This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Montani, V. Articles by Kohn, L. D. Search for Related Content PubMed PubMed Citation Articles by Montani, V. Articles by Kohn, L. D.

Regulation of Major Histocompatibility Class II Gene Expression in FRTL-5 Thyrocytes: Opposite Effects of Interferon and Methimazole¹

Cell Regulation Section, Metabolic Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases (M.S., S.-i.T., K.S., C.G., G.N., J.S., M.S., B.F., L.D.K.), and Experimental Immunology Branch, National Cancer Institute (D.S.S.), National Institutes of Health, Bethesda, Maryland 20892; the Department of Surgery, Johns Hopkins University (M.S.), Baltimore, Maryland 21287; and the Departments of Cancer Biology and Medicine, Harvard School of Public Health and Harvard Medical School (A.M.R.), Boston, Massachusetts 02115

Address all correspondence and requests for reprints to: Dr. Leonard D. Kohn, Metabolic Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, Building 10, Room 9C101B, National Institutes of Health, Bethesda, Maryland 20892-1360. E-mail: lenk{at}bdg10.niddk.nih.gov

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Aberrant expression of major histocompatibility complex (MHC) class II antigens is associated with autoimmune thyroid disease; aberrant expression duplicating the autoimmune state can be induced by interferon- (IFN). We have studied IFN-induced human leukocyte antigen (HLA)-DR gene expression in rat FRTL-5 thyroid cells to identify the elements and factors important for aberrant expression. Using an HLA-DR 5'-flanking region construct from -176 to +45 bp coupled to the chloramphenicol acetyltransferase reporter gene, we show that there is no basal class II gene expression in FRTL-5 thyroid cells, that IFN can induce expression, and, as is the case for antigen-presenting cells from the immune system, that IFN-induced expression requires several highly conserved elements on the 5'-flanking region, which, from 5' to 3', are the S, X₁, X₂, and Y boxes. Methimazole (MMI), a drug used to treat patients with Graves’ disease and experimental thyroiditis in rats or mice, can suppress the IFN-induced increase in HLA-DR gene expression as a function of time and concentration; MMI simultaneously decreases IFN-induced endogenous antigen presentation by the cell. Using gel shift assays and the HLA-DR 5'-flanking region from -176 or -137 to +45 bp as radiolabeled probes, we observed the formation of a major protein-DNA complex with extracts from FRTL-5 cells untreated with IFN, termed the basal or constitutive complex, and formation of an additional complex with a slightly faster mobility in extracts from cells treated with IFN. MMI treatment of cells prevents IFN from increasing the formation of this faster migrating complex. Formation of both complexes is specific, as evidenced in competition studies with unlabeled fragments between -137 and -38 bp from the start of transcription; nevertheless, they can be distinguished in such studies. Thus, high concentrations of double stranded oligonucleotides containing the sequence of the Y box, but not S, X₁, or X₂ box sequences, can prevent formation of the IFN-increased faster migrating complex, but not the basal complex. Both complexes involve multiple proteins and can be distinguished by differences in their protein composition. Thus, using specific antisera, we show that two cAMP response element-binding proteins, activating transcription factor-1 and/or -2, are dominant proteins in the upper or basal complex. The upper or basal complex also includes c-Fos, Fra-2, Ets-2, and Oct-1. A dominant protein that distinguishes the IFN-increased lower complex is CREB-binding protein (CBP), a coactivator of cAMP response element-binding proteins. We, therefore, show that aberrant expression of MHC class II in thyrocytes, induced by IFN, is associated with the induction or increased formation of a novel protein-DNA complex and that its formation as well as aberrant class II expression are suppressed by MMI, a drug used to treat human and experimental autoimmune thyroid disease. Its component proteins differ from those in a major, basal, or constitutive protein-DNA complex formed with the class II 5'-flanking region in cells that are not treated with IFN and that do not express the class II gene.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
MAJOR histocompatibility complex (MHC) class II molecules are heterodimeric transmembrane glycoproteins that are encoded by the human leukocyte antigen-D (HLA-D) region on chromosome 6 and play a central role in immune function (1, 2, 3). Class II antigens are usually expressed on antigen-presenting cells such as B cells, macrophages, or dendritic cells; the class II molecules present antigenic peptides to CD-positive T lymphocytes, causing T cell activation (1, 2, 3). Class II expression is normally not evident on epithelial cells, such as thyrocytes (1, 2, 3, 4, 5, 6); abnormal or aberrant expression of class II molecules on thyrocytes is associated with autoimmune thyroid diseases (ATD) (4, 5, 6).

One form of ATD is Graves’ disease, in which autoantibodies develop to the TSH receptor (TSHR) and induce hyperthyroidism by mimicking the action of TSH. Whereas numerous attempts to develop a Graves’ disease model by immunizing animals with the extracellular domain of the TSHR have largely failed (7, 8, 9, 10, 11, 12, 13, 14, 15, 16), immunizing mice with fibroblasts transfected with the human TSHR and a MHC class II molecule, but not either alone, has induced ATD with the major humoral and histological features of Graves’ disease (17): stimulating TSHR antibodies, TSH binding-inhibiting Igs that are different from stimulating TSHR antibodies, increased thyroid hormone levels, thyroid enlargement, and thyrocyte hypercellularity. These results support the hypothesis that aberrant expression of MHC class II molecules and acquisition of antigen-presenting ability on a thyroid cell may activate T and B cells normally present in an animal, thereby allowing normal immune tolerance to be broken. Understanding the basis for aberrant class II expression in thyrocytes may, therefore, be a potentially important aspect of understanding the pathogenesis of ATD and Graves’ disease.

