Taking it in: Bacterial Endocytosis

In my high school text books, bacteria were primarily defined in terms of what they were not. “Bacteria don’t have a nucleus”, “bacteria don’t have mitochondria”, “bacteria are not capable of complex membrane trafficking” and so on. But such boundaries seem to blur as more and more “eukaryote specific” properties pop up in some corners of the prokaryotic world, with bacterial endocytosis being the latest discovery. The simple finding that some bacteria can ‘eat’ the same way as your and my cells do, could have huge implications for our understanding of the evolution of eukaryotes.

To understand what is so special about endocytosis in bacteria, we’ll first have to take a look at what bacteria normally do and don’t do. If a bacterium wants to take up particles from its surroundings, it does so in non-specific ways.  Small molecules or peptides can be taken up passively via channels in the membrane, or actively via importing ‘pumps’. Normal proteins are far too big to be taken up in this way, so the only way bacteria can gobble up a protein is if it’s completely smashed to bits (by proteases for example).

Eukaryotes have more sophisticated importing mechanisms, allowing them to import molecules as large as entire proteins. When eukaryotic receptors sense a protein the cell wants to import, Part of its cell wall invaginates as membrane coat proteins line the newly formed pit. The invagination gets deeper and deeper until the membrane closes and an internal vessicle is formed.  This vesicle now contains part of the extracellular fluid and any (macro)molecules that where floating around at the time. The vesicle can now be internalized for further processing in the endosomes. The entire process is known as endocytosis, which you can see beautifully animated using electron microscropy pictures in the video below.

A while ago I wrote a guest post on Lab Rat’s blog on the discovery of typical eukaryotic membrane coat proteins in some bacteria. This finding of compartments and membrane coat proteins in the bacterial branch of Planctomycetes was suspicious of course: do these bacteria actually use all this machinery for endocytosis? A month ago, the exciting answer was published in PNAS with one of the most simple and elegant experiments I have seen in a long time. To show that the Planctomycete Gemmata obscuriglobus (literally ‘strange ball’) is taking up proteins via endocytosis, Lonhienne  and colleagues incubated the bacteria with green fluorescent protein. Within 5 minutes of incubation the cells were glowing like christmas trees, proving that the bacteria had taken up the complete protein! A normal bacterium would cleave the GFP to bits and pieces, and import the remaining peptides into the cell. There’s no way the GFP would still be fluorescent if Gemmata had taken up the GFP in this way. In other words, Gemmata carefully took in the light bulbs it found on its doorstep, whereas other bacteria would first smash them before dragging in the pieces.

When incubated together with fluorescent proteins (GFP), Gemmata obsuriglobus happily gobbled it up and started glowing!

The next step was looking where the fluorescent proteins ended up. In eukaryotes, a specific organelle called the endosome is the first stop for endocytosed proteins. From there it is decided whether the protein will be degraded, returned to the membrane or passed on for further processing in the Golgi apparatus. The team labeled the GFP proteins in the cell with little gold particles, so that  they would show up as distinct black spots on electron microscopy pictures. As you can see in the picture below, most of the proteins are localized in a special compartment: the paryphoplasm, which seems to be an expansion of the bacterial periplasm. The authors discovered that proteins got degraded in this paryphoplasm. Moreover, the team found invaginations and vesicle like structures within the cell that are normally associated with endocytocis. That certainly sounds like complex membrane trafficking to me!

Most GFP proteins (little black dots) localize to the paryphoplasm (marked with "P").

So where does that leave us? Endocytosis, comparmentalization and membrane sorting are no longer exclusive to eukaryotes. This could mean that the entire endocytosis machinery evolved before the latest eukaryotic ancestor did. If this is what happened, we would expect to find the system conserved and retained in a few bacterial phyla, such as the planctomycetes. Bacteria in this phylum have other ‘strange’ and eukaryote-like properties (for a bacterium at least): they reproduce via budding, surround their genetic material with membranes, synthesize sterols and don’t have peptiodglycan in their cell walls. The origin of eukaryotes may well lie with an ancestor that has many of the properties planctomycetes have today. If it’s true that planctomycetes are the prokaryotes most closely related to eukaryotes, Carl Woese’s famous tree of 1991 may need to be redrawn. That would be about time too, because that thing is getting close to ancient for a field that is moving so quickly as evolutionary biology! It’s interesting to note that this scenario of eukaryotic evolution is also in direct conflict with the hypothesis that eukaryotes arose after the endosymbiosis of an archaeon by a bacterium, so expect to see some fireworks as experts debate how to integrate these findings into our current understanding of evolution..

I’m not an expert on eukaryotic evolution, but I can tell you that the discovery of a bacterium with so many distinct ‘eukaryotic’ features will impact on our ideas and views of the history of life on this planet. Before the dust settles, the song “Take it in” by Hot Chip strikes me as wonderfully appropriate:

Guest post: Don’t judge a cell by its coat
Bacterial Compasses
Evolving Molecular Machines