I recently received three lines of a transcriptional reporter strain from NemaMetrix. The lines were made using MosSCI, so should contain a single copy of the transgene in the same location in the genome. The PCR results support this (same size insert) and expression of this fluorescent reporter appears generally the same in juveniles and adults of all three lines with one exception: One of the three lines does not show expression in the gametes and eggs while the other two do (and this is what I would expect based on gene expression data on Wormbase). The difference is most noticeable on plates covered with eggs: bright red in two lines but completely dark in the third. How can this be? If the transgene were silenced, I would expect it not to express at all, yet it is expressed in the hypodermis and other tissues the same as the other two lines. I think it highly unlikely that there could be a random mutation in some transcription factor controlling embryonic expression of the gene as I would expect that to be lethal, but I am in the process of out-crossing the strain to see if the phenotype is passed along. Is there another explanation that I haven’t considered? I am stumped.
Thank you!
The obvious thing to come to mind is repression of transgene expression in the germline. It is a well-known phenomenon, and it is my understanding (from what I’ve read and heard) that under conditions that can (sometimes) avoid inducing it it’s somewhat variable and inconsistent, so you could imagine two very similar lines not showing it while a third did.
The problem with this theory is that I’ve sort of skipped a word, above: “repression of repetitive transgene expression in the germline”. This isn’t something you’d typically expect to see with MosSci (single-copy insertions are not repetitive). Still, I could just imagine it happening that one of your three insertion events managed to get a multicopy tandem repeat at the locus? Or maybe get additional insertions off site? None of this would necessarily be detectable by PCR (the PCR would work, would not give any additional bands, and it’s very unlikely you’d be doing a quantitative PCR likely to detect a two or three fold difference in template.
It’s even conceivable that additional sequence, that isn’t an intact transgene, elsewhere in the genome could silence a cleanly generated single-copy integrated transgene, by co-suppression.
If this is a phenomenon of germline silencing, it may be possible to relieve it (at least temporarily) with RNAi targeting mut-7 or rde-2. It may also be possible to induce germline silencing of one of the currently non-silenced versions of the transgene by crossing it to the silenced one.
Apparently, germ-line silencing is also observed with some MosSCI single-copy insertion lines. Very interesting paper from the Mello lab on the mechanisms here:
https://www.cell.com/fulltext/S0092-8674(12)00764-7
Single-copy transgenes generated by MosSCI don’t necessarily escape silencing in the germ line. We generated multiple vrk-1::FP lines: we repeatedly observed specific silencing in the germ line when GFP was used (expression in soma was not silenced), whereas we got stable expression with mCherry and Dendra as FPs (data in http://dx.doi.org/10.1016/j.ydbio.2016.01.010). Having said this, I also want to point out that this is the only time we have observed silencing of a MosSCI GFP line (and we have made quite a few). More on the mechanism of germ-line silencing here: https://doi.org/10.1016/j.cell.2016.05.072