possible mating combinations of the DA (A'E - ) and WF (AUEu) inbred strains and various matings of the R16 (A'E-), R33 (A'EU), and R34 (A'E - ) congenic ...
BIOLOGY OF REPRODUCTION 48, 977-986 (1993)
Differential Genomic Imprinting of Major Histocompatibility Complex Class I Antigens in the Placenta of the Rat' AMAL KANBOUR-SHAKIR, 3'4 HEINZ W. KUNZ, 3 and THOMAS J. GILL III2 '3 Department of Pathology,3 University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Magee Womans Hospital,4 Pittsburgh, Pennsylvania 15261 ABSTRACT The RTI.A and Pa (class I) loci in the major histocompatibility complex (MHC) of the rat are genomically imprinted in the placenta, since only the paternally derived antigens are expressed. To determine whether the RTI.E (class I) locus, which defines the end of the MHC opposite RT.A, is genomically imprinted in the placenta, two different types of crosses were studied: all possible mating combinations of the DA (A'E - ) and WF (AUEu) inbred strains and various matings of the R16 (A'E-), R33 (A'EU), and R34 (A'E - ) congenic recombinant strains. The results showed that the Eu locus was not imprinted. Thus, genomic imprinting can be different at two endogenous loci that are part of a multigene family in tight-linkage disequilibrium. The lack of E" imprinting is not due to the locus' being monomorphic, since the monomorphic Pa locus is imprinted, and it is most probably not due to its distance from the imprinted RT1.A and Pa loci (0.4 cM). We suggest that the control of the imprinting may lie in the physical structure of the DNA proximate to the imprinted locus.
INTRODUCTION The major histocompatibility complex (MHC) encodes antigens that are important in determining tissue compatibility as well as various components of the immune response system. Linked to the MHC is a region having genes that are involved in reproduction and development [1] and genes encoding MHC class I antigens that are organ-specific and appear at different times during development or during oncogenesis [2]. This region is well defined in the mouse (Q/TL) [3] and in the rat (grc) [4-6], and there is evidence that a similar region exists in humans [7-10] and in other species [11-13]. The immunological and developmental roles of this highly conserved region of the genome are closely intertwined in the evolution of viviparity and of placentation. In studies on the expression of MHC antigens on the placenta, we have utilized matings of the DA and WF strains of rats and examined the placentas by immunohistochemistry and by electron microscopic immunocytochemistry using a variety of specific monoclonal antibodies [14, 15]. We have shown that the class I RT1.A and Pa antigens are differentially expressed on the plasma membrane of the basal trophoblast cells. RT1.A is expressed in syngeneic matings but not in allogeneic matings, even though it is present in the cytoplasm of the basal trophoblast cells in both types of matings. By contrast, Pa is expressed on the surface of the basal trophoblast cells in both allogeneic and syngeneic matings. None of the class I antigens is present in the labyrinthine trophoblast, and the class II antigens are not expressed at all in the placenta. Using reciprocal natural Accepted December 18, 1992. Received September 24, 1992. 'This work was supported by grants from the National Institutes of Health (HD 08662 and HD 09880) and by the Pathology Education and Research Foundation. 2 Correspondence. FAX: (412) 648-1916.
977
matings and the appropriate embryo transfers, we have demonstrated that the expression of the RT1.A and Pa antigens in the placenta is genomically imprinted and that only the paternal antigen is expressed. The present study was undertaken to extend our previous work on the genomic imprinting of the RT1.A and Pa loci [15] by examining imprinting at the RT1.E locus, which defines the opposite limit of the MHC. Two separate sets of crosses were used: matings of the DA (AaE-) and WF (AUEu) inbred strains, which were used in our previous studies, and matings of the R16 (AaE-), R33 (AaEU), and R34 (AaE-) congenic recombinant strains, in which the Eu locus segregates in the absence of an Au locus. MATERIALS AND METHODS Animals All of the rats used were from our colony at the University of Pittsburgh School of Medicine (Table 1). They were housed in a room controlled for temperature and humidity under a 12L: 12D cycle, and they were fed laboratory chow and water ad libitum. One female was placed in a cage with two males, and Day 1 of pregnancy was ascertained by the appearance of a copulatory plug. The fetuses were studied at Day 17 of gestation. The DA and WF strains were inbred by brother x sister matings in excess of 40 generations. The R16 strain (Table 1) arose from a recombination in a mating between a heterozygous animal from the BIL intercross (RT1.ABDEgrc: l/ n, Il/a, Il/a, -/u, -/+) and an ACP animal (a,a,a,-, +), both of which were inbred in excess of 40 generations. Since the R16 male would be homozygous grc- and therefore sterile [16], the R16 strain was propagated by mating a heterozygous male (a/n, a/a,a/a, -/u, -/+) with a homozygous female (a,a, a, -, -). The heterozygous male, which
978
KANBOUR-SHAKIR ET AL.
