| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Cell Biology, Physiology, and Immunology, Faculty of Medicine, Campus of Bellaterra, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain;
Immunology Division and
Internal Medicine Division, University Hospital "Germans Trias i Pujol," Badalona (Barcelona), Spain; and
§
Immunology Division, Research Centre, Almirall-Prodesfarma SA, Barcelona, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In experiments aimed at cloning autoreactive T cells from tissues affected by autoimmune diseases, we (M. Catálfamo, unpublished observation), and others (7), had difficulties in eliciting clear responses to autoantigens both from the initial bulk cultures or from "autoreactive" clones. One possibility is that autoimmune responses in human disease are driven by T cells bearing low-avidity TCRs. This would also imply that most T lymphocytes bearing high-avidity TCRs to peripheral Ags are eliminated from the repertoire, and this process would occur mainly in the thymus. These results led us to examine whether peripheral and sequestered self-Ags (in general and in particular self-Ag targets of autoimmune responses, i.e., autoantigens) are produced in situ by a population of thymic cells. This question has not been experimentally addressed until recently and, when investigated, it has only focused on a few Ags.
Preliminary experiments (8, 9, 10) have detected pancreatic islet and thyroid self-Ags in human thymi. Based on these results, we decided to examine the more general issue of whether expression of peripheral self-Ags in the thymus is a normal process. This was addressed by determining the thymic expression of a panel of self-Ags whose access to the thymus is markedly different. These include: 1) self-Ags expressed constitutively in all cell types: ß-actin and glyceraldehyde phosphate dehydrogenase (GAPDH)3; 2) self-Ags of restricted tissue expression but present in the circulation at high (albumin), medium (thyroglobulin (Tg)) and low (glucagon and insulin) levels; 3) self-Ags of restricted tissue expression that are undetectable in the circulation: thyroid peroxidase (TPO), glutamic acid decarboxylase 65 (GAD65), and glutamic acid decarboxylase 67 (GAD67); and 4) classical sequestered Ags: myelin basic protein (MBP) and retinal S Ag (Ret S Ag). The H-Y Ag, which is only expressed in male tissues and generates central tolerance, provides an additional control (11). Some of the self-Ags selected for this study are targets of the autoimmune response leading to disease, i.e., GAD65, GAD67, and insulin of type I diabetes; MBP of multiple sclerosis; and TPO and Tg of autoimmune thyroiditis.
If predisposition to autoimmune disease is linked to the frequency of precursors and this is affected by thymic expression of self-Ags, disease predisposition would be influenced by the levels of self-Ag expression in the thymus, a possibility already suggested by Pugliese et al. (10) and Vafiadis et al. (9) for insulin. Therefore, it would be important to determine whether individual variability in thymic self-Ag expression is a common event.
Finally, accepting the current view that the diversity of the T cell repertoire is shaped during the initial years of life to reach a plateau around puberty (12), one would expect self-Ag expression to decrease with age. We investigated whether there is a relationship between self-Ag expression and the age of the thymic gland donors.
Given the large number of self-Ags, we predicted that the amount of self-Ags expressed in the thymus would have to be small. However, it is known that small Ag levels that are only detectable as transcripts by RT-PCR can be tolerogenic (13). We have used this technique to detect self-Ags as transcripts and, subsequently, to demonstrate the presence of their corresponding protein from cellular fractions rich in peripheral Ags.
Our results show that most peripheral and sequestered Ags are indeed expressed in normal human thymus with a tendency to decrease with age and with marked interindividual differences. Ags circulating at medium or even high levels, such as albumin and Tg, are also expressed in the thymus, indicating that endogenous synthesis is probably a requirement for self-Ags to induce tolerance in the thymus.
The above questions are crucial not only to understand tolerance but also the pathogenesis of organ specific autoimmune diseases, e.g., insulin-dependent diabetes mellitus and multiple sclerosis, in which the main defect is a failure in the maintenance of tolerance to peripheral self-Ags.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Samples of thymic tissue from 16 patients were obtained in the course of routine thoracic surgery on children with congenital heart abnormalities. The age of the donors ranged from 8 days to 13 years. Blocks (0.5 cm) were snap frozen and kept at -70°C until used. Cryostat sections were stained with hematoxilin and eosin or toluidine blue and were examined under the microscope to exclude the presence of extensive areas of adipose tissue and to ensure that both cortex and medulla tissue were present.
Tissue samples from pancreas, thyroid, liver, esophagus, stomach, and adrenal glands used as controls were obtained from cadaveric organ donors. Human brain RNA was provided by Clontech (Palo Alto, CA). Rat eyes were used as source of retinal RNA. The human cell lines M1 (fibroblasts) and U937 (monocytes) (American Type Culture Collection, Manassas, VA), and PBL were used as additional controls.
