Dynamic shim updating: a new approach towards optimized whole brain shimming

The static magnetic field within two widely spaced axial slices of the human brain was mapped in five subjects following global shimming. This revealed a first order field shift in the anterior-posterior direction between the cerebellum and cerebrum, which has implications for functional and spectroscopic magnetic resonance imaging. A new method is described called dynamic shim updating (DSU) to compensate for these field differences whereby the shim correction fields are updated in real time during multislice data acquisition to match the current imaging or spectroscopy slice. A hardware unit is presented to demonstrate the method using the first order shim corrections, which can be updated virtually instantaneously between slice acquisitions to give optimal shimming of each slice. The efficiency of the approach is demonstrated using field mapping and high speed MR imaging (echo-planar imaging), which are sensitive to field inhomogeneity.

Noninvasive assessment of hepatic triglyceride content in humans with 13C nuclear magnetic resonance spectroscopy

Hepatic lipid content was assessed noninvasively in 15 patients with hepatic steatosis by 13C nuclear magnetic resonance (NMR) spectroscopy, and compared in a double-blind fashion with histological grading and morphometric quantitation of fat in liver biopsies taken within 2 weeks of the study. The lipid content in the liver biopsies was expressed as the volume fraction of total parenchyma occupied by fat. Hepatic triglyceride content was determined by comparing the 13C NMR signal intensity in vivo with the signal intensity obtained from a lipid phantom of known concentrations. There was an approximately 30-fold increase in the 13C NMR signals of the saturated carbons (methyl/methylene [CH2]n) region of hepatic triglycerides from patients with grade 4 steatosis compared with those with grade 0, yielding a good dynamic range for measuring hepatic triglyceride content. The correlation coefficient between the morphometric and 13C NMR techniques was 0.89 (P < .01). These studies demonstrate that 13C NMR spectroscopy can be used to noninvasively assess hepatic triglyceride content in humans. This method may be clinically useful for diagnosis and follow-up of patients with hepatic steatosis.

Mechanism of free fatty acid-induced insulin resistance in humans

To examine the mechanism by which lipids cause insulin resistance in humans, skeletal muscle glycogen and glucose-6-phosphate concentrations were measured every 15 min by simultaneous 13C and 31P nuclear magnetic resonance spectroscopy in nine healthy subjects in the presence of low (0.18 +/- 0.02 mM [mean +/- SEM]; control) or high (1.93 +/- 0.04 mM; lipid infusion) plasma free fatty acid levels under euglycemic (approximately 5.2 mM) hyperinsulinemic (approximately 400 pM) clamp conditions for 6 h. During the initial 3.5 h of the clamp the rate of whole-body glucose uptake was not affected by lipid infusion, but it then decreased continuously to be approximately 46% of control values after 6 h (P < 0.00001). Augmented lipid oxidation was accompanied by a approximately 40% reduction of oxidative glucose metabolism starting during the third hour of lipid infusion (P < 0.05). Rates of muscle glycogen synthesis were similar during the first 3 h of lipid and control infusion, but thereafter decreased to approximately 50% of control values (4.0 +/- 1.0 vs. 9.3 +/- 1.6 mumol/[kg.min], P < 0.05). Reduction of muscle glycogen synthesis by elevated plasma free fatty acids was preceded by a fall of muscle glucose-6-phosphate concentrations starting at approximately 1.5 h (195 +/- 25 vs. control: 237 +/- 26 mM; P < 0.01). Therefore in contrast to the originally postulated mechanism in which free fatty acids were thought to inhibit insulin-stimulated glucose uptake in muscle through initial inhibition of pyruvate dehydrogenase these results demonstrate that free fatty acids induce insulin resistance in humans by initial inhibition of glucose transport/phosphorylation which is then followed by an approximately 50% reduction in both the rate of muscle glycogen synthesis and glucose oxidation.

