gms | German Medical Science

Background and Purpose: Valid quantification of cells and changes in cell number over time in murine or rat retina is an important tool in basic and translational research. Unblinded, subjective image analysis, however, is a risk factor for obtaining biased results.

Methods: Mice expressing histone-bound GFP under the Pdgfra promoter were used to identify astrocytes in the murine retina. An ImageJ macro based on the ITCN plug-in was developed to quantify retinal astrocytic density. Regions of interest (ROIs) were selected on collagenIV-stained flatmounts based on the retinal vasculature. The GFP signal was not visible when ROIs were selected, ruling out any potential user-dependent selection bias. The macro was then run on the GFP image of the same flatmount using the pre-defined ROIs to quantify astrocytic density in those areas.

Results: The blinded selection of ROIs along with the automated counting process grants user-independent results which can be easily exported to Excel. The specificity of the automated counting process for desired target cell populations is optimized through pre-defined inputs for cell characteristics in the ITCN-plugin. An intermediate presentation of the counted and visually marked nuclei allows for additional control of false-negative results. An automated documentation of each quantified flatmount allows for easy visualization and re-evaluation of results.

Conclusions: The presented macro provides objective and documented astrocyte counting in retinal flatmounts. This method might be similarly useful for cell counts of other labeled retinal cell types like ganglion cells or apoptotic cells.