| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 37, 38844-38853, September 10, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
**
From the
Department of Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, New York 10461, the Department of Biochemistry and Microbiology, Faculty of Sciences and Engineering, Laval University, Quebec G1K 7P4, Canada, the ||Department of Physics-INFM and Center for Excellence in Biomedical Research, University of Genova, Via Dodecaneso, 33, 16146 Genova, Italy, and the ¶G. Gaslini Institute, 16147 Genova, Italy
Received for publication, February 11, 2004 , and in revised form, June 30, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
TrHbN belongs to a family of Mb-like hemeproteins called truncated hemoglobins (trHb)1 (1). It is one of two trHbs expressed in Mycobacterium tuberculosis. TrHbN is expressed during the stationary phase of M. tuberculosis (2-4). It binds oxygen with high affinity because of a very low off rate (2). Inactivation of the glbN gene impairs the ability of stationary phase bacteria to protect aerobic respiration from NO inhibition, suggesting that trHbN may play a vital role in protecting the tuberculosis organism in vivo from NO toxicity (3). This functional assessment is supported by the observation that the NO-dependent oxidation of oxygenated trHbN (Fe(II)O2 + NO 
Fe(III) + NO3-) is very efficient with a second-order rate constant at 23 °C of 745 µM-1 s-1 (3). Introduction to TrHbs—Two distinct groups of Hbs are readily distinguished within unicellular organisms (1). The first one, occurring in bacteria and fungi, comprises the single chain flavohemoglobins that consist of an amino-terminal heme domain displaying the classical "myoglobin fold" and a carboxyl-terminal domain related to ferrodoxin reductases. The second group is the trHbs. These are widely distributed in prokaryotes, unicellular eukaryotes, and plants, forming a distinct group within the Hb superfamily. Many occur in bacteria pathogenic to man. Three groups (groups I, II, and III) can be distinguished within the trHb family. TrHbN is a group I trHb found in M. tuberculosis.
The tertiary structure of trHbN (5) as well as all other crystallographically studied trHbs (6, 7) is based on a 2-on-2 
-helical sandwich, which represents an unprecedented editing of the highly conserved 3-on-3 globin fold of animal and plant Hbs/Mbs as well as flavohemoglobins (1, 8). Two aspects of the tertiary structure of trHbN are likely to have a significant impact on ligand binding kinetics: an apolar "tunnel" linking the solvent to the distal heme pocket and a hydrogen bonding network within the distal heme pocket. Fig. 1, which is derived from the x-ray crystallographic study of the oxy derivative of trHbN (5), depicts both the tunnel and the distal heme pocket residues that can participate in the hydrogen bonding network. The presented kinetic results are analyzed and discussed in the context of these two structural features.
|
A); (ii) geminate recombination originating from ligands localized within the apolar tunnel or xenon cavities prior to either relaxation within the distal heme pocket or the localization of water within the vacated distal heme pocket (C A), and (iii) recombination from the tunnel or cavities when the distal heme pocket has relaxed to the ligand-free conformation and/or is occupied with a water molecule that hinders access to the heme (D A). In general, we attribute the C A process as rebinding from the tunnel/cavities under conditions where the distal heme pocket retains the conformational features of the initial photoproduct species; whereas, the D A process refers to rebinding from the tunnel/cavities subsequent to changes within the distal heme pocket that reduce access of the ligand to the heme-iron. Under conditions where the ligand entry from the solvent (S) into the tunnel or cavities is fast compared with the rate for D A, the ligand combination rate S A will be equivalent to D A. The present study seeks to understand observed ligand reactivity in terms of these kinetic phases.
