Vitreous
Vein
Optic
nerve
Artery
Superficia
l plexus
Intermed
iate plex
us GCL
IPL
Deep ple
xus INL
Sclera
Supplemental Figure 1
Schematic diagram illustrating development of the vascular networks in the murine
retina. The retinal vasculature forms as endothelial cells migrate from the optic nerve
onto the retinal surface at birth and progress radially to form the superficial (or inner)
plexus. Around postnatal day 7 (P7), sprouting vessels descend and advance into the
OPL where they establish the deep plexus. At P11-12 stages, sprouting vessels from
the deep plexus ascend into the IPL and ramify to form the intermediate plexus. GCL,
ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer
plexiform layer; ONL, outer nuclear layer.
Figure-S1 (Friedlander) A
B
tdTomato/DAPI
GCL
IPL
C
chx10 MAP2 DAPI
GCL
tdTomato/Stx1 (pan-amacrine neurons)
IPL
IPL
INL
INL
tdTomato/GAD (GABAergic)
tdTomato/Glyt1 (glycinergic)
INL
ONL
OPL
GCL
IPL
IPL
E
D
GS MAP2 DAPI
GFP/DAPI
GCL
Stx1
tdTomato/Calbindin (pan-horizontal cells)
Glyt1 (glycinergic)/GAD (GABAergic)
GAD
INL
INL
Glyt1
ONL
0
50
100
Percentage of amacrine cells
H
G
F
GFP CD31 NF DAPI
IPL
CD31 tomato MAP-2
GFP CD31 DAPI
IPL
j
IPL
INL
INL
OPL
INL
OPL
GCL
Supplemental Figure 2
Putative neurovascular units in the INL. (A) Whole retinal sections from mice harboring
two Cre-recombination reporters and ptf1a-Cre. (B) IHC for bipolar (Chx10; top panel)
and Mueller glia (glutamine synthetase (GS); bottom panel) is shown to highlight that
their locations in the INL are distinct from amacrine cells. (C) Cre-mediated
recombination occurs in amacrine cells in the inner margin of the INL and colocalizes
with a pan-amacrine cell marker, Syntaxin 1 (Stx1), a GABAergic amacrine cell marker
(GAD), and glycinergic amacrine cell marker (Glyt1) in cryosectioned retinas. (D) The
percentage of Cre-positive amacrine cells that colocalized with amacrine cell subtypes
was determined by counting cells in cryosectioned retinas (≥250 cells were counted for
each cell type). (E) Colocalization of Cre-mediated recombination (td-Tomato) with
calbindin positive horizontal cells at P23. (F and G) IHC on thick cut sections on ptf1aCre;R26GFP/+ mice with NF-M (F) and CD31 (G). (H) Immunofluorescence for antiMAP-2 and anti-CD31 in cryosections from P23 ptf1a-Cre; R26tdTomato/+ retinas. Scale
bar: 50 μm (A-G); 20 μm (H). GCL, ganglion cell layer; IPL, inner plexiform layer; INL,
inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer. Figure-S2 (Friedlander) ptf1a; VEGFf/+ ptf1a; VEGFf/f
VEGFf/f
DP to IM
SP to IM
ptf1a; VEGFf/f
ptf1a; VEGFf/f
D
Vertical sprouts
from DP to IM/FOV
40
30
20
DP to IM
VEGFf/f
10
E
Fold change
SP to IM
C
Vertical sprouts
B
Vertical sprouts
VEGFf/f
GS-lectin
A
2.0
1.5
* *
1.0
0.5
0.0
0
VEGFf/f ptf1a; VEGFf/f
Vegf120 Vegf164 Vegf188
Supplemental Figure 3
The retinal vasculature of the superficial and deep plexuses are unaffected
by Vegfa deletion in amacrine and horizontal cells. (A) Normal
vascularization (bidirectional arrows) is observed in GS-lectin-positive P6
ptf1a-Cre; VEGFf/f retinas and controls (VEGFf/f (no Cre), ptf1a-Cre; VEGFf/
+). (B-D) Whole-mount staining in VEGFf/f or ptf1a-Cre; VEGFf/f retinas at
P12 (B) or P15 (C and D), reveals no differences in the number of vertical
sprouts descending through the IPL from the superficial plexus or ascending
through the INL from the deep plexus (n = 4). Scale bar: 1 mm (A; upper
panels); 100 μm (A; lower panels); 50 μm (B and C). (E) qPCR analysis
showed that soluble VEGF120 and VEGF164 were the most abundant Vegf
isoforms expressed in ptf1a-Cre; VEGFf/f mice at P15 (n = 4). SP; superficial
plexus, IM; intermediate plexus, DP; deep plexus. Figure-S3 (Friedlander) GS-lectin
4M
6M
12M
ptf1a; VEGFf/f
VEGFf/f
2M
Supplemental Figure 4
Chronic intermediate plexus attenuation is
observed in ptf1a-Cre; VEGF knockout mice.
The intermediate plexus capillaries in the whole
mount retinas of VEGFf/f or ptf1a-Cre; VEGFf/f
mice at 2, 4, 6, 12 months. Scale bar: 50 μm.
Figure-S4 (Friedlander) B
CD31/DAPI
VHLf/f
GCL
INL
ONL
superficial
ICG
intermediate
VHLf/f
GCL
C
Fundus
ptf1a; VHLf/f
INL
ONL
ptf1a; VHLf/f
A
deep
composite
***
Branching point/FOV
ptf1a; VHLf/f
VHLf/f
D
***
300
***
200
***
100
0
VHLf/f
at P23
ptf1a; VHLf/f
at P23
VHLf/f
at 6M
ptf1a; VHLf/f
at 6M
Supplemental Figure 5
Blood vessels do not advance to the OPL in Vhl mutants (A) Vessels sprout
from the superficial plexus towards the OPL in controls (VHLf/f mice), but are
directed towards the IPL in ptf1a-Cre; VHLf/f mice at P5 where Vegfa is most
highly expressed. (B) In vivo imaging of the ocular fundus and indocyanine
green angiography in ptf1a-Cre; VHLf/f mice and controls revealed dense
vasculature. (C and D) The dense convoluted intermediate plexus, and
attenuated superficial and deep plexuses persisted until as late as 20
months (C). Note that the abnormally high number of branching points
persists in both groups longitudinally (D) (n = 4-5). ***P<0.001; 2-tailed
Student’s t tests. Error bars indicate mean ± SD. Scale bar: 50 μm (A and
C). GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear
layer. Figure-S5 (Friedlander) VHLf/f
ptf1a; VHLf/f
VHLf/f; Hif-1αf/f
ptf1a; VHLf/f; Hif-1αf/f
VHLf/f; Hif-2αf/f
ptf1a; VHLf/f; Hif-2αf/f
VHLf/f; VEGFf/f
ptf1a; VHLf/f; VEGFf/f
H
A
VEGF-A/DAPI
GCL
B
IPL
INL
ONL
I
VEGF-A/DAPI
C
D
J
VEGF-A/DAPI
E
F
VEGF-A/DAPI
K
G
M
Fold Change L
2 ** 1.5 1 0.5 0 Vegf120 Vegf164 Vegf188 Supplemental Figure 6
Early angiogenesis events (P12) are regulated by VHL/HIF-1α/VEGF signaling. (A-F) Combinatorial conditional
knock-out strategies were employed to show that the loss of HIF-1α (B) but not HIF-2α (C) in amacrine and
horizontal cells interferes with intermediate plexus development in haplosufficient Vhl+/- mutants compared with
controls (A). (D-F) Homozygous deletion of Vhl and Hif-1α (E) prevents the neovascularization observed in Vhl
mutants, but deletion of HIF-2α elicits no effect (F) compared with controls (D). (G) Homozygous deletion of Vhl
and Vegfa also rescues the Vhl phenotype. (H) In situ hybridization for Vegfa in Vhl mutants and controls. (I) In situ
hybridization for Vegfa in double Vhl/ Hif-1α mutants and controls. (J) In situ hybridization for Vegfa in double Vhl/
Hif-2α mutants and controls. (K) In situ hybridization for Vegfa in double Vhl/ Vegfa mutants and controls. (L)
Relative mRNA expression values from qPCR gene-profiling analysis of 84 hypoxia signaling related genes in
ptf1a-Cre; VHLf/f; VEGFf/f retinas at P12 compared with controls (harboring floxed alleles but no Cre); upregulated
genes are plotted (n = 4). (M) Fold change of Vegfa isoforms in P12 ptf1a-Cre; VHLf/f; VEGFf/f retinas at P12
compared with controls. *P<0.05, **P<0.01. ***P<0.001; Scale bars: 50 μm (A-K).
Figure-S6 (Friedlander) A
B
DT: P6-P8
P23
P0
Superficial plexus (P1-P9)
deep plexus (P7-P12)
ptf1a; R26iDTR/+
ptf1a; R26+/+
ONL
IPL
INL
INL
ONL
***
-0.6 -0.45 -0.3 -0.15
0
0.15 0.3 0.45 0.6
Distance from optic disc
INL
Scotopic
*** ***
10
50 μV
20
b-wave
μV
1500
300
1000
200
500
*
100
* **
*
** **
**
** ***
0.1
0.5
2
10
0
50
b-wave
μV
100 ms
*
a-wave
μV
400
cd·s/m2
40
30
ptf1a; R26iDTR/+
0
500 μV
0
INL
*** *** **
***
25
VEGF-A/DAPI
IPL
ptf1a; R26+/+
***
G
Intermediate plexus (P11-P17)
H
75
50
F
Scotopic
GCL + IPL
100
Photopic
Thickness (μm)
Thickness (μm)
D
tomato/DAPI
Photopic
C
ptf1a; R26RiDTR/+ ptf1a; R26R+/+
E
cd·s/m2
Flicker
μV
150
100
80
100
**
*
*
50
60
40
***
***
***
20
0
4
10
20
cd·s/m2
0
4
10
20
cd·s/m2
50 ms
0
-0.6 -0.45 -0.3 -0.15 0 0.15 0.3 0.45 0.6
Supplemental Figure 7
The genetic ablation of amacrine and horizontal cells in mice results in attenuation of
the intermediate plexus and negatively affects visual function. (A) Schematic of the
experimental design for the genetic depletion of amacrine and horizontal cells in
ptf1a-Cre; R26iDTR/+ mice. (B) DT was injected daily from P6-8. (C and D) Thinning of
the GCL/IPL and INL is observed in vivo using SD-OCT (C), and quantified (D) in
P23 ptf1a-Cre; R26+/+ and ptf1a-Cre; R26iDTR/+ mice (n = 6). (E) IHC from P23 DTtreated ptf1a-Cre; R26iDTR/+:tdTomato/+ and ptf1a-Cre; R26+/+: tdTomato/+ mice after DT
treatment at P6-8 reveal a loss of the majority of amacrine and horizontal cells. (F)
DT injections from P6-8 in P23 ptf1a-Cre; R26+/+ and ptf1a-Cre; R26iDTR/+ staged
mice induced attenuation of the intermediate plexus. (G) in situ hybridization was
performed on P12 ptf1a-Cre; R26+/+ or ptf1a-Cre; R26iDTR/+ retinas with a Vegfa
probe and counterstained with DAPI. (H) Full-field ERGs performed on ptf1a-Cre;
R26iDTR/+ mice 3 months after DT treatment at P30-32 revealed reduced b waves and
gross defects in cone-driven pathways (n = 5-6). *P<0.05. **P<0.01. ***P<0.001; 2tailed Student’s t tests. Error bars indicate mean ± SD. Scale bar: 50 μm (F, G); 40
μm (E). GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer;
ONL, outer nuclear layer. Figure-S7 (Friedlander) Supplemental Figure 8
Visual acuity in ptf1a-Cre; VEGF knockout mice is significantly impaired.
Optokinetic reflexes were measured in VEGFf/f or ptf1a-Cre; VEGFf/f mice
and significant defects were observed in 3 month old mice (n = 10). Figure-S8 (Friedlander) NF CD31 DAPI
control
ptf1a; VHLf/f ptf1a; VEGFf/f GCL
IPL
INL
MAP2 CD31 DAPI
ONL
GCL
IPL
INL
ONL
Supplemental Figure 9
Neurovascular units with the amacrine and horizontal cells and intraretinal
capillaries in ptf1a-Cre; VEGF knockout and ptf1a-Cre; VHL knockout mice.
Fluorescent immunostaining for anti neurofilament (NF) or anti-MAP2 with
CD31 in retinal cryosections from P23 ptf1a-Cre; VHLf/f mice or ptf1a-Cre;
VEGFf/f mice. The nuclei of cells were counterstained with DAPI (blue).
Scale bar: 50 μm. GCL, ganglion cell layer; IPL, inner plexiform layer; INL,
inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer. Figure-S9 (Friedlander) VEGFf/f
ptf1a; VEGFf/f
VEGFf/f
B
ptf1a; VEGFf/f
TUNEL positive cells/mm2
50
40
TUNEL/DAPI
A
30
GCL
GCL
INL
INL
ONL
ONL
20
10
0
P12
C
D
IPL
INL
OPL
ptf1a; VEGFf/f; R26td-Tomato/+
ptf1a; VEGFf/+; R26td-Tomato/+
synaptophysin/td-Tomato
P23
Supplemental Figure 10
No evidence of heightened neurodegeneration or abnormal synaptogenesis
was observed in ptf1a-Cre; VEGFf/f mice. (A) TUNEL staining of P12 ptf1aCre; VEGFf/f retinas and controls. Scale bars: 500 μm (A; upper right); 50
μm (A; lower right). (B) Quantification from (A) at P12 or P23 (n = 4). (C)
The expression pattern of synaptophysin is unremarkable in P15 ptf1a-Cre;
VEGFf/f ;R26tdTomato/+ retinas. (D) Transmission electron micrograph showing
normal synaptic ultrastructure (arrows) and synaptic vesicles in P23 ptf1aCre; VEGFf/f retinas (upper panels: controls). Scale bar: 50 μm (C); 1 μm
(D). GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear
layer; ONL, outer nuclear layer; OPL, outer plexiform layer. Figure-S10 (Friedlander) ptf1a; VHLf/f GS DAPI
Chx10 DAPI
Calbindin DAPI Calretinin DAPI
control
A
GCL
ptf1a; VEGFf/f B
C
E
F
H
I
K
L
IPL
INL
ONL
D
GCL
IPL
INL
ONL
G
GCL
IPL
INL
ONL
J
GCL
IPL
INL
ONL
Supplemental Figure 11
No differences in the topographies of interneurons and Mueller glia are
observed in ptf1a-Cre VEGF knockout and ptf1a-Cre; VHL knockout mice. (AL) Cryosections from 23-day-old ptf1a-Cre; VEGFf/f and ptf1a-Cre; VHLf/f mice
and controls were stained for calretinin (A-C), calbindin (D-F), chx10 (G-I), and
glycogen synthase (GS) (J-L). The nuclei of cells were counterstained with
DAPI (blue). Scale bar: 50 μm in A-L GCL, ganglion cell layer; IPL, inner
plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer. Figure-S11 (Friedlander) VEGFf/f
A
VHLf/f
B
ptf1a; VEGFf/f
ptf1a; VHLf/f
30
Weight (g)
Weight (g)
30
20
10
20
10
0
0
0
20
40
0
60
20
100
50
VHLf/f
D
Blood glucose (mg/dl)
Blood glucose (mg/dl)
150
ptf1a; VHLf/f
150
100
50
0
0
P12 P23 P12 VEGFf/f
E
6
HbA1C (%) VEGFf/f
ptf1a; VEGFf/f
60
Age (days) Age (days) C
40
ptf1a; VEGFf/f
VHLf/f
6
4
4
2
2
0
0
ptf1a; VHLf/f
P23 Supplemental Figure 12
Normal weight, blood glucose, and HbA1c levels in VEGF and VHL mutant
mice. (A, B) Body weights in both groups were comparable. (n = 9-10). (C, D)
Blood glucose was measured at P12 and P23 of ptf1a-Cre; VEGFf/f mice and
ptf1a-Cre; VHLf/f mice. (n = 10 each). (E) HbA1C levels in 4 month-old ptf1aCre; VEGFf/f and 6 month-old ptf1a-Cre; VHLf/f mice. (n = 8 each). Figure-S12 (Friedlander) A
VEGFf/f; Pde6brd10/rd10
ptf1a; VEGFf/f; Pde6brd10/rd10
IPL
INL
ONL
B
ptf1a; VEGFf/f; Pde6brd10/+
VEGFf/f; Pde6brd10/rd10
ptf1a; VEGFf/f; Pde6brd10/rd10
IPL
INL
ONL
ptf1a; VEGFf/f; Pde6brd10/+
VEGFf/f; Pde6brd10/rd10
ptf1a; VEGFf/f; Pde6brd10/rd10
IPL
ONL
GS-lectin (P21)
deep
intermediate
composite
IPL
INL
IPL
INL
ONL
ONL
Ptf1a; VEGFf/f; Pde6brd10/rd10
INL
superficial
VEGFf/f; Pde6brd10/rd10
D
C
Supplemental Figure 13
Loss of VEGF in amacrine and horizontal cells accelerates photoreceptor
atrophy in an animal model of retinal degeneration. (A) No degeneration
(thinning of the ONL layer) is observed in SD-OCT images of P17 ptf1a-Cre;
VEGFf/f; Pde6brd10/rd10 mice compared with controls (VEGFf/f; Pde6brd10/rd10).
(B and C) Significant ONL thinning is observed in rd10 mice with impaired
intermediate plexuses (ptf1a-Cre; VEGFf/f; Pde6brd10/rd10) compared with
rd10 mice (VEGFf/f; Pde6brd10/ rd10), or with non-degenerating controls (one
recessive rd10 allele; ptf1a-Cre; VEGFf/f; Pde6brd10/+) using OCT (B) and
histology (C). (D) The integrity of the intermediate plexus is shown in P21
ptf1a-Cre; VEGFf/f; Pde6brd10/rd10 mice (green) compared with controls
(VEGFf/f; Pde6brd10/rd10). Scale bar: 50 μm (C); 100 μm (D-F). IPL, inner
plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer. Figure-S13 (Friedlander)