GAIN: A graphical method to automatically analyze individual neurite outgrowth

BACKGROUND

Neurite outgrowth is a metric widely used to assess the success of in vitro neural stem cell differentiation or neuron reprogramming protocols and to evaluate high-content screening assays for neural regenerative drug discovery. However, neurite measurements are tedious to perform manually, and there is a paucity of freely available, fully automated software to determine neurite measurements and neuron counting. To provide such a tool to the neurobiology, stem cell, cell engineering, and neuroregenerative communities, we developed an algorithm for performing high-throughput neurite analysis in immunofluorescent images.

NEW METHOD

Given an input of paired neuronal nuclear and cytoskeletal microscopy images, the GAIN algorithm calculates neurite length statistics linked to individual cells or clusters of cells. It also provides an estimate of the number of nuclei in clusters of overlapping cells, thereby increasing the accuracy of neurite length statistics for higher confluency cultures. GAIN combines image processing for neuronal cell bodies and neurites with an algorithm for resolving neurite junctions.

RESULTS

GAIN produces a table of neurite lengths from cell body to neurite tip per cell cluster in an image along with a count of cells per cluster.

COMPARISON WITH EXISTING METHODS

GAIN's performance compares favorably with the popular ImageJ plugin NeuriteTracer for counting neurons, and provides the added benefit of assigning neurites to their respective cell bodies.

CONCLUSIONS

In summary, GAIN provides a new tool to improve the robust assessment of neural cells by image-based analysis.

Home-school differences in stress hormone levels in a group of Oxford primary schoolchildren

Twenty-three girls and thirteen boys, aged between 4 and 9 years, gave urine samples which were measured by HPLC for adrenaline, noradrenaline, cortisol and cortisone on a school day and on a home day. Home levels of noradrenaline, cortisol and cortisone were consistently higher than school levels. Cortisol was positively correlated with adrenaline and noradrenaline at home, but not at school. These results may suggest that the school day represents a stimulating but comparatively structured environment with little negative affect, while the home day is characterised by stimulation but relative uncertainty in social interactions.