Endogenous retroviruses regulate periimplantation placental growth and differentiation

Abstract

Endogenous retroviruses (ERVs) are fixed and abundant in the genomes of vertebrates. Circumstantial evidence suggests that ERVs play a role in mammalian reproduction, particularly placental morphogenesis, because intact ERV envelope genes were found to be expressed in the syncytiotrophoblasts of human and mouse placenta and to elicit fusion of cells in vitro. We report here in vivo and in vitro experiments finding that the envelope of a particular class of ERVs of sheep, endogenous Jaagsiekte sheep retroviruses (enJSRVs), regulates trophectoderm growth and differentiation in the periimplantation conceptus (embryo/fetus and associated extraembryonic membranes). The enJSRV envelope gene is expressed in the trophectoderm of the elongating ovine conceptus after day 12 of pregnancy. Loss-of-function experiments were conducted in utero by injecting morpholino antisense oligonucleotides on day 8 of pregnancy that blocked enJSRV envelope protein production in the conceptus trophectoderm. This approach retarded trophectoderm outgrowth during conceptus elongation and inhibited trophoblast giant binucleate cell differentiation as observed on day 16. Pregnancy loss was observed by day 20 in sheep receiving morpholino antisense oligonucleotides. In vitro inhibition of the enJSRV envelope reduced the proliferation of mononuclear trophectoderm cells isolated from day 15 conceptuses. Consequently, these results demonstrate that the enJSRV envelope regulates trophectoderm growth and differentiation in the periimplantation ovine conceptus. This work supports the hypothesis that ERVs play fundamental roles in placental morphogenesis and mammalian reproduction.

• development
• placenta
• sheep
• trophectoderm

Full text

Endogenous retroviruses regulate periimplantation
placental growth and differentiation
Kathrin A. Dunlap*, Massimo Palmarini
†
, Mariana Varela
†
, Robert C. Burghardt
‡
, Kanako Hayashi*, Jennifer L. Farmer*,
and Thomas E. Spencer*
§
*Center for Animal Biotechnology and Genomics, Department of Animal Science, and
‡
Image Analysis Laboratory, Department of Veterinary Integrative
Biosciences, Texas A&M University, College Station, TX 77843; and
†
Institute of Comparative Medicine, University of Glasgow Veterinary School,
Glasgow G61 1QH, United Kingdom
Edited by George E. Seidel, Jr., Colorado State University, Fort Collins, CO, and approved August 8, 2006 (received for review May 10, 2006)
Endogenous retroviruses (ERVs) are fixed and abundant in the
genomes of vertebrates. Circumstantial evidence suggests that
ERVs play a role in mammalian reproduction, particularly placental
morphogenesis, because intact ERV envelope genes were found to
be expressed in the syncytiotrophoblasts of human and mouse
placenta and to elicit fusion of cells in vitro. We report here in vivo
and in vitro experiments finding that the envelope of a particular
class of ERVs of sheep, endogenous Jaagsiekte sheep retroviruses
(enJSRVs), regulates trophectoderm growth and differentiation in
the periimplantation conceptus (embryo兾fetus and associated ex-
traembryonic membranes). The enJSRV envelope gene is expressed
in the trophectoderm of the elongating ovine conceptus after day
12 of pregnancy. Loss-of-function experiments were conducted in
utero by injecting morpholino antisense oligonucleotides on day 8
of pregnancy that blocked enJSRV envelope protein production in
the conceptus trophectoderm. This approach retarded trophecto-
derm outgrowth during conceptus elongation and inhibited tro-
phoblast giant binucleate cell differentiation as observed on day
16. Pregnancy loss was observed by day 20 in sheep receiving
morpholino antisense oligonucleotides. In vitro inhibition of the
enJSRV envelope reduced the proliferation of mononuclear tro-
phectoderm cells isolated from day 15 conceptuses. Consequently,
these results demonstrate that the enJSRV envelope regulates
trophectoderm growth and differentiation in the periimplantation
ovine conceptus. This work supports the hypothesis that ERVs play
fundamental roles in placental morphogenesis and mammalian
reproduction.
development 兩 placenta 兩 sheep 兩 trophectoderm
T
he sheep genome contains ⬇20 copies of endogenous retrovi-
ruses (ERVs) highly related to the exogenous and pathogenic
Jaagsiekte sheep retrovirus (JSRV) (1–3). Endogenous JSRVs
(enJSRVs) are abundantly expressed in the epithelia of the female
genital tract (4). In the placenta, enJSRVs are expressed in the
mononuclear trophectoderm cells of the conceptus (embryo兾fetus
and associated extraembryonic membranes) and are most abundant
in the trophoblast giant binucleate cells (BNCs) and multinucleated
syncytial plaques of the placentomes (5–7). The temporal expres-
sion of the enJSRV envelope (env) gene in the trophectoderm is
coincident with key events in the development of the sheep
conceptus (8). enJSRV env mRNAs are first detected at day 12 (5),
when the blastocyst begins the process of elongation, involving the
intense proliferation and outgrowth of mononuclear trophecto-
derm cells producing IFN-
␶
, the antiluteolytic signal for pregnancy
recognition in ruminants (9, 10).
In sheep, trophoblast giant BNCs differentiate from mononu-
clear trophectoderm cells beginning on day 14, migrate, and then
fuse with the uterine luminal epithelium, as well as each other, to
form multinucleated syncytial plaque s that ultimately form the
cotyledonary portions of the placenta (11). The BNCs derive from
the mononuclear trophectoderm cells by a poorly characterized
mechanism presumably involving mitotic polyploidy, whereas the
syncytial plaques are thought to develop by cell–cell fusion (12).
