Working with mouse models for inherited cancer syndromes, I’m accustomed to plotting Kaplan-Meier curves for cohorts of animals with different combinations of targeted tumor suppressor gene mutations. But I started out, during my postdoctoral years, with plotting Kaplan-Meier curves for cohorts of embryonic chicken or mouse neurons. To monitor the survival of sensory neurons from cranial and dorsal root ganglia, following different neurotrophin withdrawal schemes, I developed a method to map individual neurons on a grid, and then monitor their survival each day, typically for a week after treatment. My maps look like this:
I used a different colored pencil mark for each survival time point, and then plotted my Kaplan-Meier curves at the end of the experiment. This may seem like a very tedious, labor-intensive, low-tech, and low through-put method (and it is), but I can’t think of a way to automate this type of experiment. Despite their inherent beauty, fascinating behavior, and intriguing apoptotic mechanisms, embryonic neurons are fraught with several constraints and limitations, as an experimental system:
1. Early sensory ganglia contain a heterogeneous mixture that includes postmitotic neurons that have contacted their targets, postmitotic neurons that were extending axons to their targets at the time of isolation, committed neuronal precursors, neural stem cells, and glial precursors. The main reason for developing the cohort mapping method was to get around the problem of accidentally counting newly-differentiated, neurotrophin-independent neurons as mature neurons that had survived my cruel regimens of neurotrophin withdrawal.
2. Many embryonic sensory ganglia are small, contain relatively few cells, and are difficult or time-consuming to microdissect. You cannot freeze and thaw dissociated neuronal cells as if they were tumor cell lines or embryonic stem cells.
3. When comparing the responses of neurons isolated from mice harboring different combinations of targeted mutations, the number of samples from homozygous mutants might be limited, especially if the genotype is embryonic lethal. If the mutants can’t be distinguished morphologically from wild-type or heterozygous littermates, then ganglia can’t be pooled for different treatment paradigms.
So no reams of microarray data or FACS analyses from this type of experiment. Instead, more maps of neurons that look as if they were produced by Joan Miró on crack.
How unbelievably work-intensive! It’s neat that you really get to see the morphology of each neuron with this technique. I did a lot of work with adult mouse trigeminal ganglia, but we always plated them in the presence of support/feeder cells. You could hardly make out which cells were the neurons. I was doing neurotrophin withdrawal studies myself, oddly enough (in the context of HSV reactivation). How could you tell mature, neurotroph-indep neurons from newly differentiated ones? Morphology alone, or were you staining for something? Just curious. That project tanked for me, big time, after 2 years of crazy amounts of work (and insane numbers of mice). Sigh.
I used morphology alone to identify neurons, and drew their positions on the grid between 3 and 12 hours after plating the dissociated ganglia. Early on, I did some immunostaining experiments to confirm that “morphological” neurons also expressed differentiated markers.
If necessary, it’s possible to distinguish between neurons that have and have not contacted their peripheral targets, by retrogradely labeling from those targets with diI, prior to ganglion microdissection. Easier to do this in chicken embryos, through a window in the eggshell, but also possible to do with organ or slice culture in mouse.
I did a lot of work with adult mouse trigeminal ganglia
My favorite cranial sensory ganglion in rodents … allowed me to isolate enough normal peripheral nerve and ganglion tissue, as a control for Schwann cell tumors in mutant frequency assays. I’ve also microdissected trigeminal ganglia from mouse embryos at day 10; the medical students are surprised when I tell them that this is possible, as they struggle with their cadaver dissections.