Field of Science

But it's MY figure!

The postdoc and I have a minireview coming out in an ASM journal, and we're at the 'permissions' stage.

One of the figures is an explanatory diagram I drew for this article (figure on the left below).  It's similar to an explanatory diagram I drew for an article we published in Genetics a few years ago (on the right below).  Because one of the reviewers asked if our minireview figure was adapted from the published figure, in the Figure Legend we wrote '(adapted from (91)' even though it wasn't. 
.

Now the Production Editor is asking us for the formal permission from Genetics to republish this figure. I explained that it wasn't really adapted from the published one, but because the figures are similar she recommends that we request permission.

OK.....

So Genetics has a 'Get Permissions' link beside each article, and that sends me to the Copyright Clearance Center, where I specify that I want permission to republish their content in a journal.  I click on Price & Order (not a good sign). 


That takes me to another form, asking me questions I don't understand ('Duration of use'?  'Lifetime unit quantity'?).  Whatever.


I click on 'Get Price'.


I click on 'Continue'.  This takes me to a 'Login' page.



 Of course I don't have an account.  I create an account.


Now I log in with my account, but I don't see any record that they remember what I was asking for.  So I do it all again.

Several more pages.



 I click 'Add to Cart'.

My order is not complete until I click 'Checkout'.

My order is not complete until I accept the Terms and Conditions.




OK, my order is complete.  They're sending me an email confirmation.  They say it could take 15 days....

(Added later:  In fact it took less than 24 hr!)


On the other hand, I  just also requested permission to use a figure from our Nucleic Acids Research paper.  Very quick and easy - permission granted immediately.

List of RNAseq analyses to do

In case we have money for lots of RNAseq runs, here's a list of every sample that might be useful:

24 samples = 1 run:
Replicates of the samples we've already done:  MIV competence induction at t=0, t=10, t=30 and t=100:
  • 1 x KW20, 
  • 1 x sxy-, 
  • 2 x ∆HI0659, 
  • 2 x ∆HI0660.
Miscellaneous samples (3 replicates of each?):
  • KW20 in late-log (time of max 'spontaneous' competence)
  • KW20 in log phase + hydroxyurea: t=0, t=30? t=60? 
  • HI0659/0660 double mutant, in MIV at t=0 (= log phase) and t=???
  • murE749 mutant in log phase growth and stationary phase
  • Other hypercompetence mutants:  murE750, 751, 752?  in log phase growth
  • mystery hypercompetence mutant RR753 or backcross, in log phase and late-log?
  • KW20 in log phase growth as control (duplicating earlier MIV t=0?)
  • gcvA mutant in MIV?  or in murE749 background in log phase?
  • hfq mutant in MIV?  in log phase too.
  • purR? purH, crp? cya?
  • Cells in MIV + AMP.

What I've done lately

('Lately' being the 20 months since I last updated the Table of Contents of my lab notebook.)


I've been keeping a Table of Contents of my lab notebooks since I was a grad student, initially on paper but for the past 20+ years as a Word file. As can be seen from the screenshot above, each experiment has a number, and I record the date and a few words about what I was trying to do and what I found. Before I kept this blog, searching it was the easy way to find experiments on a particular problem, and it's still a very valuable resource.

One good thing about keeping a Table of Contents is that updating it forces me to go back over every experiment I've done lately.  So I've just done that for everything since May 2012, and it was very informative.  Here I'm going to write a summary of the experiments that I'd like to now follow up on, especially noting where RNAseq would be appropriate.

Experiments with the HI0659/0660 antitoxin/toxin genes:  We had found that a HI05659 knockout mutation completely prevents competence, that a HI0660 mutation doesn't affect competence, and that both genes were homologous to a known antitoxin/toxin pair.  This suggested that HI0660 encodes a toxin that blocks competence and HI0659 encodes an antitoxin that blocks this.  Now I've shown: Competence is not restored by the sxy1 or murE749 hypercompetence mutations, which act by increasing sxy expression. The HI0659 mutation does not cause a dramatic change in expression of lacZ fusions to the competence genes comA and rec2.  A HI0659/0660 double mutant transforms normally, confirming the hypothesis that HI0659's job is to block the activity of HI0660.  The HI0659 mutant grows just like wildtype in a BioScreen culture, suggesting that the toxin is either not expressed in noncompetent cells or has no activity that affects growth.  Gene expression in these mutants has been examined by RNAseq.  I'll describe these results in another post, but replication is needed.

