Differential responses of phenotypically distinct rat trophoblast cell lines to MHC class I antigen-inducing cytokines

Placenta(1994), 15, 577-590

Differential Responses of Phenotypically Distinct Rat Trophoblast Cell Lines to MHC Class I Antigen-Inducing Cytokines K. F. ROBY a, G. P. HAMLIN b, M.J. SOARES b & J. S. H U N T a'c ~Department ofAnatomy and Cell Biology and bDepartment of Physiology, Universi01of Kansas Medical Center, Kansas CioJ, Kansas 66160-7400, USA cTo whom all correspondenceshould be addressed Paper accepted30.3.1994

SUMMARY

Phenotypically distinct rat trophoblast cell lines, the Rcho-1 and R8RP.3 cells, were compared for their responses to cytokines known to induce major histocompatibility (MHC) class I antigens, tumour necrosisfactor ~ F ) , transforming growth factor(TGF-~), and interferon-~ (IFN-~I). Cell enzyme immunosorbent assays and flow cytometry experiments showed that only IFN-~ could induce RT1 class I antigens on the Rcho-1 cells. Non-adherent cells were slightly less responsive than adherent, giant cell-like Rcho-1 cells. By contrast, RT1 class I antigens on the R8RP.3 cells were induced by both TGF-~I and IFN-~I. The cytokines also had different effects on mitochondrial enzyme activity in the two lines. TNF and TGFE1 mRNAs were demonstrated in both lines by using Northern blot hybridization. Rcho-1 but not R8RP.3 cells contained two TNF messages (-2.2, 1.9 kb). Steady state levels of transcripts from the TNF gene, and, to a lesser extent, the TGF-~I gene, were increased in cultures of Rcho-1 cells that contained high proportions of giant cells. Thus, phenotypically distinct rat trophoblast cell lines do not respond identically to TNF, TGF-~I or IFN-~, transcription of cytokine genes does not prevent the cellsfrom responding to paracrine cytokine signals, and the cells contain novel TNF transcripts that might be important in cell maturation or differentiation.

INTRODUCTION Major histocompatibility complex (MHC)-derived antigens participate in T-lymphocyte recognition and destruction of infected and foreign cells, associate with growth factor receptors (Schreiber, Schlessinger and Edidin 1984; Fehlmann et al, 1985) and are thought to be important in cell to cell communication. Expression of these antigens is not 0143-4004/94/060577 + 14 $08.00/0

9 1994 W. B. Saunders CompanyLtd

578

Placenta (1994), Vol. 15

constitutive but is, instead, regulated by multiple intrinsic and extrinsic mechanisms (Singer and Maguire, 1990; David-Watine, Israel and Kourilsky, 1990). Three polypeptide growth factors, tumour necrosis factor-or (TNF), transforming growth factor-13 (TGF-131) and inteferon-'y (IFN-~/), enhance cellular expression of the MHC class I antigens (Rosa and Fellous, 1984; Collins et al, 1986; McGee et al, 1991). These cytokines, which may also influence cell growth (reviewed by Keski-Oja and Moses, 1987; Baglioni, 1992; De Maeyer and De Maeyer-Guignard, 1992), are synthesized in mammalian uteri and/or placentae during pregnancy (Tamada et al, 1990; Chen et al, 1991; Yelavarthi et al, 1991; Hunt et al, 1993; Lin et al, 1993; Chen et al, 1994b), raising the possibility that one or other of the factors may influence MHC class I antigen expression by trophoblast cells. To explore this and other questions, rat trophoblast cell lines have been generated. One, the Rcho-1 cell line, was derived from a transplantable rat choriocarcinoma (Faria and Soares, 1991). The Rcho-1 cells are capable of morphological and functional differentiation along the trophoblast giant cell lineage, including synthesis of members of the placental prolactin family (Faria and Soares, 1991). Six lines of apparently normal trophoblast cells have been generated from outgrowths of mid-gestation rat placentae (Soares et al, 1987; Hunt et al, 1989a). The lines, which include the R8RP.3 cells tested in this study, express trophoblast-specific markers but do not synthesize placental hormones (Soares et al, 1993) and give no evidence of acquiring differentiated cell characteristics during in vitro culture. In this study, we investigated expression of RT1 class I antigens by the Rcho-1 cells and evaluated their ability to respond to TNF, TGF-131 and IFN-~/with enhanced antigen expression. Comparisons were drawn with the R8RP.3 cells, which express low levels of IFN-3,-inducible antigens (Hunt et al, 1989b, 1990, 1992). Because the two phenotypically and functionally distinct lines did not respond in an identical manner, Northern blot hybridization studies were done in an effort to learn whether constitutive transcription of cytokine genes, TNF or TGF-131, might have influenced their responses Odanhaesebroeck et al, 1992; Okamoto et al, 1992).

MATERIALS AND M E T H O D S Cell lines and separation procedures The Rcho-1 cell line was established from a transplantable rat choriocarcinoma (Faria and Soares, 1991). The cells were maintained in RPMI 1640 supplemented with glutamine, antibiotics, 1 mM sodium pyrnvate, 5 x 10 -5 M 2-mercaptoethanol, and 10 per cent fetal bovine serum. For some experiments, Rcho-1 cells were grown for 24h longer after confluency was reached so as to enrich for giant cells (Faria and Soares, 1991). The giant cells adhere tightly to tissue culture flasks. Non-adherent cells Were harvested from the culture supernatant by centrifugation and adherent cells were harvested from the tissue culture flasks by brief trypsinization (0.25 per cent bovine trypsin-0.2 per cent EDTA). The R8RP.3 trophoblast cell line was derived from day 11/12 PVG.RT1 r8 x PVG.RTV 8 rat placentae (RT1 a) (Hunt et al, 1989a), was maintained in the same growth medium and was passaged by brief trypsinization. Passage 20 or less of the R8RP.3 cells was used for experiments. Rcho-1 and R8RP.3 cells were washed extensively and were tested for viability by trypan blue exclusion prior to using the cells for experiments. Viability of all preparations was consistently >98 per cent.

