help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2004-1092
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bertoldo, A.
Right arrow Articles by Kelley, D. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bertoldo, A.
Right arrow Articles by Kelley, D. E.
Related Collections
Right arrow Diabetes and Insulin
Right arrow Metabolism
The Journal of Clinical Endocrinology & Metabolism Vol. 90, No. 3 1752-1759
Copyright © 2005 by The Endocrine Society

Quantitative Assessment of Glucose Transport in Human Skeletal Muscle: Dynamic Positron Emission Tomography Imaging of [O-Methyl-11C]3-O-Methyl-D-Glucose

Alessandra Bertoldo, Julie Price, Chet Mathis, Scott Mason, Daniel Holt, Carol Kelley, Claudio Cobelli and David E. Kelley

Departments of Medicine (C.K., D.E.K.) and Radiology (J.P., C.M., S.M., D.H.), University of Pittsburgh, Pittsburgh, Pennsylvania 15213; and Department of Information Engineering, University of Padova (A.B., C.C.), Padova 35121, Italy

Address all correspondence and requests for reprints to: Dr. David E. Kelley, 807N Montefiore-University Hospital, 3459 Fifth Avenue, Pittsburgh, Pennsylvania 15213. E-mail: kelley{at}msx.dept-med.pitt.edu.


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
Insulin-stimulated glucose transport in skeletal muscle is regarded as a key determinant of insulin sensitivity, yet isolation of this step for quantification in human studies is a methodological challenge. One notable approach is physiological modeling of dynamic positron emission tomography (PET) imaging using 2-[18-fluoro]2-deoxyglucose ([18F]FDG); however, this has a potential limitation in that deoxyglucose undergoes phosphorylation subsequent to transport, complicating separate estimations of these steps. In the current study we explored the use of dynamic PET imaging of [11C]3-O-methylglucose ([11C]3-OMG), a glucose analog that is limited to bidirectional glucose transport. Seventeen lean healthy volunteers with normal insulin sensitivity participated; eight had imaging during basal conditions, and nine had imaging during euglycemic insulin infusion at 30 mU/min·m2. Dynamic PET imaging of calf muscles was conducted for 90 min after the injection of [11C]3-OMG. Spectral analysis of tissue activity indicated that a model configuration of two reversible compartments gave the strongest statistical fit to the kinetic pattern. Accordingly, and consistent with the structure of a model previously used for [18F]FDG, a two-compartment model was applied. Consistent with prior [18F]FDG findings, insulin was found to have minimal effect on the rate constant for movement of [11C]3-OMG from plasma to tissue interstitium. However, during insulin infusion, a robust and highly significant increase was observed in the kinetics of inward glucose transport; this and the estimated tissue distribution volume for [11C]3-OMG increased 6-fold compared with basal conditions. We conclude that dynamic PET imaging of [11C]3-OMG offers a novel quantitative approach that is both chemically specific and tissue specific for in vivo assessment of glucose transport in human skeletal muscle.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
DURING THE PAST two decades, several groups of investigators have used positron emission tomography (PET) for bioimaging skeletal muscle metabolism in diabetes and related conditions of insulin resistance (IR) (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). PET has the advantages of noninvasive, tissue-specific imaging of metabolism and the potential, based upon the tracer used, for isolating specific metabolic steps. A key example is that the tracer most commonly used in studies of muscle, 2-[18-fluoro]2-deoxyglucose ([18F]FDG), is restricted in its metabolism to glucose transport and phosphorylation (12). Translocation of glucose transporter proteins to the sarcolemma and within T-tubules is commonly regarded as a rate-controlling step in insulin-regulated glucose metabolism and pivotal in the pathogenesis of IR (13). Estimation of the kinetics of glucose transport as well as examination of the kinetics of glucose phosphorylation from dynamic PET imaging of [18F]FDG implicate interactive impairments of glucose transport and glucose phosphorylation in type 2 diabetes mellitus (DM) and obesity (1, 11). Yet, key methodological issues concern model configuration and the ability to segregate estimation of the kinetics of glucose transport from those of glucose phosphorylation.

A compartmental model developed, and later modified, for PET imaging of the central nervous system (CNS) (14, 15) was initially used for PET data acquired in skeletal muscle (1). More recently, Bertoldo et al. (16) proposed a muscle-specific compartmental model, a modification that takes into account movement of [18F]FDG from plasma to interstitial space and from interstitial space into tissue via transmembrane glucose transport, as well as an irreversible compartment for formation of 2-[18-fluoro]2-deoxyglucose-6-phosphate ([18F]FDG-6-P). Application of the Bertoldo et al. model to dynamic PET imaging in obesity and type 2 DM indicated marked IR of glucose transport as well as impaired glucose phosphorylation, contrasting with clear stimulation of these parameters in healthy volunteers (17, 18). Nevertheless, the fact that emission from 18F can occur at either [18F]FDG or [18F]FDG-6-P, creates implicit uncertainty as to whether compartmental modeling achieves separate estimations of glucose transport and phosphorylation.

