Myristate binding to a secondary site FAx, allosterically coupled to the heme site, not only increases optical absorbance of MnIIIheme-bound HSA by a factor of approximately three, but a
Trang 1binding to human serum albumin
Optical and NMR spectroscopy characterization
Gabriella Fanali1, Riccardo Fesce1, Cristina Agrati1, Paolo Ascenzi2,3and Mauro Fasano1
1 Dipartimento di Biologia Strutturale e Funzionale, and Centro di Neuroscienze, Universita` dell’Insubria, Busto Arsizio (VA), Italy
2 Dipartimento di Biologia, and Laboratorio Interdisciplinare di Microscopia Elettronica, Universita` ‘Roma Tre’, Italy
3 Istituto Nazionale per le Malattie Infettive I.R.C.C.S ‘Lazzaro Spallanzani’, Roma, Italy
Human serum albumin (HSA) is the most prominent
protein in plasma, but it is also found in tissues and
secretions throughout the body HSA abundance (its
concentration being 45 mgÆmL)1 in the serum of
human adults) contributes significantly to
colloid-osmotic blood pressure HSA, best known for its
extraordinary ligand binding capacity, is constituted
by a single nonglycosylated all-a chain of 65 kDa con-taining three homologous domains (labelled I, II, and III), each composed of two (A and B) subdomains The three domains have different binding capacity for a broad variety of ligands such as aminoacids (Trp and Cys), hormones, metal ions, and bilirubin Moreover, HSA has a high affinity for heme and is
Keywords
allostery; fatty acid binding; heme binding;
human serum albumin; NMR relaxation
Correspondence
M Fasano, Dipartimento di Biologia
Strutturale e Funzionale, Universita`
dell’Insubria, Via Alberto da Giussano 12,
I-21052 Busto Arsizio (VA), Italy
Fax: +39 0331 339459
Tel: +39 0331 339450
E-mail: mauro.fasano@uninsubria.it
Website: http://fisio.dipbsf.uninsubria.it/cns/
fasano
(Received 21 April 2005, revised 25 July
2005, accepted 26 July 2005)
doi:10.1111/j.1742-4658.2005.04883.x
Human serum albumin (HSA) is best known for its extraordinary ligand binding capacity HSA has a high affinity for heme and is responsible for the transport of medium and long chain fatty acids Here, we report myri-state binding to the N and B conformational myri-states of Mn(III)heme–HSA (i.e at pH 7.0 and 10.0, respectively) as investigated by optical absorbance and NMR spectroscopy At pH 7.0, Mn(III)heme binds to HSA with lower affinity than Fe(III)heme, and displays a water molecule coordinated to the metal Myristate binding to a secondary site FAx, allosterically coupled to the heme site, not only increases optical absorbance of Mn(III)heme-bound HSA by a factor of approximately three, but also increases the Mn(III)-heme affinity for the fatty acid binding site FA1 by 10–500-fold Cooper-ative binding appears to occur at FAx and accessory myristate binding sites The conformational changes of the Mn(III)heme–HSA tertiary struc-ture allosterically induced by myristate are associated with a noticeable change in both optical absorbance and NMR spectroscopic properties of Mn(III)heme–HSA, allowing the Mn(III)-coordinated water molecule to exchange with the solvent bulk At pH¼ 10.0 both myristate affinity for FAx and allosteric modulation of FA1 are reduced, whereas cooperation
of accessory sites and FAx is almost unaffected Moreover, Mn(III)heme binds to HSA with higher affinity than at pH 7.0 even in the absence of myristate, and the metal-coordinated water molecule is displaced As a whole, these results suggest that FA binding promotes conformational changes reminiscent of N to B state HSA transition, and appear of general significance for a deeper understanding of the allosteric modulation of ligand binding properties of HSA
Abbreviations
FA, fatty acid; HSA, human serum albumin; MSE, mean square error; NMRD, nuclear magnetic relaxation dispersion.