Roby et al: CytokineRegulation ofMHC Class I Expression in Rat Trophoblast Cells

579

Cytokines Recombinant mouse TNF-ot was a gift from Genentech, South San Francisco, CA. Porcine TGF-I31 was purchased from R & D Systems Inc., Minneapolis, MN, and recombinant rat IFN-7 (specific activity, 1 • 106U/mg protein) was purchased from Amgen, Thousand Oaks, CA.

Cell cultures For each cell culture experiment, duplicate plates were prepared. Test cells were dispensed into 96-well flat-bottom microplates (5 • 104 cells/well in 100 I~1). Cytokines contained in 100 p~l of the appropriate medium were added, and the plates were incubated for 24h at 37~ 5 per cent CO2. One hundred microlitres of medium were removed and replaced with fresh medium or fresh medium containing cytokines. Replacement was repeated at 48 h, but not at 72 h, when the experiments were terminated. Three to five well replicates were used for each determination. Means and standard deviations were determined for the replicates. Each type of assay was repeated a minimum of three times.

Assay of mitochondrial enzyme activity (MTT assay) Mitochondrial enzyme activity was assessed in the wells of one microplate after 72 h of incubation by using a modified M T T [3-(4,5-dimethylthiazole-2-yl)-2.5 diphenyl tetrazolium bromide] assay (Mosmann, 1983; Hunt, Atherton and Pace, 1990). This colorimetric assay measures dehydrogenase enzymes that are active only in the mitochondria of living cells, and which cleave the tetrazolium ring, reducing the M T T to a formazan product. Microplates were centrifuged for 5 min at 800 x g, then medium was removed from the wells. The cell layers were washed once with phenol red-free RPMI 1640 buffered with 25 mM Hepes. M T T substrate [1.0 mg/ml in phenol red-free RPMI 1640, 100 p~l (Sigma, St Louis MO #M2120)] was added to each well. The microplates were then incubated at 37~ for 3 h. The plates were centrifuged as before, and the M T T solution was carefully removed from the wells. Anhydrous isopropanol (100 wl, Sigma, St Louis, MO #405-7) was added, and the plates were shaken vigorously on an orbital shaker. Optical density was read at 570 nm using a Titer-Tek Multiskan microplate reader (Flow Lab), after blanking on isopropanol.

Cell enzyme immunosorbent assay (cell ELISA) A cell ELISA was performed on the second microplate as previously described (Hunt, Atherton and Pace, 1990). Cells in the wells were fixed in freshly prepared 0.05 per cent glutaraldehyde, and the fixation was quenched with 0.2 M glycine. Non-specific binding was prevented by blocking with 5 per cent Carnation Milk in phosphate buffered saline (PBS), then washing the plates three times in assay buffer [Tris-buffered saline (TBS) containing 0.02% thimersol and 0.05% Tween-20]. A mouse anti-rat MHC class I monoclonal antibody (OX18, Sera-Lab, MCA51, Bioproducts for Science, Indianapolis, IN, 1:500, ascites) or an equivalent concentration of control mouse IgG (Sigma, St Louis, MO) diluted in PBS containing 5 per cent milk, was added and the plates were incubated for 30 min at 37~ The plates were washed three times in TBS, and 100 ixl/well of the developing antibody was added (goat anti-mouse IgG conjugated to horseradish peroxidase, 1:1000, Tago Inc., Burlingame, CA). The plates were again incubated at 37~ for 30 min, then washed three times in TBS. Freshly prepared TMB peroxidase substrate (Kirkegaard and Perry, Gaithersburg, MD, 100 ixl/well) was added and the plates were

580

Placenta H994), Vol. 15

incubated for 15 min at RT. The reaction was halted by adding 50 ~l/well of 1 M H3PO4, and then the plates were read at 450 nm. Flow cytometry Rcho-1 cells separated as described above were adjusted to 1 • 107 cells/ml in PBS (pH 7.2), and were incubated with either the mouse anti-rat class I monoclonal antibody, OX18, or with normal mouse IgG. Binding was detected using fluoresceinated goat anti-mouse IgG, and labelled cells were analysed by flow cytometry as previously described (Hunt and Soares, 1988; Hunt et al, 1992). Molecular probes A 298 bp EcoRI-HindlII fragment of Mouse TNF-ot (Fransen et al, 1985; Murray and Martens, 1989) cloned into pGEM3 (Promega Biotec, Madison, WI) and a 340bp fragment of the mouse TGF-[31 gene (Derynck et al, 1986) cloned into pGEM3 were gifts from Dr Christine Marten. After linearization of the plasmids, 32p-labelled antisense RNA probes were synthesized by in vitro transcription using T7 RNA polymerase in the presence of ot-32p-GTP (DuPont/NEN Products, Wilmington, DE). The glucose-3-phosphate dehydrogenase (G3PDH) DNA probe was synthesized from a 1.2 kb EcoRI fragment of G3PDH cDNA (a gift from R. W. Allen, American Red Cross Blood Services, St Louis, MO) cloned into pGEM3z plasmid (Promega). After digestion with restriction enzymes and separation on a 1 per cent agarose gel, the insert was excised and eluted from the gel using the QIAEX Gel Extraction kit (Qjagen Inc., Chatsworth, CA). 32p-labelled probe was synthesized using the BRL Random Primer DNA labelling system (Life Technologies, Gaithersburg, MD). Northern blot hybridizations The mouse macrophage cell line, RAW.264, was grown in tissue culture flasks for 4 h in the presence of 10 t~g/ml of S. minnesota R595 LPS (Galanos et al, 1969; a gift from D. C. Morrison, Department of Microbiology, Immunology and Molecular Genetics). Total cellular RNA was isolated from these cells as well as the Rcho-1 and R8RP.3 cells using a modification of the technique of Chomczynski and Sacchi (1987). Rcho-1 and R8RP.3 cell RNA was enriched for poly(A) § RNA by one cycle of oligo-dT cellulose chromatography (Aviv and Leder, 1972). The RNA samples were denatured in 2.2M formaldehyde/50 per cent formamide and were separated by electrophoresis in 2.2M formaldehyde/1.5 per cent agarose gels. Ten micrograms of total cell RNA from the RAW.264 cells were loaded per lane and 2 ~g ofpoly(A) § RNA from the trophoblast cells were loaded per lane. RNAs were transferred to Nytran filters (Schleicher and Schuell, Keene, NH) as described by Thomas (1980). Hybridizations were carried out as described previously (Yang et al, 1993; Roby et al, 1993). The filters were exposed to Kodak XAR film with intensifying screens.

RESULTS RT1 class I expression by untreated and cytokine-treated rat trophoblast cell lines Constitutive expression of RT1 class I antigens was low in both cell lines. In 10 separate cell ELISA assays, binding of the mouse anti-rat monoclonal antibody, OX18, to untreated

581

Roby et al: CytokineRegulation ofMHC Class I Expression in Rat Trophoblast Cells

Rcho-1 and R8RP.3 cells averaged A450 0.05 and 0.09, respectively, after subtraction o f values obtained for normal mouse IgG (A45o averages o f 0.15, 0.05). Treatment with cytokines had no reproducible statistically significant influence on binding of normal mouse IgG by either line. T N F had no discernible effect on expression o f R T I class I antigens by either line [Figure 1 (a)]. While the same appeared to be true for TGF-131 [Figure 1 (b)], M T T values Cell line

TNF (U/ml)

Reho-1

0 2.5 25 250

R8RP.3

0 25

(a)

m 0

Cell line TGF-13(ng/ml! Rcho-1 0 0.1 1 10 R8RP. 3

(b)

m

0 1 0

Cell line

IFN-~/(U/ml)

Rcho-1

0 10 100 1000

R8RP.3

200 400 600" 800 1000 Relative class I expression

200 400 600 800 1000 Relative class I expression

0

100 0

200 400 600 800 1000 Relative class I expression Figure 1. Effects of(a) TNF, (b) TGF-[31 and (c) IFN-~/on expression of RT1 class I antigens by Reho-1 and R8RP.3 cells. The cell ELISA was used to assess changes after 72h. The results are given as relative class I expression and are normalized values that were calculated as described in materials and methods. for Rcho-1 cells had decreased somewhat in the Rcho-1 cells and had declined markedly in the R8RP.3 cells following exposure to TGF-131 [Figure 2Co)]. Cell E L I S A data acquired by testing the effects of a range o f TGF-131 concentrations, then normalizing to M T T values (Figure 3), indicated that while TGF-131 did not enhance Rcho-1 RT1 class I antigens, it was a highly effective dose-dependent inducer o f R8RP.3 cell antigens.

582

Placenta (1994), Vol. 15

Cell line

TNF (U/ml)

Rcho- 1

0 2.5 25 250

R8RP.3

I I I I

0 25

! 0.0

(a)

0.5

1.0

1.5

Proliferation

2.0

(A570)

Cell line TGF-13(ng/ml) Reho-1

R8RP. 3

0 0.1 1 10

! !

0 1

D 0.0

0.5

1:o

(b) 2.0

1:5

Proliferation (As70) Cell line IFN-7 (U/ml) Reho-1

R8RP.3

0 10 100 1000 0

100

! !

D I

(c) 0.5 1.0 1.5 2.0 Proliferation (A570) Figure 2. Effects of(a) TNF, (b) TGF-131 and (c) IFN-3' on Rcho-1 R8RP.3 cell mitochondrial activity,assessed by performing MTT assays for mitochondrial enzyme activity (Proliferation, 570 nm) after 72 h of culture. Data shown here were acquired from microplates that were duplicates of those shown in Figure 1 (cell ELISA assays). 0.0

Figure 1 (c) shows that treatment with I F N - ~ / e n h a n c e d RT1 class I expression by both Rcho-1 and R8RP.3 cells. A somewhat lesser degree of up-regulation was evident in the Rcho-1 cells (fivefold over control) than in the the R8RP.3 cells (eightfold over control) when comparisons were made between A450 readings of uninduced cells and cells exposed to 100 U / m l of rat I F N - 7 . This difference was evident in three separate experiments. Effects o f c y t o k i n e s on m i t o c h o n d r i a l e n z y m e activity in r a t t r o p h o b l a s t cell lines For each cell E L I S A experiment, duplicate microplates were prepared and tested by using the M T T assay, which assesses mitochondrial enzyme activity. This assay was chosen for its usefulness in comparing the effects of cytokines on cells that can endoreduplicate, as is the case for the Rcho-1 cells, or proliferate, which happens with the R8RP.3 cells. Figure 2(a) shows that T N F had no detectable influence on Rcho-1 cells whereas R8RP.3 cell mitochondrial enzyme activity was markedly decreased. As shown in Figure

Roby et al: CytokineRegulation of MHC Class I Expression in Rat TrophoblastCells

583

12 ~ A

A

~

10 8

9

6

4 2 0

.......

0

'

.

.

.

.

.

.

.

.

.

i

i

,

,

,llll

0.1

1.0 10 TGF-~ (ng/ml) Figure 3. TGF-13inducesRT1 class I antigenson R8RP.3 cellsbut not on Rcho-1 cells.The abitraryunitsshown on the • axis indicatethe ratio betweenA45o readingsfor class I antigens(cellELISA)and A57o readingsfor cell numbers (MTT assays).(---~), Rcho-1; (-A-), R8RP.3. 2(b), treatment with TGF-I31 diminished M T T values for both Rcho-1 and R8RP.3 cells. The cytokine was a less effective inhibitor of Rcho-1 than R8RP.3 cells; in the experiment shown, Rcho-1 cells treated with 1.0 ng/ml of TGF-[31 were inhibited by 34 per cent while R8RP.3 cells were inhibited by 92 per cent. Figure 2(c) shows that the responses of the two lines to IFN--/were also different. No inhibition of the Rcho-1 cell M T T reading was detected whereas mitochondrial enzyme activity in microwells containing R8RP.3 cells was markedly decreased [Figure 2(c)]. Microplates containing untreated and cytokine-treated trophoblast cells were inspected using an inverted microscope. As illustrated for IFN-7 in Figure 4, cell density was roughly correlated with M T T values [Figure 2(c)]. Effects o f IFN-7 on RT1 class I antigens expressed by n o n - a d h e r e n t and adherent Rcho-1 cells In situ, rat trophoblast cells may either retain their M H C class I antigens in the cytoplasm or express the antigens on their membranes (Kanbour et al, 1987). Therefore, induction of M H C class I antigens by IFN-7 was investigated by using flow cytometry, which specifically identifies cell membrane antigens. Non-adherent and adherent Rcho-1 cells were separated in order to learn whether there were any differences in the inducibility of the two morphologically and phenotypically distinct subpopulations (Faria and Soares, 1991). Table 1 shows that few non-adherent or adherent Rcho-1 cells were RTI class I positive prior to exposure to IFN-7. In two experiments, induction of cell surface antigens on approximately 75 per cent of the non-adherent cells was accomplished by IFN-~/. By contrast, nearly all of the adherent cells were positive for membrane RT1 class I antigens following IFN-7 treatment. The fluorescent intensity (mean channel of fluorescence) was consistently lower for non-adherent cells (average of 51 in two experiments) than for adherent cells (average of 73), indicating that lower densities of RT1 class I antigens were induced on the former than on the latter subpopulation. Background binding of mouse IgG in cell ELISA, described above, suggested that the Rcho-1 cells expressed IgG Fc receptors. This was confirmed by flow cytometry, which

Placenta (1994), Vol. 15

584

Figure 4. Morphology of IFN-~/-treated cultures of Rcho-1 and R8RP.3 cells. R8RP.3 rat trophoblast cells grown

for 72 h in medium (a), 10 U/ml of recombinant rat IFN-~ (b), and 100 U/ml of IFN-v (c), Rcho-1 cells grown for 72 h in medium (d), 10 U/ml of lFN-~ and (e), 100 U/ml of IFN-~. (x400).

Table 1. Flow cytometric analysis of MHC class I antigens on non-adherent and adherent

Rcho-I cells exposed or unexposed to IFN--/

IFN-v U/ml

Experiment 1

Experiment 2

% Positive Intensity*

% Positive Intensity

Non-adherent Non-adherent

0 100

5 74

22 42

<1 78

25 60

Adherent Adherent

0 100

17 96

20 67

1 96

25 79

* Peak channel of fluorescence. showed that both n o n - a d h e r e n t and a d h e r e n t Rcho-1 cells b o u n d normal mouse I g G [Figure 5(a) and (c)]. T h e mean channel o f fluorescence was low and was the same (channel 25) for the two subpopulations. Class I antigens on I F N - ~ - t r e a t e d n o n - a d h e r e n t and a d h e r e n t Rcho-1 cells are shown in Co) and (d). T h e F c receptor positive cells were among those that bound the anti-class I reagent, as shown by the decrease in positive cells at channel 25 and below.

Roby et al: Cytokine Regulation of MHC Class I Expression in Rat Trophoblast Cells M o u s e IgG

585

Anti-class I (b)

..

~.~..::'_.:":'.":-. ......... ": (a.). (c)

(d)

~t~

t

13

256 13 256 Figure 5. Assessment of Fc receptors [binding of mouse IgG, (a) and (c)] and M H C class I antigens [binding of a mouse anti-rat M H C class I monoclonal antibody, panels (b) and (d)] expressed on IFN-~-treated non-adherent (a) and (b), and adherent (c) and (d), Rcho-I cells using flow cytometry.

Detection of TNF and TGF-[31 transcripts in trophoblast cells by Northern blot hybridization Because constitutive synthesis of cytokines is known to influence cellular responses to the same factors, Northern blot hybridization studies were done to determine whether either Rcho-1 or R8RP.3 cells contained transcripts from the TNF and TGF-[31 genes.

i

i

<

1.9 kb

Figure 6. Identification o f T N F mRNA by Northern blot hybridization. Ten p~g of total cell RNA was loaded for the RAW.264 cells that were used as a positive control. Two micrograms of poly(A) § RNA of the other samples were loaded per lane. R83, R8RP.3 cells; Rcho-1 D2, Rcho-1 cells grown in culture for 2 days; Rcho-1 D9, Rcho-! cells grown in culture for 9 days; RAW, RAW.264 cells.

