Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effect of adenovirus E3/19K protein on cellular processes in the endoplasmic reticulum Lomas, Cyprien 1999

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

Item Metadata


831-ubc_1999-463745.pdf [ 11.41MB ]
JSON: 831-1.0089293.json
JSON-LD: 831-1.0089293-ld.json
RDF/XML (Pretty): 831-1.0089293-rdf.xml
RDF/JSON: 831-1.0089293-rdf.json
Turtle: 831-1.0089293-turtle.txt
N-Triples: 831-1.0089293-rdf-ntriples.txt
Original Record: 831-1.0089293-source.json
Full Text

Full Text

E F F E C T O F A D E N O V I R U S E 3 / 1 9 K P R O T E E S f O N C E L L U L A R P R O C E S S E S I N T H E E N D O P L A S M I C R E T I C U L U M by C Y P R J E N L O M A S B . S c , The Un ive r s i t y o f B r i t i s h C o l u m b i a , 1987 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S Facul ty o f Science Department o f Z o o l o g y W e accept this this as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A M a y 1999 © C y p r i e n Pierre-Etienne L o m a s , 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. 1 further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. • Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract The human adenovirus has developed several methods to evade the immune system. One mechanism by which it accomplishes this involves the endoplasmic reticulum (ER) retained adenovirus E3/19K protein. This protein interferes with antigen presentation by binding and retaining Major Histocompatibility Complex Class I (MHC CI I) proteins in the ER. E3/19K binds all human and all but one mouse MHC CI I molecules tested to date. Differences in mouse MHC CI I sequences were exploited to determine the structures involved in binding. Human 293 cells transfected with the mouse H-2 alleles Kd, Kb, Kk, Dd, Db, L d were infected with adenovirus 2. It was found that MHC CI I alleles could be grouped into three categories. The H-2 allelic proteins K d and K b were found to be binders; K k and D d non binders and D b and L d slow binders. Examination of a cell line transfected with the slow-binding H-2Db protein revealed that D b is expressed at a reduced level at the cell surface. To determine the cause of this, cells were subjected to conditions previously used to restore defective cell surface expression of MHC CI I including culture at reduced temperature, addition of excess P 2m and exposure to gamma interferon. All these methods were unsuccessful in increasing cell surface expression of Db. Rather than being due to a missing co-factor or unstable conformation, the accumulation of the allelic proteins in the ER in this transfectant was due to an undetermined mechanism. Because E3/19K binding quickly stabilised a mature MHC CI I conformation in the presence of tunicamycin it was suggested that it bound MHC CI I like a chaperone. It was found that ii E3/19K binding to MHC CI I did not block the association of endogenous ER resident chaperones calnexin and TAP. Peptide binding to the MHC CI I-E3/19K complex could also occur. These experiments showed that E3/19K did not associate with MHC CI I through the peptide binding groove and did not disrupt the interaction between the MHC CI I and the processing machinery found in the ER, namely calnexin and TAP, and therefore does not retain MHC molecules by making them conformationally immature. This study shows E3/19K works independently of chaperones. E3/19K may be a tool to trap MHC CI I - chaperone complexes at a specific point in their maturation. ni Table of Contents Abstract 1 1 List of Tables v i List of Figures y i i Nomenclature x Acknowledgements X 1 1. Introduction 1 1.1. Immunity • 1 1.2. Pluripotent Stem Cells 2 1.3. B Cells 2 1.4. T Cells 2 1.5. MHC 4 1.6. Cellular Mechanics 11 1.6.1. Peptides U 1.6.2. Proteasome . . . H 1.7. Antigen Presentation 13 1.7.1. Nature of the ER 14 1.7.2. Chaperones 15 1.7.3. Other chaperones • 15 1.7.4. Glycosylation 17 iv 1.8. The Golgi complex and membrane traffic 18 1.9. T Cell Receptor/Ligand Interactions 19 1.10. Viruses 22 1.11. Adenovirus 25 1.11.1. E3 region 29 1.11.2. Overview of proteins and effects of the E3 region 31 Objectives 34 2. Materials and Methods 36 3. Comparison of Differential Binding of E3/19K to MHC Class I allelic proteins 46 4. Efforts to increase the cell surface expression 76 5. E3/19K binding to MHC CI I does not exclude association with calnexin 93 6. Discussion 119 7. Conclusion 127 References 128 Appendix 1 147 Appendix 2 148 v List of Tables T a b l e 1. Summary o f the effects o f E 3 region proteins 30 Tab le 2. Reagents used 37 Tab le 3. Summary o f E 3 / 1 9 K association wi th H - 2 a l le l ic proteins 66 Tab le 4. A comparison o f the ce l l surface expression o f the H - 2 D protein at 2 6 ° C and 3 7 ° C wi th and without excess P2m 86 T a b l e 5. T h e effect o f y - I F N on M H C C I I c e l l surface express ion 89 T a b l e 6. Dete rmina t ion o f M o l e c u l a r W e i g h t ( M W ) o f proteins i n sucrose gradient fractions 160 v i List of Figures Figure 1. The M H C C I I regions for different species 5 Figure 2. The M H C C I I gene 7 F igure 3. Crys ta l structure o f M H C C I I 9 Figure 4. The interaction between M H C C I I and the T C R 21 F igure 5. T h e antigen presentation pathway 24 Figure 6. The adenovirus genome 27 Figure 7: The E 3 region o f adenovirus 29 Figure 8 The E 3 / 1 9 K protein 32 Figure 9. F A C S analysis o f 293 cells and 293 transfectants . 449 Figure 10. The rate o f transport o f H - 2 M H C CI I allelic proteins transfected into 293 cells 53 F i g u r e 11. T h e effect o f A d 2 infec t ion assessed by the co- immunoprec ip i t a t ion o f E 3 / 1 9 K wi th H - 2 M H C CI I i n 293 cells 59 v i i Figure 12. The effect o f A d 2 infection on M H C CI I expression in 293 cells and 293 transfectants 63 Figure 13. Ove rv i ew o f maturation paths for the M H C C I I complex 74 Figure 14. Intracellular accumulation o f H - 2 D b in 293 cells 80 F igure 15. The effect o f growth at 2 6 ° C on the intracellular accumulation o f H 2 D b in 2 9 3 D b cells 82 F igure 16. The effect o f infection wi th a vaccinia virus human-(3 2 m construct on P 2m expression 85 Figure 17. The effect o f growth at 2 6 ° C and infection wi th a vacc in ia virus human-(3 2m construct on escape o f intracellular stores o f H - 2 D b protein f rom the E R 87 F igu re 18. The effect o f increasing levels o f tun icamyc in on M H C C I I i n 293 and 293.12 cel ls 97 F igure 19. T i m e course o f maturation o f M H C C I I i n 293 and 293.12 cells 99 v i i i Figure 20. Calnexin association with MHC CI I precipitated from 293 and 293.12 cells 102 Figure 21. Calnexin association with W6/32 and TAP precipitates from 293 and 293.12 cells 105 Figure 22. Precipitation of MHC CI I in enriched suspension using a biotinylated peptide 110 Figure 23. Separation of 293 lysates over a 5-20% sucrose gradient 151 Figure 24. Precipitation of MHC CI I and E3/19K from 293 and 293.12 cell lysates separated over a 5-20% sucrose gradient 154 ix Nomenclature 293 human cell line derived from embryonic kidney carcinoma 293.12 293 cell line transfected with E3/19K 621 293 cell line transfected with truncated E3/19K Ad adenovirus P2m beta-2-microglobulin BiP heat shock protein in the ER BSA bovine serum albumin DMSO dimethyl sulfoxide DNA deoxyribonucleic acid DSP dithiobis (succinimidylpropionate) DTT dithiothreitol E3/19K 19 kilodalton adenovirus protein coded in early region 3 ECL enhanced chemiluminescence EDTA ethylenediaminetetraacetic acid EGF epidermal growth factor Endo H endoglycosidase H ER endoplasmic reticulum ER endoplasmic reticulum FACS fluorescence activated cell sorter Gal galactose ylFN gamma interferon H-2 mouse MHC CI I region The mouse versions are K, D and L HEPES 2-(2-hydroxyyethyl)-l-piperadineethanesulfonic acid HLA human MHC CI I region HLA A, B, and C HRPO horse radish peroxidase Ig immunoglobulin IR immune response Kd, Kb, etc Alleles d and b of the mouse H-2K gene kDa kilodalton mAb monoclonal antibody MHC Major Histocompatibility Complex MW molecular weight NRS normal rabbit sera PMSF phenylmethylsufonylfluoride R425, R426 rabbit anti MHC CI antisera SA-HRPO streptavidin horse radish peroxidase SDS sodium dodecyl sulfate TAP transporters associated with antigen processing TCA trichloroacetic acid TcR T cell receptor UPR unfolded protein response W6/32 monoclonal that detects mature CI I HLA x Acknowledgements I would like to thank my supervisor, Dr. Wilfred Jefferies for his tireless assistance and endless patience in guiding both my research and many other aspects of my life. His influence will remain with me for a lifetime. I am also grateful for the environment that he created both in and outside the lab. Many lifelong friendships were started here. A special thanks is in order for the members of my committee who invaluable contribution to this document was greatly appreciated. I must thank the members of the lab without whose contribution and assistance this work would never have been completed. In particular, special thanks go to Reinhard Gabathuler and Daphne Blew. These two individuals offered continual help and assistance. I would also like to thank the members of the department whose daily antics and thoughtful comments kept me both amused and focused. These people include Roger Lippe, Ian Haidl, Gregor Reid, Gerry Kolaitis, Michael Food, Kathy Shimizu, Nancy Carpenter, Greg Lizee, Jacqueline Tiong and Brandie Thorlakson. Special thanks to Alex Moise, Joseph Yang, and Liz Frey. I am indebted to the Department of Zoology for their financial support in the form of TA ships and for the opportunities to teach. In particular, the guidance of Ellen Rosenberg, Brian Oates and Carol Pollock was appreciated. Finally I am grateful for the support of my friends and family who supported my quirkiness throughout my degree. Keith MacCrimmon was always a phone call away. Nick Heap was especially introspective at all the right moments. Elizabeth Richards helped keep everything in perspective. My mother and father supported me in countless ways. Above all, the contribution, support and guidance of Renee leNobel were invaluable in the completion of this work. xi 1. Introduction The human body with its immune system is like a heavily fortified medieval donjon. While a great number of attacks on the integrity of the body are thwarted on a continual basis, breaches of the immune system by pathogens can result in discomfort, disease and worse. Persistent viral infections result when viruses escape detection by the immune system. Many viruses have evolved a set of mechanisms that effectively target susceptible components of the immune system (1). These viruses can propagate undetected and unchallenged by the host immune system. Persistent viral infections offer a good model to examine the immune system. Investigation of the strategies used by viruses to escape the immune system provides a useful method for understanding the components involved in antigen presentation and the interactions between them. This study investigates the role of the endoplasmic reticulum (ER) resident E3/19K protein of human adenovirus in the immune system, protein processing and maturation through the ER and Golgi, antigen presentation and viral clearing. 1.1. Immunity An immune system is present in all vertebrates and many invertebrates. It is a mechanism employed by organisms to protect themselves from pathogens and other foreign materials. This system also serves as a method to detect self from non-self. While in many lower organisms this differentiation of self versus non-self is a simple mechanical differentiation between 'in and out', in higher vertebrates the immune system is more sophisticated with multi-faceted methods for preserving the integrity of the living body (2). The immune system in humans and mice has both specific and non-specific methods for eliminating foreign substances. Non-specific mechanisms include physical barriers, temperature levels and pH to make living conditions hostile to potential pathogens. 1 Additionally, lysozymes, interferons, complement and natural killer (NK) cells are present and non-specifically respond to foreign objects. A second, more sophisticated method available to the body to defend itself are the specific responses to foreign material. These are known as the acquired immune response. Some features of the acquired immune response include immunological memory, the ability to respond to a high number and diverse set of antigens and the ability to recognize foreign antigens in the context of self antigens (3). 1.2. Pluripotent Stem Cells The acquired immune response is driven by the action of lymphocytes. Pluripotent stem cells are the progenitors of two lines of cells, the lymphoid and the myeloid. The lymphoid line can be further divided into several cell types including B and T cells. B Cells have characterizing immunoglobulins present on their surface; T Cells are distinguished by the presence of a T Cell Receptor and Cluster Differentiation marker CD3. 1.3. B Cells B Cells are specialized for immunoglobulin (antibody) production. Antibodies can be raised against almost any type of foreign antigen. Antibodies are very specific and one antibody recognizes one epitope. Immunoglobulins consist of a heavy (H) and light (L) chain. Different combinations of H and L chains can be assembled to generate a large diversity of antibodies specific for different antigens. The diversity of antigens against which antibodies can be raised is considerable. 1.4. T Cells T Cells recognize fragments of proteins bound to self MHC molecules (4). Viral (and other) proteins are degraded into peptides. These peptides are bound by major 2 histocompatibility complex (MHC) molecules and recognized by T cell receptors (TCR) at the surface of infected cells (5). Both MHC CI I and CI II molecules present foreign (and endogenous) antigens to the TCR of circulating T cells (6). The T Cell population can be further divided in CD4 + and CD8 + populations. CD4 + cells are involved in the recognition of MHC CI II with bound peptide; CD8 + cells are involved in the recognition of MHC CI I with bound peptide. Immunoglobulins and the TCR are structurally related to one another and are part of the immunoglobulin gene superfamily. Additional members of this family also include MHC CI I, CI II, CD4, CD8, (32m, MRC-OX2, ICAM-1, ICAM-2 and many others. These proteins have a characteristic 110 residue domain and a cysteine bridge (7). It is likely that these proteins evolved from a limited number of original genes. The great diversity present in both immunoglobulins and the TCR is the result of gene rearrangements. This diversity is produced by rearrangements between V, D and J segments. RAG-1 and RAG-2 genes are responsible for this genetic rearrangement (8). As the primary methods for clearing viral infections include removing infected cells from the host before the viral cycle has completed, detection of infected cells is very important. T Cells effect the killing, recognizing the processed foreign proteins presented at the cell surface of an infected cell by a MHC CI I protein (9). In cases where the immune system is unable to clear the viral infection, persistent infection occurs. The effects of the persistent infection will not become apparent until the general health of the victim declines, often as the patient succumbs to old age or another illness (such as HIV). Some of the viruses that cause persistent infections include Herpes Simplex Virus (HSV), Cytomegalovirus (CMV), Adenovirus (Ad), and Human Immunodeficiency Virus (HIV). Each has particular and differing strategies to achieve a persistent infection due to 'stealth' (10). 3 1.5. MHC The major histocopatibility complex (MHC) is one target of persistent viruses. The MHC region is found in all vertebrates and many invertebrates (11). Precursors to the MHC are also found in some invertebrates. The region codes for more than one thousand proteins including many that play key roles in the immune system. These proteins include MHC CI I , CI I I , complement, TAP 1 and 2. The MHC region is on chromosome 17 in humans and chromosome 6 in mice. There are three MHC CI I loci in both humans (HLA-A, B and C) and mice (H-2 K , D and L) (12). Polymorphism at the MHC CI I locus is extensive. With as many as one hundred alleles at each loci, the diversity found within the population is very high. MHC CI I molecules have been studied extensively, with CI I regions studied in vertebrates ranging from xenopus to human (figure 1). 4 Figure 1. The MHC CI I regions for different species M H C regions for many different species have been identif ied based on s imilar i t ies between species ( 1 2 ) . M H C CI I genes are represented by white boxes; M H C CI II genes by b lack boxes. DP DQ DK BC R h e s u s m o n k e y RbLA-D 8 " f t C h i m p a n z e e Cb LA-P-9 -U—D G-D a S h e e p OLA Dog AC FLA R a b b i t R a t 1—G—-fl-K A E A B O B "Qo" G u i n e a Q\q CPLA—IIHH [] [HK DML Qo-2 Th -LHHf OOHHHF S y r i a n h a m s t e r Hm-l Si~£i -Smh-F o w l Xfmopus XLA . i f 5 M H C CI I is found noncovalently associated wi th beta-2-microglobul in (P2m) at the cell surface o f most nucleated cel ls in the body. M H C C I I molecu les consis t o f three ext racel lu lar domains (alpha 1-3), a transmembrane region and a cy top l a smic ta i l . E a c h extracellular d o m a i n consists o f about 90 amino acids and is a lso a member o f the Ig super fami ly . T h e M H C CI I molecule is g lycosyla ted at asparagine 86 in humans wi th a second (and sometimes third) g lycosyla t ion site in mice . T h e mature M H C C I I molecule is about 46-55 kilodaltons (kDa) . The polymorphic regions o f the M H C CI I molecule are found main ly i n the alpha 1 and alpha 2 regions. The genetic organiza t ion o f the M H C C I I region reveals many exons and introns. E a c h o f the three alpha regions and the transmembrane region are encoded by a separate exon and the cytoplasmic tail is compr ised o f three exons (figure 2). 6 Figure 2. The MHC CI I gene T h e M H C CI I gene is made up o f large exon and intron regions (12). Similarities between the exons (black boxes) and introns o f mouse (K<» and L A non classical Q6) , human H L A 12.4 and f j 2 m are depicted be low. 5 'UT { «1 E l " E2 £3 5 ' U T 1 «1 <r2 * 3 TH / \ \ 3 ' UT 1 3 ~ ? r \ C ? n t \ s ' \ t U E 5 £6 E7 E8 S3—I—fi—S 5 'UT I o1 Q 6 (27.1) a 3 TH CY 5 ' U T 1 °1 HLA 12.4 x3 TM CT /5,m 5 'UT 1 . . . 3 ' U T 1—— 5 kb 7 Since antigens are presented by M H C molecules, extensive p o l y m o r p h i s m a l lows for great variabili ty in the number and types o f antigens that can be presented wi th in a population. Some a l l e le l i c forms o f M H C CI I proteins present specific antigens far more efficiently than others. Ind iv idua l s wi th certain alleles are more susceptible to some disease because their M H C CI 1 is s i m p l y unable to present the antigen to their i m m u n e system. F o r example , certain autoimmune diseases have been l inked to specific H L A alleles. Structural ly, the M H C CI I molecule consists o f two antiparallel alpha helices (alpha 1 and 2) resting on top o f two antiparallel beta pleated sheets (alpha 3) (13). The helices form a 25 A long by 10 A wide groove. The highly conserved 12 k D a protein 3 2 m is non-covalently associated wi th the alpha 3 domain . The crystal structure for M H C CI I has been determined for many a l le l ic proteins. The ribbon structure is demonstrated in figure 3. Peptide is bound in the groove between the alpha 1 and a lpha 2 regions (f ig 3a). Peptide hypervariable regions o f the M H C C I I molecule are found near the peptide b inding groove. The resolved crystal structures have demonstrated the existence o f pockets inside the groove into w h i c h the R-groups o f bound peptides fit t ightly. Peptide sequence constrains wh ich peptides fit into the groove. Different a l le l ic forms o f M H C CI I b ind sets o f peptides l imi ted in sequence by key 'anchor ' posit ions. Peptides der ived from v i ra l (and endogenous) proteins are tightly bound in the peptide b i n d i n g groove . F igure 3b illustrates the peptide b ind ing groove from the top. Th i s associat ion is a requisite step in the assembly o f a stable M H C C I I molecule (14). 8 Figure 3. Crystal structure of MHC CI I The crystal structure o f M H C CI I reveals the presence a large peptide b inding groove at the top. In a) the peptide binding groove is demonstrated l y ing between a l and a 2 regions. A l s o shown is the close association o f (3 2m. In b) and c) the peptide b inding groove is seen from above. The phys ica l constraints o f the groove permit l imi ted sets o f peptides to b ind . The constraints on the peptide include pockets wi th in the groove for amino acid R group side chains and we l l conserved anchor positions (13). 9 1.6. Cellular Mechanics 1.6.1. Peptides The abili ty o f the immune system to respond to mil l ions of potential pathogens is due to the large number o f different peptides that can be bound by the M H C CI I molecule . Some investigators have suggested that potentially mi l l i ons o f different peptides can occupy the peptide b inding groove o f M H C CI I. T h e T C R found on the surface o f T cells recognises both the po lymorph ic region o f the M H C C I I molecule and the peptide contained wi th in the groove. Recent studies demonstrate direct contact between the T C R and the peptide occurs (15). A study in w h i c h self peptides eluted from M H C CI I were sequenced identified many motifs o f peptides associated wi th a single al lel ic protein (16). Other studies that examined the stability o f M H C C I I molecules associated w i t h peptides showed that there is a hierarchy o f peptide sequences that confer stabil i ty (17). T h e peptide specificity o f different M H C C I I a l l e l i c proteins is i n part de t e rmined b y the side chains o f the amino acids l i n i n g the polymorphic peptide groove o f the M H C C I I molecule. Different allelic proteins o f M H C CI I have preferences for peptides ranging f r o m 8 to 11 amino acids in length (18). A m i n o acid side chains o f bound peptides ex tending into the groove dictate that certain peptide residue posit ions are regarded as anchor residues (19, 20) . F o r example , the H - 2 protein K b preferentially binds octapeptides wi th a tyrosine or phenylalanine at position 5 and a M e t or He at posi t ion 8 ( 1 8 ) . 1.6.2. Proteasome Peptides that b i n d to M H C C I I molecules are generated wi th in the cy top lasm by the proteasome. The proteasome is a barrel shaped structure made up o f many distinct subunits. 11 Viral and endogenous proteins are tagged with ubiquitin and targeted for degradation by the proteasome. The proteasome is well-conserved throughout many species ranging from Archaebacterium (21) through to humans and serves as a general cellular method to remove abnormal, short lived and otherwise redundant proteins. The proteolytic process starts with the tagging of the doomed protein with ubiquitin to mark the protein for hydrolysis by the 26S proteasome (22). The proteasome is also able to function without ubiquitin. The proteasome consists of a 20S Multi Catalytic Protease complex (MCP) catalytic core with additional subunits. The 20S core is approximately 650 kDa (23), composed of 13-15 subunits of similar size (24) stacked in a barrel shape and is found in the cytosol and the nucleus of all eukaryotic cells. It has multiple catalytic sites and comprises up to 1% of the total protein of a mammalian cell. Association of an ATPase complex results in a 26S proteasome which degrades ubiquitinated proteins in the presence of ATP (25). In another configuration of the proteasome, the two subunits, LMP 2 and LMP 7, are also found. They are encoded in the MHC region, are upregulated by y-IFN and have been shown to participate in a proteasome involved in the generation of peptides bound by MHC CI I molecules (26). Proteasome-generated peptides are presented to immature MHC CI I molecules present in the ER by the MHC-encoded transporter associated with antigen presentation (TAP) proteins (27). Encoded in the MHC CI II region, TAP 1 and 2 form a heterodimer that pumps peptides across the ER membrane in an ATP-dependent fashion. These two proteins are non-covalently associated in the ER. The TAP proteins are members of the ATP binding cassette (ABC) transporter protein family and show homology across species. The binding of TAP proteins with empty MHC CI I heavy (H) chains and P2m in the ER (28-30) is required for MHC loading with peptides (31). Cells with deficiency in the TAP proteins show reduced peptide 12 loading and antigen presentation (32-35). Studies have demonstrated that stringency o f these transporters is not high for peptides is not h igh. Peptides must be longer than 7 residues and peptides longer than 12 residues are transported wi th lower eff ic iency (36, 37). S o m e specif ici ty in peptides transported by T A P has been, shown. In rats, different alleles o f T A P transporters favour different sets of peptides (38) . L i k e most p l a sma membrane and secretory proteins, M H C C I I molecules are synthesized on E R bound ribosomes and transported across the E R membrane to the lumen o f the E R . Process ing occurs in the lumen o f the E R and continues wi th transport from the E R to the c i s , media l and trans sections o f the G o l g i to the ce l l surface. In general , proteins are transported out o f the E R when they have achieved a correct conformat ion for exit. Treat ing cel ls wi th reducing agents or otherwise al tering the condit ions found in the E R can delay or even b lock transport o f proteins out o f the E R (39, 40) . Several studies have shown that in the case o f M H C C I I molecules , exi t ing the E R can be the rate l imi t ing step for ce l l surface expression. 1.7. Antigen Presentation The maturation o f the M H C C I I into a tr imolecular complex made up o f heavy (H) chain , P 2 m and bound peptide involves several specific proteins i n the E R . Peptide is supplied by T A P 1 and T A P 2. Recent findings suggest that a 48 k D a glycoprotein cal led tapasin is also associated with the M H C C I I - P 2 m - T A P complex (41). Ca lnex in functions i n a chaperone l ike manner for glycoproteins such as transferrin (42) and is required for association o f M H C C I I wi th T A P (43). The calc ium-binding calreticulin protein also associates wi th nascent proteins fo l l owing the association o f ca lnexin . B o t h calnexin and calreticulin improve the chances o f a mature conformation by preventing aggregation and premature disulphide bond formation (44). 13 The interaction o f many different intermediates results in antigen presentation by M H C CI I at the c e l l surface. The cytokine gamma Interferon (y-IFN) can upregulate many aspects o f antigen presentation. In ce l l lines with l ow M H C CI I expression, treatment with y - I F N results in increased cel l surface expression (45). Studies in M H C CI I deficient cells demonstrate that the addi t ion o f y - I F N upregulates the association o f 3 2 m wi th M H C C I I H chains and the subsequent c e l l surface express ion (46) . In add i t ion , y - I F N upregulates the proteasome subunits encoded i n the M H C region and M H C C I I H chain production (47). 1.7.1. Nature of the ER L i k e other membrane and ce l l surface proteins, the in i t ia l assembly o f the M H C CI I c o m p l e x occurs i n the l umen o f the E R . T h e endoplasmic r e t i cu lum is an interconnected network o f membranes forming a tubular system w i t h i n the c e l l that is cont inuous with the nuclear membrane. Proteins destined for secretion or ce l l surface expression are processed in the E R soon after translation. These proteins contain a N terminal signal sequence consisting o f about 20 residues that mediates transport across the E R membrane. Once inside, the signal is qu ick ly cleaved and the nascent protein associates wi th E R resident proteins that bind to and further process the protein. F o r membrane bound proteins, a second signal sequence found on the prote in stops t rans locat ion across the membrane and causes this r eg ion to become embedded i n the E R , serving as the transmembrane anchor (48). P rocess ing i n the E R is not yet comple te ly understood. There are many types o f processing enzymes that act i n concert o n nascent proteins. These inc lude s ignal peptidase, p ro te in d i s u l p h i d e i somerase ( P D I ) , p ro te in p r o l y l i somerase ( P P I ) , E R G I C 53 (49), o l i g o s a c c h a r y l transferase, a lpha g lucos idase I, a l p h a g lucos idase I I , E R a lpha 1,2-mannosidase, thiol dependent reductase E R p 5 7 , ca lnexin , calreticulin and B i P . Different sets 14 o f molecules associate together and mod i fy the newly formed E R prote in . Modifications inc lude the addi t ion o f ol igosaccharides , o l igomer iza t ion and the format ion o f d isu lphide bridges. W h e n complete, a protein with a mature and correct conformation can exit the E R and proceed to further processing through the cis G o l g i . 1.7.2. C h a p e r o n e s Proper assembly o f E R proteins involves E R chaperone proteins. Chaperones can be defined as proteins that interact wi th immature or misfolded proteins and either help them to achieve a proper conformation or direct them to be degraded. S o m e w e l l k n o w n chaperones include the stress proteins (50) including B i P (51, 52) and ca lnexin . B i P is involved in both translocation o f nascent proteins into the E R and i n the b ind ing and retaining o f misfolded proteins in the E R , whereas ca lnex in binds and retains nascent glycoprote ins in the E R until they achieve a mature conformation (53, 54). T h e chaperone ca lnex in (55) associates w i t h nascent g lycoprote ins i n the E R i n a l ec t in - l ike manner by b ind ing to the intermediate G l C j M a n 9 G l c N A c 2 (56, 57). C a l n e x i n associat ion stabilises intermediate structures prevent ing their r ap id degradat ion (58) and enhances efficient assembly. 1.7.3. Other chaperones In addition to interacting wi th calnexin , it is clear that membrane proteins bound for the cell surface (and other glycoproteins) must also interact wi th other chaperones. These include chaperone molecu les such as prote in d i su lph ide isomerase, g lucosy lases and c is -proly l isomerase. A l s o involved directly wi th M H C CI I molecules and antigen presentation are the T A P proteins and tapasin and B i P . C a l n e x i n associates wi th immature M H C C I I (and other membrane glycoproteins) in 15 the ER. Calnexin binds immature glycoproteins through a lectin-like domain (56). The lectin action (and subsequent association) of calnexin requires a mono-glucosylated oligosaccharide. The glycosylation/de-glycosylation cycle is an effective means of generating and re-generating substrate for calnexin binding and re-binding, causing incorrectly folded or immature proteins to be retained by calnexin (59). The binding step will apply steric constraints to the bound protein, and during this step inter- and intra-chain disulphide bonds can be formed. Glucosylation of glycoproteins regulates transport of glycoproteins out of the ER (60). Once the calnexin-bound protein achieves correct conformation, the final glucose residue is removed and association with calnexin terminates (57). With MHC CI I , the correct conformation is achieved with the addition of the peptide and P2m forming a tri-molecular complex (61). Studies with reagents that inhibit some of the above mentioned chaperones demonstrate that disruption of regular maturation results in incorrectly folded or immature proteins. These proteins are either retained in the ER or they are tagged for degradation. They are not further modified in the ER and are not further processed through the Golgi. Recent studies using lactacystin to inhibit the proteasome have suggested that such tagged proteins are exported back to the cytoplasm for degradation (62). Studies with daudi cells that do not produce correctly folded MHC CI I because they are missing P2m protein showed the accumulation of heavy (H) chains in the ER when treated with a proteasome inhibitor (63). Use of the reducing agent dithiothreitol (DTT) and permeabilised cells showed that disulphide bridge formation in nascent proteins is a fairly early reaction and is mediated by protein disulphide isomerase. Preventing disulphide bridge formation prevents proper folding of MHC CI I, but is reversible upon removal of DTT (64). While formation of the disulphide bridges occurs during protein folding, it is likely that this step does not influence folding, but rather promotes rapid reshuffling of incorrect disulphide pairings and stabilising of a correct 16 one (65, 66) Excess unfolded protein in the E R can trigger the unfolded protein response ( U P R ) . The U P R is an upregulation o f E R lumenal proteins inc luding B i P and P D I in response to ma l fo lded proteins in the E R . Stresses that can cause the U P R inc lude i nh ib i t i on o f glycosylat ion by starvation or the addition o f drugs, addition o f reducing agents, expression o f fo lding mutants, or addi t ion o f c a l c i u m ionophores that reduce c a l c i u m stores in the E R . In yeast, a transmembrane serine/threonine kinase has been identified that is i nvo lved in the U P R , possibly transmitting the U P R signal to the nucleus (67). It has also been found that protein over load o f the E R can trigger N F - k B in a manner independent o f the U P R , poss ibly by causing the release o f C a 2 + f rom the E R (68). Th i s may be a generalised cellular antiviral response. 1.7.4. Glycosylation Nascent proteins are g lycosyla ted i n the E R . A m o n g s t the early act ing E R chaperone proteins are the g lycosy la t ion mediators. A n oligosaccharide assembled o n the l i p i d carrier do l i cho l phosphate is transferred to a nascent g lycoprote in . T h i s o l igosacchar ide consists o f two G l c N A c , 9 mannose and three g lucose residues ( G l c 3 M a n g G l c N A c 2 ) and is able to efficiently associate cotranslat ional ly at most A s n - X - S e r / T h r sites. S ter ic hindrance and accessibili ty o f the site seem to influence transfer o f the ol igosaccharide to the glycoprotein; different g lycosy la t ion sites on the same protein exhibi t different but consistent g lycosyla t ion patterns (69, 70) . O n c e a protein has fo lded, the sites o f g lycopro te in attachment may no longer be accessible to the cel lular glycosylat ion machinery. U p o n transfer, this o l igosacchar ide is a lmost immedia te ly set upon by E R resident proteins. Glucos idase I acts almost immediately on the outermost glucose o f the N - l i n k e d core g lycan , r e m o v i n g it. T h e second glucose is then sequential ly removed by glucosidase II. 17 Glucos idase II also subsequently removes the third glucose, l eav ing a h igh mannose form of the sugar. Replacement o f the third glucose residue is mediated by UDP-glucose .g lycoprote in g l u c o s y l l r a n s f e r a s e ( 6 0 ) r e s u l t i n g i n a m o n o - g l u c o s y l a t e d g l y c a n . A glucosylation/de-glucosylation cycle (71) seems to be one method employed to effect quality control wi th in the E R (57, 72). S tudies w i t h t u n i c a m y c i n , w h i c h b l o c k s the synthes is o f the d o l i c h o l l inked o l igosaccha r ide , show that g l y c o s y l a t i o n is an essent ial step in the p rocess ing o f a g lycoprote in . Fa i lu re to become glycosylated results in immature proteins that collect in the E R , associated wi th B i P (73). T h e above studies show that each step i n the maturation o f a g lycoprote in is essential for any subsequent steps. Studies with castanospermine reveal that inhibit ing the glucosidases has a s imi la r effect, resulting in rapid degradation o f unassembled M H C C I I (74). Indeed, n e w l y formed proteins are subject to qua l i ty cont ro l w i t h those that are mis fo lded be ing retained and degraded (75). There is some evidence that degraded proteins exit the E R and are degraded in the cytosol by the proteasome machinery (76). 1.8. The Golgi complex and membrane traffic Proteins are further modi f i ed as they travel f rom the E R into the cis G o l g i . T h e y continue through to the trans face o f the G o l g i and are packaged i n vesicles or make their way to the ce l l surface. W h i l e the 100 year anniversary o f the discovery o f the G o l g i apparatus was recendy celebrated (77), much o f the structure, function and organization o f the G o l g i have yet to be determined. The G o l g i is the location o f hundreds o f enzymes. M u c h debate has centered on the nature o f g lycoprote in maturation. Wha t is agreed upon is that proteins wi th s ignal sequences are subject to the 'export ' mechan i sm of the cel l (78). Proteins are exported from the cytosol to the lumen o f the E R where they have access to the exterior o f the ce l l through the G o l g i . Proteins make their way through the medial and trans 18 G o l g i before being packaged in vesicles that can transport them to the ce l l surface where they are either released outside the ce l l or they form part o f the outer layer o f the plasma membrane. What is less clear are the rules governing the transport o f material between the various compartments o f the E R - G o l g i system. The bulk f low theory suggests that there is a bulk flow from the E R through the G o l g i to the cell surface (79, 80). A n y resident in the E R that is not tagged with a retention signal w i l l eventually make its way through the E R , the G o l g i complex and through to the c e l l surface. Recent ly , investigators e luc ida t ing the nature o f vesicular budding in yeast have chal lenged the 'bulk f low model ' , postualt ing that proteins (known as cargo) associated wi th membrane bound vesicles are transported through to the G o l g i from the E R in response to some posit ive signal (81-83). A sugar may mediate this positive signal (84) and E R G I C - 5 3 (that has some lect in-l ike properties) is the protein mediat ing this (49, 85). .9. T Cell Receptor/Ligand Interactions Once an M H C C I I molecule has matured through the E R - G o l g i processing machinery, it presents bound peptide at the ce l l surface for recognit ion by T ce l l s . C D 8 + cytotoxic T lymphocytes ( C T L ) are involved in v i ra l clearance. Recogni t ion and subsequent destruction o f viral ly infected cells is effected by T C e l l Receptors ( T C R ) that must recognize foreign antigens in the context o f a host or ' s e l f protein (86). The antigen that stimulates a response is the bound peptide fragment derived f rom foreign protein. Studies have shown that C T L responses may be raised against a single immunodominant epitope (87, 88). T h e crystal structure o f the T C R interacting w i t h the M H C C I I molecule has a l l owed insight into the interaction. The M H C C I I peptide groove a l lows b ind ing o f the peptide i n a s ingle orientation; much o f the peptide is hidden from the T C R . Studies wi th mutants suggest that the T C R binds diagonally across the alpha helices (89), avo id ing the protruding N- termina ls o f each o f the a l and a2 alpha helices and max imis ing contact wi th the p e p t i d e - M H C C I I complex (13, 90). 19 Recent elucidation o f the crystal structure o f a T C R interacting wi th the polymorphic region o f M H C CI I and peptide shows the intimacy (and specificity) o f the contact between the two proteins. Figure 4 reveals the tight association o f the a and (3 chains of the T C R with the a l and a.2 regions o f M H C CI I and the bound peptide. 2 0 Figure 4. The interaction between MHC CI I and the TCR This figure demonstrates the tight association between the a and (3 chains o f the T C R (shown in pink and blue respectively) and the M H C CI I protein (depicted in light green) with bound peptide (ye l low) . The peptide in the peptide groove is sandwiched between the T C R and the a l and al sections o f the M H C CI molecule. C lose association o f (32m (dark green) with the a 3 regionis also observed (15). 21 1.10. Viruses M o s t v i ra l infections are cleared i n the procedures ment ioned above. In general, the viral life cyc le starts with infection, fo l lowed by transcription o f early viral genes that hijack the ce l lu la r machinery. T h e host machinery is then used to replicate the v i ra l genome and synthesize v i ra l proteins. Proteins and the genetic material are packaged to produce more viruses w h i c h are released into the surrounding area. In short, a virus turns its host cel l into a v i rus-making factory. Different ce l l types are more or less susceptible to infection by particular viruses. N o n -permiss ive c e l l types do not readi ly support in fec t ion and propagat ion o f virus whereas permissive ce l l types readily succumb to v i ra l infection and contribute to the viral life cycle. The immune system attempts to destroy v i ra l ly infected cel ls prior to the release of more v i rus . C T L s target infected cel ls and cause their i m m i n e n t destruct ion. T C e l l s recognize antigens w i th in the context o f a l l e l i c forms o f M H C molecu le s . T h i s is k n o w n as a l le l ic restriction (4). 22 Figure 5. The antigen presentation pathway The steps occurring in antigen processing are detailed here, a) Proteins are synthesised within the cell. Proteins tagged with ubiquitin are degraded by the multicatalytic protealytic complex or proteasome. The proteasome is a large (650 kD), well conserved found within the cytoplams. Peptides of 7 to 12 residues are transported into the ER through the action of the ATP dependant TAP proteins. b) MHC CI I proteins are co-translationally translocated into the ER where the associate with the chaperone calnexin. The H chain also associates with L chain (P2m) a n d other chaperones such as Erp57 and calreticulin to name a few. Tapasin mediates the association of the immature MHC CI I H chain-(P2m) complex with TAP. Peptide loading by TAP proteins results in a mature conformation accompanied with release by calnexin. The tri-molecular complex is subsequently expressed on the cell surface where it can be recognised by the TCR of passing T Cells. (91) 23 24 1.11. Adenovirus There exist at least 70 different serotypes of human adenovirus wh ich are divided into 6 groups termed A through F . The pathological effects o f adenovirus are quite variable, but in general, they can cause acute disease o f the eye, respiratory tract and gastrointestinal tract (92). W h i l e adenovirus rarely causes a chronic medical condit ion in healthy individuals , the virus has the abil i ty to evade the host immune system and cause a persistent infection. In several studies, infection by viruses in general and adenovirus in particular tended to result in a sub clinical effect (93). In addit ion, adenoviruses may become persistent and shed in feces for years post infection i n otherwise healthy indiv iduals (94). Adenovi ruses can exis t i n a latent state and cause disseminated disease i n immunocompromised patients due to react ivat ion (95). F o r example, it has been demonstrated that over 10% of patients wi th acquired immunodeficiency syndrome ( A I D S ) shed adenoviruses i n their urine (94). It has been suggested that infect ion and persis tence o f adenoviruses i n ear ly c h i l d h o o d m a y resul t i n respiratory i l lness i n adulthood (96). T h e adenovirus is a non-enveloped icosahedral structure made up o f an outer protein caps id and an inner core composed o f a t ightly packed D N A - p r o t e i n core (97). T h e protein capsid is composed o f 240 hexons and 12 pen ton bases that are noncovalently attached to fibre proteins. E a c h penton is surrounded by f ive hexons and is found at each o f the twelve 'corners ' o f the v i r i on wi th an attached protruding fiber prote in . T h e core consists o f virus D N A and core proteins. The v i r ion is approximately 600-700 A in diameter (97). The capsid enters host cells by association o f the fiber proteins wi th specific receptors on the ce l l surface (98) fo l lowed by import o f the v i r ion into the ce l l . T h e v i r i on is transported to the nucleus where the capsid is stripped away and the D N A is transcribed either i n an 'early' or a 'late' phase. T h e early phase precedes v i r a l repl icat ion; the late phase fo l lows viral repl icat ion. E a r l y genes code proteins that are able to disrupt normal ce l lu lar processes, 25 i n c l u d i n g those that i n v o l v e presenting peptides der ived from foreign proteins at the ce l l surface. Late phase proteins have functions that aid packaging the large amounts o f viral D N A and viral capsids into new vir ions. T h e genome o f adenovirus consists o f about 36,000 base pairs o f l inear, double stranded D N A . It contains mul t ip le over lapping spl iced m R N A s generated from a single m R N A precursor (F igure 6). T h e early genes are d iv ided into 6 major transcriptional units, consist ing o f the E l A , E 1 B , E 2 , E 3 , E 4 and L l . The immediate early region consists of E l A gene products inc lud ing a 289 amino acid (aa) protein and a 243aa protein that act as trans-ac t ing regula tors o f other ea r ly r eg ion genes (99) and as suppressors o f certain enhancer-dependent genes. E 1 B gene products are poor ly understood. E 2 region gene products participate in viral D N A replication. The E 3 region codes for several gene products that enhance the abi l i ty o f a v i r a l ly infected ce l l to evade the immune system. The E 4 region also codes for a trans-activating protein and little is k n o w n about the functions o f the L l gene products. 26 Figure 6. The adenovirus genome T h e adenovi rus genome is made up o f early and late t ranscr ib ing units. In this i l l u s t r a t i o n , the double l ine represents the d s D N A . A r r o w s represent transcription products and numbers refer to proteins. Rightward transcription products are depicted by the r; leftward by the 1. (100). 289R 243R 123R E1B 495R 15SR I— III4 III pVU V pVI II ( h e x o n ) L3 L2 L l ia6K 55.52K I O O . J J K pVUl * 1 14 L5 E3 \25K. airC 6 7 K gp19K. 11.6K. 75K 10.4K. 145K, 14.7K 20 30 40 50 60 ro E2A 80 90 . i .... i 100 IV«, U3K. 14.6K. U 2 K 7.1K. 133K. 34.1K 17.1K. 3 3 K E4 80K terminal protein 140K DNA polymerase E2B 72K ssONA binding 27 T h e adenovirus has deve loped several mechanisms to evade the i m m u n e system. Perhaps the best characterised is its abi l i ty to downregulate the ce l l surface expression o f M H C CI I molecules and therefore evade the immune surveil lance mechanism. V i r a l sub-groups achieve this result by different means. A d 12 b locks transcript ion o f M H C CI I m R N A , causing a repression o f c e l l surface express ion (101). A d 2 and 5 have a 25 k D a protein ( E 3 / 1 9 K ) that binds the M H C CI I molecu le and causes its retention inside the infected ce l l . Other mechanisms utilised by adenovirus include downregulat ion o f the E G F - R e c e p l o r s at the ce l l surface and protection from T N F - i n d u c e d ce l l destruction. 1.11.1. E3 region W h i l e the E 3 region has been s h o w n to be non-essential to adenovirus replication in cultured cells (102), it is the region that codes for several proteins that may confer an ability to evade the immune system o f the host (100). W h i l e the differences in sequence between the E 3 region o f adenovirus o f different serotypes is s ignif icant , there is consensus in the open reading frames ( O R F s ) generating some o f the proteins. In A d 2 and A d 5, variable sp l ic ing o f a common m R N A precursor gives rise to at least 6 different O R F s . 28 Figure 7: The E3 region of adenovirus T h e E 3 region encodes several proteins that aid in evas ion o f the immune system. S o l i d arrows show m R N A ; dotted lines show introns. Hatched bars show identified peptides and grey bars are those proposed to exist. (100). a c d e E 3 A 3.6 K 6.7 K gp19 K ED" 291 C S S t \ \ \ \ \ \ S S 785 1022 .205 1204 1681 I 3 72 ' ' 768 1 1.6 K I860 2163 1 7 4 0 -7.5 K m m 2166 2361 10.4 K 14.5 K 2173 2490 2495 2891 2157 E 3 B 3308 14.5 K 2495 2891 951 12.5 K 291 612 14.7 K S . W W N 2886 3270 2 6 8 0 ! 29 1.11.2. Overview of proteins and effects of the E3 region Several gene products o f the E 3 region have been studied. E3/6 .7 has been shown not to participate in downregulat ing the cel l surface level o f M H C C I I molecules in concert wi th E 3 / 1 9 K . E 3 / 1 9 K is the focus o f this study and the features shall be discussed in detail below. T h e protein E 3 / 1 4 . 7 K has been shown to prevent cy to lys i s o f adenovirus-infected cultured cells by tumour necrosis factor (103, 104). E3/10 .4 downregulates the E G F receptor (105). Table 1. Summary of the effects of E3 region proteins E 3 / 1 9 K B l o c k s surface expression o f M H C CI I (106) E 3 / 1 4 . 7 K Prevents l y s i s o f infected ce l l s by Tumour Necros i s Factor ( T N F ) (103) E 3 / 1 0 . 4 K D o w n r e g u l a t i o n o f E p i d e r m a l G r o w t h Factor ( E G F ) receptor (107) -a lso w o r k s i n con junc t ion w i t h 14.5 to prevent T N F cytolys is (108) E 3 / 1 4 . 5 K D o w n r e g u l a t i o n o f E p i d e r m a l G r o w t h Factor ( E G F ) receptor (107) -a l so w o r k s i n con junc t ion w i t h 10.4 to prevent T N F cytolys is (108) E 3 / 1 1 . 6 K A d e n o v i r u s Dea th Protein-induces apoptosis (109) E 3 / 6 . 7 K N o publ ished function T h e E 3 / 1 9 K protein is conserved across several serotypes. T h e E 3 / 1 9 K protein o f A d 2 is a 142 amino ac id T y p e 1 E R resident transmembrane glycoprote in . T h e protein consists o f a 15 amino ac id (aa) cytoplasmic tai l , a transmembrane region, a conserved 20 amino acid spacer region (110), and then a variable region. Th is lumenal section o f the protein is glycosylated at positions 12 and 6 1 ( 1 1 1 ) . A t positions 11 and 28 are the first pair o f cysteines that participate 30 in a d i su lph ide bond; the second spans f rom C y s 22 to C y s 83. These d i su lph ide bonds are conserved across several serotypes (112) and confer stability and contribute to the structure o f E 3 / 1 9 K . T h e conserved spacer and cysteine residues have been shown to be important for b inding to M H C CI I molecules (110, 113). Other studies have shown that truncated E 3 / 1 9 K molecules lacking the cytoplasmic tail and some o f the transmembrane regions retain the ability to bind M H C CI I molecules (114). Site-directed mutagenesis o f this protein has revealed that almost every stretch o f the lumenal domain is required for association o f this protein wi th M H C CI I (110). 31 Figure 8 The E3/19K protein The adenovirus E 3 1 9 K protein (114). The E3/19K protein is cotranslationally translocated into the ER. Upon transcription, its signal sequence is cleaved. Some processing occurs and it migrates as far as the cis Golgi before returning to the RER. ER retention is mediated by a short stretch on its cytoplasmic tail; truncation of the tail results in E3/19K at the cell surface (114). E3/19K associates cotranslationally with nascent MHC CI I molecules. By virtue of the ER retention site, it retains MHC CI I molecules in the ER preventing travel through the Golgi to the cell surface. The E3/19K molecule has been shown to associate with rodent, human, and simian MHC CI I molecules (115, 116). E3/19K shows varying affinities for different MHC CI I molecules (117, 118). While it binds to every human MHC CI I molecule tested to date, it is unable to bind to at least one mouse H-2 protein. During viral infections, the E3/19K protein of adenovirus type-2 (Ad2) binds to major histocompatibility complex (MHC) Class I proteins in the endoplasmic reticulum (ER) where these molecules are assembled. This prevents MHC CI I transport to the cell surface where they can present adenovirus peptides to cytotoxic T lymphocytes (CTL) (100, 119). Despite studies showing the E3 region not to be essential for virus growth in cultured cells (102), data supporting the role of the E3/19K protein in persistent infections is provided by experiments in which this protein is deleted from an otherwise normal virus and shown to result in the early clearance of the virus in Ad2 infections in mice (120). It is likely that E3/19K is involved in permitting adenovirus infection to spread and become persistent in humans (96). Lack of a good animal model has hampered study in this area. 33 Objectives W h i l e adenovirus may have many strategies to evade the i m m u n e system, the best studied mechanism is the retention o f M H C CI I proteins wi th in the E R away from detection by roving T cells . This study exploited the properties o f the adenovirus E 3 / 1 9 K protein to perturb the antigen processing component o f the immune system. A s E 3 / 1 9 K is an E R retained protein, study o f this protein a l lowed detailed analysis o f many functions o f the E R including antigen presentation and protein maturation. Add i t i ona l ly , the role o f chaperones in quality control and large functional complexes was studied. The objectives o f this study were to use the effects o f the E 3 / 1 9 K protein to study antigen processing occurr ing wi th in the ce l l . B l o c k i n g o f antigen presentation revealed details o f protein processing and maturat ion i n the E R . T h e ident i f ica t ion o f and the interactions between several E R resident proteins was revealed. The exact nature o f the b lock ing o f antigen presentation was explored by examining the role o f E 3 / 1 9 K in the E R . E 3 / 1 9 K was bound to many different M H C C I I proteins i n order to establish a m i n i m u m conformation or sequence o f residues for b ind ing . Instead the nature o f b inding was shown to be influenced by other factors inc luding the inherent abi l i ty o f a M H C C I I molecule to be processed by the endogenous E R maturation pathway. A d d i t i o n a l l y , it was possible to group M H C C I I a l le l ic proteins into three categories: binders (those that bound E 3 / 1 9 K ) , non-binders and a new category cal led s low-binders . S low-binders b ind M H C C I I quite strongly but require a longer period o f association to achieve this b inding . Investigation o f one s lowbinder revealed the retention o f large amounts o f protein in the E R with comparatively smal l amounts at the ce l l surface. Attempts to reverse the retention led to the examination o f many factors wh ich influence protein maturation in the E R and showed this process was more compl ica ted than or ig ina l ly suspected. T h i s i n turn suggested that E 3 / 1 9 K retention o f M H C CI I may involve more interactions than s imply M H C C I I - E 3 / 1 9 K association. 34 Regula r maturation o f M H C CI I wi th and without E 3 1 9 K and the association o f this complex w i t h E R resident chaperones was examined . E 3 / 1 9 K rapid ly promoted a stable conformation under adverse conditions acting l ike a chaperone. E 3 / 1 9 K association with M H C C I I d i d not b l o c k the interaction o f endogenous chaperones suggest ing that the s ize o f complexes in the E R were quite large. Other mechanisms such as the unfolded protein response may be invo lved in the E 3 / 1 9 K effect on infected cells . Peptide b ind ing by M H C CI I is a step in antigen presentation. The effect o f E 3 / 1 9 K b inding to M H C CI I on peptide binding was examined. E x p l i c i t peptide b inding to M H C CI I i n the presence o f E 3 / 1 9 K was demonstrated. Th i s showed that both peptide and E 3 / 1 9 K bound M H C C I I concurrent ly , reducing the l i ke l ihood that the associat ion was through the peptide b ind ing groove. Demons t ra t ion o f m u l t i p l e proteins b i n d i n g to M H C C I I suggested that large complexes o f proteins were formed i n the E R . E 3 / 1 9 K w i t h these complexes also occurred. Steric hindrance between members o f these large complexes seems l i k e l y . T o examine i f E 3 / 1 9 K and other molecules occurred as multimers, complexes were precipitated and separated on a sucrose densi ty gradient. These f indings lend further credence to the idea that other effects o f E 3 / 1 9 K remain to be discovered. 35 2. Materials and Methods 2.1 Cellular Methods 2.1.1 Tissue Culture The cell lines used in these studies are listed in table 2. M o s t experiments used either 293 cel ls or transfected 293 ce l l s . 293 cel ls are adherent human embryonic k idney cells transformed with A d 5 (121). Adherent cells were g r o w n to 85-95% confluence in D M E M supplemented wi th L - G l u (2 raM), 10 m M H E P E S ( p H 7.2), and 10% Fetal C a l f Se rum. An t ib io t i c s were omit ted f rom culture media . C e l l s were passaged us ing 0 .05% trypsin (wt/vol) in phosphate buffered saline ( P B S ) wi th 1 m M E D T A at 3 7 ° C , d i lu t ing cells 1:10 in fresh culture media. Non-adherent cells were g rown in flasks wi th the media supplemented as above. L i v i n g ce l l concentration was moni tored by trypan blue exc lus ion and cultures were g rown to a concentration o f l x l O 6 to 5 x l 0 6 ce l l s /ml before passaging at a 1:10 di lut ion. W i t h the exception noted be low, a l l cel ls were incubated at 3 7 ° C in a humid i f i ed 5% C 0 2 / 9 5 % air environment. Large scale cultures were g rown i n sealed B e l c o spinner flasks at 3 7 ° C in incubator not supplemented with C 0 2 . C e l l spinners were set at 60-70 rpm. M e d i a was supplemented wi th C a 2 + . C e l l concentrat ion was moni to red as above w i t h t rypan blue and was main ta ined between 1 x 10 6 and 5 x l O 7 c e l l s /ml . 36 Table 2. Reagents used Cell Lines 293 human embryonic kidney (ATCC CRL 1573) 293.12 293 transfected with E3/19K A549 human lung carcinoma (ATCC CCL 185) 293p39 Low passage 293 Monoclonal Antibodies (TCS) W6/32 mouse anti HLA A, B, C (ATCC HB 95) PA 2.1 mouse anti HLA A2, Aw69 (ATCC HB 117) OKT6 mouse anti transferrin receptor 34.5.8s + (34-l-2s) mouse anti D d (ATCC HB 102) 15.5.5s + (H100.27.55) mouse anti D k (ATCC HB 24) 34.1.2s mouse anti K d (ATCC HB 79) 16.3.IN mouse anti K k (ATCC HB 25) 28.14.8s mouse anti Db, L d (ATCC HB 27) Antisera R418 Rabbit anti E3/19K (122) R426 R425 Rabbit anti MHC CI I H chains (123) and calnexin Rabbit anti calnexin(124) anti transporter Rabbit anti TAP (125) Anti human (32m Rabbit anti human P2m (126) Lysis Buffers 0.6% Chaps 0.6% Chaps, PBS 1% NP40 1% NP40, 120 mM NaCl, 4mM MgCl 2, 20 m M Tris 1% N-Octyl Glucoside 1% N-Octyl Glucoside, 50mM NaP0 4 pH 7.0 PBS + 1% SDS 1% SDS, PBS w/o Mg2+ or Ca+ Media Hams F12 DMEM alpha MEM 37 2.1.2. Transfection 293 cells were transfected using Lipofectin (GibcoBRL). Genomic DNA in PBR322 was purified using the Promega MaxiPrep kit. 10 [Lg DNA was combined in 10 fold excess with a neomycin resistance marker and transfected following the manufacturers instructions. Cells were allowed to grow 36-48 hours before the selection agent, G418 (neomycin) (GibcoBRL) was added at a final concentration of 1 mg/ml. Dead cells were removed with frequent media changes. Surviving colonies were isolated with cloning rings. Individual clones were grown up and tested for gene expression. 2.1.3. Viruses Adenovirus 2 (Ad2) stock was obtained from Dr. Frank Graham through Roger Lippe. Stocks were grown in bulk on low passage 293 cells (121) in complete alpha MEM, harvested and stored at -80°C in 10% glycerol/PBS^at 2xl0 9 plaque forming units per ml (PFU). PFU were calculated using viral stock serially diluted onto A549 cells and overlaid with 2x F l l media and 1% Agarose. Plaques were scored 7 days after infection. Prior to use, virus stock was gently thawed at 4°C. Vaccinia virus containing the human P 2m gene was generously provided by Dr. J. Yewdell (National Institutes of Health, Bethesda, MD). Prior to infection cells were washed in PBS + +. Cells were infected at a multiplicity of infection (MOI) of 5. Virus was added to the cells and allowed to attach for one hour. Media was added back after the first hour and infection was allowed to proceed for approximately 18 hours. 38 2.1.4. Peptides Synthetic peptides were purchased from the Alberta Peptide Institute. Peptides were biotinylated at the N terminal using NHS-Ester biotin (Sigma). Peptides dissolved at 1 mM in NaPhosphate buffer (pH 7.5) were mixed with lOOx excess NHS-Biotin Ester for two hours. Unreacted NHS-ester was neutralised with Tris-HCl (pH 7.4) to a final concentration of 10 mM. Peptide (biotinylated and non-biotinylated) was added to growing cell cultures 24 hours prior to labelling at concentrations ranging from 5nM to 5p.M. 2.1.5. Fluorescence activated cell sorting (FACS) Adherent cells were pre-treated with versene or 0.05% trypsin and washed in FACS buffer to remove them from the plate. Cells were suspended at lxlO 6 cells/ml in DMEM supplemented with 200 mM L-Glutamine, 20 mM Hepes (pH 7.2) and 20 mM NaAzide (hereafter referred to as FACS buffer). Cell suspensions and all manipulations were performed at 4°C for the duration of the antibody labelling. Suspensions were washed twice with FACS buffer and incubated with primary antibody (200 pi of monoclonal TCS or 3 pi of ascites) for 45 minutes. Unbound primary antibody was washed out with two more FACS buffer washes and the samples were incubated with 100 jxl of 10 mg/ml stock fluorescein isothiocyanate (FITC) labelled secondary antibody (Jackson) for 45 minutes. Excess secondary antibody was washed out with two more FACS buffer washes followed by a wash with PBS with Ca 2 + or Mg + (PBS"). The cells were fixed in a 1.5% paraformaldehyde/PBS and stored in the dark at 4°C until analyzed. Analysis was performed on a Becton Dickson (BD) FACScan. 5000 gated events were collected and subsequently analyzed using the Lysis II software from BD. 39 An alternate protocol involved substituting FITC coupled secondary antibody with biotinylated secondary antibody. A third incubation using FITC coupled streptavidin was also added. 2.1.6. Animals Rabbits used for antisera generation were housed in the University of British Columbia south campus arrimal facility. They were kept under the care of Willem Schoorl according to the guidelines of the Canadian Council on Animal Care. Rabbits were initially injected with antigen in the lymph nodes and subsequently bled to raise antisera. When necessary they were sacrificed by C 0 2 asphyxiation. 2.2. Protein Techniques 2.2.1. Antibodies Monoclonal antibodies (mAbs) were used as Tissue Culture Supernatant (TCS) but in some cases ascites fluid was available. Rabbit antisera to E3/19K (R418) and to MHC CI I (R425 and R426) were donated by Dr. Sune Kvist. Anti calnexin antisera (124) was a gift from Dr. Bergeron, McGill University, Montreal, Canada and mouse TAP 1 and TAP 2 antisera were obtained from Dr. G. Butcher, AFRC, Cambridge, UK. Antisera to E3/19K was raised in rabbits using synthetic peptide coupled to KLH. 2.2.2. Metabolic Labelling An 85-95% confluent 65 mm plate with approximately Ixl0 6-2xl0 6 cells received 300 p.Ci 3 5S in 2 mis during labelling. Cell cultures were washed in methionine and cysteine free DMEM (Gibco) one hour prior to labelling and subsequently labelled with 150 \iCUm\ Pro-Mix 40 ( S-Met-Cys) (Amersham) for periods ranging from 5 min to one hour. In pulse chase experiments, the labelling media was replaced with standard media for the chase. All labelling was performed under standard cell culture conditions. Cells were lysed with a variety of lysis buffers (see table 2) supplemented with PMSF (40 p.g/ml) on ice. . Efficiency of labelling was determined by counting small aliquots (3 pi) of lysate following TCA precipitation of labelled proteins. 2.2.3. Immunoprecipitation Lysates were spun at 10,500 g for 30 minutes. Volumes of lysates were normalised to account for differences in labelling and aliquots ranging from 200 pi to 1ml were used for immunoprecipitation. All immunoprecipitation steps were performed at 4°C. Samples were pre-cleared by incubation with 3 pi Normal Rabbit Sera (NRS) for 45 minutes, followed by a 45 minute incubation with 45 pi of a 1:1 Lysis Buffer:Prot-A-Sepharose (Pharmacia) slurry. Specific antibodies were added to pre-cleared samples. In general, 100 pi of TCS from monoclonal cell lines or 1 pi of ascites fluid or sera was incubated with the samples for 45 minutes. Following specific incubation, 45 pi of a Prot-A-Sepharose slurry was added and allowed to incubate for a further 45 minutes. Beads were washed several times with buffers of varying salt concentration (three times with buffer B: 0.2% NP40, 150 mM NaCl, 2mM EDTA, 10 mM Tris, two times with buffer C: 0.2% NP40, 500 mM NaCl, 2 mM EDTA, 10 mM Tris, once with Buffer D: lOmM Tris) and finally sucked dry. Beads were then either stored at -80°C or boiled at 95°C for 5 minutes with bromomix (60% sucrose, bromophenolblue 0.1%, Tris-HCl pH 8.8 with freshly made 0.5M dithiothreitol)-20% SDS mixture. Samples were then blocked with 0.5mM 41 iodoacetimide, and separated with SDS polyacrylamide gels electrophoresis (SDS-PAGE). 2.2.4. SDS-PAGE Immunoprecipitated samples were separated on SDS-PAGE. Large gels were prepared as gradient gels with 10-15% polyacrylamide topped with a 5% stacking gel. 25 |il samples were loaded and run O/N at 18 mAmps (350 V). Gels of 10, 12 and 15% were also used. Upon complete migration of the dye front across the gel, gels were fixed (10% Acetic Acid, 30% v/v Methanol) for 30 minutes and then enhanced in 20mM NaSalicylate (known as enhance). Gels were dried and placed on Kodak X-AR radiographic film at -80°C for periods ranging from one day to several months. Alternatively, samples were separated on the BioRad Minigel (10 x 7 cm) system. Gels were made as above and samples separated according to manufacturers instructions. 2.2.5. Western Blotting As an alternative to fixing after PAGE, western blotting was performed. In addition to 1 4 C labelled molecular weight (MW) standards, biotinylated standards were also present on the gels. Following complete migration of the dye front in the electric field, the gel equilibrated in western transfer buffer () for 30 - 90 minutes. The gel was then sandwiched against a sheet of Immobilon P polyvinylidene difluoride (PVDF) membrane, submerged in western transfer buffer and blotted in an electric field for time periods between one and twelve hours. Alternatively, the BioRad mini western apparatus was used according to manufacturers instructions. After transfer, membranes were washed 2 times in western wash buffer (0.1% BSA, 0.05% Tween 20, 0.01M thimerasol in PBS), and blocked for one hour in BSA (2.5%) fortified western wash buffer. Membranes were washed 3 more times and exposed to antisera. Antisera was added at dilutions of 1:10 to 1:5000 in 45 mis total volume. Alternatively, the glass plate techniquewas used to conserve antisera. Briefly, the membrane was placed on a 42 glass plate with 100 pi of diluted antisera. Antisera was distributed over the membrane with a glass spreader every 30 minutes for two hours creating a thin film of antisera between the plate and the membrane. Following the two hour incubation, the primary antibody was washed away and the secondary goat anti rabbit coupled with biotin (Jackson) was added at a dilution of 1:10,000 for 45 minutes. Membranes were then washed several times and incubated with HRPO coupled to streptavidin. Subsequent detection was performed using Enhanced Chemiluminescence (Amersham ECL) system. Membranes were exposed to Kodak X-AR film for periods ranging from 5 seconds to overnight. Subsequently they were stored damp at 4°C for future use. 2.2.6. Tunicamycin Tunicamycin was used to block N-linked glycosylation prior to pulse-chase experiments. Cells were incubated with varying concentrations of tunicamycin in regular media for one hour. Media was replaced with labelling media (prior to addition of S-Met-Cys) supplemented with tunicamycin for one hour. Cells were then labelled in the presence of tunicamycin. Subsequent treatment of the cells is described below in section 2.2.7. 2.2.7. Cross linking Cross linking was performed during lysis with the addition of DSP [dithiobis(succinimidylpropionate)] (Pierce) to the lysis buffer. Cells were prepared, labelled and lysed as above with the exception that the lysis buffer contained DSP (200 Jig/ml). The lysis buffer used with DSP was PBS with 1% SDS. Crosslinking was stopped with 100 pi of 1M Tris-HCl pH 7.5 prior to spinning the lysate down. 43 2.2.8. Endoglycosidase H (Endo H) digestions Endo H (Boehringer Mannheim) was prepared according to manufacturers instructions with the addition of PMSF (800 (xl NaCitrate, 200 pi Endo H, 1 u.1 PMSF). Precipitates treated with Endo H were washed as above (2.2.7) except that in the final wash with Buffer D, beads were resuspended and divided into two aliquots prior to final drying. The first aliquot was incubated with 12.5 ul Endo H for 12 hours at 37°C. A second 12.5 ul was added for an additional 12 hours. The second aliquot was incubated in the same mixture without Endo H. Following a 24 hour incubation, the Endo H was washed out with Buffer D and the samples were prepared for SDS-PAGE as above. 2.2.9. Sucrose density centrifugation Sucrose density centrifugation was performed as described in the literature (127). Continuous 5% to 20% sucrose gradients were created using stock sucrose solutions. The detergents CHAPs, digitonin or N-Octyl glucoside were utilised in the lysis buffer and sucrose solutions. Gradients were made using the Pharmacia gradient maker. 300 ul of sample was placed on top of a 12 ml cushion in a Beckman SW 41 UltraClear tube and centrifuged at 4°C and 148,000g (36,800 rpm) for 28 hours in an SW41 rotor. Fractions were manually collected in 1 ml aliquots from the top of the tube. Samples were then diluted and run on SDS-PAGE. 2.2.10. Densitometry Densitometry was performed using the Molecular Dynamics scanner. Gels were prepared without enhance solution and exposed using the Phosphorlmaging (Molecular Dynamics) cassette. Computer analysis including densitometric traces were performed on the 44 resulting data. 2.2.11. Data Imaging Data were scanned and manipulated using Adobe Photoshop and MacDraw and printed on a Codonix printer. 45 3. Comparison of Differential Binding of E3/19K to M H C Class I allelic proteins 3.1 Introduction This study relates the transport rates of MHC CI I molecules to their ability to interact with the E3/19K molecule. E3/19K binds to every human allelic protein tested so far and all but one H-2 protein. Previous studies involved many groups of researchers using cell lines from different species (128-132). However, since the original E3/19K studies were performed, elucidation of the steps in antigen presentation has occurred. Many different proteins, some of which are allele specific, may be involved. Therefore the rate of maturation for different MHC CI I proteins is likely to vary when studied in different cell lines. Alleles of mouse MHC H-2 proteins were transfected into human cells to establish a model system with a consistent cellular background. 3.2 Rationale and Goals The goals of this study were to elucidate the binding requirements of E3/19K for MHC CI I. By comparing differences between those allelic forms of MHC CI I proteins that did and did not bind, it was hoped that a set of factors influencing binding would be elucidated. These data would also shed further light on the nature of adenovirus evasion of the immune system and further understanding of MHC function. 3.3 Results Previous studies had assessed the ability of E3/19K to bind to different MHC CI I allelic proteins in different cell lines from different species. Comparison of differences of E3/19K binding MHC CI I alleles in these experiments reflected many variables, some of 46 which related to differences between species. By transfecting many different MHC alleles into the same cell line, some of the variation due to species differences could be removed. At the same time, because of the allele specific nature of these effects, this also represents a limitation of the study; effects seen in 293 cells may be completely different from those seen in other cell lines. 3.3.1. Transfected H-2 allelic proteins are expressed at different levels at the surface of 293 cells In order to establish the relative level of surface expression of the H-2 allelic proteins in 293 cells, FACS analysis was performed on the transfectants. Several clones from each transfection were analyzed, and the clone with the highest expression for each transfected allele of mouse MHC CI I genes was used in all subsequent experiments. The decision to use the clone with the highest level of expression was based on the assumptions that E3/19K had an innate ability to bind particular alleles of MHC CI I and that large amounts of mouse MHC CI I proteins expressed in 293 cells would provide more ligand for E3/19K. Limiting amounts of E3/19K seemed unlikely since in adenovirus infected cells, E3/19K is expressed in excess. Data in Figure 9a shows that these transfectants express the H-2 proteins at the surface in varying levels. The level of endogenous HLA expression in untransfected 293 cells is shown on the left (labelled HLA). While some of the allelic proteins show low levels of expression, all show a detectable level of expression. As a control for transport to the cell surface, in Fig 10c the cell surface levels of the transferrin receptor show that transport to the cell surface is not disrupted. To determine whether low surface expression of some MHC CI I proteins was due to competition for factors with endogenous HLA, the level of HLA-A, B and C expression was assessed using W6/32. A comparison of the W6/32 cell surface expression in the transfected and untransfected 293 cells is shown in Figure 9b. These data showed that in every 47 transfectant except the K d transfectant, the level of cell surface expression of the W6/32 epitope was reduced. The level of reduction of W6/32 expression indirectly allowed the monitoring of the effect of the foreign MHC CI I molecule on the endogenous MHC CI I. In both the D b and K b transfectants the level of W6/32 expression was reduced to a level of 58% and 5%, respectively, of untransfected 293 cells. The K d transfectant alone showed an increase in the level of W6/32 expression; this was due to cross reactivity of W6/32 with K d (133). Comparison with levels of K d expression assessed with the K d specific monoclonal 34.1.2s and W6/32 expression in the non-transfected 293 cells suggests that the K d and endogenous expression were additive. Reduction in cell surface expression of endogenous MHC CI I proteins could reflect a competition for limited resources such as 32m or peptides. Mouse MHC CI I proteins have a higher affinity for human P2m than endogenous MHC CI I (133). Excess mouse MHC CI I may leave little P2m available for endogenous HLA alleles, resulting in lowered W6/32 expression. The phenomenon of internal competition modulating HLA expression is previously unreported. 48 Figure 9. FACS analysis of 293 cells and 293 transfectants Confluent plates of 293 cells and 293 transfectants were analyzed by FACS. Cells were harvested as described in Materials and Methods. In all cases the untransfected 293 cell data is in the first column and labelled HLA. a) Levels of cell surface expression of transfected MHC CI I was determined by incubation with allele specific antibodies as listed in table 2. Calculation of arbitrary fluorescence units (AFU) involved taking the difference between mock treated and specifically labelled FITC labelled samples. . b) Levels of endogenous HLA-A2 expression was assessed using the monoclonal antibody W6/32. c) Transferrin receptor levels were determined using the OKT9 monoclonal antibody. 49 S p e c i f i c CI I E x p r e s s i o n Allele W 6 / 3 2 E x p r e s s i o n 293 transfectant 0 K T 9 E x p r e s s i o n 3 5 0 / " 293 transfectant 50 3.3.2. Transport of H-2 allelic proteins to the cell surface of 293 cells To determine levels of expression of transfected MHC CI I proteins, monoclonal antibodies were used to precipitate MHC CI I from cells pulsed for 15 minutes and chased for 0, 30, 120, 240 and 480 minutes. During the course of the chase, the mobility of the precipitated MHC CI I decreases, reflecting modification of the proteins. MHC CI I proteins undergo processing from a 43 kDa protein at two hours to a 46 kD MW form present at four hours of the chase. The rate limiting step for transport to the cell surface is escape from the ER. To assess whether the MHC CI I proteins had left the ER and migrated to the cis-Golgi, Endo H digestion was performed on half of the MHC CI I precipitates. Glycans of glycoproteins are modified from a high mannose form to a more complex Endo H resistant form in the cis-Golgi. The analysis shows that a portion of the population of D b molecules exit the ER (as demonstrated by their insensitivity to Endo H) after two hours. However, a large proportion of the D b molecules remains Endo H sensitive after four hours, suggesting an extremely slow transport rate. The half life (t J / 2) for this protein to leave the ER is greater than four hours. In the D d transfectant, the D d molecules achieve a mature conformation after four hours. D d is present in an immature lower molecular weight form for most of the four hour chase. The higher molecular weight form appears after four hours of chase. In the samples treated with Endo H, most of the sample has disappeared from all but the 480 minute timepoint. This likely reflects an unstable D d MHC CI I protein, with most of it falling apart during the 24 hour, 37°C Endo H incubation. The Endo H resistant form visible at the final four hour chase point suggests that the mature protein is more stable. The L d molecule is processed marginally over the four hour chase. It appears that the epitope recognized by the monoclonal used may not be stable under the conditions of the 24 hour incubation with the Endo H buffer. 51 The K b molecule is partially Endo H resistant after about one hour and is therefore rapidly transported. Analysis of the K d precipitation demonstrates that an epitope recognised by the conformation specific mAb for the K d protein does not appear until 30 minutes post pulse. Endo H digestion reveals that the forms observed at 30 minutes and one hour have not exited the ER; by two hours, much of this molecule is Endo H insensitive and at the end of four hours, almost no Endo H sensitive K d molecules remain. A similar pattern is observed for K k. Endo H digestion reveals that the K k protein has left the ER by two hours post pulse, suggesting that it is even more rapidly transported than Kd. 52 Figure 10. Levels of expression and rates of transport of H-2 M H C CI I proteins transfected into 293 cells Lysates from confluent 293 cells transfected with different mouse MHC CI I (H-2) proteins were precipitated with allele specific monoclonal antibodies as listed in table 2. Cells were pulsed with Pro-Mix (35S Cys-Met) for 30 minutes and chased for 0, 30, 60 120 and 240 minutes (lanes 1 to 5 respectively). Aliquots which have been incubated with Endo H are in lanes 6-10. Levels of expression for each allele were determined. MHC CI I proteins with different mobilities reflecting differently processed forms are labelled with arrows. Precipitates were split and alternately mock treated (lanes 1-5) or treated with Endo H (lanes 6-10) for 24 hours and then separated by SDS-PAGE. Endo H resistant proteins have reduced mobility and are labelled by the upper arrow. 53 Once expression of H-2 proteins in human 293 cells was established, the effect of the E3/19K on the different proteins was examined. Initially, it was hoped that expression of the transfected mouse H-2 alleles would be uniformly strong and that this level of expression would provide a clear picture about whether E3/19K bound or not. In transfectants with MHC CI I allelic proteins that bound E3/19K, reduced cell surface expression was expected. In transfectants with proteins that did not bind, no reduction in cell surface expression was expected. For proteins that bound less effectively, it was postulated that temporary association with E3/19K may result in slower transport to the cell surface. It should be noted that the level of expression for each allele was quite variable in the 293 cells. This variable expression probably reflects the accessory proteins and chaperones available for the transfected allele. Additionally, as shown in Fig lib, the expression of the transfected allele did affect the cell surface expression of the endogenous MHC CI I alleles. Highly expressed transfectants such as K b cause a large reduction in cell surface expression of HLA A, B and C detectable by the conformation specific monoclonal antibody W6/32. This could reflect limited P2m or peptide. 3.3.3. Co-precipitation of the adenovirus E3/19K protein with M H C CI I in > 293 transfectants To address the binding capacity of H-2 proteins to E3/19K, the transfectants were infected with Ad2 at a MOI of 5 for 18 hours and metabolically labelled with a pulse of 3 5S methionine and then chased for 0-4 hours. After lysis, the H-2 allele expressed in each transfectant was immunoprecipitated and examined by SDS-PAGE and autoradiography. Controls were carried out to confirm that the infected transfectants expressed equal amounts of the E3/19K protein (appendix 1). Half of the precipitated samples were digested for 24 hours 55 at 37°C with Endo H to determine whether the MHC CI I molecules had acquired mature N-linked glycans. Results are shown in figure 11. E3/19K is detected by co-precipitation with MHC CI I in Ad2 infected cells. While confirmation of the identity of E3/19K by subsequent western blotting was not performed in the mouse MHC CI I transfectants, comparison with parallel immunoprecipitations using the 293.12 cell line helped to establish the identity of E3/19K. 293.12 cells have transfected with the EcoRI d fragment containing the E3/19K protein from Ad2 and exhibit reduced cell surface expression of MHC CI I in FACS experiments. Comparison of immunoprecipitation of MHC CI I from 293, Ad2 infected 293 cells and 293.12 cells reveals a protein corresponding to E3/19K(131). After a chase of 2 and 4 hours, the D b molecule co-precipitates a protein at about 25 kDa that corresponds to E3/19K. A corresponding band also appears at the 2 and 4 hour time points in the Endo H digested samples. In the Endo H digested samples, the band corresponding to E3/19K has the lower MW consistent with digestion of the N-linked sugars on this molecule.The three bands in the Endo H digested samples correspond to differences in sensitivity in the two carbohydrates present on E3/19K. Co-precipitating with D b is the P2m molecule, suggesting that transfected D b proteins are being assembled with human P2m. The appearance of P2m in the samples that have been exposed to 37°C for 24 hours in the Endo H treatment suggests that the D b molecule has adopted a stable conformation. The D d molecule does not co-precipitate with E3/19K. The Endo H digestion of D d reveals it exits the ER at the same rate as in the uninfected cells. An Endo H resistant protein is observed at 4 hours. The L d molecule also does not co-precipitate E3/19K during the four hour chase. However, in both infected and uninfected cells, the L d molecule remains Endo H 56 sensitive throughout the four hour chase. The K b protein co-precipitates E3/19K. E3/19K appears to co-precipitate instantaneously with the K b molecule appearing at all time points. At the 2 and 4 hour time points, the molecular weight of E3/19K has decreased marginally, suggesting carbohydrate processing. The Endo H digestion shows that two more forms of E3/19K appear after two hours; they differ in carbohydrate attachments. After two hours, a second protein at about 22 kDa appears. At four hours post pulse an additional band at the same molecular weight as the untreated precipitates appears. This band likely represents an Endo H resistant form of E3/19K. The K d molecule co-precipitates E3/19K strongly at all timepoints. K d is assembled by about 30 minutes post chase. At this point, the heavy chain and P2m is observed. The K d molecule remains Endo H sensitive throughout the entire four hour chase. Once again, two forms of partially Endo H resistant E3/19K are observed co-precipitating with Endo H sensitive K d molecules. These data suggest that E3/19K is cycled back and forth through more mature compartments such as the cis Golgi and is able to bind MHC CI I molecules in an earlier compartment and retain them there. Differences in the maturity of the carbohydrates on E3/19K bound by H-2 MHC CI I proteins suggests differing affinities for E3/19K based on its processing. The K k molecule does not co-precipitate with E3/19K. The K k is entirely Endo H resistant after about four hours. To summarize results from figure 11, the E3/19K protein clearly co-precipitates with Kb, Kd, and D b proteins. L d does not co-precipitate E3/19K during the chase times performed in this experiment. The Endo H sensitivity of Kd, Db, and K b during the four hour chase confirms that they do not exit the ER during this period and thus the E3/19K binding of these 57 molecules. The Kk, and D d molecule are Endo H insensitive after four hours. Comparison with Figure 10 shows differences between MHC CI I transport in adenovirus infected and non infected cells. E3/19K had no effect in retarding the egress of these allelic forms of MHC CI I proteins from the ER. The D d protein is able to achieve Endo H sensitivity at the same rate that it did in non-infected cells (Figure 10), becoming Endo H resistant after 4 hours. Interaction between D d and E3/19K is likely to be weak or nonexistent. 58 Figure 11. The effect of Ad2 infection on the rate of transport of transfected MHC CI I proteins in 293 cells 293 cells and 293 transfectants were infected with Ad2. Infection was at MOI of 5 eighteen hours prior to labelling with Pro Mix (35S-Met/Cys). The cells were pulsed for 30 minutes and chased for 0, 30, 60 120 and 240 minutes (lanes 1 to 5) respectively. Aliquots which have been incubated with Endo H are in lanes 6-10. MHC CI I was precipitated with allele specific antibodies as listed in table 1. Samples were divided and alternately mock treated or treated with Endo H for 24 and then separated by SDS-PAGE. E3/19K migrates at -25 kDa and is denoted by arrows. (32m also co-precipitates and is seen on the gel at 12kDa. 59 3.3.4. Adenovirus infection of 293 transfectants reduces cell surface expression of M H C CI I proteins To assess the effect of E3/19K on cell surface expression, FACS analysis of Ad2 infected transfectants was performed (figure 12). While the cell surface expression of transfected H-2 proteins is variable, when infected with Ad2, cell surface expression of many of these MHC CI I proteins is reduced. The FACS data show that in the untransfected 293 cells, infection with Ad2 causes reduction of the cell surface HLA (W6/32 epitope) to 62% of uninfected expression. The difference in cell surface expression of the transferrin receptor in infected and mock infected cells is shown in figure 12b demonstrating that the effect of adenovirus infection is not blocking all transport to the cell surface. In the transfected 293 cells, Ad2 infection causes the cell surface expression of the H-2 transfectants to be reduced as well. A comparison of levels of cell surface expression of H-2 proteins in infected and uninfected cells shows which alleles are susceptible to E3/19K binding. Figure 12 shows the greatest level of reduction of cell surface expression was highest for D b which expressed 43% of the D that the uninfected transfectant expressed. L also showed a large reduction at 44%. At the other end of the spectrum, the proteins that were affected the least were K k and Dk, which showed 93% and 95% of uninfected expression respectively. The cell surface expression of K d (62%), K b (73%) and D d (81%) were moderately inhibited. These data support the hypothesis that amongst the H-2 molecules Kd, D b and L d bind to the E3/19K molecule with the greatest affinity. However, it should be noted that while a FACS experiment represents 5000 individual events, the results are subject to some presorting. Sampled cells are gated by size and cells are selected to represent a healthy population. While size selection is performed prior to observing 61 results, the outcome of the experiment could reflect arbitrary sampling of a larger, more diverse population. 62 Figure 12. The effect of Ad2 infection on M H C CI I expression in 293 cells and 293 transfectants a) 293 cells and transfectants were alternately infected or mock infected with Ad2 at MOI of 5. Infection was started eighteen hours prior to harvest for FACS analysis. Cell suspensions were specifically labelled with MHC CI I allelic protein specific antibodies as listed in table 1. Suspensions were analyzed using flow cytometry. Arbitrary fluorescence units (AFU) are the difference between mock treated and specifically labelled FITC labelled samples for each transfectant. The first bar of each pair shows the uninfected cell surface expression; the second the level of cell surface expression in infected transfectants. 63 B u n i n f e c t e d i • ArJ2 2 9 3 transfectant 64 b) The levels of the transferrin receptor at the cell surface are compared in infected and uninfected transfectants. Transferrin Receptor Expression • mock • infected Transfected Allele Results representative of these experiments are summarized in table 3. As determined by immunoprecipitation, Kd, K b and D b all co-immunoprecipitate with E3/19K, but the products of alleles K k and D d do not. The data for L d is less conclusive. Column two shows the reduction of cell surface expression of these proteins as detected in FACS analyses. For the proteins that co-immunoprecipitate E3/19K, a reduction in cell surface expression is noted. Reduction in cell surface expression results range from 43% of the level of uninfected 65 expression to 73%. For the L d allelic protein, cell surface expression of 44% that of uninfected is observed in the presence of E3/19K (Ad infection). This is suggestive of down regulation due to association with E3/19K. Surface expression of D k is unaffected by Ad infection. Finally, in the third column, the rate of exit from the ER assessed by the Endo H sensitivity (or the difference in Endo H sensitivity) of MHC CI I proteins from infected and non-infected cells is summarized. In this column, it is observed that there are no differences in Endo H sensitivity in the K k and D d proteins in infected and non-infected cells. Table 3. Summary of E3/19K association with H-2 allelic proteins Mouse Allel ic proteins Co-IP (a) F A C S data (b) Endo H * sensitivity Binders K d + 62% + K b + 73% + Slow-binders D b +/- 43% +/-L d - 44% -Non-binders K k - 93% -D d - 81% -D k - 95% 9 * After 4 hours An overview of all the techniques employed gives an indication of the ability of each H-2 transfectant to associate with E3/19K. The highest expressing clone for each transfectant was selected and tested above. Based on each technique, the H-2 proteins tested are grouped into three categories. The binders show co-immunoprecipitation (a). The slow-binders are not observed to co-immunoprecipitate, but do result in reduced cell surface expression as seen in the FACS experiment demonstrated here (b). Finally, the non-binders have not been shown to co-precipitate and also do not reduce the cell surface expression of MHC CI I. 6 6 3.4 Discussion The strong point of this study is that it is the first in which the binding of the H-2b,d, and k proteins to the Ad2 E3/19K protein in the context of a common cellular background has been examined. This set of experiments made the initial assumption that variable cell surface expression of transfected mouse genes was likely due in part to varying transport rates through the cell. These transport rates likely reflect incompatibilities between human antigen presentation proteins and mouse MHC CI I alleles. Additionally, it is possible that accessory proteins required for mature, stable MHC CI I molecules may also be missing or of an inadequate fit. The initial assumption made in this set of experiments was that overexpression of MHC CI I alleles would provide more ligand for potential E3/19K binding limits the conclusions that may be drawn from the data. While overexpression of transfected proteins in these experiments results in clear demonstrations of E3/19K binding and non-binding for some allelic proteins, in other cases it results in unstable, short lived MHC CI I proteins. The instability observed in Endo H digestions suggest that some of the transfected alleles may be very short lived in 293 cells. The ability of E3/19K to bind these alleles is likely low and reflects limits of the expression system rather than the ability to bind these alleles. These instability seen in these MHC CI I proteins may be due to our detection system. Perhaps the monoclonals selected do not recognize or act to destabilize transfected MHC CI I proteins. Alternately, some MHC CI I-P2m-peptide complexes may be more susceptible to disruption by the detergents used in these experiments. This set of experiments revealed many problems associated with protein trafficking and re-establishment of antigen presentation by mouse MHC proteins in human cells. Further experiments which would more finely control the many variables could include a comparison 67 of unfolded or improperly folded proteins to those detected by conformation specific monoclonal antibodies. This would allow some comparison of total levels of expression to transport of mature proteins out of the transfected cell. A wider sampling of the clones of transfectants to include those which expressed mouse MHC CI I at lower levels would allow a better assessment innate affinity between MHC CI I proteins and E3/19K. Finally, to broaden the scope of the conclusions, this experiment could be repeated in other cell lines. Within the constraints outlined above, this study has identified those allelic proteins that interact with E3/19K in our model system. This study also extends those studies done earlier with Kd, K k and the hybrid MHC CI I molecules containing regions of K k spliced to regions of K d molecules (131). Furthermore, this study has identified a new sub-group of slow binding molecules. The identification of those allelic MHC CI I proteins that bind E3/19K has been determined in three ways. Co-immunoprecipitation of E3/19K with MHC CI clearly demonstrates binding. In addition to co-immunoprecipitation, the effect of E3/19K is assessed by comparing the rate of exit of MHC CI I proteins from the ER in infected and mock infected transfectants. The third method to determine binding is by detecting reduction of MHC CI I expression at the cell surface. Taken together, these methods identified which allelic proteins were binders, non-binders and slowbinders. These data are summarised in table 3. The first method shows E3/19K clearly binds to Db, K d and K b but not Kk, D d and L d. The second method confirms that E3/19K does not slow the exit from the ER of D d and Kk. Differences in Endo H sensitivity between infected and mock infected transfectants are observed for Db, K d and Kb. L d is poorly transported to the cell surface, but is more stable (i.e. does not fall apart during Endo H treatment) in Ad2 infected cells. 68 FACS analysis shows a reduction in cell surface expression for all transfected allelic proteins in the presence of E3/19K. While overall expression is of transfected alleles is quite variable, the percentage difference between infected and uninfected expression was compared. This comparison shows the level of reduction of MHC CI I cell surface expression is not as dramatic for the K k and D d alleles as for Kd, Kb, Db, and L d. While the data used here was from a single experiment, repetitions of the FACS experiments demonstrated that while overall cell surface expression varied greatly between experiments, the differences in expression due to E3/19K (between infected and non-infected) remained fairly constant, varying no more than 10%. The results above suggest that the nature of the L d association with E3/19K is different from that of other allelic proteins. The first method employed shows that over the course of a four hour chase, L d does not co-immunoprecipitate E3/19K. This could be due to the fact that the monoclonal antibody used interferes with E3/19K binding. It could also be because that over the course of the four hour chase, L d does not fold into a conformation that promotes binding to E3/19K. Figure 10 shows that during a four hour experiment, only a limited amount of this protein is detected and Endo H digestion and exposure to 37°C for 24 hours suggests that what little is there is quite unstable. FACS analysis confirms the level of cell surface expression of the L d protein is lower than that of other allelic proteins, but is still detectable. FACS analyses, which assay a steady state level of expression, demonstrate that L d does fold into a mature conformation and can be expressed on the cell surface (Figure 9a), and that this level of expression is reduced by E3/19K (figure 12). This finding allows the proposal of a new subgroup of allelic proteins that associate with E3/19K. L d (and D b) illustrate the subgroup of allelic proteins that is slowly transported out of the ER and will bind E3/19K, but only once having achieved a transport competent 69 conformation. These proteins can be called "slow-binders" because the rate at which they fold this mature conformation and are able to bind E3/19K is greatly reduced. Slow-binders are characterised by their slow maturation rate and are transported out of the ER quite slowly. It is likely the length of time that they are present in the same compartment with E3/19K that enhances the opportunity to associate and be retained. The D b protein is another member of this "slow binding" group. One hour after the beginning of the pulse, only trace amounts of E3/19K are co-immunoprecipitated; it is only after four hours that significant amounts of E3/19K are seen. In contrast, K d binding strongly to E3/19K occurs almost immediately, or shortly after translation. The E3/19K molecule co-precipitates even at time zero suggesting that E3/19K made during the pulse can bind already pre-existing K d molecules. In addition, at the 30 minute timepoint, K d is still available for the E3/19K to bind. This contrasts sharply with the D and L molecules that appear to mature (acquire P2m, peptide, dissociate calnexin) more slowly than Kd. However, once a mature conformation has been achieved, E3/19K binds the "slow-binders" with great efficiency. The conformation of D b that E3/19K binds is initially not present in the D b transfectant, but appears with subsequent maturation. This suggests that E3/19K binding is dependant on overall MHC CI I conformation rather than specific residues. This view is in contrast with previous studies that have attempted to map specific residues crucial for E3/19K binding (128, 131). In cells transfected with the MHC CI I alleles classified as slow-binders, the level of endogenous MHC CI I (or W6/32) cell surface expression is quite low. The cellular components that participate in the maturation of MHC CI I molecules (peptide, peptide transporters, calnexin, and other chaperones) may be occupied with or competed for by the slow binding H-2 MHC CI I and therefore have reduced availability for the endogenous HLA. The reduction of endogenous MHC CI I indicates that there is some limit to the number of 70 MHC CI I proteins that can be correctly folded and/or expressed by a cell. The K k protein is one of the allelic proteins which does not bind E3/19K. It is also exits the ER very quickly. The t{/2 for Endo H resistance is approximately 15 minutes; the average for endogenous MHC CI I is 45 minutes to 1 hour. Rapid exit from the ER may reflect a lack of competition for cofactors such as P2m or peptide resulting in very rapid folding. Alternately, it could reflect a reduced ability for the human chaperones present in 293 cells to recognize and retain the K k protein and ensure quality control. However, K k has limited exposure to E3/19K (reduced period of time) and the opportunity to associate may simply not occur. The inability of K k to bind E3/19K may not due to lack of affinity but rather lack of exposure. If E3/19K binding to MHC CI I is partly due to being in the right place for the appropriate period of time, then the contribution of affinity between these proteins becomes less clear. In studies where E3/19K is precipitated directly, the ratio of E3/19K to MHC CI I is always much higher than when MHC CI I is precipitated. Clearly there exists a large pool of E3/19K that is not binding to MHC CI I. Conversely, the level of E3/19K expression within a cell correlates inversely to the level of cell surface MHC CI I expression (134). While affinity between MHC CI I proteins and E3/19K has been noted, an estimation of the relative strength of this affinity has never been made. Secondly, the period of time required for maturation, and hence the length of exposure to E3/19K is influenced by many other factors including but not limited to availability of correct peptide and the ability of chaperones such as calnexin, calreticulin, tapasin, TAP, glucosyltransferases, protein disulphide isomerase and others to bind to mouse MHC CI I proteins. Our evidence shows that for the small sample of alleles expressed in 293 cells, great 71 variations in stability were observed. Our findings are limited by the sources of variability inherent in our expression system. This variability reflects the complexity of antigen presentation and protein folding in general. Some specific limitations include the level of expression of the transfected allele. As mentioned above, endogenous HLA cell surface expression was reduced when mouse allelic proteins were expressed suggesting that there is a limit to the total MHC CI I that can be processed by a particular cell. Overexpression of transfected genes may perturb regulation of this limit. Secondly, species differences in proteins involved in antigen presentation may affect the expression of stable conformations of mouse MHC CI I proteins in human cells. However, it is likely that such limitations will be present in any study of E3/19K binding MHC CI I proteins and that a better picture of MHC CI I binding by E3/19K could be obtained by using multiple expression systems. The binding of E3/19K by MHC CI I allelic proteins was examined against a common cellular background. 293 cells were chosen as they are amongst the best characterised cell lines for Ad 2 infection and are easily infected. The highest expressing clones were chosen to be representative for each allelic protein, with differences in levels of gene expression observed for some alleles. However, when infected with Ad 2, levels of E3/19K expression are well in excess (appendix 1); therefore limited access to E3/19K is unlikely a factor in this study. Differences in allele expression may influence proper folding and exit from the ER. Based on the this set of experiments, a speculation about E3/19K binding can be made. E3/19K binds to those MHC CI I molecules that it is exposed to for. MHC CI I alleles which are normally processed through the ER remain there for the period of time necessary for peptide loading and proper folding. Chaperones such as calnexin serve effect the quality control. MHC CI I molecules that fold quickly probably escape the attention of E3/19K. 72 Rapid folding may reflect bypassing the quality control mechanism altogether. Reduced affinity between chaperones and MHC CI I proteins and could result in empty or non-functional MHC CI I at the cell surface which in adenovirus infected cells, are not bound by E3/19K. Figure 13 outlines the different fates for MHC CI I. The different paths that can be taken are represented by arrows (labelled 1, 2 or 3). The length of the arrow gives an indication of the amount of time the process takes. The dashed arrow indicates an expedited path. The binders are assumed to follow the first path. They are efficiently and rapidly transported through the ER and Golgi to the cell surface. The slow-binders take the second path. These allelic proteins are less efficiently transported and spend more time in the ER. Finally, the non-binders are very rapidly transported to the cell surface and may possibly bypass the processing machinery altogether. This results in low chance of contact with E3/19K. 73 Figure 13. Overview of maturation paths for the M H C CI I complex The maturation path through the ER and Golgi is illustrated. Different potential paths are shown for each of the types of MHC CI I proteins based on their ability to bind E3/19K. Binders, non-binders and slow-binders are depicted. 1) Binders travel through the ER and Golgi in the most straightforward manner. Interaction with chaperones in the ER and Golgi ensures proper processing and maturation. 2) Slow-binders interact with the same chaperones, but the rate of transport through the ER is slower. This is depicted as a circuitious route through different ER compartements. 3) Non-binders travel quickly through the entire ER-Golgi maturation pathway, arriving at the cell surface very rapidly. This is depicted with a dashed line. 74 C o m p l e x of H L A c lass I and pept ide Cell surface 1 HLA § P2m Q Calreticulin @ ERp57 B Tapasin fl _ heavy chain 75 4. Efforts to increase the cell surface expression of H-2 D b 4.1. Introduction In the previous chapter experiments with different H-2 transfected 293 cells were discussed. Of the transfectants studied, the 293D transfectant was the most intriguing. The D b protein was unusual because while the cell surface expression was quite high, a much larger proportion of the D b molecules remained inside the cell and was Endo H sensitive. This protein was slowly transported through the ER with the majority remaining intracellular. Other proteins seemed to mature through the ER properly, indicating that mouse allelic proteins can be successfully transfected and expressed in the human 293 cells. Much of the D b that is made is still within the cell at four hours post pulse. In the last chapter it was shown that in transfected cell lines, surface expression for endogenous MHC CI I was reduced. This suggested that the H-2 proteins were competing for factors required for MHC CI I maturation. The limit for levels of cell surface expression in the 293Db transfectant could be a shortage of an essential co-factor required for proper conformation. Improperly folded H-2 Db proteins could be prevented from escaping from the ER. Previous studies in cell lines exhibiting low cell surface expression revealed many strategies to restore expression to regular levels. These studies were instrumental in determining which components of a cell were crucial for regular expression of the MHC CI I molecule. In general, the deficiencies in these cells could be broken down into two or three categories. In order to determine the nature of the incomplete expression of the D b protein, strategies previously applied to other surface expression mutants were utilised. Some of the earliest MHC CI I cell surface negative mutants characterised were those that lacked P2m. The Daudi cell line, a mouse cell line lacking p^m and consequently MHC CI 76 I cell surface expression, was restored to a normal phenotype when fused with a cell line expressing P2m (135). Other experiments also demonstrated the dependence of the H chain on the P2m for cell surface expression (136). Another effective treatment for cells with reduced cell surface expression is the incubation with y-IFN (137). y _IFN is part of a normal cellular antiviral response; many elements of the antigen presentation pathway have interferon regulatory elements in their promoters. The antiviral effects of y-IFN have been widely documented (138). The effects of y-IFN include widespread upregulation of antigen presentation elements. y-IFN reactive elements upregulate transcription of many of the factors necessary for MHC CI I gene upregulation. These factors include P2m, MHC CI I, TAPs and LMPs. Upregulation of these elements enhances the ability of infected cells to present antigens at the cell surface. In cells that present little MHC CI I at the cell surface, y-IFN increases expression. The mutant mouse cell line CMT-64 has reduced MHC CI I expression at the cell surface due to missing transporters (139). Similar results have been reported with other cell lines (140). Restoration of MHC CI I expression has been achieved by transfecting the cell line with TAP 1, TAP 2 and both TAP 1 and TAP 2 (47). In the absence of transfected TAPs, cell surface expression can be stimulated by treating these cells with y-IFN. Treatment with y-IFN upregulates the transporters, as well as P2m and MHC CI I, resulting in cell surface expression (139). Studies with the TAP 2 defective mouse cell line RMA-S and the parental line RMA demonstrated that a population of MHC CI I molecules devoid of peptide exists. These 'empty' MHC CI I molecules are expressed at the cell surface (141). Indeed, some of the 77 earliest studies with peptides and MHC CI I were done showing RMA-S cells becoming sensitized to CTL killing by the addition of extracellular peptide (142). These cells have a mutation in the TAP 2 peptide transporter and are unable to form a TAP 1/ TAP 2 heterodimer required to transport a full complement of peptides into the ER. Many of the D b and K b molecules in this cell line are never loaded with peptides. They are quite unstable and consequently expressed at a low levels on the cell surface. Growing cells at 26°C increases the thermodynamic stability of characteristically unstable MHC CI I molecules devoid of peptide and permits them to survive longer at the cell surface. The phenomenon of empty MHC CI I molecules on the cell surface was shown to be normally occurring when empty MHC CI I molecules were not only found on mutant cells growing at low temperature but also found on the surface of the non-mutated parental cell line RMA (35). One of the mechanisms above might be responsible for the lack of cell surface expression of the D b protein in the transfected 293Db cell line. 4.2 Rationale and Goals The goals of this study were to examine a MHC CI I protein that remained intracellular. Several different strategies were employed to determine if intracellular stores of the D b protein could be coerced to travel to the cell surface. These studies would shed light on the nature of antigen presentation and also protein maturation and transport. The analysis of regular transport of cell surface bound glycoproteins may allow elucidation of several of the mechanisms present for quality control in protein transport through the ER and Golgi. 78 4.3 Results 4.3.1. The D protein remains intracellular in the 293D transfectant Endo H analysis of the 293Db transfectant reveals that the D b protein is well expressed. Figure 14 shows large amounts of D b expressed imunoprecipitated from lysates of these cells. However, Endo H digestion reveals that only a small fraction of the transfected H-2 D protein is able to exit the ER. This figure shows a pulse-chase experiment. At the right side of the figure the lysates have been incubated with Endo H. Most of the lysates are Endo H sensitive; at 2 and 4 hours post chase Endo H resistant D b is observed (upper arrow). 79 Figure 14. Intracellular accumulation of H-2D b in 293 cells Conf luent 293 cel ls transfected wi th M H C C I I H - 2 D b (293Db) were pulsed for 30 minutes and chased for 0, 30, 60 120 and 240 minutes (lanes 1 to 5) respectively. M H C C I I was precipitated wi th protein specific antibodies as listed in table 1. Samples were d iv ided and alternately m o c k treated or treated wi th E n d o H for 24 and then separated by S D S - P A G E . E n d o H sensitive proteins are identified by the lower arrow; E n d o H resistant proteins by the upper arrow. 293D b . EndoH 1 2 3 4 5 1 1 2 3 4 5 80 4.3.2. Growth at 26° C does not result in the bulk of the intracellular stores of the D b protein escaping the ER In figure 15 the effects of growth at 26° are shown. The MHC CI I allele D was precipitated and half of the lysate was incubated with Endo H.Culture at 26°C results in many contaminating proteins co-immunoprecipitating with the mouse MHC CI I protein. These contaminants are characteristic of precipitates from temperature shocked cells; incubation at 37°C for 24 greatly reduces the levels observed suggesting that they are quite unstable. While growth at 26°C results in more Endo H resistant D b proteins, most of the population remains Endo H sensitive. It was reasoned, that 293 cells may lack the peptide repertoire bound by the D b protein. Only the earliest studies examining peptide transporters had suggested there existed allele specificity to the peptide transporters (38, 143). However, if the D b molecules in the 293 cell were unable to obtain the correct peptide to form a trimolecular complex, they would be unable to achieve the correct conformation to exit the ER. This would result in large amounts of the D b population remaining in the ER. Alternately, D b complexes devoid of peptide may be formed, but be thermodynamically unstable and hence very short lived. These short lived species would likely be too unstable to be detected outside of the ER. It had previously been demonstrated that unstable MHC CI I molecules lacking peptide could be stabilised and detected at the cell surface by culturing cells at 26°C (144). Table 4 shows the FACS results from growing D b cells at 26°C and 37°C. The level of D b at the cell surface grown at 26°C is comparable to the cell surface expression grown at 37°C. 81 Figure 15. The effect of growth at 2 6 ° C on the intracellular accumulation of H - 2 D b in 293D b cells 293 Db transfectants were alternately cultured at 2 6 ° C or 3 7 ° C for 24 hours pr ior to l abe l l ing . C e l l s were pu lsed for 30 minutes and chased for t w o hours at 2 6 ° C or 37°C respect ively. C e l l s were l y sed on ice and lysates precipitated w i t h a Db specific monoclonal antibody (as listed i n table 2). P r io r to S D S - P A G E , precipitates and molecular weight ( M W ) standards were either treated or m o c k treated wi th E n d o H for 24 hours. 293 D b @ 26° and 37 ° +/- Endo H 14 82 4.3.3. Excess P 2m does not promote proper folding of the the bulk of the intracellular store of the D b protein While 293 cells show no shortage of P2m normally, the extra burden of large amounts of transfected D proteins could result in insufficient quantities of P2m to supply both the transfected and the endogenous MHC CI I molecules. A comparison of the levels of endogenous HLA (W6/32) expression in both 293 and the transfectants suggested that in almost every case, the presence of an H-2 transfectant reduced the level of endogenous MHC CI I expressed at the cell surface (Figure 9). One explanation for this observation is that levels of P2m in the transfectants were limiting the total number of MHC CI I molecules that could achieve a mature conformation and make it to the cell surface. By adding more P2m to these cells, it would be possible to promote more H chain-P2m interactions and thereby upregulate the surface expression ofD b. 293Db cells were infected with the P2m-Vaccinia construct. Infected and mock infected cells were FACS analyzed (table 4). While the level of P2m was increased when this method was used, the cell surface expression was unchanged. In figure 16 we demonstrate that infection with the human-p2m-Vaccinia construct b b markedly increases the amount of P2m precipitated with D proteins. 293D cells infected or mock infected with a human P2m vaccinia construct were precipitated with anti P2m antisera. The precipitated P2m from the infected cells co-immunoprecipitates many other proteins 83 characteristic of vaccinia infections. Figure 16. The effect of infection with a vaccinia virus human-P m construct on P2m expression 2 9 3 D b ce l ls were alternately infected (+) or m o c k infected (-) w i t h a hurnan-P 2m construct expressed i n vacc in ia virus (hu-P 2 m-Vac) . C e l l s were infected 18 hours pr ior to l abe l l ing w i t h P ro M i x (35S-Met /Cys) . The cel ls were pulsed for 30 minutes and chased for 120 minutes. P 2m w a s precipitated wi th rabbit anti human-P 2 m antisera (as listed i n table 2). M H C C I I and |3 2 m are denoted by the arrows. + 85 Table 4. A comparison of the cell surface expression of the H-2 D b protein at 26°C and 37°C with and without excess P m non-infected (AFU) infected (AFU) percentage 26°C 44.8 64.4 144 37°C 58.9 68.1 116 percentage 132 106 I+l+l+l+l+l+l+l+l+l+l+l+l+l+l+l+l+l+l+A £+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+^ ^ $$$$$$$$$$$$$$$$$$$$$$$$ ++++++x++++++^  The 293Db transfectant was alternately cultured at 26°C and 37°C. Additionally, cells were either mock infected or infected with the hu-P2m-Vaccinia as described in Materials and Methods. At 18 hours post infection, the cells were harvested and FACS analyzed as described in Materials and Methods. Levels of cell surface expression are given in arbitrary fluorescence units (AFU). Calculation of AFU involved taking the difference between mock and specifically labelled FITC labelled samples. Percentage of cell surface expression is determined by comparing changes in AFU along rows or down columns, e.g Uninfected 293Db transfectants at 37°C express 132% the level of Db expressed at 26°C at the cell surface. The result of adding excess human P2m in cells cultured at both the standard 37°C and 26°C is shown in table 4. Addition of P2m results in an increase in cell surface expression of 15%. When cells cultured at 26°C are exposed to excess P2m, the level of cell surface expression increases by 43%. Figure 17 shows the analyses of the Endo H resistance of these transfectants. While the percentage of cell surface expression of D b increases, the overall amount that is Endo H resistant remains much less than the Endo H sensitive population. 86 Figure 17. The effect of growth at 26°C and infection with a vaccinia virus human-Pm on escape of intracellular stores of H-2Db protein from the ER 293Db cells cultured at 2 6 ° C and 3 7 ° C were alternately infected or mock infected with a human-P 2 m construct expressed in vacc in ia virus (hu-P 2 m-Vac). Ce l l s were infected 18 hours pr ior to l abe l l i ng wi th P ro M i x (35S-Met /Cys) . T h e cel ls were pulsed for 30 minutes and chased for 120 minutes. M H C C I I H - 2 D b was precipitated wi th a Db specific antibody (as listed in table 1). P r io r to S D S - P A G E , precipitates and M W standards were alternately mock treated or treated wi th E n d o H for 24 hours. 293Db +/- p2m-Vac @ 26° and 37° C 26< 37< • • - • M + hu-f^m-Vac + + Endo H kD 43 29 87 4.3.5. y l F N has no effect in upregulating cell surface expression of D Previous studies had demonstrated that y-IFN had been used to upregulate cell surface expression of MHC CI I molecules in cell lines that normally had low levels. In these studies, the non-specific effects of y-IFN included upregulating many components involved in antigen presentation, with the obvious final effect of increasing cell surface expression of MHC CI I. This treatment was applied to 293 D b cells to determine if some other element required for the efficient transport of D b was regulated with the other antigen processing elements. Table 5 shows the results of this treatment. 293 cells tranfected with E3/19K (293.12 cells) and the 293 D b transfectant cell lines were pretreated with y-IFN. While the effect of y-IFN can be observed on MHC CI I expression in 293 cells, treatment with y-IFN has absolutely no effect in upregulating the cell surface expression of D b in the transfectants. In 293.12 cells, which normally exhibit reduced cell surface expression, y-IFN is able to upregulate cell surface expression to levels close to that observed in 293 cells exposed to y-IFN. 88 Table 5. The effect of y-IFN on M H C CI I cell surface expression Cell Line normal (AFU) y-IFN (AFU) percentage 293 12.8 20.4 160 293.12 3.91 19.5 498 293Db 6.43 5.89 91.6 Transfectants were alternately treated or mock treated with y-IFN. Cells were cultured in the presence of y-IFN for 24 hours. FACS analysis was performed to assess the effect of y-IFN on cell surface expression. Levels of cell surface expression were detected with W6/32 in 293 and 293.12 cells and 28.14.8s in 293Db cells and are given in arbitrary fluorescence units (AFU). Calculation of AFU involved taking the difference between mock and specifically labelled FITC labelled samples. Percentage of cell surface expression is determined by comparing changes in AFU along rows, e.g y-IFN treated 293Db transfectants express 91.6% the level of Db at the cell surface expressed in untreated cells. 89 4.4 Discussion Techniques documented in the literature were attempted to increase cell surface expression in the 293Db transfectant to levels closer to the overall level of expression ofD b. In figure 15 upregulation of cell surface expression of H-2 D b does not occur when cells are cultured at 26°C. Previously, it was thought that the regulation of cell surface expression of MHC CI I relied on factors such as rate of transport to the cell surface and turnover of surface molecules. The bulk flow theory predicts that proteins deposited in the ER will eventually pass through the ER, cis and trans Golgi and either out to the cell surface or into endosomes unless they are retained by an organelle specific retention or retrieval signal. The rate of transport of MHC CI I molecules was believed to be directly related to the ability of the MHC CI I molecule to achieve a stable and mature conformation and escape the ER. In many mutants, the factors that affect this are often the cofactors such as 3 2m (Daudi) and peptide (RMA-S, CMT-64). Peptide is controlled by TAP 1 and TAP 2. The experiments in this chapter suggest that in this case, some other factor is responsible. Results reported in the previous chapter demonstrated that D b is a slow binder to E3/19K. The length of time that this molecule spends in the same compartment as E3/19K enhances its ability to be bound by E3/19K. FACS experiments show that this protein is expressed at the cell surface in uninfected cells at a level similar to that of endogenous MHC CI I. These experiments also showed that when this (or other allelic proteins) are expressed at the cell surface, the level of the endogenous MHC CI I expression goes down correspondingly. The reduction in cell surface expression likely is related to a shortage of cofactors. Results reported in this chapter suggest that the addition of an excess of co-factors still has no effect on upregulating cell surface expression. 90 It is possible that there is feedback control of total MHC CI I molecules at the cell surface. When a certain number of MHC CI I molecules is expressed on the cell surface, a signal is generated that reduces the number to follow. Rather than each allelic protein being regulated differently and by separate mechanisms, each protein expressed may have a 'set' contribution or ratio of the final cell surface population. In 293Db, the level of the endogenous MHC CI I does not drop to zero; the block in the maturation and transport of one allelic protein (in this case the D b) is clearly not blocking the maturation and transport of the endogenous allelic proteins. It has been shown that the retention of E3/19K in the ER triggers a UPR (68). The level of activation of N F - K B has been shown to be dependant on the amount of E3/19K present in the ER. It is not the binding of E3/19K with MHC CI I that is the trigger of N F - K B ; it is its appearance in the ER. An excess of D b molecules in the ER may trigger the UPR. Others show that the release of Ca 2 + from the ER into the cytoplasm seems to trigger this response (68). Further studies in yeast suggest that activation of a membrane serine/threonine kinase triggers the alternate splicing of a transcription factor, HAC1, in the nucleus (145, 146). HAC1 has an affinity for the unfolded protein response element (UPRE) and upregulates expression of genes with this element. Some known products include heat shock proteins such as BiP. In a recent review (145) it is speculated that free BiP plays a role in the regulation of this process. Excess protein in the ER may mop up all free BiP in the ER. In yeast, free BiP may associate with a membrane protein called Irelp. Dissociation of BiP allows Irelp to dimerise and autophosphorylate, sending a signal through the UPR to upregulate transcription of more BiP. Presumably, the increased levels of Bip should establish a new equilibrium of unfolded proteins to unfolded protein binding capacity in the ER. In these studies, the investigators suggest that the UPR is instrumental in preventing apoptosis. 91 The regulation of the UPR may influence the turnover of molecules at the cell surface. The UPR upregulates expression of heat shock protein in response to proteins that remain in the ER. Cell surface regulation is probably influenced by a combination of concentration, regular turnover at a specific rate and a positive signal to either keep proteins at the cell surface or to tag them for turnover. 92 5. E3/19K binding to M H C CI I does not exclude association with calnexin 5.1 Introduction The observation thatE3/19K very rapidly bound and stabilised a mature conformation of MHC CI I proteins leads to the speculation that E3/19K behaves like a chaperone. Examination of E3/19K binding to MHC CI I in the ER involving the ER resident chaperones leads to the proposal that E3/19K binding in the ER will disrupt or involve the chaperones involved in the MHC CI I processing pathway. Chaperones are molecules that bind to immature peptide sequences. They bind to proteins before they fully mature and form either a transient or permanent association with them. Chaperones are usually associated with a protein that has not yet been transported to its final destination. The role of a chaperone protein is to bind to immature and misfolded proteins and to cooperatively act with them to allow them either to achieve their correct conformation (e.g. calnexin) or to shuttle for degradation (BiP). Many chaperones act with protein intermediates in the ER. Ribosomes transcribing nascent peptides are directed to the ER by way of the signal sequence;the peptide is directed into the lumen of the ER. The newly made protein is modified by many enzymes that act in concert to further process it. The signal sequence is removed by signal peptidase. Transport through the ER membrane may be enhanced by association with BiP, which acts as a ratchet pulling the peptide into the ER. Glycans are added and modified by a series of glycosylation enzymes. Disulphide bridges are established between cysteine residues; proline isomerisation is mediated by protein prolyl isomerase (PPI). Calnexin and calreticulin act in concert as a type of quality control mechanism; once proteins have achieved a correct conformation, they are released by calnexin and permitted to progress further through the ER-Golgi maturation pathway. More recently, other participants such as ERp57 have been shown to have a role in 93 the glycoprotein quality control mechanisms in the ER. BiP/Kar2p has been implicated in playing a regulatory role in the UPR of cells (147). These proteins are involved with most, if not all, glycoproteins that travel through the ER. In addition to the concerted action of these ER resident proteins, the MHC CI I molecules that are involved in antigen presentation are acted on by a further set of specialised chaperones. This list has grown lately as the interactions among all the participants are better defined. Some chaperones involved in the antigen presentation path include calnexin and calreticulin with specialized roles. Additionally, since MHC CI I molecules are composed of a trimolecular complex that includes a peptide fragment, the TAP proteins also have been demonstrated to act as chaperones. More recently, essential roles for tapasin (148) have been elucidated. 5.1.1. Maturation of a Cell Surface Membrane Protein A cell surface membrane protein comes in contact with this progression of chaperones, mentioned above, on its way to the cell surface. If it is properly folded, the interaction with the chaperone proteins will be temporary. If it is not properly folded, there exist many chaperones that will bind the improperly folded protein and feed it to the degradation machinery of the cell. Almost any path that a peptide will take through the cell will involve the participation of chaperones. It is in this manner that E3/19K also functions as a chaperone. E3/19K binds to immature peptides and seals their fate in a manner not unlike BiP. It stabilises a mature epitope and halts its progression through the cell. In the pulse chase studies with HLA proteins, the W6/32 antibody was used to follow the rate of appearance of mature MHC CI I molecules. Acquisition of the W6/32 conformation has been used as an indication of maturity (133, 149). In contrast to the Rabbit antiserum R426, W6/32 recognises only the MHC CI I molecules that have escaped the ER. While the epitope recognised by the W6/32 epitope is coincident with a mature epitope, it should be noted that the converse is not necessarily true; recognition by W6/32 does not guarantee that a MHC 94 CI I protein is mature. Calnexin is one effector of quality control in the ER, indiscriminately binding glycoproteins in the ER until they achieve a mature conformation and are fit to be transported to the cell surface. Calnexin binds MHC CI I molecules at some point in their maturation. 5.2. Rationale and Goals The goal of this study was to determine whether the association of E3/19K with MHC CI I complex in the ER would disrupt the association of any other chaperones in the ER. This set of experiments set out to determine if an E3/19K bound protein would still associate with calnexin. This data would give information on the physical association of participants in complexes. It would also shed light on the timing of different events occurring in protein maturation and quality control. 5.3. Results 5.3.1. E3/19K preserves W6/32 binding of M H C CI I HLA in the presence of tunicamycin Many studies have demonstrated that proper glycosylation of MHC CI I molecules is required for maturation of MHC CI I molecules (150). Growth in the presence of a glycosylation inhibitor such as tunicamycin results in increased levels of improperly folded proteins in the presence of BiP in the ER and can trigger the UPR. When assessed with W6/32 the mature epitope is lost when 293 cells are grown in the presence of tunicamycin (figure 18). Unglycosylated MHC CI I glycoprotein is unable to exit the ER. An unglycosylated MHC CI I molecule is likely unable to associate with calnexin (56, 151) and therefore can proceed no further down the maturation pathway. When the 293.12 cells are treated with tunicamycin the MHC CI I loses its carbohydrates but the W6/32 epitope is still detected and MHC CI I co-precipitates with E3/19K. This finding suggested that E3/19K can bind to MHC CI I in the 95 absence of carbohydrates and is able to force it into a mature conformation. In figure 18b the effect of a tunicamycin titration on 293 and 293.12 cells is examined. In this experiment, the non-specific epitope recognized by the antisera R426 (and the identical R425) can be precipitated at all concentrations of tunicamycin. A similar result is observed in 293.12 cells. As the tunicamycin concentration increases, the mobility of the MHC CI I increases coincident with the loss of carbohydrates. Figure 18a illustrates the results of the same experiment using W6/32. When the mature conformation specific monoclonal W6/32 is used, MHC CI I ceases to be precipitated from 293 cells at ~7]ig/ml Tunicamycin. In 293.12 cells, MHC CI I can be detected at all tunicamycin concentrations. E3/19K is also observed co-immunoprecipitaring at all points. This result demonstrates that E3/19K stabilises a mature (W6/32) epitope under adverse conditions. In this case, E3/19K is able to cause conformation specific mAb to bind even in the absence of carbohydrates. 96 Figure 18. The effect of increasing levels of tunicamycin on MHC CI I in 293 and 293.12 cells 293 and the E 3 / 1 9 K transfected 293.12 c e l l l ines were cul tured i n the presence o f increasing concentrations o f tun icamyc in . C e l l s were g r o w n i n med ia supplemented wi th 0, 3.5, 7, 14 and 21 Lig /ml (lanes 1-5) respectively for 18 hours pr ior to label l ing . B o t h labell ing and chase media were also supplemented wi th tunicamycin. Ce l l s were labelled for 30 minutes and chased for t w o hours. Immunoprec ip i ta t ion was per formed w i t h a) the conformation specif ic m o n o c l o n a l W 6 / 3 2 or b) non conformation-specif ic sera R 4 2 6 f o l l o w e d by S D S -P A G E . Tunicamycin titration 293 a) W6/32 43 kD — 1 2 mm 3 4 5 293.12 1 2 3 4 5 b) R426 43 kD 293 1 2 3 4 5 i 293.12 1 2 3 4 5 97 5.3.2. E3/19K rapidly stabilises a mature epitope The rapidity of the formation of a mature epitope was investigated in figure 19. This figure illustrates R426 and W6/32 precipitated products at various time points post chase. The R426 population appears immediately in both 293 and 293.12 cells. The increased mobility of E3/19K with time reflecting modification of sugars. In figure 19b an immunoprecipitation with W6/32 is shown. In 293 cells it is apparent that the rate of transport through the ER and coincident mature conformation is 2-4 hours. The first appearance of the mature W6/32 epitope starts at about one hour post chase. In the 293.12 cells a W6/32 epitope can be observed almost immediately, with some precipitate visible at 0 minutes post pulse. Once again, this gel confirms that both P2m and E3/19K are co-precipitated with MHC CI I molecules in 293.12 cells. This experiment demonstrates that a mature MHC CI I epitope is very rapidly stabilised in the presence of E3/19K, even though the MHC CI I molecule is not transported out of the ER. 98 Figure 19. Time course of maturation of M H C CI I in 293 and 293.12 cells 293 and 293.12 cells were grown to 95% confluence and labelled using 35S-Met/Cys Pro-Mix. Cells were pulsed for 30 minutes and chased for 0, 30, 60, 120 and 240 minutes (lanes 1-5) respectively. Lysates were precipitated with non conformation-specific sera R426 or the conformation specific monoclonal W6/32. Precipitates were separated by SDS-PAGE. MHC CI I (a), (32m (b) and E3/19K (c) are indicated. 99 Because E3/19K binds to MHC CI I so quickly and almost immediately induces it to form a mature conformation, it was suggested that it was acting in a chaperone like manner. It is interesting that E3/19K quickly stabilizes a mature epitope. The following experiments examined whether the association of the 'chaperone' E3/19K would affect the association of other more regular ER resident chaperones such as calnexin. Specifically, an experiment was performed to determine if E3/19K and calnexin binding to MHC CI I was exclusive. 5.3.3. E3/19K binding to M H C CI I does not exclude calnexin association An experiment that tested if E3/19K and calnexin could bind to MHC CI I at the same time is shown in figure 20. MHC CI I molecules were immunoprecipitated with either W6/32 or R426. This experiment was done as documented in Materials and Methods except the cells were lysed in the gentle CHAPs buffer. Lysis in NP40 typically abolished any association of calnexin with MHC CI I molecules. MHC CI I was precipitated, run on a gel and then transferred to a membrane. The membrane was probed with anti calnexin antisera as described in Materials and Methods. Figure 20 shows the association of calnexin with the MHC CI I-P2m complex. In 293 cells precipitated with W6/32, no calnexin can be found associated with this complex. When precipitated with R426, calnexin is found associated with the complex at all timepoints. In the 293.12 cells, the MHC CI I-P2m-E3/19K complex does associate with calnexin at all timepoints using both W6/32 and R426. 101 Figure 20. Calnexin association with MHC CI I precipitated from 293 and 293.12 cells 293 and 293.12 ce l l s were g r o w n to 9 5 % confluence and labe l led us ing P r o - M i x . C e l l s were pu lsed for 30 minutes and chased for 0, 60, 120 and 240 minutes (lanes 1-4 respect ively) . Lysates were precipitated w i t h either the conformat ion specific monoclonal W 6 / 3 2 or non conformation-specif ic sera R 4 2 6 . Precipitates were separated by S D S - P A G E . Ge ls were blotted onto Immobi lon P V D F membranes overnight ( O / N ) . Membranes were f ixed and western blotted wi th ant i -calnexin antisera (1:10,000) us ing the glass plate incubation descr ibed in materials and methods. Spec i f i c antisera was detected us ing Horse R a d i s h P e r o x i d a s e ( H R P O ) conjugated secondary an t ibody ( Jackson) and the E C L method (Amersham) . K o d a k X A R film was exposed for 1 to 15 minutes. 102 These data confirm that when W6/32 and R426 are used to immunoprecipitate MHC CI I molecules from 293 and 293.12 cells, different subsets of these molecules are precipitated corresponding to MHC CI I molecules of different maturity. The R426 precipitate is the total MHC CI I population including both immature and mature MHC CI I molecules; the W6/32 precipitate is the subset of those that are mature. Calnexin is not expected to be associated with the mature W6/32 epitope but will associate with the R426 epitope. That calnexin also associates with the MHC CI I-E3/19K complex is not completely unexpected as this complex is retained in the ER. The observation that E3/19K does not block this association is noteworthy. 5.3.4. M H C CI I precipitated with anti TAP antisera also associates with calnexin and E3/19K The role of the TAPs has been intimately linked to empty MHC CI I molecules. Empty MHC CI I molecules are retained in the ER until they receive the correct peptide, then they are able to escape the ER and 20-80 minutes later are expressed on the cell surface (152). The association of other ER resident proteins such as tapasin have also recently been observed (153). Precipitation with anti-transporter antisera should bring down those immature MHC CI I that are accepting peptides. Figure 21 illustrates 293 and 293.12 lysates precipitated with W6/32 mAb and antitransporter antisera. This experiment consisted of a pulse followed by several chase timepoints. The precipitates were separated on a gel and subsequently transferred to a membrane for western blotting analysis. This figure demonstrates once again that MHC CI I in 293.12 cells forms a complex with calnexin. When precipitated with anti MHC CI I, calnexin still associates with the E3/19K-MHC CI I-P2m complex. Moreover, this figure reveals that, when precipitated with 103 the antitransporter antisera, c a lnex in can s t i l l be found i n the western b lo t o f the same precipitate. Add i t iona l ly , E 3 / 1 9 K can be observed co-imrnunoprecipitating wi th this complex . Th i s E R resident complex consists o f M H C C I I, E 3 / 1 9 K , P 2 m , calnexin and possibly T A P 1 and T A P 2. Th is helps establish a clearer picture o f the b ind ing process i n the E R . It suggests that this is a fair ly large protein complex . 104 Figure 21. Calnexin association with W6/32 and Tap precipitates from 293 and 293.12 cells 293 and 293.12 cells were grown to confluence and labelled as detailed in materials and methods. This figure shows an immunoprecipitation by W6/32 and anti transporter antisera followed by western blotting with anti calnexin antisera. Cells were pulsed 30 minutes and chased for 0, 30, 75, 150 and 240 minutes (lanes 1-5 respectively), a) Lysates were precipitated with the conformation specific monoclonal W6/32 or the anti transporter antisera. Precipitates were separated by SDS-PAGE on the BioRad minigel system according to manufacturers instructions. Co-immunoprecipitated MHC CI I and E3/19K are denoted by arrows labelled i and ii respectively. b) Subsequently, membranes were probed with anti calnexin antisera using western blotting techniques detailed in Materials and Methods. Calnexin is denoted by the arrow labelled iii. Visualisation of the anti calnexin antisera utilised secondary goat anti rabbit antisera conjugated to HRPO (Jackson) and the Amersham ECL detection system. ECL treated membranes were exposed to Kodak XAR film for 1 to 15 minutes. 105 ^ r - CM CO lO n 1 O T -5.3.5. Peptide binding to M H C CI I does not exclude E3/19K binding The repertoire of peptides able to associate with specific MHC CI I proteins has been fairly well characterised. In addition to hundreds of peptides eluted from various MHC CI I proteins, a set of rules dictating the topography of peptides able to bind specific proteins has been formulated (154). One such peptide that binds the HLA-A2 protein is one derived from the HIV reverse transcriptase (RT) protein. The sequence of the peptide is ILKEPVHGV. While it had been demonstrated that such peptides could be eluted from MHC CI I and then identified in mass spectroscopy experiments, it had not been shown that these peptides were bound by the peptide binding groove. In an attempt to demonstrate binding of the peptide to the MHC CI I, the observation that E3/19K co-precipitated with this complex was made. The HIV RT peptide ILKEVFVG was identified as a peptide that fit the motif of an HLA A2 binding peptide. This peptide was N-terminally biotinylated to generate a marker for binding to MHC CI I. N-terminally biotinylated peptides were incubated with MHC CI I and ultimately would be used to precipitate these complexes. The experimental procedure has been described in Materials and Methods. Figure 22 demonstrates that biotinylated peptide was able to precipitate a MHC CI I enriched lysate. It was also noted that less MHC CI I was observed when less biotinylated peptide was added. In a competition with cold peptide at lOx excess, binding to the biotinylated peptide was abrogated. Finally E3/19K could still binds peptide bound MHC CI I molecules. This demonstrated that the association of E3/19K with MHC CI I molecules did not block the peptide binding groove. In figure 22 we show an increase in the amount of acid eluted MHC CI I pulled down with biotinylated peptide corresponding to peptide concentration. A 10 fold increase in biotinylated peptide yielded an increase in the amount precipitated. In a competition with 10 fold excess cold peptide, no MHC CI I was pulled down, demonstrating the specificity of the peptide for HLA-A2. Finally cold peptide alone did not precipitate MHC CI I. 108 This experiment was performed using lysates from 293.12 cells. In 293.12 cells, the interference of E3/19K with the association of the biotinylated peptide was investigated. If E3/19K was binding to the MHC CI I molecule through the peptide binding groove, then the binding of the biotinylated peptide should be competitively inhibited. Alternately, E3/19K binding to MHC CI I molecules might alter the geometry of the peptide binding groove rendering it unable to hold the biotinylated peptide. This did not occur. When the 293.12 lysates were used, the biotinylated peptide continued to precipitate MHC CI I molecules. In addition, the precipitated MHC CI I molecules co-precipitated with E3/19K. This shows that peptide binding is not excluded by E3/19K association. It may suggest that MHC CI I molecules that have associated with E3/19K may be empty facilitating peptide binding in lysate, but this remains to be conclusively shown 109 Figure 22. Precipitation of M H C CI I in enriched suspension using a biotinylated peptide 293.12 cells were grown to 95% confluence and labelled and labelled using 35S-Met/Cys Pro-Mix (Amersham). Cells were pulsed for 30 minutes and chased for two hours. Chase media was alternately supplemented with biotinylated peptide at 5 uM and 50 nM (lanes 1 and 2 respectively), a mixture of peptide (5 uM) and biotinylated peptide (50nM) (lane 3), peptide (5 uM) (lane 4), biotin (25 uM) (lane 5) or mock treated (lane 6). At the end of the chase period, cells were washed and lysed in the presence of the crosslinking reagent DSP (200 ug/ml). Lysates were precipitated with non conformation-specific anti MHC CI I antisera R426. Precipitates were subjected to a glycine acid solution to remove Prot-A-Sepharose. The supernatant was divided into two aliquots; a 450 ul aliquot and a 50 ul aliquot. The 450 ul aliquot was precipitated with streptavidin-agarose (Pierce) (lanes 1-6 on the left). The 50 ul aliquot was precipitated with R426 a second time (lanes 1-6 on the right). Precipitates were separated on SDS-PAGE. 110 Ill 5.4. Discussion The results in this chapter demonstrated two significant points. The first was that E3/19K in 293 cells rapidly stabilised a mature MHC CI I epitope. The implications of this are discussed below. Second, the complex involving E3/19K also involves several other ER resident proteins including calnexin and at least one and possibly both TAP proteins. When testing to determine if E3/19K could bind in the absence of carbohydrates, it was contrary to expectations to observe that in a tunicamycin treatment of 293 and 293.12 cells, the mature MHC CI I epitope was lost in 293 cells but was retained in 293.12 cells. Tunicamycin blocks the transfer of the dolichol associated carbohydrate to the asparagine residue of the MHC CI I molecule. In most cases, the lack of this carbohydrate on MHC CI I molecules prevents any further maturation through the ER and they remain intracellular (124). Some studies have even shown that tunicamycin triggers the onset of apoptosis (155). 293.12 cells stably transfected with the E3/19K gene exhibit some of the characteristics reported in Ad2 infections including reduced cell surface expression of MHC CI I molecules (156). In 293.12 cells grown in the presence of tunicamycin, E3/19K was able to bind MHC CI I even when it was lacking a carbohydrate. It was also determined that the mature conformation detected by W6/32 was stabilised in this case. This observation suggested that the binding of E3/19K to the MHC CI I molecule stabilised a mature conformation. This study makes the assumption that the W6/32 epitope indicates a mature conformation of MHC CI I. Previous studies have used a W6/32 conformation as an indication of a mature conformation. However, the fact that this epitope can be precipitated from 293.12 cells which have been incubated in the presence of tunicamycin suggests that there may be many similarities between the structure stabilised by E3/19K and a mature MHC CI I molecule, but that the two conformations are not identical. It should be noted that the mature conformation referred to in this chapter is likely a stable conformation which resembles that of a mature MHC CI I protein and as such may differ from the surface bound protein in several 112 key features. When using the conformation specific monoclonal W6/32, it was observed that in 293 cells a mature conformation could be detected about 30-60 minutes post chase. This period of time required for maturation includes the participation of many other ER resident proteins including PDI (157, 158), rotamase, calnexin, calreticulin (159, 160), BiP (161-164) and many others. In the presence of E3/19K, MHC CI I molecules were able to achieve a mature (or W6/32 reactive) conformation almost immediately in 293.12 cells. This suggested that E3/19K binds to nascent MHC CI I molecules and quickly stabilize them. MHC CI I molecules could mature faster than in the absence of E3/19K, possibly bypassing the action of the concert of ER chaperones mentioned above. When precipitated with W6/32 in 293 cells, the mature MHC CI I epitope is not associated with calnexin molecules. This result was expected as the W6/32 recognizes a mature MHC CI I molecule which should no longer be accessible to calnexin. Calnexin binds to proteins until they achieve a mature conformation; in the case of MHC CI I molecules, calnexin release seems to coincide with the addition of peptide and the achievement of the correct conformation for release from the ER. When MHC CI I molecules are precipitated with R426 from the 293 cell line, MHC CI I associates with calnexin at all time points. Once again, this can be explained by the fact that R426 recognizes a wide subset of MHC CI I molecules including those that are immature. Immature MHC CI I molecules should still be associated with calnexin until they have achieved the correct conformation to leave the ER. As in 293 cells, the R426 precipitate of 293.12 cells associates with calnexin. This is due to R426 binding both mature and immature MHC CI I molecules. E3/19K co-immunoprecipitates with this complex demonstrating that E3/19K binding to MHC CI I molecules does not sterically or in any other way inhibit the association with calnexin. The MHC CI I protein with the W6/32 epitope is still able to bind calnexin in the presence of E3/19K in 293.12 cells. This is not unexpected as MHC CI I should be resident in the ER. 113 Although the initial association of calnexin with glycoproteins is mediated through the carbohydrates (56), the areas that appear to be crucial for continued association are found within the transmembrane region (165) reducing the likelihood of E3/19K sterically interfering with binding. Calnexin is readily dissociated from the MHC CI I complex in detergent. Further studies probe the associations between all of the participants of this complex. This complex appears to be loosely associated and fairly large with several participants. Having established this, it was investigated whether other ER resident proteins would also be found in this complex. Earlier it had been surmised that the binding of E3/19K might involve enveloping the alpha 3 region of the MHC CI I molecule. Key association points would be the alpha 1 and alpha 2 hypervariable regions. However, it is difficult to understand how calnexin fits into this picture. Calnexin may protrude further into the lumen of the ER than E3/19K and thereby associate with the N terminal portion of the MHC CI I molecule. One paper suggests that the areas involved in MHC CI I binding are in the transmembrane region (166). The above scenario raises the question whether calnexin associates with E3/19K as well. To further elucidate the nature of this calnexin-MHC CI I-P2m-E3/19K complex resident in the ER, it was examined whether calnexin binding was preceded by the addition of peptide. Previously it had been shown that MHC CI I molecules formed ternary complexes with TAP and calnexin and dissociation coincided with the acquisition of a peptide (28). The immunoprecipitation of 293 and 293.12 cells with both an anti transporter antisera and W6/32 addressed this. In the 293.12 cells precipitated with W6/32 the E3/19K co-immunoprecipitates with MHC CI I molecules. Surprisingly, precipitation of 293.12 lysates with the anti-transporter antisera precipitates a large complex including E3/19K. This suggests a large complex 114 involving the peptide transporter, MHC CI I, P 2m and E3/19K. The association with the peptide transporter is significant because it had been proposed that the nature of E3/19K binding to the MHC CI I molecule is through the peptide binding groove. That E3/19K and peptide transporter binding to MHC CI I is not exclusive suggest that E3/19K is may not be binding in the peptide binding groove. The experiment that shows that biotinylated peptides were able to bind to and cause the immunoprecipitation of MHC CI I molecules in a lysate demonstrates that when immunoprecipitated with the biotinylated peptide, the E3/19K proteins still associates with MHC CI I. This is significant in that it demonstrates that E3/19K binding to MHC CI I does not occur through the peptide binding groove. In addition, the association between E3/19K and MHC CI I does not sterically interfere with peptide access to the binding groove. In an earlier chapter, the nature of E3/19K binding to MHC CI I molecules was examined. One of the proposals considered was the possibility of E3/19K binding to the MHC CI I molecule through the peptide binding cleft. Binding to the cleft could interfere with the transfer of peptides from TAP to MHC CI I, possibly preventing release from TAP. With the results of this set of experiments this possibility seems unlikely. These findings also elucidate the order of the sequence of events in the ER. It appears that calnexin binds the MHC CI I glycoprotein first. Subsequent association with the peptide transporter and loading of the peptide triggers dissociation of both calnexin and the transporter. In cells not infected with Ad , the MHC CI I molecule is released for further travel through to the cell surface. In Ad infected cells, E3/19K may interfere with the transfer of peptide to the MHC CI I molecule preventing its subsequent release. In a previous chapter it is suggested that E3/19K is only able to bind those allelic proteins that remain in the ER to be loaded with peptide. Those that travel empty to the cell surface are not bound as effectively by E3/19K (i.e. K k). In the case of a slow binder, like Db, 115 the length of time that this protein stays associated with the peptide transporter and calnexin in the ER enhances the effectiveness of E3/19K binding. E3/19K binding to D b seems to take longer than for other proteins. Binding does not occur co-translationally. This means there is a large population of empty MHC CI I proteins that need to be loaded with peptide present in the ER. If E3/19K intercepts the MHC CI I molecules before they are loaded with peptide, does this excess of empty MHC CI I molecules monopolise all available peptide transporters? Previous results suggest this not to be the case. In the case of D , it seems that MHC CI I molecules which remain in the ER for a long time do not influence the maturation of other proteins. Recent studies implicate a role for tapasin in MHC CI I loading, stating that one TAP complex can service up to four MHC CI I molecules (167). In the last chapter it was noted that E3/19K proteins with the ER retrieval signal are returned to the ER (168). In the case of E3/19K mutants that were missing the ER retention signal, the passage of MHC CI I through the ER was slowed (114)! The ability of E3/19K to slow the passage through the ER, along with the inherent ability to stabilize mature epitopes suggests it shares some of the functions normally attributed to chaperones. Once the ER retention signal is removed, the E3/19K protein still delays the passage of MHC CI I through the ER (169). In addition, Gabathuler et al. (169) are unable to show mutant E3/19K and MHC CI I associating outside of the ER. In the absence of its ER retention signal, E3/19K association with MHC CI I molecules may be temporary. Schekman and his colleagues (83, 170-172) have speculated the existence of a positive signal required for macromolecules to be transported from the ER . The fact that the E3/19K molecule is not readily transported out in the absence of its ER retention signal may be an example of an ER protein that needs a positive signal to leave the ER. Gabathuler et al. (114) show that removal of the ER retention signal does slow the 116 passage of MHC CI I molecules through the ER. However, they do not clearly show whether or not the E3/19K protein without its ER retention signal (621 protein) makes its way out of the ER. It is likely that MHC CI I is not continually bound by 621 since the level of cell surface MHC CI I is not reduced in FACS experiments (173). There is no evidence to show association in any compartment other than the ER. It is conceivable that when the 621-MHC CI I complex reaches a subcompartment with physical characteristics different from the ER, dissociation occurs. E3/19K has an affinity for only those MHC CI I molecules to which it is exposed in the same compartment. This affinity is likely not high and if the MHC CI I -E3/19K complex moves to a compartment such as the cis Golgi, the complex spontaneously dissociates. Finally, it is the ER retention/retrieval signal KKMP that plays a role in retrieving the E3/19K and possibly the entire complex to the ER (168, 174), keeping them in an environment favouring association of E3/19K and MHC CI I. There is no evidence that bound MHC CI I are retrieved from a distal compartment, possibly because they will not progress past being TAP associated without a peptide. If the association of MHC CI I and E3/19K is largely dependant on the conditions found within the ER and is unable to bind outside this compartment, then the relationship between these two proteins is similar to that of a chaperone. In the last chapter it was suggested that at the cell surface there may be a type of regulation of MHC CI I molecules. A positive signal generated at the cell surface related to the concentration of MHC CI I molecules present, may feed back to the ER and limit the maturation of more MHC CI I molecules. The speedy appearance of a 'mature epitope' within the cell may have further implications. The overloading of molecules ready to leave the ER may send a signal to the cell surface that upregulates cell surface turnover. Alternately, the mature epitope of the MHC CI I molecules within the ER may help avoid the inevitable UPR. It has been speculated that the UPR is stimulated by the levels of free BiP in the ER (175) and vice versa (176). Mature conformations of retained proteins may be less likely to be bound by BiP and 117 less likely to trigger the UPR. Previously it has been proposed that the UPR helps prevent apoptosis of affected cells (176, 177), but some of the signalling triggers leukotrienes that stimulate inflammation. Infiltration of an afflicted area by immune response cells may allow a body to subsequently mount a more effective immune response to an affliction. Adenovirus infected cells may be less susceptible to such a response not only because of their characteristic ability to evade the immune system but also because of their weak stimulation of the UPR (68). The complex involving E3/19K, MHC CI I and transporter does exist. Additionally, it is observed that TAP and calnexin are also associated. The involvement of all of these components yields a very large complex. 118 6. Discussion The effect of the adenovirus E3/19K protein on both the cellular machinery involved in antigen presentation and the regular maturation of MHC CI I molecules has been examined with investigations on the role of the interactions of the many accessory proteins found in the ER, Golgi and other compartments of the cell. This study outlined the participants in the processing of cell surface proteins and helped elucidate the sequence and relative order of many of the steps required for effective antigen presentation. This study also investigated the role of E3/19K in the normal infection process of adenovirus and possibly other related viruses (adeno associated viruses). The first chapter (three) examined the ability of E3/19K to bind specific protein products of different H-2 proteins in mice. This was the first time that the binding ability of E3/19K has been assessed against a common background (human 293 cells). Previously, pairs of proteins in different cell lines were exposed to E3/19K and the retention compared. The K d and K b allelic proteins bind to E3/19K very effectively while the K k and D k proteins do not. These proteins were classified into binding and non-binding groups. The D b and L d proteins were also found to bind, albeit slowly. Previously they had been shown not to bind to E3/19K. The binding of these proteins had a half life of 3-12 hours, that contrasted greatly with the t 1 / 2 for the strong binders K d and Kb. This new class of binders was designated slow-binders. E3/19K binding to MHC CI I is based partly on an innate affinity of specific allelic proteins to E3/19K. More importantly, binding requires that E3/19K and the target protein occupy the same intracellular compartment for an extended period of time. Allelic proteins like K d bind quickly to E3/19K whereas proteins like D b bind less quickly, but equally as 119 effectively, to E3/19K. That E3/19K does not bind K k and D k likely reflects the amount of exposure that they have to E3/19K rather than a lack of affinity. As these molecules are quickly transported to the cell surface, they are not in the same compartment with E3/19K long enough for an association to occur. Unlike in the mouse, when surveying E3/19K binding to human HLA proteins, it is found that all tested allelic proteins bind. The K k and D k allelic proteins are the only two shown to date not to bind to E3/19K. It is noteworthy that the tm for K k to travel through the ER and become Endo H resistant is the shortest of all allelic proteins tested. The observation that the K k protein does not bind may be due to the fact that the allelic protein is transported out of the ER at a very rapid rate and therefore is not physically available to be bound by E3/19K. It follows that if the rate of transport out of the ER for the K k protein could be retarded, E3/19K binding may be observed. The observation that the D protein binds E3/19K strongly but with a long tm suggests that the factors that influence E3/19K binding are closely related to factors that influence transport from the ER and to the cell surface. Normally, MHC CI I proteins are translated into the ER where they immediately achieve a correct conformation through the interaction of many chaperones and ER resident proteins. Proteins that achieve a correct conformation are quickly shuttled along the ER-Golgi pathway where they are acted upon by other chaperones. Those that do not achieve a correct conformation are retained in the ER. Previous studies have shown that different allelic proteins are transported with different rates through the ER. One of the factors that influences the rate of transport out of the ER is the availability of peptides that can be bound by the MHC CI I molecule. The affinity of specific MHC CI I proteins for existing peptides and the competition for these peptides may greatly influence the rate of transport of MHC CI I molecules to the cell surface. In the case of 120 D b the inability to bind peptides that confer a stable conformation may be the reason for slow transport. For Kk, which is rapidly transported from the ER, the protein may have ready access to peptides that allow the correct mature conformation and the resulting transport to the cell surface. Another scenario is that K k may bypass the quality control mechanism of the ER and is transported empty. Several studies have demonstrated that empty MHC CI I molecules can be found on the cell surface in low numbers. These molecules are unstable and are rapidly turned over at the cell surface. Factors that influence their stability include the allelic protein which is being expressed and the cell line (and therefore cellular machinery) in which it is expressed. In the experiment with transfected 293 cells, the K k protein may bypass the chaperones responsible for quality control and proceed to the cell surface without peptides. This would explain the quick rate of passage through the ER to the cell surface and could also account for the inability of E3/19K to form any association with this protein. Unlike the K k protein, D b proceeds through the ER with a much longer half life. Comparison with K d shows that D b binds E3/19K much more slowly. This may be due to its slow rate of maturation through the ER. It could also reflect a lower affinity of D b for E3/19K. A molecule with lower affinity for E3/19K may still be able to bind based on the length of time in the same compartment. The greater exposure of D b to E3/19K offsets its lower affinity. Conversely, the fact that K d is rapidly transported through the ER is offset by its higher affinity for E3/19K. The findings of earlier researchers should be reconsidered in light of these observations. Previously, groups determining the affinity of E3/19K for specific allelic proteins may actually have been determining the amount of exposure these allelic proteins had 121 to E3/19K. When D b was identified as a non binder, its rate of transport out of the ER limited the binding to E3/19K. The inability to form a complex with E3/19K was dependant on cellular factors that influence the rate of transport through the ER rather than an overall affinity between the two proteins. The ability of D b to bind has been overlooked by other groups partly because of the longer time required for this association. Perhaps all allelic proteins are either fast binders or slow-binders. Those previously shown not to bind may not have been exposed to the conditions that would allow them to share the compartment with E3/19K for a period long enough to promote binding. In the following chapter, factors that influence the normal maturation of MHC CI I molecules were discussed. The 293D transfectant expressed large amounts of D but processed relatively small amounts to the cell surface. Much of the translated product was found within the ER where it remains for an extended period of time. The 293Db cell line was subjected to conditions that have been demonstrated to restore defective cell surface expression. TAP 2 deficient RMA-S cells grown at 26°C instead of the regular 37°C show increased levels of MHC CI I molecules expressed at the cell surface . These MHC CI I molecules were devoid of peptide in the binding cleft. The reduced temperature was believed to lower the kinetic energy of the system and reduce the rate of dissociation of these unstable complexes at the cell surface. The experiment consisted of growing 293DD at 26°C and comparing the cell surface expression to 293D grown at 37°C. The results showed no increase in cell surface expression in FACS analysis. To complement this experiment, level of P2m available to the MHC CI I molecules in the 293Db cell line was increased. The rationale for this experiment was that if [3 m was in limiting supply, then increasing this supply would result in greater numbers of D (and 122 presumably all other) MHC CI I proteins at the cell surface. Increased P 2m caused by the infection of 293D with a vaccinia construct did not increase the cell surface expression. Finally, studies with the CMT-64 cell line showed that high levels of cell surface expression could be achieved by exposing these cells to y-IFN. y l F N acted as a general upregulator of many cellular processes including the expression of TAPs, P2m and MHC CI I. When the 293Db was exposed to ylFN, it was found that the levels of D b at the cell surface did not increase. Failed attempts to upregulate the levels of D b at the cell surface demonstrated an intriguing property of MHC CI I expression. In a cell with normal levels of MHC CI I expression (endogenous HLA), levels of foreign MHC CI I molecule could not be upregulated. Attempts to provide missing cofactors or to reduce the stringency of the quality control (by reducing the temperature) were unsuccessful in increasing the level of expression. This suggests that the level of the D b protein on the surface of 293 cells is actively regulated. It is assumed that some other accessory protein may have a direct role in controlling the level of surface expression of MHC CI I. The bulk flow theory has been the prevailing theory regarding the transfer of membrane and secretory proteins through the ER. More recently, other groups including one led by Schekman have suggested the existence of positive signals that move cargo from the ER to the Golgi (83). Investigation of the regulation of cell surface molecules may be influenced by the turnover of MHC CI I molecules at the cell surface. Some investigators (114) have suggested that the number of cell surface molecules is based entirely on the rate of transport to the surface. They suggest that turnover at the cell surface occurs at a set rate; reduced populations 123 result when any particular member of the population travels to the cell surface at a rate lower than that of the turnover. The results suggest that the level of expression of any particular species at the cell surface may be concentration dependant. Once the concentration of a particular population of MHC CI I molecules reaches its saturation point, any increase in expression will not result in increased representation at the cell surface. In the experiments, any excess D b is retained in the ER. In the third section, it was observed that association of human HLA proteins with E3/19K caused rapid maturation as assessed by a conformation specific monoclonal antibody. HLA - A, B and C were very quickly recognised by W6/32 in the presence of E3/19K. Due to E3/19K retention in the ER, these molecules never achieved Endo H resistance. In short, E3/19K was binding to a misfolded and unprocessed immature MHC CI I molecule and forcing it into a conformation that resembled the mature conformation, yet preventing escape from the ER. Whether this interaction sterically interfered with the ability of other well characterised ER resident chaperones to associate with MHC CI I molecules was studied. Of particular interest was whether of calnexin and the TAP proteins could still associate with E3/19K bound MHC CI I. The normal sequence of events for MHC CI I association with ER resident chaperones was probed with E3/19K. With experiments described in that chapter, the sequence of events for the E3/19K binding to MHC CI I molecules can be elucidated. In most cases, MHC CI I is cotranslationally translocated into the lumen of the ER where it associates with P2m. Achievement of a mature conformation permits release by calnexin and egress from the ER is usually coincident with the loading of the correct peptide. With some allelic proteins, release from the ER may occur without peptide loading resulting in rapid transport to the cell surface. E3/19K associates with MHC CI I molecules in the ER before transport out of the ER. The 124 "slow-binders" suggest that this association is not necessarily co-translational for all allelic proteins as has been previously reported. E3/19K only will associate with those proteins which stay in the same compartment long enough to permit binding and E3/19K has different affinities for different allelic proteins. The exact nature of E3/19K association with MHC CI I molecules has yet to be determined. Differences in the binding affinity of different proteins have led some studies to map those residues in the MHC CI I molecule that may influence the binding to a greater degree than others. Site directed mutagenesis of E3/19K has shown that mutations in almost any region can abrogate binding. Other studies have mapped the key regions in the MHC CI I molecule to those hypervariable regions in the alpha 1 and alpha 2 regions. This led researchers in the past to suggest that E3/19K was binding in the peptide binding groove, possibly displacing peptides. Precipitation of E3/19K in a complex involving peptide transporters suggested that these two proteins were associating with MHC CI I non-exclusively. It is possible that E3/19K blocked the transfer of peptides to MHC CI I and their release from the TAP proteins. The experiments that precipitated MHC CI I with biotinylated peptide further demonstrated that MHC CI I and E3/19K did not associate with one another via the peptide binding groove. The role of E3/19K as a protein that was associating with MHC CI I as a member of a large group of proteins was investigated. The set of experiments demonstrating the association of several ER proteins in a complex with E3/19K indicated that this complex may be very large. Velocity sedimentation gradients were performed to assess the size of the complex. The sucrose gradients identified several subpopulations of the E3/19K protein with different densities. These subsets suggest that in addition to binding and retaining MHC CI I in the ER, E3/19K forms complexes with other proteins as well. E3/19K is in excess in studies that assess co-precipitation by alternately precipitating with anti-MHC CI I antibodies or anti E3/19K antisera, which suggests that much of it never associates with MHC CI I. 125 The nature of binding of different MHC CI I allelic proteins by E3/19K may reflect both the rate of transport and the path of the MHC CI I protein being bound. Proteins such as K k are not bound; these may be rapidly transported by bypassing the regular maturation process involving peptide loading. MHC CI I proteins present in the ER for long periods of time are bound effectively, for example Db. Like Kk, D b may not be maturing properly and is possibly susceptible to the quality control mechanisms present in the ER and therefore unable to exit the ER. Both proximity and time promote association with E3/19K. Human MHC CI I proteins are quickly bound by E3/19K and a mature conformation quickly stabilises. This mature conformation may result in rescue from the default degradation pathway that degrades improperly formed proteins in the ER. Additionally, this build up of mature MHC CI I molecules in the ER may trigger a UPR response. Finally, the experiments with ylFN reveal that the effect of E3/19K in the ER can be abrogated. This is significant because it makes the physiological role of E3/19K uncertain. Previously it was believed that as a result of its retention of MHC CI I in the ER, E3/19K reduced CTL killing of infected cells. However, the results of this research suggest that MHC CI I will not be reduced in an antiviral response. The nature of the restoration of MHC CI I cell surface expression is also interesting. Many groups have shown that E3/19K is normally found in excess in the ER. The response to y-IFN is unlikely to be a simple upregulation of MHC CI I levels. 126 7. Conclusion In summary, E3/19K served as a good tool to examine the nature of MHC CI I allele specificity in viral infection and the subversion of antigen presentation by viral machinery. Secondly, E3/19K proved to be invaluable in probing the cellular machinery involved in the manufacture of proteins that are processed through the ER. This includes all cell surface proteins as well as those that remain associated with membranes inside the cell. It has been demonstrated that E3/19K binds to different MHC CI I allelic proteins with different affinities. It was also demonstrated that different proteins behave differently based on the characteristics of the cell in which they are expressed. Factors such as the inherent ability of an allelic protein to be transported to the cell surface or detained within the cell proved to be as important as the affinity of specific proteins for E3/19K. Finally, examination and postulation of the effects of E3/19K on viral evasion of the immune system have led to a deeper understanding of the many mechanisms employed to this end. 127 Jt 1. McFadden , G . and K . Kane, 1994. How D N A viruses perturb functional M H C expression to alter immune recognition. Adv Cancer Res, 63: 117-209. 2 . Langman, R., The Immune System. 1989, San Diego: Academic Press, Inc. 3 . K le in , J . , Immunology: The Science of Self-Nonself Discrimination. 1982, Toronto: John Wiley & Sons. 4 . Zinkernagel , R . M . and P . C . Doherty, 1974. Immunological surveillance against altered self components by sensitised T lymphocytes in lymphocytic choriomeningitis. Nature, 251: 547-8. 5 . Haskins, K . , J . Kappler and P. Marrack , 1984. The Major His tocompat ib i l i ty Compex-Restricted Antigen Receptor on T Cells. Annu. Rev. Immunol., 2: 51-66. 6. Schwartz, R. , 1985. Associations in T-cell activation. Nature, 317: 284-285. 7 . W i l l i a m s , A . F . and A . N . Barclay, 1988. The Immunoglobul in Superfamily: Domains For Cel l Surface Recognition. Annu. Rev. Immunol, 6: 381-405. 8 . Schatz, D . , M . Oettinger and D . Balt imore, 1989. The V ( D ) J Recombination Activating Gene, R A G - 1 . Cell, 59: 1035-1048. 9 . Benacerraf, B . and H . O . McDev i t t , 1972. His tocompat ib i l i ty - l inked immune response genes. Science, 175: 273-9. 10. Wiertz, E . J . , S. Mukherjee and H . L . Ploegh, 1997. Viruses use stealth technology to escape from the host immune system. [Review] [52 refs]. Molecular Medicine Today, 3: 116-23. 11 . Steinmetz, M . and L . Hood, 1983. Genes of the major histocompatibility complex in mouse and man. Science, 222: 727-733. 12. Kle in , J . , Natural history of the major histocompatibility complex. 1986, New Y o r k : Wiley -Interscience Publication. 1-775. 13 . Bjorkman, P . J . , M . A . Saper, B . Samraoui , W . S . Bennett, J . L . Strominger and D . C . Wiley, 1987. Structure of the human class I histocompatibility antigen, H L A - A 2 . 128 Nature, 329: 506-12. 14. Townsend, A . , C . Ohlen, J . Bastin, H . Ljunggren, L . Foster and K . Karre , 1991. Association of class I major Histocompatibility heavy and light chains induced by viral peptides. Nature, 340: 443-448. 15. Garc ia , K . C . , M . Degano, L . R . Pease, M . Huang, P . A . Peterson, L . Teyton and I.A. Wilson, 1998. Structural basis of plasticity in T cell receptor recognition of a self pept ide -MHC antigen. Science, 279: 1166-72. 16. Hunt , D . F . , et al., 1992. Characterization of peptides bound to the class I M H C molecule H L A - A 2 . 1 by mass spectrometry [see comments]. Science, 255: 1261-3. 17. Deres, K . , T . N . Schumacher, K . H . Wiesm:uller, S. Stevanovhc, G . Greiner, G . Jung and H . L . Ploegh, 1992. Preferred size of peptides that bind to H-2 K b is sequence dependent. Eur J Immunol, 22: 1603-8. 18. Fa lk , K . , O . Rotzschke, K . Deres, J . Metzger, G . Jung and H . G . Rammensee, 1991. Identification of naturally processed viral nonapeptides allows their quantification in infected cells and suggests an allele-specific T cell epitope forecast. J Exp Med, 174: 425-34. 19. Rammensee, H . G . , K . Falk and O . Rotzschke, 1993. M H C molecules as peptide receptors. [Review] [111 refs]. Current Opinion in Immunology, 5: 35-44. 20 . Rammensee, H . G . , K . Falk and O . Rotzschke, 1993. Peptides naturally presented by M H C class I molecules. [Review] [136 refs]. Annual Review of Immunology, 11: 213-44. 2 1 . L u p a s , A . , P. Z w i c k l and W . Baumeister, 1994. Proteasome sequences in eubacteria. Trends in Biochemical Sciences, 19: 533-4. 22 . Hershko, A . and A . Ciechanover, 1992. The Ubiqui t in System for Protein Degradation. Annu Rev Biochem, . 2 3 . Ortiz-Navarrete , V . , A . Seelig, M . Gernold , S. Frentzel , P. Kloetzel and G . Hammerling, 1991. Subunit of the 20S proteasome (multicatalytic proteinase) encoded by the major histocompatibility complex. Nature, 353: 662-664. 129 24. Goldberg , A . L . and K . L . Rock, 1992. Proteolysis, proteasomes and antigen presentation. [Review] [92 refs]. Nature, 357: 375-9. 2 5 . Peters, J . M . , 1994. Proteasomes: protein degradation machines of the cell. Trends Biochem Sci, 19: 377-82. 26 . Monaco , J . J . and H . O . McDevi t t , 1986. The L M P antigens: a stable M H C -controlled multisubunit protein complex. Hum Immunol, 15: 416-26. 2 7 . Spies, T . , V . Cerundolo, M . Colonna, P. Cresswell, A . Townsend and R. DeMars, 1992. Presentation of viral antigen by M H C class I molecules is dependent on a putative peptide transporter heterodimer. Nature, 355: 644-6. 28 . Suh, W . K . , E . K . Mitche l l , Y . Y a n g , P . A . Peterson, G . L . Waneck and D . B . Will iams, 1996. M H C class I molecules form ternary complexes with calnexin and T A P and undergo peptide-regulated interaction with T A P via their extracellular domains. J Exp Med, 184: 337-48. 29 . Suh, W . K . , M . F . Cohen-Doyle , K . F r u h , K . Wang, P . A . Peterson and D . B . Wil l iams, 1994. Interaction of M H C class I molecules with the transporter associated with antigen processing. Science, 264: 1322-6. 30 . Ortmann, B . , M . J . Androlewicz and P. Cresswell , 1994. M H C class I/beta 2-microglobulin complexes associate with T A P transporters before peptide binding. Nature, 368: 864-7. 3 1 . El l io t t , T . , V . Cerundolo and A . Townsend, 1992. Short peptides assist the folding of free class I heavy chains in solution. European Journal of Immunology, 22: 3121-5. 32 . Cromme, F . V . , et al., 1994. Loss of transporter protein, encoded by the T A P - 1 gene, is highly correlated with loss of H L A expression in cervical carcinomas. Journal of Experimental Medicine, 179: 335-40. 33 . Ossevoort, M . A . , et al., 1993. Differential effect of transporter Tap 2 gene introduction into R M A - S cells on viral antigen processing. Eur J Immunol, 23: 3082-8. 34 . Gaskins, H . R . , J . J . Monaco and E . H . Leiter, 1992. Expression of i n t r a - M H C 130 transporter (Ham) genes and class I antigens in diabetes-susceptible N O D mice [letter; comment]. Science, 256: 1826-8. 3 5 . Attaya, M . , et al., 1992. Ham-2 corrects the class I antigen-processing defect in R M A - S cells. Nature, 355: 647-9. 36 . M o m b u r g , F . , J . Roelse, G . J . Hammerling and J . J . Neefjes, 1994. Peptide size selection by the major histocompatibility complex-encoded peptide transporter. J Exp Med, 179: 1613-23. 3 7 . Schumacher, T . N . , et al., 1994. Peptide length and sequence specificity of the mouse T A P 1 / T A P 2 translocator. Journal of Experimental Medicine, 179: 533-40. 38 . Powis, C . J . , J . C . Howard and G . W . Butcher, 1990. Variation in the biosynthesis of the rat R T l . A a classical class I antigen due to the cim system. Transplant Proc, 22: 2517-8 . 3 9 . Lodish , H . and N . Kong , 1990. Perturbation of Cellular Calc ium Blocks Exit of Secretory Proteins from the Rough Endoplasmic Reticulum. J.B.C., 265: 10893-10899. 40 . Lodish, H . and N . Kong, 1993. The Secretory Pathway is normal in Dithiothreitol-treated cells, but disulfide-bonded proteins are reduced and reversibly retained in the endoplasmic reticulum. 3.B.C., 268: 20598-20605. 4 1 . Sadasivan, B . , P . J . Lehner, B . Ortmann, T . Spies and P. Cresswell, 1996. Roles for calreticulin and a novel glycoprotein, tapasin, in the interaction of M H C class I molecules with T A P . Immunity, 5: 103-14. 42 . Wada, I., M . K a i , S. Imai, F . Sakane and H . K a n o h , 1997. Promotion o f transferrin folding by cyclic interactions with calnexin and calreticulin. Embo J, 16: 5420-32 . 4 3 . Sadasivan, B . K . , A . Cariappa, G . L . Waneck and P. Cresswell, 1995. A s s e m b l y , peptide loading, and transport of M H C class I molecules in a calnexin-negative cell line. Cold Spring Harb Symp Quant Biol, 60: 267-75. 44 . Hebert, D . N . , B . Foellmer and A . Helenius, 1996. Calnexin and calret icul in 131 promote fo ld ing , delay ol igomerizat ion and suppress degradation of inf luenza hemagglutinin in microsomes. Embo J, 15: 2961-8. 4 5 . Zhou , X . , F . Momburg , T . L i u , M . U . Abdel , M . Jondal , G . J . Hammerling and H . G . Ljunggren, 1994. Presentation of viral antigens restricted by H - 2 K b , Db or K d in proteasome subunit L M P 2 - and LMP7-deficient cells. Eur J Immunol, 24: 1863-8. 46 . Klar , D . and G . Hammerling, 1989. Induction of assembly of M H C class I heavy chains with beta2-microglobulin by interferon-gamma. Embo, 8: 475-481. 4 7 . Gabathuler, R . , G . Reid , G . Kolait is , J . Driscol l and W . A . Jefferies, 1994. Comparison of cell lines deficient in antigen presentation reveals a functional role for T A P - 1 alone in antigen processing. Journal of Experimental Medicine, 180: 1415-25. 48 . Blobel , G . , 1980. Intracellular protein topogenesis. Proc Natl Acad Sci USA, 77: 1496-500. 4 9 . A r a r , C , V . Carpentier, J .P . Le Caer, M . Monsigny, A . Legrand and A . C . Roche, 1995. E R G I C - 5 3 , a membrane protein of the endoplasmic reticulum-Golgi intermediate compartment, is identical to M R 6 0 , an intracel lular mannose-specific lectin o f myelomonocytic cells. J Biol Chem, 270: 3551-3. 50 . Ferreira , L . R . , K . Norris , T . Smith, C . Hebert and J . J . Sauk, 1994. Association of Hsp47, Grp78, and Grp94 with procollagen supports the successive or coupled action of molecular chaperones. J Cell Biochem, 56: 518-26. 5 1 . Bonnerot, C , M . S . M a r k s , P. Cosson, E . J . Robertson, E . K . Bikoff , R . N . Germain and J .S . Bonifacino, 1994. Association with B iP and aggregation of class II M H C molecules synthesized in the absence of invariant chain. Embo J, 13: 934-44. 52 . F l y n n , G . C . , J . Pohl , M . T . Flocco and J . E . Rothman, 1991. Peptide-binding specificity of the molecular chaperone BiP. Nature, 353: 726-30. 5 3 . Anderson, K . S . , J . Alexander, M . Wei and P. Cresswell , 1993. Intracellular transport of class I M H C molecules in antigen processing mutant cell lines. Journal of Immunology, 151: 3407-19. 132 54 . David, V . , F . Hochstenbach, S. Rajagopalan and M . B . Brenner, 1993. Interaction with newly synthesized and retained proteins in the endoplasmic reticulum suggests a chaperone function for human integral membrane protein IP90 (calnexin). J Biol Chem, 268: 9585-92. 55 . Degen, E . and D . Will iams, 1991. Participation of a Novel 88-kD Protein in the Biogenesis of Murine Class I Histocompatibility Molecules. J Cell Biol, 112: 1099-1115. 56 . Ware, F . E . , A . Vassilakos, P . A . Peterson, M . R . Jackson, M . A . Lehrman and D .B . W i l l i a m s , 1995. T h e molecu lar chaperone ca lnexin binds G l c l M a n 9 G l c N A c 2 oligosaccharide as an initial step in recognizing unfolded glycoproteins . J Biol Chem, 270: 4697-704. 57 . Hebert , D . N . , B . Foel lmer and A . Helenius, 1995. Glucose tr imming and reglucosylation determine glycoprotein association with calnexin in the endoplasmic ret iculum. Cell, 81: 425-33. 58 . Jackson, M . R . , D . M . Cohen, P . A . Peterson and D .B . Wil l iams, 1994. Regulat ion of M H C class I transport by the molecular chaperone, calnexin (p88, IP90). Science, 263: 384-7. 59 . Hammond, C . and A . Helenius, 1994. Folding of V S V G protein: sequential interaction with B iP and calnexin. Science, 266: 456-8. 60 . Trombetta, S . E . , M . Bosch and A . J . Parodi , 1989. Glucosylation of glycoproteins by mammalian, plant, fungal, and trypanosomatid protozoa microsomal membranes. Biochemistry, 28: 8108-16. 6 1 . Degen, E . , M . Cohen-Doyle and D . Will iams, 1992. Efficient Dissociation of the p88 Chaperone from M a j o r Histocompatibility Complex Class I molecules Requires Both 3 2-Microglobulin and Peptide. J Exp Med, 175: 1653-1661. 62 . Fenteany, G . , R . F . Standaert, W.S . Lane, S. Choi , E . J . Corey and S .L . Schreiber, 1995. Inhibition of proteasome activities and subunit-specific amino-terminal threonine modification by lactacystin. Science, 268: 726-31. 133 6 3 . Hughes , E . A . , C . H a m m o n d and P . Cresswel l , 1997. Mis fo lded major histocompatibility complex class I heavy chains are translocated into the cytoplasm and degraded by the proteasome. Proceedings of the National Academy of Sciences of the United States of America, 94: 1896-901. 64 . Braakman, I., J . Helenius and A . Helenius, 1992. Manipulat ing Disulfide Bond Formation and Protein Folding in the Endoplasmic Reticulum. Embo J, 11: 1717-1722. 65 . Gething, M . J . and J . Sambrook, 1992. Protein folding in the cell. Nature, 355: 33-4 5 . 66 . Creighton, T . E . , 1997. Protein folding coupled to disulphide bond formation. Biol Chem, 378: 731-44. 67 . Cox, J .S . , C . E . Shamu and P. Walter, 1993. Transcript ional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase. Cell, 73: 1197-206. 68 . Pahl , H . L . , M . Sester, H . G . Burgert and P . A . Baeuerle, 1996. Act ivat ion of transcription factor N F - k a p p a B by the adenovirus E 3 / 1 9 K protein requires its E R retention. Journal.of.Cell Biology., 132: 511-522. 69 . Swiedler, S., J . Freed , A . Tarent ino , J . P lummer and G . W a r t , 1985. Oligosaccharide microheterogeneity of the murine major histocompatibi l i ty antigens. Reproducible site-specific patterns of sialylation and branching in asparagine-linked oligosaccharids. The Journal of Biological Chemistry, 260: 4046-4054. 70 . Swiedler, S .J . , G . W . Hart , A . L . Tarentino, T . H . Plummer, J r . and J . H . Freed, 1983. Stable oligosaccharide microheterogeneity at individual glycosylation sites of a murine major histocompatibility antigen derived from a B- cell l y m p h o m a . J Biol Chem, 258: 11515-23. 7 1 . G a n a n , S., J . J . Cazzulo and A . J . Parodi , 1991. A major proportion of N -glycoproteins are transiently glucosylated in the endoplasmic reticulum. Biochemistry, 30: 3098-104. 134 72 . Sousa, M . C . , M . A . Ferrero -Garc ia and A . J . Parodi , 1992. Recognition of the oligosaccharide and protein moieties of glycoproteins by the UDP-Glc :g lycopro te in glucosyltransferase. Biochemistry, 31: 97-105. 73 . Dorner , A . , D . Bole and R. K a u f m a n , 1987. The Relationship of N-Iinked Glycosylat ion and Heavy Chain-b inding Protein Association with the Secretion of Glycoproteins. The Journal of Cell Biology, 105: 2665-2674. 74 . M o o r e , S . E . and R . G . Sp iro , 1993. Inhibi t ion of glucose t r imming by castanospermine results in rapid degradation of unassembled major histocompatibility complex class I molecules. J Biol Chem, 268: 3809-12. 75 . M c C r a c k e n , A . A . and J . L . Brodsky, 1996. Assembly of ER-associated protein degradation in vitro: dependence on cytosol, calnexin, and A T P . J Cell Biol, 132: 291-8. 76 . Huppa, J . B . and H . L . Ploegh, 1997. The alpha chain of the T cell antigen receptor is degraded in the cytosol. Immunity, 7: 113-22. 77 . Farquhar, M . G . and G . E . Palade, 1998. The Golgi apparatus: 100 years of progress and controversy. Trends Cell Biol, 8: 2-10. 78 . Schatz, G . and B . Dobberstein, 1996. Common principles of protein translocation across membranes. Science, 271: 1519-26. 7 9 . Mel lman, I. and K . Simons, 1992. The Golgi complex: in vitro Veritas?. [Review] [126 refs]. Cell, 68: 829-40. 80 . Pfeffer, S.R. and J . E . Rothman, 1987. Biosynthetic protein transport and sorting by the endoplasmic reticulum and Golgi . [Review] [117 refs]. Annual Review of Biochemistry, 56: 829-52. 8 1 . Schekman, R . , 1996. Polypeptide translocation: a pretty picture is worth a thousand words [comment]. Cell, 87: 593-5. 82 . Balch, W . E . , J . M . McCaffery , H . Plutner and M . G . Farquhar , 1994. Vesicular stomatitis virus glycoprotein is sorted and concentrated during export from the endoplasmic reticulum. Cell, 76: 841-52. 135 8 3 . Nishimura, N . and W . E . Balch, 1997. A di-acidic signal required for selective export from the endoplasmic reticulum. Science, 277: 556-8. 84 . Scheiffele, P . , J . Peranen and K . Simons, 1995. N-glycans as apical sorting signals in epithelial cells. Nature, 378: 96-8. 85 . Fiedler, K . , R . G . Parton, R. Kellner, T . Etzold and K . Simons, 1994. VIP36, a novel component of glycolipid rafts and exocytic carrier vesicles in epithelial cells. Embo Journal, 13: 1729-40. 86 . Zinkernagel , R . M . and P . C . Doherty, 1979. MHC-res tr i c t ed cytotoxic T cel ls: Studies on the biological role of polymorphic major transplantation antigens determining T-cell restriction-specificity, function, and responsiveness. Adv.Immunol., 27: 51-177. 87 . Takeshita, T . , et al., 1993. Role of conserved regions of class I M H C molecules in the activation of CD8+ cytotoxic T lymphocytes by peptide and purified cell-free class I molecules. Int Immunol, 5: 1129-38. 88 . Takahashi , H . , R. Houghten, S.D. Putney, D . H . Margul ies , B . Moss , R . N . Germain and J . A . Berzofsky, 1989. Structural requirements for class I M H C molecule-mediated antigen presentation and cytotoxic T cell recognition of an immunodominant determinant of the human immunodeficiency virus envelope protein. J Exp Med, 770: 2 0 2 3-35 . 89 . Sun , R . , S . E . Shepherd, S.S. Geier , C T . Thomson , J . M . Sheil and S . G . Nathenson, 1995. Evidence that the antigen receptors of cytotoxic T lymphocytes interact with a common recognition pattern on the H - 2 K b molecule. Immunity, 3: 573-82. 90 . Bjorkman, P . J . , M . A . Saper, B . Samraoui , W . S . Bennett, J . L . Strominger and D . C . Wiley, 1987. The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature, 329: 512-8. 9 1 . van Endert, P . M . , 1999. Genes regulating M H C class I processing of antigen. Curr Opin Immunol, 11: 82-8. 92 . Straus, S., Adenovirus Infections in Humans, in The Adenovirus, H . Ginsburg , Edi tor . 136 1984, Plenum: New Y o r k . p. 451-487. 9 3 . Fox, J . P . , C . E . Hal l and M . K . Cooney, 1977. The Seattle Virus Watch. V I I . Observations of adenovirus infections. Am J Epidemiol, 105: 362-86. 94 . de Jong, P . J . , G . Valderrama, I. Spigland and M . S . Horwitz , 1983. Adenovirus isolates from urine of patients with acquired immunodeficiency syndrome. Lancet, 1: 1293-1296. 9 5 . Hierholzer, J . C . , R. Wigand, L . J . Anderson, T . Adr ian and J . W . G o l d , 1988. Adenoviruses from patients with AIDS: a plethora of serotypes and a description of five new serotypes of subgenus D (types 43-47). Journal.of.Infectious.Diseases., 158: 804-813. 96 . Matsuse, T . , S. Hayashi, K . Kuwano, H . Keunecke, W . A . Jefferies and J . C . Hogg, 1992. Latent adenoviral infection in the pathogenesis of chronic airways obstruction. American Review of Respiratory Disease, 146: 177-84. 97 . Nermut, M . , The Architecture of Adenovirus, in The Adenovirus, H . Ginsburg , Edi tor . 1984, Plenum: New York . p. 5-32. 98 . Phil ipson, L . and K . Lonberg-Holm, 1969. Fate of adenovirus during the early phase of infection. J Gen Microbiol, 57: x-xi. 99 . Berk, A . J . , 1986. Adenovirus promoters and E l A transactivation. Annu Rev Genet, 20: 45-79. 100. Wold , W.S. and L . R . Gooding, 1991. Region E3 of adenovirus: a cassette of genes involved in host immunosurveillance and virus-cell interactions. [Review]. Virology., 184: 1-8. 101. Bernards, R. , P.I. Schrier, A . Houweling, J . L . Bos, A . J . van der E b , M . Zi j l s tra and C . J . Melief, 1983. Tumorigenicity of cells transformed by adenovirus type 12 by evasion of T-cell immunity. Nature, 305: 776-9. 102. W o l d , W . S . , C . Cladaras, S . L . Deutscher and Q.S . Kapoor , 1985. The 19-kDa glycoprotein coded by region E3 of adenovirus. Puri f icat ion , characterization, and structural analysis. J Biol Chem, 260: 2424-31. 137 103. Gooding, L . R . , L . W . Elmore, A . E . Tollefson, H . A . Brady and W . S . Wold , 1988. A 14,700 M W protein from the E3 region of adenovirus inhibits cytolysis by tumor necrosis factor. Cell, 53: 341-6. 104. Gooding, L . R . , I .O. Sofola, A . E . Tollefson, P. Duerksen-Hughes and W.S. W o l d , 1990. The adenovirus E3-14.7K protein is a general inhibitor of tumor necrosis factor-mediated cytolysis. J Immunol, 145: 3080-6. 105. C a r l i n , C . , A . Tollefson, H . Brady, B. Hoffman and W . W o l d , 1989. Epidermal growth factor receptor is down-regulated by a 10.400 M W protein encoded by the E3 region of adenovirus. Cell, 57: 135-144. 106. Andersson, M . , S. Paabo, T . Nilsson and P . A . Peterson, 1985. Impaired intracellular transport of class I M H C antigens as a possible means for adenoviruses to evade immune surveillance. Cell, 43: 215-22. 107. Tollefson, A . E . , A . R . Stewart, S.P. Ye i , S .K . Saha and W . S . Wold , 1991. The 10,400- and 14,500-dalton proteins encoded by region E3 of adenovirus form a complex and function together to down-regulate the epidermal growth factor receptor. J Virol, 65: 3095-105. 108. Gooding, L . R . , T . S . Ranheim, A . E . Tollefson, L . Aquino, P. Duerksen-Hughes, T . M . Horton and W.S . Wold , 1991. The 10,400- and 14,500-dalton proteins encoded by region E3 of adenovirus function together to protect many but not all mouse cell lines against lysis by tumor necrosis factor. J Virol, 65: 4114-23. 109. Hausmann, J . , D . Ortmann, E . Witt, M . Veit and W . Seidel, 1998. Adenovirus death protein, a transmembrane protein encoded in the E3 region, is palmitoylated at the cytoplasmic tail. Virology, 244: 343-51. 110. Flomenberg, P. , J . Szmulewicz, E . Gutierrez and H . Lupatkin , 1992. Role of the adenovirus E3-19k conserved region in binding major histocompatibility complex class I molecules. J Virol, 66: 4778-83. 111. Kornfe ld , R. and W . S . W o l d , 1981. Structures of the oligosaccharides of the 138 glycoprotein coded by early region E3 of adenovirus 2. J Virol, 40: 440-9. 112. F lomenberg , P . , M . Chen and M . Horwi tz , 1988. Sequence and Genetic Organization of Adenovirus Type 35 Early Region 3. Journal of Virology, 62: 4431-4437. 113. Sester, M . and H . G . Burgert, 1994. Conserved cysteine residues within the E3/19K protein of adenovirus type 2 are essential for binding to major histocompatibility complex antigens. J Virol, 68: 5423-32. 114. Gabathuler, R. and S. Kvist , 1990. The endoplasmic reticulum retention signal o f the E3/19K protein of adenovirus type 2 consists of three separate amino acid segments at the carboxy terminus. J Cell Biol, 111: 1803-10. 115. Signas, C , M . G . Katze, H . Persson and L . Phi l ipson, 1982. A n adenovirus glycoprotein binds heavy chains of class I transplantation antigens from man and mouse. Nature, 299: 175-8. 116. Paabo, S., F . Weber , O . K a m p e , W . Schaffner and P . A . Peterson, 1983. Association between transplantation antigens and a viral membrane protein synthesized from a mammalian expression vector. Cell, 33: 445-53. 117. Tanaka, Y . and S. Tevethia, 1988. Differential effect of adenovirus 2 E 3 / 1 9 K glycoprotein on the expresson of H - 2 K b and H-2Db Class I antigens and K - 2 K b and H -2Db restricted SV40 Specific C T L Mediated Lysis. Virology, 165: 357-366. 118. Severinsson, L . , I. Martens and P . A . Peterson, 1986. Differential association between two human M H C class I antigens and an adenoviral glycoprotein. J Immunol, 137: 1003-9. 119. Wold , W.S . and L . R . Gooding, 1989. Adenovirus region E3 proteins that prevent cyto lys i s by cytotox ic T cells and t u m o r necros i s f a c t o r . [Review] . Molecular.Biology.&.Medicine., 6: 433-452. 120. Ginsberg, H . S . , L . L . Moldawer, P .B . Sehgal, M . Redington, P . L . K i l i a n , R . M . Chanock and G . A . Prince, 1991. A mouse model for investigating the molecular pathogenesis of adenovirus pneumonia. Proc Natl Acad Sci USA, 88: 1651-5. 139 121. Graham, F . L . , J . Smiley, W . C . Russell and R. Na irn , 1977. Characteristics of a human cell line transformed by D N A from human Adenovirus type 5. Journal of General Virology, 36: 59-72. 122. Persson, H . , C . Signas and L . Philipson, 1979. Purification and characterization of an early glycoprotein from adenovirus type 2-infected cells. J Virol, 29: 938-48. 123. Levy , F . , R. Larsson and S. Kvist , 1991. Translocat ion of peptides through microsomal membranes is a rapid process and promotes assembly of H L A - B 2 7 heavy chain and beta 2-microglobulin translated in vitro. J Cell Biol, 115: 959-70. 124. O u , W . J . , P . H . Cameron, D . Y . Thomas and J . J . Bergeron, 1993. Association of folding intermediates of glycoproteins with calnexin during protein maturation. Nature, 364: 111-6. 125. Powis, S .J . , A . R . Townsend, E . V . Deverson, J . Bastin, G . W . Butcher and J . C . Howard, 1991. Restoration of antigen presentation to the mutant cell line R M A - S by an M H C - l i n k e d transporter. Nature, 354: 528-31. 126. Bikoff , E . K . , L . Jaffe, R . K . Ribaudo, G . R . Otten, R . N . Germain and E . J . Robertson, 1991. M H C class I surface expression in embryo-derived cell lines inducible with peptide or interferon. Nature, 354: 235-8. 127. Newcomb, J . R . and P. Cresswell, 1993. Structural analysis of proteolytic products of M H C class II-invariant chain complexes generated in vivo. J Immunol, 151: 4153-63. 128. Beier, D . C , J . H . Cox, D . R . V in ing , P. Cresswell and V . H . Engelhard, 1994. Association of human class I M H C alleles with the adenovirus E3/19K protein. Journal of Immunology, 152: 3862-72. 129. Burgert, H . - G . , J . Maryanski and S. Kvist, 1987. "E3/19K" protein of adenovirus type 2 inhibits lysis of cytolytic T lymphocytes by blocking cell-surface expression of histocompatibility class I antigens. Proceedings of the National Academy of Science of USA, 84: 1356-1360. 130. Cox, J . H . , R . M . Buller, J . R . Bennink, J . W . Yewdell and G . Karupiah , 1994. 140 Expression of adenovirus E3/19K protein does not alter mouse M H C class I-restricted responses to vaccinia virus. Virology, 204: 558-62. 131. Jefferies, W . A . and H . G . Burgert , 1990. E3/19k from Adenovirus 2 is an immunosubversive protein that binds to a structural motif regulating the intracellular transport of major histocompatibility complex class I proteins. Journal of Experimental Medicine, 172: 1653-1664. 132. Lapham, C . K . , I. Bacik, J . W . Yewdell, K . P . Kane and J . R . Bennink, 1993. Class I molecules retained in the endoplasmic reticulum bind antigenic peptides. J Exp Med, 177: 1633-41. 133. Jefferies, W . A . and G . G . MacPherson, 1987. Expression of the W6/32 H L A epitope by cells of rat, mouse, human and other species: critical dependence on the interaction of specific M H C heavy chains with human or bovine beta 2-microglobul in . European Journal of Immunology, 17: 1257-63. 134. Korner , H . and H . G . Burgert, 1994. Down-regulation of H L A antigens by the adenovirus type 2 E3/19K protein in a T-lymphoma cell line. Journal.of.Virology., 68: 1442-1448. 135. Fellous, M . , M . Kamoun, J . Wiels, J . Dausset, J . Clements, J . Zeuthen and G . K l e i n , 1977. Induction of H L A Express ion in D a u d i Cells after C e l l Fus ion . Immunogenetics, 5: 423-436. 136. Seong, R. , C . Clayberger, A . Krensky and J . Parnes, 1988. Rescue of Daudi C e l l H L A Expression by Transfection of the Mouse beta2 Microglobul in Gene. J Exp Med, 167: 288-299. 137. Gi l lo t , D . , A . N o u r i , S. Compton and R. Ol iver , 1993. Accurate and rapid assessment of M H C antigen upregulation following cytokine st imulat ion. J. 1mm. Methods, 165: 231-239. 138. Boehm, U . , T . Klamp, M . Groot and J . C . Howard, 1997. Cellular responses to interferon-gamma. [Review] [282 refs]. Annual Review of Immunology, 15: 749-95. 141 139. Jefferies, W . A . , G . Kola i t i s and R. Gabathuler , 1993. IFN-gamma-induced recognition of the antigen-processing variant C M T . 6 4 by cytolytic T cells can be replaced by sequential addition of beta 2 microglobulin and antigenic peptides. J Immunol, 151: 2974-85. 140. Seliger, B . , S. Hammers, A . Hohne, R. Zeidler, A . Knuth , C D . Gerharz and C . Huber, 1997. IFN-gamma-mediated coordinated transcriptional regulation of the human T A P - 1 and L M P - 2 genes in human renal cell carcinoma [In Process Citation]. Clin Cancer Res, 3: 573-8. 141. Townsend, A . , T . Elliott, V . Cerundolo, L . Foster, B . Barber and A . Tse, 1990. Assembly of M H C class I molecules analyzed in vitro [published erratum appears in C e l l 1990 Sep 21;62(6):following 1233]. Cell, 62: 285-95. 142. Aosa i , F . , T . H . Y a n g , M . Ueda and A . Yano , 1994. Isolation of naturally processed peptides from a Toxoplasma gondii- infected human B lymphoma cell line that are recognized by cytotoxic T lymphocytes. J Parasitol, 80: 260-6. 143. Powis, S., J . Howard and G . Butcher, 1991. The major histocompatibility complex class II-linked cim locus controls the kinetics of intracellular transport of a classical class I molecule. J. Exp. Med, 173: 913-921. 144. Ljunggren, H . G . , et al., 1990. Empty M H C class I molecules come out in the cold. Nature, 346: 476-80. 145. Shamu, C . E . , 1998. Splicing: H A C k i n g into the unfolded-protein response. Curr Biol, 8: R121-3. 146. Sidrauski, C , J .S . Cox and P. Walter, 1996. t R N A ligase is required for regulated m R N A splicing in the unfolded protein response [see comments]. Cell, 87: 405-13. 147. Menzel, R., F . Vogel, E . Kargel and W . H . Schunck, 1997. Inducible membranes in yeast: relation to the unfolded-protein-response pathway. Yeast, 13: 1211-29. 148. Lehner, P . J . , M . J . Surman and P. Cresswell, 1998. Soluble tapasin restores M H C class I expression and function in the tapasin-negative cell line .220. Immunity, 8: 221-31. 142 149. Kahn-Perles, B . , C . Boyer, B. Arnold , A . Sanderson, P. Ferrier and F . Lemonnier, 1987. Acquisit ion of H L A class I W6/32 defined antigenic determinant by heavy chains from different species following association with bovine beta (2)-microglobulin. The Journal of Immunology, 138: 2190-2196. 150. Ploegh, H . , H . O r r and J . Strominger, 1981. Biosynthesis and cell surface localization of nonglycosylated human histocompatibil ity antigens. Journal of Immunology, 126: 270-275. 151. Andersson, H . , I. Nilsson and G . von Heijne, 1996. Calnexin can interact with N -linked glycans located close to the endoplasmic reticulum membrane. FEBS Lett, 397: 321-4. 152. Kvist , S. and F . Levy, 1993. Ear ly events in the assembly of M H C class I antigens. [Review] [78 refs]. Seminars in Immunology, 5: 105-16. 153. Ortmann, B . , et al., 1997. A critical role for tapasin in the assembly and function of multimeric M H C class I - T A P complexes. Science, 277: 1306-9. 154. Rammensee, H . G . , T . Friede and S. Stevanoviic, 1995. M H C ligands and peptide motifs: first listing. Immunogenetics, 41: 178-228. 155. Dr icu , A . , M . Carlberg, M . Wang and O. Larsson, 1997. Inhibition of N-linked glycosylation using tunicamycin causes cell death in malignant cells: role of down-regulation of the insulin-like growth factor 1 receptor in induction of apoptosis. Cancer Res, 57: 543-8. 156. Burgert, H . G . and S. Kvist , 1987. The E3/19K protein of adenovirus type 2 binds to the domains of histocompatibility antigens required for C T L recognition. EMBO Journal, 6: 2019-2026. 157. K l a p p a , P. , R . B . Freedman and R. Z i m m e r m a n n , 1995. Protein disulphide isomerase and a lumenal cyclophilin-type peptidyl proly l cis-trans isomerase are i n transient contact with secretory proteins during late stages of translocation. EurJBiochem, 232: 755-64. 143 158. Freedman, R . B . , T . R . Hirst and M . F . Tuite, 1994. Protein disulphide isomerase: building bridges in protein folding. Trends Biochem Sci, 19: 331-6. 159. van Leeuwen, J . E . and K . P . Kearse, 1996. Calnexin associates exclusively with individual C D 3 delta and T cell antigen receptor ( T C R ) alpha proteins containing incompletely trimmed glycans that are not assembled into multisubunit T C R complexes. J Biol Chem, 271: 9660-5. 160. Otteken, A . and B . Moss, 1996. Calret icul in interacts with newly synthesized human immunodeficiency virus type 1 envelope glycoprotein, suggesting a chaperone function similar to that of calnexin. J Biol Chem, 271: 97-103. 161. K i m , P.S. and P. A r v a n , 1995. Calnexin and B iP act as sequential molecular chaperones during thyroglobulin folding in the endoplasmic reticulum. J Cell Biol, 128: 29-38 . 162. Nossner, E . and P. Parham, 1995. Species-specific differences in chaperone interaction of human and mouse major histocompatibility complex class I molecules. J Exp Med, 181: 327-37. 163. Sanders, S . L . and R. Schekman, 1992. Polypeptide Translocation Across the Endoplasmic Reticulum Membrane. J Biol Chem, 267: 13791-13794. 164. Haas, I . G . , 1994. B i P (GRP78) , an essential hsp70 resident protein in the endoplasmic reticulum. Experientia, 50: 1012-20. 165. Margolese, L . , G . L . Waneck, C . K . Suzuki , E . Degen, R . A . Flavel l and D . B . Wil l iams, 1993. Identification of the region on the class I histocompatibil ity molecule that interacts with the molecular chaperone, p88 (calnexin, IP90). J Biol Chem, 268: 17959-66 . 166. Carreno, B . M . , K . L . Schreiber, D . J . M c K e a n , I. Stroynowski and T . H . Hansen, 1995. Aglycosylated and phosphatidylinositol-anchored M H C class I molecules are associated with calnexin. Evidence implicating the class I- connecting peptide segment in calnexin association. J Immunol, 154: 5173-80. 144 167. Solheim, J . C . , M . R . Harr i s , C .S . Kindle and T . H . Hansen, 1997. Prominence of beta 2-microglobulin, class I heavy chain conformation, and tapasin in the interactions of class I heavy chain with calreticulin and the transporter associated with antigen processing. J Immunol, 158: 2236-41. 168. Mart ire , G . , et al., 1996. Different fate of a single reporter protein containing K D E L or K K X X target ing signals stably expressed in m a m m a l i a n cells. Journal.of.Biological.Chemistry., 271: 3541-3547. 169. Gabathuler, R. , F . Levy and S. Kvist , 1990. Requirements for the association of adenovirus type 2 E3/19K wild-type and mutant proteins with H L A antigens. J Virol, 64: 3679-85. 170. Kuehn , M . J . , J . M . Herrmann and R. Schekman, 1998. COPII-cargo interactions direct protein sorting into ER-derived transport vesicles. Nature, 391: 187-90. 171. Tisdale, E . J . , H . Plutner, J . Matteson and W . E . Balch, 1997. p53/58 binds C O P I and is required for selective transport through the early secretory pathway. Journal of Cell Biology, 137: 581-93. 172. Schekman, R. and I. Mel lman, 1997. Does C O P I go both ways?. [Review] [20 refs]. Cell, 90: 197-200. 173. Lippe , R. , E . Luke , Y . T . K u a h , C . Lomas and W . A . Jefferies, 1991. Adenovirus infection inhibits the phosphorylation of major histocompatibility complex class I proteins. J Exp Med, 174: 1159-66. 174. Townsley, F . M . and H . Pelham, 1994. The K K X X signal mediates retrieval of membrane proteins from the Golgi to the E R in yeast. Eur J Cell Biol, 64: 211-216. 175. Pahl , H . L . and P . A . Baeuerle, 1997. The ER-overload response: activation of N F -kappa B. [Review] [37 refs]. Trends in Biochemical Sciences, 22: 63-7. 176. Brewer, J . W . , J . L . Cleveland and L . M . Hendershot, 1997. A pathway distinct from the mammalian unfolded protein response regulates expression of endoplasmic reticulum chaperones in non-stressed cells. Embo Journal, 16: 7207-16. 145 177. Cox, J .S . , R . E . Chapman and P. Walter, 1997. The unfolded protein response coordinates the production of endoplasmic reticulum protein and endoplasmic ret iculum membrane. Molecular Biology of the Cell, 8: 1805-14. 178. Roche, P . A . , M . S . Marks and P. Cresswell, 1991. Formation of a nine-subunit complex by H L A class II glycoproteins and the invariant chain. Nature, 354: 392-4. 179. Anonymous, The Handbook of Biochemistry and Biophysics. 1966, Cleveland: Cleveland W o r l d Publishing. 180. McMaster , W . R . and A . F . Williams, 1979. Identification of l a glycoproteins in rat thymus and purification from rat spleen. Eur J Immunol, 9: 426-33. 181. Ha id l , I., Characterization of Dendritic Cell and Macrophage Cell Surface Proteins, . 1996, University of Brit ish Columbia. 146 Appendix 1 Lysates from 293 transfectanst were precipitated with the E3/19K specific monoclonal Tw 1.7 at 30 and 120 minutes post pulse. Arrows denote E3/19K and co-immunoprecipitated MHC CI I. Db Kb Appendix 2 Velocity Gradient centrifugation A2.1. Introduction The work described in the previous chapters suggested a complex interaction between E3/19K, MHC CI I, P2m and other ER resident proteins. Additionally, it appeared that the ability of E3/19K to bind to specific MHC CI I proteins is based partly on affinity and partly on the time resident in the same compartment within the cell. The experiments with 293Db demonstrated that there might exist different populations of E3/19K which differentially bound D b (and possibly the other proteins). Finally, there was evidence that the complex that E3/19K formed in the ER involves calnexin and the TAPs, with the possibility of other chaperones being present in the complex. To address these questions, the size and shape of the MHC-E3/19K complexes was investigated. The use of the velocity sedimentation gradients allowed the opportunity to further divide the intracellular MHC-E3/19K of a cell into subgroups based on the distance it migrated through a sucrose gradient when subjected to centrifugal force for a set period of time. In previous experiments, lysates were solubilized in relatively harsh detergents, effectively destroying anything but the strongest of associations between molecules. By using gentle detergents before separating, protein complexes could be identified by their size. A specific protein such as MHC CI I likely associates with many different proteins in various complexes throughout its maturation stages through the cell. Identification of these different intermediates would help identify the stages of the maturation of the MHC CI I molecule and would help in understanding its ability to bind to E3/19K. It was suggested that MHC-E3/19K binding is more complex than first believed. It was further hypothesised that E3/19K exists in multimers. This would shed some insight into the life cycle of this protein and the optimal conditions for this protein to associate with 148 MHC CI I. This in turn would offer insight on the evasion of the immune system by the adenovirus. A2.2. Rationale and Goals. To determine the relative size of the E3/19K complex, cell lysates were run over a continuous sucrose gradient. In a series of experiments that utilized the gentle detergents N-Octyl Glucoside, Chaps and digitonin, lysates were separated according to their size and shape. The proteins were identified with immunoprecipitation. Further experiments resulted in the characterization of attributes of these proteins in an attempt to determine the intracellular compartment they had resided at the time of labelling. A2.3. Results These experiments set out to determine the size of the MHC CI I complex by identifying the fraction or fractions in which these molecules eluted. Initial results showed that MHC CI I was present in several different sucrose gradient fractions when precipitated with W6/32. The precipitated MHC CI I had a different MW when precipitated from different fractions. In 293.12 cells the MHC CI I complex associated with E3/19K eluted in different sets of fractions than those in which the MHC CI I - P2m complex (in 293 cells) alone eluted. These results gave an estimate of the size of the complex consisting of E3/19K bound to MHC CI I. Furthermore, it was observed that E3/19K also eluted in fractions corresponding to larger complexes than the fractions containing MHC molecules. A2.3.1. Sucrose gradients separate cellular components by size In figure 23, 293 cells were labelled overnight and the lysate run on a continuous 5-20% sucrose gradient as described in Materials and Methods. The purpose of this experiment 149 was to establish that the sucrose gradient did separate proteins based on size. This experiment allowed the determination of the characteristics and separating qualities of a sucrose gradient. The higher mass proteins travelled further through the sucrose and eluted in different fractions. This was not unexpected and indicated a good method for analyzing the data in the following gels. 150 Figure 23. Separation of 293 lysates over a 5-20% sucrose gradient 293 cells were grown to confluence and labelled as above. Cells were labelled for two hours and lysed in N-Octyl Glucoside lysis buffer. 300 Lil of the lysate was layered on top of a 5-20% sucrose gradient and spun in a SW41 rotor at 36,800 rpm for 28 hours. Twelve 1 ml fractions were collected (lane 1 is the top fraction; lane 12 the bottom) and separated without precipitation by SDS-PAGE. Subsequent transfer to an Immobilon PVDF membrane followed, a) The membrane was exposed to Kodak XAR film for a period ranging from one day to three months (at -80°C). 151 A2.3.2. M H C CI I complex precipitated from specific sucrose fraction 293 lysates run over 5-20% sucrose gradients were precipitated with W6/32. This yielded MHC CI I- P2m complexes in fractions 3, 4 and 5 (figure 24). MHC CI I was found to have migrated to other fractions as well, but P 2m was not observed associated in these fractions. This suggested that P2m does not associate with all the MHC CI I present in a cell, but rather with a subset. The fraction of MHC CI I that does not associate with P 2m represented a distinct subpopulation not observed before. To elucidate differences between MHC CI I complexes with and without E3/19K, 293.12 cell lysates were separated using the same technique and immunoprecipitated with W6/32. In the 293.12 cells, the MHC CI I complex (MHC CI I, P2m, E3/19K) was found in denser fractions (fractions 5, 6 and 7). E3/19K is observed co-immunoprecipitating in these fractions. The difference in migration, and therefore their size, of the complexes in the two cells is very likely due to E3/19K. 153 Figure 24. Precipitation of M H C CI I and E3/19K from 293 and 293.12 cell lysates separated over a 5-20% sucrose gradient 293 (a) and 293.12 (b) cells were grown to confluence and labelled as above. Cells were labelled for 2 hours and lysed in N-Octyl Glucoside lysis buffer. 300 Lil of the lysate was layered on top of a 5-20% sucrose gradient and spun in an SW41 rotor at 36,800 rpm for 28 hours. Eleven 1 ml fractions were collected (lane 1 is the top fraction; lane 11 the bottom). Fractions were precipitated with W6/32 or the anti E3/19K antisera R418. Precipitates were separated using SDS-PAGE. 154 CD 00 co * t co oo r- CM DC r -0 TJ (0 (0 o (/) CM CM CO CO O) > CM > O CD CO CO CM I I I I co CD CD CO CD CM \5S In the 293.12 lysates, E3/19K did not localize exclusively in those fractions with labelled MHC CI I present. In addition to the fractions with labelled MHC CI I and P2m complexes, E3/19K seemed to migrate further through the sucrose gradient ending up in fractions where MHC CI I was not observed. A band at -25 kDa and a band at ~50 kDa was also observed in these fractions (8-10). This led to speculation that the sucrose gradient was separating different sub-populations of E3/19K complexes within the 293.12 cells. The standard E3/19K complex observed in earlier experiments consisting of P2m> MHC CI I, and E3/19K were localised to fractions 5, 6 and 7. E3/19K complexes without observable MHC CI I, which could be multimers, were found in the fractions 8, 9 and 10. The discovery of E3/19K in different fractions suggested that they may be present in different sub-populations. These sub-populations may localise with different ER resident proteins. In some complexes, E3/19K may form multimers that do not include MHC CI I and P2m. The observation of different subpopulations of E3/19K suggested these subpopulations may exhibit differing affinities and abilities to form complexes with MHC CI I. Additionally, MHC CI I with different mobilities was observed in different fractions. Variance in mobility reflects differences in the maturity of the MHC CI I; specifically in the post translational modifications of these proteins. It is possible that E3/19K has differing affinities for MHC CI I based on its maturity and its corresponding conformation. If E3/19K is in the presence of an MHC CI I molecule that is in a favourable conformation for binding, then a complex will form. If not, E319K may form complexes with other proteins, including itself, resulting in multimers. 157 A2.4. Discussion This set of experiments demonstrated the MHC CI I complex can be precipitated from different sucrose gradient fractions suggesting that the complex exists in different forms. In 293.12 cells, the complex precipitated by W6/32 consists of MHC CI I H chain, P2m, peptide and E3/19K. This complex does elute in a different fraction from the corresponding MHC CI I complex (without E3/19K) found in 293 cells. In addition finding E3/19K present in complexes with MHC-(32m complex, populations of E3/19K are likely present in the ER in other complexes. Association in these complexes is both with and more likely without MHC CI I molecules. In addition to determining which fractions of the sucrose gradient the different subpopulations of these proteins precipitate in, the sucrose gradient can be used to calculate the sedimentation coefficient (S value) of the protein complexes and the apparent molecular weight of these complexes. These S values are calculated by determining the distance migrated through a gradient of known density by a complex acted on by known centrifugal force. To convert the S value to the MW, estimation or determination of the stokes radius must be incorporated. The S value was calculated using formulae previously published (178). The sedimentation coefficient was given by the formula: S20,w = S T , m [ T l T , m ( P p - p 2 0 , w ) / T l 2 0 , w ( P p - p T , m ) ] where r| and r| 2 o ^  is the viscosity of sucrose at temperature T or water at 20°C, respectively. P T m and p 2 Q w are the densities of the sucrose at temperature T or water at 20°C. 158 p p is the density of the particle analyised. This can also be the reciprocal of the partial specific volume, 1/v. The values for the sucrose viscosity and density were obtained from standard tables (179). The partial specific volume of the protein-detergent complex was assumed to be similar to other transmembrane glycoprotein-detergent complexes. It was further assumed that these proteins bind 0.5 grams of detergent per gram of protein. S T M is given by the equation S T =ln(r 7 r ) / ( c o 2 t ) Where r 2 and r is the distance travelled from r to r , co is the angular velocity of the rotor and t is time. From the S value, the molecular weight of both the protein-detergent complex and the protein can be calculated according to the following: M =(6TIN S „ R )/(l-vp„ A ) r A 20,w e 7 ^20,w 7 with N. being Avogadro's number and R the stokes radius of the protein. The stokes radius was estimated to be 4.4 nm (180). MW estimates obtained using these sucrose gradients showed agreement with previously published values for MHC Class II complexes (181). The size of the protein and the radius of the detergent with the proteins contribute to the size of the complex. Substituting the calculated S value in the formula above, the apparent molecular weight for a complex can be determined (see table 6). The MHC CI I-P2m complex in 293 cells in the third and fourth fractions give values of 60-75 kDa. This compares well to the literature value of 60 kDa. In 293.12 cells, the MHC CI I, P2m - E3/19K complex were calculated to be in the 95-116 kDa range. This value makes it is quite possible that a second E3/19K protein is present in this 159 complex. This would make the ratio of MHC CI I to (32m to E3/19K to be 1:1:2. Table 6. Determination of Molecular Weight (MW) of proteins in sucrose gradient fractions Fraction 2 r S MW (kD) (1) MW (kD) (2) 3 9.19 4.18 93 56 4 9.89 5.57 123 75 5 10.6 7.06 156 95 6 11.3 8.62 191 116 7 11.9 10.42 230 140 8 12.7 12.69 281 170 (1) MW of protein and detergent (2) MW of protein alone The generalised calculations of the MW for complexes found in 12 different 1 ml fractions from samples layered on top of a continuous 5%-20% sucrose gradient and separated for 28 hours at 36,400 rpm (148,000g) are shown above. The S and MW of proteins found in different fractions was calculated using formulae in text, r is the distance migrated through the gradient. Viscosity and density of fractions (not shown) are derived from standard tables (179). The variables Stokes Radius (Re) was previously determined to be 4.5 nm (180). The association of E3/19K with MHC CI I as a dimer is not completely unexpected. In experiments using the E3/19K-Vaccinia vector, E3/19K is overexpressed and many bands with molecular weights approximately twice that of E3/19K are observed (data not shown). 160 Additionally, in an earlier chapter that investigates the steric interactions of E3/19K with MHC CI I, it was proposed that for E3/19K to effectively associate with the specific amino acids of the MHC CI I sequence shown to influence its binding, E3/19K may act in concert with additional factors that perturbs the 3-D structure of the MHC CI I molecule. A novel method of membrane perturbation is proposed for E3/19K to extend far enough into the lumen of the ER for E3/19K to reach the residues on the MHC CI I molecule that affect its binding. Multimers of E3/19K could achieve this. E3/19K is also present in the denser fractions. E3/19K appeared in fractions corresponding to MW of 116-170kD in fractions precipitated with MHC CI I antisera. Unlike the less dense fractions, in these fractions labelled MHC CI I is not apparent. Also present is the band that correlates to a protein approximately double the size of E3/19K. This complex could be a multimer of E3/19K. In the denser fractions, it is important to note that these complexes are still being precipitated with anti-MHC CI I antisera. While labelled MHC CI I is not visible under these conditions, it is likely still present. These fractions could be illustrating a novel interaction between multimers of E3/19K and MHC CI I that may precede labelling in the pulse/chase. E3/19K may have associated with MHC CI I molecules prior to labelling and entered into a very stable conformation. E3/19K may form very long-lived associations with MHC CI I molecules. The appearance of a band at approximately double the size of E3/19K (about 50 kDa) further suggests that the MHC CI I, P2m, E3/19K complex is larger than the one normally observed with MHC CI I and P2m and that multimers of E3/19K may be involved. In an earlier chapter, it was observed that the D b protein in 293Db does not bind to E3/19K until several hours post chase. The earlier results showed that one of the features of the D b protein is that it needs a longer time to bind to E3/19K. Additionally, in 293 D b cells, 161 several differently glycosylated subsets of the E3/19K population were observed. In the 293 D b it is the fully glycosylated version of E3/19K that first binds to the D b protein. Later time points reveal that all the other subsets are found associated with Db. This could represent a cooperative binding of E3/19K to MHC CI I. Identification of the slowbinding protein, Db, allows the proposal that in the case of some proteins such as L d, binding to E3/19K may take a very long time to occur. While the cell surface expression of L d is markedly reduced, very little co-immunoprecipitation throughout the time course of the experiments is observed. E3/19K may be a very long lived resident of the ER, and it may bind to different subsets of MHC CI I dependant on its maturity. The existence of E3/19K in the denser fractions of the sucrose gradient could be due to its participation in a larger complex with other ER resident proteins or chaperones. This may represent portions of the total E3/19K population that hitherto go undetected in the MHC CI I co-precipitation. Both these findings and the literature report that the amount of E3/19K precipitated by anti E3/19K antisera is much greater than the E3/19K that co-precipitates with MHC CI I. There probably exists a pool of E3/19K that never associates with MHC CI I, or at least does not associate within the time constraints of the experiments reported here. The E3/19K found in the denser fractions may be this subset of the total E3/19K population. E3/19K in this fraction may also be associated with the ER resident proteins such as calnexin and others. The E3/19K complex with MHC CI I is a little more complicated than originally thought. Previously it had been assumed that E3/19K bound to the MHC CI I complex as a monomer. A sub population of E3/19K is present in a larger complex that may involve multimers of E3/19K. Additionally, the ER resident chaperones such as calnexin and TAP may be present. These complexes are originally observed when precipitated with anti MHC CI 162 I sera and antibodies. However, labelled MHC CI I is not observed in these fractions. This suggests that the MHC CI I component of these complexes has not been labelled. This may reflect an association of MHC CI I with E319K that is of longer duration than the length of the pulse-chase. 163 


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