Use of relaxation-edited one-dimensional and two dimensional nuclear magnetic resonance spectroscopy to improve detection of small metabolites in blood plasma

The 1H nuclear magnetic resonance (NMR) spectra of biological samples, such as blood plasma and tissues, are information rich but data complex owing to superposition of the resonances from a multitude of different chemical entities in multiple-phase compartments, hampering detection and subsequent resonance assignments. To overcome these problems, several spectral-editing NMR experiments are described here, combining spin-relaxation filters (based on T(1), T(rho), and T(2)) with both one-dimensional and two-dimensional (2D) NMR spectroscopy. These techniques enable the separation of NMR resonances based on their relaxation times and allow simplification of the complex spectra. In this paper, the approach is exemplified using a control human blood plasma, which is a complex mixture of proteins, lipoproteins, and small-molecule metabolites. In the case of T(1rho)- and T(2)-edited 2D NMR experiments, a "flip-back" pulse was introduced after the relaxation editing to make the phase cycling of the "relaxation filter" and the 2D NMR part independent, thus enabling easy implementation of the phase-sensitive 2D NMR experiments. These methods also permit much higher receiver gains to be used to reduce digitization error, in particular, for the small resonances, which are sometimes vitally important for metabonomics studies. Both pulse sequences and experimental results are discussed for T(1)-, T(1rho)-, and T(2)-filtered COSY, T(2)-filtered phase-sensitive DQF-COSY, and T(1), T(1rho)-, and T(2)-filtered TOCSY NMR.     

1H NMR resonance assignments in a paramagnetic heme protein by two-dimensional spectroscopy: heme resonances in equine met-azido myoglobin

Specific heme protons for the majority of resonances in the downfield resolved region of equine met-azido myoglobin have been assigned using solely the two-dimensional 1H NMR experiments NOESY and COSY. Metazido myoglobin provides a useful test case for the applicability of these techniques to paramagnetic proteins for the following reasons. First met-azido myoglobin is a mixed spin-state protein, with significantly shorter relaxation times and broadened lines relative to pure low-spin systems (eg., met-cyano myoglobin). Second, met-azido hemoglobin and met-azido myoglobin are important as models for the physiological forms of hemoglobin. Third, a few sperm whale met-azido myoglobin resonances have been previously assigned, which permits a comparison of assignments for these similar proteins, and a check of the method presented here.     

Studies of individual carbon sites of proteins in solution by natural abundance carbon 13 nuclear magnetic resonance spectroscopy. Relaxation behavior.

The aromatic regions in proton-decoupled natural abundance 13C Fourier transform nuclear magnetic resonance spectra (at 14.2 kG) of small native proteins contain broad methine carbon bands and narrow nonprotonated carbon resonances. Some factors that affect the use of natural abundance 13C Fourier transform NMR spectroscopy for monitoring individual nonprotonated aromatic carbon sites of native proteins in solution are discussed. The effect of protein size is evaluated by comparing the 13C NMR spectra of horse heart ferrocytochrome c, hen egg white lysozyme, horse carbon monoxide myoglobin, and human adult carbon monoxide hemoglobin. Numerous single carbon resonances are observed in the aromatic regions of 13C NMR spectra of cytochrome c, lysozyme, and myoglobin. The much larger hemoglobin yields few resolved individual carbon resonances. Theoretical and some experimental values are presented for the natural linewidths (W), spin-lattice relaxation times (T1), and nuclear Overhauser enhancements (NOE) of nonprotonated aromatic carbons and Czeta of arginine residues. In general, the 13C-1H dipolar mechanism dominates the relaxation of these carbons. 13C-14N dipolar relaxation contributes significantly to 1/T1 of C epsilon2 of tryptophan residues and Czeta of arginine residues of proteins in D2O. The NOE of each nonprotonated aromatic carbon is within experimental error of the calculated value of about 1.2. As a result, integrated intensities can be used for making a carbon count. Theoretical results are presented for the effect of internal rotation on W, T1, and the NOE. A comparison with the experimental T1 and NOE values indicates that if there is internal rotation of aromatic amino acid side chains, it is not fast relative to the over-all rotational motion of the protein.     

