| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Immunology |
Department of Immunology,3 Roswell Park Cancer Institute, Buffalo, New York 14263
Department of Anatomy and Cell Biology,4 University of Kansas Medical Center, Kansas City, Kansas 66160
Department of Basic Sciences,5 Mercer University School of Medicine, Macon, Georgia 31207
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(IFN-). The absence of MHC class II antigens on trophoblasts is thought to play a critical role in preventing rejection of the fetus by the maternal immune system. The inability of trophoblasts to express MHC class II genes is primarily due to lack of the class II transactivator (CIITA), a transacting factor that is required for constitutive and IFN--inducible MHC class II transcription. We, therefore, investigated the silencing of CIITA expression in trophoblasts. In transient transfection assays, transcription from the IFN--responsive CIITA type IV promoter was upregulated by IFN- in trophoblasts, which suggests that CIITA is silenced by an epigenetic mechanism in these cells. Polymerase chain reaction analysis demonstrated that the CIITA type IV promoter is methylated in both the human choriocarcinoma cell lines JEG-3 and Jar and in 2fTGH fibrosarcoma cells, which are IFN- inducible for CIITA. Conversely, methylation of the CIITA type IV promoter was not observed in human primary cytotrophoblasts isolated from term placentae or in mouse or rat trophoblast cell lines. Simultaneous treatment with IFN- and the histone deacetylase inhibitor trichostatin A weakly activated CIITA transcription in mouse trophoblasts. Stable hybrids between human choriocarcinoma and fibrosarcoma cells and between mouse trophoblasts and fibroblasts expressed CIITA following treatment with IFN-. These results suggest that silencing of CIITA transcription is recessive in trophoblasts and involves an epigenetic mechanism other than promoter methylation. The fact that CIITA is expressed in the stable hybrids implies that trophoblasts may be missing a factor that regulates chromatin structure at the CIITA promoter.
gene regulation, immunology, trophoblast
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Trophoblasts do not express major histocompatibility complex (MHC) class II or class Ia antigens, even after exposure to interferon (IFN-) [6–12]. Aberrant upregulation of MHC class II expression in trophoblast cells has been reported in the placentas of women suffering from chronic inflammation of unknown etiology [13] and spontaneous recurrent miscarriages [14]. Inappropriate MHC class II antigen expression on trophoblasts could lead to maternal immune responses to the fetus by two different mechanisms: 1) stimulation of allogeneic rejection reactions against trophoblasts expressing paternally derived MHC class II antigens or 2) direct presentation of fetally derived peptides to maternal helper T cells by MHC class II-positive trophoblast cells. Thus, stringent silencing of MHC class II antigen expression may be essential for fetal maintenance.
Cells can be divided into three categories based upon the profile of MHC class II gene expression: 1) professional antigen-presenting cells, including dendritic cells, B cells, macrophages, and thymic epithelial cells, constitutively express class II [15, 16]; 2) fibroblast, epithelial, and endothelial cells do not normally express class II but are induced to do so in response to IFN- [15]; and 3) trophoblasts, plasma cells, and sensory neurons do not express class II antigens, either constitutively or after exposure to cytokines [6–12, 15]. MHC class II gene expression is regulated primarily at the level of transcription, and cell type-specific class II transcription is due to differential expression of a transacting factor termed the class II transactivator (CIITA) [17, 18]. Expression of CIITA is constitutive in mature B cells and dendritic cells and is activated by IFN- in fibroblast, epithelial, and endothelial cells [19–21]. The inability of trophoblast cells to express class II genes, even in the presence of IFN-, is due to the absence of CIITA [22–24]. Transfection of CIITA expression vectors into plasma cells, trophoblasts, and IFN--inducible cells such as HeLa cervical carcinoma cells results in constitutive class II gene expression [19–24]. Based on these and other studies, CIITA has been called the master regulator of MHC class II transcription [17, 18].
CIITA expression is also regulated primarily at the level of transcription by multiple promoters that function in a cell type-specific manner [25–29]. Constitutive transcription is mediated by the type I promoter in dendritic cells and macrophages and by the type III promoter in B and activated T cells [25, 26, 28, 30]. The cell-type specificity of the type II promoter is currently unclear [25]. IFN--inducible transcription is controlled primarily at the type IV promoter, although the type III promoter is weakly IFN- responsive in select cell types [25, 27–29].
IFN--inducible transcription from the CIITA type IV promoter is mediated by the Janus kinase 1 (JAK-1)/signal transducer and activator of transcription (STAT-1) signaling pathway [27–29, 31]. Binding of IFN- to its cognate receptor results in activation of JAK-1, which phosphorylates cytoplasmically localized STAT-1 [32]. Once STAT-1 is phosphorylated, it homodimerizes and translocates to the nucleus, where it activates transcription of the IFN regulatory factor 1 (IRF-1) gene and a host of other genes including that for CIITA [31, 32]. STAT-1, IRF-1, and the ubiquitously expressed upstream stimulatory factor-1 (USF-1) subsequently cooperate to activate CIITA transcription by binding to a gamma interferon-activating site (GAS), an interferon-regulatory element (IRE), and an E box located in the CIITA type IV promoter, respectively [27–29]. Binding of STAT-1 to the GAS in CIITA promoter IV is followed closely by acetylation of histones H3 and H4 [33], indicating that modifications of chromatin are necessary for CIITA transcription. The need for these modifications is further substantiated by the observation that IFN--inducible CIITA transcription requires the recruitment of the chromatin remodeling complex SWI/SNF to the CIITA type IV promoter [34].
