BRIEF REVIEW
Phosphorus Magnetic Resonance Spectroscopy ( 3 'P MRS) in Neuromuscular Disorders -
A
Zohar Argov, MD," and William J. Bank, MDt
Phosphorus magnetic resonance spectroscopy monitors muscle energy metabolism by recording the ratio of phosphocreatine to inorganic phosphate at rest, during exercise, and during recovery from exercise. In mitochondrial diseases, abnormalities may appear during some or all these phases. Low phosphocreatine-inorganic phosphate ratios at rest are not disease-specific, but can be increased by drug therapy in several myopathies. Phosphorus magnetic resonance spectroscopy can also record intracellular pH and thus identify disorders of glycogen metabolism in which the production of lactic acid is blocked during ischemic exercise. The measurements of accumulated sugar phosphate intermediates further delineate glycolytic muscle defects. Myophosphorylase deficiency responds to intravenous glucose administration with improved exercise bioenergetics, but no such response is seen in phosphofructokinase deficiency. The muscular dystrophies show no specific bioenergetic abnormality; however, elevation of phospholipids metabolites and phosphodiesters was detected in some cases. While phosphorus magnetic resonance spectroscopy remains primarily a research tool in metabolic myopathies, it will be clinically useful in identifying new therapies and monitoring their effects in a variety of neurornuscular disorders. Argov 2, Bank WJ. Phosphorus magnetic resonance spectroscopy (31PMRS) in neuromuscular disorders. Ann Neurol 1991;30:90-97
The biochemical investigation of neuromuscular disorders has long relied only on in vitro examination of muscle tissue obtained by biopsy. In vivo phosphorus magnetic resonance spectroscopy ("P MRS) now permits repeated, noninvasive, real-time studies of muscle energy metabolism at rest and during exercise. We highlight the basic concepts of muscle 31PMRS and address those clinical muscle disorders in which 31P MRS observations have been relevant. The analytical sensitivity of in vivo 31PMRS is limited, since metabolites cannot be measured at less than 0.3 to 0.5 mM in a 2 T magnetic field. Fortunately, the five major peaks present at such concentrations and visible in a normal resting muscle spectrum (Fig 1) are associated with energy metabolism. Three peaks identify the distinct positions of each of the phosphate atoms of ATP. The remaining peaks are those of phosphocrearine (PCr) and inorganic phosphates (Pi). The spectral distance between PCr and Pi peaks varies with pH, permitting the determination of intracellular p H of muscle cells [I]. The area under each peak is deter-
mined by the amount of each metabolite in the sampled muscle volume. This volume depends on the size of the external detection coil used in in vivo studies. Current MRS techniques eliminate most of the phosphorus signal from nonmuscle tissues. Calibration to absolute concentrations may be achieved by comparison with an external phosphate standard of known concentration. ATP concentration, however, has been traditionally used as an internal standard {2) because it was thought to remain constant in normal working muscle. A constant concentration of phosphoruscontaining compounds, however, cannot be assumed in metabolic disorders or muscular dystrophies [3). Therefore, the relative ratios of compounds are preferential measurements to single peak amounts for quantitative evaluations. Such ratios also have a bioenergetic significance. Muscle energetic state is characterized by the balance between ATP utilization and ATP production or by the cytosolic phosphorylation potential (PP) {4, 51. The PP plays an important role in the control of mitochondrial respiration and is directly propor-
From the "Department of Neurology, Hadassah University Hospital, Jerusalem, Israel, and the +Department of Neurology, Hospiral of University of Pennsylvania, Philadelphia, PA.
Address correspondence to Dr Argov, Department of Neurology, Hadassah University Hospital, Jerusalem 91 120, Israel.
Received Nov 10, 1989, and in revised form Mar 19 and Dec 11, 1990, and Jan 4 , 1991. Accepted for publication Jan 4 , 1991.
90 Copyright 0 1991 by the American Neurological Association
PI
AT P
I0
5
0
-5 p m
-10
-15
-20
Fig 1 . 3 1 PMRS spectra of normal muscle during rest and exercise. The positions of the three phosphate atoms of A T P are marked as a, /3, and y. The regions where phosphodiesters (PDE) and phosphomonoesters (PME) appear when elevated are marked by arrows. Note the fall of PCr and rise of P i while exercise is increased. ATP remains stable during such louds of work. PCr = phosphocreatine; Pi = inorganic phosphate; S = the chemical shift between PCr and P i used fur pH detmination; ppm = parts per million calibration of the spectral distance.
tional to the ratio of PCr/Pi at a given pH, both measurable by 31P MRS (see Addendum). The measured ratio of PCr to Pi reflects an average of data from heterogeneous human muscle cells that may have varying rates of oxidative metabolism. This applies both to normal muscle and to its pathological conditions. Such heterogeneity has been confirmed in the cat, where slow-twitch, oxidative muscle has significantly lower PCrIPi values than fast-twitch, glycolytic fibers {b}. A high-resolution spectrum of resting muscle can be recorded within minutes. During moderate exercise, PCr falls with a commensurate rise in Pi; ATP remains constant (see Fig 1). These rapid changes in PCr and Pi cannot be accurately represented in the time resolution of in vivo 31PMRS (about 1 minute) [73. Steadystate exercise studies, however, overcome this problem 17, 81. In such a protocol a subject performs constant muscle work (measured by an ergometer) until a stabilized PCriPi is achieved. Similar steady-states obtained at increased levels of work permit a plot of work performed versus Pi/PCr [S, 97. The initial slope of such a plot reflects the relationship of work to energy expenditure and enables the comparison of metabolic capacity between different subjects as well as changes in an individual over time or in response to treatment. At the cessation of exercise, 31PMRS spectra rapidly return to their resting state characteristics. In contrast to energy metabolism during exercise involving glycolytic activity, the process of PCr recovery is governed solely by oxidative metabolism. Postexercise recovery is therefore a good measure of mitochondrial function
{lo]. The rate of recovery in normals depends on the extent of exercise performed. The greater the endexercise PCr depletion, the slower the rate of recovery [101. In order to accurately compare rates of postexercise recovery, similar degrees of PCr depletion must be achieved at the end of exercise. Technical and physical limitations of in vivo 31P MRS must also be considered. Measurement of work rate is difficult because of equipment limitations in a high magnetic field. Also the physically measured work and its relation to biochemical work are dependent on the type of exercise performed (isometric, isokinetic, etc.). In order to make valid comparisons between normal and various disease states, maximum power (work capacity) for each muscle tested must be assessed. This can be difficult in weak patients who may be unable to perform steady-state exercise or who cannot overcome the minimal resistance of an ergometer. Diminished muscle bulk in pathological conditions can also affect the accuracy of measurements. Indeed, even the normal asymmetry between dominant and nondominant arm muscles may affect exercise studies [111.
Mitochondrial Myopathies Disorders of mitochondrial metabolism are particularly suited for 31P MRS studies since the latter primarily monitors oxidative metabolism. Features of oxidative metabolism can be measured by 31PMRS during rest, work, and recovery after exercise in patients with mitochondrial myopathies (MM), and illustrate many of the abnormalities that can be demonstrated with this technique.
Rest PCrlPi Normal muscle at rest has a low metabolic rate and a high energy capacity, resulting in a high PCr/Pi [7]. Impaired mitochondrial function reduces the energy state and decreases PCr/Pi at rest. One of the first human 31PMRS studies showed an elevation of Pi (low PCrlPi) at rest in a patient with MM [123. Subsequent studies showed that the resting state could also be normal in some patients with MM [13]. In two large series, 66 and 83% of patients with MM had low PCr/ Pi at rest [ 14,151. A low PCr/Pi at rest is, however, not a disease-specific abnormality (Table 1). Mitochondrial energy pathways may be affected in carnitine deficiency [l6} and hypothyroidism { 171. Uncoupled mitochondrial respiration has been suspected in malignant hyperthermia 1181. The majority of disorders listed in Table 1, however, do not represent a known primary disturbance in energy metabolism, and secondary mitochondrial malfunction is suggested [19]. PCr/Pi at rest can also be acutely altered in response to transient muscle injury in normal subjects after extensive exercise 1201.
Brief Review: Argov and Bank: 31P MRS in Neuromuscular Disorders
91
Table 1 . Eiuman Neuromuscular Disorakrs Associated with Low Phospbocreatine-Inorganic Phosphate Ratios at Rest
1. Mitochondrial myopathies 2. Primary carnitine deficiency 3 . State after muscle contracture in myophosphorylase deficiency
4 . Hypothyroid myopathy 5. Muscular dystrophy 6. Myotonic dystrophy 7. Motor neuron and other denervating disorders 8. Polymyositis 9. Muscle injury 10. Recurrent myoglobinuria of undetermined origin 11. Malignant hyperrhermia suscepribles
Chungar during ExerciJe The phosphorylation potential rapidly diminishes in patients with MM when metabolic demands are stressed during exercise. Although such patients perform less work than normal subjects, a greater decrease in PCr/Pi occurs at the end of exercise C141. Only 5 of 18 patients with MM could exercise at multiple steady-states, which are necessary to plot work rate versus PiiPCr. The initial slope of these plots was abnormally low in these patients fl5]. Although both studies indicate an abnormally early and rapid reduction of PCr in patients with MM during exercise, this abnormahty has not yet been quantitated. Impaired oxidative metabolism should result in enhanced glycolytic activity and a subsequently greater intracellular lactic acid accumulation. Surprisingly, a relative resistance to intracellular acidosis has been seen in MM-affected muscles 12, 12, 14, 151. The mechanism of this phenomenon is not understood; however, an adaptation of intracellular buffering systems responding to an overproduction of lactate has been proposed 114, 151.
slow rate of recovery was seen in only 30% of patients in one study {143. Data corrected ro comparable endexercise PCrlPi values, however, showed that I? (94%) of 18 patients had a prolonged recovery L151. A delay in postexercise recovery of P W P i is therefore a most sensitive 31PMRS indicator of mitochondrial malfunction. No correlation can be made between MRS findings and the site of the defect in the mitochondrial respiratory chain. There may also be variability of 31PMRS data in patients with a similar rnitochondrial defect [ 15, 191. The injection of known mitochondrial inhibitors in animals showed acute 31PMRS abnormalities during exercise and recovery after exercise but not at rest r21). ilP MRS findings correlate best with the degree of muscle function impairment in patients with MM {19}. Some mitochondrial cytopathies may show no 31P MRS abnormality, indicating that skeletal muscle mitochondria are probably not involved. Four generalized categories of 31PMRS findings in patients with MM are compiled in Table 2.
