TA cloning results

Yesterday I transformed E.coli with the ligation product of TA cloning. The results look good! All the controls worked as expected.

The CFU for my Rec1 experimental batch were 4*10^8, and the transformation efficiency was 1.15*10^-5,  plenty of white colonies to pick from. There are some small white colonies on most plates that look very distinct, they aren’t fuzzy and have a very flat white texture(see pic 1). I believe these are contaminants as they showed up in the no DNA controls. They are rare, making up around 5% of white colonies.

I plated on X-gal(for CFU) and X-gal+Ampicilin(For Transformants) LB plates. No IPTG was used. The controls I did are as follows:

“- Control” Ligation with vector but no insert DNA. Expected to see all blue colonies on ampicilin plates, found 44 blue colonies and 1 white colony, although it is a contaminant type (see pic 2)

“+ control” Ligation with vector and control insert DNA. Worked as expected, had lower CFU but higher TF than the rec1 plasmid

“No DNA control”  E.coli transformed with no DNA. I expected not to find any colonies whatsover on AMP plates, I found a single colony with a blue center and white outside(see pic3).

“Positive (plasmid) control” E.coli transformed with an equal molarity of an uncut vector containing AMP-r but not LacZ. I expected to see only white colonies on Amp plates and I did, although one plate had a little areas that looked like galaxy, white colonies diminished to the smallest visible size accumulated in a patch. I can’t explain this. The other identical plate was cleaner, but had a far lower TF than rec1. Also two very large colonies with blue centers appeared on the CFU plates. I can’t explain that either(contam?).

“Heat shock control” Same as No DNA control, only spent 50s at room temperature instead of the 42 degree bath. No transformants, very slightly lower CFU than No DNA control(CFUs were similar among all treatments)

Pic 1: Contaminant white colony. From plate 1, a rec1 CFU plate with x-gal but no Ampicilin.



Pic 2: The single white colony on the – control, ampicilin+x-gal plate


Pic 3. A lonely colony that is blue in the center and white on the outside that showed up on a x-gal/Amp no DNA control.


I re-streaked 10 white colonies from the Rec1 plates on X-gal/Amp LB plates and put them in the incubator, along the smaller white “contaminant” colony from pic 1. I’ll pick a colony from each restreak, and use the same colony to both inncolate broth and perform PCR to test for Rec1. Afterwards I’ll digest the strains that had the proper PCR amplification for rec1. I put the initial plates into the fridge, just incase none of the ten colonies I restreaked work I can always go back to them and pick new ones.


Sorry there haven’t been updates in a while. Things have been a little difficult in the lab recently. I’m working on a powerpoint presentation for Rosie, I might post some slides on the blog after I speak with her.

I mostly just want to jot down a thought I had during the last lab meeting. When I started in the lab I worked on dprA and comM, two competence genes that have significant negative effects on transformation when knocked out. We weren’t sure what the roles of these genes were. dprA for instance might protect single stranded DNA from degradation by exonucleases or it’s role might to bring uptaken DNA to recA. Because dprA is very widespread and highly conserved in bacteria the answer would relatively important. We used a double mutant approach to explore, but this approach has some limitations.

One of the difficulties in investigating the roles of these two genes is that we do not have good tools for investigating what goes on in the cytosol, mostly because it’s difficult to distinguish the uptaken DNA from chromosomal DNA. Transforming cells with a specially designed PCR product that has little homology to chromosomal and critically includes several primer sites that do not have homology to the chromosome. In a recA- background no recombination is possible so all the DNA that’s has been taken up and translocated in the cytoplasm will persist in the cytoplasm, being degraded by exonucleases. We have an effective protocol to rupture the outer membrane(useful for studying the periplasm),  adding DNAase after this step should degrade all the DNA that is not in the cytosol. Afterwards a DNA prep(perhaps with a wash to get rid of the DNAse) can recover just the DNA in the cytosol which will be mostly chromsomal

If dprA and comM roles are to protect DNA from degradation, the PCR fragments fed to the cells should on average be shorter in knockouts. By using primers unique to the ends of the PCR fragment, qPCR would allow for quantification of how many intact strands remain. Having a stepwise system of primer sites on the PCR fragment would allow investigation of questions such as how much degradation is going on and where.

