Fig 1.
General structure of CPA/AT transporters.
(a) Example of broken helical transporter (Sodium bile acid symporter: PDB 4n7w) shown as N- and C-terminal repeat. Each repeat unit is composed of a scaffold and a core subdomain. (b) Example of reentrant helical transporter (Sodium citrate symporter: PDB 5a1s) shown as N- and C-terminal repeat. (c) In structure space, both the N- and C-terminal subdomains form the scaffold and core domains, respectively. The lipid bilayer is colored red and blue to denote outside and inside. The scaffold and core domain are colored brown and purple respectively. (d) The scaffold and core domain are colored brown and green respectively. The broken and reentrant helices are shown in darker shades compared to other helices of the core domain for easy visualization.
Table 1.
Annotation of topology and subdomains for families in the CPA/AT transporters.
The orientation (Nin, Nout) describes the location of the N-termini of the protein. Comparison of classification of families from this work, TCDB and Pfam are shown. The name in bold preceding the semicolon denotes superfamily, the name following the semicolon denotes the family. Families with unassigned superfamily start with a Semicolon. H, RH and BH indicate transmembrane helix, Reentrant helix and Broken helix respectively. When a homologous template was found with an E-value better than 1*10−3 it was shown. The Pcons scores for the trRosetta models are also listed.
Fig 2.
Comparison between PSE_1 (Reentrant-helix-reentrant fold-type) and 2HCT (CPA-reentrant fold-type).
The brown dots represent the localization of the membrane a) An example of an internal symmetry with two reentrant helices (2HCT:5a1s). b) PSE_1 model. c) Superposition between the models of PSE_1 (red) and the 2HCT structure (blue).
Fig 3.
Visualization of the relationship between the families by a network linking families with significant similarities in the MSA-MSA alignments (E-value < 0.1 thin edges, E-value <0.001 thick edges) for repeat units (right) and full-length families (left). Each family/repeat is colored according to the fold type, Magenta: BART-fold, Green: CPA-reentrant, Red: CPA-broken and Blue: reentrant-helix-reentrant fold type. In the repeat unit network, a) N-terminal units are circular and C-terminal units are star-shaped. In the full protein network. b) The full proteins are square.
Fig 4.
Transition and alignments between 2HCT and PSE_1.
a) A “Dotplot” between PSE_1 and 2HCT: The blue lines represent the TM helices and the red one the reentrant helices, the color spectra show the Log of E-value of the alternative alignments. b) The panel shows an all-vs-all distance matrix between the trRosetta models and the 2HCT PDB structure 5a1s. The greyscale indicates the distance between the mutual distances between the protein residues from 4 to 10 Å. b) Schematic representation of the topology of the protein with its corresponding structure below. The TM helices (in-out) and (out-in) are colored white and grey respectively. The reentrant helices after the transition are colored red. The helix 7 that changes position after the transition is colored green. The part of 2HCT putatively lost in the truncation is colored blue. c) Corresponding protein structures.
Fig 5.
Broken-Reentrant transporter transitions by duplication of repeat unit.
(a) Sequence and topology alignment between Na_H_Exchanger_1 (CPA-broken fold-type) and DUF819 (CPA-reentrant fold-type). All the transmembrane helices are numbered sequentially and the reentrant helix is denoted as RH (b) Structure superposition of broken and reentrant transporter. The extra scaffold helix and the broken-reentrant transition are highlighted in bright colors. (c) A zoomed-in figure of broken (pink) and reentrant helix (yellow and blue) is shown. The glycines are shown in stick representation. (d) The aligned positions of the broken and reentrant N-core helix are represented as sequence motifs (e) Cartoon representation showing the events of duplication and mutation in reentrant transporters leading to the transition of broken-reentrant transporters.
Fig 6.
Change in orientation in reentrant transporters by internal duplications.
(a) Sequence and topology alignment between transporters from CPA-reentrant fold-type showing the change in orientation. All the transmembrane helices are numbered sequentially and the reentrant helix is denoted as RH. (b) Sequence similarity between N- and C-terminal repeats represented by E-values in different families with fold-types containing symmetric repeat units are shown (c) Reentrant N- and C-terminal core helix motif (d) Cartoon representation showing the events of duplication from C-terminal repeat of reentrant transporter and subsequent internal duplication leading to change in orientation.
Fig 7.
Change in orientation in broken transporters by shuffling.
(a) Sequence and topology alignments, which shows the high similarity between N-terminal and C-terminal repeats of SBF_like and Mem_trans families and vice versa. All the transmembrane helices are numbered sequentially. (b) Sequence similarity, represented with E-values, between NR-NR, CR-CR and NR-CR repeats of two different families in the BART fold-type are shown. NR and CR refer to the N- and C-terminal repeats respectively. (c) Cartoon representation showing events of shuffling of repeats leading to change in orientation.
Fig 8.
Fold-types and their evolution in CPA/AT transporters.
The four fold-types of CPA/AT transporters are shown. Topologies, type of core helix (Broken/reentrant), the orientation of the N-terminal helix as well as other data are listed. All the evolutionary events responsible for the evolution of fold-types are summarized. The TM helices (in-out) and TM helices (out-in) are colored dark grey and white, respectively. Reentrant helices (in-in) and (out-out) are colored yellow and blue respectively.
Fig 9.
An example showing the annotation of the topology of DUF819 family using our integrated pipeline.
The main steps involved in the pipeline are shown. (a) Evolution guided topology annotation: Topology predictions are mapped onto the seed MSA to obtain a multiple topology alignment. The phylogenetic tree from seed MSA is used to reorder the multiple topology alignment into a “Reordered topology alignment”. This evolution guided topology prediction is used to infer initial topologies for a family. Generation of the final topology and annotation of core and scaffold subdomains is obtained by comparing to the known structure of sodium-citrate symporter (PDB id: 5A1S). The TM helices (in-out) and TM helices (out-in) are colored dark red/grey and light red/grey, respectively. Reentrant helices (in-in) and (out-out) are colored yellow and blue respectively. The inside and outside loops are colored yellow and blue respectively. The vertical bar is colored based on the taxonomy of the sequences (Bacteria: Purple, Archaea: Dark blue Eukaryotes: Green). Scaffold subdomains and reentrant core subdomains are colored brown and green, respectively. N- and C-terminal repeats are shown as black trapezoids. ΔG values describing the hydrophobicity [65] are obtained for the representative sequence and are plotted to the aligned residues in the representative sequence. (b) Validation of broken/reentrant transporters by using the KRbias for the DUF819 family. KRbias or positive inside rule is the enrichment of inside loops compared to the outside loops of a transmembrane protein[36, 46, 47]. The number of Lys (K) and Arg(R) amino acids in the inside and outside loops are compared. The expected correct topology would show a higher KR-bias in one of the two topology models (Broken/Reentrant). The KR-bias plot is shown as a 2D scatter plot. (c) Homology modelling of the representative sequence (d) Contact prediction modelling pipeline.