| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Review |
From the Max-Planck Institute of Neurobiology,* Martinsried; the Institute for Clinical Neuroimmunology, Klinikum Großhadern, Ludwig Maximilians University, Munich; Max-Planck-Institute for Immunobiology, Freiburg, Germany
Abstract
Histological samples of autopsy or biopsy tissue provide the best available evidence that autoreactive T cells are involved in the immunopathogenesis of many autoimmune diseases. However, morphology alone does not provide information on the antigen-specific T-cell receptor (TCR) of these cells, let alone on their antigen specificity. In this review article we discuss a number of emerging possibilities for identifying TCR sequences directly from biopsy tissue. We also review the methods for expressing presumably autoreactive TCR molecules and speculate on how the expressed TCR might be used to identify target antigens. Such information should eventually provide new insights into disease pathogenesis which lead to better therapies.
. Then CD8+ T cells contact muscle fibers and orient their perforin-containing lytic granules vectorially toward the membrane of the target muscle fiber (Figure 1b)
. This "in flagranti" situation, morphologically speaking, strongly suggests that by recognizing an antigen on the target muscle fiber the T cell is activated to behave as a cytotoxic effector cell. A careful analysis of their antigen receptor repertoire provides further evidence for the pathogenic role of such T cells. To this end, a number of new techniques have recently become available, eg, microdissection of single infiltrating T cells, combined with single-cell polymerase chain reaction (PCR) amplification of their clonotypic, rearranged T-cell receptor (TCR) Vß chain.2,3
|
- or ß-chains over-represented in the T-cell receptor repertoire? Which antigen(s) do the infiltrating T cells recognize? How is antigen-recognition related to the pathogenesis and to the course of the disease? Can the recognition process be manipulated to influence disease progression? In this article we review these state-of-the-art technologies for identifying and characterizing such infiltrating T cells and their antigen-specific receptors in tissue biopsy specimens.
Figure 2
shows the overall strategy: to morphologically identify putatively pathogenic T cells in the target tissue, to molecularly clone and express both chains of their antigen-specific TCR, and then to try to identify the cognate antigen(s). None of these steps is trivial, mainly because of the high variability of TCR molecules. They are membrane-anchored ß- or 
-heterodimeric proteins of antibody-like structure. Each chain is composed of variable (V), joining (J), and constant (C) regions. The V and J regions are connected by the small diversity (D) region and/or random nucleotides (N). The genomic organization and the respective numbers of genetic elements of the TCR V and J regions are delineated in Figure 3a
and Table 1
. Each chain of the resulting receptor protein has one hypervariable and two variable loops that protrude from a compact ß-sheet structure.6-8
These loops are termed "complementarity determining regions" (CDR), because they directly contact the antigenic peptide. The high degree of variability (the total diversity of ß TCRs may exceed 1017
5
) increases the difficulties to determine the TCR sequences of the infiltrating cells, because the variety of very similar, but not identical proteins and nucleotide sequences is immense.
|
|
|
Availability and Choice of Tissue
The availability of biopsy tissue depends on the disease. In multiple sclerosis (MS), for example, a biopsy is made only in exceptional cases. Relatively few studies have directly addressed the T-cell repertoire expressed in MS lesions.9,10 Many investigations require fresh frozen rather than fixed tissue. In a recent study, the TCR ß-chain sequences of single tissue-infiltrating T cells were identified in frozen tissue of two MS patients, using a combination of microdissection and single-cell PCR.2
In contrast, tissue biopsy is routinely taken for diagnostic reasons in other diseases, eg, polymyositis. Polymyositis has the additional advantage that muscle is easily accessible, in contrast to CNS. This is probably one of the reasons why there have been numerous studies of the T-cell repertoire expressed in polymyositic muscle.11,12
A monoclonal expansion of T cells was found in a unique case of polymyositis.13,14
In other cases, striking expansions of muscle-infiltrating ß TCR+CD8+ T cells were observed.15-18
The infiltrating T cellsmade direct contact with the muscle fiber targets and secreted lytic granules toward the attacked muscle fibers,19
suggesting that the T cells recognize an antigen on the muscle fiber surface. These examples illustrate that the target tissue itself, rather than peripheral blood or cerebrospinal fluid, is the best cell source for this type of study.
Morphological Identification of Autoaggressive T Cells
The quality of the biopsy material is crucial for all types of investigations. Consequently, the appropriate preservation of the tissue is an essential prerequisite. Ideally the tissue should be frozen immediately after excision to prevent any degradation of RNA, DNA, or proteins. Most traditional fixation procedures, however, modify proteins and nucleic acids, thus hindering any further analysis.
One of the first steps in identifying T cells and T-cell receptors in tissue is to apply immunohistochemistry, using monoclonal antibodies against differentiation markers like CD3, CD4, CD8, and antibodies against TCR V- and V-ß family members. Staining with the latter antibodies may provide the first indications of (oligo)clonal expansions, but this method has obvious limitations. First, although the panel of available anti-V-ß monoclonal antibodies is quite large, currently there are only three antibodies against TCR V-, namely anti-V2,20
anti-V12,21
and V24.22
Note that V24 is not only expressed on "regular" CD4+ and CD8+ T cells, but also on human NKT cells (reviewed by23,24
). Second, not all of the available anti-TCR V-chain family are suited for immunocytochemistry. Third, most antibodies specific for a particular V family do not distinguish between the different subfamilies, which may differ by some amino acids, whereas other antibodies may recognize only one particular allele or subfamily, but not the whole family. Fourth, V-region specific antibodies do not distinguish between different clonotypes (that is, different hypervariable CDR3 regions). These limitations indicate that immunohistochemical methods may provide the first clues of clonal expansions, and may help to identify the expanded TCR V-chain family, but they are not suitable for identifying TCR clonotypes.
If expanded T-cell clones are observed in tissue samples, they are usually superimposed on a background of unrelated cells. These may include several subpopulations: 1) bystander T cells attracted non-specifically by inflammatory cytokines or chemokines, 2) intravascular T cells from blood capillaries, 3) irrelevant clonal expansions as they may occur in healthy, especially older persons,25-28 and 4) expanded T-cell clones in patients with superimposed infections. The first two groups of bystander cells and T cells from capillaries typically have random distributions of T-cell receptors and therefore provide a continuous background, whereas the latter two groups may include expanded clones with preferred TCR expression. They thus may mimic the presence of expanded autoimmune T cells. Further confusion may arise from the fact that in infectious disease, the clonally expanded T cells may be activated, as are some of the autoreactive T cells in autoimmune disease. It is therefore important to distinguish the truly autoaggressive T cells from the different types of expanded, activated, and bystander cells. This is not usually achieved by staining TCR with soluble MHC/peptide complexes, because this would require previous knowledge of the relevant antigens.29 Important distinguishing criteria can be provided by the morphology and location of the inflammatory cells in the tissue. In polymyositis, for example, autoaggressive CD8+ T cells make direct contact and invade the target muscle fibers.30 In this case it is relatively simple to identify the potentially autoaggressive T cells and distinguish them from unrelated bystander cells. In other cases, however, the topographical situation may be less informative. For example, CD4+ T cells may act as "helper" cells by secreting soluble mediators in the local milieu of the inflammatory lesion. This function does not necessarily depend on direct cell-cell contact. Further, the antigen-specific priming of CD4+ T cells may occur at other sites outside the lesion.
Characterization of Expanded TCR Chains from Tissue Samples
Several PCR-based techniques can be used to determine the DNA sequences of expanded T-cell clones in homogenized tissue samples. The basic principles of anchored, inverse, and family PCR, and of TCR-CDR3 spectratyping are outlined in Figure 3b
. Some advantages and limitations of the different techniques are compared in Table 2
. All techniques use cDNA as templates. Anchored and inverse PCR use PCR primers that hybridize to shared TCR chain sequences, irrespective of the V or J family usage. They therefore allow an essentially unbiased expansion of all TCR V-chain families. For anchored PCR (also termed rapid amplification of cDNA ends),31,32
a synthetic oligo-C sequence is first added to the N terminus of the cDNA. Then an oligo-G-containing forward-PCR primer is used for amplification. The corresponding reverse-primer hybridizes to the C region. For inverse PCR,33
the double-stranded cDNA is converted into a closed circle, and the unknown regions are amplified using a primer pair within the C region. The third method, TCR V-family PCR,34
uses a pool of many different primers that hybridize to the V regions. Therefore, a specific primer must be included for each V family. Only the reverse primer (which hybridizes to the C region) is the same for all reactions. If the chosen experimental conditions are optimal, this method can amplify minute amounts of cDNA from particular TCR sequences present in the sample. However, one disadvantage of this method is that not all V sequences may be amplified equally well, depending on the different efficiencies of the various primers. It therefore might be problematic to directly compare efficiencies between different V families, unless the primer efficiencies have been quantified in earlier experiments and are normalized. It is essential with all techniques, whether anchor, inverse, or family PCR, to clone the PCR products and to sequence a statistically significant number of clones.
|
-chains, due to the considerably large number of J- genes.
