P Nuclear Magnetic Resonance Studies Of Bioenergetic Changes In ...
Gc8326
jan/98
31P-nuclear magnetic resonance studies of bioenergetic changes in
skeletal muscle in malnourished human adults1–3
Andrew Thompson, Andrei Damyanovich, Annie Madapallimattam, David Mikalus, Johane Allard, and
Khursheed N Jeejeebhoy
ABSTRACT
In previous studies, both animals and malnour-
solely to reduced intake, malabsorption, or both, without the
ished children receiving 25% of the protein-energy intake of a
effects of a superimposed critical illness. We therefore used
control group, resulting in a 25% weight loss, had lower ratios of
nuclear magnetic resonance (NMR) spectroscopy to observe
phosphocreatine to -ATP and of phosphocreatine to inorganic
skeletal muscle in malnourished patients and normal control sub-
phosphorus, higher free ADP concentrations, and lower free ener-
jects to test the hypothesis that, in adult humans, we would
gy of ATP hydrolysis than the control group. Therefore, the effect
observe changes similar to those observed in animals.
of malnutrition on muscle energetics in adult humans was exam-
ined by using 31P–nuclear magnetic resonance spectroscopy in
Downloaded from
malnourished patients with a mean body mass index (BMI; in
SUBJECTS AND METHODS
kg/m2) of 16.4 compared with healthy control subjects with a sig-
Patients and control subjects
nificantly higher body mass index of 24.5 (P < 0.005). The mean
(± SEM) ratio of phosphocreatine (PCr) to ATP in the malnour-
Patients were individuals referred by their physicians because
ished patients was 2.28 ± 0.27, which was significantly lower
of clinical malnutrition. On the basis of a diet history and clini-
than the ratio of 3.1 ± 0.15 in control subjects (P < 0.02). The
cal and biochemical investigations, these patients were judged to
www.ajcn.org
ratio of inorganic phosphorus (Pi) to ATP in malnourished
have insufficient energy intakes, to be malabsorbing energy, or
patients was 0.33 ± 0.04, which was significantly lower than the
both; they also had considerable weight loss. In addition, sub-
ratio of 0.48 ± 0.03 in control subjects (P < 0.02), but the ratio of
jective global assessment found that they were all severely mal-
PCr to Pi was not significantly different from that in control sub-
nourished (category C). Three patients were hypoalbuminemic
by on February 9, 2010
jects. There was a significant correlation between BMI and the
and anemic (patients 1, 2, and 7) and two had lymphocyte counts
ratio of PCr to ATP (P < 0.01) and of Pi to ATP (P < 0.01). These
< 1.2
109/L (patients 2 and 8), which indicate moderate mal-
data suggest that progressive loss of BMI is associated with a rel-
nutrition (6). The characteristics of the seven malnourished
ative loss of muscle creatine and phosphorus in relation to ATP.
patients (aged 25–81 y; two men and five women) are shown in
The findings were unlikely to have been due only to atrophy of
Table 1.
fast-twitch fibers because such atrophy would have altered the
These patients were compared with 15 normal control sub-
ratio of PCr to Pi.
Am J Clin Nutr 1998;67:39–43.
jects (aged 23–62 y; seven men and eight women) who were vol-
unteers recruited from the staff and student body of the Univer-
KEY WORDS
Nuclear magnetic resonance spectroscopy,
sity of Toronto and who were healthy, taking no medications,
malnutrition, bioenergetics, ATP, phosphocreatine, inorganic
and not known to be receiving treatment of or to have any dis-
phosphorus, skeletal muscle, adults, humans
ease. The age range of the volunteers included the ages of most
of the patients studied. Both groups were studied by using 31P-
NMR spectroscopic assessment of the right lateral gastrocne-
INTRODUCTION
mius muscle, according to the protocol described below. The
Previous studies have shown that hypoenergetically fed grow-
protocol was approved by the University of Toronto Committee
ing rats receiving 25% of the protein-energy intake of control
for Human Research and informed consent was obtained.
animals, resulting in a 25% weight loss, had lower phosphocrea-
tine (PCr) concentrations, a lower ratio of PCr to ATP, lower free
energy of ATP hydrolysis ( G
), but higher free ADP concen-
ATP
1
trations in the gastrocnemius muscle (1–5) than controls. Stimu-
From the Department of Medicine and Medical Imaging, University of
Toronto.
lation studies showed that rephosphorylation of creatine was
2 Supported by MRC grant MT-10885.
slower in the muscles of hypoenergetically fed rats than in those
3 Address reprint requests to KN Jeejeebhoy, Room 6352, Medical Sci-
of controls (3). In malnourished children, Gupta et al (5) also
ences Building, University of Toronto, Toronto, Ontario, Canada M5S 1A8.
found lower PCr-ATP ratios but higher free ADP concentrations
E-mail: jeejeebhoyk@smh.toronto.on.ca.
than in controls. However, it is not clear whether similar changes
Received September 11, 1996.
occur in malnourished adult humans in whom malnutrition is due
Accepted for publication July 17, 1997.
Am J Clin Nutr 1998;67:39–43. Printed in USA. © 1998 American Society for Clinical Nutrition
39
40
THOMPSON ET AL
TABLE 1
Clinical characteristics of malnourished patients
Subject
Age
Sex
BMI (kg/m2)
Diagnosis
Medications
Hemoglobin
Lymphocyte count
Albumin
g/L
109/L
g/L
1
42
M
10.4
HIV stage IV
3TC, gancyclovir,
102
1.6
32
trimethoprim/sulfamethoxazol-
fluoxetine, sodium bicarbonate,
clarythromycin, ethambutol,
ciprofloxacin, clofazimine,
fluconazole
2
50
M
15.0
HIV stage IV
Diphenoxylate/atropine,
97
0.3
31
ciprofloxacin, fluconazole,
codeine phosphate, acyclovir,
trimethoprim/sulfamethoxazole
3
25
F
17.7
Anorexia nervosa
None
125
1.82
38
4
35
F
18.4
Anorexia nervosa
Pancreatic enzyme
140
1.70
37
supplementation
5
44
F
19.9
Gastroparesis
Ondansetron, docusate sodium,
139
1.71
42
psyllium hydrophilic mucilloid
6
77
F
22.2
Self-neglect
Cisapride, ranitidine, timolol eye
133
2.24
37
drops, hypertonic saline
eye drops
7
81
F
14.1
Gastrectomy
Sotolol, lorazepam, gastrolyte,
115
0.912
28
Downloaded from
oral magnesium supplementation,
oral citrate supplementation
NMR protocol
nourished patients or the other way around did not influence the
Spectra were acquired on a 1.5-T Sigma whole-body scanner
significance of the results. Hence, we used a mean value of the four
www.ajcn.org
(General Electric Medical Systems, Wankesha, WI) by using a
measurements to correct for partial saturation.
20–12.5-cm transmit-receive surface coil tuned to 25.85 MHz
pH was calculated indirectly by using the following formula,
and positioned over the lateral head of the right gastrocnemius
which was validated previously by using test solutions (1):
muscle. Field homogeneity in the sensitive region of the coil was
by on February 9, 2010
achieved by means of iterative gradient shim adjustments. The
pH = 6.75 + log {[(PCr Pi) 3.27]/[5.672 (PCr Pi)]} (1)
10
pulse sequence used consisted of a single hard pulse with a rep-
etition time of 2 s, a spectral width of 4000 Hz, 2048 data points,
Metabolite ratios and estimation of PCr, Pi, and free crea-
and 256 scans. Spectral processing was performed with zero fill-
tine
ing to 4096 points, exponential apodization (1 Hz), fast Fourier
transformation (7) and phasing, followed by baseline correction
To determine metabolite ratios, areas under the PCr, Pi, and
by using a polynomial spline function with which to remove the
ATP peaks were integrated by using a computer program. The
underlying broad resonance originating from bone. All peaks
areas were corrected for incomplete relaxation as described
were then identified and fitted by using a Marquardt-Levenberg
above and the PCr-ATP, Pi-ATP, and Pi-PCr ratios were calculat-
line-fitting routine (8) from which peak amplitudes, areas, chem-
ed from the respective areas.
ical shifts, and widths at half maximum were obtained. Precision
Because we showed previously that hypoenergetic feeding
and reproducibility were tested on phantoms of known concen-
resulting in a 25% weight loss did not reduce ATP concentrations
trations of inorganic phosphorus (Pi), PCr, and ATP. Spectra
(1), we calculated PCr and Pi from the ratios reported here by
were analyzed on a SPARC 20 workstation (Sun Microsystems
using a previously published value for muscle ATP of 8.2 mmol
Inc, Mountain View, CA) by using the computer program SA/GE
intracellular water/L (9). PCr and Pi values were calculated from
(SPECTROSCOPY APPLICATIONS GENERAL ELECTRIC
the ratios of PCr to ATP and of Pi to ATP by using the following
(General Electric Medical Systems).
equations:
Correction for partial saturation and pH calculation
PCr (mmol/L) = (PCr:ATP)
8.2
(2)
The spectra were corrected for partial saturation by taking the
Pi (mmol/L) = (Pi:ATP)
8.2
(3)
ratio of spectra obtained with a repetition time of 2 s to those of
fully relaxed spectra in two control and two malnourished individ-
Free creatine (FCr) was calculated in two ways. On the
uals and applying them to the data obtained. The correction was not
assumption that 1) PCr + FCr = 42.5 mmol intracellular water/L
different between control subjects and malnourished patients for
(8) in both control subjects and malnourished patients, and that
two reasons. First, there were no gross differences between the cor-
2) PCr + FCr was reduced by 15%, to 36.1 mmol/L in malnour-
rection factors for both groups. Second, the correction for partial
ished patients. This assumption was made because in surgical
saturation done either by applying the data from control subjects to
patients, Smyreng et al (10) showed that PCr + FCr was 15%
control subjects and the data from malnourished patients to mal-
lower in those who were severely malnourished.
HYPOENERGETIC FEEDING AND MUSCLE ENERGETICS
41
Free magnesium
6 mo (Table 1). Three patients (patients 1, 2, and 7) had malab-
Because the main effect of free Mg2+ concentrations is the
sorption and chronic diarrhea. Four patients (patients 3–6) were not
alteration of the equilibrium constant of the creatine kinase reac-
eating sufficient food as determined by diet history and observa-
tion (K ) and also because free Mg2+ concentrations have been
tions in the hospital. There was no evidence of neuromuscular dis-
CK
reported to vary [0.2–0.4 mmol/L (11, 12), 1 mmol/L (13), and
ease and none of the patients or control subjects had sustained any
2.5 mmol/L (14)], we used a range of values to calculate free
major orthopedic injuries to their lower limbs. The mean BMI was
ADP concentrations in two previous studies in animals (1, 3).
16.4 in the patients and 24.5 in the control subjects.
Although concentrations of free Mg2+ altered the K , the relative
CK
Metabolite ratios
differences between control rats and hypoenergetically fed rats
remained constant (1, 3) because Mg2+ was not significantly dif-
Mean PCr-ATP and Pi-ATP ratios were significantly lower in
ferent between the two groups of rats. Gupta et al (15, 16)
patients than in control subjects (Table 2). Linear regression
showed that the relative chemical shift between the
and
showed a significant positive correlation (P < 0.05) between
resonances of ATP can be used to estimate free Mg2+ in muscle.
these ratios and BMI (Figures 1 and 2). In contrast, there was no
Because the relative chemical shifts between
and
reso-
significant correlation between BMI and Pi:PCr (Figure 3).
nances of ATP were not significantly different between the con-
There was also no significant difference in pH between the two
trol subjects and malnourished patients in this study (data not
groups.
shown), the free Mg2+ content in muscle was not different
The mean FCr calculated under the assumption that the sum
between the two groups. These data were consistent with our
of PCr and FCr was identical in control subjects and in malnour-
previous observations in rats (1). Therefore, free ADP was cal-
ished patients (42.5 mmol intracellular water/L) was significant-
culated by using the value 1.0 mmol Mg2+/L.
ly higher (P < 0.02) in malnourished patients (24.2 ± 2.2
mmol/L) than in control subjects (17.6 ± 1.2 mmol/L). In con-
Calculation of free ADP
trast, when the sum of PCr and FCr was assumed to be reduced
Downloaded from
Free ADP concentrations were calculated by using the follow-
to 36.1 mmol/L in the malnourished patients, FCr was not sig-
ing equation:
nificantly different between control subjects (17.6 ± 1.2
mmol/L) and malnourished patients (17.8 ± 2.2 mmol/L).
[ATP] = [ATP][Cr]/[PCr][H+]K
(4)
When PCr + FCr was assumed to be identical in control sub-
CK
jects and malnourished patients (42.5 mmol/L), calculated free
Concentrations of ATP, PCr, and H+ were determined by using
ADP concentrations were numerically but not significantly high-
www.ajcn.org
NMR spectroscopy. The constant K
was calculated by using the
er in malnourished patients (9.2 ± 4.4 mol/L) than in control
CK
equation of Lawson and Veech (13). In this equation K
depends on
subjects (5.6 ± 0.6 mol/L). The mean difference in free ADP
CK
pH and free Mg2+. For the reasons given above, the concentration of
concentrations between the two groups was only 1.2 mol/L
Mg 2+ used was 1.0 mmol/L. Because the pH was not significantly
when PCr + FCr was assumed to be reduced to 36.1 mmol/L in
by on February 9, 2010
different between control and malnourished patients, the average
the malnourished patients. G
was not significantly different
ATP
observed pH for the group derived from the 31P-NMR spectra was
between control subjects and malnourished patients, irrespective
used to calculate K
K
and FCr values derived on the basis of the
of the assumed PCr + FCr value.
CK.
CK
assumptions given earlier were used to calculate free ADP.
Calculation of GATP
DISCUSSION
G
was calculated as follows:
Growing rats fed a hypoenergetic diet providing 25% of the
ATP
protein-energy intake of controls and studied after a 25% loss in
G
= G0
+ ln 9[FCr][Pi]/([PCr][H+])K
(5)
body weight had lower PCr-ATP and Pi-PCr ratios, higher cal-
ATP
ATP
CK
culated free ADP concentrations, and lower (G
in the gastroc-
ATP
FCr was calculated by assuming the two different values of
nemius muscle (1–3) than controls. Similar findings were noted
PCr plus FCr referred to earlier. The G0
value depends on
in malnourished children by Gupta et al (5), who used 31P-NMR
ATP
Mg2+ and pH and was calculated from Alberthy’s tables as
spectroscopy. However, rats deprived of food for 2 d (1) had a
described previously (1).
lower PCr-ATP ratio than hypoenergetically fed animals but free
ADP concentrations that were not significantly higher. In these
Statistical methods
animals, the reason for the lower PCr-ATP ratio was related to a
Data are reported as means ± SEMs. Differences between con-
trol subjects and malnourished patients were tested with unpaired t
tests (17). The relations of BMI to PCr:ATP, Pi:ATP, and Pi:PCR
TABLE 2
were analyzed by linear regression (17). All hypotheses testing was
Metabolite ratios in muscle
two-sided; the level of significance was set at P ≤ 0.05.
Control subjects
Malnourished patients
Ratio
PCr:ATP
3.11 ± 0.151
2.28 ± 0.272
RESULTS
Pi:ATP
0.48 ± 0.03
0.33 ± 0.042
Pi:PCr
6.67
7.39
Patients and control subjects
pH
7.18 ± 0.01
7.14 ± 0.01
The patients had a variety of diseases associated with malnutri-
1 x– ± SEM.
tion, which patient histories indicated had been present for at least
2 Significantly different from control subjects, P < 0.02.
42
THOMPSON ET AL
FIGURE 3. Absence of correlation (r = 0.21) between the ratio of
inorganic phosphorus (Pi) to phosphocreatine (PCr) and BMI. The cir-
FIGURE 1. Positive correlation (r = 0.76) between the ratio of phos-
cled data points represent patients who were either the oldest or who
phocreatine (PCr) to ATP and BMI. The circled data points represent
have AIDS.
patients who were either the oldest or who had AIDS.
lower total creatine (PCr + FCr) content in muscle (1). In adults
ed by other studies that showed that resting ratios were not influ-
with a low BMI, findings were similar to those in rats deprived
enced by age but that the response to exercise was reduced with
of food, with no significant differences in pH and Pi-PCr ratios
aging (18, 19).
but a significantly lower PCr-ATP ratio compared with control
The reduction in the PCr-ATP ratio in malnourished patients
Downloaded from
subjects. Although we did not measure muscle creatine directly
could not have been due to acidosis or anoxia because there was
in this study, a likely explanation for our findings is that in mal-
no spectroscopic evidence of a change in muscle pH, and exper-
nourished adults there is a fall in total creatine and phosphorus
imental malnutrition has not been associated with changes in
without a significant rise in free ADP. The finding that free ADP
blood pH, partial pressure of carbon dioxide, or HCO
(1). Pre-
3
concentrations were not significantly different between control
viously, malnutrition was shown to cause atrophy of fast-twitch
subjects and malnourished patients in our study supports this
fibers (20) but it is unlikely that our findings resulted from fast-
www.ajcn.org
conclusion. Further support for a reduction in muscle creatine is
twitch fiber atrophy because such atrophy is typically associated
the finding by Smyreng et al (10) of a reduction in total creatine
with a decrease in the Pi-PCr ratio, which was not observed in
in the muscles of malnourished surgical patients.
this study (21).
Our results potentially could have been confounded by the
The data presented here are different from those obtained by
by on February 9, 2010
fact that we had elderly subjects and AIDS patients in this study
Gupta et al (5) in malnourished children. Gupta et al’s data are
who may have had altered muscle metabolism because of age or
similar to the data from our previous work in hypoenergetically
disease. However, it is unlikely that AIDS or age altered our
fed rats, possibly because of one important fact: the studies in
results because the Pi-ATP, PCr-ATP, and Pi-PCr ratios in our
rats and children were both conducted in growing organisms
two oldest patients and the two AIDS patients (Figures 1 and 2)
whereas the data in this study were obtained in adult individuals.
were close to the regression line and are not the outliers. Hence,
The finding of a reduced PCr-ATP ratio indicating a reduction
they had ratios related to the degree of wasting, a finding that is
in PCr and perhaps in total creatine in muscle likely indicates
no different from observations in younger patients and patients
functional consequences. In other studies, increasing muscle crea-
without AIDS. These observations show that neither AIDS nor
tine was shown to enhance performance (22); therefore, it is pos-
age influenced resting NMR substrate ratios when expressed in
sible that creatine depletion may impair performance. However, to
relation to BMI. The conclusions in relation to age are support-
confirm a reduction in muscle creatine, it will be necessary to
measure it directly by using 1H-NMR together with 31P-NMR and
to determine whether creatine supplementation or overall nutri-
tional rehabilitation will improve muscle function as well as other
nutritional and biochemical outcome variables, such as immune
function, plasma protein status, and cytokine metabolism.
REFERENCES
1. Pichard C, Vaughan C, Struk R, Armstrong RL, Jeejeebhoy KN. The
effect of dietary manipulations (fasting, hypocaloric feeding and
subsequent refeeding) on rat muscle energetics as assessed by
nuclear magnetic resonance spectroscopy. J Clin Invest 1988;
82:895–901.
2. Argov Z, Chance B. Phosphorus magnetic resonance spectroscopy
FIGURE 2. Positive correlation (r = 0.70) between the ratio of inor-
in nutritional research. Annu Rev Nutr 1991;11:449–64.
ganic phosphorus (Pi) to ATP and BMI. The circled data points represent
3. Mijan de la Torre A, Madapallimattam A, Cross A, Armstrong RL,
patients who were either the oldest or who had AIDS.
Jeejeebhoy KN. Effect of fasting, hypocaloric feeding, and refeed-
HYPOENERGETIC FEEDING AND MUSCLE ENERGETICS
43
ing on the energetics of stimulated rat muscle as assessed by nuclear
13. Lawson JWR, Veech RL. Effects of pH and free Mg2+ on the K
of
eq
magnetic resonance spectroscopy. J Clin Invest 1993;92:114–21.
the creatine kinase reaction and other phosphate hydrolases and
4. Mizobata Y, Rounds JD, Prechek D, DeRosa E, Wilmore DW,
phosphate transfer reactions. J Biol Chem 1979;254:6528–37.
Jacobs DO. 31P magnetic resonance spectroscopy demonstrates
14. Wu ST, Pieper GM, Salhany JM, Eliot RS. Measurement of free
expansion of the extracellular space in the skeletal muscle of starved
magnesium in perfused ischemic arrested heart muscle. A quantita-
rats. J Surg Res 1994;56:491–9.
tive phosphorus-31 nuclear magnetic resonance and multiequilibria
5. Gupta RK, Mittal RD, Agarwal KN, Agarwal DK. Muscular suffi-
analysis. Biochemistry 1981;20:7399–403.
ciency, serum protein, enzymes and bioenergetic studies (31-phos-
15. Gupta RK, Benivic JL, Rose ZB. The determination of the free mag-
phorus magnetic resonance spectroscopy) in chronic malnutrition.
nesium level in the human red blood cell by 31P NMR. J Biol Chem
Acta Paediatr 1994;83:327–31.
1978;253:6172–6.
6. Blackburn GL, Bell SJ, Mullen JL. Nutritional medicine. Philadel-
16. Gupta RK, Moore RD. 31P NMR studies in intracellular free Mg2+ in
phia: WB Saunders, 1989.
intact frog skeletal muscle. J Biol Chem 1980;255:3987–93.
7. Arfken GB, ed. Mathematical methods for physicists. 3rd ed. Orlan-
17. Sokal RR, Rohlf FJ. Biometry: the principles and practice of statis-
do, FL: Academic Press, 1985.
tics in biological research. San Francisco: Freeman, 1981.
8. Marquardt DW. An algorithm for least squares estimation of non-
18. Zanconato S, Buchthal S, Barstow TJ, Cooper DM. 31P-magnetic
linear parameters. J Soc Ind Appl Math 1962;11:431–41.
resonance spectroscopy of leg muscle metabolism during exercise in
9. Harris RC, Hultman E, Nordesjo L-O. Glycogen, glycolytic inter-
children and adults. J Appl Physiol 1993;74:2214–8.
mediates and high-energy phosphates determined in biopsy samples
19. Coggan AR, Abduljalil AM, Swanson SC, et al. Muscle metabolism
of musculus quadriceps femoris of man at rest. Methods and vari-
during exercise in young and older untrained and endurance-trained
ance of values. Scand J Clin Lab Invest 1974;33:109–20.
men. J Appl Physiol 1993;75:2125–33.
10. Smyreng T, Larsson J, Schildt B, Wetterfors J. Nutritional assess-
20. Goldspink G, Ward PS. Changes in rodent muscle fibre types during
ment reflects muscle energy metabolism in gastric carcinoma. Ann
postnatal growth,
undernutrition and exercise. J Physiol
Surg 1983;198:146–50.
1979;296:453–69.
11. Maughan D. Diffusible magnesium in frog skeletal muscle cells.
21. Kushmerick MJ, Moerland TS, Wiseman RW. Mammalian skeletal
Downloaded from
Biophys J 1983;43:75–80.
muscle fibers distinguished by contents of phosphocreatine, ATP,
12. Gupta RK, Gupta P, Yushok WD, Rose ZB. Measurement of the dis-
and Pi. Proc Natl Acad Sci U S A 1992;89:7521–5.
sociation constant of MgATP at physiological nucleotide levels by a
22. Balsom PD, Soderlund K, Sjodin B, Ekblom B. Skeletal muscle
combination of 31P NMR and optical absorbance spectroscopy.
metabolism during short duration high-intensity exercise: influence
Biochem Biophys Res Commun 1983;117:210–6.
of creatine supplementation. Acta Physiol Scand 1995;154:303–10.
www.ajcn.org
by on February 9, 2010