The mechanism(s) responsible for silencing CIITA expression in trophoblasts is of clinical interest because inappropriate MHC class II expression may induce deleterious maternal immune responses against the developing fetus. Previous studies demonstrated that the CIITA type IV promoter is methylated in the human choriocarcinoma cell lines Jar and JEG-3 [33, 35, 36], and CIITA expression was activated in these cells by sequential treatment with the demethylating agent 5-azacytidine and IFN- [35, 36]. These observations led two independent groups to propose that methylation of the CIITA promoter plays a critical role in silencing CIITA transcription in trophoblast cells.
In the present study, we tested the hypothesis that CIITA is silenced in trophoblasts by an epigenetic mechanism(s). The methylation status of the CIITA type IV promoter was examined in human and rodent trophoblast cell lines and in purified primary human cytotrophoblasts isolated from term placentas. In addition, stable hybrids between trophoblasts and fibroblasts or fibrosarcomas were generated to determine whether silencing of IFN--inducible CIITA transcription in trophoblasts is dominant or recessive.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Jar, JEG-3, HeLa, SM9, and M-11 cells were grown as previously described [22]. Colon-26, NIH-3T3, and 2fTGH cells were grown in Dulbecco modified Eagle medium supplemented with 10% fetal calf serum (FCS) and penicillin/streptomycin. Human and mouse IFN- were purchased from PBL (Piscataway, NJ), and trichostatin A (TSA) were purchased from Wako (Richmond, VA). M-11, SM9, Colon-26, Jar, and JEG-3 cells were exposed to various concentrations (12.5–200 nM) of TSA for 24 h to generate a toxicity curve prior to examining the effects on CIITA gene expression. Based on these studies, the following concentrations of TSA were used in subsequent studies on gene expression: 25 nM (M-11), 50 nM (SM9, Jar, and JEG-3), and 100 nM (Colon-26). 5-Azacytidine was purchased from Sigma (St. Louis, MO), and cells were treated with 0.5–3.0 µM 5-azacytidine for 3–10 days.
Isolation and Purification of Human Villous Cytotrophoblasts
Cytotrophoblasts were purified as described previously [1] using modifications of other procedures [37, 38]. Approximately 40 g of term villous placental tissue was finely minced and subjected to three 30-min stages of digestion in a solution of trypsin and DNase. The resulting cell suspensions were layered over a discontinuous 5–70% Percoll gradient (Sigma-Aldrich) and centrifuged. The cell layer located between the densities of 1.053 and 1.060 was collected and resuspended in culture medium (Iscove modified Dulbecco medium containing 10% FCS, 100 µg/ml streptomycin, 100 U/ml penicillin) and further subjected to negative selection using anti-HLA class I antibody (clone W6/32; American Type Culture Collection, Manassas, VA) coupled to magnetic microbeads (Miltenyi Biotec, Auburn, CA). Because villous cytotrophoblasts lack HLA class I molecules on their surface, the cells in the flow-through fractions consisted of purified cytotrophoblasts. Purity of the cytotrophoblasts was further assessed by immunostaining with cytokeratin-7 antibodies (clone OV-TL; Dako, Carpinteria, CA).
Isolation of Genomic DNA and Southern Blot Analysis
Genomic DNA was isolated using a Blood and Cell Culture DNA Maxi kit (Qiagen, Valencia, CA) as recommended by the manufacturer. Aliquots of purified DNA (10 µg) from Jar, JEG-3, HeLa, and Raji cells were digested with BamHI, HindIII, KpnI, or XbaI, fractionated on 0.8% agarose gels, denatured, and transferred to Zetaprobe nylon membranes (BioRad, Hercules, CA). The human CIITA cDNA was radiolabeled by random primer labeling with Klenow fragment and purified using G-50 sepharose microspin columns (Pharmacia, Piskataway, NJ) using standard procedures. Membranes were hybridized with radiolabeled CIITA cDNA, washed, and exposed to x-ray film using standard techniques.