Thymus cell fractionation protocol is summarized in Fig. 1
. Thymic tissue (
4 g) was minced into
very small fragments with sharp scissors and resuspended in 25 ml of
RPMI 1640 culture medium (BioWhittaker, Walkersville, MD) supplemented
with 100 IU/ml penicillin (Normon, Madrid, Spain), 40 µg/ml
gentamicin (Normon), 2 mg/ml collagenase (type P, sp. act. 2.7 U/mg;
Boehringer Mannheim, Mannheim, Germany), and 0.05 mg/ml DNase I (Sigma,
St. Louis, MO) and was digested at 37°C under continuous stirring for
25 min. After removing the supernatant-containing cells, the process
was repeated twice.
|
Purification of RNA, cDNA synthesis, sample normalization, RT-PCR, and assessment of primer sensitivity
RNA was prepared following Chomczynski’s technique (15). Frozen samples were homogenized in lysis buffer consisting of guanidium thiocyanate (4 M; Serva, Heidelberg, Germany), N-lauroyl-sarcosine (0.5% w/v; Serva), Na citrate, pH 4 (50 mM; Sigma), and 2-ME (5%; Sigma) using an homogenizer (T25 Ultra-Turrax, IKA Labortechnik, Germany). To prevent potential cross-contamination of samples, the homogenizer probe was repeatedly rinsed in 2 M NaOH and sonicated between samples. To reduce sampling error, RNA was extracted from several blocks from each organ after histology assessment (see above). Pancreatic and liver tissue were pulverized in liquid nitrogen before RNA extraction. Each RNA sample was quantified in a spectrophotometer and its integrity tested by electrophoresis in 2% agarose and ethidium bromide staining. To avoid interference by contaminating genomic DNA, all samples were treated with DNase I (15 U; Pharmacia Biotech), DTT (5 mM; BRL, Gaithersburg, MD), RNaseOUT Recombinant Ribonuclease Inhibitor (20 U; BRL), Glycogen (40 mg; Boehringer Mannheim), and First Strand Buffer (50 mM Tris-HCl, 75 mM KCl, 3 mM MgCl2, pH 8.3; BRL) in a volume of 40 µl at 37°C for 30 min. After treatment, the RNA was precipitated, quantified, and tested for integrity as above.
For reverse transcription, 1 µg of total RNA (denatured by heating for 5 min at 68°C) was incubated with SuperScript II Reverse Transcriptase (100 U; BRL), RNaseOUT (20 U), dNTPs (1 mM final concentration; Pharmacia Biotech), Oligo d(T)20 (5 µM; Genset, Paris, France), and First Strand Buffer in a volume of 20 µl at 37°C for 1 h. The reaction product was heated to 95°C for 5 min and cooled on ice.
All cDNA samples were normalized using GAPDH expression as reference. Twenty microliters the cDNA reaction mix was diluted with ddH2O to 100 µl and serial dilutions (1/10, 1/20, and 1/40) were prepared. Two microliters from each cDNA dilution was amplified for GAPDH at 22 PCR cycles (before the reaction reached the amplification plateau). The protocol used for this PCR reaction was similar to that used for the amplification of self-Ag cDNA (see below). Dilutions that gave bands of similar intensity in ethidium bromide-stained gel were selected for subsequent PCR experiments.
PCR was performed in 20 µl with Dynazyme II DNA polymerase (0.4 U;
Finnzymes Oy, Espoo, Finland), DZ buffer (10 mM Tris-HCl, pH 8.8, 1.5
mM MgCl2, 50 mM KCl, 0.1% Triton X-100), dNTPs (0.2 mM;
Pharmacia Biotech) and primers (1 µM final) designed to span at least
one intron (Genset; sequences in Table I). PCR was performed in a thermal cycler
(Perkin-Elmer Cetus 480, Emerville, CA) with a 40-s denaturation step
at 94°C, a 40-s annealing step at different temperatures depending on
the primers (melting temperatures, tm, are
listed in Table I), a 20-s extension step at 72°C for 35
cycles, and a final extension of 7 min at 72°C. One sample containing
all reagents without cDNA was included in every run as contamination
control. Product specificity was confirmed by oligoprobe hybridization:
12 µl of the amplified products were subjected to electrophoresis in
2% agarose and transferred to nylon membranes (Hybond, Amersham, U.K.)
in 20x SSC. Membranes were rinsed for 5 min in denaturation solution
(1.5 M NaCl, 0.5 M NaOH) and for 1 min in neutralization solution (1.5
M NaCl, 0.5 M Tris-HCl (pH 7.2), 0.001 M EDTA), and the DNA were UV
cross-linked. The membranes were prehybridized directly in 2x SSC, 5x
Denhardt’s, 1% SDS, and salmon ssDNA for 1 h at hybridization
temperature (thy; thy =
tm - 15°C; see Table I). Hybridization
was conducted for 3 h at the same temperature and with the same
solution to which 1 x 106 cpm/ml of purified
radiolabeled oligoprobe had been added. The oligoprobes were labeled
with [
-32P]-dATP (Amersham) using T4 polynucleotide
kinase (New England Biolabs, Beverly, MA). Oligoprobes were designed in
our laboratory, and their sequences are shown in Table I. Filters were
washed with tetramethylammonium chloride (TMAC) washing solution (25 mM
Tris, pH 8.5, 2 mM EDTA, 1% SDS, and 3 M TMAC; Sigma). This solution
obviates the need for adjusting the washing temperature according to
the Cytosine-Guanine content of the probe so that it only depends on
the length of the hybrid (16). The 21-mer oligoprobes were washed at
61°C and the 18-mer oligoprobes at 57°C. Autoradiography was
performed at -70°C using Cronex x-ray film (DuPont, Boston,
MA).
|
Northern blot hybridization
Total RNA was extracted as described above. Poly(A)+
RNA was purified using Dynabeads Oligo (dT)25 (Dynal, Oslo,
Norway) according to the manufacturer’s instructions. Ten micrograms
of poly(A)+ RNA were subjected to electrophoresis in
formaldehyde agarose gels, transferred to nylon membranes (Hybond,
Amersham) in 20x SSC, and UV cross-linked. Membranes were hybridized
as described (17); briefly, prehybridization was performed in 0.25 M
Na2HPO4, 1% BSA, 1 mM EDTA, and 7% SDS for
2 h at 68°C. Hybridization was conducted for 16–20 h at the
same temperature and with the same solution to which 1.5 x
106 cpm/ml of purified radiolabeled probe had been added.
The probes were labeled with [
-32P]-dCTP (Amersham) by
random priming using the oligolabeling kit (Pharmacia Biotech). The
probes used (412 bp-ß-actin, 338 bp-albumin, and 257 bp-Tg) were
generated by PCR in our laboratory. The filters were washed in 20 mM
Na2HPO4, 1 mM EDTA, and 1% SDS at 68°C for
30 min. Autoradiographs were performed at -70°C using Cronex x-ray
film (DuPont).
Immunofluorescence staining
Cytospins from dispersed cell preparations of the different thymic fractions were first fixed in acetone at -20°C for 2 min, air-dried for 30 min, and then stained with mAbs to cytokeratin (Ck) 8/18 low m.w. (NCL5D3; Novocastra Laboratories, U.K.), CD2 (OKT11, European Cell Culture Collection, Witshire, U.K.), CD14 (47–3Q6, provided by Dr. R. Vilella, Hospital Clínic Provincial, Barcelona, Spain), CD68 (CD68-EBM11, Dako, Denmark), and with a control mAb (IA3, provided by Dr. R. Vilella) followed by FITC goat anti-mouse serum (Southern Biotechnology, Birmingham, AL). Other cytospins were stained with rabbit anti-S100 (18) serum (Dako), followed by tetramethylrhodamine B isothiocyanate (TRITC) goat anti-rabbit Ig (Dako), mAb anti-HLA Class II (EDU-1, provided by Dr. R. Vilella), and FITC goat anti-mouse IgG (Southern Biotechnology). Separate control cytospin slides were stained following the same protocol but omitting one of the specific Abs in each. Preparations were examined under phase contrast and UV illumination by two independent observers (X.F.-F. and R.P.-B.) under a Zeiss Axioplan UV microscope (Zeiss, Oberkochen, Germany). The results are given as a percentage of positive cells.
Statistical analysis
For statistical analysis transcription levels were classified into two categories: positive (including weak positives) and negative. Results were compared in binomial tables; the paired or unpaired test was applied as indicated. Values of p < 0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The aim of this work was to demonstrate transcription of self-nonthymic genes in the thymus by means of RT-PCR. Precautions to ensure reliability, avoid artifacts, and enable interexperiment comparisons were taken even if precise quantification was not practical or feasible. These precautions were: 1) extraction of the RNA from a number of randomly selected thymic blocks representative of the whole gland; 2) elimination of genomic DNA interference by treating samples with DNase and by using primers spanning intronic regions; 3) normalization of the cDNA input into the PCR reactions according to the expression of GAPDH; 4) confirmation of the specificity of the amplification by hybridization with a complementary oligoprobe; 5) conversion of autoradiograph OD values to number of template copies by using the results from the primer titration as reference and the OD of their positive control to calculate a correction factor for autoradiograph exposure time; 6) definition of a semiquantitative scale for the transcription level of specific genes (autoradiograph OD values corresponding to less than 10 copies per aliquot of cDNA normalized for GAPDH and PCR amplified were considered negative, values corresponding to a number of copies between 10 and 100 were weakly positive, and values corresponding to more than 100 copies were positive); and 7) all amplification experiments have been conducted at least twice. Experiments assessing self-Ag expression in thymic cell fractions were repeated four times. Results have always been consistent.
In summary, the normalization for GAPDH makes it possible to compare RT-PCR results for each given self-Ag in different tissue samples, while the definition of transcription levels allows the comparison of results from different self-Ags in each given tissue sample.
Normal human thymic glands contain transcripts from a broad spectrum of self-Ags
As expected, all thymic glands contained transcripts for the
control constitutive house-keeping genes ß-actin and GAPDH. Because
cDNAs were normalized for PCR loading according to GAPDH, not
surprisingly its expression level was similar in the 12 thymic glands.
ß-actin transcripts were also present in all glands while H-Y Ags
were predictably expressed only in male thymi. The relative expression
levels follow the hierarchy actin > GAPDH > H-Y. GAPDH
transcripts were more abundant than average peripheral Ag transcripts.
The amount of H-Y transcripts was similar to that of Tg, glucagon, and
albumin as assessed by direct inspection of gels under UV (Fig. 2). Hybridization results also
indicated that the H-Y transcription levels were, in many instances,
within the same order of magnitude as that of some peripheral Ag
transcripts (Fig. 3A).
|
|
A). The circulating
levels bore some relation with the level of thymic transcription.
Albumin, which circulates at 50 g/L, was detected in 12 out of 12
thymi, while insulin, glucagon, and Tg, which circulate at 500–5 ng/L
(107–109 lower levels than albumin), were
detected in 5, 6, and 6 out of 12 glands, respectively. Noncirculating
peripheral Ags, TPO, GAD67, and GAD65, were detected in a similar
proportion of glands: 6, 6, and 1, respectively. Sequestered Ags, MBP
and Ret S Ag, were detected in 3 out 12 glands each. These results are
assigned to one of three categories according to the defined levels of
transcripts: negative, weak positive, and positive (Fig. 3C). However, for statistical analysis the results were
considered as either positive (including weak positives) or negative,
and the peripheral Ags were clustered in groups, i.e., circulating,
noncirculating, and sequestered. The analysis showed significant
differences between circulating and noncirculating Ags, 29/48 (60%) vs
13/36 (36%) (
2 (1) test, p = 0.03), and
between circulating and sequestered Ags, 29/48 (60%) vs 6/24 (25%)
(2 (1) test, p = 0.007). Organ-specific Ags are transcribed in more tissues than expected but are not ubiquitous
To determine the tissue specificity of the peripheral Ags studied,
we analyzed the transcription of these genes in a panel of control
samples. We found that most of the Ags presented the expected tissue
distribution (Fig. 4A). TPO
transcripts were clearly detected only in thyroid; glucagon in
pancreas; GAD65 in pancreas and brain; Ret S Ag in retina; and MBP
transcripts were clearly amplified in brain but small amounts were
detected also in the thyroid and in the cell line U937. In contrast,
transcripts for albumin, Tg, insulin, and GAD67 were detected in more
tissues than expected but were not ubiquitous, and their expression
remained confined to embriologically related tissues. Results are
summarized in Fig. 4B. The cell lines expressed some
peripheral autoantigens, but this is not totally unexpected in
transformed cells.
|
Because the level of transcription, as assessed by RT-PCR, was
appreciable for albumin and Tg, we tried to demonstrate their
presence in six thymic gland samples by a Northern blot with 10
µg of poly(A)+ RNA. Results were still negative after 21
days exposure (Fig. 5). RT-PCR results
were confirmed in that albumin transcripts were detected in control
pancreatic tissue mRNA.
|
To investigate whether transcription of peripheral self-Ags in the
thymus is related to age, we included in the study samples from a range
of ages. Overall, the expression of peripheral Ags in the thymus tended
to decrease with age as it can be appreciated in the checkerboard
graphic (Fig. 3C). For statistical analysis, both weak and
clear positive results were considered positive, and donors were
stratified by age. The difference was significant; in thymi from
<2-yr-old donors, 32/54 (80%) PCRs were positive vs 16/54 (30%) in
those from >2-yr-old donors (nonpaired 2 (1) test,
p = 0.004). For some Ags, e.g., glucagon and insulin,
there are strong discrepancies among glands from donors of the same
age.
Identification of thymic cell fraction(s) that contain peripheral Ag transcripts
To identify the cell population actively transcribing peripheral Ag genes in the thymus, we analyzed the levels of their transcripts in different cellular fractions from four enzymaticly dispersed glands (TMB31, TMB33, TMB38, and TMB39). Cells were characterized by flow cytometry and cytosmear staining with mAbs to markers for the three most abundant thymus cell types: thymocytes (CD2), stromal epithelial cells (8/18 Ck), and macrophages (labeled with a mixture of mAbs to CD14 and CD68). Dendritic cells were identified by double IFL staining with mAb to HLA Class II and S-100 on cytosmears. The initial cell preparation was found so heterogeneous that the aim of the fractionation protocol had to be limited to generate fractions that differed markedly in cell composition among themselves and with respect to the starting material rather than to generate pure populations. This protocol combined with the assessment of peripheral Ag transcription level in each fraction would make it possible, by a cosegregation analysis, to assign given self-Ag transcripts to particular cell populations.
We succeeded in generating fractions of very different cellular
composition. As shown in Fig. 1B and
6a, fraction 1 was in average
100-fold enriched in Ck-positive epithelial cells when compared
with fraction 4 (thymocytes); fraction 3 was 16-fold enriched in
macrophages and around fourfold in dendritic cells when compared with
fraction 4 (thymocytes); and, finally, fraction 4 is fourfold enriched
in CD2+ thymocytes with respect to fraction 1. Fraction 2
is the preparation from which fraction 3 and 4 were derived and,
therefore, contains a mixed population. It was included in the analysis
to check the consistency of the results. Fig. 6, a and
b, show representative IFL staining of cytosmears and flow
cytometry.
|
) showed that
fraction 1, which has the highest content of epithelial cells, also
contained the highest level of peripheral self-Ag transcripts (23/32
RT-PCRs). Fraction 4 (thymocytes) and fraction 3 (enriched in
macrophages) only occasionally showed high levels of peripheral Ag
transcripts (2/32 for both) (paired 2 test,
p < 0.001). The expression of peripheral Ag bore no
relation with the distribution of dendritic cells.
|
illustrates and summarizes the
results. Their validity was supported by the finding of a similar
number of copies of GAPDH and H-Y per cell in all cellular fractions.
Given the number of steps and the variability of some of the techniques
and the unrealistic assumption of a 100% efficiency of mRNA
retrotranscription, these results should be taken as approximate and
valid only to determine the minimal order of magnitude of the copy
number. Taking into account all these shortcomings we consider that
these estimations are useful in the interpretation of the raw data. Our
results suggest that epithelial cells are responsible for peripheral Ag
expression in the thymus.
|
.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tolerance to peripheral and sequestered Ags is certainly not absolute as demonstrated by cell cloning experiments showing the presence of circulating T cells capable of recognizing peripheral self-Ags such as thyroid peroxidase (TPO) (22), collagen (23, 24), and even the sequestered Ag MBP (25, 26). Models of autoimmune diseases that are induced by the immunization of animals with self-Ags in adjuvants also prove the availability of autoreactive T cell clones in the normal repertoire (27, 28, 29, 30). Because it seemed unlikely that the thymus could harbor a representation of the whole spectrum of Ags expressed in all the tissues of the organism, it has been assumed that tolerance to Ags of restricted tissue distribution would be dependent on peripheral tolerance. Experiments conducted using transgenic mice expressing as transgenes a variety of nonself-Ags under the control of tissue specific promoters supported this view (31), and tolerance was postulated to be the result of clonal anergy and/or ignorance (32). The two-signal paradigm of lymphocytic activation (33, 34, 35) has been invoked to explain the induction of T cell anergy in the periphery, i.e., presentation of Ag by parenchimatous cells expressing MHC but not costimulatory molecules would result in clonal anergy.
Miller et al. (36, 37) pointed out that the first experiments with transgenic mice, which had suggested a dominant role for peripheral mechanisms in establishing tolerance, may have been misleading because the expression in the thymus of transgenes coding for peripheral Ags had not been checked. They and other authors (38) had results demonstrating thymic expression of such "tissue specific transgenes." The implication is that the tolerance to peripheral Ags observed in the initial set of transgenic mice was "central" tolerance. Recent experiments (4, 5) have provided direct evidence of the critical importance of precursor frequency of autoreactive T cells in the development of autoimmunity. These results call for a reevaluation of the role of central mechanisms in the maintenance of tolerance to peripheral Ags.
In this report, we provide evidence for the expression of a broad variety of nonthymic self-Ags in human thymic glands from different age donors from a few days of life to 13 years. Our results are in agreement with a number of recent reports of the unexpected detection of some peripheral Ags in human (9, 10, 39, 40, 41) and mice thymus (42, 43, 44, 45). This is the first study in which the question of whether this is a general phenomenon has been addressed by the systematic investigation of the expression of a wide variety of peripheral self-Ags in the human thymus. The results presented here indicate that human thymus does express the full spectrum of self-Ags. However, we cannot rule out that some of the neuroendocrine self-Ags detected may actually have a regulatory function in the thymus as it was suggested in the past (46, 47). Nevertheless, it is difficult to envisage a nonimmunologic function for all the self-Ags expressed in the thymus. A similar view was proposed years ago under the name of the "peptone hypothesis" (48), according to which epitopes representative of all self-Ags are expressed in the thymus to shape the T cell repertoire.
A shortcoming of this and previous studies is that peripheral Ag expression in the thymus has been assessed at the mRNA level using RT-PCR, whereas the presence of the corresponding protein has only been confirmed in a few cases. However, there are several arguments that uphold that most of these transcripts are translated into functional proteins in small but functionally relevant quantities. i) The abundance of transcripts (in the order of 1–100 copies per epithelial cell) is 2–4 orders of magnitude above the abundance attributable to "illegitimate transcription," i.e., the minimal spontaneous transcription of genes that occur at random, which is 1 copy per 100-1000 cells; amplification of those illegitimate transcripts always requires two rounds of PCR using nested primers (49, 50). The levels of transcripts we detected in some thymi were in some instances like Tg and albumin, comparable to the levels of H-Y transcripts, an Ag whose expression is known to induce central tolerance. ii) When self-Ag expression was investigated in the different thymic cell fractions, the transcripts were preferentially distributed in the fractions most enriched in epithelial cells, a candidate for T cell selection (51). Furthermore, by staining cytosmears of dispersed thymic cells enriched in epithelial cells we have detected cells that contain insulin in their cytoplasm. Further work is required to determine the cell type and confirm that the cells that stained for insulin are those that contain the transcripts detected by RT-PCR.
Based on the above arguments, we consider that self-Ag expression in the thymus is real and functional in terms of being capable of influencing thymic selection.
The abundance of self-Ags and their distribution in the organism seem to bear some relation with their level of expression in the thymus. Albumin, the more abundant self-Ag investigated, was also the only peripheral Ag whose transcripts were present in all thymic glands. From our results, we can envisage that the degree of thymus expression follows the hierarchy: peripheral self-Ags circulating at high and medium levels > noncirculating peripheral self-Ags > sequestered Ags.
Regarding interindividual variation in the expression of self-Ags, we
did not find any in actin, a bona fide constitutive gene, nor in H-Y, a
constitutive gene but expressed only in glands from male donors. In
contrast, all the other self-Ags showed strong expression differences
from thymus to thymus (see Fig. 3C). The overall expression
of self-Ags is dependent on age as demonstrated by the differences
found when the number of peripheral self-Ags expressed in the groups of
<2 years and >2 years were compared. This was expected because thymic
function is known to decline with age. This cannot be explained by the
progressive replacement of the thymus parenchyma by adipose tissue
because histology of the tissue samples had been carefully examined.
However, for some Ags there was an age-independent variability, e.g.,
insulin and TPO. Recent data showing that insulin expression in the
thymus is variable and related to allelic variations in the 5' upper
regulatory region of the gene (9, 10) are relevant to this point. This
type of polymorphism may exert an independent effect on the level of
transcription and it may be also operative in other autoantigen genes,
some of which also have VNTRs 5' to their coding region, e.g., ICA69
(52).
The function of this broad range of thymic peripheral self-Ags is not likely to relate to positive selection, driven by low affinity interactions. Therefore, a limited number of ligands should be able to select a large repertoire of TCRs. In fact, it has been recently demonstrated that a single peptide is able to select a relatively ample repertoire of T cells in the thymus (53). The generation of this limited number of ligands could easily rely on the presentation of classical thymus-derived ubiquitous self-Ags. It is therefore likely that the expression of peripheral self-Ags in the thymus is related to negative selection. Negative selection involves high-affinity interactions and would require the same array of peptides that cells can encounter in the periphery. It is unlikely that the thymic expression of self-Ags such as albumin, Tg, and even insulin (all of them circulating Ags that are available to standard APCs via endocytosis) would be required for the negative selection of class II-restricted CD4+ T cells. Nevertheless, expression of these self-Ags may be crucial for deleting class I-restricted CD8+ T lymphocytes. Peptides recognized by these cells are derived from endogenously synthesized proteins that bind to HLA class I molecules in the endoplasmic reticulum after being processed through the class I pathway. Endocytosed Ags can also enter the class I pathway (cross-presentation) (54, 55, 56), but probably this does not occur in the thymus because it could result in the fortuitous induction of tolerance to circulating proteins from pathogens. Thus, there is need to synthesize most peripheral self-Ags, both tissue-bound and circulating, to delete T cells bearing high-affinity autoreactive TCR. Another possibility, which does not exclude the former, is that the expression of peripheral Ags in the thymus serves for the positive selection of regulatory T cells responsible for the maintenance of tolerance to peripheral Ags as demonstrated in autoimmune diabetes (57) or for eye Ags (58).
The final identification of the cell type expressing the spectrum of peripheral self-Ags was not expected from the cell fractionation experiments because the cellular complexity of the human thymus precludes such a simple approach to this question. We tried to rule out some populations such as thymocytes and start defining the cell lineages expressing the self-Ags. The results seem at first difficult to interpret because transcripts of some self-Ag are present in the thymocyte-enriched population isolated by N-SRBC rosetting, a poor candidate for the induction of negative selection. However, when all the data are taken into consideration it becomes obvious that the fraction with the highest proportion of epithelial stromal cells is also the fraction with the highest levels of peripheral autoantigen transcripts. Moreover, the calculations that allowed us to estimate (within an order of magnitude) the number of mRNA molecules per a given peripheral self-Ag per cell also pointed to the stromal epithelial cells as those expressing peripheral Ag. Thymic epithelial cells are a complex population made of different subpopulations that have been classified into six categories (59). Interestingly, a recent report describes the existence of "rare" peripheral Ag-expressing cell (PAE) in mice thymus (44). Grafting thymi containing such a population to nude mice resulted in tolerance to the Ag expressed by the peripheral Ag-expressing cell.
From the results presented herein, we conclude that peripheral self-Ags are expressed in normal thymic glands, by one of the cell types present in the fraction enriched in stromal epithelial cells. This fulfills an essential premise for postulating a role for the thymus in the establishment of tolerance to peripheral self-Ags. It remains to be proven that this expression is immunologically relevant in the induction of tolerance. It also raises the question of whether thymic malfunction is a necessary, but not sufficient, condition for the development of organ specific autoimmune diseases such as type 1 diabetes and multiple sclerosis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Ricardo Pujol-Borrell, Immunology Division, University Hospital "Germans Trias i Pujol," Ctra. del Canyet s/n, 08916 Badalona (Barcelona), Spain. E-mail address:
3 Abbreviations used in this paper: GAPDH, glyceraldehyde phosphate dehydrogenase; GAD, glutamic acid decarboxylase; Tg, thyroglobulin; TPO, thyroid peroxidase; MBP, myelin basic protein; Ret S Ag, retinal S antigen; N-SRBC, neuraminidase-treated SRBC; Ck, cytokeratin; IFL, immunofluorescence.