Twim #7

Anouther very informative episode of twiM. There were two main points, the extracellular death factor in E.coli and the protein turnover in cyanobacteria allowing them to be photosynthetic during the day and nitrogen fixing at night, mostly focusing on iron recycling. This post will focuse on EDF

Extracellular death factor is a qourum sensing molecule that interacts with toxin anti toxin system in E.coli, amplifying the activity of two toxins called MazF and ChpBK. The theories behind this are a bit strange, group theoryie and I haven’t comprehended them yet. If it is related to cannabilism and bacteria programmed cell death, bacteria eating their brethren, maybe it could be tied in competence as both might be nutritional in nature, if you’re going to be cannabalistic you might as well be competent. There is an interesting paper that links it with peptidoglycan synthesis! A grad student I know in the neighboring tuberculus lab is working with toxin-antitoxin systems and I’ve learned a bit about it from him.

The really interesting thing that I got from this is that extra cellular death factor is a pentapeptide that acts as a qourum sensing molecule,  H-Asn-Asn-Trp-Asn-Asn-OH (NNWNN). Haemophilus influeenzae doesn’t have qourum sensing -I’m not sure how we know this, assuming someone tested that supernatant from late log cells doesn’t induce competence, atleast not in wildtype. However one of the theories we have about murE is that the mutations alter the proteins specifity from adding meso-DAP to adding LL-DAP creating an alterted pentapeptide. The normal Meso-MurNAC pentapeptide is found in high levels in late log and interacts with the ampR homolog to cause it to act as a repressor. An undergrad worked on this idea but nothing was ever concluded, although there should be good avenues to experiment – knocking out the gene dapF dramastically increases LL-DAP levels, from a 1:1 ratio to 5000:1(In E.coli. Some bacteria have an alternative pathway to creating meso-DAP and it’s not clear if HI does)

There’s a good post about L-forms on small things considered.

MurE information from the Ma and Redfield paper.

What we know about MurE

– Log phase growth is the unaffected in MurE mutants. The ability to recover from stationary growth is unchanged. No difference in sensitivity to osmotic shock. No differences in sensitivity to Gentamycin indicating no change to cell envelope permability. No differences in aztreonam(septation protein PBP3 inhibitor) or imipenem and mecillinam(PBP2 inhibitors) sensitivity

– MurE mutants do not constitutively take up DNA

– MurE mutants have the same specificity to the uptake signal sequence as wild type

– Mutations in essential competence genes(cya, crp, sxy, rec-2, comE, dprA?) prevent transformation in murE mutants as well. This paper is a little outdated, dprA isn’t an essential gene, its transformation frequency is just close to limit of detection

– LacZ fusions show greatly increased expression of comA and rec-2 in murE mutants, and slightly decreased sxy expression. Might be outdated data I need to check out the microarrays when I get back to the lab

Here are my initially thoughts on the murE mutant. I talked to the post doc for a while and have a little handle on it now.  We came up with two reasonablish and non mutual exclusive models for why a murE mutant could be hyper competent

Model 1) Transformation frequency increases the most after growing as cells that are growing as quickly as possible are transfered to the starvation media MIV. Perhaps this is because the cells are perceiving a signal that tells them “hey, you are not replicating anymore” and this signal is part of cell wall biosynthesis. In a murE mutant growth is not affected, but maybe the cells are getting that signal and for some reason it is only being heard by the competence regulon. What are other physiologies that could be affected by a signal that says that growth has abruptly stopped?

Model 2) MurE is the step in cell wall biosynthesis directly before the recycling pathway. If the mutation makes murE super awesome or really bad at its job this the recycling pathway pools will be affected accordingly.

Model 3) There is a transcription factor in the recycling pathway called AmpR(called that because mutations in it causes ampicillin resistance). In E. coli and Haemophilus AmpR is caled GcvA. I’ll refer to it as GcvA from now on. A brief look into the literature revealed it’s something I need to investigate fully.

EDIT: I had wrongly copied down that the transcription factor was called GcrA, not GcvA. These are two different, but related proteins, corrected. Thanks Rosie!

This pseudomonas paper will need to be read too.

Some short thoughts and assignments

– Does DNA bind to cell walls? Old paper says so but I need to find out the modern opinion. Read the Baccilus literature to see what their deal is

– There are some bacteria without cell walls, but with functional and conserved biosynthesis genes including murE. Chlamydia Papers to read.

Some first step methods to start investigating:

– If a murE mutant is complemented with WT murE(or the other way around), what happens? Does the effect go away? This is important because it will tell us if the mutation is dominant – which could mean several things like being awesome at de novo synthesis or making some strange x factor by product, or if it’s recessive and masked by a functional copy

– Clone murE into e. coli, or make an allele specific mutation of E. coli’s murE. Does it affect E.colis transformation? This question will be key, and will shape the direction of investigation.

Thats it for now. Tomorrow I’ll start gathering lab notebooks

Spare thought about rec2: Phage recombination is drastically lowered in a KO mutant. Is there a build up of phage in the periplasm? There’s a method for destroying the outer membrane and washing off everything in the periplasm. This would only be plausible if rec2 cells infected with the phage at the right temperature without needing to recombine also had lower plaques forming units(or maybe not? some reason its specific to recombination but not just infection?)


This morning the lab had a brainstorming session for potential projects I could do. The project proposal I wrote was rejected(see previous post). Here are the candidates:

1) Why is the MurE mutant hypercompetent?

MurE is an essential gene involved in the peptigoglycan pathway, and this mutant has 3 point mutations – 2 in non conserved regions and 1 in a highly conserved region. For some reason this makes it highly competent, it’s the most transformable strain we have available.

Ma and Redfield 2000. KW20 is the wild type strain, the other four are murE mutants. Transformation frequency is greatly increased in the murE mutants

We have no idea why. No one expected the cause of the phenotype to be a gene involved in cell wall synthesis. The most likely explanations have be ruled out. The cell wall isn’t any more permeable and the bacteria aren’t more sensitive to antibiotics. DNA uptake specificity is not affected. It’ a completely unexpected result that does not fit in with our current model of competence at all.

For the next few days I’ll be researching what we know about murE and learning about cell wall biosynthesis pathways to come up with some ways to approach this question. I’ll be reading E. coli papers on murE(and hoping that H.I isn’t any different), looking at old lab notebooks from different students, old microarray data, and anything else I can get a hold of.

2) Transformation in a mucous environment

How does HI transform in a more natural environment, a mucus layer? A previous student had developed a method for it so most of the tools are ready and should be very straight forwarded. I could do some interesting things, like co-cultures. However the results would be super low impact, and despite being a master in snot after completion I won’t learn any transferable skills

3) Does DNA quench competence?

Again some interest but very low impact. We have a bit of conflicted data on whether transformation is limiting or uptake, and it could be a time not quantity. The most likely answer, that cells get ‘full’ of DNA and then cant take up anymore is so boring it might not be publishable.

4) Cross-linking

This is pretty cool. Transformation is able to recombine large tracts of DNA and we have a bunch of competence genes we think are involved in protecting DNA from exonucleases.  In these mutants the transformation frequency of a marker is decreased, however we don’t know anything about the overall transformation going on. Are smaller fragments getting transformed? The reason I’d steer clear of this project is because anouther approach would be much more informative, sequencing the entire genome of transformed clones, which would not only tell us about the size of recombination tracts more exactly but also quantity. I want to work on something that will stand the test of time.

So I’ll be learning about cell walls for the next little while while all this percolates. Benchwise I’m making a rec1 clone and need to order some primers for it.

Project Proposal

Tuesday I briefly presented some of my ideas for a masters project in lab meeting. In preparation of this I wrote a project proposal document and sent it around the lab. It was very rough and incomplete, but I just wanted to get it out before the meeting. I was hoping to polish it up before putting it up here, but unfortuanetly Rosie doesn’t think it’s appropriate ( potentially very difficult to do, we’d have to invent some tools and we don’t have specialized equipment needed and even then it might not work in our organism, it’s too expansive and open ended for a masters project).  I’m moving on to other research and ideas so I’ll just post the very very rough draft here.

Davids Proposal for a Masters Project

The goal of developing a fluorescent reporter in Haemophilus influenzae is to evaluate competence gene induction and metabolism in individual cells. The competency of individual cells in a culture is thought to be heterogeneous based on observations from old congression experiments which suggest that only a minority of the cells are responsible for the majority of transformation, however these stochastic properties have not yet characterized. Additionally since fluorescence proteins can be used as a measure of metabolic activity, an indicator of the nutritional state of the cell, these reporters can be used to test the relationship between nutrition and competence. Here I will outline the research interest in single cell analysis of H. influenzae, both directly by reporters and indirectly by congression, and detail a series of experiments to test

The three main aims are as follows:

Aim 1: Characterizing congression of wildtype cultures in late log and MIV, along with in sxy-1

Aim 2: Characterizing the stochastic nature of competence, and using this information make inferences about it’s regulation

Aim 3: Determining if a relationship exists between competent cells and nutrition

Explanations of Congression

Congression is the observation that double transformants occur more often than expected, giving the interpretation that a fraction of the culture is more transformable. This is both an unexpected observation and potentially a tool for investigating the regulation of competence. This variability in phenotype is stochastic, and can result from two forms of cellular noise, intrinsic and extrinsic. Intrinsic noise is due to the random nature of chemical reactions and is exacerbated by the rarity of chemicals, which could be prominent due to haemopholis small size. Extrinsic noise is due to the differing cellular states the cells can be at, if competent cells are also more nucleotide starved perhaps cells that have recently replicated, especially if they were transferred to MIV during replication, are more competent. These two factors can be distinguished using fluorescent reporters(a). Two equivalent but differently fluorescing reporters with the same promoter will act in concert if variation is due to extrinsic factors but not if due to intrinsic factors. Interestingly stochastic effects can vary gene by gene and has been found to be correlated to function, usually interpreted as an evolutionary advantage to be heterogeneous in certain phenotypes, and this can be tested by comparing competence genes to housekeeping genes.

Nature of competence:

Approximately ten times more transformants are found after transfer to the starvation medium MIV than in late log culture. Two non mutual exclusive models with different predictions can account for this, with the truth likely falling somewhere between them.

Model 1: Ten times more cells are in MIV than late log

Prediction: Observed/Expected double transformants should increase in MIV

Model 2: Cells that are competent are ten times more competent

Prediction: Observed/Expected in MIV should be similar to late log

I’d like to test this by plating transformants. Fluorescent markers of competence induction should collaborate and resolve these predictions. I’d also like to test sxy-1 as it does not get upregulated when transferred to MIV and presumably the same proportion of cells are competent in late log and in MIV due to its stem loop being unresponsive to changes in purine abundance, and observed/expected double transformants should be the same in sxy-1 under both conditions.

Competence and Nutrition

Assuming that cells that are undergoing replication are more in need of nucleotides, and that competency is in response to nucleotide depletion, there should be some relation between the amount of DNA present in a cell and competency. Many bright stains such as syto62 are available that allow the quantification of nucleic acids in a cell. Ideally I want the use of two stains, one that binds to free nucleic pools and one that binds to incorporated DNA to discriminate between these two conditions, although I’m not aware of a stain to do this. To test the relationship of DNA availability and competency cells that have a fluorescence marker of competence can be stained along with a DNA stain, and ideally flow cytometry. This could also be tested more indirectly without the use of  fluorescent reporters, just quantifying the stains in a fluorometer and correlating with transformation frequency. Additionally I could test the relationship of constitutive metabolic rate and competence in a similar way by using a constitutive marker and a competence gene marker as constitutive markers should be dim when the cell is metabolically inactive and bright when it is metabolically active.

Although this has little to do with competence, I’m also interested in a result reported by Goodgal  that 1.6-2 cells exist in solution per colony forming unit eventually found on plates, however this wasn’t done with modern rich media.  Many of these cells may just be dead, but if they are not we are missing a large constituent of the subpopulation make up of haemophilus. A constitutive reporter gene will be able to determine if these cells are metabolically inactive

Fluorescent Reporters

Cloning fluorescent genes into a promoter of interest is a widely used technique to visualize gene expression. Candidate markers must be exceptionally bright to easily detect fluorescence in the tiny H. influenzae cells, and have a short enough lag time from translation to the maturation of the fluorophore that accurate inferences can be made about competence. Protein weight is mostly irrelevant for choosing a marker and photobleaching sensitivity depends on whether the fluorescence recovery after photobleaching(FRAP) technique will be used. In recent years new generations of fluorescent reporters have been introduced, while widely used fluorescent markers such as GFP have been optimized for use in different organisms and mutant variants constructed for use in bacteria and growth at 37 degrees. Fluorescent markers have been used successfully in haemophlius influenza[1], and modern reporters are brighter and more suitable for use in bacteria[2].

I started working with DsRED however I don’t believe it is suitable for this study due to its long maturation time, 10 hours in E.coli compared to alternative markers such as mCherrys 15 minutes[3]. I believe the best choice for studying competence in H. influenzae is mCherry due to its short maturation time and relative brightness, while td-tomato is an alternative with the greatest brightness of any spectral class but a 1 hour maturation time in E.coli[3]. If two complementary markers are needed the GFP derivative emerald would be a good choice due to its brightness, optimization for 37 degrees,  and similar maturation rate and compatibility as a dual reporter with mCherry[4]. The Venus variant of VPF is another good candidate, it has a maturation time of only 7 minutes in E.coli and nearly 80% the brightness of fluorescein, a common  and is sensitive to photobleaching which makes it ideal for FRAP

Quantifying fluorescence can be done by fluorescence microscopy, overlaying a phase contrast image and a fluorescence filter image in the same field of view and integrating fluorescence over cell area. The FRAP technique can be useful to eliminate previous fluorescence and isolating only the induction of competence.  A flourometer can be used as a control for methodology by checking if changes between conditions are proportional with those seen with microscopy.

I’m concerned that haemolipus small size will make fluoresnce hard to detect in individual cells. In E. coli 30 copies of GFP need to be present to easily see a cell. However recent advances both in markers and detection methods allow for the visualization of single fluorescent molecules. Proteins such as td-tomato are much brighter than GFP and are shifted further away from the cells background autofluoresccnce(due mostly to flavin).  Techniques like cloning a membrane targeted sequence to the marker and using a short excitation pulse(strobe photography) also help greatly with visualization. A high powered argon laser and a cooled ccd camera with high quantum efficiency are needed for effective microscopy. Flow cytometry is superior to fluorescence microscopy for detection and quantification.


I would like to thank Sunita for her continual support, advice and knowledge, Josh for his discussions and especially developing the models for MIV, and Rosie for her helpful comments and questions on previous write-ups.

(This is a very rough section I’m not sure to include/if it fits)

Does Nutrition quench Competence?

Goodgal 1961 reports that when cells are grown in media with DNA they are less transformable compared to cells grown without the added DNA. I’d like to repeat these experiments, and again see if sxy-1 is affected differently than wild type. Transformation frequency peaks around 100 minutes in MIV and then diminishes(?). Perhaps this is because if nutrient starved competent cells don’t find DNA they enter a post competent state where they are unable to transform but can still be rescued when transferred to rich media. If all cells in a culture can be made to be equally competent, and DNA ‘quenches’ competence, the observed/expected ratio of double transformants should be below one.

Previous Older Entries Next Newer Entries