NMR studies of muscle glycogen synthesis in insulin-resistant offspring of parents with non-insulin-dependent diabetes mellitus immediately after glycogen-depleting exercise

To examine the impact of insulin resistance on the insulin-dependent and insulin-independent portions of muscle glycogen synthesis during recovery from exercise, we studied eight young, lean, normoglycemic insulin-resistant (IR) offspring of individuals with non-insulin-dependent diabetes mellitus and eight age-weight matched control (CON) subjects after plantar flexion exercise that lowered muscle glycogen to approximately 25% of resting concentration. After approximately 20 min of exercise, intramuscular glucose 6-phosphate and glycogen were simultaneously monitored with 31P and 13C NMR spectroscopies. The postexercise rate of glycogen resynthesis was nonlinear. Glycogen synthesis rates during the initial insulin independent portion (0-1 hr of recovery) were similar in the two groups (IR, 15.5 +/- 1.3 mM/hr and CON, 15.8 +/- 1.7 mM/hr); however, over the next 4 hr, insulin-dependent glycogen synthesis was significantly reduced in the IR group [IR, 0.1 +/- 0.5 mM/hr and CON, 2.9 +/- 0.2 mM/hr; (P < or = 0.001)]. After exercise there was an initial rise in glucose 6-phosphate concentrations that returned to baseline after the first hour of recovery in both groups. In summary, we found that following muscle glycogen-depleting exercise, IR offspring of parents with non-insulin-dependent diabetes mellitus had (i) normal rates of muscle glycogen synthesis during the insulin-independent phase of recovery from exercise and (ii) severely diminished rates of muscle glycogen synthesis during the subsequent recovery period (2-5 hr), which has previously been shown to be insulin-dependent in normal CON subjects. These data provide evidence that exercise and insulin stimulate muscle glycogen synthesis in humans by different mechanisms and that in the IR subjects the early response to stimulation by exercise is normal.

Short echo time proton magnetic resonance spectroscopic imaging of macromolecule and metabolite signal intensities in the human brain

A novel approach is presented for imaging macromolecule and metabolite signals in brain by proton magnetic resonance spectroscopic imaging. The method differentiates between metabolites and macromolecules by T1 weighting using an inversion pulse followed by a variable inversion recovery time before localization and spectroscopic imaging. In healthy subjects, the major macromolecule resonances at 2.05 and 0.9 ppm were mapped at a nominal spatial resolution of 1 x 1 x 1.5 cm3 and were demonstrated to be highly reproducible between subjects. In subacute stroke patients, a highly elevated macromolecule resonance at 1.3 ppm was mapped to infarcted brain regions, suggesting potential applications for studying pathological conditions.

Human brain GABA levels rise after initiation of vigabatrin therapy but fail to rise further with increasing dose

Using 1H spectroscopy, we measured occipital lobe GABA levels serially in 18 patients enrolled in an ongoing open label trial of vigabatrin. Brain GABA levels were elevated twofold in patients taking vigabatrin (3 to 4 g/d) compared with nonepileptic subjects. Serial measurements suggested that brain GABA rose in proportion to vigabatrin dose up to 3 g/d. Doubling the dose from 3 to 6 g/d failed to increase brain GABA further. Serial measurements on three patients taking 6 g/d showed a gradual decrease in brain GABA in two patients over 1 to 2 years of treatment. These observations suggest that GABA synthesis may decrease at high GABA levels.

1H NMR studies of glucose transport in the human brain

The difference between 1H nuclear magnetic resonance (NMR) spectra obtained from the human brain during euglycemia and during hyperglycemia is depicted as well-resolved glucose peaks. The time course of these brain glucose changes during a rapid increase in plasma glucose was measured in four healthy subjects, aged 18-22 years, in five studies. Results demonstrated a significant lag in the rise of glucose with respect to plasma glucose. The fit of the integrated symmetric Michaelis-Menten model to the time course of relative glucose signals yielded an estimated plasma glucose concentration for half maximal transport, Kt, of 4.8 +/- 2.4 mM (mean +/- SD), a maximal transport rate, Tmax, of 0.80 +/- 0.45 micromol g-1 min-1, and a cerebral metabolic glucose consumption rate (CMR)glc of 0.32 +/- 0.16 micromol g-1 min-1. Assuming cerebral glucose concentration to be 1.0 micromol/g at euglycemia as measured by 13CMR, the fit of the same model to the time course of brain glucose concentrations resulted in Kt = 3.9 +/- 0.82 mM, Tmax = 1.16 +/- 0.29 micromol g-1 min-1, and CMRglc = 0.35 +/- 0.10 micromol g-1 min-1. In both cases, the resulting time course equaled that predicted from the determination of the steady-state glucose concentration by 13C NMR spectroscopy within the experimental scatter. The agreement between the two methods of determining transport kinetics suggests that glucose is distributed throughout the entire aqueous phase of the human brain, implying substantial intracellular concentration.

Pediatric orofacial injuries

Perioral injuries may have significant medical, dental and psychological consequences in children. Soft-tissue injuries are common-place, with the most common types being contusions, abrasions, lacerations and electrical and chemical burns. Each type requires specific care and follow through. Immediate intervention and treatment are important, and rapid neurologic assessment of a child before treatment helps the long-term prognosis. Children must also be screened for abuse and neglect. The common goal of treatment is to prevent infection, provide function and minimize scarring. Perioral injuries are preventable.

The roles of insulin and glucagon in the regulation of hepatic glycogen synthesis and turnover in humans

To determine the respective roles of insulin and glucagon for hepatic glycogen synthesis and turnover, hyperglycemic clamps were performed with somatostatin [0.1 micrograms/(kg.min)] in healthy young men under conditions of: (I) basal fasting) portal vein insulinemia-hypoglucagonemia, (II) basal portal vein insulinemia-basal glucagonemia, and (III) basal peripheral insulinemia-hypoglucagonemia. Synthetic rates, pathway (direct versus indirect) contributions, and percent turnover of hepatic glycogen were assessed by in vivo 13C nuclear magnetic resonance spectroscopy during [1-13C]glucose infusion followed by a natural abundance glucose chase in conjunction with acetaminophen to noninvasively sample the hepatic UDP-glucose pool. In the presence of hyperglycemia (10.4 +/- 0.1 mM) and basal portal vein insulinemia (192 +/- 6 pM), suppression of glucagon secretion (plasma glucagon, I:31 +/- 4, II: 63 +/- 8 pg/ml) doubled the hepatic accumulation of glycogen (Vsyn) compared with conditions of basal glucagonemia [I: 0.40 +/- 0.06, II: 0.19 +/- 0.03 mumol/(liter.min): P < 0.0025]. Glycogen turnover was markedly reduced (I: 19 +/- 7%, II: 69 +/- 12%; P < 0.005), so that net rate of glycogen synthesis increased approximately fivefold (P < 0.001) by inhibition of glucagon secretion. The relative contribution of gluconeogenesis (indirect pathway) to glycogen synthesis was lower during hypoglucagonemia (42 +/- 6%) than during basal glucagonemia (54 +/- 5%; P < 0.005). Under conditions of basal peripheral insulinemia (54 +/- 2 pM) and hypoglucagonemia (III) there was negligible hepatic glycogen synthesis and turnover. In conclusion, small changes in portal vein concentrations of insulin and glucagon independently affect hepatic glycogen synthesis and turnover. Inhibition of glucagon secretion under conditions of hyperglycemia and basal concentrations of insulin results in: (a) twofold increase in rate of hepatic glycogen synthesis, (b) reduction of glycogen turnover by approximately 73%, and (c) augmented percent contribution of the direct pathway to glycogen synthesis compared with conditions of basal glucagonemia.

Nuclear magnetic resonance studies of muscle and applications to exercise and diabetes

Natural-abundance 13C nuclear magnetic resonance (NMR) spectroscopy is a noninvasive technique that enables in vivo assessments of muscle and/or liver glycogen concentrations. When directly compared with the traditional needle biopsy technique, NMR was found to be more precise. Over the last several years, we have developed and used 13C-NMR to obtain information about human glycogen metabolism both under conditions of altered blood glucose and/or insulin and with exercise. Because NMR is noninvasive, we have been able to obtain more data points over a specified time course, thereby dramatically improving the time resolution. This improved time resolution has enabled us to document subtleties of the resynthesis of muscle glycogen after severe exercise that have not been observed previously. An added advantage of NMR is that we are able to obtain information simultaneously about other nuclei, such as 31P. With interleaved 13C- and 31P-NMR techniques, we have been able to follow simultaneous changes in muscle glucose-6-phosphate and muscle glycogen. In this article, we review some of the work that has been reported by our laboratory and discuss the relevance of our findings for the management of diabetes.

The effect of gabapentin on brain gamma-aminobutyric acid in patients with epilepsy

Gabapentin has come into clinical use as adjunctive therapy in the treatment of epilepsy. Designed to mimic gamma-aminobutyric acid (GABA), its mechanism of action remains elusive. In vivo measurements of GABA in human brain were made using 1H magnetic resonance spectroscopy. We used a 2.1-T magnetic resonance imager-spectrometer and an 8-cm surface coil to measure a 13.5-cm3 volume in the occipital cortex. GABA levels were measured in 14 patients enrolled in an open-lbel trial of gabapentin. GABA was elevated in patients taking gabapentin compared with 14 complex partial epilepsy patients, matched for antiepileptic drug treatment. Brain GABA levels appeared to be higher in patients taking high-dose gabapentin (3,300-3,600 mg/day) than in those taking standard doses (1,200-2,400 mg/day). Gabapentin appears to increase human brain GABA levels.

Contribution of net hepatic glycogenolysis to glucose production during the early postprandial period

Relative contributions of net hepatic glycogenolysis and gluconeogenesis to glucose production during the first 12 h of a fast were studied in 13 healthy volunteers by noninvasively measuring hepatic glycogen content using 13C nuclear magnetic resonance spectroscopy. Rates of net hepatic glycogenolysis were calculated by multiplying the change in liver glycogen content with liver volume determined by magnetic resonance imaging. Rates of gluconeogenesis were calculated as the difference between rates of glucose production determined with an infusion of [6,6-2H]-glucose and net hepatic glycogenolysis. At 6 P.M. a liquid mixed meal (1,000 kcal; 60% as glucose) was given, to which [2-2H]glucose was added to trace glucose absorption. Hepatic glycogen content was measured between 11 P.M. and 1 A.M. and between 3 and 6 A.M. At 11 P.M. the concentration was 470 mM and it decreased linearly during the night. The mean liver volume was 1.47 +/- 0.06 liters. Net hepatic glycogenolysis (5.8 +/- 0.8 mumol.kg body wt-1.min-1) accounted for, on average, 45 +/- 6% and gluconeogenesis for 55 +/- 6% of the rate of whole body glucose production (12.6 +/- 0.6 mumol.kg body wt-1.min-1). In conclusion, this study shows that, even early in the phase of the postabsorptive period when liver glycogen stores are maximal, gluconeogenesis contributes approximately 50% to hepatic glucose production.

Direct assessment of liver glycogen storage by 13C nuclear magnetic resonance spectroscopy and regulation of glucose homeostasis after a mixed meal in normal subjects

Despite extensive recent studies, understanding of the normal postprandial processes underlying immediate storage of substrate and maintenance of glucose homeostasis in humans after a mixed meal has been incomplete. The present study applied 13C nuclear magnetic resonance spectroscopy to measure sequential changes in hepatic glycogen concentration, a novel tracer approach to measure postprandial suppression of hepatic glucose output, and acetaminophen to trace the pathways of hepatic glycogen synthesis to elucidate the homeostatic adaptation to the fed state in healthy human subjects. After the liquid mixed meal, liver glycogen concentration rose from 207 +/- 22 to 316 +/- 19 mmol/liter at an average rate of 0.34 mmol/liter per min and peaked at 318 +/- 31 min, falling rapidly thereafter (0.26 mmol/liter per min). The mean increment at peak represented net glycogen synthesis of 28.3 +/- 3.7 g (approximately 19% of meal carbohydrate content). The contribution of the direct pathway to overall glycogen synthesis was 46 +/- 5 and 68 +/- 8% between 2 and 4 and 4 and 6 h, respectively. Hepatic glucose output was completely suppressed within 30 min of the meal. It increased steadily from 60 to 255 min from 0.31 +/- 32 to 0.49 +/- 18 mg/kg per min then rapidly returned towards basal levels (1.90 +/- 0.04 mg/kg per min). This pattern of change mirrored precisely the plasma glucagon/insulin ratio. These data provide for the first time a comprehensive picture of normal carbohydrate metabolism in humans after ingestion of a mixed meal.

In vivo regulation of muscle glycogen synthase and the control of glycogen synthesis

The activity of glycogen synthase (GSase; EC 2.4.1.11) is regulated by covalent phosphorylation. Because of this regulation, GSase has generally been considered to control the rate of glycogen synthesis. This hypothesis is examined in light of recent in vivo NMR experiments on rat and human muscle and is found to be quantitatively inconsistent with the data under conditions of glycogen synthesis. Our first experiments showed that muscle glycogen synthesis was slower in non-insulin-dependent diabetics compared to normals and that their defect was in the glucose transporter/hexokinase (GT/HK) part of the pathway. From these and other in vivo NMR results a quantitative model is proposed in which the GT/HK steps control the rate of glycogen synthesis in normal humans and rat muscle. The flux through GSase is regulated to match the proximal steps by "feed forward" to glucose 6-phosphate, which is a positive allosteric effector of all forms of GSase. Recent in vivo NMR experiments specifically designed to test the model are analyzed by metabolic control theory and it is shown quantitatively that the GT/HK step controls the rate of glycogen synthesis. Preliminary evidence favors the transporter step. Several conclusions are significant: (i) glucose transport/hexokinase controls the glycogen synthesis flux; (ii) the role of covalent phosphorylation of GSase is to adapt the activity of the enzyme to the flux and to control the metabolite levels not the flux; (iii) the quantitative data needed for inferring and testing the present model of flux control depended upon advances of in vivo NMR methods that accurately measured the concentration of glucose 6-phosphate and the rate of glycogen synthesis.

Initial observations on effect of vigabatrin on in vivo 1H spectroscopic measurements of gamma-aminobutyric acid, glutamate, and glutamine in human brain

Recent developments involving 1H nuclear magnetic resonance (NMR) spectroscopic editing techniques have allowed noninvasive measurements of gamma-aminobutyric acid (GABA) in human cerebrum. The additional information gained from GABA and macromolecule measurements permitted more precise glutamate (Glu) and glutamine (Gln) measurements. Occipital lobe GABA in 10 nonepileptic, healthy subjects was 1.0 mumol/g brain [95% confidence interval (CI) 0.9-1.1]. Vigabatrin (VGB) is a safe and effective antiepileptic drug (AED) that irreversibly inhibits neuronal and glial GABA-transaminase. GABA levels were increased in all patients treated with VGB. With a standard dose of 3-6 g/day, GABA levels were 2.6 mumol/g (95% CI 2.3-2.8). Mean occipital GABA level measured in epileptic patients not receiving VGB was 0.9 mumol/g (95% CI 0.7-1.1). Gln was increased by 1.9 mumol/g and Glu was decreased by 0.8 mumol/g in patients receiving VGB as compared with patients receiving standard medications alone.

Decreased muscle glucose transport/phosphorylation is an early defect in the pathogenesis of non-insulin-dependent diabetes mellitus

Recent studies have demonstrated that reduced insulin-stimulated muscle glycogen synthesis is the major cause of insulin resistance in patients with non-insulin-dependent diabetes mellitus (NIDDM). This reduced rate has been assigned to a defect in either glucose transport or hexokinase activity. However it is unknown whether this is a primary or acquired defect in the pathogenesis of NIDDM. To examine this question, we measured the rate of muscle glycogen synthesis and the muscle glucose 6-phosphate (G6P) concentration using 13C and 31P NMR spectroscopy as well as oxidative and nonoxidative glucose metabolism in six lean, normoglycemic offspring of parents with NIDDM and seven age/weight-matched control subjects under hyperglycemic (approximately 11 mM)-hyperinsulinemic (approximately 480 pM) clamp conditions. The offspring of parents with NIDDM had a 50% reduction in total glucose metabolism, primarily due to a decrease in the nonoxidative component. The rate of muscle glycogen synthesis was reduced by 70% (P < 0.005) and muscle G6P concentration was reduced by 40% (P < 0.003), which suggests impaired muscle glucose transport/hexokinase activity. These changes were similar to those previously observed in subjects with fully developed NIDDM. When the control subjects were studied at similar insulin levels (approximately 440 pM) but euglycemic plasma glucose concentration (approximately 5 mM), both the rate of glycogen synthesis and the G6P concentration were reduced to values similar to the offspring of parents with NIDDM. We conclude that insulin-resistant offspring of parents with NIDDM have reduced nonoxidative glucose metabolism and muscle glycogen synthesis secondary to a defect in muscle glucose transport/hexokinase activity prior to the onset of overt hyperglycemia. The presence of this defect in these subjects suggests that it may be the primary factor in the pathogenesis of NIDDM.

Contribution of hepatic glycogenolysis to glucose production in humans in response to a physiological increase in plasma glucagon concentration

The contribution of net hepatic glycogenolysis to overall glucose production during a physiological increment in the plasma glucagon concentration was measured in six healthy subjects (18-24 years, 68-105 kg) after an overnight fast. Glucagon (approximately 3 ng.kg-1.min-1), somatostatin (0.1 microgram.kg-1.min-1), and insulin (0.9 pmol.kg-1.min-1) were infused for 3 h. Liver glycogen concentration was measured at 15-min intervals during this period using 13C-labeled nuclear magnetic resonance spectroscopy, and liver volume was assessed from magnetic resonance images. The rate of net hepatic glycogenolysis was calculated from the decrease in liver glycogen concentration over time, multiplied by the liver volume. The rate of glucose appearance (Ra) was calculated from [3-3H]glucose turnover data using a two-compartment model of glucose kinetics. Plasma glucagon concentration rose from 136 +/- 18 to 304 +/- 57 ng/l and plasma glucose concentration rose from 5.6 +/- 0.1 to 10.4 +/- 0.9 mmol/l on initiation of the infusions. Mean baseline Ra was 11.8 +/- 0.4 mumol.kg-1.min-1, increased rapidly after the beginning of the infusions, reaching its highest value after 20-40 min, and returned to baseline by 140 min. Liver glycogen concentration decreased almost linearly (from 300 +/- 19 mmol/l liver at baseline to 192 +/- 20 mmol/l liver at t = 124 min) during 2 h after the beginning of the infusions, and the calculated mean rate of net hepatic glycogenolysis was 21.7 +/- 3.6 mumol.kg-1.min-1. Mean Ra during the same time period was 22.8 +/- 2.3 mumol.kg-1.min-1. Thus, net hepatic glycogenolysis accounted for 93 +/- 9% of Ra.(ABSTRACT TRUNCATED AT 250 WORDS)

Impaired net hepatic glycogen synthesis in insulin-dependent diabetic subjects during mixed meal ingestion. A 13C nuclear magnetic resonance spectroscopy study

Hepatic glycogen concentration was measured in six subjects with insulin-dependent diabetes mellitus (IDDM) and nine weight-matched control subjects using 13C nuclear magnetic resonance spectroscopy during a day in which three isocaloric mixed meals were ingested. The relative fluxes of the direct and indirect (3 carbon units-->-->glycogen) pathways of hepatic glycogen synthesis were also assessed using [1-13C]glucose in combination with acetaminophen to noninvasively sample the hepatic UDP-glucose pool. Mean fasting hepatic glycogen content was similar in the two groups. After each meal, hepatic glycogen content increased, peaking 4-5 h after the meal in both groups. By 11:00 p.m. the IDDM subjects had synthesized only 30% of the glycogen that was synthesized by the control group [IDDM subjects, net increment = 44 +/- 20 (mean +/- SE) mM; control subjects, net increment = 144 +/- 14 mM; P < 0.05]. After breakfast the flux through the gluconeogenic pathway relative to the direct pathway of hepatic glycogen synthesis was 1.7-fold greater in the IDDM subjects (59 +/- 4%) than in the control subjects (35 +/- 4%, P < 0.0003). In conclusion, under mixed meal conditions, subjects with poorly controlled IDDM have a major defect in net hepatic glycogen synthesis and augmented hepatic gluconeogenesis. The former abnormality may result in an impaired glycemic response to counterregulatory hormones, whereas both abnormalities may contribute to postprandial hyperglycemia.

In vivo measurement of phenylalanine in human brain by proton nuclear magnetic resonance spectroscopy

Disorders of the CNS are the major causes of morbidity and mortality observed in untreated subjects with phenylketonuria (PKU). A method to measure cerebral concentrations of phenylalanine (Phe) in vivo would greatly enhance the ability to investigate both the pathophysiology and the efficacy of therapy of this aminoacidopathy. Twelve image-guided localized proton nuclear magnetic resonance spectroscopic studies were performed in seven subjects with PKU using pulse sequences optimized to detect the aromatic protons of Phe. Ten control studies were also performed using a 2.1-Tesla Bruker Biospec spectrometer. Plasma Phe was measured at the time of the spectroscopic examination in the PKU patients. A Phe signal was observed in all 12 studies performed on the group with PKU, and in five studies cerebral Phe concentrations were measured to be 480 to 780 mumol/g. Plasma Phe concentrations were 0.7 to 3.3 mM (10.8 to 54.8 mg/dL) in the subjects with PKU. Human cerebral Phe concentrations can be measured noninvasively using proton nuclear magnetic resonance spectroscopy. A simultaneous measure of Phe and several other cerebral metabolites is obtained with this innovative technology. Adaptations of this technique can be used to investigate PKU and other neurometabolic disorders with modifications of current clinical magnetic resonance imaging systems.

Simultaneous determination of the rates of the TCA cycle, glucose utilization, alpha-ketoglutarate/glutamate exchange, and glutamine synthesis in human brain by NMR

13C isotopic tracer data previously obtained by 13C nuclear magnetic resonance in the human brain in vivo were analyzed using a mathematical model to determine metabolic rates in a region of the human neocortex. The tricarboxylic acid (TCA) cycle rate was 0.73 +/- 0.19 mumol min-1 g-1 (mean +/- SD; n = 4). The standard deviation reflects primarily intersubject variation, since individual uncertainties were low. The rate of alpha-ketoglutarate/glutamate exchange was 57 +/- 26 mumol min-1 g-1 (n = 3), which is much greater than the TCA cycle rate; the high rate indicates that alpha-ketoglutarate and glutamate are in rapid exchange and can be treated as a single combined kinetic pool. The rate of synthesis of glutamine from glutamate was 0.47 mumol min-1 g-1 (n = 4), with 95% confidence limits of 0.139 and 3.094 mumol min-1 g-1; individual uncertainties were biased heavily toward high synthesis rates. From the TCA cycle rate the brain oxygen consumption was estimated to be 2.14 +/- 0.48 mumol min-1 g-1 (5.07 +/- 1.14 ml 100 g-1 min-1; n = 4), and the rate of brain glucose consumption was calculated to be 0.37 +/- 0.08 mumol min-1 g-1 (n = 4). The sensitivity of the model to the assumptions made was evaluated, and the calculated values were found to be unchanged as long as the assumptions remained near reported physiological values.

Image reconstruction of sequentially sampled echo-planar data

For echo-planar imaging (EPI), failure to time-reverse alternate echoes results in aliasing in the image. We encountered image artifacts in EPI acquired on a system with sequential sampling. After examining the source of these image artifacts, we concluded that the artifacts were a result of the type of sampling method used in data acquisition and the way the time-reversal of alternate echoes was carried out prior to Fourier transformation. Two methods are demonstrated to obtain artifact-free EPI with sequential data sampling.

13C-nuclear magnetic resonance spectroscopy studies of hepatic glucose metabolism in normal subjects and subjects with insulin-dependent diabetes mellitus

To determine the effect of insulin-dependent diabetes mellitus (IDDM) on rates and pathways of hepatic glycogen synthesis, as well as flux through hepatic pyruvate dehydrogenase, we used 13C-nuclear magnetic resonance spectroscopy to monitor the peak intensity of the C1 resonance of the glucosyl units of hepatic glycogen, in combination with acetaminophen to sample the hepatic UDP-glucose pool and phenylacetate to sample the hepatic glutamine pool, during a hyperglycemic-hyperinsulinemic clamp using [1-13C]-glucose. Five subjects with poorly controlled IDDM and six age-weight-matched control subjects were clamped at a mean plasma glucose concentration of approximately 9 mM and mean plasma insulin concentrations approximately 400 pM for 5 h. Rates of hepatic glycogen synthesis were similar in both groups (approximately 0.43 +/- 0.09 mumol/ml liver min). However, flux through the indirect pathway of glycogen synthesis (3 carbon units-->-->glycogen) was increased by approximately 50% (P < 0.05), whereas the relative contribution of pyruvate oxidation to TCA cycle flux was decreased by approximately 30% (P < 0.05) in the IDDM subjects compared to the control subjects. These studies demonstrate that patients with poorly controlled insulin-dependent diabetes mellitus have augmented hepatic gluconeogenesis and relative decreased rates of hepatic pyruvate oxidation. These abnormalities are not immediately reversed by normalizing intraportal concentrations of glucose, insulin, and glucagon and may contribute to postprandial hyperglycemia.

NMR studies of human brain function

Recent advances in magnetic resonance imaging and spectroscopy make it possible to measure localized changes in human brain activity and metabolism in single subjects during sensory stimulation and cognition. Differences between stimulated and unstimulated subjects can be visualized to a resolution of mm3 in less than 1s, a significant improvement over the more established method, positron emission tomography. Magnetic resonance spectroscopy of the human brain, measuring fluxes in several cm3, has followed changes in metabolic rates during visual stimulation.

Localized 13C NMR spectroscopy in the human brain of amino acid labeling from D-[1-13C]glucose

Cerebral metabolism of D[1-13C]glucose was studied with localized 13C NMR spectroscopy during intravenous infusion of enriched [1-13C]glucose in four healthy subjects. The use of three-dimensional localization resulted in the complete elimination of triacylglycerol resonance that originated in scalp and subcutaneous fat. The sensitivity and resolution were sufficient to allow 4 min of time-resolved observation of label incorporation into the C3 and C4 resonances of glutamate and C4 of glutamine, as well as C3 of aspartate with lower time resolution. [4-13C]Glutamate labeled rapidly reaching close to maximum labeling at 60 min. The label flow into [3-13C]glutamate clearly lagged behind that of [4-13C]-glutamate and peaked at t = 110-140 min. Multiplets due to homonuclear 13C-13C coupling between the C3 and C4 peaks of the glutamate molecule were observed in vivo. Isotopomer analysis of spectra acquired between 120 and 180 min yielded a 13C isotopic fraction at C4 glutamate of 27 +/- 2% (n = 4), which was slightly less than one-half the enrichment of the C1 position of plasma glucose (63 +/- 1%), p < 0.05. By comparison with an external standard the total amount of [4-13C]glutamate was directly quantified to be 2.4 +/- 0.1 mumol/ml-brain. Together with the isotopomer data this gave a calculated brain glutamate concentration of 9.1 +/- 0.7 mumol/ml, which agrees with previous estimates of total brain glutamate concentrations. The agreement suggests that essentially all of the brain glutamate is derived from glucose in health human brain.

Analysis of macromolecule resonances in 1H NMR spectra of human brain

Macromolecule resonances underlying metabolites in 1H NMR spectra were investigated in temporal lobe biopsy tissue from epilepsy patients and from localized 1H spectra of the brains of healthy volunteers. The 1H NMR spectrum of brain tissue was compared with that of cytosol and dialyzed cytosol after removal of low molecular weight molecules (< 3500 daltons) at 8.4 and 2.1 Tesla. The assignment of specific resonances to macromolecules in 2.1 Tesla, short-TE, localized human brain 1H NMR spectra in vivo was made on the basis of a J-editing method using the spectral parameters (delta, J) and connectivities determined from 2D experiments in vitro. Two prominent connectivities associated with macromolecules in vitro (0.93-2.05 delta and 1.6-3.00 delta) were also detected in vivo by the J-editing method. Advantage was taken of the large difference in measured T1 relaxation times between macromolecule and metabolite resonances in the brain spectrum to acquire 'metabolite-nulled' macromolecule spectra. These spectra appear identical to the spectra of macromolecules isolated in vitro.