In summary, immunocytochemistry is indispensable for identifying potentially autoaggressive T cells in the target tissue, in particular when cytotoxic effector T cells are analyzed. Despite the limitations of anti-V antibody staining, it still gives a quick and broad overview by assigning V-chain family usage to morphologically identifiable cells. This aspect is particularly important because it helps decide whether in situ expanded T-cell clones are indeed attacking the target tissue, or whether they are bystanders. To identify the complete TCR sequences of the expanded clones, one of the PCR-based techniques must be used. All methods are highly sensitive, particularly if nested primer pairs are used. Anchored and inverse PCR have the advantage of yielding full-length products and virtually unbiased product distributions, but inverse PCR may require relatively high amounts of template. Moreover, the products have to be cloned into plasmids, and a statistically significant number of individual Escherichia coli clones have to be sequenced. However, the products of TCR family PCR should also be cloned before sequencing, because direct sequencing of the PCR products might reveal only the most prominent CDR3 sequence. The most straightforward indication for clonal expansion in tissue can be obtained by CDR3 spectratyping, because this technique may directly detect clonal expansions not only of particular V, but also of the corresponding CDR3 regions. For final proof of clonal expansion, the products must also be sequenced. If all primers to all V and J families are included, a comprehensive overview may be obtained. However, as mentioned above, the broad variability of J- elements still impedes investigation of the TCR -chains and hence a complete analysis.
Identification of Paired TCR Chains Expressed in Individual T Cells
In most studies reported so far, the T-cell expansions observed in autoimmune tissues were oligoclonal. Only a single case of a monoclonal expansion has been reported.14
This was a very special case, because the expanded T cells expressed a rather than ß TCR. Hence, analysis of a tissue homogenate containing oligoclonally expanded T cells with one of the PCR techniques described above would yield several - and ß-chain sequences, but it would be impossible to know which chain pairs to which ß chain. As long as this pairing is not known, however, the structure of the functional pathogenic receptor will not be known. In principle, single-cell PCR can overcome this fundamental problem. For this approach, interesting cells are identified in the biopsy sample by staining with appropriate antibodies. Next, single T cells of interest are excised directly from tissue sections under a microscope using a micromanipulator or a highly focused laser beam (laser-assisted microdissection),36,37
and the isolated cells are transferred into a PCR tube where the TCR genes are amplified by single-cell PCR (Figure 3c)
.
If the - and ß-chain sequences of the expressed TCR are not known, pools of many different V- and J-specific PCR primers, or degenerate primers, must be used. Both types of primers are of course less efficient than defined pairs of specific primers. Nevertheless, unbiased amplification of the TCR ß-chain genes from single cells has been successful.2,3,38-41
Because RNA might be unstable in a biopsy sample, the authors used rearranged genomic DNA as a template; this is much more resistant to degradation. Obvious disadvantages are that only a single copy of template is present per cell, and the conserved region is separated from the rearranged V(D)J region by a large intron (Figure 3a)
. Therefore, it was not feasible to use a single reverse-PCR primer (hybridizing to the C region), as would be used for RT-PCR. Instead, the authors used one pool of 7 reverse J primers covering all 13 different J-ß genes (Table 1)
together with 8 forward V-primer pools, each containing 2 to 5 individual primers for hybridizing to the V-ß regions. With this approach, unknown TCR ß-chains could be amplified from single cells excised from tissue sections. In up to 30% of all single cells examined a rearranged TCR ß-chain was detected.
Amplification of the TCR -chains is even more challenging. Although the number of different V elements is comparable to that of the ß chain, the -chain repertoire may select from a greater number of different J elements (50 functional J genes as compared to 13 for Jß; Table 1
). Thus, the number of primers has to be significantly increased if all possible V and J elements are to be covered. One way to restrict this variability is to stain the putative target T cells with one of the available three anti-V antibodies. After microdissection of the corresponding cell, the V chain can then be amplified using only one specific V primer instead of a pool covering all possible V sequences. This approach was successfully used to isolate a pair of and ß chains from a freshly prepared, FACS-sorted single cell.42
These authors used cDNA as the PCR template, which allowed them to ignore any variability of the J regions and simply employ reverse primers that hybridized to the C region. However, this approach might fail when frozen sections of biopsy samples (rather than fresh cells) are investigated, because of the instability of mRNA as compared to that of DNA.
