Technical Highlight - December 2014
Short description: Analyses of nine topology prediction programs identify those with the greatest prediction accuracy for specific transmembrane transport protein families, and lead the authors to propose several distinguishable evolutionary pathways for these proteins.
Proposed pathway for evolution of channel proteins and small-molecule transporters. Figure reproduced with permission from ref. 1, Copyright © 2014, Karger Publishers, Basel, Switzerland.
Many topological prediction programs have been developed for analysis of the structurally diverse integral membrane channel and carrier proteins found universally in all living organisms. The simplest of these proteins contain just one or two transmembrane segments and form non-specific oligomeric channels, whereas others span the membrane many times and function without oligomerization. The former are believed to be the evolutionary precursors of the latter. Saier and colleagues (PSI TransportPDB) compared the accuracy of nine topology prediction programs on various families of membrane transport proteins.
In an initial analysis of four families, HMMTOP, SPOCTOPUS and MEMSAT yielded best predictions, with the other six programs showing inferior performance. The authors combined these three algorithms into a new program, TMStats, and used this novel program to analyze a broad range of integral membrane transport protein types. They detailed their results according to protein family for each of the channel and carrier types analyzed.
Even the best programs underperformed in analyses of several protein families, and the causes of these errors were often family-specific. In some cases, systematic mistakes were identified, some but not others of which occurred using multiple programs. Predictions for families containing proteins that can exist in both membrane-integrated and soluble forms were especially challenging, particularly since structures of the membrane-integrated forms were often not available. The authors proposed introducing family-specific prediction algorithms, all combined into a single program with the capacity to determine which settings to use for a given query family.
Based on these analyses, the authors put forth evolutionary pathways for the appearance of different types of channels and secondary carriers. As noted above, they suggested that these evolved multiple times, starting from small oligomeric channel-forming peptides, via repeated intragenic duplication events. This was presumed to be followed by subsequent acquisition of mutations to confer substrate specificity and transport-related conformational transitions that converted channels into carriers.
A. Reddy et al. Reliability of nine programs of topological predictions and their application to integral membrane channel and carrier proteins.
J. Mol. Microbiol. Biotechnol. 24, 161-190 (2014). doi:10.1159/000363506