Hyaluronidase 2 (HYAL2) is a glycosylphosphatidylinositol-
anchored cell-surface protein that can serve as a cellular receptor
for exogenous JSRV Env as well as for retrov iral vectors
pseudotyped by enJSRV Env (13, 14). By RT-PCR analyses,
HYAL2 mRNA is first detected in the conceptus on day 16, which
is associated with the onset of BNC differentiation (5). Throughout
pregnancy, HYAL2 mRNA can be detected in the BNCs and
multinucleated syncytia of sheep placentomes but not in the mono-
nuclear trophectoderm cells of the conceptus or any cells of the
endometrium.
Of great interest for comparative physiology is that enJSRV env
expression in the developing ovine placenta is strikingly similar to
that observed for syncytin 1 and 2, products of human ERV env in
humans and primates (15–19) and possibly of two related env genes
(syncytin A and syncytin B) in mice (20). Syncytins encode highly
fusogenic retroviral envelope proteins that are expre ssed in the
syncytiotrophoblast layer generated by mononuclear cytotropho-
blast cell fusion at the maternal–fetal interface. Syncytins are
fusogenic when expressed in vitro, thereby advancing the hypothesis
that they are involved in placental morphogenesis (15–19). Thus,
circumstantial evidence gleaned from studies of primate s, sheep,
and rodents supports the concept that independently acquired
ERVs have been positively selected for a convergent physiological
role in placental morphogenesis (21, 22).
In these studies we tested the hypothesis that enJSRV Env has a
biological role in periimplantation ovine conceptus development
and placental morphogene sis by using an in vivo morpholino
loss-of-function approach (23) to block enJSRV Env production
in utero.
Results
A morpholino antisense oligonucleotide (MAO) was designed to
specifically inhibit expre ssion of enJSRV env mRNAs (MAO-env)
(Fig. 1A). MAOs inhibit RNA splicing and兾or translation by a steric
block mechanism that is RNase H-independent (23). Morpholinos
are effective only when designed to complement the nucleotide
region around the start codon and兾or possible splicing sites of a
given gene mRNA. The nucleotide sequence around the splice
acceptor and start codon of the exogenous JSRV env and the known
enJSRV loci are highly conserved, indicating that one common
MAO should inhibit splicing and translation of most enJSRV
Author contributions: K.A.D., M.P., M.V., and T.E.S. designed research; K.A.D., M.P., M.V.,
R.C.B., K.H., J.L.F., and T.E.S. performed research; M.P., M.V., R.C.B., K.H., J.L.F., and T.E.S.
contributed new reagents兾analytic tools; K.A.D., M.P., M.V., and T.E.S. analyzed data; and
K.A.D., M.P., and T.E.S. wrote the paper.
The authors declare no conflict of interest.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: ERV, endogenous retrovirus; JSRV, Jaagsiekte sheep retrovirus; enJSRV,
endogenous retrovirus related to JSRV; BNC, binucleate cell; MAO, morpholino antisense
oligonucleotide; CSH, chorionic somatomammotropin hormone; PAG, pregnancy-associ-
ated glycoprotein.
§
To whom correspondence should be addressed. E-mail: tspencer@tamu.edu.
© 2006 by The National Academy of Sciences of the USA
14390–14395
兩
PNAS
兩
September 26, 2006
兩
vol. 103
兩
no. 39 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0603836103

proviral loci expre ssing an intact env gene (6). To examine mor-
pholino effectiveness, we conducted a series of in vitro studies in
transiently transfected 293T cells (Fig. 1 B–D). MAO-env effec-
tively inhibited expression of enJS5F16 (one of the cloned enJSRV
loci) (6) Env (Fig. 1B, lane 2) in a dose-dependent fashion (Fig. 1C).
A five-mismatch (MAO-5mis) and a standard (MAO-std) control
morpholino had no effect on the expression of enJS5F16 Env (Fig.
1 B and C). As expected, MAO-env and the control morpholinos
did not affect expre ssion of enJSRV Gag (the polyprotein forming
the retroviral capsid) in transfected 293T cells (Fig. 1D), because
the retroviral Gag is produced from a full-length genomic mRNA,
whereas Env is produced only from correctly spliced mRNA (24)
(Fig. 1 A). Thus, MAO-env specifically and effectively inhibits
enJSRV env mRNA translation.
To conduct in vivo loss-of-function studie s, we injected the
morpholinos into the lumen of the ovine uterus on day 8 after
mating and determined their effect on conceptus development and
pregnancy e stablishment on either day 16 (study one) or day 20
(study two). The blastocyst hatches from the zona pellucida on day
8 of pregnancy in sheep (10). The large size of the uterine horn
(10–12 cm in length and 2–3 cm in width) makes intraluminal
injection feasible in this large-animal model. The amount of mor-
pholino used for these studies was extrapolated from work in the
mouse (25). A dose–response study to determine the minimal
amount of morpholino that can be used for in vivo studies was not
conducted. Indeed, this amount may be gene-specific, given the
relative difference s in mRNA abundance for each transcribed gene.
