First published online 3 August 2005
doi: 10.1242/dev.01960
Development 132, 3907-3921 (2005)
Published by The Company of Biologists 2005
A bHLH transcriptional network regulating the specification of retinal ganglion cells
Lidia Matter-Sadzinski1,2,
Monika Puzianowska-Kuznicka3,4,
Julio Hernandez2,
Marc Ballivet2 and
Jean-Marc Matter1,2,*
1 University of Lausanne, Eye Hospital Jules Gonin and Institute for Research in
Ophthalmology, 15 avenue de France, 1004 Lausanne, Switzerland
2 University of Geneva, Sciences II, Biochemistry Department, 30 quai
Ernest-Ansermet, 1211 Geneva, Switzerland
3 Polish Academy of Sciences, Medical Research Center, Department of
Endocrinology, ul. Banacha 1a, 02-097 Warsaw, Poland
4 Medical Center of Postgraduate Education, Department of Clinical Biochemistry,
Marymoncka 99, 01-813 Warsaw, Poland
View larger version (83K):
[in a new window]
|
Fig. 1. Several bHLH factors pattern the early retinal neuroepithelium. (A) At
stage 14, HES1 transcripts accumulate in discrete domains in the eyecup (ec)
and neural tube (nt). There is no detectable accumulation of HES1 transcripts
in the central region (red bracket) of the presumptive retina. (B,C) The first
ATH5- and NGN2-expressing cells are detected in the central retina (nr) at
stage 15. (D) At stage 16, ASH1 transcripts are not detected in retina. (E) At
stage 17, a robust accumulation of HES1 transcripts is taking place throughout
the peripheral retina (J). In the central retina, a few cells located on the
vitreous side express HES1 at a high level (arrowheads in E,I). (F,K) Most
cells in the central retina express ATH5 and those expressing ATH5 strongly
are mostly located on the vitreous side. (H) At stage 18, the sparse cells
expressing Neuro M are scattered across the central retina (arrowhead in M).
There are no cells expressing ATH5 or Neuro M in the HES1 domain (F,H,L,N).
(O) Quantification of in situ hybridization. Adjacent retinal sections were
hybridized with the indicated bHLH riboprobes at stage 18. The ATH5, Neuro M
and NGN2 domains coincide in the central retina and they abut on the
peripheral HES1 domain. ASH1 is detected in an annular sector (G, brackets) at
the interface between the HES1 and ATH5 domains. l, lens. Scale bar: 140 µm
in A; 80 µm in B,C; 100 µm in D-H.
|
|
View larger version (97K):
[in a new window]
|
Fig. 2. Growth of the retina is accompanied by changes in the patterning of
progenitor cells. (A,B,E) Between stages 18 and 26, the ATH5 domain expands in
register with the threefold increase in retina diameter. (C) HES1 transcripts
are abundant at the periphery and sparse in the central region. Scattered
cells expressing HES1 at a high level are detected in the central retina
(arrowheads in L). The HES1 and ATH5 expression domains are complementary
(B,C), but the anterior margin of the ATH5 domain overlaps the posterior HES1
region, where HES1 transcript levels are decreasing (brackets in H,I). ATH5
and NGN2 transcripts accumulate in the posterior retina (B,D). The NGN2 domain
extends beyond that of ATH5 (brackets in J,K). At stage 26, ASH1- and
ATH5-expressing cells are interspersed in the posterior retina. ASH1 extends
beyond ATH5 (arrowheads in E,F). At stage 30, ATH5 transcripts are distributed
throughout the whole retina (G), except at the ciliary margin (arrows). ATH5
transcripts are not evenly distributed across the retina. They are abundant on
the ventricular side of the proliferative zone (pz) (inset in G). At stage 30,
HES1 expression is downregulated both in the peripheral and in the central
retina (M). Sections in L and M were counterstained with Toluidine Blue. Scale
bar: 380 µm in A-G; 40 µm in L; 60 µm in M; 240 µm in H,I; 150
µm in J,K.
