What next (turning on Sxy in E. coli)?

OK, the fusion of the ppdA promoter to lacZ produces 100-fold more beta-gal when Sxy is overexpressed. What are the important things to do with it?

1. Find out how sensitive the ppdA promoter is to Sxy: The pASKAsxy plasmid produces a great deal of Sxy when its promoter is induced with IPTG, but we know that most of this Sxy aggregates into insoluble and probably nonfunctional 'inclusion bodies'. So I could try inducing with decreasing concentrations of IPTG, and measure the effect on beta-gal production. It may be that only very slight induction of Sxy will give a big induction of the ppdA promoter. Ideally the amount of Sxy would be measured too, but we don't (yet) have the antibody to E. coli Sxy that would let us easily make these measurements. The beta-gal vs IPTG assay will be easy. If I find that ppdA induction needs only a tiny bit of IPTG, I'll have more confidence that natural induction will need only modest induction of Sxy expression from its normal promoter.

2. I should also do a time course of induction, so I know how quickly the steps happen (IPTG causes sxy transcription, Sxy protein is made, Sxy causes transcription of ppdA::lacZ, beta-gal is made). I can do this with the standard (1 mM) IPTG concentration I used yesterday. Yesterday I waited 2 hours after adding IPTG before assaying beta-gal; now I'll try 1 min, 2 min, 5 min, 10 min, 30 min, 60 min, 90 min, 120 min. The results will let me optimize the assay conditions I'll use when looking for culture conditions that induce Sxy expression from its normal promoter.

What culture conditions should I test first? My previous finding that the baseline expression of this fusion is independent of Sxy and CRP means that I shouldn't bother looking for conditions that reduce expression, so I won't try adding glucose to the medium to turn off CRP. I will try adding cAMP to the medium to make sure CRP is fully active. I will do a time course, following growth of the culture in LB, looking especially at expression as the cells' growth slows at high cell density. I will transfer the cells from LB to minimal salts (with a small amount of amino acids) to mimic the inducing effect of transferring H. influenzae from rich medium to MIV. Maybe I'll try growing the E. coli in BHI before such a transfer, as BHI is a much richer medium than LB. Perhaps I should get an E. coli mutant that can't synthesize purines (or pyrimidines), to see if transferring it to minimal gives induction.

Other things I might do:

Should I try to get a strain carrying this fusion in the chromosome rather than on a plasmid? This would eliminate concern about possible variations in the number of copies of the plasmid, due to differences in culture conditions. And if the chromosomal insertion was stable I could eliminate the selective antibiotic from the test cultures. This sounds like a job for recombineering - I wonder if we can still get this working.

Should I sequence the ppdD::lacZ fusion and the hofM::lacZ fusion to see if they have mutations that would explain their failure to be induced by Sxy?

Results of the obvious next step

As I posted a few days ago, I planned to introduce a sxy-expression plasmid into my cells with CRP-S promoter fusions to lacZ, to see if the CRP-S promoters are indeed induced by Sxy. I did that yesterday, and I just finished calculating the amounts of beta-galactosidase activity these cells produced with and without induction of sxy expression.

Results: Of the three fusions I tested (all that I have), one produced almost 100-fold more beta-galactosidase when Sxy was present, one produced only slightly more, and one produced even less! The exclamation mark is because I expected them all to respond similarly.

The fusion that was induced by Sxy is to the ppdA promoter (really the promoter of a 4-gene operon). The H. influenzae homologs (not pilB but comNOPQ) are in a similar operon, which is very strongly induced in competent cells and strongly dependent on CRP and Sxy. The ComN and ComO proteins are known to be required for competence. This fusion is carried on a plasmid. The 100-fold induction when Sxy is overexpressed is what I was hoping to see. This means that the low expression when Sxy is not induced is indeed baseline, so I'm comfortable with this baseline expression not being dependent on Sxy or CRP.

The results with the other fusions are surprising. The one that produced only slightly more is a plasmid-borne fusion of lacZ to the hofM promoter (really the promoter of the hofMNOPQ operon). This is homologous to H. influenzae's comABCDE operon, and like comNOPQ it is required for competence and needs CRP and Sxy for induction. Its CRP-S promoter element doesn't look significantly worse than that of ppdA, but it was induced only three-fold by Sxy overexpression.

The fusion that produced even less beta-galactosidase when Sxy was induced is to the ppdD promoter; it's integrated into the chromosome rather than being carried on a plasmid. This is the most surprising result, as the grad-student-now-postdoc-in-waiting ('gsnpiw') used quantitative PCR to show that ppdD mRNA is induced about 100-fold by overexpression of Sxy from the same plasmid I just used. Induction of Sxy in the fusion-carrying cells not only failed to induce the ppdD::lacZ fusion, it actually made these cells quite sick; the cells with IPTG grew to less than half the density of the uninduced cells, and produced only about half as much beta-galactosidase per cell. Perhaps the chromosomal insertion in these cells has somehow made them sensitive to transcriptional activation of the ppdA gene.

Anyway, the good news is that I can use the ppdA fusion as an indicator of Sxy expression, in my search for conditions that induce Sxy and thus induce expression of CRP-S genes. I'll set aside (for now) the questions about why the other fusions don't behave similarly.

(Sorry I've been too lazy to include figures in my recent blog posts; I promise to improve.)