Doctor of Philosophy (PhD) / Background: The development of the blood supply in the primate retina has been extensively studied; however the relationship of the differentiating retina to the choroidal blood supply is less well known. The interaction of astrocytes and vascular endothelial cells promotes the development of the retinal vasculature from 14 weeks’ gestation (WG). Initially, astrocytes lead the developing capillaries from the optic nerve towards the macular area. However, neither astrocytes nor endothelial cells enter a prescribed avascular area, within which the fovea later forms. This may be attributed to expression of a factor that inhibits astrocyte and endothelial cell proliferation in the fovea. A factor found in the CNS that is already known to have these effects is transforming growth factor-β (TGF-β). Aims: This thesis investigated the relationship between retinal maturation and choroidal blood vessel supply and the possible role for TGF-β as an antiangiogenic factor in maintaining an avascular fovea during primate retinal development. Methods: Human eyes between 11 WG and 40 years were obtained with ethical approval from Prince of Wales Hospital and the NSW Lions Eye Bank and fixed and sectioned for histological procedures or prepared for polymerase chain reaction (PCR). Macaque eyes from foetal day (fd) 64 to postnatal 11years (p11y) were obtained from Bogor Agriculture University, Indonesia with the approval of the Ethics Committee of the University of Washington, Seattle, USA. Macaque eyes were also fixed and sectioned for immunohistochemistry and in situ hybridisation. RNA was extracted from human foetal retinas and used for RTPCR (Reverse Transcriptase PCR), QPCR (Quantitative PCR) and preparation of riboprobes. PCR products were analysed using both restriction digest and sequencing. RTPCR was used to identify TGF-β1, TGF-β2 and TGF-β3 in the developing human and in the developing and adult macaque retinas whilst QPCR was used to quantify the TGF-β isoforms in central compared to peripheral retina and in foetal compared to adult retina. In situ hybridisation was performed according to a standard protocol and visualised using Roche HNPP Fast Red detection set with designed riboprobes for TGF-β1, TGF-β2 and TGF-β3 (DIG RNA labelling kit). Some sections were counterstained with vimentin antibody. Immunohistochemistry was performed on human retina and choroid sections using antibodies to CD34 and Ki67 and on human and macaque retina using antibodies to synaptophysin, vimentin, GFAP, calbindin, S-opsin, RG-opsin, rhodopsin, TGF-β1, TGF-β2, TGF-β3 and their receptors TβRI and TβRII. Sections of the retina were imaged and analysed using either a Leica Confocal microscope and TCSNT software or Zeiss Confocal microscope and LSM 5 Pascal software. Data from the human retina and choroid sections corresponding to different regions (foveal, parafoveal nasal, parafoveal temporal, nasal and temporal) was collected to measure the number of Ki-67 immunolabelled mitotic endothelial cells and the area of CD34 immunolabelled choriocapillaris using Adobe Photoshop version 5.0.2, NIH software version 1.62 (measurement macros) and Excel. In the human and macaque sections the intensity of TGF-β protein and mRNA expression was captured from different regions of the retina (foveal, parafoveal nasal, parafoveal temporal, nasal, temporal, nasal to disc) to compile montages. Montages were then re-imported into LSM 5 Pascal software to measure the optical density across each montage along the ganglion cell layer, outer neuroblastic zone and photoreceptor layer collecting data in Excel for graphical representation. In addition to the montages, individual sections were assessed for co-localisation of TGF-β and TβR to various retinal cell types. Results: Analyses of choriocapillaris area and endothelial cell (EC) proliferation were able to demonstrate that the area of choriocapillaris endothelium is greater in the foveal region at all ages (14-18.5WG), that the rate of choriocapillaris EC proliferation declines dramatically over this same period and that the lowest rates of EC proliferation are at the incipient fovea. Most importantly these findings indicate that EC proliferation in the choriocapillaris does not appear to be promoted by increased metabolic activity in central retinal neurons which are more developed with higher oxygen and nutrient demands, which is the mechanism widely thought to regulate development of the retinal vasculature. PCR showed all TGF-β isoforms to be present in the human developing and adult retina. QPCR revealed that TGF-β2 was the most predominant isoform, followed by TGF-β3 with very small amounts of TGF-β1 seen. The isoforms were more abundant in developing rather than adult retina and in central rather than peripheral retina. Studies of the distribution of TGF-β protein and mRNA using immunohistochemistry and in situ hybridisation confirmed the low levels of TGF-β1 protein and mRNA observed in QPCR and demonstrated distinct centroperipheral gradients in the photoreceptor layer for TGF-β2 and TGF-β3. Relative high amounts of TGF-β in the fovea could affect vascular patterning due to TβRI seen in astrocytes which lead the blood vessels at the foveal rim at the level of the ganglion cell plexus. TGF-β2 and TGF-β3 expression is detected before formation of the foveal avascular zone (FAZ) at fd64 (~15WG) - fd73 (~17WG) with levels peaking in the foveal region at fd105 (~25WG) by the time the FAZ forms. Conclusions: This thesis has shown that EC proliferation in the choriocapillaris does not appear to be promoted by increased metabolic activity in central retinal neurons as reduced rates of EC proliferation in the ‘foveal’ chorioretinal location were observed at all ages studied between 14 and 18.5WG. The findings suggest that mechanisms regulating proliferation and growth of the choroidal vasculature are independent of differentiation in the neural retina and are therefore different to those governing the formation of the retinal vasculature. All TGF-β isoforms are expressed in developing and adult human and macaque retina with TGF-β2 being the predominant isoform. TGF-β isoforms are more abundant in central compared to peripheral retina and in developing compared to adult retina. Centro-peripheral gradients of TGF-β2 and TGF-β3 across the photoreceptor layer and TβRI on astrocytes support the presence of TGF-β in the fovea as an antiproliferative and antiangiogenic factor by helping to define the FAZ early in development, well before 23-25 WG in humans and before fd100 in macaques.

Doctor of Philosophy (PhD) / Background: The development of the blood supply in the primate retina has been extensively studied; however the relationship of the differentiating retina to the choroidal blood supply is less well known. The interaction of astrocytes and vascular endothelial cells promotes the development of the retinal vasculature from 14 weeks’ gestation (WG). Initially, astrocytes lead the developing capillaries from the optic nerve towards the macular area. However, neither astrocytes nor endothelial cells enter a prescribed avascular area, within which the fovea later forms. This may be attributed to expression of a factor that inhibits astrocyte and endothelial cell proliferation in the fovea. A factor found in the CNS that is already known to have these effects is transforming growth factor-β (TGF-β). Aims: This thesis investigated the relationship between retinal maturation and choroidal blood vessel supply and the possible role for TGF-β as an antiangiogenic factor in maintaining an avascular fovea during primate retinal development. Methods: Human eyes between 11 WG and 40 years were obtained with ethical approval from Prince of Wales Hospital and the NSW Lions Eye Bank and fixed and sectioned for histological procedures or prepared for polymerase chain reaction (PCR). Macaque eyes from foetal day (fd) 64 to postnatal 11years (p11y) were obtained from Bogor Agriculture University, Indonesia with the approval of the Ethics Committee of the University of Washington, Seattle, USA. Macaque eyes were also fixed and sectioned for immunohistochemistry and in situ hybridisation. RNA was extracted from human foetal retinas and used for RTPCR (Reverse Transcriptase PCR), QPCR (Quantitative PCR) and preparation of riboprobes. PCR products were analysed using both restriction digest and sequencing. RTPCR was used to identify TGF-β1, TGF-β2 and TGF-β3 in the developing human and in the developing and adult macaque retinas whilst QPCR was used to quantify the TGF-β isoforms in central compared to peripheral retina and in foetal compared to adult retina. In situ hybridisation was performed according to a standard protocol and visualised using Roche HNPP Fast Red detection set with designed riboprobes for TGF-β1, TGF-β2 and TGF-β3 (DIG RNA labelling kit). Some sections were counterstained with vimentin antibody. Immunohistochemistry was performed on human retina and choroid sections using antibodies to CD34 and Ki67 and on human and macaque retina using antibodies to synaptophysin, vimentin, GFAP, calbindin, S-opsin, RG-opsin, rhodopsin, TGF-β1, TGF-β2, TGF-β3 and their receptors TβRI and TβRII. Sections of the retina were imaged and analysed using either a Leica Confocal microscope and TCSNT software or Zeiss Confocal microscope and LSM 5 Pascal software. Data from the human retina and choroid sections corresponding to different regions (foveal, parafoveal nasal, parafoveal temporal, nasal and temporal) was collected to measure the number of Ki-67 immunolabelled mitotic endothelial cells and the area of CD34 immunolabelled choriocapillaris using Adobe Photoshop version 5.0.2, NIH software version 1.62 (measurement macros) and Excel. In the human and macaque sections the intensity of TGF-β protein and mRNA expression was captured from different regions of the retina (foveal, parafoveal nasal, parafoveal temporal, nasal, temporal, nasal to disc) to compile montages. Montages were then re-imported into LSM 5 Pascal software to measure the optical density across each montage along the ganglion cell layer, outer neuroblastic zone and photoreceptor layer collecting data in Excel for graphical representation. In addition to the montages, individual sections were assessed for co-localisation of TGF-β and TβR to various retinal cell types. Results: Analyses of choriocapillaris area and endothelial cell (EC) proliferation were able to demonstrate that the area of choriocapillaris endothelium is greater in the foveal region at all ages (14-18.5WG), that the rate of choriocapillaris EC proliferation declines dramatically over this same period and that the lowest rates of EC proliferation are at the incipient fovea. Most importantly these findings indicate that EC proliferation in the choriocapillaris does not appear to be promoted by increased metabolic activity in central retinal neurons which are more developed with higher oxygen and nutrient demands, which is the mechanism widely thought to regulate development of the retinal vasculature. PCR showed all TGF-β isoforms to be present in the human developing and adult retina. QPCR revealed that TGF-β2 was the most predominant isoform, followed by TGF-β3 with very small amounts of TGF-β1 seen. The isoforms were more abundant in developing rather than adult retina and in central rather than peripheral retina. Studies of the distribution of TGF-β protein and mRNA using immunohistochemistry and in situ hybridisation confirmed the low levels of TGF-β1 protein and mRNA observed in QPCR and demonstrated distinct centroperipheral gradients in the photoreceptor layer for TGF-β2 and TGF-β3. Relative high amounts of TGF-β in the fovea could affect vascular patterning due to TβRI seen in astrocytes which lead the blood vessels at the foveal rim at the level of the ganglion cell plexus. TGF-β2 and TGF-β3 expression is detected before formation of the foveal avascular zone (FAZ) at fd64 (~15WG) - fd73 (~17WG) with levels peaking in the foveal region at fd105 (~25WG) by the time the FAZ forms. Conclusions: This thesis has shown that EC proliferation in the choriocapillaris does not appear to be promoted by increased metabolic activity in central retinal neurons as reduced rates of EC proliferation in the ‘foveal’ chorioretinal location were observed at all ages studied between 14 and 18.5WG. The findings suggest that mechanisms regulating proliferation and growth of the choroidal vasculature are independent of differentiation in the neural retina and are therefore different to those governing the formation of the retinal vasculature. All TGF-β isoforms are expressed in developing and adult human and macaque retina with TGF-β2 being the predominant isoform. TGF-β isoforms are more abundant in central compared to peripheral retina and in developing compared to adult retina. Centro-peripheral gradients of TGF-β2 and TGF-β3 across the photoreceptor layer and TβRI on astrocytes support the presence of TGF-β in the fovea as an antiproliferative and antiangiogenic factor by helping to define the FAZ early in development, well before 23-25 WG in humans and before fd100 in macaques.

Purpose: To review the roles of the different vascular beds nourishing the inner retina [retinal ganglion cells (RGCs)] during normal development of the human eye, using our own tissue specimens to support our conclusions. Methods: An extensive search of the appropriate literature included PubMed, Google scholar, and numerous available textbooks. In addition, choroidal and retinal NADPH-diaphorase stained whole mount preparations were investigated. Results: The first critical interaction between vascular bed and RGC formation occurs in the sixth to eighth month of gestation leading to a massive reduction of RGCs mainly in the peripheral retina. The first 3 years of age are characterized by an intense growth of the eyeball to near adult size. In the adult eye, the influence of the choroid on inner retinal nutrition was determined by examining the peripheral retinal watershed zones in more detail. Conclusion: This delicately balanced situation of RGC nutrition is described in the different regions of the eye, and a new graphic presentation is introduced to combine morphological measurements and clinical visual field data.

Purpose: To review the roles of the different vascular beds nourishing the inner retina [retinal ganglion cells (RGCs)] during normal development of the human eye, using our own tissue specimens to support our conclusions. Methods: An extensive search of the appropriate literature included PubMed, Google scholar, and numerous available textbooks. In addition, choroidal and retinal NADPH-diaphorase stained whole mount preparations were investigated. Results: The first critical interaction between vascular bed and RGC formation occurs in the sixth to eighth month of gestation leading to a massive reduction of RGCs mainly in the peripheral retina. The first 3 years of age are characterized by an intense growth of the eyeball to near adult size. In the adult eye, the influence of the choroid on inner retinal nutrition was determined by examining the peripheral retinal watershed zones in more detail. Conclusion: This delicately balanced situation of RGC nutrition is described in the different regions of the eye, and a new graphic presentation is introduced to combine morphological measurements and clinical visual field data.