FcRn in the Yolk Sac Endoderm of Mouse Is
Required for IgG Transport to Fetus
This information is current as
of June 18, 2017.
Jonghan Kim, Sudhasri Mohanty, Latha P. Ganesan, Keding
Hua, David Jarjoura, William L. Hayton, John M. Robinson
and Clark L. Anderson
J Immunol 2009; 182:2583-2589; ;
doi: 10.4049/jimmunol.0803247
http://www.jimmunol.org/content/182/5/2583
Subscription
Permissions
Email Alerts
This article cites 25 articles, 10 of which you can access for free at:
http://www.jimmunol.org/content/182/5/2583.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2009 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
References
The Journal of Immunology
FcRn in the Yolk Sac Endoderm of Mouse Is Required for IgG
Transport to Fetus1
Jonghan Kim,2*† Sudhasri Mohanty,2* Latha P. Ganesan,* Keding Hua,‡ David Jarjoura,‡
William L. Hayton,† John M. Robinson,§ and Clark L. Anderson3*
I
gG is transported across the human placenta from maternal
circulation to the fetus, passively immunizing the otherwise immunodeficient neonate with a full complement of
its mother’s protective Abs. In its path to the fetus, IgG
traverses two cellular monolayers of the human terminal villus,
the syncytiotrophoblast and the endothelium. The transporter of
the syncytiotrophoblast is almost certainly the nonclassical
MHC class I protein FcRn4 that binds IgG with high affinity and
is driven by a transcellular endosomal pH gradient (1– 4). Presumably, then, IgG negotiates the endothelium by moving along
a constitutive endocytic pathway, down an IgG concentration
gradient. Whether villus endothelial Fc␥RIIb2 participates in
IgG conveyance across the endothelium is not yet clear (1).
Because of constraints to studying placental transport in human,
we have moved to the mouse where many of the essential cellular
and molecular elements of the human transport pathway appear to
*Department of Internal Medicine, †College of Pharmacy, ‡Center for Biostatistics,
and §Department of Physiology and Cell Biology, The Ohio State University, Columbus, OH 43210
Received for publication September 30, 2008. Accepted for publication December
8, 2008.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported in part by Grants HD38764, CA88053, and AI57530 from
the National Institutes of Health.
2
These authors contributed equally.
3
Address correspondence and reprint requests to Dr. Clark L. Anderson, Department
of Internal Medicine, The Ohio State University, 012K Davis Heart and Lung Research Institute, 473 West 12th Avenue, Columbus, OH 43210. E-mail address:
[email protected]
4
Abbreviations used in this paper: FcRn, Fc receptor (neonatal); placenta, mouse
chorioallantoic placenta or the human placenta; yolk sac, mouse yolk sac placenta;
CAV1, caveolin 1; DAPI, 4⬘,6-diamidino-2-phenylindole; DIC, differential interference contrast.
Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.0803247
be represented. However, placental anatomy is different between
the two species. While all human placental functions are accomplished by a single chorioallantoic placenta, the mouse divides
placental functions between two distinctive organs, both called
placentas. One is a labyrinthine chorioallantoic placenta not too
dissimilar from the human placenta. We refer to this organ as the
“placenta.” The second placenta in mouse is a yolk sac placenta,
which we call simply the yolk sac. In human, the yolk sac is
present only very early in gestation. Based on analogy with the rat
and with other mammals such as rabbit, the yolk sac (but not the
placenta) of the mouse is the organ exclusively responsible for IgG
maternofetal transport during gestation (2). It is perhaps responsible for the gestational transport of other macromolecules as well
(5).
As in the human placenta, two cellular layers of the yolk sac
separate maternal from fetal circulations; namely, the endodermal monolayer, which expresses FcRn in the rat and by inference in the mouse (6 – 8); and the cells of the vitelline vasculature, which are thought to transfer IgG constitutively by
transcellular vesicular transport (9). It can be inferred further
that the yolk sac expresses FcRn because newborn FcRn⫺/⫺
mice are IgG deficient (10).
Herein we report our testing of several predictions of the
hypothesis that mouse yolk sac FcRn is the maternal-fetal transporter of IgG. We first established that FcRn⫺/⫺ fetuses taken
at gestational day 19 –20 have virtually zero plasma IgG concentrations, indicating that FcRn is required for transport of
⬎99.5% of IgG from mother to fetus. We next found FcRn to
be expressed exclusively in the endoderm of the mouse yolk sac
and not in yolk sac mesenchyme or vasculature or in the placenta. We further found that while yolk sac endoderm appears
to take up IgG constitutively, transfer of endodermal IgG to the
fetal circulation requires FcRn. Our results suggest that the IgG
transport mechanisms in the mouse yolk sac and the human
placenta are similar and parallel.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
In adults, the nonclassical MHC class I molecule, FcRn, binds both IgG and albumin and rescues both from a degradative
fate, endowing both proteins with high plasma concentrations. FcRn also transports IgG from mother to young during
gestation. Anticipating that a detailed understanding of gestational IgG transport in the mouse may give us a useful model
to understand FcRn function in the human placenta, we have studied FcRn in the mouse yolk sac placenta in detail.
Analyzing day 19 –20 fetuses of the three FcRn genotypes resulting from matings of FcRnⴙ/ⴚ parents, we found that FcRnⴚ/ⴚ
fetuses showed negligible IgG concentrations (1.5 ␮g/ml), whereas IgG concentrations in FcRnⴙ/ⴚ fetuses were about a half
(176 ␮g/ml) that of FcRnⴙ/ⴙ fetuses (336 ␮g/ml), indicating that FcRn is responsible for virtually all IgG transport from
mother to fetus. Immunofluorescence and immunoblotting studies indicated that FcRn is expressed in the endoderm of the
yolk sac placenta but not in other cells of the yolk sac placenta or in the chorioallantoic placenta. IgG was found in
the endoderm of both FcRnⴙ/ⴙ and FcRnⴚ/ⴚ yolk sac placentas and in the mesenchyme of FcRnⴙ/ⴙ but was missing from
the mesenchyme of FcRnⴚ/ⴚ yolk sac placentas, indicating that IgG enters the endoderm constitutively but is moved out
of the endoderm by FcRn. The similarities of these results to human placental FcRn expression and function are
striking. The Journal of Immunology, 2009, 182: 2583–2589.