Determination of the intracellular pH of intact erythrocytes by 1H NMR spectroscopy

A method is described for determining the intracellular pH of intact erythrocytes by ¹H NMR. The determination is based on the pH dependence of the chemical shifts of resonances for carbon-bonded protons of an indicator molecule (imidazole) in intact cells. The imidazole is introduced into the erythrocytes by incubation in an isotonic saline solution of the indicator. The pH dependence of the chemical shifts of the imidazole resonances is calibrated from ¹H NMR spectra of the imidazole-containing red cell lysates whose pH is varied by the addition of acid or base and measured directly with a pH electrode. To reduce in intensity or eliminate the much more intense envelope of resonances from the hemoglobin, the ¹H NMR measurements are made by either the spin-echo Fourier transform technique or by the transfer-of-saturation by cross-relaxation method.     

1H-NMR investigation of the oxygenation of hemoglobin in intact human red blood cells.

Using improved selective excitation methods for protein nuclear magnetic resonance (NMR), we have conducted measurements of the oxygenation of hemoglobin inside intact human red blood cells. The selective excitation methods use pulse shape-insensitive suppression of the water signal, while producing uniform phase excitation in the region of interest and, thus, are suitable for a wide variety of applications in vivo. We have measured the areas of 1H-NMR resonances of the hyperfine-shifted, exchangeable N delta H protons of the proximal histidine residues of the alpha- and beta-chains in deoxyhemoglobin (63 and 76 ppm downfield from the proton resonance of 2,2-dimethyl-2-silapentane-5-sulfonate (DSS), respectively), which are sensitive to the paramagnetic state of the iron, and for which the alpha- and beta-chain resonances are resolved, and from the ring current-shifted gamma 2-CH3 protons of the distal valine residues in oxyhemoglobin (2.4 ppm upfield from DSS), which are sensitive to the conformation of the heme pocket in the oxy state. We have found that the proximal histidine resonances are directly correlated with the degree of oxygenation of hemoglobin, whereas the distal valine resonances appear to be correlated with the conformation in the heme pocket that occurs after the binding of oxygen, in both the presence and absence of 2,3-diphosphoglycerate. In addition, from the proximal histidine resonances, we have observed a preference for the binding of oxygen to the alpha-chain (up to about 10%) of hemoglobin over the beta-chain in both the presence and absence of 2,3-diphosphoglycerate. These new results obtained in intact erythrocytes are consistent with our previous 1H-NMR studies on purified human normal adult hemoglobin. A unique feature of our 1H-NMR method is the ability to monitor the binding of oxygen specifically to the alpha- and beta-chains of hemoglobin both in solution and in intact red blood cells. This information is essential to our understanding of the molecular basis for the hemoglobin molecule serving as the oxygen carrier in vertebrates.     

Chain-selective isotopic labeling for NMR studies of large multimeric proteins: application to hemoglobin

Multidimensional, multinuclear NMR has the potential to elucidate the mechanisms of allostery and cooperativity in multimeric proteins under near-physiological conditions. However, NMR studies of proteins made up of non-equivalent subunits face the problem of severe resonance overlap, which can prevent the unambiguous assignment of resonances, a necessary step in interpreting the spectra. We report the application of a chain-selective labeling technique, in which one type of subunit is labeled at a time, to carbonmonoxy-hemoglobin A (HbCO A). This labeling method can be used to extend previous resonance assignments of key amino acid residues, which are important to the physiological function of hemoglobin. Among these amino acid residues are the surface histidyls, which account for the majority of the Bohr effect. In the present work, we report the results of two-dimensional heteronuclear multiple quantum coherence (HMQC) experiments performed on recombinant (15)N-labeled HbCO A. In addition to the C2-proton (H epsilon(1)) chemical shifts, these spectra also reveal the corresponding C4-proton (H delta(2)) resonances, correlated with the N epsilon(2) and N delta(1) chemical shifts of all 13 surface histidines per alpha beta dimer. The HMQC spectrum also allows the assignment of the H delta(1), H epsilon(1), and N epsilon(1) resonances of all three tryptophan residues per alpha beta dimer in HbCO A. These results indicate that heteronuclear NMR, used with chain-selective isotopic labeling, can provide resonance assignments of key regions in large, multimeric proteins, suggesting an approach to elucidating the solution structure of hemoglobin, a protein with molecular weight 64.5 kDa.     

Individual assignments of the methyl resonances in the 1H nuclear magnetic resonance spectrum of the basic pancreatic trypsin inhibitor.

In earlier work the resonances of the 20 methyl groups in the basic pancreatic trypsin inhibitor (BPTI) had been identified in the 360-MHz 1H nuclear magnetic resonance (NMR) spectra and most of the methyl lines had from spin-decoupling experiments been assigned to the different types of amino acid residues. The assignments to the different amino acid types were now completed by studies of the saturation transfer between the denatured and the globular forms of the inhibitor and by spin-decoupling experiments in nuclear Overhauser enhancement (NOE) difference spectra. These distinguished between the methyl resonances of Ala and Thr. Furthermore, for most of the methyl resonances, individual assignments to specific residues in the amino acid sequence were obtained from measurements of intramolecular proton-proton NOE's, use of lanthanide NMR shift and relaxation probes, and comparative studies of various chemically modified forms of BPTI. These data provide the basis for individual assignments of the methyl 13C NMR lines in BPTI and for detailed investigations of the relations between the spatial structure of the protein and the chemical shifts of the methyl groups. The methyl groups in BPTI are of particular interest since they are located almost exclusively on the surface of the protein and thus represent potential natural NMR probes for studies of the protein-protein interactions in the complexes formed between BPTI and a variety of proteases.     

Interaction of a spin-labeled phenylalanine analogue with normal and sickle hemoglobins: detection of site-specific interactions through spin-label-induced 1H NMR relaxation

In a preliminary report, we have previously shown that N-[(2,2,5,5-tetramethyl-1-oxypyrrolidin-3-yl)carbonyl]-L-phenyl ala nine tert-butyl ester (SL-Phe) exhibits specific binding to hemoglobin and an antiaggregation activity more than 2 orders of magnitude greater than that of phenylalanine [Lu, H.-Z., Currie, B. L., & Johnson, M. E. (1984) FEBS Lett. 173, 259-263]. Transverse 1H NMR relaxation measurements have been used to investigate the interaction of SL-Phe with hemoglobin molecules by use of the resonances assigned to the C2 protons of the beta 2 His, the beta 143 His, and the beta 146 or beta 97 His residues as intrinsic probes. Distance calculations using the paramagnetically induced relaxation data suggest that the SL-Phe binding site is approximately 12-16 A away from the C2 protons of the beta 2 His and the beta 146 or beta 97 His residues in the (carbonmonoxy)hemoglobin tetramer; for deoxyhemoglobin, the distances are approximately 14-17 A between the SL-Phe binding site and the C2 protons of the beta 2 His, the beta 143 His, and the beta 146 His residues. Calculations using the (carbonmonoxy)hemoglobin crystal atomic coordinates only restrict the probable SL-Phe binding region to the full F and H helices of the beta-chain and a small section of the alpha-chain. For deoxyhemoglobin, the distance calculations provide greater restrictions on the probable binding region, limiting it to small sections of the beta-chain F, G, and H helices near the EF bend and to a few residues on the alpha-chain.(ABSTRACT TRUNCATED AT 250 WORDS)     

NMR investigations of duplex stability of phosphorothioate and phosphorodithioate DNA analogues modified in both strands.

Duplex formation from the self-complementary 12mer d(CGCGAATTCGCG) (Dickerson dodecamer) in which all phosphodiester linkages were replaced by phosphorothioate or phosphorodithioate linkages was studied using variable-temperature 1H and 31P NMR spectroscopy. Melting temperatures of the dodecamer, measured spectrophotometrically, showed significant decrease upon sulfur substitution (Tm 49 degrees C for the phosphorothioate and 21 degrees C for the phosphorodithioate, compared with 68 degrees C for the unmodified oligomer, in 1 M salt). Hyperchromicity observed upon melting of the dithioate was surprisingly low. NOESY spectra of the monothioate showed a cross-peak pattern characteristic for a right-handed duplex. Imino proton resonances of the duplex, shown by the mono- and the dithioate, were similar to those of the parent compound. In spite of monophasic melting curves, temperature dependence of the imino proton resonances and phosphorus resonances of the phosphorodithioate indicated heterogeneity with respect to base-pairing, compatible with the presence of a hairpin loop. Relaxation times (T1) of the imino protons in the phosphorothioate, determined by the saturation recovery method, were considerably shorter than in the unmodified oligomer. Base-pair lifetimes in the unmodified Dickerson dodecamer, determined by catalyst-dependent changes in relaxation rates of imino protons, were in the range of 2-30 ms at 20 degrees C. Strongly reduced base-pair lifetimes were found in the phosphorothioate analogue.     

