The Griffoniasimplicifolia I Isolectin B4 and Tomato Lectin Recognize Astrocytes instead of Microglia in Adult Hamster Retina

Abstract

Two lectins, Griffonia simplicifolia I isolectin B₄ (GS-I lectin) and Lycopersicon esculentum (LE) lectin, reported to be specific microglia markers in a number of animals including rodents, were used to stain hamster retinas. In the adult hamster retina, both of them failed to stain microglia but instead labeled astrocytes (in addition to endothelial cells). GS-I lectin-positive astrocytes were mainly located at peripheral regions of the retina but were variable in number and staining intensity, while LE lectin stained most astrocytes as judged by co-localization with anti-glial fibrillary acidic protein staining. To see whether astrocytes (and/or microglia) would be labeled by the lectins during postnatal retinal development, retinas from newborns all the way to adults were studied. Starting from P0, GS-I lectin selectively labeled astrocytes located adjacent to and peripheral to the tips of developing blood vessels, suggesting a functional association of glycoproteins containing α-D-galactose (the terminal sugar residue with binding affinity to GS-I) with the migration of the astrocytic front during development. The GS-I lectin staining on this population of astrocytes became reduced as the astrocytic front reached the retinal periphery, so that in the adult only astrocytes in the periphery became positive for GS-I. In contrast, LE lectin stained most astrocytes in the developing retina, starting from P2 and with maximum staining intensity achieved around P14, which continued unchanged to adulthood. During early postnatal retinal development, both lectins could label microglia, but the staining was abolished well before the retina became mature. GS-I could label microglia until P18, and the stained cells were mainly confined to the retinal periphery where blood vessels were absent, suggesting that expression of the terminal α-D-galactose residue is regulated by interaction with the developing vasculature (and/or astrocytes). LE could mark microglia till P6, but the labeling was weaker compared with GS-I. The results indicate that at least in the hamster, GS-I and LE lectin may be a more selective marker for retinal astrocytes than microglia. It would be interesting to see whether some other animals will exhibit a similar picture, and whether the expression of the sugar residues (specific for GS-I and LE lectin binding) on astrocytes instead of microglia will confer any functional consequences uniquely different from those animals with specific microglia binding.