The current study was undertaken to bolster the capacity of dynamic PET imaging to isolate the step of transmembrane glucose transport in skeletal muscle based on chemical specificity of 3-O-methylglucose (3-OMG) labeled with 11C, [O-methyl-11C]3-O-methyl-D-glucose. [11C]3-OMG enters the free glucose pool in tissue, is not further metabolized, and can be bidirectionally transported (19). An impressive number of in vitro and in vivo experiments have shown that 3-OMG is an almost ideal glucose analog to probe transmembrane transport, sharing the same transport system as and with equivalent affinity for glucose transport proteins (20, 21, 22). There have been prior in vivo PET investigations in humans using [11C]3-OMG to study glucose transport across the blood-brain barrier (23, 24, 25, 26) and an even smaller number of animal studies with this tracer in liver (27, 28). However, to our knowledge, the current study is the first to use [11C]3-OMG and dynamic PET imaging to study basal and insulin-stimulated glucose transport in human skeletal muscle.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Research volunteers

Lean, healthy, young adults with normal glucose tolerance were recruited for this investigation, and their clinical characteristics are shown in Table 1Go. Eight volunteers were studied after an overnight fast and without insulin infusion (basal), and nine volunteers were studied after a similar overnight fast, but during insulin-stimulated conditions (clamp). The two groups of volunteers were closely matched for age, gender, and body mass index and had normal fasting values for glucose, insulin, hemoglobin A1c (5.1 ± 0.1% vs. 5.1 ± 0.1%), high-density lipoprotein cholesterol (62 ± 6 vs. 65 ± 5 mg/dl), and triglyceride (76 ± 11 vs. 89 ± 13 mg/dl), indicative of normal insulin sensitivity, and as also indicated by similar and low values for homeostasis model assessment of IR. Before the study, each participant had medical and laboratory examinations to verify good health. Informed, written consent was obtained, and the University of Pittsburgh institutional review board reviewed and approved this clinical investigation.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Clinical characteristics of research volunteers

 
Hyperinsulinemic-euglycemic clamp. All research volunteers were admitted to University of Pittsburgh General Clinical Research Center on the evening before studies and received a dinner of a standard composition (7 kcal/kg; 50% carbohydrate, 30% fat, and 20% protein), then fasted overnight. Volunteers had been instructed to refrain from exercise on the day of admission. On the following morning, an iv catheter was placed in an antecubital vein for infusion of saline (basal group) or insulin and glucose (insulin group) and for later injection of [11C]3-OMG. To obtain the arterial samples for determination of [11C]3-OMG in plasma, a catheter was placed in a radial artery. Samples were obtained for determinations of insulin and fatty acids at baseline and hourly thereafter. In those randomized to receive insulin, a primed, continuous infusion (30 mU/min·m2 body surface area) was begun, and arterial glucose was measured at 5-min intervals during insulin infusions to adjust rates of infusion of 20% dextrose so as to maintain euglycemia using the glucose clamp method (29). Plasma glucose was measured using a glucose analyzer (YSI, Inc., Yellow Springs, OH); plasma insulin was measured by RIA, and plasma fatty acids were measured using a colorimetric enzymatic assay (Wako NEFA C test kit, Wako Chemicals, Richmond, VA). Euglycemic insulin infusion was maintained for 1 h before the start of dynamic PET imaging and was continued throughout PET imaging.

Tracer synthesis

The radiosynthesis of [11C]3-OMG was based upon methods previously reported (30) and yielded 1.5–2.5 GBq (40–67 mCi) final product for injection (after a 40-min synthesis) possessing chemical and radiochemical purities of 90% or greater and a specific activity of 7.4 GBq/µmol (≥0.2 Ci/µmol) or more at the time of injection. Briefly, diacetone-D-glucose (Sigma Aldrich Corp., St. Louis, MO; 10 mg) was dissolved in acetonitrile (0.5 ml). High specific activity [11C]methyl-iodide (31) was produced using a remote radiosynthesis system and was delivered to the diacetone-D-glucose reaction solution in a nitrogen gas stream, where it was collected at room temperature. The reaction mixture was heated at 75 C for 5 min, followed by acetonitrile evaporation at 120 C under a stream of nitrogen gas. Two milliliter of 1 N HCl was added to the residue, and the reaction mixture was heated for an additional 10 min at 120 C. The crude reaction product was purified by passage through an anion exchange column (AG11A8, Bio-Rad Laboratories, Hercules, CA) with sterile water, and the solution containing [11C]3-OMG was filtered through a 0.22-µm pore size sterile filter (Millipore Corp., Bedford, MA) into a sterile vial. Quality control of the final product included pH determination, analytical HPLC [Phenomenex NH2 (Phenomenex, Torrance, CA), 5 µm, 250 x 2-mm column eluted with 90:10 (vol/vol) acetonitrile/water; 0.5 ml flow rate; k' = 7] to determine chemical and radiochemical purities and specific activity, and tests to establish apyrogenicity and sterility of the product. An authentic chromatographic standard of 3-OMG (Sigma Aldrich) was used to confirm the identity of the 11C-labeled analog using both normal and reverse phase HPLC methods.

Radiolabeled metabolite assay

For a subset of subjects (n = 4), HPLC methods were used to determine the extent to which [11C]3-OMG was metabolized in vivo over time. Blood samples were collected for each subject at three times after radiotracer injection: 2, 10, and 30 min. The assay for radiolabeled metabolites of [11C]3-OMG in plasma was performed using a Phenomenex Prodigy ODS (3) analytical column eluted with acetonitrile/water (90:10, vol/vol). In-line HPLC detectors include UV (model 486, Waters Corp., Medfield, MA) to calibrate the retention time of radiolabeled [11C]3-OMG with an authentic cold standard and radio-HPLC (model Gabi, Raytest Corp., New Castle, DE) to quantify the radiolabeled peaks.

PET image acquisition

PET imaging studies of [11C]3-OMG uptake into skeletal muscle were performed at University of Pittsburgh Positron Emission Tomography Center, using the CTI ECAT HR+ PET scanner in three-dimensional imaging mode (Siemens, New York, NY; 63 parallel planes; axial field of view, 15.2 cm; slice width, 2.4 mm). The scanner gantry is equipped with a Neuro-insert (CTI PET Systems, Knoxville, TN) to reduce the contribution of scattered photon events (32). PET data were reconstructed using filtered backprojection (Fourier rebinning and two-dimensional backprojection with Hann filter; kernel FWHM, 3 mm). To allow quantitative correction of attenuation, a 10-min windowed transmission scan was performed before emission scanning using rotating rods of 68Ge/68Ga. The PET data were corrected for radioactive decay and scatter (33). The final reconstructed PET image resolution was about 6 mm.

Initial pilot studies were performed to examine optimal tracer dose, using 3, 5, or 10 mCi. A dose of 5 mCi was selected for the studies presented in this paper. The injection of [11C]3-OMG was administered over 20 sec, and a 90-min dynamic PET scan was simultaneously initiated (36 frames: 8 x 15 sec, 8 x 15 sec, 4 x 1 min, 16 x 5 min). Blood for measuring the arterial [11C]3-OMG activity curve was obtained manually, drawing 0.5-ml samples from the radial artery catheter (10 samples every 6 sec, then eight samples every 15 sec, seven samples every 1 min, 10 samples every 5 min, and three samples every 10 min) over a total time of 90 min. Each blood sample was immediately centrifuged, and 200 µl plasma were removed for immediate 11C counting using a COBRA Auto-{gamma} model 5003 {gamma}-counter (Packard Instruments, Meriden, CT).

Volunteers were positioned in the PET scanner so that the midcalf corresponded to the midpoint axial field of view. To minimize movement during scanning and compression of muscle tissue, the legs were supported by foam blocking at the ankles and knees. Regions of interest (ROIs) were drawn in the muscle compartments of the calf region, carefully avoiding large blood vessels. The tissue-time-activity data of the ROIs were converted to units of radioactivity concentration (microcuries per milliliter) using an empiric phantom-based calibration factor (microcuries per milliliter/PET counts per pixel).

Modeling [11C]3-OMG kinetics

Spectral analysis (SA). The input-output modeling approach referred to in PET literature as SA (34) was the first analytical approach used, with the goal of estimating the number of compartments needed to describe [11C]3-OMG kinetics in skeletal muscle. Briefly, if the impulse response, h(t), of the system is written as:

with ßj ≥ 0, for every j, the total activity in the ROI, C(t), is the convolution of h(t) with the arterial plasma tracer concentration, Cp(t), plus a term taking into account the vascular component present in the ROI:

where Vb is the fraction of the total volume occupied by the blood pool, and Cb(t) is the arterial blood tracer concentration calculated as Cb(t) = Cp(t)(1 – 0.3 H) (35), with H the subject’s hematocrit. The method estimates the number, M, of nonzero values of {alpha}j that, together with the corresponding ßj, best describe the data. This provides information about the number of compartments needed to describe the tracer data even if it does not provide specificity on compartment configuration, as discussed below. The method starts by using M = 1 and estimates the values of {alpha}1, ß1, and Vb, then tries M = 2 and estimates {alpha}1, ß1, {alpha}2, ß2, and Vb, and so on. To select the best SA model, the Akaike information criterion (AIC) was used, as more fully described in Parameter estimation below.

Compartmental model. As presented in Results, SA results indicated that two reversible compartments provide the best statistical fit for [11C]3-OMG kinetics in skeletal muscle. A potential consideration for the two reversible compartment model configuration would be to consider two parallel compartments, such as might exist in heterogeneous tissue (36). However, because the aim of the current study was to use data on [11C]3-OMG kinetics in skeletal muscle to specifically test the previous findings on glucose transport obtained by the Bertoldo model, we elected to use the SA findings to support a model consisting of two extravascular compartments in series, as shown in Fig. 1Go. The model has four rate constants and can be described mathematically by the following set of equations:


where Cp is the [11C]3-OMG plasma arterial concentration, Ci is the extracellular concentration of [11C]3-OMG normalized to tissue volume, Ce is the [11C]3-OMG tissue concentration, C is the total 11C activity concentration in the ROI, K1 [ml/ml/min] and k1 [min–1] are the exchange between plasma and extracellular space, and k3 [min–1] and k4 [min–1] are the transport in and out of cell. All five model parameters k1, k2, k3, k4, and Vb, are a priori uniquely identifiable (37). Because Vb estimates were negligible, we ignored the Vb parameter. From the model one can calculate the volume of distribution of [11C]3-OMG in the intracellular space, Vic [ml/ml]:

and the volume of distribution in the extracellular space, Vec [ml/ml/min]:



View larger version (10K):
[in this window]
[in a new window]
  		FIG. 1. The configuration of the two-extravascular-compartmental model is shown. Cp, The tracer concentration in plasma; Ci, the extracellular (interstitial) compartment, with the kinetic parameters k1 and k2 representing inward and outward transport between plasma and Ci; Ce, the tissue compartment, with the kinetic parameters k3 and k4 representing inward and outward transport between Ci and Ce.

 
Parameter estimation. Both SA and compartmental model parameters were estimated by weighted nonlinear least squares, as implemented in SAAM II (38). The measured PET activity, Cobs, was described as:

where e(tj) is the measurement error at time tj assumed to be independent, Gaussian, zero mean and with a variance given by (39):

where {Delta}tj is the length of the scanning interval relative to Cobs(tj), and {gamma} is an unknown proportionality constant estimated a posteriori (37) as:

WRSS() is the weighted residual sum of squares evaluated at the minimum, i.e. for p equal to the estimated :

where wj is the weight of the j-th datum [wj = {Delta}tj/Cobs(tj)]), tj is the midpoint time of the j-th PET frame, N is the number of PET frames (observations), and P is the number of parameters. Parameter precision was evaluated from the inverse of the Fisher information matrix (37).

The best SA model configuration was selected using the AIC (37, 40). The AIC is based upon the principle of parsimony, such that the model of choice is identified as one that fits the data best with the fewest parameters. The AIC value was computed as:

The model of choice was associated with the smallest AIC value and thus was the most parsimonious.

Statistics

The mean ± SEM is reported. The significance of differences has been determined using the Mann-Whitney U test; P < 0.05 was considered significant.


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
Metabolic conditions during PET imaging

The metabolic conditions attained during basal (B) and insulin-stimulated clamp (C) studies differed by study design, with a 10-fold difference in insulin concentration (B, 46 ± 7; C, 448 ± 21 pmol/liter; P < 0.001), but similar steady- state plasma glucose (B, 4.9 ± 0.1; C, 5.1 ± 0.1 mmol/liter). Plasma free fatty acids were similar in the two groups after an overnight fast (573 ± 57 vs. 479 ± 22 µmol/ml; not significant). During B studies, plasma FFA remained at these levels, but were suppressed significantly during C studies (670 ± 46 vs. 83 ± 29 µmol/ml; P < 0.001). The mean steady- state rate of glucose infusion required to maintain euglycemia was 5.86 ± 0.50 mg/min·kg (range, 3.70–7.08 mg/min·kg) during C; glucose was not infused during B studies. Typical images obtained from a B and a C study are shown in Fig. 2Go, and representative tissue activity curves are plotted in Fig. 3Go along with corresponding arterial tracer activity.



View larger version (9K):
[in this window]
[in a new window]
  		FIG. 2. Shown is tissue activity of [11C]3-OMG in the lower legs under basal and steady-state insulin-stimulated conditions. The higher intensity (yellow) color of the insulin-stimulated images denotes higher tissue activity of [11C]3-OMG.

 


View larger version (12K):
[in this window]
[in a new window]
  		FIG. 3. Tissue activity curves obtained from dynamic PET images of [11C]3-OMG in the lower legs are shown in the left panels, and the corresponding arterial tracer activity of [11C]3-OMG is shown in the right panels. The upper figures represent tissue and plasma activity under basal conditions, and the lower panels represent data obtained under insulin-stimulated conditions. Each point within the tissue activity plots represents the activity during a time frame and is decay corrected; these points also represent the mean tissue activity across 25 planes of the lower legs during the respective time frame, as described in Subjects and Methods.

 
Radiolabeled results

The radio-HPLC analysis of plasma radioactivity was consistent with minimal metabolism of [11C]3-OMG over 30 min. Average values (n = 4) of the percentage of unmetabolized [11C]3-OMG were 98.4 ± 0.3%, 98.6 ± 0.5%, and 97.8 ± 1.2% at 2, 10, and 30 min, respectively.

SA

Tissue activity was initially analyzed using SA. One-, two-, and three-exponential models were examined. The three-exponential model was rejected as implausible, because the precision of parameter estimates was not acceptable (i.e. coefficient of variation, >100%). To assess by statistical criteria whether a one- and two-exponential model provided the more robust fit to the tissue activity curve, AIC values were calculated, with more negative values denoting a better fit. During B studies, AIC values were lower for a two-exponential model compared with a one-exponential model (–5.74 ± 0.11 vs. –4.12 ± 0.11; P < 0.01). The same pattern and virtually the same AIC values were obtained when comparing the two- vs. the one-exponential model during C studies (–5.43 ± 0.27 vs. –4.33 ± 0.29; P < 0.01). By these criteria, the two-exponential SA model was identified as the most appropriate to describe the impulse response of [11C]3-OMG tissue activity during both basal and insulin-stimulated conditions. The impulse response function reflects the response of muscle to the radiotracer input. The model of choice has two ß terms. The first ß is the fastest and most likely reflects the exchange of radiotracer between plasma and interstitial space. The second ß value is slower and is mainly associated with movement of tracer between interstitial space and tissue.

Compartmental modeling

As described above, the results from SA indicated that a model with two reversible compartments best describes [11C]3-OMG tissue activity. Therefore, for compartmental modeling, we employed a model comprised of two reversible, extravascular compartments existing in series (or cascade), which is an adaptation of the Bertoldo et al. model (16). The overall fit of the model to the tissue activity curves is shown in Fig. 4Go. Parameter estimates are shown in Table 2Go together with their precision. Rate constants for [11C]3-OMG could be resolved in all cases. Numerical identification was of good quality in all studies, except for a quite small value for k4 in subject B7. Under basal conditions, there was relatively minor intersubject variability in the kinetic parameters k1 through k4. During the insulin clamps, values for k1 did not change significantly compared with those under basal conditions. There were, however, statistically significant changes in parameters k2 through k4. The mean value for k2, representing outward transport from the interstitial compartment, was reduced by approximately 46% (P = 0.02). These patterns for k1 and k2 obtained from modeling of [11C]3-OMG are entirely consistent with prior studies using the Bertoldo et al. model for estimating the effects of insulin on these rate constants obtained from [18F]FDG kinetics in skeletal muscle (17), as will be discussed later. Values for k3, representing inward transport to the tissue compartment, increased 6-fold (P < 0.001), whereas those for k4, representing outward transport from the tissue compartment, increased during insulin-stimulated conditions by approximately 2-fold. The transport rate constant was correlated with rates of glucose infusion (r = 0.84; P < 0.001). The effect of insulin on the k3 parameter is consistent with effects previously reported in analysis of [18F]FDG kinetics in skeletal muscle (17). However, it should be noted that due to differences in the kinetics of methyl-glucose and deoxyglucose and also to the potential for phosphorylation of the latter, but not the former, direct comparison of the k3 and k4 parameters between these tracers is not germane.



View larger version (7K):
[in this window]
[in a new window]
  		FIG. 4. The curve of the compartmental model fit to tissue-activity curves, determined using the arterial tracer data (as the input function) and a two extravascular reversible compartment model, as depicted in Fig. 1Go, is shown for basal and insulin-stimulated conditions. As shown in this figure, the model achieved an excellent fit for both physiological conditions.

 

View this table:
[in this window]
[in a new window]
  		TABLE 2. Parameter estimates of [11C]3-OMG kinetics in skeletal muscle determined using the compartmental model comprised of two reversible, extravascular compartments

 
Based upon the values for the individual rate constants and the changes induced by insulin, the estimated extracellular distribution volume for [11C]3-OMG more than doubled during insulin-stimulated conditions (P < 0.01), as shown in Table 3Go. However, the estimated intracellular distribution volume of [11C]3-OMG, Vic, increased to an even greater extent, rising 6-fold (P < 0.001).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Estimates of skeletal muscle [11C]3-OMG volumes of distribution in the extracellular and intracellular compartments

 

   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
In the current study we report the findings from a novel method for tissue-specific and metabolic pathway-specific quantitative assessment of glucose transport in human skeletal muscle using dynamic PET imaging of [11C]3-OMG. PET imaging of skeletal muscle substrate metabolism and blood flow has emerged as a powerful bioimaging modality for the study of insulin action and insulin resistance (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). What is unique in the current study is applying this imaging method with the use of a positron labeled 3-OMG analog, metabolism of which is limited to bidirectional transport. PET imaging of muscle uptake of [11C]3-OMG was found to be of excellent technical quality and to reveal a very clear effect of insulin stimulation. In addition, quantitative methods to extract physiological information from the kinetic patterns of tissue activity of [11C]3-OMG were used to examine the effect of insulin.

The starting point for the quantitative approaches to be discussed, namely, SA and compartmental modeling, is acquisition of dynamic PET imaging. Dynamic PET imaging refers to the process of sequential, sustained PET imaging of a tissue (ROI) starting from the time of tracer injection and acquired in time frames of various and selected lengths, so that a complete and continuous tissue-time-activity curve can be constructed and analyzed in conjunction with the plasma-time-activity curve of the injected tracer. Selected examples of these tissue and arterial activity curves are shown in Fig. 3Go. Fractional extraction of glucose by skeletal muscle is quite low during fasting conditions, approximately 1–2% (41), and increases by approximately 10-fold during moderate insulin stimulation. We observed a quite similar pattern for [11C]3-OMG. This is an important point, because it bolsters a strong set of prior observations that 3-OMG is an excellent tracer for the study of glucose transport. Animal studies with 3-OMG and glucose have shown that these share the same transport system and with equivalent relative affinities for glucose transport proteins on the extracellular and intracellular sides of the plasma membrane (20, 21, 22). This point was also verified experimentally in humans by Bonadonna et al. (42), as venous washout curves after injections of [14C]3-OMG and glucose into the brachial artery were virtually superimposed during fasting and insulin-stimulated conditions, suggesting that in skeletal muscle, the kinetics of transport revealed by 3-OMG and glucose is indistinguishable.

Because metabolism of 3-OMG is restricted to the step of glucose transport, it is somewhat easy to conceptualize that any compartmental model that might describe the kinetics of the uptake of this tracer should prominently contain this biochemistry. However, our initial approach was SA, an approach that is model-independent, yet has great merit in laying out an objective beginning for construction of a compartmental model. SA identified the interaction of two reversible compartments during both basal and insulin-stimulated conditions, but did not indicate the presence of an irreversible compartment. With regard to this last point, plasma analysis revealed no significant metabolism of [11C]3-OMG.

Given that the biochemistry of 3-OMG is restricted to glucose transport, what physiological considerations can be given to delineate a compartmental model containing two reversible compartments? We would propose two considerations. One is that the two reversible compartments exist in parallel; by this, we mean to infer that the aggregate tissue activity of the tracer is the net interaction of the heterogeneous tissue(s). Although ROIs for capturing tissue activity were placed onto skeletal muscle, with care taken to exclude sc adipose tissue, bone, and major blood vessels, the presence of oxidative and glycolytic muscle fibers within muscle is one consideration in support of a model of two parallel reversible compartments pertaining to heterogeneity in the kinetics of glucose transport based on differences in glucose transport according to fiber type and other related characteristics (43). The current experiments do not provide sufficient data to more rigorously evaluate this issue. However, another compartmental model consistent with a configuration of two reversible compartments is that the compartments exist in series or cascade, rather than in parallel. The structure of this model is shown in Fig. 1Go.

The established approach to compartmental modeling of PET imaging of glucose metabolism in various tissues has been to use a model constructed with compartments in series. Sokoloff et al. (14) first developed a model for analysis of [18F]FDG in the CNS comprised of a reversible, followed by an irreversible, compartment, with the first compartment corresponding to exchange of [18F]FDG in tissue, exchanging with blood, and with the second, irreversible compartment corresponding to the formation of [18F]FDG-6-P in tissue. This model was adapted and used for PET imaging of [18F]FDG in human brain (44) and myocardium (45). In our initial studies using dynamic PET imaging of [18F]FDG in skeletal muscle to investigate insulin resistance in type 2 DM (1, 2, 11), we used the Sokoloff model. Recently, Bertoldo et al. (16) proposed an additional modification of this compartmental model, adapted to skeletal muscle, with the initial reversible compartment corresponding to muscle interstitium, a second reversible compartment corresponding to bidirectional transmembrane transport of glucose, and an irreversible compartment corresponding to glucose phosphorylation. Applied to dynamic PET imaging data using [18F]FDG, an important finding of the Bertoldo et al. model was that insulin had a robust effect in healthy volunteers to stimulate the rate constants for inward glucose transport and glucose phosphorylation, with only minor effect on the inward rate constant for [18F]FDG from plasma to the interstitium.

Returning to the issue of dynamic imaging of [11C]3-OMG, we postulate that the two reversible compartments indicated by SA correspond to exchange of tracer between blood and interstitial space and from interstitial space to tissue mediated by transmembrane glucose transport. The main finding of this compartmental modeling is a robust effect of insulin to enhance the inward transport rate constant, k3, of [11C]3-OMG into muscle, a finding certainly consistent with known effects of insulin to enhance transmembrate glucose transport (13). The rate constant k3 increased 6-fold during insulin infusions, rates that achieved steady-state elevations of insulin. The insulin level attained was in the mid to upper physiological range for healthy volunteers, as would be attained transiently during prandial conditions. In contrast, little effect of insulin was observed on the rate constant that we attribute to the exchange of tracer between blood and interstitial space, k1.

These findings with [11C]3-OMG are concordant with previously reported findings from studies of insulin action on skeletal muscle using dynamic PET imaging of [18F]FDG, analyzed with the Bertoldo et al. model (17, 18). In those studies insulin was observed to exert a robust effect in healthy volunteers to stimulate the k3 rate constant (inward glucose transport into the tissue compartment) and to stimulate k5, the rate constant for glucose phosphorylation, yet it had a minor effect on k1, the rate constant for inward exchange from blood to interstitium (17, 18). In earlier studies, k1 was found to correlate well with tissue perfusion, as measured independently using [15O]H2O, thus solidifying the concept that this rate constant indeed corresponds to an exchange between blood and interstium (17, 18). This relationship between the kinetics of tissue perfusion and k1 for the modeling of [11C]3-OMG has not been explored and will need to be addressed in future studies. There have been prior studies using [11C]3-OMG and dynamic PET imaging to investigate glucose metabolism in the brain (23, 24, 25, 26). Interestingly, the same two extracellular compartment-four rate constant model shown in Fig. 1Go was previously found in CNS studies to be the best model to describe [11C]3-OMG kinetics in brain tissue, although the physiological meaning given to the compartments was different. In brain, the first compartment described [11C]3-OMG in tissue, whereas the nature of the second compartment was unclear and was postulated to represent passage of free [11C]3-OMG into a cerebral subcompartment (23, 24, 25).

In summary, in the current study we examined the use of [11C]3-OMG as a glucose tracer for the study of insulin action in human skeletal muscle. The findings reveal that high quality imaging can be attained and that a clear effect of insulin to modulate the amplitude and configuration of tissue tracer activity is observed. SA indicates the presence of a relatively parsimonious construct to the metabolic modeling of these data, a simplicity that is certainly consistent with the restriction of 3-OMG to the steps of substrate delivery from blood to tissues and bidirectional transmembrane transport. Our initial application indicates a robust effect of insulin to stimulate glucose transport in skeletal muscle of lean, healthy, insulin-sensitive volunteers. We conclude that dynamic PET imaging of [11C]3-OMG offers great promise as a novel and specific method for quantification of glucose transport in human skeletal muscle.


   Acknowledgments
 
We gratefully acknowledge the efforts and cooperation of the research volunteers, and the support from the staffs of the University of Pittsburgh General Clinical Research Center, and PET Center.


   Footnotes
 
This work was supported by the NIDDK, NIH (Grant DK-60555-02), the University Pittsburgh General Clinical Research Center (Grant 5MO1-RR-00056), the Obesity and Nutrition Research Center (NIDDK Grant P30-DK-46204), and NIH Grant EB-01975.

First Published Online December 21, 2004

Abbreviations: AIC, Akaike information criterion; B, basal; C, clamp; CNS, central nervous system; DM, diabetes mellitus; [18F]FDG, 2-[18-fluoro]2-deoxyglucose; [18F]FDG-6-P, [18F]FDG-6-phosphate; IR, insulin resistance; 3-OMG, 3-O-methylglucose; PET, positron emission tomography; ROI, region of interest; SA, spectral analysis.

Received June 9, 2004.

Accepted December 6, 2004.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Kelley DE, Mintun MA, Watkins SC, Simoneau JA, Jadali F, Fredrickson A, Beattie J, Theriault R 1996 The effect of non-insulin-dependent diabetes mellitus and obesity on glucose transport and phosphorylation in skeletal muscle. J Clin Invest 97:2705–2713[Medline]
  2. Kelley DE, Williams KV, Price JC 1999 Insulin regulation of glucose transport and phosphorylation in skeletal muscle assessed by positron emission tomography. Am J Physiol 277:E361–E369
  3. Kelley DE, Williams KV, Price JC, McKolanis TM, Goodpaster B, Thaete FL 2001 Plasma fatty acids, adiposity and variance of skeletal muscle insulin resistance in type 2 diabetes mellitus. J Clin Endocrinol Metab 86:5412–5419[Abstract/Free Full Text]
  4. Nuutila P, Koivisto VA, Knuuti J, Ruotsalainen U, Teras M, Haaparanta M1992 Glucose-free fatty acid cycle operates in human heart and skeletal muscle in vivo. J Clin Invest 89:1767–1774
  5. Nuutila P, Knuuti MJ, Maki M, Laine H, Ruotsalainen U, Teras M, Haaparanta M, Solin O, YkiJarvinen H 1995 Gender and insulin sensitivity in the heart and in skeletal muscles: studies using positron emission tomography. Diabetes 44:31–36[Abstract]
  6. Nuutila P, Maki M, Laine H, Knuuti MJ, Ruotsalainen U, Luotolahti M, Haaparanta M, Solin O, Jula A, Koivisto VA, Voipio-Pulkki LM, Yki-Järvinen H 1995 Insulin action on heart and skeletal muscle glucose uptake in essential hypertension. J Clin Invest 96:1003–1009
  7. Nuutila P, Raitakari M, Laine H, Kirvela O, Takala T, Utriainen T, Makimatilla S, Pitkanen O, Ruotsalainen U, Iida H, Knuuti J, Yki-Järvinen H 1996 Role of blood flow in regulating insulin-stimulated glucose uptake in humans. J Clin Invest 97:1741–1747[Medline]
  8. Peltoniemi P, Yki-Jarvinen H, Laine H, Oikonen V, Ronnemaa T, Kalliokoske K, Raitakari O, Knuuti MJ, Nuutilla P 2001 Evidence for spatial heterogeneity in insulin- and exercise-induced increases in glucose uptake: studies in normal subjects and patients with type 1 diabetes. J Clin Endocrinal Metab 86:5525–5533[Abstract/Free Full Text]
  9. Reinhardt M, Beu M, Vosberg H, Herzog H, Hubinger A, Reinauer H, Muller-Gartner H-W 1999 Quantitation of glucose transport and phosphorylation in human skeletal muscle using FDG PET. J Nucl Med 40:977–985[Abstract/Free Full Text]
  10. Routsalainen U, Raitakari M, Nuutila P, Oikonen V, Sipila H, Teras M, Knuuti J, Bloomfield PM, Iida H 1997 Quantitative blood flow measurement of skeletal muscle using oxygen-15-water and PET. J Nucl Med 38:314–319[Abstract/Free Full Text]
  11. Williams KV, Price JC, Kelley DE 2001 Interactions of impaired glucose transport and phosphorylation in skeletal muscle insulin resistance. A dose-response assessment using positron emission tomography. Diabetes 50:2069–2079[Abstract/Free Full Text]
  12. Fowler JS, Wolf AP 1986 2-Deoxy-2-[18F]fluoro-D-glucose for metabolic studies: current status. Appl Radial Isot 37:663–668
  13. Sheperd PR, Kahn BB 1999 Glucose transporters and insulin action: implications for insulin resistance and diabetes mellitus. N Engl J Med 341:248–257[Free Full Text]
  14. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M 1977 The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897–916[Medline]
  15. Huang S, Phelps ME 1986. Principles of tracer kinetic modeling in positron emission tomography and autoradiography. In: Phelps MJ, Schelbert HR, eds. Positron emission tomography and autoradiography: principles and applications for the brain and heart. New York: Raven Press; 287–346
  16. Bertoldo A, Peltoniemi P, Oikonen V, Knuuti J, Nuutila P, Cobelli C 2001 Kinetic modeling of [18F] FDG in skeletal muscle by PET: a four-compartment five rate constant model. Am J Physiol 281:E524–E536
  17. Williams K, Bertoldo A, Mattioni B, Price J, Cobelli C, Kelley D 2003 Glucose transport and phosphorylation in skeletal muscle in obesity: insight from a muscle-specific positron emission tomography model. J Clin Endocrinol Metab 88:1271–1279[Abstract/Free Full Text]
  18. Williams K, Bertoldo A, Kinahan P, Cobelli C, Kelley D 2003 Weight loss induced plasticity of glucose transport and phosphorylation in the insulin resistance of obesity and type 2 diabetes. Diabetes 52:1619–1629[Abstract/Free Full Text]
  19. Carruthers A 1990 Facilitated diffusion of glucose. Physiol Rev 70:1135–1176[Free Full Text]
  20. Narahara H, Ozand P 1963 Studies of tissue permeability. IX. The effect of insulin on the penetration of 3-methylglucose-H3 in frog muscle. J Biol Chem 238:40–49[Free Full Text]
  21. Buschiazzo P, Terrell E, Regen D 1970 Sugar transport across the blood-brain barrier. Am J Physiol 219:1505–1513[Free Full Text]
  22. O’Doherty R, Halseth A, Granner D, Bracy D, Wasserman D 1998 Analysis of insulin-stimulated skeletal muscle glucose uptake in conscious rat using isotopic glucose analogs. Am J Physiol 274:E287–E296
  23. Brooks D, Beaney R, Lammertsma A, Herold S, Turton D, Luthra S, Frackowiak R, Thomas D, Marshall J, Jones T 1986 Glucose transport across the blood-brain barrier in normal human subjects and patients with cerebral tumours studied using [11C]3-O-methyl-D-glucose and positron emission tomography. J Cereb Blood Flow Metab 6:230–239[Medline]
  24. Brooks D, Gibbs J, Sharp P, Herold S, Turton D, Luthra S, Kohner E, Bloom S, Jones T 1986 Regional cerebral glucose transport in insulin-dependent diabetic patients studied using [11C]3-O-methyl-D-glucose and positron emission tomography. J Cereb Blood Flow Metab 6:240–244[Medline]
  25. Feinendegen L, Herzog H, Thompson K 2001 Cerebral glucose transport implies individualized glial cell function. J Cereb Blood Flow Metab 21:1160–1170[Medline]
  26. Vyska K, Magloire J, Freundlieb C, Hock A, Becker V, Schmid A, Feinendegen L, Kloster G, Stocklin G, Schuier F 1985 In vivo determination of the kinetic parameters of glucose transport in the human brain using 11C-methyl-D-glucose (CMG) and dynamic positron emission tomography (dPET). Eur J Nucl Med 11:97–106[Medline]
  27. Bender D, Munk OL, Feng H-Q, Keiding S 2001 Metabolites of 18F-FDG and 3-O-11C methylglucose in pig liver. J Nucl Med 42:1673–1678[Abstract/Free Full Text]
  28. Munk O, Bass L, Roelsgaard K, Bender D, Hansen S, Keiding S 2001 Liver kinetics of glucose analogs measured in pigs by PET: importance of dual-input blood sampling. J Nucl Med 42:795–801[Abstract/Free Full Text]
  29. DeFronzo RA, Tobin JD, Andres R 1979 Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 237:214–223
  30. Kloster G, Muller-Platz C, Laufer P 1981 3-[11C]-Methyl-D-glucose, a potential agent for regional cerebral glucose utilization studies: synthesis, chromatography and tissue distribution in mice. J Labelled Compound Radiopharmaceuticals 18:855–863
  31. Dannals R, Ravert H, Wilson A, Wagner Jr H 1986 An improved synthesis of (3-N-[11C]methyl)spiperone. Int J Radiat Applied Isotopes 37:433–434
  32. Weinhard K 1998 Applications of 3D PET. In: Bendriem B, Townsend D, eds. The theory and practice of 3D PET. Boston: Kluwer Academic; 133–167
  33. Watson C, Newport D, Casey M, DeKemp R, Beanlands R, Schmand M 1997 Evaluation of simulation-based scatter correction for 3D PET cardiac imaging. IEEE Trans Nucl Sci 44:90–97[CrossRef]
  34. Cunningham VJ, Jones T 1993 Spectral analysis of dynamic PET studies. J Cerebral Blood Flow Metab 13:15–23[Medline]
  35. Cobelli C, Saccomani M, Ferrannini E, DeFronzo R, Gelfand R, Bonadonna RC 1989 A compartment model to quantitate in vivo glucose transport in the human forearm. Am J Physiol 257:E943–E958
  36. Schmidt K, Lucignani G, Moresco R, Rizzo G, Gilardi M, Messa C, Colombo F, Fazio F, Sokoloff L 1992 Errors introduced by tissue heterogeneity in estimation of local cerebral glucose utilization with current kinetic models of the [18F]fluorodeoxyglucose method. J Cereb Blood Flow Metab 12:823–834[Medline]
  37. Cobelli C, Foster D, Toffolo G 2001 Tracer kinetic in biomedical research: from data to model. Boston: Kluwer Academic/Plenum
  38. Institute S 1998 SAAM II user guide. Seattle: SAAM Institute
  39. Bertoldo A, Vicini P, Sambucceti G, Lammertsma AA, Parodi O, Cobelli C 1998 Evaluation of compartmental and spectral analysis models of [18F] FDG kinetics for heart and brain studies with PET. IEEE Transact Biomed Engin 45:1429–1448[CrossRef]
  40. Akaike H 1974 A new look at the statistical model identification. IEEE Trans Automat Contr AC 19:716–723[CrossRef]
  41. Andres R, Cader G, Zierler K 1956 The quantitatively minor role of carbohydrate in oxidative metabolism by skeletal muscle in intact man in the basal state. Measurement of oxygen and glucose uptake and carbon dioxide and lactate production in the forearm. J Clin Invest 35:671–682
  42. Bonadonna RC, Saccomani MP, Cobelli C 1993 In vivo glucose transport in human skeletal muscle: tools, problems and perspectives. Baillieres Clin Endocrinol Metab 7:929–960[Medline]
  43. Furler SM, Jenkins AB, Storlien LH, Kraegen EW 1991 In vivo location of the rate-limiting step of hexose uptake in muscle and brain tissue of rats. Am J Physiol 261:E337–E347
  44. Phelps ME, Huang SC, Hoffman EJ, Selin CE, Sokoloff L, Kuhl DE 1979 Tomographic measurement of local cerebral glucose metabolic rate in humans with (F-18)2-fluoro-2-deoxy-D-glucose: validation of method. Ann Neurol 6:371–388[CrossRef][Medline]
  45. Ratib O, Phelps M, Huang S, Henze E, Selin C, Schelbert H 1982 Positron tomography with deoxyglucose for estimating local myocardial glucose metabolism. J Nucl Med 23:577–586[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 JNMHome page
C. Spring-Robinson, V. Chandramouli, W. C. Schumann, P. F. Faulhaber, Y. Wang, C. Wu, F. Ismail-Beigi, and R. F. Muzic Jr.
Uptake of 18F-Labeled 6-Fluoro-6-Deoxy-D-Glucose by Skeletal Muscle Is Responsive to Insulin Stimulation
J. Nucl. Med., June 1, 2009; 50(6): 912 - 919.
[Abstract] [Full Text] [PDF]

Home page
 DiabetesHome page
A. Bertoldo, R. R. Pencek, K. Azuma, J. C. Price, C. Kelley, C. Cobelli, and D. E. Kelley
Interactions Between Delivery, Transport, and Phosphorylation of Glucose in Governing Uptake Into Human Skeletal Muscle
Diabetes, November 1, 2006; 55(11): 3028 - 3037.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Endocrinol. Metab.Home page
R. R. Pencek, A. Bertoldo, J. Price, C. Kelley, C. Cobelli, and D. E. Kelley
Dose-responsive insulin regulation of glucose transport in human skeletal muscle
Am J Physiol Endocrinol Metab, June 1, 2006; 290(6): E1124 - E1130.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bertoldo, A.
Right arrow Articles by Kelley, D. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bertoldo, A.
Right arrow Articles by Kelley, D. E.
Related Collections
Right arrow Diabetes and Insulin
Right arrow Metabolism

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals