Ligation of Retinoic Acid Receptor α Regulates Negative Selection of Thymocytes by Inhibiting Both DNA Binding of nur77 and Synthesis of Bim

Experimental animals and treatments

All the experiments were done using 4- to 5-wk-old NMRI or TCR-transgenic (AND) mice expressing a receptor specific for a fragment of PCC in the context of the E^k (class II MHC) molecule (39). To induce thymic apoptosis, NMRI mice were i.p. injected with anti-CD3 Ab (80 μg; BD PharMingen, San Diego, CA), and AND mice with 40 μg PCC (Sigma-Aldrich, St. Louis, MO), or with 20 μg staphylococcal enterotoxin A (SEA; Sigma-Aldrich). The amount of CD336, an RARα agonist (40), injected was 50 μg in 5 μl DMSO. The same amount of DMSO was added to each animal.

Characterization of thymocyte subpopulations

Thymocytes were isolated after 24 h of various treatments. Cells were washed twice and resuspended in ice-cold PBS containing 0.1% (w/v) sodium azide before staining with PE-labeled anti-CD4 and FITC-conjugated anti-CD8 (BD PharMingen). The cells were incubated with agitation for 30 min at 4°C, washed twice with ice-cold PBS supplemented with 1% BSA and 0.1% sodium azide, and resuspended in PBS containing 0.1% sodium azide. Unstained thymocytes treated similarly served as autofluorescence controls, whereas thymocytes stained with nonreactive FITC-conjugated goat IgG1 and PE-conjugated goat IgG1 Abs served as controls for nonspecific staining. Dual fluorescence was analyzed on a BD Biosciences FACScan (Le Pont de Claix, France) with excitation at 488 nm. Log-integrated green fluorescence (emission at 530 nm) and log-integrated red fluorescence (emission at 585 nm) were collected after combined gating on forward angle light scatter and 90° light scatter. The overlap in green and red emission was corrected using an electronic compensation network.

Separation of the CD4⁺CD8⁺ thymocyte subpopulation from the thymi of AND mice

The purification was conducted with the MACS multiparameter magnetic cell sorting system (Miltényi Biotec, Bergisch, Gladbach). Thymocyte suspension in PBS containing 0.5% BSA was first passed through 30-μm nylon mesh to remove clumps, stained with FITC-labeled anti-CD8 Ab, then incubated with MACS MultiSort anti-FITC microbeads. CD4⁺CD8⁺ cells were then positively selected on a magnetic column. Nonselected CD4⁺ and double-negative cells were also collected. The population of CD4⁺CD8⁺ thymocytes varied between 97.5 and 98.5% in the separated cell fraction.

RNA isolation and RT-PCR for detecting mRNA of nur77 and bim in thymocytes

RT-PCR for nur77 and bim was conducted following 4- and 8-h in vivo treatment, respectively, as it was described previously (41). For detection of the nur77 message oligonucleotides, 5′-TTC ATC CTC CGC CTG GCA TAC C-3′ and 5′-GTC CGA AGC TCA GGC AGT TTG C-3′ (sense and antisense for nur77); for bim variants, 5′-TCA ATG CCT TCT CCA TAC CAG-3′ and 5′-ATG GCC AAG CAA CCT TCT GAT G-3′ (sense and antisense for bim) were used. As an internal control, the β-actin message was also detected. Actin primers: sense, 5′-GGC TAC AGC TTC ACC ACC AC-3′, and antisense, 5′-GCG CTC AGG AGG AGC AAT G-3′. The reaction conditions were 30 (nur77), 35 (bim), or 23 (actin) cycles of denaturation at 94°C for 1 min, annealing at 57°C (60°C for bim) for 1 min, and extension at 72° for 1 min. Expected sizes for PCR products are 354, 591, 423, and 400 bp for nur77, bim_EL, bim_L, and β-actin, respectively.

