QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online March 24, 2005
doi: 10.1242/10.1242/dev.01732

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Related articles in Development Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by West, H. Articles by Fruttiger, M. Search for Related Content PubMed PubMed Citation Articles by West, H. Articles by Fruttiger, M.

Stabilization of the retinal vascular network by reciprocal feedback between blood vessels and astrocytes

Heloise West^1,2, William D. Richardson^1,3 and Marcus Fruttiger^1,3,*,

¹ The Wolfson Institute for Biomedical Research, University College London, Gower Street, London WC1E 6BT, UK
² MRC Laboratory for Molecular Cell Biology, University College London, Gower Street, London WC1E 6BT, UK
³ Department of Biology, University College London, Gower Street, London WC1E 6BT, UK

View larger version (133K): [in a new window]   Fig. 1. Retinal astrocytes do not proliferate indefinitely in GFAP-PDGFA mice. (A-D) Immunohistochemistry on retinal wholemounts from P7 (A,B) and P28 (C,D) animals with an anti-GFAP antibody show increased retinal astrocyte numbers in GFAP-PDGFA transgenic mice (B,D) compared with wild-type mice (A,C). However, retinal astrocytes do not proliferate out of control in transgenic mice (D). (E) Retinal astrocytes in dissociated and overnight-cultured retinae were stained with an anti-GFAP antibody and assessed for BrdU incorporation. Retinal astrocyte proliferation ceases in both wild-type and transgenic mice at about P8. Each data point represents the mean±s.d. from triplicate cultures from four animals (data points labeled with a star are significantly different at a 95% confidence level). (F,G) In situ hybridization of retinal sections from 10-day-old mice reveals that transgene mRNA (human PDGFA) can be readily detected in the astrocyte layer (black arrows) in transgenic animals (F) but not in wild-type animals (G). (H) Combined immunohistochemistry and in situ hybridization on retinal wholemounts shows that retinal astrocytes, identified with an anti-GFAP antibody (white staining), express PDGFRa mRNA (black staining, white arrows) in 14-day-old wild type animals. Scale bars: 200 µm in A-D,F,G; 20 µm in H.

View larger version (105K): [in a new window]   Fig. 2. Retinal vascularization correlates with differentiation of retinal astrocytes. Retinal wholemounts from P5 mice were analyzed for proliferation (A-G) and stained for PDGFRa (A-C,E), VEGF (G,H) or GFAP (F,I) mRNA. Retinal astrocytes were visualized by PDGFRa in situ hybridization (dark signal in A-C,E) and cell proliferation was assessed by BrdU incorporation (red nuclei in A-C,E-G). Blood vessels were stained with anti-collagen type IV antibody (green in A,C, white in H). Proliferation of retinal astrocytes is high in peripheral, avascular areas (B) but low in central, vascularized areas (C). This was quantified by counting cells in the avascular (B in panel D) or vascularized (C in panel D) retina in whole-mount preparations double labeled for PDGFRa mRNA and BrdU incorporation (data points represent the mean±s.d. from five different animals). High magnification reveals that BrdU incorporation in the peripheral retina is limited to retinal astrocytes identified by PDGFRa mRNA (E), GFAP mRNA (F) and VEGF mRNA (G) expression. VEGF mRNA (dark signal in H) is strongly expressed in avascular areas, whereas GFAP mRNA (black in I) is expressed strongly in the presence of blood vessels. Upper box in A refers to B, lower box to C. Scale bars: 50 µm.

View larger version (83K): [in a new window]   Fig. 3. Inhibition of vascular growth in vivo prevents astrocyte differentiation. In situ hybridization for PDGFRa (A-C,E,F), VEGF (G) or GFAP (H) was combined with immunostaining for anti-collagen type IV (green in A-C,E,F) and anti-BrdU (red in A-C,E,F) on retinal wholemounts. Exposure of mouse pups to 80% O2 from P0-P8 prevents retinal vascularization (B,E-H) that normally occurs in control animals (A,C). Proliferating cells in control animals are limited to vessels (arrowheads, C), whereas many retinal astrocytes are proliferating in animals lacking retinal blood vessels (E). However, retinal astrocytes are quiescent in the very center of the retina (F). Proliferation of retinal astrocytes was quantified by counting cells in whole-mount retinae (D). Data points represent the mean±s.d. (from four different animals) of cells counted in normoxic animals (C in panel D), and in hyperoxic animals in the periphery (E in panel D) or the center (F in panel D) of the retina. In the central area, VEGF mRNA expression is low (arrow, G) and GFAP mRNA expression is high (arrow, H). Box in A refers to C; boxes in B refer to E and F. Scale bar: 100 µm in A,B,G,H.

View larger version (83K): [in a new window]   Fig. 4. Avascular areas of the retina are hypoxic. EF5 staining (red) visualizes hypoxia in retinal wholemounts from P6 mice raised in normal atmosphere (A,B) and 80% oxygen (C,D). Blood vessels were stained with anti-collagen type IV antibody (green staining in A,C). Panels B and D are identical to A and C but with the green channel omitted. The arrow in D indicates a small, circular normoxic area in the centre of the retina.

View larger version (64K): [in a new window]   Fig. 5. Normoxia induces differentiation of cultured retinal astrocytes. Cultures of dissociated retinal cells from P1 animals were maintained for 4 days in vitro at 20% oxygen and 1.5% oxygen. In situ hybridization revealed high GFAP (A) and low VEGF mRNA levels in normoxic culture conditions (C), and, conversely, low GFAP and high VEGF mRNA levels in hypoxic conditions (B,D). Double immunolabeling (E,F) shows that the GFAP-positive (red) cell population is identical to the Pax2-positive (green) cell population after 24 (E) and 72 (F) hours in culture. (G) Assessment of BrdU incorporation (red, inset) in GFAP-positive cells (green, inset) reveals high proliferation in retinal astrocytes from P1 but a reduction in proliferation over time in culture, whereas retinal astrocytes from P7 animals were quiescent throughout the culture period. (H) The proliferation decrease in P1 retinal astrocytes in culture is independent of the staining method (anti-GFAP or -Pax2 labeling) used to identify retinal astrocytes. (I) However, culturing P1 retinal cells in 1.5% O2 largely prevents the decline in astrocyte proliferation seen under normoxic conditions. Scale bars: 50 µm.

View larger version (50K): [in a new window]   Fig. 6. Diagram showing interactions between different retinal cell types during retinal vasculature development. Hypoxic retinal astrocytes secrete VEGF and induce blood vessel growth. This leads to increased oxygen tension, which acts as a negative feedback on the hypoxia response in retinal astrocytes. Blood vessels also inhibit proliferation and stimulate maturation of retinal astrocytes. It is possible that these latter interactions are also mediated by oxygen.