TABLE 1. Strains used in this study. RT1 specificity Haplotype
A
B
D
E
grc
DA
a
a
a
a
-
+
WF
u
u
u
u
u
+
R16
r16
a
a
a
-
-
r16/i r33
a/n a
a/a a
a/a a
-/u u
-/+ +
r34
a
a
a
-
+
Strain
R16(a/n)* R33 R34
*Heterozygous segregant from the cross used to propagate the R16 strain.
is a segregant from the cross used to propagate the R16 strain, is designated R16 (a/n). The tR33 and R34 strains are recombinants that arose during the propagation of the R16 strain. The recombination in the R33 strain involved the Egrc region, and the recombination in the R34 strain involved only the grc region. Antibodies The development and characterization of the monoclonal antibodies (mAbs) have been described [17]: anti-Eu (mAb 70) is IgG2b; anti-Au (mAb 68), IgG1; and anti-An (mAb 42), IgG2a. The Eu antigen is defined by its reactivity with mAb 70, and the Au antigen by its reactivity with mAb 68. However, the Au antigen also carries the epitope that reacts with mAb 70: hence, mAb 70 will react with R16 and R33 in a unique manner, but its reactivity with Au in the WF strain and its crosses must be considered. The E antigen is monomorphic; i.e., it has only one serologically active epitope that has only one serological specificity. The allele of the Eu gene is null (E-), so it does not encode a serologically active antigen. Antibody Detection The presence of a maternal antibody response to the paternal component of the fetal placental antigens was evaluated in the sera of pregnant and immediately postpartum rats by cytofluorimetry through use of a Coulter Epics Profile instrument [15]. The fluorescein-labeled F(ab')2 fraction of rabbit anti-rat IgG or IgM was used as the second antibody, and the binding data were analyzed by means of an IBM PS-70 computer. The sera were also checked by virtue of lymphocyte cytotoxicity and by hemagglutination [14]. Cells from the R16 (AaE-), WF (A"Eu), and DA (AaE - ) strains were used for the assays. Immunohistochemistry: Light Microscopy The basic method has been described in detail [14], but there were some critical modifications. The frozen sections of placental tissues were cut at 10 p.m, because this thickness gave the best preservation of the tissue architecture and provided an amount of tissue sufficient to stain distinctly with the mAb. The detection system used was a modified streptavidin-biotin method (Zymed Laboratories, South
San Francisco, CA) using biotinylated mAb. Following treatment to block endogenous peroxidases and to eliminate nonspecific binding, the tissues were incubated with the biotinylated mAb for 1 h at room temperature. After washing in PBS, tissues were incubated with the steptavidin-peroxidase conjugate for 15 min at room temperature, washed three times with PBS, and incubated with the 3-amino-9ethylcarbazole (AEC)-hydrogen peroxide mixture for 15 min at room temperature. Finally, the tissues were counterstained with Mayer's hematoxylin and then placed under a cover slip with crystal mount. Immunocytochemistry: Electron Microscopy The protein A-gold technique with 15-nm gold particles was used, as described previously [14], and an average of twenty-five electron photomicrographs were examined for each measurement. The number of gold particles per micrometer of cell membrane or per square micrometer of cytoplasm was measured using computerized morphometry with a Bioquant II instrument attached to an Apple II microcomputer (R & M Biometrics, Nashville, TN). Blocking studies were performed to demonstrate that the antibodies to Au (mAb 68) and to Eu (mAb 70) detected separate molecules in the tissues. Unlabeled anti-Au was applied to the tissues in 10-fold excess of the amount of labeled antibody that was to be used later for detection, incubated for 2 h at 37°C, and then rinsed with three aliquots of PBS. Immediately thereafter, the tissues were stained with the labeled anti-Au or anti-Eu mAb. An average of twentyfive electron photomicrographs were examined in each component of the experiment, and the entire blocking experiment was performed twice, with the same results each time. Statistics Comparisons between means were made using the t-test, and p < 0.01 was taken as the level of significance. RESULTS Electron microscopic immunocytochemical staining for the Eu and Au antigens in all four mating combinations of the DA and WF strains (Fig. 1) showed that the Eu antigen of both maternal (Fig. 1A) and paternal (Fig. IC) origin was expressed on the basal trophoblast cell membrane and in the cytoplasm. Each cell stained with the labeled antibody. The syngeneic WF x WF placenta (Fig. 1, E and F) expressed both Eu and Au antigens (positive control), but the DA x DA placenta (Fig. 1, G and H) expressed neither (negative control). As expected, the maternal Au antigen was not expressed (Fig. 1B), and the paternal Au antigen was present in the cytoplasm but not on the cell membrane (inducibly suppressed; Fig. 1D). These observations were quantified by morphometric analysis (Table 2).
o
c
w
E '