In study one, pregnancy rates were not different (P ⬎ 0.10) among
Table 1. Effect of morpholinos on pregnancy and conceptus development in sheep
Morpholino
Study one: Day 16 after mating Study two: Day 20 after mating
Pregnancy
rate,* %
Conceptus
development
IFN-
␶
in
uterine flush BNCs, %
Pregnancy
rate,* %
Conceptus
development
IFN-
␶
in
uterine flush BNCs, %
MAO-std 100 (5兾5) Elongated,
filamentous
282 ⫾ 50 12 ⫾ 1 100 (5兾5) Elongated,
filamentous
ND ND
MAO-5mis 100 (5兾5) Elongated,
filamentous
232 ⫾ 50 10 ⫾ 183(5兾6) Elongated,
filamentous
ND ND
MAO-env 100 (5兾5) Growth-retarded,
filamentous
†
68 ⫾ 10 1 ⫾ 120(1兾5) Not present
‡
ND ND
All ewes received intrauterine injections of morpholinos on day 8 after mating and were hysterectomized on either day 16 or 20. ND, not determined.
*Pregnancy rate data are expressed as percentage of ewes with a conceptus recovered from the uterine lumen by flush before hysterectomy (day 16) or present
in the uterine lumen at hysterectomy (day 20) (pregnant兾total).
†
Conceptuses were growth-retarded, slightly elongated, and fragile with no detectable trophoblast giant BNCs.
‡
The single recovered conceptus was severely growth-retarded as compared with conceptuses recovered from MAO-5mis and MAO-std controls and contained
no detectable BNCs (data not shown).
Fig. 1. Design and effects of morpholinos on enJSRV Env expression in vitro.(A) MAO-env was designed to inhibit splicing and translation of enJSRV env mRNA but
not expression of full-length genomic RNA (which expresses the viral Gag). (B) 293T cells were mock-transfected (lane 1) or transfected with pSV-En2EnvFlag, a simian
virus 40-driven expression plasmid for enJS5F16 env cDNA tagged with a Flag epitope at the C terminus (lanes 2–4). Cells were then treated with MAO-env (lane 2) or
MAO-5mis (lane 3) and MAO-std as controls (lane 4). All morpholinos were complexed with the Endo-Porter delivery reagent and used at a final concentration of 80
␮
M. After 48 h, enJS5F156 Env expression was determined by immunoprecipitation (IP) and Western blot analysis (WB). Note that the full-length retroviral Env is
processed into a surface domain and a transmembrane domain (TM). (C) 293T cells were mock-transfected (lane 1) or transfected with pSV-En2EnvFlag as above. Cells
were then treated with Endo-Porter alone (lane 2), MAO-std as a control (lane 3), MAO-5mis as a control (lanes 4 – 6; 20, 40, and 80
␮
M, respectively), or MAO-env (lanes
7–9; 20, 40, and 80
␮
M, respectively). All morpholinos were complexed with Endo-Porter delivery reagent. After 24 h, enJS5F16 Env expression was determined by
immunoprecipitation and Western blot analysis as in B.(D) 293T cells were mock-transfected (lane 1) or cotransfected with pSV-En2EnvFlag and pCMV2 en56A1
expressing the full-length en56A1 clone (lanes 2–4). Cells were then treated Endo-Porter alone (lanes 1 and 2), MAO-5mis as a control (lane 3), or MAO-env (lane 4).
All morpholinos were complexed with Endo-Porter delivery reagent and used at a final concentration of 80
␮
M. After 48 h, enJSRV Env expression (Upper) was
determined by immunoprecipitation and Western blot analysis as in B and C and Gag expression by Western blot analysis (Lower).
Dunlap et al. PNAS
兩
September 26, 2006
兩
vol. 103
兩
no. 39
兩
14391
DEVELOPMENTAL
BIOLOGY

the various treatments (Table 1). However, major differences were
observed in conceptus development in ewes receiving MAO-env as
compared with the MAO-5mis or MAO-std controls. The concep-
tuses from all ewes injected with the control morpholinos, MAO-
5mis and MAO-std, were fully elongated and filamentous, which is
typical of day 16 of pregnancy (Fig. 2A). Although the conceptuses
recovered from ewes injected with MAO-env were elongated, they
were fragile and substantially smaller than those from ewes injected
with the control morpholinos (Fig. 2 A). Histological examination
found that conceptuses from ewes injected with either control
morpholino were of normal appearance, containing many mono-
nuclear trophectoderm cells (Fig. 2B). In contrast, conceptuse s
from MAO-env-treated ewes seemed to have fewer mononuclear
trophectoderm cells compared with the conceptuse s from control
ewes. In addition, most mononuclear trophectoderm cells from the
M AO-env-treated ewes displayed intracytoplasmic vacuoles
(Fig. 2B).
To quantify the effect of MAO-env on mononuclear trophecto-
derm cell growth, we measured the levels of IFN-
␶
(which is
produced solely by mononuclear trophectoderm cells) in uterine
flushings (9). Consistent with retarded growth of the conceptus
(Fig. 2 A), the relative amounts of IFN-
␶
in the uterine flushings of
MAO-env-injected ewes (68 ⫾ 10 relative units) were considerably
lower (P ⬍ 0.05) than MAO-5mis-injected (232 ⫾ 50 relative units)
and MAO-std-injected (282 ⫾ 50 relative units) ewes (Table 1).
Next, we evaluated conceptuse s for the presence of trophoblast
giant BNCs, which derive from mononuclear trophectoderm cells.
Chorionic somatomammotropin hormone 1 (CSH1; alias placental
lactogen) and pregnancy-associated glycoproteins (PAGs) are ex-
pressed only by BNCs and are useful markers for these cells (10).
