Selection of dual Valpha T cells

Incomplete allelic exclusion of TCRa gene rearrangement permits the generation of dual Valpha T cells, though the issues of their frequency and whether both alphabeta pairs participate in thymic selection have not been resolved. Both questions have been investigated using lymphocytes from mice hemizygous at the TCRa locus and consequently unable to express two rearranged TCRa genes, as background controls. The data presented show that both the frequency of dual Valpha T cells and the relative expression levels of co-expressed Valpha chains are variable and are determined by thymic selection. Possession of a Valpha chain which is inefficiently positively selected appears to increase the likelihood that a second Valpha chain will be co-expressed, whilst the relative cell surface levels of a given pair of Valpha chains differ between CD4 and CD8 subsets. Further, for some but not all Valpha pairs, dual Valpha T cells appear to express elevated levels of surface TCR. Finally, contrary to previous claims, dual Valpha T cells do not appear to be relatively frequent amongst immature thymocytes.

T cell repertoire formation displays characteristics of qualitative models of thymic selection

The use of T cell receptor elements varies between mouse strains, reflecting a balance between positive and negative selection. The presence of H-2E biases V alpha and V beta usage through major histocompatibility class II isotype preferences of V elements, and mammary tumor virus-dependent, negative selection. Quantitative models of thymic selection predict that negative selection equates to 'excess' positive selection, whereas qualitative models suggest that positive and negative selection are opposing forces. This report attempts to distinguish between the models by assessing whether, at the level of the T cell repertoire, positive and negative selection have quantitative or qualitative characteristics. The data show that the effect of bearing V alpha and V beta regions which are both preferentially (or negatively) selected in the presence of H-2E is additive or synergistic, whilst positive stimuli counteract negative ones. The data thus provide support for qualitative models of thymic selection.

Streptozotocin-induced diabetes in mice lacking alphabeta T cells

Multiple low-dose streptozotocin (MD-STZ) is widely used for the experimental induction of diabetes, but, as non-obese diabetic (NOD)-scid/scid mice have been found to display enhanced susceptibility to MD-STZ, whether or not the model is genuinely autoimmune and T cell-mediated has been unclear. Mice bearing a targeted mutation of the T cell receptor (TCR) alpha-chain were therefore used to assess whether TCR alphabeta+ cells are involved in the diabetogenic effects of MD-STZ injections. Young NOD mice lacking TCR alphabeta cells, when given five daily injections of 40 mg/kg STZ, developed diabetes at low frequency (2/12), despite the widespread destruction of pancreatic islet cells. By comparison, most normal control mice became hyperglycaemic (12/23). We conclude that whilst much of the tissue destruction observed in this model is due to the direct toxic effect of STZ, a significant amount is also due to the action of TCR alphabeta cells tipping the balance between tolerable and clinically damaging action on islet cells.

Mice lacking alpha beta + T cells are resistant to the induction of experimental autoimmune encephalomyelitis

T cells of the gamma delta subset have been found to localise to demyelinated lesions in both multiple sclerosis and experimental autoimmune encephalomyelitis (EAE) and have been implicated in the pathogenesis of both diseases. We have assessed mice carrying a targeted mutation of the T cell receptor-alpha locus which consequently lack T cell receptor (TCR) alpha beta cells but have an intact gamma delta(+)-T cell population for their susceptibility to EAE. No disease was found in any of the mutant mice, nor was any infiltration of the CNS detected. These data show that, at least in the absence of TCR-alpha beta cells. TCR-gamma delta cells are not able to elicit the pathology associated with EAE.

Non-obese diabetic mice hemizygous at the T cell receptor alpha locus are susceptible to diabetes and sialitis

To test the hypothesis that T cells carrying two T cell receptor (TCR) alpha chains play a role in autoimmunity, we backcrossed the non-obese diabetic (NOD) strain with one carrying a TCR alpha gene disrupted by homologous recombination. Mice carrying one copy of the disrupted gene are incapable of generating T cells carrying two cell surface TCR alpha chains. Our early results suggested that either dual TCR alpha T cells play a role in insulin-dependent diabetes mellitus (IDDM) induction in NOD mice or that a locus co-segregating with the disrupted TCR alpha locus protected mice from diabetes induction. From the analysis both of mice in which the region co-segregating with the disrupted TCR alpha locus is minimized and of the F1 offspring of NOD mice with the 129 strain (TCR alpha hemizygous mice), the apparent protective effect of the absence of dual TCR alpha T cells is lost; thus, such cells do not appear to play a critical role in autoimmune disease in NOD mice.