Because for bim more bands were found than expected, oligonucleotides up (5′-ATGGCCAAGCAACCTTCTGATG-3′) and down (5′-TCAATGCCTTCTCCATACCAG-3′), which recognize all bim isoforms, were used to amplify the DNA fragment by RT-PCR. PCR products were ligated into pCR-Blunt (Invitrogen, San Diego, CA), and were sequenced with M13 forward and reverse primers. Running and analysis of the sequencing reactions were done on an automated DNA sequencing apparatus (ABI 373DNA Sequencer; Applied Biosystems, Foster City, CA). Analysis of the nucleotide and deduced sequences was performed using programs from the University of Wisconsin Genetics Computer Group (GCG) software package (1994). The deduced amino acid sequences of the open reading frames were compared with the whole set of protein sequences in SWISSPROT/PIR databases using the BLAST algorithm.

Bim immunoprecipitation

Thymi removed from control and mice treated for 12 h were passed through a metal sieve to isolate thymocytes. Cells were washed twice in ice-cold PBS and resuspended in lysis buffer containing 20 mM Tris-HCl, pH 7.4, 137 mM NaCl, 10% glycerol, 2% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), and 2 mM EDTA with 5 μg/ml leupeptin, aprotinin, chymostatin, antipain, and 0.3 mM PMSF. Lysis tubes were rotated for 30 min in cold room, and cell debris was removed by centrifugation at 14,000 rpm for 20 min. The protein content of the lysate was determined by the Bradford reagent, and 2 mg protein was used for immunoprecipitation. The lysate was precleared using 50 μl protein G-Sepharose Fast Flow (Sigma-Aldrich). The supernatant was mixed with 1 μg of anti-bim Ab (14A8; Chemicon, Temecula, CA) and incubated in cold room for 2 h, followed by addition of 50 μl of protein G-Sepharose Fast Flow for another 30 min. The beads were pelleted (3000 rpm, 2 min) and washed three times with lysis buffer containing 0.5% CHAPS, and the beads were resuspended in 50 μl of 2× concentrated Laemmli buffer (42).

Western blot for nur77 and Bim

Whole cell homogenate and the immunoprecipitate were used to follow the changes in the nur77 and bim protein levels, respectively. A total of 40 μg protein, or 30 μl of the bim immunoprecipitate, was run on an 8 or 12% polyacryamide gel, respectively, and blotted onto polyvinylidene difluoride membranes using the Bio-Rad (Hercules, CA) electrophoresis and transfer system. The free binding sites of membranes were blocked with 5% nonfat dry milk powder in 20 mM Tris, 0.1 M NaCl buffer with 0.1% Tween overnight at 4°C. A total of 1 μg primary Ab was added (anti-nur77 Ab from BD PharMingen; anti-bim Ab from StressGen (Victoria, British Columbia, Canada); or 14A8 anti-bim Ab detecting all three forms of bim from Chemicon) onto the membranes in 20 mM Tris, 0.1 M NaCl, 0.1% Tween, with 1% nonfat dry milk for 2 h. The excess Ab was removed by washing for 8 min, twice with 20 mM Tris, 0.1 M NaCl, 0.1% Tween. To detect nur77 signals, peroxidase-labeled anti-mouse IgG (1:1000), and for the bim detection peroxidase-labeled anti-rat IgG (1:5000) and the ECL kit from Amersham (Arlington Heights, IL) were used. As a positive control of nur77 detection, phorbol dibutyrate (PdBU; 5 ng/ml)- and calcium ionophore (A23189, 0.5 μM)-treated thymocytes were used. Equal loading of protein was demonstrated with probing the membranes with anti-actin Ab (Sigma-Aldrich).

Oligonucleotide end labeling

The NBRE and NurRE sense and antisense oligonucleotides were annealed by heating the mixed oligonucleotides to 68°C for 15 min and gradually cooling them to 4°C. A total of 100 ng double-stranded oligonucleotide was end labeled with 2 MBq [γ-³²P]ATP using 25 U T₄ polynucleotide kinase in 70 mM Tris-HCl (pH 7.6), 10 mM MgCl₂, and 5 mM DTT at 37°C for 1 h. To purify the labeled double-stranded oligonucleotide, the labeling mixture was run on a 6% nondenaturing polyacrylamide gel (acrylamide:bisacrylamide was 30:0.8), the band was cleaved out of the gel, and the oligonucleotide was allowed to diffuse from the gel into 400 μl Tris-EDTA buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA) for 1 h at 37°C. Sequences of the oligonucleotides: NBRE sense, 5′-GAT CCT CGT GCG AAA AGG TCA AGC GCT A-3′; NurRE sense, 5′-GAT CCT AGT GAT ATT TAC CTC CAA ATG CCA GGA-3′.

