doi: 10.1242/10.1242/dev.00326
Irx4-mediated regulation of Slit1 expression contributes to the definition of early axonal paths inside the retina
Zhe Jin1,
Jinhua Zhang1,
Avihu Klar2,
Alain Chédotal3,
Yi Rao4,
Constance L. Cepko5 and
Zheng-Zheng Bao1,*
1 Department of Medicine, University of Massachusetts Medical School, 364
Plantation Street, Worcester, MA 01605, USA
2 Department of Anatomy and Cell Biology, The Hebrew University-Hadassah Medical
School, PO Box 12272, Jerusalem 91120, Israel
3 Institut de la Santé et de le Recherche Médicale U106,
Bâtiment de Pédiatrie, Hôpital de la
Salpêtrière, 47 Boulevard de I'Hôpital, 75013, Paris,
France
4 Department of Anatomy and Neurobiology, Washington University School of
Medicine, Box 8108, 660 S. Euclid Avenue, St Louis, MO 63110, USA
5 Department of Genetics, Harvard Medical School, 200 Longwood Avenue, Boston,
MA 02115, USA
View larger version (27K):
[in a new window]
|
Fig. 1. Intra-retinal axon targeting. (A) To facilitate analysis and observation,
retinas were flat-mounted by making a few cuts at the edge. (B) The cells in
the GCL and the axons in OFL can be stained and analyzed on the flat-mount
retina with the vitreal side upwards. (C) Retinal ganglion cells in ganglion
cell layer (GCL) extend axons to optic fiber layer (OFL) where they project
towards the optic disc (OD) at the center of the retina.
|
|
View larger version (126K):
[in a new window]
|
Fig. 2. mRNA expression of Irx4 and the Slit/Robo family genes in
chicken retina. In situ hybridization with various probes on retinal sections:
(A) Irx4 probe, E7 retina; (B) Irx4 probe, E12 retina; (C)
Slit1 probe, E4.5 retina; (D) Slit1 probe, E6 retina; (E)
Robo1 probe, E9 retina. Note that Irx4, Slit1 and
Robo1 are all expressed in subsets of cells in GCL. (F) Two-color in
situ hybridization on E7 flat-mount retina, with Slit1
(fluorescein-labeled, shown as red) and Irx4 probes (DIG-labeled,
shown as purple). Note that Slit1 and Irx4 are mostly
localized in two distinct cell populations in GCL. Scale bars: 100 µm.
|
|
View larger version (86K):
[in a new window]
|
Fig. 3. Irx4 specifically downregulates the expression of Slit1 in GCL.
(A) Replication-competent retroviral constructs for expressing full-length
Irx4 protein (RCAS-Irx4) or dominant-negative Irx4 protein (RCAS-DN-Irx4).
(B-G) Retinas infected with the RCAS-Irx4 virus were hybridized with the Slit1
probe (B) or Robo1 probe (D), or stained with anti-Islet 1 antibody (F). All
samples were also stained with the anti-viral GAG antibody to show the area of
infection (C,E,G). Note the RCAS-Irx4-infected areas had few
Slit1-expressing cells (marked with red broken lines). The samples
infected with the dominant-negative Irx4 virus (H,I) or the control virus
expressing only the Engrailed repressor domain (J,K) were similarly hybridized
with the Slit1 probe (H,J) and stained with the anti-viral GAG antibody (I,K).
Note the Slit1 expression was activated in the areas infected with
the RCAS-DN-Irx4 virus (arrowheads in H,I) but not the control virus. Scale
bar: 100 µm.
|
|
View larger version (131K):
[in a new window]
|
Fig. 4. Intraretinal axonal phenotype caused by RCAS-Irx4 virus infection. Retina
development is more advanced in the center than in the periphery. The axons in
the wild-type E8 retina were stained with the anti-neurofilament antibody
270.7 (A,B). (A) At the periphery, axons join and leave fascicles, producing a
`honeycomb' appearance. (B) Close to the center of the retina, the axons
appear more mature and fasciculated. (C-H) Optic vesicles were infected with
RCAS-Irx4 virus (E-H) or control RCAS-GFP (C,D) at HH stage 10-11 (E1.5) and
the infected retinas were harvested at E8. Flat-mounts of retinas were double
stained with a mouse monoclonal antibody recognizing neurofilament (270.7)
(C,E,G) and a rabbit polyclonal antibody recognizing viral antigen (anti-p27)
(D,F,H). Images in C,E and G are in the same fields as in D,F and H,
respectively. The broken white arrows in A-C,E,G indicate the direction of
axon projection toward the optic disc. (C,D) The axons in the control RCAS-GFP
virus-injected retina appear normal. (E,F) Close to the infected/uninfected
boundaries, some axons appear to turn slightly towards the uninfected area
(arrowhead in E). (G,H) In the center of the infected area, an abrupt increase
of fasciculation of the axon bundles was observed (arrowheads in G) when axons
went from an uninfected (UI) to an infected (Ib) area. The axons returned to
wild-type appearance when they went from infected (Ia) to uninfected (UI)
area, suggesting that the changes in axonal morphology in the infected area
were reversible. Scale bar: 100 µm.
|
|
View larger version (176K):
[in a new window]
|
Fig. 5. Intraretinal axonal phenotype caused by RCAS-Irx4 virus infection, viewed
at lower magnification. The retinas were infected at stage 10-11 and harvested
at E7 (A-D) or E12 (E,F). Axons were stained with an anti-neurofilament
antibody (A,C,E,F), whereas the infected area was visualized by staining with
an anti-viral GAG antibody (p27) (B,D). Images in A,C are in the same fields
as in B,D, respectively. The broken white arrows in A,C,E,F indicate the
direction of axon projection towards the optic disc. Note the axons turned to
avoid the infected area (marked with IF in A,C). However, the area that
exhibited phenotype was smaller than the infected area. Some newly infected
areas with only scattered infection (marked with and asterisk in A-D) did not
show axonal abnormality. (E) At E12, the retina was completely infected with
the RCAS-Irx4 virus (data not shown); however, the area with the phenotype did
not expand to the entire retina. Some areas still appeared normal (marked with
an asterisk), indicating that axons no longer responded to low Slit1 levels
after they had passed through the area. (F) The axons in the control virus
RCAS-GFP-infected retina appeared completely normal at E12. Scale bar: 200
µm.
|
|
View larger version (134K):
[in a new window]
|
Fig. 6. In vivo assay to test the function of Slit1 in retina. (A-C) Design of the
in vivo assay. (A) A full-length Myc-tagged Slit1 expression construct was
transfected into retinal cells by in ovo electroporation. The transfected
cells were identified using anti-Myc antibody (green), whereas the axons were
stained with anti-neurofilament antibody (red). Because some transfected cells
will be localized in the cellular GCL, immediately beneath OFL where the
retinal axons travel, we will be able to determine the effect of Slit1 on
axons by comparing the relative positions of the Slit1-transfected
cells and the axons. (B) If Slit1 is attractive to axons, axons will
preferentially overlie the Slit1-transfected cells, appearing superimposed on
the cells. (C) Alternatively, if Slit1 is repulsive to axons, the axons will
avoid the transfected cells, thus not becoming superimposed on the cells. The
actual data of Slit1 transfection are shown in D and F. Most of the
Slit1-transfected cells align with the axon bundles (blue circles). (E,G) The
control GFP-transfected cells, however, appear random with regard to the
position of the axons. The blue circles indicate the cells that align
accurately with the axons while the white circles indicate the cells that do
not align with the axons. The broken white arrows in D-G indicate the
direction of axon projection towards the optic disc. Scale bar: 150 µm.
|
|
View larger version (164K):
[in a new window]
|
Fig. 7. Slit1 can rescue the axonal phenotype caused by RCAS-Irx4 infection. Slit1
expression construct was co-electroporated with the RCAS-Irx4 construct. As
the RCAS-Irx4-transfected cells would produce infectious viral particles that
spread inside the retina, the infected area was larger than the
Slit1-transfected area. Note in the area where a large number of
Slit1-transfected cells were present (R in B; arrowheads in A,B),
axonal phenotypes caused by RCAS-Irx4 infection were largely corrected.
However, in the area that did not have Slit1-transfected cells
(marked with NR in B), the axons were unevenly distributed and excessively
fasciculated. (C) In the area where only a small number of Slit1-transfected
cells were present, axonal phenotype was partially corrected. Interestingly,
most of the transfected cells appear to align accurately with the axon bundles
(arrowheads in C). The area shown in A-C was completely infected by RCAS-Irx4,
confirmed by anti-GAG staining (D and data not shown). (E,F) As a control, an
expression construct encoding GFP was co-electroporated with the RCAS-Irx4
construct. Although a large number of GFP-transfected cells were present in
this area (visible over the staining of anti-GAG antibody, arrowheads in F),
GFP could not rescue the axonal phenotype of RCAS-Irx4 infection and the axons
appeared uneven and excessively fasciculated (E). The broken white arrows in
A-C,E indicate the direction of axon projection toward the optic disc. Scale
bar: 100 µm.
|
|
View larger version (128K):
[in a new window]
|
Fig. 8. Slit1 defines the trajectory of the early retinal axons in OFL. In situ
hybridization was carried out on E7 retinas with the Slit1 probe,
followed by immunofluorescent staining of axons. The cells positive for the
Slit1 probe appear purple in the bright-field image (B) but appear as
dark spots in the fluorescent images (A,C,D,E). Note the growth cones of the
elongating axons appear to project straight towards the
Slit1-expressing cells (arrowheads in C,D), and the early retinal
axon trajectories superimposed on the Slit1-positive cells
(arrowheads in A,B). In addition, Slit1-expressing cells did not
appear to have axons (marked with an asterisk in C,D), and axons are likely to
be extended from the cells that are negative for Slit1 expression
(marked with `?' in D). (E) Close to the optic disc, more mature axons do not
align accurately with the Slit1-expressing cells. (F) Working model
of the role of Slit1 and Irx4 in intra-retinal axon targeting.
|
|
© The Company of Biologists Ltd 2003