Approach and Objectives of the Present Study—CO-saturated derivatives of trHbN are photodissociated using laser pulses of 7 ns duration. The CO rebinding to trHbN subsequent to photodissociation is monitored on a time scale from tens of nanoseconds out to hundreds of milliseconds. Sol-gel-encapsulation of trHbN with glycerol as a bathing buffer in combination with temperature tuning is used to vary the viscosity. This approach exposes the distinct kinetic phases introduced above through a systematic tuning of conformational dynamics and relaxation. The roles of ligand and ligation-dependent conformations that are kinetically distinct are explored using encapsulation protocols. Encapsulation protocols that have been successfully used to trap nonequilibrium populations of hemoglobin (9, 10) and myoglobin (10, 11) are used in this study to generate populations of CO-saturated trHbN having not only the equilibrium conformational distribution of CO-liganded trHbN but also CO-saturated forms of trHbN having the equilibrium conformation of either the oxy or deoxy derivative of trHbN. The role of the Tyr(B10) is highlighted through the use of the Tyr(B10) Phe mutant of trHbN. The results support a model in which the conformationally plastic architecture of the distal heme pocket controls both ligand access to the heme-iron and ligand diffusion from the distal heme pocket into the adjacent apolar tunnel. The kinetics also indicate that the apolar tunnel not only functions to create high local concentrations of ligand but also greatly slow the access of water from the solvent into the distal heme pocket subsequent to ligand dissociation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Phe mutant has been previously described (2). The mutant hemoglobins were purified by the same method as described for the wild type recombinant trHbN. Horse heart Mb was commercially obtained from Sigma and used without further purification other than centrifugation to eliminate particulates. Geminate and solvent phase recombination measurements were carried out using 7-ns 532-nm pulses at 1 Hz from a Nd:YAG laser (Minilite, Continuum, Santa Clara, CA) as a photodissociation source and a greatly attenuated continuous wave 442-nm probe beam from a He:Cd laser to monitor time-dependent changes in absorption. Details of the apparatus, data collection, and data display can be found in a previous publication and citations therein (10, 12, 13). The kinetic traces are displayed on a log-log plot of normalized absorbance (proportional to the survival probability of the photoproduct) versus time. The use of the log-log plot helps expose multiple phases. In addition, exponential rebinding shows up as a near vertical line intersecting the time axis at a point that is roughly the inverse of the exponential rate constant. Kinetic phases composed of a large near continuous distribution of rates appear as a linear sloping line on this plot.
Kinetic measurements were typically carried out on solution samples contained in standard 10- or 1-mm stoppered cuvettes placed in a custom-built dry N2 purged variable temperature cuvette holder (-15 to +65 °C). Sol-gel-encapsulated samples were prepared as a thin layer lining the bottom portion of either 5- or 10-mm diameter NMR tubes as previously described (10). Encapsulated forms of CO-saturated trHbN (referred to as trHbN(CO)) were prepared using three different preparative protocols designed to trap both equilibrium and nonequilibrium forms of trHbN(CO). In the first protocol, trHbN(CO) is directly encapsulated. This protocol is designated as [trHbN(CO)] where the species enclosed within the bracket refers to the sample that is initially encapsulated. The second protocol starts with the encapsulation of the deoxy form of trHbN to which CO is added once the sample has aged. This protocol is designated as [trHbN(deoxy)] + CO, where the +CO outside of the brackets signifies addition of CO after the gelation process is complete. The third protocol designated as [trHbN(O2)] + CO is similar to the second protocol except that the oxy derivative is initially encapsulated and then subsequently converted to a CO derivative after aging. The optical spectrum was used to confirm the ligation/ligand status of the sample at all steps. These protocols were used in the attempt to generate CO derivatives of trHbN trapped with the conformational populations associated with the equilibrium CO, deoxy and oxy derivatives of trHbN. A similar approach was successfully applied to a Raman study of the Hb from Ascaris (14). Glycerol-bathed samples were prepared by replacing the aqueous bathing buffer covering the sol-gel sample with an excess volume of either CO- or N2-purged glycerol or glycerol containing buffer. Earlier studies clearly show that the added glycerol increases the viscosity of the encapsulated protein (10, 12, 13). Sol-gel samples, stored at 4 °C, were stable for months or longer with no sign of leakage, oxidation, or sample degradation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
A). No measurable geminate recombination is observed under these conditions. Under comparable conditions, the solvent phase rebinding for C-trHb and P-trHb, the trHbs from Chlamydomonas eugametos (trace a) and Paramecium caudatum (trace b), respectively, are over an order of magnitude faster than for trHbN. In contrast, the solvent phase recombination for Mb (trace d) is about a factor of 10 slower, whereas that of trHbO, the other trHb from M. tuberculosis (trace e) is almost 3 orders of magnitude slower.
|
A) (12). A second slower kinetic phase, designated as C A, is seen to emerge with decreasing viscosity (increasing temperature) (traces b-e). It can be seen that this second slower phase is close to exponential and its amplitude increases at the expense of the faster phase. The kinetic pattern, a composite of B A and C A, shows an overall slow down with decreasing viscosity (increasing temperature) because of an increase in the relative amplitude of the slower C A phase. Nonetheless, the slow exponential-like C A phase appears to get faster with decreasing viscosity (increasing temperature). It should be noted that a comparison of the solution phase (low viscosity) kinetics in Fig. 2 with the high viscosity kinetics reveals that even at its slowest the slow C A kinetic phase seen in Fig. 3 is faster than the solvent phase seen in Fig. 2 (trace c). Thus under these conditions the slowest recombination phase observed under the high viscosity conditions does not match up with the solvent phase rebinding seen under low viscosity conditions.
