BIOLOGY OF REPRODUCTION 54, 219-223 (1996)
Expression of Major Histocompatibility Complex Antigen and Timing of Invasion
by Equine Chorionic Girdle Cells Cultured on Matrigel'
Karen E. Vagnoni,3
. J
. Ginther,4 and D.P. Lunn2'3
Department of Medical Science, School of Veterinary Medicine,3 andDepartment of Animal Health
and Biomedical Science,4 University of Wisconsin-Madison, Madison, Wisconsin 53 706
ABSTRACT
Chorionic girdle cells are a highly invasive subpopulation of trophoblastic cells of the horse conceptus that adhere to the uterine
epithelium and begin to invade the endometrium on Days 34-36 (Day 0 = day of ovulation). Just prior to and during invasion (Days 3236), chorionic girdle cells express high levels of major histocompatibility complex (MHC) I, but expression of this antigen decreases by
Days 40-45 and is lost by Day 55. The mechanisms involved in the control of chorionic girdle cell invasion and altered MHCI expression
over time are not known. The objective of this study, therefore, was to determine the timing of invasion and the characteristics of MHC
expression by girdle cells cultured on Matrigel to determine whether chorionic girdle cell behavior in this environment is similar to the
behavior in vivo. Chorionic girdles from four conceptuses were collected on each of Days 30, 31, and 32 and placed in Matrigel invasion
chambers for 48 h and in additional duplicate chambers for atime period equivalent to the number of days from the day of collection to
Day 36 (6,5, and 4 days, respectively). After these culture times, the area of the filter covered by invasive cells was determined through
use of the software program NIH image. At 48 h, Day 31 and Day 32 girdle cell preparations had invaded Matrigel; but only one Day 30
girdle cell preparation showed invasion into Matrigel, and the extent of invasion was limited. Girdle cells collected on all 3 days had
invaded Matrigel by the time equivalent to Day 36. Additionally, chorionic girdles from six conceptuses were collected on Day 34 and
placed in Matrigel invasion chambers. Three of these were examined for surface antigen expression of MHCI, MHCII, and a trophoblastspecific antigen (102.1) by immunocytochemistry (ICC) every 48 h to atime point equivalentto Day 44. The remaining three were examined
for these same antigens at times equivalent to Days 36, 40, 50, and 60. Invasive Day 34 girdle cells expressed MHCI and trophoblastspecific antigen, but not MHCII, at each time examined (Days 36-60). Neither the timing of invasion nor MHCI expression by chorionic
girdle cells on Matrigel mimicked what occurs in vivo. Therefore, it is likely that the in utero environment plays a role in regulating these
two characteristics of equine chorionic girdle cells.
INTRODUCTION
called endometrial cups, and the active chorionic girdlederived cells (cup cells) produce eCG from approximately
Days 40 to 150 [2].
Mares have a gestation length of 315-360 days [2]. However, chorionic girdle cell invasion of the uterine epithelium
and stroma occurs consistently between Days 34 and 36 and
therefore is well regulated temporally. The expression of
major histocompatibility complex (MHC) I by chorionic girdle cells also appears to be developmentally regulated [58] in that just before and during the time of invasion (Days
32-36), the chorionic girdle cells express high levels of
MHCI. By Days 40-45, MHCI expression decreases, and by
Days 55-65 MHCI is absent on the differentiated chorionic
girdle cells that make up most of the endometrial cups [8].
Noninvasive trophoblastic cells generally do not express
MHCI at any time [8].
The purposes of this study were 1) to determine whether
or not chorionic girdle cells younger than Day 34 could
invade extracellular matrix and 2) to examine the expression of MHCI by cells that invade the extracellular matrix.
In vitro invasion of extracellular matrix at a time earlier than
that observed in vivo (Day 35), and altered timing of the
loss of MHCI expression in vitro compared to in vivo (Days
40-45), would suggest that the in utero environment has a
role in regulating chorionic girdle cell invasion and MHC
expression.
