Divergent Evolution of ( )8-Barrel Enzymes
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Biol. Chem., Vol. 382, pp. 1315 – 1320, September 2001 · Copyright © by Walter de Gruyter · Berlin · New York
Minireview
Divergent Evolution of ()8-Barrel Enzymes
Martina Henn-Sax1,2, Birte Höcker1, Matthias lationship), superfamilies (probable common evolution-
Wilmanns3 and Reinhard Sterner1,* ary origin) and folds (Murzin et al., 1995). Proteins are de-
1 Institut fined to have a common fold if they have the same major
für Biochemie, Universität zu Köln, Otto-
secondary structures in the same spatial arrangement
Fischer-Strasse 12-14, D-50674 Köln, Germany
2 and the same topological connections. SCOP currently
Abteilung Molekulare Genetik und Präparative
distinguishes between several hundred different folds.
Molekularbiologie, Institut für Mikrobiologie und
The (βα)8- (or TIM-)barrel is a frequently encountered
Genetik, Georg-August-Universität Göttingen,
fold, which comprises about 10% of all known protein
Grisebachstr. 8, D-37077 Göttingen, Germany
3 structures (Gerlt, 2000). (βα)8-barrel enzymes catalyse a
European Molecular Laboratory (EMBL) Hamburg
vast range of different reactions, functioning as either ox-
Outstation, c/o DESY, Notkestrasse 85, D-22603
idoreductases, transferases, hydrolases, lyases or iso-
Hamburg, Germany
merases (Pujadas and Palau, 1999; Wierenga, 2001). The
(βα)8-barrel consists of a core of eight twisted parallel β-
* Corresponding author
strands, the β-barrel, which are connected by eight α-he-
lices which form the outer layer of the structure (Figure 1).
The ()8-barrel is the most versatile and most fre- In all known (βα)8-barrel enzymes, the active site residues
quently encountered fold among enzymes. It is an are located at the C-terminal face of the β-barrel and
interesting question how the contemporary ()8- within the loops that connect the β-strands with the sub-
barrels are evolutionarily related and by which mech- sequent α-helices. Many (βα)8-barrel enzymes contain
anisms they evolved from more simple precursors. extensions to this canonical topology, either at the N- or
Comprehensive comparisons of amino acid se- C-termini of the sequence, or in loop segments (Pujadas
quences and three-dimensional structures suggest and Palau, 1999).
that a large fraction of the known ()8-barrels have The evolution of the versatile (βα)8-barrel fold has been
divergently evolved from a common ancestor. The
mutational interconversion of enzymatic activities of
several ()8-barrels further supports their common
evolutionary origin. Moreover, the high structural
similarity between the N- and C-terminal ()4 units of
two ()8-barrel enzymes from histidine biosynthesis
indicates that the contemporary proteins evolved by
tandem duplication and fusion of the gene of an an-
cestral ‘half-barrel’ precursor. In support of this hy-
pothesis, recombinantly produced ‘half-barrels’ were
shown to be folded, dimeric proteins.
Key words: Directed evolution / Enzyme fold / Histidine
biosynthesis / Protein families / TIM-barrel / Tryptophan
biosynthesis.
Introduction
The rapidly growing number of determined amino acid
sequences and three-dimensional structures allows to
investigate how the efficient and specific enzymes of Fig. 1 The (βα)8-Barrel Fold.
modern metabolic pathways are evolutionarily related, (A) View onto the C-terminal ends of the eight β-strands, which
form a cylindrical parallel β-sheet. This β-barrel is surrounded by
and how they developed from less efficient and specific
the eight α-helices.
precursors (Gerlt and Babbitt, 2001). The Structural Clas- (B) Topologic diagram of the eight (βα) modules. The active site
sification of Proteins (SCOP) database divides proteins residues of all known (βα)8-barrel enzymes are located at the
according to their amino acid sequence, structural and C-terminal ends of the β-strands and in the loops that connect
functional similarities into families (clear evolutionary re- the β-strands with the subsequent α-helices.1316 M. Henn-Sax et al.
