It's been four months since I posted about an experiment! (I discovered this because I looked back through old posts to see what experiment I should start with.) I had one experiment (A below) underway, one (B below) planned (and blogged about ) but not actually started, and one (C below) that was waiting for the RA to make a mutant.
After consulting with the RA, here are some plans:
Experiment A: The goal is to make a H. influenzae mutant strain that has two genes knocked (HI0659 and HI0660). On its own, a HI0659 knockout eliminates competence; a HI0660 knockout has no effect. Because both genes have homologs in toxin-antitoxin systems, I hypothesize that HI0659's job is to prevent HI0660 from doing something toxic when it is induced in competent cells. This predicts that a double knockout will have normal competence.
I had been going to make the double mutant myself, while the RA was on leave, but now she's back she's got this underway. Next week she is going to create the double-mutant plasmid in E. coli (she has everything ready except the electro-competent cells) I'll then transform this mutant segment into H. influenzae and test competence, with both single mutants and wildtype cells as controls.
Experiment B: I want to carefully recheck the competence phenotypes of all our hypercompetent murE mutants, under several conditions. This is basically a long series of competence assays; I just need to streak out the various strains and get to work.
Experiment C: The secondary structure of the sxy gene's mRNA regulates its expression, and the Hfq protein contributes to gene regulation by helping small regulatory RNAs (sRNAs) find and bind to their target mRNAs. So we're going to make a H. influenzae hfq knockout and test its effect on competence. The RA is making the knockout - it's at the same stage as the double-mutant knockout described above. Once she's made the knockout in E. coli (next week), I'll transform it into H. influenzae and test competence.
Experiment D: I don't think I've ever done a blog post about this - it arises out of the post-doc's experiments transforming fragments of the clinical strain 86-028NP into the lab strain Rd and sequencing the recombinants. (He doesn't seem to have posted about it either.) 86-028NP transforms about 100-fold less well than Rd. None of the recombinants acquired the full transformation defect with their segments of 86-028NP DNA, but one of them transforms about tenfold worse than Rd.
The only known Rd competence gene acquired by this recombinant is comM, and the post-doc has hypothesized that its lower competence is due to replacement of the Rd comM allele with its 86-028NP homolog. ComM increases transformation frequencies by protecting incoming DNA strands from degradation in the cytoplasm. Knockouts have normal DNA uptake but about 50-fold lower transformation frequencies. (In other species the transformation defect is more severe.) Under the post-doc's hypothesis, the 86-028NP allele would be less active than its Rd homolog.
One way to test this is to restore the Rd comM allele to this transformant (replacing the one from 86-028NP but not the other 86-028NP sequences) - if differences in comM are responsible for the differences in transformation, this should increase transformation frequencies back to the Rd level. The RA has tried to do this but the construction didn't work - I may try it again. As an alternative she's put plasmids carrying the Rd or 86-28NP alleles into the recombinant strain, and I'm going to test these strains for differences in transformation frequency.
Peter Thiel's interesting comparison of biotech with software
21 hours ago in The Curious Wavefunction