hyperlipidaemia. * Address for correspondence: Medizinische Universitdtsklinik,. Josef-Schneider-Str. 2, 8700 W~rzburg, Germany. Received 4 December 1992; ...
MAGMA (1993) 1, 55-60
31p-NMR spectroscopy of human blood and serum: first results from volunteers and patients with congestive heart failure, diabetes mellitus and hyperlipidaemia Michael Horn*, Stefan Neubauer, Michael Bomhard, Marcus Kadgien, Klaus Schnackerz and Georg Ertl Departments of Medicine and Physiological Chemistry, University of W~irzburg, Germany
31P-containing metabolites in human blood, serum and erythrocytes were measured or calculated. Phosphodiesters were found in serum, but not in erythrocytes. 2,3-diphosphoglycerate and 2,3-diphosphoglycerate/ATP ratios were increased in patients with congestive heart failure (2,3-diphosphoglycerate by 13% in mild to moderate, 31% in severe congestive heart failure, 2,3-diphosphoglycerate/ATP ratio by 9% in mild to moderate, 38% in severe congestive heart failure); phosphodiesters were increased in diabetes mellitus (by 26%) and even more so in hyperlipidaemia (by 57%). Changes of blood 3~p compounds with disease states may have diagnostic potential and should be recognized for correction of organ spectra.
Keywords: congestive heart failure, diabetes mellitus, hyperlipidaemia, 31p NMR spectroscopy, blood, serum.
INTRODUCTION
METHODS AND MATERIALS
The presence of blood inevitably contributes signal to in-vivo 31p-NMR spectra of h u m a n organs to a greater or lesser extent. Thus, definition of absolute and relative (ratios) amounts of 3~p_containing metabolites in human blood and serum is important. Blood correction of organ spectra has been previously applied based on an estimation of the 2,3-diphosphoglycerate/ATP ratio in blood from healthy individuals [1-3]. However, thorough quantitative analysis of 31p-compounds in blood and serum from volunteers and patients has been unavailable. Based on such information, appropriate blood correction of 3~p_ s~ectra from organs should be feasible. In addition, 3 P-NMR of blood and serum may by itself bear diagnostic and, possibly, prognostic potential for a variety of disease states. In this work, we characterize 31p-NMR blood spectra from volunteers and patients with congestive heart failure, diabetes mellitus and hyperlipidaemia.
Blood was drawn by venous puncture between 7.30 and 8.00 AM before patients had breakfast. Coagulation was inhibited with Na4EDTA. 2.5 ml of blood was used to obtain 31p-spectra; serum was obtained by centrifugation of 10 ml of additional blood. Samples were immediately placed on ice and transferred to the NMR unit within 30m in. Preliminary testing for stability of 31p-compounds in blood showed unchanged signals within 12h at 0°C or within 4h at 37°C. After rewarming to 37°C, one 31p-NMR spectrum was obtained (1032 acquisitions, 42min, interpulse delay 2.1 s, pulse angle 45°). Then, another 3~p-NMR spectrum of serum from the same subject was acquired using identical pulse parameters. 31P-NMR spectra were obtained on a Bruker AM 300 SWB (7.05 T magnet, Aspect 3000 computer) using a standard 10 mm multinuclear probe. Before acquiring spectra, the homogenity of the B0-field was optimized by shimming on the 2D-lock level. Data are corrected for partial saturation (1.03 for 2,3-DPG, 1.00 for PDE, y-P-, a-P- and ~-P-ATP). Saturation factors were obtained from fully relaxed spectra (TR = 24 s,
* Address for correspondence: Medizinische Universitdtsklinik, Josef-Schneider-Str. 2, 8700 W~rzburg, Germany. Received 4 December1992; accepted27 February 1993. 0968-5243 © 1993 Chapman &Hall
M HORN et al.
56 = 45°). The difference in the areas of the three P-ATP-signals is caused by some overlap with other 31p-compounds, e.g. ADP, N A D P a n d other diphosphoesters. H o w e v e r , the 7-P-ATP has the best S/N d u e to only one P-P-coupling (with fl-P-ATP). In addition, y-P-ATP is d o s e d to the excitation f r e q u e n c y thus no effects of insufficient excitation cause loss of signal. Absolute quantitation of c o m p o u n d s was achieved b y comparing resonance areas with a O , O ' - d i m e t h y l methylphosphonate (3.64 x 10 -s mol) standard (6 = 38.5 p p m versus 85% H3PO4) which was placed in a D20-containing concentric c o m p a r t m e n t a d d e d to the NMR tube. Concentrations of 2,3-DPG, PDE a n d ATP in erythrocytes w e r e calculated according to:
2,3-DPGe = 2,3-DPGB/(1 - Hkt)
(1)
7-P-ATP e = 7-P-ATPb/(1 - Hkt)
(2)
PDEe = (PDEb - (1 - Hkt) X PDEs)/Hkt
(3)
w h e r e Hkt = blood hematocrit, e, b, s = concentration of c o m p o u n d s in erythrocytes, blood and semi. Cell c o u n t a n d routine s e r u m chemistry were determ i n e d from simultaneously d r a w n blood samples (hematocrit (Hkt), erythrocytes (Coulter), glucose,
T a b l e 1.
triglycerides (GPO-PAP m e t h o d , Boehringer Mannheim Diagnostica) a n d cholesterol (CHOD-PAP m e t h o d , Boehringer M a n n h e i m Diagnostica)). The free ATP-concentration was calculated according to the m e t h o d [4-6] of Gupta et aI.
Patient groups Blood specimen were obtained from healthy volunteers (VOL; n = 13, age = 25.3 _+ 1.2) and from three groups of patients. Patients with chronic congestive heart failure (CHF) according to the N e w York H e a r t Association (NYHA) classification (class II or III = mild or moderate, CHF~, n = 9, age = 69.3 + 4.9; class IV = severe, CHF s, n = 5, age = 59.0 + 9.3). All patients h a d evidence of left ventricular dysfunction u p o n echocardiography or radio-contrast ventriculography. All 14 patients w e r e on diuretics, eight had digitalis, ten an ACE-inhibitor, five Warfarin and two an antiarrhythmic agent. Patients with diabetes mellitus (DM; n = 5, age = 63.4 _+ 6.3, one patient with type 2a, four patients with type 2b) w e r e all i n s u l i n - d e p e n d e n t for 12.7 _+ 3.9 years. These patients had n o evidence of heart disease, a s e r u m creatinine of 0.94 + 0.08, and n o excessive hyperlipidaemia (Table 1). Two h a d peripheral occlusive disease.
31p-NMR s p e c t r o s c o p i c data and results of cell c o u n t and s e r u m c h e m i s t r y . VOL
n Age
7-P-ATPb (mM) ~-P-ATPb (mM) 2,3-DPG//~-P-ATPb PDE/y-P-ATPb PDE///-P-ATPb Pi s (mM) PDEs (mM)
2,3-DPGe b (raM) 7-P-ATPe b (raM) ~ - P - A T P e b (mM) /~-P-ATPe b (mM) 2,3-DPG/y-P-ATPe b 2,3-DPG//~-P-ATPe b Hkt (%) Glucose (mg%) Triglycerides(mg%) Cholesterol ( m g % ) Free ATP (%)
13 25.3 _+ 1.2 1.68 +- 0.28 1.45 _+ 0.25 8.85 _+ 0.45 1.36 + 0,13 1.58 _+ 0.14 1.22 + 0.12 4.69 _+ 0,69 28.25 + 3,03 3.85 _+ 0.54 5.00 _+ 0.58 3.28 _+ 0,46 8.82_+ 1.02 10.24 _+ 1.01 44.87_+0.96 91.08 + 4.39 119.9_+ 15,2 195.6 -+ 8.9 20.3 _+ 5,5
CHFm 9 69.3 _+ 4.9 1.84 _+ 0.29 1.39 _+ 0.09 10.13 + 0.75 1.26 _+ 0.21 1.47 _+ 0.20 1.00 _+ 0.10 3.74 + 0.46 36.28 _+ 1.77" 4.72 _+ 0.48 5.83 _+ 0,47 3,70 _+ 0,27 8.18-+0.74 10.14 _+ 0.76 38.13-+2.13 a 108.86 + 18.7 96.0+_ 14.1 179.7 -+ 17.9 14.7 _+ 1,4
CHF~
HL
DM
5 59.0 _+ 1.2 1.61 + 0.13 1.35 + 0.13 1 2 . 6 5 + 2.48 a 1.12 -+ 0.20 1.35 _+ 0.24 0.83 +- 0.17 a 4.14 _+ 0.39 36.49 _+ 0.29 3.69 --+ 0.29 5.40 _+ 0.32 3.10 -+ 0.29 1 0 . 3 8 + 1.70 12.66 _+ 2.48 43.82-+1.87 132.8 + 36,5 ~ 106.6-+ 18.4 182.6 _+ 25.5 13.9 _+ 2.6
5 53.4 + 7.3 1.90 _+ 0.52 1.52 _+ 1.64 8,47 _+ 0.60 2.32 -+ 0.37 a 2.33 _+ 0.19 a 1.83 _+ 1.61 a 8.58 _+4.27 a 27.87 + 2.44 4.65 + 1.13 5.00 _+ 0.27 3.79 _+ 0.34 8.11 _+ 1.64 8.47 _+ 0.81 40.16+2,43 ~ 156.4 + 30,2 a 635.6_+ 110.5 a 345.6 _+ 27.5" 12.6 _+ 2.5
5 63.4 + 6,3 1.44 _+ 0.22 1.13_+0.17 9.95 _+ 0,28 3.30 + 1.24 a 4.17 _+ 1.55 a 1.57 _+ 0.14 5.69 _+ 0,89 28.89 _+ 1.82 4.10-+0.19 5.09 + 0.32 3.12_+0.17 7.71 -+ 0.30 9.86 + 0.24 35.74 -+ 4.37 a 178.8 _+ 22.9 a 217.0 -+ 18.4 a 225.6 _+ 33.1 13.0_+ 1.5
Results are mean + SE. a p < 0.05 versus VOL, unpaired one-tailed t-test. b Calculated e, b, s = concentration of compounds in erythrocytes, blood and serum.
MAGMA (1993) 1(2)
3~p-NMR-SPECTROSCOPY OF HUMAN BLOOD AND SERUM
Patients with hyperlipidaemia (HL; n = 5, age = 53.4 + 7.3) had no evidence of heart, kidney or liver disease. Classification according to Fredrickson showed four patients with type IV and one patient with type V. None of them was treated with lipidlowering agents. Patients suffering from more than one of the disease states listed above were excluded.
STATISTICAL ANALYSIS For each group of patients, data for each parameter obtained were compared with those from volunteers using the unpaired, one-tailed t-test. Calculations were aided by the StatView SE + Graphics, Statistics Utility (Abacus Concepts, Berkeley, CA, USA). All data are mean + s~. p-values of K 0.05 were considered significant.
57
RESULTS In all cases, ~P-NMR spectra of blood showed resonances for 2,3-diphosphoglycerate (2,3-DPG), phosphodiesters (PDEB), y-, ¢¢- and /~-P-ATP. In serum, peaks for inorganic phosphate (Pis) and PDE (PDE~) were visible. T~pical examples are shown in Figs 1 and 2. Other P-metabolites of blood, e.g. glucose-6-phosphate, ADP, AMP and other products of ATP degradation, are invisible due to low concentration. The signal of Pi in blood is unresolvable due to overlap with the signals of 2,3-DPG. Based on Eq. 3, erythrocytes were calculated to contain all 2,3-DPG and ATP, but no PDE. The erythrocytes of several samples were spun down and ~P-NMR-spectra were obtained. In these spectra no PDE was visible and the calculation following Eq. 3 was proved. There was no change in concentrations of free ATP among the various groups (Table 1).
DIABETES MELLIIUS
VOLUm~ ;~,3-DPG
DMMP
PDIE
o.ATP
DMMP
H'YPE~n'n)m~A
CONGeStIVE I~JLRT FAHJUIq~ (NYIIA IV)
2,S-DPG
2,3-DPG
PDE
OMMP IDMIIIIP
~ATP 'L
y-ATP a,A'rP
p-ATP
Fig. 1. Examples of 31p-NMR spectra of blood from a volunteer, a patient with congestive heart failure, one with diabetes mellitus and one with hyperlipidemia. 1032 acquisitions, 42 rain, interpulse delay 2.1 s, pulse angle 45 °. DMMP, O,O'dimethylmethylphosphonate; 2,3-DPG, 2,3-diphosphoglycerate; PDE, phosphodiesters; 7'-, ~-, /~-phosphorus atoms of ATP. M A G M A (1993) 1(2)
58
M HORN et a l. VOLUNTEER
D I A B ~ MELLITUS
POE
POE
DMMP
CONGF.hW[VEI~ARTFK1LLrp~~
IV)
PD£
PDE DINMP
Fig. 2. Examples of 31p-NMR spectra of serum from a volunteer, a patient with congestive heart failure, one with diabetes mellitus and one with hyperlipidemia. 1032 acquisitions, 42 min, interpulse delay 2.1 s, pulse angle 45 °. DMMP, O,O'dimethylmethylphosphonate; Pi, inorganic phosphate; PDE, phosphodiesters.
Absolute and relative amounts of 31p-compounds for the various groups are shown in Figs 3 and 4 and in Table 1. Although the majority of parameters were similar among groups, significant alterations of 31p_ NMR blood spectra were detectable. Both 2,3-DPG and the 2,3-DPG/ATP ratio increased in patients with heart failure, and the magnitude of this increase was related to the severity of hear failure (Fig. 3, p < 0.05 CHFs versus VOL). PDE levels in both blood and serum (Fig. 4) as well as PDE/ATP ratios (Table 1) were markedly elevated (1.57-fold for PDE, 1.71-fold for PDE/ATP ratio, both p < 0.05 HL versus VOL) in patients with hyperlipidemia. Similarly, in patients with diabetes mellitus, the PDE/ATP ratio (Table 1) was significantly increased (2.43-fold, p < 0.05 DM versus VOL).
2O
g
d 5
0 VOL
CHFm
CHFs
DM
HYP
VOL
CHFrn
CHFs
DM
HYP
14 12 a,. 10
