Analysis of MHC class I expression in equine trophoblast cells using in situ hybridization

Placenta

(1996),

17, 351-359

Analysis

of MHC

In Situ

Hybridization

J. K. Mahera, The Paper

James

A.

Class

D. B. Tresnanb, Baker

accepted

Institute

7 February

for

Animal

I Expression

S. Deacon, Health,

L. Hannah College

in Equine

Trophoblast

Cells

Using

and D. F. Antczak”

of Veterinary

Medicine,

Cornell

University,

Ithaca,

NY,

14853,

USA

1996

Down-regulation of major histocompatibility complex (MHC) genes by trophoblast cells is considered to be a primary mechanism preventing maternal immune rejection of the fetal-placental unit in mammalian pregnancy by rendering these cells, which form the primary barrier between mother and fetus, relatively non-antigenic. In situ hybridization with probes encoding human and horse MHC class I genes was used to characterize the pattern of MHC class I mRNA expression in the various forms of horse trophoblast. Strong hybridization signals were observed in the invasive trophoblast cells of chorionic girdle tissue. In contrast, no hybridization signal specific for MHC class I mRNA transcripts was observed in the descendent endometrial cup trophoblast cells. In the non-invasive trophoblast cells of the allantochorion, no hybridization signals specific for horse MHC class I mRNA transcripts were consistently detected. In parallel to the in vivo results, strong hybridization signals were observed in the small, mononuclear cells present in chorionic girdle cell explant cultures, but not in the population of large binucleate cells corresponding to endometrial cup cells. The results obtained using in situ hybridization are consistent with the hypothesis that expression of MHC class I genes may be controlled at the transcriptional level in horse invasive and non-invasive trophoblast cells, and suggest that down-regulation of MHC class I antigen expression in endometrial cup cells may be accomplished by the same mechanisms in vivo and in vitro. 0 1996 W. B. Saunders Company Ltd Placenta (1996), 17, 351-359

INTRODUCTION The question of how the fetus, which expresses both parental major histocompatibility complex (MHC) class I and class II antigens, escapes maternal immune recognition, and thus its own destruction, has been the subject of continuing study in the field of reproductive immunology (Medawar, 1953). Down-regulation of MHC antigen expression by trophoblast cells, which form the outermost layer of the placenta and are in direct contact with maternal tissues, is considered to be one of the most important mechanisms contributing to survival of the fetus (Wood, 1994). The failure of trophoblast cells to express a full complement of MHC molecules ensures that they present a poor target to effecters of the immune system, including antibodies and antigen-specific T cells that recognize foreign major and minor histocompatibility antigens. Expression of polymorphic MHC class I antigens has not been detected on any human trophoblast cell sub-population (reviewed by Hunt and Orr, 1992). However, the nonpolymorphic MHC class I molecule, HLA-G, is expressed by human extravillous cytotrophoblast cells (Sunderland, Redman ’ Present address: Molecular Biology Institute, UCLA, 405 Hilgard Ave., Los Angeles, CA 90024-1570, USA b Present address: University of Colorado Health Sciences Center, Dept. of Microbiology, B-175, Denver, CO 80262, USA ’ To whom correspondence should be addressed. 0143S4004/96/050351+09

$12.00/O

and Stirrat, 1981; Ellis et al., 1986; Hunt et al., 1990; Kovats et al., 1990; Yelavarthi, Fishback and Hunt, 1991). Although antibodies directed against paternally inherited fetal MHC molecules are detected in sera of multiparous women (van Rood, Eernisse and van Leeuwen, 1958), the incidence of maternal sensitization is low (Regan, 1988), and antibodies are rarely detected before parturition (Redman, Arenas and Sargent, 1987). The results of these studies imply that major and minor histocompatibility complex antigens are not expressed in an immunogenic form by human placental cells during gestation. Horses are unique among mammals by virtue of the fact that cytotoxic antibody responses are produced by virtually 100 per cent of pregnant mares during MHC incompatible pregnancy (Antczak et al., 1982). High titred cytotoxic antibodies directed primarily against the polymorphic MHC class I antigens of the mating stallion are regularly detected by day 60 of gestation (Antczak, Miller and Remick, 1984). A complex pattern of MHC

antigen

expression

by the invasive

trophoblast

cells

of the chorionic girdle appears to induce the alloantibody response (Crump et al., 1987; Donaldson et al., 1990; Donaldson et al., 1994). Upon differentiation of the equine chorionic girdle cells into endometrial cup cells in the endometrium, MHC class I antigen expression is permanently and irreversibly lost from the surface of endometrial cup cells by day 45 of gestation 0

1996 W. B. Saunders

Company

Ltd

352

Placenta

(Donaldson et al., 1992). The non-invasive trophoblast cells of the equine allantochorion, which form the majority of the fetal-maternal interface, do not express MHC class I antigens except in isolated patches (Donaldson et al., 1990). MHC class II antigen expression has not been detected in any horse trophoblast cells (Donaldson et al., 1990). The endometrial cup cells are the sole source of hormone, equine chorionic gonadotropin (eCG), and are considered to have an important role in establishing and maintaining a viable equine pregnancy (Allen and Moor, 1972). Down-regulation of MHC class I antigen expression by the endometrial cup cells allows evasion of the maternal cytotoxic antibody response, and may prevent premature destruction of the endometrial cup cells, which could jeopardize the outcome of pregnancy. In this paper, we describe the pattern of MHC class I mRNA expression in the various populations of horse trophoblast cells at selected times during gestation in a preliminary effort to elucidate those mechanisms that may regulate fetal-maternal immunological interactions affecting the success of pregnancy in mammals.

