Bacteria will eat anything. Their highly diverse biochemistry, and ability to adapt quickly to change means that they can adapt to take up nutrients from a range of sources, including hot acid lakes and the interior of underground thermal vents. However bacteria also predate each other, and one particular bug, Micavibrio aeruginosavorus does this by latching onto a fellow bacteria and sucking the life out of it.
There are a range of different types of predatory bacteria, and I discussed a few of them in my old blog in the post 'Bacterial Hunting Strategies'. While that post examined bacteria that hunt down their prey like wolf-packs, and the little Bdellovibrio that curls up inside a bacteria and devours it from the inside, the Micavibrio was a new one for me. It differs from other bacterial predators as rather than getting inside the cell, or killing it and then eating the remains, it simply hangs onto the slowly dying cell and sucks the life out of it, dividing around it's prey until the prey finally dies. Like a very determined and slightly horrific leech.
The Micavibrio life cycle therefore consists of two main stages. First the attack stage, where the bacteria must seek out its prey. Micavibrio were originally thought to be very fussy about which particular bacteria they feed off, but some have been found that are less specific about their prey. The specificity is very exciting from an antimicrobial point of view, as anything that targets a dangerous bacteria while not affecting the non-dangerous bacteria that live within our bodies has therapeutic potential. Particularly interesting is that the Micavibrio are not put off by biofilms; the sticky layer that bacteria such as Pseudomonas aeruginosa tend to coat themselves in. Antibiotics have real difficulties with biofilms, but the Micavibrio can swim right through them.
Examining the genome of the Micavibrio yields exciting answers as to how it functions as a leeching parasite. Despite relying exclusively upon other bacteria for survival, it has an impressive collection of metabolic enzymes, with very little of the genomic loss seen in other parasitic bacteria. It is, however, missing several genes required to make amino-acids (the building blocks of proteins) which explains its dependance on bacterial-prey. While the Micavibrio seems more than happy to produce its own energy, in order to grow, reproduce, or make any new proteins, it needs a bacterial host. The Micavibrio is also incapable of taking up amino-acids from the environment. You can grow it in a media containing all the amino-acids it needs and it will still die because it can't transport them into the cell. It needs to be attached to another bacteria to survive. Quite how it gets the amino-acids out of its prey is as yet unknown, although cytoplasmic bridges and intracellular nanotubes have been proposed.
While this is fascinating from a microbiological perspective, the synthetic biologist inside me is already asking whether this specific bacteria-killing machine could be of any medical use. Using bacteria to kill bacteria is an interesting idea, although the Micavibrio would have to be fitted with a death gene to dispose of it once the disease bacteria were removed. The problem with a death gene is simply that it is highly evolutionarily disfavored. Any bacteria that manage to loose the death gene will be at a massive advantage and would quickly spread.
However while the living antibiotics might not be immediately applicable in medical scenarios, there are plenty of industrial situations that could benefit from the specificity and hunting ability of the Micavibrio. Biofilms form in pipes, and block up nozzles in equipment, and as using antibiotics tends to leave behind small parts of the biofilm (which has an increased likelihood of being resistant) using the search-and-destroy feature of Micavibrio could be really useful. It could also (with some genetic tweaking) play a part removing contaminating bacteria from fermenting solutions, without requiring the need for antibiotic treatments.
University of Virginia Press Release
Wang Z, Kadouri DE, & Wu M (2011). Genomic insights into an obligate epibiotic bacterial predator: Micavibrio aeruginosavorus ARL-13. BMC genomics, 12 PMID: 21936919