Thyroid hormone signaling specifies cone subtypes in human retinal organoids

Abstract

Thyroid hormone in color vision development
Cone photoreceptors in the eye enable color vision, responding to different wavelengths of light according to what opsin pigments they express. Eldred et al. studied organoids that recapitulate the development of the human retina and found that differentiation of cone cells into their tuned subtypes was regulated by thyroid hormone. Cones expressing short-wavelength (S) opsin developed first, and cones expressing long- and medium-wavelength (L/M) opsin developed later. The switch toward development of L/M cones depended on thyroid hormone signaling through the nuclear thyroid hormone receptor.
Science , this issue p. eaau6348

Full text

Thyroid hormone signaling specifies cone subtypes in human
retinal organoids
Kiara C. Eldred1, Sarah E. Hadyniak1, Katarzyna A. Hussey1, Boris Brenerman1, Ping-Wu
Zhang2, Xitiz Chamling2, Valentin M. Sluch2, Derek S. Welsbie3, Samer Hattar4, James
Taylor1,5, Karl Wahlin3, Donald J. Zack2,6,7,8, and Robert J. Johnston Jr.1,*
1Department of Biology, Johns Hopkins University, 3400 N. Charles Street, Baltimore, MD 21218,
USA.
2Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA.
3Shiley Eye Institute, University of California, San Diego, La Jolla, CA 92093, USA.
4National Institute of Mental Health, National Institutes of Health, Bethesda, MD 20892, USA.
5Department of Computer Science, Johns Hopkins University, 3400 N. Charles Street, Baltimore,
MD 21218, USA.
6Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine,
Baltimore, MD 21287, USA.
7Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, MD
21287, USA.
8Institute of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, MD
21287, USA.
Abstract
INTRODUCTION: Cone photoreceptors in the human retina enable daytime, color, and high-
acuity vision. The three subtypes of human cones are defined by the visual pigment that they
express: blue-opsin (short wavelength; S), green-opsin (medium wavelength; M), or red-opsin
(long wavelength; L). Mutations that affect opsin expression or function cause various forms of
color blindness and retinal degeneration.
exclusive licensee American Association for the Advancement of Science. http://creativecommons.org/licenses/by/4.0/
*Corresponding author. robertjohnston@jhu.edu.
Author Contributions: K.C.E.: Conception, data acquisition, new reagent contribution, data analysis, and data interpretation; drafted
and revised manuscript. S.E.H.: Data acquisition and data interpretation. K.A.H.: Data acquisition, data analysis, and data
interpretation. B.B.: Data analysis and data interpretation. P.-W.Z.: New reagent contribution. X.C.: New reagent contribution. V.M.S.:
New reagent contribution. D.S.W.: New reagent contribution. S.H.: Data interpretation. J.T.: Data analysis and data interpretation.
K.W.: Data acquisition and new reagent contribution. D.J.Z.: Data acquisition and new reagent contribution. R.J.J.: Conception and
data interpretation; drafted and revised manuscript.
Competing interests: None.
Data and materials availability: RNA-seq data are available on Gene Expression Omnibus, accession no. GSE119320. All other data
and methods are in the supplementary materials. H7 stem cells are available from WiCell under a materials transfer agreement with
WiCell.
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/362/6411/eaau6348/suppl/DC1
HHS Public Access
Author manuscript
Science
. Author manuscript; available in PMC 2019 May 12.
Published in final edited form as:
Science
. 2018 October 12; 362(6411): . doi:10.1126/science.aau6348.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

RATIONALE: Our current understanding of the vertebrate eye has been derived primarily from
the study of model organisms. We studied the human retina to understand the developmental
mechanisms that generate the mosaic of mutually exclusive cone subtypes. Specification of human
cones occurs in a two-step process. First, a decision occurs between S versus L/M cone fates. If
the L/M fate is chosen, a subsequent choice is made between expression of L- or M-opsin. To
determine the mechanism that controls the first decision between S and L/M cone fates, we studied
human retinal organoids derived from stem cells.
RESULTS: We found that human organoids and retinas have similar distributions, gene
expression profiles, and morphologies of cone subtypes. During development, S cones are
specified first, followed by L/M cones. This temporal switch from specification of S cones to
generation of L/M cones is controlled by thyroid hormone (TH) signaling. In retinal organoids that
lacked thyroid hormone receptor β (
Thrβ
), all cones developed into the S subtype.
Thrβ
binds
with high affinity to triiodothyronine (T3), the more active form of TH, to regulate gene
expression. We observed that addition of T3 early during development induced L/M fate in nearly
all cones. Thus, TH signaling through
Thrβ
is necessary and sufficient to induce L/M cone fate
and suppress S fate. TH exists largely in two states: thyroxine (T4), the most abundant circulating
form of TH, and T3, which binds TH receptors with high affinity. We hypothesized that the retina
itself could modulate TH levels to control subtype fates. We found that deiodinase 3 (
DIO3
), an
enzyme that degrades both T3 and T4, was expressed early in organoid and retina development.
Conversely, deiodinase 2 (
DIO2
), an enzyme that converts T4 to active T3, as well as TH carriers
and transporters, were expressed later in development. Temporally dynamic expression of TH-
degrading and -activating proteins supports a model in which the retina itself controls TH levels,
ensuring low TH signaling early to specify S cones and high TH signaling later in development to
produce L/M cones.
CONCLUSION: Studies of model organisms and human epidemiology often generate hypotheses
about human biology that cannot be studied in humans. Organoids provide a system to determine
the mechanisms of human development, enabling direct testing of hypotheses in developing
human tissue. Our studies identify temporal regulation of TH signaling as a mechanism that
controls cone subtype specification in humans. Consistent with our findings, preterm human
infants with low T3 and T4 have an increased incidence of color vision defects. Moreover, our
identification of a mechanism that generates one cone subtype while suppressing the other,
coupled with successful transplantation and incorporation of stem cell-derived photoreceptors in
mice, suggests that the promise of therapies to treat human diseases such as color blindness,
retinitis pigmentosa, and macular degeneration will be achieved in the near future. ■
Graphical Abstract
Eldred et al. Page 2
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Temporally regulated TH signaling specifies cone subtypes. (A) Embryonic stem cell-derived
human retinal organoids [wild type (WT)] generate S and L/M cones. Blue, S-opsin; green, L/M-
opsin. (B) Organoids that lack thyroid hormone receptor β (
Thrβ
KO) generate all S cones. (C)
Early activation of TH signaling (WT + T3) specifies nearly all L/M cones. (D) TH-degrading
enzymes (such as DIO3) expressed early in development lower TH and promote S fate, whereas
TH-activating regulators (such as DIO2) expressed later promote L/M fate.