Figure 6 shows that RNA taken from lipopolysaccharide-stimulated RAW.264 cells, from Rcho-1 cells and from R8RP.3 cells contained T N F mRNA that migrated to the expected position at 1.9kb. Rcho-1 cells also contained a larger transcript (-2.2 kb) that was present in both early cultures (day 2), where undifferentiated cells predominate, and in post-confluent cultures (day 9), which contain a high proportion of differentiated cells. Values for T N F mRNA relative to G 3 P D H mRNA are shown in Figure 7. Steady state levels o f T N F mRNA were similar in R8RP.3 and Rcho-1 cells, and in the latter line were higher in day 9 cultures than in day 2 cultures.

Placenta (1994), Vol. 15

586

3.0

? 1.0 a~

0.8

o~

~

0.6

Z 0.4 r~ 0.2 0.0 Rcho- 1 D9 Reho- 1 D2 R83 RAW Figure 7. Proportion ofTNF ([3) mRNA to glucose-3 phosphate dehydrogenase (G3PDH) mRNA and TGF-I31 (In) mRNA to G3PDH mRNA in Rcho-1 cells grown for 9 and 2 days in culture, R8RP.3 cells and RAW.264 cells. Values were obtained by using a scanning densitometer. Rcho-1 D9, Rcho-1 cells grown in culture for 9 days; Rcho-1 D2, Rcho-1 cells grown in culture for 2 days; R83, R8RP.3 cells; RAW; RAW.264 cells.

i

o

i

o

oO

<:

,2.5 kb

Figure8. Identification oftransforminggrowth factor-131 mRNA by Northern blot hybridization. Ten micrograms of total cell RNA was loaded for the RAW.264 cells that were used as a positive control. Two micrograms of poly(A)+ RNA of the other samples were loaded per lane. R83, R8RP.3 cells; Rcho-I D2, Rcho-1 cells grown in culture for 2 days; Rcho-1 D9, Rcho-1 cells grown in culture for 9 days; RAW, RAW.264 cells. Figure 8 shows that all three lines, R A W . 2 6 4 cells, Rcho-1 cells and R8RP.3 cells, contained TGF-131 m R N A . Each preparation exhibited a single band at the expected position (2.5kb). Values for TGF-131 m R N A relative t o G 3 P D H m R N A are shown in Figure 7. Steady state levels o f T G F - I 3 1 m R N A were similar in R8RP.3 and Rcho-1 cells, and specific m R N A in the latter line was slightly increased in day 9 as c o m p a r e d with day 2 cultures.

DISCUSSION T h e responses o f Rcho-1 and R8RP.3 cells to M H C class I - i n d u c i n g cytokines were similar for T N F and I F N - ~ / b u t were different for TGF-131. Opposite effects were obtained when cytokine influences on cell growth were assessed; the two lines r e s p o n d e d differently to T N F and I F N - ~ / b u t similarly to TGF-131. Collectively, therefore, the data

Roby et ak Cytokine Regulation of MHC Class I Expression in Rat Trophoblast Cells

587

are consistent with an association between specific sets of phenotypic characteristics (Soares et al, 1993) and the ultimate effects of growth factors on trophoblast cells. T N F did not induce RT1 class I antigens on either Rcho-1 or R8RP.3 cells, which may have been due to failure of stimulation of Type I interferons (Hunt et al, 1990). The Rcho1 cells were also refractory to TNF-mediated growth inhibition. It does not seem likely that this was due to a lack of T N F receptors, which are expressed on essentially all cells except erythrocytes (reviewed by Pfizenmaier et al, 1992). Mouse blastocysts (Pampfer et al, 1994) and human trophoblast cells (Yelavarthi and Hunt, 1993) express T N F receptors. In other types of cells, resistance is associated with constitutive transcription of the T N F gene (Okamoto et al, 1992; Vanhaesebroeck et al, 1992). However, Northern blot hybridization studies revealed that the T N F gene was transcribed in both TNF-resistant Rcho-1 cells and TNF-sensitive R8RP.3 cells. Two species of T N F mRNA were identified in the Rcho-1 cells. A similar doublet has been observed in RNA from cycling mouse uteri (K. F. Roby and J. S. Hunt, unreported results) and more than one T N F transcript is not uncommon (De et al, 1993; Yang et al, 1993). While novel cytoldne messages are particularly abundant in mouse placentae (Craine, Guilbert and Wegmann, 1990), it remains to be determined whether or not these translate into functionally discrete proteins. Steady state levels of T N F mRNA were higher in Rcho- 1 cells that had been cultured for 9 days, which increases the proportion of giant cells. Further experimentation needs to be done in order to learn whether or not, as with mouse macrophages (Myers et al, 1989; Witsell and Schook, 1992), trophoblast differentiation might be facilitated in some manner by autocrine TNF. TGF-131, which was an efficient inducer of class I antigens of the R8RP.3 cells, had no effect on Rcho-1 antigens. Induction of M H C class I antigens is not a well described function of TGF-131 although there have been a number of reports on its ability to diminish M H C class II antigens. Possibly, this is due to the difficulty of studying induction on growth-inhibited cells, a problem that we surmounted by using duplicate microplates. Rcho-1 cells and R8RP.3 cells were growth-inhibited by TGF-f3, thus demonstrating that receptors were present and functional. Both lines contained TGF-f~I transcripts, so the presence of mRNA was not clearly associated with resistance to the growth-inhibiting property of TGF-131. It was not surprising to find TGF-131 transcripts in Rcho-1 and R8RP.3 cells; rat trophoblast cells tested by in situ hybridization have been shown to contain abundant TGF-131 mRNA (Chen et al, 1994b). IFN-~/ has been studied extensively for its effects on trophoblast cell M H C class I antigens. Under normal circumstances, expression of M H C class I antigens is related to stage of gestation and is strictly limited to specific subpopulations of trophoblast cells (reviewed by Head, Drake and Zuckermann, 1987). Trophoblast giant cells taken from mouse ectoplacental cones do not express M H C class I antigens, and antigens cannot be induced by IFN-% The antigens first appear at mid-gestation in vacuolated trophoblast cells ('glycogen cells') in the spongiotrophoblast re#on of the placenta, which is also known as the basal or junctional zone. Susceptibility to up-regulation by exogenously administered IFN-~/ develops in the same subpopulation at the same time point in mouse gestation (Mattsson et al, 1991, 1992). Neither giant cells nor labyrinthine trophoblast express significant levels of M H C class I antigens. In situ hybridization studies suggest that, in the former, this is due to lack of transcripts for the 132-microglobulin light chains that associate with M H C class I heavy chains whereas in the latter, it is due to lack of heavy chain transcripts (Jaffe, Robertson and Bikoff, 1991). Failure of M H C class I expression by certain subpopulations of mouse trophoblast cells

