| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 275, Issue 35, 27129-27136, September 1, 2000
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Biochemistry and Cell Biology and the
W. M. Keck Center for Computational Biology, Rice University,
Houston, Texas 77005-1892
Received for publication, January 20, 2000, and in revised form, June 13, 2000
Apomyoglobins from 13 different mammals were
examined for resistance to denaturation by guanidinium chloride.
Unfolding was followed by circular dichroism and tryptophan
fluorescence and analyzed globally using the two-step, three-state
mechanism first described by Barrick and Baldwin (Barrick, D., and
Baldwin, R. L. (1993) Biochemistry 32, 3790-3796).
With one exception, the rise and fall of Trp fluorescence intensity
correlates quantitatively with the native to intermediate to unfolded
steps seen in the CD curves. Although the O2 binding
properties of the holoproteins are nearly identical, the unfolding
transitions of the apomyoglobins show 600-fold differences in
resistance to guanidinium chloride denaturation. Apomyoglobins from
diving mammals, particularly from sperm whales, are the most stable,
whereas the apoproteins from pig, horse, and sheep are the least
stable, indicating selective pressure for resistance to denaturation in
the whale proteins. Sequence comparisons suggest that the key
stabilizing residues in whale globins are Ala5,
His12, Ile28, Thr51,
Ala53, Ala74, Lys87,
Lys140, and Ile142. Combinations of these
residues were substituted into pig myoglobin. The resultant multiple
mutants showed stabilities approaching that of recombinant sperm whale
apomyoglobin. Thus, comparative mutagenesis can be used to increase
heme protein stability and improve expression yields in bacteria
without compromising function.
Apomyoglobin is a compact three-dimensional unit, maintaining most
of the helical secondary structure of the holoprotein (1). NMR studies
show that native sperm whale apomyoglobin (N) has chemical shifts and
nuclear Overhauser effects similar to holomyoglobin with the
exception of residues 78-106, which form the EF loop, F helix, FG
loop, and the beginning of the G helix (2). Temperature, pH, and
denaturant-induced unfolding of the N state has been characterized structurally by many techniques and proceeds through a molten globule
intermediate containing intact A, G, and H helices, as shown in Fig.
1 (1, 3-8).
The expression of heme proteins in Escherichia coli is
governed by the stability of the apo- and reduced holoprotein, because of competition for heme with bacterial cytochromes and consumption of
O2 by oxidative metabolism. Hargrove et al. (9)
have shown that expression yields for myoglobin correlate with
resistance to both unfolding and heme loss. Tang et al. (10)
showed recently that heme loss from deoxymyoglobin at low pH occurs by
partial unfolding of the protein followed by solvation of the
prosthetic group. Four coordinate heme signals occur after cleavage of
the proximal histidine-iron bond at low pH or in the presence of high concentrations of guanidinium chloride
[GdmCl].1 The pH and
[GdmCl] needed to produce this spectral species correlate inversely
and directly, respectively, with the equilibrium stability of
apomyoglobin for a series of proteins with mutations at positions 64, 29, 68, and 107. Enhancement of apoglobin stability is clearly important for improving the production levels and the shelf-lives of
recombinant heme proteins being developed for pharmaceutical uses
(11).
Rational mutagenesis has produced large changes in apomyoglobin
stability but often at the expense of function (9, 12, 13). Replacement
of the distal histidine with apolar residues produces an apoprotein
that is 10-30 times more resistant to denaturation than the wild type
control (9). However, the increased resistance to denaturation is
achieved at the expense of decreased oxygen binding and increased
autooxidation rates.
An alternative approach is to compare the stabilities of apomyoglobins
from various mammalian species. In general, the functional characteristics of all mammalian myoglobins are very similar (Table I), but their conformational stabilities
may vary substantially. Balestieri et al. (14) first
compared GdmCl denaturation curves for apomyoglobins from cow, buffalo,
and sperm whale and reported significantly different overall
stabilities. Baldwin and co-workers (3, 15) reported that human
apomyoglobin forms the intermediate molten globule state at higher pH
values than sperm whale apoprotein. Hargrove and co-workers (7) also
showed that sperm whale apomyoglobin is significantly more resistant to
GdmCl denaturation than human, horse, and pig apomyoglobins.
The Stabilities of Mammalian Apomyoglobins Vary over a 600-Fold
Range and Can Be Enhanced by Comparative Mutagenesis*
ABSTRACT
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
View larger version (15K):
[in a new window]
Fig. 1.
Apomyoglobin unfolding. The thickness of
the ribbons indicates the amount of helical structure. Native
apomyoglobin (N) retains most of the secondary and tertiary structure
present in holomyoglobin (purple and gold) but
loses structure in the F helix (green). Addition of
denaturant unfolds the B, C, D, and E helices (gold) to give
a molten globule intermediate (I) composed of the A, G, and H helices
(purple). Further addition of denaturant results in a random
coil or unfolded state (U). This scheme was taken from Refs. 5 and
6.
Ligand binding, autooxidation, and heme loss rates for myoglobins from
different species
We have expanded and investigated these species differences more
quantitatively by using both fluorescence emission and circular dichroism to examine the unfolding of 13 mammalian myoglobins varying
in amino acid identity from 99 to 80%. The apomyoglobins were selected
by their availability and evolutionary distances from the well
characterized sperm whale apomyoglobin. Unexpectedly, the overall
equilibrium unfolding constants varied over a 600-fold range.
