Research Article

Evolutionary Analysis of Post – translational Modification Sites in Translation Elongation Factor 1A

Yosuke Kondo, Satoru Miyazaki

  1. Department of Medicinal and Life Science, Faculty of Pharmaceutical Sciences, TokyoUniversity of Science, 2641 Yamazaki, Noda-shi, Chiba 278-8510, Japan

*Corresponding author: Satoru Miyazaki, Department of Medicinal and Life Science, Faculty of Pharmaceutical Sciences, Tokyo University of Science, 2641 Yamazaki, Noda-shi, Chiba 278-8510, Japan, Tel: +81-4-7121-3630; Fax: +81-4-7121-3630; E-mail: smiyazak@rs.noda.tus.ac.jp

Abstract

Evolutionary conservation is one of the powerful analyses to predict protein functional sites. There have been many studies proposing a variety of degrees of conservation. In this paper, we propose and compare two mathematical measures that calculate the degree of conservation at each site in the molecular evolutionary process, named “the degree of conservation” and “the degree of specific conservation”. Thedegree of conservation can identify the sites which show a conserved pattern in all proteins of a proteinfamily. The degree of specific conservation is the degree that can identify the sites conserved in a subfamily but variable in the whole family. In this study, the conservation analyses were applied to a protein thathas many functions in the cell. Many researches of eukaryotic elongation factor 1A (eEF1A) revealed thateEF1A is involved in not only protein biosynthesis but also moonlighting functions such as cytoskeletalmodification, apoptosis and viral infection. And functional divergences of the EF-Tu/1A protein familyhave occurred in the evolutionary process. Our results showed that the degree of conservation can predictthe regions that are important for binding molecules of the eEF-Tu/1A proteins. The degree of specificconservation can predict the residues important for the functional divergences and post-translationalmodifications (PTMs). These results suggest that PTM sites of eEF1A are not conserved in the wholefamily but only conserved in a subfamily. It would be necessary to verify whether the conserved sitesare responsible for the moonlighting functions of eEF1A to develop new drugs that can be effective forcancers.

Keywords: evolutionary conservation, peptide elongation factor, post-translational modification

Introduction

Eukaryotic translation elongation factor 1A (eEF1A) is an abundant protein in cells and delivers aminoacyl-tRNA onto a ribosome [1]. It is said that the principal function of eEF1A is protein biosynthesis. However,eEF1A is revealed as one of the multifunctional proteins [2]. Novel eEF1A functions such as cytoskeletalmodification, apoptosis, viral infection and so on [3] have been clarified from the discovery of the actinbinding function [4]. eEF1A is also identified as a protein attaching cell membranes to induce anoikisby inactivating β1-integrin [5]. The antiadhesive function from the extracellular matrix is effective todevelop new drugs for cancers such as leukemia [6].

In addition to the moonlighting functions, many functional divergences have been observed in theeEF1A family. There are two known isoforms of eEF1A proteins, eEF1A1 and eEF1A2, in which thesequences share 92% identity [7,8]. Expression of the gene coding eEF1A depends on cell types [9]; themajority of cells express eEF1A1, neuron or muscle cells often express eEF1A2 and some of other celltypes express both eEF1A isoforms. In order to elongate a polypeptide chain by eEF1As, GDP/GTPbinding is necessary. However, eEF1As do not have identical binding affinities with GDP and GTP [10];the GTP binding affinity of eEF1A1 is higher than eEF1A2 but that of eEF1A2 is higher than eEF1A1.

Furthermore, eEF1A1 promotes apoptosis [11] but overexpression of eEF1A2 causes cancers [12]. Suchfunctional differences are important for elucidating various cell functions.There are 36 non-identical residues between human eEF1A1 and eEF1A2. It is known that some ofthese residues or corresponding residues of eEF-Tu are estimated to be important for actin binding orfibronectin binding functional divergences [13, 14]. On the other hand, identical residues were estimatedto be also important for the functional divergences because GDP/GTP binding affinity is different becauseof non-identical residues that are close to the identical residues [15]. Recent studies showed that post-translational modification (PTM) sites of eEF1A determined by mass spectrometry analysis [16-18]are located on regions close to the non-identical residues [15]. PTM residues should be important foranalyzing the functional divergences of eEF1As.

PTM sites of eEF1A is previously estimated as non-conservation sites [19]. Meanwhile, our previousstudies showed that residues involving binding molecules or functional divergences of the EF-Tu/1Afamily can be predicted by evolutionary analysis [20, 21]. Therefore, evolutionary conservation of theEF-Tu/1A protein family is useful to investigate important residues of the proteins. In this paper, wepropose degrees of conservation and discuss what degree of conservation can accurately predict the PTMsites of eEF1A. This analysis would clarify how the PTM sites of eEF1A proteins are evolved.