Plasmid Constructs, Transfections, and Luciferase Assays
DNA fragments corresponding to the human CIITA type IV promoter from -2000 to -1 and -354 to -1, respectively, were generated by polymerase chain reaction (PCR) amplification of human Raji B cell genomic DNA using the hCIITAIII-5', hCIITAIV-5' and hCIITAIV-3' primers shown below. The human CIITA type IV promoter regions were subsequently cloned upstream of the firefly luciferase gene in the pGL3-basic vector (Promega, Madison,WI) to generate pCIITAproIV(2000)luc and pCIITAproIV(354)luc. The plasmids pCIITAproIII(322)luc, pCIITAproIII(6800)luc, and pCIITAproIV(8800)luc were described previously [28, 29, 39]. HeLa, Jar, and JEG-3 cells were transfected by calcium phosphate precipitation using 5–10 µg firefly luciferase plasmid DNA as previously described [22]. The plasmid pRL-tk (1 µg; Promega), which contains the Renilla luciferase gene under the control of the HSV tk promoter, was cotransfected with all of the pCIITAproluc constructs to control for normalization of transfection efficiency. Cells were grown for 24 h in the presence of 0 and 500 U/ml IFN- and harvested for measurement of luciferase activity using a Promega Dual Luciferase kit as specified by the manufacturer. Renilla luciferase activity was measured from 5 µl of each transfected sample, and the amount of protein used in the firefly luciferase assays was normalized based on the Renilla luciferase activity.
RNA Isolation and Reverse Transcription PCR
RNA was isolated using Trizol (Invitrogen, Carlsbad, CA) as specified by the manufacturer. Reverse transcription (RT)-PCR was performed as described previously [39]. All of the primers utilized for RT-PCR analysis were described previously [20]. The numbers of PCR cycles utilized were 30–38 for CIITA, 30 for HLA-DR
, 28 for IA, and 20 for actin.
Analysis of the Methylation Status of the CIITA Promoter
Purified human genomic DNAs were digested overnight with HindIII, and aliquots were subsequently cleaved with MspI, HpaII, HhaI, AvaI, AciI, or PmlI. Mouse cell DNAs were digested with SstI followed by MspI, HpaII, HhaI, AciI, or PmlI. Aliquots (200 ng) of cleaved genomic DNA were subjected to PCR using the following primers: hCIITAIII-5' (5'-CGAAGATCTCTGCAGAAGGTGGCAGATATT-3'), hCIITAIII-3' (5'-CTAAAGCTTAGAAGCACACAGCCTCATCAC-3'), hCIITAIV-5' (5'-GGTTGGACTGAGTTGGAGAGAAACAGAGAC-3'), hCIITAIV-3' (5'-CTCCCTCCCGCCAGCTCTGGGGCCGCGGCA-3'), 5'-hsp90pr (5'-GGCGGCGATTGAGGGAAGGTTGC-3'), 3'-hsp90pr (5'-ACACCGGGACGCTGAAGCAACTGA-3'), mCIITAIII-5' (5'-CTGCAGGAGAATGTGTGTCCAATGCAATTATCATTT-3'), mCIITAIII-3' (5'-AAGCAGGCAGCCTCATCCCTCACATGCCTCTGTCTA-3'), mCIITAIV-5' (5'-GGTTGGGCTGAGATAGAGTGAAATAGAGAGAGCCAC-3'), and mCIITAIV-3' (5'-CTCCCTGCCGCCAGCTCCGTGGCTCCTAGGAGCTTG-3').
PCR of human genomic DNA was performed using the following conditions: 5 min at 95°C and 10 min at 60°C followed by 30 cycles of 90 sec at 72°C, 45 sec at 95°C, and 45 sec at 60°C. Mouse genomic DNA was amplified using 32 cycles with the same conditions as used for the human DNA. The human CIITA type III and type IV promoter primers gave rise to 322- and 354-base pair (bp) products, respectively, whereas the mouse type III and type IV promoter PCR products were 255 and 372 bp, respectively. PCR products were resolved on 1.5% agarose gels and visualized by ethidium bromide staining. For each cell type examined in these studies, at least three different preparations of genomic DNA were tested for the methylation status of the CIITA promoter.
Generation of Stable Hybrids
The cell lines to be used in these studies were first transfected with drug-resistance markers to allow for selection of double drug-resistance following fusion to ensure that the hybrids were bona fide heterokaryons. Trophoblast cells (human Jar and JEG-3 and mouse M-11) were transfected with pcDNA-3, which contains the bacterial neomycin gene under the control of the SV40 promoter, and selected for G418 resistance using 400 µg/ml G418 sulfate (Invitrogen). Cell lines that are IFN- inducible for CIITA expression (human 2fTGH and mouse NIH-3T3) were transfected with pPUR (Clontech, Palo Alto, CA), which contains the puromycin-resistance gene under the control of the SV40 promoter/enhancer, and selected with 1.0 µg/ml puromycin (Clontech). Drug-resistant clones of each of the respective cell lines were isolated using cloning cylinders, expanded, and tested for IFN--inducible expression of IRF-1, CIITA, and HLA-DR mRNAs to ensure that they maintained the parental phenotype. Cell fusions were performed by mixing 8 x 106 cells each of trophoblast and CIITA-inducible cells with polyethylene glycol (Mr 1450; Sigma) as described by Coady et al. [40]. Fused cells were exposed to both G418 and puromycin 24 h after fusion. Generation of stable hybrids differed dramatically depending upon the cell lines utilized. Several hundred drug-resistant Jar/2fTGH hybrid clones appeared approximately 3 wk after initiation of selection, and 12 colonies were isolated and expanded. The remainder were combined together to generate mixed clones. Approximately 50 clones of JEG-3/2fTGH stable hybrids were obtained, 8 were isolated and expanded, and the remainder were combined to generate mixed colonies. In contrast, only nine clones of M-11/NIH-3T3 hybrids were successfully generated and expanded.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
One potential explanation for the inability of Jar and JEG-3 choriocarcinoma cells to express CIITA following exposure to IFN- is that the CIITA gene was deleted or mutated during transformation of these cells. To address this possibility, genomic DNA was isolated from Jar and JEG-3 cells and subjected to Southern blot analysis using radiolabeled full-length human CIITA cDNA as a probe. HeLa and Raji cell genomic DNAs were included as positive controls for cells that express CIITA. DNA bands of identical sizes were detected from Jar and JEG-3 choriocarcinoma cells and CIITA-expressing HeLa and Raji cells when genomic DNA cleaved with the restriction enzymes HindIII, XbaI, KpnI, or BamHI was hybridized with radiolabeled CIITA cDNA (Fig. 1A). The sizes of the HindIII fragments correspond to those previously reported [41].
