Left-handed βαβ-units: frequency of occurrence and arrangement in protein structure

Abstract

It is well known that the overwhelming majority of the ???-units are right-handed in proteins. In this study, we have analyzed more than 1000 nonhomologous ?/?- and (?+?)-proteins and domains and found 63 left-handed ???- units. It is shown that each domain contains only one left-handed ???-unit. Most of them (78%) are located at the C-ends of the ?/?- and (?+?)-domains. Their arrangement in 3D structures of different nonhomologous domains is very similar. The other 22% are found in the ????-combinations in which the left-handed ???-unit follows the right-turned ?-module along the polypeptide chain. Possible mechanism of formation of the left-handed ???-units is also discussed.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tbsd20
Journal of Biomolecular Structure and Dynamics
ISSN: 0739-1102 (Print) 1538-0254 (Online) Journal homepage: https://www.tandfonline.com/loi/tbsd20
Left-handed βαβ-units: frequency of occurrence
and arrangement in protein structure
Anton M. Kargatov & Alexander V. Efimov
To cite this article: Anton M. Kargatov & Alexander V. Efimov (2019): Left-handed βαβ-units:
frequency of occurrence and arrangement in protein structure, Journal of Biomolecular Structure
and Dynamics, DOI: 10.1080/07391102.2019.1591306
To link to this article: https://doi.org/10.1080/07391102.2019.1591306
Accepted author version posted online: 07
Mar 2019.
Published online: 02 Apr 2019.
Submit your article to this journal
Article views: 3
View Crossmark data

LETTER TO THE EDITOR
Left-handed bab-units: frequency of occurrence and arrangement in
protein structure
Anton M. Kargatov and Alexander V. Efimov
Institute of Protein Research, Russian Academy of Sciences, Moscow Region, Russian Federation
Communicated by Ramaswamy H. Sarma
ARTICLE HISTORY Received 26 February 2019; Accepted 1 March 2019
Introduction
bab-Units are widespread in proteins and appear to act as
‘building blocks’of the so-called Rossmann’s folds (Rao &
Rossmann, 1973). They exist predominantly as right-handed
bab-superhelices in which the b-strands form a parallel
b-sheet and the a-helix is packed in the other layer. The
predominance of the right-handed form of the bab-units is
determined by several reasons (Rao & Rossmann, 1973;
Richardson, 1976; Sternberg & Thornton, 1976), however,
finally it is a result of the homochirality of L-amino acid resi-
dues in proteins. Nevertheless, there is a number of a/b-
and (aþb)-proteins and domains which have the left-
handed bab-units. They are not to be confused with the
abb- and bba-units described by Kajava (1992) since each
of the latter has the left-handed topology but is formed by
the antiparallel b-hairpin and the a-helix. In our previous
works (Efimov, 2017,2019; Kargatov & Efimov, 2018), we
have shown the relationship between mutual arrangement
of structural units in protein structure and their handedness.
In this study, we have analyzed more than 1000 nonhomol-
ogous a/b- and (aþb)-proteins and domains and found 63
left-handed bab- units. It is shown that each domain con-
tains only one left-handed bab-unit. Most of them (78%) are
located at the C-ends of the a/b- and (aþb)-domains. Their
arrangement in 3D structures of different nonhomologous
domains is very similar. The other 22% are found in the
Gbab-combinations in which the left-handed bab-unit fol-
lows the right-turned G-module along the polypeptide
chain. Possible mechanism of formation of the left-handed
bab-units is also discussed.
Materials and methods
The database of the left-handed bab-units has been com-
piled using Protein Data Bank (http://www.rcsb.org/pdb) and
the updated version of PCBOST (Gordeev, Kargatov, &
Efimov, 2010; available at http://strees.protres.ru) that
includes the updated databases of a/b- and (aþb)-domains
(Gordeev & Efimov, 2013). In total, our database includes 63
nonhomologous proteins and domains having the left-
handed bab-units (Table 1). Possible homologies were
revealed by the Blast 2 Sequences program (Tatusova &
Madden, 1999). In this work, for protein-based homology
search, the threshold used was an E-value 1e 4(¼10
4
).
Protein structures were visually examined using RasMol
molecular graphics program (Sayle & Milner-White, 1995) and
structural trees of proteins available at http://strees.protres.
ru. The secondary structure assignment has been done using
the RasMol molecular graphics program as well as by visual
inspection taking into account the u,wvalues and the
hydrogen bonding system of a protein molecule.
Results
Figure 1 shows a schematic representation of overall folds of
all the a/b- and (aþb)-proteins and domains listed in
Table 1. All the structures are oriented in a similar way so
that their N-terminal parts are located on the right and the
C-ends on the left.
