Cell Stem Cell, volume 11
Derivation Conditions Impact X-Inactivation Status
in Female Human Induced Pluripotent Stem Cells
Kiichiro Tomoda, Kazutoshi Takahashi, Karen Leung, Aki Okada, Megumi Narita, N.
Alice Yamada, Kirsten E. Eilertson, Peter Tsang, Shiro Baba, Mark P. White, Salma Sami,
Deepak Srivastava, Bruce R. Conklin, Barbara Panning, and Shinya Yamanaka
Arbitary days after starting the count
Teratoma from H9-reporter
Teratoma from H9r-3F-2
H9r 3F-2H9r 4F-1H9r 4F-2
hES cell lineshiPS cell lines
Arbitary days after starting the count
Arbitary days after starting the count
XY hES H13B
XX hES H9 reporter
XX hiPS 3S-5F-4
XY hiPS BJ-4F-1
XX hiPS K-3F-1XX hiPS K-3F-2
XX hiPS BJ-Mix
XX hES ESI03
XY hES H14
XX hES H7XX hES H9
Teratoma from K-3F-2
Teratoma from 3S-5F-4
3S-5F-4 p24K-3F-2 p22
Passage number after mixing
% of GFP positive cells
Tomoda et al. Figure S1
SUPPLEMENTAL FIGURE LEGENDS
Figure S1. Characterization of hiPSC Lines
Related to Figure 1.
(A) iPS-like colony numbers counted 1 month after introducing the indicated
combination of retroviral vectors into differentiated H9-reporter ESCs. 3F indicates
OCT4, SOX2 and KLF4. 4F indicates OCT4, SOX2, KLF4 and cMYC. No iPS-like
colonies appeared in a culture infected with a control viral vector DsRed, suggesting
that reprogramming factors were required for iPS-like colony formation from
differentiated H9-reporter cells. Red and blue bars show iPS-like colony numbers
appeared in independent two experiments.
(B) Southern blotting by KLF4 probe detects both endogenous (human and mouse) and
exogenous KLF4. All H9 ES-derived iPSC lines have different KLF4 viral integration
events, suggesting that all clones picked up are different clones. The results also
indicate that viral integration events are required for the colony formation from the
differentiated H9-reporter cells.
(C and J) Colony morphology of indicated cell lines cultured on SNL feeders.
(D, K and L) Growth curves of the indicated cell lines. Cell numbers were enumerated
when a given cell line was passed, and total cell numbers were calculated based on
(E and N) Total (endogenous + exogenous; blue) and endogenous (red) expression
level of OCT4. The expression level in the indicated cell lines was measured by
RT-qPCR. Exogenous OCT4 expression can be calculated by subtracting endogenous
OCT4 from total OCT4 expression.
(F and O) Images show differentiated cells from indicated pluripotent stem cell lines
through embryoid body formation in vitro. Expression of endoderm (alpha-fetoprotein;
AFP), mesoderm (smooth muscle actin; SMA and Vimentin) and ectoderm (beta-III
tubline; TubIII) markers were shown in green. Genomic DNA was co-stained with DAPI
shown in blue.
(G and P) Teratoma formation assays with the indicated cell lines. The pictures of neural
rosettes (ectoderm, top left), gut-like structures (endoderm, top right) and cartilage
(mesoderm, bottom left) are shown.
(H and Q) Karyotyping for indicated cell lines at indicated passage numbers. All lines
have normal chromosome number at the indicated passage numbers (46,XX) although
unidentified genetic material (arrowhead) was inserted at 1p34.3 in all H9-derived cell
lines (H), including differentiated H9-reporter but not the parental H9 ESC line. Since the
H9-reporter ESC and all H9-derived iPSC lines have the same genetic material, this
material does not strongly affect X-inactivation status in the hESC and hiPSC lines.
(I) Heat map of relative expression levels of genes involved in pluripotency and
differentiation (Jaenisch and Young, 2008) among the cell lines. Genes expressed at an
undetectable level are shown in gray.
(M) A hierarchical clustering of male hiPSC lines, female hiPSC lines, male hESC lines
and female hESC lines, fibroblasts (HDF and BJ) and differentiated H9-reporter lines
(Diff cells) with expression levels of all expressed genes in the given cell line.
Expression values of each gene among the cell lines were normalized by 75%
percentile shift normalization.