Behavior of the imino protons of the lambda OR3 17mer studied by high resolution NMR

Imino proton resonances of lambda OR3 17mer were observed with time-shared Redfield pulse method by using a JEOL 500 MHz NMR instrument. They show gradual broadening and disappearance with the elevation of temperature indicating the stepwise melting of the duplex. By the selective irradiation at each peak position nuclear Overhauser effects were observed between the imino and adenine C2H protons and between imino protons themselves. Combining these data, fifteen imino proton resonances could be assigned to each base pair except two terminal AT base pairs. Based on the assignment it can be said that AT rich regions near the terminal melt first, followed by the melting of the inside GC rich core. The two AT base pairs in the middle of the GC core are resistant to heat. Spin lattice relaxation times were also observed and the results are consistent with the melting profile.     

Some NMR Experiments and a Structure Determination Employing a {15N,2H} Enriched Protein

We present the results of studies of an aqueous sample of a highly {15N,2H} enriched protein, the SH3 domain from Fyn. Measurements of 1H relaxation and interactions between H2O solvent and exchangeable protons are given, as well as a method for increasing the effective longitudinal relaxation of solvent exchangeable proton resonances. The long-range isotope shifts are measured, for 1H and 15N, which arise due to perdeuteration. Simulations, which employed a 7 or 8 spin relaxation matrix analysis, were compared to the experimental data from a time series of 2D NOESY datasets for some resonances. The agreement between experiment and simulation suggest that, with this 1H dilute sample, relatively long mixing times (up to 1.2 s) can be used to detect specific dipolar interactions between amide protons up to about 7Å apart. A set of 155 inter-amide NOEs and 7 side chain NOEs were thus identified in a series of 3D HSQC-NOESY-HSQC experiments. These data, alone and in combination with previously collected restraints, were used to calculate sets of structures using X-PLOR. These results are compared to the available X-ray and NMR structures of the Fyn SH3 domain.     

NMR relaxation properties of 77Se-labeled proteins

A 77Se-containing moiety has been attached to cysteine residues in bovine hemoglobin, reduced ribonuclease A, and glutathione by reaction with [77Se]6,6'-diselenobis(3-nitrobenzoic acid). The resultant species contain Se-S linkages that have 77Se NMR absorptions in the range range of 568-580 ppm. Spectra have been recorded at 4.7 and 9.7 tesla (T). For labeled hemoglobin a line width of 250 Hz is seen at 4.7 T and 1000 Hz at 9.4 T. This quadrupling of line width with doubling of observational field strength is consistent with exclusive relaxation by the chemical shift anisotropy (CSA) mechanism. These line widths are greater than expected for a molecule the size of hemoglobin and indicate some aggregation at the high concentrations used. Upon dissociation and partial unfolding of the hemoglobin subunits, the line widths of the selenium resonance decrease to 35 and 120 Hz at 4.7 and 9.4 T, respectively. The spin-lattice relaxation time (T1) for the dissociated hemoglobin at 9.4 T was found to be 220 ms. Together with a value of 377 ms for the spin-spin relaxation time (T2), determined from the line width, an estimate of the CSA was made. This gave a value of 890 ppm, which is in accord with other values for Se(II) linked only by single bonds. When this value for the CSA is used, together with the CSA contribution to the line width, in estimating a correlation time for seleno(3-nitrobenzoic acid) (SeNB)-labeled glutathione, a value of 4 x 10(-11) s is obtained. For SeNB-labeled denatured ribonuclease, four distinct resonances are resolvable at 4.7 T and five resonances at 9.4 T. From T1 values for these resonances and the value of 890 ppm for the CSA, an appropriate correlation time of 0.1 ns was determined, which should result in 77Se resonances of 0.2-1.0 Hz at 4.7 and 9.4 T, respectively. Much greater apparent line widths are observed, which are attributed to microheterogeneity resulting from formation of inter- and intramolecular disulfide linkages. It is concluded that when there are no complications from protein aggregation or chemical exchange, the CSA values anticipated to exist in glutathione peroxidase or other selenoproteins should result in resonances with line widths in the range from 27 to 170 Hz, depending on field strength. These resonances should therefore be observable in the intact protein, if 77Se-enriched material is available.     

Monitoring of hypoxic metabolism in superfused plant tissues by in vivo 1H NMR

We have used a coaxial superfusion system to obtain physiologically interpretable in vivo 1H NMR spectra at 500 MHz of carrot roots, maize roots, and rice shoots in water (no 2H2O). The superfusion system was constructed from common laboratory parts, required no modification of the probe and sample loading procedure, and was inherently leak resistant. The assignment and quantitation of the in vivo 1H NMR resonances were achieved by performing two-dimensional NMR experiments in vivo, and by in vitro analysis including NMR and gas chromatography-mass spectrometry. The in vivo spectra were dominated by resonances arising from sugars, organic acids, amino acids, and ethanol. In vivo measurements of spin-lattice relaxation times and chemical shifts of beta protons of malate in carrot roots suggested that malate was located in a relatively viscous and acidic compartment. In rice shoots, the hypoxic time courses of 9 metabolites were established in vivo, and 23 in vitro. In both cases, accumulation of lactate, ethanol, Ala, and gamma-aminobutyrate as well as a decrease in Gln and Asn concentrations were observed. These findings are consistent with accelerated glycolysis and decreased tricarboxylic acid cycle activity.     

Detection of thymosin beta 4 in situ in a guinea pig cerebral cortex preparation using 1H NMR spectroscopy.

In the present work we have investigated the macromolecules that contribute to the brain 1H NMR spectrum. The cerebral cortex showed distinct resonances at the uncrowded methyl- and methylene chemical shift scale of the spin-echo 1H NMR spectrum. The peaks at 1.22 and 1.40 ppm (relative to the methyl protons of N-acetyl aspartate at 2.02 ppm) arise from cerebral macromolecules without evidence for co-resonances from low molecular weight metabolites as shown by the spin-spin relaxation decays of these resonances. In addition to these NMR signals, peaks at 0.9 and 1.7 ppm from macromolecules were detected. These resonances are from proteins, and we have identified the polypeptides that contributed to the 1H NMR peaks. Two proteins that were present at concentrations of 250 and 350 micrograms/g of dryed tissue showed 1H NMR spectra that resembled the macromolecular pattern in the cerebral 1H NMR spectrum. They were identified as thymosin beta 4 and histone H1, respectively. Thymosin beta 4 was present in soluble high speed cytoplasmic fraction and in P2 pellet, whereas histone H1 was detected in nuclear enriched fraction. A chemical shift-correlated two-dimensional 1H NMR spectrum of thymosin beta 4 in vitro revealed a coupling pattern that matched the macromolecule in the cerebral cortex which we have previously noted (Kauppinen R. A., Kokko, H., and Williams, S. R. (1992) J. Neurochem. 58, 967-974). On the basis of both one- and two-dimensional NMR evidence, subcellular distribution and high concentration, we assign the 1H NMR signals at 0.9, 1.22, 1.40, and 1.7 ppm in the cerebral cortex to thymosin beta 4.     

Water of hydration in the intra- and extra-cellular environment of human erythrocytes

The proton nuclear magnetic resonance (NMR) titration method (which requires measurement of the relaxation rate at multiple measured levels of dehydration) was applied to the analysis of human erythrocytes, a hemoglobin solution, plasma, and serum. The results allowed identification of bulk water and four motionally perturbed water of hydration subfractions. Based on previous NMR studies of homopolypeptides we designated these subfractions as superbound, irrotationally bound, rotationally bound, and structured. The total water of hydration (sum of both structured and bound water subfractions) in plasma, serum, and hemoglobin ranged from 2.78 to 3.77 g H2O/g dry mass and the sum of the three bound water subfractions ranged from 1.23 to 1.72 g H2O/g dry mass. The total water of hydration on hemoglobin, as determined by (i) spin-lattice (T1) and spin-spin (T2) NMR data, (ii) quench ice-crystal imprint size, (iii) calculations based on osmotic pressure data, and (iv) two other methods, ranged from 2.26 to 3.45 g H2O/g dry mass. In contrast, the estimates of total water of hydration in the intact erythrocytes ranged from 0.34 to 1.44 g H2O/g dry mass, as determined by osmotic activity and spin-lattice titration, respectively. Studies on the magnetic-field dependence of the spin-lattice relaxation rate (1/T1 rho) of solvent water nuclei in protein solutions and in intact and disrupted erythrocytes indicated that hemoglobin aggregation exists in the intact erythrocytes and that erythrocyte disruption decreases the extent of hemoglobin aggregation. Together, the present and past data indicate that the extent of water of hydration associated with hemoglobin depends on the amount of salt present and the degree of aggregation of the hemoglobin molecules.     

The amino 1H resonances of oligonucleotide helices: d(CGCG)

An examination of the 1H NMR assignments and exchange properties of the amino resonances of the self-complementary tetramer, d(CGCG) was undertaken with regard to buffer effects, transfer of saturation from the water resonance and temperature dependence of amino 1H line shape and chemical shift. The lack of buffer effect on visible exchangeable proton resonances is evidence for the stringent requirement for nucleo-base protonation at pH values below neutrality, which is greatly reduced in the helical state. For this reason, sharp resonances are observed for both Watson-Crick and non-Watson-Crick cytosine amino protons for base-paired regions. Considerations of monomeric exchange mechanisms for the cytosine and guanine amino protons formed the basis for successful assignment and isolation of their resonances in the helical state by presaturation of the water resonance at selected pH values. Preirradiation of the water resonance at pH less than 6 would isolate the guanine amino 1H resonances of any self-complementary oligonucleotide, to exploit its high sensitivity as a useful proble of helix in equilibrium coil premelting.     

Nondestructive Detection of Glutamate by 1H Nuclear Magnetic Resonance Spectroscopy in Cortical Brain Slices from the Guinea Pig: Evidence for Changes in Detectability During Severe Anoxic Insults

31P and 1H nuclear magnetic resonance spectroscopy (NMR) was used to study the metabolism of intact superfused cortical brain slices during normoxia and anoxia. Attention was focused on quantification of 1H NMR-detected glutamate by a water-suppressed spin-echo method, using N-acetyl aspartate as an internal concentration reference. To quantify the 1H NMR signals, the spin-spin relaxation times and saturation effects were estimated for given metabolites. In addition, absolute concentrations of metabolites were determined by biochemical methods from acid extracts of the preparations after NMR experiments. Under aerobic conditions, 1H NMR detected 79% of the glutamate determined biochemically from the brain slice extracts. During anoxia in the absence of glucose when a severe energetic failure was evident, both 1H NMR and biochemical assays gave closely matching levels for glutamate. We conclude that in the brain cortex 21% of glutamate is located in an intracellular compartment in which this amino acid does not contribute to the 1H NMR signal. However, during severe anoxia an intracellular reorganisation occurs increasing the detectability of this amino acid neurotransmitter by NMR.     

Study of the tryptophan residues of lysozyme using 1H nuclear magnetic resonance.

The identification and complete assignment of the C-2 and N-1 proton nuclear magnetic resonances (NMR) of the six tryptophan residues of hen lysozyme are reported. Identification of the resonances required a detailed examination of the spectra of the protein in H2O and in 2H2O, and involved the application of spin-echo and Carr-Purcell-Meiboom-Gill pulse sequences. Assignment was achieved by observing the effects on the NMR spectra of performing specific chemical modifications, of binding paramagnetic species (lanthanide ions and spin labels), of binding inhibitors and protons and of carrying out solvent exchange experiments. The problems involved in completion of assignment are fully discussed. In the course of performing experiments to make assignments, several interesting aspects of the behaviour of the tryptophan residues in the protein structure were observed and are discussed.     

Proton nuclear magnetic resonance and biochemical studies of oxygenation of human adult hemoglobin in deuterium oxide.