2584
Materials and Methods
Reagents
Armenian hamster anti-mouse FcRn serum and preimmune serum were
provided by Dr. A. Shaw (Washington University, St. Louis, MO) (11). A
chicken anti-caveolin 1 Ab (anti-CAV1) was produced in house (12, 13).
Alexa dye-conjugated goat anti-mouse IgG and goat anti-chicken IgG and
Prolong anti-fade mounting media were purchased from Invitrogen; FITCconjugated goat anti-hamster IgG from Jackson Immunoresearch Laboratories; and 4⬘,6-diamidino-2-phenylindole (DAPI) from Sigma-Aldrich. A
rabbit anti-rat FcRn Ab raised against the rat FcRn H chain separated by
gel electrophoresis was a gift from Dr. P. Bjorkman (California Institute of
Technology, Pasadena, CA). This antiserum in our hands immunoblots a
single band of the appropriate mobility from supernatants of cells secreting
recombinant soluble forms of rat, human, and mouse FcRn and from mouse
tissue lysates (data not shown).
Animals
Breeders of the FcRn-␣-chain-knockout strain (B6.129X1/SvJFcgrtTm1Dcr;
FcRn⫺/⫺) and its wild-type strain (C57BL/6J; FcRn⫹/⫹) were obtained
from Dr. D. C. Roopenian of The Jackson Laboratory (Bar Harbor, ME)
(14). The Ohio State University Institutional Animal Care and Use Committee approved all animal studies.
Mice heterozygous for FcRn (FcRn⫹/⫺) were produced by mating wildtype (FcRn⫹/⫹) and knockout (FcRn⫺/⫺) strains. Subsequently, heterozygous female mice were mated with heterozygous male mice, both 5– 6 mo
old, to produce fetuses in litter sizes of 5–10 with three different receptor
genotypes: FcRn⫹/⫹, FcRn⫹/⫺, and FcRn⫺/⫺.
Harvesting fetuses
At the gestational age of day 19 –20, pregnant heterozygous female mice
that had been mated to heterozygous male mice were anesthetized under
light isoflurane, bled of ⬃30 ␮l from the retroorbital plexus, and subjected
to Cesarean section to obtain near-term conceptuses. Fetal blood (⬃10 ␮l)
was taken by cardiac puncture with 28-gauge needle-attached 0.5-ml insulin syringes from the individual fetuses. Sera were harvested and stored
at ⫺80°C for determination of concentrations of IgG, albumin, and transferrin. Fetal tail tips and maternal livers were obtained for genotyping. The
placenta/yolk sac units were fixed in 4% freshly prepared paraformaldehyde (Sigma-Aldrich) in PBS for 2 h at room temperature. The tissue was
then washed extensively in PBS, incubated overnight at 4°C in 20% sucrose in PBS, embedded, and frozen at ⫺80°C.
Genotyping
The DNA, isolated from fetal tail tips and maternal liver, was genotyped
for FcRn by PCR using primers designed to distinguish the targeted (FcRn
null) allele from the FcRn wild-type allele by their different sizes on 2.2%
agarose gels. Template DNA (100 ng) was used for a 50-␮l PCR containing 200 ␮M dNTPs (Invitrogen), 1⫻ PCR buffer with 1.5 mM MgCl2
(Qiagen), and 1 unit of Taq for each reaction.
We modified a published method (14). Briefly, the oligo pair o393FGGGATGCC ACTGCCCTG and o394R-CGAGCCT GAGATTGTC
AAGTGTATT gave a 248-bp wild-type allele, whereas the targeting vector
o395F-GGAATT CCCAGT GAAGGGC vs o394R gave a 378-bp targeted
allele. Thermocycler conditions were 95°C for 10 min, 40 cycles of 94°C
for 45 s, 60°C for 1 min, 72°C for 1 min, and 72°C for 5 min. The PCR
products were resolved on 2.2% agarose gel, and genotypes were determined based on the band size, FcRn⫹/⫹ giving a band size of 248 bp, the
FcRn⫺/⫺ allele a 378 bp band, and a FcRn⫹/⫺ giving both.
Determination of serum protein concentrations
Steady-state serum concentrations of IgG, albumin, and transferrin were
determined by the sandwich ELISA as described earlier (15–17). Individual serum concentrations of IgG or albumin in the fetuses were normalized
to their transferrin concentrations, which did not differ among strains (not
shown).
Immunoblot analysis
The relative molecular mass of FcRn from different sources was analyzed
by immunoblotting using two different anti-FcRn Abs. Yolk sac and placenta tissues were obtained from C-section (see above). Intestine pieces,
consisting of duodenum and small intestine from 9- to 10-day-old FcRn⫹/⫹
or FcRn⫺/⫺ pups were removed under isoflurane anesthesia and quickly
frozen and stored at ⫺80°C until use. The freshly isolated placenta tissues
were cut into small pieces with a razor blade and homogenized with tissue
homogenizer in the presence of tissue lysis buffer (25 mM HEPES, 20 mM
Na4P2O7䡠10 H2O, 100 mM NaF, 4 mM EDTA, 2 mM Na3VO4, 1% Triton
X-100, 0.34 mg/ml PMSF, 0.01 mg/ml aprotinin, and 0.01 mg/ml leupeptin). Yolk sac samples were lysed directly without homogenization. Lysates were incubated on ice for 30 min and centrifuged at 23,000 ⫻ g for
10 min at 4°C. Postnuclear lysates containing equal amounts of protein
were boiled in SDS sample buffer for 5 min, separated by 10% SDS-PAGE,
and immunoblotted with rabbit anti-rat FcRn serum. For immunoblot using
hamster-anti-mouse FcRn serum, intestine pieces were thawed and homogenized in 60 mM octylglucoside buffer pH 7.4 (18). Proteins were then
resolved on 8 –16% gradient gel, transferred to nitrocellulose membranes
(Hybond ECL; Amersham Biosciences), blocked in 5% nonfat milk at
room temperature for 1 h, and probed with primary Abs and relevant control Abs overnight on a rocker at 4°C. Membranes were washed and incubated in FITC- or peroxidase-conjugated secondary Abs for 1 h at room
temperature. After several washes, FITC signal was collected with a Molecular Imager PharosFX systems (Bio-Rad) instrument and peroxidase
conjugates were imaged by chemiluminescence.