|
To examine the integrity of the CIITA promoters type III (PIII) and type IV (PIV) in human choriocarcinoma cells, genomic DNA from Jar and JEG-3 cells was subjected to PCR using promoter-specific primers. The PCR products from Jar and JEG-3 cells were identical in size (322 bp for the PIII and 354 bp for PIV, respectively) to those observed in HeLa and Raji cells (Fig. 1B), indicating that the CIITA promoters are also intact in Jar and JEG-3 cells. Sequencing of the PCR products from Jar and JEG-3 cells confirmed that they were amplified from the CIITA PIII and PIV, respectively (data not shown). The results shown in Figure 1 indicate that the CIITA coding region and promoters are grossly intact in the human choriocarcinoma cell lines Jar and JEG-3, which suggests that the lack of IFN--inducible CIITA gene expression in these cells is due to silencing of transcription.
CIITA PIV Is IFN- Inducible in Transient Transfection Assays of Human and Mouse Trophoblast Cells
CIITA PIV is the predominant IFN--responsive promoter in mammalian cells, although the PIII is also weakly IFN- inducible in select cell types [27–29]. To investigate the mechanism by which CIITA gene transcription is regulated in trophoblasts, transient transfection assays of human JEG-3 cells were performed using luciferase vectors containing 0.322 and 6.8 kb of human CIITA PIII sequence (Fig. 2A) or 0.354, 2.0, and 8.8 kb of human CIITA PIV sequence (Fig. 2B). HeLa cells were included as a positive control for a cell type that expresses CIITA in response to IFN-. The pCIITApIII(322)luc plasmid contains the minimal sequences necessary for B cell-specific CIITA transcription [25, 42], whereas the pCIITAproIII(6800)luc plasmid contains an upstream IFN--responsive GAS site as previously described [29], and all of the PIV constructs contain the GAS, E box, and IRE sites required for IFN--induced CIITA transcription. Cells were transfected with the luciferase vectors and subsequently incubated for 24 h in the presence of 0 and 500 U/ml IFN-.
|
As expected, IFN- had no effect in any of the cell lines on the luciferase activity from pCIITAproIII(322)luc (Fig. 2A). Similarly, luciferase activity was not significantly induced by IFN- from the 6.8-kb PIII (pCIITAproIII(6800)luc) in either HeLa or JEG-3 cells (Fig. 2A). Similar results were obtained in transfection studies with mouse NIH-3T3 fibroblasts, Colon-26 carcinomas, and M-11 trophoblasts (data not shown). These data suggest that PIII is inactive in trophoblasts, fibroblasts, and HeLa cells and are consistent with results of previous studies showing cell-type selective IFN- inducibility from the PIII [26, 43, 44].
In contrast to the results with CIITA PIII, IFN- upregulated luciferase activity an average of 8.8 ± 1.4-fold and 8.0 ± 1.8-fold in HeLa and JEG-3 cells transfected with vectors containing 0.354 kb of PIV sequence, respectively (Fig. 2B). Similarly, an average of 8.2 ± 2.0-fold and 7.5 ± 1.7-fold increases in IFN--inducible luciferase activity were observed in HeLa and JEG-3 cells transfected with 2.0 kb of CIITA PIV. In HeLa and JEG-3 cells transfected with pCIITAproIV(8800)luc, luciferase was induced 5.6 ± 1.8-fold and 6.4 ± 2.0-fold, respectively. Similar results were observed in transfections of Jar cells with pCIITAproIV(354)luc and pCIITAproIV(8800)luc, respectively (data not shown). Thus, the activation of luciferase activity by IFN- from the CIITA PIV constructs was comparable in Jar and JEG-3 cells to that in HeLa cells. Similar results were obtained in transient transfection assays of mouse M-11 trophoblasts compared with NIH-3T3 and Colon-26 cells, although the induction by IFN- was less (3- to 4-fold; data not shown). These results indicate that the IFN--responsive CIITA PIV is functionally active in transient transfection assays of trophoblast cells and that trophoblastic silencing of CIITA transcription is not due to the presence of negative regulatory elements in the 8.8-kb upstream regulatory region.
