. Monolayers were then rinsed inPBS-BSA to remove unbound molecules. After rinsing with PBS-BSA, thecoverslips were mounted in 50% glycerol (w/v) in PBS and photographed usinga Zeiss Axiophot microscope with epifluorescence optics.Electron microscopyCells were grown on Millicell HA inserts (Millipore Corporation) for 24 hprior to infection with HSV-1 (MOI5). At 18 HPI, cells were rinsed with PBSand monolayers were fixed in 2.5% glutaraldehyde in 0.1 M Na-cacodylate, pH7.3 for 1 h on ice. Cells were then rinsed and post-fixed in 1% 0504 for 1 h.These samples were rinsed, dehydrated and embedded in plastic. Specimenswere sectioned, stained and photographed using a Zeiss EM1OC transmissionelectron microscope.Analysis of HSV-1 gD in BFA treated cellsThis procedure was adapted from Saraste and others (Saraste et al. 1986).Briefly, monolayers of BFA-treated and untreated infected cells were harvestedby trypsinization, and washed extensively with cold DMEM/10% FBS. Allfurther steps were carried out in the cold. Microsomes were prepared andsubfractionated as described above. Following centrifugation, 1 ml fractions werecollected from the bottom of the tube. 0.3 ml of each fraction was diluted with0.5 ml dH2Oand centrifuged at 436,000 Xg for 10 mm in a TLA 100.2 rotor. Thesupernatants were discarded and the pellets containing the membranes weresuspended in 40 l endoglycosidase H buffer (50 mM sodium citrate (pH 5.5),0.5% SDS, 0.25% 3-mercaptoethanol, 1 mM NaN3), heated at 100°C for 5 mm,and centrifuged. Each supernatant was divided into two tubes containing 20 p.133dH2O. The material was then digested or mock-digested for 18 h at 37°C by theaddition of 1 mU of endo H or 1 of dH2O. Samples were subjected topolyacrylamide gel electrophoresis and HSV-1 gD was detected on Westernimmunoblots.Analysis of glycosaminoglycansBiochemical labeling of glycosaminoglycans was performed by amodification of procedures described previously by Bame and Esko (Bame andEsko 1989). Briefly, glycosaminoglycans were radiolabeled by incubating cells for3 days with 10 Ci/ml[35S]sulfate (carrier free) and 20 p.Ci/ml D-[6-Hlgiucosamine (38.8 Ci/mmol) in DMEM-FBS modified to contain 10 mM Na-sulfate and 1 mM glucose. The cells were washed three times with cold PBS andsolubilized with 1 ml of 0.1 M NaOH at 25°C for 15 mm. Samples wereremoved for protein determination. Extracts were adjusted to pH 5.5 by theaddition of concentrated acetic acid, and treated with 2 mg/mi pronase in 0.32 MNaCl-40 mM sodium acetate, pH 5.5, containing shark cartilage chondroitinsulfate (2 mg/mi) as carrier, at 40°C for 12 h. For some experiments portions ofthe radioactive material were treated for 12 h at 40°C with 10 mU ofchondroitinase ABC lyase, 0.5 U of heparitinase, or hyaluronidase. Theradioactive products were quantified by chromatography on DEAE-Sephacel bybinding in 100 mM NaCl, followed by elution with 0.7 M NaCl, or bycetylpyridinium chloride precipitation (Wasteson et al. 1973).For I-IPLC analysis, the glycosaminoglycan samples were de-salted byprecipitation with ethanol (Bame and Esko 1989). Following centrifugation, theethanol precipitates were suspended in 300 20 mM Tris, pH 7.4 and 50 il wasresolved by anion exchange HPLC using a 15 x 75 mm TSK DEAE-3SW column.Proteoglycans were eluted from the column using a 77 ml linear 50 mM to 700mM NaC1 gradient formed in 10 mM KH2PO4 (pH 6.0). All buffers contained340.2% Zwittergent 3-12 to extend the life of the column, as reported previously(Bame and Esko 1989). The glycosaminoglycans in the peaks were identified bydigestion of the sample with the relevant enzymes prior to chromatography.Enzyme and protein assaysNADPH cytochrome C reductase, an ER associated membrane proteinwhich mediates the reduction of cytochrome C, was measured by the followingspectrophotometric assay. Cell fractions (50tl) on ice were added to a freshlymade reaction cocktail (410 ii 0.1 M potassium phosphate buffer pH 7.5, 20 jil20% Triton X-100, 30 p1 5 mg/ml cytochrome C, 12 il 50 mM KCN) in aspectrophotometer cuvette. To start the reaction 120 p1 of 0.5 mM NADPH wasadded, the solution mixed and the increase in absorbance at 550 nm measuredover time in a Hitachi spectrophotometer (Omura and Takesue 1970). Theincrease in absorbance at 550 nm reflects the conversion of NADPH to NADP bythe enzyme.Sialyltransferase, a trans Golgi and trans Golgi network resident enzyme,was measured as described previously by Brändli and others (Brändli et al. 1988).UDPGa1-4-epimerase activity, and the conversion of UDPGa1NAc toUDPG1cNAc were measured exactly as described by Kingsley and others(Kingsley et al. 1986). Separation of nucleotide sugars on PET cellulose wasperformed as described (Kingsley et al. 1986; Randerath and Randerath 1965).Protein concentrations were determined using a Bio-Rad protein assayreagent.Purification of Golgi membranes for nucleotide sugar translocation assaysGolgi membranes were purified from L and gro29 cells according to themethod of Balch and others (Balch et al. 1984). All solutions and equipmentused were kept as cold as possible throughout the procedure. Briefly, 50 150 mm35dishes of L cells and 70 150 mm dishes of gro29 cells were grown tosubconfluency, harvested by incubation with trypsin, collected by low speedcentrifugation (600 Xg, 10 mm, 4°C), washed two times with cold mediumcontaining 10% FBS and once in homogenization buffer (250 mM sucrose, 10mM Tris-HC1 pH 7.4) prior to resuspension of the cells in five pellet volumes ofhomogenization buffer, to yield a 20% suspension. Cell suspensions werehomogenized on ice using a Dounce homogenizer with a tight fitting pestle.Homogenization was monitored by phase contrast microscopy with the aim ofmaximizing cell breakage with minimal nuclear disruption. 6 ml portions ofcrude cell extract were adjusted to 1.4 M sucrose by the addition of a equalvolume of cold 2.3 M sucrose, 10 mM Tris-HC1 pH 7.4, made 1 mMNa2EDTAfrom a 100 mM stock solution and loaded into a prechilled SW28 centrifugetube. The homogenate was gently overlayed with 14 ml of ice cold 1.2 M sucrose10 mM Tris-HC1 pH 7.4 followed by 8 ml of 0.8 M sucrose 10mM Tris-HC1 pH 7.4.The tubes were then centrifuged at 90 000 Xg for 2.5 h in a Beckman SW28 rotor.After centrifugation the white turbid band at the 1.2 M /0.8 M sucrose interfacewas removed with syringe in a minimal volume; this fraction contained theGolgi apparatus. Aliquots of this material were frozen inN2(l) and stored at -70°C. Using sialyltransferase as a Golgi marker activity it was determined that thesefractions were enriched approximately 10 fold in this activity over crude cellextract for both L and gro29 Golgi membrane fractions.Nucleotide sugar translocation assaysThe transport of nucleotide sugars was measured by a modification of aprocedure described previously (Brändli et al. 1988; Deutscher and Hirschberg1986).[‘4C]-acetate was used as a membrane impermeable standard in theseassays. Assays were performed in the presence of concentrations of nonradioactive nucleotide sugars varying from 0 to 25 p.M. 350 p.1 of assay buffer (25036mM sucrose, 150 mM KC1, 1 mM MgC12,10 mM Tris pH 7.5) containing 0.1 p.Cieach of[3H]-UDPGa1NAc (8.3 Ci/mmol)(or[3H]-CMPSA (25.1 Ci/mmol) and 0.1j.tCi [14C1 acetate (1.9 mCi/mmol) in a Beckman TLA 45 centrifuge tube waspreincubated for 10 mm at 30°C. The reaction was started by the addition of 50ill of Golgi membranes (17-25 .tg protein). The reaction mixture was incubatedfor a further 10 mm at 30°C. The reaction was terminated by the addition of 600p.1 ice cold assay buffer and immediate placement on ice. The membranes werethen pelleted by centrifugation in a TLA 45 rotor at 100 000 Xg for 20 mm at 4°C.The surface of the pellets were rinsed once with 1 ml of ice cold assay buffer andrecentrifuged to concentrate the membranes. The supernatant was decanted and500 p.1 of lysis buffer (2% SDS, 5 mM EDTA, 500 mM Tris pH 8.8) was added tothe pellets and the solubilized material analyzed by liquid scintillationspectroscopy. The relative amounts of solutes nonspecifically associated withthe pelleted Golgi membranes was corrected for by determining the amount ofradiolabeled[14C]-acetate associated with the pelleted material.The time course assays were performed in a similar fashion except thatthe concentration of non-radioactive nucleotide sugar used was 0.5 p.M and thetime of incubation was varied from 1- 20 mm.Glucose and sulfate uptake assays35 mm dishes of containing 5 X iO L and gro29 cells were incubated at37°C with medium containing 20 p.Ci[3H1-glucosamine/ ml of low sulfate/lowglucose DMEM containing 4% dialyzed FBS (dialyzed against PBS) for timesranging from 1-60 mm. At the end of the labeling period dishes were rinsed 3 Xwith PBS prior to lysis of the cells in 700 p.1 0.1 N NaOH. Samples of extract wereanalyzed for protein concentration and radioactivity. To determine the kineticsof glucose uptake in L and gro29 cells the same procedure described above wasemployed with the exception that the incubation time was 60 minutes (whichwas determined to be in the linear range for glucosamine uptake over time) andthe incubation medium contained varying concentrations of glucose from 0-300i.tM. Because glucosamine and glucose use the same translocator it was assumedthat the translocator had similar affinity for both glucose and glucosamine.Therefore these assays should reflect the relative ability of L and gro29 cells totranslocate glucose.To determine the rate of sulfate uptake in L and gro29 cells the proceduredescribed above was used with the exception that 10 p.Ci/ml[355]-sulfate wasincluded in the incubation medium.3738Chapter 1- Analysis of HSV-1 infected gro29 cells.In a previous study the mouse L cell mutant gro29 was isolated on thebasis of the ability to survive exposure to HSV-1 (Tufaro et al. 1987). This cellline was shown to be defective in the propagation of both VSV and HSV-1.Initial characterization of this cell line revealed that gro29 cells were defective inthe transport and processing of newly made G protein, the envelopeglycoprotein of VSV. Despite this defect the release of infectious VSV fromthese cells was diminished only three-fold when compared to the parental Lcells, suggesting that the secretory defect in these cells was not critical to thematuration and egress of VSV from the plasma membrane. However, the effectof this lesion on the release of infectious HSV-1 from gro29 cells was verydifferent. Although gro29 cells were infected efficiently and the replication cycleproceeded to the late stages of viral gene expression, the spread of HSV-1 fromcell to cell did not occur and underprocessed glycoproteins accumulated insidethe gro29 cell late in infection. To characterize the nature of the block to HSV-1infection in gro29 cells the following study was performed.ResultsProduction of infectious virus by gro29 cells.It has been shown previously by indirect immunofluorescence thatplaques do not form on monolayers of gro29 cells infected with HSV-1, despitethe fact that infection occurs normally and progresses to the late stages of viralgene expression (Tufaro et al. 1987). To quantify the block to viral propagationin gro29 cells, the titres of the extracellular virus produced by infected parental Lcells and gro29 cells were determined. The medium on each monolayer wasreplaced every two HPI to minimize the superinfection of cells during the39experiment. The titres were determined for each time point and then addedtogether to reflect the total number of PFLJ released from the cells during the 20h of sampling (Fig. 3). These analyses revealed that the parental L cells released200 PFU/cell while gro29 cells released 0.1 PFU/cell. There are severalpossibilities that could account for this deficiency. If viral assembly or egress wasdefective, the release of virions from gro29 cells would be impaired. If theassembly and egress of the virions was normal, it may be that the specific infectivity of the virus was diminished due to a structural defect caused by a lesion inthe gro29 cells.It has already been established that gro29 cells have suffered a defect inglycoprotein transport and processing that leads to the accumulation ofimmature forms of HSV-1 glycoproteins in HSV-1 infected cells (Tufaro et al.1987). Normal synthesis and processing of viral glycoproteins is very importantfor the propagation of HSV-1 (reviewed in Spear 1985); because of this,experiments were carried out to identify the defects in glycoprotein processingand to characterize the phenomena that contribute to the low yield of infectiousvirus from the gro29 cell line.Pulse-chase analysis of HSV-1 glycoprotein processing.HSV—1 specifies at least eleven membrane glycoproteins: gB, gC, gD, gE,gG, gH, gI, gj, gK, gL and gM (Ackermann et al. 1986b; Baines and Roizman 1993;Buckmaster et al. 1984; Heine et al. 1972; Hutchinson et al. 1992a; Hutchinsonet al. 1992b; Longnecker et al. 1987; Spear 1976; Spear 1985). gD and gB wereinvestigated in these experiments. Glycoprotein D contains three sites for Nlinked oligosaccharides (Cohen et al. 1983) and two 0-linked chains (SerafiniCessi and Campadelli-Fiume 1981), whereas gB has nine potential N-linked sites(Bzik et al. 1984a) and an uncharacterized number of 0-linked chains. TheFig. 3. Titre of virus released from HSV-1 infected L and gro29 cells.Monolayers of L and gro29 cells were infected with HSV-1 (MOI=1O). At the indicatedtimes post-infection, the medium was removed and the numbers of infectious particlesin the medium were determined by plaque assay on Vero cells as described in Materialsand Methods. The results represent the cumulative release of PFU by L and gro29 cellsover time Published in Banfield and Tufaro 1990).40!dLI0I.0eoI(‘II7uI.0.,,CCDtO1.i•Cl)(‘ICD8’0.601.011742precursor forms of these proteins (pgB and pgD) contain high-mannoseoligosaccharides sensitive to endoglycosidase H (endo H) which becomeresistant to endo H as the oligosaccharide moieties are processed to morecomplex forms (Cohen et al. 1983; Johnson and Spear 1982; Mafthews et al.1983; Serafini-Cessi et al. 1988; Wenske et aL 1982).To investigate the synthesis and processing of these proteins, pulse-chaseanalyses of the two glycoproteins in HSV-1 infected cells were performed at 5HPI (Fig. 4A). When HSV-1 infected parental L cells and gro29 mutant cellswere labeled with[35S]-methionine for 10 mm, the accumulation of label into gBand gD were similar for both cell lines. The major bands representing thenewly-synthesized gD in L and gro29 cells had the same relative mobilities inthe gel, indicating that the co- and post-translational modifications, such as theaddition of N-linked oligosaccharides, occurred normally in both cell lines.Similarly, the synthesis and processing of gB in the two cell lines wereindistinguishable at the end of the labeling period. After a 90 mm chase in theparental L cells (Fig. 4A), both gB and gD migrated more slowly in the gels,suggesting that processing of the high mannose oligosaccharides to complexforms, as well as modification by 0-linked glycosylation had occurred. Incontrast, most of the gB and gD polypeptides synthesized in gro29 cells did notincrease in size during the chase. There was no reduction in the amount oflabeled polypeptides in the cells during the chase period, indicating that the bulkof newly-made material persisted in forms that were incompletely or aberrantlyprocessed. The increase in radioactivity observed in several chase samples (Lchase, gB for example) may be due to more efficient immunoprecipitation of themore highly processed species.Several explanations could account for the failure to modify newly-madeglycoproteins in this cell line. The underprocessed species in the mutant gro2943Fig. 4. Processing of HSV-1 gD and gB in L and gro29 cells.Monolayers of L and gro29 cells were infected with HSV-1 (MOI=10). At 5 HPI, cellswere incubated with[35S1-methionine for 20 mm, rinsed and chased with unlabeledmethionine for 90 mm. HSV-1 gD and gB were immunoprecipitated from total cellextracts using monodonal antibodies and subjected to SDS -PAGE. (A)Autoradiogram showing HSV-1 gD and gB at the end of the labeling period (P) or aftera 90 mm chase (C) in L cells (L) and gro29 cells (g29) as indicated. The positions ofimmature forms of gB and gD (pgB, pgD) and mature forms (gB, gD) are indicated tothe left of the Fig. (B, top panel) Autoradiogram showing pulse-chase analysis ofHSV-1 gD. The positions of pgD and gD and the duration of chase (mm or h) areindicated. (B, bottom panel) Histogram showing the percent of HSV-1 gD (B, toppanel) that was resistant to digestion by endo H. Samples from pulse-chaseexperiments were digested with endo H and subjected to SDS -PAGE. The relativeamounts of endo H resistant gD was determined by densitometric scanning ofautoradiograms (not shown). was taken as the amount of . Solid bars, L cells; Graybars, gro29 cells. Published in Banfield and Tufaro 1990.BAgD gBL929 L9290 40L 29 L g29gDp9Db- —— —C0t0C‘UCgD gBL29 Lg2990 3hrLg29 L9295 hrL 29pgBOpgD0 30040 90 180Duration of chase (mm)45cells may represent glycoproteins that were retained in the ER or Golgi.Alternatively, the newly-synthesized polypeptides may have traversed the Golgiwithout being processed. This could arise if a component of the ER- or Golgiresident processing machinery was defective in these cells. The followingexperiments were performed to characterize further the glycoprotein fluxthrough the Golgi complex. Pulse-chase analyses of gD were repeated using alonger chase, and the rates at which newly-made glycoproteins became endo H-resistant were determined (Fig. 4B). Because endo H cleaves high-mannosechains but not complex oligosaccharides, it serves as a useful probe for theprocessing of N-linked glycoproteins (Kornfeld and Kornfeld 1985).Glycoprotein D was chosen for further analysis because all mature forms of thisprotein are endo H-resistant, whereas at least some of the high mannose N-linked oligosaccharides attached to gB remain in a sensitive form (Johnson andSpear 1982; Wenske et al. 1982). It can be seen in Fig. 4B that newly-made gDwas processed rapidly in L cells; by 40 mm post-labeling, greater than 95% of thepolypeptides were modified to a higher apparent molecular weight, and by 90mm, processing was complete. In contrast, gD synthesized in gro29 cells failed tobecome fully processed even after a 5 h chase. By 40 mm, two or three discretebands were detected that were larger than pgD, indicating that somemodifications had occurred, although a comparison of relative mobilitiessuggests that these were not the forms that accumulated in the parental L cells.As the chase proceeded (Fig. 4B, bottom), there was a rapid and completedisappearance of endo H-sensitive forms in the L cell samples, indicating thatmost of the mass of the newly-made protein was modified by the host cellprocessing machinery. In contrast, the gD synthesized in gro29 cells persisted asendo H-sensitive forms. Oligosaccharide processing was not blocked entirely,however, and 60% of the newly-synthesized gD acquired endo H-resistance46during the 5 h chase (Fig. 4B). These results indicated that the intracellulartransport of HSV-1 glycoproteins was slowed but not abolished in infected gro29cells. The large shift in molecular weight that accompanies the maturation ofgD in L cells was not observed in the gro29 cell samples. This is consistent withthe notion that these modifications occur relatively late in the