Glycolytic Disorders Impaired glycolysis or glycogenolysis blocks the ability of muscle to produce lactic acid during anaerobic exercise. Normal muscle develops cytoplasmic acidification during intense work and an intracellular p H as low as 6.0 has been observed by 31PMRS 122). Absence of intracellular acidosis, as measured by 31P MRS after ischemic exercise, confirms disorders of deficient muscle glycolysis [23, 241. Glycolytic disorders differ from disorders of glycogenolysis in that glycolytic intermediates accumulate as phosphorylated sugars in the former. These phosphomonoesters (PME) are identified to the left of the Pi peak (Fig 2). The accumulation of PME was first noted in patients lacking muscle phosphofructokinase (PFK) {24, 251. Subsequently, disorders more Qstal in the glycolytic sequence (phosphoglyceratekinase {26] and phosphoglyceratemutase EPGAM} [27}) showed a similar phenomenon. A rise in PME has also been documented in PFK-deficient
Postexercise Recmev The repletion of PCr after exercise is primarily controlled by mitochondrial oxidative metabolism, yet a
Table 2. Summa7y of 31PMRS Findings in Patients with Mitochondrial Myopatbies (MM) Normal Findings
1. MM without skeletal muscle involvement 2. Nonspecific finding of diseased muscle 3. Mitochondrial malfunctionprimary or secondary" 4. Primary mitochondrial disease
Low PCr/Pi at Rest
Prolonged PCr Recovery
Abnormal Work Kinetics
+ +
+ +
+
aEither one or both abnormahties may exist. PCr = phosphocreatine; Pi = inorganic phosphate.
92 Annals of Neurology Vol 30 No 1 July 1991
+ +
I
40 30 20 10
PME P O
0 -10 -20 -30 -40 ppm
Fig 2. Rest and exercise spectra of a patient with phosphofructokinase deficiency. Note the accumulation of phosphornonoesters (PME) daring the exercise. Pi = inorganic phosphate; PCr = phosphocreatine; ppm = parts per million.
dogs 128) and acutely blocked glycolysis induced by iodoacetate injections in rats 129, 30). In muscles with impaired glycolysis, PME rise promptly and continuously even during aerobic work, demonstrating the need for glucose as “muscle fuel” even in mild exercise. The severity of PME accumulation may be an indicator of the degree of enzymatic block 127, 30). PME levels are not increased in PFK-deficient muscle at rest and their postexercise recovery shows an exponential curve 131). The postexercise recovery of Pi in these cases is slower than that of PCr, indicating that dephosphorylation is involved in PME clearance [31). Compared with PFK deficiency, PME accumulation was lower in a patient with a partial defect of PGAM and the recovery rate of PME was faster 1271. This may be due to additional clearance of sugar phosphates by the residual glycolytic activity. The relationship between the degree of glycolytic block and PME kinetics was also confirmed in the acute animal model 1301. It has previously been assumed that intracellular p H determined the rate of PCr/Pi recovery after exercise in normal muscle, since greater acidosis was associated with slower recovery [ l o ] . In patients with myophosphorylase (MPH) deficiency, who do not develop acidosis, the recovery of PCr/Pi was proportional to the end-exercise PCr/Pi as in normal subjects [22, 32). This indicates that factors other than cytoplasmic pH are also responsible for determining the rate of postexercise recovery. The type of exercise that patients perform may determine the type of abnormalities observed by ”P MRS. For example, a diminished rise in Pi was demonstrated in PFK deficiency after a short bout of exercise and led to the suggestion that phosphates are “trapped” by the accumulating sugar phosphates [25). In contrast, a continuous and higher rise of Pi was seen in the same PFK-deficient patient during a prolonged, graded type of exercise [3 1). Clinically, patients with glycolytic disorders can perform prolonged, graded work better than short, rapid bouts of exercise. These variances must be considered when comparing 31PMRS observa-
tions of similar disorders. Lewis and colleagues 133) demonstrated a higher end-exercise Pi in MPH deficiency compared with normal subjects and suggested that oxidative metabolism is also impaired in this disorder. In PFK-deficient dogs, PCr/Pi levels were higher than control levels after similar rates of stimulated muscle work I281. Whether this indicates impaired muscle aerobic metabolism in these disorders is not yet clear. Painful exercise intolerance and muscle contracture, the main features of glycolytic muscle disorders, remain unexplained. A possible explanation has been ATP depletion during exercise C34). Biochemical examination of muscle from an MPH-deficient patient during an ischemic contracture did not, however, demonstrate loss of ATP [35). Similarly, 31PMRS showed no loss of ATP during exercise or during a contracture in an MPH-deficient patient [23, 32). In contrast, PEK-deficient muscle, both in humans { 3 1) and in dogs {ZS), showed a continuous mild loss of ATP during graded, submaximal, nonischemic exercise. This observation was not associated with exercise intolerance or a clinical contracture. In the model of acute glycolytic block in rats, a gradual loss of ATP in both biochemical t36) and in vivo 31PMRS studies {30) was found. Although PFK- and MPH-deficient patients have similar symptoms, their ATF’ kinetics during exercise differ and the causal relationship between ATP concentration and contracture remains unresolved C34, 37). The “second wind” phenomenon is typical in MPH deficiency and results in increased work capacity after a short bout of exercise and rest. It is rare, if present, in PFK deficiency. An improved end-exercise PCr/Pi has been observed in MPH deficiency after an intravenous glucose infusion 1331. The slope of work versus Pi/PCr during graded exercise is also augmented by glucose 1321. In PFK deficiency the plot of this slope remains unchanged [31) or becomes even lower [ 3 8 } under similar conditions. Blood-borne glucose may, therefore, be the mediator of the “second wind” phenomenon. Other yet untested substrates may also contribute. Two patients with a proven partial MPH deficiency showed a greater degree of intracellular acidosis during aerobic exercise than during ischemic exercise 1391. In normals, lactic acid production is greater during ischemic exercise 1391. It has been suggested that these carriers, heterozygous for MP H deficiency, represent a unique metabolic adaptation enhancing muscle use of serum glucose. 31P MRS is therefore useful in identifying several disorders of glycogen metabolism by the absence of intracellular acidosis during exercise. It can also distinguish between glycogenolytic and glycolytic disorders by the accumulation of PME. In addition, it is helpful in understanding metabolic adaptations of muscle in the presence of an enzymatic deficiency.
Brief Review: Argov and Bank: ”P MRS in Neuromuscular Disorders
93
Muscular Dystrophy PCr/Pi at rest is reduced in the forearm 140) and gastrocnemius 1411 muscles in Duchenne muscular dystrophy (DMD). The decrease in PCr/Pi in dystrophic muscles is age related, suggesting that this reflects the dystrophic process 1417. There have been conflicting reports of increased [40,41) and normal 142) intracellular p H in D M D at rest. The estimated total phosphate signal and ATP concentration remain unchanged in patients with D MD [42), in contrast to earlier biochemical data 131. This remains to be verified in a combined study of the same muscle. Several phosphate compounds, unusual for normal muscles, have been detected in DMD. An elevation of phosphodiester (PDE) is observed between the PCr and Pi peaks in arm 140) and leg 141) muscles of patients with DMD, although its exact identity is unknown. This accumulation may reflect a breakdown of phospholipids, coincident with the dystrophic process 143). The PDE serine ethanolamine is elevated in dystrophic chicken muscle 1447. Dystrophin-deficient dogs [453 and dystrophic hamsters 1467, however, show no rise in PDE, despite low PCr/Pi at rest. A rise in PDE can be seen in normal calf muscles with aging 1471 and may therefore be a nonspecific indicator of muscle breakdown. This is further supported by transient PDE accumulation in hypothyroid muscle, which disappeared after 3 months of hormone replacement 117).