Methylation of the CIITA PIV Is Not Correlated with CIITA Silencing in Human and Rodent Trophoblast Cells
The results of the transient transfection assays suggest that the mechanism responsible for silencing CIITA expression in trophoblast cells is epigenetic, such as methylation of the CIITA promoter and/or insufficient acetylation of histones. To investigate the methylation status of the CIITA upstream regulatory region, genomic DNA was isolated from human Jar and JEG-3 choriocarcinoma cells and subsequently digested with the methylation-sensitive enzymes HpaII and HhaI or the methylation-insensitive MspI. Genomic DNA from Raji B cells, HeLa cells, and 2fTGH fibrosarcoma cells were included as controls for CIITA-expressing cells. Restriction enzyme-digested DNA was subjected to PCR using primers that span regions of the CIITA upstream regulatory region containing the methylation-sensitive enzyme recognition sites (Fig. 3A). PCR products were only detected in these assays when either HhaI or HpaII failed to cleave the DNA because of methylation of the respective recognition sites. No PCR products were detected when genomic DNA from Raji or HeLa cells cleaved with either HpaII or HhaI was subjected to PCR using primers spanning the 354-bp PIV (Fig. 3B), indicating that these enzyme restriction sites were not methylated. In contrast, PCR products were clearly detected from the DNA samples of Jar and JEG-3 cells digested with HpaII or HhaI, indicating that these enzymes did not cut. Methylation of the HpaII and HhaI sites was also observed in genomic DNA isolated from 2fTGH fibrosarcoma cells (Fig. 3B) even though these cells express CIITA after exposure to IFN- [21, 28]. AciI and AvaI recognition sites were also selectively methylated in genomic DNA from Jar, JEG-3, and 2fTGH cells but not in genomic DNA from Raji or HeLa cells (data not shown). PCR bands were not observed from any of the DNA samples cleaved with methylation-insensitive MspI, indicating that this enzyme effectively cut the DNA at the CIITA PIV (Fig. 3B).
|
To ensure that the HpaII and HhaI enzymes were effective in digesting nonmethylated genomic DNA, all of the samples were subjected to PCR using primers spanning HhaI and HpaII sites within the promoter of the heat shock protein 90 (hsp90) gene, which is a constitutively active promoter that is not methylated [45, 46]. No hsp90pr PCR products were observed from Jar, JEG-3, or 2fTGH genomic DNA digested with either enzyme, indicating that they cut effectively at nonmethylated recognition sites (Fig. 3B). As an additional control for DNA integrity and quantity, all samples were subjected to PCR with primers spanning the CIITA PIII from -322 to +1 relative to the transcriptional start site, a region that does not contain any HpaII, HhaI, or MspI sites. Comparable levels of PCR products were observed from all of the samples (Fig. 3B), indicating that the DNA was intact. Based on these results, we concluded that the IFN--responsive CIITA PIV is methylated in Jar and JEG-3 choriocarcinoma cells and in 2fTGH fibrosarcoma cells, which express CIITA in response to IFN-. In contrast, this CIITA promoter is unmethylated in Raji and HeLa cells.
Methylation of the CIITA promoter in Jar and JEG-3 cells could be a reflection of 1) the normal trophoblastic phenotype, 2) the fact that the cells are transformed, or 3) long-term culturing of these cells, as previously shown for many other genes [47, 48]. To distinguish among these possibilities, the PCR assay described above was used to examine the methylation status of the CIITA PIV in primary human cytotrophoblasts isolated from a total of six different term placentas. To obtain maximal cell purity, isolated cytotrophoblasts were subjected to negative selection with MHC class I antibodies linked to magnetic beads, and purity was assessed by immunostaining with antibodies to cytokeratin. The purity of the six preparations of cytotrophoblasts examined was 90–96% (data not shown). Neither the HhaI nor HpaII sites in the CIITA PIV were methylated in primary human cytotrophoblasts (Fig. 3C). Identical results were observed with genomic DNA from all six cytotrophoblast preparations and when these preparations of cytotrophoblast genomic DNA were cleaved with the methylation-sensitive enzymes AciI, AvaI, and PmlI (data not shown).
The methylation status of the CIITA PIV was also examined in mouse A20 B cells, Colon-26 colon adenocarcinoma cells, NIH-3T3 fibroblasts, and the trophoblast cell lines M-11 and SM9. A20 cells express CIITA constitutively, whereas Colon-26 and NIH-3T3 cells are IFN- inducible for CIITA and MHC class II mRNA expression. Genomic DNA was digested with the methylation-sensitive restriction enzymes HhaI, AciI, and PmlI (Fig. 4A), and PCRs were performed using CIITA PIV primers. As observed with genomic DNA from the human primary cytotrophoblasts, none of these restriction sites were methylated at the CIITA promoter in A20, Colon-26, M-11, or SM9 cells (Fig. 4B). However, methylation of the HhaI site was detected in NIH-3T3 fibroblasts (Fig. 4B), despite the fact that these cells express CIITA in response to IFN-. In separate studies, methylation of the HhaI site was also not observed in the rat trophoblast cell lines HRP-1 or LRP-2 (data not shown). These results strongly suggest that CIITA PIV is not methylated in either primary human trophoblasts or rodent trophoblast cell lines.
|
Consistent with the lack of correlation between CIITA promoter methylation and silencing of CIITA transcription in trophoblast cells, treatment of Jar and JEG-3 cells with the demethylating agent 5-azacytidine had no effect on the ability of these cells to express CIITA in response to IFN-, even with a variety of concentrations (0.5–3.0 µM) and treatment times (3–10 days) (data not shown). Identical results were obtained in studies with SM9 mouse trophoblast cells treated sequentially with 5-azacytidine and IFN- (data not shown). In contrast, studies performed simultaneously in our laboratory demonstrated that CIITA and MHC class II mRNA expression was readily detected in L1210 lymphoma cells treated with 5-azacytidine [40], indicating that the activity of the 5-azacytidine was not compromised in these experiments. The results of these studies demonstrate that methylation of the CIITA promoter is not correlated with silencing of IFN--inducible CIITA transcription in trophoblasts.
CIITA Is Expressed in Mouse Trophoblast Cells Treated with IFN- and Histone Deacetylase Inhibitors
The acetylation status of histones bound to gene promoters plays a key role in both constitutive and inducible gene transcription [49, 50]. Histone acetylation decreases the binding affinity of histones for DNA and leads to a more "open" chromatin structure and increased access of DNA regulatory elements to their requisite transacting factors [49, 50]. The overall levels of histone acetylation are regulated by the equilibrium between the activity of two types of enzymes: histone acetyltransferases (HATs) and histone deacetylases [49, 50]. To investigate whether changes in histone acetylation alter the ability of mouse trophoblasts to express CIITA, M-11 and SM9 cells were simultaneously treated with IFN- and the histone deacetylase inhibitor TSA for 24 h. Although IFN- alone had no effect on CIITA expression in M-11 or SM9 cells, low levels of CIITA mRNA were consistently detected in M-11 cells treated with TSA alone (Fig. 5). CIITA expression was reproducibly induced in mouse trophoblast cell lines simultaneously treated with TSA and IFN- for 24 h, although the levels of CIITA mRNA were reproducibly higher in M-11 cells than in SM9 cells (Fig. 5). IFN- activated CIITA mRNA expression in Colon-26 cells, but TSA had no effect on CIITA expression in these cells, either alone or in combination with IFN- (Fig. 5). However, as shown previously [51], TSA treatment for 24 h directly upregulated MHC class II mRNA expression in Colon-26 and SM9 cells, consistent with studies showing that CIITA both recruits HATs to the class II promoters and acts as an HAT itself [52–54]. In contrast, IA mRNA was not observed in Colon-26 cells treated with IFN- until 48 h after initiation of treatment (data not shown). TSA treatment had no effect on expression of either CIITA or MHC class II (HLA-DR) gene transcription in either Jar or JEG-3 choriocarcinoma cells (data not shown). In conclusion, mouse trophoblast cells, but not human choriocarcinoma cells, express CIITA after simultaneous treatment with IFN- and histone deacetylase inhibitors.
|
Silencing of CIITA Gene Expression in Trophoblasts Is Recessive
Stable heterokaryons have previously been generated between multiple different MHC class II-expressing and nonexpressing cells to investigate the genetic mechanisms regulating CIITA and MHC class II gene expression [39, 40, 55–58]. To determine whether the silencing of CIITA expression observed in trophoblasts is dominant or recessive, stable heterokaryons between trophoblasts and CIITA-inducible cells were generated using polyethylene glycol. Individual G418-resistant clones of Jar and JEG-3 were fused to a puromycin-resistant clone of the CIITA-inducible human fibrosarcoma line 2fTGH, and a G418-resistant clone of mouse M-11 trophoblasts was fused with a puromycin-resistant NIH-3T3 fibroblast clone. Stable hybrids were selected with G418 and puromycin as described in Materials and Methods.