Deglycosylation with N-glycosidase F
N-linked oligosaccharides were removed from FcRn protein using the enzyme N-glycosidase F (Boehringer Mannheim). Briefly, protein normalized cell lysates were denatured with 5% SDS for 10 min at 100°C. Denatured lysates were incubated with enzyme diluent alone or with
N-glycosidase F (Boehringer Mannheim) in reaction buffer (pH 7.5; 0.5 M
sodium phosphate and 10% Nonidet P-40) at 37°C for 1 h. The enzyme
reaction was stopped by adding SDS sample buffer (60 mM Tris pH 6.8,
2.3% SDS, 10% glycerol, and 0.01% bromphenol blue). The deglycosylated FcRn protein was then analyzed by SDS-PAGE and immunoblotting
as above.
Immunofluorescence localization of FcRn
Sections of placenta/yolk sac units were cut at 5-␮m thickness in a Shandon cryostat (Global Medical Instrumentation) and were collected on Superfrost slides (Fisher Scientific). For immunolocalization of FcRn, the
sections were hydrated, blocked in 5% nonfat dry milk, and incubated
overnight with 1/500 hamster anti-mouse FcRn serum plus 1/400 chicken
anti-CAV1 (13, 19) in blocking buffer at 4°C. Antiserum binding was
localized indirectly by FITC-labeled goat anti-hamster IgG secondary Ab
and chicken anti-CAV1 binding by Alexa 594 dye-conjugated goat IgG
anti-chicken IgY secondary Ab. Nuclei were stained with DAPI for 10 min
and mounted under coverslips in ProLong. Controls included sections incubated with hamster preimmune serum plus secondary Ab as well as
hamster anti-mouse IgG secondary Ab alone and goat anti-chicken IgY
secondary Ab alone. For IgG distribution studies, hydrated and blocked
sections were incubated with 1/400 anti-CAV1 at 4°C overnight in a moist
chamber, washed the next morning, and then these same sections were
incubated with 1/200 Alexa 594 dye-conjugated goat anti-chicken IgY and
Alexa 488 dye-conjugated goat IgG anti-mouse IgG at room temperature
for 1 h. The slides were washed and stained with DAPI as described above.
Fluorescence and differential interference contrast (DIC) images were collected with a Nikon Optiphot microscope and a Photometrics Cool Snap fx
camera (Roper Scientific) and captured with the MetaMorph image analysis system (Universal Imaging/Molecular Devices). All images were collected within the linear response range of the camera.
Quantitative analysis of images
Noting visually that the yolk sac endoderm but not placenta showed positive FcRn immunofluorescence in FcRn⫹/⫹ but not FcRn⫺/⫺ strains, we
quantified the distribution of FcRn. Using the image analysis system, we
placed a square outline of 225 pixels (15 ⫻ 15) over the apical and basal
portions of every fourth endodermal cell, moving longitudinally from base
to tip on one side of each symmetrical villus and then back to base on the
other side and scored each enclosed image for fluorescence intensity. The
same outline was placed on mesenchyma and mesothelium areas, and the included fluorescence intensities were measured. In all, we examined basal
and apical portions of ⬃1200 endodermal areas, ⬃650 mesenchymal areas,
and ⬃200 mesothelium areas in each strain (a total of 2050 imaged areas).
Images (2050) were collected from FcRn⫹/⫹ (n ⫽ 3; ⬃95 villi) and from
FcRn⫺/⫺ (n ⫽ 3; ⬃95 villi) animals. We then averaged the fluorescence
intensity measurements for all three cellular layers for each strain and
plotted intensity per unit area as a representation of FcRn distribution.
To quantify the distribution and relative abundance of mouse IgG identified by immunofluorescence, we placed a square outline of 2000 pixels
(50 ⫻ 40), approximating the size of an average endodermal cell over
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Breeding
FcRn-MEDIATED MATERNOFETAL IgG TRANSPORT
The Journal of Immunology
2585
every fourth endodermal cell, and scored for the fluorescence intensity.
Another smaller outline (225 pixels, 15 ⫻ 15), appropriate to the smaller
size of the mesenchyme, was used to measure fluorescence intensity in the
mesenchyme. We examined ⬃500 endodermal cells and ⬃500 mesenchyme areas in both strains, each representing ⬃75 villi from three animals
from each group. We then averaged the fluorescence intensity for
endoderm or mesenchyme for each strain and plotted the intensity per unit
area representing relative IgG distribution.
Quantitative localization of placental IgG to the trophoblast
giant cell
Statistical analyses
To compare the steady-state concentrations of proteins in the fetuses,
mixed linear models were fitted to assess the fixed effects of fetuses
(wild-type, heterozygous, and knockout fetuses) on the fetal concentrations after controlling maternal concentrations. Differences in protein
concentration among three strains were considered significant at p ⬍
0.05. Analyses were performed using the SAS (version 9.1; SAS Institute). Data were presented as mean ⫾ SD or least-squares means and
95% confidence limits.
Results
FcRn is required for IgG transport to the fetus
To measure the effect of FcRn (expressed in the conceptus) on
maternofetal transmission of IgG, we compared IgG concentrations of day 19 –20 fetuses of the three different FcRn genotypes. Conceptuses of FcRn⫹/⫺ mothers mated with FcRn⫹/⫺
fathers were delivered by Cesarean section on gestational day
19 –20. The fetuses were genotyped for FcRn, and IgG concentrations of cardiac blood were measured. Fig. 1A shows that
while IgG concentrations averaged ⬃300 ␮g/ml in the FcRn⫹/⫹
fetuses, IgG concentrations were meager (1.5 ␮g/ml) in
FcRn⫺/⫺ fetuses ( p ⬍ 10⫺4). Therefore, the lack of FcRn expression in the conceptus results in a dramatic decrease in fetal
IgG concentration (⬎200-fold). FcRn⫹/⫺ fetuses manifested
IgG concentrations about one-half that of the FcRn⫹/⫹ fetuses
( p ⬍ 10⫺4) and significantly higher than FcRn⫺/⫺ fetuses ( p ⬍
10⫺4).
Albumin concentrations are also decreased in FcRn⫺/⫺ fetuses
While it is clear that FcRn binds and prolongs the lifespan of both
IgG and albumin (1), evidence is scant that albumin is transported
by FcRn from mother to fetus (2). Considering that our experimental protocol might lend evidence to the question of albumin
transport, we measured albumin concentrations in these same fetuses. Fig. 1B shows that albumin concentrations in FcRn⫺/⫺
fetuses were significantly lower (32%) than both FcRn⫹/⫹ ( p ⫽
0.0017) and FcRn⫹/⫺ ( p ⫽ 0.0032) fetuses. Albumin concentrations in FcRn⫹/⫹ and FcRn⫹/⫺ fetuses were not different ( p ⫽
0.4432). Serum concentrations of transferrin, which is not a ligand
for FcRn, did not differ among fetuses of the three genotypes (not
shown).
FIGURE 1. FcRn is required for IgG transport to fetus. Serum concentrations, determined by ELISA, of IgG (A) and albumin (B) were
normalized to transferrin levels to minimize interanimal variation. Each
point represents the fetal serum concentration of an individual preterm
fetus plotted against its maternal serum concentration. Mean values
with SD and the number of fetuses in each strain (Nf) and mothers (Nm)
are shown. A, The mean fetal IgG concentration was significantly lower
in FcRn⫺/⫺ fetuses (ⴛ) compared with both FcRn⫹/⫹ (F) and FcRn⫹/⫺
(‚) fetuses. Also, the IgG level in FcRn⫹/⫺ fetuses was about half that
of FcRn⫹/⫹ fetuses. B, Albumin concentrations of FcRn⫺/⫺ fetuses
were also significantly lower than in both FcRn⫹/⫺ and FcRn⫹/⫹
strains.
FcRn is expressed in the endoderm of the mouse yolk sac but
not in the placenta
We assessed the whereabouts of FcRn expression in the yolk
sac and placenta of the mouse using a hamster anti-mouse FcRn
Ab (11). In immunoblotting studies, this Ab recognized a band
of ⬃50 kDa, characteristic of FcRn, in lysates of neonatal intestine from FcRn⫹/⫹ but not FcRn⫺/⫺ pups (Fig. 2A). Immunoblots with hamster preimmune serum gave no such band but
did give the same bands at the buffer front seen on the Fig. 2A
blot (data not shown).
Using this same Ab for immunofluorescence studies on tissue
sections of placenta/yolk sac units, we found that the distribution
of FcRn in the mouse yolk sac was restricted to endodermal cells;
no FcRn was detected in the cells of the mesenchyme, including
the vitelline vasculature (Fig. 2Ba) or in cells of the placenta labyrinth (Fig. 2Bb). Within endodermal cells, FcRn predominated in
the apical and basal regions of these cells (Fig. 2Ba).
Quantifying the FcRn fluorescence signal by analyzing over
4000 data points from several animals, we document that the
average intensity per unit area of endoderm is significantly
higher (⬃4000 times) in FcRn⫹/⫹ than FcRn⫺/⫺ yolk sacs ( p ⬍
0.05) (Fig. 2C). The distribution of FcRn in FcRn⫹/⫹ yolk sac
was restricted to the endodermal layers; only background levels
of fluorescence were detected in the mesenchyme or mesothelial
layers ( p ⬍ 0.05).
FcRn from mouse yolk sac is more heavily glycosylated than
intestinal FcRn
We measured the mobility by SDS-PAGE and immunoblotting of
mouse FcRn in mouse placenta, yolk sac, and neonatal intestine
(Fig. 3). A predominant band approximating the mobility of FcRn,
⬃45–55 kDa, appears in samples of yolk sac and neonatal intestine
from FcRn⫹/⫹ mice but not in placenta and intestine from
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
To measure the fraction of the trophoblast layer associated with IgG, we
superimposed the IgG green on the maternal border of the trophoblast that
was morphologically identifiable by DIC, and we superimposed the red
CAV1 signal on the luminal edge of the placental fetal endothelium, also
morphologically identifiable by DIC. Along a line drawn from maternal
blood space lumen to fetal capillary lumen, we then quantified three distances: namely, the distance occupied by green (Dgreen), the distance occupied by red (Dred), and the black area in between (Dblack). Choosing
tissue areas free of ambiguity, folding, or tearing and using sections from
three animals, we made 30 measurements, calculating means ⫾ SD (␮m),
respectively, of 23 ⫾ 4, 14 ⫾ 2, and 64 ⫾ 6. Then, the ratio of Dgreen/
(Dgreen ⫹ Dblack) represents the fraction of the total thickness of trophoblast
that was occupied by IgG.
2586
FcRn-MEDIATED MATERNOFETAL IgG TRANSPORT
FcRn⫺/⫺ mice (Fig. 3A). However, FcRn from yolk sac migrates
more slowly than the FcRn of neonatal intestine, suggesting perhaps a difference in glycosylation between the two. Removing Nlinked oligosaccharides with N-glycosidase F, we found that FcRn
from both sources, yolk sac and intestine, moved with near-identical mobilities (Fig. 3B).
Endodermal FcRn is responsible for exocytosis but not uptake
of IgG
Attempting to localize the IgG transport defect between mother
and FcRn⫺/⫺ fetuses, we compared, by localizing and quantifying,
mouse IgG in sections of yolk sac from FcRn⫹/⫹ and FcRn⫺/⫺
conceptuses by immunofluorescence microscopy (Fig. 4). We
found IgG to be present in the endoderm of both strains (Fig. 4, A
and B, heavy arrows), indicating that cellular uptake of IgG is not
blocked in the absence of FcRn. However, while IgG is abundant
in the mesenchyma of FcRn⫹/⫹ yolk sac, it is almost completely
absent in the mesenchymal region of the FcRn⫺/⫺ yolk sac (Fig. 4,
A and B, thin arrows).
Quantifying the fluorescence data from images of 150 villi from
a total of six animals using MetaMorph software, we compared the
intensity of the IgG signal in endodermal cells (Fig. 4E) and the
mesenchyme (Fig. 4F) of both FcRn⫹/⫹ and FcRn⫺/⫺ yolk sac
(Fig. 4E). While there was a trend for the FcRn⫺/⫺ endoderm to
contain more IgG than the FcRn⫹/⫹ endoderm (Fig. 4E), this difference was not statistically significant ( p ⬎ 0.45). However, it is
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 2. FcRn distribution in yolk sac and placenta. A, Immunoblot of FcRn protein from mouse intestine. Three different amounts of lysate from
FcRn⫹/⫹ (lanes 1–3) and FcRn⫺/⫺ pups (lanes 4 – 6) were subjected to SDS-PAGE electrophoresis and immunoblotting with hamster anti-mouse FcRn
serum. FcRn is marked with an arrowhead. Replicate samples from three FcRn⫹/⫹ and three FcRn⫺/⫺ animals gave identical results. B, Localization of
FcRn in yolk sac and placenta. Sections of fixed yolk sac (a) and placenta (b) from day 19 –20 FcRn⫹/⫹ conceptuses were stained simultaneously with
hamster anti-mFcRn antiserum (green) and with chicken anti-CAV1 Ab (red) for endothelium. FcRn was restricted to the apical and basal portions of
endodermal cells (thick arrows) in the yolk sac (a). No FcRn signal can be found in yolk sac mesenchyme (thin arrows), in endothelium marked by CAV1
labeling (arrowheads), or in mesothelium (broken arrows), also marked by CAV1 labeling (a). The inset of a, containing a higher magnified view of an
outlined area of yolk sac, illustrates the dense green labeling in apical and basal areas of endodermal cells (arrowheads). The lack of green in b indicates
that no FcRn is expressed in the placental labyrinth. Arrowheads point to the endothelium marked by CAV1 labeling (red). The inset of b shows a higher
magnified view of a CAV1-labeled area. c and d, Black and white renditions of the corresponding DIC- and DAPI-stained images of a and b merged to
illustrate the morphology of the tissue. Hamster preimmune serum showed nothing more intense than the nonspecific staining seen with the secondary Ab
alone (data not shown). Bar ⫽ 20 ␮m. C, Quantification of immunofluorescence images. Yolk sac sections from FcRn⫹/⫹ and FcRn⫺/⫺ animals were
stained with hamster anti-mFcRn antiserum and FITC-goat anti-hamster IgG. Fluorescence images were collected from representative sites of the three
major cellular areas of the yolk sac; i.e., endodermal (ED), mesenchyme (MC), and mesothelium (MS) areas. The images were quantified, the numbers
averaged for each area (mean fluorescence intensity), and the areas compared by normogram. We noted a very faint patchy green fluorescence in the ED
of FcRn⫺/⫺ yolk sacs far below the intensity of the FcRn⫹/⫹ yolk sacs (mean fluorescence intensities of 6000 ⫾ 6 in FcRn⫹/⫹ ED, whereas mean
fluorescence intensities of MC and MS areas were ⬍30). This patchiness may indicate cross-reactivity of the Ab with another component of the ED, perhaps
similar to that described by the originator of the hamster anti-mFcRn Ab (11).
The Journal of Immunology
2587
FIGURE 3. A, Immunoblot of FcRn protein in the
mouse yolk sac. Tissue lysates of equal protein concentration from samples of yolk sac (YS), placenta (Plac),
and neonatal intestine (Int) from FcRn⫹/⫹ and FcRn⫺/⫺
conceptuses were analyzed by SDS-PAGE and immunoblot (IB) with rabbit anti-rat FcRn Ab. B, Mobility of
deglycosylated FcRn from yolk sac and neonatal intestine. The lysates were equilibrated with the appropriate
buffer, incubated with or without N-glycosidase F(⫹/⫺
N-gly F), and analyzed by immunoblotting with the rabbit anti-rat FcRn Ab. Numbers are MW markers in kDa.
IgG cannot be detected in the placental endothelium
In a fashion identical to our assessment of IgG distribution in the
yolk sac, we examined the placenta at gestational day 19 –20 for
the whereabouts of IgG by two-color immunofluorescence identifying both mIgG (green) and endothelium (red) using the CAV1
marker, superimposing both color images on DIC and DAPI images for histological correlation. We found the maternal blood
spaces of the FcRn⫹/⫹ placental labyrinth to be lined by patchy
deposits of IgG (Fig. 5A, green). Superimposition of both red and
green images with DIC plus DAPI images (Fig. 5B) of 30 trophoblast cross-sections from three placentas indicates that the green
color is associated with the luminal one-fourth (26%) of the distance from the lumen of the maternal blood space to the lumen
defined by the fetal endothelium. This result would suggest that
IgG is associated with the trophoblast giant cell alone and likely
not the underlying two layers of syncytiotrophoblast (I and II) (20,
21). The placental endothelium, identified by CAV1 (red), appears
virtually free of IgG (Fig. 5, A and C). Quantifying our visual
FIGURE 4. IgG distribution in mouse visceral yolk sac. Photomicrographs of representative tissue sections dual labeled with (green) Alexa 488 goat
IgG anti-mouse IgG and chicken anti-CAV1 Ab followed by (red) Alexa 594 goat IgG anti-chicken IgY secondary Ab. A, FcRn⫹/⫹ yolk sac in which green
signal is seen in endodermal cells (thick arrows) as well in mesenchymal areas (thin arrows). Anti-CAV1 labels both the endothelium in mesenchyme and
the mesothelium (broken arrows). The inset, which contains a higher magnification view of the small outlined area of yolk sac, shows labeling of IgG in
mesenchyme, indicated by arrowheads. B, FcRn⫺/⫺ yolk sac showing bright green staining of the endodermal cells (thick arrow) but no IgG staining in
the mesenchyme (thin arrow). The mesothelium (red) is marked with broken arrows. The inset, containing a higher magnification view of the outlined area
of yolk sac, shows with arrowheads the lack of IgG in the mesenchyme. For orientation of arrows with histologic features, the center two panels show
combined DIC and DAPI images of yolk sac from FcRn⫹/⫹ (C) and FcRn⫺/⫺ (D) conceptuses. Bar ⫽ 20 ␮m. E and F, Quantification of immunofluorescence images. Multiple images such as those in A and B were quantified as in Fig. 2C. The relative morphologic distribution and the fluorescence intensity
of IgG staining in yolk sac cells were quantified and compared (expressed as mean fluorescence intensity ⫾ SD) from recorded fluorescence intensity
measurements of both endodermal (E) and mesenchymal (F) areas of tissue sections. Matched yolk sacs from three FcRn⫹/⫹ and three FcRn⫺/⫺ conceptuses
were studied simultaneously. F, The average intensity per unit area for FcRn⫺/⫺ is 0.0193 with SD of 0.023.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
easily demonstrable that IgG was present in the mesenchymal layers of FcRn⫹/⫹ yolk sac but was undetectable in the same mesenchymal layers of FcRn⫺/⫺ yolk sac ( p ⬍ 0.003) (Fig. 4F).
2588
FcRn-MEDIATED MATERNOFETAL IgG TRANSPORT
impression by measuring red and green signal from three FcRn⫹/⫹
placentas, we document that IgG is not present in the endothelium
of the placenta (Fig. 5C).
Discussion
We have shown here in mice that FcRn⫺/⫺ fetuses at the end of
gestation have virtually no IgG in their blood; that FcRn is expressed in the yolk sac endoderm and not in the chorioallantoic
placenta; that in the course of moving through the yolk sac, maternal IgG appears to enter the endoderm in the absence of FcRn
but fails to reach the mesenchyme and the vitelline vasculature;
and that abundant IgG can be found associated with the trophoblast
giant cells of the placenta, but none appears to traverse to the
endothelium of fetal placenta. We elaborate several conclusions
from these core findings:
These findings in all affirm our hypothesis that the yolk sac
FcRn is the maternal-fetal IgG transporter in the mouse. The virtual absence of IgG in blood of the FcRn⫺/⫺ fetuses at gestational
day 19 –20 (Fig. 1A) would seem strong evidence that FcRn is the
essential transporter, but we realize that two physiological processes govern the IgG concentration in the fetus, so some discussion is necessary. We must consider both the rate of IgG transport
from mother across the yolk sac and the rate of IgG degradation
within the fetus. If we assume that the rate of IgG degradation in
the fetus approximates that of the adult where the ratio of plasma
IgG concentrations in FcRn⫹/⫹ to FcRn⫺/⫺ strains is 5:1–10:1
(15, 22), then the fetal IgG concentration ratio of 200:1 for
FcRn⫹/⫹ to FcRn⫺/⫺ in the present study indicates that much
more IgG is present in the FcRn⫹/⫹ fetuses, 20- to 40-fold more,
than can be accounted for by degradation alone. This calculation
demands that we conclude that FcRn is near completely necessary
for IgG transport across the yolk sac of the mouse. We can conclude more generally, therefore, that FcRn mediates IgG concentrations throughout the lifespan of the individual at three different
anatomic sites; protecting IgG from degradation systemically
throughout life; transporting IgG across the gut of the neonate;
and, as we affirm here, transporting IgG in utero.
We found also that FcRn⫹/⫺ fetuses manifest 48% lower IgG
concentration than FcRn⫹/⫹fetuses, indicating that two FcRn alleles are required for normal IgG concentrations in the fetus,
whether from FcRn-transport across the yolk sac or from FcRnrecycling within the fetus. This finding is at odds with the conclu-
sion of others that IgG concentrations of ␤2m⫹/⫹ and ␤2m⫹/⫺
newborns are indistinguishable (10), but it is consistent with our
evaluation of ␤2m⫹/⫺ humans who manifested 75% normal IgG
concentrations attributed to a partial lack of FcRn (23). We have
reflected that the wide range of normal values of IgG concentrations in humans may in part result from FcRn heterozygotes harboring an otherwise silent single null mutation (23).
The obvious distinction in fetal IgG concentrations among the
fetuses of three different genotypes from the same mother (Fig. 1)
indicates that mixing of fetal circulations between adjacent fetal/
placental units is nil, allowing us directly to compare IgG concentrations among different genotypes in the same litter. Furthermore,
it should be noted that we have minimized maternal effects on fetal
IgG concentrations by assessing fetuses of all three genotypes that
are the products of heterozygous, genetically identical parents.
Our immunofluorescence microscopy and immunoblotting experiments affirm our hypothesis that FcRn is expressed in the yolk
sac but not in the placenta (Figs. 2 and 3). The immunofluorescence experiments were made possible by a new hamster antimFcRn antiserum that recognizes only a single band of the appropriate mobility in samples from both neonatal intestine and yolk
sac of FcRn⫹/⫹ animals (Fig. 2). This result, along with an understanding of how the antiserum was made and first characterized
(11), would indicate that the antiserum is specific for FcRn. The
immunoblotting studies utilized two different polyclonal Abs, optimizing immunologic specificity of detection, and resolved a difference in relative molecular mass of FcRn from different tissues
by showing that the deglycosylated molecules moved identically
(Fig. 3). The functional significance of the difference in molecular
size of mouse FcRn, depending on its tissue source, is unknown.
We note that others have made the same observation in the rat (8).
Thus, we affirm what has long been implied that the yolk sac,
and not the chorioallantoic placenta, is the IgG transporting organ
of the mouse. With such a critical function being delegated to the
yolk sac, we find it remarkable that no systematic study has been
done of the division of labor between the two mouse placentas.
Within the yolk sac we find FcRn expressed exclusively in the
endodermal cells and not in the cells of the mesenchyme and
vitelline vasculature (Fig. 2). Thus, the pattern of FcRn expression
is fully consistent with observations in the rat (8). Further, these
data underscore similarities between human and mouse. FcRn
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 5. Lack of IgG in labyrinth endothelium. A, A two-color immunfluorescence photomicrograph of the labyrinth region of a chorioallantoic
placenta showing the presence of mIgG (green) and illustrating the dearth of maternal IgG in the fetal endothelium as defined by CAV1 labeling (red). The
maternal blood spaces are defined by green IgG (broken arrows). Arrowheads point to trophoblast layers I and II, which lack any green fluorescence signal.
Asterisks are placed on the characteristically large nuclei of sinusoidal trophoblast giant cells. Bar ⫽ 30 ␮m. B, Superimposed DAPI and DIC images are
provided for orientation. C, Assessing 27 mm of linear endothelium from three FcRn⫹/⫹ placentas (⬃250 images) by two-color immunofluorescence, we
quantified the amount of IgG associated with endothelial cells (green with red). The histograms show that compared with the amount of red signal
identifying endothelium (arbitrarily set at 100%), only a trace of IgG (green) colocalizes with red (⬍⬍1%) ( p ⫽ 3 ⫻ 10⫺6).
The Journal of Immunology
Acknowledgments
The authors are grateful to Dr. Derry C. Roopenian for FcRn-deficient
mouse strain breeders, to Dr. C. L. Bronson for helpful comments, to Sarah
McAllister for technical assistance, to Dr. Pamela Bjorkman for the rabbit
anti-rat FcRn H chain Ab, to Dr. Andrey Shaw for the hamster anti-mFcRn
antiserum, and to Dr. Richard Burry and the staff of The Ohio State University Campus Microscopy and Imaging Facility for helpful information
relating to imaging.
Disclosures
The authors have no financial conflict of interest.
References
1. Anderson, C. L., C. Chaudhury, J. Kim, C. L. Bronson, M. A. Wani, and
S. Mohanty. 2006. Perspective—FcRn transports albumin: relevance to immunology and medicine. Trends Immunol. 27: 343–348.
2. Brambell, F. W. R. 1970. The Transmission of Passive Immunity from Mother to
Young. North Holland Publishing Company, Amsterdam.
3. Burmeister, W. P., L. N. Gastinel, N. E. Simister, M. L. Blum, and
P. J. Bjorkman. 1994. Crystal structure at 2.2 A resolution of the MHC-related
neonatal Fc receptor. Nature 372: 336 –343.
4. Ghetie, V., and E. S. Ward. 2000. Multiple roles for the major histocompatibility
complex class I-related receptor FcRn. Annu. Rev. Immunol. 18: 739 –766.
5. Christensen, E. I., and P. J. Verroust. 2002. Megalin and cubilin, role in proximal
tubule function and during development. Pediatr. Nephrol. 17: 993–999.
6. Carpenter, S. J., and V. H. Ferm. 1969. Uptake and storage of Thorotrast by the
rodent yolk sac placenta: an electron microscopic study. Am. J. Anat. 125:
429 – 456.
7. Jollie, W. P. 1990. Development, morphology, and function of the yolk-sac placenta of laboratory rodents. Teratology 41: 361–381.
8. Roberts, D. M., M. Guenthert, and R. Rodewald. 1990. Isolation and characterization of the Fc receptor from the fetal yolk sac of the rat. J. Cell Biol. 111:
1867–1876.
9. Lin, C. T. 1980. Immunoelectron microscopy localization of immunoglobulin G
in human placenta. J. Histochem. Cytochem. 28: 339 –346.
10. Israel, E. J., D. F. Wilsker, K. C. Hayes, D. Schoenfeld, and N. E. Simister. 1996.
Increased clearance of IgG in mice that lack ␤2-microglobulin: possible protective role of FcRn. Immunology 89: 573–578.
11. Akilesh, S., G. J. Christianson, D. C. Roopenian, and A. S. Shaw. 2007. Neonatal
FcR expression in bone marrow-derived cells functions to protect serum IgG
from catabolism. J. Immunol. 179: 4580 – 4588.
12. Lyden, T. W., C. L. Anderson, and J. M. Robinson. 2002. The endothelium but
not the syncytiotrophoblast of human placenta expresses caveolae. Placenta 23:
640 – 652.
13. Takizawa, T., C. L. Anderson, and J. M. Robinson. 2005. A novel Fc␥R-defined,
IgG-containing organelle in placental endothelium. J. Immunol. 175: 2331–2339.
14. Roopenian, D. C., G. J. Christianson, T. J. Sproule, A. C. Brown, S. Akilesh,
N. Jung, S. Petkova, L. Avanessian, E. Y. Choi, D. J. Shaffer, et al. 2003. The
MHC class I-like IgG receptor controls perinatal IgG transport, IgG homeostasis,
and fate of IgG-Fc-coupled drugs. J. Immunol. 170: 3528 –3533.
15. Chaudhury, C., S. Mehnaz, J. M. Robinson, W. L. Hayton, D. K. Pearl,
D. C. Roopenian, and C. L. Anderson. 2003. The major histocompatibility complex-related Fc receptor for IgG (FcRn) binds albumin and prolongs its lifespan.
J. Exp. Med. 197: 315–322.
16. Chaudhury, C., J. Kim, S. Mehnaz, M. A. Wani, T. M. Oberyszyn, C. L. Bronson,
S. Mohanty, W. L. Hayton, J. M. Robinson, and C. L. Anderson. 2006. Accelerated transferrin degradation in HFE-deficient mice is associated with increased
transferrin saturation. J. Nutr. 136: 2993–2998.
17. Kim, J., C. L. Bronson, W. L. Hayton, M. D. Radmacher, D. C. Roopenian,
J. M. Robinson, and C. L. Anderson. 2006. Albumin turnover: FcRn-mediated
recycling saves as much albumin from degradation as the liver produces.
Am. J. Physiol. 290: G352–G360.
18. Leach, J. L., D. D. Sedmak, J. M. Osborne, B. Rahill, M. D. Lairmore, and
C. L. Anderson. 1996. Isolation from human placenta of the IgG transporter,
FcRn, and localization to the syncytiotrophoblast: implications for maternal-fetal
antibody transport. J. Immunol. 157: 3317–3322.
19. Lyden, T. W., J. M. Robinson, S. Tridandapani, J. L. Teillaud, S. A. Garber,
J. M. Osborne, J. Frey, P. Budde, and C. L. Anderson. 2001. The Fc receptor for
IgG expressed in the villus endothelium of human placenta is Fc␥RIIb2. J. Immunol. 166: 3882–3889.
20. Rawn, S. M., and J. C. Cross. 2008. The evolution, regulation, and function of
placenta-specific genes. Annu. Rev. Cell Dev. Biol. 24: 159 –181.
21. Takata, K., K. Fujikura, and B. C. Shin. 1997. Ultrastructure of the rodent placental labyrinth: a site of barrier and transport. J. Reprod. Dev. 43: 13–24.
22. Junghans, R. P., and C. L. Anderson. 1996. The protection receptor for IgG
catabolism is the beta2-microglobulin-containing neonatal intestinal transport receptor. Proc. Natl. Acad. Sci. USA 93: 5512–5516.
23. Wani, M. A., L. D. Haynes, J. Kim, C. L. Bronson, C. Chaudhury, S. Mohanty,
T. A. Waldmann, J. M. Robinson, and C. L. Anderson. 2006. Familial hypercatabolic hypoproteinemia caused by deficiency of the neonatal Fc receptor,
FcRn, due to a mutant ␤2-microglobulin gene. Proc. Natl. Acad. Sci. USA 103:
5084 –5089.
24. Simister, N. E., C. M. Story, H. L. Chen, and J. S. Hunt. 1996. An IgG-transporting Fc receptor expressed in the syncytiotrophoblast of human placenta. Eur.
J. Immunol. 26: 1527–1531.
25. Parr, E. L., and M. B. Parr. 1985. Localization of immunoglobulins in the mouse
uterus, embryo, and placenta during the second half of pregnancy. J. Reprod.
Immunol. 8: 153–171.
26. Martin, M. G., S. V. Wu, and J. H. Walsh. 1997. Ontogenetic development and
distribution of antibody transport and Fc receptor mRNA expression in rat intestine. Dig. Dis. Sci. 42: 1062–1069.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
expression in the endoderm of the mouse yolk sac and in the syncytiotrophoblast of the human placenta (18, 24) extends the functional parallel of these two organs in these two species and suggests that study of the mouse yolk sac may be a useful model for
dissection of the mechanism of IgG transport in the human
placenta.
We noted that whereas the endoderm of both FcRn⫹/⫹ and
FcRn⫺/⫺ yolk sacs contained about equal amounts of IgG, by contrast, IgG was found in the mesenchyme of only the FcRn⫹/⫹ and
not the FcRn⫺/⫺ yolk sacs (Fig. 4). Thus, in the absence of FcRn,
IgG appears to be stalled in transit from mother to fetus. IgG
clearly enters the endoderm regardless of the presence of FcRn, but
it seems unable to exit the endoderm and move to the mesenchyme
in the absence of endodermal FcRn; i.e., in the FcRn⫺/⫺ yolk sac.
Failing to reach the mesenchyme, IgG would also fail to enter the
fetal circulation, as we measured in the data of Fig. 1. This interpretation leads us to affirm a hypothesis originally formulated by
Brambell, the father of this field, in the 1950s (2). Brambell would
have said that IgG is endocytosed nonspecifically and constitutively at the apical surface of endoderm from yolk sacs of both
strains. Within the intracellular vesicles of endoderm, IgG encounters and binds to FcRn which, in complex, move to the basolateral
surface of the cell where IgG is exocytosed into the mesenchyme
tissue and eventually into the fetal vitelline circulation. Further
studies of this endoderm will show us the detailed mechanism of
the transport process.
Our observations about the chorioallantoic placenta, that FcRn
is not expressed and that maternal IgG seems not to move any
closer to the fetal circulation than the trophoblast giant cell (Fig.
5), are compatible with the general notion that in rodents and rabbits the yolk sac placenta but not the chorioallantoic placenta is the
site of IgG transport to the fetus (2, 25).
Our finding that albumin concentrations are 32% lower in
FcRn⫺/⫺ than FcRn⫹/⫹ fetuses (Fig. 1B) is noteworthy in light of
our recent discovery that FcRn binds albumin and governs its catabolic rate (1). The lower albumin concentrations in FcRn⫺/⫺ fetuses likely result from some combination of two possible effects.
First, if FcRn is expressed systemically in the albumin-producing
fetus and protects albumin from degradation, then the FcRn⫺/⫺
fetus would show a diminished albumin concentration compared
with the FcRn⫹/⫹ fetus as was seen in adult mice (1). The only
data relevant to fetal FcRn expression is a single study showing no
FcRn mRNA in the fetal proximal gut; other tissues were not sampled (26). Second, since IgG appears to be transported by FcRn to
the fetus (argument above) and since FcRn binds albumin, then
FcRn may be transporting maternal albumin to the fetal circulation. We have no direct data relevant to albumin transport by
FcRn. At the present time, we are unable to distinguish between
FcRn-mediated albumin recycling and transport in the conceptus.
2589