Other Neuromuscular Disorders Arnold and colleagues [48] reported a patient with chronic fatigue after a viral illness who had an unusual 31PMRS pattern of early and excessive acidosis during exercise. This condition is, however, controversial in both clinical and MRS data. Two patients with rnyalgia and tubular aggregates unrelated to viral illness showed similar findings 1497. We have also seen this phenomenon in a few sedentary but healthy persons. The syndrome of early acidosis, by 31PMRS criteria, remains therefore speculative since it is not necessarily abnormal or disease-specific. Patients with paramyotonia show no /'P MRS abnormality, even when the cooled muscles are stiff and paralyzed [SO]. A decreased PCr/Pi at rest can be seen in the advanced stages of myotonic dystrophy, as in many other neuromuscular disorders with advanced muscle degeneration (see Table 1). Local muscle ischemia, as seen in "compartment syndrome" { 5 l ) , may change the energy state of muscle. 31PMRS abnormalities seen in exercising skeletal muscle during congestive heart failure were thought to represent the result of diminished muscle blood flow, but may in fact be due to a chronic metabolic defect 152, 533. This must be considered when evaluating a myopathy with cardiac involvement. 94 Annals of Neurology Vol 30 No 1 July 1991
Alcoholics, with and without a history of rhabdomyolysis, demonstrate reduced acidosis during ischemic exercise, suggesting impaired glycolysis [543. Many other neuromuscular disorders, including the neurogenic muscle atrophies 1553, tested by 31P MRS showed nonspecific or normal findings [%I.
31PMRS Monitoring of Therapy Although decreased PCr/Pi at rest is a nonspecific finding, this ratio has been a valuable monitor both of progression of disease and of response to specific therapy. Eleff and associates 157) studied a patient with mitochondrial complex 3 deficiency who had the triad of "P MRS abnormalities typical of primary MM. In an attempt to bypass the biochemical defect, the patient was treated with electron acceptors (vitamins K, and C) in large doses. This resulted in improved clinical and "P MRS findings. PCr/Pi at rest rose from 1.7 to 6.1 and the rate of postexercise recovery increased markedly C57, 58). A correlation between the patient's clinical course and "P MRS findings was maintained during 5 years of treatment 119, 583. Clinical improvement associated with a rise in PCr/Pi at rest has been shown in patients with carnitine deficiency who were treated with L-carnitine [59], familial MM treated with methylprednisolone [60}, and MM treated with coenzyme Qlo161). The improved oxidative metabolism is indicative of improved mitochondrial function in these cases. Response to therapy has also been demonstrated by jlP MRS in disorders in which mitochondria are secondarily involved. Rest spectra improved in hypothyroidism treated with hormone replacement [17] and poiymyositis treated with corticosteroids (62). 31P MRS spectra at rest are particularly advantageous as objective monitors since they can be obtained in young children and severely weak individuals who are unable to perform exercise. Conclusions In vivo "P MRS offers a harmless, noninvasive measurement of muscle bioenergetics. Data are acquired promptly and repeated examinations are feasible. There are, however, technical limitations of calibration and quantification of muscle work, especially in very weak patients. The heterogeneity of skeletal muscle is also difficult to resolve 163). 31P MRS will remain predominantly a research tool in myology and will not replace in vitro biochemical studies for diagnosis. It will be useful for screening patients with exercise intolerance due to metabolic derangements. A scheme for 31P MRS testing of such patients is presented in Figure 3. "'P MRS can also contribute to further understanding of normal muscle bioenergetics and clarify adaptations that occur in the presence of impaired metabolism. Objective monitor-
The ratio of PCriPi is directly measured from the muscle spectrum. The concentration of protons {HI is calculated from the 31P MRS-determined pH. The equilibrium constant {K} is known to be 1.66 x M - ' at p H of 7 [64}. If the creatine content of the muscle is measured, then the PP can be directly calculated {14).
,Exercise ljlolerance, [Rest, Exercise, Recovery ~ r o i o c o ~ ]
I
Full Distal GlyCOlytiC
i
1.
Mixed Pattern
Normal Findmqs
Mitochondria I Malfunclion Primary and
CPT hficiency
31PMRS Glossary PCr-phosphocreatine Pi-inorganic phosphates (H2P04- and H P 0 4 - - ) PME-phosphomonoesters; in muscle, probably phosphorylated sugars and inosine monophosphate PDE-phosphodiesters; in normal human muscle, probably
i
Some PME
I
F'arlid Glycolyflc
glycerol-3-phosphorylcholine
AMP Deominasa oeticiencv?
Fig 3. A simp/ejou: chart for the evahation of exercise intolMRS. This is based on eualzrating a rest-exerciseerance by recotrey protocol and an ischemic exercise test. PCr = phosphocreatine;Pi = znorganicphosphate;PME = phosphomonoesters; CPT = cumitine palmytyl transferase; AMP = adenosine monophosphate.
ing of therapy will be most promising both in humans and in animal models.
ppm-parts p e r million; a unit for measuring the spectral distance, related t o the frequency of the recorded atom PCr/Pi-31P MRS-determined ratio that correlates with t h e phosphorylation potential ~~
Many of the observations reported here were performed while Dr Argov was a Muscular Dystrophy Association Post-Doctoral ReUnisearch Fellow in the Department of Biochemistry/B~ophys~cs, versity of Pennsylvania, Philadelphia Grants from the Muscular Dystrophy Association and the National Institutes of Health (NS 08075) supported OUT human and animal research.
References 1. Taylor DJ, Bore PJ, Styles P, et al. Bioenergetics of intact hu-
man muscle. A gated jlP NMR study. Mol Biol Med 1983;l:
Addendum The phosphorylation potential (PP) is defined as:
77-94
(11 The creatine kinase (CK) reaction that controls the cytosolic ATP concentration and its availability to mitochondria is: PCr
+ ADP + H + + A T P + Cr
where Cr is free creatine. This reaction is normally at or near equilibrium with the following equilibrium constant (Kj: K=
[ATP} x {Cr} {PCr) X {ADP) x [H}
Derived from (2) is: [ATP) =
K x [PCr) x [HI x {ADP]
m-1
which when substituted in the PP equation (1) gives:
(3)
2. Radda GK, Taylor DJ. Application of nuclear magnetic resonance spectroscopy in pathology. Int Rev Exp Pathol 1985;2: 1-58 3. Samaha FJ, Davis B, Nagy D. Duchenne muscular dystrophy: adenosine criphosphate and creatinr phosphate content in muscle. Neurology 1981;31:916-9 19 4. Gibbs C. The cytoplasmic phosphorylation potential. J Mol Cardiol 1385;17:727-731 5. Gyulai L, Roth 2, Leigh JS Jr, Chance B. Bioenergetic studies of mitochondrial oxidative phosphorylation using "P NMR. J Biol Chem 1985;260:3947-3954 6. Meyer RA, Brown TR, Kushmerick J. Phosphorus nuclear magnetic resonance of fast and slow twitch muscle. Am J Physiol 1985;248:C279-C28? 7. Chance B, Eleff S, Leigh JS Jr, et al. Mitochondrial regulation of phosphocreatineiinorganicphosphate ratios in exercising human muscle: a gated ilP KMR study. Proc Natl Acad Sci USA I98 1;78:6714-67 18 8. Chance B, Leigh JS Jr, Clark BJ, et al. Control of oxidative metabolism and oxygen delivery in human skeletal muscle: a steady state analysis of the work/energy cost transfer function. Proc Natl Acad Sci USA 1985;82:8384-8388 9. Chance B, Clark BJ, Nioka S, et al. Phosphorus nuclear magnetic resonance spectroscopy in vivo. Circulation 1985;72(suppl 4):103-110 10. Arnold DL, Matthews PM, Radda GK. Metabolic recovery after exercise and the assessment of mitochondrial function in human skeletal muscle in vivo by means of ilP NMR. Magn Reson Med 1984;1:307-315 11. Minotti JR, Johnson EC, Hudson TL, et al. Forearm metabolic asymmetry detected by 31P-NMR during submaximal exercise. J Appl Physiol 1989;67:324-329
Brief Review: Argov and Bank: "P MRS in Neuromuscular Disorders 95
12. Gadian DG, Radda GK, Ross BD, et al. Examination of a myopathy by phosphorus nuclear magnetic resonance. Lancet 1981; 2:774-775 13. Radda GK, Bore PJ, Gadian DG, et al. ”P NMR examination of wo patients with NADH-CoQ reductase deficiency. Nature 1982;295:608-609 14. Arnold DL, Taylor DJ, Radda GK. Investigation of human mitochondrial myoparhies by phosphorus magnetic resonance spectroscopy. Ann Neurol 1985;18:189-196 15. Argov Z , Bank WJ, Maris J, et al. Bioenergetic heterogeneity of human mitochondrial myopathies: phosphorus magnetic resonance spectroscopy study. Neurology 1985;37:257-262 16. Engel AG. Carnitine deficiency syndromes and lipid storage myopathies. In: Engel AG, Banker BQ, eds. Myology. New York McGraw-Hill, 1986:1663-1696 17. Argov 2, Renshaw P, Roden B, et al. The effects of thyroid hormones on skeletal muscle bioenergetics: an in vivo 31PNMR study of humans and rats. J Clin Invest 1988;81:1695-1701 18. Olgin J, Argov Z , Rosenberg H, Chance B. Non invasive evaluation of malignant hyperthermia susceptibility with phosphorus nuclear magnetic resonance spectroscopy. Anesthesiology 1988;68:507-5 13 19. Argov Z. Phosphorus magnetic resonance spectroscopy (31P NMR) as a tool for in vivo monitoring of mitochondrial muscle disorders. In: Azzi A, Drahota Z, Papa S, eds. Molecular basis of membrane-associated diseases. Berlin: Springer-Verlag, 1989: 183 - 199 20. McCully KK, Argov Z , Boden B, et al. Detection of muscle injury in humans with 3’P NMR magnetic resonance spectroscopy. Muscle Nerve 1988;11:212-216 21. Hayes DJ, Byrne E, Shoubridge EA, et al. Experimentally induced defects of mitochondrial metabolism in rat skeletal muscle. Biochem J 1985;229:109-117 22. Argov 2, Maris J, Bank W, Chance B. Postexercise recovery of phosphocreatine is p H independent: 31-P NMR study. Presented at the 30th International Congress of Physiological Science, Sept 1986, Vancouver, Canada. 23. Ross BD, Radda GK, Gadian DG, et al. Examination of a case of suspected McArdle’s syndrome by 3 1P NMR nuclear magnetic resonance. N Engl J Med 1981;304:1338-1342 24. Edwards RHT, Dawson DJ, Wilkie DR, et al. Clinical use of magnetic resonance in rhe investigation of myopathy. Lancet 1982;1:?2>-?3O 25, Chance B, Eleff S, Bank WJ, et al. 31-P NMR studies of control of mitochondrial function in phosphofructokinase-deficienthuman skeletal muscle. Proc Natl Acad Sci USA 1982;79: 7714-7718 26. Duboc D, Jenhenson P, Tran Dinh S, et al. Phosphorus NMR spectroscopy study of muscular enzyme deficiencies involving glycogenolysis and glycolysis. Neurology 1987;37:663-671 27. Argov Z , Bank WJ, Boden B, et al. Muscle 31P NMR in partial glycolytic block: in vivo study of phosphoglycerate mutase deficient patient. Arch Neufol 1987;44:614-617 28. Giger U, Argov Z, Schnall M, Chance B. Metabolic myopathy in a canine model for muscle-qpe phosphofructokinase deficiency. Muscle Nerve 1988;11:1260-1265 29. Kuwabara T, Yuasa T, Miyatake T. 31-P NMR studies on an animal model of human defective muscle glycolysis. Muscle Nerve 1986;9:138- 143 30. Argov Z , Nagle D, Giger U, Leigh JS. Muscle bioenergetics in acute glycolytic block: in vivo 3 1P-NMR study of iodoacetate injected rats. Eur J Appl Physiol 1989;58:808-815 31. Argov 2, Bank WJ, Maris J, et al. Muscle energy metabolism in phosphofructokinase deficiency as recorded by 3 I-P NMR. Ann Neurol 1987;22:46-5 1 32. Argov 2, Bank WJ, Maris J, Chance B. Muscle energy metabolism in McArdle’s syndrome by in vivo phosphorus magnetic
96 Annals of Neurology Vol 30 No 1 July 1991
resonance spectroscopy (31P NMR). Neurology 1987;37: 1720-1 724 33. Lewis SF, Hder RG, Cook JD, Nunnaly RL. Muscle fatigue in McArdle’s disease studied by 31PNMR: the effect of glucose infusion. J Appl Physiol 1985;59:1991-1994 34. Rowland LP, DiMauro S , Layzer RB. Phosphofructokinase deficiency. In: Engel AG, Banker BQ, eds. Myology. New York: McGraw-Hill, 1986:1603-16 18 35. Rowland LP, Araki S, Carmel P. Contracture in McArdle’s disease. Arch Neurol 1965;13:541-j44 36. Brumback R. Iodoacetate inhibition of glyceraldehyde-3-phosphate dehydrogenase as a model of human myophosphorylase deficiency (McArdle’s disease) and phosphofructokinase deficiency (Tarui’s disease). J Neurol Sci 1980;48:383-398 37. Ruff RL, Weissman J. Possible role of ADP in contracture of muscle with impaired myoglycolysis. Neurology 1989;39(suppl 1):360 (Abstract) 38. Haller RG, DiMauro S, Vora S, Lewis SF. Glucose impairs exercise performance in muscle phosphofructokinase deficiency: the out of wind effect. Neurology 1988;38(suppI 1):270 (abstract) 39. Bogusky RT, Taylor RG, Anderson LJ, et al. McArdle’s disease heterozygotes: metabolic adaptation assessed using 3 1P-nuclear magnetic resonance. J Clin Invest 1987;77:1881-1887 40. Newman RJ, Bore PJ, Chan L, et al. Nuclear magnetic resonance studies of forearm muscle in Duchenne dystrophy. Br Med J 1982;284:1072-1074 41. Younkin DP, Berman P, Sladky J, et al. ” P NMR studies in Duchenne muscular dystrophy: age related metabolic changes. Neurology 1987;37:257-262 42. Griffiths RD, Cady EB, Edwards RHT, Wilkie DR. Muscle energy metabolism in Duchenne dystrophy studied by j’P N M R controlled trials show no effect of allopurinol or ribose. Muscle Nerve 1985;8:760-767 43. Burt CT. Phosphodiesters and N M R a tale of rabbits and chickens. Trends Biochem Sci 1985;10:404-406 44. Pettegrew J, Minshew N, Fiet H . j’P NMR studies of normal and dystrophic chicken muscle. Muscle Nerve 1984;7:442-446 45. Bank WJ, McCully K, Giger U, et al. Muscle injury after exercise in canine muscular dystrophy, measured by 31PNMR. Neurology 1989;39(suppl 1):337 (Abstract) 46. Argov 2, Maris J, Damico L, Chance B. Mitochondrial malfunction of dystrophic hamster muscle: in vivo 31PNMR study. J Neurol Sci 1988;86:185-193 47. SatrusteguiJ, Berkowitz H , Boden B, et al. An in vivo phosphorus nuclear magnetic resonance study of the variations with age in the phosphodiesters’ content of human muscle. Mech Ageing D ~ 1988;42: v 105-1 14 48. Arnold DL, Bore PJ, Radda GK, et al. Excessive intracellular acidosis in skeletal muscle on exercise in a patient with post-viral exhaustionifatigue syndrome, a 3 1P nuclear magnetic resonance study. Lancet 1984;l: 1367-1369 49. Gross B, Glasberg M, Kensora T, et al. Early acidosis in mydgia and tubular aggregates: 3 1P magnetic resonance spectroscopy study. Ann Neurol 1988;24:165 50. Lehmann-Horn F, Hopfel D, Rude1 R, et aL In vivo P-NMR spectroscopy: muscle energy exchange in paramyotonia patients. Muscle Nerve 1985;8:606-610 5 1, Heppenstall RB, Sapega AA, Izant T, et al. Compartment syndrome: a quantitative study of high energy phosphorus compounds using 3 1P-magnetic resonance spectroscopy. J Trauma 1989;29:1113-1119 52. Mancini DM, Ferraro N, Tuchler M, et al. Detection of abnormal calf muscle metabolism in patients with heart failure using phosphorus-3 1 nuclear magnetic resonance. Am J Cardiol 1988;62:1234-1240 53. Rajagopalan B, Conway MA, Massie B, et al. Alteration of skeletal muscle metabolism in humans studied by phosphorus 31
magnetic resonance spectroscopy in congestive heart failure. Am
J Cardiol 1988;62:E53-57 54. Bollaert PE, Robin-Lherbier B, Escanve JM, et al. Phosphorus nuclear magnetic resonance evidence of abnormal skeletal muscle metabolism in chronic alcoholics. Neurology 1989;39: 82 1-824 55. Zochodne DW, Thompson RT, Driedger AA, et al. Metabolic changes in human muscle denervation: topical 3 1P NMR spectroscopy studies. Magn Reson Med 1988;7:373-383 56. Radda GK. The use of NMR spectroscopy for the understanding of disease. Science 1986;233:640-645 57. Eleff S, Kennaway NG, Buist NRM, et al. 3'P NMR study of improvement in oxidative phosphorylation by vitamins K3 and C in a patient with a defect in electron transport at complex 3 in skeletal muscle. Proc Natl Acad Sci USA 1984;81:3529-3533 58. Argov 2, Chance B, Maris J, et al. Treatment of rnitochondrial myopathy due to complex 3 deficiency with vitamins K3 and C: a follow-up study. Ann Neurol 1986;19:598-602 59. Atgov 2, Maris J, Fischbeck K, et al. In vivo study of human
60.
61.
62.
63.