To assess the ability of stable Jar/2fTGH hybrids to express CIITA and MHC class II mRNA in response to IFN-, mixed clones and four individual clones were grown for 24 h in the presence of 0 or 1000 U/ml IFN-. RNA was isolated and subjected to RT-PCR using primers for CIITA and HLA-DR. The Jar and 2fTGH clones used in the fusions retained their respective parental phenotypes; CIITA and HLA-DR mRNA were clearly expressed in IFN--treated 2fTGH clone 3 but not in Jar clone 3 treated with IFN- (Fig. 6A). Interestingly, the stable hybrids between Jar and 2fTGH expressed both CIITA and HLA-DR in response to IFN-: this was observed in both the mixed hybrid clones and four different hybrid clones (Fig. 6A; data not shown). As observed in the parental cell lines, the CIITA PIV was methylated in the Jar/2fTGH hybrids (Fig. 3B), although the CIITA gene is expressed in response to IFN-. Identical results were obtained for mixed clones and six individual clones of the JEG-3/2fTGH stable hybrids (data not shown). In addition, CIITA and MHC class II (IA) mRNA was also induced by IFN- in six clones of stable hybrids between mouse M-11 trophoblasts and NIH-3T3 fibroblasts (Fig. 6B; data not shown). These results suggest that silencing of IFN--inducible CIITA expression is recessive in trophoblast cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reporter gene assays demonstrated that the IFN--responsive CIITA PIV is functionally active in trophoblast cells in the absence of chromatin [35, 36; Fig. 2) and that trophoblasts may have the DNA-binding proteins (STAT-1, USF-1, and IRF-1) necessary for IFN--inducible CIITA transcription. The finding that the IFN--mediated upregulation of luciferase activity from pCIITAproIV(8800)luc is comparable in trophoblasts, HeLa cells, and fibroblasts (Fig. 2; data not shown) suggests that trophoblast-specific silencing of CIITA gene transcription probably is not due to the activity of negative regulatory element(s) in the 8.8-kb CIITA upstream regulatory region. Both in vivo genomic footprinting and chromatin immunoprecipitation assays indicate that neither STAT-1 nor IRF-1 are bound in vivo to the GAS and IRE sites in IFN--treated Jar or JEG-3 cells [33, 35]. Moreover, histones H3 and H4 are not acetylated at the CIITA PIV in Jar or JEG-3 cells after exposure to IFN- [33]. CIITA is weakly expressed in mouse trophoblasts simultaneously treated with IFN- and the histone deacetylase inhibitor TSA (Fig. 5). These results strongly suggest that CIITA PIV is in a closed chromatin conformation in trophoblasts and is silenced by an epigenetic mechanism in these cells.
Morris et al. [33, 35] and van den Elsen et al. [36] previously reported that CIITA PIV is methylated in Jar and JEG-3 choriocarcinoma cells and that sequential treatment of these cells with the demethylating agent 5-azacytidine and IFN- weakly activated CIITA expression. Many researchers have shown that DNA methylation can play an important role in the generation and/or maintenance of a closed chromatin conformation (reviewed in [61, 62]). These results prompted two different research groups to propose that methylation of the CIITA promoter plays a central role in silencing CIITA transcription in trophoblast cells. Although CIITA promoter methylation in Jar and JEG-3 cells was also observed in this study, the collective results suggest that promoter methylation is not the cause of CIITA gene silencing in trophoblasts. This conclusion is based on the following observations: 1) methylation of the CIITA promoter is not observed in primary human trophoblasts freshly isolated from term placentas, even though these cells do not express CIITA; 2) the PIV is methylated in human 2fTGH fibrosarcoma cells, which express CIITA in response to IFN-; 3) stable hybrids between trophoblasts and 2fTGH fibrosarcomas or NIH-3T3 fibroblasts express CIITA after IFN- treatment, even though the PIV is methylated in the human hybrids (Fig. 3B); 4) the PIV is not methylated in murine trophoblasts; and 5) treatment of human choriocarcinoma or a mouse trophoblast cell line with the demethylating agent 5-azacytidine had no effect on IFN--inducible CIITA expression, despite testing multiple concentrations and treatment times. Taken together, these results provide strong support for the conclusion that methylation of the CIITA promoter is not responsible for silencing CIITA gene transcription in trophoblasts.
These findings are similar to previous observations on the regulation of MHC class I gene expression in human trophoblast cells. The MHC class I genes HLA-A and HLA-B are also silenced in trophoblasts, even after IFN- treatment [9, 11, 63, 64]. Early studies demonstrated that the HLA-A and HLA-B loci were methylated in Jar cells and that treatment with 5-azacytidine activated MHC class I expression [65–67]. However, in a subsequent study, methylation of the MHC class I genes was not observed in primary human trophoblasts isolated from term placentas, despite the fact that these cells also lack MHC class I expression [68]. These studies and ours are consistent with a scenario in which methylation of the HLA-A, HLA-B, and CIITA gene promoters in Jar and JEG-3 choriocarcinoma cells is a consequence of either the transformed phenotype of these cells or long-term cell culture.
Correlations between CIITA promoter methylation, transcriptional silencing, and activation of expression by 5-azacytidine treatment have also been reported in human teratocarcinoma and erythroleukemia lines [69] and mouse cell lines derived from transformed pro-B and pre-B cells [39]. These results implied that CIITA may be silenced by promoter methylation in multiple cell lineages during development. However, the methylation status of the CIITA promoter has not yet been characterized in the normal cellular counterparts of these tumor types. This is an important question because de novo methylation of tissue-specific genes has been observed in established cell lines, even though these genes were not methylated in the corresponding normal, nonexpressing tissues [47, 48]. DNA damaging agents alleviate silencing of CIITA expression in several transformed mouse cell lines derived from pro-B and pre-B cells, but not in Jar, JEG-3 cells or K562 cells [39]. These studies suggest that the modes of silencing CIITA transcription may differ between trophoblast/choriocarcinoma and lymphoma lines.
