Abstract
Retinoblastoma tumor suppressor (Rb) promotes cell cycle exit, survival, differentiation, and tumor suppression in the retina. Here, we show it is also essential for vascularization and lamination. Despite minimal effects on Hif1a target expression, intraretinal vascular plexi did not form in the Rb−/− murine retina. Deleting adenovirus E2 promoter binding factor 3 (E2f3), which rescues starburst amacrine cell differentiation, or E2f2, had no effect, but deleting E2f1, which promotes neuronal cell cycle exit and survival, restored retinal vasculature. We specifically linked cell loss to the defect because removing Bax rescued rod and bipolar neurons and the vasculature, but not cell cycle exit. Despite rescuing Rb−/− neurons, Bax deletion exacerbated a delay in outer retina lamination, and exposed a requirement for Rb in inner retina lamination. The latter resembled Sem5 or FAT atypical cadherin 3 (Fat3) mutants, but expression of Sem5/Fat3 pathway components, or that of Neogenin, which perturbs migration in the Rb−/− cortex, was unchanged. Instead, lamination defects correlated with ectopic division, and were E2f1-dependent, implicating the cell cycle machinery. These in vivo studies expose new developmental roles for Rb, pinpoint aberrant E2f1 and Bax activity in neuronal death and vascular loss, and further implicate E2f1 in defective lamination. Links between Rb, angiogenesis and lamination have implications for the treatment of neovascularization, neurodegeneration and cancer.
Similar content being viewed by others
Introduction
Angiogenesis is a critical step in development and disease and is regulated by pro-angiogenic and anti-angiogenic factors1. In mice, the retinal vasculature consists of three interconnected parallel vascular plexi. A superficial vascular plexus (SVP) in the nerve fiber layer (NFL) develops from the optic nerve head and progresses radially to the peripheral retina between postnatal day 0 (P0) and P8. Subsequently, vessels sprout vertically into the retina. Around P7, sprouting vessels descend and advance into the outer plexiform layer (OPL) to establish the deep vascular plexus (DVP). Around P11, the DVP vessels ascend into the inner plexiform layer (IPL) and form the intermediate vascular plexus (IVP)2,3,4. Previous studies revealed that Frizzled-4 (Fzd4), Lrp5, Norrin, and Tetraspanin 12 (Tspan12) are required for intraretinal vascular development3,5,6, and retinal Hif1a is required for IVP development7. Retinal neurons form the neurovascular unit to interact with endothelial cells4,8. Retinal ganglion cells (RGCs) are essential for the SVP development9,10, photoreceptors are important for the development of the intraretinal vascular plexus11,12, and amacrine and horizontal cells are critical for developing and maintaining the intraretinal vasculature4.
The retinoblastoma tumor suppressor (Rb) plays a major role in regulating cell cycle and other cellular processes by interacting with adenovirus E2 promoter binding factors (E2fs)13. The Rb/E2f pathway plays critical roles in angiogenesis1. For example, Rb binds Hif-1α and enhances its transcriptional activity14. Atypical E2fs (E2F7/8) bind Hif-1α to stimulate Vegfa induction15. As such, deletion of Rb1, or E2f7/8, or Hif1 in mice all result in vascular defects in the placenta and early embryonic lethality16,17,18.
The retina comprises three distinct nuclear layers (GCL, ganglion cell layer; INL, inner nuclear layer; and ONL, outer nuclear layer), separated by two synaptic layers (OPL and IPL). Retinal lamination is guided by many different cues19, such as Dscam20,21, FAT atypical cadherin 3 (Fat3)22, semaphorins (Sema), and plexins23. Proper development of retinal lamination is important for physiological retinal responses and function. Recent studies have revealed that Rb can regulate neuronal migration and cortical lamination24,25.
Previously we reported that Pax6 alpha enhancer Cre (α-Cre)-mediated Rb gene knockout (KO) in mouse retina cause ectopic cell division, cell death, and differentiation defects26,27. α-Cre is active from embryonic day 10 (E10) in the peripheral progenitors of the temporal and nasal retina28,29. Employing this model we now implicate Rb in formation of the two intraretinal vascular plexi, as well as lamination of the outer and inner retina. We show that the angiogenesis defect is related to E2f1 and Bax-induced retinal cell death, whereas the lamination defects are cell death-independent and instead correlate with ectopic cell division.
Results
Rb is required for the development of intraretinal capillaries and retinal lamination
We examined Rbf/f;α-Cre and Rbf/f retinas between P7 and P60. To mark the RbKO area, we utilized Cre reporter Z/Red mice which express the red fluorescent protein upon Cre-mediated recombination30. Retinal whole-mount staining revealed that at P18 and later time points, the density of retinal blood vessels was much lower in the peripheral RbKO area of Rbf/f;α-Cre retinas than that in the same areas of the Rbf/f retinas (wild-type (WT) control), and the WT areas of Rbf/f;α-Cre retinas (Fig. 1a). Killing RGCs in the embryonic retina blocks the SVP formation in the postnatal tissue, as does Math5 deletion, which prevents the genesis of 95% RGCs9,10. Surprisingly, in the Rbf/f;α-Cre retina, where 80% RGCs are deleted in the P0 peripheral retina26, the SVP formed normally (Fig. 1b–d). This finding suggests that the reduced RGC density in the periphery and/or that the RGC signals in the central retina are sufficient to promote peripheral vascular outgrowth. The SVP remained intact at later stages (Fig. 1b–d), thus peripheral RGCs are also not required to maintain these vessels. At P18 the RbKO area had a normal SVP, but almost entirely lacked the IVP and DVP (Fig. 1c), as confirmed by histological staining of retinal sections (Fig. 1d). In WT P18 retinal sections, isolectin B4-positive (IB4+) cells were found in the NFL, IPL, and OPL, corresponding to the SVP, IVP, and DVP, respectively. However in RbKO retinal sections, IB4+ cells were only found in the NFL, but not in the IPL and OPL (Fig. 1d). In the P60 RbKO retina, the IVP and DVP were still absent, indicating that the defect was not transient due to delayed development (Fig. 1b).
We also examined the lamination of Rb-deficient retinas. At P0, the IPL had already formed between neuroblast layer (NBL) and GCL in WT and Rb−/− retinas (Fig. 1e); thus, although Rb is required for maturation of a subset of amacrine cells27, most of the IPL forms in its absence. At P8, the OPL had formed between the ONL and INL, in the WT but not Rb−/− retina (Fig. 1e). At P18, the OPL had also formed in the Rb−/− retinas (Fig. 1d), suggesting that Rb deficiency delayed OPL formation, likely due to fewer rod and bipolar cells26. This phenotype is similar to the migration and lamination defects in the developing Rb−/− cortex25. We concluded that Rb is required for the development of intraretinal capillary plexi and formation of the OPL.
A subset of angiogenesis regulators is down-regulated following Rb loss
The α-Cre transgene is restricted to retinal cells, indicating a non-cell autonomous effect of Rb loss on angiogenesis. As the DVP and IVP originate from the SVP at about P7 and P112,4,31, respectively, the defects should begin at around P7. Notably, P7 is the peak time for ectopic cell division and cell death of the Rb−/− retina26. As Rb can bind Hif-1α and enhance its transcriptional activity14, α-Cre-mediated Hif-1α KO retina lacks the IVP7, Rb loss may directly reduce Hif-1α activity, and thus the vascular defects. However, gene expression analysis did not support this notion.
We identified 677 RbKO-related deregulated genes (DEGs) from the microarray data of P8 RbKO retinas (Gene Expression Omnibus (GEO) accession: GSE86372)32. Gene enrichment analysis by Enrichr33,34 indicated that the most enriched pathways of RbKO-related DEGs included the cell cycle and DNA replication, consistent with Rb function, and phototransduction, explained by rod death, but not the Hif1 or Vegf pathways (Fig. 2a, b). Reverse transcription-polymerase chain reaction (RT-PCR) revealed that, compared to WT retinas, while Epo was reduced in Rb−/− retina at P7, other Hif-1α targets such as Vegfa, Id2, Vegfr2, Bnip3, and Cxcr4 were unchanged at P7 and P18 (Fig. 2c, d). One possibility is that the Rb−/− retina has higher expression of E2f7 and E2f8 (Fig. 2b, e, f)27, which can enhance Hif-1α transcriptional activity35, thus compensating for the effect of Rb loss on Hif1a activity. On the other hand, non-Hif1a targets such as Norrin, Fzd4, and Tie2, which are important for intraretinal vascular capillaries2,6, were reduced in Rb−/− retina at P7 (Fig. 2b, c).
Inactivating E2f1, but not E2f2 or E2f3, rescues the vascular and lamination defects of RbKO retina
E2f1 can suppress36 or promote37 angiogenesis in different conditions; however, we have not found any retinal vascular or lamination defects in E2f1KO, or E2f2KO, or E2f3f/f;α-Cre mice. We reported that in the RbKO retina, E2f1 mediates ectopic division and cell death, E2f3 disrupts starburst amacrine cell (SAC) differentiation, and while E2f2 does not cause defects in the RbKO retina, it mediates the ectopic division and death of cones of Rb/p107 double KO (DKO) retina27,38,39. To define whether any E2f contributes to the vascular defects, we crossed Rbf/f;α-Cre mice with E2f1−/−, or E2f2−/−, or E2f3f/f mice (Fig. 3a, b).
We used the AngioTool software to analyze the vessel coverage, average vessel length, and lacunarity of vascular plexi. This analysis confirmed that the RbKO retina had a normal SVP, but the vascular density and average vessel length were much reduced in IVP and DVP (Fig. 3c–f). Deleting E2f3 rescued Calretinin+ SAC processes in the Rb−/− IPL (Fig. 3a), as before27, but neither that nor E2f2 loss affected angiogenesis (Fig. 3b–f). In stark contrast, removing E2f1, which does not rescue the SAC defect (Fig. 3a), completely reversed the loss of both IVP and DVP, and returned vessel coverage and length as well as lacunarity to WT levels (Fig. 3b–f). Even removing one E2f1 allele had a marked effect (Fig. 3b–f). E2f1KO also restored Fzd4 and Tie2 mRNA levels in the Rb−/− retina at P7 (Fig. 2b), which may account for the rescue of retinal vascular defects. These data are reminiscent of the dose-dependent effects of E2f1 on abnormal cell division and death in the RbKO retina38.