588

Placenta (1994), VoL 15

might also be due to an inability to respond to the class I-inducing agent, IFN-% Recent experiments in our laboratory have shown that messages from the IFN-~/receptor gene are first identified at mid-gestation, and are prominent only on spongiotrophoblastic cells (Chen et al, 1994a). Therefore, it seems possible that IFN-~/ is the usual inducer of trophoblast cell M H C class I antigens, and that IFN-~ receptor development might indicate the point during gestation at which trophoblast cells will express these multifunctional antigens. IFN-'y induced higher expression of RT1 class I antigens on both of the trophoblast cell lines tested in this study. Hence, it seems likely that their normal counterparts reside in mid- to late-gestation placentae. This is known to be the case for the R8RP.3 cells, which were generated from day 11 placentae (Hunt et al, 1989a), Consistent with the idea that cells in different paths might be influenced differently by endogenous cytokines, Rcho-1 cells were not growth-inhibited by IFN-~/whereas dramatic inhibition was observed for the R8RP.3 cells. The cell lines used in these experiments, while not perfect model systems, showed clearly that phenotypically distinct trophoblast cells respond differently to various cytokines known to be produced in the uterus. Examination of placenta from transgenic mice with targeted disruptions in their cytokine genes might permit additional insights. However, transgenic mice deficient in TGF-13 (Shull et al, 1992) and IFN--/(Dalton et al, 1993) are fertile, and no evidence has been presented for aberrations in their placentae. Possibly, production of functionally similar growth factors such as other members of the TGF-13 family and Type I interferons obscure potentially important effects. TNF-deficient strains have not yet been reported. This might be due to a requirement for T N F during pregnancy, which has now been illustrated in two mouse models (Tartakovsky and Ben-Yair, 1991; De Kossodo et al, 1992).

ACKNOWLEDGEMENTS The authors appreciatethe excellenttechnicalassistance of R. Hickmanand W. Dalziel.

REFERENCES Aviv, H. & Leder, P. (1972) Purificationof biologicallyactive globin messengerRNA by chromatographyon oligothymidylicacid-cellulose.Proceedingsof the NationalAcademy of Science USA, 69, 1408-1412. Baglioni, C. (1992) Mechanisms of cytotoxicity,cytolysisand growth stimulation by TNF. In: TumorNecrosis Factors (Ed.) Beuder, B. pp. 425-438. New York:RavenPress. Chen, H,-L., Yang, Y., Hu, X.-L., Yelavarthi, K. K., Fishbaek, J. L. & Hunt, J. S.(1991) Tumor necrosis factor-alpha mRNA and protein are present in human placental and uterine cells at earlystages of gestation. AmericanJournal of Pathology, 139, 327-335. Chert, H.-L., Kamath, R., Pace, J. L., Russell, S. W. & Hunt, J. S. (1994a) Expressionof the interferon-~ receptorgene in mouseplacentas is related to stage of gestation and is restrictedto specificsubpopulationsof trophoblastcells. Placenta, 15, 109-121. Chen, H.-L., Yelavarthi, K. K., Hu, X.-L. &Hunt, J. S. (1994b)Identificationof transforminggrowth factor131mRNAin virginand pregnantrats by in situ hybridization.Journal ofReproductiveImmunology, 25, 221-233. Chornczynski, P. & Sacehi, N. (1987) Single step method RNA isolationby acid guanidinethiocynatephenol chloroformextraction.Analytical Biochemistry, 162, 156-159. Collins, T., LaPierre, L. A., Fiefs, W., Strominger, J. L. & Pober, J. S. (1986) Recombinanthuman tumor necrosis factorincreases in mRNAlevelsand surfaceexpressionof HLA-A,Bantigens in vascular endothelial cells and dermal fibroblastsin vitro. Proceedingsof the NationalAcademyof Sdence USA, 83, 446-450. Crainie, M., Guilbert, L. & Wegmann, T. G. 0990) Expressionof novel cytokinetranscripts in the murine placenta. Biologyof Reproduction,43, 999-1005.