Trang 2responsible for the transport of lipophilic compounds
and drugs and of medium and long chain fatty acids;
among them, myristic acid is a stereotypic ligand to
investigate fatty acid binding and transport properties
of HSA [1–8]
Fatty acids (FAs) are required for the synthesis of
membrane lipids, hormones and second messengers,
and serve as an important source of metabolic
energy Although the binding of fatty acids to human
and bovine serum albumin has been thoroughly
inves-tigated over many years, their binding mode and
thermodynamics are still objects of debate By
combi-ning biochemical and biophysical approaches, a
com-mon consensus view has been reached on there being
three high-affinity fatty acid binding sites, and at least
three further low affinity sites have been envisaged
NMR studies on tryptic and peptic fragments of
bovine serum albumin have localized two high affinity
sites in domain III and one in the N-terminal half of
the protein Structural X-ray diffraction studies have
demonstrated that HSA is able to bind up to seven
equivalents of long chain FAs at multiple binding
sites (labelled FA1 to FA7; Fig 1) with different
affinity In sites FA1–5 the carboxylate moiety of
fatty acids is anchored by electrostatic⁄ polar
inter-actions; on the contrary, sites FA6–7 do not display
a clear evidence of polar interactions that keep in
place the carboxylate head of the fatty acid, thus
suggesting that sites FA6–7 are low-affinity fatty acid binding sites [1,6,7,9–17]
The fatty acid binding site FA1, located in subdo-main IB (Fig 1), acts as the heme binding site as well, with the tetrapyrrole ring arranged in a d-shaped cavity limited by two tyrosine residues (Tyr138 and Tyr161) that provide p-p stacking interaction with the porphyrin and supply a donor oxygen (from Tyr161) for the ferric heme iron Ferric heme is secured by the long IA-IB connecting loop that fits into the cleft opening Heme propionates point toward the interface between domains I and III and are stabilized by salt bridges with His146 and Lys190 residues [6,8]
HSA undergoes pH- and allosteric effector-depend-ent reversible conformational isomerization(s) Between
pH 2.7 and 4.3, HSA shows a fast (F) form, character-ized by a dramatic increase in viscosity, much lower solubility, and a significant loss in helical content Between pH 4.3 and 8, in the absence of allosteric effectors, HSA displays the normal (N) form that is characterized by heart-shaped structure Between
pH 4.3 and 8, in the presence of allosteric effectors, and at pH greater than 8, in the absence of ligands, HSA changes conformation to the basic (B) form with loss of a-helix and an increased affinity for some lig-ands, such as warfarin [5,18–23]
Fatty acids are effective in allosterically regulating ligand binding to Sudlow’s site I and to the heme cleft Myristate regulates HSA binding properties in a com-plex manner, involving both competitive and allosteric mechanisms The structural changes associated with FAs binding can essentially be regarded as relative domain rearrangements to the I-II and II-III inter-faces This allosteric regulation is not observed for short FAs (e.g octanoate) that preferably bind to Sudlow’s site II and displace the specific ligands (e.g ibuprofen) without inducing HSA allosteric rearrangement(s) This indicates that the hydrophobic interactions between the long FA polymethylenic tail and HSA drives allosteric rearrangements In turn, Sudlow’s site
I ligands (e.g warfarin) displace FA7, while Sudlow’s site II ligands (e.g ibuprofen) displace FA3 and FA4 Moreover, heme binding to HSA displaces FA1 [6,8,13,16,23–26]
Heme binding to HSA endows this protein with peculiar optical absorbance and magnetic spectroscopic properties that can be used to follow ligand- and pH-dependent conformational transition(s) [19–22,27] In particular, Mn(III)heme can be used instead of Fe(III)-heme in order to increase the strength of the dipolar interaction with water protons when their NMR relax-ation rate is measured [19,20] Although an even stronger dipolar interaction could be obtained using
Fig 1 Ribbon representation of the heart-shaped structure of HSA
with the seven fatty acid binding sites labeled (FA1 to FA7); sites
are occupied by myristate anions rendered with red sticks N- and
C-termini of the polypeptide chain are labeled accordingly Atomic
coordinates are taken from [6,8,13,14] The figure was drawn using
the SWISS PDB viewer (http://www.expasy.org/spdbv/).
Trang 3Mn(II)heme, the metal undergoes oxidation under
aerobic conditions in porphyrin complexes [28]
Heme regulates allosterically drug binding to
Sud-low’s site I In fact, heme affinity for HSA decreases
by about one order of magnitude upon warfarin
bind-ing Reciprocally, heme binding to HSA decreases
war-farin affinity by the same extent [19] Fe(III)heme
allosterically inhibits ligand binding to Sudlow’s site I,
possibly by stabilizing the neutral (N) state of HSA
Vice versa, ligand binding to Sudlow’s site I impairs
Fe(III)heme–HSA formation, possibly by stabilizing
the basic (B) state of HSA [5,18,23,29–31]
Here, we report the spectroscopic analysis of the