Summary
The mechanisms underlying specification of neuronal subtypes within the human nervous system
are largely unknown. The blue (S), green (M), and red (L) cones of the retina enable high-acuity
daytime and color vision. To determine the mechanism that controls S versus L/M fates, we
studied the differentiation of human retinal organoids. Organoids and retinas have similar
distributions, expression profiles, and morphologies of cone subtypes. S cones are specified first,
followed by L/M cones, and thyroid hormone signaling controls this temporal switch. Dynamic
expression of thyroid hormone–degrading and –activating proteins within the retina ensures low
signaling early to specify S cones and high signaling late to produce L/M cones. This work
establishes organoids as a model for determining mechanisms of human development with
promising utility for therapeutics and vision repair.
Cone photoreceptors in the human retina enable daytime, color, and high-acuity vision (1).
The three subtypes of human cones are defined by the visual pigment that they express:
blue-opsin (short wavelength; S), green-opsin (medium wavelength; M), or red-opsin (long
wavelength; L) (2). Specification of human cones occurs in a two-step process. First, a
decision occurs between S versus L/M cone fates (Fig. 1A). If the L/M fate is chosen, a
subsequent choice is made between expression of L- or M-opsins (3–6). Mutations that
affect opsin expression or function cause various forms of color blindness and retinal
degeneration (7–9). Great progress has been made in our understanding of the vertebrate eye
through the study of model organisms. However, little is known about the developmental
mechanisms that generate the mosaic of mutually exclusive cone subtypes in the human
retina. We studied the specification of human cone subtypes using human retinal organoids
differentiated from stem cells (Fig. 1, D to K).
Human retinal organoids generate photoreceptors that respond to light (10–14). We found
that human organoids recapitulate the specification of cone subtypes observed in the human
retina, including the temporal generation of S cones followed by L and M cones. Moreover,
Eldred et al. Page 3
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

we found that this regulation is controlled by thyroid hormone signaling, which is necessary
and sufficient to control cone subtype fates through the nuclear hormone receptor thyroid
hormone receptor β (Thr
β
). Expression of thyroid hormone-regulating genes suggests that
retina-intrinsic temporal control of thyroid hormone levels and activity governs cone subtype
specification. Whereas retinal organoids have largely been studied for their promise of
therapeutic applications (15), our work demonstrates that human organoids can also be used
to reveal fundamental mechanisms of human development.
Specification of cone cells in organoids recapitulates development in the
human retina
We compared features of cone subtypes in human organoids with those of adult retinal
tissue. Adult human retinas and organoids at day 200 of differentiation displayed similar
ratios of S to L/M cones as indicated by expression of S- or L/M-opsins (adult, S = 13%,
L/M = 87%; organoid, S = 29%, L/M = 71%) (Fig. 1, B and C, and fig. S1A). The difference
in the ratios is likely due to the immaturity of the organoid at ~6 months compared with the
terminally differentiated adult retina. We examined L/M cones with an antibody that
recognizes both L- and M-opsin proteins because of their extremely high similarity. Both S
and L/M cones expressed the cone-rod-homeobox transcription factor (CRX), a critical
transcription factor for photoreceptor differentiation (Fig. 2, A and E) (16–18), indicating
proper fate specification in organoids. Additionally, cones in organoids and retinas displayed
similar morphologies, with L/M cones that had longer outer segments and wider inner
segments than those of S cones (Fig. 2, B to D and F to H) (19). The outer segments of
cones were shorter in organoids than in adult retinas, which is consistent with postnatal
maturation (Fig. 2, D and H) (20). Thus, cone subtypes in human retinal organoids displayed
distributions, gene expression patterns, and morphologies similar to those of cones of the
human retina.
We next examined the developmental dynamics of cone subtype specification in organoids.
In the human retina, S cones are generated during fetal weeks 11 to 34 (days 77 to 238),
whereas L/M cones are specified later, during fetal weeks 14 to 37 (days 98 to 259) (21, 22).
We tracked the ratios and densities of S and L/M cones in organoids by means of antibody
staining over 360 days of differentiation. Cones expressing S-opsin were first observed at
day 150 (Fig. 2, I, L, and M). The density of S cones leveled off at day 170 (Fig. 2M), at the
time point when cones expressing L/M-opsin began to be observed (Fig. 2, J to M). The
population of L/M cones increased dramatically until day 300 (Fig. 2, K to M), when they
reached a steady-state density. The 20-day difference between S- and L/M-opsin expression
onset in retinal organoids is similar to the 20-day difference observed in the appearance of S
and L/M cones in the fetal retina (21). These observations show a temporal switch from S
cone specification to L/M cone specification during retinal development.
We next conducted RNA sequencing (RNA-seq) through 250 days of induced pluripotent
stem cell (iPSC)-derived organoid development. We found that
S-opsin
RNA was expressed
first at day 111 and leveled off at day 160, whereas
L/M-opsin
RNA was expressed at day
160 and remained steady after day 180, which is consistent with the timeline of
Eldred et al. Page 4
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

photoreceptor maturation in organoids and fetal retinas (Fig. 2N and fig. S1B). Moreover,
CRX
RNA and CRX protein were expressed before opsins in organoids, which is similar to
human development (Fig. 2N and fig. S1, B to G) (23). Thus, human organoids recapitulate
many aspects of the developmental timeline of cone subtype specification observed in
human retinas, providing a model system with which to uncover the mechanisms of these
developmental changes.