discussed for many years, and arguments in favour of ei- (βα)8-barrels with significant sequence similarity were de-
ther convergent evolution to a stable fold or, alternatively, tected. All of them contain a characteristic glycine-rich
divergent evolution from a common ancestral barrel have phosphate-binding motif at the C-terminal end of β-
been put forward (Lesk et al., 1989; Farber and Petsko, strand 7, which was identified earlier by analysing small-
1990; Raine et al., 1994; Reardon and Farber, 1995). Here er sets of amino acid sequences (Bork et al., 1995) or
we summarize recent sequence and structure analyses, three-dimensional structures (Wilmanns et al., 1991). The
as well as protein engineering studies, which provided detected enzymes include (βα)8-barrels from tryptophan
new insights into the evolution of (βα)8-barrels. The re- biosynthesis, triosephosphate isomerase and fructose
sults suggest that the members of several important su- bisphosphate aldolases. Since there are eight equivalent
perfamilies and probably a large fraction of all known β-strands in (βα)8-barrels, the identical location of the
(βα)8-barrels have a common evolutionary origin (Altami- phosphate-binding site of all of these enzymes at the C-
rano et al., 2000; Babbitt and Gerlt, 2000; Copley and terminal end of β-strand 7 suggests that this superfamiliy
Bork, 2000; Jürgens et al., 2000). Moreover, evidence for evolved from a common ancestor.
the evolution of the (βα)8-barrel fold from ancestral ‘half- Divergent evolution is further supported by the com-
barrel’ precursors will be presented (Thoma et al., 1998; bined structural and functional analyses of phosphate-
Lang et al., 2000; Höcker et al., 2001). binding (βα)8-barrels from histidine and tryptophan
biosynthesis. HisA and imidazole glycerol phosphate
synthase (HisF) catalyse two successive reactions
The Phosphate-Binding Superfamiliy in histidine biosynthesis. As a consequence, both en-
zymes bind the common ligand N’-[(5’-phosphoribulo-
The program PSI-Blast (Altschul et al., 1997) was used to syl)formimino]-5-aminoimidazole-4-carboxamide-ribonu-
perform iterative rounds of sequence comparisons be- cleotide (PRFAR), which is the product of HisA and the
tween (βα)8-barrels (Copley and Bork, 2000). The search substrate of HisF (Figure 2A). The crystal structures of
was started with N’-[(5’-phosphoribosyl)formimino]-5- HisA and HisF from Thermotoga maritima were deter-
aminoimidazole-4-carboxamide-ribonucleotide (ProFAR) mined at high resolution (Lang et al., 2000). The superpo-
isomerase of histidine biosynthesis (HisA), which con- sition of their backbone atoms yielded a root mean square
tains two phosphate-binding sites. A number of other (rms) deviation of only 1.79 Å (Figure 2B). Furthermore,
Fig. 2 Two Sequential Reactions of Histidine and Tryptophan Biosynthesis with Similar Chemistries are Catalysed by Related (βα)8-
Barrels.
(A) HisA and TrpF catalyse Amadori rearrangements of aminoaldoses into aminoketoses. HisF and TrpC catalyse the closure of the im-
idazole and indole ring to yield ImGP and IGP, respectively. The second product of the HisF reaction, AICAR, is further used in de novo
purine biosynthesis. To form ImGP and AICAR, HisF uses nascent ammonia produced by the glutaminase HisH (Beismann-Driemeyer
and Sterner, 2001).
