Download Importance of main-chain hydrophobic free energy to the stability of

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Ancestral sequence reconstruction wikipedia , lookup

Magnesium transporter wikipedia , lookup

Expression vector wikipedia , lookup

SR protein wikipedia , lookup

Metabolism wikipedia , lookup

Protein wikipedia , lookup

Protein purification wikipedia , lookup

Protein structure prediction wikipedia , lookup

QPNC-PAGE wikipedia , lookup

Interactome wikipedia , lookup

Biochemistry wikipedia , lookup

Western blot wikipedia , lookup

Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup

Metalloprotein wikipedia , lookup

Two-hybrid screening wikipedia , lookup

Protein–protein interaction wikipedia , lookup

Proteolysis wikipedia , lookup

Transcript
International Journal of Biological Macromolecules 35 (2005) 211–220
Importance of main-chain hydrophobic free energy to the stability of
thermophilic proteins
K. Saraboji a , M. Michael Gromiha b , M.N. Ponnuswamy a,∗
b
a Department of Crystallography and Biophysics, University of Madras, Guindy Campus, Chennai 600025, India
Computational Biology Research Center (CBRC), National Institute of Advanced Industrial Science and Technology (AIST),
AIST Tokyo Waterfront Bio-IT Research Building, 2-42 Aomi, Koto-ku, Tokyo 135-0064, Japan
Received 1 December 2004; received in revised form 26 January 2005; accepted 15 February 2005
Abstract
Living organisms are found in the most unexpected places, including deep-sea vents at 100 ◦ C and several hundred bars pressure, in hot
springs. Needless to say, the proteins found in thermophilic species are much more stable than their mesophilic counterparts. There are no
obvious reasons to say that one would be more stable than others. Even examination of the amino acids and comparison of structural features
of thermophiles with mesophilies cannot bring satisfactory explanation for the thermal stability of such proteins. In order to bring out the
hidden information behind the thermal stabilization of such proteins in terms of energy factors and their combinations, analysis were made on
good resolution structures of thermophilic and their mesophilic homologous from 23 different families. From the structural coordinates, free
energy contributions due to hydrophobic, electrostatic, hydrogen bonding, disulfide bonding and van der Waals interactions are computed. In
this analysis, a vast majority of thermophilic proteins adopt slightly lower free energy contribution in each energy terms than its mesophilic
counterparts. The major observation noted from this study is the lower hydrophobic free energy contribution due to carbon atoms and main-chain
nitrogen atoms in all the thermophilic proteins. The possible combination of different free energy terms shows majority of the thermophilic
proteins have lower free energy strategy than their mesophilic homologous. The derived results show that the hydrophobic free energy due to
carbon and nitrogen atoms and such combinations of free energy components play a vital role in the thermostablisation of such proteins.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Protein stability; Hydrophobic free energy; Three-dimensional structure
1. Introduction
Several organisms, mainly archaea live under extreme
environmental temperature conditions. Proteins from thermophilic organisms usually exhibit substantially higher
intrinsic thermal stabilities than their counter parts from
mesophilic organisms. Identifying and understanding the factors contributing to the stability of proteins from organisms
living under extreme conditions stand out to be a longstanding problem. Although the molecular bases of protein
thermostablisation have been the focus of many theoretical
and experimental research efforts, this subject is only partially understood. Studies of thermostability can be divided
into three categories: (i) by examining a single thermophilic
∗
Corresponding author. Tel.: +91 44 22351367; fax: +91 44 22300122.
E-mail address: [email protected] (M.N. Ponnuswamy).
0141-8130/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijbiomac.2005.02.003
protein and comparing its structure at atomic level with one
or more mesophilic homologues, (ii) systematic approach
on the analysis based on sequence and structural information
for a group of proteins in order to reach general conclusions
and (iii) large scale comparison between thermophilic and
mesophilic genome sequences. A number of examples can
be quoted for the comparison of structures of mesophilic homologues but systematic studies are very limited. The recent
progress in genome sequence projects enables one to make