Comparison of the sequences of the most stable versus the least stable apoproteins suggested residues that stabilize myoglobin but have little effect on oxygen binding. The contributions of these
amino acids to protein stability were examined by constructing single
and multiple mutants in pig myoglobin, the apoprotein that shows
the least resistance to GdmCl denaturation.
| |
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Protein Sources--
Pig myoglobin was expressed as a fusion
protein with the N-terminal 31 amino acids of cII, an efficiently
expressed regulatory protein of 
phage (16). The mutant gene was
cloned into pLcII, transformed into phage-resistant M5219 E. coli cells, and expressed by temperature induction. Purification
of pig myoglobin from inclusion bodies has been described by Smerdon
et al. (17). Pig myoglobin produced by this method is
identical to native protein in sequence, spectral characteristics, and
function (18). The pig mutagenesis and expression plasmids were a gift
from Dr. Anthony Wilkinson (University of York, York, UK). Mutants were
made using either Kunkel (14) or double-stranded (Strategene Chameleon
kit; Ref. 15) mutagenesis methods. A similar system was used to express real wild type sperm whale and human myoglobin. Real wild type sperm
whale and pig myoglobins were expressed and purified according to
Ikeda-Saito et al. (19) and Smerdon et al. (17),
respectively. Purified human myoglobin was a generous gift from Dr.
Masao Ikeda-Saito.
Recombinant sperm whale myoglobins were produced using the synthetic gene constructed by Springer and Sligar (20) for constitutive expression in E. coli. Mutants were made and purified as described in Springer et al. (21) and Carver et al. (22). The complete sequence of each new mutant gene was determined to verify the mutation and the integrity of the remaining portion of the coding region (Sequenase, version 2.0).
Bovine myoglobin was obtained from heart muscle cell lysate and purified according Wittenberg and Wittenberg (23). Sheep, horse heart, dog, and native sperm whale myoglobins were obtained in the lyophilized state from Sigma. Other cetacean myoglobins (dwarf sperm whale, pygmy sperm whale, minke whale, goosebeak whale, porpoise, and dolphin) were a generous gift from the collection of Dr. Frank N. Gurd and Dr. Jay Berzofsky. These proteins were collected from the native organisms prior to the enactment of the endangered species act and purified as referenced in Flanagan et al. (24).
Heme was removed from myoglobins by the methyl ethyl ketone method (25). The resultant apoproteins were dialyzed overnight in phosphate buffer and concentrated to 0.2-0.5 mM as determined by absorbance at 280 nm. High purity guanidinium hydrochloride was obtained from Life Technologies, Inc. GdmCl concentrations were determined by refractive index measurements (26).
Data Collection--
Apomyoglobins were diluted to 5 µM protein in 0.2 M potassium phosphate
buffer, pH 7.0. A Hamilton Microlab 500 automatic titrator was used to
titrate this sample with a solution of 5 µM
apomyoglobin in concentrated GdmCl and buffer to give denaturant concentrations ranging from 0 to 5 M in 0.2 M
increments. Total sample volume was maintained at 1800 µl for the
duration of the titration. The sample in the cuvette was stirred to
assure even mixing and equilibrated for 
80 s between addition of the
denaturant and data collection. Sample temperature was maintained at
25 °C using a refrigerated water bath.
An AVIV 62S DS CD spectrometer with a fluorescence accessory was used to collect total fluorescence emission intensity and CD absorbance values on the same sample at each concentration of GdmCl. A detailed description of a similar apparatus is given by Ramsay et al. (27). Fluorescence excitation was at 285 nm. A 320-nm cut-off filter allowed measurement of total fluorescence emission above this wavelength. Circular dichroism was measured at 222 nm. Shutters were closed between measurements to avoid photobleaching. The path length was 1 cm. To avoid corrections for GdmCl dilution of the protein signal, both the initial sample and the added denaturant contained the same concentration of protein. Measurements were made on mixtures of completely folded (initial sample in buffer) and completely unfolded proteins (titrating denaturant) that were allowed to reach equilibrium. To determine reversibility, a single apoprotein sample was titrated from 0.0 to 5.0 M GdmCl and then titrated back to 0.2 M GdmCl.
The raw data were corrected by subtracting the contributions of GdmCl
to the CD and fluorescence signals. The corrected CD data were then
scaled from 0 to 1 with respect to the initial native form and the
fully unfolded form, which was assumed to occur at 5 M
GdmCl (Equation 1). Fluorescence data (F) were scaled from 0 to 1 with respect to the initial form and the fluorescence maximum for
the intermediate state (Equation 2). The subscripts "obs,"
"initial," "max," and "final" refer to the observed value at a given [GdmCl], the value in the absence of denaturant, the maximum absolute fluorescence, and the CD value at 5 M
GdmCl, respectively.
|
(Eq. 1) |
|
(Eq. 2) |
Data Fitting-- Baldwin and co-workers (5, 28) have proposed a three state-model for apomyoglobin unfolding where KNI and KIU are the equilibrium constants for the transition from the native (N) to intermediate (I) structure and from the intermediate to unfolded (U) structure, respectively (Fig. 1). This unfolding model can be described by Equation 3 (28).
|
(Eq. 3) |
In Equation 3, Yscaled represents the
scaled CD or fluorescence emission changes; mNI
and mIU represent the effect of denaturant on
free energy differences between N and I and between I and U, respectively; and x represents GdmCl concentration. Equation 3 assumes a linear dependence of the free energy change on the
concentration of denaturant for each unfolding process (26, 29). This
model treats I as a single species, which is supported by NMR data, but
does not allow for further gradual loss of residual secondary structure
in the U state that is observed by CD but not by fluorescence. To
account for these additional changes, Equation 3 was modified by the
inclusion of a simple linear term (m3) to fit
the CD data (Equation 4).
|
(Eq. 4) |
The data were fitted using NON-LIN, a program based on nonlinear least
squares optimization (27, 30). Initially all 10 parameters were allowed
to vary. However, the m values were fairly consistent and in
subsequent analyses were fixed at the sperm whale wild type values of
5.2 and 2.8 for mNI and
mIU, respectively. This analysis assumes that
minor sequence variations do not cause significant differences in the
differential binding of GdmCl to the various conformational states.
When the fluorescence and CD curves were fitted simultaneously, most of
the parameters were reasonably well defined, but the intermediate
YI spectral weight for the CD changes in
Equation 3 was difficult to determine and showed greater variance
than expected between proteins and samples. This variability is due to
the concerted nature of the two-step denaturation reactions observed by
CD and the nearly uniform decrease in the CD222 signal for
both the N 
I and the I U transitions.
In principle, the CD change for the intermediate should be defined by the inflection point in the titration curve, but for many apomyoglobins this feature is barely detectable. However, three of the naturally occurring myoglobins (dog, goosebeak whale, and minke whale) show an initial transition at much lower GdmCl concentrations than the second transition (see Fig. 4 and Table III). In these cases, the unfolding reaction is clearly a three-state reaction, as measured by either CD changes or the broad bell-shaped fluorescence curve. For these apomyoglobins, the YI values for the CD changes are well defined, and all three apoproteins give a value of 0.43. This value was used to fix YI when fitting the CD data for all of the other native apoproteins. Fixing YI resulted in better definition of KNI and KIU, and the global fits to both the fluorescence and CD data still gave nearly random residuals (see Fig. 2).
Comparison of Wild Type Forms of Sperm Whale Myoglobin-- The original recombinant wild type sperm whale apomyoglobin constructed by Springer and Sligar (20) is not identical in sequence to native sperm whale apomyoglobin. In fact, three different wild type proteins have been generated and are compared with native myoglobin in Table II. All four proteins have nearly identical spectral, ligand binding, heme loss, and autooxidation properties (9, 31). However, the apomyoglobin stabilities of these functionally equivalent control proteins are different. The chemically identical native and real wild type sperm whale apomyoglobins show the same unfolding constants (Table II).
|
New wild type apomyoglobin, which contains an initiator methionine,
shows a less concerted unfolding reaction, but its overall stability
(1/KNIKIU) is equivalent
to that of native and real sperm whale apomyoglobin. The extra
N-terminal methionine destabilizes the native state of apomyoglobin by
promoting the N I transition but, at the same time, stabilizes the
I state by inhibiting the final I U unfolding process. The original
or "old" wild type apomyoglobin, which contains both the initiator
methionine and a D122N substitution, serves as the genetic background
for all of the sperm whale mutations that were constructed in this
study. This apoprotein has a KNI value roughly
equal to that for new wild type apomyoglobin, but the
KIU value for old wild type apomyoglobin is
2-3-fold larger, indicating that the negative charge on the native
Asp122 residue stabilizes the molten globule intermediate.
Old wild type apomyoglobin is the least stable of the control sperm
whale myoglobins, confirming the previous results of Hargrove et
al. (9).
Ionic Strength, Urea Versus GdmCl, and Stability of the Intermediate-- GdmCl was chosen as a denaturant for five reasons: 1) Both holoprotein and apoprotein unfolding experiments can be carried out reversibly with this reagent, which keeps the heme group in solution at high denaturant concentrations. Thus, the same type of titration experiment can be used to determine quantitatively the contributions of heme dissociation and apomyoglobin unfolding to the overall stability of holomyoglobins (32). 2) Fluorescence increases and decreases are readily measured at room temperature and neutral pH when GdmCl, but not urea, is used as the denaturant (see Figs. 2 and 3, discussion below). 3) Ramsay et al. (27) have shown that the same three-state model can be used to describe combined CD/fluorescence changes associated with either acid or GdmCl induced unfolding of recombinant apomyoglobin. 4) Tang et al. (10) have shown that denaturation of deoxymyoglobin correlates with parameters measured for unfolding the N state of apomyoglobin. There is a direct, inverse correlation between the pH and the [GdmCl] required for unfolding and disruption of the proximal histidine-Fe(II) bond. 5) Stock solutions of GdmCl are readily prepared and stable for 4-5 days, making routine screening of mutants, with or without heme, relatively simple and rapid (26).
In two recent papers Baldwin and co-workers (13, 33) have shown that the molten globule intermediate observed at pH 4.0 is stabilized by anions, perhaps by both simple shielding and preferential binding, as suggested by Goto and co-workers (8, 34). They argued that urea is a better denaturant than GdmCl for examining the unfolding of apomyoglobin because the latter reagent generates extremely high ionic strengths (33). However, the fitted values of mNI and mIU in Equations 3 and 4 should take into account differential anion and cation binding to the N, I, and U states. Thus, in principle, the KNI and KIU values obtained by extrapolation back to [GdmCl] = 0 should be the same as those obtained from similar extrapolations using urea as the denaturant. Pfeil, Bolen, Pace, and others (26, 35-40) have reported examples for which the unfolding equilibrium constants determined using urea and GdmCl are identical. However, they and others have also reported extrapolated values of equilibrium unfolding constants that differ significantly between urea versus GdmCl titrations, often because of specific guanidinium binding to the native or intermediate states (26, 40-44).
We examined this problem directly by comparing GdmCl versus urea induced unfolding of old wild type apomyoglobin in 0.2 M phosphate buffer at pH 7.0, 25 °C. Unfortunately, the fluorescence changes associated with urea induced unfolding near room temperature at pH 7 are small and obscured by background fluorescence and light scattering in samples at high urea concentration. The latter problem appears to be due to both impurities in the reagent and micro-aggregates of the I and U states. Similar difficulties were seen by Kay and Baldwin (12) at pH 7.8, 4 °C.
The CD curve for the urea titration of old wild type apomyoglobin was
of higher quality, showed two transitions, and had a midpoint around
3.5 M urea. These data could be fitted using
KNI and KIU values
obtained from corresponding GdmCl titrations (~0.005 and ~0.005 to
0.010, respectively) if the m values and fraction of CD
change associated with the I state were allowed to vary. The fitted
mNI and mIU values for
urea were 1.64 and 0.85, respectively, whereas those for GdmCl were 5.2 and 2.8, respectively. The fraction of CD change associated with the N
to I transition was higher in the case of urea
(YI 
0.6) than when GdmCl was used as the denaturant (YI = 0.43). Thus, the two sets of
unfolding curves are similar, with respect to the apparent unfolding
constants, the ratio of the mNI to
mIU differential binding parameters, and the
extent of unfolding for each transition. The major difference is the
potency of GdmCl for preferential binding to the unfolded states and a
slight increase in the amount of 
helical character in the I state
induced by GdmCl. The latter differences are reminiscent of the I and
I2 states reported by Baldwin and co-workers (46, 47) in
both kinetic and equilibrium unfolding experiments.
These results are also consistent with the absolute values of
KNI and KIU reported for
wild type apomyoglobin by Kay and Baldwin (12). They reported a value
of KIU 0.005 based on urea induced unfolding
of the I state at pH 4, 4 °C and an overall urea-induced unfolding
constant, KNIKIU, equal
to 0.000038 at pH 7.8, 4 °C. Assuming that the
KIU value also applies at pH 7.8, KNI can be estimated to be ~0.008. Thus, in
both their experiments and ours, there is a roughly equal free energy
contribution from the individual N I and the I U steps to the
overall unfolding process.
Finally, we analyzed combined fluorescence and CD titration curves at three different phosphate concentrations (0.005, 0.020, and 0.200 M) to examine ionic strength effects at 25 °C, pH 7. The overall unfolding process for wild type apomyoglobin appears more concerted at low phosphate concentration. However, since bell-shaped fluorescence changes are still observed, the two unfolding constants can be resolved. The value of KNI shows little dependence on ionic strength, whereas KIU decreases from ~0.015 to ~0.005 when the phosphate concentration is increased from 0.005 to 0.2 M. Thus, in agreement with Baldwin and co-workers (13, 33), there is a small, stabilizing effect of increasing ionic strength or anion concentration on the intermediate. More importantly, these results demonstrate that GdmCl does not mask the anion effects seen when urea is used as the denaturant.
Sequence Alignments and PDBs--
Amino acid sequences were
obtained from Brookhaven's Protein Identification Resource and aligned
using the Wisconsin Package Interface (version 8.0) developed by
Genetics Computer Group.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Overall Stabilities of Mammalian Apomyoglobins--
The CD and
fluorescence unfolding data for all of the naturally occurring globins
can be analyzed globally in terms of the two-step mechanism of Barrick
and Baldwin (28). The correlation between changes in fluorescence and
CD indicate that dissolution of the secondary structure is coincident
with disruption of tertiary structure interactions in the area of the A
helix where the tryptophans are located (Fig.
2). Unfolding and refolding curves are
nearly identical (data not shown), indicating reversibility and
adequate incubation times.
|
The large 600-fold variance in stability of the native apomyoglobins
was initially surprising considering the structural and functional
similarities of the holoproteins (Fig. 3
and Table III). Sperm whale
(Physeter catadon, Kogia simus, and Kogia
breviceps), goosebeak whale, and dolphin apomyoglobins are the
most resistant to GdmCl denaturation. Pig, horse, sheep, and porpoise
apomyoglobins are the least stable, and dog, bovine, human, and minke
whale apomyoglobins have intermediate unfolding constants. The
apomyoglobins of diving mammals are clearly much more stable than
those of terrestrial mammals.
|
|
The globins examined differ from the reference sperm whale protein at 2-30 different positions or 1-20% of the amino acids (Table IV). The substitutions are fairly conservative, such as Ile versus Val (positions 13, 21, 28, and 101), Ala versus Gly (positions 5, 15, 74, 121, and 129), Arg versus Lys (positions 45 and 118), and Phe versus Tyr (position 151). Most variant positions are external, and only two or three different amino acids are observed. The exceptions are residues 13, 28, 110, and 142, which are internal, and positions 66 and 145, where substitutions of 5 and 4 different amino acids, respectively, are observed. None of the amino acid substitutions is at a position previously identified as causing significant changes in oxygen binding.
|
Stable Whale Myoglobins--
The three most stable myoglobins are
those from the sperm whales, with dwarf (K. simus) > pygmy (K. breviceps) > normal (P. catadon)
(Table III). This order of stability arises from 6-fold differences in
the unfolding constants for the N I transition. The I U
transitions are virtually identical for all three species (KIU 0.01).
Dwarf and pygmy sperm whale differ from each other by only two
residues. Ser is present at both positions 51 and 132 in dwarf sperm
whale myoglobin, whereas threonine is present at these positions in the
pygmy sperm whale protein. In the crystal structure of normal
sperm whale metmyoglobin, Thr51(CD1) forms an
intra-segmental H bond to the NH of residue 54 (D4). The NMR results
show that the D helix is still intact in the N state but appears to
unfold partially during the N I transition (2). Thus, the
Thr51 (D1) Ser substitution may stabilize the N state
by strengthening the D helix. The Thr132 (H8) Ser
substitution is less likely to be an important contributor because the
H helix remains intact in the molten globule intermediate.
The unfolding titrations of goosebeak whale (Ziphus cavirostris), dog (Canis familiaris), and minke whale (Balaenoptera acutorostrata) apomyoglobins show broad, bell-shaped fluorescence changes and CD profiles with a plateau that distinguishes the intermediate state (Fig. 3). Each of these apomyoglobins have larger KNI and smaller KUI values than the sperm whale apoprotein (Table III). The increase in the relative stability of the I state is large in all three cases, and most extreme in goosebeak whale apomyoglobin. A sequence comparison reveals that the only substitutions these species have in common, which are not found in two or more species with normal unfolding profiles, are Ala at positions 5 (A3) and 129 (H5) instead of Gly. Alanines at these two positions should strengthen the A and H helices, which are still intact in the I state (Fig. 1).
Identification of Other Key Residues-- The sequences of the globins are listed in order of decreasing stability in Table IV. The less stable apomyoglobins have Gly at residues 1, 15, 35, and 74, whereas the more stable apomyoglobins have Val, Ala, His, or Ser at these positions. The more stable apomyoglobins have Glu at positions 4 and 109 instead of Asp. At position 27, the relative influence of the acidic amino acids is reversed with respect to apoprotein stability. At positions 28 and 101 the more stable apomyoglobins have Ile instead of Val, and at positions 45 and 118 the more stable apomyoglobins have Arg instead of Lys. Site-directed mutants were made at several of these positions to determine which residues have the greatest effect on apomyoglobin stability (Table V). The recombinant systems for sperm whale and pig myoglobins were chosen for mutagenesis because of their large differences in apomyoglobin stability (18, 20).
|
Position 45 (CD3)-- All known myoglobin sequences have Lys at position 45 (CD3) with the exception of the three sperm whales and aardvark, which have an Arg. Although either residue can form a salt bridge to residue 60 (Asp) in the E helix, Arg might enhance stability through its ability to form multiple hydrogen bonds with surrounding residues and water molecules. However, the sperm whale R45K mutant protein and the pig K45R mutant protein have unfolding curves very similar to those of their respective wild type proteins, and the fitted equilibrium unfolding constants are very similar with changes of less than a factor of 2.
Gly to Ala/Ser Replacements-- Ala is found at positions 15 (A13) and 74 (E17) in sperm whale myoglobins, whereas pig and other less stable apoproteins have Gly at these positions. Similar Gly versus Ala sequence differences are seen between the proteins of mesophilic and thermophilic organisms (48), and Fersht, Baldwin, and co-workers (49, 50) have examined the stabilizing effects of Ala to Gly replacements in a variety of systems including apomyoglobin. Sperm whale myoglobins also have a Ser or His at position 35 (B16), whereas less stable apomyoglobins have Gly, a helix breaker. The contributions of these residues were examined by inserting Gly residues into sperm whale myoglobin at positions 15, 35, and 74, and by introducing the sperm whale residues (Ala15, Ser35, and Ala74) into the pig protein.
Either singly or in combination with A74G, the A15G and S35G mutations do cause measurable decreases in the overall stability of recombinant sperm whale apomyoglobin, but the corresponding G15A and G35S mutations do not enhance the stability of pig apomyoglobin. Conversely, the A74G mutation has little effect on the net overall stability of sperm whale apomyoglobin, but the corresponding G74A substitution enhances the stability of pig apomyoglobin ~3-fold, primarily because of a large decrease in KNI.
Decoupling CD and Fluorescence Changes in A74G
Mutants--
In contrast with the results for the other sperm whale
mutants, changes in total fluorescence emission and CD are not
concerted for the unfolding of the sperm whale A74G mutant. The
fluorescence intermediate is formed and begins to disappear before the
CD intermediate is formed (Fig. 4).
Similar discrepancies are observed for all the multiple mutants that
include this mutation. Presumably the residue at position 74 (E17) has
an influence on the fluorescence emission of the tryptophans at
positions 7 and 14 in the nearby A helix. For this reason, equilibrium
constants were derived from fits to only the CD data for the
apomyoglobins containing this replacement. In addition, the I state for
A74G sperm whale apomyoglobin appears to have less helical structure
than that of wild type apoprotein. The fractional CD change for the
formation of the I state (YI) is 0.59 for this
mutant instead of 0.43, which is observed for all the other native,
wild type, and mutant apomyoglobins (Fig. 4). These results suggest
that complete denaturation of the E helix and EF corner may alter the
structure and extent of folding in the molten globule intermediate.
|
Construction of More Stable Pig Apomyoglobins--
Two complex
mutants of pig myoglobin were designed to enhance overall stability
based on the results from the single and double mutant studies and on
the sequence differences between all the mammalian apomyoglobins listed
in Table IV. Both recombinant proteins contain the favorable G74A
mutation and selected residues found in the three sperm whale proteins
but not in the pig protein. The KNI value of the
N12H/E27D/V28I/G74A/N140K/M142I mutant is 4-fold smaller than that of
pig wild type apomyoglobin and similar to that of the pig G74A single
mutant. KIU for this mutant is also reduced
3-4-fold, suggesting that the His12 (A9),
Lys140 (H16), and Ile142 (H18) residues
stabilize the A and H helices, respectively, in the intermediate state.
The net result is a 15-fold increase in overall stability compared with
wild type pig apomyoglobin (Fig. 5).
|
The G5A/S51T/D53A/G74A/T87K pig mutant was based on a combination of
strategies. The G5A mutation was identified as a potentially important
mutation from its presence in minke and goosebeak whales, which have
exceptionally stable I states. Ser51 is one of the two Thr
Ser substitutions responsible for the decreased value of
KNI for dwarf sperm whale apomyoglobin compared with that for the pygmy sperm whale apoprotein. The Thr87
(F2) to Lys mutation was chosen because this residue is a lysine in all
the apomyoglobins examined except that for the pig. Similarly, all of
the other stable apomyoglobins have an Ala53 residue
instead of Asp. The resulting multiple mutant has a resistance to GdmCl
denaturation, which is virtually identical to that of old wild type
sperm whale apomyoglobin (Fig. 5). Both the KNI and KIU values are similar to those of the
recombinant sperm whale control apoprotein, and the overall stability
of the multiple mutant is virtually identical to that of the whale globin.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Common Unfolding Mechanism-- The apomyoglobins from all 13 mammalian species unfold by the same three-state, two-transition reaction mechanism shown in Fig. 1. Three species (minke whale, dog, and goosebeak whale) have a combination of unstable N and stable I states that allows simple detection of the molten globule intermediate by both CD and fluorescence changes. These data confirm the validity of using fluorescence data to define the transitions when the CD curves show a more concerted transition. Only in the case of A74G replacements in sperm whale apomyoglobin was it impossible to reconcile the fluorescence profiles with those obtained from CD changes.
Physiological Relevance-- The results in Table III show that deep diving mammals have significantly more stable apoproteins than terrestrial mammals. According to the neutral corridor theory, the stability of related proteins should be similar, and the effects of single or multiple mutations along the evolutionary pathway should be compensatory unless there is a strong physiological need for resistance to denaturation (51). Under selective pressure, an accumulation of small stabilizing mutations will cause an increase in overall stability but not a change in the functional character of the protein. Bocian's group (10) has shown that acid denaturation of deoxymyoglobin appears to require an initial unfolding of the globin tertiary structure before the proximal histidine-heme iron bond is hydrated and broken. Because sustained anaerobic and acidic conditions may occur in the skeletal muscle of whales and seals during prolonged dives (52-54), their myoglobins are almost certainly under selective pressure for increased resistance to unfolding and heme loss during hypoxia and acidosis. The need to function under these extreme conditions probably accounts for increased stabilities of the sperm whale apomyoglobins compared with those observed for the myoglobins from terrestrial mammals.
Comparative Strategies for Enhancing Globin Stability-- It is difficult to link individual single amino acid substitutions to the large differences in protein stability observed among the native apomyoglobins. In general, most of the naturally occurring single substitutions shown in Table IV have relatively small effects on resistance to denaturation by GdmCl. Normal sperm whale myoglobin has Ala, Ser, and Ala at positions 15, 35, and 74, whereas pig myoglobin has Gly, a helix breaker, at all three of these positions. Of the single mutants designed rationally at these positions, only G74A causes a significant increase in overall stability of pig apomyoglobin, and no further enhancement of stability occurs when this mutation is combined with G15A or G35S. In the case of the dwarf and pygmy sperm whale proteins, subtle Thr to Ser substitutions at two exterior positions cause 4-fold increases in the stability of the N state. The apomyoglobins with the most stable I states (minke, goosebeak whale, and dog) have alanines at the 5 (A3) and 129 (H5) positions.
Two multiple mutants of pig myoglobin were designed solely on
substitution trends seen for the stable mammalian globins. The most
stable multiple mutant, G5A/S51T/D53A/G74A/T87K, enhanced the stability
of pig apomyoglobin 60-fold, making it equivalent to the old wild type
form of recombinant sperm whale apomyoglobin. The locations and
identities of the mutations that, in combination, increase the
stability of pig apomyoglobin to that of sperm whale wild type are
shown in Fig. 6. The residues are all
located far from the heme group and primarily near the ends of helices
and on the protein surface.
|
Comparison of mesophilic and thermophilic proteins is often used to suggest substitutions that are responsible for alterations in protein stability (55-57). Alternatively, selection for protein stability can be accomplished by cloning a mesophilic gene into a thermophile and selecting for protein activity (58). However, these two methods require a thermophilic version of the gene or a protein that is essential for survival of the thermophilic organism into which it is cloned. Neither of these approaches is suitable for many mammalian proteins. We have shown that analysis of naturally occurring mesophilic myoglobin variants is an effective way to select multiple modifications for large enhancements of heme protein stability without changing functionality or mutating all the different residues individually, as was done in the pioneering studies of Fersht, Serrano, and co-workers (50, 59-61).
The comparative approach shows great promise for engineering stability into heme proteins that are being developed for industrial and chemical uses. For example, Steipe and co-workers (62-64) applied this approach to constructing and producing large amounts of functional immunoglobulin V(L) domains in E. coli. A major limiting factor in the use of recombinant proteins as pharmaceuticals is expression yield. In the case of heme proteins, the key factor appears to be globin stability. The newly translated apoprotein must be stable enough to remain in solution until heme is made available by either bacterial synthesis or exogenous addition. Once formed, the holoprotein exists primarily in the deoxygenated state because of rapid consumption of O2 by the respiring bacteria, and the resistance of deoxymyoglobin to denaturation correlates strongly with globin stability (10).
The results in Table III provide a quantitative explanation for why
wild type pig and human myoglobins have not been successfully expressed
in large quantities as intact holoproteins in E. coli (18).
The pig and human apomyoglobins are either too unstable to remain in
the bacterial cytoplasm long enough to take up heme or the resultant
deoxymyoglobins denature readily at 37° C during cell growth and
harvesting. In contrast, large quantities of sperm whale holomyoglobin
can be expressed constitutively because of its much greater resistance
to unfolding (20).
| |
ACKNOWLEDGEMENTS |
|---|
We thank Eileen W. Singleton for assistance in expression and purification of the pig myoglobin mutants and Markos Moraitis for adapting NON-LIN to fit combined CD and fluorescence unfolding data. We also thank Frank N. Gurd, Jay Berzofsky, and Masao Ikeda-Saito for gifts of the cetacean and human myoglobin proteins.
| |
FOOTNOTES |
|---|
* This work was supported by National Institutes of Health Training Grant GM08280 (to E. E. S.) and U.S. Public Health Service Grant GM 35649, HL 47020, Robert A. Welch Grant C-512, and the Keck Center for Computational Biology.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
To whom correspondence should be addressed: Dept. of Biochemistry,
MS 140, Rice University, Houston, TX 77005. E-mail:
olson@rice.edu.
Published, JBC Papers in Press, June 13, 2000, DOI 10.1074/jbc.M000452200
| |
ABBREVIATIONS |
|---|
The abbreviation used is: GdmCl, guanidinium chloride.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
| 1. | Hughson, F. M., Wright, P. E., and Baldwin, R. L. (1990) Science 249, 1544-1548 |
| 2. | Eliezer, D., and Wright, P. E. (1996) J. Mol. Biol. 263, 531-538 |
| 3. | Hughson, F. M., and Baldwin, R. L. (1989) Biochemistry 28, 4415-4422 |
| 4. | Barrick, D., and Baldwin, R. L. (1993) Protein Sci 2, 869-876 |
| 5. | Barrick, D., Hughson, F. M., and Baldwin, R. L. (1994) J. Mol. Biol. 237, 588-601 |
| 6. | Eliezer, D., Jennings, P. A., Dyson, H. J., and Wright, P. E. (1997) FEBS Lett. 417, 92-96 |
| 7. | Nishii, I., Kataoka, M., Tokunaga, F., and Goto, Y. (1994) Biochemistry 33, 4903-4909 |
| 8. | Kataoka, M., Nishii, I., Fujisawa, T., Ueki, T., Tokunaga, F., and Goto, Y. (1995) J. Mol. Biol. 249, 215-228 |
| 9. | Hargrove, M. S., Krzywda, S., Wilkinson, A. J., Dou, Y., Ikeda-Saito, M., and Olson, J. S. (1994) Biochemistry 33, 11767-11775 |
| 10. | Tang, Q., Kalsbeck, W. A., Olson, J. S., and Bocian, D. F. (1998) Biochemistry 37, 7047-7056 |
| 11. | Olson, J. S., Eich, R. F., Smith, L. P., Warren, J. J., and Knowles, B. C. (1997) Artif. Cells Blood Substit. Immobil. Biotechnol. 25, 227-241 |
| 12. | Kay, M. S., and Baldwin, R. L. (1996) Nat. Struct. Biol. 3, 439-445 |
| 13. | Luo, Y., Kay, M. S., and Baldwin, R. L. (1997) Nat. Struct. Biol. 4, 925-930 |
| 14. | Balestieri, C., Colonna, G., Biovane, A., Irace, G., and Servillo, L. (1976) FEBS Lett. 66, 60-64 |
| 15. | Hughson, F. M., Barrick, D., and Baldwin, R. L. (1991) Biochemistry 30, 4113-4118 |
| 16. | Nagai, K., and Thogersen, H. C. (1984) Nature 309, 810-812 |
| 17. | Smerdon, S. J., Dodson, G. G., Wilkinson, A. J., Gibson, Q. H., Blackmore, R. S., Carver, T. E., and Olson, J. S. (1991) Biochemistry 30, 6252-6260 |
| 18. | Smerdon, S. J., Oldfield, T. J., Dodson, E. J., Dodson, G. G., Hubbard, R. E., and Wilkinson, A. J. (1990) Acta Crystallographica B46, 370-377 |
| 19. | Ikeda-Saito, M., Lutz, R. S., Shelley, D. A., McKelvey, E. J., Mattera, R., and Hori, H. (1991) J. Biol. Chem. 266, 23641-23647 |
| 20. | Springer, B. A., and Sligar, S. G. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 8961-8965 |
| 21. | Springer, B. A., Egeberg, K. D., Sligar, S. G., Rohlfs, R. J., Mathews, A. J., and Olson, J. S. (1989) J. Biol. Chem. 264, 3057-3060 |
| 22. | Carver, T. E., Brantley, R. E., Jr., Singleton, E. W., Arduini, R. M., Quillin, M. L., Phillips, G. N., Jr., and Olson, J. S. (1992) J. Biol. Chem. 267, 14443-14450 |
| 23. | Wittenberg, J. B., and Wittenberg, B. A. (1981) Methods Enzymol. 76, 29-42 |
| 24. | Flanagan, M. A., Garcia-Marino, B. E., Friend, S. H., Feldman, R. J., Scouloudi, H., and Gurd, F. R. N. (1983) Biochemistry 22, 6027-6037 |
| 25. | Ascoli, R. F., Rossi-Fannelli, M. R., and Antonini, E. (1981) Methods Enzymol. 76, 72-77 |
| 26. | Pace, C. N., Shirley, B. A., and Thomson, J. A. (1990) in Protein Structure: A Practical Approach (Creighton, T. E., ed) Chapter 13, pp. 311-330 Oxford University Press, Oxford, UK |
| 27. | Ramsay, G., Ionescu, R., and Eftink, M. R. (1995) Biophys. J. 69, 701-707 |
| 28. | Barrick, D., and Baldwin, R. L. (1993) Biochemistry 32, 3790-3796 |
| 29. | Elwell, M. L., and Schellman, J. S. (1979) Biochim. Biophys. Acta 580, 327-338 |
| 30. | Johnson, M. L., and Frasier, S. G. (1985) Methods Enzymol. 117, 301-342 |
| 31. | Brantley, R. E., Jr., Smerdon, S. J., Wilkinson, A. J., Singleton, E. W., and Olson, J. S. (1993) J. Biol. Chem. 268, 6995-7010 |
| 32. | Hargrove, M. S., and Olson, J. S. (1996) Biochemistry 35, 11310-11318 |
| 33. | Ramos, C. H., Kay, M. S., and Baldwin, R. L. (1999) Biochemistry 38, 9783-9790 |
| 34. | Nishii, I., Kataoka, M., and Goto, Y. (1995) J. Mol. Biol. 250, 223-238 |
| 35. | Bolen, D. W., and Santoro, M. M. (1988) Biochemistry 27, 8069-8074 |
| 36. | Santoro, M. M., and Bolen, D. W. (1988) Biochemistry 27, 8063-8068 |
| 37. | Santoro, M. M., and Bolen, D. W. (1992) Biochemistry 31, 4901-4907 |
| 38. | Plaza del Pino, I. M., Pace, C. N., and Freire, E. (1992) Biochemistry 31, 11196-11202 |
| 39. | Hu, C. Q., Sturtevant, J. M., Thomson, J. A., Erickson, R. E., and Pace, C. N. (1992) Biochemistry 31, 4876-4882 |
| 40. | Pfeil, W. (1998) Protein Stability and Folding: A Collection of Thermodynamic Data , Springer Verlag, New York |
| 41. | Mayr, L. M., and Schmid, F. X. (1993) Biochemistry 32, 7994-7998 |
| 42. | DeKoster, G. T., and Robertson, A. D. (1995) J. Mol. Biol. 249, 529-534 |
| 43. | Yao, M., and Bolen, D. W. (1995) Biochemistry 34, 3771-3781 |
| 44. | Yang, M., Liu, D., and Bolen, D. W. (1999) Biochemistry 38, 11216-11222 |
| 45. | Cashon, R. E., Vayda, M. E., and Sidell, B. D. (1997) Comp. Biochem. Physiol. B 117, 613-620 |
| 46. | Loh, S. N., Kay, M. S., and Baldwin, R. L. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 5446-5450 |
| 47. | Jamin, M., and Baldwin, R. L. (1998) J. Mol. Biol. 276, 491-504 |
| 48. | Argos, P., Rosmann, M., Grau, U., Sukborn, H., Frank, G., and Tratschin, J. (1979) Biochemistry 18, 5698-5703 |
| 49. | Kiefhaber, T., and Baldwin, R. L. (1995) J. Mol. Biol. 252, 122-132 |
| 50. | Serrano, L., Neira, J. L., Sancho, J., and Fersht, A. R. (1992) Nature 356, 453-455 |
| 51. | Malcom, B. A., Wilson, K. P., Matthews, B. W., Kirsch, J. F., and Wilson, A. C. (1990) Nature 345, 86-89 |
| 52. | Kooyman, G. L., and Ponganis, P. J. (1998) Annu. Rev. Physiol. 60, 19-32 |
| 53. | Snyder, G. K. (1983) Respir. Physiol. 54, 269-294 |
| 54. | Zapol, W. M., Liggins, G. C., Schneider, R. C., Qvist, J., Snider, M. T., Creasy, R. K., and Hochachka, P. W. (1979) J. Appl. Physiol. 47, 968-973 |
| 55. | Liao, H., McKenzie, T., and Hageman, R. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 576-580 |
| 56. | Hasegawa, J., Shimahara, H., Mizutani, M., Uchiyama, S., Arai, H., Ishii, M., Kobayashi, Y., Ferguson, S. J., Yoshihiro, S., and Igarashi, Y. (1999) J. Biol. Chem. 274, 37533-37537 |
| 57. | Haney, P. J., Badger, J. H., Buldak, G. L., Reich, C. I., Woese, C. R., and Olsen, G. J. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3578-3583 |
| 58. | Imanaka, T., Shibazaki, M., and Takagi, M. (1986) Nature 324, 695-697 |
| 59. | Serrano, L., Day, A. G., and Fersht, A. R. (1993) J. Mol. Biol. 233, 305-312 |
| 60. | Fersht, A. R., Matouschek, A., and Serrano, L. (1992) J. Mol. Biol. 224, 771-782 |
| 61. | Serrano, L., Sancho, J., Hirshberg, M., and Fersht, A. R. (1992) J. Mol. Biol. 227, 544-559 |
| 62. | Wirtz, P., and Steipe, B. (1999) Protein Sci. 8, 2245-2250 |
| 63. | Ohage, E. C., Wirtz, P., Barnikow, J., and Steipe, B. (1999) J. Mol. Biol. 291, 1129-1134 |
| 64. | Ohage, E., and Steipe, B. (1999) J. Mol. Biol. 291, 1119-1128 |
| 65. | Springer, B. A., Sligar, S. G., Olson, J. S., and Phillips, G. N., Jr. (1994) Chem. Rev. 94, 699-714 |
| 66. | Hargrove, M. S., Wilkinson, A. J., and Olson, J. S. (1996) Biochemistry 35, 11300-11309 |
| 67. | Goss, D. J., LaGow, J. B., and Parkhurst, L. J. (1982) Comp. Biochem. Physiol. B 71, 229-233 |
| 68. | Antonini, E. (1965) Physiol. Rev. 45, 123-170 |
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. L. Miranda, D. H. Maillett, J. Soman, and J. S. Olson Thermoglobin, Oxygen-avid Hemoglobin in a Bacterial Hyperthermophile J. Biol. Chem., November 4, 2005; 280(44): 36754 - 36761. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Picotti, A. Marabotti, A. Negro, V. Musi, B. Spolaore, M. Zambonin, and A. Fontana Modulation of the structural integrity of helix F in apomyoglobin by single amino acid replacements Protein Sci., June 1, 2004; 13(6): 1572 - 1585. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. T.J. van den Bremer, A. H. Keeble, W. Jiskoot, R. E.J. Spelbrink, C. S. Maier, A. Van Hoek, A. J.W.G. Visser, R. James, G. R. Moore, C. Kleanthous, et al. Distinct conformational stability and functional activity of four highly homologous endonuclease colicins Protein Sci., May 1, 2004; 13(5): 1391 - 1401. [Abstract] [Full Text] [PDF]
|
|
|
|
<
). The residue number and the helical position are noted above. The
consensus sequence is given below.
