Yet more Part II: Gram positives

Beyond the Proteobacteria but still Gram negative (Acidobacteria, Aquificae):

• No transformable species have been described (but Aquifex has PilT). 

Gram positive bacteria:

•  No reports for Leptospira or Borrelia or Chlamydia or Bacteroides.

 Actinobacteria (the big orange clade, Corynobacterium, Mycobacterium, ):

•  Only one species reported to be transformable, but Johnsborg speculates that many may be.

Cyanobacteria:
Several species known to be competent

Writing goals

The writing book I'm rereading (How to Write a Lot, by Paul J. Silvia) says that, in addition to scheduling specific time slots for writing, I should have a list of writing goals that I periodically update.  I haven't made such a list yet, because my scheduled writing has only had one goal at a time - the chapter for the Roth book (out any day now), the short essay on evolution for the Darwin-year ASM book (vaporware?), the big review on regulation I'm working on now.  What other writing might I want to get done?

Here's my first try at a list of writing goals:

The molecular-biology-focused review of the regulation of competence.

The shorter evolution-focused review of the regulation of competence.

A reworked version the Darwin-year essay for some other destination.

The visiting student's manuscript on Gallibacterium competence (needs only a couple of paragraphs).

The post-doc's first paper on recombination tracts.

The post-doc's first paper on uptake specificity.

A short paper reporting our old results on mucins and transformation.

Future papers (results not in yet)?

A paper on natural competence in E. coli.

A future paper on the phenotypes of the new knockout mutations.

A future paper on recombination tracts.

A future paper on mapping loci that cause strain differences in transformability.

On teaching:

A rant about the inadequacy of current genetics textbooks (for Genetics?  Or Nature Reviews Genetics?)

Reorganization of Part III

I have lots of bits of text and ideas for the final section of the review, but before I write any more I need to sort out their organization.

Distinguish (here or earlier) between the signal-transduction mechanism and the signals themselves, and between these and the output (the changed phenotype and its consequences for survival and growth).

1.  Summarize the results of the survey:

Great diversity in regulatory mechanism, diversity in signals, core set of induced genes plus wide diversity of other induced genes.  Regulatory information is available for only a few species.  Where we do have information for close relatives we usually see differences, and at greater evolutionary distances the regulatory mechanisms appear unrelated.  Also wide variation in the non-core genes in the regulon (both number and functions).


2.  Practical applications of this information:

Does the current knowledge of how competence is regulated suggest better ways of inducing competence in lab cultures?  The standard 'competence rituals' for the model organisms were developed by trial and error, long before we understood their regulation.  Are they unnecessarily troublesome?  Can new methods increase transformation efficiencies/frequencies?  Use of mutations and plasmids that increase competence?

Does this knowledge suggest ways of inducing competence in bacteria where it is apparently unregulated (Neisseria sp.), or where little is known about regulation (Helicobacter?)?  Look at this in phylogenetic context?  Or is regulation so evolutionarily variable that even comparisons within families are not very predictive?

Does it suggest better ways of testing for competence in bacteria not known to be transformable? For example, does the regulation by cAMP and CRP shown in H. influenzae and postulated (better word) in related species suggests that other members of the Past, Ent (?) and Vib families that induction/stimulation by cAMP should be checked in other members of these families.  Perhaps it should also be tested in other species where cAMP and CRP have regulatory roles, such as throughout the gammaproteobacteria.  Such analysis has been recently done for Streptococci Halvorstein 2010 Mol Micb 78:541).

3. What are the broad questions about the role(s) of competence in bacteria? 

Explain the controversy about function (nutrients, repair, recombination).  Keep this whole section and the next quite short; save the detailed exposition for the other review.  (OK to initially write the details here and then move much of them to the other article.)

What are the predictions of the different hypotheses?  If uptake is selected because incoming DNA provides templates for DNA repair, competence should be regulated by the same damage signals that induce recA, or that induce the RecA-regulated SOS response.  If uptake is selected because incoming DNA provides nucleotides and other nutrients, competence should be regulated by nucleotide pools and/or by processes that sense availability of sources of C, N and P.  If uptake is selected because incoming DNA sometimes carries beneficial new alleles that replace inferior alleles in the chromosome by recombination, then competence should be regulated by ... what? ... I've suggested that under this selection competence might be expected to be a 'when all else fails' response, induced when the cell's other stress responses have been mobilized but have failed to solve the problem.  How strictly this test (what test? the test of whether the other stress responses have worked?) is applied would probably depend on how costly DNA uptake and recombination were, considering both the physiological costs/risks of DNA uptake and the genetic costs of recombining in alleles that reduce fitness.

'Stress' is difficult to quantitate.  We can measure the decrease in growth rate or numbers of viable cells caused by a macroscopic perturbation of culture conditions.  But often our best evidence of the importance of disruptive events is the presence of evolved mechanisms to prevent them or mitigate their effects.  For example, the presence of a gene for photolyase is evidence that pyrimidine dimers is not only a tol that lets us measure the effect of unrepaired damage on viability, but evidence that this effect has been important over evolutionary time.  The logic is straightforward for processes whose benefits and costs are direct,  but where the effects are indirect the inference is always on shaky ground, with interpretations compromised by our guesses of what might matter to the bacteria.

Lack of nutrients is often considered as a form of stress.  Difficulty of drawing a line between nutrient signals and signals of other kinds of stress?  Some signals are clearly nutritional - PTS sugars, but even cAMP has subtle complications in bacteria other than Ent and relatives.  Same for phosphate and nitrogen limitation?  Signals of nucleotide depletion?  Purine syn regulated by guanine and hypoxanthine, pyrimidine by post-transcrptional effects (?), other effects on transcription likely.  Secreted autoinducers ('quorum sensing') integrate both local cell density and the physical properties of the microenvironment, such that may be activated by dense populations in well mixed cultures or by single cells in confined spaces. And the term 'stress' is usually used very loosely; it is rarely well enough defined that hypotheses can be rigorously tested.  Somebody (John Roth?) suggested that rapid growth in rich medium might be more stressful than slow growth in minimal medium - maybe any condition is stressful if the doubling time is less than the time needed to replicate the chromosome.

The direct and indirect costs of genetic processes are rarely considered explicitly, but evaluating them is essential to a true understanding of function.  When cells become competent energy and materials must be diverted from other processes (e.g. growth) to synthesize and assemble the DNA uptake machinery.  Once in place, the machinery may interfere with membrane integrity or with other membrane transport processes.  The energy costs of the uptake process remain unknown, but, at a minimum they include a cost of pilus retraction of about ???? per bp.  Transport across the inner membrane???  There are also genetic (and perhaps DNA damage) costs.  The obvious one is the cost of replacing well functioning alleles with worse ones, likely to be overrepresented in environmental DNA.  The infrequent non-homologous recombination events will almost always be deleterious.  Any process that disrupts the integrity of double-stranded DNA is likely to increase the risk of DNA damage - we have no measure of this.

4.  What does the survey of regulation tell us about function?

What do we learn from the breadth of 'competence' regulons (the 'effectors')?  Are there any consistencies in the genes that don't directly contribute to DNA uptake or transformation?  Do competence regulons overlap with other global regulons?  For which species do we have microarray (or other) surveys of what genes are regulated?  (Not only H. influenzae, B. subtilis, S. pneumoniae, right?)  Do genes that enhance transformation but don't contribute to uptake have specific other functions?

What do we learn from surveying the signals that induce competence?  Given the diversity, are there any unifying features of the regulation of competence?  Is there always a nutrient component?  A 'stress' component?  What would be the most important experiments to do now?

Do the steps by which competence is regulated tell us anything beyond what we learn from the nature of the inducing signals?  Certainly quorum sensing does, but what about other cascades?

What can we conclude about how the regulation has evolved?  Which features of regulation have been conserved in related species?  Which have been particularly labile?

More Part II: Survey of other Proteobacteria

Dichelobacter is competent; it's a gamma.

Natural competence has been investigated in the beta- and epsilonproteobacteria, especially in Neisseria and Helicobacter.  However, although much is now known about their mechanisms of DNA uptake and the proteins responsible, little is known about regulation.


Betaproteobacteria:  Neisseria meningitidis and gonohorrhoeae:  DNA uptake requires many of the same genes as in H. influenzae.  No homologs of CRP or Sxy.  Competence appears to be constitutive under typical laboratory culture conditions - transformation frequencies are typically high throughout growth in broth and in cells grown on agar plates.  Is piliation also constitutive?  Mutation hunts for genes needed for transformation have not identified any regulatory genes (refs?).  One PTS gene involved (find ref)?  Any in vivo hints that regulation exists?  (discussion material?)


Eikenella is a competent Neisseriaceae (Tonjum 1985); she later cites this paper as evidence that competence is coupled to expression of T4P.  Also Kingella (Weir and Marrs 1992).

• Burkholderia and Ralstonia(family Burkholderaceae)  Thongdee et al 2008, B. thailandensis and B. pseudomallei.  Transformation in defined medium over 6-36 hr with DNA.  Ralstonia (Bertolla et al 1997)  Transformation highest at OD=.8.
• Thiobacillus (Family Hydrogenophilaceae - other genus Hydrophilus)  Log phase cells put on agar with DNA for several hours. Yankofsky et al 1983
• Achromobacter (Family Alkaliginaceae, also Bordetella (not competent?))  Transformation on agar plates, not quantitative, no regulatory info. (Juni & Heym 1980)

Alphaproteobacteria?

• Methylobacterium organophilum (O'Connor et al. 1977)  High TF, sharp peak of competence at end of log phase.
• Bradyrhizobium
• Agrobacterium tumefaciens

Deltaproteobacteria

• No transformable species have been described. 

 Epsilonproteobacteria

• Helicobacter (Baltrus and Guillemin 2007 did a time course.  Although phases of growth were not well defined (by the authors/experiments), measures of transformation (not frequencies but something weird) rose as culture growth slowed in late log, fell , and then rose and fell again after cfu/ml had begun to decrease.  They also looked at two other strains.  But I'm not sure that their measures mean much at all.  (RRResearch post?)   In another paper (Israel 2000?) Transformation frequencies rose and fell as cultures became dense.  Tells us that competence is regulated by growth conditions, but no details. 
• Look at a new paper (PLoS Pathogens 2010).
• Campylobacter:  Find the info Erin sent.

Beyond the Proteobacteria but still Gram negative (Acidobacteria, Aquificae):

• No transformable species have been described (but Aquifex has PilT).

Lists of people to email or thank

People to email for ideas/sources/information/papers in press:

Karen Meibom

People to thank:

Bob Hancock

Wilfried Wackernagel

Begin Part II (the survey of competence regulation)

(Put elsewhere the discussion of how role of T4P in DNA uptake means we often can't separate T4P regulation from competence regulation.  In Part I?  Is there any T4P regulation independent of 'competence' regulation?)

An introductory paragraph:  Should we relate regulation to the old framework of 'early' and 'late' competence genes?  Is this still useful? Also for each species the traditional competence rituals used in the lab (if they exist), and the levels of competence achieved (with whatever DNA) and fraction of cells that are competent.  For many species this is all that's known.

Also explain that the species are considered here in phylogenetic context (not just gram+ vs gram-).  Refer to the detailed survey by Johnsborg et al (2007 review).  Should this start with H. influenzae and other Paseurellaceae?  Yes, as this provides a good framework for the Vibrio information (better than Vibrio first).  So have a couple of paragraphs about H. influenzae, then one about other Pasteurellaceae. 

Our survey begins with the Gamma proteobacteria.  Ref for phylogeny Williams et al 2010 (http://jb.asm.org/cgi/content/full/192/9/2305).  Competence is common and well studied in several families, and regulation appears to have been conserved in the clade containing the families Pasteurellaceae, Enterobacteraceae and Vibrionacea (ref Cameron 2006).  We begin with H. influenzae because its regulation is the best known.

Pasteurellaceae:  Haemophilus influenzae:  Competence is traditionally induced by transfer of exponentially growing cells from rich medium to a starvation medium lacking nucleotides and essential cofactors.  This gives high transformation frequencies (10-3-10-2 with chromosomal DNA), with most cells in the culture competent.



H. influenzae has a compact competence regulon, consisting of 25 genes under the control of promoters activated by the regulatory proteins CRP and Sxy (called TfoX in some species).  Most of these genes have been directly or indirectly implicated in DNA uptake, and several more act on DNA intracellularly; only a few (how many?) have unknown functions and none (?) have known functions unconnected to competence or transformation.  The first step in competence induction is transcription of sxy, which requires both the transcription factor CRP and elevated levels of its cofactor cyclic AMP.  As in E. coli, production of cAMP occurs when the phosphotransferase system lacks preferred sugars to transport.  However this transcription is not enough, as effective translation of the sxy transcript requires additional signals, thought to derive from depletion of purine nucleotide pools.  Sxy translation is regulated by secondary structure (all homologs produce mRNAs with long untranslated leaders).  Once both Sxy and active CRP are present, they together activate transcription of genes with the CRP-S promoter motif; these genes comprise the competence regulon.  The mode of action of Sxy is not known, but is thought to involve direct contact with CRP (Sinha 2009).  Similar CRP-S regulons are also present in other Pasteurellaceae and in the Enterobacteraceae and Vibrionaceae (see below).
(CRP thus acts both early and later...)


Other Pasteurellaceae:  Similar mechanisms regulate competence in other Pasteurellaceae (ref Maughan chapter). Although only some Pasteurellacea are known to be naturally competent in the laboratory, all sequenced genomes have the same competence genes as H. influenzae, with CRP-S motifs in their promoters.  Evidence that these regulated genes are needed for competence, or that the regulation uses the same mechanism? A. pleuro competence needs Sxy (Janine?), so does Act. act (Bhattacharjee 2007).  Evidence that cAMP stimulates competence?  Any evidence that regulation might be different?


Competence in the Enterobacteraceae?  (There are no Enterobacteria in the Johnsborg list.) Enterobacterial genomes contain homologs of many of the H. influenzae competence genes, and the complete genomes analyzed in detail appear to contain CRP-S regulons like that of H. influenzae.  However there are no reports of natural transformation for any of the Enterobacteriaceae.  One interpretation is that cells use these genes to take up DNA, but that the DNA is quickly degraded and rarely or never recombines with the chromosome. (Figure of gammaproteobacterial competence genes and phylogeny?)

(Nominalization alert)  Although several protocols have been described that allow plasmid transformation of E. coli without the standard permeabilization treatments (cold divalent cations, electroporation), these do not require any of the genes specific to natural competence.  (One exception is our finding that low-calcium transformation requires sxy...) 

Artificial overexpression of E. coli sxy does induce the CRP-S regulon genes but its toxicity precluded investigation of competence.  Comprehensive screens for culture conditions that induce the CRP-S-regulated ppdD gene (encoding the type IV pilin) have been unsuccessful.  In this paragraph describe Finkel's demonstration that E. coli can use DNA as food, dependent on the CRP-S genes.

Vibrionaceae:  cholerae, parahemolyticus, fischeri, vulnificus, others?  Marine Vibrios JH Paul old papers.  Regulated by Sxy (cholerae, fischeri, others?), which is regulated by chitin, which Vibrios break down as major nutrient source (minimal inducer for V. cholerae is dimer of GlcNac - the chitin subunit).  V. cholerae has two Sxy homologs - one is required for competence, and overexpression of it bypasses the need for chitin.  No function has been assigned to the other Sxy homolog, which is regulated by a cyclic-di-GMP-sensitive 'GEMM' riboswitch (Kulshina 2009) (Sudarsan 2008).  Expression of the competence-regulating sxy gene is subject to  translational regulation by dimers of GlcNac (Yamamoto ref).  Yamamoto et al. overlooked a very strong CRP-N site 50 nt upstream of the 35 promoter element; this site suggests that V. cholerae competence may be regulated by cAMP and CRP as in H. influenzae.  Consistent with this, Meibom et al. found that addition of glucose prevents competence induction by chitin; induction by cAMP has not been reported.  As in H. influenzae and E. coli, the activity of V. cholerae CRP is controlled by its phosphotransferase system (Karatan & Watnick 2009 http://mmbr.asm.org/cgi/content/full/73/2/310)

Read Meibom 2005 paper - they did microarrays!

The quorum-sensing regulator HapR gene also controls V. cholerae competence (Meibom et al), and cultures become more competent as density increased (but this is a weird and not very compelling experiment).  Two sequenced V. cholerae strains that  Meibom et al. found unable to develop competence are known to carry frameshift mutations in hapR, (relevance to variation in natural populations?).

Pseudomonadaceae:  Moraxella catarrhalis (Stutzmann et al. FEMS Micb. Letters 2006).  Transformation on plates for 5 hr, frequencies v. high (>10% with chromosomal insertion!).  No transformation when cells were in stationary phase before plating.

Acinetobacter:   (Averhoff and Graf 2008)  (I can't find any other refs about regulation.)  Wackernagel papers (see email)?  Palmen 1994.

Pseudomonas:  Have competence genes (Cameron 2006) but not sxy homolog. Competence regulation has been studied in P. stutzeri (Lalucat et al.  good review MMBR 2006).  Induced by transition to stationary phase and by nutritional downshifts.  Nothing known about regulatory genes?  Transformable strains common, but 10% of strains are nontransformable.  I can't find anything about natural transformation methods in P. aeruginosa or any other pseudomonas species.  Bob Hancock says there aren't any for P. aeruginosa.  Paul Rainey also says they aren't, except stutzeri (Spiers et al. Microbiology 2000).  But see Carlson et al. 1982 (JB 153:93-99)  which says also several close relatives of stutzeri, all best at transition to stationary phase.  See email from Wackernagel.

Azotobacter:  (Ref Page Can J. Micro 1983).  Competence is induced by a nutritional downshift into an iron-limited nitrogen-free medium. cAMP also induces.  TFs are high, 10-3 - 10-2. No recent work.

Xanthomonadaceae:  Xanthomonas:  1957 paper (Corey and Starr, J. Bact?) demonstrating efficient transformation of X. phaseoli in broth by a chromosomal StrR marker .  TF ~10^-2.  Cited in 1985 as 'early reports claiming transformation...'), but no confirmations.

Legionellaceae: Legionella pneumophila (Sexton and Vogel J. Bact 2004) two genes that repress competence!  TF increases with cell density, max ~ 10^-2 at OD ~ 1.0, then falls (with a plasmid-borne homologous marker). Knocking out its only exoRNase induces competence! (nutrient nucleotides???Charpentier et al. JB 2008)


Nov. 24 additions:  Yesterday I went over the Johnsborg review with the Wu et al. phylogenetic tree in hand.  Cardiobacterium (Cardiobacteriaceae) is the only other Gammaproteobacteria reported to be competent.  This is based on a single report from 1985 (Tonjum et al, behind a paywall), with no regulation information available.  I'm emailing her to get a copy.  Also Dichelobacter?

Part I of the molecular biology review

Part I:

1. Explain what this review aims to do and why it is needed:  In bacteria, competence is the ability to take up DNA.  Here we're concerned with 'natural competence' which results from expression of genetically encoded machinery for DNA uptake, not with the DNA entry brought about by artificial (mechanical?  physical?  non-genetic?) permeabilization methods such as electroporation and treatment with divalent cations.

Mechanisms of DNA uptake are quite conserved (or convergent) but their regulation is complex and very variable.  The regulation of competence hasn't been reviewed for about 15 years (Solomon and Grossman 1996, but check for more recent reviews).  Much has changed, from microarray studies and followups.  Studies of regulation have generally been interpreted in a genetic-consequence framework. We will try to take a broader view, emphasizing the importance of understanding regulation for the ongoing problem of why bacteria take up DNA.

In this review we will examine the diversity of competence regulation, looking for unifying features, particularly those related to the various benefits DNA uptake can bring.  We will also consider how the regulation observed in lab cultures is likely to affect expression of competence in the natural environment.

2.  Overview of competence and transformation:  Transformation refers to genetic changes that result from recombination of this DNA with the chromosome.  This is how competence (and the role of DNA) was discovered, and remains the easiest way to detect DNA uptake.  Whether or not transformation results from DNA uptake depends on whether the incoming DNA carries homologous sequences similar enough for recombination, whether it carries different alleles, the extent to which it is degraded during uptake and in the cytoplasm, and the activity of the cellular DNA replication and repair proteins that carry out recombination.

Competence is widespread in bacteria (but not, apparently, in Archaea?) but its distribution is locally sporadic.  Detecting competence is most easily done by assays of transformation, and transformable bacteria have been found in many families of both gram positive and gram negative bacteria,  However, within these families, many species are reported to be not transformable and, when different isolates of 'transformable' species are investigated, both transformable and nontransformable isolates are typically found.  This suggests (1) that competence may be much more common than suggested by the single-isolate surveys, and (2) that competence is very frequently lost from bacterial lineages.  The ubiquity of species with at least some competent members, and with near-complete sets of competence genes, suggests that competence is generally adaptive in the long term, but whether competence is regained by lineages that have lost it, or whether these lineages usually die out, remains an open question. 

The mechanisms of DNA uptake used by different bacteria are very similar, with all bacteria transporting a single strand of DNA into the cytoplasm with the same inner-membrane channel.  Except for Helicobacter and Campylobacter (use family name?), (use ???), all bacteria also use force-generating proteins of the type 4 pili/type II secretion system complex to pull double-stranded DNA to the cytoplasmic/inner membrane.

However the regulation of competence is much more diverse.  Different bacteria reported to regulate competence in many different ways (here list some of the extremes).  In some cases the regulation appears to be competence-specific, but in others the 'competence' regulons include not only genes that act after DNA has been taken up (affecting its degradation and recombination, but many other genes whose functions appear unrelated to DNA uptake.  Another complication is that many bacteria not known to be naturally competent have homologs of genes that are competence-regulated in other bacteria.

The function of natural competence is controversial, which is one reason for paying close attention to its regulation.  Transformation is its most widely known consequence, but DNA from the environment also provides cells with deoxynucleotides that can be recycled for DNA replication or as sources of other nutrients (C, N, P) and with DNA strands that can be used as templates for DNA repair.  The genes responsible for the regulation of competence may have evolved to optimize the benefits of DNA uptake (depending on the extent to which the regulation is specific to DNA uptake).

3. Background for the survey:  How have the components of competence regulons been identified?  First, genes identified as having roles in transformation (mutant studies), and by specific investigation of regulation.  Second, by genome-wide analysis of the genes activated under conditions that cause competence development, both physiological conditions and expression of/dependence on already identified regulators of competence.

What kinds of genes are competence-regulated (under control of competence regulators, bearing in mind that these regulators may not be competence-specific)?  We can distinguish between genes that contribute directly to DNA uptake (call these 'competence genes'?), genes affecting what happens to DNA in the cytoplasm (degradation, protection, recombination), and genes with no apparent connection to DNA uptake or transformation.  Some of these latter genes have established functions unrelated to DNA uptake, and others have no known function.  Some of the genes in all of these categories are also common or ubiquitous in bacteria not known to take up DNA.