8536d_ch09 200-220 8/2/02 1:00 PM Page 200 mac79 Mac 79: 45 Bw: glasby et al. Immunology 5e: chapter 9 T-Cell Receptor ART TO COME HE ANTIGEN-SPECIFIC NATURE OF T-CELL RESPONSES clearly implies that T cells possess an antigen specific and clonally restricted receptor. However, the identity of this receptor remained unknown long after the B-cell receptor(immunoglobulin molecule)had been identified. Relevant experimental results were contradictory and difficult to conceptualize within a single model because the T-cell receptor(tCr)differs from the B-cell antigen binding receptor in important ways. First, the T-cell receptor Interaction of aB TCR with Class II MHC-Peptide is membrane bound and does not appear in a soluble form as the B-cell receptor does; therefore, assessment of its struc- ture by classic biochemical methods was complicated, and Early Studies of the T-Cell Receptor omplex cellular assays were necessary to determine its speci a aB and y8 T-Cell Receptors: Structure and Roles ficity. Second, most T-cell receptors are specific not for anti gen alone but for antigen combined with a molecule encoded a Organization and Rearrangement of TCR Genes by the major histocompatibility complex(MHC). This prop- T-Cell Receptor Complex: TCR-CD3 erty precludes purification of the T-cell receptor by simple a T-Cell Accessory Membrane Molecules ntigen-binding techniques and adds complexity to any perimental system designed to investigate the receptor. a Three-Dimensional Structures of TCR-Peptide A combination of immunologic, biochemical, and MHC Complexes molecular-biological manipulations has vercome the roblems. The molecule responsible for T-cell specificity Alloreactivity of T Cells was found to be a heterodimer composed of either a and B or y and 8 chains. Cells that express TCRs have approx ately 10 TCR molecules on their surface. The genomic of the t-ce ne families and means by which the diversity of the component chains is identify and isolate its antigen-binding receptor. The obvi- generated were found to resemble those of the B-cell re- ous parallels between the recognition functions of T cells ceptor chains. Further, the T-cell receptor is associated on and B cells stimulated a great deal of experimental effort to the membrane with a signal-transducing complex, CD3, take advantage of the anticipated structural similarities be whose function is similar to that of the Ig-a/Ig-B complex tween immunoglobulins and T-cell receptors. Reports of the B-cell receptor. published in the 1970s claimed discovery of immunoglob- Important new insights concerning T-cell receptors have ulin isotypes associated exclusively with T cells(IgT)and been gained by recent structure determinations using x- of antisera that recognize variable-region markers(idio- crystallography, including new awareness of differences in types) common to antibodies and T-cell receptors with how TCRs bind to class I or class II MHC molecules. This similar specificity. These experiments could not be repro- chapter will explore the nature of the T-cell receptor mole- duced and were proven to be incorrect when it was demon- cules that specifically recognize MHC-antigen complexes, as strated that the T-cell receptor and immunoglobulins do well as some that recognize native antigens. not have common recognition elements and are encoded by entirely separate gene families. As the following sections will show, a sequence of well-designed experiments using Early Studies of the T-Cell Receptor cutting-edge technology was required to correctly answer questions about the structure of the T-cell receptor, the By the early 1980s, investigators had learned much about genes that encode it, and the manner in which it recognizes T-cell function but were thwarted in their attempts to antigen
identify and isolate its antigen-binding receptor. The obvious parallels between the recognition functions of T cells and B cells stimulated a great deal of experimental effort to take advantage of the anticipated structural similarities between immunoglobulins and T-cell receptors. Reports published in the 1970s claimed discovery of immunoglobulin isotypes associated exclusively with T cells (IgT) and of antisera that recognize variable-region markers (idiotypes) common to antibodies and T-cell receptors with similar specificity. These experiments could not be reproduced and were proven to be incorrect when it was demonstrated that the T-cell receptor and immunoglobulins do not have common recognition elements and are encoded by entirely separate gene families. As the following sections will show, a sequence of well-designed experiments using cutting-edge technology was required to correctly answer questions about the structure of the T-cell receptor, the genes that encode it, and the manner in which it recognizes antigen. chapter 9 ■ Early Studies of the T-Cell Receptor ■ and T-Cell Receptors: Structure and Roles ■ Organization and Rearrangement of TCR Genes ■ T-Cell Receptor Complex: TCR-CD3 ■ T-Cell Accessory Membrane Molecules ■ Three-Dimensional Structures of TCR-PeptideMHC Complexes ■ Alloreactivity of T Cells T-Cell Receptor T - - clearly implies that T cells possess an antigenspecific and clonally restricted receptor. However, the