The lamination defects were rescued by either E2f1+/− or E2f1−/−, but neither E2f2 nor E2f3 loss. These results indicate that whereas E2f2 or E2f3 activity, or disrupted SAC differentiation are not involved, E2f1 drives the vascular and lamination defects in the RbKO retina. This is slightly different from the developing Rb−/− cortex, in which either E2f1−/− or E2f3−/− rescues the lamination defects25,40. E2f1 drives both ectopic division and cell death in the RbKO retina27, thus we next sought to define which of these cellular defects contributes to vascular disruption and retinal lamination defects.
Bax contributes to cell death and retinal vasculature defects
The Bcl-2 family member Bax mediates neuronal apoptosis41, including physiological retinal apoptosis42,43 and neuronal death in the Rb/p107 DKO brain24. Thus, we tested whether Bax drives cell death in the RbKO retina. Bax−/− did not affect ectopic division in the RbKO retina at P2 or P8, and actually elevated proliferation at P18 (Fig. 4a, b). All ectopic division had ceased by P30 indicating Rb-independent means of cell cycle exit. The increase in dividing cells at P18 might result from improved survival of ectopically proliferating Rb−/− cells at earlier times, and indeed apoptosis was markedly reduced in the Rb/Bax DKO retina at P2, and P8 when cell death peaks (Fig. 4c–d). Thus, Bax drives cell death in the Rb−/− retina.
Next, we defined which death-prone Rb−/− cell types are rescued by BaxKO. Rb loss causes death of most rod bipolar and RGCs, and many rods26. Bax loss restored RGCs to WT numbers, and also suppressed rod bipolar cell death considerably (Fig. 4e, f). Rhodopsin staining and measurement of ONL thickness indicated partial rescue of rod photoreceptors (Fig. 4e, g). Thus, Bax is a major, but not the sole contributor to apoptosis in Rb−/− neurons. Its role is greater in the inner (ganglion, bipolar cells) than outer retina (photoreceptors).
Next, we asked whether BaxKO affects the disrupted vasculature of the Rb−/− retina. Strikingly, we observed both IVP and DVP in the P18 Rb/Bax DKO retina (Fig. 5a–c). Quantification in the IVP revealed that BaxKO increased vessel coverage and length, and reduced lacunarity essentially to WT levels (Fig. 5d–f). The extent of the defects was greater in the RbKO DVP; Bax loss did not restore vascularity to WT levels, but did elevate vessel coverage and length, and dramatically reduced lacunarity (Fig. 5d–f). In line with these cellular effects BaxKO rescued the reduction of Fzd4, Norrin, and Tie2 expression in P7 Rb−/− retina, and actually induced Vegfa and Cxcr4 expression considerably by P18 (Fig. 2c, d). We also observed some increase in vasculature in the Bax−/− relative to WT retina (Fig. 5c–e). These data suggest that the modest effects of Bax loss on natural neuronal pruning (Fig. 4b)43 promote vessel formation in the developing retina.
In summary, without reducing ectopic division, deleting Bax promotes survival of Rb−/− cells, most prominently in the inner retina, which correlates with increased expression of angiogenic factors and dramatic rescue of the IVP and DVP.
Inactivating Bax exposes new roles for Rb in retinal lamination
The striking rescue of cell death in the Rb/Bax DKO retina also exposed two roles for Rb in retinal lamination. The first affected OPL formation. OPL forms in the P8 WT retina, but not in the RbKO tissue until P18 (Figs. 1e and 4a, c, e). We assumed that this delay reflects the loss of rods and bipolar cells, and thus expected it to be rescued in the DKO retina (Fig. 4e–g). Unexpectedly, no OPL was evident in the DKO retina at P8 and P18 (Fig. 4a, c, e), which was confirmed by staining for synaptic vesicle proteins synaptotagmin 1 (Syt1)44 and synaptic vesicle glycoprotein 2a (SV2a)45 (Fig. 6a).
In contrast to the uniform OPL staining in P8 WT or BaxKO retina, SV2A+ synapses were disorganized in the RbKO or DKO retinas (Fig. 6a). Syt1 staining was better ordered in the P18 RbKO retina, even though the M opsin+ cones and Onecut2-positive (OC2+) horizontal cells were almost superimposed, but it remained disorganized in the DKO tissue, even though positioning of the cone and horizontal cells was restored (Fig. 6a). By P30, a Syt1+ OPL was evident in the DKO retina, although still less uniform than in other genotypes (Fig. 6a). In the Rb−/− cortex, Neogenin 1 (Neo1) induction causes migration and lamination defects25,40. However Neo1 was not induced in the Rb−/− retina (Figs. 2b and 6b). Other migration regulators that were induced in the Rb−/− cortex, including Sema3d, ApoE, CCK, Twist1, and Twistnb25, were also not altered in the Rb−/− retina (Fig. 2b).
Potentially, ectopic division directly perturbs OPL formation, because its appearance correlated with a large reduction in ectopically dividing Ki67+ cells in the DKO retina between P18 and P30, and in the RbKO retina between P8 and P18 (Fig. 4a). Furthermore, in the boundary between central α-Cre− and peripheral α-Cre+ retina, OPL disruptions coincided precisely with Ki67+ cells (Fig. 6c, arrows), disordered SV2a vesicles coincide with dense Ki67 staining (Fig. 6c).