Thyroid hormone signaling and the temporal switch between S and L/M fate
specification
Seminal work in mice identified Thr
β
2 as a critical regulator of cone subtype specification:
Thrβ2
mutants display a complete loss of M-opsin expression and a complete gain of S-
opsin expression in cone photoreceptors (
24–26
). Similar roles for
Thrβ2
have been
characterized in other organisms with highly divergent cone patterning (27–29).
Additionally, rare human mutations in
Thrβ2
are reported to alter color perception, which is
indicative of a change in the S-to-L/M cone ratio (30). To directly test the role of
Thrβ2
in
human cone subtype specification, we used CRISPR/Cas9 in human embryonic stem cells
(ESCs) to generate a homozygous mutation that resulted in early translational termination in
the first exon of
Thrβ2
(fig. S2A). Surprisingly, organoids derived from these mutant stem
cells displayed no differences in cone subtype ratio from genotypically wild-type organoids
[wild type, S = 62%, L/M = 38%;
Thrβ2
knockout (KO), S = 59%, L/M = 41%;
P
= 0.83].
The S-to-L/M ratio is high for both wild-type controls and
Thrβ2
KO organoids, likely
owing to variability in organoid differentiation. Thus, unlike previous suggestions based on
other species,
Thrβ2
is dispensable for cone subtype specification in humans (Fig. 3, A to
C).
Because
Thrβ2
alone is not required for human cone subtype specification, we reexamined
data from Weiss
et. al
(30) and found that mis-sense mutations in exons 9 and 10 affected
both
Thrβ2
and another isoform of the human
Thrβ
gene,
Thrβ1
(fig. S2A). Thus, we asked
whether
Thrβ1
and
Thrβ2
together are required for cone subtype specification in humans. To
completely ablate Thrβ function (Thr
β
1 and Thr
β
2), we used CRISPR/Cas9 in human ESCs
to delete a shared exon that codes for part of the DNA binding domain of
Thrβ
(fig. S2A).
Thrβ
null mutant retinal organoids displayed a complete conversion of all cones to the S
subtype (wild type, S = 27%, L/M = 73%;
Thrβ
KO, S = 100%, L/M = 0%;
P
< 0.0001)
(Fig. 3, D to E and H). In these mutants, all cones expressed S-opsin and had the S cone
morphology (Fig. 3, I and J). Thus,
Thrβ
is required to activate L/M and to repress S cone
fates in the human retina.
Thrβ binds with high affinity to triiodothyronine (T3), the more active form of thyroid
hormone, to regulate gene expression (31). Depletion or addition of T3 alters the ratios of S
to M cones in rodents (25, 32, 33). Because L/M cones differentiate after S cones, we
hypothesized that T3 acts through Thrβ late in retinal development to induce L/M cone fate
and repress S cone fate. One prediction of this hypothesis is that addition of T3 early in
development will induce L/M fate and repress S fate. To test this model, we added 20 nM T3
to ESC- and iPSC-derived organoids starting from days 20 to 50 and continued until day 200
Eldred et al. Page 5
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

of differentiation. We observed a dramatic conversion of cone cells to L/M fate (wild type, S
= 27%, L/M = 73%; wild type + T3, S = 4%, L/M = 96%;
P
< 0.01) (Fig. 3, F and H, and
fig. S2B). Thus, early addition of T3 is sufficient to induce L/M fate and suppress S fate.
To test whether T3 acts specifically through Thr
β
to control cone subtype specification, we
differentiated
Thrβ
mutant organoids with early T3 addition.
Thrβ
mutation completely
suppressed the effects of T3, generating organoids with only S cones (wild type + T3, S =
4%, L/M = 96%;
Thrβ
KO + T3, S = 100%, L/M = 0%;
P
< 0.0001) (Fig. 3, F to H). We
conclude that T3 acts though Thr
β
to promote L/M cone fate and suppress S cone fate.
We confirmed the regulation of L/M-opsin expression through thyroid hormone signaling in
a retinoblastoma cell line, which expresses L/M-opsin when treated with T3 (fig. S2, C and
D) (34). T3-induced activation of
L/M-opsin
expression was suppressed upon RNA
interference knockdown of
Thrβ
(fig. S2, E and F), which is similar to the suppression
observed in human organoids.
In organoids, early T3 addition not only converted cone cells to L/M fate but also
dramatically increased cone density (Fig. 3, F and K). Moreover, T3 acts specifically
through Thr
β
to control cone density (Fig. 3, G and K). Early T3 addition may increase cone
density by advancing and extending the temporal window of L/M cone generation.
Together, these results demonstrate that T3 signals though Thrβ to promote L/M cone fate
and repress S cone fate in developing human retinal tissue.
Dynamic expression of thyroid hormone–regulating genes during
development
Our data suggest that temporal control of thyroid hormone signaling determines the S-
versus-L/M cone fate decision, in which low signaling early induces S fate and high
signaling late induces L/M fate. Thyroid hormone exists largely in two states: thyroxine
(T4), the most abundant circulating form of thyroid hormone, and T3, which binds thyroid
hormone receptors with high affinity (31, 35). Because the culture medium contains low
amounts of T3 and T4, we hypothesized that the retina itself could modulate and/or generate
thyroid hormone to control subtype fates.
Conversion of T4 to T3 occurs locally in target tissues to induce gene expression responses
(36, 37). Deiodinases—enzymes that modulate the levels of T3 and T4—are expressed in the
retinas of mice, fish, and chickens (29, 38–42). Therefore, we predicted that T3- and T4-
degrading enzymes would be expressed during early human eye development to reduce
thyroid hormone signaling and specify S cones, whereas T3-producing enzymes, carriers,
and transporters would be expressed later in human eye development to increase signaling
and generate L/M cones.
To test these predictions, we examined gene expression across 250 days of organoid
development. The expression patterns of thyroid hormone–regulating genes were grouped
into three classes: changing expression (Fig. 4A), consistent expression (Fig. 4B), or no
expression (Fig. 4C). Deiodinase 3 (
DIO3
), an enzyme that degrades T3 and T4 (36), was
Eldred et al. Page 6
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

expressed at high levels early in organoid development but at low levels later (Fig. 4A).
Conversely, deiodinase 2 (
DIO2
), an enzyme that converts T4 to active T3 (36), was
expressed at low levels early but then dramatically increased over time (Fig. 4A). We
examined RNA-seq data from Hoshino
et. al
(23) and found that developing human retinas
display similar temporal changes in expression of
DIO3
and
DIO2
(fig. S3A). Deiodinase 1
(
DIO1
), which regulates T3 and T4 predominantly in the liver and kidney (43), was not
expressed in organoids or retinas (Fig. 4C and fig. S3C). Thus, the dynamic expression of
Dio3
and
Dio2
supports low thyroid hormone signaling early in development to generate S
cones and high thyroid hormone signaling late to produce L/M cones.
Consistent with a role for high thyroid hormone signaling in the generation of L/M cones
later in development, expression of transthyretin (
TTR
), a thyroid hormone carrier protein,
increased during organoid and retinal development (Fig. 4A and fig. S3A) (23). By contrast,
albumin (
ALB
) and thyroxine-binding globulin (
SERPINA7
), other carrier proteins of T3
and T4, were not expressed in organoids or retinas (Fig. 4C and fig. S3C) (23).
T3 and T4 are transported into cells via membrane transport proteins (44). The T3/T4
transporters
SLC7A5
and
SLC7A8
increased in expression during organoid differentiation
(Fig. 4A). Additionally, two T3/T4 transporters,
SLC3A2
and
SLC16A2,
were expressed at
high and consistent levels throughout organoid development (Fig. 4B). Other T3/T4
transporters (
SLC16A10, SLCO1C1,
and
SLC5A5
) were not expressed in organoids (Fig.
4C), suggesting tissue-specific regulation of T3/T4 uptake. We observed similar expression
patterns of T3/T4 transporters in human retinas (fig. S3, A to C) (23).
We next examined expression of transcriptional activators and repressors that mediate the
response to thyroid hormone. Consistent with
Thrβ
expression in human cones (45),
expression of
Thrβ
in organoids increased with time as cone cells were specified (Fig. 4A).
Expression of thyroid hormone receptor α (
Thrα
) similarly increased with time (Fig. 4A).
Thyroid hormone receptor cofactors, corepressor
NCoR2
and coactivator
MED1,
were
expressed at steady levels during organoid differentiation (Fig. 4B). Similar temporal
expression patterns were observed in human retinas (fig. S3, A and B) (23). Thus, our data
suggest that expression of Thrβ and other transcriptional regulators enables gene regulatory
responses to differential thyroid hormone levels.
A complex pathway controls production of thyroid hormone. Thyrotropin-releasing hormone
(TRH) is produced by the hypothalamus and other neural tissue. TRH stimulates release of
thyroid-stimulating hormone α (CGA) and thyroid-stimulating hormone β (TSHβ) from the
pituitary gland. CGA and TSHβ bind the thyroid-stimulating hormone receptor (TSHR) in
the thyroid gland. T3 and T4 production requires thyroglobulin (TG), the substrate for T3/T4
synthesis, and thyroid peroxidase (TPO), an enzyme that iodinates tyrosine residues in TG
(46).
TRH
was expressed in organoids and retinas, but the other players were not (Fig. 4, A
to C, and fig. S3, A to C) (23, 47, 48), suggesting that the retina itself does not generate
thyroid hormone; rather, it modulates the relative levels of T3 and T4 and expresses TRH to
signal for thyroid hormone production in other tissues.
Eldred et al. Page 7
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Therefore, the temporal expression of thyroid hormone signaling regulators supports our
model that the retina intrinsically controls T3 and T4 levels, ensuring low thyroid hormone
signaling early to promote S fate and high thyroid hormone signaling late to specify L/M
fate (Fig. 4D).
Organoids provide a powerful system with which to determine the mechanisms of human
development. Model organism and epidemiological studies generate important hypotheses
about human biology that are often experimentally intractable. This work shows that
organoids enable direct testing of hypotheses in developing human tissue.
Our studies identify temporal regulation of thyroid hormone signaling as a mechanism that
controls cone subtype specification in humans. Consistent with our findings, preterm human
infants with low T3/T4 have an increased incidence of color vision defects (49–52).
Moreover, our identification of a mechanism that generates one cone subtype while
suppressing the other, coupled with successful transplantation and incorporation of stem
cell-derived photoreceptors in mice (53–56), suggests that the promise of therapies to treat
human diseases such as color blindness, retinitis pigmentosa, and macular degeneration will
be achieved in the near future.
Materials and methods summary
Cell lines
H7 ESC (WA07, WiCell) and episomal-derived EP1.1 iPSC lines were used for
differentiation. WERI-Rb1 retinoblastoma cells were obtained from ATCC. Cell
maintenance and organoid differentiation protocols are described in the supplementary
materials.
CRISPR mutations
All mutations were generated in H7 ESCs. Cells were modified to express an inducible Cas9
element. Plasmids for guide RNA (gRNA) transfection were generated by using the
pSpCas9(BB)-P2A-Puro plasmid modified from the pX459_V2.0 plasmid (62988, Addgene)
by replacing T2A with a P2A sequence. Mutations were confirmed with polymerase chain
reaction sequencing. Gene diagrams of deletions are displayed in fig. S2A. Detailed
transfection procedures, gRNA sequences, and homology arm sequences are included in the
supplementary materials.
Immunohistochemistry
Primary antibodies were used at the following dilutions: goat anti-SW-opsin (1:200 for
organoids, 1:500 for human retinas) (Santa Cruz Biotechnology), rabbit anti-LW/MW-opsins
(1:200 for organoids, 1:500 for human retinas) (Millipore), mouse anti-CRX (1:500)
(Abnova), and mouse anti-Rhodopsin (1:500) (GeneTex). All secondary antibodies were
Alexa Fluor–conjugated (1:400) and made in donkey (Molecular Probes). Detailed methods
for fixation, microscopy, and image processing of organoids, retinas, and WERI-Rb1 cells
are included in the supplementary materials.
Eldred et al. Page 8
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Organoid age
Opsin expression time course—EP1 iPSC-derived organoids for time course
experiments were binned into 10-day increments for analysis. Organoids were binned into
day 130 [actual day 129 (
n
= 3 organoids)], day 150 [actual day 152 (
n
= 4 organoids)], day
170 [actual day 173 (
n
= 2 organoids)], day 200 [actual days 194 to 199 (
n
= 7 organoids)],
day 290 [actual day 291 (
n
= 3 organoids)], and day 360 [actual day 361 (
n
= 3 organoids)].
Quantifications of outer-segment lengths and inner-segment widths were measured in day
361 organoids (
n
= 3 organoids).
Opsin expression in different conditions—iCas9 H7 ESC–derived organoids for
Thrb2
KOs and controls were analyzed at day 200. Organoids for
Thrb
KO, control, and
wild-type + T3 were analyzed at two time points: two organoids were taken at day 199 for
each group, and one was taken at day 277 for each group. T3-treated organoids were taken at
time points between day 195 and day 200 for different differentiations. For each treatment
group and genotype, organoids were compared with control organoids grown in parallel.
RNA-seq time course—EP1 iPSC-derived organoids were analyzed at time points
ranging from day 10 to day 250 of differentiation. We took samples at day 10 (
n
= 3
organoids), day 20 (
n
= 2 organoids), day 35 (
n
= 3 organoids), day 69 (
n
= 3 organoids),
day 111 (
n
= 3 organoids), day 128 (
n
= 3 organoids), day 158 (
n
= 2 organoids), day 173 (
n
= 3 organoids), day 181 (
n
= 3 organoids), day 200 (
n
= 3 organoids), and day 250 (
n
= 3
organoids). RNA from individual organoids was extracted by using the Zymo Direct-zol
RNA Microprep Kit (Zymo Research) according to manufacturer’s instructions. Libraries
were prepared by using the Illumina TruSeq stranded mRNA kit and sequenced on an
Illumina NextSeq 500 with single 200-base pair reads.
RNA-seq time course analysis
Expression levels were quantified by using Kallisto (version 0.34.1) with the following
parameters: “−b 100 −1 200 −s 10 −t 20–single”. The Gencode release 28 comprehensive
annotation was used as the reference transcriptome (57). Transcripts per million (TPM)
values (table S1) were then used to generate graphs in Prism and heatmaps in R by using
ggplot2. The distributions of transcripts were plotted so as to identify the best low TPM
cutoff (fig. S5A). The threshold was determined to be 0.7 log(TPM + 1)—5 TPM—and this
value was used as an inflection point for the heatmaps. Heatmaps for fig. S3, A to C, were
made similarly, by using CPM values from Hoshino
et. al
(fig. S5B) (23).
Measurements and quantification
Measurements of retinal area and cell morphology were done by using ImageJ software.
Quantifications and statistics (except for RNA-seq data) were done in GraphPad Prism, with
a significance cutoff of 0.01. Statistical tests are listed in figure legends. All error bars
represent the SEM.
Supplementary Material
Refer to Web version on PubMed Central for supplementary material.
Eldred et al. Page 9
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

ACKNOWLEDGMENTS
We thank A. Kolodkin, J. Nathans, and members of the Johnston laboratory for helpful comments on the
manuscript.
Funding: K.C.E. was a Howard Hughes Medical Institute Gilliam Fellow and was supported by the National
Science Foundation Graduate Research Fellowship Program under grant 1746891. R.J.J. was supported by the Pew
Scholar Award 00027373.
REFERENCES AND NOTES
1. Viets K, Eldred K, Johnston RJ, Jr., Mechanisms of photoreceptor patterning in vertebrates and
invertebrates. Trends Genet 32, 638–659 (2016). doi: 10.1016/j.tig.2016.07.004; pmid: [PubMed:
27615122]
2. Nathans J, Thomas D, Hogness DS, Molecular genetics of human color vision: The genes encoding
blue, green, and red pigments. Science 232, 193–202 (1986). doi: 10.1126/science.2937147; pmid:
[PubMed: 2937147]
3. Vollrath D, Nathans J, Davis RW, Tandem array of human visual pigment genes at Xq28. Science
240, 1669–1672 (1988). doi: 10.1126/science.2837827; pmid: [PubMed: 2837827]
4. Wang Y et al., A locus control region adjacent to the human red and green visual pigment genes.
Neuron 9, 429–440 (1992). doi: 10.1016/0896-6273(92)90181-0; pmid: [PubMed: 1524826]
5. Smallwood PM, Wang Y, Nathans J, Role of a locus control region in the mutually exclusive
expression of human red and green cone pigment genes. Proc. Natl. Acad. Sci. U.S.A 99, 1008–
1011 (2002). doi: 10.1073/pnas.022629799; pmid: [PubMed: 11773636]
6. Wang Y et al., Mutually exclusive expression of human red and green visual pigment-reporter
transgenes occurs at high frequency in murine cone photoreceptors. Proc. Natl. Acad. Sci. U.S.A 96,
5251–5256 (1999). doi: 10.1073/pnas.96.9.5251; pmid: [PubMed: 10220452]
7. Nathans J et al., Molecular genetics of human blue cone monochromacy. Science 245, 831–838
(1989). doi: 10.1126/science.2788922; pmid: [PubMed: 2788922]
8. Ladekjaer-Mikkelsen AS, Rosenberg T, Jørgensen AL, A new mechanism in blue cone
monochromatism. Hum. Genet 98, 403–408 (1996). doi: 10.1007/s004390050229; pmid: [PubMed:
8792812]
9. Patterson EJ et al., Cone photoreceptor structure in patients with X-linked cone dysfunction and red-
green color vision deficiency. Invest. Ophthalmol. Vis. Sci 57, 3853–3863 (2016). doi: 10.1167/iovs.
16-19608; pmid: [PubMed: 27447086]
10. Nakano T et al., Self-formation of optic cups and storable stratified neural retina from human
ESCs. Cell Stem Cell 10, 771–785 (2012). doi: 10.1016/j.stem.2012.05.009; pmid: [PubMed:
22704518]
11. Zhong X et al., Generation of three-dimensional retinal tissue with functional photoreceptors from
human iPSCs. Nat. Commun 5, 4047 (2014). doi: 10.1038/ncomms5047; pmid: [PubMed:
24915161]
12. Wahlin KJ et al., Photoreceptor outer segment-like structures in long-term 3D retinas from human
pluripotent stem cells. Sci. Rep 7, 766 (2017). doi: 10.1038/s41598-017-00774-9; pmid: [PubMed:
28396597]
13. Kaewkhaw R et al., Transcriptome dynamics of developing photoreceptors in three-dimensional
retina cultures recapitulates temporal sequence of human cone and rod differentiation revealing
cell surface markers and gene networks. Stem Cells 33, 3504–3518 (2015). doi: 10.1002/stem.
2122; pmid: [PubMed: 26235913]
14. Phillips MJ et al., A novel approach to single cell RNA-sequence analysis facilitates in silico gene
reporting of human pluripotent stem cell-derived retinal cell types. Stem Cells 36, 313–324 (2018).
doi: 10.1002/stem.2755; pmid: [PubMed: 29230913]
15. Artero Castro A, Lukovic D, Jendelova P, Erceg S, Concise review: Human induced pluripotent
stem cell models of retinitis pigmentosa. Stem Cells 36, 474–481 (2018). doi: 10.1002/stem.2783;
pmid: [PubMed: 29345014]
Eldred et al. Page 10
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

16. Furukawa T, Morrow EM, Cepko CL, Crx, a novel otx-like homeobox gene, shows photoreceptor-
specific expression and regulates photoreceptor differentiation. Cell 91, 531–541 (1997). doi:
10.1016/S0092-8674(00)80439-0; pmid: [PubMed: 9390562]
17. Freund CL et al., Cone-rod dystrophy due to mutations in a novel photoreceptor-specific
homeobox gene (CRX) essential for maintenance of the photoreceptor. Cell 91, 543–553 (1997).
doi: 10.1016/S0092-8674(00)80440-7; pmid: [PubMed: 9390563]
18. Chen S et al., Crx, a novel Otx-like paired-homeodomain protein, binds to and transactivates
photoreceptor cell-specific genes. Neuron 19, 1017–1030 (1997). doi: 10.1016/
S0896-6273(00)80394-3; pmid: [PubMed: 9390516]
19. Curcio CA et al., Distribution and morphology of human cone photoreceptors stained with anti-
blue opsin. J. Comp. Neurol 312, 610–624 (1991). doi: 10.1002/cne.903120411; pmid: [PubMed:
1722224]
20. Hendrickson A, Drucker D, The development of parafoveal and mid-peripheral human retina.
Behav. Brain Res 49, 21–31 (1992). doi: 10.1016/S0166-4328(05)80191-3; pmid: [PubMed:
1388798]
21. Xiao M, Hendrickson A, Spatial and temporal expression of short, long/medium, or both opsins in
human fetal cones. J. Comp. Neurol 425, 545–559 (2000). doi:
10.1002/1096-9861(20001002)425:4<545::AID-CNE6>3.0.C0;2-3; pmid: [PubMed: 10975879]
22. Curcio CA, Sloan KR, Kalina RE, Hendrickson AE, Human photoreceptor topography. J. Comp.
Neurol 292, 497–523 (1990). doi: 10.1002/cne.902920402; pmid: [PubMed: 2324310]
23. Hoshino A et al., Molecular anatomy of the developing human retina. Dev. Cell 43, 763–779.e4
(2017). doi: 10.1016/j.devcel.2017.10.029; pmid: [PubMed: 29233477]
24. Ng L et al., A thyroid hormone receptor that is required for the development of green cone
photoreceptors. Nat. Genet 27, 94–98 (2001). doi: 10.1038/83829; pmid: [PubMed: 11138006]
25. Roberts MR, Srinivas M, Forrest D, Morreale de Escobar G, Reh TA, Making the gradient: Thyroid
hormone regulates cone opsin expression in the developing mouse retina. Proc. Natl. Acad. Sci.
U.S.A 103, 6218–6223 (2006). doi: 10.1073/pnas.0509981103; pmid: [PubMed: 16606843]
26. Applebury ML et al., Transient expression of thyroid hormone nuclear receptor TRbeta2 sets S
opsin patterning during cone photoreceptor genesis. Dev. Dyn 236, 1203–1212 (2007). doi:
10.1002/dvdy.21155; pmid: [PubMed: 17436273]
27. Suzuki SC et al., Cone photoreceptor types in zebrafish are generated by symmetric terminal
divisions of dedicated precursors. Proc. Natl. Acad. Sci. U.S.A 110, 15109–15114 (2013). doi:
10.1073/pnas.1303551110; pmid: [PubMed: 23980162]
28. Sjöberg M, Vennström B, Forrest D, Thyroid hormone receptors in chick retinal development:
Differential expression of mRNAs for alpha and N-terminal variant beta receptors. Development
114, 39–47 (1992). pmid: [PubMed: 1576965]
29. Trimarchi JM, Harpavat S, Billings NA, Cepko CL, Thyroid hormone components are expressed in
three sequential waves during development of the chick retina. BMC Dev. Biol 8, 101 (2008). doi:
10.1186/1471-213X-8-101; pmid: [PubMed: 18854032]
30. Weiss AH, Kelly JP, Bisset D, Deeb SS, Reduced L- and M- and increased S-cone functions in an
infant with thyroid hormone resistance due to mutations in the THRβ2 gene. Ophthalmic Genet
33, 187–195 (2012). doi: 10.3109/13816810.2012.681096; pmid: [PubMed: 22551329]
31. Samuels HH, Tsai JS, Casanova J, Stanley F, Thyroid hormone action: In vitro characterization of
solubilized nuclear receptors from rat liver and cultured GH1 cells. J. Clin. Invest 54, 853–865
(1974). doi: 10.1172/JCI107825; pmid: [PubMed: 4372251]
32. Glaschke A, Glösmann M, Peichl L, Developmental changes of cone opsin expression but not
retinal morphology in the hypothyroid Pax8 knockout mouse. Invest. Ophthalmol. Vis. Sci 51,
1719–1727 (2010). doi: 10.1167/iovs.09-3592; pmid: [PubMed: 19834026]
33. Glaschke A et al., Thyroid hormone controls cone opsin expression in the retina of adult rodents. J.
Neurosci 31, 4844–4851 (2011). doi: 10.1523/JNEUR0SCI.6181-10.2011; pmid: [PubMed:
21451022]
34. Liu Y, Fu L, Chen DG, Deeb SS, Identification of novel retinal target genes of thyroid hormone in
the human WERI cells by expression microarray analysis. Vision Res 47, 2314–2326 (2007). doi:
10.1016/j.visres.2007.04.023; pmid: [PubMed: 17655910]
Eldred et al. Page 11
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

35. Schroeder A, Jimenez R, Young B, Privalsky ML, The ability of thyroid hormone receptors to
sense t4 as an agonist depends on receptor isoform and on cellular cofactors. Mol. Endocrinol 28,
745–757 (2014). doi: 10.1210/me.2013-1335; pmid: [PubMed: 24673558]
36. Dentice M, Marsili A, Zavacki A, Larsen PR, Salvatore D, The deiodinases and the control of
intracellular thyroid hormone signaling during cellular differentiation. Biochim. Biophys. Acta
1830, 3937–3945 (2013). doi: 10.1016/j.bbagen.2012.05.007; pmid: [PubMed: 22634734]
37. Darras VM, Houbrechts AM, Van Herck SL, Intracellular thyroid hormone metabolism as a local
regulator of nuclear thyroid hormone receptor-mediated impact on vertebrate development.
Biochim. Biophys. Acta 1849, 130–141 (2015). doi: 10.1016/j.bbagrm.2014.05.004; pmid:
[PubMed: 24844179]
38. Ng L et al., Type 3 deiodinase, a thyroid-hormone-inactivating enzyme, controls survival and
maturation of cone photoreceptors. J. Neurosci 30, 3347–3357 (2010). doi: 10.1523/JNEUR0SCI.
5267-09.2010; pmid: [PubMed: 20203194]
39. Bonezzi PJ, Stabio ME, Renna JM, The development of mid-wavelength photoresponsivity in the
mouse retina. Curr. Eye Res 43, 666–673 (2018). doi: 10.1080/02713683.2018.1433859; pmid:
[PubMed: 29447486]
40. Bagci E et al., Deiodinase knockdown during early zebrafish development affects growth,
development, energy metabolism, motility and phototransduction. PLOS ONE 10, e0123285
(2015). doi: 10.1371/journal.pone.0123285; pmid: [PubMed: 25855985]
41. Guo C et al., Intrinsic expression of a multiexon type 3 deiodinase gene controls zebrafish embryo
size. Endocrinology 155, 4069–4080 (2014). doi: 10.1210/en.2013-2029; pmid: [PubMed:
25004091]
42. Bruhn SL, Cepko CL, Development of the pattern of photoreceptors in the chick retina. J. Neurosci
16, 1430–1439 (1996). doi: 10.1523/JNEUROSCI.16-04-01430.1996; pmid: [PubMed: 8778294]
43. Bianco AC, Salvatore D, Gereben B, Berry MJ, Larsen PR, Biochemistry, cellular and molecular
biology, and physiological roles of the iodothyronine selenodeiodinases. Endocr. Rev 23, 38–89
(2002). doi: 10.1210/edrv.23.1.0455; pmid: [PubMed: 11844744]
44. Sharlin DS, Visser TJ, Forrest D, Developmental and cell-specific expression of thyroid hormone
transporters in the mouse cochlea. Endocrinology 152, 5053–5064 (2011). doi: 10.1210/en.
2011-1372; pmid: [PubMed: 21878515]
45. Lee TC, Almeida D, Claros N, Abramson DH, Cobrinik D, Cell cycle-specific and cell type-
specific expression of Rb in the developing human retina. Invest. Ophthalmol. Vis. Sci 47, 5590–
5598 (2006). doi: 10.1167/iovs.06-0063; pmid: [PubMed: 17122153]
46. Barrett EJ, in Medical Physiology, 2e Updated Edition, Boulpaep WBE, Ed. (Elsevier, Inc.,
Philadelphia, PA, 2012), chap. 49.
47. Dubovy SR et al., Expression of hypothalamic neurohormones and their receptors in the human
eye. Oncotarget 8, 66796–66814 (2017). doi: 10.18632/oncotarget.18358; pmid: [PubMed:
28977997]
48. Martino E et al., Thyrotropin-releasing hormone-like material in human retina. J. Endocrinol.
Invest 3, 267–271 (1980). doi: 10.1007/BF03348274; pmid: [PubMed: 6776180]
49. Rovet J, Simic N, The role of transient hypothyroxinemia of prematurity in development of visual
abilities. Semin. Perinatol 32, 431–437 (2008). doi: 10.1053/j.semperi.2008.09.009; pmid:
[PubMed: 19007682]
50. Simic N, Westall C, Astzalos EV, Rovet J, Visual abilities at 6 months in preterm infants: Impact of
thyroid hormone deficiency and neonatal medical morbidity. Thyroid 20, 309–315 (2010). doi:
10.1089/thy.2009.0128; pmid: [PubMed: 20144040]
51. Yassin SA, Al-Dawood AJ, Al-Zamil WM, Al-Ghamdi MA, Al-Khudairy ZN, Comparative study
of visual dysfunctions in 6-10-year-old very preterm- and full-term-born children. Int. Ophthalmol
(2018). doi: 10.1007/s10792-018-0959-2; pmid: [PubMed: 29916121]
52. Dowdeswell HJ, Slater AM, Broomhall J, Tripp J, Visual deficits in children born at less than 32
weeks’ gestation with and without major ocular pathology and cerebral damage. Br. J. Ophthalmol
79, 447–452 (1995). doi: 10.1136/bjo.79.5.447; pmid: [PubMed: 7612557]
53. Pearson RA et al., Restoration of vision after transplantation of photoreceptors. Nature 485, 99–
103 (2012). doi: 10.1038/nature10997; pmid: [PubMed: 22522934]
Eldred et al. Page 12
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

54. Barnea-Cramer AO et al., Function of human pluripotent stem cell-derived photoreceptor
progenitors in blind mice. Sci. Rep 6, 29784 (2016). doi: 10.1038/srep29784; pmid: [PubMed:
27405580]
55. Lamba DA, Gust J, Reh TA, Transplantation of human embryonic stem cell-derived photoreceptors
restores some visual function in Crx-deficient mice. Cell Stem Cell 4, 73–79 (2009). doi: 10.1016/
j.stem.2008.10.015; pmid: [PubMed: 19128794]
56. Tucker BA et al., Transplantation of adult mouse iPS cell-derived photoreceptor precursors restores
retinal structure and function in degenerative mice. PLOS ONE 6, e18992 (2011). doi: 10.1371/
journal.pone.0018992; pmid: [PubMed: 21559507]
57. Harrow J et al., GENCODE: The reference human genome annotation for The ENCODE Project.
Genome Res 22, 1760–1774 (2012). doi: 10.1101/gr.135350.111; pmid: [PubMed: 22955987]
Eldred et al. Page 13
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Fig. 1. S and L/M cone generation in human retinal organoids.
(A) Decision between S and L/M cone subtype fate. (B and C) S-opsin (blue) and L/M-
opsin (green). (B) Human adult retina age 53. (C)iPSC-derived organoid, day 200 of
differentiation. (D to K) Bright-field images of organoids derived from iPSCs. (D)
Undifferentiated iPSCs. (E) Day 1, aggregation. (F) Day 4, formation of neuronal vesicles.
(G) Day 8, differentiation of retinal vesicles. (H) Day 12, manual isolation of retinal
organoid. (I) Day 43, arrow indicates developing retinal tissue, and arrowhead indicates
developing retinal pigment epithelium. (J) Day 199, arrow indicates outer segments. (K) Day
330, arrow indicates outer segments.
Eldred et al. Page 14
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Fig. 2. Human cone subtype specification is recapitulated in organoids.
(A to K) S-opsin (blue) and L/M-opsin (green) were examined in human iPSC-derived
organoids [(A), (C) to (E), and (G) to (M)] and human retinas [(B), (D), (F), and (H)]. [(A)
to (C) and (E) to (G)] Arrows indicate outer segments, solid arrowheads indicate inner
segments, and open arrowheads indicate nuclei. [(A) and (E)] CRX (a general marker of
photoreceptors) is expressed in S cones and L/M cones. [(B) to (D)] S cones display short
outer segments and thin inner segments in both human retinas and organoids. [(F) to (H)]
L/M cones display long outer segments and wide inner segments in both human retinas and
organoids. [(D) and (H)] Quantification of outer segment lengths and inner segment widths
(adult retina, L/M,
n
= 13 cones, S,
n
= 10 cones; organoid, L/M,
n
= 35 cones, S,
n
= 42
cones). [(I) to (N)] S cones are generated before L/M cones in organoids. (L) Ratio of S:L/M
cones during organoid development. (M) Density of S and L/M cones during organoid
development. (N)
S-opsin
expression precedes
L/M-opsin
expression in human iPSC-
derived organoids.
CRX
expression starts before opsin expression. TPM, transcripts per
kilobase million.
Eldred et al. Page 15
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Fig. 3. Thyroid hormone signaling is necessary and sufficient for the temporal switch between S
and L/M fate specification.
(A to K) S-opsin (blue) and L/M-opsin (green) were examined in human ESC-derived
organoids. (A) Wild-type (WT). (B)
Thrβ2
early termination mutant (
Thrβ2
KO). (C)
Quantification of (A) and (B) (WT,
n
= 3 organoids;
Thrβ2
KO,
n
= 3 organoids). (D) WT.
(E)
Thrβ
KO. (F) WT treated with 20 nM T3 (WT + T3). (G)
Thrβ
KO treated with 20 nM
T3 (
Thrβ
KO + T3). (H) Quantification of (D) to (G) (WT,
n
= 9 organoids;
Thrβ
KO,
n
= 3
organoids; WT + T3,
n
= 6 organoids;
Thrβ
KO + T3,
n
= 3 organoids. Tukey’s multiple
comparisons test: WT versus
Thrβ
KO,
P
< 0.0001; WT versus WT + T3,
P
< 0.01; WT +
T3 versus
Thrβ
KO + T3,
P
< 0.0001). (I) Length of outer segments. WT, L/M
n
= 66 cells;
WT, S
n
= 66 cells;
Thrβ
KO,
n
= 50 cells (Tukey’s multiple comparisons test, WT L/M
versus WT S,
P
< 0.0001; WT L/M versus
Thrβ
KO,
P
< 0.0001; WT S versus
Thrβ
KO, not
significantly different). (J) Width of inner segments. WT, L/M
n
= 78 cells; WT, S
n
= 78
cells;
Thrβ
KO,
n
= 118 cells (Tukey’s multiple comparisons test, WT L/M versus WT S,
P
< 0.0001; WT L/M versus
Thrβ
KO,
P
< 0.0001; WT S versus
Thrβ
KO, not significantly
different). (K) T3 acts through
Thrβ
to increase total cone number. Quantification of density
Eldred et al. Page 16
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

of S and L/M cones; WT,
n
= 6 organoids;
Thrβ
KO,
n
= 3 organoids; WT + T3,
n
= 3
organoids;
Thrβ
KO + T3,
n
= 3 organoids (Tukey’s multiple comparisons test between total
cone numbers, WT versus
Thrβ
KO, not significantly different; WT versus WT + T3,
P
<
0.01; WT + T3 versus
Thrβ
KO + T3,
P
< 0.0001).
Eldred et al. Page 17
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Fig. 4. Dynamic expression of thyroid hormone signaling regulators during development.
(A to C) Heat maps of log(TPM + 1) values for genes with (A) changing expression, (B)
consistent expression, and (C) no expression. Numbers at the bottom of heat maps indicate
organoid age in days. (D) Model of the temporal mechanism of cone subtype specification in
humans. For simplicity, only the roles of DIO3 and DIO2 are illustrated. In step 1,
expression of DIO3 degrades T3 and T4, leading to S cone specification. In step 2,
expression of DIO2 converts T4 to T3 to signal Thrβ to repress S and induce L/M cone fate.
Eldred et al. Page 18
Science
. Author manuscript; available in PMC 2019 May 12.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript