Methods
Ethics statement
This work was approved by the Medicine Ethics Committee of The First
People’s Hospital of Yunnan Province (2017LS[K]NO.035). All donated
embryos in this study were surplus frozen embryos from couples who
already had at least one healthy baby after in vitro fertilization clinic
treatment. The informed consent process for embryo donation com-
plied with International Society for Stem Cell Research (ISSCR) ‘Guide-
lines for Stem Cell Research and Clinical Translation (2016)’ and ‘Ethical
Guidelines for Human Embryonic Stem Cell Research (2003)’ jointly
issued by the Ministry of Science and Technology and the Ministry of
Health of People’s Republic of China. The Medicine Ethics Committee
of The First People’s Hospital of Yunnan Province is composed of nine
members, including lawyers, scientists and clinicians with relevant
expertise. The Committee evaluated the scientific merit and ethical
justification of this study and conducted a full review of the donations
and use of these samples. All donor couples signed informed consents
for voluntary donations of surplus embryos for human embryo devel-
opment study at the Department of Reproductive Medicine in the First
People’s Hospital of Yunnan Province. No financial inducements were
offered for the donations. In the process, couples were informed that
their embryos would be used to study the developmental mechanisms
of human embryos and that their donation would not affect their in vitro
fertilization cycle. The culture of all embryos was terminated at day 14
after fertilization or upon the appearance of primitive streak anlage.
Embryo thawing and zona pellucida removal
Before embryo thawing, human embryo culture medium G-2 (10132,
Vitrolife) was equilibrated in a 4-well plate (176740, Nunc) overnight.
First, 0.5 ml G-2 medium was added into the well and 0.25 ml mineral oil
(10029, Vitrolife) was used to cover the G-2 medium. Human blastocysts
(5–6 d.p.f.) were thawed using Kitazato Thawing Media Kit (VT102,
Kitazato Corporation) by following manufacturer’s instructions.
After culture in drops of the equilibrated G-2 medium for 4 h, embryos
were transferred to acidic Tyrode’s solution (T1788, Sigma-Aldrich) to
remove the zona pellucida. Once the zona pellucida vanished, embryos
were immediately transferred to G-2 medium. Embryos were washed in
G-2 medium twice and then transferred to the in vitro culture medium.
Evaluation of embryo quality
According to the Gardner’s scoring system^40 , thawed blastocysts were
given numerical scores from 1 to 6 based on their expansion degree and
hatching status. The blastocyst with expansion and hatching status
above 3 and with visible inner cell mass above grade B were included in
the study. On the basis of morphologies, healthy embryos had to meet
the two following requirements: obvious expansion during culturingm
and absence of obviously dead or broken (fragmented) cell mass during
development. Otherwise, they were excluded from this study.
In vitro 3D culture of human embryos
Embryos without zona pellucida were cultured in a low attachment
96-well plate (3474, Corning) with 1 embryo and 150 μl blastocyst-cul-
ture medium in each well. The embryo culturing conditions were as
follows: 37.2 °C, 6% CO 2 and saturated humidity. The culture protocol
was summarized in Fig. 1a. First, at 6–8 d.p.f, the culture medium was
modified in vitro culture medium 1 (mIVC1). At 8 d.p.f., 50% of mIVC1
medium was replaced by mIVC2. At 9 d.p.f., embryos were transferred
to new wells in mIVC2 including 10% Matrigel (354234, Corning). Then,
50% culture medium was replaced by new mIVC2 including 10% Matrigel
every day. mIVC1 and mIVC2 were pre-equilibrated in the incubator for
at least 6 h before use. A step-by-step protocol has also been included
at Protocol Exchange^41.
The components of the mIVC1 and mIVC2 system are modifications
of IVC1 and IVC2, established for culturing human attached embryos
on the dish^1 ,^2. mIVC1: advanced DMEM/F12 (12634-010, Thermo Fischer
Scientific) supplemented with 20% (v/v) DFBS (defined fetal bovine
serum) (bs-0003, Biosera), 2 mM l-glutamine (25030, Thermo Fisher
Scientific), 1× ITS-X (51500-056, Thermo Fisher Scientific), 8 nM
β-oestradiol (E8875, Sigma-Aldrich), 200 ng ml−1 progesterone (P0130,
Sigma-Aldrich), 25 μM N-acetyl-l-cysteine (A7250, Sigma-Aldrich),
0.22% (v/v) sodium lactate (L7900,Sigma-Aldrich), 1 mM sodium pyru-
vate (P4562, Sigma-Aldrich) and 10 μM Y27632 (S1049, Selleck). mIVC2:
advanced DMEM/F12 supplemented with 30% (v/v) KOSR (knockout
serum replacement) (A3181501, Thermo Fischer Scientific), 2 mM
l-glutamine, 1× ITS-X, 8 nM β-oestradiol, 200 ng ml−1 progesterone,
25 μM N-acetyl-l-cysteine, 0.22% (v/v) sodium lactate, 1 mM sodium
pyruvate and 10 μM Y27632.Optimizing the Matrigel concentration
To identify optimal Matrigel concentration for culturing human blas-
tocysts under 3D condition, we designed the following four group
experiments: group 1, W/O Matr: human embryos were cultured on low
attachment plate without the Matrigel embedment up to 14 d.p.f.; group
2, 25% Matr: cultured human embryos were embedded in 25% Matrigel
at 9 d.p.f.; group 3, Matr+10% Matr: human embryos were embedded
in 10% Matrigel and transferred into the new well pre-coated by 100%
Matrigel (30 μl) at 9 d.p.f.; group 4, 10% Matr: human embryos were
embedded in 10% Matrigel at 9 d.p.f. Culture medium was mIVC1 on
6–7 d.p.f., then 1:1 mIVC1:IVC2 on 8 d.p.f. and mIVC2 on 8–14 d.p.f. For
Matr+10% Matr, 30 μl original liquid of Matrigel was added into one well
of low attachment 96-well plate at 9 d.p.f. for 30 min in the incubator.
Once the Matrigel solidified, 120 μl of mIVC2 with 10% Matrigel was
added on the surface of the solidified Matrigel. A single embryo was
transferred to a new well after equilibration for 6 h.Frozen section staining and taking photographs
Embryos were fixed by 4% paraformaldehyde, washed three times with
PBS, dehydrated by 15% sucrose for 1 min and embedded in OCT. Embed-
ded embryos were sectioned by a Leica frozen slicer at a thickness of
10–12 μm. Before staining, the slides were washed by PBS to clear OCT,
and permeabilized with 0.2% Triton X-100 for 30 min at room tempera-
ture. After blocking with 3% BSA in PBS for 4 h at room temperature,
sections were incubated with primary antibodies at 4 °C overnight and
then washed three times with 0.05% Tween-20. The following primary
antibodies were used : mouse anti-OCT3/4 (Santa Cruz, SC5279,C-10,
H1612 1:400), rabbit anti-Brachyury (T) (Santa Cruz, SC20109, poly-
clonal, A0616, 1:50), rabbit anti-SOX2 (Millipore, AB5603, polyclonal,
2826070, 1:400), goat anti-SOX17 (R&D Systems, AF1924, polyclonal,
KGA0815042,1:250), rabbit anti-KLF4 (Millipore, 09-821, polyclonal,
2383578, 1:400), rabbit anti-KLF17 (Atlas Antibodies, HPA024629,
polyclonal, C117502, 1:250), rabbit anti-β-catenin (Abcam, AB32572,
E247, GR184212-37, 1:300), mouse anti-E-cadherin (Abcam, AB76055,
M168, GR299147-4, 1:100), rabbit anti-cytokeratin 7 (CK7) (Abcam,
AB181598, EPR17078, GR3214132-10, 1:300), rabbit anti-N-cadherin
(Abcam, AB12221, polyclonal, 40975, 1:200), mouse anti-OTX2 (Santa
Cruz, SC514195, D-8, G0816, 1:100), goat anti-LEFTY1 (R&D Systems,
AF746, polyclonal, CMM0111101, 1:100), goat anti-SOX1 (R&D Systems,
AF3369, polyclonal, XUV0618081, 1:500), rabbit anti-PAX6 (Bioleg-
end, 901301, Poly 19013, b267205, 1:500), rabbit anti-laminin (Sigma-
Aldrich, L9393, polyclonal, 028M4890V, 1:50), rabbit anti-TEAD4 (Atlas
Antibodies, HPA056896, polyclonal, R78063, 1:150), goat anti-FOXA2
(Santa Cruz, SC6554, polyclonal, D1216, 1:100), goat anti-OCT3/4 (Santa
Cruz, SC8628, polyclonal, G3201, 1:250), Mouse anti-hCG (Abcam,
AB9582, 5H4-E2, GR308272-2, 1:100), rabbit anti-PRDM14 (Millipore,
AB4350, polyclonal, 2897240, 1:50), goat anti-TFCP2L1 (R&D Systems,
AF5726, polyclonal, CCUG0115021,1:200), goat anti-NANOG (R&D
Systems, AF1997, polyclonal, KKJ0514091, 1:250), mouse anti-PODXL
(R&D Systems, MAB1658, 222328, JKW0218041, 1:400), mouse anti-
HLA-G (Abcam, AB52455, 4H84, GR251679-19, 1:200), goat anti-GATA6