Placentation in the horse is unique among domestic
species in that between Days 28 and 36 (Day 0 = day of
ovulation), two morphologically distinct populations of trophoblastic cells can be identified easily. Most of the trophoblastic cells of the equine conceptus are noninvasive, as is
typical of epithelial-chorial placentation [1]. However, a discrete area (chorionic girdle) of highly invasive trophoblastic
cells forms and differentiates between Days 25 and 36. The
chorionic girdle is a narrow (9 mm) band of specialized,
avascular tissue encircling the spherical conceptus [2]. Trophoblasts of the chorionic girdle adhere to the uterine epithelium by Day 35 and begin to invade the endometrial
wall [3, 4]. Within a 48-h period, chorionic girdle cells aggressively migrate though the uterine epithelium into the
endometrial stroma. At this point, the chorionic girdle cells
are no longer attached to the placenta but instead form part
of distinct and highly differentiated tissue nodules that are
buried within the endometrial stroma. These nodules are
Accepted September 6, 1995.
Received July 31, 1995.
'Research was supported by the American Quarter Horse Association and by the College
of Agricultural and Life Sciences and the Graduate School of the University of WisconsinMadison.
2
Correspondence: D. Paul Lunn, 2015 Linden Drive West, School of Veterinary Medicine, Madison, WI 53706. FAX: (608) 265-8020.
219
220
VAGNONI ET AL.
MATERIALS AND METHODS
Girdle Cell Collection
Mares were bred, and the day of ovulation was established by daily ultrasonography. On Day 30, 31, 32, or 34
(Day 0 = day of ovulation), the conceptus was flushed from
the uterus. This was achieved by manual dilation of the
cervix and infusion of 200-500 ml of sterile saline into the
uterus to create pressure such that the mare strained and
expelled the conceptus. The conceptus was caught in a sterile beaker. The chorionic girdle was dissected away from
the conceptus with sterile scissors and forceps and was then
placed in warm RPMI, 10% fetal calf serum (Sigma Chemical
Co., St. Louis, MO), 5% penicillin-streptomycin, and Hepes
(Gibco BRL, Grand Island, NY) and was transported to the
laboratory (approximately 30 min). The chorionic girdle
was washed three times by gently removing and replacing
RPMI, 10% fetal calf serum, 1% penicillin-streptomycin, and
Hepes. The chorionic girdle was finely minced, and equal
volumes of the media and girdle tissue mixture were placed
within Matrigel invasion chambers (Collaborative Biochemical, Bedford, MA), which had been prepared according to
the manufacturer's instructions.
Determinationof the Timing qf Chorionic Girdle Cell
Invasion Through Matrigel
The ability of chorionic girdle cells collected on Days 3032 to invade through Matrigel and the time needed for invasion were examined through use of Matrigel invasion
chambers. Matrigel is a soluble extracellular matrix produced by Englebreth-Holmswarm tumor cells that contains
type IV collagen, fibronectins, laminin, and elastin and is
commonly used to examine invasive capabilities of cells [9].
In a previous study [10], it was established that equine girdle
cells collected from Day 34 conceptuses readily invade Matrigel; therefore, girdle cell preparations from four different
mares collected on each of Days 30, 31, and 32 were cultured separately in invasion chambers in which 8-gm porous filters were coated with Matrigel. For each girdle cell
preparation, equal volumes of the media and girdle cell
mixture were placed in duplicate chambers to be examined
after 48 h. Additional duplicate chambers were examined
at times equivalent to the number of days from the day of
collection to Day 36 (6, 5, and 4 days for cells collected on
Days 30, 31, and 32, respectively). In addition, Day 30 girdle
cell preparations were placed in additional wells to be harvested 4 days later (Day 34). To quantify invasive cells after
these incubation times, the media and noninvasive cells
were aspirated from the upper chamber, and the upper side
of the filter was swabbed and washed three times to remove
Matrigel and any remaining noninvasive cells from the upper chamber. The filter was then stained with Diff-Quick
(Baxter Scientific Products, McGraw Park, IL), allowed to
dry, and fixed on a glass slide. Images of the underside of
the filters were captured by a video camera attached to a
microscope and directly imported into a software program
(NIH image, version 1.55F). The total area of the filter covered by girdle cells was determined. One-way ANOVAs
were conducted on the areas equivalent to 48 h and Day
36, where day of embryo collection was the factor of interest. Because treatment (daily) means were proportional to
within-treatment variances, ANOVA was conducted on
square root-transformed data [11]. Means were separated by
least significant differences if the F-test for day indicated
significance (p < 0.05). In addition, for the Day 30 observations, data were transformed as above and a one-way
ANOVA was done for the areas equivalent to Days 32, 34,
and 36, where the length of culture was the factor of interest. Means were separated by least significant differences if
the F-test for day indicated significance (p < 0.05).