Roby et ak CytokineRegulation ofMHC Class 1 Expressionin Rat Trophoblast Cells

589

Dalton, D. K., Pitts-Meek, S., Keshav, S., Figari, I. S., Bradley,. A. & Stewart, T. A. (1993) Multiple defects of immune cell function in mice with disrupted inteferon-~ genes. Science, 259, 1739-1742. David-Watine, B., Israel, A. & Kourilsky, P. (1990) The regulation and expression of MHC class I geneg. Immunology Today, 11,286-292. De Kossodo, S., Grau, G. E., Daneva, T., Pointaire, P., Fossati, L., Ody, C., Zapf, J., Piguet, P.-F., Gaillard, R. C. & Vassalli, P. (1992) Tumor necrosis factor c~is involved in mouse growth and lymphoid tissue development.Journal of ExperimentalMedicine, 176, 1259-1264. De Maeyer, E. & De Maeyer-Guignard, J. (1992) Interferon-'/. Current Opinions in Immunology, 4, 321-326. De, S. K., Chen, H.-L., Pace, J. L., Hunt, J. S., Terranova, P. F. & Enders, G. C. (1993) Expression of tumor necrosis factor alpha in mouse spermatogenic cells. Endocrinology, 133,389-396. Derynck, R., Jarrett, J. A., Chen, E. Y. & Goeddel, D. V.(1986) The murine transforming growth factor-J3 precursor. Journal of Biological Chemistry, 261, 4377-4379. Faria, T. N. & Soares, M. J.(1991) Trophoblast cell differentiation: establishment, characterization, and modulation of a rat trophoblast cell line expressing members of the placental prolactin family. Endocrinology, 129, 2895-2906. Fehimann, M., Peyron, J.-F., Samson, M., Van Obberghen, E., Brandenburg, D. & Brossette, N. (1985) Molecular association between major histocompatibility complex class I antigens and insulin receptors in mouse liver membranes. Proceedingsof the National Academy of Science USA, 82, 8634-8657. Fransen, L., Muller, R., Marmenout, A., Tavernier, J., Van der Heyden, J., Kawashima, E., Chollet, A., Tizard, R., Van Heuverswyn, H., Van Vliet, A., Ruyssehaert, M. R. & Fiers, W. (1985) Molecular cloning of mouse tumour necrosis factor cDNA and its eukaryotic expression. NucleicAcids Research, 13, 4417-4429. Galanos, C., Lunderitz, O. & Westphal, O. (1969) A new method for the extraction of R lipopolysaccharides. EuropeanJournal ofBiochemistry, 9, 245-249. Head, J. R., Drake, B. L. & Zuckermann, F. A. (1987) Major histocompatibility antigens on trophoblast and their regulation: Implications in the maternal-fetal ralationship. AmericanJournal of Reproductive Immunology, 15, 12-18. Hunt, J. S. & Soares, M.J. (1988) Expression of histocompatibility antigens, transferrin receptors, intermediate filaments, and alkaline phosphatase by in vitro cultured rat placental cells and rat placental cells in situ. Placenta, 9, 159-171. Hunt, J. S., Atherton, R. A. & Pace, J. L. (1990) Differential responses of rat trophoblast cells and embryonic fibroblasts to cytokines that regulate proliferation and class I MHC antigen expression. Journal oflmmunology, 145, 184-189. Hunt, J. S., Deb, S., Faria, T., Wheaton, D. & Soares, M. J. (1989a) Isolation of phenotypically distinct trophoblast cell lines from normal rat chorioallantoic placentae. Placenta, 10, 161-177. Hunt, J. S., Soares, M.J., Lei, M.-G., Smith, R. N., Wheaton, D., Atherton, R. & Morrison, D. C. (1989b) Products of lipopolysaccharide-activated macrophages (tumor necrosis factor-ix, transforming growth factor-J3) but not lipopolysaccharide modify DNA synthesis by rat trophoblast cells exhibiting the 80 kDa lipopolysaccharide-binding protein.Journal of Immunology, 143, 1606-1613. Hunt, J. S., Yang, Y., Wheaton, D., Kunz, H. W. & Gill, T. J. III. (1992) Analysis of the expression and immunostimulatory capacity of class I MHC antigens on rat trophoblast cells. AmericanJournal of Reproductive Immunology, 26, 129-135. Hunt, J. S., H.-L., Hu, X.-L. & Pollard,J. W. (1993) Normal distribution of tumor necrosis factor-tx messenger ribonucleic acid and protein in virgin and pregnant psteopetrotic (op/op) mice. Biology of Reproduction, 49, 441452. Jaffe, L., Robertson, E. J. & Bikoff, E. K. (1991) Distinct patterns of expression of MHC class I and 132microglobulin transcripts at early stages of mouse development. Journal of Immunology, 147, 2740-2750. Kanbour, A., Ho, H.-N., Misra, D. N., MacPherson, T. A., Kunz, H. W. & Gill, T.J. III. (1987) Differential expression of MHC class I antigens on the placenta of the rat: a mechanism for survival of the fetal allograft. Journal ofExperimentalMedicine, 166, 1861-1882. Keski-Oja, J. & Moses, H. L. (1987) Growth inhibitory polypeptides in the regulation of cell proliferation. Medical Biology, 65, 13-20. Leeuwenberg, J. F. M., van Damme, J., Jeunhomme, G. M. A. A. & Buurman, W. A. (1987) Interferon 131, an intermediate in the tumor necrosis factor a-induced increased MHC class I expression and an autocrine regulator of the constitutive MHC class I expression.Journal of ExperimentalMedicine, 166, 1180-1185. Lin, H., Mosmann, T. R., Guilbert, L., Tuntipopipat, S. & Wegmann, (1993) Synthesis of TH2-type cytokines at the maternal-fetal interface. Journal oflmmunology, 151, 4562-4573. Mattsson, R., Holmdahl, R., Scheynius, A., Bernadotte, F., Mattsson, A. & Van der Meide, P. H. (1991) Placental MHC class I antigen expression is induced in mice following in vivo treaunent with recombinant interferon-gamma.Journal of ReproductiveImmunology, 19, 115-129. Mattsson, R., Mattsson, A., Holmdahl, R., Scheynius, A. & Van der Meide, P. H. (1992) In vivo treatment with interferon-gamma during early pregnancy in mice induces strong expression of major histocompatibility complex class I and II molecules in uterus and decidua but not in extra-embryonic tissues. Biology of Reproduction, 46, 1176-1186. McGee, D. W., Aicher, W. K., Eldridge, J. H., Peppard, J. V., Mestecky, J. & McGhee, J. R. (1991)

