Study of correlation between protein sequence similarity and organisms optimal growth temperature

AS. HERNÁNDEZ BERTHET; AA. APTEKMANN; RM. GONZÁLEZ LEBRERO; ME NOGUERA; EA. ROMAN

Congreso; XLVIII Reunión Anual de la Sociedad Argentina de Biofísica; 2019

Sociedad Argentina de Biofísica

Structural and functional information of proteins are codified in their amino acidsequences. Within a protein family, there are proteins that come from different kinds oforganisms: some growth at high temperatures (termophiles), some at middle(mesophiles) and others at low temperatures (psychrophiles). Usually occurs that optimalprotein function temperature (OPFT) is close to the optimal growth temperature (OGT) ofthe organisms that come from. These behaviour is reflected in the sequence differencesbetween orthologs that could affect their function, stability and conformational flexibility.But which of these are more associated to the OGTs? Is there any region more coupledthan others? In this work we seek to study the sequences of an Archaea?s database tocompare the differences between these sequences and the OGT.To this end, we aligned the sequences of Adenylate kinase, SPT5 elongation factor andCopper-exporting P-type ATPase protein families. Then we compared each possible pairof sequences and quantify their differences, using different substitution matrices. Wemade comparisons in windows of different size throughout the alignment, or takingpositions near in space (using a representative structure of the family as a reference) ortaking combinations of random positions of the alignment. Then we made a linear fit ofthe data and analyzed the correlation?s coefficient ( R ) and Pearson?s significance (p).We obtained values of R for each family and analyze those that outliers of more than twoor three standard deviations from the average. We obtained positions of significantcorrelation for each family, and compared the correlation of these positions withbibliographic data.We found that in the studied families are positions or combinations that have positivecorrelation with the organisms OGT. Our preliminary results suggest that uponrefinement, this method may be used to predict the OGT of a given organism only fromsequence data.