Content area
Full Text
EVOLUTIONARY BIOLOGY
Molecular machines containing related protein subunits are common in cells. Reconstruction of ancient proteins suggests that this type of complexity can evolve in the absence of any initial selective advantage. See Letter p.360
Organisms and cells are bewilderingly complicated, and the molecular machines that perform many basic cellular functions are often giant, multi- subunit, multifunctional protein complexes with tangled evolutionary histories. It is generally assumed that such complexes arose by the stepwise accretion of individual proteins, each addition representing a selective advantage by adding to or refining the machine's performance. But on page 360 of this issue, Thornton and colleagues1 argue against this standard explanation in one particular instance - that of a ring-shaped protein complex in fungi. The authors show how evolutionary processes entailing loss of function rather than gain might initially drive a system towards complexity, independently of selection*.
Many cellular molecular machines contain several proteins that self-assemble into a multi-subunit ring. In simple cases, rings are homo-oligomeric; that is, all the subunits are identical and thus probably the products of a single gene. In more complex (hetero- oligomeric) examples, the protein molecules in the ring are different, but may be related. Often, hetero-oligomeric rings seem to have arisen from homo-oligomeric complexes after a gene encoding a single subunit became duplicated, producing two or more gene copies (called paralogues), with each copy subsequently evolving to encode a slightly different protein.
The question that Thornton and colleagues address is, why does this happen? Do the structurally distinguishable subunits and/or their specific pattern of assembly confer improved or additional functions on the protein complex, with selection for their enhanced performance being the evolutionary driving force? Alternatively, might neutral processes be responsible, at least initially?
To understand how the latter possibility might come about, imagine that a gene encoding, for example, eight subunits of a homo- oligomeric ring undergoes duplication (Fig. 1). The two paralogous gene copies that this duplication produces - and the protein subunits they...