secretorypathway, after the acquisition of endo H resistance.Analysis of gD transport in HSV-1 infected cells.The data obtained from the pulse-chase analysis suggested that thetransport of glycoproteins through the secretory pathway was abnormal. Toinvestigate the accumulation of glycoproteins in the organelles of the secretorypathway during HSV-1 infection, total microsomes were prepared from infectedcells at 13 HPI and were subjected to centrifugation on sucrose step gradients.Because the membranes of the ER and Golgi cisternae have different densities,fractions enriched for different organelles and for different Golgi cisternae can beisolated by this technique. Samples of each fraction were subjected to SDS-PAGEand blotted onto nitrocellulose membranes. These gradients included allmembranes in the microsomal fraction which separated at densities from 1.24(fraction 1) to 1.08 (fraction 15). Furthermore, the location of the intermediatecompartment that exists between the ER and Golgi complex was determinedusing an antibody to the protein p58, which has been shown to be located in thiscompartment and in cis Golgi elements (Jantti and Kuismanen 1993; Saraste etal. 1987). In gradients similar to those shown, the highest concentration ofprotein was in fraction 5, and the peak of the ER enzyme NADPH cytochrome Creductase was in fraction 4 (data not shown).HSV-1 gD was analyzed using a monoclonal antibody and an alkalinephosphatase-based detection system. In the L cell samples (Fig. 5A), the majority47Fig. 5. Western blot analysis of HSV-1 gD in membrane fractions of L and gro29 cells.Parental L and gro29 cells were infected with HSV-1 (MOI=1O) and incubated for 13 h.Cell monolayers were harvested and the total microsome fraction was prepared asdescribed (Materials and Methods). Total microsomes were subjected to centrifugationin sucrose step gradients to fractionate membrane components of different densities.Fractions were solubilized and subjected to SDS-PAGE followed by electroblotting ontoa nitrocellulose membrane. The high density membranes characteristic of the ER are inthe lower numbered fractions, and the low density membranes characteristic of theGolgi complex are in the higher numbered fractions. The highest concentration ofprotein was in fraction 4, and the peak of the ER enzyme NADPH cytochrome Creductase was shown to be in fraction 5 in gradients similar to those shown. The HSV-1gD present in these fractions was detected using an anti-gD monoclonal antibody andan alkaline phosphatase staining procedure. Western blots showing gD in themembranes of (A) L cells and (B) gro29 cells are shown. The arrowhead to the left of thelanes denotes a background band that was also present in uninfected cell samples (datanot shown). The position of pgD is indicated to the left of the figure and the samplecontaining predominantly mature gD (fraction 14) is identified above the lane. Theconcentration of p58, a protein that has been shown to reside in the intermediatecompartment between the ER and Golgi complex, was determined on separate blotsusing an anti-p58 antibody (data not shown). p58 was enriched in fractions 4-6, with apeak in fraction 5. Published in Banfield and Tufaro 1990).I•__466906d4cLELLL6Lc£V(8cd)4Sb49of pgD was contained in four fractions (4-7) and decreased in concentrationtowards the top of the gradient (right). Mature forms of gD were detectable byfraction 4 and were the only forms detectable in fraction 14. The distribution ofgD in gro29 cell fractions was strikingly different (Fig. 5B). Whereas there was apeak of pgD in fractions 4 and 5, little if any mature gD fractionated in the gro29cell gradient. It was evident that gro29 fractions contained less gD overall thandid L cell fractions, although the differences were not quantified in theseexperiments. When these same blots were probed for the presence of the Golgimarker p58, a protein that is enriched in the intermediate compartment between the ER and Golgi complex (Jantti and Kuismanen 1993; Saraste et al.1987), p58 was equally abundant in the gro29 cell and L cell fractions, suggestingthat the underrepresentation of gD in the gro29 cell fractions was not an artifactof the isolation procedure. This suggested that gD was underrepresented in thecellular organelles contained in these gradients. Because it has been shown thatthe rate of synthesis of gD was normal in these cells (Fig. 4), it may be that thedefect in gro29 cells caused gD to localize into abnormal cellular structures thatwere lost during this isolation procedure.Analysis of virus egress.Based on the results demonstrating an inhibition of the transport andprocessing of HSV-1 glycoproteins, and the under representation of gD in thesecretory organelles, it was reasoned that the ability of HSV-1 particles to betransported out of the cell might also be impeded. This follows from a currentmodel of HSV-1 egress in which newly-formed virions are thought to traversethe Golgi complex en route out of the cell. To investigate this, parental L cellsand gro29 cells were infected with HSV-1 and labeled with[35S]-methionine for18 h. Extracellular virions were harvested from the medium by centrifugationand subjected to fractionation on Dextran T1O gradients. The TCA-insoluble50radioactivity was measured (Fig. 6A) and the titre of infectious particles wasdetermined for each fraction (Fig. 6B). It can be seen in Fig. 6A that L cellsreleased a large amount of virus which peaked in fraction 17. In contrast, therewas only a small peak of material detected in the gro29 cell medium (0.2% of Lcell), indicating that virus production was impeded. An analysis of the PFUpresent in each fraction revealed that the peak of radioactivity (Fig. 6A, fractions15, 16 and 17) corresponded to the peak of infectious virus (Fig. 6B). Fractions15-17, comprising the peak in each gradient, were pooled and subjected to SDSPAGE (Fig. 6C). A limited number of radioactive bands were detected, consistentwith these samples being highly enriched for virions. No radioactive polypeptides were detected in the gro29 samples during the exposure times used. Thesedata support the conclusion that gro29 cells are unable to release virions afterinfection.Analysis of cell-associated virus.One way to account for the paucity of extracellular virions is if HSV-1particles were unable to assemble in the mutant gro29 host cells. To investigatethis, nucleocapsids were isolated from total cells and fractionated on sucrosegradients. Fractions were analyzed for TCA-insoluble radioactivity to detect thepeak of nucleocapsids (Fig. 7A) and were subjected to SDS-PAGE to detect nudeocapsid proteins (Fig. 7B,C). In Fig. 7A, a small peak of nucleocapsids wasevident in fractions 21-28 for both L and gro29 cells. There were fewernucleocapsids isolated from gro29 cells, although the reduction was minorwhen compared with the reduction in virion release. Polyacrylamide gelanalysis of the gradient fractions (Fig. 7C) revealed that the patterns of newlymade polypeptides in the nucleocapsid fractions isolated from parental L cellsand gro29 cells were similar. The polypeptides denoted with asterisks were51Fig. 6. Comparison of HSV-1 particles released from L and gro29 cells.Parental L cells and gro29 cells were infected with HSV-1 (MOl =10). At two HPI, themedium was removed and replaced with labeling medium containing[35S]-methionine.At 18 HPI, virions released into the medium were harvested by centrifugation,resuspended in buffer and centrifuged through a 5-40% dextran T-10 gradient asdescribed (Materials and Methods). The gradients were fractionated and samples wereretained for further analysis. Samples from the top of the gradient are shown to the left.(A) Total TCA-insoluble radioactivity in each sample. CPM; CPM x io. (B) numberof PFU in the fractions indicated. PFU; PFU x iO. (C) Autoradiogram showingradioactive polypeptides present in the three peak fractions (16,17, and 18) from eachgradient. To prepare the samples, the fractions were pooled, pelleted by centrifugation,solubiized and subjected to SDS-PAGE. The polypeptides in the gel were electroblotted to nitrocellulose membranes and viral proteins were identified using an anti-gDmonoclonal antibody and an alkaline phosphatase detection system (data not shown).The blot was then exposed to film for 3 days to detect radioactive polypeptides in thepellets (shown in panel C). No signal was detectable for the gro29 sample. Thepositions of gD, nucleocapsid proteins (asterisks), and ovalbumin (45Kd) and carbonicanhydrase (29Kd) markers are shown. Titres of fractions were determined by H.Meadows. Published in Banfield and Tufaro 1990.c2.I g299D45... —1614120— 108064Fraction numberBc) 40000.200100F2 4 6 8 10 12 14 16 17 18 19 20Fraction numberC2953Fig. 7. Nucleocapsid assembly in L and gro29 cells.Parental L cells and gro29 cells infected with HSV-1 (MOI=10) were labeled with 50j.iCi/ml[35S1-methionine from 2-18 HPI. Cell monolayers were harvested and proteinextracts were prepared with Nonidet P40-Sodium deoxycholic acid extraction buffer.The extracts were sonicated, 0.5 M Urea was added and insoluble material was pelletedfrom the sample by low-speed centrifugation. The resulting extracts were layered ontoa 1040% sucrose gradient and centrifuged. The gradients were fractionated into 35samples and analyzed. (A) Radioactivity in TCA-precipitable material for each fraction.The bar over fractions 22-29 represents the peak of nucleocapsids. Open circles; L cells,closed circles; gro29 cells. (B,C) Autoradiograms showing radiolabeled polypeptidespresent in every other fraction. L cells (B) and gro29 cells (C). Two nucleocapsidproteins that fractionate with nucleocapsids and whole virions are indicated (asterisk).Published in Banfield and Tufaro 1990.AFracton# 1(1/A5 10 15 20 25 30 35Fraction number--4 45 9 13 17 21 25 29 33:I.- 1a — ———— —I..’ — -55specific to infected cells and fractionated consistently with nucleocapsids andvirions. By comparing these gels to those published previously (Rixon et al.1990), it was determined that the upper and lower bands represent VP21 andVP22a. The material sedimenting in gro29 fraction 35 represents urea-insolublematerial that was not efficiently removed in the step prior to centrifugation.Although the nucleocapsid fractionation procedure gave an accurateassessment of the relative amounts of nucleocapsids present in the cells, it wasof some interest to determine the number of intracellular particles that werealso infectious. Cells were lysed in the absence of detergent, nuclei wereremoved, and extracts were fractionated on Dextran T10 gradients. Total TCAinsoluble radioactivity in each fraction was determined, as was the number ofinfectious particles in selected fractions (Fig. 8A,B). In addition, samples weresubjected to electrophoresis in polyacrylamide gels to detect the polypeptidessedimenting in the gradient (Fig. 8C). In L cells (Fig. 8B), fraction 12 contained60% of the infectious particles, and 85% of the infectious particles were infractions 10-14. gD was detectable in these fractions, as were the twonucleocapsid proteins (asterisks). gro29 cell extracts fractionated differently,however. Whereas the peak of infectious particles was in the same fraction asin the L cell gradient (fraction 12), the peak of radioactivity was at fraction 14 inthe gro29 cells gradient (Fig. 8A). Analysis of the polypeptides in these fractions(Fig. 8C) revealed that a low number of polypeptides sedimented in the gradient,as expected for fractions enriched in virions. An analysis of the polypeptides inthe L and gro29 cells revealed several differences. There were alterations in thepolypeptides in the gB/gC region of the gel, which were not analyzed further.The gro29 fractions were enriched in immature forms of gD (pgD), whereas Lcell fractions contained mostly mature gD. The identities of these polypeptideswere confirmed by immunoprecipitation and western blots (data not shown).56Fig. 8. Detection of intracellular virions.Parental L cells and gro29 cells were infected with HSV-1 (MOI=10) and incubated for18 h in the presence of 50 i.tCi/ml[35S]-methionine. Cell monolayers were harvestedand whole cell extracts were prepared by gentle homogenization in the absence ofdetergents. Cell debris and nuclei were pelleted from the extracts, which were thencentrifuged for I h at 22,000 RPM through a 5-40% Dextran T10 gradient. The gradientswere fractionated and samples were retained for further analysis. (A) Radioactivity inTCA-precipitable material. The arrow denotes the peak of infectious particles. CPM=counts per mm. x iO Closed circles=L cells; Open circles=gro29 cells.. (B) Thenumber of PFU in each fraction was determined by limiting dilution analysis andplotted as the percent of total infectious particles in the sample applied to the gradient.Solid=L cells. Striped=gro29 cells. (C) A sample of every other fraction was subjectedto SDS polyacrylamide gel electrophoresis and the radioactive polypeptides weredetected by fluorography. The L cell fractions are shown in the left lanes (fractions 2-18); the gro29 fractions are shown in the right lanes (fractions 2-18). The positions of gDand pgD, as determined by immunoprecipitation and western blotting (not shown), areindicated. The relative positions of two nucleocapsid proteins are indicated (asterisk).Titres of fractions determined by H. Meadows. Published in Banfield and Tufaro 1990.C 2 1o!14 18*:j119Dc760504030C.)A201000 2 4 6 8 10 12 14 16 18 20Fraction numbera.002 4 6 8 10 12 14 16 18 20Fraction number6 10V14 18 2 658Furthermore, most of the mass of radiolabeled gD was associated with thevirion-enriched fractions, suggesting that a large portion of the intracellular gDwas embedded in viral envelopes. It is clear from these results that newly-formed particles were present in the cytoplasm of infected cells, and it appearedthat they were enveloped inasmuch as viral glycoproteins fractionated with thepeak of infectivity. The apparent differences in the sedimentation observedbetween L and gro29 cells were not investigated further.Although this procedure for isolating intracellular viral particles resultsin contamination of the sedimenting material with cellular polypeptides, it isuseful for determining the infectivity of the intracellular particles. Todetermine the number of PFU in each fraction, samples of each gradient fractionwere diluted in growth medium and used to inoculate monolayers of Vero cells.L cell samples contained 270 PFU/cell, whereas gro29 cell samples contained 1.5PFLT/cell. These data and the data regarding total intracellular particles (Fig. 8)suggest that the lesion in gro29 cells inhibits or destroys the infectivity of theparticles that are formed inside the cells, resulting in a decreased specificinfectivity for the intracellular virions.Immunofluorescence and electron microscopic analysis of infected cells.Based on the observations described above, the intracellular location ofvirions and glycoproteins was investigated. Cells were grown on coverslips,infected with HSV-1 and prepared for indirect immunofluorescence at 13 HPI.An anti-gD monoclonal antibody was used to detect gD on the cell surface andinside the cells after permeabilization with detergent (Fig. 9). Diffusecytoplasmic staining was evident in infected L cells (Fig. 9A), with many discretepatches of gD scattered throughout the cytoplasm. These patches may representvirions in cytoplasmic vacuoles that were in transit to the cell surface prior to59Fig. 9. Immunofluorescence analysis of HSV-1 gD in HSV-1 infected L and gro29cells.Parental L and gro29 cells were grown on glass coverslips and infected with HSV-1(MOI=5). At 13 HPI, monolayers were fixed and prepared for immunofluorescence asdescribed (Materials and Methods). The distribution of HSV-1 gD in permeabilizedcells (A,B) or on the cell surface (C,D) of L cells (A,C) and gro29 cells (B,D) wasdetected by indirect immunofluorescence using an anti-gD monoclonal antibodyfollowed by a rhodamine-conjugated second antibody. Published in Banfield andTufaro 1990.61egress. In some cells, gD was concentrated in a juxtanuclear region which maycontain the Golgi complex. These observations are consistent with biochemicaldata demonstrating that gD traversed the Golgi en route out of the cell (Fig. 5),and with the accumulation of gD-containing intracellular particles in thecytoplasm (Fig. 8).The localization of gD in gro29 cells was very different from that of theparental L cells (Fig. 9B). In these cells, gD was concentrated in a juxtanuclearregion and there was very little of the diffuse cytoplasmic staining visible in Lcells. The majority of the cytoplasmic gD appeared to be in small roundcytoplasmic vacuoles. There was also some reticular cytoplasmic stainingevident, consistent with gD being present in the ER. This data suggested that theintracellular distribution of gD was perturbed in these cells.To determine whether there was a block to the appearance of viral glycoproteins on the surface of gro29 cells, cells were infected, fixed and preparedfor immunofluorescence without permeabilizing the cells. Comparisons of thecell surface staining (Fig. 9C,D) indicated that gD was transported to the cellsurface in both L cells and gro29 cells. To determine the relative abundance ofgD on the surface of gro29 cells compared with L cells, samples of infected cellswere analyzed by flow cytometry in a fluorescence-activated cell analyzer (F.Tufaro, data not shown). Monolayers of L and gro29 cells were infected withHSV-1 (MOI=1O) and harvested by rinsing in growth medium containing 10mM EDTA. Mock infected cells were also harvested for use as controls for nonspecific binding. Cells were rinsed and then stained with an anti-gDmonoclonal antibody followed by a fluorescein-conjugated second antibody.The analysis of io L cells revealed a single population of cells with an averagefluorescence 28-fold higher than the fluorescence in the uninfected cell controls.By contrast, gro29 cells exhibited a bimodal distribution with 80% of the cells62eliciting an average fluorescence intensity equivalent to 9% of the L cellpopulation. The fluorescence of the remaining 20% of the gro29 cells wasindistinguishable from the fluorescence intensity of L cells. Based on theseresults it was concluded that the flux of individual viral glycoproteins from theER to the cell surface was perturbed only slightly in the mutant cell line whencompared with the flux of viral particles out of the cells.To determine the intracellular location of virions, infected L cells andgro29 cells were examined by electron microscopy. In infected L cells, largecytoplasmic vacuoles containing many virions were observed (Fig. bA). Anenlargement of a single vacuole is shown in Fig. bC. These structures were notcommon in L cells and are shown to allow for comparison to the structures seenin gro29 cells. Infected gro29 cells contained numerous small vacuoles in thecytoplasm, many of which also contained enveloped viral particles (Fig. 1OB,D).Whereas the large vacuoles in L cells were filled with particles, the vacuoles ingro29 cells were irregular in shape and contained few virions, and freenucleocapsids were often located adjacent to these vacuoles (Fig. IOD).Fig. 1OE shows a confocal immunofluorescence micrograph of infectedgro29 cells at 13 HPI stained with anti-gD antibody and a rhodamine conjugatedsecond antibody. In this procedure, decoration of the nuclear membrane andendoplasmic reticulum was