The studies of the stable hybrids between trophoblasts and fibroblasts or fibrosarcomas (Fig. 6) indicate that silencing of IFN--inducible CIITA transcription is recessive in trophoblasts. These results contrast with those of studies in which suppression of constitutive transcription from the CIITA PIII was dominant when B cells were fused with trophoblast cells [40], plasma cells [55, 57], fibroblasts, or epithelial cells [58]. Luciferase assays demonstrated that the CIITA PIII was completely inactive in trophoblasts, whereas PIV was responsive to IFN- (Fig. 2). Thus, the CIITA PIII and PIV may be silenced by different mechanisms in trophoblasts.
One explanation for the recessive nature of the silencing of IFN--inducible CIITA transcription is that trophoblast cells may lack a novel essential transacting factor(s) whose expression is dominant in the stable hybrids. Based on the observation that histones are not acetylated at the CIITA promoter in Jar or JEG-3 cells [33] and that TSA partially alleviates silencing of IFN--inducible CIITA expression in mouse trophoblasts (Fig. 5), this putative factor may play a crucial role in controlling the chromatin structure at the CIITA PIV. One possibility consistent with this model is that trophoblast cells have reduced or defective histone acetyltransferase activity. Alternatively, trophoblasts may have defects in components of SWI/SNF, the ATPase-dependent chromatin remodeling complex that is required for IFN--inducible CIITA transcription [34]. Loss of BRG1/BRM, the ATPase subunit of SWI/SNF, abrogates IFN--inducible CIITA expression [34]. However, preliminary Western blotting studies indicate that BRG1/BRM is expressed in both mouse and human trophoblast cell lines (data not shown). An alternative explanation for the recessive nature of CIITA silencing is that trophoblasts contain a factor that represses CIITA transcription, perhaps by recruiting histone deacetylases. In this scenario, expression of this putative repressor factor would be extinguished in the hybrids. The cell type-specific factor BLIMP-1/positive regulatory domain 1-binding factor 1 downregulates CIITA transcription in plasma cells, most likely by recruiting histone deacetylases [70, 71]. A third but less likely possibility is that one allele of CIITA is transcribed in IFN--treated 2fTGH cells and the stable hybrids, and the other allele is silenced by methylation. IL-2 and IL-4 transcription is regulated by a similar mechanism in T cells [72, 73].
This study provides evidence that methylation of the CIITA PIV is not responsible for silencing IFN--inducible CIITA transcription in trophoblasts. The lack of histone acetylation at the CIITA promoter in IFN--treated Jar and JEG-3 cells [33] and the fact that CIITA transcription is activated in mouse trophoblasts treated with IFN- and TSA (Fig. 5) are consistent with an epigenetic mechanism of CIITA silencing in trophoblasts. These observations and the fact that IFN- induces CIITA expression in stable hybrids of trophoblasts and fibroblasts/fibrosarcomas support the hypothesis that trophoblasts lack a factor that functions to induce an open chromatin structure at the CIITA promoter.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Shawn P. Murphy, Department of Immunology, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263. FAX: 716 845 8906; shawn.murphy{at}roswellpark.org
Received: 12 March 2003.
First decision: 27 March 2003.
Accepted: 8 May 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-interferon in cultured human trophoblasts. Am J Obstet Gynecol 1987 157:429-434[CrossRef] receptor in first trimester placenta to trophoblasts but lack of stimulation of HLA-DRA, -DRB, or invariant chain mRNA expression of IFN-. J Immunol 1992 149:2675-2680[Abstract] mediated by the transactivator gene CIITA. Science 1994 265:106-109 mutants reveals that defects in CIITA or RFX result in defective class II MHC and Ii induction. Immunity 1994 1:687-697[CrossRef][Medline]-inducible promoter (PIV) of CIITA. Hum Immunol 2000 61:850-862[CrossRef][Medline] requires cooperative interaction between Stat1 and USF-1. Immunity 1998 8:157-166[CrossRef][Medline], STAT1, and TGF-ß-regulated expression of the class II transactivator gene. J Immunol 1998 160:233-240 expression is involved in IFN- induction of the class II transactivator and class II MHC genes. J Immunol 1996 157:1559-1568[Abstract]-induced chromatin remodeling at the CIITA locus is BRG1 dependent. EMBO J 2002 21:1978-1986[CrossRef][Medline] and IRF-1 in astrocytes and macrophages: CIITA promoter activation is not affected by TNF-. J Neuroimmunol 1999 99:195-204[CrossRef][Medline]-inducible CIITA expression and can function as nonprofessional antigen presenting cells in endocytic processing and CD4+ T-cell activation. Glia 2001 36:391-405[CrossRef][Medline]