590

Placenta (1994), Vol. 15

Transforming growth factor-[3 enhances secretory component and major histocompatibility complex class I antigen expression on rat IEC-6 intestinal epithelial cells. Cytokine, 3, 543-550. Mosmann, T. (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays.Journal oflmmunologicalMethods, 65, 55-63. Murray, L. & Martens, C. (1989) The abnormal T-lymphocytes in lpr mice transcribe interferon-gamma and tumor necrosis factor-c~genes spontaneously in vivo. EuropeanJournal oflmmunology, 19, 563-565. Myers, M. J., Pullen, J. K., Ghildval, N., Eustis-Turf, E. & Sehook, L. B. (1989) Regulation of I1-1 and TNF~ expression during the differentiation of bone marrow derived macrophages. Journal oflmmunology, 142, 153-160. Okamoto, T., Watanabe, N., Yamauchi, N., Tsuji, Y., Tsuji, N., Itoh, Y., Neda, H. & Niitsu, Y. (1992) Endogenous tumor necrosis factor exerts its protective function intraceUularly against the cytotoxicity of exogenous tumor necrosis factor. CancerResearch, 52, 5278-5281. Pampfer, S., Wuu, Y.-D., Vanderheyden, I. & De Hertogh, R. (1994) Expression of tumor necrosis factor-c, (TNF-c~) receptors and selective effect of TNF-c~ on the inner cell mass in mouse blastocysts. Endocrinology, 134, 206-212. Pfizenmaier, K., Himmler, A., Sehutze, S., Scheurich, P. & Kronke, M. (1992) TNF receptors and TNF signal transduction. In TumorNecrosisFactors(Ed.) Meutler, B. pp. 439-472. New York: Raven Press. Roby, K. F., Deb, S., Gibori, G., Szpirer, C., Levan, G., Kwok, S. C. M. & Soares, M.J. (1993) Decidual prolactin related protein identification, molecular cloning, and characterization. Journal of Biological Chemistry, 268, 3136-3142. Rosa, F. & Fellous, M. (1984) The effect of gamma-interferon on MHC antigens. Immunology Today, 5, 261262. Schreiber, A. B., Schlessinger, J. & Edidin, M. (1984) Interaction between major histocompatibility complex antigens and epidermal growth factor receptors on human cells.Journal of CellBiology, 98, 725-731. ShuU, M. M., Ormsby, I., Kier, A. B., Pawiowski, S., Diebold, R. J., Yin, M., Alien, R., Sidman, C., Proetzel, G., Calvin, D., Annunziata, N. & Doetschman, T. (1991) Targeted disruption of the mouse transforming growth factor-[3 gene results in multifocal inflammatory disease. Nature, 359, 693-699. Singer, D. S. & Maguire, J. E. (1990) Regulation of the expression of class I MHC genes. CriticalReviews in Immunology, 10, 235-257. Soares, M.J., Schaberg, K. D., Pinal, C. S., De, S. K., Bhatia, P. & Andrews, G. K. (1987) Establishment of a rat placental cell line expressing characteristics of extraembryonic membranes. DevelopmentalBiology, 124, 134-144. Soares, M. J., Faria, T. N., Hamlin, G. P., Lu, X.-J. & Deb, S. (1993) Trophoblast cell differentiation: expression of the placental prolactin family. In Trophoblast Cells: Pathwaysfor Maternal-Embryonic Communication (Eds.) Soares, M.J., Handwerger, S. & Talamantes, F. pp. 43-65. New York: Springer-Verlag. Tamada, H., MeMaster, M. T., Flanders, K. C., Andrews, G. K. & Dey, S. K. (1990) Cell type-specific expression of TGF-[31 in the mouse uterus during the periimplantation period. MolecularEndocrinology,4, 965972. Tartakovsky, B. & Ben-Yair, E. (1991) Cytokines modulate pre-implantation development and pregnancy. DevelopmentalBiology, 146, 245-352. Thomas, P. S. (1980) Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proceedingsof the NationalAcademy of Science USA, 77, 5201-5205. Vanhaesebroeek, B., Decoster, E., Van Ostade, X., Van Bladel, S., Lenaerts, A. Van Roy, F. & Fiers, W. (1992) Expression of an exogenous tumor necrosis factor (TNF) gene in TNF-sensitive cell lines confers resistance to TNF~ cell lysis.Journal oflmmunology, 148, 2785-2794. Witsell, A. L. & Sehook, L. B. (1992) Tumor necrosis factor-c~ is an autocrine growth regulator during macrophage differentiation. Proceedingsof the NationalAcademy of Science USA, 89, 4754-4758. Yang, Y., Yelavarthi, K. K., Chen, H.-L., Pacel, J. L., Terranove,-P. F. & Hunt, J. S. (1993) Molecular, biochemical and functional characteristics of tumor necrosis factor-c~ produced by human placental cytotrophoblastic cells.Journal of lmmunology, 150, 5614-5624. Yelavarthi, K. K. &Hunt, J. S. (1993) Analysis of p60 and p80 tumor necrosis factor-c~ mRNA and protein in human placentae. AmericanJournal of Pathology, 143, 1131- 1141. Yelavarthi, K. K., Chen, H.-L., Yang, Y., Fishback, J. L., Cowley, B. D. Jr. & Hunt, J. S. (1991) Tumor necrosis factor-alpha mRNA and protein in rat uterine and placental cells. Journal oflmmunology, 146, 38403848.