Tuesday, April 07, 2009

On the Evolution of Bacterial Chromosomes


This is a story about two cultures; the old biologists who grew up with the 'phage group and bacterial genetics, and the younger biologists who didn't.

It's also a story about science journalism and the reporting of science in the 21st century.

We've known about plasmids in bacteria for a very long time. Plasmids are small circular DNA molecules that carry a number of genes, such as those for antibiotic resistance, or sex. Some of them are present in multiple copies while others are present in only a single copy. In the case of single-copy plasmids, their replication is coupled to that of the chromosome and the daughter plasmids segregate to the daughter cells just like the newly replicated chromosomes do.

Genes can hop from chromosomes to plasmids and back again. This phenomenon was discovered in the 1950's by Jacob and Adelberg (1959). Several well-known plasmids carrying certain chromosomal genes were studied, including a famous one known as F-lac—an F plasmid containing the lac operon.

By the time the first E. coli Bible was published in 1987, there were dozens and dozens of examples of gene transfer between chromosomes and plasmids (Holloway and Low, 1987).

During the 1970s and 80s, the DNA contents of many difference species of bacteria were published. It soon became apparent that certain classes of bacteria (e.g. Rhizobiaceae) contained large plasmids called megaplasmids. Sometimes it was difficult to tell the difference between a plasmid and a chromosome (bacterial chromosomes are usually circular).

As a general rule, plasmids were dispensable. The bacteria could be "cured" of a plasmid and still survive. When the plasmid acquired essential genes, as they did from time to time, they became chromosomes. Some species of bacteria had two or more chromosomes. It was part of general knowledge that plasmids could evolve into chromosomes as described in a 1998 review by Moreno.
Animal intracellular Proteobacteria of the alpha subclass without plasmids and containing one or more chromosomes are phylogenetically entwined with opportunistic, plant-associated, chemoautotrophic and photosynthetic alpha Proteobacteria possessing one or more chromosomes and plasmids. Local variations in open environments, such as soil, water, manure, gut systems and the external surfaces of plants and animals, may have selected alpha Proteobacteria with extensive metabolic alternatives, broad genetic diversity, and more flexible and larger genomes with ability for horizontal gene flux. On the contrary, the constant and isolated animal cellular milieu selected heterotrophic alpha Proteobacteria with smaller genomes without plasmids and reduced genetic diversity as compared to their plant-associated and phototrophic relatives. The characteristics and genome sizes in the extant species suggest that a second chromosome could have evolved from megaplasmids which acquired housekeeping genes. Consequently, the genomes of the animal cell-associated Proteobacteria evolved through reductions of the larger genomes of chemoautotrophic ancestors and became rich in adenosine and thymidine, as compared to the genomes of their ancestors. Genome organisation and phylogenetic ancestor-descendent relationships between extant bacteria of closely related genera and within the same monophyletic genus and species suggest that some strains have undergone transition from two chromosomes to a single replicon. It is proposed that as long as the essential information is correctly expressed, the presence of one or more chromosomes within the same genus or species is the result of contingency. Genetic drift in clonal bacteria, such as animal cell-associated alpha Proteobacteria, would depend almost exclusively on mutation and internal genetic rearrangement processes. Alternatively, genomic variations in reticulate bacteria, such as many intestinal and plant cell-associated Proteobacteria, will depend not only on these processes, but also on their genetic interactions with other bacterial strains.
Given this context, I was interested in a recent press release: Evolutionary origin of bacterial chromosomes revealed. "Hmmm," I thought., "I wonder what new mechanism has been discovered?"

Imagine my surprise to read ...
Most bacteria have only one chromosome. The Rhizobiaceae is an unusual bacterial family in that all of its members have either two chromosomes or one chromosome and very large plasmids. Until this study, it was not clear how such multichromosomal architectures had evolved.

João Setubal, associate professor at the Virginia Bioinformatics Institute and the Department of Computer Science at Virginia Tech, commented: "Thanks to the efforts of the Agrobacterium Genome Sequence Consortium and the wider research community, we have sufficient sequence data available from different bacterial species to allow the inference of a general model for bacterial genome evolution. It appears that the transfer of genes from chromosomes to large plasmids mediates second chromosome formation."
That's not new. The idea that large megaplasmids in Rhizobiaceae could become plasmids by acquiring essential genes has been around for three decades, at least. Surely these workers known their history? The press release must be an exaggeration of what's in the paper.

So I looked up the paper (Slater et al., 2009). These workers sequenced the genomes of a number of related bacterial species containing chromosomes and plasmids. They announce the "surprising" discovery that genes can transfer between chromosomes and plasmids.
While it has long been known that gene transfer can occur between organisms, the picture that emerges from our study shows a group characterized by composite genomes in which genes of all classes are not only migrating between organisms, but also intracellularly among chromosomal and plasmid replicons.
It sounds like they never heard of F-lac or any of the other F′ or R′ plasmids. It sounds like they are completely unaware to the fact that transfer of genes from chromosomes to plasmids is an old established fact.

The authors propose a "general model for bacterial genome evolution" in which plasmids evolve into chromosomes.

This is not an isolated phenomenon. There seem to be lots of cases where today's scientists are unaware of the history of their field. A consequence of this ignorance is that the wheel is being constantly reinvented, with all the associated hype of a modern breakthrough.

Another example is the recent "discovery" of regulatory RNAs. Bacterial and 'phage examples have been known for forty years.

Why is this happening? Why do reviewers let it pass?

[Image Credit: Jessica Snyder Sachs]

Holloway, B. and Low, K.B. (1987) F-Prime and R-Prome Factors. in Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. F.C. Neidhardt ed. vol.2.

Jacob, F. and Adelberg, E.A. (1959) Transfer of Genetic Characters by Incorporation in the Sex Factor of Escherichia coli. Comptes Rendus 249:189-191.

Moreno, E. (1998) Genome evolution within the alpha Proteobacteria: why do some bacteria not possess plasmids and others exhibit more than one different chromosome? FEMS Microbiol 22(4):255-275. [PubMed]

Slater, S.C., Goldman, B.S., Goodner, B., Setubal, J.C., Farrand, S.K., Nester, E.W., Burr, T.J., Banta, L., Dickerman, A.W., Paulsen, I., Otten, L., Suen, G., Welch, R., Almeida, N.F., Arnold, F., Burton, O.T., Du, Z., Ewing, A., Godsy, E., Heisel, S., Houmiel, K.L., Jhaveri, J., Lu, J., Miller, N.M., Norton, S., Chen, Q., Phoolcharoen, W., Ohlin, V., Ondrusek, D., Pride, N., Stricklin, S.L., Sun, J., Wheeler, C,, Wilson, L,, Zhu, H., and Wood, D.W. (2009) Genome Sequences of Three Agrobacterium Biovars Help Elucidate the Evolution of Multi-Chromosome Genomes in Bacteria. J. Bacteriol. 2009 Feb 27. [Epub ahead of print] [PubMed] [DOI: 10.1128/JB.01779-08]


  1. Ha! More evidence for Epicurean Evolution!

  2. Recombination between different DNA molecules in the same cell? I'm shocked!

    Honestly, how can you work in the field of bacterial genetics and molecular biology and not know about the early work on F' factors?

  3. Probably because when they were starving students they couldn't afford the outrageously expensive required undergraduate textbook. :)

  4. Then how did they pass the tests to earn their outrageously expensive BS in biology?

  5. ...or pass their qualifying exams?

  6. This is definitely an odd one Larry, you would have thought at least one reviewer would have been familiar enough with the bacterial genomics literature to point that out, especially if key papers had never been cited.

    Its also odd that they cite lateral gene transfer between organisms as being "known for some time" (which to eb fair it has) even though transfer between genomes within cells is far more common and has been known about for much longer. I'm sure that early work on the evolution of bacterial chromosomes shaped much of the research into endosymbiotic gene transfer and the evolution of organellar genomes for instance.

  7. Why? Textbooks abbreviate, and nobody writes synthesizing survey books. (Larry ... ?)

  8. Being a relatively young researcher myself, I'd say that the main reason is simply too much literature that needs to be read. Reading literature is usually the first thing to be moved to the end of the to-do list. We all know that we should read and think more about our research.
    If / when literature is researched and read, it tends to be the recently published works. Huge amounts of original scientific papers are published, and they are getting more from year to year. It is extremely difficult to keep up with the recent work - there is no time to actually read stuff that is decades old.

    So, yes: knowledge is lost and will be rediscovered again and again. This is surely not the first example and that process is not limited to science. On occasion, a senior colleague will remember the original works, but often none may be around.

    But the positive thing about it: the scientific finding is independently reproduced, often even using completely different approaches. This is of big value nowadays - usually reproducing something is considered a waste of time.

    PS: However, a review from 1998 is not really that old.

  9. Being a microbiology teacher, it really puzzles me how someone working in bacterial genetics cannot mention conjugation of F plasmids and F' formation as the obvious example of gene transfer between chromosome and plasmids.

    Perhaps the problem is with modern "wholesale biology". People working on genome sequencing, microarrays etc. forget the small and simple facts of their own field. And they also tend to forget that these powerful techniques are just tools and not the end itself.
    They drown in the huge amount of data trying to make some sense out of it.

  10. Sven says,

    Being a relatively young researcher myself, I'd say that the main reason is simply too much literature that needs to be read.

    No, that's not part of the problem. You don't learn the fundamental concepts of a field by reading all the papers that have ever been published. You learn them from colleagues, courses, and textbooks.

    One of the mistakes we're making these days is teaching undergraduates that they can jump right into the primary literature and understand everything they need to know. In my own department, for example, we are watering down the teaching of fundamental concepts in our introductory biochemistry course in order to tell students about some hot new research in membrane biology.

    We shouldn't be surprised when those students become researchers who know nothing about the evolution of metabolic pathways or how flux is controlled in a pathway.

    My colleagues think it's a bad idea to require a biochemistry textbook in a course because everything in the textbook is old fashioned and all that conceptual "detail" just confuses the students.

  11. "Honestly, how can you work in the field of bacterial genetics and molecular biology and not know about the early work on F' factors?"

    Especially people who work on Agrobacterium. T-DNA transfer into plants works the same way as does F factor transfer between bacteria.

    In the authors' defense, I am sort of guessing that the evolutionary processes they are discussing assume transfer of genes between genomic elements, and that they are not claiming to have discovered (or rediscovered) recombination. I also don't know that the wide-spread nature of gene movement is something that one would get from an understanding of F'-lac (a pretty contrived element, to be sure) or R plasmids (more commonly associated with transposable elements than with your run-or-the-mill gene).

    The great thing about teaching plant molecular biology is that I have a built-in reason to revisit the very old bacterial genetics literature, to show how that field drove the discovery process when it comes to Agrobacterium.

  12. Interdiciplinary studies. You learn one field (like math), then, afterwards, apply it in a different field that interests you, like biology. But, since you did not spend much of education in immersion in your current field of endeavour, you may not appreciate the subtle difficulties in developing good research topics.

    "João Setubal, associate professor at the Virginia Bioinformatics Institute and the Department of Computer Science at Virginia Tech, commented ..."

    Perhaps he came to biology from computer science.

    ~~> http://staff.vbi.vt.edu/setubal/overview.html

    " Ph.D., Computer Science, University of Washington, 1992 "

  13. Folks - that's my paper, and I'm afraid that you have completely missed the point of the manuscript (it appears that no one actually read the paper). OF COURSE we are aware of gene transfer within bacterial replicons. The manuscript ASSUMES that the reader understands this basic point of bacterial genetics (perhaps, we were too generous). The manuscript describes HOW a particular sequence of gene transfers led to evolution of a SECOND CHROMOSOME in the Rhizobiaceae, including a delineation of the LINEAGE of second chromosomes IN THIS CLADE (see Figures 4 and 5). The reason that no reviewer made the naive comments that you folks made is that they ACTUALLY READ THE MANUSCRIPT and understood both the context and the results.
    Perhaps next time you will claim that a paper describing a new E. coli strain actually reports the discovery of P1 transduction!
    Please, people, read the literature before you comment. This sort of intellectual laziness reflects poorly on you.

  14. Steve Slater says,

    Folks - that's my paper, and I'm afraid that you have completely missed the point of the manuscript (it appears that no one actually read the paper).

    Welcome to my blog.

    Now that you're here, perhaps you could answer a few questions?

    In the introduction to your paper, which I have read, you say ...

    While the sequencing of S4 and K84 was motivated by the need to have full genomic sequences for at least one biovar II representative and at least one biovar III representative, we have found that their genomes, as well as those of C58 and other Rhizobiales species, enabled us to infer a general model for bacterial genome evolution. Crucial for this inference is the complex (for bacteria) replicon architecture of all three Agrobacterium genomes. The data provided here and additional evidence (40, 46) support our model as a generalized mechanism of genome evolution among bacteria that harbor multiple chromosomes.

    I'm curious about this "general model" of bacterial genome evolution. Could you summarize in a few sentences how this model differs from previous models of chromosome evolution?

    While you're here. What do you think of the press release and the quotation from your colleague, João Setubal? Is that an accurate reflection of your position?

    It sure sounds like someone who has just discovered that genes can be transferred from chromosome to large plasmids.