Dual T cell receptor alpha chain T cells in autoimmunity

Allelic exclusion at the T cell receptor alpha locus TCR-alpha is incomplete, as demonstrated by the presence of a number of T lymphocyte clones carrying two expressed alpha chain products. Such dual alpha chain T cells have been proposed to play a role in autoimmunity, for example, because of a second TCR-alpha beta pair having bypassed negative selection by virtue of low expression. We examined this hypothesis by generating mice of various autoimmunity-prone strains carrying a hemizygous targeted disruption of the TCR-alpha locus, therefore unable to produce dual alpha chain T cells. Normal mice have a low but significant proportion of T cells expressing two cell-surface TCR-alpha chains that could be enumerated by comparison to TCR-alpha hemizygotes, which have none. Susceptibility to various autoimmune diseases was analyzed in TCR-alpha hemizygotes that had been backcrossed to disease-prone strains for several generations. The incidence of experimental allergic encephalomyelitis and of lupus is not affected by the absence of dual TCR-alpha cells. In contrast, nonobese diabetic (NOD) TCR alpha hemizygotes are significantly protected from cyclophosphamide-accelerated insulitis and diabetes. Thus, dual alpha T cells may play an important role in some but not all autoimmune diseases. Furthermore, since protected and susceptible NOD mice both show strong spontaneous responses to glutamic acid decarboxylase, responses to this antigen, if necessary for diabetetogenesis, are not sufficient.

The T cell response of HLA-DR transgenic mice to human myelin basic protein and other antigens in the presence and absence of human CD4

Analysis of HLA class II transgenic mice has progressed in recent years from analysis of single chain HLA class II transgenes with expression of mixed mouse/human heterodimers to double transgenic mice expressing normal human heterodimers. Previous studies have used either HLA transgenic mice in which there is a species-matched interaction with CD4 or mice which lack this interaction. Since both systems are reported to generate HLA-restricted responses, the matter of the requirement for species-matched CD4 remains unclear. We have generated triple transgenic mice expressing three human transgenes, DRA, DRB, and CD4, and compared HLA-restricted responses to peptide between human-CD4+ (Hu-CD4+) and Hu-CD4- littermates. We saw no difference between Hu-CD4+ and Hu-CD4- groups, supporting the notion that for some responses at least the requirement for species-matched CD4 may not be absolute. Evidence for positive selection of mouse T cell receptors in HLA-DR transgenic mice came both from the acquisition of new, HLA-restricted responses to various peptides and from an increased frequency of T cells using the TCR V beta 4 gene segment. An important goal with respect to the analysis of function in HLA transgenic mice is the clarification of mechanisms which underpin the recognition of self-antigens in human autoimmune disease. As a first step towards 'humanized' disease models in HLA transgenic mice, we analyzed the responses of HLA-DR transgenic mice to the human MPB 139-154 peptide which has been implicated as an epitope recognized by T cells of multiple sclerosis patients. We obtained T cell responses to this epitope in transgenic mice but not in nontransgenic controls. This study suggests that HLA transgenic mice will be valuable in the analysis of HLA-restricted T cell epitopes implicated in human disease and possibly in the design of new disease models.

Random activation of a transgene under the control of a hybrid hCD2 locus control region/Ig enhancer regulatory element

Locus control regions such as those of human CD2 and beta-globin differ from classical enhancers in that, whereas the former confer high level, copy-dependent, position-independent expression to linked genes in transgenic mice, the latter do not, expression levels being dependent on the site of integration. We report that the position independence of the CD2 locus control region is modified by coupling it to the immunoglobulin heavy chain enhancer. Whilst in the majority of transgenic lines the Ig heavy chain enhancer has little or no effect on T cell expression of the hCD2 transgene, in others transgene expression is non-specifically extinguished in a proportion of lymphoid cells. The transgenic locus chromatin appears inaccessible to DNase I in these cells, which do not express the gene. Furthermore, mice homozygous for the hybrid hCD2-Ig heavy chain enhancer construct contain T cells with both an active and an inactive transgene. The 'decision' to express or repress the gene appears to be a random process which involves each chromosome separately, occurs at early stages in differentiation and is heritable by daughter cells. These data suggest the possibility that stochastic decisions might control a number of biological processes.

HLA-DR and H-2E transgenes differentially mediate TCR-specific positive selection

The use of HLA transgenic mice in models of immunity and disease assumes that human MHC molecules are able to contribute toward the positive selection of the mouse TCR repertoire. As an initial step towards analysis of this we have compared the relative ability of DR alpha/E beta or E alpha/E beta complexes to induce T cell receptor (TCR) positive selection in H-2Ea and HLA-DRA transgenic mice lacking endogenous E alpha. The results show that, like E alpha/E beta, the hybrid DR alpha/E beta complexes are capable of mediating positive selection of V beta 2+, V beta 6+, and V beta 10+ cells. However, differences were found between the effects of the two transgenes. Thus, while V beta 6+ cells were efficiently selected in both H-2Ea and DRA transgenic mice, positive selection of V beta 10+ cells was less apparent in the DRA transgenic mice. Variation between Ea and DRA transgenic mice is consistent with the notion that this process is dependent on differential binding of endogenous peptides to the E alpha/E beta and DR alpha/E beta complexes. Furthermore, contrary to expectations, in neither set of mice was positive selection limited solely to the CD4+ subset. Thus, examples were found in which V beta-specific positive selection was confined to either the CD4+ or CD8+ subsets, and others in which both subpopulations were concomitantly increased. In the case of V beta 2 positive selection, H-2Ea transgenic mice showed expansion of these cells in both the CD4+ and CD8+ subpopulations while in DRA transgenic mice this occurred predominantly in the CD8+ subpopulation.

Mouse mammary tumor virus-mediated T-cell receptor negative selection in HLA-DRA transgenic mice

Products of specific mouse Mtv genes expressed in association with mouse MHC class II products cause the deletion of T cells expressing particular TCR V beta gene segments. These endogenous deletion ligands have been termed superantigens due to their ability to negatively select entire T-cell families, as defined by V beta-chain usage. In most cases, deletion is preferentially effected through interaction of the Mtv ligand with H-2E products. Although human DR alpha shares only 75% identity with the E alpha chain of H-2E, it has previously been shown to substitute for the mouse homologue in its capacity to induce the deletion of V beta 11- and V beta 17a-bearing T cells. In the present study, we have undertaken a more comprehensive analysis of the interaction of mixed DR alpha/E beta pairs with various endogenous Mtv integrants in various mouse backgrounds, leading to negative selection of particular V beta families. We show in this paper that transgenic DR alpha/E beta can also efficiently interact with products of Mtv-7, causing deletion of both V beta 6+ and V beta 7+ cells. Deletion of V beta 11+ T cells in DRA transgenic mice carrying Mtv-8 and -9, however, was less efficient than in control H-2Ea transgenic mice. These data and those from other MHC transgenic mouse studies show that while the class II alpha chain can influence the interaction with superantigen, it is the identity of the beta chain that seems to be critical.(ABSTRACT TRUNCATED AT 250 WORDS)

Synthesis and use of an internal amino acid sequencing standard peptide

A sequenceable internal standard has been made that can be added to samples destined for automated amino acid sequencing and that provides a means for continuously monitoring instrument performance. The 41-residue standard peptide is composed of two unnatural amino acids, norleucine and succinyl-lysine, and has the structure [norleucine-(succinyl-lysine)4]8-norleucine. The phenylthiohydantoin derivatives of the latter two amino acids are well separated via reversed-phase HPLC from the phenylthiohydantoin derivatives of the 20 naturally occurring amino acids. In addition, the standard peptide can be readily synthesized and it has excellent solubility, stability, and sequencing characteristics. The placement of norleucine at every fifth residue in this peptide permits accurate repetitive yield and carryover calculations following runs that are as short as six cycles. Use of this internal standard permits the early detection of suboptimal instrument performance and is especially helpful in differentiating between two common reasons (blocked NH2-terminus versus an instrument problem) for the failure of an automated protein sequencer to provide an NH2-terminal sequence.

Anergy and suppression in B-cell responses

Two main ideas have been put forward to explain the unexpectedly low anti-hapten antibody titres which can result from pre-priming a mouse with carrier before hapten-carrier immunization. The first involves the interaction of a network of idiotype-specific suppressor T cells, the second instead arguing for the role of intrinsic B-cell anergy. This paper proposes that the data available can equally be interpreted as reflecting the suboptimal interaction between T and B cells at differing stages of maturity, provided that memory B cells can be divided into two subsets. Further, it is suggested that these considerations must be taken into account in the analysis of B-cell anergy in receptor transgenic mice.

The implications of the failure to generate autoantibody-producing hybridomas from rat erythrocyte-immunized mice

Injection of mice with rat erythrocytes (RRBC) has long been thought to provide an experimental model in which suppressor T cells (Ts) control autoimmunity. The basis of this is that whilst mice immunized with RRBC produce an antibody response, of which a proportion cross-reacts with autologous red cells, the RRBC-immunized recipients of RRBC-primed spleen cells make no, or little, autoantibody, and secondly because the transfer of this autoantibody-specific suppression can be abrogated by T-cell depletion of transferred spleen cells. Here an alternative explanation of these phenomena is described.

The influence of adjuvants on the generation of autoantibody and specific suppression in rat erythrocyte-immunized mice

A number of adjuvants were investigated for their ability to modulate either the autoimmune response induced in mice by immunization with rat erythrocytes (RRBC) or the ability of RRBC-primed spleen cells to suppress the induction of anti-red cell autoimmunity in recipient mice. The inability of the agents used to do so is discussed on the background of models used to explain the generation of suppression in this system.

The influence of I-E on the generation of autoantibody and specific suppression in rat erythrocyte-immunized mice

A proportion of the antibodies produced by mice in response to the injection of rat erythrocytes (RRBC) cross-react with autologous red cells. When spleen cells from mice so immunized are transferred to naive syngeneic recipients, the recipient mice produce high anti-RRBC antibody titres but little or no autoantibody. This phenomenon has been attributed to the action of suppressor T cells. To date, the only mouse strain which consistently fails to demonstrate specific suppression of the autoantibody response is the SJL, which lacks the I-E molecules suggested to be important in the generation of suppressor T cells. The results presented here show that some, but not all, I-E negative strains of mice are capable of exhibiting transferable suppression of RRBC-induced antibodies.

Shuffling of amino acid sequence: an important control in synthetic peptide studies of nucleic acid-binding domains. Binding properties of fragments of a conserved eukaryotic RNA binding motif

We have used synthetic peptides to study a conserved RNA binding motif in yeast poly(A)-binding protein. Two peptides, 45 and 44 amino acids in length, corresponding to amino and carboxyl halves of a 90-amino acid RNA-binding domain in the protein were synthesized. While the amino-terminal peptide had no significant affinity for nucleic acids, the carboxyl-terminal peptide-bound nucleic acids with similar characteristics to that for the entire 577 residue yeast poly(A)-binding protein. In 100 mM NaCl, the latter peptide retained over 50% of the intrinsic binding free energy of the protein, as well as, similar RNA versus DNA binding specificity. However, shuffling of the sequence of this 44 residue peptide had surprisingly little effect on its nucleic acid binding properties suggesting the overriding importance of amino acid composition as opposed to primary sequence. Deletion studies on the 44 residue peptide with the "correct" sequence succeeded in identifying amino acids important for conferring RNA specificity and for increasing our understanding of the molecular basis for nucleic acid binding by synthetic peptides. The shuffled peptide study, however, clearly indicates that considerable caution must be exercised before extrapolating results of structure/function studies on synthetic peptide analogues to the parent protein.

Synthetic peptide analogs of DARPP-32 (Mr 32,000 dopamine- and cAMP-regulated phosphoprotein), an inhibitor of protein phosphatase-1. Phosphorylation, dephosphorylation, and inhibitory activity

Synthetic peptides based on the threonine phosphorylation site and proposed inhibitory site of DARPP-32 (dopamine- and cAMP-regulated phosphoprotein, Mr = 32,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) were prepared and analyzed as substrates for cAMP-dependent protein kinase and protein phosphatases-1c, -2Ac (the catalytic subunits of protein phosphatase-1 and 2A, respectively) and -2B, and as inhibitors of protein phosphatase-1c. Studies of the kinetics of phosphorylation of the peptides by cAMP-dependent protein kinase indicated an important role in facilitating phosphorylation for the region COOH-terminal to the phosphorylatable threonyl residue. Studies of the dephosphorylation of the phosphopeptides demonstrated that they were effectively dephosphorylated by protein phosphatase-2A and -2B and poorly dephosphorylated by protein phosphatase-1. The active inhibitory region of phospho-DARPP-32 was analyzed by determining the effects of synthetic phosphopeptides on the activity of protein phosphatase-1c. Phospho-D32-(8-48) and phospho-D32-(8-38) inhibited protein phosphatase-1c with IC50 values of 2 x 10(-8) and 4 x 10(-8) M, respectively, compared with an IC50 of 8 x 10(-9) M for intact phospho-DARPP-32. Phospho-D32-(9-38) was equipotent with phospho-D32-(8-38); however, further NH2-terminal deletions resulted in marked reductions in IC50 values. An analog of an active DARPP-32 phosphopeptide containing a phosphoseryl residue in place of the phosphothreonyl residue also exhibited a much reduced IC50. These data identify the essential inhibitory region of phospho-DARPP-32 as residues 9-38, which contains the phosphorylation site (Thr34). This region exhibits extensive amino acid sequence identity with phosphatase inhibitor-1, a distinct inhibitor of protein phosphatase-1. Kinetic studies of the inhibition of protein phosphatase-1c by phospho-D32-(9-38), a potent inhibitor, as well as by phospho-D32-(10-38), a weak inhibitor, indicated a mixed competitive/noncompetitive mechanism of inhibition, as has been previously found for both intact phospho-DARPP-32 and intact phospho-inhibitor-1. These findings support the hypothesis that a 30-amino acid domain in the NH2-terminal region of phospho-DARPP-32 is sufficient for the inhibition of protein phosphatase-1.

Epitope-specific regulation and the literature

The antibody response to a hapten conjugated to a protein can, in a number of situations, be suppressed by prior immunization with the carrier protein alone. The study of this finding appears to have been handicapped by previous literature on the subject having been consistently misquoted. It is to be hoped that recognition of this will allow a better understanding of the limitations to such suppression, and hence the mechanism by which it may act.

p10 single-stranded nucleic acid binding protein from murine leukemia virus binds metal ions via the peptide sequence Cys26-X2-Cys29-X4-His34-X4-Cys39

The RNA binding protein of 56 residues encoded by the extreme 3' region of the gag gene of Rauscher murine leukemia virus (MuLV) has been chemically synthesized by a solid-phase synthesis approach. Since the peptide contains a Cys26-X2-Cys29-X4-His34-X2-Cys39 sequence that is shared by all retroviral gag polyproteins which has been proposed to be a metal binding region, it was of considerable interest to examine the metal binding properties of the complete p10 protein. As postulated, p10 binds the metal ions Cd(II), Co(II), and Zn(II). The Co(II) protein shows a set of d-d absorption bands typical of a tetrahedral Co(II) complex at 695 (epsilon = 565 M-1 cm-1), 642 (epsilon = 655 M-1 cm-1), and 615 nm (epsilon = 510 M-1 cm-1) and two intense bands at 349 (epsilon = 2460 M-1 cm-1) and 314 nm (epsilon = 4240 M-1 cm-1) typical of Co(II)----(-)S- charge transfer. The ultraviolet absorption spectrum also indicates Cd(II) binding by the appearance of a Cd(II)----(-)S- charge-transfer band at 255 nm. The 113Cd NMR spectrum of 113Cd(II)-p10 reveals one signal at delta = 648 ppm. This chemical shift correlates well with that predicted for ligation of 113Cd(II) to three -S- from the three Cys residues of p10. The chemical shift of 113Cd(II)-p10 changes by only 4 ppm upon binding of d(pA)6, indicating that the chelate complex is little changed by oligonucleotide binding.(ABSTRACT TRUNCATED AT 250 WORDS)

Synthesis of the p10 single-stranded nucleic acid binding protein from murine leukemia virus

The p10 murine leukemia virus (MuLV) protein is a basic single-stranded nucleic acid binding protein encoded by the extreme 3' region of the gag gene of MuLV type C. It contains the Cys-X2-Cys-X4-His-X4-Cys sequence shared by all retroviral gag polyproteins. A similar sequence is found in the gene 32 single-stranded DNA binding protein of bacteriophage T4 and is believed to be the zinc binding region of the protein. Solid phase synthesis of p10 was carried out based on the known primary structure of the native protein, with the exception that the acetamidomethyl (Acm) derivative of cysteine was incorporated at all three cysteine positions. The structure of the synthetic p10 was confirmed by direct amino-acid sequencing, as well as by amino acid analysis and FAB mass spectrometry of endoproteinase Lys-C peptides derived from p10. A Chou and Fasman analysis of the primary sequence predicts that p10 contains 9% beta strand and/or sheet and 36% alpha helix. Circular dichroism experiments carried out on the Acm derivatized peptide gave somewhat different results, in that they suggest that p10 contains approximately 70% random coil, less than 30% beta strand and/or sheet and less than 10% alpha helix. With a Ka of greater than 10(8) M-1 for single-stranded RNA, the synthetic peptide binds as tightly as the p10 protein does when isolated directly from infected HTG-2 cells. The Acm groups can be removed from the synthetic p10 peptide by the use of mercuric acetate, followed by treatment with dithiothreitol to sequester the mercuric ion.(ABSTRACT TRUNCATED AT 250 WORDS)