Now that the authors have established that 183C is in the DRG neurons of the mouse embryo, they’d like to determine what exactly these microRNAs do. To do that, they need some fancy genetic techniques that we’ll need to get a handle on to understand these experiments.
This is some complicated stuff (at least by my standards; my genetics class was painful). There will be jargon, but I’ll explain it, and it’s really quite fascinating if you take your time with it. So let’s start with the general picture and then work our way down to the details. The authors’ strategy was, first, to genetically engineer some mice that would not express 183C in certain types of neurons. How do you get that kind of control? Essentially, you place some markers before and after the DNA sequence coding for 183C, called loxP sites. Tasty on bagels. There’s an enzyme (Cre) that can bind loxP sites quite well, and cut out the DNA in between the loxP sites (called the “floxed” gene, I can’t make this stuff up). If you can give your mice the gene for Cre, and ensure that gene only gets expressed (produces the enzyme) in the neurons you want, then you’re good to go.
To do that, you hijack a neat little property of genetics. DNA doesn’t just get transcribed into RNA and then translated into protein constantly. That would be inefficient and almost certainly kill you and these poor mice. Instead, a gene typically needs to be bound at some site (called a promoter) to a protein called a transcription factor, which gives RNA the go-ahead to assemble using the DNA as its template (or it might inhibit this transcription process, or adjust its rate). Different cells, at different times in their cell lifetimes, have different transcription factors lying around, so if you can design your mouse mutant such that the sequence coding for Cre comes right after a promoter that corresponds to a transcription factor unique to your preferred cells, you can do your science thing.
The authors don’t explain exactly which cell types each of their promoters of choice correspond to, but the Jackson Laboratory clarifies some of the known expression patterns of the transcription factors these authors use for their genetic control. Wnt1-Cre mutants are predicted to abolish expression of floxed 183C in the “midbrain and developing neural tube.” Initially I wasn’t sure how to rationalize this, considering that evidently the DRGs develop from the neural crest, not the neural tube – however, I took a peek at the source that the authors referenced as their precedent for the Wnt1-Cre construct, which notes that while Wnt1 is restricted to the midbrain in early development, it later manifests in the dorsal spinal cord, which is where we’d expect DRGs. TH-Cre corresponds to “catecholaminergic cells” (cells that produce a class of neurotransmitters including dopamine, epinephrine, and norepinephrine). The authors’ source for this one is behind a paywall that my university journal subscriptions can’t overcome, alas. And TrkB<sup>CreERT2/+</sup> is a mess. To get even more control over these mice’s genes (on the temporal level), the Cre protein is modified so that it can only access the nucleus (and thus wreak its havoc on floxed genes) when the mouse is given a drug called tamoxifen – that’s what the “ERT2” part indicates. The “TrkB” part means that Cre gets expressed in TrkB-rich cells, which are “low-threshold mechanosensory” neurons that are “lightly myelinated.” Translation? These are neurons that give the mouse its sense of touch, and they are extremely sensitive to stimuli; this myelination business refers to the degree to which the neurons are “insulated” by glia, so lightly myelinated neurons are less insulated and do not conduct action potentials relatively fast. The explanation behind how exactly that works is fascinating, but not something I think is essential for this post.
On top of this, just to make sure the effects on the microRNA they were looking for were results of Cre’s removal of floxed DNA in the cell types they wanted, the authors threw in another strain of mice with the ROSA26<sup>Tomato</sup> (NSFW) mutation. The babies of these mice and the strains of mice with those Cre mutations described above have the neurons of interest (those modified by Cre) fluorescing red. Hence “tomato.” Biologists are weird and I love them.
The qPCR technique we considered last time could in theory establish the complete absence of expression of 183C, if you used some controls (that is, if you ran qPCR on 183C in the cells of interest relative to a control RNA, and you compared the mutant with Cre and floxed 183C gene against mice without that combination). But the downside of qPCR is that it only tells you the average expression level across the entire tissue sample you analyze. You don’t learn anything about how the RNA expression varies across the landscape of the tissue of interest, which can be a problem if the cell types you want to analyze don’t necessarily form discrete clusters that you can harvest (like the DRG).
So instead they used “in situ hybridization” (ISH), which is a nice little nod to the Latin nerds out there. You take sections of the mouse’s nervous system, “fix” it chemically to hold the RNA in place and make the cells more accessible to complementary DNA or RNA fluorescent probes (similar to the ones used in qPCR, so it’s all coming full circle here), and throw those probes in. Then use a microscope to see the magic.