The proton nuclear magnetic resonance spectrum of human adult deoxyhemoglobin in D2O in the region from 6 to 20 ppm downfield from the proton resonance of residual water shows a number of hyperfine shifted proton resonances that are due to groups on or near the alpha and beta hemes. The sensitivity of these resonances to the ligation of the heme groups and the assignment of these resonances to the alpha and beta chains provide an opportunity to investigate the cooperative oxygenation of an intact hemoglobin molecule in solution. By use of the nuclear magnetic resonance correlation spectroscopy technique, at least two resonances, one at approximately 18 ppm downfield from HDO due to the beta chain and the other at approximately 12 ppm due to the alpha chain, can be used to study the binding of oxygen to the alpha and beta chains of hemoglobin. The present results using approximately 12% hemoglobin concentration in 0.1 M Bistris buffer at pD 7 and 27 degrees C with and without organic phosphate show that there is no significant line broadening on oxygenation (from 0 to 50% saturation) to affect the determination of the intensities or areas of these resonances. It is found that the ratio of the intensity of the alpha-heme resonance at 12 ppm to that of the beta-heme resonance at 18 ppm is constant on oxygenation in the absence of organic phosphate but decreases in the presence of 2,3-diphosphoglycerate or inositol hexaphosphate, with the effect of the latter being the stronger. On oxygenation, the intensities of the alpha-heme resonance at 12 ppm and of the beta-heme resonance at 18 ppm decreases more than the total number of deoxy chains available as measured by the degree of O2 saturation of hemoglobin. This shows the sensitivity of these resonances to structural changes which are believed to occur in the unligated subunits upon the ligation of their neighbors in an intact tetrameric hemoglobin molecule. A comparison of the nuclear magnetic resonance data with the populations of the partially saturated hemoglobin tetramers (i.e., hemoglobin with one, two, or three oxygen molecules bound) leads to the conclusion that in the presence of organic phosphate the hemoglobin molecule with one oxygen bound maintains the beta-heme resonance at 18 ppm but not the alpha-heme resonance at 12 ppm. These resluts suggest that some cooperativity must exist in the deoxy quaternary structure of the hemoglobin molecule during the oxygenation process. Hence, these results are not consistent with the requirements of two-state concerted models for the oxygenation of hemoglobin. In addition, we have investigated the effect of D2O on the oxygenation of hemoglobin by measuring the oxygen dissociation curves of normal adult hemoglobin as a function of pH in D2O andH2O media. We have found that (1) the pH dependence of the oxygen equilibrium of hemoglobin (the Bohr effect) in higher pH in comparison to that in H2O medium and (2) the Hill coefficients are essentially the same in D2O and H2O media over the pH range from 6.0 to 8.2...     

Water proton magnetic resonance studies of normal and sickle erythrocytes. Temperature and volume dependence.

The temperature and cell volume dependence of the NMR water proton line-width, spin-lattice, and spin-spin relaxation times have been studied for normal and sickle erythrocytes as well as hemoglobin A and hemoglobin S solutions. Upon deoxygenation, the spin-spin relaxation time (T2) decreases by a factor of 2 for sickle cells and hemoglobin S solutions but remains relatively constant for normal cells and hemoglobin A solutions. The spin-lattice relaxation time (T1) shows no significant change upon deoxygenation for normal or sickle packed red cells. Studies of the change in the NMR linewidth, T1 and T2 as the cell hydration is changed indicate that these parameters are affected only slightly by a 10-20% cell dehydration. This result suggests that the reported 10% cell dehydration observed with sickling is not important in the altered NMR properties. Low temperature studies of the linewidth and T1 for oxy and deoxy hemoglobin A and hemoglobin S solutions suggest that the "bound" water possesses similar properties for all four species. The low temperature linewidth ranges from about 250 Hz at -15 degrees C to 500 Hz at -36 degrees C and analysis of the NMR curves yield hydration values near 0.4 g water/g hemoglobin for all four species. The low temperature T1 data go through a minimum at -35 degrees C for measurements at 44.4 MHz and -50 degrees C for measurements at 17.1 MHz and are similar for oxy and deoxy hemoglobin A and hemoglobin S. These similarities in the low temperature NMR data for oxy and deoxy hemoglobin A and hemoglobin S suggest a hydrophobically driven sickling mechanism. The room temperature and low temperature relaxation time data for normal and sickle cells are interpreted in terms of a three-state model for intracellular water. In the context of this model the relaxation time data imply that type III, or irrotationally bound water, is altered during the sickling process.