Cyclin E can bind and sequester cyclin-dependent kinase 5 (Cdk5), which plays important roles in neuronal migration and lamination by phosphorylating doublecortin (Dcx)46,47,48. As an E2f target, cyclin E increased in the Rb−/− retina, while Cdk5 and Dcx levels had barely changed (Figs. 2b and 6b). In P8 BaxKO retina, Dcx and Cdk5 were most prominently expressed in the cell bodies and dendrites of horizontal and amacrine cells, while cyclin E was most evident in the OPL and less so the IPL and cytoplasm of some cells (Fig. 6d). In P8 Rb/Bax DKO retina the expression pattern of Dcx and Cdk5 was similar, but cyclin E was induced in many retinal cells (Fig. 6d). These data are consistent with the notion that cyclin E in ectopically dividing cells may interfere with Cdk5 function and contribute to the lamination defects in the RbKO and Rb/Bax DKO retina.
In addition to delayed OPL formation, there was an unanticipated “outer misplaced plexiform layer” (OMPL) in P8 DKO retinas, which split the INL (Figs. 4a, c, e and 6a, e). It was not observed in either the RbKO or BaxKO retinas (Fig. 6a). Excess cell production cannot explain this phenomenon since, at best, Bax loss restored the missing retinal cell types (Fig. 4f, g), retinal thickness was similar at all time points between DKO and control retinas, and Bax deletion had no effect on the numbers of Rb−/− death-resistant cone, horizontal, amacrine, and Müller cells (Fig. 6e–g). Many amacrine cells flanked the OMPL, as indicated by Calretinin and Ap2α staining (Fig. 6e, h). A similar OMPL was observed in Fat3−/−, Sema5A−/−, and Sema5B−/− retinas, in which amacrine cells are bipolar, leading to new synaptic contacts with bipolar cells22,23. Indeed, some amacrine cells in the DKO INL were bipolar and had ectopic dendrites that stratified in the OMPL (Fig. 6h, arrows).
In view of the phenotypic similarities between Rb/Bax DKO and Fat3 or Sem5 KO retinas, we assessed the mRNA levels of Fat3 and Sem5A & B, as well as those of Dachsous (Dchs), a Fat3 ligand, and Four jointed box 1 (Fjx1), which modifies Fat activity. The levels of Dchs, Fat3, and Sema5A were not different between WT, RbKO, BaxKO, and DKO retina at three different time points, and while Fjx1 and Sema5B were slightly increased in DKO retinas, this was also the case in the BaxKO control which has no OMPL (Fig. 6h). The occurrence of the OMPL in the DKO, where ectopic division is enhanced, and its rescue in the Rb/E2f1 DKO retina, where ectopic division is blocked, suggest that it may also be linked to excess activity of cell cycle machinery. Irrespective, these data expose a new role for Rb in coordinating retinal lamination with phenotypic similarities to Fat3-deficient and Sema5-deficient retinas
Discussion
Rb is critical in the retina as it promotes cell cycle exit, neuronal survival, SAC differentiation, and tumor suppression26,27. Here we exposed new roles in retinal angiogenesis and lamination. The vasculature was rescued by deleting E2f1 or Bax, pinpointing neuronal loss rather than ectopic division in the phenotype. Rescuing death-prone neurons with Bax deletion exposed new roles for Rb in OPL and INL formation. These data expand our insight into the multi-faceted functions of Rb in retinal development.
To separate defects linked to division or death, we focussed on Bax because it mediates the death of Rb/p107 DKO neurons in the brain4. Indeed, Bax loss rescued most rod bipolar and RGCs, and many rods, but had no major effect on ectopic cell division. Rb loss delays the formation of OPL from P8 to P18; we assumed this defect was due to rod and bipolar cell loss. However, while deleting Bax rescued many Rb−/− rods and bipolar cells, it did not rescue the OPL defect. Despite the disorder, synaptic proteins were present, and Neogenin, which is induced and disrupts lamination in the Rb/p107-null cortex24,25, was not elevated in the Rb-null retina.
The OPL defect may be related to ectopic cell division because E2f1KO promoted cell cycle exit and rescued OPL defects, but deleting Bax, which increased ectopic division due to extended cell survival, further delayed OPL genesis. Moreover, at the boundary of Cre expression, sporadic OPL disruption correlated perfectly with interspersed ectopically dividing cells. The eventual appearance of a narrow OPL correlated with delayed cell cycle exit at P18 or P30 in the Rb-null or Rb/Bax-null retina, respectively. Potentially, over-active cell cycle machinery disrupts formation of the synaptic layers. One candidate, cyclin E, can sequester Cdk5 which is important for lamination48, and we confirmed high levels of this E2f1 target in ectopically dividing Rb-null and Rb/Bax-null cells. Future work will address whether reducing cyclin E or elevating Cdk5 ameliorates the OPL defect.
The unnatural OMPL that split the Rb/Bax-null INL at P8 resembles the phenotype of Fat3−/−22, and Sema5A−/−; Sema5B−/−23 retinas, suggesting that Rb may affect these adhesive and repulsive cues19. We did not observe major changes in their expression. Cell cycle enzymes might also contribute to this phenotype as the OMPL separated post-mitotic amacrine cells in the inner INL from ectopically dividing amacrine cells in the outer INL (Fig. 4a).
Our work also exposes a critical role for Rb in the formation of intraretinal vasculature. In theory, Rb could influence angiogenesis through its positive effects on Hif114, and/or by promoting survival of cells essential for vessel formation. Our results support the latter because rescuing death-prone Rb−/− neurons with Bax deletion restored the IVP and DVP. Hif1 targets were relatively unaffected by Rb loss, perhaps due to induction of E2f7/8, known to augment Hif1 activity15, and instead other angiogenic regulators, such as Norrin, Fzd4, and Tie2, were down-regulated.
Excitatory retinal neurons, including RGCs and photoreceptors, can drive and regulate retinal angiogenesis4,49. Specifically, RGCs are required for the SVP development9, and photoreceptors are required for the DVP development50. Surprisingly, Rb deficiency did not affect the SVP development, even though Rb−/− mouse retinas lose most RGCs26. In prior studies RGCs were depleted from the entire retina9,10, but in our case cell death is confined to the periphery, so conceivably the signals in the intact central retina are sufficient to drive the SVP development in the periphery.
Clearly, this was not the case for the IVP and DVP, which showed dramatic loss in the Rb-null retina. There is a very close relationship between the number of photoreceptors and vessel profiles in the DVP50. As there are still many rods in the Rb−/− retina, their reduction may not explain all the vascular defects. Rod bipolar cells may affect the IVP development because these cells are essentially absent in Rb−/− P18 retina26. E2f1KO rescued all the cell death of rods and bipolar cells, and completely rescued the vascular defects. Bax loss rescued about 80% of rod bipolar cells and 50% of rods, but rescued about 80% of IVP coverage and 50% of DVP coverage which, while not definitive, suggests a potential role for the bipolar cells.
Amacrine and horizontal cells are required for generating and maintaining the intraretinal vasculature4. However, in our study, amacrine and horizontal cells survive Rb deficiency, but intraretinal vascular plexi still could not develop, indicating that these inhibitory neuron-regulated pro-angiogenic factors are not sufficient for the retinal vascular development.
Our work exposes new roles for Rb in the developing retina, pinpointing aberrant E2f1 and Bax activity as key drivers of defects in neuronal survival and angiogenesis, and further implicating E2f1 in aberrant lamination. The Rb/E2f pathway is commonly involved in many diseases, such as most human cancers13, retinal degeneration51, diabetic retinopathy52, and neuronal degeneration53. Abnormal cell survival, proliferation, angiogenesis, and lamination are common features of these diseases. Thus, our findings will enhance our understanding of the pathogenesis and optimize the treatment strategies for these diseases in the future.
Materials and methods
Mouse strains and genotyping
Mice were treated according to institutional and national guidelines. All procedures were performed in compliance with the Association for Research in Vision and Ophthalmology statement for the use of animals in ophthalmic and visual research. α-Cre mice (P. Gruss), Rbf/f mice (A. Berns), E2f1–/– mice (M. Greenberg), E2f2–/– mice (G. Leone), E2f3f/f mice (G. Leone), Z/Red (Jackson Laboratory, stock#005438), and Bax−/− (Jackson Laboratory, stock#002994) were maintained on a mixed (NMRI × C57/BL6 × FVB/N × 129sv) background.
Bax-null males are infertile, to generate Rbf/f;Bax−/−;α-Cre mice we first generated Rbf/f;Bax+/−;α-Cre males and Rbf/f;Bax−/− females, and inter-bred them. Of 96 pups we obtained 50 with α-Cre and 48 with Bax−/− alleles as predicted, but only 5 pups (5.2%) with Bax−/−;α-Cre, far less than the expected 24 pups (25%, Table 1), suggesting that α-Cre and Bax- alleles may be on the same chromosome. We screened pups for a crossover event to generate α-Cre and Bax- alleles on the same chromosome. We obtained Rbf/f;Bax−/−;α-Cre females, and bred them with Rbf/f;Bax+/− males to generate experimental littermates.
Mice of different genotypes were compared within the same litter and across a minimum of three litters. We have not noted any phenotypic differences in separate litters. Genotyping was performed as before26,27,54, and the primers for genotyping Z/Red mice are primers oIMR3847 and oIMR4110 for transgene (208 bp), and oIMR7338 and oIMR7339 for internal positive control (324 bp). The primers for genotyping Bax−/− mice are primers oIMR0661 and oIMR0662 for mutant allele (507 bp), and oIMR0661 and oIMR0663 for WT allele (304 bp). The sequences of the above primers are listed in Jackson Laboratory genotyping protocols (www2.jax.org).
Histology, immunofluorescence, and measurements
Eyeballs were fixed in 4% paraformaldehyde for 1 h at 4 °C, embedded in OCT (TissueTek 4583), frozen on dry ice, and cut into 12–14 µm sections on Superfrost slides. For immunohistochemistry, the retinal sections were dried at room temperature and incubated in blocking solution (0.5% normal donkey serum, 0.03% Triton X-100 in 1× phosphate-buffered saline (PBS)) for 1 h, then were incubated with primary antibodies such as active caspase-3 (Cell Signaling Technology, 9661), Ap2α (Santa Cruz, SC-8975), Brn3 (Santa Cruz, SC-6062), Calbindin (Sigma, C9484), Calretinin (Santa Cruz, SC-11644), Cdk5 (Santa Cruz, SC173), Cone arrestin (Millipore, AB15282), cyclin E (Upstate, 07-687), Dcx (Abcam, ab18723), glutamine synthetase (Millipore, MAB302), Ki67 (BD Science Pharmingen, 550609), M opsin (C.M. Craft and X. Zhu, University South California), Onecut2 (R&D System, AF6294), phospho-histone H3 (Santa Cruz, SC-8656), protein kinase Cα (Sigma, P5704), rhodopsin (Santa Cruz, SC-57433), SV2a (R. Janz, The University of Texas-Houston Medical School), and Syt1 (Abcam, ab13259). Vascular endothelial cells were labeled by fluorescein isothiocyanate (FITC)-conjugated isolectin B4 (Sigma, L2895). Antigen retrieval was performed as described by boiling sections in citric acid (H-3300, Vector Lab)27. Primary antibodies or labeled cells were visualized using donkey anti-mouse, donkey anti-rabbit, and donkey anti-goat antibodies conjugated with Alexa-488, Alexa-568, or Alexa-647 (1:1000; Molecular Probes). Nuclei were counter-stained with 4′, 6-diamidino-2-phenyindole (DAPI; Sigma) and mounted with Mowiol mounting medium.
For whole-mount staining, eyeballs were enucleated and incubated for 30 min in 4% paraformaldehyde in PBS. With a dissection microscope, a circumferential incision was made around the limbus, followed by removal of the anterior segment, lens, and vitreous body. The retinas were incubated at 4 ℃ with FITC-conjugated isolectin B4 (Sigma, L2895) and DAPI in PBS for 1–2 days. After briefly washing with PBS, radial cuts were made to divide the retina into four quadrants to flatten the retina, and flat retinas were mounted with Mowiol.
Stained sections and slides were analyzed using a Zeiss Axio Imager Z2 fluorescence microscope and Nikon C1si confocal microscope. Image J 1.50b with cell counter plugin (https://imagej.nih.gov/ij/) was used for cell counting following the online guide. The positive cells of active caspase-3, Ki67, and cell-type markers (including Brn3, PKCα) were counted manually. The thickness of ONLs was measured by the microscope program. For vascular blood vessel analysis, representative images were analyzed using the AngioTool software (https://ccrod.cancer.gov/confluence/display/ROB2/Home) to quantify the vessel coverage (percentage of area covered by IB4+ endothelial cells), average vessel length, and lacunarity (distribution of the gap sizes surrounding the object).
Microarray dataset selection and analysis
The dataset GSE86372 at NCBI GEO database (http://www.ncbi.nlm.nih.gov/geo/) was used to compare WT vs. RbKO mouse P8 retinas, which include three WT and three RbKO mouse P8 retinas. The data were analyzed by GEO2R from the GEO website. The genes, of which expression fold changes are >2 or <0.5, and adjusted p < 0.05 were selected as the RbKO-related DEGs. Totally 677 DEGs were identified. The heatmap was generated using Heatmapper55. The function enrichment of DEG was performed using Enrichr33,34, the pathways with adjusted p < 0.05 were chosen to report.
RNA extraction, RT, and quantitative real-time PCR
Total RNA was isolated from dissected peripheral retina using the TriPure isolation reagent (Roche, USA) or RNeasy mini kit (Qiagen) followed by digestion with RNase-Free DNase (DNA-freeTM, Thermo Fisher Scientific) to remove DNA contamination. First-strand cDNA was synthesized from 0.2–0.5 µg of total RNA using the RT reagent kit with gDNA Eraser (TaKaRa, China) or SuperScript II first-strand synthesis system (Invitrogen). PCR primers are listed in Supplementary Table 1. Real-time quantitative PCR was performed using the qTOWER 2.2 PCR machine (Analytik Jena, Germany) or C1000 touch Thermal Cycler (Bio-Rad, USA). Tests were run in duplicate on three separate biological samples with EvaGreen PCR Supermix (SsoFastTM, Bio-Rad Laboratories, Singapore) or SYBRGreen PCR Master Mix (Applied Biosystems). PCR consisted of 40 cycles of denaturation at 95 °C for 15 s, and annealing and extension at 55 °C for 30 s. An additional cycle (95 °C, 15 s) generated a dissociation curve to confirm a single product. Values obtained for test RNAs were normalized to β-actin mRNA levels.
Statistical analysis
All data were presented as mean ± SD. Statistical analysis was undertaken using the GraphPad Prism software (GraphPad Prism Software, Inc., San Diego, CA, USA). The results were analyzed by one-way analysis of variance (ANOVA) followed by Bonferroni correction for multiple comparisons. The threshold for significance was set at p < 0.05.
References
Schaal, C., Pillai, S. & Chellappan, S. P. The Rb-E2F transcriptional regulatory pathway in tumor angiogenesis and metastasis. Adv. Cancer Res. 121, 147–182 (2014).
Fruttiger, M. Development of the retinal vasculature. Angiogenesis 10, 77–88 (2007).
Ye, X., Wang, Y. & Nathans, J. The Norrin/Frizzled4 signaling pathway in retinal vascular development and disease. Trends Mol. Med. 16, 417–425 (2010).
Usui, Y. et al. Neurovascular crosstalk between interneurons and capillaries is required for vision. J. Clin. Invest. 125, 2335–2346 (2015).
Junge, H. J. et al. TSPAN12 regulates retinal vascular development by promoting Norrin- but not Wnt-induced FZD4/beta-catenin signaling. Cell 139, 299–311 (2009).
Ye, X. et al. Norrin, frizzled-4, and Lrp5 signaling in endothelial cells controls a genetic program for retinal vascularization. Cell 139, 285–298 (2009).
Caprara, C. et al. HIF1A is essential for the development of the intermediate plexus of the retinal vasculature. Invest. Ophthalmol. Vis. Sci. 52, 2109–2117 (2011).
Ivanova, E., Toychiev, A. H., Yee, C. W. & Sagdullaev, B. T. Intersublaminar vascular plexus: the correlation of retinal blood vessels with functional sublaminae of the inner plexiform layer. Invest. Ophthalmol. Vis. Sci. 55, 78–86 (2014).
Sapieha, P. et al. The succinate receptor GPR91 in neurons has a major role in retinal angiogenesis. Nat. Med. 14, 1067–1076 (2008).
Edwards, M. M. et al. The deletion of Math5 disrupts retinal blood vessel and glial development in mice. Exp. Eye Res. 96, 147–156 (2012).
Toner, C. K. et al. Vision-fair neuropsychological assessment in normal aging, Parkinson’s disease and Alzheimer’s disease. Psychol. Aging 27, 785–790 (2012).
Lahdenranta, J. et al. An anti-angiogenic state in mice and humans with retinal photoreceptor cell degeneration. Proc. Natl. Acad. Sci. USA 98, 10368–10373 (2001).
Burkhart, D. L. & Sage, J. Cellular mechanisms of tumour suppression by the retinoblastoma gene. Nat. Rev. Cancer 8, 671–682 (2008).
Budde, A., Schneiderhan-Marra, N., Petersen, G. & Brune, B. Retinoblastoma susceptibility gene product pRB activates hypoxia-inducible factor-1 (HIF-1). Oncogene 24, 1802–1808 (2005).
Weijts, B. G. et al. E2F7 and E2F8 promote angiogenesis through transcriptional activation of VEGFA in cooperation with HIF1. EMBO J. 31, 3871–3884 (2012).
Li, J. et al. Synergistic function of E2F7 and E2F8 is essential for cell survival and embryonic development. Dev. Cell 14, 62–75 (2008).
Wu, L. et al. Extra-embryonic function of Rb is essential for embryonic development and viability. Nature 421, 942–947 (2003).
Dunwoodie, S. L. The role of hypoxia in development of the Mammalian embryo. Dev. Cell 17, 755–773 (2009).
Hoon, M., Okawa, H., Della Santina, L. & Wong, R. O. Functional architecture of the retina: development and disease. Prog. Retin. Eye Res. 42, 44–84 (2014).
Yamagata, M. & Sanes, J. R. Dscam and Sidekick proteins direct lamina-specific synaptic connections in vertebrate retina. Nature 451, 465–469 (2008).
Fuerst, P. G., Bruce, F., Rounds, R. P., Erskine, L. & Burgess, R. W. Cell autonomy of DSCAM function in retinal development. Dev. Biol. 361, 326–337 (2012).
Deans, M. R. et al. Control of neuronal morphology by the atypical cadherin Fat3. Neuron 71, 820–832 (2011).
Matsuoka, R. L. et al. Class 5 transmembrane semaphorins control selective mammalian retinal lamination and function. Neuron 71, 460–473 (2011).
Svoboda, D.S., Paquin, A., Park, D.S. & Slack, R.S. Pocket proteins pRb and p107 are required for cortical lamination independent of apoptosis. Dev. Biol. 384, 101–113 (2013).
McClellan, K. A. et al. Unique requirement for Rb/E2F3 in neuronal migration: evidence for cell cycle-independent functions. Mol. Cell. Biol. 27, 4825–4843 (2007).
Chen, D. et al. Cell-specific effects of RB or RB/p107 loss on retinal development implicate an intrinsically death-resistant cell-of-origin in retinoblastoma. Cancer Cell 5, 539–551 (2004).
Chen, D. et al. Rb-mediated neuronal differentiation through cell-cycle-independent regulation of E2f3a. PLoS Biol. 5, e179 (2007).
Marquardt, T. et al. Pax6 is required for the multipotent state of retinal progenitor cells. Cell 105, 43–55 (2001).
Baumer, N. et al. Pax6 is required for establishing naso-temporal and dorsal characteristics of the optic vesicle. Development 129, 4535–4545 (2002).
Vintersten, K. et al. Mouse in red: red fluorescent protein expression in mouse ES cells, embryos, and adult animals. Genesis 40, 241–246 (2004).
Xu, Q. et al. Vascular development in the retina and inner ear: control by Norrin and Frizzled-4, a high-affinity ligand-receptor pair. Cell 116, 883–895 (2004).
McCurdy, S. R., Pacal, M., Ahmad, M. & Bremner, R. A CDK2 activity signature predicts outcome in CDK2-low cancers. Oncogene 36, 2491–2502 (2017).
Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016).
Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinform. 14, 128 (2013).
Baba, T. et al. VEGF 165 b in the developing vasculatures of the fetal human eye. Dev. Dyn. 241, 595–607 (2012).
Qin, G. et al. Cell cycle regulator E2F1 modulates angiogenesis via p53-dependent transcriptional control of VEGF. Proc. Natl. Acad. Sci. USA 103, 11015–11020 (2006).
Engelmann, D. et al. E2F1 promotes angiogenesis through the VEGF-C/VEGFR-3 axis in a feedback loop for cooperative induction of PDGF-B. J. Mol. Cell. Biol. 5, 391–403 (2013).
Sangwan, M. et al. Established and new mouse models reveal E2f1 and Cdk2 dependency of retinoblastoma, and expose effective strategies to block tumor initiation. Oncogene 31, 5019–5028 (2012).
Chen, D., Chen, Y., Forrest, D. & Bremner, R. E2f2 induces cone photoreceptor apoptosis independent of E2f1 and E2f3. Cell Death Differ. 20, 931–940 (2013).
Andrusiak, M. G. et al. Rb/E2F regulates expression of neogenin during neuronal migration. Mol. Cell. Biol. 31, 238–247 (2011).
Xiang, H. et al. Bax involvement in p53-mediated neuronal cell death. J. Neurosci. 18, 1363–1373 (1998).
Mosinger Ogilvie, J., Deckwerth, T. L., Knudson, C. M. & Korsmeyer, S. J. Suppression of developmental retinal cell death but not of photoreceptor degeneration in Bax-deficient mice. Invest. Ophthalmol. Vis. Sci. 39, 1713–1720 (1998).
White, F. A., Keller-Peck, C. R., Knudson, C. M., Korsmeyer, S. J. & Snider, W. D. Widespread elimination of naturally occurring neuronal death in Bax-deficient mice. J. Neurosci. 18, 1428–1439 (1998).
Fox, M. A. & Sanes, J. R. Synaptotagmin I and II are present in distinct subsets of central synapses. J. Comp. Neurol. 503, 280–296 (2007).
Wang, M. M., Janz, R., Belizaire, R., Frishman, L. J. & Sherry, D. M. Differential distribution and developmental expression of synaptic vesicle protein 2 isoforms in the mouse retina. J. Comp. Neurol. 460, 106–122 (2003).
Chae, T. et al. Mice lackingp35, a neuronal specific activator of Cdk5, display cortical lamination defects, seizures, and adult lethality. Neuron 18, 29–42 (1997).
Tanaka, T. et al. Cdk5 phosphorylation of doublecortin ser297 regulates its effect on neuronal migration. Neuron 41, 215–227 (2004).
Odajima, J. et al. Cyclin E constrains Cdk5 activity to regulate synaptic plasticity and memory formation. Dev. Cell 21, 655–668 (2011).
Sapieha, P. Eyeing central neurons in vascular growth and reparative angiogenesis. Blood 120, 2182–2194 (2012).
Pennesi, M. E., Nishikawa, S., Matthes, M. T., Yasumura, D. & LaVail, M. M. The relationship of photoreceptor degeneration to retinal vascular development and loss in mutant rhodopsin transgenic and RCS rats. Exp. Eye Res. 87, 561–570 (2008).
Arsenijevic, Y. Cell cycle proteins and retinal degeneration: evidences of new potential therapeutic targets. Adv. Exp. Med. Biol. 854, 371–377 (2016).
Wang, Y. et al. E2f1 mediates high glucose-induced neuronal death in cultured mouse retinal explants. Cell Cycle 16, 1824–1834 (2017).
Folch, J. et al. Role of cell cycle re-entry in neurons: a common apoptotic mechanism of neuronal cell death. Neurotox. Res. 22, 195–207 (2012).
Chen, D. et al. Division and apoptosis of E2f-deficient retinal progenitors. Nature 462, 925–929 (2009).
Babicki, S. et al. Heatmapper: web-enabled heat mapping for all. Nucleic Acids Res. 44, W147–W153 (2016).
Acknowledgements
We thank A. Berns for Rbf/f mice, M. Greenberg for E2f1−/− mice, P. Gruss for α-Cre mice, G. Leone for E2f2–/– mice and E2f3f/f mice, C.M. Craft and X. Zhu for M opsin antibody, and R. Janz for SV2a antibody. This study was supported by grants to D.C. from the National Natural Science Foundation of China (81371022, 81570860), and grants from the Canadian Institutes for Health Research, Canadian Cancer Society Research Institute, the Krembil Foundation, and the Foundation Fighting Blindness Canada to R.B.
Author information
Authors and Affiliations
Contributions
D.C. and R.B. designed the study and interpreted data. Y.Z., R.W., L.Z., Y.C., S.L., C.L., Y.W., L.X., J.Z., and D.C. performed the experiments. D.C. and R.B. wrote the paper and all authors contributed to editing.
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Edited by A Verkhratsky.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Zhou, Y., Wei, R., Zhang, L. et al. Rb is required for retinal angiogenesis and lamination. Cell Death Dis 9, 370 (2018). https://doi.org/10.1038/s41419-018-0411-6
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41419-018-0411-6
- Springer Nature Limited
This article is cited by
-
Mapping transgene insertion sites reveals the α-Cre transgene expression in both developing retina and olfactory neurons
Communications Biology (2022)