Interferon- (IFN) is the most potent inducer of class II gene expression; its ability to induce class II expression is well studied in antigen-presenting cells of the immune system (1, 2, 3, 18, 19). In those cells, its action is transcriptionally mediated and involves conserved sequences present in all class II MHC promoters characterized to date (1, 2, 3, 18, 19). From 5' to 3', these are the S (also called W or H), X₁, X₂, and Y boxes (Fig. 1). Despite the fact that IFN can induce aberrant class II antigen expression when given in vitro to human (20) or rat FRTL-5 thyroid cells in culture (21, 22, 23) and can mimic the autoimmune state of thyrocytes in ATD (4, 5, 6), little is known about the transcriptional mechanism by which IFN modulates class II genes in thyrocytes. We, therefore, initiated a study of HLA-DR gene expression in rat FRTL-5 thyroid cells to define elements and factors important for IFN-induced aberrant expression in thyrocytes and, by inference, ATD. We additionally evaluated the effect of methimazole (MMI) on IFN-induced class II expression, as it is an agent used to treat Graves’ disease (24) and other forms of autoimmunity (25, 26, 27) and is capable of suppressing MHC class I gene expression in FRTL-5 thyroid cells (28).

View larger version (13K): [in this window] [in a new window]   Figure 1. Nucleotide sequence of the 5'-flanking region of the -176 bp HLA-DR-CAT construct used in these experiments. The S, X1, X2, and Y boxes or elements, which are conserved in all class II MHC promoters characterized to date, HLA-DR, -DQ, and -DP (1–3, 18, 19), are underlined and their 5'- and 3'-termini in DR are numbered. The more restricted S box site noted in some reports (2, 18) is shown by a dashed line over the more extensive S box used herein or in other reports (30). The 45-bp addition that links the 5'-flanking region to the CAT reporter gene is noted.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Materials
Highly purified bovine TSH was obtained from the Hormone Distribution Program of the NIDDK, NIH (30 U/mg), or was a previously described preparation with the mol wt, amino acid, and carbohydrate composition of authentic TSH (29). MMI was obtained from the Sigma Chemical Co. (St. Louis, MO), rat and human recombinant IFN were purchased from Life Technologies (Grand Island, NY)m and rabbit polyclonal antibodies against the components of the activating transcription factor (ATF)/cAMP response element-binding protein (CREB), Fos/Jun, nuclear factor-B (NF-B), Stat, IFN response element (IRF), Ets, and Oct families of proteins were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). [-³²P]Deoxy-CTP (3000 Ci/mmol) and [¹⁴C]chloramphenicol (50 mCi/mmol) were purchased from DuPont-New England Nuclear (Boston, MA); [-³²P]ATP (6000 Ci/mmol) was obtained from Amersham (Arlington Heights, IL). Synthetic oligonucleotides were purchased from Operon Technologies (Alameda, CA). The source of other materials was Sigma Chemical Co. unless otherwise noted.

Plasmids
HLA-DR promoter constructs were obtained from Dr. L. H. Glimcher, Harvard School of Public Health and Department of Medicine, Harvard University Medical School (Boston, MA); their construction and characteristics were previously described (30). An additional chloramphenicol acetyltransferase (CAT) plasmid containing -38 to +45 bp of the 5'-flanking region of HLA-DR was constructed by PCR, using the HLA-DR-CAT chimera containing -176 to +45 of 5'-flanking region as template and the following primers: a 5'-primer with a 5'-SphI restriction site, 5'-ACATGCATGCGGTCAGACTCTATTACACCCCAC-3', and a 3'-primer, 5'-CTAGTCTAGTTTGGGAGTCAGTAGAGCTCG-3', with an XbaI restriction site (30). The PCR products were purified by phenol-chloroform extraction, digested with XbaI and SphI, purified from a 2% agarose gel with Jet Sorb (Genomed, Frederick, MD), dephosphorylated with calf intestinal alkaline phosphatase (New England Biolabs, Beverly, MA), and religated (DNA ligation kit, Takara Biochemical, Berkeley, CA), into the pCAT-Basic Vector (Promega, Madison, WI) at the SphI and XbaI sites as previously described (30).

The thyroglobulin (TG)-CAT construct, pTG-688-CAT, was derived by substituting the TG promoter insert from our previously described pTG-CAT chimera (31) into the HLA-DR-CAT chimera from which the class II promoter insert was excised. The vector containing the CAT reporter gene but no insert was the control.

DNA was prepared and purified by CsCl gradient centrifugation (32). The sequences of all constructs were confirmed by a standard method (33).

Cell culture
FRTL-5 rat thyroid cells (Interthyr Research Foundation, Baltimore, MD; American Type Culture Collection CRL8305) were a fresh subclone (F1) with the properties previously described (28, 34, 35). They were grown in Coon’s Modified F-12 Medium containing 5% heat-treated, mycoplasma-free calf serum, 1 mM nonessential amino acids, and a mixture of six hormones (6H) containing bovine TSH (1 x 10^-10 M), insulin (10 µg/ml), cortisol (0.4 ng/ml), transferrin (5 µg/ml), glycyl-L-histidyl-L-lysine acetate (10 ng/ml), and somatostatin (10 ng/ml). Cells were diploid and between their 5th and 25th passage. Fresh medium was added every 2 or 3 days, and cells were passaged every 7–10 days. In some experiments, cells were grown to near confluence in 6H medium, then maintained for 6–8 days, before experiments were initiated, in 5H medium with no TSH.