Materials and Methods

Construction of the multiple sequence alignment and phylogenetic tree

From UniProtKB/Swiss-Prot [22], we collected 984 entries, which (1) are annotated as ‘Classic translationfactor GTPase family. EF-Tu/EF-1A subfamily’, (2) do not include ‘X’ in the sequence and (3) are not afragment. The sequences were aligned by MAFFT 7 program[23]. Distances of sequence pairs were com-puted by maximum likelihood method[24] using the Jones-Taylor-Thornton model[25] as a substitutionmatrix and the Dayhoff method[26] for computing equilibrium frequencies. From all combinations of thedistances, a phylogenetic tree was written by unweighted pair group method with arithmetic mean[27].

Preparation for defining degrees of conservation

Let M = (mij) denote a given multiple sequence alignment (MSA) and here mij denote an amino acidsymbol in sequence i of site j on the MSA. Let Xj = {m1j,m2j , . . . ,mnj} be a multiset of amino acidsymbols. Let Fjdenote a field of sets of Xj. And, let Fjtbe an element of Fj and a group by node t ina phylogenetic tree.

Mathematical formulation for a degree of variability

be a normalized substitution matrix, where Smax, Smin, S (y, y) and S (y, z) are the maximum, the minimum, a diagonal element and an off-diagonal element in an amino acid substitution matrix, respectively,and w (y) is a weight of sequence y. The Gonnet matrix [28] was used as a substitution matrix. Theweight was computed by Sibbald and Algos algorithm[29] and the iteration number was 100,000.

Mathematical formulation for degrees of conservation

Data collection of functional sites in eEF-Tu/1A

Binding residues were collected from three-dimensional structures of eEF-Tu/1A described in our previousstudy [21]. Actin-binding residues were obtained from site-directed mutagenesis data [30]. PTM siteswere obtained from the PhosphoSitePlus database [31].

Results

Predicted sites by degrees of conservation

The multiple sequence alignment constructed from the sequences in the EF-Tu/1A family consisted of739 sites. The degrees of conservation and specific conservation were calculated at each site. In orderto determine a threshold of each degree of conservation, we used ROC (receiver operating characteristic)curves.

The ROC curve of the degree of conservation was created from proximity of binding molecules orions in three-dimensional structures of the EF-Tu/1A proteins as described in our previous paper [21].Because Figure 1A shows that the threshold is 0.973, predicted sites were defined as the sites in whichthe degree of conservation is higher or equal to the threshold. The predicted sites shown in Figure 1Bare mainly located on the left side of the three-dimensional structure.

The ROC curve of the degree of specific conservation was constructed from actin binding residues[13]. Because Figure 2A shows that the threshold is 0.190, predicted sites were defined as the sites inwhich the degree of specific conservation is higher or equal to the threshold. The predicted sites shownin Figure 2B are mainly located on the right side of the three-dimensional structure.

In order to clarify the difference of the predicted sites between each degree of conservation, we investigated overlapping predicted residues shown in Figure 3A. Hundred residues were overlapping residuesthat were predicted both degrees of conservation. Figure 3B also shows that the degrees of conservationand specific conservation mainly predicted the left and right side of the three-dimensional structure, respectively.

The degrees of conservation and PTM sites

PTM sites of eEF1A1 in human, mouse, rat and rabbit and eEF1A2 in human, mouse and rat were shownin Figure 4. In order to analyze which degree of conservation can accurately predict the PTM sites, thetrue positive rate and the false positive rate were calculated as shown in Table 1. The true positive rateof the degree of specific conservation is 0.827. This value is higher than that of the degree of conservation.

Meanwhile, the false positive rate of the degree of specific conservation is higher that of the degree of conservation. In addition, we created ROC curves of PTM sites and each degrees of conservation showning Figure 5. The AUC (0.773) of the degree of specific conservation is higher than the AUC (0.649) ofthe degree of conservation.

Discussion

In this paper, we investigated the conserved sites of eEF1A using two degrees of conservation. One isthe degree that can identify the sites which show a conserved pattern in all proteins of a protein family.As shown in our previous study [21], the degree of conservation can predict close sites from bindingmolecules in the three-dimensional structures of eEF-Tu/1A. eEF1A has the conserved face and thevariable face in the three-dimensional structure [15]. These faces were decided from variable residuesbetween eEF1A1 and eEF1A2. The conserved face is similar to the regions that are determined by thedegree of conservation shown in Figure 1B. This shows that the degree of conservation can predict theconserved face of eEF1A.

The other degree of conservation is a specific conservation that can identify the sites that are onlyconserved in a subfamily of a protein family. In other words, the degree of specific conservation canidentify the sites conserved in the subfamily but variable in the whole family. In this study, the subfamilyis specified as the group that contains vertebrate eEF1A1 and eEF1A2. The whole family is specified asthe group that contains whole eEF-Tu/1A family. Therefore, the degree of specific conservation predictsfew non-identical sites between eEF1A1 and eEF1A2. However, the degree of specific conservation canmainly predict close sites from the non-identical residues shown in Figure 2. This suggests that the degreeof specific conservation can predict the residues that are important for functional divergences betweeneEF1A1 and eEF1A2. In addition, the regions that are predicted from the degree of specific conservationare similar to the variable face described above. Because there are many PTM sites in the variable face,the specific conservation may be useful for predicting PTM sites.

As shown in Table 1 and Figure 5, our results show that the PTM sites were predicted well by thedegree of specific conservation than degree of conservation. In addition, Table 1 shows that the truepositive rate is high but the false positive rate is high in the degree of specific conservation. Therefore,the degree of specific conservation can accurately predict the known PTM sites that are registered inthe database of PTM sites. This shows that the degree of specific conservation is superior to the degree of conservation to predict PTM sites of eEF1A proteins. These results suggest that one degree ofconservation is inadequate to predict all important residues of eEF1A.

In conclusion, various degrees of conservation may be useful when we predict functional sites ofmultifunctional proteins like eEF1A proteins. In addition, our results suggest that PTM sites of eEF1A1and eEF1A2 are not conserved in the whole family but conserved in the subfamily.

Abbreviations

eEF: eukaryotic elongation factor; MSA: multiple sequence alignment; PTM: post-translational modification

Conflict of Interest

The authors declare no conflict of interest.

Table 1. Accuracy for predicting PTM sites by degree of conservation or specific conservation.

True positive rate False positive rate
C_ 0.395 0.291 C_ 0.827 0.376

Reference

  1. Andersen G, Nissen P, Nyborg J. Elongation factors in protein biosynthesis. Trends Biochem Sci. 2003; 28: 434-441.
  2. Ejiri S. Moonlighting functions of polypeptide elongation factor 1: from actin bundling to zincfinger protein R1-associated nuclear localization. BiosciBiotechnolBiochem. 2002; 66: 1-21.
  3. Mateyak MK, Kinzy TG. eEF1A: Thinking Outside the Ribosome. J Biol Chem. 2010; 285: 21209-21213.
  4. Murray J, Edmonds B, Liu G, et al.Bundling of actin filaments by elongation factor 1 alphainhibits polymerization at filament ends. J Cell Biol. 1996; 135: 1309-1321.
  5. Itagaki K, Naito T, Iwakiri R, et al. Eukaryotic translation elongation factor1A induces anoikis by triggering cell detachment. J Biol Chem. 2012; 287: 16037-16046.
  6. Matsunaga T, Fukai F, Miura S, et al., Combination therapy of an anticancerdrug with the FNIII14 peptide of fibronectin effectively overcomes cell adhesion-mediated drugresistance of acute myelogenousleukemia. Leukemia. 2008; 22: 353-360.
  7. Knudsen S, Frydenberg J, Clark B, et al. Tissue-dependent variation in the expression of elongation factor-1-alpha isoforms – isolation and characterization of a cdna-encoding a novel variantof human elongation-factor-1-alpha. European J Biochem.1993; 215: 549-554.
  8. Lund A, Knudsen S, Vissing H, et al. Assignment of human elongation factor1 alpha genes: EEF1A maps to chromosome 6q14 and EEF1A2 to 20q13.3. Genomics.1996; 36:359-361.
  9. Lee S, Francoeur A, Liu S, et al. Tissue-specific expression in mammalian brain, heart, andmuscle of S1, a member of the elongation factor 1 alpha gene family. J Biol Chem. 1992; 267:24064-24068.
  10. Kahns S, Lund A, Kristensen P, et al. The elongation factor 1 A-2 isoformfrom rabbit: cloning of the cDNA and characterization of the protein. Nucleic Acids Res. 1998; 26:1884-1890.
  11. Duttaroy A, Bourbeau D, Wang X, et al. Apoptosis rate can be accelerated or decelerated byoverexpression or reduction of the level of elongation factor-1 alpha. Experimental Cell Res. 1998;238: 168-176.
  12. Anand N, Murthy S, Amann G, et al. Gene encoding protein elongationfactor EEF1A2 is a putative oncogene in ovarian cancer. Nature Genet. 2002; 31: 301-305.
  13. Soares DC, Barlow PN, Newbery HJ, et al. Structural models of humaneEF1A1 and eEF1A2 reveal two distinct surface clusters of sequence variation and potential differences in phosphorylation. PLoS ONE.2009; 4.
  14. Balasubramanian S, Kannan TR, Hart PJ, et al. Amino acid changes in elongation factorTu of Mycoplasma pneumoniae and Mycoplasma genitalium influence fibronectin binding. InfectImmun. 2009; 77: 3533-3541.
  15. Soares DC, Abbott CM, Highly homologous eEF1A1 and eEF1A2 exhibit differential post-translational modification with significant enrichment around localised sites of sequence variation.Biology Direct.2013; 8.
  16. Jakobsson ME, Matecki J, Nilges BS, et al. Methylation of human eukaryoticelongation factor alpha (eEF1A) by a member of a novel protein lysine methyltransferase familymodulates mRNA translation. Nucleic Acids Res. 2017; 45: 8239-8254.
  17. Jank T, Belyi Y, Wirth C, et al. Protein glutaminylation is a yeast-specificposttranslational modification of elongation factor 1A. J Biol Chem. 2017; 292: 16014-16023.
  18. Hamey JJ, Wilkins MR. Methylation of elongation factor 1a: Where, who, and why? TrendBiochem Sci. 2018; 43: 211 – 223.
  19. Cavallius J, ZollW, Chakraburtty K, et al. Characterization of yeast ef-1: Non-conservationof post-translational modifications. BiochimicaetBiophysicaActa (BBA) – Protein Structure andMolecular Enzymology. 1993; 1163: 75 – 80.
  20. Kondo Y, Kwon Y, Miyazaki S. Detection of key residues involving functional divergence into thetranslation elongation factor Tu/1A family using quantitative measurements for specific conservation of protein subfamilies. J ComputSciSyst Biol. 2014; 7: 054-061.
  21. Kondo Y, Miyazaki S. Protein functional site prediction using a conservative grade and a proximategrade. J Data Mining Genomics Proteomics. 2015; 6: 175.
  22. Consortium TU.Uniprot: a hub for protein information. Nucleic Acids Res 2015; 43: D204-D212.
  23. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvementsin performance and usability. MolBiolEvol. 2013; 30: 772-780.
  24. Kishino H, Miyata T, et al. Maximum-likelihood inference of protein phylogeny and theorigin of chloroplasts. J MolEvol.1990; 31: 151-160.
  25. Jones D, Taylor W, Thornton J. The rapid generation of mutation data matrices from proteinsequences. ComputApplBiosci.1992; 8: 275-282.
  26. Dayhoff MO, Schwartz RM. 1978 Chapter 22: A model of evolutionary change in proteins. In: inAtlas of Protein Sequence and Structure.
  27. Sneath PHA, Sokal RR. 1973 Numerical taxonomy: the principles and practice of numerical classification. San Francisco: W. H. Freeman.
  28. Benner S, Cohen M, Gonnet G. Amino-acid substitution during functionally constrained divergent

evolution of protein sequences. Protein Eng. 1994; 7: 1323-1332.

  1. Sibbald P, Argos P. Weighting aligned protein or nucleic acid sequences to correct for unequal

representation. J Mol Biol. 1990; 216: 813-818.

  1. Gross SR, Kinzy TG. Improper organization of the actin cytoskeleton affects protein synthesis atinitiation. Mol Cell Biol. 2007; 27: 1974-1989.
  2. Hornbeck PV, Zhang B, Murray B, et al. PhosphoSitePlus, 2014:mutations, PTMs and recalibrations. Nucleic Acids Res. 2015; 43: D512-D520.
  3. Berman H, Westbrook J, Feng Z, et al. The protein data bank. Nucleic AcidsRes. 2000; 28: 235-242.
  4. Crepin T, Shalak VF, Yaremchuk AD, et al., Mammalian translationelongation factor eEF1A2: X-ray structure and new features of GDP/GTP exchange mechanismin higher eukaryotes. Nucleic Acids Res. 2014; 42: 12939-12948.

 

Received:  July 18, 2018;
Accepted: August 17, 2018;
Published: August 21, 2018

To cite this article:

Kondo Y, Miyazaki S. Evolutionary Analysis of Post-translational Modification Sites in Translation Elongation Factor 1A. Japan Journal of Medicine. 2018: 1:6.

©Kondo Y, et al. 2018.