Immunostaining for CSH1 and PAG proteins found 10 ⫾ 1% or
12 ⫾ 1% BNCs in the conceptuses from control ewes (MAO-5mis
and MAO-std, respectively), whereas BNCs were very scarce (1 ⫾
1%) or not present at all in conceptuses recovered from MAO-
env-injected ewes (Fig. 2 C and D). Therefore, in vivo enJSRV Env
knockdown in the conceptus retarded trophectoderm growth and
prohibited differentiation of trophoblast giant BNCs.
Next we confirmed that retarded development of conceptuse s of
MAO-env-treated ewes was associated with a reduction in enJSRV
Env (Fig. 3). As shown in Fig. 3A, rhodamine-labeled morpholinos
were observed in trophectoderm cells of the conceptus as well as in
luminal epithelium and superficial glandular epithelium of the
endometrium. In conceptuse s recovered from ewes injected with
MAO-5mis and MAO-std controls, abundant enJSRV Env protein
was observed at the apical surfaces of the mononuclear trophec-
toderm cells, as well as the luminal epithelium and glandular
epithelium of the endometrium. As assessed by immunofluores-
cence analysis, enJSRV Env expre ssion was almost completely
diminished in the trophectoderm of day 16 conceptuses recovered
from ewes injected with MAO-env and was also substantially
decreased in the uterine luminal epithelium (Fig. 3B). Very little or
no background binding was observed in negative controls in which
rabbit IgG was substituted for the primary antibody or the primary
antibody was omitted in the procedure (data not shown). Immu-
nofluorescence analyses are required for evaluation of enJSRV Env
abundance with the rabbit anti-JSRV Env antibody because it does
not recognize the Env protein under Western blot conditions (M.P.,
M.V., and T.E.S., unpublished results). No difference s in enJSRV
Env were observed in the middle to deep uterine glandular epi-
thelia, which was expected because of the inability of the morpho-
linos to be delivered to those cells (data not shown). As expected
from in vitro results (Fig. 1), expre ssion of enJSRV Gag protein in
the conceptus trophectoderm and endometrial epithelia was not
affected by the morpholinos (Fig. 3C). Thus, the observed alter-
ations in conceptus development, trophectoderm growth, and BNC
differentiation in MAO-env-treated ewes could be directly corre-
lated with the inhibition of enJSRV Env by the MAO-env
approach.
In study two, we determined whether conceptuse s from MAO-
env-treated ewes could establish pregnancy. Ewes were injected
with morpholinos on day 8 after mating, and effects were assessed
on day 20. Early pregnancy loss occurred in almost all MAO-env-
injected ewes, but not in control ewes (Table 1). Pregnancy rates of
100% and 83% were not different (P ⬎ 0.10) in control ewes
receiving MAO-std or MAO-5mis control morpholinos, respec-
tively, but were substantially reduced (P ⬍ 0.025) to 20% in
MAO-env-treated ewes (Table 1). The single recovered conceptus
from a MAO-env-injected ewe was severely growth-retarded and
contained no detectable BNCs (data not shown). Most of the ewes
injected with MAO-env exhibited estrus at day 17–18 after mating,
which is indicative of early pregnancy loss because of inadequate
IFN-
␶
production by the conceptus. At hysterectomy, a normal
elongated, filamentous, and implanting conceptus was observed in
the ligated uterine horn of ewes injected with control morpholinos
(data not shown).
To complement the in vivo data from studies one and two, we
isolated mononuclear trophectoderm cells from day 15 conceptuses
and cultured them in vitro. As asse ssed by RT-PCR analysis, the
cultured mononuclear trophectoderm cells expressed enJSRV env
Fig. 2. Effects of morpholinos on periimplantation conceptus trophoblast
growth and differentiation. MAO-std, MAO-5mis, or MAO-env was injected into
the uterine lumen on day 8 after mating, and conceptuses were recovered on day
16 (see Materials and Methods for experimental details). (A) Morphology of the
conceptuses was examined by using an inverted microscope. Micrographs are
shown at the same magnification. Note the retarded growth in the conceptus
recovered from a MAO-env-treated ewe. (B) Portions of the conceptuses were
fixed in paraformaldehyde, embedded in paraffin, sectioned, and stained with
hematoxylin and eosin. (Width of each field of view is 420
␮
m with the Inset at 85
␮
m.) (C and D) Trophoblast giant binucleate cells (BNCs) in conceptuses were
detected by pregnancy-associated glycoproteins (PAGs) (C) and CSH1 (alias pla-
cental lactogen) (D) in the conceptus. Immunoreactive PAG and CSH1 proteins
were detected in paraformaldehyde-fixed, paraffin-embedded sections of con-
ceptuses by using a rabbit anti-ovine PAG or anti-ovine CSH1 antibody. (Width of
each field of view is 420
␮
m with the Inset at 210
␮
m.) Data are representative of
conceptuses from all ewes. MTC, mononuclear trophectoderm cell.
14392
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0603836103 Dunlap et al.

and gag mRNAs as well as IFN-
␶
mRNA, whereas CSH1 and PAG
mRNAs were not detected (data not shown). Morpholinos were
effectively delivered to ⬎95% of the trophectoderm cells in vitro
(Fig. 4A). Immunofluorescence analyses found that treatment of
trophectoderm cells in vitro with MAO-env inhibited expre ssion of
enJSRV Env, whereas the MAO-std and MAO-5mis controls had
no effect on enJSRV Env abundance (Fig. 4B). As expected, none
of the morpholinos affected enJSRV Gag abundance (Fig. 4B).
Trophectoderm cell number was reduced (P ⬍ 0.05) by 33% in
cultures treated with the MAO-env, whereas MAO-std and MAO-
5mis controls had no effect on cell proliferation (Fig. 4C).
Discussion
These results demonstrate an essential role for ERVs in placental
morphogenesis by in vivo experimentation. When a morpholino
Fig. 4. Effects of morpholinos on in vitro ovine trophectoderm growth. Mono-
nuclear trophectoderm cells were isolated from day 15 conceptuses. (A) Morpho-
lino delivery. Rhodamine-labeled MAO-std was complexed with Endo-Porter
aqueous delivery reagent and added to cells in culture. Fluorescence microscopy
was used to visualize the labeled MAO in cells. [Width of each field of view is 870
␮
m(Left) and 90
␮
m(Right).] (B) Effect of morpholinos on enJSRV Env and Gag
protein in cultured trophectoderm cells. Cells were grown on glass slides and
mock-treated or treated with MAO-std, MAO-5mis, or MAO-env for 48 h. Immu-
nofluorescence analysis determined that synthesis of enJSRV Env, but not Gag,
protein was inhibited in cells treated with MAO-env but not the other morpho-
linos. Results are representative of three experiments. (Width of each field of view
is 140
␮
m.) (C) Effect of morpholinos on trophectoderm cell growth. Cells were
grown in culture dishes until 30% confluency and mock-treated (no morpholino)
or treated with MAO-std, MAO-5mis, or MAO-env for 48 h. Cell number was
reduced (P ⬍ 0.05) by 33% in cultures treated with the MAO-env relative to
control morpholinos. Results are from three independent experiments, and data
are expressed as the percentage of cell number in mock-treated cultures.
Fig. 3. Delivery and effectiveness of morpholinos in vivo. MAO-std, MAO-5mis,
or MAO-env was injected into the uterine lumen of sheep on day 8 after mating,
and the conceptuses were removed on day 16 (see Materials and Methods for
experimental details). (A) Portions of the conceptuses were frozen in optimal
cutting temperature (OCT) compound and sectioned. Sections were rinsed in PBS,
and a coverslip was affixed by using DAPI-containing mounting medium. Fluo-
rescence microscopy was used to detect the rhodamine-labeled morpholino
(orange兾red) and DAPI nuclei (blue). (B and C) Conceptuses and uteri were
sectioned and analyzed for enJSRV Env protein by immunofluorescence analysis
using a rabbit antiserum toward the JSRV Env (B)orGag(C) with a FITC-labeled
secondary antibody. (Width of each field of view is 140
␮
m.) Data are represen-
tative of conceptuses from all ewes. LE, luminal epithelium; sGE, superficial
glandular epithelium; S, stroma; Tr, trophectoderm.
Dunlap et al. PNAS
兩
September 26, 2006
兩
vol. 103
兩
no. 39
兩
14393
DEVELOPMENTAL
BIOLOGY

loss-of-function approach in utero was used, enJSRV Env knock-
down caused a reduction in trophectoderm outgrowth during
blastocyst elongation and formation of the conceptus. In sheep,
pregnancy recognition and establishment involve elongation of the
spherical blastocyst to a filamentous conceptus between days 12 and
16 and production of IFN-
␶
by the conceptus (26, 27). IFN-
␶
is a
developmentally regulated gene that is expressed only in the
mononuclear trophectoderm cells of the sheep conceptus (28) from
days 10 to 20 with a peak on day 16 (29–31). IFN-
␶
is antiluteolytic
and acts on the endometrium to inhibit development of the
luteolytic mechanism, thereby maintaining corpus luteum function
and ensuring continued production of progesterone, which is
necessary for pregnancy (32). Here, a reduced amount of IFN-
␶
was
found in the uterine flushes of MAO-env-treated ewes containing
a growth-retarded conceptus with fewer mononuclear trophecto-
derm cells. Thus, the pregnancy loss observed before day 20 in
MAO-env-treated ewes is likely attributable to an inability of the
growth-retarded conceptus to produce sufficient IFN-
␶
to abrogate
development of the endometrial luteolytic mechanism, resulting in
luteolysis and a return to estrus (33).
The in vivo and in vitro loss-of-function studie s presented here
strongly support the hypotheses that enJSRVs play a fundamental
role in mononuclear trophectoderm cell outgrowth and differen-
tiation of trophoblast giant BNCs during the periimplantation
period of pregnancy (8). Unfortunately, little is known of the
cellular and molecular mechanisms regulating trophectoderm pro-
liferation and differentiation during early pregnancy in ruminants
(34). Further, the cellular and molecular mechanism(s) whereby
enJSRV Env has biological effects within cells is unknown. It is of
interest that JSRV env (the exogenous counterpart of enJSRVs)
encodes a functional structural protein that is a dominant oncop-
rotein, a unique feature among oncogenic retroviruse s (35). Thus,
it is possible that enJSRV and JSRV Env proteins share common
mechanisms by which they influence the cell cycle and cell prolif-
eration, but only the Env of the exogenous JSRV evolved to become
truly oncogenic. It is also possible that enJSRV Env is essential for
trophoblast giant BNCs and formation of multinucleated syncytia
by eliciting cell–cell fusion similar to proposed actions of syncytins
in humans and mice (15–20). However, results of study one clearly
indicate that enJSRV Env influences the mononuclear trophecto-
derm cell growth and differentiation during conceptus elongation
that precedes the formation of multinucleated syncytia. None of the
enJSRV Env cloned to date elicits syncytia formation in vitro when
ovine, human, or mouse cell lines expressing ovine HYAL2 are
used (M.P., M.V., M. Golder, K.A.D., and T.E.S., unpublished
results). However, cloning of all sheep enJSRV loci will be neces-
sary to investigate enJSRV Env structure, function, and distribution
among Artiodactyla, given that the sheep genome has not been
sequenced. Available data based on zoo blots suggest that enJSRVs
are present in most Caprinae (36); however, enJSRV locus-specific
PCR analyses will be needed to assess the evolution and distribution
of the biologically relevant proviruses.