Transient expression analysis
Transient transfections in FRTL-5 cells were performed as previously described (36, 37), using one of the following procedures. In one, cells were cultivated in 6H medium to approximately 80% confluence, harvested, washed, and resuspended (1.5 x 10⁷ cells/ml) in 0.85 ml electroporation buffer [272 mM sucrose, 7 mM sodium phosphate buffer (pH 7.4), and 1 mM MgCl₂]. Plasmid DNA was added (20 µg CAT chimera together with 2 µg pRSV-luciferase) (38), which was used to measure the efficiency of transfection. Cells were pulsed (330 V; capacitance, 25 µfarad), plated (6 x 10⁶ cells/10-cm dish), and cultured in 6H medium plus 5% calf serum supplemented or not with IFN. At the times noted, cells were harvested for CAT and luciferase assays. The second procedure differed as follows. FRTL-5 cells were grown to 80% confluence in 6H medium, then maintained 6 days in 5H medium plus 5% calf serum. Cells were returned to 6H medium for 12 h and transfected as described above. Twelve hours later the medium was changed to fresh 5H medium with 5% calf serum supplemented or not with IFN. Cell viability was approximately 80% in all experiments. CAT activity was measured as previously described (36, 37); values were normalized to luciferase activity measured using the Promega (Madison, WI) assay system.

Cellular extracts
Cell extracts were made by a modification of the method of Dignam et al. (39). Briefly, FRTL-5 cells were harvested by scraping after being washed twice in ice-cold PBS and pelleted. The pellet was resuspended in 2 vol Dignam buffer C [20 mM HEPES (pH 7.9), 1.5 mM MgCl₂, 0.42 M NaCl, 25% glycerol, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonylfluoride, 1 µg/ml leupeptin, and 1 µg/ml pepstatin]. The final NaCl concentration was adjusted on the basis of cell pellet volume to 0.42 M. Cells were lysed by repeated cycles of freezing and thawing. The extracts were centrifuged at 100,000 x g at 4 C for 20 min. The supernatant was recovered, aliquoted, and stored at -70 C.

Electrophoretic mobility shift assays (EMSA)
Oligonucleotides used for EMSA were synthesized (Operon Technologies) or were purified from 2% agarose gel using either QIAEX (Qiagen, Chatsworth, CA) or Jet Sorb (Genomed) after restriction enzyme treatment of the chimeric CAT constructs described above. They were dephosphorylated with calf intestinal alkaline phosphatase, labeled with [-³²P]deoxy-CTP using Klenow or with [-³²P]ATP using T4 polynucleotide kinase, then purified on an 8% native polyacrylamide gel (36, 37, 40, 41, 42).

EMSAs were performed basically as previously described (36, 37, 41, 42). Binding reactions were carried out in a volume of 20 µl for 30 min at room temperature. The reaction mixtures contained 1.5 fmol [³²P]DNA, 3 µg cell extract, and 1.5 µg poly(deoxyinosinic-deoxycytidylic) in 10 mM Tris-Cl (pH 7.9), 1 mM MgCl₂, 1 mM dithiothreitol, 1 mM EDTA, and 5% glycerol. Where indicated, unlabeled double or single stranded oligonucleotides were added to the binding reaction as competitors and incubated with the extract for 20 min at room temperature before the addition of labeled DNA. Similarly, in experiments using antiserum, extracts were incubated in the same buffer containing antiserum or normal rabbit serum for 20 min at room temperature before being processed as described above. After incubations, reaction mixes were subjected to electrophoresis on 3.5% or 5% native polyacrylamide gels at 160 V in 0.5 x Tris-borate-EDTA at room temperature for 1.5–2 h. Gels were dried and autoradiographed at -80 C overnight unless otherwise noted.

Other methods
The protein concentration was determined by Bradford’s method (Bio-Rad, Richmond, CA); recrystallized BSA was the standard.

For fluorescence-activated cell sorter analysis, single cell suspensions were prepared and stained as previously described (17, 43), except that HLA-DR antigen was detected using a class II-specific murine monoclonal antibody (44); Leu-4 was used as a background control.

Statistical significance
All experiments were repeated at least three times with different batches of cells. Values are the mean ± SE of these experiments as noted. Significance between experimental values was determined by two-way ANOVA and were significant at P < 0.05 when data from all experiments were considered.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Expression of HLA-DR in FRTL-5 thyroid cells requires IFN and the same 5'-promoter elements important for IFN induction of HLA-DR in lymphocytes
We used the HLA-DR -176 bp minimal promoter coupled to CAT, as previously described and used in studies of immune cells (30), to evaluate expression of the HLA class II gene in FRTL-5 thyrocytes. HLA-DR is not expressed in transiently transfected FRTL-5 thyrocytes, compared to the vector control, unless the cells are treated with rat recombinant IFN (Fig. 2A, first and second sets of open and closed bars). The action of rat IFN is not duplicated by human IFN and is associated with an increase in endogenous class II expression measured by flow cytometry (see below, Fig. 8), i.e. its action is specific and appears to reflect effects on the endogenous class II antigen.

View larger version (18K): [in this window] [in a new window]   Figure 2. Class II expression in FRTL-5 thyroid cells measured using the -176 bp DR-CAT construct and 5'-deletions thereof (A) or the -176 bp DR-CAT construct with mutations within the S, X1, X2, and Y boxes (B). On the left at the top of the figure is a diagrammatic presentation of the -176 bp DR-CAT chimera with the locations of the S, X1, X2, and Y boxes noted by black boxes, and the locations of the 5'-termini of the deletions noted. On the right at the top of the figure, the mutations made in S, X1, X2, and Y boxes are presented. Transient transfections were performed, as described in Materials and Methods, in FRTL-5 cells grown to near confluence in medium with TSH (6H medium) and treated with 100 U/ml IFN for 24 h after transfection. Results are expressed relative to the vector control in the absence of IFN (first open bar in each panel), after CAT activities were corrected for both luciferase activity and cell protein. These corrections in all cases resulted in less than 5% changes in activity. Results are the mean ± SD of three separate experiments performed on three different batches of cells. *, A statistically significant increase (P < 0.01) in DR promoter activity induced by IFN; **, a statistically significant decrease (P < 0.01) in IFN-induced DR promoter activity when the -176 bp DR minimal promoter contained mutations in the S, X1, X2, and Y boxes. The same results were obtained using an alternative protocol involving cells maintained in medium with no TSH.

