Posttranscriptional Regulation of Human Leukocyte Antigen G During

1 downloads 0 Views 1003KB Size Report
peptide from the 1 domain of HLA-G and is HLA-G spe- cific [34]. The mouse ...... IL, Fauchet R, Alsat E, Le Bouteiller P. Methylation status and tran- scriptional ...
BIOLOGY OF REPRODUCTION 62, 1543–1550 (2000)

Posttranscriptional Regulation of Human Leukocyte Antigen G During Human Extravillous Cytotrophoblast Differentiation1 James Copeman,3 Robin N.N. Han,3 Isabella Caniggia,3,4 Michael McMaster,6 Susan J. Fisher,6,7 and James C. Cross2,3,4,5 Samuel Lunenfeld Research Institute,3 Mount Sinai Hospital, Toronto, Ontario, Canada M5G 1X5 Department of Obstetrics and Gynaecology4 and Department of Molecular and Medical Genetics,5 University of Toronto, Toronto, Ontario, Canada M5G 1X5 Department of Stomatology6 and Departments of Anatomy, Pharmaceutical Chemistry, and Obstetrics, Gynecology and Reproductive Sciences,7 University of California San Francisco, San Francisco, California 94143 ABSTRACT Human maternal tolerance to a semiallogenic fetus may be maintained, in part, by the unusual expression pattern of antigen-presenting molecules in placental trophoblast cells. Extravillous cytotrophoblast (EVC) cells, which invade the maternal decidua, express high levels of human leukocyte antigen G (HLA-G), a nonclassical, major histocompatibility complex (MHC) class I molecule. HLA-G transcripts have been detected in tumors and other tissues, yet protein accumulation is rare. We show that, within EVC cells themselves, the mRNA is more broadly expressed than the protein. Specifically, accumulation of HLA-G protein was markedly delayed during EVC cell differentiation. To elucidate this mechanism, we performed a comprehensive analysis comparing the expression of HLA-G and proteins essential for MHC class I expression at the cell surface. The transporter for antigen processing proteins TAP1 and TAP2, as well as tapasin and b2-microglobulin, appeared to be coordinately expressed throughout EVC cell columns. Strikingly, they all accumulated well in advance of the HLA-G protein but concurrently with its mRNA. A similar delay in the accumulation of the HLA-G protein was observed in vitro, using cultures of chorionic villi. We conclude that posttranscriptional regulation of HLA-G is fundamental to EVC cell development and is achieved independently of the peptide loading system. This represents a novel mechanism of MHC class I regulation.

INTRODUCTION

Expression of major histocompatibility complex (MHC) class I molecules, and their subsequent interaction with T cells, natural killer cells, and other cell types, are fundamental to immune competence and self-tolerance. During pregnancy a contradictory situation seems to occur at the fetal-maternal interface. The fetus is semiallogenic, yet it avoids rejection by the maternal immune system without seriously compromising maternal immune function. This may be the result, in part, of the unusual pattern of MHC class I expression at the point of contact between the mother and fetus: the trophoblast cells of the placenta. The first-trimester human placenta is made up This work is supported by a grant from the Medical Research Council of Canada to J.C.C. J.C. is a Research Fellow and J.C.C. a Scholar of the MRC Canada. 2 Correspondence: James C. Cross, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, ON, Canada M5G 1X5. FAX: 416 586 8588; e-mail: [email protected] 1

Received: 17 August 1999. First decision: 27 September 1999. Accepted: 22 December 1999. Q 2000 by the Society for the Study of Reproduction, Inc. ISSN: 0006-3363. http://www.biolreprod.org

of branching chorionic villi [1]. These are composed of a stromal core surrounded by two layers of trophoblast. The inner layer is of progenitor villous cytotrophoblast (VC) cells, which fuse to form the outer layer of syncytiotrophoblast. At discrete points on villus tips, VC cells proliferate, forming columns of extravillous cytotrophoblast (EVC) cells. In the distal portion of these columns, the EVC cells differentiate to adopt an invasive phenotype. They penetrate the maternal decidua where they remodel the maternal vasculature. A cross section of a column therefore reveals a spectrum of EVC cell differentiation stages. Most human trophoblast cells express no MHC molecules, presumably reducing the risk of allorecognition. However, EVC cells in the distal parts of columns express high levels of the placental-specific nonclassical class I molecule, human leukocyte antigen G (HLA-G) [2–4]. While the HLA-G protein has been detected only in EVC cells, endothelial cells of chorionic capillaries, and thymic epithelial cells [5–7], the mRNA has been detected in a range of tissues and tumors [8–12]. This indicates that a critical posttranscriptional mechanism restricts its expression. Given the delicate immunological balance at the fetalmaternal interface, efficient regulation of HLA-G expression may be of great clinical relevance. Preeclampsia is a common disease of pregnancy that complicates 5–7% of all pregnancies and is one of the world’s leading causes of maternal and neonatal morbidity [1, 13]. The disease is associated with abnormal differentiation of EVC cells [14– 17] and reduced levels of HLA-G [13, 16, 18]. While many of the EVC cell differentiation defects may be explained by the low oxygen tension that characterizes preeclamptic placentas, the reduced expression of HLA-G cannot [19]. Reduced HLA-G and hypoxia may therefore be independently associated with the preeclamptic phenotype. The mechanisms regulating HLA-G expression are unknown. Among trophoblast cells the protein is clearly restricted to EVC cells [2–4], whereas mRNA expression has been reported in VC as well as EVC cells [20, 21]. Although the VC cell signal was observed only inconsistently, it suggested that HLA-G expression might be regulated posttranscriptionally. Surprisingly, direct comparison of protein and mRNA expression by colocalization or analysis of serial histological sections has never been described. A number of molecules are essential for full cell surface expression of class I molecules: b2-microglobulin (b2m) is the binding partner of class I molecules [22]; the transporter for antigen processing proteins TAP1 and TAP2 together transport peptides into the endoplasmic reticulum [23, 24]; and tapasin facilitates the loading of peptide onto class I molecules [25–29]. Regulation of any of these factors would influence the maturation of HLA-G [25, 30] and so