a comparative study of these thermophilic and mesophilic
organisms [1].
There has been a growing interest in understanding
the mechanism of stabilization of thermophilic proteins
from these organisms. Understanding the physiochemical
principles of thermostability will, no doubt, aid in the
comprehension of protein folding and protein interaction
mechanisms. Theoretical and experimental approaches
212
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
have been undertaken to examine the stability of proteins.
Comparison of the sequences and tertiary structures of
homologous proteins from thermophiles, mesophilies and
thermophobes has formed the basis of theoretical efforts
[2,3]. Indeed, one review revealed many different physical
and chemical reasons such as hydrogen bonding, hydrophobic packing, secondary structure propensity and helix dipole
stabilization wherein the researchers reported the enhanced
thermostablisation [4].
In recent years, several works have been carried out theoretically and experimentally to trace the secrets of thermostablisation through mutation studies [5–9] and also based
on the analysis of amino acid composition. Fukuchi and
Nishikawa [10] showed the amino acid composition on protein surface and interior of thermophilic and mesophilic bacteria. They observed the reduction in the number of charged
residues and rich in polar residues in mesophilic bacteria
and concluded that the bias of amino acid composition of
thermophilic protein is due to the residues on protein surfaces, which may be due to extreme environment. Akke and
Forsen [11] showed that the electrostatic interactions between
charges on the surface of a protein could have significant effects on protein stability.
With regard to helix stabilizing factors and stabilization of
thermophilic proteins, Facchiano et al. [12] made the analysis
on 13 thermophilic proteins and showed that the helices of
thermophilic proteins are more stable than mesophilic homologues. Gromiha et al. [13] studied the relationship between
stability changes caused by buried mutations and changes
in 48 amino acid properties; this provides the correlation of
hydrophobicity with the stability of proteins.
The intramolecular interactions, namely hydrophobic,
electrostatic, van der Waals and hydrogen bonds play an important role in the stability of protein structures [14–16,24].
Several investigations have been carried out to understand
the mechanism for the thermostability of proteins. Gromiha
et al. [17] made a comparative analysis on the relation between thermostability and amino acid properties for a family of meso and thermophilic proteins wherein the Gibbs
free energy change of hydration and shape play a dominant
role in thermostability of proteins. The mutational study by
Hasegawa et al. [18] agreed with the results of Gromiha et
al.; they analysed the increased stability of mesophilic cytochrome c through five substitutions and observed that the
−GhN may contribute to the stability.
Szilagyi and Zavodsky [19] made a systematic study on
25 protein families consisting of 64 mesophilic and 29 thermophilic proteins and concluded that different protein families adapt to higher temperatures utilizing different sets of
structural devices and the number of ion pairs increased with
the increase in growth temperature. Querol et al. [4] found the
relationship between thermal stability and conformational
characteristics of proteins. The thermostability of 16 different
families of mesophilic and thermophilic proteins has been
examined by Vogt et al. [20] and a good correlation evinced
between the thermostability of the familial members and the
number of hydrogen bonds, as well in the fractional polar surface. The statistical analysis on 18 families of thermophilic
and mesophilic proteins by Kumar et al. [21] showed
the increase of the salt bridges and side-chain–side-chain
hydrogen bonds in majority of the thermophilic proteins;
the occurrence of residues Arg and Tyr are more frequent in
thermophilic proteins. Kumar and Nussinov [22] made the
analysis on fluctuations, ion pair contributions and stabilities
in NMR conformer ensembles and found that the overall
stabilizing contribution of ion pair is conformer population
dependent. Recently, Gromiha [23] analyzed the medium and
long-range contacts in mesophilic and thermophilic proteins
of 16 different families and explained the fact that thermophiles prefer to have contacts between residues through
hydrogen bonds; apart from hydrophobic contacts and also
between polar and non-polar residues in thermophiles than
mesophilies. Ponnuswamy and Gromiha [24] made the
investigations on the conformational stability of folded
proteins where the hydrophobic force drives the polypeptide
chain to the folded state overcoming the entropic factor,
while the other factors, especially hydrogen bonds and van
del Waals attraction, define the shape and keep it from falling
apart.
Recently, Yano and Poulos [25] compiled the factors that
are reported to be important for increased protein stability. It
has been mentioned that electrostatic interactions, cation–pi
interactions, aromatic and hydrophobic interactions and other
factors would enhance the stability [26,27]. From this diverse
collection of studies, it is difficult to come to a general conclusion about the structural features underlying the increased
thermal stability of proteins from thermophilic microorganism. The contradictions and the limited understanding are
the consequences of the limited data available and the nonuniform approach of the contributing researchers. Though the
proteins can be engineered or engineer themselves in vivo to
achieve greater stability by utilizing one or more of these
strategies, it is clear that no single and preferred mode has
yet to be established.
The aim of present work is to combine the different free
energy components of a set of thermophilic and mesophilic
proteins to assess the contributions from different stability
factors into a unified model. We compute the major free
energy components of hydrophobic, electrostatic, hydrogen
bonding, van der Waals and disulfide bonding interactions
of the folded state of proteins, and also the conformational
entropy of the unfolded state of the corresponding proteins.
Here an in depth statistical analysis of parameters was made
and investigated the importance of each interaction towards
protein thermostability.
2. Materials and methods
2.1. Data set
Recently, Kumar et al. [21] constructed the data set of
36 thermophilic and mesophilic proteins from 18 different
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
families; Szilagyi and Zavodsky [19] built a data set
representing 25 families and Vogt et al. [20] collected a set
of 56 globular proteins from 16 different families. In this
work, we construct a non-redundant data set of 23 families
of thermophilic protein and its mesophilic counterpart from
the previous studies [19–21]. These families span an entire
spectrum, containing proteins from moderately thermophilic
organisms and their mesophilic homologues. Here we
select one pair from each family. The three-dimensional
structures of all these proteins have been taken from Protein
Data Bank [28]. For a given protein, the PDB files contain
coordinates for the structure observed in a crystallographic
asymmetric unit. This may not reflect the true biochemically
relevant oligomeric state. We choose only one subunit
from subunits having identical amino acid sequences and
mutations are ignored. Also the structurally most similar
thermophile–mesophile pair having the best resolution was
chosen, so that the observed differences can be expected
to be mostly due to thermostability. The PDB codes for
all the proteins along with the resolution, average living
temperature (TL ) and rms deviation for each family are given
in Table 1. It has been reported that the average living temperature has a direct relationship with melting temperature
of proteins [17,29]. As the number of samples for TL is more
than Tm , and TL is widely used to understand the stability of
thermophilic proteins, we have used TL for the present study.
2.2. Hydrophobic free energy (HFE)
The hydrophobic free energy (Ghy ) of each protein was
evaluated by using the method of Eisenberg and Mc Lachlan
[30]. In this method, the change in free energy for transfer of
an amino acid residue to water is given by
GR = i σi Ai
(1)
where the sum is taken over all atoms i, Ai are the accessible surface areas, σ i atomic solvation parameters for the
five classes of atoms namely, carbon, neutral nitrogen and
oxygen, charged nitrogen, charged oxygen and sulfur which
are determined by a least-squares fit of Eq. (1) based on the
method of Ponnuswamy and Gromiha [24]. The σ values
are C: 12.02, N/O: −5.86, N+ : −19.46, O− : −34.98, and S:
35.51 cal/(mol Å2 ). These atomic solvation parameters perform better and explain the protein stability than other values
available in the literature [31].
The hydrophobic free energy of folded protein was expressed as
Ghy = i σi [Ai (folded) − Ai (unfolded)]
(2)
where Ai (folded) and Ai (unfolded) represent, respectively,
the accessible surface areas (ASA) of each atom in the folded
and unfolded states of the protein. The accessible surface areas of each atom in the folded state were computed using
the program NACCESS [32]. According to Shrake and Rupley [33], the ASA of amino acid residue X in extended
213
state is computed using the average of ASA of residues
present in the sequence Gly–X–Gly. The hydrophobic free
energy of each protein was calculated separately for the five
classes of atoms and due to side-chain–main-chain atoms
contributions.
2.3. Electrostatic free energy (EFE)
In our present approach, the method of Ponnuswamy and
Gromiha [24] was adopted to compute the contribution from
electrostatic free energy (Gel ). The actual electrostatic free
energy of the folded state is taken as the sum of energy due
to ion pairs [E1 ] and charge helix dipoles [E2 ]
Gel = E1 + E2
(3)
The ion pairs were defined using a simple distance criterion; two oppositely charged residues were considered as an
ion pair if their closest oppositely charged atom were closer
to each other than a predefined cutoff distance. The cutoff distance was chosen as ≤4 Å [34]. Each ion pair on the surface
of the protein is responsible for stability by about 1 kcal/mol
[2,35] whereas such ion pair is buried to contribute around
3 kcal/mol [36]. In combining these experimental results, we
follow the expression for ion pairs as
E1 = 3Nbi + 1Nex
(4)
where Nbi and Nex are the number of buried and exposed ion
pairs of a protein.
The charge helix dipole interactions are obtained by finding the positive charge residue within 4 Å distance from the C
cap–1 and negative charge residue with in 4 Å distance from
the N cap+1 [37]. Based on site directed mutational studies
[37,38], the charge helix dipole interactions could contribute
to the stability of the protein by about 1.6 kcal/mol and hence,
we compute the charge helix dipole interactions in a protein
with expression
E2 = 1.6Nch
(5)
where Nch is the number of charge helix dipole interactions
of a protein.
2.4. Hydrogen bond free energy (HBFE)
The hydrogen bond is one of the most important interatomic interactions in protein folding. The hydrogen-bonding
free energy (Ghb ) has been computed from the information
about the number of hydrogen bonds in a protein. The interactions that qualify as hydrogen bonds must be between
the listed donor and acceptor atoms, and have acceptable geometries [39]. The number of hydrogen bonds present for all
proteins was calculated by using the HBPLUS routine [39,40]
with the following default parameters (D refers to the donor
atom; A, the acceptor; H, the hydrogen atom; and AA the
atom covalently bound to A): maximum distance for D–A,
3.9 Å and for H–A, 2.5 Å; minimum angle for D–H. . .A and
214
S. no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
a
b
Protein family
Thermophilic organism, PDB id and
resolution (Å)
No. of residues
and TL (◦ C)a
Mesophilic organism PDB id and
resolution (Å)
No. of residues
and TL (◦ C)
rms Deviation
(Å)b
Citrate synthase
Malate dehydrogenase
Rubredoxin
Cyclodextrin glucanotransferase
Pyrococcus furiosus: 1AJ8-1.9
Thermus flavus: 1BDM-2.5
Pyrococcus furiosus: 1CAA-1.8
Thermoanaerobacterium thermosulfurigenes:
1CIU-2.3
Thermus aquaticus: 1EFT-2.5
Pyrococcus furiosus: 1GTM-2.2
Bacillus stearothermophilus: 1LDN-2.5
Bacillus thermoproteolyticus: 1LNF-1.7
Bacillus stearothermophilus: 1PHP-1.65
Thermus thermophilus: 1TFE-1.7
Thermotoga maritime: 1TMY-1.9
Pyrococcus furiosus: 1XGS-1.75
Thermomyces lanuginosus: 1YNA-1.55
Bacillus stearothermophilus: 1ZIN-1.65
Bacillus thermoprotelyticus: 2FXB-2.3
Thermus thermophilus: 2PRD-2.0
Thermus thermophilus: 3MDS-1.8
Bacillus stearothermophilus: 3PFK-2.4
Bacillus stearothermophilus: 1EBD-2.6
Thermoactinomyces vulgaris: 1THM-1.37
Thermotoga maritime: 1HDG-2.5
376; 100
332; 72.5
53; 100
683; 60
Chicken Heart: 1CSH – 1.6
Porcine: 4MDH – 2.5
Desulfovibrio vulgaris: 8RXN-1.0
Bacillus circulans: 1CDG-2.0
435; 37
333; 37
52; 35.5
686; 35
1.68
0.94
0.69
0.7
405; 71
419; 77.5
316; 52.5
316; 52.5
394; 52.5
145; 72.5
118; 90
295; 100
193; 50
217; 52.5
81; 52.5
174; 72.5
203; 72.5
319; 52.5
455; 52.5
279; 60
332; 82.5
E.Coli: 1EFU(C)-2.5
Clostridium symbiosum: 1HRD- 1.96
Plasmodium falciparum: 1LDG –1.74
Bacillus cereus: 1NPC-2.0
Saccharomyces cerevisiae: 1QPG-2.4
E.coli: 1EFU(B)-2.5
E.coli: 3CHY-1.66
E.coli: 1MAT-2.4
Bacillus circulans: 1XNB-1.49
Sacchromyces cerevisiae 1AKY –1.63
Clostridium acidurici: 1FCA-1.8
E.coli: 1INO-2.2
Homo sapiens: 1QNM-2.3
E.coli: 2PFK-2.4
Pseudomonas putida: 1LVL-2.45
Bacillus amyloliquifaciens: 1SUP-1.60
E.coli: 1GAD-1.80
385; 37
449; 33.5
316; 37
317; 30
415; 27.5
282; 37
128; 37
263; 37
185; 35
218; 27.5
55; 28
175; 37
198; 37
300; 37
458; 27.5
275; 35
330; 37
1.5
1.38
1.25
0.86
1.28
1.24
1.39
1.39
1.14
1.22
1.27
1.10
1.17
0.87
0.88
1.9
1.24
Bacillus stearothermophilus: 1BTM-2.8
Clostridium thermocellum 1XYZ-1.4
251; 52.5
320; 60
Homo sapiens: 1HTI-2.8
Cellulomonas fimi: 2EXO-1.8
248; 37
312; 30
1.24
1.38
EF–TU and EF–TU–TS complex
Glutamate dehydrogenase
Lactate dehydrogenase
Thermolysin and neutral protease
3-Phosphoglycerate kinase (PKG)
EF–TS and EF–TU–TS complex
Che Y
Methionine aminopeptidase
Endo-1,4-b Xylanase
Adenylate kinase
Ferredoxin
Inorganic pyrophosphatase
Manganese superoxide dismutase
Phosphofructokinase
Reductase
Subtillisin
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
Triose phosphate isomerase
Glycosyltransferase B (␤-glycanase)
TL , the average living temperature.
rms deviation, the root mean square deviation for C␣ between the two protein structures of a family.
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
Table 1
Data set used in this study showing 23-protein families and its thermophilic and mesophilic counterparts
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
for D. . .A–AA, 90.0◦ . Baker and Hubbard [41] have recommended these values after extensive analysis.
However, the interaction between the charged residues has
already been considered as ion pairs, the possible number of
hydrogen bonds between the residues has to be excluded from
the total number of hydrogen bonds in a protein. Accordingly,
the actual number of hydrogen bonds to be included in free
energy computation is
NHB = Nhb − (Nbi + Nex )
(6)
It has been reported that the free energy due to hydrogen bond
is approximately 1 kcal/mol [42], and hence, the Ghb is taken
to be
Ghb = 1NHB
(7)
2.5. Disulfide bond free energy (SSFE)
After analyzing the characteristics of disulfide bonds in
a set of proteins, Thornton [43] suggested an approximated
value of 2.3 kcal/mol, a probable free energy contribution
to a disulfide bond (Gss ). We use this value to compute the
free energy contribution from disulfide bonds. The number of
disulfide bonds present in a protein was calculated by using
the program HBPLUS [39] with the distance criteria S–S,
3.0 Å.
2.6. van der Waals free energy (VDWFE)
van der Waals interactions are calculated between the
atoms separated by at least three bonds (1–4 interactions).
van der Waals free energy (Gvw ) was calculated using the
sum of Lennard–Jones potentials over all 1–4 interactions
with AMBER [44] library files.
215
We have normalized the hydrophobic and all other free
energy terms by number of amino acid residues in each protein.
3. Results and discussion
The values of various average free energy contributions
calculated in this study are given in Table 2.
3.1. Hydrophobic free energy
The hydrophobic free energy contributions for each protein molecule was calculated according to different atom
types such as C, N/O, N+ , O− , S and due to the main- and sidechain atoms. The conclusion is that most of the thermophilic
proteins having lower free energy than its mesophilic parts
except the sulphur atom case. In this spectrum, main- and
side-chain carbon atoms and the main-chain nitrogen atoms
consistently show low energy state in all the thermophilic
proteins.
The plot of the average hydrophobic free energy due to
the main- and side-chain carbon atoms is shown in Fig. 1.
The hydrophobic free energy due to the main-chain nitrogen atoms is illustrated in Fig. 2. An interesting feature observed between the thermophilic and mesophilic families is
that all the 23 thermophilic families showing lower free energy than its mesophilic pair. The roles of hydrophobic profile
of the carbon and main-chain nitrogen atoms help to categorize the thermophilic and mesophilic and thus the thermal
stability. Further, the data presented in Table 2 showed that
the difference in hydrophobic free energy is significantly high
for the families, Methionine aminopeptidase, Ferredoxin and
Glycosyltransferase. From the analysis on the structures of
Fig. 1. Contributions of side- and main-chain carbon atoms to hydrophobic free energy.
216
No.
Average free energy [kcal/mol]
Thermophilic proteins
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
a
Mesophilic proteins
Difference (thermophilic–mesophilic)
PDB
HFE
EFE
HBFE
SSFE
Total
PDB
HFE
EFE
HBFE
SSFE
Total
HFE
EFE
HBFE
SSFE
Total
1AJ8
1BDM
1CAA
1CIU
1EFT
1GTM
1LDN
1LNF
1PHP
1TFE
1TMY
1XGS
1YNA
1ZIN
2FXB
2PRD
3MDS
3PFK
1EBD
1THM
1HDG
1BTM
1XYZ
−0.659
−0.643
−0.579
−0.688
−0.62
−0.671
−0.652
−0.604
−0.642
−0.538
−0.696
−0.704
−0.612
−0.584
−0.625
−0.629
−0.691
−0.621
−0.613
−0.637
−0.599
−0.598
−0.685
−0.047
−0.037
−0.057
−0.033
−0.017
−0.059
−0.013
−0.034
−0.036
−0.028
−0.039
−0.04
−0.031
−0.049
−0.025
−0.045
−0.054
−0.011
−0.057
−0.024
−0.038
−0.027
−0.053
−0.912
−0.858
−0.679
−0.958
−0.852
−1.021
−0.835
−1.003
−0.876
−0.917
−0.856
−0.847
−1.005
−0.912
−0.506
−0.799
−0.887
−0.912
−0.868
−0.982
−0.991
−0.912
−1
0
0
0
0
0
0
0
0
0
−0.016
0
0
−0.012
0
0
0
0
0
−0.005
0
0
0
0
−1.618
−1.538
−1.315
−1.679
−1.489
−1.751
−1.5
−1.641
−1.554
−1.499
−1.591
−1.591
−1.66
−1.545
−1.156
−1.473
−1.632
−1.544
−1.543
−1.643
−1.628
−1.537
−1.738
1CSH
4MDH
8RXN
1CDG
1EFU
1HRD
1LDG
1NPC
1QPG
1EFU
3CHY
1MAT
1XNB
1AKY
1FCA
1INO
1QNM
2PFK
1LVL
1SUP
1GAD
1HTI
2EXO
−0.651
−0.626
−0.583
−0.66
−0.595
−0.654
−0.661
−0.564
−0.651
−0.506
−0.702
−0.611
−0.631
−0.55
−0.543
−0.623
−0.65
−0.638
−0.593
−0.639
−0.598
−0.596
−0.575
−0.03
−0.025
0
−0.028
−0.06
−0.032
−0.044
−0.026
−0.048
−0.078
−0.053
−0.014
−0.011
−0.045
0
−0.017
−0.047
−0.028
−0.025
−0.019
−0.036
−0.02
−0.022
−0.966
−0.805
−0.635
−0.926
−0.794
−0.958
−0.968
−0.997
−0.822
−0.957
−0.93
−0.859
−0.962
−0.839
−0.491
−0.697
−0.874
−0.93
−0.841
−1.076
−0.939
−0.863
−0.99
0
0
0
−0.003
0
0
0
0
0
0
0
0
0
0
0
0
0
0
−0.005
0
0
0
−0.023
−1.647
−1.456
−1.218
−1.617
−1.449
−1.644
−1.673
−1.587
−1.521
−1.541
−1.685
−1.484
−1.604
−1.434
−1.034
−1.337
−1.571
−1.596
−1.464
−1.734
−1.574
−1.479
−1.61
−0.008
−0.017
0.004
−0.028
−0.025
−0.017
0.009
−0.04
0.009
−0.032
0.006
−0.093
0.019
−0.034
−0.082
−0.006
−0.041
0.017
−0.02
0.002
−0.001
−0.002
−0.11
−0.017
−0.012
−0.057
−0.005
0.043
−0.027
0.031
−0.008
0.012
0.05
0.014
−0.026
−0.02
−0.004
−0.025
−0.028
−0.007
0.017
−0.032
−0.005
−0.002
−0.007
−0.031
0.054
−0.053
−0.044
−0.032
−0.058
−0.063
0.133
−0.006
−0.054
0.04
0.074
0.012
−0.043
−0.073
−0.015
−0.102
−0.013
0.018
−0.027
0.094
−0.052
−0.049
−0.01
0
0
0
0.003
0
0
0
0
0
−0.016
0
0
−0.012
0
0
0
0
0
0
0
0
0
0.023
0.029
−0.082
−0.097
−0.062
−0.04
−0.107
0.173
−0.054
−0.033
0.042
0.094
−0.107
−0.056
−0.111
−0.122
−0.136
−0.061
0.052
−0.079
0.091
−0.054
−0.058
−0.128
PDB, protein data bank code; HFE, hydrophobic free energy; EFE, electrostatic free energy; HBFE, hydrogen bond free energy; SSFE, disulphide bond free energy.
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
Table 2
Contribution of free energies to the thermophilic and its mesophilic counterpartsa
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
217
Fig. 2. Contributions of main-chain nitrogen atoms to hydrophobic free energy.
mesophilic and thermophilic proteins in these families, we
observed that the main- and side-chain carbon atoms tend
to move interior to the protein in thermophiles. Hence, the
burial of carbon atoms to the interior of thermphilic proteins
influenced the subtle difference in hydrophobic free energy.
Fig. 3 shows the average hydrophobic free energy of each
family where 16 thermophilic and 7 mesophilic proteins having lower free energies. This study reveals that most of the
thermally stable proteins possess lower hydrophobic free energies than its mesophilic counterpart.
3.2. Electrostatic free energy
The number of ion pairs using a distance limit of 4 Å was
calculated wherein the thermophilic proteins show more ion
pairs than mesophilic ones, an agreement with the results
of Szilagyi and Zavodsky [19]. The normalized electrostatic
free energy arising out of ion pairs and charge helix dipoles
are lower for the 17 thermophilic families (Fig. 4). The
exceptions of six mesophilic families are in lower energy
state because of its excess helical content and thus the charge
helix dipoles contribute such lower electrostatic free energy.
3.3. Hydrogen bond free energy
The component of normalized hydrogen bond free energy
is observed to be dominant as that of the hydrophobic term
in all cases. In 16 thermophilic families, the hydrogen bond
free energy is lower than its mesophilic component. Although
seven mesophilic proteins possess lower energy values, difference is minimal when compared to its thermophilic part.
3.4. Free energy due to disulfide bonds and 1–4 van der
Waals interactions
In our data set, most of the proteins do not possess disulfide bonds in both thermophilic and mesophilic parts. Out of
23 sets of families, three thermophilic and two mesophilic
Fig. 3. Normalized hydrophobic free energy contribution for each family.
218
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
Fig. 4. Electrostatic free energy due to ion pairs and charge helix dipoles per residue.
proteins having one disulfide bond and one mesophilic protein contains two disulfide bonds. A study of the average
disulfide bond free energy shows two thermophilic and two
mesophilic proteins have lower energies. Only one family has
the both thermophilic and its mesophilic part having disulfide
bond for which the average disulfide bond free energy is same.
The free energies due to the 1–4 van der Waals interactions show similar trend as in other cases. The average 1–4
van der Waals free energies was lower for 17 thermophilic and
6 mesophilic proteins. It is noticed that the energy due 1–4
van der Waals free energy is linearly related to the number
of residues and the correlation of 1–4 van der Waals free energy with number of residues for thermophilic and mesophilic
proteins is found to be 0.9703 and 0.9794, respectively.
3.5. Combination of energy terms
The calculated energy terms are combined together with
different combinations. Since the VDWFE values are relatively low with respect to the other four-energy terms and are
directly proportional to the total number of residues, the contribution due to VDWFE is excluded and other energy terms
are combined.
3.5.1. Two-energy factor
The possible combination of two-energy factors implies
the lower energy state of thermophilic families in most of the
cases. Table 3 shows the lower energy state in each family
for the different possible combinations of two-energy terms.
Almost in each combination 70% of lower energy state falls
in the region of thermophilic proteins.
While considering all the six possible combinations, the
following 11 thermophilic proteins having lower energetic
contribution than their mesophilic counterparts: Malate
dehydrogenase, Cyclodextrin glucanotransferase, Glutamate
dehydrogenase, Thermolysin and neutral protease, Adenylate
kinase, Ferredoxin, Inorganic pyrophosphatase, Manganese
Table 3
Comparative table showing possible combination of two free energy components for all 23 set of familiesa
Family
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Combination of energies
HFE +
EFE
EFE +
HBFE
HBFE +
SSFE
HFE +
SSFE
HFE +
HBFE
EFE +
SSFE
T
T
T
T
T
T
M
T
M
M
M
T
T
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
M
T
T
T
T
T
M
T
M
T
T
M
T
T
M
T
T
T
M
T
M
T
M
T
M
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
T
T
T
T
M
T
M
T
M
M
M
T
T
T
T
T
T
M
T
T
T
T
T
a T represents thermophilic proteins have lower energy state and M represents mesophilic proteins have lower energy state.
superoxide dismutase, Reductase, Glyceraldehyde-3phosphate dehydrogenase and Triose phosphate isomerase.
Only three mesophilic proteins seem to have lower energy
state in all the six combinations.
3.5.2. Three-energy factor
When energetic combinations are made with three-energy
factors, the above trend of higher stability seems to be
in reality in thermophilic families. Table 4 illustrates the
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
219
Table 4
Comparative table showing possible combination of three and all four free energy components for all 23 set of familiesa
Family
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
a
Combination of energies
HFE + EFE + HBFE
HFE + EFE + SSFE
HFE + HBFE + SSFE
EFE + HBFE + SSFE
HFE + EFE + HBFE + SSFE
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
T
T
T
T
T
T
M
T
M
M
M
T
T
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
M
T
T
T
T
T
M
T
T
M
M
T
T
T
T
T
T
M
T
M
T
T
T
See Table 3 Footnotes.
three-term combinations and in each combination 74% of
the thermophilic proteins are in lower energy states.
While considering the four possible combinations, in
addition to the families showing higher stability in the twoenergy factor case, the thermophilic proteins from other five
families viz., Rubredoxin, EF–TU and EF–TU–TS complex,
Methionine aminopeptidase, Endo-1, 4-b Xylanase and
Glycosyltransferase B also show higher stability in all combinations. Along with the three mesophilic proteins showing
lower energy state in two-term combination, here the other
two mesophilic proteins are observed as in lower energy state.
3.5.3. Four-energy factor
In the case of combination using all the four-energy terms,
the tendency as in three-energy term combinations is maintained in both thermophilic and mesophilic proteins.
3.6. Comparison with other studies
Vogt et al. [20] reported that an increase in hydrogen
bonding and fractional polar surface increase the stability
of thermophilic proteins. Xiao and Honig [45] found that
electrostatic interactions are more favorable in thermophiles;
Kumar et al. [22] showed that salt bridges and side chain-side
chain hydrogen bonds increase the stability in most of the
thermophilic proteins. Recently, Yano and Poulos [25]
compiled the factors that are reported to be important
for increased protein stability. It has been mentioned that
electrostatic interactions, cation–pi interactions, aromatic
and hydrophobic interactions, etc. would enhance the
stability [26,27]. In this work, we found that the hydrophobic
free energy due to main-chain carbon and nitrogen atoms
increased the stability of thermophilic proteins.
4. Conclusions
The experimentally determined three-dimensional structures of thermostable proteins is still small in number and
the lack of structural information may lead the unfriendliness of choosing the mesophilic counterpart from the same
family of thermophilic protein where the structure is known.
Because of these difficulties, there is a barrier in extensive
statistical surveys, even though a large number of families
have been examined in the present work; the data set chosen
is comparatively larger than the previous efforts [17,20,21].
The comparative analysis on the thermophilic and their
mesophilic counterparts through the free energy tool revealed
important factors for the stability of thermophilic proteins.
We found that the hydrophobic free energy of carbon and
main-chain nitrogen atoms play an important role in thermostablization and combinations of different free energies
are lower for thermophilic cases than its mesophilic counter
parts.
Acknowledgement
K.S. acknowledges the Council of Scientific and Industrial
Research (CSIR), Govt. of India, for the award of Senior
Research Fellowship.
220
K. Saraboji et al. / International Journal of Biological Macromolecules 35 (2005) 211–220
References
[1] S. Chakravarty, R. Varadarajan, Biochemistry 41 (2002) 8152.
[2] M. Perutz, H. Radidt, Nature 255 (1975) 256.
[3] P. Argos, M. Rossmann, U. Grau, H. Zuber, G. Frank, J. Tratschin,
Biochemistry 25 (1979) 5698.
[4] E. Querol, J.A. Perez-Pons, A. Mozo-Villarias, Protein Eng. 9 (1996)
265.
[5] M.A. Arnott, R.A. Michael, C.R. Thompson, D.W. Hough, M.J. Danson, J. Mol. Biol. 304 (2000) 657.
[6] J. Chen, Z. Lu, J. Sakon, W.E. Stites, J. Mol. Biol. 303 (2000)
125.
[7] J. Hasegawa, S. Uchiyama, Y. Tanimoto, M. Mizutani, Y. Kobayashi,
Y. Sambongi, Y. Igarashi, J. Biol. Chem. 275 (2000) 37824.
[8] R.R. Biekofsky, S.R. Martin, J.E. McCormick, L. Masino, S. Fefeu,
P.M. Bayley, J. Feeney, Biochemistry 41 (2002) 6850.
[9] H. Takeshita, T. Yasuda, T. Nakajima, K. Mogi, Y. Kaneko, R. Iida,
K. Kishi, Eur. J. Biochem. 270 (2003) 307.
[10] S. Fukuchi, K. Nishikawa, J. Mol. Biol. 309 (2001) 835.
[11] M. Akke, S. Forsen, Proteins 8 (1990) 23.
[12] A.M. Facchiano, G. Colonna, R. Ragone, Protein Eng. 11 (1998)
753.
[13] M.M. Gromiha, M. Oobatake, H. Kono, H. Uedaira, A. Sarai, J.
Prot. Chem. 18 (1999) 565.
[14] K.A. Dill, Biochemistry 29 (1990) 7133.
[15] G.D. Rose, R. Wolfenden, Ann. Rev. Biophys. Biomol. Struct. 22
(1993) 381.
[16] C.N. Pace, B.A. Shirely, M.A. McNutt, K. Gajiwala, FASEB J. 10
(1996) 75.
[17] M.M. Gromiha, M. Oobatake, A. Sari, Biophys. Chem. 82 (1999)
51.
[18] J. Hasegawa, S. Uchiyama, Y. Tanimoto, M. Mizutani, Y. Kobayashi,
Y. Sambongi, Y. Igarashi, J. Biol. Chem. 275 (2000) 37824.
[19] A. Szilagyi, P. Zavodsky, Struct. Fold Des. 8 (2000) 493.
[20] G. Vogt, S. Woell, P. Argos, J. Mol. Biol. 269 (1997) 631.
[21] S. Kumar, C.J. Tsai, R. Nussinov, Protein Eng. 13 (2000) 179.
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
S. Kumar, R. Nussinov, Proteins 43 (2001) 433.
M.M. Gromiha, Biophys. Chem. 91 (2001) 71.
P.K. Ponnuswamy, M.M. Gromiha, J. Theor. Biol. 166 (1994) 63.
J.K. Yano, T.L. Poulos, Curr. Opin. Biotechnol. 14 (2003) 360.
M.M. Gromiha, S. Thomas, C. Santhosh, Prep. Biochem. Biotechnol.
32 (2002) 355.
A. Paiardini, G. Gianese, F. Bossa, S. Pascarella, Proteins 50 (2003)
122.
H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H.
Weissig, I.N. Shindyalov, P.E. Bourne, Nucleic Acids Res. 28 (2000)
235.
E.A. Gaucher, J.M. Thomson, M.F. Burgan, S.A. Benner, Nature 425
(2003) 285.
D. Eisenberg, A.D. Mc Lachlan, Nature 319 (1986) 199.
D. Eisenberg, M. Wesson, M. Yamashita, Chem. Scr. 29A (1989)
217.
S.J. Hubbard, J.M. Thornton, ‘NACCESS’, Computer Program, Department of Biochemistry and Molecular Biology, University College, London, 1993, available from:
http://wolf.bi.umist.ac.uk/unix/naccess.html.
A. Shrake, J. Rupley, J. Mol. Biol. 79 (1973) 351.
D.J. Barlow, J.M. Thornton, J. Mol. Biol. 168 (1983) 867.
L.R. Brown, A. DeMarco, R. Richarz, G. Wagner, K. Wuthrich, Eur.
J. Biochem. 88 (1978) 87.
A.R. Fersht, J. Mol. Biol. 64 (1972) 497.
H. Nicholson, W.J. Becktel, B.W. Matthews, Nature 336 (1988) 651.
D. Sali, M. Bycroft, A.R. Fersht, Nature 335 (1988) 740.
I.K. Mc Donald, J.M. Thornton, J. Mol. Biol. 238 (1994) 777.
I.K. Mc Donald HBPLUS, London, UCL, 1992.
E. Baker, R. Hubbard, Prog. Biophys. Mol. Biol. 25 (1994) 97.
A. Ben-Naim, J. Phys. Chem. 95 (1991) 1437.
J.M. Thornton, J. Mol. Biol. 151 (1981) 261.
W.D. Cornell, P. Cieplak, C.I. Bayly, I.R. Gould, K.M. Merz, D.C.
Ferguson, T. Spellmeyer, J.W. Fox, Caldwell, P.A. Kollman, J. Am.
Chem. Soc. 117 (1995) 5179.
L. Xiao, B. Honig, J. Mol. Biol. 289 (1999) 1435.