From an evolutionary point of view, we speculate that the
protection of the host against related exogenous retroviruse s was a
driving force influencing the fixation of ERVs in the genome of
various mammals (21, 22). For example, enJSRVs interfere with
exogenous JSRV at both early and late steps of the replication cycle
(14, 37). After the fixation of ERVs in the germ line of the host,
their expre ssion in the placenta may have favored conceptus
development and increased reproductive efficiency (22, 38, 39).
Given that enJSRVs are important for conceptus growth and
differentiation, it is likely that some of the host mechanisms
governing the se reproductive processe s may have been lost later
during evolution. In conclusion, available evidence obtained in vivo
in sheep and in vitro in primates and rodents strongly supports the
hypothesis that independently acquired ERVs were positively se-
lected for a convergent physiological role in placental morphogen-
esis (22). The enormous structural variability of placentae among
major taxa supports a model where retroviruses have conferred
increased diversity and functionality during evolution (21).
Materials and Methods
Morpholino Design. Morpholino oligonucleotides were designed and
synthesized by Gene Tools (Philomath, OR). The MAO-env had
the sequence GCTTC GGCAT CCTGT GGAAA AACAC and
targeted to the enJSRVs env RNA overlapping the splice donor and
acceptor region (see Fig. 1 A). The MAO-5mis control morpholino
had the sequence GGTTC GCCAT CCTCTGCAAA AAGAC
(italic type indicates differences from MAO-env). The MAO-std
had the sequence CCTCT TACCT CAGTT ACAAT TTATA and
targeted to a splice site mutant of Homo sapiens hemoglobin
␤
-chain (HBB) gene (GenBank accession no. AY605051). All
morpholinos were synthesized with a 5⬘ rhodamine modification for
convenient detection.
In Vitro
Transfection Studies. pSV-En2EnvFlag expre sse s, under the
control of the simian virus 40 promoter, the Env of enJS5F16 (6)
tagged with a Flag epitope at the C terminus. pCMV2en56A1
expresses the full-length enJS56A1 locus (6). Human 293T cells
were transfected with pSV-En2EnvFlag or cotransfected with
pSV-En2EnvFlag and pCMV2 en56A1 by using Lipofectamine
(Invitrogen, Carlsbad, CA). After 3 h, cells were washed with PBS
and incubated with Endo-Porter aqueous delivery reagent (Gene
Tools) (8
␮
l兾ml medium) and 20–80
␮
M MAO-env, MAO-5mis,
or MAO-std. After 48 h, transfected cells were lysed, and cell lysates
were analyzed for the presence of enJSRV Env or Gag by immu-
noprecipitation and兾or Western blotting employing an anti-Flag
antibody (Sigma–Aldrich, St. Louis, MO) and an anti-JSRV p23
(Matrix Science, London, U.K.) as previously described (37).
In Vivo
Studies. All animal experiments were approved by the
Institutional Animal Care and Use Committee of Texas A&M
University. Suffolk cross-bred ewe s were mated at estrus (day 0)
and on day 1 to rams of proven fertility. On day 8 after mating, the
ewes were subjected to a midventral laparotomy. The base of the
uterine horn ipsilateral to the corpus luteum was double-ligated by
using nonabsorbable umbilical tape to prevent migration and
growth of the conceptus through the uterine body into the con-
tralateral uterine horn. This surgical procedure does not affect
conceptus implantation or fetal development in sheep (40). MAO-
std, MAO-5mis, and MAO-env (100 nmol) were complexed with
Gene Tools Endo-Porter delivery reagent (50
␮
l) and diluted to a
1-ml final volume with OPTI-MEM (Invitrogen). The complex was
then introduced into the lumen of the uterus (n ⫽ 5–6 ewes per
morpholino) via the uterotubal junction by using a 1-ml syringe
fitted with a 20-gauge catheter. The volume of the uterine lumen of
one horn is ⬇250–500
␮
l. After the morpholinos had been dis-
charged into the uterine lumen, the catheter was withdrawn, and the
uterine horn was gently massaged to distribute the morpholinos
throughout the uterine lumen. The outside of the uterus was rinsed
with sterile 5% (vol兾vol) glycerol in saline to prevent the formation
of adhesions and placed back in the body cavity.
For study one, the morpholino-injected ewes were hysterecto-
mized on day 16. The uterine horn injected with the morpholinos
was flushed with 10 ml of sterile PBS. If the conceptus was present,
its morphology was recorded (spherical, tubular, or elongated). The
conceptus was immediately removed from the uterine flush with a
pipette, and the volume of the flush was recorded. Photomicro-
graphs of the conceptus were obtained by using an inverted
microscope fitted with a digital camera. Portions of the conceptus
were then placed in optimal cutting temperature (OCT) compound
(Miles, Oneonta, NY), frozen in liquid nitrogen, and stored at
⫺80°C. Another portion of the conceptus was fixed in freshly
prepared 4% (wt兾vol) paraformaldehyde in PBS and embedded in
paraffin wax. The uterine flush was clarified by centrifugation
(5,000 ⫻ g for 15 min at 4°C), aliquoted, and stored at ⫺80°C. The
14394
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0603836103 Dunlap et al.