|
A, the second slower phase with decreasing viscosity (here shown by varying the glycerol content at the same temperature). The samples that were initially encapsulated as either the oxy or deoxy derivatives (traces c and d, respectively) reveal a third still slower exponential phase, designated as D A, in addition to the two phases observed for [trHbN(CO)]. The deoxy turned CO sample shows the largest fraction of the slowest phase. The viscosity/temperature dependence of the three phases is seen in Fig. 5 for the CO recombination traces from [trHbN(deoxy)] + CO. With decreasing viscosity/increasing temperature the slowest phase, D A, gets faster and exhibits a large increase in its fractional contribution to the overall kinetics. The time course for this phase is very close to that of S A observed under low viscosity conditions as seen in aqueous solution (Fig. 2).
|
|
|
Phe mutant under these same encapsulation/solution conditions but at 45 °C. It can be seen from trace a, that mutating the B10 tyrosine to a phenylalanine results in a kinetic trace resembling that seen (trace a of Fig. 3) for the encapsulated wild type protein under higher viscosity conditions (100% glycerol vis à vis 75% glycerol). Trace b in Fig. 7 emphasizes that under these conditions, the wild type shows the second phase while the Phe(B10) mutant does not. Thus this substitution inhibits or delays (with respect to viscosity) the transition from the B A to the C A process.
|
Phe mutant sample shows the onset of a second phase; however, even at the high temperature of 45 °C, this C A phase remains substantially faster than the corresponding C A phase seen for the wild type, see traces d and e in the inset of Fig. 7. It is important to note that this mutant is the only sample that continues to display a prominent fast geminate phase at 45 °C. A comparison of the kinetic traces for the encapsulated samples of the Tyr(B10) Phe mutant, as a function of encapsulation protocol, i.e. [trHbN(deoxy)] + CO, versus [trHbN(CO)], revealed minimal differences. In solution (not shown) this mutant exhibits a small geminate phase and a solvent (bimolecular) phase that is approximately a factor of 30 faster than that of the wild type.
In contrast, the CD1 phenylalanine to tryptophan mutation (trace c) causes the appearance of the slowest phase (D A) as seen for [trHbN(O2)] + CO and [trHbN(deoxy)] + CO. Similar but less pronounced changes (less of a contribution from the slowest phase) were also observed for the Phe(CD1) Leu mutation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The Apolar Tunnel
The x-ray crystal structure of trHbN shows that the very short size of the CD-E segment (4-6 residues, the D helix is totally deleted) pulls the E-helix next to the heme, thus creating a clustering of residues in the region of the distal E-7-mediated pathway to the solvent. The resulting crowding by distal residues blocks access to the distal site cavity through the classical E7 residue gate, typically achieved in vertebrate Hbs by His(E7) (15). A similar pattern is observed for trHbs from Paramecium and Chlamydomonas (P-trHb and C-trHb, respectively) (6). Instead, several trHbs including trHbN have a "tunnel-like" extended network of cavities linking the solvent to the distal heme pocket as shown in Fig. 1. In trHbN this cavity network is composed of two roughly orthogonal branches, one long and one short, converging at the heme distal site from two distinct protein surface access sites. The two branches display inner diameters in the 5-7 Å range for a ligand accessible volume of 330-360 Å3 (5). Residues lining the cavity network are hydrophobic and are substantially conserved throughout the trHb family, suggesting that the apolar tunnel plays a functional role in many trHbs. The dimensions and polarity suggest that it is suited for small nonpolar ligand diffusion or storage.
The nerve tissue Hb from Cerebratulus lacteus (16) and the recently characterized neuroglobins and cytoglobins from higher organisms, display large protein matrix cavities that are structurally unrelated to those of the trHbs. Thus ligand entry/exit routes comparable with those seen for trHbs have not been recognized in other Hb-related proteins. In Mbs there are small protein cavities (often referred to as xenon cavities) originating from packing defects that are accessible to xenon atoms (17, 18) and small nonpolar molecules (19).
Recent studies indicate that xenon cavities in vertebrate Mbs, acting as docking sites for potential heme ligands, can have an effect on ligand binding kinetics (12, 20-29). It has been proposed that these xenon cavities play a functional role in modulating both ligand binding kinetics and multiligand reactions such as NO-mediated oxidation reactions of Mb-O2 (30, 31).