(B) Experimental evidence for evolutionary relatedness of HisF, HisA, TrpF and TrpC. The residual HisA activity of HisF, and the interconver-
sions of HisA and TrpC into TrpF are indicated by broken arrows. The rms deviations resulting from pairwise structural superpositions using
all main chain, non-hydrogen atoms are shown. The percentages of identical residues in the corresponding structure-based alignments are
given in brackets. The calculations were carried out with the program ALIGN-PDB (Cohen, 1997). Abbreviations: AICAR: 5-aminoimidazole-
4-carboxamide ribotide; CdRP: 1-(o-carboxyphenylamino)-1-deoxyribulose 5-phosphate; HisA: ProFAR isomerase; HisF: synthase subunit
of ImGP synthase; HisH: glutaminase subunit of ImGP synthase; IGP: indole glycerol phosphate; ImGP: imidazole glycerol phosphate; PRA:
phosphoribosyl anthranilate; PRFAR: N’-[(5’-phosphoribulosyl)formimino]- 5-aminoimidazole-4-carboxamide-ribonucleotide; ProFAR: N’-
[(5’-phosphoribosyl)formimino] -5-aminoimidazole-4-carboxamide-ribonucleotide; TrpC: IGP synthase; TrpF: PRA isomerase.Enzyme Evolution 1317 Fig. 3 Synopsis of a Four-Fold Structure-Based Sequence Alignment of the Amino-Terminal (HisA-N, HisF-N) and the Carboxy-Ter- minal Halves (HisA-C and HisF-C) of HisA and HisF. The arrows represent β-strands, the cylinders α-helices. β-strands and α-helices 1 – 4 correspond to HisA-N and HisF-N, β-strands and α-helices 5 – 8 correspond to HisA-C and HisF-C. β-strands 1’ and 5’, and α-helices 2’, 4’, 6’, 8’ are extensions to the limit (βα)8-barrel fold (Lang et al., 2000). Invariant residues are shown in upper case letters and residues that are identical in three of the four sequences in lower case letters. The invariant aspartate residues (D) that are essential for catalysis are located at the C-terminal ends of β-strands 1 and 5. the two aspartate residues of HisA and HisF that are im- tionary network that links the (βα)8-barrels HisA, HisF, portant for catalysis are located at equivalent positions at TrpF and TrpC (Figure 2B). the C-terminal ends of β-strands 1 and 5 (Figure 3). There- fore, both enzymes were tested for their mutual residual activities. Whereas HisA does not show detectable HisF Evolution of the Enolase Superfamily activity, HisF catalyses the HisA reaction, albeit with low efficiency (Lang et al., 2000). The sequence similarities between the members of the In analogy to HisA and HisF, phosphoribosyl anthrani- enolase superfamily are often low and their substrates late (PRA) isomerase (TrpF) and indole glycerol phos- are chemically quite diverse. However, their similar phate synthase (TrpC) are (βα)8-barrel containing en- three-dimensional structures and catalytic mechanisms zymes that catalyse two successive reactions in the indicate a common evolutionary origin (Babbitt and biosynthesis of tryptophan. They bind the common lig- Gerlt, 2000). Members of the enolase superfamiliy con- and 1-(o-carboxyphenylamino)-1-deoxyribulose 5-phos- sist of two domains: a larger (βα)7β barrel domain, which phate (CdRP), which is the product of TrpF and the sub- is a modified version of the (βα)8-barrel, and a mixed α/β strate of TrpC (Figure 2A). The rms deviation of the domain that is formed by the N- and C-terminal parts of backbone atoms of TrpF and TrpC from Escherichia coli the sequence (Babbitt et al., 1996). The mixed α/β do- is 2.03 Å (Figure 2B). In an attempt to establish TrpF ac- main is an important determinant of the substrate speci- tivity on the scaffold of TrpC, Altamirano et al. (2000) used ficity and caps the barrel domain at the C-terminal ends a combination of rational design and directed evolution. of the β-strands, where the residues that are essential In a first step, the N-terminal extension to the limit (βα)8- for catalysis are located. The mechanistic similarity with- barrel of TrpC was removed and several loops at the C- in the superfamiliy is the abstraction of the α-proton of a terminal end of the β-barrel were designed to be similar to carboxylate anion substrate, which is assisted by elec- the corresponding loops in TrpF. Subsequently, random trostatic stabilization of the resulting enolate intermedi- mutagenesis and selection in a trpF deficiency strain ate by a metal ion. The enolate intermediate is then con- were performed. With this approach a TrpC variant with verted to different products via different mechanisms in high TrpF activity but lacking TrpC activity was isolated. the various active sites. However, the chemical nature The establishment of TrpF activity on the TrpC scaffold and the location of the residues essential for catalysis and the residual HisA activity of HisF indicate a common within the barrel are highly conserved. Within the eno- evolutionary origin of (βα)8-barrel enzymes within histi- lase superfamily, therefore, new enzymatic activities ob- dine and tryptophan biosynthesis. viously evolved by retaining a crucial step in the catalyt- Both HisA and TrpF catalyse mechanistically similar re- ic mechanism while changing substrate specificity. An actions, namely Amadori rearrangements of an aminoal- interesting member of the enolase superfamily is an en- dose into an aminoketose (Figure 2A), and the superposi- zyme from Amycolaptosis sp., which acts both as N-acyl tion of their backbone atoms yielded an rms deviation of amino acid racemase and as o-succinylbenzoate syn- 2.55 Å (Figure 2B). There is strong experimental evidence thase (Palmer et al., 1999). These two reactions are con- for a close inter-pathway relationship between these en- siderably different with regard to the substrate and the zymes. Using random mutagenesis and selection in a overall chemical mechanism. The recently solved struc- trpF-deficiency strain, HisA variants were generated that ture of o-succinylbenzoate synthase from Escherichia catalysed the TrpF reaction. Moreover, one of these vari- coli shows that most interactions between the bound ants retained significant HisA activity (Jürgens et al., product and the active site are either indirect via water 2000). A closer analysis revealed that a single amino acid molecules or via hydrophobic interactions (Thompson et exchange in the active site region was sufficient to inter- al., 2000). It was speculated that this plasticity within the convert the substrate specificity from HisA to TrpF, al- active site contributes to the dual substrate specificity of though the enzymes share a sequence identity of only the homologous enzyme from Amycolaptosis sp., which about 10%. might examplify ‘evolution in action’ (Babbitt and Gerlt, Taken together, these experiments suggest an evolu- 2000).
1318 M. Henn-Sax et al.
Evolutionary Links between Superfamilies binding of the same ligand or a particular catalytic activi-
ty. In summary, numerous (βα)8-barrels within and be-
Enolase catalyses the formation of phosphoenolpyruvate tween pathways appear to have a common evolutionary
from 2-phosphoglycerate. However, it does not bind the origin.
phosphate group of the substrate at the end of β-strand
7, in contrast to the phosphate-binding superfamily. Nev-
ertheless, through a sequence family of unknown struc- Evolution of the ()8-Barrel Fold from
ture and function, significant sequence similarities Ancestral ‘Half-Barrels’
between the enolase and the phosphate-binding super-
families were detected (Copley and Bork, 2000). Circular Which of the contemporary (βα)8-barrels is most similar
permutation of the sequence might have occurred in the to the putative common ancestor? The sequences and
course of divergence of these superfamilies, because β- structures of HisA and HisF, which have substrates with
strand 3 of the enolase superfamily aligns with β-strand 5 two phosphate moieties (Figure 2A), show an internal
of the phosphate-binding (βα)8-barrels. This similarity is duplication that is not observed in other (βα)8-barrels
equivalent to a conserved metal-binding residue in β- (Fani et al., 1994; Thoma et al., 1998; Lang et al., 2000).
strand 7 of the enolase superfamily with the conserved It follows that HisA and HisF have retained ancestral fea-
glycines of the phosphate-binding (βα)8-barrels. Pro- tures, which were lost during the evolution of other bar-
bably, the conformations of these sites make them rels. The superposition of the N- and C-terminal halves
favourable locations for ligand-binding residues, even if HisA-N, HisF-N, HisA-C, and HisF-C from Thermotoga
the ligands are different. Further rounds of PSI-Blast
searches that were performed with relaxed inclusion
thresholds detected significant similarities of the phos-
phate-binding superfamily to members of the phospho-
enolpyruvate/ pyruvate superfamily and to additional
(βα)8-barrels, which are involved in leucine and lysine
biosynthesis, and in gluconeogenesis (Copley and Bork,
2000).
Horowitz (1945) speculated that anabolic pathways
evolved by several duplication and diversification events,
starting with the gene encoding the last enzyme of the
contemporary pathway (‘retrograde evolution’). The ap-
peal of this model is that no new ligand binding site has to
be invented upon evolving new enzymatic activities. Al-
though the Horowitz model is probably not correct in a
strict sense (Roy, 1999), (βα)8-barrel enzymes that catal-
yse successive reactions within the same pathway al-
most certainly evolved upon retention of a common lig-
and binding site. However, there are also (βα)8-barrel
enzymes from different pathways that bind the same lig-
and. For example, transaldolase, the α-subunit of trypto-
phan synthase, triosephosphate isomerase and fruc-
tose-1,6-bisphosphate aldolase all catalyse reactions
with D-glyceraldehyde 3-phosphate as one of the prod-
ucts. This may indicate a common ancestor for these en-
zymes with a D-glyceraldehyde 3-phosphate binding
site. Similarly, both enolase and pyruvate kinase bind
phosphoenolpyruvate (PEP) and Mg2+, and neither
shows the standard phosphate-binding motif. Probably,
both enzymes may have arisen from a common PEP and
Mg2+-binding ancestor. Other (βα)8-barrel enzymes, for
example HisA and TrpF, and the members of the enolase
superfamily, use similar mechanisms to catalyse reac-
tions of different substrates (Babbitt and Gerlt, 2000; Jür- Fig. 4 Model for the Evolution of the (βα)8-Barrel Fold.
The first tandem gene duplication and fusion generates two
gens et al., 2000). This suggests that these enzymes are
identical, fused half-barrels that then adjust to form the ancestral
derived from a common ancestor with a broader sub- (βα)8-barrel. Further gene duplications and diversifiactions lead
strate specificity (Jensen, 1976). One has to conclude, to HisA, HisF, and to other contemporary (βα)8-barrels. Adapted
therefore, that (βα)8-barrel enzymes are derived from an- with permission from Lang et al., Science 289, 1546 – 1550.
cestors with similar functions, which can either be the © (2000) American Association for the Advancement of Science.Enzyme Evolution 1319
maritima yielded close overall similarities with rms devi- References
ations of only 1.4 to 2.1 Å (Lang et al., 2000). The result-
ing four-fold structure-based sequence alignment re- Altamirano, M.M., Blackburn, J.M., Aguayo, C., and Fersht, A.R.
vealed a number of conserved residues that are located (2000). Directed evolution of new catalytic activity using the
α/β-barrel scaffold. Nature 403, 617 – 622.
at structurally identical positions in all four half-barrels
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z.,
(Figure 3). Among them are the aspartate residues es- Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-
sential for catalysis, which are located at the ends of β- BLAST: a new generation of protein database search pro-
strands 1 (in HisA-N and HisF-N) and β-strands 5 (in grams. Nucleic Acids Res. 25, 3389 – 3402.
HisA-C and HisF-C). These structural data suggest that Babbitt, P.C., Hasson, M.S., Wedekind. J.E., Palmer, D.R., Bar-
HisA and HisF evolved by tandem duplication and fusion rett, W.C., Reed, G.H., Rayment, I., Ringe, D., Kenyon, G.L.,
of a gene encoding an ancestral half-barrel comprising and Gerlt. J.A. (1996). The enolase superfamily: a general
strategy for enzyme-catalysed abstraction of the alpha-pro-
only four (βα) units. To test this hypothesis, HisF-N and
tons of carboxylic acids. Biochemistry 35, 16489 – 16501.
HisF-C were produced recombinantly in E. coli, purified
Babbitt, P.C., and Gerlt, J.A. (2000). New functions from old
and characterized (Höcker et al., 2001). Separately, scaffolds: how nature reengineers enzymes for new functions.
HisF-N and HisF-C are proteins with well-defined sec- Adv. Protein Chem. 55, 1 – 28.
ondary and tertiary structures, but are catalytically inac- Beismann-Driemeyer, S. and Sterner, R. (2001). Imidazole glyc-
tive. Upon co-expression in vivo or joint refolding in vit- erol phosphate synthase from Thermotoga maritima: quater-
ro, HisF-N and HisF-C assemble to the catalytically fully nary structure, steady-state enzyme kinetics and reaction
active HisF-NC complex. These findings support the no- mechanism of the bi-enzyme complex. J. Biol. Chem. 276,
20387 – 20396.
tion that HisA, HisF and probably a large fraction of the
Bork, P., Gellerich, J., Groth, H., Hooft, R., and Martin, F. (1995).
known (βα)8-barrels evolved from an ancestral half-bar- Divergent evolution of a beta/alpha-barrel subclass: detection
rel (Figure 4). This half-barrel was probably not a of numerous phosphate-binding sites by motif search. Protein
monomeric protein. Instead, in order to shield the hy- Sci. 4, 268 – 274.
drophobic face of the putative half-barrel from the sol- Cohen, G.E. (1997). ALIGN: a program to superimpose protein
vent, it may have formed homo-dimers. Along these coordinates, accounting for insertions and deletions. J. Appl.
lines, isolated HisF-N and HisF-C molecules form Cryst. 30, 1160 – 1161.
Copley, R.R., and Bork, P. (2000). Homology among (βα)8-bar-
oligomers, predominantly of the (HisF-N)2 and (HisF-C)2
rels: implications for the evolution of metabolic pathways. J.
type (Höcker et al., 2001).
Mol. Biol. 303, 627 – 640.
Fani, R., Lio, P., Chiarelli, I., and Bazzicalupo, M. (1994). The evo-
lution of the histidine genes in prokaryotes: a common ances-
Conclusions and Outlook tor for the hisA and hisF genes. J. Mol. Evol. 38, 489 – 495.
Farber, G.K., and Petsko, G.A. (1990). The evolution of
The present post-genomic era offers the unprecedented alpha/beta barrel enzymes. Trends Biochem. Sci. 15, 228 –
opportunity for comprehensive comparisons of protein 234.
Gerlt, J.A. (2000). New wine from old barrels. Nature Struct. Biol.
folds and functions. Such comparisons help to under-
7, 171 – 173.
stand the evolution of enzymes, which is important for Gerlt, J.A., and Babbitt, P.C. (2001). Barrels in Pieces? Nature
both basic and applied research. The presented com- Struct. Biol. 8, 5 – 7.
parisons of both sequences and structures clearly sup- Höcker, B., Beismann-Driemeyer, S., Hettwer S., Lustig, A., and
port a common evolutionary origin of a large fraction of Sterner, R. (2001). Dissection of a (βα)8-barrel enzyme into two
the known (βα)8-barrel enzymes, which are used by na- folded halves. Nature Struct. Biol. 8, 32 – 36.
ture to catalyse numerous and quite diverse biochemical Horowitz, N.H. (1945). On the evolution of biochemical synthe-
sis. Proc. Natl. Acad. Sci. USA 31, 153 – 157.
reactions. Since few amino acid exchanges can be suf-
Jensen, R.A. (1976). Enzyme recruitment in evolution of new
ficient to interconvert the catalytic activities of (βα)8-bar-
function. Annu. Rev. Microbiol. 30, 409 – 425.
rels, this fold appears to be particularly suited for en- Jürgens, C., Strom, A., Wegener, D., Hettwer, S., Wilmanns. M.,
zyme design. Finally, the surprising finding that and Sterner, R. (2000). Directed evolution of a (βα)8-barrel en-
(βα)8-barrels were derived from ‘half-barrels’ will moti- zyme to catalyze related reactions in two different metabolic
vate the search for ancestral domains within other ap- pathways. Proc. Natl. Acad. Sci. USA 97, 9925 – 9930.
parent single-domain protein folds (Petsko, 2000; Gerlt Lang, D., Thoma, R., Henn-Sax, M., Sterner, R., and Wilmanns,
and Babbitt, 2001). M. (2000). Structural evidence for evolution of the β/α barrel
scaffold by gene duplication and fusion. Science 289,
1546 – 1550.
Lesk, A.M., Branden C.I., and Chothia, C. (1989). Structural prin-
Acknowledgements ciples of alpha/beta barrel proteins: the packing of the interior
of the sheet. Proteins 5, 139 – 148.
We thank Professor Kasper Kirschner for discussions and com- Murzin, A.G., Brenner, S.E., Hubbard, T., and Chothia, C. (1995).
ments on the manuscript. Work in the authors’ laboratories is SCOP: A Structural Classification of Proteins database for the
supported by the Deutsche Forschungsgemeinschaft (grants to investigation of sequences and structures. J. Mol. Biol. 247,
R.S. and M.W.) and the Deutsche Bundesstiftung Umwelt (a 536 – 540.
grant to R.S.). M.H.-S. is a fellow of the Hans-Böckler-Stiftung. Palmer, D.R., Garrett, J.B., Sharma, V., Meganathan, R., Babbitt,1320 M. Henn-Sax et al. P.C., and Gerlt, J.A. (1999). Unexpected divergence of en- R. (1998). A histidine gene cluster of the hyperthermophilic zyme function and sequence: ‘N-acylamino acid racemase’ is Thermotoga maritima: sequence analysis and evolutionary o-succinylbenzoate synthase. Biochemistry 38, 4252 – 4258. significance. Extremophiles 2, 379 – 389. Petsko, G.A. (2000). Design by necessity. Nature 403, 606 – 607. Thompson, T.B., Garrett, J.B., Taylor, E.A., Meganathan, R., Pujadas, G., and Palau, J. (1999). TIM barrel fold: structural, Gerlt, J.A., and Rayment, I. (2000). Evolution of enzymatic ac- functional and evolutionary characteristics in natural and de- tivity in the enolase superfamily: structure of o-succinylben- signed molecules. Biologia Bratislava 54, 231 – 254. zoate synthase from Escherichia coli in complex with Mg2+ Raine, A.R., Scrutton, N.S., and Mathews, F.S. (1994). On the and o-succinylbenzoate. Biochemistry 5, 10662 – 10676. evolution of alternate core packing in eightfold beta/alpha- Wierenga, R.K. (2001). The TIM-barrel fold: a versatile frame- barrels. Protein Sci. 3, 1889 – 1892. work for efficient enzymes. FEBS Lett. 492, 193 – 198. Reardon, D., and Farber, G. K. (1995). The structure and evolu- Wilmanns, M., Hyde, C.C., Davies, D.R. Kirschner, K., and Jan- tion of α/β barrel proteins. FASEB J. 9, 497 – 503. sonius, J.N. (1991). Structural conservation in parallel β/α- Roy, S. (1999). Multifunctional enzymes and evolution of biosyn- barrel enzymes that catalyse three sequential reactions in the thetic pathways: retro-evolution by jumps. Proteins 37, 303 – pathway of tryptophan biosynthesis. Biochemistry 30, 309. 9161 – 9169. Thoma, R., Schwander, M., Liebl, W., Kirschner, K., and Sterner,
You can also read