Received for publication April 23, 1998. Accepted for publication July 27, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. V. Misharin, Y. Nagayama, H. A. Aliesky, B. Rapoport, and S. M. McLachlan Studies in Mice Deficient for the Autoimmune Regulator (Aire) and Transgenic for the Thyrotropin Receptor Reveal a Role for Aire in Tolerance for Thyroid Autoantigens Endocrinology, June 1, 2009; 150(6): 2948 - 2956. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Muixi, M. Carrascal, I. Alvarez, X. Daura, M. Marti, M. P. Armengol, C. Pinilla, J. Abian, R. Pujol-Borrell, and D. Jaraquemada Thyroglobulin Peptides Associate In Vivo to HLA-DR in Autoimmune Thyroid Glands J. Immunol., July 1, 2008; 181(1): 795 - 807. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Gavanescu, B. Kessler, H. Ploegh, C. Benoist, and D. Mathis Loss of Aire-dependent thymic expression of a peripheral tissue antigen renders it a target of autoimmunity PNAS, March 13, 2007; 104(11): 4583 - 4587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Narendran, A. M. Neale, B. H. Lee, K. Ngui, R. J. Steptoe, G. Morahan, O. Madsen, J. A. Dromey, K. P. Jensen, and L. C. Harrison Proinsulin is encoded by an RNA splice variant in human blood myeloid cells PNAS, October 31, 2006; 103(44): 16430 - 16435. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.M. de Yzaguirre, J.S. Hernandez, P.F. Navarro, P.L. Nieva, M. Herranz, M.F. Fraga, M. Esteller, A. Juarranz, and J. Fernandez-Piqueras Epigenetic silencing of E- and N-cadherins in the stroma of mouse thymic lymphomas Carcinogenesis, May 1, 2006; 27(5): 1081 - 1089. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kralovicova, T. R. Gaunt, S. Rodriguez, P. J. Wood, I. N.M. Day, and I. Vorechovsky Variants in the Human Insulin Gene That Affect Pre-mRNA Splicing: Is -23HphI a Functional Single Nucleotide Polymorphism at IDDM2? Diabetes, January 1, 2006; 55(1): 260 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. N. Pichurin, C.-R. Chen, G. D. Chazenbalk, H. Aliesky, N. Pham, B. Rapoport, and S. M. McLachlan Targeted Expression of the Human Thyrotropin Receptor A-Subunit to the Mouse Thyroid: Insight into Overcoming the Lack of Response to A-Subunit Adenovirus Immunization J. Immunol., January 1, 2006; 176(1): 668 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Garcia, K. R. Prabakar, J. Diez, Z. A. Cao, G. Allende, M. Zeller, R. Dogra, A. Mendez, E. Rosenkranz, U. Dahl, et al. Dendritic Cells in Human Thymus and Periphery Display a Proinsulin Epitope in a Transcription-Dependent, Capture-Independent Fashion J. Immunol., August 15, 2005; 175(4): 2111 - 2122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Takase, C.-R. Yu, R. M. Mahdi, D. C. Douek, G. B. DiRusso, F. M. Midgley, R. Dogra, G. Allende, E. Rosenkranz, A. Pugliese, et al. Thymic expression of peripheral tissue antigens in humans: a remarkable variability among individuals Int. Immunol., August 1, 2005; 17(8): 1131 - 1140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Liston, D. H.D. Gray, S. Lesage, A. L. Fletcher, J. Wilson, K. E. Webster, H. S. Scott, R. L. Boyd, L. Peltonen, and C. C. Goodnow Gene Dosage-limiting Role of Aire in Thymic Expression, Clonal Deletion, and Organ-specific Autoimmunity J. Exp. Med., October 18, 2004; 200(8): 1015 - 1026. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Gallegos and M. J. Bevan Central Tolerance to Tissue-specific Antigens Mediated by Direct and Indirect Antigen Presentation J. Exp. Med., October 18, 2004; 200(8): 1039 - 1049. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-K. Kim, K. V. Tarbell, M. Sanna, M. Vadeboncoeur, T. Warganich, M. Lee, M. Davis, and H. O. McDevitt Prevention of type I diabetes transfer by glutamic acid decarboxylase 65 peptide 206-220-specific T cells PNAS, September 28, 2004; 101(39): 14204 - 14209. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. I.M. Tree, G. Duinkerken, S. Willemen, R. R.P. de Vries, and B. O. Roep HLA-DQ-Regulated T-Cell Responses to Islet Cell Autoantigens Insulin and GAD65 Diabetes, July 1, 2004; 53(7): 1692 - 1699. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Gotter, B. Brors, M. Hergenhahn, and B. Kyewski Medullary Epithelial Cells of the Human Thymus Express a Highly Diverse Selection of Tissue-specific Genes Colocalized in Chromosomal Clusters J. Exp. Med., January 20, 2004; 199(2): 155 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Faideau, J.-P. Briand, C. Lotton, I. Tardivel, P. Halbout, J. Jami, J. F. Elliott, P. Krief, S. Muller, C. Boitard, et al. Expression of Preproinsulin-2 Gene Shapes the Immune Response to Preproinsulin in Normal Mice J. Immunol., January 1, 2004; 172(1): 25 - 33. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yan, T. Devos, L. Yu, G. Xia, O. Rutgeerts, J. Goebels, C. Segers, Y. Lin, M. Vandeputte, and M. Waer Pathogenesis of Autoimmunity After Xenogeneic Thymus Transplantation J. Immunol., June 15, 2003; 170(12): 5936 - 5946. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-P. Armengol, C. B. Cardoso-Schmidt, M. Fernandez, X. Ferrer, R. Pujol-Borrell, and M. Juan Chemokines Determine Local Lymphoneogenesis and a Reduction of Circulating CXCR4+ T and CCR7 B and T Lymphocytes in Thyroid Autoimmune Diseases J. Immunol., June 15, 2003; 170(12): 6320 - 6328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Friday, S. L. Pietropaolo, J. Profozich, M. Trucco, and M. Pietropaolo Alternative Core Promoters Regulate Tissue-specific Transcription from the Autoimmune Diabetes-related ICA1 (ICA69) Gene Locus J. Biol. Chem., January 3, 2003; 278(2): 853 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Murphy, M. Biondo, B.-H. Toh, and F. Alderuccio Tolerance established in autoimmune disease by mating or bone marrow transplantation that target autoantigen to thymus Int. Immunol., January 1, 2003; 15(2): 269 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Zheng, J.-X. Gao, H. Zhang, T. L. Geiger, Y. Liu, and P. Zheng Clonal Deletion of Simian Virus 40 Large T Antigen-Specific T Cells in the Transgenic Adenocarcinoma of Mouse Prostate Mice: An Important Role for Clonal Deletion in Shaping the Repertoire of T Cells Specific for Antigens Overexpressed in Solid Tumors J. Immunol., November 1, 2002; 169(9): 4761 - 4769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. S. Wong, A. K. Moustakas, L. Wen, G. K. Papadopoulos, and C. A. Janeway Jr. Analysis of structure and function relationships of an autoantigenic peptide of insulin bound to H-2Kd that stimulates CD8 T cells in insulin-dependent diabetes mellitus PNAS, April 16, 2002; 99(8): 5551 - 5556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. SALAMA, A. N. CHAUDHRY, J. J. RYAN, E. EREN, J. B. LEVY, C. D. PUSEY, L. LIGHTSTONE, and R. I. LECHLER In Goodpasture's Disease, CD4+ T Cells Escape Thymic Deletion and Are Reactive with the Autoantigen {alpha}3(IV)NC1 J. Am. Soc. Nephrol., September 1, 2001; 12(9): 1908 - 1915. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Armengol, M. Juan, A. Lucas-Martin, M. T. Fernandez-Figueras, D. Jaraquemada, T. Gallart, and R. Pujol-Borrell Thyroid Autoimmune Disease : Demonstration of Thyroid Antigen-Specific B Cells and Recombination-Activating Gene Expression in Chemokine-Containing Active Intrathyroidal Germinal Centers Am. J. Pathol., September 1, 2001; 159(3): 861 - 873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Whalen, J. Marounek, P. Weiser, M. C. Appel, D. L. Greiner, J. P. Mordes, and A. A. Rossini BB Rat Thymocytes Cultured in the Presence of Islets Lose Their Ability to Transfer Autoimmune Diabetes Diabetes, May 1, 2001; 50(5): 972 - 979. [Abstract] [Full Text]
|
|
|
|
|
|
J. Diez, Y. Park, M. Zeller, D. Brown, D. Garza, C. Ricordi, J. Hutton, G. S. Eisenbarth, and A. Pugliese Differential Splicing of the IA-2 mRNA in Pancreas and Lymphoid Organs as a Permissive Genetic Mechanism for Autoimmunity Against the IA-2 Type 1 Diabetes Autoantigen Diabetes, April 1, 2001; 50(4): 895 - 900. [Abstract] [Full Text]
|
|
|
|
 |
|
 P. Romagnani, F. Annunziato, E. Lazzeri, L. Cosmi, C. Beltrame, L. Lasagni, G. Galli, M. Francalanci, R. Manetti, F. Marra, et al. Interferon-inducible protein 10, monokine induced by interferon gamma, and interferon-inducible T-cell alpha chemoattractant are produced by thymic epithelial cells and attract T-cell receptor (TCR) {alpha}{beta}+CD8+ single-positive T cells, TCR{gamma}{delta}+ T cells, and natural killer-type cells in human thymus Blood, February 1, 2001; 97(3): 601 - 607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-M. Feng, I. M. Givogri, E. R. Bongarzone, C. Campagnoni, E. Jacobs, V. W. Handley, V. Schonmann, and A. T. Campagnoni Thymocytes Express the golli Products of the Myelin Basic Protein Gene and Levels of Expression Are Stage Dependent J. Immunol., November 15, 2000; 165(10): 5443 - 5450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Annunziato, P. Romagnani, L. Cosmi, C. Beltrame, B. H. Steiner, E. Lazzeri, C. J. Raport, G. Galli, R. Manetti, C. Mavilia, et al. Macrophage-Derived Chemokine and EBI1-Ligand Chemokine Attract Human Thymocytes in Different Stage of Development and Are Produced by Distinct Subsets of Medullary Epithelial Cells: Possible Implications for Negative Selection J. Immunol., July 1, 2000; 165(1): 238 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Aiello, M. Noris, G. Piccinini, S. Tomasoni, F. Casiraghi, S. Bonazzola, M. Mister, M. H. Sayegh, and G. Remuzzi Thymic Dendritic Cells Express Inducible Nitric Oxide Synthase and Generate Nitric Oxide in Response to Self- and Alloantigens J. Immunol., May 1, 2000; 164(9): 4649 - 4658. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. S. Wong, A. K. Moustakas, L. Wen, G. K. Papadopoulos, and C. A. Janeway Jr. Analysis of structure and function relationships of an autoantigenic peptide of insulin bound to H-2Kd that stimulates CD8 T cells in insulin-dependent diabetes mellitus PNAS, April 16, 2002; 99(8): 5551 - 5556. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