Because older embryos are larger, results could have
been biased in that more cells were potentially plated per
well in preparations from older embryos. To compensate
for this, for each observation, the area of the filter covered
at 48 h was divided by the area of the filter covered at the
time equivalent to Day 36 for that preparation. This was
done because it is likely that the area of the filter covered
by cells at the time equivalent to Day 36 reflected the initial
number of cells allotted to each chamber for each preparation. Because these data represented a wide range of percentages, they were transformed through use of the inverse
sine fraction before ANOVA was conducted [11]. The means
were separated as described above.
Determination of the Expression of MHC Antigen and
Trophoblast-Specific Antigen Using Immunocytochemistry
(ICC)
Three Day 34 conceptuses were collected, and each chorionic girdle cell preparation was placed in five Matrigel
invasion chambers as described above. One filter was harvested every 48 h through a time point equivalent to Day
44, and at the time of harvest the cells and Matrigel from
the upper chamber were removed as previously described.
The filter, while still in its housing, was placed in liquid
nitrogen and stored at -70°C. Cells that invaded through
Matrigel, as indicated by their presence on the underside of
the frozen filter, were examined by ICC for expression of
MHCI, MHCII, or a trophoblast-specific antigen that is recognized by the antibody 102.1 [121. Just before ICC evaluation, filters in their housing were placed in acetone for 5
min and then allowed to dry. The filter was cut away from
its housing and then cut into four pieces, and each piece
was fixed on a glass slide with Polymount (Baxter). Through
use of these four pieces, the cells on each filter for each day
were analyzed for expression of MHCI, MHCII, and trophoblast-specific antigen; an isotype control antibody was also
included in the analysis. To prepare sections for ICC analysis, a Pap pen (The Binding Site, San Diego, CA) was used
221
GIRDLE CELL INVASION AND MHC EXPRESSION ON MATRIGEL
to circle filter sections, and each section was incubated in
PBS. To determine expression of MHCII, the monoclonal
antibody CVS10 [131 was used as a pure supernatant that
was incubated with the filter section for 2 h. The section
was then washed in PBS and incubated for 1 h with Affinipure goat anti-mouse peroxidase (Jackson Immunoresearch, Westgrove, PA; 1:300 in PBS). A 3-amino-9-ethylcarbazole horseradish peroxidase kit (Vector Laboratories,
Burlingame, CA) was used to visualize positive-staining
cells, and cells were counterstained with hematoxylin (Baxter). Staining of cells by the remaining three antibodies was
determined by means of the Vectastain Elite ABC kit (Vector). For this method, cells present on filter sections were
incubated with monoclonal antibodies as a pure supernatant that recognized the following antigens (specific antibody in parentheses): MHCI (Mac 291), equine trophoblast
antigen (102.1), and an isotype control antibody [8, 12, 14].
After a 20-min exposure to normal serum, the primary antibody was incubated with sections for 1 h. Sections were
then washed and incubated in biotinylated anti-murine immunoglobulin antibody for 30 min and in avidin-biotin
complex for 30 min with two washes in PBS between these
steps. The 3-amino-9-ethylcarbazole horseradish peroxidase kit was used to detect positive-staining cells, and cells
were counterstained with hematoxylin.
An additional three Day 34 conceptuses were collected,
and each chorionic girdle cell preparation was placed in
four Matrigel invasion chambers as described above. One
filter each was harvested at a time equivalent to the number
of days from the day of collection to Days 36, 40, 50, and
60, that is at 2, 6, 16, and 26 days, respectively. Filters were
either frozen as described above or used immediately after
harvest for ICC. All filters were analyzed by ICC for MHCI,
MHCII, trophoblast-specific antigen, and an isotype control
antibody with the Vectastain Elite ABC kit as described
above.
Every cell on each filter section in which the nucleus was
identified by hematoxylin staining was examined to determine expression of antigens. Included in all of the ICC procedures were additional sections of control tissues from
equine lymph nodes (MHCI and MHCII positive, 102.1 negative) and endometrial cups (MHCI and MHCII negative,
102.1 positive).
TABLE 1. Invasion of Matrigel by Day 30, 31, and 32 girdle cells.
Productionof eCG by Girdle Cell Preparations
The supernatants from the lower wells (media exposed
to invasive cells) of Day 30, 31, and 32 girdle cell preparations-collected at times equivalent to 48 h and Day 36were stored at - 20°C and later analyzed for eCG (the primary secretory product of cup cells) by means of an ELISA
(Diosynth, Oss, The Netherlands). The relationship between the concentration of eCG in the supernatant and the
area of the filter covered by cells was explored through
regression analysis. In addition, the supernatant from the
Determinationof the Expression of MHC and TrophoblastSpecific Antigen Using ICC
Day
area a
area 2b
arcsin (area 1/area 2)c
30
31
32
0.02 Y
0.77XY
1.90
0.37
1.10
0.71
2.65
0.52
0.02 ¥
0.42Xy
0.79X
0.13
SEM
aMean of the square root of area (mm2) after 48 h of incubation (n = 4).
bMean of the square root of area (mm2) after incubation to an equivalent of Day 36
(n = 4).
c Inverse sine fraction of the square root of area 1 (mm2) divided by the square root
of area 2 (mm2) for each observation (n = 4).
XY
Means within columns with different superscripts differ (p < .01).
lower wells of the Day 34 chorionic girdle cell preparations
cultured through Day 44 was collected, stored, and analyzed for eCG.
RESULTS
Determinationof the Timing of Chorionic Girdle Cell
Invasion Through Matrigel
At 48 h, the square root of the mean filter area covered
by invasive girdle cells collected on Day 31 was 0.77 mm 2 ;
for cells collected on Day 32 it was 1.90 mm2 . However, at
48 h, only one Day 30 girdle cell preparation showed invasion (0.02 mm2 , Table 1). By the time equivalent to Day
36, all girdle cell preparations showed invasion, with square
root mean filter areas of 1.10 mm 2 (Day 30), 0.71 mm 2 (Day
31), and 2.65 mm 2 (Day 32); there was no effect due to day
(p > 0.05). Analysis of the areas from the Day 30 girdle cell
preparations examined at 48 h (Day 32), 4 days (Day 34),
and 6 days (Day 36) showed significance (p < 0.005). The
Day 36 area (1.10 mm2 ) was greater (p < 0.05) than either
the Day 34 area (0.10 mm2 ) or the Day 32 area (0.02 mm 2 ),
which were not different from each other.
Similar results were noted when the 48-h data were adjusted to account for differing initial cell numbers and then
analyzed; the invasive capacity (as determined by filter
area) of Day 30 girdle cells was less than that of cells collected from Day 32 tissue (p < 0.05; Table 1). The invasive
capacity of cells collected from Day 31 tissues was not different from the invasive capacity for either Day 30 or Day
32 tissues (p = 0.22 andp = 0.14, respectively).
Invasive cells were binucleate with prominent nucleoli
and were often highly vacuolated, consistent with cup cell
morphology [4]. All cells stained positive for MHCI and trophoblast-specific antigen for each time examined, that is,
culture times equivalent to the time of collection to Days
36-60. For both MHCI and trophoblast-specific antigen,
staining was granular in appearance. Staining for the trophoblast-specific antigen did not appear to weaken over
222
VAGNONI ET AL.
ts -
.
I
6-
"'
I
IO
a)
4-
2-
0
I
35
40
45
Day
FIG. 1. The concentration of eCG (ig/ml) produced by Day 34 chorionic girdle cells
cultured inMatrigel invasion chambers to times equivalent to Days 36-44. Shown
are the mean and SEM from three preparations.
time. However, staining for MHCI for two of the three chorionic girdle preparations appeared weaker on Day 60 compared to Day 50. Positive staining for MHCII was not seen
for invasive girdle cells.
Production of eCG by Girdle Cell Preparations
The concentration of eCG in the supernatant of Day 3032 chorionic girdle cell preparations after culture for 48 h,
or to the equivalent of Day 36, was positively correlated
with the area of the filter covered by invasive girdle cells
(r2 = 0.88). Equine CG was not detected in supernatant
from cultures where invasion through Matrigel did not occur (the Day 30 preparations and some Day 31 preparations
after 48 h). By 48 h of culture and through the time equivalent to Day 44, eCG was detected in the supernatant of
Day 34 chorionic girdle cell preparations cultured on Matrigel. The mean eCG accumulative concentrations (tg/ml)
for the various days were 2.7 (Day 36), 5.0 (Day 38), 4.5
(Day 40), 5.7 (Day 42), and 6.0 (Day 44) (Fig. 1).
DISCUSSION
An invasive capacity within 48 h was demonstrated by
girdle cells collected on Days 31 and 32 but not for cells
collected on Day 30. Although Day 30 girdle cells did not
invade initially, they did become invasive by a time equivalent to Day 36, and by this time the extent of invasion was
similar (no significant differences) for all 3 days. These data
suggest that as early as Day 30, girdle cells possess the genetic information necessary for invasion and do not after
this time require additional information for invasion. The
delay in invasion by Day 30 girdle cells, compared to Day
31 and 32 cells, may be explained by the difference in maturity of these cells or by a difference in the in utero environment on Day 30 compared to Days 31 and 32 of gestation.
The conclusion of the present study is that the environment in utero has a role in the regulation of invasion by
equine chorionic girdle cells. This conclusion is consistent
with observations in other species showing that degradation
of the extracellular matrix is a tightly controlled process [151.
A role for the uterine influence on inhibition of placental
invasion in the present study is supported by a series of
experiments in the mouse that tested in vivo invasiveness
of trophoblast in a variety of sites; the extent and duration
of invasion by trophoblastic cells were greater in extrauterine sites and in nonpregnant uteri compared to pregnant
uteri [16]. These results suggest that the pregnant uterus may
control invasion in an inhibitory manner. Circulating hormones are not believed to play a role in invasion regulation,
because invasiveness of trophoblast in an extrauterine site
remained uninfluenced by pregnancy [16].
Results of a previous Matrigel study indicated that chorionic girdle cells rely on matrix metalloproteinase activity
for invasion into extracellular matrix [10]. Factors implicated
in regulating matrix metalloproteinase activity include tissue
inhibitors of metalloproteinase, transforming growth factor
S3,
and hCGs [17-21]. In the human and mouse, the decidua
expresses high levels of tissue inhibitors of metalloproteinases in vivo, and this limits trophoblast invasion of the uterine wall [9, 171. Similar observations have been reported in
vitro, where exogenous tissue inhibitor of metalloproteinase completely inhibits human cytotrophoblast invasion
[18]. Likewise, decidua-derived transforming growth factor
3 plays an inhibitory role in the in vivo control of human
trophoblast invasion [19]. Expression of tissue inhibitors of
metalloproteinase in the equine uterus has not been described; however, transforming growth factor-3S mRNA increases in the equine endometrium around the time of chorionic girdle cell invasion [22], suggesting that this factor
may regulate chorionic girdle cell invasion.
Human CG regulates human trophoblast invasion in vitro
[211. However, the effect of eCG on chorionic girdle cell invasion is unknown. Two separate reports [23, 24] have established that initiation and transcription of the genes for eCG
begin before, or at least by, Days 30-33, a time prior to chorionic girdle cell invasion in vivo and before eCG can be detected in the serum or in supernatants from cultured girdle
cells [231. The production of eCG in vivo, therefore, appears
to be developmentally regulated and possibly to be delayed
in relation to the timing of mRNA expression for this hormone.
An apparent delay in cellular function described for eCG pro-
GIRDLE CELL INVASION AND MHC EXPRESSION ON MATRIGEL
duction is similar to that described in the present study, which
is that chorionic girdle cells have the ability to invade extracellular matrix at least by Day 31 whereas they do not invade
the endometrium until days later.
Expression of MHCI by chorionic girdle cells that invaded Matrigel, as determined by ICC staining with MAC
291, was observed on all cells between the times equivalent
to Days 36-60, and changes in intensity of staining for this
antigen appeared to decrease from Day 50 to 60. Kydd et
al. [81, using the same antibody (Mac 291) employed in the
present study, showed that in vivo, chorionic girdle tissue
stained strongly for MHCI on Days 33 and 34. However, in
vivo expression of MHCI was down-regulated, but still
weakly expressed, with the differentiation of these girdle
cells into endometrial cup cells between Days 40 and 45, so
that by Days 55-65 no MHCI was detected [8]. Because
Matrigel provides an extracellular matrix and therefore
more closely mimics an in vivo environment, it was expected that MHCI expression in this system would mimic in
vivo observations. However, MHCI expression by these
cells cultured in this environment did not appear to reflect
what occurs in vivo, that is, a down-regulation of MHCI
expression between Days 40 and 45.
Expression of MHCI was not lost when chorionic girdle
cells were cultured on Matrigel through Day 60. This observation contrasts with findings by Donaldson et al. [7],
who reported that MHCI expression was absent by Day 51
on chorionic girdle cells cultured on plastic. Differences between these studies may be explained by the different in
vitro environment (Matrigel vs. plastic) or the different antibodies used to detect MHCI expression by cultured cells,
as Donaldson et al. [71 did not use Mac 291 to detect MHCI
expression by cultured chorionic girdle cells.
The morphology of cells that invaded through Matrigel
was similar regardless of the age at which they were collected. The cells were binucleate and highly vacuolated,
consistent with the morphology of cup cells in vivo [3]. Invasive chorionic girdle cells from Days 30-32 and Day 34
placed in Matrigel invasion chambers produced eCG, and
the area of the filter covered by invasive cells was a good
indication of the amount of eCG produced. This result is
consistent with the observation that chorionic girdle cells
produce eCG in tissue culture systems [4].
In summary, Day 34 chorionic girdle cells that invade
Matrigel produce eCG and express MHCI and trophoblast
antigen, but not MHCII, through at least Day 60, whereas in
vivo studies have indicated a reduction in MHCI on Days
40-45. Chorionic girdle cells collected on Days 31 and 32
invade extracellular matrix on or before Day 34, suggesting
that invasion can occur earlier in vitro than in vivo (Day 35).
These observations indicate in utero regulation of MHC expression and invasion by chorionic girdle cells.
223
ACKNOWLEDGMENTS
The trophoblast-specific antibody 102.1 was generously supplied by Dr. Doug Antczak.
Statistical analysis was performed under the guidance of Dr. Rick Nordheim, University of
Wisconsin, Madison.
REFERENCES
1. Mossman HW. Vertebrate Fetal Membranes: Comparative Ontogeny and Morphology;
Evolution; Phylogenetic Significance; Basic Functions; Research Opportunities. NewJersey: Rutgers University Press; 1987.
2. Ginther OJ. Reproductive Biology of the Mare: Basic and Applied Aspects. 2nd ed. Cross
Plains, WI:Equiservices; 1992.
3. Enders AC, Liu IKM. Trophoblast-uterine interactions during equine chorionic girdle cell
maturation, migration, and transformation. AmJ Anat 1991; 192:366-381.
4. Allen WR, Moor RM. The origin of the equine endometrial cups. I. Production of PMSG
by fetal trophoblast cells. J Reprod Fertil 1972; 29:313-316.
5. Crump A, Donaldson JM, Kydd JH, Allen WR, Antczak DF. Expression of major histocompatibility complex (MHC) antigens on horse trophoblast. J Reprod Fertil Suppl 1987;
35:379-388.
6. Donaldson WL, Zhang CH, OriolJG, Antczak DF. Invasive equine trophoblast expresses
conventional class I major histocompatibility complex antigens. Development 1990;
110:63-71.
7. Donaldson WL, OriolJG, Plavin A,Antczak DF. Developmental regulation of class I Major
Histocompatibility Complex antigen expression by equine trophoblastic cells. Differentiation 1992; 52:69-78.
8. Kydd JH, Butcher GW, Antczak DF, Allen WR. Expression of major histocompatibility
complex (MHC) class I molecules on early trophoblast. J Reprod Fertil Suppl 1991;
44:463-477.
9. Mignatti P, Rifkin DB. Biology and biochemistry of proteinases in tumor invasion. Physiol
Rev 1993; 73:161-195.
10. Vagnoni KE, Ginther OJ, Lunn DP. Metalloproteinase activity is important for equine
chorionic girdle cell invasion. Biol Reprod 1995; 53:800-805.
11. Steel RGD, Torrie JH. Principles and Procedures of Statistics: A Biometrical Approach.
New York: McGraw-Hill Book Company; 1980.
12. Oriol JG, Poleman JC, Antczak DF. A monoclonal antibody specific for equine trophoblast. Equine Vet J Suppl 1989; 8:14-18.
13. Lunn DP, Holmes MA, Duffus WPH. Equine T lymphocyte MHCII expression: variation
with age and subset. Vet Immunol Immunopathol 1993; 35:225-238.
14. Kydd JH, Antczak DF, Allen WR, Barbis D, Butcher G, Daris W, Edington N, Grunig G,
Holmes MA, Lunn DP, McCullochJ, O'Brien A, Perryman LE,Tavernor A,Williamson S,
Zhang C. Report of the first international workshop of equine leucocyte antigens, Cambridge U.K. Vet Immunol Immunopathol 1994; 42:1-60.
15. Kleiner DE, Stetler-Stevenson WG. Structural biochemistry and activation of matrix metalloproteases. Curr Opin Cell Biol 1993; 5:891-897.
16. Lala PK, Graham C. Mechanisms of trophoblast invasiveness and their control: the role
of proteases and protease inhibitors. Cancer Metastasis Rev 1990; 9:369-379.
17. Alexander CM, Werb Z. Proteinase and extracellular matrix remodeling. Curr Opin Cell
Biol 1989; 1:974-982.
18. Librach CL, Werb Z, Fitzgerald ML, Chiu K, Corwin NM, Esteves RA, Grobelny D, Galardly
R, Damsky CH, Fisher SJ.92-kD type IV collagenase mediates invasion of human cytotrophoblasts.J Cell Biol 1991; 113:437-449.
19. Graham CH, Lala PK. Mechanisms of control of trophoblast invasion in situ. J Cell Physiol
1991; 148:228-234.
20. Graham CH, Lala PK. Mechanisms of placental invasion of the uterus and their control.
Biochem Cell Biol 1992; 70:867-874.
21. Yagel S, Geva TE, Solomon H, Shimonovitz S, Reich R, Finci-Yeheskel Z, Mayer M,
Milwidsky A. High levels of human chorionic gonadotropin retard first trimester trophoblast invasion in vitro by decreasing urokinase plasminogen activator and collagenase
activities. J Clin Endocrinol &Metab 1993; 77:1506-1511.
22. Lennard SN, Steward F, Allen WR. Transforming growth factor B1 (TGFB1) expression
in the endometrium of the mare during pregnancy. J Reprod Fertil Abstract Series 1993;
11:37.
23. McDowell KJ, Adams MH, Baker CB. Chorionic gonadotropin a and subunit RNAs are
present in equine placental membranes by Day 30 of pregnancy. J Equine Vet Sci 1993;
13:515-518.
24. Stewart F, Allen WR. Gene expression in the equine conceptus around the time of implantation. J Reprod Fertil Abstract Series 1991; 7:39.