View larger version (30K): [in this window] [in a new window]   Figure 8. Effect of IFN and MMI on MHC class II antigen expression in FRTL-5 cells. FRTL-5 cells were grown to near confluence in TSH, maintained for 8 days in 5H medium, then treated with 100 U/ml IFN, 5 mM MMI, or both for the last 48 h, as described in Fig. 7. Cells were stained with a fluorescein isothiocyanate-conjugated class II-specific monoclonal antibody to RT1.B (clone OX-6, Sera Labs, UK) for 60 min at 4 C, washed twice with Dulbecco’s PBS, and subjected to laser flow cytometry.

In summary, unlike lymphocytes (2, 18, 19, 30), there is no constitutive or basal expression of the HLA-DR promoter-CAT chimera in thyrocytes. Like monocytes, however (2, 18, 19), IFN increases HLA-DR promoter expression and requires the same highly conserved 5'-flanking region elements, S, X₁, X₂, and Y, that are present in all class II genes for this effect.

The effect of TSH on IFN-induced HLA-DR gene expression in FRTL-5 thyroid cells
TSH has been reported to enhance the ability of IFN to increase class II antigen expression in rat FRTL-5 thyroid cells when measured after 48 h of IFN treatment (21). We, therefore, evaluated the effect of IFN on DR gene expression in FRTL-5 thyroid cells maintained with and without TSH, as this might have influenced subsequent experiments.

In the presence of TSH, the effect of IFN on class II gene expression was measurable within 12 h and maximal at 48 h (Fig. 3A); in the absence of TSH, it was also measurable by 12 h, was maximal at 24 h, and decreased slightly, but significantly, by 48 h (Fig. 3B). Consistent with these data, IFN-induced endogenous class II antigen expression, measured by flow cytometry (data not shown), was higher at 48 h in FRTL-5 thyroid cells transfected with the HLA-DR-CAT chimera and maintained with TSH than in cells maintained without TSH. The present data are thus consistent with and might explain why IFN-increased antigen expression was higher after 48 h of treatment in the presence of TSH (21), i.e. TSH delays the ability of IFN to maximally increase gene expression. Both studies suggest that the effect of TSH on IFN activity is a kinetic phenomenon.

View larger version (21K): [in this window] [in a new window]   Figure 3. Effect of IFN on class II expression in FRTL-5 thyroid cells measured in cells maintained with (A) or without (B) TSH in the medium. Transient transfections were performed, as described in Materials and Methods, using the -176 bp DR-CAT chimera. In A, FRTL-5 cells were grown to near confluence in 6H medium and treated with 100 U/ml IFN for the noted times starting 12 h after transfection. In B, an alternate procedure was used, as detailed in Materials and Methods. Thus, cells were grown to 80% confluence in 6H medium, maintained for 6 days in 5H medium, returned to 6H medium for 12 h, and transfected as described; this procedure is required to ensure cell viability comparable to that using the first method in transfections by electroporation. Twelve hours later, the medium was changed to fresh 5H medium, supplemented or not with IFN, and CAT activity was measured at the times noted thereafter. Cell viability was approximately 85% in all experiments. Results are expressed relative to the vector control in the absence of IFN (first open bar in each panel) after CAT activities were corrected both for luciferase activity and cell protein. These corrections in all cases resulted in less than 5% changes in activity. Results are the mean ± SD of three separate experiments performed on three different batches of cells. In A and B, * denotes a statistically significant increase (P < 0.01) in DR promoter activity induced by IFN. In B, ** denotes a statistically significant decrease (P < 0.05) in IFN-induced DR promoter activity at 48 vs. 24 h.

The effect of MMI on IFN-induced HLA-DR gene expression in FRTL-5 thyroid cells

View larger version (23K): [in this window] [in a new window]   Figure 4. Effect of IFN on class II expression in FRTL-5 thyroid cells as a function of IFN concentration (A) and in the presence of MMI (B). Transient transfections with the -137 bp DR-CAT chimera were performed as described in Materials and Methods. In A, FRTL-5 cells were grown to near confluence in 6H medium and treated with 100 U/ml IFN for the noted times starting 12 h after transfection; CAT activity was measured 48 h after IFN addition. In B, FRTL-5 cells were grown to near confluence in 6H medium and treated with 100 U/ml IFN for the noted times, starting 12 h after transfection. In a duplicate set of cells, 5 mM MMI was added simultaneously with IFN addition or 24 h after the addition of IFN; CAT activity was measured 48 h after the addition of IFN. Cell viability was approximately 84 ± 4% in all experiments. Results are expressed relative to the vector control in the absence of IFN after CAT activities were corrected for both luciferase activity and cell protein. These corrections in all cases resulted in less than 5% changes in activity. Results are the mean ± SD of three separate experiments performed on three different batches of cells. In A, * denotes a statistically significant increase (P < 0.01) in DR promoter activity induced by IFN. In B, ** denotes a statistically significant decrease in IFN-induced DR promoter activity with 24 or 48 h of MMI exposure. The same results were obtained using the alternative transfection protocol involving cells maintained in medium with no TSH.

MMI prevents the ability of a maximally effective concentration of IFN to increase the activity of the -137 bp DR-CAT chimera as a function of time. Thus, CAT activity induced by 100 U/ml IFN was progressively decreased 24 and 48 h after MMI addition (Fig. 4B). The MMI concentration in this experiment (5 mM) was previously shown to be maximally effective to suppress IFN-induced MHC class I expression in these cells (28); however, the effect was evident at lower MMI concentrations and was dependent on the concentration of MMI (Fig. 5). MMI had no effect on promoter in the absence of IFN (Fig. 5). The actions of both IFN and MMI were specific; thus, neither affected the control vector, and each had opposite effects on a TG-CAT chimera (Table 1): IFN decreased TG-CAT activity, and MMI reversed the IFN-induced decrease in TG-CAT activity. Also, MMI alone increased TG-CAT activity, but not HLA-DR CAT activity (Table 1).