Experiments with the hypercompetence mutation murE749 and peptidoglygan recycling mutations:  I first remade all the strains and rechecked the MIV-induced competence phenotypes of our set of four peptidoglygan recycling mutants, in a wildtype background with and without added cyclic AMP and in sxy1 and murE749 backgrounds.  There was quite a bit of variation, but all were approximately normal on replication.  The occasional dramatic differences may just be noise, or could result from some sensitivity to the details of the experiments.  The only consistent difference is that the gcvA mutation (a putative regulator) was about 3-5-fold less competent than the others.  We should definitely do RNAseq of the murE749 mutant, and maybe of the gcvA mutant too.

Phenotype of a hfq mutant:  hfq encodes an RNA-binding protein that interacts with many regulatory small RNAs, so we and others have hypothesized that it plays a role in regulating the translatability of sxy mRNA.  The RA made a knockout, which I found has normal growth but a consistent 10-fold reduction in transformation in late-log, in MIV and in a sxy1 hypercompetent background.  The drop was more extreme in an overnight culture - this should be retested. This mutant is a good candidate for RNAseq analysis

Effect of hydroxyurea:  The small molecule hydroxyurea (HU) specifically inhibits the enzyme ribonucleotide reductase, which converts NTPs (RNA precursors) to dNTPs (DNA precursors).  Thus HU depletes dNTP pools and stalls replication forks (demonstrated in E. coli).  If H. influenzae competence is a response to blockage of DNA replication fork progression, HU should induce competence.  But it doesn't.  If competence protects H. influenzae cells from the harmful effects of stalled replication forks, competent cells should be less sensitive to HU.  But they aren't: cells that become transformed show the same sensitivity to HU as non-competent cells, and mutations that cause hypercompetence do not reduce sensitivity to growth arrest or killing by HU.

BUT. these is a connection between competence and HU.  One competence-induced gene turns out to provide substantial protection against the harmful effects of HU.  dprA is in all characterized competence regulons, where it coats incoming DNA and promotes homologous recombination.  it's also present in many more species not know to ever become competent, although no non-competence function is known.  A dprA knockout has about the same effect on HU sensitivity as a mutation in recBC, which is well established as serving mainly to help DNA replication recover from stalling.  But an E. coli dprA mutant isn't more sensitive to HU.

Isolating more hypercompetent mutants:  Highly desirable, but a badly executed series of experiments.  It did produce lots of EMS-mutagenized cells stored at -80°C.  I should first test these for induced novR mutations, and if seen then redo the selection for hypercompetent mutants.   These mutants, if obtained, would be good candidates for RNAseq analysis.

Planning more RNAseq experiments

The postdoc and former RA generated some great RNAseq data, which I'll write about in another post.  But we have some money that needs to be spent on sequencing in the next couple of months, so we need to decide which additional RNA seq runs we should do.  And then I'm going to grow the cultures and prep the RNAs.



We have data sets showing how RNA levels change after transfer to competence-inducing MIV medium for several Haemophilus influenzae strains: wildtype (2 expts), sxy- (2 expts), HI0659- (1 (antitoxin?, 1 expt) and HI0660- (toxin?, 1 expt).  For each we have samples at t=0, t=10, t=30 and t=100 minutes.  (The figure shows a comparison between wildtype and sxy- at t=0 and t=10; the red circles are CRP-regulated genes and the blue ones are competence genes.)  We need to do at least one replicate of the HI0659 and HI0660 cultures.  If we also did another replicate of everything, that would be a full 24 sample run (one lane?) for the sequencer, and enough data that we could do proper statistical analyses.

But I also want to get RNAseq data for strains with other mutations, especially the hypercompetence-causing mutation murE749 in exponential growth.  This would be a single condition, replicated once or twice, so 2 or 3 samples total.  I might be able to squeeze this in with the run described above; depending on what other experiments we plan to do with these strains, two replicates of some might be enough.  Or it might be better to do a second run since we have the funds, doing a more comprehensive analysis of other conditions and other mutants too.

One condition I'd like to examine is 'late-log' growth, where wildtype cells develop moderate levels of competence.  I want to see if these levels are comparable to those in the HI0659 antitoxin mutant, which shows no competence at all although it appears to have (compared to wildtype cells) only a slight decrease in competence gene expression at t=10, no decrease at t=30, and a moderate decrease at t=100.  It's possible that the toxin acts only by decreasing mRNA levels of other competence genes, but the disproportion between its absolute competence defect and modest RNA defect makes me wonder if it also does something else.

We have three other hypercompetence mutants with mutations in murE - I don't know if it would be worth doing one sample of each of these in exponential growth.  We also have a hypercompetent mutant of unknown genotype - RNAseq might find the mutation as well as show us the RNA changes.  Might it be worth testing a crp mutant or cya mutant, to confirm our understanding of cAMP/CRP regulation?  Or the purR or purH mutant, or cells whose MIV-induced competence development has been blocked by adding AMP?  Or cells whose DNA replication has been blocked by hydroxyurea (depletes dNTP pools).  Or the hfq mutant, which has 10-fold lower competence.

Research direction and goals (pondering)

Depending on when I count from, I've spent the past twenty or thirty years trying to get people to think rigorously about whether bacteria have any processes that evolved to promote random recombination of chromosomal alleles or genes.

And I've largely failed at this.  A few people think my ideas are reasonable, but the great majority of microbiologists and evolutionary biologists continue to comfortably assume that genetic exchange happens in bacteria because it's an evolutionary good thing.  I'm still pretty sure they're wrong, but I think I've done almost all I can to change their minds.

I'm at a decision point.  The Canadian Institutes of Health Research (CIHR) has funded the bulk of my work, but recent proposals have failed and I don't have any compelling projects to propose.  The funding environment has moved on, leaving little support for 'pure' research, especially where others think your perspective is wrongheaded. So I don't think I'm going to submit a proposal, at least not this year.

I'm not going to stop doing research.  Nor am I going to shift from benchwork to bioinformatics, at least not yet and not all the way.  I have good research space and facilities, a small pure-science grant that will support my new grad student, and funds remaining in my previous CIHR grant that can probably be extended for another year.  And I have lots of projects, lots of questions I want to answer about H. influenzae competence.

So my tentative plan is to relax a bit (i.e. not write that damned CIHR proposal) and get back in the lab.  I'm also teaching two courses this term, so I certainly won't be slacking off, but I see no point in putting this extra pressure on myself yet again.

An important paper about the interactions between influenza and H. influenzae infection

ResearchBlogging.org Wong SM, Bernui M, Shen H, & Akerley BJ (2013). Genome-wide fitness profiling reveals adaptations required by Haemophilus in coinfection with influenza A virus in the murine lung. Proceedings of the National Academy of Sciences of the United States of America, 110 (38), 15413-8 PMID: 24003154

Haemophilus influenzae is a bacterium that causes respiratory tract diseases, but it's often confused with the virus that causes influenza (also a respiratory tract disease).  The modern confusion arises from the similarity of the names, but the name similarity arises from an older confusion about the cause of influenza.

The influenza virus wasn't identified until after the big 'Spanish influenza' pandemic at the end of World War I.   At the time of the epidemic the cause of influenza was still being sought, and a small heme-requiring bacterium informally called 'Pfeiffer's influenza bacillus' (now Haemophilus influenzae) was a likely suspect, since it was found in the lungs of many influenza victims.  We now know that influenza itself is caused by a family of small RNA-genome viruses, and that H. influenzae commonly causes a secondary pneumonia, especially in people whose respiratory tracts have been weakened by other diseases or by old age.

A recent paper from Brian Akerley's group set out to identify the bacterial genes that contribute to this effect.  Their hypothesis was that some H. influenzae functions that are needed for normal infection (in the absence of influenza virus) would not be needed when the virus was present, i.e. that the virus infection allows H. influenzae to take shortcuts.

Their strategy was to infect mouse lungs with a mixed population of H. influenzae mutants carrying transposon insertions in many different genes, and then examine which mutations become lost during the infection.  Cells that have mutations in genes that don't matter during infection will do just fine, but those with mutations in genes needed during infection will be unable to grow and so their DNA will be missing from the final population.  These experiments compared a healthy population of mice with mice that had been lightly infected with influenza A virus 5 days before the H. influenzae infection.  These authors had previously examined genes needed for single infection (Gawronski et al. 2009), and another group (Lee et al. 2010) had shown a few years previously that preinfection of mice with influenza increased the severity of H. influenzae infection, but this new work used a much less virulent strain of H. influenzae.

About 30% of the H. influenzae genes were excluded from the analysis, most because they were either essential or contributed to growth in lab culture medium, and some because they were duplicated or too small to analyze. This left about 1200 genes whose roles in infection and coinfection could be analyzed.

Most of these genes were found to not be important for either type of infection.  (The researchers' criteria for 'important' were quite stringent, so this doesn't mean they make no contribution.)  But 85 genes were required for both types of infection, with another 24 required only in single infection and another 18 required only in the coinfection with influenza virus.

The infection after influenza virus is probably a good model for similar human infections.  But the significance of the simple infection is less clear.  Since the bacteria are quickly cleared from the lung, it's not clear what's being evaluated.  There's also the possibility of chance effects playing a big role here, which might explain why the genes identified by this experiment are not very consistent with those found by a very similar experiment reported by this lab a few years ago (Gawronsky et al 2009).

I went through this paper in the hope that it would give us evidence of the in vivo importance of competence genes, but it doesn't.  None of the competence-inducible (CRP-S regulated) genes are important for the conditions the authors investigated.  In fact, knockouts of quite of few of them are enriched in the recovered lung samples, suggesting that these genes may do more harm than good.

I'll write a separate post considering where my research should be going.

Back to blogging

I haven't posted anything in ages, partly because I haven't done an experiment in ages.  I'm not planning any experiments right now, but I do need to do a lot of thinking and writing about research because it's once again grant proposal time.  March 3 is the deadline for CIHR Operating Grant proposals.


I've been struggling to get started, mainly because I couldn't identify an angle that reviewers would find persuasive. The more I learn about grant writing the more I realize how important it is to get the reviewer excited about your proposal; enthusiastic reviewers will overlook problems that bored reviewers would jump on.  However a discussion today with the postdoc has suggested an angle that might work well - it's at least good enough to get going with.
Topic:  In vivo role of the Haemophilus influenzae CRP-S regulon 
Background:  This regulon controls a wide range of phenotypes important for pathogenesis: biofilm formation, motility, DNA uptake, gene transfer, DNA replication...).  It's active in vivo.  We know a lot (but not enough) about its in vitro regulation.  We've made and phenotyped knockouts of every gene. 
Goals:  Develop genetic tools for in vivo studies of the H. influenzae CRP-S regulon (e.g. control and inducible m-cherry fusions).  Use in vitro and simulated-RT studies to characterize the effects of known respiratory tract conditions on regulon expression. In collaboration with an established animal-research lab (chinchilla ear or mouse lung model), carry out in vivo investigations of regulon expression and effects on pathogenesis.. one of the established animal models.  
This angle builds very well on all our work under the previous grant (we're running on a no-cost extension so I guess it still counts as our current grant) and we can submit this proposal as a renewal rather than as a new grant.

One thing we'll need to do is identify a collaborator in whose lab the in vivo work can be done.  Sending a senior grad student or postdoc to such a lab will be much more cost effective than setting up our own facilities.