To define pairs of TCR - and ß-chains we recently stained TCR in biopsy sections of a polymyositis patient with an anti-V12 antibody, then isolated stained cells by laser microdissection, and amplified the TCR - and ß-chains from these V12+ single cells (C.K. Schneider et al, unpublished results). We successfully amplified pairs of and ß chains, thus demonstrating the usefulness of this approach in case the chain can be stained. An additional advantage of V staining is that it shows that the particular V chain is expressed on the cell surface. This may be of interest, because two chains (or rarely, two ß chains) are synthesized in a significant number of T cells,22,43-45
and it is not known which chain is expressed on the cell surface.
"Reconstruction" of TCR cDNA
Once a pair of co-expressed - and ß-TCR chains is identified, a functional heterodimer has to be reconstructed. Only amplification by anchored and inverse PCR yields products that cover the full-length sequences of TCR-DNA. The primer pairs used for family PCR, CDR3 spectratyping, or single-cell PCR mostly yield relatively short fragments because they are designed for high specificity and sensitivity. Therefore, they hybridize to sequences somewhere in the middle of the V regions, in the J regions, or at the beginning of the C regions. Parts of the V and C regions will be missing in such PCR products, and some products of single cell PCR reactions may not even contain complete J sequences. To express the TCR proteins, it is therefore necessary to reconstruct the complete sequences from fragments. Such a reconstruction from single-cell PCR is outlined schematically in Figure 3d
. A typical single-cell PCR fragment ranging from the middle of the V to the middle of the J regions (designated "b") is ligated to a fragment covering the remainder of the J and the beginning of the C region ("c"), and inserted into a cloning plasmid that already contains the beginning of the V regions ("a") and the rest of the C regions ("d"). Fragments "a" and "d" may be obtained from the cDNA of a unrelated TCR with identical V and C regions. Alternatively, the hypervariable regions of an unrelated TCR with identical V and C regions may be replaced by site-directed mutagenesis. The precise strategy depends on the number and lengths of the fragments, restriction sites, and other features.
Expression of Recombinant TCR Molecules
There are basically three strategies for producing the reconstructed autoreactive TCR molecules in a recombinant expression system. First, they can be designed as soluble proteins, ie, as small monomeric (glyco-)proteins, analogous to Fv- or Fab-antibody fragments. These monomeric TCR proteins may carry tags that can be used for detection, oligomerization, or coupling to solid matrices. Second, the TCR can be expressed at the surface of immortalized T cell lines. Here the recombinant TCR is integrated into the CD3 complex of the recipient cells and coupled to their signaling machinery. Third, the TCR can be expressed in vivo in a transgenic animal model.
Soluble TCR Molecules
The most straightforward approach is to express the TCR of interest as a soluble molecule. Different ß- or -TCR molecules of murine or human origin have been generated using a broad spectrum of expression protocols. Eukaryotic expression systems often yield correctly folded and processed proteins, but their amount is relatively small. For example, TCR molecules were expressed in myeloma, COS, Drosophila cells, and Baculovirus-transfected insect cells.46-51
In the oxidizing milieu of the periplasm of E. coli, folding and formation of disulfide bonds may also be achieved, but again with comparably low yields.52
In contrast, expression in the cytosol of bacteria may yield much larger amounts of protein, but the TCR chains have to be refolded in vitro from insoluble inclusion bodies.53-57
Considerably different constructs have been used. The TCR chains were either co-expressed to directly form heterodimers, expressed as single-chain constructs with the variable domains linked covalently by peptide linkers, or they were expressed separately and later folded and assembled in vitro. In all constructs the membrane-spanning regions were truncated and replaced by cleavable lipid linkers46
or peptide tags, such as oligo-histidine residues, which facilitate purification or detection.51,56
To support dimerization of the chains leucine-zipper domains were introduced,50
or the TCRs were fused to other large proteins, eg, antibody domains47,48
or the CD3
chain.49
These different expression strategies have been used successfully to produce soluble TCR molecules and even protein crystals, allowing structural analysis of ß TCRa (reviewed by6-8
).
In principle, soluble TCR molecules could be used to identify the corresponding antigen. The main obstacle to this approach, however, is the low affinity of TCR-ligand interactions. Many studies have addressed the kinetics and thermodynamics of TCR-ligand binding, using soluble TCR molecules or MHC/peptide complexes (reviewed by7,58 ). The results show that the interactions of the monomeric compounds are generally of low affinity and that antigen presentation is dominated by the high avidity gained by oligomerization and by the kinetics of the interaction. Therefore, the classic methods for detecting and analyzing molecular interactions such as affinity chromatography or library screening cannot be applied. These methods require high binding affinities, because they imply washing steps during which the TCR-ligand complexes would dissociate and be lost.