1543

1544

COPEMAN ET AL.

could provide a mechanism for its posttranscriptional regulation. Expression of TAP1 protein and mRNA has previously been reported to be confined to EVC cells [31, 32]. Consequently the hypothesis was proposed that restricted peptide loading, caused by an absence of TAP1, was the mechanism of HLA-G posttranscriptional regulation. However, direct comparisons of TAP1 and HLA-G expression to test this hypothesis have not been described. TAP2 mRNA has been detected in whole placentas but has not been shown in any trophoblast cell types specifically [33]. Tapasin and b2m have never been examined in a placental context. We therefore set out to perform the first comprehensive analysis in human chorionic villi of the relative expression patterns of HLA-G mRNA and protein, as well as proteins essential for its accumulation at the cell surface. Specifically we wished to determine whether the expression of HLA-G was posttranscriptionally regulated within the cell lineage that accumulates this protein, and whether the limited expression of any other protein could account for such regulation. MATERIALS AND METHODS

Antibodies

The mouse monoclonal antibody 4H84 was raised to a peptide from the a1 domain of HLA-G and is HLA-G specific [34]. The mouse monoclonal antibody HCA2 (a gift from Dr. H. Ploegh, Harvard Medical School, Boston, MA) recognizes HLA-G and HLA-A [25, 35, 36]. HLA-A has never been observed in first-trimester human trophoblast cells [3, 5, 37], so in this cell type HCA2 can be considered HLA-G specific. BBM1 is a mouse monoclonal antibody recognizing b2m [38] (a gift from Dr. J. Chamberlain, University of Toronto, ON). All other antibodies were gifts from Dr. P. Cresswell (Yale University, New Haven, CT). RING4C is a rabbit antiserum [39], and 148.3 is a mouse monoclonal antibody [40], both recognizing TAP1; 435.3 is a mouse monoclonal recognizing TAP2 [41]. TAP2B is a rabbit antiserum raised to the C-terminal peptide of the TAP2B allele. MEPsinC is a mouse monoclonal antibody raised to the C-terminal peptide of tapasin.

Fast 3,39-diaminobenzidine tablet sets (Sigma, Oakville, ON, Canada), the necessary incubation times being determined for each combination of antibodies. Sections were counterstained in 0.1% Toluidine Blue in 10 mM sodium citrate, pH 4.6, dehydrated in graded ethanol and xylene, and mounted in Cytoseal 60 (Stephens Scientific, Riverdale, NJ). Images were captured using a Leitz DMRD microscope (Leica, Postfach, Germany) with a CCD camera and a computer with Northern Eclipse (Empix Imaging, Mississauga, ON, Canada) software. The background of each image was neutralized using a Boolean function. This reduces noise very effectively, but leaves the image dark. We therefore used Adobe Photoshop (Mountain View, CA) software to increase the brightness and contrast of the whole of each image. In Situ Hybridization

Riboprobes were generated from a template of a 450base pair (bp) PvuII fragment from the 39 untranslated region of the HLA-G cDNA [4] cloned into pBluescript KS1. This fragment has been used in numerous other studies and has always been found to be specific [4, 20, 32, 42, 43]. Antisense and sense control RNA probes were transcribed in the presence of [35S]dUTP using either T3 or T7 promotors (RNA transcription kit; Promega, Madison, WI) according to manufacturer’s instructions. Hybridization procedures were as described previously [44]. Briefly, prehybridized sections were hybridized overnight at 558C with the labeled probes at a concentration of 1 3 106 counts/min per ml. Sections were washed with hybridization buffer (558C, 10 min), RNase A (378C, 30 min), twice with double-strength SSC (single-strength SSC is 0.15 M sodium chloride and 0.015 M sodium citrate; room temperature, 30 min), four times with double-strength SSC (658C, 30 min), and with 0.1-strength SSC (658C, 15 min). Sections were dehydrated in graded ethanol and air dried at room temperature. Slides were coated in NBT-2 photographic emulsion (Eastman Kodak, New Haven, CT) and incubated in the dark for 14 days at 48C before being developed. Sections were counterstained with Carazzi’s hematoxylin, then dehydrated, mounted, and examined as above.

Immunohistochemistry

Chorionic villous samples were obtained from elective first-trimester terminations at Mount Sinai Hospital, Toronto. Samples were fixed overnight at 48C in 4% paraformaldehyde. They were washed in PBS (2 times, 4 h each, and overnight); they were then dehydrated in graded ethanol and xylene and embedded in paraffin. Sections (5 mm) were cut onto slides coated in 3-aminopropyltriethoxysilane. Sections were dewaxed in xylene and rehydrated in graded ethanol to PBS. Antigen retrieval was performed by incubating the sections in 10 mM citrate buffer (8.2 mM sodium citrate/1.8 mM citric acid) and heating in a microwave at 80% power for 5 min and then 4 times for 2 min each, with 2 min cooling in between. After sections were allowed to cool, endogenous peroxidase activity was blocked by incubating sections for 30 min in 3% hydrogen peroxide in PBS. Sections were then washed in 1% BSA in PBS for 20 min. Primary antibody incubations were performed overnight at 48C. Secondary antibody incubations were performed for 60 min at room temperature. After each incubation, unbound antibody was washed off. All antibody incubations and washes were performed in the presence of 1% BSA. Antibody binding was visualized using Sigma