(R) Percentages of GFP-positive cells after culturing mixed population of Xa/Xi cells
(GFP negative) and Xa/Xa cells (GFP positive) on SNL feeders. At indicated passage
numbers after the mixing, % of GFP-positive cells in an SSEA3-positive population was
analyzed by flow cytometer. The appearance of Xa/Xa hiPSCs after several passages
on SNLs might be attributable to selection, rather than reactivation during culture.
However, Xa/Xa cell lines did not proliferate faster than Xa/Xi cell lines (Figure S1D, K
and L) and did not out compete Xa/Xi cell lines in co-cultures on SNLs (Figure S1R).
Figure S2. X/A Ratio Is a Strong Predictor of X-Inactivation Status
Related to Figure 3 and 4.
(A) Graph showing relationship between X/A ratios (x-axis) and probability (y-axis) of
X-reactivation. We used logistic regression on known cases (deposited Xa/Xi and Xa/Xa
cell lines) to examine if X/A expression ratio is a good predictor of X-inactivation status,
Xa/Xi or Xa/Xa. The model indicates there is a strong correspondence between
expression ratio and probability of X-reactivation. The coefficient for X/A ratio in the
logistic regression was 15.2, p-value 0.014, The solid red line indicates the fitted
probabilities, and the dashed lines indicate the 95% confidence interval.
(B) Graph showing the model can predict X-inactivation status with 100% accuracy
using our pluripotent stem cell lines as test cases. We plotted X/A ratios of our lines in
which X-inactivation status was examined using other conventional methods (methods
used in Figure 2) along the model shown in A. Using the probability cutoff of 0.5, eg
P(Xa/Xa) >0.5 to assign a predicted state of Xa/Xa, and a predicted state of Xa/Xi if
P(Xa/Xa) <0.5, the model predicted X-inactivation status in our cell lines with 100%
accuracy. X/A ratios from Xa/Xi lines are shown as black dots and from Xa/Xa are
shown as red dots.
(C) Recalculated model. We recalculated the model with deposited Xa/Xi and Xa/Xa
lines and our Xa/Xi and Xa/Xa lines, the increase in sample size allowed us to obtain
more precise estimate of the relationship between the expression ratio and
X-inactivation status (these estimates were then used in Figures 3A, 4B, 4C and S3D).
The coefficient for X/A ratio in the logistic regression was 30.36=7, p-value = 0.008.
According to the model, we expect at log2 ratio of 0.144 the P(Xa/Xa) = 0.05, at log2
ratio of 0.241 the P(Xa/Xa) = 0.5, at log2 ratio of 0.338 the P(Xa/Xa) = 0.95, and at
log2 ratio of 0.389 the P(Xa/Xa) = 0.99.
Figure S3. Early Passage Xa/Xi hiPSC Lines Generated on Non-SNLs Are
Converted into Xa/Xa Lines after Several Passages on SNLs
Related to Figure 4.
(A and B) Heat maps of relative expression levels of the X-linked genes under indicated
conditions which details are described in Figure 4A. Male hiPSC and Xa/Xa hiPSC lines
are used as controls.
(C) X-linked gene expression ratios plotted on the X using the hiPSC line 297C1
generated on hFibs (see details at Figure 4I).
(D) X/A ratios from four additional hiPSC lines initially generated on MEFs. Four early
female hiPSC lines (at p4 or p7) independently generated on MEFs were obtained and
cultured them on MEFs or on SNLs. RNA was extracted after the cell lines were
passaged on the indicated feeders for > p15. X/A ratios were calculated from DNA
microarray data obtained using the RNA samples.
(E) Graph showing the proportion of cells with 0, 1, or 2 sites of PGK1 nascent mRNA
accumulation in indicated conditions. One cell line (hiPS14) that exhibited the highest
X/A ratio on SNLs (D) was further analyzed by FISH and showed that majority of cells
on SNLs but not on MEFs have two PGK1 foci.