Nuclear extract from thymus

Thymi were excised and passed through a metal sieve to obtain single cell thymocyte suspension. Thymocytes were washed twice in PBS, then suspended in 400 μl of buffer A containing 10 mM HEPES, pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 10 mM 2-ME, 1 mM dithiotreitol, and 1 mM Na₃VO₄ with protease inhibitors: 5 μg/ml of leupeptin, antipain, and chymostatin; 50 μg/ml aprotinin; 0.3 mM PMSF; and 1 mM 6-aminohexanoic acid. Nuclear pellet was obtained by centrifugation at 5000 rpm for 5 min at 4°C following 5-min incubation of the cells on ice. Pelleted nuclei were resuspended in buffer C containing 20 mM HEPES, pH 7.9, 25% glycerol, 420 mM NaCl, 1.5 mM MgCl₂, 10 mM 2-ME, 1 mM DTT, 1 mM Na₃VO₄, and 1 mM EDTA with protease inhibitors as in buffer A, and incubated on ice for 15 min. The nuclear debris was removed by centrifugation at 14,000 × g for 15 min, and the supernatant was aliquoted and stored at −70°C.

EMSA

For NBRE and NurRE gel shift assays, 10 (NBRE) and 4 μg (NurRE) nuclear proteins were preincubated in the binding buffer (5 mM MgCl₂, 0.1 mM EDTA, 0.75 mM DTT, 7.5% glycerol, 0.05% Nonidet P-40) in the presence of 0.25 μg/ml BSA and 0.08 U poly(dI:dC) for 15 min on ice. For competition analysis, 1.5 μg anti-nur77 mAb (BD PharMingen) or 100-fold molar excess of cold oligonucleotide was added to the preincubation mix. After 15 min, 50,000 cpm labeled oligonucleotide was added, and the mixture was incubated for a further 15 min on ice. The complexes were resolved by electrophoresis in 6%, nonreducing polyacrylamide gel (acrylamide:bisacrylamide 30:0.8) in 0.25× Tris-borate-EDTA buffer (90 mM Tris, 90 mM borate, 2 mM EDTA, pH 8.3, set with solid borate). The gel was dried at 80°C for 2 h, and the result was visualized by autoradiography.

Detection of the in vivo rate of thymic apoptosis by TUNEL method

Thymi from 4-wk-old NMRI and AND mice were carefully removed, washed with physiological saline, Formalin fixed, embedded in paraffin, and stained with the APOPTAG in situ apoptosis detection kit (Intergen Discovery, Purchase, NY), following the manufacturer’s instructions.

Detection of the in vitro rate of apoptosis of the DP thymocytes exposed to SEA

Thymocyte suspensions were prepared from thymus glands of 4-wk-old male AND mice by mincing the glands in RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% charcoal-treated FCS (Life Technologies, Grand Island, NY), 2 mM glutamine, and 100 IU penicillin/100 μg streptomycin/ml. Thymocytes were washed three times and diluted to a final concentration of 10⁶ cells/ml before incubation at 37°C in a humidified incubator under an atmosphere of 5% CO₂/95% air. After being exposed to various treatments for 18 h, the percentage of apoptotic cells within the DP population was determined by the 7-amino actinomycin D method (43), after gating out the DP thymocytes stained positively with FITC-labeled anti-CD4 (BD PharMingen) and Cy5-conjugated anti-CD8 (purified and labeled in our lab) Abs.

Stimulation of RARα inhibits anti-CD3- or Ag-induced, but not bacterial superantigen-driven apoptosis of thymocytes

Previous studies in our laboratory have demonstrated that anti-CD3-mediated death of thymocytes in culture can be efficiently blocked by retinoids, and the effect is mediated via the retinoic receptor α (33, 34). To investigate further the role of RARα in the regulation of negative selection of thymocytes, three different in vivo models of negative selection were studied: anti-CD3-, Ag-, and bacterial superantigen-induced negative selection.

For studying anti-CD3-driven apoptosis, 4-wk-old NMRI mice were treated i.p. with anti-CD3 mAb. As we have previously reported, injection of a single dose of anti-CD3 mAb resulted in a significant loss in both the thymic weight and the CD4⁺CD8⁺ thymocyte population within 24 h (Fig. 1⇓A). Simultaneous injection of CD336, a specific agonist of RARα (40), completely prevented both the loss in the thymic weight and the loss of DP thymocyte population.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Ligation of the RARα protects against anti-CD3- and specific Ag-induced, but not against superantigen-driven thymocyte depletion. Mice were treated i.p. with 80 μg anti-CD3 mAb, 40 μg PCC, or 20 μg SEA with or without 50 μg CD336. Control animals were vehicle treated with 5 μl DMSO in physiological saline. After treatment (note: anti-CD3 and SEA, 24 h; PCC, 72 h), thymi were removed, and their weight was measured to follow thymic involution. Thymocytes were isolated, and the different populations were analyzed with flow cytometry. On the bar plots: bar 1, DMSO-treated control; 2, anti-CD3; 3, anti-CD3 with CD336; 4, PCC; 5, PCC and CD336; 6, SEA; and 7, SEA- and CD336-treated samples.

For investigating specific Ag-driven apoptosis, the AND TCRαβ-transgenic mice carrying a receptor specific for a fragment of PCC were used. In these mice, because of the presence of the TCR transgene, >95% of the thymocytes develop into CD4⁺ T cells, resulting in significant decrease in the DP subpopulation, and increase in the ratio of CD4⁺ and DP thymocyte subpopulations as compared with wild-type mice (39). Injection of a single dose of PCC resulted in a time-dependent decrease in the thymic weight, with no change in the first day, and reaching a minimum (46% weight loss) at day 3 following PCC injection (data not shown). Analyzing at day 3, the significant loss in thymic weight was accompanied by the selective loss of CD4⁺CD8⁺ thymocyte population that again was prevented by simultaneous injection of a single dose of CD336 (Fig. 1⇑B).

It has been shown that the toxins released by Staphylococcus aureus stimulate particularly those T cells that bear certain Vβ TCR (44). Because the TCR transgene contains Vβ3, and the transgene is expressed already at the double-negative stage (39), all thymocytes of the AND mice interact with the superantigen SEA. Injection of a single dose of SEA resulted in a time-dependent decrease in the thymic weight, with no change, or in some cases increase on the first day, and reaching a minimum (63% loss in total cellularity) at day 4. Because it was previously indicated that following injection of SEA thymic cell death may be initiated or modified by cytokines such as TNF-α produced by activated peripheral T cells (45), and SEA induced a significant cell proliferation of peripheral Vβ3⁺ T cells already at day 1, resulting in a 50% increase in the Vβ3⁺ cells of the spleen, we decided to test the effect of retinoids at an early time point as well. Despite no loss in thymic weight, a significant loss in the DP thymocyte population was already detected at day 1, which was accompanied by the concomitant increase in the percentage of double-negative cells that compensated the thymic weight for the loss of DP cells (Fig. 1⇑C). Although the loss of DP cells could have been the result of the down-regulation of CD4⁺ and CD8⁺ receptors known to occur following TCR ligation, the loss of DP cells via apoptosis was confirmed by the increased TUNEL positivity on thymic sections (data not shown). However, unlike in the previous models, simultaneous injection of CD336 could not prevent the loss of the CD4⁺CD8⁺ thymocytes induced by superantigen injection either at this early, or at later stages (data not shown).