well resolved, and gD was clustered in discretelocations of the cell. It is likely that this intense staining derived from thevirion-containing vacuoles visible in the electron micrographs (Fig. 1OF). Thesevacuoles may represent the intermediate compartment between the ER andGolgi as the description of this structure (Saraste and Kuismanen, 1984) isconsistent with this hypothesis as is the presence of immature viralglycoprotein.63Fig. 10. High resolution microscopy of HSV-1 infected L and gro29 cells.Parental L and gro29 cells were grown on Millicell HA inserts (Millipore) for 24 h priorto infection with HSV-1 (MOI=5) or mock-infection. Following inoculation, freshDMEM containing 10% FBS was added and the infection was allowed to proceed for afurther 18 h. For electron microscopy, cell monolayers were rinsed, fixed inglutaraldehyde, embedded and sectioned. (A) HSV-1 infected L cells; (B) HSV-1infected gro 29 cells; (C, D) Higher magnification image of region 1 indicated in A andB, respectively. (E) Confocal image of HSV-1-infected gro29 cells prepared as in Fig. 5;(F) HSV-1 infected gro29 as in B prepared from a different sample showing HSV-1particles in a cytoplasmic location. Nu, nucleus; Cyt, cytoplasm; v,virus particle; NC,nucleocapsid. Electron microscopy performed by H. Meadows. Published in Banfieldand Tufaro 1990.65To confirm that the structures observed in the infected gro29 cellscontained both viral glycoprotein and virus nucleocapsids double fluorescencelabeling experiments were performed. Fig. 11 shows infected gro29 cells pulselabeled for 30 mm from 6.5 HPI to 7 HPI with 5-bromo-deoxyuridine (BrdU) andstained with anti-BrdU antibodies and FITC-conjugated ricin at 13 HPI. In thisdouble-label immunofluorescence experiment, gro29 cells showed a strongcytoplasmic BrdU signal (Fig. hA) that co-localized with the ricin-binding sites(Fig. 11B). These results indicated that the viral DNA in the gro29 cell cytoplasmwas associated with viral glycoproteins, and was likely in or near Golgi or Golgiderived membranes. The presence of a strong BrdU signal in the nucleus ofthese cells suggests that the newly-synthesized viral DNA can take longer than 6h to be transported out of this compartment.Brefeldin A treatment of HSV-1 infected cells.It is well established that BFA induces alterations in glycoproteinprocessing and transport in mammalian cells (Doms et al. 1989; Gamou andShimizu 1988; Kato et al. 1989; Kato et al. 1987; Lippincott-Schwartz et al. 1990;Misumi et al. 1986; Perkel et al. 1988; Perkel et al. 1989; Urbani and Simoni1990). It is also known that changes in glycosylation can have a profound effecton the propagation of HSV (Campadelli-Fiume et al. 1982; Johnson and Spear1982; Olofsson et al. 1988; Peake et al. 1982; Pizer et al. 1980; Serafini-Cessi et al.1983). Because gro29 modifies viral glycoproteins abberently it was of someinterest to examine the effect of BFA on HSV-1 infection to see if anyconclusions could be made regarding the two transport blockages.66Fig. 11. Immunofluorescence localization of cytoplasmic virions.Monolayers of cells growing on glass coverslips were infected with HSV-1 at anMOI=1O. Infected cells were incubated in medium containing BrdU from 6.5-7 HPI tolabel viral DNA. At 7 HPI the BrdU was removed and the cells received fresh medium.13 HPI infected monolayers were processed for indirect immunofluorescencemicroscopy as described in the materials and methods section. Intracellularglycoconjugates were detected using FITC conjugated ricin. BrdU labeled DNA wasdetected using a monoclonal antibody against BrdU followed by incubation with arhodamine conjugated secondary antibody. The figure shows the same cells labeledwith anti-BrdU antibody (A) and with ricin (B).68Brefeldin A alters the processing of HSV-1 glycoproteins in infected cells.To study the effects of BFA on viral glycoprotein processing in HSV-1 infected L cells, cytoplasmic membranes were isolated and the HSV-1 gD containedin those membranes was incubated with endo H, and analyzed byimmunoblotting. Monolayers of infected cells were treated with 3 j.ig/ml BFAfrom 2 HPI and harvested at 13 HPI. Total microsomes were isolated from thecells and centrifuged in discontinuous sucrose gradients to separate heavy fromlight membrane fractions. Samples of membranes were collected from thegradients and the proteins contained in the fractions were subjected to digestionwith endo H. Following SDS-PAGE of the digested and mock-digested samples,the polypeptides were transferred to nitrocellulose and HSV-1 gD was detectedusing an anti-gD monoclonal antibody. Fig. 12 shows the results of this assay fortwo fractions obtained from BFA-treated and untreated cells. These fractions (Aand B), contained most of the mass of protein loaded on the gradient and alsocontained the highest proportion of endo H-sensitive forms of HSV-1 gD. Thesefractions are the equivalent of fractions 4 and 5 in Fig. 5. In untreated cells (Fig12, NO BFA), most of the mass of pgD was sensitive to digestion byendoglycosidase H. There was also a substantial amount of material which wasresistant to endo H digestion (gD), representing polypeptides that had traversedthe medial Golgi cisternae en route to the plasma membrane. Fraction B, whichincluded membranes of lower density than those in fraction A, also contained amixture of endo H-resistant and sensitive forms of gD.In BFA-treated cells, by contrast, both fractions (A and B) containedpartially-processed forms of this glycoprotein, as judged by relative mobilities.The major species of gD exhibited a lower relative mobility than the endo H-sensitive forms accumulating in untreated cells. A small amount of thismaterial was endo H-sensitive, and the digested species had lower relative69Fig. 12. Effects of BFA on the processing of HSV-1 gD during HSV-1 infection in Lcells.L cells were infected with HSV-1 (MOI=1O), incubated without BFA (NO BFA) or with 3jig/mi BFA (+BFA), and were harvested at 13 HPI. Total microsomes were preparedand fractionated further into ER- and Golgi-enriched samples on sucrose gradients asdescribed (Materials and Methods). Fractions were collected from the bottom of thetube, and samples of each fraction were subjected to digestion with endo H and wereanalyzed by SDS-PAGE and western immuno-blotting. HSV-1 gD was detected using amonoclonal antibody and an avidin-biotin alkaline phosphatase detection system. Thetwo fractions shown contained most of the mass of the protein and had the highestactivity of the ER marker NADPH cytochrome C reductase (data not shown). Thesefractions are the equivalent of fractions 4 and 5 shown in Fig. 5. These fractions also hadthe largest proportion of endo H sensitive forms. The positions of several molecularweight markers representing 7lKd and 44Kd proteins are illustrated. The immature gD(pgD) and mature forms (gD) are indicated. The positions of the major endo H-sensitiveand -resistant forms are indicated by arrowheads. Published in Cheung et al. 1991.70NO BFA +BFASample AEndoH — + — +A B— + — +71k =pgDall7gDB44k71mobilities than the endo H-sensitive species detectable in untreated cells.Although the oligosaccharide structures in the fractions from BFA-treated cellswere not characterized further, the observation that gD was largely resistant toendo H digestion indicated that most of the N-linked oligosaccharides on theHSV-1 glycoproteins were processed to intermediate forms by 13 HPI. Theseresults could be explained if BFA induced the redistribution of Golgi processingenzymes into the ER of HSV-1 infected cells. This would explain the detectionof the intermediate forms of gD in these ER enriched fractions. Thisinterpretation is consistent with the BFA-induced perturbations that have beenshown to occur in uninfected cells as well (Doms et al. 1989; LippincottSchwartz et al. 1989).These data also suggested that the lesion affecting viral glycoproteinprocessing in the mutant gro29 cell was dissimilar to the effects seen when Lcells were treated with the drug BFA in that viral glycoprotein processing occursto a greater extent in the BFA treated cells. It was thus of some interest toexamine the effects of BFA on viral glycoprotein processing in the mutant gro29cell as these results may indicate whether the transport block observed in thegro29 cells is before or after the BFA effect.Two dishes of gro29 cells were infected with HSV-1 at an MOI=1O. At 11-IPI BFA was added to one of the dishes at a concentration of Cellswere harvested at 18 HPI. Fig.13 shows an immunoblot detecting glycoprotein Dfrom infected gro29 cell membrane fractions in the absence (-) or presence (+) ofBFA. In the untreated gro29 sample the immature form of gD, pgD,predominated. By contrast, in the BFA treated samples the majority of the gDwas in an intermediate form similar to that seen in the BFA treated L cellsample (Fig.12).Fig.13. Effect of BFA on the processing of HSV-1 gD in gro29 cells.Two dishes of gro29 cells were infected with HSV-1 at an MOI=1O. The dishes were thenincubated or mock incubated with 3 tg/ml BFA from I HPI to 18 HPI when cells wereharvested, homogenized and a membrane fraction prepared by differentialcentrifugation. Samples of the membrane preparations were subjected to SDS PAGE ona 10% gel. The protein in the gel was transferred to nitrocellulose and HSV-1 gDdetected using a monoclonal antibody and an alkaline phosphatase based detectionsystem.72+74Because viral glycoproteins are processed to a greater extent in the BFAtreated gro29 cells than in untreated gro29 cells, these results suggested that theblock to transport in gro29 cells occurred prior to the transport block caused byBFA. Moreover, these results indicated that the enzymes required for theacquisition of more highly processed forms of gD were present and functional inthe infected gro29 cell, and as such the lesion in these cells with respect toherpesvirus glycoprotein processing was not due to the lack of functionalglycosylation enzymes during herpesvirus infection. These data also indicatedthat the regions of the secretory apparatus normally affected by BFA weresusceptible to the effects of this drug in gro29 cells because of the increasedmodification of gD observed in BFA treated gro29 cells. It is likely that thefailure to glycosylate HSV-1 glycoproteins in gro29 cells in the absence of BFAwas likely due to an inability of these molecules to reach the cellularcompartments which contain these activities.75DiscussionThese data describe the maturation and transport of HSV-1 glycoproteinsand virions in the mutant mouse cell line gro29. It has been shown that therelease of infectious virus from HSV-1 infected gro29 cells was diminished2000-fold (Fig. 3) due to a specific block in viral egress (Fig. 6). Although theassembly and envelopment of nucleocapsids occurred with high efficiency ingro29 cells, the low specific infectivity of the intracellular virus indicates thatthe maturation of the newly-formed virions into infectious particles wasimpaired (Figs. 7,8).It has also been determined that the viral particles that accumulateintracellularly in gro29 cells contain HSV-1 glycoproteins in the viral envelope(Fig. 8, 11). Analysis of the processing of HSV-1 gB and gD during infectionindicates that these, and probably all HSV-1 encoded glycoproteins, weresynthesized normally in gro29 cells and were modified by the addition of N-linked oligosaccharide moieties in the ER (Fig. 4A). The rate of oligosaccharideprocessing of the HSV-1 glycoproteins was lower in infected gro29 cells (Fig.4B), as most of the newly-synthesized glycoproteins were slow to convert toendo H-resistant forms (Fig. 4B). Moreover, subcellular fractionation of gro29cells revealed that newly-made gD was under-represented in the membranes ofthe Golgi complex at 13 HPI (Fig. 5). Despite this impediment,immunofluorescence experiments (Fig. 9D) and flow cytometry indicated thatHSV-1 glycoproteins were transported to the cell surface during infection (Fig.9D).It has been shown previously that gro29 cells are defective in thetransport and processing of glycoproteins (Tufaro et al. 1987). This property76likely accounts for the slow processing of the HSV-1 glycoproteins observed inthis study. What could account for the extreme block to virion egress in this cellline? Several models have been proposed for the maturation and egress ofvirions in HSV-infected cells (see Fig. 2). In one model, HSV-1 virions beginassembly in the nucleus, acquire an envelope as they bud into the perinuclearspace, are carried to the ER and Golgi apparatus and then to the plasmamembrane in vesicles similar to those which carry newly-made membraneglycoproteins and secreted proteins (Johnson and Spear 1982). Budding occurs atthe nuclear membrane, where immature glycoproteins are more prevalent thanthe processed mature forms (Compton and Courtney 1984). Because virionsreleased from infected cells contain mature glycoproteins, it is likely that theenvelope glycoproteins are modified in the Golgi apparatus while resident inthe virus membrane (Johnson and Spear 1982; Spear 1985). A second modelproposes that cytoplasmic vacuoles and not the nuclear membrane are the finalsites of virion envelopment (Nii et al. 1969; Rodriguez and Dubois-Dalcq 1978).This second mechanism is thought to operate in the envelopment of two otherherpesviruses, varicella-zoster virus (Jones and Grose 1988) and pseudorabiesvirus (Whealy et al. 1991). Both models predict that perturbations in thesecretory apparatus of infected cells could affect the maturation and transport ofglycoproteins and virions. The observed block to virus egress without acommensurate block in glycoprotein transport is unique to this cell line, andsuggests that virions and glycoproteins have different cellular requirementsduring the HSV-1 lifecycle.Somatic cell mutants defective in glycosylation enzymes can effect thetransport and processing of HSV-1 glycoproteins, and in some instances the cellsfail to produce normal amounts of infectious virus. Immature forms of several77HSV-1 glycoproteins including gB and gD accumulated when mutant BHK cellsdefective in N-acetylglucosaminyltransferase I activity were infected with HSV-1(Campadelli-Fiume et al. 1982). The release of infectious HSV-1 particles fromthese cells was relatively normal, however, in contrast to the results for gro29.A different mutant cell line, which is defective in the glycosyltransferases thatadd terminal sugars to glycoproteins (Vischer and Hughes 1981), displayedaltered HSV-1 glycoprotein forms upon infection with HSV-1 but was reducedonly three—five fold in the release of viral particles (Serafini-Cessi et al. 1983).The fact that gro29 cells have a novel phenotype with regard to virion egressindicates that gro29 harbors a lesion distinct from those characterizedpreviously.Many of the described observations suggest that the properties of gro29cells are similar to those induced in HSV-1 infected cells treated with theionophore monensin (Johnson and Spear 1982). Monensin has been shown tointerfere with the processing of N-linked oligosaccharides (Wenske et al. 1982),the addition of 0-linked sugars to HSV-1 glycoproteins, the transport of HSV-1glycoproteins from the Golgi apparatus to the plasma membrane, and the egressof virions (Johnson and Spear 1982). A number of criteria have shown thatprocessing of N-linked oligosaccharides, the transport of glycoproteins and theegress of virions are impeded in gro29 cells. The addition of 0-linked sugars tothe HSV-1 glycoproteins was not investigated directly in this study, although theabsence of the larger forms of the glycoproteins indicated that thesemodifications do not occur efficiently in gro29 cells (Johnson and Spear 1983).The subcellular fractionation of the secretory organelles of gro29 cells (Fig. 5)suggested that the failure to add 0-linked sugars may arise from inefficienttransport preventing the bulk of the newly-made glycoproteins from entering78the trans Golgi and the trans Golgi reticulum, because gD was found associatedwith “early” elements of the secretory pathway.One of the most striking phenomena observed in HSV-1-infected gro29and monensin-treated cells was the accumulation of virions in cytoplasmicvacuoles (Fig. 10). These vacuoles may represent intermediates in the pathwayof virion egress. The electron micrographs (Figs. lOB and D) suggested thatprogeny virions produced in gro29 cells may be fusing with the membranes ofthe vacuoles in which they were contained. It is possible that themaldistribution of gD in the membranes of gro29 cells may allow the cells to besuper-infected from within, resulting in excessive losses of progeny virionswhile they are in the process of being transported out of the cells. Alternatively,it may be that virions which are defective in some way are targeted to thesevacuolar structures. Support for this notion comes from the observation thatthe virions contained in the vacuoles in gro29 cells were altered in theirmorphology when compared with L cell virions (Figs. 1OC and D). It is notknown whether vacuoles of this type are Golgi-derived, as suggested previously(Johnson and Spear 1982), and it is not clear why they are abundant in gro29 cellsand monensin-treated cells.Several important distinctions between the phenotypes of monensintreated cells and gro29 cells have been identified. Treatment of HSV-1 infectedcells with toxic concentrations of monensin does not block virion egress aseffectively as observed for gro29 cells (Johnson and Spear 1982). Moreover,HSV-1 glycoproteins were not detectable on the cell surface of monensintreated cells when analyzed by cell surface iodination (Johnson and Spear 1982),whereas HSV-1 glycoproteins on the surface of infected gro29 cells was readilydetected by immunofluorescence (Fig. 9D) and flow cytometry. The capacity of79gro29 cells to transport at least a portion of their glycoproteins to the cell surfacelikely accounts for their ability to grow in culture. These results suggest thatthe targets for monensin are distinct from those affected by the lesion in gro29cells. It may be that the mutant phenotypes common to monensin-treated andgro29 cells arise in all cells deficient in secretion.What accounts for the difference in the trafficking of individualglycoproteins versus intact virions? Virus glycoproteins encoded by HSV-1 canexist as membrane-resident proteins anchored in cellular organelles orembedded in the virus envelope. Consequently, their cytoplasmic andtransmembrane domains can reside in the virion or in the cytoplasm andorganellar membranes of the infected cells. Although the molecular featuresthat regulate the rate of virus transport through the cell have not beencharacterized, it is likely that the glycoproteins embedded in the viral envelopeinfluence the trafficking of the viral particle. It may be the case that theinteractions of the virion with the secretory organelles are impaired in thegro29 cells. The fact that the requirements for virion and glycoproteintransport are differentiated in gro29 cells suggest that they are differentiated innormal cells as well.It is worth noting that other L cell mutants have been isolated that areunable to support the propagation of animal viruses. The mouse L cell lineCL3, a ricin-resistant derivative, was unable to support the growth of Sindbisvirus (Gottlieb et al. 1975; Gottlieb et al. 1979). Sindbis is an enveloped RNAvirus that encodes two glycoproteins. The cleavage of one of these, PE2 to E2,occurs prior to assembly and is required for virion formation. This cleavagewas blocked in CL3 cells and may account for the failure of Sindbis virus to budfrom the plasma membrane. When Sindbis virus-infected cells were treated80with monensin, virus assembly takes place and enveloped particles can befound in monensin-induced cytoplasmic vacuoles (Johnson and Schlesinger1980). The observations that CL3 cells, gro29 cells and cells treated withmonensin exhibit defects in the processing of glycoproteins and in thematuration and egress of diverse families of enveloped viruses argues stronglythat common cellular components facilitate these events.To further the understanding of the virus-host interactions governingthe processes of viral maturation and egress, the effect of the fungal metaboliteBrefeldin A (BFA) on the propagation of HSV-1 in culture was investigated.BFA is a macrocyclic lactone that causes the rapid redistribution of Golgicomponents into the ER (Doms et al. 1989; Lippincott-Schwartz et al. 1989). It isthought to act by releasing coat proteins from Golgi derived vesicles and therebystimulate the fusion of the Golgi membranes with an intermediate “recycling”compartment (reviewed in Pelham 1991). Incubation of cells with BFA resultsin the loss of a discernible Golgi structure and blocks transport of proteins intopost-Golgi cellular compartments (Misumi et al. 1986; Oda et al. 1987). Otherprocesses such as endocytosis, protein synthesis and lysosomal degradationappear unaffected at the low concentrations of BFA used in these studies(Lippencott-Schwartz et al. 1991; Misumi et al. 1986; Nuchtern et al. 1989;Strous et al. 1993). Removal of BFA results in the rapid flux of Golgicomponents out of the ER and reorganization of the Golgi (Doms et al. 1989;Lippincott-Schwartz et al. 1989).The results presented with respect to treatment of HSV-1 infected cellswith BFA indicated that the block to secretion elicited by this drug is clearlydifferent than the block to secretion observed in gro29 cells. These differencesare indicated by the ability of BFA treated L cells to process HSV-1 glycoproteins81to a greater extent than in gro29 cells (Fig. 12). In infected L cells treated withBFA, gD became more substantially processed to the point of becoming resistantto cleavage by endo H (Fig. 12). In contrast to this, gD synthesized in infectedgro29 cells remained in an endo H sensitive form and did not display thedecrease in electrophoretic mobility observed in the gD synthesized in infected Lcells.The effects of BFA on the egress of HSV-1 have been well characterized(Chatterjee and Sarkar 1992; Cheung et al. 1991). Cheung and others were ableto determine that release of virions into the extracellular medium was blockedby as little as 0.3 jtg/ml BFA when present from 2 h post-infection.Characterization of infected cells revealed that BFA inhibited the formation ofinfectious viral particles without affecting the formation of nucleocapsids.Electron microscopic analyses of BFA-treated and untreated cells demonstratedthat viral particles were enveloped at the inner nuclear membrane in BFAtreated cells and accumulated aberrantly in this region. Viral particles thatentered the cytoplasm of BFA-treated cells lacked an envelope, suggesting thatBFA prevented the transport of infectious particles into the cell cytoplasm. Byremoval of BFA at 18 HPI, it was demonstrated that the BFA-induced block toviral propagation was not fully reversible, despite the observation that thesecretion of human growth hormone synthesized in these same cells from anexpression plasmid was restored within 15 mm of BFA removal. These findingsindicated that the BFA-induced retrograde movement of molecules from theGolgi complex to the ER early in infection arrests the ability of the host cell tosupport the maturation and egress of enveloped viral particles. Furthermore,the effects of BFA on HSV propagation were not fully reversible, indicating thatthe maturation and egress of HSV-1 particles relies on a series of events which82cannot be easily reconstituted after the block to secretion is relieved. Takentogether these results suggested that the impediment to secretion effected byBFA occurs earlier in the secretory pathway than the lesion in gro29 cells.The action of BFA and the mutation in gro29 cells were shown to bedistinct, therefore it was of some interest to determine if there were anyobservable effects of BFA on viral glycoprotein processing in gro29 cells (Fig.13).It was shown that the viral glycoprotein gD became more fully processed in thepresence of BFA likely due to the retrograde movement of Golgi residentproteins to the endoplasmic reticulum in these cells. These data allowed theformation of two conclusions regarding the nature of the mutation in gro29cells. First, gro29 cells were susceptible to the effects of this drug suggesting thatthe molecular target for BFA is present and functional in gro29 cells. This isperhaps not surprising because of the fact that most cells are sensitive to at leastsome of the effects of BFA (reviewed in Peiham 1991), and therefore themolecular targets of this drug are likely essential for cellular function. Second,the fact that more extensively glycosylated gD was identified in gro29 cellsindicated that the enzymes required for the further maturation of theglycoprotein found in infected gro29 cells were present in these cells, andfurthermore were located in a BFA sensitive compartment, likely the Golgiapparatus.83Chapter 2- Characteristics of uninfected gro29 cells.A detailed analysis of the uninfected gro29 cell phenotype was required inorder to determine the primary defect in this mutant cell and how it relates toherpesvirus infection. The discovery of alterations in cellular functions ingro29 might lead to a better understanding of the requirements for HSV-1 egressin culture.ResultsLectin sensitivity of gro29.It is well established that alterations in glycoprotein processing (Stanley1984) or protein transport into or out of cells (Colbaugh et al. 1988; Colbaugh etal. 1989; Laurie and Robbins 1991; Robbins et at 1984; Robbins et al. 1983; Roffet al. 1986) can result in resistance to specific lectins. Because gro29 cells showedalterations in transport and processing of proteins, their sensitivity to a varietyof lectins was tested. Cells were incubated at low density in increasingconcentrations of lectin. The lectin concentration resulting in a 90% decrease inthe number of colonies formed (D10) was determined (Michaelis et al. 1992). Inthese assays, clones of gro29 cells were 32-fold and 16-fold resistant to the toxinsricin and modeccin compared with parental L cells. They were not significantlyaltered in resistance to the other lectins tested, including wheat germ agglutinin(WGA), L-PHA, concanavalin A (ConA), or RCA120. The substantial crossresistance to both ricin and modeccin is a rare phenotype that has been reportedfor only one other cell line (Laurie and Robbins 1991).To determine whether the lectin-resistance phenotype was dominant orrecessive, somatic cell hybrids were formed between clones of gro29 (gro29neo),which were resistant to G418, and a parent cell line expressing thymidine kinase84(LTK) (Michaelis et al. 1992). Hybrid cells were selected in HAT mediumcontaining 0.48 mg/ml G418. After several weeks, cells were tested for lectinsensitivity. Results from multiple fusion experiments indicated that the lectinresistance phenotype was recessive when gro29 was fused to parental cells,suggesting that gro29 had a loss-of-function mutation. The hybrid cells alsoregained the ability to propagate herpes simplex virus.Fluorescence detection of lectin binding molecules.Could lack of lectin binding molecules in gro29 cells account for theresistance to these toxins? An example of this is the MDCK cell mutant,MDCKIIRCAr, which is resistant to the lectin RCA120 (binds Gal) and isdefective in the translocation of UDPGaI into the lumen of the Golgi (Brãndli etal. 1988). This cell line shows a large decrease in the binding of fluorescentlytagged RCA120 to cell surface molecules. Based on these observations, lectinbinding molecules in L and gro29 cells were examined.L and gro29 cells growing in monolayers were labeled with differentfluorescently labeled lectins to characterize their intracellular glycoconjugates.Cells were labeled immediately after fixation and permeabilization withfluorescein conjugated ricin, WGA, and RCA120 in separate incubations. Figs.14A and 14B shows the binding of ricin (Gal and Ga1NAc) to parental L andgro29 cells. In parental L cells (Fig. 14A), ricin bound to a juxtanuclear regionresembling the Golgi complex and to small, punctate spots distributedthroughout the cytoplasm. By contrast, the binding of ricin to the mutant cellswas strongly reduced compared to the parental L cells (Fig. 14B). Longerexposures of gro29 cells revealed binding to small, punctate cytoplasmic spotssimilar in appearance to those in L cells. This contrasts markedly with thepattern of staining of WGA (sialic acid), which bound substantially to Golgi-likeFig. 14. Fluorescence analysis of lectin-binding molecules in L and gro29 cells.Parental L and gro29 cells were grown on glass coverslips in DMEM/ 10% FBS for 3days. Monolayers were then fixed with 2% formaldehyde and permeabiized withTriton X-100 as described (Materials and Methods). The distribution of ricin bindingmolecules (A,B) and WGA molecules (C,D) in L cells (A,C) and in gro29 cells (B,D) wasdetermined using fluorescein-conjugated lectins. Published in Michaelis et al. 1992.859887regions of both L and gro29 cells (Fig. 14C,D). A similar pattern to that of WGAwas observed when cells were stained with RCA120 although the intensity ofstaining was much lower (not shown).The observed decreased ricin binding by fluorescence microscopysuggested that the Golgi has fewer ricin binding molecules. As mentionedabove this lectin binds to Gal and Ga1NAc. Gal usage in gro29 cells appeared tobe normal from the analysis of the transport and processing of VSV G protein(Michaelis et al. 1992; Tufaro et al. 1987), and the fluorescence microscopy ofRCA120 binding appeared normal in comparison to L cells. Taken together,these data suggested that Ga1NAc addition to macromolecules was reduced ingro29 cells.This conclusion was also supported by the analysis of PrV egress in gro29cells (Whealy et al. 1992); where it was shown that the PrV glycoprotein gp5O,which contains no N-linked oligosaccharides but has 0-linked oligosaccharidemodifications (Petrovskis et al. 1986; Whealy et al. 1990b), failed to become 0-glycosylated efficiently and completely. The fact that 0-linked carbohydratemoieties are linked to serine or threonine through a Ga1NAc residue (Roseman1970), suggested that the lack of available GalNAc in the gro29 cell could resultin the lack of this modification.Analysis of glycosaminoglycan synthesis in gro29 cells.Due to the apparent block to the addition of Ga1NAc to macromoleculesin gro29, it was hypothesized that the synthesis of the glycosaminoglycanchondroitin sulfate might be defective as this molecule contains GalNAc in itsrepeated disaccharide unit (Kjellen and Lindahi 1991). To test this hypothesis,monolayers of gro29 and parental L cells were incubated with[35S]-sulfate and[3H1-glucosamine to label glycosaminoglycans. After three days of labeling,88glycosaminoglycan fractions from cell extracts were prepared and analyzed byanion exchange HPLC. Fractions eluting from the HPLC column were collectedand counted by liquid scintillation spectroscopy. It can be seen in Fig.15 that thecontrol L cell glycosaminoglycans resolved into two major sulfated peakscontaining predominantly heparan sulfate (HS) and chondroitin sulfate (CS)and three major glucosamine labeled peaks containing hyaluronic acid (HA),heparan sulfate (HS) and chondroitin sulfate (CS), respectively. Analysis oflabeled gro29 proteoglycans showed that there was no detectable chondroitinsulfate associated with the gro29 cells. In contrast to these results gro29 wascapable of synthesizing the proteoglycan heparan sulfate although the amountwas decreased when compared to L cell heparan sulfate synthesis. It may be thatthe decrease in heparan sulfate observed was due to the reduced amount offunctional Golgi membranes found in gro29 cells (Michaelis et al. 1992).One hypothesis that might explain the results observed with respect toglycosaminoglycan synthesis is that uptake of either glucosamine or sulfate bygro29 cells does not occur efficiently, and this affects chondroitin sulfatesynthesis to a greater extent than heparan sulfate synthesis. To ensure thatgro29 cells were capable of efficient glucose and sulfate uptake, the uptake ofglucose and sulfate from the medium was measured directly.To determine the rate of glucose uptake by L and gro29 cells dishes of cellswere incubated in a fixed concentration of radiolabeled glucosamine andvarying concentrations of unlabeled glucose for a period of 1 h. At the end ofthe labeling period cells were harvested, and transported glucosamine measuredby liquid scintillation spectroscopy. Results were corrected for differences inprotein concentration, and the number of pmoles of glucose transported wasdetermined. As shown in Fig.16 the rates of glucose uptake of L and gro29 cellswere similar, although less glucose appeared to have been taken up by the gro2989Fig. 15. Analysis of glycosaminoglycari synthesis in L and gro29 cells.Monolayers of L and gro29 cells were incubated in low glucose! low sulfate mediumcontaining[35S]-sulfate and[3H1-glucosamine for three days to labelglycosaminoglycans. After this incubation glycosaminoglycan chains were isolated fromthe cells and analyzed by anion exchange HPLC as described (Materials and Methods).Fractions eluting from the column were collected and radioactivity in the fractionsdetermined by liquid scintillation spectroscopy. The results are presented in the figureas CPM/fraction, glucosamine and sulfate counts are plotted independently.HA=hyaluronic acid, HS=heparan sulfate, CS=chondroitin sulfate. L cell glucosamineclosed boxes; L cell sulfate open diamonds; gro29 glucosamine open boxes; gro29sulfate closed diamonds. Glycosaminoglycan analysis by S. Gruenheid, F. Tufaro, andK. Schubert.600050004000E C3300020001000ProteoglUcenAnelUsisofLendgro29Cells80——I—-—Lce11g1ucosmine—*—---—Lcellsulfote—D——gro29glucosominegro29sulleteHAHSCs0204060FrctlonNumber91Fig. 16. Measurement of glucose uptake in L and gro29 cells.35 mm dishes of L and gro29 cells were rinsed 3 times with low sulfate/low glucosemedium prior to the addition of low sulfate/low glucose medium containing 0,25,50,100 or 300 M glucose and 20 jiCi[3H]-glucosamine. Dishes were incubated for 60 mm (60 mm was shown to be within the linear range of glucosamine uptake versus time inseparate experiments (data not shown)) at 37°C 5% C02,when they were rinsed 3 timeswith ice cold PBS followed by collection of the cell lysate in 0.1 N NaOH. Proteinconcentrations of the cell lysates were determined and the radioactivity in the sampledetermined by liquid scintillation spectroscopy. The specific radioactivity of eachsample was determined and the kinetics of glucose uptake versus glucose concentrationin the incubation medium was plotted. L cell open boxes, gro29 cell closed diamonds.92Glucose Uptake in L and gro29 Cells12—0—-— L cell10 gro29 cellci)4-, 4.E0.20• . • I • I • I • I • I •0 50 100 150 200 250 300 350 400[Glucose] (jiM)93cells per .tg of protein. Nonetheless, efficient uptake of glucose was observed ingro29 cells. The uptake of[35S1-sulfate by L and gro29 cells over time wasmeasured (Fig. 17). No significant difference was observed in the uptake ofsulfate between L and gro29 cells. These results suggested that the failure tosynthesize chondroitin sulfate was not a result of the failure to transport glucoseor sulfate into gro29 cells.These data indicated that chondroitin sulfate was not being made ingro29 cells. As mentioned above the fundamental difference between heparansulfate and chondroitin sulfate glycosaminoglycans is that the chondroitinsulfate repeating disaccharide unit is glucuronic acid (GalA) linked to GaINAcwhereas the primary structure of heparan sulfate consists of repeating units ofGalA linked to G1cNAc or N-sulfoglucosamine (G1cNSO3)(for review seeKjellen and Lindahl 1991). The results of these experiments suggested that therewas a deficiency in the incorporation of GalNAc into chondroitin sulfate ingro29 cells.Investigation of Ga1NAc metabolism in gro29 cells.The observed lectin resistance, lectin binding, deficiency in 0-linkedglycosylation and the failure to synthesize chondroitin sulfate strongly suggesteda defect in the metabolism of Ga1NAc in gro29 cells. Before Ga1NAc can beincorporated into macromolecules of cells growing in culture a number ofreactions must occur. First, glucose (Glu) from the medium must be transportedinto the cell (Fig. 18). The Glu is converted to UDP-Ga1NAc by numerousenzymatic reactions in the cell cytoplasm (116, 117). Many of the reactionsconverting Glu to UDP-GalNAc are common to the de novo biosynthesis ofmany cellular metabolites (DelGiacco and Maley 1964; Glaser 1959; Maley andLardy 1956; Maley et al. 1968; McGarrahan and Maley 1962; Richmond 1965),and as such defects in this pathway would be predicted to be lethal or to have94Fig. 17. Rate of sulfate uptake in L and gro29 cells.35 mm dishes of L and gro29 cells were rinsed 3 times with low sulfate/low glucosemedium prior to the addition of low sulfate/low glucose medium containing CPM[35S] sulfate. Dishes were incubated for 0,5, 10,30 and 60 mm at 37°C, 5% CO2 prior tobeing rinsed 3 times with ice-cold PBS followed by collection of the cell lysate in 0.1 NNaOH. Protein concentrations of the samples were determined and the amount ofradioactivity in the sample determined by liquid scintillation spectroscopy. The amountof radioactivity transported into the cells versus time is plotted. L cell open boxes,gro29 cell closed diamonds.95Sulfate Uptake in L and gro29 Cells4000—El--— L cellgro29 cell3000. 2000‘4-Cl)E0 20 40 60 80Time (mm)Fig. 18. Synthesis of UDPGa1NAc.The biosynthesis of UDPGa1NAc and UDPGa1 from glucose, galactose, and Nacetylgalactosamine is outlined. Note in particular the position of the enzyme UDPGa1-4-epimerase in the biosynthetic pathway. P=phosphate.9697UDPGa1-4-EpimeraseUDPGIc UDPGaIGal-i -PGIc-i -PGal a GalactoseGlucos ‘ GIc Glc-6-PI4-”4-Fru-6-PCELL+ CYTOPLASMI GalNAc N-AcetylGIcNAc-6-P GalactosamineGaINAc-1 -PI_I4-”4-UDPGIcNAc 4 UDPGaINAcUDPGa1-4-EpimerasePLASMA MEMBRANE98broad consequences affecting cellular metabolism. Upon synthesis thenucleotide sugar is translocated into the organelle in which the incorporationreaction is to take place (Fig. 19).gro29 cells appear to be unable to incorporate Ga1NAc intomacromolecules, so any defect in the metabolism of this sugar should be specificto the metabolism of Ga1NAc and not grossly affect the metabolism of othersugars. The addition of sugars such as glucosamine (Glc), galactose (Gal),mannose (Man), N-acetylglucosamine (GlcNAc) and sialic acid (SA) appearrelatively normal in gro29 cells (based on the analysis of HSV-1 and VSVglycoproteins (Banfield and Tufaro 1990; Michaelis et al. 1992; Tufaro et al. 1987)and the ability to synthesize heparan sulfate Fig.15), therefore biochemicalpathways having these sugars as intermediates should not be affected in gro29cells. Furthermore, the enzymes which transfer GalNAc from UDPGa1NAc to amacromolecular substrate tend to be specific to each acceptor substrate, so it isunlikely that all of these gene products are affected in gro29 cells. Based on thesehypothesis experiments were carried out in an attempt to determine the natureof the defect to Ga1NAc metabolism observed in gro29 cells.Rescue of chondroitin sulfate synthesis in gro29 cells.A Chinese hamster ovary (CHO) cell line has been described with areversible defect in 0-linked glycosylation (Kingsley et al. 1986). It was foundthat this mutant cell line, idiD, was defective in the enzyme UDP-Gal 4-epimerase, and that addition of Ga1NAc to the medium restored the ability ofthese cells to synthesize 0-linked carbohydrate chains (mammalian cells cansynthesize UDPGa1NAc directly from Ga1NAc via a salvage pathway (Maley etal. 1968)), thereby bypassing the epimerase function (see Fig. 18). It was thereforeof some interest to determine if the addition of Ga1NAc to the medium of gro2999Fig. 19. Translocation of UDPGa1NAc into Golgi membranes.The translocation of UDPGaINAc into the lumen of a Golgi vesicle is outlined. 1)UDPGaINAc is translocated into the Golgi lumen by an antiport exchangingUDPGa1NAc for UMP. 2) Ga1NAc is transferred from UDPGa1NAc to a macromolecularacceptor through the action of a specific UDPGa1NAc transferase. 3) The UDP releasedfrom UDPGa1NAc by the action of the transferase is dephosphorylated by aphosphatase to form Pi and UMP. 4) A permease enables the Pi to exit the Golgi. 5) TheUMP generated by the action of the phosphatase can now be used to exchange foranother nucleotide sugar. Adapted from Hirschberg 1987. Pi= inorganic phosphate.100CytosolUMPUDPGaINAc5GolgiMembrane101cells restored the ability of these cells to synthesize chondroitin sulfate as wouldbe predicted if the epimerase function were defective in gro29.Cells were grown for two days in medium supplemented with 10 p.M Gal,100 p.M Ga1NAc (Kingsley et al. 1986), prior to labeling and HPLC analysis ofglycosaminoglycans as described above. Fig. 20 shows the results of sulfateincorporation into heparan sulfate and chondroitin sulfate by gro29 cells in thepresence or absence of Gal and GalNAc. As can be seen in Fig. 20 the addition ofGal and Ga1NAc to the medium of gro29 cells restored the ability of gro29 cells tosynthesize chondroitin sulfate. These data indicated that gro29 cells containedthe enzymes involved in the synthesis of chondroitin sulfate and that the defectin these cells was likely due to the availability of GalNAc.Measurement of UDPGa1-4-epimerase activity in cell extracts.To determine if the defect in gro29 cells was the lack of UDPGa1 4-epimerase activity, as suggested by the ability of exogenous Ga1NAc to restorechondroitin sulfate synthesis, this enzyme activity was measured. The enzymeconverts UDPG1c to UDPGa1 reversibly, and UDPG1cNAc to UDPGa1NAcreversibly (Maley and Maley 1959; Filler et al. 1983)(see Fig. 18). The UDPGa1 4-epimerase activity was measured in L and gro29 cell extracts by a two-stepspectrophotometric assay which quantitates the conversion of UDPGa1 toUDPG1c over time (Kingsley et al. 1986). The results of this experiment areshown in Fig. 21. It is clear from these results that the UDPGa1 4-epimeraseactivity is normal in gro29 cells.To insure that the epimerase activity measured above was not specific forthe conversion of UDPGa1 to UDPG1c, the conversion of radiolabeledUDPGaINAc to UDPG1cNAc in L and gro29 cell extracts was measured. Themeasurement of the accumulation of radiolabeled UDPG1cNAc over time wasfacilitated by separating the nucleotide sugars by thin layer chromatography on102Fig. 20. Rescue of chondroitin sulfate synthesis in gro29 cells.Monolayers of gro29 cells were incubated in medium with or without 10 ji.M Gal, 100I.LM GaINAc for two days prior to incubating the cells in low sulfate medium with [3S5]sulfate (with or without 10 jiM Gal and 100 jiM GalNAc) for three days to labelglycosaminoglycans. After this labeling period a glycosaminoglycan fraction wasprepared from the cells and analyzed by anion exchange HPLC and subsequent liquidscintillation spectroscopy of fractions eluted from the column. These data are presentedhere as CPM/fraction. HS=heparan sulfate; CS= chondroitin sulfate. Control= gro29without Gal/Ga1NAc preincubation, closed boxes; gro29 cells +Gal/Ga1NAcincubation, closed diamonds. Glycosaminoglycan analysis performed by F. Tufaro andK. Schubert.400E L)Proteog1Jc8nAn6lysisofgro29Cells+1—Gl6ncIGlNAc800204060FrectionNumberFig. 21. Measurement of UDPGa1-4-epimerase activity in L and gro29 cell extracts.Samples of L and gro29 cell extracts were assayed for UDPGa1-4-epimerase activity asdescribed (Kingsley et al. 1986). Control samples were boiled for 5 mm prior toperforming the assay. UDPG1c production is directly proportional to absorbance at 340nm and is measured over time. L cell open boxes, gro29 cell closed diamonds, Lcontrol closed boxes, gro29 control open diamonds.104105UDPGaI-4-Epimerase Assay020—— L cellgro29 cell—*‘---- Lboiled0 gro2led...•0.10•UI0-D< 0.050.00•0 10 20 30Time (mm)106PET cellulose (Kingsley et al. 1986; Randerath and Randerath 1965). Fig. 22shows the results from the experiment. As can be seen in the graph, there wereno significant differences detectable between the epimerase activity from theextracts of L and gro29 cells.These results indicated that the UDPGa1-4-epimerase activity in gro29cells was normal. The addition of exogenous Ga1NAc to the medium of gro29cells was able to stimulate chondroitin sulfate synthesis perhaps by raising theconcentration of intracellular GaTNAc to a level which could overcome theobserved defect. The addition of GalNAc to the medium of gro29 cells did notrestore the ability of gro29 cells to bind fluorescently labeled ricin nor did itrestore the ability of gro29 to propagate HSV-1, as might be expected if theUDPGa1-4-epimerase were defective.Measurement of nucleotide sugar translocation into Golgi membranes.It is well established that lectin resistant cell lines can be defective in thetransport of specific nucleotide sugars into the Golgi apparatus. For example,the CHO cell mutant Lec2 is resistant to the lectin WGA (binds sialic acid), andwas found to be defective in the translocation of CMP-sialic acid into the lumenof the Golgi (Deutscher et al. 1984). Also the MDCK mutant, MDCKIIRCAr, isresistant to the lectin RCA120,which binds Gal, and is defective in thetranslocation of UDP-galactose into the lumen of the Golgi apparatus (Brändli etal. 1988). Because of the lectin resistant phenotype, and the failure toincorporate GalNAc into macromolecules in gro29 cells, the translocation ofUDP-Ga1NAc into Golgi membranes isolated from L and gro29 cells wasmeasured (see Fig. 19). Considering that gro29 cells are capable of incorporatingsialic acid into macromolecules (Tufaro et al. 1987, also see Fig.14D), themeasurement of CMP-sialic acid translocation into Golgi membranes was usedas a control.107Fig. 22. Conversion of UDPGa1NAc to UDPG1cNAc L and gro29 cell extracts.Samples of L and gro29 cell extracts (containing 240 pg of protein) were incubated at37°C in the presence of[3H1-UDPGa1NAc. At 0,5,7, 10 and 12 mm samples of thereactions were spotted on a PEI cellulose TLC plate which was pre-loaded with 50nmole each UDPGa1NAc and UDPG1cNAc. Nucleotide sugars were resolved asdescribed (Kingsley et al. 1986; Randerath and Randerath 1965), the UDPG1cNAcexcised from the TLC plate and radioactivity determined by liquid scintillationspectroscopy. The amount of[3H]-UDPGa]NAc converted to[3H]-UDPG1cNAc isplotted versus time of incubation. L cell open boxes, gro29 cell closed diamonds.108UDPGaINAc to UDPGIcNAc Epimerase Assay0.8-—0--—— L cell extractgro29 cell extract0.6-t3zC-,CDD 0.4-0£a.0.2-0.0 . I I I0 5 10 15Time (mm)109The translocation assays were performed as described in the materials andmethods section,[14C]-acetate was used as a membrane impermeable standardthroughout these experiments. To ensure that the membrane preparationswould be stable over the entire assay period, the rate of nucleotide sugar uptakewas first determined (Fig. 23). Golgi membranes from L and gro29 cells wereincubated at 30°C in buffer containing a fixed specific activity of[3H1-CMP sialicacid or[H]-UDPGa1NAc, for various lengths of time. As can be seen from thesedata, the rate of uptake of CMP sialic acid (Fig.23A) and UDPGa1NAc (Fig. 23B)appear linear over the sampling period although the rate of CMP sialic acidtransport is reduced in the gro29 cell membranes. From these data it wasdetermined that the kinetics of nucleotide sugar translocation would bemeasured using an incubation period of 10 mm when the rate of translocationwas linear.Fig. 24 shows the kinetics of nucleotide sugar translocation in L and gro29membrane fractions. L and gro29 Golgi membranes were incubated in varyingconcentrations of nucleotide sugar for 10 mm at 30°C before determination ofthe amount of nucleotide sugar translocated into the lumen. The amount ofCMP-sialic acid translocated into the Golgi versus the CMP-sialic acidconcentration in the reaction was considerably lower in gro29 cells as mighthave been predicted from the membrane stability assays (Fig. 23A), however it isclear from Fig. 14D that sialic acid is incorporated into macromolecules in gro29cells and furthermore the glycoprotein of VSV is capable of becomingoversialated in gro29 in relation to VSV infected L cells (Michaelis et al. 1992). Ithas been shown that the majority of the CMP-sialic acid translocated into Golgimembranes from CHO cells during these assays is incorporated into indigenousmacromolecular acceptors during the assay (Deutscher et al. 1984; Sommers andHirschberg 1982), thus it is possible that the decrease in CMP-sialic acidFig. 23. Rate of nucleotide sugar translocation into L and gro29 Golgi membranes.Preparation of Golgi membranes and procedures for nucleotide sugar translocationassays are described (Materials and Methods). Golgi derived membranes from L andgro29 cells were incubated in the appropriate buffer for various times at 3O’C beforetermination of the reactions. All reactions were performed in duplicate. (A) Time courseof CMPSA translocation into Golgi membranes. (B) Time course of UDPGa1NActranslocation into Golgi membranes. L cell membranes open boxes, gro29 cellmembranes closed diamonds.110111Time Course of CMP Sialic AcidTra nslocation into Goig i Membranes0.04-C —D-—-— L cell membranes152 0.03-0)0.02-C-)0.01--E0.00_0 10 20 30Time (mm)Time Course of UDPGaINAcTrarislocation into Golgi Membranes0.04C—D-——— L cell membranes B00.03gro29 membranes0)C.)0.020D0.01E0.0.000 5 10 15 20Time (mm)112Fig. 24. Kinetics of nucleotide sugar translocation into L and gro29 cell Golgimembranes.Preparation of Golgi membranes and procedures for nucleotide sugar translocationassays are described (Materials and Methods). L and gro29 membranes were incubatedin various concentrations of the appropriate nucleotide sugar for 10 mm at 30°C beforetermination of the reactions. All reactions were performed in duplicate. (A) Kinetics ofCMPSA translocation into Golgi membranes. (B) Kinetics of UDPGa1NAc translocationinto Golgi membranes. L cell membranes open boxes, gro29 cell membranes closeddiamonds.113CMP Sialic Acid Translocation into Golgi Membranes180-E160 —D——— gro29 cell membranes0140- L cell[CMPSA] (pM)UDPGaINAc Translocation into Golgi MembranesE 800-700 - —D-——- gro29 cell membranes BL cell membranes600-I• 30[UDPGaINAc] (pM)114translocation is due to the reduced number of macromolecular acceptors presentin the gro29 Golgi compartment because of the reduced Golgi capacity of thesecells (Michaelis et al. 1992). Fig 24B shows the rate of translocation of UDPGa1NAc versus UDPGa1NAc concentration for L and gro29 cells. The ability ofgro29 membranes to translocate UDPGa1NAc is indistinguishable from L cellsup to a substrate concentration of 5 p.M UDPGa1NAc. At a substrateconcentration of 25 p.M the L cell UDPGa1NAc translocator appeared to approachVmax whereas the gro29 translocator did not. These data suggested that thegro29 UDPGa1NAc translocator may be a “better” pump than the L cell pump. Itis not known if the differences observed between the L and gro29 UDPGa1NActranslocator are physiologically relevant.A formal possibility which may explain these results is that theUDPGa1NAc pump is functional in gro29 cells but it is not located in theappropriate intracellular compartment. Unfortunately no reagents exist todetect the protein or proteins which comprise the pump, so it is not yet possibleto determine if this is the case.A growing body of evidence suggests that the concentration and relativeratio of nucleotide sugar concentrations can have profound effects on proteintransport and glycosylation in vitro (Davidson and Balch 1993). Moreover,other molecules such as uridine monophosphokinase have been shown toinfluence the translocation of nucleotide sugars into the lumen of the Golgi invitro (Hiebsch and Wattenburg 1992). It is conceivable that gro29 cells have adefect in modulating the availability of nucleotide sugars such that under“normal” circumstances (i.e. uninfected cells) the cells are capable of sustaininga basal level of protein transport and glycosylation to permit cell survival butupon infection with HSV-1 the tremendous increase in glycoprotein synthesissaturates the Golgi capacity of these cells quickly and irreversibly, resulting inthe inability to produce progeny virus.115116DiscussionThe most striking phenotype observed in uninfected gro29 cells was thesubstantial cross-resistance to the toxic lectins ricin and modeccin (Michaelis etal. 1992). Ricin and modeccin are potent cytotoxins comprising two disulfidelinked peptides, the A chain and B chain. The B chain binds to galactose and/orN-acetylgalactosamine-containing receptors on the cell surface, and the lectinenters the cell by receptor-mediated endocytosis. Following intracellularvesicular transport, ricin and modeccin are translocated to the cytosol, wherethey catalytically inactivate the 60 S ribosomal subunit. However, there isevidence indicating that modeccin and ricin recognize different receptors on thecell surface, have different requirements for endosomal acidification, and havedifferent ribosomal targets (Stanley et al. 1990). Based on these differences, it ispossible that cells resistant to both toxins are defective in transporting the toxininto the cytoplasm from the cell surface. Alternatively, the cell-surface receptorsfor modeccin and ricin may share a requirement for a specific oligosaccharidemodification, the loss of which may disrupt the affinity of the two lectins fortheir respective receptors.The possibility that gro29 cells are altered in a carbohydrate modificationis supported by results showing that gro29 cells were lacking most of the ricinbinding molecules normally present in the parental L cells (Fig. 14). Althoughmodeccin, ricin and RCA120 bind to galactose-containing structures, it isprobable that sugar residues adjacent to those in glycoprotein oligosaccharidesinfluence the extent of binding (Olsnes et al. 1978). It may be the case that ricinand modeccin recognize predominantly GalNAc containing structures in L-cells,whereas RCA120 recognizes distinct Gal structures, because there was noapparent block to the addition of Gal to N-linked oligosaccharides in gro29 cells117(Michaelis et al. 1992; Tufaro et al. 1987; Whealy et al. 1992). The reducedlabeling of gro29 cells with fluoresceinated ricin may result from a failure to addGalNAc to newly synthesized molecules. In support of this possibility was theobservation that gro29 cells were severely impaired in their ability to synthesizechondroitin sulfate chains (Fig.15), which are composed of alternatingdisaccharides of glucuronic acid and Ga1NAc. By contrast, the synthesis ofheparan sulfate chains, composed of hexuronic acid and N-acetylglucosamine,was reduced yet apparently normal (Fig. 15). Furthermore, it has beendetermined that gro29 cells were deficient in their ability to 0-glycosylate viralglycoproteins (Banfield and Tufaro 1990; Whealy et al. 1992), this is interestingbecause the first sugar added to serine or threonine in an 0-linked glycan chainis Ga1NAc, further suggesting that a defect in the metabolism of Ga1NAc may beresponsible for the lesion observed in gro29 cells.One way to explain the observed cross-resistance to modeccin and ricin,the inability to synthesize chondroitin sulfate and the inability to 0-glycosylateviral glycoproteins is a reduced UDPGa1-4-epimerase activity in gro29 cells. Thisenzyme converts UDPG1c and UDPG1cNAc to UDPGa1 and UDPGa1NAcrespectively. A defect in this enzyme would alter the UDPGa1NAc availabilityin cells. A cell line has been characterized which displays a defect in this enzyme(Kingsley et al. 1986) . This cell line, termed idiD, shows a defect in theformation of 0-linked oligosaccharide moieties on glycoproteins. The additionof Gal and GalNAc to the culture medium of these cells restores their ability toO-glycosylate proteins. To determine if the epimerase was defective in gro29cells the effect of adding exogenous Gal and GalNAc on the synthesis ofchondroitin sulfate was determined (Fig. 20). Addition of these sugars to themedium of gro29 cells restored the ability of gro29 cells to synthesize118chondroitin sulfate, although it also modified the elution profile of heparansulfate from an anion exchange HPLC column. Nevertheless, these datasuggested that the epimerase might have been defective. This enzyme activityfrom L and gro29 cell extracts was measured and found to be normal (Figs. 21and 22). These data suggested that the ability of additional Gal and GaINAc torestore the synthesis of chondroitin sulfate was not due to the lack of afunctional epimerase, however, it also indicated that the levels of these sugarsinside gro29 cells can have an effect on glycosaminoglycan synthesis. Moreover,the ability of gro29 cells to synthesize chondroitin sulfate upon addition ofexogenous Gal and GalNAc indicated that these cells harbored the enzymaticmachinery required to synthesize this complex glycosaminoglycan chain.Another possibility which might explain the apparent Ga1NAc deficiencyin gro29 cells is if the UDPGa1NAc translocator were missing or defective inthese cells. The UDPGa1NAc translocator is an antiport which translocatesUDPGa1NAc from its site of synthesis in the cytoplasm into the lumen of theGolgi apparatus where incorporation of GalNAc into macromolecules takesplace (see Fig. 19). A defect in this translocator would result in an under-representation of GalNAc in complex molecules synthesized by enzymes of theGolgi complex. There is precedent for such a lesion. Cell lines with defects insugar nucleotide transporters have been isolated, and show resistance to avariety of lectins. In particular, the MDCKII cell mutant MDCKRCAr shows adefect in UDPGa1 transporter activity (Brändli et al. 1988). Interestingly, thesecells are resistant to RCA120,but not to ricin. Moreover, in the CHO mutantLec8, which shows 100-fold resistance to killing by WGA compared with theparental CHO cells (Stanley 1981), UDPGa1 transport is reduced to 3-5% ofnormal, whereas the transport of UDPGa1NAc is unaffected (Deutscher and119Hirschberg 1986). These data also indicate that distinct activities exist for thetranslocation of UDPGa1 and UDPGa1NAc.The transport of UDPGa1NAc into Golgi membranes isolated from L andgro29 cells was measured (Figs. 23B and 24B). The transport of UDPGa1NAc intogro29 cell Golgi was normal up to a nucleotide sugar concentration of at least 5p.M. At higher concentrations of UDPGa1NAc it appeared that the UDPGa1NActranslocator was more efficient in gro29 cells than in L cells. Why this might bethe case or whether or not this observation is physiologically relevant is notclear. The activity of this molecule may be up-regulated in an attempt tocompensate for the decrease in availability of Ga1NAc in gro29 cells. There is noevidence to support this notion.The transport of CMP sialic acid into Golgi membranes was measured as acontrol because it appeared that the metabolism of sialic acid was relativelynormal in gro29 cells (Fig. 14)(Michaelis et al. 1992; Tufaro et al. 1987).Interestingly, the rate and kinetics of transport of this nucleotide sugar weresignificantly reduced in gro29 cell membranes as compared to the control L cellmembranes (Figs. 23A and 24A). This result may be due to the observation thatmost of the CMP sialic acid transported into Golgi membranes from normalcells is incorporated into indigenous macromolecules very rapidly in theseassays (Deutscher and Hirschberg 1986). The difference in CMP sialic acidtranslocation observed between L and gro29 cells may be due to fewermacromolecular acceptors in the gro29 Golgi, because this assay measures boththe soluble and incorporated nucleotide sugars translocated into the lumen ofthe Golgi membranes. Another possibility to explain this discrepancy is that asialyltransferase activity is reduced in the gro29 cell membranes, which wouldresult in the slower incorporation of sialic acid into indigenous macromolecules120and in turn be reflected in the results of the assay. These questions could beaddressed by determining the relative amounts of acid insoluble radioactivity(incorporated sialic acid) versus the acid soluble radioactivity (CMP-sialic acid)in the Golgi membranes after the assay.Another explanation for the apparent defect in Ga1NAc metabolism ingro29 cells is that the UDPGa1NAc translocator is not located in the appropriatecellular compartment. This could result in the lack of available UDPGa1NAc forthe enzymes which perform the macromolecular synthesis reactions requiringthis substrate. The Golgi apparatus is a highly ordered, compartmentalizedstructure and, as such, any perturbations in its organization may have drasticeffects on the biochemical capabilities of this organelle. Unfortunately, themolecule(s) which comprise the UDPGa1NAc translocator have yet to beidentified and therefore no reagents are available to test this hypothesis directly.A further possibility to explain the decrease in GalNAc incorporation intomacromolecules in gro29 cells is a lack of transferase enzymes required for thesefunctions. Many enzymes would likely be required to form the various Ga1NAclinkages found in higher eukaryotic cells. It is clear that the enzymesresponsible for the incorporation of GalNAc into chondroitin sulfate are presentin gro29 cells, but perhaps another enzyme deficiency exists which has a generaleffect on GalNAc metabolism. The enzyme Ga1NAc:polypeptide Nacetylgalactosaminyltransferase which transfers GalNAc from UDPGa1NAc toserine or threonine on 0-linked glycoproteins has been characterized in somedetail (Elhammer and Kornfeld 1986; Elhammer et al. 1993) and the geneencoding this activity has recently been cloned (Homa et al. 1993). However,many questions regarding the substrate specificity and the precise intracellularlocation of this enzyme remain unclear. It would be of some interest to121investigate the nature of this molecule in gro29 cells. In particular it will beinteresting to study the expression and intracellular location of this enzyme inrelation to L cells as this may provide further clues as to the origin of the defectin gro29 cells. It may be that this enzyme has weak activity, lower levels ofexpression, or it may be aberrantly localized in gro29 cells resulting in theobserved phenotype. The possibility exists that over-expression of this enzymein gro29 may be able rescue some of the defects in these cells. This work shouldnow be possible due to the recent availability of reagents required for this study.Even though multiple transferases would likely have to be defective to result inthe phenotype demonstrated by gro29, the paucity of information regardingthese molecules suggests that the examination of UDPGa1NAc:polypeptide Nacetylgalactosaminytransferase in gro29 cells may be fruitful.It has been observed that the level and ratio of the nucleotide sugars inthe cytosol can have profound effects on the glycosylation of glycoproteins invitro. Davidson and Balch (Davidson and Balch 1993) have shown that in semipermeable NRI< cells increasing concentrations of UDPG1cNAc can inhibit thesialylation of VSV G protein. In addition, it was determined that increasing theconcentration of UDPGa1 in the assay blocked this inhibition and that theinclusion of UTP in their reactions stimulated the synthesis of more fullyglycosylated forms of G protein. A conclusion to be drawn from this study isthat the appropriate balance of individual nucleotide sugars is critical to theprocess of glycosylation in vitro. In another in vitro study, Hiebsch andWattenberg (Hiebsch and Wattenburg 1992) discovered that the enzyme activityof uridine monophosphokinase (UMPK) stimulates the transport of nucleotidesugars into Golgi membranes. UMP is an inhibitor of the translocation ofnucleotide sugars containing UDP into the Golgi apparatus (reviewed in Perez122and Hirschberg 1985). Therefore the effect of UMPK on the translocation ofnucleotide sugars is likely due to the conversion of UMP to UDP by this enzymethereby effectively decreasing the concentration of this inhibitor in the assay.Taken together, these data indicate that molecules not directly involved inprotein glycosylation or secretion can have significant effects on these processes.gro29 cells may be missing such a molecule and this results in the inability ofgro29 cells to efficiently incorporate Ga1NAc into macromolecules.It may be possible that the majority of the ricin-binding moleculesdetected in fluorescence experiments (Fig. 14) is irrelevant to the intoxicationprocess, and that gro29 cells express functional ricin receptors on the cell surface.If this were the case, it may be that toxin uptake into the cytoplasm is defectivein these cells. Previous studies have shown that following endocytosis, ricinmolecules are found in Golgi elements and that ricin cytotoxicity is enhanced bylow concentrations of nigericin or monensin, that disrupt Golgi structure andfunction, and by swainsonine and tunicamycin, that inhibit N-glycosylation(Ghosh and Wu 1988; Yoshida et al. 1990). Other studies have indicated thattreatment of cells with BFA, that causes recycling of Golgi elements into the ER,increases the resistance of the cell to ricin and modeccin (Yoshida et al. 1991).Furthermore, hybridoma cells secreting anti-ricin antibody are relativelyresistant to ricin (Youle and Colombatti 1987), suggesting that ricin interactswith the secretory pathway prior to entering the cytosol. Although it is notpossible from these studies to make conclusions regarding the precise role ofGolgi trafficking in the intoxication process, these results nevertheless implicatethe Golgi complex, and provide evidence that ricin, and by analogy modeccin,associate with the Golgi complex prior to entry into the cytoplasm. The relativelack of functional Golgi membranes in gro29 cells may interfere with the lectin123intoxication pathway, thereby preventing the efficient entry of the toxins intothe cytoplasm.Extensive cross-resistance to the two toxins is not common inmammalian cells, having been reported in only one variant of murine L cells,called LEFIC cells, which were selected originally for resistance to modeccin andPseudomonas exotoxin (Laurie and Robbins 1991). LEFIC cells are 95- and 19-fold resistant to modeccin and ricin, and show delays in the movement ofindigenous proteins along the secretory pathway. Examination of the endocyticpathway in LEFIC cells has shown that uptake and acidification-dependentactivities within early endosomes, and the delivery of endocytosed lysosomalhydrolases to lysosomes appears normal, but there is some evidence thattransport from the Golgi to lysosomes may be impaired. Based on these results,it was suggested that toxin resistance may arise in LEFIC cells due to a failure todeliver a protein required for intoxication from the Golgi to the late endosomein a timely manner, ultimately leading to toxin degradation in lysosomes.Alternatively, a defect in a cellular compartment such as the trans Golginetwork or the late colgi compartments may affect intoxication in these cells.Although any similarity between LEFIC cells and gro29 cells is still at thephenotypic level, it is interesting that both cell lines show significantdeficiencies in protein secretion and lectin intoxication.124General DiscussionThe analysis of PrV infection in gro29 cells has yielded much informationabout the defect in gro29 cells (Whealy et al. 1992). PrV encodes twoglycoproteins which have useful properties for the study of glycosylation andsecretion. One of these glycoproteins, gil, a HSV-1 gB homolog, is cleaved by acellular protease in the trans Golgi or the trans Golgi reticulum after it hasoligomerized and been transported out of the endoplasmic reticulum (Whealyet al. 1991; Whealy et al. 1990). The other glycoprotein is gp5O, a HSV-1 gDhomolog, which is modified by the addition of 0-linked oligosaccharides but notby N-linked oligosaccharide moieties (Petrovskis et al. 1986; Whealy et al. 1991).The post translational modifications of these glycoproteins were monitored bypulse-chase experiments at various times after infection of gro29 cells with PrV.At 4 HPI gil was slow to become cleaved and the majority of the gil synthesizedduring the pulse remained intact. The 0-glycosylation of gp5O occurred slowlyand incompletely in comparison to the PrV infection in the control parental Lcells at 4 HPI. By contrast, at 6 HPI no detectable gil was cleaved after a 2 hourchase and much less 0-glycosylation of gp5O occurred during the chase thanoccurred at 4 HPI. These data suggested that a cellular component required forinfection became limiting or saturated in gro29 cells as the infection proceeded.These results are consistent with the observation that during infection of gro29cells with HSV-1 the virus was released from gro29 cells early in infection butthis release slowed down drastically as the infection progressed (Fig. 3). Theseresults may also explain why the uninfected gro29 cells can grow well in culture;there may be enough activity of the defective component to support normal cellgrowth but not enough to support viral infection.125A deficiency in the amount of Golgi or a modification of Golgi structurecould also have profound effects on the lifecycle of herpes simplex viruses.Disruption of this organelle could hinder the maturation and egress of theseviruses because they rely heavily on the Golgi apparatus to transport andprocess viral glycoproteins and virions (Fig. 2). The same lesion which affectsthe lectin intoxication pathways described above could also influence therelease of herpes simplex viruses from the gro29 cell. Moreover, a deficiency ofgro29 cells to incorporate GalNAc into macromolecules may be responsible forthe inability of herpes viruses to traverse the secretory pathway; failure of gro29cells to modify a molecule required for the interaction of the virus with thesecretory pathway could result in the observed phenotype.At present, it is not known if the phenotype of gro29 cells resulted from asingle genetic defect, and an effective selection procedure for revertants has notbeen developed. An alternative approach to this problem will be to determinewhether gro29 cells represent a new complementation group by fusion to othertoxin-resistant cell lines (Laurie and Robbins 1991; Stanley 1983b; Stanley 1987;Stanley et al. 1990). This should be possible because cell hybridization studieshave shown that the gro29 phenotype is recessive (Michaelis et al. 1992).Unfortunately, the majority of the well-characterized toxin-resistant cell linesare derived from CHO cells; one of the few cell lines in culture which are nonpermissive to herpes simplex virus infection. Nonetheless, fusion of gro29 tothese cell lines should be useful for determining whether the modeccin andricin resistance phenotypes arose from a single genetic defect. It is alsointeresting to note that the toxin resistance phenotype of gro29 cells arosewithout prior exposure to lectin. The mutation(s) in gro29 cells may impingeon a cellular component common to the pathways of lectin intoxication and126herpes virus maturation and egress. The fact that the toxin resistancephenotype exhibited by gro29 is rare coupled with the block to herpesviruspropagation being unique to gro29 suggests that it is possible that gro29 cellsharbor a defect in a component which facilitates both processes.How did the selection procedure employed to isolate cells deficient inthe propagation of herpes simplex viruses generate gro29? Other mutant celllines have been isolated by this procedure and none of them show the secretiondefect observed in gro29 cells. All of the other cell lines isolated in this mannerhave been defective in the synthesis of heparan sulfate glycosaminoglycanchains (F. Tufaro, unpublished observation). One such cell line termed gro2Chas been characterized in detail (Gruenheid et al. 1992). These cells showedapproximately a 10 fold reduction in the ability to be infected by HSV-1.Regardless of the reduction in infectivity of gro2C these cells were readily killedby the virus. Re-selection of these cells by the same procedure to isolate mutantcell lines which were more resistant to HSV-1 than gro2C yielded cell lines thatfailed to synthesize chondroitin sulfate in addition to heparan sulfate (Banfieldet al. manuscript in preparation). These results suggested that the ability ofcells to survive infection in a mixed population of cells was linked to thebiosynthesis of glycosaminoglycan chains. This was not surprising in that ithad been determined (partially from the characterization of these cell lines)that heparan sulfate can act as a receptor for the binding of HSV-1 to cells, andfurthermore, in the absence of heparan sulfate, chondroitin sulfate can act as areceptor for the virus (Gruenheid et al. 1992; Leduc et al., manuscript inpreparation). The inability of gro29 cells to synthesize chondroitin sulfate andtheir reduced ability to make heparan sulfate may have been enough of a127selective advantage to isolate these cells from a mixed population consistingprimarily of cells able to synthesize glycosaminoglycan chains normally.It may be possible to isolate the defective gene that confers the lectinresistance in gro29 cells by taking advantage of the fact that gro29 cells do notexpress ricin binding molecules abundantly. Pools of cDNA clones from an Lcell expression library can be transfected into gro29 cells and a pool containingpositive clones detected by examination of transfected cells with fluorescentconjugated ricin and epifluorescence microscopy. A positive pool of clones canthen be diluted further until a single clone can confer the ability of gro29 cells tobind ricin. Once isolated, the ability of this cDNA to confer susceptibility toproductive HSV-1-infection, lectin intoxication and to restore chondroitinsulfate synthesis in gro29 cells can be examined. Although this protocol maynot isolate a gene capable of restoring all of the observed defects in these cells itwill be interesting to see if any genetic link between these various phenotypesexist. Regardless of the molecular basis of the gro29 phenotype, it is intriguingthat this cell line is able to survive in culture. It suggests that the mutation ingro29 was leaky or occurred in a gene product whose function was not requiredfor growth in culture medium. It is predicted that it should be possible tointerfere with protein trafficking or GalNAc metabolism such as to leave thecell viable, while at the same time destroying the ability of herpes simplex virusto mount a productive infection.128ReferencesAckermann, M., J. Chou, M. Sarmiento, R. A. Lerner, and B. Roizman.“Identification by antibody to a synthetic peptide of a protein specified by adiploid gene located in the terminal repeats of the L component of the herpessimplex virus genome.” T. Virol. 58 (1986a): 843-850.Ackermann, M., R. Longnecker, B. Roizman, and L. Pereira. “Identification,properties, and gene location of a novel glycoprotein specified by herpes simplexvirus 1.” Virology 150 (1986b): 207-220.Addison, C., F. J. Rixon, J. W. Palfreyman, M. O’Hara, and V. G. Preston.“Characterization of a herpes simplex virus type 1 mutant which has atemperature-sensitive defect in penetration of cells and assembly of capsids.”Virology 138 (1984): 246-259.Addison, C. F., F. J. Rixon, and V. G. Preston. “Herpes simplex virus type 1 UL28gene product is important for the formation of mature capsids.” T. Gen. Virol. 71(1990): 2377-2384.Al-Kobaisi, M. F., F. J. Rixon, and V. G. Preston. “The herpes simplex virus UL33gene product is required for the assembly of full capsids.” Virology 180 (1991):380-388.Baghian, A., L. Huang, S. Newman, S. Jayachandra, and K. G. Kousoulas.“Truncation of the carboxy-terminal 28 amino acids of glycoprotein B specifiedby herpes simplex virus type 1 mutant ambl5ll-7 causes extensive cell fusion.”J. Virol 67 (1993): 2396-2401.Baines, J. D. and B. Roizman. “The UL1O gene of herpes simplex virus 1 encodesa novel viral glycoprotein, gM, which is present in the virion and in the plasmamembrane of infected cells.” J. Virol. 67 (1993): 1441-1452.Baines, J. D., P. L. Ward, G. Campadelli-Fiume, and B. Roizman. “The UL2O geneof herpes simplex virus 1 encodes a function necessary for viral egress.” I. Virol.65 (1991): 6414-6424.Balch, W. E., W. G. Dunphy, W. A. Braell, and J. E. Rothman. “Reconstitution ofthe transport of protein between successive compartments of the Golgimeasured by the coupled incorporation of N-acetylglucosamine.” ll 39 (1984):405-416.Bame, K. J. and JD Esko. “TJndersulfated heparan sulfate in a Chinese hamsterovary cell mutant defective in heparan sulfate N-sulfotransferase.” 1. Biol.Chem. 264 (1989): 8059-8065.129Banfield, B. W. and F. Tufaro. “Herpes simplex virus particles are unable totraverse the secretory pathway in the mouse L cell mutant gro29.” I. Virol 64(1990): 5716-5729.Barker, D. E. and B. Roizman. “The unique sequence of the herpes simplex virus1 L component contains an additional translated open reading frame designatedas UL49.5.” 1. Virol. 66 (1992): 562-566.Baucke, R. B. and P. G. Spear. “Membrane proteins specified by herpes simplexvirus. V. Identification of an Fc-binding glycoprotein.” 1. Virol. 32 (1979): 779-789.Beswick, T. S. L. “The origin and the use of the word herpes.” Med. Hist. 6 (1962):214-232.Brändli, A. W., G. C. Hansson, E. Rodriguez-Boulan, and K. Simons. “Apolarized epithelial cell mutant deficient in the translocation of UDP-galactoseinto the Golgi complex.” J. Biol Chem. 263 (1988): 16283-16290.Brown, S. M., D. H. Ritchie, and J. H. Subak-Sharpe. “Genetic studies withherpes simplex virus type 1: The isolation of temperature-sensitive mutants,their arrangement into complementation groups and recombination analysisleading to a linkage map.” I. Gen. Virol. 