View larger version (25K): [in this window] [in a new window]   Figure 5. Effect of MMI on IFN-increased class II expression in FRTL-5 thyroid cells as a function of MMI concentration. Transient transfections with the -137 bp DR-CAT chimera were performed as described in Materials and Methods. FRTL-5 cells were grown to near confluence in 6H medium and treated with 100 U/ml IFN for the noted times starting 12 h after transfection. In duplicate sets of cells, the noted concentrations of MMI were added simultaneously with IFN, and CAT activity was measured 48 h later. Cell viability was approximately 85 ± 5% in all samples. Results are expressed relative to the vector control in the absence of IFN or MMI after CAT activities were corrected both for luciferase activity and cell protein; these corrections in all cases resulted in less than 5% changes in activity. Results are the mean ± SD of three separate experiments performed on three different batches of cells. **, A statistically significant decrease in IFN-induced DR promoter activity caused by MMI. The same results were obtained using the alternative transfection protocol, involving cells maintained in medium with no TSH.

View this table: [in this window] [in a new window]   Table 1. Effect of IFN- and MMI on the CAT activity of HLA-DR- and rat thyroglobulin-CAT chimeras

IFN treatment of FRTL-5 thyrocytes induces the formation of a specific protein/DNA complex with the DR 5'-flanking region; MMI inhibits its formation

View larger version (29K): [in this window] [in a new window]   Figure 6. EMSA of the ability of a 32P-radiolabeled DR-5'-flanking region probe to form protein-DNA complexes with extracts from FRTL-5 cells maintained with and without TSH and treated or not with IFN. The probe was excised by restriction enzyme treatment of the -137 to +45 bp DR-CAT chimera and is diagrammatically represented at the bottom of the figure. Extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH or maintained for 8 days in 5H medium; duplicate cultures of each were treated with 100 U/ml IFN for the last 48 h before the experiment was terminated. Cell extracts made from each set of cells were incubated with the 32P-radiolabeled probe containing -137 bp of the DR 5'-flanking region, and EMSA was performed as described in Materials and Methods. In this experiment, the autoradiogram was exposed overnight at -70 C.

View larger version (52K): [in this window] [in a new window]   Figure 7. The effect of MMI on the ability of IFN to increase the formation of protein-DNA complexes with the 32P-radiolabeled DR 5'-flanking region probe. The radiolabeled probe is the same as that used and diagrammatically presented in Fig. 6; it contains -137 bp to +45 bp of the 5'-flanking region of the DR-CAT chimera. Extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH, maintained for 8 days in 5H medium, then treated with 100 U/ml IFN, 5 mM MMI, or both for the last 48 h before the experiment was terminated. Cell extracts were incubated, and EMSA was performed as described in Fig. 6 and Materials and Methods. The arrows denote the upper and lower complexes seen in Fig. 6. In this experiment, the autoradiograms were exposed for 72 h at -70 C. Lane 5 contains extract from control cells; extracts from cells treated with IFN or MMI are noted at the top of the panels. The same results were obtained if cell extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH and treated with 100 U/ml IFN for the last 48 h before the experiment was terminated.

View larger version (37K): [in this window] [in a new window]   Figure 9. The ability of unlabeled oligonucleotide competitors to prevent the formation of protein-DNA complexes with the 32P-radiolabeled DR 5'-flanking region probe. FRTL-5 cells were grown to near confluence in TSH, maintained for 8 days in 5H medium, then treated with 100 U/ml IFN for the last 48 h, as described in Figs. 6 and 7. The ability of extracts to form protein-DNA complexes with the 32P-radiolabeled -137 HLA-DR probe used in those figures was evaluated in the presence of a 100-fold excess of the unlabeled oligonucleotides diagrammatically presented on the bottom of the figure and termed -137, -111, -90, and -38 or a 100-fold excess of oligonucleotides -176 DR Mut. S, Mut. X1, Mut. X2, and Mut. Y containing the mutations detailed in Fig. 2. The unlabeled oligonucleotides, like the radiolabeled probe, were obtained by restriction enzyme excision of the CAT chimeras used in Fig. 2B; they were incubated together with cell extracts for 20 min before addition of the radiolabeled probe. EMSA was performed as described in Figs. 6 and 7. The arrows denote the upper and lower complexes seen in Figs. 6 and 7. In A, lane 1 contains the probe with no extract as a control. The same results were obtained if cell extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH and treated with 100 U/ml IFN for the last 48 h before the experiment was terminated.

Supporting the importance of the Y box as a binding site, at least for the IFN-induced lower complex, and distinguishing it from the basal or constitutive complex were competition studies using small oligonucleotide fragments comprising the sequences of the S, X₁, X₂, or Y boxes (Fig. 10). Thus, formation of the IFN-induced lower complex was specifically prevented by a 200-fold excess of a smaller, double strand oligonucleotide, -87 to -55 bp, containing the Y box sequence (Fig. 10, bottom), but not a 200-fold excess of an oligonucleotide with the sequence of the X₁-X₂ or S boxes, when each was included in the gel shift assay in vitro (Fig. 10, lanes 7 vs. 3–6 and 8–11). A 500-fold excess of the X₁-X₂ or S boxes, when each was included in the gel shift assay in vitro, was also unable to prevent formation of either complex (data not shown). Also unable to inhibit formation were 200-fold (Fig. 10) or 500-fold (data not shown) excesses of the sense or antisense single strand oligonucleotides with the Y, X₁-X₂, or S box sequences of the minimal HLA-DR promoter (Fig. 10 and data not shown). Inhibition was not duplicated by single or double strand oligonucleotides with the sequence of the S plus X₁ boxes, -141 to -92 bp (data not shown), or with the sequence of the HLA-DR promoter between -182 to -144 bp, which served as a negative control (Fig. 10, lanes 12–14). Thus, the Y box element is particularly important in the formation of the IFN-induced complex (Fig. 9A, lane 5, and Fig. 10, lane 7), whereas the Y box element plus the 3'-sequence to -38 bp are particularly important in the formation of the basal or constitutive complex (Fig. 9A, lane 5). This may be explained by the existence of different protein components in each complex, as will be shown below.