MATERIALS

AND

METHODS

Animals Horses and ponies of the Equine Genetics Center herd were used. Pregnancies were established by artificial insemination. The day of ovulation was determined by visualization of the ovaries using real-time transrectal ultrasound.

Tissue

acquisition

and fixation

A non-surgical approach (described in Antczak et al., 1987) was used to recover fetal sacsas a source of chorionic girdle for histological secretions before the time of firm attachment of the conceptus to the uterine wall at approximately day 36 of gestation. After attachment, samples of horse placental tissues were obtained at post mortem or during surgical hysterotomy between days 38-74 of gestation. A summary of tissues tested using in situ hybridization is provided in Table 1. Tissues were placed in disposable embedding cassettes and fixed in fresh 4 per cent paraformaldehyde (PFA) (w/v) (Baker, S898-07) in 1 X PBS (130 rnM NaCl, 7 ITIM NaHzPO,, 3 mM NaH,PO,, pH 7.0) overnight at 4°C. The tissues were then dehydrated in an equal volume of the following solutions: 1 x PBS, 0.85 per cent saline, ethanol (EtOH) : saline (1 : 1 v/v), 70 per cent EtOH in saline, 85 per cent EtOH, 9.5 per cent EtOH, and two changes of 100 per cent EtOH followed by two changes of toluene. The tissues were embedded in paraffin by replacing the toluene with a 1 : 1 mixture of toluene : paraffin at 60°C followed by two changes of paraffin at 60°C under vacuum. Paraffin-embedded tissue blocks were prepared in stainless steel molds (Fisher Scientific, Pittsburgh, PA, USA) using Histoprep (Fisher Scientific) embedding rings. Ten micron

(1996),

Vol.

17

sections were cut onto aminoalkylsilane-treated glass slides (Sigma, A-3648) and stored at 4°C with desiccant.

Maintenance

and harvesting

of cell lines

Chorionic girdle tissue was obtained from day 30-36 equine conceptuses recovered from mares by non-surgical lavage (Antczak et al., 1987, Table 1). The chorionic girdle was dissected free of allantochorion and cultured using conditions previously described (Allen and Moor, 1972). Small fragments of chorionic girdle were cultured in complete medium (DMEM, 10 per cent heat-inactivated fetal calf serum (FCS), 1 per cent non-essential amino acids, 2 mM L-glutamine, 100 U penicillin/ml and 5 x 10 p4 mg Vitamin C/ml) at 37°C in 8 per cent CO, in plastic flasks or in uncoated 2-, 4-, or 8-well plastic chamber slides (Nunc, Naperville, IL, USA). The adherent trophoblast cells were harvested from plastic flasks by trypsinization for 24 min at 37°C using 0.25 per cent trypsin, 1 mM EDTA in HBSS (Gibco, Grand Island, NY, USA). The cell pellet was resuspended at a concentration of 2.5 x lo5 cells/ml in DME, containing 20 per cent FCS. Twenty to 30 000 cells were centrifuged onto APTS (Sigma, St. Louis, MO, USA) coated slides using a cytospin centrifuge (Shandon, Pittsburgh, PA, USA). The cytospun preparations were fixed in 4 per cent paraformaldehyde (PFA) (w/v) in 1 x PBS for 20 min at room temperature (RT). The cells were washed twice for 5 min in 1 X PBS at RT, dehydrated through an alcohol series in 0.85 per cent saline, and stored at - 70°C until use. Fixation of cells growing in plastic chamber slides took place directly on the slide after the plastic chambers were removed and the slides were washed in 1 X PBS for 5 min. Thereafter, fixation in paraformaldehyde was performed exactly as described above.

Probes

used in in situ hybridization

A 1.4 Kb human cDNA encoding the HLA-A2.2Y gene (Holmes et al., 1987) was subcloned into the pGEM-3Z vector (Promega, Madison, WI, USA). A 1.5 Kb cDNA encoding a horse MHC class I gene (Barbis et al., 1994) was cloned into pcDNA 1 vector (Invitrogen, San Diego, CA, USA). Linearized plasmid DNA (1 pg) was transcribed by 15-20 U of SP6 or T7 RNA polymerase (Promega, Madison, WI, USA) in 40 mM Tris-HCl (pH 7.5); 6 mM MgCl,; 2 KIM spermidine; 10 mM NaCl; 10 mM DTT; 500 FM UTP; and 20 U RNasin ribonuclease inhibitor in the presence of 300 uCi [35S]UTP (1200 Ci/mmol; New England Nuclear, Wilmington, DE, USA) at 37°C for 60 min. Probe length was adjusted to less than 500 nucleotides by hydrolysis in 80 pM NaHCO,-120 pM Na,C03 at 60°C for 1 h. Probe length was determined by Northern blotting and autoradiography after electrophoresis of intact and hydrolyzed samples in non-denaturing 1 per cent agarose gels. Unincorporated nucleotides were removed by EtOH precipitation with 1 ~1 10 mg/ml tRNA, 1 ul 1 M