identity of this receptor remained unknown long after the B-cell receptor (immunoglobulin molecule) had been identified. Relevant experimental results were contradictory and difficult to conceptualize within a single model because the T-cell receptor (TCR) differs from the B-cell antigenbinding receptor in important ways. First, the T-cell receptor is membrane bound and does not appear in a soluble form as the B-cell receptor does; therefore, assessment of its structure by classic biochemical methods was complicated, and complex cellular assays were necessary to determine its specificity. Second, most T-cell receptors are specific not for antigen alone but for antigen combined with a molecule encoded by the major histocompatibility complex (MHC). This property precludes purification of the T-cell receptor by simple antigen-binding techniques and adds complexity to any experimental system designed to investigate the receptor. A combination of immunologic, biochemical, and molecular-biological manipulations has overcome these problems. The molecule responsible for T-cell specificity was found to be a heterodimer composed of either and or and chains. Cells that express TCRs have approximately 105 TCR molecules on their surface. The genomic organization of the T-cell receptor gene families and the means by which the diversity of the component chains is generated were found to resemble those of the B-cell receptor chains. Further, the T-cell receptor is associated on the membrane with a signal-transducing complex, CD3, whose function is similar to that of the Ig-/Ig- complex of the B-cell receptor. Important new insights concerning T-cell receptors have been gained by recent structure determinations using x-ray crystallography, including new awareness of differences in how TCRs bind to class I or class II MHC molecules. This chapter will explore the nature of the T-cell receptor molecules that specifically recognize MHC-antigen complexes, as well as some that recognize native antigens. Early Studies of the T-Cell Receptor By the early 1980s, investigators had learned much about T-cell function but were thwarted in their attempts to Interaction of TCR with Class II MHC–Peptide ART TO COME 8536d_ch09_200-220 8/2/02 1:00 PM Page 200 mac79 Mac 79:45_BW:Goldsby et al. / Immunology 5e:
8536d_chog_200-220 8/2/02 9: 49 AM Page 201 mac79 Mac 79: 45_Bw Glasby et al. Immunology 5e Classic Experiments Demonstrated the T-Cell Receptors Were Isolated by Using Self-MHC Restriction of the T-Cell Receptor Clonotypic Antibodies By the early 1970s, immunologists had learned to generate Identification and isolation of the T-cell receptor was accom- cytotoxic T lymphocytes(CTLs) specific for virus-infected plished by producing large numbers of monoclonal antibod- target cells. For example, when mice infected with lym ies to various T-cell clones and then screening the antibodies phocytic choriomeningitis(LCM)virus, they wou to find one that was clone specific, or clonotypic. This ap CTLs that could lyse LCM-infected target cells in vitro. Yet proach assumes that, since the T-cell receptor is specific for these same CTls failed to bind free LCM virus or viral anti- both an antigen and an MHC molecule, there should be sig- gens. Why didnt the CTls bind the virus or viral antigens di- nificant structural differences in the receptor from clone to tly as immunoglobulins did? The answer began to emerge clone; each T-cell clone should have an antigenic marker in the classic experiments of R. M. Zinkernagel and P. C. similar to the idiotype markers that characterize monoclonal Doherty in 1974(see Figure 8-2). These studies demon- antibodies. Using this approach, researchers in the early strated that antigen recognition by t cells is specific not for 1980s isolated the receptor and found that it was a het viral antigen alone but for antigen associated with an MHc erodimer consisting of a and B chains molecule(Figure 9-1). T cells were shown to recognize anti- When antisera were prepared using aB heterodimers iso- gen only when presented on the membrane of a cell by a self- lated from membranes of various T-cell clones, some antis MHC molecule. This attribute, called self-MHC restriction, era bound to aB heterodimers from all the clones, whereas distinguishes recognition of antigen by T cells and B cells. In other antisera were clone specific. This finding suggested that 1996, Doherty and Zinkernagel were awarded the Nobel the amino acid sequences of the tCR a and B chains, like Prize for this work. those of the immunoglobulin heavy and light chains, have Two models were proposed to explain the MHC restric- constant and variable regions. Later, a second type of TCR tion of the T-cell receptor. The dual-receptor model envi- heterodimer consisting of 8 and y chains was identified. In sioned a T cell with two separate receptors, one for antigen human and mouse, the majority of T cells express the ap het and one for class I or class II MHC molecules. The altered-self erodimer; the remaining T cells express the y8 heterodimer model proposed that a single receptor recognizes an alter- As described below, the exact proportion ofT cells expressing ation in self-MHC molecules induced by their association aB or y8 TCRs differs by organ and species, but aB T cells with foreign antigens. The debate between proponents of normally predominate these two models was waged for a number of years, until ar legant experiment by ) Kappler and p. Marrack demon- The TCR B-Chain Gene Was Cloned by Use strated that specificity for both MHC and antigen resides in a of Subtractive Hybridization single receptor. An overwhelming amount of structural and functional data has since been added in support of the In order to identify and isolate the TCR genes,SM.Hedrick nd M. M. Davis sought to isolate mRNA that encodes the a and B chains from a TH-cell clone. This was no easy task be cause the receptor mRNA represents only a minor fraction of (b) (c) the total cell mRNA. By contrast, in the plasma cell, im- munoglobulin is a major secreted cell product, and mRNAs encoding the heavy and light chains are abundant and easy to 其-x The successful scheme of hedrick and davis assumed that peptideAgO the TCR mRNA-like the mRNAs that encode other integral membrane proteins-would be associated with membrane- Self MHC bound polyribosomes rather than with free cytoplasmic ri- bosoms. They therefore isolated the membrane-bound target cell target cell polyribosomal mRNA from a TH-cell clone and used reverse ranscriptase to synthesize -P-labeled cDNA probes(Figure 9-2). Because only 3% of lymphocyte mRNA is in the Killing No killing No killing membrane-bound polyribosomal fraction, this step elimi ated 97% of the cell mrNA FIGURE 9-1 Self-MHC restriction of the T-cell receptor(TCR). A Hedrick and Davis next used a technique called DNA sub particular TCR is specific for both an antigenic peptide and a self- tractive hybridization to remove from their preparation the MHC molecule. In this example, the H-2 CTL is specific for viral pep- [P]cDNA that was not unique to T cells. Their rationale for tide a presented on an H-2 target cell (a) Antigen recognition does this step was based on earlier measurements by Davis show not occur when peptide B is displayed on an H-2 target cell(b) nor ing that 98% of the genes expressed in lymphocytes are com when peptide A is displayed on an H-2 target cell (c) mon to B cells and T cells. The 2% of the expressed genes that
T-Cell Receptor CHAPTER 9 201 Classic Experiments Demonstrated the Self-MHC Restriction of the T-Cell Receptor By the early 1970s, immunologists had learned to generate cytotoxic T lymphocytes (CTLs) specific for virus-infected target cells. For example, when mice were infected with lymphocytic choriomeningitis (LCM) virus, they would produce CTLs that could lyse LCM-infected target cells in vitro. Yet these same CTLs failed to bind free LCM virus or viral antigens. Why didn’t the CTLs bind the virus or viral antigens directly as immunoglobulins did? The answer began to emerge in the classic experiments of R. M. Zinkernagel and P. C. Doherty in 1974 (see Figure 8-2). These studies demonstrated that antigen recognition by T cells is specific not for viral antigen alone but for antigen associated with an MHC molecule (Figure 9-1). T cells were shown to recognize antigen only when presented on the membrane of a cell by a selfMHC molecule. This attribute, called self-MHC restriction, distinguishes recognition of antigen by T cells and B cells. In 1996, Doherty and Zinkernagel were awarded the Nobel Prize for this work. Two models were proposed to explain the MHC restriction of the T-cell receptor. The dual-receptor model envisioned a T cell with two separate receptors, one for antigen and one for class I or class II MHC molecules. The altered-self model proposed that a single receptor recognizes an alteration in self-MHC molecules induced by their association with foreign antigens. The debate between proponents of these two models was waged for a number of years, until an elegant experiment by J. Kappler and P. Marrack demonstrated that specificity for both MHC and antigen resides in a single receptor. An overwhelming amount of structural and functional data has since been added in support of the altered-self model. T-Cell Receptors Were Isolated by Using Clonotypic Antibodies Identification and isolation of the T-cell receptor was accomplished by producing large numbers of monoclonal antibodies to various T-cell clones and then screening the antibodies to find one that was clone specific, or clonotypic. This approach assumes that, since the T-cell receptor is specific for both an antigen and an MHC molecule, there should be significant structural differences in the receptor from clone to clone; each T-cell clone should have an antigenic marker similar to the idiotype markers that characterize monoclonal antibodies. Using this approach, researchers in the early 1980s isolated the receptor and found that it was a heterodimer consisting of and chains. When antisera were prepared using heterodimers isolated from membranes of various T-cell clones, some antisera bound to heterodimers from all the clones, whereas other antisera were clone specific. This finding suggested that the amino acid sequences of the TCR and chains, like those of the immunoglobulin heavy and light chains, have constant and variable regions. Later, a second type of TCR heterodimer consisting of and chains was identified. In human and mouse, the majority of T cells express the heterodimer; the remaining T cells express the heterodimer. As described below, the exact proportion of T cells expressing or TCRs differs by organ and species, but T cells normally predominate. The TCR -Chain Gene Was Cloned by Use of Subtractive Hybridization In order to identify and isolate the TCR genes, S. M. Hedrick and M. M. Davis sought to isolate mRNA that encodes the and chains from a TH-cell clone. This was no easy task because the receptor mRNA represents only a minor fraction of the total cell mRNA. By contrast, in the plasma cell, immunoglobulin is a major secreted cell product, and mRNAs encoding the heavy and light chains are abundant and easy to purify. The successful scheme of Hedrick and Davis assumed that the TCR mRNA—like the mRNAs that encode other integral membrane proteins—would be associated with membranebound polyribosomes rather than with free cytoplasmic ribosomes. They therefore isolated the membrane-bound polyribosomal mRNA from a TH-cell clone and used reverse transcriptase to synthesize 32P-labeled cDNA probes (Figure 9-2). Because only 3% of lymphocyte mRNA is in the membrane-bound polyribosomal fraction, this step eliminated 97% of the cell mRNA. Hedrick and Davis next used a technique called DNA subtractive hybridization to remove from their preparation the [ 32P]cDNA that was not unique to T cells. Their rationale for this step was based on earlier measurements by Davis showing that 98% of the genes expressed in lymphocytes are common to B cells and T cells. The 2% of the expressed genes that H-2k CTL TCR H-2k target cell Killing Viral peptide A Self MHC H-2k CTL TCR H-2d target cell No killing Viral peptide A Nonself MHC H-2k CTL TCR H-2k target cell No killing Viral peptide B Self MHC (a) (b) (c) FIGURE 9-1 Self-MHC restriction of the T-cell receptor (TCR). A particular TCR is specific for both an antigenic peptide and a selfMHC molecule. In this example, the H-2k CTL is specific for viral peptide A presented on an H-2k target cell (a). Antigen recognition does not occur when peptide B is displayed on an H-2k target cell (b) nor when peptide A is displayed on an H-2d target cell (c). 8536d_ch09_200-220 8/2/02 9:49 AM Page 201 mac79 Mac 79:45_BW:Goldsby et al. / Immunology 5e:
8536d_chog_200-220 8/2/02 9: 49 AM Page 202 mac79 Mac 79: 45_Bw Glasby et al. Immunology 5e 202 PART I1 Generation of B-Cell and T-Cell Responses THcell clone B cell IGURE9-2 Production and identification of a cdnA clone en- coding the T-cell receptor. The flow chart outlines the procedure used by S. Hedrick and M. Davis to obtain P]cDNA clones correspond. ing to T-cell-specific mRNAs. The technique of DNA subtractive hy. bridization enabled them to isolate[ P)cDNA unique to the T cell The labeled TH-cell cDNA clones were used as probes(inset)in Southern-blot analyses of genomic DNA from liver cells, B-lym- MRNA MRNA phoma cells, and six different TH-cell clones (a-f). Probing with CDNA clone 1 produced a distinct blot patten for each T-cell clone, 97% in free 3% in whereas probing with cDNA clone 2 did not. Assuming that liver cytoplasmic membrane-bound cells and B cells contained unrearranged germ-line TCR DNA, and polyribosomes that each of the T-cell clones contained different rearranged TCR Reverse genes, the results using cDNA clone 1 as the probe identified clone anscriptase I as the T-cell-receptor gene. The cDNA of clone 2 identified the gene for another T-cell membrane molecule encoded by DNA that 2P] cDNA d does not undergo rearrangement. Based on S. Hedrick et al, 1984, Nature 308: 149 represented the T-cell receptor, all were used as probes look for genes that rearranged in mature T cells. This ap- proach was based on the assumption that, since the ap T-cell receptor appeared to have constant and variable regions, its hydroxyapatite genes should undergo DNA rearrangements like those ob- column served in the Ig genes of B cells. The two investigators tested DNA from T cells, B cells, liver cells, and macrophages by Southern-blot analysis using the 10 [P]cDNA probes ≈ identify unique T-cell genomic DNA sequences. One clone CDNAS Hybrids with showed bands indicating DNA rearrangement in T cells but specific CDNAs common not in the other cell types. This cDNA probe identified six to t cells T-cell clones different patterns for the DNA from six different mature T- B cells cell lines(see Figure 9-2 inset, upper panel). These different em a b d patterns presumably represented rearranged TCR genes Such results would be expected if rearranged TCR genes oc 10 different cur only in mature T cells. The observation that each of the CDNA clones 三≡≡≡ six T-cell lines showed different Southern- blot patterns was Use as probes in Southern probed with cDNA clone 1 consistent with the predicted differences in TCR specificity eat The cDNA clone I identified by the Southern-blot analy ses shown in Figure 9-2 has all the hallmarks of a putative TCR gene: it represents a gene sequence that rearranges, is Probed with cDNA clone 2 expressed as a membrane-bound protein, and is expressed only in T cells. This cDNA clone was found to encode the B chain of the T-cell receptor. Later, cDNA clones were identi is unique to T cells should include the genes encoding the T fied encoding the a chain, the y chain, and finally the 8 chain cell receptor. Therefore, by hybridizing B-cell mRNA with These findings opened the way to understanding the T-cell their TH-cell [32p]cDNA, they were able to remove, or sub- receptor and made possible subsequent structural and func- onal studies tract. all the cdna that was common to b cells and t cells. The unhybridized [P]cDNA remaining after this step pre ably represented the expressed polyribosomal mRNA that was unique to the TH-cell clone, including the mRNA aB and y8 T-Cell Receptors ncoding its T-cell receptor. Structure and roles Cloning of the unhybridized [PlcDNA generated a li- brary from which 10 different cdNA clones were identified. The domain structures of aB and y8 TCR heterodimers To determine which of these T-cell-specific cDNA clones are strikingly similar to that of the immunoglobulin
represented the T-cell receptor, all were used as probes to look for genes that rearranged in mature T cells. This approach was based on the assumption that, since the T-cell receptor appeared to have constant and variable regions, its genes should undergo DNA rearrangements like those observed in the Ig genes of B cells. The two investigators tested DNA from T cells, B cells, liver cells, and macrophages by Southern-blot analysis using the 10 [32P]cDNA probes to identify unique T-cell genomic DNA sequences. One clone showed bands indicating DNA rearrangement in T cells but not in the other cell types. This cDNA probe identified six different patterns for the DNA from six different mature Tcell lines (see Figure 9-2 inset, upper panel). These different patterns presumably represented rearranged TCR genes. Such results would be expected if rearranged TCR genes occur only in mature T cells. The observation that each of the six T-cell lines showed different Southern-blot patterns was consistent with the predicted differences in TCR specificity in each T-cell line. The cDNA clone 1 identified by the Southern-blot analyses shown in Figure 9-2 has all the hallmarks of a putative TCR gene: it represents a gene sequence that rearranges, is expressed as a membrane-bound protein, and is expressed only in T cells. This cDNA clone was found to encode the chain of the T-cell receptor. Later, cDNA clones were identified encoding the chain, the chain, and finally the chain. These findings opened the way to understanding the T-cell receptor and made possible subsequent structural and functional studies. and T-Cell Receptors: Structure and Roles The domain structures of and TCR heterodimers are strikingly similar to that of the immunoglobulins; is unique to T cells should include the genes encoding the Tcell receptor. Therefore, by hybridizing B-cell mRNA with their TH-cell [32P]cDNA, they were able to remove, or subtract, all the cDNA that was common to B cells and T cells. The unhybridized [32P]cDNA remaining after this step presumably represented the expressed polyribosomal mRNA that was unique to the TH-cell clone, including the mRNA encoding its T-cell receptor. Cloning of the unhybridized [32P]cDNA generated a library from which 10 different cDNA clones were identified. To determine which of these T-cell–specific cDNA clones 202 PART II Generation of B-Cell and T-Cell Responses Hybridize mRNA 97% in free cytoplasmic polyribosomes 3% in membrane-bound polyribosomes 32P Reverse transcriptase [32P] cDNA mRNA cDNAs specific to T cells Hybrids with cDNAs common to T cells and B cells Separate on hydroxyapatite column 10 different cDNA clones Liver cells B-cell lymphoma abcdef Probed with cDNA clone 1 Probed with cDNA clone 2 T-cell clones TH-cell clone B cell Use as probes in Southern blots of genomic DNA FIGURE 9-2 Production and identification of a cDNA clone encoding the T-cell receptor. The flow chart outlines the procedure used by S. Hedrick and M. Davis to obtain [32P]cDNA clones corresponding to T-cell–specific mRNAs. The technique of DNA subtractive hybridization enabled them to isolate [32P]cDNA unique to the T cell. The labeled TH-cell cDNA clones were used as probes (inset) in Southern-blot analyses of genomic DNA from liver cells, B-lymphoma cells, and six different TH-cell clones (a–f). Probing with cDNA clone 1 produced a distinct blot pattern for each T-cell clone, whereas probing with cDNA clone 2 did not. Assuming that liver cells and B cells contained unrearranged germ-line TCR DNA, and that each of the T-cell clones contained different rearranged TCR genes, the results using cDNA clone 1 as the probe identified clone 1 as the T-cell–receptor gene. The cDNA of clone 2 identified the gene for another T-cell membrane molecule encoded by DNA that does not undergo rearrangement. [Based on S. Hedrick et al., 1984, Nature 308:149.] 8536d_ch09_200-220 8/2/02 9:49 AM Page 202 mac79 Mac 79:45_BW:Goldsby et al. / Immunology 5e:
8536d_chog_200-220 8/2/02 9: 49 AM Page 203 mac79 Mac 79: 45_Bw Glasby et al. Immunology 5e 203 thus, they are classified as members of the immunoglobulin contains a short cytoplasmic tail of 5-12 amino acids at the superfamily(see Figure 4-19). Each chain in a TCR has two carboxyl-terminal end. domains containing an intrachain disulfide bond that spans B and y8 T-cell receptors were initially difficult to inves 60-75 amino acids. The amino-terminal domain in both tigate because, like all transmembrane proteins, they are in chains exhibits marked sequence variation, but the sequences soluble. This problem was circumvented by expressing of the remainder of each chain are conserved. Thus the TCr modified forms of the protein in vitro that had been engi domains-one variable(V) and one constant(C)-are struc- neered to contain premature in-frame stop codons that pre- turally homologous to the V and C domains of immuno- clude translation of the membrane-binding sequence that globulins, and the TCR molecule resembles an Fab fragment makes the molecule insoluble (Figure 9-3). The TCR variable domains have three hy The majority of T cells in the human and the mouse ex- variable regions, which appear to be equivalent to the press T-cell receptors encoded by the ap genes. These recep- complementarity determining regions( CDRs)in immuno- tors interact with peptide antigens processed and presented globulin light and heavy chains. There is an additional area of on the surface of antigen-presenting cells. Early indications hypervariability(Hv4)in the B chain that does not normally that certain T cells reacted with nonpeptide antigens were contact antigen and therefore is not considered a CDR. ti some light was shed on the problem when In addition to the constant domain, each TCR chain con- products of the CDI family of genes were found to present tains a short connecting sequence, in which a cysteine residue carbohydrates and lipids. More recently, it has been found forms a disulfide link with the other chain of the het- that certain yo cells react with antigen that is neither erodimer. Following the connecting region is a transmem- processed nor presented in the context of a MHC molecules brane region of 21 or 22 amino acids, which anchors each Differences in the antigen-binding regions of aB and yo chain in the plasma membrane. The transmembrane do- were expected because of the different antigens they recog- mains of both chains are unusual in that they contain posi- nize, but no extreme dissimilarities were expected. However, tively charged amino acid residues. These residues enable the the recently completed three-dimensional structure for a y8 chains of the TCR heterodimer to interact with chains of the receptor that reacts with a phosphoantigen, reported by signal-transducing CD3 complex. Finally, each TCR chain Allison, Garboczi, and their coworkers, reveals significant L chains a-chain阝 chain H chain H chain sequence Transmembrane region Tm) Cytoplasmic COOH tail (CT) (282) (248 FIGURE 9.3 Schematic diagram illustrating the structural similar- IgM H chains are connected to one another by a disulfide bond in the ells. The TCR a and B chain each contains two domains with the im- chains by disulfide links between the C termini of the L chains and hoglobulin-fold structure The amino-terminal domains (Va and the Cu region. TCR molecules interact with CD3 via positively VB)exhibit sequence variation and contain three hypervariable re- charged amino acid residues(indicated by +) in their transmem- gions equivalent to the CDRs in antibodies. The sequence of the con. brane regions. Numbers indicate the length of the chains in the TCR stant domains(Ca and CB) does not vary. The two TCr chains are molecule. Unlike the antibody molecule, which is bivalent, the TCR is connected by a disulfide bond between their constant sequences; the monovalent
thus, they are classified as members of the immunoglobulin superfamily (see Figure 4-19). Each chain in a TCR has two domains containing an intrachain disulfide bond that spans 60–75 amino acids. The amino-terminal domain in both chains exhibits marked sequence variation, but the sequences of the remainder of each chain are conserved. Thus the TCR domains–one variable (V) and one constant (C)–are structurally homologous to the V and C domains of immunoglobulins, and the TCR molecule resembles an Fab fragment (Figure 9-3). The TCR variable domains have three hypervariable regions, which appear to be equivalent to the complementarity determining regions (CDRs) in immunoglobulin light and heavy chains. There is an additional area of hypervariability (HV4) in the chain that does not normally contact antigen and therefore is not considered a CDR. In addition to the constant domain, each TCR chain contains a short connecting sequence, in which a cysteine residue forms a disulfide link with the other chain of the heterodimer. Following the connecting region is a transmembrane region of 21 or 22 amino acids, which anchors each chain in the plasma membrane. The transmembrane domains of both chains are unusual in that they contain positively charged amino acid residues. These residues enable the chains of the TCR heterodimer to interact with chains of the signal-transducing CD3 complex. Finally, each TCR chain T-Cell Receptor CHAPTER 9 203 contains a short cytoplasmic tail of 5–12 amino acids at the carboxyl-terminal end. and T-cell receptors were initially difficult to investigate because, like all transmembrane proteins, they are insoluble. This problem was circumvented by expressing modified forms of the protein in vitro that had been engineered to contain premature in-frame stop codons that preclude translation of the membrane-binding sequence that makes the molecule insoluble. The majority of T cells in the human and the mouse express T-cell receptors encoded by the genes. These receptors interact with peptide antigens processed and presented on the surface of antigen-presenting cells. Early indications that certain T cells reacted with nonpeptide antigens were puzzling until some light was shed on the problem when products of the CD1 family of genes were found to present carbohydrates and lipids. More recently, it has been found that certain cells react with antigen that is neither processed nor presented in the context of a MHC molecules. Differences in the antigen-binding regions of and were expected because of the different antigens they recognize, but no extreme dissimilarities were expected. However, the recently completed three-dimensional structure for a receptor that reacts with a phosphoantigen, reported by Allison, Garboczi, and their coworkers, reveals significant NH2 NH2 Vα VL VL CL CL VH VH Cµ Cµ α-chain β-chain S S S S S S S S S S NH2 NH2 + + + αβ T-cell receptor B-cell mIgM L chains H chain H chain Connecting sequence Transmembrane region (Tm) Cytoplasmic tail (CT) Cµ Cµ Vβ Cµ Cµ Cα Cβ S S S S S S S S S S Cµ Cµ S S S S S S S S S S S S S S S S S S S S S S S S COOH (248) COOH (282) FIGURE 9-3 Schematic diagram illustrating the structural similarity between the T-cell receptor and membrane-bound IgM on B cells. The TCR and chain each contains two domains with the immunoglobulin-fold structure. The amino-terminal domains (V and V) exhibit sequence variation and contain three hypervariable regions equivalent to the CDRs in antibodies. The sequence of the constant domains (C and C) does not vary. The two TCR chains are connected by a disulfide bond between their constant sequences; the IgM H chains are connected to one another by a disulfide bond in the hinge region of the H chain, and the L chains are connected to the H chains by disulfide links between the C termini of the L chains and the C region. TCR molecules interact with CD3 via positively charged amino acid residues (indicated by ) in their transmembrane regions. Numbers indicate the length of the chains in the TCR molecule. Unlike the antibody molecule, which is bivalent, the TCR is monovalent. 8536d_ch09_200-220 8/2/02 9:49 AM Page 203 mac79 Mac 79:45_BW:Goldsby et al. / Immunology 5e:
8536d_ ch09 200-220 8/2/02 1:00 PM Page 204 mac79 Mac 79: 45 Bw: ldsby et al. Immunology 5e 204 PART II Generation of B-Cell and T-Cell Responses differences in the overall structures of the two receptor types, TABLE 9.1 Comparison of ap and Yo T cells pointing to possible functional variation. The receptor they studied was composed of the y9 and 82 chains, which are Feat cB T cells 8 Tcells those most frequently expressed in human peripheral blood. a deep cleft on the surface of the molecule accommodates Proportion of CD3 cific. This antigen is recognized without MHC presentation. TO TCR V gene germ- Small The most striking feature of the structure is how it differs om the aB receptor in the orientation of its V and C re- CD4/CD8 gions. The so-called elbow angle between the long axes of the phenoxy V and C regions of y8 TCR is 111 in the aB TCR, the elbow 60% angle is 149, giving the molecules distinct shapes(Figure 9-4). The full significance of this difference is not known, but it could contribute to differences in signaling mecha- CD4*CD8+ nisms and in how the molecules interact with coreceptor CD4 CD8 60% molecules MHC restriction CD4. MHC MHc The number of y8 cells in circulation is small compared class Il restriction with cells that have ap receptors, and the V gene segments of y8 receptors exhibit limited diversity. As seen from the data CD8+ MHC in Table 9-1, the majority of y8 cells are negative for both CD4 and CD8, and most express a single y8-chain subtype. Ligands Peptide+ MHC Phospholipid In humans the predominant receptor expressed on circulat- ing y8 cells recognizes a microbial phospholipid antigen, 3-SOURCE:DKabelitz et al,1999,Springer Seminars formyl-1-butyl pyrophosphate, found on M. tuberculosis and 2155,p. 36 other bacteria and parasites. This specificity for frequently encountered pathogens led to speculation that y8 cells may function as an arm of the innate immune response, allowing rapid reactivity to certain antigens without the need for a processing step. Interestingly, the specificity of circulating Y8 ffinity would be selectively activated and amplified to deal cells in the mouse and of other species studied does not par- with the pathogen. allel that of humans, suggesting that the y8 response may be directed against pathogens commonly encountered by a given species. Furthermore, data indicating that y8 cells can secrete a spectrum of cytokines suggest that they may play a Organization and rearrangement regulatory role in recruiting aB T cells to the site of invasion by pathogens. The recruited ap T cells would presumably of TCR Genes display a broad spectrum of receptors; those with the highest The genes that encode the aB and y8 T-cell receptors are ex- pressed only in cells of the T-cell lineage. The four TCR loci (a, B, Y, and 8) ed in the germ lin hat is remarkably similar to the multigene organization of the immunoglobulin(Ig) genes(Figure 9-5). As in the case of Ig genes, functional TCR genes are produced by re- arrangements of V and J segments in the a-chain and y- V domains chain families and V, D, and J segments in the B-chain and 8-chain families. In the mouse, the a-B, and y-chain gene segments are located on chromosomes 14, 6, and 13, respec 公N tively. The d-gene segments are located on chromosome 14 A Domains between the Va and Ja segments. The location of the8-chain gene family is significant: a productive rearrangement of the chain gene segments deletes Ca, so that, in a given T cell, y8 TCR aβTCR the aB TCR receptor cannot be coexpressed with the y8 Mouse germ-line DNA contains about 100 Va and 50 Ja FIGURE9. Comparison of the y8 TCR and ap TCR. The dif- gene segments and a single Ca segment. The 8-chain gene ference in the elbow angle is highlighted with black lines. [From family contains about 10 V gene segments, which are largely T. Allison et al, 2001, Nature 411: 820. 1 distinct from the Va gene segments, although some sharing
differences in the overall structures of the two receptor types, pointing to possible functional variation. The receptor they studied was composed of the 9 and 2 chains, which are those most frequently expressed in human peripheral blood. A deep cleft on the surface of the molecule accommodates the microbial phospholipid for which the receptor is specific. This antigen is recognized without MHC presentation. The most striking feature of the structure is how it differs from the receptor in the orientation of its V and C regions. The so-called elbow angle between the long axes of the V and C regions of TCR is 111°; in the TCR, the elbow angle is 149°, giving the molecules distinct shapes (Figure 9-4). The full significance of this difference is not known, but it could contribute to differences in signaling mechanisms and in how the molecules interact with coreceptor molecules. The number of cells in circulation is small compared with cells that have receptors, and the V gene segments of receptors exhibit limited diversity. As seen from the data in Table 9-1, the majority of cells are negative for both CD4 and CD8, and most express a single -chain subtype. In humans the predominant receptor expressed on circulating cells recognizes a microbial phospholipid antigen, 3- formyl-1-butyl pyrophosphate, found on M. tuberculosis and other bacteria and parasites. This specificity for frequently encountered pathogens led to speculation that cells may function as an arm of the innate immune response, allowing rapid reactivity to certain antigens without the need for a processing step. Interestingly, the specificity of circulating cells in the mouse and of other species studied does not parallel that of humans, suggesting that the response may be directed against pathogens commonly encountered by a given species. Furthermore, data indicating that cells can secrete a spectrum of cytokines suggest that they may play a regulatory role in recruiting T cells to the site of invasion by pathogens. The recruited T cells would presumably display a broad spectrum of receptors; those with the highest affinity would be selectively activated and amplified to deal with the pathogen. Organization and Rearrangement of TCR Genes The genes that encode the and T-cell receptors are expressed only in cells of the T-cell lineage. The four TCR loci (,, , and ) are organized in the germ line in a manner that is remarkably similar to the multigene organization of the immunoglobulin (Ig) genes (Figure 9-5). As in the case of Ig genes, functional TCR genes are produced by rearrangements of V and J segments in the -chain and - chain families and V, D, and J segments in the -chain and -chain families. In the mouse, the -, -, and -chain gene segments are located on chromosomes 14, 6, and 13, respectively. The -gene segments are located on chromosome 14 between the V and J segments. The location of the -chain gene family is significant: a productive rearrangement of the -chain gene segments deletes C, so that, in a given T cell, the TCR receptor cannot be coexpressed with the receptor. Mouse germ-line DNA contains about 100 V and 50 J gene segments and a single C segment. The -chain gene family contains about 10 V gene segments, which are largely distinct from the V gene segments, although some sharing 204 PART II Generation of B-Cell and T-Cell Responses V domains C domains TCR 111 TCR 147 FIGURE 9-4 Comparison of the TCR and TCR. The difference in the elbow angle is highlighted with black lines. [From T. Allison et al., 2001, Nature 411: 820.] TABLE 9-1 Comparison of and T cells Feature T cells T cells Proportion of CD3 90–99% 1–10% cells TCR V gene germ- Large Small line repertoire CD4/CD8 phenotype CD4 ~60% 1% CD8 ~30% ~30% CD4CD8 1% 1% CD4– CD8– 1% ~60% MHC restriction CD4: MHC No MHC class II restriction CD8: MHC class I Ligands Peptide MHC Phospholipid antigen SOURCE: D. Kabelitz et al., 1999, Springer Seminars in Immunopathology 21:55, p. 36. 8536d_ch09_200-220 8/2/02 1:00 PM Page 204 mac79 Mac 79:45_BW:Goldsby et al. / Immunology 5e: