An evolutionary trace method defines binding surfaces common to protein families

TLDR: The evolutionary trace method is a systematic, transparent and novel predictive technique that identifies active sites and functional interfaces in proteins with known structure and provides an evolutionary perspective for judging the functional or structural role of each residue in protein structure.

About: This article is published in Journal of Molecular Biology.The article was published on 1996-03-29 and is currently open access. It has received 1221 citations till now. The article focuses on the topics: Protein superfamily & Protein family.

Figures

Figure 3. Evolutionary trace of the complete family of SH2 domains mapped onto the 2.7 Å resolution structure of Src SH2 domain (Waksman et al., 1993). Each row displays the protein in sequential 90% rotations about the z-axis, and successive rows show traces that arise from partitions A1 to G1, as defined in Figure 1. The internal positions that are conserved (dark blue) and class-specific (red), and the external class-specific (yellow) and conserved (cyan, though none in this Figure) positions are distributed inhomogeneously and form a single localized cluster on face A. At low PIC, partitions A and B, the trace is specific and points to only three residues, R32, H58 and Y59. To these core residues K57, K60, I71 and Y87 are added contiguously in partition C, L94 and R74 in partition D and C42 in partition E. The other faces remain essentially free of trace signal until partition G1. There, the scattered appearance of trace positions suggests that the useful limit of functional resolution has been reached at that higher PIC value. Partition F1 is not shown because it is equivalent to partition E1.

Figure 8. ET analysis carried out over a restricted SH3 domain family (see Figure 6). PIC choices, Abl SH3 orientation and color codes are unchanged from Figure 7, and the same region, on face A, is picked out by ET as functionally important. The face A cluster now prominently extends onto face D, consistent with a possible functional role distinct from ligand binding, in SH3 domains from tyrosine kinases.

Figure 4. Results of an ET analysis performed over a subfamily of SH2 domains, consisting of the sequences between fg kabl and fg nck in Figure 1. PIC values, orientation of the Src SH2 domain and color codes are as in Figure 3. The trace finds the same spatial region as it did in Figure 3 to be functionally important, though full definition of the cluster (partition C2) and the appearance of scattered signal (partition D) occur earlier, at a lower PIC value.

Figure 7. Evolutionary trace of the SH3 domain family mapped onto the 2.8 Å resolution structure of Abl SH3 domain (Musacchio et al., 1994), conventions follow those of Figure 3 and the traces correspond to the PIC values in Figure 6. A single domain is defined by evolutionary tracing to be functionally important, its core comprises position P133 (B1 and C1), and the adjacent W120, Y136 (D). The cluster is well defined in partitions E1 and G1, and scattered signal appears in partition H1. Faces B1 and D1 show side views of this cluster.

Figure 1. Sequence identity dendrogram of SH2 domains obtained as described in Methods. Vertical lines A to G represent the different partition identity cutoffs (PICs) that are used to define partitions. For each PIC, a partition of the entire family is generated by grouping together sequences that branch off from a node, shown in red, to the right of a PIC line. As PIC increases, from A to G, partitions comprise more groups, each with fewer sequences. Thick colored lines are used in this example to display the different groups in partition A. Six nodes, in black, give rise to singular groups and effectively drop out of the evolutionary trace analysis. ET analysis was performed over this entire tree, see Figure 3, and over the subtree shown in purple, see Figure 4.

Figure 10. ET analysis on the nuclear hormone receptor family of DNA binding domains (zinc fingers). Two zinc fingers homodimerize and are complexed to a palindromic DNA hormone response element, in yellow (Luisi et al., 1991). A to C Consecutive 90° rotations about the x-axis. ET mapping follows the usual color code. No functionally important cluster occurs except in C, the DNA binding surface, where conserved and classspecific positions form a single cluster per ZnF domain. Upon homodimerization these clusters create a large area of functionally important positions at the DNA binding interface.

Frequently asked questions

Q1. What are the contributions mentioned in the paper "An evolutionary trace method defines binding surfaces common to protein families" ?

The SH2 and SH3 modular signaling domains and the DNA binding domain of the nuclear hormone receptors provide tests for the accuracy and validity of their method. 

Q2. What is the importance of assessing the signal to noise problems that are inherent in sequence analysis?

As proteins undergo neutral drift during evolution, it is important to assess the signal to noise problems that are inherent in sequence analysis. 

Q3. What is the purpose of the ET method?

The ET method exploits the information inherent in a family of homologous proteins by dividing it to maximize functional similarity within groups and functional variation between groups. 

Q4. How many different branches of the SH2 domain are available?

The availability of sequences among divergent SH2 domains ranges from 13, in the ksrc group of sequences (identity greater than 90%), to many single representatives of distinct evolutionary branches such as shc, gagc and kcsk. 

Q5. What is the way to cluster proteins?

In the absence of functional information for every sequence of a large evolutionary family, clustering proteins by sequence identity will produce reasonable groups containing proteins with similar functions. 

Q6. What is the reason why the thyroid hormone receptor group fails to appear in the trace?

which is surrounded by functionally important residues and has been shown to be functionally important to binding, prominently fails to appear in the trace, due to a single mutation, A462S, in the thyroid hormone receptor group. 

Q7. What is the pattern of emergence of cluster 1?

The pattern of emergence of cluster 1 is characteristic of a functionally significant region, where residues are conserved within each cluster, but can vary between them. 

Q8. What is the common application of ET analysis?

ET analysis has been applied to the family of Ga proteins and to functional subgroups within the ZnF family of nuclear hormone receptors (unpublished results). 

Q9. What is the significance of cluster 1?

In both partitions, cluster1 clearly stands out as a region of coalescing conserved and class specific residues on a background that is otherwise free of signal. 

Q10. How many sequences were identified in the case of the SH2 domains?

In the case of the SH2 domains, 85 sequences of approximately 100 residues each were identified, and aligned to generate a sequence identity dendrogram, using standard techniques (see Methods). 

Q11. What is the ligand subsite of py?

Å of pY + 2, a ligand subsite that frequently contains a negatively charged side-chain (E in the YEEI high-affinity peptide of src). 

Q12. What are the main features of partitions?

Such partitions can group together sequences based on any number of functional characteristics and need not follow the groupings the authors use here, which are closely related to patterns of divergent evolution. 

Q13. What is the significance of the evolutionary trace?

Their results show that in both the SH2 and SH3 domains, and in the DNA binding domains of nuclear hormone receptors, the evolutionary trace identifies the ligand binding site. 

Q14. What would happen if solvent accessibility was used as the only guide to distinguish structural from functionally?

If solvent accessibility was used as the only guide to distinguish structurally from functionally important positions, the unique functional characteristics of some protein clefts would be masked and the contiguity of the functional epitope would be disrupted. 

Q15. What is the simplest explanation for the morphology of ZnF domains?

This is expected, since different ZnF domains dimerize either as homodimer or heterodimers, and each mode involves distinct, non-overlapping regions. 

Q16. What is the difference between the two groups of zinc fingers?

positions in the homodimer interface, shown in Figure 10, appear neutral because they are neither conserved nor class-specific among zinc fingers that heterodimerize.