Bioinformatic analyses of transmembrane transport: novel software for deducing protein phylogeny, topology, and evolution

Abstract

During the past decade, we have experienced a revolution in the biological sciences resulting from the flux of information generated by genome-sequencing efforts. Our understanding of living organisms, the metabolic processes they catalyze, the genetic systems encoding cellular protein and stable RNA constituents, and the pathological conditions caused by some of these organisms has greatly benefited from the availability of complete genomic sequences and the establishment of comprehensive databases. Many research institutes around the world are now devoting their efforts largely to genome sequencing, data collection and data analysis. In this review, we summarize tools that are in routine use in our laboratory for characterizing transmembrane transport systems. Applications of these tools to specific transporter families are presented. Many of the computational approaches described should be applicable to virtually all classes of proteins and RNA molecules.

Fig. 1.

Phylogenetic tree of 309 prokaryotic members of the TSUP family. The tree is based on a CLUSTAL X (neighbor joining) multiple alignment (online suppl. fig. S1).

Fig. 2.

Average hydropathy (top, solid line) and similarity (bottom, dotted line) plots for the TSUP family of transporters. A modified AveHAS program was used to derive the plots as described in this review article.

Fig. 3.

Alignment of the first half of one TSUP family member (Tcr1) with the second half of another (Ama2). All family members are homologous to each other throughout their lengths. The modified IC and GAP programs were used with 500 random shuffles, a gap penalty of 8 and a gap extension penalty of 2. The aligned sequences gave a comparison score of 19 SDs.

Fig. 4.

Phylogenetic (Fitch) trees for the APC superfamily [Jack et al., 2000]. The trees were generated with (A) the SFT1 program and (B) the SFT2 program. A The tree presents the relationships of all proteins of the APC superfamily as of June, 2008. These include all 16 families of the APC superfamily as indicated by their three letter abbreviations as defined in TCDB. Numbers refer to the individual TC numbers of the proteins within the various families. Bootstrap values are provided adjacent to each branch. The same convention is also used for figures 4B, 5D and 6B below. B A FITCH tree of the entire APC superfamily generated with the SFT2 program. The tree reveals the phylogenetic relationships of the 16 families relative to each other.

Fig. 5.

Phylogenetic trees generated with four programs for the AAAP family [Young et al., 1999] within the APC superfamily. A CLUSTAL X-TREEVIEW-neighbor joining tree. B Parsimony (Protpars) tree based on the CLUSTAL X-generated multiple alignment. C The corresponding tree generated with the SFT1 program. D The tree generated with the SFT2 program. The SFT1-based tree (C) and the SFT2-based tree (D) show the eight subfamilies of the AAAP family. Note: this tree resembles the AAAP cluster for the APC superfamily tree shown in figure 4A. All members in TCDB included in the eight AAAP subfamilies (C) were averaged to provide the positions of each of these subfamilies relative to each other (D).