353

Maher et al.: Equine MHC Class I In Situ Hybridization Table

Tissue or cell type” Chorionic

girdle

Endometrial cups + / allantochorion

Non-pregnant endometrium Pregnant endometrium + allantochorion Chorionic girdle cell explant cultures

I. Tissues used for in situ hybridization

experiments

Gestational age or days in culture”

Number of horses sampled

Day Day Day Day Day Day Day Day Day Day Day Day

1 1 1 2 1 2 1 1 1 2 1 1

Human MHC Class I Horse MHC Class I

1 1 1 7

Horse MHC Class I 18s RNA eCG-P Human MHC Class I Horse MHC Class I eCG-P 18s RNA

30 31 32 33 38 42 43 54 55 60 73 42

Day 42 Day 55 Day 73 2, 11, 13, 25, 44, 64, and 71 days in culture

Probes

Human MHC Class I Horse MHC Class I eCG-/3 18s RNA

Horse MHC Class I

a Samples were 5 pm sections of paraformaldehyde-fixed, paraffin-embedded equine trophoblast and uterine tissues or cells from chorionic girdle explant cultures maintained in vitro. b This column indicates the age of the pregnancy (day 0 is the day of ovulation detected by daily real-time transrectal ultrasound examination) or the number of days in culture at the time the tissues or cells were fixed for in situ hybridization. ’ eCG-P, beta subunit of equine chorionic gonadotropin. Table

2. Oligonucleotide

probes used for in situ hybridization”

Probe

Orientation

DNA Sequence (5’ to 3’)

eCG-J3 eCG-f3

Antisense Sense

GCCGGATGGAAGCAAAGCGCAGCTCACGGTAGGTGCACAC GTGTGCACCTACCGTGAGCTGCGCTTTGCTTCCATCCGGC

a The eCG-P antisense oligonucleotide was a gift from Dr John Nilson’s laboratory at Case Western Reserve University and the sense nucleotide was derived from the published sequence (Sherman et al., 1992). M&l,, 10 ~1 7.5 M ammonium acetate and 2.5 volumes of EtOH at - 70°C overnight or by G-50 Sephadex spun column chromatography. The specific activity of RNA probes resuspended in 50 ~1 of 1 M DTT was approximately 1 X 10’ cpm/pg. A 170 bp polymerase chain reaction (PCR) product encoding a conserved portion of the bovine 18s ribosomal RNA gene subcloned into PBS+ / - vector (Stratagene, La Jolla, CA, USA) was obtained from Dr Joanne Fortune of Cornell University for use as a positive control RNA probe. 35Slabelled sense and antisense RNA transcripts were produced as described above from T3 and T7 bacteriophage polymerase promoters. DNA sense and antisense oligonucleotide probes (40mers) specific for the beta subunit of eCG-P were derived from the published sequence (Sherman et al., 1992; Ellis and Martin,

1993) and synthesized by a service laboratory at the Cornell Biotechnology facility (Table 2). Oligonucleotides were labelled at the 3’ end with 35S-dATP using terminal deoxynucleotidyl transferase (DNA 3’-End Labelling System, Promega, Madison, WI, USA). Unincorporated nucleotides were removed using a G-25 Sephadex spun column. The specific activity of the 3’-tailed oligonucleotides was greater than 1 x lo9 cpm/pg.

In situ

hybridization

Cytospins were hybridized using a modification of the method of Wilkinson and Green (1991). Slides containing cultured cells were warmed to RT for 10 min, fixed in acetone for 10 min at 4°C dried 5-15 min at RT and refixed in 4 per cent

