Get access

Mapping glutamate responses in immunocytochemically identified neurons of the mouse retina

Authors

  • Daniel Sun,

    1. Department of Optometry and Vision Science, University of Auckland, Auckland 1020, New Zealand
    Search for more papers by this author
  • Michael Kalloniatis

    Corresponding author
    1. Department of Optometry and Vision Science, University of Auckland, Auckland 1020, New Zealand
    2. Department of Optometry and Vision Sciences, University of Melbourne, Parkville 3010, Australia
    3. Max-Planck-Institut für Hirnforschung, D-60438 Frankfurt am Main, Germany
    • Department of Optometry and Vision Science, University of Auckland, Private Bag 92019, Auckland 1020, New Zealand
    Search for more papers by this author

Abstract

The mammalian retina contains as many as 50–60 unique cell types, many of which have been identified using various neurochemical markers. Retinal neurons express N-methyl-D-aspartate (NMDA), α-amino-3-hydroxyl-5-methylisoxazole-4-propionic acid (AMPA), and kainic acid (KA) receptor subunits in various mixtures, densities, and spatial distributions. Ionotropic glutamatergic drive in retinal neurons can be mapped using a cation channel permeant guanidinium analog called agmatine (1-amino-4-guanidobutane; AGB). This alternative approach to physiologically characterize neurons in the retina was introduced by Marc (1999, J Comp Neurol 407:47–64, 407:65–76), and allows the simultaneous mapping of responses of glutamate receptor-gated channels from an entire population of neurons. Unlike previous AGB studies, we colocalized AGB with various macromolecular markers using direct and indirect immunofluorescence to characterize the glutamate agonist sensitivities of specific cell types. Activation with NMDA, AMPA, and KA resulted in AGB entry into neurons in a dose-dependent manner and was consistent with previous receptor subunit localization studies. Consistent with the various morphological phenotypes encompassed by the calbindin and calretinin immunoreactive cells, we observed various functional phenotypes revealed by AGB labeling. Not all calbindin or calretinin immunoreactive cells showed ligand-evoked AGB permeation. A small proportion either did not possess functional glutamate receptors, required higher activation thresholds, or express functional channels impermeable to AGB. AMPA and KA activation of bipolar cells resulted in AGB permeation into the hyperpolarizing variety only. We also studied the glutamate ligand-gating properties of 3[α1-3]-fucosyl-N-acetyl-lactosamine (CD15) immunoreactive cells and show functional responses consistent with receptor subunit gene expression patterns. CD15-immunoreactive bipolar cells only responded to AMPA but not KA. The CD15 immunoreactive amacrine cells demonstrated an identical selectivity to AMPA activation, but were also responsive to NMDA. Finally, localization of AGB secondary to glutamate receptor activation was visualized with a permanent reaction product. J. Comp. Neurol. 494:686–703, 2006. © 2005 Wiley-Liss, Inc.

Get access to the full text of this article

Ancillary