TCR Expression on the Surface of Lymphoid Cell Lines
Expression of a TCR on the surface of a living cell has several advantages to soluble TCR. First, the TCR molecules are oligomerized on a cell surface and may interact with antigens that are oligomerized themselves, for example, bound to a microtiter well or to the surface of an antigen-presenting cell. Therefore, the low affinity of the TCR may be overcome by the high avidity of the oligomerized molecules. Second, immortalized cells can be grown essentially without limitations. Third, and most importantly, TCR activation triggers the production of cytokines, which may be measured by ELISA. Each of these steps amplifies the initial signals. Therefore the overall sensitivity is drastically increased.
The recombinant expression of TCR molecules on the surface of immortalized cells was first described for a TCR ß-chain in a recipient cell that lacked the endogenous ß-chain.59
The recombinant ß-chain and the endogenous -chain were co-expressed as heterodimers. In later experiments two recombinant - and ß-chains were expressed so that the recipient cells expressed both the recombinant and the endogenous receptors.60,61
Besides other cell lines, such as TG4062
or DS23,63
the commonly used recipient cell lines are the human T-lymphoma Jurkat and its TCR - or ß-chain deletion mutants 18.B3,64
J79,65
RT3-T5.3,59
JBN,61
31–13,66
and the murine T-hybridoma line 58-ß-, which lacks endogenous TCR chains.67
58-ß- is derived from DO-11.10, which has been generated by fusing chicken-ovalbumin specific T-cell blasts from BALB/C mice to the AKR thymoma BW5147.68
58-ß- was later transfected with human CD4 to facilitate the recognition of human HLA class II molecules, and with the mouse CD3 chain to improve signal transduction.69
To improve the recognition of murine class-I MHC molecules, 58-ß- was also co-transfected with murine CD8 - and ß-chains.70
Different plasmids and retroviruses were successfully used to transfect immortalized cell lines like Jurkat and 58-ß-. Recently, TCR molecules were expressed in spleenocytes, peripheral blood lymphocytes, and T-cell lines (reviewed by71
).
TCR expression can be monitored by FACS analysis using V-region specific antibodies. If appropriate antibodies are unavailable, anti-CD3 antibodies may be used, provided that the recipient mutant cell line lacks the endogenous pendant of the TCR chain to be transfected. This indirect detection strategy exploits the fact that TCR-CD3 complexes can only reach the cell surface if all components are expressed concomitantly. Recognition of the TCR antigen is typically assessed by measuring secreted IL-2. Although this assay might not be as sensitive as a cytotoxicity assay, Jurkat and 58-ß- cells grow fast and may be expanded to almost unlimited cell numbers. This is a great advantage to using "real" T cells that are notoriously difficult to grow and can rarely be expanded to more than a few million cells. Further, transfectants can be easily cloned, again in contrast to "real" T cells. It is always advisable to pick single clones during the selection of stable transfectants, because the number of transfected plasmids and the position of vector insertion into the genome of the recipient cell may vary considerably. Analysis of individual clones allows the selection of those with the highest TCR expression levels. Moreover, site-directed mutations may be easily introduced into transfectants by standard techniques. Mutations in the CDR regions help exclude superantigen-like or non-specific activation and thus prove specificity of antigen recognition.72
Many investigators have chosen Jurkat or its derivatives as recipient cells for transfection of human TCR chains, because Jurkat is of human origin and therefore all molecules of the CD3 complex fit to the transfected TCR chains. Further, Jurkat is one of the best-characterized mammalian cell lines. One disadvantage is that there are no sublines lacking both TCR chains. Therefore, transfected Jurkat cells may express not only their own endogenous TCR and the transfected TCR heterodimers, but hybrids of transfected and endogenous chains might also reach the surface. Such transfectants are hardly suited for investigating unknown antigens, because it would be difficult to distinguish the signals of the transfected, parental, and hybrid TCR molecules. In contrast, 58-ß- cells and their derivatives lack both endogenous TCR chains67
so that no hybrid receptors are formed. Furthermore, even if there were no antibodies to any of the transfected TCR chains available, surface expression might be detected by using anti-CD3 antibodies. It might be a disadvantage that 58-ß- is a murine cell line, but it has been shown that human-mouse chimeric TCR molecules consisting of human Vn(D)nJ regions and murine C regions69,73-75
as well as full-length human ß- or -TCR molecules may be expressed in the context of murine CD3 molecules.62,72
The specificity of the human TCR is preserved in the context of murine CD3.
Recombinant TCR Molecules in Transgenic Animal Models
An obvious way to validate the destructive potential of a presumably autoaggressive human TCR in vivo is to use transgenic techniques to insert the TCR into living animals. However, if the target antigen is not known, it is also not known whether the transgenic mice express it. Several TCR transgenic mice that are useful models for human autoimmune diseases have been described so far (reviewed by76-79 ). In these cases, the antigen(s) recognized by the transgenic TCR were known in advance. Many of these animals expressed TCR molecules of murine origin, but mice with appropriate genetic backgrounds for investigating whether symptoms similar to the human disease might be induced also expressed human autoreactive TCR molecules. These animals expressed the antigen either endogenously, were crossed with mice transgenic for the antigen, or were challenged with the antigen. So far no animal model has been described in which a pathogenic TCR was first detected in a patient and later expressed in a transgenic animal without previous knowledge of the target antigen. TCR-transgenic animals are more useful for studying the pathogenesis and therapy of autoimmune diseases than for identifying unknown antigens.
Using TCRs to Identify Unknown Target Antigens
Both ß- and -T cells have been observed in tissue lesions of patients with autoimmune diseases. The mechanism of antigen recognition of these two types of T cells is quite different. CD8+ ß T cells recognize class I MHC molecules complexed with peptides of 8 to 10 amino acids in length, and CD4+ ß T cells recognize class II MHCs with much longer peptides (reviewed in7,8
). In contrast, TCRs have recognition patterns similar to antibodies; they may recognize a range of epitopes, including conformational epitopes of proteins, or carbohydrates, lipids or low molecular weight compounds alone or bound to MHC or MHC-like proteins such as CD1. This recognition is independent of the CD4 or CD8 coreceptors (reviewed in80-82
). In vivo, additional costimulatory receptors and accessory molecules are involved.83
The methods for identifying ß- and -T-cell antigens will be discussed separately.
ß-TCR Molecules
To identify the antigens of ß-TCR molecules, two different components, the antigenic peptide and the HLA-molecule, must be determined. Each individual expresses several relevant polymorphic HLA molecules; it is not known initially which HLA allele presents the relevant antigenic peptide. Immunohistochemistry may reveal if the putatively pathogenic T cells in the biopsy tissue express CD4 or CD8 molecules, and may thus point to MHC class II or class I restriction. The restriction molecule may be further narrowed down by using genetic and epidemiological data, which may indicate an association of the autoimmune disease with a certain HLA type. Although none of these links are strict, they may provide a clue. If the HLA restriction molecule is not known, autologous EBV-transformed B cells, isolated from the same donor as the TCR, are probably the best source of antigen-presenting cells, because they carry all of the relevant HLA molecules of the patient. To formally determine the specific HLA restriction molecule, human typing cell lines or HLA-transfected mouse fibroblasts may be used as antigen-presenting cells. It should be noted that TCR-transfected human cells (like Jurkat) and mouse cells (like 58-ß- cells) may be activated specifically by human and mouse antigen-presenting cells, provided that the correct MHC/peptide complex is expressed. However, we found that higher IL-2 levels may be obtained if 58-ß- cells are stimulated by HLA-transfected mouse fibroblasts, presumably because the mouse fibroblasts express the appropriate murine costimulation molecules.
The determination of peptide specificities of TCR molecules has a long tradition in molecular immunology. However, most of these experiments were conducted with previously known antigen specificities, ie, by using peptide derivatives of known parent peptides. If an interesting TCR is cloned from a biopsy sample, the situation is different in that the investigator initially does not have any clue about the antigenic peptide. If the HLA restriction is known (which is usually not the case), certain hints may be obtained from the known peptide-binding motifs of the restricting HLA molecule. A list of HLA peptide-binding motifs may be found under http://syfpeithi.bmi-heidelberg.com. On the basis of known anchor positions, a tailor-made peptide library may be constructed in which only amino acids suspected to face the TCR are permutated, but the anchor positions are fixed. Another strategy is to use random libraries containing fixed amino acids facing the TCR. This latter approach revealed the recognition motif of a human CD8+ T-cell line of known fine-specificity.84 Different types of peptide libraries have been described, mostly chemically synthesized pools of peptides but also phage libraries (reviewed in85-87 ). The basic dilemma of all peptide libraries is that the more comprehensive they are (ie, the more peptides they include), the smaller the concentrations of individual peptides, which finally may be too diluted to even be detected.
-TCR Molecules
-TCR molecules do not depend on antigens presented by MHC molecules, but they resemble antibodies in their antigen-recognition properties (reviewed in84-86
). While the methods for identifying potentially pathogenic -T cells in tissue and for analyzing their TCR are essentially the same as for ß-T cells, the methods for identifying target antigens are different. One possibility is to construct a soluble TCR that might be used for affinity chromatography or for screening protein expression libraries. However, both methods depend on high affinities of the receptor ligand pairs. Up to now the affinity of only one -TCR/antigen pair has been measured.88
The reported dissociation constant of 0.1 µmol/L corresponds to a relatively high affinity compared to ß TCRs, but it is still more than two orders of magnitude smaller than the affinity of a typical antibody-ligand complex. For another -TCR/antigen pair the affinity was not high enough to be measured.89
Alternatively, a TCR may be expressed on the surface of immortalized T-cell lines, which may be used as a highly sensitive tool to screen candidate antigen preparations, eg, from candidate tissues, cultured cells, or those obtained by protein-chemical methods. For example, we recently investigated the antigen recognition pattern of a TCR previously identified in muscle tissue of a patient with polymyositis.13,14,72
This is one of the few examples, if not the only known case, of a monoclonal T-cell expansion in the target tissue of a patient with autoimmune disease.11
We reconstructed the cDNA of this TCR, and expressed it in 58-ß- cells. The transfectants expressed the human -TCR in the context of the murine CD3 signaling-complex on their cell surface and were activated by appropriate antibodies. Importantly, they were also activated by extracts of cultured human myoblasts and rhabdomyosarcoma cells (Figure 4A)
. Recognition was mediated by the CDR3 region of the TCR, because control transfectants with defined exchanges in their TCR sequences could be activated by anti--TCR antibodies, but they no longer recognized muscle antigen (Figure 4B)
. The data suggest that this TCR indeed recognizes an autoantigen expressed in the target tissue. Although the target antigen has not yet been molecularly defined, this example illustrates that the strategies discussed here may help in the search for unknown (auto-) antigens recognized by infiltrating T cells in human target tissues.
|
Recent technical progress has provided a series of tools that might eventually allow us to identify the target antigens recognized by infiltrating T cells. Potentially pathogenic T cells are first identified in tissue, detected if possible "in flagranti". Next, the genes of their antigen receptors are cloned. The TCR proteins are then expressed in recombinant expression systems, which finally can be used for screening peptide- or protein-expression libraries. The achievement of each of these individual steps signifies a major breakthrough. Here we have described some recent techniques and provided several examples for each of the individual steps of TCR characterization, including identification and cloning of TCR chains of monoclonally expanded infiltrates or of TCR chains from individual, presumably pathogenic cells, recombinant expression of these molecules, and attempts to identify the target antigens of ß or TCRs. However, the successful combination of all of the successive steps, starting from a patient’s biopsy and ending with the clear-cut identification of the target antigen(s), has not yet been accomplished. One of the bottlenecks is that T-cell infiltrates are usually not monoclonal but polyclonal. Therefore, elaborate techniques like single-cell PCR are required to detect the corresponding TCR - and ß-chain pairs. Moreover, identification of the TCR -chain sequence is especially problematic, although recent results indicate that it is in principle possible. This raises the hope that the successful identification of both TCR chains may be reliably achieved in near future. Another problem that needs to be overcome is the low affinity of many TCR molecules to their antigens, but published work has shown at least some receptors might have sufficient affinity to be analyzed in vitro. Despite these problems, the identification of T-cell antigens therefore seems to be almost within our reach. This is a worthwhile goal, for not only will we eventually be able to explain the pathogenesis of autoimmune diseases, but also to develop new diagnostic and therapeutic approaches.
Acknowledgements
We thank Judy Benson for helpful comments on the manuscript.
Footnotes
Address reprint requests to Klaus Dornmair, Ph.D., Max-Planck-Institute of Neurobiology, D-82152 Martinsried, Germany. E-mail: dornmair{at}neuro.mpg.de
Supported by Hermann und Lilly Schilling Stiftung and Deutsche Forschungsgemeinschaft grants SFB 571-A1 and 571-A3.
Accepted for publication May 28, 2003.
References
T-cell receptors and their recognition of non-peptide antigens. Mol Immunol 2002, 38:1051-1061[Medline]
T-cell receptor repertoire in acute multiple sclerosis lesions. Proc Natl Acad Sci USA 1992, 89:4588-4592ß variable and joining gene expression by muscle-infiltrating lymphocytes in the idiopathic inflammatory myopathies. J Immunol 1994, 152:2569-2576[Abstract]
/ receptor. N Engl J Med 1991, 324:877-881[Abstract]
/ T-cell receptor. J Exp Med 1992, 176:1785-178912.1 expression on CD8+ T cells: evidence that V participates in shaping the peripheral T-cell repertoire. J Exp Med 1991, 174:639-648 chains: dual receptor T cells. Science 1993, 262:422-424 chain. Science 1989, 243:217-220ß versus T lineage commitment. Immunity 1998, 9:649-655[Medline]
and ß genes from a single T cell without the establishment of a T-cell clone. Clin Exp Immunol 2001, 123:340-345[Medline]
chains on the surface of normal murine T cells. Eur J Immunol 1995, 25:1617-1623[Medline]
ß heterodimers. Proc Natl Acad Sci USA 1991, 88:8077-8081 and T-cell receptor extracellular segments. Proc Natl Acad Sci USA 1994, 91:11408-11412ß T-cell receptor interactions with syngeneic and allogeneic ligands: affinity measurements and crystallization. Proc Natl Acad Sci USA 1997, 94:13838-13843ß murine T-cell receptor in Escherichia coli. J Immunol Methods 1997, 206:163-169[Medline]
ß with an antigenic Tax peptide from human T lymphotropic virus type 1 and the class I MHC molecule HLA-A2. J Immunol 1996, 157:5403-5410[Abstract]
domain expressed in bacteria: does it dimerize in solution? J Exp Med 1996, 184:1251-1258ß T-cell receptors and coreceptors CD8 and CD4 in their recognition of peptide-MHC ligands. Trends Immunol 2002, 23:408-413[Medline]
and ß T-cell receptor genes. Nature 1986, 320:232-238[Medline]
- and ß-gene products as tool for the production of human T-cell receptor-specific monoclonal antibodies. Hum Immunol 1992, 35:141-148[Medline]
ß-human T3 receptor complexes. Nature 1987, 325:125-130[Medline]
2 homodimer. J Immunol 1992, 148:3469-3477[Abstract]
cytoplasmic domain mediates CD2-induced T-cell activation. J Exp Med 1992, 176:139-145- and ß-chain transcripts reveals a nonfunctional -mRNA of BW5147 origin. Eur J Immunol 1989, 19:2269-2274[Medline]
TCR derived from a polymyositis lesion. J Immunol 2002, 169:515-521 T cells: a right time and a right place for a conserved third way of protection. Annu Rev Immunol 2000, 18:975-1026[Medline]
T cells: functional plasticity and heterogeneity. Nat Rev Immunol 2002, 2:336-345[Medline]
T cells that recognize an inducible MHC class Ib molecule. Science 2000, 287:314-316 TCR interaction can be modulated by the glycosylation of the ligand. J Immunol 1999, 163:288-294This article has been cited by other articles:
 |
|
 |
|
  N. Schwab, C. G. Bien, A. Waschbisch, A. Becker, G. H. Vince, K. Dornmair, and H. Wiendl CD8+ T-cell clones dominate brain infiltrates in Rasmussen encephalitis and persist in the periphery Brain, May 1, 2009; 132(5): 1236 - 1246. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Junker, J. Ivanidze, J. Malotka, I. Eiglmeier, H. Lassmann, H. Wekerle, E. Meinl, R. Hohlfeld, and K. Dornmair Multiple sclerosis: T-cell receptor expression in distinct brain regions Brain, November 1, 2007; 130(11): 2789 - 2799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Hohlfeld and H. Wekerle Autoimmune concepts of multiple sclerosis as a basis for selective immunotherapy: From pipe dreams to (therapeutic) pipelines PNAS, October 5, 2004; 101(suppl_2): 14599 - 14606. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Skulina, S. Schmidt, K. Dornmair, H. Babbe, A. Roers, K. Rajewsky, H. Wekerle, R. Hohlfeld, and N. Goebels Multiple sclerosis: Brain-infiltrating CD8+ T cells persist as clonal expansions in the cerebrospinal fluid and blood PNAS, February 24, 2004; 101(8): 2428 - 2433. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