18 (1973): 329-346.Buckmaster, E. A., U. Gompels, and A. Minson. “Characterization and physicalmapping of an HSV-1 glycoprotein of approximately 115 X 10 (3) molecularweight.” Virology 139 (1984): 408-413.Bzik, D. J., B. A. Fox, N. A. DeLuca, and S. Person. “Nucleotide sequencespecifying the glycoprotein gene, gB of herpes simplex virus type 1.” Virology133 (1984a): 301-311.Bzik, D. J., B. A. Fox, N. A. DeLucca, and S. Person. “Nucleotide sequence of aregion of the herpes simplex virus type 1 gB glycoprotein gene: mutationsaffecting the rate of virus entry and cell fusion.” Virology 137 (1984b): 185-190.Cai, W., B. Gu, and S. Person. “Role of glycoprotein B of herpes simplex virustype 1 in entry and cell fusion.” T. Virol. 62 (1988a): 2596-2604.Cai, W., S. Person, C. DebRoy, and B. Gu. “Functional regions and structuralfeatures of the gB glycoprotein of herpes simplex virus type 1.” T. Mol. Biol. 201(1988b): 576-588.Cai, W., S. Person, S. C. Warner, J. Zhou, and N. A. Delucca. “Linker-insertionnonsense and restriction site deletion mutations of the gB glycoprotein gene ofherpes simplex virus type 1.” T. Virol. 61(1987): 714-721.130Campadelli-Fiume, G., M. Arsenakis, F. Farabegoli, and B. Roizman. “Entry ofherpes simplex virus I in BJ cells that constitutively express viral glycoprotein Dis by endocytosis and results in degradation of the virus.” I. Virol. 62 (1988a): 159-167.Campadelli-Fiume, G., E. Avitabile, S. Fini, D. Stirpe, M. Arsenakis, and B.Roizman. “Herpes simplex virus glycoprotein D is sufficient to inducespontaneous pH-independent fusion in a cell line that constitutively expressesthe glycoprotein.” Virology 166 (1988b): 598-602.Campadelli-Fiume, G., R. Brandimarti, C. Di Lazzaro, P. L. Ward, B. Roizman,and M. R. Torrisi. “Fragmentation and dispersal of Golgi proteins andredistribution of glycoproteins and glycolipids processed through the Golgiapparatus after infection with herpes simplex virus 1.” Proc. Nati. Acad. Sci.(USA) 90 (1993): 2798-2802.Campadelli-Fiume, G., L. Poletti, F. Dall’olio, and F. Serafini-Cessi. “Infectivityand glycoprotein processing of herpes simplex virus type 1 grown in a ricinresistant cell line deficient in N-acetylglucosaminyl transferase I.” T. Virol. 43(1982): 1061-1071.Campadelli-Fiume, G., S. Qi, E. Avitabile, L. Foa-Tomasi, R. Brandimarti, and B.Roizman. “Glycoprotein D of herpes simplex virus encodes a domain whichprecludes penetration of cells expressing the glycoprotein by superinfectingherpes simplex virus.” T. Virol. 64 (1990): 6070-6079.Chan, W. L. “Protective immunization of mice with specific HSV-1glycoproteins.” Immunology 49 (1983): 343-352.Chatterjee, S. and S. Sarkar. “Studies on endoplasmic reticulum-Golgi complexcycling pathway in herpes simplex virus-infected and brefeldin A-treatedhuman fibroblast cells.” Virology 191 (1992): 327-337.Cheung, P., B. W. Banfield, and F. Tufaro. “Brefeldin A arrests the maturationand egress of herpes simplex virus particles during infection.” T. Virol. 65 (1991):1893-1904.Chou, J. and B. Roizman. “The terminal a sequence of the herpes simplex virusgenome contains the promoter of a gene located in the repeat sequences of the Lcomponent.” T. Virol. 57 (1986): 629-637.Claesson-Welsh, L. and P. G. Spear. “Oligomerization of herpes simplex virusglycoprotein B.” T. Virol. 60 (1986): 803-806.131Cohen, G. H., V. J. Isola, J. Kuhns, P. W. Berman, and R. J. Eisenberg.“Localization of discontinuous epitopes of herpes simplex virus glycoprotein D:use of a non-denaturing (“native”-gel) system of polyacrylamide gelelectrophoresis coupled with Western blotting.” I. Virol. 60 (1986): 157-166.Cohen, G. H., M. Katze, C. Hydrean-Stern, and R. J. Eisenberg. “Type-commonCP-1 antigen of herpes simplex virus is associated with a 59,000-molecular-weight envelope glycoprotein.” T. Virol. 27 (1978): 172-181.Cohen, G. H., D. Long, J. T. Matthews, M. May, and R. J. Eisenberg.“Glycopeptides of the type-common glycoprotein gD of herpes simplex virustypes 1 and 2.” T. Virol 46 (1983): 679-689.Cohen, C. H., M. Ponce de Leon, H. Diggelmann, W. C. Lawrence, S. Vernon,and R. J. Eisenberg. “Structural analysis of the capsid polypeptides of herpessimplex virus types 1 and 2.” T. Virol. 34 (1980): 521-531.Cohen, G. H., W. C. Wilcox, D. L. Sodora, D. Long, J. Z. Levin, and R. J.Eisenberg. “Expression of herpes simplex virus type 1 glycoprotein D deletionmutants in mammalian cells.” J. Virol. 62 (1988): 1932-1940.Colbaugh, P. A., C. Y. Kao, S. P. Shia, M. Stookey, and R. K. Draper. “Three newcomplementation groups of temperature-sensitive Chinese hamster ovary cellmutants defective in the endocytic pathway.” Somatic Cell Mol. Genet. 14 (1988):499-507.Colbaugh, P. A., M. Stookey, and R. K. Draper. “Impaired lysosomes in atemperature-sensitive mutant of Chinese hamster ovary cells.” T. Cell Biol. 108(1989): 2211-2219.Compton, T. and R. J. Courtney. “Virus-specific glycoproteins associated withthe nuclear fraction of herpes simplex virus type 1-infected cells.” T. Virol. 49(1984): 594-597.Dales, S. and Y. Chardonnet. “Early events in the interaction of adenoviruseswith HeLa cells. IV. Association with microtubules and the nuclear porecomplex during vectorial movement of the innoculum.” Virology 56 (1973):465-483.DallOlio, F., N. Malagolini, V. Speziali, G. Campadelli-Fiume, and F. SerafiniCessi. “Sialyated oligosaccharides 0-glycosidically linked to glycoprotein C fromherpes simplex virus type 1.” J. Virol. 56 (1985): 127-134.Darlington, R. W. and L. H. Moss. “Herpesvirus envelopment.” T. Virol. 2 (1968):48-55.132Davidson, H. W. and W. E. Baich. “Differential inhibition of multiple vesiculartransport steps between the endoplasmic reticulum and trans Golgi network.” LBiol. Chem. 268 (1993): 4216-4226.DebRoy, C., N. Pederson, and S. Person. “Nucleotide sequence of a herpessimplex virus type 1 gene that causes cell fusion.” Virology 145 (1985): 36-48.DelGiacco, R. and F. Maley. “Hexosamine metabolism: II. Acid-soluble productsin rat liver following perfusion with D-glucosamine-1-14C.” I. Biol. Chem. 239(1964): 2400-2402.Deutscher, S. L. and C. B. Hirschberg. “Mechanism of galactosylation in the Golgiapparatus. A Chinese hamster ovary cell mutant deficient in translocation ofUDP-galactose across Golgi vesicle membranes.” T. Biol. Chem. 261 (1986): 96-100.Deutscher, S. L., N. Nuwayhid, P. Stanley, E. I. Briles, and C. B. Hirschberg.“Translocation across Golgi vesicle membranes: a CHO glycosylation mutantdeficient in CMP-sialic acid transport.” 39 (1984): 295-299.Doms, R. W., C. Russ, and J. W. Yewdell. “Brefeldin A redistributes resident anditinerant Golgi proteins to the endoplasmic reticulum.” T Cell Biol 109 (1989): 61-72.Dowdle, W. R., A. J. Nahmias, R. W. Harwell, and F. P. Pauls. “Association ofantigenic type of herpes simplex virus hominis with site of viral recovery.” LImmunol. 99 (1967): 974-980.Eisenberg, R. J., D. Long, L. Pereira, B. Hampar, M. Zweig, and G. H. Cohen.“Effect of monoclonal antibodies on limited proteolysis of native glycoprotein Dof herpes simplex virus types 1 and 2 by use of monoclonal antibody.” T. Virol.41 (1982): 478-488.Eisenberg, R. J., D. Long, M. Ponce De Leon, J. T. Matthews, P. G. Spear, M. G.Gibson, L. A. Lasky, P. Berman, E. Golub, and G. H. Cohen. “Localization ofepitopes of herpes simplex virus type 1 glycoprotein D.” T. Virol. 53 (1985): 634-644.Elhammer, A. and S. Kornfeld. “Purification and characterization of UDP-Naetylgalactosaminyltransferase from bovine colostrum and murine lymphomaBW5147 cells.” J. Biol. Chem. 261 (1986): 5249-5255.Elhammer, A., R. A. Poorman, E. Brown, L. L. Maggiora, J. G. Hoogerheide, andF. J. Kezdy. “The specificity of UDP-Ga1NAc:polypeptide Nacetylgalactosaminyltransferase as inferred from a database of in vivo substratesand from the in vitro glycosylation of proteins and peptides.” T. Biol. Chem. 268(1993): 10029-10038.133Forrester, A. J., H. Farrell, G. Wilkinson, J. Kaye, N. Davis-Poynter, and A. C.Minson. “Construction and properties of a mutant herpes simplex virus type 1deleted for glycoprotein H sequences.” 1. Virol. 66 (1992): 341-348.Friedman, D. I., E. R. Olson, C. Georgopoulos, K. Tilly, I. Herskowitz, and F.Banuett. “Interactions of bacteriophage and host macromolecules in the growthof bacteriophage 2..” Micobiol. Rev. 48 (1984a): 299-325.Friedman, H. M., G. H. Cohen, R. J. Eisenberg, C. A. Seidel, and D. B. Cines.“Glycoprotein C of HSV-1 functions as a C3b receptor on infected endothelialcells.” Nature 309 (1984b): 633-635.Fuller, A. 0. and W. C. Lee. “Herpes simplex virus type 1 entry through acascade of virus-cell interactions requires different roles of gD and gH inpenetration.” T. Virol. 66 (1992): 5002-5012.Fuller, A. 0. , R. E. Santos, and P. G. Spear. “Neutralizing antibodies specific forglycoprotein H of herpes simplex virus permit viral attachment to cells butprevent penetration.” T. Virol. 63 (1989): 3435-3443.Furlong, D., H. Swift, and B. Roizman. “Arrangement of herpesvirusdeoxyribonucleic acid in the core.” 1. Virol. 10 (1972): 1071.Gage, P. J., M. Levine, and J. C. Glorioso. “Syncytium-inducing mutationslocalize to two discrete regions within the cytoplasmic domain of herpessimplex type I glycoprotein B.” T. Virol. 67 (1993): 2191-2201.Gamou, S. and N. Shimizu. “Glycosylation of the epidermal growth factorreceptor and its relationship to membrane transport and ligand binding.” LBiochem. 104 (1988): 388-396.Georgopoulou, U., A. Michaelidou, B. Roizman, and P. Mavromara-Nazos.“Identification of a new transcriptional unit and gene product within theunique sequences of the short component of the herpes simplex virus Igenome.” T. Virol. 67 (1993): 3961-3968.Ghosh, P. C. and H. C. Wu. “Enhancement of cytotoxicity of modeccin bynigericin in modeccin-resistant mutant cell lines.” Exp. Cell Res. 174 (1988): 397-410.Gibson, W. and B. Roizman. “Proteins specified by herpes simplex virus. VIII.Characterization and composition of multiple capsid forms of subtypes I and 2.”J. Virol. 10 (1972): 1044-1052.134Glaser, L. “The biosynthesis of N-acetylgalactosamine.” T. Biol. Chem. 234 (1959):2801-2805.Gottlieb, C., J. Baenziger, and S. Kornfeld. “Deficient uridine diphosphate-Nacetylglucosamine:glycoprotein N-acetylglucosaminyltransferase activity in aclone of Chinese hamster ovary cells with altered surface glycoproteins.” I. Biol.Chem. 250 (1975): 3303-3309.Gottlieb, C., S. Kornfeld, and S. Schlesinger. “Restricted replication of twoaiphaviruses in ricin-resistant mouse L cells with altered glycosyltransferaseactivities.” I. Virol. 29 (1979): 344-351.Gruenheid, S. , L. Gatzke, H. Meadows, and F. Tufaro. “Infection andpropagation of herpes simplex virus in a murine L cell mutant defective inheparan sulfate proteoglycan synthesis.” T. Virol. 67 (1992): 93-100.Grüter, W. “Experimentelle und klinische Untersuchungen uber densogenannten Herpes corneae.” Ber. Dtsch. ophthal. Ges. 42 (1920): 162-167.Heilman, C. J., Jr., M. Zwieg, J. R. Stephenson, and B. Hamper. “Isolation of anucleocapsid polypeptide of herpes simplex virus types 1 and 2 possessingimmunologically type-specific and cross-reactive determinants.” I. Virol. 29(1979): 34-42.Heine, J. W., R.W. Honess, E Cassai N., and B. Roizman. “Proteins specified byherpes simplex virus. XIII. The virion polypeptides of type 1 strains.” I. Virol. 14(1974): 649-651.Heine, J. W., P. G. Spear, and B. Roizman. “Proteins specified by herpes simplexvirus. VI. Viral proteins in the plasma membrane.” T. Virol. 9 (1972): 431-439.Herold, B.C., D. WuDunn, N. Soltys, and P. G. Spear. “Glycoprotein C of herpessimplex virus type I plays a principal role in the adsorption of virus to cells andin infectivity.” T. Virol. 65 (1991): 1090-1098.Hiebsch, R. R. and B. W. Wattenburg. “Vesicle fusion in protein transportthrough the Golgi in vitro does not involve long- lived prefusionintermediates. A reassessment of the kinetics of protein transport as measuredby glycosylation.” Biochem. 31 (1992): 6111-6118.Highlander, S. L., W. Cai, S. Person, M. Levine, and J. C. Glorioso. “Monoclonalantibodies define a domain on herpes simplex virus glycoprotein B involved invirus penetration.” T. Virol. 62 (1988): 1881-1888.135Highlander, S. L., S. L. Sutherland, P. J. Gage, D. C. Johnson, M. Levine, and J. C.Glorioso. “Neutralizing antibodies specific for herpes simplex virus glycoproteinD inhibit virus penetration.” 1. Virol. 61 (1987): 3356-3364.Hirschberg, C. B. “Topography of glycosylation in the rough endoplasmicreticulum and Golgi apparatus.” Ann. Rev. Biochem. 56 (1987): 63-87.Hoggen, M. D. and B. Roizman. “The isolation and properties of a variant ofherpes simplex producing multinucleated giant cells in monolayer cultures inthe presence of antibody.” Am. T. Hyg. 70 (1959): 208-219.Homa, F., T. Hollander, D. J. Lehman, D. R. Thomsen, and A. P. Elhammer.“Isolation and expression of a cDNA clone encoding a bovine UDPGalNAc:polypeptide N-acetylgalactosaminyltransferase.” I. Biol. Chem. 17 (1993):12609-12616.Honess, R. W. and B. Roizman. “Proteins specified by herpes simplex virus. XI.Identification and relative molar rates of synthesis of structural andnonstructural herpes virus polypeptides in the infected cell.” J. Virol. 12 (1973):1347-1365.Hope, R. G., J. Palfreyman, M. Suh, and H. S. Marsden. “Suiphated glycoproteinsinduced by herpes simplex virus.” J. Gen. Virol. 58 (1982): 399-415.Hutchinson, L., H. Browne, V. Wargent, N. Davis-Poynter, S. Primorac, K.Goldsmith, A. C. Minson, and D. C. Johnson. “A novel herpes simplex virusglycoprotein, gL, forms a complex with glycoprotein H (gH) and affects normalfolding and surface expression of gH.” T. Virol. 66 (1992a): 5603-5609.Hutchinson, L., K. Goldsmith, D. Snoddy, H. Ghosh, F. L. Graham, and D. C.Johnson. “Identification and characterization of a novel herpes simplex virusglycoprotein, gK, involved in cell fusion.” J. Virol. 66 (1992b): 5603-5609.Isola, V. J., R. J. Eisenberg, G. R. Siebert, C. J. Heilman, W. C. Wilcox, and G. H.Cohen. “Fine mapping of antigenic site II of herpes simplex virus glycoproteinD.” J. Virol. 63 (1989): 2325-2334.Johnson, D. C., R. L. Burke, and T. Gregory. “Soluble forms of herpes simplexvirus glycoprotein D bind to a limited number of cell surface receptors andinhibit virus entry into cells.” T. Virol. 64 (1990): 2569-2576.Johnson, D. C. and V. Feenstra. “Identification of a novel herpes simplex virustype 1-induced glycoprotein which complexes with gE and bindsimmunoglobulin.” J. Virol. 61 (1987): 2208-2216.136Johnson, D. C., M. C. Frame, M. W. Ligas, A. M. Cross, and N. D. Stow. “Herpessimplex virus immunoglobulin G Fc receptor activity depends on a complex oftwo viral glycoproteins, gE and gI.” I. Virol. 62 (1988): 1347-1354.Johnson, D. C. and M. W. Ligas. “Herpes simplex viruses lacking glycoprotein Dare unable to inhibit virus penetration: quantitative evidence for virus-specificcell surface receptors.” I. Virol. 62 (1988): 4605-4612.Johnson, D.C. and M. J. Schlesinger. “Vesicular stomatitis virus and Sindbisvirus glycoprotein transport to the cell surface is inhibited by ionophores.”Virology 103 (1980): 407-424.Johnson, D. C. and P. G. Spear. “Monensin inhibits the processing of herpessimplex virus glycoproteins, their transport to the cell surface, and the egress ofvirions from infected cells.” J. Virol. 43 (1982): 1102-1112.Johnson, D. C. and P. G. Spear. “0-linked oligosaccharides are acquired by herpessimplex virus glycoproteins in the Golgi apparatus.” 32 (1983): 987-997.Jones, F. and C. Grose. “Role of cytoplasmic vacuoles in varicella-zoster virusglycoprotein trafficking and virion envelopment.” T. Virol. 62 (1988): 2701-2711.Karger, A. and T.C. Mettenleiter. “Glycoproteins gill and gp5O play dominantroles in the biphasic attachment of pseudorabies virus.” Virology 194 (1993): 654-664.Kato, S., S. Ito, T. Noguchi, and H. Naito. “Effects of brefeldin A on the synthesisand secretion of egg white proteins in primary cultured oviduct cells of layingJapanese quail (Coturnix coturnix japonica).” Biochim. Biophys. Acta 991 (1989):36-43.Kato, S., T. Noguchi, and H. Naito. “Secretion of egg white proteins in primarycultured oviduct cells of laying Japanese quail (Coturnix coturnix japonica).”Poult. Sci. 66 (1987): 1208-1216.Kingsley, D.M., K. F. Kozarsky, L. Hobbie, and M. Krieger. “Reversible defects in0-linked glycosylation and LDL receptor expression in a UDP-Gal/UDP-Ga1NAc4-epimerase deficient mutant.” 44 (1986): 749-759.Kjellen, L. and U. Lindahi. “Proteoglycans: structures and interactions.” Annu.Rev. Biochem. 60 (1991): 443-475.Kornfeld, R. and S. Kornfeld. “Assembly of asparagine-linked oligosaccharides.”Annu. Rev. Biochem. 54 (1985): 631-664.137Koyana, A. H. and T. Uchida. “Inhibition of multiplication of herpes simplexvirus type 1 by ammonium chloride and chloroquine.” Virology 138 (1984): 332-335.Kwong, A. D. and N. Frenkel. “Herpes simplex virus-infected cells contain afunction(s) that destabilizes both host and viral mRNAs.” Proc. Nati. Acad. Sci.(USA) 84 (1987): 1926-1930.Laemmli, U. K. “Cleavage of structural proteins during the assembly of the headof bacteriophage T4.” Nature 227 (1970): 680-685.Laurie, S. M. and A. R. Robbins. “A toxin-resistant mouse L-cell mutantdefective in protein transport along the secretory pathway.” T. Cell. Physiol. 147(1991): 215-223.Ligas, M.W. and D.C. Johnson. “A herpes simplex virus mutant in whichglycoprotein D sequences are replaced by 13-galactosidase sequences binds to but isunable to penetrate into cells.” T. Virol. 62 (1988): 1486-1494.Lippencott-Schwartz, J., L. Yaun, C. Tipper, M. Amherdt, L. Orci, and R. D.Klausner. “Brefeldin A’s effects on endosomes, lysosomes, and the TGN suggesta general mechanism for regulating organelle structure and membrane traffic.”Cell 67 (1991): 601-616.Lippincott-Schwartz, J., J. G. Donaldson, A. Schweizer, E. G. Berger, H. P. Hauri,L. C. Yuan, and R. D. Klausner. “Microtubule-dependent retrograde transport ofproteins into the ER in the presence of brefeldin A suggests an ER recyclingpathway.” 60 (1990): 821-836.Lippincott-Schwartz, J., L. C. Yuan, J. S. Bonifacino, and R. D. Klausner. “Rapidredistribution of Golgi proteins into the ER in cells treated with brefeldin A:evidence for membrane cycling from Golgi to ER.” çll 56 (1989): 801-813.Liu, F. and B. Roizman. “The herpes simplex virus 1 gene encoding a proteasealso contains within its coding domain the gene encoding the more abundantsubstrate.” J. Virol. 65 (1991a): 5149-5156.Liu, F. and B. Roizman. “The promoter, transcriptional unit, and codingsequence of herpes simplex virus 1 family 35 proteins are contained within andin frame with the UL26 open reading frame.” T. Virol. 65 (1991b): 206-212.Long, D., G. H. Cohen, M. I. Muggeridge, and R. J. Eisenberg. “Cysteine mutantsof herpes simplex virus type 1 glycoprotein D exhibit temperature-sensitiveproperties in structure and function.” I. Virol. 64 (1990): 5542-5552.138Long, D., T. J. Madara, M. Ponce De Leon, G. H. Cohen, P. C. Montgomery, and R.J. Eisenberg. “Glycoprotein D protects mice against lethal challenge with herpessimplex virus types I and 2.” Infect. Immun. 37 (1984): 761-764.Longnecker, R., S. Chatterjee, R. J. Whitley, and B. Roizman. “Identification of aherpes simplex virus I glycoprotein gene within a gene cluster dispensable forgrowth in cell culture.” Proc Natl Acad Sci (USA) 84 (1987): 4303-4307.Longnecker, R. and B. Roizman. “Clustering of genes dispensable for growth inculture in the S component of the HSV-1 genome.” Science 236 (1987): 573-576.Luria, S. E. and M. Delbruck. “Mutations of bacteria from virus sensitivity tovirus resistance.” Genetics 28 (1943): 491-511.MacLean, C. A., S. Efstathiou, M. L. Elliott, F. E. Jamieson, and D. J. McGeoch.“Investigation of herpes simplex virus type 1 genes encoding multiply insertedmembrane proteins.” I. Gen. Virol. 72 (1991): 897-906.Maley, F. and H. A. Lardy. “Formation of UDPG1a and related compounds by thesoluble fraction of rat liver.” Science 124 (1956): 1207-1208.Maley, F. and G. F. Maley. “The enzymic conversion of glucosamine togalactosamine.” Biochem. Biophvs. Acta 31(1959): 577-578.Maley, F., A. L. Tarentino, J. F. McGarrahan, and R. DelGiacco. “The metabolismof D-galactosamine and N-acetyl-D-galactosamine in rat liver.” Biochem. T. 107(1968): 637-644.Martin, S., B. Moss, P.W. Berman, L. A. Lasky, and B. T. Rouse. “Mechanisms ofantiviral immunity induced by a vaccinia recombinant expressing herpessimplex virus type 1 glycoprotein D: cytotoxic T cells.” T. Virol. 61 (1987): 726-734.Matthews, J. T., G. H. Cohen, and R. J. Eisenberg. “Synthesis and processing ofglycoprotein D of herpes simplex virus types 1 and 2 in an in vitro system.” IVirol. 48 (1983): 521-533.McGarrahan, J. F. and F. Maley. “Hexosamine metabolism: I. The metabolism invivo and in vitro of D-glucosamine -1-C14 and N-acetyl-D-glucosamine-1-C14 inrat liver.” I. Biol. Chem. 237 (1962): 2458-2465.McGeoch, D. J., C. Cunningham, G. McIntyre, and A. Dolan. “Comparativesequence analysis of the long repeat regions and adjoining parts of the longunique regions in the genomes of herpes simplex viruses 1 and 2.” T. Gen. Virol.72 (1991): 3057-3075.139McGeoch, D. J., M. A. Dairymple, A. J. Davison, A. Dolan, M. C. Frame, D.McNabb, L. J. Perry, J. E. Scott, and P. Taylor. “The complete DNA sequence ofthe long unique region in the genome of herpes simplex virus type 1.” 1. Gen.Vir. 69 (1988): 1531-1574.McGeoch, D. J., A. Dolan, S. Donald, and F. J. Rixon. “Sequence determinationand genetic content of the short unique region in the genome of herpes simplexvirus type-i.” T. Mol. Biol. 181 (1985): 1-13.McGeoch, D. J., A. Dolan, and M. C. Frame. “DNA sequence of the region in thegenome of herpes simplex virus type I containing the exonuclease gene andneighboring genes.” Nucleic Acids Res. 14 (1986): 3435-3448.Michaelis, C., B. W. Banfield, S. Gruenheid, Y. Tsang, R. Lippe, W. A. Jefferies, B.W. Wattenburg, and F. Tufaro. “Toxin resistance and reduced secretion in amouse L-cell mutant defective in herpes virus propagation.” Biochem. Cell Biol.70 (1992): 1209-1217.Mindel, A. Herpes Simplex Virus. I ed., The Bloomsbury series in clinicalscience., ed. J. Tinker. London: Spinger-Verlag, 1989.Misra, V. and E. L. Blewett. “Construction of herpes simplex viruses that arepseudodiploid for the glycoprotein B gene: a strategy for studying the function ofan essential herpes virus gene.” T. Gen. Virol. 72 (1991): 385-392.Misumi, Y., K. Miki, A. Takatsuki, G. Tamura, and Y. Ikehara. “Novel blockadeby brefeldin A of intracellular transport of secretory proteins in cultured rathepatocytes.” J. Biol. Chem. 261 (1986): 11398-11403.Morgan, C., H. M. Rose, M. Holden, and E. P. Jones. “Electron microscopicobservations on the development of herpes simplex virus.” T. Exp. Med. 110(1959): 643-656.Muggeridge, M. I., V. J. Isola, R. A. Byrn, T Tucker J., A. C. Minson, J. C. Glorioso,G. H. Cohen, and R. J. Eisenberg. “Antigenic analysis of a major neutralizationsite of herpes simplex virus glycoprotein D, using deletion mutants andmonoclonal antibody resistant mutants.” I. Virol. 62 (1988): 3274-3280.Muggeridge, M. I., W. C. Wilcox, G. H. Cohen, and R. J. Eisenberg. “Identificationof a site on herpes simplex virus type I glycoprotein D essential for infectivity.”J. Virol. 64 (1990): 3617-3626.Nii, S. ,C. Morgan, and H. M. Rose. “Electron microscopy of herpes simplexvirus. II. Sequence of development.” T. Virol. 2 (1969): 517-536.140Nuchtern, J. G., J. S. Bonifacino, W. E. Biddison, and R. D. Klausner. “BrefeldinA implicates egress from endoplasmic reticulum in class I restricted antigenpresentation.” Nature 339 (1989): 223-226.Oda, K., S. Hirose, N. Takami, Y. Misumi, A. Takatsuki, and Y. Ikehara.“Brefeldin A arrests the intracellular transport of a precursor of complement C3before its conversion site in rat hepatocytes.” FEBS. Lett. 214 (1987): 135-138.Olofsson, S., J. Blomberg, and E. Lycke. “O-glycosidic carbohydrate-peptidelinkages of herpes simplex virus glycoproteins.” Arch. Virol. 70 (1981): 321-329.Olofsson, S., M. Milla, C. Hirschberg, Clercq E. De, and R. Datema. “Inhibition ofterminal N- and 0-glycosylation specific for herpesvirus-infected cells:mechanism of an inhibitor of sugar nucleotide transport across Golgimembranes.” Virology 166 (1988): 440-450.Olsnes, S., K. Sandvig, K. Eiklid, and A. Pihl. “Properties and action mechanismof the toxic lectin modeccin: interaction with cell lines resistant to modeccin,abrin, and ricin.” J. Supramol. Struct. 9 (1978): 15-25.Omura, T. and S. Takesue. “A new method for simultaneous purification ofcytochrome b5 and NADPH-cytochrome c reductase from rat liver microsomes.”I. Biochem. 67 (1970): 249-257.Para, M. F., M. L. Parish, A. G. Noble, and P. G. Spear. “Potent neutralizingactivity associated with anti-glycoprotein D specificity among monoclonalantibodies selected for binding to herpes simplex virions.” T. Virol. 55 (1985): 483-488.Para, M. F., K. M. Zezulak, A. J. Conley, M. Weinberger, K. Snitzer, and P. G.Spear. “Use of monoclonal antibodies against two 75,000 molecular-weightglycoproteins specified by herpes simplex virus type 2 in glycoproteinidentification and gene mapping.” T. Virol. 45 (1983): 1223-1227.Peake, M. L., P. Nystrom, and L. I. Pizer. “Herpesvirus glycoprotein synthesisand insertion into plasma membranes.” T. Virol. 42 (1982): 678-690.Peiham, H.R.B. “Multiple targets for brefeldin A.” II 67 (1991): 449-451.Perez, M. and C. B. Hirschberg. “Translocation of UDP-N-acetylglucosamine intovesicles derived from rat liver rough endoplasmic reticulum and Golgiapparatus.” I. Biol. Chem. 260 (1985): 4671-4678.Perkel, V. S., A. Y. Liu, Y. Miura, and J. A. Magner. “The effects of brefeldin-A onthe high mannose oligosaccharides of mouse thyrotropin, free alpha-subunits,and total glycoproteins.” Endocrinology 123 (1988): 310-318.141Perkel, V. S., Y. Miura, and 1. A. Magner. “Brefeldin A inhibits oligosaccharideprocessing of glycoproteins in mouse hypothyroid pituitary tissue at severalsubcellular sites.” Proc. Soc. Exp. Biol. Med. 190 (1989): 286-293.Person, S., R. W. Knowles, G. S. Read, S. C. Warner, and V. C. Bond. “Kinetics ofcell fusion induced by a syncytia-producing mutant of herpes simplex virus type1.” J. Virol. 17 (1976): 183-190.Petrovskis, E. A., J. C. Timmins, M. A. Armentrout, C. C. Marchiolli, R. J. YanceyJr., and L. E. Post. “DNA sequence of the gene for pseudorabies virus gp5O, aglycoprotein without N-linked glycosylation.” T. Virol. 59 (1986): 216-223.Piller, F., M. H. Hanlon, and R. L. Hill. “Co-purification and characterization ofUDP-glucose 4-epimerase and UDP-N-acetylglucosamine 4-epimerase fromporcine submaxillary glands.” T. Biol. Chem. 258 (10 1983): 10774-10778.Pizer, L. I., G. H. Cohen, and R. J. Eisenberg. “Effect of tunicamycin on herpessimplex virus glycoproteins and infectious virus production.” I. Virol. 34 (1980):142-153.Pogue-Guile, K. L. and P. C. Spear. “The single base pair substitution responsiblefor the syn phenotype of herpes simplex virus type 1, strain MP.” Virology 157(1987): 67-74.Pressman, B. C. “Biological applications of ionophores.” Annu. Rev. Biochem.45 (1976): 501-530.Preston, V. G., J. A. Coates, and F. J. Rixon. “Identification and characterizationof a herpes simplex virus gene product required for encapsidation of virusDNA.” J. Virol. 45 (1983): 1056-1064.Quinn, T. C. “Epidemiologic and serologic evidence for a role of herpesvirusesin HIV infection.” In Herpesviruses. the immune system, and AIDS., ed. L.Aurelian. 1-20. 1 ed., Vol. Boston/Dordrecht/London: Kiuwer AcademicPublishers, 1990.Randerath, K. and E. Randerath. “Ion-exchange thin layer chromatography. XIV.Separation of nucleotide sugars and nucleoside monophosphates on PETcellulose.” Anal. Biochem. 13 (1965): 575-579.Read, G. S., S. Person, and P. M. Keller. “Genetic studies of cell fusion induced byherpes simplex virus type a.” I. Virol. 35 (1980): 105-113.Richmond, J. E. “Role of sugar nucleotides in the incorporation of sugars intoglycoproteins.” Biochemistry 4 (1965): 1834-1838.142Rixon, F. J., M. D. Davison, and A. J. Davison. “Identification of the genesencoding two capsid proteins of herpes simplex virus type I by direct amino acidsequencing.” J. Gen. Virol. 71 (1990): 1211-1214.Robbins, A. R., C. Oliver, J. L. Bateman, S. S. Krag, C. J. Galloway, and I.Mellman. “A single mutation in Chinese hamster ovary cells impairs bothGolgi and endosomal functions.” T. Cell Biol. 99 (1984): 1296-1308.Robbins, A. R., S. S. Peng, and J. L. Marshall. “Mutant Chinese hamster ovarycells pleiotropically defective in receptor mediated endocytosis.” T. Cell Biol. 96(1983): 1064-1071.Rodriguez, M. and M. Dubois-Dalcq. “Intramembrane changes occurring duringmaturation of herpes simplex virus type 1: freeze fracture study.” J. Virol. 26(1978): 435-447.Roff, C. F., R. Fuchs, I. Meliman, and A.R. Robbins. “Chinese hamster ovary cellmutants with temperature sensitive defects in endocytosis. I. Loss of function onswitching to the nonpermissive temperature.” T. Cell Biol. 103 (1986): 2283-2297.Roizman, B. and W. Batterson. “Herpesviruses and their replication.” InFundamental Virology, ed. B. N. Fields and D. M. Knipe. 607-636. 1 ed., Vol. 1.New York: Raven Press, 1986.Roop, C., L. Hutchinson, and D. C. Johnson. “A mutant herpes simplex virustype 1 unable to express glycoprotein L cannot enter cells, and its particles lackglycoprotein H.” 1. Virol. 67 (1993): 2285-2297.Roseman, S. “The synthesis of complex carbohydrates bymultiglycosyltransferase systems and their potential function in intercellularadhesion.” Chem. Phys. Lipids 5 (1970): 270-297.Saraste, J. and E. Kuismanen. “Pre- and post-Golgi Vacuoles operate in thetransport of Semliki Forest virus membrane glycoproteins to the cell surface.”Cell 38 (1984): 535-549.Saraste, J., G. E. Palade, and M. G. Farquhar. “Temperature-sensitive steps in thetransport of secretory proteins through the Golgi complex in exocrine pancreaticcells.” Proc. Nati. Acad. Sci. USA 83 (17 1986): 6425-9.Sears, A. E., B. S. McGwire, and B. Roizman. “Infection of polarized MDCK cellswith herpes simplex virus 1: two asymmetrically distributed cell receptorsinteract with different viral proteins.” Proc. Nati. Acad. Sci. USA 88 (1991): 5087-5091.143Serafini-Cessi, F. and G. Campadelli-Fiume. “Studies on benzhydrazone, aspecific inhibitor of herpesvirus glycoprotein synthesis: Size distribution ofglycopeptides and endo-b-N-acetylglucosaminidase H treatment.” Arch. Virol 70(1981): 331-343.Serafini-Cessi, F., F. Dall’Olio, N. Malagolini, L. Pereira, and C. CampadelliFiume. “Comparative study on 0-linked oligosaccharides of glycoprotein D ofherpes simplex virus types 1 and 2.” 1. Gen. Virol. 69 (1988): 869-877.Serafini-Cessi, F., F. Dall’Olio, M. Scannavini, and C. Campadelli-Fiume.“Processing of herpes simplex virus-I glycans in cells defective in glycosyltransferases of the Golgi system: relationship to cell fusion and virion egress.”Virology 131 (1983): 59-70.Sherman, G. and S. L. Bachenheimer. “DNA processing in temperature-sensitive morphogenic mutants of HSV-1 .“ Virology 158 (1987): 427-430.Sherman, G. and S. L. Bachenheimer. “Characterization of intranuclear capsidsmade by ts morphogenic mutants of HSV-1.” Virology 163 (1988): 471-480.Shieh, M. T., D. WuDunn, R. I. Montgomery, J. Esko, and P. G. Spear. “Cellsurface receptors for herpes simplex virus are heparan sulfate proteoglycans.” LCell Biol. 116 (1992): 1273-1281.Showalter, S. D., M. Zweig, and B. Hampar. “Monoclonal antibodies to herpessimplex virus type 1 proteins, including the immediate-early protein ICP4.”Infect. Immun. 34 (1981): 684-692.Smibert, C. A. and J. R. Smiley. “Differential regulation of endogenous andtransduced B-globin genes during infection of erythroid cells with a herpessimplex virus type I recombinant.” T. Virol. 64 (1990): 3882-3894.Sodora, D. L., G. H. Cohen, and R. J. Eisenberg. “Influence of asparagine-linkedoligosaccharides on antigenicity, processing, and cell surface expression ofherpes simplex virus type I glycoprotein D.” T. Virol. 63 (1989): 5184-5193.Sodora, D. L., G. H. Cohen, M. I. Muggeridge, and R. J. Eisenberg. “Absence ofasparagine-linked oligosaccharides from glycoprotein D of herpes simplex virustype 1 results in a structurally altered but biologically active protein.” T. Virol. 65(1991a): 4424-4431.Sodora, D. L., R. J. Eisenberg, and G. H. Cohen. “Characterization of arecombinant herpes simplex virus which expresses glycoprotein D lackingasparagine-linked oligosaccharides.” T. Virol. 65 (1991b): 4432-4441.144Sommers, L. W. and C. B. Hirschberg. “Transport of sugar nucleotides into ratliver Golgi.” T. Biol. Chem. 257 (1982): 10811-10817.Spear, P. G. “Membrane proteins specified by herpes simplex viruses. I.Identification of four glycoprotein precursors and their products in type 1-infected cells.” I. Virol. 17 (1976): 991-1008.Spear, P. G. “Glycoproteins specified by herpes simplex viruses.” In Thherpesviruses, ed. B. Roizman. 315-356. vol. 3. New York: Plenum PublishingCorp, 1985.Spear, P. G. “Entry of alphaherpesviruses into cells.” Sem. Virol. 4 (1993): 167-180.Spear, P. G. and B. Roizman. “Proteins specified by herpes simplex virus. V.Purification and structural proteins of the herpesvirion.” I. Virol. 9 (1972): 143-159.Stanley, P. “Selection of specific wheat germ agglutinin resistant (WGAR)phenotypes from Chinese hamster ovary cell populations containing numerouslecR genotypes.” Mol. Cell Biol. 1 (1981): 687-696.Stanley, P. “Lectin-resistant CHO cells: selection of new mutant phenotypes.”Somat. Cell Genet. 9 (1983a): 593-608.Stanley, P. “Selection of lectin-resistant mutants of animal cells.” MethodsEnzymol. 96 (1983b): 157-184.Stanley, P. “Glycosylation mutants of animal cells.” Annu. Rev. Genet. 18 (1984):525-552.Stanley, P. “Biochemical characterization of animal cell glycosylation mutants.”Methods Enzymol. 138 (1987): 443-458.Stanley, P., S. Sallustio, S. S. Krag, and B. Dunn. “Lectin-resistant CHO cells:selection of seven new mutants resistant to ricin.” Somat. Cell Mol. Genet. 16(1990): 211-223.Strous, G. J., P. van Kerkhof, G. van Meer, S. Rijnboutt, and W. Stoorvogel.“Differential effects of brefeldin A on transport of secretory and lysosomalproteins.” J. Biol. Chem. 268 (1993): 2341-2347.Tartakoff, A. M. and P. Vassalli. “Plasma cell immunoglobulin secretion arrest isaccompanied by alterations of the Golgi complex.” I. Exp. Med. 146 (1977): 1332-1345.145Tartakoff, A. M. and P. Vassalli. “Comparative studies of intracellular transportof secretory proteins.” T. Cell. Biol. 79 (1978): 694-707.Torseth, J. W., G. H. Cohen, R. J. Eisenberg, P. W. Berman, L. A. Lasky, C. P.Cerini, C. J. Heilman, S. Kerwar, and T. C. Merigan. “Native and recombinantherpes simplex virus type 1 envelope proteins induce human immune Tlymphocyte responses.” T. Virol. 61 (1987): 1532-1539.Towbin, H., T. Staehelin, and J. Gordon. “Electrophoretic transfer of proteinsfrom polyacrylamide gels to nitrocellulose sheets: Procedure and someapplications.” Proc Natl. Acad. Sci. USA 76 (1979): 4350-4354.Tufaro, F., M. D. Snider, and S. L. McKnight. “Identification and characterizationof a mouse cell mutant defective in the intracellular transport of glycoproteins.”J. Cell Biol. 105 (1987): 647-657.Urbani, L. and R. D. Simoni. “Cholesterol and vesicular stomatitis virus Gprotein take separate routes from the endoplasmic reticulum to the plasmamembrane.” J. Biol. Chem. 265 (1990): 1919-1923.Vischer, P. and R. C. Hughes. “Glycosyl transferases of baby-hamster-kidney(BHK) cells and ricin-resistant mutants. N-glycan biosynthesis.” Eur. T.Biochem. 117 (1981): 275-284.Wadsworth, S., R. J. Jacob, and B. Roizman. “Anatomy of herpes simplex virusDNA. IL Size, composition, and arrangement of inverted terminal repetitions.”J. Virol. 15 (1975): 1487-1497.Wasteson, A., K. Uthne, and B. Westermark. A novel assay for thebiosynthesis of sulphated polysaccharide and its application to studies on theeffects of somotomedin on cultured cells. Biochem. J. 136 (1973): 1069-1073.Weber, P. C., M. Levine, and J. C. Glorioso. “Rapid identification of nonessentialgenes of herpes simplex virus type 1 by Tn5 mutagenesis.” Science 236 (1987):576-579.Weinheimer, S. P., P. J. McCann ifi, D. R. O’Boyle II, J. T. Stevens, B. A. Boyd, D.A. Drier, G. A. Yamanaka, C. L. Dilanni, I.C. Deckman, and M. G. Cordingly.“Autoproteolysis, of herpes simplex type I protease releases an active catalyticdomain found in intermediate capsid particles.” I. Virol. 67 (1993): 5813-5822.Wenske, E. A., M. W. Bratton, and R. J. Courtney. “Endo-beta-Nacetylglucosaminidase H sensitivity of precursors to herpes simplex virus type 1glycoproteins gB and gC.” T. Virol. 44 (1982): 241-248.146Whealy, M. E., J. P. Card, R. P. Meade, A. K. Robbins, and L. W. Enquist. “Effectof Brefeldin A on alphaherpesvirus membrane protein glycosylation and virusegress.” I. Virol. 65 (1991): 1066-1081.Whealy, M. E., A. K. Robbins, and L. W. Enquist. “The export pathway of thepseudorabies virus gB homolog gil involves oligomer formation in theendoplasmic reticulum and protease processing in the Golgi apparatus.” T. Virol.64 (1990): 1946-1955.Whealy, M. E., A. K. Robbins, F. Tufaro, and L. W. Enquist. “A cellular functionis required for Pseudorabies virus envelope glycoprotein processing and virusegress.” T. Virol. 66 (1992): 3803-3810.Wilcox, W. C., D. Long, D. L. Sodora, R. J. Eisenberg, and G. H. Cohen. “Thecontribution of cysteine residues to antigenicity and extent of processing ofherpes simplex virus type 1 glycoprotein D.” T. Virol. 62 (1988): 1941-1947.Wildy, P., W. C. Russell, and R. W. Home. “The morphology of herpes virus.”Virology 12 (1960): 1044-1052.Wittels, M. and P. G. Spear. “Penetration of cells by herpes simplex virus doesnot require a low pH-dependent endocytic pathway.” Virus Res. 18 (1991): 271-290.WuDunn, D. and P. G. Spear. “Initial interaction of herpes simplex virus withcells is binding to heparan sulfate.” T. Virol. 63 (1989): 52-58.Yao, F. and R. J. Courtney. “A major transcriptional regulatory protein (ICP4) ofherpes simplex virus type 1 is associated with purified virions.” T. Virol. 63(1989): 3338-3344.Yoshida, T., C. C. Chen, M. S. Zhang, and H. C. Wu. “Disruption of the Golgiapparatus by brefeldin A inhibits the cytotoxicity of ricin, modeccin, andPseudomonas toxin.” Exp. Cell. Res. 192 (1991): 389-395.Yoshida, T., C. H. Chen, M. S. Zhang, and H. C. Wu. “Increased cytotoxicity ofricin in a putative Golgi-defective mutant of Chinese hamster ovary cell.” Exp.Cell Res. 190 (1990): 11-16.Youle, R. J. and M. Colombatti. “Hybridoma cells containing intracellular antiricin antibodies show ricin meets secretory antibody before entering the cytosol.”I. Biol. Chem. 262 (1987): 4676-4682.Zarling, J. M., PA Moran, LB Lasky, and B Moss. “Herpes simplex virus (HSV)specific human T cell clones recognize HSV glycoprotein D expressed by arecombinant vaccinia virus.” J. Virol. 59 (1986): 506-509.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items