Nature of amino acid substitutions in homologous proteins during evolution A. S. Kolaskar and V. Ramabrahmam School of Lite Sciences, University of Hyderabad, Hyderabad-S00 134, bndia (Received 11 May 1981) Replacement substitutions of mitochondrial cytochrome cand x and chains of huemulobin hace heen studied ‘ay considering the structural soilrity among amino acid residues a the secondary and tertiary structural leels Secondary siruetural similarity explains ~ 70, while tertiary structural similarity explains ~ 50", of observed Peplacentents for most of the cases. These structural simlarties could not account forall she replacement substitutions, Te study was extended 19 consider the composition of edlons, won the chemical nature a polarity ‘ofthe replacing und replace residues. These alo could not individually account for ll he affected replacements In general, mo property of mina acid residues is consereed for substitutions ovcurring at any single position luring evolution of proteins, Keyword Introduction The study of replacement substitutions between sequences of homologous proteins of different species has been looked to for information about evolutionary pathways', and a genetic basis has been given by Fitch and Margoliash? for these replacement substitutions. The sequence differences between homologous proteins have also been used to estimate replacement site divergence of corresponding genes by determining the minimum number of base changes needed to generate the observed amino cid replacements and multiple base changes within individual codons? However, the limitations in detecting certain replacement substitutions precludes the accurate estimation of rates of fixation of base changes® In spite of these limitations, many groups have successfully made use of these replacement substitutions in establishing molecular phylogenies and taxonomic relationships” '”. Thus, the question arises as to why nature has accepted only certain mutations out of the large number possible. It may be that only those mutations which have a minimal effect on the three- dimensional structure of the homologous protein are accepted. This can be seen ftom the erystal structure data of various cytochrome c!#? and haemoglobins??*" Structural data indicate that the local three-dimensional structure undergoes little change, which probably helps to increase the efficiency of the protein, or (0 conserve itn new environment. Replacement substitutions’ in homologous proteins therefore provide valuable structural information Two levels of structural similarity among amino acid residues were considered in the present study. First, the secondary structure, which includes preference or indifference to a particular secondary structure by an amino acid residue, and second, the tertiary structure. Amino acid residues that are similar at the level of tertiary structure are termed conformationally similar, and have 0141 -$130/821030180-05803.00 O19? Buterworth & Co. (Publishers) Ls 150 Int, J, Biol. Macromol, 1982, Vol 4, April Proteins amino acs: haemoglobin eytechrome ¢ similar probability distributions in the (@,-plane?2. This, structural similarity is found to be an independent property of amino acid residues, We have studied the replacement substitutions from secondary and tertiary structure points of view in ‘mitochondrial cytochrome ¢, and the and B-chains of haemoglobin from various species, This study indicated that the observed replacements can not all be explained fon a structural basis, We extended the study by considering the polarity, chemical nature and composition of codons of replaced and replacing residues. Individually, none of thes similarities could explain all the replacements observed in the species considered for cytochrome cand haemoglobin. We also looked for the conservation of specific properties of residues, such as those mentioned above, at particular position in the sequence, where a replacement takes place. ‘The results were inconclusive, suggesting that the substitutions uccepted by nature during evolution can not be explained by considering any single property that we examined of the amino acid residues. Method ‘The amino ucid sequences of mitochondrial eytochromec, zzand fi-chains of haemoglobin for the species considered {sce Tables 1 and 2 below) were taken from Handbook of Biochemistry and Molecular Biology”. The replacement substitutions of these sequences were compared as follows. For eytochome c, the reference species ranged from Rhesus monkey to Pacific lamprey: for a and f- chains of haemoglobin, the reference species ranged from Rhesus monkey to kangaroo. Replacements in the sequences of other species were studied with respect to the sequence of reference species. In Table 1 the first row corresponds to replacements observed in the sequence for human eytochrome © when the reference species variedAmine acid substitutions in proteins during evolution: A. 8, Koluskur and V, Raanabrabonn “Table 1 (a) Percentage of replacements in cytochrome e by residues similar in secondary seuetural ality é e3 fable gig th? siisetier, 222228 z sminuee FEES ES EF E28 E22 é Mon wa Sha Ree DST HOH w hes monkey Sehaag nyo nh 5 8M 2D Hone - sO BoM Bw MT Donkey - 00 37 G7 1 GTS Ge OS GO Con oo 88 wwe DR A Ciel = 2 ITS 3 wie Fo Bw M6 15 35 Elepan sea * io 93 ‘or 5 a0 50 8) 3 Se ad bee @ © 6 1 aw We 3 3 Ba - % 30 40 5038 88 @ Rabbi BG FS 8 Ho 8 9S 5 UO 1H HTT BS Kangoo - - HO Ha No we TW TS Chicken - 50100 67 30 G8) 42 UF ON MTT Te Em So 1 er 5067 SO A King pene SG 0 oh aS Be HOE OT sy Pekin duck eee wa we D Pisum 8 ONG 2 Rew Soupping ure - - Te oe Rates - - - St ee mS oF ball oe Soe ‘una So SSS te Bonito - So Ci - Dog sh Soe “Table (by Percentage of replacements by conformationaly similar residues in eytochrome ¢ Reference species peey Elephant seal Pekin duck Rabo: Species studied Rhesus monkey Hore - 2 9 Donkey - * Cow - Camel - 0 Elephant seal — Dog at a Rabbit - Kangaroo — - Chicken - - Emu - - 067 King penguin — Pekin duck - So Pigson| - Snapping turtle ~ - Ratlesnake — So £286 on -weyewsne Oe BR 8 a » leeks & Carp - oo Dog fish - - Int. J. Biol. Macromol, 1982, Vol 4, April 151Amin acid substitmions in proteins during evolution: A. S. Kolaskar and ¥. Ramabrakomam ‘Table 2_ (al Percentage of replacements by residues similar in secondary steuctural afinity Haemoglobin 2-chain| Haemoglobin f-chain| z 25 ; Gye oe : 5 g e222» § 4 Reference species ze E keewnespics Fes eee PFs Kangaroo 15 18 TS TE 76 73 19 80 Kangaroo Pe ee Expat 63 73 78 10 82 67 Ow Horse. slow 5056 5 68 44 50 62 — Hone 6 TL 67 6 7D oe 7 Pig 28 97 $8 8 a0 Pie 4g nO Sheep 59 56 70 a7 Sheep ws bo Dog 30 7 2 78 Dos 38H Rabe 76 a Rabbit B03 HO Mouse NB » 7 Mouse CS7BL nm Rsv monkey %0 — = = = hess monkey 68 “Table 2. (by Percentage of replacements by conformational similar residues Haemoglobin chain Haemoglobin fin a8 g Be = 24¢ 06 2 25 e ed shit 2 52S 2225.83 Reference species Bz S25 FE S peterence species 23287820 Kangaroo 32M 37 38 36 2027-27 Kangaroo 50 a0 36 Elephant uM Hose slow 17 25 38 $2 33 25 23 — Hone a 5 33 0 0 > he - 93 4 oo Shesp = 30 3 36 2 — = Dog aoe aoa wR Mouse NB 24 = > — Moue cs7t Rhee monkey ao oo Rhesus monkey - from Rhesus monkey to Pacific lamprey. Similarly, when examining replacements in the eytochrome ¢ sequence of cow, the reference species varied from camel to Pacific lamprey. In the haemoglobin comparisons, the reference species are given in separate columns in Tuble 2 for and Fechains. The amino acid residue in the sequence of the reference species which can be replaced is termed the replaced residue and the residue Which occupies that position in the sequence of other species is the replacing residue Each replacement was studied by taking into consideration the following characteristics of amino acid residue preferences and similarities, Secondary structure preference Levit’s classification of secondary _ structural preference of amino acid residues was used, since this classification has no overlap and clearly states whether a particular amino acid residue favours is indifferent to or breaks w particular secondary structure. replacing residue that prefers or is indifferent to the secondary structure preferred by the repluced residue is considered to have similar secondary structural preference 152. Int. J. Biol, Macromol, 1982. Vol 4, April Conformational similarity Conformationally similar residues were taken from ‘Table 2 of our previous paper*®, This table gives a set of amino ucid residues having a high probability 10 adopt similar main chain conformations. Thus, if the replacing, residue is conformationally similar to the replaced residue, a minimal change in the tertiary structure is expected, Composition of codons Residues that have tvo bases in common, itespective of the position in respective codons, were considered to be similar in this study. Chemical nature Classification of the chemical nature of residues was adapted from Stryer®* Polarity ‘The residues are divided as hydrophobic, neutral polar or polar, as deseribed by Lehninger" A simple algorithm was developed to study theseAmino acid substitutions in proteins during evolution: A. S. Kolaskar and ¥; Remabrahnan features of replacement substitutions in. mitochondrial eytochrome « and the « and chains of haemoglobin in dliflrent species. Results and discussion Tables 1 and 2 give the percentage of observed replacement substitutions for respective comparisons when a particular property or similarity of amino acid residues is considered. For example, the value 67 in the third column ofthe ist row of Table Ja signifies that ifthe sequence of human cytochramecis compared with that of hhorse as the reference species, 8 out of 12 replacements (67%) are such that the replacing residues have similar secondary structural affinity to the replaced resi Similar results are given in Tuble 2afor the and f-chains ff haemoglobin. These Tables show that similarity in secondary structural afinity of the replacing and replaced residues explains between 100 and 20% of replacement substitutions, although for most eases the percentage is around 65-75. In tis study, we considered substitutions throughout the polypeptide chain, so that the results should not be interpreted as meaning that the secondary structural regions are conserved. Indeed, study of the exystal structure of Tuna and Bonito cytochrome. ¢ indicates that the secondary structural region is not the same in both species, Tuna eytochrome ¢ consists of five tchelices while Bonito cytochrome ¢ consists of three ‘chalice and two 3,9 helices. At position 61, which lis in schelical region of both proteins, the residue was Asm in Tuna and Gln in Bonito; Asn is an achelx breaker and Gln is an achelix preferrer. Thus, the secondary structural regions difer and the replacing and replaced residues do not prefer the same secondary structure We then examined the replacement substitutions by conformationally similar residues. The results are given in Tables 1h und 2h for eytochrome ¢, and « and f-chains of haemoglobin, respectively. Similarity of the tertiary structure explains a considerable numberof replacements: the percentage of replacements explained varies between 20 and 60. Analysis of crystal structure data indicates, ‘hat almost all the replacing residues ean adopt a similar main chain conformation to that of the replaced residue. ‘This conclusion is derived from our observation that when the (g,)-maps are compared the replacing and replaced residues have very low diserepancy values between the irids which represent these observed conformations, ‘Thus, the replacing amino acid residues are those which have a minimal effect on the three-dimensional structure of the protein Since none of the 1wo structural similarities could fully explain the replacements, we studied the replacement Substitutions from the other three aspects mentioned in the Method to see whether consideration of any one of these properties could explain the replacements, ‘When the composition of codons of replacing and replaced residues Was considered, ic. genetic similarity, i could not explain all the replacements, only ~70% Similarity in chemical nature or polarity explained < 507, of replacements" Since no. single feature could explain all the replacements, we examined the data to see whether any of Bate avaible from he publisher. these properties is conserved in a position affected by replacements duringevolution, For example, residue 89 of the eytochrome « sequence for dillerent species, going from Pacific lamprey to man is Gly, Ser, Gly. Giy. Gly. Gly, Lys, Ala, Alt, Se, Ser, Ser. Ser. Gly. Asp. Ala, Gy. Gly, Gly, Gly, The, Thr, Glu, Glu, Several observations can'be made for this example (1) The residues Ser, Asp and Gly prefer chain reversals, Ala, Lys and Glu preler 2-heix and Thr prefers sheet. Giy is an achelx breaker and Ala is a breaker of chain reversals, Thus similarity at the secondary structural evel is not maintained (2) Conformational similarity is not preserved. Lys, Gu, Ala, Asp, Set, Tr are ll conformationally diferent from Gly. (3) Codons for Lys and Thr differ from those for Gly by not having two bases in common, irrespective of position in the codon, Thus, the codons for the replacing and replaced residues are not similar. (4) Amino acids having diffrent chemical natures occur atthe same position: Ala and Gly are aliphatic, Thr and. Ser are hydroxy amino acid residues. Asp and Glu are acidic and Lys is basi (5) The polarity of the residue varies: Gly. Ser and The ‘are neutral polar, Asp, Glu and Lys are polar and Ala is hydrophobic “Thus, none ofthese properties is conserved at reside 89 of eytochrome c. Similar observations are made at other positions inthe eytochrome.csequenceswellas the and chains of haemoglobin, During evolution, therefore. the femphasis seems to be on conservation of the three- dimensional structure of homologous proteins rather than conservation of local structure, chemical nature or polarity. This may refletspecies-characteristc functions Of these homologous proteins in addition to their common function ‘The only partial explanation of replacements between sequences of any two species by individual properties of amino acid residues, and the non-conservation of any of these. properties in replacement substitutions. during evolution, shows that this type of study alone cannot provide enough information to establish an evolutionary pathway, Phylogenetic trees constructed using data from Tables J and 2and methods suchas those developed either by Dayhoff", or the modified method’, may provide a better basis. Generation of such phylogenetic trees isin progress, and. will include not only the amino acid substitutions. but will also indirectly account for the Additional functions characteristic of u given species, incorporated during evolution, “Tis study also points to the need for such studies, 0 ‘id in understanding whether micto-evolution and maero-evotion are in phase, ie. to establish whether the ‘changes observed in proteins nucleic acids and hormones, eorrespond one to One with morphological changes of species References 1 Sehwara. RM. and Daybolt M.O. Scene 1978, 199,395 2 Fitch WM. and Margolin. ETsolutonary Blogs (ES {T Dolwamsky, M_K. Hee and WC. Steel. Appision entrs-CroteEdocutional Divson/ Meret Corporation, ‘New York 190, Vol pp. 67-109 3 Wikon AC. Culoson 88 and White. Aaa Re, Bam, orn, 373 Int. J. Biol, Macromol, 1982, Vol 4, April 153Amino ec substitutions in proteins during evolution: A. 8. Kolaskar and ¥, Ramabrahmam 4 Dickerson Rand Go in “Protein: Structure, function snd fokiton Menlo. Park, RensmintCamminesPabtshing, ‘Calor, 580 5) Ehirstinds A, Poskony. J. W. Mania Lawn, RM OWonra Ge Spee R.A” Do Ric, 3K Forges BG Weismann. Mt Shion, JL, Blech AE. Siti, 0 Baral FE, Shoulders. Cand Proudfoot Ni Cel 98,2 so Margossh. E.Prc. Natl Asal. So USA 196% $0,672 Sith E Land Margolish E Fl. Pro 1964.23, 1283 Smith, E Land Margolis E Pl Pre. 19.23, 1276 MurgSlish and Sith EL a"Evoling Genes and Protas’ (Eds'¥. Bron and H.} Vogl) Asageme Pres, New York, 1965.22 10 Fok WM. and Margolin, See 1967, 188,279 HL Nolun.C. and Margotiash Emr Ree, Biche, 164,37, 727 12 Dayhofe MOL Sete July 1903 6 Hh Sith Cin The Enya Ea PD, Boyer Academie Press ‘Now York, 1970, Vo pp 27339 14 Dickerson, RE J. Ma rol 197, 1,26 Walls." Sig“Chemsiry” of Macromolecules) MTP Tetrmtioal Review of Snes, Brocersty Seis One (Ed H Guteund} Butersorts, London, 1978 Vol lepp I-36 154 Int. J. Biol. Macromol, 1982, Vol 4, April as % Dayhof MO. MeLauglin, PJ, Saker. W.C and Hun LT. Naturatsonschon 1973, 62 13 ‘Barnabus. Mithew, . Ae Ratna Pariah M. Vand Barnabas, Sind Biche. Bip 197 18.386 Dickeson Ri Eand Fnkovche fe inThe Enrymes (Ed, PD. Boyer Academic Pros, New York, 97S. Vol 1h pp 397-847 Seavasun, AR J Pept, Proc Res. 180.6. Porte Me Fe J Atl fl 1988 136 Pete MF, Kendron J. Cand Walon, H.C. J- Mol Biol 1965.18.60 ‘Kolar. S and Ramabrabramy,V, ft J. Bu, Maram Handbook of Biochemistry and Molecalur Biology. Proteins I. (E.G. D. Fusmanl. CRC Pres. 976 Lest Bloch 197 AT 27 Stjer Lia “Bochemisiry, WH. Fresman, San 1955 ehnings. A. Ln ‘Bochemisy 2 Ed. Worth New York “Alas of Protein Sequence and Structure. (Ed. MO. Dayal} Nation Biomedical Research Foundation, Washington, B.C wwe Burnabss.J, Goodman, M.and Moors,G. WJ. Mal Bi 197, 18