Figure 2.4. Birthdate of Müller cells in the mouse retina. A. Sections showing a 3H-thymidine-labeled rods (r), bipolar cells (b), amacrine cell (a), and Müller cells (m) . B. Time course of Müller cell generation relative to that of other cell types in the mouse retina. Müller cells appear to be one of the last cells to become postmitotic in the developing retina (Young, 1985). (Copyright 1985 John Wiley & Sons, Inc., reprinted with permission.)
early stage and a second much later in retinal development. Another explanation is that Müller cells are produced throughout development, but those born early either die or are transformed into other kinds of glia (Ohira et al., 1984; Reichenbach and Wohlrab, 1986). It has also been proposed that Müller cells are born early, but re-enter the cell cycle and are detected in 3H-thymidine labeling experiments only at later developmental stages. At present, there is no unequivocal experimental evidence to support these possibilities (cf. Robinson, 1991).
Perhaps the most plausible explanation for the observed discrepancy between the 3H-thymidine labeling data and immunocytochemical results is that the antigenic markers used are not reliable identifiers of Müller cells in the developing retina. It is possible that retinal (neuron-Müller cell) precursor cells indeed express some of the antigens (Sheffield and Li, 1987; Tout et al., 1999; Seiler and Turner, 1989; Lemmon and Reiser, 1983; Willbold and Layer, 1992a, b), but as the retinal cells differentiate, the antigen is lost in neurons whereas its expression is maintained in Müller cells. This would lead to exclusive antigen localization in mature Müller cells.
There is experimental evidence to support this argument. For example the Müller cell marker, carbonic anhydrase (Fig. 2.5), is expressed by virtually all cells in embryonic tissue, but is found only in Müller cells in the adult chick retina (Linser and Moscona, 1981a). Similarly, a glucocorticoid receptor essential for glutamine synthetase induction, is found on all embryonic retinal cells but is expressed only by Müller cells in the mature chick retina (Gorovits et al., 1994). These examples illustrate the point that some well-accepted Müller cell-specific markers are probably expressed by all cells in the embryonic retina, but their expression becomes restricted to Müller cells in the adult. Therefore, some commonly used "Müller cell markers" may not be useful indicators of Müller cell differentiation during development. It appears that negative regulatory elements acting in conjunction with neuron-specific transcription factors are responsible for turning off the expression of carbonic anydrase and glucocorticoid receptor genes in retinal neurons during or after their differentiation, whereas the genes remain active in Müller cells (cf. Schoenherr and Anderson, 1995). Moreover, cellular retinaldehyde binding protein (CRALBP), the most reliable marker for Müller cells, is not expressed until postnatal stages in the retina, when Müller cells are born. It is likely, therefore, that the 3H-thymidine-birth data are more reliable, and this helps support the idea that Müller cells are generated late in retinal development.
Was this article helpful?