354

Placenta

PFA in 1 X PBS for 20 min at RT. Slides were then washed twice in 1 X PBS for 5 min at RT and acetylated in 300 ml of 0.1 M triethanolamine HCl containing 0.76 ml of acetic anhydride for 10 min at RT. The slides were washed in 1 x PBS followed by 0.85 per cent saline for 5 min at RT, dehydrated through an EtOH series containing 0.85 per cent saline, and air dried. Slides containing paraffin-embedded tissue sections were warmed at RT for lo-20 min, dewaxed in two changes of xylene for 10 min each, rehydrated in 100 per cent (2 min), 100, 95, 80, 60 and 30 per cent (v/v) EtOH in saline for several seconds each, then in saline (5 min) and PBS (5 min). The sections were then fixed in 4 per cent PFA (20 min), washed in PBS twice (5 min each), treated with 20 ug/ml proteinase K in Tris-EDTA (50 UIM : 5 mM, 5 min), washed in PBS (5 min), post-fixed in 4 per cent PFA (20 min), dipped in water and immersed in 0.25 per cent (v/v) acetic anhydride in 0.1 M triethanolamine HCl for 10 min. The slides were transferred through PBS (5 min), saline (5 min), 30 per cent (< 1 min), 60 per cent (5 min), 80 per cent (< 1 min), 95 per cent (< 1 min), and two changes of 100 per cent (< 1 min) (v/v) EtOH in saline and air dried at RT (30 min). Each slide (one slide=24 tissue sections) was hybridized with either 3 X 10’ cpm of oligonucleotide probe or 5 x lo6 cpm of RNA probe in 50 ul of solution A (50 per cent deionized formamide, 0.3 M NaCI, 20 mM Tris-HCl (pH 8.0) 5 mM EDTA, 10 InM sodium phosphate buffer (pH 8.0), 10 per cent Dextran sulfate, 1 X Denhardts, 10 mM DTT, and 0.5 mg/ml yeast tRNA). Hybridization was allowed to proceed overnight (16-20 h) at 42 or 50°C in a sealed humid chamber containing absorbent paper soaked in a solution of 50 per cent formamide and 5 X SSC. Washes for oligonucleotide probes were as follows: 2 x SSC, 10 mM DTT for 50 min at RT to dislodge coverslips; 2 X SSC, 10 mMDTT for 60 min at 42°C; 2 X SSC for 1 min at RT, then overnight in a solution of 2 X SSC containing 0.05 per cent Triton X-100. The next day the slides were washed twice in 2 x SSC for 5 min at RT, once in 0.5 x SSC for 5 min, and dehydrated through an EtOH series containing 0.3 M ammonium acetate. For RNA probes the washes were as follows: 5 X SSC, 10 UIM DTT at 50°C for 30-60 min to dislodge coverslips; 50 per cent formamide, 2 X SSC, 10 ITIM DTT at 65°C for 30 min (high stringency wash); NTE buffer [0.5 M NaCl, 10 mM Tris-HCl (pH 7.5), 5 mM EDTA] three times at 37°C for 10 min each; RNase A (20 ug/ml) in NTE buffer at 37°C for 30 min; NTE buffer at 37°C for 15 min; repeat high stringency wash; 2 X SSC followed by 0.1 X SSC at RT for 15 min. The tissue sections were hydrated in an EtOH series Containing

0.3 M

XrUnOniUm

aCetaR!.

Autoradiography Dried slides were apposed Kodak, Rochester, NY,

to Kodak XAR-5 USA) overnight

film (Eastman to determine

(1996),

Vol. 17

exposure times under emulsion. Slides were dipped in Kodak NTB2 emulsion (International Biotechnologies Inc., New Haven, CT, USA) diluted 1 : 1 with 0.6 M ammonium acetate and dried overnight at RT. The slides were exposed for 7-14 days in light tight boxes at 4°C. The slides were developed in Kodak Dektol developer diluted 1 : 1 with distilled water for 2 min, stopped in distilled water for 10 set, fixed in Kodak fixative for 5 min at 15-17°C and rinsed in distilled water for 5 min. The sections were lightly counterstained with either a 0.01 per cent solution of toluidine blue or 5 per cent eosin Y, rinsed in distilled water, air dried, and photographed using a Nikon FA 35 mm camera attached to a Nikon Diaphot-TMD inverted microscope using Kodak Gold 200 ASA film for colour prints. Darkfield photomicrography and Nomarski optics were done using Zeiss 35 mm cameras attached to a Zeiss Universal transmitted light microscope and Axioplan microscope, respectively, using Kodak Technical Pan film for black and white prints.

lmmunohistochemistry Frozen sections of chorionic girdle and endometrial cup tissue were assayed using indirect immunoperoxidase labelling with anti-horse MHC class I specific monoclonal antibodies, as described (Spear et al., 1985; Donaldson et al., 1990).

RESULTS Detection chorionic

of MHC class girdle cells

I mRNA

transcripts

in

An intense hybridization signal was observed on sections of day 30 and 32 chorionic girdles hybridized with horse and human MHC class I RNA probes [Figure l(c)]. The pattern of expression of MHC class I mRNA transcripts was identical to the pattern of antigen expression detected by cross-reactive anti-human and anti-horse MHC class I monoclonal antibodies [Figure l(a) and (b)].

Detection endometrial

of

MHC class cup cells

I mRNA

transcripts

in

Sections of early, ma,ture and late endometrial cup tissue (representing days 42, 60 and 73 of gestation) were hybridized with human MHC class I sense and antisense RNA probes. The age distribution of samples was expanded to represent days 38, 42, 43, 54, 55, 60 and 73 of gestation when the horse MHC class I riboprobes were used. These time points represent the limits of the endometrial cup lifespan that can yield interpretable results from histological sections. In situ hybridization of mature endometrial cups with either the human or the horse riboprobe revealed a pattern of expression of MHC class I mRNA transcripts that coincided

Maher

et al.: Equine

MHC

Class I In Situ Hybridization

355

Figure 1 Immunohistochemical labelling and in situ hybridization of horse chorionic girdle. (a) and (b) Monoclonal antibody A131, specific for human MHC class I antigens (Spear et al., 1985), showing representative labelling of chorionic girdle cells by anti-human and anti-horse MHC class I monoclonal antibodies (Donaldson et al., 1990). (a) Bar= 114 ,um. (b) Bar= 11 pm. (c) Horse MHC class I specific antisense RNA probe. Bar=28 pm. (d) Negative control, horse MHC class I specific sense RN.4 probe. Bar=28 pm. cg, chorionic girdle cells; ale, allantochorion trophoblast.

Figure 2. Immunohistochemical labelling and in situ hybridization brightfield photomicrograph of horse endometrial cup tissue hybridized Bar=46 pm. (c) High-power brightfield photomicrograph of the (d) High-power brightfield photomicrograph of horse endometrial cup the grain density within the boxed areas (c) and (d). Bar=28 pm. ec,

of mature endometrial cups. (a) Monoclonal antibody with the human antisense MHC class I RNA probe specimen from (b), compare with immunohistochemical tissue hybridized with a negative control human sense endometrial cup cells; gl, maternal endometrial gland

A131. Bar=28 pm. (b) Low-power [see (c) for a higher magnification]. labelling of (a). Bar=28 pm. MHC class I RNA probe. Compare epithelium.

356

Figure 3. In situ hybridization of day 54 mature endometrial cups using equine chorionic gonadotropin and RNA probes derived from horse MHC class I cDNA clone 8-9. (a) Brightfield photomicrograph of antisense oligonucleotide probe. Note the endometrial cup trophoblast cells surrounding the maternal Bar=44 pm. (b) Darkfield photomicrograph of (a). Bar=44 pm. (c) Darkfield photomicrograph of horse class I antisense RNA probe. Bar=44 pm. The pattern of hybridization is the inverse of that shown in (b). epithelium; ly, maternal lymphocytes.

exactly with detectable antigen expression. Strong hybridization signals were observed in maternal cells (maternal endometrial gland epithelium and lymphocytes; Figure 2) hybridized with MHC class I antisense probes, while no hybridization signal was observed in cells in which antigen could not be demonstrated immunohistochemically [mature endometrial cup cells; Figure 2(a) and (c)l. A weak positive hybridization signal was detected within newly invaded cup cells (day 38 of gestation) hybridized with the horse MHC class I antisense probe (data not shown). This finding is in agreement with the immunohistochemistry results previously obtained by Donaldson et al. (1992), in which the intensity of labelling of immature endometrial cup cells by horse MHC class I specific monoclonal antibodies decreased between days 38 and 43 of gestation. Identical results were obtained using the horse and human MHC class I RNA probes, which suggests that both RNA probes detected the same population of MHC class I mRNA transcripts. Using the positive control antisense eCG-P oligonucleotide probe a strong hybridization signal was observed in endometrial cup cells, which are the sole source of eCG, from samples collected on days 38, 42,43, 54, 55 and 73 of gestation [Figure 3(a) and (b)]. No hybridization signal was observed in maternal cells (uterine gland and endometrium epithelium and maternal leukocytes). The pattern of expression of eCG-P mRNA transcripts detected in the endometrial cups was the inverse of that obtained with the MHC class I antisense probe [Figure 3(c)]. The positive control 1% RNA antisense riboprobe detected 18s RNA transcripts within all cell types in mature endometrial cup tissue from days 42, 43, 54, 55, 60 and 73 of gestation (data not shown).

Detection transcripts

of

MHC class I and in allantochorionic

eCG-P mRNA trophoblast

cells

Hybridization signals were not consistently observed from allantochorionic trophoblast cells interdigitating with maternal

beta subunit (KG-pa) specific oligonucleotide probes horse endometrial cup tissue hybridized with the eCG-P gland: one cell is delimited by arrows as an example. photomicrograph cup tissue hybridized with the MHC ec, endometrial cup cells; gl, maternal endometrial gland

endometrium hybridized with horse or human MHC class I or eCG-/3 antisense probes (data not shown). An intense hybridization signal was observed from all cell types within the allantochorion membrane hybridized with the positive 18s RNA antisense probe (data not shown).

Comparison of hybridization patterns chorionic girdle and endometrial cup in vitro and in vivo

in cells

The pattern of MHC classI and eCG-P hybridization detected in chorionic girdle cells cultured in vitro parallels the pattern of hybridization detected during differentiation into endometrial cup cells in vivo. A strong signal was observed in mononuclear, dividing cells harvested from chorionic girdle cell cultures maintained in vitro for 11-71 days and hybridized to the horse MHC classI antisense probe [Figure 4(a) and (b)], but not to the eCG-P antisense probe [Figure 4(c) and (d)]. In contrast, an intense hybridization signal was detected within the descendent, terminally differentiated endometrial cup cells hybridized to the eCG-P antisense probe, but not to the horse MHC class I antisense probe. No hybridization signals were observed when the cultures were hybridized with the corresponding negative control eCGj3 [Figure 4(e) and (f )] and MHC class I sense probes (data not shown). Strong hybridization signals were observed within all cell types in the chorionic girdle cell cultures hybridized to the positive control 18s RNA antisense probe (data not shown).

DISCUSSION

The finding that endometrial cup and allantochorionic trophoblast cells lack MHC class I mRNA by in situ hybridization is consistent with the hypothesis that class I genes are downregulated at the transcriptional level. Studies done in mice (Tanaka et al., 1983; Morello et al., 1985; Drezen, Babinet and Morello, 1993) and humans (Drew et al., 1993) suggest that

Maher

et al.: Equine

MHC

Class I In Situ Hybridization

357

Figure 4. Day 34 chorionic girdle explants maintained in culture for 25 days and hybridized using horse MHC class I and equine chorionic gonadotropin beta subunit (&G-p) gene probes. (a) High-power, brightfield photomicrograph showing a representative example of chorionic girdle and endometrial cup trophoblast cells maintained in vitro and hybridized with the horse MHC class I antisense RNA probe. The black dots over the two chorionic girdle cells shown represent the hybridization signal. Bar=29 pm. (b) Darkfield photomicrograph of (a). Bar=29 pm. cg, chorionic girdle cell; ec, endometrial cup. (c) Brightfield photomicrograph of cultured trophoblast cells hybridized with the eCG-B antisense oligonucleotide probe. The two large, polygonal cells (indicated by arrows) display the endometrial cup trophoblast phenotype, while the small, round cells represent chorionic girdle trophoblast. Nomarski optics, Bar= 18 pm. (d) Darkfield photomicrograph of (c). A strong hybridization was generated by the endometrial cup cells (indicated by arrows), but not by the mononucleate chorionic girdle cells, This is the reverse of the pattern obtained with the horse MHC class I probe in (b). Bar= 18 pm. (e) Brightfield photomicrograph of cultured trophoblast cells hybridized with the negative control eCG-B sense oligonucleotide probe. The round, mononucleate cells are chorionic girdle cells and the large, polygonal cell is an endometrial cup cell. Bar= 18 pm. (f ) Darkfield photomicrograph of (e). Bar= 18 pm.

down-regulation of MHC class I gene expression occurs at the transcriptional, rather than the post-transcriptional level in cell types that do not normally express class I antigens. In human syncytiotrophoblast, for example, a portion of the mechanism effecting transcriptional down-regulation of classical MHC class I genes through enhancer A binding has been reported (Boucraut et al., 1993). Although hybridization can be used to describe the pattern of mRNA expression in individual cell types, nuclear run-on transcription assays (Weber, Jelinek and Darnell, 1977) are necessary to assess the transcriptional state of MHC gene promoters definitively. These assays have not been done with horse trophoblast cells. The intriguing possibility exists that non-polymorphic MHC class I or class Ib molecules are expressed by horse

placental cells, similar to the non-polymorphic MHC class I molecule, HLA-G, expressed by human trophoblast cells (Kovats et al., 1990). However, this view is not supported by our findings. Although non-polymorphic MHC class I or class Ib molecules, such as TL in mice (Bradbury et al., 1988) and CD1 (reviewed by Calabi and Bradbury, 1991) are sufficiently divergent in the alpha 1 and alpha 2 domains from classical class I molecules that antigen expression may not have been detected by the panel of monoclonal antibodies used by Donaldson et al. (1992), the alpha 3 domain of these molecules is highly conserved (Calabi and Bradbury, 1991; Hughes, 1991). It seems likely, then, that this feature of non-classical class I or class I-like molecules present on endometrial cup cells would have been recognized by monoclonal antibodies

358

produced in our laboratory that appear to recognize framework determinants (Kydd et al., 1994). In support of this argument, expression of HLA-G cell surface antigen was detected on human extravillous cytotrophoblast cells using antibody W6/32, which recognizes a monomorphic determinant of HLA molecules (Sunderland, Redman and Stirrat, 1981). Unfortunately, a comparable antibody that recognizes a monomorphic determinant of equine MHC class I molecules is not available. At the mRNA level, no transcripts were detected in cup cells using either the human or horse MHC class I probes. In contrast, HLA-G mRNA transcripts were readily detected in human villous cytotrophoblast cells by in situ hybridization using a full-length cDNA probe encoding the polymorphic MHC classI gene HLA-A2 (Hunt et al., 1990), which was also used in our experiments. However, extensive screening of the equine trophoblast cDNA library produced in our laboratory with a probe specific for the conserved regions of the class I alpha 3 domain is required to address the nature of expression of non-classical MHC class I and class Ib genes in horse placental cells. MHC class I antigen expression is permanently and irreversibly down-regulated on endometrial cup cells (Donaldson et al., 1992). The maternal lymphocytes, which cluster around the endometrial cups, may be a source of cytokines that could mediate local upregulation of MHC antigen expression on maternal cells. Interleukin-2, interferon-y, and TNF-a, which

Placenta

(1996),

Vol. 17

induce MHC gene transcription (Singer and Maguire, 1990), can be detected in endometrial tissues and equine peripheral blood lymphocytes using reverse transcriptase PCR (Grunig and Antczak, 1995). Despite the presence of these cytokines, equine endometrial cup cells appear to resist the induction of MHC mRNA and antigen expression. We speculate that a change in the chromatin configuration of class I genes leading to a transcriptionally inactive state in endometrial cup cells could be responsible for this stringent level of control. The in situ hybridization results using chorionic girdle explant cultures paralleled those obtained for MHC class I and eCG immunoreactive protein using immunohistochemistry (Donaldson et al., 1992). The primary finding contributed by in situ hybridization (i.e. that MHC class I mRNA transcripts were not detected in the antigennegative endometrial cup cells) suggests that polymorphic, classical MHC antigen expression by endometrial cup cells may be regulated by the same mechanisms in vitro and in vivo. In conclusion, characterization of the pattern of MHC class I mRNA expression in horse placenta provides insights into the mechanisms that trophoblast cells may use to downregulate MHC antigen expression. Studying these mechanisms in diverse species may contribute to understanding of the fetal-maternal immunological interactions affecting the success of mammalian pregnancy.