Whereas the kinetic results to date indicate that the small volume xenon cavities in Mb represent a side branch off of the linear ligand diffusion pathway between the solvent and the distal heme pocket (24, 28, 29), the crystallography suggests that the larger volume apolar tunnels (both branches) in trHbs are linearly positioned between solvent and distal heme pocket (5, 6). The positioning, the hydrophobic makeup, and the volume for these tunnels suggest important roles both in controlling substrate access to the heme and in ligand storage or accumulation within the protein. Such a role was invoked to explain the biphasic kinetic pattern observed for the rebinding of CO to P. caudatum trHb under low CO concentrations (13). A more widespread role for these tunnels is also suggested by their presence in other systems. For example, protein matrix tunnels, connecting the surface to active sites in methane monooxygenase hydroxylase and in carbon monoxide dehydrogenase (32), allow for internal substrate translocation in several other proteins (33, 34).
Hydrogen Bonding Networks within the Distal Heme Pocket
One major difference between trHbs and vertebrate Hbs/Mbs is the presence of multiple residues within the distal heme pocket capable of forming a H-bonding network that may or may not include heme-bound ligands (1). Several spectroscopic studies show clear indications not only of hydrogen bonding networks but also of their role in stabilizing heme-bound dioxygen (2, 4, 35-38). Nearly all trHbs have a Tyr(B10) instead of the almost invariant Leu(B10) seen in vertebrate Hbs/Mbs (1). In trHbN the heme-bound O2 is stabilized through direct interaction with the Tyr(B10) phenolic oxygen atom and hydrogen bonding of this oxygen atom to Gln(E11) (5). The crystal structures of ferric cyano-met C. eugametos and aquo-met P. caudatum trHbs show distal site networks based on Tyr(B10)-Gln(E11)/Thr hydrogen bonds (6). Unlike trHbN, which has a leucine at the E7 site, these other ferric trHbs have the potential of an additional hydrogen bond arising from the interaction between their polar Gln(E7) and suitable heme-bound ligands.
Our current hypothesis is that these H-bond networks not only stabilize the bound O2 but also modulate the positioning and dynamics of distal pocket residues that participate in the control of: (i) ligand access to the iron, (ii) ligand diffusion within the distal heme pocket, and (iii) ligand diffusion out of the distal heme pocket into the adjacent tunnel and back from the tunnel into the distal heme pocket. The network is anticipated to be ligand dependent because the hydrogen bonding may or may not include the heme-bound ligand depending on the nature or presence of the ligand. As a consequence, the modulation of the kinetics through the hydrogen bonding network is also expected to be ligand dependent.
A Framework for Discussing TrHb Kinetics
In a previous work (13), the following linear kinetic scheme was used to describe the binding of ligands to trHbs where the solvent (S) and heme (A) are linked through an apolar tunnel (C) connecting the solvent to the distal heme pocket (B).
 | (Eq. 1) |
For trHbN and other Hbs where the cavities or tunnel are not a blind pouch, but are instead part of the linear pathway between solvent and heme, one uses a single reaction coordinate of the kind shown in Fig. 8. Each possible functionally distinct conformation of the protein has associated with its own reaction coordinate. Thus C is used to indicate the situation where the ligand is in the tunnel but the conformation of the distal heme pocket is still that associated with the initial photoproduct; whereas D is the other extreme signifying a ligand in the tunnel but with fully relaxed distal heme pocket as would exist in the equilibrium deoxy population.
|
B A path and the other the C(D) B A path.
Three Kinetic Phases Are Observed for CO Recombination to trHbN
Kinetic traces for CO recombination to different samples of trHbN exhibit one, two, or three distinct phases. The composition of each trace is a function of viscosity/temperature, sample history (in the case of encapsulated samples), and specific mutations. In the subsequent discussion, the following issues will be addressed: (i) an assignment for the individual kinetic phases, and (ii) a model for the kinetics that accounts for the dependence of the kinetics on viscosity/temperature, sample history, and mutagenic modification of the distal heme pocket. The schematic shown in Fig. 8 will be used to facilitate the ensuing discussion.
The Fast Phase: Rebinding from within the Distal Heme Pocket (B A)—The kinetic traces from [trHbN(CO)] at -15 °C in 100% glycerol and [(Tyr(B10) Phe)trHbN(CO)] at -15 °C in 75% glycerol show what is close to a single highly nonexponential rebinding phase that is complete within a few microseconds of photodissociation. Other samples prepared under different conditions or these samples under reduced viscosity conditions also show this phase but with reduced amplitude. It is clear that the amplitude for this phase dramatically decreases with decreasing viscosity. For trHbN, the amplitude for this phase ends up being close to zero for aqueous buffers. The kinetic trace associated with this phase when the amplitude is at or near 100% is very similar to that seen for Mb mutants, where the barrier for B A, Barrier I, is lowered because of the changes in E11 and/or the barrier for B C, Barrier II, is increased because of the dissociated ligand being constrained to remain within the distal heme pocket by the large aromatic side chain of E11 (12). The time scale and the similarity of the trHbN traces to those from the above Mb mutants as well as to traces from other Mbs and Hbs at higher viscosity/lower temperature make it clear that this phase arises from geminate rebinding from within the distal heme pocket. This rebinding process is labeled B A in Fig. 8.
The nearly linear behavior of B A on the log-log plot is consistent with there being a distribution of rate constants on the time scale of the rebinding (39). Similar behavior under higher viscosity conditions has been attributed primarily to a distribution of kinetically distinct conformational substates with each substate having a slightly different kinetic barrier (39).
The Slow Kinetic Phases: Rebinding from within the Apolar Tunnel (C A)—As the viscosity is reduced, the kinetic traces from those samples that exhibit the single B A process begin to display a second distinctly slower phase. This slower phase is still faster, by approximately a factor of 10, than the rebinding process observed in solution. Under comparable viscosity conditions, a similar phase is observed for samples of Mb mutants that have reduced barriers for the B A process and are without residues that significantly block access into or out of the distal heme pocket (12). In the case of Mb and Mb mutants this phase is attributed to rebinding arising from CO localized within the xenon cavities. However, trHbN does not have the well delineated xenon cavities observed for Mb. Instead it has a long branched apolar tunnel that links the solvent to the distal heme pocket. It is therefore viewed as reasonable to assign this second slower exponential phase seen in trHbN to the rebinding of CO localized within the apolar tunnel that is designated as C(D) B A in Fig. 8, which in subsequent discussion is abbreviated to C A. A similar assignment was made for the fast C A phases seen for C-trHb and P-trHb (13).
The Role of Conformational Relaxation: D A—Samples encapsulated as deoxy or oxy prior to CO addition exhibit a slower third kinetic phase designated as D A. This phase is comparable with the solvent rebinding phase observed for trHbN in aqueous solution in the high CO concentration limit. The time scale for that rebinding is also within experimental error consistent with the association rate of the binding of CO to deoxy trHbN (2). These observations coupled with the added observation that this phase increases in amplitude with decreasing viscosity/increasing temperature are all compatible with assigning this slowest phase to rebinding from within the tunnel under conditions where: (i) the heme environment has fully relaxed to a conformation or distribution of conformations associated with the deoxy form of trHbN and (ii) the rate of CO entry into the tunnel from the solvent is not rate-limiting.
The Role of Conformation in Determining Kinetic Patterns
Kinetic traces from CO-saturated samples of encapsulated trHbN are strongly dependent upon sample history. In particular, the fraction of the trace originating from B A, C A, and D A is a function of the initial encapsulated species. The sample initially encapsulated as trHbN(CO) shows the highest fraction of fast phases; whereas the sample initially encapsulated as trHbN(deoxy) prior to addition of CO shows the highest fraction of the slowest phase. The [trHbN(O2)] + CO sample displays an intermediate composition of phases. In all cases the fraction of slow phase increases relative to the preceding faster phase as the viscosity decreases/temperature increases.
The observation that the kinetic trace of the [trHbN(O2)] + CO sample slowly, over a period of a month, evolves to resemble the kinetic trace from the encapsulated equilibrium species, [trHbN(CO)], demonstrates that the sol-gel-encapsulation protocol is successfully trapping unrelaxed conformations and that the kinetic traces are conformation dependent. The most straightforward interpretation is that the sol-gel locks in the conformation of the initially encapsulated species as has been observed for HbA (9, 40-43), Mb (10, 11), transferrin (44), and Ascaris Hb (14). Thus the sol-gel results indicate that ligationand ligand-specific conformations give rise to different kinetic traces under high to moderate viscosity conditions and the resulting nonequilibrium conditions can be maintained for an extended period within the sol-gel.
The presented results show that under conditions where the ligation-dependent conformation is relaxing very slowly, the fractional composition of the kinetic trace is still highly dependent upon viscosity/temperature. It is observed for all samples that as the viscosity decreases, the amplitudes of the slower phases increase relative to the preceding faster phase but the actual composition of the overall kinetic trace is still sample dependent.
This behavior is most directly interpreted using the concept of a hierarchy of conformations and conformational substates (31, 45, 46). The basic idea that has been extensively developed and refined through studies on Mb, is that there are ligation (and ligand dependent) conformations of trHbN. These conformations provide the overall scaffolding that determines the conformation-specific range of accessible conformational substates associated with the localization positions for side chains of the internal amino acid residues. The implication is that the bounded range of accessible relaxed and unrelaxed conformations associated with the architecture of the distal heme pocket is different for [trHbN(CO)], [trHbN(deoxy)] + CO, and [trHbN(O2)] + CO. Importantly, these conformational dependent kinetic behaviors were found to disappear in the Phe(B10) mutant.
Structural Determinants of the Kinetic Phases
It is known primarily from detailed kinetic studies and from low temperature crystallography studies on photodissociated liganded derivatives of mutant Mbs, that the positioning and size of key residue side chains within the distal heme pocket strongly dictate the partitioning of the kinetics into the different phases (12, 24, 25, 27-29, 47). Kinetic differences attributed to differences in the proximal strain have been reported for sol-gel-encapsulated forms of MbCO (10); however, these effects are much smaller that those arising from mutagenic changes in the distal heme pocket architecture. Furthermore, the Raman frequency of the iron-proximal histidine stretching mode for trHbN(deoxy) that is already near the upper limit for either deoxy or photoproduct forms of Mb and Hb (4) is indicative of a favorable strain-free proximal environment. In addition the frequency of this Raman band is the same for both deoxy trHbN and the nanosecond photoproduct from the CO derivative of trHbN both in solution (49) and in the sol-gel either with or without added glycerol.2 These observations all but eliminate proximal relaxation as a significant contributor to the origin of conformation dependent or relaxation dependent kinetic traces from trHbN. As a result, processes associated with the distal heme pocket will be focused upon to account for kinetic effects attributed to relaxation.
Partitioning of the Multiple Kinetic Phases
The ratio of amplitudes for the B A and C/D A kinetic phases is determined by the ratio of the rates for CO rebinding from within the distal heme pocket and CO escape from the distal heme pocket into the adjacent tunnel (B C). The partitioning of the rebinding phases from the tunnel into C A and D A is determined by the relative rates of rebinding from the tunnel versus relaxation within the distal heme pocket. In the following discussion we explore how these processes might be related to the properties and behavior of specific elements of structure.
B A
Overall the B A process is consistent with barrier I being low (see Fig. 8). The very large amplitude for the B A process for trHbN under high viscosity conditions indicates that under those conditions, the rebinding is faster than the B C process. This could occur if the B C process were modulated by the presence of a large side chain blocking access from the distal heme pocket into the tunnel. However, the transition for the relative amplitude of the B A from 100% to near 0% with decreasing viscosity is not consistent with side chains blocking access between the distal heme pocket and the adjacent tunnel/cavity B C. The low amplitude for B A relative to C/D A for trHbN under low viscosity conditions reflects very rapid diffusion of the dissociated CO from the distal heme pocket. Indeed, several studies show that side chain motions are required to facilitate this diffusive process (50-52). The question thus arises as to what drives the viscosity dependent B C process in trHbN.
The Role of the B10 Side Chain in Facilitating B C
Mb mutants having an aromatic side chain at the B10 position exhibit evidence of a rapid B C process (12, 13, 20, 24, 25, 27, 29, 51, 53). Recent time-resolved crystallographic studies of the Mb(Leu(B10) Phe) mutant directly reveals how small motions of the phenyl side chain facilitate the rapid nanosecond diffusion of CO from the distal heme pocket into the adjacent Xe4 cavity (53). Consistent with the B10 side chain enhancing the B C process are the observations reported in several of the above listed references that the amplitude for the B A process is very low for Mb mutants having the B10 leucine replaced with Phe, Tyr, or Trp. Together these findings all show that the aromatic B10 side chains enhance the B C process. Thus it is reasonable to propose that the low amplitude for B A in trHbN under low viscosity conditions to the motions of the tyrosyl side chain of Tyr(B10) with the caveat that the architecture of the distal heme pockets of Mb and trHbN, whereas similar in the overall positioning of residues, still exhibit clear differences in the distances between side chains.
To the extent that the motions of the B10 side chain of trHbN facilitate ligand diffusion, the enhancement of the amplitude of the B A process for the Phe(B10) mutant suggests that the tyrosyl-mediated interactions play a role in the process. One such interaction could be the hydrogen bonding between the B10 tyrosyl side chain and the side chain of Gln(E11). Such an interaction could serve to anchor the B10 side chain and thereby limit its position (as well as that of the E11 side chain). Alternatively, this interaction could position the side chain of Glu(E11) to either facilitate B C directly or inhibit B A through steric effects near the iron. In contrast the side chain for Phe(B10) should have greater access to different loci within the distal heme pocket and not influence the side chain of E11.
The Origin of D A: Relaxation of the Aromatic B10 Side Chain in TrHbN and Mb and the Role of E11 in Limiting Side Chain Relaxation
The viscosity dependence of the D A recombination phase for the encapsulated samples and its evolution into the solvent phase recombination observed in aqueous solvent indicate that this phase arises from a viscosity-dependent relaxation initiated upon ligand dissociation. For the trHbN Phe(B10) mutant, there appears to be only a single C/D A recombination phase that is a factor of 10 faster than the D A phase seen for the wild type trHbN. This observation suggests that the relaxation of the B10 tyrosyl side chain plays a role in the partitioning of the rebinding between C A and D A.
In Mb mutants having an aromatic B10 residue, the rebinding from the solvent or from the xenon cavities is typically very slow. For Mb(Leu(B10) Trp) and Mb(E7H Gln/Leu(B10) Tyr), the CO rebinding to the heme from the xenon cavities is even slower than from the wild type (12, 13). This slow rebinding is attributed to the observed relaxation of the aromatic B10 side chain to a position that blocks access of the ligand to the heme. This relaxation is observed even under cryogenic conditions (27) that accounts for the persistence of the slow D A phase observed for the above two Mb mutants even under high viscosity conditions. In contrast, the C A or D A process in trHbN and other trHbs (with the exception of trHbO that has a G8W residue blocking access to the heme (7, 35, 54)) remains relatively rapid under all conditions. The implication of those observations is that the aromatic side chain in these trHbs does not relax to the position comparable with that in the Mb mutants where the relaxation results in greatly hindered access to the heme site (24).
The viscosity pattern for trHbN as well as for C-trHb and P-trHb suggests that relaxation of the tyrosyl side chain of Tyr(B10) is not as extensive as occurs in Mb mutants having either a tyrosine or tryptophan B10 residue. A plausible origin for this difference is the E11 residue. For these trHbs, the E11 residue is either Gln or Thr, whereas for the Mb mutants it is Val. The crystallography shows that the side chains of the E11 residues in these trHbs can get close enough to the tyrosyl side chain of Tyr(B10) (5, 6) to form a hydrogen bond. It is thus plausible that this hydrogen bonding limits the extent of tyrosyl relaxation subsequent to dissociation and thus minimizes the relaxation induced increase in Barrier I that occurs in the Mb mutants.
The Origin of Conformation-dependent Kinetic Patterns: Ligand-dependent Hydrogen Bonding Networks
The variation in kinetic pattern for the different sol-gel-encapsulated CO derivatives of trHbN shows that the ligand and ligation-dependent conformations exhibit different kinetics. These differences can be explained using an extension of the same arguments given above to account for the differences between the Mb mutants and trHbN. Here the added factor is that both the presence and absence of a ligand as well as the identity of the ligand determine the hydrogen-bonding network that dictates the relaxation properties of the B10 tyrosyl side chain or the E11 side chain. Earlier studies have revealed that there is conformational plasticity within the distal heme pocket of the CO derivative of trHbN as reflected in multiple Fe-C or CO stretching frequencies (4). The two dominant species exhibit frequencies of populations having the tyrosyl proton either close to or distant to the bound CO. The crystallography of trHbN showed that for ligands such as dioxygen that are capable of strong hydrogen bonding, the bound ligand participates in a hydrogen bonding network that can consist of not only the B10 side chain but also E11 (and E7 in cases where E7 is a glutamine) (5, 6). Thus the positioning of the B10 and E11 side chains could depend upon ligation as well as whether the bound ligand is prone to hydrogen bonding as is dioxygen but not CO. Furthermore, it is likely that the orientation of the bound ligand will impact the positioning of the E11 side chain and thus influence its capacity to participate in the hydrogen bonding network. The sensitivity of the kinetics to the distal heme pocket architecture is strongly reflected in the kinetic patterns of not only the different encapsulated CO forms of trHbN but also of the B10 and CD1 mutants of trHbN. The latter mutants likely influence the positioning of the heme with respect to the B10 and E11 side chains.
C/D A: The Role of Water
For native or wild type vertebrate Mb, CO recombination is slowed by water molecules rapidly occupying the vacated distal heme pocket giving rise to a slow D A phase (48, 55-57). This effect is apparent in the CO recombination for Mb in a trehalose glass (12). Under extreme drying, the recombination consists of a B A and C A process with the latter occurring on the order of a millisecond. When the glassy matrix is allowed to take up water, the third phase appears. This slowest phase is comparable in its time course with solvent phase rebinding or the CO combination rate in rapid mixture experiments. The time scale of this slow phase is attributed to the requirement that the water in the distal heme pocket be displaced in order for ligand binding to occur. The time scale for water to enter a vacated distal heme pocket is on the order of tens to hundreds of nanoseconds based on direct time resolved studies (55). This rapid access is attributable to the E7 gating mechanism that links the solvent or surface waters to the distal heme pocket (57).
The C/D A recombination phases for trHbN, P-trHb, and C-trHb are either comparable with or faster than the C A phase in Mb (13). The time course for these fast trHb phases fall within the range of time scales for the recombination from Mb mutants that have distal heme pocket residues with side chains that are both nonpolar and moderate to small in size. It follows that the fast time course for the C/D A recombination phases for trHbN, P-trHb, and C-trHb stems from the absence of steric hindrance as a result of either large residue side chain or water molecules blocking access to the heme from the tunnel.
Because the trHbs in question have polar residues within the distal heme pocket, the issue arises as to why water does not appear to be a factor in slowing trHb recombination as it does for Mb. As noted above, water accesses the distal heme pocket of Mb via the E7 gate. In contrast, the x-ray data on liganded derivatives indicate that solvent waters may access the distal heme pocket of the above trHbs through a long apolar tunnel (5, 6). The large body of existing data supports the following sequence of events subsequent to photodissociation for Mb. When the CO is heme-bound the distal heme pocket is water-free. Upon dissociation, the CO first occupies the water-free distal heme pocket and then diffuses into the adjacent xenon cavities. Once the distal heme pocket is vacated, water can rapidly enter via the E7 gate. The presence of water within the distal heme pocket either slows down the re-entry of the ligand from the xenon cavities or limits ligand access to the heme. In contrast to the Mb situation, the pathway for water entry and ligand migration is one and the same for the trHbs. Thus water must compete with the dissociated ligand not only for occupancy of the distal heme pocket but also for the apolar tunnel. The apolar nature of the tunnel is likely to favor occupancy by the apolar ligand over that of the polar water molecule. The net result is that despite the moderately polar environment within the distal heme pocket of the trHbs, the tunnel pathway results in a relatively slow flow of water from the solvent into the distal heme pocket. The kinetic patterns are consistent with water not being able to access the distal heme pocket on the time scale of ligand rebinding from within the apolar tunnel. The absorption spectra of the Met forms of these trHbs indicate that water does indeed access the distal heme pocket as anticipated based on the side chain composition of the distal heme pocket. Our results indicate that even if water can access the distal heme pocket, it does so at a slow rate that makes water much less of a contributor to the functional properties compared with most other globins.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The kinetic patterns also indicate that the apolar tunnel linking the solvent to the distal heme pocket not only functions as a potential reservoir that can accumulate nonpolar ligands, but also can limit the rate at which water can occupy the polar distal heme pocket when vacated by the dissociated ligand. In this capacity, the tunnel provides a mechanism that allows for fast to very fast recombination for trHbs without interference from water despite a polar environment within the distal heme pocket.
The kinetic pattern under low viscosity conditions reveals a very fast escape of the dissociated ligand from the distal heme pocket with fast recombination from the tunnel. This combination in association with the apolar tunnel that favors nonpolar over polar substrates makes trHbN well suited to functions requiring both rapid binding of nonpolar substrates and rapid elimination of polar product.
The extreme modulation of the rebinding pattern for trHbN through viscosity may have functional significance. M. tuberculosis is known to produce large amounts of trehalose in support of cell wall synthesis. It is also likely that this organism as with many other similar simple organisms increases the production of trehalose in response to stress. High concentrations of intracellular trehalose is known to be the basis for long term survival of organisms adapted to anhydrobiosis by virtue of its glass forming capabilities. It is plausible that the response of trHbN to the increased viscosity associated with high trehalose concentrations is to greatly limit ligand access to the heme, ligand escape from the distal heme pocket, and ligand dissociation. Such an effect would significantly minimize autoxidation of the heme and enhance the stability of NO or O2 bound derivatives during the periods of enhanced concentrations of intracellular trehalose.
|
FOOTNOTES |
|---|
** To whom correspondence should be addressed. Tel.: 718-430-3591; Fax: 718-430-8819; E-mail: jfriedma{at}aecom.yu.edu.
1 The abbreviations used are: trHb, truncated hemoglobins; Hb, hemoglobin; Mb, myoglobin.
2 U. Samuni and J. M. Friedman, unpublished results.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