View larger version (41K): [in this window] [in a new window]   Figure 10. The ability of oligonucleotides with the sequences of the S, X1-X2, and Y boxes of HLA-DR to inhibit the formation of protein-DNA complexes between extracts from FRTL-5 thyroid cells treated with IFN and the 32P-radiolabeled HLA-DR 5'-flanking region probe. The probe is the same as that used in Figs. 6, 7, and 9; it contains -137 to +45 bp of the 5'-flanking region of the DR-CAT chimera. Oligonucleotide competitors are diagrammatically outlined on the bottom along with the probe; the sequence of each can be derived from Fig. 1. Additionally, an oligonucleotide representing -182 to -144 bp of the 5'-DR flanking region was used as a negative control. Extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH and treated with 100 U/ml IFN for the last 48 h before the experiment was terminated. Cell extracts made from each set of cells and extracts were incubated with or without 300 fmol of the noted unlabeled single or double strand oligonucleotides (a 200-fold molar excess over probe) before being incubated together with the radiolabeled probe (see Materials and Methods). EMSA were performed as described in Figs. 6, 7, and 9. The arrows denote the upper and lower complexes seen in Figs. 6, 7, and 9. The same results were obtained if cell extracts were from FRTL-5 cells grown to near confluence in TSH, maintained for 8 days in 5H medium, then treated with 100 U/ml IFN for the last 48 h, as described in Figs. 6, 7, and 9.

Characterization of proteins in the protein-DNA complexes formed with extracts from FRTL-5 thyroid cells treated with or without IFN
Although we observed that formation of the two complexes was still markedly attenuated in competition experiments comparing the ability of unlabeled oligonucleotides with mutations in each of the S, X₁, X₂, and Y boxes to inhibit complex formation (Fig. 9B), competition by an oligonucleotide with a single site mutated resulted in small residual complexes, each with a different mobility in the region of the upper complex (Fig. 9B, lanes 2–5 vs. 1). This suggested there might be multiple protein components in the constitutive complex, consistent with studies in antigen-presenting cells from the immune system (1, 2, 3), as will be noted in Discussion. Additionally, differences in the ability of the oligonucleotides containing the Y box or the Y box plus the 3'-sequence to -38 bp (Figs. 10 and 9A) to compete for the IFN-increased lower complex vs. the basal, upper complex, respectively, suggested that each might involve different proteins. We pursued this question using antisera against specific transcription factors or coregulators.

An antiserum specific for ATF-1 eliminated formation of the upper, basal, or constitutive complex and reduced the intensity of the IFN-increased lower complex, both of which are evident in incubations of extracts from FRTL-5 cells treated with interferon and the DR probe containing -137 bp of the 5'-flanking region (Fig. 11, lane 6). A specific antiserum to ATF-2 appeared to have a lesser effect on the upper complex but a slightly greater ability to decrease the intensity of the lower complex (Fig. 11, lanes 4 vs. 6) when analyzed quantitatively after densitometry; however, the significance of this was unclear. Nevertheless, this was not the case for antisera to ATF-3, ATF-4, or two cAMP response element-binding proteins, CREB-1 or CREB-2 (Fig. 11, lanes 5, 7, 2, and 3, respectively), suggesting that these were specific effects of the antisera, and these were specific components of the upper or basal complex in particular. In the case of all negative antisera, here and discussed below, the efficacy of the antiserum was validated in control experiments showing inhibition and/or shifts using radiolabeled oligonucleotides with consensus sequences for each factor.

View larger version (32K): [in this window] [in a new window]   Figure 11. The abilities of specific antisera directed at CREBs, NF-B subunits, IRF-1 or -2, Stat-1 family members, Jun family members, or AP-1/Jun homodimers to modulate the ability of a 32P-radiolabeled DR 5'-flanking region probe to form protein-DNA complexes with extracts from FRTL-5 cells untreated with IFN. The probe was excised by restriction enzyme treatment of the -137 bp DR-CAT chimera and is diagrammatically represented at the bottom of Fig. 6. Extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH and treated with TSH plus 100 U/ml IFN for the last 2 days of culture. Extracts were incubated with the 32P-radiolabeled probe after a 20-min preincubation with the noted antiserum. EMSA were performed as described; the autoradiogram was exposed 72 h at -70 C. Antisera were obtained from Santa Cruz Laboratories. In lanes 2–11, the antisera used were the following: CREB-1 (C-21), sc-186; CREB-2 (C-20), sc-200; ATF-2 (C-19), sc187; ATF-3 (C-19), sc-188; ATF-1 (FI-1) sc-241; ATF-4 (Z-5), sc-244; IRF-1 (M-20), sc-640; IRF-2 (C-19), sc-498; and CBP (A22), sc-369. Other negative antisera in this series were: CREB-1 (240), sc-58; CREB-1 (24H4B), and sc-271; and CREB-1 (X-12), sc-240. In lanes 12–21, the antisera used were the following, respectively: p65 (A) G, sc-109G; p50 (NLS) G, sc-114G; p52 (H-27), sc-298; Stat-1 p84/91 (C-136), sc-464; Stat-1 p91 (C-24), sc-345; c-Jun/AP-1 (N)-G, sc-45-G; c-Jun (H-79), sc-1694; Jun-B (210), sc-73; and Jun D (329)-G, sc-74-G. Other negative antisera in this series were Stat-1 p91 (C-111), sc-417; c-Jun (KM-1), sc-822; c-Jun (79), sc-4–113; c-Jun/AP-1 (D)-G, sc-44-G; Jun B (N-17)-G, sc-46-G; Jun B (C-1), sc-1671; and AP-2 (c-18), sc-184. The arrows denote the upper and lower complexes shown in Fig. 6.

Several antisera were effective in altering the formation of the upper or basal complex, as evidenced in studies using extracts from control cells (Fig. 12). Thus, an antiserum against c-Fos caused a shift in the migration of the complex (Fig. 12, lane 5), and antisera directed toward Fra-2, Oct-1, and Ets-2 all inhibited formation of the complex and resulted in complexes with slightly slower or faster migrations when added to incubations containing extracts from cells that had not been exposed to IFN (Fig. 12, lanes 5–8 vs. 2). Antisera against Fra-1 (Santa Cruz preparations sc-605 or sc-183), Ets-1 (Santa Cruz preparations sc-111 or sc-350), or Oct-2 (Santa Cruz preparation sc-233) had no effect on either of the complexes using either extract (data not shown), indicating specificity. The same results were obtained using extracts from cells treated with IFN, i.e. there was no clear indication that the lower IFN-increased complex was altered by these antisera. However, it is evident that the migration of some of the new complexes will overlap the mobility of the IFN-increased lower complex (Fig. 12, lane 1 vs. 5–8). We included the Y box double strand oligo as a competitor to eliminate any lower complex and better track the mobility shifts in this experiment (Fig. 12, lane 4).

View larger version (46K): [in this window] [in a new window]   Figure 12. Ability of specific antisera directed at c-Fos, Ets, and Oct-1 family members to modulate the ability of the 32P-radiolabeled HLA-DR 5'-flanking region probe to form protein-DNA complexes with extracts from FRTL-5 cells treated with TSH but no IFN. The probe was the -137 bp DR-CAT-derived fragment used in Figs. 6, 7, and 9–11. Extracts were from FRTL-5 rat thyroid cells grown to near confluence in TSH; they were incubated with the 32P-radiolabeled probe after a 20-min preincubation with the noted antiserum. EMSA were performed as described in Figs. 6, 7, and 9–11; antisera were obtained from Santa Cruz Laboratories. In lanes 5–8, the antisera used were the following, respectively: c-Fos (K-25)-G, sc-253-G; Fra-2 (Q20), sc-604; Oct-1 (C-21), sc-232; and Ets-2 (C20), sc-351. The solid arrow denotes the major upper complex defined in Fig. 6; the dashed arrows denote either a supershifted complex in lane 5 or faster migrating complexes in lanes 5–7. Lane 4 shows an incubation with a 200-fold excess (300 fmol) of the Y box double strand oligonucleotide that inhibits formation of the lower complex (Fig. 10) and serves as a marker of upper complex migration. Autoradiography was performed for 36 h at -70 C.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Aberrant MHC class II expression on nonimmune cells has been associated with multiple autoimmune diseases not only ATD (4, 5, 6), i.e. diabetes and rheumatoid arthritis (56, 57). The hypothesis emerged that aberrant class II expression allowed cells to become antigen-presenting cells, interact with T cells, and initiate an immune response (4, 5, 6, 58); the hypothesis has, however, been controversial (58). In ATD, it was possible that aberrant class II expression was a secondary response to cytokines such as IFN, which were produced by lymphocytes infiltrating the tissue (58). More importantly, there was little direct evidence that class II was critical in the initiation of autoimmune thyroid disease (58). In a recent study, however, the hypothesis became more attractive (17). Thus, immunization of mice with fibroblasts transfected with the human TSHR and a MHC class II molecule, but not with either alone, induced ATD with the major humoral and histological features of Graves’ disease (17). These data established that aberrant expression of MHC class II molecules on cells expressing the TSHR could induce ATD and that acquisition of antigen-presenting ability as a result of aberrant class II expression can activate T and B cells normally present in an animal, thereby allowing normal immune tolerance to be broken.

The role of MMI as an immunosuppressant able to decrease MHC class II levels on thyrocytes can also be construed as controversial. Thus, three reports did not find that MMI could decrease MHC class II antigen expression in thyrocytes (44, 59, 60), whereas a fourth (61) did. The positive study (61) used a continuously cultured, functioning, human thyroid-hybridoma cell line. The negative studies (44, 59, 60) used human thyrocytes from normal individuals and/or Graves’ patients; however, cells were maintained in high concentrations of FCS for a number of days before testing, a condition known to not only alter or eliminate some thyroid functional responses, i.e. iodide uptake, but also alter thyroid gene expression, as evidenced, for example, in studies of dog thyrocytes by the Dumont group. The possible importance of this, particularly with respect to MHC class II, is evident in one of the negative reports (44). Thus, this report showed that iodide suppressed MHC class II in FRTL-5 thyroid cells, but not in human thyrocytes. As it has recently been shown that iodide can decrease class II in the thyroids of patients being prepared for surgery and in cultured normal human thyroid cells maintained in a functional state (62), it is a reasonable possibility that there might have been a response problem in the human thyroid cells in the three negative studies.

Under these circumstances, it seemed important to examine the control of class II expression in normal, functioning thyrocytes to understand the mechanisms underlying aberrant class II expression, as it was a potentially important aspect of understanding the pathogenesis of ATD. It also seemed reasonable to see what actions MMI might have on class II expression in thyrocytes where MMI was effective in suppressing class I gene expression (28) and where iodide could suppress class II and class I expression (28, 44). We thus used FRTL-5 thyroid cells as our thyroid culture system, as it has been used in MMI and class I promoter studies (28, 36, 37, 52, 53) as well as class II studies (21, 22, 23, 44). We used the human HLA-DR promoter, as it has been extensively studied in lymphocytes (2, 18, 19, 30), because regulation of all class II genes is usually coordinate (2, 18, 19, 30), and regulation of class II levels occurs primarily at the level of gene transcription (2, 18, 19, 30). We used IFN as an inducer of class II gene expression in FRTL-5 cells, as this, too, is well described (21, 22, 23) and mimics the ability of IFN to cause aberrant class II or increased class I expression on human thyrocytes (20, 62). Regardless of whether IFN is a primary factor in autoimmune disease expression, its role as an amplifier of the autoimmune response in ATD seemed clear in studies suggesting a "vicious cycle" among T cells, monocytes, and Graves’ or normal thyrocytes (63, 64, 65).

In the present report we establish that the HLA-DR promoter-CAT construct is normally not expressed, as is the case for the endogenous class II gene, in FRTL-5 (21, 22, 23) or human thyrocytes (4, 5, 6, 20, 44, 59, 60, 61, 62). We show, however, that IFN induces HLA-DR gene expression, as it does endogenous class II expression in FRTL-5 (21, 22, 23) or human thyrocytes (5, 6, 20, 44, 59, 60, 61, 62), and that the same elements, S, X₁, X₂, and Y boxes, are required for IFN to increase MHC class II gene expression in thyrocytes as in antigen-presenting cells from the immune system (2, 18, 19, 30). A novel finding is that IFN increases the formation of a specific, multicomponent, protein-DNA complex that is different from the basal, constitutive complex formed with the 5'-flanking region in the absence of aberrant class II gene expression. Thus, it is distinguished from the basal complex by the dominance of CBP as a component of the complex, by a particularly strong or direct interaction with the double strand Y box element of the class II promoter, as evidenced in oligonucleotide competition experiments, and by the profound effect of MMI to prevent IFN from inducing or enhancing its formation. MMI only attenuates the formation of the basal or constitutive complex.

In summary, we provide reasonable evidence that MMI can suppress aberrant MHC class II gene expression in thyrocytes as well as class I expression (28) and strengthen the supposition that MMI acts as an immunosuppressant in functioning thyrocytes. The clinical relevance of aberrant class II and abnormal class I expression in Graves’ disease as well as the importance of suppressing both MHC genes to achieve clinical remission have recently been supported by the observation that iodide suppresses both class II and class I expression, which are elevated in Graves’ thyroids (62). Iodide is an agent that, like MMI, can be used to treat Graves’ patients, albeit transiently, and has been used extensively to prepare Graves’ patients for thyroidectomies.

The present report has other novel findings relative to understanding the basis for aberrant class II expression in thyrocytes and the ability of IFN to regulate growth and differentiation in FRTL-5 cells (21, 22, 23). Thus, it has been reported that the CRE binding motif of the X₂ site is required for aberrant expression of MHC class II antigens in simian virus 40 large T antigen-transformed COS cells (66), as point mutations in the X₂ box decrease transcription 20-fold. Evidence also exists that multiple CREBs can interact with the CRE-like X2 site (67). The present data in thyrocytes are consistent with this, in that they demonstrate that ATF-1 and/or ATF-2, two CREBs, are components of the complexes. The novel finding, however, is the demonstration that CBP is a dominant component of the faster migrating IFN-increased complex and possibly contributes to the IFN-induced increase in the intensity of the upper, basal complex. CBP is a critical coactivator of binding proteins that bind to the CRE, links CREBs with basal transcription factors, and is critically involved in the regulation of cell growth and differentiation (54, 55, 68). Signal transduction by the JAK-STAT pathway and its regulation by IFN also involve CBP (68).

Work in human B cells identified an X box binding protein, hXBP-1, and showed that it could bind to the 3'-end of the x box and X-Y interspace, that it was a leucine zipper class protein with structural similarities to the c-fos and c-jun protooncogenes, that it formed a heterodimer with c-Fos, and that it was important for basal class II expression in Raji cells (69, 70). This protein regulates HLA-DR, but not HLA-DQß, gene expression (70), i.e. it may exhibit gene and tissue specificity. It is not yet known whether this protein is present in FRTL-5 thyroid cells; nevertheless, this observation would be consistent with our demonstration of c-Fos in the basal or constitutive complex. As noted earlier, a downstream octamer sequence in HLA-DR (2, 18, 19) may explain the Oct-1 interaction and its involvement in the basal complex.

In a separate study (71), we show that CIITA, a protein-protein binding factor rather than a transcription factor (67), is a mediator of the effect of IFN to cause aberrant class II expression in FRTL-5 cells. We also show that a Y box protein, which can act as a negative regulator of TSHR and MHC class I regulation in FRTL-5 thyroid cells (72, 73), is a regulatory factor of class II expression and the IFN-increased protein-DNA complex. Human YB-1 can down-regulate class II gene expression in human glioblastoma cells (74, 75). These last data support our hypothesis that coordinate suppression of MHC class I and class II genes as well as the TSHR reflects the involvement of common regulatory elements and trans factors in their expression (52).

In summary, we have shown that IFN appears to have select effects regulating the interaction of specific transcription factors with the HLA-DR 5'-flanking region in FRTL-5 cells and results in a novel, hitherto undescribed, complex with the class II gene. We show that MMI can suppress IFN-induced aberrant class II expression and that this is associated with its ability to inhibit the formation of a specific and novel IFN-increased complex with the class II gene. As a first approximation, we have been able to identify some of the transcription factors involved in the multimeric basal or constitutive complex formed with the HLA-DR 5'-flanking region in FRTL-5 thyroid cells that do not express the class II gene as well as a novel component protein in the IFN-increased complex.

	Footnotes