|
|
View larger version (103K):
[in a new window]
|
Fig. 3. Activity of the electroporated ATH5 promoter in stage 22-23 retina. (A)
When controlled by the ubiquitous CMV promoter, GFP and lac reporters
are both expressed in the electroporated peripheral retina. (B-F) The
ATH5-promoter/lac and the CMV-promoter/GFP reporter plasmids were
electroporated alone (B,C), in combination with a NGN2 expression vector (D,E)
or in combination with NGN2 and HES1 expression vectors (F). (B) GFP-positive
cells are distributed throughout the peripheral (p) and central (c) retina,
whereas lac+ cells are confined to the central region
(arrowhead). (C) lac+ cells (arrowheads) are sparse in the
central retina. (D) Overexpression of NGN2 increases the proportion of
lac+ cells in the central, but not in the peripheral
retina (arrowhead in E). (F) No lac+ cells were detected
when both NGN2 and HES1 were overexpressed in the central retina. Data
presented in each panel are representative of at least five independent
experiments. l, lens. Scale bar: 170 µm in A; 120 µm in B; 30 µm in
C; 40 µm in D,F; 80 µm in E.
|
|
View larger version (42K):
[in a new window]
|
Fig. 4. Transcriptional analysis of ATH5-expressing single cells. Stage 22-23
(E3.5) retinal cells were transfected with an ATH5-promoter/GFP-reporter
plasmid either singly or in combination with a vector expressing NGN2. They
were cultured for either 24 (E4.5) or 48 hours (E5.5). Stage 26 (E5) retinal
cells were transfected with the ATH5-promoter/GFP-reporter plasmid and
cultured for 24 hours (E6). Individual GFP-positive cells were collected and
processed for single-cell RT-PCR using the primers listed in Table S1 (see
supplementary material). (A) Representative transcriptional profiles obtained
with a set of 39 cells from the five groups generated by the experiment, as
identified by the colour code in B. RT-PCRs of total RNA isolated from E8
retina (NRE8) were used as positive controls for each set of primers. (C)
Ratios of HES1-, Neuro M-, Delta 1- and ASH1-positive cells to the total
number of cells tested for expression of these genes.
|
|
View larger version (34K):
[in a new window]
|
Fig. 5. Co-expression of RGC-specific genes and bHLH transcription factors in
newborn RGCs and RGC precursors. Cells were transfected with an
ATH5-promoter/GFP-reporter plasmid at stage 26 (E5) and cultured for 24 hours.
(A, right) Transcriptional profile of a newborn RGC. This neuron-like
GFP-positive cell (left) expresses ATH5, Neuro M, ß3 and BRN3C, but not
ASH1. (B) ATH5-expressing cells do not always co-express Neuro M, ß3 and
BRN3C. (C) Colocalization of ß3 promoter activity and HES1 or Neuro M
expression. Cells were transfected with a
ß3-promoter/lacZ-reporter plasmid at stage 24. After 24 hours in
culture, lacZ-expression was revealed and cells were processed for in
situ hybridization with (a) HES1- or (b,c) Neuro M-specific riboprobes.
Arrowhead in a indicates a double-labelled cell.
|
|
View larger version (37K):
[in a new window]
|
Fig. 6. Regulation of the ATH5 promoter during retinogenesis. (A) Retinal cells
isolated at stages 22 (E3) to 37 (E12) were transfected with an
ATH5-promoter/CAT-reporter plasmid singly or in combinations with ATH5 and/or
NGN2 expression vectors. Cells were assayed for CAT activity 24 hours after
transfection. ATH5 transcription is passing through three phases in the course
of retinogenesis. During the first phase (HH22-HH24), the promoter is weakly
active and responds strongly to NGN2 overexpression, except in the presence of
ATH5. During the second phase (HH25-HH30), upregulation of promoter activity
coincides with a transient increase in ATH5 mRNA (curve). ATH5 and NGN2 both
enhance promoter activity and ATH5 becomes dominant over NGN2. The third phase
(HH34 and beyond) sees a decrease in ATH5 mRNA and is marked by the inability
of either proneural protein to transactivate the promoter. (B) Retinal cells
isolated at stages 22-23 and 29-30 were transfected with an
ATH5-promoter/lacZ-reporter plasmid singly or in combinations with
NGN2 or ATH5 expression vectors. lac+ cells were counted
after 24 hours in culture. The number of lac+ cells
obtained upon transfection with a control SV40-promoter/lacZ-reporter
plasmid at each stage is set at 100 and cell numbers are given relative to
this value. (C) Schematic representation of promoter activity as revealed by
X-gal and CAT assays. Approximately 30% of cells express ATH5 at stages 22-23
and 29-30. The horizontal arrows indicate average promoter activity as
measured by CAT assay, the open arrowhead marks the threshold for X-gal
detection. At stage 22-23, promoter activity is low and only one in 30
ATH5-expressing cells is detected by X-gal. Overexpression of NGN2 increases
promoter activity 10-fold but only six out of 30 ATH5-expressing cells are
stained with X-gal. At stage 29-30, the whole population of ATH5-expressing
cells is stained with X-GAL upon ATH5 overexpression. (D) At stage 22-23, most
cells are weakly stained with X-GAL (arrowheads in a). Overexpression of NGN2
strongly enhances promoter activity (b) and the number of X-gal stained cells.
At stage 29-30, cells display strong promoter activity (c) and overexpression
of ATH5 enhances staining intensity (d).
|
|
View larger version (38K):
[in a new window]
|
Fig. 7. HES1 exerts a dominant-negative effect upon the ATH5 promoter. (A) Retinal
cells at stages 22-23 or 29-30 were transfected with an
ATH5-promoter/CAT-reporter plasmid alone or with different combinations of the
ATH5, NGN2 and HES1 expression vectors. NGN2 and HES1 expression vectors were
co-transfected in different ratios, as indicated. (B) Peripheral and central
regions of retina were dissected at stage 22-23. They were electroporated with
NGN2 and control expression vectors and cultured as explants for 24 hours. The
presence of ATH5 mRNA was detected by northern blot hybridisation.
Overexpression of NGN2 upregulated ATH5 expression in the central but not in
the peripheral retina. (C) Retinal cells at stages 24 or 29-30 were
transfected with an ATH5-promoter/lacZ-reporter plasmid singly or in
combinations with ATH5 or NGN2 expression vectors. lac+
cells were revealed and processed for in situ hybridization with a
HES1-specific riboprobe. Overexpression of NGN2 increased the relative number
of double-labelled cells, indicating that the NGN2 protein can activate the
ATH5 promoter in cells that express HES1 (a), unlike the ATH5 protein (b).
|
|
View larger version (85K):
[in a new window]
|
Fig. 8. ATH5 expression is upregulated during the last S phase. (A) Retinal cells
isolated at stages 22-23 and 28-29 were transfected with an
ATH5-promoter/lacZ-reporter plasmid singly or in combinations with
NGN2 or ATH5 expression vectors and pulse-labelled with
[3H]-thymidine at the end of a 24-hour culture period. (Left) The
number of lac+ cells counted when the reporter plasmid was
transfected alone is set at 1. (Right) At stage 22-23, overexpression of NGN2
enhances promoter activity in proliferating cells (a) and increases the pool
of nonradioactive cells whose ATH5 promoter is upregulated (b). At stage
28-29, lac+ cells whose promoter is strongly upregulated
are unlabelled (d). The detection of double-labelled cells (c) and their
increased number upon ATH5 overexpression indicate that ATH5 promoter activity
is upregulated during the S phase. (B, left) A retina at stage 29-30 was
pulse-labelled for 45 minutes with BrdU and chased for 15 minutes. Transverse
sections were hybridized with an ATH5-specific riboprobe. Most BrdU-positive
cells are in S phase and their nuclei are located on the vitreous side (vi) of
the pz. ATH5 transcripts accumulate on the ventricular side (ve) of the pz in
the region where cells are in the G1 and G2 phases of the cell cycle. A few
BrdU-positive nuclei are located in this region (arrowheads). (Right)
Schematic of mitosis in the pz. Scale bar: 40 µm.
|
|
View larger version (27K):
[in a new window]
|
Fig. 9. Interacting transcriptional patterns as retinal cells go through three
consecutive phases during the conversion of progenitors into newborn RGCs.
|
|
© The Company of Biologists Ltd 2005