Northern Analysis

Northern hybridization of RNA preparations from cultured cytotrophoblast cells was performed as previously described [4]. Briefly, purified cytotrophoblast cells were prepared from a pool of placentas ranging from 8 to 10 wk gestation [45, 46]. Cells were either processed immediately or cultured for up to 48 h on a thin layer of Matrigel (Collaborative Biomedical Products, Bedford, MA). Total RNA was extracted from cells using a guanidine isothiocyanate method, separated by formaldehyde-agarose gel electrophoresis (10 mg per lane), and transferred to Nytran membranes (Schleicher and Schuell, Keene, NH) for analysis by Northern hybridization. Probe templates were the 450-bp PvuII fragment from the 39 untranslated region of the HLA-G gene [4] and the full coding regions of the TAP2 and b2m genes [47, 48]. Probes were synthesized by random priming, using [32P]dCTP and the Klenow fragment of DNA polymerase I, and hybridized to the membranes overnight. Membranes were then washed in 0.3-strength SSC, 0.1% SDS at 688C, and signal was detected by incubation with x-ray film.

PLACENTAL REGULATION OF ANTIGEN PRESENTATION

1545

FIG. 1. HLA-G, b2m, and TAP2 mRNA in cultured cytotrophoblast cells. RNA from cytotrophoblast cells, cultured for the times indicated, was hybridized with probes to HLA-G, b2m, or TAP2.

Villous Explant Cultures

Villous explant cultures were established as previously described [49], using first-trimester placentas obtained as above. Briefly, small fragments of villi were placed on membranes precoated with Matrigel and cultured overnight in Dulbecco’s modified Eagle’s medium (DMEM)-Ham’s F12 (Life Technologies, Burlington, ON, Canada). The next day, and every 48 h thereafter, the medium was replaced with DMEM-Ham’s F-12 including an antibody to transforming growth factor (TGF)-b3 (R&D Systems, Minneapolis, MN) at 10 mg/ml. Cultures were maintained for 5 days, then fixed, embedded, sectioned, and stained as above. RESULTS

Delayed Expression of HLA-G Protein with Respect to mRNA During EVC Cell Differentiation

In a preliminary analysis, we examined purified trophoblast cells, which can be induced to differentiate in vitro by culture on Matrigel [46]. Previously, HLA-G mRNA was detected by Northern hybridization at similar levels following trophoblast purification and after such culture [4]. We have replicated this result, showing that, if anything, the mRNA level was reduced during culture (Fig. 1). In contrast, the expression of the HLA-G protein at the cell surface has been reported to increase from 25% at purification to 60% after a 12-h culture period [4]. Although the purification procedure itself may affect cell surface proteins, these expression patterns are consistent with posttranscriptional regulation of HLA-G. To perform a more detailed and accurate assessment of this possibility, we performed a histological examination of placental samples. Initially we compared the staining patterns of two antibodies recognizing HLA-G, i.e., HCA2 and 4H84, at different concentrations. We examined eight sets of serial sections from six first-trimester placentas (Fig. 2). The cytotrophoblast cell column shown derived from the proliferation of VC cells on the left, underlying the column, and subsequent differentiation of EVC cells as the cells moved distally, to the right. As the serial sections examined were adjacent to one another and only 5 mm thick, we could perform incisive intersection comparisons of HLA-G expression as detected by the different antibodies. HLA-G protein detection was restricted to the enlarged cells of the distal portion of the column, with a complete absence of staining proximal to this. An up-regulation of HLA-G expression therefore occurred at a midpoint in columns. Crucially, the two antibodies (raised to different polypeptides) at the two different concentrations detected the HLA-G protein up-regulation at comparable positions. This suggests that the increase in signal strength at a column midpoint was not an artifact of antibody sensitivity but, rather, reflected a true increase in protein accumulation. Interesting-

FIG. 2. Expression of HLA-G protein in trophoblast columns. Adjacent 5-mm sections from a 10-wk postfertilization placenta were examined by immunohistochemistry with the antibodies HCA2 (A: undiluted, B: 1:2 dilution) and 4H84 (C: undiluted, D: 1:2 dilution), both recognizing HLAG in trophoblast cells. One representative trophoblast column is shown, growing from left to right, with comparable fields of view in each panel (3345). S, Stroma. Presence of HLA-G protein was detected as brown staining.

ly, the villous stroma never showed any staining with the HCA2 antibody (Figs. 2–6). HCA2 recognizes HLA-A and HLA-G [36]. HLA-A is never expressed in trophoblast cells, so the antibody can be considered HLA-G specific in this cell type. In contrast, cells of the villous stroma express classical MHC class I molecules, so a staining of stromal cells with this reagent would be predicted. However, our results are consistent with previous studies, which also failed to show clear staining of stromal tissue with HCA2 [5, 37]. The reason for these consistent observations is currently uncertain, although they may indicate a lower sensitivity of HCA2 to HLA-A molecules, compared to HLAG, in immunohistochemical analysis. Next, we employed a similar analysis to compare the expression of HLA-G protein and mRNA. Four pairs of serial sections from each of two placentas were examined by immunohistochemistry with the antibody HCA2 and by in situ hybridization with an HLA-G-specific riboprobe (Fig. 3). As before, HLA-G protein expression was restricted to the enlarged EVC cells of the distal portion of the column. In contrast, the HLA-G mRNA was detected in the proximal as well as the distal region of the column, with a possible low level of signal above background in the underlying VC cells. We inconsistently observed a low level of signal in the VC and/or syncytiotrophoblast layers (data not shown). However, the level of signal was always much less than that observed in columns. A sense control riboprobe never showed any signal above background (data not shown). Up-regulation of the mRNA occurred at a more proximal position in columns than that of the protein (compare Fig. 3, B and D), therefore at an earlier stage of EVC cell differentiation. Hence, some form of posttranscriptional regulation of HLA-G must be operating in this cell lineage. Onset of HLA-G Protein Expression Appeared Later Than That of TAP1, TAP2, Tapasin, and b2m During EVC Cell Differentiation

It has previously been reported that TAP1 protein expression coincides with that of HLA-G [31]. We therefore

1546

COPEMAN ET AL.

FIG. 3. Relative expression of HLA-G protein and mRNA. Adjacent 5mm sections from a 6-wk postfertilization placenta were examined by immunohistochemistry with the antibody HCA2 recognizing HLA-G (A, B) and by in situ hybridization with an HLA-G-specific antisense riboprobe (C, D). One representative trophoblast column is shown, growing from left to right. The boxed area in A is enlarged in B, using magnifications of 3325 and 3650, respectively. The comparable field of view from the adjacent section, analyzed by in situ hybridization, is shown under lightfield (C) and darkfield (D) illumination. Presence of HLA-G mRNA was detected as black and white dots, respectively.

investigated whether the restricted onset of expression of TAP1, or other molecules essential for the accumulation of MHC class I molecules at the cell surface, might provide a mechanism for the posttranscriptional regulation of HLAG. Once again in a preliminary analysis we examined purified and cultured trophoblast cells by Northern hybridization for the expression of TAP2 and b2m mRNA (Fig. 1). Both transcripts were detected directly after purification. Whereas levels of b2m mRNA appeared similar before and

after the culture period, levels of TAP2 mRNA appeared to increase. However, purified cytotrophoblast cells are heterogeneous with respect to their stage of differentiation [4]. For a more accurate assessment, and to investigate whether any such an increase could be related to the accumulation of the HLA-G protein, we analyzed placental sections, comparing the expression patterns of HLA-G and proteins essential for its expression at the cell surface. Trophoblast columns on sections from six placentas were examined using one antibody to HLA-G and to tapasin and two antibodies to TAP1 and to TAP2 (Fig. 4). Again the expression of HLA-G was restricted to the distal portion of the column. A series of adjacent sections showed that TAP1, TAP2, and tapasin were also up-regulated within columns. Indeed, the position of up-regulation was similar for these three proteins, suggesting that their expression is coordinated. Strikingly, staining for TAP1, TAP2, and tapasin appeared to be closer to the base of the column (to the left) than for HLA-G, in the small partially differentiated EVC cells. To investigate this possibility more closely, we performed staining for TAP1, TAP2, and tapasin, each on a section directly adjacent to one stained for HLA-G (Fig. 5A). Again, HLA-G expression was restricted to the distal region of the column. In contrast, the up-regulation of TAP1, TAP2, and tapasin occurred at a more proximal stage than that of HLA-G. Up-regulation appeared to be more gradual than that of HLA-G, with the suggestion of a low level of expression in the progenitor VC cells underlying the column. The accumulation of TAP1, TAP2, and tapasin proteins therefore occurred earlier in EVC cell differentiation than that of HLA-G. Finally, we investigated whether a restricted expression of b2m might affect the regulation of the HLA-G protein. Adjacent sections were examined for the expression of HLA-G, b2m, and tapasin (Fig. 5B). Again, HLA-G expression was limited to a distal portion of the column, whereas tapasin was up-regulated in a more proximal region

FIG. 4. Expression of HLA-G and factors essential for peptide loading. Adjacent 5-mm sections from a 10-wk postfertilization placenta examined by immunohistochemistry with the antibody HCA2 recognizing HLA-G (A, B), with the antibodies 435.3 (C) and TAP2B (D) recognizing TAP2, with the antibodies 148.3 (E) and RING4C (F) recognizing TAP1, and with the antibody MEPsinC (G) recognizing tapasin. One representative column is shown, growing from left to right. The boxed area in A (3300) is enlarged in B (3470), and the comparable field of view shown in the adjacent sections (C–G).

PLACENTAL REGULATION OF ANTIGEN PRESENTATION

1547

FIG. 5. Expression of HLA-G relative to TAP1, TAP2, tapasin, and b2m. A) Pairs of adjacent 5-mm sections from 6.5-wk (top panels) and 6-wk (middle and bottom panels) postfertilization placentas were examined by immunohistochemistry to detect the proteins indicated. Comparable fields of view from adjacent sections are shown side by side. In each case one representative column is present, growing from left to right. Antibodies used were HCA2 recognizing HLA-G, 148.3 recognizing TAP1, 435.3 recognizing TAP2, and MEPsinC recognizing tapasin. Top 3500; middle and bottom 3300. B) Adjacent 5-mm sections from a 10-wk postfertilization placenta were examined as for A, using the antibodies HCA2 recognizing HLA-G, BBM1 recognizing b2m, and MEPsinC recognizing tapasin. 3400.

and more gradually. Detection of b2m clearly matched the tapasin expression pattern, as distinct from that of HLA-G. HLA-G Protein Expression Occurred Later Than That of TAP1, TAP2, Tapasin, and b2m in Villous Explant Culture

Growth and differentiation of EVC cells can be mimicked in vitro by explant culture of chorionic villi [49, 50]. Tips of villi from freshly isolated placenta are placed on Matrigel and cultured in serum-free medium. In the presence of activin, or factors that interfere with TGF-b signaling (such as antibodies to the TGF-b-binding protein endoglin), VC cells proliferate and form outgrowths [49, 51]. Cells differentiate within these outgrowths to adopt an invasive, EVC-like phenotype and begin to express HLA-G. Outgrowths are therefore analogous to trophoblast columns in the placenta. The precise sequence of differentiation events can be influenced. In EVC cell columns in vivo, the a5 integrin subunit is not expressed by progenitor VC cells but is up-regulated in a proximal region of columns. This position is similar to that described above for TAP, tapasin, and b2m. In villous explant cultures treated with an antibody to endoglin, the a5 integrin subunit is highly expressed throughout outgrowths, even in their most proximal cells, probably due to an acceleration of early differentiation events in vitro [49]. We examined villous explant cultures in which outgrowth had been induced using a function-perturbing antibody to TGF-b3. This disrupts the same pathway as the antibody to endoglin [51] and might, therefore, induce more proximal expression of the proteins we are examining. Indeed, tapasin and b2m were detected throughout outgrowths (Fig. 6), even in the most proximal cells, with no further distal up-regulation. The same result was obtained for TAP1 and TAP2 (data not shown).

In contrast, the HLA-G protein was up-regulated only midway through the outgrowths, markedly later in EVC cell differentiation than that of the proteins essential for its cell surface expression. Thus, the in vitro system replicated the discordance between expression patterns originally observed in situ.

FIG. 6. Expression of HLA-G relative to TAP1, TAP2, tapasin, and b2m in villous explant cultures. Pairs of adjacent sections were examined by immunohistochemistry to detect the proteins indicated. Comparable fields of view from adjacent sections are shown side by side, in each case showing one representative outgrowth, growing from left to right. Antibodies used were HCA2 recognizing HLA-G, MEPsinC recognizing tapasin, and BBM1 recognizing b2m (H). Top panels 3680; bottom panels 3820.

1548

COPEMAN ET AL.

FIG. 7. Model of expression of HLA-G, TAP, and tapasin in differentiating EVC cells. Levels of expression are related to position of EVC cells in a diagrammatic trophoblast cell column (growing from left to right). Expression of b2m may be more widespread.

DISCUSSION

In this study we have compared the onset of expression of HLA-G mRNA and protein, as well as that of the TAP1, TAP2, tapasin, and b2m proteins, essential for the cell surface accumulation of all MHC class I molecules. We observed that TAP1, TAP2, tapasin, and b2m, together with the HLA-G mRNA, are all up-regulated at a very early stage in EVC cell differentiation (Fig. 7). In contrast, the accumulation of the HLA-G protein is restricted to the distal portion of EVC cell columns. Together, these results reveal three important aspects of antigen-presenting capability during EVC cell development. First, HLA-G expression is posttranscriptionally regulated. Second, proteins essential for MHC class I expression at the cell surface are coordinately expressed during EVC cell differentiation. Third, the HLA-G protein accumulates markedly later than all other such proteins. Although several published studies have examined HLA-G and TAP1 mRNA and protein expression in the placenta [20, 21, 31, 32], such investigations have failed to make the observations described here. The previous studies examined only semiserial or unrelated sections. In contrast, our use of adjacent sections, probed independently for different molecules, provides a more accurate assessment of the relative expression patterns. Moreover, the in situ localization of transcripts and proteins permits analysis of early events in the differentiation of EVC cells. As purified cytotrophoblast cells are heterogeneous with respect to differentiation stage [4], our observations could never have been made using purified cells alone. This is the first description of posttranscriptional regulation of HLA-G expression in EVC cells. Previous studies suggesting that HLA-G might be posttranscriptionally regulated were based on the inconsistent detection of HLA-G mRNA in VC cells [20, 21]. In contrast, our results show that this regulation is a fundamental part of the differentiation of EVC cells, one of the only cell types to accumulate the protein. Accordingly, we have seen posttranscriptional regulation of HLA-G in every cell column examined. To exclude the possibility that the restricted detection of the HLA-G protein was the result of low antibody sensitivity, we have used two antibodies at different concentrations to detect HLA-G (Fig. 2). No additional areas of staining were

observed. Moreover, staining with the HLA-G-specific antibodies 4H84 and HCA2 provided a marked increase in signal at a midpoint in the column, as noted previously [37]. This is in contrast to the more gradual up-regulation of TAP1, TAP2, tapasin, and b2m, which was detected throughout the proximal portion of the column. These distinct patterns argue against a sensitivity artifact. Recently, posttranscriptional regulation of MHC class I expression has also been observed in the bovine placenta [52], suggesting that this mechanism might be widely used. It is notable that TAP1, TAP2, and tapasin shared similar patterns of expression, and that these were concomitant with that of the HLA-G mRNA. This coordinate expression of molecules involved in MHC class I antigen presentation is consistent with descriptions of shared transcriptional regulatory mechanisms for many of these factors, such as interferon-gamma activation site elements in their promotors and responsiveness to the PML gene product [53, 54]. b2m was detected in the same regions as TAP1, TAP2, and tapasin, but its expression could be more widespread. It is a secreted molecule and therefore might not accumulate in cells sufficiently for detection by the immunohistochemistry methods employed here, unless a protein to which it binds, such as MHC class I heavy chain, TAP1, or TAP2 [22, 55], is also present. Our data suggest that, during early EVC cell differentiation, a dedicated mechanism prevents the accumulation of the HLA-G protein in a posttranscriptional manner. We can conclude that this mechanism acts independently of the peptide loading system for two reasons. First, the molecules essential for peptide loading and class I maturation accumulate significantly earlier in EVC cell differentiation than the HLA-G protein. This difference is supported by the use of multiple antibodies, and by replication and modification in vitro. The expression of TAP1, TAP2, tapasin, and b2m therefore cannot limit HLA-G accumulation. Second, the antibodies we have used to detect HLA-G, i.e., HCA2 and 4H84, both recognize their target as free MHC class I heavy chain (4H84 also recognizes HLA-G as a heterodimer with b2m) [25, 34, 35]. They would be expected to detect the accumulation of unloaded HLA-G molecules that would occur if the loading of peptide were inhibited. Such an accumulation would be expected at the same position within cell columns as the appearance of HLA-G transcripts. Clearly, this is not what is observed. Instead, possible mechanisms of posttranscriptional regulation include regulation of translation, rapid protein degradation, or presence of protein in a form undetectable by the antibodies used here. There is no direct evidence to discriminate between these possibilities at this time. However, in the recent study by Blaschitz and colleagues [5], two different antibodies raised to HLA-G provided a more widespread signal than had been anticipated. Staining was present in the proximal region of cell columns but became more faint in the distal regions as EVC cells differentiated. It is currently unclear exactly what epitopes are being recognized by these antibodies [34, 56]. However, an intriguing hypothesis, which would resolve these observations with ours, is that the antibodies detect the HLA-G protein in a sequestered or alternative form. Crucially, our replication of results in an in vitro system suggests that determination of the posttranscriptional regulatory mechanism will be tractable to an experimental approach. Clearly it is critical to determine whether the mechanism of regulation is specific to HLA-G or is of application to MHC class I molecules in general. HLA-C is also be ex-

PLACENTAL REGULATION OF ANTIGEN PRESENTATION

pressed in EVC cells [37, 57]. In contrast to the characteristic distal up-regulation of HLA-G that we observe, expression of HLA-C was reported to be detected throughout cell columns, even when the antibody used recognized only the mature, cell surface form [37]. This observation is consistent with our findings that TAP, tapasin, and b2m, which are required for cell surface expression of HLA-C, may be expressed throughout cell columns and are up-regulated at an early stage of EVC cell differentiation. The specificity of posttranscriptional regulation could be further scrutinized by analysis of HLA-E, which is likely to be expressed in cell columns but has not yet been described due to lack of suitable reagents. In conclusion, we have shown that factors essential for the maturation and peptide loading of MHC class I molecules are expressed in advance of the appearance of the HLA-G protein. Differentiating EVC cells may therefore be in a state of readiness when the HLA-G protein is expressed, permitting a very rapid accumulation of this nonclassical class I molecule at the cell surface. The implied existence of a novel posttranscriptional mechanism of MHC class I regulation may be of broad significance. HLA-G transcripts have been detected in a range of tissues, apparently without accumulation of the protein, yet presumably in the presence of fully competent peptide loading machinery [8–12]. Our data reveal the existence of a mechanism that could explain these remarkable discrepancies. The evolution of such a system suggests that the ectopic expression of HLA-G is actively prevented.

11.

12. 13. 14. 15.

16. 17.

18.

19. 20.

ACKNOWLEDGMENTS We thank Lindsay McWhirter, Jeanne Zielonka, and Christine Botsford for collecting clinical samples; Peter Cresswell, Mary Pan, Jaana Karttunen, Naveen Bangia, John Chamberlain, Danny Schust, and Hidde Ploegh for antibodies.

21.

22.

REFERENCES 1. Cross JC. Trophoblast function in normal and preeclamptic pregnancy. Fetal Maternal Med Rev 1996; 8:57–66. 2. Ellis SA, Palmer MS, McMichael AJ. Human trophoblast and the choriocarcinoma cell line BeWo express a truncated HLA class I molecule. J Immunol 1990; 144:731–735. 3. Kovats S, Main EK, Librach C, Stubblebine M, Fisher SJ, DeMars R. A class I antigen, HLA-G, expressed in human trophoblasts. Science 1990; 248:220–223. 4. McMaster MT, Librach CL, Zhou Y, Lim KH, Janatpour MJ, DeMars R, Kovats S, Damsky C, Fisher SJ. Human placental HLA-G expression is restricted to differentiated cytotrophoblasts. J Immunol 1995; 154:3771–3778. 5. Blaschitz A, Lenfant F, Mallet V, Hartmann M, Bensussan A, Geraghty DE, Le Bouteiller P, Dohr G. Endothelial cells in chorionic fetal vessels of first trimester placenta express HLA-G. Eur J Immunol 1997; 27:3380–3388. 6. Crisa L, McMaster MT, Ishii JK, Fisher SJ, Salomon DR. Identification of a thymic epithelial cell subset sharing expression of the class Ib HLA-G molecule with fetal trophoblasts. J Exp Med 1997; 186: 289–298. 7. Mallet V, Blaschitz A, Crisa L, Schmitt C, Fournel S, King A, Loke YW, Dohr G, Le Bouteiller P. HLA-G in the human thymus: a subpopulation of medullary epithelial but not CD83(1) dendritic cells expresses HLA-G as a membrane-bound and soluble protein. Int Immunol 1999; 11:889–898. 8. Shukla H, Swaroop A, Srivastava R, Weissman SM. The mRNA of human class I gene HLA-G/HLA 6.0 exhibits a restricted pattern of expression. Nucleic Acids Res 1990; 18:2189. 9. Onno M, Guillaudeux T, Amiot L, Renarde I, Drenou B, Hirel B, Girr M, Semana G, Le Bouteiller P, Fauchet R. The HLA-G gene is expressed at a low mRNA level in different human cells and tissues. Hum Immunol 1994; 41:79–86. 10. Pangault C, Amiot L, Caulet-Maugendre S, Brasseur F, Burtin F, Guil-

23. 24. 25. 26.

27. 28.

29.

30. 31. 32.

1549

loux V, Drenou B, Fauchet R, Onno M. HLA-G protein expression is not induced during malignant transformation. Tissue Antigens 1999; 53:335–346. Paul P, Cabestre´ FA, Le Gal FA, Khalil-Daher I, Le Danff C, Schmid M, Mercier S, Avril MF, Dausset J, Guillet JG, Carosella ED. Heterogeneity of HLA-G gene transcription and protein expression in malignant melanoma biopsies. Cancer Res 1999; 59:1954–1960. Real LM, Cabrera T, Collado A, Jimenez P, Garcia A, Ruiz-Cabello F, Garrido F. Expression of HLA-G in human tumors is not a frequent event. Int J Cancer 1999; 81:512–518. Taylor RN. Review: immunobiology of preeclampsia. Am J Reprod Immunol 1997; 37:79–86. Zhou Y, Damsky CH, Chiu K, Roberts JM, Fisher SJ. Preeclampsia is associated with abnormal expression of adhesion molecules by invasive cytotrophoblasts. J Clin Invest 1993; 91:950–960. Zhou Y, Damsky CH, Fisher SJ. Preeclampsia is associated with failure of human cytotrophoblast to mimic a vascular adhesion phenotype. One cause of defective endovascular invasion in this syndrome? J Clin Invest 1997; 99:2152–2164. Lim KH, Zhou Y, Janatpour M, McMaster M, Bass K, Chun SH, Fisher SJ. Human cytotrophoblast differentiation/invasion is abnormal in pre-eclampsia. Am J Pathol 1997; 151:1809–1818. Caniggia I, Grisaru-Gravnosky S, Kuliszewsky M, Post M, Lye SJ. Inhibition of TGF-beta3 restores the invasive capability of extravillous trophoblasts in preeclamptic pregnancies. J Clin Invest 1999; 103: 1641–1650. Hara N, Fujii T, Yamashita T, Kozuma S, Okai T, Taketani Y. Altered expression of human leukocyte antigen G (HLA-G) on extravillous trophoblasts in preeclampsia: immunohistological demonstration with anti-HLA-G specific antibody ‘‘87G’’ and anti-cytokeratin antibody ‘‘CAM5.2.’’ Am J Reprod Immunol 1996; 36:349–358. Genbacev O, Zhou Y, Ludlow JW, Fisher SJ. Regulation of human placental development by oxygen tension. Science 1997; 277:1669– 1672. Yelavarthi KK, Fishback JL, Hunt JS. Analysis of HLA-G mRNA in human placental and extraplacental membrane cells by in situ hybridization. J Immunol 1991; 146:2847–2854. Chumbley G, King A, Holmes N, Loke YW. In situ hybridization and northern blot demonstration of HLA-G mRNA in human trophoblast populations by locus-specific oligonucleotide. Hum Immunol 1993; 37:17–22. Williams DB, Barber BH, Flavell RA, Allen H. Role of b2-microglobulin in the intracellular transport and surface expression of murine class I histocompatibility molecules. J Immunol 1989; 142:2796– 2806. Howard J. Supply and transport of peptides presented by class I MHC molecules. Curr Opin Immunol 1995; 7:69–76. Powis SJ. Major histocompatibility complex class I molecules interact with both subunits of the transporter associated with antigen presentation, TAP1 and TAP2. Eur J Immunol 1997; 27:2744–2747. Grandea III AG, Androlewicz MJ, Athwal RS, Geraghty DE, Spies T. Dependence of peptide binding by MHC class I molecules on their interaction with TAP. Science 1995; 270:105–108. Ortmann B, Copeman J, Lehner PJ, Sadasivan B, Herberg JA, Grandea AG, Riddell SR, Tampe´ R, Spies T, Trowsdale J, Cresswell P. A critical role for tapasin in the assembly and function of multimeric MHC class I-TAP complexes. Science 1997; 277:1306–1309. Lehner PJ, Surman MJ, Cresswell P. Soluble tapasin restores MHC class I expression and function in the tapasin-negative cell line .220. Immunity 1998; 8:221–231. Copeman J, Bangia N, Cross JC, Cresswell P. Elucidation of the genetic basis of the antigen presentation defects in the mutant cell line .220 reveals polymorphism and alternative splicing of the tapasin gene. Eur J Immunol 1998; 28:3783–3791. Schoenhals GJ, Krishna RM, Grandea III AG, Spies T, Peterson PA, Yang Y, Fry¨h K. Retention of empty MHC class I molecules by tapasin is essential to reconstitute antigen presentation in invertebrate cells. EMBO J 1999; 18:743–753. Schust DJ, Hill AB, Ploegh HL. Herpes simplex virus blocks intracellular transport of HLA-G in placentally derived human cells. J Immunol 1996; 157:3375–3380. Clover LM, Sargent IL, Townsend A, Tampe´ R, Redman CWG. Expression of TAP1 by human trophoblast. Eur J Immunol 1995; 25: 543–548. Roby KF, Gershon D, Hunt JS. Expression of the transporter for antigen processing-1 (Tap-1) gene in subpopulations of human trophoblast cells. Placenta 1996; 17:27–32.

1550

COPEMAN ET AL.

33. Roby KF, Fei K, Yang Y, Hunt JS. Expression of HLA class II-associated peptide transporter and proteasome genes in human placentas and trophoblast cell lines. Immunology 1994; 83:444–448. 34. McMaster M, Zhou Y, Shorter S, Kapasi K, Geraghty D, Lim KH, Fisher S. HLA-G isoforms produced by placental cytotrophoblasts and found in amniotic fluid are due to unusual glycosylation. J Immunol 1998; 160:5922–5928. 35. Stam NJ, Vroom TM, Peters PJ, Pastoors EB, Ploegh HL. HLA-Aand B-specific monoclonal antibodies reactive with free heavy chains in Western blots, in formalin-fixed, paraffin-embedded tissue sections and in cryoimmuno-electron microscopy. Int Immunol 1990; 2:113– 125. 36. Sernee MF, Ploegh HL, Schust DJ. Why certain antibodies cross-react with HLA-A and HLA-G: epitope mapping of two common MHC class I reagents. Mol Immunol 1998; 35:177–188. 37. Hutter H, Hammer A, Blaschitz A, Hartmann M, Ebbesen P, Dohr G, Ziegler A, Uchanska-Ziegler B. Expression of HLA class I molecules in human first trimester and term placenta trophoblast. Cell Tissue Res 1996; 286:439–447. 38. Brodsky FM, Bodmer WF, Parham P. Characterization of a monoclonal anti-b2-microglobulin antibody and its use in the genetic and biochemical analysis of major histocompatibility antigens. Eur J Immunol 1979; 9:536–545. 39. Spies T, Cerundolo V, Colonna M, Cresswell P, Townsend A, DeMars R. Presentation of a viral antigen by MHC class I molecules is dependent on a putative peptide transport heterodimer. Nature 1992; 355: 644–646. 40. Meyer TH, van Endert PM, Uebel S, Ehring B, Tampe´ R. Functional expression and purification of the ABC transporter complex associated with antigen processing (TAP) in insect cells. FEBS Lett 1994; 351: 443–447. 41. van Endert PM, Tampe´ R, Meyer TH, Tisch R, Bach JF, McDevitt HO. A sequential model for peptide binding and transport by the transporters associated with antigen processing. Immunity 1994; 1: 491–500. 42. Schmidt CM, Ehlenfeldt RG, Athanasiou MC, Duvick LA, Heinrichs H, David CS, Orr HT. Extraembryonic expression of the human MHC class I gene HLA-G in transgenic mice. Evidence for a positive regulatory region located 1 kilobase 59 to the start site of transcription. J Immunol 1993; 151:2633–2645. 43. Guillaudeux T, Rodriguez AM, Girr M, Mallet V, Ellis SA, Sargent IL, Fauchet R, Alsat E, Le Bouteiller P. Methylation status and transcriptional expression of the MHC class I loci in human trophoblast cells from term placenta. J Immunol 1995; 154:3283–3299. 44. Han RNN, Han VKM, Buch S, Freeman BA, Post M, Tanswell AK. Insulin-like growth factor-I and type I insulin-like growth factor re-

45. 46.

47.

48.

49. 50. 51. 52. 53. 54. 55.

56.

57.

ceptor in 85% O2-exposed rat lung. Am J Physiol 1996; 271:L139– 149. Fisher SJ, Cui TY, Zhang L, Hartman L, Grahl K, Zhang GY, Tarpey J, Damsky CH. Adhesive and degradative properties of human placental cytotrophoblast cells in vitro. J Cell Biol 1989; 109:891–902. Librach CL, Werb Z, Fitzgerald ML, Chiu K, Corwin NM, Esteves RA, Grobelny D, Galardy R, Damsky CH, Fisher SJ. 92-kD type IV collagenase mediates invasion of human cytotrophoblasts. J Cell Biol 1991; 113:437–449. Powis SH, Mockridge I, Kelly A, Kerr LA, Glynne R, Gileadi U, Beck S, Trowsdale J. Polymorphism in a second ABC transporter gene located within the class II region of the human major histocompatibility complex. Proc Natl Acad Sci USA 1992; 89:1463–1467. Suggs SV, Wallace RB, Hirose T, Kawashima EH, Itakura K. Use of synthetic oligonucleotides as hybridization probes: isolation of cloned cDNA sequences for human beta 2-microglobulin. Proc Natl Acad Sci USA 1981; 78:6613–6617. Caniggia I, Taylor CV, Ritchie JWK, Lye SJ, Letarte M. Endoglin regulates trophoblast differentiation along the invasive pathway in human placental villous explants. Endocrinology 1997; 138:4977–4988. Genbacev O, Schubach SA, Miller RK. Villous culture of first trimester human placenta: model to study extravillous trophoblast (EVT) differentiation. Placenta 1992; 13:439–461. Caniggia I, Lye SJ, Cross JC. Activin is a local regulator of human cytotrophoblast cell differentiation. Endocrinology 1997; 138:3976– 3986. Ellis SA, Sargent IL, Charleston B, Bainbridge DR. Regulation of MHC class I gene expression is at transcriptional and post-transcriptional level in bovine placenta. J Reprod Immunol 1998; 37:103–115. Boehm U, Klamp T, Groot M, Howard JC. Cellular responses to interferon-g. Annu Rev Immunol 1997; 15:749–795. Zheng P, Guo Y, Niu Q, Levy DE, Dyck JA, Lu S, Shelman L, Liu Y. Proto-oncogene PML controls genes devoted to MHC class I antigen presentation. Nature 1998; 396:373–376. Solheim JC, Harris MR, Kindle CS, Hansen TH. Prominence of b2microglobulin, class I heavy chain conformation, and tapasin in the interactions of class I heavy chain with calreticulin and the transporter associated with antigen processing. J Immunol 1997; 158:2236–2241. Bensussan A, Mansur IG, Mallet V, Rodriguez AM, Girr M, Weiss EH, Brem G, Boumsell L, Gluckman E, Dausset J, Carosella E, Le Bouteiller P. Detection of membrane-bound HLA-G translated products with a specific monoclonal antibody. Proc Natl Acad Sci USA 1992; 92:10292–10296. King A, Boocock C, Sharkey AM, Gardner L, Beretta A, Siccardi AG, Loke YW. Evidence for the expression of HLA-C class I mRNA and protein by human first trimester trophoblast. J Immunol 1996; 156:2068–2076.