Reprints Available Directly from the Publisher Photocopying Permitted by License Only Differential Effects of X-irradiation and Cyclosporin-a Generation of Cyclosporin-a-induced Autoimmunity

Cyclosporin A (CsA), a potent inhibitor of T-cell activation, has been shown to have several effects on thymocyte maturation, thymic stromal cells, and the generation of autoreactive T cells. In Lewis rats, the combination of lethal irradiation, syngeneic bone marrow transplantation, and a 4-week course of CsA administration causes the development of an autoimmune disease (CsA-AI) resembling allogeneic graft-versus-host disease. This occurs upon withdrawal of CsA, provided the thymus receives irradiation and is present during CsA treatment. In this study, the separate effects of irradiation or CsA treatment on thymic stromal cells and thymocytes, compared to the combinatory effects, were examined using immunohistochemistry and tricolor flow cytometric analysis. CsA treatment causes an involution of the thymic medulla and a strong reduction of the cell number of thymocytes and stromal cells residing in the medulla. However, within the remaining medullary area, changes in cell distribution and antigen density on these cells were not observed. Irradiation on the other hand causes a strong depletion of thymocytes. The thymocyte population is recovered within 2 weeks and a cortical and medullary region can be distinguished. CsA treatment in combination with irradiation results in a strongly inhibited recovery of the medulla during CsA treatment, whereas the cortex recovers to normal size and morphology. The presence of the medullary IDC and epithelial cells is reduced proportionally to the small size of the medulla. However, the distribution of these stromal cells is normal. During the CsA administration, the thymuses from irradiated and CsA-treated rats are very similar to thymuses from CsA-treated rats. In conclusion, no changes specific for irradiation plus CsA treatment have been observed. Regarding the distribution and size of medullary stromal cells and residing thymocytes, thymuses from irradiated and CsA-treated rats hardly differ from the thymuses from rats treated only with CsA. Therefore, irradiation seems essential in the generation of CsA-AI by eliminating suppressor-cell circuits in the periphery.


ne mar- row t
ansplantation (BMT).Although CsA is also very useful in inhibiting the progression of several autoimmune diseases, surprisingly, autoimmune phenomena may rise upon withdrawal of CsA treatment under certain conditions.Rats treated with CsA after lethal irradiation and consecutive syngeneic BMT develop GVHD-like lesions after CsA administration is stopped (Glazier et al., 1983); in humans, too, similar lesions have been reported after autologous BMT  (Jones et al., 1989).The crucial role of the thymus in the induction of this CsA-induced autoimmunity (CsA-AI) has unequiv- ocally been established (Sorokin et al., 1986; Wodzig  et al., 1993).

Immunosuppression with CsA not only effec- tively blocks the proliferation of mature peripheral T cells, but also blocks thymocyte maturation (re- viewed by Ritter and Boyd, 1993).Maturation of cortical CD4 + CD8 + (double-positive) thymocytes 128 L.J.J. BEIJLEVELD et al.   into medullary CD4 + and CD8 + (single-p sitive) thymocytes is strongly inhibited (Gao et al., 1988;  Jenkins et al., 1988; Kanariou et al., 1989).Immu- nohistochemical analysis of CsA-treated thymuses from rodents shows a relatively small medulla, in which the loss of medullary thymocytes is accom- panied by a loss of medullary stromal cells (Tanaka   et al., 1988; Schuurman et al., 1990; Ritter and   Ladyman, 1991) and formation of large areas de- nuded of MHC class-II positive epithelial cells (Bruijntjes et al., 1993).The medullary stromal cells, consisting of epithelial cells, macrophages (M)), and interdigitating cells (IDC), play an important role in negative selection of autoreactive cells.The absence or functional alteration of medullary stromal cells has been suggested to lead to the escape of possible autoreactive cells to the periphery and the development of autoimmunity (Schuurman et al., 1990;  Beschorner and Armas, 1991a; De Waal et al., 1992).

Although upon CsA administration in the rat autoreactive T cells are generated (Beschorner et al.,  1988b) and in the mouse the peripheral occurrence of T cells bearing self-reactive T-cell receptors, as determined y the use of mAbs recognizing specific "forbidden Vfs", has been demonstrated (Jenkins et   al., 1988), only the combination of CsA and irradi- ation evokes CsA-AI in the rat.Lethal irradiation, in terms of complete bone marrow depletion, has numerous effects on the immune system, but syngeneic BMT normally leads to an almost complete recovery of the thymus (Mulder and Visser, 1988;  Mulder et al., 1988; Murawska et al., 1991).Irradi- ation wipes out almost the whole thymocyte popu- lation, whereas the thymic stromal cells are relatively radio-resistant.Shielding of the thymus during irradiation prevents disease both in rat (Beschorner et al., 1987) and in mouse (Cheney and   Sprent, 1985).It has been suggested that both whole-body irradiation, including the thymus and the CsA treatment have effects on the thymus essential for the induction of CsA-AI.Alternatively, post-CsA prevention of thymic recovery by thymec- tomy in unirradiated rats consistently evokes CsA-AI (Beschorner et al., 1991b).This implies that the effect of irradiation can be replaced by thymectomy after the CsA treatment, suggesting that a normal thymic microenvironment after CsA is necessary to reestablish self-tolerance through autoregulatory cells.

In our rat model, the combination of CsA and irradiation evokes CsA-AI, so this study was de- signed to determine the differential effects of CsA and irradiation on the thymus, as compared to the combinatory effects.The attention was focused on (1) the presence and immunohistochemical proper- ties of the thymic stroma with special reference to the medullary IDC and epithelial cells and (2) the number and phenotype of thymocytes during and after CsA treatment.Furthermore, the disease- inducing effects of the separate treatments in com- bination with thymectomies were determined.


RESULTS


Thymus Morphology and Thymocyte Number

The thymus was morphologically assessed for the relative size of the medulla and for thymocyte density

ithin th
cortex and medulla with hema- toxylin and eosin sections.The thymocyte number was determined by counting single-cell suspensions in a Birker hemocytometer.

Total body irradiation with 8.5 Gy caused a collapse of the thymic architecture and a strong reduction of thymocytes within 48 hr.Whereas the medulla eemed only slightly affected, the cortical thymocytes had completely disappeared.One week after irradiation, there was hardly any cortex or medulla detectable, but as soon as 2 weeks after irradiation, a new cortex was formed together with a small medulla.

Two days after irradiation, the total number and viability of thymocytes were very low.During the 2 weeks following irradiation, thymocyte cell number increased from 1.2 x 107 at week 1 to + 40 x 107 at week 2. Thereafter, the cell number increased much slower (Fig. 1).This thymocyte recovery was in harmony with morphological restoration of the thy- mus.There was little difference in total thymocyte number between SYN+ or SYNrats, but the differences in the cortex/medulla ratio were drastic.

Two weeks after irradiation, the medulla in the SYN+ thymus was very small compared to SYN- and AMCbut resembled the AMC+ thymus.As soon as 1 week after the daily administration of 7.5 mg CsA/kg, a reduction of the thymic medullary area was noticed in AMC+ thymuses.Two weeks after the initial administration, the reduction was maximal and remained the same at least until 1 week after the 4-week course of CsA-administration had been terminated, both in SYN+ and AMC+ thymuses.The cortical area recovered within 2 weeks after irradiation and seemed to be unaffected by CsA in SYN+ thymuses.post BMT in the medulla, as was the expression of MHC class and II, ICAM-1, ED1, and ED8.Therefore, the expression of these markers is very indicative for the presence of IDC in SYN+ thymuses, as well as AMC+ thymuses from 2 weeks after irradiation onwards.In SYN-thymuses as soon as 2 weeks after irradiation normal distribution and expression of medullary IDC antigens was seen (Fig. 2).

A partial recovery of the medullary size in SYN + and AMC+ thymuses was observed at 2 weeks ater cessation of the CsA treatment together with a r covery of medullary stromal cells expressing ED1, ED8, OX6, OX18, OX39, and OX62.AMC+: open circles; SYN-: closed squares; SYN+: closed A panel of mAbs recognizing thymic epithelium circles) at the time indicated.Cell numbers were assessed using a Birker Hemocytometer.From week 6 and later when CsA-AI was used to examine different subsets of epithelial developed, two SYN+ animals free of macroscopical signs cells.The mAbs HIS37 and RCK 102, both pan (SYN+) and two animals with clear signs of dermatitis epithelial cell markers in the thymus, visualized (SYN + CsA-AI: dotted line) were assessed, three different areas with regard to the presence and morphology of epithelium: (1) the medulla with Interdigitating Cells epithelial cells most abundantly present in the cor- ticomedullary region, (2) the cortex with the fine We used a set of mAbs reactive with IDC in the reticular epithelium, and (3)the epithelial-free areas thymic medulla.Most of these mAbs, however, are closely located to the capsule surrounding the thy- not specific for IDC, but additionally react in the mus and the protruding septae.Furthermore, mAbs thymus with epithelial cells (OX6, OX18, 1A29), M) HIS38 and ED19 recognize cortical epithelial cells, (ED1, ED8, 1A29, OX6, OX18), or thymocytes whereas HIS39 and ED20 are specific for medullary (OX18).The diffuse staining patterns of the mAbs epithelium.The epithelial cells are also detected by recognizing epithelial cells or Mq besides IDC in the the earlier mentioned mAbs OX6, OX18, and 1A29.thymic medulla made it difficult to distinguish be-

In SYN + rats, the cortical restoration seemed tween the different cell types.Especially mAbs normal.HIS38 as well as ED19 clearly showed OX39 and OX62, which in the thymic medulla the cortical area, which was very extensive after 2 recognize specifically IDC, appeared to be very weeks.However, a combination of changes in useful for the identification of this cell type.thymic architecture was observed (see Table 1).As mentioned before in SYN+ rats, the thymic Epithelial-free areas (EFA) close to the septae and architecture was perturbed by irradiation, and the capsule became very flat.Furthermore, after 2 medulla recovery was strongly inhibited by the CsA weeks, vast areas void of epithelium developed treatment.As in SYN-thymuses, the effects of surrounding the medulla.The medulla remained irradiation were very drastic; 1 week after irradia-relatively small as determined by the medullary tion a clearly detectable medulla was absent.A1-epithelium markers HIS39 and ED20.Thymuses though the IDC antigens were present, it was from SYN+ rats showed no difference in the impossible to determine whether it concerned cell epithelium cell pattern compared to AMC+ thydebris or viable cells.Morphologically, normal cells muses at 2 weeks after irradiation, except for the expressing OX6, OX18, OX39, or OX62 were not absence of the flat EFAs close to the septae and detectable in the medulla.However, in SYN+ rats, capsule in the AMC+ thymuses.Overall, these a small medulla could be detected as soon as 2 changes were especially clearly visualized by the weeks after irradiation (Fig. 2).The medulla of mAb OX6, recognizing MHC class II present on SYN+ thymuses was of comparable size as in epithelial cells, MG and IDC (Fig. 3).In SYN- AMC+ thymuses.OX62-positive cells and OX39-rats, irradiation damage seemed restored within 2 positive cells were present in a normal distribution weeks.Typically, the EFAs close to the septae and Present recovery capsule were very flat as in SYN+ thymuses, but the epithelial-free areas around the medulla were

t observed.
Macrophages

The presence of macrophages was examined with mAbs ED1 (monocytes, M(, and IDC) and ED2 ( ortical M)).

In SYN+ and SYN-thymuses 48 hr after irradiation in the cortex, the relatively radio- resistant Mq were packed very densely against each other due to the disappearance of cortical thymocytes.As soon as 2 weeks after irradiation, ED1-and ED2-positive M were normally distrib- uted throughout the cortex, both in SYN + and SYN-thymuses.In AMC+ thymuses ED2 distri- bution in the cortex was no disturbed by CsA.ED2-positive cells never were observed in the medulla, as identified by the localization of the morphological difference between cortical and medullary thymocytes, in any of the thymus sec- tions of all four groups studied.In the EFAs surrounding the medulla, ED1-and ED2-positive cells were present in a normal distribution, com- parable to the ED1 and ED2 distribution in the cortex.The distribution of EDl-positive Mq in the medulla was normal 2 weeks after irradiation and seemed unaffected by

he CsA trea
ment.


Thymocytes

The thymocytes have been studied with OX19, R73 (both expressed low in the cortex and high in the medulla), HIS45 (specific for medullary thymo- cytes), OX8, and W3/25 (in the cortex, all thymo- cytes are positive, whereas in the medulla, approximately two-thirds are W3/25-positive and one-third is OX8-positive).The antigen expression recognized by these mAbs was not changed in any of the observed thymic sections at any time.The only difference was observed in the distribution.In SYN+ thymuses 2 weeks after irradiation, a normal cortex and a small medulla could be detected in AMC+ thymuses.In the epithelial-free areas in the vicinity of the medulla, as formed by the CsA treatment, and in the epithelial-free areas close to the septae and capsule, the thymocytes had a cortical phenotype.In SYN-and AMC-thymuses, the distribution was normal.Overall, the distribu- tion described was according to the stromal-cell distribution defining cortex and medulla.After ces- sation of the CsA treatment, both SYN+ and AMC + thymuses gradually recovered as stated before, and the distribution of cortical and medul- lary thymocytes resembled the normal situation.At the onset of disease, a disturbance in thymic mor- phology was observed affecting cortex and medulla (as described by Beschorner et al., 1988a).This was associated with a reduction of thymocyte number (Fig. 1) and a ecrease of thymic size.

By tricolor flow cytometry, we studied the thymo- cyte populations with respect to the antigens CD4, CD8, and TCR(x[.In SYN+, the number of TCRc[high positive thymocytes remained very low during CsA administration (Fig. 4), whereas in SYN-the strong depletion of the thymocytes was overcome as soon as 2 weeks after irradiation (Fig. 4).In AMC+, the TCR(-high positive thymocytes also diminished upon CsA administration (Fig. 4).Tri- color analysis showed that in SYN+ and AMC+ thymuses, the medullary CD4+TCRc[-high and CD8+ TCRc[-high thymocytes disappeared, whereas the cortical CD4+CD8+TCRc[intermedi- ate thymocytes were unaffected (Fig. 5).After the CsA treatment was stopped, the number of CD4+ TCR([-high and CD8+TCR(x[-high cells started to normalize in SYN + and AMC + thymuses.As CsA- AI developed, the total number of thymocytes was reduced, but the relative number of CD4+ TCRc]-high and CD8+TCR(]-high cells remained the same.All these changes in thymocyte FIGURE 3. Thymuses from AMC-(A, B), AMC + (C, D), SYN-(E, F), and SYN + (G, H) rats at 2 and 4 weeks after initiation of the experiment.Sections were stained with mAb OX6 recognizing MHC class II (RTIB) present on epithelial cells and IDc.(A, C, E, and G) Week 2; (B, D, F, and H) eek 4. E: epithelial-free areas void of MHC class II; C: cortex with fine reticular epithelium; M: medulla with IDC and epithelium.In (C and D) (AMC + and (G and H) (SYN + ), arrows are indicating the small medullary remnants.ing a life gate, 10,000 cells per sample were analyzed.Per group, two thymuses were analyzed and the average relative amount of TCR(-high posi- tive thymocytes is shown.AMC- thymuses were assessed from week 0, thymuses of AMC+, SYN-and SYN+ rats were tested from week until week 8.

AMC AMC * SYN + CD8 ICD4 FIGURE 5.By tricolor flow cytometry, the CD4, CD8, and TCR( expressions of thymocytes, isolated from thymuses at week 4, were assessed.By using a life gate, 10,000 cells per sample were analyzed.The first column shows the analysis of a control thymus (AMC-);

in the second column, a CsA-treatd thymus (AMC+) is analyzed., and in the third column, a thymus from an irradiated and CsA-treated rat (SYN + is analyzed.In the upper panels, the total expression of TCR( is shown.The middle panels show the dot p ot analysis of CD4 (FITC) and CD8 (PE) for the same thymocyte populations.Per qua rant, the TCR( expression is depicted in the lower panels as CD4-CD8+ in quadrant 1, CD4+CD8+ in quadrant 2, CD4-CD8-in quadrant 3, and CD4+CD8-in quadrant 4. Notice the presence of TCR0-high positive cells in quadrant and 4 of the AMC-thymus and the absence of these cells in the AMC + and SYN + thymuses due to the strong decrease of CD4 and CD8 single-positive thymocytes, rather than not expressing TCR( by these cells.

population were in accordance with the changes observed by immunohistochemistry.

The Effects of Thymectomies Performed After Irradiation or CsA treatment To evaluate the effects of an abrogated recovery after irradiation or CsA treatment only, thymecto- mies were performed.Because the presence of the tromal microenvironment recovers in the second week after irradiation, the possible generation of autoreactive T cells during this period was evaluated by thymectomies, preventing subsequently the gen- eration of regulatory T cells in the thymus.At days 7, 10, and 14 after irradiation and consecutive BMT each time, three SYN-rats were thy

ctomized an
observed daily for the development of CsA-AI.None of these animals developed macroscopical signs of CsA-AI within a period of 3 months after thymectomy.

In order to test the alternative hypothesis that the effect of irradiation can be replaced by thymectomy after the CsA treatment, two groups of 12 AMC+ rats were thymectomized at day 28 directly after a continuous course of CsA administration.Again, no macroscopical signs of CsA-AI developed within a period of 3 months after thymectomy.


DISCUSSION

The induction of CsA-AI requires the combination of lethal irradiation and syngeneic bone marrow reconstitution and the treatment with CsA for a given period.Early after irradiation, the thymus morphology is very much perturbed.Because the thymocyte population has almost completely disap- peared, stromal cells are densely packed.Therefore, the presence and distribution of specific stromal cell types are dif icult to determine immunohistochemi- cally.Two weeks after irradiation the thymus is repopulated with thymocytes and a morphologically distinguishable cortex and medulla are formed.Whereas the cortex morphology and size are nor- mal, the restoration of the medulla is strongly inhibited by the CsA treatment.The presence of the medullary IDC and epithelial cells is reduced pro- portionally to the small size of the medulla; how- ever, the distribution of these stromal cells in the shrunken medulla is normal.

The effects of CsA on the thymus are still obscure.CsA causes a strong involution of the thymic me- dulla and a reduction of IDC in a dose-dependent manner (Damoiseaux et al., 1993).There is increas- ing evidence that CsA interferes with the early steps in T-cell activation by inhibiting the transcription of genes involved in cyt0kine production (Schreiber and Crabtree, 1992).Similar events occur in the maturation steps from cortical double-positive thymocytes into single-positive medullary thymocytes.A cascade of activation events is supposed to be essential in which several cytokines are produced (reviewed by Ritter and Boyd, 1993).It has been shown that the medullary stroma is highly dependent upon the presence of TCRc-high thymocytes (Shores et al., 1991; Thomson, 1992).A cytokine- driven network of factors produced by stromal cells and thymocytes might be very important for the maintenance of the medulla.The development of medullary IDC probably is dependent on the production of cytokines, such as GM-CSF and IL-2.Therefore, an effect on thymocyte maturation probably also involves an influence on IDC maturation and turnover.On the other hand, the thymocytes are highly dependent on the stromal cells for their selection.It has been suggested that the absence of stromal cells in CsA-treated thymuses is responsible for the leakage of autoreactive cells to the periphery.In this study, we showed that the antigenic density and distribution of these stromal cells are not dis- turbed.Thus, the role of CsA in inhibiting negative selection is not explain

by t
is activity of CsA.It has been described that CsA also inhibits apoptosis (McCarthy et al., 1992; Yasutomi et al., 1992).Therefore, it seems likely that thymocytes that are no longer allowed to become activated due to the presence of CsA escape apoptosis and leave the thymus unselected.Although CsA treatment causes the formation of "forbidden" possible autoreactive T cells (Jenkins et al., 1988; Bryson et al., 1991), CsA treatment alone is not sufficient to cause disease (Beschorner et hl., 199 lb).

After irradiation and BMT, the presence of the thymus is essential during the first 2 weeks of CsA administration for the development of CsA-AI (So- rokin et al., 1986;Wodzig et al., 1993).During this period, the effects of irradiation must be overcome, and a strong influx of thymocytes is observed.The presen e of medullary IDC is hard to determine and it is tempting to suggest that a lack of negative selection might occur during this period due to the absence of IDC.However, IDC are relatively radio- resistant (Murawska et al., 1991) and thymectomies of SYNrats at 7, 10, or 14 days after irradiation and BMT did not induce autoimmune phenomena.Hence, the T cells that mature early after irradiation in SYNrats do not form an autoreactive population capable of inducing autoimmunity as they do in SYN+ rats.

Our data show morphological changes in thymic architecture, especially in the distribution of cortex and medulla.However, it seems that there are no essential differences found in thymuses from irradi- ated and CsA-treated rats compared to CsA-treated rats, regarding the presence of thymic stromal cells.Cells essential for thymic selection are present, and the phenotype of the stromal cells and thymocytes is normal.At the onset of disease, the thymus seems to be one of the organs affected in CsA-AI.Thymus morphology shows that this effect probably is not mediated by corticosteroids as induced by a stress response, but that the thymus is one of the target organs of CsA-AI (Beschorner et al., 1988a).Al- though we know that the thymus is allowed to recover after irradiation, albeit in an altered fashion due to CsA administration, very often no thymic remnants are found in the late acute phase of disease.In SYN+ rats, thymectomies performed directly after the cessation of CsA administration do not influence the course of disease, indicating the essential role for the thymus early after irradiation during CsA treatment, bat not in the period there- after.Inability to recover after CsA cessation, how- ever, is not an explanation for development of disease as stated by others (Beschorner et al.,  1991b), because in our study thymec

y after
he cessation of CsA treatment in AMC+ rats do not cause CsA-AI.Because irra- diation is essential to evoke CsA-AI, it is likely that there is an additional effect on the peripheral im- mune system that normally suppresses effectively autoreactive T cells leaking from the thymus during CsA treatment.The role and charact ristics of this peripheral counterpart are under current investiga- tion and may elucidate the contributions the thymus provides in the induction of CsA-AI.


MATERIALS AND METHODS


Animals

Female, specific pathogen-free Lewis (LEW, RT1I) rats were used.The animals were obtained from our own breeding colony and were fed ad libitum.Animals were 6 weeks of age at the start of the experiment.For the collection of the thymus and bone marrow, rats were killed by cervical disloca- tion under ether anesthesia.

Protocol for the Induction of CsA-AI

The experimental protocol has been described before (Bos et al., 1988).In brief, rats were given 8.5 Gy at 0.5 Gy/min using a R6ntgen irradiator (Philips MG320, Hamburg) 1 day prior to synge- neic BMT.

t rats were given 6
107 viable nu- cleated syngeneic bone marrow cells in 1.0 ml BSS intravenously into a tail vein.CsA, a kind gift from Sandoz Pharma Ltd. (Basel, Switzerland), was dissolved in olive oil at a concentration of 7.5 mg/ml.Rats received 7.5 mg CsA/kg-day, admin- istered subcutaneously for 28 days.


Thymectomies

While under ketamin and xylazine anesthesia, rats were intubated and maintained on a respirator.The thorax was opened by mediastinal incision and the thymus was entirely removed.

Experimental Design In the first experiment, the immunohistochemistry of the thymus was studied.Rats treated according to the previously mentioned protocol (SYN+) were compared to three different age-matched control (AMC) animals: rats irradiated and sy geneic bone marrow reconstituted and treated with olive oil only (SYN-), rats only treated with CsA (AMC+), and rats administered olive oil only (AMC-) (Table 2).

Two rats from each group were sacrificed at 48 hr after irradiation and consecutive BMT and thereafter weekly starting from day 7 for up to 8 weeks.In the second experiment, the thymo

te populati
n was studied by flow cytometry.Rats were treated similarly.The thymuses were collected at the same time points and prepared for flow cytometry.

The third experiment was designed to evaluate the disease-inducing effects of the separate treat- ments in combination wit

thymectomies.Thymecto
ies were performed on three different time points (days 7, 10, and 14) after irradiation and consecutive BMT (SYN-rats) or directly after ces- sation of a 4-week course of CsA administration (AMC+ rats).


Antibodies

The mouse anti-rat monoclonal antibodies (mAb) used for immunohistochemistry are specified in Table 3. 1A29 was kindly provided by Prof. M. Miyasaka (Tokyo).The ED series was kindly pro- vided by Dr. C.D. Dijkstra (Amsterdam).The set of HIS mAbs was kindly provided by Dr. J. Kampinga   (Groningen, the Netherlands).W3/25 and the OX mAbs were obtained from the European Collection of Animal Cell Cultures (Salisbury, UK).OX62 was kindly provided by Dr. M. Brenan (Oxford), and R73 was kindly provided by Prof. Th.H(inig (Wirzburg, Germany).RCK102 was obtained from Organon Teknika (the Netherlands).


Immunohistochemistry

The thymus was snap frozen in isopentane and used for cryosections.Four-to six-tm-thick thymus cryosections were

t, air dried, a
d stored at -20C.All thymus sections were stained using a two-step immunoperoxidase technique after acetone fixation.Sections were incubated for 60 min with mAb in the appropriate dilution.After washing in PBS, slides were incubated with a 1:200 dilution of rabbit anti-mouse Ig peroxidase conjugate (Dako, Denmark) in PBS with 0.5% bovine serum albumin (w/v) (BSA, Sigma) and 3% normal rat serum (v/v) for 30 min.After washing with PBS, the sections were stained for peroxidase activity with 3,3'- diaminobenzidine-tetrahydrochloride (Sigma, St. Louis) (0.5 mg/ml, in Tris-HCI buffer (pH 7.6), containing 0.02% (v/v) H202).Control sections were incubated in the same way omitting the first- step mAb.Subsequently, the sections were counter- stained with hematoxylin, dehydrated, and covered with Entallan (Merck, Germany).


Flow Cytometry

The thymus was teased apart and passed through a 100-t-mesh nylon gauze and collected in BSS sup- plemented with 2% heat-inactivated new born calf  II (RTIB), present on IDC, epithelium, and M CD8, present on all cortical and a subset of the medullary thymocytes MHC class (RTIA), present on all cells with high expression on epithelium, MG and IDC CD5, pan T-cell marker present on all thymocytes Interleukin-2 receptor, present on T blasts and IDC Integrin-like structure, present on IDC T-cell receptor All types of keratin, not species-specific CD4, present on all cortical thymocytes, a subset of medullary thymocytes and CTES: cluster of thymic epithelial staining CTES Pan epithelial cells CTES II Subscapsular/perivascular and the majority of medullary epithelail cells CTES III Cortical epithelial cells including thymic cells