DEVELOPMENTAL DYNAMICS 237:758 –767, 2008

PATTERNS & PHENOTYPES

Transcription Factors CTCF and Pax6 Are

Segregated to Different Cell Types During
Retinal Cell Differentiation
M. Valeria Canto-Soler,1* Hu Huang,1 M. Soledad Romero,1 and Ruben Adler1,2

We have hypothesized that the transcription factor CTCF may influence retinal cell differentiation by
controlling Pax6 expression, because (1) CTCF has been shown to repress Pax6 expression in some tissues,
and (2) Pax6 blocks the differentiation of retinal progenitor cells as photoreceptors and promotes their
differentiation as nonphotoreceptor neurons. Our results show that, as predicted by this hypothesis, CTCF
and Pax6 become segregated to different retinal cell types. The factors are initially coexpressed in the
undifferentiated neuroepithelium, but already at that time they show complementary periphery-to-fundus
gradients of distribution. As the retina laminates, Pax6 becomes restricted to ganglion and amacrine cells,
and CTCF to the bipolar/Muller cell layer and the outer nuclear layer. Polymerase chain reaction analysis
of laser capture microdissection samples and dissociated cells showed that both immature and
differentiated photoreceptors are CTCF (ⴙ)/ Pax6 (ⴚ). Functional studies are now under way to further
analyze the role of CTCF in retinal cell differentiation. Developmental Dynamics 237:758 –767, 2008.
© 2008 Wiley-Liss, Inc.

Key words: retinal development; retinal differentiation; retinal gradients; photoreceptors; laser capture microdissection;
single cell cDNA

Accepted 26 November 2007

INTRODUCTION repression, transcriptional activation, tility includes the use of different com-
CTCF (CCCTC-binding factor) is an hormone-responsive gene silencing, binations of zinc fingers for binding to
evolutionary conserved 11 zinc-finger methylation-dependent chromatin in- different sequences, as well as post-
phosphoprotein, first identified as a sulation, genomic imprinting, and a translational modifications that deter-
transcription factor that binds to si- central role in the enhancer blocking mine switches between repressive and
lencer elements in the chicken c-myc activity of insulator elements (reviewed activating functions (Filippova et al.,
and lysozyme genes (Baniahmad et al., by Ohlsson et al., 2001; Klenova et al., 1996; El-Kady and Klenova, 2005). In
1990; Lobanenkov et al., 1990). Its di- 2002; Ladomery and Dellaire, 2002; the chick, CTCF is a single gene with
verse functions include transcriptional Dunn and Davie, 2003). CTCF’s versa- four major mRNA species (Klenova et

ABBREVIATIONS AC amacrine cells BC bipolar cells ED embryonic day GC ganglion cells GCL ganglion cell layer HC horizontal cells
INL inner nuclear layer IR inner retina LCM laser capture microdissection ONL outer nuclear layer OPL outer plexiform layer PGCL
putative ganglion cell layer PONL putative outer nuclear layer

1
The Department of Ophthalmology, The Johns Hopkins University School of Medicine, Baltimore, Maryland
2
The Department of Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, Maryland
Grant sponsor: NIH; Grant number: EYO4859; Grant number: Core Grant EY1765; Grant sponsor: Research to Prevent Blindness, Inc. (New
York, NY); Grant sponsor: the Knights Templar Eye Foundation, Inc. (Schaumburg, IL).
*Correspondence to: Valeria Canto-Soler, The John Hopkins School of Medicine, 600 N. Wolfe St., 519 Maumenenee, Balti-
more, MD 21287-9257. E-mail: [email protected]
DOI 10.1002/dvdy.21420
Published online 25 January 2008 in Wiley InterScience (www.interscience.wiley.com).

© 2008 Wiley-Liss, Inc.

 CTCF AND PAX6 EXPRESSION DURING RETINAL CELL DIFFERENTIATION 759

al., 1993). Although the open reading become strongly expressed in photore- lines and tissues (Klenova et al., 1993)
frame predicts a protein of 82 kDa, ceptor progenitors in which Pax6 is but, to the best of our knowledge, the
CTCF migrates to an apparent 130-kDa down-regulated. 120-kDa isoform has not been previ-
molecular mass due to the electro- We have investigated these predic- ously reported.
phoretic properties of its N- and C- tions at different stages of development, As also illustrated in Figure 1D, com-
terminal domains (Klenova et al., using whole retina extracts, laser cap- parative analysis of various develop-
1997). In addition to the 130-kDa form, ture microdissection (LCM) samples, mental stages showed different tempo-
isoforms of 97, 80, 73, 70 and 55 kDa individual cells isolated from retinal ral patterns of isoform expression in the
have been observed in different chicken dissociates, and tissue sections. CTCF retina. The-130 kDa isoform was down-
tissues and cell lines (Klenova et al., mRNA and protein were detected in the regulated after ED12 (Fig. 1D); in con-
1993). chicken retina at all the stages studied. trast, the 70-kDa isoform, which has
Our interest in CTCF was stimulated CTCF and Pax6 mRNA and protein been suggested to interact with the 130-
by its reported capacity to regulate the showed complementary patterns of ex- kDa isoform (Klenova et al., 1997), ap-
expression of Pax6, a homeobox gene pression throughout retinogenesis; all peared progressively up-regulated be-
whose involvement in various aspects of photoreceptor subtypes were CTCF (⫹) tween ED12 and ED18 (Fig. 1D). The
retinal development has been investi- and Pax6 (⫺). These results provide cir- 97- and 120-kDa bands were weak at
gated in several laboratories (Schedl et cumstantial evidence in support of the ED6, increased between ED8 and
al., 1996; Marquardt et al., 2001; Zhang hypothesis that CTCF may influence ED12, and returned thereafter to low
et al., 2003; Philips et al., 2005; Curto et retinal cell differentiation, and particu- levels similar to those at ED6 (Fig. 1D).
al., 2007), including our own (Belecky-Ad- larly photoreceptor cell differentiation, The 55-kDa band was the weakest one
ams et al., 1997; Toy et al., 2002; Canto- through the down-regulation of Pax6. throughout development and became
Soler and Adler, 2006). CTCF has been Functional experiments are under way practically undetectable after ED12
shown to down-regulate Pax6 expression to test this hypothesis. (Fig. 1D).
by interacting with a repressive element
located upstream of the Pax6 P0 pro- In Situ Hybridization
moter (Li et al., 2004; Wu et al., 2006); RESULTS Analysis of CTCF and Pax6
accordingly, CTCF down-regulation en- Analysis in Retinal Extracts mRNA Distribution in the
hances Pax6 expression in rabbit corneal
epithelial cells, while CTCF overexpres- Expression of CTCF mRNA in Chick Embryo Retina
sion in transgenic mice suppresses Pax6 the embryonic chick retina. Initial in situ hybridization showed a
and leads to severe microphthalmia, sim- CTCF mRNA expression was detected distribution of CTCF mRNA that ap-
ilar to that of small eye Pax6 mutants (Li in retinal extracts by reverse transcrip- peared complementary to the previ-
et al., 2004). Pax6 is diffusely expressed tase-polymerase chain reaction (RT- ously reported distribution of Pax6
in the undifferentiated retinal neuroepi- PCR) at all the stages studied between mRNA at several stages of chick retinal
thelium; as cells become postmitotic and embryonic day (ED) 6 and 18 (Fig. 1A). development (Belecky-Adams et al.,
arranged in a characteristic laminar pat- Real-time quantitative PCR (QPCR) 1997); both transcription factors, there-
tern; however, Pax6 remains strongly ex- showed a fairly constant level of CTCF fore, were analyzed in parallel in this
pressed in putative ganglion, amacrine, mRNA relative to the housekeeping study. At early (ED 3– 4) stages, the
and horizontal cells, while becoming con- gene glyceraldehyde-3-phosphate dehy- undifferentiated retinal neuroepithe-
spicuously down-regulated in photorecep- drogenase (GAPDH; Fig. 1B). This per- lium showed mRNA for these two fac-
tor progenitors in the outer nuclear layer sistent level of expression as the retina tors distributed in complementary gra-
(ONL), and in the cells occupying the changes from an undifferentiated neu- dients: periphery-low to fundus-high for
Muller/bipolar region of the inner nuclear roepithelium to a highly heterogeneous CTCF, and periphery-high to fundus-
layer (INL; Belecky-Adams et al., 1997). tissue is consistent with the possible in- low for Pax6 (Fig. 2A,B). By ED6, CTCF
Based on these observations, we hypoth- volvement of CTCF in a variety of dif- and Pax6 signals appeared fairly wide-
esized that Pax6 downregulation could be ferent developmental phenomena. spread throughout the undifferentiated
necessary for photoreceptor differentia- neuroepithelium (Fig. 2C,D). However,
tion; this hypothesis was supported by Expression of CTCF protein at in the vitreal region of the retina, pre-
the finding that Pax6 overexpression in sumably occupied by recently “born”
vitro inhibited the differentiation of reti-
different stages of retinogenesis. ganglion cells, CTCF signals appeared
nal progenitor cells as photoreceptors, In agreement with a previous report weaker and Pax6 labeling was stronger
while stimulating their differentiation as (Klenova et al., 1993) analysis of than in the rest of the neuroepithelium
nonphotoreceptor neurons (Toy et al., chicken heart tissue by Western blot (Fig. 2C,D). As retina lamination be-
2002). Taken together, these observations showed two CTCF isoforms of 130 and came evident by ED8, CTCF mRNA sig-
would suggest that functional interac- 55 kDa (Fig. 1C). As shown in Figure nals were more intense in the ONL and
tions between CTCF and Pax6 could be 1D, the retina contained not only the the outer part of the INL than in the
relevant to retinal cell differentiation and 130- and 55-kDa isoforms, but also ganglion cell layer (GCL) and the inner
predict that these transcription factors three additional isoforms of approxi- part of the INL (Fig. 2E). Pax6 distribu-
should be distributed in complemen- mately 70, 97, and 120 kDa. The 70- tion again showed a complementary
tary patterns within the retina and, and 97-kDa isoforms have been previ- pattern, with signals being stronger in
more specifically, that CTCF should ously described in some chicken cell the GCL and the inner part of the INL
760 CANTO-SOLER ET AL.

than in the outer part of the INL and in

the ONL (Fig. 2F). These patterns ap-
peared more sharply delineated by
ED18, when the lamination of the ret-
ina is already well established; CTCF
mRNA appeared restricted to the ONL
and outer region of the INL, with very
low or undetectable levels in the GCL
and inner region of the INL (Fig. 2G);
Pax6 expression was predominantly
confined to the GCL and the inner re-
gion of the INL, but lighter Pax6 signal
was also seen in horizontal cells, partic-
ularly at higher magnification (Fig. 2H;
and data not shown).

Distribution of CTCF and Pax6

mRNA in LCM Samples and
Isolated Photoreceptor Cells
LCM-isolated ONL and inner
retina samples.
Putative photoreceptor cells are already
postmitotic by ED8, when they occupy
the ONL, which is clearly demarcated
from the rest of the retina by an incipi-
ent outer plexiform layer (OPL). LCM
was, therefore, used to isolate the ONL
layer, as well as an inner retina (IR)
fraction that extended from the OPL to
the vitreal surface of the retina. After
cDNA synthesis and global PCR ampli-
fication, possible cross-contamination
was investigated by PCR analysis of
mRNAs known to be restricted to the
ONL (e.g., IRBP) or the IR (CRALBP,
CAC, Vsx2/Chx10 and Prox1). ONL and
IR samples showing no signs of cross-
contamination (n ⫽ 4) were processed
for PCR detection of CTCF and Pax6
mRNA. As expected, both CTCF and Fig. 1. CTCF mRNA and protein are expressed in the chicken retina throughout development.
Pax6 were detected in the highly heter- A,B: CTCF mRNA was investigated at different developmental stages by reverse transcriptase-
ogeneous IR samples; on the other polymerase chain reaction (RT-PCR, A) and real-time quantitative PCR (B). C,D: Western blot
hand, ONL samples, containing only analysis. C: CTCF protein isoforms present in embryonic day (ED) 18 heart. D: CTCF isoforms
detected in the chicken retina at different stages of development.
photoreceptor progenitors, were posi-
tive for CTCF but devoid of Pax6 (Fig.
3A).
Rods, double cones, and red and green the cells was corroborated by PCR am-
Analysis of CTCF and Pax6 single cones were identified by mor- plification of an mRNA expressed in
mRNA expression in isolated phological criteria, including an elon- both rods and cones (Crx), and of pho-
gated shape, the presence of a rod-like toreceptor type-specific visual pig-
photoreceptors. or cone-like outer segment and, in the ment mRNAs (rhodopsin for rods, io-
The data described in previous sec- case of single cones, a brightly colored dopsin for double cones and red cones,
tions did not elucidate whether CTCF oil droplet whose pigmentation is and green opsin for green cones; Fig.
mRNA is expressed in all or only in characteristic of particular cone sub- 3B). CTCF expression was detected in
subsets of photoreceptors. This ques- types (Bowmaker et al., 1997; Lopez et all photoreceptor sub types analyzed
tion was investigated by PCR analysis al., 2005; Huang et al., 2007; Wahlin (Fig. 3B). Signals were seen in 4/6
of globally amplified cDNA from indi- et al., 2007). Blue and violet cones rods, 4/6 red cones, 5/6 green cones,
vidual rods and cones isolated from were not included in this analysis be- and 5/6 double cones. It is unclear at
dissociated ED20 chicken retinas. cause of their scarcity. The identity of this stage whether individual varia-
 CTCF AND PAX6 EXPRESSION DURING RETINAL CELL DIFFERENTIATION 761

al., 2006). It appeared therefore impor-

tant to determine whether the distribu-
tion of CTCF and Pax6 proteins
matched that of their cognate mRNAs
at all the developmental stages in-
cluded in this study. Immunocytochem-
ical analysis at ED3– 4 showed that
CTCF was highly expressed in the fun-
dal region of the retina with lower ex-
pression levels at the periphery, while
Pax6 conversely showed stronger sig-
nals at the periphery than in the fundus
(Fig. 4A–C). These gradients were
much less conspicuous by ED6. In the
fundal region, in particular, CTCF and
Pax6 signals overlapped with a fairly
diffuse distribution pattern; however,
putative ganglion cells showed stronger
signals for Pax6 than for CTCF, while
the opposite was true in the ONL (Fig.
4D,D⬘). The periphery-to-fundus gradi-
ents were no longer observed at ED8,
when overlapping expression in cells lo-
cated in the middle region of the INL
(Fig. 4E⬘) was accompanied by the
emergence of a pattern of complemen-
tary laminar distribution of CTCF and
Pax6. Thus, CTCF signals were strong
and Pax6 signals were almost undetect-
able in the ONL and the outer region of
the INL, whereas signals for CTCF
were very weak and those for Pax6 were
strong in ganglion cells and the inner
region of the INL (Fig. 4E,E⬘). These
laminar differences became even
clearer as the retina matured at ED12
(not shown) and ED18 (Fig. 4F–H).
Given the heterogeneity of the INL, the
cellular distribution of CTCF within
this retinal region was further investi-
gated by colabeling with antibodies
against Lim1, Lim3, or AP2␣, tran-
scription factors known to be expressed
in horizontal, bipolar and amacrine
Fig. 2. CTCF and Pax6 mRNA show complementary patterns of distribution throughout retino- cells, respectively. This analysis corrob-
genesis. A,B: Transverse sections of embryonic day (ED) 4 chicken embryo eyes. CTCF (A) and orated that CTCF is expressed in hori-
Pax6 (B) showed complementary fundus-to-periphery patterns of distribution. C–H: Transverse zontal and bipolar cells (Fig. 4I–N) but
sections of embryonic chicken retina showing distribution of mRNA for CTCF (C,E,G) and Pax6
not in amacrine cells (Fig. 4I–Q). It is
(D,F,H) at ED6 (C–D), ED8 (E,F) and ED18 (G,H). Retinal sections are shown with the pigment
epithelium at the top and the vitreal surface of the retina at the bottom of the pictures. Note that worth noting that horizontal cells were
CTCF signals were stronger in the outer nuclear layer (ONL) and the outer region of the inner the only cell type in which CTCF and
nuclear layer (INL), whereas Pax6 signals were stronger in the ganglion cell layer (GCL) and the Pax6 could be concomitantly detected at
inner region of the INL. The asterisk indicates horizontal cells. f, fundus; p, periphery. this stage with the techniques used in
this study (Fig. 4F–H, asterisks; and
tion reflects real differences between Comparative Analysis of F⬘–H⬘).
the cells or represents limitations of CTCF and Pax6 Protein
the technique, as previously reported
Expression DISCUSSION
by others (Chiang and Melton, 2003;
Tietjen et al., 2003; Wahlin et al., It has been recently reported that the The results of this study can be sum-
2004). Pax6 PCR products were not expression of several genes involved in marized as follows: (1) CTCF mRNA
observed in any of the photoreceptors retinal cell differentiation is regulated and protein were identified in whole
studied (not shown). posttranscriptionally (Decembrini et retina extracts at all the stages stud-
762 CANTO-SOLER ET AL.

Fig. 3. Photoreceptor precursors and differen-

tiated photoreceptors express CTCF but are de-
void of Pax6. A: Hematoxylin and eosin (H&E)
panels show transverse sections of embryonic
day (ED) 8 chicken retina from which outer nu-
clear layer (ONL) and inner retina (IR) samples
were isolated by laser capture microdisection
(LCM). Polymerase chain reaction (PCR) de-
tected both CTCF and Pax6 mRNA in IR sam-
ples, while only CTCF was detected in ONL
samples. B: Differential interference contrast
microscopy images of single photoreceptors
isolated from dissociated ED20 chick retinas.
Individual photoreceptor subtypes were iden-
tified as described in the Results section. Note
that CTCF mRNA was detected in all photore-
ceptor subtypes analyzed.

Fig. 4. CTCF and Pax6 protein

show complementary patterns of
expression throughout retinogen-
esis. A–E: Transverse sections of
chicken embryo optic cups at em-
bryonic day (ED) 4 (A–C), ED6
(D,D⬘) and ED8 (E,E⬘) processed
for double-labeling immunocyto-
chemistry for CTCF (red) and Pax6
(green). F–Q: Transverse sections
of ED18 chicken retina processed
by double-labeling immunocyto-
chemistry for CTCF (red) in com-
bination with Pax6 (F–H and F’–
H’) Lim1 (I–K), Lim3 (L–N), AP2␣
(O–Q) (green). Retinal sections are
shown with the pigment epithe-
lium at the top and the vitreal sur-
face of the retina at the bottom of
the pictures. Note CTCF expres-
sion in horizontal (asterisks) and
bipolar cells and its exclusion from
amacrine cells. AC, amacrine
cells; BC, bipolar cells; GC, gan-
glion cells; GCL, ganglion cell lay-
er; HC, horizontal cells; INL, inner
nuclear layer; ONL, outer nuclear
layer; PGCL, putative ganglion cell
layer; PONL, putative outer nu-
clear layer.
 CTCF AND PAX6 EXPRESSION DURING RETINAL CELL DIFFERENTIATION 763

ied; (2) Western blots showed several 2004), the expression of RZF in hu- that the cells encounter as they mi-
CTCF retinal isoforms, including one man photoreceptors (Sharma et al., grate toward their definitive laminar
that has not been reported so far in 2004), and the synergistic interactions positions within the retina (Adler,
other tissues; bands corresponding to between Sp4 and Crx (Lerner et al., 2000). Horizontal cells represented a
all isoforms showed characteristic 2005). We detected CTCF among sev- noteworthy exception, because they
temporal variations in intensity; (3) eral zinc-finger transcription factors did show signals for both CTCF and
CTCF and Pax6 mRNA showed com- found in developing photoreceptors by Pax6. There is an interesting correla-
plementary patterns of distribution at suppression subtractive hybridization tion between this “hybrid” phenotype
all the stages studied: (i) At ED3– 4, (Huang and Adler, unpublished obser- of horizontal cells and their peculiar
CTCF and Pax6 showed complemen- vations), and its presence has now migratory behavior (Edqvist and Hall-
tary periphery-to-fundus gradients of been verified with several complemen- book, 2004), because after being
distribution; (ii) At ED6, the vitreal tary techniques. “born” they move first toward the in-
surface of the retina already showed Although QPCR analysis showed a ner retina (an environment in which
weaker signals for CTCF and stronger fairly constant relative level of CTCF amacrine and ganglion cells become
signals for Pax6 than the rest of the mRNA in the retina throughout devel- rich in Pax6 and poor in CTCF), and
neuroepithelium; (iii) At ED8, CTCF opment, a much more dynamic behav- only subsequently return to the
mRNA was abundant and Pax6 ior was observed with other methods. boundary between the INL and the
mRNA scarce or undetectable in the Examples are the bands correspond- OPL (where cells are rich in CTCF
ONL and the outer region of the INL, ing to five different CTCF isoforms de- and poor in Pax6). It must be noted
while Pax6 signals were strong and tected by Western blot analysis, which that, at this time, it is uncertain
CTCF signals low or undetectable in showed isoform-specific variations in whether these complementary pat-
ganglion cells and the inner region of intensity during development. Their terns of expression are peculiar to the
the INL; (iv) This complementary biological significance in the retina is chick, or are also found in other spe-
laminar distribution became even bet- still unclear, as is the role of CTCF cies (Burke et al., 2002a).
ter defined by ED18; (4) PCR analysis isoforms in other tissues (Klenova et As mentioned in the Introduction
showed that photoreceptor precursors al., 1993; Dunn and Davie, 2003). It section, our interest on possible inter-
as well as differentiated rods, double has been suggested that the smaller actions between CTCF and Pax6 de-
cones and red and green single cones forms of CTCF could act as modula- rived from the hypothesis that Pax6
were positive for CTCF but devoid of tors of the full length protein (Klenova down-regulation is necessary for pro-
Pax6 mRNA; (5) CTCF and Pax6 pro- et al., 1997; Dunn and Davie, 2003), genitor cells to develop as photorecep-
tein also showed complementary pat- but further analysis will be necessary tors (Belecky-Adams et al., 1997; Toy
terns of distribution at all the stages to determine the validity of this and et al., 2002), and from reports that
studied; (6) Within the heterogeneous other possible functions. CTCF can regulate Pax6 expression in
INL, colocalization of CTCF immuno- A very dynamic pattern of expres- corneal and retinoblastoma cells by
reactivity with that for Lim1 or Lim3 sion, involving CTCF and Pax6, was interacting with a repressor element
but not with Ap2␣ demonstrated that also observed by in situ hybridization located upstream of the Pax6 P0 pro-
CTCF is present in horizontal and bi- and immunocytochemistry. Although moter (Li et al., 2004, 2006). In the
polar cells, but is absent from differ- opposite fundus-to-periphery CTCF context of this hypothesis, it is rele-
entiated amacrine cells. and Pax6 gradients were observed at vant that it was possible to demon-
Zinc-finger proteins are the largest ED3– 4, both factors were detectable strate the complete absence of Pax6
family of transcriptional regulators in together in the undifferentiated reti- in CTCF (⫹) photoreceptors and their
the human proteome (Ladomery and nal neuroepithelium until approxi- precursor cells, by PCR analysis of iso-
Dellaire, 2002; Ganss and Jheon, mately ED6. Their coexpression ap- lated single cells and LCM samples.
2004; Merzdorf, 2007). Their possible peared transient, however, and was These data are consistent with the
involvement in photoreceptor differ- no longer detectable after ED8; as the possibility that CTCF up-regulation
entiation has been recognized in Dro- layers of the retina became estab- may influence the development of ret-
sophila for some time (Moses et al., lished at this time, CTCF and Pax6 inal precursors as photoreceptors
1989; Moses and Rubin, 1991; Della et became segregated into a complemen- through Pax6 down-regulation. It is
al., 1993; Begemann et al., 1997; Lai tary laminar pattern. Even at ED6, not inconceivable, however, that
et al., 1997; Dickson, 1998; Lai and Li, moreover, signals for Pax6 were stron- CTCF could also influence photorecep-
1999). Their analysis in vertebrate ger in putative ganglion cells located tor differentiation through other
retinas includes reports of the down- at the vitreal surface of the retina, mechanisms. For example, CTCF
regulation of IRBP by Mok2 (Arranz whereas CTCF signals appeared more binding sites are often found in close
et al., 2001), the embryonic distribu- intense in the region of putative pho- proximity to thyroid hormone re-
tion and function in lens development toreceptors (future ONL). Taken to- sponse elements (TREs), and syner-
of the muscleblind gene (Kanadia et gether, these changes are consistent gistic interactions between CTCF and
al., 2003a,b), the interaction of Nrl with a model proposing that the fate of thyroid hormone receptor (THR) are
with the zinc-finger protein Fiz1 (Mit- developing retinal cells (including the necessary for the repression of several
ton et al., 2003), the repression of the type of transcription factors that they genes (Lutz et al., 2003; reviewed by
rhodopsin and IRBP promoters by express) is influenced by position-de- Burke et al., 2002b). Interestingly,
Kruppel-like factor 15 (Otteson et al., pendent microenvironmental factors THRs are expressed in retinal progen-
764 CANTO-SOLER ET AL.

TABLE 1. List of Primers Used in This Study

Gene Accession no. Application Forward primer Reverse primer

CAC Z14957.1 LCM GGCAACATCTTGTCTCCCATA GCTCTCCCAAACAGCCAATA
Vsx2/Chx10 AF178671 LCM TTCTGAGCCTGAAGGGAATG CACAGCGTGGGCTATAATCT
CRALBP BX935536 LCM TAACCCCAACCACCTCCATA CTGGCTGTGCAGACAGTGTT
CRX AF 285171 Single cell PCR ATGCCACAAGCACACAGGCA GCATCGGCAAACACAGGAAGCA
CTCF NM 205332 RT-PCRSingle cell AAGGGCAAACGTGGCCGAAA AGGCTGAGCTTCCTCCTCTT
PCRn In situ probe
synthesis LCM
CTCF NM 205332 QPCR AGAGCCTCAGTCTTGCTGAT CCTTCCCAGCTTTCAAACCC
GAPDH Single cell PCR AGGGTAGTGAAGGCTGCTGCTGA ATGCAGAACTGAGCGGTGGT
Green opsin NM 205490 Single cell PCR ATCCCTGCAGAGCTGCACAT AGCAGTCCTTGTGCTGACACGA
Iodopsin NM 205440 Single cell PCR TCCCGCATGGTGGTGGTGATGAT AACCGGAAGCTGCGGTGTGGA
IRBP LCM CGATGCCGGACATCAAGTTC ACCCACCACTGGTCTGTTCT
Pax6 D87837 Single cell PCRLCM CACCACTTCCACAGGTCTCAT TCGTGCCGAACCGATATAAT
Prox1 U46563 LCM ATAACCTGCGCCAGTAATCG AGACCAAGACGTGGCAATTT
Rhodopsin D00702.1 Single cell PCR ACACGCTGAAGCCGGAGATCAA TTCTTGCCGCAGCAGAGGGTTGT

a
LCM, laser capture microdissection; RT-PCR, reverse transcriptase-polymerase chain reaction; QPCR, real-time quantitative PCR.

itor cells and differentiating photore- bin/primer3/primer3_www.cgi) ac- mM). Individual rods, double cones,
ceptors, and play a role in the control cording to the published chicken and green and red single cones, iden-
of photoreceptor genesis and differen- CTCF sequence (NM_205332). Real- tified by morphological criteria de-
tiation (Ng et al., 2001; Yanagi et al., time QPCR was performed in 96-well scribed in the Results section, were
2002; Roberts et al., 2006; Mader and plates with the Bio-Rad IQ5 system, captured with a capillary pipette, and
Cameron, 2006; reviewed by Harpa- as described in (Huang et al., 2005) microgram amounts of cDNA were
vat and Cepko, 2003). Additional with some modifications. Each 20 ␮l synthesized and exponentially ampli-
analysis of the Pax6 gene sequence of reaction contained 10 ␮l of 2⫻ fied from each cell (Dulac, 1998; Wah-
with Mapper (Marinescu et al., SYBR Green Supermix, 100 –250 nM lin et al., 2004). Cell identity was ver-
2005a,b; http://mapper.chip.org) re- of primer mix, and 20 ng of a cDNA ified by PCR amplification of cell type-
vealed several TREs in the vicinity of template. QPCR conditions included specific visual pigment genes as
the active CTCF binding sites, making an initial denaturing step for 3 min described in the Results section, with
it possible to speculate that CTCF at 95°C, followed by 35–38 cycles primers listed in Table 1.
may repress Pax6 expression in devel- (95°C for 15 sec, 55–58°C for 20 sec,
oping photoreceptors by synergistic and 72°C for 25 sec). Primers were LCM
interaction with THRs. optimized by melting curve profiles
The anterior chamber, lens, and vitre-
and agarose gel analysis. For quan-
ous body were removed from enucle-
EXPERIMENTAL tification, a standard curve was gen-
ated ED8 chick embryo eyes, and the
erated from a series of cDNA dilu-
PROCEDURES resulting eyecups were infiltrated at
tions. Relative expression of each
4°C in 6.75, 12.5, and 25% sucrose in
Experimental Animals gene was calculated using the second
0.1 M phosphate buffer for 10, 20, and
derivative maximum values from the
White Leghorn chick embryos from 30 min, respectively, equilibrated in
linear regression of threshold cycle
B&E Eggs (York Spring, PA), staged 2:1 25% sucrose: OCT for 1 hr, and
(CT) number versus log concentra-
as in Hamburger–Hamilton (Ham- frozen by immersion in a mixture of
tion of the amplified gene. The
burger and Hamilton, 1951), were dry ice and methylbutane. Seven mi-
housekeeping gene GAPDH was
used. crometer cryosections were mounted
used for normalization.
onto slides with charged PEN-foil
RT-PCR and Real-Time membranes (Leica Microsystems,
Single Cell PCR Bannockburn, IL), fixed in 70% cold
QPCR Enucleated eye cups form ED20 ethanol for 30 sec, briefly rinsed in
Total RNA was isolated from chick chicken embryos were incubated in DEPC-treated water, stained with he-
embryo retinas at ED6, 8, 10, 12, 15, Ca2⫹ and Mg2⫹ free HBSS (CMF) for matoxylin for 10 seconds, and dehy-
and 18 with Trizol (Invitrogen, 20 min at 37°C; after removing the drated in 70% and 100% ethanol for 30
Carlsbad, CA), reverse-transcribed retinal pigment epithelium, the retina and 60 sec, respectively. Outer nu-
into cDNA with Supercript III (In- was dissociated after a 5-min incuba- clear layer and inner retina samples,
vitrogen) and primed with oligo(dT) tion in CMF containing 1 mM ethyl- delimited as described in the Results
at 50°C for 60 min. CTCF primers enediaminetetraacetic acid (EDTA) section, were separately collected us-
(see Table 1) were designed with and 5 U/ml papain, preactivated with ing an LMD6000 LCM microscope
Primer3 (http://frodo.wi.mit.edu/cgi- the reducing agent L-cysteine (2.7 (Leica Microsystems). From each sam-
 CTCF AND PAX6 EXPRESSION DURING RETINAL CELL DIFFERENTIATION 765

ple, cDNA was synthesized and expo- tic acid) for 1 hr on ice; briefly rehy- 500). Sections were then profusely
nentially amplified according to Dulac drated in 95%, 75%, and 50% ethanol; washed in 1⫻ PBS, followed by over-
(1998) and Wahlin et al. (2004) with washed in 1⫻ phosphate buffered sa- night incubation with CTCF antibody
some modifications. line (PBS) and subsequently im- (1:4,000) and visualization with Alexa
mersed in 5, 10, and 15% sucrose in 594 goat anti-mouse secondary anti-
phosphate buffer for several hours body.
Western Blot Analysis each, followed by overnight incubation
Chicken retina and heart tissues were in 20% sucrose and 1-hr incubation in
In Situ Hybridization
homogenized in cold lysis buffer (150 a 2:1 ratio of 20% sucrose and OCT
mM NaCl, 50 mM Tris pH 8.0, 2 mM Tissue-Tek (Ted Pella; Redding, CA), In situ hybridization was carried out
EDTA, 1% Triton X-100) containing and by snap freezing on dry ice/meth- as in Canto-Soler and Adler (2006)
“Complete Mini™” protease inhibitor ylbutane. Immunohistochemistry was Pax6 sense and antisense probes
(Roche, Indianapolis, IN), incubated carried out in 10-␮m cryostat sections. were generated from cDNA gener-
for 30 – 45 min on ice, and centrifuged For double labeling with CTCF (mono- ously provided by Dr. Olof Sundin
at 12,000 rpm for 15 min and the su- clonal antibody kindly provided by Dr. (Johns Hopkins University, Balti-
pernatant containing the crude pro- Loukinov, National Institute of more, MD). CTCF probes were gen-
tein lysate saved for analysis. Protein Health) and Pax6 (rabbit polyclonal erated by PCR amplification of a
concentration was measured with the antibody kindly provided by Dr. Reed, 345-bp segment from embryonic
DC Protein Assay Kit (Bio-Rad, Her- Johns Hopkins University), sections chick retina cDNA using the primers
cules, CA) according to manufactur- were blocked and permeabilized in listed in Table 1. Probes were la-
er’s instructions. Equal amounts of 10% goat serum /0.25% Triton X-100 beled with digoxigenin RNA labeling
protein were electrophoresed on Nu- for 1 hr at room temperature and in- mix, and hybridization was detected
PAGE Novex 4 –12% bis-tris gels (In- cubated overnight in a primary anti- with an alkaline phosphatase-la-
vitrogen), and transferred into a nitro- body mixture (CTCF 1:4,000/Pax6 beled anti-digoxigenin antibody and
cellulose membrane using an iBlot 1:1,000) in 2% goat serum containing the nitroblue tetrazolium/5-bromo-4-
Dry Blotting System according to the 0.05% Triton X-100 at 4°C. Antibody chloro-3-indolyl phosphate (NBT/
manufacturer’s instructions (Invitro- binding was detected with Alexa 594 BCIP) alkaline phosphatase enzy-
gen). Membranes were incubated in goat anti-mouse and Alexa 488 goat matic reaction (Roche; Indianapolis,
primary antibody (CTCF monoclonal anti-rabbit (Molecular Probes Inc., IN).
antibody kindly provided by Dr. Eugene, OR; 1:1,000). Double immu-
Loukinov, National Institutes of nostaining involving CTCF and a sec-
NOTE ADDED IN PROOF
Health, 1:2000) overnight at room ond monoclonal antibody (Lim3, Lim1
temperature, followed by incubation or AP2␣; NICHD-funded Develop- M. V. C-S., H. H., and M. S. R.
with an alkaline phosphatase-conju- mental Studies Hybridoma Bank would like to express their whole-
gated secondary antibody (Sigma; maintained by The University of hearted admiration and gratitude for
1:3,000) and Vector’s alkaline phos- Iowa, Iowa City, IA) was performed Dr. Ruben Adler, who passed away
phatase detection kit (Vector labs, using tyramide amplification accord- after final acceptance of this article.
Burlingame, CA). Gels run in parallel ing to the manufacturer’s instructions His sudden and unexpected death
were processed with a primary anti- with some modifications (TSA-PLUS filled us with the deepest sorrow for
body against chicken tubulin losing not only a superb mentor but a
Fluorescein System; Perkin Elmer,
(NICHD-funded Developmental Stud- true role-model for many aspects of
Boston, MA Perkin Elmer). Briefly,
ies Hybridoma Bank maintained by our lives.
sections were incubated in 1% H2O2
The University of Iowa, Iowa City, IA) for 30 min at room temperature to
as a loading control. As an additional quench endogenous peroxidase activ- ACKNOWLEDGMENTS
loading control, gels were stained with ity, blocked and permeabilized in 10% The authors thank Olof Sundin,
Coomassie blue after the transfer goat serum/0.25% Triton X-100 for 1 Dmitri Loukinov, and Randall Reed
step. hr at room temperature, and incu- for kind gifts of the Pax6 cDNA con-
bated overnight at 4°C in primary an- struct and CTCF and Pax6 antibod-
tibody diluted in 1⫻ PBS (Lim1, 1: ies respectively. They also thank the
Immunohistochemistry 2,500 and Lim3, 1:200) or in 2% goat members of the Adler lab for helpful
Chick embryos were euthanized by de- serum containing 0.05% Triton X-100 discussions and Jane Cione for sec-
capitation before ED10, and by halo- (AP2␣, 1:350). Control reactions retarial assistance. R.A. was funded
thane overdose thereafter. Whole showed that, at these concentrations, by the NIH and is the Arnall Patz
heads were fixed at ED3– 4, and enu- the antibodies were detectable after Distinguished Professor of Ophthal-
cleated eyes (devoid of cornea, lens, tyramide amplification but undetect- mology.
and vitreous body) were used thereaf- able with fluorescent secondary anti-
ter. At ED12, 15, and 18, moreover, bodies. Antibody binding was detected
the dorso-fundal region of the optic with sequential incubations with bio- REFERENCES
cup was isolated with surgical blades tinylated goat anti-mouse (1:1,000), Adler R. 2000. A model of retinal cell dif-
for fixation. Tissues were fixed in cold horseradish peroxidase–streptavidin ferentiation in the chick embryo. Prog
acetic ethanol (75% ethanol - 25% ace- (1:1,000) and fluorescein-tyramide (1: Retin Eye Res 19:529 –557.
766 CANTO-SOLER ET AL.

Arranz V, Dreuillet C, Crisanti P, Tillit J, across the neuroepithelium during reti- Ladomery M, Dellaire G. 2002. Multifunc-
Kress M, Ernoult-Lange M. 2001. The naldevelopment.Development131:1343– tional zinc finger proteins in develop-
zinc finger transcription factor, MOK2, 1351. ment and disease. Ann Hum Genet 66:
negatively modulates expression of the El-Kady A, Klenova E. 2005. Regulation of 331–342.
interphotoreceptor retinoid-binding pro- the transcription factor, CTCF, by phos- Lai ZC, Fetchko M, Li Y. 1997. Repression of
tein gene, IRBP. J Biol Chem 276:11963– phorylation with protein kinase CK2. Drosophila photoreceptor cell fate through
11969. FEBS Lett 579:1424 –1434. cooperative action of two transcriptional
Baniahmad A, Steiner C, Kohne AC, Ren- Filippova GN, Fagerlie S, Klenova EM, repressors Yan and Tramtrack. Genetics
kawitz R. 1990. Modular structure of a Myers C, Dehner Y, Goodwin G, Neiman 147:1131–1137.
chicken lysozyme silencer: involvement PE, Collins SJ, Lobanenkov VV. 1996. Lai ZC, Li Y. 1999. Tramtrack69 is posi-
of an unusual thyroid hormone receptor An exceptionally conserved transcrip- tively and autonomously required for
binding site. Cell 61:505–514. tional repressor, CTCF, employs differ- Drosophila photoreceptor development.
Begemann G, Paricio N, Artero R, Kiss I, ent combinations of zinc fingers to bind Genetics 152:299 –305.
Perez-Alonso M, Mlodzik M. 1997. diverged promoter sequences of avian Lerner LE, Peng GH, Gribanova YE, Chen
muscleblind, a gene required for photo- and mammalian c-myc oncogenes. Mol S, Farber DB. 2005. Sp4 is expressed in
receptor differentiation in Drosophila, Cell Biol 16:2802–2813. retinal neurons, activates transcription
encodes novel nuclear Cys3His-type Ganss B, Jheon A. 2004. Zinc finger tran- of photoreceptor-specific genes, and syn-
zinc-finger-containing proteins. Develop- scription factors in skeletal develop- ergizes with Crx. J Biol Chem 280:20642–
ment 124:4321–4331. ment. Crit Rev Oral Biol Med 15:282– 20650.
Belecky-Adams T, Tomarev S, Li HS, 297. Li T, Lu Z, Lu L. 2004. Regulation of eye
Ploder L, McInnes RR, Sundin O, Adler Hamburger V, Hamilton HL. 1951. A se- development by transcription control of
R. 1997. Pax-6, Prox 1, and Chx10 ho- ries of normal stages in the development CCCTC binding factor (CTCF). J Biol
meobox gene expression correlates with of the chick embryo. J Morphol 88:49 – Chem 279:27575–27583.
phenotypic fate of retinal precursor cells. 92. Li T, Lu Z, Lu L. 2006. Pax6 regulation in
Invest Ophthalmol Vis Sci 38:1293– retinal cells by CCCTC binding factor.
Harpavat S, Cepko CL. 2003. Thyroid hor-
1303. Invest Ophthalmol Vis Sci 47:5218 –
mone and retinal development: an
Bowmaker JK, Heath LA, Wilkie SE, Hunt 5226.
emerging field. Thyroid 13:1013–1019.
DM. 1997. Visual pigments and oil drop- Lobanenkov VV, Nicolas RH, Adler VV,
lets from six classes of photoreceptor in Huang H, Frank MB, Dozmorov I, Cao W,
Paterson H, Klenova EM, Polotskaja AV,
the retinas of birds. Vision Res 37:2183– Cadwell C, Knowlton N, Centola M,
Goodwin GH. 1990. A novel sequence-
2194. Anderson RE. 2005. Identification of
specific DNA binding protein which in-
Burke LJ, Hollemann T, Pieler T, Renka- mouse retinal genes differentially regu-
teracts with three regularly spaced di-
witz R. 2002a. Molecular cloning and ex- lated by dim and bright cyclic light rear-
rect repeats of the CCCTC-motif in the
pression of the chromatin insulator pro- ing. Exp Eye Res 80:727–739.
5⬘-flanking sequence of the chicken
tein CTCF in Xenopus laevis. Mech Dev Huang H, Wahlin K, McNally M, Irving N, c-myc gene. Oncogene 5:1743–1753.
113:95–98. Adler R. 2007. Developmental regulation Lopez R, Lopez-Gallardo M, Busturia I,
Burke LJ, Zhang R, Lutz M, Renkawitz R. of muscleblind-like (MBNL) gene expres- Anezary L, Prada C. 2005. Spatial and
2002b. The thyroid hormone receptor sion in the chicken embryo retina. Dev temporal patterns of growth and differ-
and the insulator protein CTCF: two dif- Dyn (in press). entiation of cone oil droplets in the chick
ferent factors with overlapping func- Kanadia RN, Johnstone KA, Mankodi A, retina. J Neurosci Res 79:401–411.
tions. J Steroid Biochem Mol Biol 83:49 – Lungu C, Thornton CA, Esson D, Tim- Lutz M, Burke LJ, LeFevre P, Myers FA,
57. mers AM, Hauswirth WW, Swanson MS. Thorne AW, Crane-Robinson C, Bonifer
Canto-Soler MV, Adler R. 2006. Optic cup 2003a. A muscleblind knockout model C, Filippova GN, Lobanenkov V, Renka-
and lens development requires Pax6 ex- for myotonic dystrophy. Science 302: witz R. 2003. Thyroid hormone-regu-
pression in the early optic vesicle during 1978 –1980. lated enhancer blocking: cooperation of
a narrow time window. Dev Biol 294:119 – Kanadia RN, Urbinati CR, Crusselle VJ, CTCF and thyroid hormone receptor.
132. Luo D, Lee YJ, Harrison JK, Oh SP, EMBO J 22:1579 –1587.
Chiang MK, Melton DA. 2003. Single-cell Swanson MS. 2003b. Developmental ex- Mader MM, Cameron DA. 2006. Effects of
transcript analysis of pancreas develop- pression of mouse muscleblind genes induced systemic hypothyroidism upon
ment. Dev Cell 4:383–393. Mbnl1, Mbnl2 and Mbnl3. Gene Expr the retina: regulation of thyroid hormone
Curto GG, Lara JM, Parrilla M, Aijon J, Patterns 3:459 –462. receptor alpha and photoreceptor pro-
Velasco A. 2007. Modifications of the ret- Klenova EM, Nicolas RH, Paterson HF, duction. Mol Vis 12:915–930.
ina neuronal populations of the het- Carne AF, Heath CM, Goodwin GH, Marinescu VD, Kohane IS, Riva A. 2005a.
erozygous mutant small eye mouse, the Neiman PE, Lobanenkov VV. 1993. CTCF, The MAPPER database: a multi-genome
Sey(Dey). Brain Res 1127:163–176. a conserved nuclear factor required for op- catalog of putative transcription factor
Decembrini S, Andreazzoli M, Vignali R, timal transcriptional activity of the binding sites. Nucleic Acids Res 33:D91–
Barsacchi G, Cremisi F. 2006. Timing chicken c-myc gene, is an 11-Zn-finger pro- D97.
the generation of distinct retinal cells by tein differentially expressed in multiple Marinescu VD, Kohane IS, Riva A. 2005b.
homeobox proteins. PLoS Biol 4:e272. forms. Mol Cell Biol 13:7612–7624. MAPPER: a search engine for the com-
Della NG, Senior PV, Bowtell DD. 1993. Klenova EM, Nicolas RH, U S, Carne AF, putational identification of putative
Isolation and characterisation of murine Lee RE, Lobanenkov VV, Goodwin GH. transcription factor binding sites in mul-
homologues of the Drosophila seven in 1997. Molecular weight abnormalities of tiple genomes. BMC Bioinformatics 6:79.
absentia gene (sina). Development 117: the CTCF transcription factor: CTCF mi- Marquardt T, Ashery-Padan R, Andrejew-
1333–1343. grates aberrantly in SDS-PAGE and the ski N, Scardigli R, Guillemot F, Gruss P.
Dickson BJ. 1998. Photoreceptor develop- size of the expressed protein is affected 2001. Pax6 is required for the multipo-
ment: breaking down the barriers. Curr by the UTRs and sequences within the tent state of retinal progenitor cells. Cell
Biol 8:R90 –R92. coding region of the CTCF gene. Nucleic 105:43–55.
Dulac C. 1998. Cloning of genes from single Acids Res 25:466 –474. Merzdorf CS. 2007. Emerging roles for zic
neurons. Curr Top Dev Biol 36:245–258. Klenova EM, Morse HC III, Ohlsson R, genes in early development. Dev Dyn 236:
Dunn KL, Davie JR. 2003. The many roles Lobanenkov VV. 2002. The novel BO- 922–940.
of the transcriptional regulator CTCF. RIS ⫹ CTCF gene family is uniquely in- Mitton KP, Swain PK, Khanna H, Dowd M,
Biochem Cell Biol 81:161–167. volved in the epigenetics of normal biol- Apel IJ, Swaroop A. 2003. Interaction of
Edqvist PH, Hallbook F. 2004. Newborn ogy and cancer. Semin Cancer Biol 12: retinal bZIP transcription factor NRL
horizontal cells migrate bi-directionally 399 –414. with Flt3-interacting zinc-finger protein
 CTCF AND PAX6 EXPRESSION DURING RETINAL CELL DIFFERENTIATION 767

Fiz1: possible role of Fiz1 as a transcrip- Philips GT, Stair CN, Young Lee H, Wrob- differentiation of photoreceptor and non-
tional repressor. Hum Mol Genet 12:365– lewski E, Berberoglu MA, Brown NL, photoreceptor neurons. Invest Ophthal-
373. Mastick GS. 2005. Precocious retinal mol Vis Sci 43:3522–3529.
Moses K, Ellis MC, Rubin GM. 1989. The neurons: Pax6 controls timing of differ- Wahlin KJ, Lim L, Grice EA, Campochiaro
glass gene encodes a zinc-finger protein entiation and determination of cell type. PA, Zack DJ, Adler R. 2004. A method
required by Drosophila photoreceptor Dev Biol 279:308 –321. for analysis of gene expression in iso-
cells. Nature 340:531–536. Roberts MR, Srinivas M, Forrest D, Mor- lated mouse photoreceptor and Muller
Moses K, Rubin GM. 1991. Glass encodes a reale de Escobar G, Reh TA. 2006. Mak- cells. Mol Vis 10:366 –375.
site-specific DNA-binding protein that is ing the gradient: thyroid hormone regu- Wahlin K, Moreira E, Huang H, Yu N,
regulated in response to positional sig- lates cone opsin expression in the Adler R. 2008. Molecular dynamics of
nals in the developing Drosophila eye. developing mouse retina. Proc Natl Acad photoreceptor synapse formation in the
Genes Dev 5:583–593. Sci U S A 103:6218 –6223. developing chick retina. J Comp Neurol
Ng L, Hurley JB, Dierks B, Srinivas M, Salto Schedl A, Ross A, Lee M, Engelkamp D, 506:822– 837.
C, Vennstrom B, Reh TA, Forrest D. 2001. Rashbass P, van Heyningen V, Hastie
A thyroid hormone receptor that is re- ND. 1996. Influence of PAX6 gene dos- Wu D, Li T, Lu Z, Dai W, Xu M, Lu L. 2006.
quired for the development of green cone age on development: overexpression Effect of CTCF-binding motif on regula-
photoreceptors. Nat Genet 27:94 –98. causes severe eye abnormalities. Cell tion of PAX6 transcription. Invest Oph-
Ohlsson R, Renkawitz R, Lobanenkov V. 86:71–82. thalmol Vis Sci 47:2422–2429.
2001. CTCF is a uniquely versatile tran- Sharma S, Dimasi D, Higginson K, Della Yanagi Y, Takezawa S, Kato S. 2002. Dis-
scription regulator linked to epigenetics NG. 2004. RZF, a zinc-finger protein in tinct functions of photoreceptor cell-spe-
and disease. Trends Genet 17:520 –527. the photoreceptors of human retina. cific nuclear receptor, thyroid hormone
Otteson DC, Liu Y, Lai H, Wang C, Gray S, Gene 342:219 –229. receptor beta2 and CRX in one photore-
Jain MK, Zack DJ. 2004. Kruppel-like Tietjen I, Rihel JM, Cao Y, Koentges G, ceptor development. Invest Ophthalmol
factor 15, a zinc-finger transcriptional Zakhary L, Dulac C. 2003. Single-cell Vis Sci 43:3489 –3494.
regulator, represses the rhodopsin and transcriptional analysis of neuronal pro- Zhang X, Heaney S, Maas RL. 2003. Cre-
interphotoreceptor retinoid-binding pro- genitors. Neuron 38:161–175. loxp fate-mapping of Pax6 enhancer ac-
tein promoters. Invest Ophthalmol Vis Toy J, Norton JS, Jibodh SR, Adler R. tive retinal and pancreatic progenitors.
Sci 45:2522–2530. 2002. Effects of homeobox genes on the Genesis 35:22–30.