amount of IFN-
␶
in the uterine flush was quantified by semiquan-
titative Western blot analysis as described in ref. 41.
For study two, the morpholino-injected ewes were hysterecto-
mized on day 20. The uterine horn injected with the morpholinos
was not flushed but rather opened along the mesometrial border to
expose the conceptus. Portions of the conceptus (if present) and
sections of the uterine horn containing the conceptus were frozen
in optimal cutting temperature (OCT) compound or fixed in 4%
(vol兾vol) paraformaldehyde for histological analysis.
Histology, Immunohistochemistry, and Immunofluorescence. Effec-
tive delivery of the rhodamine-labeled morpholinos was analyzed by
fluorescence microscopy. Cryosections of the uteri and conceptuses
were prepared, and a DAPI-containing mounting medium was
used to visualize nuclei. The enJSRV Env and Gag proteins were
evaluated in frozen conceptus tissue sections by immunofluores-
cence staining using a rabbit antiserum toward the JSRV Env or
Gag as described previously (6, 37). Negative controls included
substituting rabbit IgG in place of the primary antibody as well as
removal of the primary antibody. Filter checks were used to ensure
specific signals. Immunoreactive CSH1 and PAG proteins were
assessed in the conceptus by using standard immunohistochemical
procedures with antibodies to ovine CSH1 (rabbit polyclonal anti-
body was kindly provided by Russ V. Anthony, Colorado State
University, Fort Collins, CO) and ovine PAG (rabbit anti-
recombinant PAG3 polyclonal antibody was kindly provided by
Jonathan A. Green, University of Missouri-Columbia, Columbia,
MO). Rabbit IgG was used in place of the relevant primary antibody
as a negative control. The number of BNCs was quantified by
determining the number of CSH1- and PAG-immunostained BNCs
relative to mononuclear trophoblast cells in at least five nonse-
quential sections of each conceptus from each ewe. The nuclei were
visualized after applying hematoxylin counterstain. Sections from
all sheep and treatment groups were analyzed in duplicate in
the same run, and images were captured by using standardized
procedures.
In Vitro
Ovine Mononuclear Trophectoderm Cell Culture and Prolifer-
ation Assay. Conceptuses from day 15 pregnant ewes were recov-
ered by sterile flush. The inner cell mass was removed by dissection,
and trophectoderm cells were isolated and cultured by using
methods described previously (42). Rhodamine-labeled MAO-std
(100 nmol) was complexed with Endo-Porter aqueous delivery
reagent (6
␮
l兾ml of medium) and added to cells in culture.
Fluorescence microscopy was used to visualize the labeled MAO in
cells 24 h after treatment by using a Zeiss Stallion Double Detector
Imaging system (Carl Zeiss Microimaging, Thornwood, NY).
For immunofluorescence analyses, cells were grown on two-well
Lab-Tek Coverglass Chambered slides (Nalge Nunc, Rochester,
NY) and mock-treated (no morpholino) or treated with MAO-std,
MAO-5mis, or MAO-env (100 nmol) complexed with Endo-Porter
aqueous delivery reagent for 48 h. Immunoreactive enJSRV Env
and Gag proteins were analyzed by immunofluorescence staining
using a rabbit antiserum toward the JSRV Env or Gag as described
previously (6, 37); rabbit IgG was used in place of the primary
antibody as a negative control. Cells were sequentially imaged by
using a Cy3 filter set followed by differential interference contrast
optics using either a Plan-Apochromat (10⫻; n.a., 0.45) or a
C-Apochromat (63⫻ water correction; n.a., 1.2) objective lens. The
experiment was independently repeated three times.
To determine the effects of morpholinos on trophectoderm cell
proliferation, cells were grown in six-well culture dishes until 30%
confluency and mock-treated (no morpholino) or treated with
MAO-std, MAO-5mis, or MAO-env (100 nmol complexed to
Endo-Porter aqueous delivery reagent) in triplicate for 48 h. A
colorimetric assay using Janus green dye was used to assess cell
numbers (43). The experiment was independently repeated three
times.
Statistics. Pregnancy rate data were analyzed by
␹
2
test. All quan-
titative data were subjected to least-squares ANOVA by using the
General Linear Models procedures of the Statistical Analysis
System (SAS Institute, Cary, NC). Statistical models accounted for
sources of variation, including the main effects of morpholino type
and, where appropriate, histological section (BNC quantification)
or replicate (proliferation assays). The following data are presented
as least-squares means with SEM: (i) data for IFN-
␶
in uterine flush
and the number of BNCs in Table 1 and (ii) data for cell number
in Fig. 4C.
We thank F. W. Bazer, J. C. Neil, and O. Jarrett for useful discussion and
R. V. Anthony and J. L. Green for provision of reagents. This work was
supported in part by the Wellcome Trust (to M.P.) and grants from the
National Cancer Institute (to M.P.) and National Institute of Environ-
mental Health Sciences (to T.E.S. and R.C.B.). M.P. is a Wolfson–Royal
Society Research Merit Awardee.
1. DeMartini JC, Carlson JO, Leroux C, Spencer T, Palmarini M (2003) Curr Top
Microbiol Immunol 275:117–137.
2. Palmarini M, Sharp JM, De las Heras M, Fan H (1999) J Virol 73:6964–6972.
3. Palmarini M, Mura M, Spencer TE (2004) J Gen Virol 85:1–13.
4. Palmarini M, Gray CA, Carpenter K, Fan H, Bazer FW, Spencer TE (2001) J Virol
75:11319–11327.
