Structural studies by proton magnetic relaxation of stereochemical probes. II. Allosteric effects in human methaemoglobin.

The haem-iron accessibility to solvent molecules in human aquomet- and fluoromethaemoglobin was studied by the magnetic relaxation of protons from a stereochemical probe (methanol in deuterated solutions) in its dependence on allosteric effects induced by inositol hexaphosphate and pH between 5.5 and 8.5. The exchange of methanol with bulk solvent was observed only when inositol hexaphosphate was bound to aquomethaemoglobin, which is consistent with a widening of the haemcrevice compared to the conformation in the absence of inositol hexaphosphate. An increase in alkalinity in the physiological range of the Bohr effect results in a gradual impedence of the solvent dynamics inside the haem-pocket. The fast-relaxation phase of methyl protons indicates that a large number of methanol molecules are under the strong influence of the protein; this effect is considerably smaller with inositol hexaphosphate bound to aquomethaemoglobin. The hypothesis which implies a proton from the coordinated water molecule is responsible for the observed relaxation rates has been critically discussed. The model with a water molecule exchanging between a position next to the sixth-ligand site of the haem-iron and the bulk solvent is further substantiated experimentally. This model has been found to be the simplest and most self consistent in the interpretation of all these proton magnetic relaxation data.

A proton magnetic relaxation study of ferric myoglobin and haemoglobin in water/ethanediol solutions.

Structural alterations of the haem vicinity of the high-spin derivatives of bovine ferric myoglobin (metmyoglobin) and human haemoglobin and the changes of the interaction with inositol hexaphosphate induced by ethanediol were monitored by solvent-proton magnetic relaxation. On addition of ethanediol up to 60% the fluoromet derivatives exhibit a gradual increase in the accessibility of the haem for the molecules from the solvent. In aquomethaemoglobin solutions with more than 25% ethanediol there is no unique explanation of proton magnetic relaxation. Ethanediol enhances the binding of inositol hexaphosphate to methaemoglobin, but the structural consequences of this binding on the haem-pockets seem to be diminished. The mechanisms of the observed structural and functional alterations of myoglobin as well as haemoglobin tetramer are discussed here.

Haem accessibility in cytochrome P-450 from rabbit liver. A proton magnetic relaxation study by stereochemical probes.

Cytochrome P-450 was solubilized from phenobarbital induced rabbit liver and purified by affinity chromatography. The longitudinal proton magnetic relaxation rates of this ferric, low-spin sample (as confirmed by ESR) in 20% glycerol aqueous solution are very large compared with low-spin methaemoglobin and myoglobin derivatives. Similarly high rates were measured in a deuterated solution using the aliphatic protons of glycerol as stereochemical markers, which strongly suggests that the haem iron in cytochrome P-450 is much more accessible to the solvent than in harmoglobin or myoglobin. Type I substate (Spasman) produced small but significant increases in NMR rates both in the H2O and in the 2H2O solution, while binding of aniline (Type II substrate) doubled the rates.

Comparison of advanced glycation endproducts on haemoglobin (Hb-AGE) and haemoglobin A1c for the assessment of diabetic control.

