I previously introduced the reader to the enhanced complexity that is seen in eukaryotic cells and asked why it is that eukaryotes are like this. We then considered an experiment using GFP that helped us to appreciate one way in which the eukaryal and bacterial cell plans are different: yeast, but not bacteria, were able to efficiently fold an artificial multi-domain protein. Let us there pause to consider this in more detail.

Large, complex proteins are actually modular structures built from more than one simpler protein domain (fold). A domain is an “independently folding” structure that has a rather specific conformation and is often associated with particular functions. As a consequence, multi-domain proteins entail a division of labor, where different domains handle different jobs, all integrated into the overall function of the protein. For example, a complex protein could be composed of three domains, where one domain binds DNA, another domain binds and hydrolyzes ATP, and yet a third domain binds another protein. It is this modular design that allows this protein to link interactions with DNA and other proteins to energetic flow.

To make the artificial multi-domain protein mentioned above, we simply take the single domain GFP and, as one example, link it together with the single domain enolase. In other words, you open your toolbox of a thousand or so protein folds, pull out the GFP-like fold and connect it to the TIM-barrel fold (a common enzyme fold).

Or, for those who are more visual, you simply take this:

And use a short chain of amino acids to connect it to this:

Bottom line? Yeast, a simple eukaryote, can fold it while bacteria cannot. Why is that? Before answering that question, let’s consider whether this is an exception or a rule.

Researchers from the Stockholm Bioinformatics Center recently conducted an extensive survey of protein databases (Ekman, D, Bjorklund, AK, Frey-Skoyy, J, and Elofsson, A. 2005. Multi-domain proteins in the three kingdoms of life: orphan domains and other unassigned regions. JMB 348: 231-243). Their basic finding was that 65% of eukaryotic proteins have multiple domains compared to only 40% for prokaryotes. When the cut-off value for a domain is 100 amino acids, the following breakdown is obtained:

Eukaryotes: Single-domain (35%), Two-domain (20%), and Three or more domains (45%).

Prokaryotes: Single-domain (60%), Two-domain (20%), and Three of more domains (20%).

Previously, we noted that eukaryotic proteins are typically larger than prokaryotic proteins and now we can see why: eukaryotes are just better at making multi-domain proteins. Thus, it is not surprising that they could synthesize the artificial two-domain GFP protein when bacteria could not. But there is more.

There is a particular type of multi-domain protein known as the multi-domain repeat. In this case, a single type of domain is repeated multiple times within the same amino acid chain. For example, imagine a protein that had three or more GFP-like folds all linked together. Not only are eukaryotes better at making this type of protein, but it is also more commonly found in multicellular organisms, where approximately one out of every ten multicellular proteins is a multi-domain repeat. In fact, with this criterion alone, unicellular yeast are more like bacteria than humans.

But before chasing down that rabbit, let’s ponder why it is that eukaryotes seem to be better at making multi-domain proteins…..