5. Dunlap KA, Palmarini M, Adelson DL, Spencer TE (2005) Biol Reprod 73:271–279.
6. Palmarini M, Hallwirth C, York D, Murgia C, de Oliveira T, Spencer T, Fan H (2000)
J Virol 74:8065– 8076.
7. Sanna E, Sanna MP, L oddo C, Sanna L, Mura M, Cadelano T, Leoni A (2002) Eur
J Histochem 46:273–280.
8. Dunlap KA, Palmarini M, Spencer TE (2006) Placenta 27(Suppl A):5135–5140.
9. Roberts RM, Ealy AD, Alexenko AP, Han CS, Ezashi T (1999) Placenta 20:259–264.
10. Guillomot M (1995) J Reprod Fer til Suppl 49:39–51.
11. Wooding FB (1984) Am J Anat 170:233–250.
12. Wooding FB (1992) Placenta 13:101–113.
13. Miller AD (2003) Cur r Top Microbiol Immunol 275:179–199.
14. Spencer TE, Mura M, Gray CA, Griebel PJ, Palmarini M (2003) J V irol 77:749–753.
15. Blaise S, de Parseval N, Benit L, Heidmann T (2003) Proc Natl Acad Sci USA
100:13013–13018.
16. Blond JL, Lavillette D, Cheynet V, Bouton O, Oriol G, Chapel-Fernandes S,
Mandrand B, Mallet F, Cosset FL (2000) J Virol 74:3321–3329.
17. Frendo JL, Olivier D, Cheynet V, Blond J-L, Bouton O, Vidaud M, Rabreau M,
Evain-Brion D, Mallet F (2003) Mol Cell Biol 23:3566–3574.
18. Mallet F, Bouton O, Prudhomme S, Cheynet V, Oriol G, Bonnaud B, Lucotte G,
Duret L, Mandrand B (2004) Proc Natl Acad Sci USA 101:1731–1736.
19. Mi S, Lee X, Li X, Veldman GM, Finnerty H, Racie L, LaVallie E, Tang XY,
Edouard P, Howes S, et al. (2000) Nature 403:785–789.
20. Dupressoir A, Marceau G, Vernochet C, Benit L, Kanellopoulos C, Sapin V,
Heidmann T (2005) P roc Natl Acad Sci USA 102:725–730.
21. Harris JR (1998) BioEssays 20:307–316.
22. Prudhomme S, Bonnaud B, Mallet F (2005) Cytogenet Genome Res 110:353–364.
23. Summerton J (1999) Biochim Biophys Acta 1489:141–158.
24. Vogt VM (1997) in Retroviruses, eds Cof fin JM, Hughes SH, Varmus HE (Cold
Spring Harbor Lab Press, Cold Spring Harbor, NY), pp 27–69.
25. Luu KC, Nie GY, Salamonsen LA (2004) Proc Natl Acad Sci USA 101:8028–8033.
26. Spencer TE, Bazer FW (2004) Reprod Biol Endocrinol 2:49. Available at www.rbe-
j.com兾content兾2兾1兾49.
27. Roberts RM, Ealy AD, Alexenko AP, Han CS, Ezashi T (1999) Placenta 20:259–264.
28. Farin CE, Imakawa K, Hansen TR, McDonnell JJ, Murphy CN, Farin PW, Roberts
RM (1990) Biol Reprod 43:210–218.
29. Ashworth CJ, Bazer FW (1989) Biol Reprod 40:425–433.
30. Farin CE, Imakawa K, Roberts RM (1989) Mol Endocrinol 3:1099–1107.
31. Guillomot M, Michel C, Gaye P, Charlier N, Trojan J, Martal J (1990) Biol Cell
68:205–211.
32. Spencer TE, Ott TL, Bazer FW (1996) Proc Soc Exp Biol Med 213:215–229.
33. Mann GE, Lamming GE (1999) Reprod Domest Anim 34:269–274.
34. Cammas L, Reinaud P, Dubois O, Bordas N, Germain G, Charpigny G (2005) Biol
Reprod 72:960–967.
35. Fan H (2003) Jaagsiekte Sheep Retrovir us and Lung Cancer (Springer, Berlin).
36. Hecht SJ, Stedman KE, Carlson JO, DeMartin i JC (1996) Proc Natl Acad Sci USA
93:3297–3302.
37. Mura M, Murcia P, Caporale M, Spencer TE, Nagashima K, Rein A, Palmarin i M
(2004) Proc Natl Acad Sci USA 101:11117–11122.
38. Villarreal LP (1997) J Virol 71:859–865, and erratum (1998) 72:6277.
39. Stoye JP, Coffin JM (2000) Nature 403:715–717.
40. Bazer FW, Roberts RM, Basha SM, Zavy MT, Caton D, Barron DH (1979) J Anim
Sci 49:1522–1527.
41. Satterfield MC, Bazer FW, Spencer TE (2006) Biol Reprod 75:289–296.
42. Johnson GA, Bazer FW, Jaeger LA, Ka H, Garlow JE, Pfarrer C, Spencer TE,
Burghardt RC (2001) Biol Reprod 65:820– 828.
43. Raspotnig G, Fauler G, Jantscher A, Windischhofer W, Schachl K, Leis HJ (1999)
Anal Biochem 275:74–83.
Dunlap et al. PNAS
兩
September 26, 2006
兩
vol. 103
兩
no. 39
兩
14395
DEVELOPMENTAL
BIOLOGY