Glycation process in vivo results in two different products: early and advanced glycation endproducts (AGEs). The mechanism of early product formation has been well described, with HbA1c as the best-studied example. The finding that advanced glycation endproducts are also formed on haemoglobin suggests that HbA1c is a precursor for Hb-AGE formation. HbA1c has been well established as an important indicator for glycaemia monitoring, but the diagnostic role of Hb-AGE has not yet been clarified. A question is whether HbA1c and Hb-AGE are competitive or complementary parameters. In our study, Hb-AGE was quantified by the competitive ELISA technique using polyclonal anti-AGE-RNase antibodies to detect AGE immunoreactivities of proteins precipitated in red cell hemolysate. Results are expressed as AGE units/mg Hb. Hb-AGE was analysed in three groups of patients divided according to HbA1c values as follows: group I (n = 25) HbA1c < 7%, Hb-AGE = 6.93 (5.7-7.3) U/mg; group II (n = 25) HbA1c = 7-10%, Hb-AGE = 8.62 (7.7-10.2) U/mg; and group III (n = 25) HbA1c > 10%, Hb-AGE = 12.47 (10.8-13.9) U/mg (median (interquartile range)). A close relation between the amounts of red cell HbA1c and Hb-AGE was observed in all diabetic subjects (n = 75) r = 0.77, P < 0.001. Patients with HbA1c level > 8% were considered to be in poor glycaemic control and those with HbA1c < 8% in good control. In the well-controlled subgroup (n = 33), HbA1c and Hb-AGE were less tightly correlated (r = 0.37, P <0.001). However, in those patients with a higher level of HbA1c = 12.55 (8.9-13.3)% (n = 42), the related Hb-AGE was 11.5 (10.3-12.8) U/mg Hb, yielding a more significant correlation (r = 0.51, P < 0.001). The content of Hb-AGE did not correlate with age (r = 0.09), diabetes duration (r = 0.05) or severity of retinopathy and/or nephropathy. The observed difference may reflect a different kinetic rate of HbA1c production and subsequently the rate of Hb-AGE formation. The discrepancy in the correlation between HbA1c and Hb-AGE suggests that they are complementary rather than opposed parameters. The amount of haemoglobin-linked AGEs does not correlate with the presence or absence of retinopathy and/or nephropathy. It seems that Hb-AGE represents only the metabolic status, equally in the subjects with and without diabetic microangiopathy.

Detection of autoantibodies against advanced glycation endproducts and AGE-immune complexes in serum of patients with diabetes mellitus.

Advanced glycation of protein causes their immunogenicity. The evidence that advanced glycation endproducts (AGEs) have antigenic properties has led to a hypothesis that the AGE structure found in vivo may exert an autoimmune response. In the present study, we showed the sera of diabetic patients as well as of nondiabetic individuals to contain autoantibodies to epitopes of AGE structures. Contrary to what might be expected, we observed lower AGE antibody titers in diabetic subjects, and postulated that the antibodies against AGEs form immune complexes in vivo, hampering their determination. The existence of immune complexes containing AGE moiety was established by two independent criteria: (a) serum AGE-immune complexes (AGE-IC) were detected by enzyme-linked immunosorbent assay (ELISA) using an immunochemical bridge; and (b) soluble AGE-IC were precipitated from serum by polyethylene glycol and analyzed. We demonstrated the presence of circulating AGE-IC in sera, predominantly in the sera of diabetic subjects. We also found an inverse correlation between serum AGE level and AGE-IC (r=-0.8, P<0.000), indicating the serum level of AGEs to decline with an increasing presence of AGE-IC. The content of AGE in soluble immune complexes was significantly higher in diabetic patients than in control subjects (3.51+/-1.9 vs. 1.89+/-1.0 microgEq/ml (P<0.00004), and correlated inversely with free antibodies (r=-0.26, P<0.01). Interactions of AGE autoantibodies with AGE as a continuously produced antigen result in the formation of AGE-immune complexes that may play a role in the atherogenic processes.

Ion-exchange chromatography separation of the detergent and the solvent from immunoglobulins after solvent-detergent treatment.

For inactivation of lipid-enveloped viruses during the immunoglobulin production, the solvent-detergent (S/D) method was applied. Tri-n-butyl phosphate (solvent) and Triton X-100 (detergent) were removed from S/D treated immunoglobulins by ion-exchange chromatography on Q-Sepharose Fast Flow (FF). During the chromatographic procedure immunoglobulins remained bound on a Q-Sepharose FF, whereas solvent and detergent were eluted by washing with starting buffer. Elution of immunoglobulins was achieved by increasing the ionic strength of the starting buffer. The final immunoglobulin preparation contained less than 10 microg/ml of Triton X-100 and less than 2 microg/ml tri-n-butyl phosphate. It was confirmed that the S/D procedure did not cause a significant change in polymers and specific antibodies content. Immunoglobulin classes were also not affected by the same procedure.

Removal of detergent and solvent from solvent-detergent-treated immunoglobulins.

The solvent-detergent (S/D) method was applied for inactivation of lipid-enveloped viruses during the production of immunoglobulins. Amberlite XAD-7 resin was used for removal of solvent (tri-n-butyl phosphate, TnBP) and detergent (Triton X-100) after the performed S/D inactivation procedure. The S/D reagents from the immunoglobulin preparation were adsorbed on Amberlite XAD-7, while immunoglobulins passed through the column and retained their biological activity. Using the method developed here, the final immunoglobulin preparation contains less than 1 ppm of Triton X-100 and less than 2 ppm TnBP.

Bisalbuminemia in two Croatian families.

BACKGROUND: Bisalbuminemia is a dysproteinemia characterized by the occurrence of two albumin fractions on serum protein separation by electrophoresis on cellulose acetate sheets. Bisalbuminemia may occur as a hereditary trait or as analytical interference with some drugs, especially penicillin. METHODS: Two patients with the finding of bisalbuminemia are presented. Both patients (patient 1 was a 4-1/2-month-old male infant, and patient 2 was a 15-year-old boy) were admitted for respiratory infection. RESULTS: Bisalbuminemia was detected by serum protein electrophoresis and confirmed by isoelectric focusing in pH gradient gel (pH range 4.0-6.5). This finding was supported by simultaneous detection of abnormal albumin in the mother of patient 1, while the father had normal albumin. The abnormal fast albumin in both patients had an increased relative mobility of 1.08 when measured from the sample application position. CONCLUSIONS: The results presented are the first description of albumin mutations in Croatia (that according to the CISMEL group could be classified as ZC/HZ), and present the first step in identification prior to determination of structural change and amino acid sequence in the albumin molecule.

Characterization of F VIII concentrates produced by two methods incorporating double virus inactivation.

The trend toward the production of high purity factor VIII concentrates for clinical use is still in progress. Although all plasma derivatives must undergo viral inactivation procedures, the possibility of transmission of viral diseases is not completely eliminated. In order to reduce such risk, we have included double virus inactivation in the procedure of factor VIII concentrate production. In a scale-up procedure for isolation of factor VIII cryoprecipitate, two methods were used. The first is based on the chromatographic purification of factor VIII after pasteurization of cryoprecipitate solution and solvent/detergent (S/D) inactivation of viruses. The second is based on multistep precipitation of factor VIII by sodium chloride and glycine. Viral inactivation was performed by combination of S/D treatment and heating of final freeze-dried product 30 min at 100 degrees C. The typical yield of factor VIII activity in the freeze-dried product was about 20% for the first method, and 25-30% for the second. Electrophoretic analyses of both factor VIII preparations by SDS-PAGE and IEF show very low content of contaminant proteins, in accordance with observed 400-650-fold increase of their specific activity over plasma. Both factor VIII products were stable in the liquid state for more than 24 h at room temperature. The final products, after double viral inactivation, are considered to be suitable for clinical evaluations.

In vitro and in vivo metabolisms of K-48.

Metabolic pathways of the oxime K-48 have been elucidated by means of in vitro and in vivo experiments. K-48 exposure to rat liver microsomal fraction resulted in the formation of a hydroxylated derivative, in addition to a small molecular fragment. The in vivo metabolism in rats was investigated after intramuscular administration of 50 mumol oxime. K-48 was present in the rat serum in unchanged form. Similarly, the analysis of rat cerebrospinal fluid indicated the sole occurrence of unchanged K-48. In contrast, unchanged K-48 was not found in the rat urine, where only the metabolite generated by epoxidation on the alkyl chain connecting the two pyridinium rings was present. The presence of unchanged K-48 in the serum and cerebrospinal fluid facilitates quantitative determination using HPLC separation and ultraviolet absorbance detection.

Resonance Raman studies of nitric oxide binding to ferric and ferrous hemoproteins: detection of Fe(III)--NO stretching, Fe(III)--N--O bending, and Fe(II)--N--O bending vibrations.

The nature of bonding interactions between Fe(III) and NO in the ferric nitrosyl complexes of myoglobin (Mb), hemoglobin A (HbA), and horseradish peroxidase (HRP) is investigated by Soret-excited resonance Raman spectroscopy. On the basis of 15NO and N18O isotope shifts, we clearly identified the Fe(III)--NO bond stretching frequencies at 595 cm-1 (ferric Mb X NO), 594 cm-1 (ferric HbA X NO), and 604 cm-1 (ferric HRP X NO). The Fe(III)--N--O bending vibrations are located at 573 cm-1 (ferric Mb X NO) and 574 cm-1 (ferric HRP X NO), which are very similar to the Fe(II)--C--O bending modes at 578 cm-1 in Mb X CO and HbA X CO. However, the Fe(III)--NO and Fe(II)--CO stretching frequencies differ by approximately equal to 90 cm-1, indicating a much stronger iron-axial ligand bond for the [Fe(III) + NO] system, which is isoelectronic with the [Fe(II) + CO] system and, hence, presumably also has a linear Fe(III)--N--O linkage (in the absence of distal steric effect). The unusually strong Fe(III)--NO bond may be attributed to the pi bonding involving the unpaired electron in the pi (NO) orbital. The N18O isotope shift data indicate that the widely accepted assignment of the Fe(II)--NO stretching vibration at approximately equal to 554 cm-1 in ferrous nitrosyl Mb/HbA is incorrect; instead, we assign it to the Fe(II)--N--O bending mode. The validity of the assignment of Fe(II)--O2 stretch at 567 cm-1 in oxy-HbA by Brunner [Brunner, H. (1974) Naturwissenschaften 61, 129-130] is now in doubt. Literature data are presented to suggest that it is the Fe(II)--O--O bending vibration.

Proton magnetic relaxation dispersion in human fluoromethaemoglobin solutions.

The solvent proton spin-lattice relaxation time of high spin Fe3+ (S=5/2) human A fluoromethaemoglobin aqueous solutions was measured at 14 Larmor frequencies in the range from 2.2 to 96 MHz. The observed paramagnetic relaxation rates are analysed in terms of the Solomon-Bloembergen theory, with the g-tensor value of 2 based on the consideration of the protein tertiary structure. From the H2O (pH 6) haemoprotein solution relaxation data, tau(c) =(9.3+/-0.3) X 10(-10) sec. If the total relaxation rates are corrected for the "outer-sphere" paramagnetic contribution, tau(c)=(6.5+/-0.4) X 10(-10) sec. The latter correction is obtained from the p.m.r. of the non-exchangeable aliphatic protons of C2H4(OD)2 added to the D2O-solution of fluoromethaemoglobin. Assuming that single proton transfer is taking place through the protein channel along the axis normal to the haem (g=2), the protein "binding" site is at a distance of 3.93 to 3.98 A from the haem Fe3+ ion.

Proton magnetic resonance relaxation in Pseudomonas aeruginosa cytochrome oxidase solutions.

We have measured the temperature and frequency dependence of solvent proton magnetic relaxation rates in solutions of Pseudomonas aeruginosa cytochrome oxidase (EC 1.9.3.2) in its native low spin oxidized, its reduced, and its carbonyl reduced derivative. In solutions of the native oxidized enzyme, a large paramagnetic enhancement of the proton NMR relaxation rates, propagated to the solvent by the fast exchange mechanism, is observed. The ratio (T1/T2)pmg = 1.25 +/- 0.10 at 24 MHz demonstrates that dipole-dipole interaction of the neighboring paramagnets is the dominant relaxation mechanism. Measurements of proton relaxation in solutions of cytochrome oxidase from which the hemes D have been extracted demonstrates that hemes C do not contribute to the observed paramagnetic effects. The electron spin relaxation time of the ferric hemes D of 3.2 +/- 0.4 ns is calculated from the frequency dispersion data. This is the longest value reported for hemoprotein solutions so far. These features of a low spin ferric hemoprotein are similar to those found recently both for the microbial and for the microsomal cytochrome P-450. The calculated distances between the exchanging proton(s) and heme D iron ions demonstrate the high accessibility of the environment of heme D from the solvent side, also for molecules not penetrating the inner coordination sphere.

Rat tissue collagen modified by advanced glycation: correlation with duration of diabetes and glycemic control.

Collagenous proteins are especially prone to nonenzymatic glycation, because they contain several dibasic amino acid residues with free amino groups, have a very slow turnover rate, and are exposed to ambient levels of glucose. The aim of this study was to determine the time-dependent course of advanced glycation process in diabetic rats in relation to glycemic control and duration of diabetes, compared to age-matched controls. Immunochemical assay with antibodies to advanced glycation end products (AGE) was first developed to qualitatively detect and quantify the AGE formed in rat tendon and aortic collagen. Individual collagen samples were extracted by extensive pepsin and collagenase digestion. The amount of AGE was measured by competitive ELISA and results were expressed as AGE U/mg collagen. Diabetic rats showed a significant increase in AGE content in aortic collagen at 20 weeks (n = 6, 206.6 +/- 16.7 U/mg collagen) compared with that measured at 4 and 12 weeks (n = 6, 110 +/- 12.8 U/mg collagen, and n = 13, 184.9 +/- 12.3 U/mg collagen at 4 and 12 weeks, respectively; p < 0.001 between 20 weeks and 4 weeks; p < 0.01 between 20 weeks and 12 weeks). The amount of AGE in tendon collagen of diabetic rats increased from 1.9 +/- 0.38 U/mg at 4 weeks to 11.2 +/- 6.1 U/mg collagen at 20 weeks, p < 0.001. Considerable disparity was observed in the intensity of glycation between aortic and tendon collagen. AGE-content per mg of aortic collagen was several-fold to that found in tendon collagen (p < 0.001). To investigate the effect of glycemic control on the advanced glycation process, total aortic AGE-collagen content was compared between untreated diabetic rats (D; n = 13, 184.9 +/- 12.3 U/mg) and diabetic rats treated for 12 weeks with insulin (DI; n = 6, 133.9 +/- 10.7 U/mg), or phlorizin (DP; n = 6, 132.4 +/- 8.9 U/mg), or by a combination of insulin/phlorizin (DIP; n = 6, 124.3 +/- 6.5 U/mg). In spite of therapy used, all groups of diabetic animals had a significantly higher aortic AGE-collagen content than those in the nondiabetic control group (C: n = 8, 104.6 +/- 14.9 U/mg) of the same age (D, DI, DP, DIP vs. C, p < 0.001). Comparison between the mean levels of glycated hemoglobin (D: 5.62 +/- 0.38 % vs. C: 1.7 +/- 0.05%) and mean AGE levels in the studied group of animals yielded a very good exponential correlation (r = 0.89, p < 0.001). Glycation-derived late-stage collagen modification was detected by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and by immunoblotting confirmed to contain (an) AGE-structure(s). Our study provides strong immunochemical evidence of AGE formation in vivo during hyperglycemia, and of their temporal association with structural alterations of extracellular matrix proteins. The advanced glycation process is retarded and reduced in intensity, but not completely abolished, by glycemia regulation with, or independently of, insulin.

Iron-carbon bond lengths in carbonmonoxy and cyanomet complexes of the monomeric hemoglobin III from Chironomus thummi thummi: a critical comparison between resonance Raman and x-ray diffraction studies.

Soret-excited resonance Raman spectroscopy yields direct information regarding the iron-carbon bonding interactions in the cyanomet and carbonmonoxy complexes of hemoglobin III from Chironomus thummi thummi (CTT III) in solution. By isotope exchange in cyanide (13CN-, C15N-, and 13C15N-) and carbon monoxide (13CO, C18O, and 13C18O), we have assigned the Fe(III)-CN- stretching at 453 cm-1, the Fe(III)-C-N- bending at 412 cm-1, the Fe(II)-CO stretching at 500 cm-1, the Fe(II)-C-O bending at 574 cm-1, and the C-O stretching at 1960 cm-1. The resonance Raman data, in conjunction with those obtained from heme model complexes with well-known Fe-C bond distances, strongly suggest that the Fe(III)-CN- bond (approximately 1.91 A) is longer (hence weaker) than the Fe(II)-CO bond (approximately 1.80 A). This result disagrees with those of x-ray crystallographic studies [Steigemann, W. & Weber, E. (1979) J. Mol. Biol. 127, 309-338] in which the Fe-C bond lengths were reported as 2.2 A in cyanomet and 2.4 A in carbonmonoxy CTT III. Based on Badger's rule and normal mode calculations, the x-ray data would lead to the prediction of 279 cm-1 for the Fe(II)-CO stretching frequency in CTT III . CO, which was not observed. On the other hand, we estimate the Fe-CO bond as approximately equal to 1.82 A, which is very similar to the 1.80-A value in human Hb . CO crystals. Furthermore, we have used isotope shift data to estimate the Fe-C-O angle as 169 +/- 5 degrees, somewhat larger than the 161 degrees value found by Steigemann and Weber. We therefore conclude that there must be errors in the x-ray crystallographic refinement for the ligand geometry in carbonmonoxy and cyanomet CTT III.

Selective pigment medium for Streptococcus agalactiae.

Increased awareness of the importance of Streptococcus agalactiae (Group B streptococcus) in neonatal morbidity and mortality has demonstrated the need for a selective and differential primary plating medium. In this report we describe a selective pigment medium containing colistin and nalidixic acid that allows presumptive identification of S. agalactiae from the primary plate, even when specimens are polymicrobic. A preliminary study compared the value of three media for pigment production by 137 known S. agalactiae strains. This was followed by a clinical laboratory trial using the best of these pigment media at an obstetric hospital in which 581 specimens from neonates, female reproductive tracts, and urinary tracts were screened. Subsequent to the clinical laboratory trial, an epidemiological investigation using the selective pigment medium was conducted on a mother/infant population. Of the 1,331 specimens screened for S. agalactiae, 75 (5.6%) were positive on 5% sheep blood agar and 79 (5.9%) were positive on the selective pigment medium.

The haem-accessibility in leghaemoglobin of Lupinus luteus as observed by proton magnetic relaxation.

Using the solvent-protons' longitudinal magnetic relaxation rates (p.m.r.) for Lupinus luteus leghaemoglobin derivatives the accessibility of the haem has been evaluated by our "stereo-chemical p.m.r. titration" method with nonexchangeable protons of aliphatic lower alcohols in otherwise deuterated solutions. The haem in leghaemoglobin is more accessible and its protein environment more flexible compared with vertebrate haemoglobins. The correlation time in aquometleghaemglobin aqueous solution has been determined by measuring the frequency dispersion of the p.m.r. rates between 6.1 and 93 MHZ. Taking into account the measured value of tauc = (7.7 +/- 0.5 x 10(-10) s the iron-to-proton inter-spin distances have been calculated. The significance of these distances as well as the electronic g-factor anisotrophy for elucidation of fine structural details of the haem-environment are discussed.

Comparison of microsomal and solubilized monooxygenases from rat and rabbit by proton magnetic relaxation.

The paper presents results of a comparative study of the haem environment, by proton magnetic relaxation, in P-450 and P-448 monooxygenases from rat and rabbit, induced by phenobarbital and 3-methylcholanthrene, in both species. It was established that the method yields information on the accessibility of the haem iron for solvent molecules (protons), both in microsomes and in solubilized samples of various degrees of purification, i.e. association. The state of micelles in the solutions does not alter the haem iron accessibility. A slight difference was found for the microsomes suspended in a phosphate vs. pyrophosphate buffer, but this is without any consequence with regard to the species and form differences. The correlation time for the highly purified LM2 fraction of rabbit P-450 could not be determined more precisely than before for a sample of lower purity, because the relaxation rates are frequency independent. The correlation time for the rat P-448 monooxygenase was determined by dispersion measurements to be (4.1 +/- 0.4) x 10(-11) s. It was found that the PMRx behaviours of rabbit and rat monooxygenases are more alike in microsomes than in the partially purified solubilized form. The solubilization produces a pronounced alteration of the PMRx temperature dependence only for the rat 3-MC induced monooxygenase P-448. For the P-450 form the haem iron becomes less accessible on solubilization, both for the rabbit and the rat liver monooxygenases, whereas in case of rat liver P-448 the accessibility is considerably enhanced on solubilization. There is a substantial structural specificity of the haem environments from the two animal species, the one from rat being tighter. The reduced, NO-bound rabbit liver monooxygenase was studied also, but the results are not yet conclusive, except the fact that the unpaired spin from NO is thoroughly shielded from the solvent compared with the haem iron from the original sample. The following series of increased haem-iron accessibility emerges from the PMRx studies known so far: rat (P-448) less than rabbit (P-448) less than rat (P-450) less than rabbit (P-450) in microsomes, and rabbit (P-448, with 3-MC bound?) less than Pseudomonas putida (P-450) rat less than (P-448), less than rat (P-450) less than rabbit (P-450) from solubilized samples. For the latter, it appears that increased enzymic specificity goes along with a closing of the haem cleft.