- Email: [email protected]
Diagnosis of muscular glycogenosis by in vivo natural abundance 13C NMR spectroscopy
Neuromtttcular Dtsorders, Vol. I. No L pp. 99-101. 1901 Printed in Gtmlt tlrttain
DIAGNOSIS
OF MUSCULAR ABUNDANCE
$3.00 ÷ 0.00 C 1991 Pergamon Pr¢~ pk 0~8~6~91
GLYCOGENOSIS 13C N M R
BY IN VIVO NATURAL
SPECTROSCOPY
PHILIPPEJEHENSON.* DENTSDuBoc.* GILLESBLOCH.MICHELFARDEAU~and ANDRI~SYROTA Service Hospitalier FredericJoliot. CEA. Orsay; and * Service de cardiologie. H6pital Cochin and INSERM UI53, Paris, France (Received22 November 1990: accepted 18 April 1991)
Abstract--Natural abundance ~-~CNMR (nuclear magnetic resonance) spectroscopy was used to distinguish patients suffering from muscle glycogenosis type V (McArdle's disease) from normal subjects by measuring their muscle glycogen content at rest. Proton-decoupled °C spectra were obtained in 10-15 min from calf muscles at rest. The ratio of the glycogen/creatine signal areas was 12.9 ± 1.7 in four McArdle's disease patients and 2.0 + 0.7 in seven normal subjects. This technique thus allows the non-invasive diagnosis of muscle glycogenosis. Key words: Muscle, glycogenoses, McArdle's disease, magnetic resonance (MR), carbon-13 ("C), glycogen, creatine, diagnosis.
INTRODUCTION
MATERIALSAND METIlODS
Carbon- 13 N M R spectra of calf muscles were McArdle's disease is a muscular glycogenosis (type V) characterised by exercise intolerance obtained in a two Tesla superconducting magnet and rhabdomyolysis with frequent myoglob- (21.6 MHz for I~C) with two concentric singleinuria [I], It is due to a lack of muscle phos- loop surface coils (7 cm diameter for I~C, phorylase activity leading to glycogen storage and II cm for tH decoupling) placed 0.5 cm away to absence of lactic acid production during from the skin. Spectra were acquired with a exercise. Phosphorus 31 (~lp) nuclear mag- 180 degree pulse at the coil centre repeated netic resonance (NMR) has been shown to be every 200 ms. Ten minute spectra (3000 scans) useful for the non-invasive diagnosis of muscular were recorded with tH decoupling during glycogenoses involving glycogenolysis or gly- the 40 ms acquisition time (2 W average power). Four patients (age: 45 4- 19 years) colysis [2,3]. Patients suffering from McArdle's disease show little variation in muscle pH during .w!th biopsy-proven myophosphorylase defexercise, since lactic acid is not produced. How- mtency and seven age-matched normal subjects ever, this ~P NMR test requires a rela- were studied far from any intense muscular tively intense exercise and thus is not easily activity. applicable to uncooperative patients. Human muscular glycogen has been shown to be obserRESULTS vable in vivo by carbon 13 (~~C) NMR [4], in spite of the low NMR sensitivity and low natural The muscle glycogen content was evaluated abundance (I.1%) of t~C, because the glycogen using total creatine (creatine + phosphoC~ peak is well separated from the other peaks creatine) as an internal reference which can be which are mainly due to lipids. We used natural assumed to be constant in muscle, as discussed abundance ~JC NMR, as preliminarily re- below. ported [5], to distinguish patients suffering Typical spectra are shown in Fig. I, The ratio from glycogenosis from normal subjects by of the area of the glycogen C~ peak at 100.5 ppm measuring their muscular glycogen content at to that of the creatine peak at 157 ppm (Fig. 2) rest. was 2.0 4- 0.7 in normal subjects (n -- 7, range 1.3-3.3) and 12.9 4- 1.7 in McArdle's disease (n = 4, range 10.5-14). No overlap occurred * Author to whomcorrespondenceshould be addressed, between the two groups. 99
100
P. JEHENSON et al.
by comparison ~'ith a leg-shaped phantom of known concentration has been proposed. This Difficulties of absolute quantita;ion in NM R assumes that only muscle tissue is observed and, have often been circumvented either by looking therefore, is not generally applicable as adipose and connective tissues may abound in limbs of normal or diseased subjects. McArdle GLY We proposed to use total creatine (Cr). i.e. creatine plus phosphocreatine (PCr), as an internal tissue-specific reference for t~C muscle spectra. Creatine is mostly concentrated in CR muscle, shows little variation in human muscle X 75 .j~ biopsies [6,7] and its guanidino carbon peak at 157 ppm is well separated from other peaks (its is B .,~ glycogen carbon C,). in addition, it is absent from spectra of fat tissue [8] so that negligible Normal contamination occurs and the observed Cr actually reflects the amount of observed muscle tissue. However, its relatively low concentration X 50 (similar to that of glycogen)compared with that of lipids, and even though higher than that of A many other compounds present in the muscle. rcquires good quality spectra. The disadvantage of Cr is that it has much longer relaxation times PPM than glycogen and is, therefore, partly saturated Fig. I. 'tlydrogen-decoupled "C muscle spectra obtained in at the repetition rate used (sec below). Creatinc 15 min in a normal subject (A) and in a myophosphorylasclevels can be assumed to be normal in muscular deficient patient (B), showing the much higher glycogen content in the patient. CR = crcatine, GLY = glycogen glycogenoses at rest as about 6 0 0 of muscle Cr is carbon I. GLY and CR peaks are magnified above the in the form of PCr [6,7], the concentration of corresponding spectra. which is normal [2,3]. The rest of muscle crcatine (free creatine) is in equilibrium with PCr via the GLYCOGEN/ CREATINE creatine-kinase reaction. Creatine was also itssumed or measured to be normal by others in this 1SJ McARDLE's and other myopathies [9]. The measured glycogen/Cr ratio of 2 was comparable with that of about 2.7 determined in biopsies of leg muscles from normal subjects by a biochemical assay [6,7]. As mentioned above, some saturation of Cr (but not of glycogen) occurs at the repetition rate used. This affects the measured signal area of Cr and could be compensated by measuring its "'saturation factor'" but the latter could not be obtained yet with enough precision #1 vivo. Applying this correction would probably tend to increase the Cr signal and hence the difference between Fig. 2. Individual ratios of the glycogen/creatine signal areas NMR and biochemically measured glycogen/Cr in normal subjects and patients suffering from McArdle's ratios in normal subjects; it would, howdisease. ever, bring the results of the two methods at relative variations only or by comparison with closer for patients. Anyhow, in agreement with peaks assumed to be constant: in .ltp muscle what was expected [10,11], a high muscle spectra, ATP or total phosphate is usually taken glycogen content was found at rest in patients as a reference of known concentration. Problems suffering from McArdle's disease, and the large are encountered in finding similar references for gap with normal subjects provides full other elements and ~C in particular. Quanti- confidence in this measurement for diagnostic tation of glycogen in human gastrocnemius purposes. DISCUSSION
1
.
.
.
.
.
.
.
.
I
.
.
.
.
.
.
.
.
.
l
•
Muscle Glycogenosis and "C NMR
In conclusion. *'C NMR measurement of the glycogen,'Cr ratio in the muscle showed a clear separation between normal subjects and patients suffering from type V glycogenosis. Potential clinical applications include the distinction between muscle glycogenoses and other diseased or normal subjects, thus contributing to the noninvasive diagnosis ofhuman metabolic myopathies. Acknowh'dgements--This work was partly supported by the Association Fran~aise contre les Myopathies (AFM).
5.
6.
7. 8.
REFERENCES
1. McArdle B. Myopathy due to a defect in muscle glycogen breakdown. Clin Sci 1951: i0: 13-35. 2. Ross B D. Radda G K. Gadian D G. Rocker G. Esiri M. Falconner-Smith J. Examination of a case of suspected McArdle's syndrome by "P nuclear magnetic resonance. N EnglJ Med 1981: 304: 1338-1342. 3. Duboc D. Jehenson P. Tran-Dinh S, Marsac C. Syrota A. Fardeau M. Phosphorus NM R spectroscopy study of muscular enzyme deliciencies involving glycogenolysis and glycolysis. Neuroh~gy 1987; 37: 663-671. 4. Avison M J, Rothman D L. Nadel E, Shulman R G.
9.
10. II.
101
Detection of human muscle glycogen by natural abundance *'C NMR. Proc Natl Acad Sci I,'SA 1988; 85: 1634-1636. Jehenson P. Duboc D, Bloch G, Syrota A. Evaluation of glycogen content in normal and diseased human muscle by in vivo ~'C NMR. Radioh~gv 1990: 177(P): 165 (abstract). Harris R C. Hultman E, Nordesjo L O. Glycogen. glycolytic intermediates and high-energy phosphates determined in biopsy samples of musculus quadriceps femoris of man at rest. Methods and variance of values. Scand J Clin Lab Invest 1974: 33: 109-120. Sullivan M J, Green H J. Cobb F R. Skeletal muscle biochemistry and histology in ambulatory patients with long-term heart failure. Circuhtthm 1990: 81: 518-527. Canioni P, Alger J, Shulman R G. Natural abundance carbon-13 nuclear magnetic resonance spectroscopy of liver and adipose tissue of the living rat. Biochemistry 1983: 22: 4974~.1980. Arnold D L, Taylor D J, Radda G K. Investigation of human mitochondrial myopathies by phosphorus magnetic resonance spectroscopy. ,4nn Neuro11985: 18: 189-196. Dreyfus J C. Les glycog,~noses musculaires. Etude biochimique. Ann Anat Pathol(Pari.v) 1973: 18(2): 169194. Di Mauro S. Metabolic myopathies. In Winken P J. Bruyn G W e d s . Diseases of Musch'. Hamlhook o[ CTinh'al m'uroh~e.v, Vol 41. Amsterdam: Elsevier. North-|lolland Biomedical Press, 1979.
DIAGNOSIS
OF MUSCULAR ABUNDANCE
$3.00 ÷ 0.00 C 1991 Pergamon Pr¢~ pk 0~8~6~91
GLYCOGENOSIS 13C N M R
BY IN VIVO NATURAL
SPECTROSCOPY
PHILIPPEJEHENSON.* DENTSDuBoc.* GILLESBLOCH.MICHELFARDEAU~and ANDRI~SYROTA Service Hospitalier FredericJoliot. CEA. Orsay; and * Service de cardiologie. H6pital Cochin and INSERM UI53, Paris, France (Received22 November 1990: accepted 18 April 1991)
Abstract--Natural abundance ~-~CNMR (nuclear magnetic resonance) spectroscopy was used to distinguish patients suffering from muscle glycogenosis type V (McArdle's disease) from normal subjects by measuring their muscle glycogen content at rest. Proton-decoupled °C spectra were obtained in 10-15 min from calf muscles at rest. The ratio of the glycogen/creatine signal areas was 12.9 ± 1.7 in four McArdle's disease patients and 2.0 + 0.7 in seven normal subjects. This technique thus allows the non-invasive diagnosis of muscle glycogenosis. Key words: Muscle, glycogenoses, McArdle's disease, magnetic resonance (MR), carbon-13 ("C), glycogen, creatine, diagnosis.
INTRODUCTION
MATERIALSAND METIlODS
Carbon- 13 N M R spectra of calf muscles were McArdle's disease is a muscular glycogenosis (type V) characterised by exercise intolerance obtained in a two Tesla superconducting magnet and rhabdomyolysis with frequent myoglob- (21.6 MHz for I~C) with two concentric singleinuria [I], It is due to a lack of muscle phos- loop surface coils (7 cm diameter for I~C, phorylase activity leading to glycogen storage and II cm for tH decoupling) placed 0.5 cm away to absence of lactic acid production during from the skin. Spectra were acquired with a exercise. Phosphorus 31 (~lp) nuclear mag- 180 degree pulse at the coil centre repeated netic resonance (NMR) has been shown to be every 200 ms. Ten minute spectra (3000 scans) useful for the non-invasive diagnosis of muscular were recorded with tH decoupling during glycogenoses involving glycogenolysis or gly- the 40 ms acquisition time (2 W average power). Four patients (age: 45 4- 19 years) colysis [2,3]. Patients suffering from McArdle's disease show little variation in muscle pH during .w!th biopsy-proven myophosphorylase defexercise, since lactic acid is not produced. How- mtency and seven age-matched normal subjects ever, this ~P NMR test requires a rela- were studied far from any intense muscular tively intense exercise and thus is not easily activity. applicable to uncooperative patients. Human muscular glycogen has been shown to be obserRESULTS vable in vivo by carbon 13 (~~C) NMR [4], in spite of the low NMR sensitivity and low natural The muscle glycogen content was evaluated abundance (I.1%) of t~C, because the glycogen using total creatine (creatine + phosphoC~ peak is well separated from the other peaks creatine) as an internal reference which can be which are mainly due to lipids. We used natural assumed to be constant in muscle, as discussed abundance ~JC NMR, as preliminarily re- below. ported [5], to distinguish patients suffering Typical spectra are shown in Fig. I, The ratio from glycogenosis from normal subjects by of the area of the glycogen C~ peak at 100.5 ppm measuring their muscular glycogen content at to that of the creatine peak at 157 ppm (Fig. 2) rest. was 2.0 4- 0.7 in normal subjects (n -- 7, range 1.3-3.3) and 12.9 4- 1.7 in McArdle's disease (n = 4, range 10.5-14). No overlap occurred * Author to whomcorrespondenceshould be addressed, between the two groups. 99
100
P. JEHENSON et al.
by comparison ~'ith a leg-shaped phantom of known concentration has been proposed. This Difficulties of absolute quantita;ion in NM R assumes that only muscle tissue is observed and, have often been circumvented either by looking therefore, is not generally applicable as adipose and connective tissues may abound in limbs of normal or diseased subjects. McArdle GLY We proposed to use total creatine (Cr). i.e. creatine plus phosphocreatine (PCr), as an internal tissue-specific reference for t~C muscle spectra. Creatine is mostly concentrated in CR muscle, shows little variation in human muscle X 75 .j~ biopsies [6,7] and its guanidino carbon peak at 157 ppm is well separated from other peaks (its is B .,~ glycogen carbon C,). in addition, it is absent from spectra of fat tissue [8] so that negligible Normal contamination occurs and the observed Cr actually reflects the amount of observed muscle tissue. However, its relatively low concentration X 50 (similar to that of glycogen)compared with that of lipids, and even though higher than that of A many other compounds present in the muscle. rcquires good quality spectra. The disadvantage of Cr is that it has much longer relaxation times PPM than glycogen and is, therefore, partly saturated Fig. I. 'tlydrogen-decoupled "C muscle spectra obtained in at the repetition rate used (sec below). Creatinc 15 min in a normal subject (A) and in a myophosphorylasclevels can be assumed to be normal in muscular deficient patient (B), showing the much higher glycogen content in the patient. CR = crcatine, GLY = glycogen glycogenoses at rest as about 6 0 0 of muscle Cr is carbon I. GLY and CR peaks are magnified above the in the form of PCr [6,7], the concentration of corresponding spectra. which is normal [2,3]. The rest of muscle crcatine (free creatine) is in equilibrium with PCr via the GLYCOGEN/ CREATINE creatine-kinase reaction. Creatine was also itssumed or measured to be normal by others in this 1SJ McARDLE's and other myopathies [9]. The measured glycogen/Cr ratio of 2 was comparable with that of about 2.7 determined in biopsies of leg muscles from normal subjects by a biochemical assay [6,7]. As mentioned above, some saturation of Cr (but not of glycogen) occurs at the repetition rate used. This affects the measured signal area of Cr and could be compensated by measuring its "'saturation factor'" but the latter could not be obtained yet with enough precision #1 vivo. Applying this correction would probably tend to increase the Cr signal and hence the difference between Fig. 2. Individual ratios of the glycogen/creatine signal areas NMR and biochemically measured glycogen/Cr in normal subjects and patients suffering from McArdle's ratios in normal subjects; it would, howdisease. ever, bring the results of the two methods at relative variations only or by comparison with closer for patients. Anyhow, in agreement with peaks assumed to be constant: in .ltp muscle what was expected [10,11], a high muscle spectra, ATP or total phosphate is usually taken glycogen content was found at rest in patients as a reference of known concentration. Problems suffering from McArdle's disease, and the large are encountered in finding similar references for gap with normal subjects provides full other elements and ~C in particular. Quanti- confidence in this measurement for diagnostic tation of glycogen in human gastrocnemius purposes. DISCUSSION
1
.
.
.
.
.
.
.
.
I
.
.
.
.
.
.
.
.
.
l
•
Muscle Glycogenosis and "C NMR
In conclusion. *'C NMR measurement of the glycogen,'Cr ratio in the muscle showed a clear separation between normal subjects and patients suffering from type V glycogenosis. Potential clinical applications include the distinction between muscle glycogenoses and other diseased or normal subjects, thus contributing to the noninvasive diagnosis ofhuman metabolic myopathies. Acknowh'dgements--This work was partly supported by the Association Fran~aise contre les Myopathies (AFM).
5.
6.
7. 8.
REFERENCES
1. McArdle B. Myopathy due to a defect in muscle glycogen breakdown. Clin Sci 1951: i0: 13-35. 2. Ross B D. Radda G K. Gadian D G. Rocker G. Esiri M. Falconner-Smith J. Examination of a case of suspected McArdle's syndrome by "P nuclear magnetic resonance. N EnglJ Med 1981: 304: 1338-1342. 3. Duboc D. Jehenson P. Tran-Dinh S, Marsac C. Syrota A. Fardeau M. Phosphorus NM R spectroscopy study of muscular enzyme deliciencies involving glycogenolysis and glycolysis. Neuroh~gy 1987; 37: 663-671. 4. Avison M J, Rothman D L. Nadel E, Shulman R G.
9.
10. II.
101
Detection of human muscle glycogen by natural abundance *'C NMR. Proc Natl Acad Sci I,'SA 1988; 85: 1634-1636. Jehenson P. Duboc D, Bloch G, Syrota A. Evaluation of glycogen content in normal and diseased human muscle by in vivo ~'C NMR. Radioh~gv 1990: 177(P): 165 (abstract). Harris R C. Hultman E, Nordesjo L O. Glycogen. glycolytic intermediates and high-energy phosphates determined in biopsy samples of musculus quadriceps femoris of man at rest. Methods and variance of values. Scand J Clin Lab Invest 1974: 33: 109-120. Sullivan M J, Green H J. Cobb F R. Skeletal muscle biochemistry and histology in ambulatory patients with long-term heart failure. Circuhtthm 1990: 81: 518-527. Canioni P, Alger J, Shulman R G. Natural abundance carbon-13 nuclear magnetic resonance spectroscopy of liver and adipose tissue of the living rat. Biochemistry 1983: 22: 4974~.1980. Arnold D L, Taylor D J, Radda G K. Investigation of human mitochondrial myopathies by phosphorus magnetic resonance spectroscopy. ,4nn Neuro11985: 18: 189-196. Dreyfus J C. Les glycog,~noses musculaires. Etude biochimique. Ann Anat Pathol(Pari.v) 1973: 18(2): 169194. Di Mauro S. Metabolic myopathies. In Winken P J. Bruyn G W e d s . Diseases of Musch'. Hamlhook o[ CTinh'al m'uroh~e.v, Vol 41. Amsterdam: Elsevier. North-|lolland Biomedical Press, 1979.