Volume 100, Issue 4 pp. 424-431
Free Access

Failure to remove autoreactive Vβ6+ T cells in Mls-1a newborn mice attributed to the delayed development of B cells in the thymus

M. Touma

M. Touma

Department of Cell Science, Graduate School of Science and Technology,

Search for more papers by this author
K. J. Mori

K. J. Mori

Department of Molecular and Cellular Biology, Faculty of Science, Niigata University, Niigata, Japan

Search for more papers by this author
M. Hosono

M. Hosono

Department of Cell Science, Graduate School of Science and Technology,

Search for more papers by this author
First published: 25 December 2001
Citations: 9
Dr M. Hosono, Department of Cell Science, Graduate School of Science and Technology, Niigata University, Ikarashi, Niigata 950–2181, Japan.

Summary

Clonal deletion of autoreactive T cells in the thymus is one of the major mechanisms for establishing tolerance to self-antigens, and self-reactive T cells bearing Vβ6 T-cell receptors are usually deleted before their maturation in Mls-1a mice. However, these T cells develop transiently in the neonatal thymus, and migrate to the periphery. In order to understand the mechanisms which permit these potentially auto-toxic T cells to generate, we investigated in vivo the physiological or functional properties of the elements involved, such as neonatal T cells, antigens and antigen-presenting cells (APC). Confirming the previous findings that each of these elements per se is already completed in function in neonates, we investigated the possibility of the absence or immaturity of particular APC with Mls antigens of their own products in the neonatal thymus. In the search for the cellular and histological changes occurring in the newborn thymus, we found that the elimination of Vβ6+ T cells progressed in parallel with the development of thymic B cells. Involvement of B cells in purging the autoreactive T cells from the newborn thymus was shown by prevention of the deletion of Vβ6+ T cells after the removal of B cells by the treatment of neonates with anti-immunoglobulin M antibodies. The restricted and stable expression of CD5 on the thymic B cells, but not on the splenic cells, suggests that these B cells are not postnatal immigrants from the periphery. Finally, it is concluded that the deficiency in the deletion of self-reactive T cells in the thymus of Mls-1a neonates is due to the delayed development of B cells.

Introduction

Clonal deletion is one of the fail-safe mechanisms involved in the prevention of generating self-reactive T cells, being understood under the advantages of both superantigens, some of which are encoded within endogenous mammary tumour virus (Mtv) genes, and T cells expressing the antigen-corresponding Vβ T-cell receptor (TCR) gene segments. 1–3 As a well-known case, T cells bearing Vβ6 TCR are eliminated during the differentiation and maturation processes in the thymus of Mls-1 (Mtv-7)-bearing (Mls-1a) mice just before their entry into the single positive stage of development.4,5

However, purging the self-aggressive cells from the T-cell pool could not be completed during the first few days after birth. 5–7 Thus, autoreactive Vβ6+ T cells are generated in the Mls-1a neonatal thymus 7–9 and a proportion of them are peripheralized in an anergic state. 7 Exploration of the mechanisms which allow the autoreactive T cells to develop and mature is important for providing a clue to understanding the generation of organ-specific autoimmune diseases in the neonatally thymectomized mice. 10–12

In order to clarify the mechanisms which allow the generation of self-aggressive T cells in neonates, attempts have been made based on suspicions of abnormality in at least one of the three components, developing T cells, antigen-presenting cells (APC) and the antigens involved, and have resulted in the finding that deletion-resistance of T cells and/or immaturity of APC in neonates are unlikely.7,13 Therefore, the significance of the antigenic properties still remains to be solved. The same temporary goal is also seen in the report by Signorelli et al., 14 who used appropriate model mice, including transgenic mice, in terms of molecules or functions different between the newborn thymus and the adult one. Although these authors concluded that defective clonal deletion in neonates is most likely due to the lack of antigen, the mRNA level of a group of Mtv genes in the neonatal thymus was found to be similar to that in the adult one, somewhat contrary to the report of the low-level expression of other Mtv genes in the neonatal thymus, 13 and all these Mtv genes dealt with were accompanied by a loss of the corresponding Vβ-bearing T cells in the thymus of young adults. In addition to the ontogenic changes of the Mtv transcript level in the thymus, both the differences in the amount of Mtv-mRNA among the different type of cells 15–18 and the fact that T-cell deletion was induced not only by the professional APC, such as dendritic cells (DC) and B cells, 19–21 but also by CD8+ T cells, 22 make it difficult to define what decides the appearance or disappearance of the autoreactive T cells.

Despite this large amount of information, it may not be easy to focus on and define the elements engaged in the deletion or survival of the autoreactive T cells at the site at the right time in the neonates. It may be fruitful to observe in detail the phenomena occurring in the developing thymus, where all the components required are functionally constructed. Here, we investigated the elements changing before and after the deletion of Vβ6+ T cells in the thymus of Mls-1a mice. In agreement with the previous reports that T cells in the early postnatal thymus are capable of receiving and transducing the deletion signals and that Mls antigens are present in the neonatal thymus, our results provide direct evidence that thymic B cells are indispensable for removing self-reactive T cells by the fact that B-cell depletion by anti-immunoglobulin M (IgM) antibody-treatment prevented Vβ6+ T-cell elimination in the newborn thymus. The stable and restricted expression of CD5 on the thymic B cells revealed their origin in the thymus, showing that the thymus is really the central organ for induction of self-tolerance by deletion of the autoreactive T cells.

Materials and methods

Mice

BALB/c (H-2d, Mls-1b), AKR/J (H-2k, Mls-1a) and B10.BR (H-2k, Mls-1b) mice were originally purchased from Japan SLC. Inc. (Shizuoka, Japan). (BALB/c × AKR/J)F1 (CAKF1) hybrids and (BALB/c × B10.BR)F1 hybrids were produced and maintained in our animal facility, and the date of birth was referred to as day 0 of age.

Antibodies and reagents

Culture supernatants of the following antibody-producing hybridomas were used as the primary antibodies for the indirect cell staining; 44-22-1 for anti-Vβ6 (rat IgG), F23.2 for anti-Vβ8.2 (mouse IgG) and RA3.6B2 for anti-B220 (rat IgG). Both biotin-labelled anti-rat IgG and biotin-labelled anti-mouse IgG (KPL, Gaithersburg, MD) were used as the secondary antibodies, and fluorescein isothiocyanate (FITC)-streptavidin or phycoerythrin (PE)-streptavidin (PharMingen, San Diego, CA) were used for the third reagents in cell staining. Biotin-conjugated monoclonal antibody for CD5 (Ly-1), FITC-anti-TCRβ, FITC-anti-CD11c (PharMingen) and FITC-anti-IgM (µ-chain) (Caltag, Burlingame, CA) were also used for cell staining. Affinity-purified goat anti-mouse IgM (µ-chain) (Chemicon, Temecula, CA) was used for B-cell-targeting by intravenous (i.v.) injection into newborn mice. Staphylococcal enterotoxin B (SEB) was purchased from Sigma Co. (St. Louis, MO) and dissolved in phosphate-buffered saline (PBS).

Cell preparations

Single-cell suspensions from the thymus and spleen from any aged mice were prepared by mincing the organs on a 200-mesh stainless-steel screen. For the flow cytometric analyses, all samples were washed with PBS and resuspended in staining buffer (PBS containing 2% fetal calf serum and 0·02% NaN3). Red blood cells, when necessary, were removed by suspending the lymphoid cells in a haemolytic solution. In the case of an inoculation, cell suspensions from the thymus and spleen were prepared in Eagle's minimum essential medium (MEM, Nissui Seiyaku Co., Tokyo, Japan) without any serum supplement, as previously described. 9 Cells were washed three times in MEM before use.

Flow cytometric analysis

Approximately 2 × 106 cells were stained by several combinations of primary antibodies as described above and, if necessary, followed by treatment with appropriate antibodies conjugated with fluorescein dyes. All samples were analysed with a fluorescence-activated cell sorter (FACScan) flow cytometer and data were processed with lysis ii software (Becton Dickinson, Mountain View, CA). Dead cells were gated out by means of forward and sideways scatter and propidium iodide staining. Between 10 000 and 50 000 cells were analysed per sample in each experiment.

Tolerance induction

Detailed methods for neonatal tolerance induction have been previously described. 9 Briefly, a 50-µl cell suspension aliquot was prepared from the spleen of CAKF1 mice, and was injected via the facial vein into each newborn mouse within 24 hr of birth. The inoculum used as the tolerogen usually contained 2 × 107 spleen cells. Two days after the injection, unless otherwise mentioned, thymus cells obtained from the recipients were used for the flow cytometric analysis.

Local host-versus-graft reaction (HVGR) of popliteal lymph node swelling assay

To assess the Mls-immunogenicity of the thymocytes and spleen cells of CAKF1 mice, we used the popliteal lymph node (PLN) swelling assay for the local HVGR, as previously described.9,20 Briefly, 20 µl of 2 × 107 cells in MEM without serum supplement was injected into each hind footpad of Mls-negative, major histocompatibility complex (MHC)-identical (BALB/c × B10.BR)F1 hybrid mice at the age of 7 weeks. For a negative control, MEM only or Mls-negative cells were injected instead. Four days after the injection, the recipients were killed and the PLN were carefully removed and weighed to an accuracy of 0·01 mg.

B-cell depression by anti-IgM treatment

Ten to eighty micrograms of goat anti-mouse IgM dissolved in PBS was i.v. injected into CAKF1 mice within 24 hr after birth. Five days later the recipients were killed and their thymi were used for the flow cytometric analysis and for the histological analysis.

Histological observation

Immediately after each thymus was embedded in optimal cutting temperature (OCT) compound (Miles Lab. Inc., Naperville, IL), the tissue block was frozen and stored at −80° until cryostat sections were cut. Frozen sections of 5 µm thickness were air-dried and fixed with cold acetone for 10 min. Immunohistological observations were carried out as previously described. 9 For TCR-bearing cells, the sections were washed with PBS and incubated overnight with the rat anti-mouse Vβ6 primary antibody at 4°. After the primary antibody was washed out with PBS, specimens were treated with biotin-conjugated goat anti-rat IgG for 1 hr at room temperature, then followed by a further incubation with FITC-labelled streptavidin for 1 hr. For detection of the B cells, FITC-anti-IgM was used. These specimens were examined by fluorescence microscopy.

Statistical analysis

The significance between the experimental and control groups was determined using the Student's t-test. Differences were considered to be significant if the P-value was less than 0·05.

Results

Presence of Mls antigen in an immunogenic form in the newborn thymus

Developmental profiles of Vβ6+ T cells in the thymus of either Mls-1-positive (Mls-1a) or Mls-1-negative (Mls-1b) mice are shown in Fig. 1. Irrespective of the Mls-genotype, Vβ6+ T cells showed a slight increase in the few days after birth, reaching a plateau on day 3 of life, after this it started to decrease in the Mls-1a mice toward the background level within the next few days. In order to specify the elements causing this kinetic profile of Vβ6+ T-cell development, we tested first Mls-antigen-presenting activity, or functional combination of the elements required, i.e. a complex of MHC class II and Mls antigens on APC in the newborn thymus by means of the MHC-syngeneic local HVG assay. Mls-immunogenic activity on the day of birth, showing a significant response (1·9 times the PLN-swelling in comparison with the medium control), was comparable to the activity observed at 5 days and more after birth, when the clonal deletion has progressed or is almost completed. Thus, it is suggested that the Mls antigens in the neonatal thymus are in a functional form to stimulate T cells and that the Mls-immunogen is not responsible for the deletion of Vβ6+ T cells.

Kinetic profiles of Vβ6+ T-cell development in the thymus of Mls-1-positive and -negative newborn mice. Thymus cells from newborn BALB/c (Mls-1b; ▴), AKR/J (Mls-1a; ○) and their hybrid (BALB/c × AKR/J)F1 (CAKF1) (Mls-1b/a; ●) mice are analysed for Vβ6 expression in the TCRαβ-positive cell population several days after birth, as indicated. Each symbol indicates a mean value of about five mice with a bar of SD.

Deletion susceptibility of Vβ6+ T cells appearing in the newborn

The susceptibility for deletion of Vβ6+ T cells in the 2-day-old Mls-1a thymus was tested by an injection of Mls-bearing adult spleen cells, because it is well known that Vβ6+ T cells are eliminated when Mls-bearing spleen cells are injected into Mls-negative neonates.5,9, 23,24 Thus, on the day of birth CAKF1 (Mls-1a) mice were inoculated with spleen cells from adult syngeneic mice. Within 2 days after the cell injection, when the autoreactive T cells are not yet deleted in the thymus of control mice, Vβ6+ T cells were significantly reduced from 8·0% in the control mice to 2·2% in the cell-injected mice, while Vβ8.2+ T cells of the antigen-unrelated control were not affected at all. In contrast, Vβ6+ T cells in the Mls-1a newborn thymus which has Mls-immunogen, as mentioned above, were not eliminated by an injection of the Mls-negative BALB/c spleen cells, suggesting that the antigens on the neonatal thymocytes could not be presented by the Mls-negative spleen cells to the corresponding T cells. Furthermore, the presence of the deletion-inducing APC has been demonstrated by an injection of bacterial superantigen, SEB, resulting in an early and strong elimination of the corresponding Vβ8+ T cells in the newborn thymus (data not shown). Based on all the above data, it is suggested that Mls antigens in the immunogenic form may not be in the form for clonal deletion, i.e. a possibility of a close association of these antigens with a special type of APC, which may appear in the thymus a few days after birth.

In the course of removal of the Vβ6+ T cells in the 2-day-old thymus by an injection of syngeneic adult spleen cells into newborn mice, it was found that the age of the spleen cell donors was critical. Spleen cells from CAKF1 mice during a weaning period are not fully effective to remove Vβ6+ T cells, or immature in their tolerogenicity. On the other hand, the Mls-immunogenicity in these spleen cells, as assessed in the local HVG reactivity, has been found to be as effective as those of adults (data not shown). These findings indicate that the element(s) which is required for the clonal deletion develops after those for the positive immune reaction, suggesting an independent generation of tolerance inducibility from that of immunogenicity.

B cells accumulating in the medulla may be related to the Vβ6+ T-cell deletion in the neonatal thymus

In order to search for the responsible cells, we analysed first by flow cytometry the developmental profiles of DC and B cells, both being well-known potent APCs, and tested which are more closely related to the deletion of Vβ6+ T cells in the developing thymus of Mls-1a mice. As indicated in Fig. 2, both of them correlated with the decreasing number of Vβ6+ T cells during the first week of life. Thus, DC (CD11c+) as well as B cells (B220+ IgM+) appear to be fundamentally comparable as a function of Vβ6+ T-cell elimination, though the slope of the correlation lines is different from each other. Another type of cell, categorized as B220+ IgM, was found to be rather constant. However, B cells seem to participate more deeply than DC based on the following two observations. First, the artificial removal of Vβ6+ T cells in the 2-day-old thymus by injecting adult spleen cells into newborn mice produces a conspicuous increase in the number of B cells and less or only marginal change in DC ( Fig. 3). Second, the ontogenetic profile of the B-cell development shows a sharp increase between day 3 and day 5 of age, during which time almost complete removal of the Vβ6+ T cells was achieved ( Fig. 4). During these days, the accumulation of B cells in the medulla, especially the outer medulla, became prominent ( Fig. 4b). Therefore, in the next experiments, an attempt was made to prevent the generation of intrathymic B cells by an i.v. injection of anti-mouse IgM antibodies into newborn mice.

Inverse correlation between the number of Vβ6+ T cells and that of DC/B cells in the neonatal thymus of Mls-1a mice. The percentage of Vβ6+ T cells in the TCRαβ-positive cells was individually plotted against the percentage of DC (CD11c+○) or B cells (B220+, IgM+●) in the thymus from CAKF1 mice of several ages within the 1st week of life. The straight lines were obtained by the least squares method, and the correlation coefficient (r) between Vβ6+ T cells and DC or B cells was −0·91 (P < 0·001) or −0·76 (P < 0·001).

Prominent increase in the number of B cells accompanied by a decrease of Vβ6+ T cells in the thymus of 2-day-old Mls-1a mice receiving spleen cells of adult syngeneic mice. Within 24 hr after birth CAKF1 (Mls-1-positive) mice were i.v. injected with media, as a control, or with 2 × 107 spleen cells from adult syngeneic mice. Two days after the injection, their thymus cells were analysed by flow cytometry for Vβ6-bearing T cells, DC and B cells. The data were expressed as a mean value against those of the medium-injected controls with a bar SD of the values. n, number of mice used.

B-cell development in the thymus of neonatal mice. The developmental profile of B cells is inverse to the kinetic profile of autoreactive Vβ6+ T cells in the early thymus of Mls-1a mice as shown in (a) where percentages of IgM+ cells coexpressing B220 in the thymocytes from CAKF1 mice were individually plotted (●) several days after birth, and simultaneously, the developmental kinetics of Vβ6+ T cells (○) appearing in Fig. 1 is overlapped. Each closed symbol refers to each mouse, or a mean of nine mice with a bar of SD. Thymi from 1- or 6-day-old CAKF1 mice were prepared for immunohistology (b) by staining their sections with FITC-conjugated anti-IgM antibodies: C, thymus cortex; M, medulla; magnification × 80; and for flow cytometry (c) in double staining with both anti-B220 and anti-IgM antibodies. The percentage of B220+ IgM+ cells and B220+ IgM cells in the quadrants are indicated at the top of each quadrant.

Removal of B cells in the developing thymus allows the prolonged generation of Vβ6+ T cells

A single injection of a given amount of anti-IgM antibodies into neonates just after birth caused the disappearance of B cells in the thymus ( Fig. 5a,b) as well as in the periphery, and simultaneously prevented the deletion of Vβ6+ T cells in the CAKF1 thymus 5 days after birth, when the deletion should have been almost completed ( Fig. 5c). The proportion of B cells is dependent on the amount of anti-IgM antibodies injected and is also highly correlated with the number of Vβ6+ T cells in the thymus of these mice ( Fig. 6). B-cell depletion only affected the development of Vβ6+ T cells and there was no increase in the number of Vβ8.2+ T cells, rather, the latter showing a tendency to decrease because of the increase in Vβ6+ T cells. These rescued Vβ6+ T cells were functional as seen in the anti-Mls-reactivity of the thymus cells in the syngeneic local GVH reaction, resulting in an increased swelling of PLN by a factor of 2·9 in comparison with those from the antibody-untreated mice.

Treatment with anti-IgM antibodies suppresses the development of B cells and the deletion of Vβ6+ T cells in the thymus of newborn Mls-1a mice. On the day of birth, CAKF1 mice were i.v. injected with 40 µg of anti-IgM antibodies. Five days later, mice were killed and their thymi were used for immunohistology (a) and flowcytometric analysis (b and c). Thymus sections were stained with either FITC-conjugated anti-IgM or anti-Vβ6 antibodies (a); C and M are for the cortex and medulla, respectively, and magnification, × 60 for IgM, × 120 for Vβ6, and thymocytes were stained with both anti-B220 and anti-IgM antibodies (b) or anti-Vβ6 and anti-TCRαβ antibodies (c); the percentage of B220+ IgM+ cells and B220+ IgM cells in the quadrants are indicated at the top of each quadrant in (b) and the figures on the histogram indicates percentage of Vβ6 positive cells in the TCRαβ+ population in (c).

Removal of B cells prevents the deletion of autoreactive Vβ6+ T cells in the thymus of neonatal Mls-1a mice treated with anti-IgM antibodies. On the day of birth, CAKF1 mice were injected with anti-IgM antibodies (10–80 µg/mouse) or with PBS, as controls. Five days after the injection, thymocytes were routinely analysed for the proportion of Vβ6+ T cells, Vβ8.2+ T cells, as an unrelated population of T cells, and B cells. The percentage of B cells in the thymus was individually plotted against that of Vβ6+ (circles) or Vβ8.2+ (squares) cells in the TCRαβ-positive cells. The proportion of Vβ6+ T cells, but not of Vβ8.2+ T cells, shows an inverse correlation with that of B cells, the correlation coefficient (r) being −0·7 (P < 0·001).

B cells in the early thymus may not be postnatal immigrants from the periphery

Whether the thymic B cells in the neonates are really of thymus origin is an important question to understand the thymus as a central organ of self/non-self discrimination. During the few weeks after birth, the majority of thymic B cells are large in size in comparison with splenic ones and about 90% express CD5 ( Fig. 7a). In the following weeks, thymic B cells change their CD5 expression pattern. Thus, indicating a stable existence of CD5+ B cells in the thymus during the suckling period, while B cells in the periphery were consistently negative in CD5 ( Fig. 7a). B cells in the early thymus may not be of extrathymic origin, and if they originated in the periphery, they would consist of B cells expressing no CD5. Indeed, when adult spleen cells were injected into newborn mice, small and CD5-negative B cells became prominent in the recipient's thymus ( Fig. 7b). These observations suggest that B cells in the neonatal thymus are not postnatal immigrants from the periphery.

B cells in the thymus of suckling mice are distinct from those in the periphery in their expression of CD5. B cells were analysed by gating on IgM-positive cells after double staining with anti-IgM and anti-CD5. Histograms represent expression of CD5 on B cells and small insertions (boxes) show the size of their cells in the forward-scatter analysis. (a) Ontogenic profiles of thymic and splenic B cells. (b) Demonstration for immigration of the CD5-negative peripheral B cells in neonatal thymus receiving adult spleen cells by means of the experimental protocol shown in Fig. 3. Thymic B cells were analysed for CD5 expression and the forward scatter, as inserted.

Discussion

Much interest has been paid to the cellular and molecular basis for the incompleteness of elimination of autoreactive T cells during the first few days after birth in the mouse, because neonatal thymectomy retards the removal of potentially autoreactive T cells in the periphery, as demonstrated in the endogenous superantigen (SAg)-bearing mice treated by thymectomy at 3 days, 7 and often induces organ-specific autoimmune diseases. 10–12 The mechanism(s) which allows these autoreactive T cells to be generated in the limited periods after birth are not yet well understood. Although several searches for the generation of autoreactive T cells are consistent in the responsibility of the nature of SAg-presentation in the newborn thymus,13,14 it remains to be clarified.

Our interesting finding is that the developmental profile of Mls-tolerogenicity in the thymus was dissociated from that of the Mls-immunogenicity, i.e. the former demonstrated by the deletion of Vβ6+ T cells appears only after day 3 of life (refs. 5–7 and Fig. 1), while the latter has been already generated on the day of birth and may not change through the 1st week, as seen in nominal antigens. 25 This suggested the involvement of a unique type of APC for tolerogenicity, and/or their limited localization to an anatomical site, and may be reflected by the accumulating information that the amount of Mtv transcripts differs among types of cells, including cells other than the well-known professional APC; for example, the appreciable level of transcription of Mtv-7 in B cells, DC and CD8+ T cells16,17 and of Mtv-1, -6, -8 and -11 in B cells but not in DC. 18 However, one might think that the involvement of a unique APC in the Vβ6+ T-cell deletion is unlikely, because it is often demonstrated that the Mtv-7 transcripts could be intercellularly transferred. Indeed, when Mtv antigens are produced elsewhere, as seen in the case of the bacterial superantigen, SEB, 14 they would be presented by conventional APC to induce the deletion of the corresponding T cells in the thymus. 26 In the present report using SEB as Signorelli et al. used, 14 showed the presence of functional APC in the newborn thymus, which would participate in the deletion of T cells if the corresponding antigens are appropriately provided.

Although the Mls antigens participating in the immunogenicity of the newborn thymus may not be intercellularly transferable to bind APC, our present data are not against the idea of their ability to be transferred to an appropriate APC as reported by others.22,26 This idea only came from the deletion of the corresponding T cells more than 2 weeks after the experiments started, both in the neonatal tolerance induction experiments and in the radiation-reconstitution experiments under the donor–recipient combination between I-E+ Mls and I-E Mls+ strains of mice. 26 In contrast to this, within a few days after birth the Mtv gene products may not be transferable, irrespective of their quantity or quality. Our present data were obtained from 2-day-old mice, probably when the Mtv transcripts are restricted in a given type of cells and not yet fully extended throughout the thymus. Finally, the responsible cells in the early deletion of Vβ6+ T cells are targeted by the treatment of neonates with anti-IgM antibodies which removed B cells in the thymus ( Fig. 5).

Cells involved in the removal of autoreactive T cells in the late ontogeny may not be a single type, while in neonates, thymic B cells are probably sufficient to remove the corresponding Vβ6+ T cells, especially when they are only the cells of Mtv producers. Indeed, B cells are thought to be one of the main producers of Mtv-7 transcripts in adult mice.16,19 Direct involvement of thymic B cells in clonal deletion was shown in a previous report, 20 where thymic B cells were administered into the thymus of both Mls-1-negative and I-E-negative mice and the intercellular transfer of their Mls-products was not demonstrated within a short administration period of 1 week. On the other hand, DC caused only an anergy state of unresponsiveness, 20 and anergic T cells, which generally could not proliferate in vivo, are gradually removed from the periphery in the growing individuals. 7 This would also result in their clonal elimination in adult life. Such a functional difference between B cells and DC in the negative selection, both of which have a similar level of Mtv transcripts, 17 remains unclear.

As discussed above, the level of mRNA of Mtv genes depends on the type of cell, so that the effect of artificial disruption of B-cell development on the specificity of T cells varies among Vβ segments used. Thus, Gollob and Palmer 27 demonstrated that in the B-cell-suppressed mice, Vβ11+ T cells directing to Mtv-9 SAg were not deleted and retained in the periphery, while removal of Vβ6+ T cells directing to Mtv-7 SAg were not affected by the removal of B cells. The latter case may be seemingly contrary to our data presented here, but the deletion of Vβ6+ T cells in the periphery is probably due to the intercellular transfer of the Mtv-7 gene products by other than B cells. The case of the µm-targeted, B-cell-deficient mice may be similar, where Vβ6+ T cells are deleted normally in the periphery of adults as seen in wild-type, although unexpectedly, the gene-targeting rather promoted the Vβ6+ T-cell removal in the early thymus when the deletion has not yet started in the wild-type. 28 Endogenous SAgs provide a good model for the central control system of the thymus in self-recognition. B-cell engagement in the developing thymus to remove autoreactive T cells and their delayed development in neonates allows the transient appearance of auto-toxic T cells in the periphery. Presently, the biological significance of the appearance of the self-reactive T cells in early life is still an open question.

Acknowledgments

We thank Drs K. Onoe (Hokkaido University, Sapporo), M. Inaba (Kansai Medical University, Moriguchi) for providing hybridomas, J. Yodoi (Virus Institute, Kyoto University, Kyoto), T. Abo (Faculty of Medicine, Niigata University) for their encouragement through the study, and S. Araya for supporting this research. Thanks are due to Dr S. Odani (Faculty of Science, Niigata University) for his kind advice on molecular research. Part of this work was supported by a grant-in-aid for the Scientific Research from the Ministry of Science, Education and Culture of Japan.

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.