Stereochemical analysis of these folds results in findings
as follows:
1. There is only one left-handed bab-unit in each domain.
2. There are three types of left-handed bab-units: (i) simple
bab-units (the left and middle columns); (ii) split
bab-units in left-handed abCd-units (the right column);
(iii) bab-units in Pbab-combinations in which the left-
handed bab-unit follows the right-turned P-module (the
bottom row, the right column; see also Kargatov &
Efimov, 2018).
3. Most often (50%), the left-handed bab-units occur in
a/b-proteins and domains containing 7S a/b-motifs
(Efimov, 1997; Gordeev & Efimov, 2013). The 7S (seven-
segment) a/b-motif is a structural motif composed of
four b-strands and three a-helices which are arranged so
that the b-strands form a parallel b-sheet and the a-heli-
ces are packed on both sides of the b-sheet. All the
bab-units in these motifs are right-handed (in Figure 1,
CONTACT Alexander V. Efimov efimov@protres.ru Institute of Protein Research, Russian Academy of Sciences, 142290 Pushchino, Moscow Region,
Russian Federation
ß2019 Informa UK Limited, trading as Taylor & Francis Group
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS
https://doi.org/10.1080/07391102.2019.1591306

they are highlighted in gray; see the middle column,
three upper rows and the right column, the upper row).
4. All the left-handed bab-units of our database are located
at C-ends of a/b- and (aþb)-domains and in the C-ter-
minal parts of the Pbab-combinations.
5. All the left-handed bab-units are arranged in a similar
way so that their a-helices are packed into the
‘bottom’a-helical layer of the three-layered structures
or into the a-helical layer of the two-layered
Gbab-combinations. The G-module consists of con-
nected elements of the b-strand-loop-b-strand type in
which the b-strands do not form H-bonds between
each other and arranged so that the overall fold of
the G-module resembles a clip or the Greek letter G
(they are highlighted in gray in Figure 1;seethebot-
tom row, the right column).
Table 1. Database of nonhomologous left-handed bab-units.
No. PDB (chain) Name of protein Position of domain Position of left-handed bab-unit
01. 1D6T (A) Ribonuclease P 22–110 44–87
02. 1DAR (A) Elongation factor G 490–597 522–568
03. 1DHR (A) Dihydropteridine reductase 7–230 171–230
04. 1E5K (A) MobA 6–171 93–127
05. 1E6U (A) GDP-fucose synthetase 3–246 156–246
06. 1EA5 (A) Acetylcholinesterase 24–506 417–506
07. 1ENO (A) EACPR 19–292 226–292
08. 1EZI (A) Acylneuraminate cytidylyltransferase 3–202 97–136
09. 1FO8 (A) N-acetylglucosaminyltransferase I 109–275 205–242
10. 1FWK (A) Homoserine kinase 5–114 45–87
11. 1FX2 (A) Receptor-type adenylate cyclase GRESAG 4.1 899–1104 1020–1067
12. 1FXO (A) Glucose-1-phosphate thymidylyltransferase 2–217 102–139
13. 1GA8 (A) Galactosyltransferase LgtC 1–204 98–130
14. 1H41 (A) Alpha-glucuronidase 31–149 110–149
15. 1H7E (A) CMP-KDO 2–217 90–130
16. 1I24 (A) Sulfolipid biosynthesis protein SQD1 2–294 202–294
17. 1I52 (A) CDP-ME synthetase 7–207 100–137
18. 1IM4 (A) Dbh bypass polymerase 2–167 140–167
19. 1J5E (I) Ribosomal protein S9 3–89 25–67
20. 1JTV (A) 20-Alpha-hydroxysteroid dehydrogenase 2–255 177–255
21. 1JU3 (A) Cocaine esterase 5–341 277–341
22. 1JV1 (A) UDP-N-acetylglucosamine pyrophosphorylase 102–398 245–282
23. 1K47 (A) Phosphomevalonate kinase 1–123 43–93
24. 1KBZ (A) dTDP-glucose oxidoreductase 1–215 145–210
25. 1KWS (A) Glucuronosyltransferase I 75–297 187–216
26. 1LL2 (A) Glycogenin-1 3–184 96–125
27. 1MG7 (A) Early switch protein XOL-1 45–148 86–123
28. 1NNS (A) L-asparaginase II 2–190 108–151
29. 1O7O (A) Alpha-1,3-galactosyltransferase 128–334 219–247
30. 1OMZ (A) Alpha-GalNAcT EXTL2 66–266 145–178
31. 1OYS (A) Ribonuclease PH 27–148 57–119
32. 1P5H (A) Formyl-coenzyme A transferase 8–200 119–200
33. 1PKP (A) Ribosomal protein S5 82–147 104–128
34. 1QBB (A) Chitobiase 236–335 297–335
35. 1QG8 (A) SpsA protein 3–208 91–126
36. 1QGV (A) Spliceosomal protein U5-15kD 6–132 87–132
37. 1QV9 (A) Mtd 3–184 114–184
38. 1QYC (A) Phenylcoumaran benzylic ether reductase PT1 6–215 147–215
39. 1RHY (A) IGPD 4–90 33–64
40. 1RHY (A) IGPD 97–186 120–161
41. 1RRE (A) ATP-dependent protease La 595–694 623–668
42. 1S4N (A) Glycolipid 2-alpha-mannosyltransferase 122–381 241–274
43. 1UEK (A) CDP-ME kinase 1–110 47–83
44. 1W58 (1) Cell division protein FtsZ homolog 1 39–338 185–257
45. 1XGK (A) Nitrogen metabolite repression regulator NmrA 7–219 144–214
46. 1XHB (A) Polypeptide GalNAc transferase 1 117–346 203–237
47. 1Y1P (A) Aldehyde reductase II 13–274 199–274
48. 1Z6Z (A) Sepiapterin reductase 4–255 187–255
49. 2BKA (A) Tat-interacting protein 30 19–227 160–227
50. 2BO4 (A) Mannosylglycerate synthase 2–211 94–129
51. 2BTO (A) Tubulin btubA 4–377 201–278
52. 2C42 (A) Pyruvate-ferredoxin oxidoreductase 260–392 356–392
53. 2CBI (A) Hyaluronidase 64–176 140–176
54. 2FR1 (A) Erythromycin synthase eryAI 1472–1642 1606–1642
55. 2FYB (A) Beta-1,4-galactosyltransferase 1 176–331 242–269
56. 2I5E (A) Hypothetical protein MM_2497 1–160 87–118
57. 2NN6 (A) Polymyositis/scleroderma autoantigen 1 44–160 75–139
58. 2OI6 (A) Bifunctional protein GlmU 5–216 97–133
59. 2VEO (A) Lipase A 52–399 354–399
60. 3BRE (A) Probable two-component response regulator 202–338 295–338
61. 3D4J (A) Diphosphomevalonate decarboxylase 9–141 63–117
62. 3OH8 (A) Nucleoside-diphosphate sugar epimerase 148–365 297–365
63. 4EI7 (A) Plasmid replication protein RepX 17–361 178–258
2 A. M. KARGATOV AND A. V. EFIMOV

Discussion
In our database (http://strees.protres.ru), there are 1301 proteins
and domains (among them 388 are nonhomologous) contain-
ing five-segment (5S) a/b-motifs, 870 proteins and domains
(294 are nonhomologous) containing seven-segment (7S)
a/b-motifs, 926 proteins and domains (401 are nonhomologous)
containing abCd-units (they are considered as representatives of
(aþb)-proteins; see also Gordeev & Efimov, 2013), and 146
nonhomologous proteins having different combinations of
bab-units and G-modules (Kargatov & Efimov, 2018)(among
them 13 proteins contain 14 left-handed bab-units). As a rule,
each protein or domain contains from 3 to 5–7bab-units. The
overwhelming majority of them are right-handed and only 63
bab-units found in nonhomologous proteins of our database
are left-handed. The main result of this work is that most left-
handed bab-units are located at the C-ends of a/b-or
(aþb)-domains (78%) and the other 22% of the units are
found in the Gbab-combinations in which the left-handed
bab-unit follows the right-turned G-module (they can also be
classified as (aþb)-domains). It can be concluded that forma-
tion of the left-handed bab-unit depends on its arrangement in
protein structure. Possible reasons for this are discussed below.
In accordance with our hypothesis, 5S and 7S a/b-motifs
as well as abCd-units can act as nuclei or ‘ready-made’build-
ing blocks in protein folding (Efimov, 1997). Alternatively, the
structural motifs may be regarded as the starting structures
in protein modeling. In proteins, these structural motifs tend
to be located at the edges of two- or three-layered struc-
tures with additional a-helices and/or b-strands arranged on
one definite side of each motif (see Figure 1). The lager pro-
tein structures can be obtained by stepwise addition of
a-helices and/or b-strands to the starting structural motif tak-
ing into account a restricted set of rules inferred from known
principles of protein structure. Thus, it looks like the starting
motif grows preferably in one definite direction (Efimov,
1997), from the starting structural motif (at the right part of
each domain in Figure 1) to the C-end (on the left). At the
Figure 1. Schematic representation of overall folds of proteins and domains having left-handed bab-units. The structures are viewed end-on with a-helices shown
as circles and b-strands as rectangles. The near connections are shown by double lines and the far ones by single lines. The left-handed bab-units are highlighted
in black. The 5S and 7S a/b-motifs as well as abCd-units are highlighted in gray. N and C are N- and C-ends. On the right of each structure, PDB-entries of proteins
containing them are shown. For additional information, see Table 1.
Figure 2. Possible pathways of formation of the left-handed bab-units at the C-terminal parts of two a/b-domains that occur most often in proteins (see Figure
1). N and C are the N- and C-ends. Concave shaded lines show the corners formed by the ‘bottom’a-helical layer and the b-sheet.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 3

final stages of domain growing the left-handed bab-unit can
be formed. Possible mechanisms of its formation are shown
in Figure 2.
It seems likely that its a-helix is arranged first into the
‘bottom’a-helical layer of a domain. Such packing of the
a-helix closes the overall domain fold into a cycle giving rise
to a more stable and cooperative structure (Efimov, 2010).
On the other hand, such packing results in that accessible
surface area (ASA) of both, the a-helix and a corner formed
by the ‘bottom’a-helical layer and the b-sheet (shown by
concave shaded lines in Figure 2) is decreased and that is
favorable. Such arrangement of the a-helix is more favorable
if the surfaces of the corner and a-helix are hydrophobic.
Many a/b-domains are finished growing at this stage (see
the corresponding structural trees at http://strees.protres.ru
and in Efimov (1997), Gordeev and Efimov, (2013)). If a
domain has an additional b-strand joined to the a-helix, it is
packed into the b-sheet that results in formation of the left-
handed bab-unit (Figure 2). The other structures that contain
the left-handed bab-units are the Gbab-combinations in
which the left-handed bab-unit follows the right-turned
G-module. In the Gbab-combinations, the right-turned
G-module clips together the b-strands of the left-handed
bab-unit making the obtained closed structure more stable
and cooperative. Combinations of this kind and features of
their amino acid sequences have been studied in our previ-
ous paper (Kargatov & Efimov, 2018).
Funding
This work was supported by the Russian Foundation for
Basic Research (Project No. 17-04-00242).
Disclosure statement
No potential conflict of interest was reported by the authors.
References
Gordeev,A.B.,Kargatov,A.M.,&Efimov,A.V.(2010).PCBOST:Proteinclassi-
fication based on structural trees. Biochemical and Biophysical Research
Communications,397(3), 470–471. doi:10.1016/j.bbrc.2010.05.136
Gordeev, A. B., & Efimov, A. V. (2013). Modeling of folds and folding
pathways for some protein families of (aþb)- and (a/b)-classes.
Journal of Biomolecular Structure and Dynamics,31(1), 4–16. doi:
10.1080/07391102.2012.691341
Efimov, A. V. (1997). Structural trees for protein superfamilies. Proteins:
Structure, Function, and Genetics,28(2), 241–260. doi:10.1002/
(SICI)1097-0134(199706)28:2<241::AID-PROT12>3.0.CO;2-I
Efimov, A. V. (2010). Structural motifs are closed into cycles in proteins.
Biochemical and Biophysical Research Communications,399(3),
412–415. doi:10.1016/j.bbrc.2010.07.089
Efimov, A. V. (2017). Structural motifs in which b-strands are clipped
together with the G-like module. Proteins: Structure, Function, and
Bioinformatics,85(10), 1925–1930. doi:10.1002/prot.25346
Efimov, A. V. (2019). Handedness of structural units depends on their
mutual arrangement in protein structure. Journal of Biomolecular
Structure and Dynamics. doi:10.1080/07391102.2019.15.80220
Kajava, A. V. (1992). Left-handed topology of super-secondary structure
formed by aligned a-helix and b-hairpin. FEBS Letters,302(1), 8–10.
doi:10.1016/0014-5793(92)80271-H
Kargatov, A. M., & Efimov, A. V. (2018). Unique combinations of bab-units
and G-like modules in proteins and features of their amino acid
sequences. Molecular Biology, (Moscow),52(1), 36–41. doi:10.1134/
S0026893318010077
Rao, S. T., & Rossmann, M. G. (1973). Comparison of super-secondary
structures in proteins. Journal of Molecular Biology,76(2), 241–256.
doi:10.1016/0022-2836(73)90388-4
Richardson, J. S. (1976). Handedness of crossover connections in beta
sheets. Proceedings of the National Academy of Sciences of the United
States of America,73(8), 2619–2623.
Sayle, R. A., & Milner-White, E. J. (1995). RASMOL: Biomolecular graphics
for all. Trends in Biochemical Sciences,20(9), 374–376.
Sternberg, M. J. E., & Thornton, J. M. (1976). On the conformation of pro-
teins: The handedness of the beta-strand-alpha-helix-beta-strand unit.
Journal of Molecular Biology,105(3), 367–382.
Tatusova, T. A., & Madden, T. L. (1999). Blast 2 sequences –A new tool
for comparing protein and nucleotide sequences. FEMS Microbiology
Letters,174(2), 247–250. doi:10.1111/j.1574-6968.1999.tb13575.x
4 A. M. KARGATOV AND A. V. EFIMOV