ACKNOWLEDGEMENTS We would like to thank Dr Nick Holmes for the gift of human MHC class I cDNA, Dr John Nilson for the gift of &G-p specific antisense oligonucleotides, Dr Joanne Fortune for the gift of the plasmid DNA containing the bovine 18s RNA sequence, Dr Janet Rossant and MS Valerie Prideau for teaching the in situ hybridization assay to Dr Maher, and Drs David Shalloway and Willie Mark for many helpful discussions. This work was supported in part by USPHS grants lF32 HD-07436 (to J.K.M.), HD-15799 (to D.F.A.), by the Harry M. Zweig Memorial Fund for Equine Research, and by the Dorothy Russell Havemeyer Foundation, Inc.

REFERENCES Allen, W. R. & Moore, R. M. (1972) The origin of the equine endometrial cups. I. Production of PMSG by fetal trophoblast cells. 3’ournal of Reproduction and Fertzlity, 29, 3 13-3 16. Antczak, D. F., Bright, S. M., Remick, L. H. & Bauman, B. E. (1982) Lymphocyte alloantigens of the horse. I. Serologic and genetic studies. Tissue Antigens, 20, 172-187. Antczak, D. F., Miller, J. M. & Remick, L. H. (1984) Lymphocyte &antigens of the horse. II. Antibodies to ELA antigens produced during equine pregnancy. 3ournal of Reproductive Immunology, 6, 283-297. Antczak, D. F., Oriol, J. G., Donaldson, W. L., Poleman, C., Stenzler, L., Volsen, S. G. & Allen, W. R. (1987) Differentiation molecules of the equine trophoblast. 3oburnaZ of Reproduction and Fertility, (Suppl.), 35, 371-378. Barbis, D. P., Maher, J. K., Stanek, J., Klaunberg, B. A. & Antczak, D. F. (1994) Horse cDNA clones encoding two MHC class I genes. Immunogenetics, 40, 163. Boucraut, J., Hawley, S., Robertson, K., Bernard, D., Loke, Y. W. & Le Bouteiller, P. (1993) Differential nuclear expression of enhancer A DNA-binding proteins in human first trimester trophoblast cells. 3ournal of Immunology, 150, 3882-3894. Bradbury, A., Belt, K. R., Neri, T. M., Milstein, C. & Calabi, F. (1988) Mouse CD1 is distinct from and co-exists with TL in the same thymus. ZSl’lBO ff, 7, 3081-3086. Calabi, F. & Bradbur&A. (1991) The CD1 system. A review. Tissue Antigens, 37, l-9.

Grump, A., Donaldson, W. L., Miller, J., Kydd, J., Allen, W. R. & Antczak, D. F. (1987) Expression of major histocompatibility complex (MHC) antigens on equine trophoblast. 3ournal of Reproductzon and Fertility, (Suppl.), 35, 379-388. Donaldson, W. L., Oriol, J. G., Pelkaus, C. L. & Antczak, D. F. (1994) Paternal and maternal major histocompatibility complex class I antigens are expressed co-dominantly by equine trophoblast. Placenta, 15, 123-135. Donaldson, W. L., Oriol, J. G., Plavin, A. & Antczak, D. F. (1992) Developmental regulation of class I major histocompatibility antigen expression by equine trophoblastic cells. Dzfirmtiation, 52, 69-78. Donaldson, W. L., Zhang, C., Oriol, J. G. & Antczak, D. F. (1990) Invasive equine trophoblast expresses conventional class I major histocompatibility complex antigens. Development, 110, 63-71. Drew, P. D., Lonergan, M., Goldstein, M. E., Lampson, L. A., Ozato, K. & McFarlin, D. E. (1993) Regulation of MHC class I and µglobulin gene expression in human neuronal cells: factor binding to conserved cis-acting regulatory sequences correlates with expression of the genes. &Lmal ofZmmmunoZogy, 150, 3300-3310. Drezen, J, M., Babinet, C. & Morello, D. (1993) Transcriptional control of MHC class I and p,-microglobulin genes in viva. 3ournal of Immunology, 150, 2805-2813. Ellis, S. A. & Martin, A. J. (1993) Nucleotide sequence of horse µglobulin cDNA. Immunogenetics, 38, 383. Ellis, S. A., Sargent, I. L., Redman, C. W. G. & McMichael, A. J. (1986) Evidence for a novel HLA antigen on human extravillous trophoblast of a choriocarcinoma cell line. Immunology, 59, 595-601.

Maher

et al.: Equine

MHC

Class I In Situ Hybridization

Grunig, G. & Antczak, D. F. (1995) Horse trophoblasts produce tumor necrosis factor alpha but not interleukin 2, interleukin 4, or interferon gamma. Biology of Reproduction, 52, 531-539. Holmes, N., En&, P., Wan, A. M., Denney, D. W. & Parham, P. (1987) Multiple genetic mechanisms have contributed to the generation of the HLA-A2/A28 family of class I MHC molecules. 3ournal of Immunology, 139, 936941. Hughes, A. L. (1991) Evolutionary origin and diversification of the mammalian CD1 antigen genes. Molecular Biology 6 Evolution, 8, 185-201. Hunt, J. S., Fishback, J. L., Chumbley, G. & Lake, Y. W. (1990) Identification of class I MHC mRNA in human first trimester trophoblast cells by in situ hybridization. 3ournal of Immunology, 144, 44204425. Hunt, J. S. & Orr, H. T. (1992) HLA and maternal-fetal recognition. FASEB 3., 6, 2344-2348. Kovats, S., Main, E. K., Librach, C., Stubblebine, M., Fisher, S. J. & DeMars, R. (1990) A class I antigen, HLA-G, expressed in human trophoblast. Science, 248, 220-223. Kydd, J., Antczak, D. F., Allen, W. R., Barbis, D., Butcher, G., Davis, W., Duffus, W. P. H., Edington, N., Grunig, G., Holmes, M. A., Lunn, D. P., McCulloch, J., O’Brien, A., Perryman, L. E., Tavernor, A., Williamson, S. & Zhang, C. (1994) Report of the First International Workshop on Equine Leucocyte Antigens, Cambridge, U.K., July 1991. Veterinary Immunology and Immunopathology, 42, 3-60. Medawar, P. B. (1953) Some immunological and endocrinological problems raised by evolution of viviparity in vertebrates. Symposia of the Society oj Experimental Biology, 11, 320-338. Morello, D., Duprey, P., Israel, A. & Babinet, C. (1985) Asynchronous regulation of mouse H-2D and beta-2 microglobulin RNA transcripts. Immunogenetics, 22, 441452. Redman, C. W. G., Arenas, J. & Sargent, I. L. (1987) Maternal antifetal immune responses during human pregnancy. In Reproductive Immunology: Materno-Fetal Relationship Colloque Inseam (Ed.) Chaouat, G. pp. 25-32. Regan, L. (1988) A prospective study of spontaneous abortion. early pregnancy loss: mechanisms and treatment. In: Proceedings of the 18th Study

359 Group of the Royal College of Obstetricians and Gynaecologists, October I987 (Ed.) Beard, R. W. & Sharp, F. pp. 2342. Sherman, G. B., Wolfe, M. W., Farmerie, T. A., Clay, C. IM., Threadgill, D. S., Sharp, D. C. & Nilson, J. H. (1992) A single gene encodes the B-subunits of equine luteinizing hormone and chorionic gonadotropin. Molecular Endocrinology, 6, 951-959. Singer, D. S. & Maguire, J. E. (1990) Regulation of the expression of class I MHC genes. In CRC Critical Reviews in Immunology, pp. 235-257. Spear, B. T., Kornbluth, J., Strominger, J. L. & Wilson, D. B. (1985) Evidence for a shared HLA-A intralocus determinant defined by monoclonal antibody 131. Joioumnal of Experimental Medicine, 162, 1802-1810. Sunderland, C. A., Redman, C. W. G. & Stirrat, G. M. (1981) HLAA,B,C antigens are expressed on nonvillous trophoblasts of the early human placenta. 3oumal of Immunology, 127, 26142615. Tanaka, K., Ozato, K., Jay, G., Parnes, J. R., Ramanathan, L., Seidman, J. G., Chang, K. S. S. & Appella, E. (1983) Control of H-2 antigen and pa-microglobulin gene expression in mouse trophoblast cell clones. Proceedings of the National Academy of Sciences, USA, 80, 5597-5601. van Rood, J. J., Eernisse, J. G. & van Leeuwen, A. (1958) Leukocyte antibodies in sera from pregnant women. Nature, 181, 1735-1736. Weber, J., Jelinek, W. & Darnell, J. E. (1977) The definition of a large viral transcription unit late in Ad2 infection of HeLa cells; mapping of the nascent RNA molecules labeled in isolated nuclei. Cell, 10, 611-616. Wilkinson, D. G. & Green, J. (1991) In situ hybridization and the three-dimensional reconstruction of serial sections. In: Post-implantation Mammalian Embryos: A Practical Approach (Ed.) Copp, A. J. and Cockroft, D. L. pp. 151-171. Oxford University: IRL Press. Wood, G. W. (1994) Is restricted antigen presentation the explanation for fetal allograft survival? Immunology Today, 15, 15-18. Yelavarthi, K. K., Fishback, J. L. & Hunt, J. S. (1991) Analysis of HLA-G mRNA in human placental and extraplacental membrane cells by in situ hybridization. 3ownal of Immunology, 146, 2847-2854.