Bioinformatics approaches for classification and investigation of the evolution of the Na/K-ATPase alpha-subunit

Abstract

Background: Na,K-ATPase is a key protein in maintaining membrane potential that has numerous additional cellular functions. Its catalytic subunit (α), found in a wide range of organisms from prokaryotes to complex eukaryote. Several studies have been done to identify the functions as well as determining the evolutionary relationships of the α-subunit. However, a survey of a larger collection of protein sequences according to sequences similarity and their attributes is very important in revealing deeper evolutionary relationships and identifying specific amino acid differences among evolutionary groups that may have a functional role.

Results: In this study, 753 protein sequences using phylogenetic tree classification resulted in four groups: prokaryotes (I), fungi and various kinds of Protista and some invertebrates (II), the main group of invertebrates (III), and vertebrates (IV) that was consisted with species tree. The percent of sequences that acquired a specific motif for the α/β subunit assembly increased from group I to group IV. The vertebrate sequences were divided into four groups according to isoforms with each group conforming to the evolutionary path of vertebrates from fish to tetrapods. Data mining was used to identify the most effective attributes in classification of sequences. Using 1252 attributes extracted from the sequences, the decision tree classified them in five groups: Protista, prokaryotes, fungi, invertebrates and vertebrates. Also, vertebrates were divided into four subgroups (isoforms). Generally, the count of different dipeptides and amino acid ratios were the most significant attributes for grouping. Using alignment of sequences identified the effective position of the respective dipeptides in the separation of the groups. So that ²⁰⁸GC is apparently involved in the separation of vertebrates from the four other organism groups, and ⁴¹DH, ⁴³¹FK, and ⁴⁵¹KC were involved in separation vertebrate isoform types.

Conclusion: The application of phylogenetic and decision tree analysis for Na,K-ATPase, provides a better understanding of the evolutionary changes according to the amino acid sequence and its related properties that could lead to the identification of effective attributes in the separation of sequences in different groups of phylogenetic tree. In this study, key evolution-related dipeptides are identified which can guide future experimental studies.

Keywords: Decision tree; Evolution; K-ATPase; Phylogenetic tree; α-subunit of Na.

Fig. 1

The phylogenetic tree of Na,K-ATPases for all organisms. Different symbols and colors were used to distinguish organisms and the type of isoforms. The scale indicates the number of amino acid substitutions per site

Fig. 2

The phylogenetic tree of different pumps in fungal sequences. Different colors were used to distinguish different kind of pumps

Fig. 3

The phylogenetic tree for 378 sequences of ssu Rrna (16S/18S rRNA) from various organisms of three life domains which were used for construct NAK phylogenetic tree. Different symbols and colors were used to distinguish organisms and the type of isoforms. The scale indicates the number of amino acid substitutions per site

Fig. 4

The phylogenetic tree of Na,K-ATPase for vertebrate organisms. Different symbols and colors were used to distinguish organisms and the type of isoforms. The hollow shapes indicated the isoform was predicted by the phylogenetic tree in this study. Circle, triangle, square and diamond shapes were used to show α1, α2, α3 and α4, respectively. The scale indicates the number of amino acid substitutions per site

Fig. 5

The changes of similarity rate of different isoforms within and among different groups of vertebrates. Am: amphibious, Fi: fish, Bi: bird, Ma: Mammalia, Re: reptile

Fig. 6

The Decision tree of different isoforms of α-Na,K-ATPase for vertebrate organisms. The tree generated using Random Forest model with information gain criteria when run on FCDS dataset. aa: amino acid

Fig. 7

The Decision tree of α-Na,K-ATPase for all organisms. The tree generated using Decision tree model with information gain criteria when run on FCDS dataset. aa: amino acid