11/06/2012 11:18 am ET | Updated Jan 06, 2013

Inter-Kingdom Horizontal DNA Transfer in All Directions: Infectious Bacteria Evolve by Acquiring Protein Domains From Eukaryotic Hosts

Earlier blogs have covered horizontal DNA transfer between bacteria and from bacteria and fungi to animals. This blog will present some of the evidence for horizontal transfer in the reverse direction: DNA acquisition by infectious bacteria from eukaryotic hosts.

Bacteria pick up eukaryotic sequences encoding different characters, such as nutrition, but the most notable use of "higher" organism DNA is to infect and manipulate eukaryotic cells in the service of bacterial survival and multiplication.

Bacteria take control of eukaryotic cells by injecting regulatory proteins into their hosts. The molecular injection systems are related to similar structures that bacteria use to acquire DNA from the environment.

I became aware that DNA transfer can be a two-way process a couple of months ago when my University of Chicago colleague Howard Shuman told me about some of his work on Legionella pneumophila, the bacterium that unexpectedly caused Legionnaire's Disease in 1976 at an American Legion convention in Philadelphia.

The Philadelphia infection was traced back to bacterial populations in the air-conditioning system at the hotel where the convention was held. Legionella has a complex lifestyle. In water supplies, Legionella can live either as multicellular biofilms on inorganic surfaces or as intracellular endosymbionts or parasites of amoeba and other eukaryotic microbes.

Inhaling micro-droplets of water containing Legionella can lead to pneumonia, where the bacteria encase themselves in intracellular compartments called vacuoles in lung cells. It is not yet clear whether free-living or intracellular Legionella is the infectious agent.

Some of the molecules Legionella injects into a microbial or mammalian eukaryotic host cell are called "effector proteins." Effector proteins alter control of host cell biology by changing control and signaling pathways and thus restructuring the cytoskeleton that organizes the interior of the cell. Clearly, the injected proteins have to be able to interact in very specific ways with molecules in the host cell's control circuitry. Where do they get this specificity?

The answer is that Legionella takes the specific information from its hosts. Legionella effector proteins contain functional "domains" (segments) that are not found among other bacteria but rather among eukaryotes. For example, one injected effector protein is a regulatory enzyme called a protein kinase; it has a key domain structure that clearly relates to domains of similar molecules in protists, plants and animals. The organization and evolution of proteins as systems of domains is covered in an earlier blog.

The presence of eukaryotic domains in these effector proteins indicates that Legionella acquired them from eukaryotic hosts, most probably their amoebal and other microbial hosts. The external origin of the DNA encoding these proteins is confirmed because it has a lower GC content than the clearly bacterial DNA in the Legionella genome.

It should not be surprising that Legionella has the ability to acquire external DNA. We have known for many years that it has an active DNA uptake system of the kind related to the effector injection apparatus. Intriguingly, this DNA uptake system is SOS-inducible, just like the DNA repair and mutagenesis functions discussed in the previous blog. Intercellular DNA transfer may play a role in damage repair.

Of even greater interest is the discovery that Legionella can become "hypercompetent" for taking up DNA when grown at 30 C, a temperature more similar to that experienced when growing in amoeba rather than in mammalian host cells at 37 C. Since the same kind of multimolecular system serves to acquire DNA and also transfer effector proteins into eukaryotic hosts, it is tempting to hypothesize that infection and horizontal DNA transfer occur at the same time.

In 2008, my colleague Howard Shuman and his student Karim de Felipe looked at the sequenced bacterial genomes to find out how common it was for them to have eukaryotic protein domains. Although they never published their results, the data are quite interesting. Eukaryotic domains can be found in some bacteria that are only known to be free living, but infectious bacteria that are either pathogens or symbionts are far more likely to have them. Only infectious bacteria have the largest repertoires of eukaryotic domains (> 4% of all proteins). These include bacteria and "mycoplasma" (wall-less bacteria) that infect plants as well as animals and eukaryotic microbes. The mycoplasma plant pathogen Phtyoplasma asteris is the champ with >12% of its proteins related to eukaryotes.

Other investigators have noticed that some of effector protein eukaryotic domains can be found in the large DNA viruses that infect amoeba. These viruses are remarkable for containing sequences from all three kingdoms of life: bacteria, archaea and eukarya. This discovery has led some molecular evolutionists to propose that amoeba and their DNA viruses compose a "melting pot" where different sequences mix for evolutionary innovation.

As a microbiologist, I like to remind readers and audiences that bacteria are the most successful cell biologists on the planet. In asking how they acquired the necessary information, we see that interkingdom DNA transfer and genomic incorporation of useful coding sequences have been integral aspects of infectious bacteria developing that cell biological expertise.