Table S1. All Cell Lines Used in This Study
Place Source Factors In
3S-5F-4 G F, 44 y OSKM + + N SNLs 0.42(p15)
K-3F-1 G F, 36 y OSK + ND N SNLs -0.01(p5)/0.5(p22)
K-3F-2 G F, 36 y OSK + + N SNLs 0.02(p5)/0.53(p22)
BJ-4F-1 G M, Neo OSKM + ND ND SNLs 0.01
BJ-Mix G M, Neo OSKM + ND ND SNLs -0.05
H9r-3F-2 G F, H9 ES OSK + +
H9r-4F-1 G F, H9 ES OSKM + ND
H9r-4F-2 G F, H9 ES OSKM + ND
3M4F1 G F, 44 y OSKM + ND N MEFs 0.38(p14)/0.24(p16)
3M4F3 G F, 44 y OSKM + ND N MEFs 0.42(p16)/-0.03(p22)
201B2* K F, 36 y OSKM + + N SNLs -0.11(p5)/0.42(p68)
201B6* K F, 36 y OSKM + + N SNLs 0.01(p5)/0.31(p26)/
201B7* K F, 36 y OSKM + + N SNLs 0.36(p18)/0.41(p47)
253G1* K F, 36 y OSK + + N SNLs 0.13(p5)/0.38(p26)
253G4* K F, 36 y OSK + + N SNLs 0.25(p18)/0.23(32)
TIG107-3F1 K F, 81 y OSK + + N SNLs 0.08(p5)/0.41(p21)/
TIG120-3F7 K F, 6 y OSK + + N SNLs -0.11(p5)/0.23(p23)
TIG107-4F1 K F, 6 y OSKM + + N SNLs 0.12(p5)/0.55(p16)
TIG118-4F1 K F, 12 y OSKM + + N SNLs 0.26(p10)/0.37(p20)
TIG120-4F1 K F, 6 y OSKM + + N SNLs 0.08(p5)/0.33(p18)
TIG108-4F3 K F, 40 y OSKM + + N SNLs 0.35(p24)
1488-4F1 K F, 36 y OSKM + ND N SNLs 0.15(p27)
1503-4F1 K F, 73 y OSKM + ND N SNLs 0.54(p28)
923S1 K F, 36 y Epi ND ND N SNLs 0.40(p15)
923S2 K F, 36 y Epi ND ND
923S3 K F, 36 y Epi ND ND N SNLs 0.66(p15)
923M2 K F, 36 y Epi ND ND N MEFs 0.08(p15)
923M3 K F, 36 y Epi ND ND
297C1* K F, 36 y OSKM + + N hFibs 0(p7)/0.13(p19)
297C2* K F, 36 y OSKM + ND ND hFibs -0.03(p7)/-0.1(p19)
297F1* K M, 56 y OSKM + + N hFibs 0.05(p14)/-0.04(p17)
hiPS21 UCLA F OSKM ND ND ND MEFs 0.17(p30)/0.15(p36)
hiPS24 UCLA F OSKM ND ND ND MEFs -0.07(p17)
hiPS6 UCLA F OSKM ND ND ND MEFs 0.12(p15)
hiPS12 UCLA F OSKM ND ND ND MEFs -0.05(p16)
hiPS14 UCLA F OSKM ND ND ND MEFs -0.07(p19)
hiPS28 UCLA F OSKM ND ND ND MEFs 0.02(p15)
H9*** WiCell F - + + N SNLs 0.18
WiCell F - + +
H7*** WiCell F - + ND ND SNLs 0.15
ESI03** WiCell F - + ND ND SNLs 0.03
H1 WiCell M - + ND ND SNLs -0.01
H13B WiCell M - + ND ND SNLs -0.03
H14 WiCell M - + ND ND SNLs -0.07
G F, H9 ES - - ND
Cell line names (Line), where given hiPSC or hESC lines were generated or obtained
(Place), age and gender for each individual from which donor fibroblasts were obtained
(Source), reprogramming factors used (Factor), in vitro and in vivo differentiation
potentials (In vitro and In vivo), results of karyotyping (Karyotype), types of feeder cell
used for deriving hiPSC lines or culturing hESC lines during this study (Feeder) and
expression ratios between X chromosome and autosomes at indicated passage number
if indicated in parentheses (X/A). In the first column, “Diff H9-reporter’ indicates
differentiated cells from H9-reporter ES cells from which H9r iPSC lines were generated.
In the second column, G indicates Gladstone, K indicates Kyoto University, UCLA
indicates University of California, Los Angeles and WiCell indicates the National Stem
Cell Bank. In the third column, “F” indicates female, and “M” indicates male. “Neo”
indicates neonatal. In the fourth column, O, S, K and M indicates OCT4, SOX2, KLF4
and cMYC, respectively. Epi indicates episomal vector reprogramming. In the fifth and
sixth column, + indicates that a given cell line can be differentiated into cell types of the
three germ layers in vitro and in vivo, respectively. “ND” indicates “not determined”. In
the seventh column, “N” indicates normal karyotype, and “ND” indicates “not
determined”. In the eighth column, SNLs indicate leukemia inhibitory factor
(LIF)-expressing immortalized mouse embryonic fibroblasts, MEFs indicate mouse
primary embryonic fibroblasts, and hFibs indicate autologus human primary fibroblasts.
In the ninth column, numbers in parentheses indicates passage numbers when X/A
ratios were analyzed. * Cell lines 201B2, 201B6, 201B7, 253G1, 253G4, 297A1, 297C1,
297C2 and 297F1 were as reported (Nakagawa et al., 2008; Takahashi et al., 2009;
Takahashi et al., 2007). ** Class II (Xa/Xi with XIST expression) and *** Class III (Xa/Xi
without XIST expression) hESC lines. hESC lines were categorized based on
X-inactivation status in undifferentiated state (Silva et al., 2008). No Class I (Xa/Xa)
hESC lines in this study.
Table S2 Primers Used in This Study
Sequence (5’ to 3’) Description
AGCGAACCAGTATCGAGAAC Total human
ATCACCCTCTCAGGAAAGTCTAAG WDR44 SNP
TAAGGAAACCACAGTGTAGCTTTGACAGA FRMPD4 SNP
GGGCCTATTCCTCTCTACCTTTAAGGACATTTA TSPAN6 SNP
meWDR44 F ATTTAGGGATTTTAATTAGAGATTT Bisulfite
meWDR44 R CCTAACTCTACAATTCCAAAAAAAC
EXTENDED EXPERIMENTAL PROCEDURES
OCT4-Reporter Plasmid Construction
A BamH1 fragment, including an EGFP-IRES-puromycin-resistance gene cassette from
pCAG-EGFP-IP vector (from Dr. Niwa, RIKEN CDB, Japan), was subcloned into a Bgl2
site of pCRXL-human OCT4 promoter plasmid (Takahashi et al., 2007).
Establishment of H9-Reporter Cell Lines
To establish the H9-reporter ESC line, H9 ESCs were cultured under feeder-free
conditions (Takahashi et al., 2007). We introduced the OCT4-reporter plasmid into the
H9 ES line with Nucleofection (Amaxa). Briefly, H9 ESCs were treated with 0.05%
trypsin at 37ºC for 1 min and pipeted to form a mixture of small clumps and single cells.
The OCT4-reporter plasmid (3 g) was introduced into 2.0X106 H9 ESCs with 100 l of
solution V, according to the manufacturer’s instructions. After nucleofection, the cells
were seeded on two Matrigel-coated wells of a 6-well plate with MEF cell-conditioned
medium supplemented with 4 ng/ml of bFGF and 10 M of ROCK inhibitor. At 3 days
after nucleofection, 325 ng/ml of puromycin was added into the medium to select
desired clones. Two weeks after selection, we picked five colonies and evaluated them
with expression pattern of GFP and OCT4 by immunostaining and flow cytometry. The
clone with the highest OCT4-GFP expression was selected for generating the
H9-reporter ESC line.
Generation of the Differentiated H9-Reporter Cells
To generate differentiated H9-reporter cells, 8-day-old embryoid bodies (EBs) were
seeded on a gelatin-coated plate and cultured in EB medium (Takahashi et al., 2007) for
1 month. Differentiated cells were re-seeded after treatment with 0.05% of trypsin and
removal of large clumps of cells with a Cell Strainer (BD Bioscience). Cells were grown
under standard fibroblast culture conditions for another 12 days and then used to derive
H9r iPSC lines.
In Vitro Differentiation
EB differentiation was carried out using standard protocols (Takahashi et al., 2007).
The details of the protocol for endothelial cell differentiation are prepared for publication.
Briefly, 6-day-old EBs were stained using a specific antibody against VEGFR (R&D
systems) conjugated with phycoerythrin (PE), and then VEGFR+ cells were purified
using a magnetic activated cell-sorting system against PE (Miltenyi Biotec). The purified
cells were cultured for additional 2 weeks, and then endothelial cells were purified again
using a specific antibody against CD31 (BD Bioscience).
RNA Isolation and PCR
Total RNA was purified with Trizol reagent (Invitrogen) and treated with Turbo DNA-free
kit (Ambion) to remove genomic DNA contamination. Before extracting total RNA from
hiPSC lines on human fibroblasts, we removed the human fibroblasts by sorting
SSEA4-positive cells. Total RNA (1 gwas used for reverse transcription reaction with
SuperScriptIII (Invitrogen) and random hexamer primers, according to the
manufacturer’s instructions. PCR was performed with rTaq (NEB). TaqMan probes and
TaqMan Gene Expression Master Mix (Applied Biosystems) were used for examining
expression levels of GAPDH, WDR44, XIST and NANOG. All PCR primers except for
TaqMan probes are listed in Table S2.
For immunostaining, cells that were cultured on a Lab-Tek Chamber Slide (Nunc) were
fixed with 4% paraformaldehyde in PBS for 10 min at room temperature. After washing
with PBS, the cells were treated with 0.1% TritonX-100 in PBS for 10 min at room
temperature. The cells were blocked with 5% bovine serum albumin (Sigma Aldrich) for
30 min and then incubated with primary antibodies for 1 h. Primary antibodies against
H3K27me3 (1:100, Abcam), NANOG (1:100, Abcam) and RNA polymerase II (1:50, H5,
Covance) were used. After the incubation with the primary antibodies, cells were
washed with PBS twice and incubated with secondary antibodies for 30 min. Secondary
antibodies used were Alexa488-conjugated goat anti-mouse IgM (1:500, Invitrogen),
Alexa488-conjugated goat anti-rabbit IgG (1:500, Invitrogen) and Alexa555-conjugated
goat anti-mouse IgG (1:500, Invitrogen). After the incubation with secondary antibodies,
cells were washed with PBS twice and mounted on a glass slide with a mounting
medium (DAPI II, Abbott).
Allele Specific Expression (SNPs sequencing)
We employed primers for WDR44, FRMPD4 and TSPAN6 mRNA listed in Table S2.
Genomic DNA (1 g) was treated with bisulfite and purified using an EpiTect kit (Qiagen),
according to the manufacturer’s recommendations. The promoter region of the human
WDR44 was amplified by PCR with the primer set listed in Table S2. The PCR products
were subcloned into pCR2.1-TOPO. More than 20 clones of each sample were verified
by sequencing with the M13 universal primers.
X Chromosome Probe Design, Synthesis, and Labeling
X chromosome painting probes were generated from every other 1-mb interval through
in silico selection from non-repeat masked genomic sequences. Selection criteria
included melting temperature matching and homology filtering. In silico–selected
sequences were then sent for massively parallel de novo synthesis by Agilent
Technologies, and chemically synthesized oligonucleotide libraries were generated.
These libraries were amplified by PCR, followed by the introduction of fluorescent
labeling. Labeling was conducted using Kreatech’s Platinum Bright: Nucleic Acid
Labeling Kit. USP spectrum orange with an emission of 565 was used as the fluorescent
dye. The specificity and robustness for the probes were confirmed in male metaphase
cells (data not shown).
X painting was performed first. hiPSCs and hESCs that were cultured on a Matrigel
coated coverslip were fixed with MeOH for 5 min and then washed with PBS twice. The
cells were fixed again with 1% PFA for 5 min, washed with PBS three times for 5 min,
and with 2X SCC three times for 5 min each. Genomic DNA was depurinated with 1.5 N
HCl for 1.5 min and then washed with PBS twice. Co-denaturation and slide
hybridization was performed using ThermoBright (Abbott) or PTC-100 Programmable
Thermal Controller (MJ RESEARCH). The cells were hybridized with 10 l hybridization
solution that contains 7 l of CEP Hybridization Buffer (Abbott), 1 l of the painting
probes and 2 l of dH2O at 78° C for 5 min to denature and then at 37° C for 4–6 h. The
cells were washed with pre-heated Wash buffer 1 (0.4X SSC and 0.3% NP-40 in dH2O)
at 73° C for 2 min, followed by Wash buffer 2 (2XSSC and 0.1% NP-40 in dH2O) at
room temperature for 1 min. The coverslip was mounted on a slide glass, sealed with
mounting medium (Abotte) and stored at -20°C.
Immunostaining was performed with a specific antibody against RNA polymerase II.
The coverslips were carefully removed from the glass slides, and the slides were
washed with PBS three times for 5 min each. The remaining procedure for the
immunostaining was performed as described in the immunostaining section.
For RNA FISH, we used XIST and PGK1 probes (Sterallis FISH probes) obtained from
Biosearch Technologies. Each probe is a mixture of 48 oligonucleotides, and the probes
were designed using the company’s website. Cells were harvested using Accutase
(Milipore) to make a single-cell suspension and then fixed with 4% PFA for 10 min. The
fixed cells were stored in 70% EtOH at -20°C at least for overnight. We cytospun the
fixed cells on a glass slide with a Cyotospin (Shandon) at 800 rpm for 3 min. The FISH
experiments were performed using the company’s protocol
(http://www.singlemoleculefish.com/protocols.html), but we used mounting medium
(Abotte) to mount the stained samples. FlSH signals were visualized using a CFI-Plan
Apo TIRF100x oil objective mounted on a Nikon Ti 2000 inverted microscope equipped
with a spinning disk confocal unit (CSU-22; Yokogawa). Optical sections were acquired
with an EMCCD camera (Hamamatsu). Images were recorded using the open source
MicroManager v1.2 software (UCSF) and processed using ImageJ (NIH) to analyze and
count the cell numbers. We only counted bright nuclear PGK1 dots.
We calculated X/A ratios (Deng et al., 2011; Kharchenko et al., 2011; Lin et al., 2011;
Nguyen and Disteche, 2006; Yildirim et al., 2012) with raw data of selected genes from
a microarray for each cell line at each passage number with Genespring GX10 without
any transformations. For the calculation, we selected X-linked genes of which
expression levels are at top 60% among several cell lines. We also selected autosomal
genes with same criteria. X/A ratios were calculated using median of expression values
of the top 60% of the selected genes. We chose this calculation method because this
method allowed us to discriminate between reported Xa/Xa and Xa/Xi lines the most
clearly we had tried. We used cell lines in which X-inactivation status was characterized
as controls (Hanna et al., 2010; Lengner et al., 2010; Tchieu et al., 2010).
Statistic Analyses for X-Reactivation
We used logistic regression and a training data set with deposited Xa/Xi and Xa/Xa lines
to examine the relationship between X/A expression ratio and X-reactivation status.
The model was fit using R (R Development Core Team, 2011) and the “brglm” package
(Ioannis Kosmidis, 2007). We assumed that X-reactivation status is a Bernoulli
response with status indicated by a 1 if X-reactivation occurs, and a 0 if X-reactivation
fails to occur. X-reactivation is assumed to occur with some probability, p. In other
??/?? ~ ???????.
We modeled this probability of reactivation, p, using logistic regression and the log2
ratio of X/A expression, or “Ratio” as a predictor.
1 ? ?? ? __ ? ____????
We found the log2 ratio of X/A expression to be statistically significant predictor of the
odds of reactivation, __ ? =15.2, p-value 0.0144.
We then tested the model’s ability to predict X-inactivation status on data from our cell
lines that had been experimentally verified as Xa/Xi or Xa/Xa. We found that using a
predicted probability cutoff of ?̂ ? 0.5, we predicted the X-inactivation status with
100% accuracy. We then combined our training and test data sets and re-fit the model
to obtain a more precise estimate of the correlation between the X/A ratio and the
log-odds of X-reactivation, the resulting estimate of the Ratio coefficient was __ ? = 30.37,
p-value 0.008. We tested a few additional covariates in the model, such as array type
and feeder, but these did not explain any further variation in the data, and were not
R Development Core Team (2011). R: A language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/.
Ioannis Kosmidis (2007). brglm: Bias reduction in binary-response
Deng, X., Hiatt, J.B., Nguyen, D.K., Ercan, S., Sturgill, D., Hillier, L.W.,
Schlesinger, F., Davis, C.A., Reinke, V.J., Gingeras, T.R., et al. (2011).
Evidence for compensatory upregulation of expressed X-linked genes in
mammals, Caenorhabditis elegans and Drosophila melanogaster. Nature
genetics 43, 1179-1185.
Jaenisch, R., and Young, R. (2008). Stem cells, the molecular circuitry of
pluripotency and nuclear reprogramming. Cell 132, 567-582.
Kharchenko, P.V., Xi, R., and Park, P.J. (2011). Evidence for dosage
compensation between the X chromosome and autosomes in mammals.
Nature genetics 43, 1167-1169; author reply 1171-1162.
Lin, H., Halsall, J.A., Antczak, P., O'Neill, L.P., Falciani, F., and Turner, B.M.
(2011). Relative overexpression of X-linked genes in mouse embryonic stem
cells is consistent with Ohno's hypothesis. Nature genetics 43, 1169-1170;
author reply 1171-1162.
Yildirim, E., Sadreyev, R.I., Pinter, S.F., and Lee, J.T. (2012). X-chromosome
hyperactivation in mammals via nonlinear relationships between chromatin
states and transcription. Nat Struct Mol Biol 19, 56-61.