To exclude the possibility that thymocytes in the SEA-injected mice die because of indirect effects caused by cytokine production from SEA-activated mature peripheral T cells, we decided to test whether CD336 could also inhibit SEA-induced death tested in vitro, as in our previous experiments increasing concentrations of CD336 could inhibit the in vitro apoptosis of both PdBU- and calcium ionophore-treated thymocytes (34), and anti-CD3-treated T cells (41). Because among the thymocytes from the AND mice the proportion of the apoptosis-sensitive DP population is relatively small (Fig. 1⇑), and SEA induces proliferation of the CD4⁺ cells, the percentage of apoptotic cells was determined within the gated DP rather then in the whole thymocyte population (Table I⇓). Increasing concentrations of SEA (from 0.1 to 10 μg/ml) induced increasing rate of apoptosis of DP cells. Addition of CD336 in concentrations, at which it inhibited anti-CD3-mediated death (34, 41), however, failed to inhibit SEA-induced death tested at 10 μg/ml concentration of SEA. These data imply that retinoids selectively interfere with the anti-CD3- and Ag-driven negative selection.

Stimulation of RARα does not affect anti-CD3- or Ag-induced expression of nur77, but suppresses its DNA-binding activity

Because nur77 was shown to be critical in mediating TCR-induced death of thymocytes (11, 12, 13, 18, 19, 20), we decided to test whether retinoids could affect the induction of nur77. As shown in Fig. 2⇓, both anti-CD3 and PCC treatment induced the expression of nur77 on both the mRNA and protein levels within 4 h following both anti-CD3 and PCC injection. Following in vivo induction, the protein appeared on Western blots with multiple bands ranging in size from ∼70 up to 90 kDa, similarly to the in vitro stimulation of thymocytes by PdBU and calcium ionophore. This wide range of nur77 protein species is caused by heavy phosphorylation of the protein demonstrated previously (18). Ligation of RARα by CD336 had no effect either on the level or on the major phosphorylation pattern (visible on Western blots) of nur77 induced by either stimulus.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Ligation of RARα does not prevent nur77 induction initiated by anti-CD3 cross-linking or specific Ag (PCC). Mice were injected with 80 μg anti-CD3 mAb, 80 μg anti-CD3 mAb together with 50 μg CD336, 40 μg PCC, or 40 μg PCC together with 50 μg CD336. Control mice were treated with 5 μl DMSO in physiological saline. Four hours later, thymi were excised; one-half was homogenized for Western blot analysis (A), while the other half was used for total RNA extraction and RT-PCR (B). Note that because of the usage of anti-mouse IgG secondary Ab, in addition to nur77, the Igs present in the thymic extracts are also visible on the blots.

To test whether retinoids could interact with the DNA-binding ability of nur77, the binding of nur77 to both NBRE and NurRE was investigated using nuclear extracts of thymocytes following in vivo PCC or anti-CD3 treatment for 4 h. An increased intensity of shifted band for both response elements (nur77-DNA complex) was detected in EMSA when thymocytes were treated with PCC (Fig. 3⇓, A and B, lane 2). This was not detected in the control thymi treated with vehicle or with CD336 alone (not shown). The identity of the shifted band (nur77-DNA complex) was confirmed by using anti-nur77 Ab and a chase experiment with cold probes. The nur77 mAb used in our experiments blocked protein-DNA complex formation, proving the presence of nur77 in the shifted band (46). Costimulation of RARα with CD336 prevented formation of both types of nur77-DNA complexes induced by PCC (Fig. 3⇓, A and B, lane 3). Because the level of nur77 was essentially unchanged between the PCC- and the PCC and CD336-treated cell extract (Fig. 2⇑), these data suggest that ligation of RARα inhibits the DNA-binding activity of nur77 to both NBRE and NurBE. Similar results were received when anti-CD3 stimulation was used (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Ligation of RARα decreases the DNA-binding activity of nur77. For electromobility shift assays, mice were injected with 40 μg PCC, or 40 μg PCC together with 50 μg CD336. Control mice were treated with 5 μl DMSO in physiological saline. Nuclear extracts were prepared from thymi removed 4 h posttreatment. EMSA was performed with 10 μg of nuclear proteins for NBRE and with 3 μg of nuclear proteins for NurRE. To check the specificity of the gel shift, 100 times higher molar concentration of cold NBRE, or NurRE oligonucleotides, or anti-nur77 mAb was used. A, EMSA using NBRE oligo as a probe; B, EMSA using NurRE as a probe. Lanes 1, 4, and 7, Control; 2, 5, and 8, PCC treated; 3, 6, and 9, PCC- and CD336-treated samples; and lane 10, free oligonucleotides.

Superantigen does not induce the expression of nur77 in the CD4⁺CD8⁺ thymocyte population

Next, we investigated whether injection of SEA can affect the expression levels of nur77. Similarly to the treatment with anti-CD3 or PCC, a single injection of SEA resulted in a time-dependent increase in the expression of nur77 in the thymus at both mRNA (Fig. 4⇓A) and protein levels (Fig. 4⇓B). The induction of nur77, however, did not occur in the apoptosis-sensitive DP thymocyte population, but was detected in the pooled double-negative and CD4⁺ cells (Fig. 4⇓C). These data imply that superantigens mediate apoptosis in the CD4⁺CD8⁺ thymocyte population by a mechanism that does not involve up-regulation of nur77.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Superantigen treatment does not induce nur77 in the CD4⁺CD8⁺, apoptosis-sensitive thymocyte population. TCR-transgenic mice were treated with 20 μg SEA. Time course of nur77 induction in total thymus homogenates determined on both mRNA (A) and protein levels (B). C, Expression of nur77 mRNA in the separated DP or double-negative and CD4⁺ simple-positive thymocyte populations 4 h after SEA treatment with or without CD336 (50 μg), respectively. Nur77 expression was detected by RT-PCR using total RNA extract.

Stimulation of RARα inhibits anti-CD3- or Ag-induced expression of bim

Because it has been previously shown that both bim_EL (23 kDa) and bim_L (19 kDa) are expressed in the mouse thymus (31), and were suggested to be involved in the negative selection of thymocytes (28), we decided to investigate whether ligation of TCR could affect the expression of bim. First, direct Western blot analysis was used (Fig. 5⇓A), with which increased expression of both bim_EL and bim_L was detected following both anti-CD3 and PCC induction.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Ligation of RARα prevents bim induction during negative selection driven by both anti-CD3 mAb and specific Ag. Mice were treated for 12 h with i.p. injected anti-CD3 mAb (80 μg), or PCC (40 μg) with or without 50 μg CD336. Control animals were injected with 5 μl DMSO in physiological saline. Bim levels were determined by direct Western blot (A) or following immunoprecipitation (B). Note: we were unable to detect bim_EL expression following immunoprecipitation, as during electrophoresis it runs together with the Ig L chains used for immunoprecipitation.

Because the amount of bim expressed by the thymus was very low, the data obtained by direct Western blot analysis were confirmed by combined immunoprecipitation/Western blot analysis as well, using the 14A8 mAb (31). Stimulation of TCR by either anti-CD3 mAb or the specific Ag resulted in an increased expression of bim_L detected 12 h following injection, which again was inhibited by simultaneous addition of CD336 (Fig. 5⇑B).

In accordance with the protein data, following TCR stimulation by anti-CD3 Ab or PCC, an increase was detected in the bim mRNA levels as well, which was prevented by injecting CD336 (Fig. 6⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Bim mRNA is induced in the thymus following anti-CD3 or PCC treatments. Mice were treated for 8 h with i.p. injected anti-CD3 mAb (80 μg), or PCC (40 μg) with or without 50 μg CD336. Control animals were injected with 5 μl DMSO in physiological saline. Bim expression was detected by RT-PCR using total RNA extract. Sequencing confirmed that the 591-bp cDNA band corresponds to bim_EL, while the 423-bp cDNA band to bim_L. The third band seen is the result of an alternative amplification of bim_L caused by binding of the forward primer to an unexpected additional target sequence.

Although the induction of bim was readily detected in both anti-CD3- and PCC-treated thymi, no induction of bim was seen following SEA treatment (data not shown).