64.
lipid rnyopathies by "P NMR spectroscopy. Ann Neurol 1986;18:119 (Abstract) Heiman-Patterson TP, Argov 2, DiMauro S, et al. 31PNBMR studies during methylprednisolone treatment of a familial steroid-responsive mitochondrial disorder. Neurology 1989;39 (suppl 1):337 (Abstract) Nagashwi Y , Takahashi M, Yorifuji S, et al. Long-term coenzyme Q l O therapy for a mitochondrial encephalomyopathy with cytochrome c oxidase deficiency: a 31P NMR study. Neurology 1989;39:399-403 Bank WJ, Argov Z. The value of 31PNMR in the diagnosis and monitoring the course of human myopathies. Ann N Y Acad Sci 1988;508:448-450 Challiss RA, Blackledge MJ, Radda GK. Spatial heterogeneity of metabolism in skeletal muscle in vivo studied by 3 l-spectroscopy. Am J Physiol 1988;254:C417-422 Lawson JW,Veech RL. Effects of p H and free M 2 + on the Kcqof the creatine reaction and other phosphate hydtolyses and phosphate transfer reaction. J Biol Chem 1979;254:6528-6537
Brief Review: Argov and Bank: 31P MRS in Neuromuscular Disorders
97
Phosphorus Magnetic Resonance Spectroscopy ( 3 'P MRS) in Neuromuscular Disorders -
A
Zohar Argov, MD," and William J. Bank, MDt
Phosphorus magnetic resonance spectroscopy monitors muscle energy metabolism by recording the ratio of phosphocreatine to inorganic phosphate at rest, during exercise, and during recovery from exercise. In mitochondrial diseases, abnormalities may appear during some or all these phases. Low phosphocreatine-inorganic phosphate ratios at rest are not disease-specific, but can be increased by drug therapy in several myopathies. Phosphorus magnetic resonance spectroscopy can also record intracellular pH and thus identify disorders of glycogen metabolism in which the production of lactic acid is blocked during ischemic exercise. The measurements of accumulated sugar phosphate intermediates further delineate glycolytic muscle defects. Myophosphorylase deficiency responds to intravenous glucose administration with improved exercise bioenergetics, but no such response is seen in phosphofructokinase deficiency. The muscular dystrophies show no specific bioenergetic abnormality; however, elevation of phospholipids metabolites and phosphodiesters was detected in some cases. While phosphorus magnetic resonance spectroscopy remains primarily a research tool in metabolic myopathies, it will be clinically useful in identifying new therapies and monitoring their effects in a variety of neurornuscular disorders. Argov 2, Bank WJ. Phosphorus magnetic resonance spectroscopy (31PMRS) in neuromuscular disorders. Ann Neurol 1991;30:90-97
The biochemical investigation of neuromuscular disorders has long relied only on in vitro examination of muscle tissue obtained by biopsy. In vivo phosphorus magnetic resonance spectroscopy ("P MRS) now permits repeated, noninvasive, real-time studies of muscle energy metabolism at rest and during exercise. We highlight the basic concepts of muscle 31PMRS and address those clinical muscle disorders in which 31P MRS observations have been relevant. The analytical sensitivity of in vivo 31PMRS is limited, since metabolites cannot be measured at less than 0.3 to 0.5 mM in a 2 T magnetic field. Fortunately, the five major peaks present at such concentrations and visible in a normal resting muscle spectrum (Fig 1) are associated with energy metabolism. Three peaks identify the distinct positions of each of the phosphate atoms of ATP. The remaining peaks are those of phosphocrearine (PCr) and inorganic phosphates (Pi). The spectral distance between PCr and Pi peaks varies with pH, permitting the determination of intracellular p H of muscle cells [I]. The area under each peak is deter-
mined by the amount of each metabolite in the sampled muscle volume. This volume depends on the size of the external detection coil used in in vivo studies. Current MRS techniques eliminate most of the phosphorus signal from nonmuscle tissues. Calibration to absolute concentrations may be achieved by comparison with an external phosphate standard of known concentration. ATP concentration, however, has been traditionally used as an internal standard {2) because it was thought to remain constant in normal working muscle. A constant concentration of phosphoruscontaining compounds, however, cannot be assumed in metabolic disorders or muscular dystrophies [3). Therefore, the relative ratios of compounds are preferential measurements to single peak amounts for quantitative evaluations. Such ratios also have a bioenergetic significance. Muscle energetic state is characterized by the balance between ATP utilization and ATP production or by the cytosolic phosphorylation potential (PP) {4, 51. The PP plays an important role in the control of mitochondrial respiration and is directly propor-
From the "Department of Neurology, Hadassah University Hospital, Jerusalem, Israel, and the +Department of Neurology, Hospiral of University of Pennsylvania, Philadelphia, PA.
Address correspondence to Dr Argov, Department of Neurology, Hadassah University Hospital, Jerusalem 91 120, Israel.
Received Nov 10, 1989, and in revised form Mar 19 and Dec 11, 1990, and Jan 4 , 1991. Accepted for publication Jan 4 , 1991.
90 Copyright 0 1991 by the American Neurological Association
PI
AT P
I0
5
0
-5 p m
-10
-15
-20
Fig 1 . 3 1 PMRS spectra of normal muscle during rest and exercise. The positions of the three phosphate atoms of A T P are marked as a, /3, and y. The regions where phosphodiesters (PDE) and phosphomonoesters (PME) appear when elevated are marked by arrows. Note the fall of PCr and rise of P i while exercise is increased. ATP remains stable during such louds of work. PCr = phosphocreatine; Pi = inorganic phosphate; S = the chemical shift between PCr and P i used fur pH detmination; ppm = parts per million calibration of the spectral distance.
tional to the ratio of PCr/Pi at a given pH, both measurable by 31P MRS (see Addendum). The measured ratio of PCr to Pi reflects an average of data from heterogeneous human muscle cells that may have varying rates of oxidative metabolism. This applies both to normal muscle and to its pathological conditions. Such heterogeneity has been confirmed in the cat, where slow-twitch, oxidative muscle has significantly lower PCrIPi values than fast-twitch, glycolytic fibers {b}. A high-resolution spectrum of resting muscle can be recorded within minutes. During moderate exercise, PCr falls with a commensurate rise in Pi; ATP remains constant (see Fig 1). These rapid changes in PCr and Pi cannot be accurately represented in the time resolution of in vivo 31PMRS (about 1 minute) [73. Steadystate exercise studies, however, overcome this problem 17, 81. In such a protocol a subject performs constant muscle work (measured by an ergometer) until a stabilized PCriPi is achieved. Similar steady-states obtained at increased levels of work permit a plot of work performed versus Pi/PCr [S, 97. The initial slope of such a plot reflects the relationship of work to energy expenditure and enables the comparison of metabolic capacity between different subjects as well as changes in an individual over time or in response to treatment. At the cessation of exercise, 31PMRS spectra rapidly return to their resting state characteristics. In contrast to energy metabolism during exercise involving glycolytic activity, the process of PCr recovery is governed solely by oxidative metabolism. Postexercise recovery is therefore a good measure of mitochondrial function
{lo]. The rate of recovery in normals depends on the extent of exercise performed. The greater the endexercise PCr depletion, the slower the rate of recovery [101. In order to accurately compare rates of postexercise recovery, similar degrees of PCr depletion must be achieved at the end of exercise. Technical and physical limitations of in vivo 31P MRS must also be considered. Measurement of work rate is difficult because of equipment limitations in a high magnetic field. Also the physically measured work and its relation to biochemical work are dependent on the type of exercise performed (isometric, isokinetic, etc.). In order to make valid comparisons between normal and various disease states, maximum power (work capacity) for each muscle tested must be assessed. This can be difficult in weak patients who may be unable to perform steady-state exercise or who cannot overcome the minimal resistance of an ergometer. Diminished muscle bulk in pathological conditions can also affect the accuracy of measurements. Indeed, even the normal asymmetry between dominant and nondominant arm muscles may affect exercise studies [111.
Mitochondrial Myopathies Disorders of mitochondrial metabolism are particularly suited for 31P MRS studies since the latter primarily monitors oxidative metabolism. Features of oxidative metabolism can be measured by 31PMRS during rest, work, and recovery after exercise in patients with mitochondrial myopathies (MM), and illustrate many of the abnormalities that can be demonstrated with this technique.
Rest PCrlPi Normal muscle at rest has a low metabolic rate and a high energy capacity, resulting in a high PCr/Pi [7]. Impaired mitochondrial function reduces the energy state and decreases PCr/Pi at rest. One of the first human 31PMRS studies showed an elevation of Pi (low PCrlPi) at rest in a patient with MM [123. Subsequent studies showed that the resting state could also be normal in some patients with MM [13]. In two large series, 66 and 83% of patients with MM had low PCr/ Pi at rest [ 14,151. A low PCr/Pi at rest is, however, not a disease-specific abnormality (Table 1). Mitochondrial energy pathways may be affected in carnitine deficiency [l6} and hypothyroidism { 171. Uncoupled mitochondrial respiration has been suspected in malignant hyperthermia 1181. The majority of disorders listed in Table 1, however, do not represent a known primary disturbance in energy metabolism, and secondary mitochondrial malfunction is suggested [19]. PCr/Pi at rest can also be acutely altered in response to transient muscle injury in normal subjects after extensive exercise 1201.
Brief Review: Argov and Bank: 31P MRS in Neuromuscular Disorders
91
Table 1 . Eiuman Neuromuscular Disorakrs Associated with Low Phospbocreatine-Inorganic Phosphate Ratios at Rest
1. Mitochondrial myopathies 2. Primary carnitine deficiency 3 . State after muscle contracture in myophosphorylase deficiency
4 . Hypothyroid myopathy 5. Muscular dystrophy 6. Myotonic dystrophy 7. Motor neuron and other denervating disorders 8. Polymyositis 9. Muscle injury 10. Recurrent myoglobinuria of undetermined origin 11. Malignant hyperrhermia suscepribles
Chungar during ExerciJe The phosphorylation potential rapidly diminishes in patients with MM when metabolic demands are stressed during exercise. Although such patients perform less work than normal subjects, a greater decrease in PCr/Pi occurs at the end of exercise C141. Only 5 of 18 patients with MM could exercise at multiple steady-states, which are necessary to plot work rate versus PiiPCr. The initial slope of these plots was abnormally low in these patients fl5]. Although both studies indicate an abnormally early and rapid reduction of PCr in patients with MM during exercise, this abnormahty has not yet been quantitated. Impaired oxidative metabolism should result in enhanced glycolytic activity and a subsequently greater intracellular lactic acid accumulation. Surprisingly, a relative resistance to intracellular acidosis has been seen in MM-affected muscles 12, 12, 14, 151. The mechanism of this phenomenon is not understood; however, an adaptation of intracellular buffering systems responding to an overproduction of lactate has been proposed 114, 151.
slow rate of recovery was seen in only 30% of patients in one study {143. Data corrected ro comparable endexercise PCrlPi values, however, showed that I? (94%) of 18 patients had a prolonged recovery L151. A delay in postexercise recovery of P W P i is therefore a most sensitive 31PMRS indicator of mitochondrial malfunction. No correlation can be made between MRS findings and the site of the defect in the mitochondrial respiratory chain. There may also be variability of 31PMRS data in patients with a similar rnitochondrial defect [ 15, 191. The injection of known mitochondrial inhibitors in animals showed acute 31PMRS abnormalities during exercise and recovery after exercise but not at rest r21). ilP MRS findings correlate best with the degree of muscle function impairment in patients with MM {19}. Some mitochondrial cytopathies may show no 31P MRS abnormality, indicating that skeletal muscle mitochondria are probably not involved. Four generalized categories of 31PMRS findings in patients with MM are compiled in Table 2.
Glycolytic Disorders Impaired glycolysis or glycogenolysis blocks the ability of muscle to produce lactic acid during anaerobic exercise. Normal muscle develops cytoplasmic acidification during intense work and an intracellular p H as low as 6.0 has been observed by 31PMRS 122). Absence of intracellular acidosis, as measured by 31P MRS after ischemic exercise, confirms disorders of deficient muscle glycolysis [23, 241. Glycolytic disorders differ from disorders of glycogenolysis in that glycolytic intermediates accumulate as phosphorylated sugars in the former. These phosphomonoesters (PME) are identified to the left of the Pi peak (Fig 2). The accumulation of PME was first noted in patients lacking muscle phosphofructokinase (PFK) {24, 251. Subsequently, disorders more Qstal in the glycolytic sequence (phosphoglyceratekinase {26] and phosphoglyceratemutase EPGAM} [27}) showed a similar phenomenon. A rise in PME has also been documented in PFK-deficient
Postexercise Recmev The repletion of PCr after exercise is primarily controlled by mitochondrial oxidative metabolism, yet a
Table 2. Summa7y of 31PMRS Findings in Patients with Mitochondrial Myopatbies (MM) Normal Findings
1. MM without skeletal muscle involvement 2. Nonspecific finding of diseased muscle 3. Mitochondrial malfunctionprimary or secondary" 4. Primary mitochondrial disease
Low PCr/Pi at Rest
Prolonged PCr Recovery
Abnormal Work Kinetics
+ +
+ +
+
aEither one or both abnormahties may exist. PCr = phosphocreatine; Pi = inorganic phosphate.
92 Annals of Neurology Vol 30 No 1 July 1991
+ +
I
40 30 20 10
PME P O
0 -10 -20 -30 -40 ppm
Fig 2. Rest and exercise spectra of a patient with phosphofructokinase deficiency. Note the accumulation of phosphornonoesters (PME) daring the exercise. Pi = inorganic phosphate; PCr = phosphocreatine; ppm = parts per million.
dogs 128) and acutely blocked glycolysis induced by iodoacetate injections in rats 129, 30). In muscles with impaired glycolysis, PME rise promptly and continuously even during aerobic work, demonstrating the need for glucose as “muscle fuel” even in mild exercise. The severity of PME accumulation may be an indicator of the degree of enzymatic block 127, 30). PME levels are not increased in PFK-deficient muscle at rest and their postexercise recovery shows an exponential curve 131). The postexercise recovery of Pi in these cases is slower than that of PCr, indicating that dephosphorylation is involved in PME clearance [31). Compared with PFK deficiency, PME accumulation was lower in a patient with a partial defect of PGAM and the recovery rate of PME was faster 1271. This may be due to additional clearance of sugar phosphates by the residual glycolytic activity. The relationship between the degree of glycolytic block and PME kinetics was also confirmed in the acute animal model 1301. It has previously been assumed that intracellular p H determined the rate of PCr/Pi recovery after exercise in normal muscle, since greater acidosis was associated with slower recovery [ l o ] . In patients with myophosphorylase (MPH) deficiency, who do not develop acidosis, the recovery of PCr/Pi was proportional to the end-exercise PCr/Pi as in normal subjects [22, 32). This indicates that factors other than cytoplasmic pH are also responsible for determining the rate of postexercise recovery. The type of exercise that patients perform may determine the type of abnormalities observed by ”P MRS. For example, a diminished rise in Pi was demonstrated in PFK deficiency after a short bout of exercise and led to the suggestion that phosphates are “trapped” by the accumulating sugar phosphates [25). In contrast, a continuous and higher rise of Pi was seen in the same PFK-deficient patient during a prolonged, graded type of exercise [3 1). Clinically, patients with glycolytic disorders can perform prolonged, graded work better than short, rapid bouts of exercise. These variances must be considered when comparing 31PMRS observa-
tions of similar disorders. Lewis and colleagues 133) demonstrated a higher end-exercise Pi in MPH deficiency compared with normal subjects and suggested that oxidative metabolism is also impaired in this disorder. In PFK-deficient dogs, PCr/Pi levels were higher than control levels after similar rates of stimulated muscle work I281. Whether this indicates impaired muscle aerobic metabolism in these disorders is not yet clear. Painful exercise intolerance and muscle contracture, the main features of glycolytic muscle disorders, remain unexplained. A possible explanation has been ATP depletion during exercise C34). Biochemical examination of muscle from an MPH-deficient patient during an ischemic contracture did not, however, demonstrate loss of ATP [35). Similarly, 31PMRS showed no loss of ATP during exercise or during a contracture in an MPH-deficient patient [23, 32). In contrast, PEK-deficient muscle, both in humans { 3 1) and in dogs {ZS), showed a continuous mild loss of ATP during graded, submaximal, nonischemic exercise. This observation was not associated with exercise intolerance or a clinical contracture. In the model of acute glycolytic block in rats, a gradual loss of ATP in both biochemical t36) and in vivo 31PMRS studies {30) was found. Although PFK- and MPH-deficient patients have similar symptoms, their ATF’ kinetics during exercise differ and the causal relationship between ATP concentration and contracture remains unresolved C34, 37). The “second wind” phenomenon is typical in MPH deficiency and results in increased work capacity after a short bout of exercise and rest. It is rare, if present, in PFK deficiency. An improved end-exercise PCr/Pi has been observed in MPH deficiency after an intravenous glucose infusion 1331. The slope of work versus Pi/PCr during graded exercise is also augmented by glucose 1321. In PFK deficiency the plot of this slope remains unchanged [31) or becomes even lower [ 3 8 } under similar conditions. Blood-borne glucose may, therefore, be the mediator of the “second wind” phenomenon. Other yet untested substrates may also contribute. Two patients with a proven partial MPH deficiency showed a greater degree of intracellular acidosis during aerobic exercise than during ischemic exercise 1391. In normals, lactic acid production is greater during ischemic exercise 1391. It has been suggested that these carriers, heterozygous for MP H deficiency, represent a unique metabolic adaptation enhancing muscle use of serum glucose. 31P MRS is therefore useful in identifying several disorders of glycogen metabolism by the absence of intracellular acidosis during exercise. It can also distinguish between glycogenolytic and glycolytic disorders by the accumulation of PME. In addition, it is helpful in understanding metabolic adaptations of muscle in the presence of an enzymatic deficiency.
Brief Review: Argov and Bank: ”P MRS in Neuromuscular Disorders
93
Muscular Dystrophy PCr/Pi at rest is reduced in the forearm 140) and gastrocnemius 1411 muscles in Duchenne muscular dystrophy (DMD). The decrease in PCr/Pi in dystrophic muscles is age related, suggesting that this reflects the dystrophic process 1417. There have been conflicting reports of increased [40,41) and normal 142) intracellular p H in D M D at rest. The estimated total phosphate signal and ATP concentration remain unchanged in patients with D MD [42), in contrast to earlier biochemical data 131. This remains to be verified in a combined study of the same muscle. Several phosphate compounds, unusual for normal muscles, have been detected in DMD. An elevation of phosphodiester (PDE) is observed between the PCr and Pi peaks in arm 140) and leg 141) muscles of patients with DMD, although its exact identity is unknown. This accumulation may reflect a breakdown of phospholipids, coincident with the dystrophic process 143). The PDE serine ethanolamine is elevated in dystrophic chicken muscle 1447. Dystrophin-deficient dogs [453 and dystrophic hamsters 1467, however, show no rise in PDE, despite low PCr/Pi at rest. A rise in PDE can be seen in normal calf muscles with aging 1471 and may therefore be a nonspecific indicator of muscle breakdown. This is further supported by transient PDE accumulation in hypothyroid muscle, which disappeared after 3 months of hormone replacement 117).
Other Neuromuscular Disorders Arnold and colleagues [48] reported a patient with chronic fatigue after a viral illness who had an unusual 31PMRS pattern of early and excessive acidosis during exercise. This condition is, however, controversial in both clinical and MRS data. Two patients with rnyalgia and tubular aggregates unrelated to viral illness showed similar findings 1497. We have also seen this phenomenon in a few sedentary but healthy persons. The syndrome of early acidosis, by 31PMRS criteria, remains therefore speculative since it is not necessarily abnormal or disease-specific. Patients with paramyotonia show no /'P MRS abnormality, even when the cooled muscles are stiff and paralyzed [SO]. A decreased PCr/Pi at rest can be seen in the advanced stages of myotonic dystrophy, as in many other neuromuscular disorders with advanced muscle degeneration (see Table 1). Local muscle ischemia, as seen in "compartment syndrome" { 5 l ) , may change the energy state of muscle. 31PMRS abnormalities seen in exercising skeletal muscle during congestive heart failure were thought to represent the result of diminished muscle blood flow, but may in fact be due to a chronic metabolic defect 152, 533. This must be considered when evaluating a myopathy with cardiac involvement. 94 Annals of Neurology Vol 30 No 1 July 1991
Alcoholics, with and without a history of rhabdomyolysis, demonstrate reduced acidosis during ischemic exercise, suggesting impaired glycolysis [543. Many other neuromuscular disorders, including the neurogenic muscle atrophies 1553, tested by 31P MRS showed nonspecific or normal findings [%I.
31PMRS Monitoring of Therapy Although decreased PCr/Pi at rest is a nonspecific finding, this ratio has been a valuable monitor both of progression of disease and of response to specific therapy. Eleff and associates 157) studied a patient with mitochondrial complex 3 deficiency who had the triad of "P MRS abnormalities typical of primary MM. In an attempt to bypass the biochemical defect, the patient was treated with electron acceptors (vitamins K, and C) in large doses. This resulted in improved clinical and "P MRS findings. PCr/Pi at rest rose from 1.7 to 6.1 and the rate of postexercise recovery increased markedly C57, 58). A correlation between the patient's clinical course and "P MRS findings was maintained during 5 years of treatment 119, 583. Clinical improvement associated with a rise in PCr/Pi at rest has been shown in patients with carnitine deficiency who were treated with L-carnitine [59], familial MM treated with methylprednisolone [60}, and MM treated with coenzyme Qlo161). The improved oxidative metabolism is indicative of improved mitochondrial function in these cases. Response to therapy has also been demonstrated by jlP MRS in disorders in which mitochondria are secondarily involved. Rest spectra improved in hypothyroidism treated with hormone replacement [17] and poiymyositis treated with corticosteroids (62). 31P MRS spectra at rest are particularly advantageous as objective monitors since they can be obtained in young children and severely weak individuals who are unable to perform exercise. Conclusions In vivo "P MRS offers a harmless, noninvasive measurement of muscle bioenergetics. Data are acquired promptly and repeated examinations are feasible. There are, however, technical limitations of calibration and quantification of muscle work, especially in very weak patients. The heterogeneity of skeletal muscle is also difficult to resolve 163). 31P MRS will remain predominantly a research tool in myology and will not replace in vitro biochemical studies for diagnosis. It will be useful for screening patients with exercise intolerance due to metabolic derangements. A scheme for 31P MRS testing of such patients is presented in Figure 3. "'P MRS can also contribute to further understanding of normal muscle bioenergetics and clarify adaptations that occur in the presence of impaired metabolism. Objective monitor-
The ratio of PCriPi is directly measured from the muscle spectrum. The concentration of protons {HI is calculated from the 31P MRS-determined pH. The equilibrium constant {K} is known to be 1.66 x M - ' at p H of 7 [64}. If the creatine content of the muscle is measured, then the PP can be directly calculated {14).
,Exercise ljlolerance, [Rest, Exercise, Recovery ~ r o i o c o ~ ]
I
Full Distal GlyCOlytiC
i
1.
Mixed Pattern
Normal Findmqs
Mitochondria I Malfunclion Primary and
CPT hficiency
31PMRS Glossary PCr-phosphocreatine Pi-inorganic phosphates (H2P04- and H P 0 4 - - ) PME-phosphomonoesters; in muscle, probably phosphorylated sugars and inosine monophosphate PDE-phosphodiesters; in normal human muscle, probably
i
Some PME
I
F'arlid Glycolyflc
glycerol-3-phosphorylcholine
AMP Deominasa oeticiencv?
Fig 3. A simp/ejou: chart for the evahation of exercise intolMRS. This is based on eualzrating a rest-exerciseerance by recotrey protocol and an ischemic exercise test. PCr = phosphocreatine;Pi = znorganicphosphate;PME = phosphomonoesters; CPT = cumitine palmytyl transferase; AMP = adenosine monophosphate.
ing of therapy will be most promising both in humans and in animal models.
ppm-parts p e r million; a unit for measuring the spectral distance, related t o the frequency of the recorded atom PCr/Pi-31P MRS-determined ratio that correlates with t h e phosphorylation potential ~~
Many of the observations reported here were performed while Dr Argov was a Muscular Dystrophy Association Post-Doctoral ReUnisearch Fellow in the Department of Biochemistry/B~ophys~cs, versity of Pennsylvania, Philadelphia Grants from the Muscular Dystrophy Association and the National Institutes of Health (NS 08075) supported OUT human and animal research.
References 1. Taylor DJ, Bore PJ, Styles P, et al. Bioenergetics of intact hu-
man muscle. A gated jlP NMR study. Mol Biol Med 1983;l:
Addendum The phosphorylation potential (PP) is defined as:
77-94
(11 The creatine kinase (CK) reaction that controls the cytosolic ATP concentration and its availability to mitochondria is: PCr
+ ADP + H + + A T P + Cr
where Cr is free creatine. This reaction is normally at or near equilibrium with the following equilibrium constant (Kj: K=
[ATP} x {Cr} {PCr) X {ADP) x [H}
Derived from (2) is: [ATP) =
K x [PCr) x [HI x {ADP]
m-1
which when substituted in the PP equation (1) gives:
(3)
2. Radda GK, Taylor DJ. Application of nuclear magnetic resonance spectroscopy in pathology. Int Rev Exp Pathol 1985;2: 1-58 3. Samaha FJ, Davis B, Nagy D. Duchenne muscular dystrophy: adenosine criphosphate and creatinr phosphate content in muscle. Neurology 1981;31:916-9 19 4. Gibbs C. The cytoplasmic phosphorylation potential. J Mol Cardiol 1385;17:727-731 5. Gyulai L, Roth 2, Leigh JS Jr, Chance B. Bioenergetic studies of mitochondrial oxidative phosphorylation using "P NMR. J Biol Chem 1985;260:3947-3954 6. Meyer RA, Brown TR, Kushmerick J. Phosphorus nuclear magnetic resonance of fast and slow twitch muscle. Am J Physiol 1985;248:C279-C28? 7. Chance B, Eleff S, Leigh JS Jr, et al. Mitochondrial regulation of phosphocreatineiinorganicphosphate ratios in exercising human muscle: a gated ilP KMR study. Proc Natl Acad Sci USA I98 1;78:6714-67 18 8. Chance B, Leigh JS Jr, Clark BJ, et al. Control of oxidative metabolism and oxygen delivery in human skeletal muscle: a steady state analysis of the work/energy cost transfer function. Proc Natl Acad Sci USA 1985;82:8384-8388 9. Chance B, Clark BJ, Nioka S, et al. Phosphorus nuclear magnetic resonance spectroscopy in vivo. Circulation 1985;72(suppl 4):103-110 10. Arnold DL, Matthews PM, Radda GK. Metabolic recovery after exercise and the assessment of mitochondrial function in human skeletal muscle in vivo by means of ilP NMR. Magn Reson Med 1984;1:307-315 11. Minotti JR, Johnson EC, Hudson TL, et al. Forearm metabolic asymmetry detected by 31P-NMR during submaximal exercise. J Appl Physiol 1989;67:324-329
Brief Review: Argov and Bank: "P MRS in Neuromuscular Disorders 95
12. Gadian DG, Radda GK, Ross BD, et al. Examination of a myopathy by phosphorus nuclear magnetic resonance. Lancet 1981; 2:774-775 13. Radda GK, Bore PJ, Gadian DG, et al. ”P NMR examination of wo patients with NADH-CoQ reductase deficiency. Nature 1982;295:608-609 14. Arnold DL, Taylor DJ, Radda GK. Investigation of human mitochondrial myoparhies by phosphorus magnetic resonance spectroscopy. Ann Neurol 1985;18:189-196 15. Argov Z , Bank WJ, Maris J, et al. Bioenergetic heterogeneity of human mitochondrial myopathies: phosphorus magnetic resonance spectroscopy study. Neurology 1985;37:257-262 16. Engel AG. Carnitine deficiency syndromes and lipid storage myopathies. In: Engel AG, Banker BQ, eds. Myology. New York McGraw-Hill, 1986:1663-1696 17. Argov 2, Renshaw P, Roden B, et al. The effects of thyroid hormones on skeletal muscle bioenergetics: an in vivo 31PNMR study of humans and rats. J Clin Invest 1988;81:1695-1701 18. Olgin J, Argov Z , Rosenberg H, Chance B. Non invasive evaluation of malignant hyperthermia susceptibility with phosphorus nuclear magnetic resonance spectroscopy. Anesthesiology 1988;68:507-5 13 19. Argov Z. Phosphorus magnetic resonance spectroscopy (31P NMR) as a tool for in vivo monitoring of mitochondrial muscle disorders. In: Azzi A, Drahota Z, Papa S, eds. Molecular basis of membrane-associated diseases. Berlin: Springer-Verlag, 1989: 183 - 199 20. McCully KK, Argov Z , Boden B, et al. Detection of muscle injury in humans with 3’P NMR magnetic resonance spectroscopy. Muscle Nerve 1988;11:212-216 21. Hayes DJ, Byrne E, Shoubridge EA, et al. Experimentally induced defects of mitochondrial metabolism in rat skeletal muscle. Biochem J 1985;229:109-117 22. Argov 2, Maris J, Bank W, Chance B. Postexercise recovery of phosphocreatine is p H independent: 31-P NMR study. Presented at the 30th International Congress of Physiological Science, Sept 1986, Vancouver, Canada. 23. Ross BD, Radda GK, Gadian DG, et al. Examination of a case of suspected McArdle’s syndrome by 3 1P NMR nuclear magnetic resonance. N Engl J Med 1981;304:1338-1342 24. Edwards RHT, Dawson DJ, Wilkie DR, et al. Clinical use of magnetic resonance in rhe investigation of myopathy. Lancet 1982;1:?2>-?3O 25, Chance B, Eleff S, Bank WJ, et al. 31-P NMR studies of control of mitochondrial function in phosphofructokinase-deficienthuman skeletal muscle. Proc Natl Acad Sci USA 1982;79: 7714-7718 26. Duboc D, Jenhenson P, Tran Dinh S, et al. Phosphorus NMR spectroscopy study of muscular enzyme deficiencies involving glycogenolysis and glycolysis. Neurology 1987;37:663-671 27. Argov Z , Bank WJ, Boden B, et al. Muscle 31P NMR in partial glycolytic block: in vivo study of phosphoglycerate mutase deficient patient. Arch Neufol 1987;44:614-617 28. Giger U, Argov Z, Schnall M, Chance B. Metabolic myopathy in a canine model for muscle-qpe phosphofructokinase deficiency. Muscle Nerve 1988;11:1260-1265 29. Kuwabara T, Yuasa T, Miyatake T. 31-P NMR studies on an animal model of human defective muscle glycolysis. Muscle Nerve 1986;9:138- 143 30. Argov Z , Nagle D, Giger U, Leigh JS. Muscle bioenergetics in acute glycolytic block: in vivo 3 1P-NMR study of iodoacetate injected rats. Eur J Appl Physiol 1989;58:808-815 31. Argov 2, Bank WJ, Maris J, et al. Muscle energy metabolism in phosphofructokinase deficiency as recorded by 3 I-P NMR. Ann Neurol 1987;22:46-5 1 32. Argov 2, Bank WJ, Maris J, Chance B. Muscle energy metabolism in McArdle’s syndrome by in vivo phosphorus magnetic
96 Annals of Neurology Vol 30 No 1 July 1991
resonance spectroscopy (31P NMR). Neurology 1987;37: 1720-1 724 33. Lewis SF, Hder RG, Cook JD, Nunnaly RL. Muscle fatigue in McArdle’s disease studied by 31PNMR: the effect of glucose infusion. J Appl Physiol 1985;59:1991-1994 34. Rowland LP, DiMauro S , Layzer RB. Phosphofructokinase deficiency. In: Engel AG, Banker BQ, eds. Myology. New York: McGraw-Hill, 1986:1603-16 18 35. Rowland LP, Araki S, Carmel P. Contracture in McArdle’s disease. Arch Neurol 1965;13:541-j44 36. Brumback R. Iodoacetate inhibition of glyceraldehyde-3-phosphate dehydrogenase as a model of human myophosphorylase deficiency (McArdle’s disease) and phosphofructokinase deficiency (Tarui’s disease). J Neurol Sci 1980;48:383-398 37. Ruff RL, Weissman J. Possible role of ADP in contracture of muscle with impaired myoglycolysis. Neurology 1989;39(suppl 1):360 (Abstract) 38. Haller RG, DiMauro S, Vora S, Lewis SF. Glucose impairs exercise performance in muscle phosphofructokinase deficiency: the out of wind effect. Neurology 1988;38(suppI 1):270 (abstract) 39. Bogusky RT, Taylor RG, Anderson LJ, et al. McArdle’s disease heterozygotes: metabolic adaptation assessed using 3 1P-nuclear magnetic resonance. J Clin Invest 1987;77:1881-1887 40. Newman RJ, Bore PJ, Chan L, et al. Nuclear magnetic resonance studies of forearm muscle in Duchenne dystrophy. Br Med J 1982;284:1072-1074 41. Younkin DP, Berman P, Sladky J, et al. ” P NMR studies in Duchenne muscular dystrophy: age related metabolic changes. Neurology 1987;37:257-262 42. Griffiths RD, Cady EB, Edwards RHT, Wilkie DR. Muscle energy metabolism in Duchenne dystrophy studied by j’P N M R controlled trials show no effect of allopurinol or ribose. Muscle Nerve 1985;8:760-767 43. Burt CT. Phosphodiesters and N M R a tale of rabbits and chickens. Trends Biochem Sci 1985;10:404-406 44. Pettegrew J, Minshew N, Fiet H . j’P NMR studies of normal and dystrophic chicken muscle. Muscle Nerve 1984;7:442-446 45. Bank WJ, McCully K, Giger U, et al. Muscle injury after exercise in canine muscular dystrophy, measured by 31PNMR. Neurology 1989;39(suppl 1):337 (Abstract) 46. Argov 2, Maris J, Damico L, Chance B. Mitochondrial malfunction of dystrophic hamster muscle: in vivo 31PNMR study. J Neurol Sci 1988;86:185-193 47. SatrusteguiJ, Berkowitz H , Boden B, et al. An in vivo phosphorus nuclear magnetic resonance study of the variations with age in the phosphodiesters’ content of human muscle. Mech Ageing D ~ 1988;42: v 105-1 14 48. Arnold DL, Bore PJ, Radda GK, et al. Excessive intracellular acidosis in skeletal muscle on exercise in a patient with post-viral exhaustionifatigue syndrome, a 3 1P nuclear magnetic resonance study. Lancet 1984;l: 1367-1369 49. Gross B, Glasberg M, Kensora T, et al. Early acidosis in mydgia and tubular aggregates: 3 1P magnetic resonance spectroscopy study. Ann Neurol 1988;24:165 50. Lehmann-Horn F, Hopfel D, Rude1 R, et aL In vivo P-NMR spectroscopy: muscle energy exchange in paramyotonia patients. Muscle Nerve 1985;8:606-610 5 1, Heppenstall RB, Sapega AA, Izant T, et al. Compartment syndrome: a quantitative study of high energy phosphorus compounds using 3 1P-magnetic resonance spectroscopy. J Trauma 1989;29:1113-1119 52. Mancini DM, Ferraro N, Tuchler M, et al. Detection of abnormal calf muscle metabolism in patients with heart failure using phosphorus-3 1 nuclear magnetic resonance. Am J Cardiol 1988;62:1234-1240 53. Rajagopalan B, Conway MA, Massie B, et al. Alteration of skeletal muscle metabolism in humans studied by phosphorus 31
magnetic resonance spectroscopy in congestive heart failure. Am
J Cardiol 1988;62:E53-57 54. Bollaert PE, Robin-Lherbier B, Escanve JM, et al. Phosphorus nuclear magnetic resonance evidence of abnormal skeletal muscle metabolism in chronic alcoholics. Neurology 1989;39: 82 1-824 55. Zochodne DW, Thompson RT, Driedger AA, et al. Metabolic changes in human muscle denervation: topical 3 1P NMR spectroscopy studies. Magn Reson Med 1988;7:373-383 56. Radda GK. The use of NMR spectroscopy for the understanding of disease. Science 1986;233:640-645 57. Eleff S, Kennaway NG, Buist NRM, et al. 3'P NMR study of improvement in oxidative phosphorylation by vitamins K3 and C in a patient with a defect in electron transport at complex 3 in skeletal muscle. Proc Natl Acad Sci USA 1984;81:3529-3533 58. Argov 2, Chance B, Maris J, et al. Treatment of rnitochondrial myopathy due to complex 3 deficiency with vitamins K3 and C: a follow-up study. Ann Neurol 1986;19:598-602 59. Atgov 2, Maris J, Fischbeck K, et al. In vivo study of human
60.
61.
62.
63.
64.
lipid rnyopathies by "P NMR spectroscopy. Ann Neurol 1986;18:119 (Abstract) Heiman-Patterson TP, Argov 2, DiMauro S, et al. 31PNBMR studies during methylprednisolone treatment of a familial steroid-responsive mitochondrial disorder. Neurology 1989;39 (suppl 1):337 (Abstract) Nagashwi Y , Takahashi M, Yorifuji S, et al. Long-term coenzyme Q l O therapy for a mitochondrial encephalomyopathy with cytochrome c oxidase deficiency: a 31P NMR study. Neurology 1989;39:399-403 Bank WJ, Argov Z. The value of 31PNMR in the diagnosis and monitoring the course of human myopathies. Ann N Y Acad Sci 1988;508:448-450 Challiss RA, Blackledge MJ, Radda GK. Spatial heterogeneity of metabolism in skeletal muscle in vivo studied by 3 l-spectroscopy. Am J Physiol 1988;254:C417-422 Lawson JW,Veech RL. Effects of p H and free M 2 + on the Kcqof the creatine reaction and other phosphate hydtolyses and phosphate transfer reaction. J Biol Chem 1979;254:6528-6537
Brief Review: Argov and Bank: 31P MRS in Neuromuscular Disorders
97