myristate-dependent conformational changes of the N
and B states of Mn(III)heme–HSA, by optical
absorb-ance spectroscopy and NMR spectroscopy, that show
allosteric interaction(s) between FAs and Mn(III)heme
with HSA Interestingly, FAs increase Mn(III)heme
affinity to HSA, whereas warfarin and FA7 ligands
were reported to behave in the opposite way with
respect to ferric heme binding to HSA [19,21,31]
Additionally, the affinity of Mn(III)heme for HSA and
the spectroscopic properties of the Mn(III)heme–HSA
adduct in the presence of myristate are similar to those
of the B conformational state of HSA, suggesting that
myristate binding to one or more modulatory sites
possibly drives the N to B state HSA transition
Results
In the absence of myristate, at pH 7.0 (i.e HSA in the
N conformational state), Mn(III)heme binds to fatty
acid-free HSA with a dissociation constant KH
2.0· 10)5m (Fig 2A) Although the binding curve
does not reach saturation and therefore the KH value
should be considered as a lower limit, it is worth to
note that it is two order of magnitude larger than that
measured for Fe(III)heme [32] In the presence of
1.0· 10)4m myristate, the optical absorbance
spec-trum of Mn(III)heme–HSA displays a characteristic
shoulder at 440 nm with well-defined isosbestic points
(Fig 2B)
In the presence of myristate, the expression for
HSA-bound Mn(III)heme concentration could not be
solved analytically Four major features are evident in
optical absorbance difference (DA) curves: (a) at low
HSA concentrations, the curves are depressed by
myri-state, indicating that myristate affinity for FA1 is
higher than that of Mn(III)heme in the absence of
myristate, and Mn(III)heme binding to FA1 is
pre-cluded by competition equilibrium (left column of
Scheme 1) (b) Maximal DA values are clearly
increased in the presence of myristate, thereby
indica-ting that binding of myristate to a modulatory site FAx increases the signal yield of the complex The DA value for 1.0· 10)5m Mn(III)heme–HSA–myristate complex can be estimated about A10*¼ 0.33, by nor-malizing the value observed at 1.0· 10)4m myristate and 30 lm HSA (0.255) to full Mn(III)heme binding, based on the molar fraction of the Mn(III)heme–HSA adduct that gave similar spectral data at pH 10.0 (see below) Furthermore, (c) at intermediate HSA concen-tration the binding curves rapidly rise and appear to
Fig 2 (A) Binding isotherms for Mn(III)heme binding to fatty acid-free HSA and to the HSA–myristate complexes, at pH 7.0 and 25.0 C; open triangles: no myristate; solid triangles: 5.0 · 10)6M myristate; open circles: 1.0 · 10)5M myristate; solid circles: 2.5 · 10)5M myristate; crossed diamonds: 5.0 · 10)5M myri-state; open diamonds: 7.5 · 10)5M myristate; solid diamonds: 1.0 · 10)4M myristate The continuous lines were obtained by numerical fitting of the data Values of the dissociation equilibrium constants obtained according to Scheme 1 are given in Table 1 (B) UV-visible spectral changes observed for a solution of 1.0 · 10)5M Mn(III)heme titrated with HSA (0–3.0 · 10)5M ) in the presence of 1.0 · 10)4M myristate, at pH 7.0 and 25.0 C The arrows indicate the increase of HSA concentration.
Trang 4reach saturation for HSA concentrations well lower
than in the absence of myristate This indicates that
binding of myristate to the modulatory site also
increa-ses the affinity of Mn(III)heme for FA1 (Scheme 1,
central column) Finally, (d) at high HSA
concentra-tion and intermediate myristate concentraconcentra-tions (1.0–
5.0· 10)5 m) the binding curves decline, suggesting
that unbinding of myristate from FAx occurs,
accord-ing to equilibrium of the framed reaction in Scheme 1
A kinetic model was set up to numerically fit the
optical absorbance data reported in Fig 2A The
mini-mal core of the model was based on the competition
between Mn(III)heme and myristate for binding to
FA1 (defined by the parameters KH and KM) and on
the allosteric modulation of FA1 properties by
myri-state binding to FAx (defined by the parameters KM*,
KHM, and possibly KMM1KM, if myristate binding to
FA1 is also modulated; see Experimental procedures
for explanation of the notations for the equilibrium
constants) However, binding of myristate to
addi-tional FA sites must also be considered, to take into
account the decrease in free myristate concentration at
increasing concentrations of HSA; this requires the
further set of parameters KMS1 to KMS5 For the sake
of simplicity, these constants were bound to a fixed
affinity ratio series, with KMS1 as a free parameter and
KMSn¼ KM ⁄ 10(n)1) ⁄ 2, n¼ 2–5; this is in general
agreement with the estimates reported in the literature
[1,9,10,33] Two further free parameters (in addition to
KH, KM, KHM, KM* and KMS1) completed the model:
the asymptotic absorbance in the absence of myristate
(A10) and the absorbance of 1.0· 10)5 m
Mn(III)-heme–HSA–myristate complex (A10*) However, this
simplified model did not adequately fit the
experimen-tal data (MSE¼ 4.9 · 10)5); in particular, it could not
reproduce the peak followed by partial decline observed at intermediate myristate concentrations, par-ticularly evident for 1.0–5.0· 10)4m myristate, and in general the right part of the curves (at high [HSA]) In order to qualitatively reproduce this feature, positive cooperation must be introduced between at least one
of the additional FA binding sites and FAx, so that the Mn(III)heme–HSA–myristate adduct releases myri-state from FAx, as free myrimyri-state concentration van-ishes, and the optical absorbance signal declines Several sets of parameters gave good fits to the experimental data, yielding almost identical curves (MSE¼ 3.0 ± 0.1 · 10)5): an example of a set of fit-ting curves is displayed as continuous lines in Fig 2A All these solutions indicate a value for KMS1 in the range between 1.5· 10)6 and 3.0· 10)6m (and thus values of the dissociation constants for the 5 additional
FA sites ranging from 2· 10)6to 2· 10)8m) and sug-gest that the affinity of myristate for FAx is modulated
by additional site no 3 or 4, with dissociation constant
in the order of 8.0· 10)8to 1.3· 10)7mand a 50–200-fold decrease in FAx affinity when the coupled site releases myristate
Very similar fits were obtained, whether or not the affinity of FA1 for myristate was assumed to change when FAx is occupied The strength of cooperative coupling between accessory sites and FAx could also change over a wide extent (10–500-fold decrease in
KM* when additional FA site no 3 or 4 releases myri-state) producing equally good fits However, the set of estimated dissociation constants for FA1 and FAx changed quite markedly depending on the assumptions regarding cooperativity among FA binding sites The best fitting values for the parameters of the model (Scheme 1) are reported in Table 1 for two nicely
(Myr)P(…)–(…)
KM*
KMSn
↔
(Myr)P(Myr)–(Myr)n
(…)P(…)–(…) KM
*
Sn
↔ (…)P(Myr)–(Myr)n
KHM
(Hem)P(…)–(…) KM
H
Sn
↔ (Hem)P(Myr)–(Myr)n
Scheme 1 Allosteric and competition equilibria involving Mn(III)heme and myristate binding to HSA Binding sites are indicated with the notation (FA1)P(FAx)–(FAS), where FA1 is the heme binding site, acting as myristate binding site as well, FAx is a different myristate binding site allosterically coupled to FA1, and FAS are n secondary myristate binding sites, with different affinities, allosterically uncoupled to FA1.
P ¼ protein, HSA; Myr ¼ myristate; Hem ¼ Mn(III)heme Values of the dissociation equilibrium constants are given in Table 1 and in the text The framed transition is associated with a change in the optical absorption spectrum (see text).
Trang 5fitting models: no change in FA1 affinity for myristate,
depending on FAx binding (KMM¼ KM), or a similar
change in FA1 affinity for both Mn(III)heme and
myr-istate (KMM⁄ KM¼ KHM⁄ KH); in both cases 100-fold
decrease was assumed in KM* when additional FA site
no 3 releases myristate By inspection of the model
parameters (Table 1) it is clear that the assumptions
strongly affect the estimated affinity of FAx for
myri-state (KM*) and, as a consequence, the magnitude of
the allosteric modulation of FA1 (KHM⁄ KH¼
5.4· 10)2 vs 5.0· 10)3) The estimate of KMalso
dif-fers by about one order of magnitude, but the
differ-ence is smaller for the estimate of KMM, i.e FA1
affinity for myristate with occupied FAx, which
pre-sumably is the relevant dissociation constant for
com-petition between Mn(III)heme and myristate with the
latter in excess
The same model was also applied to data obtained
at pH 10.0 (Fig 3) The model is over-defined, and
several sets of parameters give comparable fits; the
results obtained by fixing KM to the value observed
at pH 7.0 are displayed in Table 1 The consistent
aspects, relatively independent of the model
assump-tions, are the following: (a) FA1 affinity for
Mn(III)-heme (KH) increases by at least one order of
magnitude with respect to pH 7.0, but both FAx
affinity and allosteric modulation of FA1 are reduced
(b) Cooperation of accessory sites and FAx is almost
unaffected Finally, (c) the asymptotic absorbance of
the Mn(III)heme–HSA complex (A10) becomes
com-parable to that of the Mn(III)heme–HSA–(FAx +
myristate) complex (A10*), and the latter is not
altered by the change in pH Again, the occurrence
of well-defined isosbestic points indicate that the
binding equilibrium occurs through only two forms,
the HSA-free and the HSA-bound Mn(III)heme
(Fig 3B)
Table 1 Values of the thermodynamic dissociation constants ( M ) for myristate and Mn(III)heme binding to HSA at pH 7.0 and 10.0 (Scheme 1 and see text) Assumptions: a KM ¼ K M KM* · 100 for unoccupied FAS3 b K M ⁄ K M ¼ K HM⁄ K H KM* · 100 for unoccupied FAS3.
Constant
pH
Fig 3 (A) Binding isotherms for Mn(III)heme binding to fatty acid-free HSA and to the HSA–myristate complex, at pH 10.0 and 25.0 C; open triangles: no myristate; solid diamonds: 1.0 · 10)4M myristate The continuous lines were obtained by numerical fitting
of the data Values of the dissociation equilibrium constants obtained according to Scheme 1 are given in Table 1 (B) UV-visible spectral changes observed for a solution of 1.0 · 10)5M Mn(III)-heme titrated with HSA (0–3.0 · 10)5 M ) in the presence of 1.0 · 10)4M myristate, at pH 10.0 and 25.0 C The arrows indicate the increase of HSA concentration.
Trang 6Mn(III)heme–HSA was titrated with myristate in
order to follow the conformational transition(s)
associ-ated to the fatty acid binding As shown in Fig 4,
binding of myristate to Mn(III)heme HSA causes the
appearance of a shoulder at 440 nm that disappears on
increasing myristate concentration due to the
displace-ment of Mn(III)heme from FA1 Here, the equilibrium
occurs through three different forms, Mn(III)heme–
HSA in the absence of myristate, Mn(III)heme–HSA
with myristate bound to site(s) other than FA1
(spec-trum with the shoulder at 440 nm), and free
Mn(III)-heme; therefore, no isosbestic points are observed
A consistent behavior has been observed by
measur-ing the paramagnetic contribution of Mn(III)heme to
the solvent water proton NMR relaxation rate (Eqn 1
in Experimental procedures) Figure 5 shows the
relax-ivity of fatty acid-free Mn(III)heme–HSA observed at
10 MHz, 25.0C, as a function of the myristate
con-centration The relaxation rate increases, apparently
with a varying slope, up to sevenfold molar excess of myristate, while it starts to decrease when myristate concentration is further increased
An overview of the conformational changes due to both fatty acid binding and pH may be obtained by plot-ting1H-NMR relaxation rate data vs pH for the differ-ent Mn(III)heme⁄ HSA ⁄ fatty acid ratios Figure 6 shows the pH dependence curves of the observed relaxation rate measured at 10 MHz, where this parameter is most affected, for Mn(III)heme–HSA and Mn(III)heme– HSA-myristate at 1 : 1 : 3, 1 : 1 : 4.5, and 1 : 1 : 6 molar ratios Values of pK for the three titration steps, obtained at the different Mn(III)heme–HSA-myristate ratios, have been determined using Eqn (2) (Experimen-tal procedures; Table 2)
Fig 4 (A) Absorbance change measured at 440 nm for a solution
of 1.0 · 10)5M Mn(III)heme–HSA as a function of myristate
con-centration Data were obtained at pH 7.0 and 25.0 C (B) UV-visible
absorption spectra of a solution of 1.0 · 10)5M Mn(III)heme–HSA
in the absence (continuous line) and in the presence of
4.5 · 10)5M (dotted line) and 1.0 · 10)4M myristate (dashed line).
Fig 5 Change of the relaxivity measured at 10 MHz of a 1.0 · 10)3M solution of Mn(III)heme–HSA as a function of myri-state concentration Data were obtained at pH 7.0 and 25.0 C.
Fig 6 Water proton relaxation rates measured at 10 MHz and 25.0 C, as functions of pH, for fatty acid-free Mn(III)heme–HSA (solid squares), Mn(III)heme–HSA–myristate at 1 : 1 : 3 (solid tri-angles), 1 : 1 : 4.5 (open diamonds), and 1 : 1 : 6 molar ratios (open circles) The continuous lines were calculated according to Eqn (2) Results of the fitting are given in Table 2 Under all the experi-mental conditions, the Mn(III)heme–HSA concentration was 1.0 · 10)3M
Trang 7Contributions to relaxation differ, depending on the
conformational state of HSA and on the occupancy of
the myristate binding sites The relaxation rate change
is highest between pH 5.5 and 8.0, where HSA is in
the native form (N state) The relaxivity of the
Mn(III)heme–HSA complex increases with myristate
concentration It should be noticed that at pH lower
than 5.5, myristate is expected to be in the protonated
form that is not able to bind HSA [34] Conversely,
between pH 8.3 and 11.9 (i.e where HSA is in the B
form), the contribution of
Mn(III)heme–HSA-myri-state to paramagnetic relaxation does not differ
signifi-cantly from that of fatty acid-free Mn(III)heme–HSA
As myristate binding appears to enhance the
relaxiv-ity of Mn(III)heme–HSA, we attempted to gain more
information from the analysis of NMRD profiles at
various myristate concentrations Figure 7 shows
NMRD profiles of fatty acid-free Mn(III)heme–HSA
and of Mn(III)heme–HSA-myristate obtained at
1 : 1 : 3, 1 : 1 : 4.5, and 1 : 1 : 6 molar ratios at
pH 7.0 Note that NMRD profiles are significantly
dif-ferent in the high field region whereas at the low
frequency limit they are almost coincident NMRD
profiles of Mn(III)heme–HSA as a function of
myri-state concentration were also measured at pH 10.0 in
order to check whether any change occurred for the B
state of HSA as well As shown in Fig 7, the NMRD
profiles of Mn(III)heme–HSA at pH 10.0 do not
appear to be affected by myristate
Optical absorbance spectra are suggestive of different
coordination modes of Mn(III)heme in the different
conformational states of HSA [35], therefore we
meas-ured the paramagnetic contribution to the 17O-NMR
linewidth at pH 7.0 and 10.0 as a function of myristate
concentration (Fig 8) For paramagnetic
metallopro-teins, the width of the17O NMR resonance is affected
by the presence of the paramagnetic metal through the
exchange of water molecules directly coordinated to the
metal center, according to Eqn (3) (Experimental
pro-cedures) [20] Unlike protons, 17O nuclei are negligibly
affected by dipolar coupling with nearby unpaired
electrons, and the paramagnetic broadening of the 17O resonance is diagnostic of the occurrence of a direct coordination bond between water and Mn(III) [20,36]
Table 2 pK values of pH-dependent water proton relaxation rates
measured at 10 MHz and 25.0 C of fatty acid-free Mn(III)heme–
HSA and of Mn(III)heme–HSA-myristate pK values were obtained
by fitting data in Fig 6 according to Eqn (2).
Mn(III)heme ⁄ HSA ⁄ fatty
1 : 1 : 0 6.60 ± 0.04 9.40 ± 0.16 11.80 ± 0.05
1 : 1 : 3 6.30 ± 0.01 8.00 ± 0.02 12.10 ± 0.03
1 : 1 : 4.5 5.43 ± 0.01 7.32 ± 0.01 11.40 ± 0.02
1 : 1 : 6 5.40 ± 0.01 7.40 ± 0.02 11.50 ± 0.03
Fig 7 NMRD profiles of fatty acid-free Mn(III)heme–HSA (solid squares) and of Mn(III)heme–HSA–myristate at 1 : 1 : 3 (solid tri-angles), 1 : 1 : 4.5 (open diamonds), and 1 : 1 : 6 molar ratios (open circles) at pH 7.0 (A) and at pH 10.0 (B) Under all the experimental conditions, Mn(III)heme–HSA concentration was 1.0 · 10)3M Data were obtained at 25.0 C.
Fig 8 Paramagnetic contribution to the linewidth of the 17 O water resonance of 1.6 · 10)3M solution of Mn(III)heme–HSA as a func-tion of myristate concentrafunc-tion Solid squares: pH 7.0, HSA N state; open circles: pH 10.0, HSA B state Data were obtained at 25.0 C.
Trang 8As shown in Fig 8, the linewidth change is significant
(about 20 Hz) for the protein in the N state, and
becomes larger (about 40 Hz) in the presence of
satur-ating concentration of myristate On the other hand,
this contribution is almost negligible for HSA in the B
state, and starts to increase to about 20 Hz in the
pres-ence of high myristate concentration
Discussion
Myristate binding to HSA affects the Mn(III)heme
binding properties The results presented here indicate
that current views – seven FA binding sites, FA1
involved in ipsosterical competition with heme binding,
FAx allosterically coupled to FA1, a scale of affinity
ratios of about half a decade among FA sites, with
dis-sociation constants in the range 10)6)10)8m, and
sev-eral possible allosteric cross interactions among FA
sites [1,3,6,9–16,33] – allow us to numerically model the
experimental results with good accuracy In particular,
modeling indicates that binding of myristate to FAx
not only increases optical absorbance of
Mn(III)heme-bound HSA by a factor of 3, but also increases FA1
affinity for Mn(III)heme by 10–500-fold (depending on
the assumptions about possible similar changes in
affin-ity of FA1 for myristate) This brings the value of HSA
affinity for Mn(III)heme, with myristate bound to FAx,
in the range of HSA affinity for Fe(III)heme [32]
Fur-thermore, modeling indicates that positive cooperation
between an accessory FA site (with affinity 0.8–
1.5· 10)7m for myristate) and FAx is needed to
account for the shape of the DA curves (Fig 2A)
At pH 10.0 (i.e where HSA is in the B state),
Mn(III)heme binds more strongly to HSA than at
pH 7.0 (i.e where HSA is in the N state) even in the
absence of myristate, with KH 10)6 m (Fig 3)
Moreover, in the presence of saturating concentrations
of myristate, the tendency of the curve to become
sig-moidal is much attenuated, suggesting a substantial
impairment of allosteric modulation by myristate
bind-ing to FAx Numerical analysis of the data, usbind-ing the
same models that fit the data at pH 7.0, indicate that,
independently of the model assumptions, both FAx
affinity for myristate and allosteric modulation of FA1
are reduced, whereas cooperation of accessory sites
and FAx is almost unaffected Furthermore, the
asymptotic absorbance of the Mn(III)heme–HSA
adduct (A10) becomes comparable to that of the
Mn(III)heme–HSA-(FAx+myristate) complex (A10*),
whereas the latter is not altered by the change in pH
Taken together, these observations strongly suggest
that the conformational changes produced by changing
the pH from 7.0 to 10.0 (i.e shifting the HSA
confor-mation from the N to the B state) is very similar to that induced by myristate binding to site FAx Indeed, the HSA affinity for Mn(III)heme and the absorbance
of Mn(III)heme–HSA increase by factors of about 10 and 3, respectively, and myristate effects become much attenuated Still, the same interaction(s) that at pH¼ 7.0 produces marked differences among absorbance curves at various myristate concentrations appear to fully account for the small reshaping of the curve pro-duced by 1.0· 10)4mmyristate at pH¼ 10.0
Myristate binding to HSA determines conformational changes that open the FA1 cavity allowing Mn(III)-heme binding and consequently myristate displacement Actually, addition of up to three moles of long-chain FAs is reported to enhance the binding of Sudlow’s site
I (i.e FA7) ligands, and this behaviour is usually explained by a cooperative effect established by FA binding to domain III (i.e to FA4 and FA5) [26,37–39]
On the other hand, myristate bound at the limit of subdomain IA (i.e to FA2) was suggested to be func-tionally linked to Sudlow’s site I [25] It should be noticed that binding of more than three equivalents of myristate decreases warfarin affinity for Sudlow’s site
I, as Fe(III)heme does [19,21,31,40] Sudlow’s site II (i.e FA4) ligands do not appear to be effective in modulating Sudlow’s site I ligands and heme binding properties [21,26]
The marked variation in the optical absorbance spectrum of Mn(III)heme–HSA induced by myristate binding at pH 7.0 might be explained in terms of a change in the coordination sphere of Mn(III) [35] Although structural data for Mn(III)heme–HSA are not available yet, evidence for a Mn(III)-coordinated water molecule was gained by 17O-NMR linewidth measurements, that showed a transverse relaxation rate different from Fe(III)heme–HSA, where no Fe(III)-coordinated water molecule(s) were observed [20] On the other hand, both X-ray structures deposited in PDB so far for Fe(III)heme–HSA display the Tyr161 residue as the only axial ligand for Fe(III) [6,8] In the absence of myristate, Mn(III)heme–HSA in the N state has a water molecule coordinated to the metal (Fig 8) that could provide a source for paramagnetic relaxa-tion of the solvent water bulk This is at difference with Fe(III)heme, due to the different affinity of the metals for phenolic oxygen ligands [6,8] Nevertheless, this contribution is not evident from the NMRD pro-file, which is almost superimposable to that of Mn(III)heme–HSA in the B state This finding could indicate that there is a water molecule coordinated at both pH but that its exchange is limiting the relaxivity Therefore, the binding of myristate seems to markedly increase the exchange rate and induce a relaxivity
Trang 9enhancement, although at pH 10.0 the possible increase
in the exchange rate by myristate is incapable to
induce a significant increase of the relaxivity
At pH 10.0 (i.e when HSA is in the B state),
17O-NMR linewidth measurements show no evidence of
water molecules coordinated to Mn(III)heme, as already
observed in the case of Fe(III)heme–HSA Two
hypo-theses should be taken into consideration: either the
absence of water molecules in the coordination sphere
of the metal ion, or the presence of one water molecule
with a very slow exchange rate Myristate binding to
HSA might increase the exchange rate, thereby
produ-cing a small broadening, but this is only observed
at pH 7.0 The structural similarity of Mn(III)heme
vs Fe(III)heme and the structural evidence of a
penta-coordinated Fe(III) atom, with no water molecules
coordinated to it, favour the first hypothesis: in this
case, upon deprotonation at pH 10.0 the phenolic
Tyr161 oxygen becomes more nucleophylic and
displa-ces the Mn(III)-coordinated water molecule with the
consequent quenching of the paramagnetic relaxation
Conclusions
The conformational transition(s) driven by myristate
binding to HSA may be efficiently monitored by
taking advantage of the optical and relaxometric
pro-perties of the Mn(III)heme label Mn(III)heme
binds to FA1 in the fatty acid-free HSA with
KH 2.0 · 10)5m; myristate not only competitively
binds to FA1, but also binds to a different site(s) and
induces conformational changes that lowers the
equi-librium constant for Mn(III)heme binding to the FA1
site by a factor of 10–500 (depending on possible
modulation of myristate binding to FA1) This
con-formational change(s) also favours the exchange of the
Mn(III)-coordinated water molecule with the solvent
bulk At pH¼ 10.0, Mn(III)heme binds to HSA with
higher affinity even in the absence of myristate,
releas-ing the metal-coordinated water molecule
As a general remark, NMRD data prove a valuable
complement to X-ray crystallography to add dynamic
information to structural data, and to provide
thermo-dynamic description of the binding equilibria As an
addition to conventional optical methods, NMRD
pro-vides a useful hint to follow environment changes that
involve the coordination sphere of the paramagnetic
metal
Experimental procedures
All reagents were purchased from Sigma-Aldrich (St Louis,
MO, USA), were of highest purity available, and were used
without further purification HSA was essentially fatty acid-free according to the charcoal delipidation protocol [41–43] and used without any further purification Absence of
MALDI-TOF mass spectrometry Mn(III)heme was pre-pared as previously reported [28] The actual concentration
of the Mn(III)heme stock solution was checked as bis-imidazolate complex in sodium dodecyl sulfate micelles
556 nm) [44] Mn(III)heme–HSA was prepared by adding
NaCl) The final solution of Mn(III)heme–HSA was
Mn(III)heme was present in the reaction mixtures The act-ual concentration of the HSA stock solution was deter-mined by using the Bradford method [45]
The sodium myristate 0.1 m solution was prepared by adding 0.1 m fatty acid to NaOH 0.1 m The solution was
The sodium myristate solution was then mixed with
the desired fatty acid to protein molar ratio The Mn(III)-heme–HSA-myristate complex was incubated for one hour
at room temperature with continuous stirring [6] Sample
pH was changed by adding a few lL of 0.1 m HCl or NaOH solutions
Binding experiments of Mn(III)heme to HSA-myristate and titrations of Mn(III)heme–HSA with myristate were investigated spectrophotometrically using an optical cell with 1.0-cm path length on a Cary 50 Bio spectrophotome-ter (Varian Inc., Palo Alto, CA, USA) In a typical experi-ment, a small amount of a solution of Mn(III)heme in
sol-vent mixture of DMSO-aqueous 0.1 m phosphate buffer
This solution was titrated with HSA by adding small
buffer and recording the spectrum after incubation for
a few min after each addition Difference spectra with respect to Mn(III)heme were taken and the binding iso-therm was analyzed by plotting the difference of absorb-ance between the maximum and the minimum of the two-signed difference spectra against the protein concentra-tion [27]
Data have been numerically analyzed using the matlab language (The MathWorks, Natick, MA, USA) according
to Scheme 1, with the following dissociation equilibrium
myristate binding to site FA1 and competing ipsosterically
bind-ing to the HSA–myristate complex, with myristate bound
Trang 10to FAx Accordingly, the dissociation constant for
myri-state binding to FAx with FA1 occupied by Mn(III)heme
in the model to take into account subtraction of myristate
by additional binding sites no 1 to 5 essentially uncoupled
to FA1 and⁄ or FAx Fatty acid binding sites (FA1 to FA7)
are numbered according to literature [6,7,14]
and at variable pH were obtained on a Stelar
Spinmaster-FFC fast field cycling relaxometer (Stelar, Mede, Italy) with
measurements was ± 0.5%
1
H nuclear magnetic relaxation dispersion (NMRD)
pro-files were recorded at variable concentration of myristate
by measuring water proton longitudinal relaxation rates
pro-ton Larmor frequencies) with the field cycling relaxometer
described above
to the solvent water longitudinal relaxation rate referenced
to a 1.0 mm concentration of paramagnetic agent) were
determined by subtracting from the observed relaxation
for the buffer at the experimental temperature, divided by
Unbound water protons relax by means of diffusion–
controlled dipolar interaction (outer sphere contribution,
to the metal ion or bound to the protein in close proximity
of the paramagnetic center the dipolar interaction is
modu-lated by the reorientation of the macromolecule with
respect to the applied magnetic field The latter term is
des-cribed by Eqn (1):
ð1Þ
water molecules close to the metal centre [M] is the
longitudinal relaxation time of localized water protons
[20,36]
a function of pH was analyzed according to Eqn (2):
i
ð2Þ
17
recor-ded at 7.0 T on a Bruker Avance 300 spectrometer (Bruker
Biospin, Rheinstetten, Germany), equipped with a 5 mm
inner diameter tunable broadband probehead, by using a
iso-topic abundance of 2% Experimental settings: spectral width 6.0 kHz, 90 pulse 16 ls, acquisition time 0.47 s, 128 scans, no sample spinning [20] Paramagnetic contributions
presence of Mn(III)heme–HSA at different myristate
2M)
by Eqn (3):
2Mþ sM
ð3Þ
water molecule, [M] is the concentration of the paramag-netic metal ion, and q is the number of water molecules coordinated to it The oxygen transverse relaxation time
interac-tion, that occurs only in the presence of direct oxygen-water coordination [36]
In all figures, error bars have been omitted for clarity as all errors have been observed to be less than 2% of the measured values
Acknowledgements
This work was supported by the Italian Ministry for Instruction, University and Research Part of the work has been performed at the Bioindustry Park Canavese, Colleretto Giacosa (TO), Italy
References
1 Spector AA (1975) Fatty acid binding to plasma albu-min J Lipid Res 16, 165–179
2 Sudlow G, Birkett DJ & Wade DN (1975) The charac-terization of two specific drug binding sites on human serum albumin Mol Pharmacol 11, 824–832
3 Hamilton JA, Cistola DP, Morrisett JD, Sparrow JT & Small DM (1984) Interactions of myristic acid with
Acad Sci USA 81, 3718–3722
4 He X & Carter DC (1992) Atomic structure and chemis-try of human serum albumin Nature 358, 209–215
5 Peters T Jr (1996) All about albumin: biochemistry, gen-etics and medical applications Academic Press, Orlando,
FL, USA
6 Zunszain PA, Ghuman J, Komatsu T, Tsuchida E & Curry S (2003) Crystal structural analysis of human serum albumin complexed with hemin and fatty acid Struct Biol 3, 6